STANDARD METHODS

OF

CHEMICAL ANALYSIS
A Manual of Analytical Methods and General Reference for the Analytical
Chemist and for the Advanced Student

BY

WILFRED W. SCOTT, Sc.D.

Late Professor of .Chemistry, University of Sotrthern California. Author of "Qualitative
Chemical Analysis," "Technical Methods of Metallurgical Analysis,"
':r'I1U!.!ganic Quantitative Chemical Analysis." .

FIFTH EDITION
EDITED BY

N. HOWELL FURMAN, Ph.D.

Professor of Chemistry, Princeton University

IN COLLABORATION WITH EMINENT SPECIALISTS

IN TWO VOLUMES, ILLUSTRATED

VOLUME TWO-SPECIAL SUBJECTS

LONDON

1'HE

TECHNICAL

PRESS

LTD.

LATE OF AVE MARIA LANE, LUDGATE HILL

GI.OUCESTER ROAD, KINGSTON HILL, SURREY

CQPYRIGHT,

By D.
1939 By D.

fj9':z,h

1917, 1922, 1925,

VAN NOSTRAND COMP-'\NY

VAN NOSTRAND COMP1\NY-, INC.

'All Rights' Reserved.

This book, or any parts thereof, may n,ot
be reproduced in any form without
written permission from th~ publish~r.

Fifth Edition, Published March 1939

Reprinted, May 1940, March 1942

FebnlUTY 1943, March 1944, March 1[11451

plttN'l'ED IN THE U oft.

s7 A.

lANCASTER PRESS.. INC., LANCASTER, PA.

CONTRIBUTORS
VOLUME TWO

N. Howell Furman, Ph.D. Professor' of Chemistry, Princeton University,

Princeton, N. J. Editor in Chief. Contribution-Chapter on Potentiometric Methods; revision of material contributed by the late Professor
W. W. Scott.
Herbert AD~i:pr, :B.~ in Chemistry. President of tile Rubberoid Company,
N. Y. Contribution-Chapter on bituminous substances, including
Asphalts, Tars and Pitches.
J. Bryte Barnitt, Ch.E. Consulting Chemical Engineer, Pittsburg, Pa. Contribution-Chapter 011 Sampling.
\
Robert C. Beckstrom, A.B., A.M. Consulting Petroleum Engineer, Tulsa,
Okla. Contribution-Petroleum and' Its Products, Interpretation of
Tests.
Wallace R. Brode, Ph.D. Associate Professor of Chemistry, Ohio State
University, Columbus, Ohio. Contribution-Chapter on Spectrographic
Analysis.
.
Walter J. Brown, B.S. Chief Analyst, National Lead Company Research
Laboratories, Brooklyn, N. Y. Contribution-Chapter on Non-Ferrous
Alloys.
.:'
Emile M. Chamot, B.S., Ph.D. Professor Emeritus of Chemical Microscopy
and Sanitary Chemistry, Cornell University, Ithaca, N. Y. Contribution
-Co-authOl; of the chapter on Chemical Microscopy.
Dudley K. French, B.S. Head of Dudley K. French and Associates, Chicago,
.Ill. Contribution-Water Analysis.
Henry A. Gardner, D.Sc. Director, Scientific Section, National Institute of
Paint and Varnish Research. Contribution-Co-author chapter on
Paints.
* Augustus H. Gill, S.B., A.M.; Ph.D., Sc.D. formerly Professor of Technical
Chemical Analysis, Massachusetts Institute of Technology, Cambridge,
Mass. Contributions-Chapters on Oils, Fats and Waxes, and on Gas
Analysis.
W. N. Greer, Ph.D. Leeds and Northrup Qpmpany, Philadelplia, Pa.
Contribution-Chapter on Electrometric Hydrogen Ion Measurement.
Frank E. Hale, Ph.D. Director of Lal:)Oratories, Department of Water Supply,
Gas and Electricity, New York, N.;,y. Contribution-Microscopical
Examination of Water.
'
L. T. Hallett, Ph.D: Chief, Department of Quantitative Microanalysis,
Eastman' Kodak Co., Rochester, N. Y. Contribution-Chapter' on
Quantitative Microchemical Analysis.
v

CONTRIBUTORS TD VOLUME TWO

VI

W. C. Hanna. Chief Chemist California Portlanc;l Ce~ent Co., Colton, Calif.

Chairman Working Committee on JYIethods of Chemical Analysis of
Cement, A.S:T.M. Contribution-Chapter on Cement.'
..,_
1. M. Kolthoff, Ph.D. Professor and Head of Division of Analytical Chemistry,
University of Minnesota, Minneapolis, Minn. Advisory Editor. Contributions-Chapter on Conductometric Methods; abstracted many of his
own papers for Vol. 1.
.
C. P. Long, Research Chemist, representing the Chemical Staf( of .the Procter
and Gamble Co., Ivorydale, O. Contribution-Chapter on Soap: Methods'
for Determination of Moisture ..
Arthur R. Maas, Ph.C., Pharm.M. Professor of Pharmacology and Toxicology, University of Southern California. Contribution-Chapter on
Poisons.
F. R. McCrumb, B.S., M.S. Chemical Director, W. A. Taylor Co., Baltimore,
Md. Contribution-Co-author chapter 'on Determination of pH by the
Colorimetric Method.
Clyde W. Mason, Ph.D. Professor of Chemical Micros'copy, Department of
Chemistry, Cornell University, Ithaca, N. Y. Contribution-Co-author
of chapter on Chemical Microscopy.
L. E. Salas, M.A., B.Sc. Research and Analytical Che.mist, London, England.
Contribution-Chapter on Rubber, Gutta Percha, Balata.
John A. Schaeffer, A.M., Ph.D. President of Franklin and Marshall College,
Lancaster, Pa. Former Chief Chemist, Eagle Picher Lead Co.. Contribu- ,
tion-Co-author, chapter on Paints.
* Wilfred W. Scott, A.B., A.M., Sc.D. Late Professor of Chemistry, University of Southern California, Late Chief Editor of Standard Methods of
Chemical Analysis. Contributor of a number of cllapters in Volumes
.
I and II.
Atherton Seidell, Ph.D. Chemist, National Institute of Health, Washington,
D.C. Contribution-Chapter on Solubility. I>
C. G. Storm, Ph.D. Chief, Explosives Section, Ammunition Division, Office
of Chief of Ordinance, U. S. A. Contribution-Chapter on Explosives.
Raymond B. Stringfield, A.B., B.S. Consulting Chemical Engineer, Secretary
Dental Plastics Co., Los Angeles, Calif. Formerly Chief Chemist, Goodyear Tire and Rubber Co. of California. Contribution-Chapter on
Rubber Compo'llnding Ingredients.
G. G. Sward, B.S., M.S. Chemist, National Paint, Varnish and Lacquer Assoc.
Inc. Contribution-Co-author chapter on Paints.
W. A. Taylor, Ph.D. President, W. A. Taylor Co., Baltimore, Md. Contribution-Co-author of chapter on Determination of pH by the Colorimetric Method.
Joseph Winlock, A.B. Chief Metallurgist, Edw. G. Budd Manufacturing Co.,
Philadelphia, Pa. Contribution-Chapter on Metallography.
Thomas A. Wright~ Chemical Director, Lucius Pitkin Co., New York. Advi.
sory Editor. Contribution-critical review of various chapters. Vol. II.
Revision of Fire Assay Material in Volume I.

* Deceased.

CONTENTS
SPECIAL SUBJECTS

PART 1. SAMPLING, PRELIMINARY TREATMENT

AND APPLIED ANALYSIS
STANDARD METHODS OF SAMPLING SOLIDS, LIQUIDS AND GASES
METHODS FOR SAMPLING'SOLIDs.-General outline. Introduction. Sampling unit;
collection of the gross sample; carload lots, wheelbarrow lots, bags, barrels, etc.; crane
bucket lots, carts, shoots, etc.; shiploads, storage piles, 1301-1304. Sampling of
powdered material; sampling wet material; sampling deliquescent materials; kilns,
roaste.fs, dryers, furnacE:ls, non-ferrous metals and a.lloys, etc., 1305--1306. Procedure
for reducing the gross sample. Reduction of the gross sample, long pile and alternate
shovel method. Cone and quartering, 1307-1310. Apparatus for reduction and
preparation of the sample-crushers, grinders, 1310-1313. Mixing and dividing sample;
the use of standard sieves, 1314-1316. Containers for shipment to laboratory; drying
ovens for moisture determination; references, 1316-1318.
SAMPLING LIQUIDs.-General outline for -sampling liquids, 1318. Sampling liquids
in quiescent state, 1319. Sampling liquids in motion. Sampling special liquids,.
1320-1323. Liquids containing insoluble matter in suspension; liquids composed of immiscible layers; liquids containing crystals due to supersaturation; volatile liquids,
1323-1324. References, 1324.
SAMPLING GAsEs.-General outline for sampling gases, 1324. General considerations
and apparatus-sampling tubes for drawing samples from gases flowing through pipes,
collecting a representative sample, the design of apparatus, 1325--1326. Taking a
"grab" sample; taking a continuous sample, apparatus and methods of sampling,
1326-1328. References, 1329. Standard specifications for sieves for testing purposes,
A. S. T. M., 1329-:1333.
DECOMPOSITION OF THE

S~MPLE

Decomposition of the material, 1334. Reagents used-I, Water; II, Acids; III,
Fusions, 1335-1337. Properties and special uses of reagents, 1337-1338.
DETERMINATION OF MOISTURE, WATER OF CRYSTALLIZATION
.
AND WATER OF CONSTITUTION
Method of loss in weight, 1338"':1339. Adsorption method, 1339-1340. Distillation
method: water in petroleum, 1340-1341; Brown-Duvel method for grains, 1841-1342;
toluene-moisture method, 1342-1343; xylene method, 1343; Kingman mllthod, 13431346. Determination of moisture in gases, 1347.
ALLOYS-I.

ANALYSIS OF NON-FERROUS ALLOYS

Classification of alloys, 1348. Preliminary examination; qualitative scheme, 13481349. Methods of procedure, 1349-1350. Difficulty of complete separation of elements,
vii

viii

CONTENTS

1350. Limit of accuracy, 1350. Sampling, 13.50-1351. Analysis of brasses and

bronzes-analysis of manganese bronze, 135l. Determination of copper, 1352; l~ad, 1354;
tin, 1356; iron, 1357; manganese, 1358; aluminum, 1359. Analysis of bronze bearing
metals, 1362. Determination of copper, 1362; tin, 1362; lead, 1364-1365; iron, 1365;
nickel, 1365-1366; zinc, 1366; phosphorus, 1366-1369; antimony, 1369-1370; sulfur,
1370-1371; arsenic, 1371; cadmium, 1371-1372. Analysis of copper, nickel, zinc alloys,
1373. Determination of tin, 1373; copper, 1373; lead, ~373; iron, 1373-1374; mangaIlese,
1374; nickel, 1374-1375; zinc, 1375. Analysis of bismuth bronze, alloy of copper, tin, bismuth, nickel and zinc, i375. Determination of tin, 1375; lead, 1375; copper, 1376; bismuth, 1376; nickel, 1377; zinc, 1377; iron, 1377; aluminum, 1378; manganese, 1378.
Copper-nickel alloys, 1378. Monel metal-determination of copper, 1378; nickel, 1378;
iron, 1379; sulfur, 1379; silicon, 1379; manganese, 1380. Tin and lead base alloys,
1381. Determination of lead, 1381; copper, 1382; antimony, 1382; tin,13S3; arsenic, 13S4:
Rapid methods for control work, 1384; arsenic, antimony, tin, 1385. The determination of lead, bismuth, iron, aluminum and nickel in tin base bearing metal, i387-1390.
Determination of antimony, zinc, copper in solder, 1390-1391. Rapid method for
control work, 1391-1393. Determination of tin in battery plate met~l, 1393. Zinc
base die casting alloys--determination of copper and aluminum, 1394; magnesium, 1395;
cadmium and iron, 1396; lead and tin, 1396. Fusible metals, 1399. AnalysiS of
'Vood's alloy--determination of lead, 1399; tin, 1399; bismuth, 1;~99; cadmium and
copper, 1400; antimony, 1401; iron, 1401; zinc, 1401; arsenic, 1402. Silver alloys-analysis of silver solders-dctel'ruination of silver, 1402; copper, 1402; cadmium, 1403;
zinc, 1403; tin, 1404; lead, 1404; iron, 1404; nickel, 1405. Heat resisting alloys, 1405.
Analysis of nichrome. Determination of nickel, 1405; chromium, 1406; iron, 1407;
manganese, 1407; carbon, 1408; silicon, 1408; sulfur, 1409. Corrosion resisting alloys,
1409. Analysis of stellite. Determination of cobalt, 1409; iron, 1411; chromium, 1411;
molybdenum, 1411; 'manganese, 1412; silicon; 1412.
A. S. T. M. method for the ' chemical analysis of pig lead, 1413; determinatioll of
silver; bismuth; arsenic; remaining metals; determination of antimony and tin; determination of copper and cadmium; iron; zinc; nickel and cobalt, 1413-1420.
ALLOYS-II.

THE CHEMICAL ANALYSIS OF IRON AND STEEL

A. S ..T. M. Methods for steel, cast iron, open hearth iron and wrought iron-sampling, 1422-1426; total carbon by Combustion, 1426-1430; graphite in cast iron, 14-30;
manganese by the bismuthate method, 1431-4; manganese, by the persulfate method,
1434-5; phosphorus by the molybdate-magnesia method, 1435-8; phosphorus by the
alkalimetric method, 1438-1440; sulfur by the gravimetric method, 144(}-1442; sulfur
by the evolution method, 1442-1444; silicon-sulfuric acid method, 1444-1446; perchloric acid method, 1446; copper--electrolytic method, 1445-1448; thiosulfate-iodide
method, 1448-1449; nickel-by the blyoxime method, 1450-1452; by the cyanide
titration method, 1452-1454; chromium by the persulfate oxidation method, 1454-1457;
vanadium--electrolytic separation method, 1458-9; by the ferrous sulfate-permanganate
method, 1459-1461; molybdenum-by the a-benzoinoxi-me method, 1461; by the sulfideoxide method, 1462; by the colorimetric method, 1464; tungsten by the cinchonine
method, 1465; cobalt, 1466; titaniUIh, 1468; zirconium, 14'70.
,
Other methods. Manganese---:afte~ oxidation by periodate, 1470; modified persu~
fate-arsenite method, 1471; chromium and vanadium-indicator methods, 1473-1477;
"ftel' oxidation with pcrchloric 8cid, 1477-1479; in the same sample, 1479. Sdenil1nl-

CONTENTS

ix

standard method of the Carpenter Steel C~., 1480. Molybdenum Maag-McCollam

colorimetric procedure, 1481. Titanium, zirconium or aluminum--electrolytic separation, ,1482-1484.

ALLOYS-III. FERRO-ALLOYS
A. S. T. M. methods. Ferrosilicon-determination of silicon, 1485. Ferromanganesemanganese by the modified bismuthate method, 1486; carbon, 1488; phosphorus, 1488;
sulfur, 1490; silicon, 1490. Ferrochromium-determination of chromium after sodium
peroxide fusion, 1490; carbon, 1493; silicon, 1493. Ferrovanadium-determination of
vanadium by persulfate method, 1494; carbon, 1495; phosphorus-modified molybdatemagnesia method, 1495; sulfur-oxidation method, 1497; silicon, 1497; aluminUIIiether-hydrochloric acid method, 1498. Tungsten metal-determination of tungsten,
1499; carbon, 1500; phosphorus, 1500; suliur, 1502; silicon, 1502. Ferrotungstendetermination of tungsten, 1503; silicon, 1504; sulfur, 1504; copper, antiI~ony and tin,
1504--1506. Ferromolybdenum-determination of carbon, 1506; molybdenum, 1507;
~
sulfur, 1508; phosphorus, 150?; copper, 1509.

BfTUMINOUS SUBSTANCES INCLUDING ASPHALTS, TARS AND PITCHES

PART I. Examination of bituminous substances, Table I, 1510; distinguishing
char,cteristics of bituminous substances, Table II, 1511-1512; physical characteristics,
1513. Fracture; streak; specific gravity-hydrometer method; pyknometer method,
1515; analytical bala~ce method, 1517. Viscosity-Engler method, 1518; Saybolt
method, 1518; Float test, 1518-1520. Penetration (hardness), 1520. Ductility, 1521.
Thermal tests, 1523. Fusing or softening-point-ring and ball method, 1523; cube
method, 1525. VoIatile matter, 1527-1530. Evaporation test; 1530--1533. Flashpoint, 1533. Fixed carbon, 1533. Distillation test-for road oils, 1533; for cut-back
asphaltic products, 1536; for coal-tar products, 1538. Solubility tests-in carbon disulfide: where constituents are not to be examined further, 1539; constituents to be
examined further-water, 1541; sulfur, 1541; carbonates, 1541; non-mineral insoluble;
iron, Al, Ca, Mg, Si0 2, etc., 1542; carbon disulfide extract, 1542; evaporation at
atmospheric pressure, 1543; evaporation under vacuum, 1543; evaporation in stl'eam of
CO 2, 1545. Carbenes,1546. Soluble in petroleum naphtha, 1547. "Free carbon"(A) por'ous thimble method, 1549; (B) asbestos mat method, 1549; (C) paper thimble or
filter-paper method, 1550. Chemical tests. Water, 1551; substances distilling at
low temperatures, 1551; at high temperatures, 1552; substances not fusing at 300 0 F.,
1554. Oxygen (in non-mineral matter), 1554. Solid paraffins, 1555-1557.-(A) Sulfonation ~esidue insoluble in con'c. sulfuric acid, 1557; (B) sulfonation residue insoluble
in water, 1558. S~onifiable constituents, 1559. Diazo reaction, 1560. Anthraquinone reaction, 1562.
PART II. Examination of bituminous substances combined with discrete aggregates.
(A) Physical tests of finished product, 1563. Resistance to displacement, 1563. Tensile
strength, 1564. Compressive strength, 1565. (B) Separation of finished product into
its component parts, 1567. Sepamtion of bituminous matter and discrete aggregatehot extraction method, 1567; cold extraction method, 1569; centrifugal extraction
method, 1570--1572.
PART III. Examination of bituminized fabrics, 1573. Classes of products, 1573;'
table showing construction, 1573; (A) physical tests of the finished product, 1574.
Weight per unit area, 1574; Thickness, 1577; resistance to heat, 1.577; (B) separation of

CONTENTS

the Jini.ehecl product into ite component parts, 1578. Moisture, 1578. Anaiysis of
mturated fabrics, 1578. Analysis of saturated and coated fabrics, 1580. (a) Mineral
matter, 1581; (b) bituminous matter, 1582; Cc) fibrous matter, 1584; compilation of
results, 1585.
PART IV. Examination of bituminous-solvent compositions, 1586. (A) Physical
tests of finished product, 1586. (B) Separation of the finished product into its com~
ponent parts, 1586. Estimation -and recovery of the solvent-steam distillation method,
1586--1589. (C) Estimation, recovery and examination of pigm,ent and filler, 1589.
(D) Estimation and recovery of the base, 1589-1500.
PART V. Examination of bituminous dispersions, 1591. (A) Physical tests of
finished product, 1591. Settlement test, 1591; demulsihility test, 1591 j effects of freezing, 1592. (B) Separation of the dispersion into its component parts, 1592. Distillation residue, 1593; water and volatile oils, 1594-1596.
CEMENT

Specifications, 1597-1599. A. S. T. M. methods for chemical analysis-Sa.mpling,

1599; determination of loss on ignition, 1600; insoluble residue, 1601; sulfuric anhydride,
1601; silica, 1601; ferric oxide and alumina, 1602; calcium oxid~, 1603; magnesia, 1604;
blank determination, 1604. Tentative methods: rapid determination of magnesi~
acid-alkali method, 1605-1607; titration of magnesium o:xyquinolate, 1607-1609.
Phosphorus, 1609; manganese, 1611; J. Lawrence Smith's Ipethod for alkalies, 1612~1614.
Rapid methods for chemical analysis of cement, 1614; silica, 1614; ferric oxide and
alUmina, 1614; calcium oxide, 1614; magnesia., 1615; other components, 1615. Rapid
method for the determination of carbonates in raw materials, 1616. Analysis of limestone, lime, raw mix, etc., 1617. Free lime in ccme.!_lt!_ 1618. A table of analyses, 1619.
COAL AND COKE

A. S. T. M. methods. Standard method of sampling coal, 1620. For aU determinations except moisture-time of sampling, 1620; size of increments, 1620; collection
of gross sample, 1620; quantity represented, 1620; crUshing, 1620; mixing and reduction,
1623; for determination of total moisture, 1624. Standard method of sampling coke,
1624. For all determinations except moisture, 1625-1627; for determination of total
moisture, 1628. Standard methods of laboratory sampling and analysis of coal and cOk;e,
1629. Preparation of laboratory samples, 1629. Standard specifications for sjeves
for testing purposes, 1630; method of sampling, 1632-1636; methods of analysis, 1636.
Determination of moisture, 1636; determination of ash, 1637; deteimination of volatile
matter, 1638; determination of fixed carbon, 1640; determination of sulfur, 1640; (A)
Eschka method, 1640; (B) bomb washing method, 1642; (C) scidium peroxide fusibn
method, 1643; determination of phosphorus in ash-method I-for a11 cases,. 1644;
method 2-when titanium is low, 1644. -Determination of fusibility of ash, 1645-1649;
ultimate analysis, 1649. Carbon and hydrogen, 1649-1652; nitrogen, 1652. Calori"metric determination, 1654. Apparatus, 1654; procedure, 1655; calculation of results,

1657-1659.
EXPLOSIVES

Introduction, Ift60. Black powder--sampling, moisture, nitrates,' suilur, char(loal,

ash, calculation of results, 1660-1662. Nitroglycerin dynamites-..:..sampIing, quaIiu.tive
examination, moisture, extraction with ether-reflux con~enser methodj suction method,

CONTENTS

xi

evaporation of ether extract, nitroglycerin, sulfur, resins, oils, etc., extraction with water
and determination of nitrates, extraction with acid, determination of starch, insoluble
residue and ash, 1662--1667. Ammonia dynamite, 1667. Gelatin dynamite, 1668;
sampling, analysis, nitrocellulose, 1668. Low freezing dynamite-moisture, nitrotoluenes, modified Kjehldahl method for nitrogen, separation of nitrocompounds from
nitroglycerin, nitrosugars,' nitrochlorhydrins, nitropolyglycerin, nitroglycol, 1669-1673.
"Permissible" explosives-ingredients, qualitative analysis, tests, mechanical separation
(if solid ingredients, moisture, extraction with ether, with water, nitrates, chlorates,
perchlorates, gum arabic, sugar, extraction with acid, extraction with acetone, insoluble
residue and ash, 1673-1679. Nitrostarch explosives-general nature, moisture, extraction with petroleum ether-oils, sulfur, etc., extraction with water-nitrates, gums, etc.,
insoluble residue-nitrostarch, charcohl, cereal products, etc., nitrostarch, charcoal,
1679-1681. Trinitrotoluene (TNT)-solidification point, ash, mqisture, insoluble,
acidity, nitrogen, 1681-1682.' Picric acid----solidification point, moisture, sulfuric acid,
ash, insoluble in water, soluble lead, nitric acid, 1682--1683. Ammonium picratemoi~re, insoluble material, ash, free ammonia or free picric acid, 1683-1684. Tetrylmelting point, moisture, insoluble in benzene, acidity, 1684. Mercury fulminatepreparation of fiample, mercury fulminate content, acidity, insoluble matter, free
mercury, chlorides, 1685. Lead Azide, acidity-lead content, 1685-1686. Blasting
caps'and electric detonators-preparation of sample, moisture, analysis of composition
containing mercury fulminate and. potassium chlorate, analysis of compositions containing nitrocompounds, 1686--1688. Primers-variations in composition, types of
primer composition, preparation of the sample, qualitative examination, quantitative
analysis, 1688-1690. Nitrocellulose-general, preparation of sample, nitrogen, solubility in ether-alcohol: (a) guncottQn, (b) pyrocellulose, solubility n. acetone, ash, stability .test: heat test with potassium iodide-starch paper, stability test at" 134.5 C.,
preparation of methyl violet test paper, viscosity, 1690-1694. Smokeless powder.
Nitrocellulose powders-moisture and volatiles, moisture, diphenylamine, ash, solubility
in ether-alcohol, in acetone, stability test at 134.5 C., "surveillance test" at 65.5 C.,
nitrogen, 1695-1698. Nitroglycerin smokeless powders, 1698-1699. Table of typical
compositions of explosives, 1700 .
......
FIXED OILS, FATS AND WAXES. PETROLEUM.
OILS, FATS AND WAXES
Introductory. Examination of an unknown oil, 1701. Petroleum products-(a)
burning oils;-flash test; fire test; significance of the tests; specific gravity, significance;
dis,tillation test (A. S. T. M.); apparatus; procedure; accuracy; correction for barometric
pressure. Significance of the test, 1702-1712. Determination of sulfur; A. S. T. M .
... lamp method-1tpparatus; procedure; detection of acidity; sulfuric acid test; mineral
salts; determination
water; color, 1712-1717. (b) Lubricating oils-viscosity;
Engler apparatus; viscosity by means of the Saybolt viscosimeter, A. S. T. M.; MacMichael's viscosimeter; absolute viscosity; specific gravity; evaporation test; cold test;
cloud and pour points, A. S. T. M. method-1tpparatus; procedure; flash point-open
cup, A. S. T. M.-1tpparatus, procedure, 1717-1736. Detection of soap; caoutchouc;
test for fatty oils; gumming test; carbon residue test-Conradson carbon residuescope; appara~us; special proced~re; tolerances. Gasoline test; microscopical test; heat
test; Herschel'S demulsibility test; friction tests; color, 1736--1741. Liquid fuels.
(a) Gasoline-doctor test (sodium plumbite); corrosion test, A. S. T. M. method;

or

xu

CONTENTS

,distillation of gasoline; significance of test; octane nUIlfber, 1742-1743. ,Fuel oils.

Flash point, Pensky-Martens closed tester, A. S. T. M.; flash point by tag closed tester,
A. S. T. M.; 1744--1756. Sulfur, bomb method, A. S. T. M.,apparatus; procedure; water,
A. S., T. M. method-apparatus; sample; procedure; accuracy; sediment-apparatus;
extraction procedure. Water and sediment by centrifuge, A. S. T. M. method. apparatus; procedure; accuracy; calorific power; 1756-1764. Animal and vegetable
oils-specific gravity; refractive inde~; Valenta test; Eliadin test; Maumene test;
references, 1764-1767. Iodine number or value-Hanus method; thiosulfate solution;
Hubl's method-manipulation; oxidized oils, iodine number of; bromine ,number or
value; saponification value-solutions required; apparatus; blank determination;
procedure; percentage of fatty oil, 1767-1773; detection of unsaponifiable oils-(I) from
saponification number; (2) by grawmetric methods; identification of unsaponifiable
~atter; hydrocawon oils; tar oils; rosin oUs; paraffin; ceresene; higher alcohols; cholesterol; lactones. Test for animal or vegetalbe oils; test for antifluorescents; acetyl value,
1773-1777. Special tests for certain oils-Bechi's test for cotton-seed oU; Halphen's
test for cotton-seed oil; heX!1bromide test for linseed oil; Bellier's qualitative test for
peanut oil; Renard-Tolman quantitative test; Bach's test for rapeseed oil; Lieb!jfmannStorch test for rosin oil; Baudoin's or Camoin's test for sesame'Oil; Villavecchia test; free
acid test; spontaneous combustion test. Mackey's apparatus; drying, test upon glass;
titer test, references, 1777-1785. Edible fats-butter; examination of the fat; preservatives; color; lard; water; cottonseed oil. Hardened oils. Waxes. Miscellaneous oils
and lubricants-table, references, 1785--1:789. U. S. Government specifications for
petroleum products-gasoline; burnihg oils; fuel oils; greases; lubricating oils; characteristics of the fatty acids from some oils-tables; characteristics of some waxes-table;
viscosimeter chart; factor tables; reagents, 1790-1801. Junker's calorimeter-liquid
fuels. Determination of tetraethyl lead in gasoline; nitrobenzene in oils for floors,
1802--1805.
PETROLEUM AND ITS PRODUCTS-INTERPRETATION OF TESTS
Introductory, 1806. Crude oils-physical tests; distiJlation; gasoline, naphtha-tests
and interpretation; U. S. Bureau of Mines test; A. S. T.,iM. test for gum content, 18071816. Kerosene, mineral seal and gas oil-specific gravity; other tests, 1816-1817.
Lubricants-specific gravity; color; viscosity; cloud test; pour test; flash and fire test;
carbon residue; demulsibility; fatty acid; rosin; neutralization number; evaporation
test; crankcase dilution, A. S. T. M. test, 1817-1822. A. S. T. M. methods for the
analysis of grease-sample; solvents; procedure, ash, fillers,..etc.; free alkali and free
acid; water, 1823-1829. Waxes-melting point, A. S. T. M., 1830--1833. Heavy oils
and asphalts-introductory; asphalts (references). Fuel oil, tests; A. S. T. M. method
for thermal value-apparatus; procedure; computation of results, 1833-1837. Road oils,
1837. Shale oils, 1838.
PAINTS
Introductory, 1839. Analysis of paint vehicles-"-Composition of liquid part; percentage of liquid by ignition method; percentage of liquid by extraction methods;
separation of vehicle components; water; direct distillation for volatiles; detection of
resinates; detection of various oils, 1840-1842. Analysis of paint oils-specifications for
lin!\eed oil; analysis of linseed o~l-loss on heating at, 105-1106 C.; foots, general tests;
speCific gmvity; acid number; saponification numher; unsaponifiable matter; iodine
turnber; ash; time of drying on glass; reagents for testing, 1842-1847. Examination of

CONTENTS

xiii

tung oil (Chinese wood oil)-gelation tests; Browne heat test; Worstall heat test;
analysis of various oils, 1847-1850. Examination of turpentine-A. S. T. M. methodscolor; specific gravity; refractive index; distillation; polymerization, 1850-1851. Analysis
of varnish-flash point; acid number; ash; solvent; fixed oils and resins; separati<1n of
polymerized oils and resins, 1851-1853. Other materials, 1854. The analysis of paint
pigments-classification of pigments. Analysis of white pigments-basic sulfate of
lead-volumetric determination of lead; volumetric determination of zinc; total lead
sulfate; super sublimed white lead, 1854-1858. Basic carbonate of lead-total lead
(gravimetric); total lead (volumetric); carbon dioxide; acetic acid; metallic lead. Zinc
lead and leaded zinc-moisture; lead and zinc; total soluble sulfates (absence of BaSO.;
pressure of BaS04); soluble zinc sulfate; sulfur dioxide; calculations, 1859-1862. Zinc
oxide-moisture; zinc oxide; total soluble sulfur; sulfur dioxide (Rigg Method); zinc
sulfate, 1862. Lithopone-moisture; barium sulfate; zinc sulfide; zinc oxide, 18621863. Titanium pigments-barium sulfate; calcium sulfate; titanium dioxide, 18631864. Antimony oxide, 1864. Zirconium pigments, 1864-1865. Silica or silexChina clay-asbestine-moisture 1866. Whiting (Paris white)-moisture; loss on
ignition; calcium; magnesium; carbon dioxide; sulfates, 1866. Barytes and blanc
fixe-moisture; loss on ignition; barium sulfate; soluble sulfates; earbon dioxide, 18f>7.
Analysis of a composite white paint-insoluble residue; total lead; alumina and iron
oxide; zinc; calcium and magnesium; sulfate; sulfide; carbon dioxide; calculations,
1867-1869. Red and brown pigments. Red lead :lnd orange mineral-moisture;
, organic color; total lead and insoluble residue; determination of lead peroxide and
true red lead, 1869-1871. Vermillion, 1871. Iron oxides, 1872. Blue pigments,
1872. Ultramarine blue-moisture; silica; aluminum oxide; sodium oxide; total
sulfur; sulfur as sulfate, 1872-1873. Prussian blue-moisture; nitrogen; iron and
aluminum oxides; slllfuric acid; commercial analysis, 1873. Sublimed blue leadtotal lead; total sulfur; lea~ sulfate; lead sulfite; lead sulfide; lead carbonate; lead
oxide; zinc oxide; carbon and volatile matter, 1874. Yellow and orange pigments,
1874. Chrome yellows-American vermillion-basic lead chromate-moisture; in-'
soluble residue; lead; chromium; zinc, calcium and magnesium; sulfuric acid, 1874-1875.
Green pigments. Chrome green-moisture; insoluble residue; lead; iron, alumina and
chromium; calcium and magnesium; sulfuric acid; nitrogen, 1875-1876. Black pigments-moisture; oil; carbon; ash; analysis of ash, 1876-1877.
PAPER AND PAPER-MAKING MATERIALS

T. A. P. P. 1. methods. Isolation of cellulose by chlorination method, 1878-1880.

Chlorine consumption of pulp, 1880-1881. Alpha cellulose in pulp, 1881-1883. Pitch
in wood pulp, 1883-1884. Cuprammonium disperse viscosity of pulp, 1884-1888.
Moisture in pulp by toluene method, 1889-1891. Methoxy groups in pulp, 1891-1893.
Weighing, sampling and testing wood pulp for moisture, 1893-1896. Ash in pulp;
1896-1897. One per cent caustic soda solubility of pulp, 1897-1899. Sampling paper
for testing, 1899. Conditioning paper for testing, 1900. Paraffin in paper, 1901.
Active sulfur in paper, 190f-1902. Amount of coating of mineral coated paper, 19021903. Resin in paper, 1903-1904. Moisture in paper, 1904-1905. Ash in paper, 1905.
Proteinaceous nitrogen in .paper, 1905-1906. Starch in paper, 1906-1908. Mineral
filler in paper, 1909. Total acidity of paper, 1910. Alpha cellulose in paper, 1911-1913.
Copper number of paper, 1913-1914. Acid soluble iron in paper, 1915. Hydrogen ion

xiv

CONTENTS

concentration (PH) of paper extracts, 1916--1919. Arsenic in paper, 1919-1922. Zinc

pigments in paper, 1922-1924. Titanium pigments in paper, 1925-1928. Analysis of
bisulfite cooking liquor, 1929-1931. Analysis of casein, 11931-1935. Analysis of salt
cake,. 1935-1940. Analysis of rosin, 1940-1942.
I
.
POISONS, THEIR DETECTION AND ESTIMA-:DION. WITH SPECIAL
REFERENCE TO 'ORGANIC POISONS
Introductory, classification of poisons, general procedure, groupings of poisons, 19431944. Group I-yellow phosphorus; hydrocyanic acid, procedures for detection and
estimation; chloroform, qualitative and quantitative determination; ethyl alcohol and
methyl alcohol, detection methods for each; carbolic acid (phenol), detection and estimation; chloral hydrate, detection, decomposition, 1944-1951. Group II-alkaloids,
glucosides and other organic compounds-methods of analysis;. list of or{anic poisons,
1951-1953. Acetanilid, tests; veronal, tests, extraction from bladder contents; n~cotine,
distribution, detection; strychnine, distribution, detection; brucine, detection; atropine,
detection; cocaine, detection; codeine, detection; narcotine, 'detection; morphine, detection; opium, detection; ptomaines, detection, 1953-1~62. Group III--;-preparation
of solution; detection of arsenic and estimation; determination of carbon monoxide in
air; blood solution, determination in blood; special reagents, 1962-1965.

RUBBER, GUTTA

P~RCHA

AND BALATA

Introductory-the rubber hydrocarbon, 1966--1967. Crude rubber, 1967.. Examination of raw rubber-general information; evaluation of latex; raw rubber, 1967-1968.
Technical analysis of crude rubber-samples; moisture; resin; acid value of acetone extract;
ash; copper; manganese; organic insoluble; protein, 1968-1974. Determination of the
rubber hydrocarbon-(a) Spence's method; (b) precipitation method; (c) Fendler's
method; (d) tetrabromide method; (e) nitrosite method, 1974-1976. Analysis of rubber
solutions-solvent; proportion of solids to solvent, 1976--1977. Analysis of vulcanized
rubber-moisture; water soluble constituents; full scheple for analysis of vulcanized
rubber, 1977-1979. Inspection of acetone extract. Deiled examinatioI). of the acetone
extract--total sulfur; saponifiable and unsaponifiable matter, 1979-1980. Unsaponifiable matter, 1980. Saponifiable matter, 1981. I Accelerators, 1981. Examination of
the chloroform extract, 1982. Alcoholic potash extracti"on-determination of sulfur and
nitrogen in the extract, 1982-1983. Determination of total fillers, 1983. Determination
of sulfur-total sulfur; Steven's method; Kay and Sharp"S' -method; bromine-nitric
acid method; zinc oxide-nitric acid method; sulfide sulfur-Steveri's method; sulfate sulfur; sulfite sulfur; calculation of sulfur of vulcanization. Determination of
selenium; determination of chlorine, 1983-1986. Qualitative test for glue; quantitative
determination, 1986--1987. Determination of cellulose-Epstein and Moore's method,
1987-1988. Determination of inorganic fillers-list of fillers; separation of organic
matter-Rothe's method. Determination of barium; mercury; arsenic; antimony; tin;
zinc; iron, chromium, aluminum; calcium; magnesium and alkali metals; silica and
silicates; vermillion, 1988-1991, Hard rubber (ebonite), 1991. Reclaimed rubber,
1992. Gutta' percha and balata,-sampling; washing loss; de~rmination of moisture;
resin; gutta percha and balata hydrocarbons; dirt, 1992-1994. Separation of balata
and rubber, or gu~ta percha and rubber-Charlescm's method, 1994. Physical tests,
1991H997. Bibliography, 1997-1.998.

CONTENTS

xv

RUBBER COMPOUNDS AND COMPPUNDING INGREDmNTS

Introductory, 1999. Standard methods for the analysis of rubber goods-acetone
extract; chloroform extract; alcoholic alkali extract; free sulfur; total sulfur; ash; sulfur
in ash; special determinations, 1~99-2000. Preparation of sample, 2000. Reagents,
2001. Methods of aJ:illlysis-specific gravity; acetone extract; chloroform extract;
alcoholic-alkali extract; free sulfur; total sulfur; alternative method; ash; sulfur in~h;
barium sulfate; total antimony; antimony in ash; free carbon; detection of glue; nitro~l1'
calculated as glue; unsaponifiable; hydrocarbons, A, B; rubber hydrocarbons by joint
rubber insulation committee method; cellulose; barium carbonate; calculations; rubber
as compounded; rubber by volume; results, 2002-2012. Interpretation of results-:;specific gravity, 2012. Extracts, 2012-2013. Sulfur, 2013-2014. Ash-analysis of
ash,2014. Free carbon, 2015. Calculation of probable composition of stock, 2015-2016.
Analysis of raw materials, 2016. Raw rubber, 2016. Reclaim, 2016. Accelerators and
antioxidants, 2016--2017. Softeriers, 2017. Fillers and reinforcing agents, 2018.
Carbon black, 2018. Zinc oxide, 2019. Color pigments, 2019. Solvents, 2019.

SLAG ANALYSIS
Introduction, composition of slags, 2020. Decomposition of the sample and analysis
-general reverberatory slag; chilled blast furnace slag-determination of silica; lime;
iron, 2021-22. Determination of other constituents in slags-alumina; manganese;
zinc; 'magnesia; copper-?olorimetric; electrolytic and iodide methods, 2022-2025.'

METHODS OF ANALYSIS/OF SOAPS AND SOAP PRODUCTS

Introduction, 2026. Preparation of samples for analysis-cakes or bars; powders,
flakes and chips; paste soap products; liquid soap, 2027. Methods of analysis-moisture
and volatile matter (oven method), 2028; distillation method, 2028; total fatty acids,
fatty anhydride, combined alkali and anhydrous soap-soaps containing no coconut
oil, 2029; soaps high in coconut oil, 2030; chlorideS; 2030; unsaponified and unsaponifiable
matter, 2031; unsaponifiable matter, 2032; free alkali or free fatty acid, 2032; sodium
carbonate, 2033; determination of carbon dioxide, 2034; determination of silica present
as alkali silicates, 2034; borax, 2035; perborates, 2036; phosphates, 2036; sulfates,
2037; total alcohol insoluble matter, 2037; water insoluble matter, 2038; glycerine, starch
and sugar, 2038; volatile hydrocarbons, 2040; titer test, 2042; iodine value, 2043; saponification value, 2043; acid value, 2044; rosin in fatty acids, 2044.

WATER ANALYSIS
General considerations, 2046. Sanitary analysis-o~ganic nitrogen; chlorine; oxygen
consumed. Physical tests-turbidity; color; odor-cold, hot; table; taste, 2047-2049.
Chemical tests. Free ammonia; albuminoid ammonia; organic nitrogen; nitrogen as
nitrite; nitrogen as nitrate; oxygen consumed; chlorine as chlorides; total solid residue,
2049-2055, Interpretation of results, 2055-2057: Mineral analysis. Outline of
procedure; silica, iron; colorimetric; aluminum, calcium, magnesium; manganese,
persulfate method; bismuthate method; sulfates-benzidine method; sodium, potassium
and lithium; alkalinity, hydroxyl ion concentration; carbon dioxide-evolution method;
acidity; hydrogen-ion concentration; free carbonic acid; chlorine; nitrates; ammonia arid
its compounds; total mineral residue; hydrogen sulfide; ~il; dissolved oxygen by Winkler's
method, 2057-2080. Methods for the determination of small amounts of lead, zinc,

xvi

CONTENTS

copper and tin, 20RG-2081. Hardness-total hardness;, preparation of solutions; operation; table; magnesium chloride; calcium sulfate; lime and soda value; methods of
reporting and interpretation, 2081-2088. Field assay of w!1rer, 2088. Boiler room control methods-introduction; solutions; determination of pliosphate; Buromin method;
sulfate; turbidimetric method; volumetric method, 2088-2093.
Rapid, methods for water analysis-special reagents; methods of procedurechloride; carbonate, bicarbonate and hydroxyl; sulfate; vlkaline earths; alkali del
termination, 2093-2097.
Irrigating waters, 2098.
MICROSCOPICAL EXAMINATION OF WATER ~UPPLY
Importance of the determination; table of odors of organisms, 2099-2101. Sampling
for analysis, 2102. Sedgwick-Rafter method of examination-apparatus required;
procedure; tabular outline identification of forms; li'lt of organisms; i4entification of
forms; control of microscopic organisms, 2102-2112.
Tabl~ of chemicals required for treatment of different genert, 2113. References,
2114-2115.
Bacteriological examination of water-sampling; apparatus needed; adjustment of
reaction; sterilization; nutrient broth; sugar broth; nutrient gelatin; nutrient agar;
eosin methylene blue agar; endo medium; brilliant green, lactose, peptone bile, 211&-2123:
Examination of water for total bacteria; examination forB. coli; subclassification of B.
coli. Coli-aerogenes group-reaction classification; presumptive test; confirmed test;
completed test; selection of coli-aerogenes tests; examination for B. typhi, etc., 2124-2133.
MISCELLANEOUS ANALYSES
Acetone-reagents; sa~ple; procedure; specific gravity; acidity; water; color; residue;
miscibility with water; KMn04 test; aldehydes. Analysis of acetone by Messinger's
method, 2134-2137. Alcohol-detection and estimation-iodofor.m test; Berthelot's
test; ethyl acetate test; Determination-distillation method; evaporation method;
analysis of grain alcohol (ethyl alcohol, spirits of wine, cologne spirits)-per cent ethyl
alcohol; non-volatile; acidity expressed as acetic, acid; aldehydes; qualitative; quantitative-reagents; procedure; furfural; esters expressed, as ethyl acetate; fusel oil; nitrates;
sulfur compounds, 2137-2141. Specific gravity tables, 2142-2146. Methyl alcoholdetection by method of Riche and Bardy; by oxidation method; in presence of ethyl
alcohol; analysis of mixtures of methyl acetate; methanol and water, 2147-2148. Formaldehyde-Leach's test; fIehner's test; morphine sulfate teat; determination, 2149.
Glycerol. Acetin method-reagents; method; calculation. Bichromate processreagents; method, 2150-2154. Light oil in coke oven gas-apparatus; procedure;
distillation; calculation; volume of gas required, 2154-2156. Starch and starch productssampling; preliminary examination; moisture; ash;,acidity; alkalinity; test for alkaline
starch or alkali; viscosity; converted starches-sampling, moisture ash; total acidity;
viscosity; added materials; sizing strength, 215&-2158. Tannic acid-reagents; total
astringency-Lowenthal-Proctor method, 2158-2159.
Bleaching powder-sampling; available chloride; chlorates; total chlorides; bases
and silica; free lime; quality;. wood and coarse impurities; proximate analysis for sand
and grit, 2159-2161. . Analysis of clay-chemical analysis; alkaJies; titanium; state of
silica. 2161-2164. Analysis ,of glass-density; sample !}nd determination o~ silica; lead
oxide; ferric oxide and alumina; lime; magnesia; alkalies; boric 'oxide; manganese dioxide;
sulfur; barium oxide, separation from lime; factors, 2164-2169. Glass-making ma~ls

CONTENTS

xvii

-potash; feldspar, kaolin, China clay; barium carbonl}te; lim'c, 2169-2174. Printing
inks-separation of oil and pigment; analysis of the oil; analysis of the pigmcnt: (a)
black inks; (b) blue inks; (c) red inks; (d) green inks, 2174-2180. Dctermination of
peroxides-hydrogen peroxide-Kingzett's iodide ,method; arsenious acid method of
Jamieson,2180-2181. Sodium sulfide-complete analysis-water insoluble; total iodine
titration; sulfides; thiosulfate and sulfite; alkalinity, polysulfide sulfur and sodium
sulfate; sodium sulfhydrate liquor-calculations; sodiuin chloride; notes, 2181-2187.

PART II.

SPECIAL TECHNIQUES, APPLICABLE TO

DIVERSE SUBSTANCES
ACIDIMETRY AND ALKALIMETRY

Indicators-classification; special uses; stock solutions, 2191-2192. lntimate standards; preparation of pure sodium carbonate, 2193. Preparation of standard acidsstandard sulfuric acid; standardization; hydrochloric acid from constant-boiling acid;
gravimetric determination; benzoic acid standard; standard alkali solution; standardizatio.n against potassium acid phthalate, 2194-2198. Preparation of reagents by specific
gravity method, 2198. Standard burett{)S, 2199.
TITRATION OF ACIDS AND ALKALIES
Methods of weighing acids-dilute acids, non-volatile under ordinary conditions;
concentrated acids, fuming or volatile under ordinary conditions; Lunge-Ray pipette;
Dely weighing tube; snake weighing tube; Blay-Burkhard weighing burette, 2200-2203.
Analysis of muriatic acid-total acidity and HCI; impurities in commercial hydrochloric
acid, free chlorine; nitric acid or nitrates; sulfuric acid and sulfates; arsenic; barium
chloride; total solids and silica, 2203-2205. Determination of specific gravity, 2205.
Table of specific gravity for H('jl, 2206-2209. Analysis of hydrofluoric acid-special
apparatus; total acidity and hydrofluosilicic adid; sulfuric acid; sulfurous acid; residue,
2209-2211. Complete analysis of nitric acid-determination of total acidity; sulfuric
acid; hydrochloric acid; lower oxides determined as nitrous acid; nitric acid; iodine; free
chlorine; total non-volatile solids, 2211-2214. Ferrous sulfate method for the direct
determination of nitric acid-reagents; standardization of ferrous sulfate reagent; general
procedure. Determination of nitric in sulfuric acid; evaluation of nitric acid or nitrates;
Determination of nitric acid in oleum or in mixed acids; determination of nitric acid in
arsenic or phosphoric acids, 2214-2219. Determination of nitrous acid or nitrite by
per~anganate method, 22'19. Specific gravity table for nitric acid, 2220-2223.' Volumetric determination of phosphoric acid and its salts, 2224-2226. Specific gravity
table for phosphoric acid, 2227. Analysi:! of sulfuric acid-procedure for titration with
standard NaOH, 2228. Determination of impurities in: sulfuric acid-residue on heating; lead; iron; arsenic; antimony; zinc; selenium; sulfur dioxide; hydrochloric acid;
nitric acid; fluorine, 2229-2233. The analyses of oleum, or fuming sulfuric acid, and of
mixed acid-total acids; lower oxides; sulfuric acid and free sulfuric anhydride; nitric
acid; calculation of results, 2233-2237. Table of specific gravity for sulfuric acid,
2238-2243. Method 9f analysis of chlorosnlfonic acid, 2244.' Volumetric estimation
of free acid in the presence of iron salts, 2245. Organic acids. Formic acid, 2246.
Acetic anhydride-acidity method; aniline method, 2247-2248. Acetic acid-impurities;
formic acid; furfurol; acetone; sulfuric, sulfurous, hydrochloric acids in acetic acid;
metals in acetic acid, 2248-2251. Acetates, 2251. Specific gravity table 'for acetic
acid, 2252. Carbonic acid-free; bicarbonate; carbonate, 2252. Citric acid, 2253.

xviii

CONTENTS

Oxalic acid, 2253. Phenol (carbolic acid), 2253. Tartar~c acid-acid potassium tartrate, Oudeman's method; Goldenberg method for total tartaric acid; impurities iri
tartaric acid, 2254-2255.
ALKALIES
Analysis of sodium hydroxide-procedure;' determination of impurities; Sodium
carbonate; sodium chloride; sodium sulfate; sodium silicate; 'sodium aluminate; insoluble
matter; water, 2255---2257. Strength of caustic liquors by the Itydrometer, 2257.
Potassium hydroxide, 2257. Carbonate in presence of bicarbonate, 2257. Standard
methods of the Solvay Process Company: analysis and testing of impurities in caustic
soda. Reagents and indicators; standard solutions; methods of analysis-silica; ferric
oxide and alumina; calcium carbonate; magnesium carbonate; manganese; copper,
2257-2261. Analysis of sodium carbonate, soda ash-procedure: total alkali;, bicar7
bonate; sodium carbonate; sodium chloride; insoluble m;tter; iron, 'alumi~a, lime,
niagnesia; sodium sulfate, sulfite, silicate, sulfide; loss on ignition, 2261-2262. Modified
sodas, 2263. Analysis of sodium bicarbonate; mixtures of hydroxides and carbonates;
mixtures of carbonates and bicarbonates, 2264-2267. Tables of specific gravity, NaOH
and NH.OH, 2268-2269. Analysis of aqua animonia, 2270. Analysis of alkaline
mixtures, 2271-2273.

DETERMINATION OF pH BY THE COLORIMETRIC METHOD

Introductory; meaning of pH; the pH sgale, 2274-2278. Buffer ation; buffer systems, 2278-2281. The use of indicators-introductory; table of indicators; essentials
of the method; color comparison, 2281-2284. Colorimetric determination of buffered
materials, 2284. Unbuffered or slightly buffered materials, 2284. Salts-effect on pH;
tables; 2285-2287. Effect of proteins; table for gelatine, 2287-2288. Effect of temperature, 2288. Specific effects-dichromatism; effect of suspensions, emulsions, etc.,
on color, 2289. Effect of chlorine, 2290. Colorimetric pH eqUipment-Taylor slide
comparator; long range slide comparator; Roulette comparator; the Hellige comparator,
2291-2296.

ELECTRO METRIC HYDROGEN I<;lN MEASUREMENT

Potentiometer principle, 2297. Electrode systems-reference electrodes; calomel
electrode; silver chloride electrode, 2298-2300., Measuring electrodes-hydrogen
electrode; limitations; platinizing electrodes; hydrogen gas, 2300. Quinhydrone ~lec
trode; limitations, 2301. Antimony electrode, 2302. Glass electrode, 2303. Working
'
'_
formulae, 2303-2304.

tY"POTENTIOMETRIC METHODS
Single electrode potentials, 2305. Oxidation-r~duction electrodes; reference electrodes, 2306-2308. Potentiometric titrations, 2308. Neutralizatipns, 2309-2312.
Precipitations and complex formations, 2312-2314. Oxidation-reduction reactions,
2314-2315. Determination of chromium'and vanadiu~ in steels, 2315-2318. Simplified titration methods, 2318. Pinkhof-Tteadwell me~hod, 2318. The differential
method, 2318. Bimetallic electrodes-dissimilar electrodes; polarized systems, 2320.
Vacuum-tube voltmeters, 2321. Micro methods, 2322.
I

CONDUCTOMETRIC ME-TEODS
The principles of conductometric titrations, 2323.

Mobility

of ions, table,

2324.

CONTENTS

:xix

Change in conductance during titration-various cases, 2324-2325. Performance of

the titrations-apparatus, 232q-2328. Application of the method to acid-base titrations
--strong acid with it strong base; intermediate weak acids and bases; very weak acids
and bases, 2328-2331. Mixtures of acids, 2332. Replacement titrations, 2332. Precipitation and complex-formation, 2333. Applications, 2335.

GAS ANALYSIS
Sampling-tubes; pumps; containers for samples, 233~2339. Measurement of large
gas volumes':_meters; Pitot tube; rotameter; Thomas electric gas meter; orifice meter;
anemometer; flow-meter; Venturi meter, 2339-2342. Measurement of gas in small
quantities-Hempel gas burette; separatory funnel and graduate, 2342-2343. Standards
of measurement; specific gravity, 2343-2344. Absorption apparatus, tubes and pipettes,
2344. Examination of the gases-detection and determination of the various gases;
tables of groups, 2345-2348. Analysis of gases-analysis by Orsat apparatus-manipulation; carbon dioxide, oxygen, carbon monoxide, hydrocarbons; accuracy, 2349-2351.
Eliott apparatus-description; manipulation; carbon dioxide, oxygen, carbon monoxide;
accuracy; time required, 2351:-2353. Hempel's apparatus-description; manipulation;
determination of oxygen in air-by phosphorus; by potassium pyrogallate; by explosion
with hydrogen; analysis of manufactured or city gas-carbon dioxide, illuminants,
oxygen, carbon monoxide, methane and hydrogen (a) by Hinman's method, and (b~ by
Hempel's method; calculation (a) Hinman's method; (b) Hempel's method; nitrogen;
acc~acy and time required; not~s, 2353-2359. The properties of the. various gases and
the methods for their determination-group I-ammonia, nitrous anhydride, nitrogen
dioxide, 2360-2361. Group II-chlorine, hydrochloric acid, cyanogen, hydrocyanic
acid, hydrogen sulfide, sulfur dioxide, silicon fluoride, carbon dioxide, 2362-2366.
Group III-phosphine, arsine, stibine, 2366. Group IV--oxygen, ozo~e, 2367-2370.
Group V-carbon monoxide, 2370-2373. Group VI-nitric oxide, 2373. Group VIIhydrogen, methane, ethylene, acetylene, nitrous oxide, nitrogen, carbon oxysulfide, and
the noble gases, 2373-2380. Applications of gas analysis and interpretation of results'chimney and flue gases-carbonic acid indicators; temperature; composition of coal;
tables and chart-combustion efficiency; loss due to carbon monoxide; smoke, 2380-2388.
Producer and fuel gases. Blast furnace gas. Tables; determination of dust, 2388-2389.
City gas. Candle power; calorific power-Sargent calorimeter; Junkers calorimeter;
llIlanipulation; calculations; Thomas recording meter. Calculation of B.T.U. Sulfur;
hydrogen sulfide; ammonia; naphthalene; carbon dioxide; specific gravity; table of
composition of commercial gases, 2389-2399. Sulfuric acid gases-burner gases; sulfur
dioxide; nitrogen oxides; oxygen, 2399. Mine gases-methane; carbon monoxide,
2400-2402. Electrolytic galles, 2402. Acetylene-determination of oxygen, sulfurcontaining gases, phosphine, and acetylene, 2402. Atmospheric air-moisture; carbon
dioxide; carbon monoxide, 2403-2407. Natural gas, 2407, 2409-2410. Determination
of moisture in gases, 2408. Determination of gasoline vapor in gaseous mixtures, 2408.
Calculations,2411-2414. Determination of suspended matter in gas, 2414-2417. Determination of nitrogen by the nitrometer, 2418-2419. Compressed gases and their
handling,2420. Reagents,2421-2424. Tables, 2424-2433.

CHEMICAL MICROSCOPY
Advantages of microscopical metho.ds, 2434. Studies of single substances-identification of substances; optical properties, 2434-2437. Studies of mixtures of substances--;

xx

CONTENTS

various advantages of trw microscopic method, 2438. H{tndling small amounts of

material-technique; filtration; decantation; distillations, 2439. Microscopical qualitative analysis, 2440. Methods of applying reagents, 2441.1 Useful group reagentspotassium mercuric thiocyanate; cesium chloride; cesium, \lulfate; potassium nitrite;
silver nitrate; barium chloride, 2442-2443. Identity tests, for anions and cationscations, 2443-2444. Scheme for qualitative analysis of 'common non-ferrous alloys,
2445. Anions, 2445. Analysis of materials of unknown compositiop'-crystallographic
examination, 2446. Reasons for failure to obtain a positive test, ,2447. Microscopic
quantitative analysis; criteria; preparation methods; analysis by estimation, 2447-2449.
Analysis by counting particles---counting one constituent onlYi counting all constituents,
2449-2451. Analyses based on measurements of area-areal and linear analysis, 2452.
Determination of physical properties, 2453. Measurements; particle-size determinations, 2454-2456. Apparatus for microscopical analysis, 2456. Bibliography, 24572459.
QUANTITATIVE MICROCHEMICAL ANALYSIS
Introductory,2460. The microlabcratory, 2461. The microchemical balance, 2461.
Weighing, notes, 2463. Sampling, 2464. Drying and weighing samples-hygroscopic
substances; liquids; gases, 2464-2469. Treatment of rubber tubing; micro Neubauer
crucible, 2469. Filtering tube, 2470. Removal of traces of precipitate. Immersion
filter,2470-2471. Burettes. Standard solutions; titration. Cqlorimetry. Heating of
combustion tubes; electric furnaces. Equipn;.ent,2472-2475. 'Determination of carbon
and hydrogen in organic material-apparatHs; absorption tubes; procedure; notes;
2475-2482. Determination of carbonate carbon-apparatus; procedure; notes, 24822484. Nitrogen by the micro Dumas method-apparatus: carbon dioxide generator;
combustion tube; nitrometer. Procedure; calculation; according to Pregl; according to
Nieder! and Trautz; notes, 2485-2491. Nitrogen by micro Kjehldahl method-apparatus; procedure; notes, 2492-2494. Determination of halogens in organic compounds-I,
according to Pregl-apparatus; procedum; notes; volumetric determination of iodinereagents; procedure; notes. Volu~etl'ic determination off chlorine and bromine-solutions; procedure; notes; titration using adsorption iridicatdl'S; notes, 2494-2499. Micro
determination of phosphorus in organic substances-;-preparations of solutions; methods
of oxidizing organic material; treatment of solution with molybdate; volumetric method;
procedure wl).en arsenic is present, 2499-2501. Colorimetric determination of phos-.
phorus and arsenic-reage,r;tts; procedure; notes, 2501-2503. ,rhe microanalytical determination of arsenic-oxidation of the substance; gravimetric detsf!.Ilination; volumetric determination-procedure; arsenic in cadavers. Gutzeit method, 2503-2507.
The micro determinatio~ of organic sulfur-apparatus; procedure; determination by
fusion; notes; volumetric determination of sulfur-Tel' Meulen hydrogenation methodsolutions; apparatus; procedure; notes, 2507-2510. Microelectroanalysis-cell of Clark
and Hermance; apparatus; procedure; determination of wckel; zinc; tin; traces of metals.
Traces of zinc in aluminum solutions; lead in the. presence of zinc; traces of copper in
nickel soluti?ns. Cadmium; mercury, 2510-2516. The Pregl cell-apparatus; procedure;
determination of copper in alloys;, deteI;mination of mercury; notes; gold; silver, 25162518. Copper in organic or biological material-ashing; treatment of ash. The Biazzo
method for cOPPCl:-standards; notes, 2519. The carE,alnate colorimetric method, 2520.
Gmvimetric determination of nickel and copper-Brimck's method; procedure; copper,
2g:,:,. Vollmetric determination of iron-reagen~; procedure; notes, 2522. Determina-

CONTENTS

xxi

tion of acetyl-method of Elek and Harte; method of Robinson and Bailey, 2523-2526.
The determination of methoxy-ethoxy-modified Zeisel procedure-apparatus; reagents;
procedure; calculations; notes, 2527-2530. The determination of molecular weight-for
solids by rise in boiling point method, 2530. Table of constants useful in ebullioscopy,
2532. Method of Rast, 2533. Molecular weights of liquids"':""micrq Victor Meyer
method-apparatus; determination; calculation; notes, 2534-2536. Micro gas analysisapparatus; collection and storage of samples; preparation of samples; preparation of
beads; ope~ation of apparatus; specific determinations: oxygen; water vapor; carbon
monoxide; ethylene; acetylene; hydrogen chloride; ammonia; nitrogen and inert gases;
acetaldehyde and similar compounds; the determination of combustible gases; nitrous
oxide; methane; hydrogen-oxygen, 2537-2545. References-applications; special
methods of analysis; physical methods; determination of organic radicals; fats; specific
compounds or elements, 2545-2547.
c

METALLOGRAPHY
Introductory, the microscope, pho(ographic materials; preparation of specimens for
microscopical examination, 2548-2552. Etching sample for examination; reagents used
in etching, 2552. Equilibrium diagrams; photomicrograph of pure iron; eutectic
definition, 2553-255G. Copper-zinc alloYf\; brass, 2557. Iron-carbon alloys; characteristics of carbon steel, 2558-2563. Influence of mechanical work; slip bands; Neumann
lines; critical or exaggerated grain growth, 2563-2566. Heat treatment of steel, 2566.
Case hardening, 2567. Influence of some elements on iron and steel, 2568;-2569.

METHODS FOR THE DETERMINATION OF SOLUBILITY

Introduction, 2570. Apparatus for the determination of the solubility of solids by
the analytical method, 2571-2579. Separation of saturated solution and undissolved
solid, 2579-2580. Analysis of the saturated solution; identification of the solid phase,
2580-2582. Transition temperatures, 2582. Synthetic method, 2582. Freezing-point
method,2584. Volume change method, 2584. Titration method, 2585. Distribution
coefficients, 2586. Electrolytic conductivity method, 2586. The solubility of gases in
liquids, 2587-2591.

SPECTROGRAPHIC ANALYSIS
Introduction,2592. Apparatus, 2594-2598. Light sources; flame, arc, spark, other
light sources, 2598-2600. Spectrographic standards, 2600. Tables-spectral lines
arranged according to elements, 2602-2603; wave lengths, 2604-2607. Quantitative
analysis; comparison with standards, internal standards, measurement of intensities
of lines, control of conditions, plates, 2609-2614. Absorption spectra,. 2614-2616.
Special note-patent situation, 2617; references, 2617.

INDICES TO VOLUMES I AND IT

Inside of front cover-Table of atomic weights.
Inside of rear cover-Table of solubilities.

Author
{ Subject

PART I
SAMPLING,
PRELIMINARY TREATMENT
4.ND
APPLIED ANALYSIS

STANDARD METHODS OF SAMPLING 1

GENERAL OUTLINE FOR SAMPLING SOLIDS

INTRODUCTION
An accurate and uniform sampling procedure, applicable to all solid mate~
rials from the viewpoint of both producer and consumer, would necessitate
the standardization of an infinite number of details, and probably result in a
very cumbersome and impractical manipulation. It, therefore, appears advisable to first consider the various steps of the process of sampling, attempting
their standardization and then applying this standardization to each particular
product or group of related products.
The process of sampling is divided into three major operations:
1. The collection of the" gross" sample.
2. The reduction of the " gross" sample to a proper 'and convenient size
for transportation to the laboratory.
3. The preparation of the sample for analysis.
These essential points and the accuracy with which they are performed
determines the value of the subsequent analytical results. Experimental data
and general mathematical deductions make it possible to standardize these
operations with a fair degree of accuracy.

SAMPLING UNIT 2
The sampling unit may be defined as that portion of the material which is
chosen in such a manner that there is a high probability that it will contain
the different sized particles of the material in the proportion in which they
occur in the entire bulk of the material. Providing there are no great uncontrollable irregularities in the mateJOial, the sampling unit, if taken under
carefully prescribed conditions for each different class, should be properly
representative of the material. However in order to obviate the necessity of
specifying in great detail, and to guard against unforeseen sectional variations
in the material, it is more expedient to select a number of sampling units,
these depending upon .the size of the shipment to be sampled, and combine
these sampling units into one large sample called the" Gross Sample."
The character of the material, size of the various particles, uniformity of
composition and the prominence of anyone or more constituents are factors
governing the determination of the amount to be taken as the sampling unit.
Without considering a mathematical discussion of the modulus of precision
of a condition which would be .impossible of duplication and in view of the
Chapter by J. B. Barnitt .
See: Chemical Analysis of Aluminum, Aluminum Research Labs., 1935, pp. 1-3.
Sampling and Evaluating Secondary Metals, T. A. Wright, Am. lnst. Min. & Met.
Engineers, Tech. Pub. No. 81 (1928). Also A. S. T. M. 1935 Proceedings, on Testing
in the Precious Alloy Field.
1301
1

1302

STANDARD METHODS OF SAMPLING

experimental data compiled on the sampling of coal, lit is safe to adopt, as the'
general sampling unit, an amount of material equal to approximately 500 times
the weight of the largest particle, If 'this is impractical, due to the presence
of very large particles, and the material cannot qe reduced by passing it
through a crusher, it will be necessary to reduce, bY- hand, the large particles"
at the various sampling points 8,0 that the correct proportion of large particles
can be included in the sampling unit. In some cases, however, a different
'
bulk of sampling unit will be found necessary.

COLLECTION OF THE GROSS SAMPLE

The amount of material to be included' in the gross' sample depends more
upon the size of the individual particles than upon the size of the shipment
since it is simply a multiple of, the sampling unit. When the shipment is .
large and frequent sectional variations do not prevail, and it is being transferred by containers with a capacity approximately equal to the sampling
unit, it is advisable to' take the entire conte;nts of a definite percentage of the
containers as increments of the gross sample.
When the accumulated gross sample is too large to handle conveniently
as a unit, it is permissible to reduce the gross sample in small sections by the
standard procedure and subsequently combine the reduced portions in the
correct ratio.
/'
'
1. FROM CARLOAD LOTS
(a) When the Material is to be Unloaded from Flat-bottomed Cars.Divide the top surface of the car into eight equal areas and, from each of these
sections, take approximately the equivalent of one sampling unit in the follow- '
ing manner: From a stable bank of the material, beginning from the bottom,
take with a shovel, at regular and equal intervals from the bottom'to the top
of the bank, ten equal portions of the material as/specified. Combine these
sampling units 'for the gross sample.
I
I
The stable bank may be obtained:
1. By digging down to the bottom of the; car at the center of each section.
2. As the center of each section of"the car is reached during the ordinary
process of unloading by shoveling or otherwise.
(b) When the Material is Being Transferred from Or'to.the Car by Means
of Wheelbarrows, Tram-cars, Wagons, etc.-Take the specified increment
according to the material in question from a specified number of conveying
containers. Combine these increments for the gross sample. When at any
point a particle is encountered which is larger than the specified increment
it should be broken down and a portion of it included in the'sample.
(c) When the Material is Dumped from a Hopper Car into a Hopper, Bin
or Pile.-Under these conditions, the sample is taken from the resulting coneshaped pile by beginning at the bottom outer edge of the pile and taking the
specified increment, according to the character of the material, every two feet
on a straight line. to the apex of the cone. This.operation is repeated on each
quadrant of the pile, The entire procedure is repeated after. each car has
been du'uped and all increments combine~ for the gross sample.

STANDARD METHODS OF SAMPLING

2.

1303

FROM WHEELBARROWS, BARRELS, 'BAGS, TRAM CARS, CARTS, TRAYS, ETC.

(a) From a specified number of containers, take the specified increment

and combine them for the gross sample.
(b) Take the entire contents of every nth container and combine them for
the gross sample. 3
(c) If the material is fairly uniform 4 and composed of small particles,
sample the container by taking ten equal specified increments uniformly distributed on two right angle diagonals across the surface of the container .
. Sample the specified number of containers in this way and combine the several
portions for the gross sample.

3.

FROM CRANE BUCKETS, CARS, CARTS AND SIMILAR CONTAINERS,

CHUTES, ETC.

(a) If one container amounts to less than one sampling unit, take every
10th load and combine these for the gross sample.
(b) If one container amounts to approximately 1-3 sampling units, take
. every 20th load and combine these for the gross sample.
(c) If one container amounts to more than three sampling units, take every
50th load and combine these for the gross sample. Note: When very large
shipments are being sampled by the above procedures very large gross samples
will result which would be difficult or inconvenient to reduce to the, laboratory
sample. In this case the gross sample may be reduced periodically as follows.
At regular intervals, representing a definite number of tons, the gross sample is
quartered, one quarter reserved and three quarters returned to the main
bulk. The total combined quarters thus reserved are mixed by shoveling
into a cone-shaped pile and then further reduced by the Standard Procedure
to the laboratory sample.
(d) When the material is composed of particles one inch or less in size,
take a specified increment from every nth container.
(e) When the material is being dumped into a crusher from which it flows
or is flowing, from a drying kiln or other process kiln, the material may be
sampled by taking specified increments from the outflowing stream at stated
regular intervals.

4.

FROM SHIPHOLDS, SCOWS, HOPPERS, HOPPER-CARS, BINS, STORAGE

PILES, ETC.

In this case the material may be sampled in any of several ways.

(a) According to the previous section 3, while loading or unloading.
(b) According to the section 1, for carloads while unloading or loading.
(c) While using material from a storage pile, daily or periodic samples may
be taken as follows: On ten vertical lines from the bottom to the top of the
fac'e of the pile distributed at uniform distances across the face, take ten equal

3 In cases where experience indicates the procedure to be justifiable, it may be permissible to take a portion of the contents of every nth container.
4 Relatively few materials are sufficiently uniform to be sampled in this manner.
A more thorough method of sampling is used for materials containing precious
metals. .

1304

STANDARD METHODS OF SAMPLING

specified increments at equal distances from bottom to top.

mcrements for the gross sample over any r]""j,."r] "",.jAr]

Combine these

SAWMETI-IOD

( iii

i ! )

'\! : : : i?
Templet

SAMPLE PUNCH
,
PUNCI-I METI-IOD
("

mr~

(Top View

Ib

J.{'Dia.

1
FIG. 151.

Apparatus for Sampling Solids.

Seet. 1. Thief for sampling solids >" diameter or smaller. OpeningH to correspond. Both tubes hollow and to be made from is" sheet brass (hard). Inside tube
to fit snugly into outside tube. Point to be steel or iron.
Sect. la. Saw Method of Sampling Pigs. The bars are sampled in sets of five,
according to Template as shown. Saw is sharpened on Emery Wheel to size and shape
here shown.
I
Sect. lb. Punch Method of Sampling Pigs. 6 The' bars selected as samples are
placed in a line, with every other bar, bottom side up. The sampling is done according
to template in sets of five bars each, as indicated above. The punch must be driven
almost through the bar. If a large sample is desired, the bars are turned over and
sampled on the other diagonal.
I
Sect. Ie. Sample Punch.
6

Lear!

id

best sampled when molten (" gum-drop" method) rather than in' pig form.

STANDARD METHODS OF SAMPLING

5.

1305

FOR POWDERED MATERIAL

When the material is a fine powder or is composed chiefly of material under2 mm. and contains a small amount of unsegregated partiCles up to 10 mm.
in the largest dimension, the sampling unit is taken by means of a specially
designed thief (Fig. 151) or by means of a suitable shovel, dipper, etc.
(a) From large shipments, as shipholds, scows, barges, cars, etc. On a
line through the center of the holder lengthwise, take portions with the thief
every five feet. Repeat this operation on each of two lines parallel with and
half way between the first line and each side of the holder. Combine all the
portions for thE:) gross sample. If the shipment is composed of more than
one holder, combine the portions from each holder into a composite grt)s~
sample.
(b) From tram cars, carts, buggies, etc.
, Take a specified portion from each h"Older with the thief or shovel and
combine 'all portions for the gross sample.
(c) From barrels, bags, wheelbarrows, etc.
Take a portion from every nth container and combine the portions for the
gross sample .
. (d) From.small piles (not more than six feet in height).
With the thief, take one portion vertically at the peak of the pile and one
vertically at each of ten points uniformly distributed along a line around the
pile half-way between the peak and the edge of the pile. Combine these

portions for the gross sample.

(e) FroJ)l a well-mixed bulk of material (from mixers, packing machines,
etc.).
After allowing a specified quantity, according to the condition of the
material and type of apparatus, to flow past the sampling point, take the
specified quantity for a sample. Periodic samples may be taken in the same
manner, as often as desired.
6.

FROM WET MATERIAL

. Wet materials arrange themselves in several classes:

When the material is thoroughly wet but not dripping, take the sample
by any of the foregoing suitable procedures, keeping the collected portions
in moisture-tight containers.
(a) When water is dripping from the material, take the sample as prescribed
for the particular material by one of the foregoing procedures and allow the
total gross sample to .drain, the drainage and remainder of the sample being
weighed and recorded. If further elimination of water is necessary either by
air drying, artificial drying or pressing, before the gross s~mple is reduced
to the laboratory sample, the total water eliminated must be calculated and
taken into account in the final evaluation of the material to the original wet
condition. In case this water carries some of the material in solution, it will
be necessary to ascertain the llmount of this by analysis .and include the results
in the final evaluation.
(b) When fine solid material is non-uniformly suspended in a liquid several
vertical sections of the material may be taken by means of a specially designed
thief and combined for the gross sample. (See general outline for sampling
liquids.)

1306

STANDARD METHODS OF SAMPLING

7.

FROM HYGROSCOPIC OR DELIQUESCENT MATERIALS

(a) When such material is contained in airtight con,tainers which may be

readily opened and closed, take a sample by meanf1 of a thief (Sect. 1), when
the material does not contain resistant particles, larger than Y2 inch in croSs
section or, otherwise, by means of a small scoop or shovel. These increments
should be placed and retained in an air-tight receiver until 'ready for 'the
.
analysis. 7
~
(b) When such material can be sampled while being transferred to the
final container for shipment, small and equal increments at uniform and regular
intervals, are taken directly from the stream by means of a suitable dipper
and combined for the gross sample.

8.

FROM KILNS, ROASTERS, DRYERS, 'FURNACES, ETC.

This class of material constitutes a special case,' and, consequently the

size of sampling unit and manner 15y which' it may be tl}ken .will be specified
according to the size of particles and charac:;ter and conditioJ of the material.
9.

SAMPLING NON-FERROUS

M~TALS i~m

ALLOYS 8

(a) In the Form of Pigs, Billets, Ingots, Sheets, Slabs, etc.-One pig, etc.,
.should be taken to represent each ton of metal in the lot and each of these
should be sampled by one of the following methods after thoroughly cleaning
the surface to be sampled.
,/
.
1. By sawing completely through the specimen as illustrated by Sect. la,
Fig. 151. The sawdust from all the specimens' is then thoroughly mixed and
quartered down on a clean surface and the required amount of sample drawn.
2. By punching or drilling completely through if the equipment permits,
or halfway through from two opposite sides as illustrated by Sect. le, Ib,
Fig. 151; In this case the holes shall be spaced along a diagonal line from one
corner of the specimen to the other. Sampling in this wanner may be so
arranged that one or more holes are made in each! of several specimens or a
group in such positions that they represent consec6.tive positions on the diagonal of a single specimen. (See illustration in Sect. Ie.) These punchings or
drillings are carefully melted in a clean graphite crucible, and either granulated
by carefully pouring into distilled water and thoroughly drying, or by casting
into thin slabs which may be sawed completely through in several places and
the sawdust treated as described above. Drillings may'be- chipped and mixed
if convenient.
.
(b) In the Form of Sheets.-In sampling brass discs, ana sheets, when
they can be crated so that the edges are flush, recourse can be made to a portable milling machine operated by a 7i H.P. 'electric motor run transversely
across the edges, so adjusted that 'only a very thin milling is taken from each
disc or sheet. These are coned and quartered for the sample. By this method
the serviceability of the disc or sheet is not destroyed, and the completeness

Readily oxidizable material must be stored in similar fashion (iron

See E. Keller, Bur. Mines Bulletin, 122 (1916), 105 pp. For steels,
Methods of the Chemists of the U. S. Steel Corporation for Sampling
of Iron Ores, etc. Pub!. by Carnegie Steel Co., PittSburgh, Pa. See
E 30-36, pp. 2-6, Methods of Chemical Analysis of Metals (1936).
7

pyrites, etc.).
ores, etc., see
and Analysis
A. S. T; M.,

STANDARD METHODS OF SAMPLING

1307

of the sampling depends only on the accuracy of adjusting the edges in the
crate.
PROCEDURE FOR REDUCING THE GROSS SAMPLE
By combining the several sampling units, the gross sample is o'btained
which should now be so large that large single chance particles of material
could be entirely foreign matter or entirely pure substance and affect only
slightly if at all the final sampTe. Increasing the size of the gross sample
increases its accuracy, but cost and convenience of collection and reduction
of sample, limit the degree to which this can be carried to advantage. The
reduction of the gross sample to the laboratory sample is an operation which
must be performed with accuracy and precision .. Automatic machinery and
labor and time saving devices may be used for this operation adhering closely
to the following general scheme with the exception that some materials of
uniform comparison need not be finally reduced as small as 6 mm.-4 meshwhile it may be necessary to finally reduce other material to .14 mm.-IOO
mesh-or finer.
In'sampling large shipments, the gross sample may become excessively
large and unwieldy. This may be obviated by periodically reducing the gross
sample by the Standard Long Pile and Alternate Shovel Procedure, and subsequently compositing these reduced portions for the final gross sample.
REDUCTION OF THE GROSS SAMPLE'
(Approx. 1000 lbs.) crushed to about 22.6 mm.-I mesh; thoroughly mixed
and halved by the Long Pile and Alternate Shovel procedure.
Discard

72
72
72
72

One-half crushed to about 16 mm.-1.5 mesh; mixed and halved by

the Long Pile and Alternate Shovel procedure.
One-half crushed to about 11.3 mm.-2 mesh; mixed and halved by
the Long Pile and Alternate Shovel procedure.
One-half crushed to about 4 mm.-5 mesh; mixed by rolling on canvas; halved by Cone and Quartering procedure.
One-half ground to about 2 mm.-IO mesh; mixed by rolling on
canvas; halyed by riffling (or sampling machine) or by the quartering procedure.
One-half mixed and riffled down to about 8 lbs.

The entire eight pounds is then further reduced and, the required sample
taken according to the character of the particular material.
A portion for the determination of moisture may be taken at some point
during the redu~tion according to the character of the sample and manner
in which the moisture test is made. When a moisture sam pie is wanted it is
,necessary that the sampling units be collected and stored under conditions
whereby moisture would not be lost or absorbed. The'mixing, crushing and
halving operations, preceding the taking of the sample for moisture should
be accomplished as rapidly as possible.
In case the gross sample contains an excessive amount of water which would
be lost in the subsequent operations necessary to reduce it to the laboratory

1308

STANDARD METHODS OF: SAMPLING

sample, it should be kept in a tight container during the collection of the

various increments, after which it is weighed lmd allowed to. air-dry, or it
may be dried artificially; and reweighed. It is then in a condition to
reduced to the laboratory sample according to tl)e above scheme, taking into
account, of course, the moisture thus driven off from the origInal gross sample
in the final calculation for the evaluation of the material.

be

HEDUCTION OF GROSS SAMPLE (LONG VILE AND ALTERNATE

SHOVEL)
This procedure is based upon the method recommended and used by the
U. S. Bureau of Mines.

--- H

~Ii-

-=-

..

~-.~~~
FIG. 152.

Method of Sampling Ore.

STANDARD METHODS OF SAl\IPLI NG

1309

1. Th oroughl y mix thc entirc grol'!'; sample, after crushing in a su itable

manner until a ll pieces are a pproximately onc inch or [cs;; ill any dim cnsion,
by shoveling it into a cone-shaped pile, depositing each shovelful on thc apex
of the pil e (Fig. 152, A and B ).
2. Shovel all of the material into a pil e a bout th e width of thc shovcl an d
approximately 10 feet long acco rding to th e total amo un t of material, s preading
each shovelful uniformly oyer the whole length of the pile, beginning altorna ~ely from opposite ends (Fig. 152, C).
3. Divide the long pile into two equal portions by beginning on one side
of the pile, at either end, and take successive shovelful s, adva ncing eac h ti lll e
by the width of the shovel around the pil e, combining the fir;;t and eycry
alternate shovelful into a neat cone-shaped pile as in X o. 1 and discarding the
second and every alternate shovelful (Fig. 152, D and E).
4. Repeat the aboye operation on the reserwd por tion until a portion of
about 500 Ibs. is obtained.
5. Crush the material until no piece is greater than approximately thl'eefou rths inch in any dimension, a nd then repeat the above operation (X o. 3) .
6. Crush the ma terial until no piece is larger than app roximately one-half
inch in any dimension and then repeat the above operation (X o. 3).
7. A portion of abo ut 125 Ibs. is no,, obtained which is further reduced
by the" Cone a nd Quartering Procedure."
REDUCTIOX OF GROSS SAl\IPLE (COXE AND QUARTERIXG )
1. Mix the materi al amounting to abo ut 125 Ibs. by crushing to pass it
4 mm.-5 mes h-sieye and shoyeling into a neat cone (Fig. 1:52, P and G).
2. Flatten th e co ne by prel:;sing the apex vertically c!o\yn \yith the shoyel or
board so that, when quartered, each quarter \yill con tain the materia l originally
contained therein.
3. Divide the fl attened pile into equal quarters by passin g a straight edge
board vertically t,,ice through ce nter of th e pile at right angles, each time
drawing half of th e pi le a few inches to
one side (Fig. 152, J[ a nd J ) .
4. Discard two opposite quarters and
brush away all fine particles from the
exposed surface (Fig . 152, J ).
5. C ru sh the remaining quarters to
pass a 2 mm.-10 mes h-sieve. Mix
thoro ughly by shoveling into a neat co ne.
6. Repeat operations Xo. 3 and 4 01'
riffle to about 20 pounds.
7. Grind the retained sample to pass
a .85 mm.-20 mesh-sieve. Mix the
material thorough ly on a clean canvas FIG. 153. 9 Jail" and Toggle Crusher for
by raising alternate opposite corners of
Coarse and l\Iedium Eine C:rincling.
the cam'as, th us rolling the material
from sid e to sid e one hundred times (Fig. 152, ]( and L ).
8. Treat the sample further as required by the particular material being
sampled.
9 Court esy of Sturtevant Mill Co.

1310

STANDARD METHODS OF SAMPLING

It is, of course, impossible to reduce hygroscopic or deliquescent gross

samples to the laboratory sample according to the above scheme unless they
can be dried or otherwise transformed into a stable condition. When it is
necessary to reduce such material it should be done rapidly and by hand,
under the most favorable atmospheric conditions in regard to humidity, etc.
When a large sample is taken for analysis the extent of the reduction of the
sample will depend upon the size of the sample taken for analysis.
APPARATUS FOR REDUCTION AND PREPARATION OF THE
SAMPLE
Any sampling system which does not properly control the ratio of the size
of largest particle to the size of sample can not be depended upon to produce a
representative sample. The proper ratio for almost all material will be obtained by following the procedure outlined under the heading" Procedure for
Reducing the Gross Sample." This reduction process is greatly facilitated
by the use of suitable crushers, grinders, ritHe samplers and mixing devices.
The following examples of such apparatus are therefore given as an aid to the
designing and equipping of the sampling room, the size and capacity of which
will be determined by the number, character, and size of the gross samples to
be handled.

FIG. 154. 10

Crusher and Grinder in Cross Section.

Crushers.-A jaw and toggle crusher is shown in Fig. 153. This is a satisfactory crusher for coarse, medium and fine crushing and permits a considerable range of adaptability to suit varying conditions and requirements.
Simple adjustments allow the capacity to be varied from about 200 pounds
per hour for the production of fine material to 700 pounds per hour for coarser
material whim considering only one of the several sizes of this type of crusher.
The parts are readily accessible for cleaning and repairing. By means 0, this
type of crusher it is possible to reduce the material to about 4 mesh.
10

By courtesy of Sturtevant Mill Co.

STANDARD METHODS OF SAMPLING

FIG. 155.11

FIG. 156.
11

Roll Grinder for Brittle Material.

Roll Grinder with Case Lifted to Show Interior.

By co urttiY of Sturtevant Mill Co.

1311

1312

STANDARD METHODS OF SAMPLING

Disc Grinder Ready to Run.

FIG. 157.1 2

FIG. 158. 12

Disc Grinder Open for Clea"ning.

FIG. 159. 13
12
13

Cruo;her Open for Cleaning.

By courtesy of Sturtevant Mill Co.

By courtesy of Sturtevant Mill Co.

,/

STANDARD METHODS OF SAMPLING

1313

Grinders.-Grinders are necessary for reducing the material finer than

4 mesh and may be of the roll, disc or coffee mill type, according to the character of the material.
. Roll grinders are especially adapted for brittle material. Figures 155 and
156 illustrate a type of roll grinder which may be readily cleaned and has a
capacity of 100 to 1,000 Ibs. per hour when producing a reduction to a fineness
of 60 mesh and 2 mesh respectively. Disc grinders have a somewhat greater
range of adaptability than the roll grinder and will reduce brittle or tough
material to a high degree of fineness. The grinding is accomplished between
a stationary and a revolving steel or iron disc, the output being small and fine
or large and coarse according to the adjustment. Figures 157 and 158 illustrate a satisfactory type of disc grinder.

FIG. 160. Combination Crusher and Sampler.

The coffee mill type of grinder is adaptable to soft or tough materials

which have a tendency to stick to the rolls of the roll grinders or to clog the
discs of some types of disc grinders. Many kinds of coffee mill type grinders
have been developed from which selection can be made according to the
character of material under consideration. Figures 155 and 157 illustrate very
good coffee mill types. When reducing material to extreme fineness the ball
mill, special small laboratory grinders, the bucking board and mortar and
pestle are always applicable for small samples.
Another type of crusher and grinder which has been developed particularly
for sampling coal is shown by Fig. 154. This crusher will reduce the material
and deliver an accurate ten per cent sample in one operation. By repeating
the operation on the aliquot the sample may be further reduced.
For many materials this crusher would serve to entirely reduce the gross
sample to the laboratory sample with a very small amount of labor.

1314

STANDARD METHODS

OF SAMPLING

MIXING AND DIVIDING SAMPLES

'1

Large samples (500 lbs. or more) are llsually mixed and divided by the
" Long Pile and Alternate Shovel '1 procedure \}r (500--;100 lbs.) by th~ "Cone ;
and Quartering" procedure.
, ,
'
Small samples (less than 100 lbs.) are usually mixed by rolling on a clean
canvas, whereby the particles are given a rolling motion in alt.ernately opposite
directions, the dividing being done by quartering 'or, more accurately and
rapidly, by passing it through some form of a(fi~e. A convenient and serviceable type of riffle is illustrated by Fig. 161. Another and more sirriple type is
shown by Fig. 162.
'

"

t.
SC; - -----------.f5u
iCSlot

-- .

- ------:t--" a "

RIFFLE SAMPLER

FIG. 161.

FIG. 162.

FIG. 161. Riffle Sampler. "a," cover to fit tightly over one of the boxes with
slot in rounded tops-also fitted with two lugs ~hich fit in ends of riffle and help to
f
rock evenly.
,
FIG. 162. The cut shows the apron of the cylinder of this machine open for receiving the sample. After the sample is poured in, the apron's position is shifted, a
being moved to a' . The cylinder is then revolved counterclockwise. The closed sides
of the riffles plow through and thoroughly mix the sample, and no sample can be discharged through the riffle while the cylind~r is revolving in this direction if the level of
the sample is below the axis of the cylinder. Aftel'tbe sample has been mixed the
cylinder is rotated one turn clockwise; the sample in'the cylinder is then cut by the
planes of the riffle and half of it is discharged into the receiving tray. The sample
remaining is again mixed by revolving the cylinder counterclockwise. By alternately
changing the direction, the sample is alternately mixed and halved until twice the'size
of sample required remains in the cylinder~ The tray is then emptied o{ the discarded
material, and the sample remaining, mixed and halved, and the sample caught in the
tray is bottled' and becomes the official sampl~, while the sample remaining in the cylinder is bottle~ and held as the reserve sample. The interior of the machine is easy of
access and should be brushed clean after each sampling operation.

THE USE OF STANDARD ,SIEVES

,/-"

This section is based upon the specificjrtions.issued by the U. S. Bureau of

Standards (1912) and adopted by the-Amero lnst. of Mining Engineers (1917).

STANDARD METHODS OF SAMPLING

1315

The essential quantity in the definition of a sieve of a particular size is,

primarily, the size of the openings through which the particles to be sifted
are to pass.
A series of sieves of different sizes should have the openings vary uniformly, according to the size of the particles,desired to separate. Inasmuch
as most of the grinding of substances to fine particles is chiefly with the object of increasing the surface upon which some chemical action is to take place,
the opening of such a series of sieves should vary in such a way that the square
or fourth powers of the width of the openings shall form a geometrical series.
/"

Designation of
Sieve, l\1etric
System

Customary
Usage

Opening

Mm.

mm.
128
90.5 mm.
mm.
64
45.3 mm.
32.0 mm.
22.6 mm.
16.0 mm.
11.3 mm.
8.0 mm.
5.66 mm.
4.00 mm.
2.83 mm.
2.00 mm.
1.41 mm.
1.00 mm.
.85 mm.
.71 mm.
.59 mm.
.5 mm.
.42. mm.
.36 min.
.29 mm.
.25 mm.
.21 mm.
.17 mm.
.14 mm.
.125 mm.
.105mm.
.088mm.
.074 mm.
.062 mm.
.052 mm.
.044 mm.

1 mesh
2 mesh
5 mesh
10 mesh
20 mesh
30 mesh
40 mesh
50 mesh
60 mesh
70 mesh
80 mesh
100 mesh
120 mesh
150 mesh
170 mesh
200 mesh
250 mesh
280 mesh
325 mesh

Inch

128.0
5.04
90.5
3.56
2.52
64.0
45.3
1.78
32.0
1.26
22.6
0.891
0,630
16.0
11.3
0.445
8.0
0.315
0.223
5.66
4.00
0.157
2.83
0.111
2.00
.0.079
1.41
0.0555
1.00
0.0394
0.85
0.0335
0.0280
0.71
0.59
0.0232
0.50
0.0197
0.42
0.0165
0.36
0.0142
0.29
0.0114
0.25
0.0098
0.21
0.0083
0.17
0.0067
0.14
0.0055
0.125
0.0049
0.105
0.0041
0.088
0.0035
0.074
0.0029
0.062
0.0024
0.052
0.0021
0.044 0.0017

Number of Mesh per

Linear
Centimeter

1.0
1.4
2.0
2.75
3.9
5.0
7.0
8.0
9.0
10.0
12.0
14.0
16.0
20.0
23.0
27.0
31.0
39.0
47.0
59.0
67.0
79.0
98.0
110.0
127.0

Inch

2.54
3.56
5.10
7.00
9.9
12.7
17.8
20.3
22.9
25.4
30.5
35.6
40.6
50.8
58.4
68.6
78.7
99.1
119.4
149.9
170.2
200.7
248.9
279.4
323.0

Diameter of Wire

Mm.

9.5
9.5
6.4
5.26
4.85
4.11
3.05
2.67
2.00
1.48
1.00
0.81
0.56
0.59
0.43
0.40

OAO

0.41
0.33
0.29
0.26
0.21
0.185
0.16
0.15
0.116
0.089
0.064
0.061
0.053
0.040
0.039
0.035

Inch

0.375
0.375
0.25
0.207
0.192
0.162
0.120
0.105
0.079
0.058
0.039
0.032
0.022
0.0232
0.0169
0.0157
0.0157
0.0161
0.0130
0.0114
0.0102
0.0083
0.0073
0.0063
0.0059
0.0046
0.0035
0.0025
0.0024
0.0021
0.0016
0.0015
0.0014

While the sampling process does. not require extremely accurate sieves
it is advisable to use those which conform approximately to a standard specification. Consequently whenever reference is made to a sieve it will be understood to comply with the following specifications, which have been adopted
by a conference of representatives of various scientific and technical societies,
government bureaus, and private fir-ms, held at the U. S. Bureau of Standards.

1316

STANDARD METHODS OF SAMPLING

This sieve scale is essentially metric. The siE)Ve~having an opening of 1 mm:

is the basic one and the sieves above and belo:w this in the series are related
to it by using in general the square root of
2 (1.4142), 0):' the fourth r()()t of 2 (1.1892),
as the ratio of the width of one opelling to
the next sllla,lle~ opening. The first ratio i~
used for openings between 1 mm. and 128
mm. while the fourth root of 2 is used as the
ratio for opehings below l.mm. to give more
Outlet for
sieves in that part of the scale.
Burned
Gases

DRYING OVEN

CONTAINERS FOR SHIPMENT

TO LABORATORY' .

Samples in which the moisture content

FIG. 163. Dryer for Coarse Sam'pies. The outlet for air at the top is important should always be collected and
may be connected with a 'chimney
or any other device which will fur- . shipped in moisture-tight containers. A
nish a suitable draft. The sample galvanized iron or tin can with a screw tor
is spread on tared pans, weighed, wnich is seared with' a, rubber gasket and
and dried at 100 to 150 C. above adhesive tape is best adapted to this purroom temperature, and weighed pose. Glass fruit-jars sealed with rubber
again. The drying should be continued until the loss in weight is not gaskets may be used, but require very care'ful packing to avoid breakage in transit.
more than 0.1 per cent per hour.

J-Ieating units

FIG. 164.
FIGS. 164 and 165.

FIG. 165.

Drying oven, triple walled, with a layer of air entirely sur

rounding the inner chamber, the temperature of! this layer being practically the samE
as that of the air in the chamber. This approaches an adiabatic construction whict
has been approved by the Bureau of Standards as the best type of design for calorimeter!
and other constant temperature devices. By this construction together with a heav)
outer insulation, heat exchanges between the inner chamber and the room outside are
practically entirely prevented, and as a result of uniformity of temperature is secured
in the heated space. The drying chamber sh()uld be entirely shielded from direct
,~'1diation from the heating units, and into eVCfOY'JJart of which heat should be carried
by convection current~.

STANDARD METHODS OF SAMPLING

1317

Samples in which the moisture content is of no importance need no specid

protection from loss of moisture.

DRYING OVENS FOR MOISTURE

DETERMINATION
Samples may be dried ?y means of a forced
draft of air at a slightly elevated temperature,
or by heating at 100-1100 C. in Ii, well-ventilated oven until a constant weight is obtained.
The former procedure is particularly adapted to
material which has a tendency to undergo undesirable chemical or physical changes upon
prolonged heating at elevated temperatures or
for rapidly removing extraneous moisture preparatory to grinding. The latter procedure is
us.ually employed for quantitative determinations bn small finely ground samples. In either
case a good system of ventilation and a definite
uniform temperature are the chief considerations.

FIG. 166.14 '.

High Temperat1.lre O;ens"for
operation from room temperature to 260 0 C.

FIG. 167. 15 Vacuum oven. Supplied with a vacuum chamber which may be easily
removed, permitting the oven proper to be used as a regular drying oven when the
vacuum feature is not desired. Moisture test of flour can be made in five hours at
75 0 C., while in an ordinary oven at 100 C. from ten to 'twelve hours are required.
The same conditions apply for organic tests in general, i.e., for fertilizers where the tests
must be made at the. lowest possible temperature to avoid driving off nitrogen compounds; for powder where there is great danger of igniting the sample; for milk where
there is great danger of charring; foi' operations involving essential oils, etc..
'Vhen tests are to be conducted in an atmosphere of hydrogen, nitrogen, carbon
dioxide or other gas, it is only necessary to pass a current of the desired gas through
the vacuum chamber.
14 By courtesy of Arthur Thomas Co.
15 By courtesy of Central Scientific Co.

1318

STANDARD METHODS O;E SAMPLING

Figllre 163 illustrates an oven for- drying)arge samples by means of a cur-.

rent of heated air. Figures 164 and 165 show!.a type 'of oven with a good
ventilating system. FigurE) 166 shows another type qf oven, while 167 illustrates a vacuum sirying oven which may be nece~sary for special work.
.

REFERENCES
"Method of Least Squares Applied to Estimating Errors in Coal Analysis." Technical
Paper 171, U. S. Dept. of the Interior..
. .
Bailey, E. G., "Accuracy in Sampling Coal." J. Ind. Eng. Chem., 1, 161 (1909).
,
Weld, F. C., "Accuracy in Sampling." J. Ind. Eng. Chem., 2, 426 (1910).
Camp, J. M., "The Methods of the U. S. Steel CorporationJor the Commercial Sampling
and Analysis of Iron Ores." J. Ind. Eng. Chern., 1, 107 (1909). Professional Paper
No. 48, U. S. Geo!. Survey.
"
Pope, Geo. S., "Sampling Coal Deliveries." Bull. No. 63 and Technical Paper No. 133,
U. S. Bur. of Mines. "The Sampling of Coal," Amer. Soc. for Testing Materials,
.
Standards (1918), 673, 679.
.
Wolf, A. G., "Sampling of Mine Floors." Eng. Min. J., 106, 537,671 (1918).
.
Daman, "An Automatic Ore Sampler." Eng. Min. J., 103, 188 (1917).
Maxwell, "Mill Check on Mine Sampling." Eng. Min. J., 104,884 (1917).
Herr, "Observations on Sampling." Eng. Min. J., 102, 1015 (1916).
Pearce, "Ore Sampling of a Custom Mill." Eng. Min. J., 104,66 (1917). "Preparation ..
of Laboratory Samples." J. Ind. Eng. Chem., 9, 100 (1917).
Circular No. 39, U. S. Bur. of Standards ..
"Standard Screen Scale for Testing Sieyes." BulL.Amer" lnst. of Mining Eng., June,
1917.
" . .
Kerr, C. H., "Standard Testing Sieves." Trans. Am. Ceram. Soc., 15,375 (1913).
.
Hoover, T. J., "Standard Series of Scre.ens for Laboratory Testing." Eng. Min. J.,
90,27 (1910).
Fairlie, A. M., "Standard Sieves and Screens." Metal Ind., 12, 113 (1914).
"Mechanical Preparations of Solids for Chemical Treatment." Chem. Eng.! 9, 171.

GENE~

OUTLINE FOR

SAMPLI~G

LIQUIDS

In general, a uniform and representative sample of liquid material is a

feasible possibility because of the usual absence of segregated constituents,
the ease with which the separate components may be intimately mixed, and
the degree of comminution to which a liquid may be subjected without producing a separation of the individual components.
When the liquid is thin, non-viscous, and does not contain immiscible
constituents, a homogeneous condition usually exists, and a sample from any
part of the bulk is sufficiently representative of the whole. When viscous or
immiscible materials are present, a heterogeneous condition usually exists and
extreme care must be exercised to obtain a thoroughly representative sample.
The proper selection of a representative sample of a liquid therefore, involves
a consideration of the physical laws of liquids, their chemical activity, the

STANDARD lVlETH?DS OF SAMPLING

1319

iniscibility of all components, and the interference df any insOluble materials

carried by the liquid.
The sampling of liquids is qivided into the three following classes; samplin,g
liquids in quiescent state, sampling liquids in motion, sampling special liquors,
e.g., immiscible liquids, volatile liquids, liquids carryitlg sediment, etc.

SAMPLING LIQUIDS IN QUIESCENT STATE

All liquids not in motion may be sampled by obtaining portions at specified

points from the top to the bottom of the liquid by means of a suitable sampling
apparatus ..
1. FROM STORAGE TANKS, TANK CARS, BOATS, EVAPORATING KETTLES, VATS,
CRYSTALLIZERS, MIXERS, SETTLER:;;, ETC.

(a) If the liquid has been thoroughly agitated or is known to be reasonably

uniform in composition, transfer several portions to a container by means of
a suitable dipper, the several dippings being taken pl:omiscuously throughout
tbe mass of liquor.
---Glass NOZZle

"Au

---'?-

Pinch Cock

"

I~
I'
II
II

I:

.if W/'Plpe __ >11

II

Lenqfh up to
/01.0"

.j'Thick

-:r-:;~:::=:;::;t==;~=
I
I
---;r
I

Wrought Iron

Dlsc

I
I

/V

----:-4 61ass Tubing

I
I

I
I
I
I

~
I
I

-<-- -

--_1. Glass Boffle

: 4 "Oiam. -8 "High
I

; ...

,f~Bu

,!,,'V~~~=~~_Y.
------5"-----J<'IG. 168.

Sampling Liquids Appar4tus.

1320

STANDARD METHODS OF SAMPLING

(b) If 'the character and conditiQn of th-e liquid is such as too permit the
formation of zones of different composition, obtain Isamples by means of a
suitable apparatus as shown by Fig. 168,' whereby PJoper proportions, of the
liquid are taken at every point from the top to the \:lOttom. Comb!ne these
samples for the gross sample and, after thoroughly mixing, 'bottle tl;le required
samples (Fig. 168). By slowly lowering the samplirig bottle into the liquid
with the outlet (A) open, a portion of the liquid from every point passes through
(B) into the bottle, whereas if outlet ,(A) is kept closed untiIthe desired depth
is reached, the entire sample may be taken at any desired point.
Lead, wood, or any other material which is resistant to the liquid may b~
substituted for the iron parts of this device, according to the character of the
liquid being sampled. Under some conditions it will be necessary to substitute
a ground glass stopper for the rubber one. Openings of different sizes can be
attached at (A) to govern the rate at which the sample will flow into the bottIe.
(c) Take a sample from the upper, middle, and lower levels of t~e liquid by means
"
MGlas
of a sampling device similar to that shown
Rod
by Fig. 168, by keeping the outlet tube A
closed until the container has been lowered
Made of }(Pyrex Tubing
to the desired depth.
Length
30"
48"
8/J."

2.

FROM DRUMS, CARBOYS, BARRELS, ETC.

(a) By means of a thief (Fig, 169) take

proportional fractions from each container.
'Combine and mix these increments for the
composite gross sample. Bottle the amount
GrOUnd,$(
required for the sample. If the contents
Glass
Stopper
of the container are such as to preclude
thorough mixing or if the material has a
Rubber Stopper
tendency to form strata of varying compoFIG. 169. Sampling Liquids
sitions,
the open thief should be lowered
Apparatuf3.
into the liquid at such a rate as to keep
the levels of liquid inside and outside of the thief very nearly equal, in order
to include portions at all points from,the top to th~ bottom of the liquid.

SAMPLING LIQUIDS IN MOTION,3.

FROM A PIPE OR CONDUIT CARRYING A CONTINUOUS FULL FLOW UNDER "

A CONSTANT OR VARYING' HEAD

(a) A continuoUf; or intermittent sample may be taken by inserting a

small pipe into the line on the discharge side df the pump or propelling force
of the flow. This sampling tube should extend one-half of the distance to
the center of the flow, witli the inner open end turned at an angle of 90 and
facing the flow of the liquid. The sampling tube may be supplied with a
stopcock in order to regulate the flow into a receiver; The amount drawn
dUl'ihg th.e period of taking a continuous san.wl constitutes the gross sample.
In this instance the stopcock should be adfusted at the beginning of the sampling p3riod, to give the desired flow and no readjustment made during the

"

STANDARD METHODS OF SAMPLING

1321

sampling period. Intermittent samples may be taken by opening the stopcock for a definite~period of time at definite intervals. The combined intermittent samples then constitute the gross sample. '
)
4.

FROM A PIPE OR CONDUIT CARRYING A VARYING FRACTION OF FLOW

. (a) By means of a sampling weir similar to the apparatus shown by Fig.

171, a definite fraction of varying flow can be collected qver a required period

p
~

One size larger than

pipe above the elbow

FIG. 170.

Apparatus for Sampling Liquids in Motion.

of time. This gross sample may be mixed and a portion taken for the final
sample or by connecting two or more of the above devices in series; a small
representative sample may be collected automatically, e.g., if one million gallons of liquid per day are passing through a pipe or conduit line at an irregular
rate, it could be sampled by conq.ucting the stream to a series of four sampling
weirs, each of which separates 5% or 1/20 of the flow. A final six gallon
sample would then represent the entire million gallons. The size of each weir
should be such as to give a suitable overflow and the apron over which the
flow passes for division should be fixed in a horizontal position.
(b) Bl means of the arrangement illustrated by Fig. 170, an approximate
sample ratio can be continuously diverted, from a non-uniform flow. This
device consists ,of a pipe A, inserted into the flowing stream at an angle of
70-80 0 provided with an opening B facing the flow and extending to the top
of pipe P. This opening B is constructed as shown in the cut (Fig. 170), the
width above the center of pipe P being equal to the diameter of pipe A, then
tapering from the center to the bottom of the pipe P. As the flow passes pipe
A, an approximately constant sample ratio is diverted through pipe A and is
s:1mpled by a small pipe C, the opening of which faces the flow at a point
halfway between the center and bottom of pipe A. Pipe C is fitted with a
stopcock whereby the amount of the sample flow through C may be regulated.
Pipe A reenters pipe P at some point which will give sufficient positive flowhead through A. While this arrangement w~ll not necessarily produce an
. accurate sample ratio, an increased flow through P will increase the flow-head
through A and consequently increase the amount of sample withdrawn at C
and an approximately relative sample ratio will be maintained with one regulation of stopcock C:
(C) With large pipes and large flows, the above arrangement may be modified as shown by Fig. 172. Three pipes, 1, 2, and 3, bent at 90~ angles with
openings facing the flow at different levels are in~erted in pipe P and connected

1322

STA]'iJ"DARD METHODS OF; SAMPLING

to a common pipe;. The flow through A is then sampled by a pipe C as

described above. An approximately relative sahlple ratio is thus maintained
with one regulation of stopcock C.

fIG, 171. Apparatus for Sampling Liquias,

FIG, 172,

Apparatus for

I
Sampli~Liquids
...1

in Motion,

,>

STANDARD METHODS OF SAMPLING

1323

SAMPLING SPECIAL LIQUIDs'

5.

LIQUIDS CONTAINING INSOLUBLE .MATTER IN SUSPENSION

(a) When the sediment remains in. reasonably uniform suspension d.uring
the period required Tor taking the sample, the liquid is thoroughly mixed and
the sample oQtained by one of the foregoing procedures.
(b) When the sediment settles rapidly or is impossible of uniform distribution it is necessary to take special precautions to secure the components of
the sample in the proper proportion. This can be accomplished with a fair
degree of accuracy by mixing the liquid thoroughly and then rapidly taking
a complete column of the liquid by means of the special thief or device shown
by Fig. 168 or 169. By repeating this operation a number of times, a fairly'
representative sample will be obtained. Another procedure of possible application under certain conditions would be to take several portions at as many
uniformly distributed points from the top to the bottom by means of the thief
(Fig. 168 or 169). These portions are then combined and mixed for the gross
slmwle.
(c) To accurately sample a liquid containing insoluble matter, it is necessary
to filter the entire liquid, weigh the insoluble material and filtrate separately
and then take separate samples of the sediment and liquid, recording the
ratio of sediment to liquid for use in the final evaluation of the original material. This, however, is impracticable except in such instances wherein the size
of sample allows of economical handling in the manner described.

6.

LIQUIDS COM~OSED OF hUqSCIBLE LAYERS

The most practical and accurate procedure in this case is to effect a separation of the layers and combine portions of each layer in the proper ratios. If
this is impossible, the material may be transferred to a container of. uniform
cross section, the depth of each layer measured, and then the proper portion
taken from each layer, for a composite sample, by means of a sampling thief
similar to those shown in Figs. 168 and 169.

7.

LIQUIDS CONTAINING CRYSTALS DUE TO SUPERSATURATION

Frequently conditions are such that a portion of the liquid or certain

constituents thereof have crystallized in the container. Oleum and acetic
acid are common examples of this class. When practical the material should
be warmed sufficiently to dissolve completely or melt the crystals before
sampling by one of the foregoing procedures. Otherwise, if the crystals are
small and remain in suspension the mass should be sampled as a liquid containing insoluble matter in suspension.
8.

VOLATILE LIQUTDS

Ca) Volatile liquids are best sampled from

a, continuous flow while the

liquid is being discharged or by means of a siphon from containers. If the
liquid is uniform in comppsition or can be made so by agitation in some manner,
a portion may be run to the bottom of the sampling bottle until the bottle
overflows through another tube from the top into another bottle or to the

132.4

STANDARl? METHODS OF $AMPLING

waste pipe, to such an extent as to leave it liquid if the first bottle containing
the maximum amount of the volatile constituent (see Fig. 173). 'Fhe sample
bottle should then be quickly stoppered and sealed for transportation to the
laboratory.

FIG.

17~.

Apparatus for Sampling Liquids.

REFERENCES
"Sampling Creosote Oil." Amer. Soc. /Testing Materials (1918), p. 721.
"Liquid Sampling Specifications." Am. Gas Light J., 103,391.
"Standard Methods for the Sampling and Analysis of Commercial Fats and Oils."
J. Ind. Eng. Chern., 11, 116 (1919).

GENERAL OUTLINE FOR SAMPLING GASES

The proper sampling of gases often presents greater difficulties than the
analysis. This is particularly true when it is desirable to obtain a representa~
tive salJlple of a gas during a considerable period of time'asjt passes through a
pipe or flue from the source of production. Gases tra_vel through straight
pipes and conduits in an irregular succession of waves with a spiral motion,
the maximum veJocity being at. the center of the pipe and the minimum near
the walls. The motion of these waves and point of maximum velocity.is
altered in a varying degree by every bend and oblltruction throughout the line
of travel. Where gases are encountered \vhich are subject to temperature
changes either above or below normal, this factor will also considerably interfere with proper sampling. The temperature of gases will vary throughout
the cross-section of the pipe, usually being hotte~t at points of maximum
velodty and coldest next to .the walls an<!-m dead pockets. Theoretically if it
were possible to discharge the total gas' to be sampled into a large holder and
proVjde time for it to thoroughly diffuse, it wOllld then be in the most ad-

STANDARD METHODS OF SAMPLING

1325

vantageous condition whereby a true and representative sample could be

obtained by simply withdrawing a portion. Unfortunately this is not often
the condition under which ga~es are to be sampled, being usually sampled from
a flow possessing varying velocity, temperature, and composition. Under such
conditions it is practically impossible to determine a point of approximate
average velocity, temperature or composition from which a representative
instantaneous sample can be crawn.

:" Melal Tube,

, LeadorA/~Cu
Alloy

, -- Portland
Cement
Threads iTmmnrrni',-L
10 Attach
loSlack-"""

61a55 Tube
fa Sample
ConTaIner..'

FIG. 174.

Apparatus for Sampling Gas.

Consequently it becomes necessary to collect a sample over an extended

period whereby the number of series of gas waves sampled is such that there
is a high. probability that a true anq. representative sample will be obtained.
Under some 90nditions, useful and more definite data regarding composition
and gas-flow is obtained by taking frequent" grab" samples. It is, therefore,
quite evident that the subject of gas sampling naturally divides itself into
two phases for consideration that of continuous sampling and" grab," or instantaneous, sampling.

1.

GENERAL CONSIDERATIONS AND ApPARATUS

(a) Sampling-tube for Drawing Samples from Gases Flowing through

Pipes.-Without discussing the merits of the various forms of sampling tubes
such as the single-opening tube, and the various types of perforated tubes all
of which are more or less useful under certain conditions, the multiple type
originated by A. H. White-Ii Gas and Fuel Analysis "-is apparently best
adapted for securing uniform samples (see Fig. 174). This tube should be
inserted so that the longest tube reaches the center of the flowing gas.
(b) Collecting a Representative Continuous Sample.-In order that the
sample should be truly representative of the total gas flowing during a definite
period, it is necessary to constantly draw a definite proportion of the gas.

1326

STANDARD METHODS OF SAMPLING

Great elaboration of apparatus and equipment would, be noocSSBIry if the rate

of drawing the sample were adjusted to vary directly with the velocity of the
gas, in which case the sample would accurately repre~ent the gas. The most '
practical approach to this ideal condition would be Ito take the continuous
sample under a constant pressure, whereby the rate of" sampling will vary
directly with the pressure of the gas being sampled. f
Figure 175 illustrates a common form of apparatus whereby a stre;lm of
gas at constant pressure is drawn from a gas line. This apparatus 'consists
of a pressure-gage B to indicate any obstruction of the sampling tubes, a
bubbling-bottle C to give a visual control of the rate of -the stream, a gas:From Sampling Tube

To Aspirator

FIG. 175.

Apparatus for Sampling Gas.

meter D which may be omitted if not needed, a pressure control, E and a

pressure gage F on the line to the aspirator. When aspiration is necessary the
pressure at the sampling-tube as shown by gage B should be only a few tenthsof-an-inch of water. This may be regulated by the depth to which the tube
is immersed in the water of the regulator E. When the gas is under pressure,
the aspirator and E and F become urlllecessary.
(c) The Design of Apparatus.-The design of apparatus-and the sampling
procedures-must take into account the solubility and chemical activity of
the gas. Saturated magnesium chloride solution, glycerine and water mixtures, and various oils have been used with varyjng success in overcoming
the solubility of gases in the solutions over which they are collected. Further
investigations along these lines for specific gases will probably make possible
the use of other liquids than mercury for gas'es having high solubility factors.
The corrosive character of the individual gases will determine the type of conI
tainers, etc., which may be used.
2.

TAKING A "GRAB" SAMPLE

(a) For a." grab," .or instantaneous, sample the ordinary t,vo-bottle aspira-

tor may be used (see Fig. 176) using care not to draw the sample so fast as to
produce a reduced pressure at gage B (Fig. 175).
3.

TAKING A C.ONTINUOUS; SAMPI,E

(a) For continuous sampling the apparatus illustIT'.ted by Pig. 17i will
be found convenient.
This apparatus consists of two large galvanized metal tanks, one in the
form of an aspirating-bottle in which the s_n'i.ple is taken, and the other,
slightly larger, which acts as a reservoir. A sample of gas taken under const/tnt pr\_'ssure and at a uniform rate over any specified period of time, which

STANDARD METHODS OF SAMPLING

1327

may be varied at will, is obtained by filling the tanks with water, oil or other
liquids ac(wrding to the character of the gas, until the level reaches the top
of the aspirating-tank. A syphon acts
Pinch Cock
as the outlet and a uniform rate, if the
pressure of th~ gas in constant, is obtained by attaching it to a wooden
float. The speed of outflow may be
regulated by the length of the syphon
and the opening or size of the outletnozzle.
White rubber
tubill9

""
",.

",.
"
"

~::;_:t~.::~:

FIG. 176.

Apparatus for Sampling Gas.

FI<}. 177.

Apparatus for Sampling Gas.

By arresting the downward, movement of the float and syphon by means

of a shoulder or suitable device, a small amount of water remains in the aspirator; this acts as a seal and automatically discontinues the intake of gas.
By closing the stopcocks A and B the operation is completed.
These tanks may be constructed to crate, in compact form for transportation, a 20-liter glass aspirating-bottle, as shown in Fig. 178.
This bottle is to be used when the gas to be sampled would attack the
metal aspirating-tank. If a layer of some non-absorbing oil is placed on the
surface of the water the original composition of the gas is unaffected.
(b) When mercury must be used because of the solubility of the gas in
water, an apparatus !'!imilar to that illustrated by Fig. 179, originated by
Thomas Gray, J. Soc. Chern. Ind., 32, 1092, will be found convenient.
" It consists essentially of an ordinary gas-sampling tube N, attached to a
reservoir J, which is counterpoised by a vessel R floating in a tank Q, the float
being connected with [by means of a cord which passes over the pulleys P.
" The float R may be a tinned-iron vessel or glass bottle containing the
necessary quantity of water or lead shot to establish the balance. The height
of the tank Q must be greater than the length of the sampling-tube N, the
overflow-tube I should be slightly longer than. N, and the length of the wider
tube H, which merely serves to prevent loss of mercury by splashing, should
slightly exceed that of I.
" The jet V, under pressure which may be regulated by adjustment of the
overflow W, through which the excess of water from U escapes to the waste,
delivers sufficient water during the specified period to raise the level in the
tank to a height equal to the length of the sampling-tube.

1328

STANDARD METHODS OF 'SAMPLING

" As the water-level rises, the reservoir J is stl:1adily lowered, drawing the
gas uniformly into the sampling tube, and the mercury, thus displaced, escapes'
through the overflow-tube I to the bbttle G. By means of an aspirator a
rapid current of gas is drawn along the tube F-which is connected to the flue
or gas-main by mean~ of the sampling tube (Pig. P9), a suitable filt~l b(jing
interposed, if necessary, to retain any solid' particles. The .conne'ction A,
between the sampling
tube and the tube F,
p
is preferably made of
capillary tubing 3 to
4 inches long, to prevent the diffusion of
u
the gas backwards from
N to F ~ If desired, a
small non-return mercury valve may be in~
serted at this point.

H ,
Q

FIG. 178. Apparatus for

Sampling Gas.

FIG. 179.

Apparatus for Sampling,Oas.

" To start the apparatus, the tube F is connected with the gas supply and
a rapid current of gas is drawn through F by means of the aspirator.
"The rubber tube connecting the jet V 'with the water-su~ply is then
closed by the clip S, and the water is allowed to run into T and to overflow
freely through W the height of which has previously been adjusted to the desired level. Water is now run from the.'tank Q until the top of the overflowtube I is at the lev:el of the shoulder of the gas tube N, and J and N are now
filled with mercury by means of a small funnel in f: ~Th~ capillary A is finally
filled with mercury by raising the reservoir J, the lower stopcock of the sampling-tube is closed, and J is t.hen lowered t.o its former position and filled to
the overflow-tube with mercury; the operation of sampling may now be commenced by opening the lower stopcock of the gas tube N and starting the flow
of water from the jet. The aspirator and/the water-supply to T remain in
operation during the whole period of the test and the apparatus requires no
further attention.
" The period of sampling may be. varied in a number of different ways:
(1) "Jets may be constructed to deliver the requisite quantity of water in
specified times; these are easily drawn from glass t~bing in a blowpipe-flame,
the final adjustment being effected by al~irng the head of water in T.
(2) "Tanks of various diameters may be employed.
(3) "The height of the sampling tube N may be varied."
,

STANDARD METHODS OF SAMPLING

1329

REFERENC;ES
Sinnatt, F. S., "A Convenient Apparatus for Obtaining an Ayerage Sample of Gas and
for Regulating the Flow of a Gas into an Evacuated VesseL" Analyst, 37, 250.
Gray, Thomas, "An Automatic Gas Sampling Apparatus." J. Soc. Chem. Ind., 32,
1092.
'White, "Gas and Fuel Analysis." International Chemical Series.

The author desires to acknowledge his indebtedness to J. S. qoye for his

assistance in preparation of this chapter.

STANDARD SPECIFICATIONS FOR SIEVES FOR TESTING

PURPOSES 16
A. S. T. M.

DESIGNATION:

E 11-26 17

Sieves for precision testing have a variety of uses, including both dry
screening and wet screening of material. The primary object is to separate
materials according to the size of particle. Since particles are three dimensional and the openings of screens are areas having two dimensions only, the
method of separation by means of screens is at best an approximation. In
many instances, however, it is the best practical means at hand, and will give
useful and reliable results if care is taken in securing representative samples,
in the preparation of the sample for testing, in the details of procedure and
in the method of reporting results. In the Appendix are indicated those points
which must be considered in specifying sieving tests for any individual material.
An essential in each method, however, is a series of sieves of known characteristics. The important consideration in sieves is the size of opening and
with the size of opening known it is possible to translate the results obtained
in one series of sieves into terms of another series. The importance of the size
of opening in the sieve series covered by these specifications is emphasized by
designating the sieves in terms of opening in microns. This new scheme of
designation should_accordingly cause no confusion, giving as it does a direct
indication of the size' of opening. As such, it is not open to the objections to
the various arbitrary number designations heretofore in use and it is also free
from the objection to the designation by means of mesh in that it is independent
of the size of wire.
IS Under the standardization procedure of the Society, these specifications are under
the jurisdiction of the A.S.T.M. Committee E-1 on Methods of Testing.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
11 These specifications are issued under the fixed designation Ell; the final number
indicat.es the year of original adoption as standard or, in the case of revision, the year'
of last revision. Issued as Tentative, 1925; Adopted, 1926.
.

1330

STANDARD METHODS OF SAMPLING

With the size of opening known, the methOd of re<:ording results is simplified. It is recommended that in reporting sieve analyses in graphic form
the sizes of sieve opening he arranged as ordinates with the largest size at the
top and that the percentages be plotted as abscissas .
1. Scope.-These specifications cover sieves for use in the usual cases of
precision testing for the fineness of materials (classification of materials according to size). The sieves have been designated according to the size of openings
in terms of microns, the corresponding United States Standard Sieve Series
designation in arbitrary numhers being given for the convenience of those
familiar with the latter scheme of designation.
..
PHYSICAL REQUIREMENTS
2. Cloth and Frame.-\Vire cloth for standard sieves shall be \"oven (not
twilled, except that the cloth of 44, 53 snd 62-micron sieves may be tw;]Jed
until further notice) from .brass, bronze, or other suitable wire and mounted
on the frames without distortion. To prevent the material being sieved from
catching in the joint between the cloth and the frame, the joint shall be smoothly
filled with solder, or so made that the material will not catch. The sieve
frames should be circular, about 20 em. (8 in.) in diameter and about 5 em.
(2 in.) or 2.5 em. (1 iII.) between the top of the frame and the cloth.
X"OTE.-The fjjze specified ahove is for the regular standard test~ng sieve, but for
special purprn:;e.'> ot.})cr t-;izf'S may be used, ItS for example, the 3-in. sieve which is designed
especially for use on laboratory baln,nccs IlS in t.J~e ~esting of pigments.

3. Permissible Variations.-The average opening between the adjacent

warp and the adjacent shoot wires, taken separately 1 shall be that given in
column 3 (or 4) of Table I, within the" Tolerance in Average Opening'" given
in column 7. The average diameter of the warp and of the shoot 'wires, taken
separately, of the cloth o[ any given sieve shan be that gh'en jn column 5 (or 6)
of Table I within the <t Tolerance on 'Vire Diameter" given in columns 8 and 9,
The ma.ximum opening between adjacent parallel wi!cs shall not exceed the
nominal width of opening for that sieve by more than the "Tolerance in
,Maximum Opening 11 given in column 10 of Table I.
WORKlIIA.NSHIP

4. The Sieves sha.n be free from imperfections in the sieve cloth or its
mounting; as for example, punctured, loose or wavy cloth, imperfections in
soldering, etc. ,
MARKlNG

5. The siev,", shall bear a label upon which the manuI8Cturer'. nominal
opening in metric units (preferably in microns) with the equivalent in inches,
together with the size of wire should be given.
NarE.-It h;'re('ommcnded that, until the new mi('run designation shall have been
thoroughly est.abli~bed, in addition the corresponding 'United States Standard Sieve
Scrjc~ number be given,

STANDARD METHODS OF SAMPLING

1331

TABLE I
REQUJREMENTR FOR SIEVE OPENINGS AND 'VIRE DIAMETERS WITH PERMIsSIBLE

VARIATIONS

Designation

4760 micron
4000
"
"
3360
2830
"

2380
2000
1680
1410
\190
1000
840
710
.590
500
420
3050
297
250
210

177
149
125

lO!>
88
74
62
53
44

Correl'lJond ing . S.
Standard
Sieve Series
Number

"
"
"
"
"
"
"
"
"
"
"
"
"
"
"
"
"
"

"
"
"
"
"

"

Sieve

Wire

in.

mm.

in.

Average
Opening,
per cent
7

Tolerance on
Wire Diameter, per cent

Tolerance
in
Maximum
Openin~.

per cent

Under

Over

10

3
3
3
3

15
15
15
IS

30
30
30
30

0.0661 69
0.0555 .61
0.0469 0.54
0.0394 0.48

8.

0.0331
0.0299
0.0272
0.0240
0.0213
0.0189

3
:J
3
3
3
3

IS
15
15
15
15
15

30
'30
30
30
30
30

10
10
10
10
.10
10

0.0331
0.0280
0.0232
0.0197
0.0165
0.0138

0.42
0.37
0.33
0.29
0.25
0.22

0.0165
0.0146
0.0130
0.0114
0.0098
0.0087

5
.5
5
5
5
5

15
15
15
15
15
15

30
30
30
30
30
30

2.5
25
25
25
25
205

0.188
0.162
0.140
0.119
0.102
0.086

0.0074
0.0064
0.005.1)
0.0047
0.0040
0.0034

6
6
6
6
6
6

15
IS
15
15
15

35
35
3.5
35
35
35

40
10
10
40
40
40

0.074
0.063
0.053
0.046
0.041
0.036

0.0029
0.0025
0.0021
0.0018
0.0016
0.0014

8
8
8
8
8
8

15
15
15
15
15
15

35
35
35
35
35
35

60
60
60
90
90
90

6
7

0.187
0.157
0.132
0.111

8
10
12
14
16
18

2.3.~

2.00
1.68
1.4\
1.19
1.00

0.09:17 0.84
0.0787 0.76

O.~~
O.2r:

Tol~ance

10
10
10
10

4.76
4.00
::L36
2.83

[)

D~meter

mm.

20
0.84
25 , 0.71
30
~:.59
35
50
0.42
40
405
0.35

O~ning

50
60
70
80
100
120

0.210
0.177
0.149
0.125

0.0117
0.0098
0.0083
0.0070
0.OOS9
0.0049

140
170
200
230
270
325

0.105
0.086
0.074
0.062
0.0.53
0.044

0.0041
0.0035
0.0029
0.0024
0.0021
0.0017

NOTE.-Sieves having larger

1.27
1.12
1.02
0.92

0.050
0.044
0.040
0.036

openings~than

c::..s

those included in the table may he

specified, but in specifying sieves of larger epenings the actual size of the opening (in
millimeters or in incheS) shall be stated. A tolerance of 3 per cent on average openings nnd of 10 per ce~t in maximum openings shall be permitted.

,\METHODS OF TESTING
6. (a) To determine .whether the cloth conforms to the requirements of
these specifications, the wire diameter shall be ,measured in each direction by
means of a micrometer-microscope or projection apparatus or other suitable
apparatus of requisite precision, care being taken in all cases to avoid diffraction effects in measuring to the edge of a wire. Micrometer calipers and steel

scales should not be used for this purpose.

The mesh shall be determined by

1332

STANDARD METHODS OF'SAl}1PLING

counting the number of openings for a distance of at least 6 in. in both dire~
tions of the sieve, using a scale, either. glass or st~el, and a microscope. The
use of scales with special rulings for obtaining mesh by counting "interference bands" is recommended in cases where lhrge numbers of any of the
sieves with the smaller openings are to be measure1d. - From the wire diameter
and the mesh the average opening'shall be computed.
NOTE.-No general rule can be given for the number of observations to be taken,
except that they be sufficient to cause no change in the result if say 3 or 4 additional
measurements are made.

(b) Measurements of large openings shall be made by means of the micrometer-microscope or the projection lantern or other suitable apparatus in
the same way as wire diameters are measured.
NOTE.-In order to insure the securing of standard -sievesit is possible to obtain
from the manufacturer or dealer sieves which have been certified by the National
Bureau of Standards.

APPENDIX
LIST OF POINTS TO BE CONSIDERED IN DRAWING UP SPECIFICATIONS FOR A
SIEV:ING TEST

I.

ApPARATUS.

1. Enumeration of sieves to be used.

2. Auxiliary apparatus required for collection and reduction of gross
sample, sample containers, and apparatus other than the sieves
used in making the tests:

II.

SAMPLING.

1.
2.
3.
4.

Unit of sampling.
Method of collecting representative gross sample.
Size of gross sample.
"
Cautions to be taken in handling sarri'ple to prevent contamination
or change.
5. Reduction of gross sample.

III.

PREPARATION OF THE SAMPLE FOR TESTING.

1.
2.
3.
4.
5.

IV.

Drying.
Mixing.
Special treatment, if any is necessary.
Selection of test portion.
Size of test portion.

PROCEDURE.

A. Dry Scree,nin(l:
1. Order in which the sieves are to be used.
2. Detailed description of method of shaking.
3. Detailed description of pffiGedure other than method of
shaking.
.
4. End point.

STANDARD METHODS OF SAMPLING

1333

/
B. Wet Screening:
, 1. Liquid to be used.
2. Description of any preliminary separation by classification or
other means.
3. Order in which screens are to be used.
4. Detailed description of procedure including amount and
method of adding liquid, method of stirring or 'other
agitation.
5. End point.
V. METHOD OF REPORTING RESULTS.
VI.

ALLOWABLE
MATERIAL.

VARIATIONS

BETWEEN

DUPLICATE

TESTS

ON

THE

SAllIE

DECOMPOSITION OF THE SAMPLE

Decomposition of the Material.-In the instructions for the solution of

the material special procedures are often given owing 'to the predomiuance of
some particular compound or compounds present. Directions are given for
example for oxides, others for sulfides, and again others for silicates, etc: The
chemist is frequently at loss to know what procedure to select as the mere
inspection of the material frequently will furnish him no information as to its
predominating trait.
.
Owing to the variables in ores and alloys it is not possible to have any set
procedure applicable to all conditions. ' It is often necessary 'to combine two
or more methods to effect solution of certain refractory ores. Acid extractions
are frequently followed by fusion methods for the solution of the acid insoluble
.' material, likewise fusions require' subsequent water extraction followed by
acids. In general, ores should be finely ground, before treatment by wet or
fusion method. The following general rules, however, will be of value.
General Procedure.-It is advisable to test the solubility of a small amount
of the material in a test tube, using water if the material is a salt or hydrochloric acid if an ore, and dilute nitric acid if an alloy or metallic product,
In examining an ore the sample should be finely powdered. About 0.5 to 1.0
gram is taken and treated with 10 to 20 ml. of concentrated hydrochloric acid.
A pear-shaped flask, of pyrex glass, is convenient f~r this decomposition. The
solution is heated gently until no further action is evident. If the ore contains sulfur as sulfide, the addition of nitric acid and bromine are recommended, nitric acid in most cases follows the addition of hydrochloric acid,
but in case of sulfur determination, nitric 'acid and bromine are added first
to avoid loss of sulfur as H 2S. For' alloys dilute. nrtric acid precedes HeI.
Additio~al Hel and HNO a may be necessary to complete the possible action
of these acids. Dilute sulfuric acid is now added and the solution evaporated
to near dryness. This generally effects complete decomposition. If a residue
remains upon adding Hel and water and heating, it is advisable to filter off
the solution and fuse the residue by one of the ,fluxes given under" Fusions,"
page 1336. It is often advisable to start witli a fusion of the original oresodium and potassium carbonate for high silicon ores, sulfates of barium,
strontium and calcium; potassium acid sulfate fusions made preferably in a
pyrex glass flask for oxides of metals. In case of refractory oxides decompaler
to the molten mass.
position is assisted by adding a small piece of_ .filter
/,
I

Original chapter by Wilfred W. Scott.

1334

DECOMPOSITION OF THE SAMPLE

1335

Reagent to
Materials
Effect Solution
I. Water will dissolve the alkali salts, ammonium, lithium, potassium and
sodium acetates, bromides, chlorates, chlorides, iodides, nitrates,
nitrites, sulfates. Also arsenates, arsenites, borates, carbonates,
oxalates, phosphates, and tartrates. The halides of silver, mercury, and lead are insoluble in water. The arsenates, arsenites,
borates, carbonates, oxalates, phosphates and tartrates of the
heavy metals require acid treatment to effect solution. Sulfates
of Ba, Sr, Ca, Hg", Pb are insoluble in water.
II. Acids
1. Hydrochloric acid decomposes carbonate and oxide ores of B, Ca, Fe,
Mn, Mg, P, Sn (reducing agt.), Ti, U, Zn (SnCh assists Fe sol.).
2. Hydrochloric acid together with or'followed by an oxidizing reagent,
i.e., HNO a, KCIO a, etc. Alloys of AI, Sb, Bi, CUi Cr, Co and
Ni steels, ferro silicon, Ti and V steels, sulfide ores of Cu, Ce,
Pb, Mo, U, Zn.
3. Nitric acid, dilute or concentrated. Alloys of Bi, Cd, Co, Cu, Pb,
Mn, silico manganese, ferro manganese, P and Ti steels, ores
of Cd, Cu, Mo, Co and Ni.
4. Aqua Regia.-The metals and alloys of Au, Ir, Hg, V, Pt. Ores of
Cd, Hg, Rh, W.
5. Sulfuric acid.-Ores of AI, Be, Mn, Pb, Th, Ti, U.
6. Sulfuric acid with HF, silicates, Cu ores and alloys.
7. Hydrofluoric acid, usually in conjunction with H 2S04, HCI or HNOa
useful for dissolving ores containing silicates where Si02 is not
to be determined. An HF-HNOa (5: 5: 200 H 20) is an excel.
l!Jnt solvent for antimony-tin alloys.
8. KBr, Br, HNO a, HCl combinations to dissolve pyrites of Pb, Ni, Fe,
As, etc. Reactions 1-8 (excepting 7) may be made in glass;
7 in platinum.
9. Constant-boiling hydriodic acid, sp.gr. 1.70, attacks many of the
slightly soluble substances: HgS with brisk evolution of H 2S;
Sn02 with evolution of SnI4; the hot acid transposes the silver
halides, the majority of the "insoluble" sulfates (Pb; Ca; Sr;
Ba; Cr); dissolves CaF 2 ; CrCl a (anhydrous), etc. 2
10. Perchloric acid is a powerful oxidizing and also a dehydrating agent
when hot and concentrated. The solubility of the majority of
the perchlorates in water makes the acid a very desirable solvent.
The perc hI orates in general are soluble in organic solvents such
as alcohols, ketones, aldehydes, ethers, and esters. Anhydrous
perchloric acid is a dangerous substance because it interacts with
oxidizable matter to cause violent explosions. With proper precautions, especially the avoidance of evaporation 'of alcoholic
, solutions of perchloric acid, the use of the aqueous solutions is
without hazards. 3
2 Caley, J. Am. Ghem. Soc. 54,3240,4112 (1932); 55, 3947 (1933) Caley and Burford,
Ind. Eng. Chern., Anal. Ed. 8, 63 (1936).
3 Perchloric acid, The G. F. Smith Co. Columbus, O.

1336

DECOMPOSITION OF THE SAMPLE

III. Fusions
j
1. The insoluble residue remaining from an aCid treatment of a substance
may be brought into solution, generally, by fusiotJ. with .an appropriate flux.
The fusions are made in crucibles (20 m!. capaGity) of silica, iron, nickel,
silver, " palau" (gold-palladium 80 : 20) or platinum. Potassium qisulphate
fusions may be made in silica or glass as well as iii platinum; sodium carbonate
fusions may be made 'in iron, nickel, "palau"," and platinum; hydroxide or
peroxide fusions may be made in iron, nickel or silver .crucibles; sodium carbonate potassium nitrate fusions may be made in platinum, but hydrochloric
acid should not be used in dissolving the fusion as the chlorine liberated will
attack the platinum.
2. In making fusions the sample is well mixed with six to twenty times its
weight of the fusion mixture and placed over a thin layer of the flux in the
crucible and then covered by a lid of the same material as that of the crucible.
Heat is applied cautiously until the mixture melts, the heat being raised during
the action. When the melt has reached a quiet fusion and the ,resulting
molten mass has become clear and homogeneous, the action is complete'.
(a) A stiff platinum rod is placed in the molten mass and held in place
until this has cooled, sufficiently to solidify around the wire. By reheating
the crucible the mass of fusion may be lifted out, cooled and together with
the crucible placed in the beaker in which the subsequent treatment is made.
(b) The molten mass may be poured into a crucible lid of sufficient capacity
and when cooled placed in the beaker together with the crucible.
(c) The molten mass is cooled by rotating the crucible or flask in such a
way that it solidifies in a layer on the sides.
3. Water is now added to the material and the mixture heated until the
mass disintegrates. The beaker is covered by a clock glass and the ~dditional
solvent now added. This may be hydrochloric or sulfuric acids, ammonium
carbonate or any other appropriate solvent.
The following fluxes are commonly used- I

L Sodium carbonate.-Decomposes silica~'es of 'AI, Ca, Ce, Cr, etc.; halides of

Ag; sulfates of Ba, Pb.
2. Sodium and potassium carbonates.-Silicates of AI, F, Fe, Ni, Se, Te, etc.
The mixture fuses at a lower temperature than carbonate alone.
3. Sodium carbonate with oxidizing agent, i.e., KQIOa, KNO a, Na 202, ZnO,
MgO.-Ores of Sb, As, Cr, Fe, Mo, V, Zr, sulfides, pyrites, etc.
4. Sodium peroxide.-Ores of Sb, As, Cr, Mo, Ni, V, U, Sn. Certain alloys
and steel for chromium, etc.

5. Potassium acid sulfate (with H 2S04).-Ores of AI, Sb, Cu, Cr, Co, Fe,
Mn, Ni, Rh, Ta, Ti, W. Steel with Cr, Mo, V and W, Th, phosphates,'
slags, oxides of metals of basic characrer. Silica rem~ins insoluble. In
a bisulfate fusion in a 250-ml. flask, 25-50 m!. of water and 10 m!. Hel
are added and the mixture boiled until the cake dissolves. If any dark
residue remains, tlie solution is decanted off and the residue again
fused with additional flux. As stated in pag}\ 1334, certain refractory
oxides require the addition of a redugng agent, such as carbon from a
filter paper, to assist the acid sulfate flux. It is better to make two
or more fusions on resulting residues remaining from an acid extraction
of the melt, rather than to increase the amount of flux in'the first charge.

DECOMPOSITION OF THE SAMPLE

1337

, 6. NaOH, KOH, basic flux (with or without KF).-Oxidized Sb ore, ores of

Cr, Sn, Zn, Zr. Sulfides and sulfates in sulfur determinations.,
7. Other reagents.-NH4Cl and CaCOa-alkali silicates, KCN--!tin oxide ore,
CaO, NazOz. Organic compounds for halide determination.
Reactions 1, 2, 3 may be carried out in iron, nickel or platinum crucibles,
5 in silica (porcelain) or platinum, 4, 6, and I7 in iron, nickel and silver.

Reagents.4--Ammonium acetate, NH4 CzH a0 2, is a very effective solvent of

lead salts, especially the sulfate. It is prepared by adding strong acetic acid
to strong ammonium hydroxide until the solution is just acid.
Ammonium hydroxide, NH40H, is a powerful solvent of chloride and bromide of silver.
Ammonium sulfide, (NH4 hS, readily dissolves the sulfides of arsenic, antimony, and tin. This reagent'decomposes slowly, losing NHa and forming
,(NH4 hS2. It should be kept in a dark cool place. To prepare the solution
pass a rapid current of pure hydrogen sulfide through a solution of ammonia
in a reagent bottle .
. Hydrochloric acid, HCI, is a powerful solvent and when mixed with HNO~
in the proportion (2-HCI : 1-HNOa) it will dissolve gold and platinum due
to the chlorine which is thus liberated in the nascent state.
Hydrofluoric aciq" HF, is used to decompose silicates, the 'silica being
volatilized.
Nitric acid, HN0 3 , is a powerful solvent and oxidizing agent. The fuming
variety is a more active oxidizer and desulfurizer.
Sulfuric acid, H 2S0 4, is a powerful solvent and is generally used in a dilute
form made by mixing concentrated acid (1.84 sp.gr.) with water in the pro
portion I-H 2S04: 5H 20.
Argol, KHC 4H 40 6, potassium acid tartrate commercial grade, is a powerful
reducing agent and also a basic flux.
Borax, cryst., Na 2B 40 7 10H 20, is an acid flux sometimes used as a cover
in place of N aCl. To dehydrate, fuse it in an iron or chalk lined clay crucible,
pour the fused mass on a clean surface and pulverize when cold. Melting
point, borax glass, 742 0 C.
Charcoal is a reducing agent and desulfurizer.
Iron, metallic, a basic flux, and desulfurizer. Nails or wire about Ys inch
dia. generally used. Melting point above 1500 0 C.
Lead, metallic, a basic flux. It is also used as a collector of the precious
metals in the assay of gold and silver ores. Granulated lead is used in scorification assays and sheet lead in bullion assays. Its silver content should be
determined. Melting point, 327 0 C.
,
Lead flux.-Sixteen parts sodium bicarbonate, 16 parts potassium carbonate, 8 parts flour and 4 parts borax glass. This mixture serves as a flux,
reducing agent, and desulfurizing agent.
Litharge, PbO, is a basic flux and an oxidizing and desulfurizing agent.
Its silver content should be determined. Melting point, 884 0 C.
Silica, Si02, is an acid flux. Powdered lime-glass can be used as a substitute. Melting point, Si0 2, 1775 0 C.
4

W. J. Crook, Lefax, Philadelphia.

1338

DETERMINATION OF WATER

Potassium bisulfate, KHS0 4, is a decomposing agent' and, an &cid flux.

Silica is not made soluble by this reagent. Melting point, 200 0 C., K 2S0 4 ,
1072 0 C.
Potassium carbonate, K 2 COa, acts same as Na2qOa and is .frequently mixed
with it. A mixture in proportion to the molecular weights of the two substances is an excellent flux for decomposing certainlsilicates, days, etc., which
are difficult to handle by either carbonate alone. Melting point, K 2COa,
OO~C.

Potassium cyanide, KCN, is a powerful reducing and desulfuriiing nux and

is frequently used in the fire assay ,of Pb, Sn, Bi and Sb. Its solution is a
.
solvent of gold and silver. Fuses at red heat.
Potassium hydroxide, KOH, acts same as NaOH and is used for the same.
.
purposes. Melting point, 360 0 C.
Potassium nitrate; KNO a, is a basic flux and also a powerful oxidizing agent.
Melting point, 3390 C.
Sodium bicarbonate, N aHC0 3, is a desulfurizing agent, a basic flux and
sometimes an oxidizing agent. The corresponding potassium salt may also
be used.
Sodium carbonate, N a 2 C0 3, is use_d alone or in conjunction with other
reagents for the decomposition of silicates, etc. Melting point, 849 0 C.
Sodium chloride, NaCI, a neutral flux, used as a cover in assay. Melting
_
point, 804 0 C.
Sodium hydrate, NaOH, is used principally for the decomposition of sulfides and sulfates in sulfur determinations. Sometimes used to decompose
certain silicates, oxides .and certain organic compounds. It is a basic flux.
Melting point of NaOH, 318 0 C.
Sodium nitrate, NaN0 3, is a decomposing and also an oxidizing agent.
KNO a is also used for same purpose. Melting point, 316 0 C.
Sodium thiosulfate solution, N a 2S20a, is a solvent for AgCl.
Oxidizing Agents.-Those in most general use are oxygen, chlorine, bromine, potassium permanganate, potassium bichromate, nitric acid, potassium
chlorate, sodium nitrate, hydrogen peroxide itnd ~mmonium nitrate.
Reducing Agents.-Hydrogen, H, is used lin the form of a gas which should
be dry and free 'from impurities. Hydrogen sulfide, H 2S, is prepared' by
adding dilute H 2S04 to pure iron sulfide. If pure iron sulfide is not at hand,
it can be prepared by fusing iron nails with' sulfur in J,pe proportion of about
1 part iron to 2 parts 'sulfur; by weight. Sodium sulfite, N a~03, is used for
the reduction of ferric solutions. It separates arsenious sulfide, which is soluble
in it, from the sulfides of antimony and tin, which are insoluble in it.' Stannous
chloride, SnCI2, is frequently used for redu~ing iron solutions. Solutions of
sugar, tartaric acid and many other organic compounds will serve as reducing
agents.
THE DETERMINATION OF MOISTURE, WATER O}<' CRYSTALLIZATION AND WATER OF CONS'!:ITUTION

1. Method of Loss in Weight.-A suitabl~, weighed sample which may

range from it few tenths of a gram up to lOo-'g, depending on the nature of
the material is either h~ated at a suitable temperature (usually 100-1050 C.)

DETERMINATION OF WATER

1339

cooled in a desiccator and weighed, or else allowed to stand, with weighing

after stated intervals, over a suitable drying agent such as anhydrous magnesium perchlorate, P 20 S, sulfuric acid (conc.), CaS04, CaCl 2 and the like,
until the weight of the residue is constant. The percentage loss in weight is
calculated. Details vary from substance to substance; the reader is therefore
referred to the detailed table of contents for the determination of water in various materials, as for example, paper,
soap, various ores, etc.
With crystallized salts, a temperature of 100-105 C. is
frequently insufficient to drive ~ff all of the water of crystallization. For example: Crystallized copper sulfate loses
4/5ths of its water of crystallization at 120-125 C. and the
remaining 1/5 only at 240-250 C. at atmospheric pressure
and humidity. Some salts melt in their own water of crystallization' and undergo hydrolysis after which it may be
impossible to complete the drying without decomposing the
salt; this is true of common (potash) alum. In such instances the bulk of the water must be driven off at room
temperature or at somewhat higher temperature, after which
the drying may frequently be completed at 100 C.
Some salts, notably acid salts, like Na 2HP0 412H 20 lose
their water of crystallization at 100 C. or below. Upon
igniting the dehydrated residue at full burner temperature
the salt is transformed into a " pyro " salt with loss of " water
of constitution." Example: 2Na2HP04=H20+Na4P207.
This is one of the kinds of changes that make it difficult
in many cases to interpret fully the significance of loss on
ignition at high temperatures. Oxidation of the lower
oxides in the non-volatile residue is another complication
that frequently occurs. For these reasons the direct determination of moisture or water of combination is frequently advisable whenever the simpler loss-in-weight determination c'an not be used.
2. Absorption Method.-The material is heated in a
boat in a combustion tube as indicated in Fig. 180. The
evolution of the water-and its transmission to the absorption
vessel are aided by a stream of dried gas, either air or other
gas inert toward the material in question. The gas stream
is dried by the sa,me.. drying agent that is to be used for
trapping the water that is evolved on heating. The drying
agent is chosen according to the products that may be expected on heating the material strongly. For example, in
determining moisture in a carbonate, sulfuric acid \vould
absorb the water snd not the CO 2 evolved. If ammonia were evolved, as
for example -in investigating double ammonium phospnates of magnesium,
manganese, etc., lime would be a good absorbent for the water. The weight
of water from a given. weight of sample is found by weighing the absorption
tube before and after the heating. Particular care must be taken to drive all
of the water from the end of the combustion tube into the absorbing vessel.

1340

DETERMINATION OF WATER

, The boat and combustion tube may be of any convenient material for the
purpose at hand-hard glass, quartz, porcelain or p)atinum.
Penfield Method.-A variant of the evolution and weighing method that
is frequently used in mineral analysis was devised by Penneld. 5 The 'rock or
mineral powder is placed in a bulb of hard glass; the tube connected to ~he
terminal bulb is widened out to form bulbs (one or tWb) at a distance of several
centimeters from the terminal bulb. The open end of the tube is connectEld ..
to a capillary tube by rubber during the heating of the powder. The water
that is driven off is caused to condense by cooling the bulbs; t.tte capillary
prevents loss of water. Finally the bulb with the mineral residue is removed
by sealing off the tube in a flame, and the part containing the water is weighed,
and the water is then removed from it and the weight of the glass is determined.
DISTILLATION METHODS FOR MOISTURE

In the distillation methods, the moisture is removed from the saDlple by

the distillation by suitable means of heat of a solvent immiscible with water.
The sample mayor may not be soluble in the distilling medium. The distilling medium has a boiling point higher than water, is condensed, and collected with the water in a suitable trap or measuring tube. The excess of the
condensed solvent is .returned to the distilling vessel or is drawn off from the
measuring tube.
Marcusson 1, la, 2 was one of the first to determine moisture by distilling'~
the samples with a liquid immiscible with water and measuring the water in
the manner used in the methods described in this chapter. The sample was
introduced into the distilling flask into a measured quantity of toluene 2 or
xylene. 1a The mixture was distilled into a receiver constructed at the bottom
to a calibrated narrow tube until several hundred cubic centimeters were
recovered. The Dean and Stark, Bidwell and Sterling, and Kingman are
some of the many variations of the method.
Some of the advantages of distillation methods over oven or hot plate
methods are: the actual collection and measurerrient of the water present
separated from other volatile constituents; little or no opportunity for oxidation such as exists in oven or hot plate methods; little or no effect due to
humidity such as existil in oven 'or hot plate drying where the sample is cooled
in a desiccator before final weighing; and the use of a larger and more representative sample.
The methods which are described are the most common of the distillation
methods in use.
1. Water in Petroleum Products and Other Bituminous Materials. 3 , 4, 6 _
This method is essentially the method described by E. W. Dean and D. D.
Stark. 6 The paper contains a comprehensive bibliography on moisture methods, particularly those of the distillation type.
" The method is suitable for a variety of materials, but is especially applicable tQ petroleum, fuel oil, road oil, coal tar, water-gas tar, coke-oven tar,
and other petroleum products or bituminous materials."
For distilling apparatus, either: a glass flask or a metal still may be used.
Glass is generally used for petroleum prod~c.ts..f;ind the metal still for road
:; Penfield, Am. J. Sci. (series 3), 48, 31 (1894).
Compiled by Mr. C. P. Long of the Chemical Division of Procter and Gamble Co.

DETERMINATION OF WATER

1341

materials and tars. An electric heater is preferred for glass, although a gas
burner may be used. A ring gas burner is used for the metal still. Dean and
Stark prefer an electric'heater. 6 7
The apparatus, made according to A. S. T. M, specifications, may be purchased from chemical apparatus supply houses. The condenser is a glass reflux
type of prescribed length so arranged as to discharge the condensate in a
Dean-Stark trap. The trap is 10 illl. capacity, tapered from the 2. ml. mark
to 0 ml. Above 2 ml. the inside diameter is 18 to 19 mm. The part below
the 2 ml. mark is calibrated in 0.1 ml. divisions.
The solvent used when testing petroleum products or bituminous materials
derived from petroleum shall be gasoline, free from water, and show the following distillation range:
5% shall distill at a temperature not below 90 C. nor above 100 C.
90% shall distill below 210 C.
The solvent used when testing bituminous materials,derived from coal tar,
w.ater-gas tar, etc., shall be a coal-tar naphtha or light oil and shall show the
following distilling range:
98% shall distill between 120 C. and 250 C.
Procedure: For a regular Dean and Stark trap the sample size should be
such that not over 10 ml. of water are collected. For samples high in moisture
a longer trap may be secured or the sample size reduced. The volume of the
sample is accurately measured, transferred into the distilling vessel, and mixed
with 100 ml. of solvent by swirling, taking care to avoid loss of material. The
solvent can best be added to the still by rinsing the graduate with several,
portions. The connections between the distilling vessel, condenser and trap
are made with tight-fitting corks. The heat is applied and regUlated so that
the condensed distillate falls from the end of the condenser at the rate of 2
to 5 drops per second. The distillation should be continued at the specified
rate until no water is visible on any part of the apparatus except in the bottom
of the trap. A persistent ring of condensed water in the condenser tube shall
be removed by increasing the distillation rate for a few minutes.
The volume of the water in the trap at room temperature is meaf?ured and
calculated to the percenta'ge of the original volume of the sample.
2. Brown-Duvel Method for Grains.-The apparatus was described in
1907 by Brown and Duvel. 8 The apparatus was patented 9 by them in order
to allow free use of it by the public. The various components of the apparatus, accessories, details of operation and effect of variation in certain details
are covered in a bulletin written by D. A. Coleman and E. G. Boerner. lO Since
this bulletin covers all the specifications in detail for the flasks, thermometers,
and all other parts of the apparatus and accessories as well as details of the
methods for various grains, it should be available before the apparatus is placed
in use.
The apparatus includes a heating chamber, arranged so several samples
may be handled at the same time, and a tank containing cold water through
which condensing tubes pass. The various parts are fastened on a suitable
frame. Gas burners are the usual source of heat.

1342 .

DETERMINATION OF WATER

The distilling medium is a good grade of mIneral engine oil, free lfrom water,
0.900 specific gravity, viscosity 342 to 513 [seconds Saybolt at 20 C., and
open cup flash point, 3500 to 400 0 F. The fir~-point should not be over 450 0 F.
Grains are placed in the distilling flask whol~ in a weighed amount specified.
for each grain. With the standard thermometer in a definite position in a
specified quantity of oil, heating is continue4 until the thermometer in 'the
oil reaches the temperature designated for the kind of grain under test. Then
the flame is extinguished. The water is measured when the thermometer has
dropped to 1600 C. in a graduate under the small layer of oil which distills
over with the water.
The endpoint, i.e., flame extinguishing temperature for the g:r.:ain, is standardized or set against drying in a water-jacketed oven under standard conditions.u
3. Toluene-Moisture Method (Bidwell-Sterling).1. 2__The receiving and
measuring tube used is a modification of the Dean and Stark tube. In order
to increase the accur~cy of the measurement of the water layer, the calibrated
lower end consists of a narrow tubular portion similar to a burette calibrated
in 0.1 ml. divisions. The volumes Illay be read. to 0.01 ml. especially if a
hand-lens is used. Toluene (B.P. 111 0 C.) is used as distilling solvent rather
than xylene (B.P. approximately 1400 C.) since the lower distilling temperature
of the solvent, apparently eliminates decomposition of many of the organic
materials under investigation.
The method was tried by Bidwell and Sterling on a large variety of organic
materials, such as raisins, flours, butter, dried milk, leather,1~ molasses, grass.
The method is satisfactory for moisture in essential oils,13 soaps (especially
liquid and jelly soaps), glycerine,14 soybeans, and other grains, fats, mineral
oils and all sorts of mixtures and emulsions.
The receiving tubes and also complete apparatus including condensers with
standard ground glass connections may be purchased from apparatus supply
houses. A 250-ml. Erlenmeyer or Soxhlet flask is satisfactory for the dis. tilling flask although th~ size of the flask ma,y be increased with large Ramnle
sizes. As source of ,heat, electric heaters so 'djusted that the heating element
is enclosed and non-glowing, and Argand gas burners are satisfactory.
The toluene used should have been previously heated under a reflux condenser connected to the flask by a Bidwell-Sterling tube until surplus mois'ture
has ceased to distill off. The receiving tube should be carefully calibrated
and before using should be absolutely clean. It isadvis_able to clean the tubes
and condenser with chromic-sulfuric acid cleaning solution followed by water
and alcohol rinses and oven drying.
Procedure.-Weigh sufficient sample to -give 2 to 5 gms. of water into the
distilling flask. Add sufficient toluene to cover the sample completely,' usually
about 75 ml. Connect the receiving tub~ and reflux condensers. Fill the
tube through the top of the condenser. Protect the side arm from the heat
of the burner by a piece of asbestos board vertically sUl?pended.
With water flowing ,through the condenser, heat gently to avoid foaming,
continue boiling slowly at about 2 drops per second until nearly all the water
has distilled over. Then increase the rat(} _to abo.U't 4 drops per second. A
boiling piece is often an advantage. Fo.r' frOme samples Bidwell and Sterling
recommend sufficient dry san.d to cover the bottom of the distilling flask.

DETERMINATION OF WATER

1343

When the water has apparently all been distilled over to the receiving tube,
wash down the condenser by pouring toluene in at the top. . Continue distillation a short time to determine if further water distills over. If it does,
repeat the ~ashing of the condenser. Any drops of water adhering to the
sides of the condenser tubing usually can be loosened by a copper wire through
the tpp of the condenser.
Allow the receiving tube to come to room temperature or bring to a definite
temperature by immersing in a bath of this temperature. Read the volume
of the water. The volume of water multiplied by the specific gravity at the
temperature at which the volume is read will give the weight of water in gms.
When used on glycerine or a soap high in glycerine, the distillate may be
corrected for its glycerine content by separating the aqueous layer and determining the glycerine content by the bichromate method. (See Methods for
Analysis of Soap.) Alcohol and acetone cause very unsatisfactory and unreliable results.
.
In order to prevent soaps from foaming, a carefully measured amount of
concentrated hydrochloric acid in sufficient quantity to break the soap into
fatty acids is very desirable. The water content of the acid must be known
and proper correction made in the amount of 'water recovered. Soaps containing much carbonate should be handled by the xylene-sodium acetate method.
Glass wool floated in the liquid will often prevent foaming without the addition of acid.
Soybeans and grains are placed in the flask whole. Distillation is continued
until water ceases to come off. The toluene in the distilling flask is allowed to
cool somewhat. The beans or grain are then strained out on a coffee strainer.
The solvent is returned to the distilling flask while the beans are given a rinse
with a small amount of petroleum ether and allowed to drain until practically
free from solvent. The sample is then quickly ground in a coffee mill and
returned to the distilling flask and distillation continued. until water ceases to
distill over.
4. Xylene Method.-Marcusson in 1904 2 used toluene as distilling agent
in a distillation method for lubricating greases. In 1905 1. 1.\ he describes a
similar method for water in oils, fats, and soaps in which xylene was used as
the distillation medium. In a book 18 by Marcusson 0927) on methods for
analyses for fats 'and oils, xylene is used as distillation solvent and the procedure is essentially that described in the previous references. If soaps are
present so that foaming will occur and carbonates are absent, dry oleic acid
or potassium bisulfate is added.
Rogers 12 and Bidwell and Sterling 1 pointed out that many organic materials which showed definite evidence of decomposition when xylene was used,
could be safely handled when toluene was used. In the soap methods,19 a
xylene method with anhydrous sodium acetate tel prevent foamingis described.
The general procedure with xylene is identical with the Bidwell-Sterling
procedure using toluene. The approximately 30 C. higher boiling of xylene
is often an advantage on materials that do not decompose with the liberation
of water at the distilling temperature and that hold the water rather tightly.
5. Kingman Method;-The method was described by W. A. Kingman. ls
The method had been used by him for several years. It differs from the other
distillation methods described in this section by the use of a ,non-inflammable

1344

DETERMINATION OF WATER

solvent, tetrachlorethane (also called commerciatly acetylene tetrachloride).

The solvent is heavier than water so that the former is the lower layer in the
measuring receiver. The Kingman measuring tube. is eql,lipped with a stopcock at the lower end so the solvent may be draWn off during distillation in
order to keep the water.from running back to the1distilling flask (Fig r 181).
I

Flask

Bore o~ graduated tube to be of

such sjz_/! that 10 ml. graduated as
;speci{ie'd occupies about 200 mm.

Glass Stop Cock

FIG. 181.

Kingman Mbisture Apparatus.

The method is an official method of the American Ch~inical Society 16 and

the American Oil Chemists' Society 17 for moisture in fats.
The apparatus used is the regular Kingman set.,..up which may be obtained
with all glass joints (Fig. 181). The distilling solvent is tetrachlorethane.
Benzene is used to aid in m!lasuring.
Determination.-Weigh from 50 to 200 gms. of the sample into a 500-m!'
Erlenmeyer flask. For samples containing less than 1 % moisture use 200
gms. ; for samples containing from 1 to 5% moisture, 100 gms.; and for those
containing larger quantities, smaller amounts. Add about 100 m!. of acetylene
tetrachloride to the sample in the distillatiolJ- _flask all'd distill. Continue the
distillation until from 25 to 40% of the soWeTIt has distilled over, drawing off
solvep.t from time to time. At first the bOlling temperature will be about

DETERMINATION OF WATER

1345

80 C., rIsmg gradually to 95-100 C., when practically all the moisture is
driven off, then going up rapidly to about 144-145 C. During the rapid rise
the heating should be slowed down; otherwise a too rapid evolution of vapor
ensues which heats the condenser and causes the water already distilled to boil
away from the receiver tube. Allow the apparatus to cool and then wash
down the condenser and connecting tube into the graduated tube of the apparatus with a little coal-tar benzene. When the moisture has settled sufficiently, read the column of water, making due allowance for meniscus, and
calculate the percentage.
The method is not as satisfactory for soaps as the toluene or xylene methods.
REFERENCES
1. G. L. Bidwell and W. F. Sterling, J. Ind. Eng. Chem., 17, 147 (1925).
1a. J. Marcusson, Mitt. klg. Materialprlifungsamt, 23, 58 (1905).
2. J. Marcusson, Mitt. klg. Materialprlifungsamt, 22,48 (1904); Z. angew. Chem.,
. .18, 754 (19Q5).
3. A. S. T. M. Standards, 1933, Part II, Non-metallic Materials, page 891, A. S. T. M.
D95-30.
4. A. S. T. M. Standards, 1933, Part II, Non-metallic Materials, page 827, A. S .. T. M.
'D128-27.
5. A. S. T. M. Standards, 1933, Part II, Non-metallic Materials, page 814, A. S. T. M.
DI58-28.
.
6. E. W. Dean and D. D. Stark, J. Ind. Eng. Chem., 12, 486 (1920).
7. E. W. Dean, J. Ind. Eng. Chern., 10,823 (1918).
8. A Quick Method for the Determination of Moisture in Grain, Bclletin 99, Bureau
of Plant Industry, U. S. :qept. of Agriculture (1907). Reissued as Bureau of
Plant Industry, Circular 72, under title" A Moisture Tester for Grain and Other
Substances and How to Use It."
9. U. S. Patent No. 848616, March 26, 1907.
10. The Brown-Duvel Moisture Tester and How to Operate It, D. A. Coleman and
E. G. Goerner, Bulletin 1375, U. S. Dept. of Agriculture, 1926.
11. Air-oven and Water-oven Methods Specified in the U. S. Standards for Determining the Moisture Content of Grain, U. S. Dept. of Agriculture, Bureau of
Agricultural Economics, Service and' Regulatory Announcements, No. 147, April,
1935.
12. Rogers, Bureau of Chemistry, Bulletin 137, 172 (1910).
13. R. M. Reed, Oil and Soap, 9, 67, 1932.
14. L. F. Hoyt & P. C. _Clark, Soap, 7, 25, 1931.
15. W. A. Kingman, J. Ind. Eng. Chern., 18,93 (1926); Determination of Water Content of Liquid Glue (1926).
16. Standard Methods for Sampling and Analysis of Commercial Fats and Oils, J. Ind.
Eng. Chern., 18, 1348 (1926).
17. Standard Methods for Sampling and Analysis of Commercial Fats and Oils, A. C. S.,
and A. O. C. S. Lefax edition.
18. J. Marcusson, Die Untersuchung del' Fette und Ole, Dritte Auflage, Verlag von
Wilhelm Knapp, Halle (Saale) (1927).

19. See chapter on Methods for Anal:ysis of Soap and Soap Products.
Calculation of Percentage of Material Determined in the Original Sample
to the Dry Basis, or if Determined on a Dried Sample to the Original Form
Including Water. Frequently a chemist is required to report the percentage of
constituent both on the dried sample and in its original form containing wa~er.
The following formulae' will be of assistance in these calculations:

DETERMINATION OF WATER

1346

#.
III

Let a = percentage of material found (determined on the dry basis, or

the original sample, as the case may be).
b = percentage of total dry material with water expelled.
c = per cent of water in the original sample, 100 -c = b.

100Xa
.
Then --b- =per cent calculated from wet baSIS to dry.

al~:

= per cent calculated from'dry basis to wet.

DETERMINATION OF WATER

1347

DETERMINATION OF MOISTURE IN GASES

The gas to be tested is passed through a dehydrating agent such as phosphorus pentoxide, P 20 S, alumina, A1 20 3 , or lime, CaO, contained in a weighed
U-tube. The following facts should be borne in mind in selecting the dehydrating agent: (a) It should not absorb the gas; (b) it should not react chemica,lly with the gas. For example, lime or alumina could not be used for
determining moisture in sulfur dioxide, nor could phosphorus pent oxide be
used in determining moisture in ammonia. For the former, phosphorus pentoxide is best, and for the latter, lime. Alumina that has been carefully heated
to 1400 is useful for determining moisture in neutral gases. It should be
remembered that gases dried over calcium chloride will give up moisture to
concentrated sulfuric acid, and these in turn will give up moisture to phosphorus pentoxide: Professor Morley has even determined the amount of
moisture that is left after this latter treatment.
Procedure. The volume of the gases required for the test varies widely
according to the percentage of moisture in the gas, 1000 ml. to 10,000 ml. are
generally required. For minute amounts of moisture it may be necessary to
lead the gas over the dehydrating agent for a given length.. of time, using a
manometer 'or difference gauge, or a gas meter. The absorption' tube is
wejghed before and after the test and the increase in weight taken as the
moisture content of the gas. See Chapter on Gas Analysis.

ALLOYS, I

ANALYSIS OF NON-FERROUS ALLOYS

Classification of Alloys.-There is wide diversification among alloys commonly used, both in properties and in composition. In general, they may be
divided into two groups, ferrous (containing iron) and non-ferrous (not containing iron). The great majority of alloys are non-ferrous. Those that do
contain ir<;m are called " iron alloys," to distinguish them from steel. Most
of these latter are corrosion-resisting, heat-resisting and electrical resistance

alloys, and resemble alloys that are strictly non-ferrous.

Alloys are utilized for various purposes requiring for instance :-mecnanical
strength, fatigue, resistance, corrosion resistance, electrical conductivity,lightness, heaviness, toughness, fusibility, infusibility, anti-frictional properties,
etc. Naturally, one alloy cannot meet all such requirements physically, and'
care must also be exercised in the selection of alloys for economic reasons.
Many alloys are known to the trade' by names 2 which designate their
prime use in industry; others are otherwise designated. Some alloys of dissimilar composition may be put to the' same purpose, while alloys of similar
composition may be unsuited for a common use. Again, the same alloy may
be put to several 'purposes. Hence it is appar\)nt that a classification based
upon the trade name of an alloy or upon the purpose to which the alloy may
be put would cause considerable confusion.
The following classification is adopt!l.d from W. Campbell: 3
. Brasses
Bronzes
Copper-nickel alloys
Copper-nickel-zinc alloys
Aluminum bronzes
Copper-manganese alloys
Copper-silicon alloys
Aluminum alloys

Magnesium alloys
Tin base alloys
Lead base alloys
Zinc base alloys
Fusible metals
Gold alloys
Silver alloys
Platinum 'alloys

I
(

Resistance alloys
Heat resistance alloys
Corrosion-resistance -alloys
Stainless and corrosion resistance alloys
Cobalt alloys
Hard alloys
Pyrophoric alloys

Preliminary Examination.-No single reagent will act as a common solvent

for all alloys. Among the solvents used may be mentio~a nitric acid, hydrochloric acid, sulfuric acid, sodium hydroxide, and fusion with alkaline fluxes.
At times a mixture of acids is required.
The following scheme with proper modific::tion will serve for the identification of most alloys.
NITRIC ACID METHOD:

Treat one or more grams of the alloy filings or drillings with concentrated HN0 3 or
HNO s (1 : 1) until ,the metal is completely decomposed or in. solution, adding water if
necessary to assist solution.
'
1 Chapter contributed by W. J. Brown, Chief Analyst, National' Lead Company.
2 "Proprietary Alloys," Metal Industry, London, 193ft
3" A List. of Alloys," by W. Campbell, published by the A. S. T. M., 1930.
1348

ANALYSIS OF -'NON-FERROUS ALLOYS

1349

(a) If the metal dissolves, antimony and tin are absent; omit their tests.
(b) If the metal forms a turbid .~olution, antimony and tin may be present. To a
portion, add several volumes of water and boil. (a) If a clear solution is formed treat
the entire solution in the same way, and omit tests for Sn and Sb.
(c) If the residue is insoluble, Sb and Sn are indicated, evaporate the entire solution
to dryness, gently ignite to expel water, add HNO" heat, dilute and filter. If the
filtrate is turbid, repeat the filtration until clear.
Precipitate.-Sn02; Sn,(PO,),; Sb 20,;
Filtrate.-Contains the H 2S group
(without tin and antimony) and (NH 4).S
5, all white and. traces of the copper
group metals.
gro,Up.
a. Digest from 10 to 20 minutes with
Add from 1 to 3 ml. of concentrated
strong K.S, Na 2S or (NH 4).S solution and H 2S0 4 and evaporate to fumes, cool, dilute
filter.
and filter.
b. Melt with Na 2CO, and sulfur, cool,
digest with water, and filter.
As 20

Precipitate.Traces of copper
group. If sufficient in
quantity: dissolve in
in HNO,; add to filtrate containing
(NH 4)2S group.

Filtrate.-Contains Sn, Sb, As, in

solution. Just acidify with dilute HCI
and filter.
a. Analyze the
precipitated sulfides.
b. Test the filtrate for phosphates.

Precipi ta te.PbSO" white. Treat

with NH 4C 2H,O.
solution, filter, and
add K 2Cr04. A yellow precipitate insoluble in KOH
proves lead present.

Filtrate.-Contains the H 2S group

(without Sn, Sb, and
Pb and free from
HNO,) and the
(NH,).S group.
Into a portion of
the solution pass
in H 2S. A precipitate ,shows the pres_ _ _ _ _ _ _ _"'-_ _ _ _ _::--_ _' ence of the H 2S group metals.

a. If the H,_s group is not present, as shown by a test on a portion of the solution,
make a test for the (NH,).S group.,
b. If the H.8 group is present, precipitate the metals with H.S, filter and wash.
Precipitate-Contains the H 2S group,
sulfides (without lead, tin and antimony).
If mercury is present use (NH')2S to
separate the sub-groups ..
If mercury is absent digest the precipitate with 20% KOH (2 to 3 grams dissolved
in 10 mi. of water), filter and wash.
Precipitate.Copper group.
Analyze by procedure outlined for
the insoluble H 2S
subgroup.

Filtra t e.-Arsenic group in solution. Acidify, filter

and analyze precipitate by the procedure, for the soluble
H 2S subgroup.

Filtrate.-Contains the (NH 4).S gro)lp

a. Test for Phosphates. If found present consult notes on this group in Part II.
b. Make a test for the group, by making a small portion ammoniacal, and passing in H.S, if not present. A precipitate
.other than sulfur indicates the presence
of the metals of this group. Analyze the
precipitate by the first procedure on the
metals of the (NH,).S group.
c. If a test is made for magnesium,
analyze the filtrate by procedures for the
alkalies.

Methods of Procedure.-Alloys are too complex and vary too much in

composition for any single method of procedure to suffice for' a complete quantitative analysis. For alloys soluble ,in nitric acid and containing a moderate
amount of tin and a small amount of antimony, the scheme for the identification of an alloy may be followed in a more or less general way. For alloys
containing considerable tin or antimony, solution is best effected with a hydrochloric acid-nitric acid solution or a hydrochloric acid solution with an oxidizing
agent such as potassium chlorate. Upon evaporation, lead may -be separated

1350

ANALYSIS OF NON-FERROUS Ai-LOYS

as chloride. Subsequent treatment'-will depend upon just wh;at metals are

present or desired to be determined. A 'proceaure followed for the determination of one constituent may be unsuited for the determination of another
constituent in the same alloy. Thus in a light aluminum alloy, a sepa-ration
of copper, nickel, calcium and magnesium may be made by-a treatment of the
,alloy with sodium hydroxide and these metals I may be determinect in the
insoluble portion, while determinations for other constituents are made by
dissolving the alloy in acids.

It is the exception rather than the rule that a complete analysis of an alloy
containing several elements is carried out on a single weighed portion. Several
determinations are made, the choice of method depending not alone upQn the
metals desired or the metals associated with them, but also upon the extent to
which they may be present.
Difficulty of Complete Separation of Elements.-In maki!.J;g an exact metal
analysis the chemist should have in mind that many of the separations are
incomplete. Solutions may require further treatment for the recovery of unprecipitated portions and precipitates may be contaminated, thus necessitating
a second and even a third separation. It is only in this manner that the accuracy of an analysis can be assured.
Limit of Accuracy.-In general an analysis is conducted in duplicate and
results expressed in percentages of the alloy should not differ by more than
1 unit in the number at the right o~ the decimal point. This applies to components that are present in appreciable amount, say 5, 10 or more per cent.
Usually a determination for a component present in such an amount is made
on 1 g. of sample and results that differ by 1 unit in the tenths position represent a good percentage accuracy, whereas the same difference would represent
a poor percentage accuracy when a component is present in small amount,
as.for example 1,0% or less. On the other hand, the same percentage accuracy
is not to be expected when a minor constituent is present only, to the extent
of about 0.05%; in such a case duplicates as close as 0.04 and 0.05% would
I
be satisfactory.
A metal that is a main component of one a1l6y may be an impurity present
in very small amount in some other alloy. It is quite apparent the same method
of procedure should not be followed in botp. instances. In order to determine
a very small amount of an impurity witp. suitable precision a large sample
would be analyzed (10 g. or more depending upon the ease of handling and
the precision desired) or a more sensitive method used'Or both. It is a common
practice in the analJ:sis of metals to weigh out charges as large as 200 and 500
grams.
The sum of the components of an alloY,should, for exact work, total within
0.2% of 100%. When, for commercial reasons, a total of 100% is expressly
desired, the difference plus or minus between the summation and 100% must
be proportionately taken from or added to the several components. (This,
of course, is not strictly accurate, since the individual errors may vary in
amount as well as in direction). Occasionally, a main constituent is obtained
by difference so that the summation is exactly 100%. In reporting percentages of major components only four significant figures ,~re retained.
Sampling.-Frequently, to save time, a bUsy analyst will chip a piece from
aq, ear of it pig of metal or rasp a few grams of filings from an edge of a bar.

ANALYSIS OF NON-FERROUS ALLOYS

1351

Most alloys are not sufficiently homogeneous to be sampled in this manner.

In a molten mass of alloy there are generally constituents that have a wide
range in their melting points and upon cooling, the more infusible ones solidify
and attain appreciable crystal or particle size while the bulk of the mass is
still liquid. If the cry"stals are lighter or heavier than the liquid metal, they
will tend to segregate to the top or bottom. It is advisable to stir thoroughly
while in a completely molten condition before casting and if the analyst is
taking his own sample from a wixing pot, it should be cast in a quick-cooling
form.
Samples submitted to the laboratory may be in a variety of shapes, sizes
and conditions. When in the form of a pig, ingot or bar, they may be drilled,
sawed or filed. Filings represent only the portion that is treated and unless
the whole bar is rasped this method of sampling is to be avoided. Drillings,
as a rule are ioo coarse for portions of 1 g. to be weighed out. However,
drillings from certain low-melting alloys may be melted under rosin, and after
stirring, cast into a small bar which is sampled by sawing. Sa wings generally
call be weighed out in 1 g. portions and this method is to be preferred for the
sampling of most alloys. For the correct method of sampling see" Sampling
Metals and Non-Ferrous Alloys" under chapter on Standard Methods of
Sampling.
Even with correct sampling care should be exercised when weighing out
the portion for analysis. In the case of a 20% antimonial lead alloy, cubic
crystals of antimony form. These crystals are hard and brittle and will
pulverize when cut with a hack-saw blade. Consequently, when such an
alloy is sampled, the sawings obtained will be a mixture of coarse particles
and fine dust. The coarse particles have an appreciably lower and the dust
a very considerably higher antimony content than that of the original alloy.
Sawings of this nature are screened and the fine and coarse material weighed.
Each portion may then be weighed out proportionately and combined or run
separately and the results calculated back to the original basis. It may be
possible to weigh without screening if a large portion is weighed out, but in
that event a different procedure is followed. The material is quartered down
to approximately the weight desired by a riffle sampler.
Before weighing out a portion for analysis, pass a magnet through the
sawings, drillings or filings to remove any iron that may have come from the
cutting tooI. Do not, however, pass a magnet when the alloy itself is magnetic.

ANALYSIS OF BRASSES AND BRONZES

. A b;ass, strictly speaking, is an alloy of copper and zinc; a bronze is an
alloy of copper and tin. In both alloys copper is the predominating metal.
The modern brass may have other constituents in addition to copper and zinc,
and the bronze other constituents in addition to copper and tin. In many
bronzes the tin may be wholly displaced by another metal or metal~.

1352

ANALYSIS OF NON-FERROUS ALLOYS

Practically all of the following methods given for the analysis of a brass
or a bronze call for a preliminary separation of ltin as meta-stannic acid' by
means of solution and digestion of the alloy in nitric acid. The meta-stannic
acid .so obtained is subject to contamination b~ such alloy constituents as
copper and iron and will quantitatively contain such elements as phosph.6rus,
arsenic and vanadium provided the ratio of tin to these elements in' the allpy
is high enough. In addition to the above drawbacks, tin is not always quantitatively precipitated as meta-stannic acid, particularly when the alloy contains more than 0.25% of iron. In ordinary analyses of materjal containing
but little iron, correction or provision is made for antimony and phosphorus
when they are present, and the contaminants and the slight solubility losses
are ignored. This procedure results in compensating errors in the determinations of copper and tin; for the impurities make up for dissolved tin while the
dissolved tin plates out to a considerable extent with the copper. In umpire
analyses, or with alloys high in iron or antimony, the above weaknesses of the
method must be taken into account and provision must be made for the complete recovery of all tin, as well as the complete recovery of all contaminating
elements carried down with the tin. 4
ANALYSIS OF MANGANESE BRONZE (A BRASS)

5.

General Considerations.-Methods are given for the determination of

copper, tin, lead, iron, manganese and aluminum. Zinc is taken by difference.
For methods of determining small amounts of zinc and also nickel, phosphorus, antimony, sulfur, arsenic, see" Analysis of Bronz&Bearing Metals."
DETERMINATION OF COPPER BY THE;, ELECTROLYTIC METHOD
Apparatus for Electrolysis. Electrodes.-The electrodes rec0mmended are
of the stationary and not of the rotating type, as the stationary require less
operator's time and there is less chance f.or errbneous results to be obtained
than with the other kind. Rapid and accurate results are obtained with
stationary electrodes made from platinum gauze.
Cathodes.-Platinum cathodes may be formed either from plain or perforated sheets, or from wire gauze, and may be either open or closed cylinders.
They should give a,depositing surface, counting both sides, of approximately
100 sq. cm. Gauze cathodes are recommended, and are best.made from gauze
containing approximately 400 meshes per sq. cm. (50 meshes per linear inch).
The cathode should be stiffened by doubling the gauze for about 3 mm. at the
top an~ at the bottom of the cylinder. The cylinder should be approximately
30 mm. in diameter and 30 mm. in height. The' stem should be made from
1.14 mm. or 1.29 mm. wire, flattened and welded the entire wjdth of the gauze;
the height over all should be approximately 135 mm., and the gauze should
be sand blasted.
A. S. T. M. Methods of Chemical Analysis of Metals, p. 157, December, 1936.
A. S. T. M. Methods of Chemical Analysis of Metalf/\ p. 133, December, 1936.
These methods, exclusive of that for aluminum, are-approved as American Standard by
the Al1lerican Standards Association.
. ._F6 For Specifications for Manganese-Bronze Ingots, see 1936 Book of A. S. T. M.
Standards, Part I, p. 563.
4

ANALYSIS OF NON-FERROUS ALLOYS

1353

Anodes.-Platinum anodes may be of the spiral- type when used in the

determination of copper by electrolysis, or in the electrolytic determination of
lead when it is pJ:esent in amounts not over 0.2%. When used in electrolytic
determination of copper and lead together in samples containing O.!?% lead
or over, the anodes should be of wire gauze. Spiral anodes should be made
from 1.02 mm. or 1.14mm. wire formed into a spiral of 7 turns having a height
of approximately 51 mm. and a diameter of 16 mm., the length over all being
approximately 145 mm. Platinum gauze used in making anodes should contain approximately 400 meshes per sq. cm. (50 meshes per linear inch). The
gauze should be formed into closed cylinders. approximately 12 mm. in diameter. and 30 mm. in height. The cylinders should be stiffened by doubling
the gauze for about 3 mm. at the top and at the bottom. The stem should
be made from 1.02 mm. or 1.29 mm. wire, flattened and welded the entire
width of t:le gauze; the height over all should be approximately 137 mm., and
the gauze should be sand blasted.

Method.-In a 150-ml. beaker dissolve 1 g. of bronze in 10 ml. of HN0 3,

(sp.gr. 1.42). When the action has ceased, evaporate the solution to dryness,
and bake thoroughly on the edge of a hot plate. Moisten thoroughly with
HN0 3 (sp.gr. 1.42), and after digesting for a short time dilute to 50 ml. with
distilled water, neat to boiling, and allow to stand and settle for about one
hour, keeping the temperature just below the boiling point. Filter on double
7 cm. papers, keeping the solution hot, and receive the filtrate in a 200-ml.
electrolysis beaker. Wash with boiling hot water. Add to the filtrate 5 ml.
of H 2S0 4 (sp.gr. 1.84), and evaporate until copious fmnes of H 2 S0 4 are evolved,
Dilute to about 100 ml. with distilled water, add 1.5 Inl. of HN0 3 (sp.gr. 1.42).
insert electrodes, cover with 'a pair of split watch-glasses and electrolyze.
For each solution use a current density over night of 0.5 amp. per sq. dm.
(4.5 amp. per sq. ft.) or use a current density of 4.0 amp. per sq. dm: (36 amp.
per sq. ft.) for 2Yz hours. The latter case requires the use of gauze cathodes.
When the solution is colorless, wash down cover gla~ses, electrodes, and sides
of beaker, raising the level of the liquid slightly ~nd continue the electrolysis
about 15 minutes, noting whether or not copper is deposited on the newly
exposed surface of the platinum. If no copper appears, transfer about 1 ml.
of the colorless solution to tIle cavity_ of a porcelain test plate, and add a few
drops of fresh H 2S water. If the slightest discolor::ttion occurs continue the
electrolysis, repeating the test. Remove the cathode quickly, rinse it in distilled water and then dip it in two successive baths ()f alcohol. Shake off the
excess alcohol and ignite the remainder by bringing it to the flame of an alcohol
-lamp. Keep the cathode moving continually as the alcohol burns. Weigh as
metallic copper.
Accuracy.-Duplicate determinations should check within 0.10% of copper.
NOTEs.~1. In alloys of the type of manganese bronlle, which carry considerable
iron, it is impossible to remove the tin quantitatively as metastannic acid (H 2Sn03},
notwithstanding the preliminary baking.- Upon testing (by the addition of HzS water),
the liquid in the pits of the test plate may be colored yellow, due to the tin present, but
there should be no darkening due to the presence of copPer. If the cathode appears
dark, as though tin had been deposited on it, dissolve the coating in 10 m!. of HNO,
(sp.gr. 1.42)" filter off the 'tin as before, add to the filtrate I) m!. of H 2S0 4 (sp.gr. 1.84),
and redeposit copper by electrolysis. It is rarely necessary thus to purify the deposited
copper.

1354.

ANALYSIS OF NON-FERJtOUS ALLOYS

2. If zinc is to be determined in the alloy ,be careful to keep the solution.quantitative

when the electrodes are removed, catching all washiQ_gs in the beaker containing the
electrolyte.
.
o
3. A slight amount of copper is retained by the ntetastannic acid (H 2Sn03) , which
should be recovered in very accurate work buUs generally neglected in routine analysis.
4. In the case of alloys containing larger amounts, of tin and less than 0.2% of iron,
I
. .
a preliminar:y baking is not essential.
DETERMINATION OF LEAD BY THE ELECTROLYTIC METHOD
(FOR SAMPLES C~NTAINING UP TO

0.2%

OF' LEAD)

Method.-In a 250-ml. beaker dissolve 5 g. o,f bronze in 2& ml. of H~Oa

(sp.gr. 1.42). Allow the solution to bake dry on the edge of a hot plate or by
standing over night on the steam bath. Add 17 ml. of HNO a (sp.gr. 1.42), and
after digesting for a short time, add 85 ml. of hot water and heat to boiling
for a few minutes. Allow the solution to stand for several hours at a temperature just below the boiling point. Filter off the precipitated H 2 SnOa on double
closely woven filter papers, being careful to keep the solution hot thtoughout
the process of filtration. Wash with boiling hot water. Receive the combined
filtrate and washings in a 250-ml. beaker. Dilute to about 150 ml. with distilled water and insert electrodes. In the positive terminal use a sand-blasted
platinum gauze electrode such as is used for the-cathode in the determination
of. copper; in the negative terminal insert an electrode such as is used for the
anode in the determination of coppe'r. Cover with a pair of split watch glasses
and electrolyze. For each solution use a current density of 1.25 to 1.5 amp. _per
sq. dm. (11.25 to 13.5 amp. per sq. ft.) for one hour. Wash down the cover
glasses, electrodes and sides of beaker, raising the level of the liquid slightly, and
continue the electrolysis for 15 minutes. If no darkening of the newly exposed
surface of the platinum can be detected the lead has been entirely deposited.
Continue the electrolysis until no darkening of the platinum can be detected
when the current has been passed for 15 minutes after the level of the liquid
has been raised.
!
When the lead has been entirely deposited, r~move the anode quickly, rinse
it in distilled water and then in alcohol, and dry it at 210 C. for one-half
hour. Weigh as Pb0 2 , using the factor for lead 0.8643 instead of the theoretical value.
Accuracy.:-Duplicate determinations should check within 0.01% of lead.
Results obtained by,this method are likely to be about 0.Q1 % too high. For.
exceptionally accurate or investigation work redissolve the coating from the
electrode and determine the lead as sulfate. See the" Determination of Lead
as Sulfate."
NOTES.,-l. It is impossible to remove the tin quantitatively as H 2Sn03 in the case
of alloys containing over 0.25% of iron, and traces of tin are sometimes found in the
'Pb0 2 deposit. Small amounts of manganese also :are occasionally deposited with the
PhOz. Tin and manganese are rarely present in the Ph0 2 in amounts which necessitate
a purification of the deposit.
2. In this and the following method a slight amount of lead may be retained by the
H 2Sn0 3 This should be recovered in very accurate work, ~ut is neglected in routine
analysis.
;3. See note 4 under "Determination of Cop'per'by the,Electrolytic Method;"
4. See note 2 under" Determination of Lead oy the Electrolytic Method, Analysis of
Bronze Bearing Metal."

ANALYSIS OF

NON-F~RROUS

ALLOYS

1355

DETERMINATION OF LEAD AS SULFATE

(FOR SAMPLES CONTAINING UP TO

0.2%

OF LEAD) (OPTIONAL)

Solutions Required. Lead Acid.-Mix 300 m!. of H 2S0 4 (sp.gr. 1.84), and
1800ml. of distilled water. Dissolve 1 g. of lead acetate, c.p.,\in 300 ml. of
distilled water and add this to the hot solution, stirring meanwhile. Let
stand at lea~t 24 ~ours and siphon through a thick asbestos filter.
Dilute Alcohol for Washing.-Mix equal parts of denatured alcohol and
distilled water.
Method.-Proceed exactly as in " Determination of Lead by the Electrolytic Method" up to the point where the precipitated H 2SnOa is filtered and
wash~d.

.
To the combined filtrate and washings add 120 m!. of lead acid, and evaporate until copious fumes of H 2S0 4 are evolved. Cool, add 105 ml. of distilled.
water to dissolve the salts and to make the acid concentration the same as
in lead acid, heat to boiling, aud allow to cool and settle for 5 hours, or over
night if convenient. Filter on a weighed porcelain Gooch crucible and wash
with lead acid. Wash out the lead acid with dilute alcohol, set the gooch
crucible inside a porcelain crucible, dry and ignite for 5 minutes at the full
heat of a Tirrill burner, cool and weigh as PbS0 4 , which contains 68.33% lead.
Accuracy.-Duplicate determinations should check within 0.01% of lead.
Results obtained by this method are likely to be about 0.01 % too low.
NOTEs.-l. It is impossible toremove the tin quantitatively as H 2SnO, in the case of
alloys containing over 0.25% of iron, but by the preliminary baking and long standing,
most of the H 2SnO, is precipitated.
2. See Note 4 under" Determination of Lead by the Electrolytic Method."
3. For the aetermination of lead as a sulfate in alloys containing over 0.2% see,
"Analysis of Bronze Bearing Metal."
DETERMINATION OF COPPER AND LEAD SIMULTANEOUSLY
0.2% OF LEAD)
Method.-In a 150-m!. beaker, dissolve 1 g. of brass in 10 ml. of HNO a
(sp.gr. 1.42). When the action has ceased, evaporate the solution to dryness,
and bake thoroughly on the edge of a hot plate. Add 10 m!. of HNO, (sp.gr.
1.42), and after digesting for Ii short time add 50 m!. of hot distilled water, heat
to boiling, and allow to stand and settle for about one hour, keeping the temperature just below the boiling point. Filter on double 7 cm. filter paper,
keeping the solutionhot, and receive the filtrate in a 200-ml. electrolysis beaker.
Wash with boiling water, dilute to about 100 m!. and insert electrodes. In
the positiye terminal insert one of the platinum gauze anodes previously
described, and in the negative terminal insert a platinum gauze cathode.
Cover with a pair of split watch glasses and electrolyze. For each solution use
a current density of 3 to 5 amp. per sq. dm. (27 to 45 amp. per sq. ft.). After
about 45 minutes, the lead will have been entirely deposited on the anode as
Pb0 2 ; without interrupting the current add to the electrolyte 3 to 4 m!. of
H 2S0 4 esp.gr. 1.84), and continue the electrolysis. When the solution is colorless, wash down the cover glasses, electrodes and sides of the beaker, raising
the level of the liquid slightly, and continue the passage of the current for
(FOR SAMPLES CONTAINING OVER

1356

ANALYSIS OF NON-FERROUS ALLOYS

about 15 minutes, noting whether or llQt copper is deposited on the newly

exposed surface of the platinum. If no copper ~ppears, transfer about 1 m!.
of the colorless solution to the cavity of a porcelain test plate, and ,add a few
drops of fresh H 2S water. If the slightest disco~oration occurs, continue the~
electrolysis until there is no discoloration whatever upon repeatil).g the test.
Remove the solution from the electrodes quickly without interrupting the
current. Rinse the electrodes .in distilled water and then dip them in t\vo'
successive baths of alcohol. Dry the anode at 210 C. for 30 minutes. Weigh
as Pb0 2, using the factor for lead 0.8643 instead of the theoretical value.
Shake off the excess alcohol from the cathode and ignite' the remainder... by
~ringing it to the flame of an alcohol lamp. Keep the cathode moving continually while the alcohol burns. Weigh as metallic copper.
Accuracy.-Duplicate copper determinations should check within 0.10%
.of copper. Duplicate lead determinations should check within 0.02% of lead.
Lead results obtained by this method are likely to be about 0.01 % too high.
See notes under the " Determination of Lead by the Electrolytic Method."
DETERMINATION OF TIN
Solutions Required. Dilute Hydrochloric Acid.-Mix 500 m!. of HCI
Csp.gr. 1.18), and 500 m!. of distilled water.
.
Dilute Sulfuric Acid.-Mix 300 ml. of H 2S0 4 Csp.gr. 1.84), and\600 mL of
distilled water.
Dilute Ammonia.-Mix 100 m!. of NH 40H Csp.gr. 0.90), and 900 m!. of
distilled water.
; Ammonium Acetate Solution for Washing.-Dissolve 10 g. of ammonium
acetate in 300 ml. of distilled water. Make slightly acid with acetic .acid and
saturate with H 2S gas.
Method.-In a 150-ml. beaker dissolve 2 g.. of bronze in a mixture of 10 m!.
of HCI Csp.gr. l.18), and 5 ml. of HNO g Csp,gr. L42). Dilute to about 75 m!.
with distilled water. Add NH 40H Csp.gr. ,0.90)/until the basic salts of copper
have dissolved and the solution has becorp.e a deep blue. Boil and allow to
settle, and filter on a closely woven filter paper. Wash with dilute ammonia
and with hot water. Dissolve the precipititte on the filter with hot dilute HC!.
Dilute the solution to about 100 m!. Add NH 4 0H Csp.gr. 0.90) until a permanent precipitate ~orms. Heat the solution to boilillg, allow to settle, filter.
and wash as before. Dissolve the precipitate on the filter with boiling hot
dilute H 2S0 4, washing the paper very thoroughly with this a.cid. Add NH 40H
Csp.gr. 0.90) cautiously until the precipitate which forms at first dissolves
rather slowly. Allow the solution to stand for some hours and if any lead
sulfate forms filter it off. Dilute the solution to about 200 m!. and saturate
it with H 2S gas. Filter the precipitated tin sulfide off on double papers and
wash with ammonium acetate solution, retainIng the filtrate and washings for
the iron determination. Dry the precipitate and place it with the papers in a
porcelain crucible. which projects part way through a hole in a piece of asbestos
board. 'Heat slowly until any free sulfur has been driven off, but without
allowing the sulfur to burn: When the !luJfur has been expelled, place the
crucible <)n a triangle and ignite slowly' at first and finally at the full heat of the

ANALYSIS OF NON-FERROUS ALLOYS

1357

burner. If the precipitato weighs more than 20 mg. heat to constant weight,
using a blast lamp Weigh as Sn02, which contains 78.77% of tin.
Accuracy.-Duplicate determinations should check within 0.06% of tin.
NOTEs.-l. If, during ignition, the sulfur is allowed to burn, some tin sulfate may
be formed, causing high results. On the other hand, low results may be caused by
too high heat, which causes Epattering and volatilizes some tin sulfide.
2. Tin and iron are best determined on the same sample, retaining the filtrate from
the tin sulfide precipitate to use for the iron determination.
3. For the determination of tin in alloys containing more tin than is present in this
.
alloy see" Analysis of Bronze Bearing Metal."
4. In alloys containing an appreciable amount of antimony and in the presence of a
large excess of tin, the antimony as determined by the method given under "Analysis
of B,-onze Bearing Metal" should be calculated to Sb 20. and a corresponding amount
deducted from the weight of the ignited Sn02 before calculating to tin. Similarly
arsenic found in the alloy should be calculated to As 20 S and a corresponding correction
made.
DETERMINATION OF IRON
'Solutions Required.-See " Determination of Tin," and the following:
Dilute Sulfuric Acid for Reductor.-Mix 50 m!. of H 2S0 4 (sp.gr. 1.84),
and 1000 m!. of distilled water. This solution is used boiling hot.
Potassium Permanganate.-Dissolve 0.2 g. of KMn04 in water, filter
through asbestos, and dilute to 1000 m!. with distilled water. Standardize
against 0.020 g. portions of pure sodium oxalate. Each milliliter is equivalent
to approximately 0.00033 g. of iron.
.
,
Method.-Proceed exactly as in the" Determination of Tin" through the
filtration and washing of the precipitated tin sulfide. Combine the filtrate
and washings from the tin sulfide precipitate and boil until all H 2S is expelled.
Add HNO a (sp.gr. 1 42) until the iron is oxidized, which is shown by the solution becoming clear and of a yellowish color. Add about 5 g. of NH 4CI, then
NH.OH (sp.gr. 0.90) until a permanent precipitate forms. Boil, allow to
settle, filter on a loosely woven filter paper and wash with dilute ammonia
and with hot water. Dissolve the precipitate on the paper with hot dilute
Hel and add NH 40H (sp.gr. 0.90) to the solution again until a permanent
precipitate forms. Boil, allow to settle, filter on a loosely woven filter paper,
and wash with dilute ammonia and with hot water until all NH 4 CI is removed.
Dissolve the ,precipitate and wash the filter paper with 100 m!. of dilute H 2S0 4
for reductor. Pass the solution through a Jones reductor, wash first with
150 m!. of. the H 2S0 4 for reductor and then with 100 m!. of distilled water.
Titrate with the stand!!-rd KMn04 solution.
Accuracy.-Duplicate determinations should check within 0.03% of iron.
NOTEs.-l. A blank determination should be made on corresponding amounts of acid
and water passed through the reductor and the results should be corrected accordingly.
About 0.5 ml. of the permanganate will be required to give a permanent coloration to
the solution.
2. A small quantity of liquid should always be left in the reductor funnel, and air
should never be allowed to enter the body of the reductor.
3. A description and further details of use of reductor see "Estimation by Reduction
with Jones Reductor and Oxidation by Standard Permanganate Solution" and i, Notes on
Reductor Technique" in chapter under Molybdenum.
.
4. For the determination of iron in alI'oys containing more tin than is present in thil;l
alloy see" Analysis of Bronl!!e Bearing Metal."

1358

ANALYSIS OF NON-FERROUS ALLOYS

DETERMINATION OF MANGANESE B'Y THE PERSULFATE

METHOD
Solutions Required.-Solution for Dissolving. Mix 500 m!. of H 2S0 4
(sp.gr. 1.84), 200 m!. of HNO a (sp.gr. 1.42) and ~700 m!. of distilled water.
Silver Nitrate.-Dissolve 1.33 g. of AgNO a in, 1000 m!. ()f distilled water.
Stock Sodium Arsenite Solution.-To 15 g. of arsenious oxide (As 20 a) in
a 300-m!. Erlenmeyer flask add 45 g. of Na 2 CO a and 150 m!. of distilled. water.
Heat the flask and contents on the steam bath until the As 20 a is dissolved ...
Cool the solution, filter and make up to 1000 m!. with distilled' water.
Standard Sodium Arsenite Solution.-':"Mix 200 m!. of the stock sodium
arsenite solution with 2500 m!. of distilled water,. and standardize against a ,
steel or iron of known manganese content as determined by the Bismuthate
Method, or standardize against one of the Bureau of Standards'.' standar.d
irons or steels. One milliliter of this solution should be equivalent to approximately 0.00035 g. of manganese.
Ammonium Persulfate.-Dissolve 6 g. of ammonium persulfate in 100 m~.
of distilled water. (This solution should be made up as needed as it deteriorates r a p i d l y . ) .
I
Method.-Adjust the size of the sample to the expected percentage so that
the portion taken does not contain more than 0.002 g. of manganese. ~lace
it in a 250-m!. Erlenmeyer flask with 24 ml. of the" solution for dissolving"
and heat until solution is complete; Boil until the' oxides of nitrogen are
expelled. Dilute to 100 m!. with boiling hot water, add 20 ml. of AgNO a
solution and 30 m!. of the persulfate solution, then remove to a moderately
warm place to stand for a few minutes for com,plete development of the permanganic acid. Cool to below 25 C., and titrate with the standard arsenite
.
solution to the disappearance of the pink color.
Accuracy.-Duplicate determinations should check within. 0.02% of manganese.
NOTES.-1. Large amounts of permanganic acid fare unstable; on that account
samples which contain large amounts of manganese sh'ould have correspondingly small
samples weighed out. For instance, with metal containing 1.5% of manganese use a
0.10 g. sample, while if metal contains 0.75% manganese use a 0.20 g. sample.
2. The solution mQ.st be cold when it is titrated, for high temperatures cause low
results to be obtained.

DETERMINATION OF MANGANESE BY THE BISMUTHATE

METHOD (OPTIONAL)
Solutions Required. Nitric Acid (1 : J-).-Mix 250 m]. of HNO a (sp.gr.
1.42). and 750 m!. of distilled water. .
. .
Nitric Acid for Washing.-Mix 30 m!. of HNO a (sp.gr. 1.42), and 970 m!.
of distilled water.
'
Ferrous Ammonium Sulfate.-Dissolve 12.4 g. of ferrous ammonium sulfate crystals in 950 ml. of distilled water, and add 50 ml of H 2S0 4 (sp.gr. 1.84).
Potassium Permanganate.-Dissolve 1 g. of KMn04,in 1000 m!. of distilled
water. Allow it to stand for about one week aud' then filter it through asbestos. Standardize against 0.10 g. portions of pure sodIum oxalate.

ANALYSIS OF NON-FERROUS ALLOYS

1359

Method.-In a 250-ml. Erlenrn,eyer flask dissolve 1 g. of bronze in 50 ml.

of HNOa (1 : 3). Without filtering off the H 2Sn03, cool and add 0.5 g. of
sodium bismuthate. Heat for a few minutes until the .purple color has disappeared, with or without the precipitation of manganese dioxide. Add a little
ferrous~ammonium sulfate solution until the solution becomes clear, and boil
until the oxides of nitrogen are expelled. Cool, add an excess of sodium bismuthate and agitate for a few minutes. Add 50 ml. of HNO a for washing and
filter through an alundum crucible or an asbestos pad. Wash with 50 ml. of
the HNO a for washing. Add from a pipette or a burette 10 ml. of ferrousammonium sulfate solution and titrate with KMn04. Owing to the" presence
of considerable copper, the end point is somewhat different from the normal
.
pink color. . .
In exactly the same manner carry through a blank determination, using the
same amounts of HNO a and sodium bismuthate as was done with the regular
sample. Finally add exactly 10 m!. of ferrous-ammonium sulfate solution and
titrate with KMn04. The difference between the two titrations is due to the
manganese. Since one manganese as permanganate oxidized five irons, the
iron value of the permanganate multiplied by the factor (Mn/5Fe) (or 0.1967)
gives the value in terms of manganese~
Accuracy.-Duplicate determinations should check within 0.02% of manganese.
NOTES.-l. Instead of employing the method of reducing the permanganic acid by
means of standardized ferrous ammonium sulfate solution and titrating the excess of
this reagent, it is possible to reduce the permanganic acid by standard sodium arsenite
solution. See "Determination of Manganese by the Persulfate Method."
2. The filtrate from the bismutnate must be perfectly clear, as the least particle of
bismuthate carried into the filtrate will vitiate the results.
DETERMINATION OF ALUMINUM IN SMALL QUANTITIES

(a) General Method for Accurate Work

Solutions Required. "Electrolytic Solution."- Mix 300 ml. of HNOa
(sp.gr. f.42), 1700 m!. of distilled water and 500 ml. of H 2S0 4 (sp.gr. 1.84).
Acidified H 2 S Solution for Washing.-Dilute 10 ml. of HCI (sp.gr. 1.18)
with 1000 ml. of distilled water and saturate with H 2S.
Phenylhydrazine Solution Jor Washing.-Add saturated 802 water to a
few ml. of phenylhydrazine until the crystalline sulfite first formed redissolves
and then add ph~nylhydrazine drop by drop with vigorous agitation until the
odor of S02 is no longer perceptible; dilute 5-10 ml. of this solution with 100
ml. of hot water and allow to remain at the boiling point for a few minutes to
get rid of the excess S02.
2% NH 4 CI Solution for Washing.-Mix 30 ml. of HCI (sp.gr. 1.18) with
200 m!. of distilled water, add methyl red, neutralize with NH 4 0H until the
solution changes to a distinct yellow and then dilut~ to 1000 mI. with water.
Method.-Dissolve 1 g. of the brass or bronze in 35 ml. of " electrolytic
solution," dilute with 100 ml. of distilled water and remove copper by electrolysis. Evaporate the electrolyte to fumes of SOa. (I~ much tin is present,
7

A. S. T. M. Methods of Chemical Analysis of Metals, p. 170, December, 1936.

1360

ANALYSIS OF NON-FERROUS ALLOYS

boil the sulfuric acid solution vigorously for a few minutes.) Cool and dilute
with 100 ml. of water. Filter off the PbS0 4 and aJ)y tin that may be precipItated at this point, washing with hot water. Dilute the filtrate to at least
200 ml.
Pass H 2S into the filtrate for 30 minutes.. This will precipitate any copper,
lead, tin, arsenic or antimony which may be presept. Allow the sulficles to
settle and filter, washing with acidified H 2S water. Boil the filtrate to onehalf volume to expel H 2S, add 3 ml. HNOa (sp.gr. 1.42) and boil for a few
minutes to oxidize the iron.
Precipitate the iron and aluminum according to the following (Blum,
J. Am. Chem. Soc., 38, 1282, 191.6). Add 10 ml. of HCI (sp.gr. 1.18) and
just neutralize with NH 4 0H, using methyl red as indicator. Boil for tw.o
minutes and filter, washing with hot 2% NH 4 Cl. Dissolve the precipitate
from the paper with boiling hot HCI (1 : 1) and wash alternately with hot
water and HCI (1 : 1) into the beaker in which the precipitation was made.
Reprecipitate the iron and aluminum following the same procedure as before.
Dissolve the precipitate again with hot HCI (1 : 1), dilute to 100 ml. with
water, make just ammoniacal, then only add acid enough to hold iro~ in
solution. Add sufficient saturated solution of ammonium bisulfite to reduce
the iron (5 to 20 drops). Add a few drops of methyl orange, quickly bring ,to
neutrality with NH 40H and then add 6 to 7 drops of HCl (1 : 1) in excess.
Finally add from 1 to 3 ml. of phenylhydrazine and stir until the precipitate
becomes flaky. The supernatant liqu6r should be plainly acid to litmus and
the precipitate may be colored owing to organic matter. Filter and wash with
phenylhydrazine wash solution. Ignite in a weighed platinum crucible, finishing with the blast for five minutes. Weigh as AI 20 a. The weighing should
be made rapidly and with the crucible covered as Al 20 a is somewhat hygro
f
SCOpIC.
NOTES.-1. For small amounts of aluminum a 5 g. sample should be used. In this
case dissolve in 60 ml. of "as~ay solution" of the followil).g proportions: 7 vols. HN0 3
(sp.gr. 1.42). 10 vols. H,S04 (sp.gr. 1.84), 25 vols. H,O. .
2. When phosphorus is present, the final precipitatd will be AI,03 plus P,Os. In
this case it is necessary to fuse the residue with Na;CO" dissolve with hot water, acidify
with HNO, and determine phosphorus by the alkalimetric method. The phosphorus is
calculated to P,Os and subtracted.
3. For accurate work the AI 20, residue is treated with HF and H 2S0 4 for the removal
of possible Si0 2
.
,
4. Ammonium bisulfite solution may be prepared by saturating cold NH 40H (1 : 1)
witjl SO, until yellow.
..
5. This method is especially satisfactory for small amounts of aluminum in brass
or bronze since no reagents are used which may contaminate the precipitate with aluminum, iron or !'ilica. A precipitate with phenylhydrazine gives positive indication of
the presence of aluminum.
'

(b) Rapid Method for Control Work

Solutions Required. 2.5% Sodium dydroxide Solution.-Dissolve 2.5 g.
NaOH (free from aluminum) in 1000 ml. water.
.
Sodium Sulfide Solution.-Dissolve 150 g. NaOH (free/from aluminum) in
1000 m!. of water, saturate 500 m!. of this solu!JOll with H 2S and mix the remaining 500 m!. of solution with it.
.

ANALYSIS OF NON-FERROUS ALLOYS

1361

Dilute Hydrochloric Acid (1 : 3).-Mix 200 mi. of HCI (sp.gr. 1.18) with
600 mi. of water.

Acidified H 2S Water.-Dilute 10 mi. of HCI (sp.gr. 1.18) with 1000 mi. of

water and saturate the solution with H 2S.
2% NH4 CI Solution.-Mix 30 mi. of HCI (sp.gr. 1.18) with 200 mi. of
water, add methyl red, neutralize with ammonia until the solution changes to
a distinct yellow and then dilute to 1000 mi. with water.
Method.-Dissolve 2 g. of the sample in 20 mi. of HCI (sp.gr. 1.18) and
5 mi. oLHN03 (sp.gr. 1.42). Boil the solution to expel chlorine and dilute
with 50 mi. of water. Nearly neutralize the cold solution with 2.5% NaOH
solution (or stronger if much free acid is present) and pour it slowly with
constant shaking into a 500-mi. volumetric flask containing 100 mi. of sodium
sulfide solution. Dilute to the mark with 2.5% NaOH solution and mix thoroughly. Filter on a large, dry, No. 42 Whatman filter paper or its equivalent
and keep the paper well filled with the solution lest iron be oxidized and
dissolved. Reject the first 20-25 mi. and catch exactly 250 mi. which will
represent 1 g. of sample. In routine analysis of material containing not
more than 1 % of aluminum the. aliquot portion may be gathered in a 250-ml.
graduate.
Transfer the aliquot portion to a 600-m!. beaker, neutralize the solution
with dilute HCI (1 : 3) and finally add 25 mi. excess. Digest at 40-60 C.
for one hour and filter and wash the paper and precipitate with acidified H 2S
water.
Boil the filtrate and washings to expel H 2S, add a few drops of methyl
red indicator and then dilute ammonia (1 : 2) until the solution is just distinctly yellow. Boil for two minutes and filter at once through a small filter.
Wash the beaker, paper and precipitate two or three times with hot 2% NH 4 CI
solution and discard the filtrate. Dissolve the precipitate in 20 mi. of hot
dilute HCI (1 : 3), wash the filter thoroughly with small portions of water and
reserve it for the second filtration. Dilute the filtrate to 50 mI., add methyl
red and precipitate with dilute ammonia as before. Filter, wash with hot 2%
NH 4CI solution and ignite and weigh.
The ignited residue is prone to carry Si0 2 and mU$t be purified before
weighing as follows. Add one or two drops of water, one drop of diluted
H 2S0 4 and 1 to 5 ml. HF, Evaporate to dryness, 'increase the heat slowly
and finally heat with a blast lamp or its equivalent. Weigh as Al 20 3. The
weight of A1 20 3 corrected for the blank and multiplied by 52.91 gives the per
cent of aluminum.
NOTEs.-The first Al(OHl, precipitate will carry down some NaCl and much of any
silica that wa~ dissolved in the NaOH solution: hence it is not safe to omit the second
precipitation and HF treatment. There is no attack on the glassware during the short
contact with the weak cool alkaline solution. It is, of cour"e, preferable that the NaOH be
free from aluminum. The most satisfactory test for it lies in running an analysis with
non-ferrous alloys containing no AI. Direct test, by acidification followed by precipitation. with ammonia is sufficient if no precipitate is obtained, but is of doubtful value in
case one appears, for it may contain other elements, such as iron, which do not affect the
results.
If phosphorus is known to be present the method should be followed through the first
addition of ammonia. If aluminum is indicated, re-acidify the solution with HCl, add
macerated paper, two drops of methyl orange and 10 ml. of a solution of diammonium

1362

ANALYSIS OF NON-FERROUS ALLOYS

phosphate (10%). Render the solution just ammonIacal, then just restore the pink color
with dilute Hel (1 : 3), heat to boiling and add 30 mi. of 25% ammonium acetate. Boil
for five minutes, filter on an 11 cm. No. 42 Whatm!j.n paper lind wash with hot 5%
NH 4NO. solution until 5 ml. of the washings no longer give a test for chlorides. Ignite
in platinum or porcelain, heat at about 1100 C. for JO minutes and weigh as AlP0 4
This method is not strictly accurate on account of t,he uncertain composition of the
phosphate but is sufficiently accurate for all but the most painstaking analyses.
.,
0

ANALYSIS OF BRONZE BEARING METALS

General Considerations.-The determinations given below cover most of

the constituents commonly found in a bronze.
,
For the determination of manganese and aluminum see" Analysis of Manganese Bronze."
DETERMINATION OF COPPER BY THE ELECTROLYTIC METHOD
Method.-I.n a 150-m!. beaker,' dissolve 1 g. of the bronze in 10 m!. of
HNO a (sp.gr. 1.42). When the action has ceased, boil until the oxides of
nitrogen are expelled, add 50 m!. of hot distilled water and proceed as in
" Determination of Copper by the Electrolytic Method, Analysis of Manganese Bronze."
Retain the papers containing the precipitate of H 2S.nOa for subsequent use
in the tin determination.
. ,
NOTEs.-l. Copper, zinc and tin are conveniently determined on the same sample,
tin being first removed, and any zinc tested for aft~{. copper has been deposited by
electrolysis.
,
2. If the solution becomes cool during the :qrst filtration, some of the H 2SnO, may
dissolve.
"
3. See note 3 under" Determination of Copper by the Electrolytic Method, Analysis
of Manganese Bronze."
,
4. For alloys containing over 0.25% of iron it is advisable to remove the tin as
H 2SnO, by a preliminary baking. See" Determination of Capper Analysis of Manganese
Bronze."
, " , ,
DETERMINATION OF TIN
(FOR SAMPLES FREE FROM PHOSPHORUS)
Method.-See the" Determination of Copper by the Electrolytic Method."
Use the papers containing the meta-stannic acid which was filtered off in
the "Determination of Copper by the Electrolytic Method." Ignite the
papers with the precipitate while moist in a porcelain or platinum crucible,
slowly at first, and finally at the full heat of the burtler. If the precipitate
8 A. S. T. M. Methods of Chemical Analysis '()f'Metals, p. 157, December, 19;36.

ANALYSIS OF NON-FERROUS ALLOYS

weighs more than 20 mg., heat to constant weig!lt, using a
as Sn02, and calculate to Sn.

bl~st

lamp.

1363
Weigh

NOTES.-l. This method is not accurate for. alloys which contain phosphorus, or
which contain over 0.25% of iron. See" Determination of Tin, Analysis of Manganese
Bronze."
2. The ignited tin oxide contains small amounts of iron and copper as well as any
phosphorus, antimony or arsenic which may be present in the alloy. Qualitative tests
should be made for phosphorus, antimony and arsenic. In the absence of antimony
and arsenic the following" Qualitative Test for Phosphorus" is here made use of to detect
- phosphorus and to estimate its amount, if any is found.

QUALITATIVE TEST FOR PHOSPHORUS

Solutions Required. Ferric Chloride.-Dissolve 25 g. of ferric chloride in
100 ml. of distilled water, and add 25 ml. of HCI (sp.gr. 1.18).
Method.-Dip a piece of the alloy into a few milliliters of ferric chloride
for about ten seconds, and rinse it in running water. Alloys containing
phosphorus are darkened noticeably where they have been dipped; alloys
containing over 0.25% of phosphorus are rendered almost black by this test.
In alloys containing up to 0.20% of phosphorus, a correction for its presence
in the ignited tin oxide may be made according to the following empirical
method. From the phosphorus content of the alloy, which has been determined on a separate sample, calculate the weight of P 20 li and subtract twothirds of this weight from the weight of the ignited tin. oxide cofltaining phosphorus. The remainder is considered as pure Sn02. The factor for converting
phosphorus to % X P 20 5 is 152.6%, or approximately 172 times the phosphorus
content of the sample.
If the alloy contains over 0.20% of phosphorus, a special method for the
determination of tin should be used.
In alloys which contain over 0.25% ,of iron, use the method for the" Determination of Tin, Analysis df Manganese Bronze."
DETERMINATION OF TIN FOR SAMPLES CONTAINING
PHOSPHORUS
Solutions Required. Yellow Ammonium Sulfide.-Saturate 150 m!. of
NH 4 0H (sp.gr. 0.90) with H 2S gas, and then add 50 m!. more of NH 4 0H.
Dissolve in this solution 3 or 4 g. of sulfur flour, and about 1 g. of NH 4 C!.
Make up this solution fresh, and filter immediately before using.
Dilute Yellow Ammonium Sulfide for Washing.-Mix 20 m!. of yellow
ammonium sulfide, prepared as above, and 400 m!. of distilled water.
Ammonium Acetate for Washing.-Dissolve 10 g. of ammonium acetate in
300 m!. of distilled water, make slightly acid -with acetic acid and saturate
with H 2S gas.
Method.-See the" Determination of Copper by the Electrolytic Method."
Use the papers containing, the freshly filtered meta-stannic acid which
was filtered off in the" Determination of Copper by the Electrolytic Method."
Transfer the papers with the precipitate to a 150-mI. beaker, and cover with
40 to 50 m!. of yellow ammonium sulfide. Warm for about 15 minutes, or

1364

ANALYSIS OF NON-FERROUS ALLOYS

until the meta-stannic acid has dissolved. Filter and wash thoroughly with
dilute yellow ammonium sulfide. Acidify the cRmbined filtrate and washings
cautiously with 50% acetic acid. Warm, and allow the precipitated tin sulfide
and sulfur to settle. Filter on double papers I and wash with ammonium
acetate solution. . Dry the precipitate, a~d proceed exactly as in " Determination of Tin, Analysis of Manganese Bronze." I
NOTEs.-l. The ignited tin oxide will contain any arsenic or antimony that was"
present in the sample. For its correction see Note 4, "Determination of Tin," Analysis
of Manganese Bronze.
2. In the case of alloys containing smaller amounts of tin and over 0.25% of iron
proceed as in "Determination of Tin, Analysis of Manganese Bronze."
DETERMINATION OF LEAD BY THE ELECTROLYTIC METHOD
Method.-Proceed exactly as in " Determination of Copper by the Electrolytic Method" up to the point where the tin has been separated as H 2SnO .
To the filtrate add 9 ml. of HNO. (sp.gr. 1.42), dilute to a volume of about
150 ml. and electrolyze for 2 hours, using a current of 2 amperes. See" Determination of Lead by the Electrolytic Method, Analysis of Manganese
Bronze."
NOTEs.-l. See Notes 1 and 2 under "Determination of Lead by the Electrolytic
Method, Analysis of Manganese Bronze.~!
2. Samples/or Analysis.-When lead is less than 0.10%, a 5 or 10 g. sample is taken.
The 5 g. sample should be dissolved in 25 m!. of RNO.. The 10 g. sample should be
dissolved in 45 m!. of RNO.. When the lead content is greater than 15% dissolve the
sample in RNO. as described under Method, transfer the solution to a graduated fla~k
and choose aliquot portions for analysis in order not to exceed the above mentioned limit
of lead to be deposited on the anode as Pb0 2 The lead peroxide will tend to flake off
the electrode when the amount deposited becomes excessive.
DETERMINATION OF LEAD M..S SULFATE
.
(OPTIONAL) I
Solutions Required.-See " Determination of Lead as Sulfate, Analysis bf
Manganese Bronze."
Method.-In a 150-m!. beaker dissolve'l g. of bronze in 10 ml. of HNO.
(sp.gr. 1.42). When dissolved, boil until the oxides 'of nitrogen are expeilec),
add 50 ml. of boiling water, allow to stand, filter and wa~h 'with hot water.
(For' further particulars see" Determination of Copper, Analysis of Manganese Bronze.")
.
To the filtrate add 40 ml. of lead acid and evaporate to fumes of H 2S0 4
Cool, add 35 ml. of water to dissolve the salts, heat to boiling, allow to cool
and settle for 5 hours and proceed as in " Determination of Lead as Sulfate,
Analysis of Manganese Bronze."
,
NOTEs.-1. Copper and lead may be determined on the same samples if so desired.
In such a case, after filtering off thfl lean sulfate and washing with lead acid, and before
washing with dilute alcohol, remove the filtrate, add 3 m!. of RNO. (sp.gr. 1.42), dilute
to 120 mI. and determine copPflr by electrolysis. See.the" Determination of Copper by
the Electrolytic Method." ,
-'

ANALYSIS OF NON-FERROUS ALLOYS

1365

2. Copper and lead may also be determined simultaneously by the electrolytic

method. ee" AJ1Il.lysis of Manganese Bronze."
\
3. For amounts of lead less than 0.2% proceed as in (. Determination of Lead as
Sulfate, Analysis of Manganese Bronze."

DETERMINATION OF IRON
'Solutions Required. Yellow Ammonium Sulfide and Dilute Yellow Ammonium Sulfide for Washing.-See " Determination of Tin."
Other Solutions.-See " Determination of Iron, Analysis of Manganese
Bronze."

Method.-Proceed exactly as in " Determination of Copper" up to thepoint where the tin has been separated as H 2Sn03. '
Retain both the filtrate and the precipitate. Transfer the papers with the
precipitate to a ISO-ml. beaker, and cover them with 40 to 50 ml. of yellow
ammonium sulfide. Warm about 15 minutes, or until the meta-stannic acid
is dissolved. Filter and wash thoroughly with hot, dilute yellow ammonium
sulfi-de.
Dissolve the small black residue on the filter with hot dilute HCI, receiving
the solution in a small beaker. Add 5 ml. of- HN0 3 (sp.gr. 1.42) and boil to
oxidize sulfur and iron. Add this small solution to the original filtrate from
the meta-stannic acid. Add about 5 g. of NH 4 Cl, then NH 4 0H (sp.gr. 0.90)
and proceed as in the" Determination of Iron, Analysis of Manganese Bronze."
DETERMINATION OF NICKEL BY THE DIMETHYLGLYOXIME
'METHOD
Solutions Required. Dimethylglyoxime Solution.-Dissolve 5 g. of dimethylglyoxime in 500 ml. of grain alcohol, and filter before using.
Sodium Potassium Tartrate Solution.- Dissolve 200 g. of the salt in 1 liter
of distilled water.
Dilute Ammonia for Washing.-Solution containing 10 m!. of NH 40H
(sp.gr. 0,90) per liter of water.
Method.-In a 150-ml. electrolytic beaker, dissolve a 1 g. sample of the
bronze in 10 m!. of HN0 3 (sp.gr. 1.42) and remove tin and copper exactly as
described in the" Determination of Copper by the Electrolytic Method."
Evaporate the solution'to about 100 mI., filter, add about 1 g. of NH 4 CI, then
IJ 10 ml. of sodium potassium tartrate solution, and nearly neutralize with NH 4 0 H
(sp.gr. 0.90). Heat nearly to boiling, add 5 ml. of dimethylglyoxime solution
for each 0.01 g. or fraction of nickel present and then NH 4 0H in slight excess.
Let stand at least ,30 minutes on the steam bath, filter and wash with dilute
ammonia solution. For accurate work, the precipitate should be dissolved in
hot HCI (1 :'1) and reprecipitated as before. Filter on a weighed Gooch
crucible, wash with dilute ammonia solution, and then with hot water; dry
at 110 to 120 0 C. for about two hours and weigh as the glyoxime salt which
contains 20.32% of nickel. If it is preferred, instead of by weighing the glyoxime salt, the nickel can be determined electrolytically as follows: Dissolve
the red precipitate of nickel glyoxime in hot HCI (1 : 1), add 10 ml. of H 2S0 4
(1 : 1) and evaporate to strong fumes. Dilute, neutralize with NH 40H (sp.gr.

1366

ANALYSIS OF NON-FERROUS ALLOYS

0.90), and add 20 ml. in excess. The final volume should be about 150 ml:
Electrolyze with a current of 0.3 ampere per solution. Test the solution for'
completion of deposition with fresh H 2S water as is done in the (( Determination of Copper by the Electrolytic Method." %move the cathode, wash with
water and dry with alcohol as in the" Determination of Copper by the Electrolytic Method." Weigh the deposit as metallic, nickel.
DETERMINATION OF ZINC
After the copper has been. completely dep0sited by electrolysis, quickly
remove'the cathode and wash it thoroughly with distilled water, catching
the washings in the beaker containing the electrolyte. Heat the solution and
saturate with H 2S gas. Filter off any precipitate, and boil the solution to
expel H 2S. Make the solution barely alkaline with NH 40H, and add 25 ml.
of 85% formic acid. Dilute the solution with distilled water to 300 mI., heat'
to boiling, and saturate with H 2S gas. Filter, wash with hot water and dissolve the precipitate with hot HCI (1 : 1) and transfer the solution to a weighed
platinum or porcelain dish or crucible. 'Add a few drops of H 2S0 4 (sp.gr. 1.84)
and evaporate the solution until copious fumes escape. If the solution is not
clear and colorless, cool, add a few milliliters of HNO a (sp.gr. 1.42) and again
evaporate the solution until fumes of H 2S0 4 come off freely. Repeat the
treatment with HNO a, if necessaryvuntil the organic matter is destroyed, and
the solution is colorless. Remove the excess of H 2S0 4 by heating the dish
cautiously, and finally heat to dull redness. Weigh as ZnS04 and calculate
to Zn.
DETERMINATION OF PHOSPHORUS BY THE ALKALIlYIETRIC
METHOD

Solutions Required. Nitric Acid for Dissolving.-Mix 1000 m!. of HNO a

(sp.gr. 1.42) and 1200 ml. of distilled water.
/
Nitric Acid for Washing.-Mix 20 m!. of HNO a (sp.gr. 1.42) and 1000 ml.
of distilled water.
Ammonium Moll/bdate Solution No. i.-Place in a beaker 100 g. of 85%
molybdic acid, mix it thoroughly with 2{0 ml. of distilled water, add 140 m!.
of NH 40H (sp.gr. 0.90), filter and add 6Q ml. of HNO a (sp.gr; 1.42).
Ammonium Molybdate Solution No. 2.-Mix 400-mL._of HNO a esp.gr. 1.42)
and 960 ml. of distilled water.
.
\When the solutions are cold, add solution No. 1 to solution No.2, stirring
constantly; then add 0.1 g. of ammonium phosphate dissolved in 10 mI. of
distilled water and let stand at least 24 hOurs before using.
'
t
Potassium Nitrate, 1% Solution.-Dissolve 10 g. of potassium nitrate in
'.
1000 mI. of distilled water.
Phenolphthalein Indicator.-Dissolve 0.2 'g. in 50 m!. of 95% ethyl alcohol
and 50 ml. of distilled water.
.
Standard Sodium Hydroxide.-Dissolve 3.3 g. of purified NaOH in 1000 ml.
of distilled water, add a slight excess of 1% soluti9n of barium hydroxide,
let stand for 24 hours, decant the liquid, .a.ud-stll.lldardize it against a steel of
known phosphorus content as determined by the molybdate-magnesia method,

ANALYSIS OF NON-FERROUS

ALLOY~

1367

so that 1 ml. will be equivalent to 0.01 % of phosphorus on the basis of a 1 g.

sample (see Notes). Protect the solution from carbon dioxide with a sodalime tube.
Ferric Chloride.-Dissolve 1.00. g. of ferric chloride (phosphorus free) in
100 ml. of distilled water.
Standard Nitric Acid.-Mix 5 m!. of HN0 3 (sp.gr. 1.42) and 1000 ml. of
distilled water. Titrate the solution against standardized NaOH, using
phenolphthalein as indicator: and make it equivalent to the sodium hydroxide
by adding distilled water.
Method.-In a 400-mI. beaker, dissolve a 1 g. sample in 10 mI. of HN0 3
(sp.gr. 1.42) and heat ,until brown fumes are driven off; then add 15 m!. of
HCI (sp.gr. 1.18) and evaporate to dryness. Add 10 ml. of HCI (sp.gr. 1.18)
and again evaporate to dryness, but do not bake.
Add 15 m!. of HCI (sp.gr. 1.18) and heat to solution; dilute with 50 m!.
of distilled water and bring to a boil. Take off the hot plate and wash down
the cover glass and sides of the beaker, add 20-mesh c.p. zinc, 2 or 3 g. at a
time until the solution becomes colorless, that is, until all the copper, tin and lead are precipitated. Filter, receiving the filtrate in a 400-m!. beaker, and
wash well with hot water.
Add 3 m!. of the ferric chloride solution to this filtrate and make faintly
ammoniacal; bring to a boil, let settle and filter. Dissolve this precipitate
in hot dilute HCI and again make ammoniacal; heat to boiling, let settle and
filter, washing well with hot water.
Dissolve this precipitate in as small a quantity of hot dilute HN0 3 as
possible, receiving the solution in a 300-m!. Erlenmeyer flask; wash filter free
of iron with hot dilute HN0 3 and hot distilled water. Evaporate to about
15 mI., cool, 'make ammoniacal, then neutralize with HN0 3 (sp.gr. 1.42) and
add 5 mI. in excess. Cool to 80 C., and add 60 ml. of the ammonium molybdate solution. Let stand one minute, then shake three minutes and let settle.
Filter on an 11 cm. filter (using a small quantity of paper pulp on the filter)
and wash free of iron with 2% nitric acid wash, then free of acid with the 1 %
potassium nitrate solution.
Place the paper in the thoroughly washed acid-free flask in which the
original precipitation was made, add 25 m!. of distilled water and an excess
of the standard sodium hydroxide solution. By using a long glass rod, the
paper can be pulped and the yellow ,precipitate dissolved in a very few seconds.
Wash off the rod and the sides of the flask and add a few drops of t~e phenolphthalein solution and titrate the excess of standard NaOH with the standard
HN0 3 The strength of the. standard NaOH and the standard HN0 3 in
terms of phosphorus can be determined by using a standard steel or preferably
a standard phosphor bronze in which the phosphorus has been determined
by the molybdate-magnesia method.
.
Accuracy.-Duplicate determinations should check within 0.01 % of phosphorus.
NOTES.-l. The ammonium molybdate solution should be kept in a cool place and
should always be filtered' before using.
2. All distilled water used in titration should be freed from carbon dioxide by boiling
or otherwise.
'

1368

ANALYSIS OF NON-FERROUS ALLOYS

3. Bureau of Standards Steel No. 19 (a) or a bronze of known phosphorus content

are recommended as suitable materials for standardizatio~ of the sodium hydroxide
solution.
4. (Gravimetric method, optional.) It is consi!lered by some that more accUl;ate
results can be obtained by a gravimetric determination of phosphorus as Mg 2P 207.
This can be carried out by dissolving the yellow molybdate precipitate in 2(4 ml. of
NH 40H (1 : 1) to which has been added 2 g. of citrip acid, precipitating with magnesia
mixture, and proceeding as usual, finally weighing as Mg 2P 20 7.
DETERMINATION OF PHOSPHORUS BY THE FERRIC-ALUM
METHOD 9 (Optional)
Solutions R~quired. Ferric Chloride.-Dissolve 100 g. of ferric chloride
(phosphorus free) in 100Ill!. of distilled water.
Dilute Ammonia.-Mix 100 Ill!. of NH 4 0H .esp.gr. 0.90), and 900 m!. of
distilled water.
Dilute Hydrochloric Acid.-Mix 500 m!. of HCI (sp.gr. 1.18), and 500 IllI.
of distilled water.
Dilute Sulfuric Acid for Dissolving,-Mix 200 m!. of H 2S0 4 (sp.gr. 1.84),
and 800 m!. of distilled water.
Dilute Sulfuric Acid for Reductor.-Mix 50 mI. of H 2 S0 4 (sp.gr. 1.84),
and 1000 m!. of distilled water. This solution is used boiling hot.
Ammonium Molybdate.-Solution No. 1. Plac!l in a beaker 100 g. of 85%
molybdic acid, mix it thoroughly'with 240 m!. of distilled water, add 140 m!.
of"NH 40H (sp.gr. 0.90), filter and add 60 m!. of HN0 3 (sp.gr. 1.42).
Solution No.2. Mix 400 m!. of HN0 3 (sp.gr. 1.42), and 960 m!. of distilled water.
When the solutions are cold, add solution No.1 to solution No.2, stirring
constantly, then add 0.1 g. of ammonium phosphate dissolved in 10 m!. of
distilled water, and let stand at least 24 hours before using.
Acid Ammonium Sulfate.-Mix 25 m!. of H 2 S0 4 (sp.gr. 1.84), and 1000 m!.
of distilled water, and then add 15 m!. of NH40H (sp.gr. 0.90).
Ferric Alum.-Dissolve 200 g. of ferric aminonium sulfate crystals in 1950
m!. of distilled water. Add 50 m!. of H 2S0 4'(sp.gr. 1.84), and 80 m!. of phos'ph<l{ic acid, 85%.
I
Potassium Permanganate.-Dissolve from 3.0 to 3.2 g. of KMn04 in 1000
m!. of distilled water. Allow the solution to stand for about one week, and
then filter it through an asbestos filter. (Standardize .by using about 0.200 g.
portions of pure sodium oxalate.
.
Method.-In a 400-m!. casserole dissolve 1 g. of sample in 10 m!. of HN0 3
(sp.gr. 1.42). Add 20 mI. of HCI (sp.gr. 1.18), and evaporate to dryness.
Moisten with HCI, evaporate to dryness again, and bake to dull redness.
Moisten with,HCI again, add 3 m!. of ferric chloride solution, and dilute to
about 200 m!. with distilled water. Add N!I 4 0H (sp.gr. 0.90) until the basic
salts of copper have dissolved and the soluti~n has become a deep blue. Boil,
allow to settle, and filter on a loosely woven filter paper. Wash with dilute
amm6nia and, with hot water .. Dissolve the precipitate on the filter with hot
dilute HCI, dilute the solution to about 200 mI., add NH 40H (sp.gr. 0.90)
until the precipitate which forms at first .dills~lves rather slowly, and saturate
9 A. S. T. M. Methods of Chemical Analysis of Metals, p. 154, December, 1936.

ANALYSIS OF NON-FERROUS ALLOYS

1369

with HzS gas. Filter off and reject the. preci~itate: B~il the filtrate to expel
H.S and add HN0 3 (sp.gr. 1.42) untll the 1ron 1S oX1dized. Add NH 40H
(sp.~r. 0.90) until the ~olut~on is alkalin~. Boil ~nd filter on a loosely woven
filter paper. Wash w1th d~lute amrr.lOma and w1th hot water. Dissolve the
precipitate on the filter w1th hot d1lute H 2S0 4, rec(\iving the solution in a
350-m!. Erlenmeyer flask. Add NH 40H Csp.gr. 0.90) until the iron is entirely
precipitated, and then add HNO a Csp.gr. 1.42) cautiously until the solution
just becomes clear. Bring the solution to a temperature of about 80 C., and
add 40 ml. of ammonium molybdate at room temperature .. Allow to stand
for one minute, shake or agit~te for ~ minu.tes, filter ~I). a 9 cm. paper, and wash
very thoroughly (about 25 t1~es) w1th aCld a~momum sulfate. Dissolve the
precipitate on the paper usmg 50 ml. of d1lute ammonia. Add sufficient
H2S04 to adjust the acidity to 2.5% by volume and immediately pass the
solution through a Jones reductor, which has the reductor tube prolonged
and reaching nearl;y to the bottom of the flask, dipping into 50 ml. of fe_rricalum solution. Wash through the reductor with 150 ml. of the H 2S0 4 for
reductor and follow with an additional 100 ml. of liistilled water. Titrate
with 'standard KMn04'
'
By this method the molybdenum which in passing through the reductor
R&S M.f'.D reduced f'.ntirt'Jy to the form corresponding to the oxjde .A102 0J' js
partially oxidized by the ferric alum, an equivalent amount of iron being
reduced to the ferrous condition. The resultant solution is not rapidly oxidized
through contact with air during the titration, and the, calculations can therefore be made on the M0 20 a - 2MoO a basis.
Accuracy.-Duplicate determinations should check within 0.01 % of .ph6sphorus.
NOTEs.-The ammonium-molybdate solution should be kept in a cool place and
should always be filtered before using.
A blank determination should be made on corresponciing amounts of acid and
water, passing through the reductor into the usual amount of ferric-alum solution in the
flask.
A small quantity of liquid should always be left in th() reductor funnel, and air
should never be allowed to enter the reductor.
For a description of the Jones reductor consult, "Determination of Iron, Analysis of
Manganese Bronze."

DETERMINATION OF ANTIMONY BY THE VOLUMETRIC

METHOD
Solutions Requirea. Potassium Permanganate.-bissolve 0.3 g. of potassium permanganate in 1000 ml. of distilled water. Standardize by dissolving
25 mg. of pure antimony in 15 ml. of boiling sulfuric acid and proceed according
to the method from this point.
Method.-In a 250-ml. beaker, dissolve 5 g. of br()llze in 25 ml. of HNO a
Csp.gr. 1.42). When the action has ceased, boil until the oxides of nitrogen
are expelled, add 125 ml. of boiling water and allow to stand and settle for
one hour or longer, keeping the temperature just below the boiling point.
Filter on double 9 cm. closely woven papers, keeping the solution' hot. Wash
with boiling water. Discard the filtrate.

1370

ANALYSIS OF NON-FERROUS ALLOYS

Transfer the papers containing the tin al1d antimony to a 350-ml. Erlenmeyer flask and add 25 ml. of HNO a '(sp.gr. L~2), 5 g. of ammonium persuifate, and 15 ml. of H 2S0 4 (sp.gr. 1.84). Boil to strong fumes of SOa. If
the solution is brown, cool and add 5 ml. of HNO a (sp.gr. 1.42) and a little
more (1 g. or less) persuifate, if necestlary, arid boil again to strong fumes.
of SOa. When the ,solution is colorless, cool, ~dd 20 ml. of distilled water
20 ml. of HCI (sp.gr. 1.18) and 1 g. of sodiuin sulfite. Boil gently for 10
minutes to expel S02 completely. Dilute with 200 ml. of distilled water and
cool to 10 to 12 C. in ice or under running water.
Titrate with standard permanganate solution until the appearance ot a
decided pink color.
'
NOTES.-l. In alloys containing a large excess of tin, the antimony is precipitated
quantitatively with the meta-stannic acid.
.
2. Ammonium persulfate and, nitric acid are used to destroy the filter paper, and
subsequent boiling with sulfuric acid effects the solution of the tin.and antimony. A
mixture of fuming nitric and sulfuric acids might be used to accomplish.,the same results,
but the persulfate mixture is much more efficient.
.
3. If the solution becomes cool during the first filtration, some of the meta-stannic
acid may dissolve.
4. If arsenic is present, correction should be made for it.

DETERMINATION OF SULFUR
Method.-Weigh out four 5 g. samples of bronze into 250-ml. beakers.
Dissolve each in 25 m!. of HNO a (sp.gr. 1.42) and boil until the disappearance
of red fumes. Add 125 ml. of boiling water and allow to settle for one hour,
or longer, keeping the temperature just below the boiling point. Filter on
double 12.5 cm. closely woven papers, keeping the solution hot. Wash with
boiling water. Discard the meta-stannic acid precipitates, combine the four
filtrates in a tall 700-ml. beaker and add 0.5 g. of sodium carbonate, and
evaporate the solution until syrupy. Cool somewhat, then dilute to the lip
with warm water. There should be present a ll}'yer abo}lt 6 mm. in thickness,
composed of basic salts of copper. Electrolyze, using large stationary electrodes, at a current of 4 amperes, or using the Frary solenoid at a current of
about 10 amperes. The lead must be removed at the anode as the copper is
taken out at the cathode. When copper and lead are out, evaporate the
electrolyte to small volume, cover the beaker, add 75 ml. of HCI (sp.gr. 1.18)
and boil down to small volume. Add 75 mL mo1'elICI and evaporate .to
dryness to dehydrate any silica present. Moisten with about 0.5 ml. of HCI
and take up with 20 ml. of hot water. Filter into a sman beaker, Heat the
filtrate to boiling and add 10 m!. of 5% BaCh solution, drop by drop, with
stirring. Allow to settle at least five hours, filter on a small ashless paper,
and wash with hot water. Ignite cautiously in a small porcelain or platinum
crucibl~, and then heat to a bright redness for 20 minutes.
Cool and weigh
the BaS04 and calculate to sulfur.
'
.
NOTES.--;-l. Throughout the entire analysis, the utmost precautions must be taken tU
prevent contamination by sulfur in any form, H 2S0 4 fumes, HzS and S02 gases especially.
All heating with gas burners must be avoided. Run a blank'with each analysis, adding
the reagents and subjecting it to the same ope_!}l.tlons as the beakers containing the
dl'illings. Subtract the amount of BaS04 founa in the blank from that in the deter-

ANALYSIS OF NON-FERROUS ALLOYS

1371

minations. If this blank amounts to 0.002 g., it is a sign of poor work or impure reagents and the analysis should be repeated. 'Vash out all beakers, funnels, etc., with
distilled water before using.
2. It is necessary to keep the solution containing the meta-stannic acid hot during
filtration. If the solution becomes cool, some of the meta-stannic acid may dissolve.

DETERMINATION OF ARSENIC
Solutions Required. Ammonium Carbonate.-Dissolve 150 g. of ammonium carbonate in one liter of distilled water.
\
Method.-Dissolve the sample (5.00 to 10.00 g.) in a 100-ml. beaker in
10 ml. of HNO,,(sp.gr. 1.42) and 30 ml. of HCl (sp.gr. 1.18). When solution
is complete, cool, add 35 rol. of H 2S0 4 (sp.gr. 1.84), evaporate to the appearance of copious fumes of H 2S0 4, and continue the fuming for from ten to fifteen
minutes. Cool, take up in -HCI (sp.gr. LIS), transfer to an SOO-ml. flask
ha.ving a two-hole stopper with' separatory funnel and outlet t'ube connecting
with a vertical condenser outside. (The part of the outlet tube within the
distilling flask should have a steam outlet hole about 1 cm. from the-open end.)
Rinse the beaker with HCl (sp.gr. 1.18) and dilute the solution with the acid
to make a total volume of 350 to 450 m!. Add 10 g. of cuprous chloride
(Cu 2CI 2), connect the flask with the condenser, and gradually raise the contents
of the flask' to boiling. Catch the distill~te in a tall 400-ml. beaker containing
100 ml. of cold water. The end of the condenser should extend several millimeters into the water in the beaker and the beaker should be surrounded with
ice water in order to avoid any possible volatilization of arsenious chloride.
After about, 100 m!. of distillate has passed over, remove the flame from the
distilling flask, pour 3 to 5 ml. of hypophosphorus acid through the separatory
funnel into the.distilling flask, and redistill as before, catching the distillate
in the same beaker. The volume of the second. distillate should be 50 to 100
ml. Pass a rapid current of H 2S through the distillate for one hour, allow to
settle, and filter on a weighed Gooch crucible: 'Vash twice with HCl (1 : 1),
/,w.,.'l' t~;""'i'i.\5W N'.._:t.._"1;0 ..%\1& ;<rate.", t}l'.",,%' t.._:.._W6l5' w':t}i' 8k-ml.._ml, <'NN~ }03.',,_" j}.._ID.,OS w':t.._"1;0 Cs..!.
Dry at 110 0 C. for 10 minutes, and weigh after cooling. Dissqlve the As 2S a
\"ith (NH4)z'CO a solution, wash thoroughly with water; dry for 30 minutes
at 110 0 C. and reweigh after cooling. The difference in ,,,eights ,multiplied
by 0,609 gives the weight of arsenic.
NOTE.~Blank determinations should be made on the reagents carried along as in
the method and the proper cOlTections made.

THE DETER:\lINATION OF CADlIIIUlII IN BRASS

General Considerations.-From the results of a series of experiments on
known synthetic mixtures, it ,vas found that the follo\ving method is satisfactory even in the presence of mod,erate amounts of arsenic, lead, bismuth, tin
and antimony.
Method.-Prepare a sufficient amount of sawdust or drillings from the brass
to be t.ested to furnish at least...,a twenty-gram sample. For the assay weigh ten
gramf:; of this sample into a 500-ml. Erlenmeyer flask and dissolve in concentrated nitric acid. Add 50 m!. concentrated H 2S0 4 and fume strongly. Cool

1372

ANALYSIS OF NON-FERROUS ALLOYS

and dilute to 200 ml. with water, boil for a few minutes and allow to cool.
After two or three hours filter off any insoluble matter, collecting the filtrate in a
tall 600-ml. beaker. Discard the residue tin the paper and dilute the filtrate to
400 ml. Heat to boiling and pass a rapid ftream of H 2S through this hot solution for thirty minutes. (It is well to tho.roughly mix the copper sulfide with
the solution by means of rubber-tippeq ro(i so as to keep the precipitate from
being forced out of the solution by the gas.)
At the end of the half hour filter off the precipitated sulfides on an 18Yz
cm. paper, receiving the filtrate in a tall fiOO-ml. beaker." Wash the sulfides
once with hot water. If proper precipitation has been made, this filtration
will be rapid and easy to make and the filtrate will have no blue color. Return
the precipitated sulfides with the filter paper to the beaker i~ which precipita-'
tion was made and add 400 ml. of 10% H 2S0 4 Boil vigorously for a half hour
and filter hot on a 19Yz cm. paper. Wash the residue thoroughly with boiling
water. Discard the residue on the paper. All of the cadmium is now present
in these two solutions.
Cool and pass a rapid stream of H 2S through the solution for ten minutes.
Add ammonia slowly to each solution until zinc sulfide begins to precipitate.
When this is certain, continue the passage of the gas for five minutes more and
then filter off the precipitates fjeparately on 12Yz cm. paper. Discard the
filtrates. Combine the two papers containing mixed sulfides and boil with
100 ml. of 10% H 2S0 4 for ten minutes, stirring frequently; Filter and wash.
Cool and dilute to 300 ml. Pass in H 2S and add ammonia drop by drop until
cadmium is entirely precipitated. Pass gas for about twenty minutes. Filter
the precipitate on an 11 cm. close texture. paper. Return the paper to the
beaker and boil with 50 ml. of 10% H 2S0 4 Filter and wash. Dilute to 300 ml.
and pass in H 2S for twenty minutes. A drop or two of ammonia wilL be suffi,
cient to start the precipitation of cadmium.
Filter the cadmium sulfides on a weighed Gooch crucible. Wash twice
with cold water. Dry at 110 C. for two hours and reweigh.
Weight CdSXO.778= )Veight cadmium.
The final cadmium sulfide precipitate will have a characteristic brigh't, yellow color
and will be entirely free of other sulfides if the foregoing directions be carefully followed.
As a special precaution when arsenic or tin are suspected in the brass the final sulfide may be washed out of the crucible after weighing and treated with yellow ammonium
sulfide, (the addition of NH 4Cl will tend. to prevent cadmium sulfide being carried
into the ammoIJ.ium sulfide solution as a colloid,) - Filter and wash the residue and
dissolve the cadmium sulfide in 10% H 2S0 4, reprecipitate and weigh again.

ANALYSIS OF NON-FERROUS ALLOYS

, 1373

ANALYSIS OF, COPPER, NICKEL, ZINC ALLOYS

ANALYSIS OF GERMAN SILVER
(ALLOY OF COPPER, NICKEL AND ZINC)
DETERMINATION OF TIN, COPPER AND LEAD
Proceed exactly as in the determination of these elements under" Analysis
of Manganese Bronze."
DETERMINATION OF IRON
Solutions Required. Stannous Chloride.-Dissolve 450 g. of the salt in
380 ml. of concentrated HCI (sp,gr. 1.18) to which some water has been added
and dilute to 2 liters,
Mercuric Chloride.-Make a saturated solution of the salt in hot water,
allow to cool to room temperature and filter.
" Manganese Sulfate.-Dissolve 160 g. of the salt in water and dilute to
1750 m!. To this add 330 ml. of phosphoric acid (sp.gr. 1.70) and 320 m!. of
sulfuric acid (sp.gr. 1.84). This solution is to obviate the deterious action of
hydrochloric acid upon the standard potassium permanganate. The phosphoric acid allows the formation of iron phosphate, which, being nearly colo'rless,
renders the end reaction more distinct.
Potassium Permanganate.-Make a solution containing 0.3163 g. of salt
per liter. This is approximately N/I00.
Standardization of above Permanganate Solution.-Take measured quantities of a carefully standardized ferric chloride solution, containing 5 g. of
metallic iron and 100 ml. concentrated hydrochloric acid to a liter in No. 1
beakers. About 5 m!. of this solution should prove adequate. Add from a
burette stannous chloride solution until the iron is almost reduced, told by the
color change. Then heat to boiling and reduce iron completely with 811CI 2,
using a slight excess of the reagent. Add 5 m!. of HgCh solution to oxidize
excess of 8nC1 2 Pour solution into a 600-ml. beaker containing 6-8 ml. of
the Mn80 4 solution in 400 m!. of water. Titrate with KMn04 immediately
until a faint pink is obtained. Calculate potassium pefmaI\~anate in terms of
metallic iron.
Treatment of Sample.-Weigh out a 5 g. sample of drillings from which
all steel chips have been removed by a magnet and transfer to a 250-ml. beaker.
Dissolve in as little nitric acid (sp.gr. 1.42) as possIble (25 ml. should be sufficient), heat until all nitric oxide fumes have been dliven off and then dilute
to 100 ml. with distilled water. Add 3 to 4 g. of ammonium chloride crystals,
heat to boiling, and add a slight excess of ammonia to precipitate all the iron
as hydroxide.
A small amount of zinc.is usually precipitated at this point as hydroxide
and also, if manganese is present in the alloy, it will come down as manganese
hydroxide.
After the precipitate has settled down, filter through filter paper and wash
with ammonium hydroxide 1 : 4 until all soluble salts are removed (or until'

1374

ANALYSIS OF NON-FERROUS ALLOYS

the blue copper color has been washed from the filter paper). Treat the pre-'
cipitate on the filter paper with a little hot HGI (1 : 1), catching the resulting
solution in original beaker. Wash the filter p~per alternately with hot water
and HCI (1 : 1) until the iron has been cOIfpletely removed. Dilute the
hydrochloric acid solution in the beaker to about 75 m!. and add 3-4 g. of ammonium chloride. Heat to boiling and repre~ipitate the iron with a slight
excess of ammonium hydroxide. Filter and w'ash as before. (A double precipitation is usually sufficient to remove zinc, etc.) Dissolve the precipitate
in a small amount of hydrochloric acid (1 : 1). Add stannous chloride to
reduce the iron-2 or 3 drops are usually sufficient.
Heat almost to boiling and add 1 or 2 drops more of stannous chloride to
be sure an excess is present. Run in an excess of mercuric chloride solution
(about 5 m!.) to oxidize the stannous chloride.
.
Pour into a 600-ml. beaker containing 6-8 ml. of the manganese sulfate
solution in 400 m!. of water.
.
Titrate with standard potassium permanganate solution until a slight pink
opalescence is obtained.
- ,
DETERMINATION OF MANGANESE
Proceed exactly as in "Deterniination of Manganese by the Persulfate
Method" or "Determination of, Manganese by the Bismuthate Method,"
under" Analysis of Manganese Bronze."
DETERMINATION OF NICKEL
Solutions Required. Standard Solution of Potassium Cyanide.-Dissolve
26 g. of pure potassium cyanide in 1000 m!. of distilled water and add 1.25 g. of
silver nitrate. This solution is standardized against a solution of known metal
content.
Potassium Iodide.-Dissolve 100 g.lof c.p. potassium iodide crystals in
1000 m!. of distilled water.
,
Sodium Pyrophosphate.-Make a saturated solution of c.p. sodium pyrophosphate crystals in distilled water.
I
Method.-See the" Determination of Copper, -J,ead and Tin."
Transfer the electrolyte and washings from the electrolytic determination
of copper to a 400-m!. beaker. Add 3 to 4 g. of ammonium chloride and then
50 to 75 m!. of a saturated solution of sodium pyrophosphate, depending'
upon the amount of zinc present. Ne.utralize with ammonium hydroxide
(sp.gr. 0.90) and then add enough to make the solution only slightly but yet
distinctly alkaline. A large excess of ammop.ia will hinder 9r entirely prevent
the reaction. Cool, and do not let the temperature get much above 20 C.
or the result will be irregular. Add 2 m!. of a 10% Ilolution of potassium
iodide and titrate with a standard solution of potassium cyanide containing
silver nitrate until the white cloud caused by the formation of silver iodide
has entirely disappeared and one drop of th:e: cyanide causes a clear solution.
This is the end point which is very sharp ind distinct.

'

ANALYSIS OF NON-FERROUS ALLOYS

1375

XOT'S.-Tbe method of estimating nickel by means of adding potassium cyanide to

an ammoniacal solution of nickel containing silver iodide in suspension !'O that the solution
rcmaint.'<i turbid until the nickel W88 aU converted into the double cya.nide of nickel
8.n~ potassium,'after which 8. single drop of the .cyanide in eXC(!f!8 .cleared up ~he solution,
was originally propose:d by T. Moore and modified by \V. B. Price for use 1D the determination of nickel in German rover, etc.
..".~
The following elements interfere with the metboo: ~fanianese, zinc. iron. aluminum
and especially copper, which, owing to its forming cYIlnides., would render the method
valueless. Cobalt, if present, will be cstimn.ted fL'I;nickel: The iron, aluminum nnd
tine may be kept in solution so that they do.lnot interfere by too ure of organic adds
6uch 8.8 tartaric or citric or with sodium pyroph06pha.te.

DETERMINATION OF ZINC
Zinc is uruaUy taken by difference.
In case an aetua. determination is desired proceed exactly as in " Determination of Zinc,-Analysia of Bronzo Bearing IVletals.JI

ANALYSIS OF BISMUTH BRONZE

(ALLOY OF COPPER, TIN, BISMUTH, NICKEL AND. ZINC)
General Considerations.-Bismuth will interfere with the electrolytic determination of copper and must be removed before this determination can be
made.

DETERIIIIXATIOX OF TIN
Proceed exactly as in

Determination of Tin, Analysis of Bronze 'J;learing

~1eta.1. "

, DETEiufINATroX OF LEAD

filtrate from

To the
the H,SnO, separation, add 8 ml. of H,SO. (sp.gr.
1.84) and evaporate until copious fumes of H_ 2 SO, are evolved. Cool, a.dd
30 mi. of distilled water and. boil until all bismuth sulfate is dissolved. Cool,
filter on a weighed Gooch crucible and wash '"ith cold dilute H,sO. (10%).
Ytlter as soon &8 possible, otherv.-ise some basic bismuth sulfate will separate out.
Set the Gooch crucible inside a porcehdn crucible, dry, ignite for 5 minutes at
tbe fuJI heat of a Tirrill burner, cool and weigh as PbSO. which contains 68.33%
of lead.
NOTE.-A small amount of lead still reme.iDB in the solution to be recovered later.
(See Note: "Determination of Bismuth.")

1376

ANALYSIS OF NON-FERROUS ALLOYS

DETERMINATION OF COPPER

To the filtrate from the PbS0 4 separation add 0.11 g. of iron wire dissolved in
10 ml. of HNO a (sp.gr. 1.42) and thel/ add NH49H (sp.gr. 0.9) in sufficient
excess to hold the copper in solution. NoW add 10 ml. of a saturated solution
of ammonium carbonate, bring to boiling, filter I and wash with hot water
containing a small amount of (NH 4 )2CO a Reselive the filtrate. Wash the
precipitate back into the original beaker and dissolve any remaining precipitate on the paper with 20 ml. of hot RNO. (1 : 2) catching the resulting
solution in the original beaker. Wash several times with hot water. Warm I
until the precipitate is completely dissolved and make a second separation
of iron and bismuth with ammonium hydroxide and ammonium carbonate.
Again dissolve the precipitate with nitric acid and make a third separation.
Hold the residue.
Combine the three filtrates and washings from the separation of iron and
bismuth, which should not exceed 400 ml., make just acid with H 2S0 4 (1 : 1),
add 10 ml. in excess, and evaporate down to a volume of 100 ml. Cool, add
1.5 ml. of HNO. (sp.gr. 1.42) and electrolyze for copper. (See" Determination
of Copper by the Electrolytic Method, Analysis of Manganese Bronze.")
DETERMINATION OF BISMUTH
Treat the precipitate containing tile bismuth and ii:on on the paper with a
little hot HCl (1 : 1) catching the resulting solution in a clean beaker. Wash
the paper alternately with hot water and hot dilute HCl (1 : 1) until the bismuth has been completely removed. Neutralize the free acid with NH 4 0H
(sp.gr. 0.9), make slightly acid with HCl and pass H 2S. Filter and wash with
H 2S water. (Discard the filtrate unless it is desired to run zinc or nickel on
this portion.)
Wash the precipitate of bismuth sulfide back into the beaker and pour
through the filter paper 30 ml. of hot HN0 3 (1 : 25 catching the solution in the
same beaker. Wash a few times with hot water! Heat the solution until the
precipitate is completely dissolved. Filter' through the same filter paper and
wash with hot dilute HNO. (1 : 50). If a black insoluble residue still remains,
transfer the residue to a small porcelain crucible, dry, ignite over a low flame,
cool, dissolve in 1 ml. of HNO. (sp.gr. 1.42) and combine with the main filtrate.
Boil the combined solutions down to a volume of 60 ml. and-allow to stand and
settle for one hour keeping the temperature just below the boiling point. If a
precipitate of H 2SnO. forms, filter, wash, dry, ignite and weIgh as Sn02' Add
the weight found to that already obtained (see" Determination of Tin.").
Evaporate the filtrate to a volume of about 50 mI., make just neutral with
NH 4 0H (1 : 3) and then add 5 ml. of HCl (1 : ,9) in excess. If a precipitate of
BiOCI forms, cautiously heat until it goes into solution. Dilute to a volume of
400 ml. with boiling hot water and allow to stand in a warm place several hours.
Filter oil a weighed porcelain Gooch crucible and wash with hot water. Finally
wash once with 95% alcohol but transfer the filtrate to' a' clean beaker before
doing so. Dry at 110 C. for 30 minutes, cooLa.nd weigh as BiOCI. Multiply
by Q.8025 for bismuth.
. _. -

ANALYSIS OF NON-FERROUS ALLOYS

1377

NOTE.-The filtrate from the BiOCI separation contains a small amount of lead that
did not precipitate out as PbSO. (see "Determination of Lead"). This may be recovered in the folloWing maimer: make the solution just alkaline with ammonium
hydroxide, pass R 2S, filter and wash. Dissolve the precipitate from the filter paper
with 15 m!: of hot RNO, (1 : 2) and wash with hot water. Add 12 m!. of RNO, (sp.gr.
1.42), dilute to a volume of 150 m!. with distilled water and electrolyze for lead (see
"Determination of Lead by the Electrolytic Method, Analysis of Manganese Bronze").
Add the weight of lead found to that already obtained.
DETERMINATION OF NICKEL
Method.-Remove the tin and lead as in "Determination of Tin" and
" Determination of Lead." Remove the copper by electrolysis (see" Determination of Copper by the Electrolytic Method, Analysis of Manganese
Bronz~ "). Continue as in " Determination of Nickel, Analysis of German
Silver."
N OTE.-1. If the amount of lead is small, the copper and lead may be removed at the
same time by electrolysis (see" Determination of Copper and Lead Simultaneously by
the Electrolytic Method, Analysis of Manganese Bronze"). It is not necessary to
evaporate to dryness.
2. The presence of a moderate amount of bismuth will not prevent the electrolytic
precipitation of copper. The deposit, however, will be black and possibly spongy, and
may slough off in washing, in which case, filter and wash and combine washings and main
filtrate. If the spongy material has dropped off into the electrolyte it will dissolve.
In such a case filter both electrolyte and washings through the same paper, wash, combine filtrate and washings, pass R 2S, filter and wash, boil to expel H.S, oxidize l "he iron
with RN0 3 (sp.gr. 1.42) and proceed ~ directed above.
DETERMINATION OF ZINC
Method.-Remove the tin, lead and copper as directed under" Determination of Nickel." Continue as in " Determination of Zinc, Analysis of Bronze
Bearing Metal."
DETERMINATION OF IRON
Method.-Proceed exactly as in "Determination of Iron, Analysis of
Bronze Bearing Metal," up to the point where the H 2SnOa has been separated
and any iron present in the H 2SnOa has been recovered and combined with the
main filtrate. Adq. 5 g. of ammonium chloride crystals, and a slight excess of
NH 40H (sp.gr. 0.9). Boil, allow to settle, filter on a loosely woven filter paper
and wash with dilute ammonium hydroxide and with hot water. Treat the
precipitate on the paper with a little hot HCl (1 : 1) catching the lesulting
solution in original beaker. Wash the filter paper alternately with hot water
and HCl (1 : 1) until the iron has been completely removed. Neutralize the
solution with NH.OH (sp.gr. 0.9), make acid with 5 m!. of Hel (sp.gr. 1.18)
in excess, dilute to a volume of 100 ml., warm and pass H 2S. Filter and wash
with hot water. Boil the filtrate to expel H 2S, add a few m!. of HNOa (sp.gr.
1.42) to oxidize the iron, dilute to a volume of 75 mI., add 3 to 4 g. of ammonium
chloride crystals, and a slight excess of NH.OH (sp.gr. 0.9) and continue as in
" Determination of Iron, Analysis of German Silver."

1378

ANALYSIS OF NON-FERROUS ALLOYS

DETERMINATION OF ALUMINUM

Proceed exactly as in" Determination of AluJinum, Analysis of Manganese

Bronze."

DETERMINATION. OF MAfOANESE
Proceed exactly as in" Determination of Manganese, Analysis of Manganese
Bronze."
.

COPPER NICKEL ALLOYS

ANALYSIS OF MONEL METAL 10
(ALLOY OF Ni, CU, AND lRON)
DETERMINATION ~F COPPER AND NICKEL

Electrodes.-See " Analysis of Manganese Bronl'le."

Method.-Dissolve one gram of driIIings in 15 mI. of nitric acid (1 : 1) in a
200-mI. covered beaker. Add 30 mI. of sulfuric acid (1 : 1), and .evaporate
until fumes of sulfuric acid are given off. Dilute with 125 ml. of water, heat on
hot plate until sulfates are dissolved. Add 1.5 mI. of nitric acid, cool ahd plate
out copper over night with a current of 0.3 amp. After copper is all plated out
pour the solution and water used in washing off cathode.. into a 500-mI. beaker.
The cathode is dipped in alcohol and heated over ap. alcohol lamp and when cool,
weighed. The increased weight is equal to copper. The solution in the beaker
is covered, made alkaline with ammonia, and 75 mI. added in excess. To 30 mI.
of cold water add about 2 grams of sodium 'peroxide. Add this solution to the
alkaline nickel solution and let stand on the steam bath for one-half hour.
Repeat the addition of sodium peroxide solution, then cool the alkaline nickel
solution, and when cold plate nickel over night with a current of 0.5 amp.
Wash the cathode with water and alcohol. Heat over an' alcohol lamp and
when cool weigh as nickel and cobalt. The nickel and cobalt contain a small
amount of manganese, so dissolve the nickel and cobalt in nitric acid and determine the manganese by the bismuthate method. This is deducted, the result
being nickel anet cobalt.
Monel Metal contains a small amount of {cobalt. If nickel, exclusive of
cobalt, is desired dissolve the deposit with hot nitric acid, transfer to a 250 mI.
graduated flask, dilute to the mark with cold water, mix and draw off exactly
50 ml. Proceed as described in" Determination of Nickel, Analysis of Bronze
I
Bearing Metals."
10 See "Tentative Methods of Ch,emical Analyski"nf Nickel," A, S. T. M. Methods of
Chemical Analysis of Metals, p. 226, December, 1936.

ANALYSIS OF NOr<-FERROUS ALLOYS

1379

....

NOTE.-In very accurate determinaticms the solution, after plating out,the nickel
and cobalt, is evaporated down, made ~cid with sulfuric acid, make alkaline with
ammonia, sodium peroxide added and replated, same as above.
DETERMINA,rION OF IRON

tl

Solutions Required. Cuprous C/lloride Solution.-Place into a 2-liter

heavy glass bottle, 400' g. of NaCI, 100 m!. of HCI (sp.gr. (1.18) and 20 g. of
Cu 2 Cb. Add 2 liters of warm water; shake well; place into the solution a few
copper rods and let stand over night in ~ cool place. After shaking, the solution
is ready for use. One milliliter of Cu2Cl2 should reduce 0.003 to' 0.005 g. of iron.
Standard Ferric, Chloride Solution.--Dissolve 5 grams of standard iron wire
in 100 m!. of hydrochloric acid (1 : 1), then 5 grams of potassium chlorate are
added to oxidize the iron. Boil until all the free chlorine is driven off, then cool
and make the solution up to two literG
Method.-Dissolve one gram of drjIlings in 25 ml. of nitric acid (1 :-'1), in a
300-m!. Griffin lip beaker, add 50 m!. Of cold water and make strongly alkaline
with ammonia. Boil a few minutes, filter and wash precipitate with hot water
two or three times. The precipitate is redissolved in hot hydrochloric acid
(1-3) to which a few drops of sulfurous acid are added. After the precipitate is
dissolved, boil the solution a few minutes, add a few drops of nitric acid;Iet solution cool for 5 or 10 minutes, then make alkaline with ammonia, filter and wash
with hot water. Repeat redissolving ttrid ammonia precipitation. After final
reprecipitation, dissolve the iron hydrOxide in 30 ml. of hot hydrochloric acid
(1-3) into a 24-oz. flask, cool and add ;300 m!. of cold distilled water and titrate
\vith a standard cuprous chloride solution, using 2 drops of a saturated solution
of potassium sulfocyanate as inside indicator. The cuprous chloride solution is
standardized each time before using, \#"ith standard ferric chloride.
NOTE.-If the ferric hydroxide is dissolved in dilute- sulfuric acid and the iron reduced with zinc, the iron titrated with potassium permanganate is just as accurate as
above method, providing 5 grams of the drillings are taken.
DETERMINATION OF SULFUR AND SILICON

Method.-Weigh 10 g. of drilling. into an 800-m!' casserole, add 3 g. of

KCI0 3 , cover and then add 200 ml. of concentrated HNO a ,Place on steam bath
until action has ceased and then evaporate to dryness'on the hot plate. Cool,
add 50 mi. of HCI (sp.gr. 1.18) and a~ain evaporate to dryness. Repeat this
operation. Take up with 150 m!. of water, add 8 m!. of HCI (sp.gr. 1.18), heat
to boiling, filter and wash, taking care that the filtrate and washings do not
exceed 300 m!. in volume. Reserve t}1e residue. Heat the filtrate to boiling,
add 25 m!. of 20% BaCl 2 solution, and gllow to stand in a warm place over night.
Filter through a small double ashless filter paper and wash with very dilute
HCI, and finally with hot water. Ignite in a small platinum crucible and then
heat to bright redness for twenty minutes. Cool and weigh as BaS04. Multiply by 0.13735 for sulfur.
Transfer the residue to a weighed platinum crucible, dry, ignite and weigh
as 8iO,. Multiply by 0.4672 for silicon.
Run a blank analysis on the cheDlicals used. If this blank amounts to
0.004 g., obtain fresh reagents and repeat the analysis.
R'-m" \\\ ~'U"P\.\.'t'b.(ft"

1380

ANALYSIS OF NON-FERROUS ALLOYS

DETERMINATIO:!i OF MANGANBSE

Solutions Required. Ferrous Ammonium SJlfaie.-35.6 grams per liter.

Add about 1 mL of sulfuric acid. One mL equals 0.001 g. of Mn.
Potassium Permanganate.-2.90 grams per liter. One mL equals 1 mL of
I
ferrous ammonium sulfate.
The ferrous ammonium sulfate solution changes .rapidly, therefore, it is
standardized against permanganate every day . Run 25 mL of ferrous ammonium sulfate solution into a flask which contai~s a sample that has been
titrated, then titrate to a faint pink with permanganate. This gives the value
in permanganate of the ferrous ammonium sulfate.
Method.-Dissolve one gram of drillings in 10 mL of concentrated nitric
acid in a No.2 tall 250 mL beaker. Add 30 mL'of warm water and add ahout
0.5 g. of sodium bismuthate, heat a few minutes, then add enough sulfur~)Us
acid to clear the solution and boil until nitrous oxide,is all driven off. Cool to
about 15 C., add an excess of sodium bismuthate and let stand a few minutes.
Filter through asbestos felt on a Gooch crucible into a 500 mL Erlenmeyer
suction flask; apply suction gently, wash with 3% nitric acid .. Run into the
flask a measured amount of ferrous ammonium sulfate solution and titrate. the
excess at once with standard potassium permanganate, deducting 0.5 mL of
permanganate, which is used in neutralizing the color of copper and nickel
solution. The number of mL of permanganate used less 0.5 mL is subtracted
from the original amount of ferrous a;mmonium sulfate taken and the remainder
multiplied by 0.001 gives the per cent of manganese.
DETERMINATION OF CARBON

The usual carbon combustion apparatus is used with a platinum combustion

tube. In addition to the usual absorption train there is placed next to the
platinum combustion tube a 5 inch "U "-tube filled with glass beads and
moistened with chromic acid solution, with a loose plug of glass wool in the
forward end. This serves to oxidize any sulfurous anhydride and to abs9rb
the sulfuric anhydride formed from the sulfur lin the drillings during combustion. This tube is washed out daily and fresh glasswood inserted. The
forward end of the platinum combustion tube contains a plug of loosely
wound platinum gauze, 5Yz inches long, completely filling the bore of the tube.
and a similar roll 2 inches long, fitted with: a loop is pushed in after the boat.
Fill the alundum boat with ignited alumina and make a "V "-shaped
depression in the middle with a spatula, pressing the materia] against the sides
of the boat. Ignite the boat and its contents strongly to expel any carbon
dioxide present and burn off any carbonaceous matter that the alumina may
contain. Weigh out 3 g. of the drillings which have been broken down so as
to pass through a 10-mesh sieve and transfer it to the boat.
The absorption is made in potassium hydrqxide (sp.gr. 1.27) placed into a
Geissler potash bulb or with" Ascarite " contained in a Schwartz drying tube,
with a calcium chloride tube attached. Burn the carbon by starting a slow
current of oxygen and lighting one of the burners at the forward end of the tube,
and as soon as the tube is red hot increase the heat slowkv by lighting the other
burners. The valve contrpIIing the supply. olO'X~gen must be regulated so as
to keep a slow current of gas passing through the absorption apparatus at all

ANALYSIS OF NON-FERROUS ALLOYS

1381

times, and not to exceed three bubbles in a second. Monel Metal containing
manganese, or high in silicon, is burned at a temperature between 11250 C. and
1200 0 C., for tw;enty minutes.
Analysis of Aluminum and its Alloys.-See Chapter on Aluminum, Volume I.
Analysis of Beryllium-Copper Alloys.-See Chapter on Beryllium, Volume I.
Analysis of Magnesium and its Alloys.-See Chapter on Magnesium,
Volume 1.

TIN BASE AND LEAD BASE ALLOYS

ANALYSIS OF ALLOYS OF .LEAD, COPPER, ANTIMONY AND TINll
These methods apply particularly to white metal-beal}g alloys 12 (known
commercially as " Babbitt Metal ") and to similar lead ana tin base alloys.
Two sets of methods are here given. The first method is somewhat slow
but is recommended as giving the more accurate results where the analysis of
these alloys is of comparatively infI:equent occurrence. The second method
is rapid and is suitable for control work, giving good results after the analyst
has become familiar with it.
GENERAL METHODS 13'
A. DETERMINATION OF LEAD, COPPER, ANTIMONY AND TIN
DETERMINATION OF LEAD

Solutions Required. Mixed Acid.-Dissolve 20 g. of KCI in 500 m!. of

wa.ter, add 400 ml. of HCI (sp.gr. 1.18), mix and add 100 m!. of HN0 3 (sp.gr.
1.42).
A:cohol-Hydrochloric Acid Mixture.-Mix 400 ml. of 95% ethyl alcohol
and 100 ml. of HCI (sp.gr. 1.18).
Acid Ammonium Acetate.-To 500 ml. of NH 40H (sp.gr. 0.90), add 500 m!.
of water and then acetic.acid (80%), until neutral to litmus; then add sufficient
acetic acid to make the solution 2% acid with acetic (80%) .
. Method.-Dissolve 1 g. of the finely divided alloy by boiling in 70 to 100 m!.
of "mixed acid" solution in a covered beaker. Add more "mixed acid"
if a complete solution of the alloy is difficult to obtain, and continue boiling
until evaporated to about 50 ml. Add 5 m!. of HCI (sp.gr. 1.18), cool in ice
water until the bulk of PbCl 2 has crystallized out, then add slowly, with constant stirring, 50 ml. of alcohol (95%), continue stirring for a few minutes an~
A. S. T. M. Methods of Chemical Analysis of Metals, p. 184, December, 1936.
For Specifications for Babbitt Metal, see 1936 A. S. T. M. Standards, Part I, p. 700.
13 G. W. Thompson's Method, J. Chern. Ind., 15. 179 (1896).
11
12

1382

ANALYSIS OF NON-FERROUS ALLOYS

I

cool in ice water for 10 minutes. Add 50 m!. more 'alcohol from a pipette, allow
to stand in ice water for 20 minutes and filter through 9 cm. paper into an
800-m!. beaker. Wash by decantation three timEfs with alcohol-hydrochloric
acid mixture and finally wash the paper twice with the same mixture. Wash
the PbCl 2 from the paper back into the beaker; w~sh paper several times with
hot water, catching washings in beaker with the bulk of the chloride. Finally
wash paper with hot acid ammonium acetate soluti'on. Heat until all PbCl 2 is
dissolved. Add 15 m!. of saturated solution of K 2 Cr 2 07; heat until precipitate
is of good orange color; filter on weighed Gooch crucible; wash with water,
alcohol and ether; dry at 100 C. and weigh. Calculate percentage of lead
by the empirical factor 63.75.,
NOTEs.-1. With proper manipulation, the PbCI. should contain consistently all but
0.003 g. of lead. When, however, the lead content is low as in the high tin Babbitts, it
is not necessary to use 70 to 100 ml. of mixed acid or to add alcohol after 50lution of the
alloy. In such cases, weigh th_e alloy directly into a 600-ml. beaker, add about 15 ml.
of HCI (1 : 1) warm, and then add HNO s (sp.gr. 1.42), a little at a time, until the alloy
is dissolved and the tin completely oxidized.
2. During the heating of the acid ammonium acetate to dissolve the PbCl., the
solution should remain perfectly clear, any turbidity indicating the .presence of tin or
antimony, as even 1 mg. of tin or antimony will cause a slight but distinct turbidity .
DETERMINATION OF COPPER AND AN'TIMONY

Solutions Required.-Potassium Hydroxide.-Dissolve 100 g. of KOH in

500 m!. of water.
/
Potassium Cyanide.-Dissolve 3.5 g. of KCN in 1800 m!. of water and
standardize against copper of known purity.
Sodium Thiosulfate.-Dissolve 24.8 g. of Na 2S 2 03' 5H 20 in 1000 mL of
water, and allow to stand for 24 hours. Standardize against Antimony Metal
c.p., using same quantity of reagents and same procedure as under (c) Determination of Antimony.
Method.-See " Determination of Lead."
Evaporate the filtrate from PbCl 2 filtration by ljoiling in the loosely cover~d
800-m!. beaker, and finally .take to dryness ,on a {water-bath. Add 10 m~. of
KOH solution and after a few minutes add 20 ,m!. of 3% H 2 0 2 Add mere
KOH, if solution is acid, until an alkaline reaction is shown by litmus paper.
Heat on the water-bath for 20 minutes, add 10 g. of ammonium oxalate, 10 g.
of oxalic acid, and 200 m!. of water and heat to boiling. Pass in H 2S for 45
minutes with solution ,near boiling; filter at once; anCl' wash precipitate with
hot dilute solution of oxalic acid saturated with H 2S, catchIng washings with
filtrate.
Wash the precipitate of copper and antimony sulfides from the filter paper
back into the beaker with the least amount of water possible; treat ~ith 10 m!.
of KOH solution; heat on water-bath until the undissolved matter is distinctly
black, filter through same paper into a 500-mJ.fErlenmeyer flask. Wash the
precipitate with hot water containing a small a~ount of K 2S.

(a) Determination of Copper by Potassium Cyanide Titration

Dry and ignite the precipitate with paper in a small/casserole, dissolve in
nitric acid (sp.gr. 1.42), boil to expel nitrous fumeS, neutralize with Na 2 C0 3 , add
a few drops of NH 40H and titrate with ,standard KCN solution.

ANALYS,IS OF NON-FERROUS ALLOYS'

1383

(b) Determination of Copper by Electrolytic Method

See" Determination of Copper by Potassium Cyanide Titration."

Dry and ignite the sulfide precipitate in a small casserole, dissolve in
10 ml. of HNO a (sp.gr. 1.42), boil until oxides of nitrogen are expelled and add
50 ml. of distilled water. Transfer to a tall 200-ml. beaker, wash out casserole,
add 5 nil. of H 2S0 4 (sp.gr. 1.84) and determine copper by electrolysis. See
" Analysis of Manganese Bronze."
NOTES.-1. On filter after heating with KOH solution, the copper remains as
sulfide with a small amount of lead sulfide which failed to be precipitated as chloride.
If it is desired to determine this lead, it can be done by separation from the copper by ,
~
ordinary methods.
. 2. If the amount of antimony and copper is small (less than 10 mg.), tH ead which
failed to be precipitated as chloride may also fail to come down as sulfid on passing
H 2S through the oxalic solution. It will remain in the filtrate from the sulfides and be
deposited electrolytically with the tin on the cathode. This can be prevented by
adding an oxalic acid solution of a pure antimony salt containing about 100 mg. of
antimony just before passing H 2S. In this case antimony must be determined on a
sepamte portion. (See" Determination of Antimony in Solder," General Method.)
3. If the amount of copper present is from 8 to 10 mg. or more, it should be determined by the electrolytic method. In this case, the small amount of lead precipitated
as sulfide with the copper is deposited on the anode as Pb0 2 and may be weighed,
calculated to lead by factor 0.8662 and added to that obtained under "Determinations
of Lead."

(c) Determination of Antimony

See" Determination of Copper by Potassium Cyanide Titration."

Add 50 ml. of HCI (sp.gr. 1.18) to the KOH solution of antimony sulfide,
and boil down to small volume until all arsenic has been expelled as arsenious
chlQride.
Add 25 ml. of HCI (sp.gr. 1.18) and 1 g. of KCIO a and boil until colorless.
Filt.er into a 500-ml. Erlenmeyer flask, through glass wool if sulfur has'l3eparated,
wash out original flask with HCI (sp.gr. 1.18), cool, add 1 g. of KI, 1 ml. of
.
CS 2, and titrate with standard thiosulfate solution.
NOTES.-l. When arsenic is present to the extent of 0.3%, the addition of 50 mI.
of HCI (sp.gr. 1.18) to the KOH solution and subsequent boiling may not completely
expel the arsenic. In such case add a gram or two of sodium thiosulfate and repeat with
the addition of HCI and boiling .
2. If sulfur is present it will dissolve in the CS 2 and impart a color to the indicator
which interferes with the reading of the end-point. To eliminate this condition titrate
with the Na 2S 203 solution to within 0.5 ml. of the end-point, pour the solution into a clean
flask, add the CS 2 and continue the titration.
DETERMINATION OF TIN

Apparatus for Electrolysis. Electrodes.-See "Determination of Copper

by the Electrolytic Method, Analysis of Manganese Bronze."
Platinum wire spiral anodes.
Method.-See "Determination of. Copper and Antimony."
Concentrate the filtrate and washings from the precipitated sulfides, if necessary, to a volume of 200 ml. and add 5 g. more of oxalic acid in case the amount
of tin is over 0.5 g. Electrolyze until the solution reacts alkaline, using a
current of about 0.5 ampere. Remove the cylinder; wash twice with distilled
water and once with 95% ethyl alcohol; dry at 1100 C. and weigh.

1384

ANALYSIS OF NON-FERROUS ALLOYS

J

NOTE.-l. If the electrolysis is carried out over night, it will usually be found by
morning that the solution has become alkaline and it milY be taken for granted that the
tin has all been precipitated on the cathode. The best results are obtained by regulating
the current, or the time, so as to render the solution alkaline only a very short time
before the cathode is to be removed. The anode shouldl touch the bottom of the beaker
to insure proper agitation of the electrolyte. Ifhe catJiode should be 0.25 inch above.
2. Appreciable amounts of nickel, zinc and iron shbUld not be present.
I

B.

DETERMINATION OF ARSENIC

Solutions Required. Ferric Chloride.-Solution of specific gravity 1.43.

Ammonium Carbonate.-Dissolve 75 g. in 500 mI. of water.
Method.-Weigh 1 g. of sample into a 250-mI. distillation flask and add 10
mI. of FeCls solution,' 60 mI. of HCI (sp.gr. 1.18),20 mI. of water and 5 g. of KCI.
Connect the flask with a condenser, heat slowly until solution is complete and
boil down to as small a volume as possible, catching distillate in a tall 400-mI.
beaker. Cool flask, add 50 m!. HCI (sp.gr. 1.18) and redistill as before, catching distillate in same beaker. Cool distillate and pass H 28 through for one
hour, allow to settle and filter the arsenious sulfide on a Gooch crucible. Wash
twice with HCl (1 : 4), four times with cold water, three times with alcohol
and four times with C8 2 Dry at 110 C. for 10 minutes, and weigh after
cooling. Dissolve the arsenious sulfide ,with (NH 4)zC0 3 solution, wash thoroughly with water, dry for 30 minutes at 110 C. and reweigh after cooling.
The difference in weights multiplied by 0.6091 gives the percentage of arsenic.

"
RAPID METHODS FOR CONTROL WORK 14
A.

DETERMINATION OF LEAD AND COPPER

Solutions Required. Lead Acid.-See "Determination of Lead as Sulfate, Analysis of Manganese Bronze."
Dilute Alcohol for Washing.-Mix equal parts of 'denatured alcohol and
distilled water.
.
METHOD:

(a) Determination of Lead

In a covered 300-mI. Erlenmeyer flask dissolve 1 g. of the alloy in 20 mI.

of H 2 S0 4 . (sp.gr. 1.84); heat the solution nearly to boiling until the metal is
completely decomposed and the PbS0 4 is white (this'IDay -take 30 minutes or
more) and finally boil for several minutes. Allow to cool, but not below 60 C.,
and then add slowly 50 mI. of water while the solution is agitated. Heat to
boiling for several minutes in order to insure complete solution of antimony
sulfate. Allow the PbS0 4 to settle out until the solution is clear, not letting
the temperature fall below 60 C. If the liq1,lid does not clear quickly, it
must be heated longer. When clear, pour the solution through a weighed
porcelain Gooch crucible with asbestos mat, decanting the solution as completely as possible without allowing more than a very small amount of PbS0 4
14

See "The Analysis of Lead, Tin, Antimony and CoppeF," by D. J. Demorest,

J. Ind. and Eng. Chern., 5, 842 (1913); "Rapid Analysis of Alloys for Tin, Antimony and
Arsenic," by F. A. Stief, J. Ind. and Eng. Chem"..7,"'211 (1915); and W. B. Price and
R .. K. Mea?e. "Technical Analysis of Brass," p. 182, John Wiley and Sons, N. Y., 1917.

ANALYSIS OF NON-FERROUS ALLOYS

1385

to go over into the crucible. Now add 10 ml. more of H 2S0 4 (sp.gr. 1.84) to
the PbS0 4 in the original flask, and boil for several minutes. Cool, add slowly
30 ml. of water, and again heat to boiling for a few minutes; allow the solution
to cool to about 60 C. and completely transfer the PbS0 4 to the Gooch crucible.
Wash with lead acid, retaining the filtrate and these washings for the copper
determina,tions. Remove the beaker containing these solutions and wash out
the lead acid with dilute alcohol; set the Gooch crucible inside a porcelain
crucible; dry and ignite for five minutes at the full heat of a Tirrill burner; cool
and weigh as PbS0 4, which contains 68.33% lead.
(b) Determination of Copper

To the filtrate from the PbS0 4, which contains the copper and which has
been transferred to an Erlenmeyer flask, add NH 40H (sp.gr. 0.90) until the
solution is slightly alkaline; then add 2 ml. of H 2S0 4 (sp.gr. 1.84); heat nearly
to boiling; add 2 g. of N a 2S03 and when this is entirely dissolved add 1 g. of
KCNS dissolved in 10 m!. of water. Shake well and allow the precipitated
CuCNS to settle for one hour while the solution is kept hot. Filter on a close
filter paper, wash with cold water and ignite paper and precipitate in a porcelain
crucible. Proceed by one of the two following methods:
(1) Dissolve in HN0 3 (sp.gr. 1.42), add 5 fil. H 2S0 4 (sp.gr. 1.84); evaporate'
until fumes of H 2S0 4 are evolved. Dilute to about 100 ml. with distilled
water: Add 1.5 ml. of HN0 3 (sp.gr. 1.42) and determine copper by electrolysis.
See the" Determination of Copper by the Electrolytic Method, Analysis of
Manganese Bronze."
(2) Dissolving the ignited precipitate in HN0 3 (sp.gr. 1.42), as in (1);
boil to expel nitrous fumes, neutralize with N a 2 C0 3 and determine volumetrically with cyanide as in A of " General Method."
NOTE.-If the amount of copper is very small it may escape detection by this
method, in which case it should be determined as in A of "General Method."

B.

DETERMINATION OF ARSENIC, ANTIMONY AND TIN

Apparatus. Arsenic Still.- A condenser is made from glass tubing in the

form of a letter S, about 18 in. long and Yz in. in inside diameter, tapering to
about 74' in. at the upper end and to about Ys in. at the lower end. One curve is
nearly filled wit~;ater, and is submerged in cold water in a 500-ml. beaker.
The lower end <lips into about 75 ml. of water in a 300-ml. beaker and the
upper end is connected by a delivery tube with a 300-ml. Florence flask, closed
with a rubber stopper which is fitted with a delivery tube and with a thermometer reaching to about 1 in. above the surface of the liquid in the flask.
Solutions Required. Iodine.- Dissolve 10.7 g. of iodine in 50 ml. of distilled water which contains 20 g. of KI in solution and dilute to 1000 m!. with
distilled water. Standardize against pure tin having exactly the same conditions for titration as are specified for the analysis of the alloy. Each milliliter
is equivalent to approximately 0.00500 g. of tin or 0.00315 g. of arsenic.
Potassium Permanganate.-Dissolve 2.7 g. of KMn04 in distilled water,
filter through asbestos, and dilute to 1000 ml. with distilled water. Standardize
against pure antimony having exactly' the ~ame conditions fo~ titration as are

1386

ANALYSIS OF NON-FERROUS ALLOYS

specified for the analysis of the alloy._ . Each milliliter is equivalent to approximately 0.00500 g. of antimony or 0:00465 g.lof iron.
Ferrous Ammonium Sulfate.-Dissolve 12.4 g. of ferrous ammonium sulfate crystals in 950 mL of distilled water. and add 50 mL of H 2 S0 4 (sp.gr. 1.84).
Stare h.-To 1000 mL of boiling water add ~ cold suspension of 6 g. of
starch in 100 mL of distilled water; cool, add a ~ew drops of chloroform and
mix thoroughly.
.

(a) Determination of Arsenic

Method.-In a 300-mL Florence flask, dissolve 0.5 g. of the alloy in exactly
8 mr. of H 2S0 4 (sp.gr. 1.84). The metal must be finely divided, preferably in
the form of thin foil-like shavings, although sawings or very fine drillings may
be used. Heat the solution to boiling; cool, add about 5 mL of water and a
bulk of about 0.5 mL of clean granulated pumice stone and boil the covered
solution very gently for about five minutes, or until the strong odor of SO! can
no longer be detected. Finally, being careful to have 5 mL of water present,
cool the solution to about 18 C. and cautiously add 20 mL of HCI (sp.gr. 1.18).
Insert in the flask a rubber stopper fitted with a thermometer and delivery tube
and connect the delivery tube with the" is "condenser. Heat the solution to
gentle boiling for from 10 to 15 minutes, keeping the vapor temp_erature at
107 C. for at least 5 minutes. Wash out the condenser into a 300-mL beaker
and add an excess of about 2 g. of NaHCO a Bring the volume of the solution
to about 200 mr.; warm to about 27 ~C., add 5 mL of starch solution and titrate
with standard iodine solution to the appearance of a deep blue color.
N OTE.-A blank determination should be made on corresponding amounts of reagents
treated as above, and the results should be cO,rrected accordingly.
.

. (b) Determination of Antimony

Cool the solution which remains in the flask after the arsenic distillation;
add about 130 mL of cold, recently-boiled distilled water to which has been
added about 3 mL of HCl (sp.gr. 1.18) and titrate with standard KMn04
solution to the appearance of a decided per:qJ.anerit pink color. From a burette
add an excess of standard ferrous ammonium sulfate solution until the pink
color is discharged and then titrate with standard KMn04 solution to the
reappearance of a pink color.
(c) Determination of Tin

-~

If the sample does not contain 14% (or 0.07 g.) of antimony, add enough
dissolved SbCl a to the solution to bring the contents of antimony up to about
0.07 g. for the purpose of insuring perfect 'reduction of the tin and uniform
conditions for the titration. Then add exactly 6 mL of H 2S0 4 (sp.gr. 1.84)
and 60 ml. of HCl (sp.gr. 1.18) and add about!6 in. of clean, soft No. 14 pure
iron wire cut into 2 in. lengths and cleaned with dilute HCI just before using.
Boil gently for 30 minutes, add about 6 in. more wire and boil 30 minutes longer.
Remove from heat, close flask loosely with a rubber stopper and allow about
two minutes for all air to be expelled by hydrogen and acid vapors. Close
flask tightly and quickly place it in cold wa1&rj"'cooling the solution to about
20 C. Transfer the solution quickly to a' 500-roL beaker (leaving the iron

ANALYSIS OF NON-FERROUS ALLOYS

1387

wires in the flask) and rinse flask and contents with 150 mI. of cold, recentlyboiled distilled water. Add rinsings to main solution in the beaker, bring
final volume to about 300 to 350 mI., add 5 ml. of starch solution and titrate
quickly with standard iodine solution ~o the appearance of a strong blue color.
NOTES.-A blank determination should be made on corresponding amounts of
reagents treated as above, and the results should be corrected accordingly.
If the copper content of the alloy is as high as about 3% or over, it prevents a good
end-point in the tin titration when the tin percentage is low. Under such conditions,
just before titrating, add about 1 g. of KI crystals, stir until nearly dissolved and titrate
immediately. If allowed to stand too long at this point, trouble may be encountered
due to the precipitation of some of the other constituents.
Other forms of pure iron may be employed, but the wire as above specified has been
found convenient.
For accurate work it is desirable to reduce and titrate the solution of tin in an
atmosphere of CO 2

THE DETERMINATION OF IMPURITIES IN TIN BASE AND

LEAD BASE ALLOYS 15
General Considerations.-The impurities that are usually present in tin
base and lead base alloys vary from 0.5% to traces. The following methods
are given on the assumption that an impurity does not exceed 0.10%. When
an impurity is present to a greater extent, a smaller amount of sample may be
taken.
THE DETERMINATION OF LEAD, BISMUTH, IRON, ALUMINUM
AND NICKEL IN TIN BASE BEARING METAL
DETERMINATION OF LEAD

Solutions Required. Potassium Hydrosulfide Stock Solution.-Weigh

200 g. of KOH into an ~.90-ml. beaker, dissolve in about 800 m!. of distilled
water and pass H 2S for about one hour. Make up to a volume of 2000 m!.
with distilled water, mix and allow to stand over night.' Filter off portions
as required for use.
Potassium Hydrosulfide Wash Solution~-Mix one part of the potassium
hydrosulfide stock solution with two parts of distilled water.
Lead Acid.-See "Determination of Lead as Sulfate, Analysis of Manganese Bronze."
Acid Ammonium Acetaie.-See "Determination of Lead, Analysis of
Alloys of Lead, Copper, Antimony and Tin" under General Methods.
Method.-Weigh 5 g. of the sawings into a 600-m!. beaker, add 30 m!. of
HCI (sp.gr. 1.18), warm on a hot plate and add HNO a (sp.gr. 1.42), a few drops
at a time, until the alloy is dissolved and the tin is completely oxidized. Dilute
to a volume of 100 m!. with cold water and make strongly alkaline with a cold
saturated solution of KOH. Avoid a large excess of the KOH solution. Immediately add 70 m!. of the potassium hydrosulfide stock solution, stir and
allow to stand in a warm place until the precipitate settles and the supernatant,
solution is a clear lemon yellow. Filter and wash with the warmed potassium
hydrosulfide wash solution.
15 Methods of the National Lead Company Research LaboratOrIes.

1388

ANALYSIS OF NON-FERROUS ALLOYS

Throw the filter paper and 'precipitate back into the original beaker, add
5 ml. of HNO. (sp.gr. 1.42), heat until the paper,is pulped and the sulfide dissolved, add 40 ml. of lead acid and evaporate to fumes of H 2S0 4 Just before
the solution begins to fume it will become dark colored due to the action of the
H 2S0 4 on the filter paper. At this point add a; \:lrop or two of HNO. (sp.gr.
1.42) to the hot solution and continue to fume. Repeat the addition of HNO.
until the solution shows no brown coloration even upon strong fuming. Cool,
add a few mI. of water and take to fumes a second time. Again add a few IllI. of
water and fume. Finally cool, add 35 mI. of water, bring to boiling, cool, and
allow to stand several hours. Filter and wash with lead acid.
Dissolve the PbS0 4 with 30 mI. of hot ammonium acetate solution, wash
with hot water, dilute to a volume of 150 mI., bring to boiling, add 10 ml. of a
. saturated solution of K 2 Cr 207 and allow to stand in a warm place several
hours. Filter on a weighed porcelain gooch crucible and wash several times
with hot water and once with alcohol. Dry at 110 C. for one hour. Cool
and weigh as PbCr04. Multiply by 0.6375 for lead.
\
Accuracy.-Duplicate determinations should check within 0.008% of lead.
N oTE.-Ir: p~eparing the Gooch crucible it is advisable to fit a small filter pape~ in
the bottom of the Gooch crucible before adding the asbestos and treat exactly as in the
determination.
DETERMINATION OF BISMUTH

Method.-Dissolve 10 g. of sawings with 50 wI. of HCI (sp.gr. 1.18) and

sufficient HNO a (sp.gr. 1.42) to oxidize the tin, dilute to a volume of 150 mI.,
make alkaline with a saturated solution of KOH, add 80 ml. of the KSH stock
solution, stir, allow to stand, filter and wash with the KSH wash solution.
(See" Determination of Lead.")
Wash the precipitate back into the beaker, place under the funnel and pour
through the filter 15 ml. of hot HNOa (1 : 2). Bring to boiling, filter through
the same paper and \vash with hot water. If a black residue still remains,
transfer to a small porcelain crucible, dry, ignite ~t a low heat, dissolve in 2 ml.
of HNO a (sp.gr. 1.42) and transfer to the main solution. Evaporate to a vol- .
ume of 50 ml. and allow to stand'in a warm;place for one hour. If a precipitate'
of H 2Sn03 forms, filter and wash with hot HNO. (1 : 50). Discard the residue.'
Make the solution just neutral with NH 40H (1 : 2), add 5 ml. of HCI
(1 : 9) in excess, dilute to a volume of 400 ml. with..hot water and allow to
stand in a warm place at least two hours, preferably over- night. Filter on a
weighed porcelain Gooch crucible, wash several times with hgt water, once with
alcohol, dry at 110 C. for one hour, cool and weigh as BiOCI. Multiply by
0.8024 for bismuth.
Accuracy.-Duplicate determinations should check within 0.005% of
bismuth.
I

NOTES.-1. When a relatively large amount of a lead base alloy or a tin base alloy
containing considerable lead is dissolved in a HCl solution, lead chloride will separate out
before the alloy is dissolved, if insufficient solution is present, and retard further solvent
action. For the solution of alloys containing lead and sj.lbsequent procedure see
"Determination of Anti'mony, Zinc and Copper in Solder."
2. If a precipitate of H 2SnO. forms when the-sutfides are dissolved in HNO., antimony may be present in solution, and precipitate out with the BiOCI. In that case,

ANALYSIS OF NON-FERROUS ALLOYS

1389

filter the BiOCI on a paper, dissolve in 10 ml. of hot HCI (1 : 1), wash, dilute to 100 mI.,
pass H 2S, filter, wash, treat the sulfides, with 10 ml. of the KSH stock solution mixed
with an equal quantity of water, filter and wash. Proceed as described above.
3. See Note under "Determination of Lead."
DETERMINATION OF IRON

Method.-Dissolve 12.5 g. of alloy as described under" Determination of

Bismuth." Add about 2 g. of KCI and evaporate to near dryness. Add 20
ml. of HCI (sp.gr. 1.18), dilute to 50 ml. with water, bring to boiling and pass
H 2S. As the sulfides begin to form, keep adding hot water from time to time
until the volume of solution is finally brought up to 400 ml. Continue passing
HzS for 30 minutes. Transfer to a 500-ml. graduated flask, cool to room t,emperature, make up to the mark with a 2% solution of HCI through which H 2S
has been passed, mix and filter off 400 ml. through a dry paper discarding the
first 15 or 20 ml. of solution.
Boil the filtrate to expel H 2S, add 3 m!. of HNO a (sp.gr. 1.42) to oxidize the
iron, make just alkaline with NH 40H (sp.gr. 0.9), using methyl red as an indicator; boil for two minutes, filter and wash several times with a 2% solution of
NH 4 CI, and once with hot water. (The filtrate may be used for zinc. See
" Determination of Zinc in Solder.") Dissolve the precipitate from the filter
with 15 ml. of H 2S0 4 (1 : 4), wash with hot water and evaporate to fumes of
H 2S0 4, Cool, add 25 ml. of water, bring to boiling, cool and allow to stand
for any lead to separate that may noLPave been precipitated by the HzS.
Filter and wash with cold dilute H 2se)'(1 : 9). To the filtrate add a slight
excess of a strong solution of KMn04, pass through a Jones reductor and titrate
with approximately 0.006 N KMn04. (Bee" Determination of Iron, Analysis
of Manganese Bronze.")
Accuracy.-Duplicate determinations should check within 0.003% of iron.
NOTEs.-l. See Note 1 under "Determination of Bismuth," and "Determination of
Zinc, Determination of Antimony, Zinc and Copper in Solder."
.
2. If iron only is desired, the solution may be made slightly alkaline and allowed to
stand an hour or so in a warm place. The addition of
methyl red may be dispensed with.
DETERMINATION OF ALUMINUM

Apparatus Required. Anodes.-Platinum wire

spirals.
Cathodes.-The cathode is metallic mercury
contained in a glass receptacle (Fig. 182) in which
has been fused at A a small piece of fine platinum
wire looped at the outer end. Contact is made
by means of a piece of copper wire. About 12 m!.
of mercury is used forming a column about 1.5
inches high. The apparatus will conveniently
hold, exclusive of the mercury, 125 mI. of solution. FIG. 182. Apparatus for DeMethod.-Proceed exactly as in " Determinatermination of Aluminum.
tion of Iron" up to the point where liny lead that
may not have been precipitated as a sulfide has been separated as a sulfate.
Make the filtrate just neutral with NH 4 0H (1 : 2) and then acid with 10
ill!. of H 2S0 4 (1 : 9) in eXeeS!). Dilute to a volume of 80 mI., transfer to the

1390

ANALYSIS OF NON-FERROUS ALLOYS

receptacle containing the mercury as ()a~hode, and e~ectrolyze over night using a
current of 0.5 ampere. Transfer the electrolytelto a 400-ml. beaker, wash the
mercury with water, combine electrolyte and washings, add 5 g. of NH 4 CI,
make just neutral with NH 40H using methyl rep' as an indicator, boil for two
minutes, filter and wash with hot water. Ignite in a weighed porcelain crucible
at first over a low, then over a strong Bunsen flatp_e. Cool and weigh as Al 20 a
Multiply by 0.529 for aluminum.
.
Accuracy.-Duplicate determinations ,should check within 0.003% of
aluminum.
NOTEs.-l. See Note 1 under "Determination of Bismuth."
2. Phosphorus, if present, will interfere. For its correction see "Determination of
Aluininum, Analysis of Manganese Bronze," Method (a).
\
DETERMINATION OF NICKEL

Method.-Proceed exactly as in " Determination of Iron" up to the point

where 400 ml. of solution has been filtered off from the sulfides.
Boil to expel H 2S, add 3 ml. of HNO a (sp.gr. 1.42) to oxidize iron, evaporate
to a volume of about 100 mi., add 10 m!. of a 20% solution of sodium potassium
tartrate, nearly neutralize with NH 4 0H (sp.gr. 0.9), heat nearly to boiling,
add 10 m!. of a 1 % solution of dimethylglyoxime and then NH 40H in slight
excess. Allow to stand in a warm place at least 30 minutes, filter on a weighed
Gooch crucible and wash with 1% solution of NH 4 0H. Dry at 110 C. for
two hours. Cool and weigh as NiC sH 14 N 40 4. Multiply by 0.2032 for Nickel
(see " Determination of Nickel, Analyses of Bronze Bearing Metal ").
Accuracy.-Duplicate determinations should check within 0.001% of
nickel.
NOTEs.-l. The filtrate from the iron and aluminum separation with NH 40H may
be used for the determination of nickel in which case the addition of sodium potassium
tartrate is unnecessary.
2. See Note under "Determination of Lead."
I
I

DETERMINATION OF ANTIMONY, ZINC ANP COPPER

IN SOLD,ER 16
(ALLOY OF LEAD AND TIN)
DETERMINATION OF ANTIMONY'17

A. General Method
Solutions Required. Mixed Acid.-Mix 400 ml. of HCI (sp.gr. 1.18) and
500 ml. of water and then add 100 ml. of'HNO a (1.42).
Potassium Hydroxide Solution.-Dissolve 100 g. of KOH in 500 ml. of
water.
i
Copper Solution.-Dissolve 2.0 g. of CuS04'5H 20 in 500 ml. of water.
KSH Stock Solution.-See "Determination of Lead, etc., in Tin Base
Bearing Metal."
I

Methods of the National Lead Company Research Laboratories.

See "Specifications for Solder Metal," 1936-'BOok of A. S. T. M. Standards, part I,
p.695.
16
17

ANALYSIS OF NON-FERROUS ALLOYS

1391

Potassium Iodide Solution.-Dissolve 20 g. of KI in 500 mI. of water.

Sodium Thiosulfate.-8ee "Determination of Lead, Analysis of. Alloys
of Lead, Copper, Antimony and Tin," General Methods.
Method.-Weigh 5 g. of the sawings into a 250-ml. beaker, add 150 ml.
of the mixed acid and heat over a low flame on the hot plate. When PbCl 2
begins to form, decant the clear solution into a 400-ml. beaker, again add 150 ml.
of the mixed acid to the undissolved alloy and continue heating until the solution is complete. Two 150 mI. additions of the mixed acid should suffice to
completely dissolve the alloy without the formation of appreciable PbCh in the
hot solution. Combine the two solutions, add about 2 g. of KCI, evaporate
to a volume of 100 mI., cool in ice water and separate the PbCl 2 by decantation,
washing three times with cold HCI (1 :'2). Evaporate to a volume of 40 ml.
Cool, allow to stand a few hours or preferably over night, decant the clear
solution into a 600-mI. beaker, wash the PbCh by decantation with cold HCI
(1 : 2) pouring the washings through a small filter, evaporate the solution to
near dryness, make just alkaline with KOH solution, add 12 g. of oxalic acid
and lO mI. of copper solution, dilute to a volume of 400 mI., bring to boiling,
and pass H 28 for 45 minutes, keeping the solution hot and the volume constant
during the passage of the H 2S. Filter and wash with a hot dilute solution of
oxalic acid saturated with H 2S.
Wash the precipitate from the filter paper back into the beaker, add 10 ml.
of KSH stock solution and 10 mI. of KOH solution, and digest on the water
bath for 5 or 10 minutes. Filter through the same paper into a 500-ml. 'Erlenmeyer flask and wash with hot water containing a little of the KSH solution.
Discard the residue.
'
Proceed exactly as in " Deter~ination of Antimony, Analysis of Alloys of
Lead, Copper, Antimony and Tin," General Methods.
Accuracy.-Duplicate determinations should check within 0.012% of
antimony.
B. Rapid Method for Control Work
Solutions Required. N/20 Potassium Permanganate Solution.-Dissolve
1.58 g. of KMn04 in 1 liter of water. For its preparation and standardization
see " Determination of Antimony, Analysis of Alloys of Lead, Copper, Antimony and Tin," Rapid Methods for Control Work.
Method.-Weigh 3 g. of the sawings into a 500-mI. Florence flask, add 20
mI. of H 2S0 4 (sp.gr. 1.84), place over a free flame and heat until completely
decomposed. Cool, add 50 mI. of distilled water and boil down to a volume of
40 mI. If 80 2 is not completely expelle'd, dilute slightly and continue boiling
until it is. The volume at this stage should be 35 to 40 m!. Cool, add 15 mI.
of HCI (sp.gr. 1.18) and boil vigorously for five minutes. Add 200 mI. of cold
water, immerse in cold tap water until the solution is cold and titrate with
Nj20 KMn04 solution to the first tinge of a permanent pinl):.
Run a blank determination on a mixture composed of pure tin and pure lead
approximating the composition of the sample. Correct for same.
Accuracy.-Duplicate determinations should check within 0.01% of
antimony.
NOTE.-The antimony content of the metals used for the blank should be known.
The best brands of tin on the market may carry a small -amount of antimony; the best
brands of lead may carry as a maximum 0.004% of antimony.

1392

ANALYSIS OF NON-FERROUS ALLOYS

DETERMINATION OF:

ziNc

Solutions Required. Citric Acid Solution.LDissolve 250 g. of the crystals

in 500 mL of water.
I
Standard Zinc Solution.-Dissolve \,xactly Q.1 g. of U. S. Bureau of Standards pure zinc in 5 mL of HCl (sp.gr. 1.18) and dilute to exactly 1000 mL with
water.
\
.
Potassium Ferrocyanide Solution.-Dissolve 2 g. of K4Fe(CN) 6' 3H 20 in
100 ml. of water.
.
Method.1 8-Weigli 6.25 g. of the sawings into a 250-mL beaker and proceed
exactly as in " Determination of Antimony" up to the point where the solution
is evaporated to near dryness. Add 2(}mL of HCl (sp.gr. 1.18), dilute to 50 mL
with water, bring to boiling and pass H 2S, adding hot water from time to time
during the passage of the H 2S until the volume is brought up to 400 mL Continue passing H 2S for 30 minutes. Transfer to a 500-mL graduated flask, cool
to room temperature, make up to the mark with a 2% solution of HCI through
which H 2S has been passed, mix and filter off 400 mL through a dry paper discarding the first 15 or 20 ml. of solution.
To the filtrate add 5 ml. of H 2S0 4 (sp.gr. 1.84) and evaporate fumes of
H 2S04. Cool, add 35 ml. of water, bring to boiling, cool, and allow to stand.
Filter and wash with cold dilute H 2S0 4 (1 : 9). Discard the residue. Bring
the filtrate to boiling, add 3 ml. of HNO s (sp.gr. 1.42), add 2 g. of NH 4Cl, make
slightly alkaline with NH 40H (sp.gr. 0.9), again bring to boiling and allow to
stand in a warm place for an hour or so. Filter and wash.
Boil the filtrate to expel NH 40H, add 10 ml. of citric acid solution, dilute
to l:I. volume of 200 mI., again bring to boiling and cautiously add CaCO g in
small portions at a time until about 1 g. of calcium citrate separates. Pass
H 2S through the solution until cold and allow to stand several hours, part of
the time on a water bath. Filter and wash with a 2% solution of ammonium
thiocyanate.
Wash the precipitate back into the beakeri with as little water as possible,
set the beaker under the funnel and pour tfn:ough the filter 15 mL of HCl,
(1 : 1). Replace with a clean beaker; bring the solution to boiling, filter
through the same paper and wash with hot water. Keep the volume below
100 m!. Reserve as solution No. 1. At the same time the sample is being run
a blank determination should be carried along. Reserve as solution No.2.
Cool both solutions to room temperature.
Transfer solution No.1 into a 100-mL Nessler tube, ~dd2 mL of the ferrocyanide solution, make up to mark and mix. Transfer solution No.2 into
another 100-mL Nessler tube, add 2 mL of the potassium ferro cyanide solution,
make up to mark, mix and add the standard zinc solution, a few drops at a
time, until the turbidities match. Mix between each addition of the standard.
Comparison is best made by setting the tubes over a sheet of fine print and
looking down the tubes. (See" Determination of Zinc, Tentative Methods of
Chemical Analysis of Pig Lead.")
Accuracy.-Duplicate determinations should check
,. within 0.001
. % of zinc.
18

See "Determination of Small Quantitiea.,of.Zinc," by M. Bodansky, J. Ind. and

Eng. Chern., 13, 696 (1921).

ANALYSIS OF NON-FERROUS ALLOYS

1393

NOTE.-The turbidimetric test is very sensitive for amounts f,)f zinc not exceeding
0.0005 g. If the ~ample contains over 0.01 % of zinc take such e,ll aliquot portion as
will not contain more than this amount.
DETERMINATION OF COPPER

Solutions Required. Standard KeN Solulion.-DissolVe 4 g. of KCN in

1000 all. of water. Standardize against pure copper exactly as in the determination. Each m!. of the sta~dard solution should equal about 0.001 g. of
copper.

KSH Slock dnd Wash Solutions.-See "Determinatioll of Lead, etc. in

Tm Base Bea:ring MetaL"
Method.-Weigh 5 g. of the, sawings into a 250-m!. beaker and proceed
exactly as in " Determination of Antimony J1 up to the point where the lead
has been separated as PbCI 2 Dilute the filtrate to a volume of 150 mI., make
alkaline with a Mturated solution of KOH, add 70 m!. of KSH stock solution,
stir, allow to settle, filter and wash with KSH wash solution. Throw the filter
back into the beaker, add 5 m!. of HNO, (sp.gr. 1.42) and 5 m!. of H,SO.
(sp.gr. 1.84) and evaporate to fumes of H 2S0 4 Remove carbonaceous matter
with HNO. (sp gr. 1.42) and again fume. Cool, add 50 m!. of water, bring to
boiling, cool, allow to stand, filter and 'wash with cold water. (See It Determination of Lead, Analysis of Lead, Bismuth, etc. in a 'Tin Base Bearing
Meta!.") Discard the residue.
'Varm the filtrate, pass H 2S for 15 minutes, filter and wash with H 2S water.
Wash the precipitate back into the same beaker, add 10 m!. o' KSH stock
solution, digest all the hot plate for a few minutes, filter through the same paper,
and wash hvo or three times with KSH wash solution and twice 'with H 2S water.
Transfer the-residue to a small porcelain crucib1e, dry, ignite, cool, add 2 ~l.
of HXO. (1 : I) ~nd evaporate to dryness. Cool, add I mi. of HNO, (I : I),
neutralize with p. saturated solution of sodium carbonate, add 4 drops of
NH.OH (sp.gr. 0.9) ard titrate with standard KeN solution to the disappearance of the blue color.
Accuracy.-Duplicate determinations should check within 0.004% of copper.
NOTE.-For anlounts of copper exceeding 0.010 g. weigh out tt smaller charge or
detennine t.he copper electrolytically. (See" Determination of Copper, Analysis of
Alloys of Lead, Tin, Antimony and Copper," General Methods.)

DETERlIUNATIOX OF TIN IN BATTERY PLATE METAL

I

\AL~O"I )'1 L~.h.Y> .h.ND

Awn:1o.\mn)

Apparatus_Required.-Insert in, a No.6 sulfur-free rubber stopper a bent

outlet tube haVIng an outside diameter of about 3.5 mm. and a capacity of
about 12 m!. A bent pipette will answer for the purpose. ,Vhen the stopper is
inserted into the JIlouthl of a 500-mI._Erlenmeyer flask_the short end of the tube
should just pass thr,ough the stopper while the other erid should reach to a level
~ in. from the bo~tom of the flask and be... ! in. distant from it. 19
I

lDCraig, A., "Notes on Chemical Analysis," p. 114, The Chemical Publishing Co.,

1924,

1394

ANALYSIS OF NON-FERROUS ALLOYS

Solutions Required. Standard Iod,ine Solution and Starch Indicator.-See

" Analysis of Alloys of Lead, Copper, Antimoh)t and Tin," Rapid Methods for
Control Work.
.
Sodium Bicarbonate Solution.-'- Dissolve 1.Q g. of N aHCO a in 100 mI. of
recently boiled distilled water contained in a 150 mI.-beaker.
Method.-Weigh 5-10 g. of sawings into a 30,0-mI. cone flask, add 25-30 ml.
of H 2S0 4 (sp.gr. 1.84) and heat on a bare hot plate until completely decomposed.
Cool, add 100 ml. of water and 20 ml. of HCI (sp.gr. 1.18), introduce four
pieces of iron and heat on the hot plate over a low flame until the iron is almost
but not entirely dissolved. A reduction of 30 minutes should suffice. Filter
off the precipitated metals through glass wool into a 500-ml. Erlenmeyer flask
in which has been placed four pieces of iron and wash flask and funnel with hot
water. The volume at this stage should be about 250 ml. Connect up the
outlet tube and reduce as before. When reduction is complete, set the bicarbonate solution so that the free end of the outlet tube is immersed in it. When
gas bubbles start to pass through the NaHCO a solution, remove both flask and
beaker from the stove, allow to cool at room temperature for about 10 minutes
and then cool the flask in cold tap water. Remove the outlet tub.e, withdraw
the undissolved iron with a magnetized steel rod or file, wash both removed
iron and rod with water, add 5 ml. .of starch solution, and titrate with the
standard iodine solution to the first tinge of blue.
Run a blank by treating 10 g. of lead sawings exactly as the alloy was treated
and make a correction for same. /
Accurac:y,.-Duplicate titrations should not disagree more than 0.1 ml.
NOTES.-l. Ordinary strap iron used for strapping shipping cases may be used for
the tin reduction. Each piece should be 2 in. long and Y2 in. wide, weigh about 2 g.
and be free from grease, lacquer and rust.
2. The free end of the outlet tube should be kept immersed in the NaHC0 3 solution
during the whole cooling operation.

ANALYS~S

OF .ZINC BASE DIE-CASTI:J:iG ALLOY 2Q

ALLOY OF ZINC, COPPER AND ALUMINUM

A.

21

DETERMINATION OF COPPER' AND ALUMINUM

Electrodes for the Electrolysis of Copper.-See " Determination of Copper

by the Electrolytic Method, Analysis of Maftganese Bronze."
Apparatus for the Electrolytic Separatio~ of Zinc.22_.See " Determination
of Aluminum, Determination of Lead, Bismuth, etc. in Tin Base Bearing
Metal."
Methods of the National Lead Company, R,esearch L'boratories.
For specifications, see 1935 Book of A. ~ . ..T;-M. Tentative Standards, p. 369.
22 See "Determination of Aluminum and Magilesium in Zinc Base Die Casting
Alloys," by C. M. Craighead, J. Ind. and Eng. Chern., Anal. Ed., 2, 188 (1930).
20

21

ANALYSIS OF NON-FERROUS ALLOYS

1395

Solutions Required. Acidulated H 2 S Wash Water.-Dilute 10 ml. of

H2S04 (sp.gr. 1.84) to 500 ml. with water and pass H 2S for 10 minutes.
Method.-Weigh 4.0 g. of. the sawings into a 300-ml. beaker, add 100 ml.
of water and 15 ml. of H 2S0 4 (sp.gr. 1.84). (Add the acid in small doses.)
When decomposition is complete, filter and wash with hot water. Hold the
residue. Through the filtrate pass H 2S, filter and wash with acidulated H 2S
wash water. Throw both filters into the beaker the filtration was made from,
add 5 ml. of HNO s (sp.gr. 1.42) and 5 ml. of H 2S0 4 (sp.gr. 1.84) and evaporate
to fumes of SOs, adding a few drops of HNO a (sp.gr. 1.42) from time to time
until the solution shows no evidence of a brown coloration. Finally cool,. take
up with 100 ml. of water, bring to boiling, add 3 ml. of HNO s (sp.gr. 1.42),
wash into a 200-ml. lipless beaker, cool, dilute to a volume of 150 ml. and
electrolyze over night with a current of 0.5 amperes. Weigh as metallic copper
Run in duplicate.
Boil the filtrate from the CuS separation to expel H 2S, oxidize any iron
present with a few ml. of H 20 2 and continue boiling until the solution has
attaIn.ed a volume of about 75 mL Cool; transfer to the receptacle containing
the mercury and electrolyze over night with a current of 1.5 amperes.
Transfer the electrolyte to a 400-mL beaker, wash the anode and the mercury with water, combine electrolyte and washings, add 5 g. of NH 4Cl make
just neutral with NH 40H (1 : 1) using methyl red as an indicator, boil for 2
minutes, filter, wash, ignite over a blast to constant weight and weigh as AbOa.
Multiply by 0.5291 for AI. Run in duplicate.
NOTES. -1. Should the copper content of the alloy be low (0.1 % or less) the determination may be made by a cyanide titration.
2. If the separation of zinc is incomplete a second separation of the Al(OH)a will
have to be made. The presence of zinc may readily be detected by passing H 2S through
a portion of the filtrate.
3. A small amount of iron may be present with the Al 20 a. In such case, triturate,
add 5 ml. of HCI (sp.gr. 1.18), evaporate to dryness, take up with 5 ml. of HCI (sp.gr.
1.18), transfer to a IOO-ml. graduated flask, make up to the mark, mix, tap off 25 ml.
into a 100-ml. graduated Nessler tube, add 2 ml. of NH 4CNS (1: 5), make up to the
mark, mix and match agail1llt a standard solution of iron (1 ml. =0.0001 g. of iron)
under similar conditions.

B.

DETERMINATION OF MAGNESIUM

Method.-Weigh 10 g. of the sawing'l.into a 400-ml. beaker, add 100 ml. of

water and 20 mL of H 2S0 4 (sp.gr. 1.84). (Add the acid in small doses.) When
decomposi~ion is complete, filter and wash with hot water.
Discard the
residue. Evaporate the filtrate to a volume of about 75 ml. and proceed as in
" A " for the electrolytic separation of aluminum and zinc.
After separating most if not all of the zinc, transfer the electrolyte and
washings to a 600-ml. beaker, add about 5 g. of tartaric acid (in solution), 5 g.
of NH 4Cl and 10 ml. of a saturated solution of microcosmic salt. Make the
solution alkaline with NH 40H (sp.gr. 0.9) adding sufficient NH 4 0H to have
present 15 ml. for every 100 ml. of solution. Stir vigorously and allow to stand
over night. Filter and wash with cold dilute NH 4 0H (1 : 9), ignite and weigh
as Mg 2Pi>7. Multiply by 0.2184 for Mg. Run in duplicate.

1396

ANALYSIS OF NON-FERROUS ALLOYS

NOTE.-Shouid the solution after standing over night appear clear or should transparent,crystals form, the magnesium has hot' been coIJtpletely precipitated. A vigorous
stirring at this point should completely precipitate the magnesium. Allow to stand
several hours or preferably over night and proceed as .outlined above.
C.

DETERMINATION OF CADMIUM AND IRON

Weigh 30 g. of the1sample into a 600-ml. beaker, add 100 ml. of water'and

34 ml. of H 2S0 4 (sp.gr. 1.84). (Add the acid in small doses.) When decomposition is complete, add 10 g. of Na 2S04, dilute to a volume of 350 ml. with
water, warm and pass H 2S for 30 minutes. Allow to stand in a warm place for
2 hours and again pass H 28. Filter and wash. ,Reserve the filtrate (Filtrate
No.1).
Throw the paper containing the precipitate back into the original beaker,
add 10 ml. of HNO a (sp.gr. 1'.42) and 10 ml. of H 2S0 4 (sp.gr. 1.84) and evaporate to fumes of SOa. Clear up any discoloration due to carbonizing of the
paper by adding a few drops of HNO a (sp.gr. 1.42) to the hot solution. Cool,
add 5 ml. of water, bring to boiling, cool, allow to stand an hour or so, filter and
wash with cold H 2S0 4 (1 : 9). Discard the residue.
To the filtrate add 5 m!. of HNO a (sp.gr. 1.42), dilute to a volume of 150 ml
with water and electrolyze for the removal of copper as described in " A."
Remove and wash the electrodes, combine electrolyte and washings in a 600-ml.
beaker and evaporate to fumes of :;;Oa. Cool, add 7 g. of Na 2S04, dilute to a
volume of 300 ml. with water, warm and pass H 2S as before. Allow to stand
as before. Again pass H 2S for 30 minutes, filter and wash with H 2 S water.
Reserve the filtrate (Filtrate No.2).
Dissolve the precipitate by pouring 10 ml. of hot HNO a (1 : 1) through the
paper, wash with hot water, add 5 ml. of H 2S0 4 (sp.gr. 1.84) and ev~porate to
strong fumes of SOa. Cool, add a few ml. of water and again evaporate to
fumes. Repeat the treatment with water and subsequent fuming to remove the
last traces of nitric acid. Finally cool, take up with 50 ml. of water, bring to
boiling, cool, neutralize with NH 40H (1 : 2) and then make acid with H 2S0 4
(1 : 9) having 3 ml. in. excess. Dilute to ~a volume of 100 ml. with water and
electrolyze over night using a current of ,0.5 affi_i)eres. Remove the cathode,
rinse in water and then in 95% alcohol, dry at 110 C. for 10 minutes, cool and
weigh as Cd. Run in duplicate.
,
Combine the reserved filtrates No.1 and No. 2,jJoil to expel H 2S, oxidize
the iron present by adding 10 m!. of H 2 0 2 to the boiling olution, continue
boiling to expel H 20 2, cool and add a slight excess of a saturated solution of
KMn04 to insure removal of the H 2 0 2 Pass through a Jones reductor and
titrate with N/20 KMn04 solution for irqn, Run a blank through the reductor
and correct for same. Run in duplicate.
NOTE.-Weigh out small pieces of metal. Sawings, drillings or' filings are likely to
be contaminated with iron from the tool used, to take the sample .
.

D.

DETERMINATION OF LEAD AND TIN

/

Solutions Required. Lead Acid.-See _:"j)etermination of Lead as Sulfate, Analysis of Manganese Bronze."
. -

ANALYSIS OF NON-FERROUS ALLOYS

1397

Ferric Chloride Solution.-Dissolve 25 g. of FeCh 6H 20 in 100 ml. of

water and 10 ml. of HCI (sp.gr. 1.18); dilute to 500 ml. with water and mix.
Method.-Weigh 75--100 g. of sample into a 400-ml.lipless beaker, add 100
m!. of water and sufficient HNO s (sp.gr. 1.42) to affect solution. Avoid having
a large excess of acid. Add the acid in 10 ml. doses allowing acid present to
react with the alloy until it is almost spent before adding a subsequent dose.
Heating the solution after the first violent reaction subsides accelerates solution
of the alloy and hastens neutralization of the acid. If the above procedure is
carried out properly complete solution of the alloy may be attained without
having more than 10-15 m!. of free acid present. Finally cool, add 20 m!. of
HNO s (sp.gr. 1.42), dilute to a volume of 300 ml. with water and electrolyzeover night with a current of 1.0 ampere using a spiral anode and a cylindrical
cathode. (See" Method A.") Transfer the cathode to a clean beaker and
hold in reserve. Rinse the anode in water and then in 95% ethyl alcohol, dry
at 110 0 C. for 10 minutes, cool and weigh as Pb0 2 Multiply by 0.8643 for
lead. (Reserve the water washings.) (To insure complete precipitation
of the lead combine electrolyte and washings, add 10-15 ml. of HN0 3
(sp. gr. 1.42 and again electrolyze.) Dissolve the deposit of Pb0 2 into
a 200-ml. lipless beaker, with 10 m!. of HNO a (sp.gr. 1.42) and a few m!. of
H 20 2, add to ml. of lead acid and evaporate to' fumes of S03. Cool,
dilute to a volume of 50 ml. with water, bring to boiling, cool and allow to stand
several hours. Filter and wash with lead acid. Dissolve the PbS0 4 with
about 30 m!. of hot ammonium acetate solution containing 2 ml. of free acetic
acid for each 100 ml. of solution, dilute to a volume of 150 ml., bring to boiling
and add 5 ml. of a saturated solution of K 2 Cr 207. Allow to stand in a warm
place several hours, filter on a prepared Gooch crucible, wash with hot water
and finally with 95% ethyl alcohol. Dry at 1050 C. for one hour, cool and
weigh as PbCr04. Multiply by 0.6375 for lead. Run in duplicate.
~fter removing the lead, combine electrolyte and washings.
Filter and
wash with a hot solution of HNO s (1 : 49). Reserve the residue (Residue
No.1). Neutralize the filtrate with NH 40H (sp.gr. 0.9) and make acid with
HCI (sp.gr. 1.18) having 5 nil. of free HCI for every 100 ml. of volume. Cool
and pass H 2S. Filter and wash with dilute HCI (1 : 49) through which H 2S
has been passed. Discard the filtrate. If an appreciable amount of zinc has
been precipitated as a sulfide throw the paper (and precipitate) back into the
original beaker, add 10 m!. of HNO a (sp.gr. 1.42) and 10 m!. of li 2S0 4 (sp.gt.
1.84) and evaporate to fumes. Destroy organic matter by adding HNO a
(sp.gr. 1.42) to the hot solution. Cool, add 100 m!. of water, bring to boiling to
dissolve salts, neutralize with NH 40H, acidify with HCI as above and make a
second precipitation with H 2S. Filter and wash. Reserve the residue (Residue
No.2).
Dissolve'the deposit of copper from the cath~de held in reserve with the
least quantity of HNO a (sp.gr. 1.42), transfer to a 600-ml. beaker, add 10 mI.
of ferric chloride solution, make ~lightly alkaline with NH 40H (sp.gr. 0.9),
dilute to a volume of...300 m!. with water, bring'to boiling and allow to stand a
half hour or so in a warm place. Filter and wash with hot water. If the
precipitate retains copper, dissoJve the Fe(OH)a into the beaker the filtration
was made from with 20 m!. of hot HCI (1 : 1); wash with hot water. Make a
second separation of iron with NH 40H and filter through the same paper.

1398

ANALYSIS OF NON-FERROUS ALLOYS

. Should the precipitate still hold copper, make a third separation. When
finally the precipitate of Fe(OH)3 is freed of copper, transfer it, together with
the two residues held in reserve, to a 300-ml. cOI!e flask, add 10 ml. of HNOa
(sp.gr. 1.42) and 20 ml. of H 2S0 4 (sp.gr. 1.84) and evaporate to fumes. Destroy any organic discolQration with HNO a (sp.gt. 1.42). When the solution
shows no discoloration even upon strong fuming copl, add a few ml. of water and
fUlhe again.~ Repeat with the addition of water and subsequent fuming two or
three times to rid the solution of the last traces of nitric acid. Finally cool,
add 100 ml. of water and 20 ml. of HCl (sp.gr. 1.18), reduce with iron, and
proceed as described in "Determination of Tin in Battery Plate Metal."
Titrate with approximately N /20 KMn04 solution. Carry along a blank ,and
correct for same. Run in duplicate.
NOTE.-In general, the Pb0 2 results come high.
by the gravimetric method.

E.

Report the lead figure ob.tained

DETERMINATION OF MANGANESE

Solutions Required.-See " Determination of Manganese by the Persulfate

Method, Analysis of Manganese Bronze,;"
Method.-Weigh 10 g. of the drillings into a 400-ml. beaker, add 80 ml. of
the" solution for dissolving" and warm until completely in solution. Transfer
to a 125-ml. graduated flask, cool to'room temperature, make up to the mark
with water, mix thoroughly and measure off 50 ml. of the solution into.a 300-ml.
Erlenmeyer flask. Boil until the oxides of. nitrogen are expelled, dilute to
100 ml. with boiling hot water, add 20 ml. of AgNO a solution and 30 ml. of
(NH 4)2S208 solution and let stand over a hole on the steam bath until a full
permanganate color has developed. Cool to below 25 C. and immediately
titrate with standard arsenite solution to the disappearance of the pink color.
0
:Run in duplicate.
NOTES.-l. One g. of the metal requires 2.75 ml.jof the "solution for dissolving"
for solution.
.
2. Should the aliquoted portion contain morlJ than 0.0015 g. of Mn, a precipitate of
hydrated Mn02 will form. In that case, meas1lre off a smaller portion, add sufficient
"acid mixture" to bring the acid mixture content up to 24 ml. and proceed as described
above.
.
I

. 3. If the solution is allowed to stand too long a time, the color will fade. Warming
for 15 minutes should convert all the Mn to permanganic aCid. ~ ~
4. Owing to the rigid specifications called for in this class of alloys and the tediousness
of chemical analysis for impurities, especially tin, spectrographic methods are preferable.
See "Tentative Method of Test for Quantitative Spectrochemical Analysis of Zinc
Alloy Die Castings for Minor Constituents and Impurities," 1935 Book of A. S. T. M.
Tentative Standards, p. 1488. "
'

ANALYSIS OF NON-FERROUS ALLOYS

1399

FUSIBLE METALS
Fusible metals are mixtures of lead and tin to which has been added bismuth, cadmium or both to lower the melting point. They may contain in
addition small amounts of impurities, such as copper, arsenic, antimony,iron
and zinc.
ANALYSIS OF WOOD'S ALLOY 23
(ALLOY OF LEAD, TIN, BISMUTH AND CADMIUM) ....
DETERMINATION OF LEAD

Method.-Proc~ed exactly as in " Determination of Lead, Alloy of Lead"

'
Copper, Antimony and Tin," General Methods.
DETERMINATION OF TIN

Method.-Weigh 1 g. of the sawings into a 300-ml. cone flask, add 20 ml.

of H 2S0 4 (sp,gr. 1.84) and heat on a hot plate until completely decomposed.
Cool, add 100 ml. of cold water and 20 ml. of HCI (sp,gr. 1.2), reduce with iron
and proceed as described in " Determination of Tin in Battery Plate Metal."
DETERMINATION OF BISMUTH

Reagents Required. Fusion MixtuTe.-Weigh 100 g. of Na 2C0 3, 140 g. of'

K 2C0 3 and 100 g. of sulfur into ,a large porcelain mortar. Grind and mix
thoroughly, transfer to a glass-stoppered bottle and reserve for use.
KSH Wash Solution.-See " Determination of Lead, Bismuth, etc. in Tin
Base Bearing Metal."
Method.-Weigh 1 g. of the sawings into a 250-ml. beaker, add about 10 ml.
of water, heat, and add sufficient HNO a (sp,gr. 1.42), a little at a time, until
th~ alloy is decomposed. Evaporate to dryness, digest with 10 ml. of HN0 3
(sp,gr. 1.42), dilute to 50 ml. with water, bring to boiling and allow to stand,
and settle for one hour keeping the temperature just below the boiling point.
Filter, keeping the solution hot, through a small close-textured paper to which
a little paper pulp has been added and wash with hot dilute HN0 3 (1 : 49).
Reserve the filtrate.
Transfer the paper and precipitate to a small porcelain crucible, dry and
ignite. Mix the residue with 4 g, of fusion mixture, cover with 4 g. of the fusion
mixture and heat until in. quiet fusion; heat 20 minutes longer. Cool, leach
with as little water as possible, remove and wash the crucible. Combine
washings. with the main solution. The supernatant solution should be of a
clear yellow color. Filter and wash with warm (not hot) KSH wash solution.
(The filtrate may be used for the determination of antimony. See" Determination of Antimony.")
Dissolve the residue with 15 nil. of hot HNO a (1 : 2), wash with hot water,
boil to a volume of 25 ml. and allow to stand for one hour keeping the temperature just below the boiling point. If an insoluble residue is present, filter on
a small close-textured paper to which a little paper pulp has been added and
23

Methods of National Lead Company Research Laboratories.

1400

ANALYSIS OF NON-FERE,OUS. ALLOYS

wash with hot dilute HN0 3 (1 : 49). Discard the residue. Combine the filtrate with the original filtrate from the H 2Sn03 beparation:
Boil the combined filtrates down to a volume of 100 mI., make just
neutral with NH 40H (1 : 3) as shown by a piec~ of litmus paper and then add
5 mI. of HCI (1 : 9) in excess. Dilute to a volume of 400 mI. with boiling hot
. water and allow to stand in a warm place for seyeral hours. Filter and wash
with hot water. Reserve the filtrate.
.
Dissolve the precipitate from the filter with hot HCI (1 : 1) and ~ash
alternately, first with hot water and then with hot HCI (1 : 1). Evaporate
the solution down to a volume of about 1 mI., dilute. to a volume of 400 mI.
with boiling hot water and allow to stand as before. Filter on a weighed porcelain Gooch crucible and wash with hot water. Transfer the filtrate to a clean
beaker and reserve. Finally wash the precipitate on the Gooch once or twice
'with 95% alcohol, dry at 110 C. for 30 minutes, cool and weigh as BiOCI.
Multiply by 0.8024 for bismuth.
NOTEs.-l. During the solution of the alloy, lead nitrate may separate out and
precipitated H.Sn03 may coat undissolved alloy and prevent further solution. In such
a case, dilute with sufficient hot water to dissolve lead nitrate and detach the H.Sn03
coating from undissolved alloy with a glass rod.
.
2. The fusion of the H.SnO. is best conducted by placing the small porcelain crucible
inside a larger one. Cover both crucibles. Set the covers on the crucibles in an inverted
position, first breaking off the small ring handles of the covers before placing them on
the crucibles.
/
3. Start the fusion at a low heat and gradually bring the temperature up. If
started at too high a temperature the melt will froth over the top.
4. If the melt is leached with an excessive amount of water, the sulfides, especially
iron sulfide, will not separate but will remain in solution in a colloidal state. In such a
case, twaporate to a volume of about 50 ml., add about 20 ml. of KSH stock solution,
stir, and allow to stand in a warm (not hot) place until the supernatant solution is of a
clear yellow color. The addition of a few grams of NH.Cl accompanied with vigorous
stirring will accelerate the precipitation of the colloidal sulfides. (Do not boil.)
DETERMINATION OF CADMIUM ,t.\ND COPPER

Electrodes.-See " Determination of Coppe'r by the Electrolytic Method,

Analysis of Manganese Bronze."

Method.-Combine the two filtrates frbm the BiOCI separations, evaporate

to a volume of about 40 mI., add 10 ml. of. H 2S0 4 (1 : 1), bring to boiling, cool,
allow to stand 30 minutes or so, filter and wash with cold water. Discard the
residue. Dilute the filtrate to a volume of 300 ml~ 'warm and pass H 2 S for
30 minutes, allow to stand in a warm place an hour or so and again pass H 2S
for 15 minutes. Filter and wash two or three times with R 2S water containing
a little (NH 4).S04. Throw the paper containing the precipitate back into the
beaker the filtration was made from, add '10 mr. of HN0 3 (sp.gr. 1.42) and 5
mI. of. H 2S0 4 and evaporate to fumes, taking the necessary steps to remove
carbonaceous matter. (See" Determination' of Lead, Bismuth, etc. in Tin
Base Bearing MetaL") Cool, take up with a few mI. of water and fume.
Again .cool, take up with water and fume. Finally cool, dilute to a volume
of 40 mI., bring to boiling, cool, allow to stand at leaJt two hours, filter and
wash with cold dilute H 2S0 4 (1 :'9). Discard the residue. Neutralize the
filtrate with NH 4 0H (1 : 2) and then make' :reid with 3 ml. of H 2 S0 4 (1 : 9)
in excess, dilute to a volume of 150 mI. and electrolyze over night with a

ANALYSIS OF NON-FERROUS ALLOYS

1401

current of 1.0 ampere, using a platinum spiral for ariode and a platinum
cylinder for cathode. Remove the cathode, rinse in water and in 95% alcohol,
dry at 110 C. for 10 minutes, cool and weigh as cadmium plus copper. '
Dissolve the deposit from the cylinder with 5 ml. of HNO a (sp.gr. 1.42)
and wash the electrode free from cadmium with water. Combine the electrolyte and washings and evaporate to a small volume. Wash into a small
porcelain crucible and evaporate to dryness. Take up with 1 m!. of HNO a
(1 : 1) and proceed as in " Determination of Copper in Solder."
Having determined the amount of copper present deduct from the weight
of copper plus cadmium. The difference is cadmium.
NOTE.-Do not lose any of the electrolyte in removing the cathode.
DETERMINATION OF ANTIMONY

(a) General Method.-Dissolve 5 g. of the sawings in dilute HNO a and

proceed as in " Determination of Bismuth" up to the point where the H 2SnOa
precipitate after fusion and .leaching, has been filtered and washed.
Warm the alkaline sulfide filtrate, add KClO a, a little at a time, and HCl
(sp.gr. 1.18) until the solution is strongly acid and the sulfur has been oxidized.
The solution should be clear and distinctly yellow. Dilute to a volume of
300 m!. with water and boil down to a volume of 200 m!. When the chlorine
has been expelled, add 10 m!. of copper solution (2.0 g. of CUS04' 5H 20 dissolved in 500 m!. of water), make slightly alkaline with NH 4 0H (sp.gr. 0.9), then
just acid with a saturated solution' of H 2 C 2 0 4, add 12 g. of solid H 2 C 2 0 4, dilute
to a volume of 400 m!. and proceed as in "Determination of Antimony,
Analysis of Alloys of Lead, Copper, Antimony and Tin," General Methods.
NOTE.-Care must be exercised in rendering the alkaline sulfide solution acid with
Hel. If the acid is not added in small doses, especially when the solution is hot, the
r~pid evolution of H 2S will make the solution froth over the top of the beaker.

(b) Rapid Method.-Dissolve 5 g. of the sawings in dilute HNO a and proceed as in " Determination of Bismuth" up to the point where the H 2SnOa has
been filtered and washed.
Transfer the paper containing the H 2SnOa and HSbO a to a 350-mI. Erlenmeyer flask and, proceed as in "Determination of Antimony, Analysis of
Bronze Bearing Meta!."
DETERMINATIN OF IRON

Proceed as in " Determination of Iron, Determination of Lead, Bismuth, etc.

in Tin Base Bearing Meta!."
DETERMINATION OF ZINC

Proceed as in " Determination of Zinc in Solder."

1402

ANALYSIS OF NON-FERROBS ALLOYS

J

DETERMINATION OF ARSENIC

Proceed as in " Determination of Arsenic, Anaiysis of Alloys ~f Lead, Tin,

Antimony and Copper," General Methods.

SILVER ALLOYS
ANALYSIS OF SILVER SOLDERS

2,4

(ALLOY OF SILVER, COPPER, CADMIUM AND ZINC)

DETERMINATION OF SILVER

Solutions Required. Dilute Hydrochloric Acid Solution.-To 3 m!. of HCI

(sp.gr. 1.18) add 97 m!. of water. One milliliter of this solution should precipitate about 0.038 g. of silver.
Method.-Weigh 1 g. of the sawings into a 250-m!. beaker, add 20 mt of
HNO. (1 : 4), gently warm until solution is complete and, then boil to expel
the oxides of nitrogen. Dilute to a volume of 175 mI., heat to boiling and add,
drop by drop with constant stirring, sufficient dilute HCI solution to precipitate
all the silver present. Avoid a large excess of the precipitant. Cool in a dark
place. When the solution is cold and the precipitate has settled, test the clear
solution with a drop of the dilute HCI solution to determine whether or not the
precipitation is complete. Filter on a weighed porcelain Gooch crucible, and
wash with cold dilute HNO. (1 : 200); finally wash once 'with cold water.
I
Reserve the filtrate.
Place the Gooch crucible inside a porcelain crucible, dry and heat until the
material begins to show signs of fusion. Cobl and weigh as AgCl. Multiply
by 0.7526 for Silver.
N OTES.-I. If tin is present, as shown by a precipitate of metastannic acid when dissolving in nitric acid, it should be filtered off as in "Determination of Tin," before pro- .
ceeding with the determination of silver. (See note under '''Deter_mination of 'Tin.")
2. If it is desired to determine the silver by a combination assay'method, dissolve
an amount of sample which will contain about 0.3 g, of silver in dilute RNO. (1 : 4),
precipitate the silver as AgCl, filter and wash and proceed as in "Assay of Copper Bullion" in chapter under Fire Assay for Gold and Sil:er.
DETERMINATION OF COPPER

I
.
Method.-Transfer the filtrate from the AgCI separatIon to a 400-m!'
beaker, add 5 ml. of H 2S0 4 (sp.gr. 1.84), evaporate to fumes of H 2S0 4 , cool,
24 The methods given under" Analysis of Silver Solders" are 'A, S. T. M. methods,
excepting those for cadmium and zinc. See A. S. T. M, Methods of Chemical Analysis
.. -- '
.
of Metals, p, 193, December, 1936,
25 See Specifications of Silver Solders" 1936 Book'6fA. s. T. M. Standards, Part I,

p.69I.

ANALYSIS OF

NO~-FERROUS

ALLOYS

1403

dilute with 50 m!. of water, bring to boiling, transfer to a 200-m!. lipless beaker;
add 3 m!. of HNO a (sp.gr. 1.42), cool, dilute to 150 m!. with water and proceed
as in " Determination of Copper, Analysis of Manganese" Bronze."
DETERMINATION OF CADMIUM

Method.-Transfer the electrolyte and washings to a 600-m!. beaker,

evaporate to fumes of H 2S0 4, cool, add 7 g. of (NH 4)2S04, dilute to a volume of
300 m!., bring to boiling and make two separations of the cadmium from zinc'
with H 2S. Dissolve the final precipitate with 10 m!. of HNO a (1 : 1), add
5 m!. of H 2S0 4 and evaporate to fumes of SOa. Proceed as outlined in " Determination of Cadmium and Iron, Analysis of Zinc Base Die Casting Alloy." "
DETERMINATION OF ZINC 26

Solutions Required. Standard Potassium Ferrocyanide Solution.-Dissolve 21.7 g. of K 4 Fe(CN)63H 20 in water, filter, dilute to 1000 m!. with water
and mix. Standardize by weighing out 0.2 to 0.25 g. of pure zinc into a 400-m!.
beaker, dissolve with 7 m!. of HCI (sp.gr. 1.18), dilute to 200 mI., add about 7 g.
of NH 4 CI and titrate as in the" Determination of Zinc" in the sample. Deduct 0.2 m!. from the burette reading as a correction for end-point. ' Each
milliliter should approximately equal 0.005 g. of zinc.
Uranyl Nitrate Indicator.-Dissolve 5 g. of U02(NO ak 3H 20 in 100 ml. of
water.
Method. 27-Combine the two reserved filtrates from the cadmium sulfide
separations with the electrolyte and washings, boil to expel dissolved gas, add
5 m!. of HN0 3 (sp.gr. 1.42) and evaporate to a volume of about 300 m!.
Make slightly alkaline with NH 40H (sp.gr. 0.9), allow to stand for any iron
present to settle out, filter and wash first with dilute NH 40H (1 : 4) and
then with hot water. Dissolve the precipitate with a little hot Hci (1 : 1)
and make a second separation of iron with NH 40H. Combine the two filtrates
from the iron separations, neutralize with HCl (sp.gr. 1.18), make slightly
alkaline with NH 40H (sp.gr. 0.9), and pass H 2S for 15 minutes. While
the H 2S is passing, make slightly acid with 80% acetic acid and continue
passing H 2S for 15 minutes longer. Filter on a close-textured paper to
which a little paper pulp has been added and wash two or three times with
H 2S water to which a little acetic acid has been added. Transfer the paper
containing the precipitate to the beaker the filtration was made from, add
20 m!. of HCl (1 : 2), open up the filter so that the precipitate will come in
contact with the acid,' and allow to stand in a warm place until the ZnS has
completely dissolved. Dilute to a volume of 200 m!. with water, add about 7 g.
of NH 4CI, pass H 2S for five minutes, heat to 80 C. and titrate with the standard
K4Fe(CN) 6 solution until a drop of the solution added to a drop of the indicator
on a spot plate gives a slight tinge of brown. Deduct 0.2 m!. and calculate to
per cent of zinc.
'
26 See "Rapid Ferrocyanide Method for Determining Zinc in Ores" in chapter
under Zinc.
27 For the titration with ferrocyanide solution using an inside indicator see" Titration
in Acid Solution-Beparating of Zinc as Sulfide" in chapter under Zinc.

1404

ANALYSIS OF NON-FERROUS ALLOYS

NOTES.-l. If the sample contains more than 25% of zinc, take such an aliquot
portion as will not contain more than 0.25 g, .It is advisable to aliquot before separating
t
the zinc as a sulfide.
2. If nickel and manganese are present separate the .zinc as a sulfide in a formic acid
solution see "Determination of Zinc, Analysis of Bronz, Bearing Metal."
.

I
DETERMINATION OF TIN

Method.-Dissolve a separate 5 g. sample wiih 100 mI. of HNO a (1 : 4),

boil to expel the oxides of nitrogen, dilute to a volume of 250 mI., bring to
boiling and allow to stand for about one hour keeping the temperature just
below the boiling point. Filter on double 7 cm. ashless papers, being careful
to keep the solution hot during the filtration and wash with hot water. Reserve the filtrate for the determination of lead. Transfer the paper and precipitate to a weighed porcelain crucible, dry, ignite, cool and weigh as Sn02.
Multiply by 0.7877 for Tin.
NOTE.-Every precaution must be taken to avoid contamination by chlorides in
order to prevent the precipitation of Agel during the determination of tin.
DETERMINATION OF LEAD

Solutions Required.-See " Determination of Lead as Sulfate, Analysis of

Manganese Bronze."
.
Method.-Heat the filtrate from the H 2SnOa separation to boiling and
precipitate the silver as in " Determination of Silver." Filter on a closelywoven paper and wash with col.d dilute HNO a (1 : 200). Add 80 m!. of lead
acid and evaporate to fumes bf H 2S0 4. To insure complete removal of
chlorides, cool, add a few m!. of water and fmrie again. Cool, add 70 m!. of
water, heat to boiling and proceed as in " Determination of Lead as Sulfate,
Analysis of Manganese Bronze."
Transfer the filtrate from the PbS0 4 separation to a Clean beaker before
the final washing with alcohol and reserve for the determination of iron.
NOTE.-In order to avoid "spitting" during the lvaporation with lead acid it is
suggested to heat over a burner holding the beaker with a pair of tongs and twirling
.
as the evaporation is carried on.
DETERMINATION OF IRON

Solutions Required.-See " Determination of Iron,.Analysis of Manganese

Bronze."
,
Method.-To the filtrate from the PbS0 4 separation ad_d 3 mI. of HNO a
(sp.gr. 1.42) and remove copper by electrolysis. Remove the electrodes, wash
them with water, transfer the electrolyte and washings to a 600-mI. beaker, boil
to expel dissolved gases and add 5 m!. of HNO a (sp.gr. 1.42). Cool, add 5 g: of .
NH 4Cl and sufficient NH 40H (sp.gr. 0.9) to make the solution alkaline and to
hold up the zinc, bring to boiling and allow to st:and in a warm place an hour or
so. Filter and wash with hot water. Dissolve the precipitate through the
filter paper with hot dilute HCl (1 : 1) and make a second separation of iron
with NH 4 0H. Combine the two filtrates and reserve for the determination
of nickel. Dissolve the final precipitate throu_gh the paper with 100 m!. of
dilute H 2S0 4 for reductor, add a few drops of 11 strong 'solution of KMn04 until

ANALYSIS OF NON-FERROUS ALLOYS

1405

a distinct color persists, bring to boiling, pass through the reductor and proceed
as in " Determination of Iron, Analysis of Manganese Bronze."
,
DETERMINATION OF NICKEL

. Solutions Required: Dimethylglyoxime Solution.- Dissolve 5 g. of dimethylglyoxime in 500 mI. of grain alcohol and filter before using.
Method.-Evaporate the combined filtrates from the Fe(OH)3 separations
to a volume of about 100 mI., make barely acid with HCI (sp. gr. 1.18), heat to
boiling, add 5 m!. of dimethylglyoxime solution for each 0.1 g. or fraction of
nickel present and proceed as in" Determination of Nickel, Analysis of Bronze
Bearing Metal."

HEAT RESISTING ALLOYS

ANALYSIS OF NICHROME

28

(ALLOY OF NICKEL, CHROMIUM, IRON AND MANGANESE)

DETERMINATION OF NICKEL

Electrodes.-See "Determination of Copper, Analysis of Manganese

Bronze."
Solutions Required. Nitro-Hydrochloric Acid.-Mix 10 m!. of HNO a
(sp.gr. 1.42), 30 m!. of HOI (sp.gr. 1.18) and 40 ml. of water.
Dilute Acetic Acid.-Mix 1 m!. of acetic acid (sp.gr. 1.04) and 3 m!. of water.
Potassium Thiocarbonate.-Saturate 125 m!. of a 5% solution of KOH
with H 2S, add 125 m!. of KOH solution (5%) and 10 ml. of CS2 and heat
moderately. Decant the dark red liquid from undissolved CS 2 and preserve
in a well closed flask.
Sodium Dimethylglyoxime.-Dissolve 3 g. of d!methylglyoxime in 100 m!.
of NaOH (3%), made fresh each time used.
Method.-Weigh 1 g. of the alloy into a 600-ml., beaker dissolve in 20 ml.
of nitro-hydrochloric acid, add 20 ml. of H 2S0 4 (1 : 1) and evaporate to fumes
of H 2S0 4 Cool, add 100 mI. of water and allow the solution to digest at a low
heat until all salts have dissolved. Filter on a small paper and wash with hot
water. If an insoluble residue is present wash with dilute H 2S0 4 (1 : 99).
Transfer the residue to a platinum crucible, dry, ignite at a low heat, add 2 m!.
of HF and 2 drops of H 2S0 4 (sp.gr. 1.84), evaporate to dryness, heat over a low
flame to expel free H 2S0 4, cool and fuse with a little Na 2CO s. Dissolve the
melt in HCI, and add to the main nickel solution.
Transfer the' nickel solution to a 250-m!. graduated flask, make up to the
mark with water at room temperature and mix. Draw off exactly 50 ml. into
28

A. S; T. M. Methods of Chemical Analysis of Metals, p. 203, Dec. 1936.

1406

ANALYSIS OF NON-FERROUS ALLOYS

a 600-m!. beaker, add 25 ml. of a 20% solution of tartaric acid (enough to hold
all the iron and chromium in solution), make alkltline withNH 4 0H and then
make acid with 15 ml. of HCI (sp.gr. 1.18) in excess. Add 20 to 30 ml. of the
sodium dimethylglyoxime solution, neutralize with INH 40H and then make ivst
acid with the dilute acetic acid. Stir the solution vigorously and allow to stand
at a temperature of from 75 to 90 C., with occasio,ll'al stirring, for 30 minutes.
Filter on a weighed Gooch crucible, wash with hot water, dry at 110 C. to
constant weight, cool and weigh as nickel dimethylglyoxime. Multiply by
0.2032 for nickel.
,
If greater accuracy is desired, use a larger aliquot portion'with proportionally
larger amounts of reagents and filter the nickel dimethylglyoxime precipitate on
a 15 cm. paper and wash 18 or 20 times with hot water. Dissolve the precipitate off the paper with boiling dilute HNO a (1 : 3) and wash the filter thoroughly with hot water. Add 10 ml. of dilute H 2S0 4 (1 : 1) and boil the liquid
gently until strong fumes of SOa are evolved. Cool, add 10 ml. of HNO a
(sp.gr. 1.42) and repeat the evaporation. Rinse the cover and 'sides of beaker
with a fine jet of water and fume the solution again to insure the expulsion of
every trace of HNO a. Cool, add approximately 50 m!. o'f cold water and boil
the contents of the beaker for several minutes, which should result in a perfectly clear solution.
.,
Make the solution alkaline with 25 m!. of NH 40H (sp.gr. 0.9) in excess,
insert electrodes, and electrolyze with a current of from 1 to 2 amperes until
the solution has become 'colorless. Continue the electrolysis for at least 15
minutes longer, and then test the solution by adding 1 or 2 drops of it to a
solution of potassium thiocarbonate. A pink color indicates the presence of
nickeI. Remove the cathode quickly, rinse it in water and then in absolute
alcohol, dry for a few minutes at a temperature of 80 to 100 C., cool and
weigh as metallic nickel.
NOTE.-If 'an alcoholic solution of dimethylglyoxime was used to precipitate the
nickel, the contents of the beaker should be digested at a temperature just short of the
boiling point until the odor of alcohol can no longer be ,detected before filtering.
DETERMINATION OF 'CHROMIUM

Solutions Required. Ferrous Ammonium Sulfate Solution.-Dissolve

'39.25 g. of FeS04(NH 4)2S0A6H 2 0 in water, 'add 20 ml. of H 2 S0 4 (1 : 1), dilute
to 1000 m!. and mix. ,
_ .
Standard Potassium Permanganate Solution (0.1 N).--,For"its preparation
and standardization against Na 2 C 20 4see" Potassium Permanganate, Standard
Solutions" in chapter under Reagents. Each milliliter should equal 0,001734
g. of chromium.
Method.-Treat 1 g. of the sample in a covered 600-ml. beaker with 10 ml.
of perchloric acid (60%). Heat to strong fumes:of perchloric acid to aid in the
solution of the alloy, and continue the heating for 3 or 4 minutes after the alloy
has dissolved completely to insure oxidation of the larger part of the chromium
to the sexivalent state. Add 100 ml. of hot water and heat to boiling for several
minutes to expel chlorine. Remove from the plate and/add 20 mI. of dilute
H 2S0 4 (1 : 1) and 5 ml. of HNO a (sp.gr. 1.4~):_: 'Dilute to 250 ml. with hot
water, add 10 ml. of AgNO a (0.5%), 3 g. o( ammonium persulfate ana 5 or 6

ANALYSIS OF

NON-F~RROUS

ALLOYS

1407

drops of KMn04 solution (2.5%). Boil for 5 minutes or longer, add 20 ml. of
NaCI solution (10%) and boil for 10 minutes after the permanganic acid or
any Mn02 formed has dissolved completely, in order to make sure that all
chlorine has been expelled. Cool to room temperature and dilute with cold
water to 400 ml. Treat with 2 ml. of H 3 P0 4 (sp.gr. 1.72), add from a burette
ferrous ammonium sulfate solution until the solution turns green and then
add 5 ml. in excess. Titrate this excess back with 0.1 N' potassium permanganate.
In a clean flask, place 50 ml. of the ferrous ammonium sulfate, add 10 ml.
of H 2S0 4 (1 : 1) and titrate with the standard KMn04 solution. From this
titration calculate the number of ml. of KMn04 solution necessary to titrate ~
the amount of ferrous ammonium sulfate solution add~d to the sample. The
difference between the number of ml. of KMn04 solution calculated and the
number used for the titration of the sample represents the chromium present
and the percentage is found by multiplying this volume by 100 times the
chromium titre of the KMn04 solution.

NOTE.-The strength of the ferrous ammonium sulfate may change from day to day.
Its strength in terms of KMn04 solution should be ascertained each time the determination is made.
DETERMINATION OF IRON

Solutions Required. Stannous Chloride Solution.-Dissolve 50 g. of

SnCh 2H 20 in 100 ml. of Hel (sp.gr. 1.18) and dilute to 1000 ml.
Standard- Potassium Permanganate Solution.-See "Determination of
Chromium." Each milliliter equals 0.005584 g. of iron.
Method.-For alloys containing 10% or more of iron, use a 1 g. sample,
and for alloys containing 2% or less of iron, use a 3 g. sample. Treat the sample
in a 500-m!. Erlenmeyer flask fitted with a Meyer bulb or trap with 20 to 50 m!.
of perchloric acid (60%) at a temperature of 203 C. or higher until the alloy
has dissolved completely, and continue the boiling of the acid for five minutes
longer to aid in the complete oxidation of the chromium to the sexivalent state.
The tall flask and bulb \or trap) are used to prevent the loss of FeCb by volatilization, which happens if a covered beaker is used. Add approximately
200 in!. of hot water, an excess of about 10 ml. of NH 40H (sp.gr. 0.9) and 5 g. of
ammonium persulfate. Boil for 5 minutes, filter and wash thoroughly with hot
water. Disi?olve the Fe(OHh in hot Hel (1 : 4) and repeat the precipitation
with NH~OH and ammonium persulfate, and the filtration. Dissolve the
Fe(OHh in the least possible amount (15 to 25 ml.) of boiling, dilute Hel
(1 : 4) and wash the filtrate thoroughly with water. Heat to boiling, add
stannous chloride solution slowly and with stirring to discharge the yellow color
of the ferric chloride and to provide a few drops in excess. Finish the determination by the Zimmerman-Reinhardt method. 29
0

DETERMINATION OF MANGANESE
Solutions Required.-See " Determination of Chromium."
Method.-Decompose 1. g. of the sample with 20 ml. of perchloric acid
(60%) exactly as outlin;ed in the determination of chromium. Add 200 ml. of
29 Blair, A. A., "The Chemical Analysis of Iron," 8th Ed., p. 248.

1408

ANALYSIS OF NON-FERROVS ALLOYS

warm water, an excess of about 10 m!. of NH 40H (sp.gr. 0.9) and 5 g. of ammonium persulfate. Boil for 5 minutes, filter on twp superimposed 11 cm. filter
papers, supported on a Buchner funnel, gentle suction being employed. Wash
the precipitate thoroughly with hot water. Dissolve the precipitate of Fe(OH)a
and Mn02 in. 75 m!. of hot HNO a (sp.gr. 1.135) ~nd a .few drops of H 2SO a.
Wash the filter thoroughly with hot water. Repeat the precipItation with
NH 40H and ammonium persulfate, and the filtration. Dissolve the precipitate in 50 m!. of hot HNO a (sp.gr. 1.135) and a few drops of H 2SO a, and
wash the filter with HNO a of the same strength. Transfer to a 300-m!. Erlenmeyer fl!j.sk and boil for several minutes to expel oxides of nitrogen. Trea~
cautiously with 0.5 g. of sodium bismuthate and heat to boiling for 1 or 2
minutes. Clear with a sufficient excess of H 2SO a and boil for 2 or 3 minutes
longer. Cool to 15 C., add 1 to 2 g. of sodium bismuthate, shake vigorously
for 1 minute, dilute with an equal volume of cold water and filter at once on an
ignited asbestos or alundum filter. Wash with ce>ld HNO a (3%). Add from a
burette 20 to 25 m!. (depending on the amount of permanganic acid) of ferrous
ammonium sulfate ,solution and titrate the excess with standard KMn04
solution.
To ascertain the amount of manganese present proceed exactly as in " Determination of Chromium."
DETERMINATION OF CARBON

See" Determination of Carbon by the Direct Combustion Method, Standard Methods of Chemical Analysis of Plain Carbon Steel," in the chapter under
Iron and Steel.
DETERMINATION OF INSOLUBLE RESIDUE

Method.-Treat 5 g. of the sample (if low in insoluble residue 10 g. should

be taken) with from 200 to 400 m!. of HCI (sp.gr. 1.18) at a temperature of from
60 to 70 0 C. until the reaction appears to be complete and then add 1 to 2 m!.
of HNO a (sp.gr. 1.42). Boil for 2 minutes, dilute rith an equal volume of hot
water, filter and wash with hot dilute HCI (1 : 19 .
Ignite the residue at a low heat in a platinum crucible, treat with HF aild a
few drops of H 2S0 4 (1 : 1), and evaporate 'just to fumes of H 2S0 4 , If 'any
residue remains, add 10 m!. of water, filter on a 9 cm. paper and wash with hot
. water. Ignite any residue r'emaining on the paper and weigh. Calculate to
per cent o'f insoluble.
Accuracy.-In the case of ,an insoluble residue of 0.05%-:- duplicate determinations should check within 0.02% insoluble residue.
I

DETERMINATION OF' SILICON

Method.-Treat a two to five factor weight sample (one factor weight

equals 0.4672 g.) in a covered 300-m!. porcelain~casserole with 15 to 35 m!. of
perchloric acid (60%) at a temperature of 203 0 C. (the boiling point of the acid)
until the-alloy has dissolved completely, and continue boiling for 10 minutes
longer. Add 100 ml. of warm water, boil for several minutes and filter on an
11 cm. paper of close texture. Wash six or eight times 'with hot water, once
wit,h H 2SO a (10%) to reduce any chromic acidileTd in the filter, and then thorI
oughly with hot water.

ANALYSIS OF NON-FERROUS ALLOYS

1409

Transfer the paper and precipitate to a small platinum crucible, dry, and
ignite, at first at a dull red heat and finally at 1050 to 1100 C. Cool and weigh.
Add 2 drops of H 2S0 4 (1 : 1) and several ml. of HF and evaporate until all the
H 2S04 has been expelled. Ignite the crucible again at 1050 to 1100 0 C., cool
and weigh.
The difference between the first and second weights multiplied by 100 and
divided by the weight of sample taken gives the percentage of silicon in the
sample.
0

DETERMINATION OF SULFUR

See" Determination of Sulfur by the Evolution-Titration Method, Standard

Methods of Chemical Analysis of Plain Carbon Steel," in the chapter under
Iron and Steel with the following modifications:
1. Due to the length of time (30 minutes) required for solution of the alloy,
closer observation is necessary than in the analysis of steel.
2. During the solution of the sample, care must be exercised in order that
absorbing solution is not drawn back into the flask.
3. There is danger of contamination from rubber hose connections and
rubber stoppers.
4. The use of a gas flame is a possible source of sulfur contamination and
an electric heater is recommended.
. Accuracy.-In the case of a sulfur content of 0.05%, duplicate determinations should check within 0.003%.

CORROSION RESISTING ALLOYS30

AN ALYSIS OF STELLITE

31

(ALLOY OF COBALT, CHROMIUM AND MOLYBDENUM)

32

Electrodes.-See " Determination of Copper by the Electrolytic Method,

Analysis of Manganese Bronze."
DETERMINATION OF COBALT

Method.-Dissolve 1 g. of the sample in 30 ml. of HCI (sp.gr: 1.18) and

30 ml. of HN0 3 (sp.gr. 1.42). Evaporate to a volume of 20 mI., add 100 mI.
.of HN0 3 (sp.gr. 1.42) and evaporate to dryness. Do not bake. Add 50 mI.
30 See" Analysis of Stellite, Akrite and Similar Complex Alloys," Chern. Abs., 23,
5128 (1929).
31 Johnson C. M., "Rapid Methods for the Chemical Analysis of Special Steels,
Steel-Making Alloys, and Graphite," p.322, John Wiley and Sons, N. Y., 1914.
32 For methods of separation and estimation of W, Mo, U, V, Ti, Zr, As, P and Cr,
see "Analytical Methods for Certain Metals," Department of the Interior, Bureau of
Mines, Bull., 212, p. 18, 1923.

1410

ANALYSIS OF NON-FERROUS ALLOYS

of HCI (sp.gr. 1.18), evaporate to a volume of 20 ml., transfer to a lOOO-ml.

wash bottle flask and dilute to 200 m!. with water. Add Na 2 02, a gram at a
time, until a black precipitate forms and tends ito settle, and then add 10 g.
more of Na 202 and 10 g. of Na 2CO g. Bring to boiling, continue to 150il for
10 minutes, cool, filter on double 15 cm. papers arid wash with water containing
a little Na202' Dissolve the precipitate througn the paper ,with 60 m!. of hot
HCl (1 : 1) into the original flask and make a second separation of cobalt and
iron with Na 202 and Na 2CO g. Repeat the peroxidations until the final filtrate
is free of yellow color. Three peroxidations should suffice. Discard the filtrates. Finally dissolve the precipitate with 60 ml. of hot HCl (1 : 1) .into
the original flask and wash with hot water. Transfer the papers to a porcelain
crucible, dry, ignite, cool, add 5 m!. of HCl (sp.gr. 1.18) to dissolve the remaining
residue and add to the main cobalt and iron solution.
Transfer the combined solutions from the flask to a 600-m!. beaker and
make a basic acetate separation of the iron. (See" Separation of Manganese
from Iron and Aluminum, Basic Acetate Method" in chapter under Managnese.) Dissolve the precipitate with hot HCl (1 : 1) and make a second
basic acetate separatioll.. Combine the' two filtrates and washings. Reserve
the precipitate for the determination of iron.
Boil the combined filtrates from the basic acetate separations down to a
volume of 500 mI., make just alkaline with NH 4 0H (1 : 1), pass H 2S for about
20 minutes and then while the H 2S is passing, make slightly acid with acetic
acid and continue passing H 2S for 15 minutes longer. Filter but do not wash
the precipitate. Wash the precipitate from the paper back into the beaker,
unfold the filter and place it on the cover glass with the side of the filter that
came in contact with precipitate exposed to view. Place cover glass and filter
on the beaker, add to the contents of the beaker 30 ml. of a mixture composed
of equal parts of HNO g (sp.gr. 1.42) and HCl (sp.gr. 1.18) and heat on tb.e hot
plate until the acid fumes have dissolved from the filter any adhering CoS.
Do not heat too long a time or the filter will disintegrate.. Wash dow~ the
/.
filter and cover glass. Discard the filter. ;
To the solution add 10 mI. of H 2S0 4 (sp.gr. 1.84) and evaporate to fumes of
SOg. Cool, add 2 ml. of water and bring to fumes a second time to remove the
last traces of HNO g and HCI. Cool, add 1l00 m!. of water and boil to dissolve
salts. Transfer the solution to a 600-ml. beaker, cQ2l, make alkaline with 60
ml. of NH 40H (sp.gr. 0.9) in excess, dilute to a volume-of.- about 400 mI. and
electrolyze over night with a current of 1.0 ampere. Remove the cathode;
wash twice with water and once with 95% alcohol; dry at 110 C., cool and
weigh as metallic cobalt.
NOTEs.-1. When manganese is present, a solution containing 30 ml. of water and
2 g. of Na 202 is added to the alkaline solution just: before plating out the cobalt.
2. If any nickel is present it will plate out along with the cobalt. In that case dissolve both metals from the cylinder with RNO, (sp.gr. 1.42) and determine the nickel
by the glyoxime method. (See "Precipitation of Nickel by Dimethylglyoxime" in
chapter under Nickel.) Deduct for nickel. The difference i,s cobalt.
3. Johnson folds a small square of cheese cloth in with the,filter at the apex to prevent
the alkaline solution from tearing the filter when~~ring after a peroxidation.

ANALYSIS OF NON-FERROUS ALLOYS

1411

DETERMINATION OF IRON

Dissolve the precipitate obtained from the separation of the cobalt with
hot HCI (1 : 1), add NH 4 0H (sp.gr. 0.9) until a permanent precipitate forms,
boil, allow to settle, filter and wash with dilute ammonia and hot water until all
NH 4 Cl is remov~d. Proceed as in " Determination of Iron, Analysis of Manganese Bronze."
NOTE.-RUIi a blank on chemicals used.
DETERMINATION OF CHROMIUM

Solutions Required. Ferrous Ammonium SUlfate and Standard KMn04' _

-See" Determination of Chromium, Analysis of Nichrome."
Method.-'-Fuse 0.5 g. of the sample with 8 g. of N a 202 in a 35-ml. nickel
crucible. Keep the fusion molten for 5 minutes. Cool, place the crucible on
its side, in a 375-m!. porcelain casserole, cover with a watch glass and cautiously
add sufficient cold water to cover about half the cr_ucible. When the melt is
completely dissolved, remove the crucible and wash it with hot water. Combine the solution and washings, filter and wash thoroughly with water containing a little Na 202. Discard the residues if free from undecomposed alloy.
(See " Sodium Peroxide Fusion, Chrome Iron Ores" in chapter under Chromium.) Boil the filtrate for 30 minutes to expel the excess of peroxide, make
acid with 50 m!. of H 2S0 4 (1 : 3) in excess, cool, dilute to a volume of about 500
m!. with cold water, add a measured excess of ferrous ammonium sulfate solution, more than sufficient to completdy reduce the chromium and titrate back
with the standard KMn04 solution.
Proceed as in " Determination of Chromium, Analysis of Nichrome."
DETERMINATION OF MOLYBDENUM

M~thod.-Dissolve 0.5 g. of the sample in a 250-ml. beaker with 30 ml. of

HCI (sp.gr. 1.18) and 30 mi. of HNO. (sp.gr. 1.42), evaporate to a volume of
20 mI., add 5 ml. of H 2S0 4 (sp.gr. 1.84) and evaporate to fumes of H 2S0 4
Cool, add 2 m!. of water and fume a second time. Cool, dilute to 50 mI. with
water, bring to boiling to dissolve salts, filter and wash with hot water. Reserve the filtrate. If the residue contains undissolved alloy, transfer to a
porcelain crucible, dry, ignite at a low heat, cool, transfer to a small nickei
crucible and fuse with 10 times its weight of a mixture of equal parts of Na202
and Na 2CO a Cool, leach with least amount of water, filter and wash. Discard the residue. Boil the filtrate 30 minutes to expel the excess of peroxide,
make slightly acid with H 2S0 4 (1 : 1) and boil to expel CO 2, Combine with
the main filtrate.
Bring the combined solutions to boiling and pass H 2S for 15 minutes. The
volume at this point should not exceed 175 m!. Transfer to a 250-m!. pressure
flask, stopper and immerse in hot water for 45 minutes. Let stand over night.
Filter on a 9-cm. filter and wash with cold dilute H 2S0 4 (1 : 50) through which
H 2S has been passed. Dissolve any adhering precipitate in: the flask and on
the tube used for passing the H 2S gas with 10 ml. of hot HNO. (sp.gr. 1.42),
wash into a 250-mi. beaker, add the filter paper containing the precipitate, then
add 5 m!. of H 2S0 4 (sp.gr. 1.84) and evaporate to fumes of H 2S0 4, taking the

1412

ANALYSIS OF NON-FERROUS ALLOYS

necessary steps to remove carbonaceous matter and the last traces of nitrio
acid. (See" Determination of Lead, Bismuth, e1)c. in Tin Base Bearing
MetaL") Cool, add 75 ml. of water and boil to a clear solution. Partially
reduce the molybdenum with 5 g. of granulated \zinc' and proceed as in
" Determination of Molybdenum in Wulfenite or Molybdenite" in chapter
under Molybdenum. II
I
DETERMINATION OF MANGANESE

33

Solutions Required. Ferrous Ammonium Sulfate and oj N KMn04

Solution.-See " Determination of Chromium, Analysis of Nichrome." (Each
milliliter of 0.1 N KMn04 equals 0.0027465 g. of manganese.)
Method.-Weigh 2 g. of the sample into a 600-ml. beaker, add 30 ml. of
HCl (sp,gr. 1.18) and 30 ml. of HN0 3 (sp.gr. 1.:t2) and heat until dissolved.
If the residue remaining contains undissolved alloy, dilute, filter and wash.
Transfer the residue to a porcelain crucible, dry, ignite at a low heat, cool,
transfer to a small pure iron crucible, fuse with 10 times its weight of N a 202,
cool, leach with least amount of water, remove and wash the crucible, make the
solution acid with HN0 3 (sp.gr. 1.42) ana boil to expel H 20 2. Combine this
recovered portion with the main filtrate and evaporate the combined solutions
to a syrupy consistency. Take up with 50 ml. of HN0 3 (sp.gr. 1.42), evaporate
as before, add 100 ml. of HN0 3 and boil down to a volume of 80 ml. Add 3 g.
of KCI0 3 in portions of 0.5 g. each ana boil at least 5 minutes after the last
addition of the KCI0 3 Add 100 ml. of water, filter through a thin asbestos pad
with suction and wash thoroughly with dilute HN0 3 (1 : 200). Transfer. the
pad and precipitate to the original beaker being sure to recover any precipitate
adhering to the funnel.
Now add 100 ml. of water, 50 ml. of H 2S0 4 (1 : 3) and a measured am6unt
of ferrous ammonium sulfate solution, slightly more than sufficient to dissolve'
the Mn02 (about 20 ml. should suffice). When the Mn02 is completely dissolved, titrate the excesS of ferrous ammonium sulfate solution with the standard KMn04 solution.
"
Ascertain as in " Determination of Chromium, Analysis of Nichrome" the
numper of milliliters of standard KMn04 solution that would be required to
titrate an amount of ferrous ammonium sulfate'solution equal to that added to
dissolve the Mn02. The difference between the'number of milliliters of KMn04,
solution calculated and the number used for the titration-of the, sample represents the manganese present and the percentage is found by multiplying this
volume by 50 times the manganese titre of the KMn04 solution:
DETERMINATION OF SILICON
Method.-Proceed as in " Determination of Sili~on, Analysis of Nichrome."
33 See "Ford-Williams Method for Manganese" in chapter under Manganese; also,
Proc. A. S. T, M., Part I, p. 598, 1929.

ANALYSIS OF NON-FERROUS ALLOYS

1413

TENTATIVE METHODS OF CHEMICAL ANALYSIS OF PIG LEAD.

A. S. T. M. METHOD

34

SERIAL DESIGNATION: B 35-36T 35

A.

DETERMINATION OF SILVER

Weigh 100 g. of the sample into a 3-in. scorifier and heat in a muffle furnace
until the assay" covers." Pour into an iron mold and allow to cool. Free the
resulting lead button from PbO, scorify again and pour as before. The button
last obtained should not weigh over 20 g. and can be cupelled directly. Weigh
the silver button obtained upon an assay balance.
NOTE.-If the amount of silver is large, the button should be examined for gold, in the
usual manner.

B.

DETERMINATION OF BISMUTH

Solutions Required. Sodium Carbonate.-Dissolve 100 g. Na 2 C0 3 in a

liter of distilled water.
Ammonium Carbonate.-Make a half saturated solution.
Alkaline Sulfide Wash Solution.-Dissolve 200 g. of KOH in a liter of
distilled water and mix one part of this solution with 4 parts of H 2S water.
Method. (a) Determinationfor Ordinary Amounts of Bismuth.-Dissolve
20 g. of the sample in a 400-ml. beaker with 100 ml. of HNO a (1 : 4), with the
aid of heat. When solution is complete, add dilute ammonia (1 : 2) with
constant stirring, drop by drop from a burette, until a faint opalescence appears.
If an actual precipitate is formed, redissolve by the addition of a small amount
of HNO a (1 : 4) and repeat the addition of ammonia. Now add 5 ml. of lICI
(1 : 9), fill the beaker with hot water, bring to boiling, and allow to stand on a
steam bath for two hours. The assay, while standing, must not reach the
boiling temperature. Filter through a 7-cm. paper. Transfer the precipitate
completely to the paper by means' of a "policeman" and wash twice with
hot water. Carefully examine the filtrate, washings, and any decanted liquid
and reject if clear. Dissolve the precipitate by dropping around the edges,
from a 5-ml. pipette, 5 ml. of boiling HCI (1 : 9), receiving the sulution in the
original beaker. Wash the paper thoroughly with hot water, fill the beaker
with water, bring to boiling, and allow to stand as before. Filter the bismuth
oxychloride upon a weighed Gooch crucible, wash thoroughly with water, once
with alcohol, once wIth ether, and dry upon the hot plate. Cool and weigh.
NOTE.-If time permits it is convenient to allow the assay to stand over night. In
that case, the precipitate of bismuth oxychloride generally settles so completely that the
clear supernatant solution can be decanted.
34 This method is reproduced through the courtesy of the American Society for
Testing Materials, 260 S. Broad St. Philadelphia, Pa.
35 These methods are issued under the fixed designation B 35; the final number indicates the year of original adoption as standard or, in the case of revision, the year of
last revision. Issued as Tentative, 1919; Adopted, 1920; Revised, 1924; Reverted
Tentative, 1936. See A. S. T. M. Methods of Chemical Analysis of Metals, p. 174,
December, 1936.

1414

ANALYSIS OF NON-FERROUS ALLOYS

\

(b) Determination for Amounts of _B_ismuth Smaller than Can be Determined by (a).-Dissolve 100 g. of the sample in 500 ml. of dilute HNO g (1 : 4).
When solution is complete, allow to cool and add N a 2 CO g solution little by
little until a slight permanent precipitate has for,med. Then add 50 ml. of
the Na 2 CO g solution, bring to boiling, allow to stand warm until the supernatant liquor is clear again, filter and reject the iJiltrate. _Dissolve the precipitate without washing by slo,vly pouring hot HNO g (1 : 4) around the edges
of the filter paper, using no more acid than necessary. Wash the paper once
with hot 'water and determine bismuth in the filtrate as described in (a).
NOTE:-When the sample contains a small amount of bismuth, it is often difficult to
ascertain when the correct amount of ammonia has been added to the nitrate solution.
In this case, place a small piece of litmus paper in the solution, and add the ammonia very
gradually until the litmus p,aper indicates a neutral reaction.

(c) Determination of Bismuth in Samples Containing Appreciable Amounts

of Tin and Antimony.-Use a piece of litmus paper as in Note under (b), and
after the first precipitation of bismuth oxychloride has been filtered off and
washed (see (a proceed as follows: Dissolve the bismuth by dropping around
the edges of the paper 10 ml. of boiling HCI (1 : 2), receiving the solution in
the original beaker. Discard the paper after washing. Dilute the solution'
to about 200 ml. with fresh H 2 S water and then pass H 2S gas through the hot
solution for 15 minutes. Filter and wash with hot water. Remove any tin
or antimony present by washing threfYtimes with alkaline sulfide wash solution.
Wash the precipitate again with hot water, place it, together with the filter
paper, in a 100-ml. beaker, add 20 ml. of HNO a (1 : 4), boil until sulfides are
completely dissolved and the paper well pulped. Filter the solution, receiving
the filtrate in the original beaker, and wash well. Determine bismuth in the
original beaker, and wash well. Determine bismuth in the filtrate as in (a).
NOTEs.-When the sample cc;mtains comparatively large amounts of tin or antimony,1
the residue left after the solution of the lead in the nitric acid obscures the opalescence
found upon the addition of ammonia.
_
I
If the original sample contains more than 0.25 per gent of bismuth, it is preferable
to use only a IO--g. charge.
:
C.

DETERMINATION' OF ARSENIC

Method.__f_Dissolve 111.11 g. of the sample in 550 m!. of HNO g (1 : 4).

When solution is complete wash into a graduated liteJ': flask, add 75 ml: of
H 2S0 4 (1 : 1), cool, and make up to the mark with water. -Transfer to a large
beaker, rinsing out the flask with 25 ml. of water. Mix thoroughly, allow to
settle and filter off 900 mI., equivalent to a 100-g. charge. Evaporate in a large
porcelain dish until only enough H 2S0 4 is left to moisten the residue. When
cool, wash into a small distilling flask with 60 ml. of HCI (sp.gr. 1.18) and 20
ml. of water, cleaning the dish carefully. Add.IO'g. of ferrous sulfate and distill,
boiling to as small a volume as possible. When cool add 50 ml. more of HCI
(sp.gr. 1.18) and redistill. Pass H 2S gas through the cold distillate for 45
minutes. Filter, and weigh the As 2S g on a Gooch crucible, washing with cold
water, alcohql and CS 2 After drying and weighing, redis~olve with (NH 4)2COg
solution and reweigh the Gooch crucible, calculatiD-g-the
. _,.
.loss in weight to arsenic .

NOTE.-The 25 ml. of water added is equivalent to the volume of precipitated lead

sulfate.

ANALYSIS OF NON-FERROUS ALLOYS

D.

1415

DETERMINATION OF REMAINING METALS

'Solutions Required. Tartaric Acid.-Dissolve 50 g. of tartaric acid in

250 mI. of distilled water to which has been added 250 ml. HCl (sp.gr. 1.18).
. Method.-Dissolve 222.23 g. of the sample in 1100 ml. of HNO a (1 : 4),
using a 1300-mI. beaker. When solution is complete examine for color and
turbidity. If clear, wash the solution at once into a 2000-ml. graduated flask.
In case of a residue (Note 1), however, dilute to about 1100 ml. and allow to
stand until the supernatant liquor is clear. Decant as much as possible into
a 2000-mI. flask, filter the remainder and receive the filtrate in the same flask.
Wash the precipitate well and then place it, together with the filter paper, in,a
100-mI. beaker and add 20 mI. of the tartaric acid mixture. Heat to boiling
and when the paper is well pulped allow to digest warm for 30 minutes. Now
add 50 ml. of hot water, filter and wash. (Note 2.) Carefully dry the residue
and ignite. If any appreciable residue remains, brush it into a small silver
dish containing I g. of molten KOH. Fuse for 5 minutes, and after cooling'
dissolve in as little hot water as possible and add to the above mentioned tartaric
acid filtrate. Render this solution just alkaline with ammonia and then just
acid with HCI, and saturate hot with H 2S gas. After digesting for 30 minutes
on the steam bath, pass H 2S through the solution again for 15 minutes. Filter
and wash with slightly acidified H 2S water. Reject the filtrate. Wash the
sulfides from the paper into the original beaker and add 5 m!. of KOH solution
(1 : 5) for every 25 mI. volume present. Digest hot for 5 minutes and filter
through the original paper into a small flask graduated to 110 m!. After washing with H 2S water containing a little of the KOH solution, cool the filtrate and
JIlake up to the mark. Mix and reserve 100 ml. as alkaline sulfide solution No.
1. The precipitate may be discarded.
Add slowly to the main solution in the 2000-ml. flask 150 mI. of H 2S0 4
(1 : 1). After cooling and filling l.1.p to the mark, pour into a clean 3-liter
flask provided with a rubber stopper.' Rinse the flask out with 50 mI. of water,
which is equivalent to the volume of lead sulfate present and is added to the
portion. After mixing thoroughly by shaking, allow the precipitate to settle
and filter off 1800 ml. of the liquid. This is equivalent to a 200-g. charge.
Place this in a No.9 porcelain evaporating dish and evaporate, first over a free
flame and later on the hot plate until only enough H 2S0 4 is left to moisten the
residue remaining. Add 50 ml. of water and, after digesting warm for a short
time, wash the solution into a 250-ml. beaker, cleaning the dish carefully.
Allow the solution to digest on the steam bath for 4 or 5 hours. (Note 3.)
Then filter, wash and. evaporate the filtrate to 200 to 250 ml. Place any
residue (Note 4), together with.the filter paper, in a 100-ml. beaker and treat
with 20 ml. of the tartaric acid mixture. Boil for 5 minutes, dilute with 50 mI.
of hot water and filter: Make the fi.ltrate alkaline with ammonia and just acid
with HCl, and. obtain the tin, antimony and arsenic as previously described,
reserving the whole of the alkaline sulfide solution as solution No.2. Reject
the sulfide residue.
To filtrate from the lead- 'sulfate, add ammonia until the neutrai point is
reached, arid then for every 50 mI. of the solution present add 2 ml. of HCI
(sp.gr. 1.18). Pass H 2S gas into the hot solution until saturated, digest for
30 minutes on the steam bath and again pass H 2S gas into the solution. Filter

1416

ANALYSIS OF NON-FERROUS ALLOYS

and wash with H 2 S water slightly acidified. (N ote 5.) Separate tin, antimony
and arsenic in the precipitate with KOH solution as usual, obtaining an alkaline
sulfide solution No.3. (Note 6.)
To sum up, three alkaline sulfide solutions hare been obtained, containing
tin, antimony and arsenic, a precipitate of metallic sulfides containing copper,
.
lead, etc., and a solution containing iron, zinc, nickel, etc.
NOTES.-l. A residue indicates the presence of antimony, tin, possibly arsenic, or
sulfur as lead sulfate.
2. It has been found that even this treatment occasionally fails to dissolve stannic
acid completely.
.
3. If it is preferred, allow the solution to stand over night to insure the complete
solution of all soluble salts.
4. Any residue of lead sulfate may contain some tin, antimony, or possibly arsenic.
5. The filtrate will contain any iron, zinc, nickel, cobalt and manganese; while in
the precipitate will . be found any copper, cadmium, lead, silver, bismuth, tin, antimony
and arsenic.
6. When separating the sulfides of arsenic, antimony and tin from sulfides of copper,
lead, etc" it is necessary to wash all the sulfides back into the beaker in which they were
precipitated. These sulfides sometimes cling so tenaciously to the paper that in dislodging them more water than the 25 m!. specified is required. In this case allow the
sulfides to settle and then decant the clear supernatant liquor through the filter until
the volume is reduced to 25 mI. Before rejecting the decanted fluid always test with
'
H 2Swater.
In washing sulfide precipitates with water, much trouble is experienced from the
tendency of the precipitate to pass through the filter in .the colloidal form. This is
particularly true in washing sulfides tl{at have been ,digested with KOH. Time and
trouble will be saved by washing all the sulfides precipitated from mineral acid solutions
with H 2S water containing a little of the acid in which they were precipitated, The same
is true of sulfides precipitated in or filtered from an alkaline solution. ,
When working with alkaline solutions in which tin is to be determined, avoid the
use of Jena or other glass that contains zinc. The zinc content of the glass may influence
the result.
'

DETERMINATION OF ANTIMONY AND TIN

Solutions Required. Potassium Iodide.- Dissolve 100 g. in a liter of distilled water,
'
I
Standard Sodium Thiosulfate.-Dissolve 24.8 g, of Na2S 20g5H 20 in 1000
ml. of distilled water, and allow to stand, for 24 hours. Standardize against
Antimony Metal c.p., using the same quantity of reagents and same procedure
as under method. Each milliliter is equivalent to approximately O.OO~ g. of
antimony.'
"- _,_
Method.-Wash the alkaline sulfide solutions Nos. 1, 2,. and 3 into a 6oo-ml.
beaker, and acidify with 5 m!. of HNO g (sp.gr. 1.42) and 20 m!. of HCI (sp.gr.
1.18). Evaporate the solution to dryness on' the steam bath. To the residue add 200 m!. of water, 10 g. of oxalic acid and 10 g. of ammonium
oxalate and heat the solution until it is clear. , Then pass H 2S gas through the
boiling solution for 45 minutes. Filter off the' precipitate, consisting of arsenic
and antimony sulfides, and wash with hot water. Determine tin electrolytically
in the filtrate, continuing the electrolysis until all the oxalic acid is decomposed
and the solution becomes alkaline. Dissolve the deposit on the cathode with
a small amount of HCI.and examine qualitatiyely for ,tin.
.
Dissolve the sulfides of arsenic and antHii15ny in KOH as usual, collecting
the filtrate in a 500-m!. Erlenmeyer flask. Add 50 ml. of Hel (sp.gr. 1.18)

ANALYSIS OF NON-FERROUS ALLOYS

1417

and boil the solution until about 30 m!. are left. Expel the arsenic as chloride.
Now oxidize the solution with a pinch of KClO s and boil until no more chlorine
remains. Cool and add 5 m!. of potassium iodide solution. Titrate the liberated iodine with 0.1 N sodium thiosulfate solution, using carbon disulfide as an
indicator.
DETERMINATION OF COPPER AND CADMIUM
Solutions Required. Sodium Chloride (Solution No. l).-Dissolve 1 g. of
NaCl in 100 m!. of distilled water.
Sodium Chloride (Solution No. 2).-Dissolve 10 g. of NaCI in 100 m!. of
distilled water.
.
Standard Potassium Cyanide.-Dissolve 2 g. of KCN in a liter of distilled
water and standardize against a known amount of copper as treated in the
analysis.
.
Sodium Carbonate (Solution No. l).-Dissolve 50 g. of Na 2 CO s in 1 liter
of distilled water .
. Sodium Carbonate (Solution No. 2).-Dissolve Na2CO~ in distilled water to
saturation.
.
Method. (a) Determination Where Copper Exceeds 0.0025%.-Place the
filter containing the sulfides in a 100-m!' beaker and add 20 m!. of HNO a
(1 : 4). Heat with occasional stirring until the paper is thoroughly pulped
and the sulfides are completely dissolved. Filter into a 250-m!. beaker. Dry
the residue, which generally contains a small amount of copper, ignite in a
porcelain crucible, boil with 5 ml. of HNO s (1 : 1), and wash into the main
portion, keeping the volume below 100 ml. Render it strongly alkaline with
ammonia, and add 5 g. of potassium cyanide, then saturate it in the cold with
H 2S gas. (N ote 1.) Filter the solution and evaporate the filtrate to a volume
of 20 to 30 m!. in a 4-in. porcelain casserole. Boil until solution is complete.
Add 20 m!. of H 2S0 4 (1 : 1) and evaporate the solution under a hood until dense
fumes of H 2S0 4 escape. Cool, dilute, and warm until copper sulfate is all
dissolved. Now filter, if necessary, into a 200-ml. beaker, render just alkaline
with ammonia, make acid by the addition of 3 m!. of HNO s per 100 m!. of
solution, and electrolyze for copper.
Dissolve the precipitate of sulfides in the usual manner with 20 m!. of HNO a
(1 : 4). Add 1 mI. of NaCl solution No.1 to the solution, still containing the
pulped filter, and digest for 30 minutes. Filter off the AgCI, wash and reject.
Make the filtrate, not exceeding 100 mI., alkaline with a slight excess of Na 2 CO a
and add 5 g. of KCN. Digest for 1 hour. Filter and wash with Na 2 CO a
solution No. 1. Reject the precipitate of bismuth. Now add a few milliliters
of ammonium sulfide ~olution to the filtrate to precipitate any cadmium as
yellow cadmium sulfide. Filter upon a weighed Gooch crucible and weigh as
cadmium sulfide. (Note 2.)
N OTES.-I. Copper remains in the solution, while lead, silver, bismuth and cadmium
are precipitated.

2. If an appreciable amount of cadmium sulfide is found, it should be converted

to and weighed as cadmium sulfate, according to method for cadmium under "Determination as Sulfate" in chapter under Cadmium.

(b) Determination Where Copper is Less than 0.0025%.- Place the

paper containing the sulfides in a porcelain cz:ucible, dry carefully and ignite.

1418

ANALYSIS OF NON-FERRO,US ALLOYS

When the carbon has been all burned off,- cool arid dissolve the residue in 5 to
10 m!. of HNO a (1 : 1). After evaporating to a voiume of 1 to 2 m!. add 1 m!.
of H 2S0 4 (1 : 1). Then evaporate the solution until fumes appear, cool, dilute,
add a few drops of N aCI solution No.2, and filter off the lead sulfate and silver
chloride. Again evaporate the filtrate until fume~ ~f H 2S0 4 ~ppear, and when
cold dilute and neutralize with Na 2 CO a solution I No.2. Then add about
six drops of concentrated ammonia and titrate the solution with standard KCN
solution until the blue color is discharged.
The cadmium can be obtained by making this solution, titrated for copper,
strongly alkaline, diluting a little, and adding 5 g. of KCN. Saturate the
solution cold with H 2S .gas, filter, discard the filtrate and treat the precipitate
for cadmium as described in Method (a).
DETERMINATION OF IRON
Method.-Evaporate the filtrate containing iron, zinc, etc., to 100 m!. and
oxidize with a few drops of HNO a Separate the iron with ammonia as usual,
making two separations, and'receive the :fiitrate in a 500-ml. Erlenmeyer flask.
Redissolve the iron hydroxide with hot HCl (1 : 1) or dilute H 2S0 4 and determine the iron volumetrically by any of ,the standard methods.
/

DETERMINATION OF ZINC
Solutions Required. Nitric Acid (1 : 4).-Mix 200 ml. of HNO a (sp;gr.
1.42) with SOO m!. of distilled water.
Sulfuric Acid (1 : l).-Carefully pour,. with stirring, 500 m!. of' H 2S0 4
(sp.gr. 1.84) into 500 m!. of distilled water.
Acidulated Hydrogen Sulfide Water.-Add 20 m!. of HCl (sp.gr. LIS) to
1000 m!. of distilled water and saturate with hydr'ogen sulfide.
Ammonium Thiocyanate Solution (2%}.-Dfssolve 20 g. of NH 4 C1\fS in
1000 ml. of distilled water.
;
Hydrochloric Acid (1 : 3).-Mix 100 m!. of HCI (sp.gr. US) and 300 m!.
of distilled water.
Standard Zinc Solution (0.1 mg. of zinc per ml.):;:-Dissolve exactly 0.1 g.
of U. S. Bureau of Standards pure zinc in 5 ml. of HCI {sp;gr._1.1S) and dilute
to exactly 1000 m!. with distilled water.
Potassium Ferrocyanide Solution.-Dissolve 34.S g. of -K 4 Fe(CN)6' 3H 20
in 1000 ml. of distilled water.
Method.-Dissolve 222.23 g. of the sample in 1100 m!. of HNO a (1 : 4),
using a 1300-m!. beaker. When the lead is disilblved, transfer the solution to a
2000-m!. graduated flask and add slowly 150 ml. of H 2S0 4 (1 : 1). Cool, fill
the flask to the mark and then pour the solution into a clean 3000-ml. flask
prbvided with a rubber stopp~r. Rinse tHe measuring flask with exactly 50 ml.
of water, which is equivalent to the volume of lead s~lfate which is present.
Mix the solution thoroughly by shaking, allow__the-precipitate to settle and filter
through a dry filter until 1800 mI. of filtrate has been obtained.

ANALYSIS OF NON-FERROUS ALLOYS

1419

Place exactly 1800 ml. of filtrate (equivalent to a 200-g. charge) in a No.9

porcelain evaporating dish and evaporate the solution to approximately 100 ml.
Transfer the solution to a 600-mL beaker, neutralize with ammonia, and then
add 5 ml. of HCI (sp.gr. 1.18) for every 100 ml. of solution. Warm the solution
and pass in a rapid current of hydrogen sulfide until it is saturated. Digest for
30 m:inutes on the steam bath, add an equal volume of water and again saturate
with hydrogen sulfide. Filter and wash with acidulated H 2S water.
Discard the precipitate and evaporate the filtrate in glassware containing no
zinc until the volume of the solution is approximately 100 ml.
,
Neutralize the solution with ammonium hydroxide, add 5 g. of citric acid(
and warm until the acid is dissolved. Add small portions of calcium carbonate
to the hot citric acid solution until about 1 g. 'of calcium citrate has separated
and then pass in a rapid current of H 2S as the solution is allowed to cool.
Allow the solution to stand for from 2 to 4 hours, part of the time on a water
bath, until the supernatant liquid is clear.
Collect the precipitate on a filter, wash with a 2%' solution of ammonium
thiocyanate and then dissolve the precipitate in hot dilute hydrochloric acid
{I : 3). If the solution has a reddish color (due to iron) the'zinc must be reprecipitated as above. If the solution is clear, evaporate it to dryness on the
steam bath, take up the residue in 3 ml. of HCI (sp.gr. 1.18), add 20 ml. of
water and filter if not perfectly clear.
Transfer the solution (Note 2) to a IOO-ml. Nessler jar and dilute to 95 ml.
Prepare other Nessler jars containing 3 ml. of HCI (sp.gr. 1.18) and definite
volumes of standard zinc solution. Dilute to 95 roI. Add 5 ml. of potassium
ferrocyanide solution to each jar, mix quickly, and compare the turbidities by
viewing longitudinally as the jars are held over a sheet of fine print. Add more
of the standard zinc solution from a burette to the jar which approximates the
turbidity of the unknown most closely, until the turbidities match each other
and calculate the percentage of zinc on the basis of a 200-g. sample' or the
aliquot portion taken.
NOTES.-1. All glassware that contains zinc must be avoided and'in umpire work a
blank test should be carried along with the test,
2. The whole solution can be used if the lead contains no more than 0.00025 % of zinc.
If more zinc is present it is best to take such an aliquot'portion of the solution as will give
approximately 0.5 mg. of zinc and then to add enough Hel to provide 3 ml.
3. For further details concerning the turbidmetric test, consult the "Determination
of Small Quantities of Zinc" by Mr. Bodansky, in the J. Ind. Eng. Chem. 13, pp. 696697 (1921).
4. The addition.of calcium carbonate with the formation of a precipitate of calcium
citrate serves the purpose of giving a clear filtrate, and prevents the loss of colloidal
sulfide.
.

DETERMINATION OF NICKEL AND COBALT

Solutions Required: Hydrogen Sulfide Wash Water.-To each 100 mI. of
hydrogen sulfide water add 20 ~l. neutral ammonium acetate.
. Method.-Render the filtrate just alkaline with ammonia and saturate with
H 2S. Heat to boiling and then make just acid with acetic acid, add 20 ml. of
neutral ammonium acetate solution, and boil until the sulfides of nickel and
cobalt separate out. Filter and wash with warm H 2S wash water. Dry the

1420

ANALYSIS OF NON-FERROUS ALLOYS

precipitate and paper in a porcelain crucible and carefully ignite. If there is

an appreciable amount of residue after ignition, di~solve by boiling with 10 ml.
of aqua regia, wash into a 250-ml. beaker, add 10 ml. of H 2 S0 4 (1 : 1), evaporate until fumes appear, cool, dilute to 200 mI., make alkaline with ammonia,
and add 15 ml. of concentrated ammonia., Then ~lectrolyze the solution and
weigh the nickel and cobalt as such.
NOTEs.-If the amount of nickel and cobalt is small it can be weighed as oxide.
If the filtrate from the nickel and cobalt sulfides shows a brown color, it indicates
that the precipitation has not been complete. In this case render the solution ammoniacal and repeat the above process.

ALLOYS, IT
THE CHEMICAL ANALYSIS OF IRON AND STEEL

INTRODUCTION
The former division of the subje.ct of steel analysis into the. fields of plain
carbon steels and alloy steels is no longer followed"in the better treatises in this
field. Due to the cooperation of the National Bureau of Standards and the
various industrial laboratories and societies for the standardization of methods
of testing, the referee methods in this field are now well standardized, and for
this reason the subject of steels and ferroalloys is now covered by reproducing
the methods of the American Society for Testing Materials. There are included in this chapter some selected methods for various elements, under the
caption" Other Methods" (p. 1470). Those who desire a still fuller selection
of methods should consult one of the more extended treatises in this field. 2
The present selection should serve for all practical purposes.

TENTATIVE METHODS OF CHEMICAL ANALYSIS OF STEEL, CAST

IRON, OPEN-HEARTH IRON AND WROUGHT IRON 3

A. S. T. M.

DESIGNATION:

E 30-36 T

PREFACE
The following methods have been compiled as standard procedures for use
in referee analyses. These methods, however, are not intended to preclude
the use of other methods (or apparatus) that give results within the permissible
tolerances. In any case, the analyst should check the method and technique
1 Methods from the literature; arranged by N. H. Furman.
2 Lundell, Hoffman and Bright, Chemical Analysis of Iron and Steel, J. Wiley &
Sons, Inc., 1932. New York.- C. M. Johnson, Chemical Analysis of Special Steels,
John Wiley & Sons, Inc., New York, 1930: Methods of the Chemists of the U. S. Steel
Corporation, published by the Carnegie Steel Co., Pittsburgh, Pa. W. W. Scott,
Technical Methods of Metallurgical Analysis, D. Van Nostrand Co., New York, 1923.
F. Ibbotson, The Chemical Analysis of Steel-Works Materials, Longmans Green & Co.,
New York, 1920.
3 Under the standardization procedure of the Society, these methods are under the
jurisdiction of the A. S. T. M. Committee E-3 on Chemical Analysis of Metals. Reproduced with the permission of the American Society for Testing Materials, 260 S. Broad St.
Philadelphia, Pa.
4 This is a Tentative Standard and under the Regulations of the Society is subject to
annual revision. Suggestions for revision should be addressed to the Headquarters of
the Society, 260 S. Broad St., Philadelphia, Pa. Issued, 1936. Accepted for publication
as tentative by Committee E-lO on Standards, August 26, 1936.
1421

1422

CHEMICAL ANALYSIS OF lRON AND STEEL

that is used by means of a National Bureau Gf Standards' standard sample that

has a composition comparable with that of thE! material under test. A list of
these Standard Samples is given in the Bureau's Supplement to Circular 398.
Concentration of Acids.-Throughout the~e metho"U~, the use of c_oncen.
trated acids is intended, unless otherwise specified. For example, hydrochloric
acid, or HCI, refers to the concentrated acid ($p.gr. 1.18) as furnished by the
manufacturers. Diluted acids will be designated as in the following example,
" diluted HCI (2 : 98) " will mean 2 volumes of the concentrated acid of.sp.gr.
1.18 diluted with 98 volumes of water.
Water.-The term" water" as used in these methods refers only to distilled'
.
water.
.
PRECAUTIONS
Attention is directed to certain precautions to be observed in the use of the
reagents employed in the analytical procedures described in thege methods.
Perchloric acid has become valuable reagent in the analysis of ferrous
alloys. While perchloric acid of 70% strength may be boiled with i91Punity
at approximately 200 0 C., it cannot b;l too strongly emphasized that contact
of the boiling undiluted acid or hot vapor of perchloric acid with organic matter,
or even easily oxidized inorganic matter such as compounds of trivalent
antimony, will lead to serious explosions. If oxidizable substances are present,
nitric acid must always be added a~ a precautionary measure before evaporqting
to fumes of perchloric acid. Do not use rubber thumb stalls around fuming
perchloric acid. Handle with beaker tongs. Perchloric acid evaporations
should be done in a hood with good draft. Hoods with wood frames,in which
a great deal of HCI0 4 is fumed should be washed occasionally with water..
Hydrofluoric acid burns are both painful and dangerous, as inf.ection and
necrosis of bone tissue may occur. If hydrofluoric acid has touched the skin
wash immediately with plenty of watel' and soak the exposed part in a strong
solution of borax.
.
Nitric dioxide vapors are cumulatively harmful to the respiratory system
so solutions of metals in nitric acid should be !hade in a well-ventilated hood.
Hydrogen sulfide is a dan~erous poison which is fatal if inhaled in sufficient
concentrations. Precipitations and filtrations involving this reagent should
be done under the hood.
I

SAMPLING
(a) Rolled and Forged Steel Products.-Different parts of a piece of steel
:vary in composition. For this reason, a sample from a single piece must be
carefully selected if it is to be representative of that piece. If the analysis is
to be representative of the melt, a number of representative pieces should be
sampled and analyzed separately. In any c,se, the sample flhall be so selected
as to be thoroughly representative, and large enough to suffice for all of the
required determinations.
It_ should be. remembered that the composition of the steel is changed by
certain operations and that samples should be taken from the steel in its original
I
condition.
"
Preparation of Samples.-Samples shaH..'Consist of drillings or chips cut by
some machine tool without the application of water, oil, or other lubricant,

/'

CHEMICAL ANALYSIS OF IRON AND STEEL

1423

and shaH be free from scale, suriace II\etal, grease, dirt, or other foreign subIf the samples are taken by drilling, a drill not less than ~ nor more
than %: in. in diameter shall be used. Samples shall be uniform, thoroughly
mixed, and free from dust. Chips too coarse to pass a No. 20 sieve aTe not
recommended, nor shall long 'curly urillings which will not pack closely for the
carbon determination be used.
In referring samples to other analysts for check analyses, ipieces of the
original full-size section, when possible, should be submitted rather than
cuttings, unless the latter are specifically requested.
,
Location of Samples: Large Sections.-For large sections, iJlcluding blooms,
billets, slabs, rounds, squares, shapes, etc., samples shall be taken at any point
midway between the outside and the center of the piece by drilling paranel to
the axis. In cases where this is not practicable, the piece shall be drilled on the
side, see Fig. 183 (b) and (c), but the drilling. shall not be collected until they
represent the portion midway between the outside and the center. The
" tension test specimen may be used for sampling if it conforms to the abo've
conditions.
~
..
Bored Forgings.-For bored forgings, samples shall be taken midway
between the inner and outer surface of the wall.
Thin Af aterial.-For thin material or material of small cross-section, such
as plates, shapes, bars, etc., if the method described above for large sections
is not applicable, the sample shall be taken by machining off the entire crosssection, or, if this is not possible, by drilling entirely through the material at a
point midway between the outside and the center, See Fig. 183 (a).
Sheets Rolled Longitudinally.-For sheets rolled from slabs or bars longitudinally, the specimen for sampling shall be cut 2 in. in width and' across the
full width of the sheet as rolled. The specimen shall be cleaned by pickling
or grinding and then folded once or more by bringing the ends together and
closing the bend. The sample for analysis shall be taken in the middle of this
length by milling the inside sheared edges or drilling entirely through from the
nat surface. Sampling by milling is preferable. For sheets of a light gage
more than one specimen may be taken and stacked together before folding.
Sheets Rolled Transversely.-For sheets rolled from slabs or bars transversely,
the specimen shall be cut from the side of the sheet, half way between the middle
and end as rolled, 2 in. in width and 18 in. in length. If the sheet is No .. 20
stances.

(c)
(b)
FIG. 183.-Location of Samples for Check AnalysiS.
(~)

"

1424

CHEMICAL ANALYSIS OF lRON AND STEEL

gage (24 oz. per sq. ft. or 0.037 in. in thickness) or lighter, the specim~n shall
be cut from the full length .of the sheet as rolle'd. The specimen selected shall
be cleaned by pickling or grinding and then folded. once or more by bringing
the ends together and closing the bend. The s~mple for analysis shall be taken
in the middle of this length by milling the inside sheared edges or drilling entirely through from the fiat surface. Sampling by milling is preferable.
Sheets Not of the Full Size, Rolled.-Sheets cut from larger sheets alld not
of the full size rolled shall be sampled by milling or drilling the sheet in a
sufficient number of places so that the flam pIe is representative of the entire
sheet. The sampling may be facilitated by folding the sheet both ways.
.
(b) Pig !ron.-In the absence of a special agreement between the manufacturer and the purchaser, or whenever it is necessary to sample pig iron in the
solid state, the sampling shall be carried out in the following manner: Not less
than 3 pigs shall be taken to represent any lot or shipment and for lots of more
than 30 tons one pig shall be taken for each 10 tons of iron. The pigs shall be
selected by some means such as the knotted rope system, that eliminates the
element of personal choice; and 3 to 7 pigs taken in tile order selected, shall
constitute a unit sample. For boat o'r barge shipments arrangements between
the seller and the purchaser may be made, if desired, in regard to taking fewer
pigs per ton. If for any reason it is desirable to fracture the pigs, they may be
broken in two, and one portion of each of the original pigs reserved for the
sample for analysis. Any loose sand or other deleterious matter on the reserved
.
portions shall be removed, conveniently by brushing.
The sample for analysis shall be collected after the proper discard, by drilling
each pig or portion of pig with a properly sharpened and hardened, fiat beaded,
%- or ;!i-in. highspeed tool-steel drill in a direction at right angles to the long
axis of the pig and at such a feed as to form a minimum amount of fine drillings.
The first drillings shall be discarded and only the drillings collected after the
drill is cutting into clean metal shall be reserved for the sample. The drilling
shall extend to within U in. of the opposite surface of the pig; and suitable
precautions shall be taken to collect all the ,drillings, fine as well as coarse
particles, and to avoid contamination of the sample in any way. Any pigs
or portions of pigs found too hard to drill readily may be annealed by heating
to a dark red color and cooling in air.
The sample for analysis shall be cOIlj.posed of equal portions by weight of
the drillings from the pigs forming the unit saml).le, and the weight of the
combined sample shall not be less than 75 g. The drillings.shall be thoroughly
mixed, as.. by gently grinding in a suitable mortar until all pass the No. 20
(840-micron) sieve. 5 Unless accurate determinations of combined and graphitic
carbon are required, the sample shall be carefully divided into two or more
portions by pouring in a cone-shaped pile on glazed paper and halviIlg or
quartering in the usual manner. These portions, of which one shall be reserved
for the manufacturer and one for the pllrchaser, shall be placed in a suitable
container, fully labeled for identification, and dispatched to the laboratories
without unnecessary delay and with full instructions as' to the determinations
required.
6 See Standard Specifications for Sieves for-Testing Purposes (E 11-26), 1936 Book
of A. S.;. M. Standards, Part II, p. 1413. Reproduced onpp. 1329-1333 of this Volume.

CHEMICAL ANALYSIS OF IRON AND STEEL

1425

When combined. and graphitic carbon must be determined with the highest
degree of accuracy, equal portions by weigJ:tt of the drillings from the pigs or
parts of pigs representing a unit sample shall be mixed to form a sample of at
least 100 g. which shall then be separated into coarse and fines by sieving as
described under (c), the sampling of gray-iron castings.
(c) Gray-Iron Castings.-In accordance with the Standard Specifications
for Gray-Iron Castings (A. S. T. l'4. Designation: A 48) of the American
Society for Testing Materials,6 three test bars, 1.20 in. in diameter shall be cast
in sand from each heat after one quarter and three quarters of the heat has been
poured.
.
One bar from each set having been broken, one end of each next to the
fracture shall be thoroughly cleaned !1nd the outer skin removed for a sufficient
distance from the fracture and down to clean metal. Chips then shall be
taken by means of a lathe or milling machine ll-cross the whole face of the bar
until not less than 100 g. has been collected. The same amount shall be
taken from each bar. The bar shall be so clamped as to' permit the attachment. or use of any suitable device for collecting every part of the sample and
the machine shall be run slowly enough to reduce to a minimum the danger of
loss of fine particles (avoid drafts). Horizontal drilling may also be employed.
Preparation of Samples from Castings for Analysis.-The entire gross sample
shall be weighed and then sifted on a No. 80 (177-micron), and if need be a No.
120 (125-micron), tight-fitting sieve (with cover). 7 Both sieves and cover m,ust
be tight-fitting to avoid loss of fine graphite by dusting. As an alternate procedure (a) the drillings may be sifted through two sieves of such sizes that not
under 10% nor more than 20% of the entire sample remains on the larger
sieve, and not under 10% and not over 20% shall pass through the smaller
sieve or (b) sieves may be used of such sizes that concordant results may be
obtained.
The two (or three) portions so obtained shall be separately weighed. Each
one shall then be thoroughly mixed without any loss of material and divided
by weight into two (or three) exactly equal portions, each of which shall be
placed in a clean, glass-stoppered bottle or other suitable receptacle and
appropriately labeled. Of the three sets of the subdivided sample, one shall
be placed in a clean, glass-stoppered bottle or other suitable receptacle and
appropriately labeled. Of the three sets of the subdivided sample, one shall
be retained by the manufacturer, one sent to the purchaser and one held in
reserve.
Before weighing the sample for analysis, the contents of each bottle shall
be thoroughly mixed. Carbon is determined on each of the sieved portions,
and the carbon present in the original drillings is calculated from the ratio
of the weights of the separate sieved fractions to the original iron.
As a general rule, the very fine material is richest in carbon, the coarse
material next, while the intermediate sizes are the poorest. For routine
analyses, the aim shoulct be to obtain the iron in as small particles as possible,
as for example by milling. In such case, the samples for graphite and carbon
may be taken- by forming a corie, flattening it out, and taking small dabs with
61936 Book of A. S. T. M. Standards, Part I, p: 457.
7

See Standard Specifications for Sieves for Testing Purposes (E 11-26),'1936 Book of

A. 'S. T. M. Standards, Part II, p. 1413.

1426

CHEMICAL ANALYSIS OF IRON AND STEEL .

a spatula from at least ten, representative portions for each weighed portion.
Less care is needed in taking portions for the p.etermination of the other constituents; grab samples from the container after gentle tumbling are satisfactory.
(d) Malleable Iron.-Proceed as in the sarilpling of pig iron.
'
(e) Wrought Iron.-As it is desirable that ~he sample represent the entire
cross-section of the specimen, sampling by millin& is preferable. Specimens
shall be freed frdm any mill scale by pickling, gtinding, or other- suitable means.
It is necessary to secure all of the millings, for the various sized particles will
differ in composition. For routine work, samples for carbon, manganese,
phosphorus, sulfur, and silicon may be taken from the thoroughly mixed.
millings. For the determination of slag and oxides, or for umpire determinations of the other constituents, the entirll sample shall be weighed and then
sifted without dusting on a No. 35 (500-micron) sieve. The two fractions
shall be weighed and in taking samples for analysis, portions of these two wellmixed fractions must be taken in the same proportion which they bear to the
gross sample.
TOTAL CARBON BY THE DIRECT-COMBUSTION METHOD
The method of direct combustion of the -metal in oxygen is recommended,
the carbon dioxide obtained being absorbed in soda-asbestos with suitable
purifying and protecting trains following the furnace. Owing to the diversity
of apparatus by which correct reSults may be obtained in the determination of'
carbon, the recommendations are intended to indicate what is acceptable rather
than to prescribe definitely what shall be used.
. Apparatus and Reagents. Combustion Apparatus.-Any combustion apparatus heated by electricity (preferably) or gas which will heat the sample to a
temperature of 1000 0 to 1100 0 C. may be used. Electric furnaces, heated by
silicon carbide rods giving temperatures up to 13700 C. are very useful for
determining carbon in refractory alloys. Combustion tubes may be of porcelain, sillimanite, clay, quartz or platinum arid obviously must be gas tight.
Quartz is liable to devitrification when used_( intermittently at temperatures
above 1000 0 C., and may then become ~porous. Tubes 30 in. in length with
l~-in. inside diameter and drawn out at one end are convenient for the small
Gooch crucibles of fused silica that are used in the determination of graphite.
CataZyzers.-Though materials such as ironized asbestos, copper oxide,
platinized <@artz or asbestos, or platinum gauze are often put in the exit end
of the combustion tube for the purpose of insuring complete combustion of
carbon -to the dioxide, it is questionable whether they are of any value, save
as baffles for holding back iron oxide dust and sulfur trioxide, for they soon
become fouled. When such baffles are used the fixed sulfur shall be burned
out every 200 determinations or oftener, by drawing the exit end of the combustion tube into the hot zone while a curren,t of air or oxygen is passed through
it. 'With high sulfur steels it is best to omit baffies in the exit end of the tube
and insert a specia~ sulfur dioxide absorber or purifi~r between the tube and
the carbon dioxide absorber.

Boats and Covers.-The boats and covers may'be of porcelain, quartz,

alundum, clay, platinum or nickel, and s.b.6uld preferably receive a lining of
granular alundum or any other material" found to be suit;tble for.the purpose.

CHEMICAL ANALYSIS OF IRON AND STEEL

1427

Nickel boats shall not be made of sheet nickel containing more than 0.10%
of carbon. Before use, new boats and covers shall always be preheated in
oxygen until no more carbon dioxide is given o.ff or until a constant blank is
obtained. In order to p!event injury to thEl tube from spattering, a platinum
or nickel coveJ," open at both ends and allowing free access of oxygen is desirable.
Material for Lining Boats.-" RR Alundum" (AI203) alkali-free, specially
prepared for carbon determinations, and 50-mesh or finer, is entirely satisfactory for the lining of boats. Ignited low-silica chrome ore, or " Zirkite "
(natural oxide of zirconium) properly sized and freed from materials causing
a blank may be employed. Ferric oxide has been recommended as permitting
combustion at low (800 0 to 900 0 C.) temperatures but is not recommended for
high temperatures. Quartz sand, owing to its liability to fuse or to slag with
oxides of iron, causing bubbles. of gas to be enclosed, is objectionable.
Oxygen.- The purity of the oxygen shall be not less than 99.5%. Organic
matter of any kind is an undesirable impurity. It is usually absent, and it
suffices to pass the oxygen through an absorbent such as soda-asbestos followed
byaI.lhydrone. If carbonaceous matter is suspected, the oxygen shall be passed
through a tube that is loosely packed with asbestos or copper oxide and heated
to about 450 0 C. before it is passed through the soda-asbestos.
Combustion of the Sample. Size of Particles of Sample.-The finer the
chips (excluding dust, which causes low values on a hot boat), the better, except
with alloys that burn too vigorously. Drillings or millings sized between 14 to
60 mesh are satisfactory.
Distribution of the Sample in the- Boai.-In general, the sample shall be
p3:cked in a small groove or furrow that has been made in the beddin'g material
in the boat. If the material burns too rapidly, satisfactory regulation of the
speed of combustion' may sometimes be obtained by spreading the sample
somewhat loosely over the bedding material or by covering with a thin layer
of the bedding material.
Addition of Accelera{ors to the Sample.-Shot tin, Ys-in. pellets, is an
'excellent accelerator, one pellet being placed on each end of the layer of drillings.
Millings of open-hearth iron, or other iron with 0.01 to 0.03% carbon are
satisfactory aids in the combustion of large drillings and refractory alloys,
Red lead, copper oxide, lead and powdered copper are also used. Red lead,
unless previously ignited at 500 0 C. and kept in ground-glass stoppered bottles,
will cause erratic blanks.
Preheating of Sample.-If the sample is allowed to come to the temperature
of the furnace before the oxygen is admitted, it usually bursts into a bright
flame and burns completely. A period of 72 to 1,72 min. of preheating suffices.
Rate of Adding Oxygen.-The rate at which oxygen is admitted is also a
factor in the velocity of combustion. Assuming the combustion apparatus has
been heated to the temperature range above that recommended,;t is possible,
if the material is closely packed and if oxygen is admitted at too rapid a rate,
that the combustion may be so violent as to cause excessive spattering of fused
oxides and such fluidity of the molten slag that the boat or other container may
be injured or destroyed. Sufficient oxygen, however, shall be run in to insure
a current of gas through the absorber at all stages of the combustion. When tin is
employed as an accelerator, combustion is very rapid and it is therefore necessary to increase the flow of oxygen during combustion. When the sample has

1428

CHEMICAL ANALYSIS OF IRON AND STEEL

burned (1 to 2 min.) the flow of oxygen may be reduced to a rate of 150 mi.
per min.
.
I
."
Purification of the Carbon Dioxide.-The purifiers that foll~w the combustion tube must remove finely divid~d oxides of ir9n, oxides of sulfur or selenium,
dry the gases before they enter the weighed absorber, and protec~ the absorber
from outside effects. Iron oxide is screened out ~uring the passage of the gases
through the liquids or columns of solids that precede the weighed absorber.
The small amount of 80 2 that is given off from low-sulfur steels can be satis-

C D

(a) Apparatus used at the National Bureau of Standards.

<

A.
B.
C.
D.
E.
F.

Oxygen tank with needle or reducing valve.

Preheating furnace, 450 0 C, with platinized asbestos.
Ascarite tower.
/
Mercury manometer.
Combustion tube and electric furnace, 1000---11500 C.
Special absorption tube which contains asbestos in the left arm, and in the right
arm H 2S0 4 saturated with H 2Cr04.
G. Tower with 20-30 mesh Anhydrone. .
H. Weighed tower with Ascarite and Anhydrone.
I. H 2S0 4 guard tube and gage.

02;nno
r
~n __
~E:::~
rrn

4 .

10

(b) Apparatus used in a representative steelworks laboratory.

1.
2.
3.
4.
5.
6.
7.
S.
9.
10.

Mercury valve.
Bottle containing concentrated H 2S0 4.
Tower, with gooseneck top, containing CaC!..
Tower with stick NaOH.
Electric furnace with combustion tube and adapters.
Bottle with aqueous solution of KMn04 (5%) inserted only for steels with more
than 0.05% sulfur.
Tower with granulated zinc.
Bottle with concentrated H 2S0 4.
Tower with P 206.
Weighed tower with Ascarite.
I

FIG. 184.-Apparatus for the Determinatio]l-of-Carbon by the Direct-Combustion

Method.

CHEMICAL ANALYSIS OF IRON AND STEEL

1429

factorily removed by H~S04 that has been saturated with chromic acid. Highsulfur steels need other absorbents such as an aqueous solution of chromic acid
(50%), a solution of KMn04 (5%), or heated platinized silica gel that will
convert the dioxide to trioxide. The sulfur trioxide that is so formed is not
removed by anyone absorbent, but is condensed and absorbed during its
passage through the liquids or columns of solids in the train.
Absorbing Bulbs.-No special types of absorbing bulbs are recommended,
although the Fleming, Miller, Turner, and the Midvale (sometimes called
Stetzer and Norton) bulbs have proved satisfactory. When filled, the tubes
shall not weigh over 200 g. and they shall always be weighed filled with oxygen
and against a like counterpoise. Open bulbs, such as the Midvale, lose oxygen
by diffusion. They shall be filled with oxygen before weighing when not in
continuous use, and the same time interval shall be held between weighings.
Absorbents.-The most desirable absorbent for carbon dioxide is 20 to 30
mesh soda asbestos (Ascarite) followed by Anhydrone, Mg(CI0 4 )2. The latter
is needed to absorb the water which is formed during the reaction and is not
held. by the unused Ascarite.
General Arrangement of Apparatus.-Figures 184 (a) and (b) show two
typical arrangements of the apparatus. The apparatus and arrangement may
be modified, provided satisfactory results for the carbon determination will be
obtained. .For example, with oxygen free from carbonaceous matter the preheater may be omitted; other forms of absorption towers may also be substituted for the special absorption tube.
Procedure. Carbon Steel (Sulfur Content less than O.l%).-After having
properly set up and tested the apparatus, spread 1.36 to 2.73-g. of the sample
on the bed material in the boat so that the particles are in intimate contact.
Cover the sample with a suitable cover and introduce the boat into the hot.
combustion tube. Close the tube and allow the sample to heat for 1 to 2 min.,
depending upon the size of the particles. Then admit the oxygen, at a rate
of 300 to 400 ml. per min. while combustion is going on. When combustion
is complete, slow the current so that the gases leave the absorbents at the rate
of about 150 mL per min. for 6 to 8 min. in order to sweep out the carbon
dioxide. Withdraw the absorption tube filled with oxygen, place it in the
balance case for 10 min.,8 open momentarily, and weigh against a similar tube
used as a counterpoise. The increase in weight represents carbon dioxide.
Multiply this by 27.27 and divide by the weight of the sample to obtain the
percentage of carbon in the steel. Remove the boat from the tube and examine
the melt for evidences of incomplete combustion. If the drillings are not
thoroughly fused in a solid pig, the determination shall be rejected. The
proper correction for a blank run on every material used in the determination,
except the sample, shall be determined with special care in umpire analyses,
or in analyses of very low-carbon material.
High-Sulfur Steels.-Proceed as with carbon steels of low-sulfur content
but insert a special 80 2 oxidant in the train such as (1) a tube of platinized
silica gel heated to 440 0 C. and followed by a tube containing ironized asbestos
and anhydrone, (2) a tube containing about 25 m!. of an aqueous solution of
8 The tube will warm up when CO 2 is absorbed.
It is not necessary to wait until it
reaches room temperature if it is in continuous use, provided the same time interval is
maintained, and approximately the same amount of CO 2 is absorbed.

1430

CHEMICAL ANALYSIS OF IRON

AND' STEEL
i

chromic acid (50%), or (3) an aqueous solution of KMn04 (5%) followed by

suitable desiccants.
j
, Alloy Steels.-Proceed as with carbon steel8. Although most of the lowalloy steels burn perfectly at 11000 C. without the addition of an accelerator,
many alloy steels require an accelerator to obtai~ complete combustion at this
temperature.
,
If a steel burns with difficulty, either place a small pellet (Ys in.) of pure tin
on each end of the sample (or spread 1 g. of pure tin millings over the sample),
or mix with 1 to 2 g. of fine millings of open-hearth iron and proceed as usual.
With high-chromium high-nickel steels (18% chromium, 8% nickel; 20%
chromium, 20% nickel, etc.), the principal sources.of error in carbon determinations are (1) combustion at too low temperatures and (2) omission of, or faulty
corrections for, blank runs. These steels, if burned at the temperatures generally used for carbon determinations (that is, 1050 0 to 11000 C.), require the
aid of an accelerator, such as tin, open-hearth iron, or red lead. More certain
combustion of all carbon is obtained if a temperature of 1250 0 C. or higher is
employed and tin or open-hearth iron is used as an accelerator. Preheating is
necessary, and burning at higher pressures of oxygen seems advantageous;
that is, a sufficiently rapid stream of oxygen should be maintained during the
burnin(J so that it 2'S bubblin(J freely at the/exit end Of the train. After the burning
is completed the rate of flow of the oxygen is reduced to about 150 ml. per min.
and continued for about 8 to 10 min. If sulfur exceeds 0.06%, proper precautions shall be taken to eliminate oxides of sulfur.
Selenium Steels.-Alloy steels containing selenium evolve volatile selenium
oxides during combustion; therefore, a tower packed with asbestos fibers shall
be placed in the train between the exit end of the combustion tube and the
,C0 2 absorbing tube. An aqueous solution of KMn04 (5%) as used for highsulfur steels is also satisfactory for this purpose.
Pig, Cast, and Malleable lron.-Proceed as with carbon steels using a I-g.
sample. Special care shall be used in obtaining a representative sample for
the determination of both total carbon and g"aphite (see Sampling). Precautions for sulfur are of no particular moment for cast irons, except, of course,
for continuous combustions of irons of very high sulfur content.
Alloy Cast Irons.-High-silicon and most alloy cast irons require a temperature of at least 11000 C. for complete combustion. Accelerators are desirable for high chromium-nickel irons.
.
OPen-Hearth and Wrought Iron.-Proceed as with 'carbon steels using a
3- to 4-g. sample and pay special attention to proper blank corrections. With
very-low-carbon material, as open-hearth iron, the use of small absorption
tubes (Schwartz glass-stoppered, 10 cm.; U-shape, containing Ascarite and
Anhydrone) tends to reduce errors caused by variations in temperature and
humidity.
I

GRAPHITE IN CAST IRON

Special Solutions Required. Diluted Nitric ,Acid (sp.gr. 1.20, approximately).-Mix 380 ml. of HN0 3 and 620 ml. of water~
Ignited Asbestos.-Gently ignite prepa.reii- asbest6s under oxidizing conditions until all carbonaceous matter has 'been 'consumed. Prepared asbestos is

,CHEMICAL ANALYSIS OF IRON AND STEEL

1431

asbestos (finer than lO-mesh) that has been digested with HCI and then
thoroughly washed with water.
Potassium Hydroxide Solution (sp.gr. 1.10, approximately).-Dissolve 120
g. of stick KOH in 1000 ml. of water.
Diluted Hydrochloric Acid (1 : ZO).-Mix 50 m!. of HCI and 1000 m!. of
water.
Procedure.-Dissolve 1 to 3 g. of the sample (see Sampling) in 50 m!. of
HNOa (sp.gr. 1.20) and heat gently on the steam bath, with occasional stirring,
until action ceases. If silicic acid separates add 5 m!. of HF to facilitate filtration. Collect the residue, conveniently by suction on ignited asbestos contained,
in a fused-silica Gooch crucible of such diameter as will fit in the combustion
tube. Wash the residue thoroughly with hot water, then with a hot solution
of KOH (sp.gr. 1.10), hot water, then with diluted, HCI (1 : 20), and finally
with hot water. Dry at a temperature not exceeding 1500 C., and determine
the graphite by direct combustion at approximately 900 0 C. in the apparatus
used for the determination of total carbon, taking care to close the combustion
tube'immedjately after inserting the sample.
MANGANESE BY THE BISMUTHATE METHOD '
(ABSENCE OF COBALT)

Special Solutions Required. Diluted Nitric Acid (sp.gr. 1.135).-Mix 280

m!. of HNO s with 720.ro!. of water.
Diluted Nitric Acid (3 : 97).-Boil 40 ml. of HNO a under the hood for 1
min., cool somewhat, and pass in a current of purified air for 5 min. Mix
30 m!. of this acid with 970 m!. of water. Add 1 g. N aBiO a, shake, and allow
to settle.
Asbestos.~The asbestos for this work shall be shredded, digested with
HCI, thoroughly washed with water, then digested on the steam bath with
diluted HNO a (1 : 3), and finally thoroughly washed with water to remove
chlorides and to oxidize any ferrous iron contained in the asbestos. In making
th~ asbestos filter-pad, it is advisable to have a thin, but tight, bed and as
much surface as possible .. This insures rapid filtration, and the filter may be
used until it becomes clogged with bismuthate (about five to six times).
Bismuthate.-The sodium bismuthate shall contain enough active oxygen
to correspond to at least 75% NaBiO a. Manganese and chlorides shall not
exceed 0.0005 and 0.001 %, respectively.
Ferrous Ammonium Sulfate Solution.-Dissolve 12 g. of Fe(NH 4)2(S04)2.
6H 20 in a cold mixture of 950 ml. of water and 50 ml. of H 2S0 4.
Standard Potassium Permanganate Solution (0.03 N, approximately).Dissolve 1 g. of KMn04 in 1000 ml. of water. Let age for one week or more.
Filter through purified asbestos. Standardize against the National Bureau
of Standards' standard :sample 40c of ~odium oxalate as follows: Transfer 0.1
g. of the standard sodium oxalate to a 400-m!. beaker. Add 100 m!. of diluted
H 2S0 4 (5 : 95) which has been previously boiled for 10 to 15 min. and cooled
to 27 0 3 C. Stir until the oxalate has dissolved.' Add 40 to 42 ml. of MKn04
(0.03 N) at a rate of 25 to 35 ml. per min., stirring slowly. Let stand until
the pink color disappears. (If the pink' color should persist because the
permanganate is too strong, discard, and begin again, adding a few milliliters

1432

CHE;MICAL ANALYSIS OF IRON AND STEEL

less 'of the KMn04 solution.) Heat to 55 to '600 C. and complete the titration
by adding permanganate until a faint pink coldr persists for 30 sec. Add th6'
last 0.5 to 1 ml. dropwise, with particular care. to allow each drop to become
decolorized before the next is introduced. Determine the excess of permanganate required to impart a pink color to1the solution by matching the
color in another beaker containing the sam~ a~ount of the specially treated
diluted H 2S0 4 (5 : 95) at 55 to 60 C.
Standard Sodium Arsenite Solution (0.035 N, approximately).-Dissolve
2 g. of sodium arsenite in water, filter if the solution is not clear, and dilute to
1 liter. The solution is preferably standardized on a standard steel of approximately the same manganese content as the sample. A clear greenish
yellow color which does not change when another drop of arsenite is added is
a satisfactory end point. As an alternate standardization procedure, 20-ml.
portions of the standard solution of KMn04 (0.03 N).may be titrated in a solution, containing 1 g. of iron as ferric nitrate, under the same condition of acidity
and dilution as in the above method. .The theoretical titer of the arseilite
solution cannot be used because manganese is not quantitatively r~duced to
the bivalent stage.
Procedure. Carbon Steel.- Disso~ve 1 g. of the sample in 50 ml. of diluted
HNO a (sp.gr. 1.135) and boil to expel oxides of nitrogen. Remove from the
heat, cool somewhat, add about 0.5 g. of sodium bismuthate, and boil for 2 to
3 min. (With high-carbon steels, about 1 g. of bismuthate, or an amount
sufficient to effect either a pink color or a precipitate of Mn02, should be added.)
Clear the solution of permanganic acid or manganese dioxide present by adding a
few drops of a saturated solution of S02 (freshly prepared) or other suitable
reducing agent free from manganese and chlorides, and boil to expel oxides
of nitrogen and sulfur. Cool to 15 C. or lower, add 0.5 g. of sodium bismuthate (or an amount equal to 26 times the weight _of manganese present)
and agitate for 1 min. Add 50 ml. of HNO a (3% and free from nitrous acid)
and filter through asbestos (or 'through glass frit or alundum crucibles, sufficiently tight to prevent passage of bism~thati particles). Wash with HN0 3
(3%) until the washings run through colorless. The filtrate must be clear and
absolutely free from particles of bismuthate. , Titrate by either method (a)
or (b) as follows:
(a) Ferrous Suljate-Permanganate Titration.-Treat the filtrate with 2 ml. of phosphoric acid (85%) and then from a burette.J!.dd enough of the FeS04
solution to completely discharge (1 to 2 ml. excess) the color of the permanganic
acid. Titrate with the permanganate solution to the appearance of a faint
pink color. In exactly the same manner carry through a blank determination
using the same amounts of acid and bismuthate as was done with the sample.
Finally, add the exact volume of ferrous ammonium sulfate solution which
was employed and titrate with the standard ,KMn04 solution. The difference
between the volumes of KMn04 solution, required in the two titrations represents the manganese in the sample.
(b) Arsenite Titration.-After filtering off the bismuthate, immediately
titrate with the arsenite solution to a clear greenish/yellow color which does
not change upon addition of another drop of._.:,._arsenite.
.
Alloy Steels.-Steels containing nicker;' molybdenum, and less than 1%
of chromium shall be treated like carbon steels, although it is preferable in the

CHEMICAL ANALYSIS OF. IRON AND STEEL

co

1433

ruost accurate work to remove chromium if it is present as an essential constituent of the steel (over 0.50% chromium). This may be done by the bicarbonate separation described for chromium-vanadium steels in the next
paragraph. An alternative procedure for high chromium alloys consists in
dissolving the alloy in HCI-HNOa (1 : 1) (or HClO 4), adding 20 ml. of HCI0 4,
fuming' strongly for 10 min. after the appearance of the red color to oxidize
chromium, diluting, precipitating iron and manganese by NH 40H-(NH 4)2S 20s,
filtering, and. washing. The washed precipitate is then dissolved in 50 ml. of
diluted HNO a (sp.gr. 1.135) and treated as with carbon steels.
Chromium-Vanadium, Stainless, and Similar Steels.-Dissolve 1 g. of the
sample contained in a 300-m!. covered Erlenmeyer flask in 20 m!. of diluted
H2S04 (1 : 9). When action is complete, dilute to 100 m!. with boiling water.
Add from a burette sodium bicarbonate solution (8%) until a per~anent
precipitate is formed (approximately 36 ml.) and then 4 ml. more. For high
chromium (18%) steel use a 6-m!. excess. Cover the flask, boil for 1 min.,
and let the precipitate settle. Filter rapidly, conveniently through a cone and
papet; containing a little paper pulp, and wash the flask and precipitate four or
fiye times with hot water. 9 If the precipitation has ooen properly performed
there will be no more precipitate than can be conveniently handled on an ll-cm.
paper. The filtrate will become cloudy in the funnel stem and in the receiving
vessel, owing to oxidation and hydrolysis. Heat the filtrate to boiling and
oxidize with small portions of HNO a, adding a total of 12 ml. Evaporate to
about 25 ml. Add about 0.5 g. of sodium bismuthate and boil 2 to 3 min.
Clear the solution by adding a suitable reducing agent, boil and complete the
determination as with carbon steels. lO
Cast Iron.-Transfer 1 g. of the sample to a small Erlenmeyer flask, add
25 mL of diluted HNO a (1 : 3), heat, and boil 2 to 3 min. after action ceases.
Cool, filter and wash the flask and paper with 25 ml. of diluted HNO a (1 : 3).
9 The bicarbonate precipitate does not ordinarily contain more than small amounts
(about 0.005%) of manganese. In very accurate work this should be recovered by
transferring the precipitate to a beaker, dissolving it in aqua regia, fuming with sulfuric
acid, and then carefully adding sodium peroxide in excess and boiling. Chromium and
vanadium are thus separated from manganese and iron, which should be .filtered off,
dissolved in 5 ml. of HNO a (1 : 1) and added to the filtrate from the bicarbonate precipitation. A drop of sulfurous acid added to the nitric acid aids in the solution of the
manganese dioxide.
10 Alternately the zinc oxide separation of chromium and manganese may be used in
place of the bicarbonate method. The procedure is as follows: Transfer 1 g. of the sample
to a 400-ml' beaker, add 25 ml. of diluted HNO. (sp.gr. 1.135) and heat gently. If the
alloy is not attacked by HNO a dissolve in 25 ml. of diluted H 2S0 4 (1 : 5) and oxidize
with a few milliliters of HNO a. When the sample is dissolved, dilute to 200 ml., nearly
neutralize with NH 40H and then add 5-ml. portions of a freshly prepared suspension of
ZnO (prepared by thoroughly shaking 50 g. of the finely powdered reagent with 300-m!.
of water) until the iron is precipitated and a slight excess of ZnO is present. Shake
thoroughly after each addition of the precipitant. When sufficient ZnO has been added,
further addition of the reagent causes the brown precipitate to appear lighter in color.
A sufficient excess of ZnO is also indicated by a slightly white and milky supernatant
liquid. Allow the precipitate to settle, filter through a rapid filter paper, and wash well
with cold water. The ZnO precipitate will retain small amounts of manganese (about
0.005%). A.dd 12 m!. of HNO" and evaporate to about 30 ml. Add about 0.5 g. of
sodium b'.~muthate and boil 2 to 3 min. Clear the solution by adding a suitable reducing
agent, ~,oil, and complete the determination as with carbon steels.

1434

CHEMICAL ANALYSIS OF IR,ON AND STEEL

Wash the paper a few more times withhot water, and then complete the determination as with carbon steels, including the preliminary oxidation and boiling
with bismuthate.
,
OPen-Hearth Iron and Wrought Iron.-Proc'eed as with carbon steels using
2 to 3 g. of the sample. It is advisable, befor~ filtering the analysis sample,
to treat the asbestos pad with a weak solution ofi KMn04 and wash it free from
the latter with diluted HN0 3 (3%). Very low' manganese irons shall not be
filtered through pads used previously for higher manganese alloys. Colorimetric determinations by the periodate method on 0.5- to I-g. samples of openhearth iron are applicable for routine work.
MANGANESE BY THE PERSULFATE METHOD
(PRESENCE OF COBALT)

Special Solutions Required. Acid Mixture.-Add 100 ml. of H 2S0 4 to

525 ml. of water slowly and with stirring, cool, and add 125 ml. of H aP0 4
(85%) and 250 m!. of HNO a
Silver Nitrate (O.8%).-Dissolve 8 g. of AgNO s in 1000 ml. of water.
Ammonium Persulfate (25%).-Dissolve 25 g. of (NH4)2S208 in 85 m!: of
water. The solution must not be kept longer than 12 hr. If the salt is of less
than 95% strength, the necessary equivalent must be used.
Standard Sodium Arsenite.-f3ee Bismuthate Method. To standardize,
dissolve 1 g. of electrolytic or open-hearth iron of a low and known manganese
content (as National Bureau of Standards' standard sample 55a of ingot iron)
in 30 m!. of the acid mixture. Dilute to 100 ml. with warm water, add exactly
20 ml. of a standard solution of KMn04 (0.03 N), and then add just enough
H 2SO a to reduce the KMn04. Boil for a few minutes to remove oxides of
nitrogen. Add 10 m!. of the AgNO a solution and 10 m!. of the (NH 4)2S20S
solution, boil, cool, and titrate as in the procedure to be described. To obtain
the manganese titer, divide the amount of manganese present in the added
KMn04 and open-hearth iron by the mil1iliter~ of arsenite required.
Procedure. Carbon Steel: 1. Visual Tiiraiion.-Transfer 1 g. of the
sample, for a steel containing 1% or less t of manganese, or 0.5 g. of the sample,
for a steel of higher manganese content, to a 500-m!. Erlenmeyer flask, and add
30 ml. of the acid mixture. Heat until solution is complete, and boil until
oxides of nitrogen have been expelled. I Add 10D ml. of hot water, 10,ml. of
AgNO a solution and 10 m!. of (NH 4)2S208 solution> Hej1t to boiling and boil
briskly for 1 to 1.5 min. Cool to 30 C. or lower, add 75- m!. of water, and
titrate rapidly with standard arsenite solution to a clear yellow end point which
does not change upon the addition of more arsenite. If the solution is not
titrated rapidly with arsenite, part of the manganese may be reoxidized by the
persulfate during the titration and thus yielp high results.
2. Potentiometric Titration.-Proceed to the titration stage as in 1, cooling
the solution to 15 C. To titrate, bring the needle (or light-beam) on the
the solution to 15 C. To titrate, bring the n!ledle (or light-beam) on the scale,
and then add the standard arsenite solution. As the end point is approached
(rapid fading of the pink color) the needle will_remairl stationary, or move very
slowly. Continue the addition of the arfle~ite dropwise until a shar:!> break
of four to five divisions is obtained. The potentiometric titration 0': per-

CHEMICAL ANALYSIS OF IRON AND STEEL

1435

Jl].anganic acid with arsenite is, like the visual, empirical. It is, therefore,
necessary to standardize the arsenite solution against a like weight of a standard
steel which has approximately the same manganese content as the sample
being analyzed and has been treated exactly as the latter.
While the potentiometric titration offers no particular advantage over the
visual method for steels containing less than 1% of manganese, it is claimed
that more reproducible end points can be obtained potentiometrically with steels
of high manganese content or in turbid solutions when graphite has not been
separated. The apparatus consists essentially of three parts, a motor stirrer,
electrodes, and a potentiometric unit. In the Larrabee apparatus the electrodes consist of gold and platinum rods with small vanes at the bottom. In'
the Kelley apparatus platinum and calomel electrodes are used; it is desirable
to insert a salt bridge (N a 2S04) solution between the calomel electrode and
the titrating solution in this apparatus.
Alloy Steels: Nickel, Chromium-Nickel, Chromium-Vanadium, and
Chromium-Vanadium-Molybdenum Steels.-Proceed as with carbon steels.
C}1romium Steels (over 4% chromium).-Separate the greater part of the
chromium by the bicarbonate separation,ll using a 6 ml. excess for high chromium (18%) steels. To the filtrate add 30 ml. of the arsenite acid mixture,
evaporate to 125 ml. and complete the 'determination as directed for plain
carbon steels. With high carbon (0.8%) high chromium steel, test the bicarbonate precipitate for manganese.n
Chromium-Tungsten Steel.-Treat 0.5-g. of the sample with 50 ml. of
diluted H 2S0 4 (1 : 9) and 3 m!. of H 3 P0 4 Heat until all action ceases. Add
40 mt of water and 5 m!. of HN0 3 , boil until the carbides have dissolved, and
proceed as with carbon steels. With chromium-tungsten steels the arsenite
solution is preferably standardized on a steel of similar type of known manganese
content. Tungsten steels with high carbon (0.8%) and high chromium (4.0%)
~ontents may not be completely decomposed by the H 2S0 4-H aP0 4 attack. In
this case treat the sample with 10 ml. of a mixture of equal parts of HCI and
HNO a and 4 to 5 drops of HF. When action ceases add 5 m!. of HCI0 4
(60%) and evaporate just to dryness. Cool, add 5 mI. of H 2S0 4 and fume
until HCI0 4 is expelled. Cool, add 50 mI. of water, 5 m!. of H 3 P0 4 (sp.gr;
1.73), dilute to 100 mI. and heat until salts dissolve. If Mn02 appears, dissolve
it with a few drops of H 2S0 3 Complete the determination as with carbon
steels.
,
Cast lron.-Dissolve as directed for carbon steels, dilute, filter through
a rapid paper, wash with hot water, dilute to 125 mI., and complete as with
carbon steels.
Open-Hearth lron.-Proceed ,as with carbon steels, but use a 2- to 3-g.
sample.
Wrought lron.-Treat 1 g. of the sample as directed for carbon steels.
PHOSPHORUS BY THE MOLYBDATE-MAGNESIA METHOD
Special Solutions Required. Diluted Nitric Acid (sp.gr. 1.135).--Mix 280
ml. of HN0 3 and 720 m!. of water. .
11

See determination of Manganese in Chromium ,Steels by the Bismuthate Method,

p.1433.

1436

CHEMICAL ANALYSIS OF IRON AND STEEL

Diluted Nitric Acid (2 : 100).-Mix 20 ml. of' HNO a and 1000 ml. of water.
Potassium Permanganate (2.5%).-Dissolve 25 g. of KMn04 in 1000 ml.
of water.
Ammonium Molybdate.-Place in an 800-m!. besker 65 g. of c.p. ammonium
molybdate crystals, (NH4)6Mo70244H20, 225 g. of\NH 4NO a, 15 ml. of NH 40H
(sp.gr. 0.90), and 600 ml. of water. Stir and heat igently. When the crystals
have dissolved, filter (without washing) and dilute to 1000 ml. with water.
Magnesia Mixture.-Dissolve 65 g. of MgS04 and 75 g: of (NH 4)2S04 in
500 ml. of water. Add NH 40H in slight excess and let stand overnight; filter
if a precipitate appears. Make just acid by adding a very sligl}t excess of
H 2S0 4 , dilute to 1 liter, and keep in a glass-stoppered bottle.
Diluteil Ammonium Hydroxide (1 : 20).-Mix 100 ml. of NH 40H (sp.gr.
0.90) and 2000 m!. of water.
Potassium Nitrate (1%).-Dissolve 10 g. of KNO a in 1000 m!. of water.
Procedure. Carbon Steels.-Transfer 4 g. of the sample 12 to a,300-ml.
Erlenmeyer flask, dissolve the sample in 125 ml. of diluted HNOa (sp.gr. 1.135),
oxidize the organic matter by adding 10 ml. of the solution of KMn04 (2.5%)
and boil for 3 to 5 min. If no precipitat,e forms, add more KMn04 and boil
again. Dissolve the precipitate by the dropwise addition of a saturated solution of Na 2SOa or H 2SO a, and boil for a few minutes to expel oxides of nitrogen.
Cool to about 75 C. and add 85 mr. of molybdate reagent. Stopper the flask,
shake for 10 min., and allow to stand for 2 hr. or more. Filter through a tight
filter paper. Wash the flask, precipitate, and paper with the cold diluted
HNb a (2 : 100) and then five times with KNO a solution.
Set the iltrate and washings aside after thorough mixing, and note whether
further separation of phosphomolybdate occ~rs. Dissolve the precipitate on
the filter and in the flask in 20 ml. of diluted NH 40H (1 : 1) to which has been
added 2 g. of citric acid, catch the solution in a 250-ml. beaker, wash the filter
several times with diluted NH 40H (1 : 20), then with hot water, and finally
several times with diluted HCI (1 : 20). If the ammoniacal solution 9f the
phosphomolybdate is not clear at this point, heat it to boiling, filter through
the same paper, wash the paper with hot watef, ignite in a small platinum
crucible, and fuse any residue with as little Na 2 CO a as possible (not over 0.5 g.);
extract the cooled melt with hot water, cooi, filter, and add the water solution
to the ammoniacal solution.
Render the ammoniacal solution acid \vith HCI, add 20 ml. of the acid
magnesia: mixture, cdol in ice water, and then add N'H 40B;_slowly until the
solution is faintly ammoniacal. Stir vigorously for 5 min. or until a crystalline
precipitate appears,1a and finally add 5 to 10 ml. of NH 40H. The volume of
the solution at this point should not exceE;ld 100 ml. Allow the ,solution to
stand in a cool place for 4 to 6 hr., or, preferably, overnight. Filter and wash
the precipitate moderately with diluted NH 4qH (1 : 20). D_issolve the precipitate on the filter in 20 ml. of diluted HCI (1 : 1), catching the solution in
the original beaker. Wash the filter thoroughly with warm diluted HCI
12 If the percentage of phosphorus is less than 0,05%, it is advantageous to treat two
or more 4-g. samples as described, and combine the yellow precipitates either by filtering
through the same paper or by' combining the precipitates after filtering on separate papers.
13 If'the amount of phosphorus is very small,jp"may take 30 min. to 1 hr. before the
precipitate begins to appear.

CHEMICAL ANALYSIS OF IRON AND STEEL

)437

(1 : 1). Add 0.5 to 1 g. of NH4Br14 and gently boil the HCI solution to It
volume of 5 to 10 ml. (but not to dryness) to eliminate arsenic. Dilute to a
volume of 50 to 75 mI., add 0.1 to 0.2 g. of citric acid, 2 to 3 mi. of the acid
Illagnesia mixture, and cool in ice water. Make ammoniacal, stir as above,
and allow to stand in a cool place fpr 4 to 24 hr.
Filter, wash with diluted NH 40H (1 : 20), and ignite the precipitate carefully and at as low a temperature as possible until the carbon has been destroyed
and the residue is white. Finally, ignite to constant weight at 1000 to 1100 C.
Dissolve the ignited precipitate in 5 mi. of diluted HNO a (1 : 1) and 20 ml. of
water. If no residue remains, the ignited precipitate may be regarded as
Mg2P207. If a residu~ remains, filter, wash with hot water, ignite, and weigh.
Treat with a few drops of HF, evaporate to dryness, ignite, and reweigh.
Subtract any loss of weight from the original weight of the pyrophosphate, and
calculate the percentage of phosphorus on the basis of this purified magnesium
pyrophosphate.

Alloy.Steels: Nickel, Chromium-Nickel, Stainless and Similar Alloy Steels

Contai[ling Neither Tungsten nor Vanadium.-Proceed as with carbon steels.
High-chromium steels (stainless), high-chromium high-nickel (18% chromium,

8% nickel; 20% chromium, 20% nickel), and other steels in this group that do
not dissolve in diluted HNO a (sp.gr. 1.135), shall be treated as follows: Treat
3 g. of the sample in a 500-ml. Erlenmeyer flask with 75 ml. of a mixture Qf
equal parts of HCI and HNO a, and heat gently. When decomposition is
complete, add 25 ml. of HCI0 4 (60%) and evaporate until fumes of HCI0 4 are
evolved. Continue the heating for 5 min. to oxidize chromium 'and to dehydrate silica. Cool somewhat, add 40 ml. of water, filter, and wash the flask,
paper, and silica with 55 ml. of diluted HNO a (sp.gr. 1.2). Add KMn04 to
the filtrate, l;lOil, add H 2SO a to destroy oxides of manganese and to reduce all
of the chromium, and proceed as usual. As an alternative method, the sample
m~y be decomposed by heating with diluted HNO a (1 : 1) and adding small

portions of either HF or HCI.

Austenitic Manganese Steels (over 10% manganese).-Transfer 4 g. of the
sample to a 300-ml. Erlenmeyer flask, add 70 ml. of diluted HNO a (sp.gr. 1.135)
and digest until action ceases. Add 35 ml. of HCI0 4 (60%) and evaporate
just to fumes. Add HF dropwise until all of the silica is in solution and then
an excess of 5 drops. Fume so that the HCI0 4 refluxes on the sides of the flask
for 25 to 30 min. Cool, add 50 mL of water and 10 mi. of HNO a Add a few
drops of KMn04 (2%) and boil until oxides of manganese are precipitated.
Dissolve the precipitated manganese with H 2SO a and boil a few minutes to
expel oxides of nitrogen. Cool to 70 C. and precipitate with molybdate, etc.,
as with carbon steels.

Chromium- Vanadium Steels or Other Stee~s Containing Vanadium but No

Tungsten.-Proceed as with carbon steels until the solution is ready for the
addition of the molybdate reagent. At this point 'cool to 10 C., add 5 ml. of
a solution of ferrous sulfate (100 g. of .FeSO.j 7H 20, in 1000 m!. of diluted
H 2S0 4 (5 : 95 and 2 to 3 drops of H 2SO a Mix, add 85 ml. of molybdate,
shake for 10 min., and allow to stand for 4 hrs. or, preferably, overnight. Filter,
and finish the determination as directed for carbon steel~.
14 With a precipitation temperature of 25 C., arsenic, when present in small amounts.
is not precipitated, and therefore the HBr treatment can be omitted.

1438

CHEMICAL ANALYSIS OF IRON AND STEEL

High-Speed Steels or Other Steels Containing Tungsten and Vanadium.Decompose 4 g. of the sample in 125 ml. of dihtted .RN0 3 (sp.gr. 1.135), add
30 ml. of HCI and evaporate to dryness. Take up the residue in 20 ml. of
diluted HCI (1 : 1), dilute to 100 mL with hot w:j,ter, and filter off the tungstic
acid.l 5 Evaporate the filtrate twice with.20~ml. portions of HN0 3 to expel
the HCI, taking the second evaporation just to a, syrup; add 65 ml. of diluted
HN0 3 (sp.gr. 1.135), and filter the solution if it is not entirely clear. Cool
to 10 C., add 5 ml. of a solution of FeS04 (100 g. in 1000 ml. of diluted H 2S0 4
(5 : 95 and 2 to 3 drops of H 2S0 3 Mix, add 85 ml. of molybdate reagent,
shake for 10 min., allow to stand for 4 hrs., or preferably, overnight. Filter
and finish the determination as directed for carbon steels.
Cast Iron.-Dissolve 2 to 3 g. of the sample in a covered casserole in 30
ml. of diluted HN0 3 (1 : 1). When solution is complete, add 10 ml. of diluted
HCI (1 ': 1), evaporate to dryness, and bake for 15 to 20 min. on the hot plate.
Cool, drench the residue with HCI, dilute to 50 ml. with hot water, and warm
until the salts are in solution. Filter without delay, wash with diluted HCI
(1 : 20), and evaporate the filtrate to syrupy consistency. In the meantime,
place the paper containing the graphite lj.nd silica in a platinum crucible, burn
the carbon in a good oxidizing atmosphere, cool, and add 5 to 10 drops of HCl
and 1 to 2 ml. of HF. Evaporate just to dryness, take up any residue 16 in 5 ml.
of HCI, and add the solution to the main solution which is being evaporated.
When the solution is of syrupy consistency, transfer it to a 300-ml. Erlenmeyer
flask by alternate washing with diluted HNO a (1 : 1) and hot water, using
not more than 30 ml. of the former and 70 ml. of the latter. Nearly neutralize
with NH,OH, and add 1.00 ml. of molYbdate reagent. Shake for 10 min. lj.nd
allow to stand for 4 to 6 hrs., or preferably, overnight. Complete the determination as with carbon steels.
Open-Hearth Iron and Wrought Iron.-Proceed as with carbon steels.
PHOSPHORUS BY THE ALKALI METRIC METHOD
I
Special Solutions Required.-In addition to' those listed urider the molybdate-magnesia method, the following special solutions will be required:
Standard Sodium Hydroxide.-Transfer 6.5 g. of C.P. NaOH to a I-liter
flask. Add 1000 ml. of water, and a sli9ht excess of a solution of Ba(OH)2.
15 This precipitate may contain a small amount of phos~orus, and in standardization
work it shall be treated as follows: Transfer the bulk of the precipitate to an Erlenmeyer
flask with a jet of water, and then treat the paper with 25 mi ..of-hot diluted NH,OH
(1 : 4) containing 0.5 g. of citric acid. Catch the filtrate in the flask containing the
remainder of the precipitate. Wash the paper with hot water, then a few times with
hot diluted HCl (1 : 20), holding the volume of the filtrate and washings to about 75 mi.
Slightly acidify the solution with HCI, add 25 ml. of the acid magnesia mixture and about
10 mi. of NH,OH in excess. Add a few glass beads, cool the solution in ice water, stopper
the flask, and shake thoroughly. for 1 hr. AlloW:: to stand overnight at about 0 C.
Filter, and wash a few times with diluted NH,OH (1 : 20). Dissolve the precipitate in
a little hot diluted HNO" and add it to the original solution in which phosphorus is to
be determined, or determine the phosphorus separately by precipitating it with molybdate reagent and titrating with standard alkali and acid. This recovery usually amounts
to less than 0.001 % and need not be made in routine analys!)s.
16 If the percentage of titanium or zirconium is high, this residue should be fused
with Na.CO" extJlacted with hot water, cooled;-fittered, and the water extract added
to the main solution.

CHEMICAL ANALYSIS OF IRON AND STEEL

1439

Stopper and let stand for 24 hrs. Decant the liquid, and standardize it against
a steel of known phosphorus content, as determined by the molybdate-magnesia
method. Appropriate National Bureau of Standards' standard steels are
recommended for this purpose. The solution may also be standardized by use
of the National Bureau of Standards' standard sample 84 of acid potassium
phthalate and the ratio 23 NaOH to 1 phosphorus. Protect the alkali solution
from carbon dioxide by means of a soda-lime or soda-asbestos tube.
Standard Nitric Acid.-Mix 7 m!. of clear HNO a with 1 liter of water and
standardize against the standard alkaF solution, using phenolphthalein as the
indicator. If desired, the acid solution may be rendered equivalent to the
NaOH by dilution with water.
Phenolphthalein Indicator.-Dissolve 0.29 g. of reagent in 100 m!. of ethyl
alcohol (50%).
.
Procedure. Carbon Steels.-Transfer 2 to 3 g. of the sample to a 300-m!.
Erlenmeyer flask, dissolve 17 the sample in 65 m!. of diluted HNO a (sp.gr. 1.135)
and oxidize organic matter by adding 10 m!. of the KMn04 (2.5%) and boiling
for ~ to 3 min. If no precipitate forms, add more 'KMn04 and boil again.
Dissolve the precipitate by the dropwise addition of a saturated solution of
Na2SOa or H 2SO a and boil for a few minutes to expel oxides of nitrogen. Adjust
the volume to 60 m!. and the temperature to 45 C., and add 50 m!. of the
molybdate reagent. Stopper the flask, shake for 10 min., and allow the precipitate to settle. Filter through a 9-cm. close-texture paper. Wash the
flask, precipitate, and paper twice with 5-m!. portions of diluted HNO a (2 : 100)
and then five times with 5-m!. portions of a solution of KNO a (1 %). Finally,
wash the paper about ten times (until free from acid), directing the jet.of KNO a
solution around the edge of the 'paper and then spirally down. Return the
paper and precipitate to the flask, add 25 m!. of water and a 2- to 5-ml: excess
of standard NaOH, both free from carbon dioxide, and shake or stir until the
'precipitate is dissolved. Dilute to about 150 ml. with water free from carbon
dioxide, add 3 drops of the phenolphthalein indicator, and discharge the pink
color with standard HN Oa.
Alloy Steels: Nickel, Chromium-Nickel, Stainless, and Similar Alloy Steels
Containing Neither Tungsten nor Vanadium.-Proceed as with carbon steels.
With high-chromium nickel, or other steels insoluble in diluted HNO a, treat
2 g. of the sample with 50 m!. of a mixture of equal parts of HNO a and HCI,
add 4 to 5 drops of HF, heat till action ceases and then add 20 m!. of HCI0 4.
Evaporate to fumes, fume 5 to 10 min., filter, etc. (as in the molybdatemagnesia method)/8 and finish by the alkalimetric method. With steels of
appreciable titanium content (over 0.25% titanium) it is best to remove titanium before precipitating the phosphomolybdate.
Austenitic Manganese Steel (greater than 10% manganese).-Transfer
2 g. of the sample to a 400-m!. beaker, add 35 m!. pf diluted HNO a (sp.gr. 1.135)
and digest until action ceases. Add 20 m!. of HCI0 4 (60%) and evaporate
just to fumes. Add HF dropwise until all of the" silica is dissolved and then
add an excess of 5 drops. Fume so that HCI0 4 refluxes on the sides of the
beaker for 20 to 25 min. Cool, add 50 m!. of water and 10 m!. of HNO a. Add
17

3 to

To prevent the possibility of silica precipitating in steel of high silicon content,

4 drops of HF may be added while the sample is dissolving.

18

See p. 1437.

1440

CHEMICAL ANALYSIS OF IRON AND STEEL

a few drops of KMn04 (2%) and boil until oxides of;manganese are precipitated.
Dissolve the precipitated manganese with H 2SO a and boil for a few minutes to
expel oxides of nitrogen. Transfer the solution to a 300-ml. Erlenmeyer flask
and adjust the volume to 60 ml. and the temperature to 45 C.19 Add 50 ml.
of ammoniacal molybdate solution, shake 10 min} and let stand for 20 min.
Filter, wash, and titrate as directed for carbon steels.
Chromium- Vanqdium Steel, or Other Steels Containing Vanadium but No
Tungsten.-Proceed as with carbon steels until the solution is ready for' the
addition of molybdate reagent. At this point cool to 10 C., add 5 m!. of a
solution of ferrous sulfate (100 g. FeS04' 7H 20 in 1000 ml. of diluted H 2S0 4
(5 : 95)) and 2 to 3 drops of H 2SO a. Mix, and then add 50 ml. of molybdate
reagent. Shake for 10 min., allow to settle 1 hr., filter, and finish as directed
for carbon steel.
High-SPeed Steels, or Other Steels Containing Tungsten and Vanadium.-'
For the decomposition of the sample (2 g.) and the preparation of the solution
for the precipitation of the ammonium phosphomolybdate proceed as in the
gravimetric method for the determination of phosphorus in these steels. 20
Precipitate the phosphorus in a cool reduced solution and finish the determination as described for chromium-vanadium steels by the alkalimetric method
(see preceding paragraph).
.
Cast fron.-Dissolve 0.5 to 2 g. of the sample in 65 ml. of diluted HNO a
, (sp.gr. 1.135). Filter through a paper of loose texture and catch the filtrate in
a 300-ml. Erlenmeyer flask. Wash tne paper a few times with diluted HNO a
(2 : 100) and then with about 50 ml. of hot water.21 Add 10 ml. of KMn04
. to the filtrate and boil for 3 to 5 min. Dissolve the precipitated oxides by the
dropwise' addition of a saturated solution of Na 2SOa or H 2SO a and boil for a
few minutes to expel oxides of nitrogen. Cool to 45 C., add 50 m!. of molybdate reagent and proceed as with carbon steels.
OPen-Hearth fron.-Proceed as with carbon steels. The result shall be
corrected by that obtained in a blank run.
Wrought fron.-Proceed as with carbon steelsf using 1 g. of the sample.
In wrought iron phosphorus will be present as phosphide and as phosphate
(in slag inclusions). If a differentiation b~tween the two forms is desired,
determine total phosphorus in one sample; 'determine the phosphate that is
left after phosphine has been driven off (by' attack with a non-oxidizing acid
as HCn from another sample, and the phosphorus occurring as phosphide is
then obtained by difference. The accuracy of the results obtained by this
phosphate-phosphide method is somewhat doubtful, but the 'method is ap-'
parently the only one available at present.
SULFUR BY THE GRAVIMETRIC METHOD
Special Solutions Required. Barium Chloride Solution (10%).-Dissolve
100 g. of BaCI 2 2H 2 0 in 1000 ml. of water.
19 If the solution at thin point contains silica, filter, wash with water, boil until the
volume of the filtrate is approximately 60 mi., and again cool to 45 C.
20

See p. 1438.

21 For treatment of the residue when the iron contains appreciable amounts of titanium or zirconium see the Determination of Phospilorous in Cast Iron by the Molybdate-Magnesia Method, p. 1438.

CHEMICAL ANALYSIS OF IRON AND STEEL

1441

Diluted Hydrochloric Acid (1 : 99).-Mix 10 ml. of HCl (sp.gr. 1.19) and

990 ml. of water.
Procedure. Carbon Sleel.-Dissolve 4.57 g. of the sample in 50 ml. of
HN0 3 22 in a covered beaker or flask. In case solution is slow or difficult,
HCI may be added dropwise at intervals. With steels that dissolve too rapidly,
it is necessary to place the cooled acid in a covered beaker and add the sample
in small portions. When solution is complete, add 0.5 g. of N a 2C0 3 and carefully evaporate to low volume (about 10 ml.)23 in a sulfur-free atmosphere.
Cool, add 30 ml. of HCI, and evaporate just to dryness. Add 30 ml. more of
HCI, and evaporate to syrupy consistency.24 Add 10 ml. of HCI, 25 ml. of
water, 5 g. of 20- to 30-mesh zinc (free from sulfur), and warm on the steam'
bath until the iron is reduced to the ferrous state and the evolution of hydrogen
has nearly ceased. Remove the silica and undissolved zinc by filtration through
a tight filter paper, wash with approximately 75 ml. of diluted HCI (1% by
volume), warm to 60 to 70 C. and add 10 ml. of the solution of BaCh (10%).25
Stir vigorously and let stand 18 to 24 hrs. Filter on a 9-cm. paper of close
texture and discard the filtrate. Wash once or twice with cold diluted HCl
(1 : 50_0) and then with hot water until free from chlorides. Reserve the
precipitate and evaporate the washings to dryness. Dissolve the slight residue
in 1 ml. of diluted HCI (1 : 1), dilute to 25 ml. with hot water, add 1 ml. of
BaCh (10%), and digest at 70 to 80 C. for several hours, avoiding any undue
loss by evaporation. Filter on a small paper of close texture and wash with
,hot water until free from chlorides.26 Ignite both papers in platinum and treat
with one drop of H 2S0 4 (1 : 1) and 1 mI. of HF. Evaporate to dryness, ignite,
and weigh. Duplicate blanks shall be carried through all steps of the determination. The corrected weight of the ignited BaS04 multiplied by three is
equal to the percentage of sulfur.
Alloy Steels.-For steels which can be dissolved in HN0 3, or HN0 3 plus
snmll portions of HCI, proceed as with carbon steels. High chromium-nickel
steels shall be dissolved in a 400-ml. tall-form beaker in 150 ml. of a mixture of
equal parts of HN0 3 and HCI. The acids shall first be thoroughly mixed in
another container and, when thoroughly red, added to the beaker containing
the sample. If the reaction proceeds too rapidly the beaker and contents
should be chilled by placing them in ice water.
Alternately, the following Meineke method (solution in acid copper potassium chloride) may be used for the chromium-nickel alloys.
Meineke Method.-Dissolve the sample (4.57 g.) in 250 ml. of copper
potassium chloride solution (dissolve 300 g. of 2 KCI CuCI 2 2H 20 in a mixture
22 The addition of 2 to 5 m!. of bromine is claimed to effect more complete recovery of
sulfur.
23 The use of a coarse screen of Ys in. wire, triangles, or an asbestos pad on the hot
plate permits of more rapid evaporation without the danger of spattering.
24 The HN0 3 can also be expelled by adding about' 50 ml. of HCIO" evaporating to
fumes of HCIO., and fuming for about 5 min. The solution is then cooled, diluted to
75 ml., reduced with zinc, filtered, BaCb added, and the procedure continued as described.
25 The solution should preferably contain not more than 2% by volume of HCI at the
time of the precipitation with BaCI 2 Ordinarily, there will be no hydrolysis of iron
during the filtration and washing of the undissolved zinc or the precipitation with BaCI 2
Should this occur, the solution mu:Jt be cleared by the addition of HCI, having due regard
to the final permissible acidity.
26 This recovery of BaSO. ordinarily represents from 0.001 to 0.003% sulfur.

1442

CHEMICAL ANALYSIS OF IRO,N AND STEEL

of 1000 mL of water and 75 mL of HCl) at Ioom temperature and with mechanical stirring (slow).27 Filter through purified asbestos. Wash two or three
times with diluted HCl (2 : 98). Return the residue and asbestos pad to the
beaker and treat with 20 mL of HNO a. Heat an~ add KClO a until all carbonaceous matter is destroyed. Graphite may be ignored. Evaporate the
solution to dryness, take up in 10 mL of HCl and a:gain evaporate to dryness.
Take up in 10 mL of diluted HCl (1 : 9) and 20 rill. of water, filter through
paper and wash with 50 ml. of hot water. Boil and add 5 ml. of BaC1 2 (10%).
Digest for about 2 hrs., filter; and finish the determination as directed in the
HNO a method (see preceding paragraph).
Tungsten Steels.-Transfer 4.57 g. of the sample to a 600-ml. beaker or
Erlenmeyer flask and treat with 75 mL .of HNO a If reaction is slow, heat
gently; if it is vigorous, cool in ice water; If reaction is very slow, add 5 ml.
of HCl or a drop or two of HF occasionally as the solution is heated on the
steam bath. 2s When HNO a alone is used, the particles of steel tend to become
covered with a film of tungstic acid. Stirring or rubbing these coated granules
speeds up the decomposition. When reaction is complete, digest until the
residue is pure yellow and contains no dark material, adding fresh portions of
acid if necessary. Evaporate to syrupy consistency. Cool, add 30 mL of
HCl, and again evaporate to syrupy consistency. Add 100 mL of hot water
and boil until soluble salts are in solution. Filter and wash the residue with
diluted HCI (1 : 10), catching the filtrate and washings. in a 600-mr. casserole.
Evaporate to dryness, add 30 mL of'HCl, and again evaporate to dryness.
Cool, add 60 mL of diluted HCl (1 : 1), warm until salts are in solution, and
add 50 mL of boiling water. Add 10 m!. of cinchonine solution (125 g. of
cinchonine dissolved in 1000 mL of diluted HCl (1 : 1)), and let stand overnight. Filter and wash with cinchonine wash water (30 m!. of cinchonine
solution diluted to 1000 mI.). Evaporate the filtrate and washings until a
slight .film begins to form and proceed with the reduction by zinc as with carbon
steels. In this determination it is particularly important that corrections be
made as determined by blanks carried through all' steps of the method. As
cinchonine sometimes contains sulfur, this solutiodshould be measured so that
the same amount is used in the blank as in the determination. If the cinchonine
contains excessive quantities of sulfates, wash the crystals on a Buchner funnel
with water until the washings no longer give a precipitate with barium chloride.
OPen-Hearth Iron.-Proceed as with carbon steels.
Cast Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon steels.
SULFUR BY THE EVOLUTION METHOD
Apparatus.-Figure 185 shows a typical apparatus for this determination.
Special Solutions Required. Diluted Hydrochloric Acid (1 :' l).-Mix 500
ml. of HCI and 500 m!. of water.
Ammoniacal Zinc Sulfate.-Dissolve 200 g. of ZnS047H 20 in 1000 mL
of water, and add 1000 m!. of NH 40H (sp.gr. 0.90). Let stand 24 hrs. and filter.
27 Selenium is also obtained with sulfur in the residl.% hence *ith selenium s~,eels, all
operations should be conducted under a good hood. -28 Alternately, the sample may be treated with 150'mi. of a mixture of Hel and HNO a
as in alloy steels.

CHEMICAL ANALYSIS OF IRON AND STEEL

1443

Ammoniacal Cadmium Chloride.-Dissolve 10 g. of GdCb in 400 ml. of

water and add 600 ml. of NH 40H (sp.gr. 0.90) ..
Starch Solution.-To 500 ml. of'boiling'water, add a cold suspension of 5 g.
of soluble (or arrowroot) starch in 25 m!. of water, cool, add a cool solution of
5 g. of NaOH in 50 ml. of water; then add 15 g.
.
of K1 and mix thoroughly.
Potassium Iodate.-Dissolve 1.12 g. of KIO a
and 12 g. of KI in 1000 ml. of water. For gen'eral work the theoretical sulfur titer of this
solution shall be used; for specialized work on
one kind of material, the solution may be standardized against 1ike material. The theoretical
titer is based on standard sodium ox'alate and is
obtained as follows: To 300 ml. of water in a
600-m!. flask, preferably glass-stoppered, add 10
ml. of HCl (sp.gr. 1.19) and 1 g. of KI. Cool
and add 25 ml. of KMn04 (0.03 N) which has
been standardized against sodium oxalate. Swirl
gently, stopper, and let stand for 5 min. Titrate
the liberated iodine with thiosulfate solution
. (approximately 0.03 N) until the color nearly
fades. Then add 5 m1. of starch solution and
Electrfc:Healer
continue the titration until the blue color is just
destroyed. In a new flask, repeat the experiment with the sole difference that 25 m!. of the
iodate solution is substituted for the standard
permanganate solution. Calculate the normality of the iodate solution, and dilute it if deFIG. 185.-Apparatus for
sired so that its normality is 0.03J2, or 1 m!. is Determination of Sulfur by the
equivalent to 0.01% of sulfur on a 5-g. sample. Evolution Method.
For standardization against like materials, 5-g.
portions of the proper standard are carried through all steps of the method.
The following equations represent the reactio~s that are involved:
2KMn04+ lOK1+ 16HCI--)-51 2+ 12KCl+2MnCI 2+8H 20
KIO a+5KI+6HCI--)-31 2+6KCI+3H 20
12 + 2Na2S20a--)-Na2S406+ 2NaI
The normality of the iodate solution is then found by dividing the volume of
N a 2S';Oa solution required in the second titration by the volume required in
the first, and multiplying by the normality factor of the solution of KMn04.
If the solution is then diluted so that the normality is 0.0312, 1 m!. represents
0.01 % of sulfur on a 5-g. sample.
Procedure. Carbo'f!. Steel.-Place 5 g. of the sample in the dry flask and
connect the latter as shown in Fig. 185. Place 10 ml. of the ammoniacal solution
of zinc sulfate or cadmium chloride and 150 ml. of watBr in the tumbler. Add
80 ~l. of the diluted HCI (1 : 1) to the flask through the thistle tube, heat the
flask and its contents gently until the solution of the iron is complete, then boil
the solution for 30 sec. The heat shall be so adjusted that there is a rapid,
steady evolution of gas. Disconnect the delivery tube, leaving it in the solu-

1444

CHEMICAL ANALYSIS OF IRON AND STEEL

tion, and remove the tumbler. Add 2 m!. .Qf starch solution, then 40 m!. of
diluted HCI (1 : 1) and titrate immediately with th iodate solution to a
permanent blue color.
NOTEs.-Direct sunlight must be avoided if cadmium sol~tions are used, otherwise
low results are obtained.
.
This method succeeds best when the evolution of gas is rapid and the conditions are
kept constant.
Trouble may occur if the absorbing solution becomes too hot. This condition can
be avoided by keeping the absorbing solution in a water bath.

Some analysts prefer to add the diluted HCI at a temperature of 80 to 90 C. and to

heat the flask on an electric heater instead of over gas. The use of concentrated HCI
necessitates a condensing arrangement and does not offer any general advantages over
the diluted acid for carbon steel.
With most carbon steels, the results for sulfur which are obtained by the
evolution method and the use of the theoretical sulfur titer check the gravimetric results within 0.002%. Some steels dissolve too slowly and some
(for example, certain steels containing high sulfur or high carbon) do not yield
all of their sulfur. In such cases the samples must be annealed as follows:
Place 5 g. of the sample in a 20-ml. porcelain crucible. Cover with a ;!.i-in.
layer of graphite, Acheson No. 38, cover the crucible and heat at 6850 C. for
20 min. Transfer the cooled annealed steel and graphite to the evolution flask
and proceed as in the regular method.
Alloy Steels.-Many alloy steels yield all their sulfur when analyzed by
the same procedure as is used for carbon steels, but some yield only a p!l.rt.
Hydrochloric acid of 2 : 1 strength, or the concentrated acid (sp.gr. 1.19) with
some alloy steels, will give slightly more evolved sulfur than does the 1 : 1 acid.
For example, some molybdenum steels yield most of their sulfur with concentrated HCI. For high silicon steels, concentrated HCI plus a small amount
of HF (0.5 m!.) has been recommended. The evolution method, however, can
be used only if experiment has shown that all sulfur is evolved as H 2S in the
particular type of steel under analysis. The evolution method is not satis~
factory for the determination of sulfur in selenium steels.
Cast Iron.-Most cast irons do not give up all their sulfur as hydrogen sulfide. The amount that is give:6. off can often be increased by annealing the
sample, although with many irons annealing yields no increase in the amount
of evolved sulfur. Annealing is done as directed in the second preceding
paragraph.
,
With high-silicon cast irons, use hot (70 0 C.) diluted HCr.....(1 : 1), heat
rapidly to boiling, and then simmer. If the solution froths badly, add about
0.5 m!. of HF.
Concentrated HCI has been recommended for certain alloy cast irons, in
which case suitable condensing arrangements must be provided. Even then
care must be taken to avoid excessive neutralization 6f the alkaline absorbent.
Open-Hearth Iron and Wrought Iron.-Proceed as with carbon steels.
SILICON BY THE SULFURIC ACID MET,HOD

Procedur:e. Carbon Steels.-Transfer the sampte- (4.672 g. if silicon is

under 0.5%, and 2.336 g. if it is over this percentage) to a 300-m!. casserole,

CHEMICAL A:NALYSIS OF IRON AND .STEEL

1445

treat with 100 m!. of diluted H 2S0 4 (1 : 4), and cover.29 Warm gently. When
reaction ceases, remove and rinse the cover, and evaporate the solution, taking
care to avoid spattering, until copious fumes of H 2S0 4 are given off. Replace
the cover and continue the fuming for 2 to 3 min., but no longer since insoluble
sulfates may be formed. Cool somewhat, and add 100 ml. of warm water
(40 0 to 50 0 C.) at one time. Stir until salts are in solution, heating gently if
necessary but never boiling. Filter immediately using a rapid ashless filter
paper for high-silicon steels, and a tighter paper for low-silicon steels. Transfer,
all the residue to the paper, scrubbing the casserole thoroughly with a rubbertipped rod. Wash the paper and residue alternately with hot diluted HCI
(5 : 95) and hot water until salts of iron have been removed. This usually
requires six washings with 5-ml. portions of each solution. Transfer the paper
and residue to a platinum crucible, heat carefully until the carbon is gone,
and then cover and ignite for 10 to 15 min. at 10500 to 11000 C. Cool in a
dsiccator and weigh. Add enough diluted ,H 2S0 4 (1 : 1) to moisten the Si0 2
and then 3 to 5 ml. of HF. Evaporate to dryness, carefully heat until H 2S0 4
is gone, and then ignite at 1000 0 C. Cool in a desiccator and weigh. If the
factor w~ight was taken the loss in weight multiplied by 10 represents the
percentage of silicon.
NOTE.-In the 'above and succeeding methods some silica passes into the filtrate.
The amount left in solution depends chiefly on the silicon content of the alloy, that is,
about 0.005% silicon for a 0.1 % silicon steel, approximately 0.02% for a 0.4% silicon
steel and about 0.05% or more for a 4.5% silicon steel. Hence in umpire work it is
necessary to evaporate the filtrate and washings to fumes of H 2S0 4, dilute, filter, and
ignite the washed paper and contents with the first portion.
Alloy Steels.-Proceed as with carbon steels if the sample dissolves in
diluted H 2S0 4, If it does not, use the acid mixture (given for carbon steels)
or the perchloric acid method. so The latter is especially suited for highchromium alloys.
Tungsten Steels.-Transfer 2.336 g. of the sample to a beaker or casserole
provided with a cover glass. Add a mixture of 20 ml. of HCl and 20 ml. of
HN0 3, and warm until all the steel has dissolved. Add 60 ml. of warm diluted
H 2S0 4 (1 : 1), and evaporate until copious fumes of H 2S0 4 appear. Cool
somewhat, cautiously add 5 ml. of HCl, and swirl gently until well mixed.
Add 125 ml. of warm water (40 0 to 500 C.), and proceed as in the sulfuric acid
method. Ignite the tungstic acid' and silica together at a temperature of about
10000 C. and weigh. Treat with HF and H 2S0 4 , ignite at 800 0 C., and again
weigh.
Cast Iron.-Proceed a_s with carbon steel, using 2.336-g. of the sample.
With gray or mottled pig iron it is desirable to use a finely divided sample and
to boil vigorously with slightly more acid at the start, as the particles tend to
become coated with gelatinous silica.
29 More rapid and equally satisfactory determinations can be made by dissolving
the sample in 40 ml. of diluted HCI (1 : 1), evaporating to low volume, then adding the
H 2S0 4 and evaporating to fumes, etc. Another alternate procedure consists in dissolving
the sample in 80 m!. of an acid mixture (600 mI. of water, 200 mI. of H 2S0 4 (sp.gr. 1.84),
100 mI. of HNO s (sp.gr. 1.42) and 100 m!. of HCI (sp.gr. 1.19, heating until solution is
complete, and then evaporating to copious fumes of H 2S04, etc. In this method 5 mI.
of Hel is added to the partially cooled residue before dilution with warm water.

so See p. 1446.

1446

CHEMICAL ANALYSIS OF IROllT AND STEEL

Open-Hearth Iron.-Proceed as with carbon steels using 10- to 20-g.

samples and corresponding amounts of H2S04~ F~me gently for 12 to 15 min.
instead of 1 to 2 min. as with carbon steels. Umpire analyses of open-hearth
I
iron must be made in platinum dishes.
Wrought Iron.-Dissolve 4.672 g. of the sample'i';o. 80 ml. of the acid mixture
(as given for carbon steels), evaporate, and fume for 1 to 2 min. Cool somewhat and cautiously add to the warm solution 5 ml. of HCl as the solution is
swirled gently. When well mixed, add 125 ml. of warm water, 'stir, and complete the determination as directed for carbon steels.
SILICON BY THE PERCHLORIQ ACID METHOD
Procedure. Carbon Steels.-Transfer 4.672 g. of the sample for steels
containing less than 0.5% silicon and 2.336 g. for steels containing more than
0.5% silicon to a 400-ml. beaker. Add 40 ml. of HN0 3 (sp.gr:" 1.20).31 Heat
cautiously until solution is complete. Then raise the cover and add 40 inl. of
HCI0 4 (60 to 70%) for the 2.3-g. sample, or 60 ml. for the 4.7-g. sample.
Evaporate to fumes, cover the beaker, and 9'ontinue to heat at such a rate that
HCI0 4 refluxes on the sides of the beaker for 15 to 20 min. Cool ~omewhat,
add 125 ml. of hot water, 'stir until salts are in solution, crushing up any lumps
of silica with a flattened rod. Filter immediately and proceed as in the sulfuric
acid method. The washing of the paper must be thoro~gh, for residual HCI04
held in the silicic acid tends to cause 'popping and loss of material upon subsequent ignition.
Alloy Steels: Nickel, Chromium-Nickel, Chromium-Vanadium, and Similar
Low Alloy Steels Containing no Tungsten and Less than 5% of Chromium.Proceed as with carboI\ steels.
High-Chromium Steels (Stainless, 18% chromium, 8% nickel, etc.).Proceed with the perchloric acid method as directed for carbon steels after
dissolving in a solution of equal parts of HCl and HNO a
Cast Iron.-Proceed as with carbon steel, using! 2.336 g. of the sample for
i
white irons and 0.9344 g. for gray irons.
Alloy Cast Irons.-For high chromium-nickel cast irons proceed as with
carbon steels but dissolve the sample in a mixture of 20 ml. of HCI and 20 ml.
of HNO a
COPPER BY THE ELECTROLYTIC OR GRAVIMETRIC METHOD
Special Solutions Required. Diluted Sulfuric Acid (1 : 9).-Slowly stir
100 ml. of H 2S0 4 into 900 ml. of water.
Acidulated H 2 S Water.-Saturate 500 mt of diluted H 2S0 4 (1 : 99) with.
H 2S.
Sodium Hydroxide (5%).-Dissolve 5Q. g. of raOH in 1000 ml. of water.
Procedure. Copper Steels: (a) Preliminary Separation.-Transfer 5 g. of
the sample to a 600-ml. beaker, add 200 ml. of diluted H 2S0 4 (1 : 9), and heat
gently until action ceases. Dilute to 350 ml. with water, heat to boiling, and

Forhigh-sillcon alloys, dissolve the sample in 40 m!. of diluted HCI (1 : 1), care'fully oxidille with HNO a, and then add the HCI0 4 -Sfe p. 1422 for precautions when
using HCI0 4
'
31

CHEMICAL ANALYSIS OF IRON AND STEEL

1447

saturate with H 2S as the solution is allowed to cool during the course of 25 min.
Allow the precipitate to settle, filter on paper or paper pulp, and wash a few
times with diluted H 2S0 4 (1 : 99) that has been saturated with H 2S.
,Transfer the residue and p~per to a 25- or 35-ml. tall-form porcelain crucible,
ignite at a temperature not exceeding 550 C., and fuse with 2 to 4 g. of alkaline
pyrosulfate. Dissolve the cooled melt in 25 ml. of diluted HCl (1 : 9), dilute
the solution tq 100 ml.,2 neutr!J.lize with a solution of NaOH (5%), and add
approximately 0.3 ml. in excess. Boil for 3 min., digest for about 30 min.,
filter, and wash five or six times with a cool solution of NaOH (0.5%).
Dissolve the precipitate in 15 to 25 ml. of hot diluted HNO. (1 : 3), wash the
paper with hot water, add 5 ml. of H 2S0 4 and evaporate until fumes of H 2S0 4 appear. Cool-the solution, dilute to 40 mI., add an excess of several milliliters
of NH 40H, and heat to boiling. Allow the precipitate to settle, filter into a
250-mI. beaker, and wash with hot water. If more than 4 mg. of iron, tin,
chromium, or other elements precipitable by NH 40H, appear to be present,
dissolve the precipitate in 15 to 25 ml. of hot diluted HNO s (1 : 3), and repeat
the precipitation with NH 4 0H. Filter and combine the filtrates. The further
treatment depends on whether the copper is to be determined by electrolysis,
or by precipitating as the sulfide and igniting to the oxide. The former method
is to be preferred in analyses of the highest accuracy, and if over 0.25% of
copper is present.
(b) Determination of Copper by Electrolysis: Apparatus.-Ditect current
must be provided. The most satisfactory electrodes are platinum cylinders
made from 50- to 6O-mesh, sand-blasted gauze, stiffened by doubling the gauze
for about 3 mm. at the top and, bottom and supported by a platinum stem
which is flattened and welded the entire width of the. gauze.
Procedure.-Transfer the ammoniacal solution to a 250-ml. beaker, neutralize with diluted H 2S0 4 (1 : 1), and add an excess of 5 ml. of diluted H 2S0 4
-(1 : 1) and also 4 mL of diluted HNO s (1 : 1). Dilute the solution to 200 ml.
Adjust the 'anode and weighed cathode, cover the beaker with split watch
glasses, and electrolyze with a current of 0.5 amp. until the solution becomes
colorless (about ,2 hrs.). Rinse the cover glasses and the exposed stems of the
electrodes and sides of the beaker. Continue the electrolysis for 30 min. and
test for complete deposition (0.3 to 0.5 ml. of the electrolyte should not give a
brown color with 0.3 to 0.5 fil. of freshly prepared H 2S-water). When deposition is complete, lower the beaker while washing the cathode with a stream of
water. When free from' electrolyte, detach the cathode, dip it in alcohol,
shake off the excess, and dry for 1 min. at 100 C. Cool in a desiccator and
weigh. The increase in 'weight multiplied by 20 represents the percentage of
copper in the material.3s
(c) Determination of Copper by Weighing as Oxide. (Copper less than
O.25%).-Neutralize the ammoniacal solution obtained in (a) with diluted
H 2S0 4 (1 : 1), add 4 ml. of the diluted acid in excess, and adjust the volume to
about 100 ml. Heat to boiling and saturate with H 2S as the solution cools.
32 If an appreciable amount of silica has separated, it may be removed by filtration
before proceeding further.
33 Deposits of copper may be removed by immersing the cathode in diluted RNO,
(1 : 1), rinsing with water, then boiling with fresh RNO s for 5 to 10 min., rinsing with
water and then strongly igniting for 10 to 25 min. over one or two large Meker burners.

1448

CHEMICAL ANALYSIS OF IRON AND STEEL

Allow to settle, filter, transfer all the precipitate to the paper, and wash thoroughly with diluted H 2S0 4 (1 : 99) that has been siJ,turated with H 2S. Transfer
the paper and precipitate to a small porcelain or quartz crucible that has been
weighed with cover., He~t ~nder good oxidiz.in~ conditions until carbon has
been destroyed, and then'lgmte to constant weIght at 900 0 to 1000 0 C. Cover,
cool over a good desiccant, and weigh as CuO. If a 5-g. sample' was used, the
weight of CuO multiplied by 16.represents the percentage of' cop'per in the
material.
Carbon Steels.-Use a 10-g. sample and proceed as with copper steels.
Alloy Steels.-Most alloy steels may be analyzed as described for carbon
steels.
Tungsten Stee~s.-Treat 5-g. of the sample with 100 ml. of diluted HCI
(1 : 1). Carefully add 20 ml. of diluted HNO s (1 : 1), and boil gently until the
tungstic acid becomes bright yellow. Dilute the solution to 150 m!. with
boiling water, digest for a few minutes, filter and wash with diluted HCI (1 : 9).
Any copper retained in the tungstic acid is recovered as follows: Treat the
residue with NH 4 0H, add 5 g. of tartaric acid and sufficient H 2S0 4 to have an
excess of 5 m!. per 100 ml. of solution. Treat with H 2S, filter and combine
with the main sulfide precipitate.
Add 15 ml:of H 2S0 4 to the filtrate, evaporate just to fumes of H 2S0 4 , cool,
and add 100 ml. of water. If a small.residue of tungstic acid separates, filter
and wash with diluted H 2S0 4 (5 : 95). Add 5 g. of tartaric acid to the clear
filtrate, neutralize with NH 4 0H, ad.just the acidity to 5% by volume with
H 2S0 4, heat to boiling, and pass a rapid stream of H 2 S into the solution for 10
to 15 min. Let the precipitate settle, filter, wash, and complete the determination as directed for ({arbon steels.
Cast Iron.-Dissolve 5 g. of the sample in 100 ml. of diluted H 2 S0 4 (1 : 5).
When solution is complete evaporate to fumes of H 2S0 4 , cool somewhat, dilute
to 100 ml. with warm water, heat until salts have dissolved, and filter. Wash,
the residue with hot water, dilute the filtrate to 500 ml. with hot water, and
proceed as with carbon steel. In very accurate work the residue shall be
examined for copper.
Wrought Iron and OPen~Hearth Iron.-Proceed as with copper steels,
using 10-g. of the sample.
COPPER By THE THIOSULFATE-IOI?IDE METHOD
Special Solutions Required. Sodium Thiosulfate (50%).-Dissolve 100 g.
of Na 2S 2 0S5H 20 in 100 ml. of water and filter if a residue remains.
Standard Sodium Thiosulfate.-Dissolve 5 g. of Na 2S20S5H 20 in 500 mL
of recently distilled water, add 0.1 g. of Na 2 CO a and dilute to 1 liter. To
standardize, transfer 0.05 g. of pure metallic c(\pper to' a 150-Il1L beaker, cover
and dissolve the metal in 4 to 5 ml. of diluted HNO a (sp.gr. 1.20). Boil gently
to expel oxides of nitrogen. Cool, add NH 4 0H until the solution just turns
blue, add acetic acid, etc., and complete the titration as described in the method.
Starch Solution.-See the determination of Sulfur by the Evolution Method.
Procedure. Copper Steels.-Transfer 5 g. of the' sample to a 600-mL
beaker, add 100 m!. of diluted H 2S0 4 (1 : 9),.a&rd'"heat gently until action ceases.
Dilute to 250 mI., heat to boiling, add 10 mL of the solution of Na2S20a 5H 20

CHEMICAL ANALYSIS OF IRON AND STEEL

"

1449

(50%) and continue the boiling for 5 to 10 min. or until the precipitate settles
rapidly. Filter immediately, transfer ali the precipitate to the paper, and wash
the paper and. precipitate well with diluted H 2S0 4 (1 : 99) that has been
saturated with H 2S. Place the paper and precipitate in a porcelain or silica
crucible, dry and ignite at a low temperature (520 0 to 5500 C. i conveniently in a
muffle furnace) until all carbon is destroyed. Cool, transfer the contents of the
crucible to a 250-ml. beaker. Add 5 to 6 ml. of diluted HN0 3 (sp.gr. 1.20) 34
to the crucible, warm gently and pour upon the residue in the beaker. Rinse
the crucible with a little water and warm the beaker and contents until the
copper oxide has dissolved. Carefully evaporate the solution to a volume
of 2 to 3 ml. in order to expel most of the acid. 34 Cool, "add 90 ml. of water,35
and add either 10 ml. of a solution of ammonium bifiuoride (20%), or 1 g. of
sodium fluoride (to prevent interference by ferric iron). Then add NH 4 0H
until the solution just reacts alkaline to litmus. Cool the solution to room
temperature. Acidify with acetic acid (80%) and add 1 ml. of the acid in
excess. Add 3 to 4 g. of KI which has been dissolved in a little water, stir well,
and immediately titrate with the standard thiosulfate. When the brown tints
have nearly disappeared, add 5 ml. of starch solution and continue the titration
until one drop changes the color from, blue to yellowish white, and remains
permanent for 15 to 20 sec.
Carbon Steel.-Use 10 g. of the sample and proceed as with copper steel.
Alloy Steels.-Most alloy steels may be analyzed as described for copper
or carbon steels.
Tungsten Steels.-Treat 5 g. of the sample with 75 ml. of diluted HCl
(1 : 1). Carefully add 15 ml. of diluted HN0 3 (1 : 1) and boil gently until
the tungstic acid becomes bright yellow. Dilute the solution to 100 ml. with
boiling water, digest for several minutes, filter, and wash with diluted HCI
(1 : 9). For recovery of copper held by tungstic acid, see the determination of
copper in tungsten steels by the electrolytic method. 36
Add 10 ml. of H 2S0 4 to the filtrate and evaporate to fumes of sulfuric acid.
Cool, dilute to 200 mI., heat to boiling, add 60 ml. of the thiosulfate solution,
boil for 5 min., and complete the determination as described for copper steels.
Cast Iron.-Dissolve 5 g. of the sample in 100 ml. of diluted H 2S04 (1 : 4).
When solution is complete, evaporate to copious fumes of sulfuric acid. Cool
somewhat, dilute to 100 ml. with warm water, heat until salts have dissolved,
filter, and wash the residue with hot water. Dilute the filtrate to 250 mI., heat
to boiling, add 20 mI. of thiosulfate solution, continue the boiling for 5 min. and
complete the determination as described for copper steels.
Wrought Iron and OPen-Hearth Iron.-Use 10 g. of the sample and proceed
as described for copper steels.
34 Alternately, the residue on the paper may be dissolved in hot diluted HN0 3 (1 : 2)
and the solution caught in the original beaker, and 5 mi. of diluted sulfuric acid (1 : 1)
added. Then evaporate the solution to copious fumes, cool, dilute to 50 mi., treat with
HF, NHDH, etc., and complete the titration with thiosulfate as directed.
35 Since molybdenum interferes partly in this method, it is necessary, if the steel
contains more than 0.25% of molybdenum, to separate the latter from copper by treatment at this point with NaOH as directed in the electrolytic or gravimetric method.
36 See p. 1448.

1450

CHEMICAL ANALYSIS OF IRON AND STEEL.

NICKEL BY THE DIMETHYLGLYOXIME METHOD

SPecial Solutions Required. Dimethylglyoxime Solution (1%).-Dissolve

10 g. of dimethylglyoxime in 1000 m!. of ethyl or methyl alcohol, or
diluted NH 40H (13 : 7). Ten milliliters of this solution allows sufficient excess
to precipitate completely 0.015 g. of nicker. The !alcoholic solution keeps almost indefinitely; the ammoniacal solution does not keep as well, and no more
of it should be prepared than.can be used in a week. Occasionally NH 40H
contains impurities which cause incomplete precipitation of nickel.
Tartaric Acid Solution (25%).-Dissolve 250 g. of tartaric acid in 600 ml.
of water, filter, add 19 m!. of HN0 3, and dilute to 1 liter.
Procedure. ' Nickel Steels (0.05 to 3.5% nickel).-Transfer 1 g. of the
sample 37 to a 400-m!. beaker, cover, and treat with 60 ml. of dilutea HCl
(1 : 1).38 Warm until decomposition is complete, and then cautiously add 10
m!. of diluted HN0 3 (1: 1). Boil until iron and carbides are oxidized and
oxides of nitrogen have been expelled. Dilute to 200 m!. with hot water. Add
20 m!. of a solution of tartaric acid (25%), neutralize with NH 40H and add
1 m!. in excess. Filter if any insoluble matter (such as silica) is present, and
wash the paper and residue with hot wat,er containing a little NH 40H and
NH 4Cl. Add HCI until slightly acid, warm to about 60 to 80 C. and add
20 m!. of an alcoholic or ammoniacal solution of dimethylglyoxime (1%).
Add NH 40H until slightly alkaline, and digest for 30 min. at about 60 C.39
(a) Gravimetric Method.-Filter through a weighed Gooch crucible under
light suction, but do not allow the mat to run dry. Wash the precipitate
thoroughly with hot water. Add 5 ml. of dimethylglyoxime reagent to the
filtrate and washings, stir, and allow to stand to,determine whether precipitation
is complete. Dry the precipitate for 1 hr. at 110 to 120 C., and weigh as the
salt NiC8H1404N4 which contains 20.32% of nickel. Porous crucibles or
crucibles containing glass frit may also be used. If glass frit crucibles are used,
it is advisable to allow a thin mat of the precipitate'to form before strong suction
is applied. The precipitate may also be caught on filter paper and ignited at
about 800 C. to the oxide, NiO. In this case, ca~e shall be taken to prevent
sublimation of the glyoxime before it is decomposed. This is done by enclosing
the wet paper and precipitate in one or two wet ashless papers and igniting
so that the papers char before taking fire.
If appreciable cobalt (over 1%) or copper (over 4%) is present, add sufficient glyoxime to take care of them as well a~ the nickel, and preferably reprecipitate the nickel as follows: When the precipitate has been, washed,' discontinue the suction, place the original beaker under the funnel, and add 25 ml.
of a hot mixture of 20 ml. of diluted HCl (1 : 1) and 5 ml. (jf HN0 3. After
1 min. apply suction until dry, repeat the trea~ment with 25 m!. of acid mixture,
37 If the steel contains less than 1 % of nickel, 2- or 3-g. samples may be used with
correspondingly larger amounts of tartaric acid. For Ejteels containing more than 3.5%
of nickel, either a sample equivalent to about 0.035 g. of nickel, or a 'suitable aliquot
portion of a larger sample may be used.
38 If the percentage of chromium is less than 0.5%, the sample may be dissolved in 5a
mI. of hot diluted HN0 3 (1 : 3).
39 If the amount of nickel is small (under 0.2%), or if much cobalt is present, the
solution should be allowed to stand at room temperature overrtight and filtered cold.
In routine analyses invqlving more than 2 mg. of ni~el..and no appreciable amounts of
cobalt, the solution mayi be filtered within 5 min. after adding the glyoxime and stirring.

CHEMICAL ANALYSIS OF IRON AND STEEL

1451

"

drain, and wash thoroughly with 50 ml. of hot water containing 1 g. of tartaric
acid. Nearly neutralize the absolutely clear solution with NH 4 0H and precipitate with dimethylglyoxime and NH 40H as before. A glyoxime precipitate
contaminated by cobalt is a darker red than a pure nickel dimethylglyoxime.
(b) Electrolytic Method.-Occasionally in umpire analyses, and with highnickel steels the determination is best completed by the electrolytic method
as follows: Filter the nickel dimethylglyoxime precipitate on a 12 or I5-cm ..
paper and wash thoroughly 18 to 20 times with hot water. Dissolve the
precipitate in hot diluted HNO a (1 : 3) and wash the filter thoroughly with hot
water. Add 20 ml. of diluted H 2S0 4 (1 : 1) and evaporate to strong fumes of
sulfuric acid. Cool somewhat, add 10 ml. of HNO., and repeat the evaporation,
to fumes. Rinse the cover and sides of the beaker with water and fume the
solution again to insure the expUlsion of every trace of the HNO a Cool, add
, 50 ml. of cold water, and heat until salts are dissolved.
Neutralize with NH 4 0H and add an excess of 25 ml. of NH 40H. Dilute
to 175 ml. and electrolyze with a current of from 1 to 2 amp. 'using a weighed
platSnum gauze cathode and a spiral platinum anode. Continue the electrolysis
until the solution has become colorless (5 to 8 hI's.). The solution may be
tested for complete electrolysis by adding one or two drops of it to 1 ml. of
a solution of potassium thiocarbonate. 4o A pink or red color indicates the
presence of nickel. Without interrupting the current, wash the cathode with
cold water as it is withdrawn from the solution. Dip the cathode into a beaker
of water and then into alcohol, dry for a few minutes at 100 C., cool, and weigh.
The increase in weight of the cathode multiplied by 100 and divided by
the weight of the sample taken gives the percentage of nickel in the sample.
In very accurate work, dissolve the deposit in warm HNO a, wash the cathode
with water, then alcohol, dry for a few minutes at 80 C., and reweigh.
High-Chromium High-Nickel Steels (20% chromium, 20% nickel; 18%
chromium, 8% nickel, etc.).-Transfer 0.35 to 0.5 g. of the sample to a 400-ml.
beaker, and treat with 20 ml. of diluted HCl (1 : 1) and 20 ml. of diluted HNO a
(1 : 1). Heat until solution ensues, add 15 ml. of diluted H 2S0 4 (1 : 1), and
evaporate just to light fumes (for low-carbon, low-silicon alloys this latter step
may be omitted). Cool somewhat, and add 100 ml. of water. Warm until salts
dissolve, filter, and proceed as directed for nickel steels, sufficient dimethylglyoxime being added to precipitate all of the nickel (20 to 40 ml. of a 1 %
solution).
Carbon Steels, and Other Steels with less than 0.05% Nickel.-Transfer
5 g. of the sample to a 400-ml. beaker, add 40 ml. of diluted HCl (1 : 1). Heat
until solution ensues and then carefully add 15 ml. of diluted HNO a (1 : 1).
Evaporate to a volume of about 15 ml. and add 50 ml. of diluted HCl (sp.gr.
1.12). ,Transfer. to a 200-ml. separatory funnel, rinsing the beaker with
several 15-ml. portions of the diluted HCl (sp.gr; 1.12). Cool to 10 C., add
120 ml. o(lether, and carefully shake for 1 to 2 min. in a stream of cold water.
Let settle for several minutes and then draw off the lower clear solution into
the original beaker. Gently heat the solution in the beaker to expel the ether
(avoid free flames), add 0.3 g. of KCIO a, boil until the chlorate is decomposed,
40 Prepared by saturating one-half of a solution df KOH (5%) with H.S, adding the
other half, and heating moderately with one twenty-fifth of its volume of es 2 The dark
red liquid is decanted from the undissolved eS 2 and kept in a well-closed flask.

1452

CHEMICAL ANALYSIS OF IRo.N AND STEEL

dilute to 1()0 mI., and add 3 g. of tartaric acid. Make the solution alkaline
with NH 40H and filter. Acidify with Hel and1complete the determination
as directed for nickel steels.
Cast Irons.-Dissolve 5 g. of the iron in 40 l mL of diluted HCr (1 : 1),
carefully add about 15 ml. of diluted HNO a (1 l, 1) to oxidize the iron, and
evaporate to dryness. Drench the hot dried ml\ss with 10 mL of HCl and
then dilute with 75 ml. of hot water. Filter, wash with diluted HCl (sp.gr.
1.12) and evaporate the filtrate to a syrupy consistency. Add 50 mL of diluted
.HCl (1 : 1), transfer to a 200-mL separatory funnel, rinse the beaker with
several small portions of the diluted HCl (1 : 1), add 12'0 ml. of ether, and
complete the determination as directed for carbon steels.
High-Nickel Chromium Alloy Cast Irons (15% nickel, 6% copper, 2%
chromium, etc.).-Transfer a 2.5-g. sample to a 400-mL beaker or flask and
treat with a mixture of 25 ml. of HCl and 25 mL of HNO a When solution is
. complete add 30 mL of HCI04, 5 to 10 drops of HF, and fume for 10 to 15 min.
after the chromium has been oxidized. Cool somewhat, add 100 ml. of water
and heat to boiling. Filter and wash well with diluted HCI (5 : 95), catching
the filtrate and washings in a 250-ml. volumetric flask.41 Mix the contents,
cool to room temperature, adjust to the mark and mix thoroughly. Pipette
out 50-ml. aliquot portions and proceed by the glyoxime-electrolytic method
as directed for nickel steels. Either dissolve and reprecipitate the nickel glyoxime, or determine any occluded copper in the deposit and correct for same.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon steels.
Nickel Wrought Iron.- Proceed as with 'Dickel steels, but dissolve the first
glyoxime precipitate and again precipitate nickel with dimethylglyoxime as
directed in the procedure with cobalt steels:
NICKEL BY THE CYANIDE TITRATION METHOD
Special Solutions Required. Sulfuric-Citric, Acids.-Dissolve 200 g. of
citric acid in a cooled mixt!}re of 100 ml. of H 2S04 and 900 ml. of water.
Standard Silver Nitrate Solution.-Dissolve 5.789 g. of AgNO a in water
and dilute to exactly 1000 ml. Each 1 mt of this solution is equivalent to
0.001 g. of nickel.
Standard Potassium Cyanide Solution._:_Dissolve 4.5 g. of KCN in 1000
ml. of water containing 1 'g. of KOH. Standardize ;gainst a steel of known
nickel content by applying the method described in the procedure below. The
cyanide changes with age so its titer should be checked frequently.
Procedure. (a) Direct Titration.-This,method is not applicable to steels
containing copper, cobalt or more than 1% of tungsten. Copper reacts like
nickel, and divalent cobalt not only consumes cyanide but also causes trouble in
obtaining an end point. A fairly satisfactory: end point, r~presenting both
nickel and cobalt, can be obtained by first oxidizing the cobalt in ammoniacalcitrate solution to the trivalent state with NaCIO a
41 The residue sometimes conntins appreciable amounts o'f nickel. This may be
recovered 9Y igniting, treating with HF, evaporating. and fusing the residue with a smali
amount of K 2S20 7, dissolving and adding to the main solution.

CHEMICAL ANALYSIS OF IRON AND STEEL

1453

Nickel Steels.-Transfer 1 g. of the sample to a 400-ml. beaker, treat with

20 ml. of diluted HNO a (1 : 1) and heat until solution is complete. 42 Add 10
ml. of a solution of (NH 4)2S20g (15%), and boil for 5 min. Cool, and add 50
ml. of the sulfuric-citric acid solution. Immediately add NH 40H until the
solution assttmes a greenish tint, dilute to 200 ml. and complete the neutralization ~s follows: Add exactly 2 ml. of the standard AgNO a, and, if no precipitate
appears, add diluted HCI (1 : 10) until a precipitate of AgCI forms. Then add
diluted NH 40H (1 : 1), dropwise, with constant stirring until the precipitate
just dissolves. Add 3 ml. of NH 40H (sp.gr. 0.90), 2 ml. of a solution of KI
(10%), and titrate with standard KCN with continuous stirring until the
solution becomes perfectly clear., Determine the volume of KCN equivalent
to exactly 2 ml. of the standard AgNO a Subtract this quantity from the
volume of KCN and multiply by the nickel titer of the cyanide solution.
(b) Following a Preliminary Precipitation by Dimethylglyoxime.-This
method may be applied to steels containing cobalt, copper and tungsten.
The silver and cyanide solutions are the same as described in (a) but instead
of standardizing these solutions against a standard steel the theoretical titer
of AgNO a is taken as standard and the equivalent of the KCN is found by
titration in a water solution containing no iron or citrate.
Nickel Steels.-Precipitate the nickel in a I-g. sample as described for
nickel steels by the dimethylglyoxime method. 43 Dissolve the washed precipitate with a hot mixture of 20 ml. of diluted HCI (1 : 1) and 5 ml. of HNO a
in a 400-ml. beaker.44 Evaporate the solution to 50 ml. or until free from
dimethylglyoxime and oxidizing gases. Cool the solution, dilute ,to 150 mI.,
add exactly 2 ml. of the standard AgNO a, neutralize with NH 4 0H, and complete the titration as described in (a).
(c) Following a Preliminary Extraction with Ether.-This method is not
applicable to steels containing cobalt, or more than 1% of tungsten.
Nickel Steels.-Transfer 1 g. of the sample to a 150-ml. beaker, cover, and
treat with 20 ml. of diluted HCI (sp.gr. 1.12). When reaction ceases, carefully
add 4 ml. of diluted HNO a (1 : 1), and boil to expel oxides of nitrogen. Cool,
transfer the solution to a 200-ml. separatory Tunnel, and rinse the beaker with
diluted HCI (sp.gr. 1.12). Cool to 10 C., add 40 ml. of ether (perform this
operation away from open flames or hot plates), shake gently for a few minutes,
let settle for 2 min., and draw off the acid layer into a 250-ml. beaker. Add
5 ml. of diluted HCI (sp.gr. 1.12) to the ether portion. Cool, shake, let settle
for 1 min., draw off the acid layer arid add it to the main extract. Heat gently
to expel dissolved ether, add 0.2 g. of KCIO a, and boil until the chlorine is
driven off. Dilute to 100 ml. with water, neutralize with NH.OH, add an
excess of 3 to 4 ml. and boil for a few minutes. Filter and wash with hot water.
To the filtrate, add 10 mi. of HCI, heat just short of boiling, and treat with H 2S.
Filter and wash with acidulated H 2S water. Boilto expel H 2 S, dilute to 200 mI.,
neutralize with NH 4 0H, and proceed with the titration as described in (a).
42 If the steel contains more than 1 % of chromium, decompose the sample in 20 ml.
of diluted H 2S0 4 (1 : 4), oxidize by carefully adding 5 ml. of HNOa, boil for 5 min., cool,
and add the sulfuric-citric acid mixture.
4a See p. 1450.
44 If the steel contains more than 4% of copper, or 2% of cobalt, the nickel should be
reprecipitated with glyoxime. See dimethylglyoxime method.

1454

CHEMICAL ANALYSIS OF IRO,N AND STEEL

Carbon Steels.-Transfer 5 g. of the sample to a 400-ml. beaker, add 40 mL

of diluted HCI (1 : 1). Heat until solution ensu~s, then carefully add 15 ml.
of diluted HNO z (1 : 1). Evaporate to a volumE< of about 15 mI., add 50 mI.
of diluted HCI (1 : 1), and extract with 120 ml. o{ ether as directed for carbon
steels by the dimethylglyoxime method. 45 Heat tlle acid extract geIftly to expel
dissolved ether, add 0.3 g. of KCIO a and complete t};le determination as described
in (c).
Cast Irons.-Dissolve 5 g. of the iron in 40 mI. of diluted HCI (1 : 1), carefully oxidize with HNO s, and evaporate to dryness. Dilute, filter and extract
with ether as directed for carbon steels by the dimethylglyoxime method. 45
Complete the determination as described in (c).
'
CHROMIUM BY THE PERSULFATE OXIDATION METHOD
Special Solutions Required. Sulfuric-Phosphoric Acid Mixture.-'-Mix 320
ml. of diluted H 2S0 4 (1 : 1), 80 ml. of H s P0 4 (85%) and 600 ml. of water.
Silver Nitrate (1%):-Dissolve 10 g, of AgNO a in 1000 ml. of water.
Ammonium Persulfate (150/0).-Make up as needed by dissolving.15 g. of
the salt in 100 ml. of water.
Standard Potassium Permanganate Solution (0.06 N, approximately).-Dissolve 2 g. of KMn04 in 1000 ml. of water. Let age for 1 week or more. Filter
through purified asbestos and standardize against ,the National Bureau of
$tandards' standard sample 40c of sodium oxalate as follows: Transfer 0.2 g.
of the standard sodium oxalate to a 400-ml. beaker. Add 150 ml. of diluted
H 2S04 (5 : 95), previously boiled for 10 to 15 min. and then cooled to 27 C.
3 C. Stir until the oxalate has dissolved. Add 40 to 42 mI. of the KMn04
(0.06 N) at a rate of 25 to 35 ml. per min., while stirring slowly. Let stand
until the pink color disappears. (If the pink color should persist because the
permanganate is too ~trong, discard, and begin again, adding a few less milliliters of the KMn04 solution.) Heat to 55 to 60~ C. and complete the titration
by adding permanganate until a iaint pink colof persists for 30 sec. Add the
l
last 0.5 to 1 ml. dropwise, with particular care to allow each drop to become
decolorized before the next is introduced. Determine the excess of permanganate required to impart a pink color to the solution by matching the color
in another beaker containing the same amount of the specially treated diluted
H 2S0 4 at 55 to 60 C. One milliliter of exactly 1 l't_KMn04 is equivalent to
0.01734 g. of chromium.
"
Standard Ferrous Sulfate.-Dissolve 32 g. of FeS04lNH 4)2S04'6H 20 in
a cool mixture of 50 ml. of H 2S0 4 and 950 ml. of water. To obtain the ratio
of the FeS04 solution to the KMn04 solution, take 25 ml. of the former, ~I.ilute
to 350 ml. with cool diluted H 2S04 (5 : 95), add 2 ml. of H aP0 4 and titrate
with KMn04 solution to a faint permanE;nt pink tint. Det~rmine the blank
on the same volume of water and acids, deduct, and calculate the volume of
KMn04 solution which is equivalent to 1 mI. of the FeS04 solution. The
ratio of the FeS04 solution to KMn04 must be determined daily unless the
ferrous 'solution is kept under hydrogen. Stronger fllutions of FeS04 and
KMn04 may, of course, be prepared for use ~th high-chromium steels. Ap<

oJ,

45

See p. 1451.

CHEMICAL ANALYSIS OF IRON AND STEEL

1455

proximately 0.1 N solutions are desirable with material containing over 3%

of chromium.
Procedure. Chromium Sieels.- With steels, containing less than 2% of
chromium, transfer 2 g. 'of the sample to a (JOO-ml. beaker and add 60 m!. of acid
mixture. Heat until action ceases, cautiously add 10 ml. of diluted HNO a
(1 : 1), and boil until all carbides are dissolved and oxides of nitrogen are
expelled. With steels containing between 2 and 5% of chromium, dissolve 1 g.
of the sample, evaporate until salts separate, dilute with about 50 ml. of warm
water, add 5 to 10 m!. of HNO a, and again carefully evaporate until salts
separate. If carbides still persist, filter, wash the paper with warm water,
ignite the paper and residue, fuse with sodium carbonate, and add the solution
of the melt to the main solution. After the sample is completely dissolved
dilute to 300 m!. with hot water and add 5 mi. of the AgNO a (1 %) and 20 ml.
of the (NH 4hS20S (15%). Boil the solution for 8 to 10 min. If the color of
KMn04 does not develop add more AgNO a and (NH 4)2S20S and again boil for
10 min. ,Add 5 m!. of diluted HCI (1 : 3) and continue the boiling for 5 min.
after the pink color has just disappeared. If the KMn04 is not destroyed by
boiling for 10 min., or if a precipitate of Mn02 remains, add 2 to 3 mi. more of
diluted HCI (1 : 3) and boil as before. The total period of boiling after the
addition of (NH4) 28 20 8 shall not be less than 15 min. ; 30 min. will do no harm.46
Cool the solution, dilute to 400 mI., and add a measured volume of the standard
FeS04 solution (25 ml. for less than 1.5% chromium steel, 50 ml. for 1.5 to 3% chromium, etc.). Stir and titrate with the standard KMn0447.until an end
point is obtained which is permanent upon continued stirring for 1 min.48 The
titration must be corrected for dilution effect and color interference. The
correction ,may be made (1) empirically; 49 (2) by titrating the same volume
4& Chromium (and vanadium) may also be oxidized to Cr6 (and V 5) with hot concentrated HCIO,. See p. 1422 for precautions when using HCIO,. Unless special precautions are taken, oxidation of chromium does not quite reach 100%. The oxidation procedure is as follows: Transfer the sample (depending on the chromium content, 0.5 g. for
stainless and 2 g. for 1 to 3% chromium steels) to a 500-ml. tall-form beaker (Pyrex) or
500-mI. Soxhlet flask (Pyrex) and cover with a watch glass. If a flask is useq the watch
glass should be raised by a small glass hook to avoid subsequent formation of a liquid
seal. A stirring rod placed inside the vessels, will also prevent bumping tendencies.
Add 20 to 25 mI. (depending on the weight of the sample) of perchloric acid (60%). A
technical grade of acid is'satisfactory and cheaper. Warm until the steel is'dissolved and
boil gently for 20 min. Cool rapidly, by immersing the vessel in cold water for onefourth of a second, withdraw rapidly, and then immerse a second time and swirl for 8
to 10 sec. Add an equal volume of water and boil 3 min. to remove chlorine. Cool,
dilute ,to 200 to 250 ml., and titrate with FeSOcKMn04 as directed.
,7 In the absence of tungsten, the end point may be obtained by potentiometric titration with standard FeSO, solution. In this case, the ferrous solution is standardized
with recrystallized K 2Cr 207. Vanadium is included in this titration. If present, it
must be determined and subtracted from the apparent chromium value (vanadium,
per cent XO.339 = chromium, per cent). With tungsten steels the potentiometric procedure is modified by omitting H.PO. and oxidizing tungsten to tungstic acid with HNO.
(when H.PO. is added to hold tungsten in solution the potentiometric end point with
FeSO 4 is diffictfit to detect).
48 In the absence of vanadium the first end point is permanent, but if vanadium is
present, the end point will fade at first, owing to the slow oxidation of vanadium,from
the quadrivalent to the quinquevalent stage in a cold solution.
49 In this correction the dilution effect is ignored and the volume of KMnO. used in
overcoming the green color is taken as equivalent to 0.6% of the chromium present.
The correction is usually applied to the titer of the solution, as, for example, by using the
titer 0.01744 g. of chromium instead of 0.01734 g. per milliliter of 1 N solution.

1456

CHEMICAL ANALYSIS OF IRON, AND STEEL

of FeS04 solution in & solution of like volume and acidity and containing the
same amounts of the coloring elements iIi their fillal valencies; or (3) by a
second titration of the final solution. The last is: the most convenient 'and
satisfactory in occasional analyses, and may be performed by boiling the
solution which has just been titrated for 10 ,min. in \order to destroy the slight
excess of permanganate, cooling to room temperature, and then titrating with
KMn04 to the color that was originally taken as th~ end point. The sQlution
can be used for the determination of vanadium. 50
Chromium in 2-g. sample, per cent = (A X B) - (C - D) X 50 E
where A=the volume of FeS04 solution,
B=the volume of KMn04 solution which is equivalent to 1 ml. of FeS04
solution,
C =the volume of KMn04 solution,
D = the correction for the end point, and
E =the chromium titer of the KMn04 solution.
If the solutions are equivalent and a 2-g. sample is used, calculate the
chromium as follows:
Chromium in 2-g. sample, per cent = A - ( C - D) X 50 E

High-Chromium High-Nickel Steels.-Transfer 0.5 g. to a 600-mI. beaker

- and treat with 75 m!. of diluted H 2S0 4 (1 : 4). When solution is complete
evaporate to'salts, cool, dilute to 70 mI., and dissolve the iron salts. Oxidize
the iron by the cautious addition of HNO a Add 3 to 4 drops of HF and again
evaporate just to salts. Cool, dilute to 300 ml. with hot water, add 10 ml. of
the AgNO a solution, 15 g. of (NH4) 2S208, and boil the solution for 10 to 12 min.
Add 5 mI. of diluted HCI (1 : 3) and complete the determination as in chromium
steels. Stronger solutions of FeS04 and KMn04 are more convenient for use
with high-chromium steels.
The visual end point in high-chromium steels can be detected more easily
when oxidation-reduction indicators are used, that is, ortho-phenanthroline
ferrous complex for the FeS04-K 2Cr 207. With o-phenanthroline (1 to 2 drops
of the 0.025 Molar indicator) the end point in the FeS04-KMn04 titration is
indicated by the change in color from pink to clear green (permanent for 60
sec.). With diphenylaminesulfonic acid, the chromic acid is titrated with the
FeS04 solution to a clear green end point. The FeS04 solution "in this case is
standardized either on a standard chromium steel or pureK 2 Cr 207. Vanadium,
if present, will also be titrated as in the potentiometric method. When
o-phenanthroline is used the vanadium may be determined in the same solution,
after the initial end point is obtained, by reducing the acidity of the solution
with sodium acetate and titrating slowly at 50 C. with the standard KMn04
(0.05 N) to the green end point.
Carbon Steels and Other Steels with less than. 0.15% Chromiilm.-Transfer
10 g. of the sample 51 to a 500-mI. Erlenmeyer flask, add 110 ml. of diluted
60 See the Determination of Vanadium in Chromium-Vanadium Steels by the Ferrous
Sulfate-Potassium Permanganate Method, p. 1459.
51 Larger or smaller samples may be taken, but the volume' of acid shall be varied
accordingly. A good rule to follow is to use the ~quivalent of 1 m!. of concentrated
H 2S0 4 for each 1 g. of steel and then 1 m!. in excess.

CHEMICAL ANALYSIS OF IRON AND STEEL

1457

H 2SP4 (exactly 10% by volume). Heat to boiling, boil until reaction is complete and then dilute with 100 m!. of boiling water. Add a solution of NaHCO a
(8%) from a burette until a permanent precipitate appears (approximately
36 m!. with carbon steels) and then 4 m!. in excess. Boil for 1 min., let settle,
filter on a rapid filter, and quickly wash the flask and precipitate two or three
times with hot water. If the precipitation has been properly performed, there
will be no more precipitate than can be conveniently handled on an ll-cm.
paper. The filtrate will become cloudy in the funnel stem and in the receiving
vessel on account of oxidation arid hydrolysis. Ignite the residue in a nickel
or iron crucible (free from chromium) and fuse with ten or twelve times its
volume of Na 202 (free from chromium). Dissolve the cooled melt by immersing it in 100 m!. of cold water, remove the crucible, add 1 g. of Na 202 and boil
for 5 to 10 min. or allow to stand on the steam bath for 30 min. Filter through
an asbestos pad, 52 preferably on a small Buchner funnel, and wash with a cold
solution of NaOH (2%) containing 1% of Na 2S04. Dilute to a measured
volume and compare the color with a standard solution of K 2 Cr 20r containing
approximately the same concentration of chromium and alkali.
To prepare the standard solution, dissolve 0.283 g. of K 2Cr 207 in water and
dilute to exactly 1000 m!. Each 1 m!. shall correspond to 0.1 mg. of chromium.
When these solutions are used, they shall be made alkaline to compare with the
solution under test. Solutions containing from 2 to 10% of NaOH and 1 mg.
of chromium per 100 m!. are suitable for the colorimetric comparisons.
If the color of the unknown solution is too deep for convenient colorimetric
comparison, the solution may be boiled thoroughly to decompose all the peroxide, acidified, and one of the titration procedures applied.
Cast Iron (less than 0.15% chromium).-Proceed as with carbon steels.
Cast Iron (over 0.15% chromium).-Transfer 2 g. of the sample to a BOO-m!. .
beaker and add 60 ml. of the sulfuric-phosphoric acid mixture. Heat until
action ceases, add 15 m!. of diluted HNO a (1 : 1) and boil until the oxides of
nitrogen are expelled. Evaporate until salts separate, dilute with about 50 m!.
of warm water and digest until salts are dissolved. Filter and wash the paper
with warm water. If chromium is less than 0.75%, the residue will be practically free from chromium. To recover any insoluble chromium, ignite the
paper and insoluble matter until all carbon is consumed. Treat the residue with
HF and H 2S0 4 , fuse with Na 2 CO S , and add the solution of the melt to the main
solution. To the latter add 5 ml. of the AgN0 3 , dilute to 300 m!. with boiling
water and complete the determination as directed for chromium steels.
High-Nickel Chromium Alloy Cast Irons (15% nickel, 6% copper, 2%
chromium, etc.).-Ttansfer 2 g. of the sample to a 500-m!. Erlenmeyer flask
and add 20 m!. of a mixture of equal parts of HCI and HNO s. Heat until
action ceases. Then ad.d 15 m!. of HCI0 4 (60%), 5 drops of HF and evaporate
to fumes on a hot plate.' Fume 1 to 2 min. over an open flame and then 10 min.
more on a hot plate. Cool somewhat, add 50 ml. of water and transfer to a
600-m!. beaker. Add 20 m!. of :fi2S04, dilute to 300 m!. and complete the
determination by the persulfate-silver nitrate method.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon steels.
62 Asbestos is more satisfactory than filter paper.
If paper is used, it shall first be
thoroughly washed with a solution of NaOH (5%) in order to remove soluble organic
matter.

1458

CHEMICAL ANALYSIS OF IRQ:N AND STEEL

VANADIUM BY THE ELECTROLYTIC SEPARATION METHOD

j

. Special Solutions Required. Diluted Sulfuric Acid (1 : 9) .-Slowly stir

1O~ ~l. of H 2S0 4 into ?OO ~l. of ?ool w.ate~, cool, f1nd the. n dilute to 1000 ml.
It IS Important that thIS aCId be of the mdlCated strength.
Sodium Bicarbonate (8%).-Dissolve 80 g. of N!tHCOa in 1000 ml. of water.
Potassium Ferricyanide.-Prepare a very dilute soJution as needed, by
dissolving a crystal the size of the head of a pin in 25 ml. of water.

Standard Potassium Permanganate.-Prepare as described under the

determination of chromium in chromium steels by the persulfate oxidation
method.
Procedure. Vanadium Steels.-Transfer 2 g. of the sample to a 300-ml.
Erlenmeyer flask, and add 30 ml. of diluted H 2S0 4 (exactly 10% by.volume).
When action is complete, dilute to 100 ml. with boiling water and heat to
boiling. Agitate gently while adding N aHCO a solution from a burette until
a permanent precipitate is formed, and then add 4 ml. more. Cover the flask,
boil for 1 min., and let the precipitate settle. Filter rapidly, conveniently by
moderate suction through a cone and paper containing some paper pulp, and
wash the flask and precipitate four or five times with hot water. The filtrate
will become cloudy in the funnel stem and receiving vessel because of oxidation
and hydrolysis of the iron, but this is orno consequence. Place the paper and
precipitate in the original flask, add 20 ml. of HNO a, sMke until the paper has
broken up, and add 5 ml. of H 2SO:' Heat over a free flame until fumes of
H 2S0 4 are given off. Cool, and, if organic matter is still present, ad.d m~e
HNO a and repeat the treatment. Finally cool, wash down the flask, and
evaporate until it is certain that HNO a has been completely expelled. Cool,
dilute to 40 ml., sa neutralize the solution with NH 4 0H, add H 2S0 4 till acid and
then add an excess of 0.5 ml. per 100 ml. of solution. Transfer the solution to
the electrolyzing apparatus, rinsing the flask with small portions pf water.
The electrolysis may be performed with a mercury cathode in (1) a special
modified funnel. 54 (2) in an ordinary beaker withfa sealed-in platinum wire, or
(3) .in an ordinary beaker with the cathode wire enclosed in a glass tube extending into the mercury. In any case, the cathode wire shall be embedded in
5a If tungsten.is present, the solution must be filtered, the tungstic acid washed with
diluted H 2S0 4 (1 : 99) its vanadium content determined colorimetrically and added to
that found by titration. The colorimetric test 'shall be made as follows. Dissolve the
tungstic acid in a solutipn of NaOH, or fuse the ignited impure oxide with Na 2CO a and
extract the melt with water. Filter if not clear, and dilute to .75_ to 100 m!. Add
H.P0 4 (sp.gr. 1.37) until acid, then a 5-m!. excess, and let stand for 1 to 2 hrs. Compare
the yellow solution of vanadotungstic acid with a standard prepared by adding from a
burette a standard solution of vi1lladotungstic acid to water until the intensity of color
in the solution is the same when the solutions are of equal volume. The standard solution
of vanadotungstic acid shall be prepared as follows: Dissolve_2.5 g. of sodium tungstate
and enough sodium vanadate to give exactly 0.05 g. of vanadium in 100 m!. of water.
Dilute to 200 mI., add 25 mI. of H.P0 4 (sp.gr. 1.37)" and dilute to exactly 500 m!. Ammonium salts cannot be used as they give rise to turbid solutions., If only ammonium
vanadate is available, dissolve it in water, expel all ammonia by boiling with a slight
excess of NaOH, and acidify with HaP04.
The vanadium held by the tungsten approximates 0.01 mg. of vanadium per 0.01 g.
of tungsten and, in routine analyses, the correction is made by calculation, that is,
.
0.018% vanadium is added for an 18% tungsten ~el.'
54 Melaven cell, see A. S. Melaven, "Electro1ytic Cell for Use with the Mercury
Cathode," Industrial and Engineering Chemistry, Analyti?al Edition 2, p. 180 (1930).

CHEMICAL ANALYSIS OF IRON AND STEEL

1459

about 200 g. of mercury, and the solution electrolyzed using a current density
of approximately 0.16 amp. per sq. cm. as the solution is stirred or agitated.
Continue the electrolysis until iron is absent as indicated by a ferricyanide
test on a small drop of the electrolyte. This should not require more than 45
min. When all the iron has been removed, draw off the electrolyte, and wash
the mercury two or three times with water while the current is continued. lili
In these operations care shall be taken to prevent any amalgam from passing
into the electrolyte. li6 Add 2 to 3 ml. of diluted H 2S0 4 (1 : 1), heat to 70 0
to 80 0 C., and add KMn04 solution until a strong pink colorappears. Heat to
boiling, and pass a current of S02 into the solution until the vanadium.is
reduced (2 to 5 min.),li1 Continue the boiling, and pass a rapid stream of CO 2
(free from O 2) until the solution is free from S02. This may be ascertained by
passing the gas issuing from the flask into 5 m!. of water containing a drop of
diluted H 2S0 4 (1 : 1) and enough KMn04 to give a faint pink tint. Cool the
solution to 60 0 to 80 0 C., and titrate with the KMn04 (0.03 N). Repeat the
reduction and titration until concordant results are obtained. Correct the
titrlttion by a blank determInation (usually amounting to about 0.1 m!.) of
the KMn04 (0.03 N) on a solution of like volume and acidity. To obtain the
percentage of vanadium, multiply the corrected volume by the vanadium titer
of the solution, multiply by 100, and divide by the weight of the sample.
Carbon Steels.-Proceed as with vanadium steels but treat 5 g. of the sample
with 60 m!. of diluted H 2S0 4 (exactly 10% by volume).
Cast lron.- Proceed as with. carbon steels.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as w;ith carbon steels.
VANADIUM BY REDUCTION WITH FERROUS SULFATE AND
TITRATION WITH PERMANGANATE li8
Special Solutions Required.-See the special solutions required in the
determination of Chromium by the Persulfate Oxidation Method.
Procedure. Chromium-Vanadium Steels.-After the determination of
chromium, by the persulfate oxidation method, add 5 m!. of H aP0 4 (85%),
unless the solution already contains it, and then add 15 m!. of a solution of
lili To clean the mercury; transfer it to a large flask, add diluted HNO a (1 : 9) con- .
taining a little NaN0 2 and agitate the mercury by drawing air through it for several
hours. Remove the acid, add a fresh portion of acid, and agitate again. Draw off the
second portion of acid, add water and repeat the agitation. Repeat the treatment with
water to remove all acid.
66 With chromium steels it is usually necessary, because of the brittle chromium
amalgam, to filter the electrolyte through a rapid filter and wash with water.
57 The direct use of a solution of H.SOa or an alkali sulfite is unwise unless it has been
freshly prepared, for after a lapse of time they contain oxidizable bodieE other than
sulfurous acid or sulfite. Sulfur dioxide is most conveniently used from a cylinder of
theJiquefied gas, or it may be obtained as wanted by heating a flask containing a solution
of H.SOa or a sulfite to which diluted H.s0 4 (1 : 1) is added.
.
58 Where apparatus for potentiometric titrations is available, oxidation with HNO s
is a very satisfactory method for the determination of vanadium, particularly for
chromium-tungsten-vanadium steel. The sample is treated with 100 ml. of diluted
H 2S0 4 (1 : 4) and then with HNO a to oxidize iron (and tungsten). After dilution to
200 ml. and addition of 40 ml. of HNO a (sp.gr. 1.42), the solution is boiled for 35 to 45
min., cooled, and titrated potentiometrically with FeS04.

1460

CHEMICAL ANALYSIS OF IRON AND STEEL-

FeS04 (0.03 N approximately) (1 ml. will reduce 0.0015 g. of vanadium) if

vanadium is under 0.8%, and proportionately- more if the percentage exceeds
this amount. If enough has been added, a drop of the solution will give immediately a blue color with a drop of fresh KaFe(qN)6 solution. Stir the solution thoroughly, add 8 ml. of a freshly prepared solution of (NH 4)2S20s (15%),
and stir for 1 min. Titrate with the KMn04 (0.03 N) to a definite pink tint'
which does not fade upon continued stirring for 11 min. Subtract the same
blank as was determined in the determination of chromium in chromium
steels. 59
If a correction for the end point has not been determined in a prior determination of chromium, or if the chromium value is unknown and an empirical
correction is therefore impossible, the blank shall be obtained at this point as
follows: Boil the solution for 10 to 12 min. to destroy the slight excess of KMn04,
cool to room temperature and titrate with the KMn04 solution to the same
tint as was obtained before. The volume required represents the blank. 60
The vanadium may then again be reduced by an excess of FeS04 and the
solution treated with (NH 4)2S20s, etc., as a check on the first. The percentage
of vanadium is calculated by multiplying the corrected volume of FeS04
solution by the vanadium titer of the solution, dividing by the weight of sample,
and multiplying by 100.
Chromium-Tungsten-Vanadium Steels.-Transfer 2 g. of the sample to
a 600-mL beaker, and add 100 ml. of diluted H 2S0 4 (1 : 9). Heat until action
ceases, add 10 ml. of diluted HNO{'(1 : 1) arid 4 ml. of diluted HCI (1 : 1)
and boil the solution gently for about 30 min. or until the tungsten has been
oxidized to yellow tungstic acid. Stir frequently to break up the film adhering
to the sides and bottom of the beaker. Dilute 'to 100 ml., filter through a light
filter paper and wash with diluted H 2S0 4 {I : 200).61 Dilute the filtrate to
300 mI., add 5 ml. of H aP0 4 and then KMn04 until the solution is pink.. From
this point proceed with the addition of FeS04 as described for chromiumvanadium steel.
Carbon Steels (less than 0.05% vanadium).-IT'ransfer a lO-g. sample to a
500-ml. Erlenmeyer flask, make a bicarbonate separation, fuse, and dissolve
the melt as described for the determination of chromium in carbon steels by
the persulfate oxidation method. 62 Acidify with H 2S0 4 and H aP0 4 and complete the titration as with chromium-vanadium steel. As an alternate procedure, 10 g. of the sample may be treated with cupferron as described in the
following section for chromium-nickel steels.
.
High-Chromium High-Nickel Steels.-Treat 10 g. of the'sample (less than
0.05% vanadium) in a 400-ml. beaker with 250 ml. of diluted H 2S0 4 (8 : 92) and
59 According to the Methods of the Chemists of the U. '8. Steel Corporation for the
Sampling and Analysis of Alloy Steels, p. 47 (1921), the correction amounts to 0.02%
+ 1.8% of the percentage of chromium present. For example, if Cr = 1.5%, the blank
due to chromium is 0.018X1.5 or 0.03%, and the total correction is 0.02+0.03 or 0.05%
of vanadium.
'
60 The blank may also be obtained by means of a standard steel of known vanadium
content and containing about the same amount of chromium as the test sample.
61 The vanadium content of the tungstic acid may be determined by igniting, fusing
with N a 2CO 3,' extracting with water and proceeding as described in the colorimetric test
given under the determination of Vanadium in Vanadium Steels by the Electrolytic
Separation Method, p. 1458.
. - 62 See p. 1456.

/'

CHEMICAL ANALYSIS OF IRON AND STEEL

1461

heat gently until action ceases. (For vanadium steels, use a 2-g. sample and
150 m!. of diluted H 2S0 4 (5 : 95).) Cool the solution to 15 C. and add a
goodly amount of paper pulp. Then add dropwise with constant stirring a
cold, freshly prepared 6% solution of cupferron (ammonium nitrosophenylhydroxylamine, C 6 H 6N NO ONH4) until the precipitate just assumes a reddish
brown color. Filter through an ll-cm. paper containing some paper pulp, and
wash 10 to 12 times with cold diluted sulfuric acid (1 : 99). Transfer the
paper and contents to the original beaker, add 20 m!. of HNO a and 10 ml. of
H 2S0 4. Evaporate to fumes of H 2S0 4. Cool, add 10 ml. of HNOa and again
evaporate to copious fumes of H 2S0 4. Cool, dilute to 300 ml. add 3 m!. of
H aP0 4 (85%) and then KMn04 until the solution is pink. From this point
proceed with the addition of FeS04 as described for chromium-vanadium steel.
Cast Iron (less than 0.05% vanadium).-Proceed as with carbon steels.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon steels .
. MOLYBDENUM BY THE a-BENZOINOXIME METHOD 6a
SPecial Solutions Required. a-Benzoinoxime (2%)._:_Dissolve 10 g. of
a-benzoinoxime in 500 m!. of ethyl alcohol. Filter if not clear.
Boric Aci!I (4%}.-Dissolve 4 g. of boric acid in 100 m!. of water.
Procedure. Molybdenum Steels.-Transfer 1 to 3 g. of the sample to a
600-ml. beaker, add 50 ml. of diluted H 2 S0 4 (1 : 6) and warm until action
ceases. Carefully add just enough HNO a to decompose carbides and to oxidize
iron and molybdenum. Add 2 to 4 drops of HF, mix and then add 10 m!. of
the boric acid solution. 64 Boil for a few minutes and filter if the solution is
not perfectly clear. 65
Dilute to 100 ml. with water, cool to 25 0 C., and add sufficient FeS04
(0.5 g. is usually sufficient) to reduce vanadic and chromic acids. Cool to
5 C., stir, and sl6wly add 10 ml. of the a-benzoinoxime solution and 5 ml.
extra for each 0.01 g. of molybdenum present. Continue to stir the solution,
add just enough bromine water to tint the solution a pale yellow, and then
add a few more milliliters of the benzoin reagent. Allow the beaker and contents to remain in the cooling mixture 10 min. with occasional stirring, stir in
a little macerated filter pulp, and filter through a rapid paper. If the first
50 m!. or so are not entirely clear, this portion shall again be filtered. Wash
.the precipitate with 200 ml. of a cold, freshly prepared solution containing 25
to 50 m!. of the prepared reagent and 10 mL of H 2S0 4 in 1000 m!. On standing,
needlelike crystals will appear in the filtrate if sufficient reagent has been
employed.
Transfer the precipitate and paper to a platinum crucible and cautiously
dry. Char, without flaming, and ignite at 500 to 5250 C. Coo~! weigh, and
63 H. B. Knowles, "The Use of a-Benzoinoxime in the Determination of Molybdenum," Bureau of Standards Journal of Research, 9, No.1, July, 1932, Research
Paper 453.
114 With larger samples (5 to 6 g.) of silicon steel it is best to evaporate the solution,
dehydrate, and 1"emove silica before treatment with the benzoin reagent.
65 Alternately, the sample may be dissolved in a mixture of Hel and HN0 3 , HCI0 4
added and the solution evaporated to fumes of HCI0 4, diluted, sufficient H 2SO a added
to reduce the chromium, and, after boiling out the excess 80 2 and filtering off any silica,
cooled and molybdenum then precipitated.

1462

CHEMICAL ANALYSIS OF IRON AND STEEL

repeat the heating until the weight remains ~onstant. Treat the ignited residue
with 5 mi. of NH 40H, digest, and filter through a small paper. Wash well
with diluted NH40R (1 : 99). Ignite the paper and contents in the original
crucible, cool, and weigh. The difference in weifhts represents the Moo a
present.

Tungsten is also precipitated. If present, the a)nmoniacal filtrate shall be

treated as follows: Add 5 mi. of diluted H 2S0 4 (1 : 1) and evaporate to fumes of
H 2S0 4. Cool, dilute to 25 mi. with water, add 1 to 2 mi. of cinchonine solution
(125 g. of cinchonine dissolved in 1000 qJ.I. of diluted HCI (1 : 1)). Digest at
80 to 90 C., preferably overnight. Filter through a tight paper containing
a little paper pulp and wash with cinchonine wash solution (30 mi. 'of the
cinchonine solution diluted to 1000 mI.). Transfer the paper and conttlnts to
a platinum crucible, char the paper and ignite at 750 to 850 C. 'until constant
weight is obtairwd. Cool, weigh, and subtract from the weight of MoO a
obtained previously. In very accurate work any residue obtained here shall
be dissolved and tested for molybdenum by the colorimetric method. 66
With high-molybdenum medium-tungsten steel (8% molybdenulll, 2%
tungsten) the ignited oxides may be weighed, dissolved, and molybdenum
determined by the MoSs-MoOs (after reprecipitation of the sulfide) method.
Tungsten is then obtained by difference. I
If molybdenum is present in very smail amounts (carbon steels), the ignited
a-benzoinoxime precipitate shall be dissolved in NH 40H and the molybdenum
determined colorimetrically. 66
.Cast Iron.-Transfer 1- to 5-g. of the sample to a 600-mi. beaker, treat with
100 mi. of diluted H 2S0 4 (1 : 4) and warm. When action ceases, add HNO a
drop wise until rapid effervescence ceases (usually 2 to 5 mi.) and then add 2 to
3 drops in excess. Evaporate the solution to fumes of H 2S0 4, cool somewhat,
and add 100 mi. of water. Warm until salts are dissolved and filter through
a rapid paper. Wash the paper with warm water. Dilute the filtrate to 150
mi. and cool to 25 C. Add sufficient FeS04 to reduce any chromium or
vanadium which may have been oxidized by the' above treatment, cool to
5 C. and complete the determination as directed (or molybdenum steels.
OPen-Hearth Iron.-Proceed as with molybdenum steels.
Wrought Iron.-Proceed as with molybdenum steels.
MOLYBDENUM BY PRECIPITATION AS 'SULFIDE AND
WEIGHING AS OXIDE
"
Procedure. Molybdenum Steels (Absence of Tungsten).-Transfer 2 to
10 g. of the sample (approximately 0.03 g. molybdenum) to a 600-mi. beaker,
treat with 100 mi. of diluted H 2S0 4 (1 : 5), and warm. When action ceases,
add 20 mi. of a solution of (NH4) 2S 208 (25%) and boil the solution for 8 to 10
min. to oxidize the molybdenum and part of the irbn. Cool somewhat, add 5 g.
of tartaric acid, neutralize with NH 40H, add diluted H 2S0 4 (1 : 1) until
acidified and then 10 mi. in excess for each 100 mi. of solution. Heat to boiling
and pass in a rapid stream of H 2S for 10 min. Dilute with an equal volume of
hot water, and pass in the gas for 5 min. D~~.at-at 50 to 60 C. for 1 hr.
66

See p. 1464.

CHEMICAL
ANALYSIS
OF IRON AND STEEL
.
,

1463

Filter and wash the sulfur and sulfides with diluted H 2S0 4 (1 : 99) saturated
with H 2S.67 Place the paper and precipitate in the original beaker,68 add
5 mL of H 2S0 4 and 20 ml. of HNO a, cover, and heat to fumes of H 2S0 4 , Cool
somewhat, add 10 rul. of HNO a and again evaporate to fumes. If the solution

is not clear and of a light color, .repeat the treatment with HN 0 3

Cool, dilute to 100 mI., and add a slight excess (10 to 12 drops) of a solution
of NaOH (20%). Heat to boiling and set aside for 5 min. Filter and wash
the paper and residue with hot water. Heat the filtrate to boiling and treat
with H 2S for 10 min. Add diluted H 2S0 4 (1 : 1) until acidified and then a
4 mL excess per 100 mL of solution. Treat with H 2S for 5 min. and digest at
50 to 60 C. for 1 hr. Filter through a tight 9-cm. filter paper, and wash
thoroughly with diluted H 2S0 4 (1 : 99) saturated with H 2S.69 Transfer to a
small porcelain crucible, heat carefully until carbon is destroyed and then at
500 to 525 C. until constant weight is obtained. 70 Test the ignited oxide for
impurities by treating with NH 4 0H. If copper is indicated, determine its
amount colorimetrically, calculate to CuO and deduct. If a residue remains,
6lter., wash with water, ignite, weigh, and deduct from the weight of Mo0 3 .n
The factor for molybdenum in MoO a is 0.667.
Tungsten Steels.-Dissolve 2 to 10 g. of the sample (approximately 0.03 g.
of molybdenum) in 100 mL of diluted HCI (1 : 1), cautiously add 20 mL of
diluted HN0 3 (1 : 1), and then boil gently until.the tungstic acid becomes bright
yellow. Dilute to 150 mI., heat to boiling, filter, and wash the residue with
diluted HCl (1 : 9). Reserve the precipitate. Add 15 mL of H 2S0 4 to the
filtrate, evaporate to fumes of H 2S0 4 , cool, and add 100 ml. of water. Digest
until soluble salts are in solution. . If any tungstic acid separates, filter through
a small filter, wash with a little diluted H 2S0 4 (1 : 99), and combine with the
reserved tungstic acid precipitate. Add 5 g. of tartaric acid to the clear filtrate
67 In umpire analyses the filtrate shall be boiled to expel H 2S and its molybdenum
content determined colorimetrically, or the unprecipitated molybdenum (usually not in
excess of 0.5 mg.) shall be recovered as follows: Boil the filtrate to expel H 2S and to
reduce the volume to about 450 ml. Add 20 ml. of the persulfate solution, boil for 8
to 10 min., and then pass in a rapid stream of H 2S for 10 to 15 min. Digest for 1 hr.,
filter, wash, and combine with the main precipitate.
68 An alternate method for routine analyses consists in directly igniting the impure'
sulfide at 500 to 525 C., weighing and then correcting the weight of the ignited residue
as follows: To the crucible add diluted HCI (1 : 1) and heat until the residue is dissolved.
Transfer the solution to a small beaker, dilute to about 100 mI., and add NaOH in slight
excess (10 drops of a 20% solution). Boil for 1 min., allow the precipitate to settle, and
filter. Dissolve the precipitate in diluted HCI (1 : 1), and precipitate with NaOH as
before. Wash the paper and precipitate thoroughly with water to remove alkali salt;
ignite the paper and precipitate in the original crucible, and weigh. The difference
between the two weights represents MoO a.
69 Precipitation of molybdenum is usually complete, but it is well to test the filtrate
by boiling to expel H 2S, oxidizing with bromine water, boiling to expel bromine, and again
gassing with H 2S.
70 Molybdenum oxide volatilizes at temperatures above 5000 C. but the rate is very
slow at temperatures below 600 0 C. The heating may be done in a muffle with pyrometer, or in a "radiator." This is a 50-mL porcelain crucible containing a disk of
asbestos board, ,4 mm. thick on the bottom and fitted with a nichrome triangle which is
bent to fit the inside of the crucible and supported by bending the end wires over the rim.
The crucible shall be placed so that the bottom is 8 cm. above the top of a Tirrill burner
and heated by a flame 12.5 cm. high.
71 If small amounts of molybdenum are in question, the determination shall be
checked by testing the solution of the oxide as described in the colorimetric method.
0

1464

CHEMICAL ANALYSIS OF IRON AND STEEL

and neutralize with NH 40H. Add H 2S0 4_u ntil aciqified, then 5 mI. per 100 mI.
in excess, and pass in H 2S as in the absence of tungsten.
Some molybdenum is always carried down by ,the tungstic aaid and shall
be recovered as follows: Dissolve the combined turigstic acid residues in a hot
solution of NaOH (5%), and wash the papers with ~ little water and then with
a little hot diluted H 2S0 4. Add 5 g. of tartaric acid, then H 2S0 4 until the
sblution contains 5 ml. per 100 mI., and precipitate with H 2S as in the absence
of tungsten. Filter, wash, combine with the main sulfide precipitate, and
complete the determination as described for molybdenum steels.
Cast Iron.-Transfer 2- to 5-g. of the sample to a 600-ml. beaker; treat
with 100 mI. of diluted H 2S0 4 (1 : 4), and warm. When action ceases, add
diluted HNO a (1 : 1) dropwise until rapid effervescence ceases (usually 5 to
10 ml.) and then add three to five drops in exces_s. - Evaporate the solution
to fumes of H 2S0 4 Cool somewhat, add 100 mI. of warm water, stir, and heat
until salts are dissolved. Filter through a rapid filter paper catching the
filtrate in a 600-ml. beaker. Wash the paper well with hot water. To the
filtrate add 5 g. of tartaric acid and neutralize the solution with NH 40H:!
Add 10 ml. of diluted H 2S0 4 (1 : 1) for each 100 ml. of solution, heat to boiling,
and pass in a rapid stream of H 2S for 10 min. Dilu~ with an equal volume of
hot water, and pass in the gas for 5 min. Digest at 50 to 60 C. for 1 hr.,
filter, and complete the determination as described for molybdenum steels.
Carbon Steels.-Proceed as with molybdenum steels.
OPen-Hearth Iron.-Proceed as with molybdenum steels.
Wrought Iron.-Proceed as with ~olybdenum steels.
MOLYBDENUM BY THE COLORIMETRIC METHOD
Special Solutions Required. Sodium Thiocyanate (5%).-Dissolve 50 g.
of the salt in 1000 ml. of water.
Stannous Chloride.-Transfer 350 g. of SnCl 2 ,2H 2 0 to a SOO-ml. Erlenmeyer flask, add 200 ml. of diluted HCI (1 : 1), ahd boil gently until solution
, is practically complete. Transfer the solution to a I-liter bottle, dilute to
1000 m!. with freshly boiled water, add a few pieces of metallic tin, and stopper.
Butyl Acetale.-Saturate technical butyl acetate with NaCNS and SnCh
by shaking, and keep in a qark bottle.
Standard Molybdenum Solution.-Dissolve 0.5 g. of pure Na 2 Mo0 4 2H 20
in 1000 ml. of water containing 5 ml. of H 2S0 4, and standardize by reducing
100 ml. of the solution in a Jones reductor and titrating with KMn04. Each
1 ml. of the solution should contain approximately 0.0002 g. of molybdenum.
The standard- solution may also be made from National Bureau of Standards'
standard sample 71 of calcium molybdate.
_
Procedure.-Transfer 0.5 g. of the sample (for steels and irons containing
0.25% or less of molybdenum) to a 150-ml. Erlenmeyer flask. (For alloys
containing 0.25 to 0.50% of molybdenum use 0.2 g. of the sample, and if over
0.5% 0.1 g. of the sample.) Add 20 ml. of HCI0 4 (60%) per 1 g. of sample and
warm until the sample has dissolved. Heat to Roiling, cover, and fume until
all carbonaceous matter has been destroyed, __ Goal somewhat, add 25 ml. of
water, and boil for a few minutes to expel free clliorine.

CHEMICAL ANALYSIS OF IRON

. AND STEEL
,

1465

Add.2 g. of tartaric acid and a slight excess of a solution of NaOH (10%).

Heat at about 80 0 C. for a few minutes. Remove from the source of heat,
neutralize with diluted H 2S0 4 (1 : 1) and then for each 8 ml. of solution add
an excess of 2 ml. of the H 2S0 4 (1 : 1), which will give a solution of 10% H 2S0 4
by volume. Cool to room temperature and transfer the solution to a 250-ml.
separatory funnel, rinsing out the flask with H 2S0 4 (10%). Add 10 mI. of a
solution of NaCNS (5%). Shake vigorously for 30 sec., treat with 5 ml. of
the SnCl 2 solution and shake well for 1 min. Cool to room temperature, add
20 ml. of treated butyl acetate, or 50 ml. of treated ethyl ether,72 stopper, and
shake well. Allow the extract to separate and draw off the lower layer which
may be discarded. Transfer the extract 73 into a Nessler, Julian, or Mlori":
meter tube (for example, Kennicott) and compare with a standard similarly
carried through all steps of the procedure. 74 The standard steel may be prepared by adding portions of a standard molybdenum solution to a molybdenumfree steel, or a'standard molybdenum steel may be used.
TUNGSTEN BY THE ACID DIGESTION-CINCHONINE METHOD
Special Solutions Required. Cinchonine.-Dissolve 125 g. of cinchonine in
1000 ml. of diluted HCI (1 : 1).
Cinchonine Wash Solution.-Dilute 30 ml. of the above solution to 1000 ml.
Procedure. Tungsten Steels.-Transfer 2 g. (for steels containing less than
5% of tungsten use 5 g.) of the sample to a 400~ml. beaker,75 cover, and treat
with 50 ml. of HeI. Warm gently. When decomposition' is complete, cease
heating, and scrub with a policeman to detach carbides and tungsten. Gradl1allyadd 10 ml. of diluted HNO~ (1 : 1).76 Digest at 100 C. with occasional
stirring until the tungstic acid is bright yellow and free from black particles.
Dilute to 150 ml., add 5 ml. of cinchonine solution and a small amount of paper
pulp. Digest at 90 0 to 95 0 C. for 30 min. or longer as t1.e solution is stirred
occasionally.77 Decant the clear solution through an ashless paper containing
a little ashless paper pulp. Wash by decantation with two or three 30- to
40-ml. portions of hot cinchonine washing solution, transfer the residue to the
paper,78 and wash the paper and residue thoroughly with the washing solution.
72 For satisfactory color comparisons the concentration of molybdenum should not
exceed 0.05 mg. per milliliter of solvent.
73 On very low molybdenum material it is of advantage to add another portion of
NaCNS and SnCh, mix in the separatory funnel and again extract.
7qf the extract is turbid, it should be filtered through a layer of glass wool.
75 Smooth (unetched) beakers should be used because tungstic acid has a tendency to
stick rather tenaciously to the vessel.
,
76 Less vanadium in vanadium steels is retained in the tungstic acid when iron and
tungsten are oxidized with H 20 2 If the latter is to be used add to the partially cooled
solution 6 ml. of perhydrol (30%) diluted with 14 ml. of water and boil gently until the
volume has been reduced to about 25 ml. Dilute to 200 ml., add cinchonine, etc.
77 Small amounts of tungsten separate slowly in the presence of considerable iron.
Consequently, with less than 2% of tungsten, let the solution stand for 18 to 24 hrs. It
has been stated that a more rapid precipitation of tungsten by cinchonine results if the
iron is reduced to the bivalent state. This may be obtained conveniently by treating
with sulfur dioxide.
78 Dissolve any W0 3 that still adheres to the beaker, by adding a few drops of
NH 40H. EVaporate just to dryness and then add 2 to 3 ml. of diluted HCI (1 : 9) and
0.5 ml. of cinchonine solution. Heat. to boiling, digest a few minutes, and pour onto
the filter.

1466

CHEMICAL ANALYSIS OF IRO:N" AND STEEL

Transfer to a weighed platinum crucible, and ignite at as low a tem~erature

as possible until carbon is gone. Treat with "1 ot 2 drops of diluted H 2S0 4
(1 : 1) and 1 to 3 ml. of HF. Evaporate to dryness, ignite at 7500 to 850 0 C.,
cool, and weigh as "impure WO s." Add 4. g. of Na 2 CO S , and heat gradually
until fusion is complete. Cool, take up the melt in \100 ml. of water, and' filter.
Thoroughly wash the crucible and residue, and reserve the filtrate. Transfer
the residue to the crucible, ignite, fuse with 1 g. of Na2CO S, cool and take up in
25 ml. of water. Filter and thoroughly wash the crucible, paper, and residue
with .hot water. Combine the filtrate with' the reserved "filtrate. A.gain
transfer the residue to the crucible and ignite. Cool, weigh, correct for the
residue obtained from 5 g. of Na 2COs,79 and subtract the corrected weight
from the weight of " impure WO s."
If the combined filtrates show a yellow color, evaporate to 100 m!. and
determine chromium colorimetrically (as under the determinatiop. of chromium
in carbon steels). Calculate to Cr 20s, and deduct from the weight of " impure
WO s." Divide the solution into three equal parts. In one aliquot portion
determine molybdenum oxide, MoOs, by the colorimetric method. In the
second aliquot portion determine vanadium pentoxide, V 20 5, by the colorimetric
method,80 adding one-third as much chrom~te to the comparison solution as was
found in the colorimetric test. In the. third aliquot portion separate tin,
tantalum, columbium,sl and the like by aCidifying, adding NH 4 0H in moderate
excess and boiling. . Filter, wash thoroughly, ignite, and weigh. Add the
weights of the oxides found in the three aliquot portions, multiply by 3, and
subtract from the weight of the" impure WO s."
The fip.al corrected weight of WO s multiplied by the factor 0.793, divided
by the weight of sample taken, and multiplied by 100, represents the percentage
of tungsten in the steel.

Carbon Steels and Other Steels Containing less than 0.2% Tungsten.Transfer 5 g. of the sample to a 600-ml. beaker, add 75 ml. of dih-\ted H 2S0 4
(1 : 6) and warm until action ceases. Carefully add just enough diluted
HNO s (1 : 1) to decompose carbides and to oxidize;the iron. Add about 5 mg.
of molybdenum (conveniently, '25 ml. of tlie sodium molybdate solution).82
Dilute to 150 mI., cool to 50 C. and precipitate with a-benzoinoxime. Determine tungsten subsequently in the mixed o~ides as directed by precipitation
with cinchonine as described for tungsten st~els.
.
COBALT BY THE zrNC OXIDE-a-NITROSO-,B-NAPHTl!PL METHOD
Special Solutions Required. Zinc Oxide Suspension.-Add 50 g. of the
finely powdered reagent to 300 ml. of water and shake thoroughly.
79 The residue obtained from the c.p, grade or Na 2CO S usually exceeds 1 mg.
In
analyses of 2-g. samples of a high-speed tool steel (18.23% tungsten, 3.51 % chromium,
and 0.97% vanadium), the correction for the impurities in the Na.CO a averaged 1 mg.
and the impurities in the "impure WO s " averaged' 5.8 mg. (representing 0.23% of
tungsten). The weights of Fe.O., Cr.Os, and V.05 that were obtained ranged from 1.4
to 4.4 mg., 0.4 to 0.5 mg., and 1.4 to 2.7 mg., respectively.
80 See the Determination of Vanadium in Vanadium Steels by the Electrolytic
Separation Method, p. 1458.
l
81 If a search is to be made for tantalum and columbium, the residue insoluble in
Na.CO s should also be examined, because part of theSi[may be contained in it.
82 See the standard molybdenum solution described under the determination of
Molybdenum by the Colorimetric Method, p. 1464.

CHEMICAL ANALYSIS OF IRON AND STEEL

1467

a-Nitroso-{3-Naphthol.-Dissolve 1 g. of the dry reagent in 15 mI. of glacial

acetic acid and filter.
Procedure. Cobalt Steel.-Transfer 1 g. of the sample to a 400-mI.
beaker, add 25 mI. of diluted HCI (1 : 1), heat, and when decomposition is
comp'lete, add 5 mI. of diluted HNO a (1 : 1) to oxidize the iron. If tungsten is
present the digestion with HNOa-HCI shall be continued until all of the tungsten
has been converted to yellow tungstic acid. Evaporate until salts begin to
separate (about 5 mI.). Add 100 mi. of hot water, and digest on the steam bath
for about 5 min. Dilute the solution to about 200 mI., and add a freshly prepared suspension of zinc oxide in portions of about 5 mI. until the iron is precipitated and a slight excess of zinc oxide is present. Shake thoroughly after'
each addition of the precipitant and avoid a large excess. When sufficient zinc
oxide has been added, further addition of ,the reagent causes the brown precipitate to appear lighter in color upon thorough shaking. A sufficient excess
is also indicated by a slightly white and milky supernatant liquid. Allow the
precipitate to settle for a few minutes and filter the solution through a 12.5-cm.
rapid filter paper.sa Wash the beaker and the precipitate on the filter three
times with cold water. Reserve the filtrate and washings. When the filter
has drained, transfer the paper and precipitate to the beaker in which the
precipitation was made, add 12 mI. of HCI, and stir the paper to a pulp. The
iron should now be in solution; if it is not, add more HCI, avoiding a large excess.
Dilute the solution to 200 mI., and repeat the precipitation with zinc oxide.
Filter on a 15'-cm. paper, and wash four or five times with cold water.
For routine work, a single precipitation will often suffice. In this case
take a 2-g. sample, dilute the solution to exactly 500 m!. after the addition of
zinc oxide, mix thoroughly, and filter through a dry filter into a 250-mi. measuring flask (equals I-g. sample). With one precipitation by zinc oxide, the percentages obtained for cobalt in high-speed steels will be from 0.1 to 0.3 too low
as the result of retention of cobalt by the bulky precipitate.
To the combined filtrates and washings, from the zinc oxide separation,
add 10 mi. of HCl, and adjust the volume to about 400 mi. Heat the solution
to boiling, add 8 ml. of a-nitroso-{3-naphtholllolution plus 3 ml. in addition for
every 0.01 g. of cobalt present. Allow the solution to cool for 30 min. or more,
and filter through a rapid filter paper. Transfer all of the precipitate to the
filter and wash with hot diluted HCI (1 : 3) and then thoroughly with hot water.
Transfer the wet paper and precipitate to a weighed porcelain crucible,
heat gently at first, preferably in a muffle furnace, and finally ignite to constant
weight at 750 0 to 8500 C. Heating above 900 0 C. has a tendency to convert
C0 30 4 to CoO. Cool, and weigh as C0 3 0 4 which contains 73.4% of cobalt.
In very accurate work in which more than 0.01 g. of cobalt is involved, the
oxide shall be reduced in hydrogen, cooled in an atmosphere of hydrogen, and
the cobalt weighed as metal. 84
Nickel accompanies cobalt almost completely in the zinc oxide separation.
Hence, in very accurate work, when nickel predominates, or much of it is
83 A little finely divided zinc oxide may pass through the paper at first.
This is
unobjectionable, because zinc is not precipitated by a-nitroso-{1-naphthol.
84 With high molybdenum or copper steels (over 1%), the ignited Co a0 4, or cobalt
metal may contain small amounts (approximately 0.5 mg.) of these elements. Suitable
corrections may be made after solution of the residue and colorimetric determinations
of the contaminants.

1468

CHEMICAL ANALYSIS OF IRON AND STEEL

present, the ignited cobalt oxide should be dissolved in HCI and cobalt again
precipitated with a-nitroso-{1-naphthoI.85
I
A blank shall be taken through all steps of the determination. it. I-g.
sample of National Bureau of Standards' standard sample 10d of bessemer steel,
or sample 50a of chromium-tungsten-vanadium,ste~l is satisfactory for this
purpose. It is especially important that the same quantity of the a-nitroso-{1naphthol be used in the blank run as in the determination. A little macerated
paper added to the blank after the a-nitroso-{1-naphthol 'reagent facilitates
filtration and washing.
Carbon Steels (and Other Steels Containing less than 0.10% Cobalt)..Dissolve 10 g. of the sample in HCI and cautiously oxidize with just enough
HNO a Extract the iron with ether and wash the ether extract once with
diluted HCI. Warm the ether-extracted acid solution to expel residual eth~r,
and oxidize with KCI0 3 Dilute to 200 mI., and precipitate twice with ~inc
oxide as directed for cobalt steels. In material containing very little cobalt,
it is advantageous to combine the extract acid solution obtained in e~her
separations of a number of separate lO-g. samples.
..
Cast Iron.-Proceed as with carbon steels.
Open-Hearth Iron.-Proceed as with carbon steels.
Wrought Iron.-Proceed as with carbon, steels.
TITANIUM

BY

THE

CUPFERRON-COLORIMETRIC

METHOD

Special Solutions Required. Cupferron (6%).-Dissolve 2 g. of the salt

in 35 m!. of cold water. This reagent shall be prepared as needed.
Standard Titanium Solution.-The standard titanium solution is usually
made from potassium titanium fluoride, K 2 TiFs H 2 0, as follows: Recrystallize
the c.p. salt twice from hot water in a platinum dish. Filter, dry at room
temperature, and preserve in a glass-stoppered bottle. Transfer 2.70 g. to a
platinum dish, add 100 mL of H 2S0 4, and evaporate 3 times to strong fumes of
the acid, each time washing down the inside of the d;~h with water. Cool, and
pour the solution rapldly and with 13tirring into 900 mL of water. If the solution
is properly made, it 'lhould be perfectly clear. 'Cool to room temperature, and
transfer to a glass-stoppered bottle. Standardize by taking 50-m!. portions,
precipitating with NH 40H in a boiling 250-mL solution, filtering, igniting at
about 1200 C., and weighing as Ti0 2 Each 1 ml. of the solution should
contain 0.5 mg. of titanitIm.

Procedure. Titanium Steels.-Transfer 0.5 to 1 g. of the sample to a 400m!. beaker, add 100 mL of diluted HCI (1 : 4), cover, and heat gently until
action ceases. Cool the solution to 15 to 2Qo C., and add dropwise a 'cold,
freshly prepared, solution of cupferron (6%), wit!J. constant stirring, until the
precipitate just assumes a reddish brown color. Additional cupferron only
causes more iron to be precipitated. Add a goodly amount of ashless paper
pulp and filter through an ll-cm. rapid filter paper. Wash 12 to 15 times with
cold diluted HCI (1 : 9). Transfer the paper and residue to a 50-m!. platinum
crucible, dry, and ignite at a temperature (less than 5000 Q.) just sufficient to
85 Tests on a 0.5-g. sample of a steel containing.J.o% of pobalt and 6% of nickel
.
showed but 0.1 mg. of nickel in the first precipitate.

CHEMICAL ANALYSIS OF IRON AND STEEL

1469

destroy the carbon of the filter paper. Fuse the contents of the crucible with
1 g. of K 2S 20 7 and dissolve the cooled melt in 25 m!. of diluted H 2S0 4 (10 : 90).
Transfer the solution to a Camp or a Nessler comparison tube, and dilute
to a suitable volume containing 10 m!. of H 2S0 4 per 100 m!. Add H aP0 4
dropwise until any iron color is destroyed. To a second tube, add reagents
K 2S0 4, H aP0 4, etc. in amounts that are present in the unknown, and dilute
with diluted H 2 S0 4 (5 : 95) to the same volume. Cool both to room temperature. Add 3 m!. of H 2 0 2 (3%) to each and then a standard solution of Ti(S04)z
to the blank run and a similar volume of diluted H 2S0 4 (5 : 95) to the unknown
until the colors match when the solutions are well mixed. The unknown then
contains the amount of' titanium that was added to the blanls: run. If a
colorimeter of the plunger type is used, a suitable standard solution of peroxidized titanium may be prepared and the colors matched by changing the depths
of the columns that are viewed. The amounts of titanium in the two solutions
are then inversely proportional to the heights of the columns.
If the steel contains an appreciable amount of copper, filter the original
H 2SO.4 solution of the sample through a small filter paper containing some
paper pulp. Wash well with hot diluted H 2S0 4 (1 : 9). Cool the filtrate to
15 to 20 C. and precipitate with cupferron as described in the first paragraph.
Transfer the paper containing the acid insoluble material to a 250-m!. beaker,
add 25 m!. of diluted HNO a (3 : 7) and heat until the copper has dissolved.
Add 50 m!. of hot water and a slight excess of NH 4 0H. Heat to boiling, filter,
and wash the paper and precipitate with hot water. Burn off the paper at as
low a temperature as 'possible and add the residue to the ignited cupferron
precipitate. Fuse the combined residues with K 2S 20 7 , dissolve in diluted
H 2S0 4 (1 : 9) and determine calorimetrically as previously described. If the
steel contains vanadium (more especially when small amounts of titanium are
sought), transfer the ignited cupferron precipitate to a 100-m!. platinum dish,
add 5 m!. of HF (48%), 10 ml. of HCl0 4 and evaporate to a volume of 5 m!. or
less. Cool somewhat, dilute to 50 ml. and add an excess of 5 m!. of the NaOH
(10%). Boil for several minutes, let settle, and filter on a tight 9-cm. filter
paper. Wash the paper and precipitate with hot water, ignite, fuse with a
small amount of K 2S 20 7, and complete the determination for titanium as
described above.

Carbon Steels, OPen-Hearth Iron, and Wrought Iron (less than 0.05%
titanium).-Treat 5 g. of the sample with 150 ml. of diluted HCI (1 : 4) and
proceed as described for titanium steels.
Cast Iron (total titanium).-Treat 5 g. of the sample with 100 ml. of diluted
HCI (1 : 2), cover, and warm. When all action has ceased, cool to 10 C.,
add 1 m!. of a solution of cupferron (6%), filter on a paper of close texture, and
wash the insoluble matter with water. Transfer the paper and residue to a
platinum crucible, dry, and ignite under good oxidizing conditions in an uncovered crucible until all carbon is gone. Treat with 1 to 2 m!. of HF and 1 m!.
of diluted H 2S0 4 (1 : 5), and evaporate to dryness. Fuse the residue with 1 to
2 g. of NazCO g Dissolve the melt in about 50 m!. of water, digest for 15 min.
at 90 to 95 C., filter, and wash with water. Ignite the residue in platinum,
and fuse with 1 to 3 g. of K ZS Z0 7 Cool, treat the melt with 25 m!. of diluted
H 2S0 4 (1 : 9), transfer to a Camp or Nessler comparison tube, and determine
the titanium calorimetrically as described for titanium steels.

1470

CHEMICAL ANALYSIS OF IRON ANI) STEEL

ZIRCONIUM BY THE CUPFERRON-PHljSPHATE, UETHOD
J

Procedure. Zirconium Steels.-Transfer 2 to 3 g. of the sample to a

250-ml. beaker, add 100 m!. of diluted HCl (1 : 4), cover, a~d heat until action
ceases. Dilute to 150 ml., cool to 15 to 20 C., precipitate with cupferron,
filter, and fuse the ignited precipitate with K 2S 20 7 as direc,ted for titanium in
titanium steels. Dissolve the cooled melt in 10::! ml. of diluted .H 2S0 4 (1 : 9),
and filter through a small paper to separate any siliceous matter that may be
present. Wash with diluted H 2S0 4 (1 : 9). To the filtrate add 2 ml. of ~02
(30%) and a solution of 3 g. of (NH 4)2HP0 4 in 25 ml. of water. Stir vigorously
and let stand for 1 to 2 hr. at a temperature of 60 to 65 C. An excess of
,Peroxide must be present at all times, and with amounts of zirconium, under
0.01 % the solution should be allowed to stand at room temperature overnight.
Filter through a 9-cm. filter paper containing ashless pulp, and wash thoroughly
with a cold solution of NH 4 NO s (5%). Transfer the paper and residue to a
platinum crucible, and ignite very carefully so that the paper chars but does
not flame. When the paper has charred, gradually increase the temperature
until all carbon is gone, and then heat at about 1050 C. for 15 min. Cool in a
desiccator and weigh as ZrP 20 7
In very accurate analysis, the pyrophosphate shall be tested for titanium
by fusing with 4 to 5 g. of K 2S 20 7, and dissolving the melt in 40 ml. of diluted
H 2S0 4 (1 : 9) containing 5 ml. of H 20 2 (3%). If titanium is present, it is
determined colorimetrically, calculated to Ti 2P 2 0 9, and deducted from the
weight of ZrP 2 0 7

Carbon Steels, Open-Hearth' Iron, Wrought Iron and Cast Iron.-Treat

5 g. of the sample with 150 ml. of HCl (1 : 4) and proceed as for zirconium steels.

OTHER METHODS

86'

VOLUMETRIC DETERMINATION OF MAGNANESE":""AFTER

PERIODATE OXIDATION 87
Chloride must be absent. 'Cobalt interferes as in the bismuthate method.
When periodate is added to a solution of cobaltous sulfate, the solution becomes
dark brown, owing to oxidation of the cobaltous salt. In steel containing 1.13%
of manganese and 0.5% of cobalt, 1.32% of manganese-:was obtained. Cerium
also interferes because it is oxidized to a eerie salt.
Procedure for Steel or Iron, Free from Chromium.-Dissolve 1 gram of
steel containing not more than 0.15 mg. of chromium in 4 nil. of concentrated
sulfuric acid and 25 ml. of water. To the hot solution; add, cautiously, 1 m!.
-'-

Selected from the literature by the editor.

81 Willard and Thompson, Ind. Eng. Chern., Anal. Ed., 3, 399 (1931).

86'

CH.EMICAL ANALYSIS OF IRON AND STEEL

1471

or more of concentrated nitric acid to oxidize the ferrous ir.on and carbonaceous
matter, and boil to remove nitrous fumes. Any graphite present will do no
harm and will be filtered out later. Dilute to 50 or 75 mI., add 3 m!. of 85%
phosphoric acid (or an equivalent amount of more dilute acid) and 0.3 gram of
sodium or potassium periodate. Boil gently for 15 minutes to oxidize the
manganese, dilute to 150 mI., cool to room temperature, and add slowly, with
constant stirring, 4 or 5 grams of mercuric nitrate (Hg(NO ah2H 20) dissolved
in a little water. Filter immediately through a fairly large asbestos filter into
an excess of standard ferrous sulfate, wash with cold water four or five times,
and titrate back with standard permanganate.
Procedure for Steel or Iron Containing Not Over 0.1 % of Chromium.Dissolve 1 gram in a mixture of 15 mI. of water and 15 mI. of 85% phosphoric
acid. Oxidize the ferrous iron and carbonaceous matter by adding carefully to
the hot solution 1 or 2 mI. of concentrated nitric acid, and boil to remove nitro'us
fumes. Dilute to 100 mI., add 0.3 gram of sodium or potassium periodate, boil
gently 15 minutes to oxidize the manganese, dilute to 150 mI., and cool to
room temperature .. Precipitate the perioQ.ate by adding slowly, with constant
stirring, 2 to 3 grams of mercuric nitrate dissolved in a little water, and filter
immediately through a fairly large a~bestos filter into excess of standard ferrous
sulfate containing 10 mI. of 50% sulfuric acid to prevent precipitation of
mercuric phosphate. Wash with cold water four or five times and titrate back
the filtrate and washings with standard permanganate. The filtration will be
slower than when sulfuric acid is present during precipitation, but the time
required, including washipg, should not be more than 3 minutes.
Procedure for Iron Ore and Other ~)Xide Ores.-Place 1 gram in a 250 mI.
Pyrex beaker, and add 15 mI. of 85% phosphoric acid and a few drops of con- .
centra ted sulfuric acid. Stir until all the ore is free from the bottom of the
beaker, cover, and heat until fumes of suffuric acid are given off, taking care
that it does not froth over or cake on the bottom. If the ore is not dissolved
by this time, keep it hot for a longer time, stirring occasionally, but not allowing
the temperature to rise, because the beaker would be attacked. Cool until the
mass begins to be viscous, then add quickly 100 mI. of water, ~nd heat. Everything should dissolve except some gelatinous silica. From here on the procedure is the same as for steel containing 0.1 % of chromium.
Procedure for Bronze.-Dissolve 1 gram in a mixture of 3 ml. of concentrated nitric acid, 10 mI. of 85% phosphoric acid, and about 7 or 8 mI. of water.
Then dilute to 50 or 75 mI. and proceed as above. Owing to the deep blue color
of the copper salt, the back titration with permanganate is conveniently carried
out electro metrically or by use of alphazurine indicator.
NOTE.-If the amount of manganese is between 15 and 30 mg., the amount of periodate in the above procedures should be increased to 0.5 gram.

A MODIFIED PERSULFATE-ARSENITE METHOD .FOR

MANGANESE 88
The permanganic acid formed by the oxidation of the sample with ammonium persulfate, in the presence of phosphoric acid with silver as catalyst, is
88 E. B. Sandell, I. M. Kolthoff, and J. J. Lingane, Ind. and Eng. Chew., Anal. Ed.,
7, 256, July 15, 1935.

14'72

CHEMICAL ANALYSIS OF IRON AND STEEL

titrated with a reducing solution containing equivalent amounts of sod~m

arsenite and sodium nitrite instead of with the bustomary sodium arsenite
alone. With the mixed reducing solution, heptavalent manganese is reduced
to the divalent condition, and the solution therefore becomes colorless at the
end point instead of yellow or brown as when arSenite alone is used for the
titration. Small amounts of chromium, vanadium, nickel, and molybdenum
do not interfere. Silver must be precipitated as the chloride before the titration
can be made.
'
Procedure for Manganese in Steels. Reagents.-(l) Acid mixture (recommended by Bright and Larrabee): Water, 525 ml.; sulfuric acid, concentrated
(sp.gr.= 1.84), 100 ml.; phosphoric acid, 85%, 125 ml.; and nitric acid, concentration (sp.gr. = 1.42), 250 ml. First add the sulfuric acid to the water, cool,
and then add the nitric acid and phosphoric acids. (2) Silver nitrate solution,
0.1 M. (3) Ammonium persulfate solution, 25 grams (of 95% salt) in 80 ml.
of water. Ammonium persulfate slowly decomposes in solution and therefore
this solution should not be kept for more than 2 or 3 days. (4) Sodium chloride
.
solution, 0.2 M. (5) Sulfuric acid, l2 N.
Standard Solutions.-(l) Sodium arsenite-sodium nitrite, 0.05 N. Dissolve
2.5 grams of pure arsenic trioxide in 25 mL of 4 :N" sodium hydroxide solution,
dilute to 200 mI., add slightly more than enough dilute sulfuric acid to neutralize
the hydroxide, and then a slight excess of sodium bicarbonate to neutralize the
acid (the solution should finally be neutral to litmus paper). Dissolve 0.85
gram of sodium nitrite in the solution thus obtained and dilute to 1 liter. (2)
Potassium permanganate, 0.05 N. Prepare and staI:\dardize against pure dry
sodium oxalate in the usual manner.
Standardization of Sodium Arsenite-Sodium Nitrite Solution.-Pipette
10 mL of standardized 0.05 N potassium permanganate solution (for the best
results the amount of permanganate'taken should correspond approximately to
the amount of manganese in the material under test) into a 250-ml. Erlenmeyer
flask, and add 1.0 gram of electrolytic iron of known manganese content and 30
ml. of the sulfuric-phosphoric-nitric acid mixture./ Warm to hasten solution,
and finally boil for. 2 to 3 minutes to expel oxides of nitrogen. Then add 50 ml.
of cold water, 5 mL of silver nitrate solutidn, and 10 ml. of ammonium persulfate solution. Heat to boiling, and boil fpr 30 to 45 seconds. Immediately
cool the flask and its contents to 20 to 25 O. by placing in water. Add 5 ml.
of sodium' chloride solution and 10 mL of 12 N sulfuric acid. Then titrate with
arsenite-nitrite solution. Run in the reducing solution at a uniform rate not
exceeding 5 or 6 ml. per minute, swirling the liquid in the flask continuously
until the solution has become pale pink, and then continue the addition more
slowly, allowing 5 seconds between drops. When very near the end point as
indicated by the very pale pink color, allow approximately 10 seconds between
drops or fractions of a drop. This slow additiqn is required only for the last
few drops. The end point is reached when the color changes abruptly to white
(suspended silver chloride). If the titration has been correctly performed, a
drop of 0.05 N permanganate solution added after the end point has been
reached will give a coloration persisting for at least 3 minutes.
Determination.-Dissolve a 0.9- to 1.1-gram sample' of steel in 30 mL of
the acid mixture and then proceed as directed iff the standardization (beginning
}Vith the second sentence).


CHEMICAL ANALYSIS OF IRON AND STEEL

1473

NOTEs.-If an accuracy greater than 0.01 % of manganese is not required, the addition of iron in the standardization may be omitted. Then standardize as follows:
Transfer 30 ml. of acid mixture to a 250-ml. Erlenmeyer flask and add 50 ml. of
water, 5 ml. of silver nitrate solution, and 10 ml. of ammonium persulfate solution. Boil
for one minute, cool to room temperature, add 10 ml. of 0.05 N standard potassium permanganate, precipitate the silver, and titrate with the arsenite-nitrite solution as described above.
An equivalent amount of pure ferrous or ferric sulfate (but not ferrous ammonium
sulfate or ferric alum) as free as possible from manganese, may be used instead of electrolytic iron in the standardization. In any case the manganese content of the iron or
iron salt added must be determined colorimetrically, preferably with periodate as reagent
(7, 8), and a correction applied if necessary.
When the manganese content of the sample falls outside the range 0.3 to 0.8% (using
a I-gram sample) the arsenite-nitrite solution should be standardized against a volume of
standard potassium permanganate that contains roughly the same quantity of manganese
as the sample. No more than 15 mg. of manganese may be present in the solution when
diluted to 100 ml. previous to titration. In other words, using a I-gram sample, not
more than 1.5% of manganese may be present in the material analyzed if the above
directions are to be followed.
Cast irons may be analyzed by the above procedure if the graphite, left after dissolving the sample, is filtered off before the oxidation.

CHROMIUM AND VANADIUM

89

INDICATOR METHODS

Preparation of Indicator.-A 0.01 M solution of diphenylamine sodium

sulfonate is prepared by dissolving 3.2 grams of the barium salt (obtainable
from the Eastman Kodak Co., Rochester, N. Y.) in a liter of water, adding to
this solution a slight excess of sodium sulfate, and decanting or filtering the
solution.
Preparation of Oxidized Indicator.-The volume of the 0.01 M indicator
solution specified 'in a given experiment is placed in a small beaker, 5 ml. of
water, 3 or 4 drops of concentrated sulfuric acid, and 3 or 4 drops of 0.1 N
potassium dichromate are added, and then very dilute ferrous sulfate (0.01
to 0.02 N) is added until the purple color, which appears on the addition of the
first few drops of ferrous sulfate, just turns to a bluish green. As this purple
color begins to disappear, the ferrous sulfate should be added in parts of a drop,
in order to have no excess present in the oxidized indicator solution. This
bluish green solution is added to the solution to be titrated. In the experiments
described in this paper the oxidized indicator was prepared in separate samples
for each titration.
A stock solution of the oxidized indicator is often more convenient and may
be prepared as follows:
One hundred milliliters of 0.01 M diphenylamine sodium sulfonate and 25
m!. of concentrated sulfuric acid are diluted to 900 m!. in a liter volumetric
flask. To.this solution 25 ml. of 0.1 N potassium dichromate are added slowly
with frequent shaking, followed byO.l N ferrous sulfate with repeated shaking,
until one drop causes a visible change in color from bluish green to a clear deep
green in the liquid when viewed. through the neck of the flask. This will require
approximately 6.5 ml. of ferrous sulfate.
Standardization of Ferrous Sulfate Solution for Vanadium Determinations.
-Weigh out 0.04 to 0.06 gram samples of pure potassium dichromate into
89

Willard and Young, Ind. and Eng. Chem., Anal. Ed., 5, pp.

154, 158 (1933).

1474

CHEMICAL ANALYSIS OF IRON AND STEEL

400 mI. beakers and dissolve in a little water, or measure out that volume of
standard dichromate solution which will require 35 tb 50 ml. of 0.025 N ferrous
sulfate for titration. Add 5 ml. of sulfuric acid (sp.~r. 1.5), 5 ml. of phosphoric
acid (sp.gr. 1.37), and dilute to 200 m!. Then add 0.3 m!. of 0.01 M indicator,
oxidized, and titrate with approximately 0.025 N .f~rrous sulfate. No blank
correction is required in this standardization. 90
,
Procedure for Vanadium in Tungsten Steels.-Add' 25 to 30 mI. of. wa ter
and 5 ml. of sulfuric acid (sp.gr. 1.83) to a I-gram sample of the steel weighed
into a 400-ml. beaker. Warm gently until the steel is completely decomposed
and the tungsten separates out as-a black powder. Rub loose from the beaker
all tungsten and to the boiling hot solution add 5 ml. of hydrofluoric acid
(48%), then nitric acid (sp.gr. 1.42), at first sl<y, using 5 m!. in all., Boil 2
minutes. A clear green solution is obtained. If the solution is not P!'ldectly
clear at this point, as occasionally happens with steels especially difficult to
decompose, boil 2 or 3 minutes longer. Dilute to approximately 100 mI., add
1 gram of ammonium persulfate, and boil 5 minutes. Add 5 ml. of 0.1 N
ferrous sulfate, dilute to between 175 and. 200 ml., and cool to room temperature. Add to this solution 0.1 N potassium permanganate from a buret until
a distinct color persists for 2 minutes, to be sure all the vanadium is oxidized.
At this point either procedure A or B may' be used. Procedure A is preferable.
A. Add 0.05 M sodium nitrite slowly from a buret until the solution becomes green, then 5 ml. excess of the ,nitrite solution, followed by 2 grams of
urea. Stir the solution thoroughly and allow it to stand for 5 minutes.
'
B. Add 5 ml. of 0.1 M sodium azide and boil vigorously (in hood) for 5
minutes to remove all the hydrazoic acid. Cool to room temperature.
To the solution thus obtained add 3 ml. of hydrofluoric acid (48%) and 0.3
ml. of 0.01 M diphenylamine sodium sulfonate which has been oxidized, and
titrate at once with 0.025 N ferrous sulfate standardized as directed below.
The color change of the indicator at the end point is from purple to green. The
correction to be applied for the indicator may be determined from the following
table and should be subtracted from or added to the volume of ferrous sulfate
'
used in the titration. 90
Vanadium in Steel

Indicator Corrections in
0.025 N FeSO.
~

O-{}.85 , ... , ...................... '................ ,., -0.10

~:~5::1~

::::: :'::::::: ::::: ::::::::::::::::: ::::::~>:,;

tg:~g

Standardization of Ferrous Sulfate for Chromium Plus Vanadium Determinations.-Place in a 600 ml. beaker that tlmount of potassium dichromate,
either in solid form dissolved in a little water or in the form of a standard
solution, which will require approximately 35 to 50 ml. of 0.05 N ferrous sulfate.
Add 5 m!. of sulfuric acid (sp.gr., 1.5), 5 m!. of phosphoric acid (sp.gr. 1.37) and
dilute to 300 ml. Then add 0.3 mI. of 0.01 M indicator, oxidized, and titrate
with approximately 0.05 N ferrous sulfate. No blank correction for the indicator is required here.
Procedure for Chromium Plus Vanadium in,Tungste'n Steels.-Add 25 to
30 ml. of water and 5 m!. of sulfuric acid (sp.gr.'1.83) to a I-gram sample of the
90 Willard and Young, Ind. Eng. Chern., Anal. Ed., 5, 154 (1933).

CHEMICAL ANALYSIS OF IRON AND STEEL

1475

steel weighed into a 600-ml. beaker. Warm gently until the steel is completely
decomposed and the "tungsten separates out as a black powder. Rub loose
from the beaker all tungsten and to the boiling hot solution add 5 ml. of hydrofluoric acid (48%), then nitric acid (sp.gr. 1.42), at first slowly, using 5 ml. in
all. Boil 2 minutes. A clear green solution is obtained. If the solution is
not perfectly clear at this point, as occasionally happens with steels especially
difficult to decompose, boil 2 or 3 minutes longer. Dilute to approximately
300 mI., heat to 60 C. or. higher, add 10 ml. of silver nitrate solution (2.5
grams of silver nitrate per liter), 5 grams of ammonium persulfate, and boil
vigorously for 10 minutes. A few small pieces of broken porcelain in the solution will prevent bumping. Anyone of the three following procedures may be
used to destroy the permanganic acid formed during the oxidation process.
Procedure A is preferable.
A. Add 0.1 M sodium azide, drop by drop, from a pipette to the boiling
hot solution until all permanganate has been reduced, limiting the excess of
azide used to 1 or 2 drops. Boil for 1 or 2 minutes to remove the hydrazoic acid.
B. Add 0.05 M sodium nitrite, drop by drop, from a pipette to the boiling
hot solution until all permanganate has been reduced, limiting the excess of
nitrite used to 1 or 2 drops. Then add 1 gram of urea at once and stir the solution thoroughly.
C. Add 5 ml. of 1 to 3 hydrochloric acid to the hot solution and boil vigorously for 10 minutes to reduce the permanganate and remove all chlorine.
To the solution from A, B, or C, cooled to room temperature, add 3 ml.
of hydrofluoric acid (48%) and 0;3 ml. of 0.01 M diphenylamine sodium sulfonate, oxidized, and titrate at o,nce with 0.05 N ferrous sulfate standardized
as directed below. Immediately after the addition of the indicator the solution
is a yellowish brown. As the ferrous sulfate is added this yellow color disappears gradually, being replaced first by a purplish red and then by a clear
purple. The color change at the end point from purple to green is very sharp.
An indicator blank -of -0.30 ml. of 0.05 N ferrous sulfate should be applied
to the volume of ferrous sulfate used in the titration.
The usual procedures for determining chromium and vanadium in steels
that do not contain tungsten may be carried out using oxidized diphenylamine
as the indicator.
Preparation of IndicaJor.-A 1/10% solution of diphenylamine is prepared
by dissolving 0.1 gram of the indicator in 10 ml. of sulfuric acid (sp.gr., 1.83)
and diluting this solution with 90 ml. of glacial acetic acid. The method of
preparing portions of oxidized indicator solution from the 0.1 % solution is the
same as used for diphenylamine sulfonic acid, except that 5 m!. of phosphoric
acid (sp.gr., 1.37) instead of water are added to a measured volume of the
indicator in a very small beaker. No sulfuric acid is added, but only 3 or 4
drops of dichromate, followed by very dilute ferrous sulfate.
Procedure for Vanadium in Steels without Tungsten.-Add 30 to 40 m!. of
water to a 4- or 5-gram sample in a 600 ml. beaker. Run in a measured volume
of sulfuric acid (sp.gr., 1.83) from a burette, allowing 1.5 ml. for each gram 9f
steel and 3 ml. in excess. After the steel has been completely decomposed,
boil until a considerable quantity of salts separates out, in order to assist in
decomposing carbides. Dilute with 30 to 40 ml. of water and heat until the
salts have dissolved. To the boiling hot solution add nitric acid esp.gr., 1.42)

1476

CHEMICAL ANALYSIS OF IRON. AND STEEL

in small portions until the oxidation of ferrous sulfate is complete (3 to 5 m!.

of acid are sufficient). Avoid any appreciable ex~ess. Boil the solution to
destroy oxides of nitrogen, cool to room temperature, add 25 mi. of phosphoric
acid (sp.gr., 1.37), and dilute to 300 m!. Add 0.1 N potass~ permanganate
from a burette until an excess is present. Th,en add 3 or 4 drops more and
let the solution stand for 2 minutes to be sure that t,he color persists and all of
the vanadium is oxidized. Either procedure A or 'B may be used to 'destroy
excess permanganate. Procedure A is preferable.,
,A. Add 0.05 M sodium nitrite slowly from a burette until the solution
becomes green, then 5 mi. excess of the nitrite solution, followed by 2 grams of
urea. Stir the solution thoroughly and allow it to stand for 5 minutes.
B. Add 5 mi. of 0.1 M sodium azide and boil vigorously (in hood) for 5
minutes to remove all hydra zoic acid. Cool to room temperature.
To the solution thus obtained, add 15 grams of crystallized sodium a'cetate.
This should be the correct amount of acetate to react with the 3 mi. excess of
sulfuric acid used in dissolving the steel and with the slight amount of nitric acid
present. If too much acetate is added so that a permanent preCipitate forms,
a drop or two of sulfuric acid will cause the, solution to clear. As soon as the
acetate has dissolved, add 0.5 mi. of 0.1 % diphenylamine, oxidized, and titrate
with 0.025 N ferrous sulfate standardized by the method given below.
Standardization of Ferrous Sulfate Solution for Vanadium Determinations.
-Use sufficient standard potassium dichromate solution to require 35 to 50
mi. of 0.025 N ferrous sulfate. Dilute to 300 mI., add 0.5 mi. of 0.1 % diphenylamine, oxidized, and titrate with the 0.025 N ferrous sulfate.
Recommended Procedure for Chromium Plus Vanadium in Steels Without
Tungsten.-Use a sample of suitable size, varying from 2 grams with low
chromium to 0.25 gram with st.ainless steels. Place it in a 600 m!. beaker, add
15 mi. of water, 15 mi. of phosphoric acid (sp.gr., 1.37), and run in a measured
volume of sulfuric acid (sp.gr., 1.83) from a burette. Allow 1.5 ml, of acid for
each gram of steel and 3 mi. excess. After the steel has'been completely decomposed, boil until a considerable quantity of salts separates out, in order to
assist in decomposing carbides. Dilute with l 20 fiI. of water and heat until
the salts have dissolved. Add nitric acid (sp.gr., 1.42) in small portions to the
hot liquid until the violent oxidation of ferrous sulfate is over (2 to 3 mi. of
acid are sufficient). Avoid any appreciable excess. Boil the solution to destroy oxides of nitrogen, dilute to 300 mI., heat to l)Qiling and add 10 mi. of
silver nitrate containing 2.5 grams of silver nitrate per 1iter,_and 1.5 grams of
ammonium persulfate. If no permanganate tinge appears in -the solution on
boiling add more ammonium persulfate. In no case will more than 2 to 2.5
grams be required if samples of the weights suggested above are used. Boil
the solution for 10 minutes to decompose the excess persulfate, and use either
procedure A or B to reduce the permanganate formed. A is preferable.
A. Add 0.1 M sodium azide, drop by drop, froin a pipette to the boiling hot
solution until all permanganate has 'been reduced, limiting the excess of azide
alided to 2 or 3 drops. Boil for 2 minutes to remove the hydra zoic acid. A
reddish color which may develop in the solution during the boiling period may
be disregarded.
'
B. Add 0.05 M sodium nitrite, drop by dr_op;from a pipette to the boiling
hot solution until all permanganate has been reduced, limiting the excess used

CHEMICAL

A~ALYSIS

OF IRON AND STEEL

1477

to 2 or 3 drops. Then add 1 grain of urea at once and boil the solution for
5 minutes. A reddish color which may develop in the solution during the
boiling period may be disregarded.
To the solution from A or B, cooled to room temperature, add that quantity
of crystallized sodium acetate which is required to react with the excess of
sulfuric acid used in dissolving the steel (1 m!. of concentrated sutfuric acid is
equivalent to 4.8 grams of sodium acetate trihydrate). As soon as this has
dissolved, add 0.5 m!. of 0.1 % diphenylamine, oxidized, and titrate with 0.05 N
ferrous sulfate. The ferrous sulfate should be standardized by the procedure
given under vanadium in steels. without tungsten. That volume of standard
dichromate should be taken which will require about 35 to 50 m!. of ferrous
sulfate.
DETERMINATIONS OF CHROMIUM AND VANADIUM AFTER OXIDATION
WITH PERCHLORIC ACID 91

J?lace samples of 0.5 to 2 graIl_ls, depending on the chromium content, in a

500-ml. tall-form lipped beaker, or better still in a 500-m!. Soxhlet flask, and
add 20 to 25 m!. (depending on the size of the sample) of 70% perchloric acid,
commercial or C.P. Cover the vessel and warm until all the metal has dissolved, taking care not to let the action become too violent. Most steels will
dissolve in less than 5 minutes. Boil the solution 15 to 20 minutes, noting the
change in color if much chromium is present, and continuing the boiling at least
10 minutes longer. The perchloric acid, should condense and run down the
sides of the beaker or flask. Cool somewhat, add an equal volume of water, and
,boil 3 minutes to expel chlorine. After cooling, the solution is diluted to a suitable volume and is ready for titration as describedlater.
.
Cast iron or alloys high in carbon should be dissolved in dilute perchloric acid,
either with or without nitric acid, and evaporated to fumes, because the action
of the hot concentrated acid is too violent. If it is necessary to use 4- or 5-gram
samples, add 50 or 60 ml. of acid. This is desirable in cast iron OJ; steel very
low in chromium. Boiling for 1 or 2 hours may be necessary to oxidize all the
graphite, and further addition of perchloric acid may become necessary.
Procedure for Ferrochromium.-An alloy of this type requries different
treatment, because when poiled with 70% perchloric acid, the chromium trioxide formed, which is fairly insoluble even in the hot acid, tends to coat the
particles of ferrochromium and prevent further oxidation. Weigh 0.2 gram
of the powdered ~lloy into a 500-mI. tall lipped beaker or Soxhlet flask. Add
20 ml. of hydrochloric acid, sp.gr. 1.18; and heat for 15 minutes. Add 15 ml.
of 70% perchloric acid and boil 30 minutes after all hydrochloric acid is expelled.
Cool somewhat, add a few cubic centimeters of water to dissolve the chromium
trioxide and reoxidize by boiling 15 minutes after the water is expelled. Add a
few cubic centimeters of water and look for any undissolved particles of metal.
If any are present, another oxidation may dissolve them. Otherwise pour off
most of the solution, add a few cubic centimeters of 70% per9hloric acid to the
residue, and boil till solution is complete. Add the main decanted liquid,
evaporate off the water, and reoxidize for 5 minutes. Dilute and boil off
chlorine.
91

H. H. Willard and R. C. Gibson, Ind. Eng. Chern, Analyt. Ed., 3, 88 (1931).

1478

CHEMICAL ANALYSIS OF IRON AND STEEL

Procedure for Chromic Oxide and Chromite.-Chromic oxide, which is not

attacked by the usual acids, is readily oxidized to chromic acid by boiling 70%
perchloric acid, a 0.15-gram sample requiring 15 minutes. Ch#imite is more
resistant; if ground to pass a 200-mesh sieve it requires 60-90 minutes.
Weigh a 0.5-gram sample into a tall, 500-ml. lip~ed beaker, or better still,
Soxhlet flask. Add 20 ml. of 70% perchloric acid and boil 30 to 60 niinutes.
Cool somewhat, then add a few cubic centimeters of water to dissolve the
chromium trioxide which coats the particles. If any un attacked ore remains,
evaporate off the water and boil the concentrated acid again for 20 or 30 minutes. If this treatment does not decompose all the ore, pour off most of the
solution, ad~ to the residue 10 ml. of perchloric acid, and boil until the ore is
dissolved. Add the main decanted solution, evaporate off the water, oxidize
5 minutes, dilute, and boil off the chlorine. The solution is diluted to a s~itable
volume and when cool, is ready for titration.
Titration of Chromic Acid.-After the chromium has been oxidized as
already described, it may be titrated by any of the usual methods.
(1) Excess of standard ferrous sulfate is added and the excess is titrated
back by means of standard permanganate, allowing 1 minute for the end point
if vanadium is present. This method is so well known as to need no description.
In the absence of vanadium, the end pointjs made much sharper by adding 0.5 .
ml. of a 0.1 % solution of erioglaucine A, better known as alphazurine (Schultz
506) as recommended by Knop.92 T.lui change from green to red after the
addition of an excess of 0.03 ml. of 0.1 N permanganate in 300 ml. if very' distinct even in the presence of the chromic salt: The color lasts only a few
seconds.
(2) Excess of standard arsenite is added, which is then back-titrated with
permanganate in the presence of hydrochloric acid and a trace of iodide as
catalyst, as described by Kolthoff and Sandell. 93 If vanadium is absent,
alphazurine as indicator improves the end point. In testing this method 2-gram
samples were used, phosphoric acid was not really necessary, and vanadium
was determined in the same solution by titration/with ferrous sulfate using
diphenylbenzidine as indicator, following the ,directions of the authors. Good
results were obtained. The amount of acetate to be added is discussed in
connection with the next method.
(3) The chromic acid (and vanadic acid, too, if present) is titrated directly
with standard ferrous sulfate either electro metrically ina._strongly acid solution,
or by the use of diphen'ylbenzidine as indicator in a solution containing phosphoric acid and sufficient sodium acetate to react with all tJ{e perchloric acid
present, a!lcording to the method of Willard and Young. 94 The procedure is as
follows: After boiling off the chlorine, cool, djlute the perchloric acid solution
to 200 to 400 ml., add 15 ml. of phosphoric acid (sp.gr. 1.37) and, stirring
constantly, a weight of sodium acetate trihydrate equivalent to. the excess of
perchloric acid used in dissolving the steel. On:e gram of iron was found to
require about 5.4 ml. of 70% perchloric acid for conversion into ferric perchlorate, and 1 ml. of this acid contains 1.17 grams of perchloric acid, equivalent
to 1.58 grams of crystallized sodium acetate. Thus, if 20 IJll. of acid are added,
92 Knop, Z., anal. Chern., 77, 125 (1929).
__ _
93 Kolthoff and Sandellt- Ind. Eng. Chern., Anal. :Ed., 2, 140 (1930).
94 Willard and Young, Ind. Eng. Chern., 20, 764, 769 (1928).

CHEMICAL ANALYSIS OF IRON AND STEEL

1479

the excess will require about 23 grams of acetate. If the steel was first dissolved in nitric acid, the excess of perchloric acid will be a little greater. The
proper amount of acetate has been added when more would form a permanent
white precipitate 'of ferric phosphate. If this occurs, add dilute sulfuric acid,
with constant stirring, until the solution clears. Unless the acidity is reduced
in this way, the color- change at the end point will be slow. Add 0.6 to 0.8 ml.
of 0.1 % diphenylbenzidine ~olution (in phosphoric or acetic acid), and allow
5 minutes, but not more than 10, for the purple color to develop. It wiIi
appear brownish until most of the chromic acid is reduced. If no color appears, add a few drops of dilute sulfuric acid. Titrate with ferrous sulfate to a
clear green, approaching the end point carefully. Add a correction of 0.04 m!.
of 0.1 N ferrous sulfate for each 0.5 m!. of indicator.
For the electrometric method add 25 to 30 ml. of sulfuric acid (sp.gr. 1.5)
and titrate chromic plus vanadic acids with ferrous sulfate. If much vanadium
is present, a much sharper end point is obtained by cooling the solution to 5 C.
Whichever method is used, it has already been shown that the ferrous
sulfate should be standardized in the same way and under the same conditions
of volume and acidity at about the time it is used.
Titration of Vanadium.-This may be determined by any of the usual
methods.
Silica has been dehydrated. by the boiling perchloric acid and after the
titration it may be filtered off and determined. 95
CHROMIUM AND VANADIUM IN THE SAME SAMPLE

96

Procedure, Perchloric Acid'Method.-Weigh a sample of suitable size,

varying from 2 grams with low chromium to 0.25 gram with stainless steels,
into a 500 ml. lipped beaker of tall form or into a 500 m!. Soxhlet flask. If
the latter is used, the watch glass should be placed on very thin glass hooks or
on a bent platinum wire about 1 mm. in diameter to prevent the forlP-ation of a
liquid seal whe~ the contents of the flask are heated. Add 20 to 25 ml. of 70%
perchloric acid (depending on the size of the sample) and heat very slowly to
prevent the reaction from becoming violent. After the steel has dissolved
(usually only 3 to 5 minutes will be required) boil the solution for 15 to 20
minutes. The longer period is better for high-chromium steels. Cool the
flask and contents a moment in the air and then in a stream of running water.
Add 25 m!. of water, rinse off the watch cover into the beaker, and boil the
solution for 3 minutes to remove all chlorine. Dilute the liquid to 250 to 300
mL and cool in running water. If the steel contains more than a trace of
vanadium so that two end points are to be determined in the titration with
standard permanganate, 15 ml. of phosphoric acid (sp. gr., 1.37) are added
most conveniently at this point. Add to the solution at room temperature a
measured excess of 0.1 N ferrous sulfate, and then two or three drops of 0.025 M
o-phenanthroline ferrous complex. Much of the indicator precipitates, but
sufficient of it is dissolved for the first end point and all of it dissolves on heating
the solution before the second titration is commenced.
Titrate immediately after the solution of the indicator with 0.05 N potassium permanganate until the color of the solution changes from pink.to a clear
95 Willard and Cake, J. Am. Chern. Soc., 42, 2208 (1920).
il6 Willard and Young, Ind. Eng. Chern., Anal. Ed., 6, 48 (1934).

1480

CHEMICAL ANALYSIS OF IRON AND STEEL

green. The end point is very sharp and there is, almost no return of the pink
color for a minute or more. Add sufficient crystallized i5~dium acetate to react
with the free perchloric acid present. The amount of this ~alt requil?d may
be estimated approximately by considering that 5.4 ml. pI 70% perchloric acid
are used to decompose and oxidize each gram of steel ~nd that 1.6 grams of
the acetate will be required for each cubic centimeter of the acid remaining.
If preferred, the acetate may be added in small p'ortions to the solution while
it is being heated, until the addition of further acetate would cause the f0rmation
of a permanent precipitate of ferric phosphate. Heat the solution to 50 C.,
using a thermometer as stirring rod, and titrate the vanadyl at once with the
0.05 N Plilrmanganate. The oxidizing agent should be added slowly, as only
a small volume of it will be required and also because the reaction between
vanadyl and permanganate ions is not instantaneous. The color change at
the end point is the same as in the first titration and the end point is considered as reached when there is no return of a pale pink shade during an interval of a minute. It is sometimes advisable to add a drop more of indicator
,before titrating the vanadium.
To determine the normality of the ferrous sulfate solution, add sufficient
of it to require 35 to 50 ml. of the permanganate to 250 ml. of water containing
10 to 15 ml. of 70% perchloric acid. Add 2 dI:ops of 0.025 M o-phenanthroline
ferrous complex and titrate at once with the standard permanganate solution.
DETERMINATION Oli' SJ_<;LENIUM IN STAINLESS STEELS

97

Dissolve a 5 gram sample in 60 ml. of a mixture of equal parts of concentrated HNO s and concentrated HCI in a 50D-ml. Erlenmeyer flask. The initial
reaction is usually very rapid, and can be controlled by cooling the flask. After
the initial reaction has subsided, solution of the sample is completed on the hot
plate.
Add 35~l. of 70% HCI0 4 and evaporate to dense fumes, continue fuming
until the condensation of the vapor is observed near thl neck of the flask.
Cool, add 100 ml. of cold water and warm slightly to effect solution of salts.
Filter through a No. 42 Whatman paper to remove silica, washing well with
HCI (1 : 1) and hot water. The volume of the filtrate should be held, to 150
to 175 ml.
,
Add 125 ml. concentrated HCI and 100 ml. water freshly ~aturated with S02.
Heat to 70 C. and allow to stand at this temperature for three hours. A
higher temperature tends to increase the contamination of the' precipitated
selenium by occlusion.
Cool to room temperature and filter the met.allic selenium on a weighed
Gooch crucible. ' The asbestos should be of the 'amphibole variety and should
be thoroughly washed with warm HCI (1 : 1) and water freshly saturated with
80 2, After filter has been prepared, dry and ignite at 700 C., cool and weigh.
Filter the solution containing the selenium through the crucible, wash
precipitate free of iron with warm dilute HCI (1 : 1). Wash free of acid with
warm water, followed by two alcohol (96%) washing and one ether washing.
Dry for one hour at 100 C. and weigh.
.
I
97 Standard Procedure of the Carpenter Steel Co., Raa'i;g Pa., through the Courtesy
of W. J. Boyer, Asst. Chief Chemist.

CHEMICAL ANALYSIS OF IRON AND STEEL

1481

Record the gain in weight as "impure selenium." To check the purity,

ignite to constant weight at a temperature of 700 0 C. If this weight is within
1.0 mg. of the ignited tare weight, and the asbestos mat shows no marked
discoloration, consider the selenium pure. Otherwise, subtract the ignited
weight of the crucible after volatilizing the selenium, and record the difference
as pure selenium. Divide the weight of selenium by weight of sample taken,
times 100, to calculate the per cent selenium.
NOTE.-This method runs on an average 0.02% higher as compared to the volumetric
procedures recently published. 98

THE RAPID DETERMINATION OF MOLYBDENUM IN STEEL

99

Solutions Necessary. Hydrochloric-sulfuric Acid Mixture.-:I:I 20, 1450 mI.,

H 2S0 4 , sp. gr. 1.84, 450 mI., HCI, sp. gr. 1.19, 100 mI.
Nitric-sulfuric Acid Mixture.-H 20, 750 mI., HNO a, sp. gr. 1.42, 350 ml ..
H 2S0 4 , sp. gr. 1.84, 225 mI.
Potassium SUlfocyanate Solution.- KCNS, 50 grams, H20, 1000 mI.
. Stannous Chloride Solution.-SnCI2, 250 grams, HCl, sp. gr. 1.19, 200 mI.,
H 20, 800 mI. Dissolve the stannous chloride in the hydrochloric acid, boil
until clear, and add the water. Add a few pieces of metallic tin to prevent
oxidation.
Procedure.-Weigh a 0.5-gram sample of steel and place in a 400 ml.
beaker. Add 10 ml. of nitric-sulfuric acid mixture. Heat gently until the
sample is in solution, then evaporate carefully and rapidly on a hot plate to
cQpious fumes. Do not use a cover glass. Some trouble due to spattering
may be experienced at first; however, a little practice in regulating the temperature of the hot plate will ~eadily overcome this difficulty. To obtain concordant results all nitrates must be driven off, which necessitates steady
fuming.
.
Cool the contents of the beaker; add exactly 30 ml. of hydrochloric-sulfuric
acid mixture and heat until the salts dissolve. The amount of hydrochloric
a~id used in this operation is very important; too much will cause a fading of
the color even before the ether extraction can be made, while not enough of
. this reagent may present difficulties in the solution of the salts. Then add
30 mI. H 20.
Cool the solution to room temperature; add from a burette 5 ml. of potassium
sulfocyanate solution. Stir well and add 10 mI. of stannous chloride solution
from a burette and again stir thoroughly. It is very important that all of the
iron be reduced before the extraction is made; therefore, after the addition of
the stannous chloride care should be taken that none of the unreduced solution
is left on the sides of the beaker.
Extra,ction with Ether.-Transfer the acid solution to a separatory funnel
of suitable size; add 20 mI. of ether and shake well. Return the acid solution
98 Determination of Selenium in 18-8 Stainless Steels, by G. G. Marvin, with W. C.
Schumb. Ind. & Eng. Chern., Anal. Ed., 8, 109 (1936). Determination of Selenium in
Steel, by W. C. Coleman and C. R. McCrosky. Ind. & Eng. Chern., Anal. Ed., 8, 196
(1936). Gravimetric Determination of Selenium in Alloy Steels, by Louis Silverman.
Ind. & Eng. Chern., Anal. Ed., 8, 132 (1936).
99 By O. L. Maag and C. H. McCollan, The Timken Steel & Tube Company, Canton,
Ohio, Ind. Eng. Chern., 17,524 (1925).

1482

CHEMICAL ANALYSIS OF IRON AND STEEL

to the original beaker and draw off the ether solution into a c~ean, dry 50 ~IDl.
graduated cylinder. Shake out the acid mixture with Isuccessive 10 ml. portions of ether until all color is removed, if one extraction ~oes not remove same. >
The volume of ether used for the sample should be apprpximately the same as
is required to give a corresponding color with the standard. Transfer the
combined ether solutions to a graduated matching tube.,
Color Comparison.-For this operation a Kennicott-Campbell Hurley
colorimeter may be used. Illumination is supplied by a Wratten Safelight No.
2, made especially for dark-room purposes. This fixture carried a 50 watt bulb,
which gives sufficient light and does not materially affect the atmospheric
temperature and thus cause a rapid deterioration of the standard due to
evaporation of the ether. It is essential that the light be so placed that the
illumination is equal on both mirrors.
Place the matching tUbe containing the unknown in the colorimeter, and
adjust the leveling device until the depth of color in the tube containing the
standard equals that of the unknown. Take the reading in milliliters of the
standard solution in the gradua~ed matching tube. From this value,the
molybdenum content of the unknown steel canbe calculated.
Standard Soiution.- Weigh two 0.5-gram Ijamples of standard steel. Run
through in the usual manner and make the ether extract up to exactly 50,ml. in
the standard matching tube. Place in the colorimeter and compare. If the
color tints match, one sample may be diluted to exactly 100 ml. as a standard for
low percentage molybdenum, or the two may be united to make a standard for
steels with a higher molybdenum content. Transfer to the leveling tube of
the colorimeter.
The top of the leveling tube should be covered when not in use, in order to
increase the time that a standard can be used. Since the deterioration Ofi a
standard is generally due to evaporation of tlie ether rather than breaking
down of the salt, its life is largely dependent on the room temperature. In this
laboratory the standard is renewed every two hours, two samples of standard
steel being kept fuming slightly on the cooler portion odhe hot plate, to be used
whenever the standard in use is suspected.
Calculation of Results.-If a steel containing 0:23% molybdenum is used
as a standard, a I-gram sample will contain 0.0023 gram of molybdenum; on
dilution to 100 m!. each cubic centimeter will contain 0.000023 gram of molybdenum. If 14; ml. of the s~andard are used to match the color of the unknown
steel, the unknown must contain 14 XO.000023, or 0.000322 gram: .
Since the weight of the unknown used is 0.5-gram, it will contajn twice this
amount per gram, or 0.064%.
Ml. standard (14) Xgrams/ml. (0.Q00023)
Grams sample (0.5)

% Mo

(0.064).

ELECTROLYTIC DETERMINATION OF TITANIUM, ZIRCONIUM

OR ALUMINUM IN CHROMIUM STEELS AND ALLOYS 100
The practice of adding 0.5 to 3% titanium or zir90nium t6 " Enduro" and
other high chromium alloys during the last few yearrsjias created a demand for
100 By H. A. Kar, Research Chemist, The Timken Steel & Tube Co., Canton, OhiO,
Metals and Alloys, (1935).

CHEMICAL ANALYSIS OF IRON AND STEEL

1483

a short and accurate method for determining the exact amount of these elements
held in the steel. . By electrolysis the interfering elements are deposited while
titanium, zirconium and aluminum, if present are left in the solution.
Apparatus.-The mercury in this method is used as the cathode and the
power is supplied by a l/l6-in. thick platinum wire sealed in a glass tubing.
About Y2 in. of the platinum wire is left outside the lower end of the sealed glass
tube to be dipped in the mercury from the side of the beaker and about 2 in.
above the upper end of the glass tubing to be. connected to the negative pole
of the circuit.
A platinum plate, about 134 in. in diameter, welded to a Ys-in. thick platinum wire and sealed in a glass tube, is used as the anode. The anode is connected to the positive pole of the circuit, and lowered about Y2 in. below the
surface of the solution. The plate is provided with a hole about % in. diameter
in the center for the stirring rod.
The mercury is constantly agitated during electrolysis by a " T "-shaped
glass rod passing through the hole in the plate and attached to motor driven
stirrer. The" T " end of the rod reaches the bottom of the mercury and is
adjusted to ope,rate without touching the beaker or electrodes.
Procedure.-Transfer one gram sample (two grams if titanium is less than
0.5%) to a 200-ml. beaker, add 20 to 40 ml. of 10% sulfuric acid, and heat slowly
until action is complete. Filter into a 200 m!. tall electrolytic beaker containing
mercury about one inch deep, wash a few times with water, and reserve the
paper. The filtrate and washing should not exceed 75 m!. in volume
Set the beaker containing the solution in a trough of circulating water and
electrolyze at 4 amp. for lY2 to 2 hrs. As soon as electrolysis begins, transfer the
reserve filter paper to a platinum crucible and ignite. Cool, add a little potassium bisulfate and fuse. Add 5 ml. water and 3 drops of sulfuric acid in the
crucible and heat to dissolve the melt. Cool and combine with the main
solution.
At the end of a set period, withdraw a drop of the electrolyte and treat
with a drop of fresh 1 % solution of potassium ferrocyanide; if a blue color does
not develop, raise the electrodes above the surface of the electrolyte and wash
with cold water. Shut off the power and get the beaker out of the water
trough. Filter, wash about 5 times with cold Y2% sulfuric acid and proceed
with the filtrate by either 'of the following two methods.
1. Add enough sulfuric acid to the filtrate to make it 10% acid by volume,
cool and treat slowly with 6% solution of cupferron while stirring, until furthr
addition does not cause a precipitate. Allow this to settle for 5 min. Filter on
suption and wash 5 times with 10% cold hydrochloric acid and 10 times with
cold water. Transfer the paper to a platinum crucible, dry and ignite. Volatilize the silica with hydrofluoric' acid in the presence of sulfuric acid and again
ignite at 10000 C. in an electric muffie for one-half hour. Cool and weigh.
2. Treat the filtrate with excess of ammonia, add a couple of crystals of
ammonium persulfate and boil for 2 min. Add paper pulp in the beaker and
stir. Filter and wash with ammoniacal hot water. Transfer the paper to a
platinum crucible and ignite; Volatilize the silica with hydrofluoric and sulfuric acid and ignite at 1000 0 C. in an electric muffile for Yz hr. Cool and
weigh.

1484

CHEMICAL ANALYSIS OF IRON AND STEEL

Fuse the residue left in the crucible with a little potassIum bisulfate, cool
the melt and dissolve in 20 m!. of 20% sulfuric acid. Transfer the solution
to a 200-m!. beaker, add 2 grams of citric acid, and stir to .dissolve. Render
the solution decidedly alkaline with ammonia and saturat~ for 5 min. with
H 2S. Filter and wash 10 times with 5% ammonium citrate saturated with
ammonium sulfide. Transfer the paper to a crucible and J.gnite. Cool and
weigh. The difference in weight is titanium or zirconium oxide. If titanium
is sought multiply by 60, and if zirconium is sought multiply the weight of
oxide by 73.9.
.
.
This method can be applied to all kinds of steel alloyed with titanium or
zirconium.
Determination of aluminum in nitralloy or other aluminum alloy steels is
made by this method, proceeding exactly as described above and precipitating
aluminum by ammonia after iron, chromium and other interfering elements
are deposited.
The determinations of titanium and zirconium in ferrotitanium and ferrozirconium are also made by this method, precipitating either one by cupferron
after electrolysis is compJeted.
This procedure works well for uranium in steels, if vanadium is absent.
Mercury does not deposit the vanadium as it does the iron, manganese, chromium, cobalt, nickel, molybdenum, tin and copper.
The mercury can be cleaned and used again after every determination either
by shaking it in a separatory funnel with concentrated hydrochloric acid, or
pressing through a chamois skin.

ALLOYS, III
FERRO-ALLOYS
TENTATIVE METHODS OF CHEMICAL ANALYSIS OF
FERRO-ALLOYS 1
A. S. T. M. DESIGNATION: A 104-36 T

FERROSILICON
DEl'ERMINATION OF SILICON BY FUSION WITH SODIUM PEROXIDE AND
DEHYDRATION WITH HYDROCHLORIC ACID
Transfer 0.5000 g. (use 0.4000 g. for 90% silicon) of the finely powdered
sample (passing a No. 150 sieve 3) to a 50-m!. pure iron crucible made from No.
20 gage (0.32 in. thickness) ingot iron, add approximately 10 g. of sodium
peroxide. Mix thoroughly with a platinum or iron rod and carefully clean the
rod of adhering particles by scraping with another rod. Cover the mixture
with a layer of about 2 g. of additional peroxide. Heat the crucible and contents on a hot plate for 5 to 10 minutes to expel any water in the peroxide which
would cause spattering in the subsequent fusion. Carefully fuse over a'low
flame by holding the crucible with a pair of tongs and slowly revolving it
around the outer edge of the flame until the contents have melted down quietly.
When the fusion is molten rotate the crucible carefully to stir up any unattacked
particles on the bottom or sides, the crucible and contents being maintained at
a low red heat. Just before completion of the fusion, which requires only three
or four minutes, increase the temperature to bright redness for a minute. If
the reaction has proceeded too violently, due to too rapid heating, to the use of
insufficient sodium peroxide, or to the lack of thorough mixing, appreciable loss
will occur and the work should be repeated.
Cover the crucible and allow it to cool until the fusion has solidified, and
tap the crucible, while still warm, several times on an iron plate, which will
loosen the fused mass in a solid cake. When the crucible and cake have completely cooled, transfer the cake to a large (275-ml.) gold or preferably platinum
dish. Cover the dish with a clock glass and cautiously add 50 ml. of cold water.
1 Under the standardization procedure of the Society, these methods are under the
jurisdiction of the A. S. T. M. Committee E-3 on Chemical Analysis of Metals. Reproduced through the courtesy of the American Society for Testing Materials, 260 S. Broad
St. Philadelphia, Pa.
2 This is a Tentative Standard and under the Regulations of the Society is subject
to annual revision. Suggestions for revision should be addressed to the Headquarters
of the Society, 260 S. Broad St., Philadelphia, Pa. Issued as Tentative, 1925; Adopted,
1927; Revised, 1934; Reverted to Tentative, 1936.
3 For detailed specifications for this sieve, see Standard Specifications for Sieves for
Testing Purposes (A. S. T. M. Designation: Ell) of the American Society for Testing
Materials, 1936 Book of A. S. T. M. Standards, Part II, p. 1413.
1485

1486

FERRO-ALLOYS,

When ,the reaction ceases, wash the small amount of adhering matter in the
crucible into the dish with a little water. Cool the solution and carefully a,dd
an excess of hydrochloric acid. Evaporate to dryness, preferably on a steam
bath, but do not bake at a temperature above qOo C. If platinurii or gold
dishes are not available for solution of the fused cake, the latter can be disintegrated with water in a pure nickel dish and th,e contents then transferred
to a porcelain dish containing sufficient hydrochloric acid to provide an excess
of acid. It is not desirable to dissolve the fusions directly in porcelain, because
of the action of the alkaline solution.
When the residue is dry, allow the dish to cool. Add 10 ml. of hydrochloric
acid (sp.gr. 1.13), cover the dish and heat gently for a few minutes. Dilute,
with 100 ml. of hot water, digest for a few minutes and filter through an 11-cm.
paper. Wash about eight times with hot 5% hydrochloric acid, three times
with warm 5% sulfuric acid and a few times with hot water. Reserve this
paper until the second evaporation has been completed.
Evaporate the filtrate, to dryness and bake the covered casserole or dish
at 110 0 C. for one hour. Heating at a higher temperature is unnecesSRry and
also disadvantageous. Cool, add 10 ml. o~ hydrochloric acid (1 : 1), digest on
the steam bath for ten minutes, add 100 ml. of warm water, filter immediately
and wash about eight times with cold 1 % hydrochloric acid, twice with cold 5%
sulfuric acid and a few times with cold water.
Place the paper and silica from the second dehydration in a large partially
covered platinum crucible and char the paper without inflaming. Add the
paper containing the silica from the first dehydration, char as before and then,
ignite carefully until the carbon has been burned off. Great care should be
exercised in igniting the papers as the current of air produced by a burning
filter is sufficient to carry finely divided silica out of the crucible. When the
carbon of the filter paper has been completely burned, gradually heat the tightly
covered crucible in a muffle at 1100 0 to 11250 C. for 30 minutes. Cool in a
desiccator, weigh and check for constant weight by ten minutes additional
heating at the above temperature. Add sufficient sulfuric acid (1: 1) to
moisten the silica and then 15 to 20 ml. 6f pufe hydrofluoric acid (48%).
Evaporate to dryness, ignite and weigh. Tne loss represents silica which contains 46.72% of silicon. A blank, not omitting the same amount of hydrofluoric
acid, should be carried through all stages of ,'the operation.
N OTE.~The silica remaining in the filtrate after the second. dehydration is practically
negligible, averaging about 0.0006 to 0.0010 g. in hydrochloric dehydration and 0.0010
to 0.0015 g. in sulfuric dehydration.
FERROMANGA~ESE
DETERMINATION OF MANGANESE BY THE MODIFIED BISMUTHATE METHOD 4

Reagents Required: Dilute Nitric Acid (sp.gr.: 1.135).-Mix 280 ml. of nitric
acid (sp.gr. 1.42) and 720 ml. of distilled water. Test for proper specific
gravity.
Dilute Nitric Acid (3: 97).-Mix 30 ml. of recently' boiled nitric acid
(sp.gr. 1.42) and 970 ml. of distilled water. Add a little sodium bismuthate
and, shake well.
/
4 J. of Ind. Eng. Chemistry, 16, 58 (1924).

FERRO-ALLOYS

1487

As~estos.-Asbestos for this work should be digested in hot nitric acid

(sp.gr. 1.42) and then washed free from acid with distilled water.
Sodium Bismuthate.-This reagent should be of 80% grade and its oxidizing
power tested as follows: Shake 0.5 g. of the reagenJ; with 4 g. of potassium iodide
and a little water in a stoppered flask. Add 15 ml. of hydrochloric acid and
allow to stand in the dark, with occasional shaking until the bismuthate has
entirely decomposed. Dilute to 300 ml. and titrate with 0.1 N sodium thiosulfate, using starch as an indicator at the end. One milliliter of 0.1 N Na 2S 203
equals 0.0140 g. of NaBi0 3.
Standard Potassium Permanganate Solution.-See the Determination of
Chromium in :Ferrochromium. 5 One milliliter of 0.1 N KMn04 equals 0.001099 /
g. of manganese.
Method: (A) Solution of the Sample.-All samples should be crushed to
approximately a size to pass a No. 100 sieve and dried at 105 to 110~ C.
With high-carbon 80% ferromanganese, dissolve 0.25 g. of the prepared
sample in 60 ml. of nitric acid (sp.gr. 1.42) in a 750-ml. Erlenmeyer flask, provided with a clock-glass cover or trap to prevent loss by spraying.
Add 2 g. of potassium nitrate and boi! gently until the volume of the solution
has been reduced to about 15 ml. Cool, add 10 ml. of H 2S0 4 (1 : 1) and
evaporate to strong fumes of sulfuric acid. Cool somewhat, dilute the solution
with 200 ml. of HN0 3 (sp.gr. 1.135), heat to boiling, remove from the heat and
add sodium bismuthate a little at a time until a precipitate of manganese
dioxide has formed. When this occurs, add sulfurous acid dropwise until the
precipitate has dissolved and an excess of about 1 ml. has been provided to
reduce any chromium to the trivalent form. Boil the solution about five
minutes to completely expel oxides of nitrogen, dilute to 250 ml. with water,
cool the flask and solution in crushed ice and treat as described later. Highcarbon ferromanganese containing 1 % or more of chromium may not dissolve
completely in nitric acid. In this case, dilute the solution after treatment with
nitric acid, filter, and fuse the washed residue with a little sodium carbonate.
Dissolve the melt in a small volume of nitric acid (sp.gr. 1.135), and add ,it to
the main solution. Treat with potassium nitrate, etc., as previously described.
With low-carbon ferromanganese or manganese metal dissolve 0.2500 g.
of the sample in 250 ml. of nitric acid (sp.gr. 1.135) in a covered 750-ml. Erlenmeyer flask and boil the solution gently for a few minutes to expel oxides of
nitrogen. Treat with bismuthate and sulfurous acid as described above.
With silico-manganese (60 to 70% of manganese) and manganese-silicon
(20 to 25% of manganese) treat 0.3000 g. of the former and 1.000 g. of the latter
with 5 ml. of hydrofluoric acid in a large platinum dish equipped with a gold
or platinum cover. (Some spiegeleisens are completely soluble in nitric acid in
which case the sample can be dissolved directly in 250 ml. of nitric acid (sp.gr.
1.135); When the reaction moderates, add nitricacid (sp.gr. 1.42) a few drops
at a time until the sample has completely dissolved. Add 10 ml. of nitric acid
((sp.gr. 1.42) and evaporate to a syrupy consistency. Again add 10 ml. of
nitric acid (sp.gr. 1.42) and evaporate as before to expel hydrofluoric acid.
Dissolve the residue in nitric acid (sp.gr. 1.135), transfer to a 750-ml. Erlenmeyer flask, dilute with nItric acid (sp.gr. 1.135) to 250 ml. and treat as described before with sodium bismuthate, etc.
6 See p. 1490.

FERRO-ALLOYS

1488

(B) Oxidation of Manganese.-If the preceding 4irections have been

followed, the manganese will be present in a concentration of approximately
0.001 g. per 1 m!. of nitric acid (sp.gr. 1.135). This c9ncentration of manganese
and nitric acid, together with a temperature of 10 to 15 C. are the conditions
which insure maximum stability of the permangalii.c acid, formed as further
described. To insure complete oxidation of the manganese to permanganic
acid, it is essential that the sodium bismuthate (80% of NaBiO s) be used in
the ratio of at least 26 g. to every gram of manganese in solution. Add approximately 7.0 g. of sodium bismuthate (80%) to the flask, agitate briskly for
one minute, dilute with 250 m!. of cold water, and filter immediately on a layer
of acid-washed asbestos supported on a 2-in. alundum or perforated porcelain
plate resting in a large glass funnel. The filter can be washed free from manganese more readily if not allowed to run dry during the filtering and washing.
Wash the filter and residue with cold 3% nitric acid until the washings are
entirely colorless and immediately treat the filtrate and washings as described
in the following paragraph.
.
Weigh accurately 9.000 g. of ferrous ammonium sulfate, taken from a.bottle
of the well-mixed salt, and add to the filtered' solution of permanganic acid. Stir
briskly and as soon as the reduction is complete and the salt has dissolved,
titrate the excess of ferrous salt with 0.1 N potassium permanganate. The
manganese value of the ferrous salt is obtained in the following manner:
Two hundred and fifty milliliters 6f nitric acid (sp.gr. 1.135) contained in
a 750-ml. Erlenmeyer flask are treated with approximately 2.0 g. of sodium
bismuthate and the solution is vigorously agitated for one minute, diluted with
250 ml. of cold water, and immediately filtered on a layer of acid-washed
asbestos as previously described.. Five (5.0000) grams of the ferrous ammonium
sulfate are added to the filtrate and the resulting solution is immediately titrated.
with 0.1 N KMn04 until a faint pink color appears. The excess of permanganate used to cause an end point color must be determined by watching the
color in another beaker containing the same amounts of acid and cold water.

nbt

NOTES.-l. Chromium "\yhen less than 2% does

cause any interference if the
reactions are rapidly done in cold solutions. Larger amounts interfere to some extent
and should be separated prior to the final oxidation with bismuthate.
2. The gravimetric determination of manganese by the pyrophosphate method
will give high values if the precipitate of manglj.nese ammonium phosphate is allowed
to stand more than ten hours before filtration.
3. Volumetric methods such as Volhard's, Ford-Williams" and Julian's do not permit
the use of a theoretical titre. Each necessitates the use of an empirical titre obtained
by titration of material of known manganese content which has been carried through
all steps of the method.

DETERMINATION OF CARBON BY THE DIRECT-COMBUSTION METHOD

See the. Determination of Carbon in Ferrochromium by the ;Direct-Combustion Method. 6 A furnace temperature of 1150 to. 1175 C. suffices, and highsilicon material should be covered with 1 or 2 g. of ingot iron.
DETERMINATION. OF PHOSPHORUS BY THE ALKALI;METRIC M:mTHOD

Solutions Required: Dilute Nitric Acid-(sp.gr. 1.135).-Mix 280 m!. of

HNO a (sp.gr. 1.42) and 720 m!. of distilled wafer.
6

See p. 1493":

FERRO-ALLOYS

1489

Ammonium Molybdate.-See the Determination of Phosphorus in Ferrovanadium by the Modifled Molybdate-Magnesia Method. 7
Standard Sodium Hydroxide.-Dissolve 6.5 g. of purifled NaOH in 1000 ml.
of distilled water, add a slight excess of a 1 % solution of Ba(OH)2, let stand
for 24 hours, decant the liquid and standardize against Bureau of Standards
Standard Benzoic Acid as follows: Dissolve 0.6 g. of benzoic acid in 15 m!. of
alcohol in a small Erlenmeyer flask and titrate in the cold with phenolphthalein
indicator, taking precautions to prevent access of carbon dioxide from the air
during titration. The ratio of phosphorus to sodium hydroxide is taken as
1 to 23 in calculating the value of the sodium hydroxide. Protect the solution
from carbon dioxide with a soda-lime tube.
Standard Nitric Acid.-Mix 10 m!. of HN0 3 (sp.gr. 1.42) and 1000 ml. of
distilled water. Titrate the solution against the standard hydroxide solution,
using phenolphthalein as an indicator. Adjust the solution so that it is equivalent to the sodium hydroxide solution.
Dilute Ammonium Hydroxide (sp.gr. O.96).-Dilute 400 m!. of NH 4 0H
Csp.gr.0.90) with 600 ml. of water. The strength of this solution should be
established by titration or test with a hydrometer.
Method.-Dissolve 2.000 g. of the sample in 70 m!. of nitric acid Csp.gr. 1.42),
evaporate to dryness and bake for two or three hours at 120 0 C. High-silicon
alloys may require the addition of a little hydrofluoric acid for complete decomposition. In this case, several evaporations with 15 ml. portions of nitric acid
Csp.gr. 1.42) should be made to remove hydrofluoric acid before .the baking
treatment. Moisten the residue with 25 ml. of dilute hydrochloric acid (1 : 1),
digest a few minutes,8 add 75 ml. of ~vVater, heat until the salts are dissolved and
then filter to remove silica. Evaporate the filtrate to a syrupy condition, add
15 m!. of nitric acid (sp.gr. 1.42) and again evaporate to a small volume. Again
add 15 ml. of nitric acid (sp.gr. 1.42) and evaporate to a small volume. Add
100 m!. of nitric acid (sp.gr. 1.135), heat to boiling and add a 1.5% solution of
potassium permanganate until oxides of manganese are precipitated. Dissolve
the oxides of manganese by addition of sulfurous acid and boil to expel the
oxides of nitrogen. To the cold solution add 40 m!. of NH 4 0H (sp.gr. 0.96) and
50 m!. of ammonium molybdate solution and shake for 5 minutes. Allow to
settle for 10 minutes; for material containing less than 0.02% of phosphorus,
allow to stand 30 minutes. Filter and wash 10 times with a 1 % solution of
KN0 3 Return the precipitate and paper to the precipitating vessel, add 1 to 3
m!. in excess of standard N aOH solution and 25 m!. of water, both free from carbon dioxide, and shake or stir until the precipitate is dissolved. Dilute to 100
m!. with water free from carbQn dioxide, add 3 drops of a 0.2% solution of phenolphthalein, and discharge the pink color with standard acid. Finish the titration by adding standard alkali until the reappearance of the pink color. With
very low phosphorus material this titration should be corrected by a blank
carried through all steps of the method.
NOTE.-If vanadium is present in amounts over 0.08% it should be reduced by
ferrous sulfate and sulfurous acid before precipitation with molybdate.
7 See p. 1495.
8 Addition of a little sulfite or sulfurous acid facilitates solution after the baking
treatment.

FERRO-ALLOYS

1490

DETERMINATION OF SULFUR BY THE OXIDATION METHOD

Proceed as in the. Determination of Sulfur by Ithe Gravimetric Method

as described in the Tentative Methods of Chemiclj-l Analysis of Steel, Cast
Iron, Open-Hearth Iron and Wrought Iron (A. S. T. M. Designation: E 30-36 T)
of the American Society for Testing Materiais. 9
'
.

NOTEs.-Routine determinations of sulfur can be made by the evolution met~od

provided concentrated hydrochloric acid (sp.gr. 1.19) is used, preferably in all-glass
apparatus of the Pulsifer type.
DETERMINATION OF SILICON BY THE NITRO-SULFURIC ACID METHOD

Solutions Required: Nitro-Sulfuric Acid.-Pour 160 ml. of sulfuric acid

(sp.gr. 1.84) slowly and with stirring into 660 ml. of water and add 180 ml. of
nitric acid (sp.gr. 1.42).
Method.-Dissolve 0.9344 g. of the sample passing a No. 100 sieve 10 with
50 ml. of the nitro-sulfuric acid in a 250-ml. porcelain casserole and evaporate
until copious fumes of sulfuric acid are evolved. After the solution has cooled
somewhat add 125 ml. ot distilled water and.5 ml. of HCI (sp.gr. 1.19). JIeat
with frequent stirring until all salts are dissolved and immediately filter on a
9-cm. paper. Wash the precipitate with cold dilute HCI (5 : 95) and hot water
alternately to complete the removal of sol'\lble s'alts, and finally with hot water
until free from acid. Transfer the filter to a platinum crucible, char the' paper
carefully, and finally ignite with the crucible covered over a blast lamp or in a
muffle furnace at 1050 0 to 11000 C. for at least 10 minutes; cool in a desiccator
and weigh. Add sufficient dilute sulfuric acid (1 : 1) to moisten the silica
and then a small amount of hydrofluoric acid. Evaporate to dryness, ignite
and weigh. The difference in weights divided by 2 and multiplied by 100
represents the percentage of silicon in the alloy.
NOTE.-For silico-manganese and manganese-silicon, fuse 0.4672 g. of the former or
0.2336 g. of the latter with 8 g. of sodium peroxide in a pure iron crucible. Dissolve the
fusion and evaporate to fumes of sulfuric acid. When acidifying the solution of the
fusion with sulfuric acid it is necessary to add some sulfujous acid to dissolve the manganese dioxide which separates out.
FERROCHROMIUM
DETERMINATION OF CHROMIUM BY FUSION WITH S.ODIUM PEROXIDE
I

Solutions Required: Potassium Permanganate, 0.1 N.-D.issolve 3.2 g. of

potassium permanganate in 1000 ml. of distilled water, allow to age for at least
ten days and filter through purified asbestos. Standardize the permanganate
solution against 0.2 g. samples of Bureau of Standards Standard Sodium
Oxalate by the procedure given on p. 1431.
Standard Ferrous Ammonium Sulfate.-Diss6Ive 43 g. of the salt in 250
ml. of cold water, add 40 ml. of sulfuric acid (1 : 3), dilute to one liter with
distilled water, and mix thoroughly. Large amounts of the solution can be
,
For detailed specifications for these sieves, see Standard Specifications for Sieves
for Testing Purposes CA. S. T. M. Designa~ion: E 11) of the American Society for Testing
Materials, 1936 Book of A. S. T. M. Standards, Part II;p. 1413.
9

See p. 1440.

10

FERRO-ALLOYS

1491

conveniently mixed by means of a stream of carbon dioxide from a cylinder.

As the ferrous ammonium sulfate solution gradually weakens in reducing
power it is necessary to standardize it daily or at the same time the ferrochromium is analyzed. The solution can be standardized against either potassium dichromate or sodium oxalate.
The first method is somewhat simpler and more satisfactory for routine
work as the conditions prevailing in both the standardization and in the analysis
are practically identical. No" blank" corrections are afterwards necessary
where the amounts of dichromate, ferrous sa:It and permanganate are approximately those prevailing in the analysis. The second method is employed
chiefly for checking the accuracy of the results obtained with potassium dichromate and in cases where only. occasional 'analyses are made and pure dichromate is not readily available.
To prepare pure potassium dichromate, recrystallize at least twice the
purest grade of the salt obtainable, dry the crystals at 1500 C., grind to a fine
powder and again dry at 150 0 C. to constant weight. Keep the prepared salt
in a ground-glass stoppered bpttle. Transfer 1.0000 g. of the pure dichromate,
which contains approximately the same amount of chromium as an 0.5000-g.
sample of 70% ferrochromium, to an SOO-ml. beaker. Dissolve in cold water,
acidify with 40 mI. of sulfuric acid (1 : 3) and dilute with cold water to 500 m!.
Add a slight excess of the ferrous sulfate solution, approximately 190 mI., and
titrate the excess ferrous salt with 0.1 N permanganate to the first faint permanent darkening of the clear green color.
'
Example.-The calculations involved are illustrated as follows: In a standardization,
190.0 m!. of the ferrous solution was used and 5.06 m!. of 0.1 N permanganate was
required to titrate the excess.
~ ml. of 0.1 N K 2Cr 207 contains 0.004903 g. of K 2Cr 207
1 g. of K 2Cr 207 equals 203.94 m!. of 0.1 N K 2Cr 207
then:
203.94+5.06 = 209.0 m!. of equivalent 0.1 N solution reduced by the ferrous solution.

2~:OO = 1.100, the 0.1 N normality factor.

In standardizing through sodium oxalate, it is necessary to add the ferrous
ammonium sulfate solution by means of an accurately calibrated pipette or
burette instead of an automatic pipette. Transfer 100 ml. of the appl'Oximately
0.1 N ferrous solution to a 600-ml. beaker, dilute to 300 ml. with cold 5%
sulfuric acid, add 2 m!. of phosphoric acid and titrate immediately with 0.1 N
KMn04 to a faint permanent pink color. Determine the" blank" on the same
volume of water and acids and deduct. When this method is used it is necessary in the actual analysis to determine a " blank" to counteract the influence
of the green color of the chromium sulfate upon the permanganate end point.
This " blank" can be determined in the solution used in the analysis and is
obtained as pescribed later.
Method.-Transfer 0.5000 g. of the sample of ferrochromium (dried at
1050 to 1100 C.) to a 30-m!. heavy-walled porcelain crucible. Samples of lowcarbon ferro chromium should be crushed to pass through a No. 40 screen, while
high-carbon ferrochromium and chromium metal should be reduced to pass a
No. 100 screen. Add approximately 8 g. of sodium peroxide and thoroughly

1492

FERRO-A~LOYS

mix the contents of the crucible with a small rod, being careful to clean the rod,
which can be done conveniently by scraping with another rod. Cover the
mixture with 1 to 2 g. of sodium peroxide. Carefully fuse the c'ontents of the
covered crucible, preferably in an electric muffle heated to 600 0 to 700 0 C.
Fusion for five minutes at a low red heat after the mass has melted will insure
complete decomposition. When the charge has melted, the crucible should be
given a slight rotary motion to stir up .any unattacked particles.
Place the cooled crucible and cover in a600-ml. beaker and dissolve the
contents in 200 ml. of water. 'Whim the melt has diss91,ved, remove the crucible
and cover, add 1 g. of sodium peroxide and boil the solution ten minutes to
completely decompose the excess of peroxide.,
'
Allow the precipitate to settle and filter through a pad of ignited asbestos,
preferably in a Buchner funnel. An asbestos rather than a paper filter should
be used, for the aJkaline solution extracts reducing matter from paper which
would cause low results for chromium. T.horoughly wash the precipitate with
hot water, and at no time allow the precipitate to run dry during the filtering
and washing. If desired, the precipitate can be subsequently dissolved in
dilute sulfuric acid and tested for chromium by adding a slight excess of sgdium
peroxide; the amount of chromium retained is negligible if the precipitate is
filtered and washed as directed.
Cool the filtrate, acidify with 50 m!. of sulfuric acid (1 : 1), again cool to
room temperature, transfer to an SOO-m!. beaker, dilute, with cold water to
500 ml. and add a measured excess of ferrous ammonium sulfate. The 'point
at which reduction is complete can be told by changes in the deep green color
which is developed. The ferrou. salt can be added either as an approximately.
0.12 N solution or in weighed portions from a bottle of the well-mixed salt.
If the solution of ferrous salt is used, approximately 180 m!. \vill be required
for a 70% alloy; if the salt is added, 8 g. will usually be sufficient. Titrate
the excess of ferrous salt with standard O~10 N potassium permanganate. The
first faint permane:r:t darkening of the clear green color is taken as the end point,
which is quite sharp and well delined to the practiced eye. If the ferrous salt
has been standardized against dichromate, subtra;ct the required number of
milliliters of 0.1 N'permanganate from the number of milliliters of 0.1 N ferrous
solution (obtained by multiplying the number of milliliters of the approximately
0.1 N ferrous solution by the, 0.1 N normality factor). The difference represents the number of milliliters of 0.1 N ferrous solution which multiplied by
q.001733 g. gives the number of grams of chromium present.
,"Vhere the ferrous solution has been standardized against sodium oxahite,
it is necessary to subtract a " blank" from the permanganate titr~tion. This
(/ blank." is obtained as follows: When the end point is reached and the readings
have been noted, add about 5 m!. in excess _of the ferrous solution-and oxidize
the_ excess of ferrous iron by' adding 8 mI. of a 15% solution of ammonium
persulfate. Stir the liquid vigorously for one minute. Carefully add O.l,N
permanganate to the solution with constant stirring until the same end point
is obtained as before. The volume of permanganate required, u~uaJly 10.2 to
0.4 mI., constitutes the H blank" to be deducted from the volume of the permanganate titration." For example, ,in. the determination of chrotPium in ,an
0.5000-g. sample, 190.0 m!. of a 0.11 N ferrous solution and 11.0.0 m!. of an
0.1l'{ prmanganate (11.40-0.4 m!. blank) were employed.

FERRO-ALLOYS

1493

The calculations are as follows:

190.0 X 1.100 =209.0 m!. of 0.1 N ferrous solution
209.0-11.0 =198.00 ml.
198.0XO.001734X2X100=68.65% of chromium
The above calculations are given for exactly 0.1 N permanganate; if the
permanganate is not 0.1 N, the volume used ml,lst be converted to the equivalent of 0.1 N by multiplying by the proper factor.
DE~ERMINATION

OF CARBON BY THE DIRECT-COll-IBUSTION METHOD

Determine carbon by the direct-combustion method described in the Tentative Methods of Chemical Analysis of Steel, Cast Iron, Open-Hearth Iron
and Wrought Iron (A. S. T. M. Designation: E 30-36 T) of the American
Society for Testing Materials,ll with the following modifications:
1. Ferrochromium alloys require a furnace temperature of 1150 0 to 12300 C.
in order to yield all of the carbon in the material.
2. "The determination is made as given in the reference except that the
weighed sample of ferrochromium, 0.5000 g. for 4 to 8% carbon and 3 to 1 g.
for 0.10 to 2.00% carbon, is mixed with 1 g. of copper oxide and covered with
2 g. of pure ingot iron drillings in order to facilitate complete combustion.
3. On account,of the small weight of sample used, the blanks are important
and must be carefully determined.
DETERMINATION OF SILICON

Low-carbon ferrochromium and chromium metal can be decomposed by

means of dilute sulfuric acid, but the high-carbon alloy has to be fused with
sodium peroxide. For low-carbon ferrochromium, transfer 0.9344 g. (a double
factor weight) to a 250-ml. porcelain casserole, add 60 ml. of sulfuric acid
(1 : 4) and warm the covered casserole gently until the reaction is complete.
For high-carbon ferrochromium, transfer 0.9344 g. of the sample passing a No.
100 sieve to a 30-ml. pure iron crucible made from No. 20 gage (0.032 in.
thickness) ingot iron. Add approximately 8 g. of sodium peroxide, mix thoroughly and fuse as directed under the determination of chromium. When
the fusion has solidified, tap the covered crucible, while still warm, on an
iron plate to loosen its contents in a solid cake which is then transferred to a
250-ml. casserole. Rinse the crucible with a little water, transfer the solution
to the casserole, cover, and~ carefully add 60 ml. of sulfuric acid (1 : 4) after
which the cover glass is rinsed and removed.
Evaporate the solution obtained by either of the above procedures until
fumes of sulfuric acid are given off. If chromium sulfate solutions are fumed
at too high temperatures or for any great length of time or with too]ittle free
acid present, considerable difficulty will occur in dissolving the salts. The
fuming should, therefore, be made at a temperature not much higher than that
required for evolving the acid. Allow the casserole and contents to partly
cool (if permitted to become entirely cold, difficulty may be experienced in
getting chromium salts to dissolve), add 100 ml. of warm water while stirring
11

See p. 1426.

1494

FERRO-ALLOYS

to prevent the residue from caking on the bottom and boil the solution gentiy
for several minutes to dissolve all salts. Filter, lwashing the paper several
times with cold dilute hydrochloric acid (2%), and then with warm water.
Transfer the filter and contents to a platinum cruciqle, char the paper, carefully
ignite to remove carbon, and then heat the covered crucible over a blast lamp
or in a muffle furnace at 10500 to 1l0J o C. for at least 15 minutes. Cool in a
desiccator and weigh. Add sufficient dilute sulfuri~ acid (1 : 1) to moisten the
silica and then a small amount of hydrofluoric acid. Evaporate to dryness,
ignite and weigh. The difference in weights represents silica which contains
46.72% of silicon.
NOTE.~Crucibles made from pure ingot iron contain only traces of silicon and the
amount of silica present in the sodium peroxide used is usually negligible, but "blanks"
should be run on new lots of crucibles and of peroxide. All nickel crucibles, even the
German "Rein" nickel, contains appreciable amountE! of silicon.

FERROVANADIUM

.'

DETERMINATION OF VANADIUM BY THE .AMMONIUM PERSULFATE METHOD

Solutions Required: Ammonium Pers~lfate (15%).-Mak~ up asneeded by

dissolving 15 g. of the salt in 100 ml. of distilled water. This salt should be
95% pure and its purity tested by adding 30 ml. of a standard ferrous sulfate
solution to 0.3 g. of the persulfate, then 200 ml. of hot' water and titrating the
excess ferrous salt with potassium permanganate. One milliliter of 0.1 N
ferrous solution equals 0.01141 g. of (NH 4)2S20S.
Potassium Ferricyanide Indicator.-Dissolve 0.1 g. of the salt in 100 ml. of
water. This solution should be made up fresh daily,
Standard Ferrous Ammonium Sulfate (Approximately 0.1 N).-See the
Determination of Chromium in Ferrochromium. 12
Potassium Permanganate (0.1 N).-See the Determination of Chromium in
Ferrochromium. 12 One milliliter of 0.1 N KMnO, is equivalent to 0.005096 g.
of vanadium in the following method.
.
I
Method.-Treat 0.5000 g. of the alloy; dried at 1050 to 110 C., in a 250-ml.
covered beaker with a mixture of 60 ml. of sulfuric acid (1 : 3) and 25 ml. of
nitric acid (1 : 1). If the alloy fails to decompose readily because the silicon
content is high, it may be necessary to add a smail amount of hydrofluoric acid.
In this event, a 200-;ml. platinum dish should be substituted for the beaker.
When the reaction has abated, rinse the cover glass and evaporate the solution
until copious fumes of sulfuric acid are evolved. Cool, adg 100 ml. of water
and heat until the salts are ,dissolved. Transfer the solution to an 800-ml.
beaker, dilute ttl 400 ml. and to the cooleq. solution add an approximately 0.1 N
solution of potassium permanganate until a strong pink color has developed
which remains permanent for thirty seconds. t Next reduce the vanadium by
the use of an approximately 0.1 N solution of ferrous ammonium sulfate added
until a drop of the solution withdrawn from the beaker and placed on a drop
of potassium ferricyanide indicator shows ferrous iron present-about 35 ml.
of 0.1 N ferrous solution. Add 5 ml. of the ferrous lammonium sulfate in
excess, stir the, solution at least one minute, ~dize the excess of ferrous sulfate
12

See p. 1490.

FERRO-ALLOYS

1495

with 8 ml. of 15% ammonium persulfate and st'it the solution vigorously for
one minute. Titrate the solution with 0,1 N potassium permanganate solution,
added with constant stirring until a faint pink color appears which remains
permanent for 30 seconds. Determine the " blank" by dissolving 0.4 g. of
" ingot iron" in 60 ml. of H 2S0 4 (1 : 3) and 5 ml. of nitric acid (sp.gr. 1.42), and
putting the solution through all of the operations of the analysis. If the alloy
contains an appreciable amount of chromium, a weight of pure K 2 Cr 2 07
equivalent to the amount of chromium in the sample is dissolved in the acids
with the ingot iron.
DETERMINATION OF CARBON

See the Determination of Carbon in Ferrochromium by the Direct-Combustion Method. 13 A furnace temperature of 1050 to 1100 C. will suffice for
ferrovanadium and complete combustion except in very high-silicon material
can be secured without the addition of ingot iron and copper oxide.
DETERMINATION OF PHOS~HORUS BY THE MODIFIED MOLYBDATE-MAGNESIA
METHOD

Solutions Required: Ammonium Molybdate.Solution No. 1.-Place in a beaker 100 g. of 85% molybdic acid, mix thoroughly with 240 ml. of distilled water, add 140 ml. of NH 40H (sp.gr. 0.90),
fllter, and add 60 ml. of HN0 3 (sp.gr. 1.42).
Solution No. 2.-Mix 400 ml. of HN0 3 (sp.gr. 1.42) and 960 ml. of distilled
water.
When the solutions are cold, add Solution No. 1 to Solution No.2, stirring
consantly, then add 0.1 g. of ammonium phosphate dissolved in 10 ml. of
distilled water, and let stand at least 24 hours before using.
Magnesia Mixture.-Dissolve 50 g. of MgCI 2 6H 2 0 and 200 g. of NH 4 CI in
1000 ml. of distilled water containing a few milliliters of hydrochloric acid.
Method.-Transfer 4 g. of the sample to a platinum dish of 150 to 200-ml.
capacity, add 70 ml. of dilute HN0 3 (5 : 6) and heat the covered dish until
vigorous action ceases. Add 1 or 2 ml. of hydrofluoric acid, stir the solution,
evaporate to dryness, moisten with 10 ml. of nitric acid (sp.gr. 1.42) and again
evaporate to dryness. Moisten the cooled residue with 10 ml. of HNO s (sp.gr.
1.42), add 50 ml. of water and wash the contents of the dish into a 400-m!.
beaker, rubbing the dish well with a policeman. Remove the last traces of
adhering salts with 5 m!. of HCI (sp.gr. 1.19) and hot water. Add 10 ml. of
HN0 3 (sp.gr. 1.42) and evaporate to a syrupy consistency. Cool, add 25 mL
of HNO s (sp.gr. 1.42) 60 ml. of water, 1 g. of boric acid (to take care of any
residual fluoride) and heat gently until solution occurs. Filter through a rapid
paper into a 300-ml. Erlenmeyer flask and wash with hot water. Heat the
solution to boiling, add about 12 mL of potassium permanganate solution
(1.5%) and continue the boiling until manganese dioxide precipitates. Dissolve
the precipitate by the addition of sufficient sulfurous acid added dropwise.
About 1 m!. usually suffices. Boil for a few minutes to completely expel
oxides of nitrogen. Add to'the cold solution 15 ml. of ammonium hydroxide
(sp.gr. 0.90) and cool the solution to 15 C.
13

See p. 1493.

1496

FERRO-ALLOYS

Next reduce the vanadium in the cold (15 0 C.), solution by ferrous sulfate
added in approximately 1.5 g. excess.. The use 6f~ a 50% solution of the salt
furnishes a convenient way of adding the necessary amount. One gram of
quinquevalent vanadium requires 5.5 g. of FeS0 4 17H 20 for reduction to the
quadrivalent state.\ A 4-g. sample of 30% vanadiub alloy, including an excess
of 1.5 g., would require 8.1 g. of ferrous sulfate. Too large'an excess of ferrous
sulfate should be avoided. The addition of ferrous sulfate will change the
color of the solution to a deep blue. Add 2 drops of a saturated solation of
sulfurous acid, 100 mj.. of molybdate reagent, shake for 10 minutes and allow
to stand at 150 C. for 4 to 6 hours, or preferably overnight. It is w.ell to let
the solution stand in a refrigerator. Filter ~n a paper of close texture, keeping
as much of the precipitate in the flask as possible, and wash the flask and the
precipitate with not more than 50 ml. of cold, dilute nitric acid (1 : 9~). Set
the filtrate and washings aside after thorough mixing and see whether further
separljotion of phospho molybdate occurs. Dissolve the precipitate on the filter
in 20 ml. of dilute ammonium hydroxide (1 : 1) to which have ,been added 2 g.
of citric acid, catch the solution in the original flask containing the bulk of the
yellow precipitate and finally wash the filter several times with dilute hydrochloric acid (5 : 95). If the ammoniacal solution of the phospho molybdate is'
not clear at'this point, heat it to boiling, ~lter through the same paper, and
wash the paper with hot wll.ter. In this case, and whenever appreciable amounts
as such elements as titanium or zirc_onium are known to be present In the
original material, preserve the paper and ignite it in a small platinum crucible
with the paper reserved below.
Render the ammoniacal solution acid with hydrochloric acid, add 20 ml.
of magnesia mixture and then ammonium hydroxide (sp.gr. 0.90) slowly until
a crystalline precipitate appears, and then finally an excess of 3 to 5% by
volume. The volume of the solution at this point should not exceed 100 ml.
Allow the solution to stand in a cool place for 4 to 6 hours, or preferably
overnight.
Filter, keeping as much of the precipitate as pqsible in the flask, and wash
the flask and the' precipitate moderately with dilute ammonium hydroxide
(1 : 20). Set the filtrate and washings aside after thorough mixing and see
whether further precipitation occurs. Dissolve the precipitate on the filter
in 25 ml. of dilute hydrochloric acid (1 : 1), catching the solution in the original
flask containing the bulk of the precipitate. Wash th~ filter thoroughly with
dilute hydrochloric acid (5 : 95). In case the original material contains appreciable amounts of the elements mentioned above (titanium or zirconium),
add the paper to the one reserved above, ignite, fuse any.residue with as little
carbonate as possible (not over 0.5 g.), extra~t the cooled melt with hot water,
filter, and add the water solution to the hydrochloric acid solution. Transfer
the solution to a 200-ml. beaker. In case the amount of arsenic is known to be
very small, add 0.5 to 1 g. of ammonium bron:{ide and boil the hydrochloric
acid solution to a volume of 5 to 10 ml. Where large amounts of arsenic were
originally present, treat with hydrogen sulfide, digest, filter, and wash with
acidulated hydrogen sulfide water. ,Boil to expel hyd~ogen sulfide and to
reduce the volume to 50 to 75 ml. Add 0.1 to 0.2 g. of citric acid, 2 to 3 ml. of
magnesia mixture, make ammoniacal as above', and allow to stand in a cool
place for 4 to 24 hours.

FERRO-ALLOYS

1497

Filter, wash with dilute ammonium hydroxide (1 : 20), and ignite the
precipitate carefully and at as Iowa temperature as possible until the carbon
has been destroyed and the residue is white. Finally, ignite at approximately
1000 C. to constant weight. Dissolve the ignited residue in 5 ml. of dilute
nitric acid (1 : 1) and 20 ml. of water. Filter, wash the residue with hot water,
ignite and weigh. Treat with a few drops of hydrofluoric acid, evaporate to
dryness, ignite and re-weigh. Subtract any loss of w~ight from the original
weight of the pyrophosphate, and calculate on the basis of this purified magnesium pyrophosphate.
.
DETERMINATION OF SULFUR BY THE OXIDATION METHOD

Transfer 4.57 g. of the sample to a 600-ml. beaker and dissolve it in 75 ml.

of HNO a (sp.gr. 1.42). The reaction is apt to be very rapid and the acid should
be added in small portions to the covered beaker which is preferably kept
.cooled by immersion in ice water. Effect complete solution by gently warming
if necessary.
Add 30 ml. of HCI (sp.gr. 1.19) and evaporate to dryness on the steam-bath.
Add 30 m!. more of HCI (sp.gr. 1.19), again evaporate to dryness, and bake for
30 minutes at 105 C. Moisten the residue with 30 ml. of HCI (sp.gr. 1.19)
and evaporate on the steam-bath to a volume of 5 ml. Add 5 ml. of HCl
(sp.gr. 1.19), dilute to 150 ml. and filter. Wash with warm water.
Adjust the volume of the solution to 250 ml. which should now contain'
about 3 to 4% by volume of HCI. 'Warm to 75 C., add 30 ml. of BaCh
(10%), and allow to. stand for 24 to 40 hours.
Filter, with as little transfer of the precipitate as possible, through a paper
of close texture. Wash the paper two or three times with cold 1 % HCI and
seven times with warm water. Ignite the paper and contents in a platinum
crucible. Treat the ignited residue with 5 ml. of HCI (sp.gr. 1.19) and transfer
the resulting solution to the original beaker containing most of the precipitate
of BaS04. Digest the combined portions of BaS04 so as to dissolve any barium
vanadate. Adjust to 15Q-ml. volume and 2% acidity, add 8 ml. of 10% BaCh
solution and allow to stand for 24 hours. Filter on a paper of close texture,
wash three times with cold 1 % HCI and 10 times with warm water. Ignite
the paper in platinum, cool in a desiccator and weigh. A blank should be
carried through all steps of the determination. The corrected weight of the
ignited barium sulfate multiplied by three is equal to the percentage of sulfur.
NOTES ...;__If desired, the washings can be kept separate and examined for BaS04 as
follows: Evaporate the washings to dryness. Dissolve the slight residue in 50 mL of hot
dilute HCI (2 : 98), add 2 Iril. of BaCh solution (10%), digest at 70 to 80 C. for two hours,
avoiding any undue evaporation. Filter on a small paper of close texture and wash as
above. The recovery of BaS04, which is ordinarily obtained, represents approximately
0.001 to 0.002% of sulfur.
In high-silicon alloys the residue from the initial acid digestion should be evaporated
with hydrofluoric and nitric acids; fused with sodium carbonate and the water extract
of the melt added to the main solution.

DETERMINATION OF SILICON BY THE NITRO-SULFURIC METHOD

Solutions Required: Nitro-Sulfuric Acid.-Pour 160 ml. of sulfuric acid,

esp.gr. 1.84) slowly and with stirring into 660 m!. of water, and add 180 ml. of
nitric acid (sp.gr. 1.42).

1498

FERRO-ALLOYS

Method.-Dissolve 0.9344 g. of the sample passing a No. 100 sieve of low

or medium-silicon ferrovanadium in 40 mr. of the nitrosulfuric acid in a 250-m!.
porcelain or platinum dish and evaporate the solution until" copious" fumes
\
.'
of sulfuric acid are evolved.
Ferrovanadium containing 4% or more of silicon is not completely sGluble
in nitric acid and sulfuric acid. In this case, fus~ 0.9344 g. of the alloYI thoroughly mixed with 8 g. of sodium peroxide, in a 30-m!. crucible of pure iron.
Five minutes heating at a dull red heat after the contents of the crucible have
melted will insure complete decomposition of the sample. Allow the fusion to
solidify, but before it has entirely cooled tap the covered crucible on an iron
plate which will loosen the contents in a solid cake. Transfer the melt to a
250-m!. porcelain casserole, rinse the crucible with warm water and cautiously
add the solution to the covered casserole. Next add 150 m!. of sulfuric acid
(20% by volume) and evaporate to copious fumes of sulfuric acid. Cool the
sulfuric acid solution obtained in either procedure, add 125 m!. of water and
heat for a few minutes with frequent stirring until all salts are dissolved. Immediately filter on a 9-cm. paper. Wash the precipitate with cold dilute HCl
(5 : 95) and cold water alternately to complete the removal of iron salts, and
finally with hot water until free from acid: Transfer the filter to a platinum
crucible, char the paper without inflaming,and finally ignite the covered cr.ucible
in a muffle furnace at 1100 0 C. for 15--20 minutes. Cool in a desiccator and
weigh. Add sufficient dilute H 2S0 4 (1 : 1) to moisten the silica and then a
small amount of hydrofluoric acid. Evaporate to dryness, ignite and weigh.
The difference in weights divided by 2 and multiplied by 100 represents the
percentage of silicon in the alloy.
DETERMINATION OF ALUMINUM BY THE ETHER-HYDROCHLORIC ACID METHOD 14

Dissolve 2 g. of the sample in 25 to 30 m!. of dilute nitric acid (sp.gr. 1.135)

(Note 1). When action ceases, add 5 m!. of hydrochloric acid and evaporate
to dryness. Add 10 m!. of hydrochloric acid, evaporate to dryness and bake.
the residue at a moderate temperature. Dissolve the residue in 10 m!. of
hydrochloric acid, add 30 m!. of water, boil and filter. Wash the residue with
hot dilute hydrochloric acid (1 : 9) followed,by water (Note 2).
Evaporate the solution to dryness on a steam-bath, add 20 m!. of hydrochloric acid and warm gently until complete solution is obtained. Cool the
solution in ice water and pass hydrochloric acid gas intQ it to saturation. Add
20 ml. of ether and cdntinue passing the gas into the cold -solution until it is
. again saturated. Allow the mixture to stand for several hours, keeping it cold
by immersing the container in ice water. If the solution becomes warm or if
it is not allowed to stand, the separation of hydrated aluminum chloride is not
quite complete.
Filter through asbestos (small Biichner-asgestos pad washed once with
cold ether-hydrochloric acid mixture immediately before use), wash, by decantation, with a cold mixture of equal parts of strong hydrochloric acid and
ether which has been saturated with hydrochloric acid gas. The aluminum
I
chloride retains a little vanadium.
Dissolve the precipitated aluminum chloridE}..in the .least possible amount
of dilute hydrochloric acid (1 ; 9) back into the'QFiginal beaker, from which all
14 Gooch-Havens, Am. Journal of Science, 4,11-416.

FERRO-ALLOYS

1499

of the precipitate need not have been removed. Evaporate the solution to
near dryness on a steam-bath, add 10 m!. of strong hydrochloric acid and pass
hydrochloric acid gas into the solution to saturation, add 15 ml. of ether and
continue passing hydrochloric acid gas to saturation. keeping the solution cold
by means of ice water. Let stand several hours, filter on asbestos and wash
with the ether-hydrochloric acid mixture as before.
Dissolve the aluminum chloride in dilute hydrochloric acid (1 : 9) into a
250-m!' beaker, add a few grams of ammonium chloride, make faintly alkaline
with ammonia, boil, filter (using paper pulp), wash with hot water, dry, ignite /
and weigh as AI 20 a.
According to published data, this method gives a complete separation of
aluminum from iron, beryllum, zinc, copper, mercury and bismuth. A little
vanadium is carried down in the first separation, but two chloride precipitations
seem to give a practically complete separation. The proportion of vanadium
retained in the final alumina is hardly appreciable when working on a 35%
vanadium alloy.
The effect of chromium is not known; presumably some of it would be
included in the AhOa. This could be corrected for by fusing the Al 20a with
sodium peroxide and estirnnting the Cr 2 0a by color.
NOTEs.-1. Dilute nitric acid (sp.gr. 1.135) is recommended for solution. While this
requires rather long digestion at steam-bath temperature, solution seems to be quite
complete even when the alloy is high in silicon., In some cases where silicon is very
high it is necessary to use a little hydrofluoric acid with the nitric acid. In this case
solution should be effected in a platinum dish, the solution evaporated to dryness and
the dry residue transferred to a beaker before treatment with hydrochloric acid. A little
dilute hydrochloric acid may be used to loosen any oxides that adhere to the dish.
2. The siliceous residue may retain alumina; it should be ignited in platinum. Silica
should be expelled by HF+H 2SO. and the residue fused with KHSO.. After dissolving
the fusion in water plus a little HCI, the liquid is made faintly alkaline with ammonia
and boiled. Any iron and alumina precipitate is removed by filtration, washed with
water, dissolved in a little hot dilute HCI and added to the solution of the alumina obtained after the first ether-hydrochloric acid separation.
FERROTUNGSTEN AND TUNGSTEN METAL
DETERMINATION OF TUNGSEN

Solutions Required: Cinchonine Solution.-Dissolve 125 g. of cinchonine

in a mixture of 500 m!. of HCI (sp.gr. 1.19) and 500 ml. of distilled water.
Cinchonine Wash Solution.-Dilute 30 mL of the above solution to one liter.
Method.-Treat 1 g. of the material passing a No. 100 sieve in a large (50-m!.)
platinum crucible with 5 ml. of HF; add HNO a drop by drop until the metal dissolves. Add 3 to 4 mLof H 2S0 4 and evaporate on a steam-bath until HNOa+HF
is expelled. Shake gently over a small Bunsen flame, or heat slowly in a sheet
iron "radiator" (described by Hillebrand) 15 until H 2S0 4 fumes strongly.
Cool, transfer to a 250-mL beaker with water, finally wiping the crucible with a
little filter paper. A little WO a sticks to the crucible; it can not be removed
by wiping. Reserve the crucible. Dilute the contents of the beaker to about
150 m!. with water, add 10 roL of hydrochloric acid (sp.gr. 1.19) and boil for a
few minutes. R~roove from the source of heat, add 10 ro!. of cinchonine solution
16 U. S. Geological Survey Bulletin No. 700, p. 33 ..

.. gric;:ltnal

a')"~~;;

Bapatltl..

1500

Lnmny,
"'-FERRO-ALLOYS

and digest the solution at 80 to 90 C. with occasional stirring, for 30 to 45

minutes..
I
Add some paper pulp to the solution, stir the solution and when the tungsten
precipitate has settled, filter on a 9-cm. paper c,ontaining a little "ashless"
filter pulp. Wash thoroughly with hot cinchonine solution and finallyseveral
times with warm 1% hydrochloric acid. Gently ignite the filter and residue of
crude WO a in the crucible in which it was originally treated until the carbon is
consumed. Add a few drops of nitric acid and dry in the radiator. Ignite the
covered crucible for 5 minutes in the full heat ot a Bunsen burner, cool, and
weigh. Ignitions may be made at about 750 C. in an electric muffle. As
WO a is slowly but steadily volatilized at temperatures above 750 0 Q.,.serious
error may result from continued ignition of the tungsten trioxide at the temperature that is specified. Add about 5 g. of Na 2CO a and fuse, running the
fusion around the side of the crucible to remove all WO a Dissolve the fusion
in hot water, add alcohol, heat, filter and wash with hot water. Place the filter
in the crucible and ignite; add a little N a 2 CO a and fuse again. Dissolve the
fusion in water, filter and wash very thoroughly with hot water to remove the
last traces of Na 2CO a ; ignite in the sam~ crucible as at first, cool, anD weigh.
The difference between the weight of crucible plus residue and the weight of
crucible plus impure tungstic oxide is WO a XO.7931 = W.'
NOTES.-At the best, direct determination of tungsten in high-grade metal is subject
to inherent errors. Accuracy within 0.2% is all that can be expected by ariy method.
DETERMINATION OF CARBON

See the Determination of Carbon in Ferrochromium by the Direct-Combustion Method. 16 The alloy ground to pass a No. 100" sieve and mixed with
2 g. of pure iron is burned in an electric furnace heated at 1050 0 to 1100 0 C.
A " blank" on the ingot iron must be determined and deducted.
I

DETERMINATION OF PHOS/HORUS

Treat 1 g. of the finely powdered sample in a platinum dish, fitted with a

gold cover, with 15 mI. of HNO a (sp.gr. 1.42), add 3 mI. of HF and warm gently.
When action subsides add 3 ml. more of HF. After action subsides, boil,
remove cover and if decomposition is not complete add more HF and boil again.
When solution is complete, wash off the cover and evapora~e at a 10": heat to
a volume of about 10 mI., then add 3 or 4 drops of concentrated permanganate
solution and continue evaporation until crusts of WO a begin to form at the
edges, that is to a volume of about 6 m!. Add 5 mI. of H 2S0 4 and evaporate
on the stove at a low heat until HF and HNO a are expelled-and H 2S0 4 fumes
are given off. (Strong heat causes spattering and also causes hard over-baked
crusts to form on the bottom of the dish whi~h resist subsequent treatment.)
Cool, add 25 inI. of water and boil (by agitating over a Bunsen flame) until all
soluble salts are dissolved. Discharge pink color, due to excess permanganate,
by adding sulfurous acid drop by drop. The pink color may not be very evident
but the 80 2 is added even so to reduce higher oxides of manganese. Boil for
a minute or two after adding the 80 2 Ac!cGi':20 g. of pure t~rtaric acid and,
16

See p. 1493.

FERRO-ALLOYS

'1501

when this is dissolved and the solution is cooled to a temperature of about"

50 C., add 20 ml. of NH 40H (sp.gr. 0.90) diluted with an equal volume of
water. The precipitated tungstic acid should dissolve completely, giving a
clear solution. The solution becomes hot from the reaction between H 2S0 4
and NH 4 0H. While it is still hot, add 10 ml. of magnesia mixture and transfer
it from the platinum dish to a 6-oz. glass-stoppered bottle. Set the bottle in
ice water and when it is thoroughly cooled add four or five glass beads about
6 mm. in diameter. Stopper it tightly and shake in an efficient shaking
machine for at least ten minutes. The agitation should be violent. The
beads aid in starting the formation of the magnesium precipitate; after agitation
add 15 ml. of strong NH,OH and retur:r!" the bottle to the ice-water tank and
put into a refrigerator to stand overnight. Phosphorus separates as magnesium
ammonium phosphate free from tungsten but containing possibly basic magnesium compounds. After standing overnight filter the solution through a
9-cm. paper containing a little paper pulp and wash the bottle and paper
thoroughly by small additions of ammonia wash water (1 part NH 40H to
3 patts H 20). Do not attempt to remove all the precipitate from the bottle
but. remove the beads to the filter. The precipitate consists of magnesium
ammonium phosphate and arsenate together with silica and other impuritieS: Tin, tungsten, molybdenum, vanadium and titanium are eliminated
}f precipitation with magnesia mixture in the presence of tartaric acid.
Dissolve the magnesia precipitate in hot dilute HCI and evaporate. to a small
volume, say 7 or 8 ml. Add 10 ml. of strong HCI and 0.5 g. of KBr and evaporate to dryness. Take up with 5 ml. of HNO a and evaporate to dryness again.
The KBr serves to reduce Asv to As"'. The addition of strong HCI and boiling
followed by evaporation to dryness eliminates As. Dissolve residue in 60 ml.
of HNO a (sp.gr. 1.135), filter into a 250-ml. shaking bottle; add 40 ml. of
NH 40H (sp.gr. 0.96), cool to 33 C., add 30 ml. of molybdate solution (Blair)
and finish the determination by the alkalimetric method (Handy) described in
Blair's Chemical Analysis of Iron.
NOTEs.-The usual proportions of HNO a and HF are reversed in order to provide a
constant excess pf HNO a to oxidize P. The procedure given takes a little longer than
when the sample is treated with HF first and HNO a is added a little at a time, but the
solution is finally complete. It is necessary to keep the platinum dish covered after
action begins, as the reaction is somewhat violent.
The platinum dishes recommended are:

8 cm. in diameter at top;

7.8 cm. in diameter at bottom;
4 cm. high.
They have flat bottoms and are wire-rimmed at the top to give additional stiffness.
They weigh 58 to 60 g. each and hold about 175 mI. The covers are made of pure gold
(for economy), "dished" like a crucible cover to fit the top of the dishes closely. The
dishes have small lips to aid pouring. The "tongue" of the cover overlaps the lip.
Ordinary round-bottomed dishes may be used, but the manipulation is much more
difficult; there is a greater tendency to spattering and danger of local baking or overheating in evaporating to fumes of H 2SO,. If the separated WO a is oyerheated locally
it does not dissolve readily in NH,OH. In flat-bottomed dishes, the WO a is spread in a
thin layer and heat is applied evenly all over the bottom.
A good shaking machine should be used. The magnesia precipitate may be started
by shaking the bottles by hand, but it is a tiresome job. The solution must be cold, say
go or 100 C., when shaking begins.

1502'

FERRO-ALLOYS

Permanganate solution is added to insure complete' oxidation of phosphorus, as in

steel analysis. The color of the permanganate_gradually fades in the hydrofluoric nitric
solution, but after evaporating to fumes and adding water the solution is usually slightly
~n~
.
The amount of tartaric acid is limited to 1.20 g. si~ce ammonium tartrate retards
the formation of the magnesia precipitate. Complete precipi~ation can only be obtained by brisk agitation and by keeping the solution very1cold, followed by long standing
in a cold place. By this method determinations started at 3 p.m. may be completed by
noon the next day.
DETERMINATION OF SULFUR AND SILICON

Solutions Required~ Cinchonine Solution.-Thoroughly wash cinchonine

reagent with hot water on a BUchner funnel, dry, and dissolve 125 g. of the
washed cinchonine in a mixtm:e of 500 ml. of HCI (sp.gr. 1.19) and 506 ml. of
distilled water.
Cinchonine Wash Solution.-Dilute 30 ml. of the above solution to one liter.
Barium Chloride Solution.--:-One hundred grams of BaCI 2 2H 20 to 1000 ml.
SULFUR

Fuse 2 g. of the finely powdered metal, intimately mixed with 8 g. vf dry

Na 2 C0 3 and 2 g. of powdered KN0 3, in:a large platinum crucible (40 mI.).
The fusion is best done over an alcohol blast flame' or in an electric muffle to
avoid contamination with sulfur contained in illuminating gas. Leach the
fusion with 100 mI. of distilled water in a 250-ml. beaker. Filter into.a 12-cm.
porcelain casserole and wash the residue several times with hot distilled water.
The solution contains practically all of the sulfur and most of the tungsten
and silica. The residue contains some of each, but not more than traces of
sulfur.
Render the liquid acid with 20 ml. of HCI and evaporate to dryness on a
steam-bath, thus precipitating tungstic acid. Mter the residue is dry, heat
it to 1100 to 1200 C. in an oven to dehydrate silica, take up with 5 mI. .of HCI
and 50 mI. of water, add 3 ml. of a strong cinchonine solution and boil, or heat
just short of the boiling point, for 30 minutes. ~et the solution stand for at
least three hours and preferably overnight. By {means of the cinchonine, all
W0 3 is rendered insoluble. Filter through: an 11-cm. filter paper (free from
ash) which has been liberally treated with an emulsion of filter paper pulp and
wash the residue with dilute cinchonine solution. Reserve the residue. The
filtrate contains all the sulfur; heat it to boiling and add 10 ml. of barium chloride
solution, boil for ten minutes, allow it to stand overniglit, filtl2r on a 9-cJll. filter
paper (free from ash), wash repeatedly but not excessively with hot water,
ignite in platinum and weigh as BaS04. If the work is properly carried out
every trace of W0 3 is removed from the solutions and there is no danger of
including Ba W0 4 with the BaS04. If one suspects that the BaS04 is not quite
pure, it should be fused with a little Na 2 C0 3 , lejl-ched with water and filtered;
the filtrate should be acidulated with HCI, a few drops of cinchonine solution
. added and the solution evaporated to dryness on a steam-bath. The dry
residue should be dissolved in a little water, a few drops of HCI added, and
any residue filtered off. In the clear filtrate BaS04 should be again precipitated
as before. The weight of BaS04XO.1374=S.
' .
.
" Blanks" on all reagents must be made parallel with the determination;
this is important since all sodium carbonate ootainable contains sulfur and

FERRO-ALLOYS

1503

there is sometimes more or less of it in cinchonine. All evaporations should

be made over steam and the operations conducted in a place free from sulfur
gases.
NOTE.-Evolution of sulfur in a stream of hydrogen saturated with hydrochloric
.
acid as in Johnson's Method is optional.
SILICON

The residue of tungstic oxide from which the solution for the determination
of sulfur was filtered contains most of the silica. The residue from leaching
the original fusion contains the rest.
Wash the residue from Na 2 CO a fusion from the filter paper with a fine jet
of water into a small casserole, add an excess of HCI and evaporate to dryness
on a steam-bath, heat the dry residue at 110 to 120 C. to dehydrate the silica,
drench with hydrochloric acid, dilute to 100 ml. with hot water, heat until
salts are in solution, and filter on a small filter paper (free from ash). Wash
thoroughly with hot dilute HCI (1 : 100) and finally with hot water. Reserve
the filter.
Wash the residue of tungstic oxide, silica and filter pape~ pulp from the
filter into a small casserole, add 5 ml. of HCI and heat for a few minutes, then
filter again through the same paper and wash it thoroughly with hot water.
This second washing is necessary to remove sodium salts completely. (The
original washing sufficed to remove all of the small amount of sulfur present.)
Transfer the well-washed filter and residue to a platinum crucible, add the
small filter containing the little silica recovered from the N a 2CO a fusion residue
and ignite both to constant weight in a platinum crucible. On account of
the presence of filter paper pulp, the ignited residue is porous and friable; if
paper pulp had not been used the WO a would be dense and not easily susceptible
to the subsequent treatment. Cool the ignited residue of WO a and Si0 2 and
weigh it. Add 2 or 3 drops of H 2S0 4 and 5 or 6 ml. of pure HF. Digest at a
gentle heat for some time, and then slowly evaporate off the HF. When
H 2S0 4 fumes are evolved, cool and again add HF, digest as before, evaporate
off the HF, ignite gently to expel H 2S0 4 and heat strongly for ten or fifteen
minutes. Cool and weigh. The loss in weight after expulsion is Si0 2 which
multiplied by 0.4672 equals silicon.
0

FERROTUNGSTEN

DETERMINATION OF TUNGSTEN

Proceed as in the Determination of Tungsten in the analysis of Ferrotungsten and Tungsten MetalY
NOTES.-l. Fuming with perchloric acid can be substituted for fuming. with sulfuric
acid if desired. In this case, add 10 ml. of perchloric acid (60%) after the treatment
with HF and HNO a and evaporate to strong fumes of perchloric acid. Cool, add 50 ml.
of water, transfer the solution to a 400-mI. beaker and proceed as in the case of sulfuric
acid.
2. If the alloy contains molybdenum, the weighed tungsten oxide may contain a
few hundredths of a per cent of molybdenum as MoO a If extreme accuracy is desired,
the oxide should be tested colorimetrically. If the molybdenum content of the alloy
17 See p. 1499.

FERRO-ALLOYS

1504,

is unknown, one of the sodium carbonate'extracts of the fused impure WO s can be treated
with tartaric acid, mineral acid and hydrogen sulfide and then, if appreciable molybdenum is indicated, the amount shollid be determined inl the sodium carbonate extract
obtained in the duplicate run.

3, At the best, direct determinations of tungsten in .p.igh-grade metal is subject to

inherent errors. A_ccuracy within 0.2% is all that can bel expected by any method:
DETERMINATION OF SILICdN

Fuse 0.9344 to 1.0416 g. of the sample as described under Determination

of Sulfur and Silicon in the Analysis of Ferrotungsten and Tungsten Metal,18
Leach the fusion with water, acidify with HCl and evaporate to dryness and
bake at 110 0 C. Add 20 ml. HCl (sp.gr. 1.19) and warm until all iron salts
appear to have dissolved. Add 200 ml. of boiling water and a little paper pulp.
Stir well and filter. Wash the tungstic acid thoroughly with hot dilute HCl
(1 %). Ignite until carbon is gone, add a drop or two of H 2S0 4 , evaprate,
and then ignite first at a low te.mperature and finally at 1000 0 to 1100 0 ~for
15 to 30 minutes. Cool the ignited WO s and Si0 2 and weigh it. Add 5 or-6
ml. of pure HF and a drop or two of dilute H 2S04 (1 : 1), evaporate, and again
ignite at 750 0 q. until constant weight is obtained. The loss in weight, corrected for any Si0 2 in the reagents, multiplied by 100 and divided by the weight
of sample taken gives the percentage of silicon.
DETERMINATION OF SULFUR

Transfer 5 g: of the sample to a 300-ml. platinum dish provided with a

platinum cover. Add 100 ml. of HNO a (sp.gr. 1.42) and add HF a few drops
at a time, while heatirrg, until the alloy has dissolved completely. Rinse and
remove the cover. Add 1 g. of NaNO a and 30 ml. of HCI0 4 (60%) and evaporate to strong fumes of HCI0 4 Cool, rinse down with water, and again
evaporate to fumes.
Add 100 ml. of water, 10 m}. of cinchonine solution (125 g. of cinchonine,
sulfur-free, dissolved in 1 liter of 1 : 1 HCl) and' some ashless paper pulp.
Stir well, digest for 15 minutes at 60 0 to 70 0 C. ana filter. Wash the tungstic
aci<j. with hot HCl (2%) containing 10 ml. of cinchonine solution to 1 liter.
Precipitate the sulfur as described under Determination of Sulfur in Ferr<;>molybdenum. 19
DETERMINATION OF COPPER, ANTIMONY AND l'IN

Mix 5 g. of the finely powdered metal with 25 g. of Na 202 in a 50-ml. iron

crucible. Fuse carefully over a free flame by holding the crucible with It pair
'of tongs and slowly revolving it around the outer edge of the flame until the
contents have melted down quietly. Do not raise the temperature rapidly
enough to cause spattering. When the charge, is molten, keep the crucible
at a low red heat and impart a slight rotary motion to stir up any unattacked
particles. . At the end of 4 to 6 minutes increase the temperature to bright
redness for a minute. Cool the crucible with the cover tightly on and as soon
as the melt has solidified, tap the still slightly warm crucib,le several times on an
18
19

See p. 1502.
See p. Ui08.

FERRO-ALLOYS

1505

iron plate in order to loosen the melt in ~ solid cake. Transfer the melt to a
600-ml. beaker, cover, treat with 300 ml. of warm water, rinse the crucible
and add the rinsings to the beaker. Heat to boiling, cool somewhat, add 20 g.
of tartaric acid and then 50 m!. of dilute H 2S0 4 (1 ; 1). Heat until a clear
solution is obtained. If iron scale remains insoluble, decant the liquid, dissolve the scale in strong acid, using a drop of HNO a if necessary, and add the
solution to the main solution. If unattacked tungsten metal remains, filter,
and make a second fusion on the small residue.
(a) Separation of Copper from Antimony and Tin.-Adjust the volume of
the solution to a volume of 400 ml., heat to from 65 to 70 C. and pass a rapid
stream of H 2S through the solution for 30 mInutes. Filter, and wash the
precipitated sulfides with'dilute H 2S0 4 (5 : 95) containing about 1 g. of tartaric
,acid per 100 ml. and saturated with H 2S. Return the paper and sulfides to
the beaker in which the precipitation was made and treat with 50 ml. of I}OHK 2S solution (prepared by saturating 1 volume of a 10% solution of KOH
with H 2S and then adding 3 volumes ,of the solution). Add 1 g. of N a 2 02 and
heat to boiling in the course of 5 minutes and as the solution is occasionally
stirred. Dilute with an equal volume of water, filter into a 400-ml. beaker,
'and wash with a diluted K 2S solution (1 ; 10); reserve the filtrate (A).
(b) Copper.-Dissolve the residue and determine copper colorimetrically
or electrolytically as described under Determination of Copper in Ferromolybdenum. 2o
.
(c) Separation of Tin and Antimony.-Treat the reserved filtrate (A) with
HCI until the solution is acid and then add 1 ml. per 100 ml. in excess. Filter,
and wash with acidulated hydrogen sulfide water. Transfer the paper and
sulfides to a 400-ml. beaker, treat with 10 ml. of HCI and add KCIO a, a few
crystals at a time, as the solution is warmed to from 35 to 40 C. Dilute to
2GO m!. and boil gently to expel chlorine. Add 5 g. of oxalic acid, heat to
from 60 to 70 C. and pass in a rapid stream of H 2S for 20 to 30 minutes.
Filter, and wash thoroughly with hydrogen sulfide water containing 0.5 ml.
of sulfuric acid, and a few crystals of oxalic acid per 100 m!. Reserve the
filter, paper and precipitate (B) for the determination of antimony.
(d) Determination of Tin: Standard Solution of Iodine.-Dissolve 3.97 g.
of iodine in a solution of 7 g. j)f KI in a little water. When solution is complete,
dilute to 1 liter and mix thoroughly. One ml. of the solution should represent
approximately 0.00185 g. of tin and the exact titer is found by dissolving 0.02 g.
of pure tin in HCI, treating with lead as in the method, and titrating. For
the small amounts of tin involved, the theoretical titer as based on standardization against pure arsenious oxide can be used.
Apparatus for Reduction of Tin.-The apparatus for reduction is shown in
Fig. 186. At the point A a small hole about 1 mI\l. in diameter shall be blown
in the glass. The glass rod B shall be drawn out at the end and covered with
rubber tubing so as to form a light fitting stopper for the end of the tube C.
Treat the filtrate with 5 ml. of H 2S0 4 and evaporate to heavy fumes of the
acid. Cool somewhat, and dilute with 100 m!. of water. Add 50 m!. of a ,
solution of ferric sulfate (1 %) and heat to boiling. Stir vigorously and add
dilute NH 4 0H (1 ; 1) until the solution is alkaline and then add 3 to 5 m!. in
excess. Let settle, filter, and wash with hot water. Dissolve the precipitate
20

See p. 1509.

1506

FERRO-ALLOYS

in 80 mI. of hot dilute HCl (1 : 1) and wash the filter with hot water, collecting
the solution (not over 150 m!. in volume) in a 300-ml:. Erlenmeyer flask. Add 1
to 2 g. of tes~ lead and close the flask
with a one-hole rubber stopper carrying
a bent deli\fery tube, as slJ.own in
Glass Rod
Fig. 186.
I
Boil gently for 20 minutes. At the
end of this period immerse the end" of
6the delivery tube in a small beaker"containing about 50 m!. of a strong solution
of NaHCO a and insert the rubber-tipped
tube at C, Fig. 186. Remove the flask
and beaker from the hot plate ant! cool
in a stream of cold wate~. When 'cool,
remove the stopper, add 5 m!. of freshly
prepared starch solution and titrate
FIG. 186. Apparatus for Reduction
with the standard solution of iodine to
of Tin.
a permanent blue tint. Subtract the
volume required in a blank run and multiply by the tin titer of the solution.
(e) Determination of Antimony.-Spread the reserved filter paper and precipitate (B) side down, on a small watch glass and place this over a beaker
containing a few centimeters of HCl and a' drop or two of bromine. Warm the
beaker until decomposition of the sulfide is complete. Rinse cover and paper
with warm dilute HCI (1 : 9), keeping the volume downto 25 m!. Evaporate
the solution to about 5 m!. on the steam bath and then add 1 g. of pure ferrous
sulfate, 0.5 g. of potassium chlori~, and 40 ml. of HCI. Evaporate the solution
on the steam bath to a volume of about 5 ml. Dilute with 10 ml. of water,
add 1 m!. of HNO a and heat until the iron is oxidized. Dilute to 100 mI. heat
to boiling, and add dilute NH 40H until in definite excess. Let settle, filter,
and wash with warm water. Dissolve the precipitate in 10 to 15 ml. of hot
dilute HCl (1 : 3) containing 0.1 g. of tartaric aci4~ Dilute to 100 mI., heat
and add ammonium bisulfite drop-by-drop to reduce the iron. Boil to expel
the excess of S02, cool and pass in a rapid stream of H 2S for 15 to 30 minutes.
Filter through a weighed Gooch crucible and wash,successively with acidulated
hlYdrogen sulfide water, alcohol, ether and carbon disulfide. Heat in a radiator
and add 'a drop of HNO a, from time to time to oxidize the sulfide. Finally
'
heat at about 600 0 C. and weigh as Sb 20 4
FERRO MOLYBDENUM
DETERMINATION OF CARBON BY THE DrRECT-COMBUSTION METHOD

Determine carbon by th~ direct-combustion mpthod described in the Tentative Methods of Chemical Analysis of Steel, Cast Iron, Open-Hearth Iron and
Wrought Iron (A. S. T. M. Designation: E 30-36 T) of the American Society
for Testing Materials 21 with the following modifications:
1. It is advisable to use a flux such as ingot iron or red lead, particularly
with low-carbon alloys. If fluxes are added, great_care shall be taken in ob'taining a proper correction for the blank.
21 See p. 1426.

FERRO-ALLOYS

1507

2. Because of the high sulfur maximum, special provision must be made for
the removal of oxides of sulfur in the exit gases. This can be done (a) by
passing the gases through platinized silica gel heated to 50.0. 0 C. and then
through a column of ironized asbestos; (b) by passing through a strong solution
of chromic acid or potassium permanganate and then through a desiccant
such as magnesium perchlorate; or (c) by inserting a 6-in. plug of ironized
asbestos in the exit end of the combustion tube and burning out the accumulated
sulfate regularly.
DETERMINATION OF MOLYBDENUM

Solutions and Apparatus Required: Ferric Phosphate Solution.-Dissolve

IDa g. of ferric sulfate in 1 liter of water to which 150. m!. of phosphoric acid
~sp.gr.

1.73) and 20. m!. of dilute sulfuric acid (1 : 1) have been added.
Standard 0.1 N Potassium Permanganate.-Prepare and standardize the
solution as described in the Tentative Methods of Chemical Analysis of Ferro
Alloys (A. S. T. ,M. Designation: A 104-36 T) of the American Society for
Testing Materials. 22 One ml. of a 0.1 N solution of KMn04 represents 0..00.67 g.
of sodium oxalate, and 0.00.32 g. of molybdenum.
Jones Reductor.-The reductor should contain a column of 20. to 3o.-mesh
amalgamated zinc 74 in. in diameter and 10 in. in length, resting on enough
asbestos to prevent particles of zinc from passing into the receiving flask. To
amalgamate the zinc, treat it with a solution of HgCl 2 (2%) for a few minutes
and then wash thoroughly with water. When idle, the reductor should always
be kept full to above the top of t~e zinc with distilled water.
METHOD

'Transfer 0..5 g. of sample to a 15o.-ml. beaker, cover, treat with 10. ml. .of
dilute HNO a (1 : 3), and warm to complete solution. If solution is difficult,
add a drop or two of HF. When reaction is complete, cautiously add 10. m!. of
dilute H 2S0 4 (1 : 1), and evaporate to fumes of the acid. Cool, add 40 m!. of
water and warm. Rinse and remove the cover. Filter and wash the paper
and silica 12 to 15 times with hot water, 3 to 4 times with hot dilute NH 40H
\ 1 : 3), and finally 4 to,,5 times with hot water, the washings being allowed to
run into the main filtrate. Discard the silica. Cool, add NH 40H until it is
difficult to avoid a red tint, and heat the still acid solution to boiling. Pour
the solution slowly and with vigorous stirring into 75 ml. of nearly boiling
dilute NH 40H (1 : 5) contained in a 5o.o.-ml. beaker. Rinse the beaker with
a little water, and then with a little hot dilute NH 40H and add to the main
solution. Add a little paper pulp, filter into a 5o.O-m!. beaker, and wash the
residue with hot water. Set the filtrate aside. Dissolve~the precipitate in a
slight excess of hot dilute H 2S0 4 (1 : 4), and riearly neutralize and pour into
NH 40H as before. Filter into the reserved filtrate. Dissolve the precipitate
and repeat the operation.
Treat the three combined ammoniacal filtrates ,"-1th 3 g. of powdered
tartaric acid. Stir until dissolved, and saturate the warm solution with H 2S.
If a precipitate appears, filter and wash with a dilute solution of (NH 4 )2S.
22

See p. 1490.

1508

FERRO-ALLOYS

Warm the filtrate, cover the beaker, arid add dilute 'H 2S0 4 (1 : 1) until the
solution contains 10 IIil. in excess for each 100 ml. Oflsolution. Let the solution
stand at the side of the steam bath (about 40' C.) for 15 minutes, and let Iilettle.
Filter, and wash the sulfide thoroughly with dilut~ H 2S0 4 (1 : 99) containing
.
10 g. of tartaric acid per liter and saturated with II'2S.23
Transfer the MoS a and paper to a 250-ml. beaker,. cover, and treat with (i) ml.
of H 2S0 4 and 10 mL of HNO a Cautiously heat 'until fumes of H S0 4 are
t
given off. Let cool, add 5 ml. of HNO a, again heat to fumes, and repeat the
treatment until the yellow color due to organic matter has disappeared. Cool,
rinse and remove the cover, rinse the inside of the beaker, and add a solution
of KMn04 (2.5%) very cautiously until ~ permanent red tint is obtained.
Again evaporate to fumes of H 2S0 4. Cool and add 75 ml. of water.' Boil
for a few minutes, add 2 g. of pure zinc shot (0.002% o~ iron or under), and
continue the boiling until any copper has been reduced to metal. Filter
through a filtering crucible and wash with hot water.
If the Jones reductor has been standing idle, pass 100 ml. of warm (40 to
50 C.) dilute H 2S0 4 (5 : 95) through it, then a little water, and discard the
solution. Add 35 ml. of the phosphate solution to the receiver, and then
enough water so that the tip of the reductor dips well beneath the solution
when the receiver is connected with the reductor. Draw the cool or slightly
warm solution of the molybdenum, which should be about 100 ml. in volume
and contain about 5 ml. of H 2S0 4, through the reductor as the receiving flask
is given slight rotary motion. Just before the surface of the liquid reach-es the
zinc, add a 50-ml. portion of water, and finally rinse by adding two more portions, each time just before the surface of the solution reaches the zinc. Close
the stopcock while a portion of the last washing solution remains in the reductor
funnel. Disconnect, and raise the reductor as a little water is allowed to run
through the stem, rinse the outside of the stem, and proceed with the titration
with the standard 0.1 N solution of KMn04. Run a blank, subtract the volume
of 0.1 N solution required, and multiply the remainder by the molybdenum
titer of the solution.
DETERMINATION OF' SULFUR

Transfer 2 to 5 g. of the sample to a 500-ml. casserole and dissolve it in 30

to 75 m!. of HNO a (sp.gr. 1.42). The reaction is likely to be very rapid and the
acid should be added cautiously to the covered casse/ole which is preferably
kept cooled by immersion in ice water.
....
Add 30 ml. of HCI '(sp.gr, 1.19) and evaporate to dryness, on the sand or
steam bath. Add 30 ml. more of HCI (sp.gr. 1.19), again evaporate to dryness,
and bake for 30 minutes at 105 C. Take up with 25 to 50 ml. of HCl (sp.gr.
1.19) and when solution is complete, dilute with water, and filter to remove
silica. Evaporate the filtrate to from 15 to 25 m!. to expel most of the 'free
acid, and dilute to 100 ml. with warm water. Add a little paper pulp, filter,
and wash, first with dilute HCl (1 : 99) and then with water. IJilute to 200 ml.,
add 20 ml. of a solution of BaCl 2 (10%), and stir vigorously for several minutes.
Allow to stand overnight and filter on a paper of close texture. Wash the
paper and BaS04, 18 to 20 times with cold Hel (1 %), 3 or~' times with NH 40H
(5%) to remove ammonium phosphomolybdat_;,jgnite and weigh., A blank
should be run on all reagents used.
-_
23 Test filtrate for molybdenum by the colorimetric method, p. 1464.

FERRO-ALLOYS

1509

DETERMINATION OF PHOSPHORUS

Dissolve 2 g. of the finely ground sample in 20 ml. of HNO s, (sp.gr. 1.42)

and 20 ml. of HCI (sp.gr. 1.19). Add 20 ml. of dilute H 2S0 4 (1 : 1) and evaporate to fumes of sulfuric acid. Cool somewhat, add 75 ml. of dilute HCI
(1 : 5), heat nearly to boiling and filter. Wash the filter with hot dilute HCI
(1 : 5) and dilute the filtrate and washings to 250 ml. Add NH 40H in slight
excess, heat to boiling, boil for 1 to 2 minutes, and let the precipitate settle.
Filter through l1-cm. paper of loose texture and wash the precipitate with
hot water. Discard the filtrate. Dissolve the precipitate in 25 ml. of hot
dilute HCI (1 : 1) and reprecipitate with NH 40H. Filter, wash a few times
with hot water, and discard the filtrate. Dissolve the precipitate in 100 ml. of
hot dilute HNO s (1 : 3)1 catching the filtrate in a 300-ml. Erlenmeyer flask.
From this point proceed as in the Determination of Phosphorus by the Alkalimetric Method in the analysis of Ferromanganese.24
DETERMINATION OF COPPER

Transfer 5 g. of sample to a 400-ml. beaker and dissolve in a slight excess of

dilute HNO s (1 : 3), adding a few drops of HF if solution is difficult. Cautiously add 25 ml. of H 2S0 4 (sp.gr. 1.84), and evaporate to fumes of sulfuric
acid. Cool, treat with 150 ml. of water, filter into a 400-ml. beaker, and wash
the residue with water. Nearly neutralize with NaOH and pour into 100 ml.
of a warm solution of NaOH (10%). Filter, and discard the filtrate. Dissolve
the precipitate in warm, dilute H 2S0 4 (1 : 3) and repeat the precipitation and
filtration. Dissolve the precipitate in warm, dilute H 2S0 4 (1: 3). Cool
and treat with 3 g. of powdered tartaric acid. When this has dissolved, add
NH 4 0H till slightly alkaline, then dilute H 2S0 4 (1 : 1) till just acid, and finally
2 m!. in excess. Add 5 g. of Na2S0a, heat nearly to boiling and, slowly and
with stirring, add a solution of 1 g. of KCNS in as little water as possible. Add
a little paper pulp and digest with occasional stirring for 30 minutes at about
90 C. Filter and wash thoroughly with hot water containing a little tartaric
acid and KCNS. Ignite paper and residue in a porcelain crucible at about
700 0 C. Brush the residue into a 100-ml. electrolysis beaker, add 5 ml. of
HNO a to the crucible, warm for a few minutes, and pour into the beaker.
Heat until the oxide has dissolved and evaporate to 1 to 2 ml. Add 50 m!. of
water and determine the copper by electrolysis (see the Determination of
Copper in the Tentative Methods of Chemical Analysis of Steel, Cast Iron,
Open-Hearth Iron and Wrought Iron CA. S. T. M. Designation: E 30-36 T)
of the American Society for Testing Materials 25 if the amount is high, or by
the colorimetric method if the amount is small.
24
25

See p. 1488.
See p. 1446.

l.

EXAMINATION OF BITUMINOUS SUBSTANCES,

INCLUDING ASPHALTS, T,ARS ~ND PITCHES 1
PART I
EXAMINATION OF CRUDE, REFINED AND BLENDED
BITUMINOUS SUBSTANCES
The methods ordinarily used for examining bituminous substances and their
mixtures may be grouped under four headings, viz.: .
(1) Physical characteristics.
(3) Solubility tests.

(2) Thermal tests.

(4) Chemical methods of examin~tion.
TABLE I

Description

For
Purposes
of Identification

Physical Characteristics:
Fracture ......................
Streak ........................
Specific gravity ................
Viscosity ............ , .........
Penetration (Hardness) .........
Ductility .......................

YES
YES
YES
....
YES
....

Thermal Tests:
Fusing or softening-point .......
Volatile matter ................
Evaporation test ...............
Flash-point ....................
Fixed carbon ..................
Distillation test ................

Adaptibility for
a Given

Gauging
the Uniformity
of Supply.

Purpose

Yes
....
YES
YES
YES

. ...
. '"
Yes
Yes
YES
Yes

YES
Yes
Yes
....
YES
Yes

YES
YES
YES
YES

YES
Yes
Yes
Yes

YES

Yes

Solubility Tests:
Solubility in carbon disulphide ...
Carbenes ............. \ ........
Solubility in petroleum naphtha ..
Free carbon in tars .............

YES
Yes
YES
YES

YES
... ,
Yes
Yes

YES
Yes
Yes
Yes

Chemical Tests:
Water ........................
Oxygen in non-mineral matter ....
Solid paraffins ....... ' j '
Sulfonation residue ............
Saponifiable matter ............
Diazo reaction .................
AnthraqUinone reaction .........

Yes
YES
YES
YES
YES
YES
YES

....

....

0:"

-0- .

....
....
....

. ...
....
. ...
. ...

y~~
. ...

... Yes
Yes

YES
YES
...

YES
. ...

.
. ...

...
YES
. ...

....

* (a) Purity;
1

(b) Care exerciSed in its preparatipo.; ("1;) _Intrinsic value.

By Herbert Abraham.
.

1510
-"

....

. ...
....
. .. ,
'"

,
....
....
....
. ...
. ...
YES

'"

'"

YES
Yes
Yes
Yes
. .. ,
Yes

- .

. ...
'" .

. ...
Yes
Yes
YES
Yes

ABa
Criterion
of the
Quality*

. .. ,
....

Purposes
of
Factory
Control

. ...

. ...

EXAMINATION OF BITUMINOUS SUBSTANCES

uotlO'BaU

ouolI!rib~JqtUV

nQMoualI

"

,S!

000)

00)

ZZZ

ZZ

0 0

0 0

0 0 ) 0 0 0 ) oooco) oooco):Bo) :B!'l:B) ~lS~!o

zz
zzz zzzzz zzzzz ~z ~~~ ><><~~z

I
Izz Izz I zz:z; Izzzzz Izzzzz Izz I III III III I :J!S gs !
'zzz

O."la

j
~

0 0

.&

,,:B
15><

--a "'a

0 0 0

j
'.::l

.e1 1"
~-<
-<

z~~

0 0 0

0 0

0 0

'""

11

,J
1"1

0 0

t!l

i:g
<2

J~rJ

"";=
.o~

,.ij ~

!:.il
Z""
,!,

11
d

1511

0 0

~
0"
!:.:;::

Il<

~~~

f-<

&!~

"

~1Il
:00"

:0
,.!l""

i1:a

Ii

~~><~~

"

1512

EXAMINATION OF BITUMINOUS SUBSTANCES

OOl) '';[ om;

lU Al!soOSJA Ja(~ua

((tII

1~:iS I~:iS I
8

-jj -jj

:~:
I ; :~ I :I : ~: .: I :I: ..:
: : I:

:..

:: : . . :

EXAMINATION OF BITUMINOUS SUBSTANCES

1513

In general, these tests may have one or more of the following objects in view:
(a) To serve as a means of identification.
(b) To ascertain the value of the substance for a given purpose.
(c) To gauge its uniformity of supply.
(d) As an aid to factory control in its preparation, refining or blending.
(e) As a criterion of its quality.
The most important methods of testing are given in Table I.
Table II gives a list of the principal bituminous substances, together with
such of their physical and chemical characteristics as will enable them to be
distinguished, one from another.

PHYSICAL CHARACTERISTICS
Fracture
This is ascertained upon cleaving the specimen by subjecting it to a sharp
blow, and examining the cleavage surface. Only hard and "brittle" bituminous substances will yield to this test, including the hard asphalts and asphaltites. The fracture may either appear conchoidal (rounded and curved like a
shell), or hackly (jagged, irregularly and rough).
Streak
This represents the color of the powder which is left behind on drawing a
piece of the solid bituminous material across the surface of unglazed porcelain.
Hard bituminous materials only will yield to this test. 'The streak may be
classified as white (where no streak is visible), yellowish, yellowish brown,
brown, brownish black and black.
Specific Gravity
Hydrometer Method.-Where speed is essential and great accuracy not
required, the specific gravity of fluid bituminous materials may be determined
with a hydrometer having its scale sub-divided to unity in the third place of
decimals. Usually a series of hydrometers are used, ranging respectively from
0.800 to 0.900, 0.900 to 1.000, 1.000 to 1.080, 1.070 to 1.150, 1.150 to 1.230, and
in such dimensions as to enable them to be used in a 100 ml. cylinder approximately 300 mm. long having an inside diameter of not less than 32 mm. as
illustrated in Fig. 187. The hydrometer shall conform to the following requirements as to dimensions:
Dimension

Length of stem ....... , ....... . 125 mm.

Length of bulb ............... . 105 mm.
Length of scale .............. . 80mm.
6mm.
Diameter of stem ............. .
Diameter of bulb ............. . 22mm.

Permissible
Variation

6 mm.
5 mm.
4 mm.
O.5mm.
1 mm.

1514

EXAMINATION OF BITUMINOUS. SUBSTANCES

Most hydrometers are adapted to read at 60 F./60 F., or in other words,

the instruments are calibrated for water at 60 F. taken as unity. The standard
temperature for' testing bituminous materials is 77 F., and they, should accordingly be brought to this temperature when tested with the hydrometer.
For correcting the reading to water at 77 F., it shoJld be multiplied by 1.002,
I'
as follows:
Sp.gr. at 77 F./77 F. =~p.gr. ,at 77 F./60 F. X 1.002.
In running the test, the bituminous material is brought to 77 F., immediately poured into the hydrometer jar, and then the hydrometer allowed to

"A---:
.

100-

--.

6mm.

-1:-

1I ;!

101
102

I:

103

104

..
..
..
..

lOS
106 :
107

,I

.j

--~-

I:

(;)

-{---

(;)

1<\

3~"

I
:

_.i.
I
FIG.

187.-Hydrometer for Liquid Bitumir!ous Substances.

sink into it until it comes to a definite resting-point, whereupon it is raised

slightly, and allowed to. sink a second time. The reading is then noted. The
hydrometer must not be pushed below the point at which it comes to rest until
after the second reading, has been taken, whereupon it should be pushed a slight
distance below the end-point to observe whether or not it will fise. If it fails
to do so, it is evident that the bituminous material is too viscous to be tested
by the hydrometer method, and some other method should be employed.
Care should be taken that the hydrometer does not touch the sides or bottom.
of the cylinder when the reading is taken, also thlJ.t the surface of the liquid is
free from froth or bubbles.
The hydrometer has also been reco.mmended' for ascertaining the specific
gravity of hard asphalts, which are melted and tested while fluid at elevated
temperatures, the results being converted to 60/60 F. by suitable conversion
tables.
I
For converting specific gravity into. degr~es .aume and vice versa, the
following formulae may be used:

EXAMINATION OF BITUMINOUS SUBSTANCES

For liquids

lig~ter

1515

than water:
0B

e
140
aum =Sp.gr. 600/600 F. -130

Sp.gr. 60/60 F. =

130+10~aume

For liquids heavier than water:

Pyknometer Method.-This method has been standardized as follows:

(A) The specific gravity of road oils, road tars, asphalt cements and soft
tar pitches shall be expre'ssed as the ratio of the weight of a given volume of the
material at 25 C. (77 F.) to that of an equal volume of water at the same
temperature and shall be expressed thus:
Specific Gravity 25 C. (77 F.)/25 C. (77 F.)
The determination of specific gravity shall be made with a pyknometer or
weighing bottle (Fig. 188), which shall consist of a straight-walled glass tube
approxima,tely 70 mm. long and 22 mm. in diameter, carefully
ground to receive an accurately fitting solid glass stopper with a
hole of 1.5 to 1.7-mm. bore in place of the usual capillary
opening. The lower, part of the stopper is made concave in
order to allow all air bubbles to escape through the bore.
: i
The depth of the cup-shaped depression shall be about 4.8 mm.
!
at the center. The stoppered tube should have a capacity of
:
:I'
I
about 24 ml. and when empty should weigh not over 35 g.
~
tr
Before making a determination, the pyknometer with stopI
II
:'-._ ........I:
per shall first be calibrated by weighing it clean and dry upon
an analytical balance. This weight is called a. It shall then
FIG.188.be filled with freshly boiled distilled water at a temperature of Pyknometer
25 C. (77 F.), the stopper firmly inserted, all surplus moisture or Weighing
wiped from the surface with a clean dry cloth and again weighed. Bottle.
This weight is called b.
When determining the specific gravity of road oils or road tars which flow
readily, the material shall be brought to a temperature of, 25 C. (77 F.) and
poured into the pyknometer until it is full, with care to prevent the inclusion
of air bubbles. The stopper is then firmly inserted and all excess of material
forced through the opening is carefully removed with a clean dry cloth. The
pyknometer and contents are then weighed and this weight is called c. The
specific gravity of the material shall be calculated from the formula:

.
c-a
SpeCl'fi c GravltY=-b-a

1516

EXAMINATION OF BITUMINOUS SUBSTANCES

When determining the specific gravity of tar and asphalt products which are
too viscous for the method described in the preceding p,aragraph, a small amount
of the material shall be brought to a fluid condition by the gentle application of
heat, care being exercised to prevent loss by evaporation. When sufficiently
fluid, enough is poured into the clean dry pykometer to about half fill it. Precautions shall be taken to keep the material from touching the sides of the tube
above the final level and to prevent the inclusion of air bubbles. The tube
should be slightly warmed before filling. The pyknometer and contents are
then cooled to room temperature and weighed with the stopper. This weight
is called c. The pyknometer is next removed from the balance, filled with
freshly boiled distilled water, and the stopper firmly inserted. It is then completely immersed for not less than thirty minutes in a beaker of distilled water
maintained at 25 C. (77 F.) after which it is removed, and all surplus water is
wiped off with a clean cloth. It is immediately weighed. This weight is
called d. The specific gravity of the material shall be calculated from the
formula:
'c-a
Specific Gravity
(b-a)-(d-c)
When making the specific gravity determination it is important that:
(a) Only freshly boiled distilled water f?hall be used.
(b) When weighing the pyknometer completely filled, the temperature of
its contents shall be within 10 C. (1.8 F;) of 25 C. (77 F.).
(c) Precautions shall be taken to prevent expansion and overflow of the
contents from the heat of the hand when wiping the surface of the pyknometer.
(d) The presence of all air bubbles shall be eliminated in filling the pyknometer and inserting the stopper.
(e) Weighings sh~ll be made quickly after filling the pyknometer and shall
be accurate to 1 mg. A number of trial fillings and catch weights may be
necessary to obtain the desired degree of accuracy.
.
(f) To prevent breakage of the pyknometer when cleaning it out 'after a
determination has been made upon a very viscous 'or semi-solid material, it
will be found advisable to warm it in an oven a't not over 100 C. until most of
the material may be poured out and then to swab it with a piece of soft cloth
or cotton waste. When cool it may be finally rinsed with carbon dis.ulfide,
benzol or other solvent and wiped clean.
The limit of accuracy of the test is 0.005 specific gravity.
(B) The specific graVity of creosote oil fractions is determined in a similar
manner by means of a pyknometer, but in this case the results are expressed
at 38 C. (100 F.)/15.5 C. (60 F.).
(0) The pyknometer method may also be used for finding the specific
gravity of hard and brittle bituminous substances, including hard asphalts of
high fusing-point, asphaltites, asphaltic pyrobitumens, non-asphaltic pyrobitumens and pyrobituminous shales. Approximately 3.5 grams of the material
ground to 60-mesh are carefully weighed and introduced into a 50-m!. pyknometer, with about 30 m!. of distilled water. A vertical condensing bulb is attached to the pyknometer with a small section of rubber tubing, the open end
being connected with an aspirator to maintain.a partial vacuum. The pyknometer is then boiled on a water-bath to expel alltIle 'air from the sample. The

EXAMINATION OF BITUMINOUS SUBSTANCES

1517

inside of the" condensing tube is then washed back into the pyknometer, which,
is cooled to the desired temperature, stoppered, filled to the mark with water
a,t the same temperature and weighed. The specific gravity may then be
calculated from the foregoing formula.
Analytical Balance Method.-This method is adapted for substances which
can be melted and cast in a mold to form a briquette which may be handled on
cooling. It has be_en standardized as follows:
.
The determination of specific gravity shall be made with an analytical
balance equipped with a pan-straddle or other stationary s~pport (Fig. 189).

FIG. 189.-Analytical Balance for Ascertaining the Specific Gravity.

The test specimen shall be a cube of the material measuring approximately

:72 in. to the edge. It shall be prepared by melting a small sample of the
material by the gentle application of heat, care being exercised to prevent loss
by evaporation, and pouring when sufficiently fluid into a :72-in. brass cubical
mold, having the form illustrated in Fig. 194, which has been amalgamated
with mercury and which is placed on an amalgamated brass plate. Precautions
should be tak~n to prevent the inclusion of air bubbles. The hot mat.eriaI
should slightly more than fill the mold and when cool the excess may be cut off
with a hot spatula. The specimen shall be removed from the mold when cooled
to room temperature.
The balance shall first be tared with a piece of fine waxed silk thread sufficiently long to reach from the hook on one of the pan supports to the straddle or
rest. The test specimen shall then be attached to the thread, so as to be
suspended about 1 in. above the straddle from the hook on the pan support,
and weighed. This weight is called a and shall be accurate to 0.1 mg. The
specimen, still suspended by the thread, shall then be weighed completely
immersed in freshly boiled distilled water at 25 C. (77 F.)l C. (1.8 F.),

1518

EXAMINATION OF BITUMINOUE! SUBSTANCES

adhering air bubbles being first removed with a fine wire. This weight is
called b and shall be accurate to 0.1 mg. The specific gravity of the material
shall be calculated from the formula:
Specific Gravity = a

~b I

I
The limit of accuracy of the test is 0.005 specific gravity.

Viscosity
Engler Method.-This method as described on p. 1-718 is adapted for examining liquid or semi-liquid bituminous substances.
Saybolt Method.-The details of this method will be found on p. 1719.
Float Test.-This instrument is used largely for testing the viscosity or consistencyof semi-solid bituminous materials. The range of the float test is limited,
and it cannot be used with very fluid bituminous materials or with hard solids. It
accordingly fills the gap between the Engler viscosimeter on one hand, and the
needle penetrometer on the other. The test is not affected by the presence of
'
finely divided mineral matter or free carbon.
The procedure has been standardized as follows:
"~61,-

>

r----...;;A~

".

9;.of2.Dmm.......

-.~
~

_.L-~

~~
~
~

- ""1.J.4~(JJmm.j,
'.---'"

Weigh'......
f"b'

~~ ... ~&%M:I

Float.

(Aluminu") \. /,Sfandanl

uotO.!O

,a.

l]'};lli

mm..

...

..J
lapen:u

9.70 / , '

Thrwti

I";'"'

'{~1a

ftJmalt

~n.;C\e

Weigh/"', fll2JZI

::!:;_<S

UO~()'IO j "'/0'1
mm.

Mm.

Colla ..
(Brass)
W,;ght qf Fi.af, :rl.90~ 0.20,.
WeigH qf C4/hlj 9.I/J t ().2()g.
'SfandanJ

Supporl

,p....mti!.)'.

FIG. 190.-Float Test Apparatus.

ExAMINATION OF BITUMINOUS SUBSTANCES

1519

The Hoat (Fig. 190) shall be made of aluminium or aluminium alloy and
shall be in accordance with the following requirements:

Weight of float, g ..........................

Total height of float, mm ...................
Height of rim above lower side of shoulder, mm ..
Thickness of shoulder, mm ..................
Diameter of opening, mm...................

Minimum

Normal

37.70
34.0
26.5
1.3
11.0

37.90
35.0
27.0
1.4
11.1

The collar shall be made of brass and shall be

following requirements:

Weight of collar, g......... ' ................

Over-all height of collar, mm ................
Inside diameter at bottom, mm ..............
Inside diameter at top, mm .................

In

Maximum

38.10
36.0
27.5
1.5
11.2

accordance with the

Minimum

Normal

Maximum

9.60
22.3
12.72
9.65

9.80
22.5
12.82
9.70

10.00
22.7
12.92
9.75

The top of the collar shall screw up tightly against the lower side' of the
shoulder.
The assembled Hoat and collar:, with the collar filled flush with the bottom
and weighted to a total weight of 53.2 g., shall Hoat upon water with the rim
8.5 1.5 mm. above the surface of the water. Dimensions of the apparatus
additional to 'those required above are given in Fig. 190. The thJrmometer
shall be graduated in either Centigrade or Fahrenheit degrees as specified, the
ranges being _2 0 to +80 0 C. or +30 0 to + 1800 F., respectively. The diam-.
eter of the bath and the depth of water shall be' at least 185 mm.
The brass collar shall be placed with the smaller end on a brass plate which
. has been previously amalgamated with mercury by first rubbing it with a dilute
solution of mercuric chloride or nitrate, and then with mercury.
The sample shall be completely melted at the lowef,t possible temperature
that will bring it to a sufficiently fluid condition for pouring, excepting creosote
oil residues, which shall be mixed and poured at a temperature of 1000 to 1250 C.
It shall be stirred thoroughly until it is homogeneous and free from air bubbles.
The sample shall then be poured into the collar in any convenient manner until
slightly more than level with the top.
Asphalt and Asphalt Products.-Asphalt and asphalt products shall be
cooled to room temperature, placed in water maintained .at 50 C. for five
minutes, after which the surplus material shall be removed by means of a
sp~tula, or steel knife, which has been slightly heated. The collar and plate
shall then be placed in a tin cup containing ice water maintained at 5 C.,
1 C., and left in this bath for at least fifteen minutes.
Tar Producis.-Tar products shall be immediately immersed in ice water
maintained at 5c C. for five minutes, after which the surplus material shall be
removed by means of a spatula or steel knife, which has been slightly heated.

1520

EXAMINATION OF BITUMINOUS SUBSTANCES

The collar and plate shall then be placed in a tin cup containing ice water
maintained at 5 C., 1 C., and. left in this bath for at least fifteen mihutes.
The bath shall be filled with water and the water heated to the temperature
at which the test is to be made. This temperature shall be accurately maintained and shall at no time throughout the test be allowed to vary more than
0.5 C. from the temperature specified.
.
After the material to be tested has been kept in the ice water for not less
than fifteen minutes nor more than thirty minutes, the collar with its contents
shall be removed from the plate and screwed into the a:luminium fioat and
immersed in water at 5 C. for one minute. Any water shall then be removed
from the inside of the fioat and the latter immediately floated in the warm batli.
As the plug of material becomes warm and fluid, it is forced upward and out of
the collar until the water gains entrance into the saucer and causes it to sink.
The time in seconds between placing the apparatus onthe water and when
the water breaks through the material shall be determined by means of a' stop
watch, and shall be taken as a measure of the consistency of the material under
examination.
NOTE.-Speciai precautions should be taken to insure the collar fitting tightly into the
float and to see that there is no seepage of water between the collar and float during the
~~
.
Penetration (Hardness)
Penetration is defined as the consistency of a bituminous material, expressed
as the distance that a standard needle vertically penetrates a sample of the
material under known conditions of loading, time and temperature. Where
the conditions of test are not specifically mentioned, the load, time and temperature are understood to be 100 g., 5 seconds, 25 C. (77 F.), respectively,

and the units of penetration to indicate hundredths of a centimeter.

The container for holding the material to be tested shall be a fiat-bottom,
cylindrical metal or glass dish, 55 mm. (2 136 in.) in diameter and 35 mm. (1%
in.) deep. The needle (Fig. 191) for this test shall ~e made from a cylindrical
s'4q' to 9~""""'~..

(1.C: Zoe mm. .

'-.., I
T

ae

To'

. ()J4 ~OJ6m,'

b"
I1<6.35mm..J
.
'
- ....-............... :C~"::.(;~ .. ....--..-...... ~ ..~
UpplW.

FIG. 191.-Needle for Penetration Test.

steel rod approximately 50.8 mm. (2 in.) long, and having a diameter of 1.00
to 1.02 mm. This shall be symmetrically tapered at one end to a cO,ne approximately 6.35 mm. (7i in.) in height and whotle angle shall be within the
range if 8 40' and 9 40'. After tapering, the point shall be "blunted" by
grinding off to a truncated cone, the smaller base of which shall be from 0:14
to 0.16 mm. in diameter. The finished needle shall be hardened and highly
polished.
.
The water bath shall be maintained at a temperature not varying more than
0.1 C. from 25 C. (77 F.). The volume of _WJl.ter shall be not less than 10
liters and the sample shall be immersed to a depth of not less than 10 cm.

EXAMINATION OF BITUMINOUS SUBSTANCES

1521

(4 in.) and shall be supported on a perforated shelf not less than 5 cm. (2in.)
from the bottom of the bath. Any apparatus which will allow the needle to
penetrate without appreciable friction, and which is accurately calibrat~d to
yield results in accordance with the definition of penetration, will be acceptable.
The transfer dish for container shall be a small dish or tray of such capacity
as will insure complete immersion of the container during the test. It shall
be provided with some means which will insure a firm bearing and prevent
rocking of the container.
The sample shall be completely melted at the lowest possible temperature
and stirred thoroughly until it is homogeneous and free from air bubbles. It
shall then be poured into the sample container to a depth of not less than 15 mm.
(% in.). The sample shall be protected from dust and allowed to cool in an
atmosphere not lower than 18 C. (65 F.) for one hour. It shall then be
placed in the water bath along with the transfer dish and allowed to remain
one hour.
In making the test, the sample shall be placed in the transfer dish filled
with .water from the water bath of sufficient depth to completely cover the
container. The transfer dish containing the sample shall then be placed upon
the stand of the penetration machine. The needle loaded with specifieli weight
shall be adjusted to make contact with the surface of the sample. This may
be accomplished by making contact of the actual needle-point with its image
reflected by the surface of the sample from a properly placed source of light.
Either the reading of the dial shall then be noted or the needle broyght to zero.
The needle is then released for the specified period of time, after which the
penetration machine is adjusted to,measure the distance penetrated .
.At least three tests shall be made at points on the surface o(the sample not
less than 1 cm. (% in.) from the side of the container and not less than 1 cm.
(% in.) apart. After each test the sample and transfer dish shall be returned
to the water bath and the needle shall be carefully wiped toward its point with
a clean, dry cloth to remove all adhering bitumen. The reported penetration
shall be the average of at least three tests whose values shall not differ more than
four points between maximum and minimum.
.
When desirable to vary the temperature, time and weight, and, in order to
provide for a uniform method of reporting results when variations are made, the
samples shall be melted and cooled in air as above directed. They shall then
be immersed in water or brine, as the case may require, for one hour at the temperature desired. The following combination!! are suggested:
At 0 C. (32 F.) 200-g. weight, 60 seconds.
At 46.1 C. (115 F.) 50-g. weight, 5 seconqs.
Ductility
This test has been standardized as follows:
The ductility of a bituminous material is measured by the distance to which
it will elongate before breaking when two ends of a briquet of the material of
the form described, are pulled apart at a specified speed and at a specified
temperature. Unless otherwise specified, the test shall be made at a temperature of 25 C.O.5 C. (77 F.O.9 F.) and with a speed of 5 cm. per minute
(5.0%).

1522

EXAMINATION OF BITUMINOUS SUBSTANCES

The mold shall be similar in design to that shown in Fig. 192. Dimensions
shalL be as given with the permissible variations indicated. The mold 'shall
be made of brass, the ends, band b' , being known as clips, and the parts, a and

1<-- --.: -1/.9cmtO.lcm.

I

-- - - _",

f.

-.,_,
>.J.IS~
J.25,,,,,

.~--'-'''''I..,.......J-".1.._'--'''''L-..~;;L...J;;J Io.99 to 1.01 em

The opening in the end of each clip, as indicated by "x," shall be half an ~lIipse
having a transverse axis of 3.2 cm. 0.05 cm. and half of the longitudinal axis shall
be 1.45 to 1.55 cm.
FIG. 192.-Mold for Ductility Test Specimens.

a', as sides of the mold. The dimensions of. the mold shall be such that, when

properly assembled, it will form a briquet having the following dimensions:

Total length ...................................
Distance between clips ........... :. . . . . . . . . . . . ..
, .... Width at mouth of clip ...... " .................
Width at minimum cross-section (halfway between
clips) .....................................
Thickness throughout. . . . . . . . . . . . . . . . . . . . . . . . ..

7.45 to 7.55 cm.

2.97 to 3.03 cm.
1.98 to 2.02 cm.
0.99 to 1.01 cm.
0.99 to 1.01 cm.

The water bath shall be maintained at the specified test temperature ;varying
not more than 0.10 C. (0.18 0 F.) from this temperature. The volume of water
shall be not less than 16' liters and the sample shall be immersed to a depth of
not le:;;s than 10 cm. and shall be supported on a per~orated shelf not less than
5 cm. from the bottom of the bath.
I
For pulling the briquet of bituminous material apart, any apparatus may
be used that is so constructed that the briquet will be continuously immersed in
water as specified, while the two clips are pulled apart at a uniform speed, as
specified, without undue vibration.
The bituminous material to be tested shall be completely melted until
thoroughly fluid by heating it in an oil bath maintained at- the minimum
temperature needed to properly liquefy the sample. It _shall then be
strained through a No. 50 sieve and, after a thorough stirring, poured into the
mold. The mold shall be assembled on a brass plate and, so as to prevent the
material under test from sticking, the surface of the plate and interior surfaces
of the sides aa' , of the mold shall be thoroughly amalgamated.~ The plate
upon which the mold is placed shall be perfectly flat and level so that the bottom
surface of the mold will touch it throughout. In filling the mold, care shall be
taken not to disarrange the parts and thus distort the briquet. In filling, the
2 The amalgamation may best be effected by immersing the clean mold in a solution
of mercury bisuIphate containing free metallic mercury, .a'p'd 'so as to come in contact with
the latter. Instead of mercury, the metal mold, preferably of stainless steel, may be
moistened with glycerol.

EXAMINATION OF BITUMINOUS SUBSTANCES 1523

material shall be poured in a thin stream back and forth from end to end of
the mold until it is more than level fiull. It shall be left to cool to room temperature and then placed in the water bath maintained at the specified temperature of test for 30 minutes, after which the excess bitumen shall be cu't off
by means of a hot straight-edged putty knife or spatula so that the mold shall
be just level full.

NOTE.-When paving asphalt cements are being tested, the oil bath shall be maintained at a temperature of from 125 to 1500 C. (257 to 3020 F.).
The brass plate and mold, with briquet, shall then be placed in the water
bath and kept at the specified temperature for at least 1,Y2 hours, when the'
. briquet shall be removed from the plate, the side pieces detached, and the
briquet immediately tested. The rings.,at each end of the clips shall be attached to the pins or hooks in the ductility machine and the two clips pulled
apart at a uniform speed as specified until the briquet ruptures. A variation
of 5% from the speed specified will be allowed. The distance through which
the clips have been pulled to produce rupture shall then be measured in centimeters. While the test is being made, the water in the tank of the ductility
machine shal! cover the sample both above and below it by at least 2.5 cm. and
shall be kept continuously at the temperature specified within 0.5 C.
(O.9 F.).
A normal test is one in which the material between the two clips pulls out
to a point or thread until rupture occurs at the point where the thre'ad has
practically no cross-sectional area. The average of three normal tests shall be
taken and reported as the ductility of the sample.
If the bituminous material comes in contact with the surface of the water
or the bottom of the bath, the test shall not be considered normal.
NOTE,-When :the specific gravity of the bituminous material to be tested is below
0.98 or above 1.01, the specific gravity of the water bath in the ductility machine shall
be made the same as the material to be tested by the addition of either methyl alcohol
or sodium chloride.
If a normal test is not obtainable on three successive tests, the ductility
shall be reported as being unobtainable under th~ conditions of the test.

THERMAL TESTS
Fusing or Softening-Point
Ring-and-Ball Method.-This has been standardized as follows:
The softening of bituminous materials generally takes place at no definite
moment or temperature. As the temperature rises, they gradually and imperceptibly change from a brittle or exceedingly thick and slow-flowing material
to a softer and less viscous liquid. For this reason the determination of the

1524

EXAMINATION OF BITUMINOUS SUBSTANCES

softening-point must be made by a fixed, arbitrary and closely defined method

if the results obtained, are to be comparable.
I
, ,
The apparatus shall consist of the following:
ta) A brass ring 15.875 mm. (% in.) in inside diameth and 6.35 mm. Oi in.)
deep; thickness of wall, 2.38 mm. U2 in.); permissible variation on inside diameter and thickness of ring 0.25 mm. (0.01 in.). This ring shall be attached in a
convenient manner to a No. 13 B. & S. gauge brass wire (diameter 1.83 mm.
=0.072 in.). See Fig. 193.
No,13'8'S Gag"
Brass Wire. ...

-~

"

...,i..IIIIIIL'._.....:..t..J.W......

r f

--..~

, 6r$ Rin9 .nd Wire.

Assembly
of Apparatus.

Proper Po.ilion,
of B,an.

FIG. 193.-Apparatus for Ring-and-Ball 'Method,

(b) A steel ball 9.53 mm.

(Ys

in.) in' diameter weighing between 3A5 and

.
(c) A glass vessel, capable of being heated, not less than 8.5 cm. (3.34 in.)
in diameter and measuring 10.5 cm. (4.13 in.) in depth from-the bottom of the
flare. (A BOO-m!' beaker, low form, meets this requirement.) "-.
The sample shall be melted at the lowest possible temperature to avoid loss
of volatile constituents and stirred thoroughly, avoiding incorporating air
bubbles in the mass, and then poured into the ring so as to leave an excess on
cooling. The ring, while being filled, should rest on a brass plate which has
been amalgamated to prevent the bituminous material from adhering to it.
After cooling, the excess material shall be cut off cleanly with a slightly heated
knife.
FaT Substances Fusing at 80 0 C. (1760 F.) aT Below.- U se a thermometer which
shall be graduated in ~ither Centigrade or Fahrenheit degrees as may be specified,
the ranges being -2 to +80 0 C., or 30 to 180 0 JT.,..,respectjvely. Fill the glass
vessel to a depth of substantially 8.25 cm. (3.25 in:) with freshly boiled, dis

3.55 g.

EXAMINATION OF BITUMINOUS SUBSTANCES

1525

tilled water at 50 C. (41 0 F.). Suspend the ring containing the sample in the
water so that the lower surface of the filled ring is exactly 2.54 cm. (1 in.)
above the bottom of the glass vessel and its upper surface is 5.08 cm. (2 in.)
below the surface of water. Place the ball in the water but not on the specimen.
Suspend the thermometer so that the bottom of the bulb is level with the
.bottom of the ring and within 0.635 cm. (U in.), but not touching the ring.
Maintain'the temperature of the water at 5 C. (41 F.) for fifteen minutes. 3
With suitable forceps, place the ball in the center of the upper surface of the
bitumen in the ring, thus completing the 'assembly as in Fig. 193. Apply the
heat in such a manner that the temperature of the water is raised 5 C. (9 0 F.)
each minute. The temperature recorded by the thermometer at the instant
the bituminous material touches the bottom of the glass vessel shall be reported
as the softening point. No correction shall be made for emergent stem. The
rate of rise of temperature shall be uniform and shall not be averaged over the
period of the test. The maximum permissible variation for any minute period
after the first three shall be 0.5 C. (0.9 0 F.). All tests in which the rate of
rise in .temperature exceeds these limits shall be rejected.
FOT Substances Fusing above 80 0 C. (176 0 F.).-The thermometer shall
be graduated in either Centigrade or Fahrenheit degrees as specified, the ranges
being 30 to 1600 C., or 85 to 320 0 F., respectively. The same method as given
above shall be employed, except that U. S. P. glycerin shall be used instead of
water, and the starting point of the glycerin bath shall be 320 C. (89.60 F.).
The bath shall be brought to this temperature and thoroughly agitated, 'then
the apparatus and specimens shall be placed in the bath which shall be maintained under agitation at the starting temperature for fifteen minutes, after
. which the assembly shill be completed. by placing the ball on the center of the
specimen and the test carried on as above. In applying the heat, the ring
apparatus shall be placed off the center of the container and the burner placed
midway between the center and edge of the beaker away from the specimen.
Rigid adherence to the prescribed rate of heating is absolutely essential in
order to secure accuracy of results: A sheet of paper placed on the bottom of
the glass vessel and conveniently weighted will prevent the bituminous material
from sticking to the glass vessel, thereby saving considerable time and trouble
in cleaning. The limit of accuracy of the test is 0.5 C. (0.9 0 F.).
Cube Method.-This method is restricted to testing tar-pitches, and has
been standardized as follows:
The softening of pitch takes place at no definite moment or temperature.
As the temperature rises, pitch gradually and imperceptibly changes from a
brittle or exceedingly thick and slow-flowing material to a softer and less
viscous liquid. For this reason the determination of the softening-point must
be made by a fixed, arbitrary and closely defined method if the results obtained
are to be comparable.
The apparatus shall consist of the following:
(a) A mold suitable for forming a 12.7-mm. (Y2-in.) cube of pitch. (A
recommended,type is shown in Fig. 194).
3 NOTE BY AUTHOR.-It will be found more convenient, and the accuracy of test will
in no way be sacrificed, if the initial temperature of the water is maintained 25 F. below
the softening point of the substance under examination.

1526

EXAMINATION OF BITUMINOUS SUBSTANCES

(b) An L-shaped right-angled hook made of No. 12 B. and S. gauge copper

wire (diameter 2.05 mm. =0.0808 in.) the foot of whidh shall be 2.54 cm. ~1 in.)
long.
t
(c) A glass vessel, capable of being heated, not le,ss than 8.5 cm. (3.34 in.)
in diameter and measuring 10.5 cm. (4.13 in.) in depth from. the bottom the
flare. (A 600-ml. beaker, Griffin low form, meets this requirement.)
.
(d) A thermometer shall be graduated in either Centigrade or Fahrenheit
degrees as specified, the ranges being -2 to +800- C. or +30 to + 180 F.,
respectively.

ot

Slandani
Svpporf..
\.

SIt/e.

Front

I-!ethod of Pll<lng Cub.

on Wire Hook.

tflItR"""---

. W.l. Pilth

Mol~

(Spetill)

-'"

":III'
I

_l

- A... mbly
Seetion
t~ro.9h Sr...

""'If

FIG. 194.-Apparatus for Cube-in-Water Method.

The pitch shall be formed into a 12.7-mm. (Y2'-in.) cube, truly shaped and
with sharp edges, either by melting and pouring, or softe~ing and pressing,
into a mold. In all cases an excess of pitch shall be used and the surplus material shall be cut off cleanly with a slightly heated knife. The harder pitches
specified can ordinarily be molded at room temperature, the softer pitches in
water at about 4 C. (39.2 F.). If they ar~_nmlted,. they should first be
thoroughly stirred, avoiding incorporating air bUbbles in the mass, and then

EXAMINATION OF BITUMINOUS SUBSTANCES

1527

poured into the mold so as to leave an excess on cooling. The mold should
rest on a brass plate and the surface of the plate and the interior surfaces of the
mold should be amalgamated to prevent the pitch from adhering to them.
For Pitches Fusing between 110 and 170 F.-Assemble the apparatus as
shown in Fig. 194. Fill the glass vessel to a depth of substantially 9.5 cm.
(3.75 in.) with freshly boiled, distilled water at 15.5 C. (60 F.). Place the
cube of pitch on the wire as shown in Fig. 194 and suspend it in the water so
that its lower edge is exactly 2.54 cm. (1 in.) above the bottom of the glass
vessel and its upper edge is 5.08 cm. (2 in.) below the surface of the water.
Allow it to remain in the water for fifteen minutes before applying heat. Suspend the thermometer so that the bottom of the bulb is level with the bottom
edge of the cube of pitch and within 0.635 cm. Oi in.), but not tou_ching the
cube. . Apply the heat in such a manner that the temperature of the water is
raised 5 C. (9, F.) each minute. The temperature recorded by the thermometer at the instant the pitch touches the bottom of the glass vessel shall
be reported as the softening-point. No correction shall be made for emergent
stem .. The rate of rise of temperature shall be uniform and shall not be
averaged over the period of the test. The maximum permissible variation
for any minute period after the first three shall be 0.5 C. (0.9 F.). All
tests in which the rate of rise in temperature exceeds these limits shall be
rejected.
For Pitches Fusing below 110 F.-Use the same method as given above,
except that the water when placed in the glass vessel shall be at a temperature
of 4 C. (39.2 F:). The cube shall be allowed to remain fifteen minutes in
this water before applying the heat.
The use of freshly distilled water is essential, as otherwise air bubbles may
form on the cube and retard its sinking. Rigid
adherence to the prescribed rate of heating is
absolutely essential in order to secure accuracy
of results. A sheet of paper placed on the
bottom of the glass vessel and conveniently
weighted will prevent the pitch from sticking to
the glass vessel, thereby saving considerable
time and trouble in cleaning. The limit of
accuracy of the test is 0.5 C. (0.9 F.).
For Pitches Fusing above 170 F.-The
heating is performed in an air bath in the apparatus illustrated in Fig. 195. The cube should
~
be suspended in line with the observation win- 'F10. 195. -CUbe-lll. A:Ir Me th0 d
dows, and thethermometer bulb brought to the for High Fusing-point Substances.
same level. The temperature is raised 9 F.
per minute, and recorded by the thermometer when the cube drops 1 in.

Volatile Matter
This test is used for identifying various bituminous materials. Thus in the
case of asphalts, the volatilization test will often serve to identify soft native
asphalts, which contain larger percentages of volatile matter than soft residual
or blown petroleum asphalts. Cut-back products also carry a large percentage
of volatile constituents.

1528

EXAMINATION OF BITUMINOUS. SUBSTANCES

The test may also be used to determine the adaptability of a bituminou.s

substance for certain definite purposes, where it becorhes necessary to heat it to
high temperatures, as for example in the paving industry or in .manufacturing
bituminized roofings and floorings. It likewise servE>~ as a valuable adjunct for
gauging the uniformity of supply and for purposes <)r factory control. . It also
furnishes an indication of the weatherproof properties of the material. Other
thing being equal, bituminous substances showing the smallest percentage of
volatile matter will prove most weatherproof on exposure to the elements. It
should be noted, however, that the volatility test al0Ile must not be taken'as the
final criterion as to whether or not a bituminous ~ubstance is weatherproof,
since other factors should also be taken into consideration. The volatility'
test may be regarded as an accelerated test, showing the loss of volatile constituents exclusive of water which will take place upon exposure to the weather in
a relatively thin layer, for a long time.
The following method has been I}-dopted as standard,

FIG. 196.-Electrically-heated Volatility Oven.

This test covers the determination of the loss in 'weight jexclusive of water)
of oil and asphaltic compounds when heated as hereinafter prescribed. The
material under examination shall, therefore, firSt:--he tested for water and if
water is found to be present, it shall be removed by suitable methods of dehy-

EXAMINATION OF BITUMINOUS SUBSTANCES

1529

dration before the material is subjected to the loss on heating test; or another
sample shall be obtained which is free from water.
The oven shall be rectangular iIi form with double walls and heated by
electricity, as illustrated in Fig. 196. Its interior dimensions shall be as follows:
height, exclusive of space occupied by the heating element, not less than
29.26 cm. (11.5 in.), width and depth, each, at least 5.08 cm. (2 in.)' greater than
the diameter .of the revolving shelf. The oven shall be provided with a perforated metal circular shelf approximately 24.8 cm. (9.75 in.) in diameter. A
recommended form of aluminum shelf is shown in Fig. 197. This shelf shall

i'-;r.1<

Section A-B.

Top View,

FIG. 197.-Shelf for Volatility Oven.

be placed in the center of the oven, with respect to all dimensions of the interior
of the same, suspended by a vertical shaft and provided with mechanical means
for rotating it at the rate of 5 to 6 revolutions per minute. One side of the
oven shall be hinged and equipped to serve as a tight fitting door which shall
contain a window at least 4 in. square, with double glass, through which a
thermometer, located in front of and level with, the revolving shelf, may be
read without opening the door. One round vent hold shall either be placed in
the top and bottom of the oven, or one only near the top, and one only near the
bottom at each side of the oven. Each hole shall be not less than 1.27 cm.
(0.5 in.) nor more than 1.60 cm. (0.63 in.) in diameter and they shall both be
left open during the operation of the oven.
The thermometer shall be graduated in Centigrade degrees, the range being
155 to 170 0 C.
The container in which the sample is to be tested shall be of metal or glass,
cylindrical in shape, and shall have a flat bottom. Its inside dimensions shall
be substantially as follow,s: diameter, 55 mm. (2.17 in.); de'pth, 35 mm. (1.38 in.).
NOTE.-The American Can Company's 3-oz. Gill style flat-bottom, seamless ointment box, deep pattern, fulfills these requirements.

1530

EXAMINATION OF

BITUMINOU~

SUBSTANCES

The sample as received shall be thoroughly stirred and agitated, warming,

if necessary, to insure a complete mixture before the portion for analysis is
removed.
Fifty grams of the water-free material to be tes.t~d shall be weighed into a
tared container conforming to the foregoing requireinents. The oven shall be
brought to a temperature of 163 C. (325 F.), and the box containing the sample placed in one of the recesses of the revolving shelf. The. oven shall then be
closed and the shelf rotated 5 to 6 revolutions per minute during the entire test.
The temperature shall be maintained at 163 C. 1 C. (325 F. 1.8 F.)
for 5 hours after the sample has been introduced and the oven has again reached
that temperature. The sample shall be removed from the oven, cooled and
weighed, and the loss due to volatilization calculated. During the 5-hour
period the temperature shall not vary more than 1 C. (1.8 F.). All
tests showing a greater variation in temperature shall be rejected.
Under ordinary circumstances a number of samples having about the same
degree of volatility may be tested at the same time. Samples varying greatly
in volatility should be tested separately. Where extreme accuracy is required,
not more than one material should be tested, at one time and duplicate samples
of it should be placed-simultaneously in the oven. Such duplicates shall check
within the limits of accuracy given below. Results obtained, on samples
showing evidences of foaming during the test shall be rejected.
Up to 5% loss in weight, the resulti} obtained may be considered as correct
within 0.5. Above 5% loss in weight the numerical limit of error increases
0.01 for every 0.5% increase in loss by volatilization, as follows:
Volatilization
Loss, Per Cent

Numerical
Correction

True Volatilization
Loss, Per Cent

5.0
5.5
6.0
10.0
15.0
25.0
40.0

0.50
0.51
0.52
0.60
0.70
0.90
1.20

4.50- 5.50
4.91- 6.01
5.48~ 6.52
9.40-10.60
14.30-15.70
24.10-25.90
38.80-41.20

NOTE.-If additional periods of heating are deSired, it .is recommended that they be
made in successive increments of five hours each. When the penetration or other characteristics of the sample after heating are required, melt the residue in the container at
the lowest possible temperature and thoroughly mix by stirring, taking care to avoid
incorporating air bubbles in the mass. Then bring it to the standard temperature and
test as prescribed.
- - -

Evaporation Test
This test is used principally for testing road oils, in determining the socalled asphalt content, and is carried out by evaporating the specimen under
carefully controlled conditions until the residue sl:).ows a penetration of exactly
100 at 77 F. (100 g., 5 sec.). The percentage by' weight of residue is recorded
and furnishes an indication of the quantity of constituents present which may
be depended upon to contribute to 'the durability of the pavement. It will
serve to differentiate between straight-distilled and cut-back products. This
test has been standardized as follows:
'. _
I
This method of test covers the determinatiQJ1..():f"percentage of residue having
a specified penetration at 100 g., five seconds, 25 C. (77 F.), obtained by

EXAMINATION OF BITUMINOUS SUBSTANCES, 1531

heating a road oil or a semisolid asphalt having a penetration of more than
100, at a temperature of 249 to 260 0 C. (480 to 5000 F.), When the penetration
of the residue is not otherwise. stated it shall be understood to be 100. The
residue obtained is available for testing as desired.
This method of test covers the determination of percentage of residue having
a specified penetration at 100 g., 5 seconds, 25 0 C. (77 0 F.), obtained by heating
a road oil or a semi-solid asphalt having a penetration of more than 100, at a
temperature of 249 to .2600 C. (480 to 5000 F.). When, the penetration of the
residue is not otherwise stated it shall be understood to be 100. The residue
obtained is available for tests as desired.
The apparatus shaH consist of a container, heating bath, hot plate, and
thermometer,. with necessary accessory apparatus. The container in which
the sample is to be tested shan be a flat-bottom, cylindrical seamless tin box,
70 mm. (2% in.) in diameter and 45 mm. (1% in.) in depth.
NOTE.-The American Can Co.'~ 6-oz. Gill style flat-bottom, seamless ointment box,
deep pattern, fulfills these requirements.
(Ii) The heating bath shall be a cast-iron air-bath permitting the immersion
of the container to a depth of 1 ~ in. through an opening l6 in. larger in diameter

.
'_

';;'.-..

.,

.....# ..

~ . ~~ .

//-/~'--'----"

FIG.

198.~Cast-Iron

Air-Bath.

than the ~ontainer. It shall support the container ~ in. above the hot plate
and with at least 3i in. free air space between the sides of the container and of
the air-bath below the opening. A suitable air-bath is shown in Fig. 198.

1532

EXAMINATION OF BITUMINOUS SUBSTANCES

(b) The air-b&th shall be heated upon a suitably mounted hot plate, heated
either electrically qr by. means of a gas flame. The plate shall pe capable of
maintaining the sample continuously at the required temperature, and apparatus necessary to fulfill this requirement, such aE1 ~ rheostat. OJ' gas pressure
regulator, shall bE~ provided.
Tbe tbermometer snall con'iorm to tbe 'iollowing requirementB. 'T'nBBB
specifications cover a special thermometer graduated to either Centigrade or
Fahrenheit degrees as specified, the ranges being -6 to' +400 0 C., or +20 to
+760 0 F., respectively. The sample as received shall be thoroughly stirred
and agitated, to insure a complete mixture before the portion for testing is
removed.
I~
One hundred grams (100.000.10 g.) of the material to be tested '8hall be
weighed into a tared container, which shall then be placed in tbe air-bath in
position to be heated. The thermometer shall be supported i:n the sample
equidistant from the sides of the container and with the bottoIil of the .bulb
neither more than U in. above nor touching the bottom of the cOrltainer. The
bulb shall be completely immersed in the sample throughout the heating. An
assembly of the apparatus is shown in Fig. 199.

FIG. 199.-,Assembly of Apparatus for E':aporation Test.

The sample sbould be heated as rapidly a~ possible, to prevent foaming, to

a temperature of 249 0 C. (480 0 F.) and during the/evaporation, the temperature
0
shall be maintained between 249 0 C. (480 F.) :and 260 0 C. (500 F.). The
sample shall be stirred with the thermometer from time to time to prevent local
overheating and, to maintain a homogeneous samvle, all cakes of hardened
bitumen which form at the sides of the container shall be fluxed in the sample.
An experienced operator can judge approximately Iwhat percentage or
residue he should obtain to secure the desired-'penetration. When it is supposed that the residue will show the required penetratio)l, the bitumen on the

EXAMINATION OF BITUMINOUS SUBSTANCES

1533

thermometer which may be readily scraped off shall be returned to the container, which then shall be removed from the air-bath, cooled and weighed.
The penetration of the residue shall then be determined with the exception that
the 6-oz. container in which the evaporation has been conducted, shall be used
instead of the 3-oz. container specified in the test for" penetration."
It frequently is necessary to make several trials before a residue of the
required penetration is obtained. If it is determined to be greater than that
required, all water shall be removed from the container and the surface of the
sample, and the heating and determination of penetration may be repeated as
before. o.rdinarily a residue shall be considered as satisfactorily obtained when
its penetration is within 15 of that desired, and its percentage by weight of the
original sample shall be calculated. When it is necessary to determine more
'. precisely the percentage of residue having the specified penetration, such a
percentage shall be computed by interpolation between percentages of two
residues, one having a penetration greater and one having a penetration lower
than that specified. The percentage shall be reported as:
P!!rcentage of residue of ........ penetration (determined ........ ) stating, first,
the specified penetration, and second, the penetration actually determined for the sample
tested or calculated by interpolation.

Certain types of road oil will readily form rings of hard asphalt at the side
of the container. Great care should be taken 1{hat this material be completely
fluxed in the sample before the penetration of the residue is determined. Duplicate determinations should not differ by more than 1.0% with the same operator
nor more than 2.5% between different laboratories.
Flash-Point
A number of flash-point testers have been proposed, including the PenskyMartens, the Cleveland and the Tag types, which are fully described on p.1744 ff.
Fixed Carbon
The same procedure is followed as for testing coal (" Volatile Combustible
Matter"), as described on p. 1638 and p. 1639.
Distillation Test
The value of this test is to ascertain the adaptability of bituminous materials for a given use, generally for road treatment; also for gauging the llniformity of supply, for purposes of factory control, and most important of all, as a
criterion of the quality. This test is generally applied to tar products as an
equivalent of the volatility test becomes of value in identifying the kind used
(upon determining the specific gravity of the fractions distilled), as a means of
distinguishing a cut-back tar from a straight-distilled tar (upon determining
the specific gravity of the fractions, their viscosity, alsQ the fusing-point of the
residue), and for detecting the presence of abnormal amounts of naphthalene.
This test has been standardized as follows:
(a) For Road Oils (Asphaltic and Coal-tar Pitch), etc.-The apparatus
consists of a flask, condenser tube, shield, receivers and thermometers as
specified.

1534 'EXAMINATION OF BITUMINOUS SUBSTANCES

Flask: The distillation flask, Fig. 200, shall be It side neck distilling flask,
having the following dimensions:
I
Diameter of bulb, outside .................. . 8~ mm. 1.5 mm.
Diameter of neck, inside ........ , .......... . 2,2 mm. 1.0 mm.
Diameter of tubulature, inside .............. . 10.0 mm. 0,5 mm.
Height of flask, outside ....... , ............ . 13~ mm.1.5 mm.
Vertical distance bottom of bulb, outside, to
horizontal tangent at tubulature inside, .. 93 mm. 1.5 mm.
Length of tubulature ........ " ., .......... : 220 mm. 5.0 mm.
Angle of tubulature ....................... . 73 deg. 2 deg.
1.0 to 1.5 mm.
Thickness of tubulature wall ............... .
~ Inside Diameter
~->-I 22 mm.I.O mm.

T-~-

FIG. 2oo,-Distillation Flask.

Condenser Tube: The condenser tube shall be a suitable form of tapered glass
tubing of the following dimensions:
Outside diameter of small end .............. . 12.5 mm.1.5 mm.
Outside diameter of large end .............. . /28.5 mm. 3.0 mm.
Length ............................... ~ .. . 360.0 mm. 4.0 mm.
Length of tapered part ................... , .. 100.0 mm. 5.0 mm.

Shield: A g~lvanized iron shield, lined with Ys-in. asbestos, of the form and
dimensions shown in Fig. 201 shall be used to protect the flask from air currents
and to prevent radiation. The cover (top) may be of transij.f!. board masie in
two parts, or it may be of galvanized iron lined with Ys-in. asbestos.
Receiver: The distillates shall be collected in tared Erlenmeyer flasks having
a capacity of 50 to 100 ml.
Thermometer: The thermometer shall be graduated in either Centigrade or
Fahrenheit degrees as specified, the range being from 0 to 400 0 C. or 30 to
760 0 F., respectively.
(
The sample as received, shall be thoroughly stirred and agitated, warming
if necessary, to insure a complete mixture before the portion for analysis is
removed. The material may be tested for distillation without dehydration, if water is
present not to exceed 2.0%. If water is prEl_se'iit in excess of 2.0%, the
bituminous material shall. be dehydrated before testing, by distilling 500 ml.

EXAMINATION OF BITUMINOUS SUBSTANCES

1535

in an 800-rol. copper still provided with a water-cooled condenser, the distillate being caught in a separatory funnel. When all the water has been ex-

I
I

r---136 mm.3 mm.~

Cover in Two Parts
:

.,

, _ r-:o-

\! Il\!ij

45 mm.

~ 2mm.

, '..:

"

'I
I,
"I,

I,
'I
I,

I'
Ill"
Flanged open-end cylinder
made of 22 gage galvanized
iron with {'~ in. asbestos lining
riveted to metal.

'

D!~l
!,
I

1
,

:t,

't,I

",

"
,"
"'I

I,

I,
I,

Two mica wlndows are

provided at right angles
to the end slot.

FIG. 201.-Shield.

pelled, the distillate is allowed to settle, the water drawn off and the oils
returned to the residue in the still after the contents have cooled below 212 0 F.
Thermctfftftf

Shi'eld

i?eIorl..

/~l.~.i

''l:_~r '~;~2~;;:'"
20M"h
~Chimn~'y

FIG. 202.-Distillation Apparatus Assembly for Road Oils.

The flask shall be supported on a tripod or rings over two sheets of 20-roesh
gauze, 150 mm. square, as shown in Fig. 202. It shaH be connected to the

1536

EXAMINATION OF BITUMINOUS SUBSTANCES

condenser tube by a tight cork joint. Th~ thermometer shall be inserted

through a cork in the neck with the top of the bulb l~lVel with the lowest point
of juncture of the tubulature and neck of the flask.
.
The axis of the flask through the neck shall be v~rtical.
The distance from the bulb of the thermometen to the outlet end of the
condenser tube shall be not more than 600 nor less than 500 mm. The burner
should be protected from drafts by a suitable shield ~~ chimney.
One hundred grams (100 g.O.l g.) of the flample shall be weighed into the
flask, the apparatus assembled and heat applied so that the first drop comes
over in from five to fifteen minutes. The distillation shall be conducted at the
rate of between 50 and 70 drops per minute and the distillate collected in
weighed receivers.
NOTE.-In testing creosote oils: 80 to 120 drops per minute are specified.

The condenser tube shall be warmed whenever necessary to prevent

accumulation of solid distillates. The fractions shall be collected at the
points designated by the specifications. The receivers shall be changed
when the thermometer indicates the maximum temperature for each fraction.
When the maximum specified temperature of the test is indicated by the thermometer, the -flame shall be removed and any oil which has condensed in the
condenser tube shall be drained into the last fraction.
The residue shall remain in the flask with the cork and thermometer in
position, until no vapors are visible and it shall then be weighed. If tests of
the residue are required, the flask shall then be inclined so that the residue will
flow around the sides, thus collecting any condensed vapqrs that ,may be on the
sides of the flask, after which the residue shall be poured into a suitable receptacle and covered. If the residue becomes so cool that it cannot be poured readily
from the flask, it shall be reheated to a temperature not exceeding 125 C. by
holding the bulb of the flask in a suitable bath and not by the application of
flame. For weighing the receivers and fractions, a balance accurate to atleast 0.05 g. shall be used.
/
During the progress of the distillation the thermometer shall remain in its
original position. No correction shall be mac;le for the emergent stem of the
thermometer. The results of the distillation test shall be reported in percentages by weight of water-free material.
,
(b) For Cut-Back Asphaltic Products.-In testing cut-back asphaltic products, the foregoing method is modified in the following particulars:.
The condenser shall consist of a 250-mm. standard glass jacketed condenser
(Fig. 203).
The following dimensions are recommended:
Length of jacket, excluding the necks. . . . . .. ,250
Outside diameter of adapter of condenser tube. I 23
Length of adapter. . . . . . . . . . . . . . . . . . . . . . .. 75
Outside diameter of condenser tube proper. . . . 12.5
Over-all length of condenser tube, including
adapter ............................. 475

mm.
mm.
mm.
mm.

5
1
5
0.5

mm.
mm.
mm.
mm.

mm 25 mm.
i
The adapter sh~ll be ofthe curved design of.Jieavy wall (1 mm.) and reinforced top glass, with an angle of approximatelY 105, and with a diameter at

EXAMINATION OF BITUMINOUS SUBSTANCES

1537

the large end of approximately 18 mm. The outlet end shall be ground to an
angle of 45 5 with the inside vertical.
The receivers shall be graduated cylinders, of uniform diameter, with a
pressed or molded base and a lipped top. The over-all height shall be not less
than 24,8 cm. (9% in.) nor more than 26.0 cm. (lOX in.). The cylinder" shall

FIG. 203.-Distillation Apparatus Assembly for Cut-Back Asphaltic Products

be graduated in single milliliters to contain 100 ml., and the graduated portion
shall be not less than 17.78 cm. (7 in.) nor 'more than 20.32 cm. (8 in.) in length.
Each fifth graduation,shall be distinguished by a longer line, and the graduations
shall be numbered from tlie bottom up at intervals of 10 ml. The graduations
shall not be in error by more than 1 ml. at any point on the scale.
The flask shall be supported on a tripod or ring over two sheets of 20-mesh
gauze, 150-mm. square as shown in Fig. 203. It shall be connected to the
condenser tube by a tight cork joint. The thermometer shall be inserted
through a cork in the neck with the bottom of the bulb U in. from the bottom of
the flask. The axis of the flask through the neck shall be vertical. The distance from the neck of the flask to the outlet end of the condenser tube shall be
not more than 600 nor less than 500 mm. The burner should be protected
from drafts by a suitable shield or chimney.
The adapter shall be adjusted over the end of the condenser tube so as to
conduct the distillate into the receiver, and the top of the receiver shall be
covered closely during' the distillation with a piece of blotting paper or its
equivalent, which shall be cut so as to fit the adapter tightly. The adapter,
shall extend into. the receiver at least 2.54 cm. (1 in.) but not belowthe 100-ml.
mark. Unless the laboratory air temperature is between 12.8 and 18.3 C.
(55 and 65 F.) the receiver shall be immersed up to the 100-ml. mark in a
transparent bath maintained between these temperatures. The condenser tube
shall be clean .and dry .
Two hundred milliliters (calculated from the specific gravity of the material
at 15.5 C. (60 'F.)) of the sample shall be weighed into the flask, the apparatus
assembled and heat applied so that the first drop comes over in from 5 to 15 min.

1538

EXAMINATION OF BIT-qMINOUS

~UBSTANCES

The distillation shall be conducted at the rat.El of between 50 and 70 drops per
minute. Should the sample foam the distillation rate Will have to be reduced,
but the normal rate shall be resumed as soon as possible. If excess foaming
persists the distillation may be more easily controlleCliby applyi~g the flame
near the edge of the bulb instead of at the center of. samb. The distillate shall
be collected in the specified receivers, and the volume ot distillate at all specified
temperatures recorded. The volume of any separat'ed water shall also be
recorded. When the maximum specified temperatur{l of the test is indicated
by the thermometer, the flame shall be removed and the residue poured imme'diately into a 6-oz. tin box placed on its cover to prevent two rapid cooling at
the bottom. Any Qil which may remain in the condenser tube shall be drained
into the last receiver.
As soon as no further vaporization is apparent, the residue shall be stirred
to insure homogeneity, and then poured into the neCessary apparatus for the
required tests. During the progress of the distillation the thermometer shall
remain in its original position. No correction shall be made for the emergent
stem of the thermometer. Temperatures to be observed in the distillation test
shall be corrected for the effect of the altitude of the laboratory in which ~he
test is made.
Tll.p-_r.p.Rl..Uj$ n.f j.be. 11i.'<f..i11.ajjOJJ lm,j..s.b.sll be_rPPfl_rjeD in percentage by vo)u.me
of water-free material.
.
(c) For Coal-tar Products.-In testinfl coal~tar products, the foregoing
method shall be modified as follows: The thermometer :Shall be inserted through
a cork in the neck of the flask, with the bottom of its bulb 12 to 13 mm. from
the surface of the oil. The exact location of the th~rmometer bulb shall be
determined by placing a vertical rule graduated in divistons not exceeding 1 mm.
back of the flask, when the latter is in position for Ule test, and sighting the
level of the liquid and the point for the bottom of the thermometer bulb.
The following fractions are usually reported:.
I
For Bituminous
Road Oils:

Up to 170 C.
170 to 235 0 C.
235 to 270 0 C.
270 to 3000 C.
Residue.
0

For Cut-back
Asphalts:
0

Up to 190 C.
Up to 225 0 C.
Up to 3150 C.
Up to 3600 C.
Residue.

For Coal-tar
Products:

Up to 210 0 C.
210 to 235 0 C.
235 to 270 0 C.
270 to 315 0 C.
315 to 355 0 C .
. ~Residue
...._ .

EXAMINATION OF BITUMINOUS SUBSTANCES

1539

SOLUBILITY TESTS
Solubility in Carbon Disulfide
This test is useful for purposes of identification, for ascertaining the adaptability of a bituminous substance for a given purpose, for gauging its uniformity
of supply, and as a criterion of its quality (i.e., purity) and consequently its
intrinsic value. Crude bituminous materials are often purchased on the basis
of the percentage soluble in carbon disulfide. The presence of non-mineral
matter insoluble in carbon disulfide may be regarded as just that much inert
material, and in certain cases as an indication that the material has been carelessly prepared or overheated in its process of manufacture. The mineral
matter may be regarded as an adulterant. In the case of native asphalts, the
larger the percentage soluble in carbon disulfide, the greater will be their
intrinsic value. 'The percentage and composition of the mineral matter will
often indicate the source of the native asphalts. Asphalts derived from' petroleum are substantially free from mineral constituents, and with the possible
exception of the harder grades, contain little to no non-mineral matter insoluble
in carbon disulfide ..
With a native asphalt containing over 10% of mineral matter, it is necessary
to separate the portion soluble jn carbon disulfide, before ascertaining its
physical characteristics, fusing-point, and sometimes fixed carbon, in which case
the soluble constituents should be recovered as will be described.
Two methods will be consid!'lred, depending upon whether or not the
constituents are to be subjected to a detailed analysis.
. Where the Constituents are Not to be Examined Further.-In this case
the method recommended follows along the lines o the one standardized by the
A. S. T. M. which is substantially as follows, differing however somewhat in
phraseology: The sample shall be representative and if it contains more than
2% of water it shall be dehydrated by distillation in a copper still, the water-free
distillate being returned to the residue. If the material is hard and brittle, it
may be ground 4 and dried at a temperature below the temperature of volatilization of the material, in a shallow nickel or iron dish.
A Gooch crucible, approximately 4.4 cm. in width at the top, tapering to
3.6 cm. at the bottGm, with a depth of 2.5 cm. shall be set in the filter tube
inserted in the stopper of the filtering flask. The flask shall be connected with
the suction pump. Before suction shall be applied, the crucible shall be filled
'with asbestos 5 suspended in water which shall be allowed to settle partly in the
crucible. A light suction shall be applied to draw off the water, leaving a firm
mat of as.bestos in the crucible. More suspended asbestos shall be added and
the process repeated until a felt is built up that barely transmits light. The
felt shall then be thoroughly washed with water, dried in a drying oven, and
ignited over a Bunsen burner: The crucible shall then be cooled in a desiccator
and weighed. An amount of materiaL which shall contain approximately 1
gram of bitumen shall be weighed into a tared Erlenmeyer flask.
4 Where it is not desirable to crush the rock or sand grains, a lump should be placed in
the drying oven until it is heated through and softened at the lowest possible temperature,
whereupon it may be crushed into a thin layer and dried as described.
6 Asbestos (amphibole) cut in pieces not exceeding 1 cm. in length shredded and
shaken up with water.

1540

EXAMINATION OF BITUMINOUS [SUBSTANCES

(a) Method Used Where but Little Finely-Divided Insoluble Matter is

Present.-One hundred ml. of chemically pure carbon disulfide shall be added
to the flask in small portions with continued agitation until all lumps disappe'1r
and nothing adheres to the bottom. The flask shalll be corked and set aside
for fifteen minutes.
I
The Gooch crucible shall be set up again with the sucti9n flask and the
carbon disulfide solution carefully decanted through the asbestos felt, with
or without light suction as may be found necessary. No sediment shall be
allowed to go onto the filter. A small amount of carbon disulfide shall be used
to wash down the sides of the flask and then the precipitate shall be brought
onto the felt and the flask scrubbed with a feather if necessary to remove all
precipitate. The contents of the crucible shall be washed with carbon disulfide until the washings are colorless. Suction shall be applied to the cru~ible
to remove the carbon disulfide. The crucible shall be dried in the oven at 100
to 125 C. for twenty minutes, cooled in the desiccator and weighed.
0

The weight of the residue in the crucible shall be recorded. . . . . . . . . .. (1)

Ignite the residue in the crucible at a red, 'heat to a clean ash, 10001, add a
few drops of ammonium carbonate solution, re-ignite to a low req heat, cool .
and weigh the residue ..................
(2)
In case any insoluble matter adheres to.the flask, it shall be dried, weighed
and the increase 'over the original weight recorded ............. : . . . . .. (3)
The filtrate containing the soluble matter shall be evaporated, the bituminous residue burned, and the weight of ash recorded ................. (4)
Constituents Soluble in Carbon Disulfide equal the weight of material (dry)
taken for analysis, minus the sum of items (1), (3) and (4). Calculate this
in per cent.
Non-Mineral Constituents Insoluble in Carbon Disulfide equal item (1),
minus (2), plus item (3). Calculate this in per cent.
Mineral Constituents equal the sum of items (2) and (4). Calculate this
in per cent.
/
I

(b) Method Used Where a Substantial Quantity of Finely-Divided Insoluble Matter is Present.-This procedure should be followed where the filter
clogs unduly, or where the mineral matter passing through the filter exceeds
0.5%. The material should be weighed out in the same way as in the foregoing,
and treated with 100 mI.,of chemically pure carbon disulfide.
The flask shall be loosely corked and shaken at intervals. until all large
particles of material have been broken down. The flask shall be left undisturbed for forty-eight hours to permit the contents to settle. Th~ solution
shall be decanted into a similar tared flask, care'being taken to disturb as littlE)
of the residue as possible. The first flask shall be treated again with fresh
carbon disulfide as before and left undisturbed f,or forty-eight hours. The
solution from the second flask shall then be carefully decanted upon the Gooch
crucible without use of vacuum, and this shall be followed by ~h~ solution from
the first flask. The filter shall be washed with fresh carbon disulfide. The
residue remaining in each flask shail be shaken again with frlsh carbon disulfide
and allowed to settle for twenty-four hours. .T.he--soluti6n from both flasks
shall then be decanted through the filter and the' residues remaining in the

EXAMINATION OF BITUMINOUS SUBSTANCES

1541

flasks shall be washed again with carbon disulfide, the solution decanted and
the process repeated until the washings are practically colorless. The temperature shall be maintained between 20 and 25 0 C.
The crucible and both flasks shall be dried at 100 to 125 0 C. for twenty
minutes, cooled in a desiccator and weighed. The procedure from this'point
on is exactly the same as in the preceding. The per cent constituents soluble in
carbon disulfide, non-mineral constituents insoluble in carbon disulfide, and
mineral constituents shall be calculated.
The author finds that in the preseljlce of large quantities of finely divided
insoluble matter, the method may be materially shortened by adding a weighed
quantity (about twice the weight of bituminous material) of freshly ignited,
long-fibered amphibole to the bituminous substance in the first flask. On
shaking with carbon disulfide, the asbestos serves to dilute the insoluble matter,
preventing the latter from clogging the pores of the filter, and accordingly
reducing the time of filtration.
Where the Constituents are to be Examined Further.-Extract 25 to 50 g.
of the moisture-free bituminous substance with carbon disulfide as previously
described, increasing the quantities of carbon disulfide proportionately. In
certain cases an extractor described on p. 1568 may be conveniently employed.
Residue Insoluble in Carbon Disulfide: This contains the non-mineral constituents insoluble in carbon disulfide (e.g., insoluble bituminous matter, free
carbon and vegetable matter derived from associated soil) together with the
mineral constituents retained by the filter (e.g., clay with 2 molecules of water in
chemical combination, 6 silica, silicates, calcium and magnesium carbonates,
calcium sulfate with half a molecule of water in chemical combination, iron
pyrites, etc.). Dry in an oven at 220 0 F. to constant weight and mix well to
obtain a uniform sample. If the mineral matter is coarse, pulverize sufficient
of the well-mixed material in a mortar to 200 mesh or finer for use in carrying
out the examination described in sections (a) to (e) inclusive.
(a) To determine the 'water in chemical combination with the clay and
calcium sulfate, heat 10 g. to bright redness in a stream of dry hydrogen or
illuminating gas, in an electric combustion apparatus, catching the water in
a tared calcium chloride bulb, which is reweighed at the end of the operation.
The gain in weight represents the water which should be calculated in percentage.
(b) To determine the sulfur present in sulfides, weigh 25 g. into a small flask
closed with a stopper carrying a thistle tube filled with hydrochloric acid and
an outlet tube leading into a beaker of bromine water. Introduce the acid and
boil until all the hydrogen sulfide is expelled into the bromine water, which
serves to oxidize it into sulfuric acid. Expel the excess of bromine by boiling
and precipitate with barium chloride in the usual manner. Calculate the
percentage of sulfur present.
(e) To determine the carbon dioxide present in carbonates, weigh 1 g. into a
Schroetter or ;Mohr alkalimeter and treat with hydrochloric acid in the usual
manner. The loss in weight represents the sum of the carbon dioxide (liberated
from the carbonates) and the hydrogen sulfide (liberated from the iron pyrites).
6 The water of hydration in clay is held tenaciously, for none is given up during one
hour's ~eating at 293 C., although it is expelled on heating to redness.
0

1542

EXAMINATION OF BITUMINOUS SUBSTANCES

Subtract the weight of hydrogen sulfide (ascertained in b) to arrive at the weight

of carbon dioxide present, and calculate its percentag~.
(d) To determine the non-mineral constituents insoluble in carbon disulfide, '
weigh 5 g. into a porcelain dish, treat with an excess of concentrated hydrochloric
acid, evaporate to dryness on a water bath and thenl dehydrate the silica by
heating in an air-oven at 220 F. for two hours. Boill up with water, filter on
a weighed Gooch crucible and wash thoroughly with boiling water until the
filtrate no longer shows a reaction for calcium salts. Dry at 220 F. until no
further loss occurs and then weigh th,e rettidue, which represents the nonmineral constituents insoluble in carbon disulfide, together with clay and its
water of hydration, silica, silicates, calcium sulfate and its water of hydration,
etc. Ignite to redness until all the carbonaceous matter is consumed and
reweigh. The loss in weight represents the sum of the non-mineral constituents
insoluble in carbon disulfide, also the total water of hydration (origipally
associated with the clay and calcium sulfate). Subtract the per cent water of
hydration (ascertained in a) to arrive at the per cent non-mineral insoluble in
carbon disulfide.
(e) To determine the iron, aluminum, calcium, magnesium, silica, sulfates,
etc., weigh 2.5 g. into a platinum crucible and ignite until all the carbonaceous '
matter is consumed. Then fuse with potassium-sodium-carbonate, dissolve
the melt in water, acidify with hydrochloric acid, evaporate to dryness to
dehydrate the silica, take up with boiling water, filter on a weighed Gooch
crucible to separate the silica, which is 'then washed with boiling water until
free from salts, ignited and weighed. The filtrate is diluted to exactly 250 ml.
at 77 F. whereupon 200 ml. are drawn off, oxidized by boiling with nitric acid,
and the iron and aluminum precipitated with an excess of ammonia, filtered,
washed, ignited and weighed together as Fe203 and Ah03. Calcium, magnesium and S03 are then determined in the filtrate in the usual manner. The
remaining 50 mI. of the original filtrate are then reduced with metallic zinc and
sulfuric acid, and the iron titrated with N/l0 KMn04. The Al 20 3 is then
calculated by difference. The percentage of clay present in the original material is calculated from the following formula:
,I
Ah03' 2Si02 2H2b.
Gypsum is calculated from the following formula:
CaS042H20.
Potassium and sodium are determined in a separate portion.
Carbon Disulfide Extract: This contains the soluble bituminous constituents
(together with any chemically combined mineral constituents, e.g., iron present
in certain asphalts and fatty-acid pitches, lead present in the sludge asphalts,
copper present in certain fatty-acid pitches, etc.,F also any colloidal mineral
constituents not retained by the asbestos filter ,(e.g., clay, silica, calcium and
magnesium carbonates, etc.). Cool the solution to exactly 77 F., measure
its volume' and then dilute with carbon disulfide at 77 F. to the next higher
10 ml. mark. Thoroughly agitate the liquid and with a/pipette transfer an
7 It is contended that in natural rock asphalts" a.-portion .of the mineral matter is
chemically united with the asphalt in the form of salts of sulfonated acids.

>'

EXAMINATION OF BITUMINOUS SUBSTANCES

1543

aliquot portion into a tared crucible or dish. Evaporate the solvent over a
steam bath, then incinerate the residue and ignite at white heat to a clean ash,
until no further loss in weight occurs. Moisten with ammonium carbonate
solution, re-ignite at a low heat (not beyond incipient redness) and reweigh.
The ash represents the anhydrous clay, silica, calcium and magnesium carbonates, together with the mineral constituents originapy in chemical combination with the bituminous matters. Calculate the mineral matter associated
with the bituminous constituents in total carbon disulfide extract. Note that
any water of hydration originally combined with the clay is not ascertained
by the foregoing procedure, but the amount present is usually so small that it
may be disregarded without vitiating the results. From the weight of the
residue insoluble in carbon disulfide, calculate the soluble bituminous constituents by difference. Evaporate an aliquot portion of the well-mixed carbon
disulfide extract to exactly the calculated weight.
It has been found that when the solvent is expelled by heating over a waterbath for one hour, tlie fusing-point is increased 2 to 3 C.; also that the retention
of 1 to 2% of very fine mineral matter does not change the characteristics of the
recovered bituminous constituents to any appreciable extent.
The following variations have been proposed for recovering the bituminous
constituents in their unaltered state:
(a) Evaporation at Atmospheric Pressure.-From the weight of the extracted mineral matter, calculate the bituminous matter by difference, and
evaporate the carbon disulfide extract to exactly this weight. This may be
conveniently performed by distilling and condensing most of the carbon disulfide over an incandescent light or, an electric stove in a large distillation retort
connected with a condenser and receiving flask. The retort shall have a
thermometer inserted, with the bottom of the bulb approximately 1 in. from
the bottom of the retort. The concentrated solution is transferred to a tared
dish, evaporated dry on a steam bath with constant stirring,-which may be
conveniently accomplished with a motor-driven agitator, and the last traces
of solvent expelled in an oven at 105 C. until the residue attains the calculated
weight.
An alternate procedure consists in distilling the filtered extract (containing
a calculated amount of soluble constituents equal to approximately 100 g.) to
a bulk of about 150 mI., then transferring the solution to a spherical flask of
300 mi. capacity. The evaporation is completed on a water-bath which is
heated from 40 0 C. up to the boiling-point, at which it i~ maintained for 5
hours, during which period the contents are stirred at intervals with a tared
glass rod. Finally the flask is heated in an electrically heated oil-bath, from
an initial temperature of 1050 C. to a final temperature of 1300 C., which is
kept constant for a period ranging from 4 to 6 hours, during which the contents
are kept stirred with the tared glass rod. The oil-bath is kept agitated with a
motor driven paddle, which likewise causes the flask to revolve . The time of
heating in the oil-bath may be determined by dissolving asphalt of known
penetration in carbon disulfide .and ascertaining the time required at 130 C.
to bring its penetration back to its original figure.
.
(b) Evaporation under Vacuum.-The solvent is first removed by distillation as in the foregoing, until the temperature of the residue in the flask reaches
3000 F. The distillation flask is then connected to a filter pump and the

1544

EXAMINATION OF BITUMINOUS SUBSTANCES

distillation continued under a vacuum of 180 mm. mercury, raising the temperature to 500 0 F. at the rate of 5 to 6 F. per minute. If the original penetration is greater than 150 at 77 F., the distillation temperature should not
exceed 425 F. The residue in the flask should thenl be wei~ed. This 'procedure will give satisfactory results, provided. the bituminous substance does
not contain any appreciable quantity of constituents' {rolatile, at 400 F.
The following modification is claimed to give very accurate results. Dissolve the substance in C.P., carbon disulfide previously ,dried by contact with
CaCho The extract should contain 5 to 6% of the soluble constituents, and
not exceed 240 m!. in volume, which may be attained by evaporation of' by
dilution. A shallow, circular brass or nickel plate as illustrated in Fig. 204 i.s

FIG. 204.-Shallow Plate for Evaporation.

supported horIzontally (using a spirit level) in a vacuum-dessicator. About
60 ml. of the extract are poured into the plate and distributed uniformly over
its bottom, whereupon vacuum is applied to the dessicator (180 mm. mercury)
and maintained for 172 p.ours. This induces the evaporation of most of the
carbon disulfide. Any bubbles or blisters in the residue are pricked with a
knife, and the entire layer scratched with the point of the knife in criss-cross
directions as illustrated, so as to break it up into ridges, thereby exposing the
interior of the mass. It is then evacuated another 72 hour. The plate is then
warmed for 10 to 20 minutes at 70 C. in an oven to soften the film of bituminous substance, which is thereupon scraped 6~t with a spatula. Then
another 60 m!. of the extract are introduced in the vessel and treated in exactly
the same manner. This operation is repeated four times, or until all of the
original extract (240 m!.) has been evaporated. In this manner, about 15 to
16 g. of the asphalt is collected and utilized for further tests.
This same metho,d may be adapted to bitumindUs emulsions, in which case
j~ is modified as follows: About 35 g. of the emii'ision are broken by adding a

EXAMIN:ATION OF BITUMINOUS SUBSTANCES

1545

few drops of concentrated HCI. The separated ~lumps are freed from acid by
kneading in several portions of water with a glass rod. The asphalt is then
dissolved in about 240 ml. of carbon disulfide, allowed to stand in a separatoryfunnel until most of the water rises to the surface, whereupon the solution is
drawn off into a flask and shaken repeatedly with granular CaCh at 15 minute
intervals. The carbon disulfide solution is then poured off and treated as
described above to recover the dissolved asphalt.
(c) Evaporation in a Stream of CO 2.-In this case the carbon-disulfide or
benzol extract containing 50 to 75 g. of the asphalt is distilled in a wide-mouthed
250 m!. flask in a form of apparatus shown in Fig. 205. The flask is fitted with

FIG. 205.-Apparatus for EVaporation in Stream of CO 2 ,

a goose-neck vapor line of approximately 10 mm. internal diameter, connected

with a condenser. The CO 2 inlet consists of a copper tube of approximately
5 mm. internal diameter, bent so that its lower end forms a circle slightly less
than the diameter of the mouth of the flask, which is perforated with l2 in.
holes along its bottom to direct the gas downwardly. A thermometer is
inserted through the same cork so that the lower end of its bulb reaches >i in.
from the bottom of the flask and in the center of the circle of the copper tubing.
The flask is immersed up to its neck in an oil-bath, which prevents. foaming
during the distillation. The rate of flow of CO 2 is measured by a differential
gage or manometer-connected by means of tees to a l.mm. capillary tube, which
is previously calibrated. The gas is drawn from a cylinder through a pressureregulating valve, and is bled from this through a needle-valve connected to
the flow-meter. When the temperature of the asphalt reaches 300 0 F., the
CO 2 is admitted slowly, about 150 ml. per minute, for only a few seconds at a
time. This is repeated at 3 to 50 1<. intervals and serves to agitate the contents
of the flask and minimizes foaming. When the internal temperature reaches
3250 F., the gas is admitted at the rate of 875 ml. per minute and held for
exactly 15 minutes, whereupon the asphalt should be removed and poured
directly into tins or testing molds. A maximum temperature of 300 0 F. is
recommended for asphalts softer than 100 penetration at 77 0 F.; likewise 350 F.

1546

EXAMINATION OF BITUMINOUS ,SUBSTANCES

for asphalts having a fusing-point of over 150 0 F. (R. and B.). It is important
that the extracted asphalt solution be distilled on the I same day that it is dissolved, otherwise considerable alteration may occur if tlhe solution is allowed to
stand for more than 10 hours, regardless of wh,ether benzol or carbon disulfide
is used, and will result in a change in the physical charadteristics of the extracted
I '
asphalt.
A modification of the procedure consists in distilling off the solvent in' a
stream of CO 2 under reduced pressure.
A rapid method for separating asphalt from asphaltic limestones and dGlomites consists of the following: Weigh out 10 g. of the rock asphalt and add
75 m!. of a mixture of 30 ml. concentrated RCI (sp.gr. 1.19) and 45 ml. ethyI
ether. Stir 10 minutes, until the carbonates dissolve~ ,add 75 m!. water and
warm over a water-bath until the ether is expelled. Then filter and wash with
warm water. Dry the filter at 105 0 C. and extract the filter with.. carbon disulfide to recover the asphalt. The extract may be treated by any of the procedures described above.
The soluble bituminous constituents, after being separated in their pure
state, as previously described, may be then examined further for their physical,
and chemical characteristics. Due allowance should be made for the fact that
any bituminous matter insoluble in carbon disulfide will be retained mechanically by the extracted mineral constituents, which with asphaltic products is
usually relatively unimportant, but in the~ case of coal-tar products, will amount
to a considerable item.
Carbenes
The expression" carbenes" has been applied to that portion of bituminous
substances soluble in carbon disulfide but insoluble in carbon tetrachlor.ide.
This test is of value in identifying bituminous substances, gauging their uniformity of supply, for purposes of factory control, and as a criterion of their
quality. Certain hard native asphalts and asphaltites; particularly grahamite,
normally contain a percentage of carbenes, whereas petroleum asphalts do not
show carbenes unless they are overheated or over-bI6wn. If more than 0.5%
is present in petroleum asphalts, their quality is: to be regarded as questionable.
Carbenes are found in tars and pitches in varying amounts.
Although carbenes are found in grahamite a:r;J.d certain hard natural asphalts
when tested as such, they disappear upon fluxing to a softer consistency. With
petroleum asphalts, tars and pitches, the carbenes are of diffe:rent character,
since they are insoluble in fluxes and do not disappear upon being' so treated.
This test is carried out by following the same procedure as in determining
the solubility in carbon disulfide, but replacing the latter with carbon tetrachloride. The carbon tetrachloride must be free from carbon disulfide, which
may be insured by distilling it under a dephlegmatpr, discarding any distillate
below 76 C. The solvent is then filtered througli calcium chloride, and any
free hydrochloric acid removed by blowing dry air through it.
The procedure has been standardized as follows:
4n amount of material which shall contain approximately one gram of
bitumen shall be weighed into a tared. Erlenmeyer, :6~sk. 6ne hundred milliliters of chemically pure carbon tetrachloride .shalt be added to the flask in
small portions with continued agitation until all lumps disappear and nothing

EXAMINATION OF BITUMINOUS SUBSTANCES

1547

adheres to the bottom. The flask shall be corked and set aside in subdued
light for at least twelve hours.
The Gooch crucible shall. be set up again with the suction flask and the
carbon tetrachloride solution carefully decanted through the asbestos felt, with
or without light suction as may be found necessary. No sediment shall be
allowed to go onto the filter. A small amount of carbon tetrachloride shall be
used to wash down the sides of the flask and then the precipitate shall be brought
onto the felt and the flask scrubbed with a feather if necessary to remove all
precipitate. The contents of the crucible shall be washed with carbon tetrachloride until the washings are colorless. Suction shall be applied to the crucible to remove the carbon tetrachloride. The crucible shall be dried in the oven
at 100 to 125 C. for twenty minutes, cooled in the desiccator, and weighed.
In case insoluble matter adheres to the flask, it shall be dried and weighed,
and the increase in weight over the original weight shall be added to the weight
of insoluble matter in the crucible.
The proportion of bitumen soluble in carbon tetrachloride shall be reported
on .the 'basis of total bitumen taken as 100:
Proportion of Bitume.n Soluble in} = Bitumen Soluble in Carbon Tetrachloride
Total Bitumen
Carbon TetrachlorIde.
The difference between the p~rcentages soluble in carbon disulfide and
carbon tetrachloride respectively, represents the per cent of "carbenes."
Soluble in Petroleum Naphtha
This test is employed mainly for purposes of identification. It is also used
to a certain extent for determining the adaptability of bituminous substance
for a given use, for gauging the uniformity of supply, and for purposes of factory
control. As a general principle, the harder the bituminous product, the smaller
will be the percentage that dissolves in petroleum naphtha. Asphaltites are
relatively insoluble in this menstruum. Mineral waxes, peat-, lignite- and
shale-tars or pitches are largely soluble. The solubility of native and petroleum
asphalts varies, depending largely upon their hardness, and also in the case of
petroleum asphalts upon the extent to which the distillation has been driven.
Coal-tar pitches are relatively insoluble in 88 naphtha.
The portion soluble in petroleum naphtha has been termed "petrolenes"
by some, and "malthenes" by others, whereas the non-mineral constituents
insoluble in naphtha are generally referred to as "asphaltenes."

It is important that the petroleum naphtha should be derived from petroleum composed entirely of open-chain hydrocarbons, and test exactly 88
Baume, equivalent to a.specific gravity of 0.638 at 60 F.{60 F. At least 85%
by volume should distill between 95 and 150 F. -The density and character
of the naphtha is important, since l).eavy distillates, or products derived from
petroleum containing unsaturated or cyclic hydrocarbons, will exert a greater
solvent action upon the bituminous substance.
The results will be more consistent if the petroleum spirits is first washed
. with fuming sulfuric acid to remove the aromatic constituents. There appears
to be no difference in the results if the precipitation is carried out at any temperature between 0 and 32 C. As the boiling-point of the petroleum spirits

1548

EXAMINATION OF BITUMINOUS .SUBSTANCES

employed in making the test increases, the quantity of precipitate decreases.

The fraction below 105 0 F. appears to give the most teliable results.
Asphaltenes show increasing solubility in solvents in the ()rder of their
surface tension (e.g. ether, benzol, carbon disulfide add pyridine). EthElr has
been recommended as a substitutefor petroleum naphtha, because it is a homogeneous substance, not requiring standardization, and in addition has good
flocculating properties and exerts a greater solvent action on hydroxy acids
present in certain asphalts.
This method is performed in the same manner as for determining the portion
, soluble in carbon disulfide, petroleum naphtha being substituted for the latter.
Hard bituminous substances should be powdered; liquid bituminous substances
flowed in a thin layer over the bottom of the flask; and semi-solid to semi-liquid
substances heated until fluid and distributed in a thin.layer to present a greater
surface to the solvent. It is advisable not to use.a stirring rod, as this causes
the bituminous substance to adhere to the inner surface of the flask and to the
rod itself. The operation should take place at room temperature, and away
from the direct rays of the sun. The introduction of a weighed portion of
long-fibered asbestos to the solution will assist in its filtration.
The percentage of asphaltenes varies considerably; thus, with asphalts all
having the same R. and B. fusing-point of 1400 F., the following are obtained:
extracted asphalt from Trinidad asphalt 37%, Mexican residual asphalt 20%,
California residual asphalt 12%, Colombian residual asphalt 16%, Illinois
residual asphalt 12%, Texas residual asphalt 9 to 17%.

" Free Carbon"

This test is generally used for testing tars and pitches for the presence of nonmineral matter insoluble in hot benzol, although sometimes carbon disulfide is
used for the purpose. The test is of value for purposes of identification, for
ascertaining the adaptability of the tar or pitch for E!- given purpose, and for
gauging its uniformity of supply. Tars and pitches ,containing large percentages of insoluble matter, known as "free carbon," are objectionable for certain
manufacturing purposes, since the free carbon acts as so much inert matter.
The term "free carbon" is a misnomer, since it is not elemental carbon, but a
complex mixture of hydrocarbons of high molecular weight, containing 90.0
to 91.7% carbon, 3.4 to 4.0% hydrogen, 1.0 to 1.2% nitrogen, 2.5 to 3.3% oxygen and 0.7 to 1.4% sulfur, on the ash-free basis. The presence.. of hydrogen
has been explained by the great absorptive power of carbon injts pure state,
which retains hydrocarbons tenaciously, as well as hydrogen, which is not driven
off at temperatures as high as 800 0 C. Free carbon is more soluble in aniline
or pyridine than in benzol. or carbon disulfide. Selenium oxychloride exerts
the greatest solvent action upon it, but unfortuna'tely the residue cannot be
freed from this solvent. It is also partially decomposed by digesting with hot
fuming nitric acid.
Marcusson has found that the benzol-insoluble constituents of vertical- and
horizontal-retort coal-tars, amounting to 7% and 24% respectively, consisted
of oxy-acids 8.6% and 0.5%; pyridine soluble resins.-73% a~d 16.3%; pyridine
insoluble resins 18.4% and 32%; and partly coked material 0% and 51.2%

EXAMINATION OF BITUMINOUS SUBSTANCES

1549

respectively. The following method of extraction should be used. Tars must

be dehydrated before extracting, and pitches, if sufficiently hard.
(A) Porous Thimble Method.-The extractor shall be of a form shown in
Fig. 206, or of a similar form in which the substance is subjected to direct

.'Filf.rPap"..tIJ
Alundum

Cup

Courtesy A.S.T.M.

FIG. 206.-Extraction Apparatus.

Courtesy A.S.T.M.

FIG. 207.-Extraction Apparatus.

washing by the vapors of the boiling solvent. The filtering medium shall be a
flat-bottom, 30 by 80 mm. RA-98 alundum thimble. The thimble 'shall be
'suspended in the extraction flask either by a wire basket hung from two small
hooks on the under surface of the metal coyer of the flask, or it shall be supported by making perforations near the upper edge of the thimble and suspending from the cover by German silver or platinum wire.
An amount of material which shall contain 10.0 g. 0.1 g. of the substance
shall'be w~ghed into a 100-ml. beaker; 50 ml. of pure benzol shall be adiied and
the solution stirred thoroughly. The solution shall be transferred at once to
the weighed alundum thimble. The beaker shall be rinsed clean with pure
benzol and the washings added to the thimble. The thimble shall then be
covered with a lid of alundum ware and placed immediately in the extraction
apparatus. The extractor shall contain a suitable quantity of pure benzol and
shall be heated sufficiently to boil the solvent. The extraction shall be continued until the solvent descending fro~ the thimble is colorless. The thimble
shall then be dried at 1050 C.5 C., cooled in a desiccator, and weighed.
(B) Asbestos Mat Method:-The filtering medium shall consist of a No.3
Gooch, Coors porcelain or equivalent crucible approximately 3.5 cm. in diameter at the top, tapering to 2.2 cm., at the bottom with a depth of 4.0 em., containing a mat of acid-washed medium fiber asbestos approximately 2 mm. in
thickness.
The Gooch crucible shall be placed in the suction apparatus and filled with
acid-washed medium fiber asbestos suspended in water. Gentle suction shall
be applied and more of the suspension, if necessary, shall be added to 'make a
mat approximately 2 mm. in thickness. With the suction still on, the pad
shall be washed with water until all small particles of asbestos are removed.

1550

EXAMINATION OF

BITUMINOU~

SUBSTANCES

The crucible shall be dried at 1050 C.5 C., placed in a d~siccator, and
weighed.
I
An amount of material which shall contain 10.0 g. o.i g. of the substance
shall be weighed into a 125-ml. Erlenmeyer flask; 510 ml. .of pure beniol shall
be added and the solution stirred thoroughly and br:ought to a boil. The hot
solution shall be carefully poured into the weighed: prepared Gooch crucible
without suction until the mat is covered. Then gentle suction shall be a:rJpliecl,
to the crucible and the remaining solution added, taking care that the mat is
covered with solution at all times. The Erlenmeyer .flask shall be rinsed
clean with pure benzol and the washings added to the crucible. Benzol shall
be added to the crucible until the descending solvent is colorless. The crucible
shall then be dried at 1050 C.5 C., cooled in a desiccator, and weighed.
(C) Paper Thimble or Filter-paper Method.-A paper thimble or else two
thicknesses of S': & S. No. 575 or Whatman No.5 hardened filter-paper (15 cm.
in diameter, arranged in cup shape by folding) may be used in connection with
an extractor of the form shown in Fig. 207. In this case the procedure is
similar to that outlined in the Porous Thimble Method (A).
The insoluble residue of'''free carbon" in~ludes the mineral ash, which may
be ascertained by incineration. Some tars (e.g. blast-furnace tars) contain
substantial amounts of ash, which would ,otherwise introduce serious errors.
It has been found that the amount of free carbon will increase with the time
the solvent remains in contact with the tar or pitch. The amount of such
increase varies with the particular solvent employed. In the case of carbon
disulfide or chloroform, the increase is less than with benzol or toluol (or mixtures of these two) and the amount apparently reaches a maximum III about
120 hours.
A method has been proposed in which 5 g. of the pitch is heated with 200 ml.
tetralin in an autoclave at 240-250 0 C. under a pressure of 12-13 atmospheres
for 2 hours. Upon cooling, the liquid is filtered through a Gooch crucible .and
washed successively with 50 ml. tetralin and 100 ml. benzol. The residue is
dried for 2 hours at 150 0 C. in an atmosphere of CO 2 nd weighed. cAnthracene
oil (freed from crystallizable substances at 15 C.) has also been proposed, in
which case, 5 g. of the substance are digested with 5 g. anthracene oil under a
reflux condenser for 4 to 5 hours, cooled, dilut'ed with 500 ml. benzol, filtered,
washed with benzol, etc., as in the preceding.. Aniline has also been suggested
as extracting medium, involving the digestion of 5 g. of substance with 25 ml.
aniline over a stea'm bath for Y2 hour, ,pouring the liquid 'On a por~)Us clay plate
to absorb the solvent, washing the residue with pyridine, followed by benzol,
'
drying and weighing.

EXAMINATION OF BITUMINOUS SUBSTANCES

1551

CHEMICAL TESTS
Water
The estimation of water is made in some cases for purposes of identification,
and in others as a criterion of the quality. Native asphalts and tars are examined in tliis way to ascertain whether they exist in the crude or the dehydrated
state. This test is also used for dehydrating bituminous. substances to render
them suitable for further examination, where the presence of water would interfere.
Substances Distilling at Low Temperatures.-This method is adapted to
crude petroleum, tars, creosote oil and other' fluid bituminous substances
distilling at comparatively low temperatures. The apparatus is set up as
shown in Fig. 208. The copper still is provided with a removable flanged top
and yoke of the form and approximate dimensions shown in Fig. 209
.-TJ1 t rmommr

1........

Cpnnf!ng I
)

Tubl

3875,"' ..... ->1

. ..
t

''r------I

Separol11ry
Fun"el,~ ....

( C"pac/fy

120cc)

6radvaf~d

inJ cc.-

~ .. -.. 3.7.a2....~

(9(Om",.1: fi.lmm)

Courtesy A.S.T.M.

]fIG. 208.-Assembled Apparatus for Water Test.

Courtesy A.S.T.M.

FIG.

209.-Coppe~Still.

The condenser consists' of a copper trough carrying a straight-walled glass

tube. The separatory funnel has a total capacity of 120 ml. with the outlet
graduated in fifths of a milliliter.
Pour 200-500 m!. of bituminous material into the still and weigh. Clamp
the top in place, using a paper gasket moistened with lubricating oil. Apply
heat with the ring burner supported just above the level of the bituminous
material at the beginning 0:( the test, and then gradually lower it as the water
distils over. Continue the distillation until the vapor temperature reaches
205 0 C. Collect the distillate in the separatory funnel, and let it stand until a
clean separation of water takes place. Read off the volume of water, calculate
its weight, and figure the per cent present in the crude bitumino11s material.

1552

EXAMINATION OF BITUMINOUS SUBSTANCES

Draw off the water, and return any light oil to the bituminous matter in the still.
The dehydrated materjal should then be used for fuhher tests.
.
To prevent frothing and spattering, it has been recommended that the still
be surrounded with a cylindrical vessel, closed at the bottom and open at the
top, of a somewhat greater diameter than the still. 'The vessel is so adjusted
that its upper rim is a little lower than the level of th~ bituminous substance in
the retort, whereupon it is fill~d with water until it overflows. Heat is then.
applied to the shallow layer of tar above the water level, with the ring burner.
The water jacket absorbs heat which would otherwise be transmitted to the
lower level of the tar and cause bumping. The level of water falls gradually by
evaporation, permitting the zone undergoing dehydration to fall slowly. When
the water level has fallen below the bottom of the still, the contents are practically dehydrated.
A modification of the foregoing method has been suggested which consists in
providing the cover of the copper retort with two openings, one connected with
a receptacle holding 100 m!. of tar and provided with a stop-cock to control the
flow of tar into the retort. A weighed quantity of tar is introduced into the
receptacle and enough run into the retort to cover the bottom, and the remainder is allowed to drip slowly into the retort. This prevents the tar
frothing over.
'
Substances Distilling at High Temperatures.-This method is adapted to
asphalts and other bituminous substa]lces compara~ively free from volatile
constituents, and incapable of distilling without suffering decomposition.
Substances Fusing below 300 0 F.-When it is desired to determine the percentage of moisture without using the residue for other purposes, a convenient
method consists in distilling it with a solvent. The following procedure has
been standardized:
The apparatus shall consist of a metal still or glass flask, heated by suitable
means and provided with a reflux condenser discharging into a trap connected
to the still or flask. The trap serves to collect and measure the condensed
water and to return the solvent to the still. The t:ype of distilling apparatus
used is not an essential feature of this method, but glass has been generally used
for petroleum products and the metal still for ,road materials and tars.
(a) The metal still (Fig. 210 (a)) shall be a vertical cylindrical vessel, preferably of copper, having a faced flange at the, top to which the head is tightly
attached by means of a clamp. The head shall be of metal, preferably of brass
or cop~r, and be provided with a tubulation 1 in. in inside diameter.
(b) The glass flask (Fig. 210 (b)) shall be of the short-neck, round-bottom
type, made of well-annealed glass, having an approximate capacity of 500 m!.
The burner used with the metal still shall be,a ring gas burner 4 in. (100 mm.)
in inside diameter. With the glass flask, an ordinary gas burner or electric
heater may be used as the source of heat. The cOIfdenser shall be pf the watercooled, reflux, glass-tube type, having a condenser jacket not less than 400 mm.
(15U in.) in length with an inner tube 9.5 to 12.7 mm. (% to 72 in.) in outside
diameter. The end of the condenser to be inserted in the trap shall be ground
off at an angle of 300 from the vertical axis of the condense,r.
The trap shall be made of well-annealed glass .constr,~cted in accordance
with Fig. 210 (c) and shall be graduated as showh-;:--from 0 to 10 m!. in 0.1 m!.
divisions. The error of any indicated capacity shall not be greater than 0.05 ml.

EXAMINATION OF BITUMINOUS SUBSTANCE~

The outside diameters should be preferably 2.5 to 3.5 mm.
than the inside diameters specified.

1553

Uz to Ys in.) greater

tA...

!
i

..:1_

-1f.

-~--

A- ................ 45to55mm. _
B .. (l1f4idt Diarmhr) 14 .. 16 mm. -

C-(I",id, Di'am,ftr)f2 .. 16mm._

-l"

l" "

D... .. .. .. 150"nOmm._ 6"" ~l

E_ ...h .......
25" JIm",. , ... 'l.~
F- ..........40__100 "I08mm. _ 4 " ....itH-(Imid, Diam,hr) 18 " I~ mm. u

to,

'1-'"
21"
'f"

.-

i" II tM

(b)

Courtesy A.S.T.M.

FIG. 21p.-Apparatus for Water Test.

(a) The solvent used when testing petroleum products or bituminous materials derived from petroleum shall be gasoline, free from water, and shall conform to the following distillation requirements:
'

5% shall distill at a temperature not below 1940 F. (900 C.) nor above 212 0 F. (1000 C.)
90% shall distill below 410 F. (210 C.)
0

(b) The solvent used when testing bituminous materials derived from coal
tar, water-gas tar, etc., shall be a coal-tar naphtha or a light oil and shall conform to the following distillation requirements:

98% shall distill between 248 0 F. (1200 C.) and 482 0 F. (250 0 C.)
The sample shall be thoroughly representative of the material to be tested
and the portion of the sample used for the test shall be thoroughly representative
of the sample itself. Deviation from this requirement shall not be permitted.
N OTE.-The difficulties in obtaining proper representative samples for this determination are unusually great so that the importance of sampling cannot be too strongly
emphasized.
Procedure:When the sample to be tested contains less than 10% of water,
exactly 100 ml. of the material to be tested shall be placed into the still or flask
and thoroughly mixed with an equal volume of solvent by swirling, proper care
being taken to avoid any loss of material. If the material is measured by vo1-

1554

EXAMINATION OF BITUMINOUS SUBSTANCES

ume, an accurate 100 m!. graduated cylinder sha,ll' be used and, the contents
transferred to the still by rinsing with one 50-ml. I1ortion of solvent, followed
by two successive 25-ml. portions of solvent, the cylihder being allowed to drain
each time. When the sample to be tested contains, more than 10% of water,
the volume of material used shall be decreased to t~at which will yield somewhat less' than 10 m!. of water.
'
i

NOTE.-In special cases where the water content exceeds 10% and it is not desirable
to reduce the size of the sample to that which will yield somewhat less than 10 mI. of
water, a distilling tube receiver graduated from 0 to 25 m!. may be used. This tube
shall be graduated from 0 to 2 ml. in 0.1 mI., from 2 to 5 ml. in 0.2 ml. and from 5 to
25 ml. in 0.5 ml.

The connections between the still or flask, trap and condenser shall be made
by means of tight-fitting corks as shown. The end of the condenser inserted in
the trap shall be adjusted to that position which will allow the end to be submerged to a depth of not more than 1 mm. below the surface of the liquid in
the trap after distillation conditions have been established. When the metal
still is used, a heavy paper gasket moistened with the solvent shall be inserted
between the lid and flange before attaching the clamp. A loose cotton plug
shall be inserted in the top of the condenser tube to prevent condensation of
atmospheric moisture in the condenser tube.
Heat shall then be applied and so regulated that the condensed' distillate
falls from the end of the condenser at the rate Qf from two to five drops per
second. The ring burner used with the metal still should be placed about 3 in.
above the bottom of the still at the beginning of the distillation and gradually
lowered as the distillation proceeds.
The distillation shall be continued at the specified rate until no water is
visible on any part of the apparatus except at the bottom of the trap. This
operation usually requires less than an hour. A persistent ring of condensed
water in the condenser tube shall be removed by increasing the rate of distillation for a few minutes.
The v:olume of condensed water: measured in thy'trap at room temperature
multiplied by 100 and divided by the volume of the sample used shall be the
percentage of water and shall be reported as " ...... per cent water by volume,
A.S.T.M. method."
,
The accuracy to be expected with this method is that duplicate determinations of water should not differ from each other by more than one division on
the t r a p . ,
. Substances Not Fusing at 300 0 F.-In this case the material "is comminuted
by powdering (to about 60 mesh) or shaving, and a weighed quantity spread in
a thin layer on glass and maintained in an oven at 105 0 C. for one hour, or until
the weight becomes constant. If the substance is oxidizable in air, it should be
heated in an atmosphere of nitrogen or illuminating gas. Cool in a desiccator,
reweigh and calculate the .per cent moisture.
I
Oxygen (in Non-mineral Matter)
There being no satisfactory direct method for deterJliining oxygen, it is
computed by subtracting the sum of the perc~tages .of hydrogen, carbon,
nitrogen, sulfur, water and ash from 100%. TKe-result so obtained is affected

EXAMINATION OF BITUMINOUS SUBSTANCES

1555

by all the errors incurred in the other determinations, and especially by the
change in weight of the ash-forming constituents on ignition. Iron pyrites
will absorb oxygen from the air and change to ferric oxide, increasing the weight
of ash, and thereby causing a negative' error in the oxygen, equivalent to threeeighths of the pyritic sulfur. Any calcium carbonate present will tend to
a!bsorb sulfur combined with the bituminous constituents. On the other hand,
there is always a loss on ignition' of "water of composition" from the clayey
and shaley constituents, also carbon dioxide from carbonates, etc., which tend
to compensate for the absorption of oxygen.
A more correct value is obtained by making the corrections indicated in the
following formula:
.
.

Corrected Oxygen = 100 - [( C.- C') + (H - H') + N + H 2 0+

8'+ Corrected
Ash]
,
,
Where C equals the total carbon; C' the carbon of carbonates; H the total
hydrogen less hydrogen of water; H' hydrogen from water of composition in
clay, shale, etc.; N the nitrogen; H 20 the moisture as found at 1050 C.; 8' the
sulfur' not present as pyrite or sulfate; "Corrected Ash," the mineral constituents
originally present, which for most purposes may be calculated with sufficient
accuracy by adding to the ash as found, five-eighths of the weight of pyritic
sulfur, the CO 2 of carbonates, and the water of composition of clay, shale, etc.
Solid Paraffins

Until recently, it was considered that the presence of paraffin was an indica,tion of the quality of asphaltic products, and many specifications stipulated
the maximum percentage permissible. It is now generally conceded, however,
that there is no rational relation between the solid paraffins in asphaltic
products and their quality. The determination of paraffin is therefore of
value only for purposes of identification. Traces of solid paraffins are found in
asphaltites, natural asphalts, and in asphalts produced from strictly asphaltbase petroleums. On the other hand, more or less paraffin is present in asphalts derived from non-asphaltic and semi-asphaltic petroleums. It is absent.
in tars and pitches derived from high-temperature distillation processes.
The following method does not give absolute figures, since it merely discloses the paraffin hydrocarbo!lS which are solid at room temperature, without
taking the liquid paraffins into consideration; nevertheless the results are of
value for purposes of comparison. Weigh 50 g. of the material in a tared 6-oz.
glass retort, and slowly distil until nothing but a residue of coke remains. The
distillation should take in the neighborhood of forty-five minutes from the time
the first drop comes over. The distillation is conducted'by heating at the rate
of 40 to 50 0 F. per minute to 450 0 F., followed by a rise of 150 F, per' minute to
5000 F. The distillate is caught in an Erlenmeyer flask, and weighed. Either
5 or 10 g. of the well-mixed distillate, depending upon the quantity of solid
paraffins present in the crude material, are transferred into a large test tube
and dissolved in 25 ml. of absolute ethyl ether and 25 ml. 'of absolute ethyl
alcohol. A similar mixture containing 25 ml. each of ether and alcohol is
made up, and this together with the oil solution is cooled separately to exactly

1556

EXAMINATION OF BITUMINOUS SUBSTANCES

0 F. for one-half hour in a mixture of ice and salt (to which if necessary calcium
chloride may be added). The oil solution is then :r:apidly filtered through a
weighed Gooch crucible similarly maintained at 0 F. by a jacket of ice and salt,
and w~shed with 50 ml. ?f the co?led ether-alcoho~\ ~ixt~re. A sim~le and
convement apparatus consIsts of an Illverted bottle 6 In. III dIameter, havmg,the
bottom cut off, and attached to the same rubber stopper which supports the
funnel holding the Gooch crucible. The space between the bottle, the crucible,
and the supporting funnel is packed with the ice and salt mixture: The Gooch
crucible is then removed, the outside wiped clean, placed on a tared glass and
dried in an oven at 80 C. until the last traces of ether and alcohol are evaporated. The residue is weighed, and the perce~tage of solid paraffins calculated
in the original 50 g. of substance taken for analysis.
Various substitutes have been suggested for the ether-alcohol mixture, including methyl-ethyl-ketone (butanone) containing 1.3% water, having a
specific gravity of 0.812 at 20 C.; also nitrobenzene used at 32 C.
It is a mooted question whether the solid paraffins ascertained by the
foregoing method are originally present as such in natural asphalts and asphaltites, or are generated during the distillation process through the destructive
decomposition of other non-paraffinic components.
An alternate method used with success qy the author -consists in dissolving
10 g. of the substance in the smallest amount of benzol, applying heat if necessary, adding 200 ml. of warm 88 naphtha, filtering through dried fuller's earth
and washing with additional warm 88 naphtha until the filtrate is clear. This
removes the dark-colored asphaltic substances, which remain adsorbed by tire
fuller's earth. The filtrate containing the paraffin is distilled to a small bulk,
evaporated to dryness on the water bath under-reduced pressure, the residue
dissolved in 25 ml. of ether and 25 ml. of alcohol, and treated as described
above.
Instead of removing the dark-'colored asphaltic constituents by means of
fuller's earth, an alternate procedure consists in extracting them by shakipg the
88 naphtha solution three successive times with 20 illl. concentrated sulfuric
acid in a separatory funnel, followed with N KOB: (in 50% alcohol-water
solution) and finally with water until neutral to phenolphthalein.
Another method for separating paraffin wax from petroleum residues consists in mixing the residue with an excess of full.er's earth, cooling and extracting
either with: (1) cold benzol (which removes both- the oil:y matters and th.e asphalt), or (2) cold gasoline (which removes the oily matters only) and then with
cold benzol (which removes the asphalt). The residue of fullei's earth is then
treated with hot water which separates the adsorbed paraffin wax. This
procedure is precisely the reverse of the prec_edi~g, which may however be
explained by the fact that in this case the process is carried on in the cold.
According to Holde, the refractive index of the ,solid paraffins will indicate
whether the original substance was ozokerite or paraffin, or a mixture of the
two. When tested at 90 C. on the Zeiss or Abbe refractometer, ozokerite will
show a refractive index below 15.0, whereas the solid paraffins derived from
petroleum, shale, ligniteJ etc., will test between 15.0 and 30.0. When tested
in the solid state, commercial paraffin waxes show_a refractive index ranging
from 1.500 to 1.540 at temperatures between 31J" and 130 F. Furthermore,

EXAMINATION OF BITUMINOUS SUBSTANCES

1557

para.ffin wax has a lower molecular weight than ceresine or ozokerite-for

melting points between 50 and 60 C., the former'tests between 330 to 400,
and the latter above 450.
Sulphonation Residue

(A) Sulfonation Residue Insoluble in Concentrated Sulfuric Acid.-This

method of test has been proposed for separating saturated and unsaturated
hydrocarbons, as a means of destinguishing various asphaltic products (e.g .
native asphalts and petroleum asphalts) from tars and pitches, thereby serving
for purposes of identification. Two alternate procedures are available, viz.:
(a) Portion Soluble in 88 Petroleum Naphtha.-The portion soluble in 88
petroleum naphtha is brought to exactly lOO ml., either by adding more 88
naphtha or else by evaporation, so that the quantity of substance carried in
solution will be approximately 5 g. This is then shaken in a 500-ml. separatory
funnel at 77 F. for exactly three minutes, with 30 m!. of a mixture of concentrated sulfuric acid and fuming sulfuric acid, having a specific gravity of 1.84
at i1 F. The funnel is allowed to stand quietly overnight, whereupon the
acid is drawn off and the oils unacted upon treated with another 30 ml. of the
acid. This time a few hours' standing should effect a sharp separation. If
the. second acid layer is strongly colored, the treatment should be repeated a
third time. The naphtha solution is washed successively with water, a 5%
solution of sodium carbonate and finally with water. The solution is evaporated to dryness' over a steam bath and the residue weighed. This is equal
to the saturated hydrocarbons present in the portion soluble in 88 petroleum
J}.aphtha. As a guide in evaporating the last traces of naphtha from the saturated hydrocarbons, a blank test should be run on 100 ml. of the 88 naphtha,
whereupon the portion unacted upon is mixed with 0.75 g. of a non-asphaltic
petroleum residuum and evaporated on the steam bath alongside of the sample
under test, until the former is reduced to exactly its original weight.
The results are expressed as the percentage of saturated hydrocarbons
. present in the portion soluble in carbon disulfide. This is calculated in the
following manner: If a represents the percentage soluble in carbon disulfide, b
the percentage soluble in 88 naphtha and e the percentage of saturated hydrocar:bons in b; then the saturated hydrocarbons present in the portion soluble in
carbon disulfide will equal (be/a) X 100.
(b) Distillate at 315-355 C.-This test expresses the percentage of saturated hydrocarbons in the distilla~e between 315 and 3550 C. obtained upon
subjecting the bituminous substance to the flask method of distillation. It is
used to differentiate tars and pitches among themselves as well as from mineral
waxes, asphalts (native and pyrogenous) and asphaltites. The figures for coaltar pitches have already been given. The author cites the following additional
figures: wood-tar pitch 0%, saponifiable fatty-acid pitches 0%, un~aponifiable
fatty-acid pitch~s 0%, residual asphalt from Mexican asphaltic petroleum 86%,
wurtzilite asphalt 87% and gilsonite 85%.
The method of determining the sulfonation residue consists in distilling
sufficient of the material under examination by the flask method to obtain at
least 10 m!. of distillate between 315 0 and 3550 C. Exactly lO m!. of this
fraction are measured into a Babcock milk bottle, and 40 m!. of 37 normal

1558

EXAMINATION OF BITUMINOUS SUBSTANCES

sulfuric acid added, 10 m!. at a time. The bottle and its contents are shaken
for two minutes after each addition, and when all the acid has been added, the
bottle is kept at a constant tempe,rature of 98-lOqo C. for one hour, during
which it is shaken vigorously every ten minutes. At the end of the hour, the
bottle is removed, cooled, filled to the top of the graduations with ordinary
sulfuric acid, and whirled for five minutes in a Babcock separator. The unsulIonated residue multiplied by 2 gives the per cent by volume directly '(each
graduation being equal to 1/200 mI.).
'
It is important that the acid should be of the proper strength. A mixture
of fuming sulfuric acid and ordinary concentrated sulfuric acid should be prepared to contain exactly 80.07% of 80a, which constitutes 37 normal acid. If
the sulfonation residue is dark in color, it should be treated with an excess of a
10% sodium hydroxide solution, and if completely soluble in this reagent, the
test is regarded as negative.
(B) Sulfonation Residue Insoluble in Water.-Treat 3 g. of the carbon disulfide or benzol soluble portion of the substance, after evaporation of the solvent,
with 6 ml. of' concentrated sulfuric acid (95%) in a stout test-tube and 'l,gitate
continuously for three-quarters of an hour while heating in boiling water. The
contents of the test-tube are then poured into 500 ml. of cold water. Browncoal tar and brown-coal-tar pitch are practically insoluble in cold water a11d
partly soluble in hot water. Coal-tar pitch, wood-tar pitch, oil-gas-tar pItch
and naphthol pitch are completely soluJ:.>le in cold water, forming a deep black
solution, which upon filtering leave little to no residue. Lignite-tar pitch, peattar pitch, bone-tar pitch, rosin pitch, fatty-acid pitch, wool-fat pitch and
montan pitch are partly soluble in cold water. Asphalts (both natural, as
well as those derived froni petroleum) and asphaltites are insoluble in cold
water, forming a copious insoluble precipitate. Let the solution stand fO,r two
hours and without agitating the precipitate, decant carefully through a weighed
Gooch crucible of about 7 cm. diameter. Wash the residue with hot water,
and finally transfer the precipitate to the filter. Continue washing with water
until the filtrate no longer gives an acid reacti<;m wi~h methyl orange, then dry
at 105 C. and weigh. Marcusson used this method for determining the percentage of coal-tar pitch in admixture with refined Trinidad asphalt. He
found that upon treating the C8 2 soluble portion of this asphalt with sulfuric
acid as outlined above, a small amount of soluble sulfo-derivatives were formed,
so that an average correction of 4% must be made to obtain 'accurate figures.
The foregoing test h~s been modified for the ostensible purpose of yielding
more accurate results, as follows: 2 g. of the substance are weighed in a test-tube
3.5 cm. in diameter, which is immersed 12 cm. in a 30% solution of salt, maintained at 107-108 C., whereupon 20 g. of concentrated sulfuric acid (95%) are
added. A mechanical glass stirrer is introduced and caused to revolve at the
rate of 300 turns per minute and the digestion cOhtinued for 2Y2 hours. The
mechanical stirring is claimed to break up the lumps and result in a better
filtration. Carefully introduce water in the test-tube, to a point 12 qm. from
the bottom, and continue the stirring for another 15 minutes, then transfer the
contents of the tube into 1 liter of boiling water. Let stand over night and
carefully filter through two layers of tared filter. paper in'to a 9 cm. porcelain
filter, using suction. Wash the residue with boiling water until the filtrate is

EXAMINATION OF BITUMINOUS SUBSTANCES

1559

clear and no longer reacts acid with methyl orange. The filter containing
the residue is placed on a Petri dish and dried for 2 hours at 105 C., cooled in
a dessicator and weighed, This represents the asphaltic constituents present,
which, however, are contaminated with a certain proportion of insoluble sulphoderivatives of the asphalt. An examination of a range of petroleum asphalts,
as well as Trinidad asphalt, indicated that the weight of insoluble asphaltic
constituents with their associated insoluble sulfo-derivatives, as ascertained
by this method, must be corrected by deducting 20%, to obtain the net weight
of asphaltic constituents actually present. Similarly, an examination of a
number of coal-tar pitches showed that in no case was more than 0.1 % of '
insoluble sulfo-derivatives present. No correction need therefore be made for
pitches. Any free carbon present in the pitch (ascertained by the insoluble
matter separated upon originally dissolving the substance in carbon disulfide)
should be taken into account by using the following formula:
o

% Asphalt=80%(100-C)'E
where: C=percent free carbon present in the mixture; S=grams residue obtained after sulfonation; E=grams mixture (after having been freed from
insoluble constituents and carbonaceous matter) taken for the sulfonation
process.

Saponifiable Constituents
Certain bituminous substances, such as montan wax, rosin pitch, and fattyacid pitch, are often composed largely of saponifiable constituents. Others,
including pine tar, pine-tar pitch, hardwood tar, hardwood-tar pitch, peat tar,
lignite tar, Done tar, bone-tar pitch and other forms of fatty-acid pitches, contain smaller percentages. This test is also used for gauging the uniformity of
supply, and in the case of fatty-acid pitches, as a criterion of the quality.
The following procedure has been devised by the author for specifically examining
bituminous materials or admixtures of bituminous materials with animal or vegetable
oils and fats, since the customary' methods do not adapt themselves especially well,
due to the formation of troublesome emulsions. The bituminous material is first freed
from insoluble constituents, including any mineral matter, by boiling with benzene
under a reflux condenser, cooling and filtering through a Gooch crucible. . The insoluble
constituents are dried at 100 C. and weighed. Sufficient of the bituminous substance
should be taken to yield approximately 5.0 g. of extract. The benzene' solution is
evaporated or distilled to 50 ml. and 50 ml. of the saponifying liquid added from a
pipette. This should consist of a 10% solution of caustic potash, prepared by dissolving
100 g. of anhydrous potash in 500 ml. of 95% ethyl alcohol, and diluting to a litre with
90% benzene. The liquid is allowed to stand overnight to permit any carbonate to settle,
and the clear solution decanted. After the saponifying agent is added, the mixture is
boiled under a reflux condenser, for ~-1 hour, and the contents of the flask while still
warm poured in a separatory funnel containing 150 ml. of boiling water and 25 ml. of a
10% solution of potassium chloride. Add 250 ml. of benzene, agitate vigorously, and
allow the funnel to rest quietly in a warm place until the solvent separates. If an emulsion forms which refuses to se:parate on standing, add 200 ml. more benzene and 100 ml.
95% ethyl alcohol and stand III a warm place overnight. This will invariably effect a
more or less complete separation of the solvent. From this point on the method is
illustrated by the following tabular outline:

1560

EXAMINATION OF BITUMINOUS SUBSTANCES

Saponify as described.
Draw off the soap solution as completely as possible.
Decant the benzol layer, leaving the intermediate layer in the separatory funnel.
Aqueous Soap Solution.
Exhaust with 200 ml. portions of benzol:
Aqueous Soap
Solution.

Benzol Layer.

Intermediate Layer.
'<

Combined Benzol
Extracts.
Combine and exhaust with 100 ml. pOi'!
tions of 50% alcohol.
Benzol Solution.

Combined Alcoholic
Extracts.
Combine and exhaust with benzol.
Combined Benzol
Extracts.

Combine, evaporate to a'small bulk, complete the evaporation at 1000 C., cool
and weigh the Unsaponijiable Consti/
tuents.

Alcoholic Soap Solution.

,--=-----Combine----

Transpose with dilute hydrochloric acid, warm and exhaust with benzol. Separate
the aqueous solution containing the glycerol and mineral salts. Evaporate the combined benzol extracts to a small bulk, and then complete the evaporation of solvent at
1000 C. Cool and weigh. Weight equals the free acids derived from the saponijiable
constituents.
In the case of bituminous materials that are more or ,less completely saponifiable,
the intermediate layer is apt to be absent. In this, case the process will simplify. itself
considerably. The foregoing procedure will separate the unsaponifiable constituents
in practically an ash-free state.

Diazo Reaction
This test is used for identifying bituminous substal1.ces carrying phenols,
including wood tar and wood-tar pitch, oil-gas- and wate'r-gas-tar.s~ and pitches,
shale tar, peat- and lignite-tars and pitches, bone-tar, bone-tar pitch and the
:various coal-tar pitches.
This reaction is carried out by boiling 2 g. of the bituminous substance with
20 ml. N. aqueous caustic soda, for approximately five minutes. After 'cooling,
the liquid is filtered. If the filtrate is dark colored, it may be lightened by
shaking with finely pulverized sodium chloride and filtering. It is then cooled
in ice to 10 Q., and 'a few drops of freshly prepared diazobenzolchloride solution
added (prepared by dissolving 1 g. aniline chlorhydrate in 10 ml. water and 3 ml.
25% hydrochloric, acid, and then adding drop by drop, a saturated solution of
0.5 g. sodium nitrite in water). To avoid the preparation 'of diazobenzolchloride (which does not keep), the use of paradiaz6benzol-sulfonic acid (C 6 H 4 N 2 SOa) may be substituted, since it is a more stable chemical and will keep well

EXAMINATION OF BITUMINOUS SUBSTANCES

1561

in' stock. If phenols are present" a more or less fugitive red coloration will
result, sometimes accompanied by a reddish precipitate.
Assuming that the bituminous substance gives the diazo reaction, the
question will often arise whether the product is a straight-distilled pitch, or an
asphalt "cut-back" with a high boiling-point distillate containing phenolic
bodies, derived from coal tar, lignite tar, etc. Marcusson has worked out a
method applicable under these circumstances, which consists in dissolving 10 g.
of the bituminous substance in 15 ml. of benzol, and pouring the solution into
200 ml. of 88 petroleum naphtha. The resulting precipitate is filtered and
washed with petroleum naphtha and dried. It is then boiled for fifteen minutes
with N /2 alcoholic caustic potash under a reflux condenser to extract the '
phenols. The liquid is cooled and filtered, the alcohol evaporated, and the
residue dissolved in water. Sodium chloride is added to clarify 'the liquid and
remove any substances imparting a dark color, the solution is filtered and the
filtrate treated for the diazo test described above. If a straight-distilled pitch
containing phenols is present, a positive reaction will be obtained. If the
origip.al substance gives the diazo test, but the residue treated in the above way
does not, then the admixture of high boiling-point oils containing phenolic
bodies with a substance free from phenols (e.g., asphalts, etc.) is established.
It is claimed that the presence of as little as 10% of pitch containing phenols
may be detected in this manner.
Where bituminous substances contain calcium carbonate, the phenolic
bodies present combine with the lime, forming insoluble calcium phenolate
which yields but a faiht diazo reaction. However, on treating such substances
with a solvent in the presence of hydrochloric acid, the calcium phenolate is
decomposed, and the diazo reaction becomes much more delicate.
A still more sensitive test for ascertaining the presence of phenols (e.g. tars
or pitches) in asphalts consists in the following: A reagent is prepared by dissolving 0.2 g. paranitraniline in 20 ml. water and 5 ml. 20% sulfuric acid, whereupon 0.3 g. sodium nitrite are added. Hard asphalts (3 to 4 g.) are ground in
a mortar with 25 ml. water made alkaline with NaOH; then filtered and the
filtrate acidified with H 2S0 4 and 3 drops of the foregoing reagent added. The
solution is finally made alkaline with NaOH, whereupon the presence of phenols
will develop a more or less intense red coloration. It is claimed that the presence of 1 % of tar or pitch may thus be detected, inasmuch as a 1 : 25,000
solution of phenol will produce a blood-red color, and a 1 : 250,000 solution a
pinkish-red color. Pure asphalts, on the other hand, yield a light yellow color
or leave the reagent unchanged.
The Dutch Standardization Committee have adopted a combination of
paradiazobenzol-sulfonic acid with Millon's reagent for detecting the presence
of phenols in substances containing colophonium, as follows: 10 g. of the substance are boiled 20 minutes with 25 ml. N aqueous NaOH and filtered. To
5 ml. of the filtrate add a few drops of paradiazobenzol-sulfonic acid. If no
red coloration results, phenol (tar or pit()h) is absent, but if a coloration occurs,
tar or pitch may be present, but it, may also be caused by colophonium. In
this event, the remainder of the filtrate is almost neutralized with nitric acid
and evaporated to 5 ml. Then add 5 ml. freshly prepared Millon's reagent and
place in a glass beaker with boiling water. If no coloration results within half
an hour, the mixture is free from phenols.

1562

EXAMINATION OF BITUMINOUS SUBSTANCES

Anthraquinone Reaction

The anthraquinone reaction is used for 'detecting dnthracene in ta,.r products

produced at high temperatures, including oil-gas-ta~ and pitch, wd'ter-gas-tar'
and pitch, and the various coal-tar pitches. l'his test is therefore valuable for
I
purposes of identification.
The tar or pitch is first subjected to distillation in accordance with the letort
method, the offtake and condensing tube being kept warm to prevent the accumulation of any solid distillate. The distillate passing over between 270 0 and
355 0 C. is caught separately and examined for anthracene in the following
manner. The fraction is heated until it is thoroughly fluid to secure a uniform
sample, and 5 g. weighed out, while hot. After cooling, 10 m!. of abf'olute
ethyl alcohol are added, the solids allowed to crystallize and the liquid decanted.
One to 2 g. of solid substances containing the anthracene are dried on a waterbath, transferred to a 500-ml. flask connected with a return condenser, 45 m!.
of glacial acetic acid added, and the contents boile<). for two hours. Thefollowing mixture is then added drop by drop through a separatory funnel, viz.: 15 g.
of anhydrous chromic acid dissolved in 10 m!. of glacial acetic acid, and 10 m!.
of water. The boiling is 'continued for another two hours, the flask cooled, and
400 m!. cold water added. This treatment oxidizes the anthracene to anthraquinone, which on cooling separates as a s6lid mass. This is filtered, washed
with hot water, then with a hot 1 % solution of caustic soda and again with hot
water. The residue of anthraquinone is then dried and its weight multiplied
by 0.856 to obtain the corresponding weight of anthracene. From 0.25 to
0.75% of anthracene is found in coal tars, and a correspondingly larger percentage in coal-tar pitches.
A color reaction for establishing the presence of anthracene consists in boiling for Yz to 1 hour the crystals of anthraquinone (1 part) with zinc dust (2
parts) and 50% NaOH solution (30 parts), whereupon an intense red-colored
solution is obtained, which on filtering in contact with air becomes decolorized.

ExAMINATION OF BITUMINOUS SUBSTANCES

1563

PART II
EXAMINATION OF BITUMINOUS SUBSTANCES COMBINED WITH
DISCRETE AGGREGATES
(A) PHYSICAL TESTS OF FINISHED PRODUCT

Resistance to Displacement
This test has been proposed for examining fine-aggregate paving mixtures
(e.g. the wearing-course of sheet-asphalt pavements) and has also been adapted
for testing coarse-aggregate paving mixtures. The procedure when testing
fine-aggregate mixtures consists in weighing the dry sand and filler to 0.1 g.;
heating to 3500 F. and mixing thoroughly. The melted asphaltic binder heated
to 3500 F. is then weighed. to 0.1 g. and thoroughly mixed with the aggregate
by means of a spatula. Approximately 100 g. of the completed mixture are
weighed into a cylindrical forming mold 2-in. in diameter and 4-in. high. The
mold and contents are heated to 250-260 0 F. as determined by the thermometer
used for stirring the mixture. Remove and insert a cylindrical plunger into

I+----r::~.~.

of Compression
Machine

(diameter 2-in.)

t===k--- Clamp

.....__----Cirt;ular Washer
(opening 1.7S.in.)
._----Test Specimen
(2.in. diam.b y I-in high)

FIG. 211.-Apparatus for Testing Resistance to Displacement.

the mold, and then tamp by first giving 60 blows with a tamper weighing 500 g.,
followed by 15--20 blows with a tamper weighing 1400 g. The mold and contents are placed in a compression machine to which is applied a load of 3000-lbs.

1564

EXAMINATION OF BITUMINOUS SUBSTANCES

per sq. in. and the briquette allowed to cool under pr:essure. The compressed
briquette should measure 2-in. in diameter and appfl>ximately I-in. in height.
Let stand at room temperature over night and then, maintain in a water-bath,
at 140 F. for one hour before testing. Two such briquettes should be prepared, each specimen being tested by placing it top lend down in the testing
mold and inserting the plunger as illustrated in Fig. I 211. A circular washer
having a circular opening 1.75 in. in diameter is clamped on the lower ,end of
the mold and the load then applied at a rate which will cause the testing head
to be depressed I-in. in 25 seconds. As the pressure increases, the briquette
will be forced through the orifice of the circular washer. The reading is observed on a spring dyn!Lmometer and the maximum is recorded. The average
of two specimens is taken as a measure of the" stability" of the mixture. This
test indicates the resistance of the pavement to displacement and may be used
to proportion mixtures of sand, filler and asphaltic cement giving optimum
results.
For testing coarse-aggregate paving mixtures a larger mold is recommended
in conjunction with a testing ring % in. thick having a minimum orifice 5:J,i
in. in diameter for a depth of Ys in., after which it tapers out to a diameter of
6 in. at the bottom of the ring.
Samp~es of the finished mixture taken from the testing-plant are warmed to
250-260 F. and formed into briquettes as described above. Finished pave
, ments may be tested by cutting out cores 2 in. in diameter and approximately
134 in. high, which are then placed directly in the mold arid tested at 140 F.
Tensile Strength
Any standard testing machine may be used provided it is accurate to w1thin
1 % of the lowest load to be applied. Special specimen holders of steel, as
shown by Fig. 212 (a), shall be used. The specimen holders shall be attached
to the heads of the machine by links held so that the pull is central, to avoid
any transverse strain. For hot-molded materials an~ for plastic cold-molded
materials, the test specimen shall be molded to the fo,rm and dimensions shown
in Fig: 212 (b).
.
NOTE.-To permit the use of one mold for all molding compounds the following
variations in dimensions of the test specimen will be permissible:
For hot-molded compounds ................... " 5%
For cold-mold~d compounds ................ , .. .,~. lO%
All specimens shall be conditioned before testing by drying in an oven at
50 C.5 C. for 48 hrs., followed by cooling to room temperature in a dessicator. Five specimens shall be tested at room temperature. The speed of
the head of the testing machine shall be such that the load can be accurately
weighed, but shall not exceed 0.050-in. (1.27 mm.) p,er minute with the machine
running idle. All tests shall be made at normal room temperature of about
20 C. (68 F.).
.
The report shall include the following:
(a) The breaking load of each specimen in pounds or kilograms;
(b) The thickness and width of each specimen,in. inche's or millimeters as
measured by a micrometer at the center of the- sP6.Cimen, that is, the point of
minimum section;

EXAMINATION OF BITUMINOUS SUBSTANCES

1565

(c) The ultimate tensile strength in pounds per square inch or in kilograms
per square millimeter of each specimen, calculated from the minimum area
measured at the center of the test specimen before the load is applied;
(d) The character of the material tested, with description of the fracture
and its location .wit.h respect to the middle point of the specimen.
(e) The speed in inches or millimeters per minute at which the head of the
testing machine traveled during the test.

-r,:: . ....,- -----

.. : "i'
~~~

Compressive Strength

This method has similarly been standardized as follows:

Any standard testing machine may be used provided it is accurate to
within 1 % of the lowest load for which it is used. One end of the specimen

1566

EXAMINATION OF BITUMINOUS SUBSTANCES

shall bear upon an accurately centered spherical bearing block, located whenever practicable at the top, and the metal bearing plate~ shall be directly in
J
contact with the ends of the test specimen.
The test specimen shall be molded in the form of a cylmder having a height
of 1.25 in. (31.8 mm.) and a diameter of 1.125 in. (28.6 mm.) (Fig. 213). The
ends shall be flat and perpendicular to the axis, bein~
ground to this condition if neces.sary.
_.
NOTE 1.-To permit the use of one mold for all molding"
compounds the following variations in dimensions of the test
specimen will be permissible:
'

!
' . ' -r,"

For hot-molded compounds .................... , 5%

For cold-molded compounds ..................... 10%

NOTE 2.-Due to the size and shape of this specimen, it is

sometimes difficult to completely cure it within the relatively
short time which would ordinarily be used and with the temperatures ordinarily used. This is particularly true when hand
_:i:..
molds, that is, molds heated by conduction from the platens,
Courtesy A.S.T.M. are used. It is, therefore, recommended that special pre!)au-'
tions be taken to insure thoroughly cured parts, experiments
FIG. 213.-Compres- indicating that 20 to 30 minutes may be necessary in some cases
sion Test Specimen. with hand molds.
:;

T""""

All specimens shall be conditioned before testing by drying in an' oven at

50 C.5 C. (122 F.9 F.) for 48 hrs., followed by cooling to room temperature in a desiccator. Five specimens shall be crushed at room temperature.
Additional tests shall be made at elevated temperatures, the actual temperatures selected depending upon the use that is to be made of the material.
Five specimens shall be tested at each temperature, selected. The speciinen
shall be kept at that temperature a sufficient length of time to become of uniform temperature throughout 'and shall be maintained at that temperature
during the test.
The load shall be applied on the ends of the, spec~fnen and the cross-head
speed of the testing machine shall be such that the load can be accurately
weighed but shall not exceed 0.050 in. (1.27 mm.) per minute when the machine
is running idle. For the best results use the slowest possible speed.
The report shall include the following:
(a) The dimensions of, each specimen in inches or in millimeters;
(b) The Joad on each specimen in pounds or kilograms at theJirst sign of
failure;
(c) The ultimate compressive strength in pounds per square-inch or kilograms per square millimeter of each specimen, ,calculated from the measured
area of each specimen before the load is applied;
(d) The general character of the material teste~; with description of how it
acts under the applied load;
I
(e) The speed in inches or millimeters per minute at which the head traveled
during the test. The time required for testing each specimen should be recorded.

EXAMINATION OF BITQMINOUS SUBSTANCES

1567

(B) SEPARATION OF FINISHED PRODUCT INTO ITS

COMPONENT PARTS
Separation of the .Bituminous Matter and Discrete Aggregate
Bituminized aggregates are separated into their bituminous and discrete
components for the combined purposes of ascertaining the percentage and
nature of the mineral constituents and for examining the physical and chemical
characteristics of the bituminous binder, with the object of its identification
or duplication .
. Hot Extraction Method.-This has been standardized as follows:
~ Bituminous grouts 'shall be heated in an oven or on ahot plate in a pan or
other suitable container at the lowest possible temperature to prevent 6ver-.
heating and volatilization, and when sufficiently fluid,. shall be thoroughly
stirred to insure a uniform sample, whereupon 10 to 30 g. shall be taken for
analysis.
_
. Asphalt mastics or mastic cake shall be warmed on a hot plate or in a hot
oven until soft enough to be broken up or stirred, so that a representative sample
for analysis may be taken. The amount taken for analysis will depend upon
the amount of coarse gravel or stone in tlie mixture. The larger the gravel or
stone, the larger will be the sample required for accuracy. The size of samples
to be taken shall be as follows: Where all particles pass a No. 10 sieve, 10 to
30 g.; where 25% of the aggregate is retained on a No. 10 sieve, 50 g.; where
50% of the aggregate is retained on a No. 10 sieve, 100 g ; and where 75% of
the aggregate is retained on a No. 10 sieve, 200 g.
(a) FOT Analysis of 10- to jO-g. Samples.-In cases where a 10- to 30-g.
sample is sufficient, the analysis shall be carried out by means of the glass
extractor illustrated in Fig. 207.
An ordinary fat-free Whatman or S. and S. filter thimble, 60 mm. in length
by 26 mm. outside diameter, shall be dried for 30 minutes in an oven at 212 0 F.,
allowed to cool in a desiccator, and then weighed in a suitable weighing bottle.
The weighed sample shall be placed in the thimble and a disc of wool felt or
a plug of absorbent cotton shall be placed over the, sample in the top of the
thimble to distribute the solvent uniformly and prevent splashing. After
placing in the extractor, 40 to 50. ml. of carbon disulfide shall be poured over
the sample, whereupon the thimble containing the sample shall be suspended
under the condenser by a fine wire bail. The flask shall be cautiously heated by
a steam-bath or electric heater just enough to vaporize the solvent. Cold
water is circulated through the condenser. The heat evaporates the carbon
disulfide in the flask. This condenses upon the condenser and drops back upon
the sample through which it filters, thus dissolving out the bitumen which
collects in the bottom of the flask.
The extraction should be discontinued when the carbon disulfide drops
colorless from the filter. The time of extraction will dep(;jnd upon the nature of
the bitumen and mineral aggregate in the sample and upon the degree of heat
applied, the coldness of the water in the condenser and other factors. In
some cases extraction may be complete in one hour, in others four or five hours
may be necessary.
When the solvent comes through clear, the filter shall be removed and washed
with a fine jet of carbon disulfide from a washing bottle to wash out any bitumen

1568

EXAMINATION OF BITUMINOUS SUBSTANCES

that may be retained at the top of the p!l;per and to break up any channels
that may have been formed by the carbon disulfide passing through. If the
washings show any color, the thimble shall be put back and extraction continued until the solvent again becomes colorless. .Itlshall"then be removed,
dried carefully, at a low temperature at first to prevent 'ignition of the absorbed
carbon disulfide, and finally to constant weight at 100 C. (212 F.), cooled
and weighed.
'
.. ,
An aliquot portion of the solution in the flask shall be rinsed into a weighed
porcelain or silica evaporating dish or crucible and the solvent burned off under
a hood. The residue shall be ignited over a flame or in a muffle and the ash
weighed and the weight added to tl:).at of the mineral matter in the filter paper.
This is to correct for the fine mineral matter which will be carried through the
paper by the sol:vent. Should there be a consid'erable amount of ash recovered
in this way, and if it is found that the
mineral matter is calcium or other carbonate, it shall be recarbonated by repeated
treatment with ammonium carbonate solution and finally ignited at a dull red heat.
Sec.tionD-D
Ordinarily, however, the mineral matter
Sec.tionC-C.
going through the paper will be so small in
amount that the difference caused by ignition may be neglected. The corrected loss
in weight on the original sample represents
the percentage of asphalt present.
(b) For Analysis of 50- to 500-g. SamSection B-~.
1<.. -. 6~-~
ples.-The apparatus for analysis of samples containing coarse aggregate shall be .the
large extractor shown in Fig. 214, consisting
of a large brass cylinder, through the bottom of which projects a Hk.candle-power
incandescent carbon filament bulb to supply
heat to the extra!ction apparatus proper,
which is held in the upper portion of the
FIG. 214.-Extraction Apparatus.
cylinder. This apparatus is composed of a
cylindrical brass vessel for holding the solvent, a cylindrical wire basket made of 80-mesh wire cloth suspended in the
cylinder, and an inverted 'conical condenser which serves as a top.
A large filter paper, 12 or 13 in. in diameter, shall be fitted inSIde the wire
basket of the extractor by folding once more than in ordinary filtering, or by
wrapping it over a form which fits inside the basket (a cylindrical bottle of
proper size makes a good form) and placing it inside the basket.
The basket with contained filter paper shall bel dried and weighed. The
sample shall be weighed and packed in the filter paper in the basket. Care
should be taken not to pack all coarse particles in one place and th~ fine particles
in another, Qut to have them mixed together in uniform proportions.
The sample shall be covered with a disk of felt or wad of absorbent cotton
to insure even distribution of the dropping solvent,_thus preventing it from
forming a channel through the sample. The baske~shall be suspended in the
extractor and 150 to 200 ml. of carbon disulfide poured over the felt or cotton.

@)

tV

EXAMINATION OF BITUMINOUS SUBSTANCES

1569

The condenser shall be placed over the top and water circulated through it.
Current shall be started through the electric lamp underneath the extractor
and the extraction carried on exactly as in the smaller glass extractor, but on a
larger scale. The time for extraction will vary from 3 to 12 or more hours,
depending upon the nature of the sample.
To determine when extraction is complete, the condenser shall be raised
and the basket lifted out to observe if drippings are clear. One or two drops
caught upon white filter paper should leave but a light stain.
The drying and weighing of the basket, burning off of the solution for CCl'recti on, and calculation of the weight of mineral matter shall be determined 8S
in the foregoing. Note that any water in chemical combination with the mineral constituents is ,not ascertained by this procedure, but the amount present
is usually so small as to be negligible.
Cold Extraction Method.-Weigh out 1000 g. of the well-disintegrated
sample into an 800 m!. beaker and add 500 m!. carbon disulfide. Stir five or
six times during the first 24 hours thoroughly to break up the aggregate, and
then let stand quietly for 48 hours, after which the solution shall be carefully
decanted and filtered through a series of three 100 ml. funnels, using absorbent
cotton as the filtering medium. The aggregate left in the beaker shall be subjected to a second washing, conducted as described. The extracts shall be
combined and placed in a quart jar, which shall be tightly covered and allowed
to stand quietly for 48 to 72 hours, after which the solution shall be filtered
through an asbestos mat in a
Gooch crucible to remove the last
traces of sediment. The mineral
ingredients shall then be dried and
weighed as 'previously described.
In the case of mastic block and
similar bituminized'aggregates containing a high percentage of finelydivided mineral matter, there is apt
to be a loss of fine material, caused
by carbon disulfide solution creeping behind the folds of the filter
paper and over its edge, rather than
through the paper. In the case of
mastic the following procedure
is recommended: Two 32-cm. No.
5 Whatman filter papers, after FIG. 215.-Method of Folding Filter-papers
for Extraction Test.
being dried in an oven and cooled
in a desiccator, are counterpoised.
These papers are folded together and fastened with a paper-clip in the
manner shown in Fig. 215, .whereupon they are placed in a glass funnel
7-in. in diameter, the top edge of which is ground flat. The funnel is covered
with an 8-in. diameter glass plate, in the center of whioh a %-in. diameter hole
has been drilled. To make certain that the plate completely closes the funnel,
it is advisable to secure the plate to the funnel with gelatine cement, prepared
by dissolving 10 g. of gelatine in 80 m!. water and adding 20 g. glycerin. The
mastic block is warmed in an oven and broken into small pieces. Two 75 g.

1570

EXAMINATION OF BiTUMINOUS SUBSTANCES

represent~tive samples should be taken alld run in- duplicate. One sample is
introduced into the counterpoised filter and- the fUllIl;el covered with the glass
plate as described. Carbon disulfide is introduced directly into the funnel.
through a 1500 ml. separatory funnel as shown in Fip. 216. The addition of
carbon disulfide may be ~ade automatic, so that
when the level of the liquiddn the filter falls below a
predetermined height, a fresl\. supply will be admitted,
and this may continue overnight without attention.
When the filtrate comes through colorless, the carbon
disulfide supply is stopped and the paper allowed
to dry, first in air, and finally in an oven at 1200 C.
for one hour. After cooling in a ~esiccator, the filter
and contents are weighed, the outside pap~r being
placed on the weight-pan as a counterpoise. The
weight of asphalt extracted is ascertained by subtracting the weight of the aggregate from the weight
of the sample taken, and its percentage calculated.
For products other than mastics, the following
weights of material are recommended: surfacings
containing aggregate all passing 10-mesh: two 50 g.
samples; surfacings containing up to 25% U-in. or
%-in. aggregate: two 150 g. samples; surfacings containing up to 25% %-in. aggregate: two 300 g.
samples; surfacings containing up to 50% %-in.
aggregate: two 500 g. samples.
Another variation consists in covering the
coarsely broken sample with carbon disulfide. and
allowing same to stand quietly overnight, whereupon
it will be found that most of the soluble material
has gone into solution. The residue is then w.ashed
with 8 to 10 portions o carbon disulfide. Fine
mineral matter is removea. from the combined extracts by filtering or 'centrifuging.
.
Centrifugal Extraction Method.-The most efficient apparatus of this type is illustrated in Figs.
217 and 218. It con'sists of a 1/5 h.p. vertical motor
, a, capable of making 1100 revolutio_ns per minute at
FI~ofl~xt!~b~~a*:s[~r 110 volts, with either direct or alternating current.
Its shaft projects into a cylindrical copper vessel b,
having a concave bottom and draining into the spout- c. A circular brass plate
d, 9Yz in. in diameter supports an inverted iron bowl e, 8Yz in. in diameter by
2 156 in. high, having a 2-in. circular hole at the top. A brass cup f is fastened to
the inner side of the bowl, -having a circle of Ys i~. holes for the 'admission of
solvent, and terminating in a hollow axle which- fits snugly through a hole in
the center 9f the brass plate d. A felt ring g, % in. wide and about 0.090 in.
thick (cut from No. 80 roofing felt) is firmly pressed against the bowl by the
milled nut h for which the hollow axle is suitably threadecr. The axle in turn
fits snugly over the shaft of the motor, to wh~ch:tt -is secured by a slot and
cross-pin.

EXAMINATION OF BITUMINOUS SUBSTANCES

.

1571

Procedure.-The filter ring shall be dried for 30 minutes in an oven at 212 F.,
allowed to cool in a desiccator and then weighed.
The required quantity (10 or 25 g.) of the disintegrated sample shall be
accurately weighed 'on a tared watch glass or weighing scoop. The sample shall
then be placed in the bowl of the rotarex machine, care being taken to evenly
distribute the mixture around the periphery of the bowl to prevent vibration
during the extraction. The filter ring and bowl plate s~aH then be placed on
0

FIG. 217.-Centrifugal Extractor.

FIG. 218.-Centrifugal Extractor.

the bowl and fastened down fairly tight. The bowl shall then be placed in
position in the rotarex machin~ and from 40 to 50 ml. of carbon disulfide po_ured
into the bowl through the funnel opening. After allowing the material to
digest for about 15 minutes, the motor shall be started, slowly at first, and then
the speed increas,ed sufficiently to cause the dissolved bitumen to flow from the
spout in a thin, stream. The dissolved bitumen or filtrate shall be collected in
a beaker or other receptacle. When the first charge has drained, the motor
shall be stopped and a fresh charge of carbon disulfide added and again allowed
to digest for 15 minutes and the machine started as before. This operation
shall be repeated from three to five times or until the solvent runs clear from
the spout showing all bitumen has been dissolved from the sample, this usually
requiring three to four washings.
When the last addition of solvent has drained off, the bowl shall be removed
and placed with the bowl plate and filter ring on a sheet of glazed manila paper
and allowed to dry at room temperature. The aggregate shall then be brushed
,from the bowl onto the paper together with any aggregate adhering to the bowl
plate,and filter ring. The aggregate shall then be transferred to a tared watch
glass or pan and together with the filter ring dried in an oven at 212 0 F.,
'
cooled in a desiccator and weighed.
In making the correction for the mineral" matter suspended in the filtrate,
add sufficient carbon disulfide to bring the same to a definite volume and a
multiple of 100 ml. The whole filtrate shall then be thoroughly shaken so that
the mineral matter is uniformly dispersed throughout the filtrate and an aliquot

1572

EXAMINATION OF BITUMINOUE\ SUBSTANCES

part of 100 ml. then taken and transferred into a weighed porcelain or silica
evaporating dish and the solvent burned off under a hood. The' residue shall
then be ignited over a gas flame or in a muffle until it ~s entirely free from carbon,
cooled and weighed and the total ash correction det~rmined as follows:
_
T o t a I as h -

Total
. filtrate X,Ash in gram~, recovered in
III ml.
aliquot part
AI'lquot part III
'I
m.

Should there be a considerable amount of ash recovered in this manner,

and if it is found that the mineral matter is calcium or other carbonate, it should
be recarbonated by repeated treatment with ammonium carbonate solution
and finally ignited at a dull red heat.
An alternate procedure consists in using a centrifugal tube in the following
manner: The apparatus to be used in this test shall be any suitable type of
centrifugal machine capable of being rotated rapidly (1800 r.p.m,) surmounted
by a head carrying an equal number of metal tube shields, slightly larger than
the glass tubes used, and provided with rubber cushions, the whole being encased in a metal shield. In place of glass tubes aluminum or bronze tubes may
be used.
Procedure.-The required quantity (IO'or 25 g.) of the disintegrated sample
selected at random throughout the sample shall be accurately weighed on a
tared watch glass or weighing scoop." After weighing, the sample shall be
transferred to a numbered and weighed tube. The tube shall be filled with a
definite quantity of carbon disulfide approximately three-fourths full and then
inserted in the metal tube shield of the centrifuge machine. The tubes shall
be so placed in the centrifuge machine as to have equal weights opposite each
other around the head of the centrifuge. After allowing the material in the
tube to digest for about 15 minutes, the centrifuge machine shall be started,
slowly at first, gradually bringing it up to full speed and whirling from 5 to 10
minutes. The machine shall then be stopped, the tupe removed and the solvent
and dissolved bitumen decanted into a numbered flask. The tube shall again
be filled with carbon disulfide, mixing the soli~ matter at the bottom of the tube
with the solvent by stirring, the material allowed to digest for 15 minutes,
whirled from 5 to 10 minutes, the solvent decanted and the bitumen dissolved
as before. This operation shall be repeated ,from three to five times or until
the solvent in the tube, after whirling from 5 to 10 minut~, is clear and colorless.
The tube shall now be removed from the machine and the solvent allowed to
evaporate from the aggregate at room temperature and the material then dried
in an oven at 212 0 F. to constant weight, cooled in a desiccator and weighed.
The solvent and dissolved bitumen in the flask shall be rinsed into a weighed
porcelain or silica evaporating dish and the solvent burned off under a hood.
The residue shall then be ignited over a gas fla,me or in a muffle until it if! I
entirely free from carbon, allowed to cool in a desiccator and the ash weighed
and its weight added to the weight of the mineral matter in the. tube. Should
there be a considerable amount of ash recovered in this manner, and if it is
found that the mineral matter is calcium or other carbo,nate, it shall be recarbonated by repeated treatment with ammonium carbonate solution and
finally ignited at a dull red heat. Ordinarily the mineral matter in the solution
is found to be so small that the difference caused by ignition may be neglected.

EXAMINATION OF BITUMINOUS

SU~STANCES

1573

PART III
EXAMINATION OF BITUMINIZED FABRICS
This caption includes the following groups of products, viz.:.
Q-Prepared roofings.
R-Composition shingles.
S- Deck and porch coverings.
T-Bituminized fabrics for constructing built-up roofs.
U-Bituminized fabrics for constructing waterproofing membranes.
V-Electrical insulating tape.
W-Waterproof papers for wrapping and packing.
X-Waterproof papers for insulating against heat or cold;.
'Y-Felt-base floor coverings (surfaced with linseed oil and pigment
composition) .
Z-Expansion joints for pavements.
These are constructed as shown in Table III, where the index a indicates that
asphaltic compositions have been used, and t signifies that coal tar (pitch) et al.,
have been used.
TABLE III

Paper

Burlap

Duck

Burlap Paper
and
and
Rag or
Light
Asbestos Cotton
Felt
Fabric
-..,--1----.--1---,- - - - - - a
a
a
a
a
Light
Cotton
Fabric

Rag-Felt

Asbestos
Felt

------1--------------- - - - - - - - - --Single Layered:

Saturated only.. WX WX U U
Coated only (one
or two sides). W
U U S
Saturated and
coated....... X
U U QS ... V

TUY TU T

QR

Laminated
(Bituminated) :

Layers saturated
only. . . . . . . .. ....
Layers saturated
and coated. . .
Layers unsaturated ........ W W
One layer unsaturated and
others saturated. . . . . . .. w w

Q QT

UZ U

Q
W W

......... ... ... ... ...... .... .... ... .... ...

W W

NOTE.-Letters in heavy type indicate the more important groups of products.

1574

EXAMINATION OF BITUMINOUS SUBSTANCES

f

The finished products falling in this cl~ss include sheet roofings, floor coverings, waterproof membranes, sheathing and -insulating papers, expansion
joints involving the use of woven or felted fabrics, ~lectrical insulating tap~,
and certain tYpes of wall board. As these are constructed in many different
ways, 'it will obviously be impracticable to describ~ in detail the analytical
methods applicable to each. The ones which follow have been devised spe~ifically for examining prepared roofings, but with these as a starting-point,
others may readily be evolved for testing floor coverings, waterproof membranes, sheathing and insulating papers, etc.
(A) PHYSICAL TESTS OF THE FINISHED PRODUCT

Weight per Unit Area

(a) For Saturated Felted and Woven Fabrics.---,-The width of each roll
taken shall be measured to the nearest 3\ in. The minimum and maximum
width so obtained shall be reported.
Each roll taken shall be weighed intact to the nearest Xt' lb. and the .minimum and maximum gross weight shall be reported.
Each roll sampled shall be stripped. AI} the wri1ppers and packing material
shall be weighed together to the nearest U lb. and the average weight thereof
per roll shall be reported. Each roll shall then be unwound, observing the
workmanship and finish while so doing~ The length and width of each roll
shall thereupon be measured to the nearest Xt' in. and the square feet of material
contained in each roll calculated.
The rolls shall be rewound, fastened with pieces of light string and then
each roll shall be weighed to the nearest Xt' lb. In the case of felted fabrics,
calculate the weight of each roll in pounds per 100 sq. ft. and report the maximum and minimum, likewise recording the average for the rolls sampled, which
shall be regarded_ as the corresponding weights of the entire shipment. In the
case of woven fabrics, calculate the weight of each roll)in ounces per square yard,
and report the minimum, likewise recording tlfe average for the rolls sampled,
which shall be regarded as the corresponding weights of the entire shipment.
From the rolls examined, select the one who,Be weight per unit area is nearest
the average weight per unit. area of the lot. The roll so selected shall be laid
flat, the first convolution or two carefully unwound, and with a knife and
straight edge the sheet shall be cleanly cut across i1t right aIlgles to the edges.
A section measuring 30 in. in the direction of the roll's length spall be removed.
If the material is surfaced with sand or other fine material" any detached
particles belonging to each 30-in. section shall be retained. The width of each
section shall be measured to the nearest lz in. 'EaQh section, together with any
detached surfacing, shall be weighed to the neare~t gram. The weight in
pounds per 100 sq. ft. or the weight in ounces per sq. yd.. shall be calculated
from the formulas:
Weight of 30-in. section in grams
. sec tIOn
' "m mc h es
Pounds per 100 sq. ft. = 1.0582 X W'Id thof 30-111.
I

Weight of 3O.:in. section in grams

Ounces per sq. yd. = 1.5238X I' dth 0 /"30'
t'"
h
-111. sec IOn In lllC es
W

EXAMINATION OF BITUMINOUS SUBSTANCES

1575

This must check within 1% of the average weight per unit area of shipment.
If the sample fails to do this, then additional samples shall be cut from the

same roll, until one is obtained which does, which sample shall be kept for
further examination.
NOTE.-As a referee method or in case any dispute arises regarding the properties
as may be ascertained from the particular sample selected, a 30-in. sample shall be taken
and examined separately from each roll sampled as above.

If the material is surfaced with sand or other finely comminuted material,

the surfacing shall be swept with a moderately stiff brush from all the 3D-in.
sections taken. All the material thus removed shall be caught and weighed'
to the nearest ounce. From the aggregate areas of all the sections taken, the
average weight of detached comminuted surfacing shall be calculated in pounds
per 100 sq. ft.
The 3D-in. section selected for examination shall be cut into sections for
further tests in the manner shown in Fig. 219.

r--f"T---'-rr---1T
r------, I I
I :
L------JGIGrl
, l
L
n
J liZ 1
,:

: ,------1 1 I '
;sI---Y3---L1..lGII
1- _____ _' 1 13
I
, ,'I I
r--F4---JGI GI-J
r------J 4 151

I!
I/~' bmml@l1lm
III!

1:

I:

I !
L ____ _,.J I I
I :
L_Y..? __ LL.L ____J....':(.

/2" ...... l>l

..., ........ 18"

6 """"*"'4 ~"*27~
JO' ... -.- -. __ '--"--'>1

k.............--.-... (s" .. -.-. _....;.1.:.-

k--

FIG. 219.-Sampling Bituminized Fabrics.

(b) For Smooth-roll and Mineral-Surfaced Roll-Roofing.-Each roll

selected shall be weighed to the nearest U lb. The wrappers, packing materia),
nails, and cement from each roll shall be weighed together to the nearest U lb.
and the average weight thereof per "square" shall be recorded. The quantity
and character of nails and cement shall be recorded.

1576

EXAMINATION OF BITUMINOUS, SUBSTANCES

Each roll shall be unwound and the workmanship and finish of the roofing
observed. Recor!i the weight of any loose surfacing that may fall off in so
doing.
,
The length and width of each roll shall be measured to the nearest >i-in.,
and the square feet of material contained in each roll shall be calculated. From
these values, the average area of the rolls examined ,shall be calculated.
From the above net weight, exclusive of loose surfacing, of roofing per 108
sq. ft. contained in each roll shall be calculated and the minimum weight per
108 sq. ft. of the lot.
The average weight per 108 sq. ft. for the rolls examined shall be calculated.
This shall be regarded as the average weight per 108 sq. ft. of the lot .... CIa)
From the rolls examined, the one whose weight per 108 sq. ft. is nearest the
average weight of the lot shall be selected. The first convolution or two shall
be unrolled, and with a knife and straight edge, the sheet shall be cut across
at right angles to the edges. A section measuring exactly 30 in. in the direction
of the roll's length shall then be removed. The width of this section shall be
measured to the nearest l2 in. and weighed, neglecting any loose s:urfacing.
The weight in pounds per 108 sq. ft. shall then be calculated from the following
formula:
= Weight of 30-in. section in ounces 324 Clb)
P oun d s per 108 sq. f t. W'd h f 30 .
t'"
h X . .
1 to-In. sec IOn m mc es
The weight so determined shall check within 1.5% of the average weight per
108 sq. ft. of the lot CIa). If this sample fails to check, then additional samples
shall be cut from the same roll, until one is obtained which does, which sample
shall be kept for further examination.
NOTE.-As a referee method, or in case any dispute arises regarding the properties as
may be ascertained from the particular sample selected, a 30-in. sample shall be taken and
examined separately from each roll sampled.
(c) For Mineral-surfaced Shingles.-Each bundle selected shall be weighed
to the nearest >i lb. The packing material shall be w~ighed to the nearest ~ lb.
and the weight recorded separately for each bundle. The average weight of
packing material per" square" shall be recorded.
The shingles in each bundle, "a," shall be' counted and the workmanship
and finish of the shingles shall be observed. ,Record the weight of any loose
surfacing that may fall off in so doing.
From the above the net weight of each bundle, exclUSIve of loose surfacing,
shall be calculated, "b," The dimensions of two representative shingles shall
be accurately measured and the net area of material in each shingle calculated
in square feet, "c."
The area in square feet of material contained in each bundle (aXc) shall
be calculated. The weight of roofing material per ,108 sq. ft. contained in each
bundle

(_b_ X 108) shall be calculated, and the minimum weight per 108

aXc
sq. ft. recorded. This shall be regarded as the minimum weight per 108 sq. ft.
of the lot.
The average weight Cd) per 108 sq. ft. for th~, bundles' examined shall be
calculated. This shall be regarded as the averagE! Feight 'per 108 sq. ft. of the
entire shipment. , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. CIa)

EXAMINATION OF BITUMINOUS SUBSTANCES

The average weight per shingle

(C1~)

1577

shall be calculated ......... (lb)

From the bundles selected, a definite number of representative shingles,

whose weight shall fall within 1.5% of the average weight ascertained in (Ia)
shall be taken for further examination. The number of shingles so selected
shall represent as closely as possible 6 sq. ft. of material, and shall be selected
as far as possible from different bundles.
NOTE.-As a referee method, or in case any dispute arises regarding the properties as
may be ascertained from the particular sample selected, then a similar sample shall be
taken and examined separately from each bundle of shingles sampled.

Thickness
The thickness of eacR specimen shall be measured at 10 equally spaced
points, I in. from a cut edge, with a micrometer gage having flat bearing surfaces at both contact points of not less than Vz in. in diameter and preferably
of tl).e spring-and-dial type. The readings for each roll sampled shall be averaged and the minimum, maximum and average thickness calculated to the
nearest 0.001 in.
Resistance to Heat
(a) For Asphalt-saturated Fabrics.-Ftom each specimen, 12 by 6-in. strips
shall be cut at K-1 and K-2 in Fig. 219., care being taken not to disturb any of
the detached surfacing. Each strip shall be weighed and suspended in the
center of an air-oven maintained at 221 F. (105 C.)5 F. (3 C.) with a thin
wire fastened through holes punctured near one edge.' The thermometer shall
be inserted in the oven to such a depth that its bulb will be in line with the
center of the specimens. The strips shall be kept in the oven for exactly five
hours, then cooled and removed carefully, and each specimen weighed. The
average loss shall be calculated as a percentage and the average percentage of
moisture deducted. The final figure shall represent the average loss on heating,
exclusive of moisture.
(b) For Coal-tar Saturated Felt Only.-From the rolls making up the sample, strips 6 in. wide shall be cut across the sheet. Sufficient of these shall be
taken to make up a sample of 250 g.5.0 g. These strips then shall be rolled ,
up and placed in the wire basket of the extraction apparatus described on p.1568.
The sample shall then be covered with a disk of soft filter paper to insure an even
distribution of the solvent.
The carbon disulfide extract containing the bitumen shall be transferred to
a 500-mI., short-neck; round-bottom flask. The flask shall be equipped with
a Hempel column, 16 cm. in length, 15 mm. in inside diameter and filled to a
depth of 5 cm. with glass beads. The Hempel column shall be connected with
a water-cooled condenser. A -2 to +80 C. thermometer shall be placed in
the Hempel column so that the top of the bulb shall be opposite the bottom of
the tubulature. Distillation for the removal of carbon disulfide shall be continued by means of a steam bath until drops cease to fall from the condenser.
The distillation then shall be carried out over a small flame, heating carefully
and observing the thermometer until the distillation ceases and the temperature
of the vapor as indicated by the thermometer shows no further rise on slightly
increasing the flame.

1578

EXAMINATION OF BITUMINOUS SUBSTANCES

I

When the contents of the flask are sufficiently cool to be poured without
appreciable vaporizing, 100 g.O.l g. shall be distilleli as on p.1533.
Fractions shall be taken at 210 0 C. and 235 0 C. and the distillation stopped
at that point. The per cent of distillate by weight at each teni"perature shall
be calculated on the basis of the weight of sample tar.!en for distill!ition.
(c) For Asphalt Roll-roofings and Shingles.-Two; strips shall be cut from
the sample of roofing or shingles, each approximately ,8 by 8 in. They shall be
weighed and suspended vertically in the same direction as the material would
be applied to the roof, in the center of an air-oven maintained at 1760 F. (80 0 C.)
5 F. (3 0 C.) with a thin wire fastened through holes punctured near one
edge. The internal dimensions of the oven shall not he less than 12 by 12 by
12 in. An electrically heated oven is recommended. The thermometer shall
be inserted in the oven to such a depth that its bulb will be in line with the
center of the specimens. The strips shall be maintained at the prescribed
temperature for exactly two hours, then cooled in a desiccator and each specimen weighed. The average loss shall be calculated as a percentage. Any
change in appearance of the specimen shall be recorded, such as flowing, sagging,
blistering or absorption of the coatings; also' sliding of granular surfacing.
(B)

SEPARATION OF FlNISHE;D PRODUCT INTO ITS

COMPONENT PARTS

SEPARATION OF BITUMINOUS MATTER, MINERAL MATTER AND FIBROUS

CONSTITUENTS

Moisture
From each specimen, 2-in. strips shall be cut, as shown at A-I and A-2 in
Fig. 219. The' 2-in. strips shall be cut into I-in. squares. About '50 g.,
selected at random, shall be accurately weighed and distilled with 100 ml. of a
coal-tar distillate of which 5 to 10% boils below ioO o (ij. (212 0 F.) and of which
at least 90% distills up to 180 0 C. (356 0 F.), in an apparatus of the form and
under thc"'conditions prescribed on p. 1553. The flask containing the solvent
and sample shall be heated in a paraffin bath at a temperature of'302 to 3380 F.
(150 to 170 0 C.) until no further water passes over. The distillate shall be
allowed to stand in the graduated receiver until the water separates, whereupon
the volume of water shall be measured and its weight calCUlated. From this,
the average percentage of moisture in the fabric, as received, shall be calculated.
Analysis of Saturated

F~brics

From each roll sampled, a 2-in. strip (within l2 in.) shall be cut across the
specimen, as shown at H in Fig. 219.
/
Each strip shall be extracted with carbon disulfide in an extractor shown in
Fig. 219, the .extraction being continued for several hours after the drippings
hav:e become- colorless. The desaturated fabric shall be removed and heated in
a ventilated oven at 225 0 F. (107.1 0 C.), cooled in a desiccator, any adhering
comminuted surfacing being brushed off and reta!9-oo,- and weighed as rapidly
as possible. The heating shall be repeated until the weight of the fabric re-

EXAMINATION OF BITUMINOUS SUBSTANCES

1579

mains constant, as determined by two consecutive weighings taken not less

than ten minutes apart, which shall show a further loss of not more than 0.1 %.
This will give the weight of the fabric in the moisture-free state. WhEjre a coa1tar pitch saturant has been used, the moisture-free weight of the desaturated
fabric shall be corrected for the carbonaceous matter retained mechanically in
its interstices, by means of the method described in the following paragraph.
The following colorimetric method shall be used to ascertain the carbonll::ceous matter derived from a coal-tar pitch saturant and retained by the desaturated fabric:
(a) About 15 g. of an unsaturated fabric of the same general character as
the one under examination, shall be macerated by boiling in water, disintegrat.:
ing with a rotary egg-beater and picking the" fibres apart with needles. The
fibres shall be filtered through fine cloth and dried at a temperature of 2250 F.
(107.1 0 C.) to constant weight. One gram of the fibres shall be accurately
weighed into a flask and diluted to exactly 100 m!. with distilled water at room
temperature. About 50 g. of glass beads shall be added and the contents of
the'flask shaken vigorously until the fibres are reduced to a homogeneollS pulp
in uniform suspension.
(b) A distilled coal tar, having approximately 10 to 25% of insoluble carbonaceous matter, shall be procured and the carbonaceous matter extracted
with benzol until it is free from soluble matter; then dried at a temperature of
225 0 F. (107.1 0 C.) to constant weight. One gram of the purified carbonaceous
matter shall be accurately weighed and diluted to exactly 100 m!. at room temperature with a starch solution of a consistency sufficient to carry' the carbonaceous matter into temporary suspension. (A 12.5% solution is recommended.)
(c) The liquid carrying the fibres, obtained as described in Paragraph (a),
shall be titrated with the suspension of caz:bonaceous matter, obtained as
described in Paragraph (b), and ~rom time to time a field prepared from a drop
of the well agitated mixture shall be examined under a microscope at 100 diameters magnification until the color exactly matches a field prepared from the
desaturated fabric under examination, when both are reviewed side by side under
parallel conditions. The end-point is fairly sharply defined. The burette
reading gives directly the percentage of carbonaceous matter adhering to the
moisture-free fabric under examination.
(d) The weight of moisture-free fabric shall be corrected by deducting the
weight of adhering carbonaceous matter.
The total comminuted surfacing plus any filler present in the bituminous
saturant represents the sum of the detached comminuted surfacing, the amount
brushed off from the surface of the desaturated fabric, plus the amount recovered upon evaporating and igniting the bituminous extract obtained as
above.
,
The weight of bituminous saturant represents the difference between the
weight of the original fabric, and the sum of the weights of the moisture-free
desaturated fabric, plus moisture, plus the total comminuted surfacing and
filler present.
Ash.-A representative sample shall be secured by cutting from each strip
of de saturated felt a piece about Yz in. in diameter as shown at 1-1, 1-2, and 1-3
(Fig. 219). About 25 g. selected at random from all the specimens sampled in

1580

EXAMINATION OF BITUMINOUS SUBSTANCES

this manner shall be accurately weighed, and incinerated i.p. a weighed porcelain
or quartz crucible either over an open flame or in a muffle, until all the carbon is
consumed. A few drops of ammonium carbonate solution shall be added, the
sample ignited gently, and weighed. The percentage pf ash shall be calculated'
on the basis of the moisture-free felt.
I
I

NOTE,-As a referee method, or in case greater accuracy is desired, the three portions
taken from each specimen roll shall be weighed and ignited separately. The minimum,
maximum and average of ash present shall be calculated on the basis of the moisture-free
felt.

Analysis of Saturated and Coated Fabrics

Since the fabrics and bituminous matter may be assembled in many different ways, the resultant products are too numerous to itemize. It is im- .
practical, therefore, to give analytical methods applicable to each. The ones
which follow have been devised specifically for examining prepared roofings
and composition shingles composed of roofing-felt, saturated and coated on both
sides with asphalt, and surfaced on the top with either powdered or granular
mineral matter, and on the under side with powdered mineral matter. These
methods are typical ones, and with slight modifications and a little ingenuity
may be adapted to other forms of bituminized fabrics.
Asphalt roll-roofings and shingles may be divided into two general types as
shown in Fig. 220.

a'lffI.

rnmm71TTmm71nr __

'....VPE
.~

~C b-

-Fine dusfinfj nnisn

--.Topbif~minOlJscoafing
--Btfumlnlzedfelt

-'r&Hom hifuminous coafin~'

--,eFine dusfing finiSh

(carse or moderafely coarse

/f!mbedded mineralmafk"
"tYpE ii-1B---roPbifu"!inouscoafing
V
i-<
---Btfummrzed felt
. b,.
:.-Bqflombl~umin_Ot;spoafing
--Finedusfmg fimsh

FIG,'220.-Types of Asphalt Roll-roofings and Shingles.

Type X represents a layer of felt, saturated and coated with-bituminous

matter, having the surface coatings dusted with fine mineral matter (e.g.,
talc), This type embraces what is commerciall~ known as "smooth-surfaced
roll-roofing."
Type Y is similar to Type X, but wit,h a top surfacing of either: (1) coarse
mineral matter (e,g., granular slate) embracing what is commercially known as
"mineral-surfaced roll-roofing" and" mineral-surfaced shingles"; or (2) moderately coarse mineral matter (e,g., granular talc) embracing what is commercially known as "roll-roofing surfaced with granular talc."
The mineral matter, bituminous matter and fibrous m'tter may be distributed in the following manner:

EXAMINATION OF BITUMINOUS SUBSTANCES

1581

A-MINERAL MATTER

1. Detached ..... , .............. '}Fine Mineral Matter (e.g., finely ground talc or

2.
3.
4.
5.
6.

mica), Type X (on top and bottom), also Type

Y (on bottom only).
Embedded in the top coating ..... Moderately Coarse Mineral Matter (e.g., coarsely
.
ground talc) Type Y (on top only).
Embedded in the bottom coating Coarse Mineral Malter (e.g., crushed slate,
crushed brick or tile), Type Y (on top only).
Admixed with the top coating}M
b
(Types X and Y). . . . . . . . . . . .
ay or may not e ~resent. If present, conAdmixed with the bottom coating
s~s.ts of. very fine mmeral matter (e.g., clay,
silIca, lImestone, slate flour, shale, etc.).
.
(Types X and Y). . . . . . . . . . . .
Ash present in desaturated felt (i.e., ash on incineration).
B-BITUMlNOUS MATTER

1. Contained in the top coating.

2. Contained in the bottom coating.

3. Saturant of the felt.
C-FIBROUS MATTER

1. The desaturated felt.

The test strips for the separation of prepared roofing and shingles into their
component parts shall be prepared as follows:
From the samples representing the average weight a number of strips shall
be cut, each measuring approximately 2-in. in width by 8 in. in length. These
shall be weighed, the length and width measured to within 1 mm. or lz in., and
the weight calculated per 108 sq. ft. Any of the 2 by 8-in. strips whose weight
'varies more than 1.5% from the average weight of the shipment, shall be
rejecteq, and the remaining strips shall be retained for further examination.
NOTE.-If a strip is cut to exactly 2 by 7.46 in. its weight in grams divided by two
gives directly the weight in pounds per 108 sq .. ft.

Detached Mineral Matter (A-l).-Remove the detached mineral particles

from both sides of the 2-in. strips with a jet of compressed air operating under
a nozzle pressure of 25 lb. per sq. in. and reweigh in grams. Make a correction
to include any loose mineral matter. Calculate the weight in pounds per 108
sq. ft. of the detached mineral matter (A .. l) . ....................... " (2)
Moisture-free Felt (C-l).-Eight of the 2 in. by 8-in. strips shall be extracted with benzol (boiling-point 80 to 82 C.), or C.P. carbon disulfide in a
suitable extractor or centrifuge, and the extraction continued until the drippings
have become colorless. The desaturated felt shall be removed from the extractor, being careful not to lose any adhering mineral matter, placed in a tared
weighing bottle provided with a ground-glass stopper, heated in a ventilated
oven at 225 F. (107 C.)5 F. (3 C.) for thirty minutes~ the stopper inserted, then cooled and weighed. The heating shall be repeated until the
weight of the felt remains constant, as determined by two consecutive weighings
taken not less than ten minutes apart, which shall show a further loss of not
more than 0.1 %. Any adhering mineral matter shall be brushed off the felt,
weighed and retained for further examination. The weight of said adhering
mineral matter shall be deducted to obtain the weight of the moisture-free felt.
The extract and accompanying mineral matter shall be retained for further

1582

EXAMINATION OF BITUMINOUS .SUBSTANCES

examination. From the weights so obtained and the respective areas of the
strips of desaturated felt" the weight of moisture-free f~lt shall be computed in
pounds per 108 sq. ft., from the following formulas and the results averaged:" (3)
.

d
108
it
oun s per
sq..
,

flit

Weight of moisture-free
in grams
2212
A
f h
. .
.
X.
rea 0 t e stnps m square centIm~ters

or
Pounds per 108 sq. ft.

Weight of moisture-free felt in grams X34 3

.
Area of the strips in square inches

The extracted felt shall be retained for examination as to weight, thickness,

strength, etc.
Ash in Moisture-free Felt (A-6).-The strips of desaturated felt obtained
as described shall be cut into .squares about 2 cm. on each side and thoroughly
mixed. Pieces amounting to about 10 g. shall be selected at random and dried
at 225 0 F. (107 0 C.) for thirty minutes. They shall be cooled in a desiccator,
accurately weighed and then incinerated in a weighed crucible either over an
open flame or in a muffle until all the carbon has been consumed. A few drops
of ammonium carbonate solution shall be added, the sample ignited gently,
cooled and reweighed. The percentage of ash. shall be calculated on the basis
of the moisture-free felt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. (4)
Bituminous Saturation (B-3) in Moisture-free Felt.-From the 2 by 8-in.
strips, cut 40 strips about %: by 2 in. and weigh, this being a convenient size
to handle. These strips shall then be separated into three horizontal sections.
The small arrows a and b, to the left of the various types of roofing iUustrated in
Fig. 220., indicate approximately where the layers should be separated. The
top bituminous coating, with the embedded mineral matter, shall be removed
first.
NOTE.-The purpose shall be to remove the entire coatings with some of the saturated felt adhering and no coatings left on the central section of saturated felt. The
operator may accomplish this in any way that is most conv,enient. Care shall be taken
to preserve the mineral surfacing with the layers to' whicH it belongs if it becomes detached during the operation. A convenient manner to effect the separation consists
in warming the strips in an air-oven at a temperature of approximately 1760 F. (80 0 C.);
then with a knife, the front and back coatings, respectively, shall be peeled off, care being
taken to remove as little as possible of the saturated felt, and at the same time to make
sure that all of the coatings and surfacings are stripped from the felt.

These three horizont~l sections thus obtained shall b~ \veighe.d separately

and the total weight should agree within 1% with the original weight of the
strips which have been separated. The saturated felt (middle section) shall
be extracted with benzol (boiling point 80 0 to 820 C.) or c.p. carbon disulfide,
in a suitable extractor or centrifuge and the extraction continued until the
washings have become colorless. The desaturated felt shall be removed from
the extractor, placed in a tared weighing bottle p~ovided with a ground-glass
stopper, heated in a ventilated oven at 225 F.5 F. (107 C.3 C.). for
30 minutes, the stopper inserted, then cooled and weighed. The heating shall
be repeated~ntil the weight of the felt remains constant as,determined by two
consecutive weighings taken with an intervening.:,...heating period of not less
than 10 minutes, which shall show a further lo~ in weight of not more than
0.1 %. Calculate the weight of bituminous saturation by difference. The

EXAMINATION OF BITUMINOUS SUBSTANCES

1583

saturant may be recovered by combining the extract and washings in a 200-ml.

distilling flask, provided with a dropping funnel to add the solution during
distillation, to which is attached a water-cooled condenser and a receiver, the
latter having a vented connection to a vacuum pump. The distilling bulb is
immersed in a bath having a temperature of about' 212 0 F. (100 0 C.) if the
solvent were carbon disulfide and a temperature of about 275 0 F. (135 0 C.) if
benzol or chloroform were used as the solvent. Distillation shall be continued
at this temperature at atmospheric pressure until no more solvent is evolved,
after which vacuum shall be applied very gradually until a vacuum of 0.5 in.
mercury pressure is attained. Great care shall be exercised in applying the
vacuum gradually to prevent foaming. This vacuum shall be continued for
one hour after the specified vacuum has been reached. The distilling flask
shall then be disconnected and the residue represents the felt saturant and shall
be weighed.
.
NOTE.-Use the residue of bituminous saturation recovered from the felt for
'
examining its physical and chemical characteristics.

Calculate the per cent of bituminous saturation (B-3) carried by the dry
felt .............................................................. (5)
Calculate the weight in pounds yer 108 sq. ft. of bituminous saturation
(B-3) present in the dry felt, i.e. (5) X (3) . . . . . . . . . . . . . . . . . . . . . . . . . . .. (6)
Bituminous Matter in the Top and Bottom Coatings Respectively (B-1 and
B-2).-Take six of the 2 in. by 8 in. strips, from which the detached mineral
matter has been removed, and split them lengthwise by tearing the felt midway
between the points a and b. Weigh and extract the top and bottom sections,
respectively, in a suitable extractor or centrifuge. Desiccate and weigh the
dry felt in each section. Separate the total embedded and admixed mineral
matter recovered from each section, dry at 225 0 F. (170 0 C.), cool in a desiccator
and weigh. Calculate the weight of bituminous saturant present, i.e., weight
of felt recovered on extraction X (5). From the original weight of the top and
bottom sections, respectively, subtract the combined weights of dry felt,
bituminous saturant, embedded and admixed mineral matter. The difference
represents the weight of bituminous matter in the surface-coating carried by
tl.J.at particular section.
Calculate the weight in pounds per 108 sq. ft. of bituminous matter (B-1)
in the top coating .................................. '" ........... (7a)
Calculate the weight in pounds per 108 sq. ft. of bituminous matter (B-2)
in the bottom coating. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. (7b)
Calculate the weight in: 'pounds per lOS sq. ft. of total mineral matter embedded and admixed with the top coating ............. .' ............ , (Sa)
Calculate the weight in pounds per 108 sq. ft. of total mineral matter embedded and admixed with the bottom coating ....................... (8b)
To separate the bituminous constituents in the coatings for further examination and also to check the percentage of mineral matter, both detached and
embedded in the top and bottom coatings, with that determined above, the
detached mineral matter shall be brushed off from six weighed 2 by 8 in. strips.
The strips shall again be weighed and the percentage calculated based on the
original weight of strips. The outer portion of the top and bottom surface
coatings, respectively, shall then be scraped off after the strips.have been heated

1584

EXAMINATION OF BITUMINOUS SUBSTANCES

in an oven at 212 0 F. (1000 C.) for 30 minutes. The coatings shall be scraped
off by holding a dull knife at right angles to the strip of roofing supported on a
firm level surface and drawing the blade sideways under moderate pressure.
Care shall be taken to avoid scraping entirely throu~h the surface coating.
This is important. The scrapings shall then be weighed 'and dissolved in benzol
, (boiling point 80 to 82 0 C.) or c.p. carbon disulfide; and the mineral matter
separated by filtering or centrifuging and washing with successive portions of
solvent. The mineral matter shall then be dried at 2250 F. (107 0 C.), cooled in
a desiccator and weighed. Calculate the percentage of mineral matter l::J{tsed
on the original weight of roofing strips taken, which should check with that
obtained previously. The bituminous matter in the scrapings shall be calculated by difference and reduced to a percentage basis on the original roofing.
strips used. The bituminous coating shall be recovered by evaporating off the
solvent as described. Both surface coatings should be treated separately in
this manner.
NOTE.-Use the separated bituminous coatings Jor examining their physical and
chemical characteristics.

Coarse or Moderately Coarse Embedded ,Mineral Matter (A-2), also Fine

Admixed Mineral Matter'(A-4) on the Top Coating of Type Y.-The mineral
matter recovered from the top sections of Type Y shall be boiled with 100.ml.
water and washed through a 55-mesh copper wire screen with sufficient water
to remove the fine mineral matter. The total aqueous suspension of mineral
matter which passes through the screen shall then be recovered by filtering
the washings through a weighed Gooch crucible, then dried at 225 0 F. (107 C.)
and reweighed. This represents the fine mineral matter admixed with the top
coating. The mineral constituents retained on the 55-mesh screen represent
. the coarse or moderately coarse mineral matter embedded in the top coating.
This shall likewise be dried and weighed.
Calculate the weight in pounds per 108 sq. ft. of coarse or moderately coarse
mineral matter (A-2) embedded in the top coating of/Type Y ......... ;. (9)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter (A-4)
admixed with the top coating of Type Y (*) ......................... (10)
Fine Embedded and Admixed Mineral Matter (A-2 and A-4) in the Top
Coating of Type X j Likewise Fine Embedded 'and Admixed Mineral Matter
(A -3 and A-4) in the Bottom Coating of Types X and y.-rrake five of the 2-in.
by 8-in. strips, from which the detached mineral matter has been washed off with
water, and scrape off the outer layer of the top and bottom coatings, respectively,
by means of moderately rough sandpaper or a knife blade. Enough of the
surface shall be scraped to remove every vestige of the fine embedded mineral
matter, and at the same time care should be taken not to cut completely through
the surface coatings into the saturated felt underneath. Then scrape off about
5.0 g. of the top and bottom surface coatings, respectively, with a sharp knife
as described, taking care not to scrape entirely through the surface coatings.
Ignite the 5.0-g. portions separately in a tared crucible until all the carbon has
been consumed, add a few drops of ammonium carbonate solution, ignite to a
. _, ~
.
dull red heat and weigh.

* Up to 2% admixed mineral

matter probably adventitious and not filler.

EXAMINATION OF BITUMINOUS SUBSTANCES

1585

Calculate the percentage by weight of .fine mineral matter admixed with the
top coating (Type X) ............................................ (l1a)
Calculate the percentage by weight of fine mineral matter admixed with the
bottom coating (Types X and Y) ................................. (lIb)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter admixed
with the top coating (Type X) i.e. (7 a) X (lla) .................... " (12a)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter admixed
with the bottom coating (Types X and Y) i.e. (7b) X (lIb) . ........... (12b)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter embedded in the top coating (Type X) i.e. (8a) - (12a) . . . . . . . . . . . . . . . .. (13a)
Calculate the weight in pounds per 108 sq. ft. of fine mineral matter embedded in the bottom coating (Types X and Y) i.e. (8b) -(12b) . ...... (13b)
Compilation of Results.-The following items shall be reported in pounds
per 108 sq. ft ..:
Type X

TypeY

Minera~

Matter:
Fine mineral matter embedded in top coating ........... Item (13a)
Coarse or moderately coarse mineral matter embedded in
top coating ....................................... ... .......
Fine mineral matter embedded in bottom coating ....... Item (13b)
Fine mineral matter admixed with top coating .......... Item (12a)
Fine mineral matter admixed with bottom coatin~ ...... Item (12b)
Bituminous Matter:
Contained in top coating .......... : ..... : ............ Item (7a)
Contained in bottom coating ............. , ........... Item (7b)
Saturant of the felt ................................. . Item (6)
Fibrous Matter:
Felt after desaturation ............................... Item (3)
Net weight of roofing material. .....................
Sum
Check total ..... : ................................ (Item 1)
Detached mineral matter .......... : ................. Item (2)
Packing material, nails and cement .....................
Weight
Gross weight per 108 sq. ft .........................
Sum
Per Cent Ash:
From desaturated felt. , ............................. Item (4)

.0.0

Item (9)
Item, (13b)
Item (10)
Item (12b)
Item (7a)
Item (7b)
Item (6)
Item (3)
Sum
Item (1)
Item (2)
Weight
Sum

Item (4)

If desired, the results may also be expressed in

% by weight.

1586

EXAMINATION OF BITUMINOUS, SUBSTANCES

PART IV
I

EXAMINATION OF BITUMINOUS-SOLVENT COMPOSITIONS

I

(A)

PHYSICAL TESTS OF FINISHED PRODUCT

The following tests have been proposed, concerning which the particulars
will be found elsewhere.
Specific Gravity.
Viscosity.
Plasticity and Mobility.
Flash-point.
Spreading Capacity and Workability.
Draining Test.
Time of Drying.
Hiding Power.
Color.
Gloss.
Hardness, Abrasion and Adhesion.
Water Absorption.
Resistance to Heat.<
Resistance to Oil.
Resistance to Acids and Alkalies.
Dielectric Strength.
(B)

SEPARATION OF FINISHED PRODUCT INTO ITS

COMPONENT PARTS
I

Estimation and Recovery of the Solvent

Steam Di~tillation Method.-This has been standardized as follows: The

bituminous mixture is distilled in a current of steam, the solvent is condensed
and separated from the water. The steam-g~nerator shall be made of either
metal or glass, with a capacity of from 2 to 4 liters, suitable for continued use in
the production of steam: If of glass, it shall be fitted ~ith two outlets with
suitable connections for rubber tubing. In the case of a metal generator, a
large opening for filling and a water gage shall be additions:! parts of the
apparatus. The generator shall be supplied with suitable pinch cocks or valves
so that steam may be blown off to the atmosphere until the test is ready. The
bath shall be of metal of sufficient capacity to permit immersion of the distilling
flask to a depth of not less than 10 cm. Heat for :the steam generator shall be
supplied by a suitable gas generator or electric hot plate. The bath may be
heated by any convenient means.
The distilling flask shall be a short ring-neck, round-bottom flask of lOOO-ml.
capacity. It shall be fitted with a three-hole n~bber stopper; with a steam
distilling tube which will reach to within % in.-(t2:J mm.) of the bottom of the
fl:1sk and project from the top at a convenient distance for connection to the

EXAMINATION OF BITUMINOUS SUBSTANCES

1587'

generator; a vapor outlet tube which extends from beneath the rubber stopper
to a point sufficiently above the distilling flask that will permit convenient
connection to the condenser; and with a thermometer. The steam tubing
should be not less than 2 nor more than 4 mm. in internal diameter and the
vapor outlet tube should be not less than 5 mm. in internal diameter.
The condenser shall consist of a 196 -in. (14.29 mm.) outside diameter No. 20
Stubbs' Gauge seamless brass tube, 22 in. (55.88 cm.) long. It shall be set
at an angle of 750 from the perpendicular and be surrounded with a cooling
bath 15 in. (38.1 cm.) long, approximately 4 in. (10.16 cm.) wide by 6 in. (15.24
cm.) high. The lower end of the condenser tube shall be cu.t off at an acute
angle, and curved downward for a length of 3 in. (7.62 cm.) 'and slightly backward so as to insure contact with the wall of the graduate at a point 1 to lU in.
(2.54 to 3.18 cm.) belowthe top of the graduate when it is in position to receive
the distillate.
A separatory funnel having a capacity of not less than 500 m!. should be
provided. Accessories consist of suitable ring stands for supporting the steam
generator, distilling flask, bath for distilling flask, separatory funnel, and a
thermometer.

Courtesy A.S.T.M.

FIG. 221.-Assembly of Distillation Apparatus.

The apparatus shall be assembled as shown in Fig. 221. The steamgenerator shall be filled with water and heat applied. The bath shall be filled
with a high-flash-point oil and raised to approximately 140 C. (284 F.)
Five hundred milliliters of the sample shall be weighed into the round-bottom
flask. The connection shall be made from the steam-generator to the steam
delivery tube, the end of which shall be within Y2 in. of the bottom of the distilling flask. The outlet from the distilling flask shall be connected to the
condenser and the separatory funnel placed in position at the outl~t of the
condenser to receive the distillate. The end of the bulb of the thermometer in
the steam-distilling flask shall be placed within Y2 in. (12.7 mm.) of the bottom
of the djstilling flask.
When the temperature of the sample in the distilling flask reaches 130 0 C.
the outlet of the steam generator shall be closed, thus forcing the steam to pass
through the sample. The flow of steam shall be adjusted so that the distillate
'is collected at the rate of approximately 6 to 10 m!. per minute. The distillation
shall be stopped when 100 m!. of the distillate contains not more than 0.5 ml.
of solvent, as determined by measuring the amount of oil in 100 m!. of distillate.

1588

EXAMINATION OF BITUMINOUS f3UBSTANCES

When the distillation is finished, the water shall be separated from the distiJlate
and the distillate measured and retained for further tests, if required b;} the
specifications.
,
In some cases, the distillate does not separate read~ly from the water, and
this separation can be facilitated by tl;le addition of sodium chloride, which will
result in a sufficient difference in gravity to produce a !clear separation of the'
two layers.
The results shall be reported in per cent by weight or volume as required
by the specifications, based on the weight of the sample taken.
'
NOTE.-In the distillation of varnishes the end point differs with the various types
and compositions. In all cases, however, the rate of distillation slows down as the end
point is approached; and, when the end point is reached, acrolein (formed in the decomposition of the varnish base) is evolved and may be detected by its characteristic odor.
In addition to the evolution of acrolein, the following phenomena are suitable indi- .
.
cators of end point:
(1) When the end point of the solvent is considerably lower than the initial boiling
point of the base, the end point of the distillation is indicated by the temperature reaching
a maximum and then starting to fall consistently.
(2) When the end point of the solvent and the initial boiling point of the base are too
close, or overlap, the above temperature drop does not occur. The end point of the
distillation is then indicated by a sudden foaming up of the residual base in the distillation
flask; or, if the distillation has been accompanied by foaming, the end point is indicated
by a breaking of surface tension with a consequent cessation of foaming.

The specific gravity of the 'Volatile matter (solvent) recovered by distillation

shall be determined in accordance with the procedure described below.
The difference between 100% and the percentage of non-volatile matter
determined above shall be calculated and recorded as the percentage of volatile
matter by weight. ,
,
.
The percentage of volatile matter by volume shall be calculated from the
formula:

AC

X=-

where
x=the percentage of volatile matter by volume;
A =the specific gravity of the original varnis4;
B = the specific gravity of the volatile matter (solv~nt) recovered by.
distillation,
C = the percentage of volatile matter by weight.
With proper care and attention to detail in making this test, differences
occurring between different laboratories should' not exceed 2% for varnishes
containing solvents which have end points not over 235 C. (455 F.), and
s~ould not exceed 4% when the end point of the so~~ent is as high as 307.2 C.
(585 F.).
To correct the loss of distillate for water-soluble constituents, weigh out
another 100 g. into a 250-ml. flask and distil without steam over an electric
stove. Continue the distillation until the residue ip. the flask reaches a temperature of 200 C. This gives somewhat lowel: resnlts than the first method,
but the distillate should be tested for water-sohible substances to correct the

EXAMINATIGN OF BITUMINOUS SUBSTANCES

1589

results obtained by the previous method. Turpentine dissolves to the extent

of 0.3 g. for each 100 ml. of water condensed.
.
If the residue is to be used for further exarrtination, any water introduced
into it by the steam distillation may be removed by distilling it twice with ten
times weight of a mixture of 3 parts benzol and 1 part alcohol.

(C)

ESTIMATION, RECOVERY AND EXAMINATION OF

PIGMENT AND FILLER

Dilute 100 g. of the well-mixed material with 500 m!. of benzol in an SOO-ml.
stoppered flask. Either centrifuge or let stand in a warm place until the pigment or filler has settled, then carefully decant the supernatant liquid into a
clean flask of large capacity. The pigment or filler is shaken up with 250 ml.
more benzol, allowed to stand in a warm place until it settles, and the, supernatant liquid decanted into the second flask. Repeat the treatment with
benzol until the vehicle has been completely extracted from the pigment. The
prevention of skin formation during this process may be attained by the addition of anti-oxidants such as phenol or hydrochinone (dissolved in ether),
which are subsequently expelled on heating the pigment and base. The combined extracts are allowed to stand quietly to recover any pigment that may
have been carried over with the benzol, and then carefully decanted through a
weighed Gooch crucible provided with an asbestos filter. The residues in the
flask and on the Gooch crucible are washed with benzol as before, and combined
with tM balance of pigment or filler; which is then dried at 110 C. and weighed.
The combined extracts are retained for further examination.
The pigments or filler recovered from the previous test are subjected to a
qualitative or quantitative analysis for purposes of identification likewise to
granularmetric tests.

(D)

ESTIMATION, RECOVERY AND EXAMINATION

OF THE BASE

If no pigments or fillers are present, the base is recovered as described, and

its percentage by weight ascertained directly. If pigments or fillers are present,
the combined extracts are distilled to a small bulk, transferred to a tared dish,
and evaporated in an oven at 110 C. exactly to the calculated weight of the
base, by subtracting the weights of solvent and pigment or filler from the original
weight of material taken for examination. When oxidizable substances are
present, the final evaporation should take place in an atmosphere 'of illuminating
gas.
.
The base will contain the bituminous constituents (with the exception of
any" free carbon" associated with coal-tar pitch, or the like, which will be
separated with the pigments and fillers), likewise animal and vegetable oils or
fats, resins, and m~tallic bases and dryers. It may be separated into its component parts as follows:

1590

EXAMINATION OF BITUMINOUS, SUBSTANCES

METHOD OF ANALYZING THE SEPARATED BASE

Dissolve 50 g. in 150 ml. benzol. Add 10 ml. dilute nitric acid (1 : 1j and boil under a reflux condenser
for one-half hour to decompose any metallic soaps (i.e., driers,.etc.). Add 150 ml. water, boil under reflux
condenser, transfer to a separatory funnel, draw off the aqueous layeri' boil with another 100 rul. water,
and repeat if necessary until all the metals are removed.'

I
I

Benzol Solution:
I I
Aqueous Extrac~;
Distil to 100 ml., add 300 ml. of the saponifying liquid *, boil under reflux condenser
Contains the
for one hour, and separate the unsaponilUi.ble and saponifiable constituents as described. metallic bases as
nitrates. Examine
qualitatively and
Saponifiable Matter:
Unsaponifiable Matter:
Aqueous Layer: then quantitativeIy for lead, manSeparate the fatty and resin
Examine a small portion. If
Determine per- ganese, cob al t,
higher alcohols are present, sepa- acids as described.
centageglycerol. zinc, calcium, and
rate the balance into:
Multiply this by magnesium.
10 to estimate
(N:B.-Tlie last
pet cent of vege- three used for hartable or animal dening rosin. The
Hydrocarbons:
Resin Acid.:
Higher Alcohols, Fatty Acid.:
oils or fats (triEtc.:
Contain the
Include acids
Include acids glycerides) pres- mettallic dryers
bituminous subContain cho- derived from derived from en t in the origi- should not ~e
found by ignitl n,
stances (i.e., as- lesterol, etc., de- vegetable and rosin and the nal substance.
since the lead will
phalt, coal-tar rived from wool animal oils or fossil resins.
be reduced to
pitch, unsap. grease, also the fats, also from
metal by the ormatter derived unsaponifiable fatty-acid pitch,
ganic matter, and
from fatty-acid consti tuen ts
volatilized.)
pit~h, etc.).
originally pres"
ent in resins (4
to 8%).

* The saponifying liquid shall consist of a 10% solution of KOH dissolved in equal
parts of 95% C2H 50H or 90% benzol.
/
.

EXAMINATION OF BITUMINOUS SUBSTANCES

1591

PART V
EXAMINATION OF BITUMINOUS DISPERSIONS
(A)

PHYSICA-L TESTS OF FINISHED PRODUCT

Settlement Test

The following apparatus will be required:

(a) Cylinders.-Two glass cylinders of 500-ml. capacity with pressed or '
molded glass 'base and cork or glass stoppers. The outside diameter shall be
5.0 cm.0.5 cm. and the cylinders shall be graduated at each 5-ml. interval
to the 500-m!. mark.
(b) Glass Pipette.-A syphon, glass tube pipette, 60-m!' capacity, form
optional.
A 500-m!. sample, :representative of the emulsion, shall be placed in each of
two glass cylinders. The cylinder shall be stoppered air-tight and stood aside
unmolested, at laboratory air temperature for 5 days. After standing for this
5-day period, approximately the first 55 ml. of emulsion shall be removed by
means of the pipette or syphon from the top of each cylinder without disturbing
the balance of their contents. Exactly 50 g. of each of the two samples, after
each has been thoroughly mixed separately, shall be weighed into separate
600-ml. low-form glass beakers and the asphaltic residue determined by
evaporation at 3250 F. (163 0 C.) fQr 3 h:r:. in the apparatus described.
After removal of the first sample, approximately the next 390 ml. shall be
syphoned off from each of the cylinders. The residue remaining in the cylinders
shall be mixed thoroughly and exactly 50 g. shall be weighed out from each of
them and the amount of asphaltic residue (all sediment, if any, included) shall
be determined by evaporation as described above for the two top samples.
The numerical difference between the .average percentage of asphaltic
residue from the two top samples and the average percentage found in the two
bottom samples shall be recorded.
Demulsibility Test
This test indicates the rapidity with which the dispersion "breaks" during
use, and may be ascertained in accordance with the following procedure:
The following apparatus and reagents will be reQ.uired:
(a) Sieves.-Three No. 14 sieves of the U. S. Standard Sieve Series, of iron
wire cloth, unframed, approximately 5-in. square.
(b) Beakers.-Three glass beakers of 600-m!. capacity each.
(c) Glass Rods.-Three glass rods, rounded ends, approximately 156 in. in
diameter.
(d) Burette.-A 50-ml. glass burette graduated in 0.1 ml.
(e) Calcium Chloride~olution (0.02 N).
(f) Calcium Chloride Solution (0.10 N).
The percentage of residue shall be determined by distillation as described
on p. 1593.

1592

EXAMINATION OF BITUMINOUS, SUBSTANCES

The weight of each assembly of beaker, rod and sieve shall be recorded.
Exactly 100 g. of the emulsifi~d asphalt shall be wJighed into each of three
600-ml. tared beakers. Over a period of approximately 2 min., 35 ml. of 0.02 N
CaCh solution (if quick-setting emulsion is being te~ted) or 50 m!. of 0;10 N
CaCh solution (if mixing type emulsion is being testeii) shall be added to each
beaker. from a burette. While adding the solution qf CaCh, the contents of
the beaker shall be stirred continuously and vigorously, kneading lumps against
the sides of the beaker to insure thorough mixing of the reagent with the emulsion. This operation shall be performed after bringing the weighed sample of
emulsion and the reagent to the standard temperature of 77 F.l.O F.
(25 C.0.5 C.).
.
One of the wire sieves shall be fitted over a beaker or other suitable vessel
and the mixture of emulsion and reagent poured through the sieve. The
beaker, containing the sample and glass rod, shall be rinsed with distilled water.
All lumps shall be kneaded and broken up and the washing of the beaker, ~-od
and sieve shall be continued until there is no longer any appreciable color imparted to the wash water. Mter washing as directed, the beaker, rod and sieve
used in each individual t.est sha]} be placed in a drying oven and dried at 325 F.
(163 C.) to constant weight.
The total weight thus obtained less the total tare weight of the beaker, rod
and sieve shall be the weight of the residue by the demulsibility test. The
ratio of the average weight in grams fro:gJ. three tests of each individual sample
of emulsified asphalt, A, to the weight in grams of residue per 100 g. of emulsion,
B, obtained in the test for residue by distillation described in Section 6 -multiplied by 100, shall be recorded as the percentage of demulsibility of the sample
tested:
Percentage Demulsibility =

~ X 100

.
Effects of Freezing
Approximately 400 g. of the emulsion shall be pla'ced in a clean metal container, such as a I-pt. press-top tin. The emulsion in the closed container
shall be exposed to a temperature of 0 F. (-17.7 C.) for twelve consecutive
hours. At the expiration of the freezing period, the emulsion shall be permitted
to thaw by exposure of the container to the temperature of the laboratory.
Mter the first operation of freezing and thawing, the procedure shall be
repeated twice, so that the emulsion will have been subje~ted to. tPree cycles of
freezing and thawing.
After the third cycle, the emulsion may be homogeneous or may have separated into distinct layers whi<;h cannot be ren,dered homogeneous by stirring
at laboratory temperature.
The result of this test shall be reported as ,either "Homogeneous" or
"Broken."
J'
(B)

SEPARATION OF THE DISPERSION INTO ITS

COMPONENT PARTS
I

The following products are likely to be pres_w.t, viz:: water, ammonia,

various chemicals, bituminous matter, animal ami-vegetable oils or fats, other
forms of non-bituminous organic matter and mineral matter.

EXAMINATION OF BITUMINOUS SUBSTANCES

r"

Distillation Residue.-The following method

has been standardized:
The following apparatus will be required:
(a) Iron Still.-The iron still shall be approximately 6 by 3,Yz in. in inside diameter with
adjustable ring burner with holes on the inner
periphery to fit around the outside of the still.

1593

1"

18 '8

NOTE.-A modification of this still is shown in Fig.

222. It consists of the regular still with an expansion
chamber superimposed thereon. Two additional ring
bUrlters are required, one approximately 6 in. in inside
diameter with the holes bored on the inside periphery,
and one approximately 2 in. in diameter with the holes
bored on top.

(b) Connecting Apparatus.-A connecting

tube, tin shield, condenser trough, condenser
tube "and graduated cyli~der, as shown in Fig. 223.
(c) Thermometer.-A thermometer graduated
from 30 0 to 580 0 F. (0 "to 300 0 C.).
NOTE.-The details of the assembly of apparatus
for the distillation test are illustrated in Fig. 223.
Still for
Exactly 200 g. of a well-mixed and representa-.
Foaming
tive sample of the emulsion shall be placed in the
previously weighed iron still (incluaing lid, clamp,
thermometer and gasket, if gasket is used). A gasket oJ oiled paper may be
used between the still and its cover or the joint ground to a tight fit. The
cover shall be clamped securely on the still. The thermometer shall be inserted
MtaChirtedioirtt

~~
;;

t;

~ _l::,,~
~ ~"'~

j!J-

1"
2 I.D.

it-b==- -----------1~J

'I~t_-------:
----------:t.:..-

.! Detail of Lid for Slil/

8

l! mm. glass tubing

- _

_j_

7"I.D.,'

11_

k----

Glass to glass iomt

wit" Tubber h,bing

'--_:::::.=,--=--..:~::;_":'

_,....

tL4'~!1~9~==:lt
,4!

7"

8_

13"8 -"--~
Detail of Tin Shield

FIG. 223.-Apparatus Assembly for Distillation Test of Emulsified Asphalts.

1594

EXAMINATION OF

BITUMINOU~

SUBSTANOES

through the small hole in the cover, using a cork stop'per, so that the end of the
bulb is
in. from the bottom of the still. The ring burner shall be placed
around the still and the heat applied by this means to the top of the still.
Just enough heat from a bunsen burner shall also be\applied to the connecting
tube to prevent condensation of water in this tube.
.
After practically all the condensate has been removed from the still and
the temperature of the residue has reached 250 0 F. (121 0 C.), the position of
the heat from the ring burner shall be lowered to midway of the still and held
there until the thermometer reaches 349 F. (176 C.). The burner shall then
be rapidly lowered to within in. of the 'bottom of the still and the temperature
increased to and maintained at 5000 F. (260 0 C.) for 15 min. This latter period
of heating is l).ecessary to insure a smooth homogeneous residue in the still.
At the expiration of the heating period at the maximum temperature, the still '
and accessories sha,V'again be weighed and the percentag~ residue ?alculate4, ~nd
reported. The cover shall then be removed from the stIll and sUItable portIOns
of the residue shall be poured immediately into suitable molds and containers
for making the required tests. The'residue in the molds and containers shall
be permitted to cool, uncovered, to laboratory room te'mperature and thereafter
tested for specific gravity, penetration, ductility, fusing-point, solubility in
carbon disulfide and ash.
It should be noted that the distillation residue will carry any non-volatile
dispersing agents present in the emulsiqn.

x:

x:

NOTE.-When it appears impossible to distill an emulsified asphalt in .the still

described above due to excessive foaming of the emulsion, then the modified still shown
in Fig. 222 should be substituted for the still shown in Fig. 223 and the following procedure followed: Place the 6-in. burner around the larger diameter of the still near its top.
This serves as a support. Place the 4-in. burner immediately beneath the flare and the
2-in. burner not less than 2 in. below the bottom. Distillation is started with only the
2-in. burner lighted. Practically all of the distillate should be over in about 45 min.
When the distillation apparently stops, light the two larger ring burners and adjust to a
low flame. Distillation resames and when it stops again increase the heat by adiusting
the flame of the 2-in. burner. When the temperature cim b~ read upon the thermometer,
increase the rate of heating by raising the flame on botH the 2 and 4-in. burners and
bring the temperature to 500 F. (260 C.). If any evidence is noted of the emulsion
beginning to foam over in the delivery tube, remove the 2-in. burner quickly and raise
a pan of water so as to immerse the still bottom to a depth of about 2 in, for a moment
which will check the foaming. Upon resumption of heating watch delivery tube carefully
and repeat treatment if necessary.
When the residue has reached and remained at 500 F. (260 0 C.) fQr.15 min., proceed
as described for the regular still. While the distillation should be completed in not less
than 1 hr. nor more than 174: hr. from the first application of heat to the still, the maximum stated is not mandatory as a longer time may be required in some cases to avoid
foaming of the emulsion into the condenser. If the, residue in the still prior to pouring
the ductility and penetration specimens appears granular or heterogeneous in any way,
stir with a spatula until the material runs from the spatula in strings' instead of drops,
and then pour. Should the residue then appear granuiar and heterogeneous, the test
shall be rejected and the distillation repeated as previously described with the exception
that the still temperature at 500 F. (260 C.) shall be maintained for a longer time than
15 min. until the desired condition of the residue is obtained.

Water and Volatile Oils.-The following methoti_has peen standardized:

This method of test determines water exiating-in a .sample of bituminous
emulsion by distilling the sample with a volatile solvent.

EXAMINATION OF BITpMINOUS SUBSTANCES

1595

(a) The apparatus shall consist of a metal still or glass flask, heated by
suitable means and provided with a reflux condenser discharging into a trap
connected to the still or flask. The trap serves to collect and measure the
condensed water and to return the solvent to the still.
-J, I-<-H Inside Oi.m.
The type of distilling apparatus used' is not an essential
feature of this method.
(b) Metal Still.-The metal still shall be a vertical
cylindrical vessel, preferably of copper, having a faced
flange at the top to which the head is tightly attached
by means of a clamp. The head shall be of metal, preferably of brass or copper, and be provided with a tubulation 1 in. in inside diameter. .
'
D
(c) Glass Still.-The glass flask shall be of the short
neck, round-bottom type, made of well-annealed' glass,
having an approximate capacity of 500 ml.
(d) Heat Source.-The burner used with the metal
still shall be a ring gas burner 100 mm. {4 in.) in inside
diameter. With the glass flask, an ordinary gas burner
or electric heater may be used as the source of heat.
(e) Condenser.-The condenser shall be of the waterE 0
....L..
cooled, reflux, glass;tube type, having a condenser jacket
C Inside__"j ~
Oi.m,
not less than 400 mm. (15U in.) in length with an inner
tube 9.5 to 12.7 mm. (Ys to Y:2 in.) in outside diameter.
A= 45 to 55 mm.
B = 14 to 16 mm.
The end of the condenser to be inserted in the trap shall
c= 12 to 16 mm.
be ground off at an angle of 300 from the vertical axis of
0=235 to 255 mm.
E = 23 to 38 mm.
the condenser.
.
F=186 to 194 mm.
(f) Trap.-The trap shall be made of well-annealed
H= 78 to 79 mm.
glass constructed in 'accordance with Fig. 224, and shall
FIG. 224.
be graduated from 0 to 25 ml. in O.l-ml. divisions. The
tolerance of the graduations between 0 and 2 ml. shall be 0.5 ml. and b~
tween 2 and 25 ml. shall be O.l ml. The outside diameters should be preferably 2.5 to 3.5 mm. Uz to VB in.)' greater than the inside diameters specified.

!u

The solvent used when testing bituminous emulsions shall be a coal-tar

naphtha or a light oil and shall conform to the following distillation requirements:
98 per cent shall distil between 248 F. (120 C.) and 482 0 F. (250 C.)
0

The sample shall be thoroughly representative of the material to be tested

and the portion of the sample used for the test shall be thoroughly representative of the sample itself. 'Deviation from this requirement shall not be permitted.
NOTE.-The difficulties in obtaining proper representative samples for this determination are unusually great so that the importance of sampling cannot be too strongly
emphasized,
When the material to be tested contains less than 25% of water, exactly a
lOO-g. sample shall be placed in the still or flask. When the material contains
more than 25% of water, the sample shall be 50 g. The sample to be tested
shall De thoroughly mixed with an equal volume of solvent by swirling, proper
care being taken to avoid any loss of material.

1596

EXAMINATION OF

BITUMINOU~

SUBSTA,NCES

The connections between the still or flask, trap and condenser shall be made
by means of tight-fitting corks as shown in Fig. 21Q (a) and (b). The end of
the condenser inserted in the trap shaH be adjllsted to that position which will
allow the end to be submerged to a depth of not :qJ.pre than 1 mm. (0.04 in.)
below the surface of the liquid in the trap after distillation conditions have been
established. When the metal still is used, a heavy paper gasket moistened
with the solvent shall be inserted between the lid and flange before attaching
the clamp. A loose cotton plug shall be inserted in the top of the condenser
tube to prevent condensation of atmospheric nlOisture in the condenser tube.
Heat shall then be applied and so regulated th"at the conden~ed distillate
falls from the end of the condenser at the rate of from 2 to 5 drops per sec.
The ring burner used with the metal still shall be placed about 3 in. above
the bottom of the still at the beginning of the distillation and gradually lowered,
as the distillation proceeds. The distillation shall be continued at'the specified
rate :Intil no water is visible on any part of the apparatus except at the bottom
of the trap. This operation usually requires l(lss than 1 hr. A 'persistent ring
of condensed water in the condenser tube shall be removed by increasing the
rate of distillation for a few minutes.
The volume of condensed water measured in the trap at room temperature
multiplied by 100 and divided by the weight of sample used, shall be the
percentage of water and shall be reported as ", " . per cent water by weight."
The accuracy to be expected with this lilethod is that duplicate determinations of water should not differ from each other by more than one di:vision
on the trap.
The sum of the percentages of Distillatioll Residue and Water, deducted
from 100 represents the percentage of Volatile Oils.

CEMENT
SPECIFICATIONS AND CHEMICAL ANALYSIS 1
The physical tests ordinarily applied to Portland cement are as follows:
Setting time.
S_oundness.
Tensile strength.2

Che~ical analysis is als~ made, particular attention being paid to the determination of magnesia, sulfuric anhydride, insoluble residue, and loss on ignition.
As a general rule, however, it may be said that so far as the consumer is concerned, more attention is paid to the physical tests than to chemical analysis.
Standard specifications covering the requirements for cement, both chemical
and physical, have been adopted by the American Society for Testing Materials,
and by the U. S. Government." The former are generally recognized by cement
users as the standard requirements, while the latter are used by the various
branches of the federal government. The two sets of specification may be very
different.
The specifications and methods of making the chemical tests follow.
o

SPECIFIOATIONS 4
.

Definition.-Portland cement is the product obtained by finely pulverizing

clinker produced by calcining to incipient fusion an intimate and properly
propprtioned mixture of argillaceous and calcareous materials, with no additions subsequent to calci~ation excepting water and calcined or uncalcined
gypsum.
Chemical Limits.-The following maximum limits shall not be exceeded by
amounts greater than the respective tolerances indicated as allowable in the
chemical determinations:
'
Tolerance
Limits
Loss on ignition, %, _..... _. . . . . . . . . . . . . . . . . .. 4.00
0.25
0.15
Insoluble residue, % ..... , , , , . . . . . . . . . . . . . . . .. 0.85

0.10
Sulfuric anhydride (S03), % ............ , .. , . " 2.00
0.40
Magnesia (MgO), %........................ " 5.00
1 By W. C. Hanna, Chief Chemist and Chemical Engineer, California Portland
Cement Co., Colton, California.
2 See A. S. T. M. Method C 77-37 for the Physical Tests.
1937 Supplement to
Book of A. S. T. M. Standards, pp. 61-73.
3 The federal specifications are issued by the Director of Procureme_nt, Washington,
D. C., and sold by the Superintendent of Documents, Government Printing Office,
Washington, D. C. The specifications now in effect covering,cement are as follows:
88-0-158. September 30, 1936. Cements, Hydraulic; General Specifications (Methods
for Sampling, Inspection, and Testing).
S8--C-191a. September 30, 1936. Cement; Portland.
8S-O-201. September 30, 1936. Cement; Portland, High-Early-Strength.
S8--C-206. September 30, 1\}36. Cement; Portland, ModeratecHeat-of-Hardening.
S8-0-181a. May 22,1936. Cement; Masonry.
4 Standard Specifications for Portland Cement.
A. S. T. M. Designation: C 9-37.
1597

CEMENT

1598 '

The residue on a standard No. 200 sieve shall not exc~ed 22% by' weight.
A pat o,f neat cement shall remaJn firm and- hard, and show no signs of
distortion, cracking, checking, or disintegration in the steam test for soundness.
The cement sh:}ll not develop initial set in less than ~5 minutes when the
Vicat needle is used or in less than 60 minutes when the Gillmore needle is used.
Final set shall be attained within 10 hours.
'
The average tensile strength in pounds per square inch bf not less than three
standard mortar briquettes composed of one part cement and three parts
standard sand, by weight, shall be equal to or higher than the following:
Age at Test,
Days

Storage of Briquettes

7
28

1 day in moist air, 27 days in water .. " .... .

1 day

in moist air, 6 days in water ........ .

Tensile Strength,
lb_ per sq. in.

275
350

The average tensile strength of standard mortar at 28 days shall be higher

than the strength at 7 days:
'
Packages and Marking.-The cement shall be delivered in packages as
specified with the brand and name of the manufacturer plainly marked thereon,
unless shipped in bulk. When shipped in bulk, this information 'shall be contained in the shipping advices accompanying the shipment. A bag shall
contain 94 lb. net. A barrel shall contain 376 lb. net. All packages shall be
in good condition at the time of inspe~tion.

The cement shall be stored in such a manner as to permit easy acce~s for
proper inspection and identification of each shipment, and in a suitable weathertight building which will protect the cement from dampness.
,
Inspection.-Every facility shall be provided the purchase:r_ for careful
sampling and inspection at either the mill or at the site of the work, as may be
specified by the purchaser. At least 12 days from the tjme of sampling shall
be allowed for the completion of the 7-day test, and at :least 33 days shall be
allowed for the completion of the 28-day test. The cement shall be tested in
accordance with the methods hereinafter prescrilled. The 28-day test need
not be made if waived by the purchaser.
Rejection.-The cement may be rejected if it fails to meet any of the requirements of these specifications.
_
Cement remaining in sto;age prior to shipment for a period' greater than 6
months after completion of the tests shall be retested and shall be_rejected if
it fails to meet any of the requirements of these specifications.
Cement failing to meet the test for soundness.in steam may be accepted
if it passes a retest using a new sample at any time within 28 days thereafter.
The provisional acceptance of the cement at the miF shall not deprive the
purchaser of the right"Of rejection on a retest of soundness and time of setting
at the time of delivery of cement to the purchaser.
Packages varying more than 5% from the specified weight may be rejected;
and if the average weight of packages in any shipment, as sho'}!n by weighing
50 packages taken at random, is less than that spe3~d; the entire shipment
may be rejected.
. -

1599

CEMENT

Methods of Testing.-The cement shall be sampled and tested in accordance with the Standard Methods of Sampling and Testing Portland Cement
(A. S. T. l\f. Designation~ C 77) of the American Society for Testing Materials.

STANDARD METHODS OF SAMPLING AND TESTING

PORTLAND CEMENT 5

A. S. T. M.

DESIGNATION:

C 77-37 6

SAMPLING
Size and Number of Samples.-l. Samples for purpose of tests shall weigh
at least 4 lb.

2. Test samples shall be either individual or composite samples, as may be

ordered, and one test sample shall represent not more than 300 bbl., unless
otherwise specified by the purchaser.
Methods of Sampling.-3. When cement is sampled in cars, trucks, boats,
warehouses, etc;, one sample shall be taken from one sack in each 40 sacks (one
sample per 10 bbl.), and combined to form one test sample. In the case of
truck samples where the cement is being trucked from one mill, it is permissible
to combine the samples from several trucks to form a test sample representing
not more than 300 bbl. When sampling bulk shipments, representative samples shall be taken from well distributed points.
Sampling Bulk Storage.-4. The cement may be sampled at bulk storage
points by any of the following methods:
.
(a) From the Conveyor Delivering to Storage.-At least 4 lb. of cement
shall be taken from at least each 300 bbl. passing over the conveyor. This
may be secured by taking the entire test sample at a single operation, known
as the "Grab Method," or by combining several portions taken at regular
intervals, known as the" Composite Method." When obtaining a composite
sample this shall be secured by combining approximately equal weights taken
at regular intervals, each poftion representing approximately 10 bbl. Automatic samplers may be used in obtaining samples.
5 Under the standardization procedure of the Society these methods are under the
jurisdiction of the A. S. T. M. Committee C-1 on Cement.
A Manual of Cement Testing to supplement these methods has been prepared and
published as information, not a part of these methods, see Proceedings, Am. Soc. Testing
Mats., Vol. 32, Part I, p. 316 (1932).
6 These methods are issued under the fixed designation C 77; the final number
indicates the year of original adoption as standard or, in the case of revision, the year
of last revision. Adopted, 1904; Issued in Amended Form Under Present Designation,
1930; Revised, 1932, 1937. Reproduced with the permission of the American Society
_ for Testing Materials.

1600

CEMENT

(b) From Storage by Means of Proper Sampling Tubes.-Tubes inserted

vertically may,be used for sampling cement to a maximlJ.ffi depth of 10 ft.
Tubes inserted horizontally may be used where the construction of the storage
permits. Samples shall be taken at points well distributed lover the storage.
(c) From Storage at Points of Discharge.-Sufficient cement shall be drawn
from the discharge openings to obtain samples representative of the cement,
as determined by the appearance at the discharge openings df indicators placed
on the surface of the cement directly above these openings before the drawing
of the cement is started. One 4-lb. sample shall be taken for at least each
300 bbl.
.
Preparation of Sample.'-5. The sampling shall be done by or under the
direction of a responsible representative of the purchaser.
6. Samples shall be shipped and stored in moisture-proof, air-tight containers.
Before testing, samples shall be thoroughly mixed and then passed through a
No. 20 sieve in order to' break up lumps and remove foreign materials.
CHEMICAL ANALYSIS

Number of Determinations.-7. (a) When cement is sampled in accordance

with Section 3 at least.one determination of loss on ignition, insoluble resi9.ue
content and magnesia content shall be made for each 10 test samples taken, and
at least one determination of sulfuric anhydride (S03) conten~ shall be made for
each 3 test samples. In case of truck deliverres and car deliveries the sulfuric
anhydride (SP3) content shall be determined on each test sample. When the
total number of test samples taken is less than either of the groups above sptcifled at least one determination shall pe made for each chemical requirement.
(b) When cement is sampled in accordance with Section 4 at least one
determination of loss on ignition, insoluble residue content and magnesia
content shall be made for each 15 test samples and at least one determination
of sulfuric anhydride (80 3) content shall be made for each 3 test samples;
provided that at least one determination for each chemica) requirement shall
be made on each bin sampled.
(c) The chemical determinations may be lIlade either on composites of the
original test samples or may be made on individual test samples selected at
random at the option of the purchaser.
(d) The chemical determinations shall be made in accordance with the procedure described in the following Sections 8 to 17 (Note).'
NOTE.-This method of chemical analysis is not considered as possessing the highest
obtainable accuracy, but is an acceptance method to be followed in making acceptance
tests on cements covered by specifications requiring that tests be made in accordance
with the analytical procedure contained in this standard.

Loss on Ignition.-8. Heat a I-g. sample of cement.in a weighed covered

platinum crucible of 20 to 25-ml. capacity, as follows, using either Method 1 or 2
as specified.
(a) Method I.-Place the crucible in a hole in an asbestos board, clamped
horizontally, so that about three fifths of the crucible projects below the board,

7 Thi~

method of chemical analysis is very similar to that p,repared by the National

Bureau of Standards and adopted by the U. S. Department-'o(the Interior, Bureau of
Reclamation, in Specifications No. 566. Very few revisions have been made.

CEMENT

1601

and heat at a full red heat (900 0 to 1000 0 C.) for 15 min.; check the loss in
weight by a second heating for 5 min. Take care to wipe off particles of' asbestos that may adhere to the crucible when withdrawn from the hole in the
board. Greater neatness and shortening of the time of heating may be secured
by making a hole to fit the crucible in a circular disk of sheet platinum and
placing this disk over a somewhat larger hole in an asbestos board.
(b) Method 2.-Place the crucible in a muffle furnace at any temperature
between 900 0 and 1000 0 C. for 15 min. and check the loss in weight by a second
heating for 5 min.
Insoluble Residue.-9. Add 25 ml. of cold water and 5 ml. of HCl (sp.gr.
1.19) to a I-g. sample of cement. Heat the solution and grind the material with
the flattened end of a glass rod until it is evident that decomposition of the
cement is complete. Dilute the solution to 50 mI. and digest on a steam bath
for 15 min. Filter the residue, wash six times with hot water, and digest the
filter paper and contents in about 30 mI. of a sodium carbonate solution (5%),
the liquid being held at a temperature just short of boiling for 15 min. Filter
the remaining residue, wash twice with hot water, then with a few drops of hot
diluted HCI (1 : 9), and finally eight to ten times with hot water, then ignite
at a red heat and weigh as the insoluble residue. .
Sulfuric Anhydride.-lO. Add 25.mI. of cold water and 5 mI. of HCI (sp.gr.
1.19) to a I-g. sample of cement. Heat the solution and grind the material with
the flattened end of a glass rod until it is evident that decomposition of the
cement is complete. Dilute the solution to 50 mI. and digest on a steam bath
for 15 min., filter, and wash the residue thoroughly with hot water. Dilute the
solution to 250 mI., heat to boiling, add slowly, drop by drop from a pipette, 10
mI. of a hot barium chloride solution (10%), and continue the boiling until the
precipitate is well formed. Then digest. the solution on the steam bath for at
least 3 hr.; preferably overnight. Filter the precipitate, wash and place the
paper and contents in a weighed platinum crucible and slowly char and consume
the paper without flaming. Then ignite, and weigh the barium sulfate. The
weight obtained multiplied by 34.3 gives the percentage of sulfuric anhydride.
The acid filtrate obtained in the determination of the insoluble residue may be
used for the estimation of sulfuric anhydride instead of using a separate sample.
Silica.-l1. (a) Transfer a 0.5-g. sample of the finely powdered cement to an
evaporating dish, preferably of platinum for the sake of celerity in evaporation,
moisten with 10 mI. of cold water to prevent lumping add 5 to 10 ml. of HCI
(sp.gr. 1.19) and digest with the aid of gentle heat and agitation until solution is
complete. Solution may be aided by light pressure with >the flattened end of a
glass rod. Evaporate the solution to dryness on a steam bath. Treat the
residue without further heating at first with 5 to 10 mI. of HCl (sp.gr. 1.19) then
diluting to half strength or less, or pour at once upon the residue a larger volume
of acid of half strength. Then cover the dish and digest for 10 min. on the
bath or a hot plate. Dilute the solution with an equal volume of hot water,
immediately filter, and wash the separated silica thoroughly with hot water
and reserve. Again evaporate the filtrate to dryness, baking the residue in an
oven 1 hr. at a temperature of 105 0 to 1100 C. Then take up the residue with
10 to 15 mI. of diluted HCI (1 : 1) and heat on the bath or hot plate. Dilute
the solution with an equal volume of hot water and catch and wash the small
amount of silica it contains on another filter paper. Reserve the filtrate and

1602

CEMENT

washings. Transfer this paper and the r~sel'yed paper containing the residues
to a weighed platinum crucible, dry, ignite first over aibunsen burner until the
carbon of the filter paper is completely consumed without flaming and finally
over a burner or in an electric oven at 11000 to 12000 C, ~mtil the weight remains
constant.
I

(b) Treat the silica thus obtained, which will contain small amounts of
impurities, in the crucible with a few drops of water, about 10 m!. of HF and
one drop of H 2S0 4 and evaporate to dryness over a low flame. Finally, heat the
,small residue at 1050 0 to 11000 C. for a minute or two, cool, and weigh. The
difference between this weight and the weight previously obtained represents
the amount of silica. Washing of the silica precipitates can be made more
effective by using a hot solution of diluted HCI (1 : 99), and then completing
the washing with hot water. Consider the weighed residue, remaining after
volatilization of the silica, as combined aluminum and ferri'c oxides and add it
to the result obtained in the determination of these oxides (Section 12).
Ferric Oxide and Alumina.-12. Treat the reserved filtrate (Note 1), which
should have a volume of about 250 mL, with HCI, if necessary, to insure a total
of 10 to 15 mL of the acid, add a few drops 'of methyl red indicator and then
heat to boiling. Then treat with NH 40H (Note 2), drop by drop, until one
drop changes the color of the solution to a ,distinct yellow. Bring to boiling
the solution containing the precipitated iron and aluminum hydroxides, boil
for 1 or 2 min., and then filter. Wash the precipitate onc(l by decantation and
slightly on the filter with a hot ammoni~m nitrate solution (2%). Sef\ aside
the filtrate and transfer the precipitate and filter paper to the same beaker in
which the first precipitation was effected. Then dissolve the iron and aluminum
hydroxides in hot diluted HCI (1 : 3), make up the solution to about 100 mI.,
and reprecipitate the hydroxides as i1,). the first case. Then filter the solution
and wash the precipitate with two lO-mL portions of a hot ammonium nitrate
solution (2%). Combine the filtrate and washings with that set aside and
place the precipitate in a weighed platinum crucible, slowly heat by a bunsen
or Meker burner until the papers are charred, and finally ignite to constant
weight at 1050 0 to 1100 0 C. with, care to pr~vent /reduction, and weigh as
Al 2 0 a FezOa.
I

NOTES.-l. If a platinum evaporating disn has been used for the dehydration of
silica, iron may have been partially reduced. At thil1 stage, add about 3 ml. of saturated

aqueous solution of bromine to the filtrate and boil the filtrate to eliminate the excess
bromine before adding the methyl red indicator.
'2. The NH 40H used to precipitate the hydroxides must be free 01 contamination
with carbon dioxide.
_

Ferric Oxide.-This method for determining ferric oxide is not affected by

any titania or vanadium oxide that may be in the cement. It is therefore
understood that these are counted as Al 20 a in this method. If the precipitations of aluminum and iron are made as directed, little if any manganese will'
'
be retained in the precipitate.
Solutions Required.-13. The following solutions will be required:
(a) Stannous Chloride.-Dissolve 2 g. of the crystallized salt in four times
its weight of a mixture of three parts of water and one ofiHCI (sp.gr. 1.19).
Add scraps of iron-free granulated tin and boil u_ntil~he solution is clear. Keep
the solution in a closed dropping bottle containing-metallic tin.

CEMENT

1603

(b) Phosphoric Acid Mixture.-Add slowly and with stirring 150 ml. of
orthophosphoric acid (sp.gr. 1.71), and 150 ml. of H 2 S0 4 (sp.gr. 1.84), to 500 ml.
of water and then dilute to 1000 ml. with water and mix.
(c) Diphenylamine Indicator.-Dissolve 1 g. of diphenylamine in 100 ml.
of H 2S0 4 (sp.gr. 1.84).
(d) Standard Potassium Bichromate Solution.-Dissolve 2.457 g. of
K 2 Cr 207 in distilled water and dilute to 1 liter. Standardize against standard
iron wire of known iron content in the same manner as directed for the ferric
oxide determination in cement (Section 14), using a weight of iron wire which'
will yield a titration approximately equal to that required by the cement sample
in question. .
Method.-14. Add 20 m!. of cold water and 10 m!. of HCI (sp.gr. 1.19) to a
I-g. sample of cement. Heat the solution and grind the material with the
flattened end of a glass rod until it is evident that decomposition of the cement
is complete. Dilute the solution to 50 ml. and heat to boiling. Then treat
the solution with 5 m!. of HC:! (sp.gr. 1.19) and then with a solution of stannous
chloride, added drop by drop while stirring until the solution is decolorize<;l,
then add one drop more. Cool the solution to room temperature, rinse the
inside of the vessel with water and add all at once 10 m!. of a cool saturated
solution of mercuric chloride. Then stir the solution vigorously for 1 inin.,
treat with 15 ml. of phosphoric acid mixture and three drops of indicator and
dilute to 150 m!. with cold water. Then titrate with a standard potassium
dichromate solution having a titer no greater than 0.004 g. of Fe 20a. The
end point is taken as the point when a drop causes an intense deep-blue coloration that remains unchanged on the further addition of the standard K 2 Cr 2 07
solution.
Calcium Oxide.-15. Add a few drops of NH 40H to the combined filtrates
obtaided in the alumina and ferric oxide precipitation (Section 12) and bring
the solution to boiling. Add to the boiling solution 25 ml. of a boiling ammonium oxalate solution (5%), the boiling being continued until the precipitated
calcium oxalate assumes a well-defined granular form. Then allow the solution
to stand for about 20 min. or until the precipitate has settled, filter, and moderately wash with an ammonium oxalate solution (0.1 %). Reserve the filtrate
and washings. Place the filter paper containing the precipitate, wet, in a
platinum crucible and burn off the paper over the small flame of a bunsen
burner. Then ignite the precipitate until the oxalate is converted to oxide
(Note), redissolve in HCI and .make up the solution to 100 ml. with. water.
Add NH 40H in slight excess and bring the solution to boiling. If a small
amount of alumina separates at this point, filter it out, weigh, and add the
amount to that found in the original alumina determination (Section 12).
Then reprecipitate the lime by ammonium oxalate, allow to stand until settled,
filter, and wash, taking care to avoid an excess of the washing solution. Combine the filtrate with that already reserved. Weigh the precipitate as calcium
oxide after ignition and blasting to constant weight in a weighed, covered
platinum crucible.
NOTE.-If the ignited precipitate is darkly colored, it may contain an appreciable
amount of manganese. The contamination of the second precipitate of calcium oxalate
by manganese should be as low as possible. The following procedure may be used to
reduce the contamination: add 5 mI. of cold water cautiously to the precipitate and then

1604

CEMENT

cautiously add 10 mI. of diluted HOI (1 : 1). Heat the mixture. If the resulting solution
is colored, digest it on the hot plate gently for 10 min. or longer until the color disappears
or until it is apparent the color will not become lighter on further heating. After a
little practice it is possible to dissolve hard-burnt lime in this way without explosive
spattering.
I
If the second precipitate of calcium oxalate is so fine as tol cause difficulty in filtering
or if the precipitate causes bumping, the volume of solution may be increased to 150 or
200 ml., instead of 100 mI. prior to precipitation.

Magnesia.-16. In this method for determining magnesia, most of any man. ganese in the cement is precipitated with the magnesium. It is understood
in this method that manganese so precipitated shall be calculated as magnesium.
Acidify the filtrates reserved in the determination of calcium oxide (Section
15) with HCI and concentrate to about 150 mI. Add to this solution about
10 ml. of diammonium phosphate solution (25 g. of the salt per 100.ml. of
water) and boil the solution for several minutes. Then cool the solution by
placing in a beaker of ice water. After cooling, ~dd NH 40H drop by drop,
while stirring constantly, until the crystalline ammonium-magnesium orthophosphate begins to form, and then in moderate excess (5 to 10% of the volume
of the solution), the stirring being continued for several minutes. Set the
solution aside for at least 8 hr. in a cool atmosphere, and then filter. Redissolve
the precipitate thus obtained in hot diluted HCI (1 : 4), make up the solution
to about 100 mI., add 1 ml. of the solution of diammonium phosphate, and
NH 40H drop by drop, while stirring coi"lstantly until the precipitate is ag~in
formed as described, and the ammonia is in moderate excess. Then cool,
allow to stand for about 2 hr., filter, and wash with two lO-ml. portions of wash
solution (200 m!. of concentrated NH 4 0H, 100 g. of ammonium nitrate, and
800 m!. of wat'er), place in a weighed platinum crucible, slowly char the paper,
and carefully burn off the resulting carbon. Ignite the precipitate at Do temperature between 11000 and 12000 C. to constant weight, taking care to avoid
bringing the pyrophosphate to melting. The weight of magnesium pyrophosphate obtained multiplied by 72.4 gives the p.erQfmtage of magnesia.
Blank Determination.-17. Make a blank determination for sulfuric anhydride, silica, ferric oxide and alumina, ferric oxide, calcium oxide, and
magnesia, using the same procedures as outlined in Sections 8 to 16, and using
approximately the same amounts of reagents. Correct accordingly the results
previously determined. .
_
Balances.-18. The balances used in the chemical determinations shall conform to the following requirements:
(a) Capacity not less than 100 g. in each pan;
(b) The two arms of beam to be equal to withi'i'J. one part in 100,000;

(c) Capable of reproducing results within 0.1 mg.;

(d) "Sensibility reciprocal" 8 not more than 0.2 mg. per division of the

graduated scale.
8 The sensibility reciprocal is a measure of the sensitivity of a balance, and is the
weight required to move the position of equilibrium of the beam, pan, pointed, or other
indicating devi~e of a scale a definite amount at the canamty or at any lesser load. For
a complete definition of sensibility reciprocal, see Bureau -of-Standards Handbook, M 85,
pp.80--82.

CEMENT

1605

(e) The weights used in the chemical determinations shall conform to the
requirements of the National Bureau of Standapds specifications for Class
" S " weights as contained in Bureau of Standards Circular No.3.

TENTATIVE METHOD OF CHEMICAL ANALYSIS OF

PORTLAND CEMENT 9
A. S. T. M.

DESIGNATION:

C 114-37 T

10

METHOP A. ACID-ALKALI METHOD FOR RAPID

DETERMINATION OF MAGNESIA 11
Solutions Required.-1. The following solutions will be required:
(a) Trinitrobenzene Indicator.-Dissolve 1.0 g. of C.P. 1-3-5 trinitrobenzene in 100 m!. of ethyl alcohol (95 to 100%).
(b) Phenolphthalein I~dicator.-Dissolve 0.4 g. of C.P. phenolphthalein in
100 m!. of ethyl alcohol (50%).
(c) Standard Sodium Hydroxide Solution.-Dissolve 13.0 g. of sodium hydroxide in water, add 10 m!. of a freshly filtered saturated barium hydroxide
solution, aI!d dilute to 1 liter with dIstilled water which has been recently boiled
and cooled. Shake the solution from time to time for several hours, and filter
into a tubulature bottle fitted with an autpmatic burette. Protect the solution
from carbon dioxide in the air, with soda-lime tubes at the top of the bottle
and the burette.
The solution as prepared is approximately Pi N, and shall be standardized
against acid potassium phthalate or benzoic acid acidimetric standards furnished by the National Bureau of Standards (Standard Samples Nos. 84 and
39), using the methods in the certificates accompanying the standard samples.
The exact normality of the solution shall be determined, and its MgO equivalent
calculated on the basis that 1 m!. of a normal solution is equivalent to 0.02016 g.
of MgO.
(d) Standard Acid Solution.-Dilute 32.0 m!. of HCI Csp.gr. 1.19) or
26.0 m!. of HNO a (sp.gr. 1.42) to 1 liter with distilled water. This solution
9 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee C-1 on Cement.
10 This is a Tentative Standar:d and under the Regulations of the Society is subject to
annual revision. Suggestions for revision should be addressed to the Headquarters of
the Society, 260 S. Broad St., Philadelphia, Pa. Issued, 1934; Revised, 1935., 1937.
Reproduced with the permission of the American Society for Testing M~terials. .
11 This procedure is a modification of the method originally published by Fred P.
Diener in Concrete, 39, No.6, December, 1931. The changes, which are in minor details
and do not affect the underlying principles of the method, were made by the author in
cooperation with the Working Committee on Methods of Chemical Analysis of Cement
of the A. S. T. M. Committee C-1 on Cement for use in the investigations sponsored by
the working committee. The theory of the method is discussed in the original article.

1606

CEMENT

will be slightly in excess of 73 N, and shall be standardized against the standard

sodium hydroxide solution, .using phenolphthalein'indicatof' and adjusted by
dilution until the two solutions are of equal strength.
CI
(e) Potassium Iodide-Starch Paper.-Mix equal volu~es of a 5% potassium iodide solution and a 1% soluble starch solution, and I dip strips of filter
paper into the mixture.
,Method.-2. (a) To 5.0 g. of cement in a BOO-m!. beaker shall be added
about 50 m!. of water and 50 ml. of HCI (1 : 1), and the solution boiled until the
cement is completely in solution, excepting possibly some silica which may
separate out (Note 1). Then 5 m!. of saturated bromine water shall be added
and the boiling continued until the excess bromine is completely expelled, as
indicated by the absence, of a blue coloration on potassium iodide-starch paper
when held in the issuing steam. The solution shall then be diluted to about
250 m!. with hot water and a 20% solution of sodium hydroxide added to the
boiling solution from a pipette, until the excess acid is neutralized and a permanent precipitate just forms. Then, continuing the boiling, HCI (1 : 1) .
shall be carefully added from a pipette until the precipitate dissolves and the
solution assumes a yellow ferric chloride color, avoiding much excess acid.
The sides of the beaker shall be washed down with hot water and the boiling
continued for a minute or two, andthen the beaker shall be removed from the
burner.
'
(b) To the solution shall be cautiously added 4.0 g. of calcium carbonate
(weighed to the nearest 0.1 g.). This amount should be sufficient to completely
neutralize the acid, with some excess. The solution shall then be diluted to
about 400 ml. with hot water and boiled vigorously for 10 min. The solution
shall be removed from the burner, and a 0.5% potassium permanganate solutron
carefully added to the hot solution from a burette, with vigorous stirring, until
the supernatant liquid shows a faint pink color after settling for a moment.
Avoid use of more permanganate than necessary to produce a faint .. pink color.
Then the boiling shall be continued for 2 min., 2.0 ml. of ethyl alcohol added,
and the boiling continued for a few minutes or until the pink color has disapI
peared (Note 2).
(c) The contents of the beaker shall then be ,transferred to a 500-m!.
volumetric flask, cooled to room temperature, diluted to the mark, and mixed
by shaking. The solution shall be filtered through a dry paper, and after
discarding the first 10 or 20 m!. of filtrate, 200 m!. of the filtrate shall be transferred with a pipette to a 500-m!. Erlenmeyer flask. The solution in the flask
shall be heated to boiling, aitd 3.0 m!. of trinitrobenzene indicator added.
Then standard sodium hydroxide sloution shall be added 1.0 m!. at a time with
intermittent boiling for a few seconds, until the magnesium is completely precipitated and a definite pink color persists on boiling. The total amount of
alkali used shall be recorded (Note 3).
Cd) The solution shall be transferred to a 250-m!. v.olumetric flask, cooled
to room temperature, diluted to the mark, and mixed by shaking. The solution
shall be filtered through a dry paper, into a graduated cylinder, discarding the
first 10 or 20 m!. of filtrate, until 125 m!. of filtrate are secured. The 125 m!.
of filtrate shall be transferred to a flask, a few drops of phenolp)lthalein indicator. added, and the solution titrated with standard acid solution to the disap. -'-:pearance of the pink color.

CEMENT

1607

(e) Calculation.-If ,the acid and alkali solutions are of equal strength,
the quantity of alkali in milliliters minus twice the quantity of acid in milliliters
is equivalent to the MgO in a 2-g. sample:

MgO, per cent =

100(a-2b).
.
2
X MgO eqUIvalent of 1 m!. standard alkalI,

where a = the quantity of standard alkali in milliliters;

b = the quantity of standard acid in milliliters.
(f) Blank Determination.-A blank determination shall be made, using
the same procedure as outlined above, and using approximately tIie same
amounts of reagents. The titration results shall be corrected accordingly.
, NOTEs,-L Some samples of cement, particularly those containing considerable iron
or manganese, may require longer boiling or the addition of slightly more HCI to effect
complete solution.
2. The purpose of the addition of potassi'um permanganate is to remove ms.nganese
from the sample, which otherwise would be precipitated and reported as MgO. The
permanganate in a hot solution oxidizes the manganese to insoluble MnO, and is itself
reduced to MnO" The ethyl alcohol precipitates the excess permanganate. In samples
containing little or no manganese, this portion of the procedure may be omitted.
3. Enough standard alkali should be added to completely precipitate the magnesium
without too large an excess which might cause some precipitation of calcium. A guide
to proper procedure is the amount of back titration with acid, which should be held
between 1.0 and 2.0 mI. (after being doubled). A little experience with the color change
of the trinitrobenzene indicator will enable the operator to hold the back titration between these limits.
METHOD B. RAPID DETERMINATION, OF MAGNESIA BY
TITRATION OF MAGNESIUM OXYQUINOLATE 12
Solutions Required.-3. The following solutions will be requir!)d:
.
(a) 8-Hydroxyquinoline (1.25%).-Dissolve 25 g. of 8-hydroxyquinoline in
60 m!. of glacial acetic acid. When the solution is complete dilute to 2 liters
with cold water. One milliliter of this solution is equivalent to 0.0016 g. of
MgO.
(b) Sodium Thiosulfate Solution (0.1 N).-Dissolve 25 g. of Na 2 S 2035H 20
in 200 m!. of water and dilute to 1 liter. Standardize this solution either on
sodium oxalate through a permanganate solution, or on arsenious oxide through
a solution of resublimed iodine. Standard Samples Nos. 40 and 83 of the'
12 This procedure is a ,modification of the procedure originally published by J. C.
Redmond in the 'Bureau of Standards Journal of Research, 10, No.6, June, 1933. The
revision was made after correspondence-with Mr. Redmond and for use in investigations
sponsored by the Working Committee on Methods of Chemical Analysis of Cement of
the A. S. T. M. Committee C-l on Cement. The revised procedure differs from the
original one mainly in two respects: (1) The hydrates of iron and aluminum are precipitated twice instead of once; and (2) the amounts of 8-hydroxyquinoline reagent and
standard bromate solution used are variable instead of fixed.

The original article should be consulted for a discussion of the theory and experimental basis of the method. The method was developed primarily for use by the National Bureau of Standards for the detection of cements whose magnesia content did
not meet the requirements of Federal Specifications Board Specification No. la. For
use in the investigations of the working committee, the method was so revised that the
magnesia content of any cement might be determined rapidly and correctly.

1608

CEMENT

National Bureau of Standards are convenient for this.purpose. One milliliter

of exactly 0.1 N sodium thiosulfate solution is equivalent to 0.000504 g. of MgO.
(c) Potassium Bromate-Potassium Bromide Soluti~n (0.2 N).-Dissohre
20 g. of KBr and 5.57 g. of KBrOa in 200 m!. of waterl and dilute to 1 liter.
Obtain the ratio of this solution to the standard sodium thiosulfate solution as
follows: To 200 m!. of water in a 400-ml. beaker add exa~tly 25 m!. of bromate
solution. Add -20 ml. of HCI (sp.gr. 1.19), stir, and add immediately 10 ml.
of KI solution. Mix well and titrate at once with the sodium thiosulfate
solution until nearly colorless. Add 2 m!. of starch solution, and titrate to the
disappearance of the blue color.
(d) Potassium Iodide Solution (25%).-Dissolve 25 g. of KI in 100 ml. of
water.
(e) Starch Solution.-To 500 m!. of boiling water add a cold suspension
of 5 g. of soluble starch in 25 m!. of water, cool, add a cool solution of 5 g. of
NaOH in 50 ml. of water, add 15 g. of KI and mix thoroughly.
Method.-4. (a) To a 0.500-g. sample of cement in a 400-ml. beaker shall be
added 10 m!. of water and 10 ml. of HCl (sp.gr. 1.19). This shall be heated
gently and any coarse particles ground with the flattened eng. of a glass rod
until de'composition is complete. Hot water shall then be added to make a
total volume of 100 ml.
(b) Three drops of methyl red (0.02 alcoholic solution) shall be added to
the solution and then NH 40H (sp.gr. 0.90) added until the solution is distinctly
yellow. This solution shall then be heated to boiling and boiled for 1 to 2 min.,
then removed from the hot plate and allowed to stand until the precipitate has
settled. The solution shaH then be filtered without delay and the precipitate
washed twice with a hot solution (2%) of ammonium chloride. The precipitate
shall then be transferred with the filter paper to the beaker and dissolved in
10 ml. of HCl (1 : 1). This shall be diluted to about 100 ml. and heated to
boiling. The hydrates shall then be reprecipitated, filtered, and washed as
above. To the filtrate shall be added 1 ml. of NH 40H (sp.gr. 0.90). The
filtrate shall then be heated to boiling and while boiling 25 m!. of a hot solution
(4%) of (NH4)2C204H20 shall be added and the boiling continued for 2 min.
The digestion shall be continued on the steam bath for 10 to 15 min.
(e) The solution shall be cooled to 70 C. and 10 to 25 m!. of the 8-hydroxyquinoline reagent added (Note 1) and then 4 ~l. of NH 40H (sp.gr. 0.90)
added per 100 ml. of solution. The solution shall be stirred on a mechanical
stirring machine for 15 mih. and set aside until the precipit'ate...has settled
(Note 2). The solution shall then be filtered and the precipitate washed with
hot dilute NH 40H (1 : 40). The precipitate shall then be dissolved in 50 to
75 ml. of the hot dilute HCl (1 : 9), the resulting solution diluted to 200 m!.
and then 15 m!. of HCI (sp.gr. 1.19) added. The solution shall then be cooled
to 25 C. and 10 to 35 ml. of standard bromate solution (Note 3) added from
a pipett~. The solution shall then be stirred and aliowed to stand for about
30 sec. to insure complete bromination. To this solution shall then be added
10 m!. of the KI solution. The resulting solution shall be stirred well and then
titrated with the standard sodium thiosulfate solution until the color of the
iodine becomes faintly yellow. At this point 2 m!. of.:Jhe starch solution shall
be added and the solution titrated to the disappeara'nce of the blue color.

CEMENT

1609

(d) Calculation.-The percentage of magnesia in a 0.5-g. sample may be

'calculated as follows:

Percentage of MgO= [(B-C)XAJ X200,

where A =the weight of MgO in grams per milliliter of standard sodium thiosulfate solution;
B = milliliters of standard sodium thios.ulfate solution equivalent to
25 ml. of standard bromate solution, and
C = milliliters of sodium thiosulfate solution required in the titration.
(e) Blank Determination.-A blank determination shall be made, using the
same procedure as outlined above, and using approximately the same amounts
of reagents. The titration results shall be corrected accordingly.
NOTES.-1. An excess of the 8-hydroxyquinoline reagent should be used. If it is
small, the result of the determination of magnesia may be low; on the other hand, if
large, the result may be high. The following guide may be used:
Approximate Content of
1\lagnesia, .%

Approximate Amount of
R<;agent Required, ml.

o to 1.5, ........................................
1.5 to 3.0 ..................................... : . ..
3.0 to 4.5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
4.5 to 6.0 ....................................... "

10
15
20
25

2. The precipitate should be filtered within an hour. Prolonged standing may cause
high result.
3. The amount of the standard bromate solution used may be as follows:
Approximate Content of
Magnesia, %

Amount of Standard
Bromate Solution, ml.

O.to 1 ............................................
1 to 2 ........................... , ............. '"
2to3 ............................................
3to4 ...........................................
4to5 ............................................
5to6 ............................................

10
15
20
25
30
35

PHOSPHORUS

Special Solutions Required: 5. (a) Ammonium Molybdaie.-Dissolve 100 g.

of molybdic acid (85% MoOa) in a mixture of 400 ml. of water and 100 ml. of
NH 4 0H (sp.gr. 0.90) and cool the solution to room temperature. While stirring
constantly, pour this solution very slowly into a cooled mixture of 500 ml. of
water and 400 ml. of HNO a (sp.gr. 1.42). No permanent precipitate should
form, which does not redissolve with stirring. Add 1 drop of a saturated
solution of sodium-ammonium phosphate (NaNH 4 HP0 4 ) to the solution, shake
it thoroughly, and allow it to stand at least 24 hr. before use. Filter portions
of the solution 'as needed.
(b) Magnesia Mixiure.-Dissolve 50 g. of MgCI 2 6H 2 0 and 100 g. of
NH 4 Cl in 500 ml. of water. N:a:ke the solution slightly alkaline with NH 40H
'and allow it to stand at least 12 hr. Filter the solution, acidify it with H,CI
and dilute to 1 liter.
Procedure. 6. (a) Weigh a 1 to 3-g. sample of cement (Note) into a 250-ml.
beaker. Treat the cement with 50 to 75 ml. of water and then with 10 to 15 ml.

CEMENT

1610

of HNO a (sp.gr. 1.42). Boil the mixture until the solution is as complete as
possible. If the cement contains an appreciable ampunt of manganese as
shown by the presence of a red or brown residue, aq<;i a few milliliters of a
potassium nitrite solution (5%) to aid in the soltltion. Boil the solution until
all nitrous fumes are completely expelled, taking care w;>t to allow the volume
of the solution to become so small as to cause the precipitation of gelatinous
silica. Filter the solution into a 500-ml. Erlenmeyer flask and wash the re&idue.
The volume of the solution should now be about 150 ml.

NOTE.-The amount of sample and reagents used depend on the content of phosphorus in the cement. The minimum rcquirements are sufficient if the cement contains
0.5% P 20 5 or more. The maximum amounts are required if the content of P 20 6 is 0.1 %
or less.
(b) Add 10 g. of ammonium nitrate to the solution and heat until the salt
is dissol ved. Heat the solution to 85 0 C. and add 50 to 100 ml. of the solution
of ammonium molybdate. Stopper the flask and shake it vigorously for 5 min.,
then set it aside for 30 min. Filter the precipitate, rinsing the flask five times
and washing the filter paper eight times with a cold solution of potassium
1\\t'.i:'O,t~

\1%>.

(c) Put the original flask under the funnel containing the precipitate.

Pour
25 ml. of diluted NH 40H (1 : 4) over the precipitate and' around the upper
edges of the filter paper un til the preci pita te is entirely dissolved . Wash the
paper three times with cold water, twice with diluted HCI (1 : 1), and finally
three times with cold water. Acidify the solutio!). with HCI and add to it 4 or 5
drops of a saturated solution of citric acid to hold traces of iron, if any, in
solution. Cool the solution to room temperature and transfer it to a 400-ml.
beaker. The volume of the solution should now be about 150 ml. '
(d) To the solution add 25 ml. of magnesia mixture, and then while stirripg
constantly add drop by drop 50 ml. of NH 40H (ep.gr. 0.90) at such a rate that
5 min. are required for the addition. Allow tlle mixture to stand 4 hr. or
longer. Filter the precipitate and wash it a fe'" time~ with diluted NH 4 0H
(1 : 20). Dissolve the precipitate into the original beaker by pouring 25 ml.
of diluted HCI (1 : 1) on the precipitate. Wash the' filter paper thoroughly
with diluted HCI (1 : 20). With the washings jl.nd water, dilute the solution
to 50 to 100 ml. and add to it 1 to 2 ml. of magl].esia mixture. While stirring
constantly, add NH 40H slowly to the solution tl,S before. Allow the mixture
to stand 2 hr. or longer. Filter the precipitate, scrub and.riJIse the beaker, and
wash the precipitate ten 'times with diluted :NH 4 0H (1 :20}.~ ._Ignite the
precipitate in a weighed crucible at 1100 0 to 1200 0 C. to constant weight. Multiply the weight of the magnesium pyrophosphate (Mg 2P 20 7) by 0.6379 to get
the weight of the phosphorus in the sample as p.__O
(e) Blank Correction.-Make a blank determination, using the same procedure as outlined above, and using approximately the same. amo~nt of reagents. Correct the weight of magnesium pyrophosphate (Mg 2P 20 7) accordingly.
(j) If a determination of phosphorus is required, the weight of the ferric
oxide and alumina (AI 20 a+Fe 20a) as obtained by the method described in
Section 12 of the Standard Methods of Sampling and ~esting Portland Cement
(A. S. T. M. Designation: C 77) of the American ~(Jt!iety for Testing Materials
shall be corrected for the amount of the P 20. fOllnd.

CEMENT

1611

MANGANESE

Special Solutions Required: 7. (a) Standard Sodium Arsenite Solution.Dissolve in 100 ml. of water 3.0 g. of N a 2CO S and then 0.9008 g. of As 20 S,
heating the mixture until the solution is as complete as possible. If the solution
is not clear or contains a residue, filter the solution. Cool it to room temperature, transfer it to a volumetric flask, and dilute it to 1 liter. Each milliliter
of this solution is approximately equivalent to 0.003 g. of Mn 20s.
(b) Dissolve 0.58 g. of KMn04 in 1 liter of water and standardize it against
0.0300 g. of sodium oxalate'oxidimetric standard furnished by the National
Bureau of Standards (Standard Sample No. 40) according to the directions
furnished with the sodium oxalate. Put 30.0 ml. of the KMn04 sol~tion in a
250-ml. Erlenmeyer flask. Add 60 ml. of diluted RNO a (1 : 4) and 10 ml. of
a potassium nitrite solution (5%) to the flask. Boil the solution until the
nitrous acid is completely expelled. Cool the solution and add sodium bismuthate and finish by titrating with. the standard solution of As 20 a in the
manner described below for the determination of manganese of cement.
'(c) The value of the standard solution of As 20 a in terms of Mn 20a is:
0.03 XO.236 X30
aXb

where a=the number of milliliters of the solution of KMn04 required by

0.0300 g. of Na2C204, and
b=the number of milliliters of the standard solution of As 20 S required
by 30.0 ml. of the s,olution of KMn04.
Procedure.-8. (a) Weigh a 1.0 to 3.0-g. sample of cement (Note 1) into a
250-nil. beaker and treat it with 5 to 10 ml. of water and then with 60 to 75 ml.
of diluted RNO a (1 : 4). Boil the mixture until the solution is as complete as
possible. Add 10 m!. of a: sodium nitrite solution (5%) to the solution and
boil it until the nitrous acid is completely expelled (Note 2), taking care not to
allow the volume of the solution to become so small as to cause the precipitation
of a gelatinous silica. There may be some separated silica which may be
ignored; but if there is still a red or brown residue, use more solution of sodium
nitrite to effect a complete decomposition and then boil again to expel the
nitrous acid. Filter the solution into a 250-ml. Erlenmeyer flask and wash the
filter paper with water.
NOTES.-l. The amount of cement taken for analysis depends on the content of
manganese, varying from 1 g. for about 1% of Mn 20. to 3 g. for 0.25% or less of Mn.O .
2. When sodium nitrite is added, the expulsion of nitrous acid by boiling must be
complete. If any nitrous acid remains in the solution, it will react with the added sodium
bismuthate and decrease its oxidizing value. If there is any manganese in the cement,
the first small quantity of sodium bismuthate should bring out a purple color.

(b) The solution should have a volume of 100 to 125 ml. Cool it to room
temperature. To the solution add a total of 0.5 g. of sodium bismuthate in
small quantities, with intermittent shaking. After the addition is completed,
shake the solution occasionally for 5 min. and then add to it 50 ml. of cool'
diluted RNO s (1 : 33) which has been previously boiled to expel nitrous acid.
Filter the solution through a pad of ignited asbestos in a Gooch crucible or a
carbon filter with the aid of suction. Wash the residue four times with the cool

, 1612

CEMENT

diluted HNO a (1 : 33). Titrate the, filtrate immediately with the. standard
solution of As 20 a, The endpoint is reached when a yllllow color is obtained
further addition of
free of brown or, pur.ple tints and it does not change upon
I
As 20 3. Calculate the manganese in the cement as Mn2,03.

(c) Blank Correction.-Make a blank determination, using the same procedure as outlined above and using approximately the same amounts of reagents.
Correct the titration results accordingly.
(d) If a determination of manganese is required in addition to a determination of magnesium by Section 16 of Standard Method C 77, manganese shall
be eliminated from the combined filtrates obtained in the alumina and ferric
oxide determination before calcium oxide is p:r;ecipitated (Section 15 of Standard
Method C 77) in the following manner:
Acidify the combined filtrates with HCI and evaporate them to a volume of
about 100 ml. Add 40 ml. of saturated bromine water to the hot solution and
immediately add NH 40H until the solution is distinctly alkaline. Boil the
solution for 5 min. or more, making certain that the solution is at all times
distinctly alkaline. Allow the precipitate to settle, filter and wash with hot
water. Discard the precipitate. Acidify the filtrate' with HCl and boil until
all the bromine is expelled. Make the solution alkaline with NH 40H and
proceed in accorda~ce with Section 15 of Standard Method C 77.

J. LAWRENCE SMITH'S METHOD FOR ALKALIES

For the determination of total alkalies in hydraulic Gement or clinker, weigh
5 grams of the finely pulverized sample and mix with 1 gram NH 4 CI by grinding
the two together in an agate mortar. Add 4 grams C.P. calcium carbonate
and continue the grinding until a thorough mixture has been secured. Place
about U inch of C.P. calcium carbonate in the bottom of a large fusion crucible,
and then add the above mixture. Cover with about U inch of C.P. calcium
carbonate.
Mount the crucible through a hole in an asbestos board about ,Ys inch thick,
so that about U inch of the crucible projects above the board. Cover the
crucible with a well-fitting cover and set a pla.tinum crucible containing some
water on the cover. This serves to keep 'the cover cool, so as to prevent loss
of alkali by volatilization, and water should be kept in the crucible through the
entire ignition period. Heat the crucible at a very/low temperature at first,
and gradually raise the temperature to a full Bunsen flame. This gradual
heating should require about 1 hour, and at no time should the heating be
.strong enough so that the NH 4 Cl fumes can be seen issuing from the crucible.
Finally heat for a full hour at the full flame of the Bunsen burner.
Remove the fusion with hot water and transfer..tO'"'a '375 ml. casserole. Allow to digest for several hours with about 150 ~l. of hot water. If possible,

CEMENT

1613

this digestion should be carried on. over nigh,t. Then boil the solution for.,a
minute, allow to settle and pour the clear liquid through a 12.5 cm. paper into
a 600-ml. beaker. If there are any lumps in the casserole, grind them, with an
agate pestle, add about 50 ml. water,. and boil for a few minutes. Allow to
settle, decant the clear liquid through the same paper, add 30 ml. water and
bring to boiling temperature, allow to settle and decant the clear liquid through
the same paper. Repeat this washing by decantation three times and finally'
wash the residue into the filter. Rinse the casserole thoroughly and wash the
residue in filter once with hot water.
Evaporate the filtrate to about 50 mI., add a few grams of solid ammonium
carbonate and boil for a minute. Filter through an 11 cm. paper into a 250~ml.
beaker, washing six times with cold water containing a little concentrated
ammonia and about 10 gms. ammonium carbonate per liter. To the filtrate
add 10 m!. saturated ammonium oxalate solution and 5 ml. 1-1 ammonia
water and evaporate down to about 10 to 20 ml. Filter through a 9 cm. paper
into a platinum dish, washing five times with hot water. Add 2 or 3 ml.'
concentrated HCl to the solution in the dish, and evaporated to dryness.
Ignite the residue in the dish cautiously over a free flame until the fumes of
ammonium salts are driven off and finally heat at a temperature short of dull
redness. Cool the 'dish, take up the residue in about 50 ml. of hot water, add
1.5 ml. 10% .BaCh solution. Digest on water bath for about 30 minutes,
filter, and wash 5 times with hot water. Evaporate the filtrate to about pO ml.
Add 2 or 3 grams solid ammonium carbonate, boil for a minute, filter, and wash
with ammonia~carbonate wash water .
.To the filtrate add half a gram solid ammonium carbona\,c, 2 ml. saturated
ammonium oxalate solution and a few ml. or 1-1 ammonia water. Evaporate
the solution down to about 10 ml. and filter through a 9 cm. paperinto a clean
platinum dish, washing 5 times with hot water. Add 1 ml. concentrated HCl
and evaporate to dryness, Heat at a temperature short of dull redness until
all ammonium salts are driven off, cool in a dessicator, and weigh. After
weighing, dissolve the salts in about 50 ml. hot water, and filter through a 9 cm.
paper into a 150 ml. beaker. Wash the paper 5 times with hot water, and then
place it in the platinum dish. Ignite until the paper is consumed, cool, and
weigh. The difference between these two weights gives the weight of the
alkalies as sodium and potassium chlorides.
To the solution in the 150-ml. beaker add a few ml. of 1-1 HCI and evaporate to 10 ml. Then add 6 ml. of platinum solution containing 5 g. HiPtCI 6 6H2 0 per 100 ml. for every 0.10 gms. of mixed chlorides weighed, and 3 ml. in
excess. Evaporate on a water bath to a thick syrup. Carry this evaporation
, as far as possible without allowing the solution to go to dryness. The syrupy
solution should solidify on cooling. Then add about 100 ml. 01 85% ethyl
alcohol, stirring the solution thoroughly. The potash in the sample will now
be precipitated as yellow potassium platinic chloride, while the sodium is in
solution.
Allow the precipitate to settle and filter the clear liquid through a weighed
Gooch crucible. Wash the precipitate in the beaker 5 times by decantation
with 80% ethyl alcohol. Transfer the precipitate to the crucible with the
alcohol and scrub the beaker. Wash the crucible and precipitate 6 times with

1614

CEMENT

the alcohol. Dry the crucible for several hQurs at 1050 C., cool, and weigh
as potassium platinic chloride (K 2 PtCI s).
j '
Calculate the K 20 and N a 20 in the sample' as follows:
Weight K2PtCI6XO.193S""weight\K20,
Weight K. 2 PtCI 6 XO.3067 = weight KCI,
Weight mixed chlorides-weight KCI=~eight NaCI,
Weight NaCIXO.5303 = weight Na 2 0.
Make a blank determination, using the same procedure as outlined above
and using approximately the same amounts of reagents and correct the determination of alkalies accordingly.
.

RAPID METHOD FOR CHEMICAL ANALYSIS OF

PORTLAND CEMENT
Before submitting the cement to a 'chemical analysi~ it should be passed
through a No. 50 test sieve to free it from pieces of clinker too large to be quickly

decomposed by acid.
Silica. I 3_Weigh 0.5 gram of cement into a tall form 200-m!. beaker, add
12 to 15 ml. of 60% perchloric acid and cover with a watch glass. Boil the
mixture until the dense white fumes in the beaker clear up and the acid refluxes
down the sides of the beaker for about ten minutes. Take care not to allow
the mixture to evaporate to dryness as it may not be possible to free the silica
from dehydrated impurities. Tak~ care not to a1l9w hot concentrated peichloric acid to come in contact with organic :matt~r as such a contact may
result in an explosion. Allow the mixture to cool to room temperature. Wash
the watch glass and the sides of the beaker with hot water increasing the volume
of the mixture to about 50 ml. and add 5 ml.. hydrochloric acid, specific gravity
1.19. Boil the mixture for one or two minutes: Filter the residue and wash it,
twice wit.h hot dilute hydrochloric acid (1 : 20) and then _;y_ith hot water.
Reserve the filtrate for the determination of ferric oxide and aluinina. Ignite
the residue in a tared platinum crucible, at first slowly with lid on until the
paper is well charred, then more strongly until the residue is white, and finally
at 1000-11000 C. for 45 minutes. Weigh as SiD 2
Ferric Oxide and Alumina.-Follow the A. S. T. M. Standard Method C 77,
page 1602.
;
Calcium Oxide.-To the filtrate (volume 300 to 400 ml.) add 2 or 3 ml.
ammonium hydroxide, specific gravity 0.90. Put in a rubber-tipped stirring
rod to prevent bumping. Bring the solution to a boil and, keeping the boiling
up, add slowly 25 ml. of a boiling 5% solution of ammonium,ilxalate. Continue
13 See also" A Rapid Method for the Determtnatioii'1)f Silica in Portland Cement,"
Earl E. Maczkowski, J. of Res. of the National Bur. of Standards, 16, No.6, June 1936.

CEMENT

1615

the boiling for 5 to 10 minutes and then allow the mixture to stand for 15 to 20
minutes. Filter, receiving the filtrate in a 600-mI. beaker. Wash the precipitated calcium oxalate eight or ten times with hot water. Reserve the
filtrate and washings for the determination of magnesia. Wash the precipitate
back into the original beaker. Add about 250 mI. of hot water and 20 mI. of
sulfuric acid (1 : 1). Heat the solution just short of boiling and titrate it
with standard potassium permanganate solution, adding the filter paper when
the titration is nearly completed. Calculate the per cent of CaO in the sample.
Potassium Permanganate Solution.-Make up the solution on the basis of
5.634 grams to one liter and standardize it with National Bureau of Standards
Standard Sample of sodium Oxalate or with high grade sodium oxalate specially
made by chemical wanufacturers for this purpose. The latter is more economical but should pe checked with a sample from the National Bureau of
Standards. Dissolve 0.7500 gram of sodium oxalate in 250 mI. of hot water
and add 20 mI. of sulfuric acid (1 : 1). Heat the solution just short of boiling
and titrate it with the permanganate solution. If the reading is 62.8 mI., the
standard solution is exactly of the required strength and a reading in analysis
gives directly the pefcentage of calcium oxide in a half gram sample. If the
reading in standardi2lation is not 62.8, it is more satisfactory to use a table of
corrections than to attempt to adjust the strength of the solution because the
solution deteriorates rapidly and forms a sediment of manganese dioxide.
The percentage of calcium oxide is equal to 62.8a+8 where a is the reading in
analysis and 8 is the reading in standardization.
Magnesia.-Acidify the filtrate with hydrochloric acid and boil it down to
about 250 mI. Whep. the solutio~ is cooled to room temperature, add 30 mI.
5% solution of sodium ammonium phosphate and 30 mI. of concentrated
ammonia water. Stir the mixture vigorously until a crystalline precipitate is
formed. Stand the mixture several hours, preferably overnight. Filter the
precipitate and wash it with 5% ammonia water. Ignite it in a tared crucible,
at first slowly until the paper is burned off and then at 1000-1100 C. for 20
to. 45 minutes. Weigh as Mg 2P 207 and calculate the percentage of MgO in
the sample.
Other Components.-Use the A. S. T. M. Standard Method C 77 and
Tentative Method C 114 which may be modified in order to gain speed. 14
NOTE.-This method in 'principle is similar to methods generally used by cement
plant laboratories for routine wor~. The rapid methods ~ive fairly accurate. results
and serve their purpose well, espeClally when they are occasIOnally checked agamst the
official methods of the A. S. T. M. The use of perchloric acid for the dehydration of
silica "was first proposed by Willard and Cake, J. Am. Chern. Soc., 42, 2208 (1920), and
has been improved for tlse on cement by numerous analysts.
14 See also: "DeterrJlination of Sulfuric Anhydride in Portland Cement by Means of
the Wagner Turbidimeter," Robert B. Rudy, J. Res. National Bur. of Standards, 16,
No.6, June 1936. "Determination of Sulfur Occurring as Sulfide in Portland Cement,"
Harry A. Bright, J. Ree. National Bur. of Standards, 16, No.2, February 1937.

CEMENT

1616

RAPID METHOD FOR THE DETERMINATION

bF CARBONATES IN
.
.
.
RAW MATERIALS

Weigh out 0.500 gram of very finely ground sample into a 500-ml. Erlenmeyer flask. Add 50 mL of 0.200 N hydrochloric acitl and boil for about 5
minutes, using a reflux condenser to prevent the loss of acid. Wash the condenser down and cool the flask in a stream of cold water. Add a few drops of
0.5% phenolphthalein indicator and titrate with 0.200 N sodium, hydroxide
solution to the first pink coloration which may be momentary. Shake the
flask vigorously during the titration and add the last portion of the alkali
dropwise. Add x ml. excess of sodium hydroxide. 'If the content of magnesium is high and the red color fades rapidly, it may be necessary to add the
excess alkali in several small amounts with boiling between additions until a
permanent red color is obtained. See below in regard to the required amount of
x. Call the total volume of sodium hydroxide s. Transfer the contents of the
flask to a 200-ml. test tube, add the rinsings, and bring the volume to 100 mL
Heat the mixture to boiling and then allow it to stand until the precipitated
magnesium hydroxide settles. If the precipitate is large, it may be necessary
to filter it on a dry filter paper. Draw 50 ml. of the clear solution with a
pipette. . Put it in the original flask and titrate with the standard hydrochloric
acid to the disappearance of the red color." Call the volume of acid used in this
back titration h.
On the assumption that all the calcium and magnesium in the sample
exist as carbonates, values are calculated thus:
% MgCO a =1.58 (x-2h),
% CaCO a =2.0 (60+2h-s).
This assumption is applicable to high grade limestone but is in general not
so to raw mix used in cement manufacture on account of the insoluble silicates
of calcium and magnesium. In the case of raw mif the factors should be
greater than 1.68 and 2.0 and may be established by an accurate determination
of total magnesium and total calcium in a standard sample of raw mix. The
factors may be calculated thus:
F

t f M CO _ total Mg as MgCO a
ac or <;>r g 3 X _ 2h
.,,~,

_ total Ca as CaCO a
F ac t or for C a CO a 60+2h-s
,

A factor thus obtained is valid only if the raw mix in analysis is similar
to the standard sample of raw mix. If the composition of raw ~ix va.ies,
total magnesium and total calcium should be determined from time to time
and the factors be changed accordingly.
If the content of magnesia is consistently much lower than the limit allowed
by specifications, the determination of magnesium carbonate,may be omitted.
Stop the determination of calcium carbonate at the a.,Epearance of the first pink
coloration. x and h will be zero in the formulae given above for ca~cium carbonate. When magnesium carbonate is determined, its content decides the

OEMENT

1617

amount of excess sodium hydroxide, x, to be. used. It should be so as to make

h equal to 0.2 to 0.5 ml. If insufficient sodium hydroxide is used magnesium
may not be completely precipitated. If too much is used, calcium may be
partially precipitated with magnesium hydroxide.
NOTE.-This method has been in common use for about thirty years.

ANALYSIS OF LIMESTONE, LIME, RAW MIX, ETC.

If the ratio of lime to silica, alumina, and ferric oxide in a sample is equal
to about two or greater, it is convenient to ma.ke a sinter in a 10 ml. platinum
crucible at about 1300 C. for analysis. A blast lamp with illuminating gas
and compressed air is good. If properly made the sinter is easy to remove from
the crucible. If a little material adheres to the bottom of the crucible, it may
be removed with a few drops of hydrochloric acid. The sinter may be dissolved in acid and analyzed like Portland cement.
If the ratio of lime to silica, alumina and ferric oxide is much less than two,
the mixture is liable to fuse and such a fusion is very difficult to remove. To
overcome this danger, it is practical to add a little chemically pure calcium
carbonate to the mixture before the ignition. Its exact weight should be
known only when a determination of calcium is involved.
The determination of ferric oxide may be high due to platinum dissolved
from the crucible. The presence of platinum may be recognized when stannous
chloride is added to reduce ferric ions. Platinum will cause the colorless solution to turn yellow in a few minutes. To eliminate platinum, first remove
silica either by boiling with perchloric acid as in the method for the rapid
analysis of Portland cement, or by a single evaporation in hydrochloric acid to
dryness. Then precipitate ferric oxide and alumina once with ammonia and
dissolve the precipitate in hydrochloric acid after it is thoroughly washed
with hot water. Then fQllow the usual procedure, starting with stannous
chloride.
This method of decomposition may also be applied to silicate rocks when
only an approximate analysis is desir~d. It is necessary to decrease the weight
of sample to about 0.2.gram and to use about one gram of pure calcium carbonate. A good sinter may also be obtained by the ignition of a mixture of
0.5 gram of sample, 0.5 gram of calcium carbonate, and 0.5 gram of sodium
carbonate at about 1000 0 C. However, the presence of sodium is undesirable
as the,precipitated calcium oxalate is contaminated by occluded sodium oxalate
to the extent of about 0.003 gram unless the calcium oxalate is dissolved in
acid' and re-precipitated. For quick work the cal dum is precipitated once
and corrected for both the addition of pure calcium carbonate and the occluded
sodium oxalate, the approximate amount of which is known from experiments.
NOTE.~This

method has been in common use for about thirty years.

1618

CEMENT
DETERMINATION OF FREE LIME IN CEMENT

'Revised Procedure. 1s-The determination is based on the solution, of the

uncombined lime in a hot solution of glycerol and alcohol and the subsequent
titration of the dissolved lime with an alcoholic solution of ammonium acetate
in accordance with the reaction
.
:I
CaO + 2CH aCOO NH4 = (CH 3 COO) 2Ca +:H20 + 2NH 3
Prepare a solution consisting of 1 part by volume of U. S. P. glycerol and
5 parts by volume of absolute ethyl alcohol or anhydrous alcohol denatured i!l
accordance with Formula 3-a or Formula 2-b of the Bureau of Internal Revenue.
To each liter of this solution add 2 ml. of an indicator prepared by dissolving
1 gram of phenolphthalein in 100 'ml. of absolute alcohol. It is essential that
the glycerol-alcohol solution be neutral to the indicator. If the solution is
colorless, add a dilute alcoholic solution of sodium or potassium hydroxide
until the pink color appears, and just remove this by a drop of an alcoholic
solution of ammonium acetate. If the initial color is pink, just remove it by
the alcoholic solution of ammonium acetate.
A standard alcoholic solution of ammonium acetate, approximately 0.2 N,
is required. Dissolve 18 grams of crysta,lline ammonium acetate in 1 liter of
absolute ethyl alcohol or anhydrous alcohol denatured by Formula 3-a or
Formula 2-b. Standardize this solution by titrating against pure calcium oxide.
The calcium oxide should be freshly prepared by calcining pure calcite or
calcium oxalate in a platinum crucible at 900 0 to 1000 0 C. to constant weight.
Ordinary C.P. CaO should not be used. Pulverize a few grams of the calcite
or calcium oxalate to an impalpable powder and.heat as before. Weigh about
0.1 gram and transfer to a 200-ml. Erlenmeyer flask containing 60 ml. of the
neutralized glycerol-alcohol solution. Disperse the calcium oxide in the solution by shaking. Insert a water or air reflux condenser in the neck of the flask,
and boil the mixture for 20 minutes. Then remove the condenser and immediately titrate the solution, while near boiling, with the ammonium acetate
solution. Replace the condenser and boil the solution again for 20 minutes.
Repeat the titrations at 20-minute intervals. .The titration is complete when
no further color appears in the solution during continuous boiling for 1 hour.
Then'calculate the value of the solution in terms of grams of lime equivalent to
each milliliter of ammonium acetate solution.
Grind the cement to be tested to an impalpable powder. Weigh a I-gram
sample and transfer it to a 200-ml. Erlenmeyer flask contl!.Jning 60 ml. of the
glycerol-alcohol solution.' Shake the flask 'to disperse the cement-in the solution.
Attach the reflux condenser and boil the mixture on the hot plate. Titrate,
as in the standardization, at 20-minute intervals, until no further color appp,ars
after boiling for an hour. In the final titrations, if there is doubt of the exact
end point, add a drop of indicator to the quiescent liquid in the flask, 'carefully
observing the point at which the indicator strikes the solution. If no color
appears the end point has been reached. Then calculate the uncombinl!d lime
content of the cement from the known lime value of the amnionium acetate
solution used.
NOTE.i6-The determination of free lime in fresh clinker is 'generally made at a
Portland cement plant for the control of the burning _BI'6cess. A large amount of free
15 Method of W. Lerch and R. H. Bogue, Ind. Eng. cnein., Anal. Ed., 2, 296 (1930);
Paper No. 23 of the Portland Cement. Association at the National Bureau of Stand.ards.
16 By W. C. Hanna.

CEMENT

1619

lime renders cement unsound. Less value is attached to the method when applied to
aged cement because the method does not differentiate between calcium oxide and calcium hydroxide. The latter results from the reaction of calcium oxide lW.d calcium.
silicates and aluminates with atmospheric water and has no harmful effect on the properties of cement.
EXAMPLES OF CEMENT

8i0 2 17 to
Al 20 S
3 to
o to
Fe20S ..................
CaO ................... 57 to
MgO ..................
o to
80 s ...................
1 to
Loss on ignition .... : .... 0.5 to
Insoluble residue ........
o to

27
10
7
70
5
2.5
4
1

--- --- - - - - - - --- ---

21.16' 22.61 20.08 20.54 26.66 24.66 28.91

5.65 5.30 6.54 4.75 2.86 5.60 10.96
2.23 5.23 2.40 6.05 1.76 0.38 4.68
63.36 59.26 65.80 61.50 64.10 61.87 34.64
4.20 3.55 1.40 3.89 1.27 0.90 14.82
1.61 1.81 2.50 1.54 1.43 1.51 1.04
0.97 1.21 1.12 0.70 1.60 3.84 4.50
0.09 0.09 0.12 0.09 0.11

A. Extremes for Portland cement, from cement literature.

B.' 8tandard Portland cement for general use, manufactured in California. Average
bf results by ten analysts, reported by the Working Committee on Methods of
Chemical Analysis, Committee C-1, A. S. T. M.
C. Low heat Portland cement for mass concrete, manufactured in California for use
by the City of Pasadena, California, in Morris Dnm. Average of results by ten
analysts, reported by the Working Committee on Method of Chemical Analysis,
Committee C-1, A. 8. T. M.
D. High -early strength Portland cement for rapid hardening, manufactured in California. Analysis by O. D. Guire, Jr., California Portland Cement Co., Colton,
California.
E and F. Sulfate-resisting Portland cement for resistance to corrosion by sulfate water,
manufactured in California. Analysis by O. D. Guire, Jr., California Portland
Cement Company, Colton, California.
G. White Portland cement manufactured in Pennsylvania. Analysis given in Meade's
"Portland Cement."
H. Natural cement manufactured in Rosendale District; New York. Analysi; by
C. Richardson, quoted in Eckel's "Cements, Limes and Plasters."
EXAMPLES OF CEMENT RAW MATERIALS

~l-I

Si0 2 . . . . . . . 55.52 65.86 8.55 91.28 0.74 0.54 17.04 1.74 15.37
AI 20 s ................. 14.98 9.05 3.08 2.74 0.25 0.11 6.90 0.90 9.13
Fe 20s ............ _... 5.48 3.69 78.67 0.40 0.09 0.07 2.13 0.28 0.25
CaO ................. . 7.30 6.35 4.63 2.10 55.65 32.60 37.53 49.84 25.50
MgO ................. 3.85 2.52 1.08 0.45 0.85 1.54 2.17 1.75 12.35
1.12
43.29
SOs .................. 0.78
Loss on ignition ....... 6.22- 8.32 3.82 2.39 42.40 21.94 32:88 46.01 35.40

A.

Clay found in California.

B. Shale found in California.

Iron ore found in Mojave Desert, California.

Sand found in Nevada.
E. Limestone found in California.
.
F. Gypsum found in California.
Analyses A to F, inclusive, by O. D. Guire, Jr., California Portland Cement Co.,
Coiton, California.
G. Cement rock found in Lehigh Valley, Pennsylvania.
H. Marl found in Indiana. Analysis by S. B. Newberry.
I. Natural-cement rock found in Rosendal District, N ew York. AnalysiS by L. C. Beck.
Analyses G, H, and I given in Eckel's '~Cements, Limes and Plasters."

C.
D.

STANDARD METHODS FOR THE SAMPLING AND

ANALYSIS OF COAL AND. COKE
I

STANDARD METHOD OF SAMPLING COAL

A. S. T. M.

DESIGNATION: D

21-16

It is imperative that every sample be collected and prepared carefully and

conscientiously and in strict accordance with the standard methods described
herein, for if the sampling is improperly done, the sample will be in error, and
it may be impossible or impracticable to take another sample; but if an analysis
is in error, another analysis can easily be made of the original sample.
Gross samples of the quantities designated herein must be taken whether
the coal to be sampled consists of a few tons or several hundred tons, because
of the f?llowing cardinal principle in sampling coal that must be recognized
and understood; that is, the effect of the chance inclusion or exclusion of too
many or too few pieces of slate or other impurities in what, or from what,
would otherwise have been a representative sample will cause the analysis to
be in error accordingly, regardless of the tonnage sampled. For example, the
chance inclusion Or exclusion of 10 lb. too much or too little of impurities in
or from an otherwise representative sample of 100 lb. would cause the analysis
to 'Show an error of approximately 10 in ash percentage, and in heat units of
approximately 10%, whereas for a 1000-lb. sample, the error would be approximately one tenth as large, being the flame whether the sample is collected from
a I-ton lot or from a lot consisting of several hundred tons.
When this method of sampling' is to be employed as a part of any contract
Dr .agreement, t.he ioJJowing provisions sbaJJ be specificaJJy agreed to by tbe
parties to such contract or agreement:
I
(a) The place at which the coal is to be sampled.
(b) The approximate size of the sample required when the standard conditions

do not apply.
(cl The number of samples to be taken or the amount of coal to be represented by each sample when the standard conditions .dp _not apply.

FOR ALL DETERMINATIONS EXCEPT TOTAL MOISTURE

Time of Sampling.-The coal shall be sampled when it is being loaded into
or unloaded from railroad cars, ships, barges, or wagons, or when discharged
1 Under the standardization procedure of the Soci~ty, this method is under the

jurisdiction of the A. S. T. M. Committee D-5 on Coal and Coke. Reproduced through

the courtesy of the American Society for Testing Materials, 260 S. Broad St., Philadelphia, Pa.
2 This method is issued under the fixed designation D 21; the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1915; Adopted, 1916. This method has been approved
as American Tentative Standard by the American Standards Association.
1620

COAL AND COKE

1621

from supply bins, or from industrial railway cars, or grab buckets, or from any
coal-conveying equipment, as the case may be. If the coal is crushed as
received, samples usually can be taken advantageously after the coal has passed
through the crusher. Samples collected from the surface of coal in piles or
pins, or in cars, ships or barges are generally unreliable.
Size of Increments.-To collect samples, a shovel or specially designed tool,
or mechanical means shall be used for taking equal portions or increments.
For slack or small sizes of anthracite, increments as small as 5 to 10 lb. may
be taken, but for run-of-mine or lump coal, increments should be at least
10 to 30 lb.
Collection of Gross Sample.-The increments shall be regularly and systematically collected, so that the entire quantity of coal sampled will be
represented proportionately in the gross sample, and with such frequency that
a gross sample of the required amount shall be collected. The standard gross
sample shall not be'less than 1000 lb., except that for slack coal and small
sizes of anthracite in which the impurities do not exist in abnormal quantities
or in pieces larger than % in., a gross sample of approximately 500 lb. shall be
considered sufficient. If the coal contains an unusual amount of impurities,
such as slate, and if the pieces of such impurities are very large, a gross sample
of 1500 lb. or more shall be collected. The gross sample !3hould contain the
same proportion of lump coal, fine coal, and impurities as is contained in the
coal sampled. When coal is extremely lumpy, it is best to break a proportional amount of the lumps before taking the various increments of a sample.
Provision should be made for the pteservation of the integrity of the sample.
Quantity Represented.-A gross sample shall be taken for each 500 tons or
fraction thereof, or in case of larger tonnages, for such quantities as may be
agreed upon.
Crushing.-After the gross sample has been collected, it shall be systematically crushed, mixed, and reduced in quantity to convenient size for transmission to the laboratory. The sample may be crushed by hand or by any
mechanical means, but under such conditions as shall prevent loss of coal or
the accidental admixture of foreign matter. Samples of the quantities indicated in Table I shall be crushed so that no pieces of coal and impurities will
be greater in any dimension, as judged by eye, than specified for the sample
before division into two approximately equal parts.
The method of reducing by hand the quantity of coal in a gross sample
shall be carried out as prescribed in Section 6, even should the initial size of
coal and impurities be less than indicated in T9:ble 1.
TABLE

Weight of Sample to be
Divided. lb.

1000 or over ................ .

500 ....................... .
250 ....................... .
125 ....................... .
60 ....................... .
30 ....................... .

I
Largest Size of Coal and Impurities
Allowable in Sample before Division. in.

1
~

Yz

-h or to pass a 4760-microD
(No.4) sieve

1622

COAL AND COKE

'~
0

I:lIJ

>::

~
S
~

Ul

'-

""...c:
0

-+"

<l)

:g

"2<:(l

11...,
<:(l

Ul

to

C'I
C'I

ci
~

z'"

",,,2

--"
~:~~

:!:Cc,,,-

=~~=
"."......

.G:= ....

III

..
a

iii

...

:;;
:c

i
f

..'"t:
;:!

i:

COAL AND COKE

1623

Hand Preparation.-The progressive reduc~ion in the weight of the sample

to the qoantities indicated in Table I shall be done by the following methods,
which are illustrated in Fig. 225;
(a) Mixing and Reduction by DiscardiI1g Alternate Shovelfuls.-The
alternate-shovel method of reducing the gross sample shall be repeated until
the sample is reduced to approximately 250 lb., and care shall be observed
before each reduction in quantity that the s/tmple has been crushed to the
fineness prescribed in Table I.
.
The crushed coal shall be shoveled into a ccnical pile (Fig. 225) by depositing each shovelful of coal on top of the precec:iing one, and then formed into
a long pile in the following manner: The saplpler shall take It shovelful of
coal from the conical pile and spread it out in a straight line (Fig. 3A or
8A) having a width equal to the width of the shovel and a length of 5 to 10 ft.
The next shovelful shall be spread directly ov~r the top of the first shovelfui,
but in the opposite direction, and so on back and forth, the pUe being occasionally flattened, until all the coal has been formed into one long pile (Fig.
225~ 4 O[ 9\, The samvler shall thell discard half of this J!ile . J!roceedin~ as
follows:
Beginning on one side of the pile, at either end, and shoveling from the
bottom of the pile, the sampler shall take one shovelful (shovelful No.1,
Fig. 3B or 8B) and set it aside; advancing aloJlg the side of the pile a distance
equal to the width of the shovel, he shall tae a second shovelful (shovelful
No.2) and discard it; again advancing in the same direction one shovel width,
he shall take a third shovelful (shovelful No.3) and add it to the first. The
fourth (shovelful No:4) shall be taken in a like manner and discardej:l, the fifth
(shovelful No.5) retained, and so on, the sampler advancing always in the
same direction around the pile so that its size will be gradually reduced in a
uniform manner. When the pile is removed, about half of the original quantity of coal should be contained in the new pile formed by the alternate shovelfuls which have been retained. (Fig. 5A or lOA show the retained halves,
and Fig. 5B or lOB the rejected halves.)
(b) Mixing and Reduction by Quartering.__,.-After the gross sample has been
reduced by the alternate-shovel method to approximately 250 lb., further
reduction in quantity shall be by the quartering method. Before each quartering, the sample shall be crushed to the fineJless specified in Table I.
Quantities of 125 to 250 lb. shall be mixeo thoroughly by coning and reconing (Figs. 12 and 13); quantities less than 125 lb. shall be placed on a
suitable cloth, measuring about 6 by 8 ft., mixed by raising first one end of
the cloth and then the other (Fig. 18, 24 pr ~O), so as to roll the coal back
and forth, and after being mixed thoroughly shall be formed into a conical
pile by gathering together the four corners of the cloth (Fig. 19, 25 or 31).
The quartering of the conical pile shall be done as follows:
The cone shall be flattened, its apex being pressed down vertically with
a shovel, or board, so that after the pile has been quartered, each quarter will
contain the material originally in it. The fl::J.ttened mass, which shall be of
uniform thickness and diameter, shall then be m.arked into quarters (Fig.
14, 20, 26 or 32) by two lines that intersect at nght angles directly under
a point corresponding to the apex of the origir~al cone. The diagonally opposite quarters (BB in Fig. 16, 22, 28 or 34) shall then be shoveled away and

1624

COAL AND COKE

discarded and the space that they occupied brushed Clean. The coal remaining shall be crushed, mixed, coned, and quartered successively until the sample
is reduced to the desired quantity.
(c) The 30-lb. quantity (Fig. 29) shall be crushed\to 136 -in. size or to pass
a 4760-micron (No.4) sieve, mixed, coned, fiattenea. and quartered. The
laboratory samples shall include all of one of the quarters, or all of two opposite quarters (Fig. 34), as may be required. The laboratory sample shall be
placed immediately in a suitable container and sealed in such a manner as to
preclude tampering.
Mechanical Preparation.-Only such mechanical means as will give equally
representative samples shall be used in substitution for the hand method of
preparation herein standardized.
FOR THE DETERMINATION OF TOTAL MOISTURE
Collection of Moisture Sample.-Tl).e special moisture sample .shall weigh
approximately 100 lb., and shall be accumulated by placing in a waterproof
receptacle with a tight-fitting and waterproof lid small equal parts of freshly
taken increments of the standard gross sample. The accumulated moisture
sample shall be rapidly crushed and reduced.'mechanically or by hand to about
a 5-lb. quantity, which shall be immediately placed in a container and sealed
air-tight and forwarded to the laboratory without delay.
Use of Standard Sample.-Only when equally representative results will be
obtained shall the standard gross sample be used instead of the special moisture
sample fol' the determination of total moisture.

STANDARD METHOD OF SAMPLING COKE FOR ANALYSIS

A. S.' T. M. DESIGNATION: D 346-36 4

It is imperative that every sample be collected and prepared carefully and
conscientiously and in strict accordance with the standard procedure described
herein, for if the sampling is improperly done, the sample will be in error, and
it may be impossible or impracticable to take another sample; whereas if an
analysis is in error, another analysis can readily be 4tade of the original sample.
3 Under the standardization procedure of the Society, this method is under the jurisdiction of the A. S. T. M. Committee D-5 on Coal and Coke. Reproduced through
the courtesy of the American Society for Testing Materials, 260 S. Broad St. Philadelphia,
Pa.
I
4 This method is issued under the fixed designation D ~46; the final number indicates
the year of original adqption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1932; Adopted in Amended Form, 1935.

COAL AND COKE

1625

Gross samples of not less than the quantities designated herein must be
taken whether the coke to be sampled consists of a few tons or several hundred
tons in order to minimize the effect of the chance inclusion or exclusion of too
many or too few pieces of non-representative material.
FOR ALL DETERMINATIONS EXCEPT TOTAL MOISTURE
Time of Sampling.-(a) The coke shall be sampled while it is being loaded
into or unloaded from railroad cars, ships, barges, or, trucks, or when discharged
from supply bins, railroad cars, grab buckets, or other coke-conveying equip- .
ment.
(b) Samples collected from the surface of coke in piles, bins, cars, ships or
barges are, in general, unreliable because of size segregation and should not be
used for determining conformance to specifications unless the purchaser and
seller so agree. In case it is necessary to collect a sample of coke from the
surface of loaded shipments, nine equal increments shall be taken about 1 ft.
below the surface. The nine sampling points shall be located as shown in
Fig. 226.
Size of Increments.-To collect samples, a shovel or specially designed tool
or mechanical means shall be used for taking equal portions or increments.
The gross sample'shall consist of not less than 25 nor more than 50 increments
of approximately equal quantity, except that when samples are collected from
the surface of loaded shipments the gross sample shall consist of 9 increments
of approximately equal quantity.
Collection of Gross Sample.-:The increments shall be regularly and sys- .
tematically collected, so that the entire quantity of coke sampled will be
represented proportionately in the gross sample, and with such frequency that
a gross sample of the required amount will be collected. The standard gross
sample shall be not less than the quantities given in the following table:
\

MINIMUM WEIGHTS OF GROSS SAMPLES OF COKE

Run-of-oven, Blast-furnace, Foundry, Water-gas, and any coke containing a

range of size of pieces made from uncrushed or co~rsely rushed coal, b
except coke breeze ................................................ ,
Run-of-oven, Blast-furnace, Foundry, Water-gas, and any coke containing a
range of size of pieces made from crushed coal," except coke breeze ......
Closely-sized coke made from uncrushed or coarsely crushed coal b free of coke
breeze ...........................................................
Closely-sized coke made from crushed coal c free of coke breeze ...... , ......
Coke breeze (all passing a 72 ()r %;-in. square-hole sieve) ...................

500 lb."
250 lb."
250 lb."
125 lb."
125 lb.

a In case the pulverization of the coal is not known, take quantities designated for
coke made from uncrushed or coarsely crushed coal.
b More than 10% on a 74'-in. square-hole sieve.
c Not less than 90% passing through a U-in. square-hole sieve.

Quantity Represented.-A gross sample shall be taken for each 250 tons or
fraction thereof, or in case of larger tonnages, for such quantities as may be
agreed upon. Each lot of coke arising from a different source or known to be
of different quality or size shall be sampled separately.
Crushing.-The entire gross sample shall be crushed, mixed, and reduced
in quantity to convenient size for transmission to the laboratory. The sample

COAL AND COKE

1626

shall be crushed preferably by means of jaw or roll crushers, or in case mechanical crushing means are not available, the cokE) samplf) shall be crushed on a
chilled-iron or hard-steel plate by impact of a tamper, hard bar or sledge, avoiding all rubbing action, as otherwise the ash content may! be materially incre;tsed
by the addition of iron from the sampling apparatus, evJn though hardened iron
or steel is used. The crushing shall be done under s~ch conditions as ~hall

...
L.. length of car
~:L <:-;L
L
,1d~ ~h~~:L
1d~ I-<-:L
iO~ ~.L ~.L
l(i~ ~:L
Th~ I"<-;L
~~ ~iO~
J_

. 10'"

10

16""

"

=tJ
is:1~
IJ/

~'!'

.:0'"
.-g
::: ....

_~- is:~f
is:l~

is:1~

tl\
FIG. 226.-Location of Sampling Points from Exposed Surface of Car.

prevent loss of coke or accidental admixture of foreign matter. Samples of

the quantities indicated in Table I shall)e crushed so that no pieces of coke
and impurities will be greater in dimension, as judged by eye, than specified
for the sample before division into two approximately equal parts.
TABLE I.-WEIGHTS OF COKE SAMPLES WITH CORRESPONDING CRUSHING SIZES.

Largest Size of
Coke and Impurities
Allowable in Sample
Before. Division

Weight of Sample to be
Divided, lb.

250 or over. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 in.

l~g '. '. :::::::::::::::::::': :::::::::::::::::::! ::::::::: ~ :~:

30 ....................................................
7.:i: in.
.
.
~

The method of reducing by hand the quantity of coke in a gross sample

shall be carried out as described in Section 6, eVen should the initial size of coke
and impurities be less than indicated in Table I.
Mixing and Reduction.-T'he progressive reduction i"ii the.weight of the
sample to the quantities indicated in Table I shall be done by the following
methods which are described and illustrated in Plate V of the Standard Method
of Sampling Coal (A. S. T. M. Designation: D 21) of the American Society for
Testing Materials."
'
(a) The alternate-shovel method of reducing the gross sample sl;tall be
repeated until the sample is reduced to approximately 125 lb., and care shall
be observed before each reduction in quantity that the sample has been crushed
to the fineness specified in Table 1.
The crushed coke shaH be shoveled into a conical pile (Eig. 225, 2 or 7) 5
by depositing each shovelful of coke on top of the preceding one, and then
1\

Page 1622.

_,-'

COAL AND COKE

1627

formed into a long pile in the following manner: The sampler shall take a
shovelful of coke from the conical pile and spread it out in a straight line
(Fig. 3A or 8A) having a width equal to the width of the shovel and a length of
5 to 10 ft. The next shovelful shall be spread directly over the top of the first
shovelful, but in the opposite direction, and so on back and forth, the pile being
occasionally flattened, until all the coke has been formed into one long pile (Fig.
3B or 8B). The sampler shall then discard half of this pile, proceeding as follows:
Beginning on one side of the pile, at either end, and shoveling from the
bottom of the pile, the sampler shall take one shovelful (shovelful No.1, Fig. 4
or 9) and set it aside; advancing along the side of the pile a distance equal to the
width of the shovel, he shall take a second shovelful (shovelful No.2) and '
discard it; again advancing in the same direction one shovel width, he shall take
a third shovelful (shovelful No.3) and add it to the first. The fourth (shovelful
No.4) shall be taken in a like manner and discarded, the fifth (shovelful No.5)
retained, and so on, the sampler advancing always in the same direction around
the pile so that its size will be gradually reduced in a uniform manner. When
the Rile is removed, about half of the original quantity of coke should be contained in the new pile formed by the alternate shovelfuls which have been
retained. (Fig. 5A or lOA show the retained halves, and Fig. 5B or lOB the
rejected halves.)
(b) After the gross sample has been reduced by the alternate-shovel method
to approximately 125 lb., further reduction in quantity shall be by the quartering method. Before each quartering, th~ sample shall be crushed to the fineness
specified in Table 1.
Quantities of 60 to 125 lb. shall be mixed thoroughly by coning and reconing
(Figs. 12 and 13) ; quantities less than 60 lb. shall be placed on a suitable cloth,
measuring about 6 by 8 ft., mixed by raising first one end of the cloth and then
the other (Fig. 18, 24 or 30), so as to roll the coke back and forth, and after
being mixed thoroughly shall be formed into a conical pile by gathering together the four corners of the cloth (Fig. 19, 25 or 31). The quartering of the
conical pile shall be done as follows:
The cone shall be flattened, its apex being pressed down and worked outward
by means of a shovel, so that after the pile has been quartered, each quarter '
will contain the material originally in it. The flattened mass, which shall be of
uniform thickness and diameter, shall then be marked into quarters (Fig. 14,
20, 26 or 32) by two lines that intersect at right angles directly under a point
corresponding to the apex of the original cone. The diagonally opposite
quarters eBB in Fig. 16, 22, 28 or 3Ll) shall then be shoveled away and discarded
and the space that they occupied brushed clean. The coke remaining shall be
crushed, mixed, coned, and quartered successively until the sample is reduced
to approximately 30 lb.
.
~
(c) The 30-lb. quantity shall be crushed to 7!'-in. size, mixed, coned, flattened and quartered. The laboratory samples shall include all of one of the
quarters, or all of the two opposite quarters, as may be required. The laboratory satnple shall be placed immediately in a suitable container and sealed in
such a manner as to preclude tampering.
Reduction by Mechanical Means.-Only such riffles or other mechanical
means as will give equally representative samples shall be used in substitution
for the hand method of reduction in quantity herein standardized.

COAL AND' COKE

1628

FOR THE DETERMINATION OF TOTAL MOISTURE

Sample for Total Moisture.-For determining tbtal moisture, a special

moisture sample weighing approximately 50 lb. shall be ,taken, except in the case
of coke breeze, in which case the moisture sample shall weigh approximately
25 lb. The moisture sample shall be taken when the1coke is being load~d or'
unloaded as the case may be, and shall be accumulated by placing in a water~
proof receptacle with a tight-fitting lid small equal parts of the freshly-taken
increments of the standard gross sample as previously described. The special"
moisture sample, without any preliminary crushing, shall preferably be dried
to constant weight at a temperature of not less than 104 nor more than 200 0 C. 6
In case it is impracticable to dry the entire sample, the following procedures
may be used:
Far Coke Appearing Dry.-The special moisture sample shall be crushed
rapidly to Yz-in. size and reduced mechanically or by hand to about a 5-lb.
quantity, which shall be immediately placed in a container and sealed air-tight
and forwarded to the laboratory without delay.
Far Coke Appearing Wet.-The special moisture sample shall be spread on
tared pans, weighed, and air-dried or dried In a warm place or on a warm or
heated surface until the coke appears dry, and weighed again. The sampling
shall be completed as described in the preceding paragraph for coke appearing
dry. This loss in weight divided by weight of sample, multiplied by 100, is the
percentage of air-drying loss and should be corrected as follows for the moisture
found in the sample sent to the laborat~ry:
Percentage
total moisture
(as received)

'percentage
100-percentage of air-drying loss X moisture in air100
dried sample

percentage

+ of air-drying
loss.

Report.-Since in the report of the analysis, a brief description should be

given of the method of taking the sample, by such characteristic expressions as
" Belt Sample," " Top-of-car," etc., information as to ,how the sample is taken
shall be included in the description which accompani~s the sample sent to the
laboratory.
'
6 Experiments made at the U. S. Bureau of Mines have shown that results checking
within 0.5% are obtained between these temperature limits. See A. C. Fieldner and
W. A. Selvig, "The Determination of Moisture in Coke," U. S. Bureau of Mines Technical
Paper No. 148 (19~7).

1629

COAL AND COKE

STANDARD METHODS OF LABORATORY SAMPLING AND

ANALYSIS OF COAL AND COKE 7
A. - S. T. M.

DESIGNATION:

D 271-33; s

PREPARATION OF LABORATORY SAMPLES

Apparatus: (A) ~or Coal: The apparatus used for sampling coal shall consist
of the following:
(a) At"r'Drying Oven.-An oven for air-drying wet samples. A suggested
form is shown in Fig. 227. Such an oven is not essential but is economical
where many samples are to be dried,s

...;!L

FIG. 227.-Drier for Coarse Samples. The outlet for air at the'top may be connected with a chimney or any other device which will furnish a suitable draft. (Bulletin
No.9, Geological Survey of Ohio, p. 312.)
7 Under the standardization procedure of the Society, these methods are under the
jurisdiction of the A. S. T. M. Committee D5 on Coal and Coke.
These methods comprise a revision and combinati.on of the former Standard Methods
of Laboratory Sampling and Analysis of Coal (A. S. T. M. Designation: D 22-24) and
of Laboratory Sampling and Analysis of Coke (A. S. T. M. Designation: D 37-24)
which methods were accordingly discontinued in 1927.
8 These methods are issued under the fixed designation D 271; the 'final number
indicates the year of original adoption as standard or, in the case of revision, the year
of last revision. Adopted, 1916, 1918; Revised, 1921, 1924, 1927, 1929, 1930, 1933.
These methods have been approved as American Standard by the American Standards
Association.
9 For det/tils of air-drying oven, see F. M. Stanton and A. C. Fieldner and W. A.
Selvig, "Methods of Analyzing Coal and Coke," U. S. Bureau of Mines Technical Paper
No.8, p. 2 (1929). Reproduced through the courtesy of the American Society for
Testing Ma~rials, 260 S. Broad St. Philadelphia, Pa.

1630

COAL AND COl\:E

(b) Pans for Air-Drying Wet Samples.-Galvanized iron pans 18 by 18 in.

by 1.5 in. in depth.
I.
(~) Balance or Solution Scale.-A balance or Bc.aJenfJ,ving fJ, cfJ,pfJ,city of 5 kg.
sensitive to 0.5 g. for weighing the galvanized iron pans with samples.
(d) Crusher.-A jaw crusher suitable for c~Ushing boarse samples to pass a
I
4760-micron (No.4) sieve.
(e) Grinder.-A roll crusher or coffee-mill type lof grinder suitable for
reducing the material passing a 4760-micron (No.4) sieve to pass an 840micron (No. 20) sieve. To reduce the moisturEj loss while crushing, a coffeemill type of grinder should be entirely enclosed and have an enclosed hopper
and receptacle capable of holding 10 lb. of coal.
(f) PUlverizer.-An Abbe ball mill, planeta.ry disk crusher, chrome-steel
bucking board, or any satisfactory form of pulv~rizer for reducing the material
passing an 840-micron (No. 20) sieve to pass a 250-micron (No. 60) sieve.
The porCelain jars for the ball mill should be all proximately 9 in. in diameter
and 10 in. high. The flint pebbles should be SlUooth, hard and well rounded.
(g) Large Rijjie Sampler.-A large riffle sampler with Yz or %-in. divisions
for reducing the coal passing the 4760-micron (No.4) sieve to 10 lb. (see Fig.
228).
I
(h) Small Rijjie Sampler.-A small riffle ~ampler'With Kor %-in. divisions
for reducing the material passing the 840-micrOll (No. 20) and the 250-Iilicro~
(No. 60) sieves to a laboratory sample (see Fig. 229).
(i) Sieve.-An 8-in. 250-micron (No. DO) sieve with cover and receiver.
NOTE.~The sieve designations employed are those of the Standard Specifications
for Sieves for Testing Purposes CA. S. T. M. Designation: E 11) of the American Society
for Testing Materials,10 and all sieves shall conform to the detailed requirements of
these specifications. See p. 1329.

(j) Containers.-Samples in which the moistllre content is important should

always be shipped in moisture-tight containers. A galvanized-iron or tin can
with an air-tight friction top or a screw top which is sealed with a rubber gasket
and adhesive tape is best adapted to this purpOSe. Gl~ss fruit jars sealed with
rubber gaskets may be used, but require very careful p'acking to avoid breakage
in transit. Samples in which the moisture contl'int is of no importange need no
special Pl'otection from loss of moisture.
(B) For Coke: The apparatus used for sampling coke shall consist of the
following:
.....
(a) Pans for Total Moisture Delermination.-Galvanized "iron pans 24 by
24 in. by 4 in: in depth.
(b) Balance or Solution Scale.-A balance or scale having capacity of
10 kg. sensitive to 1 g. for weighing the galvani?;,ed-iron pans with samples.
(c) Crusher.-As described under Section A.
(d) Roll-Crusher.-A hard-steel roll crusher suitable for reducing the
material passing a- 4760-micron (No.4) sieve tl) pass an 840-micron (No. 20)
sieve.
(e) Pulverizer.-An Abbe ball mill, hard-steel roll crusher or hard-steel
diamond mortar for reducing the product pa~sing an 840jilicron (No. 20)
sieve to Dass a 250-micron (No. 60) sieve. The por~elain j~rs for the ball mill

10 193::$ Book of A. S. T. M. Standards, P~rt II, p. '1244:

COAL AND COKE

FIG.

FIG.

228.-Large Rime Sampler.

229.-8maU Riffle Sampler.

1631

1632

COAL AND COKE

should be approximately 9 in. in diameter and 10 in. high. The flint pebbles
should be smooth, hard and well rounded. The roll!! of the hard-steel roll
crusher should revolve at the same speed.
(f) Large Riffle S"ampler.-Same as in Sectibn A (g) f9r reducing coke passing
the 4760-micron (No.4) sieve to 10 lb. (see Fig; 228). ' I
(g) Small Riffle Sampler.-T,he same as Section.A (h) (see Fig. 229).
(h) Sieve.-The same as Section A (i).
I

(i) Containers.-The same as Section A (.i).

U) Oven, Stove or Hot Plate.-An oven, stove or hot plate for drying coke
samples in the determination of total moisture. If an oven is used it should
have openings provided for natural ventilation and should be capable of being
regulated between 104 and 200 0 C. If the coke is dried on a stove or hot plate a
thermometer should be placed in it, and care exercised that the temperature
does not exceed 200 0 C. at any point in the pan of coke.
Method of Sampling: (A) For Coal Appearing Dry: (a) If the sample is
coarser than will pass completely a 4760-micron (No.4) sieve and larger in
amount than 10 lb., quickly crush it with the jaw crusher to pass a 4760-micron
(No.4) sieve and reduce it on the larger rime sampler to 10 lb.; 11 then crush it
at once to pass an 840-micron (No. 20) sieve !by passing through rolls or an
enclosed grinder, and take, without sieving,. a 50-g. total moisture sample,
immediately after the material has passed through the crushing apparatus.
This sample should be taken with a spoon from various parts of the product
passing an 840-micron (No. 20) sieve, and should be placed directly in a rubberstoppered bottle.
(b) Thoroughly mix the main portion of the sample, reduce on the small
rime sampler to about 200 g., and pulverize to pass a 250-micron (No. 60) sieve
by any suitable apparatus without regard to loss of moisture. After all the
material has been passed through the 250-micron (No. 60) sieve, mix and divide
it on the small rime sampler to about 50 g. Transfer the final sample to a 4-oz.
rubber-stoppered bottle. Determine moisture in both the 250-micron (No. 60)
sieve sample and the 840-micron (No. 20) sieve sample in accordanCE) with
Sections 6, 7, and 8 under the Determination of Moisture.
(c) Calculation.-Calculate the analysis of the coal passing the 250-micron
(No. 60) sieve which has become partly air-dried during sampling, to the drycoal basis, by dividing each result by 1 minus its content of moisture. Compute the analysis of the coal" as received" from the dry-coal analysis by multiplying by 1, minus the total moisture found in the sample passing_l!n 840-micron
(No. 20) sieve.
.
(B) For Coal Appearing Wet: (a) Spread the sample on tared -pans, weigh,
and air-dry at room temperature, or in the special drying oven, shown in Fig.
227, at 10 to 150 C. above room temperature, and weigh again. The \frying
should be continued until the loss in weight is not more than 0.1 % per hour.
Cbmplete the sampling as described in Section 3 for iiry coal.
(b) Calculation.-Correct the moisture found in the air-dried sample
passing an 840-micton (No. 20) sieve to total moisture "as received," as
follows:
I
11

5 lb.

If the sample is crushed to pass a 3360-micron (No,j).sieve it may be reduced to

COAL AND COKE

1633

100 - percentage of air-drying loss

'. .
.
.
100
,
X (percentage of mOIsture III alr-dned sample passmg
an 840-micron (No. 20)..sjeye)+(percentage of air-drying loss) = (total. moisture "as
received")
.
(c) Calculate the analysis to "dry-coal" and "as-received n bases as
described in Section 3 for-dry coal, using for the" as-received" calculation the
total moisture as found by the formula in Paragraph (b) in place of the moisture.
found in the coal passing an 840-micron (No. 20) sieve.

NoTEs.-Freshly mined or wet coal loses moisture rapidly on exposure to the air of
the laboratory, hence the sampling operations between opening the container and taking
the total-moisture sample palOsing an 840-micron (No. 20) sieve must be conducted with
the utmost dispatch and with minimum exposure to air.
The accuracy of the method of preparing laboratory samples should be checked
frequently by resampling the rejected portions and preparing a duplicate sample. The
ash in the two samples should not differ more than the following limits:
No carbonates present ................................... 0.4%
Considerable carbonate and pyrite present .................. 0.7%
Coals with more than 12% ash, containing considerable carbonate and pyrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1.0%

(C) Ball-mill Method jor Coal Appearing Wet or Dry: (a) This method of
sampling 12 does not require a total moisture sample of the coal passing an
840-micron (No. 20) sieve as do methods A and B. The coal is first air-dried
to bring it to a condition of approximate equilibrium with the air to minimize
moisture change during the preparation of the sample for analysis. After
air-drying all operations are performed with the utmost dispatch to prevent
moisture change. Fine grinding of the sample is done in an air-tight ball mill.
Total moisture is computed from the air-drying loss and the residual ,moisture
in the sample prepared for analysis.
(b) Spread the sample on tared pans, weigh, and air-dry in the special
mositure oven shown in Fig. 227, at 10 to 15 C. above room temperature until
the loss in weight between two successive weighings, made 6 to 12 hours apart,
does not exceed 0.1 % per hour. Record the loss in weight as" air-drying loss."
(c) Immediately after the last weighing, quickly crush the entire sample by
means of a roll-crusher adjusted so the product will pass an 840-micron (No. 20)
sieve. Then, without sieving, quickly reduce the coal on the small riffle sampler
to about 200 g. Put this 200-g. portion at once in the porcelain jar (I-gal.
capacity) of an Abbe ball mill, sealed air-tight by means of a rubber gasket, and
grind to pass a 250-micron (No. 60) sieve. The jar shall be slightly more than
half full of well-rounded flint pebbles, about ~ in. in diameter, and rotated .
at 60 to 75 r.p.m. Bituminous coals require from 10 to 20 minutes and anthracite about 1 hour of rotation to be pulverized to the designated fineness.
(d) When pulverization is complete, pour the contents of the jar on a Yz-in.
square-mesh sieve, and separate the sample from the pebbles by shaking the
sieve over a rubber mixing-cloth. Brush carefully the pebbles and inside of
the jar to remove any adhering pulverized coal. A convenient size of sieve is
one approximately 17 <in. square. Then quickly pass the sample through a
12 This method of sampling is used by the U. S. Bureau of Mines, see F. M. Stanton,
A. C. Fieldner and W. A. Selvig, "Methods of Analyzing Coal and Coke." U. S. Bureau
of Mines Technical Paper No.8, p. 2 (1929).
.

COAL AND COKE

1634

250-micron (No. 60) sieve. To minimize moisture change, keep the sieve
covered while the sample is being sieved. 'Usually !lome coarse particles,
amounting to 0.2 to 0.4% of the sample, remain on the sitive. As these particles
are likely to consist largely of ash-forming substances, reduce them on a chromium-steel bucking board to pass the sieve and then add'1{0 what passed through
the sieve. Quickly mix the sample on a rubber mixing-cloth and reduce .to
about 50 g. by passing through the small riffle sampler. 'Put this 50-g. portion
which constitutes the laboratory sample, in a rubber-stoppered, wide-mouthl
4-oz. bottle. To minimize moisture change, the mixing and reducing of the
pulverized sample after removal from the ball mill shall be done with utmost
dispatch. The total time elapsing from the opening of the porcelain jar to the
stoppering of the laboratory sample shall not exceed 3 minutes;
Calculation.-As all the analytical determinations are made on the air-dried
sample passing a 250-micron (No. 60) sieve, calculate the analysis to "asreceived" and " dry-coal" bases from the analysis of the air-dried coal in the
following manner:
CALCULATION FROM ANALYSIS OF "AIR-DRIED" COAL TO COAL" AS RECEIVED"

All figures expressed in per cent

.
" as reCClve
. d" =mOlS
.t
(loo-air-drying loss) + . d . 1
M Olsture
,ureX
100
alr- rymg 08&.
.
.
.
/ (100-air-drying loss)
VolatIle matter "as receIved" = volatIle matterX
100
.
.
FIxed carbon "as

recel~ed"=fixed

carbon X

(lOO-air-drying loss)
100

h (100-air-drying loss)
"
. d"
Ash as receIve =as X
100
.

. d"
If X (100-air-drying loss)
Sulfur " as receIve
=su ur

100'
.
I

.
(100-air-drying loss)'
.
.
Hydrogen" as receIved" = hydrogen X
100
-jJ1/9 alr-<irymg loss.
C arb on

"

b
(100-air-drying loss)
. d"
as receIve =.car onX.
100
..

.
. .
(100-air-drying loss)
.
NItrogen "as recClved"=mtrogenX
,
100
.
(100-air-drying loss)
.
.
'.
Oxygen "as recClved" = oxygen X
100
' +8/9 alr-<irymg loss.
.
.
.
(100-air-drying loss)
CalorIes" as receIved" = calOrIes X
100
~

,
CALCULATION FROM ANALYSIS OF "AIR-DRIED" COAL TO "DRY COAL"

All figures expressed in per cent

Volatile matter in "dry coal" = volatile matter X 100 100. t
.
-IDOlS ure
Fixed carbon in "dry coal" = fixed carbon X 100 100. t ~~-mols ure

COAL AND COKE

Ash 10 "dry coal" = ash X

1635

100.
.
100-mOlsture

Sulfur in "dry coal" = sulfur X lQO 100. t .

-mens ure
Hydrogen in "dry coal" = (hydrogen-l/9 moisture) X 100 100. t .
-mOlS ure
Carbon in "dry coal" = carbon X 100 100. t .
,
-molS ure
Nitrogen in

"~ry coal" = nitrogen X 100 -mOlS

100. t
.
ure

Oxygenin"drycoal"=(oxygen-8/9moisture)X

100. t .
1oo -molS ure

Calories in "dry coal" = calories X 100 100. t

.
-mOlS ure

(D) For Coke: (a) Total Moisture Determination.-Dry the entire sample
receiveod at the laboratory, without any preliminary crushing, to constant weight
at a temperature of not less than 104 nor more than 200 0 C.13
Calculate the loss in weight to percentage of moisture, which shall constitute
the total moisture in the coke as received at the laboratory.
The alJowable difference in duplicate determinations by the same analyst is
0.5%.
(b) Reduction of Sample.-Crush the dried sample mechanically with a jaw
or roll crusher, or by hand on a chilled iron or hard-steel plate by impact of a
hard bar or sledge, avoiding all rubbing action, as otherwise the ash content will
be' materially increased by the addition of iron from the sampling apparatus,
even though hardened iron or steel is used. Continue the crushing until all the
sample passes through a 4760-micron (No.4) sieve, mix and reduce on the large
riffle sampler to not less than 51b.; again crush the 5-lb. sample sufficiently fine
to pass an 840-micron (No. 20) sieve; mix and reduce on the small riffle sampler
to 200 g. Transfe! this 200-g. portion to the porcelain jar of an Abbe ball mill
and pulverize to pass a 250-micron (N o. ~O) sieve. When pulverization is
complete, pour the contents of the jar on a %-in. screen and separate the sample
from the pebbles by shaking the screen. Pass all the material through a 250micron (No. 60) sieve, pulyerizing any coarse particles by impact in a hardsteel diamond mortar, and mix with the remainder of the sample. Reduce the
sample through the small riffle sampler to about 50 g. and transfer to a rubberstoppered glass bottle.
(c) In case a ball mill is not available for fine grinding, quarter the 5-lb.
sample passing an 840-micron (No. 20) sieve to 200 g. and pulverize to pass a
250-micron (No. 60) sieve by means of a hard steel roll crusher in which the
rolls revolve at the same speed, or by impact hl a hard-steel diamond mortar.
The use of rubbing surfaces such as a <iisk pulverizer or a bucking board is
never permissible for grinding coke.
13 Experiments made at the U. S. Bureau of Mines'have shown that results checking
within 0.5% are obtained between these temperature limits. See A. C. Fieldner and
W. A. Selvig, "The Determination of Moisture in Coke," U. S. Bureau of Mines Technical Paper No. 148 (1917).

1636

COAL AND COKE

NOTE.-The accuracy of the method of preparing laboratory samples should bp

checked ~requently by resampling the reje~ted portions and prepar,ng a duplicate sample.
The ash III the two samples should not differ more than 0.4%. .

FIG. 230.-Toluene or Glycerin and Water Oven for Determining Moisture (Technical

Paper No.8, Bureau of Mines, p. 9 (1929.

METHODS OF ANALYSIS
DETERMINATION OF MOISTURE

(For Coal and Coke)

Apparatus: (a) Mdisture Oven (For Coal).-For determining the moisture

of coal, the oven shall be so constructed as to have a unifofJn temperature in all
parts and a minimum of air space. It may be of
~
F
the form shown in Fig. 230. Provision shall be
made for renewing the air in the oven at the rate
of two to four times a minute, with the air dried
by passing it through concentrated H 2S0 4
(For Coke).-For determining'the .moisture of
coke, an ordinary drying oven with openings for
natural air circulation and capable of -temperature regulation between limits of 104 and 110 c.
FIG. 231.-Capsule for Use in . may be used.
.'
.
Determining Moisture.
,(b) Capsules with Covers.-A conyement form,
which allows the ash determination to be made
on the same sample, is a porcelain capsule, Ys in. deep and 1~ in. in diameter; or a fused silica capsule of similar shape. This shall be used with a wellfitting flat aluminum cover, illustrated in Fig. 231.
Platinum crucibles or glass capsules with ground-glass caps' may also be
used. They should be as shallow as possible, cQIlSiStent with convenient
I

han~ling.

COAL AND COKE

1637

Method: (A) For Coal pr Coke Passing 250-micron (No. 60) Sieve: Ca)
Heat the empty capsules under the conditions at which the sample is to be
dried, place stopper or cover on capsule, cool over concentrated H 2S0 4 Csp.gr.
1.84) for 30 minutes, and weigh. Dip out with a spoon or spatula from the
sample bottle approximately 1 g. of the sample; put this quickly into the capsule, close, and weigh at once.
(b) An alternate procedure (more open to error), after transferring an
amount slightly in excess of 1 g., is to bring to exactly 1 g. in weight (0.5 mg.)
by quickly removing the excess weight of the sample with a spatula. The
utmost dispatch must-be used in order to minimize the exposure of the sample
until the weight is found.
After removing the covers, quickly place the capsules in a preheated oven
(at 104 to 1100 C.) through which passes a current of air dried by concentrated
H 2S0 4 (the current of dry air is not necessary for coke). Close the oven at
once and heat for 1 hour. Then open the oven, cover the capsules quickly and
place them in a desiccator over concentrated H 2 S0 4 When cool, weigh.
The percentage of moisture in the sample passing a 250-micron CN o. 60)
sieve shall be used to calculate the other results to a dry basill.
(B) For Coal Passing 840-micron (No. 20) Sieve: Use 5-g. samples, weighed
with an accuracy of 2 mg., and heat for 1;\1 hours i the procedure is otherwise
the same as with the sample passing a 250-micron (No. 60) sieve as described in
Section 8.
NOTE.-For methods of greater accuracy see Report of Subcommittee II on Moisture, of the Joint Committee on AnalysL'3 of Coal, Proceedings, Am. Soc. Testing Mats.,
Vol. XIV, Part I, pp. 419-423 (1914).
.
DETERMINATION OF ASH

(For Coal and Coke)

Apparatus: (a) Gas or Electric Muffle Furnace (for Coal).-For determination of ash of coal, the muffle shall have good air circulation and be capable of
having its temperature regulated between 700 an4 750 C.
(b) Gas or Electric Muffle Furnace or Meker Burner (for Coke).-For
determination of ash of coke, the muffle shall have good air circulation and be
capable of having its temperature regulated to not exceed 950 0 C.
. (c) Porcelain Capsules.-Porcelain capsules, Ys in. deep and I%; in. in
diameter, or similar shallow dishes or platinum crucibles.
Method: (A) For Coal: Place the porcelain capsules containing the dried
coal from the moisture determination in a cold muffle furnace, or on the hearth
at a low temperature, and gradually heat to redness at such a rate as to avoid
mechanical loss from too rapid expulsion of volatile matter. Finish the ignition
to constant weight (0.001 g.) at a temperature betwee~ 700 and 750 0 C.
Cool in a desiccator, and weigh as soon as cold.
0

NOTEs.-Before replacing the capsules in the muffle for ignition to constant weight,
the ash should be stirred with a platinum or nichrome wire. Stirring once or twice before
the first weighing hastens complete ignition.
The result obtained by this method is "uncorrected" ash. For" corrected" ash
see the preliminary report.i4 The actual mineral matters in the original coal are usually
very different in weight and composition from the weight of the "uncorrected" ash.
14 Report on Fixed Carbon and A&h, Proceedings, Am. Soc. Testing Mats., Vol. XIV,
Part I, p 426 (1914).

1638

COAL AND COKE

(B) For Coke: Place the capsules containing the dried coke from the moisture determination in a muffle furnace or over a burnell, and heat to redness at
such a rate as to avoid mechanical loss. Finish the ignition to constant weight
(0.001 g.) at a temperature not exceeding 950 0 C. Cool in a desiccator and
weigh.
' \ "
NOTEs.-Befote replacing the capsules in the muffle for ignition to constant weight
the ash should be stirred with a platinum or nichrome wire. Stirring once or twice
before the first weighing hastens complete ignition.
Test the ash for unburned carbon, by moistening it with alcohol; any carbon remaining will show as black particles.
DETERMINATION OF 'vOLATILE MATTER

(For Coal and Coke)

Apparatus: (a) Platinum Crucible with Closely Fitting Cover (for Coal).The crucible shall be of not less than 10 nor more than 20-ml. capacity; of not
less than 25 nor more than 35 mm. in diameter; of not less than 30 nor more than
35 mm. in height.
(b) Platinum Crucible with Closely Fitting Cover (for Coke).-The crucible
shall be of lO-m!. capacity, with capsule cover having thin flexible sides fitting
down into crucible. Or the double-crucible method may be used, in which the
sample is placed in a 10 or 20-ml. platinum. crucible, which is then covered with
another crucible of such a size that it will fit closely to the sides of the outer
crucible, and its bottom will rest 7i to Y2 in. above the bottom of the outer
crucible.
(c) Vertical Electric Tube Furnace; or a Gas or Electrically Heated Muffle
Furnace (for Coal or Coke).-The furnace may be of the form as shown in
Fig. 232. It shall be regulated to maintain a temperature of 950 0 C. 20 C.
in the cr:ucible, as shown by a thermocouple kept in the furnace. If the determination of volatile matter is not an essential feature of the specifications under
"
which the coal or coke is bought, a Meker burner may be used.
Method: (A) For Coal and Coke, Usual Method: (a) Weigh 1 g. of the
sample in a weighed platinum crucible, close with a cover, and place on platinum
or nichrome-wire supports in the furnace chamber, which shall be at a temperature of 950 0 C. 20 C. Mter the more rapid uischarge of volatile matter has
subsided, as shown by the disappearance of the luminous flame, or in the case of
coke after heating 2 or 3 minutes, tap the cover lightly to more perfectly seal
the crucible and thus guard against the admission of air. After heating for
exactly 7 minutes, remove the crucible from the furnace and, without disturbing
the cover, allow it to cool. Coke should be cooled in a desiccator. Weigh
as soon as cold. The loss of weight minus moisture equals the volatile matter.
"(b) Modificationfor Sub-Bituminous Coal, Lignite, and Peat.-Mechanical
losses are incurred on suddenly heating peat, sub-bituminous coal, and lignite;
therefore, they must be subjected to a preliminary' gradual heating for 5 minutes; this is best done by playing the flame of a burner upon the bottom of the
crucible in such a manner as to bring about the discharge of volatile matter at
a rate not sufficient to cause sparking. After the preliminary heating, transfer
the crucible to the volatile-matter furnace and }wat: Ior 6 minutes at 950 0 C.
as in the regular method described in Paragraph (a) .

COAL AND COKEl

1639

NOTEs.-The cover 'should fit closely enough so thllt the carbon deposit from bi.,
tuminous and lignite coals does not burn away from the under side.
Regulation of temperature to within the prescribed limits is important.

FIG. 232.-Electric Tube Furnace for Determining Volatile Matter. For 1l0-vcIt
alternating current, 60 ft. of Nichrome Wire, No. 17. B. & S. gage will give the required
temperature. :rhe temperature must be controlled by an external resistance. (U. S.
Bureau of Mines Technical Paper No.8, p. 12 (1929.

(B) For Coal.and Coke, Using Meker Burnt!r: Weigh 1 g. of the sample in a weighed platinum crucible and close with a cover or, in the case
of coke, with another crucible. Place in the flame of a No.4 Meker burner,
having approximately an outside diameter at the top of 25 mm. and
giving a flame not less than 15 cm. high. The temperature should be

1640

COAL AND COKE

950 0 C. 20 C., as determined by placing a thermopouple through the perforated cover, which for this purpose may be of nickel I or asbestos. The junction of the couple should be placed inc contact with the center of the bottom
of the crucible; or the temperature may be indicated by the fusion of pure
K 2Cr04 in the covered crucible (fusion of K 2Cr04, 9~8 C.).15 The crucible
shall be placed in the flame about 1 cm. above the top of the burner and the
heating continued 7 minutes. Where the gas pressure is variable it is well to
use a U-tube attachment to the burner.
DETERMINATION OF FIXED CARBON

(For Coal and Coke)

Calculate fixed carbon as follows:

100- (moisture+ash+volatile matter) = percentage of fixed carbon
DETERMINATION OF SULFUR

(For Coal and Coke)

A. Eschka Method

Apparatus: (a) Gas or Electric Muffle Furnace, or Burners.-For igniting

the sample with the Eschka mixture and for igniting the BaS04.
(b) Porcelain or Platinum Crucibles or Capsules.-Porcelain capsules, Ys in.
deep and 1% in. in diameter, or porcelain crucibles of 30-m!. capacity, high or
low form, or platinum crucibles of similar size shall be used for igniting the
sample with the Eschka mixture.
(c) Porcelain, platinum, alundum, or silica crucibles of 10- to 15-m!.
capacity, shall be used for igniting the BaS04.
Solutions and Reagents: (a) Barium Chloride.-Dissolve 100 g. of BaCh
2H 2 ()ein 1000 ml. of distilled water.
(b) Saturated Bromine Water.-Add an excess of/bromine to 1000 ml. of
distilled water.
..
(c) Eschka Mixture.-Thoroughly mix two parts (by weight) of light
calcined MgO and one part of anhydrous Na 2C0 3. Both materials should be
as free as possible from sulfur.
(d) Methyl Orange.-Dissolve 0.02 g. in 100 ml. of hot distilled water and
filter.
,
(e) Hydrochloric Acid:-Mix 500 ml. of HCI (sp.gr. 1.l9);and 500 ml. of
distilled water.
.
(f) Normal Hydrochloric Acid.-Dilute 80 ml. of HCI (sp.gr. 1.19), to
1 liter with distilled water.
,
(g) Sodium Carbonate.-A saturated solution, approximately 60 g. of
crystallized or 22' g. of anhydrous N a2C0 3 in 100 ml. of distilled water.
(h) Sodium Hydroxide Solution.-Dissolve lOD g. in 1 liter of' distilled
water. This solution may be used in place of the Na2C03 solution.
Method:PrepaTation of,Sample and Mixture.-Thoroughly mix on glazed
paper 1 g. of the sample and 3 g. of Eschka mixture. 'rf{m~fer to a porcelain
capsule, Ys in. deep and I%; in. in diameter, or a PJ>rcelain crucible of 30-ml.
.-'

15

U. S. Bureau of Mines Reports of Investigations, SeriarNo. 2917 (1929).

COAL AND COKE

1641

capacity, high or low form, or platinum crucible of similar size, and cover with
about 1 g. of Eschka mixture.
.
(a) Ignition.-On account of the amount of sulfur contained in artifidal
gas, the crucible shall be heated over an alcohol, .gasoline or natural. gas flame
as described in Paragraph (b), or in a gas or electrically heated muffle, as described in Paragraph (c) for coal and in .Paragraph (d) for coke. The use of
artificial gas for heating the sample and Eschka mixture is permissible only when
the crucibles are heated in a muffle.
..
,
(b) Heat the crucible, placed iIi a slanting position on a ~riangle, over a
very low flame to avoid rapid expulsion of the volatile matter, which tends to
prevent complete absorption of the products of combustion of the sulfur .. Heat'
the crucible slowly for 30 minutes, gradually increasing the temperature and
stirring after all black particles have disappeared, which is an indication of the
completeness of the procedure.
(c) (For Coal).-Place the crucible in a cold muffle and gradually raise the
temperature to 8000 C. , 25 0 C. in about 1 hour. Maintain this maximum
temperature for about 1Yz hours.
(d) (For Coke).-Place the crucible in a warm muffle and gradually raise
the temperature to 8000 C. 25 0 C. in about 30 minutes. Maintain this
maximum temperature until on stirring all black particles have disappeared.
(e) Subsequent Treatment.-Remove and empty the contents into a 200ml. beaker and digest with 100 ml. of hot water for Yz to % hour, with occasional
stirring. Filter and wash the insoluble matter by decantation. After several
washings in this manner, transfer the insoluble matter to the filter and wash
five times, keeping the mixture ~ell agitated. Treat the filtrate, amounting
to about 250 ml., with 10 to 20 ml. of saturated bromine water, make slightly
acid with fICl and boil to expel the liberated bromine. Make just neutral to
methyl orange with N'aOH or Na 2 CO a solution, then add 1 ml. of N HCl.
Boil again and add slowly from a pipette, with constant stirring, 10 ml. of a 10%
solution of BaCI 2 2H2 0 .. Continue boiling for 15 minutes and allow to stand
for at least 2 hours, or preferably over night, at a temperature just below boiling.
Filter through an, ashless filter paper and wash with hot distilled water until a
AgNO a solution shows no precipitate with a drop of the filtrate. Place the
wet filter containing the precipitate of BaS04 in a weighed platinum, porcelain,
silica or alundum crucible, allowing a free access of air by folding the paper
over the precipitate loosely to prevent spattering. Smoke the paper off
gradually and at no time allow it to burp with flame. After the paper is practically consumed, raise the temperature to approximately 9250 C. and heat to
constant weight.
The residue of MgO, etc., after leaching, should be dissolved in HCI and
tested with great care for sulfur. When an appreciable amount is found this
should be determined quantitatively. The amount of sulfur retained is by
no means a negligible quantity.I6
(f) Blanks and Corrections.-In all cases a correction must be applied either
(1) by running a blank exactly as described above, using the same amount of all
reagents that w\lre employed in the regular determination, or more surely (2)
by determining a known amount of sulfate added to a solution of the reagents
16

Journal Am. Chemical Soc., 21, 1125 (1899).

COAL AND COKE

1642

aft~r these have been put through the prescribed series of operations. If this
latter procedure is adopted and carried out, say, once a ~eek or whenever a new
supply of a reagent must be used, and for a series of solutions covering the range
of sulfur content likely to be met with in the samplesj it is only necessary to
add to or subtract from the weight of BaS04 obtained from a sample, whatever
deficiency or excess may have been found in the appropriate" check" in order
to obtain a result that is more certain to be correct than if a " blank" correction
as determined by the former procedure is applied. This is due to the fact that
the solubility error for BaS04, for the amounts of sulfur in question and the .
conditions of precipitation prescribed, is probably the largest one to be considered. BaS04 is soluble 17 in acids and even in pure water, and the solubility
limit is reached almost immediately on contact with the solvent. Hence, in
the event of using reagents of very superior quality or of exercising more than
ordinary precautions, there may be no apparent" blank," because the solubility
limit of the solution for BaS04 has not been reached or at any rate not exceeded.
(g) Calculate sulfur content as follows:
P

t
f Sulf _ (Weight of BaS04-blank)X13.74
ercen age 0
urWeight of sample

NOTE.-As shown in the preliminary report,l~ the Atkinson and sodium-peroxide

methods give results in close agreement with the Eschka method. Register 19 has
shown that if 5% of nitrogen is present in the gases contained in the bomb calorimeter
the sulfur of a sample is almost completely oxidized to H 2S0 4 and the washings of the
calorimeter may be used for the determination of sulfur.
.

B.

Bomb Washing Method 20

Method: (a) Ignition.-Sulfur is determined in the washings from the

oxygen-bomb calorimeter following the calorimetric determination. The type
of bomb, amount of water in the bomb, oxygen pressure, and amount of sample
taken shall be the same as specified under the calorimetric determination. The
bomb shall stand in the calorimeter water for not less than 5 minutes after firing.
(b) Subsequent Treatment.-Remove the bomb frefm the calorimeter water
and open the valve carefully so as to allow the gases'to escape at an approximately even rate so the pressure is reduced to at,mospheric in not less than one
minute. Bombs equipped with valves other than needle valves, such as compression valves, shall be provided with a device ISO the valve can be controlled
to permit a slow and uniform release of the gases. Open the bomb and examine
the inside for traces of unburned material or sooty deposit. ~ If-the~ are found,
the determination shall be discarded. Wash carefully all parts -of the interior
of the bomb, including the tray, with a fine jet of distilled water containing 1 m!.
per liter of a saturated solution of methyl orange, until no acid reaction isobserved. It is essential to wash through the valve opening in the case of
bombs equipped with compression valves, or other types of valves with large
openings, as considerable spray may collect in such :valve openings.
Journal Am. Chemical Soc., 32, 588 (1910); 33, 829 (1911).
Journal of Industrial and Engineering Chemistry, 5, 5 (1913).
19 Ibid., 6, 812 (1914).
I.
20 W. A. Selvig and .A:. C. Fieldner, "Check Determinations of Sulfur in Coal and
Coke by the Eschka, Bomb-washing, and Sodium Peroxige-Fiision Methods," Industrial
.
-and Engineering Chemistry, 29, 729--733 (1927).
17

18

COAL AND COKE

1643

Collect the washings in a 250-ml. beaker and titrate with standard alkali
solution to obtain the "acid correction" for the heating value, as specified
under the calorimetric determination. Add 1 ml. of NH 40H (sp.gr. 0.90), heat
the solution to boiling, and filter through qualitative filter paper. Wash the
residue and filter paper thoroughly five or six times with hot distilled water.
To the filtrate and washings, amounting to about 250 ml., add 1 ml. of saturated
bromine water and sufficient HCI to make it slightly acid. Boil the solution to
expel the excess bromine. _Adjust the acidity, precipitate, and determine the
sulfur as specified under the Eschka method.

C.

Sodium Peroxide Fusion Method

21

Apparatus: (a) Combustion Bomb.-The Parr coal sulfur bomb, or its

equivalent, shall be used. The bomb shall have an inner surface which is not
attacked by the chemicals on ignition of the charge.
(b) Reagents.-Sulfur-free powdered sodium peroxide (Na 202), powdered
c.p. potassium perchlorate (KCl0 4 ) or potassium chlorate (KClO a), and powdered c.p. benzoic acid shall be used.
Method: (a) Place 1 g. of KCI0 4 or KClO a in a dry sulfur bomb and break
up any lumps that occur. Add a 0.5-g. sample and mix thoroughly with a
"glass rod. Then add one measure (about 15 g.) sodium peroxide, close the
bomb and mix thoroughly by shaking. In the case of cokes or anthracites, or
coals excessively high in ash which fail to ignite or fuse properly (as indicated
by the fusion being honey-combed in appearance), add 0.3 g. of benzoic acid
to the bomb at the time the chlorate and sample are added. It'should be noted
'that a mixture of potassium chlorate and organic matter alone produces a
mixture of extremely explosive properties. One of the important functions
of the sodium peroxide is to provide a diluent, thus slowing down the reaction,
so care should be taken that it is not omitted in the charge. Potassium perchlorate is fully equal if not superior to the KCIO a, and is without turbulence
in its reaction.
Fasten the cover securely to the bomb and ignite the charge by applying a
sharply pointed flame from a blast lamp to the bottom of the bonib for a brief
period, or by electri.c ignition, according to the type of bomb used. Place
the bomb inside a piece of steel pipe when the charge is ignited to prevent
possible injury to the operator if the bomb should burst. Allow one minute
for complete combustion to take place after ignition, then cool under the tap or
in a vessel of water.
Remove the cover from the bomb, place the bomb on its side in a 400-ml.
beaker and wash off the cover with a fine jet of hot distilled water. Place a
watch glass over the beaker and cautiously add about 100 ml. of hot distilled
water. After the contents of the bomb have dissolved, remove and rinse it
carefully with the distilled water. Add slowly, concentrated HCI to the
neutral point, th~n add 1 to 2 ml. of the acid in excess. Filter through qualitative filter paper into a 600-mI. beaker and wash thoroughly five or six times
with distilled water. Dilute the filtrate to approximately 400 ml. and precipi21 W. A. Selvig and A. C. Fieldner, "Check Determinations of Sulfur in Coal and
Coke by the Eschka, Bomb-washing, and Sodium Peroxide Fusion Methods," Industrial
and Engineering Chemistry, 29, 729-733 (1927).

1644

COAL AND COKE

tate the sulfur with barium chloride and determine as specified under the
Eschka method.
I
(b) Blank Correction.-A blank correction shall be applied by running a
blank on the reagents used.
I,
DETERMINATION OF PHOSPHORUS IN ASH

(For Coal and Coke)

Solution Required: Molybdate Solution.-Dissolve 65 g. of molybdic acid

(85%) in a mixture of 143 ml. of NH 4 0H (sp.gr. 0.90) and 1"42 ml. of water.
Add this solution slowly, with constant stirring, to 715 ml. of HNO a (sp.gr.
1.20). If the solution is cloudy, add two drops of a solution of ammonium
phosphate (20%) and allow the precipitate which forms to settle. Filter the
solution into bottles and, if necessary, re-filter just before using. If the molyb.
dic acid used is 100%, 56 g. should be taken instead of 65 g.
Method: Method No.I,/or all Cases: Add to the,ash from a 5-g. sample, in a
platinum crucible, 10 ml. of concentrated HNO a and 3 to 5 ml. of HF. Evaporate the liquid, ignite the residue, and fuse with 3 g. of Na 2 CO a If unburned carbon is present, mix 0.2 g. of NaNO a with the carbonate. Leach the
melt with water and filter the solution. Ignite the residue, fuse with Na 2 CO:j,
alone, leach the melt with water and filter the solution. Just acidify with
HNO a the combined filtrate held in a flask, add 3 to 5 ml. of concentrated HNO a
in excess, and concentrate to a volume 01 100 ml. Add 6 g. of NH 4 NO a, bring
the temperatur~ of the solution to 80 C., add 50 ml. of molybdate solution
and shake the flask for 10 minutes. When the precipitate has settled, filter
and wash the precipitate, until free from acid, with a KNO a solution (2%).
Place the filter paper with the precipitate in the flask, add 25 ml. of recently
boiled distilled water, and macerate the filter paper with a glass stirring rod.
Add a measured excess of standard NaOH solution and agitate the solution to
completely dissolve the precipitate. Add 3 drops of phenolphthalein solution
as an indicator and titrate the excess NaOH with a stan1ard HNO a solution.
The alkali solution may well be made equal to 0.00025 g. of phosphorus per
milliliter, or 0.005% for a 5-g. sample. Such a solution would be 0.926 of
0.2 N.22
NOTE.-The advantage of the use of HF in the initial attack of the ash lies in the
resulting removal of silica. Fusion with alkali carbonate is necessary oJ: the elimination
of titanium, which if present and'not removed, will contaminate the phosphomolybdate
and is said 1(0 sometimes retard its precipitation.

Method No.2, when Titanium is Low: When titanium is so low as to offer

no objection, decompose the ash 'as described above, Method No.1, but
carry the evaporation only to a volume of about 5 mI. Dilute the solution
with ~ater to 30 mI., boil and filter into a flask. If tlie washings are turbid
pass them again through the filter. Ignite the residue in a platinum crucible,
fuse with a little Na 2 CO a, dissolve the melt in HNO a and add the solution, if
clear, to the main one. If not clear, filter. The fusion of the residue may be
dispensed with in routine work on a given coal or coke if it is certain that th'e
residue is free from phosphorus. Add NH 4 0H until .a _.slight precipitate ensues .
22

Ulmann and Buch, Chemical Engineer, 10, 130 (1909). -

COAL AND COKE

1645

Add concentrated HNO a to just dissolve.,the precipitate, then add 3 to 5 ml.

of the acid in excess. Heat the solution, which should have a volume of a):lOut
100 mi., to 80 C. and add 50 ml. of molybdate solution. Shake the flask for
10 miJ;mtes, filter, and determine the phosphorus as described in Method No.1.
DETERMINATION OF FUSIBILITY OF ASH

(For Coal and Coke)

Apparatus: (a) Standard Furnace.-Gas-fired furnaces such as the coal-ash

fusion furnace of the Denver Fire Clay Co., the No.3 Melter's Furnace of the
American Gas Furnace Co., or their equivalent shall be used.
These gas-fired furnaces are especially suitable for fusion determination in>
that the burners are arranged on a tangent near the base of the furnace, thus
producing a rotary flame which completely surrounds the crucible in which
the ash cones are placed. The whirling flame heats the crucible uniformly and
when the furnaces are operated with an excess of gas over air, a reducing atmosphere is maintained in the crucible, which condition gives the lowest point at
which the ash fuses. > Either natural or artificial gas may be used. Air should
be supplied at uniform pressure.

FIG. 233.-Denver Fire-Clay F,urnace and No.3 Melter's Furnace, with Accessories.
The stock design of the No.3 Melter's Furnace should be modified by pro~
viding the upper cylinder with two holes in the side; a 2-in. observation hole
with its center 4 in. from the top of the cylinder (excluding cover plate), and
a I-in. thermocouple hole 90 to the right of the observation hole; the bottoms
of these holes being in the same horizontal plane. The coal-ash fusion furnace
as furnished by the Denver Fire Clay Co. is provided with suitable observation

]646

COAL AND COKE

and thermocouple holes. The interior of the furnaces are cylindrical' and
approximately 7 in. in diameter. Counter-balanced sh~et-iron canopies connected with telescopic flues to an exhaust system are placed over the furnaces
for conducting the hot gases out of the room. Two fufnaces and' accessories
I
are shown in Fig. 233.
(b) Refractory Crucibles.-The interior of the fusion furnace as arranged
for making a test is shown in Fig. 234. Crucible a is a Corundite crucible
about 3 in. in inside diameter and about 472 in. in height outside, with a wall
thickness of about 72 in. The crucible is provided with a cover, b, and shall
have two holes in the side, an observation hole 2 in. in diameter and a thermocouple hole 1 in. in diameter, 90 to the right of the observation hole. The
bottoms of the two holes shall be in the same horizontal plane. The crucible is
supported by a Corundite support, c, so that the two holes in the crucible are in
line with the correllPonding holes in the furnace wall. The ash cones, d, are
supported on Corundite disk e so that they are visible through the observation
hole of the crucible and furnace wall.

FIG. 234.--Section of Furnace Arranged for Fusion Tests.

(c) Observation Hole.-A fused silica, alundum, or refractory porcelain

tube, f, 1 Ys in. in external diameter and 6 to 7 in. lpng, is cemented in the
2 in. observation hole of the furnace, the inner end being flush with the inside
furnace wall and the other end projecting out of the furnace. A brass sleeve
carrying a single thin glass window, made from ordinary soda glass, shall be
slipped on the outer end of the observation tube to prevent the escape of
burning gas, whi~h would interfere with a convenient Qbservati~n of the cones.
When observations at high temperatures are made,. th~-cones should be viewed
through a piece of colored glass.

1647

COAL AND COKE

(d) Blowing Tube.-At furnace temperatures above 1000 C. it is very

difficult to observe the ash cones. An open refractory porcelain blowing tube,
g, Fig. 234, about U in. in inside diameter shall be inserted in the thermocouple
hole, the inner end being flush with the inside of the furnace wall. The blowing
tube shall be connected to the compressed air line by means of rubber tubing
and the air shall be let into. the tube at the time of observation by means of a
pinch-cock, thus momentarily cooling the ash cones and rendering them visible.

,.
....~
.'

FIG. 235.-Brass Cone Mold.

(e) Brass Cone Mold.-A brass cone mold for making ash cones, % in.
high and U in. at each side of the base, which is an equilateral triangle, is
illustrated in Fig. 235 .
. (f) Pyrometer. 23-The temperature measurements shall be made with a
thermocouple o! platinum and platinum-rhodium used in conjunction with a
high-resistance millivoltmeter or potentiometer; or with an optical pyrometer
of the disappearing filament type, which has been calibrated to be accurate to
10 C. up to 1400 C. and to 15 C. from 1400 to 1600 C. If a theqnocouple
is used it should be protected from the furnace gases by a glazed porcelain tube
and so placed in the furnace that the hot junction is in the immediate vicinity
of the cones. The pyrometer equipment shall be checked frequently by
mounting small pieces of pure gold-and nickel in the same manner as the cones'
With a strong reducing atmosphere, 1452 C. should be obtained for the melting
point of nickel. and 1063 C. for the gold. The pyrometer equipment should
23 For detailed information concerning the use of pyrometers see "Pyrometric Practice," U. S. Bureau of Standards Technologic Paper No. 170 (1921).

1648

COAL AND COKE

also be standardized from time to time thr.ough the temperature range for
which it is used, by a suitably equipped standardizingJ~boratory, such as that
of the U. S. Bureau of Standards.
'
--(g) If temperature measurements are made by sightin'g an optical pyrCimeter
through the glass window of the observation tube the~ follov~ing corrections
shall.be added to the observed temperatures to correct ~or absorption of light
by the glass window:
\

CORRECTION TO OBSERVED TEMPERATURE FOR ABSORPTION OF LIGHT BY A SINGLE

CLEAN GLASS WINDOW 24

Observed Temperature,
Deg. Cent.

Correction to be Added.
Deg. Cent.

800 .............. :....................................... 5

1000 ....................... ............................... 8
~200 ...................................................... 10
1400 ...................................................... 13
1600 ...................................................... 16

Preparation of Ash: Spread out 50 to 100 g. of the sample passing a 250micron (No. 60) sieve on a 6-in. fire-clay roasting dish, and completely convert
to ash in a muffle furnace at a temperature of 800 0 to GOO C. Transfer 5 to
10 g. of this ash to all agate mortar 25 and grind to pass a 74-micron (No. 200)
sieve. Then place the ash in a silica or porc~lain capsule, % in. deep and I%, in.
in diameter, and ignite for a period of twwhours in a current of oxygen, at a
temperature of 800 0 to 850 0 C. This ignition is made to insure complete and
uniform oxidation of the ash.
'
Preparation of Cones: Moisten the ignited ash with a .10% dextrin solution
and work into a plastic mass with a spatula. Mold the plastic material into
small triangular pyramids %' in. high and 7i in. wide at the side of the base.
The pyramids are made by firmly pressing the plastic material with a steel
spatula into a brass mold of'the dimensions mentioned, the mold being similar
to that shown in Fig. 235. Strike off the surface smooth and remove the cone
from the mold by applying a small knife blade at the bse. Mount the con~s
when dry in a refractory base composed of a mixture of equal parts of kaolin
and calcined alumina. Moisten the base mixture to make it workable, and
spread a part of it out on a sheet-iron plate. Then mount the cone in a vertical
position in a small hole made in the base, and put base material into the hole
around the bottom of the cone to fill the crevices and make the'{)!;lne stand firmly.
Usually five cones are mou~ted in one base in the manner shown in Fig. 236.
Dry the sheet-iron plate with the test piece on a hot plate. Ignite the cones at
a dUll red heat for 30 minutes in an open muffle to remove the carbonaceous
material.
Method of Heating: Place the test piece in the muffle crucible of the furnace
in the position shown in Fig. 234, place the loosely' fitting cover b, on the
crucible, and ignite the gas. It is necessary to let the gas burn about 10 minutes
to heat the furnace parts before the large cover plate of the furnace is replaced;
otherwise the flame is apt to blowout. During this time increase the flow of

U. S. Bureau of Standards Technologic Paper No. 170,.p. 117 (1921).

2S A mechanical agate-mortar grinder will save time wher~many determinations are
made.
24

COAL AND COKE

1649

gas and air sufficiently to cause the combustion to take place just above the
tuyeres and yet maintain a yellowish flame at least 6 in. above the opening in
the furnace cover plate. While such a flame is maintained above the furnace,
gradually increase the temperature by a suitable adjustment of gas and air to
800 C., then reduce the rate of heat increase to not less than 5 C. and not
more than 10 C. per minute. Maintain this rate until the end of the test.
It is also important that the 6 in. reducing flame be maintained at the furnace
vent throughout the test, if possible,
and at all events up to a temperature of
1450 C. Temperatures above 1450
C. require larger proportions of air to
gas; however, a strongly reducing atmosphere is not so essential at the FIG. 236.-Typical Forms of Cones Fused
in the No. ~ Melter's Furnace.
higher temperatures, as refractory
.
ashes, owing to their low iron-oxide
content, are only slightly affected by oxidizing or reduciBg atmospheres.
After a test has been completed, turn the supply of gas and air off gradually
to avoid cracking the muffle crucible.
The softening temperature is defined 'as the temperature at which the cone
has fused down to a spherical lump, as shown in cones 2 and 3 of Fig. 236.
Cone 4 has almost reached the softening temperature.
NOTEs.-Critical points other than the softening temperature, that may be observed
during the test, and which may be of value are as follows:
- The Iniiial Deformation Temperature.~The temperature at which the first rounding
or bending of the apex of the cone takes place, as shown in cone 1 of Fig. 236. Such
,
bending must not be confused with a shrinking or warping of the cone.
The Fluid Temperature.-The temperature at which the cone has spread out over
the base in a flat layer, as represented by cone 5 of Fig. 236.
ULTIMATE ANALYSIS
(For Coal and Coke)
CARBON AND HYDROGEN

Method With Gas-Heated Combustion Furnace

The determination of carbon and of hydrogen shall be made with a weighed
quantity of the sample in a 25-burner combustion furnace of the Glaser type.
The products of combustion shall be thoroughly oxidized by passing them over
red-hot CuO and PbCr04, and shall be fixed by absorbing the water in a weighed
Marchand tube filled with granular CaC1 2 and by absorbing the Co. 2 in a Liebig
bulb containing a 30% solution of KOH.
Apparatus.-The apparatus used shall consist of a purifying train in duplicate l .a combustion tube in the furnace, and an absorption train.
(a) The purifying train shall consist of the following purifying reagents
arranged in order of passage of air and oxygen through them: H 2 S0 4, KOH
solution. soda lime, and granular CaCho One of the trains is for air and one
for oxygen. In the H 2 S0 4 and KOH scrubbing bottles the air and the oxygen

1650

COAL AND COKE

shall be made to bubble through about 5 mm. of the purifying reagent. Both
purifying trains shall be connected to the combustion tube by a Y-tube, the
joint being made tight by a rubber stopper.
'
(b) The combustion ~ube shall be made of hard Jena glass, or its equivalent.
Its external diameter shall be about 21 mm., and its itotal length 1 meter.
The first 30 cm. of the tube shall be empty; following this empty space is an
asbestos plug (acid-washed and ignited) or in its place a ~oIl of oxidized copper
gauze may be used; the next 40 cm. shall be filled with" wire" Cu 0; a second
asbestos plug separates the copper oxide from 10 cm. of fused PbCr04, whi~h
shall be held in place by another asbestos plug 20 cm. from the end of the tube.
The end of the tube shall be drawn out for rubber-tubing connection with the
absorption train.
(e) The absorption train shall consist of a Marchand tube filled with granular
CaCh to absorb moisture. The CaCh should be saturated with CO 2 before
uxing. The Marchand tube shall be followed by a Liebig bulb containing a
30% KOH solution, in which any possible impurities, as ferrous iron or nitrate:;,
have been oxidized by a little KMn04. A guard tube, containing granular
CaCh and soda linie, shall be attached to the Liebig bulb to absorb any CO 2
escaping the KOH solution and any water evaporating from that solution.
Procedure.-Connect the train to an aspirator which draws the products
of combustion through the entire train. A guard tube of CaCh prevents
moisture from running back into the absorption train. Maintain the suction
constant by a Mariotte flask. The advantage of aspirating the gases through
the train rather than forcing them through by pressure is that the pressure on
the rubber connections is from the outside, so that gas-tight connections are
more easily maintained than if 'the pressure is on the inside of the tube. Make
the connections as tight as possible. The usual test for tightness is to start
aspiration at the rate of about three bubbles of air per second through the
potash bulb, and then to close the inlet for air and oxygen at the opposite end
of the train; if there is no more than one bubble per minute in the potash bulb,
the apparatus shall be considered tight.
(
Before starting a determination when the train has been idle some hours,.or
after any changes in chemicals or connections, run a blank by aspirating about
1 liter of air through the train, which is heated in the same manner as if a
determination on the sample were being made. If the Liebig bulb and the tube
containing calcium chloride show a change in weight of less than 0.5 mg. each,
the apparatus is in proper condition for use.
...._
Use a porcelain or platinum boat provided with a glass weighing tube of
suitable size, which is fitted with an accurately ground glass stopper. Weigh
the tube and empty boat. Quickly place in the boat approximately 0.2 g. of
the air-dry sample (passing a 250-micron (No. 60) sieve and finer, or better,
_passing a 149-micron (No. 100) sieve if much free impurity is present). Place
the boat at once in the weighing tube, which is quickly stoppered to prevent
moisture change in the sample while weighing, and transfer to the furnace.
Connect the absorption tubes and transfer the boat and sample from the weighing tube to the combustion tube, which should be cool for the first 30 cm. The
CuO should be red hot and the PbCr04 at a dull-red,heat. Transfer the boat
from the weighing tube to the combustion tube as-:1apidly as possible.

COAL AND COKE

1651

As soon a.s the boat is in place near the asbestos plug at the beginning of the
copper oxide ins3rt the stopper connecting with the purifying train and start
the aspiration with pure oxygen gas at the rate of three bubbles per second.
Turn on one'burner about 10 cm. back from the boat, and continue the aspiration carefully until practically all the moisture is expelled from the sample.
Then increase the heat very gradually until all the volatile matter has been
driven off. In driving off the volatile matter the heat must be ltpplied gradually
in order to prevent a too rapid evolution of gas and tar, which may either escape
complete combustion 6r may be driven back into the purifying train. Increase
the heat slowly by turning on more burners under the open part of the tube until
the sample is ignited; then the temperature may be increased rapidly, but care,
should be taken not to melt the combustion tube.
Any moisture collecting in the end of the combustion tube or in the rubber
connection joining it to the CaCl 2 tube may be driven over into the CaCl 2 tube
by carefully warming with a piece of hot tile. Continue the aspiration with
oxygen for 2 minutes after the sample ceases to glo_w, then turn off the heat and
aspirate about 1200 ml. of air. Disconnect the absorption bulbs, wipe with
a clEian cloth, and allow to cool to the balance-room temperature before weighing.
Calculate the percentages of hydrogen and carbon as follows:
fh d
-11 19 X (increase in weight of CaCb tube)
t
ercen age 0 y rogen- .
Weight of sample
Percentage of carbon =27 27X (increase in weight of KOH bulb)
.
Weight of sampl

Weigh the ash in the boat and carefully inspect it for any unburned carbon,
which would destroy the value of the determination.
Method with Electrically-Heated Combustion Furnace
(a) Apparatus.-An electrically-heated combustion furnace of the Heraeus
type is used by the Bureau of Mines. 26 It consists of three independent heaters,
two of which are provided with sheave wheels, and are mounted on a track so
that they are movable along the tube; the third heater which surrounds the
PbCr04, is stationary.
(b) The furnace as provided by the manufacturer does not include the small
stationary heater. This may be made in the laboratory by winding an alundum
tube 12 cm. in length with No. 20 nichrome II wire and enclosing it in a cylinder
packed with magnesia-asbestos. The movable heaters have very thin platinum
foil. weighing about 9 g. in all, wound on a porcelain tube of 30 mm. internal
diameter. The larger one which heats the CuO, is 350 mm. in length, and the
smaller one, which heats the sample in the boat, is 200 mm, in length.
(c) The Jena glass or fused silica combustion tube, of about 21 mm. external
diameter and 900 mm, in length, is supported by an asbestos-lined nickel trough.
The current through each heater shall ,be regulated independently by separate
rheostats, mounted on the frame of the furnace. The two platinum-wound
heaters require an average current of about 4.5 amperes at a pressure of 220
volts, although for heating rapidly a larger amperage is necessary.
.
26

U. S. Bureau ot Mines Technical Paper No.8, p. 20 (1929),

1652

COAL AND COKE

Procedure.-Purify the oxygen or air entering the combustion tube by

passing through a T,auber's drying apparatus, which Icontains the following
reagents arranged in order of the passage of air, or oxygen through them:
H 2S0 4 , for removing possible traces of ammonia, 30% ,KOH solution, granular
soda lime, and grailUlar CaCI 2 Connect one side of ithe train directly to a
Linde oxygen tank provided with a reducing valve fori regulating the oxygen
pressure; use the other side of the train for purifying the air supply.
The absorption train shall consist of a 5 in. U-tube, filled with granular
CaCb to absorb moisture. Before using, saturate the CaCh with CO 2 to avoid
possible absorption of CO 2 during a determination by any traces of CaO that'
may be present. This saturating is done most conveniently by placing a
quantity of CaCl 2 in a large drying jar, and filling the jar with CO 2 , After
standing over night, draw dry air through the jar to remove the CO 2 Keep
the treated CaCl 2 in well-stoppered bottles.
Connect the CaCl 2 tube to a Vanier potash bulb containing a 30% KOH
solution and granular CaCh.. Six to eight determinations may be made without
recharging this bulb. Connect the potash bulb to an aspirator through a guard
tube containing granular CaCh and soda lime, and a Mariotte flask. The
Mariotte flask keeps the pressure constant.
In general, the method of determination is the same as the one used
with the gas furnace as previously described. By moving the heaters toward
the end of the tube where the gases enter, and cutting in the electric current, the
air may be warmed enough to thoroughly ~dry the tube and its contents. Cut
off the current from the small heater, and move the large heater over the CuO;
keep exposed about 250 mm. of that part of the combustion tube between
the two heaters where the boat containing the sample is to be placed. Turn
the full current on the large heater to bring the CuO to a red heat. When this
temperature is reached it is necessary to reduce the current with the rheostat
to avoid melting the tube. In the meantime, weigh and connect the absorption
train and place the boat containing the sample in the exposed and cooler part
of the tube between the two heaters.
Pass the current through the shorter heater. By manipulating the rheos.tat
and by gradually pushing this heater toward the: boat, the rate of evaporation
of moisture and evolution of volatile matter may be readily controlled.
After combustion is complete, turn the electric current off the smaller heater
and move this heater back, to allow the tube to cool for the next determination.
Conduct the final aspiration of air and the weighing of the~absQrption tra~n as
described in the Sections under the gas-furnace method.
NOTE.-In place of granulated CaCl 2, concentrated H 2S0 4 may be used for collecting
the water formed by combustion. In such cases the air and oxygen entering the combustion tube and the gas leaving the potash bulb must'also be dried by H 2S0 4
Other suitable forms of absorption vessels than those indicated in the above pro!
cedure may be used.
NITROGEN
. The Kjeldahl-Gunning method is recommended for the determination oi
nitrogen. This method has the advantage over either the simple Kjeldahl or
the Gunning method, in requiring less time for th~ complete oxidation of the
organic matter, and in giving the most uniform results.

COAL AND COKE

1653

The Kjeldahl-Gunning. Method.-Boil 1 g. of the sample with 30 ml. of

concentrated H 280 4, 7 to 10 g. of K 280 4, and 0.6 to 0.8 g. of metallic mercury
in a 500-ml. Kjeldahl flask until all particles of the sample are oxidized and the
solution nearly colorless. Continue the boiling at least 2 hours after the solution has reached the straw-colored stage. The total time of digestion will be
from 3 to 4 hours except in the case of coke and anthracite coal which require
much longer digestion. The addition of a few crystals of KMn04, after the
solution has cooled enough to avoid violent reaction, tends to insure complete
oxidation. Coke and anthracite coal should be ground to an impalpable
powder, as they are very difficult to oxidize. Even if this is done the digestion
may require 12 to 16 hours.
.
After cooling, dilute the solution to about 200 m!. with cold water. If the
dilution with water has warmed the solution, cool it again and add the following
reagents: 25 m!. K 28 solution (40 g. K 28 per liter) to precipitate the mercury,
1 to 2 g. of granular zinc to prevent bumping, and finally enough strong NaOH
solution (usually 80 to 100 m!.) to make the solution distinctly alkaline. The
danger 6f loss of ,NHa may be minimized by holding the flask in an inclined
position while the NaOH solution is being added. The alkaline solution runs
down the side of the flask and forms a layer below the lighter acid solution.
After adding the alkaline solution, connect the flask at once to the condensing
apparatus and mix the solution by gently shaking the flask.
Distill the NHa over into a measured amount (10 m!.) of standard H 2S0 4
solution, to which has been added sufficient cochineal indicator for titration.
Care should be taken that the glass c()nnecting tube on the end of the condenser
dips under the surface of the sta:qdard acid. Distill the solution slowly until
150 to 200 m!. of distillate has passed over. To avoid mechanically entrained
alkali passing over into the condenser, the rate of distillation should not exceed
100 m!. per hour. Titrate the distillate with standard NHa solution (20 ml.
NH 40H solution = 10 ml. H 2S0 4 solution =0.05 g. nitrogen). Standard
NaOH or KOH solution with methyl orange or methyl red as indicator may
be used instead of NHa and cochineal.
A blank determination shall be made in exactly the same manner as described
above, except that 1 g. of pure sucrose (cane sugar) shall be substituted in place
of the sample. The nitrogen found in this blank determination shall be deducted from the result obtained with the sample.
The K 2S and NaOH'may be dissolved in ,a single stock solution. Dissolve
sufficient K 2S in the water before adding the NaOH, to make a solution in which
the quantity necessary for a nitrogen determination (80 to 100 ml.) contains 1 g.
of K 2S. Twelve grams of K 2S and 500 g. of NaOH in one liter of water, are
required for the above proportions.
OXYGEN
There being no satisfactory direct method of determining oxygen it shall
be computed by subtracting the sum of the percentages of hydrogen, carbon,
nitrogen, sulfur; water and ash from 100. The result so obtained is affected
by all the errors incurred in the other determinations and especially by the
change. in weight of the ash-forming constituents on ignition. In the case of

1654

COAL AND COKE

coal, iron pyrite changes to ferric oxide, increasing the ash and causing a negative error in the oxygen equivalent to three-eights of \the pyritic sulfur. On
the other hand, there is always a loss on ignition, of water of composition
from the clayey and shaley constituents, CO 2 from carbdnates, etc., whigh tends
to compensate the absorption of oxygen.
!
Corrected Oxygen (for Coal).-When a more correct oxygen value is desired, it may be obtained by making the corrections indicated in the following
formula:
Corrected oxygen=100-[(C-C')+(H-H')+N+H,O+S'+corrected ash]
where

C = total carbon;
C'=carbon of ca'rbonates;
H = total hydrogen less hydrogen of water;
H' = hydrogen from water of composition in clay, shale, etc., ~
. N=nitrogen;
. H 20 = moisture as found as 105 C.;
S' = sulfur not present as .pyrite or sulfate. This is usually
small. In many types of cdal it may be disregarded.
Corrected ash = mineral constituents originally present in the coal. For
most purposes this can be determined with sufficient accuracy by
adding to the ash, as found, five-eighths of the weight of pyritic
sulfur, the CO 2 of carbonates and the water of composition of,
clay, shale, etc. See alio Determination of Ash, Sections
10 and 11.
CALORIMETRIC DETERMINATION

(For Coal and Coke)

Apparatus: (a) Combustion Bombs.-The Atwater, Davis, Emerson,

Mahler, Parr, Peters, Williams, or similar bombs may be used. The bomb shall
have an inner surface and accessory fittings, includinll' crucible, of platinum,
gold, porcelain enamel, or other material which is not attacked by HNO a and
H 2S0 4 , or other products of combustion.
:
(b) Calorimeter Jackei.-The calorimeter shall be provided with a waterjacket having a cover to protect the calorimeter f~om air currents. The jacket
shall be kept filled with water within 2 or 3 C. of the temperature of the room
(except in calorimeters which are totally submerged, where the jacke_t temperature is controlled by a thermostat) and should be stirred continuously by some
mechanical stirring device.
(c) Stirring of the Calorimeter Water.-The water in the calorimeter shall
be stirred sufficiently well to give consistent thermometer readings while the
temperature is rising rapidly. The speed of stirring I3hould be kept ,constant.
A motor-driven screw or turbine stirrer is recommended and the speed should
not be excessive. This may be determined by adjusting the temperature of
the calorimete!" to equality with that of the jacket and allowing the stirrer to
run continuously for 10 minutes. If the temperature of the calorimeter rises
more than about 0.01 C. in this length of time, the raj;e_of stirfing is excessive.
Accurate results cannot be obtained when too muc1Y@.ergy is supplied by the
stirring device or when the rate of stirring is irregular. The portion .of the

COAL AND COKE

'165,5

stirring device immersed in the calorimeter should be separated from that

outside by non-conducting material, such as hard rubber, to prevent conduction
of heat from the motor or outside air.
(d) Thermometers.-Thermometers used shall have been certified by a
government testing bureau and shall be used with the corrections given on the
certificate. This shall also apply to electrical resistance or thermo-electric
thermometers. Correction shall also be made for the temperature of the
emergent stem of all mercurial thermometers, and for the" setting" of Beckmann thermometers. For accurate work, either Beckmann or special calori. metric thermometers graduated to 0.010 or 0.020 C. are required. Such ther- /
mometers should be tapped lightly just before each reading to avoid errors
caused by the sticking of the mercury meniscus, particularly when the temperature is falling. A convenient method is to mount a small electric buzzer
directly on the top of the thermometer and connect it with a dry cell and a
push button. The button should be pressed for a few seconds immediately
before each reading.
(e) Oxygen.-.The oxygen used for combustions shall be free from combustible material. The total amount of oxygen contained in the bomb for
combustion shall not be less than 5 g. per gram of the sample. But the combustion must be complete, as shown by the absence of any sooty deposit on
opening the bomb after firing.
(/) Firing Wire.-The sample in the bomb may be ignited by means of
either iron or platinum wire. If iron wire is used, it should be of about No. 34
B. & S. gage and not more than 10 cm. (preferably 5 cm.) should be used at a
time. A correction of 1600 calories per gram weight of iron wire burned snaIl
be subtracted from the observed number of calories.
(g) Standardization.-The water equivalent of a calo'rimeter can best be
determined by the use of the stap.dard combustion samples supplied by the
U. S. Bureau of Standards. The required water equivalent is equal to the
weight of the sample multiplied by its heat of combustion per gram and divided
by the corrected rise in temperature.
The calorimeter shall be standardized by the combustion of standard
samples supplied by the Bureau of Standards, and used according to the directions given in the certificates which accompany them. A standardization shall
consist of a series of not less~than five combustions of either the same or different
standard materials. The conditions as to the amount of water, oxygen, firing
wire,. method of correcting for radiation, etc., under which these combustions
are made shall be the same as for combustions of the sample. In the case of any
disagreement between contracting parties a check standardization may consist
of two or more combustions of stltndardizing samples.
Procedure: (a) Preparation of Sample.-The ground sample shall be
thoroughly mixed in the bottle and an amount, approximately 1 g., shall be
taken out and weighed in the crucible in which it is to be burned. Samples
which are likely to be blown out of the crucible should be briqueted. After
weighing, the sample should preferably be immediately placed in the bomb and
this closed. This procedure is necessary to avoid sublimation in the use of
naphthalene for standardization.
(b) Preparation of the Bomb.-The firing wire, if iron, shall be measured
and coiled in a small spiral and connected between the platinum terminals,

1656

COAL AND COKE

using, if necessary, a piece of platinum wir"e somewhat heavier than the iron
wire, to make the connection. The platinum and the iron shall both be clean.
About 0.5 m!. of water should be placed in the bottom of the bomb to saturate
with moisture the oxygen used for combustion. Whet} the crucible is put in
place in the bomb, the firing wire should touch the sample or briquet of standard
material. For the combustion of standardizing samples, iron wi!;e is preferable
to platinum.
(c) Filling the Bomb with Oxygen.-Oxygen from the supply tank shall be
admitted slowly to avoid blowing the sample from the crucible without displacing the original air content, and the pressure allowed to reach 20 atmospheres for the larger bombs or about 30 atmospheres for the smaller bombs, so
that the bomb shall contain an amount of oxygen sufficient for complete combustion, namely, at least 5 g. per gram of the sample. This method of filling
should insure sufficient nitrogen for complete oxidation of the sulfur in the coal. 27
(d) Calorimeter Water.-The calorimeter shall be filled with the required
amount of distilled water, depending upon the type of calorimeter. The
amount may be determined either by measurement in stand~rdized flask or by
weighing. The amount shall be kept the same as that used in the standardization of the apparatus.
(e) Temperature Adjustments.-The initiil temperature in the calorimeter
snaIl be so adjusted that the final temperature, after the combustion, will not
be more than 1 C., preferably about 0.5 C., above that of the jacket, unGer
which conditions the total correction for heat gained from or lost to the surroundings will be small when the rise of temperature is 2 or 3 C., and the
effect of evaporation will also be small.
" (f) Firing Current.-The eleotric current used for firing the charge shall be
obtained from storage or dry cells having an electromotive force of not more
than 12 volts. A higher voltage is liable to cause an arc between the firing
terminals, introducing additional heat, whIch cannot be measured with certainty. The circuit should be closed by means of a switch which should remain
closed for not more than 2 seconds. When possible, it is recommended that an
ammeter be used in the firing circuit to indicate when the firing wire has burned
out.
(g) Method of Making an Observation.-The bomb when ready for firing,
shall be placed in the calorimeter, the firing wires connected, the cover put in
place and the stirrer and thermometer so placed as not to be in contact with
the bomb or container. The stirrer shall then be started~and _l!:fter the. thermometer reading has become steady, not less than 2 minutes after'the stirrer is
started, temperatures shall be read at I-minute intervals for 5 minutes and the
charge then fired,. the exact time of firirg being noted. Observations of
temperature shall then be made at intervals depending upon the method to be
used for computing the cooling correction. When the temperature has reached
its maximum and is falling uniformly, a series of thermometer readings shall be
taken at I-minute intervals for 5 minutes to determine the final cooling rate.
(h) Titration.,-Mter a combustion and after allowing the gas to escape the
bomb shall be opened and the inside examined for traces of unburned material
I

S. H. Regester, "Oxidation of Sulfur Compounds of .Coal, and of Nitrogen in the

Bomb Calorimeter, and the Correction to pe Applied in I>eiermining the Heating Value
of Coal," Industrial and Engineering Chemistry, 6, 812 (1914).
27

COAL AND COKE

1657

or sooty deposit. If these are found, the observations shall be discarded. If

the combustion appears complete, the bomb shall be rinsed out thoroughly and
t!J.e washings titrated with a standard alkali solution (1 ml. = 0.02173 g. HNO a =
5 calories) using methyl-orange or methyl-red indicator, to determine the
amount of acid formed. A correction of 230 calories per gram of HNO a should
be subtracted from the total heat observed. An additional correction of 1300
calories per gram of sulfur in the sarn,ple should be made for the excess of difference in heats of formation of S02 and aquequs H 2S0 4 over the heat of formation of aqueous HNO a
'
Calculation of Results: The following method 9f calculation is recommended
to take the place of the Pfaundler or other similar formulas for calculating the '
cooling correction (radiation correction) :
Observe (1) the rate of rise, r1, of the calorimeter temperature in degrees per
minute for 5 minutes before firing; (2) the time, a, at which the last temperature
reading is made immediately before firing; (3) the time, b, when the rise of
temperature has reached six-tenths of its total amount (this point can generally
be determined by adding to the temperature observed before firing, 60% of the
expected 2~ temperature rise, and noting the time at which this point is reached) ;
(4) the time, c, of a thermometer reading taken when the temperature change
has become uniform some 5 minutes after firing; (5) the final rate of cooling,
12, in degrees per minute for 5 minutes.
Multiply the rate 11 by the time b - a in minutes and tenths of a minute, and
add to this product (subtracted if the temperature was falling at the time a)
to the thermometer reading taken at the time a. Multiply the rate 12, by the
time c-b and add this product (subtracted if the temperature was rising at the
time c and later) to the thermometer reading taken at the time c. The difference of the two thermometer readings thus corrected, provided the corrections
from the certificate have already been applied, gives the total rise of temperature due to the combustion. This multiplied by the water equivalent of the
calorimeter gives the total amount of heat liberated. Divide this result, corrected for the heats of formation of HNO a and H 2S0 4 observed and for the
heat of combustion of the firing wire, when that is included, by the weight of
the charge to find the heat of combustion 'in calories per gram. Calories per
gram mUltiplied by 1.8 give the British thermal units per pound. (See example.)
NOTE.-In practice, the time b-a will be found so nearly constant for a given
calorimeter with the usual amounts of fuel that b need be determined only occasionally.
The results shall be reduced to calories per gram or British thermal units per pound
of dry sample, the moisture being determined upon a sample taken from the bottle at
about the same time as the combustion sample is taken.
28 When the temperature rise is not approximately known beforehand, it is only
necessary to take thermometer readings at 40, 50, 60 seconds (and possibly 70 seconds
with some calorimeters) after firing, and from th~se observations to find when the
temperature rise had reached 60% of the total. Thus, if the temperature at firing was
2.135 C., at 40 seconds 3.05, at 50 seconds 3.92, at 60 seconds 4.16, and the final
temperature was 4.200, the total rise was 2.07; 60% of it was 1.24. The temperature
to be observed was then 2.14+1.24=3.38. Referring to the observations at 40 and
50 seconds, the temperatures were respectively 3.05 and 3.92. The time corresponding
to the temperature of 3.38 was, therefore: '

3.38-3.05 10 44
d
40+3.92-3.05X
=' secon s.

COAL AND COKE

1658
Example.-

OBSERVATIONS.

Water equivalent = 2550 g.

Weight of charge = 1.0535 g.
I
Approximate rise of temperature expected '13.2 C.
60% of approximate rise = 1.9 C.,
Tim~

10-21
22
23
24
25
(a) 26
(b) 27.2

(c) 31
32
33
34
35
36

Thermometer
Readings,
Deg. Cent.

15.244
15.250
15.255
15.261
15.266
15.272
17.2<>
18.500
18.498
18.497
18.496
18.494
18.493

I
\ Corrected Temperature
(or corrected Beckmann thermometer readings)

Thermometer corrections taken from .the certificate .

15.276 C.
Charge fired
18.497 C.

a The initial temperature is 15.27 C.; 60% 6f the expe"ted rise is 1.9 C.
reading to observe is then 17.2 C.

The

CALCULATION

=0.028 +5=0.0056 per minute. b-a=1.2 minutes.

The corrected initial temperature is 15.276 +0.0056 X 1.2. . . . . . . . .. = 15.283
T2 =0.007 +5 =0.0014 per minute; c-b=3.8 minutes
The corrected final temperature is 18.497+0.0014X3.8 ........... =18.502
Total rise 18.502 -15.283. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.219
Total calories 2550 X3.219. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8209
-7
Titration, etc.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

T}

Calories from 1.0535 g. sample ........................... ~ . . . . . .

8202
Calories per gram ...................................... / . . . . . .. 7785
or British thermal units per pound ......... '" ...... '............. 14013
The result obtained by the above method of computation and determination is the
total heat of combustion at constant volume, with the water in the products of combustion
condensed to liquid at the temperature of the calorimeter, that is, about 20 to 35 C.
Net heat of combustion at 20, shall refer to results corrected for latent heat of
...
.
vaporization, as follows:
Total heat of combustion in B.t.u.-1040 (total hydrogenX9) = net heat of combustion in B.t.u. per pound.
Also:
Total heat of combustion in calories-580 (total hydrogenX9) =net heat of combustion in calories per gram.
NOTEs.-For anthracite, coke, and coal of high ash content, which do not readily burn
completely, the following procedure is recommended:
I
The inside of the crucible is lined completely with ignited asbestos in a thin layer
pressed well down into the angles. The sample is then sprinkled evenly over the surface
of the asbestos. Otherwise the procedure is as previously described.
The method of computing the "cooling correction" described in U. S. Bureau of
Mines Technical Paper No.8, pp. 26-30 (1929), may also be used. I

Tolerances: The permissible differences betweep.two gr more determinations

shall not exceed the values as specified.

COAL AND COKE

1659
Permissible
Same
Le.borato!,)

A.

ON LABORATORY SAMPLES, CRUSHED TO PASS THROUGH AN

840-MICRON (No. 20) SIEVE.
MOISTURE:

Under 5% ....................................... 0.2

Over 5% ....................................... '. 0.3
B.

Differences
DIfferent
Laboratories

0.3
0.5

ON LABORATORY SAMPLES, CRUSHED TO PASS THROUGH

A 250-MICRON (No. 60) SIEVE.
1. MOISTURE:

Under 5% ...................................... 0.2

Over 5% ........................................ 0.3

0.3
0.5

2. ASH:

No carbonates present ............................. 0.2

Carbonates present. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 0.3
Coals with more than 12% of ash, containing carbonate
and pyrite .................................... , 0.5
3.

0.5

1.0

VOLATILE MATTER:

Bituminous coals ..................'. . . . . . . . . . . . . . 0.5

Lignites......................................... 1.0
Cokes ........................................... 0.2
4.

0.3

1.0
2.0

0.4

SULFUR:

Coal, Under 2% ............... ,................... 0.05

Coal, Over 2% ................................... 0.10
Coke ............................................ 0.03

0.10
0.20
0.05

5. ULTIMATJ;) ANALYSIS:

Carbon .......................................... 0.3

Hydrogen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 0.07
Nitrogen ........................................ , 0.05
6. CALORIMETRIC DETERMINATION:

Permissible differences in per cent ................. 0.3

7.

0.5

FUSIBILITY OF COAL ASH:

Permissible differences in deg. Cent.................. 30

50

ANALYSIS OF

EXPLOS~S

The methods d~scribed in this chapter cover all Q( the more common types
of explosives employed in the United States for both commercial and military
purposes. Those of chief commercial importance are black powder, nitroglycerin dynamites, including "straight" dynamites, ammonia dynamites,
gelatin dynamites, and low-freezing dynamites, "Permissible" coal mining
explosives and nitrostarch blasting explosives. Military explosives include
smokeless powder, guncotton, trinitrotoluene, picric acid, ammonium picrate,
" Amatol," tetryl and tetranitroaniline. No sharp distinction can, however,
be drawn between commercial and military explosives, as many are utilized
for both purposes.. Many tons of surplus TNT on hand at the end of the
World War were used in commercial work; nitrostarch explosives found important application for military use during the war; mercury fulminate, lead azide,
and other detonators and priming compositions are essential in every field of
explosives.
The methods described have largely been used by the writer in practical
explosives testing and analysis in connection with both Government and pri:vate work. Most of those applying to commercial explosives have been
approved by. the United States Bureau of Mines for use in its Explosives
Chemical Laboratory.2

BLACK POWDER
The composition of black powder varies to some extent, depending chiefly
on the purpose for which the explosive is to be'used. Black blasting powder
contains sodium nitrate, charcoal and sulfur; black gunpowder),s quite similar
except that potassium nitrate is generally substituted for the soaium nitrate;
black fuse powder is similar to the latter, differing mainly in its granulation.
The same general method of analysis is therefore applicable to all types of
black powder.
Sampling.-From 50 to 100 grams of the original sample is crushed in small
portions in a porcelain mortar and completely p~ssed through an 80-mesh
Chapter by C. G. Storm. Published by permission of Chief of Ordnance, U. S. A.
See Bureau of Mines Bulletin No. 51, "The Analysis of Black Powder and Dynamite," W. O. Snelling and C. G. Storm, 1913, Bulletin No. 96, "The' Analysis of Permissible Explosives," C. G. Storm, 1916, Bulletin No. 219, "Explosives, their Materials, Constitution and Analysis," C. A. Taylor and W. H.,Rinkenbach, 1923, and Tech.
Paper No. 282, "Analysis of Detonating and Priming..'Mixtures;" C. A. Taylor and
W. H. Rinkenbach" 1922.
1660
1

ANALYSIS OF EXPLOSIVES

1661

sieve, ca:r:e being taken to avoid undue exposure to the air. The separate
powdered portions are promptly bottled and the entire sample is finally well
mixed.
Moisture.-The standard method of the Bureau of Mines is to desiccate a
2-grarr. sample on a 3-inch watch glass over sulfuric acid for three days, the
loss of weight being moisture. It has been shown, however, that equally
accurate results can be obtained by drying at 70 C. in a constant temperature
oven to constant weight, for which 2-3 hours is usually sufficient. As much
as 5 hours drying at 70 C. will not cause loss of sulfur. Drying at 100 C.
gives results which are slightly high, due to loss of sulfur.
Nitrates.-About 10 grams of the finely ground sample in a Gooch crucible'
provided with an asbestos mat, is extracted with warm water by means of
suction, the water being added in 15-20 ml. portions and each portion being
allowed to stand in the crucible a short time before suction is applied. About
200 mI. of water is usually sufficient, but the last drops of filtrate should be
tested by evaporation to ensure the absence of nitrates. A blue color on the
addition of sulfuric acid containing a few crystals of diphenylamine will also
indicate the presence of nitrates.
The water extract includes a small amount of water-soluble organic material
from the charcoal in addition to tlie nitrate. It is made up to 250 m!. and an
aliquot portion (50 mI.), evaporated to dryness on the steam bath, treated with
a little nitric acid, again evaporated, heated to slight fusion and weighed.
If allowance for impurities in the nitrate is desired, a direct determination
.of nitrate may be made on a separate portion of the water extract by the
Devarda method or by means of the nitrometer, but for all practical purposes
the evaporation method is sufficient. The usual tests should be made to
determine whether sodium nitrate or potassium nitrate is present.
Some black powders contain baJium nitrate in addition to potassium nitrate.
Barium nitrate, if present, is determined in an aliquot portion of, the water
extract by precipitating with sulfuric acid and' weighing the precipitate of
barium sulfate.
The residue left in the crucible, consisting of sulfur and charcoal, is dried
at about 70 C. to constant weight (for 5 hours or over night if more convenient), the loss of weight minus the moisture content being the watersoluble portion. This result serves as a check on the evaporation result.
Sulfur.-The residue in the crucible is extracted in. a Wiley extractor or
other continuous extraction apparatus with carbon disulfide, until evaporation of a small portion of the solvent passing through the crucible shows
absence of sulfur. The excess of carbon disulfide is then allowed to evaporate
from the crucible in a warm place away from flame, and the residue finally dried
to constant weight at 100 C. The loss of weight is considered as sulfur.
Charcoal.-The dry residue in the crucible should consist only of charcoal.
Ash.-The ash in the charcoal may be determiJ,ed by ignition -over a
Bunsen burner until all of the carbon has been burned off, and weighing.
This ash also contains, of course, any non-volatile matter that may have been
present in the sulfur and nitrate. An ash content in excess of 1% (calculated
on the weight of powder taken for analysis) may be regarded as an indication
of incomplete extraction of nitrate.

ANALYSIS OF EXPLOSIVES

1662

Calculation of Results.-Since a portion of the charcoal is always dissolved

in the water extract, it is customary to express th~ \ content of charcoal by
subtracting the sum of the following from 100%:

% Moisture

(by desiccation or drying at 70~ C.).

% Nitrate (by evaporation of water extract1with HN0 3).

% Sulfur

(by loss on extraction with CS2). I

NITROGLYCERIN DYNAMIl'ES
" Straight" Dynamite
So-called "Straight" nitroglycerin dynamite has been manufactured to
only a relatively small extent in this country for a number df years. This was
due in part to wartime developments in tli'e'manufacture of both ammoniilm
nitrate and nitrocompounds, and in part to the high price of glycerin, whixh
reached its maximum about the early part of 1927. During recent years tne
use of ethylene glycol dinitrate (more commonly called nitroglycol) as a substitute for a considerable part of the nitroglycerin in various types of dynamites
and "permissible explosives" has greatly decreased the amount of straight
dynamites manufactured. "Straight" nitroglycerin dynamites are still used
to some extent, however, particularly where quick-acting dynamites of high
strength are required, as in blasting in hard rock. They consist essentially of
nitroglycerin absorbed in a " dope" composed of It combustible absorbent,
usually wood pulp, and an oxidizing material (sodium nitrate), to which has
been added a small amount of an antacid (calcium caljl)onate, zinc oxide; etc.).
The analysis is best carried out by successive e;xtractions, usually with ether,
water and hydrochloric acid.
.
Sampling.-The wrappers are removed from a number of the cartridges,
and from 3 to 5 cm. of the ends of the exposed roll of explosives rejected. The
remainder is thoroughly mixed on a large sheet df paraffined paper or in a lsrge
porcelain dish, and an average sample selected and b()ttleq-usull,lly about one
half pound. The importance of thorough mixing of the sample must not be
overlooked, in view of the fact that there is frequently a decided"tendency for
the nitroglycerin to segregate due to insufficient or unsuitable absorbent, ,so
that this liquid ingredient may not be uniformly distributed throughout the
cartridge. Also if a carefully mixed sample has bee~ allowed to stand for some
days, especially in a warm place, segregation may QCcur in the bottle, so that
it is advisable to mix the sample again before an'alysis.
Qua1itati~e Examination.-Although a qualitative analysis of a sample
known to be straight nitroglycerin dynamite is usually unnecessary, the exact
nature of the sample may be unknown, and a knowl!ldge of the composition of
some of the more complex types of dynamite is n.ecesaary before a quantitative
analysis can be properly conducted.
-

ANALYSIS OF EXPLOSIVES

1663

About 25 grams of the sample is shaken with several successive portions of

ether in a large stoppered test tube, the ether being decanted off through a
filter paper and the residue finally washed on the filter. The ether solution is
allowed to evaporate slowly on a steam bath and the filter paper spread out
on a glass plate in an oven so that the residue may dry quickly. The evaporated ether extract may contain nitroglycerin, sulfur (especially in the lower
grades of dynamite), rosin, vaseline, or paraffin oil (in ammonia dynamite),
nitro toluenes and other nitrocompounds (in low-freezing dynamites), etc.
Nitroglycerin is readily detected by shaking a drop of the liquid with one
or two m!. of concentrated H 2S0 4 and about 1 m!. of mercury in a test tube,
an evolution of brown fumes of nitric oxides being noted if nitroglycerin is'
present. A similar reaction is, however, given by nitroglycol, so that this test
will not distinguish between straight nitroglycerin and a mixture of nitroglycerin and nitroglycoI. The latter substance (ethylene glycol dinitrate) has,
when pure, almost the same nitrogen content as nitroglycerin (18.42% as
compared with 18.50%). Its presence is indicated by a continual loss in weight
of the ether extract in attempting to dry it to constant weight. This loss is
more rapid in a vacuum. Sulfur will appear as crystals in the evaporated
extract, and may be identified by removing them, washing wi~h acetic acid,
and noting the odor of S02 on heating in a flame. Rosin, vaseline, oils, etc.,
appear as a greasy scum.on the surface of the nitroglycerin or adhering to the
walls of the beaker. These substances, like sulfur, are practically insoluble in
acetic acid (70%), and may be separated from the nitroglycerin by means of
this solvent. Trinitrotoluene will app-ear in the nitroglycerin as long yellowish
needles, which may be removed, recrystallized from alcohol, and identified
'by their melting point (approx. 80 C.), or by the red color produced when the
alcoholic solution is treated with a little caustic soda solution.
The residue insoluble in ether is replaced in the test tube and treated with
water in a similar manner until all water-soluble material has been dissolved.
The water solution is tested for sodium, potassium, barium, zinc, etc., and for
nitr{ttes, chlorides, etc., using the general methods of qualitative analysis.
The residue i!3 again treated with cold dilute HCl, any effervescence being
noted as indicating the presence of a carbonate, and the resulting solution
tested for calcium, magnesium, zinc, etc., which may have been present as
carbonates or oxides for the purpose of serving as antacids.
The residue insoluble in ether, water, and cold acid may contain wood pulp,
starchy cereal products, sawdust, nitro.cellulose, ground vegetable ivory (button
waste), kieselguhr, ground nut shells, etc. It is most conveniently examined
by means of a low-power microscope, whereby its con8tituents are usually
readily determined. Starch is easily detected by heating a portion to boiling
with dilute acid, cooling and adding a few drops of iodine solution (in KI), a
blue coloration indicating starch.
Moisture.-Moisture is best determined by desiccation over sulfuric acid,
a sample of about 2 grams being spread evenly over the surface of a 3-inch
watch glass and desiccated for 3 days. Continued desiccation causes a gradual
loss of nitroglycerin, but the 3-day loss may be safely assumed to closely represent the actual moisture content. The time of the determination may be
greatly shortened by the use of a vacuum desiccator, in which case 24 hours
desiccation will give a close approximation to the true moisture content.

1664

ANALYSIS OF EXPLOSIVES

It must be remembered that in determining moisture in the presence of

nitroglycerin, some volatilization of the latter is unavpidable, and that therefore the method followed must be an empirical one. An attempt to desiccate
the sample to constant weight will show that there is undoubtedly a continual
loss of nitroglycerin. This has been demonstrated 3 I~y a series of weighings
of a sample exposed for a period of 459 days at a c~mstant temperature of
33-35 C. in an empty desiccator containing no desiccating agent. A gradual
loss resulted during the entire period, totaling 17.52%_ of the original weight
of the sample, the original moisture content of which was about 1 %.
Extraction with Ether.-Ether removes from dynamite not only the nitroglycerin, but, as has already been mentioned, sulfur, resins (present as a
component or as a constituent of the wood pulp), oils (usually from cereal
products present), etc. Nitrotoluenes, nitroglycol, paraffin, vaseline, etc.,,.are
not normal constituents of straight dynamite and are considered 'under the
type of explosive in which they are most likely to occur.
Reflux Condenser Method.-From 6 to 10 grams of the sample is weighed in
either a porcelain Gooch crucible with asbestos mat or a porous alunduJIl
filtering crucible of about 25 m!. capacity. The asbestos mat is best prepared
as follows: A .mixture of 1 liter of water and 5 grams of previously ignited and
shredded short fibre asbestos free from hard lumps and very fine material is
well shaken and about 10 ml. poured into the crucible. Suction is applied
and a smooth and perfect mat almost invariably results. The crucibles thus
prepared are dried at 100 and are ready/for use.
The sample in the extraction crucible is extracted with about 35 ml. of
ether (U. S. P.) preferably in a continuous extraction apparatus (Wiley or
similar type preferred), for about 45 minutes to 1 hour, water being continuously circulated through the condenser and the extraction tube heated on a
water bath, or electric heater, the temperature of which is so regulated thatrthe
sample in the crucible will be kept covered with ether without overflowing.
Suction Method.-If desired, the ether extraction may be carried out by
suction, the Gooch crucible being held in a carbon tjibe passing through the
stopper in a suction flask. About 100 ml. of ether in~ 6 to 8 portions is passed
through th'e crucible, each portion being allowed to stand in the crucible for
one minute before applying gentle suction. ~o more air than is necessary
should be drawn through the sample in order to avoid condensation of moisture
in the sample, which might dissolve a portion of the water-soluble salts. This
method uses considerably; more ether than the reflux condenser method and
its chief advantage is that the apparatus required is more simpYe.,
On completion of the extraction the crucible is at once placed in a drying
oven, or the excess ether may be removed by suction before drying. If ammonium nitrate is present the drying should be conducted at 70 C. for 18
hours or overnight, but otherwise 5 hours at 100 C. is ample. The loss of
weight represents all ether-soluble material plus tHe moisture in the original
sample.
'
Evaporation of Ether Extract.-The ether extract is washed out of the
extraction tube or suction flask with a little ether into a tared evaporating dish
or small beaker and the ether allowed to evaporat~ spontaneously in a warm
3 Storm, C. G., "The Analysis of Permissible Explosi"i1'"s," Bulletin No. 96, Bureau
of Mines, pages 21-24, 1916.
-

ANALYSIS OF EXPLOSIVES

1665

place, or evaporated by means 'of the" bell jar evaporator." 4 The latter
consists of a tubulated bell jar with openingiil at top 'and side, placed on a
'ground glass plate, a slow current of dry compressed air from two drying cylinders containing H 2S0 4 and soda lime respectively, entering the top opening
through a glass tube, the lower end of which extends to about one half inch
from the surface of the ether solution in the beaker, which is placed on the
glass plate. The dry air current striking tlie surface of the solution with just
enough force to cause a slight" dimple," causes rapid evaporation pf the ether,
and deposition of moisture in the beaker along with the evaporated residue is
avoided. The low temperature produced by the rapid evaporation minimizes
the loss of nitroglycerin by volatilization. From 5 to 6 hours is usually re-~
quired for complete evaporation, which should be determined by check weighings. If the bell jar method is not used, the residue, after removal of the ether,
must be desiccated over H 2S0 4 for at least 24 hours in order to remove moisture
deposited during evaporation.
Nitroglycerin.-Nitroglycerin is determined in the dried and weighed ether
extract from which all ether has been removed as above described. This
determination is best made by means of the du Pont modification of the 5-part
Lunge nitrometer (see p. 650, Vol. I). The sample is dissolved in 5-10 ml. of
pure sulfuric acid (specific gravity 1.84) and transferred to the generating
bulb of the nitrometer, the beaker and cup of the nitrometer being washed
with several further additions of acid until a total of 20-25 ml. has been used.
If the quantity of nitroglycerin present is too great, the sample, dissolved in
sulfuric acid, is transferred to a burette and an aliquot part run into the
nitrometer. The maximum amount of pure nitroglycerin used for the -determination should not exceed 0.75 gram. The determination is carried out in
the usual manner and the reading of the gas volume in the graduated reading
tube divided by .1850 to find the weight of nitroglycerin in the sample used
for the determination (pure nitroglycerin contains 18.50% N).
Sulfur, Resins, Oils, etc.-It is always preferable to carry out the extraction with ether on duplicate samples, using one sample of the extract for the
determination of nitroglycerin as above, and the other for determining sulfur,
resins, oils, etc., that may also be contained in the ether extract.
The weighed extract is redissolved in a mixture of ether and alcohol, previously neutralized with standard alkali. The solution thus obtained is
titrated wi~h standard alcoholic potash solution using phenolphthalein indicator. 1 ml. of tenth normal alkali is equal to 0.034 grams of rosin (colophony).
A large excess of the alcoholic potash is now added and the mixture heated
several hours or overnight on the steam bath to saponify the nitroglycerin.
Shake with water and ether in a separatory funnel. The ether solution contains paraffin, vaseline, or mineral oils that may be present, and is evaporated
and the residue weighed. The water solution is acidified with ,Hel, and Br
added to oxidize any sulfur. Any separated rosin is filtered off and weighed
as a check on the titration, and sulfur -determined in the filtrate by precipitation as BaS04'
Sulfur may also be separated from nitroglycerin by means of acetic acid
of approximately 70% strength, the nitroglycerin being quite soluble in acetic
4 Storm, C. G., "The Analysis of Permissible Explosives," Bulletin 96, Bureau of
Mines, page 35, 1916.

1666

ANALYSIS OF EXPLOSIVES
I

acid and the sulfur almost insoluble. The: sulfur is filtered from the solution, washed slightly 'with alcohol to remove the acetic acid solution, dried
and weighed.
If a considerable quantity of crystals of sulfur is f6und in the evaporated
ether extract, it is possible that all of the sulfur has ndt been"removed by the
ether, and in this case an extraction is made with carbon disulfide, in exactly
the same manner as the ether extraction. This extraction is made subsequent
to the extraction with water, the sulfur being determined by loss of weight
of the residue or by direct weight after evaporation of the carbon disulfide,
away from free flame.
"
Extraction with Water and Determination of Nitrates.-The dried and
weighed residue left in the crucible after extraction with ether, is extracted
with water, using a suction flask fitted with a carbon filter tube in which the
crucible is held by a short length of thin-walled rubber tubing. Cold water is
used for this extraction, as hot water would gelatinize any starch present.
A total of at least 200 ml. of water is passed through the sample, in at least 10
portions, each portion being allowed to stand in contact with the residue foria
few minutes before being sucked into the flask. An evaporation test of a few
drops of the filtrate will determine the completeness of the extraction. When
the extraction is complete, the crucible with its insoluble residue is dried for
5 hours, or overnight; at 95-100 C., and the loss of weight noted as total
water-soluble material. This includes nit,rates and other soluble salts that may
be present, together with water extract from the wood pulp, flour or other
absorbent. This soluble organic material may amount to as much as 2% of
the total sample, when cereal products are present. Calcium, magnesiul}l,
or'zinc may also be present in solution, resulting from the action of acid decomposition products of the nitroglycerin on the carbonate or other antacid present.
In routine analyses of ordinary dynamite, the loss of weight on extraction with
water is usually considered as the alkaline nitrate (sodium or potassium), but
where more exact results are desired an aliquot portion of the extract is evaporated to dryness with a little nitric acid to oxidize otganic materials, and the
residue weighed as alkaline nitrate. This weight may/be corrected for inorganic
impurities-chlorides, sulfates, iron, aluminum, calcium, etc,-determined
separately by the usual methods.
'
Nitrates may be determined by means of the nitrometer, using an aliquot
portion of water extract estimated to contain '.6 to .8 gram of NaNO a or .8
to 1.0 gram of KNO a This is evaporated on the steam batli"alm,ost to dryness
and transferred with as little water as possible, to the cup of the nitrometer.
This solution is drawn into the generator and 30 to 40 m!. of 95--'96% H 2S0 4
added slowly so as to avoid generating sufficient heat to crack the glass. The
generator is then shaken for a total time of 8-10 minutes in order to be certain
that the generation of gas is complete with the (diluted acid. The gas is
measured and the, % of nitrate calculated as in the case of nitroglycerin.
Extraction with Acid.-When starch is not present in the rejSidue, a simple
extraction of the residue insoluble in water is made with cold dilute HCl
(1 : 10), 100 ml. being drawn through the sample in the crucible in small successive portions as described under " Extraction with Wat'er." Several portions of water are then drawn through to wash out-the acid, and the residue in
the crucible dried for 5 hours at 95 to 100 C. The loss of weight is usually

ANALYSIS OF EXPLOSIVES

1667

reported as antacid, but the base dissolved may be determined by the usual
quantitative methods if desired. The acid-soluble materials generally present
are calcium or magnesium carbonate or zinc oxide.
Determination of Starch.-If starch is present in the residue insoluble in
water, it is removed together with the antacid by boiling with dilute acid.
The residue is moistened with water, scraped or washed out of the crucible into
a 500-ml. beaker, the volume brought to about 250 ml. by the addition of water
and 3 ml. of concentrated HCl, and the mixture boiled until a drop of the
solution fails to give a blue color when treated on a spot plate with a drop of a
solution of iodine in KI. This indicates that the starch has been completely
hydrolyzed to dextrin. The mixture is then filtered through a fresh crucible,
washed with water, dried and weighed, correction being made for the weight
of the asbestos mat of the original crucible.
,
The antacid dissolved in the acid filtrate is determined as already described.
The loss of weight by the boiling treatment; minus the antacid found, represents starch and other dissolved organic materials removed from cereal products
or wood pulp. The insoluble residue includes the wood pulp and the crude
'fibre of the cereal products.
Because of the impracticability of exact separations it is customary to
report all of the soluble organic material included in both. water and acid
extractions as "starch" or "starchy material," and the insoluble organic
residue as " wood. pulp and' crude fibre," or the sum of these organic materials
is often reported as " carbonaceous combustible material."
Insoluble Residue and Ash.-The insoluble residue may contain wood pulp
or sawdust, the crude fibre from various cere!l,l products such as corn meal,
wheat flour, middlings, bran, etc., ground nut shells, vegetable ivory meal,
and more rarely' inorganic material such as infusorial earth (kieselguhr), clay,
etc. These can usually be identified by microscopic examination (see Bureau
of Mines Bulletin 96, Page 74), and a determination of the ash will show
whether inorganic materials are present. A high ash content may also' indicate
incomplete water or acid extractions.
Ammonia Dynamite
So-called ammonia dynamite is essentially" straight" dynamite in which a
large part of the nitroglycerin is replaced by ammonium nitrate. The ammonium nitrate is frequently protected from moisture by a coating of vaseline
or paraffin and is usually neutralized with zinc oxide. This type of dynamite
generally contains Jess wood pulp than the corresponding grades of " straight"
dynamite, and sulfur and cereal products, such as low grade flour, are usually
present.
The determination of moisture and the various extractions are carried out
as described for "straight" dynamite. An extraction with carbon disulfide
is usually necessary to effect complete removal of the sulfur; this properly
follows the,extraction with water. The analysis of the ether extract may be
conducted as already described. In drying the residue left in the crucible
after extraction with ether, it is important that a temperature of approximately
70 C. be used, because in the presence of ZnO, the loss of ammonium nitrate
is considerable at 100 C. Pure ammonium nitrate is not appreciably affected

1668

ANALYSIS OF EXPLOSIVES

by even 24 hours heating at 100 C., but the presence of the ZnO causes decomposition at this temperature.
"1
The water extract contains sodium nitrate and ammonium nitrate together
with practically all of the zinc oxide present, the latt~r ingredient being dissolved with the ammonium nitrate, and a small amount of soluble organic
material from the flour or other absorbent. It is analyzed as follows: An
aliquot portion is evaporated to dryness in a platinum o'r silica dish on a steam
bath, the ammonium nitrate volatilized by careful heating over a burner, a
little nitric acid added to re-oxidize any nitrate that may have been reduced to.nitrite, and the residue again dried on the steam bath. The zinc oxide is now
in the form of zinc nitrate and may be separated from the sodium nitrate by
either of the following methods:
1. The residue is dried at 110-120 C. and weighed as NaN0 3 and Zn
(NOah. It is then dissolved in water, the zinc precipitated with Na 2 C0 3,
filtered, ignited and weighed as ZnO, and the NaN0 3 taken by difference; the
total NaN0 3 plus Zn(N0 3 )2 minus (ZnOX2.33) =NaN0 3
2. The residue is gently heated over a burner until evolution of oxides of
nitrogen from decomposition of the Zn(N0 3)2 has ceased, and the remaining
residue weighed as NaN0 3 and ZnO. It is then treated with water, the insoluble ZnO filte.red on a Gooch crucible, ignited and weighed, the NaN0 3 being
taken by difference.
Ammonium nitrate is determined in a ~eparate portion Qi'the water extract
by the usual method of distillation and titration.
The sum of the amounts of NH 4 N0 3, NaN0 3 , and ZnO found will be somewhat less than the total water extract owing to the presence of water-soluble
organic material from the carbonaceous absorbents.
Gelatin Dynamite
This is a form of nitroglycerin explosive in which the nitroglycerin, instead
of being absorbed in porous materials such as wood pulp, is combined with
n'itrocellulose 'in the form of a gelatinous plastic mas~. As little as 3.5% of
suitable grade of nitrocellulose" containing about 12% nitrogen will, when
heated with nitroglycerin, at about 60 C., form a jelly-like non-fluid mass
when cooled to ordinary temperature. "Blasting gelatin," used to a considerable extent where great explosive strength is required, is a stiff_colloid composed
of 90 to 93% nitroglycerin 'and 10 to 7% nitrocellulose.
All blasting explosives containing such colloids of nitroglycerin and nitrocellulose combined with an active" dope" or base, consisting of a nitrate and
combustible material, are termed gelatin dynamites. This type of explosive
is also known in some countries as " Gelignite."
Sampling.-Owing to its pasty consistency the sample of gelatin, dynamite
must be prepared by cutting portions of a number of cartridges into thin bits
with an aluminum or platinum spatula. The use of a steel spatula or knife
for this purpose is not to be recommended for reasons of safety.' An ample
quantity of sample thus prepared is well mixed and bottle9,. Owing to its
tendency to again form a solid mass upon standin_g,.:.it- should be analyzed as
soon as possible after being prepared.
.

AN AL YSIS OF EXPLOSIVES

1669

Analysis.-The principle ingredients that may be found in the different

types of. gelatin dynamite are nitroglycerin; nitrocellulose; sulfur; rosin;
sodium, potassium or ammonium nitrate; calcium or magnesium carbonate;
wood pulp, cereal products and similar carbonaceous combustible materials.
Low-freezing gelatins may also contain nitrotoluenes or other nitrosubstitution
compounds, etc.
Moisture is determined as described for "straight" dynamite, and the
extraction with ether made in the usual manner except that ether free from
alcohol (distilled over sodium) is used in order to prevent partial solution of
. the nitrocellulose. The latter is readily soluble in a mixture of ether and
alcohol, arid as ordinary U. S. P. ether contains about 4% of alcohol, there is a
possibility that an appreciable part of the 0.5% to 2.0% of nitrocellulose
present in the sample will be dissolved unless pure ether is used. The ether
extract is evaporated and analyzed as already described and the water extraction made in the usual manner. If more than 1 or 2% of sulfur was present it
will not have been completely removed by the ether, unless the extraction was
continued for a sufficiently long time. In this event, it is necessary to make
an additional extraction with carbon disulfide in the Wiley apparatus subsequent to extraction with water.
NitroceUulose.-After the extractions with ether, water, and CS 2 (if
necessary) have been made, the nitrocellulose is determined, preferably by
extraction with acetone, which is a better solvent for the purpose than a mixture of ether and alcohol. It is advisable to separate the dry residue from the
crucible, leaving the asbestos mat ihtact if possible. The residue is transferred to a small beaker, covered with acetone and allowed to stand at least
3 or 4 hours with occasional stirring. It is then filtered through the original
crucible, washed with acetone, dried and weighed, the loss of weight being
regarded as nitrocellulose. To correct for small amounts of extract from the
wood pulp or other carbonaceous material, the acetone solution may be evaporated to about 20-25 mI., and diluted gradually with a large volume (about
100 ml.) of hot water, which volatilizes the acetone, precipitating the nitrocellulose as a white flocculent mass, which is filtered, dried, and weighed.
The remainder of the analysis is conducted ,as for straight dynamite.
It will be found that the results of analysis of a gelatin dynamite do not
agree with its trade markings. For example, the usual "40% strength"
gelatin dynamite actually contains from 30 to 33% of nitroglycerin and about
1% of nitrocellulose. Weight for weight this explosive is considerably weaker
than 40% straight dynamite, which contains 40% of nitroglycerin.
Low-Freezing Dynamite
Low-freezing dynamites vary from the dynamite types already discussed
by containing an ingredient. which reduces the freezing point of the nitroglycerin. This ingredient replaces a portion of the nitroglycerin which would
be used in an equal grade of ordinary straight dynamite, ammonia dynamite,
or gelatin dynamite. While straight nitroglycerin dynamite may freeze at
temperatures as high as 8 C. (46 F.), some of the low-freezing dynamites
freeze only at temperatures considerably below 0 C. Many of this type,
'however, cannot be relied upon to resist freezing at temperatures below the
freezing point of water.

1670

ANALYSIS OF EXPLOSIVES

The additions made to nitroglycerin for this purpose. include the nitrotoluenes, nitroxylenes, nitrohydrins, nitrosugar, nitropolygl(Ycerin (tetranitrodiglycerin), and nitroglycol. Any of these substances pres~nt will be found in
the ether extract together with, and in most cases dissolved in, the nitroglycerin.
\
Moisture.-The determination of moisture may
be carried out as already described for "straight"
nitroglycerin dynamite (p. 1663), except when there
are present such substances as nitro glycol or mononitrotoluene, whose volatility would prevent an accu-'
rate determination. An apprec.iable amount of
nitroglycol, for examplE,l, would be lost in an attempt
to desiccate to even approximately constant weight
over sulfuric acid, even without vacuum. The loss
of weight on desiccation should therefore be checked
against the difference between the total loss on extraction with ether and the direct weight of the
ether extract after evaporation of the ether in a
.
bell-jar evaporator (p. 1664).
The following method for the determination of
moisture especially in dynamites containing nitroglycol was developed by the Explosives Chemical
Laboratory of the U. S. Bureau of Mines: 50 grams
of the dynamite is weighed into a 500-ml. round
bottom flask to which is attached a special moisture
collecting tube. To the top of this tube is attached
a small surface condenser. 150 to 200 m!. of carbon
tetrachloride (free from H 2 0) is added through the .imoisture collecting tube and the flask heated to
boiling on a water bath for 1 hour. The CC1 4 and
water volatilized are condensed: and fall back into
the measuring tube, the water separating on top of
the CCI 4, while the excess of the latter runs back
into the flask. The measuring tube is graduated
in 0.1 m!. and the moisture is read direct. The
tube should be standardized before use. The
assembled apparatus is shown in Fig:237.~_
Nitrotoluenes.-Trinitrotoluene is not ~'readily
soluble in nitroglycerin and separates as crystals on
evaporation of the ether from the ether extract, enaFIG. 237. Moisture De- bling it to be qualitatively separated and identified.
termination Apparatus for
Dynamites, Oils, Fats, It may be determined by difference, the nitroglycerin
being determined by means o the nitrometer. Any
Waxes, Petroleum.
dinitrotoluene present may also be dete~mined in
this manner toge_ther with the trinitrotoluene, but if mononitrotoluene is also
6 A very similar method in which a solvent lighter than water is used, preferably
benzene, is recommended for the determination of moisture in-smokeless powders containing nitroglycerin, by F. Becker and P. Dittmar, Z. ges,.JSCliiess_Sprengsto1!wesen.
29, 327, 371 (1934).
. -

ANALYSIS OF EXPLOSIVES

1671

present, the determination of the nitrogen of the nitroglycerin will be slightly

in error by about 0.5530 gram of nitroglycerin for every gram of mononitrotoluene present.6
'
Mononitrotoluene is, however, seldom present except as an impurity in the
so-called liquid di- and trinitrotoluenes used in low freezing dynamites, so
that the determination of the nitroglycerin is usually fairly accurate and the
nitrotoluenes may be calculated by difference.
The total nitrogen of the combined nitroglycerin and nitrosubstitution
compound may also be determined, the nitrogen of the nitroglycerin deducted
and the amount of nitrosubstitution compound calculated from the resulting
difference; if the identity of the nitro substitution compound has been established. A suitable modification of the Kjeldahl method which has been found
applicable to difficultly decomposable nitro compounds is as follows: 7 This
method is, of course, applicable to mixtures containing nitroglycerin.
Modified Kjeldahl Method for Nitrogen.-About 0.5000 g. of the nitrocompound is weighed into a 500-ml. Kjeldahl flask, 30 ml. of 96% H 2S0 4 and
2 g. salicylic acid added and the sample dissolved by 'heating on a steam bath
if necessary. Cool; add 2 g. zinc dust in small portions, with cooling and
rotating the flask. Continue the shaking at 15 minute intervals for 2 hours
and let stand overnight. Then heat over a small flame till fuming has ceased
(about 2 hours), cool slightly and add 1 g. HgO and boil 1-172 hours longer.
Cool and add 7.5 g. K 2S0 4 and 10 ml. H 2S0 4 and boil 172 to 2 hours more.
If the solution is not clear and almost colorless, add 1 g. more K 2S0 4 and boil
longer. Cool and add 250 m!. H 20 to dissolve the cake formed, then add 25
m!. K2~ solution (80 g. per liter H 20), 1 g. granulated Zn, and 85-90 ml. NaOH
solution (750 g. per liter H 20), and distill as usual in the Kjeldahl determination, collecting the NHa in standard H 2S0 4 solution. A blank determination
without sample is advisable.
Separation of Nitrocompounds from Nitroglycerin.-Hyde has devised a
satisfactory method for actual separation of nitrosubstitution compounds from
nitroglycerin, depending on the differences in solubility of these ingredients in
carbon bisulfide and dilute acetic acid. s Nitroglycerin is only slightly soluble
in CS2, but readily soluble in dilute acetic acid, while most nitrocompounds
are much more soluble in CS 2 and much less soluble in dilute acetic acid than
nitroglycerin. CS 2 and acetic acid are only slightly miscible. H~nce nitroglycerin and a nitrocompound may be partly separated by shaking the mixture
with CS 2 and dilute acetic acid, allowing the two solvents to separate into two
layers and drawing off one of the layers. The CS 2 layer will contain most of
the nitrocompound and the acetic acid layer most of the nitroglycerin.
6 Storm, C. G., "The Effect of Nitrotoluenes on the Determination of Nitroglycerin
by Means of the Nitrometer," Proc. 8th Int. Congo App!. Chern., Vo!' 4, 1912, p. 117;
also Bur. of Mines Bull. 41, p. 62, 1913.
7 Cope, W. C., "Kjeldahl Modification for Determination of Nitrogen in Nitrosubstitution Compounds," J. Ind. and Eng. Chern., 8, 592 (1916). See also Cope,
W. C., and Taylor, O. B., "The Determination of Nitrogen in Substances Used in Explosives," Tech. Paper 160, Bur. of Mines, 1917.
8 Hyde, A. L., "The Quantitative Separation of Nitrosubstitution Compounds from
Nitroglycerin," J. Am. Chern. Soc., 35, 1173 (1913). (See also Bur. Mines Bulletin 96,
pp. 47-50, 1916.)

1672

ANALYSIS OF EXPLOSIVES

Hyde's method involves a continuous fractional extraction 'in a rather complicated apparatus consisting of 13 long narroW extra~tion tubes, connected
with ,each other and with a condenser, reservoir and distilling flask so as to
form a closed circulating system, the CS 2 continually passing in a train of fine
drops through acetic acid in the series of extraction tubJs, carrying with it the
nitrocompound, the nitroglycerin tending to remain d1ssolved in the acetic
acid. Practically a complete separation is finally obtained, the nitro compound..
dissolved in the CS 2 collecting in the distilling flask at the end of the extraction
train and the nitroglycerin remaining in solution in the acetic acid in the tubes.
The CS 2 is evaporated and the nitrocompound weighed. Reference should be
made to the original article by Hyde for details as to construction and operation
of the apparatus.
Nitrosugars.-The nitrates of sugar, improperly called nitrosugar, ar,e used
to a considerable extent for lowering the freezing point of nitroglycerin. This
substance is soluble in nitroglycerin, being prepared with the latter by nitrating
a solution of cane-sugar in glycerin, and no method is known .for its separation
from nitroglycerin. Hoffman and Hawse 9 have reported on an optical methodfor the determination of nit'rated sugar in nitroglycerin mixtures, based on the
use of the polariscope. As an example of the application of the method, 10.65
g. of a nitrated mixture of glycerin and sugar was dissolved in 100 mI. alcohol
and its angle of rotation found to be a = 3.07. The specific rotatory power
of sucrose octanitrate having been determined as a = 56.66, the formula:'
C (concentration) =a/2a gives a result oV25.44% sucrose octanitrate in the
sample.
The result of the optical method may be roughly checked by a determination of the total nitrogen of the combined nitroglycerin and nitrosugar, assuming the nitrogen content of the nitro sugar to be 15% (theoretical for sucrose
octanitrate 15.95%), and that of nitroglycerin 18.50%.
Nitrochlorhydrins.-Dinitromonochlorhydrin has been known for years as
a partial substitute for nitroglycerin in explosives. It is a solvent for nitrocellulose in smokeless powders and has an appreciable/effect in lowering the
freezing point of nitroglycerin. Some years ago it came into use in this country as a substitute for nitrotoluenes in low freezing dynamites.
A mixture of dinitrochlorhydrin and nitroglycerin will have a lower nitrogen
content than pure nitroglycerin, the dinitrochlorhydrin containing only 14.0%
N, as compared with 18.50% N in nitroglycetin. The dinitrochlorhydrin
may be readily identified and determined quantitatively b)",treaJing the mixture containing this substance and nitroglycerin with an excess 'Of' a:Icoholic
solution of KOH, heating on the steam bath until saponification is complete,
and determining the chlorine in the solution as chloride.
It must be noted that dinitrochlorhydrin is somewhat more volatile than
nitroglYGerin and therefore in evaporating the ether from the ether extract it is
advisable to make use of the bell-jar evaporator (p. 1(64) so as to minimize its
loss during evaporation.
Nitropolyglycerin.-Nitrated polymerized glycerin-usually a mixture of
tetranitrodiglycerin and trinitroglycerin-is sometimes found ip. low-freezing
explosives. This mixture will show a lower N-con~ent than nitroglycerin,
9 Hoffman, E. J. and Hawse, V. p" "The Nitration Df:..Siicrose Octanitrate," J. Am.
Chcm. Soc., 41, 235-247 (1919).
-

ANALYSIS OF EXPLOSIVES

1673

since pure tetranitrodiglycerin contains onlY'16.19% N. The presence of the

latter substance is indicated by low solubility in dilute acetic acid (60 volumes
glacial acetic acid to 40 volumes water). One gram of nitroglycerin dissolves
in about 10.5 ml. of this acid, while 1 gram of a mixture containing 82.25%
tetranitrodiglycerin required 120 ml. of the acetic acid to completely dissolve
it. In dissolving such a mixture, it will be found that a part of the mixture
dissolves more readily than the remainder. If the more difficultly soluble
portion is separated, dried in a desiccator and its nitrogen content determined
in the nitrometer, it will be found to contain a much lower % N than the
original mixture, approximating the figure for tetranitrodiglycerin, 16.19% (an
actual trial gave 16.24% N).
If the presence of tetranitrodiglycerin is established by the above procedure
and no other substances except nitroglycerin are present, the proportions of
these two ingredients in the ether extract may be readily calculated from the
N-content as found by the nitrometer.
Nitroglycol.-Ethylene glycol dinitrate (commonly called nitroglycol) is
rather extensively used at present for lowering the freezing point of nitroglycerin explosives. The usual procedure is to nitrate a mixture of glycerin
and glycol. The product has substantially the same nitrogen content as nitroglycerin, the chief distinguishing characteristic being volatility-nitroglycol
having a considerably higher vapor pressure. A continual loss of weight in
evaporating the ether extract may therefore serve as an indication of the
prese.nce of nitroglycol.

" PERMISSIBLE" EXPLOSIVES

" Permissible" explosives are coal mining explosives which have passed
the prescribed tests of the Bureau of Mines and are recommended by the
Bureau for use in gassy and dusty mines. Their important characteristic is a
relatively low flame temperature, which is brought about by modifying the
composition of the usual types of dynamites and other blasting explosives.
The general methods of reducing the flame temperature of explosives 10 are
summarized as follows:
(a) Addition of an excess of carbon-forming less CO 2 and more CO in the

gases of explosion.
(b) Addition of free water or of solids with water
(c) Addition of inert materials.
(d) Addition of volatile salts.

('f

crystallization.

10 The thermochemical considerations involved are discussed in Bureau of Mines

Bulletin No. 15, "Investigations ,of Explosives. Used in Coal Mines," 1912, and the
details of analysis in Bureau of Mines Bulletin No. 96, "The Analysis of Permissible
Explosives," C. G. Storm, 1916.

1674

ANALYSIS OF EXPLOSIVES

The analysis of explosives of this class is therefore generally more complicated than that of the ordinary types of blasting exptosives because of the
greater variety of ingredients used in manufacture. A partial list of substances which have been found in low-flame explosives I manufactured in' this
country is shown below, arranged according to their solubility in the general
scheme of analysis:
Soluble in Ether
Nitroglycerin
Nitropolyglycerin
Nitrotoluenes
Nitromgars
Nitrochlorhydrins
Nitroglycol
Paraffin
Resins
Sulfur
Vaseline

Oils

Soluble in Water
Ammonium nitrate
"
chloride
"
sulfate
"
oxalate
"
perchlorate
Alum (cryst.)
Aluminum sulfate (cryst.)
Barium nitrate
Calcium sulfate (cryst.)
Gums
Magnesium sulfate (cryst.)
Potassium chlorate
."
nitrate
"
perchlorate
Sodium nitrate
"
chloride
"
bicarbonate
"
carbonate
Sugar
Zinc oxide

Charcoal
Clay
Coal
Corn meal.
Corncob meal
Kieselguhr
Nitrocellulose
Nitrostarch
Nitrated wood

Soluble in Acids
Aluminum
Calcium carbonate
"
silicide
Ferric oxide
Magnesium carbopate
Zinc
'
Zinc oxide

Insoluble
Peanut shell meal
Powdered slate
Rice hulls
Sawdust
Turmeric
Vegetable ivory meal
Wheat flour If
Wood pulp

Qualitative Analysis.-The qualitative examination of a " permissible"

explosive is conducted in the same manner as ha~ been described for dynamite
(see page 1662), and, in view of the greater variety of constituents that may
be present, is quite essential before a suitable scheme for q-uantit!l:tive separation can be chosen.
Tests for some of the more unusual substances not generally found in the
ordinary types of blasting explosives, and not already discussed under" Lowfreezing Explosives," are made as follows:
'
Test for Sugar.-The presence of water-soluble organic substances is indicated by an appreciable charring of the residue o~tained by evaporating a
portion of the water extract to dryness and then heating gradually over a
burner. A slight charring may result from water-soluble portions of cereal
products, wood-pulp, etc., and may be disregarded. Sugar is identified by
acidifying some of the water solution with a litt\e dilu,t~ ,HCI, lfeating to boiling,
neutralizing with KOH and then boiling with Fehling) solution. A precipitation of cuprous oxide indicates the presence of sugar.

ANALYSIS OF EXPLOSIVES

1675

Test for Gum Arabic.-Gum arabic is precipitated by the addition of a solution bf basic lead acetate to the water extract, a white, flocculent precipitate
of indefinite composition resulting (see Determination of Gum Arabic, p. 1677).
Test for Nitrostarch.-Nitrostarch is best identified by microscopic examination of the residue insoluble in water. It is easily distinguished from unnitrated starch by means of a solution of iodine in KI, which colors the starch
granules dark blue but does not affect the granules of nitrostarch.
Test for Chlorides, Chlorates, and Perchlorates.-These three substances
present in a solution may be identified as follows: Acidify slightly with nitric
acid, add exces'! of AgNO a, heat to boiling, shake well, and filter off the silver
chloride. To the filtrate add a few ml. of 40.% solution of formaldehyde
(formalin), and boil to reduce chi orates to chlorides .. This reduction is. best
carried out by heating on the steam bath for about an hour. Any chloride
thus formed is then separated by further precipitation with AgNO a and removed by filtration. The filtrate is then evaporated to dryness, the residue
transferred to a crucible and fused with dry Na 2 CO a The fused mass is
treated with dilute HNO a, when the presence of perchlorate will be indicated
by an insoluble precipitate of AgCl.
Mechanical Separation of Solid Igredients.-lt is freql!ently of advantage,
especially in connection with the interpretation of the results of analysis of an
explosive mixture containing a number of water-soluble salts, to determine the
identity of one or more of the components of the mixture by means of screening
or by a method of separation depending on variation in specific gravity of the
components. Such methods are facilitated by the fact that the ingredients
of blasting explosives are frequently.not finely pulverized in the course of
manufacture.
.
(a) By Screening.-25 to 50 grams of the sample is washed several times
with ether to remove nitroglycerin and ingredients of an oily nature, the solid
residue dried to remove adhering ether and then sifted through a set of sieves.
An examination of the portions held by the 10- and' 20-mesh screens will usually
show the presence of coars.e crystals which are large enough to be sorted out
with the aid of forceps, submitted to qualitative tests and identified witIl
certainty. A single crystal may sometimes be identified by dissolving it in a
drop of water on a microscope slide, allowing the water to evaporate and
examining the resulting crystals under the microscope. The writer has frequently identified three or four ingredients of an explosive in this manner.
(b) By Specific Gravity Separations.-This method, applied to the analysis
of explosives by Storm and Hyde,ll depends on the separation of solids from a
mixture by means of inert liquids of different specific gravities. A series of
mixtur~s of chloroform (sp.gr. 1.49) and bromoform (sp.gr. 2.83) is prepared
covering as wide a range of specific gravity as may seem desirable. Portions
of the dried sample previously extracted with ether as in (a) are added to such
liquid mixtures and the heavier salts, which settle to the bottom, separated
from the lighter ones. For example a mixture of ammonium nitrate (sp:gr.
1.74) and sodium chloride (sp.gr. 2.17) is readily separated into its components
in a liquid with a specific gravity of (e.g.), 1.90, so that the components can
be tested separately and the analyst assured that the mixture is not composed
11 Storm, C. G., and Hyde, A. L., "Specific Gravity Separation Applied to the
Analysis of Mining Explosives," Tech. Paper No. 78, Bureau of Mines, 1914.

1676

ANALYSIS OF EXPLOSIVES

of sodium nitrate and ammonium chloride-which could not be ascertained

by ordinary quantitative analysis. (For example; a I mixture composed of
16.61 % Na, 44.76% NO a, 13.00% NH4 and 25.63% 01 may contain either
61.37% NaNO a and 38.63% NH4CI, or 57.76% NH 4Npa and 42,24% Na91,
or varying proportions of all four ingredients.) The chloroform-bromoform
.
mixtures are recovered by filtering and used repeatedly. I
The specific gravities of some of the more common salts that may be found
are as follows:
'
Ammonium alum (cryst.) ...................... '.......... "
"
chloride .................................... "
nitrate ..................................... "
"
"
perchlorate ................................. "
sulfate.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ...
"
Barium nitrate ........................................ "
Calcium carbonate (ppt'd) .............................. "
"
sulfate (anhydrous) ............................. "
sulfate+2H 20 ................................. "
"
Magnesium carbonate .................................. "
"
sulfate+7H 20 ............. , ................ "
Manganese dioxide ..................................... "
Potassium alum (cryst.) ................................ "
"
chlorate ..................................... "
"
chloride ................... , ................. "
"
nitrate ...................................... "
"
perchlorate ............. /. . . . . . . . . . . . . . . . . . . . ..
"
sulfate ...................................... "
Sodium chloride ....................................... "
"
nitrate ................. ' ....................... "
"
'sulfate (anhydrous) ............................. "
"
sulfate+ 10H 20 ................................. "

1.62
1.52
1.74
1.87
1.77 .
3.23
2.72
2.97
2.32
3.04
1.68
5.03
1.75
2.33
1.99
2.09
2.52
2.66
2.17
2.26
2.66
1.46

Moisture.-The determination of moisture in all types of " permissible i,

explosives is carried out by,the method described for nitroglycerin dynamites
(page 1663). The influence of the slight volatility of nitroglycerin and of
certain nitro substitution compounds on the results of ,ihis determination has
been discussed (pp. 1664, 1670). A more serious factor in the case of many
" permissible" explosives is the presence of salts containing water of crystallization. Most salts of this type (e.g., MgS047H~O) undergo a gradual loss of
a large part of their combined water on desiccation over either H 2 S0 4 or CaCI 2 ,
thus renderinl$ it impossible to differentiate between hygroscopic moisture and
combined water. Attempts to remove the total water content by~heating at a
temperature high enough to d.rive off all of the water of crystallization are
useless on account of the increased volatilization of nitroglycerin, ammonium
nitrate, etc., at such temperatures.
In such cases it is necessary to determine all other cQnstituents by direct
methods and estimate moisture by difference, the s~lt to which the water of
crystallization belongs being calculated as containing its full quota of water;
or the crystallized salt may be calculated as anhydrous and the difference from
100% reported as " water of crystallization plus moisture."
Extraction with Ether.-The extraction with ether, the eVJlporation of the
ether, and the analysis of the ether-soluble portion are conducted as already
discussed for nitroglycerin dynamites (pp. 1664, HHtSf.

ANALYSIS OF EXPLOSIVES

1677

In drying the crucibles containing the residue insoluble in ether, a temperature of 100 C. may be used except when the residue contains ammonium
nitrate or organic nitrates such as nitrocellulose, nitrostarch, or nitrated wood.
When any of these substances are present, the residue should be dried to constant weight at 70 C. Except when salts containing water of crystallization
are present, the amount of ether-soluble material found is calculated by deducting the moisture determined by desiccation, from the difference between
the weight of original sample and the weight of the dried residue insoluble in
ether. The procedure followed when water of crystallization is present is
noted in the preceding paragraph.
Extraction with Water.-Water-soluble salts are extracted from the weighed.
residue insohible in ether as already described, the residue left in the crucibles
dried to constant weight at 95 to 100 C., cooled and weighed. The water
soluble salts in the solution are determined by the usual methods of inorganic
analysis.
Nitrates.-In determining nitrates by the nitrometer method (see p. 1666)
it must be remembered that the presence of a considerable quantity of chlorides
may interfere with the accuracy of the results. Many of the " permissible"
explosives contain sodium chloride in amounts varying from 1% to 10 or 15%.
_ M. T. Sanders 12 has shown that if the sodium chloride is present in an amount
exceefiing 15-17% of the sodium nitrate, the result is not accurate within 0.1 %.
Smaller amounts of sodium chloride do not interfere, except to increase the
amount of sludge formed in the nitrometer.
Nitrates may also be determined by the" nitron " method of Busch,13
Chlorates.-Chlorates may be determined by any of the methods described
on page 274 (reduction with S02, FeS04, or Zn) or by the formaldehyde
methodY In the latter method a portion of the sulution, containing about
0.5 g. of chlorate, is diluted to 150 ml., 5-10 ml. of 40% formaldehyde soiution,
2 ml. dilute HNO a (1 : 3), and 50 ml. of approx. tenth normal silver nitrate
added, the solution covered and heated on the steam bath for about 4 hours,
when the precipitate of AgCl is filtered off, washed, dried and weighed. This
method is accurate to .05 to .10%.

Perchlorates.-The determination of perchlorates by reduction to chlorides

on ignition with NH 4Cl in the presence of platinum is described on page 275.
Perchlorates may also be determined by means of precipitation with" nitron "
in exactly the same manner as for nitrates. The weight of nitron perchlorate
(C 2oH 16N 4HCl0 4 ) found, multiplied by 117.5 (mol. wt. of NH 4Cl0 4) and
divided by 412.5 (mol. wt. of nitronperchlorate), gives the weight of perchlorate
found, expressed as NH 4 CI0 4.
Gum.Arabic.-This substance, sometimes used as a _binder in dry explosive
mixtures-especially chlorate or perchlorate powders-is determined by precipitation with basic lead acetate solution, prepared by adding 159 g. of normal
lead acetate and 50 g. lead oxide (PbO) to 500 ml. distilled water, heating
12 Sanders, M. T., "The Effect of Chlorides on the Nitrometer Determination of
Nitrates," J. Ind. and Eng. Chem., 12, 169-170 (1919).
13 See page 639, also, for further details, "The Analysis of Permissible Explosives,"
Bureau of Mines, Bulletin 96, pages 60--{l2.
14 Storm, C. G., "The Analysis of Permissible Explosives," Bulletin 96, Bureau of
Mines, pp. 63--{l4, 1916.

1678

ANALYSIS OF EXPLOSIVES

almost to boiling, and filtering. This reagent is added to the solution containing the gum arabic until no further precipitation pccurs; the mixture is
allowed to stand for several hours, then filtered, washed with absolute alcohol,
dried at 100 O. and weighed. The weight of precipl~ate multiplied by the.
factor 0.4971 (determined experimentally) gives the weight of gum arabic
found. Ohlorides or sulfates, if present, interfere with the determination and.
must be first removed.
I
.
Sugar.-Sugar may be present as an ingredient in some "permissible"
explosives, and is always found in small amounts in the water extract if cereal,.
products such as corn meal or wheat middlings are present. A portion of the.
water extract is acidified with HOI (1 m!. conc. HOI to 100 m!. solution), heated
just to boiling, cooled, nearly neutralized with N a 2 003, an excess of Fehling's
solution added and the mixture heated until reduction is complete. The Ou 20
is filtered from. the blue liquid, dried, ignited to constant weight, and weighed
as OuO. This weight X 0.4308 equals weight of cane sugar. The result is
corrected for the result of a blank determination using distilled water instead
of the water extract. By the use of this method after first extracting with
ether, .then with water, corn meal was found to contain 2.65% and wheat
middlings 6.25-7.00% of sugar. Thus an explosive containing 25% wheat
middlings would show as much as 1.75% of sugar in its water extract.
Extraction with Acid.-As in the ordinary nitroglycerin dynamite~, the
substances removed from "permissible" explosives by acid extraction are
chiefly substances added as antacids, including calcium carbonate, magnesium
carbonate and zinc oxide. Other acid-soluble materials that may be present
include metallic aluminum or zinc, ferric oxide, manganese dioxide, and calcium
silicide. When starch is present, the residue from the water extraction is subjected to hydrolysis in boiling dilute HOI as already described (page 1667),
and the acid-soluble inorganic components determined in the filtrate by the
usual methods. An extraction with cold acid is made only when there is no
starch present.
Extraction with Acetane: Nitrocellulose and Nitrostarch.-If either nitrocellulose or nitrostarch is present, an extraction with 'acetone is made as described for gelatin dynamite (page 1669). It should be noted in connection
with the preceding steps in the analysis that in order to avoid partial solution
of these substances in ether, the ether used in the ether extraction should be
alcohol-free (distilled from sodium), and also that all drying of residues containing these materials should preferably 'be conducted at 70 Q, instead of
100 C., in order to avoid partial decomposition. It is impracticable to separate nitrostarch from nitrocellulose but they are not likely to be found together
in the same explosive .. Small amounts of nitrocellulose are detected less readily
than nitrostarch, which is easily identified by the'microscope.
Insoluble Residue and Ash.-The insoluble residue is usually carbonaceous
combustible or absorbent material and is in most chses readily identified by
means of a microscope (preferably binocular) with low power (25-50 diameters). The possible presence of any inorganic material which may have been
overlooked in the analysis is detected by means of a determination of ash, the
residue being ignited until all carbon is burned off, and thl mineral residue
weighed. This is usually not over 0.2%. If highe:r'ihan 0.5%, there is reason

ANALYSIS OF EXPLOSIVES

1679

to suspect that some such material as kieselguhr or clay is present, or that the
extractions with water or acid were not complete.

NITROSTARCH EXPLOSIVES

General Nature.-Nitrostarch explosives have been for a number of years

used to a very considerable extent in this country for commercial blasting
purposes, chiefly for quarrying. During the war, explosives of this class were
adopted by the United States for certain military purposes and proved satisfactory substitutes for trinitrotoluene as bursting charges for hand grenades,
rifle grenades and trench mortar shell.
The commercial nitrostarch explosives may contain, in addition to the
nitrostarch, any or all of the following components: oxidizing agents, as sodium
or ammonium nitrates, combustible material, such as charcoal, flour, sulfur,
etc., mineral oil, and antacids, such as calcium carbonate, or zinc oxide. Nitrostarch military explosives may consist of some such mixture as the above, or
may be composed almost entirely of nitrostarch with the addition of relatively
small amounts of oils and mat~rials used for granulating. Some types of
nitrostarch explosives proposed for military use contained water in amounts
up to 10-15% combined with a mixture of nitrostarch and a soluble nitrate.
For commercial use, nitrostarch explosives are put up in cartridge form in
paper wrappers, just as nitroglycerin explosives are prepared.
Moisture.-Moisture is determined as in the case of nitroglycerin dyna..
mites. 3 to 5 grams of the sample is desiccated over sulfuric acid for 2-3 days
(2 days is usually sufficient to give constant weight), or for 24 hours in a
vacuum desiccator with a vacuum of at least 700 mm. of mercury. Unlike
nitroglycerin, nitrostarch does not volatilize and may be desiccated to constant weight. When heated .to higher temperatures, e.g., 100. C., for any
extended time, it undergoes, like all nitric esters, a gradual decomposition with
loss of weight.
Extraction with Petroleum Ether: OilS, Sulfur, etc.-The small amount
of alcohol usually present in ordinary grades of ethyl ether, is sufficient to
cause partial solution of the nitrostarch, if ethyl ether is used for removing
oily ingredients, sulfur, etc., nitrostarch being readily soluble in mixtures of
ether and alcohol. It is therefore advisable to use petroleum Ither, which
does not dissolve nitrostarch, for this extraction.
A sample of about 10 grams of the explosive, in a Gooch crucible with
asbestos mat, is extracted with pure petroleum ether of about 0.65 specific
gravity, the excess solvent removed by suction and. the crucible with sample
dried to constant weight at approximately 70 C. The % loss of weight,
minus the moisture content, already determined, represents the percentage
of ether-soluble material present. The petroleum ether is removed by evap-

1680

ANALYSIS OF EXPLOSIVE.S

oration, the residue of ether-soluble materials dried, weighed, and its components determined by the methods used for dynamites!.
Extraction with Water: Nitrates, Gums, etc.-The dried and weighed
residue insoluble in petroleum ether is extracted with didtilled water to remove
the nitrates or other water-soluble materials. The in~oluble residue left in
the crucible is dried at a temperature of 80 C. (100 m:a,y cause some decomposition of the nitrostarch) for several hours, and weighed, the loss of weight
being total water extract and serving as a check against the sum of the components separately determined.
Ammonium nitrate, if present, is determined in the water solution by the.
usual method of distillation of the NH3 after adding an, excess of alkali. Sodium nitrate is determined by evaporating the water extract, volatilizing
ammonium salts, and weighing the residue after re-oxidizing with nitric acid
if charring has indicated the presence of any water-soluble organic materilil
(page 1667).
If the original explosive is granular in form, the presence of a binding or
agglutinating material in the water extract may be suspected. Although
numerous other substances may be used for the purpose, gum arabic is frequently employed as a binding agent in different types of dry explosives. A
qualitative test for gum arabic has been mentioned on page 1675, and its
quantitative determination may be conducted as described on page 1677, by
precipitation with basic lead acetate solution.
Insoluble Residue: Nitrostarch, Charcoal, Cereal Products, etc. Starch.A microscopic examination of the weighed insoluble residue will usually serve
to identify its components. Any un-nitrated starch or cereal product is
readily distinguished, from nitrostarch by treating with a drop of KI solution
of iodine and examining under the microscope, when the un-nitrated starch
granules will appear "blue or black and the nitrated starch colorless or yellow.
Charcoal is identified by its color.
Un-nitrated starch, if present in an amount greater than a trace, is determined by boiling with dilute H 2S0 4 or HCl until iddine solution no longer
colors a drop of the liquid blue, then filtering and w~shing thoroughly. The
residue is dried at 80 C. and weighed, the loss of weight representing starch.
Nitrostarch.-Another portion of the insolu~le residue (the analysis being.
conducted in duplicate) is extracted with acetone in. a Wiley extractor or
other continuous extracti9n apparatus, or by transferring .the residue from the
crucible to a small beaker, digesting in acetone with stirring,. ?-nd filtering
through the same crucible, washing with fresh acetone. This extraction dissolves all of the nitrostarch, leaving any charcoal or cereal products that may
be present. The residue is dried at 100 C. al!d weighed, the loss of weight
representing nitrostarch.
The nitrogen content of the nitrostarch may .be determined, if desired,
by precipitating the nitrostarch from the clear acetone solution by the addition
of water and evaporation on a steam bath. A portion of the white, floury
precipitate is then dried at 70-80 C., weighed, and its nitrogen content determined in the nitrometer (see page 650, Vol. I).
,
Charcoal.-If charcoal is present its weight may be taken direct, in the
absence of cereal products or other substances._' When the residue contains
cereal products, the material left after hydrolysis of the starch and extraction

ANALYSIS OF EXPLOSIVES

1681

of the nitrostarch will contain the crude fibre of -the cereal together with charcoal or other insoluble ingredients_ A separation of such components is usually
impracticable.
Trinitrotoluene (TNT)

Trinitrotoluene, commonly designated in this country by the abbreviation

TNT, is also known in this and other countries by such names as triton,
trotyl, tolite, trilite, trinol, tritolo, etc. The term trinitrotoluol, which is
probably more commonly used than trinitrotoluene, is incorrect according to
approved chemical nomenclature.
This explosive is of the greatest importance as a high explosive for military
use, being adaptable as a bursting charge for high explosive shell, trench
mortar shell, drop-bombs, grenades, etc., because of its powerful explOSIve
properties, relative- safety in manufacture, handling, etc., its stability, its lack
of hygroscopicity, and absence of any tendency to form sensitive compounds
with metals.
It is classified by the Ordnance Department, U. S. A., into two grades,
according to purity-Grade I with solidification point of at least 80.2 C. and
Grade II with solidification point of at least 76.0 C. Other requirements, the
same for both grades, are as follows: acidity (as H 2 S0 4), not more than 0.01 %;
insoluble in benzene or 95% alcohol, not more than 0.10%; moisture, not more
than 0.10%. Grade, I represents material purified by means of sodium sulfite.
Solidification Point.-The determination of the solidification point or
" setting point" of TNT is the best single test for purity of this compound,
and is preferably carried out as follows:
A sample of about 50 grams of TNT is placed in a I/1X6/1 test tube and
melted by placing the tube in an oven at about 95 C. The tube is then
inserted through a large cork stopper into a larger test tube about lY2/1X7",
which, in turn, is lowered into a wide-mouth liter bottle, so that the rim of the
large tube rests on the neck of the bottle. The inner test tube is provided
with a cork stopper containing 3 openings-one for a standard thermometer
graduated in 1/10 C., one for a short thermometer which is passed just through
the stopper and is used for noting the average temperature of the exposed
mercury column of the standard thermometer, and the third opening being
a small v-shaped notch at the side of the'stopper, through which passes a wire
whose lower end is bent in a loop at right angles to the axis of the tube and
which is used for stirring the molten sample of TNT.
The standard thermometer is so adjusted that its bulb is in the center of
the molten mass, and the stirrer is operated vigorously, the thermometer being
watched carefully as the temperature falls. The temperature will finally
remain constant for an appreciable time and then rise slightly, owing to the
heat of crystallization of the TNT. As this point is reached, readings should
be taken about every 15 seconds until the maximum temperature of the rise
is reached. This temperature will usually remain constant for several minutes
while crystallization is proceeding. The maximum reading, corrected for the
emergent stem of the thermometer, is taken as the solidification point of the
sample.
Ash.-About 5 grams of TNT is moistened with sulfuric acid and burned
in a tared crucible. The residue is again moistened with a few drops of nitric
acid and sulfuric acid and again ignited and the resulting ash weighed.

1682

AN AL YSIS OF EXPLOSIVES

Moisture.-A sa~ple of about 5 grams spread on a watch glass is desiccated

I
over sulfuric acid to constant weight.,
Insoluble.-A sample of about 10 grams is treated with 150 m!. of benzene or
95% alcohol, heated to boiling, and filtered while hot tlirough a tared Gooch
crucible with asbestos mat. The insoluble residue is' ~ashed, with additional
\
solvent, dried at 100 C. and weighed.
Acidity.-A 10-gram sample is melted in a large test tube or a flask and
shaken with 100 ml. of neutralized boiling water, cooled and the water decanted. A similar treatment is given using 50 m!. of boiling water, the twp
portions of water combined, cooled and titrated with tenth normal NaOH,
using phenolphthalein indicator. The acidity is calculated as % H 2S0 4 in the
original sample.
Nitrogen.-Nitrogen is not usually determined in the inspection of TNT
but when necessary it may be d~;termined by the Dumas c'ombustion method
or the modification of the Kjeldahl method described on page 1671.
Picric Acid
Ordnance Department, U. S. A., specifications for picric acid prescribe that
it shall have a solidification point of not less than 120 C.; that it shall contain
not more than the following amounts of impurities:
.
Moisture-().2% for dry material. .. / ............ "
Sulfuric acid (free and combined) . . . . . . . . . . . . . . . ..
Ash ........................................... '
Insoluble in water ............................. "
Soluble lead... . ................................
Nitric acid (free) .............................. "

22.0% for wet

0.10%
0.2%
0.2%
0.0004%
none

'Solidification Point.-Dry the sample at a temperature not exceeding 50 C.

Melt sufficient to give a 3-inch column in a 6-inch X %-inch test tube immersed
in a bath of glycerin heated to 130 C, When the sample is completely melted
remove the tube from the bath and stir the sample with a standardized thermometer graduated in 0.10. degrees, until the picric' acid solidifies. During
solidification the temperature will remain constant for a short time and then
undergo a slight rise. The highest temperature reached on this rise is recorded
as the solidification point. The test may be more accurately carried out using
the apparatus and methoq as described under trinitrotolue.Il~ (p. 1681).
Moisture.-A weighed sample of about 10 grams is spread, evenly on a
tared watch glass and dried to constant weight (about 3-4 hours] at 70 C.
Sulfuric Acid.-About 10 grams is weighed and dissolved in 250 m!. of boiled
distilled H 2 0, acidified with HCI and heated to about boiling. Hot BaCl 2
solution is added with stirring and the mixture allowed to stand at least 1 hour
on the steam bath. Filter hot on a tared Gooch crucible, wash with water,
dry at 100 C. and weigh. Calculate BaS04 found as H 2S0 4 in original sample.
Ash.-About 5 grams is weighed in a crucible, moistened with sulfuric acid,
burned carefully, and the residue ignited to burn off all carbon, The resulting
ash is cooled and weighed.
I
Insoluble in Water.-l0 grams of the sample i.;ti:eated with 150 m!. boiling
distilled water, boiled for 10 minutes, filtered wnile hot through a tared Gooch

ANALYSIS OF EXPLOSIVES

1683

crucible, washed well with hot water, and the insoluble residue on the filter
dried at 100, cooled, and weighed.
Soluble Lead.-The presence of soluble lead in picric acid is highly objectionable, because lead picrate is an extremely sensitive explosive and its presence would greatly increase the dangers involved in handling and loading
picric acid. A weighed sample of about 300 g. is digested in a 2-liter flask
with 100 mI. of a hot saturated solution of barium hydroxide in 65% alcohol.
1400 mI. of 95% alcohol is then added and the digestion continued at a temperature below the boiling point (with reflux condenser), until everything
except traces of insoluble matter is in solution. The picric acid is then allowed
to crystallize on cooling, and the solution filtered off, decanting the clear liquid'
from the crystals until 500 mI. of filtrate is obtained. This 500 mI., representing 100 g. of picric acid, is treated with 5 drops HNO a and 10 mI. of 0.1 %
HgCb solution, and H 2S passed through it for 15 minutes. Allow the precipitate to settle for 20 minutes, filter and wash with alcohol saturated with
H 2S. Dry and ignite the precipitate, then dissolve the residue in 9 mI. of
HNO a (sp.gr. 1.42) by warming, add warm water to bring the volume to 50
mI., and electrolize at 0.4 ampere and 2.5 volts, temperature 65 C., for 1 hour.
Wash the electrode by replacing the beaker with another one containing distilled water without interrupting the current.. Dry and weigh the previously
tared anode. The weight of lead peroxide foundXO.8661 gives the percentage
of soluble lead found.
Nitric Acid.-No coloration should result when a water solution of picric
acid is treated with a solution of diphenylamine in sulfuric acid.
Ammonium Picrate
Ammonium picrate, also known in this country as "Explosive D," is of
importance as a military explosive more on account of its insensitiveness to
shock and friction, than because of its explosive strength, which is less than
that of TNT. Its chief use is as a bursting charge in armor-piercing projectiles.
Military specifications require it to be prepared from picric acid of standard
purity, to contain not less than 5.64% ammoniacal nitrogen, and not more
than the following amounts of impurities:
. '
Moisture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Insoluble material. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Ash ..................................................
Free a~~oni~ (NHa) ...................................
Free PICflC aCId ...... " ................................

0.20%
0.20%
0.20%
0.025%
0.025%

Moisture.-A ;ample of about 10 grams spread on a tared watch glass is

dried at 95 C. to constant weight (about 2 hours).
Insoluble Material.-A 10-gram sample is boiled with 150 ml. of water for
10 minutes, filtered on a Gooch crucible, the residue washed with hot water,
dried at 100 C. and weighed.
Ash.-A sample of about 1 gram is saturated with melted paraffin and
burned in a tared crucible, the residue ignited to burn off all carbon, and the
ash weighed.
'Free Ammonia or Free Picric Acid.-A sample of about 5 grams is ground in a
porcelain mortar with successive 50 mI. portions of water until the sample is
entirely dissolved, each portion of the water being poured through a filter into

1684

ANALYSIS OF EXPLOSIVES
J

a 500-ml. Erlenmeyer flask. Add 3 to 5. drops of a saturated solution of

sodium alizarin sulfonate or of a 1% alcoholic solution fof methyl red and .note
whether the solution is acid or alkaline. Both indicators are yellow when
acid. In alkaline solutions sodium alizarin suifonate isl purple and methyl. red
is red. Titrate the solution with N/l0 acid or alkali. Subtract one drop
from the volume used to allow for the amount required '.to convert the solution
from neutrality to acidity or alkalinity. Calculate as follows:
01
/0
01

/0

ree ammOllla =

A XO.17B
W
'

..
'd A X2.29B
ree pICrIC aCI =
W
'

where A =ml. N/l0 solution used,

B = normality factor of solution,
W = weight of sample used.
Tetryl
Tetryl is the commercial term applied tQ the explosive trinitrophenylmethylnitramine, also improperly called tetranitromethylaniline. Its chief
use is as a " booster" charge in high explosive shell, where it serves to transmit
the 'detonating wave from the detonator or fuze to the less sensitive bursting
charge. Being in immediate contact with}he fuze it must be of a high degree
of purity, and is required by Ordnance Department specifications to have a
melting point of at least 128.5 C. and to contain not more than the following
amounts of impurities:
Moisture .............................................. 0.10%
Insoluble in Benzene .................................. , 0.40%
Acidity (as H 2SO.) ... : ................................. 0.10%
Melting Point.-The sample to be used for this test is dried overnight in a
vacuum desiccator or 2 hours at 95-100 C .and pulverized to pass a 100-mesh
screen. A capillary melting-point tube is filled to ibout U inch from the
bottom and attached to the stem of a standard thermometer so that the sample
is next to the center of the bulb. The bath is properly agitated and provisioll
made for correcting for the emergent stem of the thermometer. The temperature of the bath is raised rapidly to 120 G., then at the rate of lOin 3
minutes, the temperature ,at which the first meniscus appears across the capillary tube being noted as the melting point.
'
Moisture.-A sample of about 5 grams is weighed in a wide shallow weighing
bottle and dried over sulfuric acid in a desiccator for 48 hours, the sample
being. spread uniformly so that its depth is not over 0.5 cm. The loss of weight
is regarded as moisture.
Insoluble in Benzene.-5 grams of sample is dissolved in 75 ml. of benzene,
filtered through a tared Gooch crucible, and the residue washed with three
25-ml. portions of acetone, dried to constant weight at 100 C. and weighed.
Acidity.-':The benzene filtrate from the determination of insoluble material
is shaken with 100 mI. cold boiled distilled water in a separatory funnel, the
water layer removed and the benzene solution giv_!lri.a second extraction with
50 ml. more water. The combined water extracts' are titrated with N/20
NaOH solution, using phenolphthalein indicator.

ANALYSIS OF EXPLOSIVES

1685

Mercury Fulminate
In commercial blasting caps and electric detonators mercury fulminate may
be found intimately mixed with potassium chlorate. It is, however, used
without admixture in certain types of detonators, in the fuzes of high explosive
shell and for other military purposes. It is usually purchased under specifications which provide that it shall be at least 98% pure, shall be free from acid,
and contain not more than 2% insoluble matter, 1 % free mercury, and 0.05%
chlorine in the form of chlorides.
Preparation of Sample.-Mercury fulminate being packed and handled in a
thoroughly wet condition until dried just before use, it is generally necessary
to dry the sample before testing. This may be done by exposing in a lo,v
temperature oven at not more than 50 C. until practically dry, then in a
desiccator (not a vacuum desiccator)' over sulfuric acid or calcium chloride until
its weight is constant.
Mercury Fulminate Content.-Exactly 0.3 g. is weighed into a widemouthed Erlenmeyer flask of about 250 ml. capaqity, and 30 ml. of a 20%
solution of purest sodium thiosulfate is added quickly and the mixture shaken
for exactly 1 minute. At once titrate with N/lO hydrochloric acid using 3 drops
of methyl red indicator, the titration to be commenced 1 minute after adding
the sodium thiosulfate, and to occupy not more than 1 minute additional time.
The percentage of mercury fulminate is calculated from the volume of
standard acid required, after deducting the volume of acid required for, a blank
determination. Four molecules of HCI are equivalent to 1 mol. of m~rcury
fulminate, or 1 ml. N/I0 HCI equals 0.00711565 g. mercury fulminate. The
< reaction is assumed to be as follows:
HgC2N202+2Na2S203+2H20=HgS406+4NaOH+C2N2.
Acidity.-A 10-g. sample is extracted with 2 successive 25-ml. portions of
boiled distilled water in a Gooch crucible, and 3 drops of methyl orange solution
(1 g. per liter) added to the solution. No red tinge of color should be obtained.
Insoluble Matter.-A 2-g. sample is dissolved in 20% Na 2S203 solution,
filtered through a tare.d Gooch crucible and any insoluble washed several times
with water and dried to constant weight at 60-70 C.
Free Mercury.-The residue of insoluble matter obtained as described
above is treated with a solution of 3 g. KI and 6 g. Na 2S20a in 50 ml. H 20 by
passing the solution through the Gooch crucible. Any organic mercury compounds are thus converted into mercuric iodide, which is soluble in Na2S20a
solution. The metallic mercury remains behind on the filter, and is washed
with H 20, dried 1 hour at 80-90 C., and weighed.
Clilorides.-A 5-g. sample of fulminate is extracted in a Gooch crucible
with 2 successive 25-ml. portions of distilled water at 90-100 C. Three drops
of concentrated HNO a and 10 drops of 10% AgN0 3 solution are added to the
filtrate. If a turbidity results, the AgCI should be determined gravimetrically
or a fresh sample extracted and the filtrate titrated with a standard AgNO a
solution.
Lead Azide
Lead azide, PbN 6, is a primary or detonating explosive. Although it has
been employed in several European countries for many years as an ingredient of

,
1686

ANALYSIS OF EXPLOSIVES
I'

detonators and primers for both commercia,! and military purposes, its use in
the United States is comparatively recent. Its prin<iipal application is as a
priming charge in compound detonators, where it is superimposed on a main
charge of tetryl, TNT or other high explosive. N orrfally a crystalline compound, it is extremely sensitive when in large crystals, and hence is prepared for
practical use in the microcrystalline state, or in the form of small rounded
aggregates. Its natural white color may be slightly darkened by exposure to
light. It should be free from acidity. Its lead content may be regarded as an
indication of its purity, provided no other lead compounds are present. Pur~
PbN 6 contains 71.15% Pb. The sample taken for analysis should be dried a~
about 95 C.
Acidity.-A 10-gram sample is extracted in a Gooch crucible with 100 ml.
of previously boiled, cold distilled water in 20 ml. portions. The filtrate is
tested for acidity with methyl orange indicator.
Lead Content.-A weighed sample (approximately 1 g.) is dissolved in 50
ml. of saturated solution of ammonium acetate by warming in a 400-ml. beaker.
The solution is diluted with 200 ml. distilled water, heated to boiling, and 10 ml.
of a 10% solution of K 2 Cr 2 07 added with rapid agitation. After digesting on
the steam bath for 1 hour, the precipitate is collected in a Gooch crucible.
washed with hot water, dried at 100 C .and weighed. PbCr04XO.64109=Pb.
Blasting Caps and ElE;ctric Detonators
The commercial blasting caps and electric detonators loaded with a mixture
of mercury fulminate and potassium chlorate, which were in general use in the
United States up to a few years ago, have now been largely displaced by compound detonators in which a priming charge of mercury fulminate, lead azide,
diazodinitrophenol, hexanitromannite or other primary explosive is employed.
to initiate a main charge of tetryl, TNT, picric acid or other nitro compound.
Preparation of Sample.-In the examination of blasting caps or detonators
for either commercial or military use, .the removal of the/detonating composition
from the copper, aluminum or brass shell requires ~onsiderable precaution.
Blasting caps are emptied by squeezing the cap gently in a pair of" gas forceps,"
the jaws of the forceps being passed through a small opening in a piece of heavy
leather, rubber belting, or similar material, about 6" square, which serves as a
shield to protE)ct the hand in case of the possible 'explosion of the cap in squeezing. After each squeeze, the loosened portion of the charge is ~~aken out on a
piece of glazed paper, the cap turned slightly in the forceps and again squeezed.
The pressure on the cap should be just sufficient to slightly dent it, and in
shaking out the charge, the cap should not be tapped on the table or other
surface. With these precautions there is little "danger of an explosion: Another method of removing the charge is rolling the cap between two pieces of
hardwood board, emptying the loosened portion of the charge after each rolling.
Electric detonators 15 are opened by first cutting off the' wires or "legs"
close to the shell, then tearing off the upper portion of the shell by means of
pointed side-cutting pliers, the cap being held firmly in the fingers, and a thin
15 A safe apparatus for cutting open electric detonatof&..is_shown1n Bureau of Mines
Technical Paper No. 282, "Analysis of Detonating apdlPtiIIling Mixtures," by C. A.
Taylor and W. R. Rinkenbach, Plate 1.
-

ANAL YSIS OF EXPLOSIVES

1687

'Strip of the copper shell being torn off spirally by nipping the top edge of the
shell 'Yith the forceps. This must be done with great care, especially as the
portion of the shell containing the fulminate charge is approached. When the
greater portion of the plug which holds the wires in place has been exposed,
the plug and wires are gently pulled out, care being taken to avoid .force and
possible friction, and any adhering particles of the charge brushed off onto
glazed paper. The charge is then removed from the lower part of the shell
just as in the case of blasting caps.
The charge is removed separately from several of the caps or detonators
and each weighed in order to determine the average weight of charge as wen
as variation of same.
II Reinforced" caps, or those which contain a small perforated inner copper
capsule pressed on top of the charge, must be opened in the manner described
for electric detonators, in order to remove the inner capsule. Detonators of
this type usually contain a main charge of some nitro compound superimposed
\)y a layer of mercury fulminate, mixture of fulminate and chlorate, or lead
azide. Although a clean mechanical separation of the two layers is usually not
possible, portions can be taken from each and identified by qualitative tests
before proceeding with a quantitative examination.
Moisture.-The moisture content of the composition is determined by
desiccating to constant wlJight over sulfuric acid or calcium chloride.
Analysis of Composition Containing Mercury Fulminate and Potassium
Chlorate.-About 2-3 grams of the well-mixed composition is weighed in a
Gooch crucible provided with asbestos mat or disc of filter paper or silk, and
first moistened with a few drops of alcohol, then extracted with 200-250 m!. of
cold water in 15-20 m!. portions, using slight suction after each portion has
remained in the crucible for a few minutes. The residue in the filter is dried
to constant weight at 60-70 C. (2-3 hours), and weighed.
The water extract contains the potassium chlorate and a portion of the
mercury fulminate, which is slightly soluble in cold water. It is treated with
2 m!. of ammonium hydroxide and H 2S passed to completely precipitate the
dissolved mercury fulminate as HgS. This black precipitate is filtered off,
washed, dried and weighed. Its weight X 1.22 gives the amount of mercury
fulminate dissolved by the water. This weight added to the weight of the
dried residue insoluble in water gives the total weight of mercury fulminate in
the sample. The KCI0 3 is found by subtracting the per cent of mercury fulminate+per cent moisture from 100%.
Ana)ysis of Compositions Containing Nitrocompounds.-Trinitrotoluene,
tetryl or picric acid can be identified by melting point test, TNT melting at
about 79-80 C., tetryl at about 128 C., and picric acid at about 120-122 C.
They may be extracted from the mixture by means of ethyl ether, in which
mercury fulminate is only very slightly soluble, and the determination of KCIO a
and mercury fulminate then made as described in the preceding paragraph.
If the main charge is an organic nitrate such as nitrated vegetable ivory,
nitrostarch, nitromannite etc., such material will be left with the mercury
fulminate in'the insoluble residue after extraction with water. The mercury
fulminate is then extracted by means of a hot 20% solution of sodium thiosulfate, leaving the organic nitrate in the Gooch crucible. These materials in
the detonating composition can be readily identified by microscopic examination.

ANALYSIS OF EXPLOSIVES

1688

In detonators where TNT or tetryl compose the mainj portion of the charge,
a small amount of lead azide, with or without mercury fulminate, may "Qe used
as a priming charge for the purpose of initiating the drtonation of the nitro,
compound. It should be identified in the top portion of the charge, next
to the reinforcing cap, and will in all probability be presertt if mercury fulminate
is not found. It is practically insoluble in water and -in! ether, and will be left
in the insoluble residue. If present, fulminate is destroyed by treating the
residue, in a flask, with 25 mI. of KOH solution. This converts the lead azide
to potassium azide, KN a A slight excess of H 2S0 4 ill added and the mixture
distilled, the distillll,te, containing HN g, being collected in water. Enough
NaOH is added to the distillate to give an alkaline reaction with litmus, then
a little Pb(NO g)2, wh.en lead azide, PbN 6 will be regenerated as a white precipi:tate, which may be filtered off, washed with water, then with alcohol, dried
in the air, and tested by striking a small portion with a hammer.
Primers
Variations in Composition.-Many varieties of composition are used in
primers for small a:t'ms ammunition, and for other military purposes. The
composition must bE:) ignited by the impact of the firing pin, and must give It
flame of sufficient intensity and duration to 'ensure proper ignition of the pro,
pellant or of the detonator, depending on/ the purpose for which the primer is
employed. As primers are used with various kinds and granulation of explo,
sives, a priming composition suitable for one purpose is unsuited for another;
hence there are many types of priming compositions. A few of the types which
were in common usE:) up to comparatively recent years are shown in Table 1
below. Table 2 giV(lS a few examples of the great variety of patented type~ ot
the more modern "non-corrosive" primer compositions, in which potassiuln
chlorate is eliminated because\ the chloride formed as one of the products of
combustion was found to be responsible for corrosion of the gun barrels. "N on,
corrosive" primers have displaced many of the older ,types.
,

'l~YPES

OF PRIMER COMPOSITIONS
APPlioXIMATE COMPOSITION (PER CENT)
TABLE

,
No, 1 No.2 No.3 No.4 No.5 No.6 No:T 'No~8 No.9 No.l(j

_-_- - _Mercury fulminate ........ 31

25
Ingredien ts

Potassium chlorate ........

Sulfur ...................
Powdered glass ...........
Lead sulfocyanate ........
Copper sulfocyanate ..... ,
Barium nitrate ...........
TNT ...................
Tetryl .....' .............
Antimony sulfide .........
Lead oxide (PhO) ........ ,
Shellac .... " ............
Black powder (meal) ......

--- - -

38
-

53
-

60 50
7 , 3

11

38

31
-

31
-

30
-

36

51
9

.'

-'

44
::;;--

53
-

12

_-

26
2

3- -

25

_--53

25

17-

-,

17

--.
28
47 14
22 - 35

31

21

2
-

ANALYSIS OF EXPLOSIVES
TABLE

Mercury fulminate ........

Lead styphnate ..........
Barium nitrate ...........
Lead nitrate ...... , ... : ..
Lead peroxide ............
Barium peroxide .........
Calcium silicide ..........
Antimony sulfide .........
Lead sulfocyanate .........
Diazodinitrophenol .......
Tetracene ...............
Basic lead picrate .........
Lead dinitrosali<lylate .....
Abrasive ................

38
39
-

5
11

5
-

30-35
-

25-45
10-25
-

0-25

1689

25-40
5-10
24-44'
-

0-16
-

4-10
0-30

30
-

7
6-8
-

38-36
18
-

30-31.5
10-5.5
29-30.5
10-10.5
-

20-21

37
32
-

35-40
0-16
30-12
-

28
-

12-10
22-21

In addition to these ingredients most priming compositions are mixed with

small amounts of some binding material dissolved in water or alcohol, such as
gum arabic, gum tragacanth, glue, shellac, etc. These traces of binding materials are u.sually disregarded in the analysis of the compositions.
Preparation of Sample.-If the caps contain anvils, these must first be
carefully removed, as well as any covering of tin foil or paper. The primer
composition is then carefully removed from a number of primers and weighed
to determine the average charge. It is then carefully crushed, a little a,t a time,
and the sample well mixed. If necessary, it may be removed from the caps
by the aid of water or alcohol and the latter removed by evaporation before
weighing.
.
QUalitative Examination.-The following special tests may be made use of
in connection with a qualitative analysis of the mixture:
A small amount is burned between two watch glasses, the formation of a
mirror indicating mercury, antimony, copper or lead. The mercury mirror is
readily volatile on gentle ignition.
Extract a portion of the mixture with ether, then with water, then with
Na 2S 20a solution, then with aqua regia, retaining each of these solutions.
TNT or tetryl may be present in the ether solution and are identified by
m.p. or color test, TNT giving a deep red color with acetone and KOH.
Sulfur is detected by burning a portion of the ether-soluble material and noting
odor of S02.
The water extract is tested for KCIO a by adding H 2S0 4 , boiling, and noting
odor of chlorine. A portion is treated with HCI and FeCIa, a red color indicating thiocyanate. The usual FeS04 ring test is made for nitrates. A white
precipitate with H 2S0 4 indicates Ba or Pb.
The aqua regia solution is diluted and tested with H 2S for antimony, lead,
and copper. If the precipitate is not orange-red, lead or copper are indicated.
Dissolve in HNO a, neutralize with ~H40H; a blue solution indicates copper,
while lead is detected by the formation of a white precipitate with H 2S0 4
Any material insoluble in aqua regia may be powdered glass or other
abrasive material.
Quantitative Analysis.-The method of analysis will depend entirely upon
the ingredients indicated by qualitative tests. In general, a separation is best

1690

ANALYSIS OF EXPLOSIVES
J

effected by successive extractions with ether, water,' Na 2S203 solution (to

remove fulminate), dilute or concentrated HCI, a,nd aqua regia. The small
amount of mercury fulminate present in the water extract may be determined
by precipitation with H 2S or by, adding 10-15 ml. of th~osulfate solution and a
few drops of methyl orange and titrating with NII0 H;CI or H 2S0 4 (see page
513 of Vol. I). Other materials in the water and acid solutions are determined
by the usual analytical methods.
I
Nitrocellulose'
Genera1.-The term nitrocellulose, or more correctly cellulose nitrate,
applies to any nitration product of cellulose, ranging from products containing
in the neighborhood of 10-11 % N, which are used in the preparation of lacquers
and other commercial products, to military guncotton ~ith over 13% N. All
of these products are undoubtedly mixtures of the various nitrates of cellulose,
as indicated by the fact that there is always .some material with low nitrogen
content, soluble in ether-alcohol, in high nitrogen' guncotton, and some insoluble
material in the lower nitrated commercial products. It can usually be shown
without great difficulty that any nitrated cottori is a mixture of various nitrates
of cellulose.
The products of military importance are the insoluble guncotton of high
N-content, and the so-called" pyro " or pyrocellulose, soluble in ether-alcohol
and of about 12.60% N-content. In testing these products, the characteristics
of most importance are content of nitrogen, solubility in ether-alcohol, and
stability. Other determinations generally made are solubility in acetone and
ash.
,
Preparation of Sample.-If the sample contains a large excess of water, it
is enclosed in a clean cloth and the excess water removed by means of a press
or wringer. The pressed sample is then rubbed up in the cloth (not with the
bare hand) until lumps are removed, then spread on clean paper trays in an
air bath at .about 35-40 C. until" air-dry."
!.
Samples for stability tests and nitrogen deter;mination are treated as noted
below, the air-dry sample being suitable for determining solubility and ash.
Nitrogen.-About 1 to 1.05 g. of the air-dry sample is roughly weighed in a
tared weighing bottle, dried at 98-102 C. for 1% hQurs, cooled in a desiccator
and accurately weighed. It is then transferred to the generating bulb of a
nitrometer (Du Pont modification; see p. 650) using a total'Of ~O ml. of 94-95%
c.p. H 2S0 4, The sample must be dissolved in the acid either in the weighing
bottle or in the cup of the generator, before it is drawn into the generating
bulb, and both the weighing bottle and the cup of the generator must be
thoroughly washed out with the 20 ml. of H 2 S0 4, so that none of the sample is
losJ? The determination in the nitrometer is completed in the usual manlier
(p. 650), the result being expressed as per cent N in'the dried sample of nitrocellulose.
'
Solubility in Ether-Alcohol.-(a) Guncotton: The amount of ether-alcohol
soluble material in guncotton being usually not more than 10-12%, the determination may be made by evaporating a clear solution. Tw,o grams of air-dry
sample is placed in a clean dry cork-stoppered .J50~ml. cylinder, 67 .ml. of
95% ethyl alcohol added to thoroughly wet tlie guncotton, then 133 ml. of

ANALYSIS OF EXPLOSIVES

1691

ethyl ether (U.S.P. grade, 96%), added arid the mixture well shaken. If the
mixture of 2 parts ether and 1 part alcohol be added at once to the sample, a
gummy mass may result which dissolves with great difficulty, especially if the
solubility is unusually high.
.
The cylinder is now allowed to stand at a constant temperature of usually
15.5 C. The solubility of nitrocellulose increases as the temperature is
decreased, hence a constant temperature of digestion is important. During the
digestion, which requires at least 1 hour, the cylinder must be thoroughly
shaken at 5-minute intervals. The cylinder is now allowed to stand for at
least 4 hours, until the insoluble portion of the sample has completely settled
and the supernatant liquid is perfectly clear.
Fifty m!. of the clear solution is now drawn off with a pipette, care being
taken not to disturb the settled pulp, and evaporated in a weighed evaporating
dish on a steam bath, avoiding loss from violent boiling of the ether. When
25-30% of the solution has been evaporated, 10 m!. of distilled water is added
slowly and the evaporation continued to dryness. The effect of the water is to
leav.e the residue in a white, brittle or powdery condition, rather than a tough
film which would lose its solvent with difficulty.
The dish is finally placed in an oven at 95-100 C. for Y2 hour, cooled in a
desiccator, and weighed. The weight of the residue, corrected for the residue
in the 50 m!. of ether-alcohol and 10 m!. H 2 0 used, represents the soluble nitrocellulose in 0.5 g. of the guncotton.
(b) Pyrocellulose: The solubility of pyrocellulose may be determined in the
manner described for guncotton, but owing to the much larger amount of
soluble material present, the evaporation of the residue to constant weight
without decomposition involves considerable difficulty. Sufficient water must
be added to precipitate the soluble nitrocellulose from solution in a stringy or
fibrous condition.
The detetmination is usually conducted by either the volumetric method or
the gravimetric method.
In the volumetric method, one gram of the air-dry sample is covered with
75 m!. of 95% ethyl alcohol and allowed to stand for about 2 hours with frequent
stirring, 150 m!. of ethyl ether is then added with stirring and the agitation
continued until solution is complete. The solution is now allowed to stand at
least 4 hours with frequent stirring, during at least 1 hour of which time it is
to be kept at a temperature of 15.5 C. It is then transferred to a " solubility
tube" and allowed to stand for about 16 hours, in order that the insoluble
material may settle completely. The solubility tubes are heavy glass tubes
about 21 inches long and 1.35 inches inside diameter, tapering gradually near
the bottom to a constricted portion about 2 inches long and about 0.3 inch
inside diameter. This narrow bottom portion is graduated in one-tenth ml.
The volume of the settled insoluble material is read in ml. from the scale on
the tube. If the reading is 0.25 ml. or less, the percentage of insoluble is
considered to be 1 % or less. If an exact determination is desired the gravimetric method must be used.
In the gravimetric method the solution is prepared and settled in a solubility tube as described above and the clear liquid removed as completely as
possible by means of a narrow siphon tube of glass. Fresh alcohol and ether
are then added as before, the tube shaken and allowed to stand again until the

1692

ANALYSIS OF EXPLOSIVES

insoluble has settled, when the process may be repeated several times, depend~
ing on the amount of insoluble material present. After the last decantation,
the residue is washed from the tube to a beaker, using as small a quantity of
ether-alcohol as possible, and the mixture filtered th~ough a weighed G'C'Joch
crucible with ignited asbestos mat, the insoluble washed, free from soluble nitrocellulose with ether-alcohol, dried first at a low temperature and finally at
100 C. to constant weight, cooled and weighed. The crucible is then carefully
ignited over a free flame, cooled and weighed again. The loss of weight on
ignition is considered as insoluble nitrocellulose.
Solubility in Acetone.-A I-gram sample of air-dry pyrocellulose is treated
with about 200 m!. of acetone with frequent stirring until all gelatinous matter
has dissolved. The solution is transferred to a solubility tube (described above),
well shaken and allowed to settle for about 16 hours. If the volumetric reading
is 0.2 m!. or less, the percentage of insoluble is considered to be 0.4% or less.
A gravimetric determination may be made as described in the preceding
paragraph.
Ash.-One gram of air-dry sample is weighed in a tared crucible, moistened
with 10-15 drops of concentrated nitric acid; and digested for 2-3 hours on a
steam bath until converted to a gummy mass. The crucible is theri heated
carefully over a Bunsen. burner until the mass is completely charred, then at a
red heat until its weight is constant. The residue is the ash of the sample.
Stability Test: Heat Test with Potassium Iodide StarchPaper.-The " heat
test" or KI test, as it is commonly dffsignated, is the test most commonly
employed for determining the stability or degree of purification of nitrocellulose,
whether guncotton or pyrocellulose. This test, also referred to as the Abel test,
depends on the action of oxides of nitrogen liberated by the nitrocellulose under
the ir\fluence of heat, the gases in contact with the KI-starch paper liberating
iodine which colors the starch.
The sample is dried with great care to avoid contamination, in a clean paper
tray, at 35 to 40 C., until its moisture is reduced to the amount which will
give the minimum heat test, usually 1.5 to 2%. The/proper amount of moisture is determined as follows: During the progress of the drying, the sample
on the tray is "rubbed up " from time to time, using a piece of clean tissue
paper spread over the back of the hand. When the sample begins to adhere
to the paper, due to static electricity, a sample of 1.3 g. is weighed into a standard test tube. These tubes are 5Y2 inches long, not less than Y2 inch inside
diameter and not more than %" outside diameter, made ~6f glass about 3/64
inch (1.2 mlIi.) thick. As soon as the first sample is weighed, the tray is
replaced in the drying oven for 2-5 minutes, a second sample weighed, and this
process repeated until a series of 5 samples have been taken, the last sample
being completely dry. This serIes of samples, if properly taken, will cover the
range of moisture content giving the minimum heat test. If the sample in the
tray appears to have become too dry during the time the weighings are being
made, it may be placed in a moist atmosphere for not more than 2 hours; the
entire time of drying and making the test must not exceed 8 hours.
The tubes containing the samples are fitted with clean, fresh cork stoppers
through which pass a piece of glass rod into the enq pf which is fused a small
piece of platinum wire ,bent into a hook. The wire--iS-heated in a flame to clean
it, a piece of the standard KI starch test paper, l"X%", attached, taking care

ANALYSIS OF EXPLOSIVES

1693

that' neither wire nor paper are touched with the fingers, and the paper moistened on its upper portion by touching it with a glass rod dipped in a solution
of equal volumes of pure glycerin and water. The stoppers are then inserted
in the tubes and the tubes placed in a constant temperature water bath, so
that they are immersed to a depth of 2.25 inches. The time of placing in the
bath and the time of the appearance of the first faint yellowish discoloration
of the test paper are noted. The minimum test given by the 5 samples is
taken as the result of the test. The discoloration appears at the lower edge
of the moist portion of the paper. The temperature of the heat test bath is
65.5 C. (150 F.) for pyrocellulose, and usually 76.5 C. (170 F~) for guncotton. Pyro is usually required to stand a test of 35 minutes, and guncotton
10 minutes.
A standard test paper is absolutely essential, and is prepared as follows: 16
The paper used in preparing the test paper is Schleicher and SchUll's filter
paper 597. This is cut in strips about 6 by 24 inches, and after b~ing washed
by immersing each strip in distilled water for a short time is hung up to dry
overnight. The cords on which the paper is hung are clean and the room is
free from fumes. The washed and dried paper is dipped in a solution prepared
as follows:
The best quality Gf potassium iodide obtainable is recrystallized three times
from hot absolute alcohol, dried, and 1 gram dissolved in 8 ounces of distilled
water. Cornstarch is well washed by decantation with distilled water, dried
at a low temperature, 3 grams rubbed into a paste with a little cold water,
and poured into 8 ounces of boiling water in a 'flask. After being boiled gently
for 10 minutes, the starch solution is cooled and mixed with the potassium
iodide solution in a glass trough.
Each strip of filter paper is immersed in the above-mentioned mixture for
about 10 seconds and is then hung over a clean cord to dry. The dip=-,ing is
done in a dim light and the paper left overnight to dry in a perfectly dark room.
Every precaution is taken to insure freedom from contamination in preparing
the materials and from laboratory fumes that might cause decomposition.
When dry the paper is cut into pieces about % by 1 inch and is preserved in the
dark in tight glass-stoppered bottles, the edges of the large strips being first
trimmed off about one fourth inch to remove portions that are sometimes
slightly discolored. When properly prepared the finished paper is perfectly
white, any discoloration indicating decomposition due to contamination.
Stability Test at 134.5 C.-In addition to the KI starch test, pyrocellulose
is usually required to stand a test at 134.5 C., made as follows:
The sample is completely dried at about 40 C., and 2.5 grams placed in each
of 2 heavy glass tubes, 290 mm. long, 18 mm. outside diameter and 15 mm.
inside diameter, closed with a cork stopper through which passes a hole 4 mm.
in diameter. A strip of litmus paper or standard normal methyl violet paper,
70 mm. long and 20 mm: wide, is placed in each tube, its lower edge 25 mm.
above the sample, which is pressed down to occupy a depth of 2 inches, the
walls of the tube being wiped clean with a roll of paper. The tubes are then
heated in a constant temperature bath at 134 to 135 C., all but about 6-7 mm.
of the tube being immersed in the bath. They are partially withdrawn for
16 Storm, C. G., Proc. 7th Inter. Congress App!. Chern., 1909; J. Ind. Eng. Chern., 1,
802 (1909).

1694

ANALYSIS OF EXPLOSIVES

examination of the test papers every 5 minutes after .the first 20 minutes of
heating, and replaced at once. The time required for reddening of the litmus
paper or for turning the methyl violet paper to a salmon pink color is noted
as the time of the test. A minimum test of 30 minutes is required with the
methyl violet paper, and heating is then continlied for altotal of 5 hours, during
which time there should be no explosion.
I
The standard normal methyl ,violet paper is prepared as follows:
Preparation of Methyl Violet Test Paper.-A solution is prepared containing the following ingredients: pure rosaniline acetate prepared from 0.2500 g.
pararosaniline, .1680 g. methyl violet (crystal violet), 6 ml. c.p. glycerin, 30 mI.
water, and sufficient pure 95% ethyl alcohol to make. up to 100 ml. This
solution is placed in the angle of an inclined deep rectangular glass tray, and
large sheets of Schleicher & Schull filter paper (No. 597) cut in four strips are
dipped in it. In dipping, the strip is held by one end and dipped to within Y4"
of this end, withdrawing it slowly up the side of the tray so as to remove surplus
solutidn. The strip is then held horizontally and waved to and fro so as to
prevent the solution from running .and collecting in spots. As soon as the
alcohol has evaporated the strip is suspended yertically to dry, and when dry
is cut in strips 20 X 70 mm. These strips are bottled and kept for use in the
134.5 C. test.
Viscosity.-In the case of commercial nitrocelluloses it is frequently necessary to determine the viscosity of standard solutions. The solutions used for .
this purpose are usually as follows: 17
'
% by Weight
, Formula
A

Nitrocellulose (dried at 50 C.) .. , ...... , ............ , ... , ..

Ethyl acetate (85--88% grade). , ... , ...................... ,
Denatured alcohol, C.D. No.5 (188-190 proof) ......... j ...
Toluol (2 C. range, incl. 110.7) .... , , ........... 0'.

12.2
17.5
22.0
48.3

- -- 20.0 25,0
16.0 15.0
20.0 18.75
44.0 41.25

Formula A is used unless viscosities obtained,with it are 6 seconds or less, in

which case Formula B is used. If B gives below 3 seconds Formula C is used.
A glass cylinder 1" 0.02" inside diameter and 14" high, with marks 10"
apart at points 2" and 12/' from the top, is filled with the solution to be tested,
the cylinder stoppered and allowed to stand until all bubbles hav~ disappeared,
then brought to a temperature of 250.1 C. A steel ball 0.312" to 0.313"
~n diameter, weighing,2.035 g.O.Ol g., is then. released at the center. of the
surface of the solution, and the number of seconds required for the ball to pass
through the lO-inch column of solution between the marks on the. cylinder is
recorded ~s the viscosity of the sample with the Formula used.
17 A. S. T._ M. Standard Specification for Soluble Nitrocellulose (D 301-331).

ANALYSIS 'OF' EXPLOSIVES

1695

SMOKELESS POWDER
Nitrocellulose Powders
At the present time the smokeless powder used by all nations is composed
of either colloided nitrocellulose or a mixture of colloided nitrocellulose and
nitroglycerin with or without the addition of other ingredients. All cannon
powder used in this country is of the nitrocellulose type, small-arms powders
being of both types. The. form and size of the grains are of great variety,
depending on the arm in which the propellant is to be employed.
Physical tests made in connection with the examination of smokeless powder
include the compression test, determinations of average measurements of the
grains, specific gravity, gravimetric density, number of grains per pound, and
calculation of burning surface per pound.
Chemical tests include determinations of moisture and volatile solvent,
diphenylamine used as stabilizer, ash, material insoluble in ether-alcohol and
in acetone, and sometimes nitrogen content.
Stability tests include the 134.5 C. test, and the" Surveillance test."
Moisture and Volatiles.-A sample of the powder weighing approximately
1 gram, in the form of thin shavings cut from at least 10 grains, or of whole
grains if the powder is too small to cut conveniently, is placed in a clean, dried
and weighed 250-m!. beaker, 50 m!. of 95% (by volume) alcohol, and 100 m!.
ethyl ether added and the beaker allowed to stand under a cover':jar with
occasional stirring, until the powder is completely dissolved. This usually
requires from 1 to 2 days. Wh;en all gelatinous particles of the powder have
dissolved, the beaker is heated on the steam bath to evaporate a part of the
ether, before precipitation of the nitrocellulose with water. The amount of
ether. to be evaporated is important, since it largely determines the character
of the nitrocellulose precipitate. The presence of too much ether causes a fine
sandy precipitate; too little causes a gummy, gelatinous precipitate. A fine,
flaky, or fibrous precipitate is desirable. The proper amount of evaporation
can be best determined by practice; usually the solution may be evaporated to
about 2/3 its original volume before precipitating. When the proper volume is
obtained, 50 ml. of water is added from a graduate, with continual stirring, in
5 mi. portions. If a thick gummy precipitate forms, add a little ether until it
becomes flaky; then add the remainder of the 50 m!. of water. The heating is
continued with stirring,'until most of the ether has evaporated, and the beaker
is then left on the bath until the precipitate is just dry. It is then placed in
the 100 C. oven for 1 hour, cooled in a desiccator, and weighed as rapidly as
possible. To facilitate weighing the weights .should be placed on the balance
pan before the beaker is removed from the desiccator, so that the exact weight
can be adjusted quickly. If more than 10 seconds are consumed in this weighing, the error caused by absorption of moisture from the air is an appreciable
one. In any event a check weighing should be made after an additional 30
minutes drying at 100 C.
Absorption of moisture during weighing may be avoided by the use of a
suitable aluminum beaker provided with a tightly fitting cover, instead of an
ordinary glass beaker.

1696

ANALYSIS OF EXPLOSIVES

The final weight of nitrocellulose precipitate subtracted from the weight

of the original sample represents the weight of moisture and volatile solvent,
and is calculated as per cent of the original sample. If the powder contains
diphenylamine, this result is corrected by subtracting f~om it one fourth of the
total diphenylamine content, it having been ascertained by actual trial that
approximately this proportion of the diphenylamine is, volatilized during the
evaporation.
I
If the powder contains other organic constituents which would be more or
less completely volatilized during the evaporation process described above,
resort must be had to an indirect method of determining the content of volatil~
solvent. In such a case, the diphenylamine and other ether-soluble constituents are removed by long continued ether extraction of the finely sliced or
crushed grains of powder, the ether extract is evaporated almost to dryness in a
slow current of dry air and finally brought to constant ,weight in a sulfuric acid
desiccator. The volatile solvent is taken as' the difference between 100%
and the sum of the percentages of ether extract, nitrocellulose and moisture.
Moisture.-An approximation to the actual moisture content of the powder
can be obtained by drying a sample of not less than 2, whole grains and not
less than 20 grams for 6 hours in a vacuum oven having a vacuum of not less
than 25 inches and maintained at 552 C., cooling in a desiccator and weighing, the loss of weight being regarded as equal to the hygroscopic moisture in
the powder.
Diphenylamine.-The content of diphenylamine used as a stabilizer in
smokeless powder is most conveniently and rapidly determined by the" nitration method" as follows:
Five grams of the powder in small grains or slices is treated with 30 ml. of
concentrated HNO a in a 250-ml. beaker, covered with a watch glass and heated
on the steam bath until the powder has 'been completely decomposed. The
solution is then cooled and added to 100 ml. of cold distilled H 2 0 in a second
beaker, stirring vigorously, the first beaker being washed out completely into
the second, using additional water. This mixture is npw heated on the steam
bath until the flocculent precipitate has settled and the'liquid has a clear yellow
color. It is then cooled, filtered through a weighed Gooch crucible, the precipitate dried at 100 C. and weighed. The weighed precipitate is now dissolved by extracting with acetone, the crucible dried and weighed again, the
loss of weight being the nitrodiphenylamine produced by action of the HNO a
on the diphenylamine. This nitro diphenylamine is a mixture of nitro products,
and the empirical factor 0.4345 has been determined for converting it to its
equivalent in diphenylamine.
It is obvious that the diphenylamine found in old powders by the above
method will include any of the original diphenylamine which has been acted on
by nitric acid resulting from slow decomposition of the nitrocellulose and thereby
converted to nitro derivatives of diphenylamine. The nitration method therefore determines "Total diphenylamine." The content of free or "active"
diphenylamine is determined by the bromination method as follows: 17a
The ether- extract, obtained as described in the paragraph on Moisture and
Volatiles, is dissolved in about 100 ml. of ether, transferred tO/a 250-ml. beaker,
and 0.2 ml. bromine added. The solution is hea1j)d.-to about 75 on a steam
bath and the ether evaporated with a current of c[ry air. The residue is treated
17.

S. G. Cook, Ind. Eng. Chem. Anal. Ed., 7, 250 (1935):

ANALY$IS OF EXPLOSIVES

1697

with 40 m!. of 95% (by volume) ethyl alcohol, heated to boiling, 40 m!. of hot
water added and the beaker left on tlie steam bath for exactly 10 minutes.
The precipitate of tetrabromidiphenylamine is collected by filtering the hot
liquid through a Gooch crucible, washed with 80 m!. of boiling 47.5% (by
volume) ethyl alcohol, and finally with hot water, dried 1 hour at 100 C. and
weighed. The weight of precipitate multiplied by 0.3487 equals weight of
diphenylamine.
Ash.-The ash is determined in the manner described for nitrocellulose
(p. 1692), the sample being in the form of slices or small grains, and the digestion
with HNO s continued until decomposition is practically complete, before heating over a flame.
Solubility in Ether-alcohol.-One gram of the sample in slices or small
grains is dissolved in 150 m!. of ether-alcohol (2 : I) in the same manner as for
the determination of moisture and volatiles, and transferred to a standard
solubility tube (p. 1691), washing it in completely with fresh ether-alcohol so as
to bring the total volume to 300 m!. The insoluble material is determined as
in pyrocellulose (p. 1691).
Solubility in Acetone.-This determination is made in the same manner as
the solubility in ether-alcohol, described above.
Stability Test at 134.5 C.-This test is made on duplicate samples in the
same manner as described for pyrocellulose (p. 1693). The samples weigh 2.5
grams and are in as nearly whole grains as is consistent with this weight of
sample, large grains being split longitudinally so as to fit the ,standard tubes.
The samples are required to stand heating at 134-135 C. for 5 hours without
explosion and must not turn th.e normal methyl violet paper to salmon pink
color in less than one hour.
" Surveillance Test" at 65.5 C.-Three samples of approximately 45 grams
of powder in whole grains, or, in the case of very large grains, 5 whole grains,
are placed in 8-ounce wide-mouth glass stoppered bottles, the stoppers having
been previously ground so as to fit tightly. These bottles are then heated in
a constant temperature magazine at 65.52 C. They are observed several
times daily and the time noted when visible fumes of oxides of nitrogen appear
in any bottle. The number of days ,\rhich powder will stand this test depends
on the web thickness of the grain, as well as on actual stability and may amount
to several hundred days. The test is therefore not a laboratory test, but one
which more nearly approaches service conditions. It is of great value as an
indication of the possible" stability life" of the powder in service.
Nitrogen.-The determination of nitrogen in smokeless powder is not
usually necessary, in as 'much as the powder is usually made from nitrocellulose
of known nitrogen content, but when desired the determination is made as
follows, in the case of powders containing only colloided nitrocellulose and
diphenylamine.
An average sample of about 5 grams of the powder in slices or small grains
is dissolved in acetone (100 m!. to each 1 g. of sample). When the sample is
dissolved, the solution is added drop by drop, preferably from a burette, to
200 m!. of hot water in a beaker, the beaker being immersed in boiling water
so as to maintain its contents at about 90 C. During this addition the hot
water is continually stirred with a glass rod, so that the precipitated nitrocellulose forms stringy masses which wrap about the rod. Small accumulations

1698

ANALYSIS OF EXPLOSIVES

of the precipitate are transferred frequently from the rod to another beaker of
hot water to prevent the formation of a colloided,mass.1 When 2 g. or more of
the precipitate has been collected and the acetone has peen volatilized by the
hot water, it is removed from the beaker and dried at 35-40 C. About 1 g.
of this dry precipitate is placed in a tared weighing 'bottle, dried 1 hour at
100 C., weighed, and transferred to the cup of the n~trometer with sulfuric
acid. Part of the acid should be added to the precipitate in the weighing bottle
before transferring to the nitrometer in order to avoid loss of the dry precipitate
in handling. The determination of N is then completed as in the case of
nitrocellulose (page 1690). If the powder contains diphenylamine, a correction
is necessary for the amount of diphenylamine retained by the precipitated
nitrocellulose. This has been found to be an added correction of 0.15% N In
the case of powders containing the usual amount of 0.4% diphenylamine. This
correction compensates for. the nItrogen which becomes combined with t4e
diphenylamine, converting it to nitrodiphenylamines.
..
Instead of correcting for the effect of the diphenylamine, the latter may be
removed from the precipitated nitrocellulose, after air-drying and before final
drying at 100 C., by extraction with pure anhydrous ether. Results are quite
accurate if the determination is conducted with proper precaution.
Muraour 18 describes the following method for direct detel,"mination of
nitrate nitrogen in smokeless powders without preliminary treatment: The
powder sample is dissolyed in acetone and the solution treated with hydrogen
peroxide and sodium hydroxide with gentle warming. Sodium perborate is
then added and the mixture reduced with Devarda alloy in the presence of
additional sodium hydroxide, and finally distilled into a known volume of
standard sulfuric acid. Results are obtained which agree closely with the
theory. The method is applicable in the presence of easily nitrated materials'
such as diphenylamine, Centralite, etc., which interfere with the determination
of nitrate nitrogen by means of the nitrometer.
Nitroglycerin Smokeless Powders
I

Powders of this type are composed mainly of nitrocellulose and nitroglycerin and may contain other organic, or inorganic substances, such as vaseline, nitro-substitution compounds, substituted 'ureas or other flame-reducing
or surface-coating agents, diphenylamine, metallic nitrates, carbonates, etc.
The nitrocellvlose may bEl either high-nitration guncotton ,insoluble in etheralcohol, as in British cordite, or a low-nitration product soluble in nitroglycerin,
as in ballistite, or may be a mixture of the two varieties.
'
The method of analysis usually employed consists of (1) an extraction of
the nitroglycerin, nitrosubstitution compounds, vaseline, and other ethersoluble materials by means of anhydrous ether; (2) an extraction of the watersoluble materials; (3) determination of soluble and insoluble nitrocelluloses by
separation with ether-alcohol (2 : 1).
'
The extraction with ether is usually made in a Soxhlet apparatus, using
about 20 grams of the powder in slices or small grains, ~n a paper extraction
thimble. About 4 hours is usually required for complete/extraction. The
/~

Direct determination of nitrate nitrogen in colloidlid ,smokeless powders, Henri

Muraour. Bull. Soc. Chim. de France (4), 45, 1189 (1929):18

ANALYSIS OF EXPLOSIVES

1699

ether extract is evaporated to dryness in a tared glass dish under a bell-jar

evaporator (page 1664); and the' ether-soluble-residue weighed. To determine
whether it contains other substances than nitroglycerin, it may be poured in
small portions at a time into about 20 ml. of concentrated nitric acid (40 Be)
heated on a steam bath. The oxidizing action of the nitric acid destroys the
nitroglycerin, and the mixture is then poured into 50-100 ml. of water. Any
vaseline or similar substances separate, together with any nitro substitution
compounds in their original condition or more completely nitrated, diphenylamine in the form of a nitroderivative, etc.
These materials may be separated with more or less completeness by fractional crystallization from ether or other solvent. The exact method to be
followed depends on the nature of the materials present.
The residue insoluble in ether is dried and weighed, and then transferred
to an Erlenmeyer flask and digested in warm water until any water-soluble
materials present have been dissolved. The mixture is filtered, the residue
washed with hot water, dried and weighed. The filtrate containing the watersol.uble ingredients is examined by the usual analytical methods for inorganic
ingredients.
The nitrocellulose insoluble in water is tested for nitrogen content, solubility
in ether-alcohol and solubility in acetone, by the methods already described.
The dialkyldiphenyl ureas, known as Centralites, are used as surface coating
agents and stabilizers in some powders containing nitroglycerin and nitrocellulose. In the analysis of such powders both the nitroglycerin and Centralite
are dissolved by extraction with absolute ether, and a mixtur~ of these ingredients obtained by evaporation of the ether. Determination of the nitroglycerin in this mixture by means of the nitrometer gives low results because when
concentrated sulfuric acid is added the Centralite becomes nitrated to a tetranitrocompound. Levenson 19 has found that the Centralite in such mixtures
can be determined directly by a bromination method, as follows:
Extract 5 grams of the finely ground powder with absolute ether for about
24 hours, and evaporate the ether extract to constant weight. Transfer the
liquid resid,ue to a 250-m!. glass-stoppered Erlenmeyer flask with 50 m}. of
ethyl alcohol, add 25 m!. of standard bromate-bromide solution, cool to 20
and add 5 m!. of concentrated HCI. Stopper the flask, shake gently, and
exactly 30 seconds from the time of addition of the acid add 10 mI. of a 15%
solution of KI. Immediately titrate the liberated iodine with 0.1 N Na 2S 2 0a
solution using starch indicator. 1 ml. 0.1 N Na 2S 2 0a equals 0.0067044 gram
diethyldiphenylurea. The nitroglycerin is calculated by difference. To
prepare the bromate-bromide solution dissolve 5.568 g. KBrOa and 30 g. KBr
,in water and dilute to 1 liter. Standardize by placing 25 m!. of the solution
and 50 mI. of ethyl alcohol in a 250-ml. glass-stoppered Erlenmeyer flask, cooling
to 20, adding 5. ml. of concentrated HCI, and after 30 seconds titrating as'
above with 0.1 N thiosulfate solution. The volume of thiosulfate solution
required in the standardization minus that required for the sample represents
the Br combined with the Centralite.
19 Determination of Centralite in double-base smokeless powders, H. Levenson,
Ind. Eng. Chern., Anal. Ed., 2, 246-247 (1930).

1700

ANALYSIS OF EXPLOSIVES
TYPICAL COMPOSITIONS OF COMMONLY USED EXPLOSIVES

Black Blasting Powder

Sodium nitrate. . . . . . . . . . . . . . . . . . . . . . . . . . . .. 73
Charcoal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 16
\
Sulfur ..................................... 11 (Bur. of Mines, Bull. No. 80, p. W)
Black Military Powder
Potassium nitrate .... ,. . .. . . . . . . . . . . . . . . . . .. 75
Charcoal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 15
Sulfur ...................................... 10
Typical Dynamite formulas-4D% grades (Bur. Mines, Bull. No. 80, p. 21)
Nitro- AmmoNitro- Substi- nium Sodium
glycerin tution Nitrate Nitrate
Com.

40% straight Nitroglycerin Dynamite.

"40%-strength" Ammonia Dynamite
"40%-strength" Gelatin Dynamite ...
"40%-strength" Low-freezing Dynamite ...........................
"40 %-strength " Low-freezing Ammonia. Dynamite ................

40
22
33

NitroceUulose

Wood
Pulp

Cal-'
cium

C"tbonate

--- ------ --- ---- 44

1
15
20
42
15*
1
52
13*
1
1

30

10

17

44

20

45

15

I 13*

1
.I

* Sometimes contains also flour, cornmeal/sulfur, etc.

Granulated Nitroglycerin Powder ("Judson Powder")

Nitroglycerin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5
Combustible material t. . . . . . . . . . . . . . . . . . . . . .. 35
Sodium nitrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 60
Composed of sulfur, coal, and rosin.
COAL MINING POWDERS.

10
26
64

(PERMISSIBLE):' TYPE)

II
III
VI
IV
V
---------------1--- --------------15
10
Nitroglycerin.
.. . ...... . ... . . . 25
10
5
5
TNT ............. .. . . . . .. . .. . . ....
Ammonium nitrate.
... .. . . ., .
79
94
90
70
Sodium nitrate.
34
35
... . . . . \......... .
.... .. . .. . .. . . . ..
Sodium chloride.
9
12
10
10
........ .. . ........ .
15
Wood pulp.
17
Flour. ................................ . 25
5
Aluminum powder .......................
3
Charcoal ...............................
3
1
1
1
..
Calcium carbonate. ....... .. . . . .. .
Zinc oxide. ....... ._ ..... . .. .. . .. . . .
J
Magnesium sulfate, cryst ..
15
..
.

FIXED OILS, FATS AND WAXES

PETROLEUM

It will be remembered that the fixed oils are those which leave a permanen~
stain on paper, whereas the essential or volatile oils evaporate. The fixed oils,
if saponifiable, are glycerides of the higher fatty acids, oleic, C 17 H aa COOH,
stearic, C 17H 3S COOH, palmitic, C1sHa1COOH; and if unsaponifiable, hydrocarbons usually of the paraffin (C n H zn+2) and olefin (CnHzn) series. The fats
differ from ~he oils in having a higher melting-point, caused by a larger percentage of stearic and palmitic acid. The waxes differ from the oils or fats in that
the former are esters of monatomic alcohols. The oils are further divided
accotding to their drying power on exposure to the air, into drying, semi-drying
and non-drying oils.
The drying oils contain a large proportion of glycerides. of the unsaturated
acids, particularly linoleic and linolenic, whereas the semi-drying contain a
smaller percentage, and the non-drying little or none qf these esters.
Examination of an Unknown Oil

There being no specific tests for the various oils, as for their identification,
the analyst should, in attacking an unknown oil, ascertain all possible facts
about it, as the source, the use to which it is put, and the cost.
Certain physical properties too, may aid in the examination. The color
is of little assistance, as oils may be colored by the use of oleates or butyrates of
iron or copper. Fluorescence or " bloom" is valuable as indicating the presence
of mineral oil; this can be shown 1 by placing a few drops of the oil on a sheet
of ebonite and observing the bluish color. The odor and taste are 'particularly
valuable. Marine animal oils are detected, especially when warm, by their
strong" fishy" odor, while neatsfoot, tallow, lard, rosin and linseed oils each
have a well-marked and easily distinguishable smell. Whale oil is said to have
a " nutty," and rape oil has a harsh,. unpleasant" turnipy "taste. The turbidity, showing the presence of water, or of oils which imperfectly mix-as castor
and mineral oils-and the sediment, either stearin or dirt, are also to be noted;
these should be filtered out through paper before the oil is analyzed.
The elaidin test (page 1765) may be applied next, to allow time for the cake
to form; it will be followed by the Maumene test (page 1766) both being done in
duplicate. In making the elaidin test it is advisable to carryon an experiment
under the same conditions with a known sample of lard oil. -These two tests
will show whether the sample under examination is a drying or non-drying oil
and when the ingredients of the mixture are determined, the results of the
Maumene test can be used for calculating their relative amounts. The io'dine
test can be employed to che.ck this result.
lOuterbridge, Chern. Abstr. 5, 3350, uses an enclosed arc light giving ultraviolet
light and makes it quantitative.
Chapter contributed by the late Augustus H. Gill.

1701

1702

FIXED OILS, FATS AND WAXES

The saponification test, unless mineral or rosin oil be suspected, need rarely
be resorted to ; the reason being that it would show praptically nothing regarding
the nature of the oil. Except in the case of castor (Sapn. No. 181), rape (174),
sperm (135) and cocoanut (260), this characteristic iSj about 193.
Finally, where the importance of the case will warrant, the analyst is advised
to prepare a mixture of oils using the proper proportions indicated by the various
tests, and subject it to the more rapid tests as the specific gravity, viscosity,
Maumene and iodine number. In making out the report of analysis it shouldbe borne in mind that, excepting in the case of the special test, the results of
one test cannot be relied upon to determine the nature of an oil, but the evidence
of all the tests here given should be carefully compared and weighed before
rendering a final verdict: in consideration of the fact of the wide variation of the
characteristics of the oils, it is futile to report the quantities of oil found in a
mixture more closely than 1%.

PETROLEUM PRODUCTS
(a) Burning Oils
The tests or determinations to be made are, in the order of their importance,
flash, fire, specific gravity, distillation, sulfur, free acid, sulfuric acid, cioud
test, mineral salts and water. In some cases the color is determined as it is of
commercial importance.
.
Flash Test or Point.-By flash point we understantl the lowest temperature
to which an oil must be heated, to give off vapors which when mixed with air
produce an explosive mixture. The results of th~s test will vary according to the
quantity of air over the surface of the oil, and whether this be moving or still;
also according to the distance of the testing flame from the surface of the oil.
Furthermore; the size of this testing flame, the'length of its time of action, its
form and dimensions, and'lastly, the manner of heating the oil, will all influence
the result. 2

.
Any cause producing the rapid evolution of a large amount of petroleum
vapor tends to lower the flash point. Barometric changes are, for practical
work, negligible, each 5 mm. causing a variation of but 0.10 C.
Determination by the" New York State Board of Health Tester."-The
apparatus, Fig. 238, consists of a copper oil cup, Ii, holding about 10 oz., the
quantity usually contained in lamps, heated in a water bath by a small Bunsen
flame. The. cup is provided with a glass cover, C, carrying a thermometer, B,
and a hole for the insertion of the testing flame-a small gas flame one-quarter
..
.I
of an inch in length.
2

Engler and Haase, Z. Anal. Chern., 20, 3, 1881.

FIXED OILS, FATS AND WAXES

1703

Manipulation.-Mter describing the apparatus minutely, the regulations of

the New York State Board of Health say,3 " (2) The test shall be applied according to the following directions:
" Remove the oil cup and fill the water bath with cold water up to the mark
on the inside. Replace the oil cup and pour
in enough oil to fill it to within one-eighth of
an inch of the flange joining the cup and
the vapor-chamber above. Care must be
taken that the oil does not flow over the
flange. Remove all air-bubbles with a piece
of dry paper. Place the glass cover on the
oil cup, and so adjust the thermometer that
its bulb shall be just covered by the oil.
"If an alcohol lamp, 'be employed for
heating the water bath, the wick should pe
carefully trimmed and adjusted to a small
flaJ;lle. A small Bunsen burner may be used
in place of the lamp. The rate of heating
should be about two degrees per minute,
and in no case exceed three degrees.
" As a flash torch, a small gas jet onequarter of an inch in length should be emFIG. 238.-N. Y. Tester.
ployed. When gas is not at hand employ
a piece of waxed linen twine. The-flame in this case, however, should be small.
" When the temperature of the oil has reached 85 F. the testings should
commence. To this end insert the torch into the opening in the cover, passing
it in at such an angle as to well clear the cover, and to a distance about half-way
between the oil and the cover. The motion should be steady and uniform, rapid
and without any pause. This should be repeated at every two degrees' rise
of the thermometer until the temperature has reached 95, when the lamp
should be removed and the testings should be made for each degree of temperature until 100 is reached. After this the lamp may be replaced if necessary
and the testings continued for each two degrees.
" The appearance of a slight bluish flame which passes over the entire surface shows that the flashing-point has been reached.
" In every case note the temperature of thQ oil,before introducing the torch.
The flame of the torch must not come in contact .with the oil.
" The water bath should be filled with cold water for each separate test, and
the oil from a previous test carefully wiped from the oil cup."
For the determination with the open tester (Tagliabue''I small) reference
may be had to the author's" Short Handbook of Oil Analysis ); for the test with
the closed tester, Abel's.or Pensky-Martens, Holde-Mueller,'" Examination
of Hydrocarbon Oils"; or for the" Tag Tester."
Fire Test.-The fire test of an oil-is the lowest temperature at which it will
give off vapors which when ignited will burn continuously. It is made by continuing to heat the oil (the cover being removed in the case of a closed tester
without slipping out the thermometer) at the same rate after the flash test
a Report

of the New York State Board of Health, 1882, p. 495.

FIXED OILS, FATS AND WAXES

1704

is made and noting the point as indicated above. ;The flame is extinguished
by a piece of asbestos board and the heating discon;tinul;ld. In the case of
many illuminating oils this point is from 10 to 20 F. higher than the flashpoint.
In the case of " Mineral Sperm" (300 F. fire test oil) these tests should
be made with the Cleveland open cup. The heating should be oat the rate of
10 F. per minute, and the testing flame first applied at 230 F. and then every
seven degrees until the flashing-point is reached.
I
The most satisfactory way of making these tests is to place the watch upon
the desk and read the thermometer at the expiration of every minute, noting
down each reading in the proper column in the laboratory note-book.
Significance of the Tests.-The flash test rather than the fire test measures
the safety of an oil as the vapors producing the flash are evaporated in the fire
test. The writer has had samples of oil of 65 to 70 F. flash which gave a fire
test of 135-140 F. These, judged by the flash test are unsafe, whereas they
would be passed by the fire test.
. .
Specific Gravity: (a) By the Hydrometer.-A hydrometer jar is four-fifths
filled with the oil, a verified Baume hydrometer introduced into it, and the
depth read off to which the instrument sinks into the oil. This may be effected
by placing a strip of white paper back of the jar and noting the point at whiCh
the lower meniscus of the oil touches the scale. The temperature of the oil
is taken at the same time, and in case it be not 60 F. (15.5 C.), subtract 1
Baume from the hydrometer reading, for every 10 F. it is higher than 60, and
add 1 Baume for every 10 F. it is lower than 60 F. In practice this can be
done by Tagliabue's " Manual for Inspectors of Petroleum," which gives. the
readings at 60 F. for any gravity
from 10 to 100 Baume, between 20
and 109 F. The specific gravity may
141.5
be found by the formula 131.5+Be'
Be representing the reading Baume at
15.5 C.
/
(b) By tTie Westphal Balance.-

This is a specially constructed instrument, Fig. 239, with a glass plummet

carrying a thermometer counterbalanced by a weight. Upon immersing
the plummet in a~liquiQ. the positions
of the weights, which must be added
to restore the equilibrium, represent
the specific gravity directly. The
largest' weight represents the first
":i;;;?i~ill!!Iiiiiiii~~~.~ decimal place, the next the second,
and so on.; The instrument is placed
FIG. 239.-Wesphal Balance.
upon a level table, and by means of
the leveling screw is brought into adjustment-i.e., so that the point upon the beam is exactly opposite the point
upon the fixed part.
I
The plummet is now placed.in the vial or tJalance jar containing the oil,
cooled to 15.5 C., hung upon the balance, being careful completely to immerse

FIXED OILS, FATS AND WAXES.

1705

it in the oil, weights added to restore the equilibrium, and the specific gravity
read off as above described.
Care should be taken that the 'plummet does not touch the sides of the jar
or vial. For solid fats and some oils the specific gravity is taken at 100 C.,
using a special plummet.
Significance of the Test.-The Specific Gravity or " Gravity" as the oil
man says shows whether the sample is of paraffine or asphalt base, the latter
oils being the heavier.
Distillation Test:
A. S. T. M. METHOD D 86-35.L-KEROSENE
STANDARD METHOD OF TEST FOR DISTILLATION OF GASOLINE,
NAPHTHA, KEROSENE, AND SIMILAR PETROLEUM
PRODUCTS

A. S. T. M. DESIGNATiON: D 86-35

Apparatus: Flask.-The standard 100-ml. Engler flask is shown in Fig. 240,

the dimensions and allowable tolerances being as follows:
Tolerances

Centimeters Inches
Centimeters

Diameter of bulb, outside ......... ,..........

Diameter of neck, inside. . . . . . . . . ...........
Length of neck .............................
Length of vapor tube .......................
Diameter of vapor tube, outside ..............
Diameter of vapor tube, inside ...............
Thickness of vapor tube wall ................

6.5
1.6
15.0
10.0
0.6
0.4
0.1

2.56
0.63
5.91
3.94
0.24
0.16
0.04

0.2
0.1
0.4
0.3
0.05
0.05
0.05

Inches

0.08
0.04
0.16
0.12
0.02
0.02
0.02

The position of the vapor tube shall be 9 cm.0.3 cm. (3.55 in.0.12 in.)
above' the surface of the liquid when the flask contains its charge of 100 m1.
The tube is approximate~y in the middle of the neck and set at an angle of 75
(tolerance3) with the vertical.
.
Condenser.-The condenser (Fig. 241) consists of a 196-in. (14.29 mm.)
OD No. 20 Stubbs Gage seamless brass tube, 22 in. (55.88 em.) long. It is set
at an angle of 75 from the perpendicular and is surrounded with a cooling bath
15 in. long (38.1 em.), approximately 4 in. (10.16 em.) wide by 6 in. (15.24 cm.)
high. The lower end of the condenser tube is cut off at an acute angle, and
Bureau of Minp.s Technical Paper 323B is an earlier description.
Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S.T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
G This method is issued under the fixed designation D 86; the final number indicates
the year of original adoption as standard or, in the case of revision, the yea: of last
revision. Issued as Tentative, 1921; Adopted in Amended Form, 1927; Revised, 1930,
1935. This method has been approved as AMERICAN STANDARD by the American
Standards Association.
4

1706

. FIXED OILS, FATS AND

~AXES

curved downward for a length of 3 in. (7.62 Qm.) and slightly backward so as to
insure contact with the wall of the graduate at a pointl1 to 1~ in. (2.54 to 3.18
cm.) below the top of the graduate when it is in position to receive the distillate.
Shield.-The shield (Fig. 241) is made of approxi,$ately No. 22 gage sheet
metal and is 19 in. (48.26 cm.) high, 11 in. (27.94 cm.) lbng and 8 in. (20.32 cm.)
wide, with a door on one narrow side, with two open,ings, 1 in. (2.54 cm.). in
diameter, equally spaced, in each of the
-,.two narrow sides, and with a slot cut
!
in one side for the vapor tube. The
centers of these four openings are 8%
1,,"",
in. (21.59 cm.) below the top of the
;_l;
shield. There are also three Y2-in.
I ~~
(1.27 cm.) hol~s in each of the four
II<;.~. ...........
sides with their centers 1 in. (2.54 cm.)
~J. 9~
1QOCI11..) _
.........."'1 above the base of the shield .
Ring Support and Hard A~bestos
I
I
Boards.-The ring support is of the
ordinary laboratory type, 4 in. (10.16
em.) or larger in diameter, and is supported on a stand inside the shield.
There are two hard asbestos boards:
One 6 by 6 by ~ .in. (15.24 by 15.24
cfu. by 6.35 mm.) with a hole
in.
(3.18 cm.) 7 in diameter in its center,
the sides of which shall be perpendic.ular to the surface; the other, an asI
I
bestos board to fit tightly inside the
I
shield, with an opening 4 in. (10.16
I
cm.) in diameter concentric with the
~ . 2:56-.. -.. --.~
(6.5cm.
ring support. These are arranged as
Ouf.fldeOi4tneler)
follows: The !iecond asbestos board is
FIG. 240.-Dimensions of 100-mI.
Engler Flask.
placed on tIle ring and the first or
smaller asbestos board on top so that it
may be moved in accordance with the directions for placing the distilling flask.
Direct heat is applied to the flask only through the l~-in. (3.18 cm.) opening
'
. in the first asbestos board.
Gas Burner or Electric Heater.-(a) Gas Burner.~The, burner is so
constructed that sufficient heat can be obtained to distill the product at the
uniform rate specified below. The flame should never be so-large that it
spreads over a circle of diameter greater than 331 in. (8.89 cm.) on the under
surface of the asbestos board. A sensitive regulating valve is a necessary
adjunct, as it gives complete control of heating.
I
(b) Electric Heater.-The electric heater, which may be used in place of the
gas flame, shall be capable of bringing over the first drop within .the time specified below when started cold, and of continuing the distillation at the uniform
rate. The electric heater shall be fitted with an asbestos 'board
top VB to 74: in.
T

,"v

o.i

0
0
7 When distilling petroleum products having an en~..point above 470 F. (243.34 C.),
the hole in the asbestos board shall be 1}1 in. (3.81 em.) in diameter.

FIXED OILS, FATS AND WAXES

1707

(3.18 to 6.35 mm.) thick, having a hole lU in. (3.18 em.) in diameter in the
center. When an electric heater is employed, the portion of the shield above
the asbestos boald shall be the same as with the gas burner but the part below
the board lpay be omitted ..

"BlofH,,'!

Popu

I
I
I

~~----~------------------~

.~.

FIG. 241.-Apparatus for Distillation Test.

l'hermometers.-{n) The A. S. T. M. Low-Djstj]]atjoD Thel'mometer snall

conform to the following requirements. These specifications cover a totalimmersion thermometer graduated in either Centigrade or Fahrenheit degrees,
as specified, the ranges being 0 to 300 0 C. or 30 to 580.F. respectively.
TYPE: Etched stem, glas~..
LIQUID: Mercury.
,
RANGE AND SUBDIVISION: 0 to 300 C. in 1 C. or 30 to 580 F. in 2 F.
TOTAL LENGTH: 378 to 3M nun. (14.88 to 15.12 in.).
STEM: Plain front, enamElI back, suitable thermometer tubing.
Diameter, 6.0 to 7.0 mm. (0.24 to 0.28 in.).
BULB: Corning normal 01' equally suitable thermometric glass.
.
Length, 10 to 15 mm. (0.39 to 0.59 in.).
Diameter, 5.0 to 6.0 mm. (0.20 to 0.24 in.).
DISTANCE TO 0 C. OR 32 F. LINE FROM BOTTOM OF BULB: 100 to 110 mm. (3.94 to
4.33 in.).
DISTANCE TO 3000 C. OR 572 F. LINE FROM Top OF THERMOMETER: 30 to,45 nun.
(1.18 to 1.77 in.).
FILLING ABOVE MERC~E.Y: Nitrogen gaB.
Top FINISH: Glass ring. .
0

1708

FIXED OILS, FATS AND WAXES

GR.AD~ATION: AIl lines, figures, and letters clear cut and distinct. The first and each

succeeding 5 C. or 100 F. line to be longer than-the remaining lines. Graduations

to be numbered at each multiple of 10 C. or 20F.
~
.
IMMERSION: Total.
SPECIAL MARKING: "A. S. T. M. Low Distillation," a: serial number and the manu.
facturer's name or trade mark shall be etched on the stbm.
SCALE ERROR: The error at any point of the scale when the thermometer is standardized as provided below shall not exceed 0.5 C. or 1 F.
STANDARDIZATION: The thermometer shall be standardized immersed in the testing
bath to the top of the meroury column, at the ice point and at temperature inter~
vals of approximately 50 C. or 100 F. up to 300 0 C. or 572 0 F.
TEST FOR PERMANENCY OF RANGE: After being subjected to a temperature of 280
to 290 C. or 540 to 560 F. for 24 hours, the accuracy shall, be within the limit
specified. The test shall be made under the immersion conditions specified for
this thermometer.
CASE: The thermometer shall be supplied in a suitable' case on which shall appear
the marking: "A. S. T. M. Low Distillation, 0 to 300 C." or "A. S. T. M. Low
Distillation, 30 to 580 F." according to the type of thermometer.
NOTE I.-For the purpose of interpreting these specifications the following definitions apply:
The total length is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning I
of the enamel backing.
J
The top of the thermometer is the top of the finished instrument.
0

(b) The A. S. T. M. High-Distillation Thermometer shall conform to the'

following requirements. These specifications cover a total-immersion thermometer graduated either in Centigrade or Fahrenheit degrees, as specified,
the ranges being 0 to 400 0 C. or 30 to 760 0 F., respectively.
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGE AND SUBDIVISION: 0 to 400 0 C. in 10 C. or 30 to 760 0 F. in 2 F.
TOTAL ,LENGTH: 378 to 384 mm. (14.88 to 15.12 in.).
STEll!:: Plain front, enamel back, suitable thermometer tubing. Diameter, 6.0 to 7.0
mm. (0.24 to 0.28 in.).
"
BULB: Corning normal or equally suitable thermometric glass.
Length, 10 to 15 mm. (0.39 to 0.59 in.).
Diameter, 5.0 to 6.0 mm. (0.20 to 0.24 in.).
DISTANCE TO 0 C. OR 32 0 F. LINE FROM BOTTOM OF BULB: 25 to 35 mm. (0.98 to 1.38
in.).
DISTANCE TO 400 0 C. OR 752 F. LINE FROM Top OF THERMOMETER: 30 to 45 mm.
(1.18 to 1.77 i n . ) . '
'
FIL'I'ING ABOVE MERCURY: Nitrogen gas.
Top FINISH: Glass ring.
GRADUATION: AIl lines, figures and letters clear cut and distinct. The first and each
succeeding 5 C. or 10 F. line to be longer than the remaining lines. Graduations
to be numbered at each multiple of 10 C. or 20 F.
IMMERSION: Total.
SPECIAL MARKING: "A. S. T. 1\1:.' High Distillation" a serial number and the manufacturer's name or trade mark shall be etched on the stem.
SCALE ERROR: The error at any point of the scale up to 3700 C. or 700 0 F. when the
thermometer is standardized as provided below shall not exceed 1 0 C. or 2 0 F.
STANDARDIZATION: The thermometer shall be standardized immersed in the testing
bath to the top of the mercury column, at the ice point-and at temperature intervals of approximately 50 C. or 1000 F. up to 370." e~r 700 0 F.

FIXED OILS, FATS AND WAXES

1709

TEST FOR PERMANENCY OF RANGE: After being subjected to a temperature between

360 and 370 C. or 680 and 700 F. for 24 hours, the accuracy shall be within
the limit specified. The test shall be made under the immersion conditions specified for this thermometer.
CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T. M. High Distillation, 0 to 400 C." or "A. S. T. M. High
Distillation, 30 to 760F." according to the type of thermometer.
NOTE I.-For the purpose of interpreting these specifiGations the following defi.
nitions apply:
The total length is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning
of the enamel backing.
The top of the thermometer is the top of the finished instrument.
NOTE 2.-Under certain test conditions, the bulb of the thermometer may be
500 F. (28 C.) above the temperature indicated by the thermometer, and at an indicated temperature of 700 F. (371C.) the temperature of the bulb is approaching a
critical range in the glass. It is therefore not desirable to use this thermometer under
such conditions at indicated temperatures above 700 F. (371 0 C:) without checking the
ice point.

Graduale.-The graduate shall be of the cylindrical type, of uniform

diameter, with a pressed or molded base and a lipped top. The cylinder shall
be graduated to contain 100 mI., and the graduated portion shall be not less
than 7 in. (17.78 cm.) nor more than 8 in. (20.32 em.) long. It sh.all be gradu-.
ated in single milliliters and each fifth mark shall be distinguished by a longer
line. It shall be numbered from the bottom up at intervals of 10 m!. The
overall height of the graduate shall be not less than 9%; in. (24.8 em.), nor more
than 10% in. (26.0 em.). The graduations shall not be in error by more than
1 ml. at any point on the scale.
Procedure.-(a) The condenser bath shall be filled with cracked ice,S and
enough water added to cover the condenser tube. The temperature shall be
maintained between 32 and 40 F. (0 and 4.45 C.).
(b) The condens~r tube shall be swabbed to remove any liquid remaining
from the previous test .. A piece of soft cloth attached to a cord or copper wire
may be used for this purpose.
,
(c) One hundred milliliters of the product shall be measured in the 100-ml.
graduated cylinder at 55 to 65 F. (12.8 to 18.3 C.) and transferred directly
to the Engler flask. N one of the liquid shall be permitted to flow into the
vapor tube.
(d) The thermometer (Note) provided with a cork shall be fitted tightly
into the flask so that it will be in the middle of the neck and so that the lower
end of the capillary tube is on a level with the inside of the bottom of the vapor
outlet tube at its junction with the neck of the flask. The thermometer shall
be approximately at room temperature when placed in the flask.
NOTE.-For products having end points not higher than 482 0 F. (250 C.) or initial
boiling points not higher than 212 0 F. (100 0 C.), the Low-Distillation Thermometer
shall be used; for products having end points higher than 4820 F. (250 C.) and initial
boiling points higher than 212 F. (100 C.) the High-Distillation Thermometer shall
be used.
8

Any other convenient cooling medium may be used,

1710

FIXED OILS, FATS AND WAXES

I

(e) The charged flask shall be placed in the l}:4'-in. (3.18-cm.) opening in the
6 by 6-in. (15.24 by 15.24-cm.) asbestos board with th~ vapor outlet tube inserted into the condenser tube. A tight connection may be made by means of
a cork through which the vapor tube passes. The positibn of the flask shall be
so adjusted that the vapor tube extends into the condenser1tube not less than 1 in.
I
(2.54 cm.) nor more than 2 in. (5.08 cm.).
(f) The graduated cylinder used in measuring the charge shall be placed,
without drying, at the outlet of the condenser tube in such a position that the
condenser tube shall extend into the graduate at least 1 in. (2.54 em.) but not
below the 100-mI. mark. Unless the temperature is between 55 and 65 F.
(12.8 and 18.3 C.) the receiving graduate shall be immersed up to the 100-ml.
mark in a transparent bath maintained between these temperatures. The top
of the graduate shall be covered closely during the distillation with a piece of
blotting paper or its equivalent, cut so as to fit the condenser tube tightly.
When everything is in readiness, heat shall be applied at a uniform rate,
so regulated that the first drop of condensate falls from the condenser in not less
than 5 nor more than 10 minutes. The distillation thermometer shall be read
2 minutes after heat is applied and the indication recorded as the" correction
temperature." This figure is of significance only in cases when there is a question as to the accuracy 9 of the initial boiling point, as subsequently determined.
When the first drop falls from the end of the condenser the reading of the distillation thermometer shall be recorded as tl].e initial boiling point. The receiving cylinder shall then be moved so that the end of the condenser tube shall
touch the side of the cylinder. The heat shall then be so regulatE!d that the
distillation will proceed at a uniform rate of not less than 4 nor more than 5 mI.
per minute', The volume of distillate collected in the cylinder shall be observed
and recorded, to the nearest 0.5 mI., wh'en the mercury of the thermometer
reaches each point that is a multiple of 10 C. or the Fahrenheit equivalent of
this point (30 C., 40 C., 50 C., 60 C., etc., or 86 F., 104 F., 122 F.,
9 The initial boiling point is one of the observations which it is customary to report
as indicative of the qu~lity of gasoline. It is a point Which' js very difficult to check,
especially when duplicate tests are made at different room temperatures, and it has not
been found practical to devise an accurate method of correcting for this unavoidable
variation in operating conditions. The practical significance of initial boiling point is
not regarded as sufficient to warrant requiring that tests be conducted under regulated
conditions of room temperature. The following directions are included as an expedient
intended to obviate some of the difficulties that develop on accoul)..t of the practice of
including initial boiling point limits in specifications for the purchase and'sale of gasoline:
If the "correction temperature" is below 70 F. (21 Co) or above 80 F. (27 C.),
the observed initial boiling point, if it is below 150 F. (66 C.), may be considered
incorrect by an amount
not less than one third the difference between the "correction
temperature" and 75 0 F. (24 0 C.). In case there is a dispute between buyer and seller,
the observed initial boiling point figures obtained by all parties shall be revised according to the following formula:
Revised initial boiling point
= {Observed initial} _ Correction Temperature (deg. Fahr.) "':'75
boiling point
3
If all of the revised figures fall. within the specification limit the gasoline in question
shall be considered as passing; otherwise arrangements shall; -if possible, be made to
conduct a 0 distillation test0 ~ith the room temperature maintaIned between the limits of
70 and 80 F. (21 and 27 C.), inclusive.
0

/'

FIXED OILS, FATS AND WAXES

1711

140 F., etc.). If preferred, the reading of the distillation thermometer mlp'
be observed and recorded when the'level of the distillate reaches each lO-ml.
mark on the graduate. In case a product is being tested to ascertain whether
or not it conforms with a given specification, all necessary observations shall
be made and recorded, whether or not they are included in the series ordinarily
employed by the laboratory making the test.
No adjustment of the heat shall be made after the liquid residue in the flask
is approximately 5 ml. unless the time required to bring over the last 5 ml. of
distillate and reach the end point exceeds 5 minutes. The end point is the
maximum temperature observed on the distillation thermometer and is usually
reached after the bottom of the flask has become dry. If the bottom of the
flask is not dry the operator shall record this fact.
In case the time required to bring over the last 5 ml. of distillate and reach
the end point exceeds 5 minutes the test shall be repeated and the heat shall
be adjusted when the liquid residue reaches 5 ml. This adjustment may be
either an increase or a flecrease but must accomplish the purpose of bringing
the'period required to vaporize the last 5 ml. of distillate and reach the end point
within the limits of 3 and 5 minutes.
The total volume of the distillate colle.cted in the receiving graduate shall
be recorded as the recovery.
The cooled residue shall be poured from the flask into a small cylinder
graduated in 0.1 mi., measured whea cool and the volume recorded as residue.
, The difference between 100 ml. and the sum of the recovery and the residue
shall be calculated and recorded as distillation loss.
Accuracy.-With proper care and attention to detail, duplicate results
obtained for initial boiling point and end point, respectively, should not differ
from each other by more than 6 F. (3.3 C.). Duplicate readings of the volume
of distillate colleded in the cylinder when each of the prescribed temperature
points is reached should not differ from each other by more than 2 m!. In case
observations are made on the basis of prescribed percentage points, the differences in temperature readings should not exceed the amounts equivalent to
2 m!. of distillate at each point in question.
Correction for Barometric Pressure.-The actual barometric pressure shall
be ascertained and recorded, but no correction shall be made except in case of
dispute. In such cases the temperature points shall be corrected to 760 mm.
(29.92 in.),'by the use of the Sydney Young equation, as follows:
For Centigrade readings:

cc = 0.00012(760 -

P) (273 +tc)

For Fahrenheit readings:

C,=0.00012(760 - P)(46C +t,)

in which Cc and 9, are, respectively, corrections to be added to the observed

temperature to or tIl and P is the actual barometric pressure in millimeters of
mercury.
The following table ill a convenient approximation of the corrections as
calculated by the above equation.

1712

FIXED OILS, FATS AND WAXES

Correction 10 per 10-mm.
Difference in Pressure

Temperature Range

Deg. Cent.

Deg. Fahr.

10- 30
30- 50
50- 70
70- 90
90-110
110-130
130--150
150-170
170-190
190-210
210-230
230-250
250-270
270--290
290-310
310--330
330-350
350-370
370--390
390--410

UO- 86

Deg. Cent.

0.35
0.38
0040
0.42
0045
0.47
0.50
0.52
0.54
0.57
0.59
0.62
0.64
0.66
0.69
0.71
0.74
, 0.76
0.78
0.81

86-122
122-158
158-194
194-230
230--266
266-302
302---338
338--374
374-410
410-446
446-482
482-518
518-554
554-590
590-626
626-662
662-698'
698-734
734-770

Deg. Fahr.

0.63
0.68
0.72
0.76
0.81
0.85
0.89
0.94
0.98
1.02
1.06
1.11
1.15
1.19
1.24
1.28
1.32
1.37
1.41
1.45

Significance of the Test.-The lighter portions, for example, those between

150 and 200, burn much better than those between 250 and 290 0 ; the heavy
portions of American petroleum burn much better than those of the Russian Gifs.
The end point shall not be higher than 625 F. For lighthouse kerosene
(the highest grade) the U. S. requirements are as' follows : first drop not lower
than 160 C.; 10% over below 175, 90% over below 270, end-point not higher
than 280 and 97% recovered as distillate in the receiver.
The averages from four samples of Caucasian and terr'samples of American
oils subjected to this test were .as follows, in per ce,'nt by volume: 1
Below 1500 C.

Caucasian petroleum ........ '.' ........

American petrolmim ................. ,

8.0
16.9

150-2900 C.

86.6'
57.1

Above 2900 C.

504
26.0

The portion boiling below 150, C. is by some regarded as the gasoline or

napht\l,a portion of the kerosene.
Determination of Sulfur.-The deleterious effect ,of, the oxides of sulfur
upon hangings and bindings-as well as upon the human system-is well
known, sulfuric acid being their ultimate product. The sulfur exists in combination, partly as compounds formed from the sulfuric acid used in refining
and partly as alkyl sulfides. Its qualitative detection may be efJected by heating the oil to its boiling-point with a bright piece of sodium or potassium. If
~10 To be added in case barometric pressure is below760mm.; to be subtract,ed in
case barometric pressure is above 760 mm.

FIXED OILS, FATS AND WAXES

1713

sulfur compounds be present, a yellowish layer is formed upon the metal. After
cooling add distilled water drop by drop until the metal is dissolved, and test
for sulfides with sodium nitroprusside.
TENTATIVE METHOD OF TEST FOR SULFUR IN PETROLEUM
OILS BY LAMP METHOD 11

A. S. T. M.

D;SIGNATION: D

90-34 T

12

This method shall be used for the determination of sulfur in kerosene, _

petroleum naphtha, and other petroleum oils which can be burned completely
in a wick lamp. This method may be applied to motor fuels which are mixtures
Df gasoline and volatile non-petroleum oils.
NOTE.~When this method is applied to mixtures containing carbon bisulfide
approximately 8.5 per cent of the sulfur present as carbon bisulfide is usually determined.

Apparatus.-The apparatus shall consist of the following: (a) Absorber.An absorber of chemically-resistant glass conforming to the dimensions shown
in Fig. 242. The larger of the two bulbs of the absorber shall be filled at least
two-thirds full with pieces of chemically-resistant glass rod from 8 to 10 mm. in
length and from 5 to 6 mm. in diameter, or with perforated beads of chemically
resistant glass 5 to 8 mm. in diameter.
(b) Chimney.-A chimney of chemically-resistant glass conforming to
the dimensions shown in Fig. 242, and connected with the absorber by a cork
stopper.
.
(c) Spray Trap.-A spray trap of chemically-resistant glass conforming to
the dimensions shown in Fig. 242, and connected with the absorber by a cork
or by a rubber stopper.
(d) Lamp.-A small lamp of about 25-ml. capacity. The lamp may conveniently consist of a 25 to 35-m1. Erlenmeyer flask, preferably of the shape
shown in Fig. 242, and a cork carrying a piece of glass tubing 40 mm. in length,
from 6 to 7 mm. in outside diameter, and 3 mm. in inside diameter, with ends
ground or fused level. The wick tube should project above the cork at least
10 mm., and the cork shall have a vertical groove in its side so that air may enter
the flask to replace the oil as it is consumed.
(e) Cotton Wicking.-Two strands of clean, unused cotton wicking, 10 to 12
cm. in length and weighing about 5 to 6 mg. per em. per strand.
(!) Filter Pump.-A filter pump or other means for continuous suction.
Solutions Required: (a) Hydrochloric Acid.-Prepare a solution of HCI in
distilled water which will contain 2.275 g. of HCI per liter and check carefully
for accuracy.
11 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
12 This is a Tentative Standard and under the Regulations of the Society is subject
to annual revision. Suggestions for revision should be addressed to the Headquarters
of the Society, 260 S. Broad St., Philadelphia, Pa. Issued, 1921; Revised, 1924, 1926,
1929, 1930, 1934.

1714

FIXED OILS, FATS AND WAXES

"',Sef Screw wifh

Fiher7ip
FIG. 242.-Apparatus for Determining Sulfur in Oils.
I

NOTE.-In the case of those dimensions for which no specific.tolerance are designated
above, the permissible variation is 10 per cent to the ,nearest millimeter, provided,
.
however, that in no case the deviation be greater than 5 mm.-

FIXED OILS, FATS AND WAXES

1715

(b) Sodium Carbonate.-Diss.olve 3306 g. of Na 2C0 3 in one liter of distilled water. Ten milliliters of this sofution should neutralize exactly 10 ml.
of the HCI solution.
(c) Methyl Orang e.-Dissolve 0.004 g. of methyl orange in one liter of
distilled water.
Procedure.-The sulfur shall be determined as follows:
(a) Pass two strands of the new cotton wicking through the wick tube so
that they are not twisted but are parallel in the wick tube. Pour into the clean,
dry lamp about 15 ml. of the oil to be tested, and insert the cork containing the
wick tube. After the wick has become saturated with the oil, trim it off as
nearly as possible to the top of the wick tube with a pair of sharp scissors, finally
adjusting the wick height by touching the top of it with the tip of a finger;
Weigh the oil and lamp assembly to the nearest 0.001 g. Run a blank determi'uation at the same time and under the same conditions by burning sulfurfree alcohol in a similar lamp. Thoroughly rinse out the absorber with distilled
water, and put into it exactly 10 ml. of the sodium carbonate solution from an
accurately calibrated pipette, and then dilute the solution with 10 ml. of distilled
water.. Rinse out also the chimney and the spray trap with distilled water,
dry the chimney, and connect both to the absorber as shown in Fig. 242.
Prepare the apparatus for the blank determination in the same manner, putting
into it exactly 10 ml. of the sodium carbonate solution and 10 ml. of distilled
water. Apply gentle suction to both absorbers; light both the oil lamp and the
alcohol lamp and then place them in position under the chimneys so that the
top of each wick tube extends into the chimney not more than 1 or 2 mm. above
the edge. Use a sulfur-free flam~, such as an alcohol lamp for lighting the
lamps; do not use matches. Adjust the suction so that a steady flame 12 to
18 mm. in height and free from smoking is obtained (Note 1). The suction
shall be adjusted so that air is drawn through both absorbers at approximately
the same rate. The maintenance of proper flame height usually requires that
the wick be flush with the top of the wick tube for naphthas or motor fuels and
a little higher for illuminating oils. The room shall be free from draughts.
Continue burning for from one to one and one-half hours, or less if the sulfur
content of the oil is high. During this time the oil should be consumed at the
rate of about 2 to 2.5 g. per hour.
(b) Extinguish the flames and stop the suction on both absorbers. Weigh
the oil lamp immediately to the nearest 0.001 g., and determine by difference
the weight of oil consumed. Working with the blank first, disconnect the spray
trap and chimney and wash them thoroughly with the methyl orange solution,
using a wash bottle with a very fine jet and collecting the washings in the absorber. Use approximately 35 ml. of the solution for washing. Carefully
titrate the faintly-yellow solution with the HCI solution. During titration,
carefully agitate the contents of the absorber either by Itlte_rnate sucking and
blowing through a rubber tube held between the lips, or else by the use of a
suitable rubber-syringe bulb. As the end point is approached, draw the liquid
back and forth between the bulbs after each addition of acid, agitating as before.
When the first permanent pink color appears, the end point has been reached.
Read and record the volume of-HCl solution used (Note 2).
(c) Rinse the chimney and the spray trap used in the actual determi~ation
into the absorber to which they were connected, exactly as has been prescribed

1716

FIXED OILS, FATS AND V\;AXES

for the blank. If at this point the solution.in, the absorber should have a pink
color, too much oil has been burned and the determination shall be repeated, but
burning for a shorter time. The'temporary pink color ,that sometimes appears
-in the chimney when rinsing it out should be disregar~ed. Titrate just as,in
the blank, continuing to add the HCl solution until, the color matches that
obtained in the blank. Read and record the volume of HCl solution used.
NOTE 1.-If it is not possible to obtain a good flame free from smoking, as sometimes happens if the oil contains considerable percentages of benzol, dilute the oil with
an equal weight of sulfur-free anhydrous alcohol before putting the sample into the
lamp. If, while maintaining the specified flame height the flame still smokes, a smaller
flame is permissible. In order to get the correct, value when calculating the percentage
of sulfur present in this case, allowance shall be made for the fact that the s"ample as
burned in the lamp is one-half alcohol. This shall be done by doubling the percentage
of sulfur obtained in computation by the formula given below.
NOTE 2.-For accurate work, conditions shall be such that a minimum or zero
blank is obtained. If the titration of the blank requires more than 0.1 ml. of the HCI
solution per gram of sample burned, which corresponds to 0.01 per cent of sulfur, a
high degree of accuracy cannot be expected in the case of oiL'l of low sulfur content.

Calculate the sulfur content of the oil by substituting the proper values in
the following formula:
I
_ milliliters of HCl for blank-milliliters of RCI for sample
P ercentage 0 f su If ur grams 0 f 01'1 b urne d X 10
This formula is correct only for the standard solutions specified, pp. 1713,
1715, 1 ml. of each being equivalent to 0.001 g. of sulfur. The use.of solutions
of any other strength, such as 0.1 N, involves more complicated calculation and
is not advisable.
As a check on the determination, a sample of known sulfur content shall
be analyzed. Such,a reference sample may be made by adding a pure volatile
sulfur compound, such as carbon bisulfide, to sulfur-free anhydrous alcohol.
This check shall indicate from 85 to 100% of the sulfur present.
Accuracy.-Duplicate determinations by the sarrle operator should not
differ by more than the amounts given in the following tabulation:
Sulfur Found, per cent

Average Deviation
from Mean, i
per cent Sulfur

Under. 0.125 ............................ '...............

0.125 to 0.25 ..... ',' ............................. -" .".
0.25 to 0.50. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...
Over 0.50 ..............................................

0.0025
0.005
0.01
0.02

For different operators corresponding deviations may be double those given

above.
The barium sulfate can be precipitated and determined for rapid work by
the Jackson Turbidimeter.
The percentage of sulfur in a kerosene should, not exceed 0.05; the Pennsylvania oils contain usually 0.02 to 0.03, the Lima 0.04 to 0.05. 13
Detection' of Acidity.-Shake equal quantities of oil and warm water in a
test-tube, pour off the oil, and test the water with litmus paper! If the water be
stron~ly acid, the quantity may be determined B:sjll-"-Free Acid," page 1782.
13 Kissling, Ch., Rev. Fett und Harz. Ind., 14, 157, 1906.

FIXED OILS, FATS AND WAXES

1717

The acid in this case is most probably sulfuric, coming f.rom the refining
process.
Sulfuric Acid Test.-The object of this test is to judge of the degree of
refinement of the oil, a perfectly refined oil giving little or no color when submitted to the process. One hundred grams of oil and .40 grams of sulfuric acid,
1.73 specific gravity, are shaken together for two minutes in a glass-stoppered
bottle and the color of the acid noticed. For comparative work this color is
matched by solutions of Bismarck brown. 14
Mineral Salts.-Salts of calcium or magnesium when dissolved in the oil
diminish its illuminating power; their action is to form a crust on the wick and
prevent access of air.
Redwood 15 states that 0.02 gram of either of these salts in 1000 grams of oil
diminishes the illuminating power 30 to 40% in eight hours.
They are determined by distilling 100 to 200 m!. of the oil down to about 20
ml., evaporating and igniting this residue, and subsequently treating with hydrochloric acid. The calcium and magnesium are then determined in the usual
way'.
Determination of Water.-By rubbing the oil together with a little eosin on
a glass plate the oil will take on a pink color if water be present.
The evaporation method is approximate and applicable only to heavy oils
and greases., Its accuracy even with heavy greases is questionable.
Dilute 16 the oil with an equal volume of benzol, whirl it vigorously in a
centrifuge until the separated layer of water does not appear to increase in
volume. However, as water is somewhat soluble in any diluent used and also
in oils, a portion of the water content will fail to appear, consequently the
,method in which a diluent is used cannot be considered accurate. It.is advisable first to agitate the diluent vigofDusly with water and then to separate
with the centrifuge in order to saturate it with water before using. See also
Dean's test under Fuel Oil.
Groschuff 17 states that 100 grams of benzene will dissolve 0.03 gram of
water at 3 C. and 0.337 gram of water at 77 C., whereas petroleum products
(density 0.792) will dissolve from 0.0012 gram at 2 C. to 0.097 gram at 94 C.
While water to the extent even of 3 or 4% is apparently without influence
on the viscosity, 1% extinguishes the flame when making the flash test.
Color.-This test has lost its importance, since oils are sometimes satisfactory despite their yellow color. The determination is usually, made with
the Saybolt colorimeter,Is in which the depth of the oil is changed in a cylinder
until it matches the color of standard pl;:ttes of uranium glass.
(b) Lubricating Oils
The tests to be made are, in the order of their importance, viscosity, specific
gravity, evaporation, cold test, pour point, flash test, fire test, 'test for soap,
carbon residue test, friction test. Saponification value, tarry matter insoluble
in 88 naphtha, color and added impurities are also determined.
J. Soc. Chern. Ind., 15, 678, 1896.
Dingler, Pol. J., 255, 427, 1887.
16 Reported by Allen and Jacobs.
Bureau of Mines Technical Paper No. 25,.1912.
17 Groschuff, Chern. Abs., 5, 2550, 1911.
18 Tech. Paper 323B.

14
15

1718

FIXED OILS, FATS AND WAXES

The office of a lubricant is to prevent the attrition of axle and journal box
by interposing itself between them in a thin layer, upon wp.ich, the shaft revolves.
The ideal lubricant is that which has the greatest adhesion to surfaces and the
least cohesion among its own particles, or, as the practical man expresses it,
the most fluid oil that will do the work and stay, in plac~. The determination
of its viscosity or " body" is then of the first importande.
Viscosity is the degree of fluidity of an oil or its iriternal friction. It is
independent of the specific gravity of the oil, although this in the pipette
instruments influences the time of efflux. Within certain limits it may be
taken as a measure of the value of oil as a lubricant, by comparing the viscosity
of the oil under examination with that of other oils which have been found
to yield good results in practice.
The instruments employed for its determination may be divided into two
classes-pipette viscosimeters, giving the time of efflux, as those of Engler,
Saybolt, and others, and torsion viscosimeters, giving the retardation' due to,
the oil, those of Mac Michael and Doolittle.
In expressing viscosity, consequently, it is necessary to give the name of
the instrument with which it is determined. It is sometimes expressed. as
specific viscosity, that is, the time of the oil divided by the time of water; this
is only comparative when done with instruments of the same name, that is,
specific viscosity Engler is not the same figiIre as specific viscosity Saybolt.
Besides this manner of expressing viscosity, it is measured in absolute (C.G.S.)
units or dynes per cm. This unit of viscoity has received the name of Poise
from Poiseuille, one of the first to express this property in absolute measure.
This is possible when the diameter of the orifice, its length, the quantity and
specific gravity of the oil, its time of efflux and change of head are known.
Where it is impracticable to determine
all these data by direct measurements,
the readings of a viscosimeter may be
changed into dynes by determining the
viscosity in secqnds of standard solutions of glycerine, the viscosity of these
being determined in dynes from tables
of physical constants. Or it may be
done by use of the tables on pages 1726
and 1799.'
Engler Apparahls.-~De8cription.
The apparatus (Fig. 243) 'consists of a'
flat, brass cylindrical vessel,A, 106 mm.
in diameter and about 62 mm. deep,
holding 240 mI.,' provided with a jet 2.9
mm. in diameter and 20 mm. long.
This vessel is gilt inside and the jet, in
_~-'-_ _ _ _-J._ _~' the standard instruments, is of platinum
FIG. 243.---=-Engler Viscosimeter.
-ordinarily it is made of brass; the vessel is surrounded with a bath, B, either
of water or oil, provided with a stirrer and heated by ~ ring b~tner. The jet is
clo!3ed by the woo,den valve, P, passing through the_cOVer, and a thermometer,
, c, ;;hows the temperature of the oil; three studs show-the height to which A is

FIXED OILS, FATS AND WAXES

1719

filled and at the same time when it is level. The oil ordinarily is discharged
into the 200-ml. flask, although in case the oil or time be limited, 100 or 50 m!.
may be used and the time of efflux multiplied by 'a suitable factor. The instrument is standardized with water, 200 m!. of which at 20 C. should run out in
from 50 to 52 seconds.
Manipulation.-The instrument is thoroughly cleaned with alcohol and
ether if necessary ane! dried; any suspended matter is removed from the oil,
which is poured into it up to the level of the studs, stirred until 20 C. is reached
and the bath adjusted to the same temperature. The flask is placed beneath
the orifice, the plug raised and the time required for 200 mI. of oil to flow out is
noted; this is divided by the water value of the instrument and gives then relative or specific viscosity. If only 50 ml. are allowed to run out the time must
be multiplied by 5, and if 100 ml., by 2.35. If only 50 ml. were put in all.d 40 ml.
allowed to run out, multiply this time by 3.62 to obtain the time for 200 m!.;
if 66 ml. and 50 ml. run out, multiply by 2.79.1 9 If it be desired to express the
viscosity in absolute measure,(C.G.S. units) it can be done by reference to the
table en page 1726. It should be noted that specific viscosity obtained with
a different type of instrument, e.g., the Saybolt, is not the same as with the
Engler.
STANDARD METHOD OF TEST FOR VISCOSITY BY MEANS
OF THE SAYBOLT VISCOSIMETER 20
A. S. T_. M.

DESIGNATION: D

88-36 21

Scope.-(a) This method shall be used for determining the Saybolt viscosity of petroleum products and lubricants.
(b) The Saybolt Universal Viscosimeter shall be used only for oils with times
of flow of more than 32 sec. There is no maximum limit to viscosity to be
measured by the Saybolt Universal Viscosimeter but, in general, liqu1ds having
an outflow time of the order of 1000 sec. and higher, Saybolt, Universal, are
tested more conveniently by means of the Saybolt Furol Viscosimeter.
(c) The Saybolt Furol Viscosimeter shall be used only for oils with times of
flow of more than 25 sec. The outflow time of the Furol (Note) instrument is
approximately one-tenth that of the Universal.
NOTE.-The word "Furol" is a contraction of the phrase "fuel and road oils."
Apparatus.-The apparatus shall consist of an oil tube, bath, receiver, thermometers, timer, and withdrawal tube, conforming to the requirements specified in the following Paragraphs (a) to (g) :
.
Gans, Chem. Revue der Fett und Harz. Ind., 6, 221, 1899.
Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
21 This method 'is issued under the fixed designation D 88; the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1921; Adopted in Amended Form, 1926; Revised, 1930,
1933,1936.
19

20

FIXED OILS, FATS AND WAXES.

1720

(a) Oil Tube.-The oil tube (illustrated in Fig. 244) shall be entirely of corrosion-resistant metal, with or without plating, and shall conform to the dimensional requirements shown in Table I within the permissible variations prescribed. The lower end of the oil tube shall be I?~ovided with a nut for
locking it in place in the bath and with a cork or other suitable device to prevent
flow until the test is started. For convenience a string or its equivalent nuty
be attached to the cork for rapid removal.
The oil tube shall be standardized, and any correctIon in excess of 0.2%
shall be applied. The time of flow shall be within 1% (Note 1) of the time
as obtained with the National Bureau of Standards' master Saybolt oil tube
(Note 2).

NOTE I.-For routine testing, oil tubes having corrections as high as 2% may
be employed, provided, however, that any new oil tubes obtained after the adoption of
this method shall not have corrections higher than 1 %. For referee testing, oil tubes
having corrections higher than 1 % shall not be employed.
.
The oil may be standardized by anyone of the following methods:
(a) Determination of outflow time for American Petroleum Institute oil standard
and application of appropriate correction to obtain established value for the
oil standard.
(b) Calibration by National Bureau of Standards subsequent to January 1, 193i!.
I (c)
Comparison with a tube certified by the ,National Bureau of Standards subsequent to January 1, 1935, provided the tube shows no evidence of damage
or injury since certification.
. NOTE 2.-This tube, conforming to the A",'S, T. M. dimensional specifications, has'
outflow times calibrated against oil standards under the maintenance of the American
Petroleum Institute Committee on Viscosity Standards.
TABLE I.-DIMENSIONS OF OIL TUBES
Saybolt Universal Viscosimeter

Saybolt Furol Viscosimeter

Dimensions
Minimum,
cm.

Inside diameter of outlet tube .....

Outside diameter of outlet tube at
lower end ....................
Length of outlet tube a .
Height of overflow rim above bottom of outlet tube a . . . . . .
Outside diameter of overflow rim,
at the top a. '. \ ...
Diameter of container 11
Depth of cylindrical part of con-,
tainer a . . . . . . . . . . . . . . . . . . . . . .
Diameter of container between
bottom of cylindrical part of
container and top of outlet tube a

Normal,
cm.

Maximum. IVlinimum,
cro.
cm.

0.1750

0.1765

0.1780

0.28
1.215

0.30
1.225

0.32
1.235

12.40
b

2.955
8.8
0.9

12.50

2.975

12.60
3.30
2.995

Normal,
cm.

Maximum,
cm.

0.313

0.315

0.317

0.40
1.215

0.43
1.225

0.46
1.235

12.40

12.50

2.955

2.975

8.8

0.9

--

12.60
3.30
2.995

This dimension is identical in the Saybolt Universal and the Saybolt Furol instruments.
The minimum value shall preferably be not less than 3.2 cm.
e The section of overflow rim shall be bounded by straight lines! except that a fillet is permissible at
the junction with the bottom of the gallery,
.
a

(b) Bath.-The bath shall serve as a support to hold the oil tube in a
vertical position and as a container for the bath liquid. The bath sha.ll be
equipped with a stirring device and with means for heating or cooling. The
source of heat or refrigeration shall be more than Vi in. (3 cm.) from the oil
tube. If an external heater is used it must be mo1e-than 2 in. (5 cm.) from the
oil tube.

FIXED OILS, FATS AND WAXES

1721

The bath temperature necessary to maintain thermal equilibrium (while

the oil in the oil tube is well stirred by the oil-tube thermometer) shall be within
0.1 F. (0.06 C.) of the standard temperatures of 70 F. (21.1 C.) or 77 F.
(25 C.) or shall not exceed 100.25,122.35,130.5,141.0;181.5 or 212.0 F.
(37.9, 50.2, 54.7, 60.6, 83.1 or 100.0 C.), respectively, for the standard
temperatures tnentioned in Section 3. The level of the bath liquid shall not
be lower than U in. (0.5 em.) above the overflow rim of the oil tube. For
referee testing the bath liquid must be one which in the bath used will meet the
preceding bath-temperature conditions (Note).
NOTE.-These temperature requirements can be met with water, aqueous solutions, .
and some baths with oil. In routine testing oil is generally used as the bath medium.
This is allowable provided the temperature of the oil bath is adjusted so that the necessary condition of thermal equilibrium is maintained. It may be necessary to maintain
the oil bath at slightly higher temperatures than those specified in the preceding paragraph. Temperature differentials between oil bath and oil tube necessary to maintain
thermal equilibrium may be double those given above.
Overflow
Rim ....

fa he Drained
before Sfarfing Flow

{em.t O/em.
Irrsicle Diameter
af 6radualiorr Mark

fevaf/on

of f"/osk.

FIG. 244.-Sectional View of

Standard Oil Tube.

FIG. 245.-Receiving Flask.

(This type Of flask is recommended
as convenient and durable.)

(c) Receiver.-The receiving flask (see Fig. 245) shall be of glass with a
capacity up to the graduation mark on its neck of 60 ml.0.05 ml. at 68 F.
(20 0 C.). At the graduation mark, the inside diameter of the neck of the flask
shall be 1 cm.O.l cm. The cylindrical portion of the neck of the flask shall

1722

FIXED OILS, FATS AND WAXES

extend not less than 0:3 cm. above anq below the gr~duation mark. The
graduation mark shall be 5.8 cm.1.0 cm. above the bottom of the flask.
The maximum outside diameter shall be less than 5.5 cm.
(d) Oil-Tube Thermomeiers.-The oil-tube therrlOmeters shall conform
to the following requirements. They shall include tWQ sets of six thermometers
each, one set being graduated in Fahrenheit degrees and the other in Centigrade
degrees, the ranges being chosen to include the temper~tures commonly used in
testing. To prevent contact of the thermometer with the orifice in the oil tube
a suitable support shall be attached to the enlargement of the thermometer
stem.
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGES AND SUBDIVISIONS:
Range 66 00 to 800 F., subdivisions in 0.20 F. for tests at 70 0 and 77 0 F.
"
94 to 1080 F.,
"
" " " " " 1000 F.
"
1200 to 1340 F.,
"
. , " " " " 122 0 and 130 0 F.
0
0
"
1340 to 1480 F.,
"
" " " " " 14000 F.
1740 to 1880 F . , "
" " " "180 F.
"
"
204 to 218 F.,
"
" " " " " 210 0 F.
or
Range 19 0 to 27.C., subdivisions in 0.10 O. for tests at 21.1 0 and 25 0 C.
"
340 to 42 0 C.,
"
" . " ' " " , , 37.So C.
4900 to 5700 C.,
""
" " 50 0 and 54.40 C.
"
"
"
57 0 to 65 0 C.,
" " " " " 60 0 C.
79 to 87 C.,
"
" / " " " , , 82.20 C.
"
95 0 to 103 0 C.,
"
""
" " 98.9 0 C.
TOTAL LENGTH: 252 to 256 rom. (9.92 to 10.08 in.).
STEM: Plain front, enamel back, suitable thermometer tubing. Diameter, 6 to 7 mm.
(0:236 to 0.275 in.). The stem shall be made with an enlargement not less than 4
nor more than 7 mm. (0.16 to 0.28 in.) in length, having a diameter 2 to 3 mm.
(0.08 to 0.12 in.) greater than that of the stem, the bottom of the enlargement
being 114 rom. (4.5 in.) above the bottom of the bulb.
BULB: Corning normal or equally suitable thermometric glass.
,
Length, 25 to 35 mm. (1.0 to 1.4 in.).
Diameter, not less than 5 mm. (0.197 in.) and not greater than that of stem.
DISTANCE TO LOWEST SPECIF.IED GRADUATION FROM BOTTOM OF BULB: 135 to 150
I
mm. (5.3 to 5.9 in.).
DISTA.."1CE TO HIGHEST SPECIFIED GRADUATION .FROM Top OF THERMOMETER: 20 to
3.5 rom. (0.8 to 1.4 in.).
.
CONTRACTION CHAMBER: To be of long narrow type, top to be not more than 60 mm.
(2.36 in.) above bottom of bulb, mercury to stand in contraQ!,ion chamber at 32 0 F.
(0 0 C . ) . '
.~._
.
EXPANSION CHAMBER: To permit heating the thermometer 90 0 F. (50 0 C.) above
0
highest temperature on scale, and in all cases to permit heating to 212 F. (100 0 C.).
FILLING ABOVE MERCURY: Nitrogen gas.
Top FINISH: Glass ring. .

GRADUATION: All lines, figures and letters clear cut and distinct. They shall be in
black except for the special markings below. The ~rst and each succeeding 10 F.
or 0.5 0 C. line to be longer than the remaining lines., Graduations to be numbe~ed
at each multiple of 2 0 F. or 10 C.
IMMERSION: Total.

SPECIAL MARKING: The test temperatures specified in Section 3 shall be shown in

full at the appropriate points on the thermometer scale; th~ graduations corresponding to the test temperatures and the numbers to.l>e in red. A serial number
and the manufacturer's name or trade mark shall bc-etched on the stem.
SCALE ERROR: The error at any point of the scale shall not exceed 0.20 F. (0.10 C .).

FIXED OILS, FATS AND WAXES

1723

STANDAJl,DIZATION: The thermometers are to be standardized for the condition of total

immersion. Correction for emergent stem shall not be applied.
CASE: Each thermometer shall be supplied in a suitable case on which shall appear
the marking, "A. S. T. M., Saybolt Viscosimeter Thermometer" and the range.
NOTE.-For the purpose of interpreting these specifications the following definitions apply:
The total length is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning
of the enamel backing.
.
The top of the thermometer is the top of the finished instrument.

(e) Bath Thermometers.-Oil-tube thermometers, or other means for

measuring temperature of !Lt least equal accuracy, shall be used in the bath.
(!) Timer.-The stop watch or other timing devi'ce used shall be graduated
in divisions of 0.2 sec. or less, and shall be accurate to within 0.1 %'when tested
over a 60-min. period (Note).
NOTE.-Electricai timing devices are permissible provided they are accurate and'
capable of being read to 0.2 sec. Timing devices actuated by synchronous motors shall
be used only on electric circuits of controlled frequency.
(g) Withdrawal Tube.-The tube or pipette used for draining the gallery
shall have a smooth tip of about 3 mm. Qutside dIameter and about 2 mm.
inside dllimeter.
Temperature of Testing.-(a) With the Saybolt Universal Viscosimeter,
determinations shall be made at 70,100, 130 or 210 F. (21.e, 37.8, 54.4 or
,98.9 C.).
(b) With the SaY)bolt Furol Viscosimeter, determinations shall be made at
77, 100, 1i2 or 210 F. (25, 37.8, 50 or 98.9 C.). In tests on road and
paving materials, determinations may also be made at 140 and 180 F.
(60 and 82.2 C.).
Procedure.-(a) Viscosity determinations shall be made in a room free from
drafts and rapid changes in temperature. For .standardization and referee
tests the room shall be between 68 and 86 F. (20 and 30 C.) and the actual
temperature shall be recorded. For routine tests, temperatures up to 100 F.
may be employed without introducing errors in excess of 1 %. Determinations
shall not be made at temperatures below the dew point of the atmosphere sur, rounding the instrument.
(b) The oil tube shall first be cleaned with an effective solvent, such as
benzol, and excess solvent shall be removed from the gallery.
(c) All oil shall be passed through a 100-mesh wire strainer before it is
introduced into the oil tube. After the tube is cleaned, a quantity of the oil
to be tested, sufficient to wet the entire surface of the tube, shall be poured
into the tube and allowed to drain out (Note). The cork stopper shall be
inserted not less than >i in. nor more than % in. into the lower end of the air
chamber at the bottom of the oil tube. The cork shall fit tightly enough to
prevent the escape of air, as evidenced by the absence of oil on the cork after
it is withdrawn.

NOTE.-The plunger commonly supplied with the viscosimeter shall never be used
on instruments maintained as standards.

1724

FIXED OILS, FATS AND WAXES

(d) If the test temperature is above that_oJ: the room, the oil shall be heated
to not more than 3 F. (1.7 C.) above the temperature( of test and if the temperature is below that of the room, the oil shall be cooled to not more than 3 F.
(1.7~ C.) below the temperature of test. The oil shall\be poured into the oil I
tube until it ceases to overflow into tlle galleryt The, 6il in the oil tube shall
be kept well stirred with the oil-tube thermometer, care being taken to avoid
hitting the outflow tube. The bath temperature shall' be adjusted until the
oil temperature remains constant. After thermal equilibrium has been attained
no further adjustments shall be made in the bath temperature. The test
results shall be discarded if the indicated bath temperature varies by more
than 0.05 F. (0.03 C.) in tests at 70, 77, 100, 122 and 130 F. (21.1,
25, 37.8, 50 and 54.4 C.) or by more than 0.1O F. (0.06 C.) in tests at
140, 180 and 210 F. (60~, 82.2 and 98.9 C.).
(e) After the temperature of the oil in the oil tube has remained constant
within 0.02 F. (0.01 C.) of the desired temperature for 1 min. with constant
stirring, the oil-tube thermometer shall be withdrawn and the surplus oil
removed quickly from the gallery by means of the withdrawal tube so that the .level of the oil in the gallery"is below the level in the oil tube proper. The tip
of the withdrawal tube shall be inserted at one point in the gallery. The test
shall be started over again if the tip of the withdrawal tube touches the overflow
rim (see Fig. 244). Under no conditions shall, the excess oil be removed by rotating the withdrawal tube around the gallery.
(f) The receiving flask shall be placed<in position so that the stream of oil
from the outlet tube will strike the neck of the flask. The graduation mark on
. the receiving flask shall be not less than 10 cm. or more than 13 cm. from the
bottom of the bath. The cork shall be snapped from its position and at the
same instant the timer shall be started. The timer shall be stopped when
the bottom of the meniscus of the oil reaches the mark on the neck of the
receiving flask.
Reporting Results.-The,time in seconds as determined by the prescribed
procedure, with the proper .calibration correction, is/the Saybolt Universal
(or Saybolt Furol) viscosity of the oil at the temperature at which the test is
made. Results shall be reported to the nearest 0.1 sec. for viscosity values
-below 200 sec. and to the nearest whole second for values 200 sec. or above.
Accuracy.-With proper attention to details of method of procedure, results
in different laboratories with different operators under referee or standardization
conditions of testing, should not differ by more than 0.5%:-,
MacMichael's Viscosimeter.-This is an instrument of the torsion type
in which a disk is suspended in a cup of fluid, the latter being rotated, and the
torsion it produces on the disk noted. The disk is suspended by a gold-plated,
steel wire 10 inches long, held between two grooveu pins at the top of the standard. The brass cup is oil jacketed, the oil covering the wires 72 inch and heated
electrically, or it may be cooled by ice or brine. A bent thermometer passing
tlirough the cover indicates the temperature, which may be controlled easily
within a fraction of a degree. The graduated dial at the top of the disk is
rotary and may be easily set to 0. A dash pot filled with engine oil on the
stem of the disk damps the action of the motor. In operating/the instrument.is
levelled, the cup is filled to the mark with the fluid to1>e tested (about 100 mI.),
the temperature adjusted, the cup rotated at about 20R.P.M. and the readings

FIXED OILS, FATS AND WAXES

1725

of the dial noted. These are in degrees of angular deflection, 300 to the circumference, and noted as M. = 1/1000 poise. Water at 20 should read
10 M. The smallest, or " practical unit" is 10M.: by changing the decimal
point, practical units, absolute units or specific viscosity may be obtained at
one reading of the dial, no calculations being required. Three strengths of wire
are supplied, giving water readings of 1, 10, and 100, covering ranges of
viscosity to 5000 M. or 5 poise.
The instrument should be calibrated by solutions of pure granulated sugar
dried over sulfuric acid, the absolute viscosity of which is shown below. The
accuracy of the instrument is well within 0.5%. The instrument is very rapid,
the time required for a reading being very short. It can be applied to colloidal
or lumpy solutions or suspensions, as limes, clay, glues, gums, explosives, paints,
catsups or even jams with fruit seeds in them. The table below gives the wires
applicable to different liquids.

,.

CONSTANTS OF A SET OF TYPICAL WIRES

B.&S.
Gauge
No.

Diameter
in
Inches

Safe
Gm.-Cm.
Limit
perM
Divisions Division
M

Centi-

Centipoises Range

Materials that Illustrate

Range of Viscosities

poises

per M

-2050
1280

Minimum

Maximum

20000
12000

256000
179000

22
23

.02535
.02257

125
140

.58905
.3675

24
.25
26
27
30

.02010
.01790
.01594
.01419
.01002

158
177
199
225
315

.2295
'.13311 .
.10557
.05967
.016065

800
464
368
208
56.0

8000
4500
3500
2000
500

126000
82000
73300
46800
17600

32
34

.00795
.00630

398
503

.005355
.001836

18.7
,6.4

180
60

6760
3220

} Heavy starch solutions, double boiled

oils, rubber solutions

) greases,
Putt''', viscose,
"""" and
single boiled oils, corn
syrups, etc.
} Raw starch, viscous
oils, glycerin, etc.

VISCOSITY IN CENTIPOISES OF SUGAR SOLUTIONS 22

Per cent Sugar by Weight
T~mp. 0 C.

0 ...............
5 ...............
10 ...............
15 ...............
20 ...............
25 ... , ...........
30 ...............

20

40

60

1.789
1.516
1.306
1.141
1.005
0.894
0.802

3.804
3.154
2.652
2.267
1.960
1.704
1.504

14.77
11.56
9.794
7.468
6.200
5.187
4.382

238
156
109.8
74.6
56.5
43.86
33.78

Absolute Viscosity.-This expresses viscosity in dynes, that is, the force

necessary to produce the acceleration of 1 cm. per second on the mass of a gram.
22

Bingham & Jackson, Bureau Standards Bulletin, No. 298 (1917).

FIXED OILS, FATS AND WAXES

1726

It is independent of the instrument used; Engler numbers can be converted to

absolute viscosity by the following factors:
i
.
Engler No.

Absolute fiscosity Dynes per Sq.

em." \

1. ............................... { .. 0.01Q06Xspl'cific gravity

2 ...................................
5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
10 ...................................
20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
30 ...................................
60 ....... " ................ " ........

0.1146 Xspecific gravity

0.353 X specific gravity
0.72~
Xspecific gravity
1.46
X specific gravity
2.19
Xspecific gravity
4.38
Xspecific gravity

The Engler numbers of 5 or over are quite nearly proportional to the absolute viscosities.
Specific Gravity.-See under Burning Oils, page 1704.
Evaporation Test. 2L-The object of this test is to determine what percentage
of an oil-more especially a spindle oil-is volatile when exposed to nearly the
same conditions as it is on a bearing.
The oil is exposed upon annular disks of filter-paper 1% in. outside diamet~,
with hole % in. in diameter, which have been standing in a sulfuric acid desiccator for several days, contained in a flat watch-glass.
Manipulation.-The watch-glass and paper are weighed-to tenths of a
milligram-and about 0.2 gram of oil brotlght upon it by dropping from a rod,
and accurately ,weighed. The watch-glass is now placed in an air bath, the
temperature of which remains nearly constant at 60 to 65 C. (140 to 150 0 F.),
and heated for eight hours. It is then cooled and reweighed, the loss being
figured in per cent. No oil should be passed which gives.an evaporation of more
than 4%.
.
The following table of results upon some spindle oils shows the relat_ion of
gravity, flash point, and evaporation:
Gravity

.846

.852
.856

Flash,

298
318
348
348
336

F.

Evaporation

Gravity

7.0%
4.4%
2.0%
1.0%
1.4%

.862
.866
.870
.882
\

.'

Flash,

F.

352
366
384
364

..

~vaporation

0.9%
1.7%
0.8%
1.7%
,

NOTEs.-The temperature employed, 65 C., is approximately that attained by

, a bearing (in a spinning frame) after running two hours, thus leaving the oil exposed
to it for eight hours, assuming a ten-hour day.
,
The test is important to the insurance underwriter, because it measures the amount
of inflammable material sent into the air, and hence the liability to cause or aid conflagrations; it is important to the mill-owner, as it in'dicates the quantity of oil left
'
upon the bearing, hence serving its purpose.
The test is made upon other oils by heating them six hours in a shallow dish to
100, 1500, 220, or 300, sometimes in a draft of air or A. S. T. M. D 6-27 method as
for road materials may be used.
I
23 Waidner, Proc. Am. Soc. Test. Mat., Pt. I, 293.vU1l5.
24 Archbutt, J. Soc. Chem. Ind., 15, 326, 1896.'
-

FIXED OILS, FATS AND WAXES

1727

Cold Test.-This may be defined as the temperature at which the oil will just
flow.
M anipulation.-A 4-oz. vial is one-fourth filled with the oil to be examined,
a short, rather heavy, thermometer inserted in it, and the whole placed in a
freezing mixture. When the oil has become solid throughout, let it stand one
hour; the vial is removed, the oil allowed to soften, and thoroughly stirred
until it will run from one end of the bottle to the other. The reading of the
thermometer is now taken by withdrawing it and wiping off the oil with waste
to render the mercury visible.
The chilling-point is the temperature at which flakes or scales begin to form
in the liquid, and is determined similarly, by cooling the liquid 50 at a time.
STANDARD METHOD OF TEST FOR CLOUD AND POUR POINTS 25

A. S. T. M.

DESIGNATION:

97-3426

The cloud point of a petroleum oil is the temperature at which paraffin wax
or other solid substances begin to crystallize out or separate from solution when
the oil is chilled under definite prescribed conditions.
The pour point of a petroleum oil is the lowest temperature at which the oil
will pour or flow when it is chilled without disturbance under definite prescribed
conditions.
The test for cloud point shall be used only for oils which are transparent
in layers 172 in. thick.
The test for pour point shall' be used for all other petroleum oils and may
be used for oils on which the test for cloud point is permitted.
Apparatus.-(See Fig. 246.) The test jar, a, shall be of clear glass, cylindrical form, flat bottom, approximately 1 136 to 1/6 in. in inside diameter and'
4Yz to 5 in. high. An ordinary 4-oz. oil sample bottle may be used if it is
within the above specifications, and no test jar is available.
The thermometer, b, shall conform to the requirements of one of the following specifications:
.
(a) These specifications cover a special thermometer graduated in either
Centigrade or Fahrenheit degrees as specified, the ranges being -38 to +50 0 C.
or -36 0 to +120 0 F., respectively.
TYPE: Etched stem, glass.
LIQUID: Mercury.
.
RANGE AND SUBDIVISION: -38 to +50C. in 1C. or -36 to +120F. in 2F.
25 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
This method comprises a revision of the Standard Method of Test for Cloud and
Pour Points (A. S. T. M. Designation: D .97-33), 1933 Book of A. S. T. M. Standards,
Part II, p. 761, which method it supersedes.
26 This method is issued under the fixed designation D 97; the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1921; Adopted in Amended Form, 1928; Revised, 1930,
1933, 1934. This method has been approved as American Standard by the American
Standards Association.

1728

FIXED OILS, FATS AND WAXES

TOTAL LENGTH: 220 to 224 mm. (8.69 to 8.81 in.).

STEM: Plain front, enamel back, suitable thermometer tubilll!. Diameter, 7.0 to 8.0
mm. (0.28 to 0.31 in.).
I BULB: Corning normal or equally suitable thermometric glass.
~
Length, not over 9.5 mm. (0.37 in.).
Diameter, not greater than stem.
,
DISTANCE TO -38 C. OR -36 F. LINE FROM BOTTOM OF ULB: 120 to 130 mm. (4.73
to 5.12 in.).
i
DISTANCE TO +49'C. OR +120 F. LINE FROM Top OF THERMOMETER: 19 to 25 mm.
(0.75 to 0.98 in.).
EXPANSION CHAMBER: To permit heating to 100 C. or 212 F.
FILLING ABOVE MERCURY: Nitrogen gas.
Top FINISH: Plain.
.
GRADUATION: Ail lines, figures, and letters clear cut and distinct. Lines at multiples
of 5 C. or 10 F. to be longer than the remaining lines. Graduations to be numbered at each multiple of 10 C. or 20 F.
IMMERSION: 108 mm. or 4X: in. The words "10S-mm. immersion" on Centigrade
thermometers or "4X:-in. immersion" on Fahrenheit thermometers and a line
around the stem 108.0 mm. or 4.25 in. above the bottom of the bulb shall be etched
on. the thermometer.
SPECIAL MARKING: "A. S. T. M. Cloud and Pour," a serial number and the manufacturer's name or trade mark shall be etched on the stem.
....
SCALE ERROR: The error at any point of the scale, when the thermometer is standardized as provided below, shall not exceed 0.5 C. or 1 F.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 20 C. or 40 F. for lOS-mm. or 4U-in. immersion and
for an average temperature of 21 C. or 70 F. for the emergent mercury column.
CASE: The thermometer shall be supplied ,in a suitable case -on which shall appear
the marking: "A. S. T. M. Cloud and Pour, -38 to ,+50 C." or "A. S. T. M.
Cloud and Pour, -36 to +120 F." according to the type of thermometer.
NOTE.-For the purpose of interpreting these specifications the following definitions apply:
,
The total length is the over-all length of the finished instrument.
The diameter is that measured with a r.ing gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning
of the enamel backing.
The top of the thermometer is the top of the finished instrument.

(b) These specifications cover a special low ~loud Ind pour test thermometer
graduated 'in either Centigrade or Fahrenheit degrees as specified, the ranges
being -60 0 to +20 0 C. or -70 0 to +70 0 F.
TYPE: Etched stem, glass.
LIQUID: Toluene or other suitable liquid colored red with a permanent dye.
RANGE AND SUBDIVISION: -60 to +20 0 C. in 10 C. or -70"._,to +70 0 F. in 2 0 F.
TOTAL LENGTH: 227 to 231' mm. (8.92 to 9.08 in.).
'
STEM: Plain front, enamel back, suitable thermometer tubing. Di~~eter 7.0 to 8.0
mm. (0.28 to 0.31 in.).
BULB: Corning normal or equally suitable thermometric glass.
,
Length, 8.0 to 9.5 mm. (031 to 0.37 in.).
Diameter, 5.0 to 6.5 mm. (0.22 to 0.26 in.).
DISTANCE TO -57 C. OR -70 F. LINE FROM BOTTOM OF BULB: 120 to 130 mm. (4.73
to 5.12 in.).
!
DISTANCE TO +20 C. OR +68 0 F. LINE FROM Top OF' THERMOMETER: 35 to 45 mm.
(1.38 to 1.77 in.).
EXPANSION CHAMBER: To permit heating,to +60 C. or +140 0 F.
FILLING ABOVE LIQUID: Gas under pressure.
Top FINISH: Plain.
I
GRADUATION: All lines, figures and letters clear cut and distinct, lines at multiples
. of 5 C. or 10 F. to be longer than the remaining lines. Graduations to be numbered at each multiple of 10 C. or 20 F.

..

FIXED OILS, FATS AND WAXES

1729

IMMERSION: 198 mm .. (4~ in.). The words "108-mm. immersion" on Centigrade

thermometers, or "4U-in. immersion" on Fahrenheit thermometers and a line
around the stem 108 mm. (4.25 in.) above the bottom of the bulb. shall be etched
on the thermometer.
.
SPECIAL MARKING: "A. S. T. M. Low Cloud and Pour" a serial number and the manufacturer's name or trade mark shall be etched on the thermometer.
SCALE ERROR: The error at any point of the scale when the thermometer is standardized as provided below, shall not exceed 10 C. or 2 0 F.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 20 0 C. or 350 F. for 10S-mm. or 4U-in. immersion, and
for an average temperature of 21 0 C. or 70 0 F. for the emergent liquid column.
CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking" A. S. T. M. Low Cloud and Pour, -600 to +20 0 C." or "A. S. T. M.Low Cloud and Pour, -70 0 to +70 0 F.," according to the type of thermometer.
N OTE.-For the purpose of interpreting these
specifications, the following definitions apply:
The total length is the overall length of the
finished instrument.
The diameter is that measured with a ring
gage.
The length of the bulb is the distance from the
bottom of the bulb to tht,' beginning of the
enamel backing.
The top of the thermometer is the top of the
finished instrument.

The cork, c, shall fit the test jar, and~

. .~.
shall be bored centrally to take the test ther: 1',
mometer. The jacket, d, shall be of glass or
_,
d"
metal, shall be water tight, of cylindrical
form, flat bottom, about 4Yz in. deep, with
inside diameter % to Yz in. greater than out.f.
side diameter of the test jar.
A disk of cork or felt, e, U in. thick and k~'---ill--.l~
of the same diameter as the inside of the"
~
jacket will be required.
IL--JO:~~
The ring gasket, J, shall be about 136 in.
~_.!tiJ
thick, and made to fit snugly around the out.~
side of the test jar and loosely inside the
jacket. This gasket may be made of cork,
felt or other suitable material, elastic enough to
for Cloud
cling to the test jar and hard enough tohold its FIG. 246.-Apparatus
and Pour Test
shape. The purpose of the ring gasket is to
(as assembled for Cloud Test).
prevent the test jar-from touching the jacket.
The cooling bath, 1], shall be of a type suitable for obtaining the required
temperatures. The size and shape of the bath are optional but a support,
suitable for holding the jacket firmly in a vertical position, is essential. For
determination of pour points below 50 F., two or more baths should be at hand.
The required bath temperatures may be maintained by refrigeration if available,
otherwise by suitable freezing mixtures.

rl t:-

NOTE.-The freezing mixtl1res commonly used are as follows:

For temperatures down to 50 0 F., ice and water. ."
"
" " 10 F., crushed ice and sodium chloride.
"
"
" " -15 0 F., crushed ice and calcium chloride crystals.
" " -70 0 F., solid carbon dioxide and acetone,or gasoline.
"
"
0

1730

FIXED OILS, FATS AND

~4XES

The last named mixture may be made as follows: In a covered metal beaker chill
a suitable amount of acetone or gasoline to lOo-F., or lower, by means of an ice-salt
mixture. Invert a cylinder of liquid carbon dioxide and draw off carefully into a
chamois skin bag the desired amount of carbon dioxide, which through rapid evapora,
tion will quickly become solid. Then add to the chilled ace~one or gasoline enough of
the solid carbon dioxide to give the desired temperature.
I

Procedure.-T,he oil to be tested shall be brought to a temperature at least

25 F. above the approximate cloud point. Moisture, if present, shall be
removed by any suitable method, as by filtration through dry filter paper until
the oil is perfectly clear, but such filtration shall be made at a temperature at
least 25 F. above the approximate cloud point.
The clear oil shall be poured into the test jar, a, to a.height of not less than
2 nor more than 274: in. The test jar may be marked to indicate the proper
level.
The test jar shall be tightly closed by the cork, c, carrying the test ther~
mometer, b, in a vertical position in the center of the jar with the thermometer
bulb resting on the bottom of the jar.
The disk, e, shall be placed in the bottom of the jacket, d, and the test jar-with the ring gasket, j, 1 in. above the bottom shall be inserted into the jacket
The disk, jacket and inside of jacket .shall be clean and dry.
The temperature of the cooling bath, g, snaIl be maintained at 30 to 35 F.
The jacket, containing the test jar, shall be supported firmly in a vertical
position in the cooling bath so that not more than 1 in. of the jacket projects
out of the cooling medium.
At each test thermometer reading which is a multiple of 2 F., the test jar
shall be removed ffom the jacket, quickly but without disturbing the oil,
inspected for cloud, and replaced in the jacket. This complete operation shall
require not more than three seconds. If the oil does not show a cloud wheI.l it
has been cooled to 50 F., the test jar shall be placed in a second bath maintained at a temperature of 0 to +5 F. If the oil does not show a cloud when
it has been cooled to 20 F., the test jar shall be placed in a third bath maintained at a temperature of -'30 to -25 F.
.
/
When such inspection first reveals a distinc~ cloudiness or haze in the oil,
at the bottom of the test jar, the reading of the test thermometer, corrected
for error if necessary, shall be recorded as the cloud point.
Procedure for Pour Point.-The oil shall be poured into the test jar, a, to a
height of not less than 2 nor more than 274: in. The jar-rnay be marked to
indicate the proper level. . When necessary, the oil shall be heated. in a water
bath just sufficiently for pouring into the test jar.
_
The test jar shall be tightly closed by the cork, c, carrying the test thermometer, b, in a vertical position in the center of-the jar with the thermometer
bulb immersed so that the beginning of the capillary shall be Ys ih. below the
surface of the oil.
Heat without stirring to a temperature of 115 F. in a bath maintained at
not higher than 118 F. The oil shall then be cooled to 90 F. in air or in a
water bath approximately 77 F. in temperature. Oils on which a pour point
below -30 F. is expected shall be heated as above with the high cloud- and
pour-test thermometer in position, cooled to 60 F ...l.uHFthe low cloud~ and pourtest thermometer placed in position and the assembly-placed in the jacket. Oils

\
FIXED OILS, FATS AND WAXES

1731

on which a pour paint of above 90 F. is expected shall be heated to 115 F.

or to a temperature. 15 F. above the expected pour point, with the high cloudand pour-test thermometer in position, and the test jar immediately introduced
into the jacket.
.
The disk, e, shall be placed in the bottom of the jacket, d, and the test jar,
with the ring gasket, I, 1 in. above the bottom, shall be inserted into the jacket.
The disk, gasket and inside of jacket shall be clean and dry.
Mter the oil has cooled enough to allow the formation of paraffin wax
crystals, great care shall be taken not to disturb the mass of the oil nor to permit
the thermometer to shift in the oil. Any disturbance of the spongy network of
wax crystals will lead to low and fictitious results.
The temperature of the cooling bath, g, shall be maintained at 30 to 35F.
The jacket, containing the test jar, shall be supported firmly in a vertical
position in the cooling bath so that not more than 1 in. of the jacket projects
out of the cooling medium.
Beginning at a temperature 20 F. before the expected pour point, at each
test .thermometer reading which is a multiple of 5 F., the test jar shall be
removed from the jacket carefully and shall be tilted just enough to ascertain
whether there is a movement of the oil in the test jar. The complete operation
of removal and replacement shall require not more than 3 seconds. If the oil
has not ~eased to flow when its temperature has reached 50 F., the test jar shall
be placed in the jacket in a second bath maintained at a temperature of 0 to
+5 F. If the oil has not ceased to flow when its temperature has reached
20 F., the test jar shall be placed in the jacket in a third bath maintained at a
temperature of -30 to -25 F. For determinations of very low pour points,
additional baths should be maintained with successively lower temperature
differentials of about 30 F. In each case the test jar shall be transferred when
the temperature of the oil reaches a point 50 F. above the temperature of the
new bath. At, no time shall the cold test jar be placed directly in the cooling
medium. As soon as the oil in the test jar does not flow when the jar is tilted,
the test jar shall be held in a horizontal position for exactly 5 seconds, as noted
by a stop watch or other accurate timing device and observed carefully. If
the oil shows any movement under these conditions, the test jar shall be immediatelyreplaced in the jacket and a test for flow repeated at the next temperature 5 F. lower.
The test shall be continued in this manner until a point is reached at which
the oil in the test jar shows no movement when the test jar is held in a horizontal
position for exactly five seconds. Certain lubricating oils tend to move as a
whole and should b.e very closely observed. The reading of the test thermometer at this temperature, corrected for error if necessary, shall be recorded.
The pour point shall be taken as the temperature 5 F. above this solid point.
Special Procedure for Black Oils and Cylinder Stocks.-(a) In those cases
where it is known that a sample has been subjected to some temperature higher
than 115 F. during the preceding 24 hqurs, Or where the history of the sample in
this respect is not known, the sample shall be held in the laboratory 24 hours
before testing, unless three consecutive tests of the same sample in the same
test jar give check results. For these particular oils, results obtained by one or
the other of the alternative procedures shall be called the upper (maximum)
pour point.

1732

FIXED OILS, FATS AND WAXES

(b) The lower (minimum) pour point .shall be determined by heating a

sample with stirring to 220 F. The oil shall then be poured in the test jar.
cooled to 90 F. as before, and the pour point det~rmined as described in
Section on "Procedure for Pour Point."
I.
(c) The upper and lower pour points shall be repO:rted separately.

STANDARD METHOD OF TEST FOR FLASH AND FIRE POINTS BY

MEANS OF OPEN CUP 27
'

A. S. T. M.

DESIGNATION: D

92-33 28

Scope.-The open cup flash and fire test on all products except fuel oils and
those having an open cup flash below 175 F. shall be determined in the
Cleveland Open Cup.
Apparatus.-The Cleveland open cup, a diagram of which appears in Fig.
247, shall include the following major parts:
(a) Flash Cup.-The cup shall be made of Brass and shall conform to 't~
following requirem.ents:
Inches

Centimeters

Min.

Inside diameter immediately below filling mark 232

'
Outside diameter below flange .............. 2.ll.
32
Inside height from center of bottom to rim ... 1-/z
7
Thickness of bottom ....................... TI
Distance from rim to filling mark ............. i1
Distance lower surface flange to bottom of cup Ii.
Vertical distance upper surface flange to rim. it
Thickness of rim .......................... .JL
9
Width of lower surface of flange ............. 16

..

Normal Max.

Min.

Normal

Max.

-----------2Y:; 2ll
6.27 6.35 6.43
32
2ll
22.3.
6.75 6.83 6.91
1.
32
3.25 3.33 3.41
It. III
32
9
0.28 0.32 0.36
Ys
TI
.ll 0.91
0.95 0.99
%

l>i 1-/z 3.10 3.18 3.26

0.28 0.32 0.36
Ys
TI
3
_1_
6. 0.20 0.24 0.28
% 1.43 1.51 1.59
3.2
,

The bevelled edge of the cup shall be at an angle of approximately 45.

There may be a fillet of approximately -h in. (0.397 em.) radius inside the
bottom of the cup.
.
(b) Heating Plate.-The cup shall be supported bra Ul~tal plate 3i in.
(0.635 em.) in thickness and 6 in. (15.24 em.) in width. The Plate shall be of
brass, cast iron, wrought iron or stee1. .In the center of the plate there shall
be a plane depression 3\ in. (0.079 em.) in depth, and of just sufficient diameter
to fit the cup. There shall be a circular opening 2136 in. (5.50 em.) in diameter,
27 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
28 This method is issued under the fixed designation D 92; .the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1921; Adopted in Amended Form, 1924; Revised, 1933.
This method has been approved as AMERICAN STANDABD. by the American Standards
Association.

FIXED OILS, FATS AND WAXES

1733

cut through the plate, centering with the center of the above-mentioned depression. The plate shall be covered with a sheet of hard asbestos board ~ in.
in thickness, and of the same shape as the metal plate. There shall be cut in the

.....

~"

Filling Mark

~
+1
450 ~~

hr~--~--~~~~--~-

--'-___;--!.......::..:.-=---"-~ ~~ I =t~
"RI~
t.!."'>l +1
~~

"

5p2 Rad.

Approx. 1

~2/06(:!:~n~I

!~

l-

- _ _ _-1-

FIG.

247.~Cleveland

Open Cup.

center of the asbestos board a circular hole just fitting the cup. Heat may be
supplied from any convenient source. The use. of a gas burner, electric heater,
or alcohol lamp is permitted, but under no circumstances are products of combustion or free flame allowed to come up around the cup. The source of heat
shall be centered under the opening i.n the plate and shall be of a type that will
not produce local superheating. If a flame heater is used, it may be protected
from drafts or excessive radiation by any suitable type of shield, that does not
project above the level of the upper surface of the asbestos board.
Thermometer.-The thermometer shall conform to the following requirements. These specifications cover a special thermometer graduated in either
Centigrade or Fahrenheit degrees as specified, the ranges being - 60 to +400 0 C.,
or +20 0 to +760 F., respectively.
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGE AND SUBDIVISION: _6 to +400 C. in 2 C. or +20 to +760 F. in 5 F.
TOTAL LENGTH: 303 to 307 mm. (11.92 to 12.08 in.).
STEM: Plain front, enamel back, suitable thermometer tubing.
Diameter, 6.0 to 7.0 mm. (0.24 to 0.28 in.).
BULB: Corning normal or equally suitable thermometric glass.
Length, not over 13 mm. (0.51 in.).
Diameter, not greater than stem.

FIXED OILS, FATS AND WAXES

1734

DISTANCE TO .. _6 0 C. OR 20 0 F. LINE FROM BOTTOM OF BULB: 40 to 50 mm. (1.57 to

1.97 in.).
,
DISTANCE TO 400 0 C. OR 760 0 F. LINE FROM Top OF THERMOMETER: 30 to 45 mm.
(1.18 to 1.77 in.).
FILLING ABOVE MERCD:aY: Nitrogen gas.
Top FINISH: Red glass ring.
I
GRADUATION: All lines, figures, and letters clear cut and dis~inct. The 100 C. lines
0
or the first and each succeeding 10 F. line to be longer than the intermediate lines.
Graduations to be numbered at each multiple of 100 C. or 20 0 F.
IMMERSION: 25 mm. or 1 in. The words "25 mm. immersion" on Centigrade ther<-.
mometers or "I-in. immersion" on Fahrenheit thermometers and a line around
the stem 25.4 mm. or 1.00 in. above the bottom of the bulb shall be etched on the
thermometer.

I
I

?t 1(""-

~I
,.;t
~

m&

L1
Plale

I:

FIG. 248.-Heatipg Plat\l.

SPECIAL MARKING: "A. S. T. M. Open Flash" a serial number and the manufacturer's
name or trade mark shall be etched on the stem. .
-SCALE .ERROR: The error at any point of the scale when the thermometer is stand0
ardized as provided below shall not exceed PC. up to 372 C. or ~Y20 F. up to
700 F.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 100 F. for 25-mm. or I-in. immersion and for the
following temperatures of the emergent mercury column:
(

Thermometer Reading

100'0 C.
150 0 C~
2000 C.
250 C.
3000 C.
350 C.

Average Temperature of Emergent

Mercury Column

20U o F .............................. " 44 0 C.

30000 F .............................. " 54~ C.
400 F.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 64 C. I
500 0 F . .' ............................ ;. ~_77 C.'
600 0 F ... '" ........ " .......... _. ..-".:-: 91e.
700 0 F ............................ " ~ .1080 C.

1100 F.
129 0 F.
150 0 F.
175 F.
205 0 F.
240 0 F.

FIXED OILS, FATS ANP WAXES

TEST

FOR PERMANENCY OF

1735

RANGE: After being subjected to a temperature between

360 to 370.0 C. or 680 to 700 F. for 24 hours, the accuracy shall be within the

limit specified.
The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T. M. Open Flash, _6 to +490 C." or "A. S. T. M. Open
Flash, +20 to +760 F.," according to the type of thermometer.
NOTE.-For the purpose of interpreting these specificatiOIis the following definitions apply:
The total length is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning
of the enamel backing.
The top of the thermometer is the top of the finished instrument.' ,
CASE:

Procedure.-(a) The thermometer shall be suspended or held in a vertical

positi9n by any suitable device. The bottom of the bulb shall be U in.
(0.635 em.) 29 from the bottom of the cup, and above a point half way between
the center and back of the cup.
(b) The cup shall be filled with the oil to be tested in such a manner that the
top.of the meniscus is exactly at the filling line at room temperature. When
asphalt or other solid bituminous material is to be tested, it shall first be heated
to a temperature not less than 300 F. nor more than 350 F., to render it
sufficiently fluid. The cup shall then be filled with the hot fluid at this temperature in the same manner as with oiL The subsequent procedure shall then
be the same for both asphalt and solid bituminous material as with oiL The
surface of the oil shall be free from bubbles. There shall be no oil above the
filling line or on the outside of the apparatus.
(c) The test flame shall be a~proximately l2 in. (0.397 em.) in diameter.
NOTE.-For purposes of comparison it is recommended that a bead of suitable
light colored material be mounted in a convenient position so that the size of the test
flame can be determined. The device for applying the flame may be of any suitable
type but it is suggested that the tip be approximately is in. (0.159 cm.) in diameter at
the end and that the orifice be -h in. (0.079 cm.) in diameter. If the device for operating the test flame be mounted in such a manner as to permit automatic duplication
of the sweep of the test flame the radius of swing shall be not less than 6 in.
(d) The test flame shall be applied as the temperature read on the thermometer reaches each successive 5 F. mark. The flame shall pass in a straight
line (or on the circumference of a circle having a radius of at least 6 in.) across
the center of the cup and at right angles to the diameter passing through the
thermometer. The test flame shall, while passing across the surface of the oil,
be in the plane of the upper edge of the cup. The time for the passage of the
test flame across the cup shall be approximately one second.
(e) The oil shall be heated at a rate not exceeding 30 F. per minute temperature rise, till a point is reached approximately 100 F. below the probable
flash point of the oil. Thereafter the rate of heating shall be decreased and
for at least the last 50 F. before the flash point is reached, the rate shall be
not less than 9 nor more than 11 F. pe~ minute.
The flash point shall be taken as the temperature read on the thermometer
when a flash appears at any 'point on the surface of the oil. The true flasli
must not be confused with a bluish halO that sometimes surrounds the test flame.
29 The immersion line engraved on the thermometer stem will be is in. (0.159 cm.)
below the level of the rim of the cup.

1736

FIXED OILS, FATS AND WAXES

After determining the flash point, the heating shall be continued at the
specified rate of 9 to 11F. per minute, and applicatiop of the test flame shall
be made at the specified intervals until the oil ignites and continues to burn for
a period of at least 5 seconds. The method of application of the flame shall
be the same as for flash point. The temperature read ~t the time of the flame
application which causes burning for a period of 5 se'conds or more shall be
I '
recorded as the fire point.
Precautions.-The flash point and fire point tests shall be made in a room or
compartment free from air drafts. The operator shall avoid breathing over
the surface of the oil. It is desirable that the room or compartment be darkened
sufficiently so that the flash may be readily discernible.
Detection of Soap.-To increase the viscosity of an oil,30 resort is had to the
use of " oil pulp," " oil-thickener," or " white gelatin," usually an ol~ate of
aluminum, though other bases may be present. Its disadvantages are that it
causes the oil to chill more easily and to emulsify, thus increasing the friction.
Furthermore, it is precipitated by contact with water or steam, causing clogging
of the machinery.
.;
The test depends upon the fact that the meta phosphates of the earthy and
alkali metals and aluminum are insoluble in absolute alcoho1. 31
The test is applied as follows: five to 10 illl. of the oil to be tested are dissolved in about 5 ml. of 86 gasoline or ether, and about 15 drops of the phosphoric acid solution (Reagents, p. 1799) add~d, shaken and allowed to stand;
the formation of a flocculent precipitate indicates the presence of soap. An
idea of the kind of soap can be often gained by adding an alcoholic solution of
PtCI 4 If the precipitate becomes crystalline, it is a potash soap; if it dissolves,
soda, lime, or magnesia; if unchanged, alumina or iron.
For the accurate determination of these compounds a known weight of the
oil must be ignited, the residue determined and quantitatively examined.
Caoutchouc.-Holde 32 states that 1 to 2% of unvulcanized caoutchouc is
sometimes added to oils to increase their viscosity. This may be detected by
adding three parts of alcohol to four parts of the etherlf~l solution, whereby the
rubber material is precipitated and may be dried and weighed.
Test for Fatty Oils.-To detect small quantities of fatty oil (U to 2%)
Lux 33 recommends heating a few ml. of the oil Jor fifteen minutes with some
bits of sodium in a test-tube in an oil bath; a similar test is made with sodium
hydroxide. The temperature employed should be for light oils about 230,
for dark oils 250.34 In case fatty oil be present, the contents "of one or both
of the tubes solidify to a jelly of greater or less consistence according to the
amount of fatty oil present.
The quantitative determination of these oils, as for example in cylinder oils,
is effected after the manner of determining the saponification value (p. 1771)
or the detection of unsaponifiable oils in fatty oils (page 1773).
30 In a case which came to the writer's notice the oil would not flow out of the Say.
bolt "A" apparatus at 70, at 85 required 1167", and at 110 181".
31 Schweitzer and Lungwitz, J. Soc. Chern. Ind., 13, 1178, 1894.
32 "Examination of Hydrocarbon Oils," p. 202.
I
33 Z. anal. Chern., 24, 357, 1885.
:.. 34 Holde, Examinat~on of Hydrocarbon Oils, p. 76. - -' _
0

\
FIXED OILS, FATS AND WAXES

1737

Schreiber 35 adopts a method similar to Sweetham and Henriques, in that he

dissolves 5 grams of the oil in 25 ml. of benzol, adds 25-50 ml. N/2 alcoholic
potash, and boils for half an hour on the ~ater bath, using a 3-ft. glass tube as a
condenser.
.
Gumming Test. 36-This is designed to give an idea of the amount of a
change that may -be expected in a mineral oil when in use. These resinified
products increase the friction of the revolving or rubbing surfaces.37 It il> also a
measure of the amount that an oil will " carbonize" in a gas or gasoline engine
cylinder. It is applied after the manner of the elaidin test, by thoroughly
mixing together 5 grams of the oil in a cordial glass with 11 grams of nitrosulfuric acid and cooling by immersion in a pan of water at 10-15. Brownish
spots or, in case of a bad' oil, masses, form around the edges and become red in
the course of two hours.
As shown by long practical experience, the oil showing the least tar or gum
is the best oil; it also absorbs the least oxygen .

STANDARD METHOD OF TEST FOR CARBON RESIDUE OF

PETROLEUM PRODUCTS 38
(CONRADSON CARBON RESIDUE)
A. S. T. M. DESIGNATION: D 189-36 39
. Scope.-This method of test is a means of determining the amount of
carbon residue left on evaporating an oil under specified conditions, and is
intended to throw some light on the relative carbon-forming propensity of an
oil. The results of the test must 'be considered in connection with other tests
and the use for which the oil is intended. This test furnishes pertinent information relative to lubricants for internal combustion engines, domestic oil
fuels, and oils used in the manufacture of gas.
Apparatus.-The apparatus (see Fig. 249) shall consist of the following:
(a) Porcelain crucible, wide form, glazed throughout, or a silica crucible;
29 to 31-ml. capacity, 46 to 49 mm. (1.81 to 1.93 in.) in rim diameter.
(b) Skidmore iron crucible flanged and ringed, 65 to 82-ml. capacity, 53 to
57 mm. (2.07 to 2.20 in.) inside and 60 to 67 mm. (2.36 to 2.64 in.) outside
J. Am, Chern. Soc., 29, 74, 1907.
Gill, J. Am. Chern. Soc., 24, 467, 1902.
37 Aisinman, J. Soc. Chern. Ind., 14, 282, 1895.
38 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
This procedure is a modification of the original Conradson method and apparatus
for Carbon Test and A~h Residue in Petroleum Lubricating Oils. See Proceedings;
Eighth International Congress of Applied Chemistry, New York, September, 1912,
Vol. I, p. 131; also Journal of Industrial and Engineering Chemistry, Vol. 4, No. 11,
December, 1912.
39 This method is issued under the fixed designation D 189; the final number indicates the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1924; Adopted in Amended Form, 1928; Revised, 1930,
1936.
35
36

1738

FIXED OILS, FATS AND WAXES

diameter of flange, 37 to 39 mm. (1.46 to 1.54 in.) in height supplied with a

cover without delivery tubes and having the 'vertical {opening closed. The
horizontal opening of about 6.5 mm. (0.26 in.) shall be kept clean. The outside
diameter of the flat bottom shall be 30 to 32 mm. (1.18 to 1.26 in.).
(c) Spun sheet-iron crucible, with cover; 78 to 82 thm. (3.07 to 3.23 in.)
in outside diameter at the top, 58 to 60 mm. (approximately-2.3 in.) in height,
and approximately 0.8 mm. (0.03 in.) in thickness. Place at the bottom of this
crucible, and level before each test, a layer of about 25 m!. of dry sand, or.
enough to bring the Skidmore crucible, with cover on, nearly to the top of the
sheet-iron crucible.
(d) Triangle of bare nichrome wire of approximately No. 13 B. & S. gage
having an opening small enough to support the bottom of the sheet-iron
crucible at the same level as the bottom
of the asbestos block or hollow sheetmetal box, Paragraph (f).
(e) Circular i?heet-iron hood from 120
to 130 mm. (4%' to 5U in.) in diameter ~
the height of the lower perpendicular
side to be .from 50 to 53 mm. (2 to 2Ys
in.); provided at the top with a chimney
50 to 60 mm. (2 to 2,72 in.) in height
and from 50 to 56 mm. (2 to 2U in.)
inside diameter, which is attached to the
lower part having the perpendicular sides
by a cone-shaped member, bringing the
total height of the complete hood from
125 to 130 mm. (5Ys in.). The hood
may be made from a single piece of metal
provided the foregoing dimensions are
adhered to. As a guide for the height
of tbe flame above:tbe cbimney, a bridge
made of appr6ximately 3-mm. (Ys-in.) iron
FIG. 249.-Apparatus for Determining or nichrome wire, h, shall be attached
Carbon Residue.
having a height of 50 mm. (2 in.) above
the top of the chimney.
(f) Asbestos, block refractory ring, or hollow sheet-metal pox, 150 to 175
mm. (6 to 7 in.) in diameter if round or on a side if square, 32 to 38-mm. (lU to
1,72 in.) in thickness, provided with a metal-lined, inverted cone-shar-ed opening
through the center; 83 mm. (3U in.) in diameter at the bottom, and 89 mm.
(3,72 in.) in diameter at the top.
(g) Burner, Meker type, 24 mm. (1 in.) in diameter by 155 mm. (6 in.) in
height, suitable for either manufactured or natural gas.
Procedure.-The test shall be conducted as follows:
Place two glass beads about 0.1 in. in diameter in, and include, them in the
weight of, the tared porcelain or silica crucible in which is accurately weighed
10 g. of the r~l to be tested, free- from moisture or other suspended matter.
Place this crucible in the centJr of the Skidmore crucible. Level the sand in
the large sheet-iron crucible and set the Skidmore ;;"rucible on it in the exact

FIXED OILS, FATS AND WAXES

1739

center of the iron crucible. Apply covers to both the Skidmore and me iron
crucible, the one to the latter fitting loosely to allow free exit to the vapors as
formed.
On a suitable stand or ring, place the bare nichrome wire triangle and on it
the asbestos block or hollow sheet-metal box. Next, centre the sheet-iron
crucible in the block with its bottom r.esting on top of the triangle, and cover
the whole with the sheet-iron hood in order to distribute the heat uniformly
during the process (see Fig. 249).
.
Apply heat with a high strong flame from the Meker-type gas burner, so
that the pre-ignition period will be 10 minutes IYz minutes (a shorter time
may start the distillation so rapidly as to cause foaming or too high a flame).
When smoke appears above the chimney, immediately move or tilt the burner
so that the gas flame plays on the sides of the crucible for the purpose of igniting
vapors. Then remove the heat temporarily, and before replacing adjust by
screwing down the pinch-cock on the gas tubing so that the ignited vapors
burn uniformly with the flame above the chimney but not above the wire
bridge. Heat may be increased, if necessary, when the flame does not show
above the chimney. The period of burning the vapors shall be 13 minutes
1 minute.
When the vapors cease to burn and no further blue smoke can be observed,
readjust the burner and hold the heat as at the beginning so as to make the
bottom and lower part of the sheet-iron crucible a cherry red and maintain for
exactly 7 minutes. The total period of heating shall be 30 minutesi=2 minutes
which constitutes an additional limitation on the tolerances for the pre-ignition
and burning periods. There should be no difficulty in carrying out the test
exactly as directed with the gas burner of the type named, using city gas
(about 550 B.t.u.) with the top of the burner about 2 in. below the bottom of
the crucible. The time periods shall be observed with whatever burner and
gas is used.
Remove the burner and allow the apparatus to cool until no smoke appears,
and then remove the cover of the Skidmore crucible (about 15 minutes).
Remove the porcelain or silica crucible with heated tongs, place in the desiccator, cool and weigh. Calculate the percentage of carbon residue on the
original sample.
.
Special Procedure: SPecial Procedure for Oils Having Carbon Residues in
Excess of 5%, Such as Heavy Crude Oils, Residuums, Heavy Fuel Oils, and
Heavy Gas-Enrichment Oils.-When the carbon residue as obtained by the
procedure described above (using a 10-g. sample) is in excess of 5%, diffi-
cui ties may be experienced due to boiling over of the sample. Trouble also
may be encountered with samples of heavy products which are difficult to dehydrate. In the case, therefore; of samples showing by the usual method of
test a carbon residue in excess of 5%, the test shall be repeated in duplicate
using a 5-g. sample.

Tolerances.-Weights of oil sample shall be accurate to within 5 mg. Tests

shall be run in duplicate and repeated if necessary until the percentages of
carbon residue differ by not more than 10% from an average.
Gasoline Test.-This shows the presence of tar (still bottoms) or asphaltic'
matters.

FIXED OILS, FATS AND WAXES

1740

Mix 10 ml. of the oil with 90 ml. 86-88 gasoline (from Pennsylvania crude)'
B. pt. 30-50 C., allow to stand one hour' at 70 to 8Qo F.; not more than 10%
of flocculent or tarry matter should have settled out. This settling can be
facilitated by whirling in a centrifuge in a. graduated ,ube in which the volume
of the precipitate can be read Gff. If the test b~ applie~ to the oil before making
the flash test and then again after this test it shows t~e extent to which the oil
is changed upon heating. Other things being equal the oil which is changed
the least is the best oil. See page 1761.
Microscopical Test.-Put.a few drops of the well-mixed oil on a slide f!nd
note the nature of the suspended matter-whether carbonaceous specks, flakes
of paraffin which disappear on warming, or foreign matter. A good oil should
be practically free from all these bodies.
Heat Test.-Heat 20-25 ml. of the oil in an Erlenmeyer flask or beaker,
nearly to the flashing point and keep it at this temperature for fifteen minutes;
a satisfactory oil will darken, but remain clear even after standing a day. A
poorly refined oil changes to jet black, and forms a carbon-like precipitate.
This is usually indicative of an oil that has been refined by acid treatment ...
Herschel's Demulsibility Test. 40 Method 3Z0.31.-The test is applicable
to non-emulsifying oils, such as are used with high speed engines and turbines,
and in its present form does not apply to stt;am cylinder or other compounded
_
oils. It is made as follows:
Twenty m!. of oil and 40 ml. of distilled water are placed in a 100 m!.
cylinder having an inside diameter of 2.&-cm. and heated to 55. The oil and
water are' t~en mixed or emulsified by stirring with a paddle for five minutes at a
speed of 1500 R.P.M. The paddle is simply a metal p~ate 89 by 20 by 1.5 mm.
thick, conveniently driven by an electric motor.
The rate at which the oil settles out from the emulsion, when it is allowed
to stand at a temperature of 55 C. is a measure of the emulsibility of the oil.
There is a tendency for the rate of settling out of the oil to increase up to a
maximum and then decrease, and we may take readings every five minutes
(or every minute if the oil separates out rapidly) and ,note the maximum rate.
For example, the following readings were obtained for a certain oil:

Time

Time Since
Stopping Paddle

Reading of Cylinder
at Top of Emulsion

MI. of Oil
Settled Out

9.55 .......
10.02 .......
10.05 .......
10.10 .......

0
5
12
15
20

60
57
47
43
41

0
3
13
19
17

I
9.50 .......

Rll te of Settling
MI. per Hour

0
36
65
68
57

The maximum rate is here 68 ml. per hour, and:this is called the" demulsibility." The best transformer, motor, and turbine oils settle out in one
minute or less, so that if we take readings to even minutes and even ml., the
rate would be 1200. Oils are found with rates all the way: down from this
l
maximum rate to zero.
40

Bur. Mines Bull. 323B.

FIXED OILS, FATS AND WAXES

1741

It has been found necessary to shake the container.holding the sample of

oil if it has been allowed to stand for some time, or otherwise the oil poured off
from the top may show a different rate of settling from the oil at the bottom.
It is also necessary to protect ..the oil from sunlight which has the effect of
decreasing the rate of settling.
With some oils, under certain conditions, part of the water does not enter
into the emulsion and drops of .oil adhere to the sid.es of the cylinder. When
this phenomenon occurs, the rate of settling, indicated by the test, is apt to be
incorrectly high. The adhering drops may be a voided by using a lower speed
or less water, the total quantity of oil and water being kept equal to 60 ml.
If part of the water does not enter into the emulsion, but we do not have the
above described phenomenon, complete emulsification may sometime!! be
obtained by prolonging the stirring beyond five minutes, or by increasing the
speed.
Friction Tests.-The writer doubts if friction tests are worth the outlay
for a machine and the time expended in their execution. Without question
they do determine the relative efficiency as regards lubricating power of differe'nt oils, but the conditions under which the test is made seldom occur in
practice; the bearings upon which the oil is tested are as nearly perfect as can
be made, and the feed and load are as regular as is possible; in other words, the
conditions are ideal.
The lubricating power of an oil is so closely related to its viscosity 41 that the
author believes that results of more practical value can be obtained by the
determination of the viscosity of the oils, and subsequent observation of theif
behavior in actual use than by the longer and more troublesome friction test.
Recent experiments,42 however, have shown that of two oils of the same viscosity and other constants the coefficient of friction of one was 14% less than
the other.
In case, however, it be desired to make the friction test, the following
machines, it is believed, will be found to be most satisfactory for the purpose:
For spindle oils and light lubricating oils, the machine 43 of the Thurston
type which can be run at the highest speed and lowest pressure.
For heavy oils and railroad work, the large machine of the Thurston 44 type,
described also in Redwood Treatise on Petroleum, III, 851 (1922); also in
Archbutt and Deeley.46
Color.-This is determined with the Union Colorimeter which uses standard
glass plates, A. S. T. M. D 155-34 T.
Brannt, "Petroleum and its Products," p. 510; Woodbury, vide infra.
Trans. Am. Soc. Mech. Eng., 32, 834, 1910.
4~ Made by Olsen or Riehle Bros., Philadelphia, Pa.
44 "Lubrication and Lubricants," 1907, pp. 332-348.
4. Ib!d., pp. 355, 359.
41

42

1742

FIXED OILS, FATS AND W;AXES

LIQUID FUELS

(a) Gasoline. The tests usually made are coIOr,46! doctor test,46 corrosion,
distillation, acidity, and sulfur.
The color, acidity, and sulfur tests are made as with kerosene.
The doctor test (sodium plumbite) is made as follows: 47
Doctor Test (Sodium Plumbite)
~ Preparation of Reagents: (1) Sodium plum bite (doctor solution).-Dissolve
approximately 125 grams of sodium hydroxide (NaOH) in 1 liter of distilled
water. Add 60 grams of litharge (PbO) and shake vigorously for 15 minutes
or let stand with occasional shaking for at least a day. Allow to settle and
decant or siphon off the clear liquid. Filtration through a mat of asbestos
may be employed if the solution does not settle clear. The solution should be
kept in a tightly corked bottle and should be refiltered before use if not perfectly
clear.
(2) Sulfur.-Pure, dry flowers of sulfur.
Making of Test.-Shake vigorously tog~ther in a test tube 10 m!. of the
sample to be tested and 5 ml. of sodium plumbite solution for about 15 seconds.
Add a small pinch of flowers of sulfur, again shake for 15 seconds, and allow to
settle. The quantity of sulfur used shonld be such that practically all of it
floats on the interface between the sample and the sodium-plumbite solution.
Interpretation of Results.-If the sample is discolored, or if the yellow color
of the sulfur film is noticeably masked, the test shall be reported as positive and
the sample condemned as "sour." If the sample remains unchanged in cQlor,
and if the sulfur film is bright yellow or only slightly discolored with gray or
flecked with black, the test shall be reported negative and the sample considered
" sweet."

Corrosion Test

(CopperIDish~

Apparatus.-A freshly polished hemispherical dish of spun copper apapproximately 3Yz inches in diameter.
Making of Test.-Place 100 ml. of the gasoline to be examined in the dish
and place the dish in an opening of an actively boiling steam bath, so that the
steam comes in contact with the outer surface of the dish up to the level of the
gasoline. Leave the dish on the steam bath until all volatile. material has
disappeared.
Interpretation of the Results.-If the gasoline contains dissolved elementary
sulfur or corrosive sulfur compounds, the bottom of the dish will be. colored
gray or black.
(
If the gasoline contains undesirable gum-formihg constituents, there will
be a weighable amount of gum deposited on the dish. Acid residues will show
as gum in this test.
46 Egloff says these tests will not long be required 017 a~count,lof expense and also
that in the refining valuable anti-knock compounds areJemoved,
41 Tech. Paper 323.B.
. --

FIXED OILS, FATS AND WAXES

1743

STANDARD METHOD OF TEST FOR DETECTION OF FREE

SULFUR AND CORROSIVE SULFUR COMPOUNDS I:N
GASOLINE 48
A. S: T. M.

DESIGNATION: D

130-30 49

This method of test shall be used for the detection of free sulfur and corrosive
sulfur compounds in gasoline.
Procedure.-A clean strip of mechanically polished pure sheet copper about
Yz in. in width and 3 in. in length shall be placed in a suitable clean tube or.
sample bottle. Gasoline under test shall be added so that the copper strip is
completely immersed. The test tube or sample bottle shall be closed with a'
loosely fitting cork and held in a suitable bath at 122 0 F. (50 0 C.).
At the end of three hours the gasoline exposed strip shall be removed and
shall be compared with a similar strip of freshly polished copper.
Detection of Sulfur.-The presence of sulfur or corrosive sulfur compounds is
indicated by the corrosion or discoloration of the gasoline e.xposed strip when
compared with the fresh copper strip.
'
(a) Gasoline shall be reported as passing the test when on examination'
the exposed strip shows not more than extremely slight discoloration as compared with the fresh copper strip.
(b) Gasoline shall be reported as 'not passing the test when on examination
the exposed strip shows more than extremely slight discoloration as compared
with the fresh copper strip.
Distillation of Gasoline
Apparatus and Procedure.-The apparatus and procedure have been given
in method D 86-35, p. 1705:
Significance of Test.- R. E. Wilson 50 says, the 10% point is the best
indication of ease of starting, vapor lock (that is stopping of engine by too
volatile gasoline), and evaporation loss. The 50% point is important for
good acceleration. The 90% point is related to the dew point and controls
crank case dilution.
Octane Number is the per cent by volume of iso octane in a mixture of
iso octane and normal heptane required to match the antiknock value of a given
fuel.
48 Under the standardization procedure of the Society, this method is Under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
49 This method is issued under the fixed designation D 130; the final number indicates the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1922; Adopted in Amended Form,. 1930. This methoci
has been approved as AMERICAN STANDARD by the American Standards Association.
50 Chem. Abs. 27, 184.

1744

FIXED OILS, FATS AND WAXES

FUEL OILS

The tests usually made are flash, viscosity,. sulfur, I~ater and sediment.
STANDARD METHOD OF TEST FOR FLASH POINT BY MEANS OF
THE PENSKY-MARTENS CLOSED TESTER 61
A. S. T. M.

DESIGNATION: D

93-36

52

The A. S. T. M. standard Pensky-Martens closed tester shall be used for

determining the flash point of fuel oil unless the use of the Tag closed tester is
specified. In addition, a modified form of this me~hod as described subsequently is used for the determination of the flash point of cut-back asphalts
and other viscous materials and suspensions of solids.
Apparatus.-The A. S. T. M. Pensky-Martens tester, a diagram of which
appears ih Fig. 250, shall include the following major parts:
(a) Cup.-The cup shall be made of brass and shall satisfy the fQ.llowing
dimensional specifications:
C:;entimeters

Inches

Dimensions
Mini-

mum

Inside diameter below filling mark .........

Difference, inside and outside diameters below filling mark .......................
Inside height ............................
Thickness of bottom .....................
Distance from rim to filling mark ..........
Distan:ce lower surface flange to bottom of
cup ..................................

Normal

Maxi-

mum

Minimum

Normal

- - - - - -- -

Maximum

1.950 2.000 2.050 4.953 5.080 5.207

0.120
2.150
0.070
0.845

0.125
2.200
0.095
0.860,,

0.130
2.250
0.120
0.875

0.305
5.461
0.178
2.146

0.318
5.588
0.241
2.184

0.330
5.715
0.305
2.223

1.780 1.795 1.810 4.521 4.559 4.597

The insi~e of the cup may be turned to a slightly larger diameter above the
filling mark and the outside may be tapered above t~ flange but the wall
thickness at the upper edge shall be not less than 0.04 in~ '(0,10_~ cm.). The
flange should be approximately 0.5 in. (1.27 cm.) wide and approximately
0.125 in. (0.318 cm.) thick. It. shall be equipped with devices tor locating the
position of the lid on the cup and the cup in ,the stove. A handle, attached
permanently to the flange of the cup, is a desirable accessory.
51 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petr6leum Products .and Lubricants.
Reproduced through. the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
62 This method is issued under the fixed designation D 93; the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as Tentative, 1921; Adopted in Amen.!led Form, 1922; Revised, 1936.
This method has been approved as AME1RICAN STANn'ARD. by -the American Standards
Association.

FIXED OILS, FATS AND WAXES

1745

(b) Lid.-

(1) Stirring Devit;e.-The lid shall be equipped with a stirring device

consisting of a vertical steel shaft, not less than 0.1 in. (0.254 em.) nor more
than 0.125 in. (0.318 em.) in diameter, mounted in the center of the cup, and
carrying two two-bladed brass propellers. The blades of both propellers shall
be approximately 0.313 in. (0.795 em.) wide and shall be set at an angle of

FIG. 25Q,-A. S. T. M. Standard Pensky-Martens Tester.

approximately 45. The smaller (upper) propeller shall have an over-all

diameter of approximately 0.75 in. (1.905 cm.). The larger (lower) propeller
shall have an over-all diameter between 1.25 and 1.75 in. (3.175 and 4.445 cm.).
The thickness of the propeller blades shall be not less than 0.057 in. (0.145 em.)
nor more than 0.081 in. (0.206 em.), whic~ limits correspond respectively to

1746

FIXED OILS, FATS AND WAXES

I'

No. 15 and No. 12 B. and S. gage sheet brass. The collars on which the propeller blades are mounted shall have horizontal and vertical dimensions not
greater than 0.4 in. (1.016 cm.).
1
'
The plane of the center of the upper, propeller shall be 0.4 in. (1.016 cm.)
below the level of the rim of the cup. The plane of t~e center of the Iowel\.propeller shall be 2.0 in. (5.08 cm.) below the level of the rim of the cup. The level
of the rim of the cup is in effect the level of the plane~ part of the portion of the
lower surface of the lid inside the rim.
(2) Cover proper.-The cover proper shall be of brass and shall have a rim
projecting downward almost to the flange of the cup and fitting the outside of
the cup closely. The thickness of the cover, measured just inside the rim ,shall
be not less than 0.031 in. (0.079 cm.) nor more than 0.078 in. (0.198 cm.).
There shall be a proper locating device engaging with a corresponding locating
device on the flange of the cup.
There shall be four ope?ings in the cover, as indicated in Fig. 251.

MInimum o.938m., MaJ:/mumo.969;i?

Q531 ".
"
0.563"
0500 ..._
"
0.540 ..
n
..
0.187 ", . "--" 0 219 "
s Approxlmatety 0.75 tn,
'u
"
0.5 ". -An8/es p Equal
Angle r - Mm 135~ Ma)(. 140~
I

"

"

"

; -

"

"

y.-

It

50

D
;

IO~

"

60

15'

FIG. 251.-Cover for A. S. T. M. Standard Pensky-Martens Tester.

Opening A is an area defined by arcs of two concentric circles and the intersected lengths of two radii. The radius of the outer circle shall be not less than
0.938 in. (2.383- cm.) nor more than 0.969 in. (~4.n cm.),. The radius of the
inner circle shall be not less than 0.531 in. (1.349 em.) nor more than 0.563 in.

FIXED OILS, FATS AND WAXES

1747

(1.430 cm.). The chord of the arc of the outer circle shall be not less than
0.500 in. (1.270 cm.) nor more than 0.540 in. (1.372 cm.).
Openings Band C are equal areas, each of the same general form as opening
A but of approximately half the (angular) width. The radii of the defining
inner and outer circles shall be within the limits specified for the radii of the
two circles, arcs of which partly define opening A. The chord of the outer arc
for opening B or opening C shall be not less than 0.187 in. (0.475 cm.) nor more
than 0.219 in. (0.556 cm.). The sum of the areas of openings Band C shall
be not less than 75% nor more than 100% of the area of opening A. Openings
Band C shall be equally distant from opening A and radii drawn through each
of their centers shall be at an angle of not less than 1350 nor more than 140.
Openings A, B, and C need not conform exactly to the shape of geometrical
figures bounded by arcs of two concentric circles and intersected lengths of
radii. Their boundaries must, however, fall on or between the lines indicated
by the limiting values of the dimensional specification of the preceding text
and of Fig. 251.
Opening D is for a thermometer collar. Its center is approximately 0.75 in.
(1.905 cm.) from the center of the lid and on a radius at an angle of not less
than 50 0 nor more than 60 from a radius passing through the center of opening
C. The thermometer collar shall have an inside diameter of approximately
0.5 in. (1.27 cm.). It shall be set at an angle of not less than 10 nor more than
15 from the perpendicular.
(3) Shutier.-The lid shall be equipped with a brass shutter, approximately
0.094 in. (0.239 cm.) thick operating on the plane of the upper surface of the lid.
The shutter shall be so shaped aI).d mounted that it rotates on the axis of the
horizontal center of the lid between two stops so placed that when in one extreme
position the openings A, B, and C of the lid are completely closed and when in
the. other extreme position these orifices are completely opened.
(4) Flame exposure device.-The flame exposure device shall have a tip
with an opening 0.027 in. (0.069 em.) to 0.031 in. (0.079 cm.) in diameter.
The flame exposure device shall be equipped with an operating mechanism
which, when the shutter is in the " open" position, depresses the tip so that
the center of the orifice is between the planes of the under and upper surfaces
of the lid proper at a point on a radius passing through the center of the larger
opening ~ and apprl1ximately 0.1 in. (0.254 cm.) from the outer edge of the
opening.
NOTE.-A pilot flame for automatic relighting of the exposure flame should be
provided.
'
A bead 0.156 in. (0.397 cm.) in diameter, of some suitable material, may be mounted
on the lid so that the size of the test flame can be regulated by comparison.
The mechanism operating the shutter should be of the spring type and constructed
so that when at rest the shutter shall exactly close the three openings. When operated
to the other extreme the three openings in the lid shall be' exactly open and the tip
of the exposure tube shall be fully depressed.

(c) Siove.-Heat shall be supplied to the cup by means of a properly

designed stove which is equivalent to an air bath. This stove shall consist of
(1) an air bath and (2) a top plate on which the flange of the cup rests. '
(1) Air bath.-The air bath shall have a cylindrical interior 1.625 in.
(4.128 cm.) to 1.656 in. (4:206 cm.) deep and a diameter not less than 0.125 in.

1748

FIXED OILS, FATS AND WAXES

(0.317 em.) nor more than 0.156 in. (O.396.cm.) greater than the outside diameter of the cup. The air bath may be either a flame heated metal casting\or
an electric resistance element.

c~ting

NOTE.-If the heating element is a flame-heated metal

it shall be so designed
and used that the temperature of bottom and walls is approximately the sarrie. On
this account it should be not less than 0.25 in. (0.635 em.) thick. The casting shall be
designed so that products of combustion of the flame cannot pass up and in. contact
with the cup.
If the air bath is of the electric resistance type it shall be constructed so that all
parts of the interior surface are heated equally. This necessitates an even distribution
of'resistance wire over bottom and walls and a method of construction such that heat
is given out from. the whole core of the resistance element rather than directly from
the wire.

(2) Top plate.-The top plate shall be of metal. The total distance from
the upper surface of the plate to the bottom of the air bath shall exceed the
distance from the under surface of the flange to the bottom of the cup by not
less than 0.063 in. (0.160 em.) nor more than 0.125 in. (0.317 em.).
The top plate shall be mounted with an air gap between it and the air bath.
The top plate may be attached to the air bath by means of three screws and
spacing bushings. The spacing bushings should be of proper thickness to
define the air gap which shall be not less than 0.125 in. (0.317 cm.) nor more
than 0.187 in. (0.475 em.). The spacing bushings shall be not more than
0.375 in. (0.952 cm.) in diameter.
/
(d) Thermometers.-Two standard thermometers shall be used with the
A. S. T. M. Pensky-Martens tester. The low range, " A. S. T. M. P.M. and
Tag" thermometer shall be used for tests when the indicated reading falls
within the limits 200 to 2000 F. The" A. S. T. M. P.M. high" thermometer
shall be used for tests when the indicated reading falls within the limits 230 0 to
7000 F. For the range 200 0 to 2300 F. either thermometer may be employed,
dependihg on the convenience of the operator. The thermometers shall comply
with the requirements given in Tables I and II. The ,Low Range" A. S. T. M.
P.M. and Tag" t}:lermometer specifications cove., a special thermometer
graduated in either Centigrade or Fahrenheit degrees as specified, the ranges
being _7 to +1100 C. or +20 0 to +230 0 F., respectively. The high range
"A. S. T. M. P.M. high" thermometer specifications cover a special thermometer graduated in ei~her Centigrade or Fahrenheit degrees as specified,
the ranges being 900 to 3700 C. or 2000 to 700 0 F., respecth:ely.
Thermometers shall be mounted so that the bottom of the"bulb is 1.75 in.
(4.445 cm.) below the level of the rim of the cup (which corresponds to the level
of the lower surface of the portion of the lid inside the rim).
TABLE L-8PECIFICATIONS FOR Low RANGE "A. S. T. M. P.M. AND TAG"
THERMOMETER 63
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGE AND SUBDIVISION: _7 to +110 C. in ,0.5 C. or +20 to +230 F. in 1 F.
TOTAL LENGTH: 273 to 277 mm. (10.75 to. 10.92 in.).
63 The low range" A. S. T. M. P.M. and Tag" therm()meter ii?'the same instrument
as that specified for use with the Tag closed tester, see:::the Standard Method of Test
for Flash Point by Means of Tag Closed Tester (A. 8.""1';-1\1. Designation: D 56) of the
American Society for Testing Materials, see 1936 Book of A. S. T. M. Standards, Part II.

1749

FIXED OILS, FATS AND WAXES

STEM: Plain front, enamel back, suitable thermometer tubing. Diameter, 6.0 to 7.0
mm. (0.24 to 0.28 in.).
BULB: Corning normal or equally suitable thermometric glass.
Length, 9 to 13 mm. (0.35 to 0.51 in.).
Diameter, not greater than stem.
DISTANCE TO _7 0 C. OR +20 0 F. LINE FROM BOTTOM OF BULB: 75 to 90 mm. (2.93
to 3.54 in.).
DISTANCE TO 110 C. OR 230 F. LINE FROM Top OF THERMOMETER: 25 to 40 mm.
(0.98 to 1.57 in.).
EXPANSION CHAMBER: To permit heating the thermometer at least 50 C. (90 F.)
above highest temperature on scale.
FILLING ABOVE,MERCURY: Nitrogen gas.
Top FINISH: Glass ring.
GRADUATION: All lines, figures, and letters clear cut and distinct. The whole degree
Centigrade lines or the first and each succeeding. 5 F. line to be longer than the
remaining lines. Graduations to be numbered at each multiple of 5 C. or 10 F.
IMMERSION: 57 mm. or 2.)4 in. The words "57-mm. immersion" on Centigrade or
"2.)4-in. immersion" on Fahrenheit thermometers and a line around the stem
57.0 mm. or 2.25 in. above the bottom of the bulb shall be etched on the thermometer.
SPECIAL MARKING: "A. S. T. M. P.M. and Tag.," a serial number and the manu. facturer's name or trade mark, shall be etched on the stem.
SCALE ERROR: The error at any point of the scale when the thermometer is standardized as provided below shall not exceed 0.5 C. or 1 F., respectively.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 30 C. or 50 F. for 57-mm. or 2.)4-in. immersion and
for the following temperatures of the emergent mercury colUmn:
Thermometer Reading

20C.
40C.
70 C.
100 C.

Average Temperature of Emergent

Mercury Column

70 F ............................ " .. 20~ C.

100 F ......... , ...................... 31C.
150 F ................................ 40 C.
212.F, ............... " ............... 48 C.

70 F.
- 86 F.
104F.
118 F.

CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T .. M. P.M. and Tag., _7 to + 110 C." or "A. S. T. M.
P.M. and Tag., +20 to +230 F." according to the type of thermometer.
NOTE.-For the purpose of interpreting these specifications the following definitions apply:
The totalleIigth is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom of the bulb to the beginning
of the enamel backing.
The top of the thermometer is the top of the finished instrument.
TABLE 1I.-SPECIFICaTIONS FOR HIGH RANGE "A. S. T. M. P.M. HIGH"
THERMOMETER
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGE AND SUBDIVISION: 90 to 370 C. in 2 C. or 200 to 700 F. in 5 F.
TOTaL LENGTH: 273 to 277 mm. (10.75 to 10.92 in.).
STEM: Plain front, enamel back, sui~ble thermometer tubing. Diameter, 6.0 to 7.0
mm. (0.24 to 0.28 in.).
BULB: Corning normal or equally suitable. thermometric glass.
Length, not over 10 mm. (0.39 in.).
Diameter, not greater than stem.
DISTANCE TO 93 C. OR 200 F. LINE FROM BOTTOM OF BULB: 75 to 90 mm. (2.93 to
3.54 in.).
.
DISTANCE TO 371C. OR 700 F. LINE FROM Top OF THERMOMEll'ER: 25 to 40 mm.
(0.98 to 1.57 in.).

FIXED OILS, FATS AND WAXES

1750

FILLING ABOVE MERCURY: Nitrogen gas.

Top FINISH: Glass ring.
GRADUATION: All lines, figures, and letters clear cut and distinct. The first and each
succeeding 10 0 C. or 25 0 F. line to be longer than the remaining lines. Graduations
to be numbered at multiples of 20 0 C. or 50 0 F.
I .
IMMERSION: 57 mm. or 274 in. The words :!.57-mm. immersion" ~m Centigrade thermometers, or "274-in. immersion" on Fahrenheit therdlOmeters and a line around
the stem 57.0 mm. or 2.25 in. above the bottom of the ]:lUlb shall be etched on the
thermometer.
_ SPECIAL MARKING: "A. S. T. M. P.M. High," a serial number and the manufacturer's
name or trade mark shall be etched on the stem.
SCALE ERROR: The error at any point of the scale when the thermometer is standardized as described below shall not exceed 1 0 C. or 2Yz F.
STANDARDIZATION: The thermometer shall be standardized at intervals of approximately 50 0 C. or 100 0 F. 'for 57-mm. or 274-in. immersion and for the following
temperatures of the emergent mercury column:
Thermometer Reading

1000
150 0
200 0
250 0
3000
3500

Average Temperature of Emergent

Mercury Column

C. 200 F ................................ 61 0 C.
C. 3000 F ................................ 65 0 C.
C. 400 0 F ................................. 71 0 C.
C. 500 0 F ............ " ................ , 78 0 C.
C. 600 0 F ....... , ........ " ....... '.' ...... 87 0 C.
C. 7000 F ...................../........... 990 C.

1400 F.
149 0 F.
1600 F.
1750 F. 1950 F.
2200 F.

TEST FOR PERMANENCY OF RANGE: After being subjected to a temperature between

360 0 and 370 0 C. or 680 0 and 700 0 F. for 24 hours, the accuracy shall be within
the limit specified. The test shall be made under the immersion conditions specified for this thermometer.,
CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T. M. P.M. High, 90 0 to 3700 C." or "A. S. T. M. P.M. High,
200 0 to 700 0 F." according to the type of thermometer.
N OTE.-For the purpose of interpreting these specifications the following definitions apply:
'
The total length is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom/of the bulb to the beginning
,
of the enamel backing.
The top of the thermometer is the top of the finished instrument.

,
Procedure.-(a) All parts of the cup and its accessories shall be thoroughly
clean an1 dry before the test is started. Partic;ular care shall be taken to avoid
the presence of any gasoline or naphtha used to clean 1J!.e apparatus after a
previous t e s t . '
. -.__.
(b) The cup shall be filled with the oil to be tested up to the level indicated
by the filling mark.
(c) The lid shall be placed on the cup and the latter set in the stove. Care
shall be taken to have the locating devices properly engaged. The thermometer
shall be inserted. If it is known that too oil will flash above 220 0 F. the
" A. S. T. M. P,M. High" thermometer may be sel~cted; otherwise, it is preferable to start with the" A. S. T. M. P.M. and Tag" thermometer and change in
case a temperature of 220 to 2.30 0 F. is reached.
(d) The test flame shall be lighted and adjusted so that it is of the size of a
. '
bead -.A in. (3.97 mm.) in diameter.
(e) Heat shall be supplied at such a rate that'thetemperature read on the

FIXED OILS, FATS AND WAXES

1751

thermometer increases not less than 9 nor more than 11F. per minute. The
stirrer shall be turned at a rate of from 1 to 2 revolutions per second.
(f) Application of the test flame shall be made at each temperature reading
which is a multiple of 2 F. up to 220 F. For the temperature range above
220 F., application 'shall be made at each temperature reading which is a
multiple of 5 F. The first application of the test flame shall be made at a
temperature at least 30 F. below the actual flash point. Application of the
test flame shall be made by operating the device controlling the shutter and
test flame burner so that the flame is lowered in one-half seco:q.d, left in its
lowered position for one second, and quickly raised to its high position. Stirring/
shall be discontinued during the application of the test flame.
Flash Point.-The flash point is taken as the temperature read on the
thermometer at the time of the flame application that causes a distinct flash in
the interior of the cup. The true flash must not be confused with the bluish
halo that sometimes surrounds the test flame for the applications preceding the
one that causes the actual flash.
Barometric Pressure.-The barometric pressure shall be observed and recorded. No corrections shall be made except in case of dispute when the flash
point figure shall be correct"ed according to the following rule:
For each inch (25 mm.) below 29.92 in. (760 mm.) barometric reading add
1.6 F. to the flash point.
For each inch (25 mm.) above 29.92 in. (760 mm.) barometric reading
subtract 1.6 F. from the flash point.
Determination of Flash Point of Cut-Back Asphalts and Other Viscous
Materials and Suspensions of Solids.-The apparatus shall consist of the
following:
(a) Pensky-Martens tester as described above except that the stirrer shall
be mechanically operated to stir in a down{l,ard direction at a speed of 70 to
80 r.p.m ..
(b) Low-range thermometer as specified in Table 1.
Procedure.-The material to be tested and the tester shall be brought to a
temperature 30 F.lower than the estimated flash point. The air space between
the cup and the interior ofthe air bath shall be completely filled with water at
the temperature of the tester and material. The temperature shall be raised
throughout the duration of the test at a rate of not less than 4 nor more than
6 F. per min. With the exception of this requirement for rate of heating and
the rate of stirring of 70 to 80 r.p.m., specified in Section 6 (a); the procedure
shall be the same as that prescribed in the Standard MethOd of Test for Flash
Point by Means of the Tag Closed Tester (A. S. T. M. Designation: D 56) of
the American Society for Testing Materials. 54
Accuracy.-Duplicate tests should not differ by more than 5 F.
64

1936 Book of A. S. T. M. Standards,. Part II.

1752

~AXES

FIXED OILS, FATS AND

STANDARD METHOD OF TEST FOR FLASH POINT BY MEANS

OF THE TAG CLOSED TESTER 55
A.,S. T. M.

DESIGNATION: D 5~3656

de~ermining

Scope.-The Tag closed tester shall be used for

the flash point
of all mobile liquids flashing below 175 F., with the exception of products
classed as fuel oil which are preferably tested with the Pensky-Martens Closed
Tester as described in the Standard Method of Test for Flash Point by Means
of the Pensky-Martens Closed Tester CA. S. T. M. Designation: D 93) of the
Arperican Society for Testing MaterialsY Provision is made in later Sections
of this method for determining the flash point of lacquer solvents or dilvents
of low flash points.
Tag Closed Tester, a diagram of which appears in Fig. 252, shall conform
to the following dimensional requirements within the permissible variations
shown:
Inches

Centimeters

Dimensions

Millil11um

Depth of water surface below top of cup .... li.Depth of oil surface below top of CUp ....... 1';Depth of top of bulb of oil thermometer when
in place below top of CUp ............... 1""
Inside diameter of oil cup at top ........... 2l.hDiameter of bead on top of cover ..........
-II
Weight of oil cup ........................ -

Normal

Maximum

Minimum

Normal

Maximum

la".

II:.

1';III
64

2.74
2.90

2.78
2.94

2.82
2.98

If.
2Ys

III
32
2l'o"li

3.25
5.39
0.36

3.33 3.41
5040 5.41
0040 0.44
67 to 6~ g.

- -- -- - - -- -

32

it

"

The plane of underside of cover shall be between the to.p and bottom of the
burner tip when the tip is fully depressed.
'I
.
Thermometer.-Two thermometers ar,e required, one for measuring the
temperature of the sample under test, the other for measuring the temperature
of the bath. Each of these thermometers shall conform to the following requirements. These specifications cover a special thermometer graduated in
either Centigrade or Fahrenheit degrees as specified, the ranges being _7
to +110 C. or +20 0 to +230 F., respectively:
TYPE: Etched stem, glass.
LIQUID: Mercury.
RANGE AND SUBDIVISION: _7 0 to +110 0 C. in 0.5 0 C. or +20 0 to +230 0 F. in 1 0 F.
TOTAL LENGTH: 273 to 277 IDID. (10.75 to 10.92_in.).,
66 Under the standardization procedure of the Society, this method is under the
joint jurisdiction of A. S. T. M. Committee D-I on Paint, Varnish, Lacquer, and Reo lated Products, and Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials, 260 S. Broad
St., Philadelphia, Pa.
66 This method is issued under the fixed designation D 56; the final number indicates
the year of original adoption as standard or, in the case of revision, the year of last
revision. Issued as,Tentative, 1918; Adopted in Amended:Form, '1919; Revised, 1921,
1936.
. .-F67 1936 Book of A. S. T. M. Standards, Part II, p. 897.

FIXED OILS, FATS AND WAXES

Fw. 252.-A. S. T. M .. Tag Closed Tester.

(Arranged for Use of Gas.)

1753

FIXED OILS, FATS AND WAXES

1754

STEM: Plain front, enamel back, suitable thermometer tubing. Diameter, 6.0 to 7.0
mm. (0.24 to 0.28 in.).BULB: Corning normal or equally suitable thermometric g~ss.
Length, 9 to 13 mm. (0.35 to 0.51 in.).
Diameter, not greater than stem.
\
DISTANCE TO _7 C. OR +20 F. LINE FROM BOT.TOM OF BULB: 75 to 90 mm. (2.93
to 3.54 in.).
.
"
DISTANCE TO 110 C. OR 230 F. LINE FROM Top OF THERMOMETER: 25 to 40 mm.
.
(0.98 to 1.57 in.).
EXPANSION CHAMBER: To permit heating the thermometer at least 50 C. (90 F.)
above highest temperature on scale.
FILLING ABOVE MERCURY: Nitrogen gas.
Top FINISH: Glass ring.
GRADUATION: All lines, figures, and letters clear cut and distinct. The whole degree
Centigrade lines or the first and each succeeding 5 F. line to be longer than the
remaining lines. Graduations to be numbered at each multiple of 5 C. or 10 F.
IMMERSION: 57 mm. or 23-4 in. The words "57-mm. immersion" on Centigrade or
"27i-in. immersion" on Fahrenheit thermometers and a line around the stem
57.0 mm. or 2.25 in. above the bottom of the bul~ shall be etched on the thermometer.
SPECIAL MARKING: 58 "A. S. T. M. P.M. and Tag.," a serial number and the manufacturer's name or trade mark shall be etched on the stem.
SCALE ERROR: The error at any point of the scale when the thermometer is standardized as provided below shall not exceed 0.5 C. or 1 F., respectively.
STANDARDIZATION: The thermometer shall be standardized at the ice point and at
intervals of approximately 30 C. or 50 F. for 57-mm. or 23-4-in. immersion and
for the following temperatures of the emergent mercury column:
/

Thermometer Reading

20 C.
40 C.
70 C.
100 C.

Average T~mperature of Emergent

.
Mercury Column

70 F ....... _.. " ... " ...... " ...... " ..

100 F. _..... " " " " " " ...... " .. " . ".......
150 F. " .. " .. " ..... " ... " " .. " ..... " ....
212 F. _.......... " . . . . . . . . . . . . . . . . . ..

20 C.
31C.
40 C.
48 C.

70 F.
86F.
104 F.
118 F.

CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T. M. P.M. and Tag., _7 to +110 C." or "A. S. T. M.
P.M. and Tag., +20 to +230 F." according to the type of thermometer.
NOTE.-For the purpose of interpreting these specifications the following definitions apply:
I
/
The total length is the over-all length of the finillhed instrument.
The diameter is that measured with a ring gage.
The length of the bulb is the distance from the bottom of ~he bulb to the beginning
of the enamel backing.
The top of the thermometer is the top of the 'finished instrument.

Procedure.-(a) The test shall be performed in a ro~m or_compartment

darkened sufficiently to permit ready detection of the flash.
(b) Care shall be taken to have the tester level and steady. It shall be
surrounded on three sides by an enclosure for protection from drafts. (A
shield 18 in. square and 24 in. in height, open'in front, is suggested. Tests
made in a laboratory hood or near ventilators are n,ot to be relied upon.)
(c) Gas may be used for the test flame and for heating the water bath.
If gas is not available for the test flame, a wick of cotton cord may be inserted
58 This thermometer is identical with the low range instrument specified for use with
the A. S. T. M. Pensky-Martens Closed Tester and described in the Standard Method
of Test for Flash Point by Means of the Pensky-Martens_ Closerl,,"Tester (A. S. T. M.
Designation: D 93) of the American Society for TestiDg-Materials, see 1936 Book of
-.
A. S. T. M. Standards, Part II, p. 897.

FIXED OILS, FATS AND- WAXES

1755

in the burner tip, a small quantity of cotton waste placed in the oil chamber to
which the burner tip is attached and the chamber filled with signal, sperm, or
lard oil. An alcohol lamp may be used for heating the water bath as a substitute for gas.
The water bath thermometer shall be place4 in the collar provided for it
and the bath filled. with water until it overflows. The temperature of the
water in the bath shall be such that when testing is started it will be at least
20 0 F. (11 0 C.) below the probable flash point of the oil to be tested.
The oil cup shall be placed in its proper position in the water bath and 50 ml.
of the oil to be tested shall be measured into it, using an accurate graduate or
other measuring device for the purpose. The temperature of the oil shall be '
at least 20 F. (11 C.) below its probable flash point when the t~st is started.
Air bubbles on the surface of the oil shall be destroyed, and the cover with the
flash point thermometer in place shall then be properly attached to the bath
collar. The test flame shall be, lighted, the flame being adjusted to the size
'
of the small white bead on the cover.
'Che gas burner or alcohol lamp shall be centrally placed in the base of the
tester and lighted. The flame shall be so adjusted that the temperature of the
oil in the cup rises at the rate of 1.8 F. (1C.) per min. as closely as possible,
but in any case not faster than 2 F. (1.1 0 C.) nor slower than 1.6 F. (0.9 C.)
per min.
(a) The barometric pressure shall be recorded. If a barometer is not
available, the figure may be obtained from the nearest Weather Bureau Station
and an appropriate correction made -for difference in altitude between such
station and the laboratory.
(b) The initial temperature o(the oil shall be recorded.
(c) When the temperature of the oil is 9 F. (5 C.) below its probable Hash
point, the knob on the cover shall be turned in such a manner as to introduce
the test Hame into the vapor space of the cup, and immediately turned back
again. The time consumed in turning the knob down and back shall be about
one full second, or the time required to pronounce distinctly the words" thousand and one."
Cd) The time at which the first introduction of the test flame is made and
the temperature of the oil shall be recorded.
(e) The application of the test flame shall be repeated after each 1 F.
(0.5 C.) rise in temperature of the oil until a distinct flash in the interior of the
cup is observed. The true flash must not be confused with the bluish halo
which sometimes surrounds the test Hame during applications immediately
preceding the actual flash.
(f) The time and the temperature of the oil when the flash point' is reached
shall be recorded.
Repeat Tests.-(a) If the rise in temperature of the oil from the time of
making the first introduction of the test flame to the time at which the flash
point is observed was more rapid than 2 F. (1.1 C.), or slower than 1.6 F.
(0.9 C.) per min., the test shall be repeated, adjusting the gas burner or alcohol
lamp to the proper Fate of heating.
(b) It is not necessary to turn off the test flame with the small regulating
valve on the cover; it may be left adjusted to the proper size of flame ..
(c) After completing the preliminary test to determine the approximate

1756

FIXED OILS, FATS AND WAXES

flash point, the burner or lamp shall be removed, the oil cover lifted and the
thermometer bulb carefully wiped off. The oil cup sha:ll be removed, emptied,
and carefully wiped until dry.
(d) The temperature of the bath shall be lowered py the addition of cold
water until it is 15 F. (8 C.) below the flash point of, the oil as shown by the
preliminary test.
(e) The oil cup shall be replaced and a fresh 50-m!. sample measured into it.
The test procedure, as described above, shall then be repeated, introducing
the test flame for the first time, however, when the oil temperature is 10 F.
(5.5 C.) below the flash point obtained in the preliminary test.
(j) Oil which has once been subjected to the flash test shall be discarded.
(g) If test is to be repeated, a fresh sample shall be used.
Average Value of Flash Point.-If two or more determinations agree within
1 F. (0.5 C.), the average of these results, corrected for barometric pressure:
shall be considered th~ flash point. If two determinations do not check within
1 F. (0.5 C.), a third determination shall be made and, if the maximum variation of the three tests is not greater than 2 F. (1 C.), their average, after correcting for barometric pressure, shall be consid\lred the flash point.
Correction for Barometric Pressure.-Correction for barometric pressure
shall be made only in cases of dispute or when the barometer reading varies
more than 7'2 in. (13 mm.) from the standar.d pressure of 29.92 in. (760 mm.).
When the barometer reading is' below this standard pressure, add to the thermometer reading 1.6 F. (0.9 C.) ,for eacn1 in. (25 mm.) of barometer difference to obtain the true flash point. When the barometer reading is above the
standard pressure, deduct 1.6 F. (0.9 C.) for each. 1 in. (25 mm.) of barometer
difference to obtain the true flash point.
'
Determination of Flash Point of Lacquer Solvents or Diluents of
Low Flash Point.
Apparatus.-The apparatus shall consist of the foI1qwing:
(a) Flash Point Tester.-As described under ," Tag' Closed Tester."
(b) Thermometer.-For determining flash points below 40 F. (4.4 C.),
suitable thermometers of low range shall be used. These thermometers shall
conform to the requirements as given above except that the ranges shall be 0 F.
to 120 F. or -20 C. to +50 C. These thermometers may not be used in
the testing of materials flashing above 70 F. (21.1 C.).
Procedure.-The procedure to be followed shall be the same ,as that described previously, except that in determining flash points below 4{t F. (4.4 C.)
the bath shall be filled with brine or other low-freezing liquid instead of water.
Viscosity is determined with the Saybolt Furol viscosimeter, p. 1719.
Sulfur 59 '(A. S. T. M. Method D 129-34)
,

This method shall be used for the determination of sulfur in petroleum oils
which cannot be burned completely in a wick lamp.
Apparatus.-The oxygen bomb shall have a capacity of not less than 300
m!. and shall be of a design or construction such that JlQleaks.~hall occur at any
59 Tech. Paper 323B.
Reproduced through the courtesy of the American Society
for Testing Materials, 260 S. Broad St., Philadelphia, Pa.

1757

FIXED OILS, FATS AND WAXES

pressure or temperature generated during the test, ~nd such that when the
bomb is open the liquid contents can be easily and completely drained. The
inner surfaces shall be of materials that are chemically and physically resistant
to the process or products of combustion. The gaskets, insulating materials,
etc., shall be, as far as possible, physically and chemically resistant and in no
eveJ;lt shall they undergo any reaction which would increase or decrease the
sulfur content of the bomb liquors.
The oil cup shall be of platinum, glazed silica, or other suitable material, with
a capacity of not less than 2.5 ml. nor more than 5 ml.
If a platinum oil cup is used, the fuse wire shall be of platinum; if a glazed
silica oil cup is used, the fuse wire may be of either platinum or iron. No. 35
(B. and S. gage) is a convenient size.
(a) The distilled water and all reagents should be sulfur free, but when it is
necessary to employ reagents not sulfur free, blanks shall be run and the figures
thus obtained used to correct the results of actual determinations.
(b) The barium chloride solution shall contain 100 grams of BaCb2H 20
per liter.
Procedure.-(a) Twenty ml. of distilled water shall be placed in the bottom
.of the bomb. From 0.6 to 0.8 gram of the oil ~o be tested shall be placed in
the weighed oil cup and the weight of this charge shall be determined to an
accl,lracy of at least 0.002 gram. The cup shall be placed in the proper position in the bomb, the ignition mechanism arranged and the bomb closed.
Oxygen shall be admitted slowly until a pressure is reached as indicated by the
following table:
Capacity of
Bomb,m!.

Minimum Gage
Pressure, Atmospheres

to 350 ..................... , ....................... ,

to 400 ............................................. ,
to 450. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
to 500 ..................... , ............. , ......... ,
Above 500. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..

300
350
400
450

40
35
30
27.5
25

'The leads from the firing circuit shall be attached, the bomb placed in a
bucket of cold water, and ignited. The bomb shall be allowed to stand in the
water for 10 minutes and shall then be removed. The valve of the bomb shall
be opened, allowing the gas to escape at an approximately even rate so that the
pressure is reduced to atmospheric in not less than 1 minute. The bomb shall
be opened, and all parts of its interior, including the oil cup, rinsed with a fine
jet of distilled water. All washings, which should not amount to more than
350 ml., shall be collected in a beaker. Particular care should be taken not to
lose, oy splashing or otherwise, any of the liquid contents of the bomb. The
washings shall be filtered through a washed" qualitative" filter paper. The
filter shall be washed thoroughly. Two ml. of concentrated HCI and 10 ml.
of saturated bromine water shall be added to the filtrate. The solution shall
be evaporated to about 75 ml. on a steam bath or hot plate. Ten ml. of hot
barium chloride solution shall be added in a fine stream or dropwise to the hot
solution, stirring during the addition and for 2 minutes afterwards. The
solution shall be allowed to stand over night, or shall be kept hot for 1 hour on
the steam bath or hot plate, allowing the precipitate to settle for another hour
while cooling. The supernatant liquid shall be filtered through an " ashless

FIXED OILS, FATS ANJ? WAXES

1758

quantitative" filter paper and the precipitate washyd with water, first ,by
decantation, then on the filter, till free from chloride. ' The paper and precipitate shall be transferred to a suitable weighed crucible, dried at low heat, till
moisture is evaporated, the paper charred (wit4out flaIhing), and finally ignited
at a good red heat till the precipitate is just burned whi,t~. A satisfactory means
of accomplishing these operations is to place the crucil?le containing the wet
filter paper in a cold electric muffle furnace and to turn on the current. Drying,
charring, and ignition will usually occur at the desired rate. .
(b) After ignition is complete, the crucible shall be allowed to cool to room
temperature and weighed. The use of a desiccator is not recommended.
. (c) From the increase in weight of the crucible the percentage of sulfur
shall be calculated from the formula:
.

_ grams of BaS04 X 13.734

P ercent age 0 f suIf urf'l
d
.
grams 0 01 use

Accuracy.-The percentages of sulfur obtained by the same operator with

the same apparatus should not differ by more than 0.02+0.04 A; while the
percentages obtained by different operators should not differ by more' than
0.02+0.08 A (A represents the average of percentages obtained).

The Jacksqn Turbidimeter can be use,d for rapid work.

Water and Sediment are determined as follows:

"

Water (A. S. T. M. Method D95-30)

60

This method of test determines the water in a sample of bituminous material

by distilling the sample with a volatile solvent. This method is suitable for a
variety of materials but is especially applicable to petroleum, fuel oil, road oil,
coal tar, water-gas tar, coke-oven tar, and other petroleum products or bituminous materials.
Apparatus.-The apparatus shall consist of a metal still or glass flask, heated
by suitable means and provided with a reflux, cond/nser discharging into a trap
connected to the still or flask. The trap ~erves to collect and measure <the
condensed water and to return the solvent to the still.
The type of distilling apparatus used is not an essential feature of this
method, but glass has been generally used for petroleum products and the 'metal
still for road materials and tars.
~
3. (a) The metal still (Fig. 253, a) shall be a vertical cylindrical vessel,
preferably of copper, having a faced flange at the top to which the head is
tightly attached by means of a clamp. The head shall be of metal, preferably
of brass or copper, and be provided with a tubulation 1 inch in inside diameter.
(b) The glass flask (Fig. 253, b) shall be of the short-neck, round-botto!ll
type, made of well-annealed glass, having an approximate cap~city of 500 m!.
The burner used with the metal still shall' be a ring gas burner 4 inches
(100 mm.) in inside diameter. With the glass flask an ordinary gas burner or
electric heater may be used as the source of heat.
The condenser shall be of the water-cooled, reflux, glass tube type, having
a condenser jacket not less than 400 mm. J1J!% inches) in length with an
60 Tech. Paper 323B.
Reproduced through' the-courtesy of the American Society
. for Testing Materials, 260 S. Broad St., Philadelphia, Pa.

FIXED OILS, FATS AND WAXES

1759

inner tube 12 to 16 mm. (72 to % inch) in diameter. The end of the condenser
to be inserted in the trap shall be ground off at an angle of 30 from the vertical
axis of the condenser.
The trap shall be made of well-annealed glass constructed in accordance
with Fig. 253, c, and shall be graduated as shown from 0 to 10 mI. in 0.1 mI.'
divisions. The error of any indicated capacity shall not be greater than 0.05 mI.
The outside dIameters should preferably be 2.5 to 3.5 mm. (332 to Ys inch)
greater than the inside diameters specified.
(a) The solvent used when testing petroleum products or bituminous
materials derived from petroleum shall be gasoline free from water and shall

A_ ......._... ,.

l'

Ii"'" 21.'

45 foSSmm.
8-(In6{d, Dlam,hr; 14 ., 16m",. C-(llI6ldeJJlam,fll')12 " 16mm. _ 1.1.."

if:""

D--_.- ISO "170mm._ 6

, ..

flo

E---- .........--.25" 38mm._ I"" 1-

F- _-_-100 "'08mm. _ 4" .,

H---~---- .... --. If ,. 19 mm.. _ ~""

4J'

.I-

FIG. 253.-Apparatus for Determination of Water.

conform to the following distillation requirements, determined in accordance

with the" Tentative method of test for distillation of gasoline, naphtha, kerosene, and similar petroleum products"; A. S. T. M.-D 86. Five percent shan
distil at a temperature not below 1940 F. (90 0 C.) nor, above 212 0 F. (1000 C.);
90% shall distil below 410 0 F. (210 0 C.).
(b) The solvent used when testing bituminous materials derived from coal
tar, water-gas tar, etc., shall be a coal-tar naphtha or a light oil and shall
conform to the following distillation requirements, determined in accordance
with the It, Tentative method of test for distillation of gasoline, naphtha, kerosene, and similar petroleum products"; A. S. T. M.-D 86, 98% shall distil
between 248 F. (120 C.) and 482 0 F. (250 C.).

1760

FIXED OILS, FATS AND WAXES

Sample.-The sample shall be thoroughly representative of the material to

be tested and the portion of the sample used for theitest shall be thoroughly
representative of the sample itself. Deviation from .this requirement shall not
be permitted.
I
NOTE.-The difficulties in obtaining proper representative samples for this determination are unusually great, so that the importance of sampling can'not be too stron3ly
emphasized.

Piocedure.-When the sample to be tested contains less than 10% of water,

exactly 100 ml. of the material to be tested shall be placed into the still or
flask and thoroughly mixed with an equal volume of solvent by swirling, proper
care being taken to avoid any loss of material. If the material is measured by
volume, an accurate 100-ml. graduated cylinder shall be used and the contents
transferred to the still by rinsing with one 50-ml. portion of solvent followed
. by two successive 25-ml. portions of solvent, the cylinder being allowed to
drain each time. When the sample to be tested contains more than 10% of
water, the volume of material used shall be decreased to that which will yield
somewhat less than 10 ml. of water.
NOTE.-In special cases where the water content exceeds 10% and it is not
desirable to reduce the size of the sample to that which will yield somewhat less than
10 m!. of water, a distilling tube receiver gradUated from 0 to 25 mI. may be used.
This tube shall be graduated from 0 to 2 mI. in 0.1 mI. from 2 to 5 mI. in 0.2 mI., and
from 5 to 25 m!. in 0.5 mI.
\

The connections between the still or flask, trap, and condenser shall be
made by means of tight-fitting corks as shown in Fig. 253 (a and b). The end
of the condenser inserted in the trap shall be adjusted to that position which will
allow the end to be submerged to a depth of not more than 1 mm. below the
surface of the liquid in the trap after distillation conditions have been established. When the metal still is used, a heavy paper gasket moistened with
the solvent shall be inserted between the lid and flange before attaching the
clamp. A loose cotton plug shall be inserted in the top of the condenser tube
to prevent condensation of atmospheric moisture in the condenser tube.
Heat shall then be applied and so regulated that the condensed distillate
falls from the end of the condenser at the rate of from 2 to 5 drops per second.
The ring burner used with the metal still should be placed about 3 inches above
the bottom of the still at the beginning of the distillation and gradually lowered
as the distillation proceeds.
The distillation shall be continued at the specified rate ulitil no water is
visible on any part of the apparatus except at the bottom of the trap. This
operation usually requires less than an hour. A persistent ring of condensed
water in the condenser tube shall be removed by increasing the rate of distillation for a few minutes.
The volume of condensed water measured in the trap at room temperature
multiplied by 100 and divided by the volume of the sample used shall be the
percentage of water and shall be reported as "-per cent water by volume,

A. S. T. M.- method."
Accuracy.-The accuracy to be expected with this method is that duplicate
determinations of water should not differ from ~6li other by more than one
. division on the trap.

FIXED OILS, FATS AND WAXES

1761

Sediment
Apparatus.-Alundum (porous grade) thimbles,s1 1 inch in diameter by
weighing not less than 15 nor more than 17 grams.
Extraction apparatus (see Fig. 254) of such construction that the thimble is
completely surrounded by the vapor of the boiling
solvent. Siphon extractors must not be used.
Procedure.-Place approximately 10 grams of
the sample in the previously extracted and dried,
accurately weighed thimble and weigh to 0.01
gram, place in the extraction apparatus, and extract
with 90% benzol until the solvent dropping from
the thimble is colorless. Dry the thimble for 1
hour at 105 0 C., and weigh to O.OOOI gram. Repeat the extraction until the weight of the dried
thimbl.e and sediment is constant.
NOTE-The rate of extraction shall be such that the
mixture of oil and benzol in the thimble does not rise to
within i inch of the top.

2U inches high,

Water and Sediment by Centrifuge (A. S. T. M.

Method D 96-35) 62
This method may be used for crude mineral oils
and fuel oils. A centrifuge method for" Water.
FIG. 254.-Extraction Apand Sediment" is not entirely s~tisfactory because paratus for Determination
the amount of water obtained is nearly always lower of Sediment: a, Extraction
than the actual water content. Nevertheless, on flask; b, condenser; c, extraction thimble; d, solvent;
account of the wide use of the centrifuge for this e,
top of hot plate.
purpose, it is desirable that the method of making
the determination be standardized as far as possible. It must be clearly
understood that the reading of the centrifuge tube includes both the sediment
and the precipitated water. Accurate determination of water content, 'if desired, should be made in accordance with the Standard Method of Test for
Water in petroleum products and other Bituminous Materials (A. S. T. M.
D 95).
The sample shall be thoroughly representative of the material in question
and the portion used for the ,test shall be thoroughly representative of the
sample itself. Deviation from this rule shall not be permitted. The difficulties
in obtaining representative samples for this determination are unusually great;
hence the importance of sampling can not be too strongly emphasized.
Apparatus.-The centrifuge shall be capable of whirling at the required
speed at least two 100-ml. centrifuge tubes filled with water. It shall be of
sound design and rugged construction so that it may be operated without danger.
The tube carriers shall be so designed that the glass centrifuge tubes may be
cushioned with water, rubber, or other suitable material. The tube holders
61 Alundum thimbles listed by the Norton Co., Worcester, Mass., as 5163 RA 98,
have proved satisfactory.
62 Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.

FIXED OILS, FATS AND :WAXES

1762

shall be surrounded during the operation by a suitable metal shield or case,

strong enough to eliminate danger if any breakage oqeurs.
The preferred form of centrifuge shall have a diameter of swing e(tip to tip
of whirling tubes) of 15 to 17 inches and a speed of at le,ast 1500 r.p.m. or the
equivalent. If the available centrifuge has a" dianiet~r 'of swing varying from
these limits, it shall be run at the proper speed t() give the same centrifugal
force at the tips of the tubes as that obtained with the' preferred form of centrifuge. The proper speed shall be calculated from the, following formula in
which d represents diameter of swing (tip to tip of. whirling tubes) of the
centrifuge used:
' ,
,
r.p.m. = 1500

{i6

'\J(f'

"

The centrifuge tubes, A. S. T. M. types shown in Fig. 255, shall be made

of suitable glass and thoroughly annealed. The total capacity shall be about
125 ml. and the mouth shall be suitably constricted for closing with a cork.,
The gra~uations shall be clear and distinct, reading upward from the bottom of
the tube. Air-free distilled water at 20 C. (68 0 F.) shall be used for the calibrat~on of tubes. The length and outside diameter are optional provided they
do not conflict with the other requirements, ' The value of the divisions for each
range with limits o(~rror are given in the following tabu,lations for the A. S. T. M.
Pear-Shaped Tube and A. S. T. M. Cone-Shaped Tube: .
\
/'

A.
Range

,0 to
3 to
5 to
10 to
25 to
50 to

M. PE:AR-S:S:~PE1>

Scale Division

0.1 ml.
0.5 "
1.0 "
5.0 "
25.0 "
50.0 "

3mI........ ,,"
5 " ...........
10 " .......... ,
25 " ...........
50 " . . . . . . . . . . .
100 " ...........

A.
Range

0
0.1
0.3
0.5
1
3
5
10
25

S. T.

to 0.1 mI ... , .....

to 0.3 " ........
to 0.5 "
to 1..0 " . . . . . . . .
to 3 " . . . . . . . .
to 5 "
to 10 "
to 25 " ....... ,
to 100 " ........
co,.

o'

S. T.

M.

TUBE

Limit of Error

0.05 mI.
0.2 "
0.5 "
1.0 "
1.0 "
1.0 ",J

..

Numbered

1,2,3
mi.
4,5
"
6,8,10
"
15,20,25 "
50
"
100
"

CONE-SHAPED TUBE

Scale Division

Limit of Error

0.05 m!.
0.05 "
0.05 "
0.1 "
0.1 "
0.5 "
1.0 "
5.0 ".
25.0 "

0.02 mI.
0.03 "
0.05 "
0.05 "
0.1 "
0.2 "
0.5" "
1.0: "
1.0 "

I. _,

_NUmbered

Y2

1
" 2,3
4,5
6,8,10
15,20,25
50,75,100

mI.

"
"
"
"
"
"

The water or oil bath shall be of sufficient depth for immersing the centrifuge tubes in a vertical position to t1e 100-ml. mark.- Means shall be provJa~d
for heating this bath to 120 0 F. (49 0 C.).
.-

..

FIXED OILS, FATS 4-ND WAXES

1763

ProcEldure.-(a) Exactly 50 ml. of 90% benzol shall be measured into each

of two centrifuge tubes and exactly 50 ml. of the oil to be tested shall then be
added to each, ,The centrifuge tubes shall be tightly stoppered and shall be
shaken vigorously until the contents are thoroughly mixed. The temperature
of the bath shall be maintained at 120 0 F., and the centrifuge tubes shall be
immersed therein to the 100-ml. mark for '10 minutes.
(b) The two centrifuge tubes shall then be placed in. the centrifuge on opposite sides and shall be whirled at a
rate of 1400 to 1500 r.p.m., or the
equivalent, for 10 minutes. The commi.
bined volume of water and sediment at
the bottom of each tube shall be read
-100
and recorded, estimating to 0.1 ml. if
necessary. The centrifuge tubes shall
-75
then be replaced in the centrifuge, again
whirled for 10 minutes, and removed
-so
for. reading the volume of water and
sediment as before.
3
This operation shall be repeated
until the combined volume of water and
2
sediment in each tube remains constant for three consecutive readings.
In general, not more than four whirlings
will be required.
The combined total volume of water
o
and sediment shall be read on each
tube, estimating to 0.1 ml. if necessary.
The sum of the two readings shall be (.) Pear-Shaped Tube
(b) Cone-Shaped Tube
recorded ,as percentage of water and
FIG. 255.-Centrifuge Tubes
sediment, centrifuge method.
(A. S. T. M.).
Accuracy.-With care and proper
attention to details, duplicate determinations of water and seqiment by this
method should not differ by more than 0.2 ml., provided the centrifuge tubes
are accurate and readable to this degree.
Water by Centrifuge
Apparatus.-The apparatus shall be the same as ill the preceding method.
Procedure.-Exactly 40 ml. of 90% benzol shall be measured into each of
two centrifuge tubes and exactly 20 ml. of the oil to be tested shall then be
added to each. The centrifuge tubes shall be tightly stoppered and shall be
shaken vigorously until the contents are thoroughly mixed.
The t'Wo centrifuge tubes shall then be placed in the centrifuge on opposite
sides and shall be whirled at a rate of 1400 to 1500 r.p.m., or the equivalent;
for 10 minutes. The volume of water at the bottom of each tube shall be
read and recorded, estimating to 0.1 ml. if necessary. The centrifuge tubes
shall t,hen be replaced in the centrifuge, again whirled for 10 minutes as before,
and removed for reading the volume of water as before. This operation shall

1764

FIXED OILS, FATS AND WAXES

be repeated untii the volume of water in each tube 'remains constant for three'
consecutive readings.
The volume of water shall be read on each tube, estimating to 0.1 ml. if
necessar~, and the average perce~tage of water in th~ sample c~lculated.
Calonfic Power. 53-The calorIfic power of uncracl.{ed fuel OIl'can be found
by the formula, Cal. Power=17660+(69XA.P.I. g~avity, 0 Baume). Of a
cracked fuel oil, Cal. Power = 17780+(54XA.P.I. gravity, 0 Baume).

ANIMAL AND VEGET:A.BLE OILS

The tests most commonly employed for the identification of these oils are as
follows: specific gravity, refractive index, Valenta test, elaidin test, Maumene
test, iodine number, and saponification value.
In addition, certain special and commercial tests are applied, as Bechi test,
Baudouin test, free acid, spontaneous combustion, and drying test.
Specific Gravity.-This is usually determined either by the Westphal
balance (page 1704) or by the picnometer.
_
A two-necked flask of 50 ml. capacity, having a thermometer carefully
ground into one neck, the second one being a narrow tube bearing the mark, is
most suitable. This is filled with the oil to be examined, cooled to 15.5 0 (;.,64
the excess of oil removed and weighed. If the weighings be made to 0.5 milligram and a correction applied for the expansion of the glass-1>y the difference in
temperature = 15.5-4= 11.5 0 = -0.025% of the va~ile obtained, the determination is accurate to 0.00002.65
For the determination of the specific gravity of small quantities of oil,
satisfactory results can be obtained by weighing 1 or 5 ml. of the oil carefully
measured from an accurately calibrated pipette. Or a mixture of alcohol and
water can be made until a drop of oil will stay in any position in it, and its
specific gravity determined.
... ~
Refractive Index.-This is of the same value as the determin~tioh of specific
gravity: it has, however, the advantage that it is more rapid and uses only one
or two drops of the oil.
The apparatus preferably employed is the ADM refractometer, Fig. 256, the
prisms of which are kept at constant temperature, Ui;mally 25 0 C., by cu-culating
.'
water.
The illuminating mirror should light the cross h~irs and the telescope should
be sharply focused on them.
Faragher, Morrell & Truax, Ind. Eng. Chern., 21, 933 (1929),
Allen (Organic Analysis, 33) states that a correction of 0.00064 can be made for
each variation of 1 C.
~ '.56 Wright, J. Soc. Chern. Ind., 11, 300, 1892.
63

64

FIXED OILS, FATS AND WAXES

1765

The double prism is opened by means of the screw heads, and after carefully
cleansing the prisms with cotton and ether, a drop or two of the oil placed on
the horizontal surface of the fixed prism. The prisms are then tightly closed:
The telescope is brought into the position shown and the sector is firmly held
and the alidade (the moving part) moved forward until the field of vision shows
the boundary between li'g ht and shade just intersecting the cross hairs.
By means of the screw on the right
of the instrument this boundary line
should be made as sharp as possible.
The index of refraction is read off directly
from the sector, using a lens if necessary;
the reading is accurate to .0002.
After using, the prisms are again
carefully cleansed and a piece of filter
paper placed between them to prevent them from being scratched. The
instrument is in correct adjustment when
water at 18 gives a mean reading of
1.333. The temperature correction for
oils and fats is 0.0004 for every degree
rise.
Valenta Test.--Although considered
by some to be unreliable, yet as the
indication given by this test may be of
value, it is barely worth the trouble of
execution. It depends upon the solubility of the oil in glacial acetic acid.
Enough oil is poured into a test-tube
to fill it to the depth of about 1 in.,
the exact height being marked by the
FIG. 256.-Refractorneter.
thumb; an equal quantity of glacial acetic
acid is poured in, that is, until the acid reaches the point indicated by the thumb.
A light thermometer is placed in the tube, and it is heated until the oil dissolves
-shown by the liquid becoming homogeneous. The tube is now allowed to cool,
and the point noted at which it begins to become thoroughly turbid.
Castor oil is soluble at ordinary temperatures, while rape-seed and other
cruciferous oils are usually insoluble even at the boiling-point of the acid. The
tetnperat ures at which other oils become turbid are given on pages 1795 and
1796. .
Elaidin Test.-Although this is not a quantitative test, yet its ease of application and the conclusions which may be drawn from it render it valuable.
It depends upon the change of the liquid olein into its solid isomer ellllidin, and is
especially applicable to olive and lard oils.
M anipulation.- Five grams of the oil are weighed 67-within 2 drops-into a
cordial glass, 7 grams of nitric acid, specific gravity 1.34, are then weighed into
it,67 and two pieces of copper wire (0.6 to 1.0 gram) added. Place the glass in a
pan of cold water at about 12 C., and stir with a short glass rod about 20 to 30
"6Valenta, Dingler Polyt. J., 253, 418; also J. Soc. Chern. Ind., 3, 643, 1884.
67 Not on the analytical balance.

1766

FIXED OILS, FATS AND ,WAXES

turns, not only with a rotary movement, but also with an up-and-down motion,
so as to mix the oil and the evolyed gas thoroughly.1 When the wire has diS"solved, add a second piece and stir as before. Th~s second addition should
furnish gas enough if the liquid has been kept cool ~nd the stirring has been
\ ,
thorough.
At the end of the first hour, pure lard oil will usually show flakes of a wax-like
appearance, and upon standing without disturbance and at the same temperature for another hour, the oil will have changed to a solid white cake hard
enough to bear several ounces' weight, or admit of lifting the glass and 'contents
by the glass rod. .
.
Most of the fish and seed oils yield a pasty or buttery mass separating from a,
fluid portion, whereas olive, almoIld, peanut, lard, sperm and sometimes neat'sfoot oil, yieid a solid cake.
"
Instead of using nitric acid and copper, sulfuric acid of 46 Baume, c(\Iltaini,ng a little nitric acid and saturated at 0 C. with hitric oxide, may be
employed.
A test should always be made at the same time with an oil of undoubfe'd purity.
,

NOTEJs.-If the oil be stirred too much or too' frequently, or is too warm, it has no
opportunity to form a hard cake.
Hubl states that all attempts to make the'test a quantitative one have resulted
in failure.
Mercury can be used instead of copper.
Cailletet's method,68 in which a smaller quantity of oil is used, and sulfuric and
nitric acids allowed to act upon it in a boiling water bath, cannot, in the experience of
the writer, be depended upon to give reliable results.
/
Maumene Test. 69-While this, like the'preceding, is not a quantitative t~st~
yet the indications afforded by it are of more value in many cases than those
obtained by, quantitative methods, as, for example, the saponification value.
It depends upon the heat developed by the mixing of the oil with strong sulfuric
acid. This takes place in a small beaker 7Y:; to 9 cm. deep and of 150 ml.
capacity, packed in an agate-ware cup with dry felt br cotton waste packing.
Manipulation."":Fifty grams of the oil are weighed 70 into the beaker to
within 2 drops, and its temperature noted by a thermometer. Ten ml. of
sulfuric acid are now run gradualfy into the oil':__allowing the graduate to drain
five seconds-the mixture being stirred at the same time with the thermometer,
and the stirring continued until no further increase in, temperature is noted.
The highest point at which the thermometer remains "Constant for any appreciable time is observed, and the difference between this and' the initial temperature is the" rise of temperature." This varies with the strength bf the
acid employed, and. to secure uniformity 7'1 the, results should b~ expressed by
dividing the rise of temperature with the oil by the rise of temperature with
water, and multiplying by one hundred. This is c,alled'the " specific temperature reaction." The rise of temperature with watkt is determined in the same
manner as with oil, using the same vessel.
"
~ ?;
68 Milliau, J. Am, Chem. Soc., 15, 156, 1893.
69 80 2Cl gives similar results.
70 Not on the analytical balance.
r '_
n Tortelli, J. Soc. Chem. Ind., 23, 668, 1904, is una~e to secure uniformity in ~his
way.

FIXED

OI~,

FATS AND WAXES

176'7

.. ( NOTEs.-In performing this, test it is important that the oil and acid be of the
same temperature, attained by keeping them beside each other.

The strength of acid should be as far as possible the same; it should be determined
not by specific gravity, but by titration, as 100% and 94.3% acid have the same specific
gravity:' .,
For,concordant results the conditions should be the same, and the same apparatus
sho}lld"p! used. In case the test is to be applied to a drying oil, it should be diluted
pne-half with a mineral oil, 25 paraffin, for example, thoroughly mixing them. The
"''rise of temperature" is then, the rise of temperature of mixture minus half the rise
of temperature of 50 grams of mineral oil, multiplied by 2.
It is advisable to make a test at the same time with an oil of known purity. Results should agree within 2%. By the use of the Hubl formula, page 1769, substituting
thermal va,lues, res,ults comparable with those obtained with the iodine value can be
obtained.
Sherman, Danziger, and Kohnstamm 72 have studied this method with the idea
of eliminating the errors. Rather than dilute, the oil with a mineral oil they dilute
the acid, using one of 89%. The results obtained are a little lower for vegetable oils
and a little hjgher for animal oils than those usually'found with the strong acid as
employed by Thomson and Ballantyne. Mitchell 7~ uses an inert diluent-carbon
tetrachloride-in a vacuum-jacketed' tube and one-fifth the quantities; all oils are
diluted. Hefinds that the;results obtained'are"in close agreement with the bromine
thermal vaJues; further, that the test filll,y b,e of use in determining the degree of oxidation of fats and oils, the figures becoming greatE!r with the age of the oil.
Data upon various oils will be found on pages 1795-1796.
'
(t

~,

,~

'!-'

"References

Mauril;r:te, Compt.-Re~d.) 35, 572/1852.

Ellis, J! Soc. Chem. Irid.'; 5,'361, 1886.
. Thomsoru:and Ballantyne, J:" Soc. Chem. Ind:; 10, 234, 1891.
",I, Richmpnd, An~y.st, UJ, 58, 1895.
".,
'
Munr~~I1' Am. Pub~,Hlaltli Alli?)}O, '236, 1884.

't

Iodine Number or Value.-This is the'percentage of iodine absorbed by an

oil; the method depends upon the fact that different oils absorb different
amounts of the halogens; the process is mainly one of addition, although small
quantities of substitution products are formed. For example, the unsaturated
body olein, (C17HaaCOO)aC3H5, when brought in contact with iodine takes up
6 atoms and forms the addition product, di-iodo stearin, (C17H33I2COO)aC3H5.
Palmitirr,. (ClsH31COO)aC3H5,' when similarly treated, forms' no addition
product;' but a small quantity of the substitution product, iodo-palmitin,
(d15m~Y.OO)aH.f,and the hydrogen disphiced unites with the io'dine toforiri
hydriodic~acid. .['he quantity of hydrioclic acid thus fo'rmed is a measUI:~ of
the amount of substitution. 74 "
'

1. Hanus's Method. 75 Manipulation.-From 0.12 to 0.15 gram of a drying

oil, 0.2 and 0:3 gram of a'hon-drying oil, or 0.6 to 0.7 gram of a solid fat, is'accurately weighed into a dry 200-ml. bottle. This should be of colorless glass and
be provided with a well-ground stopper.. ~This is best effected by pouring out
about 5 grams of the' oiLinto a No.1 beak,er containing a short stirring rod, and
setting it into a watch-glass upon the pan of the analytical balance. The whole
system is weighed, the beaker removed, and several drops of oil transferred to
the bottle by dropping down the rod, being careful that no oil touches the neck.
72

J. Am. Chem.- Soc\ 24, 266, 1902.

Analyst" 26, 169, 1901.

McIlhiney, J. Am. Chem.,Soc., 16, 275, 1894.
75 This is the official K. S. T. M. metlioa.

73
74

1768

FIXED OILS, FATS AND WAXES

Eight drops are approximately 0.2 gram. The beaker is replaced in the watchglass and the system again weighed, the difference in ,weight being the amount
of oil in the bottle.
The oil is dissolved in 10 m!. of chloroform, 30 fil. of the iodine solution
(Reagents) added-best from a burette-and allowed: to stand with occasional
shaking for exactly thirty minutes; with oils of an iop.ine number of less than
100, ten minutes suffices; 10 m!. of potassium iodide solution are added; shake
thoroughly, add 100 ml. water, washing down any iodine on the stopper. Titrate the excess of iodine with N /10 sodium thiosulfate, adding the latter
gradually, with constant shaking, until the yellow color of the iodine has almost
disappeared. Add a few drops of starch solution, shake thoroughly to dissolve
out the iodine from the chloroform solution, and titrate to the disappearance
of the blue color. The blue color returns after a few minutes but the end-point
is that of its first disappearance.
At the same time at which the oil is prepared, two " blanks" should be
prepared similarly in every way to the actual tests, except in the addition of
the oil, and treated in every respect like them; the strength of the thiosulfate
solution should also be determined the same day on which this test is carried out.
Standardization of the Thiosulfate Solution.-Ten ml. of potassium iodide
and 100 m!. of water are poured into the' Erlenmeyer flask; 20 m!. of the bichromate solution, equivalent to 0.2 gram of iodine, are now measured in with
a pipette, and to this 5 m!. of concentrated hydrochloric acid added and the
mixture shaken for three minutes. It'is now titrated with the thiosulfate
solution until the yellow color of the iodine has almost disappeared; starch
paste is now added, and the titration continued until the deep-blue color of the
solution changes to a sea-green-due to CrCla,-which is usually brought about
by the addition of a single drop.

The reactions involved are:

K2Cr207+14HCl=2CrCla+2KCl+7H20+3Clz;
3CI2+6KI=6KCI+3I 2;
I
.
6Na 2S20 a+31 2 =3Na 2S406+6NaI. .
NOTES.-Wijs 76 uses iodine chloride instead of bromide; it is more troublesome
to prepare and gives results about 1.2 points higher. 77 Either of these methods has
the advantage over Hubl's-first, that the solution!! keep better, remaining practically
unchanged for several months; secondly, that the action is about sixteen times as rapid,
it being completed in fifteen minutes; thirdly, that the solutions are cheaper.
Acetic acid cannot be displaced by carbon tetrachloride as a solvent, as the last
traces of iodine are difficult to remove from it. The acetic acid used should be at least
99.5% and show no reduction with potassium bichromate and sulfuric. acid.
This method has been referred by Gill and ~imms, J:Ind. Eng. Chern. 13,
547, so that a single drop of oil 0.03 gm. suffices for a test.
Z. Hiibl's Method. Manipulation.-The oil is weighed out as in 1, into
300-ml. bottles, except that about 25% more may 'be used.
'
The oil is now dissolved in 10 m!. of chloroform, 30 m!. of iodine and mercuric chloride solution added, the bottle placed in a dark closet, and allowed to
stand, with occasional gentle shaking, for four hours. If th,!:l solution becomes
76
77

Berichte, 31, 752, 1898.

Tolman and Munson, J. Am. Chern. Soc."

...!

~5,

244, 1903.

FIXED OILS, FATS AND WAXES

1769

nearly decolorized after two hours, an additional quantity shoulq be added.

One hundred ml. of distilled water and 20 ml. of potassium iodide are added to
the contents, and the excess of iodine titrated with sodium thiosulfate. If at
this point a red precipitate (HgI 2) is formed, more potassium iodide should be
added. As the chlQroform dissolves some of the iodine, the titration can proceed until the chloroform layer is nearly colorless, then the starch solution is
added, and the operation continued to the disappearance of the blue color.
" Blanks" should be titrated as with the foregoing process, page 1768.
NOTEs.-The method was proposed by Cailletet in 1857, made use of by Mills
and Snodgrass 18 in 1883, using, however, bromine and carbon bisulfide, and described
in almost its present form by Hiibl. 19 The chief factors in its execution are (1) strength
of the iodine solution; (2) the quantity used; and (3) the length of its time of action.
1. The Strength of Iodine Solution. According to Hiibl's original memoir, the solutions can be kept indefinitely when mixed.
Fahrion 80 states that the solution deteriorated as much as from 17 to 23% in eight
days. Ballantyne 81 confirms the deterioration, but finds it much less, 5 to 8% in
thirty-eight days. This weakening of the solution is probably due to the hydriodic
acid' formed by the action of the iodine upon the alcohol. 82
The mercuric chloride acts apparently as a carrier of iodine, as the reaction takes
place very slowly without it. (Gantter.) 83 Waller 84 finds tJ:at the addition of 50
ml. HCI, specific gravity, 1.19, to the mixed iodine solution preserves it for months.
Of the other metallic chlorides, CoC1 2gives the highest true iodine value, MnCI 2, MnBr2
and NiCl 2 cause practically no addition. (Schweitzer and Lungwitz.) 85
2. The Quantity of J odine Solution Used. The mixed iodine solution as made up
should require about 53 ml. of the thiosulfate. Before using, a rough titration should
be made, and if it be much weaker than this, a proportionately larger amount added.
The action of a large excess of iodine,is to increase the substitution rather than addition;
_increase in temperature or in time produces the same effect. 86
The excess of iodine recommended is from 150 to 250%;'some observers recommend from 400 87 to 600%.88
3. Length of Time. Two hours is sufficient for olive oil, tallow, and lard, while
for linseed oil, 18 hours, balsams, and resins 24 hours should be allowed. 89

Ingle 90 has shown that the free acid formed during the process is due toth.e
action of water upon the iodochlorides. Some of these are reduced by potassium iodide with liberation of iodine and consequent reduction in the iodine
absorption. Iodine chloride is the active agent, and not hypoiodous acid.
For the calculation of the percentage of adulteration of one oil by another,
Hubl gives the following formula: 91
" Let x = percentage of one oil and y = percentage of the other oil, further,
18 J. Soc. Chem. Ind., 2, 435, 1883.
19 Dingler Polyt. J., 253, 281; also J. Soc. Chem. Ind., 3, 641, 1884.
..0 J. Chem. Ind., 11, 183, abstr., 1892.
81 Ibid., 13, 1100, abstr., 18.94.
82 J. Soc. Chem. Ind., 14, 130, 1895.
83 Ibid., 12, 717, abstr., 1893.
84 Chem. Ztg., 19, 1786, 1831, 1895.
85 J. Soc. Chem. Ind., 14, 1031, 1895.
86 J. Soc. Chern. Ind., 12,717, abstr., 1893.
87 Ibid., 14, 1031, 1895.
.
88 Holde, Mitt. kgl. Techn. Versuchs., 9, 81, 1891.
89 Dieterich, J. Soc. Chern. Ind., 12,3&1,1893.
90 J. Soc. Chern. Ind., 21, 587, 1902.
91 Dingler Polyt. J., 253, 281, 1884.

1770

FIXED OILS, FATS AND W.AXES

I'

m = iodine value of pure oil x, n of pure oil y, and I of the sample under examination, then .
.
j
100([-n) "

x=--'-----'-

m-n

He further states that the age of the oil, provide~ it be not rancid or thickened, is without influence on the iodine value. Ballantyne 92 'finds that light
and air diminish the iodine number.
As might be expected, the iodine value is inversely proportional to the cold
test.
The method, as will be seen, is a conventional one, and the best results will
be obtained by using measured quantities of reagents and 'carrying through the
process in 'the same manner every time. 93
,
,) ")',. '
...
The calculation is perhaps most easily made 'as follbws: Subtract the number
of inl. of thiosulfate used for the titration of the oil, from that obtained by
titrating the blank-this gives the thiosulfate equivalent to the iodine absorbed
by the oil. Multiply this number (of ml.) by the value of the thiosulfate in
terms of iodine, and the result is the number of grams of iodine absorbed by the
oil; this divided by.the weight of oil used and multiplied by 100 gives the iodine
number.

'
In case it be desired to recover the iodine used" reference may be had to an
article by Dieterich, abstracted in the Jour. Soc. Chem. Ind., 15, 680, 1896.
Oxidized Oils. Iodine Number of .L-To find the original iodine number of a..
semi-drying or non-drying oil which has been altered by atmospheric oxidation,
add 0.8 to the iodine number found on the altered sample for each increase of
' h
'fi
' . (k
15.5 C.
O001 10
t e specI c gravIty ta en at 15.50 .C.
.

)94

Bromine Number or Value.-The iodine method just described has, among

oth~rs, the disadvantllge .that it fails to distinguish between addition and substitution; this is sometimes of importance, and to accomplish it McIlhiney 9.
makes use of the bromine absorption.
I
Manipulation.-From 0.2~t() 0.3 gram of a'drying oil, 0.4 to 0.5 of a nondrying oil, or 1.0 to 1.2 grams'of a solid fat, are accurately, weighed into the 300
ml. bottle, as in the iodine number (page 1767).
The oil is dissolved in 10 m!. ,of carbpn tetrachloride, and 20 ml. of bromine
solution (Reagents) added, best from a burette. After allowing it to stand two
minutes by the watch, 20 or 30 ml. of,potassium iodide areadde<i, in the manner
described below, the amount depending upon the excess' of bromine. To prevent loss of bromine and hydrobromic acid, a short piece of thin and wide
rubber tubing-" bill tie tubing "-is slipped over the lip of the bottle, thus
forming a well around the stopper; some of the iodide solution is poured into
this and the bottle cooled in cracked ice. U pop. removing the stopper the
solution is sucked into the bottle, it is shaken to insure the solution of the
vapors, and the remainder of the reagent added. The iodine liberated is
titrated by sodium thiosulfate in the usual way.
92 J. Soc. Chern. Ind., 10, 31, 1891.
,
. .t,
'

93 If, for example, the water be added before the iodic;ie solutlOn, the iodine number
is changed by 0.3 per cent.
'
-' :_ .
,
94 Sherman and Falk, J. Am. Chern. Soc., 27, 608, 1895.
96 J. Am. Chern. Soc., 21, 1084, 1899.

FIXED OILS, FATS AND WAXES

1771

When this titration is finished, 5 ml. of the potassium iodate solution are
added and the titration repeated. The iodine liberated in this reaction is
equivalent to the hydrobromic acid present. Blank determinations should be
made with the reagents used, as with the iodine number.
NOTEs.-Oftentimes, particularly with resins, emulsification of the solution takes
place, masking the end-point. This can be prevented by the addition of 50 or 100 mi.
of a 10% solution of salt.
"In case ice be not at hand, the vapors will probably be completely absorbed by
passing through the iodine solution in the rubber well.
The reactions involved, in addition to those on page 1768, are:
Palmitin
(C 15H.,COO),C.H,+3Br.= (C 15H. oBrCOO)3C.H.+3HBr.
3HBr+3KI = 3KBr+3HI.
6HI + KIO. = 3I.+3H 20+ KI.
,The calculation is similar to that followed in the iodine nUinber (page 1767).
,1'he percentage of bromine found as hydrobromic acid is called the bromine substitution figure, and the total percentage absorbed, less twice the bromine substitution
figure, gives the bromine addition figure.
,
The method has the further advantages that it is rapid, the bromine solution is
permanent and inexpensive. For data upon various oils, see table on page 1795.
Saponification Value.-This is expressed by the number of milligrams of
potassium hydrate necessary to saponify one gram of the oil. It is called from
the originator" Koettstorfer 96 number or value," also" Saponification number," and must not be confounded with" Saponification equivalent" as pro,posed by Allen,97 which,is the number of grams of oil saponified by 56.1 grams
of potassium hydrate.
METHOD

98

Solutions Required: Alcoholic solution jor saponification.-Dissolve 58

grams of potassium hydrate" purified by alcohol" in 500 ml. of 95% purified
ethyl alcohol. Allow the solution to settle in a dark place. Draw off the clear
solution or filter through an asbestos filter and make up to 1 liter with 95%
alcohol. The solution so prepared shall stand at least 20 hours before it is
standardized.
Alcohol.-Purify 95% ethyl alcohol with silver oxide in the following
manner: 99
Dissolve 1.5 grams of c.p. silver nitrate in about 3 mI. of water, add to
1 liter of alcohol in a glass-stoppered bottle, and mix thoroughly. Dissolve
3 grams of potassium hydrate (by alcohol) in 10 to 15 ml. of warm alcohol.
After cooling, add slowly to the alcoholic si:ver nitrate solution; stirring slightly.
Allow the' precipitated silver oxide to settle, siphon off the clear solution, imd
distil on a steam bath.
Standard hydrochloric acid sqlution.-One-half normal solution.
96 Z. anal. Chern., 18, 199, 1879.
Commercial Organic Analysis, 2, 40.
98 Tech. Paper 323B.
99 Dunlap, J. Am. Chern. Soc., 28, 397, 1906.
97

1772

FIXED OILS, FATS AND WAXES

Phenolphthalein solution.-One gra~ phenolphthalein in 100 m!. alcohol

and water.
I
.Apparatus.-The saponification shall be carried out in a wide-mouthed,
flat-bottom extraction flask, or Erlenmeyer flask, ofl 250 to 300 m!. capacity,
fitted to a reliable condenser properly connected 'with a good cork. The
boiling shall preferably be carried on by means of all; electric hot plate.
Blank Determination.-Determination shall be made in duplicate in the
alcoholic potash solution in the following manner:
'<
Measure accurately into the flask 25 mI. of alcoholic potash solution from
a calibrated pipette. Th~ tip and outside of the pipette shall be wiped off
with a clean filter paper before the solution is deliver:ed, then rinsed out with
25 ml. of neutral alcohol. If a standard burette is used, allow 60 secon'ds
total time for drawing and draining. Connect the flask to a suitable condenser
and boil for 3 hours. Before disconnecting the flask, wash out the condenser
with a few milliliters of neutral alcohol; if a Soxhlet is used as the condenser,
the tip shall be washed off into the flask. Titrate while hot with N/2 HCl,
using 3 drops of phenolphthalein as indicator. The total number of milliliters
of N/2 HCl required for the blank represents the strength of. the alcoholic
potash solution.
Procedure.-For straight fats or oils, use 2 or 3 grams of the material, for
oils containing over 30% of fatty oils, use about 5 grams, and for oils containing less than 30% of fatty oils, use about 10 grams.
Weigh the oil accurately, by difference, p. 1768, from a small beaker into
the saponification flask. Add 25 mI. of alcoholic potash solution and 25 mI.
of neutral alcohol in the same manner as for the blank, connect to the condenser, and boil for 3 hours.' Titrate while hot. Calculate the saponification
number from- the difference between the number of milliliters of N/2 HCl
required for the determination and the average of the two blanks, using the
formula:
'fi'
b
difference, milliliters X 28.05
Sapolll catIOn num er
. h f '1
WeIg t 0 dl, grams
Petrolic ether may be used with compounded cylinder oils in the quantity
of 50 mI., providing a Soxhlet extraction flask is used to collect this petrolia
ether periodically. The Soxhlet flask should be so adjusted, through the
addition of glass rod ding or beads when necessary, that it willjust overflow
with the full quantity of petrolic ether.
The percentage of fatty oil (or fat) in a compounded petroleum product
can be calculated from the saponification nl!mber of such a product only
when the saponification number of the fatty oil is known. If the saponification
numbers of both fatty oil and compounded oil are known, the following formula
should be used:
'
'1
100
X
saponification
number
of compounded oil
P ercen t age 0 f fa tt y 01 =
.
saponification number of fatty 011
For this determination. the following values of saponification number may
be used:

FIXED OILS, FATS AND WAXES

Fatty Oil

Lard oil. ....................

Tallow ..................... :
N eat's-foot. . . . . . . . . . . . . . . . . ..
Fish .........................
Sperm .......................
Castor .......................
Rapeseed ....................

Saponification
Number

192-198
193-198
193-204
140-193
120-140
176-187
170-179

Fatty Oil

Soya bean ...................

Peanut ......... , ............
Cottonseed ...................
Blown. rapeseed ...............
Blown cottonseed. . . . . . . . . . . ..
Degras ......................

1773
Saponification
Number

189-197
18{)-197
191-197
195-216
210-225
110-210

NOT:E1s.-Many prefer to cork the flasks tightly and tie down the stoppers, thus
saponifying under pressure; others make use of a return flow condenser, oftentimes
.
'
merely a long glass tube.
McIlhiney 100 has applied the process to dark-colored substances by making use
of the principle that when ammonium chloride is added to a neutral soap solution,
and the mixture distilled, the amount of ammonia freed is e,quivalent to the quantity
of alkali combined with the fatty acids. As a description of the process is beyond
the scope of the present volume, reference must be had to the original article.
As ordinarily prepared, the alcoholic potash solution turns rapidly reddish brown,
so that it is very difficult to note the end-point. This trouble can be partially avoided
by "adding a drop or two of the'solution to the diluted indicator contained upon a tile
after the manner of the titration of iron by bichromate. As the color is probably
due to the polymerization of the aldehyde formed by the oxidation of the alcohol, it is
more satisfactory to use for the preparation of the potash solution an alcohol which is
practically aldehyde free. Alcoholic potash made up from this, using the so-called
"potash by alcohol," will give a solution which will remain water-white for weeks.
"Thymol blue" may be used as an indicator.
The writer has found, if the stock solution be kept under an atmosphere of hydrogen,
that the coloration by standing is almost entirely prevented.
This method has also been refined by Gill and Simms, J. Ind. & Eng. Chern., 13,
547 (1921) so that a single drop, 0.03 gram, suffices for a test.

Detection of Un saponifiable Oils.-The qualitative detection takes place 'by

observing the behavior of the solution obtained by boiling the oil"with alcoholic
potash when diluted with warm water. Any unsaponifiable material will
manifest itself as oily drops in the clear alcoholic solution, OF as a whitish cloud
on the addition of water.
The quantitative determination may. take place in two ways: 1. From the
saponification number. 2. By gravimetric methods.
1. From the Saponification Number.-On pages 1795 and 1796 it will be
noticed that, except for castor, rape, and sperm oils, the saponification number
averages 193. If the number found be divided by this figure, the percentage of
saponifiable matter will be obtained; this subtracted from 100 will give the unsaponifiable matter. This method gives no idea of the kind of saponifiable
matter.
2. By Gravimetric Methods.-The procedure is essentially that of Spitz and
Honig: 101 10 grams of the oil are boiled fifteen minutes under a return-flow
condenser with 50 m!. of 5% alcoholic potash; 102 40 m!. of water are added
and the boiling repeated. The liquid is allowed to cool, washed into a separatory funnel with 50% alcohol and 50 ml. of 86 gasoline, thoroughly shaken
100 J. Am. Chern. Soc., 16, 409, 1894.
For a discussion of the theory of the process,
see Lewkowitsch, J. Soc. Chern. Ind., 17, 1107, 1898.
101 Z. ang. Chern., 19, 569, 1891.
102 The potash is made by dissolving purified potash in the smallest possible quantity of water and adding absolute alcohol.

1774

FIXED OILS, FATS AND WAXES

and allowed to stand. The gasoline layer should separate clearly and quickly
from the soap solution and the latter is drawn off;' the; gasoline is washed 2 or
3 times with 50% alcohol to extract Any soap, and these washings added to the
soap solution. This latter is extracted, until upon evaporation the gasoline
leaves no stain upon paper, carJfbeing taken to 'wash t~e gasoline extracts each
time with 50% alcohol; three extractions with gasoline are usually sufficient.
The gasoline is distilled from these extracts, the residue heated until the
gasoline odor disappears, and weighed. From the appearance of the residue
some idea of the kind of unsaponifiable matter can be, obtained. This in th~ .
case of sperm oil will be mainly solid "alcohols, probably of the ethylene series.
According to Schicht and Halpern 103 this method is open to the following
errors: incomplete saponification, -incomplete extraction, solubility of soaps'
in the solvent, and the solubility of the unsaponifiable matter in the washing
solution. Their improved. method is as follows: 5 grams of fat 104 with 3
grams of solid caustic potash dissolved in a little water and 25 m!. of absolute
alcohol are boiled half an hour under a reflux condenser. After cobling 25 m!.
of 10% KCI are added and the solution is then shaken four times with 200 ml.
of petroleum eth~r distilling under 60. The 'petroleum ether is evaporated
and, without washing, the residue is dissolved in 25 ml. absolute alcohol and
the solution made slightly alkaline with normal alkali; 25 ml. of 10% KCI are
added and the shaking witl;l petrole~m ether repeated. The petroleum ether
solution is shaken with 1QO mI.. of 50% alcohol and the wash solution with
100 ml. petroleum of ether, which is afterwards washed with 100 ml. of 50%
alcohol. After combining the extracts the 'J?etroleum ether is driven off and
the residue dried and weighed.
.
NOTE.-Care should be taken to use gasoline which leaves no residue on evaporation at 100 C.
r
0

Identification of the Unsaponifiable Matter.-The unsaponifiable matter is

either liquid or solid; in case it is liquid, it may be (1) 'hydrocarbon oils, either
mineral, or formed by the distillation of waste fats, as' wool grease, or (2) tar
oils, "dead oils," etc., obtained by th~ distillation of cbal tar; or (3) rosin oils.
If it be a question of one of these three, the .specific gravity will usually
decide it; that of the hydrocarbon oils is 0.855 to 0.930, of the rosin oils 0.96
to 0.99, while the tar oils are heavier than water. Rosin oils would be shown
by the Liebermann-Storch test, p. 1781; a mixture of mineral and tar oils
would be identified by treatment with an equal quantity
nitric~acid, sp.gr.
1.45, both previously c(iloled to 15 C., and noting the rise of temperature.
Mineral oils give a very slight rise, being paraffins, whilethe tar oils belong
to the benzole series and are more easily nitrated. Hydrocarbon oils from
distilled grease oleins can be identified by their refractive index and rotatory
poWer. 1 0S
(;
Solid unsaponifiable matters may be:
(4) p,araffin.

or

Chem. Ztg., 31, 279, 1907.

For linseed and other oils, ten or twenty times this weight sp.ould be used, the
alkali being correspondingly increased.
,_'
105 Gill and Forrest, J. Am. Chem. Soc., 32, 1071; GHl.amU,\iason, J. Am. Chem. Soc.,
103

104

~OO&

FIXED OILS; FATS AND WAXES

'l775

(5) Ceresenec-refined ozokerite.

(6) Higher alcohols of the paraffin series, as cetyl, C 16H aa OH, coming from
the saponification of sperm oil and other waxes.
(7) Cholesterol, C 26 H 4a OH, and its isomers, phytosterol, sitosterol, isocholesterol, etc.
(8) Lactones, internal anhydrides 'Of oxy acids as stearlact'One,
-'
,. -'I: ,,<1.,i'

C14H29CHOHCH2CH2COOH = C14H29CHCH2CH2COO + H 20.

These may be separated by boiling for two 'hours with an equal q~~~tity of
acetic anhydride; if the substance dissolves and does not precipitate on cooling,
higher alcohols are indicated; if a mass.,of crystals separates out on cooling,
cholesterol and its isomers, or a mixtunt of these with the, higher al~9hols is
indicated; if an oily layer remains on tOPidt;is IUl indication of the presence of
paraffin or ceresenc. For the complete separation and identification of these
reference must 1?e had to Lewkowitsch, ".Analy,sis of Fats, o.ils, and W~~es,"
as .it-is beyond the limits of this chapter.. .
h
Test for Animal or Vegetable Oils.~Animal oils contain cholesterol,
C26H 4S OH, while vegetable oils contain the isomeric body phytosterol; hence
t4e isolation and iden~ification of these compounds enables one to say with
certainty as to the presence of one class of oil or the other-for example as-to
the presence of fish oil in linseed. The quantity of these bodies varies from
0.2' to 1%. The method is essentially that of Bomer .106 Fifty grams of the
oil are boiled in a flask 'with a return cooler with 75 ml. of 95% alcohol for five
minutes and the alcoHolic solution separated; this is repeated with another
portion of alcohol. The alcoholic solutions .are mixed with 15 ml. of 30%
Rodium hydroxide and evaporated on a water bath nearly to' dryness in' a
porcelain dish and the residue shaken out with ether. The ether is evaporated, the residue taken up with a little ether, filtered, again evaporated,
dissolved in 95% alcohol (by volume), and allowed to crystallize slowly,
DO'mer states that the form of the crystals is more to be relied upon tljan.3.
determination of their melting-point! Cholesterol crystallizes from alcohoL'or,
ether in leaflets or rhomboid tables containing one molecule of water of crystallization. Phytosterol crystallizes also from alcohol with one \llolecule of
water in needles forming stars or bundles. As a further means of identification, some of the esters should be made and their melting-points determined.
To this end the crystals above obtained are heated over a low flame in a
small porcelain dish covered with a watch-glass, with 2 or 3 ml. of acetic or
other acid anhydride until it boils: the watch-glass is removed and the excess
of anhydride evaporated on the water bath. The contents of the dish are
treated with a small quantity of absolute alcohol to ,prevent crystallization,
more alcohol added and the solution allowed to crystallize. The .crystals are
filtered off through a very small filter, washed with a small quantity of 95%
a:1cohol, dissolved in absolute alcohol, and recrystallized until a constant melt. ini\'7oint is obtained.
.
The following table shows the corrected melting-points of these alcohols
and their esters:
106 J. Soc. Chern. Ind., 17, 954, 1898; Tolman, J. Am. Chern. Soc., 27, 590, 1905;

J:': ':

Tolman, Bull. 107, U. S. Dept. Agriculture, 1907.

1776

FIXED OILS, FATS AND WAXES

Cholesterol

Alcohol. ......................... 148-150.8

Acetate .......................... 113-114
Benzoate. . . . . . . . . . . . . . . . . . . . . . .. 135-151
Propionate. . . . . . . . . . . . . . . . . . . . . . . 97-98

Phytosterol

136-143.8
120-137
142-148
104-105

NOTEs.-Some directions state, in isolating the cholesterol or phytosterol, to boil

with the 30% sodium hydroxide until one-fourth of the alcohol is evaporated. As a
result of repeated experiments this has been found to cut down the. yield so much that _
on a large scale 'Practically none of these bodies, particularly phytosterol, was obtained.
This agrees with the observation of Lewkowitsch that by heating cholesterol with normal
alcoholic potash, cholesterin hydrate- is obtained.
The following test will serve to differentiate between cholesterol and phytosteroP07
A very small quantity of cholesterol is warmed with 1.5 ml. absolute alcohol and a
trace of isodulcit or rhamnose (o-dimethylfurfural) added. After cooling, an equal
volume of concentrated sulfuric acid is added, so as to form a layer below the solution;
whereupon a raspberry-colored ring is produced at the zone of contact of the two liquids.
On mixing the layers while the tube is cooled in a current of cold water, the mixture
becomes intensely colored. With phytosterol the reaction fails or at most a. pink color.
Similar reactions are given by abietic acid. Or mix the pure unsaponifiable matter
with an equal quantity of pure cholesterol; if the melting point be unchanged, the unsaponifiable matter is cholesterol; lowering of the melting point indicates phytosterol.
As little as 1 % pf cotton-seed has been found in lard, and 4% in any oil has been
detected by this test.
For the means of distinguishing between drying and marine animal oils, see Halphlm,
J. Pharm. Chim., 14,391 (1901), abstracted J. Soc. Chem. Ind., 21, 74, or Chem. Centralb., 72, ii, 1097 and 1323.

Tests for Antifluorescents. 1()8__It is often desired to remove the fluorescence

or "bloom" from petroleum oils. This may be effected by refining with
chromic acid, or more easily by the. addition of a small quantity of nitronaphthalene or nitro-benzene. The latter may often be detected by the odor.
The test is made by boiling about 1 m!. of the oil with 3 m!. of 10% alcoholic potash for one to two minutes. If either of the nitro compounds be
present, a blood- or violet-red coloration is produced; ja pure mineral oil is
changed only to yellow or brownish-yellow by this trliatment. In case the
characteristic color does not appear the following'test may be applied. 109 It
depends upon the reduction of the nitro bodies to their amines.
A few I}ll. of the oil are heated with feathered tin and hydrochloric acid
in an Erlenmeyer flask for ten minutes: this can be aided by the introduction
of a piece of platinum wire. The oil is separated by a separapory funnel and
filtration through a wet filter, the filtrate treated in another separatory with
sodium hydrate until the tin hydrate redissolves and shaken out JVith 10-20
ml. of ether. The amines go into solution in the ether, giving to it a violet
color and fluorescence in the case of a-naphthylamine. These can be recognized by their odor, that of naphthylamine being very characteristic. The
latter may be recognized by dissolving in hydrochloric acid, evaporating the
latter, and upon treatment with ferric chloride obtaining an azure-blue precipitate. This changes when filtered off to purple-red and the filtrate to violet.
. Aniline can be recognized by solution in concentrated sulfuric acid and
the red and then blue color which appears on the addition of a ~mall crystal of
107 Neuberg and Rauchwerger, abstr. J. Soc. Chern. Ind,;~3, 1163, 1904.
108 Holde, J. Soc. Chem. Ind., 13,906, 1893.
.
-'
.
109 Holde, "Examination of Hydrocarbon Oils," p. 75.

/'

FIXED OILS, FATS AND WAXES

1777

potassium bichromate. Free aniline is also temporarily colored violet by a

solution of bleaching powder.

Acetyl Value.-The estimation of the acetyl value is seldom required' in oil

.analysis, it being characteristic only when triglycerides are present. For a
.description of the method and its applications, reference must be had to the
larger works, as Lewkowitsch or Allen.
SpeCial Tests for Certain Oils
Lewkowitsch says: 110 " It should be distinctly understood that color reactions taken by themselves should not be relied upon as giving a decisive answer.
At best they can only be used as a preliminary test, or as a confirmatory test .
The ease with which this test can be carried out, and its apparent reliability,
have led to an over-estimation of this very useful and important reaction; so
much so, that grave errors may be committed by those who assign to this test
an exclusive or even a paramount importance. It is altogether unjustifiable
to look upon this test, as has been done, as permitting of quantitative interpretation."
Bechi's Test for Cotton-seed OiI....This depends upon the supposition that
a substance of an aldehydic nature which reduces silver nitrate is contained
in the oil. The method is essentially that of Milliau. 11l
Fifteen grams of oil are weighed into a No.6 porcelain dish, using the coarse
scales, and heated for about ten minutes upon the water bath; a mixture of
10 m!. of 30% caustic soda and 10 ml. of the alcohol is slowly poured upon the
oil. The whole is occasionally stirred until the mass becomes clear and homogeneous, and 150 m!. of hot distilled water slowly added so as not to decompose
the soap, and'the boiling continued until the alcohol is expelled. Dilute sulfuric acid (1 : 10) is added to acid reaction, and the separated fatty acids
washed three times by decantation with cold water. A ,portion of these is
brought into a large test-tube, 15 ml. of alcohol and 2 m!. of 3% silver nitrate
solution are added, the tube is wrapped with brown paper, heJd in place by an
elastic band, and heated, with constant stirring, in the water bath until onethird of the alcohol is expelled, which is replaced by 10 m!. of water. This
heating is continued for a few minutes longer and the coloration of the insoluble
fatty acids observed. The presence of cotton-seed oil in any appreciable
proportion ca~s a mirrorlike precipitate of metallic silver, which blackens
the fatty acids \;'1 the mixture.
NOTEs.-The alcohol should, be proved free from aldehyde by a blank test. Unless
the mixture in the test-tube be thoroughly stirred while heating, it will "bump" and
eject the contents. Other methods of procedure consist in applying the test to the oil
itself, often after treatment with dilute caustic soda and nitric acid. (Wesson. ll2) The
writer had a case in which the oil gave the test while the/atty acids gave no blackening,
showing there was something in the oil itself other than cotton-seed oil which reduced
the silver nitrate. Students have no difficulty in detecting a 5% adulteration with
cotton-seed oil.
Dupont 113 thinks {hat the reduction of silver , nitrate is due rather to sulfur com-

"Chemical Technology and Analysis of Fats, Oils and Waxes," 2, 203.

m J. Am. Chem. Soc., 15, 164, 1893.
112 J. Am. Chem. Soc., 17,723,1895.
.
113 Bull. Soc. Chem. (3), 13, 696; J. Soc. Chem. Ind., 14, 811, 1895; also Charabot
and March, Bull. Soc. Chim., 21, 252, 1899.
110

FIXED OILS, FATS AND WAXES

1778

pounds .contained. in the oil; by passing steam over the oil he obtained a' product con"
taining sulfur and the oil still gave the Becni test, ThiEl work has been{repeate5i ~and
confirmEld...by the author,114 It is to. be noted that while ',the fatty acids blacken silver
nitrate they do not color cadmium, lead, or copper salts, but reduce mercury compounds. No indication of an aldehyde was noted by th~ fuchsine or ammonia. tests:
The supposition that the reducing substance is aldehydi~ in its nature finds support.
in the fact that if the oil be heated to 240 11. or ,be kept for sonie time l16 it loses tl;lis
peculiar property.
! . ,
By purifying the acids by the lead salts TO.ftelli and Ruggeri 117 are able to detect
as little as 10% of heated cotton~seed oil.
0

It is to be noted that pure lard, tung and ~ll've oil are not infrequently met
witllw'hlch give the test, consequently its indication~ cannot be considered as
conclusive.
' ';
Halphen's Test for Cotton~seed OilY5-This depends upon the observation
that this oil contains an unsaturated fatty acid which combines with sulfur,
giving a colored compound_I19 I
Procedure.-Ten ml. of the oil or melted fat are heated, in a 'large test-tube
with a; long' glass condenser tube attached, with an equal volume of amyl
alcohol.and of carbon bisulfide solution of sulfur (Reagents), at first with
ffeq~ent. a~itat~on, in a steam b, ath, and ~hen, after t~e v~~lent boiling has
ceasen; 'iIi 'a Jmne bath (105- ~ 10 0 C.) for ,forty-five' mmutes ,to three hours,
according to .the quantity of ja!iulterant present, the tube being occasionally
removed and shaken. As little as 1 % will give a crimson wine'coloration in
twenty minutes.120
>, ,
_(.

\ ~

NOTES.-If the.tnixture be heated for too long a time a

~olor due to burning is produced. The reaction se.ems to be

misleading brownish-red
peculiar to this oil; it is
more sensitive with fresh than old fats, and while, by comparison with a blank, to of
1% is noticeable, ,M of 1 % is easily detected. Cotton-seed oil which has been heated
to 250 C. does not give the test; tlie oil is then not available as food. Heating to
200 G. does not interfere with the test,121
,
T~ test is not given by an oil which has been oxidized with sulfuric acid and potassium p~manganate, although sYQh an oil gives the Bechi test.122 This shpws that the
two tests are not 'produced by the same substance. Nor .is this test 6r that of Bechi
given by an oil which has been treated with chlorine or ,sulfurous acid. 123 ff treated
with the former it is no longer edible; an oil treated with sulfurous acid and washed
with al~ohol cannot be distinguished from ordinary cotton-seed oil and does~not, as
already stated, respond to either the Halphen or Bechi ~est. In this case the test for
phytosterol_ is the ()nly means of determining if it has been added to an animal oil.
The test is also given: by kapok oil, which is used as an edible oil in China, the East
and West Indies, and in Africa; baobab oil also gives it.
-._
Lard from hogs. fed on cotton-seed meal shows this reaction strongly, as if it were
25% oil.l2{ The butter from cows similarly fed also yields the reaction.l2O
0

114 Gill and Dennison, J. Am. Chern. Soc., 24, 397, 1902.
Holde, J. S.bc. Chern. Ind., 11, 637, 1892.
. l16 Wilson, Chern. News, 59, 99, 1889.
117 J. Soc. Chemdnd., 20; 753, 1901.
118 Halphen, J. Pharm. <iJhitn.,'390, 1897. '.
119 Raikow, Chern. Ztg., 24, 562, 583, 1900.
'

.
12Oilar, Am. Chern. J., 24: 355; abstr. Anal., 26,!. 22. 1901. ,
121 Fischer and Peyan, Analyst, 30, 131, 1905; Mltsien, Z. oiIentl. Chern., 5, 135,
1899; J. Soc. Chern. Ind., 18,865.
I
122 Raikow, loc. cit.
123 Petkow, Analyst, 32 123, 1907.
124 Soltsien, Z. offent!: Chern" 7, 140, 1901.
12. Wauters, J. Soc. Chern. Ind., 19, 172, 1900.
_,.' 115

ErXED OILS, FATS AND WAXES

1779

The test may Be applied to the soaps or fatty acids, provided they are not too
deeply colored.
The amyl alcohol cannot be omitted nor substituted by ethyl alcohol without impairing the .delicacy of the test. 126 The compound.in the oil cannot be removed by
treating with animal charcoal.I24
'
-

Hexabromide Test for Linseed Oil.-The object of the test is to determirl.~

the amount of insoluble bromides of the fatty a-cids contained in the oil. Linseed gives the highest percentage 9f bromides of any of the common oils,
consequently its adulteration can 15e shown by the hexabromides .. 11.,
Fifteen grams of the oil are saponified by boiling with 15 ml. otpotassium
hydroxide solution, sp.gr., 1.35, and 15 ml. of alcohol in a flask under a reflux
condenser; 300 ml. of warm water are added and the solution distilled with
steam until the lJ,lcohol is removed .. Dilute sulfuric acid is added to excess,
the solution heated until the fatty acids are obtained as a clear oily upper
laYer; this is washed several times with hot distilled water until free from
sulfuric acid, using methyl orange as an indicator. This does not react with
fatty acids of low molecular weight which being soluble in water may redden
litmus. This washing i;:; effected in an atmosphere of inert gas, .carbonic acid
or hydrogen by stopping the flask with a three-holed stopper, carrying a siphon,
aD- entrance and an exit tube for the gas. The acids are siphoned into a small
Erlenmeyer flask and in case a few drops of water come over--an equal quantity of alcohol.added and dried upon the water bath in a stream of dry inert gas.
In order to test for the presence of unsaponified fat, 3 nil. are dissolved in
15 ml. of 95% (by volume) alcohol, and 15 ml. of aqueous ammonia are added.
If an appreciable 'amount of fat has escaped saponification, the mixture will
become turbid (Geitel).
Two grams oLmixed fatty acids 128 are dissolved in a flask in 27 ml. of Gry
ether, cooled do\~to 10 C., and 0.25 ml. of bromine allowed to run into the
solution from a very finely-drawn-out pipette, the time allowed for this being
about twenty minutes. The remaining 0.25 ml. of bromine is added somewhat .
more rapidly, within about ten minutes, the bromination thus occupying about
thirty minutes. The authors attach great value to the exact observance of
the time. The temperature should never be allowed to rise during bromination
above 5. The flask is corked and allowed to stand for two hours at 0. The
ethereal solution is next decanted through a weighed asbestos or paper filter
(Lewkowitsch) and the precipitate is washed with five lots of 5 ml. each of
dried and cooled ether. After complete draining, the precipitate is dried for
two hours at 80 to 85, and allowed to cool in a desiccator. The temperature
is designedly kept below 100, as the authors found that the color of the hexabromide becomes somewhat gray if the drying takes place at 100. The
melting-point of the hexabromides was In, whereas the melting-point of pure
hexabromide has been found to be higher. No doubt the low melting-point
is due to the drying having been carried out below 100.'
Nevertheless small traces of retained moisture cannot account for the much
larger yield of hexabromide which the authors obtained.
The yields of hexabromide obtained by these authors are as f~!l0ws:
Soltsien, loco cit., 25, Oilar, loco cit.
Utz, Rev. Fett u. Harz. Ind., 9, 125, 1902.
128 Eibner and Muggenthaler, Farben Ztg., 1912.
'

126

127

FIXED OILS, FATS AND WAXES

1780
Fatty Acids

Per cent

FattYIAcids

Per cent

Perilla oil. ................

Linseed oil, Baltic .........
Linseed oil, Dutch .........
Linseed oil, La Plata .......
Linseed oil, Indian .........

64.12
57.96
51.73
51.66
50.50

Tung oil .................

Soya bean oill ......... ,. ..
Poppy seed oil ............
Rape oil .................

nil
up to 7.78
nil
6.34
#

Bailey and Baldsiefen, J. 1. and E. Chem., 12, 1189 (1920), have devised a modification which is more elaborate and may give more accurate results.

Bellier's Qualitative Test for Peanut Oil.129

The necessary reagents are:
'
Alcoholic potash, made by dissolving 8.5 grams of pure potassium hydroxide
in 70 per cent alcohol and making up to 100 m!.
Acetic acid of such strength that 1.5 m!. will neutralize 5 m!. of the alcoholic
potash. (Acid having a specific gravity of 1.04 is approximately correct.)
Procedure.-Weigh 1 gram of the oil into a dry test-tube, add 5 ml. of the
alcoholic potash and boil gently over a small flame, avoiding evaporation as
much as possible. Boil until the oil is completely _saponified, which will take
a little over 2 minutes. Add 1.5 m!. of acetic acid, or just sufficient to neutralize the alcoholic potash (use phenolphthalein as an indicator), mix well,
cool rapidly by placing the test-tube in water at 17 to 19 C. and leave in the
water at the stated temperature for nof less than 30 minutes, shaking occasionally. Then add 50 m!. of 70 per cent al~ohol containing 1 per cent 'by
volume of hydrochloric acid Csp.gr. 1.16), shake well and again place in water
at 17 to 19 for an hour. If no peanut oil be present, the liquid will remain
clear or at most slightly opalescent; with 10 per cent, or more. of peanut oil
a flocculent, crystalline precipitate forms.
Renard l30-Tolman 131 Quantitative Test.
.
Weigh 20 grams of oil into an Erlenmeyer flask. Saponify with alcoholic
potash, neutralize exactly with dilute acetic acid, v,sing phenolphthalein as
indicator, and wash into a 500-m!. flask containing a boiling mixture of 100 m!.
of water and 120 m!. of a 2D% lead acetate so~ution. Boil for a minute and
then cool the precipitated soap by immersing the flask in water, occasionally
giving it a whirling motion to cause the soap to stick to the sides of the flask.
After the flask has cooled, the water and excess of lead c!!:_n be poured off and
the soap washed with cold water and with 99% (by volume)-alcohol. Add
200 m!. of ether, cork, and allow to stand fOf some time until the soap is disintegrated, heat on the water bath, using a reflux condenser, and"boil for about
five minutes. In the oils most of the soap will be dissolved, while{ in lards,
which contain much stearin, part will be left' undissolved. Cool the ether
sdlution of soap to 15 or 17 C. and let it stand until all the insoluble soaps
have crystallized out (about twelve hours).
Filter and thoroughly wash the precipitate with ether. Wash the soaps on
the filter back into the flask by means of a stream of hot water acidified with
The test is not thoroughly dependable.
' I
Renard, Compt. rend., 73, 1330, 1871; also Archbutt, J. Soc. Chem. Ind., 17,
1124.
. ",'131 Bull. 107, U. S. Dept. Agriculture, 1907, p. 145.
129

130

FIXED OILS, FATS AND WAXES

1781

hydrochloric acid. Add an excess of dilute hydrochloric acid, partially fill the
flask with hot water, and heat until the fatty acids form a clear oily layer.
Fill the flask with hot water, allow the fatty acids to harden and separate from
the precipitated lead chloride, wash, drain, repeat washing with hot water, and
dissolve the fatty acids in 100 ml. of boiling 90 per cent (by volume) alcohol.
Cool to 15 C., shaking thoroughly to aid crystallization.
From 5 tO,10 per cent of peanut oil can -be detected by this method, as it
effects a complete separation of the soluble acids from the insoluble, which
interfere with the crystallization of the arachidic acid. Filter, wash the precipitate twice with 10 ml. of 90% (by volume) alcohol, and then with alcohol
70% (by volume). Dissolve off the filter with boiling absolute alcohol, evap-"
orate to dryness in a weighed dish, dry and weigh. Add to this weight 0.0025
gram for each 10 ml. of 90% alcohol used in the crystallization and washing
if done at 15 C.; if done at 20 add 0.0045 gram for each 10 ml. The meltingpoint of arachidic acid thus obtained is between 71 and 72 C. Twenty times
the weight of arachidic acid will give the approximate amount of peanut oil
present. No examination for adulterants in olive oil is complete without
making the test for peanut oil. Arachidic acid has a charact~ristic structure
and can be detected by the microscope.
Bach's Test for Rapeseed OiL-According to O. Bach,132 the acids obtained
from rapeseed oil a~completely insoluble in David's alcoholic acetic acid,
in the proportion of 1 to 15, by volume; those from cottonseed, peanut, sesame,
and sunflower oil dissolve on heating. Those from the last oil separate as a
granular precipitate at 15, while from the other three they gelatinize. The
acids from olive oil are completely soluble at the ordinary temperature.
David's acid js made by mixing 22 ml. of 50% acetic acid (by volume) with
30 ml. of alcol;lOl, sp.gr. 0.817, 92.07% (by weight).
NOTE.-The author has found that Bach's observation cannot be implicitly relied
upon, as some rapeseed oils yield acids which are soluble in David's mixture.

Liebermann-Storch Test for Rosin Oil.-One or 2 ml. of the oil are shaken
with an equal quantity of acetic anhydride and gently warmed. When cool
the acetic anhydride is pipetted off and tested by the addition of 1 drop of concentrated sulfuric acid. A fine violet color is produced in the presence of rosin
oil. Tung oil and also cholesterol, which is contained in the animal fats,
,produce a similar coloration; the latter can be removed by saponifying the oil
as completely as possible and shaking out the somewhat dilute soap solution
with ether or petroleum ether. The soap solution is then acidified, setting free
the fatty acids, and these treated with acetic anhydride as if they were the oil.
Baudouin's, or Camoin's Test for Sesame Oil)33
Dissolve 0.1 gram of finely powdered sugar in 10 ml. of hydrochloric acid
(E1p.gr. 1.18), add 10 ml. of the oil to be tested, shake thoroughly for 1 minute
and allow to stand for 10 minutes. In the presence of even a very small
admixture of sesame oil, the aqueous solution is colored crimson. It should
be observed that some olive oils, especially those of African or Spanish origin,
give pink or crimson colors which can be readily differentiated from that due
to sesame oil by applying the following modification of the Villavecchia method.
132
133

Allen, "Commercial Organic AnalysiS," 2, pt. 1, 128, 1899.

Official.
,

1782

FIXED OILS, FATS AND WAXES

Villavecchia Test 134

Add 2 ml. of furfural'to 100 ml. of 95 per cent klcohol by volume and mix
thoroughly 0.1 m!. of tqis solution with 10 m!. of hyprqchloric acid (sp.gr. 1.18)
and 10 ro!. of the oil to be tested by shaking them !together for 7.i of a mjnute.
Allow the mixture to stand 10 minutes, observe the'color, add 10 m!. of water,
shake and again observe the color.
I

The furfurol may be replaced by 0.1 g. of sugar in 10 m!. Hel, sp.gr. 1.18,
and twice the quantity of oil used.
Milliau 135 saponifies as in the Bechi test and dries the acids at 105 0 and
tests them instead of the oil. Lewkowitsch 136 states that this is a needless
complication. Da Silva 131 states that this test has given colors with certain
Portuguese olive oils; also those of Bari, Brindisi and Lecce. Kreis 138 states
that the active or color-giving constituent is probably phenolic in its nature.
The reaction is given by other substances,139 as vanillin, oil of cloves, and
cinnamon; this should be borne in mind in testing oils which have been extracted from confectionery. Rancid fats prevent the coloration; it can, however, be brought about even in rancid fats by the addition of
an equal quantity of cotton-seed oil.l 40
Free Acid Test. 141,-About 10 grams of oil are weighed
(to
centigrams) into a '250-ml. Erlenmeyer flask, 60 ml. of
A
neutral alcohol (Reagents) added, the mixture warmed to
the boiling point, 'Shaken thoroughly, and titrated with
N/6 potassium hydrate, using phenolphthalein, the' flask
being frequently and thoroughly shaken. The result is
conventionally reported in per cent of oleic acid; 1.0 m!.
N/6 KOH is equivalent to 0.047 gram oleip acid. Or it
can be reported like the saponification number, in milligrams of KOH per gram of oil. For mineral oils use alcohol
diluted with its own volume of water.
Spontaneous Combustion T~st. Mackey's Apparatus.
-The apparatus,142 Fig. 21>7, consists of cylindrical copper
water bath 7 in. high and 4 in. in diameter (inside measurements), surrounded with a: Yz~in. water-jacket. The cover
is packed with asbestos and carries the draft tubes A and B,
Yz in. in diameter and 6 in. long, which cause a current of
Mad:~';~ppa- air tt;> be sucked down B and up A, 'thus_ensuring a circuratus.
lation of air in the apparatus: C is a cylinder made of
!,
24-mesh wire gaUlle 6 in. high and 1Yz in. in diameter and
supported upon a projection from the bottom of the bath. A thermometer
projects down into the center of the cylindet; if a metal condenser be con134 Ibid., 1893,505; abstr. Analyst, 19,47.
135 J. Am. Chern. Soc., 15, 162, 1893.
136 "Oils, Fats and Waxes."
137 J. Soc. Chern. Ind., 17, 275, 1898.
138 Chern. Ztg., 27, 316, 1903.
139 Gerber, Analyst, 32, 90, 1907.
140 Lauf! and Hinsmann, Chern. Ztg., 31, 1023,1908.. :__
141 Tech. Paper 323A.
_ -' ~ _ .
142 Mackey, J. Soc. Chern. Ind., 15, 90, 1896; Gill, ibid., 26, 185, 1907.

FIXED OILS, FATS AND WAXES

1783

nected to tb_fl. _water bath it can be used indefinitely without refilling and without danger of 'burning out.
Seven grams of ordinary bleached cotton wadding or " absorbent cotton"
are weighed out in a porcelain dish or on a watch-glass, and 14 grams of the oil
to be tested poured upon the cotton and thoroughly worked into it, care being
taken to replace any oil that is lost. The cotton is then placed in the cylinder,
packed about the thermometer so that it occupies the -upper 4~ in. of the
cylinder, and put into the boiling water bath. Afterthe expiration of an hour,
the bath having been kept in active. ebullition, the temperature is read. Any
oil which shows a temperature exceeding 100 0 C. in one hour, or 200 0 C. in
two hours, should be regarded as a dangerous oil, or liable to produce spon- .
t.aneous combustion. The f,.Qllowing tables show the results obtained in uSIng
'\7)'
cl
this apparatus. 143
:, Temperature

C. in

,Oil
1)4 hrs.

172

100
177-242
101-115

101

1 hr.

Olive (neutral) ............... ' ...... .

Cotton-seed ......... , ....... '........ .
Elaine, ., ............ , .... , ......... .
Olive fatty acids .............. , .. , ... .

97-98
112-128
98-103
102-114

hrs.

194-282
102-191
196

Other values obtained were:

Oil

Temp.

C.

Time,
Minutes

Iodine,
No.

Olive ................... , ...........

Lard ... ......................... ..
Oleic acid ... , ......................
Cotton-seed. '... .....................
Linseed. ,
25 Paraffi~.. ',' .' : : : : : : : : : : : : : : : : : : : :

234
234
158
234
234
97

130
75
188
70
65
135

85.4
75.2
60.5
108.9
168.1
16.2

Free Acid.
per cent

5.3

Trace
Neutral
Neutral

Besides being used for testing oils it can be applied to testing other mate,,..ials, oily waste, sawdust, or any mixtures suspected of causing spontaneous
combustion.
,
" The results 144 of the greatest practical value .obtained in the use of this
apparatus have been, first, determining the cause Memes; and, second, determining the degree of safety of the various oils used in manufacturing. Mineral
oil, as is well known, is not liable to spontaneous combustion; and a certain
percentage of animal or vegetable oil may be addeM to mineral oil without
materially increasing the danger under ordinary circumstances. This percentage varies according to the oil; with neat's-foot and first quality lard oil some
50 to 60% may be used, with cotton-seed not over 25% is allowable. The
143 Small percentages (0.5-2.0) of iron, sodium, magnesium or zinc salts 'in "red" or
olive oil may cause a degree of heating which may be dangerous. Gill, Ind. and Eng.
Chern., 16,23, 1924.
144 Richards, Tech. Quarterly, 4, 346,1891.

178~

FIXED OILS, FATS AND WAXES

claim so often made for so-called' safe' oils, said' to hav~ been changed by
special and secret processes of refining so as to be nQ longer dangerous, is easilY
exposed by this test."
Drying Test Upon Glass. 14 2_A few drops of oil are brought upon a glass
plate inclined at about 30 0 from the horizontal. A test of the oil is made from
time to time by touching it with the fingers, the time at which it does not soil
them being noted as the point when it is dry. Good oil should dry in three dtiys.
Archbutt 146 makes this test as follows: A piece of polished plate-glass 7 cm.
square by 4 mm. thick is cleaned and counterpoised on the balance; it is then
heated for an hour at 200 0 C. in an air bath to thoroughly dry it. It is taken
out, laid on a non-conductor, allowed to cool for three or four minutes, and
, the hot glass thinly painted with the oil to be tested by means of a camel's-hair
brush. When the glass is cold it is weighed. and sufficient oil added to make
it up to 0.1 gram. Two glasses are coated with the sample and two with a
standard oil, all placed on a level surface in a . large air bath at 50 C. and
heated for nine hours; one set of plates is withdrawn, cooled, and tested by the
finger. Good raw linseed is tacky, when tested by the finger when cold, in
nine hours and dry in twelve; corn oil is practically dry in fifteen hours, though
slightly tacky; cottonseed, partially dry in eighteen hours and fully dry in
twenty-one. Refined rape oil dried in forty-eight hours, and olive oil was
sticky after thirteen days.
Titer Test.-Under this rather misleading title is expressed the solidification
point of the fatty acids derived from a fat or oil; it has nothing at all to do with
titration, as might be expected. The test is extensively used for the evaluation
of fats, and according to the method provisionally adopted by the Association
of Official Agricultural Chemists is carried out as follows: 147
(a) Standard Thermometer. The thermometer must be graduated in tenth
degrees from 10 to 60, with a zero mark, and have an auxiliarr reservoir at
the upper end, also one between the zero mark and the 10 mark. The cavity
in the capillary tube between the zero mark and the 10 mark must be at least
1 cm. below the 10 mark, the 10 mark to be aboujt 3 or 4 cm. above the bulb,
the length of the thermometer being about 15 lin. over all. The thermometer is
annealed for 75 hours at 45.0 C., and the bulb is of Jena normal 16 '11 glass,
moderately thin, so that the thermometer will be quick acting. The bulb is
about 3 cm. long and 6 mm. in diameter. The stem of the thermometer is
6 mm. in diameter and made of the best thermometer tubing, with scale etche~,
on the stem, the graduation to be clear-cut and distinct; but quite fine.
(b) Determination. Saponify 75 grams of fat in a metal dish with 60 ml.
of 30% sodium hydroxide (36 Baume) and 75 ml. of 95% (by volume) alcohol
or 120 m!. of water. Boil to dryness, with constant stirring to prevent scorching, over a very low flame or over an iron or asbestos plate. Dissolve the dry
soap in a liter of boiling water, and if alcohol ~as been used, boil for forty
minutes in order to remove it, adding sufficient I water to replace that lost in
boiling. Add 100 ml. of 30% sulfuric acid (25 Baume) to free the fatty acids,
and boil until they form a clear, transparent layer. Wash with boiling water
until free from sulfuric acid, collect in a small beaker, and place on the steam
J

Arosel, J. Soc. Chern. Ind., IS, 222, 1896.

Id., 18, 347, 1899.
. ..../ ~
1<7 U. S. Dept. of Agriculture, Bureau of Chemistry Bulletin No. 107, p. 135, 1907.
1<5

1<6

FIXED OILS, FATS AND WAXES

1785

bath until the water has settled and the fatty acids are clear; then decant them
into a dry beaker, filter, using the hot-water funnel, and dry twenty minutes
at 100 C. When dried, cool the fatty. acids to 15 or 20 C. above the expected
titer and transfer to the titer tube, which is 25 mm. in diameter and 100 mm.
in length (1 by 4 in.) and made of glass about 1 mm. in thickness. Place in a
16-oz. salt-mouth bottle of clear&ass, about 70 mm. in diameter and 150 mm.
high (2.8 by 6 in.), fitted with a Mk, which is perforated so as to hold the tube
rigidly when in position. Suspend the thermometer, graduated to 0.1 C., so
that it can be used as a stirrer, and stir the mass slowly until the mercur~
remains stationary for thirty seconds. Then allow the thermometer to hang
quietly, with the bulb in the center of the mass, and observe the rise of the'
mercury. The highest point to which it rises is recorded as the' titer of the
fatty acids.
Test the fatty acids for complete saponification as follows:
Place 3 m!. in a test-tube and add 15 m!. of alcohol (95% by volume).
Bring the mixture to a boil and add an equal volume of ammonium hydroxide
(0.96.sp.gr.). A clear solution should result, turbidity indicating unsaponified
fat. The titer must be made at about 20 C. for all fats having a titer above
30 C. and at 10 C. below the titer for all other fats.
REFERENCES
Hefter, G. Technologie der Fette, Oele, und Wachsarten des Pflanzen und Tierreichs. 4 volumes, 1906.
Ubbelohde, L. Chemie, Analyse und Gewinnung der Oele, Fette und Wachse.
4 volumes, 1908.

EDmLE FATS

These include butter, lard and hardened oils.

Butter is examined for water, fat, ash, curd, and salt; these are usually
present in the following proportions:

Fat .................................. .
Water ................................ .
Salt .... , ........ , ........ ' ........... .
Curd ................................. .

Per cent

Average per cent

78 -90
5 -20
0.4-15
0.1- 5

82
12
5
1

These are determined as follows: weigh about 2 grams of butter into a

platinum Gooch crucible half filled with ignited fibrous asbestos, and dry it at
100 C. to constant weight. The loss is the amount of water. Dissolve out

1786

FIXED OILS, FATS AND WAXES

the fat by repeated treatment with petroleum ether and again dry to constant
weight. The loss represents the amount of fat. Ignite the crucible with a low
flame or in a muffie, being careful not to volatilize tl;te salt, until a light-gray ash
is obtained. The loss represents curd and the re~~due ash. By extraction of
the ash with water, and neutralization witli calcium carbonate, the salt can be
titrated with silver nitrate.
"
Examination of the Fat.-Butter is adulterated with oleomargarin~, renovated butter, and cocoanut oil. The first may be detected by testing for
'cottonseed or sesame oil either by the color tests or by the index of refraction;
this at 25 is for butter 1.459-1.462, for oleo. 1.465-1.470. Owing to the fact
that, butter. contains a l~rge per cent of volatile fatty acids (butyric, caproic,
caprylic/and capric acids, in aU about 8%), adulterants may be detected by.
determining the aJ,llount of these. The process usually employed is that of
Reichert modifi!;ld by Meissl.
Five grams of the clear fa~; filtered through absorbent cotton, are weighed
into a 250-ml. round-bottomed flask and saponified by 2 ml. potassium hydroxide 1 : 1 and 10 ml.iof '95% alcohol, under a return flow condenser for
twenty-five minutes. Tlie'alcohol is rapidly evaporated off on the water bath
until no odor of alcohol is'perceptible. Add slowly 160 ml. of recently boiled
distilled water which has been cooled to 50 or 60; warm the flask until a
clear solution of the soap is obtained. Cool to about 60 and add 8 ml. sulfuric
acid 1 : 4 to liberate the fatty acids. / Drop into the flask two bits of ~umice
(about the size j of peas) which have been heated and quenched in water, and
tie in a well-fitting cork; warm the flask until the fatty acids have melted and
are floating on the liquid. Cool to about 60 and attach the flask to a condenser, using a trap to prevent the sulfuric acid from being mechanically
carried over; 110 ml. are distilled into a graduated flask in as nearly' thirty
minutes as possible. Thoroughly mix the distillate, pour through a dry filter,
and titrate 100 ml. with N/l0 NaOH, using phenolphthalein as an indicator.
Multiply the ml. of alkali by 1.1 and calculate them to 5 grams of fat. The
Reichert-Meissl value for butter is from 24 to 34{ the average is about 28.8;
cocoanut oil gives 6-8 and other fats less than 1.
The procedure is a conventional one and should be followed as exactly as
possible. Cocoanut and other vegetable oils WQuld be shown by the fact that
the unsaponifiable matter would conta~n phytosterol; also by the Polenske
number.14s Renovated butter is best shown by the II'Spoon~ or Foam Test."
This consists in melting a third of a teaspoonful of the sample -in a tablespoon
over a small flame and stirring with a match. Increase the heat until the fat
boils briskly, stirring thoroughly several times. Oleomargarine and renovated
butter boil noisily, sputtering like a mixture 'of grease and water and produce
no foam. Butter boils with less noise and much foam sometimes rising over
the sides of the spoon. The 'pieces of curd in bJtter are much smaller than in
either of the others.
Preservatives, benzoic, boric and salicylic acids, may be examined according
to the procedure given in Woodman and Norton, " Air, Water and Food," pp.
154 and 196.
,- '
148

Z. Nahr. Gerlussm., 7, 193, 1904, also Gill, liOn-Analysis."

FIXED OILS, FATS Al'fD WAXES

1787

Color may be dE;Jtected according to Allen, " COlI).lllercial Organic Analysis,"

4th Ed., Vol. II, or Leach's" Food Analysis."
Lard is adulterated with water, 25% being added in some cases, with cottonseed oil orstearine and beef stearine.
Water is determined as with butter; cottonseed oil or stearine by the usual
tests. ,It should be borne in mind, however, that hogs fed on cottonseed meal
yield a lard which will give the Halphen test as strongly as if it contained 25%
of the oil. The iodine number and the presence of phytosterol will confirm this
test; the iodine number varies widely according to the source of the fat, but
in general it may be said it should be between 46 and 66.
Beef stearine is very difficult if not impossible of detection. For this, reference may be had to Lewkowitsch, 5th Ed., Vol. II.

HARDENED OILS
As the name denotes, these are oils which have been changed to more or
less solid fats by the addition of hydrogen, in the presence of a catalyst, usually
a compound of nickel. This betrays their presence and may be tested for as
follows: 149
Ten grams of the fat are heated on the water bath with 10 ml. of hydrochloric acid (sp.gr. 1.12) with frequent shaking for two or three hours. The
fat is removed by filtering through a wet filter, receiving the filtrate in a porcelain dish; after partial evaporation of the filtrate 2 or 3 ml. of concentrated
nitric acid are added and the evaporation continued to dryness to ensure the
destruction of the organic matter. The residue is dissolved iIi a few m!. of
distilled water, a 'few drops of a 1% solution of dimethylglyoxime in alcohol
added, and a few drops of dilute ammonia. The presence of nickel is shown
by the appearance of the red-colored nickel dimethylglyoxime. The amount
of nickel can be determined colorimetrically by comparison with solutions
containing known quantities.
The quantity of nickel is very minute, not as much as the fats take up when
cooked in nickel or nickeled dishes, and need cause no apprehension.
Hydrogenation destroys all the characteristics, particularly the color tests,
by which the different oils may be sometimes detected.
149

Kerr, J. Ind. and Eng. Chern., 6, 207, 1914.

1788

FIXED OILS, FATS AND WAXES

WAXES

These, as will be remembered, contain no glycerine; the tests applied to

them are the same as to the oils. The characteristics of the more commonly
occurring waxes are given in the table, p. 1797; sperl4 oil, wnich is really a liquid
wax, is included among the oils.
'

MISC;E~LANEOUS

OILS AND LUBRICANTS

PROPERTIES OF SOME OF T,HE MINERAL OILS

\

Oil

Black ......... , ..........

Ice machine ..............
Crank case ...............
Transformer ..............
Turbine ..................
Spindle ...................
Loom ....................
Engine ...................
Cylinder ......... _.......
Cylinder .................

Specific
Gravity,
Deg, Baume
at 60 F.

29
26-27
26-27

30
30-35
28
27-30
23-25
26-28

Flash

Viscosity
(Saybolt),
at 70 F.

Point,
OF.
/

325
325--360
455
340-380
420
320-390
360
410
525
400-575

100-120
60-100
100
400
160
58-156
203
190-210
200-300*

Cold Test,
Deg, Fahr,

5--15
Q-4
-

25

* At 212 F.
0

Belt Dressings are (1) mixtures of fats, waxes, degras or tallow with castor
or fatty oils; (2) vulcanized corn or cottonseed oil thinned with naphtha;
(3) preparations containing wood tar; or (4) preparations containing rosin,
which is undesirable. Black oils, car oils, well oil or reduced pils are crude oils
from which the naphthas and burning oils have been separated 'by distillation.
Qrank-case oils are pure mineral oils which emulsify but little with water.
Mming-mach~ne or soluble oils are lard, sulfonated oils or mineral oils held in
suspension in water by SO!lpS or alkalies, as borax or soda; the soaps used are.
either ammonium, sodium or potassium with resin, oleic or sulfofatty acids.
Rosin oils are obtained by distilling or " running:" rosin, each distillate being
called a "run" and numbered according to the times it has been distilled.
They oxidize quite rapidly and should not be used as lubricants except as soaps
in lubricating greases. Screw-cutting oils are often mixtures of 27 0 Be. paraffin
and 25% fatty 0:.1, preferably cottonseed, although lard oil' was formerly used.
Stainless oils are spindle or loom oils mixed wjthfatty oils-lard or neatsfoot.
Transformer ofls should be either pure mineral 'Or rosin oils and as free as

FIXED OILS, FATS AND' WAXES

1789

possible.fl'oII,l water, acid, alkali and sulfur. Turbine oils should be of excellent
quality, free from acid and tendency to resinify, and low in sulfur. Watch
oil is obtained from the porpoise, dolphin, or blackfish, where it exists in cavities
in the jaw and in the brain or " melon" of the fish. Lubricating greases are
mixtures of soaps of palm oil, tallow or rosin oil (with lime or soda as bases)
with various oils or fats such as rosin, tallow or mineral oil. The best are those
made from tallow by saponification with caustic soda. They may also contain
finely powdered talc or graphite. Non-fluid oils are oils or their greases stiffened with" oil pulp" or " dope," i.e., aluminum oleate or palmitate.
" The source, preparation and uses of the various oils "and greases are described in Rogers and Aubert's Industrial Chemistry, Chapters XXII, XXIV,
XXV and XXVII.
For the guidance of the analyst, the characteristics of the more frequently
occurring oils are given, the usual figures being given in italic.
The animal'and vegetable oils may be classified into
Drying.-Linseed, Chinese wOQd, poppyseed, sunflower and menhaden.
Semi-drying.-Corn, cottonseed, sesame, rape, black mustard and horse.
Non-drying.-Castor, almond, peanut, olive, cocoanut, palm, seal, cod-liver,
elaine, lard, neatsfoot, tallow, sperm and whale.,
REFERENCES

Lewkowitsch, "Technology and Analysis of Oils, Fats and Waxes." 3 vok

Gill, "A Short Handbook of Oil Analysis."
Battle, "Industrial Oil Engineerin?,"
Lockhart, "American Lubricants.'

1790

FIXED .OILS, FATS AND WAXES

.

,'0
."

:'"

:al

."

'.!ol
.C'

UInllIJXWl'mqdJIlS

:~

mnunx-em
'''''''I uOli"lllllG

mnunx'B1II

IU""

JOe! 06

r-~

. .:~
",
V
.lQ

0
.~

mmmmw
IlloO jod 9

.",
.",

FIXED OILS, FATS AND WAXES

til

Q)

til

0$
Q)

C5

~t-OOOeo")lt:)i:'<':I
t.....j~e-4c-iet:ie-i
c..
] .....
"g_.g.g.g .g.g
.' ...
~

11

~~~-.c.<t:I".'H~

t; ..... >-i ...... P""OI~'P""I

c..

1791

1792

FIXED OILS, FATS AND WAXES

J

'"

;:!

.~

g
....I

C,.)

~u!punodmoo

C'I

j, :

0>
,....,

,
,....,
:OOlQlQOOCl.QOC OOOlQOet:l
._C'l'"<1'lQ~_I!'1""':-lQt-OOlQt-o

d . . . . . . . . . . ,....j,....ie-:i 0

::q
C)
~
~

:;E

:1l....: :.......
: : : : : : :.
,

:'~~~~~~~~~~

~
~

:~

w
......

...

: : : :,: : : : :

rti

...
P

......

Ul

bIl

'.;:I

..,cd

:3

'luaa JOO

gs P"l nl!P

10100

C)

......

'W '.1 ,'9 ''1'

P"lnnpun 10rOO 'W '.1 'S ''1'

r.;

mnm!"uw

~
"..

mnlII!u!w

r.;

ummp:uW

w
E-<

: : : : : ~: : : :;~~~~~~~~~~~~~~~
:

..

..

'0000000000000000

:=
0

Ct
...... ';::
~

mUJ~ lad HO:ll

'j]W mnUI!XWI 'l!PF>V

. .. ..
..

..
..
. .
..........
, , , ..... , . .
..........

- - -

umW!U!W

I=Q

...

0......
E-<
~

C)

...
......

......

C)
~

p.,

-.

t-t-t-ooOOt-t-t-oooot-t-t-OO~t-t-t-oooo

1:,

;::

_~~oo

. :. :.

:0

:~

FIXED OILS, FATS AND WAXES

~
IV

.j

8
~C'I

llulPunodmoo

0>
.....
oo~

.....
p;:

:o.t:)o.t:)

:~...t(

Q
~

'00
'00

.",,,,

~.

w
....

...
~

l""l UO!"""OQ

u.S

...:l

~
... <5'"
A
....

w
Z

Eo<
~

.....
~

I=Q

...:l
~

...

w
Z

0
.....
Eo<
~

.............
Q

mnm!X'BW mOd

b.O

:
oS

u
G .;:::
.....
,Q
...:l ::I

tm!Jlj lOO HO:lI

'llm mn1l1!X1lUl 'AllPl'V

...:I

'luaa .lad
gg J"llnl!P 'lOIOO 'W .J.

J"llnl!pun 'lOIO.W 'J.

s .\
s 'V

:.~

1793

FIXED OILS, FATS AND W~S

1794

CHARACTERISTICS OF THE FATTY ACIDS FItOM,SOME OILS

Oil.

Refr. Ind. at
60 C.

M.pt. C.

Solidifn. Pt.
\ DC.

Iodine
Per cent.

Almond ..................

1.4461

13-14

19-11.8

93.5-96.5

Castor ...................

1.4546

13

86-88

40-43.8

31.2

145-159.4

24-27

15-20

8.4-9

17-18 (titer)

164-111

Chinese wood .............

Cocoanut .................

1.4295

Codliver, Medicinal. .......

Corn .....................

18-21.6

14-16

113-125

Cottonseed ...............

1.446

34-40

32-35

105-112

Horse ....................

37.5-39.5

33.6-37.7

84-87

Lard .....................

35
13-17

179 209.8

6-8

108-126.5

Elaine ......... i

.....

Linseed ............ , .....

17-24

1.4546
/

Menhaden ................ No figures a v ailable.

Black Mustard ............. 1.4665at20
Neatsfoot ................
Olive ...... ~ .............
Palm ..... , ..............

1.4410

9-17
28.5-29.8

16-26.5
(titer)

62-77

17-24.6

86-90

19-31
47.7-50

53.3

/36-46
I

44-

Peanut ...................

1.4461

27-35 '

22-32.5

95--103

Poppyseed ...............-.

1.4506

20-21

16.5

139

Rape ....................

1.4491

16-22

16-,JB.:__

Seal. ...................

14-33

Sesame ...................

1.4461

23-32

Sperm ......... .- .........

13-21

Sunflower. : ..............

1.4531

17-24

14-27

13-17

99-105
" 186-202

18-26

109-112

16

83-99

17-18

124

23-24

130-132

Tallow ...................
Whale ...................

--

\'

FIXED OILS, FATS AND WAXES

1795

ci

:;l
.~

o
U

'"~

...

,~

'"'"...
<0

......

""
.J.'"

60)

'"o

...a.

....

<;

11o

:;;:

"o

~
0

"''''
'"<0...
'"
<0

...

.,'"
'"...

.,o

'""" I~~
........
.,.I"'''''" ;:'" '".
'"o o

.'"

'il

,,6
o

"o
U
o

''""

00

'.."

'il

::s

1796

FIXED OILS, FATS AND WAXES

I I ~I
I I ~I

I :I :

I II:I I II: II

oi

.~

".."
]

~"

I~
I~
I

I~
U",)

t-

.....I

t-

FIXED OILS, FATS AND WAXES

.
o

:~
o

(.)

100

I~

o
co

C':)

.8;

t-

10>

10>

co

0>

20>

0>

0>

t"-

10>

~
o

1797

1798

FIXED OILS, FATS AlfD WAXES

DEGREES-ENGLER
I

Time Engler. Redwood, Redwood Adm. Saybolt Universal, Fural, Barbey

600 8001,,000

-4

2 .. 000 3 .. 000 4.000 6,,000

10,,000

Degrees.Engier

6 7 8 910

~OO

200

400

1I~
v
1000
900
80C
70 a
60 a

'Q.

1"'-

.....

'/

1..1

[,1 0,000

5,000

~</<o

4,000

'ti/" /
",'" ' /
I_cO-Eo J '

3,000

~~~~
).)

2,000

~\~e
1;

:!c}'

i~~ 00

/'

~e'

_~ ,e
'V~

.....

Q)

..

~~ ~

l,DDO

70

900
\:800
700

60

600

50

_~ 40

,8
..

IA:: 2 5,000

a,ooo

/,.x.",

I//.s"~

SS~100

.~

1/

7,000

'/I/lott <;v

<le'/

80,,000

6,000

(j",
~ '~20 a
';;; ~

II

aoo

,~o,ooo

30004000 6000 8000

9,000

400

..Q

2000

., '(\

50 a

0:

20.. 000 aOI_ooo

600 800 1000

7V

'/

"/

30
20

If
,

I
I

.'
KY

'\
//
7fg -7 ['1'1.
I
I

J,'

10
9
8
7
6

Distilled water at 20C.

G/~cerol mixtures at 20 C.
o Refined castor 011
'" Mineraloils
@ Alcohol, 40 %

.....

,,

II,

,
I

+ D.ta 'rom Ar<hbutt & Deeley

A

~,
,>

I
I

,-,

/i

I
o

,
,
,
,,
,

I
20

30

~F
40

200

".....
.....

300 400

SPO 8001000

2000

Time Engler. Redwood, Redwood Adm. Saybolt Universal, Furol. Barbey

2

5 6 7

9 10

20

30

"

"-

'0

lA

5060708090100

40 50 60

Degrees-Engler

80

~OO

2.00

:-.....

3COO 400060008000

aoo

400

600

80Q

FIG. 258.-Viscosimeter Conversion Chart 160

.(Directions for Use of Chart),
OBJECT: To find the viscosity reading of a particular oil on any of the standard
viscosimeters when its reading on one of the viscosimeters has been determined by
experiment at the same temperature.
SCALES: There are two sets of scales which must not be confused with each other.
X: The scales at bottom and left hand side go together. They/are the ones regu._
larly used in practice applying to all ordinary oils.
Y: The scales at top and on right hand side go ~ogether. They are rarely used,
applying only to very viscous products beyond range of other scales.
150 Power (1922), 377.

FIXED OILS, FATS AND WAXES

1799

NOTE.-There are two lines of scales at the bottom: (a) The lower scale applying
only to degrees Engler. (b) The upper scale applying to all other readings, including
Engler-time. In simply changing from one viscosimeter to another the scales at left
and right are not used. T~e scales at top must never be used in connection with those
::It the bottom.
APPLICATION OF CHARTS: X: For ordinary cases.-GIVEN: Saybolt Universal Viscosity = 120". PESIRED: Gase (a) Barbey reading. Case (b) EnglerDegree reading. PROCEDURE: Find point on bottom scales for Saybolt Universal
corresponding to 120" (A). Follow up vertically to curve marked Saybolt Universal
(the known instrument) (B). (a) To get Barbey reading: Follow horizontally to
curve marked Barbey (C). Drop down to bottom scale for Barbey and read (C).
Reading=265 deg. (b). To get Engler-Degree reading: Follow horizontally to curve
marked Degrees-Engler (E). Drop down to lower of the two bottom scales for EnglerDegrees and read (F). Reading=3.62 deg., i.e., 120" Saybolt=Barbey 265 deg.
= Engler 3.62.
, Y: For very viscous fluids beyond range of other scales. GIVEN: Engler-Degrees
=1100 deg. DESIRED: Saybolt Furol reading. PROCEDURE: Find point on
upper scales under Degrees-Engler corresponding to 1,100 deg. Follow vertically
down to Degrees-Engler, curve. Follow horizontally to curve tfRrked Furol. Rise
vertically .to scales under Furol and read. Reading = 3800", i1!, 1,100 deg. Engler
=,3,800" Furol.
MULTIPLYING FACTORS TO REDUCE SAYBOLT TIMES TO ENGLER NUMBERS OR TO
REDWOOD TIMES I
Saybolt Times
Seconds.

28
30
32
34
36
38
40
42
44
46

~~

55
60
65
70

Factor Saybolt
Time to Engler
Number.

0.0357

.0352
.0346

.0342
.0337
.0334
.0330
.0327
.0323
.0320
.0317
.0314
.0308
.0302
.0297
0.0293

Factor Saybolt Time

to Redwood
Time.

Saybolt Times
Seconds.

75
80
85
90
95
100
110
120
130
140
160
180
200

0.95
.95
.94
.94
.94
.93
.93
.92
.92
.91
.91

.90

.90
.89
.88
0.87

,0,

1800

Factor Say bolt

Time to Engler
Number.

0.0289
.0286
.0284

.0282
.0280
.0279
.0276
.0274
.0272
.0271
.0269
.0268
.0267
0

0.

0.0267

Factor Saybolt Time

to Redwood
Time.

0.86
.86
.86
.85
.85
.85
.85
.84
.84
.84
.84
.84

.84

.0

0.84

The Engler number is the quotient of Engler Time divided by the water value
of the instrument at 20 C. in seconds.
'
1

Reagents
The reagents used in oil analysis are few and easily obtained, A list and
their method of preparation is here given. l5l
Acetic Acid, Glacial. C.P., 99.5% pure. The determination of its strength
should be made by titration and not by specific gravity, as the 9'8% and 80%
161

Proc. Am. Soc. Test. Mat., 15, 1, 288, 1915.

1800

FIXED OILS, FATS AND WAXES

acid have the same specific gravity, 1.067. The determination of the meltingpoint gives results equally good with those obtained by titration and requires
less,time. 152 It is made after the manner of the" titer test" (p. 1784), the
tube being half filled, chilled to 10 to 11C., and further chilled by placing
the outside bottle in ice-water; the temperature of the! super-cooled acid rises
to its melting-point, where it remains sta'tionary for SOme time. The meltingpoints of acids of various strengths are as follows:
100%, 16.75 C.; 99.5%, 15.65; 99%, 14.8.
For Hanus's solution it must not reduce potassium bichromate and sulfuric acid.
Acetic Anhydride.
Alcohol: Commercial" Cologne Spirits." For the preparation of alcohol
free from aldehyde for alcoholic potash, cologne spirits are treated with silver
oxide as follows: 1Yz grams of silver nitrate are dissolved in 3 ml. of water,
added to 1 liter of alcohol and thoroughly shaken; 3 grams of potassium hydrate
are dissolved in 15 ml. warm alcohol and, after cooling, added to the alcoholic
silver nitrate and thoroughly shaken again, best in Ii tall bottle or cylinder.
The silver oxide is allowed to settle, the clear liquid siphoned off and distilled;
a few bits of pumice, prepared by igniting it and immediately quenching under
water, being added to prevent bumping. AlcollOl for use in tbe free acid deteI~
mination is prepared by placing 10 to 15 grams of dry sodium carbonate in
the reagent bottle, taking care to filter it before use.
Alcohol, Amyl.
/
Bromine. The commercial article; also a N 13 solution, made by dissolving
26.6 grams bromine in 1 liter carbon tetrachloride.
Calcium Chloride. The dry and also the crystallized salt.
Calcium Sulfate. Plaster of Paris.
Carbon Tetrachloride. Baker and Adamson's C.P. or Kahlbaum's " Tetrachlorkohlenstoff."
Chloroform. Squibb's, U. S. P.
Copper. Copper turnings or clippings, used for the generation of nitric
oxide.
I
Copper Wire. Cut in pieces of 0,3 to 0.5 graiu.
Ether. Squibb's, U. S. P.
Gasoline. Gasoline, 86 Baume.
I
Hydrochloric Acid, C.P.-Specific gravity 1.2: For N/~. HCl dilute 39 ml.
of the above acid to 1 liter' and standardize.
Iodine Solution. Fifty grams of iodine to 1 liter of alcohol. For Hanus's
solution dissolve by warming 13.2 grams iodine in 1 liter glacial" acetic acid;
cool and add 3 1111. of bromine.
,
Lead Acetate. One hundred grams of the salt to 1 liter.
Lacmoid. Three grams per liter of dilute alcohol.
Lacmoid Paper. Unsized paper dipped in above isolution.
Litmus Paper.
#
Mercuric Chloride. Sixty grams of the salt to 1 liter of alcohol.
Nitric Acid. Specific gravity 1.34.
I
Phenolphthalein. One gram of the substance to 500 m!. of alcohol.
_,-'
152 McIlhiney, et al., J. Am. Chern. Soc., 29, 1224, i907:

>'

lfIXED OILS, FATS AND WAXES

#I

1801

M eta-Pho8phoric Acid. A saturated solution of the commercial "stick

phosphoric acid" in absolute alcohol.
Pota88ium Bichromate. Dissolve 3.8633 grams of the C.P. salt in 1 liter of
water; 1 ml._is equivalent to 0.01 gram of iodine. The solution should be
tested against iron wire containing a known percentage of iron.
Pota88ium Hydroxide. N/2: Dissolve 30 grams of "potash by alcohol"
in 1 liter of alcohol. N /6: Dissolve 10 ,grams of "potash by alcohol" in 1 liter
of water and dilute to proper strength. The solution should be protected by
"stick potash" from the carbon dioxide in the air. Ten p~ cent: Dissolve
100 grams of "stick potash" in 1100 ml. of alcohol.
Pot.a88ium Iodate. A 2% solution.
Potas8ium Iodide. One hundred fifty grams of the commercial salt are
dissolved in water and made up to 1 liter of water. This should be free from
iodate, shown by yielding no coloration when acidified with strong HCl.
Silver Nitrate. Thirty grams to 1 liter+OA ml. HN0 3
Sodium.
'
. Sodium Chloride. Ordinary "coarse fine" salt for freezing mixtures.
Sodium Hydrate. 36 Baume. Dissolve 300 grams of caustic soda in
1 liter of water. .
Sodium N itropru88ide. The commercial salt.
Sodium Thi08ulfate. N/lO: Dissolve 26 grams of "sodium hyposulfite,"
in 1 liter of water.
Starch Solution. Rub up in a mortar 1 gram of potato starch, with 10 to
15 ml. of water, pour this into 200 m!. of water which is boiling actively, and
continue the boiling for a few minutes.
Sugar. Ordinarily granulated sugar.
Sulfur.. A 1.5% solution in carbon bisulfide.
Sulfuric Acid, C.P. This should be at least 99.5% pure, and its strength
be determined by titration, as 100% and 94.3% acid have the same specific
gravity.
Dilute. One part acid to ten parts water.
Nitro8ulfuric Acid, for the Elaidin Test. A liter of sulfuric acid of 46
Baume (1.47 specific gravity) is prepared by diluting 560 ml. commercial sulfuric acid to 1 liter; a few drops of nitric acid are added and nitric oxide (generated from copper and nitric acid) passed in until it is saturated. The acid
is then cooled in ice-water and the gas passed in until it is saturated at 0 C.
This is ca1led Roth's liquid.
The author wishes to acknowledge his indebtedness to Mr. Thomas T. Gray
for his careful review of this chapter. Mr. Gray's broad experience in petroleum products as Chief Chemist of Tidewater Oil Company, makes his criticism
and suggestions of special value.
1.53

Richmond, J, Soc. Chern, Ind" 9, 479, 1890.

1802

FIXED OILS, FATS AND' WAXES

i'
JUNKER'S CALORIMETER FOR LIQlJID FUELS
Directions for W orl:cing

1. Filling the lamp: Loosen the wing screw b, remove the arm with the
gauge and pour about 150-200 cm.3 of fuel into the recipient. Then put the
gauge arm on again and close the openiiIg by again tightening the screw.
2. Setting up: Place the balance by the side of the calorimeter so that th~
burner of the lamp hangs exactly in the center of the combustion chamber of
the calorimeter, after which the lamp with its appendage g is hung from one
arm of the scales and counterbalanced by putting weights on the other scale.
By using a mirror the right position to be obtained can be easily controlled.
The lamp can be turned around the appendage e, when the nut f is loosened. ,

FIG. 258a.

3. Starting: Remove the lamp again from the bala.nce, fill.the cup L beneath the burner heat with spirit and light it.' 'rhen take the hood m an(j
screw the air hose of the air-pump to the valve. When the spirit 9f small cup
is nearly burnt out, press air into the recipient by ;vigorously pushing the air
piston; the air acting upon. the fuel, it will rise into the burner and gasify on
its hot surfa~e. The gas streaming out of the nozzle ignites itself on the spirit
fiame and feeds the ga~ification, even when the spirit is completely burnt out.
If no gas streams out, the nozzle n is obstructe~.i it mfist be cleaned by
means of the special pin included with the appa;.ratUs. An ordinary pin must
not be used as in doing so damage would be done to the nozzle.

FIXED OILS, FATS AND WAXES

1803

4. Regulation: Press as much air into the apparatus that a uniform and
. free burning flame is obtained. (The development of heat should be about
1000 calories per hour, thus about 100 grams of petroleum, benzine or lubricating oil, or 130 grams of spirit must be burnt per hour.)
If the flame is too large, the screw h should be slightly slackened so as to.
let some air escape. When the flame is well regulated, the small wing screw
of the valve is tightened again, the pump is removed and the small hood m is
screwed on the thread.

5. Measuring: Now introduce the burner' into the combustion chamber

beneath the calorimeter and hang the lamp on one arm of the balance as
indicated but not before having filled the calorimeter with water so that it flows
out at the outlet.
After having introduced the lamp, the thermometer at the water outlet
begins to rise. By means of the regulating cock of the calorimeter the water
supply must be regulated thus, that the difference between the inflowing and
outflowing water is about 10-20. After having reached the state of permanency, i.e., when the thermometer does not rise any more, put a small weight
on the scale k so that the arm of the balance goes down. The balance will only
reach its state of equilibrium when a certain amount of fuel has been burnt.
The moment when the pointer of the balance reaches the point of zero,
the hose must be turned over the gauge. At the same time the cup k is charged
with 10-20 grams. After this, begin reading the two thermometers, doing
this preferably at regular intervals. The trapping. of the water and the reading
is interrupted the moment where the lamp again starts rising and the hand
goes again through o. At this point 10-20 grams of fuel have been burnt.
The higher heating value is 'calculated according to
W
H=-T

G '

where H is the heating value in calories per gram,

W the quantity of the water trapped in kilos or liters,
G the quantity of fuel consumed, in grams, ..t\
.
T the difference of temperature between infloMiig and outflowing water.
Heating value of a kilogram = 1000H.
'In order to ascertain the lower heating value, the condensed water flowing
off is caught in a small gauge, the number of milliliters of collected condensed
water obtained from 10 grams of burnt fuel is multiplied by 60 and the number
thus obtained is deducted from the upper heating value of a kilogram of fluidity,
found in the calorimeter.
6. Stopping the calorimeter: Slacken the screw h of the lamp whereby the
pressure of air will cease at once and the flame will be extinguished.
Annotations: The burner head with the small nozzle is designed for fuels
rich in carbon (as petroleum, benzine, etc.), that with the large nozzle for
fuels containing water (spirit). The more carbon, the smaller the bore.
Spirit requires the smallest pressure of air, under 200 millimeters; mineral
and lubricating oil the highest. Vegetable and animal oils must be used for
combustion only if they are completely volatilized at a boiling point of 250

1804

FIXED OILS, FATS AND WAXES

at the highest. Thus, several oils cannot be used, for instance, neat's foot oil,
rape oil, olive oil.
I
Mineral lubricating oils may be used, if they evaporate completely at 250.
If the preliminary warming is not sufficient, or th~ pressure bf the air is too
high, the burning is subject to sudden shocks; by sl~ckening the screw hand
warming up again, this inconvenience will be done away with.
When lubricating oil has been burnt, the lamp must be well cleaned with
petroleum followed by benzine . Then ,he lamp should be soaked with a
quantity of the liquid fuel which is to be tested. Only when this has been
done, the liquid fuel for a new measuring may be poured into it.
There is no danger whatever of explosion, even with benzine, as the flame
will never get back into the recipient, as the pipe and the small nozzle are
always filled with fluid. Naturally benzine should not be poured into the
recipient in the proximity of the burning lamp.

THE DETERMINATION OF TETRAETHYL-LEAD IN GASOLINE

154

Two hundred (200) ml. of the gasoline sample, measured in a volumetric

flask, are chilled in ice water for at lea~t ten minutes and then poured into a
,
500-ml. separatory funnel, supported in a ring stand. 16,:S
Place a chemically clean 600-ml. beaker under the separatory funnel. This
is done at the start to avoid loss in case a drop of the concentrated Pb solution,
subsequently formed, should seep past the stopcock. Add very slowly (dropwise at first) and with gentle agitation, 20 ml. of a solution of 1 part by volume
of C.P. liquid bromine (Br) and 2 parts of carbon tetrachloride (CCl 4). Let
stand with occasional shaking for fifteen minutes. Add ten ml. of concentrated C.P. nitric acid and agitate carefully, avoiding Joss due to gas pressure.
Allow the acid to settle and draw it off into the 600-hll. beaker. Wash down
the sides of the funnel with about 20 ml. of distilled water and draw off the
water into the beaker. The wash water at this point is to facilitate drawing
off the most concentrated Pb solution without. loss through dripping from the
stem of the funnel and without drawing off any of the gasoline layer. Repeat
the acid and water washing as outlined above, agitating_somewhat more vigorously. Make three finfil"water washes of about 50 ml. each ang shake thoroughly with these portions.
The beaker should now contain all the Pb in solution. It should contain
very little gasdline because by the above procedure it is not n.ecessary to draw
off the aqueous layers very close to the gasoli~e line. Boil the solution in a
154 By R. H. Pierson, Chief Chemist of the Gilmore Qil Co., Ltd.
155 NOTE.-Reasonabie care should be exercised to avoid contamination of the Pb
solutions from the use of wire on stopcocks and stoppers of separatory funnels, and from
metallic ring supports. Considerable amounts of fumes of Br, HBr, CC1 4 and gasoline
are evolved in the shaking procedure, and fine wire is usually used to avoid loss of
stoppers through the force of the gas pressure generated. If wirEVis used, it is well to
tie cloth over it to prevent any stray drops of acid from -carrying metal into solution.
Such metal can be removed by subsequent procedure,-b11l; it is preferable to avoid loss
of time so occasioned.
-

FIXED OILS, FATS AND WAXES

1805

good hood until the volume of liquid is reduced to 30 or 40 m!. Dilute to

100 ml. with distilled water and filter. Make the clear filtrate just ammoniacal, using litmus paper as indicator. Bring just acid to litmus with weak acetic
acid and add an excess of only five or six drops of glacial acid. Heat nearly
to boiling and add 10 ml. of a 10% solution of K 2Cr 207. Boil three or four
minutes to coagulate the PbCr04, filter hot through a tared asbestos Gooch
crucible, wash well with water, and dry in an oven at 100 C.
Finally, ignite as follow~: Place the Gooch crucible inside a wide form
porcelain crucible and heat. the outer crucible" with a small bunsen flame. The
flame should be about one inch in height and its top should just touch the
bottom of the crucible. The PbCr04 will darken slightly, which is normal.'
Excessive heating should be avoided; it is only necessary to bring about complete dehydration. Cool and weigh the Gooch crucible. Calculate to Pb or
Pb(C 2H 5)4 by the following factors:
PbCr04XO.6411 =Pb,
PbCr04 X 1.0006 = Pb(C 2H 5)4.
The 'following factors apply on the basis of the 200-m!' sample as outlined in
the procedure.
g. PbCr04 weighed X 3.2055 = g. Pb per liter of gasoline.
g. PbCr04 weighedX5.003=g. Pb(C 2H 5)4 per liter of gasoline.
Taking into account also the specific gravity of Pb(C 2H o)4, g. PbCr04Xl1.415
.=ml. Pb(C 2H 5)4 per U. S. gallon of gasoline.
It is not difficult by the above method to obtain results fl.n duplicate teeyts
which check within 0.05 ml. Pb(C 2H 5 )4 per gallon when w<i!ing on gasolines
containing ,about 2 m!. per gallon. In general, check determinations should
not show more than 2% variation.
Determination of Nitrobenzine in Oils Used for Oiling Floors
Procedure.-Shake 20 ml. of oil with 1 m!. of concentrated HOAc until
homogeneous, add about 0.2 g. of Zn powder, heat until the odor of PhN0 2
is gone, cool, decant into a separating funnel, wash the excess Zn 3 times with
1 m!. of diluted HOAc and after filtering these washings' add them to the
funnel, shake, let stand, separate the red-brown lower layer and divide into
2 parts. (1) To 1 part add 2-3 vols. of Et 20 and a " pinch" of Pb0 2 The
liquid immediately turns violet and on agitation the Et 20 absorbs this color.
(2) To the second part also add Et 20 as before and then a few drops of aqueous
K2Cr207. The aqueous layer becomes green, then yellow-red, and in the Et 20
the same violet color appears as with Pb0 2 By means of standard color
tubes which are prepared under the same conditions with knowR amounts of
PhN02 in a hydrocarbon oil the amount of PhN0 2 in the oil under investigation can be estimated. About 0.05 per cent PhN0 2 is the minimum which
can be detected. The intensity and tone of the colors are influenced by the
conditions of the reactions, so that both the amounts of liquids and the conditions must be carefully controlled. The use of a colorimeter serves to increase
the precision of the test.lo6
156 Pietro Biginelli, Atti congresso naz. chim. indo 1924, 274-6.
C. C. Davis, C. A.
19, 1391, ~ay, 1925..

PETROLEUM AND ITS PRODUCTSj TESTING AND

INTERPRETATION OF RESULTS!
, The rilodern refining methods with the growing demand of greater efficiency,
petter products and the reclaiming of former waste products places petroleum
chemistry in a very important place in the chemical industry. The testing of
crude oils .and their products determines the methods used in refining and gives
.an accurate evaluation of the oils to be refined.
Petroleums may be classed as paraffin, llsphaltic and mixed base crude.
The paraffin base petroleum is made up of the straight chain hydrocarbons
having a formula C n H 2n+2 They are saturated hydrocarbons and are not
acted upon by sulfuric acid. The asphaltic base contains the naphthenes
,(C n H2n) or cyclic compounds and aromatic hydrocarbons. The mixed c).'udes
are a combination of paraffin and asphaltic compounds.
Petroleums frequently have other compounds than the hydrocarbons mentioned, such as the compounds containing o~ygen, nitrogen, and sulfur. These
may be of industrial significance but are generally considered as the undesirable
qualities which should be removed. Foreign matter such as water, sand, clay
and asph.alt may be found as suspensions and emulsions. The information
as to their presence may avoid much trouble in the distillation as they often
form the emulsifying agents for permanent emulsions or decrease the temperatures in which " cracking" begins.
The increased demand for gasoline has increased the " cracking" of the
higher molecules and has made the tests for unsaturates, carbon in colloidal
form, fixed gases and gases containing gasoline, of great importance in most
commercial plants.
'
The reclaiming of the waste matters and the ma~ing of by-products by the
treating of the fractions containing the sulfur and nitrogen compound makes
a careful and accurate testing of these compounds of greatest importance.
The examination and evaluation of the petroleums and their products should
be done in an accurate and systematic manner, so that the combined results
will give a workable knowledge. The complexity of th~ petroleums or their
.fractions makes most tests have little value or they may'be very misleading
unless accompanied by other data. The conclusions arrived at should be
verified by other tests and a complete report made as to the results.
The following tests mentioned will be found in detail in other parts of this
text. Each test should have a definite point to determine and all data recorded in a logical manner.
1

Contribution by'Robert C. Beckstrom.

1806

PETROLEUM AND ITS PRODUCTS

1807

TESTS MADE ON CRUDE PETROLEUM AND ITS PRODUCTS,

INCLUDING THE INTERPRETATION OF RESULTS
Crude Oils
The Physical Tests.-The color of petroleum gives little information unless
other than green or black. Other colors may indicate contamination with
clays, sands or water.
Odor.-The odor of petroleum may indicate in a general way the fields from
which the oil came. A gasoline indicates a pure paraffin base crude with high
gasoline content. The aromatic and pyridine odors are common to some of
the crude oils from California, Texas, Rumania, Russia, and is found in shale
oils. The odor of pyridine is more pronounced if the sample is agitated with
potassium hydroxide solution. A slight terpene odor may be detected in some
of the Gulf Coast oils.The specific gravity or Baume gravity 2 of a crude indicates the approximate '6mount of light oils. This test should be followe'd by the distillation of
the crude oil for an accurate determination.
The test for sediment and water (page 1761) determines the need of dehydration, the types of stills used for refining or whether it will be accepted for pipe
line transportation. In the heavier crudes, water may be in such small droplets
that considerable care must be taken to complete the separation.
The fractional distillation of petroleum gives a fair determination of the
amounts of gasoline, kerosene, fuel oil, and lubricating oil. It is found that
anyone method of distillation for ,the different kinds of petroleums is unsatisfactory when fractions are identified by their boiling points alone. These
results should be supplemented with other tests. The asphaltic and mixed
base petroleums begin'to crack at lower temperatures than the paraffin base
petroleums and the products of distillation will be found to be much different.
A crude oil with emulsions or finely dispersed particles of water in it must first
be dehydrated before a distillation can successfully be made.
STANDARD METHOD OF TEST FOR DISTILLATION OF
CRUDE PETROLEUM 3

A. S. T. M.

DESIGNATION: D

285-36 4

Scope.-This method is intended for determining the percentages and distillation range of the naphtha in any crude oil of the class commercially known
llS refinable crudes.
This method does not attempt to specify what quality of
product shall be defined as naphtha nor can it be expected to duplicate the
'fi
't _
141.5
e speC! c gravl y- 131.5+BtS
3 Under the standardization procedure. of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,

.
260 S. Broad St., Philadelphia, Pa.
4 This method is issued under the fixed designation D 285; the final number indicates the y~ of original adoption as standard or, in the case of revision, the year of last
revision. ~ued as Tentative, 1928; Adopted in Amended Form, 1933; Revised, 1936.
This method has been approved as American Standard by the American Standards
Association.
2

Th

1808

PETROLEUM AND ITS

P~ODUCTS

results of commercial refining operations. It defines apparatus and procedure,

leaving selection of -numerical limits and interpretation of results to be agreed
upon by the interested parties.
Apparatus.-Flask.-The distillation flask shall'lbe of the type illustrated
in Fig. 259, the dimensions and allowable tolerances' being as follows:

-1.125" (2.!I em.)

Oulzide Dlameter

..L-

1...... 4.0"(IO.2cm.J,1

I'

Outside Diamefer

, FIG. 259.-Distillation Flask.

Tolerances

Centimeters

Inches

Diameter of bulb, outside ................

Diameter of neck, outside ...............
Length of neck .........................
Distance, top of neck to center of vapor
tube ................................
Length of vapor tube ...................
Diameter of vapor tube, outside ..........

10.2
2.9
25.4

4.0
1.125
10.0

"
6.4
2.5
17.2
6.75
0.8:-' :- 0.3125

Centimeters

Inches

O.2
0.1
0.4

0.08
0.04
0.16

0.3
::0.5
0.1

0.12
0.20
0.04

PETROLEUM AND ITS PRODUCTS

1809

The vapor tube shall be set at an angle of 755 with the neck of the flask.
Fractionating Column.-The fractionating column (see a, Fig. 260) shall be
made of a length of No. 18 iron jack chain, eno:ugh to pack under its own weight
into a column 2.54 cm. (1 in.) in diameter
and 16.51 cm. (67-'2 in.) in length, It may
conveniently be " strung" on a loop of wire
so that it hangs in loops about 2 ft. in
length, which makes the chain more convenient to handle than when it is not
looped together. The device used for supporting the column (see b, Fig. 260) shall
be made of a suitable length of wire, preferably, though not necessarily of nickelchromium, about No. 18 gage. One end
shall be wound in a spiral a little less than
1 in. in diameter, and the remaining wire
shall l:>e bent, at right angles to the plane of
the spiral and cut off at a length of about
97-'2 in. A small loop shall be bent into the
end away from the spiral and put through
a hole drilled in a strip of spring steel (see c,
Fig. 260). or other suitable material about
0.015 in. in thickness, % in. in width, and
3 in. in length, bent around a cylinder 1 in.
in diameter. When allowed to expand this
FIG. 260.-Fractionating Column
spring strip should hold firmly on the inside and
Supporting Device in Hempel
of the neck of the flask and provide a secure Flask.
'
support for the column of chain.
Thermomeier.-The A. S. T. M. Low-Distillation Thermometer, conforming to the requirements specified in Section 6 (a) of the Standard Method of
Test for Distillation of Gasoline, Naphtha, Kerosene, and Similar Petroleum
Products (A. S. T. M. Designation: D 86) of the American Society for Testing
Materials 5 shall be used.
.
Condenser.-The condenser shall consist of a 196-in. (14.29-mm.) OD No. 20
Stubbs Gage seamless bra~s tube, 22 in. (55.88 cm.) in length. It shall be set
at an angle of 75 from the perpendicular and surrounded with a cooling bath
15 in. (38.1 cm.) in length, approximately 4 in'. (10.16 em.) in width, by 6 in.
(15.24 em.) in height. The lower end of the condenser tube shall be cut off
at an acute angle, and curved downward for a length of 3 in. (7.62 cm.) and
slightly backward to insure contact with the wall of the graduate 1 to 1U in.
(2.54 to 3.18 em.) below the top of the graduate when it is in position to receive
the distillate.,.
Shield and Support.6_The shield shall be of any convenient type having a
minimum horizontal dimension of 6 in. The flask shall rest on a board of U-in.
51936 Book of A. S. T. M. Standards, Part II. See p. 1705.
6 An illustration of the condenser and of a suitable shield and support are shown
in Fig. 2 of the Standard Method of Test for Distillation of Gasoline, Naphtha, Kerosene, and Similar Petroleum Products (A. S. T. M. Designation: D 86) of the American
Society for Testi.ng Materials, see 1936 Book of A. S. T.'M. Standards, Part II.

1810

PETROLEUM AND ITS PRODUCTS

(0.64-cm.) transite or hard asbestos, in the center of which is cut a hole 3Yz in.
(8.9 cm.) in diameter. The board shall not be less than 6 in. (15.24 cm.) in
wiq.th and shall be supported by' an ordinary ring s'tand and ring.
Gas Burner or Electric Heater: (a) Gas Burner.-The burner shall be so
constructed that sufficient heat can be obtl}ined tq distill the product at the
uniform rate specified below. The flame should Inever be so large that it
spreads over a circle of a diameter greater than 5 in. (12.7 cm.) on the under
surface of the asbestos board. A sensitive regulating valve is a necessary
adjunct, as it gives complete control of heating.
:.' (b) Electric Heater.-The electric heater, which may be used in place of the
gas flame, shall be capable of maintaining the distillation at the rate specified in
Procedure (i). The electric heater shall be fitted with an asbestos board top
VB to ~ in. (3.18 to 6.35 mm.) in thickness having a hole 33--2 in. (8.9 cm.rIn
diameter in the center. When an electric heater is employed the portion of
the shield above the asbestos board shall be the same as with the gas burner,
but the part below the board may be omitted.
'
Measuring Graduates and Pipettes.-Filling charges and distillation fractions shall be measured with suitable accurate graduated cylinders or other
volumetric glassware.. A 300-ml. pipette with unconstricted outlet, similar
to Fig. 261, is a convenient device for ,measuring charges, but an accurate
graduate may be used if desired. The regular 100-ml. graduate used in gasoline
distillations as ,specified in Section 7
/ of the Standard Method of Test for
:
Distillation of Gasoline, Naphtha, Kero:~
i
sene,
and Similar Petroleum Products
E;-:
(A. S. T. M. Designation: D 86) of the
O.S"(1.?7cmJ
~~ ~
:
OJ-: C\.I
:
American Society for Testing Nraterials 1
.~ t!
:
is reasonably satisfactory for measuring
: ~
:
distillation fractions, but a graduate
--4
~ :
--{;:
of similar dimensions that can be tightly
JOO
,. I?
- -:- stoppered is preferable. One can be
~~
CS made I by '~utting down a standard
m/.
1<"5<3)"'_'~
_.'l-'o",,;.::
~-..;:
-r ~ .~ graduate to remove the lip and then
fire p()lishing.
--~~
,
C).

PI;0cedure. - (a) The condenser

:,
(\j

bath shall be filled with cracked ice

:.........
i
and enough water add~d to cover the
'~. ~
:
condenser tube. The temperature shall
a312S" (0.8 em.)~~
:
.(\j
:
be maintained between 32 and 34 F.
:, ' ,
(0 and 1.1 C.). Any other convenient
- - L __ .:L cooling
medium may be used proFIG. 261.-Measuring Pipette.
vided the~e temperature conditions are
maintaim!d.
(b) The condenser tube shall be swabbed or cleaned inside to remove any
liquid remaining from a previous test.
(c) The specific gravity of the oil to be tested shall first be determined in
accordance with the Standard Method of Test fpr _Gravity of Petroleum and
7 1936 Book of A. S. T. M. Standards, Part II:
-

-'--r. ..

:
:

_-

PETROLEUM AND ITS PRODUCTS

1811

Petroleum Products by Means of the Hydrometer' (A. S. T. M. Designation:

D 287) of the American Society for Testing Materials. 7
(d) A quantity of the oil 8. equivalent to 300 ml. at 60 F. (15.5 C.) shall
be measured or weighed into the distillation flask by any suitable means.
None of the liquid snaIl be permitted, to flow into the vapor tube.
(e) The supporting device for the fractionating column, Fig. 260, shall be
put in place and the proper quantity of iron chain dropped in carefully so that
it fills the space uniformly and without channels. Tapping the flask while the
iron chain is being added is helpful, but the column shall not be compressed
after all the chain is in place.
(f) The thermometer, provided with a cork, shall be fitted tightly into the'
flask, so that i~ will be in the middle of the neck and so that the lower end of the
capillary tube is about 'fr; in. below the level of the inside of the bottom of the
vapor tube at its junction with the neck of the flask.
(g) The charged flask shall be placed in position on the hard asbestos board,
used with the gas burner or the electric heater and connected to the condenser
with a carefully fitted cork through which the vapor tube passes. The position
of the flask shall be adjusted so that the vapor tube extends into the condenser
tube not less than 1 in. (2.54 cm.) nor more than 2 in. (5.08 cm.).
(h) A clean, dry, graduated cylinder shall be placed at the outlet of the
condenser tube in such a position that the condenser tube shall extend at least
1 in. into the graduate. The graduate shall be immersed up to the level of
about 1 in. below the tip of the condenser tube in a transparent container of
water maintained at a temperature of from 32 to 40 F. (0 to 4.4 C.). During
the distillation, the top of the graduate shall be covered closely with a piece of
blotting paper or its equivalent, cut to fit the condenser tube tightly.
(i) When everything is in readiness, heat shall be applied. The fractionating column causes some little delay between the time when the liquid starts to
boil and the time the vapor enters the condenser. It is, therefore, permissible
to apply heat vigorously until the liquid begins to boil, then to slow down the
application of heat so that the distillate begins to come over at a moderate rate.
For the first 5 to 10 m!. distilled the rate shall be from 2 to 3 m!. per minute;
thereafter, the rate of distillation shall be increased to from 4 to 5 m!. per minute.
(j) Distillation.-When, the thermometer reads . . . 9 the graduate shall be
withdrawn from beneath the condenser and the distillation discontinued. The
graduate shall be tightly stoppered and allowed to stand until all sediment and
moisture have settled out and until its contents have reached a temperature
between 55 and 65 F. The total volume in the graduate shall then be read
and recorded as D. The volume of water, if any, shall be read and recorded

~~

The percentage of naphtha in the crude oil shall be calculated from the
following formula:

D-W

Percentage of na p htha=300_ WX100

8 In case the water content of the crude oil causes bumping or in any case if the
crude oil. contains more than 2 per cent of water, the sample shall be dehydrated by
clome suitable method, without loss of naphtha, before making the determination.
9 The temperature at which the distillation shall be stopped can. best be decided by
mutual agreement between the parties concerned in the evaluation of a given crude
petroleum.

1812

PETROLEUM AND ITS PIWDUCTS

The whole procedure snaIl be repeated enough times to yield a t'otal volume
of distillate of not less than 100 pli.
j
Naphtha Distillation.-The distillates shall be Poured together (taking care
to avoid losses by evaporation and rejecting the l~yers of water, if present)
and shall be mixed thoroughly by shaking. A disthlation test of the combined
distillates shall be made in accordance with the Standard Method of Test for
Distillation of Gasoline, Naphtha, Kerosene, and lSimilar Petroleum Products
(A. S. T. M. Designation: D 86) of the American SO()iety for Testing Materials,la
except that the initial boiling point shall be recorded. and the volume of distillate
collected in the cylinder shall be observed and rectlrded to the nearest 0.5 ml.
when the mercury of the thermometer reaches 212 0 F. (100 0 C.), 221 0 F.
(105 0 C.), 284 0 F. (140 0 C.), 392 0 F. (200 0 C.) and the end point ..
Accuracy.-Results of duplicate tests in the ~rude-oil distillation should
not differ from each other by more than 0.5 per cent.
Correction for Barometric Pressure.-The actUal barometric pressure shall
be ascertained and recorded, but no correction shall be made except in case of
dispute, when the temperature points shall be corrl;lcted to 760 mm. (29.92 in.)
by the use of the Sydney Young Equation, as follows:
For Centigrade Readings:

For Fahrenheit Readings:

C =0.00012(760-P)(460+t,),
in which Ce and C, are, respectively, corrections to be added to the observed
temperature te.or t,; and P is the barometric pressU\'e in millimeters of mercury.
Sulfur may be detected by the odor of hydrogen sulfide or the mercaptans
when a crude is heated. The sulfur tests as outlinecl on pages 1713 and 1756 give
the quantitative measurements. The fact that SUlfur is found in a petroleum
does not follow that it will be equally divided in ihe distilled fractions. The
author has found sulfur in varying amounts in thb, ,iO% fractions of the same
crude oil with a slight variation in the method ot refining. The importance
of sulfur tests is not in the cr~de oil but in the prGducts of distillation.
Gasoline, Naphtha
The specific gravity, ,Baume or A. P. 1., is tak~n with 'a hydrometer (page
1704). This information without other data, SUefl as the distillation curve or
the knowledge of its being a straight run, natural !!;asoline, or blended product,
is of little value in the determination of the quaIit:y- of the gasoline.
The yellow color in a gasoline (test, page 1742) indicates improper treatment
wit~ aci? or ~odium ~lumbite solution.. A_ dar.k cplor may be caused by contammatIOn WIth heaVier hydrocarbons III pipe lme .or tank cars, and should note
be confused with colors caused' by other methods.
The sweet ethereal odors of gasoline are characteristic of the gasolines from
low sulfur paraffin base crude oil. Sour or " craQked " odor is characteristic.
I

10

1936 Book of A. S. T. M. Standards, Part II.

Se&~:
-'

1705.

PETROLEUM AND ITS PRODUCTS

1813

of a gasoline incompletely refined a'nd it is probably made from a high temperature cracking process.
The distillation test (page 1807) should be made with great care as a change
in the rate of distillation, rapid or slow heating during the process gives inac~
curate results. The distillation curve indicates a straight run, natural gasoline
or blended product. The distillation range of a motor gasoline should conform
to the following: 11
Overpoint not more than 131 F.
20% recovered in receiver not more than 221 0
50% recovered in receiver not more than 2840
90% recovered in receiver not more than 392
End point'not higher than
4370
At least 95% of distillate recovered.
0

F.
F.
F.
F.

The sodium plumbite test (page 1742) when positive indicates sulfur compounds that will cause the gasoline to turn yellow in storage. The gasoline
giving a negative test may upon standing in the sunlight later give a positive
test: This shows that the treatment of the gasoline has been incomplete and
further treatment is needed. The sodium plumbate treatment may not decrease the sulfur content and may increase it. It does, however, cause the
sulfur compounds to become more stable so as not to cause the gasoline to
become _yellow,
The testing for unsaturates (tests, pages 1766-1767) or the test by the United
States Bureau of Mines 12 may be used. The Maumene (page 1766) test is to be
used only as a check as it is unreliable for highly unsaturated mixture.
U. S. Bureau,of Mines Test._:Five ml. of the gasoline are run into a clean,
dry Babcock cream bottle and when cooled in ice water 200% by volume of
sulfuric acid (1.84 sp,gr.) is added. A rubber stopper is then placed in the
bottle and shaken with a rotary motion for several minutes. The gasoline and
sulfuric acid then are separated by adding sulfuric acid until the surface is
about level with the upper graduation mark Dn the neck of the bottle. The
bottle is aliowed to stand twelve hours or it can be placed in a centrifuge and
whirled at the rate of 500 to 1000 R.P.M. for three minutes. A reading of the
volume of gasoline remaining is then made and the loss to sulfuric acid is
calculated.
The unsaturation gives information as to the amount of acid needed for the
acid treatment. A highly unsaturated finished product may cause the deposition of carbon and gummy matter in the cylinder of the motor. This point is
not accepted by all authorities.
Corrosion and gumming test (page 1742) indicates elementary sulfur compounds in the gasoline. The gummy material may be acid residues, polymerization of highly unsaturated compounds or asphaltic bodies.
11
12

Bulletin 323A, United States Bureau of Mines.

Tech. paper 181.

1814

PETROLEUM AND ITS PROPUCTS

STANDARD METHOD OF TEST FOR GU CONTENT OF

GASOLINE 14
I

15
A. S. T. M. DESIGNATION: D, 381-36
I

Scope.-This test is a means of determining the amount of gum existent

at the time of test in gasoline and other motor fuels boiling within th.e gasoline
range. The results are indicative of the amount of gum deposition whiclJ. may
take place in service if the gasoline is used immediately but give no indication
of the gum stability of the gasoline upon storage. Results obtained by this
test are comparativ~ only, inasmuch as the amount of gum residue obtainable
from a given sample varies with the conditions of evaporation. In cases where
the sample contains other dissolved non-volatile material, such as lubricating
oil, such material may be obtained as residue along with possible gum; such
samples obviously require special treatment, depending on the nature of the
non-volatile material.
Apparatus.-The apparatus shall consist of the following:
(0) Beokers.-Berzelius-type beakers of h.eat-resistant glass, without lip,
of 100-ml. nominal capacity.
(b) Bath.-The bath shall consist ofa closed vessel provided with a reflux
condenser, a well into which the beaker fits, a coil for preheating the air as
prescribed in the Procedure, and a conical adapter to deliver th,e air into the center
of the beaker. A suitable form of apparatus is shown in Fig. 262. A mUltiple
bath having similar provisions is also satisfactory. The bath shall be filled
to within 1 in. of the top with a stable liquid having a boiling point within a
temperature range from 3200 to 330 F. (160 to 165.8 C.), ethylene glycol
containing approximately 3% water being suitable for the purpose. The bath
shall be insulated with asbestos. An electric hot plate capable of keeping 'the
liquid boiling actively is the preferred heating means. In the absence of a satisfactory hot plate the bath may be heated by a gas burner, in which case the
apparatus shall be placed in a hood with a good draft!
(c) Air Supply.-Low-pressure air, passing through a cotton or glass-wool
filter and delivered to the preheater inlet of th~ bath. A flow meter or other
air-measuring device shall be provided as shown in Fig. 262. The flow meter
shall be calibrated when attached to the bath with the bath liquid at room
temperature.
'
Procedure.-(a) The test beaker together with a sinn1ar ta~re beaker shall
be washed free from gum with a mixture of equal volumes of benzol and alcohol,
immersed for at least 12 hr. in chromic-acid cleaning solution and washed
thoroushly first with tap water and then with distilled water. Subsequent to
removal fr:om the cleaning mixture the beakers shall be handled only by means
of a clean, dry, metallic holder. After washing, both beakers shall be dried
thoroughly by heating to about 302 F. (150 C.)\ and then allowed to coolt
14 Under the st~dardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
I
15 This method is issued under the fixed designation:J)~381; the final number indicates the year of original adoption as standard or, in th'e case of r(lvision, the year of last
revision. Issued as Tentative, 1934; Adopted in Amended Form, 1936.

PETROLEUM AND ITS PRODUCTS

1815

simultaneously for the same period of time, but not less than 1 hr. in a desiccator. If the air in the balance to be used is dried with a desiccant then this
same desiccant shall be used in the desiccator; otherwise no drying agent shall
be used in the desiccator. The test beaker shall then be weighed, using its tare
(Note) on the opposite balance pan. After weighing, the tare beaker shall be
replaced in the desiccator and the test beaker shall be placed in the well of the
bath as shown in Fig. 262. The air flow shall be adjusted to a rate at the point
of measurement of 1 liter per second15%. The bath shall then be heated

''''5''O
r.1li
o.

1~4t.:> To i?efivA' Condel1$er

CoflonorG/oss
Wool Filter

I
I

I
I

I
I

O)I~

_____ 1,/
,rOD.C~pper
Silver Solder
41/ .Joints

TvbingJ/Wall

,.

Flowl1elt't;
not drl)wn

t{)

"
I

To Air $vppOi

5C(I/e

FIG. 262.-Apparatus for Determining Gum Co.ntent of Gasoline.

so that the temperature of the effluent air, measured by a thermometer placed
on the bottom of the beaker at the center, shall be from 3050 to 3200 F. (151.7 0
to 1600 C.). This is governed largely by the rate of boiling of the bath liquid
and the rate of air flow.
NOTE.. -For routine tests it is permissible to use the same beaker as a tare for more
t.han one test beaker but not for more than six test beakers, provided all are heated at
the same time.
.

1816
,

PETROLEUM AND ITS PltoDUCTS

(b) The conical adapter shall be removed and 50 ml. of the gasoline poured
into the beaker after which the adapter shall be r()placed on the air outlet so
that it is vertically above the center of the surface of gasoline. After the gasoline has evaporated, which should require from 8 to 14 minutes, the beaker shall
be left in place for an additional 15 minutes, the atr1supply being maintained.
During this 15-minute period the tare beaker shall !lJ~o be heated to about 302 0
F. C150 C.). Both test and tare beakers shall then be allowed to cool simultaneously and for the same period of time, but not 1ess than 1 hr., in a desiccator
under the same conditions as prescribed in Paragraph (a), and the test beaker
weighed as before the test. The increase in weight of the test beaker represents the amount of gum in the sample.
Results.-Duplicate determinations shall be made and results shall be
expressed in milligrams of gum per 100 ml. of gasoline.
Accuracy.-Results obtained by different operators in different laboratories
should not differ by more than the following value!$:

0 to 20 mg. per 100 m!.. . . . . . . . . . . . . . . . . 4 mg.

For gum contents of
For gum contents of
20 to 100 mg. per ,100 ml..... , . . . . . . . . . . .. 10 mg.
For gum contents above
100 mg. per 100 ml.. . . . . . . . . . . . . . . .. 20 mg.
Duplicate results obtained by a single operator should not differ by more than
one-naif tne aoove vaiues.
Vapor Pressure determinations give direct information as to the probable
behavior of motor fuels as regards easy starting Itnd "flexibility." It indicates, if the pressure is high, the blending of natural gasoline with a straight
run product. It is important in the shipping of ~asoline in tank cars 13 and
must conform with interstate shipping laws.

Kerosene, Mineral Seal and Cas Oil

. Specific Gracity.-The specific gravity, Baume or A.P.L, is determined
by the same methods as described under gasoline. The test indicates approximately where the fraction has been cut during the distillation of the crude oil.
This must be confirmed by other data such as: the distillation curve, flash and
fire points.
The dark colors show incomplete acid treJttment, percolation or contact
filtration. A" cracked" odor is the result of high temperatures or pressures
during distillation, but may be removed with acid treatment.
The viscosity tests (page 1718) must conform to the speCification for fuel
oils under working temperatures. The oils should he free from water, dirt, and
tarry matter which may clog the burners.
The cloud test (page 1727) indicates too deep a cut into the lubricating
fraction and the settling out of crystals of paraffin. A fuel oil showing a cloud
test will solidify when in storage during cold weather.
to
The distillation test determines the cause for the decrease of the illuminating
power of the oil after burning for a time. In absQrbing oils for" casing head
gas" the distillation curve may determine its usefulness as an absorbing oil.
A fraction having an over point much higher than the end point of the gasoline
is desirable.
_;,:._
,
13

Bureau of Explosive Circular MCL, No. 147.

PETROLEUM AND ITS PRODUCTS

1817

The flash and fire tests (page 1702) must meet the specifications of illuminating and burning oils. A low flash and a high fire indicates a mixture of light
fractions with the heavy fractions, making a poor fuel for the burners. The
flame will not be uniform and the higher fractions will not burn but change into
a vapor in the fire box, causing explosive mixtures. The test determines the fire
hazard and is the real measure of inflammability. Most states have a law
stating the minimum flash and fire points of illuminating oils.
The sulfur determinations (pages 1712, 1756) are of great importance in
illuminating oil as they produce disagreeable odors during the burning period,
and cause a charring of the wicks. The amount should not exceed .1 %.
Fuel oils when atomized with steam form sulfuric acids which cause rapid
deterioration of boilers and pipe stills.
Lubricants

The determination of specific gravity of a lubricating oil (page 1704) may

give information as to the nature of the petroleum from which it is made. A
high Raume gravity and a very high viscosity may indicate a lubricant that has
been mixed with soaps, resins, or oils other than mineral oils.
The color of lubricating oils indicates the source of the oil. A turbid oil
indicates finely divided droplets of water, or if cooled an amorphous wax. A
dark-colored oil should be examined under a microscope for suspended carbon.
When the oil is diluted with naphtha, the foreign matter held in suspension
settles out.
The viscosity test (page 1718) has been regarded as a most important test of
lubricants. The working temperature of the bearing should determine the
temperature at which the test is made. Most lubricants have the same viscosity at 3000 F. A lubricating oil functions only when a continuous film of oil
separates the wearing surfaces and the oil is efficient when its viscosity allows
this film to be continuous without a break in the film. The true "test for a
lubricating oil under working conditions ill the machine for which it is to be uiled.
The cloud test (page 1727) indicates the temperature at which the paraffin
wax separates from the oil solution. It shows whether the process for the
separation of wax has been well done and gives the quality of the wax which is
separated out. If the wax is amorphous in character, the oil must either be
redistilled or centrifuged to remove this wax.
The pour test (page 1730) indicates the temperature under which an oil will
flow. Knowing the working temperatures that the oil will be used will determine whether it will Qe an efficient lubricant.
The flash and fire test (page 1732) should be above the working temperatures of the bearing. High flash and burning points indicate good quality
and careful refining. Too low flash and fire test results in a large .loss from
evaporation under working conditions.
The carbon residue test (page 1737) gives information in the laboratory
approximating working conditions with the lubricants, producing more or less
carbon deposits. A large per cent of residue indicates oils. thickened by other
matter or badly refined.
.
The demulsibility test (page 1740) determines the value of a lubricant which
is to be used in forced feed lubrication and motor cylinders. A high degree of

1818

PETROLEUM AND ITS PRODUCTS

emulsification indicates an oil containing an emulsifying agent such as sulfur

compounds, salts of napthenic and other -organic acids.
The fatty acid test (page 1736) when more than 15%, unless found in a
cutting oil, will cause the formation of metallic soaps which clog the lubricating
\.
.
channels.
The test for rosin (page 1781) when found positiye indicates an adulterated
lubricating oil and is always objectionable.
Free sulfuric acid indicates poor refining methods and should not exceed
.01 % in transformer oils.
TENTATIVE METHOD OF TEST FOR NEUTRALIZATION NUMBER
OF PETROLEUM PRODUCTS AND LUBRICANTS 16
A. S. T. M.

DESIGNATION:

P 188-27 T

17

Scope.-The methods described herein are designed to indicate in petroleum products and lubricants and compound~d products the presence of organic
constituents having acid characteristics and the contamination by alkalies
and mineral acids.
Definitions.-(a) The Neutralization ,Number is the weight in milligrams
of potassium hydroxide required to neutralize one gram of oil.
NOTE.-The Neutralization Number expresses the total amount of the mineral
acid, and of the organic constituents having acid characteristics.
(b) The Alkali Neutralization Number is defined as the weight of acid
required to neutralize one gram of oil, expressed in equivalent milligrams of
potassium hydroxide.
.
(c) The Mineral Acid Neutralization Number is the number of milligrams
of potassium hydroxide required to neutralize the mineral acid content in one
gram of oil.
. Solutions Required.-(a) Aqueous Potassium IHydroxide (1 ml. =5 mg.
KOH).-Dissolve 5.1 g. of potassium hydroxide,'c.p., in one liter of freshly
boiled and cooled distilled water. Add a very small amount of barium hydroxide, sufficient to precipitate any potassiuni carbonate present. Standardize
this solution against Bureau of Standards certified benzoic acid, using phenolphthalein as an indicator according to the relation:

5 ~g. KOH
x mg. benzoic acid

56.104 g. KOH
'
122.048 g. benzoic acid-

1 ml. of KOH = 10.88 mg., benzoic acid.

This weight of

ben~oic

acid is required for standardization.

16 Criticisms of this Tentative Method are solicit~d and should be directed to~.
R. P. Anderson, Secretary of Committee D-2 on Petroleum Products and Lubricants,
250 Park Ave., New York City. Reproduced through the courtesy of the Anlerican
Society fot Testing Materials, 260 S. Broad St:, Philadelphia, Pa.
17 This is a Tentative Standard, published for the purpose of eliciting criticism and
suggestions, and as such is subject to annual revision.- .. Issued, 1925; Revised, 1926,
1927. This method has been approved as Anlericltn. lr~n!;ative Standard by the AmeriI
can Standards Association.
-

PETROLEUM AND ITS PRODUCTS

1819

Make necessary adjustments so that the value of potassium hydroxide

equals 5 mg. KOH per 1 ml.
.
NOTE.-Fit the RDlutiDn bDttle with a guard tube Df SDda lime to. prevent access
Df carbDn diDxide. The sDlutiDn shDuld be standardized at necessary intervals.
The weight Df benzDic acid shDuld be dissDlved in 50 m!. Df 95% alcDhDI and titrated
cold. FDr blank, use same amDunt Df alcDhDI and CDrrect the titratiDn.

(b) Sulfuric Acid Solution (1 ml.=5 mg. KOH).-Match a sulfuric acid

solution so. that 1 m!. af the acid will be required to. neutralize 1 m!. af the KOH
using ph en 0.1 phthalein as an indicator, titrating in boiling solutian.
(c) Neutralized 95% Alcohol.-Add a few drops af phenalphthalein and
neutralize carefully the alcahal to a very faint pink end paint with same af the
abave prepared alkali salutian.
(d) Phenolphthalein Indicator.-Dissalve 10 g. of the indicator in 1 liter of
95% alcahol, preferably ethyl. Use 1 ml. of this strength for titratian.
(e) Methyl Orange Indicator.-Dissalve 1 g. af the indicatar in 1 liter of
distilled water. Use 0.1 ml. of this strength fartitratian.
Methods.-Method A.-Neutralization Numoer of Petroleum Products.Weight or Oil: Appraximately 20 g. weighed to. 0.1 g.
Valume of Sal vent : 100 ml. of a mixture af 1 : 1 neutralized alcohol and
distilled water.
Pracedure: Agitate oil and salvent tharoughly and heat to boiling. Add
1 m!. of phenolphthalein indicator and titrate rapidly, with vigorous agitation,
to a sharp pink end point. The titration must be completed in a hot solution,
reheating same if found necessary.
The color change is noted in the alcohal water layer.
Calculation:
(Milliliters of KOH) X5
. ht 0 f 01'1 t a k en
W mg

mg.

KOH

per

g.

f'l
0

01.

Method B.-Neutralization Number of Compounded Petroleum Products.

-Weight of Oil: Approximately 10 g. weighed to 0.1 g. Volume of Solvent:
50 ml. of neutralized alcohol.
Prpcedure: Agitate oil and sal vent thoroughly and heat to boiling. Add 1
mI.. of phenolphthalein indicator and titrate rapidly, with vigorous agitation,
to a sharp pink end point. The titratian must be completed in a hot solution,
. reheating same if found necessary.
The color 'change is noted in the alcohol layer.
Method C.18-Alkali or Mineral Acid Neutralization Number in NonCOmpounded and Compounded Petroleum Products. Weight of Oil: 25 to 50 g.
weighed to 0.1 g. Volume of Solvent: Boiling distilled water, 200ml.
Procedure.-Introduce into a 250-ml. separatory funnel 25 to 50 g. of oil
weighed to 0.1 g., and add 100 ml. of boiling distilled water. Shake vigorously,
and after oil and water have separated, drain the water layer into a 500-ml.
casserole. Wash the oil in the separatory funnel twice by vigorously shaking
with 50~'!n1. portions of boiling distilled water and after separation, drain into
the casserole. To the accumulated 200 ml. of water add 1 drop of 1 % phenol18 When the Dil is acid, the acidity shall not be reparted as sulfuric or hydrochlDric
unless the presence Df these acids shall have been cDnfirmed by tests.

1820

PETROLEUM AND ITS PRODUCTS

phthalein solution. Boil. If solution turns pink, add 1 nil. of the indicator
and titrate with the sulfuric acid solution until the eixtraction is colorless.
If the addition of 1 drop of phenolphthalein causes no change in color, add
0.1 ml. of methyl orange. If color changes to red lor pink, titrate with the
KOH solution until the red color is just discharged. I
/'
An equal volume of distilled water shall be titrated with the standard
solution used, and the amount of alkali or acid required shall be subtracted
from the titration for blank correction.
Calculation: If the acid solution was required to discharge the phenolphthalein end point, the alkali neutralization number equals:
(Milliliters of acid) X 5
Weight of oil
If the alkali solution was required to discharge the methyl orange end point,
the mineral acid neutralization number equals:

(Milliliters of alk!lli) X 5
Weight of oil
I

The evaporation test (page 1726) is used in testing lubricating oils used in
high vacuum machines, transformer o_ils and air compressor oils. A high
evaporation loss indicates light oils used in blending. It may be used in testing
the dilution of " crank case" oil.
STANDARD METHOD OF TEST FOR DILUTION OF
CRANKCASE OILS 19
A. S. T. M.

DESIGNATION: D

322-35 20

Scope.-This method is intended for the <;letermination of the amount of

dilution in crankcase oils of engines when gasoline has been used as the fuel.
Apparatus.-(a) General.-The apparatus shall consist of a glass flask,
heated by a gas-burner flame, and provided with a reflux water condenser
discharging into a trap for collecting the distillate. The trap serves to collect
and measure the diluent; returning condensed water to the still. ,
(b) Trap.-The trap shall be made of well-annealed glass constructed in
accordance with Fig. 263. It shall be graduated from D to 12:5 m!. in D.I-m!.

Under the standardization procedure of the 'Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
This method comprises a revision of the Standard Method of Test for Dilution I!f
Crankcase Oils (A. S. T. M. Designation: D 322-33), 1933 Book of A. S. T. M. Sta.
ards,.Part II, p. 768, which method it supersedes.
Reproduced through the courtesy of the American ,Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.

20 This method is issued under the fixed designation D 322; the final number indicates the year of Original adoption as standard or, in the-case of revision, the year of
last revision. Issued as Tentative, 1930; Adopted,.193!f; Revised, 1935. This method
has been approved as American Standard by the American Standards Association.
19

PETROLEUM AND ITS PRODUCTS

divisions.
0.05 ml.

1821

The error of any indicated capacity shall not be greater than

(c) Flask.-The glitilS flask shall be of the short-neck, round-bottom type,

with a nominal capacity of 1 liter. '
(d) Condenser.-The condenser shall be of the water-cooled, reflux, glasstube type, having a condenser jacket not less than 400 mm. (15% in.) in length
with an inner tube 9.5 to 12.7 mm. (% to 72 in.) in outside diameter. The
end of the condenser to be inserteu in the trap shall be ground off at an angle
of 60.

(e) Burner.-The gas burner shall preferably be of the Meker type, 1 in. in
diameter, and the orifice used should be such that a narrow flame coming to a
point about 4 in. above the top of the burner can be obtained.
Procedure.-The sample shall. be mixed thoroughly, and 25 mI., shall be

PETROLEUM AND ITS PRODUCTS

1822

measured into a 25-ml. graduate and transferred to the flask. By washing

out the graduate with hot water only a negligible a~ount of oil will be left in
the graduate. Approximately 500 ml. of water shall. then be added to the flask
and the apparatus shall be assembled, after filling ~he trap with water. The
tip of the condenser shall be placed directly over t~e indentation in the trap.
Heat shall be applied with the full flame as indicated above. The rate
(Note 1) shall be such that refluxing starts within f~om 7 to 10'minutes after
the heat is applied, the water being at room temperature initially. The success
of the method depends upon heating at such a rate that boiling is continuous
and vigorous enough to prevent the oil from forming, even. momentarily, a
continuous film on top of the water. (Note 2.)
NOTE I.-As a check on the rate of distillation, a determination may be made on
a fresh oil contajning 15% of added diluent. The oil used should have a flash point
of 400 F. o~ over and the diluent used should have an initial boiling point of 300 to
320 F. and an end point of 400 to 420 F. At least 13.5% of the diluent should be
obtained by this method on boiling for 60 minutes.
NOTE 2.-Bumping with a tendency to froth over is often experienced with dirty
oils and may be prevented in several ways, namely, 'by the addition of bits of broken
glass, by the addition of steel wool, or by the addition of about 5 mI. of concentrated
HC!.
0

Readings of the amount of diluent shall be made at the following times,

taken from the time that refluxing starts: 5, 15 and 30 minutes, and each 15
minutes following, until the test is complete. Completion of the test shall be
determined on the basis of either or both of the following criteria:
. No. I.-Test is complete when the volume of diluent increases by not
more than 0.1 ml. in any 15-minute period during the course of the test.
No. 2.-Test is complete when the volume of diluent obtained in a
given time indicates completion, as follows:
Time (rom Start
of Refluxing

!
I

Test Is Complete if
Apparent Volume of
Diluent Collected Is
Equal to or Less Than

5 minutes. , ......................... ,................. 0.1 mI.a

30
" .......................... .;. . . . . . . . . . . . . . . .. 2.0 mI.
60
"
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4.0 mI.
90
a

"

............................................ 7.0 mI.

Report as "no dilution."

Generally criterion No.1 defines the completion of th~ test, but when the
test continues to a point at which any of the conditions described. under criterion
No.2 are encountered, the latter shall define the completion of the test.
When the test is complete by either of the criteria above the heat shall be
tu~ned off. After standing at least 15 minutes, to allow the distillate to settle
clear and to cool to approximately room temperature, the volume of di~t
shall be read. The percentage dilution is obtained by mUltiplying this tiial
volume by 4.
Accuracy.-Determinations by independent operators should agree within
, 20% for dilutions under 5% and within 10% for dilutions over 5%. Duplicate
determinations by the same operator should, in ~n.eral, show a closer agreement
than the above.
. -

PETROLEUM AND ITS PRODUCTS

1823

STANDARD METHODS OF ANALYSIS OF GREASE21

A. S. T. M.

DESIGNATION: D

128-27 22

Scope.-These methods of analysis permit determinations sufficiently

accurate for referee purposes of all the constituents of greases likely to be
covered by specifications. These constituents are fillers and ash, soap bases,
soap, fat, water, excess alkali or acid, petroleum products and unsaponifiable
matter. No quantitative determination of glycerin is described, but a procedure is outlined for its qualitative detection.
Sample.-The size of sample for the soap determination (Methods .(1)
and (2) for Determination of Fillers, Soap, Fat, Petroleum Oil and Unsaponifiable Matter) shall be from about 8 to 30 g., depending on the consistency of
the grease, which is chiefly determined by the percentage of soap present.
Ten to twenty grams is usually a convenient amount for No.3 Cup Grease,
while thin transmission and other greases require a larger sample. The original
sampl& should be stirred or mixed until uniform.
Samples need not be weighed more closely than 0.1 g.
Solvents.-The petroleum ether shall have an end point not higher than
200 0 F., and 125 ml. of it shall not consume more than 0.2 nil. of 0.5 N potassium
hydroxide when a blank test is made as follows:
One hundred and twenty-five milliliters of petroleum ether shall be boiled
with 10 mL of 0.5 N alcoholic potassium hydroxide and 50 ml. of neutral 50%
alcohol for 1Y2 hours on a hot plate, using a glass tube of about 7 mm. internal
diameter and 75 cm. long as a reflux condenser. After cooling, the solutions
shall be titrated with 0.5 N hydrochloric acid and phenolphthalein. Not less
~han 9.8 mI. of 0.5 N acid shall be required for neutralization. The amount of
alkali consumed in this test shall be deducted as a blank correction in the fat
determination on solution (E).23
The alcohol shall be prepared from commercial 95% grain or denatured
grain alcohol by distilling from sodium hydroxide, diluting to 50% (or 70%) by
volume with distilled ,vater and neutralizing exactly with sodium or potassium
hydroxide and phenolphthalein.
The phenolphthalein solution shall be prepared by dissolving 1.0 g. of
phenolphthalein in 50 ml. of strong redistilled alcohol, adding 50 ml. of water,
and neutralizing with 'sodium or potassium hydroxide.
Concentrated c.p. hydrochloric acid shall be used when "concentrated
hydrochloric acid" is specified. Hydrochloric acid, 10%, shall contain 10%
by weight of absolute HCl, with a permissible variation of O.5%.
21 Under the standardization procedure of the Society, these methods are under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
.
.
22 These methods are issued under the Bxed designation D 128; the final number
indicates the year of original adoption as standard or, in the case of revision, the year
of last revision. Issued as Tentative, 1922; Adopted in Amended Form, 1927. These
methods have been approved as American Standard by the American Standards Association.
23 Solutions designated by capital letters in parentheses refer to the solutions given
in Fig. 264.

1824

PETROLEUM AND ITS PRODUC'rS

Procedure.-(A) Determination of Ash.-The percentage of ash shall not

be included in the total. of the analysis.
j
NOTE.-The determination of the total ash should not in general be regarded as
of any great importance. It ,is, however, sometimes reqwred. This determination is
often unsatisfactory on account of interaction between sodium carbonate derived from
the soap, and inorganic fillers. There is always likelihood, of reaction with the porcelain
crucible itself on account of the long continued heating necessary to burn out all carbon.
Moreover, if much sodium or potassium carbonate is present, the ash is fusible and often
encloses carbon, making complete removal of the latter very difficult. Results will
always be low in the presence of easily reducible oxides of volatile metals. There is
also uncertainty as to when calcium carbqnate has been completely ignited to calcium
oxide. Ash determinations made on the same sample in different laboratories are likely
to vary widely. For these reasons, it is usually preferable to make separate determinations of the percentage of fillers and of soap, from which data a calculation is easily
made giving all the information to be gained from a direct determination of total ash.

(1) Rapid Routine Method of Ash Determination.-From 2 to 5 g. of grease

shall be placed in a weighed porcelain crucible, and the sample weighed to the
nearest 0.1 g. If lead or zinc soap is known to be absent, a platinum crucible
is more convenient. The combustible matter shall he bur~ed off slowly atld the
residue finally ignited until the ash is free of carbonaceous matter. The crucible
and contents shall then be cooled in a desiccator and weighed, reporting the
result as percentage of ash.
(2) Alternative Method of Ash Determination.-The sample shall be burned
as in method (1) until the ash is nearly free from carbon. The crucible and
contents shall then be cooled, the soluble portions dissolved in a little water,
and a slight excess of dilute sulfuric acid added, running the acid in carefully
from a pipette inserted under a small watch glass covering the crucible. The
crucible and contents shall then be warmed on the steam bath until !Jffervescence has ceased. The watch glass shall then be rinsed with water into
the crucible. The solution when tested with methyl orange for free acidity
shall show free acid present. The contents of the crucible shall be evaporated
to dryness, and the whole ignited at a low red heat,/~dding a few small pieces
of dry ammonium carbonate to drive off the excess of sulfur trioxide. After
cooling and weighing, the' result shall be reported as "percentage of ash as
sulfates."
The alternative method gives more concordant results than the first method,
but requires more time a;nd manipulation.
'- _,
(B) Qualitative Examination of Ash.-Detection of Bases."':""An easily fusible
ash, dissolving completely in water to give a strongly alkaline solution, indicates a grease containing sodium, potassium, or both. A white infusible ash,
practically insoluble in water, but imparting to it an alkaline reaction, may
indicate calcium, with or without magnesium or aluminium. Zinc is shown by
the yellow color of the ash whilE,) hot, and lead may be indicated by the pres~e
of metallic globules, or by the yellow color of the ash w~en cold.
The ash shall be dissolved in dilute nitric or hydrochloric acid, and the
presence of "the several bases confirmed by suitable chemical tests, any standard
scheme of qualitative analysis being followed.
I
(C) Quantitative Examination of Ash.-For.Jhe 'quantitative examination
of ash any standard analytical procedure shalI"be used, the choice of methods

PETROLEUM AND ITS PRODUCTS

1825

being based on the information gained from the qualitative tests. If only one
base is present, a quantitative determination is in general unnecessary.
(D) Determination of Fillers, Soap, Fat, Petroleum Oil and Unsaponijiable
Matter.-Cup, fiber and sponge greases, with or without graphite, and all
comparatively light colored greases, including axle greases, shall be exa,mined
by method (1). Dark greases containing residuum, asphaltic oils and asphalt,
tllrs, etc., shall be examined by method (2), as these ingredients usually cause
stubborn emulsions if a shaking-out process is applied at the start.
(1) Determination of Filler, Soap, Fat, Etc.-If no fillers are present, the
grease sample shall .be introduced directly {rom a weighing bottle into a separator, the weight of the sample being obtained by difference. The sample shall
then be shaken with 75 m!. of petroleum ether and 50 m!. of 10% hydrochloric acid until the grease is completely decomposed, after which the procedure given under Determination of Soap shall be followed. If fillers are
present, the sample (See Sample) shall be weighed in a small beaker, 50 m!.
of 10% hydrochloric acid added and the beaker, warmed on .the steam bath,
stirring until all soap lumps have disappeared and the upper layer is clear.
If undissolved mineral matter or other filler is present, both layers shall be
filtered through a Gooch crucible provided with a suitable mat, the beaker
and. crucible washed with water and petroleum ether, and the crucible finally
washed with strong alcohol, the alcohol washings being collected separately
and discarded. The crucible and contents shall then be dried at 1200 C. and
weighed, and the result reported as " percentage of insoluble matter" (graphite,
mica, talc, asbestos, gypsum, woo~ pulp, etc.).
Determination of Gypsum.-If gypsum is present as a filler, the results obtained in the determination described under Fillers, etc. will be too low because of
the solubility of calcium sulfate in hydrochloric acid. A sample of about 5 g.
of grease shall be weighed in a small beaker, and decomposed with 50 m!. of
petroleum ether and 25 ml. of concentrated hydrochloric acid, warming on the
steam bath until all gypsum is dissolved. The contents of the beaker shall be
cooled and poured into a separatory funnel, the beaker rinsed with water and
a little petroleum ether and the solutions in the funnel allowed to separate
clearly. The lower layer shall then be drawn into a beaker, and the petroleum
ether in the separator washed with two portions of 20 m!. each of 10% hydrochloric acid, and the washes added to the strong acid solution in the beaker.
The hydrochloric acid solution in the beaker shall then be evaporated on the
steam bath almost to dryness, diluted of 150 m!. with water, heated to boiling
and treated with 10 m1. of 10% barium chloride solution. The barium, sulfate
. shall then be filtered off and ignited in the usual way, and from the weight of
the barium sulfate, the percentage of calcium sulfate shall be calculated, reporting it as CaS042H 2 0.
NOTE.-Chaik and other forms of alkaline earth carbonates are not determined
by this method, but are considered under the determination of free alkalinity.

Determination of Soap.-The solution resulting ffom the de<!omposition of

the filler-free grease or the combined filtrate and washes from the determination of fillers, except the alcohol, which is discarded, shall be placed in

PETROLEUM AND ITS PRODUCTS

1826

a separator, allowed to clear, and the aqueous layer (A),24 which contains all
the bases as well as glycerin, shall be drawn into 'another separator. The

METHOD

~-ETHOO 1
Sample Decomposed by HCLand P.E.

Sample Dec.omposedbyKHSo..and
Extracted by PUn 50xhlet Apparatus

Filfrt,. thro~gh Gooch

I .

filtrate('Z~yers)

Insoluble$

ASBESTOnALC,~,

$qxrrrm

PE.sJ,n . (B)

Residue
(lno'!J'l'X>
DiWJrrJ.

~,~,6RAPKITt,~

Acid AquJus Sol;',. (A)

T.. co. of FItly Ac;d,P.Q, Chlorid..,

'winrndKC'

Tot.~~s~~a.I,p'o"
Wash JT,inlS wifhWalu

6,

r--L----Watei"

Traces of HCl
Wash TWI" withRE.
It

(fOTes,
So/u#OI'lS is Shown by 8randting Lin,s,

Wash Oneil WI'" Wafttr.

tunnel Fol/OW5 th. l?'ghl8ronch and

,*, LIi/hJIr.)Qlvl<lMFul1oWJ '1M L8_ff

P.E!..(C)

IJranoh

, Tra,.. 01 rattyAdd,P.o.

"rom here, MsModl

iarHllaresams

P.E.&:. P.frofetJrrf [thet:

CtO-Cthyl Cfher.

a. Petroleum Oil,.

R.stuvI (or Gty,,,.,

TNt.

Wa{er

TrsttsofHCl.

Oiscord.

Tdrt1te wrl}, KOH and SepatTTf_.

P. E. So In

I
=

(E)

ht,po,and T..". of Rol.,h Soaps-

Wa,h 3 Tim ej/Wldh50~/U:CohOI
_.

""

I , )

/ h f

Alto ollCWashlngs

Trta1ea$n,O: ~O~$

Alcoholic 501n(D

Pol.sh Soaps,T..u,clfatUo.

Pla~lnSepolT1for.
WashOmewilhP.[.
I
p.E.wa~hinCjS
AICOhotic'SOI'n.(O)

7"

Traces of Fatand P'Q.

1----<E;------oE-E---I1

PE

ChIQrid'~HC\,Gly",rin.

T .. ""ofFatty Acid, PO. and "CI.

IIttHll1v,"er(lowerjJo/ufloriIl7SepomfrJ"_

Acid Stln,(A)

P.E,Washlnqs=(c)

Whim'tra Separu/lonof? Imml$Cihft

So"n.=~E)

fafandPO.

Potash Soaps
fvapomt.inB.Qk.r,fI_
in Hoi Wafv,4cidify.Shah
OV-lfw/C4Wlfh[f(1.

Saporllf'y with I(OH,

,J
EtO.Soln.D(F)

7ifrall! and SepOI'l1M

AkoholiJ 50In.v (H)

Fatty icid and Traces Het.

Wa.1hTwio wJ/hWanr.

Pol.,h So,ps;Tra. of PD.

Wa~r
KC1, HCL

Acid

Olwzrd.

Water
HCL

DiJrar#.

1,

Acid Water
KCI, He!.
Discard.

FIG. 264.-Scheme of Grease Analysis.

petroleum ether layer (B) shall be washed three times with 25 ml. of water
remove hydrochloric acid, adding the washings to (A)! Solution (S) shall
24 Solutions designated by capital letters in parentheses refer to the solutions given
in Fig. 264.
to

PETROLEUM AND ITS PRODUCTS

1827

then be run into a 250-ml. separator. The aqueous solution (A) shall then
be washed twice with 20 mi. of petroleum ether (C), whereupon (A) is set
aside for examination for glycerin. Solution (C) shall then be washed
once with 15 mi. of water, which is rejected, and (C) sha,ll be added to (B).
If (B) and (C) are comparatively light colored, 'an approximate determination
of free fatty acids and fatty acids from soap may now be made by titrating the
solution in the separator with '0.5 N alcoholic potassium hydroxide and phenolphthalein, ,using 200 as the average neutralization value of the fatty acid (that
is, 1.0 g. of fatty acid requires 200 mg. of absolute KOH for neutralization).
If solutions (B) and (C) are dark, a few drops of phenolphthalein and sufficient
0.5 N alcoholic potassium hydroxide shall be added to make the alcoholic layer "
distinctly alkaline after vigorous shaking. If the solutions have been titrated,
a slight excess of alkali shall be added before separating. The conservative
addition of strong alcohol at this poin.t may aid in securing rapid and sharp
separation. The two solutions shall be allowed to separate sharply and the
lower alcoholic layer (D) drawn off into another separator. The upper pelroleum ether layer CE) shall be washed three times, with 30, 25 and 20 mi. of
neutral 50% alcohol respectively, adding these washes to (D). Solution CD)
shall be washed with 25 ml. of petroleum ether, after which CD) is drawn off into
a beaker and the petroleum ether added to CE). Solution CD) shall then be
evaporated to a small volume to remove alcohol, the residue of potassium soap
washed into a separator with hot water, acidified with hydrochloric acid and
shaken out twice with 50 and 25 m!. of ethyl ether (F), respectively. Solution
(F) shall be run into another separator and washed twi~e with 20-ml. portions
of water, which is discarded. Solution CF) shall then be transferred to a weighed
'beaker and evaporated to dryness on the steam bath, blowing with air to remove
. all traces of petroleum ether. The residue consisting of free fatty acid and
fatty acid from soap shall be heated for a short time on a steam bath, adding
and evaporating 5 m!. of absolute alcohol to remove the last traces of water,
and weighed. The exact neutralization value shall then be determined on as
large a sample of these fatty acids as possible. From the total fatty acid found
here, the free acid if any (see Sections on Free Alkali and Free Acid) shall be
deducted and the remainder calculated to percentage of soap, the ash analysis
being used as a guide in distributing the bases, if more than one are present.
The fatty acids may be identified to some exte~t by special tests, such as
odor, crystal form, melting point, iodine number, neutralization value, color
reactions, etc.
If the grease is appreciably oxidized, the fatty acids obtained by the ethyl
ether extraction are likely to be dark in color and hard to identify. For further
study and identification, the neutralized acids may be extracted again (qualitatively) with petroleum ether and dilute Hel, thus eliminating the small amount
of more darkly colored matter.
Detection of Glycerin.-In case it is desired to determine whether a grease
has been made from whole fats or from fatty acids, solution (A) shall be neutralized with dry sodium carbonate and sufficient excess added to precipitate
calcium or other metals. The whole mass shall then be evaporated to dryness,
the residue extracted several times with strong alcohol, the combined alcoholic
extracts filtered, and the alcohol evaporated. The residue will then contain

1828

PETROLEUM AND ITS PRODUCTS

most of the glycerin, with a little sodium chloride. The presence or absence
of glycerin in the residue shall be confirmed by suitaple qualitative tests.
Since fats on saponification yield between 10 and 11 % of glycerin, and this
analytical scheme does not include a quantitative determination of glycerin,
the total percentages in greases made from fats should always be 'less than
.
100. 25
Determination of Fat.-The petroleum ether solution (E), containing
free fat, petroleum oils, and unsaponifiable matter shall be concentrated
to a volume of about 125 ml. in a 300-ml. Erlenmeyer flask, 10 ml. of 0.5 N
alcoholic potassium hydroxide and 50 ml. of strong neutral alcohol added,
and the whole boiled on a hot plate with an air condenser for lYz hours. The
uncombined alkali shall be titrated with 0.5 N hydrochloric acid, and from the
alkali consumed, corrected for the blank determination as specified under Solvents, the percentage of free fat shall be calculated, using 195 as the average
saponification value (that is, 1.0 g. of fat requires 195 mg. of absolute KOH
for -sa ponifica tion) .
The titrated solution (G) shall be placed in a separator, the alcoholic lower
layer (H) drawn off into another separator and the remaining traces of soap
removed as described for (B) and (C), Determination of Soap, making only two
washes with 30 and 20 m!. of 50% alcohol (I). Solutions (H) and (I) shall be
combined, washed once with a little petroleum ether, which is added to solution (G), and solutions (H) and (1) then evaporated to. a small volume and
the fatty acid isolated as described for eD), Determination of Soap. The percentage of fat may be checked by weighing and titrating the free fattyacid. 26
Determination of Petroleum Oi1.-The petroleum ether solution (G), now
containing all the hydrocarbon oils and unsaponifiable matter shall be evaporated to dryness in a weighed beaker as described for (F), (Soap), the residue
weighed, and the result reported as" petroleum oils plus unsaponifiable matter."
An approximate determination of the viscosity of the petroleum oil may be
made by using a 2 or 5-ml. pipette which has been standardized against oils
of known viscosities. If a complete characterization pf the petroleum products
is required, a new sample of from 150 to 200 g. of grease shall be decomposed
as described above, except that all quantitative operations shall be omitted,
as well as the isolation of the free fatty a.cids, and the use of standard alkali and
acid is not necessary.
'
If the grease contains rosin oil, beeswax, degras, spermaceti; spermatin,
montan wax, candle pitch, and other materials containirrg-a Jll::rge amount of
unsaponifiable substances, the petroleum oils isolated from solution (G) will
contain the unsaponifiable matter, and the constants found will differ from
those of the petroleum products used in making the grease. In most cases,
no further separation is possible except in the hands of skillful and experienced
operators who can devise special methods to suit the individual conditions.
(2) Determination of Filler, Soap, Fat, Etc.-ethod for Dark Greast!tThe sample shall be weighed in a 3-in. porcelain dish, 10 g. of granulated acid
25 A close approximation to the amount of glycerin present may be calculated and
included in the analysis, by taking 11.0% of the weight of the fatty acids from soaps,
providing, of course, that the grease was made from neutral fats. I
26 Multiplying the weight of fatty acid by 1.045 giY~'a very close approximation
of the weight of fat from which it was derived. This-factor varies very little with the
molecular weight of the fat.

PETROLEUM AND ITS PRODUCTS

1829

potassium sulfate added and 10 g. of clean dry ignited sand. The dish and
contents shall be heated oh the steam bath with frequent stirring until all water
is driven off, two hours usually being sufficient. After cooling and breaking up
lumps with a small pestle, the mixture shall be transferred quantitatively to an
extraction thimble, using a little petroleum ether to wash the last traces into
the thimble, which should be already placed in a Soxhlet apparatus. The
thimble shall be extracted thoroughly with petroleum ether, the extract concentrated somewhat if necessary, and the free fatty acid and fatty acid from
soap titrated with 0.5 N alcoholic potassium hydroxide as in solutions (B) and
(C), in method (1), Determination of Soap. From this point on, the procedure is identical with that given in method (1), for soap, fat, and petroleum oil.,
The thimble shall be extracted a second time with carbon disulfide, the
extract evaporated to dryness, heated for one hour to 120 0 C. and weighed,
the results being reported as asphaltic and tarry matter. The residue in the
thimble shall be discarded.
(E) Determination of Free Alkali and Free Acid.-From 10 to 30 g. of
gre~se shall be weighed in a small beaker, dissolved as completely as possible
in 75 ml. of petroleum ether by stirring with a spatUla, the mixture washed
into a 250-ml. Erlenmeyer flask with a small amount of petroleum ether, and
the beaker rinsed with 50 ml. of 50% alcohol, the alcohol being poured into the
flask, after which a few drops of phenolphthalein solution shall be added and
the whole shaken vigorously. If the alcoholic layer, after settling for a few
seconds, is pink, 10 ml. of 0.5 N hydrochloric acid.shall be added, the solutions
boiled on a hot plate for ten minutes to expel carbon dioxide, and the excess
acid titrated back with 0.5 N alcoholic potassium hydroxidE). The free alkalinity shall be calculated in terms of hydroxide of the predominating base.
Insoluble Carbonates.-If chalk or any other form of alkaline earth carbonate, or lead carbonate, is present as a filler, it will be detected by effervescence on adding the hydrochloric acid. As the amount of such carbonates
is likely to be considerable, the volume of 0.5 N hydrochloric acid added shall
be increased sufficiently to dissolve all carbonate and leave a slight excess of
acid. The solution shall be boiled for two minutes and the excess acid titrated
back with 0.5 N alcoholic potassium hydroxide, and from the acid consumed,.
its equivalent in calcium carbonate, etc., shall be calculated, disregarding any
other forms of alkalinity which may have been present.
Free Acid.-If the original alcoholic layer is not pink, the solution shall
be titrated carefully in the cold with 0.5 N alcoholic potassium hydroxide,
shaking well after each addition. The acidity shall be calculated as oleic acid.
If soaps of iron, zinc, aluminum, or other weak bases are present, a determination of free acid is not possible, since these met:tllic soaps react with
potassium hydroxide. Up to the present, no means has been devised whereby
this determination can be made directly.
(p) Determination -of Water.-The quantity of water shall be determined
in accordance with the Standard Method of Test for Water in Petroleum Products and Other Bituminous Materials (A. S. T. M. Designation: D 95) of the
American Society for Testing Materials. 27
27

1933 Book of A. S. T. M. Standards, Part II, p. 891.

1830

PETROLEUM AND ITS PRODUCTS

Waxes

The melting point of mineral waxes may indicat;lits origin and determines
its classification. A wide range of temperature in th~ melting shows a mixture
of waxes and incomplete refining methods.
I
.

STANDARD METHOD OF TEST FOR MELTING POINT 0F

PARAFFIN WAX 28

A. S. T. M.

DESIGNATION: D 87-2229

Definition.-A. S. T. M. Paraffin Wax Melting point.-The A. S. T. M.

Paraffin Wax Melting Point is the temperature at which melted paraffin wax,
when allowed to cool under definite prescribed conditions, first shows a minimum rate of temperature change.
NOTE.-The so-called "American Melting Point" is an arbritary figure 3 F. higher
than the A. S. T. M. Paraffin Wax Melting Point.
Apparatus.- Wax Container.-Test tube 6f standard form, 25 mm. (1 in.)
outside diameter and 100 mm. (4 in.) long., It may be marked with a filling
line, 2 in. above the bottom. This test tube shall be closed in a tightly fitting
cork having two openings, 'one at. the center for the melting point thermometer
and the other for a stirrer at one side of the center. The opening for the stirrer
may be lined with glass or metal tubing to act as a guide for the stirrer.
Air Bath.-Suitable water-tight cylinder, 2 in. in inside diameter and
4,72 in. deep. This air bath shall be provided with a tightly fitting cork having
a central opening for holding the test tube firmly in a vertical position in the
I
center of the air bath.
Water Bath.-Suitable cylinder, 5Ys in. in inside diameter and 6 in. deep.
This water bath shall be provided with a suitable cover and with the guides and
fasteners necessary to hold the air bath firmly in a vEfrtical central position so
that the sides and bottom of the air bath sha~l be ~urrounded by a layer of
.water 1,72 in. thick. The water bath cover shall have a slot for introduction
of a suitable stirrer and shall have an opening for the bath thermometer so
that the latter may be suspended in a vertical position ~ in. from the outside
wall of the water bath. Air bath, water bath, and water bath cover may be
conveniently made of metal in one assembly as shown in-...Fig. 264.
Stirrer in Test Tube.~Brass or copper wire, Ys in. in dianieteJ; and about
12 in. long. A circular loop, ,72 in. in diameter, shall be formed.at one end of
this wire in sU9h a manner that the loop lies in a horizontal plane when the rest
of the wire is in a vertical position. The stirrer thus formed ,shall be passed
through the proper opening in the test-tube cork and the upper end may then
be bent into a shape convenient for holding.,

28 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products 'and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
J
29 This method is issued under the fixed designation)_87; the final number indicates the year of original adoption as standard or, in We--case of'revision, the year of
last revision. Issued as Tentative, 1921; Adopted, 1922. This method has been
approved as American Standard by the American Standards Association.

PETROLEUM AND ITS PRODUCTS

1831

TheTmomeieT.-The ,A. S. T. M. Paraffin Wax Melting Point thermometer

shall conform to the following requirements. These specifications cover a
special thermometer graduated in either Centigrade. or Fahrenheit degrees as
specified, the ranges being 27 to 71 C. or 80 to 160 0 F., respectively.

f,!

-G

1'!

t
:

:e:

~I\Z

~-----------~~:----------__'jFIG. 264a.-Apparatus for Determination of Melting Point of Paraffin Wax.

TYPE: Etched stem, glass.

LIQUID: Mercury.
RANGE AND SUBDIVISIOIJ: 27 to 71 0 C. in 0.10 C. or 80 to 1600 F. in 0.20 F.
TOTAL LENGTH: 365 to 371 mm. (14.38 to 14.56 in.).
.
STEM: Plain front, enamel back, suitable thermometer tubing. 'Diameter, 6.0 to 7.0
mm. (0.24 to 0.28 in.).
BULB: Corning normal or equally suitable thermometric glass.
Length, not over 28 mm; (1.10'in.).
Diameter, not greater than stem.
DISTANCE TO 27 0 C. OR 800 F. LINE FROM BOTTOM OF BULB: 105 to 115 mm. (4.13
to 4.53 in.).
DISTANCE TO 71 0 C. OR 1600 F. LINE FROM Top OF THERMOMETER: 25 to 40 mm. (0.98
to 1.57 in.).

1832

PETROLEUM AND ITS PRODUCTS

CONTRACTION CHAMBER: Top to be not more than 41 mm. (1.61 in.) above bottom
of bulb.
.
.
I
EXPANSION CHAMBER: To permit heating the thermometer to at least 100 C. (212 F.).
FILLING ABOVE MERCURY: Nitrogen gas.
I
Top FINISH: Plain.
,
GRADUATION: All lines, figures, and letters clear cut and distinct. The first and each
succeeding fifth line to be longer than the remaining lines. Graduations to 'be
numbered at each 1 C. or each multiple of 2 0 F.
'
IMMERSION: 79 mm. or 3% in. The words "79-mm. immersion" on Centigrade thermometers or "3Ys-in. immersion" on Fahrenheit thermometers, and a line aronnd
the stem 79.0 mm. or 3.13 in. above the bottom of the bulb shall be etched on the
stem.
SPECIAL MARKING: "A. S. T. M. Pffe. M.P.," a serial. number and the manufacturer's
name or trade mark shall be etched on the stem.
SCALE ERROR: The error at any point of the scale when the thermometer is standardized as proTided below shall not exceed 0.1 C. or 0.2 F., respectively.
STANDARDIZATION: The thermometer shall be standardized at intervals of approximately 10 C. or 20 F. for 79-mm. or 3 i-in. immersion and for an average tem- ,
perature of the emergent mercury column of 25 C. or 77 F.
CASE: The thermometer shall be supplied in a suitable case on which shall appear
the marking: "A. S. T. M. Pffe. M.P., 27 to 71 0 C." or "A. S. T. M. Pffe. M.P.,
80 to 160 F." according to the type of thermometer.
NOTE.-For the purpose of interpreting these specifications the following definitions apply:
The total length is the over-all length of the finished instrument.
The diameter is that measured with a ring gage.
The length 0; the bulb is the distance from the bottom of the bulb'to the beginning
of the enamel backing.
. /
The top of the thermometer is the top of the finished instrument.

Bath thermometer of any suitable type, accurate to 2 F. throughout the

required range.
Procedure.-An average sample of the wax to be tested shall be melted iIi
a suitable container, in a water bath whose temperature shall be not more than
35 F. above the approximate melting point of the wax sample. Direct heat,
such as a flame or hot plate, shall not be used and the wax sample shall not be
held in the melted condition any longer than neces~ary.
The test tube shall be filled with melted wax to' a height of 2 in. The testtube cork, carrying the stirrer and the melting point thermometer with the
3Ys-in. immersion line at the under surface Of the cork, shall be inserted into
the test tube for a distance of Y2 in. The lower end of the thermometer bulb
shall then be % in. from the bottom of the test tube.
The air bath being 'in its proper position in the wat;r-bath, the latter shall
be filled to within Y2 in. of the top with water at a temperature 15 to 20 F.
below the approximate melting point of the wax sample.'
~
The test tube containing the melted wax, :with wax stirrer and thermometer
in place, shall be inserted into the air bath in a central vertical position so that
the bottom of the test tube is Y2 in. from the pottom of the air bath. The
temperature of the water bath shall be adjusted by stirring if necessary, so
that it shall be lower than the temperature of the wax sample by not more than
30 F. and not less than 25 F., when the wax sample has cooled to a temperature
10 F. above its approximate melting point.
When these conditions have been obtained,~ tempe~a'ture adjustment and
stirring of the water bath shall be discontimie1L The wax shall be stirred
continuously during the remainder of the test, the stirring loop being moved

PETROLEUM AND ITS PRODUCTS

1833

up and down throughout the .entire length of the test tube in a steady motion
at the rate of 20 complete cycles per minute. The melting point thermometer
reading, estimated to 0.10 F. shall be observed and recorded every 30 seconds.
The temperature of the wax will fall gradually at first, will then become almost
constant and will then again fall gradually.
'
The melting point thermometer reading, estimated to 0.10 F., shall be
observed and recorded every 30 seconds, for at least three minutes after the
temperature again begins to fall after remaining'almost constant. The record
of temperature readings shall then be inspected and the average of the first
four readings that lie within a range of 0.20 F. shall be considered as the uncorrected melting point. This temperature shall be corrected if necessary for
error in the thermometer scale and the corrected temperature shall be reported
as the" A. S. T. M. Paraffin Wax Melting Point."
.
Accuracy.-Duplicate determinations on the same sample should differ by
not more than 0.20 F.
Heavy Oils and Asphalts
The heavy oils may be crude oils with a low Baume gravity or residues
from the distillation of mixed or asphaltic base crudes. They are used for fuel
oil, dust laying oils and road oils. Asphalts may be found in nature in a semiliquid. (Baume gravity 10) or a- mixture of sand and limestone. Artificial
asphalts are the residues of asphaltic base crude oils and may be polymerized
into a hard asphalt by blowing air through the residues at high temperatures.
See the chapter on Bituminous Substances for methods of testing emulsified
. Asphalts.
The testing of fuel oils is of little value unless the specifications for the
individual oil are given. The many types of burners and the various conditions
under which fuel oils are used make anyone set of specifications impossible to
use. The term fuel oil as now used is broad in meaning and may range from
a mineral seal distillate to heavy residues or a mixture of oils h~ving no other
apparent use .
. The gravity of fuel oils gives little information unless other information is
obtainable. The flash point is important and should be higher than the temperature at which the oil is sufficiently fluid for the particular apparatus for
which it is desired. A low flash shows a naphtha and should not be used in open
burners. The viscosity of an oil must be low enough so that it will flow readily
at low temperatures. Water and sediment (page 1761) choke the burners and
freeze in the pipes duringJow temperatures. The fuel 30 should not contain grit,
acid, fibrous or other foreign matter and shall pass through a filter of wire
gauze 16 meshes to the inch.
The sulfur content when high (over 2.5%) indicates a residue of a high
sulfur crude or a mixture of fuels obtained from acid recovery plants. A high
sulfur fuel causes rapid corrosion of all exposed metal parts of stills and boilers.
ao U. S. Bureau of Mines Bulletin 323A.

1834

PETROLEUM AND ITS PRODUCTS

STANDARD METHOD OF TEST FOR THERMAL
VALUE OF FUEL OIV i

A. S. T. M.

DESIGNATION: D

24of27 32

Apparatus.-Combustion Bombs.-The Atwater, pavis, Emerson, Mahler,

Parr, Peters, Williams or similar bombs may be used. The bomb shall have
an inner surface of platinum, gold, porcelain enamel or other material whi~ is
not attacked by nitric and sulfuric acids, or other products of combustion.
Calorimeter Jacket.-The calorimeter shall be provided with a vacuum
jacket or with a water jacket having a cover to protect the calorimeter from air
currents. In case a water jacket is used, it shall be kept filled with water within
2 or 3 C. of the temperature of the room (except in calorimeters which are
totally submerged, where the jacket temperature is controlled by a thermostat)
and this water should be stirred continuously by some mechanical stirring
device.
Stirring of the Calorimeter Water.-The water in the calorimeter shall be
stirred sufficiently well to give consistent thermometer readings while the
temperature is rising rapidly. The speed of 'stirring should be kept constant.
A motor-driven screw or turbine stirrer is r.ecommended and the speed should
not be excessive. This may be determined, by adjusting the temperature of the
calorimeter to equality with that of the jacket and allowing the stirrer to run
co~tinuously for 10 minutes. If the temperature of the calorimeter rises more
than about 0.01 C. in this length of time; the rate of stirring is excessive.
Accurate results cannot be obtained when too much energy is supplied by the
stirring device or when the rate of stirring is irregular. The portion of the
'stirring device immersed in the calorimeter should be separated from that outside by non-conducting material, such as hard rubber, to prevent conduction of
heat from the motor or outside air.
Thermometers.-The thermometers used shall have been certified by a
government testing bureau and shall be used with th,e corrections given on the
certificate. This shall also apply to electrical resistance or thermo-electric
thermometers. Correction shall also be made for the temperature of the
emergent stem of all mercurial thermometers, .and for the" setting" of Beckmann thermometers. For accurate work, either Beckmann or special calorimetric thermometers, graduated to 0.01 or 0.02 C., are required. Such thermometers should be tapped lightly just before each re1tdin~o avoid errors
caused by the sticking of the mercury meniscus, particularly. when the temperature is falling. A convenient method is to mount a small-electric ~zer
directly on the top of the thermometer and connect it with a dry cell and avush
button. The button should be pressed for a few seconds immediately before
each reading.
I
3i Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. Committee D-2 on Petroleum Products and Lubricants.
Reproduced through the courtesy of the American Society for Testing Materials,
260 S. Broad St., Philadelphia, Pa.
32 This method is issued under the fixed designation D 240; the final number indicates the year of original adoption as standard or, in tj:le_case o( revision, the year of
last revision. Issued as Tentative, 1926; Adoptedjn-Amended Form, 1927. This
method has been approved as American Standard 'by the American Standards Association.

1835

. PETROLEUM AND ITS PRODUCTS

Oxygen.-The oxygen used for combustions shall be free from combustible

,material. The bomb when filled should contain at least 5% of nitrogen to
insure complete oxidation of the sulfur. The total amount of oxyge:" contained
in the bomb for combustion shall be not less than 5 g. per gram of fuel oi!.
But the combustion must be complete, as shown by the absence of any sooty
deposit on opening the bomb after firing.
Firing Wire.-The fuel oil in the bomb may be ignited by mean!! of either
iron or platinum wire. If iron wire is used, it should be of about No. 34 B. & S.
gage and not more than 10 cm. (preferably 5 cm.) should be used at a time.
A correction of 1600 calories per gram weight of iron wire burned shall be
subtracted from the observed number of calories.
Standardization.-The water equivalent of a calorimeter can best be determined by the use of the standard combustion samples supplied by the U. S.
Bureau of Standards. The required water equivalent is equal to the weight
of the sample multiplied by its heat of combustion per gram and divided by the
corrected rise in temperature.
The cal~eter shall be standardized by the combustion of standard sample~ suppliWby the Bureau of Standards, and used according to the directions.
given in the certificates which accompany them. A standardization shall
consist of a series of not less than five combustions of each of two of the standard
~aterials which are, sucrose, naphthalene and benzoic acid. The conditions
as to the amount of water, oxygen, firing wire, method of correcting for radiation, etc., under which these combustions are made, shall be the lame as for
fuel oil combustions. In the case of any disagreement between contracting
parties a check standardization may consist of two or more combustions of
standardizing samples.
.
Procedure.-Preparation of Sample.-Approximately enough oil shall be
taken to give a thermometer rise of from 3 to 4 C. After weighing, the sample
should preferably be immediately placed in the bomb and this closed. This
procedure is necessary to avoid sublimation in the use of naphthalene for
standardization .
. Preparation .of the Bomb.-The firing wire, if iron, shall be measured and
coiled in a small spiral and connected between the platinum terminals, using, if
necessary, a piece of platinum wire somewhat heavier than the iron wire to
make the connection. The platinum and the iron shall both be clean. About
0.5 m!. of water should be placed in the bottom of the bomb to saturate with
moisture the oxygen used for combustion. When the crucible is put in place
in the bomb, the firing wire should touch the fuel oil or briquet of standard
material. For the combustion of standardizing samples, iron wire is preferable
to platinum.
Filling the Bomb with Oxygen.-Oxygen from the supply tank shall be
admitted f;llowly, to avoid blowing the oil from the crucible, until a pressure is
reached as indicated by the following table:
Capacity of Bomb, mI.

Minimum Gage Pressure,

Atmospheres

g~g i~ :gg: :::::::::::::::::::::::::::: :::::::::::::::.: : 35

40

400 to 450 ..... , ,,",.,.,.,., .,'., .... ,"',.,.,',., .. " ... 30

450 to' 500, , ., " ."."",.',., .. , .. ,.".,',, .. , ..... ,.,. 27,5
Ab~'v" 5 0 0 , . " " , . , . , " " , . " . " . " . , . , ' , . , . , ' , . , .. ,. ,',. 25

1836

PETROLEUM AND ITS PRODUCTS .

Calorimeter Water.-The calorimeter shall be filled with the required

amount of, distilled water, depending upon the type of calorimeter. The
amount may be determined either by.measurement in a standardized flask or
by weighing. The amount shall be kept the same as that used in the standardization of the apparatus.
i
Temperature Adjustments.-The initial temperature in the calorimeter
shall be so adjusted that the final temperature, after'the combustion; will not
be more than 10 C., preferably about 0.5 0 C., above that of the jacket. Under
these conditions, the total correction for peat gained from or lost to the surroundings will be small when the rise of temperature is 30 or 4 C. and the
effect of evaporation will also be small.
Firing Current.-The electric current used for firing the charge shall be
obtained from storage or dry cells having an electromotive force of not more
than 12 volts. A higher voltage is liable to cause an arc between the firing
terminals, introducing additional heat which cannot be measured with certainty.
The circuit should be closed by means of a switch which should remain closed
for not more than 2 seconds. When possible, it is recommended that an ammeter be used in the firing circuit to indicate when the firing wire has burned
out.
.
Method of Making an Observation.-/fhe bomb, when ready for firing,
shall be placed in the calorimeter, the firing wires connected, the cover put in
place and the stirrer and thermometer so placed as not to be in contact with the
bomb or cf'lntainer. The stirrer shall/then be started' and after the thermometer reading has become steady, not less than 2 minutes after the stirrer
is started, the temperatures shall be read at I-minute intervals for 5 minutes
and the charge then fired, the exact time of firing being noted. Observations
of temperature shall then be made at intervals depending upon the method to
be used for computing the cooling correction. When the temperature has
reached its maximum and is falling uniformly, a series of thermometer readings
shall be taken at I-minute intervals for 5 minutes to determine the final
I
cooling rate.
Titration.-After a combustion and after allowihg the gas to escape, the
bomb shall be opened and the inside examined, for traces of unburned material
or sooty deposit. If these are found, the observations shall be discarded.
If the combustion appears complete, the bomb shall be rinsed out thoroughly
and the washings titrated with a standard alkali solution (1 ml. = 0.02173 g.
HNOs=5 calories) using methyl-orange or methyl-red indica!or to determine
the amount of acid formed. A correction of 230 calories per gram of IWOs
should be subtracted from the total heat observed. An additional cor:t'C.ition
of 1300 calories per gram of sulfur in the oil should be made for the excess of
difference in heats of formation of S02 and aqueous H 2S0 4 over the hea~ of
formation of aqueous HN0 3
Computation of Results.-The following meth6d of computation is recommended to take the place of the Pfaundler or other similar formulas for computing the cooling correction (radiation correction) :
Observe (1) the rate of rise, 71, of the calorimeter temperature in degrees
per minute for 5 minutes before firing; (2) the time, a, at which the last temperature reading is made immediately oefore firjp~; '(3) the time, b, when the
rise of temperature has reached six-tenths of its total amount (this point can

PETROLEUM AND ITS PRODUCTS

1837

generally be determined by adding to the temperature observed before firing,

60% of the expected temperature rise, and noting the time at which this point
is reached); (4) the time, c, of a thermometer reading taken when the temperature change has become uniform some 5 minutes after firing; (5) the final rate
of cooling, r2, in degrees per minute for 5 minutes.
Multiply the rate rl by the time b-a in minutes and tenths of a minute,
, and add to this product (subtracted if the temperature was falling at the time a)
to the thermometer reading taken at the time a. Multiply the rate r2 by the
time c-b and add this product (subtracted if the temperature was rising a~
the time c and later) to the thermometer reading taken at the time c. The
difference of the two thermometer readings thus corrected, provided the corrections from the certificate have already been applied, gives the total rise of
temperature due to the combustion. This, ,multiplied by the water equivalent
of the calorimeter, gives the total amount of heat liberated. Divide this
result, corrected for the heats of formation of HNO a and H 2S0 4 observed and
for the heat of combustion of the firing wire, when that is included, by the weight
of the charge to find the heat of combustion in calories per gram. Calories
per gram multiplied by 1.8 give the British thermal units per pound.
The permissible difference between duplicate determinations made by
different analysts on different calorimeters shall not exceed 0.5%. When made
by the :;ame anafyst on the same calorimeter, the permissible difference sht1ll
not exceed 0.3%:
.
In practice, the time b-a will be found so' nearly constant for a given
calorimeter with the usual amounts of fuel that b need be determined only
occasionally.
The results shall be reduced to calories per gram or British thermal units
per pound of oil.

Road Oils
Road oils may be used as dust-laying oils, binders for sand, sandy loam and .
gravel roads. The value of a road oil depends upon its penetration into the soil,
the binding qualities of the oil, and its resistance to weather and wear.
The tests on road oils are important in as much as the importance of the
specifications for the particular duty for which they are intended. Road oils
are often placed on roads without any study as to the adaptability of the oil for
the soil or base on which it is placed.
Gravity is a common test made on road oils hut has little value. The
asphalt in a road oil acts as the binder and weather-resisting element. The
total bitumen (page 1539) is determined to give the value of the oil as a binder.
The penetration of the asphalt into the road bed determines its wearing
qualities. The viscosity should be low enough to enable to penetrate the soil
and to flow with ease and uniformly when applied. A heavy asphalt oil mixed
with a lighter oil will give greater penetration, the lighter oil slowly evaporating,
leaving a layer of high asphaltic content in the soil. The asphalt should contain
little or no paraffin wax. The wax particles combine with the soil and form
little dry particles which soon wear loose and blow away leaving small pits in
the surface which soon wear away forming large holes and ruts in the finished
roads.

1838

PETROLEUM AND ITS PRODUCTS

Asphalts either native or artificial when used for paving should be tested for
total bitumen, penetration (page 1520), viscosity ~page 1518) and carbene
(page 1546) content. If asphalts are made from petroleum residues, the wax
content should be determined. Petroleum with a cltrbene content over .5%
(part insoluble in carbon tetrachloride) indicates th'at the residue had been
overheated or overblown, making a brittle product. .

Shale Oils
The testing of shale oils is much the :same as the testing of petroleums.
Shale oils are made by the retorting of the oil shales changing the pyrobitumens
into the bitumens. The bitumens formed vary in chemical and physical
properties as the methods of retorting. High temperatures cause" cracking"
with a resulting polymerization. The nitrogen and sulfur content being high
gives a different reaction to heat than petroleums of the same gravity.
The waxes made from shales have a higher melting point than the paraffin
from petroleum. They may be separated into a white flake wax. by washing
the re-run lubricating distillate with acetone. This method can be used as a
quantitative test for waxes in shale oils.
The different'finished products, gasolines, kerosenes, fuel oils, lubricants and
waxes, are tested in the same manner as siniilar petroleum products.

ANALYSIS OF PAINTS
In reporting the results of an examination of a paint, it is advisable to give
all the analytical data as well as a resume showing the probable composition
of the paint. "This is shown in the following example:
RESULTS OF ANALYSIS

Total Pigments or Solids ......................................... 60%

Total Vehicle or J.,iquids .................. : . . . . . . . . . . . . . . . . . . . . . .. 40

Analysis of Pigment Portion

Lead Oxide (PbO) ........................................... . 37.47%
Zinc <?xid~ (Znq) ............................................ . 44.50
AlumIna, Iron, hme ........................................... . 2.90
. Magnesia (MgO) ............................................. . 1.90
Silica (Si02) .................................. '.' ............. . 4.63
Carbon Dioxide (C0 2) ........................................ . 2.50
Sulfuric Anhydride (SOa) ..................................... . 5.02
.73
Water (combined) ........................................ , ... .
99.15%

Analysis of Liquid Portion

Vehicle contained 20% volatile matter.
Volatile matter consisted of equal parts of turpentine and mineral spirits.
Non-volatile matter had:
Iodine Number .......................................... : .. 175
Acid Number. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4
Saponification Number ...................................... 188
and contained .02% ash consisting of lead and manganese o~ides.
Probable Composition of Paint
Pigment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 60%
Liquid ......................................................... 40
Pigment
Basic carbonate White lead. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Rasic sulfate White lead. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Zinc Oxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Asbestine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . ..

22%
25
43
10
100%

Liquid
Raw Linseed Oil ................................................. 80%
Mineral Spirits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10
Turpentine and Drier ................................ : . . . . . . . . . .. 10
100%
Chapter contributed by Henry A. Gardner, John A. Schaeffer and G. G. Sward.

1839 .

1840

ANALYSIS OF PAIN,J;S
SEPARATION OF PIGMENT ANID LIQUID

Composition of Liquid Part.-The vehicle or liqjJ.id portion of paints may

contain various fixed animal, vegetable or mineral qils, oleo-resinous varnishes,
turpentine, petroleum distillates, and driers,
I
It is always advisable to determine the total percentage of liquids in a :paint.
The container should be thoroughly shaken so that the contents will be uniform throughout. A portion of 4 or 5 ounces may then be removed and placed
in a screw-cap bottle. The original can of paint should then be set aside so
that settling of the pigments will take place. Unless the paint is in paste form,
this will usually be accomplished in twen,ty-four hours. A portion of the clear
liquid floating over the pigments may then be removed and ,directly examined
as outlined under Separation of Vehicle Components.
Percentage of Liquid by Ignition Method.-The percentage of liquid in the
uniform sample of paint previously obtained may be found by placing a weighed
por~ion in a porcelain crucible and slowly igniting it to burn off the organic
constituents. By carefully regulating the heat, the oil and volatile thinners
will be slowly burned off; leaving the pigment behind, which may then be
weighed, calculating the vehicle by difference. This method is rapid and works
well with pigments that are not decomposed by the ignition. When pigments
are present which show an appreciable loss on ignition, Qr blacks or blues, this
method is not to be relied upon.
/
,
Percentage of Liquid by Extraction Methods.-From 10 to 20 grams of
paint is placed in a 100-ml. weighed centrifuge tube, the exact amount being
obtained by difference. The paint is thinned with a suitable extraction liquid
and centrifuged. The supernatant liquid is then poured off and the process
repeated two or three times. The pigment is then dried, first in air, and finally
in an oven. The tube and pigment are weighed and the percentage of vO-cle
obtained by difference.
In case a centrifuge is not available, the vehicle of many paints may be
separated by simply shaking a portion of the paint in a long test-tube with the
extraction liquid, allowing the pigment to settle, repeating the extraction until
the oil is thoroughly removed.
Some operators use a Soxhlet extractor for the determination of the vehicle
of a paint. This method is rather slow and does not always give satisfactory
results.,
-..... _
It must be remembered that no method of extraction of theoil from a paint
will give absolute results. The last traces of oil cannot be removed from the
pigment, which is probably due to the fact that many pigments such as lead
and zinc react with the oil, producing small quantities of insoluble soaps which
are not completely dissolved by the solvent.
,
The choice of a solvent is important. For lin~eed oil paints, diethyl ether,
petroleum ether, 'benzene, and other solvents for the oil are suitable. If varnish is present, some experimental work on various solvents may be necessary.
There are some pigments which by reason of their low specific gravity, colloidal nature or partial solubility can never be completely' separated from oil,
either by settling, centrifuging or extraction .._Of -these the most commonly
met with are lampblack and other forms of carDon, zinc oxide and Prussian

ANALYSIS OF PAINTS

1841

blue. Colloidal pigments such as zinc oxide are very troublesome in this
respect. When these pigments, however, are present in a paint in considerable
percentage, the difficulty of their separation may be avoided by adding to the
paint three or four times its volume of fuller's earth, diluting the mixture in a
large test-tube with gasoline or petroleum ether and either centrifuging or
placing in a rack to settle. The tuller's earth carries down the colloidal pigments and the separation is sharp and easy. This method, of course, is simply
used to extract the vehicle present. The pigment resulting from the separation
cannot be used for analysis on account of admixture with the fuller's earth.
In some cases the pigments in paste colors made of lampblack and Prussian
blue cannot be separated from the liquid portion. The amount of Prussian
blue present, however, may be determined by making a Kjeldahl-Gunning
determination on a portion of the entire paint, multiplying the nitrogen found
by 4.4. For the determination of the lampblack present, a portion of the
entire paint may be boiled with an excess of alcoholic potash until all of the
oil is saponified. The mixture is then decanted through a filter and washed,
first .with hot alcohol and then with hot water. This affords a very good
separation of the liquid from the pigment of such paints. By this method,
the Prussian blue which may be present is partially destroyed, the iron content
remaining admixed with the black pigment on the filter.
Separation of Liquid Components.-Whenever possible, it is advisable to
determine the constituents of the vehicle upon the sample that has been removed from the top of the settled can of paint. A weighed portipn of this
vehicle may be placed in a tared flask and attached to a Liebig condenser.
Heating to 1800 C. or lower will drive off nearly all the volatile constituents.
The composition of the distillate may be determined by the methods given
under the Examination of Turpentine. A portion of the residue in the flask,
which consists of oil, driers, gums, etc., may be transferred to a crucible a~d
ignited. The residue may then be weighed and calculated to ash. The ash
should be analyzed for lead, manganese and other driers.
Another portion of the original vehicle may be evaporated in an atmosphere
of CO 2 (prevents oxidation) to remove the volatile constituents. A portion
of the oil residue. may then be examined for iodine number and other constants.
In most instances it would be advisable to make a saponification and extraction
of the fatty acids from this residue, determining the iodine number on the
fatty acids.
Water is best determined by distillation of the paint with a non-miscible
liquid such as toluene. _ The distillate is .collected in a graduated cylinder, the
water settling to the bottom. A better way is to use the apparatus described
by Dean and Stark. l
Direct Distillation for Volatiles.-For a direct determination of the volatile
constituents in a paint, a sample may be distilled in vacuo. - This is easily
managed wherever a vacuum pump is available and avoids the necessity 'of
overheating the oil. If, in order to obtain the fixed oils for analysis, a sample
of the clear vehicle from a settled paint is used, it should not be heated above
150 0 C .. and neither should the solvent be volatilized in such a way as to allow
the oil to be in contact with air, as the oil will oxidize rapidly while warm and
1

Ind. & Eng. Chern., 12, 486 (1920). See p. 1340.

1842

ANALYSIS OF PAINTS

its iodine number be very much lowered. :The volatile matter may also be
I
separated by steam distillation.
Detection of Resinates.-To determine whether th~ drier in a paint is of
the resinate type or linoleate type, a few drops of tlte oV liquid may be mixed
on a porcelain plate with one or two drops of acetic anhydride, subsequently
adding a drop of sulfuric acid. Upon the addition of the sulfuric acid, a flash
of purple color, turning to dark brown, will be shown ~here rosin is present.
This test is not conclusive and experience is necessary for its proper interpretation. If rosin should be present in the vehicle to a considerable extent, the
oil will have a very high acid number. The approximate percentage of rosin
present may be determined by shaking a portion of the vehicle with 95%
alcohol in a separatory funnel, subsequently separating'the alcoholic extract, .
evaporating and weighing the re'sidue.
Detection of Various Oils.-Tung (Chinese wood oil) may be detected in
the vehicle by mixing the oil with an equal volume of a saturated solution of
iodine in petroleum ether, allowing the mixture to stand in direct sunlight.
Under these conditions, a peculiar, insoluble, spongy polymer of one of the
glycerides of Chinese wood oil is shown. Fish oil can sometimes be detected
by its odor and the dark red color during saponification, but the best method
is probably that of Eisenschiml and Copthorne 2 which depends upon the
formation of insoluble bromine derivatives of the oil. The presence of soybean and other vegetable oils is in some Gases difficult to d,etect. The iodine
numbers of these oils, however, are all lower than that of linseed oil. It must
be remembered, however, that the iodine number of boiled linseed oil is lower
than that of raw oil and that the iodine number of oils extracted from many
paints is usually lower than shown by the original oil. In the presence of
considerable quantities of drier, it is always advisable to extract the fatty
a<!ids from oil and make the iodine number determination upon them.
ft1
Detection of Various Thinners.-The distillate from the paint vehicle mllif
consist of turpentine, mineral distillates, benzol and similar solvents. The
presence pf benzol is readily detected by adding a few,' drops of the distillate
to a small quantity of a mixture of concentrated nitric and sulfuric acids.
Upon heating this mixture, the characteristic odor of nitro-benzol will be
recognized if benzol is present. Mineral distillates from petroleum are easily
detected by the polymerization method given under the Examination of
Turpentine.
I

ANALYSIS OF PAINT

orts

Although linseed oil is used to the greatest extent in paints, some other
oils occasionally find application in the manufacture of special paints. The
following have been used for this purpose: soybean oil, peI:illa oil, corn oil,
cottonseed oil, sunflower oil, lumbang oil and similar~egetable oils; menhaden
.~

Ind. Eng. Chern., 2, 28 (1910).

."

ANALYSIS OF PAINTS

1843

oil, whale oil, herring oil, and similar marine animal oils of relatively high
iodine number.
There are given below specifications for, and methods for, the analysis of
linseed oil, in accordance with the latest practice developed by the Federal
Specifications Board of the Procurement Division of the U. S. Treasury Department. These methods may be followed in examining any of the other oils
mentioned above.
Raw Linseed Oil
Maximum

Foots:
Heated oil, per cent by volume .......................... .
1.0
Chilled oil, per cent by volume .......................... .
4.0
Specific gravity 15.5/15S C ................................ .
.936
4.0
Acid number .............................................. .
Saponification number ..................................... . 195.0
Unsaponjfiable matter, per cent ............................. .
1.50
Iodine number,a Wijs ...................................... .
.2
Loss on heating at 105 to 110 C., per cent ........... , ...... .
Color .................................................... .

Minimum

0.931
189.0
177.0

a When high iodine number type of raw linseed oil is specified by the purchaser, the
iodine ilUmber must be not less than 188 and the oil shall conform to all of the other
requirements.
b Not darker than a freshly prepared solution of 1.0 g. potassium bichromate in
100 ml.. pure concentrated sulfuric acid (sp.gr. 1.84).

Boiled Linseed Oil

Maximum

Time of drying on glass, hours .............................. . 18.0

.2
Loss on heating at 105 to 110 C., per cent .................. .
.945
Specific gravity at 15.5/15.5 C .............................. .
Acid number ........ , ..................................... .
7.5
Saponification number ..................................... . 195.0
1.50
Unsaponifiable matter, per cent ............................ , .
Iodine number, Wijs ....................................... .
Ash, per cent ...................... ; ...................... .
.50
Lead, per cent ............................' ................ .

Minimum

0.937 a

189.0
170.0
.05

a When quick process, not kettle-boiled, oil is called for in contract, the minimum
specific gravity shall be 0.931.

Analysis of Linseed Oil

Loss on Heating at 105 to. ~10 C.-Place 10 grams of the oil in an accurately weighed 100-ml. Erlenmeyer flask; weigh. Heat in an oven at a temperature between 105 and 110 C. for 30 minutes; cool and weigh. Calculate
the percentage loss. Make this determination in a current of dry carbon
dioxide gas.
Foots General Test.-With all materials at a temperature between 20 and
27 C., mix, by shaking for exactly one minutli in a graduated tube, 25 ml.

1844 ,

ANALYSIS OF PAINTS

of the well-shaken sample of oil, 25 mI. of acetone, ,and 10 mI. of acid calcium
chloride solution. Then clamp the tube in an, upri~ht position and allow to
settIe for 24 hours. The temperature during this period should be between
20 and 27 C. The graduated tube shall be of not ~ess than 70 ml. capacity
and shall. have at least 50 mI. graduated in 0.1 mI. The diameter of the tube
shall be such that the 50 mI. graduated portion shalll be not less than 40 cm.
or more than 60 cm. in length. The volume of the stratum lying between the
clear calcium chloride solution and the clear acetone and oil mixture is read
to 0.1 ml. or fraction thereof. This reading multiplied by four expresses the
amount of foots present as percentage by volume of the oil taken.
Heated OiI.-Heat a portion of the oil to 65 C., hold it within 2 C. of
that temperature for 10 minutes, then cool it to room temperature (20 to
27 C.). Promptly make the general foots test as described above.
Chilled Oil.-Heat a portion of the oil to 65 C., hold it within 2 C. of
that temperature for 10 minutes, then place it in a clean, dry bottle, stbpper
tightly, and place in a cracked ice and water mixture (0 C.) for exactly two
hours. At the end of this time place the bottle in a large bath of water at
25 C. and keep it there for 30 minutes, then promptly make the genel"al foots
test as described above.
Specific Gravity.-Determine at 15.5/15.5 C. by any convenient method
that is accurate within two points in the fourth decimal place.
Acid Number.-Transfer 5 to 10 grams of the oil to a 300-ml. Erlenmeyer
flask. Add 50 mI. of a mixture of equal'parts by volume of 95% ethyl alcohol
and chemically pure reagent benzol. (This mixture should be previously
titrated to a very faint pink with dilute alkali solution, using phenolphthalein
as an indicator.) Add phenolphthalein indicator and titrate at once to a faint
permanent pink color with standard sodium or potassium hydroxide solution.
Calculate the acid number (milligrams KOH per gram of oil). .
,., .
Saponification N umber.-Weigh about 2 grams of the oil in a 300Ull.
Erlenmeyer flask. Add 25 ml. alcoholic sodium hydroxide solution. Put a
condenser loop inside the neck of the flask and heat on the steam bath for one
hour. Cool, add phenolphthalein as indicator, amI/titrate with half normal
sulfuric acid. Run two blanks with the alcohOlic sodium hydroxide solution.
These should check within 0.1 ml. N/2 H 2S0 4 : From the difference between
the number of cubic centimeters of N/2 H 2S0 4 required for the blank and for
the determination, calculate the saponification number (milligrams KOH required for 1 gram of oil).,
.~
Unsaponifiable Matter.-Transfer 8 to 10 grams of the oil'to a 250-ml.
long-neck flask. Add 5 cc. of strong solution of sodium hydroxide (equal
weights of NaOH and H 20), and 50 mI. 95% ethyl alcohol. Put a condenser
loop inside the neck of the flask and boil for two hours. Occasionally agitate
the flask to break up the liquid but do not project the liquid onto the sides
of the flask. At the end of two hours remove the condenser and allow the
liquid to boil down to about 25 ml.
Transfer to a 500-ml. glass-stoppered separatory funnel, rinsing with water.
Dilute with water to 250 mI., add 100 mI. redistilled ether. Stopper and shake
for one minute. Let stand until the two layers separat~' sharp and clear.
Draw all but one or two drops of the aqueous 14'er into a second 500-ml.
separatory funne,l and repeat the process using 60 illl: of ether. After thorough

ANALYSIS OF PAINTS

1845

separation draw off the aqueous solution into a 400-ml; beaker, then the ether
solution into the first separatory funnel, rinsing down with a little water.
Return the aqueous solution to the second separatory funnel and shake out
again with 60 m!. of ether'in a similar manner, finally drawing the aqueous
solution i,nta the beaker and rinsing the ether into the first separatory funnel.
Shake the combined ether solution with the accumulated water rinsings and
let the layers separate sharp and clear. Draw off the water and add it to the
main aqueous solution. Shake the ether solution with two portions of water
(about 25 ml. each). Add these to the main water solution.
Swirl the separatory funnel so as to bring the last drops of water down to
the stopcock, and draw off until the ether solution just fills the bore of the'
stopcock. Wipe out the stem of the separatory funnel with a bit of cotton on a
wire. Draw the ether solution (portionwise if necessary) into a 250-ml. flask
and distill off. While still hot, drain the flask into a small weighed beaker,
rinsing with' a little ether. Evaporate this ether, cool and weigh. (The
unsaponifiable oil from adulterated drying oils is volatile and will evaporate on
long heating. Therefore heat the beaker on a warm plate, occasionally blowing
out with a current of dry air. Discontinue heating as soon as the odor of ether is
gone.)
Iodine Number.-Place a small quantity of the sample in a small weighing
burette or beaker. Weigh accurately. Transfer by dropping about 0.12
grams (0.09-0.15 grams) to a 500-ml. bottle having a well-ground glass stopper,
or an Erlenmeyer flask having a specially flanged neck for the iodine test.
Reweigh the burette or beaker and determine the amount of sample used.
Add 10 ml. of chloroform. Whirl the bottle to dissolve the sample. Add
10 ml. of chloroform to each of two empty bottles like that used for the sample.
Add to each bottle 25 ml. of Wijs solution and let stand with occasional shaking
for one hour. Add 10 ml. of the 15% potassium iodide solution and 100 ml.
of water, and titrate with standard sodium thiosulfate using starch as indicator.
The titrations on the two blank tests should agree within 0.1 ml. From
the difference between the average of the blank titration and the titration on
the samples and the iodine value of the thiosulfate solution, calculate the
iodine number of the samples tested. (Iodine number is centigrams of iodine
to 1 gram of sample.)
Ash.-Tare a porcelain crucible or dish. Add 10 to 25 ml. of oil, weighing
the amount added. Place on a stone slab on the floor of a hood. Ignite by
playing the flame of a burner on the surface of the oil and allow to. burn quietly
until most of the oil is burned off; then transfer to a muffle or over a flame
and continue heating at a very low temperature (not over a dull red) until all
carbonaceous matter is consumed. Cool, weigh, and calculate the percentage
of ash.
Driers.-Moisten the ash with a few drops of water and test with litmus
paper. Record whether neutral or alkaline. Wash any ash adhering to the
test paper back into the crucible. Dissolve the ash in dilute nitric acid to
which a little hydrogen peroxide .has been added. After solution is complete
make up the volume to about 50 ml. with nitric acid and water so that the
final volume will contain about 1 volume of concentrated nitric acid and 3
volumes of water. Boil to remove excess of hydrogen peroxide. Determine
manganese by the bismuthate method. See Chapter on ManganesE., Vol. 1.

1846

ANALYSIS OF PAINTS

Ash another portion of the oil and dissolve the ash as above in nitric acid
and hydrogen peroxide. Determine the lead and c\)balt by any convenient
method.
Time of Drying on Glass.-Flow the oil over Ii prfectly clean glass plate
and allow to drain in a vertical position in a well-ventilated room at a,tempera.ture between 15 and 39 C. Mter about 2 hours the film is tested at intervals
with the finger at points not less than 2;!1 cm. from the edges. The film will
be considered dry when it adheres no longer to the finger and does not rub up
appreciably when the finger is rubbed lightly acrosS the surface. With boiled
linseed oil this usually occurs in from '5 to 18 hours.
Reagents for Testing.-The following reagents will be required:
Acetone that will pass the specification of the United States Pharmacopoeia.
Acid Calcium Chloride Solution.-Saturate with calcium chloride a mixture
of 90 parts water and 10 parts concentrated hydrochloric acid (specific gravity
1.2).
.
,
Standard Sodium Thiosulfate Solulion.-Dissolve pure sodium thiosulfate
in distilled water that has been well boiled to free it from carbon dioxide in the
proportion so that 24.8'3 grams crystallized sodium thiosulfate will be present
in 1000 m!. of the solution. It is best to let this solution stand for about two
weeks before sta~dardizing. Standardize with pure resubiimed iodine. This
solution will be approximately decinormal, and it is best to leave it as it is after
determining its exact iodine value, rather than to attempt to adjust it to
exactly decinormal strength. Preserve in a stock bottle provided with a guard
tube filled with soda lime.
Slarch Solulion,-Stir up 2 to 3 grams of potato starch or 5 grams soluble
starch with 100 ml. of 1 % salicylic acid solution, add 300 to 400 ml. boiling
water, and boil the mixture until the starch is practically dissolved. Dilute
to 1 liter.
'
Potassium Iodide Solution.-Dissolve 150 grams of potassium iodidelhee
from iodate in distilled water and dilute to 1000 ml.
, Wijs Solution.-The preparation of the iodine ~bnochloride solution presents no great difficulty, but it should be done 1Vith care and accuracy in order
to obtain satisfactory results. There shall be in the solution no sensible excess
either of iodine or more particularly of chlorine over that required to form the
mono chloride. This condition is most satisfactorily attained by dissolving in
the whole of the acetic acid to be used the requisite quantity of iodine, using
a gentle heat to assist the solution, if it is found necessary. ~Dissolve iodine
in glacial acetic acid that has a melting point of 14.7 to 15-C. and is free
from reducing impurities in the proportion so that 13 g. of iodine-will be present
in 1000 m!. of solution. Set aside a small portion of this solution while pure
'and pass dry chlorine into the remainder until the halogen content of the solution is doubled. Ordinarily it will be found that by passing the chlorine into
the main part of the solution until the characteritic color of free iodine has
just been discharged, there will be a slight excess of chlorine which is corrected
by the addition of the requisite amount of the un chlorinated portion until all
free chlorine has been destroyed. A slight excess of iodine does little or no
harm, but excess of chlorine must be avoided.
I
Standard Sodium Hydroxide Solution.- PreplJ,rfra sto,ck concentrated solution of sodium hydroxide by dissolving sodium 'hydroxide in water in the pro-

ANALYSIS OF PAINTS

1847

portion of 200 grams NaOH to 200 ml. water. Allow this solution to cool and
settle in a stoppered bottl~ for several days. Decant the clear liquid from the
precipitate of sodium carbonate into another clean bottle. Add cl~ar barium
hydroxide solution until no further precipitate forms. Again allow to settle
until clear. Draw off about 175 ml. and dilu~ to 10 liters with freshly boiled
distilled water. Preserve in a stock bottle provided with a large guard tube
filled with soda lime. Determine the exact strength by titrating against pure
benzoic acid (C 6H sCOOH) using phenolphthalein as indicator. This solution
will be approxil)lately one-fourth normal, but do not attempt to adjust it to
any exact value. Determine its exact strength and 'make proper corrections ~
in using it.
Alcoholic Sodium Hydroxide Solution.-Dissolve pure sodium hydroxide in
95% ethyl alcohol in the proportion of about 22 grams per 1000 ml. Let
stand in a stoppered bottle. . Decant the clear liquid into another bottle, and
keep well stoppered. This solution should be colorless or only slightly yellow
when used; it will keep 'colorless longer if the alcohol is previously treated with
NaOH (about 80 g. to 1000 mI.), kept at about 50 C. for 15 days, and then
distilled.
Half Normal Sulfuric Acid Solution.-Add about 15 ml. sulfuric acid
(1.84 specific gravity) to distilled water, cool and dilute to 1000 ml. Determine
the exact strength by titrating against freshly standardized sodium hydroxide
or by any other accurate method. Either adjust to exactly half normal
strength or leave as originally made,. applying appropriate correction.
Examination of Tung Oil (Chinese Wood Oil)
So far as they. apply the methods given above for the analysis of linseed
oil and similar oils may be used for the examination of tung oil.' The heating
test and the quality test given below are special tests for tung oil.
Properties and Tests: Properties (A. S. T. M.).-Raw tung oil shall conform to the following requirements:
Maximum

Specific gravity at 1~:~: .. , . . . . . . . . .. . .. . . . . .. . .. .. ... 0.943

Acid number (Alcohol-Benzol) . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Saponification number ............................... " 195
Unsaponifiable matter, per cent ......................... 0.75
Refractive index at 25 C ............................. " 1.520
Iodine number (Wijs) ................................. .
Heating test, minutes (Browne heat test, see below). . . . . . . 12

Minimum

0.940
190
1.5165
163

Gelation Tests for Tung Oils.-One of the unique characteristics of tung

oil is gelation when subjected to heating. This gelation is the ,result of polymerization reactions accelerated by heat and probahly also by certain catalysts.
Browne Heat Test.-This test has been a part of the A. S. T. M. specifications for tung oil for many years. The apparatus as slightly revised in 1935
is shown in Fig. 265.
The test tubes for the oil shall be 15 cm. by 16 mm., with a mark near the
bottom to indicate 5 ml. Each tube shall be closed by a cork so perforated
that a glass rod 3 mm. in diameter may be moved freely. The tubes shall

ANALYSIS OF PAINTS

1848

each weigh 14.52.0 g. (The purpose of ,this requIrement is to control the

wall thickness.)
I
Fill a 1.000-ml. tall-form glass beaker (height 19 c~~, diameter about 9 cm.)
with cottonseed oil or other suitable oil to a height of 11 cm. Place the thermometer as shown in Fig. 265 so that the bottom willl:>e level with the bottoms
of the test tubes.
"
Use a nitrogen-filled chemical thermometer, eng~aved stem, total length
4 to 4Yz in., graduated from 210 to 310 C. in 2 C. intervals; the length
between 210 and 310 C. not less than 2Yz in. Thermometer glass shall be
well annealed. This thermometer may be suspended in the bath by a wire or
may be sealed to a glass rod as showoll in Fig. 265.

A
B

---c:.~
I

FIG. 265.-Apparatus for Browne Heat Test.

A. 1000-ml. tall-form oeaker, 9 cm. in
C. Corks.
D. Test tubes 15 cm. by 16 rom.
diameter by 19 cm. in height.
B. Support plate about 6.5 cm. wide, E. Thermometer, 'range 210 to 310 C.
made of Monel metal, aluminum or
F. Glass rods 3 rom. in diameter.
G. Guide to prevent cover from slipping.
stainless steel.

,
When the bath temperature is 293 C. (560 F.) and very slOwly rising at
this point, place the tube containing 5 ml. of the oil to be tested and the tube
containing 5 ml. of a control sample of known value into the large holes of the
cover. Note the time. Remove the source of heat for about 45 sec. and then
re-apply. Before 2 min. have elapsed the temper~ture of the bath will have
fallen to 282.2 C. (540 F.), at wliich point it sJiould be kept as steady as
possible. When the samples have been in the bath 9 min. raise the glass rods
at intervals of 15 sec. Note the time when each sample becomes firmly set.
At this period the oil will be so stiff that the entire tube may be lifted by aid
of the rod. As setting or jellying takes place :within a few s'econds of fluidity,
a good end point is afforded. Remove the sampJes..:- Heat the bath again to
293 C. (560 F.), and repeat the experiment with fresh portions of the sample.

ANALYSIS OF PAINTS

1849

No stirrer is used in the bath. A screen around the bath enables the temperature to be more easily reached. When the bath oil has become. tarry and
. viscid, it should be renewed; otherwise heating may be irregular.
High quality tung oil should gel in not over 12 minutes by this test. Raw
oiticica oil requires about 18 to 22 minutes.
Worstall Heat Test.-This test consists in heating a quantity of the oil in
an open container. Various refinements have been suggested by investigators,
such as the use of a polished nickel dish, automatic temperature control,
mechanical stirring, etc. However, the following A. S. T. M. method based
on Worstall's test,- and slightly revised in 1935, is probably as rigorous as
needed.
This test requires (1) an ordinary vitreous enamelled iron casserole having
a bottom diameter of 7.5 cm. (3 in.), (2) a wide flange tripod with a 7.5 cm.
(3 in.) opening (the object of the flange is to prevent super-heating of the sides
of the casserole), (3) a total immersion chemical thermometer, range 303000 C. and 14-15 inches in length.
Rut into the casserole 150 g. of the oil and heat so that the temperature
reaches 282.20 C.I (540 F.2) in 4 minutes, 30 seconds, stirring with
the thermometer. Turn down the flame and hold the heat as near 282.2 C.
1 (540 F.2) as possible, stirring with the thermometer, until on lifting
the latter, the oil drops with a pronounced string, showing that polymerization
has started. The time required after reaching 282.2 C.I (540 F.2)
until the string is noted, is the time of the heat test. For pure tung oils .this
will not exceed eight minutes. As soon as the oil strings, remove the lamp
CONSTANTS

OF

Sp. Gr.

Raw Linseed Oil ................

Soybean Oil ....................
Menhaden Oil ..................
Raw Tung Oil ..................
Perilla Oil ......................
Heavy Bodied Linseed Oil. .......
Lithographic Linseed Oil .. '" .....
Whale Oil ......................
Boiled Linseed Oil. ..............
(Linoleate Drier)
Boiled Linseed Oil. ..............
(Resinate Drier)
Corn Oil .......................
Cottonseed Oil ..................
Rosin Oil ..... ' ................
Lumbang Oil. ..................
Sunflower Oil ...................
Hempseed Oil. ........ -.........
Shark Oil ..... ' ................
Sardine Oil ....................
Petroleum Mixing Oil ............
Peanut Oil .....................
Oiticica Oil .....................

VARIOUS OILS
Iodine No.

Sap on. No. Acid No.

Refrac.
Index

.931
.924
.932
.942
.94
.968
.97
.924
.941

186
129
158
166
200
133
102
148
172

188
189
187
183
188
189
199
191
187

2.0
2.3
3.9
3.8
2.0
2.8
2.7
9.2
2.7

1.4867
1.4813
1.4850
1.5200
1.4874
1.4966
1.4978
1.4820
1.4895

.936

184.2

187.6

7.3

1.4895

.921
.920
.964
.927
.924
.927;910
.919
.851
.932
.969

124.8
111.7
68.9
152
124.6
149.4
132.8
134.6
28.2
102.2
148

190:1
194.3
35.5
189
189.3
191.1
158.9
177.3
52.9
188.0
189

4.1
0.9
32.4
1.0
7.5
3.9
5.2
lOA
1.1
2.2
6.0

1.4800
1.4781
-

1.4789
1.4796
1.4822
1.4815
1.4800
1.4773
1.4790
1.5141
~

ANALYSIS OF PAINTS

1850

and the thermometer, and stir with a stiff. spatula until the oil is solid.. After
stringing, a pure tung oil will require not over 40 seco:r;tds more to become solid.
When solid, allow to stand just one minute, then turn out, upside down, on
clean paper and cut at once with a clean spatula. :pure tung oil gives a gel
that is dry, not adhering to the spatula when cut, that; is firm, crumbling under
pressure of the spatula without sticking, and the cut ;portions should crumble
under pressure like dry bread crumbs. Adulterated tung oil gives a gel that
is soft, sticky, and which will not crumble.

Examination of Turpentine
RECOMMENDE]) METHODS OF THE

A. S. T.

M.

ot

Turpentine appears on the market under four types, gum spirits

turpentine, steam-distilled wood, sulfate wood and destructively-distilled wood
turpentine. The physical requirements for the first three types are identical
except for odor. The requirements of the A. S. T. M. specifications for these
types are:
;

Gum Spirits,
Steam-distilled Wood,
Sulfate Wood

Destructively
Distilled
Wood

Maximum
0

Specific gravity, 15.5/15.5 C ............... 0.875

Refractive index (nD) at 20 0 C ............. 1.478
Residue after polymerization with 38 N
H 2SO 4 :
Volume (per cent) ...................... 2.0
Refractive index (nD) at 20 0 C ...........
(This residue shall be viscous and its color
straw or darker.)
Initial boiling point at 760 mm. pressure ...... 1600 C.
Distilling below 1700 C. at 760 mm. pressure
~
(per cent) ....... " .... " ..............

l\1iuimum

l\laximum

Minimum

0.860
1.465

0.875
1.483

0.860
1.463

1.500

1500 C.
90

2
-

1570 C.
-

1.480
1500 C.
90

Color.-Fill a 200-mm., perfectly flat-b,ottom colorimet!ic tube graduated

in millimeters to a depth of from 40 to 50 mm. with the turpentine to be
examined. Place the tube in a colorimeter and place on or un_der it a No. 2
yellow Lovibond glass. Over or under a second graduated tube in the colorimeter, place a No.1 yellow Lovibond glass and run in the same turpentine
until the color matches as nearly as possible the color in the first tube. Read
the difference in depth of the turpentine in the two tubes. If this difference
is 50 mm. or more the turpentine/is" Standard" or better.
Or, compare the color of the sample in any suitable apparatus with the
color of an equal depth of a fresh solution of potassium bichromate in distilled
water containing 0.0180 g. of potassium bichromate per liter. If the sample
is as light as, or lighter than, the bichromate solutio~-;the turpentine is " Standard " or better.
.
.

/'

ANALYSIS OF PAINTS

1851

Specific Gravity.-Determine specific gravity at 15.5/15.5 C. by any

convenient method that is accurate within 2 points in the fourth figure.
Refractive Index.-Determine refractive index at any convenient temperature with an accurate instrument, and calculate the results to 15.5 C., using
the factor 0.00045 for each degree that the temperature of determination differs
from 15.5 C.
Distillation.-The distillation range of turpentine is obtained with special
apparatus and technique which are described in the A. S. T. M. and the Federal specifications. In order to obtain results comparable with those of other
analysts, the details of the test must be followed closely. The instructions.
are too long to include in a book of this character and it is therefore recommended that analysts making this test obtain copies from on{l of the sources
given above.
Polymerization.-Place 20 ml. of exactly 38/N (100.92%) sulfuric acid 3 in
a graduated, narrow-neck Babcock flask, stopper, and cool in ice-water.
Add slowly 5 ml. of the turpentine to be tested. Gradually mix the contents,
cooling from time to time, and do not allow the temperature to rise above
about 60 C; When the mixture no longer warms up on shaking, agitate
thoroughly and place the bottle in a water bath and heat at 60 to 65 C. for
about ten minutes, keeping the contents of the flask thoroughly mixed by
vigorous shaking five or six times during the period. Do not stopper the flask
after the turpentine has been added, as it may explode. Cool to room temperature, fill the flask with concentrated sulfuric acid until the unpolymerized
oil rises into the graduated neck. Centrifuge at about 1200 R.P.M. from four
to five minutes, or allow to stand for twelve hours. Calculate the amount of
'unpolymedzed residue, notice its consistency and color, and determine its
refractive index.

.ANALYSIS OF VARNISH
The testing of varnish should be largely of a physical nature. Such properties as odor, consistency, clarity, flowing, time of drying, character of finish,
hardness, resistance to moisture and abrasion, elasticity, etc., point out the
real value of a varnish. Other tests that give additional information, sometimes
a valuable nature, are as follows: Flash point, acid number, ash,
character of solvent, fixed oil and resins.
.
Flash Point.-A nickel or iron crucible of 60 mm. diameter and 40 mm.
height is filled with the varnish to within 20 mm. of the top. It is then supported in a water bath in such a manner as to be about two-thirds immersed
in the water. The water should be from 15 to 20 C. at the start and should
be heated slowly so that the temperature of the varnish, as indicated by a

0'

3 Made by mixing fuming sulfur.ic acid with co~centrated sulfuric acid.

sel' specification of the A. S. T. M. or of the Federal Specification Board.

For details

1852

ANALYSIS OF PAINTS.

thermometer suspended in it, will show a rise of about 1 degree per minute.
Test for flash at each half degree, using a very small.flame.
Acid Number.-Ten to 20 grams of the varnish ar:e weighed into a small
Erlenmeyer flask, 50 ml. neutral alcohol added, and a small funnel inserted in
the neck. Heat on the water bath for one-half hour, :With occasional shaking.
Allow to cool somewhat, add two drops of phenolphthalein indicator and titrate
with tenth-normal potassium hydroxide solution. The acid number is the
number of milligrams of KOH required to neutralize each gram of the varnish.
Ash.-Weigh in a porcelain or fused silica crucible several grams of the
varnish. Burn off over a small Bunsen flame, using great caution to avoid
boiling over and spattering. When all combustible matter is destroyed, weigh
the ash and if desired analyze it.
,
Solvent.-Steam distillation of a portion of the varnish will remove~the
solvents, leaving a residue of fixed oils and varnish resins, which may be
weighed after driving off the water. The distillate should be examined as
recommended under Methods for the Examination of Turpentine. The
amount of mineral spirits and turpentine may thus be determined.
Fixed Oils and Resins.-In the above determination, the total amount of
fixed oils and resins is obtained. It is a difficult matter, however, to determine
the exact percentage and character of resins that have been used in the manufacture of the varnish. This is due to the fact that during the process of
heating oils in the presence of resins many intricate chemical changes are
brought about, a considerable portion o( the resins being distilled off in the
form of vapors and combinations of the oil brought about that are difficult of
separation. One of the best methods, however, of separating the fixed oils
and varnish resins is carried out in the following manner.
About a half ounce of the varnish resin should be placed in a 300-ml. tared
beaker. There should then be added about 200 ml. of ice-cold petroleum e~r
and the beaker should be covered and allowed to stand, preferably in a dish
containing ice. In an hour's time the resinous ingredients will be found precipitated at the bottom of the beaker or adhering to the side thereof (With the
exception of rosin, which is largely soluble in petroleum ether). The precipitated resins should be washed with fresh portions of cold petroleum et.her two
or three times, pouring the decanted portions into a large bottle. The combined portions of petroleum ether may then be filtered through a tared filter,
adding by the aid of a stirring rod the resins contained in the beaker. The
filter paper and the beaker with the resins may then be {fried-at 100 C. and
weighed. The combined filtrates may be distilled to obtain the fixed oil which
may be examined for constants. (This fixed oil may contain Tosin.) The
amount of rosin contained in a varnish may be roughly ascertained by thoroughly shaking in a separatory funnel a portion of the varnish witlo a large
quantity of absolute alcohol. The rosin may be ol;>tained by evaporation of
the alcoholic extracts. The fixed oils after oxida~ion or polymerization, as
ca,used by the heating of the varnish during manufacture, are not 'readily
soluble in alcohol.
Separation of Polymerized Oils and Resins.-In the making of varnish,
some oils become oxidized or polymerized to a condition resembling resins.
For instance, when a varnish is examined for .re.sill~b-y the above method, it
will often be found that a considerable amount of matter insoluble in petroleum

/'

ANALYSIS

OF

PAINTS

1853

ether will be obtained even when hard resins are absent. The insoluble substance is oxidized or polymerized oil. It may be differentiated from varnish
resins by the fact that it is readily saponified by alcoholic potash. The following method by Boughton (Technologic Paper No. 65, U. S. Bureau of Standards), though involving considerable work, is probably the most accurate
method for the separation of polymerized oils and resins.
To about 4 grams of varnish in a flask add about 25 ml. of water and boil
until the volume is about 10 mI. This removes nearly all of the volatile.
Add 25 mI. half normal alcoholic potash and 25 ml. benzol and boil under a
reflux condenser for one-half hour. Evaporate the solution to about 15 mI.
and add about 10 ml. of alcohol. Transfer completely to a separatory funnel,
washing the flask with water and ether and using a policeman if necessary.
Dilute with water to about 100 mI., add 100 ml: of ether, and shake. Add a
few mI. of alcohol if necessary to make the layers separate. Draw off the
aqueous layer and wash the ether three times with water and transfer to a
tared flask for future use.
TQ the combined soap solution and washings, add an excess of hydrochloric
acid and extract twice with 50 mI. of ether. Discard the aqueous layer, wash
the combined ether extracts with water, transfer to a flask and distill off the
ether. To the dry residue add 20 mI. of absolute alcohol and 20 mI. of a
mixture of 1 volume of sulfuric acid and 4 volumes of absolute alcohol and boil
for two minutes under a reflux condenser. Completely transfer the contents
of the flask to a separatory funnel, washing the flask with water and ether.
Add 100 mI. ether and after agitation add 100 ml. of 10 per cent sodium chloride solution and again shake. Draw off the aqueous layer, extract it with
50 mI. of ether, combine the ether solutions and wash with water. Add 5Q mI.
of a fifth-normal potassium hydroxide solution and 10 ml. of aicohoi, shake
and draw off the lower layer into a second funnel. Wash the ether layer with
50 ml. of water containing 5 mI. of the potassium hydroxide solution and 5 ml.
of alcohol. .
Extract the combined aqueous portions with two 50 ml. portions of ether
and finally wash the combined ether solutions (containing the ethyl esters of
the fatty acids) with water.
Distill off the ether and boil the residue with 25 mI. of half-normal alcoholic
potash for one-half hour under a reflux condenser. Transfer completely to a
separatory funnel and extract the soap four times with ether. Wash the combined ether solution twice with water and add it to the first ether solution of
unsaponifiable matter obtained.
Unite the solution and washings containing the soaps, add an excess of
hydrochloric acid, and extract twice with ether. Transfer to a tared flask the
combined ether solutions, after washing them with water, distill off the ether,
dry the residue to constant weight at 110 C. and weigh as " fatty acids."
Report the percentage of fatty acids as percentage of oil and calculate the
percentage of resin by difference.

1854

ANALYSIS OF PAIN:rS
, .
OTHER MATERIALS I

For detailed methods other than thpse given in ~his book for the examination of shellac, resins, bitumens, and other like materials for the vehicle portion
of paints and varnishes, the following references maYi be consulted:
A. S: T. M. Standards on Preservative Coatings for Structural Materials. American
Society for Testing Materials, Philadelphia, Pa.
Physical and Chemical Examination of Paints, Varnishes, Lacquers and Colors, 9th
Edition, by Henry A. Gardner. Henry A. Gardner Laboratory, Inc., 2201 New
York Avenue, N.W., Washington, D. C.
Federal Specifications of the Procurement Division of the U. S. Treasury Depavtment.
Superintendent of Documents, Government Printing Office, Washingtonjl D. C.

THE ANALYSIS OF PAINT PIGMENTS

,/

The liquid having been extracted from the paint under examination, by
the previously outlined methods, the pigment is left ready for analysis. The
pigment can readily be classified under one of the following heads by its color,
thus shortening any preliminary examination. Many of the colors have a
white base which necessitates a determination of both the colored portion of
the pigment and any white base which may have been. used.
The general analysis of colored pigments is carried out according to the
specific method outlined for the individual colored pigments, together with the
methods for a composite white paint, provided,a qu~litative examination does
not directly reveal the identity of the pigment.!
The pigments used in the manufacture of paints are classified as follows,
in certain instances the trade names being given by which the particular pigments are known.

White Pigments
Lead Pigments

,
Basic Sulfate of Lead-Basic Sulfate White Lead.
Basic Carbonate of Lead-Old Dutch Process White Lead-.:..QuicknProcess
White Lead-Mild Process White Lead.
Zinc Lead-Leaded Zinc.
Zinc Pigments

Zinc Oxide-:-Zinc White.

Lithopone.
Titanated Lithopone.

ANALYSIS OF PAINTS
Other Opaque White Pigments

l'

Titanium Oxide.
Titanium-Barium Pigme~t.
Titanium-Calcium Pigment.
Antimony Oxide.
Lead Titanate.
Zirconium Oxide.
Silica Pigments

Silica-Silex.
Asbestine-Talcose.
China Clay-Kaolin.
Calcium Pigments

Whiting-Paris White-Chalk-Alba Whiting-Spanish White.

Gypsum-Plaster of Paris-Terra Alba-Agalite.
Barium Pigments

Barytes-Barite-Blanc Fixe-Barium Sulfate.

Barium Carbonate-Witherite ..

Red and Brown Pigments

Red Lead-Orange Mineral.
. Vermilion.
Ochres-Tuscan Red-Indian Red-Venetia'i! Red.
Umbers-Sienhas.
Blue Pigments
Sublimed Blue Lead.
Ultramarine Blue-Phthalocyanine Blue.
Prussian Blue-Antwerp Blue-Chinese Blue.

Yellow and Orange Pigments

Chrome YeHow-Lemon Chrome YeHow-Medium Chrome Yellow.
American Vermilion Orange Chrome-Basic Lead Chromate.
Orange Mineral.
Hansa Yellow.
Green Pigments
Chrome Green (mixture of Lead Chromate and Prussian Blue).
Chromium Oxide.
Green Earth.
Black Pigments
Graphite.
Carbon Black-Bone Black-Lamp Black-Drop Black-Ivory Black.
Willow Charcoal.
Black Oxide of Iron.

1855

1856

ANALYSIS OF PAINTS
ANALYSIS OF WHITE PIGMENTS

Basic Sulfate Qf Lead

Basic Sulfate White Lead.

Sublimed White Lead.

Super Sublimed White Lead

This pigment, embracing the daily analysis by the manufacturers of the

product of over five months' time, shows the following average composition:
Lead sulfate .............. : ........................... 76.68%

~:: o~~~~.. :::: ::: :: : ::: ~:: ::: :: ::::: :::: :: ::: :: :: ::

Igg
.99.70

The remaining .3 of 1% consists of moisture and ash which are rarely

determined.
The analysis of this pigment based on the following method,4 which depends
upon the above average composition, together with the volumetric determina-'
tion of the total lead and zinc contents, is rapid and accurate.
Volumetric Determination of Leads

One gram of the sample is dissolved in 100 mI. of an acid ammonium acetate
solution made up as follows:
Eighty per cent acetic acid. . . . . . . . . . . . . . . . . . . . . . . . . . . .. 125 ml.
Concentrated ammonium hydroxide .................... '. 95 ml.
Water ............................................... 100 ml.
Add this solution hot, dilute with 50 mI. water and boil until a clear solution is
obtained. Dilute to 200 ml. and titrate with standard ammonium molybdate
solution, using a fr.eshly pr~ared solution of tannic acid as an outside indicator.
A solution of ammonium molybdate containing ?:67 grams per liter usually
gives a solution where,
One mI.=O.Ol gram Pb. ,
Standardize against pure PbO, pure PbS0 4 , or clean lead foil.
For further details of this method see p. 5i1.
The gravimetric method given under Basic Carbonate-at Lead may be
preferred.
Volumetric Determination of Zinc

Dissolve 1 gram of the pigment in concentrated hydrochloric acid, adjust

the acidity to 3 mI. of concentrated hydrochloric, acid, dilute to 200, heat to
above 60 C., and titrate with standard potassiull). ferrocyanide solution, using
a fresh 5% uranium nitrate solution as an outside indicator. Calculate the
zinc to zinc oxide.
4
6

Ind, Eng. Chern" 6, 200 (1914).

I
Modification of Low's Method, "Technical Methods_of Ore Analysis," p. 149.
_,)-

ANALYSIS OF PAINTS

1857

CALCULATION OF COMPOSITION

Using the average total of 99.70%, the total lead found and the zinc oxide
content, the composition of this pigment is determined by the following calculation:
Total percentage of lead compounds present
'=total percentage found of ZnO, PbO and PbS0 4 -percentage of ZnO.
Total percentage of lead compounds present
=99.70% (average total) -:percentage ZnO.
Atomic weight lead ........"................... : .......... 207.2
Molecular weight lead oxide ......................... : . ... 223.2
Molecular weight lead sulfate . ................. '" ........ 303.2

As a hypothetical case, we can assume the presence of a 4.70% ZnO and

69.00% metallic lead.
(

d) 01 b
.
Mol, wt. PbS0 4 01 Pb f
At. wt. Pb X /0
oun
- /0 P constItuents
=% PbO present
Mol. wt. PbS0 4 -mol. wt. PbO
Mol. wt. PhO

Mol, wt. PbO 01 Pb f

d)
At. wt. Pb X /0
. oun
-

01
/0

Pb constItuents

Mol. wt. PbO - Mol. wt. PbS0 4

Mol. wt. PbSO 4

% PbS0 4 present.

Determining the percentage of lead oxide and lead sulfate present by the
above formulas we find:
303.1
)
( 207.1 X69.00 -95.00
303.1-223.1

per cent PbO = 16.68

223.1

(~~~:~ X69.00 ) -95.00

223~1-303.1

per cent PbS0 4 =78.32.

303.1
If it is necessary actually to determine the percentage of lead sulfate present,
the following procedure may be followed:
Total Lead Sulfate

Mix 0.5 gram of the sample with 3 grams of sodium carbonate. Add 30 ml.
of water and boil gently for ten minutes. Allow to stand for four hours.
Dilute with hot water, filter and wash thoroughly. All the lead sulfate is here

1858

ANALYSIS OF PAINTS
,

changed to lead carbonate, the sodium carbonate being transposed to sodium

sulfate, which is found in the filtrate.
I
The sulfate is determined in the filtrate by precipitation as BaS04. Calculate the BaS04 to PbS04. Determine the total lead hs above outlined, deduct
the lead found as PbS0 4 and calculate the residuall~ad to PbO.
The foregoing method is the one generally used in the commercial estimation
of lead and yields excellent results to the analyst who is familiar with it. How, ever, in laboratories where only occasional lead determinations 'are made, the
well-known gravimetric methods for lead and zinc will sometimes be found
preferable. The time required for ,gravimetric determinations is not much
greater and the chance of error is reduced.
.~
The method referred to consists in weighing the lead as chromate 6 and the
precipitation of the zinc from the filtrate with sodium carbonate, igniting it,
and weighing as zinc oxide.,
Super Sublimed White Lead

Super Sublimed White Lead is an extremely white basic sulfate of lead

pigment. Chemically, it contains no zinc oxide, being a true basic sulfate of
lead showing approximately 25% lead oxide, the balance being lead sulfate.
Its physical properties are similar to subiimed white lead, its distinguishing
characteristics being its extreme whiteness.
The analysis of this pigment is an exceedingly simple procedure. If it is
desired to determine the sulfate present, the method as outlined under sublimed white lead can be used. All the sulfate is calculated as lead sulfate after
determination of total lead and sulfate, the difference between the lead present
as lead sulfate and the total lead present being calculated as lead oxide.
The analysis of this pigment can be further simplified by a simple determination of the total lead present and the calculation followed for the determination of the lead sulfate and' lead oxide as outlined under sublimed white
lead. In this latter calculation, however, it must/be remembered that the
total sum of the lead constituents present represents 100%, instead of the
difference between 99.70% and the percentage of zinc oxide present in sublimed
white lead. This is due to the fact that super sublimed white lead shows the
presence of no zinc oxide or insoluble residue. By the simple substitution' of
100.00 in place of the determined per cent Pb constituellts in the formula and
the determination of the total lead present, the analysis calf be..completed in
a rapid way.
6

Ai;

outlined under Basic Carbonate of Lead.

ANALYSIS OF PAINTS

1859

BASIC CARBONATE OF LEAD

Basic Carbonate White Lead.7 Old Dutch Process White Lead.
Process White Lead

Quick

Basic carbonate white lead contains approximately 80% metallic lead and
20% carbonic acid and combined water with ~races sometimes, of silver, antimony and other metals. The formula for this compound is 2PbCOsPb(OH)2.
Total Lead (Gravimetric)
Dissolve 1 gram in 20 m!. of HNO a (1 : 1) in a covered beaker, heating till
all CO 2 is expelled; wash off cover, add 20 m!. of H 2S0 4 (1 : 1) and evaporate
to fumes of SOa, cool, add about 150 m!. of water and 150 m!. of ethyl alcohol;
let stand in cold water one hour, filter on a Gooch crucible, wash with 95%
ethyl alcohol, dry at 110 C., and weigh the PbS0 4. Calculate to PbO or to
basic carbonate. 8 Instead of determining the lead as sulfate, the sample may
be dissolved by boiling with acetic acid; then dilute to about 200 m!. with
water, make alkaline with NH 40H, then aci~, with acetic acid, heat to boiling
and add 10 to 15 ml. of a 10% solution of potassium dichromate; heat till the
yellow precipitate assumes an orange color. Let settle and filter on a Gooch
crucible, washing by decantation with hot water till the washings are colorless,
finally transferring all of the precipitate. Then wash with 95% ethyl alcohol
and then ether; dry at 110 C" and weigh PbCr04' (Any insoluble matter
should be filtered out before precipitating the lead.)
Total Lead (Volumetric)
Dissolve 0.5 gram of sample in 10 ml. of concentrated hydrochloric acid,
boil till solution is effected, cool, dilute to 40 ml. and neutralize with ammonium
hydroxide. Add acetic acid until distinctly acid, dilute to 200 m!. with hot
water, boil and titrate with ammonium molybdate as follows:
Dissolve 4.25 gram of ammonium molybdate in water and make up to one
liter. To standardize this solution, dissolve about 0.2 gram of pure lead foil
in nitric acid (pure PbO or PbS0 4 may also 'be used), evaporate nearly to dryness, add 30 ml. of water, then 5 ml. H 2S0 4 (sp.gr. 1.84), cool, and filter .
. Drop filter with PbS0 4 into a flask, add 10 ml. concentrated HCI, boil till
completely disintegrated, add 15 ml. of HOI, 25 ml. of water, and NH 40H till
alkaline. Acidify with acetic acid, dilute to 200 ml. with hot water and boil.
Titrate, using an outside indicator of one part of tannic acid in 300 parts of
~~

It should be noted that when calcium is present, it forms a more or less

insoluble molYbdate, and results are apt to be high. With samples containing
less than 10% of lead, the lead should be precipitated as PbS0 4, filtered,
redissolved and titrated as in the process of standardizing.
Routine Analysis of Whlte Pigments. American Society for Testing Materials.
This method of weighing lead sulfate is not accurate in the presence of calcium
compounds.
7

ANALYSIS OF PAINTS

1860

Carbon Dioxide
.

Determine by evolution with dilute hydrochloric acid, absorbing the carbon

dioxide in soda-lime or KOH solution. Calculate Cq2 to PbCO a, subtract PbO
equivalent from total PbO and calculate residual Pb? to Pb(OHh
Acetic Acid 9
Place 18 grams of the pigment in a 500-ml. flask, add 40 ml. of sirupy phosphoric acid, 18 grams of zinc dust and 50 ml. of water. Connect to a straight
Liebig condenser, apply heat and distill down to a small bulk. Then pass steam
into the flask until it becomes about half full of cond~nsed water, shut' off the
steam and distill down to a small bulk-this operation being conducteJ twice.
To the total distillate which was collected in a larger flask add 1 ml. of sirupy
phosphoric acid, connect to a Liebig condenser, using a spray trap, and distill
to a small volume-about 20 ml. Pass steam through till about 20;) ml. of
water condenses in the distillation flask, shut off steam and continue the distillation. These operations of direct and steam distillations are cdnducted
until 10 ml. of the distillate require only 1 drop of 0.1 N alkali to give a change
in the presence of phenolphthalein. Then titrate the total di,stillate with'
0.1 N sodium hydroxide and phenolphthalein and calculate the total acidity
as acetic acid. It will be found convenient to titrate each 200-ml. portion of
the distillate as collected.
Metallic Lead
Weigh 50 grams of the sample into a 400-ml. beaker, add a little water and
add slowly 60 ml. of 40% acetic acid and after effervescence has ceased, boil
on hot plate. Fill the beaker with water, let settle, and decant the Clear
solution. To the residue add 100 ml. of a mixture of 360 ml. of conc. NH 4 0H,
1080 ml. of water, 2160 ml. of 80% acetic acid, and boil until all solution is
complete. Fill the beaker with water, let settle and decant the clear solution.
Collect residue on 'a watch-glass, floating off everYthing but metallic lead.
Dry and weigh. Result X2 = percentage of metallic lead in sample.
The following method of A. N. Finn (unpublished) gives total basicity of
a pure white lead: Place 2 grams of pigment in an evolution flask, add a little
CO 2-free water, connect with a separatory funnel and condenser (Knorr type),
add through the funnel, finally washing down, 100 Illl. of N/4 nitric acid,
boil and absorb the CO 2 in a soda lime tube in the usual manner, (having H 2SQ4
and CaCl 2 drying tubes in train) and weigh. To the solution in 'the evolution
flask, add about 20 ml. of neutral sodium sulfate solution and titrate with
N/4 sodium hydroxide solution (carbonate-free), using phenolphthalein. CO 2
is calculated to PbCO a The amount of N/4' acid corresponding to the CO 2
is calculated and deducted from the total amount of N/4 acid neutralized by
the sample and the difference calculated to combin~d H 20, from ~hich Pb(OH)2
is computed.
9 Thompson's Method, Jour. Soc. Chern. Ind., 24, 487 (1905).
I

ANALYSIS OF PAINTS

1861

ZINC LEAD AND LEADED ZINC

Zinc lead and leaded zinc are varying compounds containing zinc oxide
and leJ1d sulfate, the former showing approximately 50% zinc oxide and 50%
lead sulfate, while the latter contains on an average 25% lead sulfate and
75% zinc oxid,e. The most widely used type is probably that with 35% basic
lead sulfate. See also pp. 509, 1060, Vol. 1.
These pigments may be analyzed by the following procedure:
Moisture.-Heat 2 grams at 1050 C. for two hours. Cool and weigh.
Lead and Zinc.-Determine the lead directly by the method given under
Basic Carbonate of Lead, and the zinc by the volumetric ferro cyanide method
as outlined under Sublimed White Lead. See also pages 509 and 1060, Vol. I.
Total Soluble Sulfates 10 (in the absence of BaS04).-Treat 0.5 gram of
the sample with 5 m!. of water, 3 grams of NH 4 CI and 5 m!. of HCI saturated
with bromine; digest (covered) on the steam bath about fifteen minutes, add
25 ml. of H 20, neutralize with dry Na 2CO a and add about 2 grams more. Boil
ten to fifteen minutes; let settle, dilute with hot water, filter and wash with
hot water; redissolve in HCl, reprecipitate as above and wash thoroughly with
hot water. Acidify the united filtrates with HCl and add a slight excess of
10% BaCl 2 solution. Let stand on steam bath for one hour, filter, wash with
hot water, ignite and weigh the BaS04. Calculate to SOa (includes 80 a formed
from S02).
Total Soluble Sulfate (in the presence of Ba804).-Treat 1 gram in a
600-m!. beaker with 10 ml. of H 20, 10 ml. of strong HCI, saturated with bromine, and 5 grams of NH 4CI, heat on a steam bath in a covered beaker for five
minutes, add hot water to make about 400 ml., boil for five minutes and filter
to separate any insoluble material. (A pure pigment should be completely
dissolved.) Wash with hot water; ignite and weigh the insoluble. Remove
lead with 'N a 2CO a as above, making a double precipitation, acidify, and to the
boiling hot filtrate add slowly, with stirring, 20 m!. of a 10% BaCl 2 solution;
let stand for two hours on the steam bath, filter, wash, ignite, and weigh as
Ba80 4. (Includes 80 a formed from 80 2 .)
Soluble Zinc Sulfate.-Heat nearly to boiling 2 grams of the sample with
150 m!. of water and 50 ml. of alcohol for thirty minutes, filter and wash with
a mixture of alcohol and water (1 : 3). Heat the filtrate to boiling and expel
most of the alcorol; then determine SOa by the usual method of precipitation
with BaCl 2. Calculate to Zn80 4 and to 80 a.
Sulfur Dioxide.-Digest 2 grams of the sample with frequent stirring in
100 ml. of freshly boiled cold water and 5 ml. of concentrated HCl; let stand
ten to fifteen minutes, add an excess of 0.01 normal iodine solution and titrate
back with 0.01 normal sodium thiosulfate solution, using starch indicator.
Report as 80 2 Run blank on reagents and make corrections.
Calculations.-Report soluble 80 a as ZnS04. Deduct ZnO equivalent of
. ,the Zn80 4 from total ZnO and report residue as ZnO. Deduct soluble 80 a
10 Report of 'Sub-committee VIII of Committee D-l, Proceedings of American
Society for Testing Materials, 14, 271-2, 1914.

1862

ANALYSIS OF PAINTS

and 80 3 equivalent to 80 2 from total 80 a, calculate remainder to Pb80 4 ; subtract PbO equivalent of Pb80 4 from total .PbO and report
remainder as PbO.
I

ZINC OXIDE
Moisture.-Heat 10 grams on watch-glass for two hours at 105 to 1.10 C.
Cool and weigh.
Zinc Oxide.-Using a 0.25 g. sample, determine zinc as outline(l under
Basic Sulfate of Lead.
Total Soluble Sulfur....._Moisten 10 g. of the pigment with alcohol, add a
few drops of Br2, or about 5 ml. of Br-water, then concentrated HCI in-excess;
boil to expel the Br2; filter off any insoluble matter, and wash the paper with
hot water.
Nearly neutralize the clear solution in a covered beaker with NaOH solution, complete the neutralization with dry ,N a 2CO a and add about 2 g. more
of this reagent. Boil 10 to 15 minutes, wash off cover, let settle, filter, and ,
wash with hot water. Redissolve the precipitate in HCI (1 : 1), reprecipitate
with Na 2 CO a as above, filter, and wash thoroughly witll hot water. Acidify
the united 'filtrates with HCl, adding/about 1 ml. in excess. Boil to expel
bromine and to the clear boiling solution add slowly with stirring an excess of
a 10% BaC1 2 solution. Let stand on a steam bath for at least 1 hour, filter on
a weighed Gooch crucible, wash thoroughly with boiling water, dry, ignite at
a dull red heat, cool, and weigh as Ba80 4 This will include soluble sulfates,
SOa formed from 80 2 and the 80 a that is formed from sulfide sulfur.
f?ulfur Dioxide (Rigg Method).-Make a suspension of 5 g. of the sample
in {fo ml. warm water (freshly boiled and then partly cooled), pour into a glass
stoppered iodine determination bottle containing 1,8 ml. of HCI and exactly
25 ml. of 0.1 N iodine solution, and shake ,until :all the oxide is dissolved.
Titrate the excess of 12 as rapidly as possible with 0.1 N thiosulfate solution.
If the thiosulfate solution is not exactly 0.1 N, a Dlank should be run.
Zinc Sulfate.-Proceed as under Zinc Lead and Leaded Zinc.

LITHOPONE
/

Lithopone, in the restricted' sense, is a mixture of equimolar quantities of

zinc sulfide and barium sulfate. The commercial pigment contains approximately 28% of the former ingredient and 72% of the latter. Lithopone is
produced in several modifications, the principal ones being those in which the
extending portion is calcium .sulfate or magnesigm silicate. ~High strength
litho pone contains from 45 to 60% of zinc sulfitle.

ANALYSIS OF PAINTS

1863

Moistu.e.-Heat 2 grams'f~r two hours at 1050 C.

Barium Sulfate.-Treat 1 gram with 10 ml. conc. HCI and 1 gram of
potassium chlorate, added in small amounts. Evaporate to one-half its volume, add 100 ml. hot water and a few ml. of dilute H 2S0 4 Boil, filter, wash
and weigh the insoluble residue, which should show only the presence of barium
sulfate. Examine the residue for silica and alumina.
Zinc Sulfide.-Determine the zinc in the pigment by the method given
under Sublimed White Lead. Calculate to zinc sulfide. From this value
subtract the zinc sulfide equivalent of the zinc oxide as determined below.
Zinc Oxide.-Digest 1 gram of the pigment with 100 ml. of 3% acetic acid
solution for 18 hours. Stir several times during digestion, especially during
the first hour. Filter, wash with 3% acetic acid and determine the zinc by
the usual method.
Soluble Salts.-Digest 2 grams with hot water and examine the filtrate for
soluble salts.

TITANWM PIGMENTS
The titanium pigments used in paints are: titanium dioxide, barium base
pigment, which is a mixture of 25 parts of titanium dioxide and 75 parts of
barium sulfate, calcium base pigment, which is a mixture of 30 parts of titanium
dioxide and 70 parts of calcium sulfate, titanated lithopone, which is a mixture
of 85 parts of lithopone and 15 parts of titanium dioxide, and lead titanate.
A mixture of titanium dioxide and magnesium silicate is sometimes found.
Barium Suifate.-Weigh Y2 gram sample into 250-ml. Pyrex glass beaker;
add 20 ml. concentrated sulfuric acid and 7 or 8 grams ammonium sulfate. Mix
well and heat on hot plate until fumes of sulfuric anhydride are evolved and
then heat directly ov~r flame to boiling for five minutes or until solution is
complete. Traces of silica, if any, remain as an insoluble residue.
Cool, take up with 100 ml. of water, boil and filter off barium sulfate and
silica, washing with 5% sulfuric acid to free residue from titanium. Use the
filtrate for the determination of titanium dioxide.
Calcium Sulfate.-Extract the calcium sulfate QY boiling 0.5 gram of the
pigment in a 250-ml. beaker with 25 ml. of hydrochloric acid (1 : 1) for several
minutes. Dilute to 150 m!. with water, boil and filter. Wash the residue
with hot hydrochloric acid solution (10 m!. of HCI sp.gr. 1.19 and 990 ml. of
water) until the washings show no. calcium, when tested with NH 40H and
(NH 4 )2C 20 4 Determine the calcium in the filtrate by the oxalate method.
. Titanium Dioxide.-Determine the titanium in the filtrate from the barium
sulfate reducing it in a Jones or similar reductor and titrating the reduced
solution with a standludized ferric ammonium sulfate or potassium permanganate solution.

1864

ANALYSIS OF PAIlITrS

Lead Titanate.*-If lead titanate is to be determined, it will be necessary

first to separate any other compounds of lead that may be present. If separate
determinations of the basic carbonate and basic sulfate of lead are desired,
first extract the former with 10% acetic acid solutidn, and then the latter with
40% neutral ammonium acetate solution; otherwise, extra~t with acid ammonium acetate solution as directed under Basic Sulfate of Lead. The lead
oxide in leaded zinc, if present, will be in the acetic acid extr8~ction. The
residue, after extraction, is dried and treated with 25 ml. sulfuric acid "and 10
grams ammonium sulfate for each 0.5 grams of residue as directed in the first
paragraph under Barium Sulfate. The solution is diluted to 200 ml~ and allowed to stand overnight. If barium sulfate is absent, and if no titanium dioxide has been precipitated, the lead sulfate may be filtered, dried and weighed
in the usual manner. It is, however, recommended that the precipitate be
redissolved in ammonium acetate solution and the lead determined as lead
chromate. Calculate the lead as PbTi0 3.
This method will account for approximately 92% of the pigment. The
remainder is uncombined sulfate, which is not distinguishable from other lead
sulfate that may be present.

ANTIMONY OXIDE

Antimony trioxide.,(Sb 20 3), a fume pigment that has recently been introduced, may be occasionally found in certain classes of mixed paints. After
being brought into solution, it may be quantitativ~ly estimated by oxidation
with permanganate or iodine.
~

ZIRCONIUM 'PIG~ENTS 11

The usual zirconium pigments, in addition ~o considerable Si0 2 usually

contain. less than 0.2% -each of Ti0 2 and Fe203. The follm'ifing Simplified
method, as modified by Mr. C. J. Kinzie, is based on the procedure of Lundell
and Knowles.
.
.
Fuse the sample, or the insoluble matter, with Na I2C0 3 in a . platinuni

* Based on method ,of D. W. Robertson, Private communication, 1937.

Physical and Chemical Examination of Paintg;-~arnishes, Lacquers and Colors.

H. A. Gardner.
.
11

ANALYSIS OF PAINTS

1865

crucible, extract the fusion with~30 m!. of H 280 4 (1 : 1) and evaporate to

strong 80 3 fumes. If. considerable residue adheres to the beaker, remove
from the heat at the first signs of 80 3 and scrape the residue from walls of
the beaker with a glass rod and wash down with H 2S0 4 (1 : 1) from a wash
bottle, then continue the heating to strong fumes as directed. Filter and treat
the residue with H 280 4 and HF. Examine the residue for un decomposed pigment and, if necessary, make a second fusion with Na 2 CO a
Treat the filtrate from the Si0 2 with NH 4 0H and filter. Reject the filtrate
and dissolve the precipitate in H 2S0 4 (1 : 1).
Adjust,the acidity of the solution to about 1% (by volume) of H 2S0 4 and
pass in H 2S to remove the heavy metals. Filter. If desired, analyze the H;S
'
precipitate by the regular methods.
Adjust' .the acidity of the filtrate from the H 2S treatment to 10% (by
volume) of H 2S0 4 , in 400 ml. of solution. Boil to remove H 2S. Cool to
between 5 and 14 C.
Add an excess of cold 6% aqueous solution of cupferron to precipitate the
zirconium, titanium and iron. An excess of the precipitant is indicated by the
formation of a fine white precipitate which redissolves instead of a curdy one,
which persists. The color of the cupferron precipitate is an excellent indication of the amount of titanium present, since zirconium precipitates are pure
'white while the titanium precipitates are yellow. A brownish tint indicates
iron or, less likely, vanadium. Filter, after a five-minute digestion, using an
ashless filter paper and a platinum cone, with gentle suction. Thoroughly
wash the precipitate with cold 10% (by volume) HCI.
Clear filtrate and washings .indicate complete precipitation of zirconium
and titanium, while cloudy or opalescent ones indicate the reverse. The latter
occasionally develops, and nO,satisfactory explanation has been found. Under
these circumstances, recovery must be made as follows: Evaporate the solution
to a small votume and destroy the organic matter by digestion with HNO s,
finally heating until fumes of H 2S0 4 appear. Cool, dilute, filter off any Si0 2
resulting from attack on glassware, and repeat the precipitation with cupferron
in a 10% H 2S0 4 solution as directed above.
Carefully dry and ignite the paper and precipitate in a tared platinum
crucible and finally ignite over a blast lamp or its ,equivalent. Cool in 1\
desiccator, and weigh as mixed oxides of Zr, Ti, Fe and rare earth. Fuse the'
mixed oxides with potassium pyrosulfate, take up in 10% (by volume) of
H 2S0 4 , and determine the iron colorimetrically.
Titanium should be determined on an original sample.
Aluminum is determined in the filtrate from the cupferron treatment.
Destroy organic matter by heating with small quantities of HNO a, finally
heating to fumes of S03. Dilute, fiiter off any silica dissolved from the beaker
and determine aluminum in the filtrate.

1866

ANALYSIS OF PAINTS
SILICA OR SILEX-CHINA

CLAY-A:~BESTINE

These pig,ments, while all true :;ilica, pigments, lare widely different in
physical structure. A microscopic examination is of great value.
The particles of quartz or silex are sharp and irregular, those of china clay
are flat, and those of asbestine are rodlike. Diatom~ceous earth may sometimes be recognized in siliceous pigments. The color, fineness, and other such
properties are more important than chemical analysis. However, if desired,
determinations of silica alumina, magnesia, etc'., may be made by recognized
~
methods.
Moisture.-Heat 2 grams at 1050 for two hours.

WHITING (PARIS WHITE), GYPSUM (PLASTER OF PARIS)

These pigments are o~ the following composition:
Whiting.-The natural form of calcium carbonate.
Paris White.-The artificial form of calcium carbonate.
Gypsum.-The hydrated form of calcium sulfate, of formula CaS042H 20.
These pigments are analyzed in the following manner:
Moisture.-Heat 2 grams at 1050 C. for two hours.
Loss on Ignition.-Ignite 1 gram at a high heat to constant weight. The
loss will be water, if carbonates are absent.
Calcium.-Treat 1 gram with dilute HCl and a few drops of HN0 3 Evaporate to dryness, dehydrate, moisten with a few drops of concentrated HCI,
dilute with hot water and deternrine the insoluble residue. Examine for
BaS04. The residue should consist qf silica.
I
. In the filtrate, precipitate and determine the iron/hydroxide and aluminum
hydroxide in the usual manner." The calcium is precipitated in the boiling
ammoniacal filtrate with 30 ml. of saturated ammonium oxalate solution,
allowing the solution to boil for one-half hour. A double precipitation is here
advisable to remove the last traces of magnesium. The calcium oxalate i~
filtered off, thoroughly w~shed and determined volumetricaJly by the perman~
ganate method, p. 211.
Magnesium.-Determine in the filtrates after removal of the calcium by
precipitation as magnesium ammonium phosphate and ignition to magnesium
pyrophosphate in the usual manner.
Carbon Dioxide.-Determine by the evolution method.
Sulfates.-Dissolve 1 gram in concentrated HCl, remove any insoluble
residue, heat to boiling and precipitate any sulfate as BaS04, determining in
the usual manner.

ANALYSIS OF PAINTS

1867

BARYTES AND BLANC FIXE

Of these two barium pigments used in the manufacture of paints, barytes
is the natural barium sulfate, while blanc fixe is precipitated barium sulfate.
Their barium s'ulfate content should be not less than 95%.
The following method may be used for the analysis of these pigments:
Moisture~-Heat 2 grams at 105 C. for two hours.
Loss on' Ignition.-Ignite 1 gram to constant weight. The loss will be
reported as loss on ignition, and will consist of free and -uncombined water,
carbon dioxide and organic matter.
Barium Sulfate.-Boil 1 gram with dilute HCl, evaporate to dryness,
moisten with HCl, add water, boil, filter and wash. Should lead be present
in the insoluble residue, as shown ty the action of H 2S, treat the insoluble
residue with a little (1 : 1) HCl and several drops of H 2S0 4 Filter, wash and
weigh the residue. Treat the ignited residue with H 2S0 4 and HF, evaporate
to dryness and ignite. The residue should show no loss as silica. The filtrate
is examined for alumina, iron, calcium and magn~sium in the usual manner.
Soluble Sulfates.-Treat 1 gram with 20 ml. conc. HCl, dilute to 200 ml.
with hot water, boil, filter, wash, add NH 4 0H until neutral, make acid with
HCl and precipitate any sulfate as BaS04. Determine in the usual manner.
Calculate to CaS04. If carbonates are present, calculate the remaining CaO
to CaCO a Any excess of oxide is reported as CaO.
Carbon Dioxide.-Determine in -the usual way. If any barium carbonate
is present, it is determined in the filtrate from the preliminary HCl treatment,
by precipitation and weighing, as BaS04. Any excess of carbon dioxide over
the barium is reported as calcium carbonate.

ANALYSIS. OF A COMPOSITE WHITE PAINT

A white paint may consist of a mixture of any of the preceding pigments,
exceptink that lead pigments and lithopone are seldom found together, owing
to their tendency to blacken with the formation of lead sulfide.
After separation from the oil and other liquids as outlined above, the white
pigment mixture may be rapidly analyzed by the following method. It is,
however, often advisable to resort to a qualitative examination .before beginning the quantitative analysis.
Insoluble Residue.-Boil 1 gram of the sample with 20 mi. (1 : 1) HCI.
Evaporate to dryness, moisten the residue with a few mi. of concentrated HCl,
allow to stand a few minutes, dilute with hot water, boil, filter and wash the
insoluble residue thoroughly with hot water. Treat the insoluble residue with
(1 : 1) HCl and 2 mi. H 2S0 4 to remove'the last traces of lead. Filter, wash
a!ld weigh the insoluble residue. Determine the silica by volatilization with

1868

ANALYSIS OF PAINTS

H 2S0 4 and HF. Any loss is reported as silica. Determine the BaS04 in the
residue by boiling with dilute Hel or by making a potassium bisulfate fusion.
The residue remaining after either of these treatments is reported as barium
sulfate.
\

Total Lead.-This constituent can best be determined on a separate sample.

To 1 gram add 10 ml. of conc. HNO a, boil, add, after,cooling, conc. H 2S0 4 and
evaporate to strong SOa fumes. Dilute with water, allow to stand several
hours, filter, wash slightly, dissolve and determine the lead volumetrically as
outlined under Sublimed White Lead.
Lead can also be determined on the combined filtrates from the insoluble
residue. Precipitate the lead in an acid solution with H 2S and deter-mine as
directed under Basic Carbonate of Lead. Reserve the filtrate for the determination of iron, aluminum, zinc, calcium and magnesium.
To determine whether both sublimed white lead and corroded white lead
are present, treat a separate portion of the paint witli boiling acetic acid, filter
and collect the insoluble residue. Determine the lead either in the filtrate or
in the insoluble residue by the volumetric method. The lead soluble in acetic
acid is the basic carbonate of lead and the lead oxide from the sublimed white
lead, while the lead sulfaj:; from the sublimed white lead remains insoluble.
Alumina and Iron Oxide.-Remove the,H 2S from the filtrate from the lead
sulfide, above, by boiling, oxidize the iron with a few drops of HNO a, and
precipitate the hydroxides with ammonia. Separate and determine the iron,
and aluminum in the usual manner.
.-'
Zinc.-Acidify the filtrate from the alumina and iron precipitation with
acetic acid, separate the zinc with H 2S and determine as outlined under Sublimed White Lead.
Calcium and Magnesium.-Determine the calcium and magnesium in the
filtrate from the precipitation of zinc sulfide in the usual manner, first testing,
however, for the presence of barium.
Sulfate.-Determine as outlined under Zinc Lead and Leaded Zincs.
Sulfide.-Should lithopone be present, separate/.the zinc oxide a'nd zinc
sulfide as outlined under Lithopone.
Carbon Dioxide.-Determine as outlined under Basic Carbonate of Lead.
Calculations.-Silica is reported as silica, except where alumina is present,
showing the presence of China clay. In this case, calculate the alumina to clay
by the following method.
'
Weight of A1 20 aX 2.5372 = weight of clay.
Weight of clayX0.4667 = weight of Si0 2 in clay.
Any difference greater than 5% may be considered silica.
Barium sulfate is reported as barium sulfate or as lithopone, if zinc sulfide
is present, according to the given composition 'of lithopone, 70% barium sulfate and 30% zinc sulfide.
Lead is reported as Basic Carbonate of Lead on the formula 2PbCO a
Pb(OH)2,
.
Calculate lead soluble in acetic acid, after determining CO 2, to basic lead
carb'onate and any residual lead to lead oxide which, together with the lead
sulfate is reported as Sublimed White Lead.
I
Should calcium sulfate be present; the portiQn...sollible in water is examined
for lime or sulfuric acid and calculated to calcium sulfate, any residual lime

ANALYSIS OF PAINTS

1869

being calculated to calcium carbonate and any residual sulfuric acid being
calculated to lead sulfate.' Any residual CO 2 after calculating ciJJcium carbonate is calculated to white lead and any residual lead is calculated to lead oxide.
Lead oxide should not be reported except in the presence of lead sulfate.
Any large percentage of magnesium indicates the presence of asbestine.

RED AND BROWN PIGMENTS

These pigments are grouped under these heads:
The Lead Oxide Pigments-The Iron Oxide and Manganese Oxide
Pigments-The Mercury Oxide Pigments

RED LEAD AND ORANGE MINERAL

These pigments in the pure form are oxides of lead, of the generally accepted form, Pb 30 4, being probably mixtures of lead monoxide, and lead
dioxide.
Two methods are given for the analysis of this pigment.
Moisture.-Dry 2 grams at 105 for two hours.
Organic Color.-Boil 2 grams with 25 m!. of 95% ethyl alcohol, let settle,
decant off the supernatant liquid; boil residue with water, decant as before and
boil residue with very dilute NH 40H. The pigment should also be extracted
with chloroform. If either the alcohol, chloroform, water or NH 40H is colored, organic coloring matter is indicated.
. Total Lead and Insoluble Residue.-Treat 1 gram with 15 mL of RNO;
(1 : 1) and sufficient hydrogen dioxide to dissolve all the Pb0 2 on warming.
If any insoluble matter is present, add 25 m!. of water, boil, filter and wash
with hot water. Insoluble contains free 8i0 2, and should be examined for
BaS04 and silicates, if appreciable. To the original solution or filtrate from
insoluble, add 20 m!. of conc. ~2S04 and evaporate to 80 3 fumes; cool and
determine lead as lead sulfate either gravimetrically or volumetrically. If the
sample contains soluble barium salts, the Pb80 4 will contain Ba80 4 and should
be treated with acid-ammonium acetate solution, the lead being determined
in the filtrate.

1870

ANALYSIS OF PAINTS
!

Determination of Lead Peroxide (P1102) and True Red Lead (Pb S0 4 ) . (Method of Diehl,12 modified by Topf l3-not appliPable when substances are
present, other than oxides of lead, that liberate iodine under conditions given.)
Weigh 1 gram of finely ground sample into a ;2DO-ml. Erlenmeyer flask,
add a few drops of distilled water and rub the mixt~re to a smooth paste with
a glass rod flattened on end. Mix in a small beaker \30 grams of C.P. " Tested
Purity" crystallized sodium acetate, 2.4 grams of C.P. potassium iodide, 10
ml. of water and 10 ml. of 50% acetic acid; stir until all is liquid, warming
gently; if necessary add 2 to 3 ml. of H 20, cool to room temperature and'pour
into the flask containing the red lead. Rub with the glass rod until.nearly all
the red lead has been dissolved; add 30 ml. of water containing 5 or 6 grams of
sodium acetate, and titrate at once with decinormal sodium thiosulfate, adding
the latter rather slowly and keeping the liquid constantly in motion by whirling
the flask. When the solution has become light yellow, rub any undissolved
particles up with the rod until free iodine no longer forms, wash off rod, add
the sodium thiosulfate solution until pale yellow, add starch solution and titrate
until colorless, add decinormal iodine solution until blue color is just restored
and subtract the amount used from the volume of thiosulfate that had been
added.
'
Calculation.-The iodine value of the sodium thiosulfate solution multiplied
by 0.942 = Pb0 2 ; the iodine value multiplied by 2.700 = Pb a0 4 ; the Pb0 2 value
multiplied by 2.866=Pb a0 4
,
The red lead may also be examined for zinc, carbon dioxide, and soluble
sulfate.
.
'
A second method for determination of the lead peroxide or true red lead
content is somewhat shorter,u
Treat 1 gram in a beaker with 15 ml. of nitric acid, sp.gr. 1.2 (110 ml. ,nitric
acid, sp.gr. 1.42 to 100 mi. of water). Stir the sample until all trace of red
color has disappeared. Add from a calibrated pipette or burette exactly 10 ml.
of dilute hydrogen dioxide (1 part of 3% hydroge~ dioxide to 3.5 parts of
water). Add about 50 mi. of hot water and stir un'til all the lead dioxide has
passed into solution. In the case of some coarsely/ground oxides the contents
of the beaker may have to be gently heated td effect complete solution. After
the oxide has completely passed into solution, dilute with hot water to about
250 ml. volume and titrate directly with a st!lndard potassium permanganate
solution, having an iron value of 0.005. Titrate to theJaint pink permanganate color. A blank titration on the hydrogen dioxide~ solution must now
be made.
_
Into a 'beaker pour 15 ml. of nitric acid of above strength and add exactly
the same amount of hydrogen dioxide (10 mI.). Dilute to 250 ml. with hot
water and titrate with standard potassium perinanganate solution to a faint
pink color.
I
The difference between the number of ml. of 'potassium permanganate required for the blank titration and the number required for the red lead titration
is the amount required for the hydrogen dioxide which was ,reacted on by the
Dingl. Polyt. Jour., 246, 196.
I
Zeitschrift fUr analytische Chemie, 26, 296.
,::.- ,
a" Analysis of Lead and Its Compounds," SChaeNer-and White and Calbeck, p. 77.

12

13

ANALYsis OF PAINTS

1871

red lead. The difference between the two amounts of potassium permanganate
required multiplied by 3.058 grams gives the percentage of red lead present.
The difference multiplied by 1.067 gives the percentage of Pb0 2 present.

VERMILION
The following portion of Walker's 15 method will suffice for the examination
of this pigment. Should the analyst desire to determine the sulfide of mercury
presep.t or make a more complete examination-reference may be made to the
original method.
True vermilion, or, as it is generally called, English vermilion, is sulfide
of mercury. On account of its cost it is rarely used in paints, and is liable to
gross adulteration. It should show no bleeding on boiling with alcohol and
water and no free sulfur by extraction with carbon disulfide. A small quantity
mixed with five or six times its weight of dry sodium carbonate and heated in
a tube should show globules of mercury on the cooler portion of the tube. The
best test for purity is the ash, which should be not more than one-half of 1%.
Makll the deterniination in a porcelain dish or crucible, using 2 grams of the
sample. Ash in a muffle or in a hood with a. very good draft, as the mercury
fumes are very poisonous. It is seldom necessary to make a determination of
the mercury.
.
Genuine vermilion is at the present time little used in paints. Organic
lakes are used for most of the brilliant red, scarlet and vermilion shades. These
organic coloring matters are sometimes precipitated on red lead, orange mineral
or zinc oxide; but as a usual thing the base is barytes, whiting or China clay.
Paranitraniline red, a compound of diazotized paranitraniline and beta-naphthol, is largely employed; but a number of colors may be used.
Paranitraniline red is soluble in chloroform. It is also well to try the solvent
action on different reds, of sodium carbonate, etc. The amount of organic
pigment present in such reds is generally very small, and when it cannot be
determined by ignition owing to the presence of lead, zinc or carbonate, it is
best determined by difference.
15 P. H. Walker, Bulletin 109, Revised, Bureau of Chemistry, U. S. Dept, of Agri.,
pp.31-33.

1872

ANALYSIS OF PAINTS
IRON OXIDES

The iron oxides and manganese oxide pigments include the ochres, umbers,
siennas, Venetian red, metallic brown, Indian red aJid Tuscan red.
In analyzing these pigments, the following cons'tituents are sought; moisture, loss on ignition, insoluble residue, iron oxide, xrtanganese dioxide, calcium
and magnesium oxides and sulfur trioxide.
Owing to the similarity of the methods used for the analysis of these pigments to those used in the analysis of iron ores, the analyst is referred to p. 469
on the Analysis of Iron Ores, or to the method of Walker.l6
.

BLUE ,PIGMENTS
The principal blue pigments used in paints are Prussian, ultramarine, blue
lead, and phthalocyanine.
Prussian blue is the double iron and-potassium salt of hydroferrocyanic and
hydroferricyanic acids.
.
Ultramarine blue is essentially a silicate and sulfide of sodium and aluminum.
Sublimed blue lead is the fume product resulting from the smelting of'lead
ores. In composition it consists ,of lead sulfate, lead sulfide, lead sulfite, lead
oxide and zinc oxide, with occasional traces of carbon. It is finding its greatest use as an inhibitive pigment for the protection of iron and steel. Its color
is a bluish gray.
Phthalocyanine blue is a copper compound of pp,thalocyanine and may be
i
identified by the'presence of' copper.

ULTRAMARINE BLUE
Moisture.-Heat 2 grams at 10fjo C. for two hours.
Silica.-Digest 1 gram with 30 ml. of conceptrated HCl, taking care to
avoid spattering. Evaporate to dryness, dehydrate, moisten with cone. HCI,
dehydrate a second time, dilute, filter, and determine the silica by volatilization
with H 2 SQ4 and HF.
Aluminum Oxide.-In the filtrate from the silica, precipitate the aluminum
hydroxide and determine in the usual manner. ~eport as aluminum oxide.
16

-,-'
Bulletin 109, Revised, Bureau of Chemistry, U. S~ Dept. Agri., pp. 33-34.

ANALYSIS OF PAINTS

1873

Sodium _Oxide.-The filtrate from the aluminum hydroxide is acidified

with H 2S0 4 Evaporate to dryness, ignite at a low red heat, and weigh the
sodium sulfate. Calculate to sodium oxide.
Total Sulfur.-Fuse 1 gram with a mixture of KN0 3 and Na 2CO a Dissolve the fused mass in HCI, boil with conc. HN0 3 for one-half hour, remove
the insoluble residue and determine the sulfuric acid in the usual way.
Sulfur Present as SuIfate.-Dissolve 1 gram in dilute HCI and boil until
all the hydrogen sulfide is removed. Filter off the insoluble residue and determine the sulfate in the filtrate .

. PRUSSIAN BLUE-(CHINESE BLUE)-ANTWERP BLUE

Moisture.-Heat 2 grams at 1050 C. for two hours. Dry Prussian blue
should contain less than 7% moisture.
Nitrogen.-Determine the nitrogen present by the Kjeldahl-Gunning
method.
Iron and Aluminum Oxides.-Ignite 1 gram at a low temperature, sufficient
to decompose all the blue, but not to render the iron difficultly soluble. Digest
the residue with (1 : 1) HCI. Any insoluble residue is examined for silica,
barium sulfate and alumina. A pure Prussian blue should show no insoluble
residue. The. filtrate is examined for alumina, iron and calcium in the usual
way.
An aliquot portion of the filtrate after the removal of the calcium is examined for the alkaline metals. Calculate any alkaline metal present to sulfate.
Sulfuric Add.-Determine the sulfuric acid in an aliquot portion after
removal of the calcium.
Commercial Analysis.-The method of Parry and Coste 17 is sufficiently
accurate to determine the Prussian blue in most instances.
If the pigment is a soda blue, the percentage of Prussian blue may be obtained with sufficient accuracy for commercial purposes by mUltiplying the
percentage of nitrogen by 3.404 or the percentage of iron (in the absence of
other iron pigments) by 2.8. The corresponding factors for potash blue are
4.4 and 3.03 respectively. Some blues, e.g., Chinese blue, may contain tin
salts. Others may contain manganese or chromium compounds. The presence of these compounds should be determined by a qualitative examination
at least.
.
17

The Analyst, 21, 225-230, 1896.

1874

ANALYSIS OF PAINTS
I'

SUBLIMED BLUE LEAD 18

Total Lead.-The total lead content'is determined by the volumetric method
for lead as outlined under Sublimed White Lead. \

Total Sulfur.-Treat 0.5 gram with 10 ml. of water and a few ml. of bromine
water. Boil gently until all the bromine has passed off. Dilute with water,
add another portion of bromine water, boil, and continue 'the treatment until
the sediment has become white in color. Add 8 ml. of nitric acid, evaporate
until the brown fumes of nitric acid have disappeared, dilute with water and
add an excess of sodium carbonate. Determine as outlined under Zinc Lead
and Leaded Zinc.
Lead Sulfate.-On a separate sample determine the sulfate directly as
outlined under Zinc Lead and Leaded Zinc, without any preliminary treatment
for the oxidation of sulfites and sulfides.
Lead Sulfite.-Boil 1}2 grams with 3 grams of s~dium carbonate. Allow
to stand, filter and thoro'ughly wash. Treat the filtrate with bromine water
as outlined under Total Sulfur and determine the combined sulfur present as
sulfate ~nd sulfite. Deduot the amount present as sulfate and calculate to
sulfite.
Lead Sulfide.-neduct the sulfur present as sulfate and sulfite from the
total sulfur and report the difference as lead sulfide.
Lead Carbonate.-Determine any CO 2 present by the evolution method
and calculate to lead carbonate. See p. 235, Vol. I.
Lead Oxide.-Deduct the lead present as sulfate, sulfite, sulfide and carbonate from the total lead and report the difference as lead oxide.
Zinc Oxide.-Determine the zinc volumetrically as outlined under Sublimed White Lead and report as zinc oxide.
Carbon and Volatile Matter.-Ignite the sample in a partially covered
crucible at a low heat f9r two hours. Report the difference in weight as carbon
and volatile matter.
.

YELLOW AND ORANGJj: PIGMENTS

Chrome Yellows-American Vermilion/Basic Lead Chromate

The pigments under this class all contain chromates, with the exception of
orange mineral, which is analyzed as under Red Lead. Frequently they contain lead sulfate and sometimes lead carbonate. A pure chrome yellow should
contain only lead chromate and insoluble lea~ _~ompoUJlds. Owing to the
18 "The Chemical Analysis of Lead and its ,eompounds," Schaeffer and White
and Calbeck, p. 35.

ANALYSIS OF PAINTS

1875

frequent use of organic colors to brighten up the pigment, it is essential that a

test be made for organic colors as outlined under VermiJion.
The analysis of these pigments is carried out in the following manner:
Moisture.-Heat 2 grams at 1050 C. for two hours.
Insoluble Residue.-Treat 1 gram with 25 ml. of concentrated HCI, boil
and during the boiling add a few drops of alcohol, one at a time. The solution
is diluted to 100 mI., ,the boiling is continued for ten minutes and any insoluble
residue is filtered off, thoroughly washed and examined for silica, barium sulfate and alumina.
Lead.-The solution is nearly neutralized with NH40H and the lead is
precipitated as PbS with H 2S. Filter off the precipitate of PbS, dissolve in
HNO a, add H 2S0 4 , boil to strong fumes and determine as outlined under
Sublimed White Lead or weigh as PbS0 4
Chromium.-The filtrate from the lead precipitation is boiled until all the
H 2S is driven off. The solution is rendered alkaline with NH 40H and the
chromium is precipitated and determined as chromic oxide. Calculate to
chromic anhydride.
If iron or aluminum is present, it will b~ necessary to separate the chromium
by oxidizing it with Na 202 to chromate, and then titrating it with ferrous
sulfate.
.
Zinc, Calcium and Magnesiunt.-Precipitate the zinc in the filtrate with
H 2S and determine:as previously outlined.
In the filtrate frop! the zinc precipitation, determine the calcium and magnesium in the usual manner.
If any carbonates are present; determine by the evolution method.
Sulfuric Acid.-Determine the total sulfate as outlined under Zinc Lead
and Leaded Zinc on p. 1861.
Calculations.-Any chromic anhydride is calculated to lead chromate, sulfuric acid to lead sulfate, if calcium sulfate is absent, and any residual 'lead
is calculated to lead oxide.

GREEN PIGMENTS
Chrome Green
Green pigments are usually mixtures of chrome yellow and Prussian blue,
though organic color is sometimes present, which may be determined by an
extraction with alcohol or chloroform.
A microscopic examination should be made to determine whether the green
is a combined precipitation product, which is of the greater value, or one mixed
after separate precipitation. A good green will show the presence ,of green

1876

ANALYSIS OF PAINTS

and _blue particles, while a poor green will show yellow and 'blue particles mixed
with green. The analysis may be carried out as fol~ows: 19
Moisture.-Heat 2 grams at 1050 C. for two hours.
Insoluble Residue.-Heat 1 gram at a low heat until the blue color has
been decomposed, keeping the temperature sufficiently low so as not to render
~ny of the iron or lead chromate insoluble. Determine the insoluble residue
as outlined under Yellow Pigments.
Lead.-Determine as outlined under Yellow Pigments.
Iron, Alumina and Chromiu:a.-A-ll the H 2S is expelled from th~ filtrate
after the lead precipitation by boiling. Add a few drops of HN0 3, boil a few
minutes and precipitate the aluminum, iron and chromium hydroxides with
NH 4 0H. Filter, wash, dissolve the precipitate in HCI, and make up the solution to a definite volume.
In one portion the three hydroxides are precipitated together with NH 4 0H
and weighed. Another portion is treated in a flask with an excess of N a 202
until the iron hydroxide has assumed its characteristic reddish-brown color.
Dilute with water, filter, wash and determine the iron in the usual way. Render an aliquot of the filtrate from the iron precipitation acid with -HN0 3 ,
precipitate the aluminum hy~roxide with N:H 40H and weigh as AI20~.
Chromium is determined :il the filtrate by titration with ferrous ammonium
sulfate as described under Yellow Pigments. Any method for the separation
of the above hydroxides may be used in place of the one. outlined.
Calcium and Magnesium.-These cO"nstituents are determined in the filtrate
from the precipitation of the above hydroxides.
Sulfuric Acid.-One gram after ignition until all the blue has been decomposed, is dissolved in 30 m!. of cone. HCI, diluted with water, boiled, filtered,
and washed. The sulfuric acid is determined in the filtrate.
Nitrogen.-Determine as outlin:ed under Prussian Blue.
I
Calculation.-The Prussian blue is determined by using the factors given
under Prussian Blue. The sulfate is calculated to lead sulfate and calcium
sulfate, should calcium be present, and the chromiu_{n to lead chromate.
I

BLACK PIGMENTS
The black pigments irtclude those which contain carbon as their essential
constituent. The introduction of many black pigments which are made from
asphaltic and coal-tar mixtures complicates their chemical analysis. For those
pigments which contain coal-tar mixtures, recourse: may be had to works 20
covering this matter thoroughly.
19 "The Analysis of Paints," Gardner and Schaeffer, __

pp.

"36-37.
Allen's "Commercial Organic Analysis," 4th:Edition; "'The Analysis of Paints ,Gardner and Schaeffer.
20

ANALYSIS
OF PAINTS
,

1877

The analysis of the simple black pigments may be carried out in the following way:
Moisture.-Dry 2 grams at 105 C. for two hours.
Oil.-Extract 2 grams, with ether in a fat-extraction apparatus.
Carbon.-Determine the carbon by difference after determining the moisture, oil and ash.
Ash.-Ignite 2 grams to a bright red heat until all the carbon is driven off.
If graphite is present, the ignition must be carried out with the aid of oxygen.
Should carbonate be present, mix the ash with a small amount of ammonium
carbonate and again ignite, thus reconverting to carbonate any oxide which
may have been decomposed.
Analysis of Ash.-The ash is boiled with concentrated HCI and the insoluble
residue determined in the usual manner. The filtrate is examined for calcium,
magnesium and phosphoric acid.
Calculate the magnesium to phosphate, any residual phosphoric acid to
calcium phosphate and a~y residual calcium to carbonate.
0

PAPER AND PAPER-MAKING (MATERIALS

I

STANDARD METHODS OF THE TECHNICAL ASSOCIATION OF

THE PULP AND PAPER INDUSTRY 1
ISOLATION OF CELLULOSE BY CHLORINATION METHQD'
Apparatus.-A special chlorination apparatus as illustrated in Fig. 266 is
required for this determination. It consists of one leveling tube and holder,
two individual glass 3-way stopcocks, one thermometer, one Hempel precision
gas burette fitted with a 3-way stopcock, a glass water jacket for the gas
burette, one Hempel gas pipette, several Jena glass crucibles (No.3, 35 ml.
capacity) equipped with a fritted glass bottom of porosity 5 to 7. In addiVon,
alundum crucibles with about the same capacity and of porosity R. A. 98 are
required.
Solutions.-Alcohol-benr~ne: This sollltion consists of 33% alcohol and
67% benzene by volume.
Calcium Chloride Solution for Gas ,Burette and Gas Pipette.-Saturate
water with CaCl. at room teinperatureibubble chlorine gas in until saturated
and filter for use. When the apparatus is not in use it is well to have it filled
with fresh CI gas, thus keeping the CaCl. solution saturated.
Test Specimen.-The specimen shall consist of about 2 grams of a'representative sample of air-dry disintegrated wood or pulp of known moisture
content.
The samples 0" wood or pulp for analysis shall be prepared according to the
procedures described in the official methods.
Procedure.-Weigh accurately approximately 2 ,grams of air-dry wood or
pulp in a tared alundum crucible contained in a weighing bottle. Extract the
crucible with contents, in the case of wood (the extraction may be dispensed
with in the case of pulps), for 6 hours, or over night if more convenient, in a
Soxhlet extractor with the mixture of alcohol and benzene. Mter evaporation
, of the solvent wash the sawdust thoroughly with hot water, using the suction
pump. Then transfer the moist material to a Jena glasS-cru~ible. The crucible is water jacketed and connected between the gas burette and the gas
pipette, as shown in Fig. 266, by means of two rubber stoppers-through which
passes a capillary glass tubing. Apply suction first at the bottom of the
crucible until the excess moisture is removed, and then at the top. The latter
removes the water from the fritted glass plate and evenly distributes the
remaining moisture throughout the entire sample.;
Pass a burette-full of chlorine gas from the burette up through the material
in the cruGible a?-d over into the gas pipette as fast as possible. The tem1 Reproduced with the permission of the Technical Association of the Pulp and
./ _
-'
Paper Industry.
2 T 201 m.
Tentative Standard--8ept. 11, 1930. -Official Standard-Aug. 15, 1932.
Corrected-March 15, 1934.
-

1878

PAPER AND PAPER-MAKING MATERIALS

1879

perature of the water in the water jacket should be in the range of 23.5 to
32 C., inclusive. During the first chlorination samples of wood absorb approximately 230 m!. of Cl gas at room temperature and atmospheric pressure.
This necessitates refilling the burette, which may be done quickly as it is
connected to a Cl tank with a 3-way stop cock. The first chlorination treatment requires from 3 to 4 minutes, after which remove the crucible from the
apparatus and wash the material with approximately 50 ml. of distilled water,
and successively with. 50 m!. of approximately 3% 80 2-water, 50 m!. of water,
then 50 ml. of freshly prepared 2% sodium sulfite solution. Transfer the
material to a 250-m!. Pyrex beaker, using a pointed glass rod. The last traces
of material are washed out of the crucible with 100 m!. of 2% N a 280 3 solution'
with the aid of a rubber policeman drawn gently over the bottom of the crucible,

Stop CI)cfr8
J\ic~=lI-..W4te"

Cliff, I'

Thermomtfer
~top

Levelinq Tube

CockC

Sv,.el'te

Wdter

Jl1clret

FIG. 266.-Apparatus for Isolation of Cellulose.

Approximately 15-m!' portions of the N a 2803 solution are added to the

crucible for each washing. When 60 m!. have been thus used, place lO-m!.
portions of the remainder in a watch glass and by means of applying gentle
suction to the top of the crucible, placed on the watch glass as before described,
remove all of the material_from the fritted glass bottom of the crucible. Cover
the beaker containing the sample with a watch glass and place in a boiling
water bath for 30 minutes. Again transfer the fibers to the glass cruCible and
wash with about 250 m!. of distilled water.
The above procedure is never sufficient to remove all the lignin, so that the
treatment with Cl ttnd subsequent treatments as outlined above must be
repeated until the fibers show only a very faint tinge of pink upon addition of
the N a 280 3 solution. The second and following treatments with Cl should
not require more than 2 or 3 minutes. Prolonged action of Cl gas, together
with the HCl formed in the secondary reactions, hydrolyze the cellulose, give
low yields and cause varying amounts of alpha-cellulose.

1880

PAPER AND PAPER-MAKING MATERIALS

I

After all of the lignin has been removed, wash the fiber thoroughly in the
alundum crucible successively with 500 ml. of hot water, 50 ml. of 95% alcohol
and finally with 50 ml. of ether. Then dry it for 27Jl hours at 105 C. in an
electrically heated oven.
I
Finally place the tared alundum crucible in the ori~nal stoppered weighing
bottle, cool in a desiccator over conc. H 2S0 4 and weig4 as cellulose.
Report.-The weight of the residue shall be reported as a percentage of the
oven-dry weight of unextracted material.
Additional Information

Roe, J. Ind. Eng. Chern. 16, 8 (1924).

Bray and Andrews, Paper Trade J. 76, No.8 (1923).
U. S. Forest Products Laboratory Standard Test.ing Manual for Pulpwood, Pulp, Stuff,
and Paper.
Bray, M. W., Paper Trade J. 87, No. 24 (1928):

CHLORINE CONSUMPTION OF PULp3

This value, obtainable during the isolation of cellulose, is a measure of the
amount of chlorine gas actually consumed by compounds other than cellulose
in the pulp. Such values for a series of pulps indicate the relative bleachabilities of the pulps and with a sufficient background of qata can be used to
predict actual bleachabilities.
/
Apparatus.-The special chlorination apparatus as described in the standard method for the isolation of cellulose is required. (See Fig. 266.)
Specimen.-The 'specimen shall consist of about 2 grams of a representative
sample of air-dry disintegrated pulp of known moisture content.
Procedure.-The procedure used for the isolation of cellulose shall be
followed as described, together with certain additional steps:
Run about 90 ml. of CI gas into the gas burette and note the exact amount,
after which pass the entire amount of gas up through the sample and over
into the gas pipette.
'
NOTE.-This method is a revision of U. S. Forest Products Laboratory Official
method No. 35 [See Paper Trade J. 87, 25, 52~53 (1928)J.
Then run the unconsumed gas back into the gas burette' and note the
exact amount. By differ~nce, obtain the total volume of ga~ used. When the
Jena glass crucible containing the chlorinated sample is placed in the suction
filter flask, wash the pulp with four 25-ml. portions of water at a temperature
of 18-23 C. Keep this filtrate in a properly labeled and covered beaker.
At each successive chlorination repeat the above procedure and correct the
total volume of Cl used to standard conditions of temperature and pressure,
i.e. 0 C. and 760 mm., taking into consideration the val?or pressure of water
vapor at the chlorination temperature over a saturated CaCl 2 solution saturated with' Cl gas. This calculation is made by means of the following
formula:
I
3 T 202 m.
Tentative Standard-8ept. 11, '1930. . offi~~i Standard-Aug. 15, 1932.
Corrected-May 15, 1935.

PAPER AND PAPER-MAKING MATERIALS

1881

V,X(p-a)

Vo

760(1+0.00367t)

Where V,=volume of gas observed at t deg. C.

Vo=volume under standard conditions.
a=vapor pressure of water vapor over saturated CaCh solution saturated with Cl gas, in mm. mercury at t deg. C.
t=the temperature of chlorination in deg. C.
p=observed barometric pressure in mm; of mercury.
The following table'is of assistance in making these calculations.
tOe.

a.

760 X(l +0.00367 t)

826.00
23.5
15.9
826.90
24.0
16.6
24.5
828.36
17.3
829.76
25.0
18.0
832.57
26.0
19.4
833.80
26.5
20.0
27.0
835.40
20.8
838.12
28.0
22.0
29.0
23.5
840.96
843.66
30.0
25.0
846.48
26.4
31.0
849.30
32.0
29.0
The weight of 1 ml. of Cl gas at standard temperature and pressure = 0.003164 gram.
Titrate the combined initial wash waters with standard alkali to determine
the weight of Cl gas that formed HCI. Then determine by difference the
weight of Cl actually consumed by the pulp.
Report.-The data shall be reported as the percentage of Cl consumed,
calculated on the oven-dry weight of the sample.

ALPHA CELLULOSE IN PULP 4

This method as described is an attempt to formulate a procedure which is
accurate and sufficiently simple so that it may be used as a routine test. The
method herein described is adapted from the tentative standard Method IV
of the American Chemical Society 5 and is the most successful method investi~
gated by the Division of Cellulose Chemistry of that Society. It is applicable
to paper making fibers provided they are not highly lignified.
NOTE.-For the determination of alpha cellulose in paper see T 429 m.
Test Specimen.-The usual precautions for sampling pulp should be rigidly
followed. The sample shall be in an air-dried condition, in an open, fluffy
form, prepared by being brushed apart while dry by means of a disintegrator
of the Gosta Hall or Koerner type. It shall be stored in an air~tight glass~
stopper~d container or fruit jar for at least 48 hours to attain a uniform mois~
T 203 m-36. ;Tentative Standard-Feb. 19, 1931. Revised-Feb. 17, 1932.
Official Standard-Aug. 15, 1932. Corrected-8ept. 15, 1936.
.
6 (Ind. and Eng. Chern., Anal. Ed. 1, 54 (1929.

1882

PAPER AND PAPER-MAKING MATERIALS

ture content throughout the material. Samples for moisture content and
alpha cellulose determinations shall be withdrawn a$ needed.
Reagent.-17.5% NaOH Solution.-The sodium hydroxide solution is
prepared by dissolving solid NaOH in an equal weight of water, and allowing
it to stand until the N a 2 C0 3 and other impurities' h~ve settled, leaving a clear
supernatant liquid. The clear liquid is decanted and diluted with CO 2-free
water until its density at 15 C. is 1.197. Such a s61ution contains 17.50.1
grams of sodium hydroxide per 100 grams of solution.
..
Procedure.-Weigh accurately in triplicate about 3 grams in a weighing
bottle and transfer to a 250-ml. Pyrex beaker. Add 30 ml. of 17.5% carbonate-free N aOH solution and allow the whole to stand for 5 minutes. Macerate
the pulp for 10 minutes with a short rod which has the end flattened to a disk
1 cm. in diameter. During this maceration add 40 ml. of 17.5% NaOH solution at 20 C. intermittently in 10-ml. portions. Then cover the beaker with
a watch glass. After 30 minutes additional mercerizing action in a watr bath
at 20 C. (total mercerization time 45 minutes) add 75 ml. of distille water
(at 20 C.) to the alkali-cellulose mixture with thorough stirring. Fi ter the
contents of the beaker immediately by means of suction on a 40-ml. Gooch
crucible having a finely perforated bottom, allowing the cellulose to form it\'!
own mat. Pour the filtrate through the mat a second and third time, if
necessary, to catch any fine material. Wash the residue in the Gooch crucible
with exactly 750 ml. of distilled water (at 20 C.) by means of suction. Disconnect the suction tube, add 40 ml. ot 10% acetic acid (at 20 C.) and allow
to soak for 5 minutes. Then apply suction to remove the acid. Wash the
alpha cellulose with distilled water until free from acid. Carefully remove it
from the Gooch crucible to a tared, flat, glass-stoppered weighing bottle. Dry
the unstoppered bottle and contents to constant weight at 105 C., making the
first weight at the end of 6 hours and then at I-hour intervals until constant.
Let'the weighing bottle and its contents cool in a desiccator for 30 minutes
and cover before weighing.
I
NOTE.-Drying of the alpha cellulose can be gr!latly expedited if the last washing is
followed by 95% ethanol and then by ether or some other highly volatile non-reactive
liquid. In each case the washing fluid is allowed, to soak into the sample for several
minutes before being removed by suction. In thi~ way the time in the oven at which
constant weight is reached may be reduced to about 2 hours instead of 6 hours or more.
The results of the two drying methods check each other within the limits of experimental
error.
~,

Determine the ash content of one of the triplicate alpha-cellulose residues

in the usual manner, and use the other residues (in the case of mibleached pulps)
for the determination of residual lignin, as below.
Modification for Unbleached Pulps.-In the case of unbleached chemical
wood pulps, correct the alpha cellulose figure for li,gnin remaining .in the residue
after mercerization by the following procedure adapted from the method developed by the U. S. Forest Products Laboratory, Madison, Wis.: 6
Treat the residue from the mercerization test with 5 ml. of water in the
original weighing tube in which it was weighed; stopper"and allow \0 stand
over night at room temperature. Then remove lhe residue to a beaker and
6

(Paper Trade J. 87, 61 (1928.

PAPER AND PAPER-MAKING MATERIALS

1883

treat with 45 ml. of H 2S0 4, sp.gr. 1.695 (76.76% by weight), at 15 C. and let
the digestion proceed at 25 C. for 16 hours. Transfer the mixture to a 2-liter
Erlenmeyer flask,. dilute with 1570 ml. of water, cover with a watch glass and
boil for 2 hours, maintaining a constant level by the addition of boiling water.
Filter the residu"through a tared Gooch or sintered glass crucible, wash with
hot distilled water and dry to constant weight at 100-105 C. Correct the
alpha cellulose figure by subtracting the percentage of lignin residue thus
obtained.
~
N OTEs.-(a) 'Time interval for steps in the procedure: An interval of 15 rniAutes should
elapse between the times of starting the alkali treatment of the individual triplicate
samples. Such a procedure allows time before proceeding to the next sample in a tripli-

cate determination for the various steps in the procedure, such as impregnation, maceration, filtration, and partial washing.
If 5 minutes are allowed between the addition cf a part of the alkali and the beginning
of maceration, it will be found that sulfite pulp swells to five or six times its original
volume, and the fibers can be much more easily macerated. The remaining H)-minute
interval, during whicll the remaining 40 ml. portion of the alkali is added, making a '
total of 75 mI. h,Jl;s been found ample time for the maceration of pulps. 'Vith'such a
scheme, for exH.'mple, one would add the alkali to the samples at 9:00 o'clock, 9:15,
and 9:30, respectively. Filtering would begin at 9:45 o'clock; 10:00, and 10:15, respectively.
(b) FiUering lIrfedium: Jena glass filter crucibles (coarse) may be used in place of the
Gooch crucibles specified in the directions.
(c) Testing for acidity: To test the acidity of the wash water on litmus paper, the
last few drops removed after the addition of water should be used.

Report.-Report results on the oven-dry basis to tenths of a per cent.

Precision.-Duplicate samples should check each other to within 0.2.
f

PITCH IN WOOD PULP'

,

'S

Preparation of Sample.-If the pulp is sufficiently wet, it shall be separated

illt(l till", layer...lld allowed to dry ill. tb.e air (l'(ernigb.t. II tb.e ~mll' i. WQ
dry for thiJ; it shall be soaked in water, then separated into layers and allowed
to dry~ overriight. The layers shall then be cut into pieces' about 2 X 4 inches
and 5 grams weighed put for each extraction. Pieces shall be- creased alternately"parallel to the shorter direction so that they may he closed up like a
camera bellows. This ,is very necessary to keep the surfaces apart in the
extractor and facilitate rapid extraction. (If preferred, the pulp may be
shredded before extraction and packed into the extractor loosely.)
Where extreme accuracy is desired the moisture shall be determined on a
sepnrate'poi-tion'of-the air-dried pulp by drying a weighed sample to constant
weightat 100-105 C., and the pitch percentages subsequently obtained shall

~e corrected to tl-i'e air-dry basis by dividing the actual results by 1X

0.90

, where

X ,)s the p~rcentage of moisture expressed as a decimal. . 1;or ordinary purposes: however, the, strips dried overnight may be considered in the air-dry
c9ndition.,
.
.

1
I.

'

,_

T 204 m. Tentative Standard-Dec. 31,1931. Official Standard-March 1, 1933.

Corrected-May 15, 1935.
.
'1

1884

PAPER AND PAPER-MAKING MATERIALS

Ether Extract.-Accurately weigh 5 grams of the pulp prepared as pre~

viously described, place it in a Soxhlet or a continuous extractor, connected
to an unweighed Erlenmeyer or Soxhlet flask, and extract with sulfuric ether
for at least 16 hours with the ether boiling rapidly.
Filter the ether solution through a wad of absorbent cotton in the bof,tom
of a glass funnel, washing the funnel and the cotton finaHy with a Httle ether,
afld collect the filtrate in a weighed Soxhlet flask. Then evaporate the .olutioD in the flask, dry the residue on the steam' bath until the ether is removed
and finally dry in the oven at not over 105 C. to constant weight. CalCltlate
the weight of residue to percentage of the a~r-dry pulp and report as et;hersoluble pitch.
Alcohol Extract.-After completion of the ether extraction place another
unweighed flask under the extractor and extract the pulp sample with nelltral
ethyl alcohol (specially denatured alcohol-No. 30 may be used) for at jeast
16 hours in a Soxhlet or a continuous extractor. Filter the alcohol sohition
through cotton into a 'weighed flask as previously described, evaporate the
alcohol, dry the residue on the steam bath and then to constant weight at
not over 105 0 C. in the oven. Calculate the weight of extract to percentage
of the air-dry,pulp and report as alcohol-soluble pitch (insoluble in ether).
Q

NOTE.-In the case of sulfite pulp time may be saved by extracting a separate portion
of the prepared sample directly with alcohol to obtain the total pitch. In this casl'l the
difference between the percentage of total pitch and of ether-BOluble pitch is the alcpholsoluble pitch insoluble in ether.

Total Pitch.-The total pitch shall ,be reported as the sum of the percent'ages of the ether-soluble pitch and the alcohol-soluble pitch insoluble in e;her.
NOTE.-E. Richter 8 and E. Opfermann 8 were among the first to call attentic!1!- ~o
the fact that the pitch in sulfite pulp is completely soluble in alcohol and a portion oJ It IS
soluble in ether. Opfermann pointed out that by extracting su1fite pulp first v.'ith Fther
and subsequently with alcohol the sum of the two extracts is the same aa the total extract
~\.-.~\,'\1_m ~'"J 'al(!,<.':.~~\ 'il.h'\1_~ ~'\1..<i \>,.~ (\l.~\\~,{ ~'f.'i';~~ '\,\>"~wp\.%<.':.'\1_ ~\\'l!.,~ \.\, \.~ ~\\~ ~~b..~. . ~l)!hJfl.
portion of the pitch that causes the most trouble in paper making. In such cases as the
writer 10 has been able to fonow through, it has been found that unbleached sulfite pulp
causing pitch troubles in making ti$Sue paper showed an ether extract greater than) %.
As there seems to be still some question as to whether it is the total pitch in the pulp, or
the ether-soluble pitch which gives trouble, it seems generally advisable to deterJIune
both forins of pitch.

'CUPRAMMONIUM DISPERSE VISCOSITY OF PULPIl

.

"

The following detailed directions for: the determination of cuprammon ium

,viscosit:v:_ of pulps is an attempt to formulate a simple procedure which is
sufficiently accurate and reproducible that it, may serve as a routine test.
\

'C, A. 8, 2803 (1914) .

z, fUr Angew. Chern. 22, 436 (1909).,

This note was written in 1931 by R. C. Griffin who was at that time chairmo.n of
the Non-Fibrous Materials Testing Committee.
,
11 T 206'm.
Tentative Standard-Feb. 14, 1932. Official Standard-April 15, J933.
Tentative Revision-April 15, 1935. This revision has been , approved by the r~p
Testing Committee. Send comments to R. G. Macdonald, 122 E. 42nd Street, ~ew
York, N. Y.
10

PAPER AND PAPER-MAKING MATERIALS

1885

The method herein described is a modification of the Shirley ..Institute method

(1). This procedure has been found applicable to bleached and
unbleached chemical pulps and to rag stocks.
'. dj,
Apparatus.-Viscosity Tube.-The viscosity tube shown in Fig.
267 is used both for dissolving the pulp and for determining the
viscosity of the solution. The tube, of heavy Pyrex glass, has an
internal diameter of 1.0 em. and a total length of 27 cm. I The capillary is 2.5 em. long and 0.09 em. inside diameter. The wide portion
is etched with three rings, A, B, and C, at heights 24.2, 12.2, and 6.2
optional,
em. above the flat end of the capillary. (The mark at B
to be used with hard unbleached pulps. In this case the constant
for the tube between the marks A and B should also be determined.)
The applicability of the procedure to pulps of widely' varying
viscosities is greatly improved by using tubes of different efflux rates
for different pulps. The following efflux rates determined with distilled water at 20 C. are specified:
Seconds
FIG. 21\7.
Bleached pulp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 15-25
Viscosity
Tube.
Unbleached pulp and high viscosity rag stock ........ 10-15

r" "1

is

These specifications are in addition to those for the dimensions and do not
aYter the dimensions.

A-~-l!-_~I

.090 em.
Int.diam.

---

FIG. 268.-Jacket.

1886

PAPER AND PAPER-MAKING ,MATERIALS

The top of the tube is closed with a _rubber stopper, carrying another
capillary which has a ground-glass stopcock. The bottom capillary is closed
with a rubber tube and clamp. Each tube has a wedge-shaped cylinder,
li'ig. 269, the bottdm of which is notched.
l'he cylinder, mad~ from 7i-inch steel or
:M:onel rod, is 2.5 cmt long.
The glass jacket, Fig. 268, supports the
viscosity tube in the thermostat. The 0acket
if! so constructed that it supports the tubf on
FIG. 269.-The Wedge.
three glass points at the bottom and is a tlidiAg fit at the top.
Calibration of Viscosity Tube.-The volume of the tube is obtained as
follows: Fill the tube, containing the wedge, with water, put the rubber stopper 12 tightly in place and force the excess water out through the capillary.
Open the stopcock and drain the contents into a burette, noting the time in
seconds for the meniscus to fall from line A to line C. Read the volume of
the water delivered to the burette and take the average of several determinations as the volume of the tube.
The tube constants C and k are determined from the formula.:
I

(1)

f=!!_(t'-~),
t

C\

where V = viscosity in centipoises;

d = specific gravity of the solution;
t = time in seconds for meniscus to fall from line A to 0;
C=tube constant;
and k=gravity constant.
The constant C is fi~st determined with standard glycerine solution of
approximately 1.21 sp.gr. (85% glycerine solution; at 20.0 C. (The sp.gr.
of the glycerine solution must be accurately determined with a pycnometer.)
The viscosity of the glycerine solution may be ob'tained from a curve drawn
from Table I. In this case constant k may ,be disregarded, since the time of
efflux is long. The formula then becomes
(2)

dt

C=-v,

from which the tube constant is calculated. The gra~ity constant k is determined by repeating the operation with water at 20.0 C. and substituting for
C in formula (1) the value found above, as "Well as d = 1 and V = 1. The formula then becomes
(3)

Reagent.-Cuprammonium Solution.-Freshly prepare the cuprammonium

solution at least every 2 months as follows: Place clean copper turnings in a
glass 'tube 24 i;ncbes long ann 4 !ffl'Cb~ i1'. Ui'i>llyet'e;,__:-un,\)\'ir.i.~ '1I~~u\-, ~\'_,'C. '\'T\-.'C
12 The rubber stopper must be replaced periodiclt"liy-by a fresh one, determining the
new volume each time.

PAPER AND PAPER-MAKING MATERIALS

1887

turnings. must be previously cleaned by washing in dilute HCI.) Pour in

cone. NH 40H (26-28% NHa) containing 2 grams of cane sugar per liter until
the tube is nearly full. Bubble air, which is passed through a wash-bottle
containing cone. NH 40H, through the solution for several hours, until the
desired Cu concentration is reached. Analyze the solution for Cu and NH a,
and make the proper adjustments. The Cu concentration shall be 14.8-15.2
grams per liter, the NHa concentration 190-210 grams per liter.
The course of the reaction can be followed by estimating the Cu content
colorimetrically by comparing with a standard, both diluted 50~fold.
Determine the Cu and NHa in the cuprammonium solution as follows:
(a) Copper: Pipette 10 m!. of the cuprammonium solution into a 250-ml.
Erlenmeyer flask, and dilute with 25 mI. of water. Boil the solution until no
more NHa is given off. Add 5 mI. of HNO a (1 : 1) and boil till red fumes are
~xpelled. Then add 2 mI. of bromine water and boil until the Br is all expelled. Remove from the heat, dilute to 75 m}. and carefully add NH 40H
until the solution becomes just alkaline (blue) but not more. Add a slight
excess (4 or 5 mI.) of glacial acetic acid, cool to tap water temperature, and
add 10 m!. of 30% KI solution. Titrate with 0.1 N thiosulfate solution until
the yellow color of free iodine is nearly gone, then add starch indicator and
continue the titration cautiously until, on the addition of another drop, the
blue color due to free I disappears.
Calculation: Copper content of cuprammonium solution in grams per liter=ml.
0.1 N thiosuIfateXO.636.

(b) Ammonia: Pipette 2 mI. of the cuprammonium solution into 50 ml. of

normal H 2S0 4 solution, keeping the tip of the pipette beneath the surface of
the liquid. Titrate the excess acid with normal N aOH, using methyl red
indicator.
Calculation: NHa content of cuprammonium solution in grams per liter = (5O-ml.
NaOH)XS.5-0.536 C.

Where C = copper concentration in grams per liter.

Care should be taken that no liquid adheres to the outside of the pipette
before it is inserted into the acid solution. The correction factor (0.536 C) is
made for the alkalinity of the copper hydroxide.
Test Specimen.-A representative pulp sample in the air-dry condition
shall be shredded by means of a mechanical shredder or a coarse steel file.
Wet pulp may be made into several handsheets, dried in the oven at 100 C.
for 3 minutes, and then shredded.
The air-dry pulp shall be placed in a stoppered bottle and the moisture
content determined on a portion of it. The weight of air-dry pulp equivalent
to 1 gram of oven-dry pulp for each 100 m!. of tube. volume shall be calculated
for .each tube and the proper amount of sample weighed out for the tube
employed. '
Procedure.-Take the calculated weight of oven-dry pulp to give a 1%
solution and loosely roll it into the form of a cigarette in a strip of hard-surfaced
paper such as glassine; then moisten this roll with 5 or 6 drops of water and
slide it from the paper into the viscosity tube. Siphon the cuprammonium
solution into the bottom of the tube by means of the rubber tubing at the

1888

PAPER AND PAPER-MAKING MATERIALS

I

bottom, until it is two-thirds full. Quickly break up the pulp roll with the
aid of a thin glass stirring rod. Add the remainder of the solution at the top,
scrape the rod 9lean and in'ser:t the stopper. Force th\l excess solution through
the capillary and close the stopcock. Wrap the viscdsity tube in black cloth,
and place it on a wheel rotating 3 or 4 r.p.m., so th~t the metal wedge falls
freely through the solution as the tube rotates. The pulp should be well
broken up so that the metal wedge will fall through the full length of the tube,
before the tube is put on the rotating wheel. Allow 15 hours rotation (over
.
night) for the complete dispersi9n of the pulp.
Place the tube in the thermostat at 20 C. till equilibrium is establi hed.
Remove the rubber tubing from the capillary and pla,ce the tube in po ition
in the jacket, which remains in the thermostat. Remove the rubber stopper,
allowing the pulp solution to flow through the capillary. Note the time in
seconds (t) for the meniscus to pass between the marks A and C, and calculate
the viscosity of the solution from the formula:

v= C
~

(t- ~)t '

in which d, the sp.gr. of the pulp solution, is(l.96; C and k are constants found.
Simplifying, the equation becomes:
where
Inspection will show when the expression kit may be disregarded, as is
usually the case with unbleached pulps of high viscosity.
The tubes may be cleaned by immersion in nitric acid (1 : 1).
Report.-Results shall be reported in terms of centipoises and represent
the average of at least two tests.
.
Precision.-Duplicate samples should check within 2.0% on bleached, and
3.0% on unbleached pulps.
/
TABLE I.-VISCOSITY OF GLYCERINE-WATER SOLUTIONS

(International Critical Tables, Vol. 5, p. 23)

I

Specific Gravity at
20.0 C.

Viscosity in Centipoises at
2(}.0~ C.

1.0000 ...................... " .... , . .

1.009~1.1014. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.734
1.1699 .............. " ............ '"
15.360
1.1848. . . . . . . . .. . .. .. . . . . . . . . . . . . . . .. 23.540
1.2057 ........................ '. . . . . .. 49.670
1.2155. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 74.780
1.2240 ........................... ! ... 110.800
1.2463 ............ ; ................ " 409.300
1.2568 ............................. " 871. 700

Literature Cited
I

1. Clibbens and Geake, "The Measurement of the Fluidity-of Cotton in Cuprammonium

Solution," J. Textile Inst., 19, 77-92 T (1928).. _,. ~
2. Committee Report of Am. Chern. Soc., Ind. Eng. Chern., Anal. Ed. I, 49-51 (1929).

PAPER AND PAPER-MAKING MATERIALS

1889

MOISTURE IN PULP BY TOLUENE METHOD 13

This method is a rapid and accurate procedure for the determination of
moisture in comparatively small laboratory samples, but is not intended to
be used in place of T 210m in acceptance tests of pulp shipments.
Since the method does not include
as water any other volatile impurities such as turpentine, unless they
are soluble in water, it is especially
valuable for ground wood pulp.
Apparatus.-The apparatus required is as follows: (1) a balance
with a capacity of about 1 kilogram
and a sensitivity of at least 0.1
gram; (2) a LiebIg condenser with
a jacket length of at least 24 inches;
(3) a' straight drying tube charged
with anhydrous CaCb; (4) a moisture trap; (5) an accurately graduated and calibrated cylinder of
100-m!. or 50-ml. capacity; (6) a
2-liter Pyrex short-necked balloon
flask; (7) a source of heat, preferably an electric heater, although a
Bunsen burner is adequate; (8) cork
stoppers and rubber tubing for connections; (9) iron supporting stand
and clamps.
The moisture trap follows the
general plan of that used in the
Dean-Stark water determination
apparatus 14 but with certain modifications. It is considerably larger
than the lO-ml. size ordinarily used
and should hold about 100 ml.
The exact size is not important.
The lower portion only is graduated and is of smaller diameter than
the upper part, since the boundary
meniscus between toluene and
water in a glass tube is much more FIG. 270.-Apparatus for the Determination
rounded and harder to reproduce

of Moisture in Pulp.
in the same shape each time than
is the air-water meniscus. A glass stopcock is sealed on at the bottom to
permit drawing off the water for measurement. The other apparatus is of
the ordinary type and needs no particular description. The drying tube at
T 208 m. Tentative Standard-Feb. 16, 1932. Tentative Revision-April 6,
Official Standard-March 15, 1934.
14 Obtainable at any chemical apparatus supply house.
13

1933.

1890

PAPER AND PAPER.MAKING ,MATERIALS

the top of the condenser is necessary to prevent condensation of moisture from

the atmosphere if the humidity is high. This assenibly is shown in Fig. 270.
Reagent.-Toluene.-The toluene should be reasonably pure, prefiablY
colorless, clear and free from visible water droplets. It is unnecessary t dry
it, especially if the above conditions are met. It may be used over a in if
redistilled and dried to the above specifications. ~t is unnecessary for the
sample in the flask to be covered by the toluene. The amount of toluene for
each determination (500 m!.) need only be measured approximately.Test Specimen.-Pulp samples should be collected in acco:r:dance with the
present standard methods for pulp sampling and should at once be placed in
airtight containers (cans or jars) neither heated nor cooled, and analyzed as
soon as possible.
The samples should not exceed about Yz pound in weight for each determination. If, as in the strip or wedge method of pulp sampling, they' must
greatly exceed this, then separate determinations should be 'made.

NOTE.-'Vhere samples are very large and must be so in order to get accurate
sampling, the method is applicable by breaking up/the large samples into,seveia}.,srnaller
ones or by using a 5-liter flask containing 1000 ml. of toluene. If this requity> more
laboratory attention than is available, the dryil}g method must be resorted to.
.
Procedure.-Weigh the closed jar or can containing the sample to the
nearest 0.1 gram, cover and all. Then open and transfer about lob to 300
grams of the pulp to the flask (which ~lready contains 500 m!. of toluene) as
rapidly as possible. Replace the cover on the container and again weigh.
Connect the apparatus, turn on the cooling water and boil the toluene until
no more water collects. Either during the boiling or at the end, if the trap
does not become completely filled with water, draw off the water through the
stopcock into an accurately graduated cylinder, cool to below 30 0 C. and
measure its volume.
When draining the water from the apparatus into the measuring cylinder,
allow time for the water to drain from the walls ana drain the last portions of
water into the cylinder by touching its lip to ,the draining tube until all water
is removed.
Since a 100-m!. cylinder is graduated only to about 1 m!., a more accurate
measurement can be made, for smaller amounts of water, by using a 25-m!'
or 50-m!. graduated cyJinder or a burette. If the amollp.t of water is small,
due either to a small sample or a very dry sample, it need not- be drained off
but can be measured on the graduations at the bottom oLthe trap. The
cylinders should be accurately calibrated for use, both for water alone and for
water with a toluene layer above it. It sometimes happens that a few drops
of toluene remain in the tube at the bottom of the trap and when the water is
drawn off this toluene is also removed. Sinoe! the water meniscus is of a
different shape with air than with toluene, the graduate should be calibrated
for both Qonditions. Trap and graduates may be calibrated as follows:
Set the ary, clean trap or graduate in a vertical position, run in water from
an accurate burette up to the first test interval graduatiorl (5 or 10 ml. for the
graduates, 1 m!. for the trap) and read the bl.lr.ette_- This is then repeated for
each interval used. The operation should also be repeated with 5 ml. of

PAPER AND PAPER-MAKING MATERIALS

1891

toluene in the trap or graduate. The calibration is conveniently represented

by graphs showing reading along one axis and actual v1)lume along the other.
To minimize the difficulties due to the meniscus assuming different shapes,
the tubes should be cleaned frequently with hot chromic-sulfuric acid cleaning
solution. A few drops of water usually collect in the lower part of the condenser tube just above the height reached by the toluene vapors. Where the
amount of sample is 100 grams or niore or the amount of water 10 ml. or more,
the amount of these drops is too small to be considered; but to minimize it,
the condenser should be cleaned frequently with hot cleaning solution.
Report.-The results shall be reported in percentage' of water by weight,
assuming each ml. of water to weigh 1 gram.
Precision.~The method is accurate to within 0.2 ml. of water. What
this represents in percentage accuracy depends on the size of sample and
amount of water. For amounts of water under 10 ml. and with accurate cali. bration of the trap with water under toluene, this may be increased to an
accuracy of 0.1 ml.
Additional Information
1. U. S. Forest Products Laboratory Bulletin: "Chemical Analysis of Pulp and Pulp
.
Woods," Method No. 13 (1928).
2. Schwalbe, Zeit. fUr Angew. Chern., 21, 408 (1908).

METHOXYL GROUPS IN PULp!5

Apparatus.-The apparatus shown in Fig. 271 is used for the determination
of the methoxyl groups. It cohsists of a 50-ml. flask (B) connected by a
ground-glass connection to a condenser the top of which forms a trap (D)
which contains a water suspension of red phosphorous sufficient to well cover
the end of the trap. At the upper end of the trap a glass tube is sealed on,
leading to a series of two bubble flasks (E). All connections are ground glass.
A preheater (X) is connected to the condenser jacket, as shown, to bring
the condenser to the proper temperature.
A source of dry, pure CO 2 is connecte!l to flask B.
Solutions.-Siluer NitTate.-The AgNO g solution is prepared by dissolving
17 grams of the C.P. salt in 50 ml. of water and making it up to a liter with
previously purified 95% ethyl alcohol. The aldehydes are removed from the
alcohol as follows: Dissolve 1.5 grams of AgNOg in 3 ml. of water, add to 1
liter of 95% ethyl alcohol and shake well. Dissolve 3 grams of N aOH in
about 15 ml. of 95% ethyl alcohol. Warm the alcohol solution gently, cool
and add to the main solution. Shake the mixture well, allow to settle for
several days, siphon off the clear solution and distill, adding a piece of pumice
to keep the liquid from bumping.
Hydriodic Acid.-If the acid as purchased is not water-white and of.sufficient sp.gr. (1.7), it must be redistilled. The acid is placed in a distillation
flask and a few grams of red phosphorous added. The sides of the flask are
protected from over-heating by setting the flask on a piece of asbestos which
16 T 209 rn.
Tentative Standard-Feb. 16, 1932. Official Standard-April 15, 1933.
Corrected-March 15, 1934. Official U. S. Forest Products Laboratory, Method No. 43,
rearranged.

1892

PAPER AND PAPER-MAKING MATERIALS

has a hole cut in the center sufficiently large to permit heating of onlq the
liquid in the flask. .The distillate coming over betiween 124-130 C. i col.~
lected and stored in a glass-stoppered brown bottle.
Test Specimen.-The specimen shall consist of a~out 0.5 gram of oven-dry
I
pulp, accurately weighed.
Procedure.-The methoxyl groups are determined according to the wellknown method of Zeisel. The substance to be examined is heated together
with hydriodic acid, whereby methyl iodide is formed. The CHaI is swept
from the reaction flask by a current of CO 2 into vessels containing an alcoholic
0.1N AgNO a solution, the CHaI being decomposed 'with the formation of AgI.

WAreR OUr,_cr-

-,

OIL BATH At 1.1D~

#

.... _ _ _

FIG. 271-.Apparatus for Determination of Methoxyl Groups.

Weigh approximately 0.5 gram of oven-dry pulp into the digestion flask
(B), add 15 m!. of redistilled HI (sp.gr. 1.70), and heat the reaction flask in a

glycerine bath at_130 C.

,
1'he temperature of the water in the condenser, used in ,this apparatus,
is maintained at 45-55 C. by means of the preheater (X) through which water
is kept flowing constantly by being connected to a water tap.
Absorb the CHaI in 35 m!. of AgNO a solution placed in the absorption
apparatus (E). Pass a current of dry CO 2 gas ,through the apparatus at a
rate of 60 bubbles per minute. Continue hen:'t:mg until the AgNO a solution,
at first milky, has become perfectly clear. This requires 15 minutes to 2 h0urs.

PAPER AND

PAPER~MAKING

MATERIALS

1893

Wash the contents of the absorption bulb into a beaker, make up to 200 mI.,
add 10 mI. of conc. HNO a and boil the mixture gently for at least 15 minutes.
After cooling this mixture filter on a tared Gooch filter. Wash the precipitate,
dry for 172 hours at 1050 C., cool in a desiccator and weigh as AgI. Calculate
to methoxyl, CHaO.
Weight AgI
Calculation: Per cent methoxyl = 13.2 X W . h
I
eIg t samp e
Report.-The data shall be reported as a percentage of the oven-dry weight
of the sample used.
Additional Information

Zeisel; Monatsh, 6, 989 (1885); 7, 406 (1886).

Manning, R. J. and Nierenstein, M., Ber. 46, 3983 (1913).
Fritz Weishut, Monatsh, 33, 1165-12 (1912).
Klemene, A., Monatsh, 34, 901-12 (1913).
Benedict and Bamberger, Monatsh, 15, 509 (1894).'
Br::y, M. W., Paper Trade J. 87, No. 25; 5!H:l8 (1928).
WEIGHING, SAMPLING AND TESTING WOOD PULP FOR
MOISTURE 16
The determination of moisture in wood pulp varies with the form in which
it is manufactured and the methods selected for moisture testing are! specified
for each given form or kind of pulp. These include: (1) baled pulp: dried
sheets, shredded pulp, hydraulic pressed laps; (2) roll pulp; (3) loose lap pulp:
wet laps, hydraulic pressed laps; (4) double-press wet-machine pulp in sheets.
NOTE.--Other conditions also affect the testing of pulp, such as the place, time and
quantity of pulp. The place of testing may be at mill during manufacture; at dock during transfer from car to ship for export or vice versa; incoming car lots or shiploads at
rec~iving p,oint. The pulp m_ay be freshly made or stored in warehouses or open piles.
Apparatus.-Sampling Tool.-For testing baled dry pulp, a boring tool
which cuts a disk about 4 inches in diameter is used. For baled shredded
pulp, a special tool (Crossley Sampling Tool made by Thompson Mfg. Co.,
Lancaster, N. H.) similar to' a cork borer is used to cut a sample 1~ inches in
diameter and to a depth of 4 inches.
Template.-For the wedge method a template is desirable with an apex
angle of approximately 4 to 9 deg. This is preferably made of brass and having
a length of about 18 inches.
Laboratory Scales.-Scales used for weighing pulp samples shall show a
sensibility of 0.1 % of maximum load. For example, if 1 kilogram (2.2 pounds)
of pulp sample is taken, the scales must show a decided deflection by the
addition of 1 gram to the load.
Drying Oven.-Any suitable laboratory drying oven may be used for
drying samples to constant weight. For the regulation of temperature, a
thermometer shall be inserted, preferably in such manner that the bulb when
suspended in air extends as close to the center of the oven as possible. In case

16 T 210 IIi-36.
Official Standard-Feb., 1926. Revised!...,_Feb. 15, 1933. Revised
-April 15, 1935. Revised-Sept. 15, 1936..

1894

PAPER AND PAPER-MAKING MATERIALS

a thermostatically controlled oven is used,. it shall be set so as to yield'a constant temperature of 215 0 F.5 (102 0 C.3") and the mercury therniom~er
inserted as before. The thermometers used in these tests should ?e tested/for
~ correctness.
I
\ .
Test Specimen.-No one method of sampling i~ adequate to the many
different forms of pulp. A brief mention of sampling methods and their application will be made but their fuller description will be included under the
procedure to which they apply.
A. Strip Method.-The strip method, while well adapted to mill sampling
of unpressed wet laps or roll pulp, is not suitable for frozen lap pulp at receiving
point.
I,j
B. Boring Method.-The boring method, while well adapted to' the
sampling of baled pulp and rolled pulp, is not suited to the sampling of wet
lap pulp, under 36% oven-dry content (40% air-dry).
C. Wedge Method.-The wedge method, or any other method which
involves the breaking open of bales, is not suitable for referee sampling of
pulp at dock, since transportation companies decline to handle broken bales,
and rebaling with an ordinary hand press is not practicable.
D. Number of Samples.-Sufficient and representative samples from
individual units must be taken to i!lsure a fair average of the lot. This is
especially necessary where grab samples from opened bales are taken.
'
Procedure.-Determination of wet Weight of Pulp.-}~.s all pulp is invoiced
on the basis of air-dry tons, rules for tIle determination of the wet weight of
the pulp are just as importa'nt as rules for sampling and testing, and the. wet
weight shall always be determined by one of the following methods:
A. Railroad Weight: Railroad weight of entire car lot where tare of empty
car is actually determined by weighing it. (This does not mean the routine
bill of lading railroad weight, but an actual weighing upon railroad scale.g
supervised. by a party or parties at interest.)
NOTE.-Each car shall be uncoupled from the train and weighed separately.

B. Certified Weight: Weight of entire car lot as Icertified by official recognized weighing bureau issuing weigh-master's dertificate of weight.
C. Weighing by Truck and Tested Scales:' Wet weight of lot by sum of
weights of truck loads passing over accurately' tested scales during loading or
unloading.
.
D. By Calculation from Average Weight of Parcels: W~t wcight of lot may
also be determined by multiplying the actual number of bales 'or rolls in the
shipment, as determined by accurate count, by the average weight of the bales
or rolls weighed and sampled. When this method is used for determining the
total weight, all the bales weighed, and none others, are to be sampled. In
domestic roll or wet bale pulp the variation in weight between individual rolls
or bales is often so great that it'may be necessary tp weigh the entire shipment.
In the case of roll pulp or shredded pulp at least 25% of the rolls, and preferably the entire shipment, shall be weighed.
The accurate weight of all bales sampled shall be ascertained by a competent weigher before sampling; all bales weighed and no othefs shall be sampled
(except in case of roll pulp and shredded pulp).;..whenever'bales are numbered
the number shall be given ,in addition "to the weight.
r

PAPER AND PAPER-MAKING MATERIALS

1895

NOTE.-Only normal bales shall be selected, A normal bale shaH be intact and
unbroken. Reject as abnormal those bales obviously damaged or whose weight differs
from the average gross weight, as determined at 'he time of testing, by more than 10%
plus or minus.

Scales: Scales will be accepted as accurate: (1) When verified. by standard

test-weights; or (2) when verified by weighing a known or measured volume
of water. Scales for weighing bales shall be provided with' 1 pound divisions.
Those equipped with counterbalances shall not be permissible. They shall
have such accuracy that they will show a variation of not more than 1 pound
when a standard 400-pound weight is placed upon them after having been
balanced.
'
. Testing at Pulp Mill.-The strip method shall be used for sampling pulp
coming from the wet machine or from driers in continuous web. Cut a 2 or
3-inch strip across the entire width of web coming from the machine. All
strips cut shall be the same width and uniform across the machine. One
sample as above described shall be taken from the wet machine for every 2000
pounds wet weight production. For web-dried pulp, one sample across the
machine shall be taken for every fifth bale or roll of production.
The sampIes taken shall be immediately deposited in l} metal can with
tightly fitting cover and the net weight obtained as soon as practicable after
taking samples. In case of unavoidable delay between time of taking samples
and weighing, special precaution shall be taken to seal the edges of cover
air-tight with friction tape or otherwise.
Weighing of Samples: Accurate laboratory scales and weights shall be used
for weighing both wet and dry .samples. Both wet weight and dry weight
shall be obtained upon the same scales with the same weights. The hot
samples shall be weighed immediately after being withdrawn from the oven
in closed containers, taking care to avoid convection currents while weighing.,
The entire bulk of the samples selected shall be dried out for the test.
After weighing the wet samples, these shall be placed in a suitable oven and
dried to constant weight at a temperature of not less than 210 0 F. (99 0 C.)
and not more than 220 0 F. (105 0 C.). A minimum allowable temperature of
at least 216 0 F. (102 C.) shall be maintained for determination of final dryness.
Two successive weighings at least three hours apart shall not show a variation
greater than 0.1 % of the wet weight of samples and the total of the minimum
weights shall be taken as the final oven-dry weight. To calculate the percentage of oven-dry pulp, divide the weight of oven-dry pulp by the wet weight
and multiply by 100. To calculate the air-dry percentage, divide the percentage of oven-dry pulp by 0.9. Multiply the wet weight' by this air-dry
percentage and divide by 100 to obtain the weight of air-dry pulp.
Testing on Dock or Receiving Point.-The moisture determination is the
same as in 2, but' the sampling is different.
A. Boring Method: This is suitable for sampling (1) baled pulp in sheets,
(2) roll pulp, (3) baled shredded pulp having 50% moisture or less and (4) hydraulic pressed laps having 64% moisture or less.
(1) Baled Pulp in Sheets:
a. Number of Bales to be Sampled: On lots of 3000 bales or less, 10% shall'
be taken, except by agreement, but never less. than 15 bales. Samples shall
be drawn from only sound and intact bales from different sections of the 'entire

1896

PAPER AND PAPER-MAKING ~ATERIALS

shipment and the chemists shall be careful to observe that no unusual conditions prevail in the selection of bales.
j
b. Details: The samples shall lfe taken by boring into each bale to a dep1lh
of 3 inches, with a special tool which cuts a di.sk about 4 inches in diametir.
Care shall be taken to see that the knife or boring tobl blade does not bend
when taking samples. The disks shall be removed an~ ten of them taken as
a sample, selected as follows:
One disk from second sheet from the wrapper.
Two disks 1 inch deep.
Three disks 2 inches deep.
Four disks 3 inches deep.
The holes to be bored shall be so located that in five successive bales they
will represent a portion extending diagonally across the bale .. Each bale shall
be bored but once, the first bale at the corner, the edges of the cut being at a
distance of 1 inch from the edge of the bale. The second cut shall then be
made half-way between the corner and the center of the bale; the third bale
shall be cut at the center; the fourth bale halfway between the cEffiter and the
corner; and the fifth bale in the opposite corner in a position corresponding
to the first. In case binding wires or straps interfere with the exact location
of cuts, borings shall be made as near the .prescribed location as possible.
(2) Roll Pulp:
/
a. Number of Rolls to be Sampled: At least 10% of the lot, but not less
than 30 rolls. Where possible to determine, the same proportion of outside,
rolls to ce~ter rolls should be maintained as coming from the machine.
b. Details: The edge of the first hole shall be 2 inches from one end; holes
in succeeding rolls shall be spaced successively one-fifth of the distance acrosi3
the roll as in bale sampling, traveling toward the opposite end of the roll.
Selection of disks shall be in accordance with bale boring as given above.
I

NOTE.-As rolls usually vary considerably in weight,

entire shipment or at least 25% of it.
.;

it is preferable to weigh the

ASH IN PULPY
Apparatus.-A crucible, such as platinum, alundum,.or ,porcelain, which
has a tightly fitting lid is required. It should have a capacity of30 ml. or more.
'
An analytical' balance sensitive to 0.1 mg.
Any suitable desiccator, such as the customary analytical desiccator which
employs sulfuric acid or calcium chloride as the drying agent, may be used.
An electric muffle furnace is recommended for igniting the pulp sample.
A Meker or other gas burner may be employed for this purpose.
A suitable drying oven with temperature control' between 100 0 C. and 105 0 C.
should also be available.
Test Specimen.-Approximately 5 grams of pulp shall be, used for each
J
determination.
17 T 211 m-36.
Tentative Standard-Feb. 15, 1933.:- Official Standard-March 15,
1934. Corrected-8ept. 15, 1936.

PAPER AND PAPER-MAKING MATERIALS

1897

The specimen shall be representative of the lot being tested. If a carload

of baled pulp is being tested, at least 10% of the bales shall be represented by
small pieces composing the portion to be ashed. Reasonable modifications will
necessarily be made in sampling lots of different sizes and different packing
methods.
Procedure.-Ignite the empty crucible and cover over a burner, cool in a
desiceator and weigh to the fourth decimal place in grams. Place the test
specimen in the crucible and, after again weighing, put the whole in the drying
oven, and dry the pulp to constant weight. The crucible and contents should
be cooled in a desiccator and the crucible covered during weighing to prevent
absorption of moisture from the air.
Place the crucible and contents in the muffle furnace (or over the burner)
and ignite until all the carbon is eliminated: Heating should take place slowly
at the start, and the crucible should be protected against strong drafts at all
times to avoid mechanical loss. Cool the crucible in the desiccator and weigh
accurately. Repeat the heating, cooling and weighing until constant weight is
reached.
Calculate the percentage of ash on the oven-dry weight of the pulp and
.multiply this result by 0.9 to obtain the percentage on the air-dry basis.
Report.-The amount of ash sh~ll be reported as percentage of the air-dry
pulp to the nearest 0.01.
Precision.-Duplicate determinations should check within 0.03.
ONE PER CENT CAUSTIG SODA SOLUBILITY OF PULP 18
NOTE.-For the determination of 1% caustic soda solubility of wood see T 4 m-36.
o

This method determines the resistance of a pulp sample to solution by hot

dilute :;tlkali. In general, the less the chemical degradation of pulp, the lower
is its alkali solubility. The bulk of the material dissolved consists of pentosans
and other carbohydrates which are less resistant to chemical action than cellulose, and of cellulose which has been partially broken down during the process
of preparation. These are the materials which in part make up the fractions
known as beta and gamma cellulose when pulp is treated in the standard
manner with 17.5% NaOH solution. A cooking or bleaching treatment which
is too drastic causes a chemical degradation which gives an inferior pulp,
deficient in such physical properties as strength and opacity. This degradation
is clearly shown by an increase in the percentage of alkali-soluble material.
The test is particularly sensitive to degradation resulting from the bleaching
operation and the percentage soluble will show an increase before any serious
loss of physical properties is apparent. When the pulp has been overbleached
enough to show a marked effect on its physical properties, the amount of
soluble material will show a marked increase, often amounting to 50% of the
original solubility. Thus this test is a. valuable control test to detect when the
cooking or bleaching processes are too drastic.
Apparatus.-The water bath employed in this determination must be so
designed as to insure the temperature of the material during treatment is
18

1934.

T 212 m-36. Tentative Standard-Feb. 15, 1933. Official Standard-March 15,

Corrected-May 15, 1935. Corrected-March 27, 1936.

1898

PAPER AND PAPER-MAKING MATERIALS

uniformly maintained at 97 to 100 C. When a new bath is used, the temperature of the material during the treatment shouid be rechecked. The
type of bath recommended is one which is covered and has holes in the t~
large enough so that beakers may be set down into the. bath until.they are su ported by the flared rim. The top of the beaker shoJld be nearly level w th
the cover of the bath. The sides of the beaker are ~ntirely surrounded by
boiling water or steam when this type of bath is used. The height of the water
in the bath should be maintained at a level above that of the liquid in the
beakers.
Tall form 200-ml. Pyrex beakers should be used. Fritted glass. crucibles
(porosity 3) are recommended for filtering the treated pulp, although~lundum
crucibles of a similar porosity are also suitable.
Caustic Soda Solution.-The 1 % NaOH solution (0.25 normal) is made by
dissolving 10 grams of solid NaOH in water and diluting to 1 liter. Determine
the NaOH content by titrating with standard acid, first using phenolphthalein
as an indicator, then continuing the titration using methyl orange indicator.
The difference between the volume used for the phenolphthalein end-point
and that used for the methyl orange end-point represents that consumed by
half the Na 2 CO a present. By subtracting ,this difference from the volume
used for the phenolphthalein end-point, the volume required to neutralize
the NaOH is obtained, and from this volume the NaOH concentration is
calculated. This shall then be adjusted to between 0.9 ana 1.1%.
'
Test Sample.-A sample which is representative of the pulp being tested
shall be air-dried and placed in an air-tight container. The moisture content
shall be determined by the standard procedure. If the pulp is in the form of
machine-dried or pressed sheets which do not absorb liquid readily" it should
first be disintegrated to slush form in water, then filtered into pads which 'are
dried without any pressing. In this way the pulp is obtained in a form which"
readily disintegrates in the alkali solution.
Procedure.-Accurately weigh on an analytical ballwce two samples of the
air-dried pulp each equal to 2 grams of oven-dry pulP' plus or minus 0.1 gram.
To each sample in a tall-form 200-ml. beaker, add 100 ml. of 1 % NaOH solution
measured with a pipette. Stir the pulp until .it is completely disintegrated,
cover the beakers with watch glasses and place:in the, water bath, which.must
be boiling steadily. Leave the beakers in the bath for e~actly 1 hour. Stir
the contents three times, 'at 10, 15, and 25 minutes after heing~p.laced in the
bath. At the end of the 1-hour period, filter the contents of each beaker by
suction on a tared crucible. First wash the pulp briefly with hot water, then
with 50 ml. of 10% acetic acid, and then wash thoroughly with hot water.
Dry the crucible and contents at 105 C., cool in a desiccator, and weigh in a
stoppered weighing bottle. Calculate the loss in( weight of the oven-dried
sample.
Report.-The results shall be reported as percentage of the oven-dry
weight soluble in the NaOH solution. This value may be corrected if desired
for hot-water solubility by subtracting from it the percentage of the oven-dry
material soluble in hot water as determined by the.standard T APPI method
(T 207 m). In this case the result shall be reported as " corrected."

PAPER AND PAPER-MAKING MATERIALS

1899

Precision.-The results of duplicate tests should never vary by more than

5% of the average result, and by careful work the deviation can be kept below
2%. The results are usually somewhat higher than those obtained by the
method previously used because with the old method of heating the temperature
of the material during treatment was often below gOo C.
.
SAMPLING PAPER FOR TESTING 19
Test Sample.-The test sample, unless otherwise specified, shall consist
where possible of at least ten sheets not less than 11 X 11 inches in size. The
sheets shall be kept smooth and flat and protected from exposure to direct
sunlight and from contact with liquids and from other harmful influences.
Particular care should be exercised in handling the sheets if optical tests or
tests for acidity are required as such tests must be made on areas untouched
by the hands.
:procedure for Sampling.-The sample sheets shall be so selected as to be
as representative as possible of the entire lot of paper. Not less than 2 nor
.more"than 5% of the total number of" units" comprising. a shipment of paper
shall be sampled, nor less than 5'nor more than 20 sets of sample sheets shall be
taken in all. The units shall be cases, frames, or bundles; where a shipment
consists of a large number of packages; or the units shall be fractions of the
packages, where only a few constitute the shipment, such fractions or units
comprising not less than 1,000 sheets each.
Sets of sample sheets from sh!Jet-cut paper shall each consist of 10 consecutive sheets; except for lots containing fewer than 10 units, when each 'set shall
consist of 5 consecutive sheets; provided that if there are reasons to believe
that the paper was cut and stacked in single or any definite number of sheets
together, each sample set shall also consist of that number. The sheets shall
be taken from a point or points over 71 inch from the top or bottom of each
frame, case or bundle.
,
From rolls a single sheet shall be taken across the first unharmed layer;"
except when the shipment comprises less than 5 rolls, when the sheets may be
taken across the first two unharmed layers.
The sample sheets shall be trimmed with their edges exactly parallel to
the machine and cross directions of the paper.
The sample sheets comprising each set shall be consecutively numbered
and sufficient specimen sheets bearing consecutive numbers shall be taken for
testing purposes, one- from each set in rotation.
Procedure for Resampling.-In case of necessity for resampling a lot of
paper, the samples shall be taken as described abov,e except that they shall,
if possible, be drawn from units different from those previously s'ampled.
19 T 400 m-36.
Official Standard-July, 1926. Revised-Nov. 22, 1934. Revised
-Sept. 15, 1935. Corrected-8ept. 15, 1936.

1900

PAPER AND PAPER-MAKING MATERIALS

J

CONDITIONING PAPER FOR TESTING 20

I
NOTE.-This method is the official standard of the Technical Association. Many
laboratories, however test at different temperatures and humidities.. This is allowable
provided the conditions employed are stated in the report. Vnless such a statemeit is
made it will be assumed that the conditions stated in the official standard prevail.
.
~
Relative Humidity and Temperature.-Whenever re~uired in a_ t~st method,
the paper sample shall be conditioned and tested in an atmosphere maintained
at 65% relative humidity and 21C. (70 F.) temperature. A tolerance is
permissible (unless for very precise work) of plus or minus 2 in the percentage
relative humidity (63 to 67%) and of plus or minus 3 C. (5.5 F.) in temperature.
Conditioning.-Each test specimen of the paper sample shall be so suspended that the conditioning atmosphere will have free access to all surfaces.
Means shall be provided for so circulating the air of the conditioning and testing
chamber that its humidity and temperature will be uniformly maintained.
The conditioning time shall be sufficient for the moisture content of the specimen to attain equilibrium with the conditioning atmosphere, this to be determined by conditioning to cpnstant weight, by very accurate weighings at
intervals of not less than one-half hour:
For work of such precision that the hysteresis in the equilibrium moisture
content may lead to an appreciable error, the moisture content equilibrium
under standard conditions shall be approached from a drier state by first reducing, if necessary, the moisture content to less than half the value under standard
conditions and then conditioning under standard conditions. For that purpose
the samples may be dried in a desiccator or other convenient means, provided
the temperature does not exceed 60 C. (140 F.).
NOTE.-With good circulation a conditioning period of .4 hours is usually sufficient.
for papers of ordinary weight and composition, but some hard-sized papers, boards and
water-re~istant specialties may requil'e 24 hours or longer.
Determination of Humidity and Temperature.-1:jhe relative humidity of
the conditioning atmosphere shall be determined:by means of either (1) a sling
psychrometer, or (2) a stationary type of psychrometer having the air circulated over the thermometer bulbs mechanically. In both cases the circulation
of air round the thermometer bulbs shall be at the rate of not less than 3 meters
(10 feet) per second and the exposure not less than 60 secouds before the readings are taken. When the sling type is used, care shall be taken> to make the
readings, especially of the wet bulb, as quickly as possible after bringing it to
rest. For a relative humidity of 65% with temperature about 70 F., at 30
inches atmospheric pressure, the wet bulb should be depressed 7.7 0 F. from
the dry bulb reading.
,
The thermo!lleters used for determining humidity and temper~ture shall
be accurately calibrated by comparison with certified standard thermometers
and any corrections found necessary applied to the readings.
NOTE.-It is recommended that thermometers approaching the following specifications be used: Range 0--50 0 C. (32-122 0 F.); graduation, 0.20 C. (0.5~F.). They should
20 T 402 m-36.
Official Standard-July, 1926. .Re.ifrse-d-Nov. 22, 1934. Corrected-8ept. 15, 1936.
-

9g~6
PAPER AND PAPER-MAKING MATERIALS

1901

be matched to within :li0 F. throughout the range used. Under ordinary conditions an
error of 1 in the percentage relative humidity corresponds to an error of approximately
:li0 F. in the wet bulb depression. Except for high altitudes, the barometric correction
is negligible.

PARAFFIN IN PAPER 21
Apparatus.-An extraction apparatus with a reflux 'condenser is necessary
for this test. The balance used for weighing shall be sensitive to 0.1 mg.
Test Specimen.-The test specimen shall consist of not less than 5 grams
of air-dry paper obtained by cutting small strips from different portions of the.
test sample in such a way as to be representative of it. The strips shall be cut
into small pieces.about 6 mm. (0.25 inches) square. It is recommended that
the test specimen be ground or shredded.
.
Procedure.-Completely extract the paraffin from the paper with chemically pure carbon tetrachloride, in the extraction apparatus. Evaporate the
extract to dryness, add 25 ml. of 0.5 N alcoholic potash, and again evaporate
to dryness. Take up the dry residue with ether and water, and transfer to a
separatory funnel. The volume of ether in the funnel should be 25 ml. and
that of the water, 150 ml. The water should contain a small amount of sodium
chloride to prevent emulsification. Shake the contents of the flask. thoroughly,
allow separation to take place, and draw off the water into a second separatory
funnel. Repeat the treatment, one or more times as may be found necessary,
with a fresh 25-ml. portion of ether. Combine the ether extracts, and wash
with fresh 100-ml. portions of water until the separated liquids are perfectly
clear. An addition of a strong solution of sodium chloride to the ether extract
may be necessary to thoroughly clarify it. Transfer the ether extract to a
weighed evaporating dish, evaporate to dryness, dry for exactly one hour at
100 C. and weigh the dried residue to an accuracy of 0.1 mg. Not less than
two determinations shall be made and the average of the results shall be computed. The percentage results of duplicate determinations of paraffin shall
agree within 0.2.
Report.-The paraffin content shall be expressed as a percenta~e of the
air-dry paper to the nearest 0.1.
ACTIVE SULFUR IN PAPER 22
Apparatus.-The'special apparatus required for this determination consists
of a 500-ml. round-bottomed distillation flask having a neck about 5 cm. (2
inches) long and 25 cm. (1 inch) in diameter, and a tu.be 10 cm. (4 inches)
long so connected to the mouth of the flask that all gases generated in the
flask will pass through a filter paper clamped between the mouth Qf the flask
and the end of the-tube, and then out through the tube.
The balance used for weighing shalL be sensitive to 1 mg. or less.
Test Specimen,-The specimen shall consist of 0.250 gram of air-dry paper,
either ground or thoroughly disintegrated by shaking in water.
21 T 405 m-36.
Official Standard-July, 1926. Revised-Jan., 1930. CorrectedSept. 15, 1936.
22 T 406 m-36.
Official Standard-July, 1926. Corrected-8ept. 15, 1936.

1902

PAPER AND PAPER-MAKING MATERIALS

Reagents.-(a) Copper sulfate solution .made by dissolving 0.786 gJ;am of

C.P. crystals of CuS045H 20 in water and diluting toll liter. 10 m!. of this
solution contain 0.002 gram of Cu.
(b) Small pieces of C.P. stick zinc, free from sulfurl and arsenic, activated
as follows: Cover the zinc with the CUS04 solution, using 10 m!. for each gram
of Zn. After allowing to stand a few minutes for deposition of the Cu, wash
the Zn with water until free from ZnS04. The Zn can' be reactivated several
times.
(c) Concentrated C.P. hydrochloric or phosphoric acid, free from S and As.
(d) C.P. lead acetate, 10% solution.
(e) C.P. sodium thiosulfate, 0.001 % solution.
(f) Sulfur-free surgical absorbent cotton prepared by bQjling in a dilute
solution of NaOH and washing thoroughly with water.
Procedure.-Transfer the weighed specimen to the distillation flask with
20 m!. of distilled water and add approximately 2 grams of activated Zn cut
in small pieces, and 10 m!. of conc. HCI or H aP0 4 Insert in the neck of the
flask a loose wad of surgical absorbent cotton about 4 cm. (17'2 inches) long.'
Clamp between the mouth of the flask and the tube a hardened filter paper
(such as Schleicher and Schulls' No. 575) which has been freshly moistened
with 10% Pb-acetate solution. Place another wad of absorbent cotton in the
tube above the filter paper. Prepare flasks under duplicate conditions, oontaining in place of the paper under test, pure, sulfur-free cotton, and measured
amounts of N a-thiosulfate solution. Set all the flasks in a steam bath for
1 hour, frequently agitating the contents. Remove the filter papers and allow
them to air-dry. Determine the percentage of sulfur present by comparing
the depth of the color developed by the S from the specimen with that developed by the known amounts of S evolved from the thiosulfate solutions. .
As 0.000001 gram of S will give a distinct stain, care must be taken to
protect the paper under test from contamination. It must be protected from
atmospheric fumes and should not be handled with the bare hands.
Report.-The amount of sulfur shall be reported Jas a percentage of the
air-dry paper.
.
.
.
NOTE.-In respect to the tarnishing effect of sulfur"a paper containing not more than
0.0008% of S is required for wrapping silverware.
Additional Information

Sutermeister, E.: "Chemistry ~f Pulp and Paper Making," p. 422, Joh; Wuey,New York.
AMOUNT OF COATING OF MINERAL COATED PAPER23
This method is suitable for the ordinary types of mineral coatings. It may
not be effective for exceptional special types, such as where' lacquers or other
materials are used to impart a high degree of water 'resistance.
Reagent.-Enzyme Solution.-Aqueous solution of approximately 1.5
grams of enzyme and 25 ml. of 0.1 N NaOH per liter.
The enzyme recommended by Sutermeister and Porter (1) is trypsin,. but
23 T 407 m.
Official Standard-July, 1926. Tentati~Revjs(on-Jan. 15, 1934.
Revised-April 15, 1935.
- .

PAPER AND PAPER-MAKING MATERIALS

1903

some of the mixtures of enzymes used commercially for desizing cotton and
degumming silk have been found to be more rapid in action, less expensive,
and more stable. The preparation Degoma D. L. made by Rohm & Haas
Co., 222 West Washington Square, Philadelphia, Pa., is particularly suitable.
Test Specimen.-The specimen for test shall consist of not less than 25
square inches.
Procedure.-Obtain the air-dry weight of the specimen. Place it in a
suitable container in such a way that the enzyme solution will have free acctss
to the entire surface of the coating. It may be laid in a fiat-bottomed tray
or rolled in cylindrical form and placed in a burette. Pour over the specimen
sufficient enzyme solution to cover it completely. Allow to stand for at least /
1 hour at 50 C., then place the paper on a pane of glass and brush off the
coating with a camel's hair brush, taking care not to dislodge the paper fibers.
Additional heating, and in some cases additional enzyme solution, may be found
necessary for very resistant coatings.
Mter the coating is entirely removed, stand the glass pane at a slight angle
and wash the paper on both sides by means of a wash bottle, holding the paper
on the glass by one corner. Dry the specimen and obtain the air-dry weight.
The difference between this weight and the original weight of the specimen is
the amount of coating material. Not less than two determinations shall be
made and the average of the results shall be reported.
Report.:_The amount of mineral coating shall be reported (1) as a percentage of the decoated air-dry paper and (2) as pounds per 500 sheets, 25X40
inches in size. The weight of the decoated air-dry paper, on this same weight
basis, shall also be included in the. report.
Precision.-Owing to the variable nature of coating materials, the precision
is variable. With most ordinary types of coatings, the percentage of coating
material found should be correct to within 0.5 to 2.
Literature Cited
1. Sutermeister and Porter, Tech. 4ssoc. Papers, XIII, No.1; 205 (1930).

RESIN IN PAPER 24
Apparatus.-A suitable extraction apparatus, such as the Soxhlet or the
Underwriters', is required for this determination. In addition, if glue or other
nitrogenous sizing agents are present, two 300-ml. separatory funnels are
necessary.
Reagent.-Acidulated Alcohol.-To 190 ml. of 95% alcohol add 1 ml. of
glacial acetic acid and 19 ml. of water, and mix.
Test Specimen.-The test specimen shall consist of not less than 5 grams
of paper obtained by cutting strips approximat!!ly Yz inch wide froni "the sample
in such a way as to be representative of it. It is recommended that the specimens be shredded or ground.
Procedure.-Obtain the air-dry weight of the specimen. If the paper is
not shredded or ground, fold th~ strips into numerous crosswise folds and place
24 T 408 m-3B.
Tentative Standard-Feb. 27,1917. Official Standard-July, 1926.
Corrected-Sept. 15, 1936.

1904

PAPER AND PAPER-MAKING MATERIALS

,
'

them in the siphon cup of the extractor. Extract with acidulated alcohol,
siphoning at least 12 times or as !llany more as may' be necessary until the
solvent siphons over colorless.
If nitrogenous sizing agents are present, they must be separated from the
resin as follows, this not being necessary in absence ofi such materials. Wash
the alcoholic extract of resin, which may contain foreign material, into a beaker
and evaporate to a few m!. on a steam bath. Cool, take up in about 25 m!.
of\ther, transfer to a 300-m!. separatory funnel containing about 150 m!. of
water to which has been added a small quantity. of sodium chloride to prevent
emulsification, shake thoroughly and allow to separate. Draw off the water
into a second separatory funnel and repeat the treatment with a fresh 25-ml.
portion of ether. Combine the ether extracts which contain the resin and any
other ether-soluble material, and wash twice, or until the ether layer is perfectly clear and the line between the ether and the water is sharp and distinct,
with 100-ml. portions of water to remove salts and foreign matter. Should
any glue, which may be extracted from the paper, interfere by emulsifying with
the ether, it may be readily removed by adding a strong solution of sodium
chloride to the combined ether extracts. Sha,ke thoroughly and draw off the
aqueous solution, repeating if necessary before washing with water.
Transfer the alcoholic or the ether extract to a weighed evaporating dish,
evaporate, to dryness, dry the residue exactly 1 hour at 100 C., and weigh to
an accuracy of 0.1 mg. Not less than t~o determinations shall be made, and
the average of the results computed. The percentage results of duplicate
determinations of resin should agree within 0.2.
Report.-The resin content shall be expressed as a percentage of the airdry paper, to the nearest 0.1 per cent.
Additional Information

"Quantitative Determination of Rosin in Paper."-C. F. Sammet, Ind. & Eng. Chern.,

5,732 (1913); Paper 13, No.1, 17 (1913).
NOTE.-8ee also T 416 m, Rosin in Paper (Qualitative).,

MOISTURE IN PAVER 25
Apparatus.-.The special apparatus required for this determination is an
air-tight container in which the specimen is dried and weighed. For the
minimum size specimen designated, a weighing bottle approximately 65 mm.
(2.56 inches) in height and 45 mm. (1.77 inches) in diameter is'suitable. For
larger specimens, proportionately larger containers should be use_d.
The oven used to dry the paper shall be equipped witb. means for ensuring
adequate temperature control and air circulation. and preferably equipped with
means of drying the air entering the, oven.
Test Specimen.-The specimen for test shall consist of not less than 2
grams of paper obtained by cutting small strips from different portions of the
test sample in such a way as to be representative of it.
NOTE.-When the paper under test is not, in 'moisture equilibrium with the surrounding atmosphere, take care to minimize the time of exposure of the test sample to
the a.tJ?osphere as much as possible, gain or loss of mois~uUl being very"rapid under this
condItion.
.
_
25 T 412 m.
Official Standard-Dec., 1926. Corrected-May 15, 1935.

PAPER AND PAPER-MAKING MATERIALS

1905

Procedure.-Place the specimen in a weighed air-tight container, close the

container and obtain the weight of the specimen. Remove the cover of the
container and dry the paper in the container, in an oven having an adequate
circulation of air, at 100-1050 C. (212-221 0 F.), for one hour. Close the container in the oven, remove to a desiccator and cool in the desiccator to room
temperature. Weigh the container and paper and repeat the entire process
until the weight is constant. All weighings shall be made to an accuracy of
1 mg. The percentage results of duplicate determinations of moisture shall
agree within 0.2.
Report.-Report the amount of moisture as a percentage, (1) of the original
weight of the paper and (2) of the oven-dry paper, to the nearest 0.1.
ASH IN PAPER 26
Apparatus.-A crucible, such as platinum, alundum or porcelain, which
will not change in weight under the ignition conditions used and with a tightly
fitting lid; a balance sensitive to 1 mg.; and a desiccator, are necessary for this
determination. An electric muffle furnace is recommended for burning the
paper.
Test Specimen.-The test specimen shall consist of not less than 1 gram
of air-dry paper obtained by cutting small strips from different portions of the
sample in such a way as to be representative of it.
Procedure.-Weigh the specimen of paper accurately in the previously
weighed crucible and ignite completely. To avoid loss of small particles of
the specimen care must be taken to heat it slowly and to protect the contents
of the crucible at all times from strong drafts. When the paper is completely
burned, as indicated by absence of black particles, remove the crucible to the
desiccator, cover and allow to remain until its temperature has reached equilibrium with that of the surrounding atmosphere. Then weigh the crucible
and contents and repeat the ignition and weighing until the weight is constant.
All weighings shall be made to an accuracy of 1 mg.
Precision.-The percentage results of duplicate determinations shall agree
within 0.2.
Report.-The amount of ash shall be reported as a percentage of the airdry paper to the nearest 0.1.
See T 211 m Ash in Pulp.

Additional Information

PROTEINACEOUS NITROGEN IN PAPER27

Apparatus.-A Kjeldahl digestion and distillation apparatus is required for
this determination. A 500-m!. Kjeldahl flask is a suitable size.
Test Specimen.-The specimen shall consist of from 3 to 5 grams of air-dry
paper, obtained by cutting small strips from different portions of the test
sample in such a way as to be representative of'it. The strips shall be cut
into pieces approximately 6 mm. (0.25 inch) square.
26

27

T 413 ffi.
T 418 ffi.

Official Standard-Dec., 1926. Corrected-April 15, 1935.

Official Standard-Dec., 1926. Corrected-May 15, 1935,

1906

PAPER AND PAPER-MAKING MATERIA.LS

Procedure.-Gunning Method.-Place .the specimen, weighed to an accuracy of 1 mg., in the Kjeldahl flask and add 10 grAms of po~dered anhydrous sodium sulfate, a small crystal of copper sulfate (about 0.2 gram) and
25 ml. of conc. sulfuric acid. Heat the flask gently uhtil frothi~g has ceased
and then digest with increasing temperature until oxi(1ation 'is complete, i.e.,
for a short time after the mixture becomes clear and colorless, or nearly so.
Cool and dilute with about 200 ml. of water. Add about 2 ml. of liquid petrolatum to prevent foaming and about 2 grams of 30-mesh granulated zinc to
prevent bumping during the distillation. Add a saturated solution of NaOH
to the contents of the flask in such amount (usually 75 ml.) that there is an
excess of 5 ml. present. This solution must be poured carefully down the
side of the flask so that it does not mix with the acid contents. The total
volume of the solution should be about 400 ml.
Immediately connect the flask to a condenser ha'ving its delivery tube just
beneath the surface of a known amount of 0.1 N sulfuric acid diluted to 100 ml.
(30 ml. of 0.1 N acid is usually sufficient). Mix the contents of the Kjeldahl
flask by shaking the flask, then heat it gradually and distill the contents for
about 45 minutes, taking care to avoid spurting. The total volume of the
distillate shQuld be about 200 ml.
Titrate the contents of the receiver flasK with 0.1 N alkali, using methyl
red indicator. The difference between the humber of m!. of 0'.1 N alkali used
and the ml. of 0.1 N acid added to the r_eceiver flask is the number of ml. of
0.1 N acid equivalent to the nitrogen present. This number multiplied by
0.0014 is the nitrogen found.
The percentage results of duplicate determinations of nitrogen shall agree
within 0.02. A blank determination shall be made on all reagents used ~nd
any nitrogen found subtracted. Report.-The amount of N shall be expressed as a percentage of the air-dry
paper to the nearest 0.01 %. If it is desired to report the .percentage of glue
or casein present, it shall be calculated by multiplying the percentage of N
I
/
found by 5.6 and 6.3 respectively.
NoTE.-As these factors vary with different kinds and grades of material, they should
be determined whenever the nitrogenous material is" available, and whenever possible
the N in the paper before addition of the nitrogenous material should be determined and
subtracted from the total N found.
STARCH IN PAPER 28
QUALITATIVE

NOTE.-This was formerly printed as a separate method T 419 m.

A positive result obtained by the following m~thod shall be regarded as
conclusive evidence of the pre.gence of starch in paper.
Boil about 0.5 gram of paper for several minutes with 10 ml. of water.
Filter off the extract, cool 'it, and add one drop of approximately O.OOlN
iodine solution. A blue coloration indicates starch. If only a faint violet
28 T 419-420 m.
Official Standard-Dec., 1926.- ~vi.sed-'January, 1930. Corrected-March 15, 1934.

PAPER AND PAPER-MAKING MATERIALS

1907

coloration is obtained, this should be disregarded, as non-starch constituents

.
. of paper sometimes ~ive such reaction.
NOTE.-It is convenient to make an approximately 0.01 N stock solution of iodine
(0.13 gram I and 2.6 grams KI in 100 ml. water) and dilute a portion of this to a pale
yellow color (about 0.001 N) each time a test for starch is made.
QUANTITATIVE

NOTE.-This was formerly printed as a separate method T 420 m.

Apparatus.-A reflux condenser is necessary for this test. The balance
used for weighing shall be sensitive to 1 mg.
Reagents.-FoT Use of Fehling's Solution.-(a) Copper sulfate (Fehling's
Solution A)-To 4 liters of distilled water add 277.3 grams of crystallized
copper sulfate (CUS04, 5H 20), and shake the mixture at intervals until the
crystals have dissolved. The CUS04 should be of high purity and as free from
Fe as possible.
(b3 Alkaline Tartrate (Fehling's Solution B)-Dissolve 150 grams of sodium
hydroxide of good quality and 500 grams of potassium sodium tartrate in
1500 ml. of distilled water.
(c) Mixed Fehling's Solution-Prepare by careful addition of the copper
sulfate solution A to an equal volume of the alkaline tartrate solution B. It
may be kept ready mixed but should in this case be carefully protected from
light and air. It is, however, preferable to keep the solutions separate until
they are needed.
(d) Molybdophosphoric Solution-Dissolve 100 grams of sodium molybdate (43% Mo) and 75 ml. of syrupy phosphoric acid (85%) in a solution
containing 275 m!. of conc. sulfuric acid and 1750 m!. of distilled water.
(e) Potassium Permanganate Solution-Prepare a N/30 solution of KMn04
by dissolving 1.10 grams in distilled water and making up to 1 liter. Standardize against sodium oxalate.
FOT Use of Benedict's Solution.-The directions for preparing this solution
are described by Joslin (Diabetic Manual Paper 197) as follows:
Copper sulfate (C.P. crystals) ..................... . 14.20 grams
Anhydrous sodium carbonate ...................... . 100 grams
Sodium citrate (5)-2 H 20) ......................... . 200 grams
Potassium sulfocyanate ........................... . 125 grams
10 ml.
5% potassium ferro cyanide solution ............... .
Distilled water to make a volume of ................ . 1000 ml.
-

With the aid of heat dissolve the carbonate, citrate and sulfocyanate in enough
water to make about 800 m!. of the mixture. Dissolve the CUS04 (accurately weighed) separately in about 100 m!. of water and pour the solution
into the other liquid with constant stirring. Add the ferrocyanide solution,
cool and dilute to exactly 1 liter. Only the copper sulfate need be weighed
with exactness.
Test Specimen.-The specimen shall consist of not less than 5 grams of
oven-dry paper obtained by cutting small strips from different portions of the
test sample in such a way as to be representative of it. The strips shall be
cut into small pieces about 6 mm. (0.25 inch) square.

1908

PAPER AND PAPER-MAKING MATERIALS

Procedure.-Place the weighed specimen of paper in a 500-ml. roundbottomed flask, add 200 ml. of water and 5 ml. of gladal acetic acid. Connect
the flask to a reflux condenser and boil the contents vigorously for 1.5 hours.
Pour the contents of the flask through a' Buchner f~nnel and wash the pulp
on the plate with 50 m!. of hot water, using suction. Add 15 m!. of conc.
HCl to the filtrate and boil under reflux for 30 minutes. Neutralize the hot
solution with solid N a 2 CO g until effervescence ceases. Cool to room temperature -and adjust to 500 ml. in a volumetric flask. Determine the starch
in the hydrolized starch solution with either Fehling's solution or Benedict's
solution. The procedure for each of these solutions follows:
1. With Fehling's Solution: Draw off 100 m!. of this hydrolized starch
solution by means of a pipette into a 400-ml. beaker. Add 25 ml. of mixed
Fehling's solution and boil for 10-15 minutes. Filter through a small Buchner
funnel containing an asbestos mat. Wash the cuprous oxide with 200 ml. of
hot water. Transfer the Cu 2 0 with the asbestos mat back to the original
beaker and add 25 m!. of the molybdophosphoric solution. Add 25 ml. of
water and filter off the asbestos through the same Buchner funnel. Wash the
mat with 200 ml. of cold water, removing every trace of blue color. Titrate
the filtrate with N/30 KMn04 solution to a faint pink after the blue color has
been discharged.
1 m!. N/30 KMn04=0.001l gram starch.
/

2. With Benedict's Solution: To approximately 10 grams of Na 2CO g in a

150-ml. beaker, add 20 mI. of water and 5 m!. of Benedict's solution. Heat to
boiling and add the hydrolized starch solution from a burette in small portions.
Keep the solution in the beaker boiling and maintain its volume ~t about
25 mI. The end-point is when the Benedict's solution changes to colorless.
1 ml. Benedict's solution = 0.002 gram starch.
As the factor for starch varies with different kinds and grades, when the
starch used in the paper is available its exact fl1Ctor iban be determined by the
following procedure:
Dry the starch at 100-105 C. for 3 hours. Weigh exactly 0.05 gram and
boil with approximately 190 ml. of 4% HCI solution for 30 minutes. Neutralize with N a 2 CO g, cool, adjust to a definite volume, and titrate against
either of the solutions as,described. When it is desired to.express the content
of starch on an air-dry basis and the actual initial moisture 'content of the
starch is not known, the average moisture figure of 12% may be used in the
conversion calculation.
'
Report.-The amount of starch shall be reported as a percentage of the
oven-dry paper to the nearest 0.01 %.
Additional Information
This method is based on that of V. Voorhees and O. Kamm. 29 The original method
was revised in, respect to the Fehling solution (suggested by Miss H. U. Kiely) and the
Benedict solution (suggested by G. P. Genberg), to avoid.use of indicators and to obtain
a sharper end-point.
29

Paper, 24; 1091 (Aug. 27, 1919).

PAPER AND

PAPER~MAKING

MATERIALS

1909

MINERAL FILLER IN PAPER ao

Test Specimen.-The specimen for test shall consist of not less than 0.2
gram of ash obtai~ed from the paper by the official method for determining
the ash of paper (T 413 m).
Procedure for Qualitative Determination.-Warm the ash in a small quantity of dil. HCI. An effervescence indicates the presence of carbonate. If
there is an insoluble residue, decant the acid solution through a filter paper and
repeat the operation twice, finally transferring the residue. to the filter paper
and washing it thoroughly.
Add an excess of NH 4 0H to a portion of the acid solution. A white,'
gelatinous precipitate shows the presence of AI. Filter off any such precipitate and add a solution of NH 4-oxalate. A white granular precipitate indicates Ca. To another portion of the acid solution add a few drops df BaOl2
solution and warm. A white precipitate will form if a sulfate compound is
present.
Dry any residue from the acid treatment of the ash, burn off the filter
paper in a platinum crucible, and add to the contents of the crucible a fusion
mixture consisting of equal parts of N a and K carbonates. Fuse to a clear
mass, cool, transfer the fusion to a beaker and heat gently with water. When
the mass is thoroughly disintegrated, filter and wash any residue thoroughly
with hot water. To the filtrate add an excess of HCI, evaporate to dryness,
and treat the residue with warm dii. HCI. A white, transparent, flaky precipitate shows the presence of silica. Filter off such precipitate and to half of
the filtrate add a solution of BaCI 2, when a white precipitate will form if an
insoluble sulfate, such as BaS04, was present in the paper. To the other half
of the filtrate add an excess of NH 40H and boil. A white gelatinous precipitate indicates AI 2 0 a
Dissolve the water-insoluble portion of the fusion with dil. HCI. Add to
the solution an excess of NH 40H and boil. A white gelatinous precipitate
shows the presence of AI. Filter off any such precipitate and add NH 4carbonate to the filtrate. A white precipitate occurs if Ba is present. Filter
and add an excess of ammonium phosphate and stir thoroughly after cooling.
A white crystalline precipitate indicates Mg.
A considerable amount of Ca and sulfate in the ash soluble in Hel indicates the presence of calcium sulfate, CaS042H 2 0 (crown filler). The pres~
ence of Ba and sulfate in the acid-insoluble portion of the ash indicates BaS04
(blanc fixe or barytes). An alkaline ash containing Ca soluble in HCI but
with no sulfate present indicates the presence of CaCO a (whiting or chalk).
The presence of Ba in ash which is soluble in HCI with effervescence indicates
BaCO a Considerable amounts of silica and Al indicate aluminum silicate'
added as china clay. Considerable amounts of silica and Mg indicate magnesium silicate derived from talc or asbestine. A microscopical examination
of the ash assists in determining the variety of silicious minerals present.
Procedure for Quantitative Determination.-The procedure for determining the amounts of the various fillers present follows that given for their
qualitative determination. The methods used for ensuring quantitative separation and determination of the various constituents are those described in
30 T 422 m.
Official Standard-Sept., 1927. Corrected-May 15, 1935.

1910

PAPER AND PAPER-MAKING MATERIALS

standard textbooks dealing with mineral analysis. When the ashing of the
paper causes chemical changes in the filler, the actual amount of filler originally
present must be calculated. The more common examples of such changes are:
loss of combined water in clay and crown filler; lossl of CO 2 in carbonates;
I
reduction of sulfates to sulfides.
Additional Information

Griffin, R. C., "Technical Methods of Analysis." McGraw-Hill, New York.

TOTAL ACIDITY OF PAPER 31
Apparatus.-The special apparatus required for this test is a grinder which
will completely disintegrate the paper without heating or contaminating it
and a steam or oil bath which can be maintained at 100 C. The grinder
shall be a " Dr. Koerner" type or its equivalent.
The balance used for weighing shall be sensitive to 1 mg.
The glassware shall be acid- and alkali-resistant.
Reagents.-(a) 0.01 N sodium hydroxide solution, conveniently made by
diluting 100 m!. of 0.1 N NaOH to 1 liter ~ith freshly boiled and cooled distilled water.
(b) Phenolphthalein indicator solution.
Test Specimen.-The specimen for test shall be cut f.rom the test sample
in such a way as to be thoroughly representative of it, and shall be completely
disintegrated in the grinder.
Procedure.-Allow the ground specimen to come to moisture equilibrium
with the atmosphere of the balance case, and weigh 5-gram portions (to nearest
1 mg.) for the acid extractions and moisture determination at the same time.
Determine the moisture content on one portion by the standard TAPPI method.
Transfer another weighed portion to a 500-m!. Erlenmeyer flask and add 250
m!. of boiling water. In some cases, the fibers absorb water slowly and tend
to float on the surface of the water. This is lavoid~d by first adding small
portions of the water and shaking well until the fibers are thoroughly saturated. After the water is added, affix to the flask a stopper containing a
narrow glass tube about 30 inches in length, which serves as a condenser.
A soil-digestion flask which has a ground-glass stopper and condensing tube
in one piece, or a rubber stopper covered with metal foil, may be used. Place
the flask in a heating bath which will maintain the contents ofthe flask at
95 to 100' C. Heat at this temperature for one hour with occasional shaking.
At the end of this period pour the contents of the flask on a Buchner funnel
(without other filtering medium) and wash the fibers remaining in the flask
into the Buchner funnel with 10 m!. of water. Apply strong suction to the
fibers, then return them to the flask. Cool the extract rapidly and titrate as
soon as it reaches room temperature, with 0.01 N NaOH, using phenolphthalein indicator. Make two additional extractions and titrations exactly as in
31 T 428 m-36.
Tentative Standard-March, 1932. Correct~d-Sept. 15, 1936.
This method has been approved by the Paper Testing Dommi~tee of the Technical
Association of the Pulp and Paper Industry. Suggesti.QrtEt-and criticisms should be sent
to R. G. Macdonald, 122 East 42nd Street, New York, N: Y.

PAPER AND PAPER-MAKING MATERIALS

1911

the first instance. Make a blank titration on 250 ml. of the water heated for
1 hour in the same bath and with the same glassware used for the extractions.
Each test result used in calculating the acidity shall be the average of not less
than two determinations. The results of duplicate determinations of total
acidity shall agree within 0.01.
Report.-Total acidity shall be expressed as a percentage of the moisturefree paper in terms of sulfuric anhydride, 80a, to the nearest 0.01. It is calculated as follows:
Let T=ml. NaOH required to neutralize the 3 extracts;
t=3 times the m!. NaOH required to neutralize the blank;
N = normality of the NaOH solution; and
W = weight of test specimen less moisture.
Then,

(T-t) XNXO.04X100 -

- per cen

t SO

a.

Additional Information

This method is that described by S. Kohler and G. Hall in The Paper Industry, 7,
No.7; (Oct. 1925), with some modifications developed at the U. S. Bureau of Standards.
Studies of tbis method and other acidity methods are reported by Wehmhofl' and by
Wehmer in Technical Association Papers (TAPPI), May, 1930 and May, 1931.
The Dr. Koerner type of grinder is described in U. S. Bureau of Standards Research
Paper No. 295.
ALPHA CELLUL08E IN PAPER 32
NOTE.-8ee also T 203 m-36.
This method is for papers free from unbleached, ground-wood, or other
highly lignified fibers, such as the ordinary types of writing and printing papers
used for record purposes.
Apparatus.-The special apparat"\ls required for this test is:
A grinder which will completely disintegrate the paper without heating
or contaminating it. The grinder .shall be a "Dr. Koerner" type or its
equivalent.
A water bath which can be maintained -at 20 C.
A 7.5-cm. Buchner funnel with a nainsook cloth filter cut to fit the funnel,
the .cloth having been previously washed to remove stiffening material, and
dried against glass. Dry the cloth to constant weight at 105 C. and weigh
it in a weighing bottle before using it.
The balance used for weighing shall be sensitive to 1 mg.
The glassware used shall be of acid- and alkali-resistant grade.
Reagents.-Sodium Hydroxide Solution.-Dissolve pure NaOH in an equal
weight of water and allow the solution to stand in a stoppered glass vessel
until it becomes clear. Filter the supernatant solution through an asbestos
mat and dilute it with water free from CO 2 to a sp.gr. of 1.197 at 15 C. The
32 T 429 m-36.
Tentative Standard-March, 1932. Corrected-8ept. 15, 1936.
This method has been approved by the Paper Testing Committee of the Technical
Association of the Pulp and Paper Industry. Suggestions and criticisms should be sent
to R. G. Macdonald, 122 East 42nd Street, New York,_N. Y.

1912

PAPER AND PAPER-MAKING MATERIALS

solution will then contain 17.5% of NaOH by weight. lhis should be checked
by titration and should be adjustedif necessary within{the limits 17.4-17.6.
Acetic Acid.-Approximately a 20% solution in water.
Test Specimen.-The specimen for test shall be cut. from the test sample in
such a way as to be thoroughly representative of it, ahd shall be completely
disintegrated in the grinder.
Procedure.-Allow the specimen to come to moisture equilibrium with thE
atmosphere of the balance. Weigh (to the nearest 1 mg.) approximately
5 grams of the disintegrated paper. ,Weigh at the same time samples for
moisture and ash determinations, and for determinations of such other noncellulose components as may be found necessary for correction of the alpha
cellulose figures (total resins, starch, glue, etc.). These determinations (lh!iJI
be made by the T APPI standard methods. When mineral filler is present, the
weight of it shall be calculated from the ash content. The weight of mineral
coating shall be determined likewise, or according to the standard TAPPI
method.
Transfer the specimens for alpha cellulose. determination to a 400-ml.
beaker, adjust the beaker and contents in a water bath at 20 C., add 50 ml.
of the NaOH solution accurately measured with a pipette, and note the time
immediately after this is added. Macerate the mixture thoroughly with a
flattened glass rod, and add a second 50-ml. portion of the N aOH solution,
this addition being made after about 5 minutes have elapsed from the time
of the first addition. Macerate the mixtU're again and let it stand with occasional stirring for a total time of 30 minutes from the time of first addition of
NaOH. At the end of this period, dilute the mixture with 250 ml. of distilled
water having a temperature of 20 C. and filter it immediately with gentle
suction, using a 7.5-cm. Buchner funnel with a nainsook cloth filter cut to fit
the funnel, the cloth having been previously washed to remove stiffening
material, and dried against glass. Dry the cloth to constant weight at 105 C.
and weigh it in a weighing bottle before using it. Wash as uniformly as
possible with 125 ml. of water at exactly 20 C. and then return the filtrate to
the funnel and allow it to pass through the pad. Then wash the pad with
750 m!. of water having a temperature of exactly 20 C. Occasional packing
of the pad with the flattened glass rod is necessary during the washing as the
pad has a tendency to shrink and permit the wash water to flow around its
edge and not through it. Follow the water with 100 m!. of 20% acetic acid
at about 20 C. and let the pad stand in this condition for'5'miu!ltes, without
application of suction. Finally apply suction and wash the pad with 1 liter
of water at about 95 C.
Transfer the pad and cloth to a Petri dish or watch glass, dry at 70 C.
for 3 hours, transfer to a weighing bottle, dry to constant weight at 105 C.
and weigh in the closed weighing bottle. Determip.ations shall be 'made of
ash and other non-cellulosic components of-the residue as may be fO'und necessary, so that the weight of the residue may be corrected for these. components.
The weight (,)f the residue, after deduction of non-cellulosic constituents,
divided by the weight of the test specimen, likewise corrected, and the result
multiplied by 100, gives the percentage of alpha celh,Ilose in the total cellulosic
content. The weight of the test specimen shall alwtrYi:l be corrected for moisture and both the test specimen and the residue shall always be corrected

PAPER AND PAPER-MAKING MATERIALS

1913

for ash. Both the test specimen and the residue shall be corrected for other
non-cellulosic components whenever they are present in significant amounts.
Not less than two determinations shall be made and the average of the results
shall be reported. Duplicate determinations shall agree within 0.3.
Report.-The amount of alpha cellulose shall be reported as percentage
of the total cellulosic content to the first decimal place.
Additional Information

Cross and Bevan 33 mention the use of NaOH of mercerizing strength in isolating
"normal" cellulose. The method described here is based on that of Jentgen 3' and
modification of his method suggested by -Parsons 35 and Ross 36 and by Burton and'
Rasch. 37
COPPER NUMBER OF PAPER38
Apparatus.-This method is for papers free from unbleached, ground-wood,
or other highly lignified fibers, such as the ordinary types of writing and
printing papers used for record purposes.
The special apparatus required for this test is (1) a grinder which will
completely disintegrate the paper without heating or contaminating it and
(2) a steam or oil bath which can be maintained at 100 C. The grinder shall
be a " Dr. Koerner" type or its equivalent.
The balance used for weighing shall be sensitive to 1 mg.
The glassware used shall be acid- and alkali-resIstant.
Reagents.-Copper Sulfate Solution.-Dissolve 100 grams of copper sulfate
(CUS04' 5H 2 0) in water and dilute to 1 liter.
Carbonate-Bicarbonate Solution.-Dissolve 350 grams of sodium carbonate
(Na 2 C0 3lOH 2 0) and 50 grams of sodium bicarbonate (NaHCO s) in water and
dilute to 1 liter.
Molybdophosphoric Acid.-Dissolve 100 grams of sodium molybdate
(Na 2Mo0 4 2H 20) and 75 ml. of phosphoric acid (83%) in a mixture of 275 m!.
of concentrated sulfuric acid and 1.75 liters of water.
Sodium Carbonate Solution.-An approximately 5% solution of Na 2 CO a
in water.

Potassium Permanganate Solution.-0.05 N.

Test Specimen.-The specim~n for test shall be cut from the test sample in
such a way as to be thoroughly representative of it, and shall be completely
disintegrated in the grinder.
Procedure.-Allow the specimen to come to moisture equilibrium with the
atmosphere of the balance. Weigh about 1.5 grams (to nearest 1 mg.) of the
33

Cross and Bevan, "Researches of Cellulose," Vol. III, 1905-1910, p. 22.

3' Jentgen, Kunsto.ffe, 1, 165 (1911).

35 Parsons, Paper Trade Journal, 82, No.8, p. 211 (Technical Section, p. 105).
36

1926.

Ross, "Research Notes" (Canadian Pulp and Paper Association, Vol. 1, p. 105).

37 Burton and Rasch, Bureau of Standards Jour. of Research, R. P. No. 295, April,
1931. Includes an illustrated description of the "Dr. Koerner" grinder.
as T 430 m-36. Tentative Standard-March, 1932. Corrected-Sept. 15, 1936.
This method has been approved by the Paper Testing Committee of the Technical
Association of the Pulp and Paper Industry. Suggestions and criticisms should be sent
to R. G. Macdonald, 122 East 4200 Street, New York, N. Y.

1914

PAPER AND PAPER-MAKING MATERIALS

ground paper. Weigh at the same time, samples for moisture and ash deter~
minations, and for determinations of such other nQn~cellulose components
(total resins, starch, glue, etc.) as may be found necessiuy for correction of the
copper number. These determinations shall be made by the TAP PI standard
methods. When mineral filler is present, the 'Weight,' ~f it shall b,e calculated
from the ash content. The weight of mineral coatiJ?g shall' be determined
likewise, or according to the Tappi standard"method. :
Immediately before use add 5.0 m!. of solution A to 95 m!. of B. Bring
the mixture to a boil in 2 minutes, and pour it over 1.5 grams of the 'finely
ground sample in a 125~ml. Erlenmeyer flask. Stir well with a glass rod in
order to distribute the fibers and to remove air bubbles. Fit the flask with
'a loosely fitting glass bulb or stopper and submerge completely in a steam Hath ...
at atmospheric pressure. Occasionally fibers tend to flo;:tt to the surface, 'there~
fore the flask should be shaken from time to time to redistribute them. Remove the flask from the steam bath at the end of 3 hours. Filter on an ashless
filter paper in a 7.5-cm. Buchner funnel, using suction. Wash by flooding with
100 m!. of 5% Na 2 C0 3 solution at about 20 C. and then by flooding with
250 m!. or hot water (about 95 C.). Transfer the fibers and filter paper to a
small beaker, add 25 m!. of the molybdate solution and macerate well with a
flattened glass rod. Transfer to a Buchner funnel again and wash thoroughly
with cold water until the blue molybdenum color is removed from the fibers.
Dilute the filtrate with water to approximately 700 m!. .and titrate it with
0.05 N KMn04 to a faint pink.
/'
The copper number is defined as the number of grams of metallic copper
in the cuprous oxide reduced from the cupric hydroxide by 100 grams of the
paper fibers. This is calculated as follows:
.
C opper N urn b er=

6.357Xm!. KMnO,XN
W
'

where N is the normality of the KMnO, and W is thlJ weight in grams of the
test specimen after deduction of the weight of the :non-cellulosic materials.
Correction of the weight of the test specimen shall alw'ays be made for moisture
and ash. Correction for other non-cellulosic components shall be made when~
ever they are present in significant amounts. N~t less than two determinations shall be made and the average of the results shall be reported. Duplicate
"- .
.
determinations shall agree within 0.1.
Report.-The copper humber shi'tll be reported on the basis_of the total
cellulosic content to two decimal places.
Additional Informatio_n

This method is essentially the Braidy 39 modification of the original Schwalbe 40

method, with modifications for its adaption to paper proposed by Scribner ,and Brode,41
and by Burton and Rasch. 42
,
39 A thorough investigation of the Braidy method was made by Clibbens and Geake,
(J. Text. Inst~, 15, T 31, 1924).
40 Die Chemie der Cellulose, p. 625; 1912.
.
41 Technologic Paper No. 354, National Bureau of Sta)l9.,!trds.
I
42 Bureau of Standards Jour. of Research, R. P . .,.295, April, 1934.
Includes an
illustrated description of the "Dr. Koerner" type of grinder~ .

PAPER AND PAPER-MAKING MATERIALS

1915

ACID-SOLUBLE IRON IN PAPER 43

The procedure commonly used for this purpose is to extract the ash of the
paper with acid but investigation has shown that the procedure described
below is preferable because: (1) it requires less time; (2) in the ashing of the
paper, some of the soluble iron may be volatilized and some of the insoluble
iron may be made soluble. The removal of the acid-soluble iron is complete.
This is considered the portion of the iron present that is potentially chemically
reactive, as distinguished from insoluble or " fixed" iron such as might occur
as silicate or other complex compound in clay filler.
Reagents.-Amnionium SUlfocyanate Solution.-Dissolve approximately
76 grams of pure NH 4SCN in water and dilute to 1 liter.
Standard Iron Solution.-Dissolve 14.04 grams of pure ferrous ammonium
sulfate, FeS04' (NH 4 )2S04',!3H 20, and 4.07 grams of pure ammonium persulfate, (NH 4)2S20S, in water containing 1 m!. of conc. H 2S0 4. Mix well and
dilute to exactly 1 liter. Dilute 25 m!. of this solution to 1 liter. When 5
grams of paper are used, 1 ml. of this standard equals 10 parts of Fe per million
parts' of paper.
Test Specimen.-The specimen for test shall be cut from the sample in
such a way as to be thoroughly representative of it and shall be either cut in
small squares (about Pi-inch) or ground. Cutting the paper in small pieces
is generaJly sufficient, but grinding of very hard, resistant paper may be found
necessary.
Procedure.-Drench 5 grams of paper with 50 ml. of concentrated HCl.
Cold acid usually suffices but ho~ acid may be required if large particles of
iron are present. After the paper is thoroughly saturated with the acid, filter
on a 7.5-cm. Buchner funnel. Wash the residue with three 50-ml. portions
of hot water, applying suction after each addition. Add a few drops of concentrated HN0 3 to the combined filtrates, make ammoniacal with concentrated NH 40H, and boil until the odor of NH3 is nearly but not quite gone.
Filter through a ;filter paper and wash the residue on the paper thoroughly
with hot water. Place the funnel containing the filter paper and Fe(OH)a in
. the neck of a 50-ml. volumetric flask, and add two successive 5-ml. portions
of dil. HCI (1: 3), with intermediate and final washing with water. Dilute
the filtrate to 150 ml. Measure 10 cc. of this solution into a 50-ml. Nessler
tube and add enough 0.05 N (or 0.1 N) KMn04 solution to com.pletely oxidize
the iron, or until there is a faint pink color to the solution. Add 10 m!. of the
NH 4SCN solution, dilute to the mark, mix, and compare immediately with
standards made in the-same way with the standard iron solution.
The standard iron comparates shall be prepared freshly for each comparison, as they fade rapidly. A blank determination on the reagents shall be
run and a correction applied if necess.ary. At least two determinations shall
be made and the average of the results shall be reported. Duplicate determinations should agree within 5 parts per million.
Report.-The results shall be reported as parts of acid-soluble iron per
million parts of paper by weight.
43

T 434 m. Tentative Standard-Jan. 15, 1934. Official Standard-April 15, 1935.

1916

PAPER AND PAPER-MAKING MATERIALS

Additional Information

American Public Health Association, "Standard Methods df Water Analysis," 1933.

(The modified procedure for paper was suggested in a communication from A. E.
Kimberly, Bureau of Standards.)
,

HYDROGEN ION CONCENTRATION (pH) OF PAPER EXTRACTS

44

Apparatus.-Grinder.-The grinder shall be capable of completely disintegrating the paper without heating or contaminating it. The type described by Burton and Rasch (1), and known as the Koerner type, is'recommended.
Glassware.-The glassware used shall be resistant to acids and alkalies. '
Pyrex glass or its equivalent is recommended.
Colorimetric and Electrometric pH Apparatus.-=-Any of the usual forms of
colorimetric apparatus employing indicators and glass comparison tubes is
suitable. The apparatus recommended for the eledrometric deter,mination is
illustrated in Fig. 2.72. In addition to this there is required a potentiometer
and galvanometer similar to the Leeds and Northrup student or precision
Wires to Potentiometer

Kel (sol n )------if--~

FIG.

272.-Electrometric pH Apparatus.

44 T 435 m.
Tentative Standard---8ept. 6, 1934. This -method -bas been approved
by the Paper Testing Committee as a tentative standM'd:- Send comments to R. G.
Macdonald, 122 E. 42nd Street, New York, N. Y.

PAPER AND PAPER-MAKING MATERIALS

1917

instruments, for measuring the voltage. In the figure, A is a saturated calomel

electrode. A platinum wire sealed through the bottom serves to connect the
mercury in the tube electrically with that in the outside arm. The electrode
is prepared by placing pure mercury in the tube to a depth of about 1 cm.,
covering this with a 2-mm. layer of calomel (preferably electrolytically prepared), adding pure KCl crystals to a depth of about 1 cm., and filling the tube
with a solution of KCl saturated at the temperature at which the electrode is
to be used. The funnel C serves as a reservoir for this solution. B is a widemouthed test-tube which contains the solution under examination. It is fitted
with a stopper bearing a glass tube D which connects this vessel with the
calomel electrode . The lower end of D is drawn out as a capillary and bent
up as shown, and the entire tube is filled with saturated KCl solution from C.
The stopper also contains a narrow glass tube E, which serves as an air vent,
and an electrode F. The electrode illustrated may be used as a quinhydrone
electrode. In such case, the solution being tested is saturated with quinhydrone by adding approximately 0.5 gram of the solid of as high purity as
possible and stirring.
Tbis electrode is the simplest one to use and is satisfactory if the pH of
the extract is less than 8, and if the buffer concentration of the extract is not
too small. If the pH is greater than 8, some .other electrode must be used,
such as the hydrogen or antimony electrode. The~e electrodes, however, do
not function well at low buffer concentrations and are" poisoned" by certain
materials which may be present in the extract. The most universally, useful
electrode is the glass electrode which can be used through the ,entire range of
pH up to about pH 11 and which is not sensitive to "poisons." The use of
the glass electrode, however, requires an especially sensitive galvanometer (3)
(4) (5). It is recommended that the glass electrode be used wherever the
equipment is available. In all cases the apparatus used is the same except the
electrode F, which must be selected to suit the type of extract being tested.
Test Specimen.-The test specimen shall be cut from the test sample in such
a 'way as to be thoroughly representative of it. It shall be completely disintegrated in the grinder.
Procedure~-Extraction.-Weigh 5 grams of air-dry paper to the nearest
0.1 gram. For the extraction, transfer the specimen to a 500-ml. flask and
add 250 ml. of boiling distilled water. The water should be of high purity,
preferably double or triple distilled. The quality of the water is particularly
important in the determination of the pH of papers having low acidity, such
as unsized papers. In such cases the water should have a pH between 6.6
and 7.0. If the fibers tend to float on the surface of the water, add water in
small portions, stir until the fibers are thoroughly wetted, then add the remainder of the water. After the water is added, affix .to the flask a stopper
containing a glass tube about 8 to 10 mm. in diameter and 75 cm. in length,
which serves as a condenser. A soil-digestion flask which has a ground-glass
stopper and condensing tube in one piece, or a rubber stopper covered with
clean metal foil, may be used. Place the flask in a heating bath which will
maintain the contents of the flask at 95 to 100 C. without boiling the water.
Heat at this temperature for 1 hour with occasional shaking. Pour the contents of the flask on a Buchner funnel without other filtering medium, and

1918

PAPER AND PAPER-MAKING MATERIALS

draw the solution through the fiber mat with strong ~uction. Disconnect the
suction just before the filtration is completed. As a few fibers may go through
the funnel with the first portion of solution, return the filtrate to the flask
and again pass it through the mat. Pour the filtrate into a clean flask, close
it tightly with a clean rubber stopper, and set aside to Icool to room temperature. The extract must be protected from the air as much as possible at all
I
times to avoid CO 2 contamination.
pH Measurement.-(a) Colorimetric: In using the colorimetric method care
must be taken that the indicators are adjusted in pH, either at the midpoint
of their range, or at frequent intervals (e.g. 0.2 pH) throughout their usual
range for more accurate work. This is necessary because paper extracts .are
usually not well buffered, ana the addition of an indicator may change the'
pH of the extract sufficiently to cause an appreciable error unless the pH of
the indicator is near that of the extract. The Acree-Fawcett isohydric indicator method (6) (7) is recommended. This procedure consists of locating the
approximate pH of the unknown solution by testing with an indicator solution
adjusted at the midpoint of its range and then making a further precision
determination with indicator solution adjusted at the approximate pH. This
may be most easily accomplished by using solutions of indicator adjusted at
intervals of 0.2 pH throughout the effective range of the indicator . The
adjustment is made by comparing the color bf 0.001 M indicator solution with
0.05 M buffer-color standards while adding to it O.Ol N HCI or NaOH as
required to match the color standard having the desired pH. Following is the
suggested procedure:
(1) Add 0.2 ml. of midpoint indicator to 10 ml. of extract in a test tube.
(2) Compare with 0.05 M buffer color standards and obtain the approximate pH.
(3) Repeat (1) by using indicator solution adjusted at the approximate pH
value obtained in (2).
(4) Compare the color standards and read to neare,flt 0.1 pH.
This method is simple and convenient, and checks the electrometric method
very closely within the limits of its accuracy. The average of the results of
at least two determinations shall be reported. 'l;'he results should agree within
0.2 pH.
.
\
(b) Electrometric: Referring to Fig. 272, flush the tube D by opening the
stopcock on C momentar.i1y with B removed. Fill the'vessel B with the
extract to the depth indicated and fit it to the stopper as shown. Run a small
amount (about 0.1 ml.) of the saturated KCI solution out through-the capillary
opening in D. Connect the wires from A and F to a potentiometer and measure the voltage. Take readings until the voltage is constant to about 1 millivolt. for a period of sev~ral minu~es. Calculate }he pH from the voltage
readmgs, E, by the followmg equatIOns:
I
.
p

0.4526-E
.
0.0591 for the qumhydrone electrode.

E-0.2358
.
pH = 0.0591 for the hydroge~ ~eitrode.

PAPER AND PAPER-MAKING MATERIALS

1919

The voltage obtained with the antimony or glass electrode, depends upon
the individual electrodes used. The pH can be calculated from the voltage
by calibrating the electrodes against buffers of known pH. In some cases
these calibrations are furnished by the manufacturers of the electrodes. The
above equations are for a temperature of 25 C. but they may be used for
measurements made at ordinary room temperature within the degree of accuracy stated below, by changing the denominator plus or minus 1/298 of its
value, respectively, for each C. above or below 25. These equations are
valid only when the saturated calomel electrode is used as the reference electrode.
,
The average of the results of at least two determinations shall be reported.'
The results should agree within 0.1 pH.
Report.-Results shall be reported as the pH of the air-dry paper. They
shall be rounded off for the colorimetric method to the nearest 0.1, and for
the electrometric method to the nearest 0.05. The report shall state which of
these methods was used.
Literature Cited
For an illustrated description of a grinder suitable for disintegrating the paper see:
1. Burton and Rasch, U. S. Bur. of Standards Jour. of Research, 6, 605 (Research Paper
295) (1931).
For detailed description of suitable electrometric methods see:
2. Clark, W; M.: "The Determination of Hydrogen Ions,'~ Williams and Wilkens Co.,
Baltimore (1928).
3. Burton, Matheson and Acree, U. S. Bur. of Standards Jour. of Research, 12, 67
(Research Paper 436) (1934).
.
4. Leeds and Northrup Company, Philadelphia, Pa.
5. American Instrument Company, Washington, D. C.
For a description of a suitable colorimetric method see:
6. Fawcett and Acree, Jour. of Bacteriology, 17, 171 (1926).
7. Fawcett and Acree, Ind. Eng. Chern., Anal. Edition, 2, 78 (1930).
8. "Modern pH and Chj.orine Control," Bulletin of W. A. Taylor and Company,
Balti.more, Md.
A study of certain of the details of this method is reported in the following articles:
9. Wehmhoff, B. L.: "The Determination of thf pH Values and Total Acidity of
Paper," Tech. Assoc. Papers, 13, p. 231, May, 1930; 14, p. 387, May, 1931.
ARSENIC IN PAPER 45
Apparatus.-The modified Gutzeit apparatus for determination of arsenic
is shown in Fig~ 273.
Two sizes of the apparatus are desirable; No.1 for use when the paper
contains less than 0.01 mg. of arsenic trioxide (As 20 a) and No.2 for use where
the paper contains more than that amount. The dimensions of apparatus
No.1 are listed to the left of the sketch; those of apparatus No.2 are listed
to the right.
Reagents.-Concentrated Nitric ACid, arsenic-free,. having a sp.gr. of at
least 1.4.
.
Concentrated Sulfuric Acid, arsenic-free, having a sp.gr. of at least 1.8.
Zinc, -arsenic-free, in the form of sticks (Bureau of Standards Standard
45

1936.

T 436 m-36.

Tentative Standard-May 15, 1935. Official Standard-8ept. 15,

1920

PAPER AND PAPER-MAKING M'ATERIALS

Sample No. 43D is recommended). Break the sticks into pieces about 1 cm.
long, rinse with water and treat with HCI until the sJrface of the Zn is clean.
Then wash with water, dry and store in a clean, wide-mouthed, glass-stoppered
\ .
bottle.
Mixed Acid.-Dilute 1 volume of As-free H 2S0 4 with 4 volumes of water,
and to this add 10 grams of N aCl for each 100 ml. of ~olution.
Ferric Ammonium Alum Solution.- Dissolve 84 grams of the salt, Fe2
(NH 4)2(S04)424H 20, in water, add 10 ml. of the mixed acid and dilute to
1 liter.
Lead Acetate Solution (for wetting glass wool).-Add to a 5% solutiQn'of
lead acetate, Pb(C 2 H a0 2)23H 2 0, sufficient. acetic acid to clarify the solutio11l'
Dimensions of
Apparatus No.1

em. long x 4 m.
bore. Constricted 6 em,
from upper end

10

Dimensions of
Apparatus No.2

~rB

7.5

em. long x
mm. bo,e.
Constricted 12, em {rom upper end.

paper.

Dry lead acetate

paper.

oz. bottle (250 ee.)

FIG. 273.-Gutzeit Apparatus (modified).

Stannous Chloride Solution.-Dissolve 80 grams of SnCh in 100 ml. of

water containing 5 mI. of As-free, concentrated HCI.
Lead Acetate Test Paper (for removal of H2S)'iSoak large she,ets of filter
paper, qualitative grade, in a 5% solution of Pb-acetate, dry, and cut into
pieces 7X5 cm. for apparatus No.1 and llX9 cm. for apparatus No.2.
Mercuric Bromide Paper.-Dry several pieces of Whatman No. 40 filter
paper at 100-105 0 C. for 1 hour and store in a desiccator 9ver fused CaC1 2.
Cut the paper into strips 2.5 mm. wide and satura~ with a 1.5% soh.ltion of
HgBr2 in 95% ethyl alcohol. Dry in a desiccator for 10 minutes and use
promptly. ~o treated strips should be stored longer than 2 hours before use.

PAPER AND PAPER-MAKING MATERIALS

1921

Standard Arsenic Solution.- Dissolve 1.000 gram of resublimed arsenious

oxide (As 2 0 a) in 25 ml. of 20% As-free NaOH solution and neutralize with
dil. H 2S0 4, using methyl orange indicator. Dilute this solution to 1 liter with
freshly distilled water, to which 10 ml. of concentrated H 2S0 4 have been added
for each liter of solution. Mix well and again dilute 10 ml. of the solution to
1 liter with water containing H 2S0 4 Finally, dilute 100 ml. of the latter
solution to 1 liter with water containing acid. One mI. of the final solution
contains 0.001 mg. (or 1 microgntm) of As 2 0 a.
Standard Stains.-Prepare 2 sets of stains by running blank determinations
to which are added definite amounts of the standard As solution, the determinations being conducted exactly as described below under Procedure.
The first set of stains should, be those made by 0.001, 0.002, 0.004, 0.006,
0.008, 0.010, 0.015 and 0.020 mg. of As 20 a, respectively. Tliis set is convenient for use with apparatus No.1 and covers the range of As content of
most papers. The second set of stains should be those made by 0.02, 0.05,
0.1, 0.2, 0.3, 0.4 and 0.5 mg. of As 20 a, respectively. This set of stains is
conyenient for use with the apparatus No.2 and for larger samples of paper.
Preservatipn of Stains.-Dip the sensitized paper containing the As stain,
as soon as prepared, in molten paraffin, free from water, and mount on a sheet
of white paper, folded back to form a cylinder. The white paper cylinder
should be dried in an oven at 1050 C. before mounting the standard stains.
It is important that the standard stains and the white paper cylinder be kept
dry, or the stains will soon fade. Place the cylinder with the stains mounted
on it in a glass test tube containing phosphorous pentoxide covered with glass
wool, close the glass tube with a rubber stopper and store the tube in a dark
,place.
r .'
Test Specimen.-Cut 5 grams 46 0' the air-dry paper under test (weighed
to the nearest mg., if arsenic is to be reported as parts by weight, or measured
accurately (0.1 in.) for area if arsenic is to be reported as parts per unit
area) into pieces from Ys to 3i inch square, and place in a porcelain casserole.
Add 10--15 mI. of concentrated HNO a and cover the casserole by setting a
watch glass inside the rim, convex side up, with a thin glass rod between the
wall of the casserole and the edge of the watch glass. (This eliminates removing the watch glass for the addition of acids during digestion of the paper.)
Heat on a hot-water bath until vigorous action ceases. Cool and add 10 ml.
of concentrated H 2 S0 4 When vigorous action ceases, heat gently on the
water bath until the mixture liquefies and is dark brown or black in color.
Then alternately cool, add concentrated HN0 3 in 5-ml. portions, and heat on
the hot plate between additions of HNO a until the solution remains colorless
to yellow when evaporated to the evolution of white fumes of SOa. To completely remove all nitric or nitrous acid, evaporate to about 10 mI., cool, dilute
with 10--15 mI. of water and again evaporate to fumes of SOa. Cool, dilute
with water, cool again, and make up to 100 ml. with water in a volumetric flask.
For a blank correction on the acids used, place in a separate casserole the
total quantity of acids used to digest and oxidize the paper sample, and heat,
until all nitric and nitrous acids are driven off and white fumes of SOa are
evolved. Cool and make up to 100 ml. in a volumetric flask.
4i This amount is suitable for most papers but may have to be varied in special cases.

1922

PAPER AND PAPER-MAKING MATERIALS

Procedure.-Place 5 or 6 strips of Pb-acetate paper in the lower tube of

the absorption tower, a loose wad of glass wool mois;tened with Pb-acetate
solution in the upper tube, and a strip of HgBr2 paper in the topmost tube
of the absorption tower of the Gutzeit apparatus. Then pipette 20 ml. of the
prepared solution of the test specimen into the flask' ~of the apparatus, add
2 m!. of ferric ammonium alum solution, 0.5 JIll. of SnCl 2 solution, 50 m!. of
mixed acid and sufficient water (in this case 27.5 ml.) t6 make a total volume
of 100 ml. Add 8-10 grams of Zn to the solution in the flask, immediately
connect the absorption tower and set the flask' in a water bath maintained at
about 20 C. Let the evolution of gas continue for 1 hour, then remove ,the
test strip of bromide paper, dip it into molten paraffin and place it in a test
tube such as used for storing the standard stains. Prepare a stain from theblank solution in the same manner as used for the solution of the paper testspecimen. Compare both stains with the set of standards, and assign to each
" unknown" stain the value of the standard stain which most nearly m~tch~s
it in color. Deduct the value of the blank stain from that obtained from the
solution of the paper test-specimen. The difference is the amount of As, calculated as As 2 0 3, in the solution tested. Multiply by 5 to obtain the a~ount
of As 20 a in the paper specimen used.
Report.-The results shall be reported as either: (a) parts of arsenic trioxide
per million parts of paper, or (b) grains of metallic arsenic per square yard
of paper.
Calculations:
Let W = weight in grams ot specimen tested;
S = area in square inches of specimen tested; and
A = value in mg. of As 20 a of matching stain (corrected for blank and for
aliquot of solution used).
'
Then
(a)

(b)

1000A
A S20 3 per ml'11'IOn partsl f paper, an d
-w=parts

.
f
.
i
d f
-s-- = grams
0 arseruc per square yar 0 paper.

15.15A

Additional Information

"Standard Methods of Chemical Analysis," 5th Ed., 'vol. I, pp. 101-108.

Griffin, R. C., "Technical Methods of Analysis," 2nd Ed., pp. 52 ahd56. McGraw-Hill,
New York, N. Y. (1927).
-.._,Hillebrand & Lundell, "Applied Inorganic Analysis," pp. 218-220. }Viley & Sons,
New York, N. Y. (1929).
Official Methods of The Association of Official Agricultural Chemists, 3d Ed., p. 306.
Williams & Wilkins, Baltimore, Md. (1930). '

I
ZINC PIGMENTS IN PAPER47
The zinc pigments used as paper fillers comprise three main classes: (1) zinc
sulfide alone, (2) high-strength zinc pigment, containing about 50% ZnS and
50% BaS04, and (3) low-strength zinc pigment, containing l~ss than 30% ZnS
_!---

47 T 438 m-36. Tentative Standard-Aug. 1, 1936, - .

PAPER AND PAPER-MAKING MATERIALS

1923

(the balance BaS04), typified by lithopone. These pigments may be used

alone or in combination with other fillers, such as clay, etc.
Reagents.-Dilute HCl (5 : 9).-To 5 volumes of concentrated HC! add 9
volumes of water. This solution contains about 15% of HCI or 157 grams per
liter.
0.3% HC1.-Dilute 10 ml. of the above solution with 500 ml. of water.
Dilute H 2 S0 4 (1 : l).-To 1 volume of water add 1 volume of concentrated
H 2S0 4
Ammonium Chloride Solution.-To 500 ml. of water add 25 ml. of concentrated NH 40H and 10 ml. of concentrated HCl.
Potassium Ferricyanide Solution.-Dissolve 1 gram of KaFe(CN)6 in 100 ml. '
of water.
Diphenylamine Indicator.-To 100 ml. of concentrated H 2S0 4 add 1 gram
of diphenylamine.
Standard Potassium Ferrocyanide Solution.-To 1 liter of water add 6.46
grams of K 4Fe(CN)63H 2 0. One ml. of this solution should be equivalent
to (}'0015 gram of Zn. Standardize by titrating a solution of known Zn contimt as described under Procedure. If it is desired to calculate the results of
Zn determination on the paper in terms of one of the three main types of ZIi
pigment listed above, the solution should be standardized against that type of
pigment. In this case dissolve a weighed amount of the pigment in the diluted
HCI, boil to remove H 2S, adjust the acidity of the solution to between 0.5
and 1.0 normal, add 2 drops of diphenylamine indicator and titrate with the
standard ferro cyanide solution. If the pigments used contain Fe or AI, which
is unusual, these must be removed before titration, as they interfere. If
merely the Zn equivalent of the standard solution is desired, a solution of
chemically pure zinc should be used. (Bureau of Standards Sample 43 is
recommended; )
Test Specimen.-The test specimen shall be large enough to contain at
least 10 mg. of Zn (0.8 g. is usually sufficient), obtained by cutting pieces about
U inch square from the sample in such a way as to be thoroughly representative of it. The oven-dry weight of the specimen shall be determined according
to T APPI Standard Method T 412 m.
Procedure.-Place the weighed specimen in a 150-rol. beaker, add 20 rol. of
water and 8 rol. of the dilute HCI, cover the beaker with a watch glass and
boil gently for about 10 minutes. Wash down the sides of the beaker and also
the watch glass with 0.3% HCl. Pour the liquid through a filter (Whatman
No.4 or 41 or S & 8 589-black ribbon) into a 600-ml. beaker, keeping as
much of the pulp in the beaker as possible. Press the pulp against the side
of the beaker with a glass rod to remove most of the solution. Wash the pulp
twice with 0.3% HCI, each time squeezing the pulp as before, and wash the
filter twice with more of the s_!troe acid (see note 1). Add 2 ml. of dilute
H 280 4 to the filtrate and boil. Keep the beaker covered with a watch glass.
Add concentrated HNO a, a few drops at a time, until all organic matter is
destroyed. Then boil the solution, keeping the watch glass on until most of
the solution is evaporated; then remove the cover and heat on the hot plate
until copious white fumes of 80 a appear and all HNO a has been removed,
especially any condensed on the sides of the beaker.
I

1924

PAPER AND PAPER-MAKING MATERIALS

Cool the solution to approximately. room temp'erature and dilute with

500 m!. of water. Heat to boiling and make distinc~ly alkaline with concentrated NH 40H. (About 5 m!. are usually sufficient.) Filter the solution
through S & S No. 589 black ribbon filter paper (ot, Whatman No.4 or 41)
and wash the precipitate well with NH 4 01 solution., About 4 washings are
usually sufficient (see note 2). Boil the filtrate until free from NHs odor,
concentrate by evaporation to 50-75 mI., add 5 ro!. of concentrated HOI,
2 drops of ferricyanide solution, 2 drops of diphenylamine indicator solution,
and titrate with the standard ferrocyanide solution to the disappeara'nce of
the purple color (see note 3). The volume of standard ferrocyanide solution
used, multiplied by its Zn equivalent, gives the amount of zinc in the samI'ie.
Report.-The amount of zinc (or the appropriate zinc filler, if kp.own) shall
be reported as a percentage of the oven~dry paper to the neares~ 0.1.
NOTES.-l. When filtering the pulp from the solution after boiling, and in the subsequent washing, the pulp should be pressed as dryas possible. This requires some. skill
and becomes quite difficult with weak, short-fibered pulps. In such cases it may be
better to use a Gooch crucible or a Buchner funnel fitted with a filter paper.
2. Where the paper tested contains considerable clay, the initial boiling with Hel
may be shortened slightly. This keeps the solution of Fe and Al at a minimum. If
a voluminous precipitate of Al(OH). is obtained When the NH.oH is added, it is very
difficult to wash the precipitate free from Zn. In this case it is best to dissolve the precipitate in HCI and reprecipitate with NH 40H, combining the two filtrates. Two solutions and reprecipitations, with a final washing with NH 4CI solution, will remove all Zn
from the Fe and AI hydroxides.
3. The volume of the solution to be titrated should be between 50 and 75 ml. The
indicator is somewhat sensitive to acidity and salt concentration, but satisfactory_ results
will be obtained if the solution is prepared as directed. Excessive amounts of NH 4-salts,
especially when only small amounts of Zn are present, may render the eno-poi,nt somewhat indistinct. Therefore, if resolution and reprecipitation have been resorted to
because of large Fe and Al precipitates, care should be taken to keep the NH 40H and
HCI used as small as possible.
The solution should be stirred actively during the titration, and when only a sJI\!111
amount of Zn is present, the standard ferro cyanide solution sn.ould be added very slowly.
In some Qases it requires some time for the color to develop after addition of the first
portions of the ferrocyanide. A false end-point usually occurs, where the blue color
changes to purple, which sometimes takes a few secon~s to appear. After this point the
ferrocyanide is added drop by drop until the purple color disappears and does not return
after stirring the solution for 20 seconds. This color change at the end-point is ordinarily
sharp and,distinct, if there are more than 10 mg. of Zn present.

Additional Information
Zinc and titanium pigments are sometimes found together in paper fillers. Ti does
not interfere with the determination of Zn as detailed above.
When one of'the three main classes of Zn pigments listed above occurs alone in a
paper and the ash of the pulp is known, the zinc pigment can be determined by direct
ignition, taking care to maintain a plentiful supply of air during combustion, to prevent
possible reduction of the Zn salt to metallic Zn with subsequent volatilization, thereby
giving low results. Only 0,5 to 1 gram of the paper should'be used; the crucible containing it should be placed in the very dull red portion of the furnace and most of the carbon
burned off. The residual carbon can be burned off at 800-850 C. with the furnace door
ajar and no cover on the crucible.
The weight of the ash, corrected for the ash of the pulp, if this i'~ known, should be
multiplied by a factor which takes into account the loss iu,weight.due to conversion of
znS to ZnO on ignition. The usual factors for the three main types of zinc pigments
are as follows:

'"

PAPER AND PAPER-MAKING MATERIALS

1925

(1) ZnS-1.l9

(2) "High-Strength" zinc pigment-l.09

(3) II Low-Strength" zinc pigment -1.06
References
Kolthoff and Pearson, flTitrntion of ,Zinc with Potassium Ferrocyanide," Ind. Eng.

Chern., Anal. Ed. 4, 147 (1932).

Steele, F. A., "The Determination of Zinc Sulfide Pigments in Paper/' Paper Trade .J.,
97, No.9, 27 (1933): .

TITANIUM PIGMENTS IN PAPER"

This method deals with the analysis of paper ash as modified by the presence of titanium pigments, i.e. (1) pure titanium dioxide, (2) titanium dioxidebarium sulfate composite, and (3) titanium dioxidecalcium sulfate composite.

I---DI,;."";,,, Bur.tt.
6.5 tm,(21 in.Jlang

I
r

FIG.

274.

Apparatus.-The only special apparatus required is a Jones reductor, shown

in Fig. 274. This consists of a glass tube about 2 inches in diameter and 21
I

u T 439 m-36.

Tentative Standard-Aug. 1, 1936.

1926

PAPER AND PAPER-MAKING M!A.TERIALS

I'

inches long with a glass stopcock at the bottom. (An ungraduated dispe~sing
burette is satisfactory.) It is connected to a liter flas~ as shown in the figure.
A means is also provided for introducing CO 2 into th~ flask through 1J, bubble
trap. The tube is, filled for about one-third of its length with 1000 grams of
a mixture of equal amounts of lO-mesh and 20-mesh Igramilated zinc resting
on a pad of glass wool. The upper portion of the tub~ contains 500 grams of
stick Zn in pieces about 2 inches long. Before use the Zn is amalgamated
as follows: Place, 250 grams of the Zn in a heavy glass flask, add a solution of
11 grams of mercuric chloride in 100 ml. of concentrated HCI, swirl the mixture
for a few minutes, pour off the liquid, and wash the Zn free from acia ..with
water.
NOTE.-When not in use the Jones reductor should be kept filled with water a~
covered to exclude dust and other extraneous matter.
Reagents.-O.l N Potassium Permanganate Solution, standardized against
U. S. Bureau of Standards sodium oxalate.
Standard Ferric Alum Solution, of which 1 ml. is equivalent to about 0.005
gram of Ti0 2 ; standardized by reduction and titration with 0.1 N KMn04.
Dissolve 30 grams of Fe2(NH4h(S04) 4' 24H 20 in 300 ml. of water acidified
with 10 ml. of H 2S0 4 and add KMn04 solution drop by drop as long as the
pink color disappears. Finally dilute the solution to 1 liter. To standardize,
dilute 50 ml. of the solution, measured from a burette, to about 100 ml. with
dilute H 2S0 4 (1 : 20), reduce in the Jones reduct or and titrltte against the
0.1 N KMn04.
Thiocyanate Indicator.-A saturated solution of NH 4SCN.
Test Specimen.-Sufficient paper shall be ashed according to TAPPI
standard T 413 m to yield the required amount of ash. The oven-dry weight
of the paper shall be determined according to T APPI standard T 412 m.
Procedure.-Qualitative.-A qualitative analysis of the ash should be made,
unless the nature of the loading material is known, sinc,e it will greatly facilitate
the quantitative analysis. The procedures for: quaJitative analysis of fillers
and ash for the commoner loading materials; such as clay (hydrated aluminum
silicate), talc (hydrated magnesium silicate), barium sulfate, barium carbonate,
calcium sulfate, and calcium carbonate, are well known and are included in
the TAPPI Standards. To test for Ti place about 0.5 gram of the ash in a
250-mI. beaker, add 20 ml. of concentrated H 2S0 4 and 10grams of (NH 4)2S04
and boil for at least 5 minutes. An insoluble residue indicates-Si0 2 or siliceous
matter. Cool the solution, dilute to 100 ml. with water, heat_to boiling, let
settle ,and filter through double Whatman No. 42 paper. Test the filtrate
with H 20 2 A clear yellow or orange co;lor indicates the presence of Ti.
NOTE.-For a more detailed qualitative procedure, se,e Am. Chern. Soc, Monograph
No. 33 "Titanium," by W. M. Thornton.
'

Quantitative.-The determination of Ti involves its reduction and titration

with a ferric salt. The details of the procedure depend upon the other materials present. The general procedure for the determinatio.n of Ti0 2 is given
in Scheme 1, which should be used when none of th~:Other ash constituents are
to be determined. Where BaS04 also is to
deterniined, use Scheme 3.

be

PAPER AND PAPER-MAKING MATERIALS

1927

The procedure for a complete analysis of the ash in the absence of BaS04 is
given in Scheme 2.
Scheme 1 : Determination of Ti0 2 Only.-Weigh out accurately into a 250-ml.
beaker a sample of ash containing not more than 0.25 gram of Ti0 2 (Usually
0.5 gram is suitable.) Add 20--30 ml. or concentrated H 2S0 4 and 10 grams of
ammonium or sodium sulfate. Cover the beaker with a watch glass and carefully boil the contents over a flame for at least 15 minutes. Cool and dilute
with water to 125-150 ml. A white precipitate may form, indicating BaS04.
Boil the solution gently and if it contains a precipitate, filter and wash the
precipitate with hot dilute H 2S0 4 (1 : 20) and finally with hot water until
free from Ti, indicated by testing with 3'% H 20 2 solution, which gives a yellow
color with the washings if Ti is present.
N oTE.-Instead of dissolving the ash with concentrated H 2S0 4, it may be fused with
Na 2CO. in a covered Pt crucible over a Meker burner, using 5-10 grams of dry Na 2CO.
for 0.5 gram of ash. Digest the melt with 20--25 ml. of concentrated H 2S0 4in about 1~5
m). of water; boil; and filter to remove any BaS04 and other insoluble matter. Wash
the precipitate free of Ti with hot dilute H 2S0 4 (1 : 20) followed by hot water.
The filtrate and washings containing the Ti salts should have a volume of
about 250 ml. Reduce and titrate as follows: Drain the Jones reductor to the
top of the fine-mesh Zn and add 200 m!. of boiling dilute H 2 S0 4 (1 : 20).
,Slowly drain the reductor and then wash at least four times with boiling water.
Drain the reductor to the top of the fine-mesh Zn, wash the hot Ti solution
into it and allow it to remain for 15--:20 minutes, keeping the reductor covered
with a watch glass. During the whole period of reduction pass a slow stream
of CO 2 through the receiving flask. Slowly drain the reductor into the receiving flask while CO 2 is continuously passed through it. When the solution
has reached the level of the fine-mesh Zn, add more of the boiling dilute H 2S0 4
and bring the level of the solution to the top of the stick Zn. Drain the reductor to the fine-mesh Zn, then wash 4 times with boiling water in the same
manner as above. Remove the flask, add 7 ml. of the NH 4SCN indicator and
titrate the solution rapidly with the standard ferric alum solution. Calculate
the titration to percentage of Ti0 2
NOTES. I.-In titrating the Ti solution, the greatest accuracy is obtained by adding
nearly all of the ferric alum before agitating; then agitate by shaking with a gentle
rotary motion which is continued until the final end-point, a light straw color, is reached.
2. Before using the reductor run a blank, using the same concentration of H 2S0 4 in
water and the identical procedure as used for the Ti determination, the reduced solution
being titrated. This blank should not require more than 0.1 ml. of the standard ferric
alum solution. If more is needed, the Zn should be reamalgamated or another lot of
ferric alum used.

Scheme 2: Analysis of Ash Containing Ti0 2 in Absence of BaS04. (Ti0 2

in mixture with clay, talc, chalk or calcium sUlfate_.)
(a) Silica.-Fuse 0.5 gram of ash with 5-10 grams of Na 2 C0 3 in a covered
Pt crucible over a Meker burner. Treat the melt with about 125 ml. of dilute
HCI (1 : 4), filter, and wash with hot water. Ignite the residue in a tared Pt
crucible at the highest heat of a Meker burner, cool in a desiccator and weigh.
Treat the ignited and weighed residue with about 5-10 ml. of HF and 3-4
drops of concentrated H 2S0 4, evaporate to dryness on a hot plate and finally
ignite to constant weight. If the amount of Si0 2 is considerable, add more

1928

PAPER AND PAPER-MAKING MATERIALS

HF and H 2S0 4 , and repeat the evaporation and ignition. The loss in weight
,
.
represents Si0 2
Evaporate the filtrate from the melt to drynes~ and bake on. the hot nlate
for about 72 hour to insure complete dehydration lof the SiG 2 Take up the
residue with warm dilute HCl (1 : 4); filter, and 'Yash with water. Dry the
residue, ignite and weigh in a tared Pt crucible. Treat with 5-10 ml. 'of HF
and 3-4 drops of concentrated H 2S0 4 , evaporate to dryness on a hot plate and
finally ignite to constant weight. The loss in weight represents soluble Si0 2
Add this "result to the previous to give the total Si0 2
If residues are present after the Si0 2 has been removed, fuse widl KHS0 4,
dissolve the melt in dilute H 2S0 4 (1 : 4) and add to the filtrate from the Si0 2
determinations.
(b) Titanium Dioxide and Alumina.-Add a slight excess of NH 40H to the
combined filtrates and washings from the Si0 2 determination, boil the solution,
filter, and wash the precipitate with 5% NH 40R. Dissolve the precipitate
in the original beaker with dilute HCl (1 : 1) and reprecipitate with NH 4 0H.
Filter through the original filter, and wash, combining the washings with the
first filtrate and washings. Dry the residue, ignite to constant weight over
a Meker burner and weigh as A1 2 0 a+Ti0 2 Subtracting the. Ti0 2, as determined in Scheme 1, gives the A1 20 a. (4.ny Fe present as an impurity will be
weighed as Fe 20a and included with the A1 20 a.)
/

NOTE.-Even if AI,03 is not present, a double precipitation is still necessary to

insure complete separation of Ti0 2, since the gelatinous precipitate tends to absorb
appreciable amounts of Ca, Mg, etc. In this case the precipitate may be dried, ignited
and weighed as Ti0 2

(c) Calcium Oxide.-Slightly acidify with HCl the combined filtrates from
the Ti0 2 and A1 20 a separation and adjust the volume to 150-200 mt Make
this solution alkaline with NH 40H, boil and add an excess of ammonium
oxalate solution. Boil gently for about 2 minutes, digest on a steam bath for
1 hour and let cool to room temperature. ;Filte~J wash the precipitate with
hot water, dry, ignite strongly, and weigh as,CaO.
If preferred, the CaC 20 4 may be washed ,free of soluble oxalates, dissolved
in dilute H 2S0 4 (1 : 10), heated to 70 C. and titrated with standard KMn04
solution.
If CaS04 is presentl calculate the CaO weighed or titrated to CaS04.
Calculation:
-CaO X2.428 = CaS04.
(d) Magnesium Oxide.-Adjust the volume of the filtrate from the Ca
separation to 150-200 ml. The solution should contain 5 to 10% of NH 4 0H.
Add an excess of (NH 4 ).HP0 4 solution with constant stirring and let stand
for at least 4 hours, preferably overnight. Filter the precipitate on a previously ignited and weighed Gooch crucible. Wash with dilute NH 4 0H
(1 : 1), dry and ignite, slowly at first and finally over a Meker burner, until
the weight is constant. (Do not fuse.) Cool in a desipcator and weigh as
Mg 2P 20 7 Calculate to MgO.
Calculation:

PAPER AND PAPER-MAKING MATERIALS

1929

ANALYSIS OF BISULFITE COOKING LIQUOR 49

Bisulfite liquor, used in the manufacture of sulfite pulp, consists essentially
of a solution of calcium bisulfite, Ca(HSO a)2, magnesium bisulfite, Mg(H80 a)2
and free sulfur dioxide, S02'
'
For control tests in the mill, a volumetric analysis is usually sufficient.
For a complete analysis, however, gravimetric 'methods are necessary. Results are usually reported in grams per 100 ml.
VOLUMETRIC CONTROL ANALYSIS

Preparation of Sample.-Into a 100-ml. volumetric flask, containing about

50 ml. of water, pipette accurately 10 ml. of bisulfite liquor, holding the tip
of the pipette just under the surface of the water. Then dilute the solution
with water to the mark and thoroughly mix.
Total S02.-Pipette 25 ml. of 0.1 N iodine solution accurately into an
Erlenmeyer flask and add about 100 ml. of water. Then pipette 10 ml. of
the prepared solution of the sample into the iodine solution and titrate the
excess of iodine with 0.1 N thiosulfate solution, adding a little starch indicator
solution near the end of the titration if desired. If starch solution is added
the end-point is the disappearance of the blue color. If no starch is used th~
end-point is the change from pale yellow to colorless. Subtract the amou~t of
0.1 N thiosulfate from the original amount of 0.1 N iodine solution (25 ml.)
and calculate the difference to grams of 80 2 per 100 ml. of the original sample.
Calculation:
1 ml. 0.1 N iodine =0.0032 gram S02.
NOTE.-In control analyses of used or partially used liquors the above procedure
may give high results due to dissolved encrusting matter from the wood. In such cases
the result obtained should be considered preliminary and a final analysis should be made
by adding to the flask an amount of 0.1 N iodine about 5 mI. less than the amount consumed in the preliminary test. Then add 10 mI. of the solution of the sample followed
by a little starch indicator and complete the titration with 0.1 N iodine until a blue
color appears which is stable for a few seconds.
, Free SOz.-Pipette 10 ml. of the prepared solution of the sample Into an
Erlenmeyer flask, dilute to about 100 ml. with water and titrate with 0.1 N
N aOH solution, using phenolphthalein solution as an indicator and continuing
the titration until a pilfk color is formed whic~ is permanent for 1 minute.
Calculate the titration to grams of 802 per 100 ml. of the original sample. .
Calculation:
1 ml. 0.1 N NaOH=0.0032 gram 80 2.
Combined S02.-Report the difference bet-ween the total S02 and the free
S02, both expressed in grams per 100 mI., as the combined S02.
NOTE.-The free 80 2 is the a~tual free 80 2 plus half of the 80 2 in the bisulfites of
calcium and magnesium, and is more properly called the "available 80 2 " as it indicates
the 80 2 in excess of the amount necessary to form monosulfites. The combined 80 2
as above calculated represents the 80 2 combined as monosulfites, or in other words is
half of the 80 2 in the bisulfites of calcium and magnesium.
49 T 604 m.
Tentative Standard-Dec. 24, 1931. Official Standard-Oct. 2, 1933.
Corrected-April 15, 1935.

1930

PAPER AND PAPER-MAKING MATERIALS

COMPLETE GRAVIMETRIC ANALYSIS

Ordinarily results shall be reported in grams perHOO m!. of the bisulfite

liquor. If actual percentages by weight are desired, the figures thus obtained
...
shall be divided by the specific gravity of the original Folution.
Silica.-Pipette 25 m!. of'the original liquor into a,small beaker or porcelain
dish and, after adding 5 mI. of concentrated HCl, e\[aporate the mixture to
dryness on the steam bath and dehydrate for 30 minutes at about 1300 C.
To the residue add 5 mI. of concentrated HCI, and then 100 mI. of hot water.
Boil and filter. Wash the residue on the paper thoroughly with hot water,
ignite over a Bunsen burner and finally over a blast lamp or Meker burner,
then cool and weigh as'SiO z.
Iron Oxide and Alumina.-Dilute the filtrate from the silica determination
to about 300 mI., then add 10 m!.. of 10% NH 4 Cl solution, bring the mixture
to a boil and add a slight excess of NH 40H. After boiling until the odor of
NH3 is nearly but not quite gone, filter the solution, wash the residue thoroughly with hot water, ignite strongly, cool and weigh as Al z0 3+Fe z03.
Calcium Oxide.-Heat the filtrate from the iron oxide and 'alumina determination to bojJjng and to it add 20 m}. of 10% (NH 4)2C 20 4 solution. After
allowing the precipitate to settle, filter the solution, wash the residue on the
paper thoroughly with hot water, then ignite to constant weight over a Meker
. burner or blast lamp and weigh as CaO.
NOTE.-For very accurate work dissolve the original precipitate of CaC 20. through
the filter paper with a slight excess of dilute HCI and wash the filter paper with hot water.
Then reprecipitate the calcium in the filtrate by boiling with an excess of NH.OH and a
few drops of the (NH.)zCzO. solution, let settle, filter, wash and ignite.
In the case of a double pre()ipitation, e>,aporate the second filtrate rapi_dly to about
25 ml. and add to the first.
Magnesium Oxide.-Evaporate the filtrate (or combined filtrates) from
the CaO determination. to about 100 ro!. and add a slight excE of HCl, followed by 25 mI. of 10% NaNH 4 HP0 4 solution. Cool the sOlution and add
25 m!. of concentrated NH 40H slowly from a burettE/or dropper with constant
stirring. Let the precipitate settle in a cold place for at least 2 hours (preferably over night), then filter on a Gooch crucible containing an asbestos mat
which has previously been ignited, cooled and weighed. Wash thoroughly
with water containing about 2.5% of NH3 (1: part concentrated NH 40H to
g'parts of water), ignite pver a Meker burner (not strong. enough to soften or
melt the pyrophosphate), cool in a.desiccator and weigh as MgzF 207. Calculate to MgO.
Calculation:
Alternative Method for Calcium and Magnesium Oxides.-Pipette 25 m!.
of the original liquor into a platinum dish, add about 1 mI. of concentrated
H 2S0 4 , evaporate the mixture to dryness, carefully ignite, cool and weigh as
the mixed sulfates, CaS04+ MgS0 4. Dissolve the residue in 25 mI. of dilute
HCl, wash into a beaker, make alkaline with NH 40H, heat/to boiling and add
sufficient (NH 4)2C ZO. solution to precipitate the lime completely. Continue
the boiling for 2 minutes and let the precipitate'd~GaC204 settle for at least

. PAPER AND PAPER-MAKING MATERIALS

1931

72 hour. Filter and wash the precipitate thoroughly with hot water, ignite
to constant weight oyer a 'blast lamp and weigh as CaO. Calculate to Ca80 4
deduct the weight from that of the mixed sulfates and calculate the remaining
Mg80 4 to MgO.
Calculations:
CaO X2.428 = CaS04.
MgS0 4 XO.3349 = MgO.
Sulfur Trioxide.-Pipette 50 or 100 ml. of the original sample into a 300-ml.
Erl,enmeyer flask and completely displace the air above the liquid by a current
of C02, which has first passed through a wash-bottle containing an approximately 1 % solution of KMn04 to remove any oxidizable sulfur-compoullds.
Then, with the current of CO 2 still passing into the flask, add 10 m!. of concentrated HCI and evaporate the solution nearly to dryness to remove S02.
Transfer the solution to a beaker, washing out the flask and adding the washings to the beaker until the total volume is 100-150 m!. Acidify this solution,
with 2 ml.. of concentrated HCl, heat to boiling, and add 10 m!. of 10% BaCh
solution drop by drop while boiling. Let the precipitate settle for several
hours, preferably overnight, filter, wash thoroughly with hot water, ignite,
cool and weigh as Ba80 4. Calculate to SOa.
Calculation:
BaSO 4 X 0.3430 k 80a.

Total S02.-Determine the total 80 2 by titrating with 0.1 N iodine and

thiosulfate as described in the volumetric method.
Calculations.-Calculate the. SOa to CaS04 and the CaO to Ca(HSO a)2.
'Calculate the MgO to Mg(HSOah Express the excess of 80 2 beyond that
required ..for the bisulfites of calcium and magnesium as free 80 2.

SOaX1.700=Ca80 4
CaO X3.606 = Ca(HSOah
MgO X 4.625 = Mg(HSOa) 2.
Ca(HSOa) 2XO.6336 = S02.
Mg(HSO a)2 X 0.6872 = S02.
ANALYSIS OF CASEIN 50
Commercial caseins are often designated according to the method of prepa. ration, as acid casein, rennet casein, etc. Acid caseins are again subdivided
according to the acid used, as sulfuric, muriatic, lactic, etc. This designation
was formerly of some value, as the older methods of preparation gave products
of different characteristics according to the acid used. With modern methods
this is not true and caseins of practically the same properties are now produced with such different acids as lactic and muriatic. The name of a casein,
except for the distinction between acid and rennet caseins, therefore gives no
hint as to its properties.
Casein is a material for which there is at present no adequate system of
testing which can be applied by both producer and consumer. This is due
60

T 607 m. Tentative Standard-May 26, 1932. Official Standard-Oct. 2, 1933.

1932

PAPER AND PAPER-MAKING lY,IATERIALS

to the fact that there is a very wide gap between the information supplied by
strictly chemical tests and that necessary to tell howl the casein wilt work in
practice. This gap can be bridged only by empirical tests made to duplicate,
as nearly as possible, the conditions of actual use, and s'uch tests cannot readily
be standardized. Both types of tests are here describJd.
Sampling.-The usual methods of sampling dried granular material in
packages shall be followed, taking equal portions from 2 to 10% of the pa'ckages according to the size of the shipment. The composite sample shall be
thoroughly mixed, quartered down to suitable size and placed in moistureproof containers.
:
Appearance.-The sample should be observed for color, odor, cleanliness
and fineness of grinding. There are no exact standards for the first three
properties and only experience can enable one to tell whether.a sample is up
to the desired quality. In general, however, a good casein should be of a
creamy yellow color, free from an excessive amount of dark orange-colored
particles and visible dirt and of a mild, characteristic odor, neither moldy,
musty nor rancid. The fineness of grinding may be determined by screening
on standard screens.
Chemical Tests.-Moisture.-For accurate work a sample of 3 to 5 grams
shaH be dried for 5 hours at 98 C. in a vacuum oven. For most work it is
sufficiently accurate to dry the sample to cpnstant weight at a tempera.wre of
lOG-105 C. in an ordinary oven. AIl weighings shaIl be made in a- glassstoppered weighing bottle, as casein takes on moisture rapidly when dry.
Fat.-A representative portion of the sample shaIl be very finely ground and
then 5 grams extracted with ether in an extractor of the Soxhlet type .. The
extraction shaH run for at least 16 hours or overnight, coIlecting the extract
in a weighed flask. After evaporating off the solvent, the extracted fat shall
be dried to constant weight at not over 100 C.
Routine Method.-A convenient method of moderate accuracy which utilizes the
Babcock milk tester is as follows:
f
Soak 2 grams of the casein in 6 mI. of water in a small beaker and when all particles
are thoroughly saturated add, with constant stirring, 9 ml. of concentrated H 2S0 4
Pour the solution into a Babcock milk test bottle (I8-gram) and wash out the beaker
with a mixture of 5 m!. of water and 5 mI. of concentrated H 2S0 4, adding the washings to
the bottle. Fill to the base of the neck with dilute H 2S0 4 (5 : 4) and whirl in a centrifuge
for 5 minutes. Then add sufficient hot dilute acid to bring the fat layer completely
within the graduated portion of the neck and whirl for 2 minutes... While still hot read
the amount of fat on the gra(luated neck. As the bottle is designea 'for~a" charge of 18
grams, this reading multiplied by 9 will give the percentage of fat in the casein.

Ash.-Weigh about 2 grams of casein into a large porcelain crucible and

ignite over a low flame until complete carboniMtion has taken place, being
careful that the contents of the crucible do not burst into flame. Then ignite'
very light gray, cool in a
over a burner or in a mume until the ash is white
desiccator and ,weigh. The final ignition should be at as Iowa temperature
as is consistent with obtaining a white ash. The procedure should not be
carried out in platinum as this metal is attacked by the phosphorus in the
casein.
'
If oxygen is available the operations may be. ~{ried 0\11, in a Rose crucible
and the final ignition assisted by a stream of oxygen.

qr

PAPER AND PAPER-MAKING MATERIALS

1933

Alkali.__:_Add about 5 m!. of distilled water to the ash and warm. Then
add a drop or two of phenolphthalein indicator and if the ash is alkaline titrate
it with 0.1 N HCI until colorless; then add 2 drops of methyl orange and complete the titration to a pink color. Calculate the total titration to Na 20.
Calculation:
1 m!. 0.1 N HCI=0.0031 gram Na 2 0.

NOTE.-The ash of a pure casein should not be alkaline to phenolphthalein or to

methyl orange. If it shows alkalinity, test it qualitatively for borax, sodium carbonate
and sodium phosphate.
Starch.-Warm, a portion of the sample with water. Cool and add a few
drops of very dilute iodine solution. A blue color indicates starch.

NOTE.-If added alkali has been found in the casein, make the solution slightly acid
before adding the iodine solution..
Total Acidity.-Weigh 20 grams of cas.ein into a beaker and add 100 m!. of
water. After soaking for a short time titrate with 1.0 N or 0.5 N NaOH,
ruhning in a little at a time and warming on the steam bath after each addition
untilthe end-point is reached. Litmus paper, moistened and dipped into the
hot solution shall be used as an indicator and the end-point shall be considered
as reached when blue litmus is turned very slightly red and red litmus is turned
very slightly blue. Report the result as percentage of NaOH required to
neutralize the casein.
, Calculation:
1 m!. 1.0 N alkali =0.04 gram NaOH.

Soluble Acidity.-Soak 20 grams of the casein in 200 mI. of water at room

temperature for at least 1 hour, filter and titrate 50 m!. of the filtrate with
0.1 N NaOH, using methyl orange indicator. Report the acidity as the number of m!. of 0.1 N alkali required to neutralize 1 gram.
Niirogen.-The nitrogen is not usually determined in casein except for
some special purpose. When it is desired the usual Kjeldahl method should
be used.
NOTE.-Authorities differ as to the factor for converting N to casein and it is known
to differ with different lots of casein. According to Browne, NXB.38 will give the percentage of pure casein: while Hopfner and Burmeister state that NXB.61 will give the
fat-, ash-, and moisture-free casein present. To convert to commercial, air-dry casein as
received, the factors have been found to vary from 7.69 to 9.10 for different samples.
It is suggested that the factor 8.00 be arbitrarily selected and assumed to apply in all
cases where it is desired to convert nitrogen to commercial air-dry casein.
Empirical Tests.-Insoluble Matter.-Weigh 100 grams of casein into a
liter beaker, add 300 in!. of water and stir thoroughly. After about 10 minutes
add a solution of 15 grams of Na 2 CO a (soda ash) in 100 m!. of water and stir
until thickening takes place. Put the beaker in a water bath at a temperature
of about 60 C. and let it stand, with occasional stirring, for 3 hours. At the
end of this period add 400 m!. of boiling water and stir thoroughly for about
5 minutes. Pour the solution upon a small brass, 200-mesh sieve which has
been weighed to 0.1 mg. and soaked in boiling water. Wash the sieve and
residue thoroughly with boiling water, then dry and weigh. Report the percentage of insoluble matter.

1934

PAPER AND PAPER-MAKING MATERIALS

Solubility (Cutting Test).-Weigh 50 grams of the ground casein into a

350-m!' beaker and add 7.5 grams of powdered borax. !Add, 250 m!. of water
at a temperature of 70 C. and heat at this temperature pn the water bath for
15 minutes, stirring constantly.. High-grade caseins will dissolve completely,
while inferior samples will show more or less insoluble I materials of v~rious
kinds. (It should be noted that a casein may not be ompletely sol!lble by
this test, yet be entirely satisfactory for coating purposes when more alkali
is used.)
Viscosity.-Viscosity tests are of considerable help in determining whether
a casein is satisfactory for coating purposes, but it. is nearly impossible to
standardize the methods employed. In the first place, there is no fixed rtllationship between the viscosity of a casein solution and that of a coating mixture made from it, and as the coating mixture is what interests the coatm;_
it is quite useless to test the casein solution only. In the sec\md place, thil
viscosity of a coating mixture varies with the kind of clay, the kind of satin
white, the proportion of the two in the mixture and the kind and amount of
solvents used in dissolving the casein. Unless standard grades of clay and
satin white can be distributed for testing purposes the best that can be done
is for each user to work up a test which suits the conditions in his own particular plant.
Clay-Carrying Strength.-In the case of casein used in the coating of paper,
strength may be considered to be the numger of parts of clay which one part of
~asein will hold on the surface of the paper with just sufficient strength so that
the coating will not be removed during the printing of the paper. (Differ~nt
mineral matters require different amounts ')f casein, but for testing purposes
clay has been generally adopted.) The following method is recommended:
Soak 50 grams of casein in 190 ml. of a solution containing 7 grams of
Na 2 CO a and warm on the steam bath until solution is complete. Add sufficient water so that the total weight of solution is 250 grams and 1 gram of
casein is present in 5 grams of solution. It makes practically no difference
what alkali is used as a solvent provided the solution i~lcomplete and not overheated for any length of time.
)
I
Into a heavy porcelain or enamel-ware cup weigh 100 grams of coating
clay which has been dried at 100 C. or higher, apd to this add 70 ml. of water.
When it has soaked for a short time, stir until free from lumps. This can be
done very conveniently by means of a copper spatula which is bent nearly at
right angles on the end to form a foot .with which lumps maY_ke pressed out
and broken up. Balance this cup with its contents of clay, water and spatula
on a balance which is sensitive to 0.1 gram with a load of 500 grams, and then
add 30 grams of the hot casein solution. Mix thoroughly and spread a little
of the mixture thinly and evenly on a 'sheet of paper similar to that used for
coating purposes. Replac~ the cup on the balan?e, add 5 grams of caseIn
solution, mix and again spread on a sheet. Repeat this process until it is
certain that enough casein 'has been added to give a coating of sufficient
strength. The sheets will then represent increments of 1 gram of casein, the
first being 6 grams per 100 of clay, the second 7 grams, etc. Mark the sheets
6, 7, 8, etc. If care is taken not to spill or waste p,ny of tlle coating mixture
during the test, the amount used for making eacwsheE;lt will not introduce any
appreciable error.
-

PAPER AND PAPER-MAKING MATERIALS

1935

The coated sheets should be dried in the air at room temperature until
fully dry and tested with sealing wax for strength. For this test heat a stick
of sealing wax on one end until it almost drips and press the hot end firmly
down on the surface of the coated paper. When it is fully cold and hard pull
it off with a .quick, vertical pull. When doing this the paper must be held
firmly; this may be done by using a large washer wh~ch will fit over the wax
and hold the paper down entirely around it. If insufficient casein is. used, the
coating will pull off without taking any fiber with it; while if excess casein is
used, the surface of the wax will be entirely covered with coating to which
fibers are adhedng. There is generally an intermediate" critical point" when
the central area of the wax shows fibers and the edges coating only. This/
would be considered just on the line between weak and strong.
The test should be reported as the number of pounds of clay held by one
pound of casein in the coating which gives the critical point. This figure is
obtained by dividing 100 by the number of grams of casein required to give
the critical point.
NOTE.-The results obtained by this test are influenced by the kind of clay used as
well as the kind of paper to which the coating miXture is applied. Of still greater importance is the thickness of coating applied and it has been proved that for accurate work
it is necessary to keep the weight per ream of coating constant within 1 pound on either
side of the desired amount. This cannot be accomplished by spreading the coating
with a brush or even by means of a hand scraper, and the only aatisfactory solution is to
construct a machine which Vlill apply a very uniform and controllable amount.
ANALYSIS OF SALT CAKE

51

The principal use of salt cake in the pulp and paper industry is in the manufacture of sulfate pulp. Salt cake is an anhydrous sodium sulfate which
contains varying amounts of impurities. The principal impurities usually are
sulfuric acid, common salt, and sulfates of Fe, AI, Ca and Mg. "Chrome Salt
Cake," obtained as a by-product in making chromic acid, has a yellow color
and contains Na 2 Cr 207 (and/or Na 2 Cr04).
A complete analysis of salt cake is somewhat complicated and involves a
considerable number of determinations, as outlined below. Since some mills
are mainly interested in the acidity and the Na 2S04 content, directions for a
. " short analysis" are also given.
Preparation of Sample.-Mix the sample thoroughly, breaking down or
grinding any lumps; quarter down to about 100 grams; grind the quartered
portion, if necessary, to about 50-mesh or finer; and place in a separate stoppered bottle.
COMPLETE . ANALYSIS ,

Moisture.-Weigh 10 grams of the prepared sample in a flat-bottomed dish

of porcelain, glass or platinum and dry to constant weight at lOG-105 C.
(Four hours is usually sufficient.) Let the dish and contents cool in a desiccator
before weighing and make the final weighing rapidly. Calcuhite the loss in
weight to percentage and report as Moisture.
51

T 619 m. Tentative Standard-Nov. 2,1933. Official Standard-April 15, 1935.

1936

PAPER AND PAPER-MAKING MATERIALS

Insoluble Matter.-Weigh 50 grams of the prepared sample into a 40(1- or

500-ml. beaker. Add about 250 ml. of boiling water and stir well for 5-10
minutes. Filter at once through a filter paper which ha~ been previously 'dried
at 105 C., cooled and weighed in a weighing bottle, collecting the filtrate in a
liter volumetric flask. Wash the residue and the filter'paper thoroughly with
hot water until a few drops of the washings give no test for sulfates. Place
the filter paper and contents in the weighing bottle and dry to constant weigJlt
at 10{}-105 C. Calculate the percentage of insoluble matter.
Cool the filtrate and washings in the liter flask to room temperature, dilute
to the mark at 20 C. and mix thoroughly. Call this Solution A.
Acidity.-Pipette 100 ml. of solution A (equivalent to 5 grams of the
sample) into a beaker or. flask, dilute with an equal volume of water, add about
1 m!. of phenolphthalein indicator, and titrate to a pink color with 0.1 N NaOH.
Calculate the titration to percentage of H 2S0 4 and report as Acidity, calculated
as H 2S0 4
Calculation: 1 m!. 0.1 N NaOH=0.0049 gram H 2S0 4
N OTE.-This titration gives the total acidity or alkali-consuming power, and includes
nat only the free H 2S0 4 present in the material but also the acidity due to sulfates of Fe
and AI, and of any bichromate. The actual free H 2S0 4 may be closely approximated
by calculating the amount of H 2S0 4 due to the vartous impurities and subtracting from
the percentage of acidity as H 2S0 4

Factors: H 2S0 4 equivalent = Al 2 0 a

X2.89
=FeO
X 1.36
= Fe 20a
X 1.84
= Na 2 Cr 207 XO.375

Chloride.-:--To the solution which has been titrated for acidity add a drop
or two of methyl orange or methyl red indicator, make very slightly acid with
a few drops of dilute H 2S0 4 , and bring back to neutrality by adding about 1
gram of powdered CaCOa and stirring. Set the flask on a white surface or
transfer the contents to a porcelain dish; add 10 drops of K 2 Cr04 indicator
solution (10%) and titrate with 0.1 N AgNO a until the brick-red coloration
due to Ag 2Cr04 remains just visible after stirring. Calculate the titration to
percentages of CI and NaCI (and also to the equivalent of Na 20 for use in the
final calculations).
'
Calculations: 1 ml. 0.1

AgNO a=0.00355 gram Cl. .

=0.00585 gram NaC!.
= 0.00310 gram N a 20.

Ferrous Iron.-To 5 grams of the original p[epared sample in an Erlenmeyer flask add 100 ml. of warm water and 5 ml. of concentrated H 2S04
Warm on the steam bath a few minutes, just long en~ugh to dissolve all soluble
matter. Cool under a water tap and titrate at once with 0.1 N KMn04.
Calculate the titration to percentages of FeO and FeS04 (and also the equivalent of Fe 20a).
Calculations: 1 ml. 0.1

KMn04=0.0072 gram FeO.

=0.0152 gra~S04'
= 0.0080 grain Fe20s..

PAPER AND PAPER-MAKING MATERIALS

1937

Ferric Iron.-pipette 100 m!. of solution A (equivalent to 5 grams of the

sample) into a 250-ml. beaker, add a few drops of concentrated HNO a, cover
with a watch glass, and boil a few m'inutes to oxidize the Fe. Add 5 m!. of
concentrated HCl or about 1 gram of'NH 4Cl. Then add slowly dilute NH 40H
in slight excess. . Boil until the odor of NHa is nearly but not quite gone, fitter,
and wash the r.esidue thoroughly with hot water. Save the filtrate. Place
the filter paper' in a weighed platinum crucible; dry in an oven; smoke off the
filter paper; ignite the residue over a Meker burner to constant weight; and
calculate the percentage of Fe 20a+ AI 20 a.
Add about 2 grams of KHS0 4 to the residue in the crucible and fuse quietly
over a gentle heat until the melt is clear. Dissolve the melt in warm water,
dilute to about 100 m!., add 5 m!. of concentrated H 2S0 4, and pass the hot
solution through a Jones reductor to reduce the Fe. Cool the reduced solution
and titrate at once with 0.1 N KMn04. Calculate the percentage of total Fe in
terms of Fe 20a and subtract the amount of Fe 20a equivalent to the FeO previously determined. The difference is the ferric Fe in terms of Fe 20a. (Calculate also the equivalent amount of Fe2(S.04)a.)

Calculations: 1 ml. 0.1 N KMn04=0.008 gram Fe 20a.

Fe 20aX2.50
= FeZ(S04)a.
Alumina.-Sub~ract the percentage of total Fe in terms of Fe Z03 from the
percentage of mixed oxides, Fe 203+ AI 20 a, and report the difference as Alumina.
(Calculate also the equivalent amount of Alz(S04)a.)

Calculation: Al 20 a X3.36 = Al.2(S04) a.

Lime.-Heat the filtrate from the Fe and Al hydroxides to boiling and add
2 m!. of concentrated NH 40H and 10 m!: of 4% (NH 4)2C 20 4 solution. Let
stand one-half hour or until the precipitate settles clear. Filter and wash
thoroughly with hot water. Char off the filter paper in a weighed platinum
crucible and finally ignite to constant weight over a Meker burner. Cool in a
desiccator and tak-:: the final weight rapidly. Calculate the weight to percentage of CaO. (Calculate also the equivalent amount of CaS04.)
Calculation: CaO X2,43 = CaS04.
NOTE.-If preferred, instead of igniting the CaC 20 4 to CaO and weighing, the CaC 20 4
may be titrated with 0.1 N KMnO. as described in the method for Analysis of Limestone.

Magnesia.-Acidify the filtrate from the lime determination with HCl, concentrate to about 150 .m!. and heat to boiling. Add 10 m!. of saturated
NaNH 4HP0 4 solution and boil for several minutes. Cool to room temperature
and add NH 4 0H drop by drop with constant stirring until the crystalline precipitate begins to form; then add a moderate excess and continue stirring for
several minutes. Let stand in a cool place overhight and filter on a Gooch
crucible, previously ignited and weighed. Wash with dilute NH 40H (1 : 10)
containing a few drops of concentrated HNO a, ignite in a muffle, cool in a
desiccator and weigh as Mg 2P 20 7 If a muffle is not available, place the Gooch
crucible in a larger platinum crucible and ignite directly over a flame. Do not
use sufficient heat to fuse or melt the precipitate. Calculate to percentage of
MgO. (Calculate also the equivalent amount of MgS04')

1938

PAPER AND PAPER-MAKING MkTERIALS

Calculations: Mg 2P 20 7XO.362 = MgO.

Mg zP 20 7 X1.08 =1\fgS04.

Chromium Trioxide.-Pipette 100 ml. of solution A (equivalent to 5 grams)

into an Erlenmeyer flask. Add 5 ml. of concentrated HCI and 10 ml. of 10%
KI solution, and titrate the free iodine with 0.1 N Na~S20a solution. When
the yellow-brown color has nearly disappeared, add a 'little starch indicator
solution and continue the titration ,to the disappearance of the blue co~ur.
From the titration calculate the percentage of CrOa and of Na ZCr 207. (Calculate also the erruivalent amount of Na 20.)
Calculations: 1 ml. 0.1 N Na 2SzOa=0.00333 gram CrOa.
.
=0.00437 gram Na 2Cr 207.
= 0.00207 gram N azO.
Total Soda.-(a) Chrome Salt Cake.-To 100 ml. of solution A in a 500-ml.
beaker add 10 ml. of alcohol and 10--15 ml. of concentrated HCI and boil down
to half volume to reduce the Cr and remove the bulk of the alcohol. To the
hot solution add a slight excess of NH 40H and 10 ml. of 4% (NH 4)zC 20 4
solution. Boil until the odor of NHa is nearly gone; let stand about 15 minutes;
filter and wash thoroughly with hot water, collecting the filtrate and washings
in.a 250-ml. volumetric flask. Discard the precipitate which contains the AI,
Cr, Fe and Ca. Cool the solution, dilute to 250 ml. at 20" C., and mix thor-~
oughly. Call this solution B.
/
.,
.
Pipette 50 ml. of solution B (equivalent to 1 gram of the sample) into a
weighed dish, preferably of platinum; evaporate to dryness on the steam bath;'
and ignite carefully to avoid spattering, gradually raising the heat until no
more fumes of NH4-salts or SOa come off. After cooling the residue moisten
it with about 2 ml. of dilute H 2S0 4 and repeat the evaporation and ignition
in order to convert any chloride to sulfate. Cool, moisten the"residue with a
freshly prepared, nearly saturated solution of (NH 4) 2CO a in wat~r; again
evaporate on the steam bath and ignite strongly-over a Meker burner. Cool in
"
a desiccator, weigh, and calculate to percentage.,
The residue is Na 2S04+ MgS0 4. Subtract! the percentage of MgS0 4,
previously determined, and calculate the difference to Na 20.
Calculation: Na zS04X0.4364=Na zO.
(b) Ordinary Salt Cake ..-In the absence of Cr, omit the preliminary boiling
with alcohol. Heat 100 mI. of solution A to boiling; add 10mI. Qf concentrated
HCI, then a slight excess of NH 40H and 10 mI. of 4% (NH 4)zC z0 4solution; and
proceed as above.
Sulfur Trioxide.-Pipette 25 ml. of solution B (equivalent to 0.5 gram of
the sample) into a 400-ml. beaker. Add a drop or two of methyl red and then
neutralize the solution carefully with very dilute NH 40H. To the neutral
solution add 300 ml. of water and 2 ml. of dilute ;HCI (1 : 18), and heat to
boiling. Add to the boiling solution 25 mL of 10% BaCl z solution, drop by
drop from a pipette. Stir well and set aside overnight at room temperature,
protected from fumes. Filter and wash with hot water until 25 ml. of the
filtrate show not more than a slight opalescence with A'gNO{ solution. Place
the filter and precipitate in a weighed platinum .cw6il5le and heat very slowly
and carefully over a low flame to char the filter paper without allowing it to

PAPER AND PAPER-MAKING MATERIALS

1939

catch fire. After completing the charring, ignite until white over an ordinary
burner, with the crucible uncovered. Cool and moisten the precipitate with a
few drops of dilute H 2S0 4. Heat very carefully to drive off the acid and finally
ignite to constant weight, cooling between weighings in a desiccator. Calculate
the weight of BaS04 to percentage of S03.
Calculation; BaS04XO.3430=S03'
Final Calculations.-From the total NazO subtract the NazO equivalent
to the NaCI and to the Na ZCr Z07; calculate and report the remaining NazO
as Na ZS04' .
From the total S03 subtract the S03 equivalent to the NaZS04, Al z(S04)3,
FeS04 and FeZ(S04)a; calculate and report any excess of S03 as free HzSO~.~
(The S03 equivalent of each of the sulfates is obtained by subtracting the corresponding oxide percentage from the sulfate percentage; e.g., 1.52% FeS04
-0.72% FeO =0.80% S03.)
Form of Report.-It is suggested that the results of a complete analysis to
be reported in the following form:
Moisture, loss at 100 0 ............................. .
Insoluble in Water ................................. .
Acidity, calculated as H 2S0 4........................ .
Chloride, C1. ...................................... .
Ferrous Oxide, FeO ................................ .
Ferric Oxide, Fe 20 .................................
Alumina, AlzO a ...............
Lime, CaO ......................................... .
Magnesia, MgO ................................... .
Chromium Trioxide, OrO ............... ; ........... .
Sulfur Trioxide, S03 ...... .' ......................... .

Per Cent

Combined as follows:
Sodium Chloride, NaC!. ... , . " .. " ............. ,
Sodium Bichromate, Na2Cr 207 ................... .
Aluminum Sulfate, Al Z(S04)3. ................... .
Ferrous Sulfate, FeSO 4
Ferric Sulfate, Fe2(S04) ........................
Calcium Sulfate, CaS04 ......................... .
Magnesium Sulfate, MgS0 4..... '................ .
Sodium Sulfate, Na 2S04 ........................ .
Free Sulfuric Acid, H 2S0 4....................... .
SHORT ANALYSIS

Prepare the sample as previously described, except that it may be quartered

down to about 25 grams before the final grinding.
Acidity.-Weigh 5 grams into a 300-ml. Erlenmeyer flask, add 100 ml. of
hot water and heat unti\all soluble matter has dissolved. Cool under a water
tap and pour into a 500-ml. porcelain dish, using .100 ml. of wat!)r in small portions to rinse the flask and complete the transfer. Add about 1 ml. of phenolphthalein indicator and titrate to a pink color with 0.1 N NaOH .. Calculate
the titration to percentage of H 2S0 4 and report as Acidity, calculated as H 2S04.
Calculation: 1 ml. 0.1 N NaOH=0.0049 gram H 2S04.
Sodium Chloride.-Follow the procedure previously described under Chloride. Calculate the titration to percentage of NaC}, and also to the equivalent
amount of Na 2S04.

1940

PAPER AND PAPER-MAKING MATERIALS

,

Calculation: N aCI X 1.215 = N a2S04.

Sodium Bichromate.-If the material contains Cr, warm a 5-gram sample
with 100 m!. of water in a 300-m!. Erlenmeyer flask until all soluble matter is
dissolved, cool to room temperature and proceed as ,desq~ibed previously under
Chromium Trioxide. Calculate the percentage of Na2Cr207 and also the
equivalent percentage of Na 2S04.
Calculation: Na 2Cr207 XO.542 = Na 2S04.
'.
Sodium Sulfate.-Treat 5 grams of the sample in a 500-m!. beaker with 100
m!. of hot water, 5 ml. of concentrated HCI and a few drops of concentrated
HN0 3 , and stir until all soluble matter is dissolved. Boil a few minutes to
oxidize the Fe. If the material contains Cr, add 10 m!. of alcohol and boil down '
to half volume. Then add a slight exc'ess of NH 40H and 10 m!. of 4%
(NH 4)2C 20 4 solution and proceed as previously described under Total Soda.
Calculate the percentage of the final residue, which contains all the N a a~
Na 2S04 and all or part of the Mg as MgS0 4. (The latter is usually very small
and is disregarded.) Subtract the Na 2S04 equivalent of the NaCI and of t4e '
Na 2Cr 207, and report the difference as Sodium Sulfate.

ANALYSIS OF ROSIN

r,'l

Rosin is the residue obtained after disti~ling off the volatile matter from
turpentine gum. It is graded according to color. The higher the temperature
to which the distillation is carried, the' darker is the resulting product. The
common commercial grades, in the order of their color, from the palest and
best to the cheapest and darkest, are: X, WW, WG, N, M, K, I, H, G, F, E,
D, and B. Grades G, F, and E are most frequently used in the paper industry;
The acid number is generally considered the best index, in selecting rosin for
making rosin size. Other tests which give valuable data for certain purposes
are saponification number, unsaponifiable matter, toluol-i,nsoluble matter, and
' /
ash.
, Sampling.-Rosin is usually received in wooden oarrels of 500 pounds
gross weight or metal drums containing a gross we~ght of 280 pounds. These
containers are stripped off, leaving a solid block of rosin. The sample for
test should be taken from approximately the center, of the block. Ordinarily it
is sufficient to sample two containers selected at random froII\._,a shipment. In
cases of dispute the number of containers to be sampled shall be,.mutually
agreed upon. The sample shall be reduced to laboratory size by 'crushing to
pass a half-inch sieve, mixing and quartering. The final laboratory sample
shall be prepared immediately before analysis and shall consist of approximately
200 grams. It shall be ground to pass a No. 10 siev~ mixed thoroughly, and
placed in a wide-mouthed bottle of such a size that it will be completely filled
I
with the powdered rosin.
NOTE.-As rosin has a marked tendency to oxidize very rapidly, samples should be
kept in lump form as much as possible.
I

Color Grade.-To determine the color grade of rosin' a-set ofstandard cubes
52

T 621 m.

Tentative Standard-Feb. 19, 1934.

-'OffiCial Btandard-ApriI15, 1935.

PAPER AND PAPER-MAKING MATERIALS

1941

must be available. These may be obtained from D. C. Campbell, Jacksonville,

, Fla.; also from Harris M. King, P. O. Box 333, Savannah, Ga.
Melt the rosin under examination and cast in a sheet iron mold to form a
cube of the same dimensions as those in the standard set. Care must be taken
to heat the rosin only just enough to pour, as overheating darkens the color.
After the cube has solidified, compare it as to color with the standards by looking through the cubes in daylight toward a neutral sky (not blue).
Acid Number.-Dissolve 3 grams of rosin by warming with about 100 ml.
of alcohol which has previously been neutralized by adding a few drops of
phenolphthalein solution and then 0.1 N NaOH until it is faintly pink. When
the ;osin has dissolved, titrate the solution toa pink color with 0.5 N NaOH,
adding more phenolphthalein if necessary. Express the result in terms of
milligrams of KOH (not NaOH) consumed per gram of rosin. This is the
Acid Number. It is also sometimes customary to report the percentage of acid.
This should be calculated as abietic acid.
Calculation: 1 ml. 0.5 N NaOH =28.06 mg. KOH=0.1511 gram abietic acid.
Saponification Number.-Weigh duplicate 2-gram portions of the powdered
rosin into Erlenmeyer flasks of about 300-ml. capacity. Add to each 25 ml. of
0.5 N alcoholic KOH solution. In a third flask place 25 m!. of the 0.5 N
alcoholic KOH (as a blank determination). Connect the flasks to reflux condensers and boil for 2 hours, swirling them from time to time to prevent. the
rosin from sticking to the sides of the flask above the liquor line. Cool and
titrate the contents of each flask with 0.5 N HCl and phenolphthalein to the
diJappearance of the pink color. From the titration of the blank subtract the
titration of each sample and report the number of milligrams of KOH consumed
by1 gram of the rosin as its Saponification Number.
Calculation: 1 ml. 0.5 N KOH = 28.06 milligrams.
Ester Number.-The ester number is the difference between the saponification number and the acid number.
Unsaponifiable Matter.-(a) In Alcoholic Solution.-Saponify 5 grams of
the sample by boiling for 2 hours under a reflux: condenser with an excess (about
50 ml.) of 0.5 N alcoholic KOH. ' Evaporate most of the alcohol, add about
100 ml. of water and extract in a separatory funnel with 25 ml. of acid-free
ether. Draw off the watery layer into a second separatory funnel, extract
this with a second 25-ml. portion of acid-free ed'er and add the ether .extract to
that in the first funnel. Wash the combined ether extracts with two 25-ml.
portions of water, adding the wash waters to the solution in the other funnel.
Pour the washed ether extract into a weighed Soxhlet flask. Finally extract
the water solution a third time with 25 ml. of ether, first using the ether to rinse
out the funnel which contained the ether extracts. Draw off the water layer
into another separatory funnel and wash the third ether extract twice with
water. Pour this washed ether extract into the flask containing the main ether
extract. Distill off the ether and dry the flask and contents at not over 1050 C.
to constant weight, cooling in a desiccator before weighing. Calculate the
percentage ,of unsaponifiable residue.
(b) In Aqueous Solution.-Saponify 5 grams of the sample by boiling for
4 hours under 'a reflux condenser with 1 gram of N a 2 C0 3 dissolved in 50-75 ml.
of water. Cool the solution and extract with acid-free ether as above.

1942

PAPER AND PAPER-MAKING

M~;rERIA~S

NOTE.-The residue as determined in (a) represents the true unsaponifiable matter.

'The residue obtained in (b), however, represents more'nearly tile rosin which is unacted
upon in determining the acid number and, since rosin size is made with soda ash, itfis
considered to be the unsaponifiable matter so far as size-makin1 purposes are concern:d,
Ash.-Ignite 5 grams of the powdered rosin in a weighed platinum crucible
to a white or light gray residue. Cool in a desiccator alul; weigh. Calculate to
percentage,
'
Toluol-Insoluble Matter (l).-The toluol-insoluble matte! is a measure of
foreign dirt-chiefly sand, chips, dirt and bark. In many cases the determination of dirt is not necessary. Where it is important, the following method shall
be used:
Place 50 grams of the freshly powdered sample in a 300-ml. beaker, add'15()
ml. of toluol, free from water and non-volatile residue, and dissolve the sample
with the aid of heat and occasional shaking. When no particles of rosin are
visible, filter at once through a porcelain Gooch crucible which has been previously prepared with a mat of pure, well washed asbestos, and has been finally.
washed thoroughly with the toluol, dried in the oyen at 1050 C. for 30 minutes,
cooled in a desiccator, and weighed. If the rosin filtrate is not clear, return it
through the crucible until clear, finally washing the residuerd the outside of
the crucible free from rosin with additional hot toluol. Dry the crucible and
contents to constant weight in the oven at 1050 to 110 C. (one hour usually
suffices), cool in a desiccator, weigh, and calculate the per:centage of toluolinsoluble matter.
/'
Literature Cited
1. Am. Soc. for Testing Mat. Standard D 269-30.

POISONS
THEIR DETECTION AND ESTIMATION
With Special Reference to Organic Poisons

Samples for toxicological examination should be placed in clean glass jars;

having airtight glass covers. The jars should be sealed, numbered and labeled,
and the sealer should affix his initials. The portions of the body to be preserved for the chemical examination are, as a rule, the stomach and contents,
a portion or all of the intestinal tract, about one fourth of the liver, both
kidneys, the spleen, the brain, the bladder contents, and occasionally the heart,
lungs and a portion or all of the spinal cord.
A preliminary examination should precede the analysis, noting the appearance, odor, color, and reaction to litmus of the samples. Opium, hydrocyanic
acid, carbolic acid, chloroform, alcohol, nitro benz ole, benzaldehyde and nicotine
may be suggested by the odor.
Salts of copper, portions of insects, or certain arsenical preparations or other
coloring matter may be suggested by the color.
The lining of the stomach should be examined; corrosion may indicate
caustic alkalies, acids, or other corrosive substances; a reddening of the lining
may indicate cyanide; a " toughened" stomach lining may indicate mercury
salts.
The color of the blood should be noted, red indicating possible hydrocyanic
acid or carbon monoxide poisoning.
The odor of such poisons as hydrocyanic acid and chloroform can often be
detected better in the brain than any other portion of the bodY.
All available information as to symptoms should be obtained and considered since this information will usually point to a particular class of poison
or even to a particular poison, Thus the work of the identification may be
lessened and, what is more important, valuable and unreplaceable material will
not be wasted in useless tests. It is usually safe to assume that but one poison
is present. However, a quantity of a poison insufficient to cause death may be
found, in which case the examination should be continued.
.
Food materials suspected of containing poisons may be examined by the
methods given but it is well to remember that in cases of intentional poisoning
the quantity of poison is usually large and the analyst should not be misled by
finding traces, as of arsenic, for instance, which might easily occur in a food
material due to the very common practice of using arsenical spray on fruits and
vegetables.
1 By Arthur R. Maas, Professor of Pharmacy and Toxicology, University of Southern
California.

1943

1944

POISONS

Classification of Poisons
Nearly all of the co~mon poisons may be placed in lone 'of three groups: I
Group I. Poisons volatilized without decomposition by distillation with ste~m
from an acid solution.
Group II. Organic poisons. Most of the poisons in this group can be Mparated from extraneous matter by treatment with hot alcohol containing
tartaric acid. Alkaloids, glucosides, bitter principles and many organic
compounds come in this group.
Group III. All poisonous metals.

NOTE. Caustic alkalies, mineral acids, and a few special poisons such as oxalic acid
do not adapt themselves to the above scheme and will have t() be treated separately ..

General Procedure
Organs of the human body should first be weighed and then a representative
weighed sample reduced to a finely divided mass (a clElan small meat grinder
answers nicely for this purpose).
.
It is advisable always to preserve a portion of the material in its original
form for possible further examination.
About one tenth of the thoroughly mixed materiaJ is used in testing for
each one of the three groups, the balance being reserved. for checking. If only
a small amount of the original material is available, tests lor aIL three groups 01
poisons may be made with the same port~on. In this case, after removal of
volatile poisons, the residue should be divided into two unequal portions, the
larger being used for non-volatile organic poisons and the smaller, togethef
with the residue from the non-volatile organic, for metttllic poisons.
Blank tests should be run upon all reagents used !Iond the chemist should
familiarize himself with the various tests before applyin~ them to the unknown.
At least three confirmatory tests, different in character, ehould be applied to the
unknown substance before the presence of a poison can be definitely established.
Whenever possible these confirmatory tests should inch.1de form, color, odor or
taste.
/
Group I
Place some distilled water in a clean round-bottomed flask connected with
a Liebig condenser. Distil about 10 ml. and test distillate with silver nitrate
solution. Absence of a precipitate indicates that tlle apparatus is clean.
Next pour out the distilled water and transfer a weighed portion of the finely
divided and thoroughly mi~ed material to the flask, ttdding enough distilled
water for free distillation. Then add tartaric acid solution gradually until the
mixture is acid after thorough shaking. The distillate should be received in a
test tube in which a little distilled water has been,placed. The distillate may
contain: .
CarboliC Acid
Yellow Phosphorus
Anilirle
Hydrocyanic Acid
Chloral Hydrate
Chloroform
Carbon Disulfide
Ethyl Alcohol
Benzaldehyde I
Methyl Alcohol
AmylJ{itrite
Acetone
AmY! Alcohol
Iodoform
Nitrobenzene

POISONS

1945

Collect the distillate in several portions. Test the first portion of 5-10 ml.
for hydrocyanic acid, chloroform, ethyl alcohol, methyl alcohol, acetone,
iodoform and nitrobenzene. The other portions of 10-20 ml. will contain the
less volatile substances, carbolic acid, etc. Observe the reaction of the distillate; if acid, test for hydrocyanic acid .. Observe the odor of the distillate.
A number of the above substances have characteristic odors. When there is
reason to believe that a certain substance is present, confirm the results by
, milking other characteristic tests. It is seldom necessary to test for all members
of the group. Test a portion of the first distillate with silver nitrate T. S.
Faiiure to obtain a precipitate eliminates hydrocyanic acid.
YELLOW PHOSPHORUS

Average dose, 1/120 grain.

Lethal dose,

1(8 to 172 gI'ains. 2

Preliminary Test: (Sherer's Test.)-Place the finely divided material in

a small flask and cover with water if necessary. Suspend in the neck of the
flask by means of a cork with a V-shaped slit two prepared strips of filter paper
in such a manner as to avoid touching the sides of the flask. One strip is
moistened with AgNO a solution and the other with a solution prepared by
adding an excess of N aOH solution to a solution of a lead salt. Warm the
flask gently to 40-50 C. on a water bath. If both strips darken, phosphorus
, may be present together with H 2S. If the AgNO a strip darkens and the Pb
strip does not, phosphorus may be present. If neither strip darkens, phosphorus
is no~ present. In case there is an indication of phosphorus, proceed asfollows:
Make a separate portion of the finely divided material quite liquid with
water to which a little H 2S0 4 has been added to neutralize any NHa present.
Transfer to a tlask or retort fitted with a long condenser terminating in a
receiver containing a dilute solution of AgNO a. The distillation is carried out
in a darkened room. Phosphorus may be detected by a phosphorescent ring
at the point where the vapors are condensing. Care must be taken to avoid
reflections of light which may be taken for phosphorescence. If a distinct
phosphorescence is observed, it is positive proof that elementary phosphorus is
present, but the absence of phosphorescence 'is not a good negative test, as
many substances like alcohol, oil of turpentine, H 2S, etc., may entirely prevent
this indication. If the preliminary test is positive and no indication is obtained
on distillation, more complete works like Autenrieth or Witthaus should be
consulted for further tests which include elementary phosphorus, phosphorous,
and hypo phosphorous acids.
HYDROCYANIC ACID

Average dose of 2% U. S. P. acid, 172 minims.

Lethal dose of 2% U. S. P. acid, 40 minims.
Lethal dose of anhydrous acid, 6/10 to 9/10 grain.
2 Where doses are given, it means average adult doses.
The lethal dose given represents the amount that would under average conditions probably be fatal to an adult.

1946

POISONS

In testing for hydrocyanic acid examine the brain the contents orAhe
r
stomach and intestines, and organs rich in blood such as the liver, and heart.
Examination should be made as soon as possible, as hydrocyanic acid rapidly
disappears through volatilization and combination wi'th sulfur of !proteins
during putrefaction forming sulfocyanates.
I
Note cautiously the odor of the material as well as the odor of the distillate.
Hydrocyanic acid has a characteristic" peach kernel" odor.
Apply the following tests:
1. Schonbein-Pagenstecher Test

In case of suspected cyanide poisoning this test should precede the distillation. A small quantity (15 grams) of the finely divided material is placed in a
small flask together with enough tartaric acid solution to moisten it thoroughly
and a piece of guaiac-copper" paper" a is suspended in the neck of the flask
in such a ~anner as not to touch the liquid. The flask and contents are then
gently warmed on the water bath. If the paper is not turned blue or bluish
green, it is positive proof that neither HCN nor a readily decomposed cyanide is
present. A positive test means only that HCN or cyanide may be present as
many substances such as Cl, HCl, NHa and Hij"Oa produce this same change in
color.

2. Prussian Blue Test

To a small portion (3 mI.) of the distillate in a test tube add a few drops of
KOH solution, then 1 or 2 drops of a FeS04 solution and 1 drop of a FeClasolution. Shake and warm gently . Acidify with dilute HCI. \ If cyanide is
present in quantity a precipitate of Prussian blue will appear immediately, but
if present only in traces the solution will acquire a blue or bluish green color
which on standing 10 to 24 hours will throw down a flocculent precipitate of
Prussian blue. The maximum sensitivity of this reaction is 1 , 5,000,000.
Since formaldehyde is the principal ingredient of most embalming fluids, the
Prussian blue test cannot be relied upon as a neg~tive lest when working with
embalmed tissues.

3. Sulfocyanate Test
To another portion of the distillate in a small evapor!!ting dish add 2 or 3
drops of KOH solution, enough yellow ammonium sulfide tocolor. ,the solution
yellow and evaporate to dryness on the steam bath. Dissolve in a-lew drops
of water and acidify with a little dilute HCI. Filter through a double filter
paper into another small evaporating dish. Add.2 or 3 drops of dilute FeCla
reddish to blood-red color
solution. If HCN is present in the distillate,
will appear, which is discharged upon the addition pf a few drops of HgCl 2
solution. The color SQ produced is due to ferric sulfocyanate. The maximum

3 "Guaiac-copper" paper is prepared by saturating strips of filter 'paper with a

freshly prepared 10% alcoholic tincture of resin of guaiac, drying and, just before using,
moistening with a 1 to 1000 CUS04 solution.
4 The author has found that hydrocyanic acid fails to,r.e!lPond
this test in the

presence of formaldehyde, due probably to the following.reacl~on:

HCOH+HCN = HOCH 2CN.

.to

POISONS

1947

sensitivity of this test is 1 : 4,000,000 and is not affected by the presence of

formaldehyde.
4. Silver Nitrate Test
Acidify 1 ml. of the distillate with dilute HN0 3 in a small test tube and add
2 or 3 drops of AgN0 3 sloution. If HCN is present, a white curdy precipitate
resembling that of AgCI is produced which dissolves in NH 40H. This reaction
cannot be due to HCI because when a very dilute solution of HCI is distilled
HCI is not found in the distillate. The maximum sensitivity of this test is
1 : 250,000.
5. Silver Cyanide Crystal Test
Acidify 1 m!. of the distillate with dilute HN0 3 in a small test tube. Place
a drop of dilute AgN0 3 on a microscope slide and carefully invert the drop over
the center of the tube so as to avoid touching the side of the tube. Allow to
stand for 15 to 20 minutes, then cover with a coverglass and examine under the
microscope, using a power of from 50 to 100 diameters. If HCN is present,
characteristic 'crystals of AgCN will appear under the microscope. (See Fig.
283.)

.6. Detection of Hydrocyanic Acid in the Presence of Potassium Ferrocyanide

Should the material contain potassium ferro cyanide (which is not a poison),
the ordinary tartaric acid distillation will show the presence of HCN even
though there is none present before distillation. Therefore, in medico-legal
cases where cyanide is found it is well to test the original material for potassium
ferrocyanide as follows:
'
Shake a small portion with water, filter and test with FeCla and dilute HCI
for'the Prussian blue reaction. If there is an indication of a ferrocyanide, distil
a fresh portion of the material with an excess of N aHC0 3 and repeat the tests
for cyanides.
7. Detection of Mercuric Cyanide
In cases of suspected mercuric cyanide poisoning a few m!. of a freshly
prepared saturated aqueous solution of H 2S should be added to .the contents of
the flask and distillation continued. If mercuric cyanide is present, HCN will
be found in this second distillation. This step is made necessary because
mercuric cyanide in dilute solution is not decomposed by either tartaric acid or
sodium bicarbonate.
8. Estimation of Hydrocyanic Acid
When HCN is found, it is always well to determine as accurately as possible
the amount present.
Acidify a weighed portion of the material with sulfuric or tartaric acid and
distil into dilute N aOH free' from chlorides. Determine the cyanide in th~
distillate by a good volumetric method. If chlorides are present in the dis'
tillate, they may be removed by one redistiilation over borax.

1948

POISONS
CHLOROFORM

Average dose, 5 minims.

Lethal dose by inhalat.ion, from 15 drops up.
Lethal dose by mouth, 2 td 8 drachm!i..!

Distribution: When inhaled, chloroform passes from the air into the bloodplasma, then to the red blood corpuscles. It can often 'be found in the brain.
Phenylisocyanide Test
To a small portion of the distillate add 1-2 drops of aniline oil and ~ few m!.
of KOH solution and heat gently. If chloroform is present, the offensiv,e,
characteristic odor of phenylisocyanide is produced. This test is sensitive to 1
in 5000. This odor is produced by other substances such as chloral and ,chloral
hydrate and for this reason other tests should be applied.
Schwarz's Resorcinol Test
To 2 ml. of a 5% solution of resorcinol add a few drops of NaOH solution and
a small portion of the distillate and heat to boiling. If chloroform is present
even in small quantity, a yellowish red color, is produced. Chloral, bromal,
bromoform, and iodoform also give this test.
N icloux' s Method (Quantitatil'e)
To 20 m!. of blood or other aqueous liquid add about 95 m!. of alcohol and
5 ml. of a 5% solution of tartaric acid in alcohol. Distil 40 ml. into 10 ml. ot
alcohol, using a condenser with an adapter dipping beneath the surface of the
alcohol so that none of the chloroform may be lost. To the distillate add 10 m!.
pf a 10% alcoholic solution of KOH (free from chlorides) and boil under a
reflux condenser for 30 minutes, cool, add 15 ml. of distilled water, neutralize to
phenolphthalein with dilute H 2S0 4 , add a few drops of neutral potassium
chromate solution and titrate with standard silver nitrate solution.
3CI= ICHCla; CHCla

+ 4KOH =

HCOOK

+ 3KCI + H 0.
2

ETHYL ALCOHOL

Lethal dose, 372 to 7 fluid ounces.

Distribution: Ethyl alcohol may be found in the brain, liv~r and ,blood.
petection

1. Lieben's Iodoform Test

Warm a small portion of the first distillate to 40 pr 50 C. Add )-2 ml. of
an aqueous iodo-potassium iodide solution and enough KOH solution to give a
distinct yellow to brownish color. If alcohol is present, a yellowish white to
lemon yellow precipitate of iodoform is formed. If the solution is very dilute,
some time may be required to form the precipitate. When fqrmed slowly the
crystals may be identified under the microscope. This test is/very delicate but
is not characteristic of ethyl alcohol ,because othe1"'primary alcohols (except

POISONS
methyl alcohol) and many secondary alcohols produce this reaction:
Fig. 281.)

1949
(See

2. Chromic Acid Test

Warm a small portion of the distillate with dilute H 2S0 4 or HCI and add
1-2 drops of a very dilute K 2 Cr 201 solution. The color of the liquid will change
from red to green and give off the odor of acetaldehyde. Many other volatile
organic compounds react in this manner.

3. Ethyl Acetate Test

Mix a small portion of the distillate with an equal volume of concentrated/
H 2S0 4. Add a very small quantity of anhydrous sodium acetate and heat.
The odor of ethyl acetate is produced if alcohol is present.
4. Vitali's Test
Thoroughly mix a small portion of the distillate in a glass dish with a small
piece of solid KOH and 2 or 3 drops of C8 2 Let stand a short time without
warming. When most of the CS 2 has evaporated add a drop of ammonium
molybdate solution and then an excess of dilute H 2S0 4. A red color is produced
if alcohol is present. Acetone and acetaldehyde will produce a similar color.
A. 5% solution of alcohol gives this test distinctly.
METHYL ALCOHOL (WOOD ALCOHOL)

Fatal dose, 1 fluid ounce.

Methyl alcohol is found in methylated spirit, which contains 10 parts
methyl alcohol and 90 parts ethyl alcohol. It is also found in some other forms
. of denatured alcohol. It may poison either from external application or internal administration. One teaspoonful has produced blindness and one ounce
death.

Detection
To 5 ml. of the first distillate add 1 ml. of 25% H 2S0 4 and 4 m!. of N/5
KMn04; allow to stand 10 minutes and filter. (Decolorize with H 2S0 3 if
necessary.) To 5 m!. of the filtrate add 1 drop FeC1 3 solution and about 5 m!.
milk. Underlay this mixture with concentrated H 2S0 4. A violet ring at line
of contact indicates formaldehyde (oxidation product of methyl alcohol).
To the balance of the.filtrate (5 m!.) add 8 drops of an aqueous solution of
resorcinol (1-200) and then carefully pour this upon 5 m!. of concentrated
sulfuric acid contained in a test tube in such a manner that the two liquids do
not mix. After standing for three minutes a rose-red ring at line of contact
indicates formaldehyde (oxidation product of methyl alcohol).
CARBOLIC ACID (PHENOL)

Average dose, 1 grain.

Fatal dose, 1-4 drachms.

Distribution: Carbolic acid is very rapidly absorbed by the skin and gastro-

1950

POISONS

intestinal tract and appears in the liver, stomach, blood, kidneys, brain and
urine in a very short time after taking.
1
1. If present in quantity, this acid may be recognized by its odor.

2. Millon's J'est

.. \

Heat'1 m!. of the distiUate with a few drops of Millon's Reagent. (See
Special Reagents.) If carbolic acid is present even in minute. quantity, a distinct red color is produced. The absence of this color is an excellent indication
of the absence of phenol but other substances such as the three cresols and
salicylic acid may produce this color and for this reason other tests should be
applied if this test is positive.

3. Bromine Test
To 1 ml. of the distillate add 1 or 2 drops of freshly prepared saturated
bromine water. If phenol is present in quantity a yellowish white precipitate
is immediately thrown down, but if highly diluted some time may be required.
The precipitate should be examined under the microscope for characteristic
crystals of tribromphenol. (See Fig. 282.)
Phenol in a dilution of 1 : 50,000 yields 01). standing a precipitate containing
some of the characteristic crystals.
Substances like salicylic aldehyde and salicylic acid also produce this precipitate and other tests should, therefore; be applied.

4. Ferric Chloride Test

To a small portion of the distillate add a very dilute solution of FeCh drop
by drop. If phenol is present to the extent of 1 : 1000, a blue-violet color is produced. Dilute Hel and H 2S0 4 change the color to yellow. This test is entirely
negative in the presence of mineral acids. .

5. Hypochlorite Test

Add a little NH 40H to a small portion of the:distillate and then 2-3 drops of
freshly prepared solution of calcium or sodium hypochlorite and warm. If
phenol is present in quantity a blue color is produced, but if very dilute only
green or blue-green color appears. The maximum sensitivity of this reaction
is 1 : 1000.

6. Nitrite Test
To a small portion of the distillate add a few drops of a dilute alcoholic solu~
tiop. of ethyl nitrite and underlay the mixture in a test tube with concentrated
H~S04. If phenol is present, a red zone will appear a.t the point of contact of the
two liquids. The maximum sensitivity of this reaqtion is 1 : 10,000.

7. Quantitative Estimation of Phenol

For the quantitative estimation of phenol the analyst ,is referred to the
Beckurts-Koppeschaar Volumetric Method as described in Autenrieth-Warren,
" The Detection of Poisons and Powerful Drugs, H' page 31.

POISONS

1951

CHLORAL HYDR.A'l'E

Average dose, 8 grains.

Fatal dose, Yz to 2 drachms.
Distribution: Chloral hydrate may be found in the stomach, blood, brain
and spinal cord.
'
Detection

Chloral hydrate gives the phenylisocyanide and resorcinol test for chloroform but the distillate does not have the characteristic odor of chloroform .
. Add a few drops of Nessler's solution to a small quantity of the second dis~
tillate and shake. If chloral hydrate is present, a yellowish red precipitate is
produced. which on standing changes to dirty yellowish green. Formaldehyde
gives a brownish precipitate with this reagent. Boil a small portion of the distillate with 0.2-0.3 g. solid sodium thipsulfate. If chloral hydrate is present, a
turbid liquid of brick red color is produced. A few drops of KOH solution will
remove the turbidity and change the color to brownish red.
Decomposition of Chloral Hydrate

Heat a small portion of the distillate with calcined MgO for 30 minutes on
the steam bath with a reflux condenser. The' chloral hydrate is decomposed
into magnesium formate and chloroform. Test as follows:
Distil a few m!. of the decomposed liquid and test the distillate for chloroform. Filter the residue, concentrate to a small bulk and divide into two parts.
To one part add 1-2 drops of HgCl 2 solution and warm. Formic acid reduces
HgCl 2 to HgCI, producing a white precipitate. To the other part add 2-3 drops
of AgN0 3 solution and warm. Formic acid reduces AgN0 3 to metallic silver,
de~ositing a mirror on the test tube.
Group II
ALKALOIDS, GLUCOSIDES AND OTHER ORGANIC COMPOUNDS

Alkaloids form salts with tartaric acid insoluble in ether and chloroform but
soluble in alcohol and water. These alkaloidal salts are decomposed by alkalies
liberating the alkaloid which is insoluble in water, but soluble in ether or chloroform. The above is the principle by which alkaloids are separated from
foreign material.
.
Meth?d of Analysis

The substance is finely hashed, if solid, and then treated with several vol~
urnes of purified alcohol 5 and made slightly acid with tartaric acid. The
mixture is placed in a flask and heated under a reflux condenser for about 30
minutes. After cooling it is filtered .and the residue washed with purified
alcohol. The alcoholic extract is evaporated on a water bath to the consistence
of a syrup. From 3 to 4 volumes of purified alcohol are then gradually stirred
in, the solution allowed to stand an hour or more, filtered and carefully eva po6 The alcohol used should be purified by treating it with aboot 0.1 % of tartaric acid,
allowing to stand 24 hours and then distilling.

1952

POISONS

rated on a water bath to a thin syrup.

50 ml. of distilled water and filtered.

This. syrup is now treated with about

I

, NOTE.-If croton oil or nitroglycerine are suspected, they should be looked for in the
residue insoluble in water.
'

The filtered aqueous acid solution should be placed in a separatory funnel

and extracted with ether, and the extractions repeated until the ether on evaporation gives no residue.
Unite the ether extractions, wash once with a small quantity of water, filter
through dry paper and evaporate. The residue may contain:
Fats and Oils
Ptomaines
Picrotoxin
Ccrlchicine
Picric Acid
Menthol
Camphor
Phenol
Cresol
Thymol

Acetanilid
Phenacetine
Chloretone
Brometone
Essential Oils
Cantharidin
Caffeine

Antipyrine
Salicylic Acid
Benzoic Acid
Veronal
Trianal
SuIfonal

Note the general appearance and taste of the residue. If bitter, test for
picrotoxin and colchicine. Veronal and trional have a bitter taste arid' give
characteristic crystals. Sulfonal is tasteless and gives prismatic crystals.
The aqueous liquid is next rendered slightly alkaline with N aOH and extracted wi_th ether as before.
The extracted substances from which the solvent has been evaporated should
be weighed so as to determine the amount present. If material so recovered
is crystalline, the weight may be taken as representing the actual amount
present. If not crystalline, the extract may be purifi'd in most cases by dissolving in dilute acid, making the solution alkalrne as/before and re-extracting
with the volatile solvent.
The residue from the evaporation of the ether:may contain any of the following substances:
PtomaineE!
Coniine
Nicotine
Aniline
Toluidine
Veratrine
Strychnine

Atropine
Hyoscyamine
Scopolamine
Cocaine
Physostigmine
Brucine
Hydrastine

Codeine
Narcotine
Pilocarpine
Sanguinarine
Berberine

Antipyrine
Quini-ne
Caffeine

Carefully_ acidify the aqueous alkaline residue remaining from the preceding
ether extraction with dilute HCI and then make alkaline with dilute NH 40H,
using litmus paper for both tests. Extract repeatedly with hot chloroform
containing 10% by volume of alcohol.

POISONS

1953

The residue from the evaporation of the chloroform may contain:

Morphine
Apomorphine
Narceine
ACETANILID

Dose, 2-10 grains.

Lethal dose, 60 grains has proven fatal, less is dangerous to a person with a
weak heart.

Phenylisocyanide Test
Dissolve a small portion of the residue in 5-6 ml. of a boiling solution of
alcoholic potash. Cool, add 2 or 3 drops of chloroform and again heat. If
acetanilid is present, the offensive odor of phenylisocyanide is produced. The
alcoholic potash decomposes acetanilid into aniline and potassium acetate.
The' former with chloroform gives phenylisocyanide.

Bichromate-Sulfuric Acid Test

To a small portion of the residue add 1-2 drops of concentrated H 2S0 4 containing a little K 2 Cr 2 07. Acetanilid produces a red color changing to brown and
finally dirty green.
'
VERONAL

Dose, 8 to 15 grains.
Lethal dose, 120 grains has proven fatal.

Distribution: Veronal should be looked for in the stomach and bladder

contents.

FIG. 275.-Atropine 1 : 500.

Wagner's Reagent.

With

FIG. 276.-Cocaine 1 : 1000.

Gold Chloride.

With

1954

POISONS

FIG. 277.-Codeine 1 : 200.

Marme's Reagent.

ftf#

FIG. 278.-Narcotine 1 : 1000.

Sodium Carbonate.

With

FIG. 280.-Strychnine 1 : 500. With

Ammonium Thiocyanate.

Wi.tl).

'tIS 1
...

..

..,

With

, t"

.-..

.. .
~

-=>=,_

,FIG. 279.-Nicotine 1 : 1000.

Gold Chloride.

FIG. 281.-Iodoform Crystals.

FIG. 282.-Tribromphenol Crystals.

POISONS

FIG. 283.-8ilver Cyanide Crystals.

1955

FIG. 284.-Morphine Crystals.

Purification and Detection

Dissolve a portion of the residue from the acid ether extraction in the
smallest possible quantity of hot water. If colored, de colorize by boiling a few
minutes with animal charcoal, filter hot and allow the solution to cool. If no
crystals are deposited, evaporate further and cool. Make the following tests
on any pure crystals obtained:
Dissolve a few crystals in a few drops of water and test with sensitive litmus
. paper. An aqueous solution of veronal has a faintly acid reaction.
Dissolve a few crystals in a few drops of dilute NaOH solution and acidify
the clear solution with dilute HCl. Veronal is precipitated.
Determ~ne the melting point on a few of. the crystals. Pure veronal melts
between 1880 and 189 0 C.
Mix a few of the crystals with 8: few crystals .of pure ver.onal and determine
the melting point of the mixture. The mixture should have the same melting
point as the unknown if the unkriown is pure veronal. Carefully sublime a
number of the crystals in a dry test tube and compare with crystals known to
be pure veronal. . A low power microscope is used for this purpose. Fuse some
of the crystals with metallic sodium in a dry test tube. Cool, dissolve carefully
in water and test for sodium cyanide by the methods given on pages 1945-1947.
Veronal contains nitrogen which reacts to form sodium cyanide.
Extraction of Veronal from the Bladder Contents
Evaporate a considerable quantity of the urine on the steam bath to 1/5 its
volume and extract several times with ether, using a large volume for each
extraction because veronal is not very soluble in this solvent. Evaporate the
extract and proceed with the residu.e as directed under "Purification and
Detection" given above.
NICOTINE

Dose, 1/20 to 1/10 grain.

Lethal dose, 2-3 drops pure alkaloid.

1956

POISONS

Distribution
Nicotine should be.sought in the stomach, lungs arid liver.
Detection
If present in large quantities, nicotine may be detected by its odor. If nicotine is suspected, the ether extract from NaOH solutioh should be evaporated
without heat. If nicotine is present, oily drops of a colorless or yellowish liquid
which reddens phenolphthalein and has a peculiar odor somewhat resembling
stale tobacco smoke will remain in the residue, If exposed to the light, the oil
becomes yellow or brown, thicker and finally resinous. If heated, it volatilizes
completely, giving off white fumes.
If a drop of Hel be placed on a watch glass and inverted over a similar watch
glass containing nicotine alkaloid, a white cloud is produced which is, however,
not so dense as that produced by coniine under like conditions and no crystars
are produced as with coniine.
Place a drop of an aqueous extract of the oily residue on a microscope slide,
add a drop of a solution of gold chloride, let stand for several minutes and
examine under the microscope. If nicotine is present, small characteristic
crystals will appear upon standing.
QUININE

Dose 15 grains.
Detection
Fluorescence Test

Dissolve a portion of the residue in a little dilute sulphuric acid.

is present, this solution will exhibit a vivid blue fluorescence.
Thalleioquin'Test

If quinine

Dissolve a portion of the residue in a few dl,'ops df very dilute acetic acid
and add few drops of saturated chlorine or bromine water and follow with
excess of ammonium hydroxide. The liquid acquires an emerald-green color.
Herapathite Test'

To a small portion of the fesidue add 20 drops of an acRimixt_ure (30 drops

of acetic acid, 20 drops of absolute ethyl alcohol and 1 drop of dilute sulphuric
acid) and heat to boiling. 'Finally add 1 drop of an alcoholic solution of
iodine (1 : 10). When the solution has stood for a few minutes, green leaflets
with a metallic luster will form.
'
STRYCHNINE

Dose, 1/60 to .1/12 grain.

Minimum lethal dose for an adult,

Y2 grain.

I
Distribution
Strychnine should be looked for in the stomach (lid contents and organs rich
in blood such as the liver and heart.

1957

POISONS
Detection

Strychnine has a very bitter taste and may be detected even in a dilution.of .
o'ne in 700,000 by the taste.

Bloxam's Test
Add 1-2 drops concentrated HNO a to a portion of the solid residue in a porcelain dish and gently warm. On adding a small crystal of KOlOa, a scarlet
color is produced if strychnine is present. NH 4 0H changes the color to brown
and a brown precipitate falls. On evaporating slowly to dryness, a dark green
residue is left. This residue is soluble in water, forming a green solution which is
changed to orange-brown by KOH ana green again by HNO a. This reaction
distinguishes strychnine from any of the alkaloids commonly occurring in cases
of poisoning.

Bichromate Test
'PlaJ!e a drop of an acid extraction of the residue on a microscope slide and
add a drop of a dilute solution of K 2 0r 2 07. Let stand for a few minutes and
examine under the microscope. If strychnine is present, characteristic crystals
will be found. Dry the crystals and add a drop of concentrated H 2S0 4, holding
the slide over a white surface. A deep blue color which changes through deep
violet, purplish red, orange to yellow is produced if strychnine is present.
NOTE.-A mixture of hydrastine and morphine causes this same color reaction but
this mixture is not possible if the extractions have been properly carried out.

Sulfocyanate Test

Place a few drops of the acid extraction on a microscope slide and add a drop
of a dilute solution of ammonium sulfocyanate. Strychnine sulfocyanate
produces characteristic needle-shaped crystals.
BRUCINE

Dose, 1/12 to

grain.

Detection

Dissolve a small quantity of the residue in 1 or 2 drops of concentrated

HN0 3 If brucine is present, a blood red color will appear, which soon changes
to yellowish red and finally to yellow. Add a few drops of a'freshly prepared
dilute solution of stannous chloride to the yellowish red or yellow solution. An
intense violet color will appear. Heat usually changes this color to yellowish
red but more stannous chloride solution restores the violet color.
Dissolve a small portion of the residue in 2 drops of dilute HOI. Place a
drop of this solution on a microscope slide and add a drop of pla~inum chloride
solution. If brucine is present, characteristic rod-like crystals separate and may
be compared under the microscope with those produced in a solution of brucine.
These crystals are formed in solutions as dilute as 1 to 20,000. The best
crystals are obtained from the more dilute solutions.

POISONS

1958

ATRQ~INE

Dose, 1/120 to 1/60 grain.

Lethal dose, 1/20 grain. Much larger doses have, been followed by recovery.

Detection
Vitali's Test
Dissolve a small portion of the dry residue from the ether extraction of solution made alkaline with NaOH in a few drops of fuming HNO a and evaporate
to dryness on the steam bath. Cool and moisten the yellowish residue with a
few drops of a 4% solution of KOH in absolute alcohol. If atropine is present,
an evanescent violet color will appear. This color reaction is also obtained with
hyoscyamine, scopolamine, strychnine and veratrine, and therefore other tests
should be applied.
Physiological Test
Dissolve a small portion of the residue in a few drops of very dilute H 2S0 4
and place a drop of this solution in a dog's or cat's eye. If atropine is present,
a noticeable enlargement of the pupil WIll be observed. (Note: Great care
must be exercised if .this test is applied: to the human eye.) One drop of a
1 : 130,000 solution ,of atropine will produce this enlargement.
Dissolve a small portion of the residue in 2 drops of dilute H 2S0 4 Place a
drop of this solution on a microscope 'Slide and add a drop of Wagner's Reagent.
If atropine is present, an amorphous 'precipitate appears and slowly forms
numbers of small characteristic crystals which may be observed under the
microscope. This test is sensitive to 1 in 8000.
COCAINE

Dose, 1/8 to 1 grain.

Lethal dose, 15 grains,

Detectionl,
The material should be examined for cocaine as soon as possible because
cocaine undergoes changes in the viscera which are said to change it rapidly into
ecgonine,
Dissolve a small portion of the residue in water and_touch it to the tongue.
.. --.
Cocaine produces a t~mporary anesthesia.
To a small portion of the residue in a porcelain dish add a few drops of
strong HNO a and evaporate carefully to dryness. Treat the residue with a few
drops of alcoholic potash and warm. If cocaine is present, a characteristic
sweet odor will be given off.
Dissolve a small portion of the residue in 1 dr 2 drops of dilute HCl. Place
a drop on a microscope slide and add a drop of gold chloride solution. Cocaine
produces afteJ; a short time very characteristic crystals, theform depending on
the dilution. This test is sensitive 1 : 20,000.
CODEINE

Dose, Y2 to 1 grain.
Lethal dose, 5 grains.

Much larger doses have been recovered from.

POISONS

1959

Detection

Nitric Acid Test

To a small portion of the residue in a porcelain dish add a few drops of a cold
25% solution of nitric acid. Codeine produces a yellow color which soon
changes to red. Concentrated nitric acid produces a reddish brown color.
Froehde's Test
To a small portion of the residue in a porcelain dish add a few' 'drops of
Froehde's Reagent (see Special Reagents) and warm very gently over a very
small flame. Codeine produces a yellowish color which soon changes to green
and finally to blue.
Formaldehyde-Sulfuric Acid Test
To a small portion of the residue in a small porcelain dish add a few drops
of concentrated sulfuric acid to which has been added a trace of formaldehyde.
Codeine produces a reddish violet color which changes
to blue-violet.
.

Marme's_Reagent Test
Dissolve a small portion of the residue in 2 drops of dilute H 2S0 4 Place a
drop of this solution on a microscope slide and add a drop of Marme's Reagent
(see Special Reagents). Characteristic crystals soon begin to form in solution
of codeine as dilute as 1 : 2900 and may be observed under the microscope.
CAFFEINE

Dose 3 grains.
Detection
Murexide Reaction
Treat small portion of the residue in a porcelain dish with a few drops of
strong hydrochloric acid and a minute crystal of potassium chlorate, and
evaporate the liquid to dryness on the water bath. When cold invert the dish
over a vessel containing a few drops of ammonia when a characteristic purple
coloration is produced.

Theobromine and, xanthine give similar reactions to caffeine with an oxidizing agent and ammonia. The purple colorations due to caffeine and theobromine are decolorized by adding alkali hydroxide solution, but that due to
uric acid is changed to blue.
Reaction to Alkaloidal Reagents
Caffeine is very imperfectly precipitated by the usual alkaloidal reagents.
No reactions result with neutral iodized potassium iodide and Mayer's solution,
which behavior distinguishes caffeine from nearly all other alkaloids except
theobromine and colchicine. From an acid solution however caffeine is
quantitatively precipitated by a solution of iodine and potassium iodide.

1960

POISONS
NARCOTINE

Dose, 2 to 15 grains.
Dete6tion
Dilute Sulfuric Acid Test
Dissolve a small portion of the residue obtained from the ether extraction of
the NaOH solution in a little dilute H 2S0 4 and evaporate in a small porcelain
dish on the steam bath. If narcotine is present, the residue has a reddish yellow
color which changes on stronger heating to crimson red. As the acid begins to
evaporate, blue-violet streaks radiate from the margin and the entire liquid
finally assumes a dirty red-violet color.
Froehde's Reaction
To a small portion of the residue in a porcelain dish add a few drops of
Froehde's reagent. If narcotine is present, a greenish color is produced which
immediately changes to cherry red. Gentle' heat hastens the appearance of
the cherry red color which is quite persistent.
Selenious Acid-Sulfuric Acid Test
To a small portion of the residue add a few drops of selenious acid-sulfuric
acid reagent (see Special Reagents). If nar,cotine is present, a greenish steel
blue color appears which, after a time, changes to cherry red. Heat destroys
the cherry red color immediately.
.
Sodium Carbonate Test
Dissolve a small portion of the residue in 2-3 drops of very dilute HCI.
Place a drop of this solution on a microscope slide and add a drop of a 5% solution of Na 2 CO a If narcotine is present, small charagteristic crystals are pre'"
cipitated in solutions as dilute as 1 to 20,000. ;These crystals should be compared under the microscope with crystals obtained from solutions of narcotine
of different concentrations.
MORPHINE

Dose, 1/20 to Y2 grain. ,

Average lethal dose, 1 to 4 grains.
larger quantity.

Recovery may follow':;,ftertaking a much

Detection

Formaldeh1/de-Sulfuric Acid Test

To a small portion of the residue from the chlorbform extraction in a small
porcelain dish add a few drops concentrated H2S0~ containing 2 or 3 drops of
40% formaldehyde solution to 5 ml. acid. A trace of morphine produces a
purple red color which changes to "iolet and finally becomes pure blue. If this
test is negative morphine is not present, but a positive reaction must be confirmed by other tests because codeine, heroin. and -apomorphine produce a
similar color.
- .

POISONS

1961

Froehde's Test

To a small portion of the residue in a porcelain dish add a drop of Froehde's

Reagent. Morphine produces a violet color which changes through blue to
dirty green and faint red. Other alkaloids of the opium group give similar
color reactions.
Iodic Acid Test
Dissolve a small quantity of the residue in dilute H 2 S0 4 , add a few drops of
iodic acid and shake with chloroform. A violet color appearing in the chloroform layer may be due to other reducing agents as well as morphine.
Ferric Chloride Test

Dissolve a small quantity of the residue in dilute HCI and evaporate to

dryness on the steam bath. Take up with a few drops of distilled water and
add a drop of neutral FeCla solution. If morphine is present, a blue color will
appear.
Nitric Acid Test

To a small portion of the residue add 1-2 drops of concentrated HNO a

Morphine gives a blood red color which gradually changes to yellow.
Marme's Test

Dissolve a small portion of the residue in 2 drops of dilute HCl. "Place one
drop of this solution on' a microscope slide and add a drop of Marme's Reagent.
If morphine is present in solution to the extent of 1 to 1000, delicate silvery
needle-shaped crystals
form slowly around the edges of the drop and in
more concentrated solutions these crystals gfoup together to form rosettes as
shown in Fig. 284.

will

OPIUM

Dose, 1 grain.
Lethal dose, 4 grains has produced death. Larger amounts have been recovered
from.
The principal constituents of opium are morphine, codeine, narcotine, narceine, and meconic acid.
Narcotine can be extracted from an aqueous solution of opium by shaking
out with ether, and evaporating the ether. Meconic acid can be obtained by
adding lead acetate to an aqueous solution of opium and filtering off the precipitate of lead meconate (the filtrate will contain morphine acetate). The precipitate of lead meconate is decomposed by treating it with dilute sulfuric acid
and gently boiling the mixture, filtering and neutralizing. On adding a drop
of ferric chloride solution, a deep red color, not easily destroyed by a solution
of mercuric chloride or hot dilute sulfuric acid, indicates meconic acid. Mercuric chloride instantly bleaches the color caused by sulfocyanates while the
boiling dilute sulfuric acid destroys that due to acetates by decomposing the
acetate and expelling acetic acid.
The solution of morphine acetate should be treated with hydrogen sulfide
to remove excess of lead, the solution filtered and the filtrate evaporated to a

1962

POISONS

soft extract on a water bath. This extract. treated with alcohol, filtered and
evaporated will give morphine acetate to which the I chara9teristic tests for
morphine should be applied.
The presence of morphine, narcotine and meconic !).?id definitely establishes
the presence of opium.
.
PTOMAINES

The consideration of ptomaines is very important in toxicology because ~f

the similarity of these substances to alkaloids. Ptomaines contain nitrogen
and are basic in character and mayor may not be toxic. They are products of
decomposition occurring especially in advanced stages of putrefaction of a dead
body. Probably they are always present in cadavers and it seems probable that
they are normal products of bacterial life processes.
Great care should be exercised especially in legal-chemical cases not to report
an alkaloid present if there is any possibility of the identifying reactions being
caused by a ptomaine. No positive report should be made until e"ery charac:.
teristic chemical test of the suspected alkaloiqs has been confirmed, and in case
there is even a remote possibility that the suspected substance may be a
ptomaine, a physiological test should be made in order to confirm the chemical
' .
examination.
A ptomaine may resemble an alkaloid chemically but usually it will hav~
entirely different physiological action. ,Ptomaines resembling to a certain
degree the following alkaloids have been isolated: atropine, codeine, coniine,
delphinine, hyoscyamine, morphine, narceine, nicotine, strychnine and veratrine.
Frequently the resemblance between a ptomaine and an alkaloid lies in one
reaction only but an animal decomposition product has been found that resembles morphine in its characteristic solubilities and also liberates ioc~ine from iodic
acid but does not give the reactions that are characteristic of morphine alone.
Ptomaines may resemble alkaloids in the following characteristics: they
give precipitates with reagents that precipitate alkaloisis, their separation with
various solvents from both acid and alkaline solution ryembles various alkaloids
and many ptomaines are strong reducing agents .and give the Prussian blue test
in common with certain alkaloids.
When it is possible to isolate the alkaloid in a pure state, its identity can be
established beyond a doubt.
<

Group ITl

-......

This group contains all metals. Arsenic however is best tested for on a
small separate portion of the original material.
Preparation of Solution
Place from 50 to 100 g. of the finely chopped m~terial in a lOOO-ml. boiling
flask. Dilute with water if necessary until a fluid mixture is obtained and add
pure concentrated HCI and crystalline KCI0 3 in the proportion of 25 ml. HCI
and 2 g. KCI0 3 to every 100 g. Allow the mixture to stand cool for 24 hours
and heat on the steam bath (preferably under a refl-qx.condetiser) with frequent
agitation and the occasional addition of KC10 3 (D,.ljg. at a time) until the contents of the flask are yellow and fluid except for fatty and white granular matter

POISONS

1903

in suspension and until the liquid does not darken on heating for half an hour
without adding KCIO a. If, during the heating, the addition of KCIO a does not
produee any visible reaction, add more concentrated HCI in the proportion of
10 m!. HCI to every 100 g. material used and continue as directed above. Care
should be used to keep the quantities of HCI and KCIO a as small as possible, as
an excess of HCI may prevent the precipitation of some of the metals by H 2S.
Expel the excess of chlorine by passing a rapid stream of CO 2 through the cool
liquid or by adding small quantities of NaH80 a until a distinct odor of 80 2 is
obtained. Filter the solution through a wet filter and wash the residue
thoroughly with hot water. If a large amount of fatty residue is left, it should
be extracted two or three times with small quantities of hot water acidulated
with HCI and these filtered extracts added to the main filtrate. If silver and
small quantities of lead, thallium, mercury or barium are suspected, this residue
should be examined qualitatively for these metals. If a large excess of HCI is
present, add a concentrated solution of Na 2 COa to such an extent that the reaction still remains distinctly acid. Place the liquid in a flask fitted with a
'two-hole rubber stopper and tube dipping below ,the liquid and pass pure H 2S
through it at the temperature of the steam bath for 2 hours. Allow to stand
cold for eight or ten hours and filter. A colored precipitate at this stage may
mean nothing, as it may be due to organic thio-compounds, and therefore the
precipitate and filtrate should be tested by any good method of qualitative
analysis for metals.
Detection of Arsenic
Place 10 grams of the finely divided material in a IS0-ml. beaker, add 10 ml.
of concentrated HNO a, cov~r and allow to stand over night. Then add 5 ml.
concentrated H 2SO.{ and warm. When the material starts to turn brown, add
concentrated HNO a 1 ml. at a time until SOa fumes are given off without the
solution turning brown. Cool; carefully add 10 ml. of water and boil down to
80 a fumes. Cool, dilute, add 0.5 g. KI, warm to 90 C., add dilute solution of
SnCh to de colorize, cool and run in the regular small Gutzeit apparatus.
DETERMINATION OF CARBON MONOXIDE IN

AIR

(Sayers, Yant and Jones, Public Health Reports, Vol. 38, No. 40, Oct. 5, 1923)
The following method slightly abridged from that of the original authors
will give approximate results where the percentage of CO in air is less than 0.2%.
Blood Solution
Saturate a small amount of defibrinated blood with CO. This is done
readily by passing house gas containing CO through blood for a minute or two
or preferably by using air containing 3 to 5% of CO. Prepare a 1 to 20 dilution
of this blood with water. Also prepare a 1 to 20 dilution of CO-free blood.
For standards set up a series of nine test tubes containing 0.0, 0.2, 0.4 . . . 1.6
ml. of CO-blood dilution, each made up to a final volume of 2 ml. with the COfree blood dilution. To e!wh of these tubes add 40 mg. of a mixture of equal
parts of tannic and pyrogallic acids. Invert the tube 4 or 5 times to mix.
These standards give a series in which the proportion of saturation with CO
varies from 0 to 80 in steps of 10%. If air is.excluded by a paraffin seal on the

POISONS

1964

surface of the liquid, these standards will be, suffici~ntly permanent for a w~k
or two, but it is little trouble to prepare them fresh for an occasional sampJ.";;.
A sample of air to be examined is obtained in a 250-ml. bottle, using an aspirator bulb and discharging the bulb at least 25 tirhes into the bottom of the
bottle. The bottle is closed with a rubber stopper ot with a well-vaselined glass
stopper if preferred. Such a sample will keep weIl.l When ready for the test,
add to the bottle 2 ml. of the 1-20 CO-free blood dilution, avoiding as much as
possible the escape of the air-gas mixture from the bottle. SIOWlY~Urn the
. bottle so that the blood solution flows over the surface of the inside of t e bottle, .
avoiding violent shaking. After 15 or 20 minutes' rotation, transfer t e blood
solution to a test tube of the same size as the standards, add 40 mg. of the
tannic-pyrogallic mixture, invert for mixing as with the standards, and, after
~tanding 15 minutes, compare with the st~ndards and determine the approximate saturation of the blood solution. This may be done with an accuracy of
5% by estimating the degree of color as equal to a standard tube or between
two such tubes. The following tables gives the approximate percentage of CO
in the air corresponding to the percentage of saturation of the blood when the
temperature at the time of making the test is not far removed from 20 C. and
the air is otherwise of fairly normal composition, say, containing at least 19%
of oxygen.
Saturation %

o....................... .'................... "

10 ..............................................
20 ............ ,.' ........... ' .. , .................
30 ........................ '.....................
40 .............................................
50 ................................... : .........
60 ........................... , .............. : ..
70 .. ' ...........................................
75 ............................ , ................

CO %
0.00
0.007
0.015
0.028
0.045
0.066
0.090
0.152
0.200

Corrections for variations in temperature and composition may be found in

the original (loc. cit.).
DETERMINATION OF

CO

IN BLOOD

Dilute 0.1 m!. of blood of a supposed victiIn of CO poisoning to 2 m!. with

0.03% solution of sodium fluoride or 0.05% solution of potassium oxalate (to
prevent coagulation). Add 40 mg. of tannic-pyrogalli'C 'acid, mixture (1-1),
invert 4 or 5 times and after II} minutes compare with standards -made up as in
the. 'method for CO in air, except that sodium fluoride or potassium oxalate
solution of the above strength is used instead of water in making up the blood
solutions. Report the percentage of saturation with CO.
In general, slight symptoms may be noted 'yhen the hemoglobin of the
blood is saturated to the extent of 25%, while 50% saturation gives violent
'
symptoms and a dangerous condition.
Special Reagents
I

MILLON'S REAGENT

Dissolve 1 part of mercury in 1 part of coli(fuming nitric acid. Dilute

with twice the volume of water and decant the clear solution after 24 hours.

POISONS

1965

FROEHDE'S REAGENT

A solution of molybdic acid in sulfuric acid, prepared by dissolving 5 mg.

of molybdic acid, or sodium molybdate, in 1 ml. of hot pure concentrated sulfuric acid. This solution, which should be colorless, does not keep long.
MARME'S REAGENT

Dissolve 30 g. of cadmium iodide and 60 g. of potassium iodide in 180 ml. of

water.
SELENIOUS-SULFURIC. ACID

REAGENT

Dissolve 5 mg. of selenious acid (H2Se20a) in 1 mI. of concentrated sulfuric

acid.
IODO-POTASSIUM IODIDE REAGENT

Dissolve 5 g. of iodine and 10 g. of potassium iodide in 100 mI. of water.

MAYER'S REAGENT

Dissolve 1.358 g. of mercuric chloride in 60 ml. of water and 5 g. of potassium

iodide in 10 m!. of water. Mix the two solutions and dilute to 100 m!.
.

RUBBER, GUTTA-PEReRA, AN:d BALATA

I

Rubber is obtained from the Latex of certain tropical plants belonging to

the Natural Orders, Euphorbiaceae, Apocynaceae, Urticaceae. The Latex iS~
" milky" juice contained in lacticiferous vessels. It is an emulsoid of rubb
in an aqueous serum and is associated with several other substances such a
resins, sugars, protein. The chief rubber-bearing plants are: IIeve a brasiliensis
(Para), Castilloa, Ficus, and the Landolphias. Guayule rubber is obtained
from Parthenium argentatum (Compositae).
.
The essential constituent of rubber is a hydrocarbon (caoutchouc). It is
a polyprene or polymerized terpene (C1oH1s)n.I The differences in the physical characteristics of various rubbers are probably due to variations in chemical
aggregation, or the value of n, i.e., the degree of polymerization: also the
amount and nature of the associated foreign matter. Harries' work on the
ozonides of rubber indicates the hydrocarbon to be an unsaturated substance
with two ethylene linkages. By bubbling ozone through a solution of caoutchouc in chloroform he obtained an ozonide (CloHlSOS)n, soluble in chloroform.

o-(~,

_....-/CHs

~Xa"CHa

HC"

I' \:\ CH
I
..
I '\. 1 ..
~2 a~ 'Y,q_

H2 C

"01"\ 0

.aHa~'o-J'This splits up on steam distillation

into laevulinic al1.lehY<ie, laevuli- '
nic acid, and a superoxiae:
Th'e formula indicates that the rubber hydrocarbon is an 8 carbon ring comppund with two double bonds and that there are two methyl groups attached
to the ring. The systematic name for the rubber hydrocarbon is " 1.5 dimethyl-cyclo-octadiene 1.5." The term octadiene signifies an 8 C-ring with
two ethylene linkages. The 1.5 at the end refers to'the position of the bonds
in the ring whilst the 1.5 at the beginning of the expression refer;; similarly to
the position of the methyl groups. Other formulae for the rubber hydrocarbon
have been proposed, and the formula is still an open questi~m
1 Many investigators prefer the formulae (C.Hs)y; '(CoI!-~r2)" etc.
Chapter by L. E. Salas.
1966

RUBBER, GUTTA-PEReRA, AND BALATA

1967

The rubber hydrocarbon is capable of forming compounds with certain

"reagents. These compounds are generally addition products (except the
nitrosite, C1oH 16 0 7 N 3). OloH16Br4, OloH I6 0, and OloH 16S 2 The process of
vulcanization is based on the combination of sulfur ,with rubber.
Rubber is a colloid, and is not strictly soluble; it forms pseudo-solutions or
swel~ings. A colloid (suspension or emulsion) is represented by suspended
particles (solid or liquid) which by carrying like (+ or -) electrical charges,
repel one another continuously, and thus refuse to settle by gravity. These
partides exhibit the" Brownian movement."
The addition of an electrolyte neutralizing the electric charges causes the
particles to coalesce (coagulation), whereupon they settle out. A colloid'
solution appears as a mass of particles which, on greater dilution, decrease in
size, disappearing as individuals to form a ,i sol" apparently limpid although
the beam of light as employed in the ultramicroscope discovers solid particles.
The first effect of the " solvent" is to swell the hydrocarbon, forming a
network or sponge.
.
'The principal solvents for rubber are gasoline, carbon-disulfide, petroleumether, benzene and its homologues, carbon-tetrachloride, chloroform, turpentine.
Crude Rubber, as it appears on the market, is the product of the coagulation
of the latex. Ooagulation is brought about by various methods, chiefly by
" smoking" as in the case of wild rubbers from the Amazon and smoked
sheet, or by means of dilute acetic aeid for plantation crepe rubber. In the
latter instance the coagulum produced is passed through rollers t<;> consolidate
it into thin sheets.
In both cases such rubbers contain, besides the hydrocarbon, about 2-3%
of substances soluble in acetone and designated as " Resins" (which contains
various ingredients, including fatty acids). Rubber also contains varying
amounts of protein (2-6%) and sugars (inositol), etc., and of course moisture
and dirt. The chief objects of analysis are:
1. To provide information in order that the manufacturer may exercise efficient

control of the raw material.

2. To test whether the manufactured articles meet specifications.
Examination of Raw Rubber.-The following information concerning a
consignment of rubber may be required:
1. Origin, botanical source, method of collection, process of coagulation, details

of packing and despatch.

2. The gross weight at time of dispatch, gross weight at time of sampling,
average tare, condition of the packages at unloading in dock.
3. The method of sampling is an important point to be decided between buyer
and seller. Storage plays an important part as regards ship consignments, for rubber stored in ships' holds Hear the engine-room would be
much drier than in other situations. (The rubber is also likely to suffer
serious deterioration if overheated.)
4. The washing loss, the content of washed dried rubber, its appearance, color,
odor, and strength after removal from the vacuum pans.

1968

RUBBER, GUTTA-PEReRA, AN:Q BALATA

5. The percentage' nitrogen content of the dried rubber, the percentage of

acetone extract and amount of ash.
6. Quantity of caoutchouc (pure).
7. Results of trial vulcanizations.
Evaluation of Latex.-The price of latex is governed entirely ):>y the actua{
percentage of dry rubber it contains (per cent by weight, or Ibs. per gallon
latex). It is important to obtain a truly representative sample. The most
convenient method for determining the quantity of rubber in latex is the trial
coagulation method. The conditions have become standardized and are stated
thus (see Stevens, I. R. I.' Transactions, Vol. V, 5,1930): "A portion of the
representative sample (10-50 ml.) is weighed out into a porcelain evaporating
basin. To this is added very diiute acetic acid (Y:r1 %) with constant stirring.
A little practice soon indicates the stage to which acidification must be carried
and the basin is then placed on the steam bath. On leaving undisturbed for
about half an hour, the rubber will set to a coherent coagulum from which
the clear serum may be poured off and replaced by clean water. The clot
is then taken to the washing rollers and creped out thinly under a good stream
of water. Drying may be effected in a steam oven, but is preferably performed at a somewhat lower temperature, e.g., 70 C. or even 50 C."
" Revertex "-a concentrated latex introduced by Dr. E. A. 'Hauser-ie
evaluated in a similar manner.
Determination of Ammonia (used as a preservative of latex).-Dilute the
latex with distilled water and titrate with NjlOO acetic acid using an external
indicator.2 Other determinations are rarely required, but examination for
traces of metals such as copper, manganese, cobalt, chromium and iron (which
might cause serious deterioration in the finished product) may be important.
Their detection is carried out on the ash and reference should be made to
special tests for traces of the metals mentioned.
Raw Rubber (wild and plantation) appears on the market as sheets, blocks,
slabs or scraps. Each kind of rubber possesses a characteristic odor due to
the contained "resins," decaying protein mat~er or the material used for
coagulation.
The vegetable refuse, earthy impurities, moisture and soluble substances
(organic acid!?, carbohydrates, tannins, etc.) may in some cases together constitute more than 50% of the whole, and which are remo~ed by the washing
and drying process (residu~l moisture about 0.5%). Plantatiori rubber is more
carefully prepared and arrives clean, and being comparatively dry (as sheet,
crepe, or blocks) requires little or no further treatment. Washed and dried
raw rubber has a density of about 0.91-0.97. It is highly elastic but hardens
when cooled down to 0 C. and softens if heated above 60 C. Above 100 C.
in the presence of air, it becomes gradually converted to a black viscous fluid.
However, rubber may be heated out of contact with air to higher temperatures
and on cooling resumes to a large degree its former condition.
Technical Analysis of Crude Rubber (for factory control and evaluation)
comprises the determination of washing loss, true rubber content, resin, ash.
The sampling of crude rubber must be carefully.c!9 ne"as tinie spent on careful

01

Hydrochloric acid and methyl red indicator may be ~ed.

,RUBBER, GUTTA-PEReRA, AND BALATA

1969

analysis may be wasted by bad sampling, Laboratory samples should not be

less than 20 lbs. One sample should be taken from each case unless the packages are many and are much alike when every third or fifth package may be
sufficient. In this case it is a good plan to wash and dry samples from the
first half of the range and compare with results obtained in the other half.
If .the results for washing loss agree within 1%, take the mean. By using a
long bladed knife or toothless" saw" a wedge-shaped sample having its apex
at the cen~er may be cut; or in the case of a prismatic block the best sample is
obtained by cutting out a quarter. A slice through the center may involve a
large error. It is sometimes found that the rubber contains so little moisture
or is so tacky that cutting may be troublesome. Under these circumstances
the knife must be wetted and the superfluous water removed afterwards. In
sampling wild rubbers containing much debris, a sample of 100 lbs. should be
passed through the factory washing process, but for plantation rubbers which
are homogeneous a sample of 10 lbs. is ample.
If the rubber is very hard it may be soaked (after weighing) for a few hours
in warm water.
After the rubber is washed it should be dried in vacuum and weighed.
Excessive washing in hot water must be avoided for at least two reasons.
The rubber may become tacky and does not release the particles of debris;
and, if vulcanization and mechanical tests are contemplated, false results may
arise due to the deterioration of the rubber.
Fine Hard Para shows about .......... , .................. . 16-18% washing loss
Plantation Smoked Sheet ................................ . 1-1.3%
"
"
Plantation Pale Crepe ........... '........................ . 0.5-1 %
"
"
Sprayed Latex Crepe (Hopkinson's Process) ................ . 8-9%
"
"

ANALYSIS OF RAW RUBBER

Chief Determinations.-Moisture, resin (acetone extract), insoluble, protein, ash, rubber, traces of coagulant.
Molsture.-Place 5 g. of rubber (cu~ into snippets with a pair of scissors)
into a flat porcelain basin (tared) and dry the material in a vacuum oven for
two hours at 60 C. (apply vacuum gradually); or in an ordinary air-oven
heated to 80 C. for three ho.urs or until weight is constant, or by desiccation
in vacuo over concentrated H 2S0 4 This process is to be preferred.
The moisture may also be determined by difference after extraction with
acetone (see next paragraph). Original weight of rubber minus the sum of
the acetone extract and residue from extract gives the amount of moisture.
Resin (Acetone Extract).-Use a Soxhlet extractor or Cottle's (Underwriters) Apparatus. Take a 3-5 g. sample of the washed and dried rubber,
cut up fine or rolled out into a thin strip.
If tacky or liable to coalesce, interleave the material with muslin or hard

1970

RUBBER, GUTTA-PEReRA, AND BALATA

f

filter paper. Charge the thimble with the. sample and the tared flask with
60 ml. pure acetone (freshly distilled over anhydrous KiC0 3, using the fraction
56-57 C.).
'
If muslin is used to enclose and interleave the rubb~r it is well to fold' over
the lower end of the roll and fasten it by a few stitches so that there is no
possibility of any of the snippets or detached particles working out betweerif
the folds.
.
Distil on water-bath or electric hot-plate for 8 hours. Detach tp.e extraction flask, distil the solvent and dry the residue at 100 C. to .constant weight.

I--- - - 8 0 - - - - - J
',/1//Q/P1/;,ns/ons in

mi//imete~s.

FIG. 285.
"

The flask should be inclined while drying in order to facilitate the removal of
the solvent.
The increase in weight of the tared flask = acetone extract.
The residue can generally be easily detached fr~m the muslin and can be
weighed (as a check) after drying in vacuo over H 2S0 4
Figure 285 shows the construction of Cottle's Extraction Apparatus (" Ur,,derwriters" design). The advantages of this extractor over the ordinary
Soxhlet are (1) that it is compact, (2) the reflux coIJ.denser is immediately above
the cup holding the sample, (3) the material being-extracted is continuously

RUBBER, GUTTA-PEReRA, AND BALATA

1971

in the vapor of the solvent at its boiling point and thus does not become chil!ed.
(A possible disadvantage is that the size of the sample is somewhat limited.)
Only glass or metal joints should be used in any extraction apparatus (i.e., no
...cork, rubber, or other extractable substance should. come into contact with the
solvent).

It should be noted that both the extract and the acetone extracted rubber
are liable to oxidation-thefefore the drying should be carried out quickly at
lowest convenient temperature. Any prolonged heating should be avoided.
Weigh immediately.
Brown glass apparatus is sometimes recommended to eliminate the effect/
of light.
.
.
The percentage of tI Resin" is geI1eraIIy characteristic of the brandPara (1.3%), Congo (4-6%), Kassai (4.5%), Borneo (10%), Ceylon Plantation (3.2%), Guayule (average 18%), Pontianak (80%). The character of the
" Resins" is important, as, technically, rubber containing hard resin is preferred. Part of the acetone extract is saponifiable and part contains unsaturated compounds. (Iodine value of rubber" resins "-Para 118, Borneo 31,
Guayule 94.) All rubbers except Para contain optically active" resins" and
this fact has been proposed as a means of detecting rubber other than Para in
a mixture. The optical activity is always dextra-rotatory and it has been
observed that the more optically active, the less. sapon~fiable is the resin.
To Determine the Acid Value of the Acetone Extract.-Take up the weighed
acetone extract with warm neutral ethyl alcohol. Titrate with N/10 alcoholic
KOH, using phenolphthalein as indicator. Calculate the number of milligrams required to neutralize 100 g. of the original rubber (=acid number).
Ash.-Heat 3 g. of the sample in a porcelain crucible of about 40 ml.
capacity. In an asbestos card, cut a hole of such a size that when the crucible
is placed in it, the base of the crucible will protrude 1 cm. below the level of
the asbestos. The crucible is placed on a tripod and heated gently at first
with a small Bunsen flame until volatile material ceases to be evolved. During
the initial stages of incineration, care must be exercised that the volatile
decomposition-products do not inflame (or losses may occur rendering the
results unreliable). Finally the temperature is raised to dull red.heat, and
ignition is continued until all the carbon 'is burnt off. The mineral matter
occurring in raw rubber varies' with the source and mode of preparation.
First quality Pale Crepe ranges from 0.10% to 0.5% ash. The color of the
ash should be noted. If white or pale yellow, and infusihle, silica, lime and
magnesia are possible cons'tituents. If highly ferruginous the ash has a brick
red color. Having now determined the moisture, "resin," ash, and washing
loss, an approximate figure for the rubber hydrocarbon may be calculated by
difference.
"'
Determination of Copper in Crude Rubber
Ash a 20-g. sample over a Bunsen burn~r and fi~ally in a mufRe furnace.
Do not expose to a higher temperature than a dull red heat. Particles of
carbon enclosed by the ash need not be burned off as any portion of the copper
existing as chlorides may be volatilized at the same time. Add 5 ml. of con..'
centra ted HCI and heat the crucible until acid boils. Transfer by means of a

..

1972

RUBBER, GUTTA-PEReRA, AND BALATA

fine stream of water and rubber policeman to a smali beaker. Boil for a'few
minutes.
, I
After the solution has cooled make it alkaline with NH 4 0H. Heaft on a
steam plate until the volume of the solution is not more than 15 mi. (a steam
plate is best for heating as there is often danger .of sp~ttering). By this time
all of the excess ammonia will' have been driven o~. Add a few drops of'
bromine water and heat to boiling. (1) Filter immediately through a Gooch
crucible into a graduated vessel (a graduated test tube which will slide inside
the usual suction filter flask serves the purpose nicely). Dilute the solution
to 30 mI.
Place 10 mi. of the solution, unless the copper is abnormally high (.01 %)
or low (.0005%), in a 50-mi. Nessler tube. If the copper concentration is
high (.01 %) use 5 mi. of solution, and if the copper concentration is low, use
20 mi. of solution. Add 22.5 mi. of 95% pure ethyl alcohol and 2.5 mi. of a
1% solution of sodium diethyldithiocarbamate in 95% pure ethyl alcohol.
Mix thoroughly.
For a standard (2) prepare a similar solution in a Nessler tube using 10 mi.
of water in place of the portion of the sample. Add a solution containing
0.00001 gm. copper per mi. from a pipette graduated in tenths ml. until. an
equivalent color is produced.
, .
Calculation:

eMI. of Standard)X(gm. of Cli'per mI.)X3

=% Cu.
XlOO
wt. of sample
(1') The bromine is added to precipitate the manganese as manganic hydroxide, manganous hydroxide being soluble in ammonium salts. Since as
much as 0.000075 gm. of manganese has no effect on the indicator, the bromine
treatment may be omitted if the amount of manganese in the rubber is known
to be less than .001 %.
(2) The standard Cu solution is made of CUS04 iIll a 5% NH 40H solution,
the strength of which is .00001 gm. Cu per mi. of solution. It is standardized
by electrolysis method using Pt electrodes.
Determination of Manganese in Crude Rubber
Ash a 10 gm. sample in a large crucible. Start the ashing process on an
open flame at about 200 0 C. slowly raising the te:rpperature 0--'600 0 C., then
place it in a muffle furnace at about 1200 0 C. until a white ash results. To the
ash in the crucible add 35 m!. H aP0 4, 35 m!. H 2S0 4 and 30 m!. HNO a and heat
to near boiling until ash is dissolved. Cool and filter through asbestos and
wash filter pad once with HNO a and then place filtrate in a 250-mi. beaker and
heat on steam plate. Add a small amount (0.2 gIn'.) of sodium periodate and
,continue h~ating until constant .color is present.' Dilute with water to a
definite volume (250-500 mI.). Compare this solution with a standard made
up as follows: Dissolve 0.406 gm. of manganese sulfate (MnS04) in 1000 m!.
of distilled water. 1 mi. of this solution is equal to .001 gm. manganese or
.01 %. This solution is used as a primary stanc!,.ar.d 'so great care should be
.
exercised' in its preparation.

RUBBER, GUTTA-PERCRA, AND BALATA

1973

Place 1 m!. of standard manganese sulfate solution in a small beaker (250

m!.) and add 35 m!. H aP0 4, 35 m!. H 2S0 4 and 30 'm!. HNO a, heat almost to
boiling (90 0 C.), and add 0.2 gm. of sodium periodate and continue heating
until a constant co'ior is present. Dilute with water to 100 m!. The standard
is then ready for comparison with the solution of the unknown.
Place an aliquot of 25 m!. of the solution .of the sample in a Nessler tube
and then measure from a burette varying volumes of the standard HMn04
solution and place in other Nessler tubes and dilute with water to the mark.
In this manner make a standard solution to match the color intensity of the
sample.
Calculations:
(m!. of standard) X (gm. Mn per m!.) X (aliquot portion) X 100 =
wt. of sample

% Mn.
0

The Organic Insoluble.'-(This material consists chiefly of protein-like

substances.) There are various methods in use. Weigh out 25 g. of rubber
and mix on rolls.
(a) Take 2 g. of the washed dried rubber in a flask. Add 10 m]. toluene
and heat for 2 hours (under a reflux condenser) upon a sand bath. Allow to
settle and filter, or better, centrifuge for half an hour. Pour off the supernatant liquid, add more toluene, mix and whirl again. The compact insoluble
is received on a tared ashless filter.
Dry and weigh. Incinerate and weigh ash. The difference = organic insoluble (Caspari).
(b) Heat 1 gram of the finely divided sample with 10 mi. of phenetol for
5 hours. Dilute with benzene. Allow to settle, filter, rinse on to a tared dish,
dry and weigh. s
(c) Warm 2 g. of rubber with 15 m!. petroleum (B.P. 200 0 C.) in a small
flask for half an hour or until swollen. Dilute with 100 m!. benzene. Fiiter
or centrifuge as in (a) (Caspari).

(d) Frank recommends that the sample be digested with cymene and centrifuged. When consolidated at the bottom of the tube it can' be washed with
benzene.
(e) Spence & Kratz dissolve the rubber in benzene to which they add 0.5%
trichloracetic acid to lower the viscosity and thus allow the insoluble to settle
out, when it may be washed wi,.th benzene, filtered and weighed.
(j) Tschirch and Schmidt heat the rubber at 800 C. with pentachlorethane
and claim to obtain complete separation of the" insoluble."
Protein (Total Nittogen).-Place 2 g. of the washed dried rubber in a 500 m!.
Kjeldahl flask made of Pyrex or other heat-resisting glass, and add 30 m!.
concentrated H 2S0 4 (or 20 m!. of concentrated H 2S0 4 and 10 m!. fuming
sulfuric acid) together with 0.2 g. of mercury or mercuric oxide. Loosely
stopper the flask with a small glass funnel, and heat at first on a sand bath
and then over a naked flame gradually increasing the temperature to vigorous
boiling until the liquid is light straw-colored and transparent.
This operation generally takes 2-3 hours.
Cool, and dilute with 200 m!. water.
a Beadle & Stevens, Analyst, 37, 13 (1912).

1974

RUBBER, GUTTA-PEReRA, AND, J3ALATA

The solution now contains all the nitrogen in the form of (NH 4 )2S04. F"it
to the flask a distillation tube containing a " splash buib." Measure into 1m
Erlenmeyer flask 20 m!. (N /10) . H 2S0 4 , or more, according to the amount of N
(NH 3 , protein) present. Allow the end of the distillatidn-tube to dip beneath
the surface of the standard acid. Detach the distillatidn-tube from the Kjeldahl flask and add 80 m!. NaO:a; (sp.gr. 1.35) cautiously down the side, keeping
the solutions from mixing until later.
Add a solution containing 2 g. sodium sulfide in 20 m!. water, and then
about 0.5 g. zinc in small pieces.
Now gently whirl the Kjeldahl flask, connect the tube and distil the ammonia into the standard (N/10) . H 2S0 4
When 100 m!. Qf the distillate has passed over, disconnect the tube and
wash the end into the standard acid.
Titrate the excess of (N/10)H 2S0 4 with (N/10)NaOH using methyl
orange as indicator.
Calculate the nitrogen and multiply the value by 6.25 to obtain the weight
of protein in the rubber. (The average N content of protein is 16.5%.
.
Wilfarth's modification of the Kjeldahl process is also recommended for the
total N in rubber. It consists in digesting 1 g. of rubber with 30 g. concentrated H 2S0 4, 7 g. K 2S0 4 and 1 g. anhydrous copper sulfate for four hours Itnd
proceeding as usual.
A blank Kjeldahl should be run on the reagents used.
Instead of distilling the ammonia evolved by the NaOH into N/1OH 2S0 4
as described above, the gas may be absorbed. by cold distilled water and the
amount determined by Nessler's reagent (See page 2049).

As stated previously, the protein mate~ial contaimid in rubber is capable of

being separated from the nitrogen-free material by a solvent method (see under
Organic Insoluble).
, Para rubber generally contains about 0.4% N, corresponding to a protein
content of about 2.5%.
I
The amount of protein depends upon the sOl,nce tind mode of preparati(m'
of the rubber and may amount to 10% and over.
Determination of the Rubber HydrocarbOil.-As previously stated; f<1C:
technical purposes it is sufficient to subtract the sum of the percentages of
washing loss (moisture and dirt) ash, organic insoluble (protein) and acetone
extract from 100 to obtain the percentage of rubber hydrQcarbon. Such a
method is far from ideal but up to the present no entirely satisfactory direct
method of determining the rubber hydrocarbon has been evolved, and those
suggested are generally somewhat tedious.
(a) Spence's Method.-Take about 1 g. of ' the acetone extracted rubber
(vacuum dried). Add 100 ml. cold benzene. Allow to swell, and make up to
200 m!. with benzene. Mix thoroughly and filter through a tube plugged with
glass-wool (or allow to settle).
'
Take an aliquot part of the solution and evaporate to dryness in a current
of CO 2. Weigh.
(b) Precipitation Method.-Weigh out 3 g. of the material and swell it in
10 m!. of toluene. Make up to 100 ml. with, tolnene. Allow to settle (or
centrifuge) .

RUBBER, GUTTA-PERCRA, AND BALATA

1975

Take up in a pipette 50 ml. of the supernatant liquid and allow it to run

into 100 ml. warm alcohol (95%), stirring meanwhile. Remove the clot by
means of a glass rod to another dish, squeeze out, and redissolve in toluene
(100 mI.).
Reprecipitate by alcohol. Squeeze out the alcohol and transfer the clot to
a tared watch-glass. Dry to constant weight, preferably in vacuo. Calculate
per cent of rubber.
If the rubber contains much resin, use the acetone extracted material.
(c) Fendler's Method.-Take 1 g. of the washed dried material. Extract
with acetone for 8 hours. Dry the sample. Dissolve the residue in petroleum
ether and make up to 100 m!. with the solvent. Allow the " insoluble"
settle out. An aliquot part of the solution is taken and evaporated, drying
subsequently in CO 2
(d) Tetrabromide Method.-(Budde's Method, Spence & Galletly's modification.) The value of this method is lessened by the fact that proteins and
resins absorb bromine.
Weigh out 0.2 g. of the original rubber sample. Extract it with acetone
for 8 hours. Place the residue in a wide-mouth bottle, cover it with 50 ml.
carbon tetrachloride and allow it to swell overnight. (The brominating solution is made by dissolving 3 ml. bromine and 0.5 g. iodine in 500 mL carbontetrachloride. )
Add 50 ml. of the" brominating solution" and allow the reaction to continue for six hours with occasional shaking. Add 50 ml. of absolute alcohol, with
stirring. Allow to settle and filter. Wash with a mixture of alcohol and
carbon-tetrachloride and finally .with alcohol. Dry at low temperature and
weigh precipitate. Ignite the precipitate. Difference in weights = Tetrabromide.

Assume the formula C 1oH 1SBr4 (wt.XO.298=rubber hydrocarbon), or the

precipitate of tetra bromide may be fused in a platinum crucible with about 8 gr.
of a mixture of Na 2 CO S and K 2 CO S (equal parts). Heat to clear fusion. Cool
and leach with water. Acidify with HNO s, and add enough AgNO s solution to
precipitate the bromide. Heat gently to agglomerate the AgBr. Filter and
determine weight of AgBr by any usual method. Calculate, 4 Br : C1oH 1S : :
amount of Br in AgBr : x.
The results obtained are often too high (substitution of Br takes place as
well as addition, and the protein in the rubber also absorbs a little Br).
(e) Nitrosite Method (Alexander, Harries).-Dissolve 1 g. of. the rubber
(acetone extracted) in benzene. Pass nitrous fumes (obtained by the action of
concentrated HNO s upon As 2 0 S (or starch)), through a drying tower containing
P 20 5, and thence through the rubber solution. The leading tube should be
wide. Continue the current for one hour.
Allow to settle and decant through a weighed" Gooch," deposit the preCipitate and wash with benzene. Pass the gas again through the filtrate and allow
to stand overnight. Decant through the" Gooch," wash with benzene or
petrol-ether and then with ordinary ether.
Dry at 75 C. and weigh.
The composition of the nitrosite is said to be CloH1607N s.
Wesson and Knorr's Method (as modified by Tuttle and Yurow).-This is

to

1976

RUBBER, GUTTA-PEReRA, ANI? BALATA

claimed as the only nitrosite method yielding sufficiently exact results. T~ke
0.2-0.5 g. of rubber (dried).
j
Extract with acetone for 8 hours. Dissolve the residue in chloroform and
pass the nitrogen-oxide gas until the solution is green, Allow the solution to
stand overnight, and filter through a " Gooch" lined with asbestos. Evapo. rate the filtrate to dryness. Dissolve the residue in acetone and filter.
Take an aliquot part of the solution in a combustion boat and evaporate to
dryness. The residue is ignited in a combustion tube and the products are
led through a train of absorption bulbs (eight in all).
The first three contain a mixture of concentrated H 2S0 4 .and K 2 Cr 207, the
fourth contains zinc dust, the fifth is packed with soda-lime, the sixth with
calcium chloride; the seventh bulb contains H 2S0 4 'and K 2Cr 207 while the
eighth holds a dilute solution of palladium chloride.
.
The fifth, sixth, and seventh absorption bulbs are weighed before and after
the combustion. The increases in weight give the total CO 2 and H 20 formed,
from which the rubber content of the specimen is calculated .
. (For details see I. R. W., 57, 17 (1917).)

ANALYSIS OF RUBBER S.OLUTIONS

Determinations Required.1. Nature and amount of solvent.
'
2. Percentage of rubber..
3. Content of total solids.

The Solvent.-Take an adequate sample and separate the solvent by steam

distillation. Separate the water and the solvent by means of a separatory
funnel. Add a suitable d,ehydrating agent (i.e., fused calci_um chloride) and
allow the solvent to stand over this for a day.
. ~
A fractionation test should be undertaken and the nature of. the solvent
investigated.
To Determine the Proportion of Solids to Solvent.-Take a representative
sample of 10 g. in a stoppered weighing bottle. Remove the stopper and place
the bottle and stopper in a clean porcelain basin and evaporate the solvent in
an electrically heated vacuum desiccator. The difference in weight = weight
of solvent.
Investigation of the residue. Cut it into snippets and extract with acetone.
The dried residue from acetone may be dissolved in acetic anhydride by Morawski's Method and examined for colophony by the.liebermann-Storch reaction
(evanescent violet coloration with concentrated H2S0 4).

RUBBER, GUTTA-PEReHA, AND BALATA

1977

Inorganic fillers are determined as follows: Weigh the sample, dilute with
petrol ether to such degree as. will permit the inorganic fillers to settle. Filter
or centrifuge. Wash the residue with the solvent. Dry and weigh.

ANALYSIS OF VULCANIZED RUBBER

The vulcanization of rubber is brought about by (1) heating with sulfur, (2) dipping
into a dilute solution of sulfuromono-chloride in C8 z (cold process for thin goods),
(3) ,by exposure to the vapors of 8.Ch (vapor cure), or (4) by the simultaneous action
of hydrogen-sulfide and sulfur-dioxide (Peachey Process).
The properties of the resulting vulcanisates vary with the percentage of 'sulfur,
the time of vulcanization and the temperature of the "cure"; also the nature of the
additions to the mixture.
.
If the sulfur is limited to about 10% soft vulcanisates are produced. When the
proportion of sulfur is raised (up to 32%) hard rubbers are formed. The change of the
rubber hydrocarbon to hard-rubber (ebonite) may be represented probably by the following equation. (Although Woo Ostwald is of opinion that the process of vulcanization
is purely physical (adsorption) it is now generally accepted that the main action is one
. of chemical combination.) Vulcanized rubber is characterized by a great increase in
resiliency, a lower temperature-coefficient of its elastic properties and a marked change
in behavior towards solvents. The process of vulcanization is irreversible.

+8: yields on vulcanization--->

A representative sample must be obtained. Remove any extraneous material such as paint, lacquer, etc., by mechanical means. Powdered talc or
starch should be removed by a brush or clean cloth. Bloom (powdered sulfur),
waxes, etc., which have risen to the surface should be included in the sample .
. Metal (wire, studs) should be separated.
I~ the rubber itself is to be examined it should be freed from any applied
fabric, but textile material actually forming part of the mixture should be
included.
If the rubber consists of distinct layers of different composition, these should
be worked up separately.

1978

RUBBER, GUTTA-PEReRA,. AND BALATA

The sample having been isolated, it is cut up into snippets with scissots.
Hard rubber (ebonite, vulcanite) is rasped with a coarselclean file and any iron
i
filings introduced are removed by a magnet.
Unvulcanized doughs and reclaim may be rolled out into 'a thin sheet on
the laboratory mill. The prepared bulk sample shoJld be ,preserved in a
stoppered wide-mouth bottle preferably of brown glass. I
The scheme of analysis is that of Weber and Henriques, with modifications.' .
It is rarely necessary to go through the whole scheme. The appearance of the
sample will give valuable information. If the rubber is Ilgl,lt or red in color it is
obvious that carbon black, pitch, tar, bitumen, and Pb compounds are absent.
In this case, the chloroform-extraction can be omitted.
Moisture.-Take a 2 g. sample and dry it to constant weight in a vacuumdesiccator containing concentrated H 2S0 4 Loss of weight = Moisture. (If
the rubber contains no material likely to be melted out, the sample may be
dried in an air oven at 80 C. or in an electrically-heated vacuum-desiccator.)
Water-Soluble Constituents.-Take a 10 g. sample in a 300-mL flask and
add 100 ml. H 20. Heat to 80 C. and allow the specimen to soak for a few
hours. If present, the following substances will be found in the solution:
Starch (dextrin), gelatine (glue), some aqcelerators (aniline), and part of.
the inorganic fillers (lime, CaSO 4).
Test for alkali; starch by iodine; size (gelatine) by Tannin. Aniline may
be detected by the violet coloration with hypochlorites. Glycerin may be
detected by evaporating a portion of the solution to dryness and heating the
residue with KHS0 4 when the odor of acrolein is obtained.
For other constituents proceed as follows: Take a 2 g, sample and extract
with pure acetone, for 8 hours.
.

,
A. Solution

B. Residue.

Extract with alcoholic KOH (IN)

I

.'

D. Residue contains
C. Solution contains
I
Sulfurized substitutes
Rubber Hydrocarbon
Mineral fillers
Chloro-sulfur substitutes
Oxidized (brown) oils
(Sulfur of vulcanization)
Polymerized oils
(Chlorine of vulcanization)
....
(Sulfur,and chlorine in above)

-.

-.

._

Note.-The mineral fillers are determined on a part of Residue D by ignition, and

the rubber-hydrocarbon is calculated by difference. The Residue D is also used to find
the sulfur and chlorine of vulcanization by method given on pages 1985-1986.
I

Full Scheme for Analysis of Vulcanized Rubber.-If the rubber is black it

may contain one or more of the following substances: Carbon black, pitch, tar,
bitumen, Pb compounds. For a systematic investigation, the following procedure may be followed: (This scheme is based on the work of Weber, Henriques and others.)
..
,..

ROBBER, GUTTA-PEReRA, AND BALATA

Take a 5 g. sample and extract with Acetone for 8 hours.
48 hours.)

1979

(Hard rubber requires

A. Extractmay B. Residue.-Extract for 4 hrs. with chloroform (immediately, withcontain:

1. "Resins"
of Rubher.
2. Added
resins.
3. Mineral

Oils.
4. Waxes.
5. Fatty Oils.
6. Resin Oils.
7. Free Sulfur.
8.M.R.
(part).
9. CelluloseEsters.
10. N aphthalene,
aniline.
11. Accelerators.
12. Antioxidants.
13. Certain
organic
colors.

(See page
1980)

out drying)
(See page 2002)

C. Extract
1.
2.
3.
4.
5.

may contain:
Pitch.
Bitumen.
Tar.
Sulfur in
the above.
M.R.
(part).

D. Residue.-Extract with N Alcoholic KOH.

(See page 1982)
/

E. Extract
1.
2.
3.
4.

F. Residue.-Weigh and extract with a

may
contain:
Sulfur
Sub.
Chloro- G. Extract
Sub.
may
Oxidized
contain:
Oils.
1. Rubber
Sulfur
hydroand Cl
carbon.
in the
2. Sulfur
above.
and Cl
in above.

.
(See page
1982)

(See page
1982)

(See,page
1982)

rubber solvent.
xylene, etc.)

(Petrol,

H. Residue.-Weigh and
extract with boiling
water.
(Wt. of F-Wt. of

H=G.)

I. Extract

K. Residue

may
may
contain:
contain:
1. Starch.
1. Minerals.
2. Dextrine. 2. Free
Carbon.
3. Gelatine.
4. Albu..3. Vegemen .
table
fibre.
4. Sulfur in
minerals.
(See page
(See page
1978)
1988)

Note.-The time required for the Acetone Extraction (and its completeness) depend
upon the apparatus and the speed of siphoning; also upon the rubber and its subdivision.
The siphon should function once in about three minutes.

Inspection of Acetone Extract.-Evaporate the solvent and dry at 90 C.

Note color and viscosity-If .pale and transparent and small in quantity,
absence of gross additions may be inferred. Fluorescence indicates the presence
of mineral-oil. Test for rosin (colo phony) and rosin oil by the LiebermannStorch reaction.
Solid W axes.-Warm a little of the acetone extract with absolute alcohol.
Cool. Waxes deposit.

1980

RUBBER, GUTTA-PERORA, AND BALATA

Detailed Examination of the Acetone Extract

Total Sulfur.-(All reagents used must be free 1 from sulfur.) Dry the
acetone extract and weigh. Take an aliquot part in an Erlenmeyer flask, add
1 g. of MgO and 10 ml. of concentrated HNQa and 1 \ml. bromine. 4
Allow to stand for a few hours. Start heating on a steam bath,-increase
'temperature on sand bath. Add more reagent if required. When oxidation
is complete, transfer to a porcelain dish and evaporate dry on a steam bath.
Add 5 ml. concentrated HCI and evaporate dry; bake gently; .repeat this
operation, Add concentrated HCI and transfer to a flask by means of hot
water. The sulfur is now in solution as sulfate. Filter hot. Make up to 100
ml. with hot water. Heat to nearly boiling and add 10% BaCb solution.
Allow to stand overnight.
Filter the BaS04. Wash with hot water, ignite in porcelain crucible,to
constant weight. Calculate S.
Saponifiable and Unsaponifiable Matter.-Take an aliquot part of the dried
extract 01 g.), add 25 ml. N/2 KOH, and heat under a reflux condenser for
two hours. Transfer to ,a separatory funnel. Cool to room temperature.
Add 20 ml. ether and shake carefully. Separate the two layers, and wash
the ether layer several times with cold water until the water no longer has an
alkaline reaction. Transfer the ether extract to a tared flask and evaporate the
ether. Dry to constant weight at 80 C. (=unsaponifiable).
The aqueous alkaline solution from'the ether extraction is now acidified
with dilute H 2S0 4 to set free the fatty acids, from the soaps.
Transfer to a separatory funnel. Shake out with ether as before and repeat.
Collect the ether fractions in a clean separatory funnel and wash free from acid
by repeated shaking with cold water.
Transfer the ether solution of fatty acids to a tared flask, evaporate ,and
weigh, (=saponifiable matter, chiefly resins and fatty acids).
If the rubber should contain cellulose-esters they will be dissolved by the
acetone and will cause the acid aqueous liquid to b,e strongly turbid, anli if
suspected the fluid is neutralized by ammonia and evaporated to a paste,
heated with 10 ml. " cuprammonium solution" and allowed to stand for 12 hrs.
in a covered vessel, shaking periodically.
'
The cellulose will dissolve and a clear solution will be obtained. Acidify
with HCl. Prepare a Gooch lined with asbestos. Filter, wash, dry, and
weigh the cellulose. (For separate determination of cellUlose see page 1987.)
The Unsaponifiable Matter in the Acetone Extract May Contain:
1.
2.
3.
4.
5.

Substances occurring in the original rubber.

Mineral oils.
Solid paraffin-hydrocarbons (paraffin-wax, ceresin, etc.).
Tar and asphalt.
Wax alcohols.

Take a known quantity of the unsaponifiable matter and extract it with hot
absolute alcohol. The residue consists of tar and aspha~t. Evaporate the
4 If certain modern accelerators have been used.:a_, ;hr~mine-nitric acid oxidation,
followed by perchloric acid, is a more effective oxidizing -treatment ("Captor," etc.).

RUBBER, GUTTA-PEReRA, AND BALATA

1981

alcoholic solution to 50 nil. Cool the flask and then place it in a freezing mixture capable of producing a temperature of minus. 5 C. The solid paraffins
will separate out. Filter thrOl,lgh an iced funnel and wash with ice-cold alcohol.
Remove the funnel from the ice-pack and dissolve the residue through the
filter with hot cMoroform into a tared flask. Evaporate the solvent and dry to
constant weight (=solid paraffins).
The alcoholic solution remaining after freezing out the solid paraffins is
evaporated to dryness on a water-bath. Dissolve the residue in CC1 4 (15 m!.)
and transfer the solution to a separatory funnel. Add 15 m!. concentrated
H 2S0 4 to the funnel and shake carefully. Separate and repeat till the H 2S0 4
extracts no more colour.
Draw off the acid. To the CC1 4 solution add distilled water and then enough
ether to dilute the CC1 4 to make it float above the water. Shake and repeat the
washing until the" CCl 4-ether solution" is acid-free. Transfer the" CCl 4-ether
solution" to a tared flask and evaporate. Dry constant at 90 C. This'
gives the weight of the liquid paraffins, and mineral oils.
1he Saponifiable Matter in the Acetone Extract comprises:
Fatty acids.
Resin acids.
Added Resins (e.g. Colophony).
Take the aqueous portion from the separation of " Saponifiable" and" Unsaponifiable." Acidify with HCl. Shake with ether. Wash the ether free
from acid. Filter the ether solution 'and wash with the solvent. Distil the
ether and dry the residue for one hour at 95 C. This gives the weight of the
fatty acids, obtained from the saponifiable oils.
To determine resin (colo phony) use Parry's Method:
Take part of the original acetone-extract or the saponifiable above. Dissolve in 20 ml. of 95% alcohol. Add dilute NaOH (made up with 1 of alcohol
+2 of H 2 0) until just pink to phenolphthalein. Heat for 10 minutes. Cool.
Transfer to a lOO-ml. graduated stoppered cylinder. Fill to the mark with
ether, and add 2 g.AgNO a Shake and allow to settle. Silver salts of abietic
acid and other organic acids form. The silver salts of the free fatty acids are
insoluble in ether, but those from colo phony (abietic acid) give silver salts
which are soluble. Therefore talfe 50 ml. of the clear solution in a separatory
funnel and shake with 20 ml. dilute HCI. Draw off the ether layer by means of
a pipette and shake twice with ether. Unite ethers and wash. Distil the
ether solution in a tared flask. Residue = Colophony. Dry at 105 C. and
weigh (X by 2).
The Detection and Determination of Accelerators is very difficult and generally impossible owing to the complex nature of the accelerators themselves,
the small amounts used, and to the great changes they undergo during vulcanization.
Some disappear"altogether (i.e. isopropyl-xanthates).
Evidence of the use of organic accelerators may be inferred by the enhanced
mechanical properties, the sub-normal sulfur content, and low vulcanizationcoefficient. For further information see Twiss and Martin (Gummi Zeitung,
3S, 1151 (1921.

1982

RUBBER, GUTTA-PEReRA, AND BALATA

Examination of the Chloroform Extract

Evaporate the extract on a steam bath. Dry to cbnstant weight in an airoven at 95 C. If the extract is fluorescent, or of c<;>lor darker than yellow,
bitumen is present. Part of the asphalt and tar may lalso be found here.
Dissolve the dry extract in benzene (warm). AUow to stand for 12 hrs.
Filter, wash several times with benzene.
I
The substances insoluble in benzene are called" Hard Asphalt." Itrould
be borne in mind that soft vulcanized rubber always yields about 1 % to chloroform. Sometimes it is soluble to the extent of 2%.
If the figure 1% is exceeded and bituminous bodies are shown to be absent
(light color of extract) it is certain that the rubber has dissolved in excessive
amount. (This indicates (1) that the material is undercured, or (2) contains
"reclaim," or (3) has been deteriorated in manufacture by excessive milling,
Jr (4) that the rubber has undergone excessive ageing.)
If the extract is brown, asphalts or " M.R." are present.
Asphaltum is much used in cable coverings to remove microporosity, and
to increase insulation.
The Alcoholic KOH,Extraction
This reagent removes the main part of the fatty oils, the white and brown
" factis " (oil-substitutes). It also dissglves a portion of the natural albumin
(casein, and wool if present). If phenol-resins are present they will also be
affected.
The residue from the chloroform is dried at 50 C. and transferred to a 200
ml. flask. Add 50 ml. benzene and allow it to stand 12 hours. Add 50 ml.
N alcoholic KOH and boil under a reflux condenser for 4 hours. Filter and
wash. Grind the residue in a mortar, leach with hot absolute alcohol, and
finally with hot water. The washings are combined with the filtrate and
evaporated dry. Add H 2 0, and transfer to separatory funnel. Acidify with
! '
-,
dilute H 2S0 4
Allow to cool, add ether, and shake. Separate th~ ether and aqueous layers,
and repeat. Combine the ether fractions in ;another funnel and wash with
cold water till acid-free. Add the" waters" to the aqueous fraction. Evaporate the ether in a tared flask to constant ,weight. Calculate per cent of
residue (= weight of fat~y acids). "Total unsaponifiabl~" minus the sum of
solid and liquid hydrocarbons gives the unsaponifiable resinous-matter.
The Determinations of Sulfur and Nitrogen in the Alcoholic KOH Extract
are made by the same methods as described for sulfur in acetone extract.
Determination of Total Fillers.-Incineration of the original rubber in an
open crucible (as described under Ash, page 1971) forms the simplest method,
but this method fails if any of the fillers are deco.mposed with loss of volatile
matter (i.e. HgS, CaCO g , free carbon, cellulose~ etc.): HgS is completely
volatilized, carbonates evolve CO 2, and sulfides change to oxides.
Take r g. of rubber. Extract with acetone, and chloroform as usual.
Dry the residue and transfer it to a 200-m!. flask (tared). '/
Add 25 ml. petroleum (B.P. 230 C.). Hea~'under a reflux condenser,
until the rubber is in solution. Cool the flask j(nd nearly fill it with benzen~.:

RUBBER, GUTTA-PEReRA, AND BALATA

1983

AllDw it to. stand until the fillers have settled. Decant and filter thrDugh a
GDDCh. Wash with hDt benzene, alcDhDI and ether. Dry the crucible at
90 0 C. to. CDnstant weight.
A centrifuge will save much time.
If an autDclave is available the sample may be heated with xylDI under
pressure.
DeterminatiDn of Sulfur

The. sulfur in rubber may Dccur in variDus fDrms. 'The determinatiDns

usually required are tDtal sulfur, free sulfur, sulfur Df vulcanizatiDn, sulfur
cDmbined in the fillers (Drganic and inDrganic).
During the. analysis sulfur appears in the variDus extracts and residues.
1. Total Sulfur (J. RDthe's MethDd).-Take 1 g. Df the rubber under test
in a rDund-bDttDm Pyrex flask. Add 30 m!. cDncentrated HNO a anti 1 g. MgO.
Heat upDn a steam-bath until fully Dxidized. Heat mDre strDngly upDn a
sand-bath to. expel brDwn fumes. 5 (If carbDn shDuld remain repeat the DxidatiDn with HNO a and heat again.) Add 5 m!. HNO a and dilute with 30 m!.
H 20, Filter and reserve the residue.
Filtrate.-EvapDrate to. dryness in a pDrcelain dish and bake the residue
for 2 hDurs at 135 0 C. MDisten with HNO a Add H 2 0 and evapDrate dry.
Add 5 ml. cDncentrated HCl. Heat and dilute to. 100 ml. with H 2 0. Filter
and wash residue. BDil the filtrate and add 10% BaCl 2 sDlutiDn. AllDW to.
stand (lvernight. Filter Dff the BaS04; and wash with distilled water. Dry,
incinerate and weigh. (BaS04 XO.1373 = S.)
.
Residue (abDve).-Fuse in.a crucible with abDut 5 g. Df a mixture Df
Na 2CO a and K 2CO a (equal parts). When clear, allDw to. CDDI, leach DUt with
hDt water and bDil the liquid. Filter.
Filtrate.-Acidify with HNO a EvapDrate to. dryness, bake at 135 0 C. fDr
2 hDurs. Add cDncentrated HCI. Heat and dilute with H 2 0. Filter Dff any
silica Dr antimDny Dxide.
EvapDrate the filtrate to. dryness, add cDncentrated HCl and evapDrate
again. Add dilute HCI and bDil; filter. Add 10% BaCl 2 and calculate sulfur
as stated abDve.
(Steven's Method, Analyst, 43, 1918)

...

0.5 g. Df the rubber sample is digested with 20 ml. HNO a (1.42). Add 0.5
g. KClO a
BDil fDr 3 hDurs under a reflux cDndenser. EvapDrate dry in a pDrcelain
basin with the additiDn Df 3 g. Df magnesium nitrate. Heat cautiDusly Dver a
naked flame. (The Mg salt mDderates the cDmbustiDn.) Any unburnt carbDn
is destrDyed by digestiDn with HNO a and KCIO a Expel excess Df acid. Add
10 ml. HCI (cDncentrated) and CDver the dish with a cIDck-glass. Gently heat
until the red fumes cease to. be eVDlved. Dilute the liquid, filter, and make up
to. 300 ml. Heat to. bDiling pDint and precipitate BaS04 with 10% BaCI 2
AllDW to. stand Dvernight. Filter and weigh.
5

See note on page 907.

1984

RUBBER, GUTTA-PEReRA, AND BALATA

(Kay and Sharp'!? Method)
j

This rapid method is carried out as follows:

0.25-0.5 g. of the finely divided sample is intimately mixed with 5 parts
by weight of pure ZnO and 4 parts of KNO a i~ a porcblain crucible and a thin
layer of ZnO as a cover. Put the lid on the crucible ~nd heat gently until the
reaction commences; then remove the flame for a time. Subsequently heat
with the full flame for five minutes. Cool. Dissolve the contents of the
crucible in dilute HCI and precipitate the sulfur as BaS04 as usual.
Bromine-Nitric Acid Method.-Method of Waters & Tuttle (Ind. Eng.
Chern., 3, 734 (1911. Take a 0.5 g. sample in a large porcelain crucible. Add
15 ml. of concentrated HNO a saturated with bromine. Cover the crucible and
allow to stand for 1 hour. Heat for one hour on the. steam bath: Remove the cover, rinse/and evaporate dry. Add 3 ml. concentrated HNO a Cover, warm
for 15 minutes. Allow to cool and add.5 g. Na 2 CO a in small portions down the
side, and replace COVer. Stir. Dry on steam bath and spread the mixture on
the sides of crucible. Fuse and destroy all carbonaceous matter~ Cool the
melt and place the crucible and its contents -in a 400-ml. beaker. Cover with
125 mI.. H 2 0. Digest for 2 hours. Filter. Wash. the residue. To .the filtr.ate add a few mls. of concentrated HCl. Cover and heat. Precipitate tlie
BaS04 as usual. (Factor for sulfur = 0.1373.) The above methods do not
distinguish between-free sulfur, sulfur co:mbined with the rubber hydrocarbon
(sulfur of vulcani~ation) and sulfur in the fillers.
If ba;ytes (BaS04) is the only sulfur-bearing filler, the zinc oxide-nitric
acid method is correct.
Zinc Oxide-Nitric . Acid Method for Total Sulfur (excluding sulfur in
barytes).-Take a 0.5 g. sample in 500-ml. Erlenmeyer flask. Add 10 mY.
ZnO-HNO a mixture made by adding 20 g. ZnO to 100 ml. concentrated HNO a
Allow the" test" to stand at room temperature for about 12 hours. And
15 ml. fuming HNO a Do not allow ignition to occur (cool if necessary).
When a solution is obtained add 5 ml. saturated bro~ine water. Evaporate
on a steam-bath to a syrup. If necessary add mo;e of the reagents. Cool
and add KCIO a Evaporate dry. Heat more strongly until all the nitrates
are decompo_sed.
Cool. Heat with 50 ml. concentrated HCl. (Any Ba salts are precipitated
here.) Filter and dilute, to 300 ml. Boil and precipitate with 10% BaC1 2
Calculate S.
.
In the presence of other sulfur-bearing mineral fillers, the solution-method
for separating fillers from non-fillers is employed in conjunction with the bromine-nitric acid method, as follows: Take a ,0.5 g. sample. Extract with
acetone. Heat to 1350 C. (until dissolved) with 25 ml. cymene in a 150-ml.
flask.
Dilute to 100 ml. with petrol-ether. Allow to stand overnight. Filter
through a "Gooch" with slight suction. Retain the residue in the flask,
wash with 10 m!. portions of petroleum-ether, then hot benzene and finally
with acetone. Again treat with the brominating solution.. Evaporate dry.
Add concentrated HCI and evaporate, and again r:ij.Cl. Add hot water, filter
and precipitate the sulfur in the filtrate by 10% 13aCl 2 solution.

RUBBER, GUTTA-PEReRA, AND BALATA

1985

If BaS04 is present, make a fusion and subtract the sulfur in the mineral
fillers from the total sulfur.
Sulfide Sulfur. Steven's Method (Chern. Analyst, 40, 275 (1915.-Arrange
a Kipp Apparatus for C02; and a 250-ml. flask with an inlet tube reaching
nearly to the bottom; a ground-in stopper should be used. The outlet tube
from the flask is connected to two absorption bulbs containing lead-acetate
solution.
Place in the flask 10 ml. concentrated HCI and then 20 ml. ether. Pass a
current of CO 2 through the apparatus until all the air is displaced. Lift the
stopper, and add 1 g. of the powdered sample. Again pass CO 2 (for 30 minutes), shaking occasionally. The H 2S evolved from the sulfides is absorbed by
the lead-acetate. Heat the generating flask to ensure' complete decomposition.
Determine the lead-sulfide in the bulbs and calculate to sulfur. If pure nitrogen
is available for sweeping out the apparatus, the procedure may be simplified
by using NaOH in the absorption bulbs in place of lead-acetate. In this case,
the sodium sulfide formed is oxidized to N a ZS04 by means of bromine. Determine BaS04 and calculate to S.
Sulfate-Sulfur.-Ignite a 1 g. rubber sample in a porcelain boat in a stream
of CO 2 Boil the residue with Na 2CO a solution. Filter, acidify the filtrate
with HCI, and boil to expel S02 from the sulfites. Precipitate the sulfate as
usual with 10% BaCI 2. Calculate S.
Sulfite-Sulfur.-Proceed as described above for sulfate-sulfur except that
instead of boiling off the S02, it is oxidized by bromine and thus converted to
sulfate.
-Precipitate the total sulfate 'by 10% BaCl z solution and calculate S. Subtract "sulfur as sulfate" from the "sulfur as sulfite+sulfate." Difference
= " sulfur as sulfite." Or t,he S02 boiled off may be absorbed in N aOR solution"
oxidized, and precipitated as BaS04.
The Sulfur of Vulvanization can now be calculated. Subtract the various
" sulfurs" determined from the total sulfur-the remainder = vulcanization
sulfur, i.e., the sulfur actually combined with the rubber-hydrocarbon.
If the rubber is a simple" rubber+sulfur " mixture, the calculation of the
sulfur of vulcanization involves the determination of the total sulfur by one
of the methods described and the determination of the free sulfur (sulfur in the
acetone extract). Subtract the latter from the former.
If the sample is a complicated mixture it is necessary to carry out the following extractions in accordance with Weber's scheme of analysis. Take a 2 g.
sample and extract with acetone, chloroform, and alcoholic KOH. This leaves
a residue containing rubber and fillers. Divide this residue into two parts;
on one part find the Total Sulfur; on the other part carry out the solvent
separation of the fillers and determine the sulfur in the latter by a fusion method.
The difference between these values will give the sulfur combined with the
rubber-hydrocarbon.
The sulfur of vulcanization can also be obtained when the nitrosite method
for determining the rubber is used. The nitrosite contains all the sulfur that
was originally chemically combined with the rubber-hydrocarbon.
Take an aliquot portion of the acetone-solution of the nitrosite compound.
Remove the acetone by evaporation and determine the sulfur in the residue

1986.

RUBBER,

GUTT~-PERCHA,

AND BALATA
I

by the method of Waters and Tuttle (Br-HNOa), followed by fusion with

alkali and KN0 3
I
The best way of arriving at the sulfur of vulcanizati.9n is the following:
Add together the free sulfur (the acetone extract), tre sulfur found in the
alcoholic KOH extract, and the sulfur in the inorganici fillers. Subtract the
sum from the total sulfur found by the method of Waters and Tuttle.
'
Determination of Selenium (Method of E. H. Shaw '& E. E. Reid).-Take
1 g. of the original rubber sample. Mix in a crucible with 0.5 sugar, 0.2 g.
NaNO a and 14 g. Na 202. Cover with a layer of Nil 202. Fuse gently. Cool,
and leach out with H 20. Boil until the H 20 2 is decomposed. Filter. Acidify'
with HCI.
The selenium has been oxidized to selenic acid. ,Add c~ncentrated HCI and
cover with a watch glass. Boil. Allow to cool to 90 0 C. Pass a current of
S02 gas (from the siphon) until the red precipitate (Se) has agglomerated,
leaving the mother liquid quite clear. Rinse the watch glass and dilute wi~h
H 20. Filter through a prepared" Gooch." Any selenium adhering to the
sides of the beaker can be removed by alcohol. Wash the precipitate in the
"Gooch," first with alcohol, then with H 20. Dry at 115 C. to constant
weight.
Determination of Chlorine (in cold vulcanization goods or mixtures containing white oil-substitute).-(Henriques' Method.) Take 1 g. of the original
sample in a porcelain crucible and mix with 2.5 g. Na 2CO a and 2.5 g. KNO a.
(These reagents should be examined for chlorine.) Fuse gently and leach out
with hot water. Heat nearly to boiling and filter. Acidify the filtrate with
HNO a. Determine the Cl gravimetrically or by the' following method. Add
25 m!. N/l0 AgN0 3 standard solution. Heat to 80 C. and determine the
excess of AgNO a by means of standard N/10 NH4CNS standard solution,
using ferric-alum as indicator. (See page 271.)
(Carius' Method)
/
Place 0.3 g. of original rubber in a Carius Tube and add 5 m!. fuming HN0 3
and 1 g. of solid AgN0 3 Seal and heat in the!furnace. Open the tube and
empty contents into a beaker.
The residue consists of a mixture of AgCI and insoluble mineral matter.
(Most rubber compounds contain mineral matter insoluble in HNO a.) Heat
the residue with dilute H 2S0 4 in a Pt crucible and add a lew-pieces of pure zinc.
Allow the reaction to continue until the zinc has dissolved. SWoccasionally.
Filter and wash with hot water. The silver has been reduced by the Zn and
the chlorine is in the filtrate. Determine chlorine .by the usual methods.
(See page 265.)
, .
Qualitative Test for Glue (Epstein & Lange's ,Method).-Take 3 g. of the
rubber sample and heat with pure cresol for 1611o~rs at 120 0 C. The rubber
dissolves. Cool and dilute with petroleum-ether. The glue ang other insolubles settle out. Allow to stand overnight. Or the liquid may be centrifuged. Filter on a prepared "Gooch," wash with warm petroleum-ether,
and then with hot benzene. Dry the crucible and contents at 90 C. Remove'
the contents of the crucible to a beaker and boil.lYltl}i'H20. Filter and add a
.
.
solution of tannic acid.

RUBBER, GUTTA-PEROHA, AND BALATA

1987

Quantitative.-The method of Epstein & Lange may be' applied quanti ta ti vely .
Determination of Glue.-This may be calculated approximately from the N
content of the original material after allowing for the natural protein of the
rubber itself. The per cent N ih rubber varies from 0.2 to 0.6. Some knowledge of the source of the rubber is necessary. But if the rubber contains accelerators or antioxidants, which is likely in modern practice, this method
would give erroneous results unless the rubber is first extracted with acetone
for eight hours. In any event the method involves a knowledge of the percentage of rubber in the sample and it is known that some of the N compounds
natural to rubber are removed by acetone.
The Kjeldahl Method is as follows: Extract 2 g. Qf the original rubber with
acetone for 8 hours. Determine the total N content as usual (running a blank
on the reagents, and using the factor 6.25).
Assume the N in the rubber to be 0.4%. Subtract this from the total N
and multiply the difference by 6.25 .
. The N in accelerators is variable; it may run as high as 20%, but is often less;
only about 1% of accelerator is used in a rubber compound. Therefore an
approximate idea of the amount of glue js possible.
Bone-glue is detected by testing the ash for phosphates. A high content
of P. indicates bone-glue.
Determination of Cellulose
Cellulose may be present in two forms.
1. As.the cotton fabric upon which the rubber was spread.
2. As a filler ingredient, in fiber form or cotton watse, etc.
To determine the relation of the tissue to the rubber material and to ascer tain the structure of the fabric as well as the weight of rubber per unit area, a
process is used similar to the solution method for Total Inorganic Fillers- (or
loading).
Take a convenient sample. If it has a thick coating of rubber, the latter
should be removed as far as possible without damaging the fabric; reserve the
rubber. Cut up the remainder of the sample into small pieces and heat in a
flask (use reflux) with petroleum or in an autoclave with benzene or xylene.
Continue the digestion until the rubber dissolves. Filter off the residue, wash
with benzene and afterwards with ether. Dry and weigh.
The material still contains the loading between the fibers.
Heat with NaOH solution. Wash with H 2 0. Heat with acetic acid and
wash with water. Dry to constant weight at 100 C. Weigh.
B'urn off the cellulose material and weigh again.
Loss of weight = cellulose (an"hydrous).
As cellulose matter in the air-dry condition generally contains 7% moisture
add 8% to the result to obtain the weight of the cellulose material as originally
used.
This method of course fails if there are other combustible materials and
volatile inorganic fillers.
Epstein & Moore's Method 6 (B. of S. Tech. Paper No. 154).-Take 0.5 g.
6 See also Standard Methods, page 2009.

1988

RUBBER, GUTTA-PEReRA, AND BALATA

I

of the sample in a flask and add 25 mi. freshly distilled .cresol; heat for four
hours at 165 0 C. under reflux. Cool and add 200 mi. petrol ether, shake continuously for a time and then allow to settle. Decant the Clear portion through
a tared "Gooch." Wash the residue in the flask with hbt benzene and then
with acetone. Add hot 10% HCl to the residue in the flask and transfer it
to the" Gooch." Wash with 10% HCI. Wash thoroughly with boiling H 20
(until free from chloride) and then with acetone until the filtrate comes through
clear and colorless. Transfer to a tared weighing bottle. Dry at 105 0 C.
to constant weight (=A). (This gives the weight of the asbestos pad+inorganic fillers and cellulose.)
(a) Transfer the contents of the weighing bottle to a flask and add 15' mi.
acetic anhydride and 0.5 mi. concentrated H 2S0 4 Digest in the steam bath
for 1 hour. Cool, and add 25 mi. 90% acetic acid. Filter through a tared
"Gooch." Wash with acetic acid and afterwards with acetone. Dry the
crucible at 1050 C. Cool and weigh (=B). Loss of weight (A -B) = cellulose.(b) Or the following procedure.
Transfer the contents of the weighing bottle (at A) to a flask a~ add 25 mi.
of " Cu-Oxam" solution. Allow the mixture to stand for 12 hours, shaking
at intervals. Filter through a tared ,I Gooch." Wash first with" Cu-Oxam,"
and then with dilute ammonia until the filtrate runs colorless. Dry the ~rucible
to com;tant weight (=B). Net weight of the residue A -net weight of the
residue B = C (wt. of cellulose dissolved by tlie " Cu-Oxam.")
The weight of the cellulose may be confirmed by precipitating the cellulose
from the blue filtrate by adding sodium chloride and then dilute sulfuric acid.
The precipitated cellulose is filtered, washed with hot H 20 until free from
chlorides (test washings with AgNO a). Dry the cellulose at 105 0 C. and weigh.,

Determination of the Inorganic Fillers

Modern rubber practice makes use of a formidable li:;>t of inorganic fillers,
in fact it is difficult to mention any low-priced material that has not been tried'
in rubber mixes.
The following list comprises substances that have proved of definite use
and that may be found in rubber material at the present time. Some of these
!,!ubstances are used for specific effects (such as, aids to vulcanization, increase
of tensile strength) whereas others are mere diluents added tQ cheapen the
product.
-..__
All the inorganic fillers are added in finely powdered form. _The most
commonly used are: Zinc-oxide, litharge, lithopone, pumice powder, chalk,
lime, clay, barium sulfate, magnesia, basic mMnesium carbonate, pentasulfide
of antimony.
The following are less common: Mercury-sulfide, arsenicals, red lead, white
lead (basic lead-carbonate), lead-sulfide, ultramarine, lead-sulfate, zinc dust,
ferric-oxide, aluminium-oxide, kaolin, calcium-sulfate, powdered glass, silica,
asbestos, talc. Before a systematic examination of the inorganic fillers can be made the
organic material must be carefully removed. The il!9i.ganic constituents may
then be determined in accordance with systematic' inorganic analysis but the
j

RUBBER, GUTTA-PERellA, AND BALATA

'1989

following will give the general procedure in a rubber laboratory, For the
purpose of separating the fillers, Rothe's Method is recommended, as this
obviates the loss by volatilization of certain ingredients if present (As, Sb, Hg),
(A rough idea of the amount of ash may be obtained by simple ignition in ~
tared porcelain crucible, and this should be carried out as a preliminary.) .
For Rothe's Method t:he amount of rubber taken should be enough to
produce 1 g. of ash. The weighed rubber is placed in a round-bottom Pyrex
flas: of about 300 mi. capacity, Add (per gram of rubber) 15 mi. concentrated
HN0 3 and 2.ml. concentrated H 2S0 4 , Heat on a sand bath for 1 hour and
then more strongly to fumes of S03. Finally heat by naked flame to boil the
H 2S0 4. Cool, and add more HN0 3 (10 mI.). .
Heat until the liquid becomes colorless. If necessary add more HNO a
Cool, and pour the strong acid solution into 50 ml. H 20. Boil the solution,.and
ru~

The residue (A) may contain PbS0 4, CaS04, BaS04, silica, silicates, and
(if Sb is present) antimony oxide and Pb antimoniate, stannic oxide. Pb
may bl:l removed by boiling the residue with several portions of ammonium
acetate (+ NH 40HI solution or the hot ammonium acetate may be passed
through the filter containing the residue. This leaching shou1d be continued
untilltle " washings" fail to contain Pb (test with K 2Cr04 or H 2S).
Wash the residue and reserve it for further treatment (see under Ba).
The "ammonillm acetate-Pb solution" may contain a little calcium. The
calcium in the filtrate from Pb is evaporated to dryness in a porcelain dish.
The ammonium salts are ignited, the residue is dissolved in dilute HCI and
Ca is determined (as on page 210) ..
Barium.-Take the residue after the removal of the PbS0 4 by ammoniumacetate (above). Ignite with a mixture of sulfur and Na 2C0 3 (equal parts).
Apply cine bunsen to the bottom and another to the top of the crucible and
continue the Iheating until sulfur ceases to burn beneath the cover. Cool.
Leach out with H 20. Filter. The residue may contain BaC0 3 (silica and
silicates) and PbS. The filtrate may contain antimony and tin (if present)reserve (B). Take the residue, add hot HCI to dissolve the Ba, Ca, and Pb,
leaving (as insoluble) the silica and silicates. [This insoluble should b!l ignited
and fused with a mixture of K 2 COa+Na 2 C03. Leach with water and .filter.
(If any residue occurs here it .should be tested for bases.) The filtrate from
the alkali-fusion may be discarded as silica is best determined in a separate
portion of the original sample.]
The acid filtrate containing the Ba and Ca is now tested for Pb and, if
foul).d, H 2S is passed through the solution to completely remove it. The
Ba and the Ca are determined in the filtrate by usual methods (see page 122).
The filtrate from residue A may contain Hg, As, Sb, Sn, .Zn, Cr. Fe, Ca,
and alkali metals, etc.
Hg.-Pass H 2S to precipitate Sb, As, Sn, Hg. Filter and transfer with
H 2S to a beaker. (Reserve, filtrate D.) Boil with dilute NaOH-the Sb, Sn,
As dissolve-and filter the i~soluble HgS: Wash the HgS and dissolve it in
'aqua regia. Boil off the chlorine, dilute, and pass H 2S. Determine the Hg
as HgS in usual manner (see page 577).
'.

1990

RUBBER, GUTTA-PEReRA, AND BALATA

As.-Acidify the alkaline. solution containing As" Sb, Sn and pass H 2S.
Filter the precipitate and wash with H 2S water.. Transfer the precipitate to
a beaker and add warm concentrated HCI. The Sb ~nd Sn dissolve. Thoroughly wash the As sulfide. Dissolve it in aqua regia and determine the As as
magnesium pyro-arsenate (see page 96).
'I
Sb.-The solution from the arsenic is now combined with the filtrate B
(above). Add NaOH and bromine (drop by drop) t~ oxidize .all the sulfur
(present as N a2S), Acidify and boil the solution to remove excess of bromine.
While still boiling addferrum-redactum to precipitate the Sb (Sn is not reduced).
Filter through a small filter besprinkled with the iron. Wash with dilute HCI.
Reserve filtrate (C) for Sn. Transfer the Sb to a beaker and dissolve it in HCI
(hot), adding a little bromine. Boil off' the excess bromine. Add KI and
H 2S0 4 to reduce the SbCl 5 to SbCb. Precipitate the Sb as Sb 2S3 and determine
it electrolytically (see page 73). If the amount of Sb is small, determine. it
gravimetrically as Sb 20 4 (see page 72) or colorimetrically by the KMn04
method:
Sn ...,-Take the filtrate C from the" antimony-tin separation." Pass H 2S
gas, filter off the tin-sulfide and determine as Sn02 (see page 959).
Zn.-Take the filtrate D reserved from the precipitation of Hg, Sb, Sn,
and As. Boil off the H 2S. Make alkaline'with NH 40H. Add formic acid.
Pass H 2S, and allow the ZnS to settle out. Filter, and wash with H 2S water
containing a little formic acid. Ignite and determine as ZnO, or by.any other
standard method described under Zinc. (See page 1058.) (Reserve filtrate'
for Fe, AI, Cr. etc.)
.
Fe, AI, Cr.-Boil off H 2S. Oxidize and precipitate Fe, AI, Cr. Separate by
standard methods (see page 467).
Ca.-Take the ammoniacal filtrate from Group III (a). Evaporate, acidify
with acetic acid, boil and precipitate the calcium as oxalate. Determine as
CaO (see page 210).
Mg and Alkali Metals.-Take the filtrate from ~a, acidify with H 2S0 4.
Ignite to decompose the NH4 salts. Dissolve, in 1% H 2S0 4. Filter into a
platinum crucible. Ignite to drive off excess of H 2S0 4. The residue may contain MgS0 4, K 2S0 4, Na 2S04.
Weigh" Total residue." Dissolve'in hot water and precipitate the Mg with
ammonium phosphate. Determine Mg as pyrophosphate (see' page 532).
Recalculate the Mg to sulfate and subtract from the" Total rellidue." Calculate the difference to Na 20.
-.....
Silica and Silicates.-(Powdered glass, pum}ce, kaolin, infusorial earth,
etc.) Silica is most conveniently determined alone on the original material. .
Take 1 g. and fuse in a Pt crucible with about 6 g. of a mixture of K 2C0 3 _
and Na 2C0 3 (equal parts). (Before using a platinum crucib.le for the fusion
of rubber compounds it is always advisable to test for Pb. If lead is. present
use a porcelain crucible.)
Cool the melt. Leach with water. Boil and filter. Wash. Acidify
with HCI. Evaporate the filtrate to dryness. ,Bake at 13$)0 C. Add 5 mi.
concentrated HCI. Evaporate dry, ap.d again bake at 135 0 C. for one hour.
Add 5 mi. concentrated HOI and dilute with H~(y' ~F'ilte,r ,and wash.

RUBBER, GUTTA-PEReRA, AND BALATA

1991

: Ignite in a platinum crucible and weigh as Si0 2 (test its purity as on page
801>.

VERMILION IN RUBBER GOODS

In absence of other metals of second group.7
Weigh out 0.5 g. sample, place in 500-ml. Kjeldahl flask, add 25 ml.
concentrated H 2S0 4, and 3 g. K 2S0 4 Boil over a bunsen burner until solution
is colorless. Bumping may be prevented by adding a few pieces of paraffin.
Cool, and wash into a 400-ml. beaker. Add water until volume is about 250
ml. Heat on hot plate and pass H 2S through hot solution until precipitation
is complete. Filter off the sulfide, wash precipitate back into the beaker, add
200 ml. of water and 2 g. Na 2SOa and boil for 72 hr. to dissolve out free sulfur.
Filter the precipitate through a weighed Gooch, wash with hot water, dry at
100-105 0 C., and weigh as HgS.

HARD RUBBER (EBONITE)

This is analyzed in the same manner as soft-vulcanized rubber.
The sample is rasped to fine powder of which 2 g. are taken and extracted
with acetone for 72 hours. The acetone extract is treated as for soft rubber, i.f).,
find weight of total extract and the sulfur in it is determined as usual. The
difference gives the" organic" acetone extract. Rubber as used for ebonite
contains about 4% resins; therefore if this figure is exceeded, it is probable
that resins have been added and these may indicate that all the added resins
have not entered the acetone extract.-Amber and shellac are insoluble in
acetone, and are frequently used in hard rubber. It is recommended that
the residue from the acetone extraction be extracted with epichlorhydrin
(B.P. 1700 C.) which will dissolve the amber, shellac, copal, and -dammar
resins if present. The extract obtained is evaporated and the weight determined. (Determine the sulfur.) Extract the residue with chloroform, and
determine the extract as usual.
The residue from the chloroform contains the rubber-hydrocarbon combined with sulfur and the inorganic fillers (also sulfur in fillers). Determine
the inorganic fillers by c,areful incineration.
7 Methods of the American Hard Rubber Company, by courtesy of D. E. Jones;
Chief Chemist.
.

1992

RUBBER, GUTTA-PEReRA, AND, BALATA

Determine the sulfur in the residue from the chloroform, and subtract the
sulfur in the mineral fillers. The difference gives the sulfur combined with the
rubber itself (i.e. sulfur of vulcanization). The inorganic constituents are .
determined in the ash in a manner similar to that descriHed for soft vulcanized
rubber.
".
The amount of rubber is found by difference.

RECLAIMED RUBBER (tc REGENERATED RUBBER")

A quantitative determination of the amount of " reclaim" in a vulcanized
rubber compound is impossible, and all that can be done is to make deductions.
For instance, "reclaim" yields a large acetone extract and therefore any
abnormally high v:alue for this extract points to either the use of reclaimed
rubber, or oil surrogates, or added resin. The almost universal use of organic'
accelerators may so lower the vulcanization coefficient of the rubber mix that
any abnormally high sulfur figure for the reclaim itself would be masked.
The examination of " reclaim" is carried out by the ordinary methods o(
analysis for vulcanized rubber.
/
Good quality "Reclaim" should not yield more than 10% to acetone.
Asphalt is found by extraction with chloroform as for soft vulcanized rubber.
(For tests on "Reclaim" see Ind. Eng. Chern., 15, 308-313 (1923) and
Circular No. 232, Bureau of Standards, Washington.)
The following table gives the highest and lowest figures for usual qualities of
commercial whole-tire reclaim:' .
Specific gravity ..................... "
Moisture (max.) ......................
Alkalinity (4 hrs.) ............. " .... "
Acetone Extract.; ...... " .......... "
Alc. KOH Extract .................. "
Chloroform Extract .................. :
Ash ................... " .... " ......
Tensile Strength ........ " .... " ......
Elongation ........ '...................

1.16 to 1.,26
1.0% .
0.15% .
7-;10%
2%
20-28%

1~-25%

600-1000 lbf'!. per sq. inch

300-500% '_,

GUTTA PERCHA AND BALATA

Gutta percha is obtained from the latex of certain plants belonging to the
Natural Order, Sapotaceae, growing wild in the Malay Pepinsula, Sumatra
and Borneo.
.-/The true gutta percha (and the best quality) is obtaIned from the tree,

RUBBER, GUTTA-PEReRA, AND BALATA

1993

Palaquium oblongifolium, although allied species give a similar, but inferior,

product.
Balata is obtained from the tree, Mimusops globosa, growing abundantly
in Venezuela and the Guianas.
'
The hydrocarbon of gutta percha and balata is somewhat similar (but with
very important .differences) to that of rubber. A larger proportion of resin
(acetone-extractable material) is associated with the two former, amounting
frequently to 50%. Gutta percha as extracted from leaves, at Tjipetir, by
the petrol-solution method and subsequent chilling, is almost white in color,
and contains only about 1 % of resin. When cold, the gutta and balata hydrocarbons are hard and tough like leather. They are thermoplastic, while
resilience (as exhibited by rubber) is almost entirely absent.
Gutta perch a is used chiefly for submarine-cable insulation material,
moldings, surgical and chemical utensils (HF containers) whilst balata is a
useful ingredient of belting. Although gutta and balata can be vulcanized
their value is not increased thereby .
. These two materials are therefore used after being carefully washed,
strained and dried. Compounding ingredients are sometimes added, and the
material is often deresina'ted in order to produce harder products.
(For a full account of Guttapercha, see Obach, Cantor Lectures, J. S. A.,
8 (1897).)
As the supply of good guttapercha is becoming restricted, the price is rising ($2.00
per lb. for best quality) and therefore adulteration or substitution is not infrequent.
The price of Balata on the other hand remains about 50 cents per lb. and there seems
at present an ample supply.'
.
Sampling.-It is almost impossible to obtain from a case of gutta percha
a truly representative sample fit for chemical analysis. The material arrives
in lumps of all ,manner of shape and is not homogeneous.
Washing Loss.-The only satisfactory way to arrive at the washing loss
is to wait until the whole consignment has passed through the factory washing
and drying processes. Calculate" washing loss."
Analysis of Gutta Percha and Balata.-The usual determinations required
are: Moisture, resin, dirt, gutta-hydrocarbon.
,
Moisture.-Cut up the sample (a thin sheet from bulk) into snippets and
mix. Take 3-5 g. in a flat dish, and dry in vacuo at 750 C. for six hours; or the
material may be warmed in a current of dry CO 2
Resin.-(Acetone-extract. )
(N ote: Acetone is the best solvent for the extraction of resins from gutta
percha and blilata. Ether dissolves a part of the gutta or balata, thus giving
high results for" resins," whilst with alcohol the gutta hydrocarbon becomes
thermoplastic on account of the relatively high boiling point of the solvent,
the snippets coalesce and the extraction is therefore incomplete.) Take a 2 g.
sample of the snippets and extract with acetone for 6 to 12 hours by the method
described for raw rubber (page 1969). Evaporate the solvent, and dry the
extract at 1100 C. to constant weight. Both the extract and the residue from
acetone are liable to oxidation. Therefore the drying is preferably performed
in a vacuum desiccator, heated electrically; or in an inert gas.

1994

RUBBER, GUTTA-PEReRA, ANI? BALATA

The resins of gutta percha and ba~ata are a mixture of white and yellow
resins. The white resins (Albanes) are soluble iri hot alcohol (not in cold).
They are crystalline and have a composition approximating to C 17H 260. The'
yellow resins (Fluavils) are amorphous and are solublelin cold alcohol as well
as hot.
This difference in solubility affords a method of separation if required.
Gutta (and Balata) Hydrocarbon.-(Rapid Method.) Boil 1 g. of the sample with 20 m!. redistilled toluene. Heat to 100 C. on a w.ater-bath until
dissolved.
.
Pour the solution into 50 ml. alcohol (95%). Allow the contents of the
vessel to stand for a few hours. Warm and agitate to clear the liquiq. Remove,
redissolve in toluene and reprecipitate by alcohol. Dry the hydrocarbon in
vacuo and weigh in a closed weighing bottle. The hydrocarbon may also be
calculated by difference as for rubber-hydrocarbon.
Dirt.-This is generally under 0.5% for washed and strained material af.ter
the factory cleaning process. The dirt consists of both organic and inorganic
debris. The inorganic matter may be determined by incinerating a 1 g. sample
of the original material.
The" organic" debris consists of fine particles of' bark. If this determination is required a solvent method must be employed and when the residue is
obtained it should be weighed (organic and inorganic dirt) and ignited to
obtain mineral (weigh). Subtract the latter from the former weight and thus
determine the organic debris. The" dirt" should be examined for ZnO,
litho pone,' antimony sulfide and carbon.
Obach gives the following range as typical of Gutta Percha.
Water .......... " . . . . . . . . . . .. .. . . . .

Dirt ................................

~~~:ocarbon}

1-1.5%
3-5%

.................... " 30-83 %

Alban. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-44%
Fluavil ........................... " /3-21 %
Ash, max........................... ./ 0.5%

The average composition of balata (ordinary quality) is as follows:

Wate"' ............ "......................

7%

Dirt ................................... 10%

Resins .. '" ............................. -39%
Balata
}
()f ---......
Hydrocarbon .. ; . . . . . . . . . . . . . . .. . . . . . . .. 43 10

After washing:
Balata Hydrocarbon. . . . . . . . . . . . . . . . . . . .. 52%

~:t~:."

. . . . . . . ::::::::::::::::::::::::: 4r~

Separation of Balata and Rubber, or Gutta Percha and Rubber.-The gutta

(and balata) hydrocarbon is insoluble in petroleum ether below 0 C., whereas
the rubber hydrocarbon and resins remain soluble. This ,fact (which forms
the basis of the process for producing leaf gutta. at.!rjipetir, Malay) is used as
the means of separation required.
-

I)

RUBBER, GUTTA-PEReRA, AND BALATA

1995

The rubber containing balata and (or) gutta percha is dissolvedjn warm
petroleum-ether. The solution is then chilled to about minus 160 C. The
balata (gutta.percha) hydrocarbon is precipitated as a white flocculent solid.
Charleson's Method for separating gutta perch a (balata) hydrocarbon from
rubber (~s in compounds for golf ball covers). Mix the sample on the mixing
rolls and then make a very thin sheet. Of this take about 3 g. and dissolve it
in a flask containing 200 m!. of petroleum ether. Warm on a steam bath with
frequent agitation. When the whole is in " solution" allow the flask to cool
to room temperature and then place it in a freezing mixture capable of producing
a temperature of -12 0 C.
The balata (or gutta) hydrocarbon is precipitated, while the rubber and
resins remain in solution.
Filter the contents of the flask through a funnel surrounded by a freezing
mixture and wash with ice-cold petroleum ether. Collect the precipitate and
dry it in an electrically heated vacuum desiccator. Weigh the dried residue.

PHYSICAL TESTS
As previously mentioned chemical analysis of rubber (raw and manufactured) should be supplemented by physical tests. Analytical determinations
are of value for the evaluation of raw rubber, yet may be inadequate for judging
quality and technical value unless accompanied by mechanical and vulcanization tests (i.e., rate of cure and quality of the vulcanizate).
The mechanical tests carried out include the determination of the stressstrain rel_ations (modulus) on stretching and compression hysteresis (repetition
of stretch and 'release). (The technique of the tensile tests, etc., are fully discussed by G. S. Whitby in " Plantation Rubber and the Testing of Rubber"
and deals in detail with the interpretation of results.)
The technique of vulcanization tests involves a "type-mixture" and the
standardization of the conditions of vulcanization so that any new consignment
of rubber may be compared under identical conditions in different laboratories
or be compared with a known quality.
Considerable care must be exercised in the preparation of the sample for test.
A standard temperature (70 0 C.) should be maintained upon the masticating
rolls. The' nip' should be fixed at a standard distance according to the size
of the batch, and a constant temperature is set for vulcanization (press or
autoclave).
The test-pieces may be either dumb-bell shaped or rings.
For mechanical testing, rings are often preferred although there is difference
of opinion respecting their relative value. The rings are generally used in a
'Schopper Machine' which possesses two rotating cylinders whose distance
apart may be increased gradually while the pull (load) and elongation are pro-

1996

RUBBER, GUTTA-PEReRA, AND BALATA

gressively recorded autographically, thus producing a 'stress-strain curve.

The advantage of ' rings' over the dumb-bell shaped te!lt-pieces lies in the fact
that the difficulties incurred with- grips (which slip, or tear the sample) are
obviated. Also, as the tensile strength and elongation at the moment of
b~eak are less important than the stress-strain Curve throughout the stretch
an autographic recording device is essential and this iS I provided on "the ringtesting machines. Nevertheless, errors due to rings exi'st due to the fact that
although the stress measured by the ring-test is the average, the elongation is
that of the inner edge only. But a more serious error may arise with testrings cut from calendered sheet which possess a distinct grain and shows its
greatest tensile strength in the direction parallel to this grain. Rings will
break at their weakest point (across the grain) whereas "dumb-bell" testpieces may be cut so as to enable mechanical tests to be carried out in any'
direction.
Ageing Tests may be conducted in a Geer oven, or by the method of Bierer
and Davis, which provide accelerated ageing under scientific conditions.
Abrasion Tests (on tire-tread material, etc.) may be carried out by running
a ball (cut from the material) in a V-shaped groove under a given load at a
fixed speed. The ball is tested to destruction. .
Permanent Set indicates the quality of the" cure." Properly vulcanized
rubber should show no permanent set after, a reasonable period of rest (10
minutes). The test is made by measuring the distance between fine gauge
marks upon the narrow portion of the dumb-bell test-piece, before and after
stretching.
.
The stress-strain curve of rubber is unique and is exactly opposite in charl;lcter to that of metals which follow Hooke's Law. Schidrowitz has demonstrated that the hysteresis loops are characteristic; strong firm rubbers show
long fiat loops which become constant after the second cycle, whereas soft
rubbers give upright loops of small area.
Attempts to determine quality numerically are not exact, but lead to
useful comparisons. Two" characteristics" are fre<juently used, viz., the
" Tensile Product" and" Slope."
.
The Tensile Product of a vulcanizate is a number obtained by multiplying
the tensile strength at break by the ratio of final length at break to the original
length of the test-portion. (Thus, in metric units, if the ultimate strength is
1.3 kg. per sq. mm., and the final length between the gauge-marks is 700%
of the original, then the Tensile Product is 9.1.)
T-he Slope is the rate at which increasing load increases the extension and
represents the resistance to stretching at high elongations.
The Friction-Pull Test measures the adhesion existing between the layers
of rubber-coated fabric. Two standard methods are employed.
(1) The dead-weight method: strips 1 in. wide are mounted in a clamp.
The ply is separated a sufficient distance to enable '''eights.to be Joaded until
the layers begin to separate. The load required to cause a separation of 1 in.
per minute is taken as the" friction-pulL"
(2) Test-machine method: The ends of the separated plies are gripped in
opposing jaws which are .drawn apart at the rate _9L2- in. per minute. The
friction-pull is recorded on a dial, or autographically.-

RUBBER, GUTTA_:PERCHA, AND BA:DATA

19'97

The conditions laid dowp' for physical-tes~ing ,and recQmmended by the

Rubber Div. of the Amer. Chem. Soc. to be adopted as standard by all Testing
Laboratories are as follows (see Rubber Chem. and Tech., 3, 179 (Apr. 1930.
Mill Data: Rolls 12 in. X6 in:; working di~ta:rice between the cheeks 1072 in.;
speed of slow roll=24 r.p.m.; gear ratio 1 : 4; temperature of rolls during
mastication = 70 C. The" nip" will vary with the volume of the batch
(i.e. for .800 ml. use a nip Ys in. wide, etc.).
.
Mixing: All ingredients'to be weighed to 0:25%. During mastication the
" nip" shall be .055 in. until the rubber runs smooth-then. open out as volume
mcreases.
Order of Adding Ingredients: Accelerators, antioxidants, blacks, fillers, .
softeners, sulfur.
'
Storage of Batch: Stack for 24 hours on a galvanized-wire netting, in the
dark at 27 C. and a relative humidity of 45%. (Variation in humidity'causes
great variation in tensile stren~th.)
,
Preparation of the Stock for Curing: Cut with a die Ys' in .. less in length and
widtll than the mold cavity-the grain of the material should run parallel
. . '
to the length of the m o l d . ,
Curing Temperature: The temperature inside the mold should be registered
by means of a thermocouple. The platens should be bored and good drainage
should be secured to avoid cold spots.
.
Time of Cure: This should start from the time the hydraulic pressure reaches
the prescribed maximum, and end at its release. The mold should be preheated for 20 minute and should be shielded by felt during the cure .
.Molds: No" preparation" of any sort need be used to prevent sticking.
The molds may be cleaned when necessary with ground emery and water.
Cooling of Cured' Test-pieces: Place the specimens immediately in cold
running water to cool for ten minutes. When cool, wipe dry and store on
wire-net in the dark.
Die-ing of Specimens: Wet with water and cut to standard size (dumb-bell
shaped specimens are recommended).
Temperature of Mechanical Test: 27 C.2 C.
Speed of Test: A Scott Testing Machine is recommended, travelling at 20
inches per minute.
Results: Breaks outside the narrow straight portion of the specimen are
discarded. Take the highest figure together with those within 5% and take
the average.
BmLIOGRAPHY
R";bber Analysis.
"Analysis of Rubber" (J. B. Tuttle).
"Indiarubber Laboratory Practice" (W. A. Caspari).
"Methods of Analysis of Raw Rubber" (D. Spence).
"Principal Methods of Chemical Analysis as used at the Netherlands Govt. Rubber
Inst."
"Rubber Goods. Methods of Physical Tests and Chern. Analysis" CU. S. Fed.
Specus. Bd.).
"Testing of Rubber Goods" (Bureau of Standards Circular No. 38).
"Chemistry Indiarubber" (Weber, 1926).
"Analys~ des Kautschuks, der G. P. und Balata" (Ditmar).
"Kautschuk'und seine Priifung" (Hinrichsen u. MemmIer).
Am. Soc. Testing Materials, Standards, Vol. II, pp. 1249-1270 (1936).

1998

RUBBER, GUT'rA-PERCHA, ANI? BALATA

Physical Testing of -Rubber.

.
' .
" "Plantation Rubber and the Testing of Rubber" (G. S; Whitby).
"Mechanical Properties of Calendered Sheets" (Van R9ssem).
"Testing of Rubber Goods" (Bureau of Standards Circular No. 38).
"Report of the Physical Testing Committee of the Divi$ion of Rubber Chemistry"
(Amer. Chern. Soc.) or Ind. Eng. Chern., May, 1925.
"Report of the Raw Rubber Specus. Committee, Rubber Division" (Amer. Chern,
Soc., Sept. 10, 1926).
I"A new mechani<;al test for Rubber Insulation" (Hippensteel).
.
"Kautschuk und seine Prufung" (Hinrichsen u. Memmler).
General Reference on Rubber.
"A Systematic Survey of the Rubber Industry" (Bedford and Winkleman, 1923).
"Estate Rubber; its preparation, properties, and testing" (DeVries, 1920) . .
"Rubber" (Dubose and Luttringer, 1918).
.
.
"The Preparation of Plantation Rubber" (Morgan and Stevens, 1929).
"Crude Rubber and Compounding Ingredients" (H. C. Pearson).
"Chemistry of Rubber Manufacture" (L. E. Weber, 1926).
"Chemistry of Rubber" (Luff, 1923).
"Technology of Rubber" (GottJob, 1927).
"Manufacture of Rubber Goods" (Hei! and Esch, 1919).
"Synthetic Rubber" (Schotz, 1926).
"Recent Progress in Rubber Chemistry and Technology" (Schidrowitz).
_ "Latex" (E. A. Hauser-Gow Lectures). Colloid Chemistry of Rubber.
"Water Absorption of Rubber" (Lowry and Kohman-J. Phys. Chern., Jan.,
1927).
"Electrical Properties of Rubber" (McPherson, Curtis & Scott, U. S. Bureau of
Standards, Sci. Paper No. 560, 1927). .
"Chemistry and Technology of Rubber ': (Davis and Blake, 1937).

RUBBER COMPOUNDS AND COMPOUNDING

INGREDIENTS 1
The analytical side of rubber chemistry falls naturally into two divisions:
1. Analysis of rubber compounds, either vuicanized or unvulcanized.
2. Analysis of raw materials, including rubber, softeners, fillers, etc.

Most of the an;:tlyses in the first class require special te~hnique, and while many
of those in the second class are common in other lines, the proper interpretation
of analytical results involves some knowledge of rubber chemistry and compounding practice.
Samples of rubber compounds are usually analyzed for one of four purposes:
1.
2.
3.
4.

To
To
To
To

check conformity to chemical specifications.

determine the composition for purposes of duplication.
determine the state of "cure" of a compound of known properties.
check the proper weighing or mixing of a batch.

Many methods have been published for the various determinations required,
but after extensive conferences and testing, the following methods for the comnion determinations have been recommended by the Division of Rubber
Chemistry of the American Chemical Society and the A. S. T. M.2 Other
methods and special determinations will be found in a subsequent chapter. "

STANDARD METHODS FOR THE ANALYSIS ~ Q~

RUBBER GOODS 3
Reasons for the Analysis
Acetone Extract.----If the acetone extraction is made on a vulcanized compound, the acetone removes the rubber resins, the free sulfur, any mineral oils
or waxes, and part of any bituminous substances or vulcanized oils that may
have been used. The percentage of free sulfur- is determined and deducted
from the total extract. The corrected figure thus obtained will at times give
1 By Raymond B. Stringfield Consulting Chemical Engineer, Sec'y Dental Plastic Co.,
Los Angeles. Formerly Chief Chemist, Goodyear Tire and Rubber Company of Callf.
2 Ind. and Eng. Chem. 16, 397 (1924).
Book of A. S. T. M. Standards, 1933.
PartII.
a Approved by the Supervisory Committee on Standard Methods of Analysis,
American Chemical flociety, March 1, 1924.
1999

2000

RUBBER COMPOUNDS

valuable information regarding the quality of the rubber present. For the
best grades of Hevea rubber, this corrected extract s~lOuld not exceed 5% of
the rubber present. A higher extract may indicate the presence of inferior or
reclaimed rubbers.
\
Chloroform Extract.-."The ,chloroform extraction removes a portion of the
bituminous substances and serves as an indication of their presence.
Alcoholic-Alkali Extract....:....Tlie purp~se of the alcoholic-alkali extraction is
to detect the presence of rubber substitutes.
Free Sulfur.-The free sulfur is defined as that which is removed during
the acetone extraction.
Total Sulfur.-This represents the sulfur that occurs in the compound either
free or chemically combined.
Ash.-The ash is the residue left after ignition, and consists 'principally
of the nonvolatile mineral fillers, together with their reaction products with
sulfur.
'
Sulfur in Ash.-The sulfur in ash consists of the sulfur from the mineral
fillers, and also part of the sulfur that was with the rubber, but which during
ignition enters into combination with one or another of the mineral fillers.
Special Determinatioris.-SpeciaI methods are given for the analysis of
compo,upds that contaip. glue, carbon, antimony, and waxy hydrocarbons.
Rubber.-Up to the present time no simple method has been devised for
the direct determination of the amount of rubber presl)nt in a vulcanized
compound. Therefore, an indirect metl(od is proposed which will give satisfactory results in all cases known to-day, except where there are found to be
'present decomposable fillers, such as carbonates, cellulose, and high percent'ages of mineral rubbers. As practically all insulating compounds contain
some of these ingredients, the said compounds when specified shall be analyzed
in accordance with Joint Rubber Insulation Committee's method, as given in
Paragraph 34.
,
When carbonates, talc, and asbestine are present, more accurate results
are obtained by the use of the Joint Rubber Insulatiqn Committee's methods
given in Paragraph 34.
I
If high percentages of mineral rubber are 'used, no accurate method is
known. If cellulose is present, the best results are obtained by the method
given in Paragraph 35.
. Preparation of Sample

---

1. Before preparing a sample for analysis, the analyst shall, by inspection,

assure himself that it has not been contaminated. The sample-shall be prepared by taking pieces from various parts of the original'sample and separating
them from foreign matter.
"
, 2. The pieces shall be ground to ~he required filfeness on a rubber mill or
,cut with scissors so as to pass a No. 14 sieve.
'
3. Hard rubber should be prepared by rasping with a coarse file, cleaning
with a magnet, and p~ssing through a No. 14 sieve.
, 4. Crude, reclaimed, or un vulcanized rubber s~all be sheeted out very thin
on an experimental mill, and shall be rolled in holland or other cotton cloth
to prevent the sample from sticking. If no mill. ~s'f(Vailable, the sample shall
be cut as fine as possible with scissors.
'.

RUBBER COMPOUNDS

2001

5. Samples of rubberized cloth sha:rr be prepared by cuttin~ into pieces

1.5 cm. square and then mixing well .
. 6. Cements: Cements shall be evaporated to dryness and the residue
analyzed as an unvulcanized sample. A separate sample shall be steamdistilled if examination of the solvent is desired.
Reagents
7. The acetone shall be chemically pure and shall be freshly redistilled
over anhydrous sodium carbonate, using the 56 to 57 C. fraction.
8. The alcoholic-alkali' solution shall be of normal strength.' It shall be
made by dissolving the required amount of alkali (either potass[um or sodium
hydroxide) in the smallest quantity of hot, distilled water and adding this to
specially purified alcohol, which shall be prepared as follows:
Dissolve 1.5 g. of silver nitrate in 3 ml. of water, and add it to 1000 ml. of
alcohol. Dissolve,3 g. of alkali in the smallest amount of hot water, cool,
add it to the alcoholic-silver' nitrate solution, and shake thoroughly. Allow
the solution to stand for at least 24 'J:!rs., filter, and distil. (Alcoliol denatured
with 10% by volume of methanol may be used in place of ethyl alcohol.),
9. The nitric acid-bromine reagent shall be prepared by adding a c'onsider::'
able excess of bromine to concentrated nitric acid and shaking thoroughly'. It
can be used immediately.
10. The zinc oxide-nitric acid solution is made by adding 200 g. of zinc
oxide to 1000 ml. of concentrated nitric acid.
11. Barium chlo~ide solution shall be made by dissolving 100 g. of crystallized barium chloride in 1 liter of'distilled water and adding 2 to 3 drops of
concentrated hydrochloric acid. If there is any insoluble matter or cloudiness,
the solution shall be heated over night on the steam bath and filtered. Care
.
should be taken not to add more acid than th~ amount specified.
12, Standard 0.1 N potassium permanganate solution shall be made as
follows:
Ditlsolve approximately 3.1 g. of potassium permanganate' in 1 liter 6f
water, and when It is' dissolved filter through an ignited asbestos pad. Weigh
out 0.25 g. of pure metallic antimony and transfer to a 600-mi.' Erlenmeyer
flask. Add 12 to 15 m!. of concentrated sulfuric acid, 10 to 12 g. of potassium
sulfate, and heat until all the antimony is dissolved. Dilute to 250 ml. with
water, add 20 mL of concentrated hydrochloric acid, cool to 10' to"15 C.,
and titrate with permanganate solution until a faint pink color is obtained.
I f 0 1 N KM '0 _ Wt. of metallic antimony':"
f
.
'
I, m.o.
n 4 - MI f
t
-wt.o antimony
0 permangana e
13. Starch-iodate paper shall be prepared by impregnating filter paper
with a solution obtained by'heating 2 g. of starch with 100 ill}. of water and,
after solution, adding 0.2 g. of potassium iodate dissolved in 5 ml. of water. '
, 14. A Gooch crucible shall be prepared in the following manner:
~,
Cut amphibole asbestos fine with shears, digest with 10% caustic soda
solution, wash with water, and then digest with concentrated hydrochloric
acid for a few hours on the steam bath. After it'has been washed comparatively free from acid by decantation, shake up the asbestos with water and use

2002

RUBBER COMPOUNDS

the 'resulting mixture in 'preparing the' pads, The Gooch crucibles should
. 1
be ignited, and are then ready for use."
~ . i5. The purity of all material shall be checked and determined by blank
analyses.
" .
\
Methods of Analysis
,

16. Specific Gravity.-This shall be determined by the use of a pycnometer,

using alcohol in place of water to eliminate the errors due to air bubbles.

A = Wt. of pycnometer filled with alcohol,

B = Wt. of pycnometer filled with sample and alcohol,
C = Wt. of sample.

Sp.gr'=C_(B_A)XSP:gr. of alcohol.
'i'.....

it ..~.

17. Acetone Extract.-The extracting apparatus used here and for other
extractions shall be of the type similar to that shown on page 1970. (See also
Industrial and Engineering Chemistry, 9, 314 (1917).) The flask shall be
heated so that the period of filling an empty siphon cup with acetone and completely emptying it will be between 2.5 and' 3.5 min.
Place 2 g. of rubber in a thimble made by folding a filter paper so that it
will fit in the extraction cup, which is suspended in a weighed extraction flask.
Extract the sample continuously for 8 hrs., unless the solution.jn ;the thimble
is still colored at the end of that time, when the extraction shall proceed for
a further period of 4 hrs. or longer. For hard rubber the extraction period
shall be a minimum of 72 hrs. Carefully note ail characteristics of the acetone
extract, both when hot and cold. Distil off the acetone on the stearit bath
at as low a temperature as possible. Loss of extract by bumping can be
avoided by means of a gentle current of air or by slightly inclining the flask.
Care must be taken to avoid allowing the flasks to stand on the steam
bath after the solvent has been removed, because/appreciable quant~ties of
free sulfur may be'lost by so doing. Dry the extraction flask and contents
in an air bath for 1 hr. at 70 C., cool, and weigh. Call the residue" Acetone
Extract, Uncorrected."
.
Wt. of extract
Percentage of acetone extract, uncorrected = Wt. sampJe X 100.

or

Save the r\lbber sample and keep for future determination.

18. Chloroform Extract-The rubber sample (Paragraph 17), without
removing the acetone from it, is suspended in a second weighed extraction
flask and extracted for 4 hrs) with chloroform. Care should be taken that
any small particles of rubber, which are often carried down into the extract, are
filtered off. Evaporate off the solvent and dry the 'residue to constant weight.,usually 1 hr. at 70 C.-cool, and weigh. The color of the chloroform solution
should be recorded. Reserve the rubber for extraction with alcoholic-alkali.

Wt. of extract
Percentage of chloroform extract = 'Wt f
. I X 100 .
-0 samp e

RUBBER COMPOUNDS

2Q03

19. Alcoholic-Alkali Extract.-Dry the rubber from the chloroform extraction at about 70 C. to remove the chloroform, transfer to a 200-mt Erlenmeyer
flask, add 50 ml. of alcoholic-alkali solution, and heat under a reflux condensrr
for 4 hrs. Filter into a 250-ml. beaker, wash with two portions of 25 ml. each of
boiling alcohol, then with three 25-ml. portions of boiling water, and evaporate
,. the filtrate to dryness. Use about 75 ml. of distilled water to transfer the
residue to a separatory funnel. Acidify the solution with 10% hydr.c;>chloric
acid, testing with Congo red paper. Extract with four 25-ml. portions of ether,
unless the fourth portion is colored, when the extraction m~ust be continued
until no further quantity can be removed. Unite the ether fractions and wash
thoroughly with distilled water until free from acid (two washings are generally
sufficient). Filter the ether solution through a plug of absorbent cotton into
a weighed flask, wash with ether, evaporate, dry to constant weight at 70 C.,
cool, and weigh.
..
,
Percentage of alcoholIc-alkalI extract

Wt. of extract
Wt f ' 1 X 100.
0 samp e

20. Free Sulfur.-Add to the flask containing th,e aC,etone extract, uncorrected (Paragraph 17). 50 to 60 ml. of distilled water and~ 2 to 3 ml. of bromine, and cover with a watch glass. (If the acetone extract indicated a large
amount of free sulfur, the amount of bromine used may he increased.) Allow
the flask to stand 0.5 hr: on the side of the stea.m bath, then heat cautiously over
the direct steam bath until the solution is practically colorless, filter into a
25Q-ml. beaker, and dilute to, about 175 ml. with distilled water. Precipitate
with barium chloride and determille in the usual manner, taking the proper
precautibns to prevent the reduction of the barium sulfate during the ignition.
Percentage oUree sulfur
o

Wt. of BaS04XO.1373
Wt.0f sampIe

X 100.

21. Total Sulfur.-Place 0.5 g. of rubber in a porcelai,n crucible of about

75 ml. capacity, add 15111l. of the nitric acid-bromine mixture, cover the crucible
with a watch glass, and let it stand for 1 hr. in the cold. Heat for an hour on
the steam bath, remove the cover, rinse it with a'little';distilled water, and
evaporate to dryness. Add 3 ml. of nitric acid, cover, warm a short time on
the steam bath, then let it cool. Carefully add in small portions, by means
of a glass spatula, 5 g. of sodium carbonate (weighed to 0.5 g.). The watch
glass is to he raised only high enough to permit the introduction of the spatula.
The carbonate is allowed to slide down the side of the crucible and is not
dropped directly into the acid. Rinse the watch glass with 2or 3 m!. of hot
distilled water and stir the J;Uixture thoroughly with a glass rod. Digest f{)r
a few minutes, spread the mixture halfway up the side 'of the crucible to
facilitate drying, and dry on a steam bath. Fuse the mixture by heating over
a sulfur-free gasoline flame.
- Place ~he crucible in an inclined position on a wire triangle and start the
ignition over a low flame. The tendency for the rubber to burn too briskly
is controlled by judicious use of the stirring rod with which the burning portion
is scraped away from the 'rest. When part of the mass- is burned white, a
.fresh portion is worked into it,' and so on untiL all of the organic matter is
-\

2004

RUBBER ,COMPOUNDS
i

destroyed. !tis necessary to hold the edge of the crucible with tongs. Toward
_the las.t .half of the operation the flame should be i~(}reased som,e:what" but it
.is _.neyer nec_essary to heat the crucible to redness. With care a cr!1cible can
be used for,at least ten or twelve fusions.
\
_Alter a fusion, allow the crucible to cool, place it! in a 400-ml. beaker, add
.suffi<liell;t distilled water to cover the crucible (about 125 ml. are required),
and.digest,on the steam bath for 2 hrs., with occasiotial stirring. If the filtra-tion.cannot he made in the same day, do not. add the water, but allow the fusion
.to stand over night'.
-'
" .

If
j

.
IjI

/
j

-';'

'11'

-/

5
4

"

/'

"

F
V

/_
0

10

200

JOO

400

,500

FIG. 286.

Filter the solution into a covered 400-ml. beaker cont:hling-5ml. of concentrated hydrochloric acid, 'and wash the residue thoroughly with hot water.
(A qualitative test for bar.ium sulfate can be made on the residue.) Now
complete the acidification of the filtrate and washings in the cold and add 2 ml.
of concentrated hydrochloric acid In excess. Cover the beaker and heat the
solution on the steam 'bath. The totai volume 'of -the solution- should not
exceed 300 ml. The solution must be acid to Co~go paper "in order to insure
the complete destruction of the carbonates. Precipitate with barium chloride
and determine the- sulfur in the usual manner."
.~.,
.
"
,
.
Wt. of BaS04XO.1373
Percentage of total sulfur 4 =
W.
f
1.' X 100.
-to_ D --samp e .
4

To correct the sulfur for occlusion use Graph -1, lirie marked 2.'

RUBBER COMPOUNDS

2005

22. Alternative Method.-When this method is used, the total sulfur

determined represents the suliur exclusive of that as barytes, if used in the
compound.
Place a 0.5-g. sample in a 500-ml. Erlenmeyer destruction flask (Pyrex
glass). Add 10 ml. of zinc oxide-nitric acid solution and whirl the flask rapidly
to moisten the sample thoroughly. If convenient, the mixture may be allowed
to stand over night at this point. By so doing the sample becomes partly
decomposed; this permits the addition of fuming nitric acid with no danger
of ignition of the sample. Add 15 ml. of fuming nitric acid and whirl the flask
rapidly to keep the sample immersed to avoid ignition by too rapid oxidation.
With some samples it will be found necessary to cool the flask under a stream
.
of tap water.
When the solution of the rubber is complete, add 5 m!. of saturated water
solution of bromine, and slowly evaporate the mixture to a foamy sirup.
(For the determination of total sulfur in unvulcanized mixtures use 3 ml. of
bromine in place of bromine water.)
If particles of organic matter remain at the end of the evaporation, add a
few m!. of fuming nitric acid, and reevaporate as before. Cbol, and add a few
crystals of potassium chlorate to assist in the 'oxidation of the sulfur and the
decomposition of any nitrates.
Place the flask on asbestos gauze and evaporate the mixture to dryness
over a Tirrill burner. Then bake the mixture at the highest temperature of
the burner until all nitrates are decomposed and no more nitrogen peroxide
fumes can be detected. (Care shQuld be taken to insure uniform penetration
of the heat throughout the contents of the flask and to remove the flask as
soon as the "baking" is complete.) When the baking is complete, cool the
flask, add 50 m!. of (1 : 6) hydrochloric acid, and heat until solution is complete.
(In case the original mixture contains barium salts, they will be precipitated
at this point. If litharge is present in the mixture, lead salts, not otherwise
removed, will be eliminated in the final washing with boiling water.) Filter
the solution, dilute to 300 mI., precipitate with barium chloride, and determine
the sulfur in the usual manner.
Percenttj,ge of total sulfur '=

Wt. of BaS04XO.1373
Wt f
I
X 100.
0 samp e

23. Ash.-Wrap a I-g. sample in a filter paper, extract with acetone for 4
and transfer to a weighed, approximately 50-m!' porcelain crucible. Ash
the sample in a muffle furnace by heating at the following rate:

~rs.,

Time, min ........ 0

Temp., 0 C........ 0

5
100

10 15 70 75
200 300 300 400

80
500

85
550

145
550

Remove crucible from the furnace, cool, and weigh.

If no furnace is available, distil off the rubber over a very small flame, not
allowing it to catch fire, and ignite gently until burned clean, cool, and weigh .
Wt. of ash
I X 100.
Percentage of ash = Wt f
. 0 samp e
&

To correct the sulfur for occlusion use Graph 1, line marked 1 (figure 286).

2006

RUBBER COMPOUNDS

24. Sulfur in Ash.-Add 3 ml. of nitric acid-bromine mixture to the ash

(Paragraph 23), cover with a watch glass and heat-for:1 hr., r.emove the cover,'
rinse it with a little distilled water, and evaporate to dryness. Complete the
determination of sulfur as described under Paragraph 21.
Percentage of sulfur

III

ash

5=

Wt. of BaS04XO.1373
W f
'I
X 100.
t.o samp e

Save the insoluble residue after filtering the solution of the fusion mixture
in water for testing according to Paragraph 25.
25. Barium SuIfate.-The barium sulfate is calculated from the barium
in the ash, which is determined as follows: Filter off the insoluble matter after
the fusion and extraction in Paragraph 24, wash back into the original beaker
with hot water, dissolve the residue in the beaker and any traces on the filter
paper with hydrochloric acid, and heat the solution on the steam bath. Filter
through the same filter as before and wash thoroughly with hot water. Nearly
neutralize the solution with ammonium hydroxide, leaving it slightly acid.
Saturate the solution with hydrogen sulfide in the cold, and when the lead
sulfide has settled, filter into a 400-ml. beaker and wash thoroughly. The
total volume should be not over 200 ml. Precipitate with 10% sulfuric acid
and determine the barium in the usual mariner. Barium sulfate determined
above is assumed to have been added as such. Obviously, if barium carbonate
is used, it must be determined in order that an undue correction shall not be
made.
.
Wt. of BaS04XO.1373
W f I X 100.
Percentage of sulfur as banum sulfate =
t.o samp e
,
26. Total Antimony.-:--When a qualitative test indicates that antimony is
present, place a 0.5-g. sample, and transfer in a Kjeldahl flask, add 25 ml. of
concentrated sulfuric acid and 10 to 12 g. of potassium sulfate, place a funnel
in the neck of the flask, and heat until the solution gecomes colorless. Coal.
and wash the funnel, dilute the solution to 100 ml. j\vith water, and transfer
to a 400-ml. beaker, dilute to 250 ml. with hot water, and precipitate the antimony with hydrogen sulfide. Filter and transfer the precipitate to a Kjeldahl
flask, add 15 m!. of concentrated sulfuric acid, 10 to 12 g. of potassium sulfate, and heat as described above until the solution is colorless. Wash the
funnel, dilute the solution to 100 ml. with water, add 1 to 2..g. of sodium sulfite,
and boil until all the sulfur dioxide is driven out. This is shown when nO"
blue colqr is obtained with starch iodate paper. Add 20 ml. of_concentrated
hydrochloric acid, dilute to 250 to 275 ml. with water, cool to 10 to 15 C.,
and titrate with standard 0.1 N permanganate solution until a faint pink
color is obtained. If iron is found to be absent, it is not necessary to precipi;.
tate the antimony with hydrogen sulfide, and the seMnd heating in. a Kjeldahl
flask can be eliminated.

8b (Paragraph 12) XmI. of permanganate 100

P ercen t age 0 f antImony =
W fIX
.
t.o samp e
I

27. Antimony in the Ash.-This is determine<!_.ib.::the ash of a I-g. sample.

Transfer the ash to a 600-ml. Erlenmeyer flask, aad 12 to 15 ml. of concentrated

RUBBER COMPOUNDS

2007

sulfuric acid and 10 to 12 g. of potassium sulfate, and qoil until solution is

complete. It may be necessary to warm part of the sulfuric acid in the crucible to transfer any adhering particles to the flask. Rinse with the remaining
portion of acid. Then complete the determination as described in Paragraph 26.
Percentage of antimony as Sb 2 0 4 in ash
=Sb (Paragraph 12)'Xml..of permanganateX1.26 100
X
Wt. of sample
.
28. Free Carbon.-Extract a 0.5-g. sample for 8 hrs. with a mixture of
chloroform and acetone in a volume ratio of 2 : 1. Transfer the sample to a
250-ml. beaker and heat on the steam bath until it no longer smells of chloroform. Add a few m!. of concentrated nitric acid and allow to stand in the cold
for about 10 min. Add 50 ml. more of hot concentrated nitric acid, taking
care to wash down the sides of the beaker, and heat on the steam bath for at
least 1 hr. At tile end of this time there should be no more bubbles or foam
on the surface. Pour the liquid while hot into the crucible, taking care to keep
as much as possible of the insoluble material in the beaker. Filter slowly with
gentle suction and wash well by decantation with hot concentrated nitric acid.
(Caution: Empty the filter flask.) Wash wi~h acetone and a mixture of equal
parts of acetone and chloroform until the filtrate is colorless. Digest the
insoluble material, which has been carefully retained in the beaker, for 30 min.
on the- steam bath with 35 ml. of 25% sodium hydroxide solution,. This
treatment with alkali can be omitted if silicates are absent. Dilute to 60 ml.
with hot distilled water and heat on the steam bath. Filter the solution
of alkali and wash well with hot. 15% sodium hydroxide solution.
Next wash the residue four times with hot concentrated hydrochloric acid.
Neutralize the last washing with ammonia and test for the presence of lead
with sodium chromate solution. If none is present, finally wash with warm
5% hydrochloric acid. Remove the crucible from the funnel, taking care that
the outside is perfectly clean, dry it in an air bath for 1.5 hrs. at 110 C., cool,
weigh, burn off the carbon at a dull red heat, and reweigh. The difference in
weight represents approximately 105% of the carbon originally present in the
form of !.amp-black or gas black.
Percentage of free carbon
=

Wt. of crucible with carbon - wt. of crucible after ignition 10

X 0
1.05 Xwt. of sample
.

29. Detection of Glue.-Extract a 2-g. sample for about 6 hrs. with acetone.
Then heat for 2 hrs. on th<t steam bath, with sufficient water to cover it, cool,
filter, and pour slowly into a 1% solution of tan.nic acid. If a permanent
cloudiness, however slight, appears, glue is present.
30. Nitrogen Calculated as Glue.-When glue is found to be present, a
correction must be made for the natural protein in the rubber. The average
amount of nitrogen is 0.4%. (See Paragraph 40 for calculations.)
Extract a 2-g. ~ample for 8 hrs. with acetone. Remove the solvent from
the sample, and transfer the latter from the filter paper to a 750-ml. Kjeldahl
flask. Add 25 to 30 ml. of concentrated sulfuric acid and 10 to 12 g. of sodium
sulfate. Place the flask on the Kjeldahl digesting apparatus. Heat gently

2008

RUBBER COMPOUNDS
I

until the first. vigorous frothing ceases, then raise the heat gradually until
the liquid boils. Continue the boiling until the solution becomes clear. Allow
the flask to cool, dilute carefully with 150 ml. of water, and allow to cool.
Add 100 ml. of 50% sodium hydroxide solution, pour~ng it carefully down the
side of the flask so that it does not mix immediately with the acid solution.
Add about 1 g. of granulated zinc to prevent bumping and a piece of paraffin the
size of a pea to diminish frothing. Connect the flask quickly with a condenser,
the delivery tube of which dips into a 500-ml. Erlenmeyer flask, containing
50 ml. of 0.1 N sulfuric acid diluted to about 100 ml. Carefully swirl the
flask to mix the contents and start to heat gently, increase the flame as the
danger of foaming over diminishes, and, finally, boil briskly until about onehalf of the liquid has passed over into the receiver. Add methyl red solution
and titrate the excess acid by means of 0.1 N sodium hydroxide. solution.
A blank determination should be made.
Percentage of nitrogen as glue
_100 (ml. H 2S0 4 Xnormality-ml. NaOHXnormality) (0.014) (6.5)
Wt. of sample
= (ml. H 2S0 4 X normality - ml. N aO H X normality ) X 4.55. 6
31. Unsaponifiable Matter.-If this dete~mination is made, the/ree sulfur
shall be determined on a separate 2-g. sample.
Add to the acetone extract obtained from two 2-g. samples (Paragraph 17)
50 mI.' of N alcoholic-alkali solution, heat on the steam bath under a reflux
condenser for 2 hrs., remove the condenser, and evaporate to dryness. Transfer
to a separatory funnel, using about 100 ml. of water, add 25 ml. of ether, and
shake. Allow the two layers to separate thoroughly, then draw off the water
layer. Continue the extraction of the water layer with fresh portions of ether
until no more unsaponifiable matter is removed, unite the ethereal layers,
and wash with distilled water. Transfer the ether to a tared flask, distil off
the ether,. dry to constant weight at 70 C., cool, and reigh.
.
Wt. 6f extract
Percentage of unsapomfiable matter = Wt f
I X 100.
.
.0 samp e

32. Hydrocarbons A.-'To the unsaponifiable matter (Paragraph .31) add

50 ml. of absolute alcohol, and heat on the steam bath for 0.5 hr. Let the
flask stand in a mixture of ice and salt for 1 hr. Filter o:ff""theseparated waxy
hydrocarbons on filter paper by applying a gentle suction. Wash with alcohol
(95% will do), which has been cooled in an ice-salt mixture. The funnel
should be surrounded by a freezing mixture.
Dissolve the precipitat!l from the filter paper with hot ~hloroform, and
catch the solution in a weighed 100- to 150-ml. beaker. Wash the flask with
hot chloroform, which is added to the same beake'r, in order to include any
undissolved waxy matter adhering to the walls or'the flask. Evaporate off
the solvent, dry to constant weight at 70 C., cool, and wei~h.
Wt. of extract i
Percentage of hydrocarbons A = Wt f" - I X 100.
: 2- :.samp e
6

Simplified formula when a 2-g. sample is used.

RUBBER COMPOUNDS

2009

33. Hydrocarbons B.-Evaporate the alcohol from the flask containing the
alcohol-soluble unsaponifiable material, add 25 ml. of carbon tetrachloride
and transfer to a separatory funnel. Shake with concentrated sulfuric acid:
drain off the discolored acid, and repeat with fresh portions of acid until there
is no longer any discoloration. After drawing off all the acid, wash the carbon
tetrachloride solution with repeated portions of water until all traces of acid
are removed. Transfer the carbon tetrachloride solution to a weighed flask,
evaporate off the solvent, and dry to constant weight at 70 C., cool, and weigh.
Wt. of extract
Percentage of hydrocarbons B = Wt f
1 X 100.
0 samp e
34. Rubber Hydrocarbons by Joint Rubber Insulation Committee's
Method'.-Add to the flask containing the rubber residue from the alcoholicalkali extraction sufficient water to make the total 125 mI., and then add 25 ml.
or- concentrated hydrochloric acid. Heat for 1 hr. at 97 to 100 C. Decant
the supernatant liquid through a hardened filter paper on a Buchner funnel,
7
in diameter, using suction. Wash the residue with 25 m!. of hot water
and decant. (While a Buchner funnel is recommended, it is permissible to
use an l1-cm. hardened filter paper with platinum cone in a 60-degree funnel.)
Perform this entire treatment with water and hydrochloric acid three times.
The rubber at this stage should be white and practically free from black specks
of undissolved fillers; if it is not, continue the acid treatment until the black
specks disappear. (If carbon is present, all the particles of rubber will be
grayish, bluish, or black, depending on the form and quantity of carbon used ..
, ,Black specks in light particles of rtili>ber usually indicate the presence of lead
sulfide, which must be removed to prevent the formation of lead sulfate on
igniting the residue C.) Add 150 ml. of hot' water to the flask and let stand
on a steam bath or hot plate for half an hour and decant through the filter
paper, repeating until the washings are free from chlorides. Transfer all the
rubber in the flask to the filter paper and dryas much as possible by suction.
Wash the rubber with 50 ml. of 95% alcohol, using suction. Transfer the
entire residue to a weighing bottle. Dry at 95 0 to 1000 C. for an hour, cool
in a vacuum desiccator under reduced pressure, and weigh. Dry for 30 min.,
cool, and weigh, repeating this process until either constant weight is reached
or the weight starts to increase. Let this weight be represented by C. Determine the ash, E, in a portion, D, of this residue, C, and sulfur, H, in the remaining portion, G. Also, determine the sulfur, F, in the ash, E. Perform the ash
determination as described under Paragraph 23, and the sulfur determination
as described under Paragraph 21. Calculations:

em.

Percentage of rubber. hydrocarbons = 100~ ( 1-~ _

E; F) .

35. Cellulose.-Treat 0.5 g. sample of rubber with 25 ml. of freshly distilled

cresol (b.p. 1980 C.) on the electric hot plate for 4 hrs. at 1650 C. Allow to
cool and add 200 m!. of petroleum ether very slowly and with constant agitation.
After the solution has settled completely, filter through a Gooch crucible and
wash three times with petroleum ether. Wash very tlioroughly with boiling
benzene and finally with acetone. Treat the contents of the flask with hot

RUBBER COMPOUNDS

2010

10% hydrochloric acid, and trarisfer the entire contents to the Gooch crucible
with the aid of a "policeman." Continue to treat with hot 10% hydrochloric
acid until the pad has been washed at least ten times. Wash the pad free from
chlorides with boiling water, and run small 'portions of acetone through it until
the filtrate is colorless. Treat with a mixture of equ~l parts of acetone and
carbon disulfide in the same manner. Wash with alc,ohol and dry for 1 hr.
and 3,0 min. at 105 C. Remove the pad from the crucible with help of {L pair
of sharp-pointed tweezers, using the underneath portion of the pad as a swa,b
to clean the sides of the crucible, and place all of this material in a tared weighing bottle. Repl&ce in the drying oven for about 10 min., cool, and weigh.
0

Weight of weighing bottle, pad, insoluble fillers, and, cellulose

-weight of weighing bottle,
= weight of pad, insoluble fillers, and c llulose.
Transfer the contents of the weighing bottle to a 50-ml. beaker and pour
over it 15 m!. of acetic anhyaride and 0.5 m!. of concentrated sulfuric acid,
and allow to digest for at least 1 hr. on the steam bath. After the mixture
has cooled thoroughly, dilute with 25 ml. of 90% acetic acid and filter through a
weighed Gooch crucible. To guard against traces of the material being carried
through, this filtration, as well as the ones to fbllow, must be very slow and only
gentle suction can be used. Wash with hot 90% acetic acid until the filtrate
comes through absolutely colorless, and/then wash about four times more.
Wash with acetone about five times. After having taken care that all of the
material has been washed out of the beaker in which the acetylation took place,
.remove the crucible from the funnel, cleau the outside thoroughly and dry for
2 hrs. at 1500 C. Cool and weigh. Original weight of crucible+weight of pad,
fillers and cellulose - weight of crucible after acetylation = cellulose.
Cellulose
Percentage of cellulose = Wt f
I X 100.
0 samp e
I

36. Barium Carbonate.-Barytes was determined by the calculation to

barytes of all barium found in the sample. O):wiously, if barium carbonate
is present, it must be determined in order that an undue correction shall not
be made. The determination is as follows: A I-g. sample, in a porcelain boat,
is placed in a combustion tube through which passes a current of carbon
dioxid((. The sample is ashed in the tube. After ignitioll: and cooling in the
atD;l,ospp.ere of carbon dioxide, the boat is removed, the residue'1inely ground
in an'agate mortar, transferred to a 250-ml. beaker, and treated with 5 to 10 g.
of ammonium carbonate, 15 to 20 ml. of strong ammonia water, and about
50 m!. of distilled water.. The mixture is boiled for 20 min., filtered, and the
precipitate thoroughly washed to remove all soluble sulfates. The residue
on the filter paper is washed back into the original beaker and about 10 m!.
of glacial acetic acid with sufficient water to make the total volume about
100 ml. are added. This is heated to boiling and filtered through the same
paper as before. Hydrogen sulfide is passed into the filtrate to precipitate
the lead, and the solution is subsequently treated as in thf:}!determination of
barytes. The final weight of barium sulfate obtajned'is 'calculated to barium
carbonate.
.

RUBBER COMPOUNDS

2011

Percentage of sulfur required to convert barium present as carbonate to sulfate

Wt. of BaS04XO.1373 1
Wt. of sample
X 00.
Calculations
37. Percentage of organic acetone extract =!"percentage of acetone extract
(uncorrected) -percentage of free sulfur.
38. Percentage of acetone extract (corrected) = percentage of organic acetone extract-percentage' of waxy hydrocarbons.
39. Percentage of waxy hydrocarbons = percentage of hydrocarbons A
+percentage of hydrocarbons B.
40. Calculations for glue: Multiply the percentage of rubber as compounded
(Paragraph 46) by 0.004 and 6.5 and call this value A. Subtract A from the
percentage of glue (Paragraph 30) and call this value B. Substitute B for
glue in Paragraph 45 in calculating the percentage of rubber hydrocarbons
and then complete the calculations.
41. Percentage of sulfur as antimony trisulfide = percentage of total
antimony X :~2 = percentage of total antimony X 0.400.
42. Percentage of total antimony as trisulfide = percentage of total antimony

Sb 2S3

X Sb

percentage of total antImony X 1.400.

43. Percentage of total sulfur corrected = percentage of total sulfur - perof sulfur as barium sulfate - percentage of sulfur as antimony trisulfide.
When barium carbonate is found', the percentage of sulfur necessary to convert
it to sulfate must be added to the total sulfur (corrected).
44. Percentage of ash (corrected) = percentage of ash -percentage of sulfur
in ash (+percentage of sulfur as barium sulfate) -percentage of antimony in
ash, calculated to Sb 20 4
45. Percentage of rubber hydrocarbon = 100 -the sum of the following
percentages, except when it is determined by the J. R. 1. C. method as in
Paragraph 34.
cen~age

Organic acetone extract

Chloroform extract
Alcoholic-alkali extract
Total sulfur corrected

Ash, corrected
Carbon
Glue
Total antimony as trisulfide

46. Rubber as Compounded.-Percentage of rubber hydrocarbon plilJ 5%

of its weight is taken as "rubber as compounded," except when the sum- 01
the percentages of the acetone extract (corrected), chloroform extract~ -and
alcoholic-alkali extract is less than the figure represented 'by the arbitrary
,5%, as in many high-grade compounds. Percentage of rubber as com:(>oundecl_
is then: Percentage of rubber hydrocarbon plus the sum of the per<le!l~!I_g().s of
the organic acetone extract, the chloroform extract, and the alcoholj~~_aJliali,
extract.
-.
..
..
-

.,.~

RUBBER COMPOUNDS

2012

47. Rubber by Volume.-Percentage of rubber by volume =

Percentage of rubber as compounded (Paragraph 40) i><
sp.gr. of compound
I.
0.94 (taken as the ,average\sp.gr. of crude rub~er)
48. Ratio of

org~nic

acetone extra~t =
1
Percentage of organic acetone extract. 100
Percentage of rubber as compounded X
.

49. Ratio of sulfur to rubber =

'
Percentage of total sulfur, corrected 100
Percentage of rubber
, as compounded X .
Statement of Results
Per cent

Per cent

Organic acetone extract

Waxy hydrocarbons
Chloroform extract
Alcoholic-alkali extract
Free sulfur
Total sulfur corrected
Ash corrected
Cellulose

Sulfur as barium sulfate

Total antimony
Sulfur as antimony trisulfide
Carbon
Glue
Rubber hydrocarbon.
Rubber as compounded
Rubber by volume
Ratio of acetone extract to rubber as compounded
Ratio of sulfur to rubber as compounded
Specific gravity

Specific Gravity.-The apparent specific gravity of an unvulcanized rubber

compound is almost invariably from 1 % to 5% low, due to the presence of air
introduced during milling, most of which escapes duting vulcanization. Due
to its large coefficient of expansion, rubber decreases rapidly in specific gravity
with increasing temperature, and raw rubber undergoes a change of state and
sudden drop in sp.gr. from approximately 0.93 ~o 0.91 at around 30 C., which
is only slowly reversible on cooling. 7

For ordinary purposes, it is sufficiently accurate to su&pend a small sample

from the beam of a balance of a Young's Gravitometer,
aftElr .balancing,
wet the surface by raising a beaker of 'alcohol under it, remove the alcohol by
use of a beaker of wash water, and weigh suspended in distilled water as usual,
thus avoiding air bubbles.
,
For factory control purposes, use of solutions of CaCl 2 or ZnCl 2 of proper
strength permits rapid checking of samples between predetermined limits.
Thus a piece of stock which should be of 1.12 sp.gr. will float on a solution of
1.13 sp.gr. arid sink in a solution of 1.11 sp.gr.
Extracts.-Hard asphalt hydrocarbon (mineral rubber), mineral oil, stearic
acid, palm oil, pine tar oil, hardwood pitch, etc., are frequently added as softeners or for other specific purposes. The analytical.estimition of these soft-

and

Pickles, India Rubber Jour. 67, 17 (1924).

.-

RUBBER COMPOUNDS

21)13

eners is not an exact matter. On the other hand, an approximation is all that
is usually required.
Plantation rubbers yield acetone extracts usually between 2.0% and 3.5%,
while various grades of wild rubbers give extracts up to 10.0%, or even higher
for resinous grades. These extracts are usually of a soft conflistency, and until
the use of added softeners became common, ~ild rubbers were largely used for
softening purposes ..
Reclaimed rubber is variable in composition, depending on the source, but
that made from whole tires usually falls within the following limits:
Specific Gravity .............................. . 1.16 - 1.24
Moisture, max................................ . 1.0%
Alkalinity, max., 4 hr. ext...................... . 0.10%
Acetone Ext .................................. . 7.0% -12.0%
Chloroform Ext ............................... 10.0% .-18.0%
Alcoholic KOH Ext ........................... . 1.00/, - 2.0%
Total S ............................ : ......... . 3.0% - 4.0%
Free Carbon ................................. . 3.5% - 7:5%
Ash ......................................... . 18.0% -35.0%

Added softeners are partially extracted from a cured stock by acetone, and
subsequent extraction with chloroform removes still more, but there is always
an appreciable amount of the softener which is not extracted, depending on the
type of softener, type of compound, cure, etc. Mineral rubbers for instance,
vary greatly, some coloring the acetone extract dark brown, and others not at
all, some carrying a large amount of sulfur and consequently combining with
little additional, while others tend to "hog" sulfur during the cure. Very little
analytical work is available to show the extraction under definite conditions.
The following is an example of the extraction of 3.0% of softener from a well
cured" pure gum" compound carrying 3.5% sulfur.

.Soften~r

Recovered in
Acet. Ext.

In CHCla
Ext.

Not
Recovered

Mineral Rubber ..............

Hardwood Pitch .............
Pine Tar Oil .... , ............
Mineral Oil. .................

44.3%
46.3
68.0
70.0

28.7%
10.7
14.0

27.0%
43.0
18.0
25.6

4.4

Saponification of the acetone extract gives some idea of the percentage of

mineral oil softeners present, but is seldom resorted to. Acetone extracts can
be checked to within 2% of the extract, and CHCla extracts to within 4%.
Alcoholic KOH extract is run only when the presence of oil substitute
(factice) is known or suspected. It removes about 50% of the substitute, and
must be corrected for an extraction of about 0.5% on the rubber, and 1.5%
of any reclaim present. When substitute is present, the acetone extract must
be corrected for about 10% of the substitute in the case of black or brown substitute, this correction being largely un vulcanized oil.
Sulfur.-The sulfur present in a rubber compound comes from two sources,
(1) sulfur added for vulcanization purposes, and (2) sulfur present in the .compounding ingredients.

201,4

RUBBER COMPOUNDS
J

The latter may be a donsiderable percentage of the total sulfur, as the following table of the per cent usually found in some of the pommon ingredients will
show.

1. .... .

Whole Tire Reclaim ......................... :

Reclaimed Friction ................................. .
Mineral Rubber .............................. : .1 ..... '.
Hardwood Pitch ....... : .............. , ... , ......... .
Lithopone .......................................... .
Golden Antimony .................................. .
Ultramarine Blue ................................... .
Mercaptobenzothiazole .............................. .
Thiocarbanilid ................................. : .... .
Diphenylguanidine .................................. .

3.5)1,
1.0
1.0- 7.5
nil
9.5
20.Q-40.0

10.0
35.0
16.0
. nil

The only o11.e of these affecting the free sulfur determination is golden
antimony. This comes in a number of grades, the -most common being that
containing 40-43% of coprecipitated CaS04 and 15~17% of free S. Most
accelerators are somewhat soluble in hot acetone, and their S would appear
as free S in an uncured stock, but during the cure these are largely decomposed,
and any residual effect on free S is negligible.
During vulcanization, sulfur combines ilf four ways, (1) with the rubber
hydrocarbon, (2) with the natural resins and proteins, (3) 'with softeners if
present, and (4) with inorganic oxides to form sulfides such as ZnS. This has
been discussed by Kelly in several papers 8 and methods of separation given.
The ordinary figure for combined sulfur (corrected total sulfur-free sulfur) includes all of these forms, and also the sulfur from various compounding ingredients as noted above. When interpreted in connection with the formula of
a compound and the processing to which it has been subjected, it is an excellent
indication of the state of vulcanization.. The percentage of combined sulfur
(CS) divided by the percentage of rubber present in a compound, gives the
coefficient of vulcanization (YC). Plotting either YC or CS against time of
cure gives a smooth curve for any given temperaturp conditions. It is thus
possible to take a sample from any point in a finished a;rticie, and by determining
its YC, estimate the equivalent cure. It should be remembered however,
that in the case of a tire or other complex article in which several stocks are
cured in contact with each other, migration of sulfur from a high sulfur stock
to one of lower sulfur content (on the rubber) is liable to occur, and also that
sulfur tends to migrate in the direction of the source of heat.
Free sulfur is at times a useful indication of the "non~blOoming" characteristics of a stock, usually less thim 1.0% FS on the rubber bei~ necessary to
prevent sulfur" bloom."
Sulfur determinations have a tendency to .run slightly high, usually not
over 0.10%. Duplicates by extremely careful work can be made to check
within 0.04% S, but on ordinary rubber work, any variation not over 0.10%
is satisfactory checking.
Ash.;---An ash determination is only an approximation of the amount of
inorganic fillers in a stock, as it is affected by several factors. The various
clays, asbestine, lime, etc., carry considerable percentages 9f combined water
which is volatilized over a range of temperature. ,Whiting tends to lose CO 2

. _,-

Ind. Eng. Chem. 12, 875; 14, 196; 16, 148.

RUBBER COMPOUNDS

2015

unless very carefully handled. During the ignition, much of the sulfur present
combines with any zinc oxide, lime, magnesia or litharge to form sulfates. \ If
the sample is all()wed to catch fire, zinc oxide will be reduced and volatilized.
In practice, ash serves as a check on the subsequent analysis of the ash, and
sometimes itself gives sufficient information on the leading of a stock.
Analysis of Ash.-In addition to the determinations of sulfur, barium
sulfate and antimony given in the standard methods, it is frequently necessary
to further analyze the ash to secure an idea of the fillers and inorganic accelerators present in the sample. Si0 2 , Fe 20a, Al 20 a, CaO, MgO, PbO and ZnO are
usually determined, but as these occur in many fillers in varying proportions,
some judgment is necessary for the proper interpretation of results, and micro;,
scopical methods are of considerable aid in determining the fillers present,
Particularly in the field of cheap inert fillers, the compounder's choice is
often affected by the location of the factory and the materials most easily
available, as similar stocks can frequently be made with widely varying compositions.

The following table shows the approximate percentages found in some comme'rcial compounding ingredients:
Asbestine-MgO 25%, SiO. 65%, Ah03 3.5%, CaO 1.5%, Ign. Loss 5.0%.
Barytes-BaSO.90-95%.
Dixie Clay-SiO. 46%, R.Oa 39%, CaO trace, MgO trace, Ign. Loss 15%.
Lead Oleate-PbO 25-30%.
Lime-Ca(OH). 91 %.
Lithopone-ZnO 29%, BaSO. 71 %.
Mag. Carb.-MgO 42%, Ign. Loss 50%.
Mag. Usta-MgO 87%, CaO 2.5%, SiO. and R.Oa 4.0%, Ign. Loss 6%.
Whiting (pptd.)-CaO 50%, MgO 3%, SiO. and R.Oa 5%, Ign. Loss 40%.

Free Carbon.-The direct determination of free carbon is quite accurate.

In addition to carbon black" the most used reinforcing filler, small percentages
of free carbon may come from any mineral rubber 01' hardwood pitch present,
but these are usually neglected.
Glue.-The presence of as little as 0.9% glue can be detected qualitatively,
but of late years, the use of glue in rubber compounds has become quite rare.
Calculation of the Probable Composition of a Stock.-The experienced
compounder will draw heavily on his.knowledge of the compounding ingredients
available and suitable to use in a stock giving properties of the sample, in
. calculating its probable composition, but the following will give an idea of
general procedure.
1. From the analysis of the ash, estimate the amounts of fillers and inorganic
accelerators used, allowing 15% ignition loss for clays, and using data such as
general appearance of the stock, state of cure, color, etc., to aid in judging the
amounts of CaO, MgO, and PbO used, if any.
2. Estimate carbon black, glue, and substitutes if any, from their special
analyses.
-3. Estimate rubber hydrocarbon by differenq_e. Subtract 3.0% of this
from the combined sulfur, and see-if the use of reclaim is indicated. The odor
of the stock is also often characteristic in connection with either reclaim or
softeners such as pine tar oil. If much CS is left over, figure it as 3.5% of
reclaim, and check this indicated reclaim content with the other extracts and
ash to see if it seems reasonable.

2016

RUBBER COMPOUNDS

4. Subtract 10% of the estimated reclaim and 3,5% of the estimated rubber
content from the acetone extract, and estimate the kiIld and amount of softeners
used from the remainder, taking into a'Ccount tne character of the extracts,
and correcting the chloroform extract if necessary for the effect of the re6laim
present.
.\
5. Figure the sulfur used as total sulfur less 3.5% of the reclaim and 6.0%
of any mineral rubber present.
.
I
6. Calculate the specific gravity of the stock from the probable composition,
and compare with the actual specific gravity, which should be within 0.02,
except in the case of compounds containing _large amounts of softeners or
sulfur, with which the actual gravity is usually several points higher than the I)
theoretical.

ANALYSIS OF. RAW MATERIALS

I

Raw Rubber.-All but a very small pircentage of the rubber used in the
United States is plantation grown, and appears upon the market quite clean
and dry, and graded (by appearance and history) accor'ding to standards established by the Rubber Exchange of New York. Except for research purposes, the chemical laboratory is .seldom called on to determine more than
moisture, acetone extract and ash. The factory, however, is interested in the
rate of cure and physical properties with given formulae, which vary WIdely
with rubber of identical appearance. For control purposes, it is common to
blend at least three lots,of rubber together on mixing mills, .and evaluate the
blend by means of physical tests on experimental batches, the formulae used
varying with the factory and the use to which the rubJ;>er is to be put. Methods
for special analytical determinations are given in the previous chapter. For
data on the chemistry and structure of hevea rubber, see Dinsmore, Ind. Eng.
Chem. 18, 1140-5 (1926), and Fisher, Chemical Reviews, VII, 1 (Mar. 1930).
Reclaim.-Reclaimed rubber is classified according to the type of rubber
scrap from which it was prepared, and is subjected when required, to the same
chemical examination as a rubber compound, with particular emphasis on
specific gravity, moistu~e, alkalinity, acetone extract and ash.. These tests
however, are mainly useful in control of material of known sourc_e. and type, and
any new material requires physical testing, both alone and in the type of compound where it is to be used, and also factory,tests in actual production compounds to determine the effect on factory handling conditions.
Accelerators and Antioxidants.-A very considerable number of substances,
mostly organic, has come into use in recent years'as additions to rl!bber compounds to accelerate the rae of vulcanization, or to retard the rate of deterioration after vulcanization. A few of these are of definite composition that permits simple tests for purity. Most, however, are compl\}x, and require extended physical tests for their evaluation, and al1l'consequently sold chiefly
without test on the strength of their manufacturer's reputation. Data with

RUBBER COMPOUNDS

2017

rf'spect to the classification of accelerators. have been published by Dinsmore

and Vogt, Rubber Chern. and Tech., 1, 410 (1928), and by Martin and
Thiollet, Rubber Chern. and Tech., 2, 356 (1929) and 3, 38 (1930). See also
Callan and Strafford, Chern. and Ind., Jan. 4, 1924. The following common
accelerators are usually subjected to chemical specifications as shown below.
Guanidines
Formulae.-

CRa

ON'\,
C-NR2
ON(
CRa H

Diphenylguanidine
(DPG)

Diorthotolylguanidine
(DOTG)

Triphenylguanidine
(TPG)

Moisture.-Max. 0.5%. Dry to constant weight at 1050 C.

Ash.-Max.0.5%. Ash one gram sample and weigh residue.
Purity.-:-Min.97.0%. Dissolve one gram sample in 25 ml. alcohol, titrate
with 0.1 N alcoholic HCI using bromphenol-blue indicator. 1 ml. 0.1 N HCI
=.0211 g. DPG, .0239 g. DOTG, .0287 g. TPG.
Fineness.-l00% through 100 mesh, min. 99% through 200 mesh. Use a
camel's hair brush lightly to assist in sifting.

Mercaptobenzothiazole
Formula.-

a)-s-n
Trade name, "Capta;c."

Moisture, Ash and Fineness, as above.

Purity.-Min. 90.0%. Dissolve one-half gram sample in 25 ml. alcohol
and titrate with 0.1 N alcoholic N aOH, using phenolphthalein indicator. 1 ml.
0.1 N NaOH=.OI68 g. mercaptobenzothiazole.
Softeners
Softeners constitute an important class of rubber compounding ingredients,
as in addition to modifying the properties of the vulcanized compound, and
in most cases reducing the cost, they greatly affect the properties before vulcanization and often make possible large savings in processing. A great many
materials are used, including petroleum products such as refined lubricating
oils, petrolatums, paraffin, and blown asphalts (mineral rubber), vegetable ojls
such as cottonseed oil, palm oil, and castor oil, wood distillation products such
as pine tar, pine tar oil and hardwood pitch, resins such as rosin, glycerol ester,
cumar, dam mar and shellac, and fatty acids such as stearic, oleic and lauric
and their zinc salts.

21)18

RUBBER COMPOUNDS

The most essential feature from the compounder"s standpoint is the uniformity of the material selected, and ordinary tests fo~ melting point, viscosity, '
etc., appropriate to the product are applied, the only features peculiar to rublter
being that copper and manganese must be absent" Ion account of the very
deleterious effect of even minute traces in rubber compoun\ls, and alkalinity
or mineral acidity must be absent on account of the effect on the rate of
vulca~ization.

Fillers and Reinforcing Agents

Reinforcing agents such as carbon black, zinc oxide, light magnesium
carbonate, and certain processed clays; color pigments such as iron oxide,
chrome green, golden antimony, and various dyes and lakes, and mineral
fillers such as whiting, barytes, clay and asbestine, are widely used in rubber
compounding for their effects on the cost and physical properties of the stocks.
In this connection, it may be mentioned that the custom has arisen in
American rubber factories of . calling many compounding ingredients "pigments," regardless of whether they are colored or not. This is chiefly due to
the practice of writing formulae in code, a certain "pigment number" representing a given ingredient, and the term even being extended to softeners such
as pine tar and solvents such as benzol.
,
On account of the great effect of particle shape, size and surface condition,
all material from new sources requires testing in actual ~ompounds before it
can be approved for use, but for control purposes on material from known
sources, customary tests for purity,and uniformity can be applied. All materials should pass 100% through 100 mesh, this in the case of reinforcing pigments
being merely to detect the presence of any lumps, as the individual particles
are far finer than any scree~ing test can grade: On carload shipments, the
screening test should be made on a representative sample of 200 pounds to
detec~ stray contamination. All materials should be dry, free from copper and
manganese, free from alkalinity and acidity, and except for lead compounds
used as such, free from lead. Other properties of in1jerest to the rubber compounder include specific gravity, particle shape and size, heat conductivity,
specific heat, reactivity with suifur, volatile matter at curing temperatures, and
tendency to agglomerate during milling. A few special tests follow.
Carbon Black.-Simultaneously determine moisture by heating a sample
one hour in an electric oven at 105 C., and total volatile by heating a one gram
sample in a platinum crucible with capsule cover in an eleCiricfurnace for seven
minutes at 955 0 C. Difference is volatile matter, which in exce~r- 5.5%-6.0%
has been shown to lower the quality of the black materially for rubber purposes.
Any black of different type from standard will show difference in covering
power, determined as follows: take sample of O~050 g. each of standard black
and of sample to be tested, and place side by side on a clean glass plate. To
each, add 2.00 g. standard zinc oxide and 30 drops of pure raw linseed oil. Mix
each by rubbing with a spatula until the resulting paste is smooth and free from
streaks. Compare colors by placing small portions of each on a microscope
slide and pressing with a second slide. Add additional zinc oxide if necessary
from a weighing bottle until the pastes show exactly the same color.
r--

Wt. ZnO for standard

:-:'-::
Wt. ZnO for sample =covermg power.

RUBBER COMPOUNDS

2019

Zinc Oxide.-Test for covering power, using 3tandard carbon black as

above. Determine sulfates by dissolving 2 grams in 50 mI. water and 10 m!.
concentrated HCI, adding 5 mI. Br water, boiling off excess Br, and precipitating
with BaC1 2 Max. 0.20% SOa. Determine lead by dissolving 5 g. in 60 m!.
1 : 1 HNO a, .diluting to 400 mI., ,and throwing out Pb0 2 electrolytically. Max.
0.20% PbO. Compare particle size and shape by grinding a few mg. in
turpentine, mounting on a slide to form a very thin film, and comparing under
the microscope with a standard sample.
Color Pigments.-Compare color and fineness by grinding 0.050 g. to 0.100
g. sample into 2.0 g. ZnO and linseed oil as in covering power test.
Solvents.-On account of their lower price range, solvents derived from
petroleum are chiefly used in this country for making rubber cements, other
solvents such as benzol, xylol, carbon bisulfide and carbon tetrachloride being
used only 'for special purposes. The petroleum solvents are held to very close
distillation specifications to control the rates of drying of the rubber cements.
Low boiling fractions' are to be avoided as much as high boiling fractions, as
they cause greater fire hazard and large loss of solvent during mixiIig and
handling. Specifications usually call for no residue on evaporation at room
temperatures, no acidity to methyl orange, sweet odor, negative doctor test,
and color not over 30 Saybolt. Typical distillation specifications for rubber
maker's naptha, using the standard A. S. T. M. method, are as follows:
52 0 C.
67
72
81
max. 90
max. 105
DP .................. max. 144

rBP. . . . . . . . . . . . . . . ..
10%. . . . . . . . . . . . . . . ..
20 .................. ,
40 ...................
60 .................. ,
80 ...................

min.
min.
min.
min.

73 0 C.
82
84
88
max. 93
max. 100
max. 128
min.
min.
min.
min.

SLAG ANALYSIS 1
The slags made in metallurgical operations consist of complex mixtures
and solid solutions of silicates, oxides, aluminates, fluorides and many other
compounds. In the molten condition some of these components are probably
in igneous solution in each other, their exact composition and the nature Qf
their association depending upon the state of equilibrium at the time the slags.
were produced. The composition of slags is important both because of its
influence upon furnace operations and because of its effects upon the characte~
of the metallic products. For these reasons it is necessary to make frequent
slag analyses in all metallurgical plants.
In iron blast-furnace slags the principal constituents are silica, alumina,
lime and magnesia, with small amounts of iron oxide, manganese oxide, calcium
sulfide and phosphoric acid. Titanium oxide is also frequently present.
In modern blast-furnace practice, which involves the use of high lime slags,
in order to keep sulfur out of the iron, the, percentage of lime and sulfur in
the slags is closely watched. In open-hearth furnace slags the amount of
phosphorus may be high enough to warrant the sale of the slag as fertilizer.
Lead blast-furnace slags consist mainly of silica, iron oxide, manganese
oxide, lime, alumina and zinc oxide, with small percentages of magnesia,
barium oxide, lead, copper and sulfur. Of these the most important to the
metallurgist are silica; iron oxide, lime, zinc oxide and lead, and determinations
of them must be made at least once a day. The other components are determined less frequently. In some of the Mexican lead-smelting plants calcium
fluoride occurs in the blast-furnace slags. The lime combined as fluoride is
determined separately, as it is not available for the saturation of silica.
Copper slags are made up mainly of silica, iron 9xide, lime and alumina.
The manganese oxide and the zinc oxide are usually less than in lead slags.
The silica and alumina, however, are generally higher. This is especially true
in the case of reverberatory slags, in which a greater viscosity can be permitted.
The percentage of copper is quite important as it measures the chief source of
loss of that metal. The percentage of copper in the slag is usually about one
hundredth of that in the, matte. The amount of lime iii- copper slags varies
a great deal more than it does in lead slags, as it is of less imp(irtance in influencing the metallurgical results. In converter slags the copper may amount
to several per. cent, and the slag then becomes an important by-product.
Composition of Slags
I
Blast-Furnace Slags.---Si0 2 =35 to 45%; FeO""0.2, to 0.5%; CaO;=32 to 46%;
A1 20.=14 to 17%; MgO = 1.0 to 1.5%; Mn02=trace to 0.1%; 8=1.5 to 2.0%; P 20 5
=0.01 to 0.96%.

Lead Blast-Furnace Slags.-8i0 2 =36.7%; FeO=29.5%; CaO=20.9%; Al 20.

=3.16%; MgO=1.69%; MnO=0.91%; 8=0.75%; BaO=0.54%; ZnO=4.19%;
Pb=1.38%; Cu=O.10%; Au =0.001%; Ag=0.70%.
. I
.

By Wilfred W. Scott.
2020

SLAG ANALYSIS

2021

Average of one year's analyses on representative samples.

Copper Blast-Furnace Slags.-8i0 2 =23 to 49%; Fe(MnO) =2.5 to 57%; Ca(MgO)
=2.5 to 25%; Al 20 a =1.5 to 18%; ZnO=nil to 23%; Cu=0.15 to 1.1%; S=ni! to 1.5%.
Decomposition of the Sample and Analysis
General Reverberatory Slag.-Usually 0.5 gram of the finely ground slag
is taken for. a determination, a larger amount when the constituents sought
are present in very small amount. The material placed in a platinum crucible
is fused with 5 to 6 grams of Na 2CO a at red heat (muffle furnace, if possible)
for 10 minutes. The mass is cooled on the side of the crucible, and the fusion
dissolved in a casserole by adding an excess of HCl, followed by a few ml. of
HN0 3 The solution is evaporated to dryness and the residue baked. The
mass is taken up with dilute HCI. Silica remains as a residue and is filtered
off and determined. The metals are determined in the filtrate.
Chilled Blast Furnace Slag.-The sudden chilling of the molten slag by
dropping into water causes a physical change in the material which enables a
decomposition by means of acids without resort to fusion, as given above.
About 0.5 gram of the finely ground chilled slag in a small casserole is
moistened with water and about 3 m!. of concentrated HCl added. All lumps
are broken up by stirring with a glass rod until a.smooth jelly results. A few
drops of HN0 3 are added to oxidize the iron and the jelly is worked up around
the sides of the casserole in an even layer to the height of about Yz inch. (This
permits rapid dehydration of the silica and reduces loss by "spitting.") The
.acids are expelled by evaporation and the residue gently baked. (If the
temperature is too high, some alumina will combine with the silica and give
high results for Si0 2 .) After cooling, about 20 mI. of concentrated H;CI are
added and the mixture boiled for a few minutes, then diluted with an equal
volume of hot water and filtered hot. Si0 2 remains as a residue, the metals
are in solution.
Silica.-The residue obtained from the decomposition is washed with hot
water until free from chlorides, the washings being added to the main filtrate.
The residue and filter are ignited, then cooled and weighed as Si0 2
NOTE.-8hould barium be present, as is the case with many lead slags, nitric acid is
omitted in the decomposition, as this would cause contamination of the silica with
barium sulfate. If nitric acid is used, it is advisable to obtain the silica by difference,
by first weighing the residue. This is placed in a weighed platinum dish, then treated
with hydrofluoric acid and a few drops of sulfuric acid. The solution is evaporated,
all the acids and the silica being expelled (as silicon fluoride). Any remaining residue is
weighed and subtracted from the first weight obtained. The loss represents the Si0 2
Copper slags carrying 40 to 45% of Si0 2 are more troublesome than those with 35 to
37% of Si0 2 in regard to alumina uniting with the silica during dehydration.
Lime.-(In presence of iron oxide and alumina) Ammonia is added to the
filtrate from silica and then oxalic acid little by little until the precipitated iron
and aluminum hydroxides just dissolve. The solution is again made ammoniacal and oxalic acid again added to dissolve iron.' The solution should
appear a light apple-green color. It is now boiled for a few minutes and the
precipitated calcium filtered off and washed with hot water until free from
oxalic acid, six or seven times being generally sufficient.

2022

SLAG ANALYSIS

The filter containing the calcium oxalate is dropped in a beaker, 150 ml. of
hot water added together with 15 ml. of (1 : 1) H 2S0 4 al1d the oxalic acid titrated
with standard potassium permanganate.
1 ml. of 0.1 N KMn04=0.0028 gram\of CaO.
The Fe value of KMn04 multiplied by 0.5 = CaO value.
NOTEs.-If preferred, iron and aluminum may be precipitated'as hydroxides, and
filtered off, calcium being determined in the filtrate. It is advisable to redissolve the
precipitate to recover any occluded lime and again precipitate the hydroxides with
ammonia, the filtrate being combined with the main filtrate.
In place of titrating the oxalate of lime, it may be ignited and the residue weighed
directly as CaO.
If oxalic acid is added in the form of a fine powder instead of a solution, the calciulJ1
precipitates in a mu~h more granular form and requires less boiling before filtering.

Iron.-(In presence of silica) Half a gram or more of the finely ground

chilled slag in a beaker is treated with 50 ml. of boiling water, the particles
stirred up and kept in suspension and about 25 ml. of concentrated HCI added.
The solution is boiled until clear. (If coke dust is present it will still be evident,
but may be negLected.) Stannous chLoride ,soLution is added, drop by drop,
until the iron is reduced (solution becomes colorless) and 2...:3 drops excess
added. The solution is cooled by placing ~he beaker in cold water. Mercuric
chloride, HgCI 2, solution is added to precipitate the excess of stannous chloride.
The iron is determined by titration ivith standard potassium dichromate,
using- ferricyanide indicator on a spot plate. See subject in the chapter on
Iron. Report as FeO.
1 ml. of 0.1 N K 2 Cr 2 07 = 0.00558 g. of Fe, 0.007i9 g. of FeO,
Fe X 1.2865 = FeO.
NOTE.-The iron may be precipitated as hydroxide, then dissolved in dilute H 2S0 4,
the solution reduced by Qoiling with test lead or reduced in cold solution with zinc and
the solution titrated with standard potassium permanganate., This method is frequently
preferred where previous fusion to decompose the ore has been required. Lime may be
determined in the filtrate from iron.
'

Determination of Other Constituents in Slags

Alumina: Phosphate M ethod.-The filtrate from the silica determination
is diluted to about 400 :t;nl. with cold water and 30 nil:-of JO% ammonium
phosphate added. Dilute ammonia water is added until a slight permanent
precipitate forms.
1.5 ml. of concentrated HCI is added and 40 ml. of 20% solution of N a 2S 203
and the mixture boiled for about two or three minutes. 15 ml. of 20% ammonium acetate and 6 ml. of concentrated acetic acid are added and the boiling
I
continued for about 15 minutes.
The precipitate is allowed to settle for 15-20 minutes, the clear solution
decanted through a filter and finally the aluminum phosphate filtered and
washed. (The presence of the acetate makes the comppund more granular.)
10 washings with hot water are sufficient.
I
The precipitate is dried, ignited and weighed as::.AlP0 4. '
The precipitate contains 41.85% Ah03.
.- - - -

SLAG ANALYSIS

2023

NOTE.-The slag may be decomposed by treating 0.5 gram in a platinum.dish with

5 ml. HNO" 6 m!. of HF and about 2 m!. of (1 : 1) H 2S0 4 The sample is taken to strong
fumes of SO" then cooled and 10 mI. of HCI added and the mixture boiled. Alumina and
iron may now be precipitated as hydroxides, and .filtered off and washed. After dissolving in HCI, precipitate AlPO. as in the method above.
Manganese.-Half a gram of the sll1g is placed in a beaker and ,about
50 m!. of water added, the solution stirred until the material is in suspension
and about 15 m!. of HCI added, followed by 5 m!. of HNO" and the solution
boiled until most of the chlorine has been expelled.
I
Hot water is added to make up to a volume of about 100 m!. The iron is
now precipitated by adding ZnO emulsion until the acid is neutralized and an
excess of ZnO forms on the bottom of the beaker.
The solution is boiled for a few minutes and the manganese titrated while
the solution is still hot (in presence of the precipitate) with standard potassium
permanganate.
The end-point is best seen by allowing the precipitate to settle slightly and
observing the clear upper stratum of the liquid.
1 m!. of 0.1 N KMn04=0.00165 g. of Mn.
The lime value of the KMn04 XO.588 = Mn.
N()TE.-In place of the method given above consult the chapter on Manganese for
the volumetric oxalic acid method.
Zinc.-To 0.5 gram (1 gram if zinc is low) of the sample in a casserole is
added 3 m!. of water, 5 m!. of HCI and 2 ml. of HNO.. When the SiO z is
completely gelatinized, about 4 grams of NH 4 CI is stirred in.
The sample is dehydrated only until the residue crumbles easily as baking
is liable to volatilize some zinc as chloride. About 30 m!. of hot water are
added and the solution brought to boiling, filtered and the residue washed with
hot water.
I I
To the filtrate ammonium persulfate and bromine are added, the amount
being governed by the manganese present, i.e., 0.03 g. of ammonium persulfate
and 10 m!. bromine for every 0.01 g. of Mn in solution. Ammonia is added
'in slight excess, the solution boiled about 2 minutes, filtered, and the residue,
Mn02, etc., washed. (It is advisable to redissolve the precipitate in a little
dilute HCI and again precipitate with 'persulfate and bromine, adding the
filtrate to the main filtrate.)
The solution is just neutralized with HCI and 5 m!. excess added, followed
by 2 grams .of ,test _lead. The solution is boiled about 15 minutes.
About 8-10 m!. of HCl are added, the solution heated to about 60 0 C.
and the potassium ferrocyanide added. When the titration is almost completed, the bluish-white color of the precipitate changes to nearly pure white.
If the ferrocyanide is made up by adding 21.63 grams of potassium ferrocyanide per liter of solution, 1 m!. win equal very nearly 0.005 gram of Zn.
Consult chapter on Zinc.
In place of the above method the rapid method given in the chapter on Zinc
may be followed.
Magnesia.-0.5 gram or more of the finely ground chilled slag, placed in
a casserole, is moistened with water, 5 ml. of HCI and a few drops of HN0 3

2024

SLAG ANALYSIS

added. .The acids after reacting with the slag are evaporated off and the
silica dehydrated. The residue is taken up with 15! ml. of HCI and about
30 ml. of water, and boiled.
Silica is filtered off and washed.
I
The filtrate is made slightly alkaline with ammorii1t, 8 ml. of (NH 4 )2S and
1 gram of (NH 4)2CO a added and the solution boiled, for a few minutes, the
precipitate then filtered off and washed several times with water containing a
little (NH 4 )2S.
The filtrate is made slightly acid with HCI and boiled down to about
50 ml. The precipitated sulfur is filtered off and the filter washed. If the
solution is cloudy, it is cleared by adding bromine water and boiling.
Ammonia is now added, and a small amount of ammonium oxalate to'
remove any CaO still remaining; after boiling, the oxalate of calcium ("lime'D
is filtered off and washed.
Magnesia is now precipitated in the filtrate by addition of Na 2HP0 4 or
another alkali phosphate, and the magnesium ammonium phosphate filtered
off, washed with 2% ammonium nitrate solution, ignited and weighed as
Mg 2P 20 7 Consult the chapter on Magnesium.
Mg 2P 2 0 7 XO.3621,= MgO.
Copper: Colorimetric Method with Ammonia.-The percentage of cgpper
normally present in blast furnace slags/should not exceed 0.3%. Converter
slags may contain as much as 2.5% copper. The colorimetric m~thod is
especially adapted for this determination owing to its simplicity and speed.
Slags which have been" chilled" are readily decomposed by HCI. Stirring
constantly during the acid action and diluting as soon as decomposition is
complete prevents the separation of gelatinous silica and hastens the solution
of copper.
Procedure: Color S,tandards.-0.2 gram of pure copper foil is dissolved in
20 m'l. of 1 : 1 HNO a in a covered beaker, warming gf.ntly. After evaporating
down to about 10 mI., the solution diluted to 50,ml. is.,transferred to a graduated
liter flask, 200 mI. of ammonium hydroxide added and the solution diluted to
1000 ml.
5, 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100 ml. portions are taken and
placed in 100-ml. Nessler colorimetric tubes, (In place of the tubes clear
white glass bottles of about 150 mI. capacity may be use9..) The volume of
each solution is diluted to 100 ml. with 1 : 5 ammonium hydroxide.
If a colorimeter is available, a standard is taken that corresponds to that
of the sample as prepared according to directions below and comparisons made
by viewing the solutions through the depth of. solutions adjusted to give the
same intensity of color.
Preparation of the Sample.-Two grams off the finely crushed sample
(100-mesh) is treated in a beaker with 10 ml. of 'water to thoroughly moisten
the powder, 10 ml. of concentrated HCI are added and the mixture stirred,
warming gently until the undissolved residue becomes flocculent, remaining in
I
suspension.
About 90 ml. of H 2S water are added or the ~me volume of water and H 2S
gas passed in until the solution is saturated .. The solution is warmed gently

SLAG ANALYSIS

2025

until the copper' sulfide coagulates. The precipitate is filtere.d through a

small filter (a Witt filter is satisfactory) and washed with H 2S water.
The sulfide is now dissolved in 5 ml. concentrated HN0 3 (in the beaker in
which the CuS was precipitated), 30 ml. 1 : 5 NH 40H solution are added, the
solution filtered through a second filter, and this washed with 1 : 5 ammonia
reagent.
The copper solution is transferred to a Nessler tube (bottle or colorimeter
tube as the case may require) and comparison made with a standard, the
solution being made up to 100 ml. with 1 : 5 NH 40H.
ELECTROLYTIC AND IODIDE METHOD FOR COPPER

For extreme accuracy the electrolytic or iodide methods are recommended.

Details of these methods may be found in the chapter on Copper. See also
the chapter on Fire Assay Methods by 1. A. Palmer.
Lead.-The determination of lead is frequently required in slags. Slags
high in silica are apt to contain lead in appreciable quantities, probably combiried as silicate. The decomposition of the slag by treatment with HF as
given in the footnote on page 2021 is recommended.
The determination of lead in slag does not differ materially from its determination in ores. Reference is made to the chapter on Lead for its estimation.
Other Constituents.-The determination of barium, titanium, sulfur,
phosphorus, chromium, cobalt, nickel, alkalies, etc., is occasionally necessary.
The procedures for these determinations may be found in the chapters dealing
with the elements in question.I
1 The introduction to this section "Slag Analysis" was written by Prof, I. A. Palmer,
Department of Metallurgy, Colorado School of Mines.
A number of the procedures appearing in this chapter on slag analysis are standard
methods used at the Washoe Smelter of the Anaconda Copper Mining Company and are
procedures originally suggested by Professor H. W. Thomson, Department of Mining,
The University of British Columbia.

METHODS FOR ANALYSIS OF' \SOAPS AND

SOAP PRODUCTS i I
Soap is the term generally applied to the alkaline salts of the non-volatile
fatty acids. From a scientific viewpoint, all metallic salts of, fatty acids are
soaps; most of them are, however, insoluble in water and, therefore, not commonly used as cleaners. Commercial soaps are made by saponification ofglycerides of fatty acids with sodium or potassium hydroxide, or by saponification of fatty acids with sodium or potassium carbonate. Various manufacturing methods are employed, but the two most common are tIle" cold process"
and" .boiled process." In the" cold process" the fat and alkali are combined
in the cold, and since no salt is added to " grain out" the soap, it, contains all
of the impurities of the fat and alkali, and \111 the glycerin. "Boiled soaps"
are made from fat by boiling in a kettle with- sodium hydroxide or carbon!l-te.
Salt is added to " grain out" the soap, which separates the soap from most of
the glycerin and impurities.
I
To soaps made by either of the above ,processes may be added various ingredients to obtain certain characteristics and qualities desirable from a viewpoint of appearance and of detergency. Perfumes, coloring matter, and medicants are added to toilet soaps; transparent soap is a combination of soap,
alcohol, sugar, and glycerin. Some soaps may contain fillers added to impart
special characteristics~ A few of these are colloidal clay, pumice, oatmeal,
bran, sawdust, milk, ground corn, blood albumin, gliadin, cellulose, etc. Medicated soaps may contain carbolic acid, thymol, petroleum, vaseline, camphor,
gelatin, iodides, chlorides, zinc stearate, zinc oxide, and mercuro-compounds.
From this brief description it can be seen that a soap received in a laboratory
may contain any of the various constituents mentio~d. The following methods include only those commonly encountered inlsoap analysis. When analyses
for fillers are desired other than those described, the usual methods for such
materials are applicable.
In compiFng these methods, in many cases the standard methods for the
sampling and analysis of ,commercial soap and soap prodllct~ 9f the American
Chemical Society 2 and of the American Oil Chemists' Society 3-are used, and
acknowledgment is given. In some cases the A. C. S.-A. O. C. S. methods
are modified or replaced by methods which have been developed for soap works
laboratories where a large number of samples 'are handled daily for control
purposes, so that methods must be rapid as well as accurate. For sampling of
I

Compiled by the Chemical Division of the Procter ~nd Gamble Co. 'Through the
courtesy of the Company and C. P. Long. The Methods approved by the A. C. S. and
A. O. C. S. Committees on the Analysis of Soaps and Soap Products have been incorporated.
2 Standard Methods for Sampling and Analysis of CommercjAl Soaps and Soap
Products, J. Ind. Eng. Chern., 14, 1159 (1922); Ind. Eng./Chem., Anal. Ed. 9,2 (1937).
3 Standard Methods for Sampling and Analysis. ofpeommercial Soaps and Soap
Products, Oil and Soap, 11,90 (1934).
2026
1

ANALYSIS OF SOAPS AND SOAP PRODUCTS

2027

shipments in commercial transactions, the methods prescribed by the two

societies referred to should be used.

PREPARATION OF SAMPLES FOR ANALYSIS

The instructions follow A. C. S.2_.A. O. C. S.3 methods in general detail.

Cakes or Bars
Samples on which an entire analysis is to be made should be run through an
ordinary food chopper. If the sample cannot be easily disintegrated and mixed,
it should be quartered by cutting at right angles in the center and portions from
all freshly cut surfaces shaved off with a knife, portions of these shavings being
used for analysis. Freshly stamped bars obtained in plant control may be
sampled with a cork borer of convenient diameter. Both ends of the sample
plug should be cut off and rejected. The samples should be preserved in airtight containers in a cool place. Excessive exposure to the air and ,handling
with the hands should be avoided as much !ls possible.

Powders, Flakes, and Chips

Samples in cartons should be removed and thoroughly mix~d on an oil
cloth. The sample should be, reduced if necessary by quartering or passing
through a riffle. It is then placed in airtight containers. Quart Mason jars
are convenient. Flakes should be broken up as much as possible with a knife
or chopper. Containers should be completely filled with the sample. When
removing porliiQns for analysis the entire sample should be mixed in the container with a spatula.

Paste Soap Products

The sample swuld be removed from its original container, thoroughly
kneaded and mixed well. Portions for analysis should be weighed quickly and
the remainder kept in an airtight container in a cool place.

Liquid Soap
The sample should be mixed carefully with a spatula or spoon. Shaking or
any action to form suds should be avoided. Unused portions should be kept
in air-tight glass containers.

2028

ANALYSIS OF SOAPS AND SOAP PRODUCTS

METHODS OF

ANALYSI~

When a determination shows non-conformity with specifications, a duplicate should be run. If the sum of all the constituents\is not reasonably close to
100%, the entire analysis should be re-run, beginning with moisture and volatile
matter and anhydrous soap. In case of heavily siliC~ted soaps, some moisture
will be retained by the silicate, and when appreciable amounts of glycerin or
solvents are present, the moisture and volatile value will be high. These facts
should be taken into consideration when interpreting final results.
,Moisture and Volatile Matter (Oven Method)
Weigh a 5.01 gram sample rapidly in a tared 4 ounce aluminum dish,
. tinned ointment box, or glass dish. Dry to constant weight in an oven at
150 0 2 C. Usually 45 minutes to 1 hour is sufficient. Soaps containing
naphtha or as much as 1 % glycerine give results too high on account of volatilization of these materials. Soaps containing silicate give results which are too
low because some water is retained by the silicate. In these cases moisture
should also be reported by difference after all other constituents have been
determined.
Some soaps will char at 1500 C. but may be dried satisfactorily at 105 C.;
usually two hours are sufficient. Soap~ from linseed and other oxidizing oils
absorb oxygen and, if .the oven is used, may gain in weight near the end of the
test. For soaps from such oils or those containing naphtha and glycerine, the
distillation method is much more satisfactory than any oven method.
Moisture (Distillation Method)

2, 3, 4

For soaps containing from 5 to 25% of moisture and volatile matter use a
20 gm. (0.04 gm.) sample. For soaps containing more than 25% moisture
and volatile matter use a 10 gm. (0.02 gm.) sample! The weighed sample is
carefully transferred to a 500 ml. short-neck round hottom flaSK or an Erlenmeyer flask. Add approximately 10 gm. of anhydrous, fused sodium acetate
to prevent violent frothing, and then follow with 100 ml. of xylene which has
previously been saturated with water by shaking the xylene with a small quantity of water and distilling., Use the xylene distillate for the determination.
Attach the flask to a graduated distilling tube receiver, 'such .a_s the BidwellSterling or Stark and Dean apparatus. The receiver should be 'calibrated at
25 C. and normally of 5 to 6 ml. capacity. It is connected to a reflux condenser (19" Liebig). Prior to starting the determination, fill the receiver
with saturated xylene by pouring in through the reflux condenser.
~o that the refl~xing ~ill be under better control,! wrap the flask ~nd the tube
leadmg to the receIver WIth asbestos cloth. Apply; heat to the flask by means
of a Argand gas burner or an electric heater and distill slowly. :The rate at the
start should be approximately 100 drops per minute. When the greater part
of the water has apparently distilled over, increase the distillation rate to 200
drops per minute until no more water is collected. rP,)lrge tJ:{e reflux condenser

, A. K. Church and J. H. Wilson, Soap, 7, 35, 193[.'-=-

ANALYSIS OF SOAPS AND SOAP PRODUCTS

2029

during the distillation with 5 ml. portions of xylene to wash down any moisture
adhering to the walls of the ~ondenser. The water in the receiver may be made
to separate from the xylene by using a spiral copper wire. Move the wire up and
down in the condenser occasionally, this causing the water to settle to the bot~
tom of the receiver. Reflux for at least two hours, after which the heat is
turned off. Adjust the water distilled over into the receiver to 25 C. Read
the volume of water and calculate the percentage of moisture' in the soap, as
follows:
Volumeinml.at25C.XO.997100_01
't
.
. h f
1
X
- /0 mOlS ure III soap.
Welg t 0 samp e
Total Fatty Acid, Fatty Anhydride, Combined Alkali, and Anhydrous Soap
(a) Soaps Containing a Small Amount of, or No Coconut OiI.-Weigh a
5.01 gram sample on a tared watch glass, transfer to a 250 ml. beaker, and
dissolve in 50 ml. of hot water. Transfer to a 250 m!. glass~stoppered extraction
cylinder, make up to a volume of 100 mI., and cool. Add 3 drops of methyl
orange and sufficient dilute H 2S0 4 (1 : 4) to give an excess of acid after decom~
posing the soap. Add 40 ml. of ethyl ether and shake gently until the contents
are well mixed. Let the cylinder stand until the ether layer has separated
clear. Draw off the ether layer into a tared 150 m!. Soxhlet flask through a
slender glass siphon by means of suction. Wash the siphon by drawing a few
m!. of ether through it. Repeat the extraction at least 3 more times, using
25 mi. of ether and shaking more vigorously after the first extraction. Silicated
soaps and flakes which show a tendency to dry out should be extracted about
6 times. Combine all extracts in the same Soxhlet flask. If the extract is not
clear transfer to another tared Soxhlet flask through a filter. Evaporate the
ether from the Soxhlet flask by warming on the edge .0J the steam bath. Rapid
evaporation will cause condensation of moisture in the flask. In case of soaps
high (40% or higher) in. coconut oil, evaporate only to about 50 ml. and proceed
as in (b). For other soaps evaporate to dryness, and finally dry for 1 hour
at 90 o 2 C. Cool and reweigh. Dissolve the extracted fatty acids in about
40 ml. of hot, neutral alcohol, add a few drops of phenolphthalein indicator and
titrate with Nil NaOH to exact neutrality.
Weight of fatty acids X 100
Weight of sample
NaOH titrationXO.9
Weight of sample

01
/0

t t l f tt
.d
0 a a yaCl .

% water equivalent to fatty acids.

% total fatty acid-% water equivalent = % "Fatty Anhydride."

NaOH titration X 3.1
Weight of sample

% combined Na 20.

NaOH titrationX4.7
Weight ?f sample

% combined K 20.

% fatty anhydride+% combined alkali = % "Real Soap" or Anhydrous Soap.

2030

ANALYSIS OF SOAPS AND SOAP PRODUCTS

(b) Soaps 1;Iigh in Coconut Oil.-If desired, this method may be used for
any soap. It is often called the" soda soap" method. 2 3
Proceed with the extraction as for boiled soaps except that the extracts are
collected in a 250 ml. beaker or Soxhlet flask. Partially evaporate the ether
to not less than 50 ml. on the edge of a steam bathi and add 50 ml. of hot,
neutral alcohol redistilled over caustic. Add several'drops of phenolphtlialein
indicator and titrate with Nil NaOH to exact neutrality. Evaporate on the
steam bath to dryness, and while the soap is in the form of a thick paste, swirl
the beaker on its side, thus distributing the soap in a thin layer along the sides
and bottom of the beaker.
.
Dry to constant weight in an oven at 1052 C. After a preliminary
drying at 105 C. for an hour, drying may be continued at 150 C. On remova1
from the oven cool in a desiccator and weigh as soon as cool.
Correct for neutral salts in t,he caustic solution by determining the amount of
salt per ml. by neutralizing 20 ml. of Nil NaOH with Nil HCI using phenolphthalein indicator and drying the residue to constant weight at 105 C, From
the weight of the residue found, subtract the' weight expected if the reagents
had been 100% pure. The difference divided by 20 gives the correction per ml.
for neutral salts. Subtract the product of the total fatty acid titration times
the factor (0.022) plus correction for neutraL salts from the weight of soda soap
to obtain total fatty acids. Calculate as for boiled soap.
If the soap is a cold process soap containing superfat, i.e. un saponified fat,
and total fatty acids are required, it will'be necessary to treat the sample after
dissolving in water with 5 ml. (or more) of 50 Be caustic potash solution.
The solutions should be mixed thoroughly. With cover glass on the beaker,
place over an opening in a steam bath, saponify for one hour with occasional
stirring. Acidify with an excess of dilute H 2S0 4 (1 : 4) and proceed as above.
If the real soap in a soap containing superfat is desired, it will be necessary
to determine the superfat including free fatty acid and unsaponifiable and to
correct the total fatty acid as determined without fur*er saponification.

Chlorides

Transfer the acid water in the cylinder from the total fatty acid extraction to
a 250 ml. beaker, neutralize with chlorine-free Nil NaOH and back titrate with
Nil H 2S0 4 to make the solution one drop on the acid side of neutral. Add 1
ml. of 10% potassium chromate indicator. Piace over a Wratten SafelightFilter No. 00 illuminated by a 200-watt Mazda light at dist"anceof.12". 'Titrate
slowly with Nl10 AgN0 3 solution, stirring vigorously. The addition of a little
gum arabic may aid in the titration.
Titration XO.585
Weight of sample

% NaCI,

TitrationXO.745 =o/c KCI

Weight of sample
0

In case total anhydrous soap is nob. to be determined, it. will be more convenient to use the following method. Dissolve 5.01 grams of soap in 40 ml.
of hot water, add 10 ml. J : 4 H 2S0 4 and heat on a"steam bath or hot plate until

ANALYSIS OF SOAPS AND SOAP PRODUCTS

2031

the fatty acids rise in a clear layer. Filter through a wet 9 cm. filter paper,
collecting the filtrate in a 250 ml. beaker. Wash the fatty acids and filter paper
with hot water until the total volume in the beaker is about 80 ml. Titrate as
described above with N/l NaOH and N/l H 2S0 4 until the solution is one drop
on the acid side of neutral. Titrate with N/lO AgNO a as described above.
A rapid method is that of Bennett.2 , 3, 5 Dissolve 5.01 grams of the soap
in 300 ml. of water, boiling if necessary to affect solution. Add an excess of
chlorine-free.magnesium nitrate solution (about 25 ml. of a 20% Mg(NOsh6H 20 solution). Without copling or filtering, titrate with N/lO AgNO a as
described above.
Unsaponified and Un saponifiable Matter (A. C. S.2_A. O. C. S.3)
The extraction cylinder used is a 250 m!. glass stoppered cylinder about 35
mm. in diameter and 300 mm. high. The solvent used is petroleum ethet which
should be of the pentane type, containing a mini~um amount of iso-pentane,
iso-;hexane and hexane and conforms to the following specifications. 2 , 3, 6
Distillation Test. *Initial boiling point-not less than 35 C. nor over 40 C.
Dry flask endpoint-not less than 50 C. nor over 60 C.
At least 95%, distilling under 55 C.
Not over 85% distilling under 40 C.
Specific gravity at l5S C.-0.630 to 0.675.
Color-water white. Doctor test-sweet. 7
Evaporation residue-not over 0.002% by weight.
Copper strip corrosion test, A. S. T. M.-D130-3G--Non-corrosive.
Unsaturated Compounds-trace only permitted.
Weigh a 5.01 gram sample into a 250 ml. beaker. Add 10'0 ml. of 50%
redistilled ethyl aleohoL Warm and stir to effect solution, keeping the temperature under 60 C. Filter off any undissolved residue on a Gooch crucible
with an asbestos or paper pad into an extraction cylinder. Wash 3 times with
hot 50% alcohol and then with 5 ml. of hot 95% alcohol. Complete the washing with a small amount of petroleum ether.
Make the volume of solution in the cylinder up to the 160 ml. mark with
redistilled alcohol. Add 50 ml. of petroleum ether. Stopper with a rubber
stopper which has been sOllked in petroleum ether overnight to remove soluble
matter. Shake vigorously for one minute and allow to settle until both layers
are clear. The volume of the upper layer should be about 40 ml. Draw off the
petroleum ether layer as closely as possible by means of a slender glass siphon
into a 500 ml. pear-shaped separatory funnel containing 20 nil. of 10% alcohol
containing a few drops of phenolphthalein. Take special care that none of the
alcohol-soap layer is drawn off. Repeat the extraction at least 6 times, using
H. C. Bennett, J. Ind. Eng. Chern., 13, 813 (1921).
Rules governing transactions between members of the National Cottonseed Products
Association, .1937-38, page 133.
* Distillation test to be made according to A. S. T. M. method D86-27.
7 R. C. Griffin, Technical Methods of Analysis, 2d edition, p. 235 (1927).
6

2032

ANALYSIS OF SOAPS AND SOAP PRODUCTS

50 ml. of petroleum ether each time. Draw all the. extracts into the same
separatory funnel. When extraction is completed, Istopper the separatory
funnel vigorously for Y2 minute to wash the extract. Draw off the alcohol layer
and swirl the funnel to collect any drops of the washing solution at the bottom.
Continue washing with two 5 ml. portions of 10% alcphol until the extract is
free from alkali and soaps as shown by the absence of color in the alcohol layer.
Transfer the washed extracts to a tared 150 ml. Soxhlet flask and rinse the
separatory funnel several times with petroleum ether. 'Evaporate the solvent
in a gentle current of dry air on the steam batp. Dry in ::n oven at 105 0 C.
for 15 minutes. Cool and weigh. Dissolve the residue in 50 ml. of warm ethyl
alcohol neutralized to phenolphthalein, titrate to the same color as the originl\l
neutral alcohol with N /25 sodium hydroxide solution and calculate to oleic acid.
Deduct this figure from the gross weight of the residue previously found. Any
blank residue from the petroleum ether must also be deducted fr6m the weight
of the residue. Report results as "Unsaponified and Unsaponifiable Matter."
Unsaponifiable Matter
Weigh 5.01 gms. of the sample into a 200 ml. Erlenmeyer flask. Add 30
ml. of redistilled 95% ethyl alcohol and add 5 ml. of 50% aqueous KOH. Boil
the mixture one hour unde~ a reflux condeiser. Transfer the solution to the
extraction cylinder and wash the flask with 95% redistilled alcohol to the 40 ml.
mark. Complete the transfer first with,warm and then cold water until the
total volume is 80 ml. Then wash the flask with a small quantity of petroleum
ether. Cool the cylinder aIi'd contents to room temperature. Add 50 ml. of
petroleum ether and proceed with the extraction as outlined under "Unsaponified and Unsaponifiable."
Weigh the residue and correct for fatty acids. Report the result as "Unsaponifiable Matter." Deduct the unsaponifiable result from that of the
unsaponified and unsaponifiable and report the difference as "Unsaponified
Matter."
I

For soaps containing lanolin, many more extractions for Unsaponified and
Unsaponifiable will be required for complete removal. For any sample, thorough and vigorous shaking is necessary in order to bring the two phases into the
most intimate contact possible.
F~ee

Alkali or Free Fatty Acid

-_

The methods outlined will not give absolutely accurate results for free alkali,
but will give results sufficiently accurate for control or practical soap analyses.
The accuracy is probably close to .01 % N a 2 0,
(a) When" Builders" (or Fillers) :Are Low in Amount and Are Not Determined.-Weigh a 20 gm. sample (.05) rapidly,and transfer to a 500 ml.
Erlenmeyer flask containing 100 ml. of hot, neutral 95% alcohol. Cover the
flask with a watch glass and dissolve the soap as rapidly as possible by heating
on the steam bath. Add a few drops of phenolphthalein and titrate rapidly
with N/4 H 2S0 4 without filtering. Take the first endpoint"disregarding any
subsequent retur~ of color. Calculate to free NazO (or NaOH) or KzO (or
KOH) as the character of the soap indicates. If -the alcoholic solution is acid
to phenolphthalein, titrate with N/10 NaOH and calculate the percentage of

ANALYSIS OF SOAPS AND SOAP PRODUCTS

2033

free fatty acid as oleic acid, or lauric acid if coconut soap. Save the solution
for determination' of carbonates.
(b) When" Builders" Are Determined. a. 2._Weigh a 10.02 gm. sample
and dissolve in 200 ml. of hot, neutral 95% alcohol. Filter through a 9 cm.
filter paper into a 500 ml. Erlenmeyer flask and wash the residue three times
with hot, neutral 95% alcohol. Protect from CO 2 and other acid fumes during
filtration. If the soap contains borax, or phosphates, it can be washed with
alcohol indefinitely without removing the apparent alkalinity. Cover the flask
with a watch glass and heat to incipient boiling. Titrate with N/4 H 2S0 4 and
report as "free N a 20 (or N aO H) " or "free KzO (or KO H) " as the character of
the soap indicates. If the alcohol is not pink, titrate with N/lO NaOH and
report "free fatty acid" as oleic acid.
If potassium carbonate is present, the free KOH results will be high due to
the solubility of KzCO a in the alcohol. For liquid soaps containing free KOH
and K 2CO a the following procedure has given satisfactory results: Weigh a 20
gm. sample of the soap solution into a 500 ml. Erlenmeyer flask. Add 10 gms.
of ~eutral dry salt. Mix the salt with the soap thoroughly by shaking. Allow
to stand 10 minutes covered by a watch glass. Add 20 m!. of hot neutral 95%
alcohol. Shake thoroughly and place on the steam bath for 10 to 15 minutes,
shaking occasionally. Filter through a neutral filter paper moistened with
neutral alcohol into another 500 m!. Erlenmeyer flask. Retain most of the
insoluble in the flask, if possible. Wash with 25 m!. of hot alcohol, pouring
around the top of the paper so as to wash the paper at the same tim'e. Give
the paper a slight additional wash with neutral alcohol after the first has drained
through. Allow to drain. Titrate with N/4 or Nj2 acid, and from this
calculate free KzO.
Sodium Carbonate

(a) If No "Builders" Are Present.-Add. from a burette 10 ml. of Nil

H2,S04 to the solution in the flask from the free alkalinity titration (a). Add 50
ml. of distilled water and boil gently for 1 hour, swirling the flask occasionally.
Remove the flask, add 10 m!. Nil NaOH and titrate with N/4 H 2S0 4 to the
colorless endpoint. If this titration is more than 7.5 ml., add another 10 ml.
portion of Nil H 2S0 4 and again boil for 1 hour. Add 10 m!. Nil NaOH and
again titrate with N/4 H 2S0 4 Calculate the per cent carbonate from the acid
titration.
(b) If" Builders" Are Present.-Dissolve the residue upon the filter paper
from the free alkalinity determination (b) with hot water into the original
beaker. If no insoluble matter other than insoluble Si0 2 from the builder is
present, the filter paper should be punctured and the entire contents of the paper
washed through. CQntinue washing until a few drops of phenolphthalein
indicator dropped on the filter paper show only a very faint pink color. Cool
the filtrate'and titrate with Nil H 2S0 4 using methyl orange indicator.' Record
the titration as the "carbonate titration" and in case no silicate is present,
calculate direct to sodium carbonate. If silicate is present, it should be determined as indicated in the" Note" under" Determination of Silica Present as
Alkaline Silicates." The above titration must then be corrected for the silica
present in calculating the carbonate.

2034

ANALYSIS OF SOAPS AND SOAP ,PRODUCTS

If potassium carbonate is present, it can best be ,determined from the COx

determination.

Determination of Carbon Dioxide (Gravimetric Method -for Carbonates)

For most cases the volumetric methods described \above are satisfactory,
but when the highest accuracy is desired, a reliable absorption method should
be used. The following method (A. C. S.-A. O. C. 'S.2. 3) has proved satisfactory.
A 250 ml. Erlenmeyer flask is placed on a gauze over a burnet. The flask is
equipped with a 2-hole rubber stopper, through one opening of which is a lO-inch
reflux condenser and through the other a thistle tube equipped at the outer eIid
with a 3-way stopcock. The lower end of the thistle tube is drawn to a sma)l
point, which is placed very close to the bottom of the flask. Then the straightaway end of the stopcock is attached to a small funnelfor the introduction of acid
to the flask. The other opening of the stopcock is attached to receive air from
a purifying train consisting of a wash bottle containing concentrated sulfuric
acid, and a second at the outer end of the train containing a 50% solution of
potassium hydroxide. The top of the reflux condenser is attached first to a
drying wash bottle containing concentrated sulfuric acid, and then to a weighed
absorbing train consisting of a suitable potasJ:l/bulb charged with 50% potassium
hydroxide, and a second containing concen,trated sulfuric acid. This train is
attached to a protective U-tube containing calcium chloride. The U-tube is
attached to an aspirator.
/
.
Procedure.-Set up the apparatus, leaving out the weighed train, and
aspirate with a slow stream of the dry carbon dioxide-free air until the apparatus
is freed of carbon dioxide. Insert the train and continue the aspiration for
Y2 hour. Check the weight of the train to determine if the air is passing
through too fast, or if the system is free from carbon dioxide. The system
must be free from leaks. Weigh out 1 or 2 gms. of the sample into the Erlenmeyer flask, cover ,vith 20 ml. freshly boiled distilled water, close the apparatus
with the train in place. Add 20 ml. dilute hydrochlori,c acid (1 : 1) thrQugh the
funnel very slowly, with no heat being applied to the flask. The rate of adding
acid should be carefully controlled so that the gas does not pass through the
train too rapidly. As soon as the acid is added, 'start aspiration gently. -When
the absorption begins to stop the gas flow, start heating gently and continue
until the contents of the flask have boiled 15 to:20 minutes. Stop heating and
continue aspirating until,the flask has cooled down. Remov_~ the train and
weigh. Calculate increase of weight as carbon dioxide. Carbon dioxide
multiplied by 2.41 equals sodium carbonate.
All connections should be made glass to glass. Apparatus with ground
glass joints is available and'is preferred. Ascarite-filled absorption bulbs and
scrubbing tower instead of those using KOH are very satisfactory and convenient.
!
Determination of Silica Present as alkaline Silicates

2. 3

When the material contains no mineral matter that is insoluble in water,*

ignite a sample of the soap containing not to exceed 0.2 gm. of silicate in a
* When water-insoluble mineral matter is present,.tak;r-;;,portion of the solution after
titration in (b), "Sodium Carbonate" equivalent to not more than 0.2 gm. silica, add
5 m!. concentrated HCI and proceed as above.

ANALYSIS OF SOAPS AND SOAP PRODUCTS

2035

platinum dish at a low temperature. When charred, extract the soluble salts
with water, return the paper and charred residue to the dish and complete the
ignition. Unite the residue in the dish and the water extract, carefully acidify
with hydrochloric acid, finally adding the equivalent of from 5 to 10 ml. conce'ntrated hydrochloric acid in excess. The dish or casserole containing the
,solution should be covered with a watch glass while adding acid so as to avoid
loss by spray.
Evaporate the acidified solution (washing off and removing the cover glass
if used) to dryness on steam bath or hot plate at a temperature not exceeding
120 0 C. Cool, moisten with concentrated hydrochloric acid, let stand 5 to 10
minutes, breaking up all lumps with a stirring rod. Add about 25 m!. of hot
water. Heat a few minutes and filter through a small ashless paper. 'Wash
thoroughly with hot water.
Evaporate the filtrate to dryness and repeat the above treatment, filtering
on a second paper*
Carefully dry the filter papers and residue at 150 0 C. in a tared platinum
crucible, ignite ~t as Iowa temperature 8 as possible until the paper is consumed,
then raise the temperature and heat to constant weight in a muffle or over a
blast lamp. Cool in a desiccator and reweigh. If extreme accuracy is desired, moisten the residue in the crucible with water, add 5 m!. hydrofluoric acid
and 4 drops of concentrated sulfuriC' acid, evaporate to dryness, and ignite as
before for about 2 minutes. Cool in a desiccator and reweigh. The difference
between this weight and the previous weight is the weight of Si0 2 9
To calculate sodium silicate (1 Na 20 : 3.25 Si0 2) multiply Si0 2 by 1.308.10
. NOTE.-In case carbonates have been determined as under (b), the filtrate may be
used for the silicate determination. In that case, add 5 mI. of concentrated hydrochloric
acid to the solution in the beakllr, evaporate to dryness and proceed as above.
Borax

Qualitative Test.-A qualitative test may be made by wetting a piece of

turmeric test pll:per with a few m!. of the Si0 2 filtrate. If the paper on drying
in air turns a brick red, a borate or perborate is present. To determine if the
coloration is due to perborate, dissolve about 2 grams of the original sample
in about 100 ml. of a 5% potassium iodide solution containing 2 m!. of starch
solution. Add 10 m!. of 1 : 4 sulfuric acid and stir. A blue solution denotes
the presence of an oxidizing agent which with a positive turmeric test can be
assumed as indicating the presence of perborate.
Quantitative Test.~To determine borate quantitatively in case silicates are
not present, weigh 1O.02 grams of the soap (5.01 grams if more than 5%
borax is present), transfer to a 250 m!. beaker and break up the SOl\p with 50 m!.
of water and 50 m!. of N /1 H 2S0 4 If the solution is not. distinctly red to
methyl orange, add N/1 H 2 S0 4 until it is acid. Do not heat longer than
* The filtrate may be collected in.a 250 mI. volumetric flask and used for qualitative
and quantitative phosphate.
S K. A. Krieger and H. S. Lukens, Ind. Eng. Chern., Anal. Ed., 8, 118 (1936) .
.9 W. F. Hillebrand, The Analysis of Silica,te and Carbonate Rocks, U. S. Geological
Survey, Bulletin 700, 102.
10 Report of the Soap Analysis Committee, A. O. C. S., Oil and Soap, 12, 10 (1935).

2036,

ANALYSIS OF SOAPS AND SOAP PRODUCTS

necessary. Filter o.ff the fatty acids through a wet paper, wash well with hDt
water, adding all washings to. the filtrate. Neutrali:ze to. methyl Drange with
Nil NaOH and nDte the reading. Add a few drDps o.f phenDlphthalein and
titrate to. the endpDint with Nil NaOH.
Add abDut 1 gram o.f neutral mannitDl and again titrate very SlDwly. Add 1
gram mDre Df the mannitDl and titrate again and repeat this prDcess until
mannitDl causes no. further actiDn Dn the endpDint. If phDsphate and bDrate
are bDth present and no. silicate, add 1 gm. Df sDdium chlDride with the mannitDl
befDre titrating; Calculate the number Df milliliters Df sDdium hydrDxide
added frDm the methyl Drange endpDint to. the final endpDint Df % Na 2B40 7
by multiplying by 0.503 Dr to. % Na zB 40 710H zO multiplying by 0.954.fDr:1O
gm. samples. In case phDsphate is present, a cDrrectiDn must be made since
part Df the phDsphate is titrated in the borate titration. For a 10 gm. sample
the following formula is used:

["Borate Titration"-(% .Na aP0 4 XO.61)]XO.503=% NaZ_B407

Since silicic acid affects the phenolphthalein and methyl orange endpDints in
the mannitol titration, it must be removed by the following method. Weigh
a 5 gm. sample in a tared platinum dish and ignite over a IDW flame until all
of the Drganic matter has been burned Dff, but do not ash the salts. CDDl the
dish and CDntents and place it in a beaker. 'Add abDut 100 ml. Df water and bDil
until the residue is remDved frDm the dish. Wash the contents Df the beaker
into. a 500 ml. distillatiDn flask, add 15 ml. Df cDncentrated hydrochlDric acid,
and distill to. dryness. PDur the distillate back into. the distillatiDn flask and
repeat. CDDI and wash the cDntents Df the flask into. the beaker cDntaining the
distillate. Filter thrDugh a quantitative filter paper and wash well. Stir the
filtrate with a stick Df caustic until nDt quite neutral to. methyl Drange. Neutralize with Nil NaOH and cDntinue titratiDn as abDve with.mannitDl. Correct calculatiDns fDr phDsphate in the same manner as abDve.
I

Perborates

Weigh a 1 gm. sample and transfer to. a 300 ml. Erlenmeyer flask with cDld
water. Make up to abDut 80 ml. Acidify immediately with 20 ml. Df Nil
sulfuric acid and titrate at o.nce with N 110 Po.tassium permanganate. One ml.
Df pDtassium permanganate is equivalent to. 0.0008 gms. Df o.xygen, 0.007694
gms. Df NaBO a 4H 2 0, and 0.00409 gms. Df NaBO a
Phosphates
A Qualitative Test for Phosphate May Be 'Made on the Filtrate from the
SiO z Determination.-If meta- Dr pyrDphDsphates are present add abDut 10 ml.
of nitric acid and hDld at near bDiling temperature' Dver night to. convert them
to. DrthDphDsphates. To. 20 ml. add 2 ml. Df cDncentrated nitric acid and 5 ml.
Df 30% NH 4 NO a so.lutiDn, phDsphate-free. Heat to. bo.iling and add 20 m!. o.f
ammDnium mDlybdate sDlutiDn. Let the sDlutiDn stand 20 minutes Dn the
steam bath. If a yellDw precipitate (Drms, a phDsphate is ,present.
Quantitative DeterminatiDn ~. 3-Pipette an llliqUDt cDrrespDnding to. 0.50
gm. Dr 1 gm. into. a 250 ml. beaker; add ammDnium hydrDxide in slight excess;

ANALYSIS OF SOAPS AND SOAP 'PRODUCTS

2037

and barely dissolve the precipitate formed with a few drops of nitric acid, stirring vigorously. Add about 15 gms. of dry ammonium nitrate or a solution
containing that amount.
To the hot solution add 70 ml. of the molybdate solution for every decigram
of phosphoric acid (P 2 0s) present. Digest at about 65 C. for an hour, and
determine if. the phosphoric acid has been completely precipitated by the addition of more molybdate solution to the clear supernatant liquid. Filter, and
wash With cold water or, preferably, 10% ammonium nitrate solution. Dissolve the precipitate on the filter with ammonium hydroxide U : 1) and hot
water, and wash into a beaker to a bulk of not more than 100 ml. Neutralize
with hydrochloric acid, using litmus paper or bromthymol blue as an indicator,
cool, and from a burette add slowly (about 1 drop per second), stirring vigorously, 15 ml. of magnesia mixture * for each decigram of phosphoric acid (P205)
present. After 15 minutes add 12 ml. of ammonium hydroxide (sp.gr. 0.90).
Let stand till the supernatant liquid is clear (2 hrs. is usually enough) filter,
wash the precipitate with dilute ammonium hydroxide (1 : 10) until the washings are practically free from chlorides; dry, burn first at a low heat and ignite
to constant weight, preferably in an electric furnace, ,at 950-1000 C.; cool
in a desiccator, and weigh as Mg 2 P 2 0 7 Calculate and report the result as
percentage of P 2 0 5 or alkaline phosphate known to be present.
Sulfates
Weigh a 50 gm. sample, dissolve in water, break up the soap with 35 ml. of
concentrated hydrochloric acid and filter off insoluble matter and fatty acids
through a wet filter paper. Wash the filter well with cold water. If silicate is
present, remove it by evaporation and filtration as in the silicate determination.
In either case make up the silicate-free filtrate to 175 ml. with water, add 5 ml.
of hydrochloric acid and boil. 'To the boiling solution, add 15 ml. to 20 ml.
of a 10% barium chloride solution, drop by drop. Continue boiling until the
precipitate is well formed and keep at a temperature of about 70 C. for one
hour or unt~l the precipitate settles well. Filter through a prepared Gooch
crucible, ignite gently, and weigh as barium sulfate. Calculate to sodium sulfate or to sulfate known to be present.

Total Alcohol Insoluble Matter

2, 3

Digest hot a 2 to 10 gm. (0.01 gm.) sample with 200 ml. of freshly boiled
ethyl alcohol neutral to phenolphthalein (94% or higher). Filter through a
counterpoised filter paper neutral to phenolphthalein, or a weighed Gooch
crucible with suction, protecting the solution during the operation from carbon
dioxide and other acid fumes. Wash the residue on the paper, or in the crucible,
with hot neutral alcohol until free from soap. Dry the filter paper, or crucible,
and residue at 100 to 1050 C. for 3 hours, cool, and weigh the total matter
insoluble in alcohol. (The matter insoluble in alcohol will contain most of the
alkaline salts, such as carbonates, borates, silicates, phosphates and sulfates,
as well as starch, and may be used for the approximate determination of these
* Magnesia mixture-Dissolve 55 gms. of crystallized magnesium chloride (MgCh6H 20) in water, add 140 gms. of ammonium chloride, 130.5 mI. of ammonium hydroxide
(sp.gr. 0.90), and dilute to I liter.

2038

ANALYSIS OF SOAPS AND SOAP PRODUCTS

constituents. These salts are not entirely insoluble in alcohol, so for accurate
determinations separate portions of the soap should qe us.ed.)
Water Insoluble Matter

2, 3

Proceed as in the determination of matter insoluble in alcohol. After

filtering and thoroughly washing the residue, extract! it with water at 60 0 C.
and wash the filter thoroughly. (When the matter 'insoluble in water is all
inorganic, boiling water may. be used for the extraction and washing.) Dry
the filter and residue at 100 to 105 C. for 3 hrs., cool, and weigh matter
insoluble in water. The nature of this matter may be determined by furtlulr
examination.
Glycerine,

Star~h,

and Sugar

The sample sizes indicated are for ordinary soap. If soaps high in glycerine
or sugar are analyzed, sample sizes should be reduced and the amount of bi..
chromate arid acid increased. Transparent soaps are frequently high in glycerine and sugar.
1. Glycerine when starch and sugar are not present: Weigh a 25 gm. sample,
transfer to a 400 ml. beaker and dissolve the soap in about 100 ml. of water.
Add about 25 ml. 1 : 4 sulfuric acid and break up the soap. Do not heat longer'
than necessary. Filter off the fatty acids, collecting the filtrate in a 250 ml.
volumetric flask. Make up to volume * and pipette 50 ml. into a 250 ml.
Erlenmeyer flask containing 0.7456 gm. of dry potassium bichromate of highest
purity (equivalent to 0.1 gm. of glycerin), 10 ml. of water, and 15 ml. concentrated sulfuric acid. Cover the flask with a watch glass and heat in a steam
bath for 2 hours at a temperature of 93 to 100 C.t Cool, and titrate with,
standardized ferrous ammonium sulfate (about 300 gms. per liter), using potassium ferricyanide as an outside inuicator, or by use of an electrometric titration
apparatus. The ferrous amm6nium sulfate solution must be of such strength
that 0.7456 gm. of potassium bichromate will be required for complete oxidation
of the iron in 19.5-20.0 ml. of the solution. It sJiould contain 100 ml. of
sulfuric acid per liter of solution. A blank test must be run exactly the same
with omission of the sample.
2 (Blank titration -Sample titration) _
l'
't t'
. - per cent g ycerme.
BIan k t I ra Ion
NOTE.-The electrometric titration apparatus (Ind, Eng. Chem"f'~nal. Ed. 9, 514
(1937)) is more rapid and in the hands of the average analyst more accurate than the
outside indicator with the spot plate. The preferred set-up is one using platinum and
tungsten electrodes. By a suitable resistance the galvanometer is adjusted to zero.
At the endpoint the needle is deflected to the end of the scale. A glass stirrer directly
connected to a suitable motor is used, By use of a wide-mouthed Erlenmeyer for
oxidation flask, there is ample room for the electrodes, stirrer, and burette tip which is,
usually, extended and bent to allow the burette to be s~veral inches to one side of the
motor. A calomel half-cell may be used instead of the 'tungsten electrode with the use
'
of N /1 Kel solution.
* If appreciable amounts of salt or soluble fatty acids are present, they may be precipitated by addition'of C.25 gm. of silver sulfate at this point. Shake well, fet settle and
filter a 50 m!. portion for analysis.
"_
'
t If the solution turns blue during oxidation, or if tRe-tiack titration is less than 4 ml.,
1.4912 gms. of potas8ium bichromate must be used or a smaller sample. taken.

ANALYSIS OF SOAPS AND SOAP PRODUCTS

2039

2. Glycerin in presence of starch but no sugar:2; 3. 11 Weigh a 25 gm. sample,

dissolve inlOO ml. of warm water (not above 70 C.), filter off the water-insoluble matter, and wash the residue twice with warm water (not above 70 C.).
Add 25 ml. of 1 : 4 sulfuric acid to the filtrate to break up the soap. Filter off
the fatty acids and: wash thoroughly, catching the filtrate in a 250 ml. volumetric flask. Determine the amount of glycerine by the method above. Make
to volume and pipette 50 ml. into the Erlenmeyer.
3. Determination of sugar (Munson and Walker Method):t Proceed as
for glycerin above, filtering off any starch or insoluble matter, but using hydrochloric acid (1 : 1) to break up the soap and boiling to invert the sugars.
Transfer to a 250 ml. volumetric flask, neutralize to litmus with N aOH, and'
make to volume. Transfer 25 ml. each of copper sulfate and alkaline tartrate
Fehling's solutions to a 400 ml. beaker and add an accurately measured 50 ml.
aliquot of the invert sugar solution. Heat so that boiling begins in four minutes, and continue boiling for exactly two minutes. Keep the beaker covered
during boiling. Filter the cuprous oxide at once in a prepared Gooch crucible
and 'Wash thoroughly with hot water (about 60 C.), and finally with 10 ml. of
alcohol and 10 ml. of ether. Dry for 30 minutes at 1050 C., cool and reweigh.
Refer to Table 9, Chapter XLII, Methods of Analysis, A. O. A. C. (1935) to
determine the amount of invert sugar equivalent to the weight of cuprous oxide.
The weight of invert sugar multiplied by 0.95 equals the_weight of sucrose.
4. Determination of glycerine in presence of sugar: 11 If starch is present, it
must be filtered off as directed in (2); then proceed as directed in (1), except boil
the ;olution when acid 20-30 minutes to secure complete inversion and to remove alcohol, increase the bichromate to 2 to 3 times the amount specified
and the H 2S0 4 to 25 ml. Determine the weight of invert sugar as directed in
(3). The oxidation of sugar is taken care of by calculating the glycerin by the
following formula for 25 gm. sample made to 250 ml. and 50 ml. aliquot.

2n (Blank titration-Sample titration)

Blank titration

17.51 W =per cent glycerine

when W is the weight of invert sugar in an aliquot equal to a 5 gm. sample, and
n is equal to number of times 0.7456 gms. of bichromate is used.
5. Determination of starch (A. C. S.-A. O. C. S.): 2, 3 Separate the waterinsoluble matter as directed in (2) using a weight of sample that will give less
than 3 gms. of starch .. Transfer the insoluble matter, 'without drying, to a 500
ml. Erlenmeyer flask with hot water. Make up to 150 ml. and add 20 ml. of
concentrated hydrochloric acid. Boil for 2Yz hours under a reflux condenser,
nearly neutralize with sodium hydroxide, and cool. Transfer to a 250 ml.
volumetric flask, make up to volume and determine reducing sugars as directed
in (3). Refer to the amount of dextrose equivalent to the weight of cuprous
oxide. This weight multiplied by 0.90 equals the weight of starch.
11

L. F. Hoyt and H. V. Pemberton, J. Ind. Eng. Chern. 14, 54 (1922); correction,

1:4, 340 (1922).

t L. S. Munson and P. H. Walker, J. Am. Chern. Soc., 28, 663 (1906). P. H.

Walker, 29, 541 (1907). Official and Tentative Methods of Analysis of the A. O. A. C.,
1935, 479, 626.
.

2040

ANALYSIS OF SOAPS AND SOAP PRODUCTS

Volatile Hydrocarbons

2, 12~ 13

This method requires a source of dry, oil-free stearh which is passed through
the sample treated with acid, sufficient to liberate the fatty acids from the soap.
The steam is next passed through strong caustic sqlution to scrub out any
volatile fatty acids while the volatile hydrocarbons I are condensed with the
steam in a suitable arrangement which allows the excess water to flow away,
leaving the volatile hydrocarbon in the measuring burette. The method may
be applied to samples containing substances immiscible with water and volatiie
with steam. For solvents heavier than water a Bidwell-Sterling tube should be
used.
"
C

E
Steam

~====~=======~
~==~

To Sewer

Steam

VOLATILE

HYDROCARBONS APPARATUS

A - Steam Trap
B - Sample Flask
C- Dropping Funnel
D-Casu.tic Scrubber-Iron
E - Kieldahl Trap
F -Condenser
G-Siphon
H-Burette
I - Expansion Bulb
J - Automatic Overflow

FIG. 287
Apparatus.~The apparatus and its arrangement are shown in the sketch.
The following are the important items, lettered to correspond to the lettering of
the sketch (Fig. 287).
. .J ~ -

Determination of Volatile Hydrocarbons in Soaps, Oil and Soap, 13, 9" (1936).
13 Report of Soap Analysis Committee A. O. C. S., 1935, Oil and Soap, 13, 7 (1936).
12

ANA;LYSIS OF SOAPS AND SOAP PRODUCTS

20flJ

Steam Trap, A, a liter round-bottom ring-neck flask equipped with a siphon

tube to the drain from the bottom of the flask and provided with a means of
regulating the steam flow into the flask.
Evolution OJ: sample flask, B, a liter round-bottom ring-neck flask. In case
large samples are desirable the size of this flask may be increased.
Caustic scrubber flask, D, a steam jacketed metal flask is preferred, but a
liter Florence flask provided with a steam coil of Ys" copper tubing around the
upper half may be used. If the glass flask is used it should be provided with a
safety bucket below it and should be renewed frequently since the strong caustic
dissolves the glass rather rapidly. This flask should be connected to the condenser by a Kjeldahl c.onnecting tube, E, or similar safety device.
The inlet tubes for the steam into the evolution and scrubber flasks should
extend nearly to the bottom of the flasks and be bent at right angles and parallel
to the sides of the flask.
Condenser, F, a 12" or longer spiral' condenser of sufficient bore so the
condensate will not readily close it.
easuring burette, H; a 10 ml. burette calibra~ed to 0.1 ml. and carrying a
bulb, I, approximately 100 ml. capacity, at the lower end.
The stoppers used should be of a good grade of rubber and should have been
thoroughly cleaned free from any surface sulfur and should be given a steam
distillation in position for several hours before use on a sample.
Insulating the flasks and tubing to reduce condensation aids distillation and
its control.
Determination.-Place 150 ml. NaOH solution (about 46 Be) and several
sticks of solid NaOH, to provide a:gainst dilution, in the scrubber flask. Rinse
_out the condenser and burette with acetone. Attach a rubber tubing to the
lower end of the burettE), fill the burette and tubing with water and raise the
outer end of the tubing so that the water level in the burette is near the top of
the scale when the water is flowing to the drain from the automatic overflow, J.
Be sure the connections a tight and that the tubing contains no air bubbles.
Place the condenser in po . ion so the lower end extends directlyinto the upper
end of the burette just ab ve the water level or connect to an adapter siphon,
G, which discharges into the burette. The cooling water should be 15.5 C. or
colder. Ice water may be desirable for low boiling hydrocarbons.
Weigh 1000.5 gms. of the soap (cut into cubes of about 1 cm. edges) or
500.3 gms. of soap powder and transfer to the evolution flask. Add about
10 gms. of gum arabic (commercial) and 100 ml. of distilled water. Place the
- flask in position with 100 ml. of 1 : 3 H 2S0 4 in a dropping funnel, C, carried in
the stopper. Connect the steam evolution, and wash flasks and condenser into
position, making sure that the stoppers are tight-fitting and held in place by
wiring. Rubber connections in the lines between the evolution flask and
condenser should be avoided.
Add the acid to the sample slowly to avoid excessive frothing. While
adding the acid, turn on the steam cautiously so adjusting the pressure by a
bleeder valve that just enough steam flows to prevent any liquid backing lito
the steam trap flask.
When all the acid has been added, turn on enough steam to cause brisk
distillation, taking care that no liquid is carried over from the evolution and
wash flasks and that the condenser water does not become warm.

20}!2

ANALYSIS OF SOAPS AND SOAP PRODUCTS

Continue the distillation until there is no increase iIi the volume of the upper
layer for 45 minutes or no small droplets can be note4 in the condensate ..
When distillation is completed, shut off and drain condenser water and allow
the steam to heat up the condenser to drive out the las~ traces of volatile hydrocarbon. Shut off the steam as soon as vapor begins tolissue from the lower end
of the condenser. Immediately open the stopcock ofl the dropping furinel to
prevent caustic being drawn into the evolution flask.,
.
Stopper the burette and allow its contents to come to room temperature or
bring them to a definite temperature by placing the burette in a water bath heid
at 25 C. for one to two hours.
Read the volume of the upper layer to the nearest .011)1l. The volume times
the specific gravity equals the weight of the volatile hydrocarbon. The specific
gravity should be determined at the temperature at which the volume is read.
A small Sprengel tube made of 3 mm. glass tubing is convenient for this purpose.
Calculation.ml. of volatile hydrocarbon X specific gravity XIOO
Weight of sample
= per cent volatile hydrocarbon.
For some samples the volatile hydrocarbon dontent may be so low that a larger
sample than 50 or 100 gms. is desirable. The size of the evolution flask may
need to be increased if larger samples are used. The amount of water in the
evolution flask and acid used should also be corr<;:spondingly increased.
Titer Test
1. Preparation of Total Fatty Matter (Fatty and Rosin Acids and Unsaponifiable Matter.-Dissolve 200 gms. of soap * in 200 ml. of water in a liter
Erlenmeyer flask, add 20 ml. of 1 : 3 sulfuric acid and heat until the fatty
matter collects in a clear layer.
The flask should be covered with a watch glass. / Much time is saved by
placing the sample in a liter beaker and using,a motor-driven stirrer to keep
constant agitation. Siphon off the acid water layer and decant the fatty matter
through a dry filter paper into tall form, lipless'lS0 ml. beaker. If droplets of
water can be seen, refilter the fatty acids through a dry filter paper.
2. Determination.-A standard titer therniometer: 14, 15 graduated at zero
and in tenths degrees from 10 to 65 C. is used or in .20 f1'om~2 C. to 66 C. _
It should be certified by the United States Bureau of Standards, or carefully
standardized against a certified thermometer. Special thermometers (graduated in tenths degree Centigrade) in 20 C. ranges from- 00_i:!0 C. may be used.
After the fatty acids are dry, cool carefully by immersing in and out of water
at about 17 C. until a cloud begins to form, stir,ring continuously with the
thermometer (about 150 r.p.m.) so that no crust forms on the sides or bottom

" * The fatty matter obtained is also used for the iodine value, saponification value,
and may be used for rosin determinations. If the soap contains superfat, it must be
saponified with glycerol-potash as in the titer determination for fats.l0
14 Official and Tentative Methods of Analysis of the Association of Official Agricultural Chemists, 4th edition, 1935, 408.
' 15 J. Ind. Eng. Chern., 18, 1349 (1926).

ANALYSIS OF SOAPS AND SOAP, PRODUCTS

2043

of the beaker. Care must be taken not to cool too rapidly or low results will
be obtained. Hold the beaker in the air and continue stirring more slowly
(about 100 r.p.m,) until the mercury remains stationary for 30 secon{ls or
begins to rise within that time. Wipe the outside of the beaker dry and care-'
fully place the beaker and contents in a suitable container (air-jacket) submerged in water, or in ice, salt, water mixtures when necessary"maintained at a
temperature of 10 C. below the titer. For titers above 30 C. the temperature
should be 20 C. Allow the thermometer to hang without any disturbance
with the bulb in the center of the fatty acids and' follow the thermometer
readings. Record the highest point to which the mercury rises as the titer of ,
the fatty acids.
When not enough soap or fatty matter is available, the official method of the
A. C. S.l5-A. O. C. 8.3 should be used. This method, which differs slightly in
details 'and requires less fatty acids, may also be used on the fatty acids obtained in the above procedure.

Iodine Value 15
The iodine value is determined on the fatty matter obtained in the above
procedure for titer test by the Wijs' method. If rosin is present this method
cannot be used.* The solutions required are Wijs' iodine monochloride solution,1. 15 standard Nj10 sodium thiosulfate, 5% potassium iodide solution,
U. S. P. chloroform, and starch indicator. Weigh exactly a sample of such size
that at least 100% to 150% of the amount of iodine absorbed remains in excess,
on a watch glass and transfer to ,a clean, dry 500 ml. glass-stoppered bottle,
washing the watch glass thoroughly with 20-30 ml. of chloroform. Add 25 ml.
of iodine mono chloride and at the same time prepare a blank with the same.
amount of chloroform and iodine mono chloride. Mix the contents gently and
immediately put the bottles in a dark place at a temperature of 75-90 F. for
exactly 30 mi.nutes. Add 20 ml. of 5% potassim}l iodide to the sample and
blank followed at once by 100 ml. of water. Titrate with Nj10 sodium thiosulfate, using starch solution as an indicator. Do not add starch until the
solutionhas a pale straw yellow color. Toward the end of the reaction stopper
the bottle and shake violently. The number of cubic centimeters of standard
thiosulfate solution used for the blank minus the amount used for the sample,
gives the thiosulfate equivalent of the iodine absorbed by the amount of sample
used in the determination. Calc!J.late to per cent iodine absorbed. Mechanical agitation using a small variable-speed motor attached to a glass stirrer is
an advantage. It is necessary to use a wide mouthed glass-stoppered bottle.

Saponification Value
Weigh exactly a 5 gm. sample of the fatty matter obtained in the titer determination and wash into a 300 ml. ErlenqJ.eyer flask with hot, neutral, redistilled

* In case the iodine value must p~ obtained on samples containing rosin, a 0.1 gm.
sample is dissolved in 20 ml. of glacial acetic acid and 20 ml. chloroform in the glassstoppered bottle. Immerse the bottle in water at a: temperature of 21.5-22.5 C. for
30 minutes. Add the Wij,s' solution as above and maintain the temperature at 21.522.5 C. for one hour. Other, operations are the same as above.
16 P. C. McIlhiney, J. Am. Chem. Soc., 29, 1222 (1907).

2044

ANALYSIS OF SOAPS AND SOAP PRODUCTS

95% alcohol and add 50 ml. of an approximately 0.7IN alcoholic-potash solution. Prepare a blank in the same manner, omitting the sample. Insert a
funn~l in the neck of the flask and saponify for exactly one hour, swirling the
contents occasionally. Cool quickly and titrate at once with N/2 hydrochloric
acid, using phenolphthalein indicator. Calculate thei,milligrams of potassium
hydroxide required to saponify one gm. of fat by multiplying the difference
between the blank and sample titrations by 5.61 and' report as saponification
value.
Acid Value
Weigh a 2 gm. sample of the fatty matter obtained in the titer determination
and transfer to a 300 ml. Erlenmeyer flask with 50 ml. of hot, neutral 95%
alcohol. Titrate with N /1 alkali, using phenolphthalein indicator. Calculate
the milligrams of potassIum hydroxide per gm. of fat by multiplying the ml. of
Nil alkali used by 28 and report as acid value.
Rosin in Fatty Acids

17

CA. C. S.L-A. O. C. S.lO

Wolff's Method, Modified-Preparation of Fatty and Rosin Acids.-Dissolve

5 gm.O.Ol gm. of the sample in 100 to 200 ml. of hot distilled water in a 250
ml. bell,ker, add a slight excess of dilute sulfuric acid (1 : 1), heat on the steam
bath until the fatty acids collect in a clear layer, cool to room temperature,
transfer to a separatory funnel, washing the beaker free from fatty and rosin
acids with small portions of ethyl ether, adding the ether washings to the separatory funnel. Add about 50 ml. of ethyl ether to the separatory funnel which is
then given a whirling movement to dissolve the fatty and rosin acids. After
the solid acids are dissolved in the ether, allow the aqueous solution to stand
for at least five minutes. Draw off the aqueous portion into another separatory
funnel designated as No.2, add about 30 ml. of ethyl ether to funnel No.2,
shake vigorously, allow to stand for at least five minutes and then draw off the
aqueous portion into another separatory funnel desigl}~ted as No.3. Transfer
the ether extract to separatory "funnel No. 1. Wash the aqueous liq~id three
more times, using for each washing about 30 ml. of ethyl ether. Then discard
the aqueous liquid and add the ether washings to.funnel No. 1. N ow add about
50 ml. of distilled water to the collective ether ,extract in funnel designated as
No.1, shake vigorously and allow to stand for five milllltes. Draw off the
aqueous solution into another separatory funnel. Repeat t'hii:r'operation, combining the water washings until the ether is free from acid, using' !llethyl orange
as indicator. Add 50 ml. of ethyl ether to the water washings, shake the
separatory funnel vigorously and allow to sta.nd for at least five minutes.
Draw off and discard the aqueous solution and wash the ether layer with small
portions of distilled water until free from acid, using methyl orange lj.S indicator.
Transfer the total ether extract to a 250 ml. Erlenmeyer flask, evaporate off the
ether slowly on a steam bath and dry one hour in an oven at 10,5 C., cool and
dissolve in 20 ml. of absolute alcohol.
First Esterification.-Add 10 ml. of a solution of one volume of concentrated
sulfuric acid (sp.gr. 1.84) and four volumes of absolute ethyl alcohol, bring to

_.-

17

Chemical Ztg., 38, 369, 382, 430 (1914); C. A.,

8, 2495

(1914).

ANALYSIS OF SOAPS AND SOAP PRODUCTS

2045

boil on the steam bath and boil exactly four minutes under a reflux condenser.
Remove from the steam bath, add to the liquid about five times its volume of 7
to 10% sodium chloride solution. Transfer to a separatory funnel, washing
the Erlenmeyer flask with small portions of ethyl ether, the ether washing being
added to the aqueous solution in the separatory funnel. Add about 30 ml.
of ethyl ether to the separatory funnel, shake vigorously, and allow to stand
for at least five minutes. Draw off the aqueous solution into another separatory funnel designated as No.2. Add about 30 ml. of ether to funnel No.2,
shake vigorously, allow to stand for at least five minutes and then draw off
the aqueous layer into another separatory funnel designated as No.3. Repeat
the extraction of the aqueous solution three more times. Then discard the
aqueous liQuid and add the combined ether solutions to separatory funnel No. 1.
Add 50 m!. of sodium chloride solution (7-10%) to the ether extract in funnel
No.1, shake vigorously and allow to stand for at least 5 minutes.
Draw off the aqueous solution into another separatory funnel. Repeat this
operation until the washings are neutral to methyl orange, all salt water washings. being combined. ' Add 50 m!. of ether to the washings in the separatory
funnel. Shake the funnel vigorously, allow to stand for at least five minutes
and then draw off and discard the aqueous salt solution. Add 50 ml. of sodium
chioride solution and wash the ether layer until the. aqueous salt solution is
neutral to methyl orange. 'Dry the ether extract by the addition of about 3
gm. of anhydrous sodium sulfate to the separatory funnel, filter into a 250 m!.
Erlenmeyer flask and evaporate off the ether slowly on the steam bath to
dryness.
.
.
'
Second Esterification.-Coollj,nd dissolve the residue in 20 m!. of absolute
ethyl alcohol and then proceed as above under "First Esterification." Add
30 ml. of neutral alcohol (94% or higher) and titr!J,te rosin or rosin soap as
desired, using phenolphthalein as indicator. (1 ml. N/2 alkali=0.173 gm.
rosin or 0.188 gm. rosin soda soap.) If the true fatty acid soap is desired, subtract the rosin soap from the total anhydrous soap (real soap) obtained under
total fatty acid determination.
The above method gives somewhat high results on all low percentages of
rosin. On the higher percentages, it is approximately correct. In all cases
where the rosin content is found to be less than 5%, the actual presence or
absence of rosin should be checked qualitativefy by the Lieberman-Storch test,
which is as follows:
One to 2 ml. of the sample of fatty acids are shaken in a test tube with 5 to
10 m!. acetic anhydride, reagent grade, with gentle warming on a steam bath.
After cooling, pour 1 to 2 m!. in a white porcelain dish and allow a drop or two
of sulfuric acid (sp.gr. 1.53) to run down the side of the vessel. If rosin is
present, a fugitive violet coloration changing to a brownish tinge is immediately
produced at the margin of contact of the reagents. The test should be checked
with a sample of fatty acids to which a small amount of rosin has been added.
(Sulfuric acid of 1.53 sp.gr. is prepared by diluting 34.7 m!. of concentrated
suifuric acid (sp.gr. 1.84), reag~nt grade, with 35.7 m!. of distilled water.)

WATER ANALYSIS
Probably at no other time has the importance of a ;{rater supply, either for
domestic or industrial purposes, been so great, as in these early years of the
twentieth century. The increasing realization of the effect of contaminating
materials, both organic and inorganic, on a municipal or private drinking supply'
-and history's record of the devastating nature of epidemics due to waterborne disease organisms, have led to the careful investigation of water for its
sanitary value and the development of materials and equipment to fight and
eliminate such contamination, and have increased many hundreds of per cent
the factor of safety to the public health. In like manner, and even to a greater
extent, has the value of water for industrial uses been a matter for careful
consideration. It is hard for the public to realize the immense quantities of
water used for industrial purposes, not only for the development of steam and
electric power, but also for purposes of manufacture.
From the standpoint of power developmert we are familiar with the heat
losses and the increased operating expenses due to scale formation in the steam
or locomotive boiler, and also to the continued rapid decrease in value of boiler
prqperty, or of power plant property, due to corrosion or rusting. Another
element of trouble which is noticed not so much in stationary boiler practice as
in locomotive boiler practice, is the element of foaming and priming of a water,
which results in much more rapidly putting the steam raiser out of active service
than either of the other ty'pes of trouble.
From the standpoint of plant deterioration due to rusting and corrosion cine
has only to look to the great mass of works on the corrosion of iron and steel
which are largely results of the growing need for some information as to the
cause and possible prevention of this particular phase pf trouble.
In the world of industry the action of a hard waten/upon soa'p consumption
has been known for centuries, and for a considerable period of time the value of
a water was determined largely by the amount of soap that it would consume
and render insoluble. This same hardness has a noticeable effect in the textile
industry, in bleaching and dyeing, in the canning industry, especially when the
water supply contains such, substances in large quantities. ,;_{n. the photographic
industry the presence of chlorides in water and certain alkalieS-isa source of
considerable trouble, and in every case, before any intelligent effort can be made
to overcome these troubles, a complete analysis of the water is necessary.
It shall be our purpose in the methods which-follow to give, where it is possible, first a system of analysis whereby a complete analysis can be made, and t6
follow this up with optional methods which, individually, are equally as good
as those occurring in the system of analysis, and in some cases more satisfactory
where the laboratory has the required equipment, adding any special methq~s
whiLh may be found available.

Chapter by D. K. French. /
2046

WATER ANALYSIS

2047

SANITARY ANALYSIS
A sanitary analysis consists in the physical examination covering turbidity,
color, odor and occasionally taste, the chemical analysis for total residue, loss
on ignition and fixed solids, noting, where possible, the odor during ignition
and also noting the appearance as regards color of the residue both before and
after ignition, the determination of free and albuminoid ammonia, nitrogen
as nitrite and nitrate, chlorine as chloride and oxygen consumed. Organic
nitrogen is frequently determined upon polluted waters.
In sanitary analysis the principal determinations relate to the various forms
and compounds in which nitrogen appears.
Organic Nitrogen.-The initial form can be determined as such, or as is
usually the case in all but highly polluted supplies, as albuminoid ammonia
which gives a very close approximate. By decomposition the organic matter
first gives nitrogen as free ammonia, then, by oxidation, nitrogen as nitrites,
and. finally the more stable form of nitrogen as nitrates is reached. Conversely
the reactions are frequently reversed through the influence of bacteria and
microscopic organisms.
Chlorine is determined and by its excess over the normal chlorine of a.general
district may indicate previous sewage contamination.
Oxygen Consumed, or " oxygen required," means the amount which carbonaceous organic compounds present consume in the presence of potassium
permanganate and acid. From these figures additional evidence is obtained as
to the sanitary character of a w!1ter, though many phases of interference can
occur.

PHYSICAL TESTS
For the physical examination, standards for turbidity and color have been
adopted.
Turbidity.-Turbidity standards are based on parts per ~illion of silica
(Si0 2) suspended in water, and the adopted standard is that of the United States
Geological Survey (A. P. H. A., p. 7). A water with a turbidity of 100 is one
which has 100 p.p.m. of silica (Si0 2) in such a state of fineness that a bright
platinum wire 1 mm. in diameter can just be seen when center of said wire is
100 mm. below the surface of the water and the observer is 1.2 meters above
the wire. The observation must be in open air, not in sunlight, and in the
middle of the day. Standards are preyared with precipitated fuller's earth (to
pass 200':mesh sieve). One gram to one liter of distilled water makes a stock
solution with 1000 turbidity. Standards for comparison are obtained by
dilution:
The Illinois Water Supply Association outlines another method (Proc.
1. W. S. A., 1914, pp. 49-51), whereby a suspension is prepared by shaking silica

WATER ANALYSIS

2048

(Si02) or fuller's earth (ground to pass a 200-mesh sieve), settling for ten hours,
and determining by evaporating and weighing the amqunt of silica (Si0 2) in a
given portion. Standards are then prepared by dilution.
Color.-All suspended matter should be removed b~ filtration. The standard designated as color 500 is obtained as follows:
I
1.246 grams potassium platinic chloride (PtCI 42KCI) \1 containing 0.5 gram of
platinum and 1 gram of crystallized cobalt chloride (CoCI 26H 20) containing
0.25 gram cobalt (Co), are dissolved in water with 100 m!. hydrochloric acid and
made to one liter with distilled water. This solution is diluted with distilleq
water for comparative purposes, but a water with a color greater than 70 should'
be diluted prior to comparison. The standards for observation should be in
100-m!. Nessler tubes with the mark 20-25 cm. above the bottom and should
be viewed vertically downwards to a white reflective surface.
Standard glass disks are used by the United States Geological Survey 2 in
place of the above standard.
Odor.-Observations should be made both on cold and hot samples. Note
should be made immediately on opening containers as some odors are very
transient and rapidly disappear.
Cold.-Shake sample violently in collecting bottle, same to be about half
full. Remove glass stopper and smell at neck of bottle.
Hot.-Use either open beaker, 400 mI.; containing 150 mI. sample well
covered and heated nearly to boiling, or sealed glass stoppered bottle or saponification flask, heating fifteen minutes just under boiling. Allow to cool slightly,
remove stopper, shake and smell. Designate odor as aromatic, grassy, earthy,
musty, fishy, putrid, disagreeable, peaty, sweetish, etc. The following table
expressing intensity of odor is copied from the American Public Health Association Standard Methods, 1913, p. 12:
Numerical
Value.

Term.

Approximate Defiyition.

None.

No odor perceptible.

Very Faint.

An odor that would not be detected ordinarily by the average

consumer, but that could 'be detected in the laboratory by
an experienced observer.

Faint.

An,odor that the consumer might detec~if his.. attention were

called to it, but that would not attract attention otherwise.

Distinct.

An odor that would be detected readily and that might cause

the water to be regarded ~ith disfavor.

Decided.

An odor that would force itself upon the attention, that might
make the water unpalatable. i

Very Strong. An odor of such intensity that the water would be absolutely
unfit to drink. A term to be used only in extreme cases.

1 Care should be taken that this be the bright yello"l\;. platinic/salt,. and not contaminated with the reddish platinous salt.
_, 2 App. made by Builders' Iron Foundry, Providence, R; 1.

WA'fER ANALYSIS

2049

Taste.-May be made on hot and cold samples. A simple statement following largely the terms applied to odor in expressing results, brackish, astringent, salty, sweetish, etc.

CHEMICAL TESTS
Free Ammonia
Apparatus.-The apparatus for this determination should be as far as possible free from joints or connections that are subject in any way to outside
contamination or to decomposition. The apparatus is composed of a distillatio'n flask 'and a condenser, with possibly a safety tube located somewhere near
the flask to avoid the possible carrying over of impurities in cc;>nnection with the
steam.
Reagents: 1. Ammonia-free Water.
2. Standard Ammonium Chloride Solution.-Dissolve 3.82 grams of ammonium cnloride in 1 liter of distilled water. Dilute 10 ml. of this to 1 liter with
ammonia-free water. 1 ml. =0.00001 gram of nitrogen.
3. Nessler's Solution.-Dissolve 50 grams of potassium iodide in the smallest
possible quantity of cold water .. Add a saturated solution of mercuric chloride
until a faint show of excess is indicated. Add 400 ml. of 50% solution of potassium hydroxide. After same has clarified by sedimentation, make up to 1 liter
with water, allow to settle and decant.
Optional Method.-Dissolve 61.75 grams of potassium iodide in 250 m!. of
redistilled water, and add a cold solution of mercuric chloride which has been
saturated by boiling with excess of salt. Pour in the mercury solution cautiously, and add an amount just sufficient to make the color a permanent bright
red. With a little practice the exact depth of color can be easily duplicated.
It will take a little over 400 m!. of the mercuric chloride solution to reach this
end-point. Dissqlve the red precipitate by adding exactly .75 gram of potassium iodide. Then add 150 grams of potassium hydroxide dissolved in 250 m!.
of water. Make up to 1 liter. Mix thoroughly and allow the precipitate
formed to settle. Pour off the supernatant liquid. Mercuric chloride increases
the sensitiveness and potassium iodide decreases it.
Operation.-Clean apparatus thoroughly as follows:
Fill a flask, which for most satisfactory results should be an 800-ml. Kjeldahl
flask, with 500 ml. of distilled water. Add a pinch of c.p. sodium carbonate and
distill first of all with no running water in the condenser jacket until free steam
blows through the apparatus. Then turn on condenser water and distill off
approximately 250 mI., testing the last 50 ml. with Nessler's solution, and this
portion should not show color in fifteen minutes' time. The flask is then emptied of the remaining water, 500 ml. of the water to be analyzed placed therein, .
and if acid, neutralized with c.p. sodium carbonate. A slight excess hurries the

..tgrlculrn'ttl f,n"" ..... ' Llbr.ary,

205oB apatla.

WATER ANALYSIS

ammonia liben1tion but also tends to cause bumping. 'The distillation is then
started, distilling 6 ml. to 10 ml. per minute, and: three (separate portions of 50
ml. each are caught in Nessler jars. After 150 ml. is distilled the flame should
be. removed. !o each 50-ml._portion add 2 ml. Nessler'r solution and. after ten
mmutes' standmg compare with standards from the staudard ammOlllum chloride solution.

FIG. 288.-Free Ammonia ApparatuS.

Albuminoid Ammonia

Reagent: 1. Alkaline Potassium Permanganate.=-Dissoive 200 grams of

potassium hydroxide and 8 grams c.p. potassium llermllnganate in 1250 ml. of
water, boil down to 1 liter and bottle while still warm.

"

WATER ANALYSIS

2051

Operation.-Add 50 m!. alkaline potassium permanganate solution and

several pieces of washed and ignited pumice to the water remaining in the flask
frop! the free ammonia determination and continue the distillation, taking off
four or five separate portions of 50 ml. each in Nessler jars. Add 2 ml. Nessler's solution to each and after ten minutes' standing compare color with standard as in the free ammonia determination.
Organic Nitrogen
While this determination is not usually made we give it for the sake of
completeness.
The portion of sample from the free ammonia determination, or a new por::
tion freed from free ammonia by distillation, is acidified with 5 ml. c.p. sulfuric
acid (nitrogen-free) and digested in a hood until colorless and H 2S0 4 fumes are
given off. A little ignited pumice will guard against bumping. a Cool, dilute
with ammonia-free water, neutralize with 10% Na 2COa solution (NHa free),
distill into Nessler tubes and compare as for free and albuminoid ammonia .
.permanent standards 4 can be made by using potassium platinic chloride, .
2 grams dissolved in water, 100 ml. conc. hydrochloric acid and made to 1 liter;
and cobalt solution, 12 grams cob alto us chloride (CoCh' 6H 20) dissolved in
distilled water, 100 ml. conc. hydrochloric acid added and made to 1 liter. The
following table represents the .amounts used, to be made to 50 ml. with distilled
watedn Nessler tubes for comparison, the 150-m!. mark being 20-25 cm. above
thEl bottom, but should be checked against Nesslerized standards and t~e Nessler
solution modified, if necessary, until the standards agree. This is accomplished
by varying the amounts of potassium iodide and mercuric chloride.

Equivalent Volume
of Standard
Ammonium Chloride,
m!'

Platinum Solution,
m!.

0.0
0.1
0.2
004
0.7

1.2
1.8
. 2.8
4.7
5.9

0.0
0.0
0.0
0.1
0.2

1.0
1.4
1.7
2.0
2.5

7.7
9.9
11.4
12.7
15.0

0.5
1.1
1.7
2.2
3.3

3.0
3.5
4.0
4.5
5.0

17.3
19.0
19.7
19.9
20.0

4.5
5.7
7.1
8.7
lOA

6.0
7.0 .

20.0
20.0

15.0
22.0

Cobalt Solution,
mI.

3 KMn04 is sometimes used at this point, but as it may cause loss of N, the practice
is not advisable.
4 Permanent Standards (Jackson, 1'ech. Quart., 13, 320 (1900.

WATER ANALYSIS

2052

Nitrogen as Nitrite
Reagent: 1. Sulfanilic Acid.-Dissolve 8 grams oftHe acid in 1 liter of acetic
acid, specific gravity 1.04. This is practically a saturated solution and keeps
I
well.
2. Naphthylamine Acetate.-Dissolve 5 grams of a-naphthylamine in 1 liter
of acetic acid, specific gravity 1.04, and filter through 'absorbent cotton (previously washed).
NOTE.-A slightly pink color resulting on standing does not materially interfere
with the use of this solution.

3. Sodium Nitrile Solution.-Dissolve 1.1 grams of silver nitrite in nitritefree water. Precipitate the silver with sodium or potassium chloride solution
and dilute to 1 liter. Dilute 100 ml. of this solution to 1 liter and then 10 ml. of
this second solution to 1 liter with sterilized nitrite-free water, adding 1 ml. of
chloroform and holding in a sterilized bottle. 1 ml. = 0.0001 milligram' of
nitrogen.
Operation.-Take 100 ml. of the water after filtration and clarificati~n,
preferably with aluminum hydroxide, to remoye possible suspended iron and
material which might interfere with color production. Add 2 ml. each of
solutions No.1 and No.2. After ten minutes' standing compare with standards
made up from the standard sodium nitrite solution (No.3).
PERMANENT STANDARDS

Cobalt Solution.-Dissolve 24 grams of cobaltous chloride (CoCI 2 6H 20) in

distilled water, add 100 m!. of conc. hydrochloric acid and make up to one liter
with distilled water.
Copper Solution.-Dissolve 12 grams of dry cupric chloride (CuCh2H 20)
in distilled water, add 100 mI. of conc. hydrochloric acid and make up to one
liter with distilled water.
The standards are used in 100-m1. tubes with the mfj,rk 12-14 cm4tbove the
bottom. The following table gives the proportions of each solution to be made
up to 100 in1.:
mL
Cobalt Solution,

.0
1.1
3.5
6.0
12.5

p,p,m, Nitrite per

100 mL of Water,

mI.
Copper Solution,

.0

.OOO~

1.1

3.0
5.0
8.0

.001
.003
.005
.010

. ",

The solutions to use for 100 m1. of water are the old ones, as fol~ows: 1 m!.
of hydrochloric acid (1 : 4), 2 m1. of sulfanilic acid (8 grams per liter), and
finally 2 ml. of naphthylamine hydrochloride (8 grams per liter with 10 ml. of
conc. hydrochloric acid), and allow color to develop twenty minutes.
NOTE.-Volume 28, page 742, J. Soc. Chern. Ind., calls attention to the possibility
of a permanent standard composed of a solution of acid..magenta (fuchsine-S, acid
fuchsine according to Weigert). According to this article, 0;2 of a gram of this dye

WATER ANALYSIS

2053

is dissolved in 50 ml. of 2/N HCl and made up to 2000 ml. with distilled water. Of
this solution 200 ml. are mixed with 50 ml. of 2/N HCl and again diluted to 2000 mI.
with distilled water. From this latter solution standard solutions can be prepared
containing various quantities, these standards being made up to 200 ml. with distilled
water after the addition of 5 ml. of 2/N HC}' Considerable work is being done on this,
but the standards have not yet been accepted in this country. However, standards
can be made by matching these solutions against standards prepared in the usual way
and their permanence is much greater than such standards.
N!trogen as Nitrate
ALUMINUM REDUCTION 5

Reagents: 1. Sodium or Potassium Hydroxide Solution.-Dissolve 250

grams of the hydrate in 1250 mI. of distilled water, add several strips of aluminum foil and allow action to pass overnight. Boil down to I liter.
2. Aluminum Foil.-Use strips of pure aluminum approximately 10 em. long, .
9 mm. wide, and ~ mm. thick, same to weigh about Y2 gram.
Operation.-IOO mI. of water is placed in a 300-ml. casserole. Add 2 mI. of
the hydroxide solution and boil down to about 20 ml. Pour the contents of the
casserole into a test-tube about 3 cm. in diameter and of approximately 100-ml.
capacity. Rinse the casserole several times with nitrogen-free water and add
the rinse water to that already in the tube, thus making the c()ntents of the tube
approximately 75 ml. Add a strip of aluminum foil. Close the tube by means
of a rubber stopper through which passes a A-shaped glass tube about 5 mm.
in diameter. Make the short end of the tube flush with the lower side of the
rubber stopper while the other end extends below the surface of distilled water
contained in another test-tube. 'This apparatus serves as a trap through which
the evolved hydrogen escapes freely. The amount of ammonia escaping into
the trap is slight and may be neglected. Allow the action to proceed for a
minimum period of four hours, or overnight. Pour contents of the tube into
a distilling flask, dilute with 250 ml. of ammonia-free water, distill and collect
in Nessler tubes and Nesslerize. When the nitrate content is high, collect
the qistillate in a 200-ml. flask and Nesslerize an aliquot portion. If the
supernatant liquid in the reduction tube is clear and colorless, the solution may
be diluted to a definite volume and an aliquot part Nesslerized without distillation.
NOTE.-Where the nitrates are very high, from 50 parts per million up, take a
smaller sample and dilute to 100 ml. before reduction.
.
Oxygen Consumed
Reagents: 1. Standard Potassium Permanganaie Solution.-Dissolve 0.4
gram c.p. salt in I liter of distilled water. I ml. is equivalent to 0.1 milligram
available oxygen.
2. Standard Ammonium, Oxalate Solution.-Dissolve 0.888 gram c.p.
ammonium oxalate in lliter of distilled water. 1 ml. is equivalent to 0.1 milligram of oxygen. The standard permanganate solution ,must be standardized
against the ammonium oxalate solution.
5 Univ. of illinois Bull. Water Survey, Series 7, p. 14 (1909); Amer. Jour. Pub.
Hygiene, 19, 536 (1909).
.

2054

WATER ANALYSIS

3. Dilute Sulfuric Acid 1:3.-Free from oxidizable matter by adding

KMn04 until a faint pink color persists after several ~ours.
Operation.-100 ml. of water are measured into a 4QO-ml. Erlenmeyer flask,
acidified with 5 mI. dilute sulfuric acid. Ten roI. of standard permanganate
solution is run in from a burette and the flask is placed irl a bath of boiling water,
the level of which is above the level of the flask contents, for thirty minutes.
Remove. Add 10 mI. of standard oxalic solution and then determine the excess
with the standard permanganate solution. Deduct from the total permanganate solution used the 10 mI. of oxalic acitl, and the remainder represents
oxygen consumed. For particularly bad waters smaller quantities of the
sample are taken and diluted to 100 mI., as it is undesirable at any time in the
course of boiling that the pink color of permanganate be completely discharged.
Chlorine as Chlorides
:Reagents: Standard Salt Solution.-16.48 grams fused c.p. sodium chloride
are dissolved in 1 liter of distilled water. 100 Illl. of this solution diluted' to
1 liter gives a standard solution, each mI. of which contains .001 gram of chlorine:
Standard Silver Nitrate Solution.-4.8 grams dried silver nitrate crystals
are dissolved in one liter of distilled water. Each ml. of this solution is equivalent to approximately .001 gram of chlorine, standardized against the Standard
Salt Solution.
,/

NOTE.-N/50solutions of both sodium chloride and silver nitrate can be used

where it is inconvenient to make too many standard solutions, using the proper factors.

Potassium Chromate.-lO% solution neutral potassium chromate.

NOTE.-A. P. H. A., page 35, recommends 5% solution of neutral potassium
chromate, adding after solution of the crystals in a few ml. of water, sufficient silver
nitrate to produce a slight red precipitation. This is filtered off, and the filtrate made
up to volume.
Operation.-100 mI. of the sample are titrate~ witi silver nitrate solution,
using 1 ml. of the potassium chromate as indicator to the first persistence of
the silver chromate red. Subtract 0.2 ml. blank from the reading. A white
porcelain dish or casserole is the preferable container, although a flint-glass
beaker over a white porcelain plate may be used. Where a chlorine is high
and more than 15 ml. of :;lilver is used, a smaller sarople~(50 ml. or 25 ml.)
should be taken and distilled water added to bring the volume up to approximately 100 mI. If the original water is noticeably colored, 25 to 30-by standard,
it may be decolorized by adding precipitated aluminum hydroxide, bringing to
a boil and filtering. Titration must always be made in the cold, however.
NOTE.-Precipitated aluminum hydroxide is prepared ]:>y dissolving potash alum in
water, precipitating by adding carefully ammonia and washing in a large jar with
distilled water, by decantation, until free from chlorine, ammonia, and nitrites. An
acid water should first be neutralized with sodium carbonate and a water containing
free hydroxides should be neutralized with sulfuric acid. Where specially accurate
work is desired, observations may be made in a dark room with a yellow light.
(A. P. H. A., page 36.) A yellow photographic glass may be used in daylight and at
night the ordipary carbon filament electric light.
.

.-r:-

WATER ANALYSIS

2055

Total Solid Residue

Evaporate 50 mI.' to dryness, in a platinum dish, at about 270 0 Fahr., and
bake for at least 30 minutes at that temperature. An ordinary water-bath
temperature will not remove water of crystallization from alkali sulfates or
calcium sulfate. Where water IS high in magnesium salts, as determined in
mineral analysis, the water-bath temperature is more satisfactory, due to the
readiness with which magnesium chloride and frequently magnesium carbonate
will decompose to oxide. As a rule, however, a temperature ffom 240 0 to 270 0
meets most Qf the conditions.
Weight in grams times 1168=grs. per gal. total solids.
Weight in grams times 20,000 = parts per million total solids.
Residues from acid waters should be ignited to a dull red heat after addition
of a drop or so of sulfuric acid, to insure complete removal of the acid itself,
which will not go off at the temperature stated. This will result in the decomppsition of all iron compounds to the oxide form, and will fix all salts, lime,
magnesium, sodium and potassium, in the sulfate form, and correction should,
be made for chlorides present, which would be converted into sulfates.
Waters high in magnesiu~ salts should be evaporated at the first specified
temperature, adding, however, a few m!. of Nj50 normal sodium carbonate
solution to insure a slight excess of sodium carbonate, correcting for the weight
of sodium carbonate added. Where the waters contain much organic matter
after weighing, they may be very gently ignited at a very dull red heat until the
carbon has been burned off. After cooling, the residue may be recarbonated
with tested ammonium carbonate solution, and again dried in the usual way.
The difference in weight after titrating for possible loss of chlorides, due to
volatilization, gives a close approximation of the organic matter present.
Similarly, waters high in magnesium chloride or nitrate compounds may be
evaporated with a few drops excess of sulfuric acid, and ignited to a dull red
heat, the residue being compared, where a complete analysis is made, with the
sum of all bases calculated to the sulfate form. This is sometimes more convenient and satisfactory than the evaporation with excess sodium carbonate.

INTERPRETATION OF RESULTS
The interpretation of the results of a sanitary water analysis is largely a
matter of experience, and it is impossible to lay down hard and fast rules covering this one matter. It is, however, possibJe to sum up the meanings of the
various determinations made, as each determination has some bearing upon the
sanitary condition.
In physical data the turbidity refers to insoluble matter in suspension. In
many cases, it is perfectly harmless, although less attractive, and frequently

2056

WATER ANALYSIS

suggests contamination, which is as apt to be present as not. High turbidities,

following rain storms or lake over-turnings, are usually ~ccompanied by B. coli,
the intestinal organism; in considerable quantity. The turbid wa~ers of the
West may cause stomach trouble until a person is accustomed to them. Color
is due, usually, to an extract of vegetable or organic fuatter, or to iron salts,
and in itself has no value save suggesting organic cqntamination. Highly
colored water may have an astringent taste, and is not looked upon with favor
by the consumer. It may cause corrosion in pipes and boilers.
Various organic matter!> are in no way determined in this analysis, the results
obtained being simply indications of certain cycles in decomposition of nitrogenous material, as no decomposition cim take place without some resulting
nitrogen compound. Free ammonia represents the first stage in this decomposition, and represents the amount of organic matter present in a partially
decomposed and decomposing state. Deep wells in glacial drift frequently 'also
contain high ammonia, however, which would in no way suggest active contamination.
'
Albuminoid ammonia represents organic substances in, an undecomposed
state, which will, however, decompose under the proper conditions. The
presence of nitrogen in such combination in large amount usually suggests the
presence of pollution of a sewage character. However, its presence usually
accompanies and varies in amount with the color and with the microscopic
organisms.
The next stage in the cycle is nitrogen" as nitrites, indicating that decomposition is actively progressing. Nitrite in surface water may indicate contamination when in considerable quantity, but'in ground water is absolutely of
no significance. (Proc. Am. W. W. Assoc., 1908, page 323.) Its presence is
due to the action of certain types of bacteria either as a product of oxidation
from free ammonia or as a product of reduction from nitrate. Ferrous compounds have also a bearing on such reduction.
The final state of decomposition is nitrogen as nitrate. This indicates the
fact that at some time in the 'past organic matter has been;present. Its presence
indicates a purified water. In large amounts it may cause itching in sensitive
persons. It is an important cause of corrosion in 'pipes and boilers.
The oxygen consumed represents the amount of oxygen required to oxidize
organic matter already in the water. It has a bearing upon the organic matter,
but there are many inorganic substances which also discharge the color of the
permanganate solution, and the result should always be consi1lered in the presence of the other determinations.
Chlorine. as chlorides, if above the normal figure for any definite location,
is a fairly good indication of sewage, as it is one of the most constant and principal constituents of sewage.
'
T~e total residue itself should not be too. high, ~~ an. excess. of inorganic
materIals would staII).p the water unfit from an mdustp~l pomt of VIew, 'and also,
from the standpoint of the individual, might make it unsatisfactory as a drinking supply for daily consumption.
With reference to standards of purity, it is impossible to make absolute
standards. We,quote as a matter of interest a table published'by the State of
Illinois, giving their suggested limits of impurities. ior-supplies in that State.
(The remarks which follow are those of the State Geological Survey.)

WATER ANALYSIS
Lake
Michigan.

None
Turbidity .................
None
Color ....................
None
Odor ....................
Residue on Evaporation ... 150.
4.5
Chlorine .................
1.6
Oxygen consumed .........
Nitrogen as:
.01
Free Ammonia .........
.08
Albuminoid Ammonia ...
.000
Nitrites ................
.04
Nitrates ...............
Alkalinity ................ ( 120.
Bacteria per mI........... 100.
Colon bacillus in one mI .... Absent

Streams.

Springs
lind
Shallow
Wells.

2057
Deep Drift
Well.

Deep Rock
Wells.

.2*
None
300.
6.
5.

None
None
None
500.
15.
2.

None
None
None
500.
15.
2.-5

None
None
None
500.
5.-100
2.-5

.05
.15
.000
.5
200.
500.
Absent

.02
.05
.000
2.00
300.
500.
Absent

.02-3
.20
.005
.5
300.
100.
Absent

.02-3
.15
.000
.5
300.
100.
Absent

10.

: Modified Nessler or Natural Water Standard equal 26 p.p.m. platinum scale.

The formation of a reasonable and just opinion regarding the wholesomeness

of a water requires that there be taken into consideration all the data of the
analysis, together with the history of the water;' the, nature of the source;
character of the soil and earth or rock strata, and the surroundings. The
interpretation of results is a task for the expert. If possible a bacteriological
examination should be made to supplement the sanitary analysis, as the latter
is more useful in determining the source of pollution and indicating the stage of
contamination.
.
Chlorine is the most permanent element shown in water analysis, as it is
never removed from water by any changes or processes of pUrification. Salt
deposits, however, in the soil must also be taken into consideration. t

MINERAL ANALYSIS
Outline of Procedure
. Evaporate 50 ml. of the sample of water to dryness and bake the residue at 270 0
F. in weighed platinum dish. Increased weight Of dish represents
total solid residue. (Can be used for 804 when sample is small.)
Ignite for organic loss.
250 m!. Titrate with N/IO acid or alkali for alkalinity or acidity. (Can be
re-used to make up volume of 500-m!. portion when water sample is
small.) Methyl orange indicator. (With a 50 ml. sample use N/50
acid.)
100,m!. Titrate with NjlO AgN03 for chlorine.

2058

WATER ANALYSIS

100 ml. Acidify, boil, precipitate with BaCh, filter and weigh for total sulfate.
(Use filtrate for Na and K when necessary.) I
100 ml. Add 2 ml. 10% Na 2C0 3, evaporate to dryness, add phenolsulfonic acid,
dilute, then excess of NH 4 0H for total nitrate,
500 ml. Evaporate to dryness (with a few ml. ,concentrated HCI when very
accurate Si0 2 figure is necessary) in No.8. R. B. dish. Bake 30
minutes, cool, add boiling Hel (concentrated); dilute and filter.
Precipitate is Si0 2 Filtrate. Add a few drops of HN0 3 , concentrate to 50 ml., cool;
and silicate impuadd NH.OH, boil and filter.
rities (also BaSO.). 1 - - - - - - - - , - - - - - - - - - - - - - - - Unless great accuracy is necessary, Precipitate (Fe, AI, Filtrate. Boil and add saturated Am. OxPOl') may be realate drop by drop, boil and filter.
it should be
ported as such or
weighed as such,
as Fe and Al after 1---------------otherwise Si0 2 can
be removed by HFa Qual. test for phos- Prec. Ca as oxalate. Filtrate Mg (and
phate has shown
Dissolvein2%H 2Mn),addlO-25ml.
and correction
same to be absent.
SO.. Titrate with
concentrated Sod.
made.
Otherwise both Fe
standard KMnO..
Phos. Solution,
and PO.''' should
then 50 ml. NH.be determined a n d ,
OH, stir well 2
weight corrected.
minutes, or more,
let stand 4 hours,
or more, filter and
wash with. 3%
NH.OH. Dry in
air oven at 50":'
60 C. Add excess
N/IO H 2SO. to dissolve ppt. Titrate
the excess acid with
N/IO NaOH.
NOTE.-For industrial purposes the original addition of HOI is not always necessary
and correction for BaSO., Phos. Mn and separation of Fe andjAI can be dispensed with
unless there is cause to suspect one to be present in material amounts.

In the matter of mineral analysis of water, it is not so hard to obtain a complete analysis of the water, including the non-incrusting or " nearly always"
soluble materials as well as the incrusting materials, as it is tQ, make nvmberless
individual or independent tests, in the hope of drawing conclusions from same.
The scheme of analysis which follows is used exclusively in the writer's laboratories, and when carried out as given, makes it possible to complete analysis of a
water, or a group of waters numbering up to ten,. in the period of eight hours
elapsed time, or twenty-f9ur hours, assuming the work is arranged in such a way
that the magnesia precipitates are allowed to stand oyernight before filtration.
On another page will be found a skeleton form for tHis complete analysis; and
this skeleton will serve as a rough guide to the more extended discussion which
will follow.
The complete analysis considers the quantitative determinations of silica,
iron and aluminum, calcium, magnesium, sodium and::potassium, as bases, and
carbonate, hydroxide, nitrate, sulfate, chloride, ahd phosphate, as radicals or

WATER ANALYSIS

2059

acids, with suggested methods for manganese, ammonia, barium, and other
materials which might possibly be present.
Prior to the starting of the analysis, the physical characteristics of the water
should be noted, turbid waters should be filtered, the suspended matter analyzed
separately when necessary, and the amount determined either by filtration and
weighing of the separated material (alundum cones are very satisfactory), or
by the difference between two residues, one of which represents the original
water and one the filtered water. The mineral analysis should represent the
filtered supply. This is due to the difficulty of getting uniform samples with
suspended matter at different times. Treatment with HzO z is also sometimes
effective when the color is high.
NOTE.-Due to the tendency of some grades of H 20 2 to ~hange both alkalinity and
chloride figures, the material used should be examined carefully. Merck's Super Oxol
(30% H 20 2) has been found thoroughly satisfactory and has not affected analytical
results.
Silica, Iron, Aluminum, Calcium, Magnesium

(Manganese, Phosphoric Acid) .

NOTE.-If from qualitative observations the water contains considerable mineral
matter, smaller quantities varying from 100 to 250 mI. may be taken, or if the sample
is apparently distilled or condensed and contains very little mineral matter, 1000 m!.
should be taken, the object being to obtain a residue neither too large nor too small.
0.4 to 0.6 gram is a good quantity to work on.
Silica
Add 5-10 ml. HCI, evaporate over free flame, then on ~~in. asbestos board,
to dryness, 500 ml. original water, using a No.8 'porcelain dish. Bake at
110-130 C. or on asbestos plate over flame for one-half hour. Moisten with
10 ml. concentrated HCl, add 50 m!. of water, boil fifteen to thirty seconds and
filter. Wash with hot water.
NOTE.-For great accuracy, evaporate twice to dryness as above, with the addition,
prior to the sample going to dryness, of 10 mI. HCl, allow to bake as above, following from
there on the usual procedure for filtration.
NOTE.-When the water is highly colored due to. o.rganic matter, a few ml. o.f HNO a
can also. be added.
e
The precipitate retained on the filter paper represents the silica or siliceous
matter, including possibly barium sulfate. Ignite and weigh.
NOTE.-If the amo.unt-is over .01 g. per liter, or 10 parts per million, moisten with a
few drops of concentrated sulfuric acid and hydrofluoric acid, expel excess acids, and
reweigh. This must be done in platinum. The loss represents silica, and should be
recorded as such, and the residue represents bases, principally barium, combineu with
sulfuric acid. This will also catch possible calcium sulfate that might be left undissolved, due to short bo.iling, to. low ililution, or conditions which would prevent its no.rmal
solubility in the original solutio.n.
'
NOTE.-When a large amount of CaS04 is suspected or indicated, extra care must be
taken to remove it completely by solutio.n.

Iron and Aluminum (Gravimetric)

T.he filtrate contains iron, aluminum, calcium, magnesium, possibly man~
ganese, and phosphate. Bring to a boil, add two or three drops cone. nitric acid

2060

WATER ANALYSIS

and concentrate to about 25-50 ml. Remove from ho~ plate or flame, add ammonium hydroxide in slight excess, boil for one or two plinutes, and filter.
The precipitate contains iron, aluminum, and possibly phosphates. Burn
and weigh as oxides of iron and aluminum, plus phosphates, and test 50 ml. of
the original water with treatment in the usual way to determine whether or not
phosphates are present. Where this precipitate of iron Il-nd aluminum oxides is
greater than 0.01 g. per liter or 10 parts per million, or where the separation
of the iron and aluminum is advisable, the precipitate should be fused with
eight or ten times its weight of potassium bisulfate, redissolved in water, the'
iron reduced to the ferrous condition with zinc, and titrated with potassium
permanganate, recording the difference in weight between the original precipitate and the iron determination as aluminum oxide.

NOTE.~Where much water is available and time is an object, an additional 500 ml.
can be carried down to approximately 50 ml. with a few drops of nitric acid, the iron
and aluminum precipitated as above mentioned with ammonia, and the precipitate
before drying redissolved in acid, reduced and titrated with potassium permanganate.
This portion can be started at the same time the original analysis is started, and win.
greatly simplify the determination and save time.
.

Total Iron (Colorimetric)

Reagents: Iron Standard.-0.7 g. cryst. ferrous ammonium sulfate is
dissolved in a small amount of distilled water, add 25 ml. dilute (1: 5) sulfuric
acid, warm slightly and oxidize completely with potassium permanganate,
make up to 1000 ml. 1 ml. = 0.1 mg. Fe.
Potassium Sulfocyanide.-2% solution.
'Potassium Permangana"te.-6.3 g. per liter.
Operation.-Instead of precipitating, or where traces of Fe are of importance, 100 ml. to 1000 ml. of the water may be carried to dryness with Hel and a
few drops of Br, taken up with 5 ml. (1 : 1) HeI, diluted to 100 mI. in a Nessler
tube, 10 ml. KeNS solution (20 g. to a liter) added and the color compared
with standards. The comparison should be made at once as the color fades.
NOTE.-It is frequently as satisfactory to add the standard iron solution from a
burette to a 100 ml. Nessler tube containing 5 m!. (1': 1) hydrochloric acid (Fe free),
10 ml. potassium sulfocyanide solution (20 g. to a liter) and sufIicient distilled water
until the color matches that of the sample.
(Ferrous Iron-Colorimetric)

(Frequently desirable in acid waters but rarely necessary.)

Reagents: Iron Standard.-0.7 g. cryst. ferrous ammonium sulfate is dissolved in one liter of distilled water containing 10 ml. dilute H 2S0 4 (Not
permanent. Should be made Up as needed.) 1 ml. =0.1 mg. Fe.
Potassium Ferricyanide Solution.-(Prepare as needed.) 0.5 g. per 100 ml.
distilled water.
Sulfuric Acid.-l : 5.
NOTE.-Prepare sample and standards at same time.

WATER ANALYSIS

2061

Operation.-Place in 100 ml. Nessler jar 50 ml. of sample, 10 ml. dilute

H 2S0 4 (1: 5), filter, if necessary, to remove suspended matter, add 15 ml. potassium ferricyanide solution and make up. to 100 ml. mark with distilled water.
Compare with standards made as follows:
Place in 100 ml. Nessler jar 75 ml. distilled water, 10 ml. dilute H 2S0 4 (1:5)
and 15 ml. potassium ferricyanide solution, and mix well. Add various amounts
of iron standard from burette, mix and compare color. Determine ferric iron
by deduction of ferrous iron from total iron.
Phosphates
Reagents: Ammonium Molybdate.-50 g. c.p. neutral salt dissolved in,
1 liter distilled water.
Nitric Acid (spec. grav. 1.07).-Dilute about 1:5 with distilled water. .
Standard Phosphate Solution.-0.5045 g. c.p. cryst. N a 2HP0 4 12H 20.
Dissolve in distilled water, 100 mI. standard HNOa added. Dilute to 1 liter.
One ml. =0.0001 gram P 20 5
Operation.-Evaporate 50 ml. water to dryness in porcelain after addition of
3 ml. HNO a (spec. grav. 1.07). Bake two hours at 212 0 F. Take up with 50
ml. distilled water, add 4 ml. molybdate solution and 2 ml. HNO a, and compare,
in Nessler tube with standards from phosphate solution made to 50 ml. and
treated with same reagents. A tube 2.5 cm. by 24 cm. to 100-ml. mark of hard,
white glass is most suitable. 6 Where waters are already colored. the evaporation
should be carried on with 3 ml. HNO a and 0.5 mI. (or more, if water is highly
,colored) of KMn04 solution (1 gram per 1000 mI.), baking at 212 0 F. for the
same time. 7 Where the phosphate is present in large enough quantities to
precipitate the gravimetric methods may be used.
Calcium

The filtrate from iron, aluminum and phosphate precipitate contains calcium, magnesium, and possibly manganese. Concentrate to about 100 ml.
Add to the hot ammoniacal solution a saturated solution of ammonium oxalate
drop by drop, or add in small portions crystals of ammonium oxalate. Allow
to boil two minutes, stirring, if necessary (on account of heavy precipitate and
tendency to bump), remove, filter and wash. (Eight complete washings are
usually sufficient.)
NOTE.-Where great accuracy is desired, the precipitate on the filter should be
redissolved in a small amount of hot, dilute, hydrochloric acid and reprecipitated with
ammonium oxalate.
Dissolve in 2% sulfuric acid and titrate with the standard solution
potassium permanganate. (N /50 KMn04 may be used.)

of

NOTE.-Where the volumetric method is to be used, eight complete washings are,

as a rule, sufficient to remove the traces of ammonia salts, which are prone to have
considerable influence upon the volumetric results, due to the possibility of traces of
ammonium oxalate still being present.
6 J. Am. Chem. Soc., 23, 96 (1901).
7 J. Ind. Eng. Chem., 5, 301-2 (1913).

2062

WATER ANALYSIS
Fe Value XO.896 = CaCOa
Fe ValueXO.5021 =CaO.
Fe ValueXO.3589=Ca.
Optional (Gravimetric)

Instead of redissolving and titrating the calcium I oxalate upon the filter
paper, it can be dried, burned and weighed either as calcium oxide or calcium
carbonate.
.
NOTE.-The burning of calcium oxalate to carbonate is not so difficult as it seems, as
an intense heat is necessary to convert it to the oxide, and if the crucible is well watched
and the flame gives just sufficient heat to carbonize and destroy the filter paper, there
will be no chance whatever of any calcium oxide being formed, or any calcium oxalate
being left. Where hypothetical combinations are used, it is very convenient to have the
calcium as carbonate without calculation. Where burned to the complete oxide it is
frequently necessary to use a blast lamp, as large precipitates require a high temPfature
to reduce completely to oxide form.
'

Magnesium
The filtrate contains magnesium (and possibly manganese). Acidify with
HCl, concentrate if necessary to 150 mI., add 25 ml. saturated solution of
ammonium sodium hydrogen phosphate (NH4N aRPO, 4R 20, microcosmic
salt), cool and make alkaline with ammonium hydroxide. Allow to stand at
least four hours.
..
The precipitftte is washed several times by decantation with ammonium
hydroxide solution (1 part NH,OR, sp.gr. 0.90 to 9 parts water), and finally
on the filter. After draining, the filter paper 'is opened out, the moisture
removed as much as possible by means of dry filter paper. The residue may ~e
dried in the room for about forty-five minutes or in the air oven at 50 to 60 C.
for fifteen to twenty minutes. When the filter has dried, ammonia will have
been expelled. Entire paper, etc., is placed in a dry beaker, N/I0 sulfuric acid
added in excess, using the methyl orange indicator, the solution diluted to 100
ml. and the excess of acid titrated with N lID sodiu~/hydro~ide.
One mi. N/lO H 2S0 4 =0.00216 gram MgO.
NOTE.-Accurate results are also obtained with the use of sodium phosphate added
direct to the 'filtrate from the calcium precipitate .'without previously acidifying with
acid, with 25 ml. to 50 m!. of ammonium hydroxide added to make strongly alkaline,
after which the solution should be very thoroughly stirred (for~a_t least two minutes),
using a rubber-ended glass rod. Allow to stand at least four hours. '-.~ _
For very rapid work in either case, if the magnesium solution after the precipitation
is cooled in ice-water, filtration can be frequently made in two hours' time.
For extremely accurate work the precipitate produced in either of the methods above
should be redissolved in a little dilute Hel and the precipitation repeated.
Optional (Gravimetric)

The ppt. can be filtered, and washed with

hydroxide. Burn and weigh as Mg 2P 20 7
Optional (Volumetric)

3%

solution of ammonium
I

The following method is purely volumetric . .../I't is used in a great many

water softening plants for control, is accurate and extremely simple.. It is not

WATER ANALYSIS

2063

. used in the general scheme of analysis because it starts with the raw water and
not a concentrated sample. A Mg determination can be made in half an hour
or less.
Solutions: (1) Lime Water.-A saturated solution is made by shaking an
excess of lime with a suitabl'e volume of water. After settling, the clear liquid
is siphoned into a bottle. Changes in temperature may cause slight precipitation in the saturated solution. To avoid this it is a good plan to add 10-20 ml.
of distilled water per liter of solution.
(2) Standard Acid.-N/50 H 2S0 4
Procedure.-l00 ml. is measured into a 250 ml. volumetric flask, neutralized
with H 2S0 4 to methyl orange end.point. The liquid is boiled down to about'
50 ml. and to the hot solution 50 ml. of lime water added. The flask is filled
to. the mark with CO 2 free water, plus 5 ml. Place on a steam bath for 10
minutes; cool, allow to settle, pipette out 50 ml. and titrate with N /50 H 2S0 4
using phenolphthalein indicator.
The relation between the lime water and the standard acid is determined by
runl1ing a blank. This figure once obtained can be used until the lime solution
is all used. It is necessary to run blanks on aU freshly made lime solutions.
It is good practice to check the lime figure from time to time.
Blank-5XmI. acid used=ml. used to ppt. Mg.
This figureX2.4=p.p.m. of Mg.
Manganese
Where necessary, manganese should be determined separately in another
portion of the water and correctio~s made. The Knorres Persulfate method is'
the most reliable for large amounts (10 milligrams Mn per liter); the Bismuthate
method for smaller amounts.
KNORRES PERSULFATE METHOD (VOLUMETRIC)

Reagents: Potassium Bisulfate c.p.

Ammonium Persulfate Solution (60 grams per liter distilled water).
Hydrogen Peroxide Solution.-Equivalent to N/I0 KMn04 .. (Approx. 5.6
mI., 3% H 20 2 diluted to 100 mI.)
Operation.-Evaporate 5 liters or more to dryness, adding first 10 ml. concentrated H 2S0 4 Ignite after adding a few crystals of potassium bisulfate
and take up in hot water. Transfer to 250-ml. Erlenmeyer flask with 5 ml.
dilute (1 : 3) H 2S0 4, add 10 ml. ammonium persulfate solution, boil twenty
minutes, cool, dissolve precipitate (manganese superoxide) in standard hydrogen peroxide solution. (If no ppt. forms no manganese is present.)
\
NOTEs.-When hydrogen peroxide solution is standardized against N/I0 KMn04
1 ml. will be equivalent to 2.747 milligram Mn.
An excess of 10-20 ml. H 20 2 solution can be added and this excess titrated with
N/I0 KMn04. '
SODIUM BISMUTHATE METHOD (COLORIMETRIC)

Reagents: Potassium Permanganate.-{).288 gram KMn04 to 1000 ml.

1 ml. = 0.1 milligram Mn.
S J. Am. Chern. Soc., 29, 1074-78, 1907.

2064

WATER ANALYStS

Sodium Bismuthate (purest).

,See method of preparation of reagent given on P3:ge 1198.
Nitric Acid.-(Spec. grav. 1.135) 3 parts concentrated HNO a to 1 part H 20
should be blown with air until free from ,oxides of nitrogen.
Sulfuric Acid.-25 ml. concentrated H 2S0 4 to 100Q ml. Add permanganate
solution to a faint but noticeable color.
'
Operation.-Evaporate 500 mI. in porcelain dish 'after adding 1 mI. dilute
H 2S0 4 in excess of that necessary to neutralize all alkali. Ignite to remove free
acid (organic matter and chlorine), cool and dissolve in 50 ml. HNOa (30 :tllt
concentrated HNO a to 1 liter), with heat if necessary. Cool again, add 0.5
gram sodium bismuthate and heat until pink color disappears, re-cool and add
sodium bismuthate in excess, filter through asbestos in Gooch crucible (asbestos
must be free from organic matter, thoroughly washed and ignited), or alundum,
crucible. Wash with nitrite-free distilled water containing 5% dilute HNO a
(30 ml. concentrated HNO a per liter). into Nessler tube, make up to 100- ml.
and match with color produced by necessary amount of standard KMn04 in
100 ml. H 2S0 4 reagent.
No. m!. standar<;l KMn04XO.2=milligrams Mn per liter.
NOTE.-The permanganate solution used for oxygen consumed (see Sanitary Method)
contains 0.139 gram Mn per liter and may be used when necessary.

No. m!. XO.278 = milligram Mn per liter.

/

Sulfates
100 m!. of the water is slightly acidified with conc. HCI and 5 m!. 10% NH 4CI
solution added, brought to a boil, and if turbid is filtered and washed four or ftve
times with boiling water. The clear or original water is now brought to ll: boil
and 10% barium chloride added drop by drop to the boiling solution in slight
excess. Boil ten minutes, stirring from time to time, if the precipitate is heavy.
Remove and allow to cool prior to filtering. The precipitate consists of barium
sulfate. Wash free from chlorides, testing with AgNOa Dry, ignite and
weigh.
BaS04 X.411 = S04.
BaS04 X .583 = CaS04.
milligrams BaS04X.340 = CaS04 grains per gallon.

BENZIDINE METHOD (OPTIONAL) 9~ -

-.

Reagents: Benzidine Solution.-Triturate in a mortar with -5 ml. to 10 m!.

water, 4 grams benzidine base. Transfer to liter flask, add 10 ml. HCI and make
to volume. 1 ml. = .002186 gram S04.
Hydroxylamine Hydrochloride.-1 % solution in water.
Operation.-Add 10 m!. hydroxylamine hydrochloride to 250 ml. water, then
add, at once stirring well, 100 m!. benzidine solutioiJ.. Allow to stand fifteen to
twenty minutes, decant through vacuum filter and wash with 10 m!. to 20 m!.
distilled water (do not let filter run dry), return filter paper to beaker, cover
with distilled water, bring to a boil and titrate with N/lO or i N/50 NaOH, using
phenolphthalein as indicator.
-' _:_9

Friedheim and Nydegger, Z. angew. Chem., 20, 9 (1907).

WATER ANALYSIS

2065

One ml. NjlO NaOH=.0048 gram 80 4 ,

One ml. Nj50 NaOH=.00096 gram 80 4
N. B. Method by Muer, J. Ind. Eng. Chem., 3, 553 (1911). When the sulfate
is 25 p.p.m. or more, the determination may be made by the turbidimeter method
direct on 100 ml. For less quantities, larger amounts of water are taken and evaporated.
Sodium, Potassium and Lithium
Sodium and Potassium

The filtrate cont(tins sodium and potassium and lithium, and may be used
for such unless the water is highly mineralized" in which case a new portion,
100 ml. to 500 ml., should be taken.
Evaporate to dryness, add saturated solution of barium hydroxide in excess,
filter, wash lvith hot water, add to the filtrate ammonium carbonate in excess and
a few drops of ammonium oxalate, boil, filter, evaporate again to dryness and'
dry at a high temperature to expel excess of ammonia salts. Redissolve, add
slight -excess of ammonium carbonate again and continue until no further precipitate is formed on such addition. Evaporate to dryness in a weighed platinum dish, remove ammonium salts by high-temperature drying, and weigh the
combined chlorides of sodium and potassium. Moisten with about 25 fil. of~'
water and a few drops of HCI and add from 1 ml. to 5 ml. of 10% solution of
platinic chloride (1 ml. to each 25 milligrams to 30 milligrams total chlorides).
Evaporate to dryness on water bath, take up and wash with 95% alcohol until
filtrate is free from color. Dry, redissolve precipitate, washing through the
filter paper in hot water. Evaporate again to dryness and weigh as K 2 PtCls.
K 2PtCIs X.16l =K.
K 2PtCl6 X .307 = KCI.
Deduct from combined weight of chlorides.

Remaining NaCIX.394=Na.

NOTE.-When separation is not necessary, the combined chlorides are calculated

as sodium chloride and reported as sodium and potassium chlorides.
See pp. 878-881 for direct methods .for Sodium.
Lithium 10

Use a large quantity of the sample. Obtain the combined chlorides of

sodium, potassium and lithium. Transfer the combined chlorides to a small
Erlenmeyer flask (50 or 100-ml. capacity) and evaporate the solution nearly,
but not quite, to dryness. Add about 30 m!. of redistilled amyl alcohol.
Connect the flask, the stopper of which carries a thermometer, with a condenser 11 and boil until the temperature rises approximately to the boiling point
of amyl alcohol (130 C.), showing that all the water has been driven off. Cool
slightly and add a drop of hydrochloric acid to convert small amounts of
lithium hydroxide to lithium chloride. Connect with the condenser and continue the boiling to drive off again all water and until the temperature reaches
See Standard Methods of Water Analysis. A. P. H. A. (1920), page 60.
The amyl alcohol may be boiled off without the use of a condenser, but the vapors
are very disagreeable.
10
11

2066

WATER ANALYSIS

the boiling point of amyl alcohol. The content of the flask at this time is
usually 15 to 20 ml. Filter through a small paper- or ia Gooch crucible into a
gr~duated cylinder and note exact quantity of filtra,t~, which determines the
subsequent correction. Wash the precipitate with small quantities of dehydrated amyl alcohol. Evaporate the filtrate and wasHings in a platinum dish
to dryness on the steam bath, dissolve the residue in wa~er, and add a few drops
of sulfuric acid. Evaporate on a steam bath and expel the excess of sulfuric
acid by ge!ltle heat over a flame. Repeat until carbonaceous matter is cOrnf
pletely burned off. Cool and weigh the dish and contents. Dissolve in a sman
quantity of hot water, filter through a small filter, wash, and return filter to
dish, ignite and weigh. The difference between the original weight of dish and
contents and the weight of tile dish and small amount of residue equals the
weight of impure lithium sulfate. The purity 0:1' the lithium sulfate should be
tested by adding small amounts of ammonium phosphate and ammonium hydroxide, which will precipitate any magnesium present with the lithium sulfate.
Any precipitate appearing after standing over night should be collected on a
small filter and weighed as magnesium pyrophosphate, calculated to sulfate, and
subtracted from the weight of impure lithium sulfate. From this weight subtract 0.00113 gram for every 10 ml. of amyl alcohol filtrate exclusive of the I
amyl alcohol used in washing residue because of the slight solubility of solid
mixed chlorides in amyl alcohol. Calculate 'lithium from the corrected weight
of lithium sulfate. Dissolve the mixed chlorides from flask and filter with hot.,
water, evaporate to dryness, ignite gently to remove amyl alcohol, filter and
thoroughly ,wash; concentrate the filtrates and washings to 25 to 50 ml.
To the weight of potassium chloride add 0.00051 gram for every 10 ml.
of amyl alcohol used in the extraction of the lithium chloride, which corrects
for the solubility of the potassium chloride ,in amyl alcohol. Calculate to
potassium.
The weight of sodium chloride is found by subtracting the combined
weights of lithium chloride and potassium chloride (corrected) from the total
weight of the three chlorides. Calculate sodium chloride'to sodium.
Alkalinity

In ordinary cases titrate with N /10 or N /50 H 2S0 4 , using methyl orange as
indicator. Special cases will be considered later.
Reagents.-Sulfuric a,cid, N /10. Methyl orange. PhcIlolphthalein.
Operation.___'::;250 ml. of 'water in 400-ml. beaker or a casserole are titrated
with N /10 H 2S0 4 , using two to five drops of methyl orange indicator (or 50 ml.
can be similarly titrated withN /50 H 2S0 4).
To colored waters add 5 to '50 m!. H 20 2 to decolorize. Start with 5 mI., and
if this is insufficient, add more, 5 m!. at a time. This totally clears some waters;
in all cases there is a great improvement. In c6rrecting for the acidity of
H 20 2, our laboratory adds 0.1 m!. to the reading for each 5 ml. H 20 2 used.
Calculate for 250-ml. sample.
I
No. m!.X4X.005=g. per liter CaCOa
No. m!. X 4,.X 58.4 X .005 = grs. per g!lJlon CaC03
Or
X 1.168 = grs. per gallon GaGOa

WATER ANALYSIS

. 2067

Distilled water, and neutral waters containing magnesium chloride and

magnesium sulfate frequently give an alkaline reaction when used with methyl
orange. In such cases from .2 to .8 ml. N /10 acid are required to discharge the
alkaline color of the methyl orange. Such a procedure would suggest to the
operator that the waters were alkaline. However, if such neutral waters are
boiled with phenolphthalein as an indicator for twenty minutes and no pink
color develops, the waters are not alkaline but neutral. The use of a blank of
.2 ml. is of no value under such conditions and it appears to the writer as much
the safest way when the titration is under 1 ml. of N /10 acid that the water be
boiled with phenolphthalein in an effort to determine absolutely whether this
water is alkaline, due to the presence of a carbonate as indicated by the methyl
orange, or whether the alkalinity is entirely due to the hydrolyzing of the cal~
cium or magnesium base present in the absence of alkali. When no pink color
is produoed the water should be pronounced neutral.
Phenolphthalein may also be used as indicator on another 250-ml. portion,
using the above procedure. This titration in connection with the methyl orange
titratjon makes possible a determination of the relation of carbonate, bicarbon~e and caustic alkalinity.

?:_~ following is adapted from a table on page 66, Standard Methods of

Water Analysis of the American Public Health Association, and is of value in
showing the relation of the various titrations. Methyl orange has been used in
place of erythrosine.
.
TABLE SHOWING RELATION BETWEEN ALKALINITY BY PHENOL~
PHTHALEIN AND THAT BY METHYL ORANGE IN PRESENCE OF
BICARBONATES, CARBONATES AND HYDRATES.
Bicarbonates.

P=O .............. .
P<!M ............ .
P=!M ............ .
P>!M ............ .
P=M ........... ,. ..

M
M-2P

o
o

Carbonates ..

Hydrates.

o
o

2P
2P
2(M-P)

2P-M

M-

M = Methyl orange alkalinity.

P = Phenolphthalein alkalinity.

PROPOSED METHOD OF TEST FOR THE DETERMINATION

OF THE HYDROXIDE ION IN . INDUSTRIAL WATER12
Scope.-There is no known method of chemical analysis by which the exact
concentration of the hydroxide ion in water may be determined. Interpreta12 Under the standardization procedure of the Society, this method is under the
jurisdiction of the A. S. T. M. ComInittee D-19 on Water for Industrial Uses. This
is a proposed draft and is published as information only. Comments are solicited and
should be addressed to the Headquarters of the Society, 260 S. Broad St., Philadelphia,
Pa.

2068

WATER ANALYSIS

tion of any titration method is complicated by the fact that when salts such
as carbonate, phosphate, aluminate, silicate or chromll-te, as well as some. types
'of organic material, are present, they are titrated to a greater or less extent
along with the hydroxide ion. The standard methqd, employing strontium
chloride, minimizes interference by sulfate, as well as; by carbonate and phosphate. A method similar in procedure, but employing barium chloride, is
'outlined in the appendix following the standard method.

STANDARD METHOD

13

This method is known as the strontium chloride method. It is based upon

the fact that since strontium carbonate and strontium phosphate are much less
soluble than strontium sulfate, it is possible toJree the solutiOli of carbonate and,
phosphate ions without precipitating sulfate. This procedure minimizes
turbidity and permits the siirect titration of the hydroxide ion.
.
Application.-This method is applicable in the examination of alkalinl
industrial waters such as boiler waters and boiler feed waters under conditions
where the water is practically free from organic matter, silicates and aluminat~s.
If these ions are present in concentrations very low as compared with the concentration of the hydroxide ions the method may be used. It is not recommended where the concentration of these ions is high.
Solutions Required.-The following solutions will be required:
(a) Acid Solution.-Hydrochloric or sulfuric acid, N/50, standardized,
prepared from concentrated acid, reagent grade.
(b) Strontium Chloride Solution.-4.5 g. SrCI 2 p'H 2 0, reagent grade, per
liter of distilled water.
f
(c) Phenolphthalein Indicator Solufion.-5, grams of phenolphthalein in 1
liter of 50% solution of ethyl alcohol.
Cd) LaMotte Purple I_ndicator Solution.-l % solution.
Procedure.-(a) 100.0 ml. of the sample containing_,not more than 150
p.p.m. of hydroxide ion shall be measured into a 250-ml. Erlenmeyer flask. If
the concentration of hydroxide is known to be higher than 150 p.p.m., a correspondingly smaller sample shall be measured.
NOTE.~Thus, if the concentration of hydroxide were approximately 300 p.p.m.,
a 50-ml. sample should be measured.
f

Edited from a method prepared by Subcommittee 8 of the Joint Research Committee on Boiler Feedwater Studies based on an investigation conducted under the
direction of this Subcommittee which was described in the research publication of the
American Society of Mechanical Engineers, "Determination of Carbonate, Phosphate
and Hydroxide in Boiler Waters," published in September, 1933; The same general
method, independently edited for use in the examination.of stationary boiler waters, is
included in the eighth edition of "Standard Methods_f.or-the Examination of Water and
Sewage," published by the American Public Health Association.
13

WATER ANALYSIS

2069

(b) To this sample shall be added quickly 1.0 m!. of the strontium chloride
solution ror each 10 p.p.m. of carbonate ion present, 1.0 mL for e~h 10 p.p.m.
of orthophosphate ion present, and 4 mL excess.
NOTE.-If the magnitude of phosphate and carbonate concentrations in the sample
is not known it will be necessary to add sufficient strontium chloride solution to precipitate what is thought to be the maximum concentrations of these ions that could
exist in the water. The excess strontium chloride will produce no interference but may
increase the turbidity of the sample by precipitation of strontium sulfate.

(c) A stopper shall be placed at once very lightly in the flask. The flask
shall then be heated to boiling with the stopper in place but not allowed to boil
fo~e than two or three seconds. After standing hot for two or three minutes the flask shall be cooled, under water if desired. The stopper shall be
in the flask at all times to prevent carbon dioxide from entering.
Cd) If phosphate is absent, two or three drops of the phenolphthalein solution
shall be added and the sample shall then be titrated with the Nj50 acid until
the phenolphthalein is colorless. The sample shall be swirled steadily during
the addition of the acid to prevent redissolving of the precipitate.
If phosphate is present, two or three drops of the LaMotte Purple solution
shall be added and the sample shall then be titrated to the endpoint of this
indicator with the N/50 acid. A comparison blank should be used if the
operator is not familiar with the endpoint of this indicator. The sample shall
be swirled steadily during the titration.
NOTE.-The LaMotte Purple indicator is recommended when phosphates are present,
since in this case the phenolphthalein endpoint at a pH value of 8 fades rapidly due to
the resolution of strontium phosphate. At the LaMotte Purple endpoint above 9.3
this does not occur so readily. If LaMotte Purple solution of the concentration employed as a pH indicator is used in place of the 1 % solution, 1 ml. is required.

Calculation.-The concentration of the hydroxide ion in terms of parts per

million shall be calculated as follows:
OH- = 340 XjXV
S

where

S =the 'volume of the sample in mI.,

j=the acid correction factor,
V = the volume of acid in ml. used in the titration.

Accuracy.-When the concentration of hydroxide ion is greater than approximately 100 p.p.m. and the concentration of carbonate or of orthophosphate is
not greater than 100 p.p.m. the values found should be accurate to within 2%.
While the accuracy on a per cent basis will decrease with decrease in concentration of hydroxide ion or increase in concentration of carbonate or orthophosphate, the values found should be accurate to within a few parts per million
unless other interfering substances such as chromates, aluminates, silicates, or
some organic materials are present.

2070

WATER ANALYSIS
TOTAL CARBON DIOXIDE !

The most accurate method for total carbon dioxide 'is that known as the
evolution method which involves the decomposition of cat bonates by an excess
of acid, the carbon dioxide evolved being absorbed in ia barium hydroxide
solution of known acid equivalent. Titration of the remaining barium hydroxide with standard acid coupled with a separately determined pH value
makes possible a calculation of carbonate and bicarbonate ions. This method
is applicable where phosphates, silicates, chromates, and aluminates are present,
but is affected by sulfides, sulfites, and other salts of volatile acids.

..

Evolution Method

14

Apparatus.-A recommended form of apparatus is illustrated in Fig. 289.

It consists of a box, open at the front, painted internally with a fiat white paint
and suitably illuminated. Upon
c
13
A
this box three 50-m!. burettes are
-0
-0
-0
mounted non-rigidly by means' of
rubber stoppers fastened to the
under side of the top. A small circulating pump sealed with mercury
is' connected as shown so that when
two 500-ml. Erlenmeyer flasks are
- 50
- 50
- 50
attached to the stoppers of burettes
Band C, a closed system is established.

Solutions Required: Acid Solutions.-(A) Sulfuric acid solution,

N/lO, standardized, prepared from
concentrated H 2S0 4 , reagent grade.
(B) Hyd~bchloric acid solution,
N/50, standardized, prepared from
concentrated Hel, reagent grade.
20 m!. iof phenolphthalein solution
shall be added to one liter of acid
before it is standardized.
Barium hydroxide solution, N/50,
prepared from Ba(OH)_2: 8H 20, reagent grade. The acid equivalent
of this solution shall be determined
FIG. 289.-Evolution carbonate determina- by using the evolution apparatus.
tion apparatus.
A 50-ml. saPlple of barium hydroxide
solution shall be measured from burette A into absorption flask E, which shall
then be securely attached to the rubber stopper on burette B. 'Thereafter 200
ml. of distilled water free from carbon dioxide shall be placed in flask D, the
pump started, and the gas circulated for about 10 minute,S. The barium
14 The Determination of Total Carbon Dioxide and of /eiirbonate and Bicarbonate
Ions in Water. A. S. T. M. D-19, Subcomm. II.

WATER ANALYSIS

2071

hydroxide :in flask E shall then be titrated with standard 0.02 N hydrochloric
acid from burette B to the endpoint of phenolphthalein. The procedure outlined also serves to calibrate the apparatus against the air in the system.
Indicator~(A) Phenolphthalein indicator solution. 5 g. of phenolphthalein cHSsolved in 1 liter of 50% ethyl alcohol. (B) Methyl orange indicator solution. 1 g. of methyl orange dissolved in 1 liter of distilled water.
Procedure; Manipulation.-(A) 50 mi. of N/50 barium hydroxide solution
shall be measured into fll\sk E from burette A, and flask E shall then be
secur~ly attached to the rubber stopper on burette B.
(B) A 200-ml. sample of the water to be tested, containing not more than
100 p.p.m. of total carbon dioxide, shall then be placed in flask D, phenolphthalein indicator added, and the flask attached securely to the rubber stopper of
burette C. The hot plate shall be put into position. The three-way cock
shall then be opened to the atn:lOsphere and the pump startt!d. When bubbles
appear in flask D the three-way cock shall be adjusted to circulate gas through
the closed system comprising flasks D and E.
(CI) With the gas circulating (to stir the solution), N/IO sulfuric acid shall
be added from burette C to the sample in flask D until the phenolphthalein
endpoint is reached. The amount of acid used shall be recorded.
(D) Two drops of methyl orange' solution shall then be added to flask D and
the titration continued to the methyl orange endpoint. The amount of acid
used shall be recorded and then 20 ml. of acid added in excess.
,
(E) With the gas still circulating, the solution in flask D shall be heated
until it just reaches the boiling point and maintained at this point for about
5 minutes. The hot plate shall tl~en be turned off and the circulating pump
stopped.
(F) The solution in flask E shall then be titrated with N /50 hydrochloric
acid to the endpoint of phenolphthalein with,constant swirling of the flask.
Calculation.-The results shall be calculated as follows:
(A) The concentration of total carbon dioxide in parts per million of CO 2
shall be calculated from the expression:
total CO 2 = (A - B) X2.2 XI
where A = volume of 0.02 N acid equivalent to volume of N /50 Ba(OH)2 used,
in mi.,
B = volume of N /50 acid used, in ml.,
f = correction factor for N /50 acid.
(B) The concentrations of carbonate and bicarbonate ions in parts per
million of CO a shall be estimated by means of the curves in Fig. 290. To use
these curves, which express the respective fractions of the total carbon dioxide
present as undissociated carbonic acid, as bicarbonate ion, and as carbonate ion,
the pH of the original sample must be known. From the determined value of
pH, the value of the fraction of the total carbon dioxide present as carbonate
ion, as bicarbonate ion, and as undissociated carbonic acid, shall be read respectively from curves a, band c, of Fig. 290. The concentrations of carbonate
ion, bicarbonate ion, and undissociated carbonic acid in terms of parts per
million, shall then be calculated from the expression

WATER ANALYSIS

2072

C03= =(A-B)X3X!Xa
HCO a =(A-B)X3.05X!Xb
H 2 C0 3 = (A - B) X3.1 X!Xc

where a, b, and c are the fractions determined respectiveLy from curves a, b, and
c of Fig. 290 and the other symbols have the meaning defined above.
I

N oTE.--8ince the preceding method of estimating ion concentrations is based upon

data for the ionization of carbonic acid at .25 0 C., the values obtained hold for this'
temperature only. At any other temperature the ion concentrations will have the other
values, which may be calculated if the dissociation constants of carbonic acid at this
temperature are known.

--- ,

1.0

0.9

I~

c\
\

.. 0.8

:!!

II

QO.7
c:

-e o~6

-;;; 0.5

IIi)

b\

\7 .
V

I
J

/\

._ 0.4

~\

50.3

1:;

aJ

..tOo2

bf

0.1
../

"

\c
\

\
~b

!/

<'
9

10

/I

"',,12

13

JIj.

PH
I
FIG. 290.-Fraction of Total Carbon,Dio:X:i.de as pH.

Accuracy.-This method will yield results for total carbon dioxide accurate
to 5% for concentrations as low as 10 p.p.m. of total CO 2 ,
Acidity
For acidity use NjlO Na 2CO. and 250 ro!. water.
Acidity due to

Indicator.

Carbonic and sulfuric acids, also Fe and Al

sulfates ................................

Phenolphthalein

Cold

Sulfuric acid, also Fe and AI sulfates .........

Phenolphthalein

Boiling

Sulfuric acid alone .........................

When desired, 20 m!. NjlO H 2S0 4 may be
added. Boil fifteen to twenty minutes, cool
and titrate, noting the excess of acidity over
the original 20 m!.

Methyl orange

_.- -

Hot or Cold.

Cold

WATER ANALYSIS

2073

Hydrogen-Ion Concentration
It is freqUently desirable to know, particularly in connection with waters
for certain industrial uses and the filtration of supplies using alum as a coagulant, the actual reaction of the water in terms of the hydrogen-ion concentration
in addition to the familiar titration with methyl orange which gives the total
available alkalinity of the water and does not indicate the actual dissociation
or pH value. The proper adjustment of the pH value along with treatment
using alum as a coagulant is necessary for economical and efficient operation
of this process. IS In the use of water for steam making and for certain manufacturing processes the added data given by this determination are making possible more accurate adjustments and more advantageous selection of supplies
and treatments.
The most accurate method for making these measurements is by means of
the, hydrogen electrode; 16 however, colorimetric methods have proved satisfactory for use with waterP The following standard buffer solutions proposed
by Clark and Lubs 18 together with their recommendations for the preparation
and use of indicators make a convenient set of standards for this purpose,
using 10 m!. of the standard buffer solution with five drops of the indicator
compared with the sample used in the same proportions.
.
N/20 NaOH per
Decigram
Indicator
pH Range
7.4
Methyl red, ................... , . " 4.4....u.0
Brom thymol blue .................. 6.0-7.4
3.2
Cresol red .... , ............. , ...... 7.2-8.8
5.3
Thymol blue ............... : ....... 8.6-10.0
4.3

For the preparation of these solutions one decigram of the dry powder is
ground in an agate mortar with the corresponding quantity of N /20 N aOH.
When solution is complete, the thymol blue and brom thymol blue should be
diluted, to a 0.04% solution and the cresol red and methyl red to 0.02% solution.
The American Petroleum Institute 19 recommends the following set of
standards.
Color Change
Name
pH Range
Yellow-blue
Brom phenol blue. . . . . . . . . . . . . .. 3.0-4.6
Yellow-blue
Brom cresol green ............... 4.Q.-5.6
Yellow-red
Chlor phenol red ....... : ........ 5.2....u.8
Brom thymol blue ............... 6.Q.-7.6
Yellow-blue
Yellow-red
Phenol red ..................... 6.8-8.4
Cresol red ... " ................. 7.2-8.8
Yellow-red
Solutions: Mj 5 Acid, Potassium Phthalate.- Make up a solution of potassium hydroxide dissolving about 60 grams of a high grade sample in about 400
15 J. Ind.'Eng. Chern., 14;395 (1922), 1038: Chern. Met. Eni' May 3,1922, Vol. 26,
No. 18; J. Am. W. W. Ass'n, Vol. 11, No.3, May, 1924.
16 Clark, The Determination of Hydrogen Ions (Williams.an Wilkins, 1920).
17 See also Index for additional discussion on colorimetric pH determination.
18 Clark (see note 2).
.
19 Code of Recommended Practice for the Treatment of Oil Field Waters.
Report
of Sub-Committee "C." Frank M. Beeson, Chairman. D. & P. E. Bull. No. 204, p. 8.
American Petroleum Institute. Jan. 2, 1930.

2074

WATER ANALYSIS

ml. of water. To this add 50 grams of resublimed anhydrid of orthophthalic

acid. Test a cool portion of the solution with plienolp~thalein. If the solution
is still alkaline, add more phthalic anhydrid; if acid, more KOH. When
roughly adjusted to a slight pink with phenolphthalin,. add as much more
phthalic anhydrid as the solution <contains and heat until all is dissolved.
Filter hot and allow to crystallize slowly. Recrystallize at least twice from
distilled water. Dry at 110-115 C. to constant w~ight. A M/5 solutioh.
contains 40.828 grams per liter.
MIS Potassium Chloride.-The salt should be crystallized three or fou'r
times and dried at 120 C. for two days. A M/5 solution contains 14.9)2
grams per liter.
MIS Acid Potassium Phosphate.-Recrystallize three times from distilled
water, dry at 110-115 C. A M/5 solution contains 27.227 grams per liter.
MIS Boric Acid, MIS Potassium Chloride.-MI5 Boric acid: Recrystallize
three times from distilled water, air dry in thin layers between filter paper aft'ti
finally in a desiccator over CaCl 2 to constant weight. A M/5 solution contains
12.369 grams per liter of boric acid and 14.912 grams of potassium chloride.
MIS Sodium Hydroxide.-Great care should be used in preparing this
solution as free as possible from carbonate. It is preferable to use a factor
rather than attempt exact adjustment. This solution should be kept in a
paraffined bottle.
PHTHALATE-NaOll MIXTURES 20
pH
4.4
4.6
4.8
5.0
5.2
5.4
5.6
5.8
6.0

50 ml. M/5 KH Phthalate

50 m!.
"
50 m!.
"
50 ml.
"
50mt
"
50 m!.
"
50 m!.
"
50 ml.
"
50 ml.
"

7.50 m!. M/5 NaOH

12.15 m!.
"
17.70 m!.
"
23.85 m!.
"
29.95 m!.
"
35.45 m!.
"
39.85 m!.
"
43.00 m!.
"
45.45 m!.
" I

50 ml. M/5 KH 2P0 4

50 ml.
"
50 ml.
"
50 m!.
"
50 ml.
"
50 m!.
"
50 m!.
"
50ml.
"
50 m!.
"
50 ml.
"
50 ml.
"

"
"
"
"
"
a"

"
"
"
"
"
"

"
"
"
"
"
"
"

" "

'f

KH 2P0 4-NaOH MIXTfRES

6.0
6.2
6.4
6.6
6.8
7.0
7.2
7.4
7.6
7.8
8.0

Dilute to 200 m!'.

"
"
"
"
"
"
"

"

5.70 m!. M/5 NaOH

8.60 m!.
"
12.60 m!.
"
17.80 ml. I
"
23.65 ml.
"
29.63 m!.
"
35.00 m!.
"
39.50 m!.
"
42.80 ml.
"
45.20 ml.
"
46.80 ml.
"

Dilute to 200 m!.

"
"
"

" " "

" " "
" " "
" " "
"" ... .._,!
"
"
"
"
"

"
"
"
"
"
" "
"
"
"

-Ii

"
"
"
"

"

BORIC

ACID.

KCI-NaOH

MIX~~RES

7.8 50 m!. M/5 H 3B0 3, M/5 KCI

2.61 m!. M/5 NaOH Dilute to 200 ml.
8.0 50m!."
"
3.97 ml.
"
,,'" "" "" ""
8.2 50 ml.
5.90 ml.
"
I"
8.4 50 ml.
8.50 m].
"
"
"
" " "
II'
8.6 50 ml.
12.00 ml.
" " " "
"
"
_.20 Clark, The Determination of Hydrogen Ions (WilliamS and Wilkins, 1920).

>

WATER ANALYSIS
8.8
9.0
9.2
9.4
9.6
9.8
10.0

50 ml. M/5 HaBOa, M/5 ReI

"
50 m l . "
50 m!.
"
"
50 ml.
"
"
50 ml.
"
"
."
50 ml.
"
.1<
50 m!.
"

16.30 ml. M/5 NaOH

21.30 ml.
"
26.70 ml.
"
32.00 m!.
"I<
36.85 m!.
I<
40.80 ml.
43.90 ml.
"

2075
Dilute to 200 m!.

"

. "

"
"
"

" " "

" " "
" "
"
" " "
" " "
I<

I<

The standard buffer solutions prepared as. described above may be used as
permanent standards. The methyl red series often deteriorates within a short
time, but if the series is checked at frequent intervals, any change may be
corrected. Standards using indicators other than methyl red have been used
for four weeks before any change in color could be detected. In order to
prevent the growth of mold, one drop of toluol may be added to each tube
before sealing. If the tubes are corked, paraffined stoppers should be used and
for accurate comparison tubes of equal diameter should be chosen.
Free Carbonic Acid 21
Reagents.-Either standard N /10 sodium carbonate or standard N /22
,!odium carbonate. For the latter dissolve 2.41 grams of dry sodium carbonat~
in one liter of distilled water which has been boiled and cooled in a carbon
dioxide free atmosphere. Hold both solutions in glass bottles protected by
tubes filled with soda-lime.
One ml. N /10 N a 2 CO a = 2.2 milligrams CO 2
One ml. N/22 Na 2 CO a =1.0 milligram CO 2
Operation.-With N /10 sodiu'm carbonate titrate 250 ml. of sample in 400ml. beaker, using phenolphthalein as indicator. First faint but permanent pink
denotes endpoint.
Using 250 ml.
No. ml. N/lO Na 2COa X8.8=COi parts per million.
No. ml. N/lO Na 2 COa X.513=C0 2 grains per gallon.
With N /22 sodium carbonate solution, use 100 ml. of sample, preferably
in 100-ml. Nessler tube, titrate and rotate the tube until faint but permanent
pink color 30 seconds without fading is produced, using phenolphthalein as
indicator.
Using 100 ml.
No. ml.N/22 Na 2CO a X10=C0 2 parts per million.
No. ml. N/22 Na 2 CO a X.583=C0 2 grains per gallon.
Chlorine
Titrate 100 ml. of water, using 1 ml. qf 10% potassium chromate as an indicator, with N /10 AgNO a to first permanent indication of the red silver chromate.
(Acid waters should be neutralized and sulfide waters boiled with a drop or so of
nitric acid and then neutralized for reliable results.)
21 For criticisms of this method see Z. Nahr. Genussm., 24, 429 (1912); also C. A.,
7, 38 (1913).
.

2076

WATER ANALYSIS
No. rul. X3.42 = grs. per gallon NaCl.
No. ml.X58.46=parts per millionlNaCL
No. mL X 35.46 = parts per million CL

NOTE.-Where qualitative test shows chlorine to be h1gh, smaller portions of the'

sample should be taken, either by certified pipette, or burette, and when the titration
with Nj10 ~gN03 is less than .2 ml., Nj50 or F'l"jlOO AgNO s should be used for accura~.
Nitrates
(Also see Sanitary Analysis)

PHENOLSULFONIC ACID METHOD

Reagent: L Phenolsulfonic Acid.-Dissolve.25 grams of pure white plienol

in 150 mL of pure concentrated sulfuric acid, add 75 ml. of fuming sulfuric a~d
(15% 80a), stir well and heat for two hours at about 100.22
2. Ammonium hydroxide 1: 1.
3. Sodium Carbonate.-l0% solution of anhydrous N a 2COa
4. Standard Nitrate Solution.-Dissolve '0.72 gram pure or c.p. potassium
nitrate in 1 liter of distilled water. Evaporate carefully 10 ml. of this solution
on water bath, moisten thoroughly with 2 ml. of solution No.1 and dilute to 1
liter. 1 ml. of this = 0.001 milligram of nitrogen.
Determination.-Carefully evaporate )00 ml. of water after the addition of
2 ml. of sodium carbonate solution. After this evaporate to dryness, cool and
add 2 ml. phenol sulfonic acid (No.1), mixing well with' a glass rod. Dilute
with 25 ml. of distilled water and add an excess of ammonium hydroxide, making up to 100 ml. volume with distilled water.
The dilute solution is then compared with the standard solution.
NOTEs.-When the chlorides are over 100 parts per million in the original sample
they should be removed with the addition of silver sulfate in the solid form and the
water should be filtered prior to evaporation. It is 'for this reason that ammonium
hydroxide is used to develop the color instead of potassium hy'(iroxide; which is frequently
recommended, as a slight excess of silver sulfate will result'in a dirty precipitate when
using potassium hydroxide, whereas the use of ammonia has no effect. Furthermore,
the filtration of a turbid nitrate solution does not result in a satisfactory color, as would
be the case without filtration.
Permanent standards can be made by procedure given on page 2053, or standards
can be made using tripotassium nitrophenoldisulfonate. The._following method is
given in an article in J. Am. Chem. Soc., 33, 381 (1911).
- .__
The theoretical amount of powdered potassium nitrate is added to the disulfonic
acid reagent in small pinches at a time (for each ml. of reagent 0.1076 gram KNO.),
stirring thoroughly after each addition. The product is then diluted, treated with
dry barium carbonate to a deep yellow color, filtered.and the precipitate washed with
boiling water to remove the barium salt which is but slightly soluble in cold water. This
extraction must be thorough. Filtrates and washings are united, the barium removed
by the addition of potassium carbonate until alkaline, the solution filtered and the
filtrate concentrated and crystallized. The solution may then be easily purified by
crystallization. In preparing the standards, however, solutions made from .known
amounts of nitrate standards will match up with this recrystallized solution, and by
means of proper dilutions the series of standards can be made. Standards made this
way will last for many months, whereas standards made from tne standard nitrate
solution are apt to lose value in a month's period and should..be made up very frequently .
22

J. Am. Chem. Soc., 33, 382,

19~1.

. -'

WATER ANALYSIS

2077

Where nitrates are high, 85 to 90 parts per million, or 5 grains per gallon
and over, colorimetric methods do not always give reliable results, and 500 m!.
of the water should be first boiled with a slight excess of acid, then made alkaline
with sodium or potassium hydroxide, reduced with 10 grams each of powdered
Zn and Fe, or 10 grams powdered AI, and distilled into an excess of N /10 or
N /100 HCI, as the case may be, and titrated back, using cochineal as indicator,
and calculating the ammonia absorbed to NO a or Ca(NO a)2 as desired. (Where
free ammonia or its compounds are present corrections must be made.)
A recent modification of this method depends upon the absorption of
ammonia into a solution of boric acid (5 grams boric acid in 100 m!. of water).'
Due to the very weak acidity of the boric acid, it is possible to titrate the
ammonia direct with standard acid, using methyl orange as an indicator, and
this has the advantage of doing away with two standard solutions. The boric
acid strength is based upon 5 grams of boric acid to 100 m!. of water to each
.2 gram of ammonia absorbed. It is stated that even the cooled condenser is
unnecessary in this absorption.23
Ammonia and Its Compounds
Place 500 m!. or less in an 800-m!' Kjeldahl flask, make alkaline and distill
into N /10, or weaker, HCI, titrate with cochineal or Nesslerize. (See Sanitary
Analysis, page 2049.)

Total Mineral Residue

Use a clean weighed platinum dish. Evaporate 50 m!. (certified pipette) to
dryness at about 1300 C. and bake for at least thirty minutes at that temperature. Ordinary water-bath temperature will not remove water of crystallization from Na 2S04 or CaS04. Weigh to the fourth decimal or .0001 gram .

Weight X 1168 = grains per gallon.

0.1 milligram=2 parts per million.
Residues of acid waters should be ignited to a dull red heat. Where the
acidity is Iowa drop or so of sulfuric acid should be added to assure the fixation
of all sodium and potassium salts as the sulfate. The ignition should be complete so that no free acid is left behind and to assure the decomposition of all iron
compounds to the oxide form. In calculating, correction must be made for the
change in the iron salts and all other compounds converted to the sulfate form
for comparison with the sulfated residue, and then the proper corrections made
to give the theoretical residue on the original water.
Residues with much organic matter, after weighing, may be gently ignited
until the carbon has been burned off, cooled, recarbonated with tested (NH 4)2CO a, dried and again weighed. The difference in weight after titrating for
possible volatilized chlorides gives approximately the organic matter present.
,

The Volumetric Determination of Ammonia. L. W. Winkler, Budapest. Z.

angew. Chern., 26, I, 231 (1913).
Determination of Ammonia by the Boric Acid Method. L. W. Winkler, Z. angew.
Chern., 27, I, 630-2 (1914).
.
The Determination of Ammonia by the Boric Acid Method. E. Bernard, Z. angew.
Chern., 27, I, 664 (1914).
23

2078

WATER ANALYSIS

Waters high in easily decomposed MgCl 2 or Ca(~03)2 should be evaporated'

with a few drops excess of H 2S0 4, or Na 2CO a, I!-nd t~e residue compared with
an addition of all bas.es calculated to the sulfate form, or corrected for added
carbonate.

NOTE.-When acid is used, ignite to a dull red heat; when carbonate, evaporate
as in the case of the original residue.
.I
Hydrogen Sulfide
Due to the fact that hydrogen sulfide is frequently very transient and often
oxidizes to sulfate in transit, it is advisable to collect this sample in a special
container at the point of sampling. Two or three bottles holding exactly 250
m!. of water each, are used, each bottle containing 50 mI. N 1100 iodine solution.
Mter filling, the bottle is sealed. ,The s~mple is titrated with standard N /1~0
sodium thiosulfate upon receipt at laboratory, at which time a blank is run,
using 50 m!. iodine solution made to mark with distilled water. The difference
between the titration of the sample and the blank repreaents hydrogen sulfide
present.
Iodine valueXO.1263X4=sulfur value grams per liter.
When hydrogen suHide is determined on sRmples Rlre8dy collected 8nd submitted, 100 m!. of the sample is oxidized with bromirte (about 10 m!. is sufficient), HCI is also added, and it is then,boiled until colorless, after which sulfates are precipitated with BaCh as already described. The difference between
the oxidized and unoxidized sulfates is calculated to E: 2S.
Oil
Frequently waters from condensing engine, or after passing heaters or oil
separators, still contain oil in small quantities. The following method has
been found most satisfactory:
Reagents: Ferric Chloride Solution.-(10 grams ot iron dissolved in 200 ml.
HCI, oxidized with HNO a and made to one liter.) I
Ammonia c.p.
I
Operation.-Add to the water taken in a large bell.ker or flask 5 m!. of the
"ferric chloride" solution and heat nearly tOI boiling; then add ammonia in
excess, to precipitate the iron (which precipitate contftins all the oil), and boil
-_ ........ _
for two minutes.
'
NOTE.-If the oil exceeds 0.4 grain per gallon, use 500 ml., or less for the determination; if below 0.4 grain per gallon use 1 liter.
Allow to stand a few minutes and filter through a 15 cm. filter paper which
has been previously extracted with ether, transferring the precipitate onto the
paper with hot water, and washing three or four timefj with hot water. Then
dry both filter and precipit~te in the water oven at 100 C. and when dry,
extract with ether in the Soxhlet in the usual way, eVll.porate the ether extract
and weigh the remaining oil.
I
Optional Method for Oil.-Shake up the sample-of water in a separatory
funnel with 25 mI. ethyl ether or 25 ml. of benzoland fleparate the extract con-

WATER ANALYSIS

2079

taining the oil. Three extractions, as direhted above, will remove all oil from
the water. The solvent is expelled by warming gently and the oil is weighed.
Dissolved Oxygen
Use the Winkler Method 24

Reagents: No. 1. Manganous SUlfate Solution.-48 grams manganous

. sulfate dissolved in 100 m!. distilled water.
No.2. Iodide Solution.-36o. grams NaOH and 100 grams KI dissolved in 1
liter of distilled water
,
No.3. Concentrated hydrochloric acid, or sulfuric acid sp.gr. 1.4 (dilute
1 : 1).

No.4. Sodium Thiosulfate Solution.-Nj100 solution is made as needed

from the Nj10 stock solution.
NOTE.-Not permanent; should be frequently restandardized against N/100 potassium bichromate.
The addition of 5 mI. of chloroform plus 1.5 grams sodium or ammonium carbonate
to Ililch liter of solution on mixing will improve its keeping quality.
N/40 sodium thiosulfate containing 6.2 grams C.P. recrystallized salt per liter may
be preferred to N/100 strength. 1 m!. of this solution is equivalent to 0.2 milligram
oxygen by weight or 0.1395 mI. oxygen by volume, standard conditions.

5. Starch Solution. 25__The starch should first be made into a thin paste with
cold water and about 200 times its weight of boiling water stirred in and ,boiled
for a few minutes. A few drolls o.f chloroform will assist in preserving this
solution.
Collection of Sample.-A small-necked, 250-m!. bottle should be used, etched
or o.therwise marked, with its exact volume previously determined. The collection should be so arranged to. exclude outside air and result in several co.ntinuous
changes o.f the co.ntents before sto.Ppering, care being taken to exclude air
bubbles.
Operation.-To sample as received add, in both cases by pipette, delivering
below surface of water and away from the air, 2 m!. solution No.}. (manganese
sulfate) and No.2 (NaqH, KI). Restopper and shake thoroughly. After precipitate has settled add 2 ml. Hel or H 2 S0 4 and again mix by thorough shaking
until precipitate has completely dissolved, transfer 100-m!. to flask, and titrate
with solution No.4 (sodium thiosulfate), using starch as indicator near end as
the color approaches a faint yello.w.
.
N = mI. N /100 thiosulfate solution.
.
V = capacity of bottle less 4 mI. (vol. sol. 1 and 2 added).
O=the amount of oxygen in parts per million in water saturated at the same temperature and pressure.
24 Ber., 21, 2843 (1888).
Also Z. Ap.a!. Chem., 53, 665-72 (1914); C.F.C.A., 8, 674
(1915).
. '
25 Hale gives the following method.
"Rub 5 grams of potato starch with cold water
to a thin paste together with 10 milligrams of mercuric iodide. Pour into one liter of
boiling water and boil half an, hour."

2080

WATER ANALYSIS

(1) Oxygen ni
p.p.m.
(2) Oxygen in m!. per liter

0.0008N X 1,000,000 81\T .26

1 0 0 ' ... ~
=.7 oxygen p.p.m.

Oxygen p.p.m. (observ~d temp. and pres.) X100

(3) 0 xygen per cent sat urat IOn = .
'
.
SaturatIOn oxygen p.p.m! (observed temp. and ;pres.)

METHODS FOR THE DETERMINATION OF SMALL AMOUNTS

OF LEAD, ZINC, COPPER AND TIN
Very frequently a determination is desired, of materials which are apt to be
present in water due to the solvent action of such water upon pipes and containers. In most cases the estimates are made by colorimetric methods if the
amounts present are exceedingly smalL As/these determinations are made I'lnly
in rare cases it seems advisable to summarize, calling attention to the fact that'
. all methods may be found in full in any of the editions of Standard Methods for
Water Analysis gotten out by th{\ American Public Health Association.
Where any or all of the metals, lead, zinc, copper and iron are apt to 'be
presertt, a large quantity (1 to 4 liters), of the water is evaporated. The metals
are separated as sulfides with ammonia and hydrogen sulfide. The precipitate
after washing is dissolved in nitric acid and refiltered to remove suspended
matter and then concentrated with H 2S0 4
The lead is removed by taking up the concentrated solution 'with 50% alcohol
(100 mI. to 150 mI.), filtering and dissolving the precipitate in ammonium acetate, after which the solution is made to volume .and divided. One-half is
saturated with hydrogen sulfide water to get an,appr6ximate idea of the amount
of lead present. To the other half add two to three drops of acetic acid, then
an excess of .hydrogen sulfide water and compare the color with standards.
This gives lead.
The alcohol is removed from the filtrate by evaporation and it is then treated
with ammonia to remove possible iron. The filtrate from-the.iron precipitate
is neutralized with 2SD4, then 2 mI. concentrated H 2S0 4 and' t gram urea
added. Copper is removed by electrolyzing (two hours with 0'.5 ampere current). If the deposit is material it may be weighed as copper after washing with
alcohol and drying. When the deposit is extremely small it should be dissolved
in nitric acid, evaporated to dryness to remove a?id taken up in water, after
which potassium sulfide solution is added and the, color compared with standards. This gives copper .

Correcting for displacement for 300-m!. bottle, .8N = .&l1N; for 275-m!. bottle,
.8N=.812N.
No correction for displacement affects result .1 p.p.m._Qxygen:
Twenty-five m!. variation in capacity of bottl!J..a:trects result :01 p.p.m. oxygen.
The above formulre are based upon N /100 thiosulfate, and titratmg 100 mi. volllme.
N = milliliters thiosulfate used.
26

WATER ANALYSIS

2081

The solution from the above is nearly neutralized with ammonia. It is

then concentrated and 2 grams potassium oxalate and 1.5 grams p<?tassium
sulfate are added and the zinc removed by electrolyzing. (Three hours with
0.3,ampere current.) This gives zinc.
Where copper only is desired it is frequently sufficiently satisfactory to concentrate the water from 50 ml. to 75 ml., after which it is acidified with 2 to 5 ml.
concentrated H 2S0 4 , depending upon whether the water is very alkaline with
carbonate of lime, etc., and then the procedure.for copper is followedP
There is no satisfactory method for the quantitative determination of small
quantities of tin. In the above-mentioned procedure, however, in case tin
should be present it would be removed with the ammonia precipitate for the
removal of iron and its presence may be avoided by dissolving the sulfides in the
original
in HNO a, in which the tin would remain behind insoluble.
, precipitation
.

HARDNESS
Total Hardness
The most accurate method for total hardness is by calculation of the calcium
and magnesium determined gravimetrically as previously outlined, calculating
the calcium as calcium carbonate and the magnesium to its calcium carbonate
equivalent in terms of parts per million. 28 However, where only the hardness is
desired, gravimetric methods are cumbersome and the following are accepted as
standard.
The standard method for the determination of total hardness, as well as
temporary and permanent, depends upon the action of the lime and magnesia in
solution upon ,!oap, the soap ,added in a very dilute solution in alcohol. Total
hardness represents the total soap acted upon by the water in its original state,
permanent hardness represents the total soap acted upon by the water after the
water in question has been thoroughly boiled and separated from the suspended
matter, and temporary hardness represents the difference between the total
hardness and the permanent hardness, and while it is supposed to represent
combined carbonates of lime and magnesia, and the permanent hardness is
supposed to represent lime and magnesia in other forms than carbonate, this i1l
rarely so due to the fact that a certain material amount of carbonate of lime
and magnesia is soluble in water, even in the absence of carbon dioxide gas.
The hardness 29 is considered entire1y due to Ca, when the amount of soap
used is calculated. The errors due to Mg and Fe present are not .considered; as
a rule the blank on distilled water equivalent to the amount of sample is not
considered.
Preparation of Solutions: The reagents used are standard soap solution and
Phelps, J. Amer. Chern. Soc., 28, 369 (1906).
C. Bahlrnann, J. Ind. Eng. Chern., 6, 209, 11 (1914).

29 Errors in the Clark Method for Determining Hardness, George G. Town, J. Am.
Water Works Assoc., 21, 1395 (1929).
27

28

WATER ANALYSIS

2082

standard calcium chloride solution, the latter being made under such conditions
that 1 ml. of the sofution is equivalent to 0.2 milligraIl} of calcium carbonate.
Standard Calcium Chloride Solution.-0.2 gram pure calcium carbonate is
dissolved in ~ small amount of dilute I!Cl, taking pai~ to avoid any loss due
to effervescence or spattering. Evaporate the solutIOn to' dryness several
times to remove excess acid. Dissolve in distilled water1and make up to 1 liter.
Standard Soap Solution. is obtained by dissolving approximately 100 grams
dry Castile soap in 1 liter 80% alcohol. This solution should stand several days.
For standardizing this solution should be diluted with alcohol (70% to 800/""), .
until 6.4 ml. when added to 20 ml. of standard calcium solution will produce\.
permanent lather. Usually less than 100 m!. of the original'soap solution will
make 1 liter of standard solution.
For plant control, and to simplify calculations, 58.3 ml. is frequently the
quantity taken, and the soap solution standardized so that 4.5 ml. =20-ml.
standard CaCh; a correction by subtracting 0.5 from all readings is made, a~d
in the resulting figure 1 ml. will be equivalent to 1 mg. CaCOa If the standard
soap solution is used and the figures to read in grains per gallon are desired,
take 41 ml. of sample and subtract a correction of 0.7 ml. from the reading
obtained.
.
For standardizing, use 250-ml. glass-stoppered bottle, add 20 ml. calcium
solution with 30 ml. distilled water. The soap solution should be added' from
a burette, approximately .2 ml. at a time, after which the bottle is shaken
vigorously until the lather formed remaIns unbroken for five minutes after
shaking and after the bottle has been placed upon its side.
NOTE.-Pure potassium oleate may be used in place of soap.30
Operation.---,Fifty ml. of the water in question are measured into a 250-)lll.
bottle, the soap solution added, approximately .2 ml. at a time, and in the same
manner as described for the standardizing of said soap solution. The following
table, copied from p. 33, Standard Methods of Water Analysis, A. P. H. A., 1913,
may be used to obtain ,the total hardness from th\"l reiults so noted:
TABLE OF HARDNESS SHOWING THE PARTS PER MILLION OF CALCIUM
CARBONATE (CaCO,) FOR EACH TENTH OF A MILLILITER OF
SOAP SOLUTION WHEN 50 ML. OF THE SAMPLE ARE USED.
m!. of Soap
Solution.

0.0
1.0
2.0

0.0

m!.

0.1

Ill!.

0.2

m!.

0.3

m!.

0.4

m!.

0.5

ml.

0.6

m!'

0.7

ml.'"-

- -- - - -- -- -- 0.0
... . ... . ... . ... . ... .
. . ...
4.8
19:5

'"

12.7 14.3
26.0 27.3
------ -- 33.8 35.1 36.4 37.7 38.0 40.3
47.1 48.6 50.0 51.4 52.9 54.;3
61.4 62.9 64.3 65.7 67.1 68.6
- -- -- -- -- -- 75.7 77.1 78.6 80.0 81.4 82.9
90.0 91.4 92.9 94.3 95.7 97.1
6.3
20.8

7.9
22.1

9.5
23.4

11.1
24.7

0.8

0.9

-~.

ml.

0.6

l(j~9

3.2
18.2
31.2

15.6
28.6

29.9

41.6
55.7
70.0

42.9
57.1,
71A

44.3
58.6
72.9

84.3
98.6

85.7
100.0

87.1
101.5

-~

3.0
4.0
5.0

32:5
45.7
60.0

6.0
7.0

74.3
88.6

30 C. BIacher, Chern. Ztg., 36, 541 (1912); J .. Soc. C~m;:-Ind.! 31, 555 (1912); C.A.,
7, 1394 (1913); C. BIacher, P. Gruenberg, M. Kissa, Cliern. Ztg., 37, 56-8 (1913); C.A.,
7,1938 (1913). L. W. Winkler, Z. Anal. Chern., 53, 409-15 (1914); C.A., 8, 29!2 (1914).

WATER ANALYSIS

2083

It is not desirable to use more than 7 m!. of soap solution for 50 ml. of the
water, and when the figures are higher, the water should be diluted with distilled
water. The reading in the table corresponding to the ml. of soap solution

used is then multiplied by the quotient 50 mil.; x m!. being equal to the amount
xm.
of water taken. In making this determination there is frequently noted a false
end-point sometimes known as the magnesium end-point. To avoid error, it
is advisable, after completing the titration, to read the burette, add 0.5 m!.
more of the soap solution and shake well. If magnesium has been responsible
for the false end-point, after such addition the later will again disappear, and
titration should be continued until a new and true end-point is reached. 'It is
advisable to determine the strength of the soap solution from time to time, as it
is very prone to change upon standing. Results should be recorded in terms of
calcium carbonate, parts per million. There are various other means of reporting. The English degree frequently noted as Clark degree, represents grains
calcium carbonate per Imperial gallon and should be multiplied by 14.3 to give
parts per million. Conversely, the result obtained in parts per million divided
by 14.3 will give Clark, or English degrees. French degrees represent parts
per 100,000 calcium carbonate and should be multiplied by 10 to give parts per
million. Conversely, division of the result obtained above by 10 will give
French degrees. German degrees represent parts per 100,000 calcium oxide
and should be multiplied by 17.8 to give parts per million calcium carbonate.
The determination of hardness is not reliable on account of the varying action
of calcium and magnesium salts, find should never be resorted to when possible
to determine these bases direct.
NOTE.-Dr. Hale claims that the soap method for hardness in skilled, hands is
accurate from 10 to 15 parts per million on waters as hard as 300 parts.

For permanent hardness the standard soap solution is used as above stated.
The water, however, is boiled gently for one-half hour, allowed to cool, made to
volume with boiled and cooled distilled water and filtered, after which the above
method is used. The difference between total hardness and permanent hardness is supposed to represent temporary hardness. The alkalinity determination given on 11 previous page is a much more accurate method of determining
temporary hardness, however, and is also much more easily carried out. When
total hardness and alkalinity are determined, permanent hardness would be
the difference between these two figures. For comparative use as against total
and permanent hardness determined as such, the results would be much different, as the alkalinity determination of all the carbonates would give a permanent
hardness representing absolutely non-carbonate hardness; whereas the determined permanent hardness would contain a material amount of combined
carbonates of lime and magnesia. The American Public Health Ass~ciation,
Committee on Standard Method of _Water Analysis, recommend that the
determination of permanent hardness by the soap solution be discontinued in
connection with softening process, as it is so unsatisfactory in general practice.
Magnesium Chloride
Frequently, when hypothetical combinations are used it is desired to check
up these calculated combinations by some chemical method. Magnesium

2084

WATER ANALYSIS

chloride is frequently produced in the course of hypothetical combinations and

its presence is as frequently a source .of much trouble iIi the determination .of a
mineral residue, owing to the ease with which it decomposes or carbonates: A
method is suggested whereby a second 50 ml. portion sirrlilar to the total mine~
residue is exactly neutralized with sufficient R 2S0 4 , th'e amount to use bejng
calculated from the total alkalinity obtained elsewhere. The solution is allowed to go to complete dryness, is baked at a temperature of 280 0 F. to 300 0 F.,
and after being cooled the chlorine is titrated. The difference between the
chlorine thus determined and the total chlorine previously determined represents chlorine lost by volatilization as magnesium chloride. In the absence
of organic matter this method is approximately accurate. Where organic
matter or other reducing material is present, however, the results are not so
satisfactory.
Calcium 'Sulfate
In a similar manner it is frequently desired to know whether or :not a water
would contain calcium sulfate, and a method of comparative satisfaction depends upon the evaporation of 250 ml. to 500 mL ofthe original water to dryness.
After cooling, 10 m!. of distilled water are added and the mineral matter loosened
from the sides of the dish and partially dissol;Ved. Ten ml. of 95% alcohol are
then added and 100 ml. of 50% alcohol. After thorough stirring and solution
this material is filtered, the. precipitate washed with 50% alcohol and the filtrate
made to volume, divided and t.ested for calcium and sulfates in the usual
manner. The method is only approximate.

LIME AND SODA VALUE

I
ill

Two very simple methods have been devised for the rapid estimation of the
amount of lime and soda-ash necessary for softening, when water treatment is
considered from the outside softening-plant point of view.

Value for Lime

Reagents: Saturated lime water (strength to be known for e-ach series of
determinations). N/IO hydrochloric acid.
Process.-Take 200 m!. of the water in questi~n; add 50 ml. saturated limewater s~lution in 250-ml. volumetric flask and heat to boiling. AllQw to cool.
Fill to the 250-ml. mark with water to replace that:lost by evaporation; filter
through a dry-folded filter and titrate 200 ml. of the filtrate with N /10 acid,
using methyL orange as an indicator.
For calculation, let a equal number of ml. N /10 calcium oxide in 50 ml.
the lime water, as determined: and let b equal the )1Umber.' m!. N /10 hydrochloric acid used in determination.
31 Drawe, Zeit. angew. Chern., 23, 52 (1910).

WATER ANALYSIS

2085

(4a-5b) X3.51 CaO will give milligrams of lime per liter required to soften
the water tested.

Value for Soda

To the neutralized 200 ml. from above titration, add 20 ml. N /10 sodium
carbonate. Heat to boiling. Tran.sfer with CO 2 free distilled water into a 250m!. flask to make up to mark with' was}lings from the dish; mix thoroughly and
filter, collecting 200 m!. of the filtrate in a beaker. Titrate with N /10 hydrochloric acid for the excess alkali. Designate the number of ml. in this titration
by c.
Formula: (20-b-ic)X33.13 Na 2C0 3 =milligrams of soda per liter necessary to soften water in question .

NOTE.-Both formulre are based upon C. P. chemicals, and corrections must be

made for the value of the commercial materials in use. These methods are valuable,
though for actual practice it- is advisable to tryout on a liter of water in question,
u,sing the calculated amounts of lime and soda for experimental purposes.

METHODS OF REPORTING AND INTERPRETATION

The manner of reporting the results of a mineral analysis of any water calls
for as much thought and uniformity as the methods of analysis themselves, and
in this department there is much less uniformity than in the case of analytical
methods. Undoubtedly, the ideal method of reporting is that which gives
results in ionic form (positive and negative radicals), in terms of parts per mil..
lion or grains per gallon. The latter term is purely American and would have
to be converted for comparison with results obtained in almost any foreign
country. Parts per million, though newer and still unfamiliar to all but professional and scientific men, is gaining gradually a strong foothold, and the
consideration of this terminology with the ionic form of reporting will be considered prior to the discussion of hypothetical combination and grains per
gallon.
Mr. Herman Stabler,32 and R. B. Dole,33 of the United States Geological
Survey, have devised and simplified certain calculations and formulas, which
greatly assist in the interpretation, comparison, and classifica,tion of waters for
Industrial and Irrigation 34 purposes. Formulas with reference to dissolved
solids will be the only ones discussed here. The following table gives reaction
coefficients:
32 Eng. News, 60, 355 (1909).
3"Water Supply Paper No. 274, p. 165, Water Supply Paper No. 254, J. Ind. Eng.
Chern., 6, 710 (1914).
.
.... U. S. G. S. W. S. Paper, 274, p. 177.

2086

WATER ANALYSIS

Reaction

Positive Radicals.

Coefficients.

Negative Radicals.

Reaction
CoeffiCIents.

Ferrous Iron (Fe) .........

Aluminum (AI) . . . . . . . . . . .
Calcium (Ca) .............
Magnesium (Mg) .........

0.0358
.1107
.0499
.0822
.0435
.0256
.992

Carbonate (CO a)
. .....
Bicarbonate (HCO a)

0.0333
.0164
.0208
.0282
.0161

Sodium (Na) .............

Potassium (K) ........ ' " .

Hydrogen (H) .............

Sulfate (SO.) ...........

Chlorine (CI) ............
,Nitrate (NO,) ' ...........

Using the above table, the parts per million of each radical multiplied by its
reacting coefficient will give its reacting value, and in the formulas which follow
this will be indicated by T prefixed to the chemical symbol of the nidical.
For checking the accuracy of the analysis, the sum of the positive reacting
values should equal the sum of the negative reacting values, and the formula,
g. E h
iYSIS.
'
The vaIue 0 I t h'IS
10 OrPos.-rNe
N
' t e percentage error 0 I t h
e ana
r Pos.+r ego

errOr should never exceed 5 for waters of 100 p.p.m. or morC? dissolved solids, and
should be 2 or less.
.
In ordinary analysis, silica, iron and aluminum are present in such small
quantities that they may, for simplicity's sake, be ignored. The following
formulas are given without comment, as full details can be found in Water
Supply Paper No. 274.
Water Soltening
For 1000 Gals. Water.-Pounds lime (90% CaD) required
.
=0.26(rFe+rA;+rMg+rH+rHC03+.0454 CO,).
Pounds soda ash i (95% N a,CO,) required
= 0.465(r Fe+r Al +rCa+r Mg+rH -rCO, -r HCO,).
NOTE.-Dr. Hale states the following. If Instead of the extended formulre of the
ionic system, I much prefer my' simple fo.rmulae.

(AlkaiinityX.44+free CO,)X.0106=lbs. CaO per 1000 gals. feed water.

Also (Total hardness-total lime) X.0047 =lbs. CaO per 1000 gals. feed water.
or total magnesia as CaC0 3
(Total hardness-alkalinity)X.OO9=lbs. Na 2 CO a per 1000 gals. feed water.
Expressed as Ca003." I
t

Foaming and Priming

Foaming coefficient F=2.7 Na.

Taking into consideration the various boilers and the action 'of various
waters in practice, the following approximate classification of waters for foaming
conditions is of value:

Non-foaming, F=60 or less.!

Semi-Ioaming, F = 60-200. /
Foaming,
F = 200 or m9re

WATER ANALYSIS

2087

Corrosion

For Acid Waters

Coefficient of corrosion C=1.008(rH+rAI+rFe+rMg-rCO.-rHCO.).

For Alkaline Waters

C=rMg-rHCO.

If C is positive, water will corrode.

If 0+.0503 On is negative, water will not corrode on account of the mineral
materials in the water.
If C is negative, but C+.0503 Oa is positive, the water mayor may not
corrode.
Scale

SiO,+2.95 Ca+1.66 Mg=scale p.p.m"

or
35

(.00833 Sm+.00833 Cm+.3 rFe+.142 rAI)+.168 rMg+.492 rCa

=scale lbs. per 1000 gallons.

There are also formulas given in the above-mentioned Bulletin on soap cost,
lime, socia, soda ash, cost hard scale (pounds per 1000 gaUons) and a hardness
coefficient of the scale formation.
Hypothetical Combinations
,
The use of hypothetical combinations in the reporting of a mineral water is
frequently of value, in that it gives a more rapid way of placing in simpler terms'
the principal materials present in the water. It should never be assumed from
the hypothetical combinations that the materials so reported are present in the
water in that particu1&1' form, but it is assumed by most, th~it the form in which
such materials are reported will represel).t the condition in which those materials
will combine, when the water.is subjected to increased pressure and iqcreased
temperature. In other words, the hypothetical combinations most generally in
use represent the way materials will appear when combined, due to the law of
mass action under steam-boiler conditions.
! i For such purpose the method which takes care of the insoluble materials or
laterials leaying the water first is the most common method in use. This
method combines as follows: Where the sum of the sulfate and carbonate radicals exceeds that of lime and magnesia as bases, the magnesia is first calculated to
carbonate'" the remaining carbonate.is combined with lime, the remaining lime
~'ith sulfate and the remaining sulfate with sodium. This also takes care
of the general condition where the carbonates alone are in excess of the combined carbonates of lime and magnesium, in which case the remaining carbonate
naturally would be calculated to sodium and all the sulfate, as well as chloride,
also, calculated to sodium: Where, however, the sum of the bases is greater
than the sum 'Of the carbonate and thfi suJfate radicals, two possible conditions
ali Can be omitted or ignored unless suspended matter, silica, etc., are present -in
large quantities. .

2088

WATER ANALYSIS

or combinations exist. Where magnesium chloride is present in the water, the

sum of the acids calculated to the soda radical should 1>e greater than the total
mineral residue. Partially decomposed magneflium chloride :is indicated in tlJ.is
way, also the fact that magnesium chloride has a lowet molecular weigHt t1tln
sodium chloride, which condition would be indicatedl in such a comparison.
When this case exists, the sulfate is first calculated to calcium, the rem~ining
calcium to carbonate, the remaining carbonate to magnesium, the remaining
magnesium to chloride, the remaining chloride to sodium. W:here the total
mineral residue is greater than the sum of the acids and where nitrates are
present, we then have nitrate of lime, which is assumed to be present rather
than magnesium nitrate, and the calculations are as follows: All the sulfate
is calculated to lime, all the magnesium to carbonate, the remaining carbonate
to lime, the remaining lime to nitrate, the remaining nitrate to sodium, and all
the chlorides to sodium. In acid waters naturally the lime and magnesia, as
well as the iron and aluminum, are calculated to the acid present in the greatest
excess. These methods of calculation will give certain materials frequently
found in 'scale formation and materials frequently supposed to cause certain
characteristic troubles in either steam or domestic usage. It.is possible in a
purely qualitative way to judge or interpret the water on the basis of the lime
and magnesium salts, for incrustation, and of the alkali salts for other troubles
in boiler practice, also from the standpoint of irrigation, the various forms of the
alkali salts as black and white alkali, without the necessity to us~ the formulas
already considered.
Field Assay of Water
Mr. R. B. Dole has published in Water Supply Paper No. 151, of the U. S.
Geological Survey, field methods for the assay of water in which tablets of known
value are used for the determination of chlorides, carbonates, sulfates and iron.
In this type of assay, a given amount of the water is taken and tablets are added
to the water until certain definite reactions take plice, when the number of
tablets used is estimated and an approximate val1-le obtained. The error in such
work varies from 3 to 15%, but the results of the test give valuable, if not
accurate, information.

BO~ER

ROOM CONTROL METHODS

Within recent yearsit has become -recognized that many things happen in a
boiler which makes it very inadvisable to depend absolutely upon pre-boiler
control of water conditions. The actual condition of the water in the boiler
brought about by operating conditions, elements oLcontamiriation, arid change
in return water, etc., have made it essential that-testing methods for use with
water .collected from the boiler itself and tested immediately in or near the
boiler room are very desirable. Many things have been published recently

WATER ANALYSIS

2089

which are slowly clearing up some of the fog which has handicapped the understanding of what goes on in the boiler. Research work relative to solubilities
at high temperatures and pressures, to the proper working of mass action in
boiler waters have established certain fairly definite rules the observance of
which can result in a fairly general freedom from boiler water troubles. The
methods from which information as to the water's condition would be gained,
must be methods calling for no special agents such as intense heat, and fall
into the classification of volumetric or turbidimetric methods. In the power
plant the presence of excesses in the boiler water of sodium carbonate, caustic
soda, the phosphate radical, the sulfate radical, and chloride radical have necessitated c\lntrol tests which must be approximately accurate, although they are
not tests that would compare with the methods already given, which should be
considered standard. In the determination of alkalinity, both caustic and
carbonate, there has been no great difference from the standard method. The
use 9f the two indicators, phenolphthalein and methyl orange, is most general,
but the _guantities tested for the most rapid results are 58.3 ml. or a fraction
thereof, and the acid used is N /50. The results read from the burette are
directly in 'grains per gallon. Where other interfering agents, like phosphates,
are present, the first titration using phenolphthalein is frequently modified by
adding a small amount of 10% barium chloride, and then without filtering, and
in the presence of the precipitate, a titration is made to the disappearance
of the phenolphthalein red. For the determination of chlorides the regular
chlorine method is used, but. due to the fact that the sample is frequently alkaline with caustic soda, the sample must be neutralized before
carrying out the titration. 'Plis chloride test gives indirect rather than direct
information. It is used to watch boiler concentration on the assumption that
the chlorides will concentrate uniformly regardless 'of treatment for calcium
and magnesium impurities, and their increase over the amount present in the
original supply will bear some sort of relationship to other constituents in the
boiler water. As a rule additional work must be done through a competent
laboratory so that several chlorine determinations can be compared in their
relationship to total solids and complete boiler water oomposition. From this
information a factor can usually be obtained for use with results of later control
tests. As the alkalinity would be variable the mathematics would be as follows:
Total Chlorine Determination as Sodium Chloride X F + M = Total Solids,
where F represent!? the factor of relationship between chlorides and concentration, and M represents methyl orange or total alkalinity.
, There is a great deal of publicity and interest in boiler water conditioning
using meta-phosphates, and the presence of slight excesses of phosphate is
important so that rapid methods have been worked out.
There are two, one a modification of a colorimetric method, the other a
modification of the molybdate precipitation, both of which are of interest. Th~
first method is that recommended by F. G. Straub. It is a modification of a
very rapid- and delicate colorimetric method, and aside from requiring five
solutions, one of which is standard, and two of which must be carefully cared
for, is available for field or boiler room control. The method follows;

2090

WATER ANALYSIS

Solutions: 36 Molybdic Acid Solution.-Dissolve cold 125 g. of pure ammo- '

nium molybdate in about 2 liters of phosphate-free disiilled water; add 75 ml.
of concentrated H 2S0 4 slowly; make up to 272 liters (roughly). ~ slight blue
coloration d~es not hin~er the ~se of this solution.
,\
.
.
.
Hydroquznone Solutlon.-Dlssolve 50 g. of c.p. hydroqumone m about 272
liters of phosphate-free distilled water and add 3 ml. ofl concentrated H 2S0 4
Carbonate-Sulfite Solution.-To 6 liters of distilled water add 1500 grams of
com. soda ash; dissolve 225 g. of sodium sulfite in 155 ml. of water and add.
Stock Solution of Phosphate.-Pure monopotassium phosphate finely
ground is dried at J05 C. for three hours, cooled, and kept in desiccator .
0.1432 gr. is dissolved in distilled water, 5 ml. of conc. H 2 S0 4 added, and made
up to 1 liter with distilled water; 1 ml. of this solution contains one-tenth of a
milligram of P04~
Sulfuric Acid Solution.-Add 300 ml. of concentrated H 2S0 4 to about
2200 ml. of distilled water and label H 2 S0 4.
Determination of Phosphate

36

Measure 50 ml. of the filtered sample of boiler water into a 250 ml. volumetric flask, and add 10 ml. of sulfuric acid.! To the flask 5 ml. of molybdic
acid are added,followed by 5 ml. of hydro quinone solution. After five minutes,
15 ml. of the carbonate-sulfite solution are Jtdded, mixed and the flask made up
to volume.
'Measure 10 ml. of standard phosphate solution by a pipette into a 250 mI.
volumetric flask and 20 ml. into a second flask; 40 ml. of distilled water are
added to the first and 30 to the second, followed by the procedure in the
determination of ph9sphate.
Place the flask containing the sample of boiler water between the two flasks
containing the standard solutions and compare colors. If the color is lighter
than the flas1i containing the 10 ml. it contains less than 20 p.p.m. of P0 4.
If it is between the one containing the 10 ml. and the lone containing 20 ml.,
it has between 20 and 40 p.p.m. If it is darker tlian the one containing 20 ml.
it contains over 40 p.p.m.
'
Different maximum and minimum values may be established by changing
the amount of standard phosphate solution used.,
The presence of certain organic matter which gives marked color or high
iron content will interfere 'with this phosphate determination. - The color is
not permanent; consequently, comparisons must be made within_15 minutes
after the addition of sodium carbonate solution.
The second method is that recommended by the Hall Laboratories in connection with their system of boiler water conditioning. In this determination,
the temperature of the sample is unimportant; in fact time can be saved by
carrying out the determination on the hot sample.
Reagents: 37 SPecial Ammonium Molybdate Reagent. Solution A.-Place
in a beaker 50 grams of 85% molybdic acid and mix it thoroughly with 120
36 Control of Boiler Water Treatment to Prevent Embrittlement.
F. G. Straub.
Mech. Eng., p. 366-7.
_.:-:":",
37 Phosphate Control Methods of the Hall System' ot Boiler Water Conditioning.
Hall Laboratories, Pittsburgh, Pa.

WATER ANALYSIS

2091

m!. of distilled water. Add to the mixture slowly and with constant stirring
70 m!. of ammonium hydroxide (NH 4 0H) sp.gr. 0.90. Filter, and then add to
filtrate slowly and with constant stirring 30 ml. of concentrated nitric acid
(HNO a) sp.gr. 1.42.
Solution B.-Add 200 m!. of concentrated nitric acid (HNO a) sp.gr. 1.42 to
480 m!. of distilled water.
When the solutions are cold add Solution A to Solution B slowly and with
constant stirring. Add to the mixture 0.05 gram of ammonium phosphate
. (NH 4 )2HP0 4 dissolved in 10 m!. of distilled water. Shake and let stand for
at least 24 hours, when the precipitate that has formed will have settled out. /
The clear solution is now ready for use.
PotassiuTT} Nitrate.-=-Crystalline c.p. grade. It is important that this
chemical be phosphate free.

Determination of Phosphate (Buromin)

Measure out approximately 125 m!. of the boiler water to be tested and
transfer it to a 250 m!. flask. Add to this one-half teaspoonful (approximately
5 grams) of the potassium nitrate and heat up to nearly the boiling point. In
order to flocculate the suspended matter, the flask is then removed from the
hot plate and allowed to stand for a few minutes. Then filter 100 m!. of the
sample. It is important that the filtered sample be entirely free of any suspended material. If the liquid first passing through the filter is not perfectly
clear, return it to the main body of the sample and continue filtration through
the same paper. For this test a good grade of filter paper, such as Munktell's
iF, is essential. The 100 ml. of filtered sample is then'heated to active boiling
temperature and poured 'at boiling temperature into the Hagan Phosphate
Gauge. 10 ml. of the special ammonium molybdate reagent is then added
and the contel1ts thoroughly mixed by shaking with a swirling motion. Air
bubbles in the capillary stem must be eliminated by opening the stop cock and
allowing little of the liquid to ,flow through. This liquid is collected and
returned to the gauge. If phosphate is present, a bright yellow precipitate will
form, and gradually settle into the capillary stem. Allow the sample to stand
for half an hour, giving the gauge an occasional swirling motion to bring all
the yellow precipitate down into the stem. The depth of the yellow precipitate
is read off in divisions of the calibrated stem.

Sulfate
The other determination is that involving sulfate and it is important from
sljveral angles. First it is of value in the control of embrittlement, and it is
still one of the inhibiting agents depended upon most strongly and also, where
other methods are employed, it is of value in connection with the use of sodium
sulfite as an absorbent of oxygen for the elimination of gas corrosion. Two
methods are of special interest' in this case, one a turbidimetric method, the
other a colorImetric methqd, which is quite new at this time, but which has
attracted a great deal of favorable attention. In the turbidimetric method
many mechanical devices are used in the obtaining of the ultimate reading
which is based upon the intensity of turbidity necessary to just make impossible

2092

WATER ANALYSIS

the detection of light passed through a standard opening. The procedure

necessary to produce the turbidity is all that will be giv~n. "
Procedure :38 Reagents.-Standard salt-acid solution, 240 grams of sodium
chloride plus 20 ml. of hydrochloric acid (sp.gr. 1.19) per liter. This solution
must be filtered until a zero turbidity is obtained. Bat-ium chloride (BaCI 22H 20) 20- to 30-mesh crystals.
I.
Method.-If the sulfate content is below 50 parts per million a 50 ml.
sample is taken and the test made in the 20 mm. cell. With a 50 ml. sample,
10 ml. of the standard salt-acid solution are added. If the sample contains
between 50 and 100 parts per million of sulfate, a 25 ml. sample is taken ih a
10 mm. cell and 5' ml. of salt-acid solution are added. To the sample is added
0.29 gram of 20- to 30-mesh barium chloride crystals measured by a standardized
cup. The solution is stirred for exactly 1 minute, the plunger tube inserted,
and readings are taken 4 to 6 minutes after the precipitation. An average of
two readings is taken and the rep-ding obtained compared with values fi'om the
standard curve to obtain the sulfate content of the sample. An initially turbid
solution must either be filtered or a blank rea~ing made, in terms of sulfate,
before the precipitation of the barium sulfate.
The second method is known as the T. H. Q. method, and is worthy of considerable attention. The indicator and full details must be obtained from the
Betz Laboratories, Philadelphia, Pennsylvania. 39
The indicator used throughout this study was manufactured in the Betz
Laboratory and considerable quantities of this material have been supplied to
the field.
Detailed directions are given below for a determination of sulfate by direct
titration using T. H. Q: as the indicator.
Materials and Reagents.-Standard barium chloride solution, the strength
varying from 1 ml. = 1 mg. of S04 to 1 ml. = 50 mg. of S04, standardized gravimetrically. An indicator composed of disodium tetrahydrozyquinone ground
with dried potassium chloride in a 1 to 300 ratio, and passjng a lOO-mesh screen.
Ethyl alcohol or alcohol denatured by formula Nq. 30
No.3-A, or'isopropyl
alcohol and bromcresol green indicator (if phosphates are present). Sodium
chloride crystals c.p.
Procedure A.-Carefully neutralize a 25 ml. sample containing up to approximately 2000 p.p.ro. of S04 with approximately 0.02 N hydrochloric acid
until just acid to phenolphthalein. The temperature of the-sample should he
below 35 C. and it is advisable to work "between 20 and 25 C~ --Add either
25 ml. of ethyl alcohol or alcohol denatured by formula No. 30 or _N o. 3-A or
isopropyl alcohol. Introduce the T. H. Q. the amount to be determined from
Table I. It is best to use a small cup to measure -the mixture. Swirl flask to
dissolve the T. H. Q.; the solution will be colored a deep yellow. Titrate with
standard barium chloride solution, the strength to beremployed depending on
the approximate sulfate content of the sample, in accordance with Table 1.
Add the standard barium chloride at a steady dropping rate with a constant
swirling of the-flask until the yellow color changes to a .rose. The rose color

qr

38 Turbidimetric Determination of Sulfate in Water.

R. .. T. SheIn, H. L. Kahler,
and E. M. Ross, W. H. & L. D. Betz, Philadelphia, Pa. .-139 Direct Titration of Sulfates.
R. T. Sheen and H. L. Kahler, W. H. &. L. D. Betz,
Philadelphia, Pa.

2093

WATER ANALYSIS
TABLE I
PROCEDURE REQUIREMENTS

Approximate SO,
Concentration
in p.p.m.

Quantity of
THQ Indicator

Dippers
THQ Indicator

1 m!. BaCh-

gms.
(2)

(3)

(4)

(1)

*
*

0.1
0.2
0.2

Up to 100

100-1,000
1,000-2,000

Y2

1
4

1 m!. BaChill p.p.ro. SO,

in mg. SO.

(5)
,'I>

40
40
160

* Subtract 0.1 ml. as a blank in titration.

'should appear, throughout the body of the solution and not as spots of color.
This point is taken as the end point.
Procedure B (sulfate range from 2000 to 30,000 p.p.m.)."--Add solid sodium
chloride to the sample in an approximate amount as given in Table II. The
procedure is the same as in A for neutralization and titration.
TABLE II
PROCEDURE REQUIREMENTS

Approximate SO,
Quantity of
Concentration THQ Indicator
inp.p.m.
(1)

2,000- 4,000
4,000-10,000
10,000-20,000
20,000-30,000

gms.
(2)

0.4
0.4
0.6
0.8

Dippers
THQ Indicator

1 m!. BaChin mg. SO,

Iml.BaChin p.p.m. SO.

(3)

(4)

(5)

2
2
3
4

10
10
50
50

400
400
2,000
2,000

Sodinm Chloride
crystals
required gros.
(6)

2
4
8
8

Procedure C (with phosphate ion present 'up to 60 p.p.m.).-Carefully

neutralize a 25-ml. filtered sample with approximately 0.02 N hydrochloric
acid until just acid (yellow range) to bromcresol green (approximately pH 4).
Follow procedure as in A or B; no correction will be required for the phosphate
ion present.

RAPID METHODS 'FOR WATER ANALYSIS

These methods are of sufficient accuracy for field 'work study of underground waters, where rapid methods are desired. The outline suggested by
F. G. Tickell in his report to the California State Mining Bureau 40 furnishes
the basis of the methods considered.
The following determinations are of special interest:
Positive Radicals-calcium, magnesium, sodium and potassium.
Negative Radicals-chloride, sulfate, bicarbonate, carbonate, hydroxide.
40 F. G. Tickell-Summary of Operations-California Oil Fields-Tenth Annual
Report, Vol. 10, No.6, Dec., 1924.

2094

WATER ANALYSIS

Colloids-iron oxide, alumina, silica.

Loss on Ignition-organic matter, CO 2
Total Dissolved Solids. .
I
Special Reagents Required: A. Standard Sodium Chloride Solution.-Dissolve 1.648 grams of dry N aCI in a graduated liter volUIp.etric flask and dilute to
exactly 1000 ml. with distilled water.
I
l'ml. contains an equivalent of 1 mg Cl.

B. Standard Silver Nitrate -Solution.-Dissolve 4.8 grams of AgNO a

crystals in a liter volumetric flask with distilled water and dilute to 1000 ml..
Standardize as follows:
Transfer 25 ml. of the sodium chloride solution to a porcelain dish. Add
0.5 ml. of 10% solution of neutral K 2 Cr04 solution and, dilute to 100 m!. with
distilled water. From a 50 ml. burette, add with constant stirring the silver
nitrate reagent until a faint, but persistent red color (Ag 2 Cr04) is obtained.
Divide 25 by the ml. AgNO a required; the result is the mg. CI equivalent of
1 ml. of the AgNO a reagent. Record this on the bottle.
Reagents A and B are used in chlorine det~rminations.
C. Standard Acid-NjlO Solution.-Details of standardization are given in
the text. Three ml. of the desk H 2S0 4 (sp.gr. 1.84-93% solution) diluted t~
1000 ml. is approximately a N /10 solution. 0.265 g. N a 2CO a will require .
exactly 50 ml. of acid for its neutralization. (Methyl orange or alizarine indi-.
cator.) In an Erlenmeyer flask place a carefully weighed (0.2-0.25 g.) dry,
pure N a 2CO a Titrate with the acid. Calculate the N /10 alkali equivalent
volume by dividing the weight of the salt used by 0.0053 and divide this by the
ml. of the acid necessary. The result is the normality of the acid. Multiply
all the acid titrations by this factor to get the true N /10 volumes.
Reagent C is used in carbonate and alkalinity determinations.
D. Barium Chloride Reagent, Nj4 Solution.-Dissolve 30.6 grams of
BaC1 2 2H 2 0 crystals in water and dilute to 1000 ml. in/~ volumetric flask using
distilled water. One ml. of this reagent is equivalent to about 0.012 g. S04.
Standardize the reagent as follows: By means of a pipette draw out 25 ml. and.
precipitate the barium as BaS04 with dilute H 2S0 4 (see determination of
barium). Filter, wash, ignite in a weighed <)rucible and weight the BaS04.
Multiply by 0.4115 to obtain the equivalent ~SO!. This relmlt divided by 25
will give the equivalent SO'4 per ml. of the reagent. Record'mg:-per ml.
E. Standard Potassium Chromate Reagent-N/4 Solution.--":pissolve 24.5
grams of the crystallized salt in water and make to 1000 ml. in a volumetric
flask. One ml. of this reagent is equivalent to,l ml. of the BaCb reagent.
To get the exact value proceed as follows: Take 25 ml. of the BaCh reagent in
an Erlenmeyer flask; neutralize with NH 4 0H; heat tc;i boiling and add 20 ml. of
the K 2Cr04 reagent from a 50 ml. burette, continue'the titration with 0.5 ml.
portions, removing the flask after each addition, allowing to settle and observing whether the clear upper layer of solution is colored yellow. Continue
until a faint yellow color due to a drop or so excess of the r"eagent over that
required by the Ba. The ml. required divided into::_25 and. multiplied by the
S04 factor of the BaCl 2 reagent gives the S04 vallie of 1 ml. of the K 2Cr04
reagent. Record the mg. S04 per ml.

,WATER ANALYSIS

2095

NOTE.-Since suspended .BaCr04 obscures the end point, the following procedure
is recommended:-As the end point is approached, filter off and examine 5 ml. portions,
returning to the main solution the portion each time until the end point is reached .

'

F. Alkali Reagent-N/IO Solution.-Dissolve 5.3 grams of pure Na 2 CO a

and 4.8 grams of stick NaOH in water and dilute to exactly 2000 m!. with
distilled water in a two liter measuring flask. Standardize against the standard
acid solution using alizarine or methyl orange indicator. Record the exact
normality factor.
G. Indicators.
Alizarine.-O~e gram pure alizarine paste mixed with 200 m!. of distilleq
water. One drop is used in the titrations. Color is violet with alkaline solution, pale lemon with neutral or acid solutions.
Methyl orange.-(See text.)
Phenolphthal~in.-(See text.)
H. Ammonium Oxalate Solution.-Saturated solution.
I. Permanganate Reagent N/IO.-(See text.) 1 m!. =0.002 g. Ca.
'Methods of Procedure: Chloride Determination.-By means of a 100 ml.
pipette transfer that volume of water to be examined to a porcelain dish. Add a
few drops of 10% potassium chromate indicator solution and titrate with silver
nitrate reagent adding this until a faint but permanent red color (Ag2 Cr04) is
obtained. Report results in- mg. Cl per liter. (Parts per million.)
Carbonate (COs) Bicarbonate (HCO a) and Hydroxyl (OH) Determination.Place, by means of a pipette, 100 mI.. of the water in a beaker or porcelain dish;
add a few drops of phenolphthalein indicator. A pink color indicates the
presence of a carbonate. A colorless solution indicates its absence.
If a pink color is obtained titrate with N /10 acid until the pink color disappears; read the burette and record the result as ml. P.
TABLE FOR CALCULATION
ReO,

co,

OR

ml. MX6.1Xf.
ml. M - 2(ml. P)
X6.1Xf.
Nil.

Nil.
2(ml. P)X3Xf.

Nil.
Nil.
Nil.

Nil.

2 (ml. P)X3Xf.
X3Xf.
ml. MX1.7Xf.

Nil.

Nil.

Indicator and Titration

P sol. colorless
ml. P less than
Yz ml. M
ml. P= Yz ml. M
ml. P greater than
Yz ml. M
ml. P=ml. M

2(ml, P) -ml. M
X1.7Xf.
2(ml. M-ml. P)

f. is the factor for the' normality of the acid.

1 ml. N/lO acid=6.l mg. RC0 3 3 mg. C0 3 1.7 mg. OR.

To the colorless solution add a couple of drops of methyl orange indicator

and continue the titration with the N /10 acid until the yellow color changes to
orange. Read the burette and record the ml. as ml. M. (M here is the toMl ml.
of acid including P.)
NOTEs.-In the table the inl. figures represent mg. of the equivalent RCOa, COa or
OR as the case may be. The ml. are converted to N/IO value by multiplying by the
factor f. and then by the figure in question. 6.1 for mg. RCO s, 3 for mg. C03, and 1.7
for mg. OR.

2096

WATER ANALYSIS

In general the CO, and OH are not found in waters. that contain appreciable amounts
of the alkaline earths eCa and Mg). The ratio of carbonate I to bicarbonate in water
containing very small amounts of Ca or Mg is influenced by the amount of dissolved
CO 2 in the solution. A high CO 2 content indicates that HCq, predominates. If this
is low the normal carbonate is present. A fresh sample may ~how HC0 3 and no CO"
but the loss of CO 2 will occur and C0 3 will be found with HC0 3

Sulfate Determination.-The volumetric procedur~ depends upon the

insoluble BaS04 forming and settling rapidly when BaCb solution is added to
the sulfate. The reagent is .added in excess and the excess determined by
titration with K 2Cr04 with formation of BaCr04. A slight excess of the
chromate'reagent becomes evident by coloring the solution yellow, best seen when.
the yellow BaCr04 has settled slightly or by filtering from BaCr04. See Note
under E.
.'
MSO , +BaCI 2 (xs) = MCI2+Ba804+(xs BaCI 2)
.BaC1 2 +K 2Cr04 (xs) =BaCr04+(xs K 2 Cr04 yellow solution).

Procedure.-Pipette 100 ml. of the water into a beaker and make faintly
acid with HCI (litmus paper) and heat to boiling. While boiling add from a
burette the standard BaCl 2 reagent ilf small portions at a time allowing to
settle with each addition until all the sulfate has been precipitated and the
reagent is. in slight excess. Record the ml. used.
Cautiously neutralize the solution with NH 4 0H (carbonate free), heat to
boiling and titrate the excess of the BaCI; with K 2Cr04 standard reagent according to the method under E of Special Reagents. See Note under E.

The 80 4 value of BaCh minus the S04 value of K 2Cr04 titrations (ml.
BaChX804 value minus ml. K 2Cr04XS04 value) is the 80 4 content of the
100 ml. sample of water. Report mg. per liter. (Parts per million.)
Optional Procedure.-Addify the solution with HN0 3 and heat to boiling.
Cool and neutralize with NaOH solution using as an indicator just sufficient
amount of litmus solution to detect end point avoiding an excess. Add a 10%
solution of AgNO a dropwise until all the chloride is nrecipitated and then a
few drops in excess; now add a slight excess of standard barium nitrate (standardized as was the BaCl 2 reagent-mol. wt. 261.4 Ba(NOa).). Titrate this'
excess with standard potassium chromate, A red color (Ag 2 Cr04) will indicate
the approximate end point. Dissolve the Ag 2Cr04 by adding NH 4 CI, heat to
boiling, and continue the titration with K 2 Cr04 'until the faint yellow color is
obtained as in the first method. Subtract the S04 value orthe chromate from
the 80 4 value of the barium nitrate to obtain the S04 of the water. Report
mg. per liter. (Parts per million.)
Alkaline Earths .
Calcium and Magnesium Determination.-Calcium and magnesium here
are determined jointly by precipitating with the mixed reagent N a 2CO a and
NaOH (the latter necessary.for the complete precipitation' of Mg). The
excess of the alkali is determined by titration with a standard acid. The two
I
are reported as calcium.
Procedure.-Place 100 ml. of the water in an Erl~meyer flask and add 25
ml. of the standard alkali reagent (F); boil, filtel',-'and wash. To the filtrate
containing the excess of the alkali reagent add a drop of the alizarine indicator

WATER ANALYSIS

2097

and titrate (best conducted in an Erlenmeyer flask) the alkali, adding a few
drops of the acid in excess. Boil to expel the CO 2 and determine the excess of
the acid by back titration with the standard alkali to the yellow neutral point,
best obtained by adding the alkali to !l faint pink and back to yellow with a
drop of standard acid.
Calculation.-If A is the total alkali reagent added, i.e., the 25 ml. and the
back titration for acid neutralization; and B is the acid titration (both being
converted to their exact equivalents-ml. in terms of 0.1 N), then A minus B is
the amount of alkali reagent required in the precipitation of Ca and Mg.
One ml. of 0.1 N (i.e., N /10) alkali reagent is equivalent to 0.002 g. Ca. '
Express the results in mg. per liter.
Magnesium Determination.-Concentrate 100 m!. of the water to about a
fourth of its volume. Acidify with HCl and expel the CO 2 by boiling. Cool,
add a few drops of phenolphthalein and neutralize the free acid with NaOH
solution. Now from a burette add a measured amount of standard NaOH
reagent (0.1 N) in amount sufficient to precipitate all the Mg present and a
slight excess. Estimate the total volume present and add twice its volume of
95% alcohol, Mg(OH)2 precipitates. Transfer to a 250 ml. vo~umetric flask,
make up to 250 ml. with 1 : 1 alcohol. Filter off 70-80 m1. and take 50 ml. of
this. Titrate the excess of NaOH with standard acid and multiply by 5.
Calculate the 0.1 N NaOH required by Mg. 0.1 N NaOH will precipitate
per m1.0.00122 g. Mg.
NOTE.-In the determination of magnesium, elements forming insoluble hydroxides,
under the conditions for magnesium hydroxide precipitation, ilhould be removed previous
to the determination of magnesium. The common interferences are iron, aluminum
and calcium.

Calcium Determination.-Concentrate 100 ml. of the water to about half

its volume if the water is comparatively soft or determine directly if hard.
Details of the oxalate-permanganate method are given in the text.
Alkali Determination.-These are calculated by difference, it being assumed
that the amount, of other radicals in the solution is negligible.
Consult table following. Multiply each radical value by its proper reaction
~oefficient and b.alance the
and - reacting values; the difference will be the
reacting value of sodium in the solution. Divide this" value by the reaction
coefficient for sodium to obtain the mg. of sodium. C!l.lculate the mg. per liter.

REACTION COEFFICIENTS
Reaction Coefficients

Positive
r

Na
Ca
Mg

0,0435
0.0499
0.0822

Negative

SO.
Cl
CO.
HCO.
OH

0.m::08
0.0282
0,0333
0,0164
0,0588

2098

WATER ANALYSIS
IRRIGATING WATERS

When a water supply is to be used for the purpose of iriigation, it is necessary

to bear in mind the fact that there are certain impuriti~s which, if present in
large enough quantities, or due to imperfect drainage, are allowed to accumulate
in the soil, will have a most detrimental action on growing crops. Due to
solubility, these are usually soda salts. They are objectionable unevenly.
Sodium carbonate seems the most dangerous and sodium sulfate the least. Not
a great deal has been published on the subject, though there is one goyernment
bulletin 41 and one article by Stabler.42 Growing out of the latter the folfowing
tables which are of value are offered.
Irrigating Waters

Alkali Coefficient
(a) When Na-.65 01 is zero or negative,

Alkali coefficient., k =

2~~O.

(b) When Na-.65 01 is positive, bu~ n9t greater than .48 S04

Alkali coefficient, k = N a ~:.~ Cl

(0) When Na-.65 01-.48 S04 is positive,

Alkali coefficient, k = N a _ .32 ~12_.43 80 4

Classification on basis of alkali coefficient:
Alkali Coefficient

Class

/
R,emarks

More than 18. .. Good .. . Have been used successfully fOf many years without special
\
care to prevent alkali accumulation.
18 to 6 ........ , Fair .. .. Special care to prevent gradual alk~ accumulation has
generally been found necessary excepton-lQose soils with
free drainage.
. .
5.9 to 1.2 ....... Poor .. .. Care in selection of soils has been found to be imperative and
artificial drainage has frequently been found necessary.
Less than 1.2.... Bad ... . Practically valueless for irrigation.

It must be borne in mind that the dangerous char~cteristics of a water have

a close relationship to the drainage, and with perfect drainage wate.rs that would
ordinarily be condemned, can sometimes be used without serious results.
41 Sonie Mutual Relations between Alkali Soils and Vegetation, Thomas H.Kearney
and Frank K. Cameron, U. S. Department of Agr., ReportrNo. 71. .
42 Stabler, Herman, "The industrial application of wlrl~ analyses."
Water-Sppply
Paper U. S. Geol. Survey No. 274, p. 179.
.

THE MICROSCOPICAL EXAMINATION OF

WATER SuPPLY
IMPORTANCE OF THE DETERMINATION
From the standpoint of palatability and of the esthetic character of water
supply there iEtno more.important examination than the microscopical analysis.
This examination discloses and measures the minute animal and plant life
that is present in all surface waters and in some well waters.
Large amounts cause an unsightly turbidity and even relatively small
quantities frequently cause complaint because of a scum produced when bath
tubs are filled with hot water or of a stain left upon the sides of,the white
porcelain. The water of swimming pools, if unnltered, may be unsightly.
Industrial enterprises may be affected, for example the staining of clothes in
laundries and interference with the manufacture of correct colors by dye
manufacturers and with the dyeing of goods by the dyers. Photography may
also be influenced.

The presence of certain types of microscopic organisms frequently serves

to identify the source of a water, or indicate from what particular reservOir a
supply may be drawn. In the case of New York City which serves to certain
districts of Manhattan Borough water from the Croton supply, to other sections water from the Catskill supply and mixtures to others, the supply in any
one section may be identified by the organisms as Catskill, Croton or a mixture.
Again samples from cellars, excavations and conduits are frequently analyzed
to determine whether the source is city water, sewage, or ground water. If
microscopic organisms are present city water or sewage must be responsible
and the chemical analysis and odor usually eliminate or indicate sewage. The
contamination of a well supply by surface waters may be indicated by, the
presence of microscopic organisms.
The amounts of microscopic organisms bear a direct relationship to the
various forms of nitrogen determined in the chemical analysis and frequently
explain changes in these determinations.1 As these growths increase free
ammonia and nitrate decrease and albuminoid ammonia increases and as the
growths decrease the reverse relationship holds.
The operation of filtel" plants, particularly the mechanical type, is much
interfered with by heavy growths of microscopic organisms, the network of the
latter interfering with the gelatinous formation of aluminum hydroxide in
mechanical filters and the bacterial jelly of slow sand filters, although certain
types of the diatoms may assist in the latter case.
I

Chap~er

contributed by F. E. Hale, Ph,D.

1 "The Significance of Nitrogen in its Various Forms in Water Supply."

F. E.
Hale, Proc. Am. W. W. Assoc., 1908, pages 323-327.

2099

2100

MICRQSCOPICAL EXAMINATION pF WATER

By far the most. important reason for determining 'microscopic organisms

is their connection with disagreeable, sometimes vile, tastes and odors iri water
supply. Those so-called littoral growths, which are attached to the banks or
bottoms of reservoirs, and which attract the quickest, ~ttention are not concerned as a rule. The trouble is caused by minute floating forms, which manufacture essential oils or perfumes like those of flowers. I Exceedingly minute
amounts produce pleasant aromatic .g~ranium or grassy odors w:hich becom'Tl
fishy, oily, pungent or vile in larger amounts or upon decay of tb.e plant growths.
Particular genera may frequently be identified by the odor by those who are
trained.
.
4
Three groups of odors are distinguished:
Aromatic (geranium) caused by Diatomaceae.
Grassy caused by Cyanophyceae.
Fishy caused by Chlorophyceae and a few Protozoa.
The following tablE.J as given by Whipple 2 in "The Microscopy of Drinking
Water" contains those orgl\nisms which have been at one time or another in
sufficient quantity in a water supply to produce Icharacteristic odors.
Group
Organism
Aromatic odor ..... Diatomaceae:

Natural Odor

Asterionella . ....... ' .. Aromatic, geranium, fishy.

Cyclotella . ........... Faintly aromatic.
Diatoma ... .......... Faintly aromatic.
M eridion . ........... Aromatic.
Tabellaria . .......... Aromatic, geranium, fishy.

Protozoa:

Cryptomonas ......... Candied violets.

Mallomonas ... ....... Aromatic, violets, fishy.

Grassy odor ....... Chlorophyceae:

Dictyosphaerium . ..... Grassy.

Cyanophyceae:

Anabaena .... ........ Moldy, grassy, nasturtium, green-ccrn.

Aphanizomenon, . ..... Grassy.
/
.

Coelosphaerium .. ..... Grassy, s'Yeet. I

Clathrocystis . ........ Grassy, sweet.
Cylindrospermurn . .... Grassy. .
Rivularia .. .......... Moldy, grassy.

Fishy odor ........ Diatomaceae:

Asterionella . ......... Fishy in large numbers.

Tabellaria . .......... Fishy in large numbeJ:.S.

Chlorophyceae:

Dictyosphaerium .. .... Faintly fishy, also grassy.

Eudorina . ........... Faintly fishy.
Pandorina . .......... Faintly fishy.
Volvox . ............. Fishy.
,

Protozoa:

Bur8ari~ . ........... Irish moss, salt marsh, fishy.

Ceratium . ........... Fishy.
"
Dinobryon . .......... Fishy, like rock-weed.
Glenodinium! ........ Fishy.
Mallomonas .. ........ Fishy, also aromatic.
Peridinium . ......... Fishy, like clam-shells.

Synura . ............. Cucumber, fishy, muskmellon, bitter taste.

Uro(Jlena . ........... Fishy and oily_;_:: .-'

With some additions by the author.

MICROSCOPICAL EXAMINATION IOF WATER

2101

In New York City's supplies, despite the diversity "of its sources, "the only
organisms which have given offense from odors have been 4sterionella, Tabellaria, Volvox, Dictyosp"haerium, Anabaena, Aphanizomenon (with admixtures of
Coelosphaerium, Clathrocystis, and Microcystis), Ceratium, Dinobryon, Perid_inium, Synura, Uroglena, and Crenothrix.
.
Asterionella; when present in 500 to 1000 standard units per cubic centimeter, produces a slightly aromatic odor. At 1000 units, rarely less, the odor
is distinctly similar to the odor of the geranium. The odor increases in intensity with increase in numbers until several thousand produce a fishy odor.
The fishy odor is also produced when smaller quantities die.
Tabellaria, and similarly Asterionella, in very small amounts produces an
earthy odor (also produced by large amounts of Synedra) , passing through
the aromatic, geranium and fishy stages with about the same relative quantities of organisms as Asterionella. At times the odor of Tabellaria has suggested illuminating gas, no other organisms being' present. In the spring of
1919 extensive complaints of fishy taste in the Catskill supply were occasioned
in New York City by only 700 units of T cibellaria.. This was the result of
chlorination with liquid chlorine at Kensico reservoir. The odor was not
noticeable in the water above the chlorination plant but appeared first just
below the plant. Chlorination killed the' organism, setting free the odorous
principle.
Anabaena and Aphanizomenon, when present in 500 to 1000 units, produce a
faintly grassy odor like freshly-cut grass. With larger numbers the odor
becomes pungent like nasturtium, or even onions. In large numbers, or
when decaying, the odor is of vile, pigpen character. In the spring of 1918
from 500 to 800 units of Aphanizomenon in the Croton supply of New York City
caused numerous complaints of an oily brown scum when bath tubs were
filled with hot water, These complaints came chiefly from high class apartment
houses.
Uroglena produces an oily fishy taste and odor, first noticeable in probably
500 to '1000 units. In larger quantities it is very disagreeable. The flavor is
that Of cod-liver oil. Chlorine has an influence on this organism also.
Synura has caused trouble in as small amount as 50 units. The odor is
vario-qsly described as like cucumber, muskmellon, etc. It leaves a bitter aftertaste. Chlorine intensifies the trouble so that only a few units are noticeable.
Dictyosphaerium, about 700 units, under influence of chlorine has produced
a grassy or pungent nasturtium.odor.
Troublesome organisms occur chiefly in surface waters: Occasionally well
waters containing iron or manganese cause trouble from growths of Crenothrix
and associated forms which clog pipes and cause an unsightly turbid discolored
water. Well strainers become clogged so as to prevent pr()per yield of water.
Microscopic organisms apparently do not affect the health. Possibly the'
taste and odor at times produce nausea or distaste for food. It would take
12,000 units of Asterionella per cubic centimeter to add a milligram of solid
matter to a glass of water.

2102

OF

MICROSCOPICAL ,EXAMINATION

WATER

:;AMPLING FOR ANALYSIS

In sampling water {or microscopical analysis precdutions should be t&ken ,
that the sample be reI,resentative and fresh. Water\ from Ii -tap should be
allowed to run to avoid heated stagnant water. Samples from a pond or
reservoir should not coJ).tain surface scum nor littoral growths. Deep samples
are best taken by the method usually employed for
dissolved oxygen samples. A quart bottle is fastened
by clamps to the side of a two-quart bottle in a water
sampler .. The bottles are fitted with two-hole rubber
stoppers. Glass tubing is arranged as follows: A
straight tube, projecting slightly above the stopper,
extends to the bottom of the quart bottle. A bent
tube is made flush with the underside of the same
stopper and connected by a short piece of stiff rubber
tubing to a bent glass tube extending through the
stopper of the two-quart bottle to its bottom. - A
straight tube, flush with the underside of the stopper
of the two-quart bottle, extends above the stopper
from eight to ten inches in order to make a difference
of head between the inlet and outlet tubes. The
bottles are lowered quickly to the required depth l;>y
a marked and measured rope. Watch is 'kept for
bubbles of air which rise to the surface of the water,
sometimes several feet away, until bubbling ceases,
which takes two to four minutes. Both bottles
must be completely filled before raising, otherwise the
FIG. 291.-Sampling sample may be from some other depth than that
Apparatus for Deep desired, The water enters the small bottle first and,
%amp\\)B.
as it only takes one half minute to lower the apparatus
to the desired depth and at least two minutes for both bottles to fill, the smaller
bottle is filled with the last water to enter whichiis from the desired depth. If
the large bottle were Jlot completely filled, the 'bottles would be filling while
being drawn up and the water of the small bottle lVould represent some unknown.
depth. (See Fig. 291.)

SEDGWICK-RAFTER METHOD OF EXAMINATION

Apparatus Required

A microscope (not necessarily high priced).

I
Eyepieces-One inch (25 mm.) and one half inch (!g,__-mm.).'
Objective-Two thirds inch (16 mm.) and one sixth-inch (4 mm.).
Ocular micrometer-Fitted to one-inch eyepiece.

MICROSCOPICAL EXAMINATION OF WATER

2103

This is ruled with a square of such size that used with a one-inch eyepiece
and two-thirds inch objective the tube length of microscope may be adjusted
so that the area viewed on the stage will be one square millimeter. The
tube length is adjusted by making the square in the eyepiece coincide with a
stage micrometer. The large square is usually divided into quarters, one
quarter into 25 small squares and one of the latter into 25 tiny squares;
each of the last represents an areal unit of measurement,.400 square microns
(20 microns to the side). A micron equals .001 mm.
A stage micrometer-graduated usually to tep.ths of a millimeter.
Sedgwick-Rafter Counting Cell-a thick slide to which is cemented a brass rim~
with internal dimensions of length 50 mm., width 20 mm., and depth 1 mm.
It has an area of 1,000 'square millimeters' and a capacity of 1 cubic centimeter. Its depth only must be exact. Several should be provided.
'Thin cover glasses, 55 mm. by 25 mm., to use as covers for the cells.
One cubic centimeter pipettes (regular bacteria pipettes will do).
Five .cubic centimeter pipettes, volumetric.
A small glass jar of distilled water-for rinsing cells and coverglasses.
Lintless cloth or handkerchiefs-for drying cells and coverglasses.
One-inch test tubes of heavy glass (bacteria dilution tubes) .
.Five-eighths-inch test tubes of heavy glass (bacteria media tubes).
Sedgwick-Rafter funnels. These are cylindrical, 2-inch diameter, sloping
toward the bottom to a neck of i-inch diameter. The distance from top
to slope is 9 inches, length of slope 3 inches, length to bottom 2i inches.
A small glass jar of distilled water-to furnish water for concentrates.
Perforated rubber stoppers-to fit one half inch tube of funnel.
Silk bolting cloth No. 15x.
Steel punch-to make i" diameter discs of bolting cloth for rubber stoppers.
White sand, 60 to 120 mesh, i.e., passing a sieve of 60 mesh to the inch and
retained on 120 mesh (Berkshire sand or ground quartz).
Support for battery of funnels.
A small wooden mustard spoon-to measure sand.
Volumetric flasks, 200 m!. capacity.
Test-tube racks.
Record books designed for microscopic examinations.
For field work, which is sometimes of importance due to destruction of
organisms in transit, a folding microscope and a metal sling filter are useful
but not necessary.
Procedure.-The actual procedure of examination is simple. Familiarity
with the various genera is essential. In the limited space of this chapter it is
impossible to picture the thousands of forms that exist. The best working
book, containing also plates, is "The Microscopy of Drinking Water" by
George C. Whipple. (Revised by Fair and Whipple.) References to other
books showing plates will be found at the end of the chapter.
Moisten the small end of rubber stopper with water so as easily to pick up a
disc of silk bolting cloth from the table and insert tightly into the funnel.
Fill the mustard spoon with prepared sand and drop into funnel making onehalf inch layer. Pour in 200 m!. of the water to be examined from a flask and

2104

MICROSCOPICAL EXAMINATION OF WATER

allow to filter, thus concentrating the organisms uponj the sand. D6 not
allow the sand to stand until drained dry. Slight suction may be carefully
applied to clogged filters. Remove funnel, hold in a slanting position and
carefully remove stopper and insert the end of funnel into ;a one-inch test tube.
Wash sand and contents into the test tube by 5 ml. of dis~illed water, running
the latter quickly around the sides of the funnel. Shake water and sand,
settle a few moments and decant liquid into the smaller (i inch) test tube.
A rack of these concentrates may be prepared at one time for examination.
Rinse three or four Sedgwick-Rafter cells and coverglasses in distilled water
and carefully dry with lintless cloth. Place cover glasses slantingly across
cell so as to leave air space at diagonally opposite ends. Run 1 ml. of concentrate into one corner of cell, air leaving at the oppol'ite corner, and slide coverglass into position to close cell. Allow cells to stand at least five minutes before
examining so that organisms will have an opportunity to settle completely,
as most of them will. This step is very important. While examination is
being made of one cell, others will be settling. Thus' at least three should be
kept going in rotation when many samples are to be examined.
.
Examine under the microscope with one-inch ocular containing micrometer
and j-inch objective and proper tube length as previously determined. Estimate the number of standard units of each organism present in an entire field
within the large square of the eyepiece ane! note on record sheet. Area only
is taken into account. Forms like Asterionella, Diatoma, Synedra, Tabellaria
may be counted and multiplied by a factor to get the number of units-thus
each arm of Asterionella may be estimated as equal to one half unit. Filamentous forms like Melosira, Anabaena, Aphanizomenon, Oscillaria' may be
counted in lengths of five units (one-t.enth of side of large square) and the tot\tl
length divided by the number of filaments laid side by side that it would take
to measure five units. The quotient multiplied by 25 gives the number of units
in th_e field. Synedra in "fine-tooth-comb form," Fragillaria, Uroglena,
Synura and bulky forms may be quickly compared wi~h a 25-unit square and
the number of units in the field estimated. The/above method of estimating
units is quicker than it reads and tends to eliminate largely the personal element
from the count, After the bottom of the slide has been counted, focus should
be changed to the top layer and Cyanophyceae, which float, estimated, that is
Anabaena, Aphanizomenon" Clathrocystis, Microcystis, Coelos.phaerium, etc.
Standard Methods of the American Public Health Association and Am.
Water Works Association require 10 fields of the 1000 to be ex~mined. In
ordinary practice, however, a careful count of five representative fields is
sufficient, choosing fields through the middle of the cell at each end, the center
and half way between. After counting, a search of the cell, particularly the
edges, is made for missed forms, especially Crustacei which swim to the edges
to get air and Protozoa like Uroglena. The latter are frequently rolling around
in the body of the cell. .Anabaena, Uroglena and Dinobryon commonly break
up and disintegrate in the absence of air and must be recognized in such condition. The process of disintegration frequently may be witnessed under the
microscope. Sometimes disintegration occurs in t.r,ansit. . '
On the record sheet, the amounts of units of each-organism are summed up
for the number of fields examined and multiplied by a factor to find the total

MICROSCOPICAL EXAMINATION OF WATER

2105

units as referred to the original water. The latter amounts are sUII).med to
find the total units of organisms per cubic centimeter.
If the quantities recommended in this chapter are used the factor is fQund
as follows:
The ratio of total cell

t~ counted fields is 1O~0 = 200.

. . 200 cc.
0
Th e concentratIOn IS -5-- = 4 .
cc.
The final factor is then

2~~ = 5.

FIG. 292.-Microscope for Microscopical Examination.

FIG. 293.-Ocular Micrometer.

Smallest square is one unit.

Large crustacea found only around the edges and not distributed through
the cell should have their total units divided by the concentration rather than
multiplied 'by the factor in referring the count to the original water.
In addition to the count an examination of the original sample in the bottle
by naked eye is often useful in detecting such forms as Uroglena, Cyclops, etc.,
and. in forming a general estimate of the amounts of organisms present.
Many errors may enter into the quantitative estimation of microscopic
organisms. The sample bottle should be inverted to distribute the forms, as
some float and others settle. Organisms may adhere to the sides of the funnel
or to the sand. Others may pass through the sand and some may disintegrate.
One of the greatest errors is caused by uneven distribution through the cell
when' organisms are not numerous. Whipple states that the total error does
not usually exceed 10%. The author of this chapter believes this figure is too
conservative and unavoidably so owing to the large personal element. However results are usually comparative at anyone laboratory.

2106

MICROSCOPICAL EXAMINATION OF WATER

Tabular Outline Identification of Forms
I

Sample . ............... "

Examined by. . . . . . . . . . . ..

Date of Collection. . . . . . . ..
Date of Examination. . . . . .

No. of Square

N\.tmber ................ .
Cqncentration ....... .... .

.
- 7

S 9

10

Total

Standard
Units'
Per C. C .

DIATOMACEAE:

A sterionella. .
1
Cyclotella . .................
Diatoma ................. '.
Fragilaria . .... ........... .
Melosira .... . , ., ., ......... 3 5
Navicula ... .. , ...... ' .' " "
Stephanodiscus . ........ ' .... 1 1
Synedra .. .................
20
Tabellaria . .................

20

4 10

24

120

1
5

5
25
4

25
125
20

List of Organisms
PROTOZOA:

CHLOROPRYCE.uJ :

Cemtium
Dinobryon ..
Glenodinium /'
Mallomonas
Peridinium
Synum
Trachelomonas
Uroglena
Vorticella

Closterium
Conferva
Eudorina
Pandorina
Pediastrum
Protococcus
ScenedelW!U8
Spirogyra
Staurastrum

Anguillula
AC4rina
TOTAL ORGANISMS:

Amorphous Matter
Miscellaneous Bodies
Sponge Spicules

ROTlFE.RA:

Cy ANOPHYCEAE:

Anuraea
Polyarthra
Synchaeta

Anabaena
A phanizomenon
Clathrocystis
'Coelosphaerium
Microcystis
Oscillaria

C.RUSTACE,A.:

Bosmina
Cyclops
Daphnia

SCHIZOMYCETES, ETC.:

Crenothrix
Mold l! yphae

OTHE.R ORGANISMS:

I,

Amorphous matter comprises organic debris due largely to broken~down

microscopic organisms and zooglea. 3 It is usually of little importance butjn
connection with consumption of dissolved oxygen ip reservoirs just below the
thermocline and at the bottom it has been shown to have importance. 4
3 The estimation should not usually include 'mineral matter, silt, precipitated iron,
etc., as indicated in "Standard Methods of the American Public Health Association,"
,
since this can serve no apparently good purpose.

4 "Thermocline Studies at Kensico Reservoir," Hale and Dowd, Jour. Ind. Eng.
Chem., Vol. 9, page 81.
-,.--

MICROSCOPICAL EXAMINATION OF WATER

1. Dinobryon, magnification 250.

2. Clathrocystis, magnificatl~n 250.

3. Aphanizomenon, magnification 250.

4. Anabaerta, magnification 250.

2107

5. Tabellaria, magnification 250.

6. Volvox! magnification lOO.
2, 3, 4. grassy to pigpen odor.
5. Aromatic geranium to fishy odor.
1, 6. Fishy odor.
D Entire figure is It single colony and dark spots are young volvoces in the colony.

2108

MICROSCOPICAL EXAMINATION OF ,WATER

Asterionella before (A) and after (B) treatment with copper sulfate, 1/5 p.p.m.
Note effect upon coloring matter. Magnification 150. Qauses geranium to fishy odor.

Fragilaria, magnificatlo""ll150.

Water concentrate as it appears under the

microscope. 1. Synedra, fine tooth comb.
2. Synedra, needle-form. 3. Asterionella.
4. Melosira.

Ulothrix, magnification 75.

/'

MICROSCOPICAL EXAMINATION OF WATER

2109

Identification of Forms.-On page 2106 is shown a section from a page of the

record book used in New York City's Water Department Laboratories, showing
the classification of organisms and the commoner forms found.
As the magnification used for counting is only 87 one must learn to identify
the forms quickly at that magnification. Occasionally it is necessary to use
the one-hali-inch ocular to make certain the identification, the latter magnification being 146. Young forms frequently differ from mature forms and
confusion arises. It may be helpful to mention certain distinguishing features
for some of these commoner forms, avoiding so far as possible technkal terms.
Asterionella (little star) in its usual form is readily recognized. Four to
eight arms connected at one end, slightly larger than the other end, and radiating from the attached ends suggest its name. The arms are usually more
slender near the center in girdle view. The coloring matter appears as regularly spaced dots or dashes strung along the arm in single row. There is
another form of Asterionella in which the arms attach at either end in zig-zag
manner with occasionally four arms arranged in star fashion. This form may
be confused with " Diatoma" or "Tabellaria," but may be distinguished by
the appearance of individual arms in valve view, Diatoma having cross striations
and Tabellaria an oval outlined in a swollen center.
It may be mentioned that Diatoms have two valves fitting together like
clams, oysters or scallops but different in that they overlap like the cover and
bottom of a petri dish or a pill box. The valve view is that of the top of the
box and girdle view that of the edge. (Different authors use different terms.
The above is according to Whipple.) .
Diatoma in girdle view has straight sides and square ends, joining at the
corners to form zig-zag chains. The valve view is distinctive from Asterionella
and from Tabellaria;'it has cross striations.
Tabellaria usually has wider arms than Asterionella or Diatoma and two
well-defined parallel marks or lines extending from each end nearly to the
middle. In some forms these are missing. The arms join in zig-zag fashion
with sometimes three or four in star shape. As mentioned before the valve
view is distinctive, being swollen and marked with an oval at the center. There
is also a distinctive short chubby form.
Synedra in common form looks like a fine-tooth comb, the needle-like
individuals being attached ,side by side at the center forming a serrated edge.
The edge distinguishes it from Fragilaria which also forms masses side by side
but the ends of the latter are square with no appearance of teeth. Frequently
Synedra exists only in individual short needles and again in very long needles.
M eridion also joins side by side but being wider at one end than the other
forms disc-like patterns. The valve view is also distinctive, like a base-ball
club and striated cross-wise.
Gomphonema is also wedge-shaped like Meridion but does not join in masses
and in valve view is shaped like a rolling pin. It has a pedicle which is frequently missing.
Many forms, once seen, are very readily distinguished, Stauroneis like an
elongated diamond with a cross marked from corner to corner, Pleurosigma,
with a double curve reminding one slightly of the letter" S," Navicula, shaped
like a little boat, Melosira, with its cylindrical cells joined end to end to form
filaments containing considerable coloring matter. Cyclotella appears usually

2110.

MICROSCOPICAL EXAMINATION QF WATER

as small perfect circles without contents whereas Stephanodiscus is a larger

circle with coloring matter conspicuous and sometimesi concentric markings
visible on the circumference. On edge a view as of a pill box is often obtainable.
The coloring matter of the Diatoms may appear from yellow to brown but
often a mere black and white effect without much suggestion of color.
.
The Chlorophyceae (green growth) have green coloring matter. The
common forms are readily distinguished and remembered from illustrations.~
Dictyosphaerium and Dimorphococcus are similar small coccus-like forms joiDd
by spider-like threads together, the former from a concave side, the latter
from a convex side. Pandorina and Eudarina are coccus-like forms set in a
jelly ball, the former with the coccus bodies grouped together closely at the
center, the latter with them separated and spaced regularly near the surface.'
Volvox is a large revolving jelly-like hollow ball with the surface dotted with
coccus forms and similar to Uroglena of the Protozoa, but differs in containing
several conspicuous larger green cells which are in reality young Volvoces.
The Cyanophyceae (blue growth) contain a blue-green coloring matter.
Anabaena appears like a string of beads. It differs from Nostoc in not having a
gelatinous sheath, from Sphaerozyga in that its spores and heterocysts are not
adjacent whereas in the latter a heterocyst is between two spores. The spores
are larger than the vegetative cells. It differs from Cylindrospermum in that
the latter has its heterocysts terminal with large spore adjacent, and the latter.
filament is sometimes tapering.
/'
'Aphanizomenon under high power has a beaded appearance and contains
a very long oval spore which is scarce. Under the ordinary power Aphanizomenon appears like a pencil mark on drawing paper and frequently appears
in attached masses of filaments.
Oscillaria is a pale blueish filament of even color without granulation, with
rounded ends, straight sides (i.e., no beaded appearance) and cross striations.
At times it may 00 seen to oscillate or wave back and forth (hence its name).
Of the Schizomycetes, Crenothrix is fairly common and occurs in well waters
containing iron or manganese. It occurs as dark brQwn masses due to discoloration with iron or manganese. The latter may be dissolved by acid when
the oblong cells become visible embedded in a gelatinous sheath, but not
touching each other, and forming a filament. Anthophysa pedicles, a Protozoon, are sometimes mistaken for it but the latter is usually branched like a
"Y" and has longitudinal striations. The coloiJ.ies rese!llble Synura, but are
"
._
colorless and generally smaller.
The Protozoa are the lowest forms of animal life, unicellular in s'trueture,
though they may aggregate in colonies. Of Dinobryon frequently only the
pencil-like outlines of the tiny cups or trumpet-like shells are visible (and
difficultly seen), one set in another, ;to form branching aggregations. The
spores of Dinobryon are sometimes mistaken for Cyrlotella. Glenodinium and
Peridinium are oval and each has a groove across the center. The former is
smaller and has a smoother outline than the latter. Synura is a small moving
yellowish ball of oval animalcules joined concentrically and closely packed.
Uroglena is a large jelly-like hollow ball of minute animalcules, the latter
embedded near the surface, similar in appearance. tg Volvox, as previously
mentioned, but lacking the large green cells. It.-roll}> around slowly in the
center of the liquid of the cell. Broken pieces have- the shape of a piece of a

MICROSCOPICAL EXAMINATION OF WATER

2111

rubber ball. It is visible to the naked eye and recognizable in the bottle.
Rotifera, Crustacea, etc. (more complicated animal forms) are readily distinguished from their pictures in most instances. Many are visible to the
.naked eye. Cyclops may readily be recognized in the bottle by its shape and.
swift darting motion. It appears about the size of a pin head shaped somewhat like a tiny fish, but wider in proportion. Bosmina (little cow) and
Acarina (little spider) each has a peculiar characteristic jerky motioh.

CONTROL OF MICROSCOPIC ORGANISMS

. Troublesome microscopic growths are controlled as follows: Reservoirs
are shut off from the service and bypassed until the forms die out under storage
which may take three weeks to three months. Aeration by fountains will disintegrate certain fragile forms like Uroglena, Anabaena, even Asterionella, and
assist in removing odors. Crenothrix may be controlled by chloramine or
liquid chlorine (Monfort & Barnes). Liquid chlorine has also been widely
employed the last 16 years to help control Synura, Uroglena, Dinobryon etc.
But the method of widest application is that of Moore & Kellerman, the application of copper sulfate in dosage ranging from .05 to 1.00 part per million by
weight according to the particular genera of microscopic organisms present.
The chemical has been usually applied by rowirig around in boats dragging
burlap bags, containing, usually, fifty pounds each, through the surface of the
water, allowing wind, wave and diffusion to mix the treated water with the
remainder. On large reservoirs launches are employed with arrangement on
either side for suspending the copper sulfate crystals in the water and the action
of the propeller causes considerably wider distribution of the chemical. In
the last few years special methods of distribution by launch have been devised,
particularly on the west coast, the crystals being distributed in solid form by a
mechanical blower by one method and by pumped water by another method.
, These methods are said to accomplish wider, more uniform, and faster distribution. Bett~r depth effect is also accomplished, the crystals dissolving as they
sink. Size of crystal is also chosen depending upon depth to be treated.
New York City has demonstrated, very satisfactorily, that dry feed of the
chemical by automatic accurate electrical apparatus to the water flowing in
an aqueduct a short time before entering the reservoir is absolutely feasible
and successful even at winter temperatures and with ice-covered reservoirs.
The feed is by means of a slowly moving shutter fed from 'a hopper. The
copper sulfate drops into floating perforated wooden box in a narrowed portion
of the aqueduct. Dosage is similar to that used in the feservoirs by the
boat method, and under such circumstances the killing dosage is known exactly.
The effect of treatment of microscopic organisms' by copper sulfate is
shown by an immediate intensification of distinctive odors, by reduction of the
number of organisms in the water through sedimentation, by the appearance of '

2112

MICROSCOPICAL EXAMINATION OF WATER

the organisms under the microscope-the coloring ma,ter being knocked to

pieces, so to speak. Sometimes there is an increase in the water bacteria which
feed upon the decayed organisms. The Cyanophyceae may produce sculh
after treatment which is of varied colors, pale blue, ye1l9w, red or brown.
Sometimes fish are killed by the treatment but this is more apt to occur
from smothering due to clogging of the gills with dead brganisms or reduction
of oxygen 'by the dead organisms. It is only occasionally that fish are killed'
in quantities, even though larger doses are usually applied than shpwn in the
following table calculated from data published by Moore & Kellerman:
Limiting Dosage of Copper Sulfate for Fish Life
Fish

Part. per Million

Trout ...... . .... . ..... . . . . . ... . .. .. . .. 0.14

Carp ...... . ..... .. ... . ... . . . ........ . 0.33
Suckers .... .. . ... .. . . ... . . . . ... .. .. . . . 0.33
Catfish . . . . .. . .. ... .. . ... . . .. . . .. . .. .. 0.40
Pickerel. . . . . . . . . . ..... . . ... .. . ... . ... 0.40
Goldfish .... . . ... . . .... . .. . ...... . . . . . 0.50
Perch . . ... , . . ...... . ...... . .. " ...... 0.67
Sunfish. . ......... . . . . . ... . . . .. . . .. . ~ . 1.33
Black Bass . ... . ....... . . ... . . .. . . . . . .'l.GO

Pound. per
Million Gallons
(Approximate)

1.2
2.8
2.8
3.5
3.5
4.2
5.5
11.1
11:'l.6 -

Experience differs with waters of different composition and with temperature as to the right dosage to apply for each organism. It is believed that it is a
waste of copper sulfate to apply an insufficient quantity. When the correct
quantity is applied the water is usually cleared within three to four days.
The following table on pages 2113-2114 is made up from data furnished .by
Moore & Kellerman, Whipple, various other sources and personal experience.
The latter are starred. Results in recent years with chlorine are also included.
The following illustration shows the corner of a bacteriological laboratory
for the microscopical examination of water.

FIG. 294.-Microscopical Apparatus, Catskill Labo~atory, Asbokan, N.

Y.

MICROSCOPICAL EXAMINATION OF WATER

2113

Chemicals Required for Treatment of Different Genera

Organisms

Odor

DIATOMACEAE:

Parts per
Million

Aromatic, ger~nium, fishy 0.12*-0.20*

Faintly aromatic
.... .
Faintly aromatic
.... .
0.25
'0.2
Aromatic
0.07
0.50
Earthy
0.36*-0.50
0.33
Aromatic, geranium, fishy 0.12*-0.50*

Asterionella t
Cyclotella tt
Diatoma t
Fragilaria t
Melosira tH
Meridion
Navicula t
Nitzchia
Synedraltt
Stephanodiscus
Tabellaria t

POUljds per

l\IiI. Gals.

1.0-1.7

Chlorine
p.p.m.

0.5*-1.0"
l.0

2.1
1.7

2.0

0.6
4.2
3.0-4.2
2.8
1.0-4.2

1.0
0.5*-Lo

CHLOROPHYCEAE:

Cladophora t
Closterium
Coelastrum t
Conferva t
Desmidium
Dictyosphaerium
Draparnaldia
Eudorina
Entomophora
Gloeocystis .
H ydrodictyon
Microspora
Palmella
Pandorina
Protococcus
Raphidium
Scenedesmus
Spirogyra t
Staurastrum .
Tetrastrum
Ulothrix
Volvox t
Zygnema

0.50
0.17
0.05-0.33
0.25
2.00

4.2
1.4
0.4-2.8
2.1
16.6

0.33
2.0-10.00
0.50 .

2.8
16.6-83.0
4.2

Faintly fishy

0.10
0.40
2.00
2.0:..10.00

0.8
3.3
16.6
16.6-83.0

Grassy

1.00
1.00*
0.12
1.50

8.3
8.3
1.0
12.5

0.20*
0.25
0.50

1.7
2.1
4.2

0.12*-0.48
0.12*-0.50"
0.12*-0.25*
0.20*-0.33
0.12
0.20
0.20-0.50*

1.0-4.0
1.0--4.2
1.0-2.1
1.7-2.8
1.0
1.7
1.7-4.2

0.5"-1.0

Grassy, nasturtium, fishy

Faintly fishy
Offensive
Very offensive

1.0

0.7-1.5
1.0

Fishy

0.3*-1.0

CYANOPHYCEAE:

Anabaena t
A phanizomenon t
Clathrocystis
Coelosphaerium t
Cylindrospermum
M icrocystis
Oscillaria t
Rivularia

Moldy, grassy, vile

Moldy, grassy, vile
Sweet grassy, vile
Sweet grassy
Grassy

0.5*-1.0
0.5*-1.0*
0.5*-1.0
0.5*-1.0
1.1

Moldy, grassy

The organisms where odors are expressed have been in sufficient quantity at one time
or another to cause the characteristic odor.
* Dosage successful in New York City's supplies.
.
t These organisms have caused trouble other than odor.
t These organisms have been affected by chlorine producing characteristic odor and
in many cases controlled by dosage ranging from 0.3 to 3 p.p.m. depending largely on
amount of organisms.
Controlled by excess caustic lime, 5 p.p.m. in the case of the diatoms.
NOTE.-Range of dosage largely due to temperature.

2114

MICROSCOPICAL EXAMINATION

OF WATER

Chemicals Required for Treatment of Different Genera (Continued)

Organisms

PROTOZOA:
Bursaria
Ceratium

tt

Chlamydomonas
Cryptomonas
Dinobryon t
Endamoeba hislolylica (cysts)
Euglena
Glenodinium
Mallomonas
Peridinium
Synura t
Uroglena

Odor

Leptothrix t
Sphaerotilis, natans
SpirophyUum t
Thiothrix tt
FUNGUS:
Achlya
Leptomitu8
Saprolegnia

Fishy
Aromatic, violets, fishy
Fishy, like clam-shells
Cucumber, muskmelon,
fishy, bitter taste
Fishy, oily, cod liver oil

. ...
2.0*-2.8

0.3*-1.0

4.2-8.3
4.2
1.5

0.3*-T.O
25.0-100.0

0.3*-1.0

0.05*-0.20*

0.4-1.6

0.3*-1.0

.....

.....
.....
0.40
0;18

Objectionable

Chlorim!
p.p.m.

4.2
4.2
4.2
4.2-16.6
1.0-2.1

5.00
0.20
Very offensive decayed, 0.33-0.5*
medicinal with chlorine
.....
Very offensive decayed
0.10
Very offensive decayed

Pounds per
Mil. Gals.

0.50
0.50*
0.50
0.50*-2.00
0.12*-0.25*

Very offensive decayed

I
I

.....
2.0

MISCELLANEOUS:
Chara t
N itella, jlexilis t
Potamogeton t
Nais t
Blood worm,
chironomus t
Blood worm, gnats
,

Million

Irish moss, salt marsh, fishy .....

I
Fishy, vile (rusty brown 0.24*-0.33
color)
0.50-1.0
Candied violets
0.50
Aromatic, violets, fishy
0.18*

CRUSTACEA:
Cyclops tt
Daphnia tt
SCHIZOMYCETES:
Beggiatoa t
Cladothrix
Crenothrix tt

Parts per

0.10-0.50
0.10-0.18*
0.30-0.80
.....
....
.....

....
16.6

1.0-3.0
1.0-3.0

41.5
1.7
2.8-4.2

0.5*

. ...

3.3
....
. ...

0.25*
0.5*-1.0

....

0.6*

3.3
1.5

0.8-4.2
0.8-1.5
?5--6.7

...

'~"-

....
. ...

1.0*
15.0-50.0

3.0

REFERENCES

George C. Whipple (Revised by Fair and Whipple), "The Microscopy of Drinking

Water.". Wiley & Sons.
Moore and Kellerman," A Method of Destroying or Preventing the Growth of Algae,
etc." Bulletin 64, Bureau of Plant Industry, U. S. Dept. of Agriculture.
Moore and Kellerman, "Copper as an Algicide and Disinfectant'in Water Supplies."
Bulletin 76, Bureau of Plant Industry, U. S. Depk-q_f"Agriculture.

MICROSCOPICAL EXAMINATION OF WATER

2115

"The Use of Copper Sulfate and Metallic Copper for the Removal of Organisms and
Bacteria from Drinking Water,-A Symposium," Vol. 19, No.4, New England
Water Works Association.
Monfort and Barnes, "Chlorarriine and Crenothrix." Journal of American Water
Works Association, Vol. 6, page 196.
Amsbary, "Treatment of Water to Prevent the Growth of Crenothrix." Jour. Am. W.
'W. Assoc., Vol. 6, page 194. .
Monfort, "Crenothrix Removal." Proe. Convention Indiana Sanitary Water Supply
Assoc., April, 1919.
W. W. Brush, Engineering News-Record, Feb. 16, 1922.
Frank E. Hale, "Taste and Odor in New York City's Supplies," Jour. Am. W. W.
Assoc., Vol. 10 (Sept., 1923), pp. 829-837.
Hudson and Gosse, "The Rotifera." Two volumes and supplement, Longman & Co.
W. Saville Kent, "A Manual of the Infusoria." Three volumes.
Wolle, "Fresh Water Algae of the United States." Two volumes.
Wolle, "Desrriids of the United States." One volume.
Wolle, "Diatomaceae of North America." One volume.
Leidy, "Fresh Water Rhizopods of North America." U. S. Geological Survey of the
Territories, Volume 12.
Ward and Whipple, "Fresh Water Biology." Wiley.
Griffith and Henfrey, "Micrographic Dictionary." London.
Gilbert M. Smith, "The Fresh-water Algae of the United States." McGraw-Hill Book
Company.

BACTERIOLOGICAL EXAMINATIO~ OF WATE~

The routine bacteriological examination is essentialiy simple. It consists
of obtaining empirically the total number of bacteria under certain standard
conditions and of testing for the intestinal group, B. coli. Usually only under
special conditions is search made for specific disease bacteria which may be
carried by water supply, such as Bacillus typhi, Bacillus paratyphi-A and B,
Spirillum cholerae, Bacillus dysenteriae-various types, Bacillus anthracis, etc.

SAMPLIN(j-

Samples must be taken in dry-steriIlzed bottles, with well-ground glass

stoppers. Two ounce capacity is usually sufficient for routine tests. The
bottles are preferably protected by covering the stoppers and necks with tinfoil
and further protected by placing in a metal screw-top can (two inch by four
inch). It is frequently convenient to use quart bottles stearn-sterilized at
fifteen pounds pressure, and capped
with tinfoil, when special samples are
to be brought in by hand,---;-the same
sample serVing for the physi~al,
microscopical and chemical analyses
after making the bacteriological
examihation. Such samples transporte<;l in leather bags of proper
size do not change temperature
more than a few'degrlles in several
hours on the hottest days though not
protected by ice. For the twoounce bottles special aluminum
cases, 'holding small quantities of ice
and fittiIlg the bags, are light and
convenierit. For shipping, the small ,
FIG. 295.-Collect!on Apparatus for Bacteri- bottles may be packed in a copperologlCal Samples.
lined wooden box containing a receptacle for the bottles and space for ice below. (See Fig. 295.)
In sampling care must be taken to take repres,e ntative, ~amples, avoiding
contamination and avoiding stagnant water. TemI>"frature'is often a guide but
it must be taken from the source or in a separate bottle with the thermometer
2116

BACTERIOLOGICAL EXAMINATION OF WATER

2117

bulb immersed in the water, a few minutes before reading. Allow tap samples to
run thoroughly, particularly in warm boiler rooms. Handle sterilized bottles
by the bottom and keep fingers away from the necks. Handle stoppers with
the tinfoil in place. Avoid splashing or over-running bottles. In streams
plunge bottle below the surface and sample by swinging hand IfP-13tream with
mouth of bottle pointed up-stream so as to avoid water washing over the hand
into bottle. Similarly in a reservoir, swing the mouth of the bottle forward
while filling. Do not rinse a sterilized bottle. These rather minute directions
are given because in a long experience they have frequently been found necessary. Experts do not always take the samples though the value of their work
depends upon the care .with which the samples are taken. Men have even
been caught reading thermometers after drawing them up twenty feet from
the surface of a reservoir.
Samples from a depth in a reservoir are best taken if possible from an
effluent pipe when regular draft is from a depth. In other cases sterilized
bottles may be lowered to the required depth and stoppers pulled out by
mea~ of special samplers. In many cases it has been found satisfactory to
take them by the two-bottle method as used for di'lsolved oxygen and described
under "Microscopic Examination of Water." In this case the quart bottle
should be thoroughly cleaned, handled as little as possible, and the w8ter
.poured into a bacteria bottle after drawing to the smface. The quart bottle
is pretty thoroughly rinsed during the process of sampling. (Fig. 291, p. 2102.)
Samples should be tested the same day they are collected, preferably within
a few hours. Expressed and iced samples will p;obably give a fairly reliable
index of B. coli content but cannot be relied on for total counts.
Apparatus and Materials Needed
Bacteria bot~les-two-ounce glass-stoppered (well ground).
Metallic containers for &bove, screw-cap, two inch by four inch.
Tinfoil, five inch wide, in rolls for, capping.
Bacteria certificates for sample data.
Containers for carrying samples-leather bags with aluminum ice cases to fit.
Containers for shipping samples-copper-lined ice chests.
Dry-sterilizer for bottles and pipettes, to operate at 1(0 0 C.
Autoclave to operate at 15 pounds steam pressure, sterilization of media.
1 ml. pipettes with two marks and also conveniently g:r-aduated to show 0.1 m!.
above the upper 1 m!. mark.
10 m!. pipettes, preferably straight tube pipettes with graduation also for 5 ml.
and 9 ml.
Pipette boxes for sterilized pipettes, 3t" square by 14;' long.
Dilution test tubes, 1/1 diameter, medium w~ight.
Media test tubes, i/l diameter, heavy weight, not lipped.
Dilution flasks, Erlenmeyer, 16-ounce capacity, for sterilized water.
Flasks, ! 'liter, 1 liter, etc., for media preparation and bulk storage.
Test tube racks for dilution tubes.
Wire cages for media tubes, 4/1 by 5/1 by 5!/1 high.
Non-absorbent cotton for plugging tubes.
Footless Smith tubes for B. coli tests, or small test tuhes inverted in large test

2118

BACTERIOLOGICAL EXAMINATION ,OF WA.TER

tubes for same, capacity in either case four times the amount'ofwater to be
tested.
Metal racks for Smith tubes.
Agateware double boilers for making media.
Agateware kettles for making media.
Agateware funnels, 6 inch, for filtering media.
Cotton flannel for filtering media.
Large beakers for making media.
Earthenware pitchers for filling tubes with media.
Side-neck burettes for filling tubes with media (side ntlck connected to funnel of
media). .
Apparatus for keeping media warm while filtering.
Gas stoves for heating media.

,
FIG. 296.-Bacteriological Apparatus. Fermentation Tubes, Petri Dishes,
Culture Media, etc.
'

Petri dishes, 10 cm. diameter and flat bottomed, with both glass covers (20 0 C.
incubation) and porous covers (37 0 C. inCUbation).
Agateware sauce pans for welting media in tubes before plantin.g;
20 0 C. incubator with thermometer, automatic regulator and safety burner.
37 0 C. incubator with thermometer, automatic regulator and safety burner.
(In many laboratories the incubators are now electrically operated.)
Burner for flaming tubes before planting.
Counting stand with engraved lines to assist in counting.
Engraver's lens for counting, B. & L. 146, magnification 2t diameters, 3t X.
Tallying register for counting.
'
Balance, similar to an apothecary's, for quick weighing.
Balance, analytical beam type, for accurate weighing.
Record books for bacteriological results.
,
Microscope with condenser and assortment of len~~ -at least t", t" and I"
eyepieces and ~", i" and 1/12" objectives. "(Satisfactory hanging drops

;u

BACTERIOLOGICAL EXAMINATION OF WATER

2119

may be made with t" eyepiece and t" objective. The 1/12" is used only
in oil immersion with condenser.)
Slides, plain and concaved (latter for hanging d;ops).
Cover slips, round or square.
Platinum needles with holders for transferring cultures.
Cleaning mixture for cleaning glassware.
Distilled water for preparation of all culture media and reagents.
Meat extract; :r5igestive Ferments Company's Bacto Beef Extract is satisfactory.
Peptone: N. Y. Ci~y Water Department Laboratories have found Digestive
Ferments Company's reliable. (Some laboratories have found Armour's
and Fairchild's satisfactory.)
Lactose and other sugars the purest obtainable.
Agar. Frequently needs soaking in water and draining before use because of
salt. Shall be dried t hour at 105 C. before weighing.
Gelatine, of light color, containing not more than a trace of arsenic, copper,
sulfides, free from preservatives, and of such a melting point that" standard
nutrient gelatin" shall not melt below 25 C. Gelatin shall be dried t hour
at 105 C. before weighing.
Litmus-reagent litmus of highest purity (not litmus cubes) or azolitmin
(Kahlbaum's) .
Certified Dyes-Basic fuchsin, Eosin yellowish, Methylene blue, Brilliant green.
General chemicals-chemically pure so far as possible.
Hellige comparator for hydrogen-ion with color discs for brown-thymol-blue,
and cresol red and phenol red.
Adjustment of Reaction.-T,he final reaction for broth, gelatin and agar
shall be between pH 6.4-7.0. For Endo medium the final reaction after the
addition of chemicals should be pH 7.8-8.2. The eosin methylene-blue agar
does not require adjustment. The increase in the hydrogen-ion concentration
during sterilization will vary slightly with the individual sterilizer in use and
the initial reaction required in order ,to obtain the correct final reaction will
ha ve to be determined.
The simplest method is the use of the Hellige Comparator which employs
discs with colored glasses. Each disc cantains the range of colors for a single
dye, 0.2 pH apart, and the colors are a very good match and accurate. Bromthymol-blue has a range 6.0--7.6, Phenol Red 6.8-8.4, and Cresol-red a range
7.2-8.8. The first covers all the ordinary media, the second Brilliant-green bile
diluted, and the latter covers Endo. Procure the dyes from any of the firms
supplying it for pH determination. The preparation of the dye solutions is as
follows according to Clark. Rub up a decigram of brom-thymol blue in 3.2 ml.
of N /20 sodium hydroxide. Dilute to 25 ml. with distilled water, making 0.4%
solution. Dilute again for use to 0.04% solution. Rub up
decigram of
Cresol-red in 5.3 ml. of N /20 sodium hydroxide. Dilute to 25 ml. with distilled water, making 0.4% solution. Dilute again for use to 0.02% solution.
With Phenol-red 5.7 ml. of N/20 sodium hydroxide is used and final solution is
.01 % strength. In testing bacteriological media, for example agar, add to
8 ml. of distilled water 2 ml. of melted ~ar, add t ml. of the brom-thymol-blue
solution, mix by pouring into one of the comparison tubes. In 'another comparison tube pour the similarly diluted media but without the dye and place
this behind the colored glasses. Comparison should be made at once and is

2120

BACTERIOLOGICAL EXAMINATION ,OF WATER

sufficiently accurate. The prepared solutions of the dyes gradually change in

pH. The solution should be. sterilized and stored in' sterile pyrex bottles.
Sterile pipettes should also be used. Always allow thcidye to drain from the
pipette and never blowout as the carbonic acili of th~ breath will influence
results. In laboratories which can afford to have th~, equipment, the dye
solutions should be checked by the electrometric method and proper corrections
applied. For this purpose standard buffer solutions covering a range of pH
should be purchased.
Sterilization.-All media and dilution water shall be sterilized in the autoclave at 15 Ibs. (120 0 C.) for 15 minutes after the pressure reaches 15 Ibs., All
air must be forced out of the autoclave before the pressure is allowed to rise.
As soon as possible after sterilization the media shall be removed from the
autoclave and cooled rapidly. Rapid and immediate cooling of gelatin and
lactose broth is imperative.

Media shall be sterilized in small containers, and these must not be closely
packed together. No part of the medium shall be more than 2.5 cm. from
the outside surface of the glass. All glassware shall be sterilized in the dry oven'
at 170 0 C. for at least 1 hour. Sample bottles may be sterilized in a steam
autoclave at 15 Ibs. pressure for\ hr.
Nutrient Broth.-To make one liter:
1. Add 3 grams of beef extract and 5 grams of peptone to 1000 ml. of
distilled water.
/
2. Heat slowly on a steam bath to at least 65 0 C., stirring until dissolved.
3. Make up lost weight with distilled water and adjust the reaction so that
final pH will be between 6.4 and 7.0 after sterilization.
4. Bring to a boil over a free flame. Cool to 25 C. and filter through filter
paper until clear.
5. Distribute in test-tubes, 10 ml. to each tube.
6. Sterilize in the autoclave at 15 Ibs. (120 0 C.) for 15 minutes after the
pressure reaches 15 Ibs.
Sugar Broths.-Sugar' broths shall be prepared in tHe same general manner
as nutrient broth with the addition of 0.5% of the reqUired carbohydrate justbefore sterilization. The removal of muscle sugar is unnecessary as the beef
extract and peptone are free from any fermentable carbohydrates. The reaction of sugar broths shall be pH 6.4-7.0. Sterilization shall be in the autoclave
at 15 pounds '(120 0 C.) for 15 minutes after the pressure reaches 15 pounds,
provided the total time of' exposure to heat is not more than onll-half hour;
otherwise a 10% solution of the required carboJ:iydrates shall be made in distilled water and sterilized at 100 0 C. for I! hours, and this solution shall be
added to sterile nutrient broth in amount suffici~nt to make a 0.5% solution,
of the carbohydrate and the mixture shall then be tubed and sterilized at
100 C. for 30 minutes, or it is permissible to add by,means of a ster~le pipette
directly to a tube of sterile neutral broth enough of the carbohydrate to make
the required 0.5%. The tubes so made shall be incubated at 37 Q. for 24 hours
as a test for sterility.
Nutrient Gelatin.-To make one lit!lr:
1. Add 3 grams of beef extract and 5 grams of p'eptone t6 1000 ml. of distilled water and add 100 grams of &elatin dried-fuf~ne-half hQur at 1050 C.
before weighing.

BACTERIOLOGICAL EXAMINATION OF WATER

2121

2. Heat slowly on a steam bath to 65 C. until all ingrec,lients are dissolved.

3. Make up lost weight with distilled water and adjust the reaction so that
final-pH shall be 6.4-7.0.
.
4. Bring to boil. Filter through cloth and cotton until clear.
5. Distribute in test-tubes, 10 ml. to each tube, or in larger containers as
desired.
'
6. Sterilize in the autoclave at 15lbs. (120 C.) for 15 minutes after the
pressure reaches 15 lbs. Cool rapidly as soon as possible.
Nutrient Agar.-To make one liter:
1. Add 3 grams of beef extract, 5 grams of peptone and 12 grams of agar,
dried for one-half hour at 105 C. before weighing, to 1000 ml. of distilled water:
Boil over a wafer bath until all the agar is dissolved.
2. Make up lost weight with hot distilled water and, adjust the reaction so
that final pH shall be 6.4-7.0 .
. 3. Filter through cloth and cotton until clear.
4. Distribute in test-tubes, 10 ml. to each tube, or in larger containers, as
desired.
5. Sterilize in the autoclave at 15 lbs. (120 C.) for 15 minutes after the
pressure reaches 15 lbs.

NOTE.-The Water Department Laboratories of New York City for many years have
used and found the following method of clarifying agar very efficient: The prepared agar
is placed in a suitable container and filtered upward by vacuum in a pressure cooker into
four one-quart jars. The mouth of the jar is covered with a filter made of a thick wad of
absorbent cotton 2 inches thick between two layers of cotton flannel, the filter occupying
thE) neck of the jar. The cotton flannel is tied securely in place on the outside of the neck.
These jars are inverted in the container holding the agar and placed in the pressure
cooker. The pressure is brought up to 15 pounds and allowed to cool. Upward filtra-
tion takes place under vacuum.

Eosin Methylene Blue Agar

1. To 1000 ml. distilled water add 10 grams of peptone, 2 grams of dipotassiumphosphate (K 2HP0 4) and 15 grams of undried agar. Boil ingredients
until dissolved and make up any loss due to evaporation with distilled water.
2. Place measured quantities (100 or 200 ml.) in flasks or bottles and sterilize
in the autoclave at 15 pounds for 15 minutes.
3. Just prior to using, melt above agar in a water bath and add the following
to each 100 ml. of agar:
lactose, sterpe 20% solution 5 m!.
eosin, yellowish, 2% aqueous solution 2 m!.
methylene blue, 0.5% aqueous solution 2 ml.
4. Mix thoroughly and pour medium into Petri dishes, allow to harden and
inoculate by streaking on the surface.
Endo Medium
FORMULA

1. Preparation of Stock Agar.-Add 5 grams of beef extract, 10 grams of

peptone and 30 grams of undried agar to 1000 m!. of distilled water.
Boil until the agar is dissolved and make up lost weight with distilled water.

2122

BACTERIOLOGICAL EXAMINATION OF WATER

Adjust the reaction so that the pH reading after sterilization will be 7.4.
Clarify if desired. Add 10 grams of lactose and dissoh,le.
Place in small flasks or bottles, 100 ml. to each, and ~terilize in autoC'laV'e
at 15 lbs. pressure for 15 minutes after the pressure has, reached 15 lbs.
_
basic
fuchsin
2. Preparation of Plates.-Prepare a 3% solution .of cJrtified
I
in 95% methyl alcohol. Allow to stand 24 hours and filter.
Melt lactose agar as prepared above and'to each 100 ml. add 1 ml. of the
3% 'basic fl1chsin solution and 0.125 gram of anhydrous sodium sulfite dissolved
in 5 ml. of distilled water. The su)Jite solution must be freshly prepared. Mix
thoroughly, pour plates with usual precautions against contamination and allow
to harden.
The medium should be light pink when hot and almost colorless when cool.
If the medium is too highly colored before incubation the strength of the basic
fuchsin solution may be varied.
FORMULA II
Dipotassium phosphate ...................... " . . .. .. 3.5 grams
Bacto peptone .......................... , . . . . . . . . .. 10.0, "
Agar (washed and dried) ....... ' ......... , . . . . . . . . .. 15.0 "
Water (distilled). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 liter
Lactose (c.p.) ......................... ~ .... " ...... 10.0 grams

To each 100 ml. of the above add:

Sodium sulfite (Anhydrous) ..........--: . . . . . . . . . . . . . .. 0.25 "
Basic fuchsin filtered 10% alcoholic solution. . . . . . . . . . . . 0.5 ml.
NOTE.-It is difficult to make a 10% solution with some basic fuchsin preparations.

Brilliant-Green Lactose Peptone Bile.-The composition for approximately

'one liter is as follows:
Dissolve 10 grams of pepto~e and 10 grams of lactose in not more than 500
ml. of distilled water.
'
.'
Add 200 ml. of fresh ox bile or 20 grams of dehydrated ox bile dissolved in
200 ml. of distilled water. The solution of dehydrated ,ox bile shall have a pH
between 7.0 and 7.5.
/
Make up with distilled water to approximately 975 ml. and adjust reaction
to 7.4 pH.
I
Add 13.3 ml. of a 0.1 % solution of Brilliant-Green in distilled water and add.
sufficient distilled water to make the volume 1000 m!. Filter through cotton,
distribute in fermentation 'tubef:l and sterilize at 15 lbs. pressure for 15 minutes
after pressure reaches 15 lbs. Total exposure to heat should not .exceed one_,
half hour. Cool rapidly after removal from autoclave.
The reaction after sterilization should be between pH 7.1 and 7.4. 6
Brilliant-Green Lactose Peptone Bile of the above formula may be obtained
mixed in dry form from the, Digestive Ferments Company, Detroit, Mich., and
has the following composition:
I
Dried oxbile . . . . . . . . . . . . . . . . . . . .. 20
grams
Peptone. . . . . . . . . . . . . . . . . . . . . . .. 10
"
Lactose . ........................ 10
((
Brilliimt-Green. . . . . . . . . . . . . . . . .. 0.0133 '~
6 The pH may be determined colorimetrically witll..Hellige equipment by use of
Phenol Red indicator in 0.01 % solution using dilutio'n -of medium of 1 : 4. Results
check. with those determined potentiometrically. The color of the medium is compensated by the equipment.

BACTERIOLOGICAL EXAMINATION OF WATER

2123

For a water inoculum of 1 m!. or less or for loop inoculations the above dry
material should be used in proportion of 4 grams to 100 ml. media water (distilled) and tubed in 10 ml. quantities. The concentration of bile in the inoculated tube will then be 2% and of Brilliant-Green 1 part in 75,000. For 10 m!.
water samples a total volume of water plus medium of 30 ml. is used and the
medium is made using 6 grams of the above dry mixture per 100 m!. of media
water, which results in the same final density of ingredients.
For routine presumptive planting on a large scale it is simpler to make up
double strength media as follows:
. Dissolve in 4 liters of hot distilled water, 160 grams of dessicated bile, 80
grams of peptone, and add 15 ml. of N /1 sodium hydroxide. Remove from
flame and add 80 grams of lactose. After solution filter through flannel, make
up to original volume and add 10.6 m!. of 1 % Brilliant-Green. .
For small tubes in which are planted 1 m!. or less of water, dilute 1500 m!.
of the above to 3000 m!. and tube in 10 ml. quantities.
For large tubes in which are planted 10 m!. of water, dilute 2500 ml. to 3750
ml. and tube in 30 ml. quantities.
For testing of 100 m!. quantities of water use the double strength media as
prepared, 100 m!. quantities of media per 100 m!. quantity of water tested.
Hesse Agar.-The composition for approximately, 1 liter is as follows:
Agar (dried). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.5 grams
-Peptone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.
"
Beef extract, Liebig's. . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
5.
"
8.5
"
Salt (sodium chloride). ,'. . . . . . . . . . . . . . . . . . . . . . . . . . .
Distilled water . .................................. . 1000.
"

Dissolve 4.5 grams of dry agar in 500 m!. distilled water by heating over a
free flame, making up loss in weight by evaporation. Into another vessel 500
ml. of distilled water is poured and to this is added 10 grams of peptone, 5
grams of Liebig's beef extract, and 8.5 grams of salt. This is heated until
all is dissolved and the loss in weight by evaporation is made up by adding
distilled water.

Add the two solutions together; boil 30 minutes; make up loss in weight
with distilled water, filter through absorbent cotton held in the funnel by
cotton flannel, passing the filtrate through several times until perfectly clear.
Test the reaction; adjust, if necessary, to pH = 7.0, and tube, using 10 m!. in
each tube. Sterilize for 15 minutes at 15 lbs. pressure in an autoclave. Cool
with running tap water and store in an ice-chest, the air of which is saturated
with moisture.
Russell Media.-To ordinary extract agar adjusted neutral to litmus add
1% of lactose and 0.1% of glucose and sufficient litmus to give a good color.
Tube and slant leaving a generous "butt" at bottom of tube for stab inoculation.

2124

BACTERIOLOGICAL EXAMINATION OF WATER

EXAMINATION OF WATER FOR TOTAL BACTERIA

The gelatine count is more particularly useful in connection with determining the efficiency of filter plarlts, since the counts ~ie higher than the agar
counts (averaging usually about ten times as high) and lin examination of weil
waters by indication of surface contamination through the presence of liquifiers.
In plating with gelatine media it is usually necessary to use 0.1 mI., or less,
as well as 1 ml., in order to obtain the right amount of colonies on a plate.
Standard methods require planting in duplicate.
Melt tubes of nLitrient gelatine or nutrient agar in boiling water,-the
latter needs thorough boiling for several minutes to soften entirely. Let cool
until luke-warm, just barely warm to the touch. Shake the sample of water
vigorously and introduce 1 mI. or less into the bottom of a petri-dish lifting the
cover on one side but not removing from over the bottom. The quantity of
water used should of course be accurately measured. Be careful not to contaminate the pipette by touching to anything other than the inside bottom of
the petri-dish. Use glass-covered petri-dishes for incubation at 20 C. and
porous covers for 37 C., in order to overcome spreaders at the higher temperature. Remove a tube of the media from the warm water, dry with a towel,
remove the cotton plug, flame the open end of tube and pour into petri-dish.
Carefully rotate dish to mix the media with the water and set on a level plate
to harden. Agar hardens so quickly tllat it must be handled and mixed
expeditiously. Incubate gelatine plates at 20 C. and agar plates at 37 C.
The gelatine plates are counted after 48 hours and the agar plates after
24 hours.
Each minute bacillus must grow where caught in the solid medium and
after sufficient time forms a colony large enough to be seen with the naked eye.
Many are however very tiny, so that a hand lens magnifying two and one-half
diameters is used in counting. A tallying machine is used to keep count, one
tally being made for every three colonies usually, to assist the speed of counting.
For accuracy there should be 30 to 300 colonies .on a ''plate, unless less than 30
grew from 1 mI. Fictitious accuracy should be avoided. The following table
should be followed in recording results:
Number of bacteria per ml.
From
1 to
50 shall be recorded as found_
"
51 "
100"""
to the nearest 5
101 "
250""""""
10
"
"
251 "
500""""""
25
"
501 "
1,000""
"
"""
50
"
1,001 "
10.000""
"
"(("
100
50;000""
"
"""
500
"
10 001 "
50;001"
100 000 "
""""
1,000
"
"
100 001"
500:000"
",,'"''
10,000
500;001" 1 000 000 " "
"
""
"
50,000
"
"
""
"
100,000
" 1,000,001" 10;000;000 " "

The counting stand is usually marked to facilitate counting, for example,

by concentric circles and lines through the center forming sectors. With high
numbers four opposite sectors may be counted, sitJIateCl at right angles to each
. .
other, and multiplied by a factor.

BACTERIOLOGICAL EXAMINATION OF WATER

2125

Examination for B. Coli

B. coli (bacillus of the colon) represents a group of bacteria inhabiting tl:e
large intestines of man and animals. Its presence in water supply signifies
possible sewage contamination and the latter means probable typhoid, dysentery, etc., germs. In fact if a drinking water becomes contaminated by
sewage, an increase in typhoid fever is almost certain to follow. Animals do
not spread typhoid though they may be a factor in paratyphoid. B. coli in a
water supply may be due to harmless sources and so of little significance.
The present Standard Methods define the coli-areogenes group as including
all aerobic ang. facultative anaerobic Gram-negative non-spore-forming bacilli
which ferment lactose with gas formation.
Various classifications of the Coli group have appeared in Standard Methods.
The following was included in the 1912 edition. One function of a classification
is to make possible a satisfactoryjdentification of dis,tinct members. N umeroui!.
members of this grouping were isolated at a hospital from pustules or sores,
autogenous vaccine made from each variety isolated, the vaccine used, and the
condition cleared up from which the original germ was isolated. For many
years pure cultures of these organisms were maintained and their characteristic
reactions persisted. This classification therefore appears to have decided usefulness.

Subclassification of B. coli

B. coli Group.-The general characteristics common to this group are:

Fermentation of dextrose and lactose with gas formation, short bacillus
with rounded ends, non-spore-forming, facultative anaerobe, gives positive test
with esculin, grows at 20 C. on gelatine and at 37 C. on agar, non-liquefying
in fourteen days on gelatine, gram-staining negative.
The group consists of four species:
B.
B.
B.
B.

communior (Durham).
communis (Escherich).
aerogenes {Escherich).
acidilactici (Hiieppe).

This group is differentiated as follows:

B. COLI' GROUP
Dextrose
Lactose +

I
I

Dulcite

Dulcite B. aerogenes
B. acidi lactici

B. communior
B. communis

' I

Saccharose +
Saccharose Saccharose +
B. communior
B. communis
B. aiJrogenes
7 AdC'pted by A. P. H. A., 1912, and recommended by the author.

Saccharose B. acidi lactici

2126

BACTERIOLOGICAL EXAMINATION: OF WATER

The "Standard Methods" in testing for the coli-aero genes group require
gas formation in lactose broth plus confirmation in accordance with the following procedure:
'
"The following classification is included in the 1936 edition of 'Standard
M'ethods! Its chief usefulness is to divide the group! into two broad types,
B. coli and B. aerogenes. The latter type represents ml\lmbers that can persist
under adverse conditions, hence less sanitary importance may sometimes be
applied to their presence. In England, particularly, more importance is
ascribed to the B. coli test. Attempt has been made to separa:te the aero genes
type into Fecal and Non-fecal grouping, but as all types are found in human
excrement this separation has not generally been found of any great importance.

.
Indol

Coli-Aerogenes Group-Reaction Classification

I'ossible Interpretation when Isolated from Water
by the Standard Methods

Reaction
Combinations
CitM.R. V.P .. rate

-- ----

+ +

+
+ +

Usually

Esch. coli

"

"

Mixture

+ - -> Intermediate
strain
, Mixture
-,+ +
+
mixture
+ -- - -1-, Always
Mixture
+
+
- + , + A. aerogenes
- - + + "
"
- - - + Extraneous form
T

-i-

Common
Source

Applies to Pure
Strains Only

Occasionally

Predominating in feces, about 50% of

group in sewage.
Non-members of group Minority in feces.
Minority in soil and
Intermediate strains
sometimes considered
sewage, rarely feces.
non-typical E. coli
Mixtures or slow-react- Soil, minority in sewing A. aerogenes
age and feces.
Atypical
Soil, sewage.

A. cloacae

Mixture
J

A. 'aerogenes

Soil, minority in sewage-and feces.

Majority in soil and
on vegetable~.
Up to 50% of total
group in sewage.
Minority in feces.
'"

"The latest edition of 'Standard Methods,' now sponsored by both the

American Public Health Association and the American Water Works Association, has been greatly changed and improved by allowing conffrmation of the
coli-aerogenes group from gas formation in lactose broth by transplanting to
several specific liquid media for confirmation by gas formation. These four
media are Brilliant-Green lactose bile, Crystal~ Violet lactose broth, Fuchsin
lactose broth, and Formate Ricinoleate broth. Confirmation may also be
obtained by typical reaction and appearance on Endo agar or Eosin MethyleneBlue agar.
'
'" Standard Methods' prescribes the following:
"The standard tests for the coli-aerogenes group shall be either the Presumptive, the Confirmed, or the Completed test. .
I
"Presumptive Test;-The formation of gas irr.:.a- standard lactose broth
fermentation tube within 24 hours at 37 C. or' the-formation of gas during a

BACTERIOLOGICAL EXAMINATION OF WATER

2127

second 24 hour period with confirmation as described in the next par;:tgraph

is presumptive evidence of the presence of memb,ers of this group.
" Confumed Test.-The formation of gas within 48 hours at- 37 C. in a
specified liquid confirmatory medium seeded from a lactose broth tube in which
gas is formed, or the appearance of aerobic lactose splitting typical colonies on
a specified solid confirmatory medium made from a lactose tube in which gas
is formed, confirms the presumption that gas formation was due to the presence
of the coli-aero genes group.
" Completed Test.-To complete the demonstration of the presence of this
group, it is necessary to show that one or more aerobic plate colonies consist of
Gram-negative N on-spore-forming bacilli, which, when inoculated into lactose
broth, form gas."
When it is desired to differentiate between the coli and the aerogenes sections
of the coli-aerogenes group, the detailed procedure sllall not follow primary
planting'in liquid media, but shall be based upon primary planting in solid
media.
'
Presumptive Test
1. Inoculate a series of lactose broth fermentation tubes with appropriate
graduated quantities of the water to be tested. Each tube must contain at
least twice a~ much medium as the portion of sample to be tested.
2. Incubate at 37 C. for 48 hours unless gas appears earlier. Examine
each tube at end of 24 hours and, if no gas appears then, again at end of 48 hours.
Record presence or absence of gas formation at each examination. More
, detailed records of amount of gas formed are not necessary for carrying out the
standard tests.
3. Formation within 24 hours of gas in the inverted vial in the fermentation
tube constitutes a positive presumptive test.
4. If no gas or only a small bubble is formed in 24 hours, incubation shall
be continued to 48 hours. If gas in any quantity is present after the second but
not the first 24 hour incubation period, the test is considered as doubtful and
the presence of organisms of the coli-aerogenes group should be confirmed.
Confirmed tubes may then be given the same consideration in the presumptive
t~st as those showing gas formation in the first 24 hour period.
5. The absence of gas formation after 48 hours incubation constitutes a
negative test (the exclusion of occasional slow gas-forming organisms is considered immaterial).
Confumed Test
The use of Endo or Eosin Methylene Blue plates or one of the following
liquid confirmatory media is permitted: Brilliant-Green lactose bile, CrystalViolet lactose broth, Fuchsin lactose broth, Formate Ricinoleate broth. The
order in which these media are named does not signify their relative values.
For confirmation all are equivalent. Selection of anyone should depend upon
correlation with a series of completed tests.
1. Endo or 'Eosin Methylene-Blue Plates.-Streak one or more plates from
a tube showing gas formation in lactose broth from the smallest amount of
water tested. It is desirable to make transfers as soon as I?ossible after gas

2128

BACTERIOLOGICAL EXAMINATION' OF WATER

I

formation occurs. In order to obtain typical results i,t is essential that plates
be streaked so as to insure presence of some discrete colonies. Transfers
should be made not later than the end of the first 24 hour incubation period if
gas is formed during this time. If at the end of 48 hours gas is formed in tubes
containing less of the sample of water than at 24 hours, transfers should be made
from these tubes. (If larger amounts are not transplanted for confirmation,
they should be recorded as confirmed, even though the smaller portions may
fail subsequently to confirm.)
Incubate the plates at 37 C. for 18 to 24 hours.
If typical colonies develop upon the plates within this period, the confirmed
test may be considered positive. If, however,. no typical colonies develop
within 24 hours, the test cannot be considered definitely negative and' the
completed test should be made.
2. Liquid Confirmatory Media.-Transfer from a lactose broth tube showing
gas to the fermentation tube containing the selected confirmatory mediulll.
It is recommended that all lactose broth tubes showing gas be confirmed to
comply with this test, but it is allowable to follow the procedure outlined for
solid media. Transfers should be made as soon as gas appears. In routine
testing it is convenient to make observations and transfers at the end of 24
and 48 hours. In making transfers the tuqe showing gas shall first be gently
shaken or mixed by rotating and the transfer shall be made by mMns of a wire
loop not less than 3 mm. in diameter; 9r it is permissible, to use the sterile
capillary tube or sterile pipette when it is considered desirable to transfer larger
amounts:
Incubate the inoculated c~nfirmatory tubes for 48 hours at 37 C.
The formation and presence of gas in any amount in the. inverted vial in
the fermentation tubes at any time within 48 hours constitutes a confirmed test.
Completed Test
Streak one or more Endo or Eosin Methylene-Blue plates from the lactose
broth tubes which show gas from the smallest amomit of water tested or from
the secondary selective medium tubes which ~how gas. If these plates are
streaked from the original lactose broth tubes they should be made at -the
same time that the secondary selective medium tubes are planted. If plates
are used for the confirmed test, completion may be continued after incubation.
Incubate.the plates at 37 C. for 18 to 24 hours.
.
Identification.-From the plates fish one or more typical colonies, or if no
typical colonies are present, fish two or more colonies considered.. most likely to
be organisms of the coli aerogenes group, transferring each to an agar slant and
,a lactose broth fermentation tube. The lactose broth fermentation tubes thus
inoculated shall be incubated until gas formation is n,oted, not to exceed 48 hours.
The agar slant shall be incubated at 37 C. for 24 hours, when a microscopic
examination for spores and a Gram stain of at least one culture shall be made,
selecting when possible the one which corresponds to one of the lactose broth
fermentation tubes that has shown gas formation. If none of the lactose broth
fermentation tubes contain gas at the end of 24 hours, all agar slants shall be
examined, and corresponding tubes examined o.n-tlre 'following day.
Results.-The formation of gas in lactose broth and the demonstration

/'

BACTERIOLOGICAL EXAMINATION OF WATER

2129

of Gram-negative non-spore-forming bacilli in the agar culture shall be considered a satisfactory completed test.
The absence of gas formation in lactose broth or failure to demonstrate
Gram-negative non-spore-forming bacilli in a gas-forming culture constitutes
a negatIve test.
When spore-forming lactose fermenting organisms are found, the culture
should be further studied to ascertain the possible presence of the coli aerogenes
group with the spore-bearing organism. This may be done by transferring the
culture to Formate Ricinoleate broth and incubating 48 hours at 37 C. If
no gas is produced only spore-forming lactose fermenters are present. If gas
is produced members of the coli-aero genes group may be present. If gas is,
produced, the presence of coli-aero genes organisms should be verified by inoculation of a tube of standard lactose broth and an agar slant. If, after 48 hours,
gas is produced in the former and no spores are present in the latter, the test
may be deemed" completed." If spores are present, for practical purposes the
organisms of the coli-aero genes group may be considered absent.
Selection of Coli-Aerogenes Tests
.

The laboratory worker shall be guided by the following basic considerations:

The Completed Test shall be applied to the smallest gas-forming portion or
portions of

_ - any sample of drinking water upon which sufficient data are not
available to justify the application of the Confirmed Test.
_ - any sample of drinking water upon which the previous records indicate
the inapplicability of the Confirmed Test.
_ - any sample of drinking water which is being examined with reference
to the U. S. Treasury Standard.
...
The Confirmed Test may be applied to the smallest gas-forming portion or to
all portions showing gas, from
_ - any sample of raw water, water in process of purification, or water
prepared for human cop-sumption where a sufficient number of prior examinations have, established a satisfactqry correlation of the Qonfirmed Test with the
Completed Test.
_ - miscellaneous samples from any other source where it is known that
the Presumptive Test is too broadly inclusive.
The Presumptive Test may be applied to the smallest gas-forming portion
or portions
_ - in examination of sewage, sewage effluents, or water showing relatively
high pollution, where fitness for use as drinking water it: not under consideration,
and
.
- - in the routine examination of raw waters in purification plants, provided that records indicate that the Presumptive Test is not too inclusive for
the. production of data statistically comparable to that obtained upon the
finished water.
The above procedure is ,too troublesome' for a laboratory handling large
numbers of routine samples. The work is too great and the time too long for
the value of the results. It was adopted because of the fact that B. welchii,
a spore-forming group (though probably of ,feca~ origin), also fo~m~ gas, _in

2130

BACTERIOLOGICAL EXAMINATION OF WATER

I

lactose media. B. welchii is obligate anaerobi.c, gram positive, esculin negative,

and a spore-former. It may easily be differentiated from B., coli by its appearance in a hanging drop. (Prepare hanging drop by placing a drop of salt solution, 0.85% sodium chloride, on a sterile cover glass and, mix with a minute
portion of the culture, taken direct from the fermentation tube. Invert this
cover slip over a hollow slide, sealed by a ring of vaseline, and' examine under
the microscope.) It is a large, non-motile bacillus, occurring in chains particularly when grown in lactose bile media, as mentioned in the 1912 edition of
"Standard Methods." In it few instances aerobic spore-forming bacilli have
been described which fermented lactose broth, but such instances are extremely
rare, and they have not been found in lactose bile.
In 1920 the lactose bile medium was improved by the additions of BrilliantGreen which dye eliminates spore-formers of the B. welchii type. Originally
the proportion of dye was 1/10,000 in 5% bile media. Extensive work by
Committees of the American Water Works Association and the American
Public Health Association has resulted in the development of a somewhat mor~
delicate medium by using 1/75,000 Brilliant-Green in 2% bile as the final dilutions after additjon of the water to be tested. The Department of Water
Supply Laboratories of New York City have since establishment in 1897 relied
upon presumptive tests for the B. coli group without confirmation to safeguard
the supply. -The media used have been improved from time to time . Since
1920 Brilliant-Green bile has been used-bec_ause of its eliminlrtion of B. welchii.
Beginning July 1st, 1933, the new formula recently standardized, 2% bile and
1/75,000 Brilliant-Green, has been used. Requirement of 10% minimum
amount of gas formed has been' retained for two reasons: first, possible inversion
of a small amount of lactose to dextrose during autoclave sterilization and,
second, the fact that vigorous forms of B. welchii may produce as much as 10%
gas in this medium containing such a reduced quantity of Brill}ant-Green.
Brilliant-Green bile gives a true comparative picture of the quality of the
water supply from the watershed to the distribution system. In freshly
polluted waters it eliminates practically none of the c,6li-aerogenes group and
may surpass lactose broth. In stored and chlorinated waters attenuated forms
of B. coli are eliminated. The author lays great sanitary stress upon this fact.
Numerous investigations including tho'se at Mt. Prospect Laboratory and those
of the late Alexander Houston of London have ~hown that in two weeks time
90-99% of B. typhi contamination of water is eliminated and in two months
time practically all B. typhi. The rapidity of gas formation-is'largely governed
by attenuation of the bacilli. Gas formation on the second day with' none or
only a small amount on the first day indicates generally contamination beyond
two weeks previous and. three days' requirement for 10% gas may mean as
much as a two months interval. The presence of the coli-aerogenes group may
mean presence of animal as well as'human contamination. There is no method
of distinguishing between the two sources. The pres~nce of human contamination means only the possibility of typhoid bacteria since less than 1 % of the
population are typhoid carriers. There is therefore a very large factor of
safety in judging the sanitation of a water from the presence or absence of the
coIi-aerogenes group and elimination of attenuated forms of B: coli is decidedly
important1. Numerous comparisons between pres1imptive and completed
tests have been made so that possible error is known. An enormous amount of

BACTERIOLOGICAL EXAMINATION OF WATER

2131

work is eliminated by the above presumptive use of Brilliant-Green bile, in

addition to saving of media, equipment, incubator space, and force. That the'
above procedure has been safe.in the case of the City of New York is shown by
the extremely low typhoid death rate of 0.3 (1937) per hundred thousand
population, and the fact that since 1907 particularly since starting of chlorination no cases of typhoid fever have been traced to the water supply nor has it
been under suspicion.
In testing for B. coli 0.1 mI., 1 mI., and 10 ml. portions of water are introduced to the bottom of the bile tubes, or in polluted ,vaters quantities varying
by decimal multiples or fractions until a negative test is obtained. At important chlorination stations 100 ml. portions of water are examined. Sewage,
.will usually show a positive test in 0.000001 ml. The plantings are made at
the same time that the plates are prepared for total counts. The tubes are
incubated at 37 C. for three days. 10% of gas within three days-is a positive
test. Tubes showing no gas in two days are discarded ..

Examination for B. Typhi, etc.

Hesse agar is retained in the text as a possible simple method of isol~tion of
B. typhi from water supply because it was successfully used to accompli'lh this
. purpose several times following its development. Thill was true by use of even
small ~mounts of water such as 1 ml. Preliminary use o(bile medi& precedes
the test on the agar.
.:( .. , "
.~ The water laboratories of London, England, hav!l. succe!,sf_ully il>olated
B. typhi aI).d B. paratyphi-B .from sewage and water by m!lthods.described in
the annual reports of former Director of Laboratories Alexander Houston.
The reports should be consulted for details, particularly, the 26th Annual
Report, Metropolitan Water Board for 1931, pages 26,27, and 31, for isolation
of B. paratyphoid-B.
This test is only mJLtie in cases of special investigation.
,',.
Direct test for B. typhi may be made from the bile tubes or from larger
quantities of water incubated in bottles of bile. 8 Dilutions from the inO.culated
bile medIa are ?nixed with 10 ml. melted Hesse agar (see Preparation of Media)
in petri dishes with porous tops. The poured plates are hardened in the ice-box
and then incubated 24 hqurs at 37 C. The colony is characteristic only when a
few colonies are on a plate. They are large size, have a dense white nucleus,
then translucent zone, then dense white seam. The colonies are perfectly
round. T:hey retain their appearance if placed in the ice-box, and as they
develop further produce another translucent area and another seam" etc.,
giving the appearance of concentric rings. Other large colonies of other 'species
usually become dense throughout. Fornis most likely to int~rfere are B. paratyphi, B. pyiJcyaneous, B. jluorescens liquefaciens, and some~ very motile forms
'
of B. coli.
From the edges of the colony direct tests should be made for 'B. typhi and
B. pata-typ~i by specific agglutination tests.
Transplants may be made to Endo media instead of Hesse agar and the
colorless colonies picked for the identifying tests.
S

In th,e latter case it is preferable to omit lactose from the bile media.

2132

BACTERIOLOGICAL EXAMINATION OF WATER

Widal Test.-To make the Widal test place a small portion. of the culture
. in a drop of normal salt solution on a cover glass, and invert it over a rubber
ring on a glass slide, using vaseline on the edges so a& to prevent evaporation
and consequent movement by currents in the drop. \
If the bacteria are motile rods resembling B. typkj ~add a drop of equal size
of highly diluted anti-typhoid serum (the latter diluted with physiological salt
solution). If in the course of half an hour the bacteria cease their motions
and agglutinate, the presence of B. typhi is practically established. In rare
cases, some strains of B. coli and allied species may respond to the Widal test,
but almost invariably in dilutions of 1-50 or less. The high-power dry lens is
best for this work.
To determine the point of highest dilution at whIch agglutination take.
place the Widal test is best made in tubes and examined macroscopically.
Place a series of small test tubes in a rack and pour into each varying dilutiom
of anti-typhoid serum. To each tube add an equal amount of 24-hour brQth
culture of B. typhi and incubate at 37 C. for at least three hours. The highest
dilution at which precipitation takes place is the highest point of agglutination.
Russell Media.-As additional rapid means of differentiating betweenB.
typhi and B. para-typhi A and B, needle stabs may be made into tubes of Russell
media (see Preparation of Media) and incubated at 37 C. The reactions are
based on the fact that the bacilli growing aerobically, that is on the slant, only
utilize the carbohydrate when present in amounts over 0.1 %, whereas, growing'
anaerobically stabbed in the butt, they must utilize the carbohydrates for their
oxygen supply and therefore ferment the trace of glucose present. The indications of the Russell media reactions are as follows.:
B. coli Group-Russell media-Slant, acid
(Glucose, acid and gas)
Butt, acid and gas (Lactose, acid and gas)
B. typhi Group}
{Slant unchanged (Glucose, acid only)
B. dysentery Group Russell media Butt, acid only (Lactose, not
;
fermented)
B. paratyphi Group}
{Slant unchanged (Glucose, aci~ and
.
gasf
Group Russell medIa B u tt ,aci'd an d gas (L ac t ose, no t
B . en ten't'd'
~ ~s
.
fermented)
B. alkaligenes Group-Russell media-Slant unchanged (Glucose, not
-ferme_pted)
Butt unchanged (Lactose, n<;>t
fermented)
In distin,guishing between typhoid and dysentery (true or Shiga type)typhoid bacilli are motile, dysentery non-motile, typhoid produces acid in
mannite, dysentery does not.
I
There also exist three types of para-dysentery (n'on-motile): Type I (ParkHiss), type II (Flexner) and type III (Strong). Para-dysentery are different
from true dysentery in that they produce acid with mannite (the same as
typhoid) but they all give positive indol, while dysentery and typhoid do not.
Indol is sometimes doubtful with Flexner. The t...hi.ee para-dysentery groups
are further distinguished among themselves by "their reactions upon maltose

BACTERIOLOGICAL EXAMINATION OF WATER

2133

and saccharose. The type I produces acid in neither, type II produces acid
in maltose but not in saccharose, type III produces acid in saccharose but not
in maltose. The usual final test for all these different types of typhoid and
dysentery is the agglutination with specific serums: The sugar fermentation
tests must be made upon freshly isolated cultures since after artificial cultivation
Park-HJss strains may ferment maltose and Flexner strains saccharose as
demonstrated by Hiss and by Lentz.
The members of the ~yphoid, para-typhoid, dysentery and para-dysentery
groups all form colorless colonies upon Endo media. This is frequently the
starting poiht for isolation.
'
REFERENCES
"Standard Methods of Water Analysis." Am. Pub. Health Assoc., Editions 1905,
1912, 1917, 1920, 1923 and 1936. (Contain extensive bibliography.)
Park and Williams, "Patbogenic Microorganisms."
Prescott and Winslow, "Elements of Water Bacteriology."
Chester, "A Manual of Determinative Bacteriology."
Bergey's "Manual of Determinative Bacteriology."
"The Determination of Hydrogen Ions," Clark.

.1

MISCELLANEOUS ANALYSES *
ANALYSIS OF ACETONE 1
PURITY
Reagents.--N/20 Na 2 S 2 0S is made up and standardized carefully against
K 2Cr 2 07. For accurate work standardized burettes must 'be used and burette
corrections and temperature corrections applied.
N/I0 Iodine is made up and is standardized at the time of making a determina.tion The standardization consists in titrating against the N /20 Na 2S 2 0S
50 ml. of iodine mixed with the same amounts of NaOH and H 2S0 4 as are used
in the determination of acetone. The titrations are made in triplicate and the
average of the three wHh burette and temperature corrections is used as th~
blank in making the calculation.
Normal NaOH is made up and standardizep against 2 N sulfuric acid.
2 N Sulfuric Acid is made up and checked against 1 N NaOH and the
strength adjusted so that 25 ml. of the acid will neutralize 50 ml. of the NaOH.
Starch Solution is prepared according to' standard procedure.
Preparation of Sample.-Approximately 1.4 grams of acetone are carefully
weighed by one of the following methods and made up to a volume of one liter.
(a) Glass ampules %" to ~" in diameter are blown and after filling with
acetone are sealed. The exact weight of sample is determined by the difference
in weight of the empty and filled ampule. Samples varying in weight more than
plus or minus 0.1 gram from 1.4 gram should be rejected. The. glass ampules
are dropped into 500 m!. glass-stoppered bottles containing 200 ml. of boiled
distilled water and broken by shaking. The capillary ,stem is then {)arefully
crushed with a glass rod to allow thorough washing outfof the acetone and the
sample and combined washings made up to one liter at 20 C.
(b) A quicker, easier method is to weigh out 1.75 m!. of acetone from a 2 ml.
pipette graduated in tenths of a ml. into a small glass-stoppered weighing bottle,
conical shape, capacity 12 to 51 ml. (as listed in Elmer and Amend's catalogue.
- This volume will weigh approximately 1.4 grams. The weighing bottle is
inverted and the mouth held under the surface of boiled distilled: 'water in a
liter beaker.' The stopper is then removed under water by means of a glass
roa with a hook on the end. The bottle is thoroughly washed out and the water
and acetone are transferred to a one liter flask and made up to the mark at 20 C"
Procedure.-Three 25 m!. aliquots from each of 3 ~asks made up according
to (a) or (b) under Preparation of Sample are pipetteq into 750-ml. Erlenmeyer
flasks containing 50 m!. of the 1 N NaOH solution. A Bureau of Standards
pipette is used for taking the samples from the flasks. Into each flask is run
with constant shaking 50 ml. of the N /10 Iodine. The flasks are then stoppered
and allowed to stand at about 20 C. (cooling necessar_y},!l wanD weather) for at

* Edited by N. H. Furman.
l'

Method of Commercial Solvents Corporation.

2134

MISCELLANEOUS ANALYSES

2135

least 20 minutes. At the end of this time 26 ml. of the 2 N sulfuric acid are
poured into the first flask which is then titrated immediately against N /20
Na2S~03. ,The solution must be acid before any thiosulfate is added, otherwise
the determination is ruined. When the color of the iodine has almost faded
out, starch solution is added and the titration continued until the blue color
disappears and the bright 'yellow color of the iodoform suspension is reached.
The mls. of thiosulfate are read, the burette and temperature corrections are
applied and the corrected reading subtnfcted from the average corrected blank.
Percentage of dimethyl ketone-by weight is then calculated from the following
formula:
Weight of acetone taken for dilution
= % CO (CHah by weight.
The following precautions must be observed:
1. Only boiled distilled water should be used in preparing the solutions and
throughout the determination.
2. Standardized glassware must be used and burette and temperature corrections must be applied to all readings.
3. Weight of sample taken and amount of iodine solution used must be as
specified. With these amounts the 35 to 40% excess of iodine, found to be
optimum for completion of the iodoform reaction, will be present. Use of the
small glass-stoppered bottle described in (b) under Preparation of the Sample
with the 2 ml. pipette insures uniformity in the weight 'of sample taken.
4. The iodine solution must be .added slowly and with constant shaking.
The Bureau of Standards calibrated burettes with fine tip and delivery time of
approximately 2 minutes have been found very satisfactory.
5. A slight excess of acid must be added just before the titration with thiosulfate. Too great an excess will give low results.
Specific Gravity.-The specific gravity is determined at 20/20 C. using a
standardized pycnometer.
'
Acidity.-Add 50 m!. of the sample to 50 m!. of neutral distilled water in a
300 ml. Erlenmeyer flask and titrate with N 1100 NaOH to a faint pink coloration using phenolphthalein as indicator.

'd
CO
m!. NaOHXNX2.2_01
2,
39.7
_ /0 aCl as
where
N =Normality of NaOH,
39.7 = Weight of 50 ml. sample eSp.gr .. 795).
Water.-Shake together one volume of the sample and three volumes of
benzene in a test tube. Any water present will be precipitated by the benzene
thus imparting a turbidity to the solution in the tube.
Color.-The sample is compared with a sample of distilled water. It
should be absolutely colorless with no cloudiness or solid matter.
Residue.-Evaporate 25 fil. of the sample to dryness in a tared porcelain
or preferably a tared platinum crucible over a water bath. Dry to constant
weight at 110 C. in an oven.
a

2136

MISCELLANEOUS ANALYSES
Weight of residue
.
..
WeIg
. ht 0 f sampIe X 100 = % ResId ue on Ev~ poratIOn.

Miscibility with Water.-A sample of acetone is shaken with ~arious proportions of water. Report any turbidity and the rati~ of acetone to water
which gave it.
.
,
KMn04 Test.-1 ml. of 0.1% KMn04 solution is added to 100 ml. of the
sample in a 200-ml. Florence flask and allowed to stand at 15.5 C. Report
the time required for the permanganate color to fade. Allow to stand at least
5 hours, if fading does not come sooner.
DETECTION OF ALDEHYDES IN ACETONE

Reagent.-Add a 10% solution of ammonia to a 5% solution of silver nitr~te

until the precipitate which is fo.rmed just redissolves. Keep in dark bottle.
Procedure.-Add 10 ml. of the sample to be tested to a test tube or small,
tall form, weighing bottle and add to this 5 ml. of the above reagent. Shake
and place immediately in a dark closet. It is necessary to shake the test tube
vigorously every five minutes during the test. When viewed against a white
background any precipitate or dark color after twenty minutes in the dark
closet would indicate the presence of aldehydes.

ACETONE

ANALYSIS OF ACETnNE BY MESSINGER'S METHOD

An amount of acetone in aqueous solution, equiv'alent to 30-40 mg., is
pipetted or added to 50 ml. of N sodium hydroxide solution contained in a
glass bottle which can be closed with a ground glass stopper. After standing
for 5 minutes, about 25% excess of a 0.1 N solution of iodine is run in from a
burette with continual shaking. It is essential to shake proPilrly or to keep the
liquid in continuous rotation. The excess of iodine is needed to complete the
reaction. 2 The bottle is then stoppered and the solution allowed to stand for
at least 10 minutes (20 minutes in cold weather).
.
Twenty-five ml. of 2 N sulfuric acid is then added from a measuring cylinder,
0.3-0.4 ml. being added in excess of the amount found necessary to neutralize
the 50 ml. of c;ustic soda solution. A 0.05 N solution of sodium thiosulfate is
then added from a burette until the yellow color just ~emaihs visible. Freshly
prepared starch solution is now added and tlie titration finished. 1 ml. of
0.1 N iodine =D.96747 mg. of acetone.
If a larger excess of sulfuric acid is added, too much thiosuate is required,
and the real amount of iodine solution required is thus reduced (see below).

__.-

L. F. Goodwin, J. Am. Chern. Soc., 42, 39 (1920).

M~SCELLANEOUS ANALYSES

2137

If the bottle is not shaken vigorously while adding the iodine solution, the
iodine cannot act completely on the acetone.

ALCOHOL
DETECTION AND ESTIMATION OF ALCOHOL
Detection of Alcohol.-Ethyl or" grain alcohol" C 2 H.OH boils at 78.4 0 C.
so that it readily distills with steam. If an appreciable
amount of alcohol is present in the liquor it may be detected by the odor of the solution or in the first fraction of
its distillate.
Iodoform Test.-A few m!. of the liquor are distilled and
to this distillate is added a small crystal of iodine or a few
ml. of aqueous solution of iodine-potassium iodide solution,
followed by sufficient potassium hydroxide to give the
liquid a distinct yellow to brownish color. The solution
is warmed gently. If alcohol is present a whitish to lemon
yellow precipitate of iodoform will appear., The odor of
iod,oform may be noted. If the crystals forrn-.slowly they
are deposited in form of perfect stars and hexa~nal plates.
This test is not conclusive of 'alcohol as other organic
materials also give Iodoform.
Berthelot'S Test.-The distillate of the liquid is vigorously agitated with a few drops of benzoyl chloride and 4
or 5 drops of lO%'solution of sodium hydroxide, until the
penetrating odor of benzoyl chloride has disappeared. The
characteristic odor of ethyl benzoate will be detected if
alcohol is present in the liquid tested.
Reaction: C 2H 50 H + C 6 H 5 COCI + KO H = C 6 H 5
CO OC 2H.+ KCl+ H 20.
Ethyl Acetate Test.-The liquid (distillate) is mixed
FIG. 297.
with the same volume of 'concentrated sulfuric acid. A very
Pycnometer.
small quantity of anhydrous (fused) sodium acetate is
added and the mixture warmed. The odor of ethyl acetate will be recognized
if alcohol was present in the liquid tested.
Reactions: C2H50H+H2S04=C2H50S020H+H20.
C2H50.S02.0H+CHa.CO.ONa=CHa.CO.OC2H5+Na.HS04.
DETER~nNATION

OF ALCOHOL

In carbonated liquids the carbon dioxide should first be expelled by pouring

the liquor back and forth from one beaker to another, or by vigorously shaking
,

2138

MISCELLANEOUS ANAL Y:El.ES

the sample in a large sepatatory funnel and drawing off the lower portion, repeating this several times if necessary. After treatm~nt the liquid should be
free from foam.
Distillation Method.-This is the most accurate m~thod for determination
of alcohol. A convenient quantity of the sample is hccurately measured' or
weighed, ~ccording to whether the per cent by weight or measure is desired.
The weight taken varies from 25 to 100 grams, according to the alcohol content.
of the sample.
The sample is placed in a 250 to 400 ml. round-bottom flask and diluted to
150 ml. The flask is connected to a condenser and 90 to 95 ml. of the liquid is
distilled into a narrow-necked flask. The distillate is made to exactly ioo ml.
with distilled water.
.
.
The specific gravity of the thoro~ghly mixed distillate is taken by means of
the pycnometer or by the Westphal balance, the temperature of the liquid
being exactly 15.6 C. From the specific gravity the corresponding percentage
of alcohol by weight or by volume, or the grams per 100 ml. in the distillate,
is ascertained from the accompanying tables.
The percentage of alcohol by weight in the sample is obtained by multiplying the per cent by weight in the distillate by the weight of the distillate,
and dividing by the weight of the sample taken. The per cent by volume irt
obtained by multiplying the per cent by volume in the distillate by 100 and
dividing by the volume of the sample taken.
Evaporation Method.-Should a distillation apparatus not be available the
following method may be used. The specific gravity of the original sample is
determined, the temperature being regulated to 15.6 C. A measured portion
of the liquor (50-100 ml.) is placed in a porcelain dish over a water bath and
three fourths of its volume evaporated off. The concentrate is diluted to
exactly its originllJ volume and its specific gravity again taken at 15.6 C. TQ,
apply the table add 1 to the or.iginal specific gravity and from this total subtract the second specific gravity. The difference is the specific gravity corresponding to the alpohol in the liquor. For example,
if'the
specific gravity of the'
,
I
original solution was 0.9989 and the de-alcoholized sample was 1.0005, then.
1.9989 -1.0005 = .9984 = 1.06% alcohol, by volume.
If the liquor is free from residue its specific gravity may be taken directly
without distillation and the alcohol content determined, provided it is a mixture
of water and alcohol.

ANALYSIS OF GRAIN ALCOHOL (ETHYL ALCOHOL, SPIRITS OF

WINE, COLOGNE SPIRITS)
Per cent Ethyl Alcohol.-The method of determining per cent of alcohol
by the distillation of the alcohol from liquors has already been discussed. In
straight grain alcohol the specific gravity may be taken directly by means of
the Westphal balance 'or by means of the pycnometer and the amount of
alcohOl by weight or volume obtained from the tables on pages 2142-6.
Proof.-This is obtained by multiplying the' per cent ,by volume of the
ethyl alcohol by 2.
_ ,
Non-volatile Residue.-100 ml. of the sample; in a large platinum (tared)
dish, are evaporated on the water bath to a moist residue. The evaporation to

MISCELLANEOUS ANALYSES

2139

dryness is accomplished in a water oven for two and a half hours at 100 C.
The weight of the residue divided by the specific gravity gives per cent of
non-volatile residue.
Acidity Expressed as Acetic Acid.-10 ml. of the alcohol are titrated with
N /10 NaOH in presence of phenolr.hthalein indicator.
1 m!. N/lO NaOH=0.006 gram acetic acid.
Distilled Solution for Determining Aldehydes, Furfural and Esters (A). To
250 ml. of the sample~30 ml. of water are added and the liquor distilled into a
250-m!. volumetric flask. When the distillate has almost reached the mark
the distillation is discontinued and the volume made up to mark by addition of'
water. The distillate contains all the aldehyde, furfural and esters. 1 ml. of
this solution is equivalent to 1 m!. of the original sample.
ALDEHYDES
QUALITATIVE EXAMINATION

Alkaline Silver Nitrate Reagent.-Three grams of AgNO a are dissolved in

a little water in a 100-ml. volumetric flask, 3 grams of pure NaOH are added
followed by 20 nil. of strong NH 40H and the whole made to 100 ml.
Test.-Ten ml. of the sample are diluted with an equal volume of water in
a glass stoppered bottle, 1 m!. of the alkaline silver reagent added, and the
bottle is closed and placed in a dark chamber for an hour. The liquid is now
filtered and the filtrate made acic;l with HNO a and a few drops of HCI added.
A precipitate of AgGI indicates the non-reduction of the silver salt and consequently a negligible amount of aldehyde in the sample.
QUANTITATIVE DETERMINATION OF ALDEHYDE

Reagents: Alcohol-Aldehyde-jree.-To 1.5 liters of 95% ethyl alcohol

are added 25 grams, of N aOH in a large distilling flask and all but 100 m!.
distilled over. To this distillate are added 2.5 grams of meta-phenylenediamine
hydrochloride, the sample being placed in a large flask with reflux condenser.
After heating on a steam bath for several hours the solution is distilled down
to about 200 m!. The distillate (rejecting the first 100 m!.) is placed in a
stoppered bottle for use.
, Fuchsin-Sulfite Solution.-To 0.5 gram of fuchsin dissolved in 500 ml.
of water are added 5 grams of S02 dissolved in water. (100 ml. H 20 saturated
with S02 at 20 C. contains 11.29 g. S02, at 15.C.=13.5 g. SOr-Seidell.
Determine by titrating with N /10 iodine sol. 1 m!. = .0032 g. S02.) Dilute
to 1000 ml. and allow to stand until colorless. The solution retains its strength
only a few days, so should be made up in small quantities at a time.
Standard Acetaldehyde Solution.---:Five grams of aldehyde ammonia are
purified by repeated extraction with ether, grinding the salt with the ether in
a mortar and decanting off the ether. The salt is dried by blowing air over
it and finally placing it in a vacuum desiccator over concentrated H 2S0 4
1.386 gram of the purified salt is dissolved in 50 m!. of the 95% alcohol reagent
prepared as stated above. To this are added 22.7 m!. of N alcoholic H 2S0 4

2140

MISCELLANEOUS ANALYSES

(49.04 grams H 2S0 4 mad~ up to 1000 ml. with 95% alcohol) arfd the solution
made to 100 ml. with 95% alcohol. To c6mpensat'i for the (NH 4)2 S04 precipitated 0.8 ml. alcohol is added. After standing all night the solution is
filtered. 100 ml. of this reagent contains 1 gram of rcetaldehyde. It retains
its strength
.
I
Reagent for Standard.-Two ml. of the above sqlution diluted with ))0%
alcohol to 100 ml. 1 ml. contains 0.0002 gram acetaldehyde. Make up Mesh
for use.
PROCEDURE FOR ALDEHYDES

Ten ml. of the solution A are diluted to 50 ml. with the aldehyde free alcohol
(diluted to 50% by volume), 25 ml. of fuchsin reagent are added and the mixture
allowed to stand for 15 minutes (all reagents and the sample should be at 15 c.f.
The solution is now compared colorimetrically with standards made up in the
same way. Nessler tubes may be used. Comparison may be made III a
colorimeter. From this the per cent aldehyde is readily calculated.
DETERMINATION OF .,FURFURAL

Standard Furfural Reagents.-One gram of freshly redistilled furfural is

dissolved in 100 ml. 95% alcohol (furfural free). This solution keeps.
One ml. of the strong solution is diluted to 100 ml. with 50% (by volume)
alcohol. 1 ml. =0.0001 gram furfural. /
Procedure.-Twenty ml. of the distilled solution (A) are diluted to 50 ml.
with 50% (by volume) of furfural-free alcohol. To this are added 2 inl. of
colorless aniline and 0.5 ml. of dilute HCI (5 : 4) and the. mixture placed in' a
water bath at 15 C. for fifteen minutes. Colorimetric comparison is now
made with standards containing known amounts of furfural, the solutions being
prepared in the same way as the sample.
DETERMINATION OF ESTERS EXPRESSED ASj ETHYL ACETATE

:Fifty ml. of the distilled solution (A) in an Erlenmeyer flask are exactly
neutralized with N/10 NaOH (phenolphthalein indicator) and about 50 ml.
excess added, the exact amount being recorded. The solution is boiled for an
hour with reflux condenser. After cooling, the excess of alkali is titrated with
N /10 acid. The ml. N /10 N aOH consumed in the sapQnification, multiplied
by 0.0088, gives the gram weight of esters calculated as ,e-thyl,acetate. This
weight divided by the specific gravity of the sample and multiplied by 2 gives
per cent.
DETERMINATION OF FUSEL OIL

Fifty ml. of the original sample, in an Erlenmeyer flask, are diluted with an
equal volume of water and 20 ml. of N 15 NaOH added. The mixture is now
saponi.fied by boiling for an hour with reflux condenser. The flask is connected
to a condenser and about 90 ml. distilled into a deep beaker. 25 ml. of water
are added to the flask and the distillation continued until the total volume
of the distillate is 115 ml. Finely ground salt (N_.],OI) is ,now added until the
solution is almost saturated and a saturated solt'tion of N aCI added until the

MISCELLANEOUS
ANALYSES
,

2141

specific gravity of the so~ution is 1.10. The solution placed in a separatory

funnel is extracted four times with carbon tetrachloride, CCI 4, using 40, 30, 20
and 10 ml. CCl 4 respectively. To the extract in a separatQry funnel 10 ml. of
KOH solution (1 : 1) are added. This solution is cooled in ice water to 0 C.
and at the same time 100 ml., accurately measured, of KMn04 solution (!:O
grams per liter) in a flask. When cooled to 0 C. the bulk of the KMnC 4
solution is added to the extract, leaving the rinsing out of the flask until later.
The mixture removed from the ice bath is shaken vigorously for five minutes,
then set aside for half an hour allowing it to warm up to room temperature
(20-25 C.).
Into a liter Erlenmeyer flask are measured accurately 100 m!. of H 2 0 2 scktion (2% stronger than the KMn04 solution) followed by 100 m!. of 2[;%
H 2S0 4 and to this mixture are added slowly the contents of the separatory
funnel, swirling to get an even distribution. (The acid solution should be
constantly in excess.) The rinsing of the separatory funnel and the flask
containing the residue of the KMn04 are added and the excess of the H 2 0 2
determined by titration with standard KMn04 solution (10 grams KMn04
per liter).
'
Blank.-The same amounts of stronger KMn04, KOH, H 2 0 2 and H 2S0 4
are mixed side by side with the sample and the residual H 20 2 determined with
the standard KMn04 to determine the blank. This is subtracted from the
. first titration and the difference multiplied by the value of the standard
KMnO.. If this is exactly 10 grams per liter as determined by titration with
N H 2 C 2 0 4 then1 ml. KMn04=0.696 gram amyl alcohol.
NOTE.-1 ml. N H 2C 20.=0.03161 g. KMnO..
tain 0.01 g. per ml. of the reagent.

(KMn04 10 g. per liter.)

The standard KMnO. should con-

DETECTION OF NITRATES IN GRAIN ALCOHOL

Fifty ml. of the sample are neutralized with N /10 NaOH (phenolphthalein
indicator) and evaporated to a moist residue. A little distilled water is added
1" m!. of phenoldisulfonic acid, and the mix made alkaline with NH 4 0H.
A bright orange color indicates the presence of nitrates. Run a blank on the
water used. See subject of nitrates in the chapter on Water Analysis.
DETERIIUNATION OF SULFUR COMPOUNDS IN GRAIN ALCOHOL

A hundred ml. of the alcohol are made slightly alkaline with N /10 N aOH
(phenolphthalein) and 5 ml. H 2 0 2 added, preferably in a large platinum dish.
The residue is ignited over an alcohol flame, then taken up with 50 ml. of water,
5 m!. dilute HCI added and 5 ml. 10% BaCI 2 The precipitat.ed BaS04 is
filtered off, ignited and weighed according to the customary procedure.
BaS04XO.1373=S.
DETERMINATION OF M~THYL ALCOHOL, WOOD SPIRIT, WOOD NAPHTHA,
WOOD ALCOHOL, ETC.

See Methyl Alcohol in section following the tables on alcohol.

2142

MISCELLANEOUS ANALYSES

PERCENTAGE OF ALCOHOL BY VOLUME XND BY WEIGHT*

GILPIN, DRINKWATER, AND SQUIBB!

Specific
Gravity at
60
ti00

1,00000
0,99984
.99968
.99953
.99937
.99923
.99907
.99892
.99877
.99861
.99849
,99834
.99819
.99805
.99790
.99775
.99760
.99745
.99731
.99716
.99701
.99687
.99672
.99658
.99643
.99629
.99615
.99600
.99586
.99571
.99557
.99543
.99529
.99515
.99501
.99487
.98996
.98984
.98971
,98959
.98947
,98934
.98922
.98909
.98897
.98885
.98873
.98861
.98849
.98837

Alcohol
per cent
by
volume.

per cent
by
weight,

0.00
0.10
0.20
0.30
0.40
0.50
0.60
0,70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2,00
2.10
2.20
2.30
2.40
2.50
2.60
2.70
2,80
2,90
3.00
3.10
3,20
3,36
.3.40
3,50
7,20
7,30
7.40
7,50
7.60
7,70
7.80
7.90
8,00
8.10
8.20
8,30
8.40
8.50

o 00

0.08
0.16
0.24
0.32
0.40
0.48
0.56
0.64
0.71
0,79
0.87
0.95
1.03
1.11
1.19
1.27
1.35
1.43
1.51
1.59
1.67
1. 75
1.83
1.91
1.99
2.07
2.15
2.23
2.31
2,39
2.47
2,55
2.64
2.72
2,80
5,77
5.86
5,94
6.02
6.10
6.18
6.26
6.34
6.42
6,50
6.58
6,67
6.75
6.83

* Bulletin No.

Specific,
Gravity at
60

Grams
per
100 ml.

0.00
0.08
0.16
0.24
0.32
0.40
0.48
0.56
0.64
0.71
0.79
0.87
0.95
1.03
1.11
1.19
1.27
1.35
1.43
1.51
1.59
1.66
1.74
1.82
1.90
1.98
2.06
2.14
2.22
2.30
2.38
2.46
2,54
2.62
2.70
2,78
5.72
5.80
5.88
5 96
6,04
6.11
6,19
6,27
6.35
6.43
6.51
6 59
6.67
6.75

600 F.

I
I:

.99473
.99459
.99445
.99431
.99417
.99403
.99390
.99376
.99363
.99349
.99335
.99322
.99308
.99295
.99281
.99268
.99255
.99241
':'99228
.99215
,99202
.99189
.99175
.99162
.99149
.99136
.99123
.99111
.99098
.99085,
.99072
.99059
.99047
.99034
.99021
.99009
.98513
.98502
,98491
.98479
.98468,
.98457
.98446
.98435
.98424
.98413
.98402
.98391
.98381
.98370
.

I,
per cent
by
volume.

Alcohol
per {'ent
by
weIght,

2 88
2.96
3.04
3.'12
3.20
3.28
3.36
3.44
3.52
3.60
3.68
3,76
3.84
3.92
4.00
4,08
4,16
4.24
4.32
4.40
4.48
4.56
4.64
4.72
4.80
4,88
4,96
5 05
5,13
5.21
~.50
6,60
5 29
5,37
6.70
6,80
5.45
5,53
6,90
5,61
7,00
7 .1~._ 5.69
11.30 ' '9:'J1
9. ig
11.40
9,27
11.50
9,35
11.60
9.43
11.70
9,51
11.80
9.59
.11.90
'12.00 I 9.67
' 12.10
9.75
9.83
12.20
9.92
12.30
10 00
12.40
12.50 f10.08
10.16
:.12.60
3.60
3.70
3.80
3.90
4.00
4.10
4.20
4.30
4.40
4.50
4.60
4.70
4.80
4.90
5.00
5.10
5,20
5,30
5.40
5.50
5.60
5.70
5.80
5.90
6.00
6.10
6.20
6,30
6'.40

65, U. S. Department of Agriculture.

Grams
per
lOQ.ml.

2.86
2.94
3.02
3.10
3.,18
3.26
3.34
3.42
3,50
3,58
3.66
3.74
3.81
3.89
3.97
4.05
4.103
4.21
4.29
4.37
4.44
4.52,
4.60
4.68
4.76
4,84
4.92
5.00
5.08
5.16
5.24
5,32'
5 40
5.48
5,56
5.64
8 97
9,05
9.13
9 21
9 29
9 36
9.44
9,5~

9.60
9.68
9 76
9.84
9.92
10.00

MISCELLANEOUS ANALYSES

2143

PERCENTAGE OF ALCOHOL BY VOLUME AND BY WEIGHT

(Continued)
Specific
Gravity at
6()o

000 F.

.98825
.98813
.98801
.98789
.98777
.98765
.98754
.98742
.98730
.98719
.98707
.98695
.98683
.98672
.98660
.98649
.98637
.98626
.98614
.98603
.98592
.98580
.98569
.98557
.98546
.98535
.98524
.98073
.98063
.9S052
.93042
.98032
.9S021
.98011
.98001
.97991
.97980
.97970
.97960
.97950
.97940
.97929
.97919
.97909
.97899
.97889
.97879
.97869
.97859
.97848
.97838
.97828
.97818

Alcohol

volume.

by
weight.

Grams
per
100 ml.

.8.60
8.70
8.80
8.90
9.00
9.10
9.20
9.30
9.40
9.50
9.60
9.70
9.80
9.90
10.00
10.10
10.20
10.30
10.40
10.50
10.60
"'10.70
10.80
10.90
11.00
11.10
11.20
15.40
15.50
15.60
15.70
15.80
15.90
16.00
16.10
16.20
16.30
16.40
16.50
16.60
16.70
16.80
16.90
17.00
17.10
17.20
17.30
17.40
17'.50
17.60
17.70
17.80
17.90

6.91
6.99
7.07
7.15
7.23
7.31
7.39
7.48
7.56
7.64
7.72
7.80
7.88
7.96
8.04
8.12
8.20
8.29
8.37
8.45
8.53
8.61
8.70
8.78
8.86
8.94
9.02
12.46
12.54
12.62
12.70
12.79
12.87
12.95
13.03
13.12
13.20
13:29
13.37
13.45
13.53
13.62
13.70
13.78
13.86
13.94
14.03
14.11
14.19
14.27
14.35
14.44
14.52

6.83
6.91
6.99
7.07
7.14
7.22
7.30
7.38
7.46
7.54
7.62
7.70
7.78
7.85
7.93
8.01
8.09
8.17
8.25
8.33
8.41
8'.49
8.57
8.65
8.73
8.81
8.89
12.22
12.30
12.37
12.45
12.53
12.61
12.69
12.77
12.85
12.93
13.01
13.09'
13.17
13.25
13.33
13.41
13.49
13.57
13.65
13.73
13.81
13.89
13.96
14.04
14.12
14.20

per cent

by

per cent

Alcohol

Specific
Gravity at

per cent

6()0

by
voiunle.

600 F,

I'

.98359
.98348
.98337
.98326
.98315
.98305
.98294
.98283
.98273
.98262
.98251
.98240
.98230
.98219
.98209
.98198
.98188
.98177
.98167
.98156
.98146
.98135
.98125
.98114
.98104
.98093
.98083
.97658
.97648
.97638
.97628
.97618
.97608
.97598
.97588
.97578
.97568
.97558
.97547
.97537
.97527
.97517
.97507
.97497
.97487
.97477
.97467
.97457
.97446
.97436
.97426
.97416
.97406

12.70
12.80
12.90
13.00
13.10
13.20
13.30
13.40
13.50
13.60
13.70
13.80
13.90
14.00
14.10
14.20
14.30
14.40
14.50
14.60
14.70
14.80
14.90
15.00
15.10
15.20
15.30
19.50
19.60
19.70
19.80
19.90
20.00
20.10
20.20
20.30
20.40
20.50
20.60
20.70
20.80
20.90
21.00
21.10
21.20
21.30
21.40
21.50
21.60
21.70
21.80
21.90
22.00

per cent
by
weight.

10.24
10.33
10.41
10.49
10.57
10.65
10.74
10.82
10.90
10.98
11.06
11.15
1'1.23
11.31
11.39
.11.47
11.56
11.64
11.72
11.80
11.88
11.97
12.05
12.13
12.21
12.29
12.38
15.84
15.93
16.01
16.09
16.18
16.26
16.34
16.42
16.51'
16.59
16.67
16.75
16.84
16.92
17.01
17.09
17.17
17.26
17.34
' 17.43
17.51
17.59
17.67
17.76
17.84
17.92

Grams

per
100 rol.

10.07
10.15
10.23
10.31
10.39
10.47
10.55
10.63
10.71
10.79
10.87
10.95
11.03
11.11
11.19
11.27
11.35
11.43
i11.51
1l.59
1l.67
11.75
11.82
11.90
11.98
12.06
12.14
15.47
15.55
15.63
15.71
15.79
15.87
15.95
16.03
16.10
16.18
16.26
Hi. 34
16.42
16.50
16.58
16.66
16.74
16.82
16.90
16.98
l7.06
17.14
17.22
17.30
17.38
17.46

..

2144

MISCELLANEOUS ANALYSES

PERCENTAGE OF ALCOHOL BY VOLUME AND BY WEIGHT

(Continued)
Alcohol

Specific
Gravity at
60

600 F.

per cent
by
volume.

per cent
by
weight.

.97808
.97798
.97788
.97778
.97768
.97758
.97748
.97738
.97728
.97718
.97708
.97698
.97688
.97678
.97668
.97242
.97232
.97222
.97211
.97201
.97191
.97180
.97170
.97159
.97149
.97139
.97128
.97118
.97107
.97097
.97086
.97076
.97065
.97055
.97044
.97033
.97023
.97012
.97001
.96991
.96980
.96969
.96959
.96949
.96937
,96926
.96915
.96905
.96894
.96883
.96872
.96861
.96850

18.00
18.10
18.20
18.30
18.40
18.50
18.60
18.70
18.80
18.90
19.00
19.10
19.20
19.30
19.40
23.60
23.70
23.80
23.90
24.00
24.10
24.20
24.30
24.4C
24..50
24.60
24.70
24.80
24.90
25.00
25.10
25.20
.25.30
25.40
25.50
25.60
25.70
25.80
25.90
26.00
26.10
26.20
26.30
26.40
26.50
26.60
26.70
26.80
26.90
27.00
27.10
27.20
27.30

14.60
14.68
14.77
14.85
14.94
15.02
15.10
15.18
15.27
15.38
15.43
15.51
15.59
15.68
15.76
19.25
19.34
19.42
19.51
19.59
19 .67 ~
19.76
19.84
19.93
20.01
20.09
20.18
20.26
20.35
20.43
20.51
20.60
20.68
20.77
20.85
20.,93
21.02
21.10
21.19
21.27
21.35
21.44
21.52
21.61
21.69
21.77
21.86
21.94
22.03
22.11
22.20
22.28
22.37

Specific
Gravity at
60

,G~:~s

14.28
14.36
14.44
14.52
14.60
14.68
14.76
14.84
14.92
15.00
15.08
15.15
15.23
15.31
15.39
18.73
18.81
18.88
18.96
19.04
19.12
19.20
19.28
19.36
19.44
19.52
19.60
19.68
19.76
19.84
19.92
20.00
20.08
20.16
20.24
20.32
20.40
20.47
20.55
20.63
20.71
20.79
20.87
20.95
21.03
21.11
21.19
21.27
21.35
21.43
21.51
21.59
21.67

cent
by
volume.

per cent

by
weight.

Grams
per
100 ml.

22.10
.97396
22.20
.97386
22.30
.97375
22.40
.97365
22.50
.97355
22.60
.97345
22.70
.97335
22.80
.97324
.97314
22.90
23.00
.97304
23.10
.97294
23.20
.97283
23.30
.97273
23.40
.97263
.97253 .
23.50
27.70
.96805
27.80
.96794
27.90
,/ .96783
28.00
.96772
28.10
.96761
28.20
.96749
28.30
.96738
28.40
.96726
28.50
.96715
28.60
.96704
.96692
28.70
28.80
.96681
28.90
.96669
29.00
.96658
'29.10
.96646
29.20
.96635
29.30
.96623
29.40
.96611
29.50
.96600
29.60
.96587
29.70
.96576
.96564
29.80
29.90
.. 96553
30.00
.96541
.96529
30.10
30.20
.96517
30.30
.96505
30.40
.96493
.96481
30.50
30.60
.96469
30.70
.96457
30.80
.96445
30.90
.96433
.96421
.31.00
. 9640~ ~
.96396
- 31. 20
.96384
31. 30
.96372
31.40

18.00
18.09
18.17
18.26
1'8.34
18.42
18.51
18.59
18.68
18.76
18.84
18.92
19.01
19.09
19.17
22.71
22.79
22.88
22.96
23.04
23.13
23.21
23.30
23.38
23.47
23.55
23.64
23.72
23.81
23.89
23.98
24.06
24.15
24.23
24.32
24.40
24'.49
24.57
24.66
24.74
24.83
24.91
25.00
25.08
25.17
25.25
25.34
25.42
25.51
25.60
25.68
25.77
25.85

17.54
17.62
17.70
17.78
17.86
17.94
18.02
18.10
18.18
18.26
18.33
18.41
18.49
18.57.
18.65
-21.98
22.06
22.14
22.22
22:30
22.38
22,45
22.53
22.61
22.69
22.77
22.85 I
22.93
23.01
23.09
23.17
23.25
23.33
23.41
23 49
23.57
23.65
23 73
23.81
23.89
23.97
24.04
24.12
24.20
24.28
24.36
24.44
24.52
24.60
24.68
24.76
24.84
24.92

600

100 ml.

Alcohol

F.

p~r

i31.1Q

. "'

2145

MISCELLANEOUS ANALYSES
PERCENTAGE OF ALCOHOL BY VOLUME AND BY WEIGHT
(Continued)
Alcohol

Specific
Gravity at
60
600 F.

per cent
by
volume,

I perbycent
weight.

Grams
per
100 mi.

22 45
22.54
22.62
26 20
26.28
26.37
26.46
26.54
26.63
26.71
26.80
26.89
26.97
27.06
27.14
27.23
27.32
27.40
27.49
27.57
27.66
27.75
27.83
27.92
28.00
28.09
28.18
28.26
28.35
28.43
28.52
28.61
28.70
28.78
28,87
28.96
29.05
29.13
29.22
29 30
29 38
29.48
29.57
29.65
33.35
33.44
33.53
33.61
33.70
33.79
33.88
33.97
34.06

21.75
21.83
21.90
25.24
25.32'
25.40
25.48
25.56
25.64
25.71
25.79
25.87
25.95
26.03
26.11
26.19
26.27
26.35
26.43
26.51
26.59
26.67
26.75
26.82
26.90
26.98
27.06
27.14
27.22
27.30
27.38
27.46
27.54
27.62
27.70
27.78
27.86
27.94
28.02
28.09
28.17
28.25
28.33
28.41
3l. 74
31.82
31.90
3l.98
32.06
32.14
32.22
32.30
32.38

Alcohol

Specific
GravIty at
60
600 F.

per ('ent
by
volume.

.90839
.96828
.96816
.96323
.9u610
.96298
.96285
.96273
.96260
.96248
.96235
.96222
.96210
.96197
.96185
.90172
.96159
.96146
96133
.:16120
.96108
.96095
.96082
.96069
.96056
.96043
.96030
.96016
.96003
.95990
.95977
.95963
.95950
.95937
.95923
.95910
.95896
.95883
.95869
.95855
.95842
.95828
.95814
.95800

.95185
.95169
.95154
.95138
.95122
.95107
.95091
.95075
.95059

27.40
27.50
27.60 \
3l.80
31.90
32.00
32.10
32.20
32.30
32.40
32.50
32.60
32.70
32.80
32.90
33.00
33.10
33.20
33.30
33.40
33.50
33 60
33.70
33.80
33.90
34.00
34.10
34.20
34.30
34.40
34.50
34.60
34.70
34.80
34.90 .
35.00
35.10
35.20
35.30
35.40
35.50
35.60
35.70
35.80
40.00
40.10
40.20
40.30
40.40
40.50
40 60
40.70
40.80

.90300
.96347
.96335
.95787
.9;3773
.95759
.95745
.95731
.95717
.95703
.95688
.95674
.95660
.95646
.95632
.95618
.95603
.95589
.95574
.95560
.95545
.9,'i531
.95516
.95502
.95487
.95472
.95457
.95442
.95427
.95413
.95398
.95383
.95368
.05353
.95338
.95323
.95307
.9;5292
.95277
.95262
.95246
.. 95231
.95216
.95200
.94519
.94.502
.94484
.94467
.94450
.94433
.94416
.94398
.94381

31. 50
31.60
I 31.70
35.90
i
II 36.00
36.10
36.20
36.30
36.40
36 .50
36.60
36.70
36.80
36.90
37.00
37.10
37.20
37.30
37.40
37.50
37.60
37.70
37.80
37.90
38.00
38.10
38.20
38.30 .
38.40
38.50
38.60
38.70
38.80
38.90
39.00
39.10
39.20
39.30
39.40
39.50
39.60
39.70
39.80
39.90
44.10
44.20
44.30
44.40
44.50
44 60
44.70
44.80
44.90
\

per cent
by
weight

Grams
per
100 mi.

- - ----25.00
25.94
26 03
25 08
26.11
25.16
28.49
29.74
29.83
28.57
28.65
29.92
28.73
30.00
30.09
28.81
28.88
30.17
28.96
30.26
29.04
30.35
29.12
30.44
29.20
30.52
29.29
30.61
29.36
30.70
29.44
30.79
29.52
30.88
30.96
29.60
29.68
31.05
31.14
29.76
29.84
31.23
29.92
31.32
31.40
30.00
30.08
31.49
30.16
31.58
30.24
31.67
30.32
31.76
30.40
31.85
31.94
30.48
30.56
32.03
30.64
32.12
30.72
32.20
30.79
32.29
32.37
30.87
30.95
32.46
3l.03
32.55
32.64
31.11
32.72
.31.18
31.26
32.81
3l.34
32.90
31.42
32.99
3l.50
33.08
31.58
33.17
31.66
33.27
34.99
37.02
35.07
37.11
35.15
37.21
35.23
37.30
35.31
37.39
35.39
37.48
35.47
35.57
35.55
37.66
35.63
37.76

2146

MISCELLANEOUS ANALYSES
PERCENTAGE OF ALCOHOL BY VOLUME AND BY WEIGHT'"
(Continued)

Specific
Gravity at
60
60oF .

.95044
.95028
.95012
.94996
.94980
.94964
.94948
.94932
.94916
.94900
.94884
.94868
.94852
.94835
.94810
.94802
.94786
.94770
.94753
.94737
.94720
.94704
.94687
.94670
.94654
.94637
.94620
.94603
.94586
.94570
.94553
.94536
.93786
.93768
.93749
.93730
.93711
.93692
.93679
.93655
.93636

Alcohol
per cent
by
volume.

per cent
by
weight.

Grams
per
100 ml.

40.90
41.00
41.10
41.20
41.30
41.40
41.50
41.60
41.70
41.80
41.90
42.00
42.10
42.20
42.30
42.40
42.50
42.60
42.70
42.80
42.90
43.00
43.10
43.20
43.30
43.40
43.50
43.60'
43.70
43.80
43.90
44,00
48.20
48.30
48.40
48.50
48.60
48.70
48.80
48,90
49.00

34.15
34.24
34.33
34.42
34.50
34.59
34.68
34.77
34.86
34.95
35.04
35.13
35.22
35.31
35.40
35.49
35.58
35.67
35.76
35.85
35.94
36.03
36.12
36.21
36.30
36.39
36.48
.36.57
36.66
36.75
36.84
36.93
40.78
40.88
40.97
41.06
41..15
41.24
41.34
41.43
41.52

3246
32.54
32.62
32.70
32.78
32.86
32.93
33.01
33.09
33.17
33.25
33.33
33.41
33.49
33.57
33.65
33.73
33.81
33.89
33.97
34.04
34 12
34.20
34.28
34.36
34.44
34.52
34.60
34.68
34.76
34.84
34.91
38.25
38.33
38.41
38.49
38.57
38,65
38.72
38.80
38.88

Specific
Gravity at
60
600 F.,

I
pe~ cent

I
I vo~me.

.94364
.94346
.94329
.94311
.94294
.94276
.94258
.94241
.94223
.94206
.94188
.94170
.94152
.94134
. 94i16
.94098
.94080
.94062
/' .94044
.94026
.94008
.93990
.93971
.93953
.93934
.93916
.93898
.93879
.93861
.93842
.93824
.93805
.93617
.93598
.93578
.93559
.93540
.93521
.93502
.93482
.93463
,

45.0045.10
45.20
45.30
45,.40
45.50
45.60
45.70
45.80
45.90
46.00
46.10
46.20
46.30
46.40
46.50
46.60
46.70
46.80
46.90
47.00
47.10
47.20
47.30
47.40
47.50
47.60
47.70
p.80
'47.90
( 48.00
48.10
49.10
49.20
49.30
49.40
49.50-.
49.60
49.70
49.80
49.90

.. See additional tables in Chern. Annual, D. Van Nostrand Co.

I

Alcohol
per cent
by
weight.

37.84
37.93
38.02
38.12
38.21
38.30
38.39
38.48
38.57
38.66
38.75
38.84
38.93
39.03
39.12
39.21
39.30
39.39
39.49
39.58
39.67
39.76
39.85
39.95
40.04
40.13
40.22
40.32
40.41
40.51
40.60
40.69
41.61
41.71
41.80
41.90
41.99
42,08
42.18
42.27
42.37

Grams
per
100 ml.

35.71
35.79
35.87
35.95
36.03
36.11
36.19
36.26
36.34
36.42
36.50
36.58
36.66
36.74
36.82
36.90
36.98
37.06
37.13
37.21
37.29
37,37
37.'45
37.53
37.1)1
37.69
37.77
37.85
37.93
38.01
38.09
38.17
38.96
39.04
39.12
39.20
39.28
39.36
39.44
39.52
39.60

MISCELLANEOUS ANALYSES

2147

THE ANALYSIS OF MIXTURES OF METHYL ACETATE,

METHANOL AND WATER3
The percentage by weight of methyl acetate is determined in duplicate by
saponification with alcoholic KOH solution using a Lunge weighing pipette
for getting the exact weight of the sample.
The specific gravity of the liquid is then determined at 20/4 C. as found in
the International Critical Tables; we have the following:
100 grams of sample
Sp.gr. at 20/40 C . = Volume of 100 grams, CA)

% MeAc by weight

0.933

= Volume of MeAc

.
III

100 grams. CB)

Then
A - B = Volume of MeOH + H 2 0 in 100 grams, (C)

100 grams-%MeAc by weight;=Weight of MeOH+H 20 in 100 grams CD)

and

=Sp.gr. of MeOH+HzO mixture in the sample.

Look up in specific gravity tables the percentages by weight of methanol and

water corresponding to the specific gravity CD/C) as calculated.
NOTE.-This procedure was checked against known mixtures of pure methyl acetate,
methanol and water and found to be fairly accurate within the range of 70 to 85% by
weight of methyl acetate. It can be used for similar mixtures of the corresponding ethyl
compounds.

METHYL ALCOHOL
DETECTION OF METHYL ALCOHOL-METHOD OF RICHE
AND BARDY4
The following method for the detection of methyl alcohol in commercial
spirit of wine depends on the formation of methylanilin violet:
Place 10 ml. of the sample, previously rectified over potassium carbonat~ if
necessary, in a small flask with 15 grams of iodine and 2 grams of red phosphorus. Keep in ice water for from ten to fifteen minutes until action has
Method of Commaccial Solvents Corporation.
4 Allen's Commercial Organic Analysis, 3d ed., 1: 80.

2148

MISCELLANEOUS ANALYSES

ceased. On a water bath distil the methyl and ethyl iodides formed into about
30 m1. of water. Wash with dilute alkali to eliminate free iodine. Separate
the heavy oily liquid which settles and transfer to a flask containing 5 ml. of
anilin. The flask should be placed in cold water, in ,dase the 'action should be

violent, or, if necessary, the reactiofi may.be stimulated by

gently warming the flask. After one hour boil the product
with water and add about 20 m1. of a 15% solution of soda;
when the bases rise to the top as an oily layer fill the flask up
to the neQk with water and draw them off with a pipette.
Oxidize 1 m1. of the oily liquid by adding 10 grams of a mixture of 100 parts of clean sand, 2 of common salt, and 3 of
cupric nitrate; mix thoroughly, introduce into a glass tube,
and heat to 90 0 C.foreight or ten hours. Exhaust the product
with warm alcohol, filter, and make up with alcohol to 100 ml.
If the sample of spirits be pure the liquid is of a red tint, but
in the presence of 1 % of methyl alcohol it has a'distinct violet
F !G. 298 .
P yc- shade; with 2.5% the shade is very distinct, and still more
nometer.
so with 5%. To detect more minute quantities of methyl
alcohol, dilute 5 m1. of the colored liquid to 100 ml. with water;
and dilute 5 ml. of this again to 400 ml. Heat the liquid thus obtained in porcelain and iminerse a fragment of white merino wool (free from sulfur) in it for half
an hour. If. the alcohol be pure the wool will remain white, but if methylated
the fiber will become violet, the depth of tint giving a fair approximate indi.
cation of the proportion of methyl alcohol present.
DETECTION OF METHANOL (METHYL ALCOHOL)
OXIDATION METHOD
Twenty-five m1. of the sample are diluted to about 100 ml. in a small distillation flask and about 10 grams of chromic acid added. Methyl alcohol, if
present, is oxidized to formaldehyde. 15 to 20 ml. of the solution are distilled
into a large test tube and a 5 to 10 ml. portion 'tested for formaldehyde by one
of the methods outlined on the following page.
DETECTION OF METHANOL (WOOD ALCOHOL) IN THE
PRESENCE OF ETHYL' ALCOHOL 5

. '

Dilute 1 ml. of the alcohol with water to 100 ml. To 2 nir. of this solution
add 2 ml. KMn04 (25 g. per 1.) and 0.4 ml. 50% H 2S0 4 After_3 minutes~e
stroy excess of KMn04 with oxalic acid, add 1 ml. of H 2S0 4 and 5 ml. of Schiff's
fuchsin bisulfite reagent. In the presence of, CH 3 0H a violet color is produced on long standing (HCHO formed). Ethyl alcohol produces no co16r.
1 % CH 3 0H may be detected. 3% gives an intense color.
Reagent.-Schiff's Fuchsin Bisulfite-0.5 fuchsin in 500 ml. of water, add
sufficient S02 to bleach the solution, make up to 1000 ml.
/; A. Kling and A. Lassieur, C. A., 18, 1801 (1924).

2149

MISCELLANEOUS ANALYSES
FORMALDEHYDE
DETECTION OF FORMALDEHYDE

Leach's Test.-The solution in which formaldehyde is suspected, obtained

by distillation if necessary, is tested as follows: To a 10 ml. portion, in a casserole, an equal volume of pure milk is added followed by 10 ml. hydrochloric
acid (sp.gr. 1.2) containing about 1 ml. of 10% ferric chloride solution per 500
of the acid. The mixture is heated to 80-90 C., the curd being broken up by
agitation. A violet-colored solution indicates formaldehyde.
Hehner's Test.-Five ml. of the solution in a large test tube is mixed with
about 50 ml. of pure milk, the tube tilted to a side and concentrated sulfuric acid
carefully added so as to run down the wall of the tube without mixing with the
milk. At the juncture of the acid and milk a violet color will appear if aldehyde
is present.
Morphine-Sulfate Test.-To 5 ml. of morphine sulfate reagent [0.5 gram
of morphine sulfate in 500 ml. sulfuric acid (sp.gr. 1.82) ] a 2 ml. portion of the
solution to be tested is added. The presence of formaldehyde is indicated by
a violet color that develops on standing.
ESTIMATION OF FORMALDEHYDE IN SOLUTION
Formaldehyde, HCHO, is a gas sold in aqueous solution (37% HCHO or
over). The solution generally c_ontains alcohol to prevent polymerization.
DETERMINATION OF FORMALDEHYDE
Hydrogen Peroxide Method. 7-To 50 ml. of normal NaOH in an' Erlenmeyer flask are added 50 ml. of neutral H 2 0 2 (3% sol.) and 3 ml. of the formaldehyde solution under examination, care being taken to add the sample with the
tip of the measuring pipette near the surface of the reagents. A funnel is placed
in the neck of the flask and the mixture heated for five minutes on the steam
bath, with occasional shaking. The solution is now cooled, the funnel rinsed
down, and the excess of N /NaOH titrated by normal H 2 S0 4 , in presence of
purified litmus indicator
1 ml. N /NaOH=0.03003 gram HCHO.
6 Carry out the tests with blanks of pure ethyl alcohol and ethyl alcohol containing a
little methyl alcohol.
.1 J. Assoc. Official Agr. Chemists, Methods of Analysis (1916), page 75.

2150

MISCELLANEOUS ANALYSES
GLYCEROL

There are two recognized methods for .the determination of glycerol:

A. Acetin method, depending upon the conversiOb. of glycerol to triacetin
by means of acetic anhydride and sodium acetate ana a quantitative saponification of the triacetin. This method is recommended by the International
Committee on glycerol .as giving results nearer the truth and should be employed for crude and refined glycerines of over 40% strength.
B. Bichromate method, which is based on the fact that glycerol is completely oxidized to CO 2 and H 2 0 by K 2 Cr 2 07 in presence of H 2S0 4 The
method is applicable for determination of glycerines in soap lyes.
.
The following procedures are taken from the report as recommended. s
ACETIN PROCESS FOR THE DETERMINATION OF GLYCEROL

Reagents Requir.ed
(A) Best Acetic Anhydride.-This should be carefully selected. A good
sample must not require more than 0.1 ml. n\)l'mal N aOH for. saponification of
the impurities when a blank is run on 7,5 m!. Only a slight color should
develop during digestion of the blank.
The anhydride may be tested for strength by the following method: Into a
weighed stoppered vessel, containing 10 to 20 ml. of water, run about 2 ~l. of
the anhydride, replace the stopper and weigh. Let stand with occasional
shaking, for several hours, to permit the hydrolysis of all the anhydride; then
dilute to about 200 mI., add phenolphthalein and titrate with N /1 NaOH.
This gives the total acidity due to free acetic acid and acid formed from the
anhydride. It is worthy of note that in the presence of much free anhydride a
compound is formed with phenolphthalein, soluble 'in alkali and acetic ILcid,
but insoluble in neutral solutions. If a turbidity i$ noticed toward the end
of the neutralization it is an indication that: the itnhydride is incompletely
hydrolyzed and inasmuch as the indicator is. withdrawn from the solution,
'
results may be incorrect.
Into a stoppered weighing bottle containip.g a known weight of recently
distilled aniline (from 10 to 20 mi.) measure about 2 mi. 0.[ the sample, stopper,
mix, cool and weigh, Wash the contents into about 200' ml:- of cold water,
and titrate the acidity as before. This yields the acidity due to the original,
preformed, acetic acid plus one-half the acid due to anhydride -(the other half
having formed acetanilide); subtract the second result from the first (both
calculated to 100 grams) and double the result, obtaining the ml. N /1 NaOH
per 100 grams of the sample. 1 m!. N /N aOH eql).als 0.0510 anhy.dride.
(B) Pure Fused Sodium Acetate.-The purchased salt is again completely
fused in a platinum,' silica or nickel dish, avoiding charring, powderE;ld quickly
and kept in a stoppered bottle or desiccator. It is most important that the
sodium acetate be anhydrous.
I
8
9

J. Ind. Chern. Eng., 3, 679 (1911).

J. Ind, Eng. Chern., 3, 683 (1911).

MISCELLANEOUS ANALYSES

2151

(C) A Solution of Caustic Soda for Neutralizing, of about N/l Strength,

Free from Carbonate.-This can be readily made by dissolving pure sodium
hydroxide in its own weight of water (preferably water free from carbon
dioxide) and allowing to settle until clear, or filtering through an asbesto~ or
paper filter. The clear solution is diluted with water free from carbon dioxide
to the strength required.
(D) N /1 Caustic Soda Free from Carbonate.-Prepared as above and carefully standardized. Some caustic soda solutions show a marked diminution in
streng~h after being boiled; such solutions should be reje~ted.
(E) N /1 Acid.-Carefully standardized.
(F) Phenolphthalein Solution.-0.5% phenolphthalein in alcohol and
neutralized.
The Method
In a narrow-mouthed flask (preferably round-bottomed), capacity about
120 mI., which has been thoroughly cleaned and dried, weigh accurately and
as rJ1pidly as possible 1.25 to 1.5 grams of the glycerin. A Grethan or Lunge
pipette will be found convenient. Add about 3 grams of the anhydro,Us
sodium acetate, then 7.5 mi. of the acetic anhydride, and connect the flask
with an upright Liebig condenser. For convenience the inner tube of this
Gondenser should not be over 50 cm. long and 9 to 10 mm. inside diameter.
The flask is connected to the condenser by either a ground glass joint (preferably) or a rubber stopper .. If a rubber stopper is used it should have had a
preliminary treatment with hot acetic anhydride vapor.
Heat the contents and keep just boiling for one hour, taking precautions
to prevent the salts drying on the sides of the flask.
Allow the flask to cool somewhat, and through the condenser tube add 50
m!. of distilled water free from carbon dioxide at a temperature of about 80 C., '
. taking care that the flask is not loosened from the condenser. The object of
cooling is to avoid any sudden rush of vapors from the flask on adding water,
and to avoid breaking the flask. Time is saved by adding the water before
the contents of the flask solidify, but the contents may be allowed to solidify
and the test proceeded with the next day without detriment, bearing in mind
that the anhydride in excess.is much more effectively hydrolyzedin hot than
in cold water. The contents of the flask may be warmed to, but must not
exceed, 80 C., until the solution is complete, except for a few dark particles
representing organic impurities in the crude. By giving the flask a rotary
motion, solution is more quickly effected.
0001 the flask and contents without loosening from the condenser. When
quite colg. wash down the inside of. the condenser tube, detach the flask, wash
of}: the stopper or ground glass connection into the flask, and filter the contents
. through an acid-washed filter into a Jena glass flask of about 1 liter capacity.
Wash thoroughly with cold distilled water free from carbon dioxide. Add
2 m!. of phenolphthalein solution (F), then run in caustic soda solution (C)
or (D) until a faint pinkish yellow color appears throughout the solution.
This neutralization must be done most carefully; the alkali should be run
down the sides of the flask, the contents of which are kept rapidly swirling with
occasional agitation or change of motion until the solution is nearly neutralized,
as indicated by the slower disappearance of the color developed locally by the

2152

MISCELLANEOUS ANALYSl.:S

alkali running into the mixture. When this point is reached the sides of the
flask are washed down with carbon dioxide-free water and the alkali subsequontly added drop by drop, mixing after each drop until the desired tint is
obtained.
I
N ow run in from a burette 50 ml. or a calculated 'excess pf N /1 N aOH (D)
and note carefully the exact amount. Boil gently for 15 minutes, the flask
being fitted with a glass tube acting as a partial cqndenser. Cool as quickly
as possible and titrate the excess of N aOH with N /1 acid (E) until the pinkish
yellow or chosen enclo-point color just remains.lO A further addition of the
indicator at this point will cause an' increase of the pink color; this must be
neglected, and the first end-point taken.
From the N /1 N aOH consumed calculate the percentage of glycerdl (including ~cetylizable impurities) after making the correction for the blank
test described below.
1 ml. N/l NaOH=O.03069 gram glycerol.
I

The coefficient of eXRansion'for normal i2olutions is 0.00033 per ml. for each
degree centigrade. A correction should be made on this account if necessary.
Blank Test.-As the acetic anhydride and sodium acetate may contain
impurities which affect the result, it is necessary to make a blank test, using
the same quantities of acetic anhydride, sodium acetate and water as in the
analysis. It is not necessary to filter ,the solution of the melt in this case,
but sufficient time must be allowed for the hydrolysis of the anhydride before
. proceeding with the neutralization. After neutralization it is not necessary
to add more than 10 ml. of the N /1 alkali CD), as this represents the excess
usually present after the saponification of the average soap lye crude. In
. determining the acid equivalent of the N /1 N aOH, however, the entire amount
taken in the analysis, 50 mI., should be titrated after dilutjon with 300 ml.
water free from carbon dioxide and without boiling.
.
Determination of the Glycerol Value of the Acetylizable Impurities.--4The
total residue lJ,t 1600 C. is dissolved in 1 or 2 ml. of w~ter, washed into the acetylizing flask and evaporated to dryness. Then add anhydrous sodium acetate
and acetic ~hydride in the usual amounts and proceed as described in the
regular analysis. After correcting for the blank, calculate the result to glycerol.
Instructions for
Calculating. the Actual Glycer.ol- Content
.

(1) Determine the apparent percentage of glycerol in th~- sample by the

acetin process as described. The result will include acetylizable impurities
if any are present.
(2) Determine the total residue at 160 0 C.'
.
(3) Determine the acetin value of the residue ,at (2) in terms of glycerol.'
(4) Deduct the result found at (3) from the percentage obtained at (1)
and report this corrected figure as glycerol. If volatile acetylizable impurities
are present these are included in this figure.
Trimethylenglycol is more volatile than glycerine and can therefore be concentrated by fractional distillation. An approximation lo the quantity can
10 A precipitate at this point is an indication of tire presence of iron or alumina, and
high results will be obtained unless a correction is made as described below.

MISCELLANEOUS ANALYSES

2153

be obtained from the spread between tlie acetin and bichromate results on such
distillates. The spread multiplied by 1.736 will give the glycol.
Bichromate Process for Glycerol Determination.

Reagents Required

(A) Pure potassium bichromate powdered and dried in air free from dust or
organic vapors, at llO to 120 C. This is taken as the standard:
(B) Dilute Bichromate S91ution.-7.4564 grams of the above bichromate
are dissolved in distilled water and the solution made up to one liter at 15.5 C.
(C) Ferrous Ammonium Sulfate.-It is never safe to assume this salt to
be constant in composition and it must be standardized against the bichromate
as follows: dissolve 3.7282 grams of bichromate (A) in 50 ml. of water. Add
50 ml. of 50% sulfuric acid (by volume), and to the cold undiluted solution
add from a weighing bottle a moderate excess of the ferrous ammonium sulfate,
and titrate back 'with the dilute bichromate (B). Calculate the value of the
ferrous salt in terms of bichromate.
(D) Silver Carbonate.-This is prepared as required for each test from 140
ml: of 0.5% silver sulfate solution by precipitation with about 4.9 ml. N /1
sodium carbonate solution (a little less than the calculated quantity of Nil
sodium carbonate should be.used as an excess 'Prevents rapid settling). Settle,
decant and wash once by decantation.
. (E) Sub acetate of Lead.-Boil a 10% solution of pure lead acetate with
an excess of litharge for one hour, keeping the volume constant, ana filter
while hot. Disregard any precipitate which subsequently forms. Preserve
out of contact with carbon dioxide .
... (F) Potassium Ferricyanide.-A very dilute, freshly prepared solution
containing about 0.1 %.
The Method
Weigh 20 grams of the glycerine, dilute to 250 wI. and take 25 mI. Add
the silver .carbonate, allow to stand, with occasional agitation, for about 10
minutes, and add a slight excess (about 5 ml. in most cases) of the basic lead
acetate (E), allow to stand a few minutes, dilute with distilled water to 100 mI.,
and then add 0.15 mI. to compensate for the volume of the precipitate, mix
thoroughly, filter through an air-dry filter into a suitable narrow-mouthed
vessel, rejecting the first 10 mI., and return the filtrate if not clear and bright.
Test a portion of the filtrate with a little basic lead acetate, which should produce no further precipitate (in the great majority of cases 5 m!. are ample,
but occasionally a crud~ will be found requiring more, and in this case another
aliquot of 25 m!. of the dilute glycerine should be taken and purified with 6 m!.
of the basic acetate). Care must be taken to avoid a marked excess of basic
acetate.
Measure off 25 m!. of the clear filtrate into a flask or beaker (previously
cleaned with potassium bichromate and sulfuric acid). Add 12 drops of
sulfuric acid (1 : 4) to precipitate the small excess of lead as sulfate. Add
3.7282 grams of the powdered potassium bichromate (A). Rinse down the
bichromate with 25 m!. of water and let stand with occasional shaking until
all the bichromate is dissolved (no reduction will take place in the cold).
Now add 50 m!. of 50% sulfuricacid (by volume) and immerse the vessel in

2154'

MISCELLANEOUS ANALYSES

boiling water for two hours and keep protected from dust and organic vapors,
such as alcohol, until the titration is completed. -Add from a weighing bottle a
slight excess of the ferrous ammonium sulfate (C)! making spot tests on a
porcelain plate with the potassium ferricyanide (F)r Titrate back with the
dilute bichromate. From the amount of bichroma~e reduced calculate the
percentage of glycerol.
1 gram glycerol = 7.4564 grams bichromate.
1 gram bichromate =0.13411 gram glycerol.
The percentage of glycerol obtained above includes any mcidizable impurities
present after the purification. A correction for the non-volatile impur.ities
may be made by running a bichromate test on the residue at 160 C.
For complete analysis see Journal of Industrial and Engineering Chemistry,
Sept., 1911, pages 679-683.
NOTEs.-(l) It is important that the concentration of acid in the oxidation mixture
and the time of oxidation should be strictly adhered to.
(2) Before the bichromate is added to the glycerine solution it is essential that the
slight excess of lead be precipitated with sulfuric acid, as stipulated.
(3) For crudes practically free from chlorides the quantity of silver carbonate may
be reduced to one-fifth and the basic lead acetate to 0.5 mI.
(4) It is sometimes advisable to add a little potassium sulfate to insure a clear filtrate.

LIGHT OIL IN COKE-OVEN GAS 11

USE OF ACTIVATED CARBON IN ITS DETERMINATION
The quantity of the light oil in coke-oven gas, brfore and after scru~bing
with straw oil, is generally determined by passing a d,efinite quantity of the gas
through straw oil contained in a train of bottles packed in ice, and subsequently
separating the light oil from the straw oil by distillation. The removal of the
light oil from the gas approaches completeness only when the flow of gas is
extremely slow, and under all conditions the method of procedure is cumbersome.
The light oil present in the gas may be determined readily and accurately
by the absorption of the' condensable vapor in a definite quantity' of gas by
activated carbon, with subsequent removal by distillation of the carbon with
U. S. P. cresol. . Carbon of activity 40 to 50, screened 8 to 14 mesh, is suitable.
Apparatus

The absorption requires a O.I-cu. ft. gas mete~ and two activated carbon.
tubes complete. The carbon tubes are cylindrical containers, 1.5 in. in diameter and 12 in. in length over all, fitted with a 20-mesh screen in one end, with
I-in. breasts and screw caps on both ends. Four connecting/caps consisting of
I-in. screw caps carrying 156 x i-in. copper tubes ~ldered in the centers and
_,
11 By Arthur L. Davis, Standard Oil Company (Indiatui), Wood River, Ill.

MISCELLANEOUS ANALYSES

2155

fitted with I6-in. rubber gaskets are used on each carbon tube during the
. absorption proper.
The distillation of the enriched activated carbon with cresol to remove the
light oil is made, using a 500-ml., round-bottom, side-neck, Pyrex distilling
flask; a 4-in. ring burner; a 200 0 C. thermometer; a "Barrett standard"
condenser; a 200-ml. separatory funnel graduated to 100 mI.; U. S. P. cresol;
and sodium hydro~de solution of specific gravity 1.10 to 1.15.
Procedure
Absorption.-Place 300 ml. of activated carbon in each of the two carbon
tubes. Screw' a connecting cap with rubber gasket to each end of each tube,
and tighten with a pair of combination pliers.
,Connect the inlet of the wet gas meter to the gas sampling tube and allow
gas to pass through for several minutes to insure a thorough displacement of air.
Turn off the gas, set the indicating hands of the meter to the zero position.
(Ca.ution: See that the hands are reasonably loose before turning so that the
driving shafts will not be twisted.) Connect the top of the first carbon tube to
the outlet of the gas meter, the bottom of the first tube to the bottom of the
second tube, and either allow the residual gas coming from the top of the second
tube to escape or be conducted away by an attached exit tube. Keep the tubes
in a vertical position.
Pass an amount of gas containing approximately 30 to 40 ml. of light oil
through the absorption tubes, at the rate of 10 to 15 cu. ft. per hr. At the
completion of the absorption note the volume of gas passed, the temperature
of the gas, and the barometric pressure.
Distillation.-Place the enriched carbon from the first tube in a 500-m!.,
l'ide-neck, Pyrex distilling flask, add 125 ml. of cresol, mix thoroughly by shaking
in the flask, insert the ther~ometer in neck of the flask, and connect to a " Barrett standard" condenser. Heat carefully with the ring burner and distil at a
, moderate rate till the temperature of 1800 C. is reached. Collect the distillate
in the 200-m!' separatory funnel, which is graduated to 100 ml. Add 125 ml. of
sodium hydroxide solution, specific gravity 1.10 to 1.15 (8 to 13%); and agitate
thoroughly. Care should be taken to release the pressure of the gas which is
often built up. Allow to stand for a few minutes, run off the sodium hydroxide
solution, and record the volume of light oil obtained.
Repeat the foregoing procedure with the second carbon tube. Record the
volume of light oil obtained separately.
The sum of the two volumes of light oil obtained above gives the total
volume of light oil from the volume of gas passed at the temperature and
barometric pressure recorded.
.
Calculation.-The amount of light oil present in the g~s (volume corrected
to 15.50 C. (760 mm.)) is calculated as follows:
Total volume of light oi1XO.264 = Gal. light oil per 1000 cu. ft. gas.
288.5
Bar. press.
Cu. ft. gasX 273+temp. 0C. X
760
.
Reprinted from J. Ind. Eng. Chern., 15, 689 (1923).

MISCELLANEOUS ANALY'SES

2156

Gal. light oil per 1000 cu. ft. X 12 = Gal. light oil per ton of coal.
coal produces approximately 12,000 cu. ft. of gas.) 1

(One ton of

Volume of Gas Required \

In order that the time required for making an absorption may be fairly
accurately determined before making a test, the volume of gas containing a cer'tain amount of condensable vapor should be known. The volume of gas
containing approximately 30 to 40, ml. of light oil is shown in the table.
In making routine analysis by this method it is advised that the determination of the gas entering the scrubbers and the gas leaving the scrubbers be
made simultaneously.
VOLUME OF COKE-OVEN GAS REQUIRED FOR CHARCOAL ABSORPTION METHOD

Gal. Light Oil

per Ton Coal

(12,000 Cu. Ft.)

Cu. Ft. Gas

Required for
Determination

0.5 ............................................... 190-250

1.0 ............................................... 95-125
1.5 .............................. ,. ................ 63-83
2.0 ............................................... 47-60
2.5 ......... , ........... '....... " ................ " 38-50
3.0 .......................................... , .. " 31-41
3.5 .......................... { .................. " 27-36
4.0 ............................................... 24-31
4.5 ...................................... , .. ,., ... 21-28

The spent activated carbon containing cresol may be revivified, but it is

more satisfactory to use fresh material for each determination.
STARCH AND STARCH PR0DUCTS
Starch is used in the paper industry for four p~tposes: (1) Engine sizing;
(2) tub sizing; (3) pasting, and (4) coating. It is necessary to divide it into
two main classes, with a method of analysis for each class. The first to be
known as raw starch and to include all starches which have not been treated
in any other way than by the purifying methods used in their manufacture.
The second to be known as converted starches with subclasses. The method
of sampling for all kinds 'of starch products to be the same, 'and,to be embodied
'
as part of the method of analysis.
Sampling.-Ten bags or barrels are taken from different 'parts of a car.
These are sampled by means of a I-inch tryer 12 inches long. One filing of
the tryer is to be taken from each bag or barrel. ' Samples so obtained are mixed,
bottled _and sealed in three jars, one for each party to the transaction, and one
for reference. For smaller deliveries than a carload 10% at least of the bags
or barrels should be sampled.
Preliminary Examination.-Raw Starch: (a) Spread out 100 grams of the
sample on a sheet of white glazed paper, and examine for fpreign material.
(b) Mix this sample with 500 ml. of water to.a smootb cream and strain
through a.200-mesh silk bolting cloth. Comparetesidue on silk with standard
sample of the same grade, treated in the same manner, at the same time.

MISCELLANEOUS ANALYSES

2157

(c) Stir filtrate from above for two seconds ',vith a circular motion, and let
stand thirty seconds. Decant carefully, and compare residue with a standard
sampls of the same grade of starch, treated in the same manner, at the same
time.
Moisture.-Dry approximately 5 grams to constant weight at 100 C.
.
Loss in weight is calculated as moisture.
Ash.-Weigh out 1 gram of the sample; transfer to a platinum crucible;
ignite to constant weight.
Acidity.-Weigh out 20 grams on a watchglass. Transfer to a 250-ml.
porcelain dish. Add approximately 200 ml. of water, and stir until starch
forms a smooth cream. Titrate with N /10 caustic soda, using phenolphthalein
as indicator until a faint pink color persists for twenty seconds.
Alkalinity.-Method is same as for acidity, except that sample should be
titrated with N /10 sulfuric acid until the pink coloration just disappears.
Test for the Presence of Alkaline Starch or Alkali.-Add a few drops of
phenolphthalein solution to water in a white porcelain dish; sprinkle in a little
of t;.he dry sample. If some of the grains turn pink, it indicates that the sample
is a mixture of raw starch and alkaline starch or alkali. If all the particles
turn pink, the material is an alkaline finished starch.
Viscosity.-Take 12 grams of the sample in 300 ml. of water, bring to boil,
then boil for ten minutes with constant stirring in an aluminum beaker over
a naked flame. Determine viscosity at 100 C. with 200 ml. of the liquid in
a Scott viscosimeter, standardized as follows: Calibrate to hold 200 ml. of
water at 100 C. Determine viscosity on first 50 ml. withdrawn. Express
result as a ratio. (Starch viscosity) -:- (water viscosity) = viscosity number.
Converted Starches.-Owing to the number and variety of the products
on the market, their classification is a matter of some difficulty. The method
.of conversion, however, may be used as a means, and thus the following classes
can be arranged: (1) Thin boiling starch, mainly used for tub sizing, converted
by treatment with weak acid solutions by the" in suspension" or " drying in "
process. (2) Partly dextrinized starches, and dextrins, made by damping
the starch with small quantities of acid, and converting at a high temperature
by the dry process. (3) Oxidized starches. (4) Acetylated starches. (5)
Alkali-treated starches, including products consisting of raw starch mixed with
certain quantities of various alkalies and alkaline salts. (6) Mixtures of starch
products and mineral fillers. (7) Mixture of starch products with glue and
casein or other sizing agents. (8) Products in solution or paste form.
Sampling, Moisture and Ash are the same as for raw starch.
Acidity and Alkalinity.-First method same as for raw starch.
Total Acidity.-Mix 10 grams of the sample in 200 ml. of water and heat
on a water bath for five minutes after reaching 75 C.; cool and titrate. This
gives the acidity within the starch granule.
Viscosity.-In determining viscosity the starches are divided into three
classes: (a) Low boiling, or laundry starch; (b) medium thin boiling; (c) high
thin boiling, with a decrease in viscosity from (a) to (c). For (a) the viscosity
should be determined in a 10% solution; for (b) in a 20% solution; for (c) in a
30% solution. Otherwise the metbod is the same as for raw starch.
Added Materials.__'!These are tested for by the ordinary methods used in
separating and identifying mineral fillers, metallic salts, glue, casein, and other

2158

MISCELLANEOUS ANALYSES

vegetable sizes. The identification of the product in relation to the classification of the process of manufacture involves a more exhaufltive.analysis including:
(1) The identification of the starch by microscopical observation. (2). Combined acid to determine the degree of acetylation of the product. (3) Solubility in cold water and reaction with Fehling's sblution)which gives a measure
of the degree to which the starch has been hydrolyzed I and dextrinized. (4)
Color reaction with iodine and other agents. (5) Technical ValuationCharacter of Size. A sample of the starch product is mixed with four or five
times its weight of water, and heated on a water bath, for at least twenty
minutes after swelling.
(a) Portions of the size are allowed to cool, and the time of setting observ~d.
This gives a figure for the permanency of the solution.
~
(b) A little of the size is rubbed out on paper and allowed to dry, when the
. surface and finish produced is observed.
(c) The solution is run out on a clean glass plate to a thin layer and allowed
to dry at room temperature. Film produced is examined and compared with,
say, a gelatin film, for physical characters.
Sizing Strength.-For paper coating the size is prepared under the proper
conditions for the particular product and mixed with definite quantities of
clay or other filler. A series of samples is made up at graduated strengths .
according to the proportions used in practice, and samples of paper are coated
by color obtained. The sizing power of .the sample is determined, in comparison with other sizing materials, treated in the same manner, at the same
time, by noting the resistance to spirit varnish, or by using the sealing wax "
test and others. In most cases, both for coating and tub sizing, the most
accurate and reliable results are only obtainable by a working trial uI\der
standard conditions.

TANNIC ACID
I

12
_~

The following procedure is applicable for the examination of pure tannic

acid solutions and tannins.
Reagents
Potassium Permanganate.-Approximately N /19 KMn04. Standardize
against N /10 oxalic acid solution.
Indigo Carmin Solution.-5 grams of pure indigo carmin dissolved in
water, 50 grams of conc. H 2S0 4 are added and the solution dilutE;ld to 1000 ml.
12 References: H. R. Proctor, "Leather Industries Laboratory Book of Analytical
and Experimental Methods."
Knecht-Rawson-Lowenthal, "Manual of Dyeing," VeL 11, pp. 802, Eighth Edition.
Leach, "Food Inspection and Analysis," p. 282.

MISCELLANEOUS ANALYSES

2159

Gelatin Solution.-20 grams of pure gelatin dissolved in hot water and

diluted to 1 liter.
Saturated NaCl Acid Solution.-5% H 2S0 4 solution saturated with NaCl.
Total Astringency-Lowenthal-Proctor Method
Standardization of the Indigo Carmin Solution.-The filtered solution should
give a yellow color free from brown when oxidized with KMn04.
Twenty-five ml. of the reagent in a porcelain casserole are diluted to 750 ml.
with water and the KMn04 reagent added drop by drop from a burette with
constant stirring until a pure yellow color is obtained. The rate of titration
should be uniform in all tests and should be carried out very cautiously as the
endpoint is approached. This is recognized by a faint pinkish rim appearing,
best seen on the shaded side of the casserole.

Procedure for Astringency.-l gram of solid (or 2-5 grams of solution sumac
extract) is dissolved in water and diluted to 1 liter. 10 ml. of this solution is
placed in a large casserole containing 750 mt of water.! 25 ml. of the carmin
solution are added and the mixture titrated with NjlO KMn04 as described for
standardization of the indigo carmin reagent. From this titration the number
of ml. KMn04 required by 25 ml. of indigo carmin reagent are subtracted and
the difference represents the KMn04 required by the sample. Multiply m!.
KMn04 required by sample by 0.004157 = A.
Astringent Non-tannins.-To 50 ml. of the solution (in an 8 oz. bottle)
made as stated in the previous determination are added 25 ml. of 2% gelatin
solution and 25 ml. of the saturated NaCI acid reagent followed by 10 grams of
.china clay. The bottle is stoppe'red and the mixture shaken for five minutes
and then filtered through a dry filter, thus removing the tannins. The filtrate
is tested with more gelatin solution to see that the tannin is completely precipitated. (A stronger gelatin solution should be used if tanning is evident and
the test repeated with 25 ml. of this.) 20 m!. of the filtrate, equivalent to 10 ml.
of the original solution, are titrated with N 110 KMn04 exactly as in the preceding test and tre ml. KMn04 consumed by the non-tannins is multiplied by
the value of the KMn04 (N/1O=0.004157) and the result recorded as B.
Tannic Acid or Tannins.-The difference between the total astringency,
calculated as tannin, and the astringent non-tannins gives the amount of
tannins, namely result A minus result B = grams tannin. Calculate to
percentage.
1 ml. NllO KMn04 = 0.004157 gram tannin.

BLEACHING POWDER
Sampling.-Twenty per cent of the packages in the shipment should be
sampled. The sample shall be taken by boring or cutting a two-inch hole
through the side of the wooden cask or iron drum midway from the ends, or
through the head near the center. The sampler, which consists of a stout iron
scoop about one and a half inches wide and eighteen inches long, shall be inserted for two or three inches, withdrawn, and the bleach removed discarded.
The sampler shall be inserted again as far as it will go, removed, the bleach

2160

MISCELLANEOUS ANALYSES

transferred to a glass fruit jar, and the jar closed until. the next cask is sampled.
The portions from the several casks shall all be mixed together to make on!:'
composite sample representing the shipment.
Available Chlorine.-Ten grams of bleaching powCler from the well-mixed
sample is weighed out and triturated in a mortar, itdc\ing small quantities
of distilled water at a time and grinding thoroughlY'8:fter each addition until
the powder is converted to a voluminous paste free from dense particles. The
liquor after each trituration is poured off into a lOOO-ml. graduated volumetric
fiallk, and the paste is finally washed into flask. Fill the flask to the mark
with distilled water, shake well and draw off an aliquot portion by means of a
50-ml. pipette for analysis, observing the precaution that the liquor drawn off
contains its proportion of suspended matter. The available chlorine is detsrmined by titration with N /10 sodium arsenite solution, using starch iodide paper .
as an indicator.
NOTE.-The NjlO sodium arsenite solution is made by drying arsenic trioxide, which
answers Krauch's requirements for purity, in a desiccator over night, 4.95 grams being
weighed out and dissolved in 200 m!. distilled water with 20 grams of pure sodium
carbonate, the solution cooled and made up to 1 liter, then standardized with resublimed
iodine.

One gram of potato starch is mixed with distilled water to a cream and thelJ.
poured slowly into 100 ml. of boiling water and the liquid allowed to boil gently
for two minutes when 1 gram of c.p. potassium iodide, dissolved in a little
distilled water, is poured in. The liquicfis then thoroughly mixed, allowed to
settle and strips of unsized paper are soaked in the solution and hung up to dry,
out of contact of fumes. .
,
Chlorates.-The solution in which the available chlorine has been determined by means of N /10 sodium arsenite is slightly acidified with sulfuric a,cid,
the solution transferred to a narrow-mouth flask, and a quantity of pure 'ammonium ferrous sulfate is added in excess of that required to reduce all chI orates
to chlorides. The flask is then stoppered with a 2-hole rubber stopper, which
is fitted with a small separatory funnel through one opening, and a small glass
tube ending in a short piece of rubber tubing (which rriay be clamped tight when
needed) is fixed in the other opening. The liquid IS then brought to boiling
and 15 ml. of 1.29 * sulfuric acid is added drop by drop, through the funnel.
When acid is all in, close opening in funnel and. clamp rubber tubing on vent,
cool, and titrate back the excess of ferrous ammonium sulfate with N /10
-...
potassium permanganate.'
.
Total Chlorides.-The slight pink color produced by the permanganate is
removed by the addition of a drop or two of N /10 ferrous ammonium sulfate,
and an excess of N /10 silver nitrate solution is added.
The silver chloride is filtered off, washed well and the excess silver nitratl)
is determined by means of N /10 potassium thiocyanate, the iron in solution
acting as an indicator. See chapter on Chlorine, Yol. 1.
NOTE.-The chlorine as chlorates plus the available chlorine subtracted from the
total chlorine as chlorides gives the chlorine present originally in the bleach as chlorides.

Determin.ation of Bases and Silica..,-To 2 grams of bleaching powder add

25 ml.1.12* HCI and 50 ml. of 3% H 2 0 2 Dige:;t.and evaporate on steam bath

* Specific gravity.

MISCELLANEOUS ANALYSES

2161

or hot plate. Dehydrate the silica and determine the silica, alumina, iron,
calcium oxide, and magnesia as in lime analysis.
Free Lime.-To 100 ml. of bleach liquor made up for the available chlorine
determination add 25 ml. hydrogen peroxide, bring to boil and oxidize the
hypochlorite. The solution is then titrated with standard. acid. At the same
time the acidity of the H 2 0 2 used should be determined and a correction maqe
for the amount used in destroying the hypochlorite. The free lime is calculated
from the acid consumption.
Quality of Powder for Producing Bleach Liquor.-Sixty grams of bleaching
powder is weighed out and added to 1000 ml. distilled water in a glass jar 4t
inches in diameter and 5 inches deep. The mixture is stirred vigorously by
means of a glass rod bent so that it just clears the bottom and sides of the jar,
driven by power from electric or water mot,or at 250 r.p.m. for exactly 15
minutes. The jar is then placed on a bench free from vibration and the time
required for the dregs to settle to the bottom noted; 25 ml. of the clear bleach
liqu'or is then measured out and tested in the usual way for available chlorine.
Determination of Wood and Coarse Impurities.-l00 grams of the powder
is triturated in a porcelain mortar, washed through a 60-mesh sieve, residue
is dried and weighed.
Proximate Analysis for Sand and Grit.-To 10 grams of the powder add
75 ml. of 1.12 hydrochloric acid, and warm until all lime salts are dissolved,
dilute to 200 ml. with distilled water, pour off the supernatant liquid, again
dilute to 200 ml. with distilled water, filter, wash, and ignite the residue.
NOTE.-If preferred, a potassium iodide-starch solution preserved with a drop of

CCI. or CS 2 may be used as the indicator in the determination of available chlorine by

testing a drop of the bleach liquor with a drop of the indicator on a pill tile rather than
the use of the pota~sjum iodide-starch paper. If desired, soluble starch, which dissolves
readily in hot water, may be used in the preparation of this indicator.

ANALYSIS OF CLAY 13
Chemical Analysis.-The sample should be finely ground, if necessary a
large quantity, 2 to 10 pounds, should be weighed and dried on steam table,
weighed again and loss in moisture calculated to perp-entage; it may then be
finely ground.
A weighed quantity, 1 to 10 grams, should now be dried at 100 0 C. to
constant weight, this loss calculated to percentage and added to that from the
first drying and is the moisture. A weighed portion, 1 to' 2 grams, of the
moisture-free powder is heated, in a covered crucible, to bright redness for 10
minutes, cooled and reweighed; loss = combined water and organic matter.
Or a weighed quantity is put in a boat which is heated in a tube in a current
)3

By J. M. Wilson, Lefax, 9-196, June, 1919. By courtesy of Lefax, Philadelphia.

2162

MISCELLANEOUS ANALySES

of dry air and drawn through a weighed Ca.C12 tube; gain in weight = combined
moisture.
I
Mix 1 gram dry clay with 5 g. Na 2CO a+t g. KNqa in a platinum crucible,
fuse 30 to 40 minutes, cool, detach melt from crucible. Wash crucibliJ with
H 20, then HCI, theJ1 H 20, pouring all washings into ,:1 porcelain dish, and add
the melt. Cov~r dish, add HCI little at a time until in considerable excess.
Evaporate dry, bake 1 hour at 1200 C. Cool, moisten well with HCl. Evaporate and bake again. Cool, moisten HCl, take up in H 20, boil 10 minutes,
filter, wash with hot HCl (1 : 1), cold H 2 0, hot HCl, cold H 2 0, then 5 to 6
times hot H 20, ignite in a weighed platinum crucible and hold.
Evaporate the filtrate and washing in a dish, bake as above, moisten with
HCl, evaporate and bake again. Take up HCl+H 20 filter and wash as above.
Burn this filter in same crucible, cool, weigh = crude Si0 2 . Moisten with few
drops' H 2S0 4 (1 : 4), then H 2 0 to make a thin mud; add enough HF to fill
crucible i full, evaporate at gentle heat; heat cautiously over lamp-to avoid
spattering-raising heat slowly to full heat of lamp. Loss in weight = Si0 2.
If any residue remains in the crucible, fuse with KHS0 4. Leach melt with
H 20 and add to main filtrate.
Boil filtrate from silica, add 1 to 3 mi. HN0 3, boil 10 minutes. Add NH 40H
in slight excess, boil 3 minutes, let settle in Warm place, decant through a filter.
Add hot water to precipitate, stir, let settle, filter, dissolve precipitate in
HCI (1 : 1). Dilute to 250 m!., boil, ~eprecipitate as above, repeat again,
wash thoroughly with hot H 20. Ignite and weigh precipitate = Fe 2 0aAl 20 3(P 20 5 Ti0 2).
'
Fuse with KHS0 4, leach with H 20, reduce in Jones reductor. Titrate F'e
with KMn04. Fe/0.7=Fe 203.
Fuse 2 to 5 grams dry sample with 5 wts. Na 2COa+t wt. KN0 3 Leach
melt with H 20, filter, wash with H 20+Na 2CO a. Save residue for Ti. Acidify
filtrate with HN0 3, boil off CO 2, add Fe 2 CIs solution (free from Phos.), boil.,
Add NH 40H, filter, take up precipitate in HNO a. Boil well and precipitate
the phosphorus with ammonium molybdate. Deter~ine phosphorus.
PX2.29=P 20 S
(AI20 a+Fe203+ P 205+Ti0 2) -(Fe20a+P205+Tj02) = AI 2 0 a.
Evaporate united filtrates from A1 2 0 3, etc., to 100 ml. Make ammoniacal,
add Br2 in excess, then 60 to 70 ml. NH 40H, boil 20 minl,!tes, filter, wash with
hot H 20. Ignite and weigh Mn a04. Mn 3 04XO.93006=M"nO;--.
Acidify filtrate from Mn02, evaporate dry, take up in HCI+H 20 (volume
150 mI.), boil, add NH 4 0H till neutral, then ammonium oxalate. Boil 10minutes, cool 2 hours, decant close as possible through filter. Dissolve precipitate in HCl, dilute to 100 mI., add ammonium oxalate, boil. Precipitate with
NH 40H, boil 10 minutes, filter after 2 hours, wash, twice with col<,i H 20, then
5 or 6 times with hot H 20. Ignite strongly, weigh while still warm, repeat
ignition and weighing = CaO. Or dissolve precipitate in least quantity HCl
(1 : 1) through filter into an Erlenmeyer flask. Wash with H 20, add 25 ml.
H 2S0 4 (1 : 4), heat to 800 C., titrate with KMn04. Fe value, KMn04/2 = CaO.
Evaporate combined filtrates from CaO to smallrvolume (15 to 20 mI.), add
1 gram ammonium phosphate (NH 4)2HP0 4, then .NH40H till solution is
neutral. Cool under tap, stir till precipitate forms, set aside 15 minutes,

MISCELLANEOUS ANALYSES

2163

stirring occasionally. Add! volume NH 40H, stir, let stand '4 hours, filter
but do not wash. Dissolve through filter in HCI into vessel in which precipita_
tion was c.onducted. Wash filter with H 20, reprecipitate with NH 40lI as
before, filter on weighed Gooch felt. Wash with (NH 4NO g, 60 g.; H 2 0, 80 ml.;
NH 4 0H (1 : 3), to 600 mI.), dry at 100 C., ignite cl1refully till a glow passes
over the precipitate, c601 and weigh Mg 2P 20 7

Alkalies.-Weigh 1 gram clay, 1 gram NH 4 CI into a platinum crucible, mix

thoroughly. Add 6 grams CaC0 3 '(free from alkalies), mix, cover with 2 grams
CaCOg. Heat covered crucible very gently till no more fumes are given off.
Raise heat slowly till lower third of crucible is at good red heat .. Maintain
this heat for 15 minutes, cool, transfer sintered mass to a porcelain dish..
Wash out crucible with water into dish, add more water till volume is 20D-300
mL Digest at 100 C., breaking lumps with a glass rod, then boil 5 to 10
minutes, add (NH 4hC0 3, allow to stand 20 to 30 minutes. Filter into a
porcelain dish, wash thoroughly with hot H 2 0 containing (NH 4)2C0 3 Evaporate nearly dry, add :N"H 40H+ammonium oxalate, filter into a platinum dish.
Evaporate dry, ignite gently to drive off ammonium salts, cool, add (NH 4 )2CO g
+ammonium oxalate, filter through very small filter into a weighed platinum
.dish. Evaporate dry, ignite gently till ammonium salts are volatilized, raise
heat till bottom of dish is a faint red, cool and weigh. Moisten with a drop or
two of HCI, heat again and weigh=NaCI+RCI. Separate 1st by PtC1 4 ,
2d by perchloric acid, 3d by titra,tion of Cl with AgNO a, or 4th by conversion'
into sulfates and determining the SOa with BaCl 2
Titanium.-Decomposition, (1) by fusion with 5 wts. Na 2CO a+! wt. RNO g,
leaching with H 2 0, and extracting Ti0 2 from residue, (2) by fusion with RHS0 4,
leaching with H 20+H 2S0 4, filtering and precipitating Ti0 2 from filtrate, (3)
by moistening with dilute H 2S0 4 and evaporating with HF to remove Si0 2
and decompose silicates.
Precipitate, add NH 40H to solution until a faint precipitate is produced,
then HCI to redissolve the precipitate and leave a slight excess. Dilute to
250 mI., add strong solution N a 2S20a and boil 20 minutes. Let settle, decant
through a filter, wash once by decantation, keeping as much of the precipitate
in the beaker as possible. When filter has drained, place beaker, in which
the precipitation was made, under the funnel and treat filter with concentrated
HCI. Allow to act 15 to 20 minutes, wash filter thoroughly with cold H 2 0,
repeat the precipitation as above, filter, wash well with hot H 20, ignite and
weigh Ti0 2
When Ti0 2 is present in small amount, it may be determined by color.
Standard Ti(S04h solution is made as follows: Fuse t gram'rutile with Na 2 CO a,
leach melt with H 2 0, filter, take up insoluble in H 2S0 4 (1 : 4), evaporate to
fumes, cool, take up with H 20+ H 2S0 4, filter if necessary, dilute till 1 ml. =
0.001 gram Ti. Decompose clay by fusion with Na 2CO a and treat the residue
with H 2S0 4 Evaporate to fumes, take up with H 20+H 2S0 4 , pour into a
Nessler tube. Fill another' Nessler tube nearly to same volume with H 2 0.
To each add 5 ml. H 20 2 and to the standard add the standard Ti(S04)2 fram
a burette until ,the brown 'colors match, when the volumes are equal. The

2164

MISCELLANEOUS ANALYSES

burette reading will then give the number of milligrams 6f Ti in the clay from
which the percentage may be calculated.
I
State of Silica.-A-Combined with bases in the clay, B-Hydrated silica,
C-Sand or silicates. Moisten 2 grams dry powder with H 2 0, add 20 ml.
H 2S0 4 (conc.), cover dish, heat 10 hours, remove cover,1 evaporate dry, cool,
take up in H 2 0, filter, wash with H 20, ignite and weigb.=cA+B+C.
Boil with Na 2 CO a solution, allow to settle, decant through a weighed Gooch
felt or alundum crucible, repeat boiling and decantation until a portion of the
filtrate gives no precipitate with NH 4 CI on standing. Transfer residue to
filter and wash thoroughly with hot H 2 0" acid, then hot H 2 0. Ignite and
weigh = C. A and B are in filtrate.
Boil 4 to 5 grams dry powder with strong solution of Na 2 CO S , filter, washwith hot H 2 0. Acidify filtrate with HCI, evaporate dry, bake at 120 0 C.;
Gool, moisten with HCI, evaporate dry, bake again. Take up in Hel and H 2 0,
filter, wash, ignite; weigh as Si0 2 =B. (A+ B+ C) -(B+ C) = A.
If A+B+C=total Si0 2 found on analysis, C is quartz. If A+B+C is
greater than total Si0 2 , C is in form of silicate and weight of Si0 2 as silicates =
A+B+C-total Si0 2 found in 1 gram.
.

ANALYSIS OF GLASS

14

As glass consists of such a variety of substances, to define it exactly is a

decidedly perplexing problem. N either can a formula be written for such a
substance, because it may contain many different radicals, and also because
glass is a mixture of silicates and not a definite compoutid. .
It is also quite a problem to write up an analysis of glass that will embrace
all glasses. The most common substances usually found in clear glasses are
Silica, Lead, Alumina, Calcium, Magnesia, Soda, ,Potash, Iron as an impurity,
and other substances found in special glasses (tnd colored glasses such as
BarIum, Borax, Manganese, Sulfur, Zinc, Tin, Copper,. ,Qhromium, Gold,
and Uranium, Fluorides, Phosphates, Lithium and Strontium are ..also occasionally found. Antimony, Arsenic, Cobalt, Selenium, Titanium, etc., are
not uncommon but are usually present in such small quantities that they are
seldom determined or even looked for.
,
In the following analysis only the substances most frequently encountered
by the glass chemist will be considered. They are-Silica, Lead, Iron, Alumina,
Calcium, Magnesia, Soda, Potash, Barium, Borax,: Manganese a~d Sulfur.
If an unknown glass is to be analyzed, it is always necessary to make a qualitative examination of the sample before any quantitative work should be attempted.
I
14 By G. E. Fisher, Lefax, 11-319.
Philadelphia.

By courtesy o( J:"

G. -Parker,

President Lefax,

,-

MISCELLANEOUS ANALYSES

2165

, Density or Specific Gravity.-The value and importance of knowing the

specific gravity of a glass is quite often overlooked but its value becomes more
evident as one continues to make the determinations and link them up with
the analysis of the glass in question. It is always advantageous to take the_
specific gravity before the glass is analyzed as it quite often gives a clue as' to
the type of glass, especially to the person who is familiar with linking gravity
and analysis. It is always interesting to calculate the specific gravity from the
percentages of constituents in the analy,sis. This can be done by any who
hav the proper tables and formula at hand.
The specific gravity of a piece of glass weighing as little as 8 grams can be
.. very accurately determined on the analytical balance. By making substitution in the equation:
D= W 1 - W 2
Wl '
where D=specific gravity, WI = weight in air, W 2 =weight in water, the density
is easily computed. Duplicate computations should check within five points
in the third decimal place.
Preparation of Sample and the Determination of Silica.-Grind the sample
of glass to an impalpable flour with a mortar and pestle. Weigh out 1 gram
of the finely powdered glass into a platinum crucible. Weigh out 4 grams of
sodium potassium carbonate fusion mixture and about half a gram of potassium
chlorate and transfer to the same crucible. Mix thoroughly and take care
not to lose any. Fuse this and keep it hot till the mass is a liquid in the bottom
of the crucible. Allow the mass to 9001 off considerably and then take up the
melt with 50 to 100 ml. of hot water containing about 10 ml. of hydrochloric
acid. If borates are present, refer to the. removal of borates under " Determination of B 2 0 a." Transfer to an evaporating dish and evaporate to dryness
on a steam bath and then bake for at least half an hour on a hot plate at 110 C.
Then take up with dilute hydrochloric acid and water and filter. Wash the
residue with hot water until free from chlorides. Save this residue. Transfer
the filtrate to the evaporating dish and evaporate an'd bake as before. Take
this up with water and hydrochloric acid and filter and wash with hot water
until free from chlorides. Combine this residue with the first one and ignite
to constant weight in a previously ignited platinum crucible. After weighing,
moisten with dilute sulfuric acid and then fill the crucible three quarters full
with hydrofluoric acid and evaporate to dryness in an iron cone, ignite and
weigh. Deduct the weight of this residue from that of the above weighing.
The difference in weights is the weight for corrected silica from which the percentage is easily calculated. Fuse the residue which remains in the crucible
with about 2 grams of potassium bisulfate, take up with dilute hydrochloric
acid and add it to the main solution. If the sodium potassium carbonate fusion
mixture is not known to be pure, it should be analyzed and corrections made
accordingly.
Determination of PbO.-Neutralize the solution, then make it just acid
with 1 drop of hydrochloric acid. Pass in hydrogen sulfide gas for several
minutes to insure complete precipitation. Allow the lead sulfide to settle,
then filter on a Gooch crucible. Wash the residue with water saturated with
hydrogen sulfide. Again pass hydrogen sulfide into this filtrate to be sure

2166

MISCELLANEOUS ANALYSES

precipitation is complete. Preserve the filtrate for the estimation of iron, etc. I
Dissolve the lead sulfide in dilute nitric acid (1 : 4)j Filter this solution to
remove the asbestos used in the Gooch. Precipitate lead sulfate from this
solution by adding dilute sulfuric acid with constant stirring. Then add 100%
by volume of 98% ethyl alcohol to the cold 'Solution\. Allow the precipitate
to settle for an hour and filter on a weighed Gooch previously ignited at a red
heat. Wash with 50% alcohol. Heat the residue sl~wly and ignite at a dull
red heat until the residue is white. Cool and weigh as lead sulfate. Calculate
the percentage of PbO by using the proper factors which may be found'in the
table below.
Determination of FezOa and AlzOa(RzOa).-Boil this solution with 2 or
3 illl. of nitric acid to remove the sulfur and oxidize the iron. Cool the solution.
Add about 5 grams of ammonium chloride and then add ammonium hydroxide
until the solution is slightly ammoniacal. Boil cautiously for about 5 minutes.
Allow the precipitate to settle. Filter and wash with hot water or hot 2%
ammonium nitrate solution. Preserve this filtrate for the determination of,
calcium. Ignite the residue in a weighed platinum crucible. Cool and weigh
as FezOa and AlzO a. Fuse this residue with potassium bisulfate and dig~st with
dilute sulfuric acid. Transfer to a 250-ml. Erlenmeyer flask. Pass in hydrogen
sulfide gas to reduce the iron. Then fit a B.unsen valve to the flask and boil to
a small volume to drive off the excess hydrogen sulfide. Cool and add a pinch
of sodium bicarbonate and titrate with stlj.ndard N /200 potassium permanganate
solution. From the amount of standard solution used calculate the percentage
of iron oxide. Deduct the percentage of iron oxide from (FezOa+AlzOa) and
the difference is the per cent AlzO a.
Determination of CaO.-Bring the solution to boiling. While hot add 10 to
20 ml. of hot saturated ammonium oxalate solution with constant stirring.
Then add about 5 m!. of ammonium hydroxide. Allow to stand for several
hours or preferably over night. Filter off the calcium oxalate and wash with
a 2% ammonia water. Preserve the filtrate for the determination of magnesia.
Ignite the residue at full. blast in a previously ignited and weighed platinum
crucible to constant weight. Weigh as calcium oxide.
Determination of MgO.-Concentrate the, solution by boiling and while
hot add 10 to 20 m!. of saturated microcosmic salt solution with constant
stirring. Then add ammonia and stir vigorously, taking care not to strike
the sides of the beaker. Allow the precipitate to settle for several hours or
preferably over night. Filter and wash with a 2% solutiOn of-ammonia water.
Dry the residue and filter paper. Ignite slowly at first and then at full heat
in a weighed porcelain or platinum crucible, until constant weight. Cool andweigh up as magnesium pyrophosphate. Calculate the percentage of MgO
by using the factor found in the table below. If large amounts are to be ignited,
it is best to ignite the residue and paper separately in a porcelain crucible.
Determination of the Alkalies.-The method in general use is that of Sullivan and Taylor. Weigh out 1 gram of finely powdered glass in a platinum
crucible of about 40 ml. capacity, moisten with water and add 2 grams of oxalic
acid crystals and enough 48% hydrofluoric acid to fill the crucible half full.
This is evaporated to dryness and warmed to a temperatule which is just high
enough to expel the excess oxalic acid. When al,H!le_acid'is expelled, the crucible is cooled and the evaporation repeated twice more with oxalic acid and water,

MISCELLANEOUS ANALYSES

2167

The amount of oxalic acid used is about 5 grams in all. After the third evaporation when all the excess of oxalic acid is expelled, the remaining oxalates are
taken up with hot water~ allowed to cool and filtered. The residue which
consists of lead oxalate and calcium oxalate is filtee;ed off. The filtrate is
evaporated to dryness in a platinum dish and then heated over a free flame
until oxalates are decomposed to carbonates. The insoluble carbonates are
filtered off and the alkali carbonates are in the filtrate. The filtrate is made
acid with hydrochloric and' evaporated to dryness to remove traces of silica.
The filtrate is evaporated to dryness in a platinum dish, the sodium chloride
and potassium chloride heated to incipient fusion, cooled and weighed. This
gives the weight of the combined alkali chlorides. The chlorides are dissolved '
in about 25 m!. of water and 4-5 m!. of chloroplatinic acid (1 gram Pt in 10 m!.)
added and the solution evaporated to a syrupy consistency. The insoluble
potassium chloroplatinate is filtered on a Gooch. Wash by decantation with
85% ethyl alcohol until the washings are colorless. The Gooch is dried at 130 0
C. to constant weight. ' The potassium chloroplatinate is computed to' potassium. chloride, and the sodium chloride found by difference from the weight of
the combined alkali chlorides. Both potassium chloride and sodium chloride
are computed to their respective oxides.
Determination of B 20 a.-Before proc~eding with any of the determinations,
boric acid must be removed. This is done by adding 25 ml. of methyl alcohol
in both dehydrations of silica. Then proceed as usual. The determination
of B 20 a must be made as follows on a separate sample of glass. Fuse.5 gram
sample with about 4 grams of sodium carbonate. Take up with 20--30 ml.
dilute hydrochloric acid and add a few drops of nitric acid to oxidize the iron.
Place the whole in a 250 ml. round bottom flask. Heat nearly to boiling and
add dry precipitated calcium carbonate in moderate excess. * Connect with a
reflux condenser and boil vigorously for ten minutes to remove CO 2 from the
solution. Filter off precipitate through a small Buchner funnel, washing several
times with hot water. Keep the volume below 100 ml. Return the filtrate
to the flask and add a small amount of calcium carbonate and again heat to
boiling. Connect up to a filter pump, through a splash trap, and continue
suction cautiously, until boiling has nearly ceased, to remove the last traces of
' \ CO 2 in the solution. Cool to room temperatureand filter if necessary. Add
4-5 drops of phenolphthalein indicator and run in slowly N /10 standard sodium
hydroxide solution until the liquid is decidedly pink. Add about 1 gram of
mannite or 10 ml. glycerol and shake. The pin!>: color will disappear. Then
run in more standard NaOH till pink and then more glycerol. Continue
this until the pink is permanent. The glycerol must be neutral. The standard
N aOH should be standardized against a standard sample of B 2 0 a The percentage B 2 0 a can then be computed.
This method is unsatisfactory when zinc and lead are present in the glass as
high results are obtained. Fuse a half-gram sample with 4 grams of sodium
carbonate. Digest with 20--30 ml. of hot water and when the 'melt is entirely
decomposed filter off any insoluble oxides. After washing, transfer the filtrate
to a 250-ml. round bottom flask. A~d about 7 m!. concentrated hydrochloric acid. Heat nearly to boiling and add a moderate amount of calcium carbonate
in excess; from here proceed as above from asterisk.*

2168

MISCELLANEOUS ANALYSES

Determination of Mn02.-The iron and alumina precipitation with ammonia also effects the precipitation of manganese. I It is then necessary to
separate these three substances. The residue containing Fe203, A1 20 3, Mn02 is
dissolved in dilute nitric acid. It is necessary that '~he solution be free from
ferrous iron. To this solution add as much sodiu'm carbonate as possible
without any precipitation taking place. This must be added very cautiously
drop by drop. Dilute and add one-half gram of sodium acetate and h<'>il.
Let settle and filter. Wash with a t% solution of ammonium a<;etate. The
filtrate should now be neutral to methyl orange. The residue contains the iron
and alumina. The filtrate contains the mangan~se. Dissolve the residue in
dilute hydrochloric acid and determine iron and alumina by the foregoing
method. The filtrate should be concentrated. Add 10 grams of sodium acetate. The solution should now have a volume of about 200 mi. Bromine
water is then added until the color is distinct. The solution is then boiled for
about 15 minutes, with small additions of bromine water. The precipitate
is then allowed to settle. Filter and wash. Dissolve'the residue in hot dilute
hydrochloric acid containing a little bisulfite. Add ammonium ptl.Osphate~
then ammonia slowly, with heating and constant stirring. Allow it to settle
a few hours, then filter and wash with a dilute solution of ammonia water.
Ignite in a weighed porcelain crucible, as manganese pyrophosphate, and compute the percentage of Mn02 from the factor in the table below.
'
Determination of Sulfur.-A separa~e sample should be used for the estimation of sulfur. Weigh out a 2 gram sample, mix,with 5 grams of sodium potassium carbonate fusion mixture and a half gram of potassium nitrate. During
fusion the charge should be protected from the flame b)\ an asbestos board
provided with a hole to accommodate the crucible. Take \lp the fusion with
water, transfer to an evaporating dish and dehydrate. Take up with dilute
hydrochloric acid, filter and wash. Dehydrate the filtrate the second time,
. take up with dilute hydrochloric acid, boil to drive off carbon dioxide, filter and
wash. To the hot filtrat.e add enough barium chloride solution to precipitate
the sulfur completely. The barium chloride shoult,f be added very slowly and
with constant stirring. Allow the barium sulfate to settle several hours and
then filter on a weighed ignited Gooch. Wash with a dilute alcohol solution.
Ignite at a temperature of about 800 0 C. Qool and weigh as barium sulfate.
Compute percentages from the table given below.
Determination of BaO and Separation of Barium from Calcium.-This
separation is made from the solution from which iron ;na alumina have been
precipitated. Boil this solution sufficiently to remove the ammonia. Evaporate the filtrate nearly to dryness. Then take up with about JOO mI. of water
and add 5-6 drops of acetic acid with enough,ammonium acetate to neutralize
any free mineral acid present. Heat the solution and add an excess of ammonium chromate. Allow the barium chromate precipitate to settle for an
hour or more. Filter on a weighed Gooch and. wash with water containing
ammonium chromate, until free from calcium. Then wash with water until
free from ammonium chromate. Dry the precipitate at 1100 C. and ignite
gently at first and finally to a dull red beat. Weigh as 9arium chromate and
compute the. percentage of BaO from table.

MI~CELLAN EO U~ ANAL YSE

2169

TABLE OF COMPUTATION FACTORS

BaS04 XO.1373=S
XO.2744=S04
.XO.3430=SO.
~ X0.4U5=S04
X 0.4202 = H 2S04
XQ.821 = Na zS04
XO.747 =K2S04
XO.892 =BaCI 2
BaCr04XO.541 .=Ba
XO.605 =BaO

PbS0 4 XO.736=PbO
XO.683=Pb
Mg 2P,07 XO.218= Mg
XO.362=MgO
Mn 2P 20 7 XO.387=Mn
XO.612 = Mn02
K 2PtC1 6 XO.306=KCI
XO.193=K 20
KCI
XO.631=K 20
NaCl
XO.530=Na 20

ANALYSIS OF THE MOST IMPORTANT GLASS-MAKING

MATERIALS 15
Potash (Potassium Carbonate) Water Insoluble. Dissolve five grams of
the sample in 100 m!. of distilled water and filter off any insoluble matter on a
tared Gooch crucible. Dry, cool and weigh.
Moisture.-Weigh out a five gram sample into a tared porcelain crucible or
a small evaporating dish and heat in an electric oven at 110 C. for several
hours. Cool and weigh and report' the loss in weight as moisture.
Chlorides.-Dissolve a five gram sample in 100 m!. of distilled water.
Filter if necessary. To the filtrate add a few m!. of nitric acid until the solution
is acid and then add silver nifrate solution until precipitation of the chlorides is
complete. Heat the solution'to boiling and stir vigorously to coagulate the
silver chloride. Let the precipitate settle (in a dark place) and filter on a tared
Gooch crucible and dry on a hot plate. Cool and weigh and calculate the
percentage of potassium chloride. Factor 0.519.
Sulfates.- Dissolve five grams of the sample in 100 m!. of distilled water.
filter if necessary. To the filtrate add hydrochloric acid until the solution
is slightly acid. Heat the solution to boiling and precipitate the sulfates by
adding barium chloride solution slowly and with vigorous stirring. Allow to
stand on a hot plate to settle. Filter on a previously ignited and weighed
Gooch crucible. Ignite at a red heat to constant weight, cool and weigh as
barium sulfate. Calculate the percentage of potassium sulfate by using the
factor 0.747.
When the Sulfates are Under l%.-A two and one-half ,gram sample is
dissolved in about 100 m!. of distilled water and then made slightly 'acid with
hydrochlorio acid. Filter if .necessary. Transfer the solution to a 250-ml.
volumetric measuring flask and dilute to 250 m!. From this solution draw off .
25 m!. with a pipette and drain into a previously ignited and weighed platinum
dish. Evaporate the solution to dryness and then heat gently (to avoid

. 1. By G. E. Fisher, Lefax, 12-561.

Philadelphia.

By courtesy of J. C. Parker, President Lefax,

2170

MISCELLANEOUS ANALYSES

spattering of the residue) to low redness or to the point where the chloridp.s
just begin to fuse. Cool and weigh and note as combined chlorides.
The precipitate in the platinum dish should then be taken up with the
smallest possible amount of water and to it add seven'teen times the weight of
the combined chlorides, of a solution of chloroplatinib acid (1 gram platinum
to 10 mI.). Evaporate this solution to a thick syrup, and take it up with
25 ml. of 95% alcohol and filter through a tared Gooch crucible and wash with
95% alcohol until the washings are colorless. It is then dried at about 160 0 C.,
cooled and weighed. The washings and residues should be preserved for the
recovery of the platinum.
K 2PtCl 6 X 0.307 = KCI.
KCI- {K2S04 r.educe~ to chlorides} = KCI as K CO .
2
3
KCI as Impunty
KCIXO.926XIOO =
0.25

01
/0

K CO
2

Combined chlorides-KCI = N aCI.

NaCIXO.905XlOO _01 N CO
, , - /0 a2 3
0.25
When the Sulfates are Over l%.-To the 25 ml. as obtained above and
heated to boiling add barium chloride in slight excess. Let it settle and filter.
Then to the filtrate add ammonium carbonate in excess to precipitate the
excess of barium. Filter. Evaporate this filtrate to dryness and ignite very
carefully to decompose the excess ammonium carbonate and fuse the combined
chlorides. Then proceed as outlined above when the sulfates are not in excess
of 1%.
The potash residues and washings should be s~ved and the platinum recovered as follows: The washings and residues are 'combined and dissolved in
hot distilled water. Add sodium formate and sodium hydrate and heat the
solution until the platinum is precipitated. Filter and ignite, cool and weigh.
Dissolve this residue in aqua regia and evaporate with successive additions of
hydrochloric acid until ,all the nitric is expelled. Then pass chlorine through
the solution and evapo'rate to a syrup to decompose the excess hydrochloric
acid. Filter and ignite and weigh and correct for the amount not dissolved.
Make up the solution with distilled water so that 10 ml. contains 1 gram of
platinum.
,
Feldspar, Kaolin and China Clay.-These are the chief sources of alumina
in glass batches and' their frequent examination i~ ~o be encouraged. They all
contain silica, iron and alumina but in different proportions. Kaolin and
China clays are examined for moisture whereas feldspar usually is not. There'fore the same methods of analysis may be carried out for these three materials.
Not more than 0.5 gram. sample should be weighed out. (This is done
because the alumina precipitate is so voluminous_ that if larger samples are
taken the precipitate is too bulky to handle weHTnd losses may occur.) This
is mixed thoroughly with about four grams of sodium potassium carbonate in a'

/'

MISCELLANEOUS ANALYSES

2171

platinum crucible and fused. Digest the fusion in dilute hydrochloric acid
and distilled water. Then transfer to a porcelain evaporating dish and evaporate to dryness and bake for about half an hour. Take it up with hot distilled
water and dilute hydrochloric acid and filter. Wash the residue with hot water
until entirely free from chlorides. Preserve the residue. Make a second dehydration on the filtrate and digest as above, filtering and washing. Combine
the residues from the two filtrations and ignite them slowly at first in a clean
platinum crucible. When the residue is snow white and to constant weight, cool
and weigh. Then moisten the residue with dilute sulfuric acid and make the
crucible three quarters full of hydrofluoric acid and evaporate to dryness.
(Sometimes it may be necessary to make a second hydrofluoric evaporation.) ,
Ignite, cool and weigh and report the loss in weight as silica. The residue
remaining in the crucible may be fused with potassium bisulfate or sodium potassium carbonate and the fusion added to the filtrate remaining from the silica.
To this solution then cautiously add ammonium hydroxide until it is just
alkaline and a dense voluminous precipitate comes down. Transfer the beaker
to a .flame and boil the solution for a few minutes. Allow the precipitate to
settle and then filter and wash with hot distilled water. Preserve the filtrate.
After the residue is thoroughly washed, place a clean beaker under the funnel
and pierce a hole in the paper and wash the residue through with hot water
and dilute hydrochloric acid. Be sure to wash the paper thoroughly because
any residue remaining and lost will make the results inaccurate. The residue
will be dissolved in dilute hydrochloric acid and reprecipitated with ammonium
hydroxide, boiled, filtered and washed -as described above. The residue is then
ignited in a tared platinum crucible, cooled and weighed and reported as (R 20 a),
4-120a' Fe 20a. This residue should then be fused with potassium bisulfate or
sodium potassium carbonate and the fusion taken up with dilute sulfuric acid.
The iron may then be reduced by a ch~ice of methods and titrated with a standard solution of potassium permanganate. The filtrate from the second precipitation of alumina should be combined with the main solution.
The solution is then brought to boiling and, while hot, add about 20 m!. of a
saturated ammonium oxalate solution and stir constantly. Then add a few
m!. of ammonium hydroxide. Allow the precipitate to settle for several hours
and then filter. Wash well with a 2% solution of ammonia water. Ignite the
residue at full blast, in a tared platinum crucible, until the weight is constant.
Cool and weigh as calcium oxide and report it as such.
Concentrate the filtrate by boiling and while hot add a few m!. of microcosmic salt solution and then ammonium hydroxide. Stir vigorously and
constantly. Allow the precipitate to settle for several hours and then filter.
Dry the residue and paper before ignit.ing. Then ignite in a tared platinum
crucible, cool and weigh as magnesium pyrophosphate. Calculate the percentage of MgO by using the factor 0.362.
If the alkalies in feldspar are desired by other than the difference method,
a separate sample may be used and the alkali determined as previously described.
For moisture in kaolins and clays a five gram sample is weighed out and
placed in a tared porcelain dish and transferred to a hot plate or oven and
heated to about 110 C. for several hours or over night. The loss in weight is
reported as moisture.

2172

MISCELLANEOUS ANALYSES

Barium Carbonate.-Weigh out a one gram sample and dissolve it completely in hydrochloric acid. When dissolved, tr&nsfer the solution, to a
porcelain evaporating dish and evaporate to dryness ,On a hot plate at 110 0 C.
This residue is then taken up in hot dilute hydrochl9ric acid and filtered and
washed with hot water until free from chlorides. Igmte in a,previously ignited
and weighed platinum crucible, cool and weigh. This residue is then
moistened with dilute sulfuric acid and then the cr'ucible is half filled with
hydrofluoric acid and evaporated and ignited; cooled and weighed. The loss in
weight is reported as silica. If any residue remains, it is fused with sodium
potassium carbonate and added to the main solution.
Separation of Iron and Alumina.-A small amount of ammonium chloride
is added to the solution and precipitation of iron and alumina is effected by
ammonium hydroxide in the usual way. After ignition and weighing, the residue
is fused in sodium pDtassium carbDnate and the irDn reduced and titrated
with a standard solutiDn ,Of pDtassium permanganate in the usual way. JteductiDn of irDn may be accDmplished in variDus ways described in the Chapter
on Iron, VD1. 1.
Separation of Barium from Strontium and Calcium.-Evaporate the filtrate
nearly tD dryness. Take up with abDut 306 m!. of distilled water and add 5
Dr' 6 drops of acetic acid with enDugh ammDnium acetate to neutralize any free
mineral acid present. Heat the .solution and add an excess of ammonium
chrDmate. AllDW the barium chrDmate precipitate tD settle fDr an hour Dr more,
filter on a weighed Gooch crucible and wash with water cDntaining ammoniuhl
chromate until free from strontium and calcium. Then wash with distilled
water until free frDm ammonium chromate. Dry the precipitate at 1100 C.
and ignite gently and finally to a dull red heat. Cow and weigh as barium
chromate. Calculate to barium carbonate by using the factor 0.7786.
Separation of Strontium and Calcium.-To the filtrate from the barium solution add about one mL of nitric acid. Heat the solution and make it alkaline
with ammonia fDllDwed by ammDnium carbonate. Strontium carbDnate and
calcium carbonate will be precipitated. Dissolve the/carbonates in hydrochlDric
acid and reprecipitate frDm a hot solution with, ammonia and ammonium
carbonate. Filter. Wash with hot water. DissDlve the precipitate in the
least amount of nitric acid and wash into an eyaporating dish. EvapDrate to
dryness anq. heat for about an hour at 160 0 C. Pulverize this mass and mix
with a few m!. of ether-alcohol solution. Make several extractions in this way
and decant' intD a flask. The residue is again dried, pulverized and washed
into the flask with ether-alcohDl solution. Digest with shiling. Wash the
residue onto a filter previDusly moistened with ether-alcohol solution. Strontium nitrate remains, which is dissolved in water. To this water solution add
dilute sulfuric acid in slight excess and 'then an'equal volume of alcohol. Allow
the precipitate to settle fDr several hours. Filter on a previously ignited and'
weighed Gooch crucible and wash several times ~ith a 50% alcohol solution.
Dry thoroughly and weigh as strontium sulfate. Compute the sulfate to
carbonate .by using the factor 0.803.
Determination of Calcium.-Calcium remaining in solution is estimated by
making the sDlutiDn slightly ammoniacal, heating and 'adding ammDnium
oxalate with vigorous stirring. AllDw the pr~cip}tate t.o settle and then filter
and wash with a very dilute ammonia soluti6n.- Ignite slowly at first, and

MISCELLANEOUS ANALYSES

2173

finally at full blast. Cool and weigh as calcium oxide. Compute to calcium
carbonate by the factor 1.7847.
Determination of M agnesium.- The magnesium is determined in the filtrate
from the calcium. Boil off the ammonia 'and, while the solution is still hot,
add a slight excess of microcosmic salt with constant stirring. Then make the
solution alkaline with ammonia and stir. Allow the precipitate to settle over
night and filter and wash with a6 dilute ammonia solution. Dry the residue
before igniting. Ignite in a previously ignited and weighed platinum crucible
until white. Cool and weigh as magnesium pyrophosphate. Calculate to magnesium carbonate by using the factor 0.75719.
If only the barium content is desired, a short method may be used. After
the silica is removed, the filtrate is treated with sulfuric acid, added a drop
at a time, and stirred vigorously. The precipitate is allowed to settle and then
filtered on a Gooch crucible which has been ignited and weighed. Wash with
water containing a few drops of sulfuric acid and finally with hot water. Ignite
gently at first and finally at a red heat. Cool and weigh as barium sulfate
and compute to barium carbonate by using the factor 0.8455.
Any residue remaining after the hydrofluoric evaporation of silica is calcu-.
lated as barium sulfate and computed to barium carbonate.
Lime.-Weigh out 0.5 gram sample into a beaker. Cover it with 10 ml.
of distilled water and then 10 ml. of hydrochloric acid and heat until clear.
Filter anc;l wash several times. Transfer the filter paper to a clean platinum
crucible and ignite. Then add about 1 gram of sodium potassium carbonate
and fuse. Take up the fusion in dilute hydrochloric acid and add it to the
main solution which is ..hen transferred to an evaporating dish and evaporated
to dryness. Take up the residue with dilute hydrochloric acid and heat and
filter. Wash with hot water until free from chlorides. , Preserve the residue
and perform a second dehydration as above. Filter and wash and add this
residue to that of the first dehydration and ignite in a previously ignited and
weighed platinum crucible. Cool and weigh as silica. Correction need not
be made on the silica.
To the filtrate from the silica add a small piece of litmus paper and add
ammonium hydroxide until it is just alkaline. Heat to boiling and filter.
Wash the residue with hot water until free from chlorides. Ignite the residue
in a tared platinum crucible. Cool and weigh as (R 20 a) the combined
oxides of iron and alumina. This residue is fused in sodium potassium carbonate or potassium bisulfate, dissolved and the iron reduced and titrated according
to the method given in the Chapter on Iron, Vol. 1.
The filtrate from the iron and alumina precipitation is heated to boiling
and about 20 ml. of a saturated solution of ammonium oxalate is added with
vigorous stirring. Allow the precipitate to settle and then filter and wash with
hot water. Transfer the precipitate and filter paper to the origin;:tl beaker and
dissolve it in 10 ml. of sulfuric acid and 40 ml. of distilled water .. Heat this
solution to 70 0 C. and titrate it with a standard solution of potassium permanganate. Compute the results to calcium oxide.
The filtrate from the calcium is concentrated by boiling. To the hot solution is added a solution of microcosmic saIt with constant stirring. It is then
made strongly alkaline with ammonium hydroxide. Allow the precipitate to
settle for several hours and then filter and wash with a 5% ammonia solution.

?174

MISCELLANEOUS ANALYSES

Dry the precipitate and paper and, when dry, separate them and ignite separately. If there is considerable residue, it is advisablEho ignite in a porcelain
. crucible. The residues are cooled and weighed, their weights combined and
computed to magnesium oxide.
I
If it is desired to determine the strontium presen't, refer to the method
outlined under the examination of Barium Carbonate. I

ANALYSIS OF PRINTING INKS

16

1. Separation of OU and Pigment.-Readily effected by mixture of 3 parts

ether, 1 part benzene. About 50 g. of ink (avoiding the hard surface film)
is placed in a weighed glass tumbler of about 300 ml. capacity, a small amount of
solvent added, and the whole stirred thoroughly until a homogeneous mixture
is obtained. The glass is then filled with the solvent to about ! in. of top,
and the whole again stirred. It is next placed in the metal cup of the centrifuging machine, and the space between the glass and metal cups filled with wate:t
in order to equalize the pressure of the liquid inside the glass. Placing a rubber
disk at the bottom of the metal cup has been found to lessen materially the
danger of breakage. The metal cup and contents are then counter-balanced,
most conveniently by either a second sample of the same ink or another sample
of ink, and then both are placed in the machine. For web-press and fiat-bed
inks 2000 r.p.m. for 10 minutes is sufficient. Where carbon black has .been
used, it has frequently been necessary to run the machine at 2600 to 2800 r.p.m.
for 20 or 30 minutes. The clear liquid is decanted through a pleated filter into
a glass bottle, a further quantity of solvent a<;lded, 'and the process repeated.
UsuaJly three treatments suffice to give practically complete separation. The
glass and contents are dried at about 90 C. and on cooling, reweighed. The
increase in weight is the pigment, which is calculated to percentage. The
amount of pigment on filter paper should be negligible if centrifuging has been
efficient.
~
2. Analysis of the Oil.-May contain linseed oil, hard gums,"rosin, rosin oil,
mineral oils, and bituminous substances. The last mentioned,. when pre'sent,
must be judged largely by their color. Sufficient of the solution from the separation of the oil and pigment to leave a residue of about 5 g. is evaporated in
a weighed beaker; 50 ml. of normal alcoholic potash is added, the beaker
covered with a watch glass, and heated on a steam ~ath for several hours, stirring
frequently to assist saponification. When the latter is complete, the watch
glass is removed and the alcohol distilled off. The residue is 'transferred to a
separatory funnel with successive portions of water, using in all about 100 mI.,
and extracted with petroleum ether until no further oil is obtainable. Four
extractions are usually sufficient. The petrol.~u.!Il.ether fractions are united in
16

By J. B. Tuttle and W. H. Smith. Tech. Paper No. 39, U. S. Bur. of Standards,

MISCELLANEOUS ANALYSES

2175

another funnel, washed with water until the wash water gives no further alkaline
reaction, filtered into a weighed beaker, the petroleum ether distilled off, and
the residue dried at 95 C., cooled, and weighed. If this unsaponifiable matter
is over 2%, it indicates the presence of something else than linseed oil and hard
gums. The wash water from the first two washings should be united with the
water layer in the first separatory funnel. This unsaponifiable matter is
tested for rosin oil by heating a small portion of the oil with 10 m!. of acetic
anhydride, allowing it to cool, and adding a drop of H 2S0 4, sp.gr. 1.63. A violet
coloration indicates rosin oil.
If the test for rosin is positive, the alkaline water solution which has been
extracted with petroleum ether is made acid with HCl (there is usually sufficient
dye present from the ink to act as indicat~), and the fatty acids which are thus
liberated are extracted with successive portions of ethyl ether. These extracts
are united, washed free from acid and salts, and evaporated in a small beaker.
A quantitative determination of the rosin may be made as follows: The fatty
acids are dissolved in 20 ml. of 95% alcohol, a drop of phenolphthalein is added,
and then strongcaustic soda (I-N aOH, 2-H 20) until the reaction is just alkaline.
The solution is heated for a few minutes, allowed to cool, and then transferred
to a 100-ml. stoppered graduated cylinder. The latter is filled to the 100 ml.
mark 'with ether, 2 g. of powdered silver nitrate crystals is added, and the
mixture shaken vigorously for 15 minutes. When the insoluble salts have
settled, .50 m!. of clear solution (containing the silver salts of rosin) is pipetted
off into a second 100-m!' cylinder, and shaken with 20 ml. dilute HCl (one acid
to two water). The ethereal layer is drawn off, and the aqueous layer is shaken
twice with ether. The ether extracts are united, washed with water, and the
'ether distilled off in a weighed beaker. The residue (rosin) is dried at 100 to
115 C., cooled, and weighed. The results are calculated on basis of the original
weight of oil.
3. Analysis ofthe Pigment. (a) Black Inks.-Ignite, at the lowest possible
temperature, a weighed quantity in a porcelain crucible (presence of Pb
prohibits Pt), the loss on ignition represents lampblack, the carbon of boneblack (should there be any present), aniline dyes, and undissolved oils or gums.
Prussian blue is decomposed by heat, part of it being volatilized, the iron remaining as iron oxide. The residue from the ignition contains any added mineral
matter of the pigment, lead or manganese from the driers, iron oxide from the
Prussian blue, or iron oxide added as such (the so-called magnetic pigment),
calcium phosphate if boneblack is present, and alkali or calcium carbonates
from the soaps present. The ash is analyzed qiIantitatively for insoluble
matter, Pb, Fe, Mn and Ca. One quarter gram of the ash is heated to dull
redness in a porcelain crucible, for a few minutes, cooled in a desiccator, and
weighed. This is transferred to a 250-ml. beaker, using conc. HCI to dissolve
any material that may stick to the crucible. About 25 ml. of conc. HCI is
added, the beaker covered with a watchglass, and after heating until as much
as will go in solu'tion is dissolved, the cover is removed and the solution evaporated to d.yness. The residue is moistened with a few drops of concentrated
HCI, 50 to 75 ml. of boiling water added, and the solution is filtered, washing
thoroughly with hot water. The filter paper and residue are ignited and
weighed, and the product is called" insoluble matter."
Fifty ml. of 10% H 2S0 4 is added to the filtrate from the previous determi-

2176

MISCELLANEOUS ANALYSES

nation and evaporated down until the solution fumes strongly. This is cooled,
diluted carefully with about 100 to 150 ml. of water;, and heated on the steam
bath until any basic ferric sulfate which sometimes separates is redissolved.
The precipitate containing the lead sulfate is now filtered off. It is dissolved
in ammonium citrate or acetate, filtered from any in$oluble. matter, the filttate
made strongly acid with H 2S0 4 and the precipitated lead sulfate filtered off
on a Gooch crucible, ignited, and weighed. The insoluble matter from the
ammonium acetate solution should be examined for Ca and Ba.
Another method for the determination of the lead is to neutralize almost all
the acid present with sodium carbonate, saturate the solution with hydrogen'
sulfide, filter off the precipitated lead sulfide, dissolve it in fairly strong nitric
acid, and determine the lead as sulfate by the addition of sulfuric acid as above .
In this case, solution in ammonium acetate is omitted. The former method
is of advantage where qualitative tests show that there is very little manganese
present, and it is desired to determine only the iron .. The iron is reduced to
the ferrous condition by passing the solution through a Jones reductor, and the
ferrous sulfate titrated with a standard solution of KMn04, which has been
standardized with sodium oxalate.
Iron is separated from manganese and other metals which may be present
by precipitation with ammonia, the precipitate being filtered off, dissolved in
HCI, reprecipitated with ammonia, and again filtered. It is now dissolved in
HCI, sulfuric acid added, and evaporated down until all the HCI is remo:ved;
the solution is diluted and the iron determined as before. Before the additi~n
of the ammonia, if H 2S has been used, the solution should be boiled until it is
removed, and nitric acid added to oxidize the iron to the ferric condition.
H 2S is now passed into the ammoniacal solution from the iron precipitation.
This is allowed to stand over night, and the precipitate, if there is any, is' .
examined for manganese. Should there be much manganese, the sulfide can
be filtered off, and the quantitative determination made by conversion into
the pyrophosphate.
If it is desired to determine the calcium, this can be done after the filtration
from the ammonium sulfide. (If phosphates are' present-as, for instance,
if boneblack is present-a basic acetate separation is required.) In either case
the lead should be separated by H 2S. The filtrate from the manganese sulfide
is heated on the steam bath until the H 2S is removed, ammonia and ammonium
oxalate are added, and the precipitated Ca oxalate is determined either as oxide
or s u l f a t e . .
'-The percentage of ash will be of great assistance in determining the nature
of the pigment. Black oxide of iron is only slightly changed on heating, being
completely oxidized to ferric oxide. Boneblack is composed largely of calcium
phosphate, yielding the greater part of its weight as ash. The presence of
any large amount of phosphoric acid will be sufficient evidence that boneblack
has been used.
{
Prussian blue should be tested for qualitatively in the dry pigment. For
this purpose 1 g. of pigment is moistened with 2 or 3 m!. of normal alcoholic
potash, heated on the steam bath until the alcohol is removed, 5 m!. of water
added, and the insoluble matter filtered off. The filtrate' is made acid with
HCI, and filtered again if necessary. When f~tiC chloride is added, a blue
precipitate will be obtained if Prussian blue is present. If blue dye obscures

MISCELLANEOUS ANALYSES

2177

the reaction, the solution is again made alkaline and filtered, made acid with
Hel as before, and copper sulfate added. The precipitate is filtered and washed
thoroughly. In this case we will have r~dish-brown copper ferrocyanide.
It is advisable, in case of doubt, to add a small amount of Pruss ian blue to the
pigment, and make a check test. In the absence of black oxide of iron we may
assume that all of the iron in the filtrate is due to the Prussian blue. The
percentage of Fe 2 0g in the ash, multiplied by 'the percentage of ash in the pigment, multiplied by the factor 1.53, will give, roughly, the amount of Prussian
blue present. The factor 1.53 is obtained by the ratio Fe7(CNhs to Fe 2 0g.
An approximate determination of the dyes can be made by extracting with
alcohol. When the presence' of oxide iron is suspected, 1 g. of pigment is
wrapped in filter paper and the dye extracted with alcohol.. When all of the
dye has been extracted, the paper and contents are dried and the nitrogen is
determined in the residue by the Kjeldahl method, From the nitrogen thus
determined the Prussian blue is calculated, using the factor 3.41. The Fe 20a
present in this amount of Prussian blue is deducted from the total Fe 2 b s found
in the ash. The remainder will be the percentage of iron from the magnetic
oxide, The formula of the latter is theoretically Fe S04, and proper calculation
should be made.
(b) Blue Inks.-A weighed quantity of pigment is ignited as under black
pigments. The ash is analyzed same as before, determining only lead, manganese, and iron if the qualitative tests show that Prussian blue is present. The
lead and manganese are reported as metallic driers, the iron is calculated to
Prussian blue, and the remainder reported as mineral filler. The presence of
ultramarine will be shown by the blue color of the ash. In this case the ash
is reported after deducting the Pb+Mn. Dyes are determined by alcohol
extraction.
(c) Red Inks.-Vermillion (mercuric sulfide) is readily detected by taking
a small quantity of pigment, covering it with 4 or 5 ml. aqua regia, and heating
gently. This is diluted with five volumes of water, filtered, and stannous chloride added to the filtrate; a grayish precipitate of mercury will be formed' if
even a very small amount is present.
Quantitative method: Dissolve the mercuric sulfide in aqua regia, and after
nearly neutralizing the diluted solution to precipitate the mercuric sulfide with
H 2S, weigh the precipitate on a Gooch crucible, observing all the precautions to
eliminate sulfur which separates during the precipitation.
The following procedure has also been found of value: One g. of the pigment
is treated with a.slight excess of ammonium sulfide. Sodium hydroxide is theri
added, while stirring. The beaker is placed upon the steam bath, adding more
alkali if necessary, until all the mercuric sulfide'has passed into solution. An
excess of alkali should be avoided. The solution is allowed to cool, filtered,
and the residue washed thoroughly. To the filtrate sufficient ammonium nitrate to reprecipitate the mercuric sulfide is' added, and it is then boiled to
expel ammonia. The precipitate is allowed to settle, and the supernatant
liquid decanted through a weighed Gooch crucible. The residual mercuric
sulfide is boiled with a little sodium sulfide solution to remove free sulfur, and is
then trltnsferred to the crucible where it is washed with hot water, until it no
longer reacts with slIver nitrate solution. It is dried at 110 and weighed. The

2178

MISCELLANEOUS ANALYSES
I

pigment is ignited, and the ash analyzed JQr lead. and manganese. The rest
I
of the ash is reported as mineraLfiller.
(d) Green Inks.-The coloring matter may be chrome green, green lake, or
dye. Some of the darker shades are obtained by tl;le addition of lampblack.
The ash of the pigment is determined as usual. Part of this ash is taken and
tested qualitatively for chromium. If present, the ash should be tested for
the following substances: PbCr04, PbS0 4, PbO, BaS04, CaS04, Fe 20a, and
Mn a04. For the determination of sulfur, 0.250 g. of the ash and 5 g. of a
mixture of equal parts of potassium nitrate and sodium carbonate are fused
in a porcelain crucible over a sulfur-free flame. The cooled mass is extracted'
with hot water and filtered. The filtrate is acidified with HCI, heated to
boiling, and 10 ml. 10% barium chloride added. After standing over night, .
the precipitated barium sulfate is filtered off, ignited, and weighed. The
solution should be sufficiently acid to prevent any significant contamination
of the barium sulfate with barium chromate.
For the determination of barium, the insoluble matter is dissolved in HCI,
the solution made nearly neutral with sodium carbonate, and H 2S is passed
into the solution until all the lead is precipitated. The lead sulfide is filtered
off, the filtrate heated to boiling, and 10 ml. of 10% H 2S0 4added. The barium
sulfate is treated as directed under the determination of sulfur.
A fresh portion of ash is mixed with sodium peroxide and fused in a nickel
crucible. The cooled melt is dissolved in hot water and filtered. Carb<tU
dioxide is passed into the filtrate, and the latter heated again on the steam
bath in order to precipitate any lead which may be held up by the caustic
alkali. Any insoluble matter which may separate is filtered off. The filtrate
- is made strongly acid with HCI, KI added, and the liberated iodine titrated
with a standard Na 2S20a solution. From the amount of thiosulfate used the
amount of CrOa present is calculated.
The two precipitates from the previous determination are combined and
used for the determination of Pb, Fe, Mn, and Ca. They are dissolved off the
filter paper with HCI, the solution is nearly neutrali~ed.with sodium carbonate,
and H 2S passed into the solution. The precipitated lead sulfide is filtered off,
dissolved in nitric acid, and determined as sulfate, as directed under black
pigments.
The filtrate from the lead sulfide is heated until all the H 2S is boiled off.
If any sulfur separates, it is filtered off. Two or three ml. of nitric acid is
added ana the solution again heated. It is then made alkaline with ammonia
and filtered. The precipitate is dissolved in HCI and reprecipitated with
ammonia, the precipitate again filtered off and the filtrate united with that
from the first precipitation. This solution is reserved for the determination
of manganese. The second precipitate of iron is dissolved in HCl, converted
into the sulfate, and the iron determined as under, black pigments.
. The united filtrates from the iron precipitatlon are saturated with H 2S,
allowed to stand over night, and then filtered. If there be a sufficient amount
of mangan~se present, it can be determined quantitatively as pyrophosphate;
otherwise qualitative identification will be sufficient.
It is safe to assume that all chromium was pre_s~nt originally as lead chromate, and it should be so calculated. The irQn..oxide should be calculated to
Prussian blue, provided there is a positive qualitative test. Any barium present

MISCELLANEOUS ANALYSES

2179

should be calculated to sulfate; if' there is any question as to its being originally
present as carbonate, the ash of the pigment is treated with very dilute HCI,
the solution filtered and the filtrate tested for barium. Barytes is difficultly
soluble in cold dilute HCI. In the absence of barytes the sulfur present is
calculated to lead sulfate. The excess of lead over that required for the lead
chromate and sulfate may be considered as drier . China clay may be present
either as an added part of the chrome green, or as the base of a green lake. In
such cases, the undetermined portion of the ash should be reported as mineral
filler. Green dyes are determined by extraction as usual. In the absence of
chrome green the pigment is ashed, and the ash analyzed for lead and manganese only, the balance being reported as mineral fillers. If h~pblack has
been used to produce a dark shade of green, it can be tested qualitatively by
taking a small portion of the pigment, treating it with strong alkali, and
,filtering through a Gooch crucible, washing first with hot water, and finally
with moderately concentrated HCI. Lampblack will show a black residue,
which will disappear on ignition. It is generally classed with the volatile
con.stituents, which are then reported as aniline dye, lampblack, undissolved
oil, etc.
:
The analytical me,thods included the essential constituents of inks: the
determination of iron, tannic acid, gallic acid, total solids, ash, acidity, etc ..
The specific gravity of the ink will indicate the amount of dissolved material
present. F. F. Rupert, in,a paper on the examination of writing inks (J. Ind.
Eng. Chem., 15, 489, 1923), emphasizes the importance of the physical tests
of color, permanence, stability and noncorrosiveness. The following tests
are suggested.
Streak Tests.:-A sheet of a 'good quality of bond paper placed on a pane
of glass and held by clamps serves for taking the streak. The streak is made
by a pipette 25 cm. long, 3.5 mm. in diameter with a file mark 62 mm. from
the tip. Ink filled to the mark (0.6 mi.) is tested by drawing the pipette over
the surface of the paper, at the same time allowing the ink to flow.
Penetration and Fluidity.-The ink should penetrate into the fibres of the
paper but shollld not pass through. A normal ink will give an oval head
(where the streak begins) and the width of the line will be uniform. A poor
ink gives a wide head and the streak narrows rapidly. Stickiness should be
observed.
Temporary Color.-The mark should be dark enough to be easily seen.
A speckled appearance indicates that the ink contains a dye which is not in
solution.
Permanent Color.-The streak is exposed to diffused light for 7 days in
an atmosphere free from chemical fumes and dust. A comparison is made
with a standard ink. The paper is now cut into crosswise strips 1 inch in
diameter. The top and bottom strips are set aside as blanks.
Exposure to Light.-Strips containing the streaks are exposed to sunlight
or ultra violet rays for a period 48 hours to two weeks or longer and the ink
thus exposed compared with the' blanks.
,
Exposure to Weather.-The exposed sample is inspected weekly.
Exposure to Water and Reagents.-Tests in water, in dilute ammonia
(1 : 10), HCI 2% solution, bleaching powder or sodium hypochlorite of 0.005
N available CI.

2180

MISCELLANEOUS ANALYSES

Stability.-Note sediment in original. Filter two portions, one into a

bottle which is later stoppered, the other portion in,to an open bottle covered
by a piece of paper. Filter each after a period of a week or more and note
sediment on filter paper.
l
Corrosion.-The effect of the ink on steel pens, Immersed in the ink for a
period of time, at least one week. The pens are we~ghed before and after the
test.
Comparison and Rating.-The system is subordinated to the knowledge of
the purpose for which the ink is to be used. For further details the reader is
directed to the article by Rupert.

DETERMINATION OF PEROXIDES 17
DETERMINATION OF HYDROGEN PEROXIDE.
KINGZETT'S IODIDE METHOD 18
The method depends upon the reactionH202+2KI+H2S04=I2+K2S04+2H20.
Procedure.-Approximately 2 grams of potassium iodide are dissolved in
200 ml. of water, conveniently in an Erlenmeyer flask, and 30 ml. of dilute
sulfuric acid (1 : 2) are added. Three good drops of N Ammonium Molybdate
are added to catalyze the reaction. Ten rol. of hydrogen peroxide solution (the
sample having been diluted to contain approximately 0.6% by weight of
H 20 2) are run in from a burette, agitating the mixture during the addition.
After standing five minutes, the liberated iodine IS titrated with N /10 thiosulfate.
DETERMINATION OF HYDROGEN PEROXIDE BY THE ARSENIOUS
ACID METHOD OF JAMIESON]9
The method is based on the reactionsAS203+2H202=As20s+2H20
and
AS203+KI03+2HCI=As20s+IQ+KCI+H20.
Procedure.-A measured quantity of N /5 solution of As 20 3, which must be
in excess of that required by the peroxide taken for analysis, is placed in a
500 ml. glass stoppered bottle, and 10 ml. of 10% NaOH agded, followed by a
Selections by W. W. Scott.
J. Chern: Soc., 1880, 792.
19 Am. Jour. ScL, 44, 150-2 (1917).

17
18

MISCELLANEOUS ANALYSES

2181

measured volume of the peroxide from a burette (15-20 ml. of the solution
made by diluting 50 ml. of the commercial peroxide to 500 ml.) with gentle
agitation of the contents of the bottle. After standing two minutes, 40 m!. of
concentrated Hel are added and the stopper inserted in the bottle. The
contents of the bottle are shaken violently, holding the stopper in place. The
stopper is now cautiously released and the confined gas allowed to escape.
Six to seven ml. of chloroform are added and the unoxidized As 20 S is titrated
with N /5 KIO s solution, shaking the closed bottle after each addition, until
the iodine color in the chloroform has disappeared.
The m!. AS 2 0 3 solution used minus the ml. equivalent of KI0 3 gives the ml.
of arsenous acid required by the hydrogen peroxide.
1 ml. N/5 AS 20 3 = 0.0034016 gram H 20 2.
This method is not influenced by the presence of organic preservatives as
is the permanganate method .which follows .
. PERMANGANATE METHOD FOR DETERMINATION OF
HYDROGEN PEROXIDE
The procedure depends upon the reaction5H202+2KMn04+4H2S04=2KHS04+2MnS04+502+8H20.
Procedure.-Fifty ml. of the commercial peroxide are diluted to 500 ml.
10 ml. of this diluted solution are taken for the test. This sample is further
diluted to 400 m!. in a beaker, 10 ml. of dilute sulfuric acid (1 : 4) are added
.and the mixture titrated with N /10 KMn04 reagent to a pink coloration.
Should the first drop or two of permanganate cause a pink color it indicates
. that an insufficient quantity of sulfuric acid is present and an additional
amount should be added .
. 1 ml. N /10 KMn04 = 0.001701 gram H 20 2.
NOTES.-If it is required to report the number of volumes of oxygen liberated by
one volume of the peroxide "per cent per volume," multiply the per cent weight of
H 20 2 by 112 and divide the result by 34.
Since H 20 2 decomposed to H 20+O, one gram molecule of peroxide sets free one
gram atom of oxygen, equivalent to 11,200 ml. Therefore 100 grams commercial
H,02=%H 20 2 X11,200 divided by 34, and 1 gram = 1/100 of this.

COMPLETE ANALYSIS OF SODmM SULFIDE 20

Sodium Sulfide.-Commercial sodium sulfide is marketed in three forms:
Fused solid, flaked solid, and crystals. For manufacturing, control analyses
may also be called for on various liquors. The same general methods can be
used in these cases by suitably adjusting the sample and aliquot sizes.
20

By courtesy of The Dow Chemical Co., Midland, Michigan.

MISCELLANEOUS ANALYSES

2182

NOTE I.-The following handling of sample may be used conveniently in many cases,
particularly with moist material: Weigh a large sample, sa:y 50 or 100 grams accurately,
dissolve, and make to 500 or 1000 mI. volume. The "water insoluble" should be run
on separately weighed samples as indicated, the filtrate being discarded. The water
insoluble material should be allowed to settle, on the l~fge made up sample, and an
aliquot giving 10-12 g. made to 500 m!. for the sulfides, sUlfite, etc., proceeding as outlined. Suitable aliquots taken from the strong made-up soluti~n can then be taken for
the Na 2CO a and NaCI determinations.
.

Water Insoluble.-Weigh accurately 10 to 12 grams .of the 60% sulfide, or

estimate an equivalent amount of other sulfide materials; transfer to a stoppered flask and dissolve in approximately 150 ml. of water and stand for several
hours or overnight in a warm place, as on a radiat()r or top of an oven. If
necessary to proceed without much delay, warm to about 70 C., stand an hour,
and filter through a 30-ml. tared Gooch crucible. A fairly thick pad must be
used. The filtrate obtained should be clear. Wash with water, and dry at
100 C. Weigh the insoluble residue.
Grams gain X 100 _at HOI I bl
. ht - /0 2 nso u 8'.
SampIe welg
Make the filtrate to 500 ml. volume for the analysis.
FeS.-This is included with the water insoluble material, Fe being present
t
as FeS. When Fe is called for, the following method may be used:
Weigh about'10 grams, dissolve in 100 ml. of water, add 50 ml. concentrated
HCI and boil off H 2S. Add a few :fil. of HNO a and then bromine-potassium
bromide solution to clear from sulfur and to oxidize the' iron. Boil off bromine,
make alkaline with NH 40H, filter, wash and ignite .

Grams Fe 20aX110 =% FeS.

Grams sample
In case AbOa is present, the precipitate (R 20 a) should be fused with KHS0 4,
dissolved, reduced, and titrated with KMn04.
'
Ordinarily these determinations are not called for unless the water insoluble
runs extremely high.
Use 25.00 ml. aliquots of the 500 ml. solution for all the subsequent determinatioI}s, unless othel' aliquots or separately weighed ~mples are called for.
A. Total 12 Titration.-Add \ known excess (100 ml.) <liN/IQ h to a 250-ml.
beaker, acidify with 15 ml. N /1 HCI, then pipette in an aliquot of the sample,
slowly, while stirring. Titrate back with N /10 N a 2S20a solution, using one ml.
of starch solution at the end after the iodine color is visibly faded. :Add a few
drops of methyl red indicator to make sure the solution is acid at the finish.
If not, double the amount of N /1 HCI called for, and repeat. 12 minus Na2S20a
= A. This is Na 2S plus NaHS plus Na2S20a plus Na 2SOa.
NOTE 2.-Methyl red indicator used here is made by heating a small amount of the
dye with water almost to a boil. From one to five mI. of the cool settled solution may
be used.
,
NOTE 3.-Starch solution used here is made by mixing 5 g. soluble starch with 10 mI.
of water, adding 20-30 mI. N/1 NaOH to clear,.tit"rif,ing to phenolphthalein neutral
point with N /1 HCI, and diluting to 500 mI.

MISCELLANEO'US ANALYSES

2183

B. Sulfides.-Place 10 to 12 mI. of ammoniacal zinc chloride reagent into

250-mI. beakers and pipette in aliquots. Mix by swirling and add enough water
to thin out the precipitate. Filter off the ZnS and wash well with warm water.
Place a known excess of N /10 12 (100 mI.) into the ,original beakers. Acidify
the iodine solution with a few mI. of 1: 1 HCl, and transfer the precipitate
to the beaker over th~ edge of the funnel. This can be done very completely,
and without difficulty. Then puncture the ti~f the paper, and wash the paper
for small amounts of adhering precipitate witli warm dilute HCI. The paper
should be left on the funnel and the acid wash collected in the original beaker.
This should be kept well stirred and small additions of HCI made till the reaction is complete. Titrate the excess 12 with Na 2S20a as above. Net 12= B. ~
This is N a 2S plus N aHS.
Save thl;l filtrates and washings from the ZnS ppt. for C and D. Any
turbidity noted, especially on the top is not due to sulfur or zinc sulfide, but to
zinc hydrate. At this point no harm is done by this occurrence. The N /1
ammoniacal zinc chloride will be hydrolyzed on diluting sufficiently.
NOTE 4.-Ammoniacal zinc chloride reagent. Dissolve 41 g, of c.p. zinc oxide in
HCI. The HCI should be in just enough excess to readily dissolve the oxide while being
,+ added slowly with stirring. Then add NH.OH slowly to make approximately neutral,
add 200 ml. of water, 150 ml. of concentrated. NH.OH, and 65 g. of NH.Cl. Make to one
liter. This is roughly a normal solution. Use to precipitate sulfides by adding a slight
excess, testing the filtrate with more reagent to be certain that precipitation is complete;
with an iodine titration of 80 to 95 ml., 10 ml. of the reagent would be sufficient. It should
not be diluted before use, as Zn(OH)2 will precipitate. During Itn analytical operation this precipitate should not be confused with ZnS. The hydrate will appear during
the neutralization, especially if much ~f the reagent was not used up as a precipitant.
Cadmium chloride reagent may be used by those who are more familiar with its use.

e and D. Thiosulfate and Sulfite.-Make the above mentioned filtrates

neutral to methyl red using first N /1 HCI and then N /10 near the end. Disregard the formation of Zn(OH)z, continuing the titration to the pink endpoint. Then add 1 mI. of starch indicator and titrate one of the two filtrates
with N /10 Iz. Call the mI. used C. This equals Na 2C 20 a plus Na 2SOa.
To the other filtrate add 5 mI. of 10% BaCl 2 and filter onto a wet asbestos pad
on a Gooch crucible; wash with water, and transfer pad and precipitate back
to the original beaker. Add an excess of N /10 I 2, acidify with a few mI. of
N /1 HCl, and stir well. Titrate back with N /10 Na 2S20a. MIlo N /10 I2
minus N /10 Na2S203=D. This is Na 2SOa. The formation of a precipitate
at this point is a test for sulfite.
.
E. Alkalinity, Polysulfide Sulfur, and Na 2S04.-Titrate aliquots with N(lO
HCl, shaking or stirring well throughout the titration, and adding a drop of
methyl orange indicator from time to time. The indicator has a tendency to
fade, the tendency being overcome as the end-point is approached. Record
the mI. to give the pink end-point as uncorrected E. The applied correction for
carbonate and sulfite will be given under" Calculations," to- follow.
N ow add two to five mI. of the a(lidJrom the burette, in excess of the neutral
point. If the solution remains perfectly clear, after several minutes standing,
no polysulfide is present. If a cloudiness appears, warm to approximately
70 C. and hold thus for an hour; filter through tared lO-mI. Gooches, being
careful to transfer, all adhering precipitate. Dry at 80 C. for two hours and
w~igh back.

2184

MISCELLANEOUS ANALYSES
. GramsX100
. ht = % Polysulfide sulfur.
,
SampIe welg'

Add 5 ml. of 1-1 HCI to the filtrate and washings from Polysulfide sulfur
and boil for three minutes. Add 10 ml. of 10% Ba8b2H 20. Let stand an
hour. Filter through No. 42 Whatman filter pap~r, or one of equal qllality,
wash well, ignite, and weigh BaS04.
g. BaS04X:6086 X 1'00 at N SO
X/'o 112 4
g. SampIe
.

Na 2C0 3.-Acid sulfide is too weakly acid to decompose'carbonates. Sodium

carbonate may be present in greater or lesser amounts. The amount of sample
to take depends on the amount known to be present or expected. ~or 2 to 8%,
5 gram samples accurately weighed are suitable. Dissolve a pair in 75 ml.
of C02-fre~ water, add 25 ml. of 25% ammoniacall\mmonium acetate solution.
(Thie solution has enough concentrated NH 4 0H aqded to just turn pink with
phenolphthalein indicator added.) After mixing, acid 20 m!. of 10% BaCb solution, stopper well, mix, and stand overnight. In the morning, filter through asbestos, using a 25- or 35-ml. Gooch crucible. 'Wash well and transfer precipitate'
and pad to a 500-ml. assay flask. The pad can readily be loosened with a
knife point. Add 100 ml. of water and 100 ml. of N/10 HCI; also several ml. of
methyl red indicator. Boil for fifteen minutes. With high carbonate; more
acid may be necessary; this will be indicated by changing of the indicator 'to
the alkaline yellow color. Cool and titrate back with N/lO NaOH. (MIs.
N/IO HCI-mls. N/10 NaOH=" T.") In additi<m to the back titration, a
further correction on the acid must be made for sulfite. Adjust the titrations
to the same size sample as follows:
g. Sample for Na 2CO a D
"'-~-=--::--:-::-=---=-",... X
= ml
g. Sample for Na 2SOa
.,

Correction for Na 2S03: Call D',

.'

N/10 RCI-eN/10 NaOH+'I)') X.0053X100 =iJ{ Na COa.

g. Sample
0
2
In many cases too much time will be lost by givirlg the overnight standing as
specified . For control purposes, the time may be shortened to as smlill a
period as ! hour, introducing a plus error usually undei' 10% of the Na 2C0 3
present. A somewhat longer standing period, sll,y 4 hours, gives the BaS
which forms to some extent, a better chance to decompose.
Calculations:
Thiosulfate. (C - D) X .0158 X 100 _ at N (~O' ,
- /0
a2"2 3.
Grams sample
SulfiteDX.0063X100 _ 01 N S(
- /0 a2 )3.
G ramssamp Ie

Sulfide and Acid Sulfide- .

-' Correct E for carbonate and sulfite by subtracting the sum of the titrations
adjusted for sample size.
'

MISCELLANEOUS ANALYSES

2185

g. Sample for B
.
S
1 f N CO X T = (N a 2C03+ N a 2S03) mls. on B baSIs-Call T'.
g. amp e or a2
3
.

E - T' = ml. corrected E, or E'.

E'-(B- E') X.0039XIOO
S
1
=
g. amp e

% Na 2S,

where E'

.
IS

less than B.

Then

. (B-E')X.0056XIOO=% NaHS.
g. Sample
When E' is greater than B, the mixture is caustic and normal sulfide, then
(E'- B) X.0040XIOO =

% NaOH

g. Sample
and
B X .0039 XIOO
g. Sample

Where E' is

l~ss

% Na 2S.

than B,

B X .0039 X 100
g. Sample

% Na 2S+NaHS, calculated as Na 2S.

The following equations and example will clarify the above calculations:
~2Nal+S.

(1)

Na 2S+1 2

(2)

(3)

NaHS+i 2 ~Nal+Hl+S.
N aHS + HCI~ N aCI + H 2S.

(4)

Na2S+2HCI~2NaCI+H2S,

with 12 -(1) H=Na 2S=2, 1 ml. N/IO 1 2 =.0039 g. Na 2S.

(2) H=NaHS=2, 1 ml. N/IO 12 =.0028 g. NaHS.
with HCI-(3) H=NaHS=I, 1 ml. N/lO HCI=.0056 g. NaHS.
(4) H=Na 2S=2, 1 ml. N/IO HCI=.0039 g. NazS;

IllustrationNa 2S

NaHS

.0039
- N 110 12
.0028

= 80.0 ml.

..0028

E'

- N /10 HCl.= 62.0 ml.

.0039
Na 2S

.0056
+

NaHS

E' is less than B - Sample = .5000 g .

..

2186

MISCELLANEOUS ANALYSES

Computing by above methodsNa2S+NaHS=6204%

Na 2S=34.3%
NaHS=20.2% I
20.2 X 39/28 = 28.1 = N aHS calc. as N a 2S.
28.1+34.3=6204. Na2S+NaHS.
80.0 X .0039 X 100 = 62 4
.5000
.

" called
It will be noted that the" total iodine titration," or reducing value,
A titration has not been used in these calculations, and that this was not called
for in duplicate. This titration gives a checK on the work. A should always
be numerically greater than B. A - B = C. After titrating C, or NazSzO a
plus Na 2SOa, an indirect valuation of Na 2S03 may be made. Remove the
blue color with a drop of N /10 Na 2S20a and titrate with N /10 NaOH recording
the ml.
A - B-(C-2/3 m!. N/10 NaOH) X.0063 X 100 _01 N SO

- 10 az a
g. SampIe
NaCIWeigh accurately samples of 2 to 4 grams and dissolve in 400 m!. of water .
.Add 1: 1 nitric acid till approximately 1.0 ml. in excess orthe neutral point hl+ve
been added. Boil gently till the odor of HzS and S02 has disappeared. Cool.
Filter off coagulated sulfur. Add a crystal or two of c.p. chromic acid to the
filtrate and bring to a boil. Precipitate chloride by addition of a slight excess
of 2% silver nitrate solution. With low chloride a few hours standing will give
a better formed precipitate. Filter the precipitate on tared Gooch crupibles,
20 or 25 ml. size, and wash free from soluble salts. Dry at 120 to 1400 C. for
two hours. Weigh back AgCl.
g. AgCIXo408X100
g. Sample

o/c
.0

NabI.
I

Sodium acid sulfide (sulfhydrate) liquor. is used in several commercial

reactions. The .methods and calculations will apply except in the case of free
hydrogen sulfide. Using the same letters for, designating various titrations, a
sample calculation is as follows:
"13.9291 grams to 5GO ml. volume and 25 m!. aliquots u.sed,fpI: sulfides, etc.
(.696 g.).
8.174 grams, direct for NazCO a
B=57.50 ml. NjlO 12
C = .38 m!. N /10 12
E=29.87 ml. N/10/HCI.
Na 2CO a =33.05 ml. N/IO HCI.
.696
.
8.174 X33.05=2.81 ml. carbonate correctIOn.
E corrected for carbonate = 27.06 mI .'

27.06X..~g:6X100 =21.7% NaHS.

MISCELLANEOUS
Sulfide as

AN~ YSES

2187

% H 2S57.50X.0017X100 =14 0501 H~

.696
. 10.

Factor N aHS 'to H 2S 34.08/56.07 = .607.

Above NaHS (21.7%) equiv. to 13.16% H 2S.
14.0.5-13.16=.89% net H 2S.

ThiosulfateCarbonate33.05 X.0053 X100 =21501 N CO

8.174
. 10 a2 3.

Summary--

% H 2S ..........................89

% HaHS ....................... 21.7

% Na 2C0 3 2.15
% Na 2S203...................... .86
NOTE 5.-Credit for above methods must be given to Sutton's Volumetric Analysis,
DuPont Publication, in J. Ind. Eng. Chem., 17, 176 (1925). Grasselli Chemical Co.,
12-18-29, and various Dow Chemical Company chemists.
-

We are indebted to A. W. Beshgetoor, Chief Analytical Chemist, the Dow Chemical

Company, for the section on the complete analysis of sodium sulfide.

PART II
SPECIAL
TECHNIQUES APPLICABLE

TO
DIVERSE SUB STAN CBS

ACIDIMETRY AND ALKALIMETRY 1

The volumetric determination of a free acid or a free base may be accomplished with rapidity and accuracy by neutralization with a known quantity of
,tandard base or alkali as the case may require. The point of neutralization or
., end-point" is accertained by means of certain compounds called indicators,
which have a different color in acid solutions than in alkaline solutions, the'
point of transition from one color to the other occurs at the point of neutraJization. This end-point may also be recognized by the electro metric method by
measuring the change of potential that occurs with the change of concentration
of the hydrogen ions in the solution. This potential change is usually large
and abrupt at the end-point.
Indicators.-The accuracy of acidimetric and alkalimetric titrations depends, to a considerable extent, on the choice of the indicator used. Indicators
are usually dyestuffs, which have one color in an acid solution and another in
alkaline solutions. The color change should occur as near as possible to the
equilibrium point reached in neutralization. The color change of certain
indicators may be due to, a difference in color of the dissociated from the undissociated compound or to the presence of certain grou'ps of radicals within
the molecule of the compound, called chromophores, which show one color ill
acid solutions and, by internal rearrangement, another color in basic solutions.
This color change depends upon the hydrogen ion concentration of the solution
rather than the absolute amount of acid present. The pH range of color
change differs with different indicators as is shown in the table given on th,e
following page. It is generally the case that indicators whose color change occurs with a hydrogen ion concentration greater than pH 7 are sensitive to
weak bases, while those whose color change occurs with hydrogen ion concentration less than pH 7 are sensatlve to weak acids. Thus methyl red wit}!
color changes between pH 4.2 and 6.3, is sensitive to weak bases, while phenolphthalein with a range of pH 8.2 to 10 is sensitive to weak acids.
Indicators may show a color change at entirely different stages of the
chemical reaction, phenolphthalein, for example, in the titration of sodium
carbonate, Na 2 CO a indicates an acid reaction as soon as one of the Na atoms is
neutralized and a drop of excess of the titrating acid liberated CO 2 Methyl
orange at this stage shows alkalinity, NaHCO a being alkaline to methyl orange
and neutral to phenolphthalein. The same difference is shown in the titration
of other polybasic compounds.
Among the indicators commonly employed in acidimetry and alkalimetry
the list given by Kolthoff and Furman 2 is of special interest; tropeolin 00,
pH 1.3 red-3.2 yellow; thymol blue, pH 1.2 red-2.8 yellow; methyl yellow,
transition 2.9-4.0, red to yellow; methyl orange, transition interval 3.0-4.4,
red to orange; brom phenol blue, pH 3.0-4.6, yellow to violet blue; congo red,
3.0-5.2, blue to red; methyl red, 4.4-6.2, red to yellow; brom cresol green, 4.0- .
1
2

By Wilfred W. Scott.
"Vol. Anal.," J. Wiley and Sons.

2191

2192

ACIDIMETRY AND ALKALIMETRY

5.6, yellow to blue; chlorphenol red, 5.0-6.6, yellow to red; brom cresol purple,
5.2-6.S, yellow to purple; sodium alizarin sulfonate, 5.5-6.S, yellow to lilac;
brom thymol blue, 6.0-7.6, yellow to. blue; azolitmin (litmus), 5.D-S.0, red to
blue; pheno.l red, 6.S-S.0, yellow to red; neutral r~d, 6.8-S.0, red to. ye~o.w
orange; rosalic acid, 6.9-S.0, yellow to. red; 'creso.l r~d, 7.2-S.8, yello.w to red;
a naphtho.l phthalein, 7.3-8.7, yellowish rose to green; curcumin, 7.4-S.6,
yellow 'to. red-bro.wn; phenol phthalein, S.2-1O, colorless to red; thymol blue,
S.D-9.6, yellow to. blue; a naphthol benzein, 9.D-11.0, co.lo.rless to blue; thymol
phthalein, 9.3-10.5, colorless to blue; alizarine yellow, '10.1-12.1, yellow to
lilac; tropeolin 0, 11.0-13.0, yellow to o.range brown! nitramine, 10.S-13.0,
colorless to red bro.wn. Certain mixed indicators give a color change more
sharply, for example a mixture of bro.m cresol purple and brom thymo.l blue,
mixed in equal parts. The indicato.r is greenish yellow at pH 6.0 and blue
at pH 6.S. A co.mprehensive table of mixed indicators is given in Koltho.ff and
Furman's work.
A list of the more common o.f the indicators is tabulated belo.w. The pH
transition interval, and the color changes are given.
Solutions of Indicators: Brom Cresol Purp1e.-0.04% alcoholic solution.
Changes color at about pH 6.
Methyl Orange.-0.1 g. per 100 ml. of water. Transition interval pH '3.4.4. Apparent pH, of end-point, or pT = 4.
Methyl Red.-The acid is dissolved)n alcohol (0.1 g. per 60 m!.) and illfide
up to 100 m!. with water. If the sodium salt is available, 0.1 g. is dissolved in
100,ml. of water. pT = 5.
Methy1 Yellow.-0.1 g. in 100 m!. of 90% alcohol. pT=4.
Phenolphthalein.-1 g. per 100 mr. of 95% alcohol. If I or 2 drops of this
solution is used per 100 m!. of solution titrated, the end-point is at pH 9. If
about ten times as much of the indicator is added per 100 J!ll. of solution .the
change appears at pH S. Cresol red (0.1 g. per 100 ml. 20% alcohol) changes
.
color at pH S.
Thymol Blue (Thymol sulfone phth~lein).-O.l ,~. in 20 ml. of hot alcehol;
dilute to 100 ml. with alcohol. The indicator has two ranges: Changes from
red to yellow at pH 1.2-2.S, pT=2.6; changes from yellow to blue, pH S.-9.6,
pT=9.
Mixed Indicators.-1.25 g. methyl red, 0.S25 g. methylene blue dissolved in
1 liter of 90% alcohol; 0.75 g. methyl red, 0.025 g. guinea green dissolved in
1 liter of 90% alcohol; 5 g. so~ium alizarin sulfonate, 0.625 -g. 'guinea green in
1 liter of water; 5' g. sodium alizarin sulfonate, 1.25 g. indigo car min in 1 liter
of water (A. H. Johnson and Jessie R. Green, Ind. Eng. Chern., Anal. Ed., 2,
2 (1930.
'

ACIDIMETRY AND ALKALIMETRY

2193

ULTIMATE STANDARDS
Constant-boiling hydrochloric acid is generally used as the ultimate standard '
acid. Benzoic acid and other acids or acid Sitlts, e.g. potassium acid phthalate,
are also used.
. Sodium carbonate is the best of the alkali standards, This salt. may be
prepared in exceedingly pure form. It is generally used as tile basic materiaL
for the volumetric standardization of the acid.

PREPARATION OF PURE SODIUM CARBONATE

'1

Bicarbonate of Soda made by the Ammonia-Soda process may be obtained

in exceedingly pure form. The impurities that may be present are silica,
ammonia, lime, arsenic, sodium chloride and sodium sulfate. With the exceptioll of silica and lime the impurities may be readily removed by recrystallizing
the bicarbonate, decanting off the supernatanf solution
containing the impurities. The bicarbonate is dried belIlQ'Co
tween large filter papers in the hot air oven (1000 C.). .
Standard Sodium Carbonate is made from this pure
sodium bicarbonate by heating at 290 0 C. to 3000 C. in an
electric oven. If a constant-temperature oven is not available a simple oven may be improvised by \l~~ of a sand bath
and a large beaker or a sheet-iron cylinder covered at the
upper end as shown in Fig. 299. A thermometer passing
through this shield registers the temperature of the material,
within a large platinum crucible. This' crucible rests upon a ,.
tri:1ngle, so that the bicarbonate is entirely surrounded by
an atmosphere of comparatively even temperature.
The sodium bicarbonate is converted to" the carbonate.
Constant weight will be obtained in about five or six hours.
When the material no longer loses weight it is cooled in a
'}j'lG. 299.
desiccator and bottled for use, preferably in several
small, glass-stoppered bottle,? For exceedingly accurate work 'the material
is analyzed and allowance made for the impurities that may still-remaiu. . The
error caused by any such impurities is so small that for all practical purposes
it may be neglected.
.
,,'
This purified sodium carbonate is an ultimate standard for acidimetric and
alkalimetric volumetric analysis.
'

2194

ACIDIMETRY AND

ALKAL~METRY

PREPARATION OF STANDAR:D ACID

STANDARD SULFURIC A;CID
I

Fifty-two per cent sulfuric acid is in equilibrium ;with the average moisture
present in the air of the laboratory; acid of this concentration is recommended
for the standard stock solution. 3
Pure 94 to 97% H 2S0 4 is diluted with sufficient-water so that its gravity is
about 1.4200 (42.7 Be.). The acid is well mixed and poured into small clel).n
and dry glass-stoppered sample bottles of about 200-m!. capacity. The bottles
are carefully sealed and placed aside for use as desired. To d,etermine the exact
strength of this standard acid a portion is standardized against the sodium
carbonate, prepared according to directions given.
.
, Method of Standardization. Procedure.-(a) Direct. From a weighing
bottle weigh out accurately about 40 X N X .053 g. (where N is 1, 0.5, 0.1 etc. the'
approximate normality of the acid). Dissolve the sodium carbonate in 50 m!.
distilled water, free of CO 2 , and titrate using methyl orange indicator;- or
one of the mixed indicators (page 2192). For 0.1 N solutions a reference solution is prepared by adding the equivalent .of ! drop of the standard acid to 100
m!. of water containing 2 drops of the indicator.
J'

Wt. of sodium carb6nate

M!. of acid used in titration X .053

N of the acid.

(b) Indirect. A catch weight of about 10 grams of the acid is weighed 0l\t
in a weighing bottle- or 100-m!. beaker (10 m!. = approximately 13 grams) aijd
placed aside for titration. The amount of sulfuric acid in the sample (weight
of sample multiplied by per cent divided by 100) is neutralized by 1.0808 times
its weight of sodium carbonate: As an excess of acid is necessary to drive out
all the carbonic acid the following formula is \j.~ed-(grams H 2S0 4-O.05)
X 1.0808 = weight of N a 2C0 3 required.
I
'
The required amount of sodium carbonat~ is weighed and transferred to a
600-m!' Erlenmeyer flask and 100 m!. of water added. The acid is carefully
poured into the flask and the rinsings of the. weighing bottle or beaker added.
The solution is boiled for 15 minutes to expel CO 2 .~ small filtering funnel
inserted in the neck of the flask prevents loss during the boiling of the acid and
carbonate mixture. The excess of acid is titrated with N /5 NaOH, using
phenolphthalein indicator, the ca1.lstic being added drop by drop until a faint
permanent pink color is obtained.
3 Ninety-three thousand pounds of sulfuric acid, y.rith an exposed surface of 1260
sq. ft. and depth of 10 in., had decreased in strength from 86% to 52.12% H 2S0 4, after
standing in a lead pan, protected from the rain, for 42 days (Sept. 9th to Oct. 21st, 1916).
Air was bubbled through a two-liter sample of this, acid for seven consecutive days,
when the solution was tested and found to contain 52.18% H 2S0 4 The average temperature of the laboratory was 74 F., the average vapor of the air (7tests) was 0.2223
gram H 20 per ,standard cubic foot. The average humidity for September and October
was 68%; the average temperature 62 F. The ave_ry.ge humidity for the past 33 years
.
was 72%; average temperature 57 F.

ACIDIMETRY AND ALl\:ALIMETRY

2195

(The sulfurIC equivalent to the N aOH added) + (weight of N a 2 CO a XO.9252)

=weight of pure H 2S0 4 present in the sample.
NOTEs.-C0 2-free water should be taken in all titrations with phenolphthalein.
The indicator contains 1 gram of the compound per liter of 95% alcohol. One mi. of
indicator of this strength is required for each titration.
Results should agree to within 0.05%.
The temperature of the acid should be observed at the time of standardization and
this noted with results on the bottles containing the standard samples. The coefficient
of expansion is .00016+ per degree F. risen in temperatUre or .000293 per degree C. per
mi. of solution.
Normal Sulfuric acid contains 49.043 grams of H 2S0 4 per liter of solution.
To make Ii liter of the normal acid the amount of the standard acid required is
100X49.043
.
calculated by the formula
t H SO .
d d grams standard aCId
per cen
2
4 In stan ar
.
necessary. The acid is weighed out in a small beaker, a slight excess being
taken (0.1 gram). The acid is washed into a liter flask and made to volume.
An' aliquot portion is standardized against the standard sodium carbonate.
The solution may now be adjusted to the exact strength",required.
Example.-If 25 m!. of the acid is found to contain 1.25 grams H 2S0 4 we
find the amount of dilution required as follows: 25 m!. of N /1 H 2S0 4 should
contain 1.226075 grams, therefore 1.226075 : 25 :: 1.25 : x, and x =

~~2~:~~~'

Then x minus 25 = the amount of water required for 25 ml. Total dilution
= dilution for 25 multiplied by the volume of. acid remaining in. the flask
divided by 25 =ml. water required to make a normal acid solution.
Fifth normal and tenth normal acids may be prepared by diluting the
normal acid to five or ten volumes as the case requires.
Gravimetric Methods. Precipitation as BaS04.-Sulfuric acid' may be
standardized by precipitating as BaS04 according to the procedure given for
sulfur. BaS04 X0.4202 = H 2S0 4,
Determination as (NH4}2S04.-To 10 ml. of the acid diluted to 50 ml. in a
large platinum dish is added NH 40H until the acid is neutralized and a faint
odor of ammonia is perceptible. The solution is evaporated to dryness on the
water bath and dried a,t 100 C. for half an hour. The residue is weighed as
(NH 4)2S04' (NH 4)2S04XO.7422 = gram H 2S0 4.

STANDARD HYDROCHLORIC ACID

Standard hydrochloric acid may be prepared from a solution of the acid
containing a known amount of Hel. The stock solution, containing an exact
amount of the HCI in solution, may be prepared by distilling dilute hydrochloric

ACIDIMETRY. AND ALKALIMETRY

2196

acid until it has attained a definite boiling point. The pressure being known
the exact amount of the acid that now distills and that remaining in the flask.
provided the original solution was pure, can be estimated from the following
\
table of Bonner and Wallace. 4
,

Boiling Point Data of

Press, Mm.

Boiling Pt.

102.209
103.967
105.564
106.424
107.859
108.584
110.007

600
640
680
700
740
760
800

..

C.
C.
C.
C.
C.
C.
C.

Her

Density, 25 0

HCl%

1.0980
1.0973
1.0968
1.0966
1.0962
1.0959
1.0955

20.638
20.507
20.413
20.360
20.268
20.222
20.155

"-

Procedure.-The dilute acid, prepared from the solution of concentrated

hydrochloric acid by diluting with an equal volume of distilled water, is pla'ced
in a flask of the distillation type with side arm. A thermometer is inserted
and the acid heated to boiling. When a constant temperature is attained the
distillate is caught for use. The barometric reading is taken at the time of
..
distillation.
'
Acid of any strength desired can be prepared by direct measurement, the
density and percentaglcl of the acid being known from the table above. The
,density multiplied "by the percentage of Hel and the result divided by 100
gives the weight of Hel per m!. of the stock solution.
Normal Hel contains 36.46 g. Hel per liter.

The weight of constant boiling Hel to weigh against brass weights in air
to give 1 liter of N solution is given by Foulk and Hollingsworth. s The following will serve:
.
Barometer
Reading
when Acid
was Collected

G. of Acid

per 1 Liter
of N Acid

770 ......................... , ......... 180.407

760 ................................... 180.193
750 .... \............................... 179.979
740 ................................... 179.766730 .............. , ................... '.179.555

Gravimetric Determination of Hydrochloric Acid by Precipitation as AgCI.Hydrochloric 3id may be standardized by precipitation with silver nitrate
solution by the procedure for determination of chrorine. Agel XO.2544 = Hel.
It is advisable to heat the sample, diluted to a con.venient volume, and add the
hot silver nitrate in slight excess of that required by Hel, the amount of the
reagent being calculated, e.g., mol. wt. Hei : mol. wt. Ag'NO a :: Wt. Hel in
sample: x.
,
4
6

W. D. Bonner and R. E. Wallace, Jour. Am. Cmhu. Soc., 52, 1750 '(i930).
J. Am. Chern. Soc., 45, 1220 (1923) ..

ACIDIMETRY AND ALKALIMETRY

2197

BENZOIC ACID S'l'ANDARD

Benzoic acid may be obtained in exceedingly pure form by melting the resublimed acid in a covered platinum dish in a constant-temperature oven, at a
temperature of 140 0 C. The acid is poured into test-tubes, cooled, and the
sticks bottled for use. The acid does not take up moisture to any appreciable
extent, even when exposed to the air for some time, so that it may be weighed
without danger of absorption of moisture.
STANDARD ALKALI SOLUTION
Standard normal sodium hydroxide is made by dissolving approximately
50 grams of NaOH sticks with 1 to '2 grams of Ba(OH)2 in 200 to 300 m!. of
water and diluting to 1000 ml. The caustic is standardized against normal
or tenth normal hydrochloric or sulfuric acid, using phenolphthalein indicator.
Other standard acids may be employed, such as oxalic, benzoic, and phthalic
acids'. See below. The solution is adjusted to the exact strength desired by
addition of distilled water. The solution should be protected from CO 2 of the
air-see Fig. 300, p. 2199.
NOTE.-The addition of Ba(OH)2 is made to precipitate the carbonate in the caustic,
as this would interfere with titrations in presence of phenolphthalein. As the presence
of barium would produce a cloudiness with H 2S0 4 it is advisable to add only an amount
sufficient to precipitate the carbonate. Milk of lime may be added in place of Ba(OHh.

Standardization of Sodium Hydroxide by Acid Potassium Phthalate.Acid potassium phthalate, KHC sH 4 0 4, can be obtained in pure form. It has
a high molecular weight, is not hygroscopic and may be weighed in an open
container without danger of change. Phenolphthalein indicator is satisfactory
for the titration.
Reaction: KHCsH404+NaOH=KNaCsH40.+H20.
204.215 g. of the salt is equivalent to 1.008 g. H or 40 g. NaOH. 1.0211 g.
is equivalent to 50 ml. of N /10 solution.
Procedure.-Weigh 2-5 grams of the acid potassium phthalate, transfer
to a 300 ml. Erlenmeyer flask and dissolve in about 50 ml. of water. Add 3-4
drops of 1 % phenolphthalein indicator and titrate with the N aOH solution to
Ii faint pink color. (See note below.)
.
The normality of NaOH is obtained by dividing the volume equivalent of
the acid potassium phthalate by the ml. NaOH. Volume equivalent = Wt.
KHC sH 40 4 -;-.0.2042 = m!. normal solution.
NOTE.-In a separate flask, with the same amount of solution and in'dicator, obtain
the same color by adding NaOH drop 'by drop. Subtract this blank from the above
titration before calculating normality.
Alkaline solutions that are to be used in presence of CO 2are best standardized against
standard acid solutions prepared by titration against Na 2CO S or against HCI prepared
from the acid of constant boiling point, approximately 20.2% HCl.s
S G. A. Hulett and W. D. Bonner, J. Am. Chern. Soc., 31, 390 (1909), and W. D.
Bonner and R. E. Wallace, J. Am. Chem. Soc., 52, 1750 (1930).

2198

ACIDIMETRY AND ALKALIMETRY

Precautions should be taken against the presence ofr CO 2 in the titrations where
phenolphthalein is used. The distilled water should be freshly boiled, and the air in
'
.
the flask should be free of CO 2
Pure NaOH may be prepared by dissolving 100 grams of the pure stick NaOH in
about 100 ml. of water, transferring to a large test tube ,a;nd allowing to settle. If the
solution is centrifuged for 20-30 minutes the sodium carbonate and other insoluble
material will settle and the clear liquid may be used at once. 7 By'means of a graduated
pipette the required amount of the clear solution is drawn out and made to volume.
40 grams NaOH per liter is a normal solution. Other normalities may be desired.
The NaOH should be kept in bottles lined with paraffine. This paraffining is done by
heating the bottle, pouring in sufficient melted paraffine and rotating the bottle so that
the paraffine flows over the sides and the entire inner surface.
The end-point with methyl orange indicator (3 drops 0.02% solution per 100 ro.l.)
is the color obtained in a solution with pH=4.2.

TO MAKE A SOLUTION OF APPROXIMATE NORMALITY

STRENGTH WITHOUT WEIGHING, FROM A SOLUTION
OF KNOWN STRENGTH AND SPECIFIC GRAVITY
Calculate the m!. of the reagent that contain the weight of the matqrial
necessary per liter. Measure out this volume and dilute to 1000 m!.
Example.-Required to make an approximate 0.1 N H 2S0 4 solution from
87% H 2S0 4 having a sp.gr. of 1.8.

A liter of normal solution contains 49.04 g. H 2S0 4 , a 0.1 N 4.'904 g. H 2S0 4

1 m!. of the acid contains 1.8 XO.87 g. "H 2S0 4. '
..
4.904 g. of H 2S0 4 is present in

1.:~~~87

3.13 m!.

Dilute 3.13 m!. of the acid to 1000 m!.

By using normal solutions ofthe acid or alkaline reagents the color changes
of the indicators commonly employed are sharp. Lunge-Berl s regard 0.2 N
solutions as the lower limit of concentration. It is occasionally advisable to
use more dilute solutions even though the color changes are not so sharp.
Borax (Na 2 B 40 7 lOH 20) is recommended by various investigators as a
suitable ultimate standard for acid standardization, using methyl red as indicator, in 0.1 N titrations, or methyl orange if the reagent is stronger than 0.2 N.
In making up normal solutions it is necessary to note the temperature of
the solution at the time of standardization, since a change of temperature will
change the value of the reagent. The following corrections' should be applied,
the table taking the initial temperature to be 20 0 C.
15 .... 0.00077
16 .... 0.90064
17. " .0.00050
18 .... 0.00034
19 .... 0.00018

20 .... 0.00000

21 .... - 0.00018
22 .... -0.00038
23 .... -0.00059
24 .... - 0.00081
25 .... - 0.00103

The corrections apply to pure water. With the reagents indicated below
the correction must be increased by the percent~ges indilJated.
7

Allen and Low, Ind. Eng. Chern., Anal. Ed.,-S;-'1~2 (1933).

"Chemisch-Technische Untersuchungsmethodik," 7th Ed., Vol. I, p. 138.

ACIDIMETRY AND ALKALIMETRY

2199

N HNO a 50%, HCl 25%, H 2SO. 45%, NaOH 40%, KOH 40%, H 2C.O. 30%,
.
Na 2CO a 40%.
N/IO HNO. 6%, HCI3%, H 2SO. 5%, NaOH 5%, KOH 4%, H 2C.O. 3%,Na.CO a
5%.

STANDARD BURETTES .
Burettes used for this work should be carefully ,checked for accuracy of
G
delivery.
For accurate titration of acids
or alkalies it is advisable to have a
titration of 75 to 100 ml. Since the
straight 100-ml. burette if graduated to twentieths of a ml. would
be too long for convenient handling,
the chamber burette is used. The
chamber located in the upper portion of the apparatus holds 75 mI.,
the lower portion drawn out into
a uniform-bore tube is graduated
in twentieths of a mI. Each tenth
of a ml. has a mark passing entirely
around the tube so that there will
be no error in reading, the eye being
held so that the mark appears to be
c
a straight line drawn across the
tube. The burette is enclosed in a
large tube filled with distilled water
Chamber
and carrying a thermometer. The
Burette
A
burette is connected, by means of
an arm at the base, with a reservoir
of standard acid. The cut, Fig.
300, shows the apparatus connected ready for use. '
If vapor is lost from the stand4.' -Wafer
ard reagents and this replaced by
Jacket
B
dry air, as in the common practice,
the solution gradually changes in
strength. A simple and ingenious
device, designed by H. W. Herig
(Gen. Chern. Co.), is shown at the
top of Fig. 300, whi<;h overcomes
F
this difficulty. The air drawn into
the reagent bottle is purified and
saturated with moisture by passing it through sodium hydroxide.
FIG. 300.
A mercury valve relieves the pressure if expansion of air in the,reagent bottle occurs due to rise of temperature.

2200

ACIDIMETRY AND ALKALIMETRY

NOTE.-The chamber burette shown in Fig. 300 was designed at the Laurel Hill
Laboratory, General Chemi9al Company.
I "
.

TITRATION OF ACIDS AND 4LKALIES

In the acid titration the sample is conveniently titrated in a white porcelain
casserole. This gives white background that enables the analyst to see the
end-point. The caustic is run into the acid, to within a few ml. of the endpoint, rapidly and then cautiously to a faint change of color-faint pink with
phenolphthalein or an orange-yellow with methyl-orange. Phenolphthalein
is generally preferred in acid titrations. CO 2-free caustic and '~/ater should
J
always be used.

METHODS OF WEIGHING ACIDS

DILUTE ACIDS NON-VOLATILE UNDER ORDINARY
CONDITIONS
Dilute acids may be weighed directly in a beaker, weighing"bottle or ordinary pipette (see directions given later) by measuring out the approximate
amount desired. Since a burette reading from 75-ml. to lOO-ml. should be
used for this work it will be necessary to take such an amount of the acid as will
require a titration between these extremes. This may be accomplished by
taking the specific gravity of the acid and referring to the table for the approximate strength. From this the volume necessary may readily be
calculated.
Example.-The case will be taken where a 75-rr,t1. to 100-ml. burette is being
used and the titration is to be made with normal caustic solution; the acid
titrated is sulfuric acid. The capacity of the:burette is 75 XO.049 = 3.675 grams
H 2S0 4 to 100 XO.049 =4.9 grams H 2S0 4 (For HCI the capacity would be
2.74 to 3.65 grams HCI and for HNO a it would be 4.73 to 6.3 grams HNO a.)
Suppo$e the sulfur~c acid has a sp.gr. of 1.1600. From the table for H 2S0 4
we find that this acid is 22.25% H 2 S0 4, then 1 ml. contains 1.16X22.25 divided
by 100=0.2581 gram H 2S0 4 , Since the capacity of the burette is 3.675 to 4.9
grams H 2S0 4 , we must weigh between 3.675 and 4.9 grams of the acid; to get
we s h ou ld ta k e 3.675
4.9 m.,
I t h at '.IS to say, 14.5 to 18.5 m I. 0 f t h e
.2581 to .2581
t h IS
acid, which will weigh 16.8 grams to 21.5
WEIGHING

grams~

CONCZNTRATED ACIDS,. FUMING OR VOLATILE

UNDER ORDINARY CONDITIO;NS

The acid must be confined during weighing-aDd until it is mixed with water
or standard caustic. The best forms of apparat~s include the following:

. ACIDIMETRY AND ALKALIMETRY

2201

Lunge-Ray Pipette.-The pipette is shown in Fig. 301. Two glass stopcocks confine the acid in a bulb. The lower part of the pipette is protected
by a ground-on test-tube. The dry pipette is weighed. Cock 2 is closed and
1 opened and a vacuum produced in the bulb by apply~ng suction at the upper end of pipette and closing stopcock 1 with
suction still on. The sample may now be drawn into the pipette
by immersing the lower end in the sample and opening the
stop-cock 2, the vacuum producing the suction. The increased
weight=acid drawn in. The pipette is emptied by'running
the acid under water.
Dely Weighing Tube. 9-This form of weighing tube has
proven to be of exceptional value, to the busy works-chemist, 2
. in the analysis of oleum and mixed acids. Both speed and
accuracy are gained by its use. The apparatus, shown in Fig.
302, consists of a long glass tube of small bore, wound in a
spiral coil.
The sample of acid is drawn into the weighed coil by
applying suction through a rubber tube attached to A and
drawing in the required amount of acid, a mark, ascertained
by a previous run being made, to indicate the point to which
the acid is drawn. The tip B is carefully wiped off with tissue
paper and the tube and sample weighed. The weight of the
FIG. 301.
tube deducted gives the weight of the sample.
Lunge-Ray
The apparatus is now inclined so that the acid runs back
Pipette.
into the crook at C to a point mluked on the wall of the tube,
in order to expel as much air as possible from this end. A rubber tube filled
with water is attached to A, the other end of the rubber tube being connected
to a bottle containing distilled water. A glass bead, such as is used in rubbertipped burettes, fitting snugly in this tube,
regulates the flow of water. The Dely tube
is now inverted, the tip being immersed in
150 ml. to 200 ml. of distilled water in a 4in. casserole-Fig. 302. By pressing gently
on the bead, water is slowly admitted in the
tube, forcing the acid before it. The acid
and water are separated by a bubble of air.
Before forcing out the last half-inch of acid,
the tube connected to the water supply is
disconnected and the weak acid from the
casserole drawn back two or three inches
into the Dely tube, then again the acid is
almost entirely expelled by water from the
reservoir and the procedure repeated. This
is to absorb the S03 gas that inv'ariably is
FIG, 302. Dely Weighing Tube in present in the bubble of air above menOperation.
tioned, which would be lost if forced out
directly by the water column. In order to facilitate this'last step it is well to
have a short rubber tube attached to the Dely tube, and a glass tip in the
9 J. G. Dely, Chemist, Gen Chern. Co.

2202

ACIDIMETRY AND ALKALIMETRY'

tube connected with the reservoir of water., The acid in the casserole, upon
washing out the Dely tube, is titrated with standard caustic according to the
procedure for titration of acids.
The tube is dried after washing, with alcohol, followed hy
ether, by heating on an asbestos m'at on a hot plate, dry air
I
being aspirated through.
Snake Weighing Tube.-The snake tube is a simple
device that may be easily made by an amateur glass-blower.
It is made out of a glass tube 8-10 ins. long, slightly thinner
than a lead pencil. One end of the tube is drawn out to
capillarity. The tube has a double bend, as shown in tpe
illustration. It is so made that it r~sts on the double bend
with the ends inclined upward to prevent the outflow of the
acid. Fig. 303.
The tube is dried with alcohol, ether and air treatment, as
in case of the Dely tube. After weighing the empty tube, acid
is drawn into it by suction ,through an attached rubber tube.
FIG. 303.
The capillary end that has dipped into the sample is wiped
Snake Tube.
dry with tissue paper. The acid and tube are weighed and
the acid estimated by difference.
The acid is run into 150 m}. of water in a casserole, the flow being regulated
by the index finger pressed against the
"
larger end of the tube. With careful
Horns fOrSuspendip9
regulation of the flow, practically no
bur.:efte
bumping occurs. With a small capillary
opening it is not necessary to place the
0 ~ y:/
finger over the larger end of the tube as
the acid flow will be slow. The tube
should be kept in motion to prevent
bumping from overheating anyone portion. Kicking back of the acid indicates
that the capillary end of the tube is too
a,'
large. When the contents of the tube
have run out, the tube is rinsed by sucking up some acid from the casserole and
allowing it to run out, repeating sev- E Capillary
eral times. Suction mlay be applied by
liibe adaplor
means of a rubber bulb attached to the
for Fuming
Acids
tube. The acid is now titrated with
standard caustic, using phenolphthalein
indicator.
Blay-Burkhard Graduated Weighing
Burette. lO-This apparatus, designed by
V. J~. Blay and W. E. Burkhard, General Chemical Company, is used for
weighing acids or other liquids. The
FIG. !?04. Blay-BurkPard Graduated
form for u;eneral use i~ shown in Fig. 304.
. -' ~ Weighing Burette.
THe burette is graduated in half m!. divi10 A similar burette of larger capacity and small weight was designed by Friedman
and La Mer, Ind. Eng. Chern., Anal. Ed. 2, 54 (1930).

ACIDS

2203

sions, from 0 to 20 ml. An apparatus half this size is used for oleum, where
a 2-ml. sample is sufficient for a determination. For the purpose of running
the sample under water a capillary tub~ (E, Fig. 304), with ground joint, is
attached to the burette. This tube is placed in the solution during titration.
The burette is provided with a glass vented stopper. (A) on the top, and a
glass cap for the tip, both having ground joints, to prevent escape of fumes
from the sample.
W. W. Scott modified the apparatus by replacing the fragile cap (A) by a
tube stopper with capillary vent (see A', Fig. 304). The vent to the air is
opened or closed by a slight turn of this stopper. By means of this tube acid
may be drawn into the burette according to the Lunge-Ray pipette procedure.
With these burettes a man can control his work very accurately and save a
great amount of time, both in weighing and manipulation.
In the analysis of strong oleum, about 50 grams of neutral Glauber salt are
placed in a casserole containing water, and the fuming acid allowed to flow
under the undissolved 'salt. The violent reaction of the acid with water is
thus ~voided. The tube E, Fig. 304, should be made of fused silica.
The glass-bulb method is still used for analysis of strong oleum. The acid
weighed in a sealed tube of known weight is mixed with water by breaking the
bulb in a stoppered bottle containing water, the acid is cooled and titrated as
usual.
I

ANALYSIS OF MURIATIC ACID

(Commer~ia1

Hydrochloric Add)

TOTAL ACIDITY AND HyDROCHLORIC ACID

The usual titration with standard caustic gives the total acidity, including,
in addition to hydrochloric acid, nitric. and sulfuric acids which may occur in the
commercial product. The acidity due to these acids is deducted from the total
acridity to find the actual HCI in the muriatic acid.
A catch weight, 10 to 15 grams of the acid, is weighed in a weighing bottle,
or a large snake tube, or the Dely tube, as in case of oleum analysis, and .the
acid allowed to mix with water in a casserole; methyl-orange indicator is added
and the acid titrated with standard normal caustic solution, the red color fading
to a lemon-yellow. A fraction of a drop of the alkali will cause the change
when the end-point has been reached.
One mI. N/INaOH=0.03647 g. HCI.
NOTE.-Hydrochioric acid may be determined gravimetrically by precipitating
the chloride with, silver nitrate-HCI+AgNOa=AgCI+HNOa, or by the t-olumetric
methods for the determination of chlorine. See Chlorine.

ACIDS

2204

DETERMINATION OF IMPURITIES IN COMMERCIAL

HYDROCHLORIC ACID. FREE CHLORINE
Five ml. of the acid are diluted to 10 mI., about 5 ml. of fresh starch solution
added and a few drops of 5% KI solution togethJr with about 1 ml. of dilute
H 2S0 4. A blue color indicates free chlorine. This color may be matched in a
Nessler tube with a standard. It is possible to det'ermine .0001 % chlorine Oil a
5-ml. sample.
NITRIC ACID OR NITRATES IN HYDROCHLORIC ACID
About 5 ml. of the hydrochloric acid is cautiously added to 75 ml. of 1l5%
H 2S0 4, the HCI being introduced under the surface of the sulfuric acid. The
nitric acid may now be titrated with standard ferrous sulfate by the procedure
for the direct determination of nitric acid and nitrates. (Method of Scott and
Bowman.) The ferrous sulfate test for nitric acid is delicate. Traces of nitric
acid produce a pink coloration; larger amolints a reddish brown to dark }lrown.
The color is permanent when an excess of ferrous sulfate has been added.
SULFURIC ACID

AND

SULFATES IN

HYDROCHLORIC

ACID

Free H 2S0 4. Fifty ml. of the sample is evaporated in a platinum dish

(steam bath) to dryness or until the HCI has been expelled. A few drops of
water are added and the material again taken to dryness (steam bath). The
residue is 'taken up with water and titrated with N /10 NaOH, using methylorange indicator. One ml.=0.0049043 gram H 2S0 4,
Total Sulfates. Fifty m!. of the muriatic acid is evaporated to abl)ut 5 to
10 m!. and then diluted to about 200 m!. and heated to boiling. Total SOa is
now precipitated by adding BaOl2 solution as in case of determination of total
sulfur. The precipitated BaS04 includes the fr~e H 2S0 4 and the combined
'
80 a. Ba80 4X 0.4202 = H 2S0 4.
ARSENIC IN HYDROCHLORIC ACID
Commercial muriatic acid niay contain arsenic. This is best determined
by the Gutzeit Method given in detail under Arsenic. 10 m!. of sample is
usually sufficient for this determination. If much arseni~ IS 'present the distillation' metMd -may be followed, using a 25 to 50-m!: sample, The distillate is
titrated with standard iodine according to procedure given for arsenic by the
iodine titration, page 99.
BARIUM CHLORIDE IN HYDROCHLORIC ACID
Fifty ml. is evaporated to dryness and then over a low flame to expel SOa.
The residue is taken up with 1 m!. of 1 : 1 HCI and 50 ml. of water. 1 m!.
H 2S04 is.added and the precipitated BaS04 filtered off and weighed. If silica
is present in the sample its weight shoulg.-l>e deducted. BaS04XO.8923
= BaCI 2.
-

ACIDS

2205

TOTAL SOLIDS AND SILICA

One hundred ml. of the HCI in a platinum dish is evaporated to dryness and
the residue ignited and weighed. 5 ml. of HF is added with a few drops of
H 2S0 4 and the solution again evaporated and ignited. The first weight = total
solids. The loss of weight in the ,second ignition = Si0 2

DETERMINATION OF S?ECIFIC GRAVITY

Control tests for strength of the common inorganic acids are generally
made in the plant by means of the hydrometer. This instrument depends on
the fact that when a solid floats in a liquid, ,the weight of the liquid displaced
is equal to the weight of the 'floating body. The hydrometer is a cylindrical
instrument, generally with a spherical bulb, weighted with lead shot or mercury.
It has a narrow stem with graduations, which indicate the gravity. The
,instrument floats vertically, after displacing its own weight of solution in which
it is placed. The instruments are adapted for varying gravities extending over
a range of ten to twenty divisions.

FIG. 305.-Hydrometer.

The Gravity tables given in the following pages have been developed with
considerable care. By means of ,these one is able to obtain quickly and with
fair degree of accuracy the strength of the acid or alkali of which the gravity is
ascertained.
The following precautions should be observed in making hydrometer tests:
1. The hydrometer should be clean and dry, and at the temperature of the
liquid, before immersing to make a reading.
'
2. The vessel in which the observation is made should be of clear glass of
suitable size and shape, to allow the hydrometer to float freely (about ~ inch
greater in diameter tllan the hydrometer bulb) and of sufficient lieight to enable,
full reading of the hydrometer (i.e., height greater than the length of the
hydrometer).
'
3. The liquid should be thoroughly mixed by means of a stirrer reaching
to the bottom of the vessel. There should be no air bubbles in the liquid or
clinging to the sides of the vessel or the hydrometer.
4. The hydrometer is slowJy immersed in the liquid, slightly beyond the
point where it floats, and is then allowed to float freely.
5. The reading is made with the line of vision horizontal to the plane and
as near as possible to this. The point is taken 'where this surface line cuts the
hydrometer scale.
6. The temperature of the liquid is taken before and after the reading and
allowance made for variation o{ the temperature from standard conditions as
indicated in the tables.
References.----:Cir. 16, 4th edition, Feb. 23, 1916, U. S. Bureau of Standards.
Sulfuric Acid Handbook by Tllos. J. Sullivan, McGrftw-Hill Book Co.
Thorpe Dictionary of Applied Chemistry, pp. 103-114. Longmans, Green and Co.

ACIDS

2206

HYDROCHLORIC ACID
By W. C.
Degrees
Baumel.

1.00
2.00
3.00
4.00
5.00
5.25
5.50
5.75
6.00
6.25
6.50
6.75
7.00
7.25
7.50
7.75
8.00
8.25
8.50
8.75
9.00
9.25
9.50
9.75
10.00
10.25
10.50
10.75
11.00
11.25
11.50
11. 75
12.00
12.25
12.50
12.75
13.00
13.25
13.50
13.75
14.00

Sp. Gr.

1.0069
1.0140
1.0211
1.0284
1.0357
1.0375
1.0394
1.0413
1.0432
1.0450
1.0469
1.0488
1.0507
1.0526
1.0545
1.0564
1.0584
1.0603
1.0623
1.0642
1.0662
1.0681
1.0701
1.0721
1.0741
1.0761
1.0781
1.0801
1.0821
1.0841
1.0861
1.0881
1.0902
1.0922
1.0943
1.0964
1.0985
1.1006
1.1027
1.1048
1.1069

Twaddell.

Degrees

Per Cent
HCl.

1.:>8
2.80
4.22
5.68
7.14
7.50
7.88
8.26
8.64
9.00
9.38
9.76
10.14
10.52
10.90
11.28
11.68
12.06
12.46
12.84
13.24
13.62
14.02
14.42
14.82
15.22
15.62
16.02
16.42
16.82
17.22
17.62
18.04
18.44
18.86
19.28
19.70
20.12
20.54
20.96
21.38

1.40
2.82
4.25
5.69
7.15
7.52
7.89
8.26
8.64
9.02
9.40
9.78
10.17
10.55
10.94
11.32
11.71
12.09
12.48
12.87
13.26
13.65
14.04
14.43
14.83
15.22
15.62
16.01
16.41
16.81
17.21
17.61
18.01
18.41
18.82
19.22
19.63
20.04
20.45
20.86
21.27

FERGUSON

Degrees
Bau~.

14.25
14.50
14.75
15.00
15.25
15.50
15.75
16.0
16.1
16.2
16.3
16.4
16.5
16.6
16.17
16.8
IG.9
17.0
17.1
17.2
17.3
17.4
17.5
17.G
17.7
17.8
17.9
18.0
18.1
18.2
18.3
18.4
18.5
18.6,
18.7
18.8
18.9
19.0
19.1
19.2
19.3

Sp. Gr.
1

Degrees Per Cent

Twaddell.
HCl.

1.1090
1.11'll
1.1132
1.1154
1.1176
1.1197
1.1219
1.1240
1.1248
1.1256
1.1265
1.1274
1.1283
1.1292
1.1301
1.1310
I.l319
1.1328
1.1336
1.1345
1.1354
1.1363
1.1372
1.1381
1.1390
1.1399
1.1408
1.1417

21.80
22.22
22.64
23.08
23.52
23.94
24.38
24.80
24.96
25.12
25.30
25.48
25.66
25.84
26.02
26.20
26.38
26.56
26.72
26.90
27.08
27.26
27.44
27.62
27.80
27.98
28.16
28.34
28.52
28.70
28.88
29.06
29.24
29.42
. 29.60
29.7S
29.96
30.16
30.34
30.52
30.70

1.1~26

1.1435
1.1444
1.1453
1.1462
1.1471
1.1480
1.1489
1.1498
1.1508
1.1517
1.1526
1.1535
I

21.68
22.09
22.50
22.92.
23.33
23.75
24.16
24.57
24.73
24.90
25.06
25.23
25.39
25.56 I
25.72
25.89
26.05.
26.22
26.39
26.56
26.73
26.90
27.07
27.24
27.41
27.58
27.75
27.92
28.09
28.26
28.44
28.61
28.78
28.95
29.13
29 30
29.48
29.65
29.83
30.00
30.18

ACIDS
HYDROCHLORIC
Degrees

Baume.
--19.4
19.5
19.6
19.7
19.8
19.9
20.0
20.1
20.2
20.3
20.4
20.5
20.6
20.7
20.8
20.9
2l.0
21.1
21.2
2l.3
21.4
21.5
21.6
21.7
21.8
21.9
22.0
22.1
22.2
22.3
22.4

Degrees

Sp. Gr.

Twaddell.

1.1544
1.1554
1.1563
1.1572
1.1581
1.1590
1.1600
1.1609
1.1619
1.1628
1.1637
1.1647
1.1656
1.1666
1.1675
1.1684
l.1694
1.1703
1.1713
1.1722
1.1732
1.1741
1.1751
1.1760
1.1770
1.1779
1.1789
1.1798
1.1808
1.1817
1.1827

30.88
31.0S
31.26
31.44
31.62
31.80
32.00
32.18
32.38
32.56
32.74
32.94
33.12
33.32
33.50
33.68
33.88
34.06
34.26
34.44
34.64
34.82
35.02
35.20
35.40
35.58
35.78
35.96
36.16
36.34
36.54

Per Cent
HCI.

30.35
30.53
30.71
30.90
31.08
31.27
31.45
31.64
31.82
32.01
32.19
32.38
32.56
32.75
32.93
33.12
33.31
33.50
33.69
33.88
34.07
34.26
34.45
34.64
34.83
35.02
35.21
35.40
35.59
35.78
35.97

~ID

Degrees

Baumti.
22.5
22.6
22.7
22.8
22.9
23.0
23.1
23.2
23.3
23.4
23.5
23.6
23.7
23.8
23.9
24.0
24.1
24.2
24.3
24.4
24.5
24.6
24.7
24.8
24.9
25.0
25.1
25.2
25.3
25.4
25.5

2207
(Continued)

S p. Gr.

1.!836
1.1846
1.1856
1.1866
1.1875
1.1885
1.1895
1.1904
1.1914
1.1924
1.1934
1.1944
1.1953
1.1963
1.1973
1.1983
1.1993
1.2003
1. 2013
1.2023
1.2033
1.2043
1.2053
1.2063
1.2073
1.2083
1 2093
12103
1.2114
1.2124
1.2134

Degrees
TWaddell.

Per Cent

36.72
36.92
37.12
37.32
37.50
37.70
37.90
38.08
38.28
38.48
38.68
38.88
39.06
39.26
39.46
39.66
39.86
40.06
40.26
40.46
40.66
40.86
41.06
41.26
41.46
41.66
41.86
42.06
42.28
42.48
42.68

36.16
36.35
36.54
36.73
36.93
37.14
37.36
37.58
37.80
38.03
38.26
38.49
38.72
38.95
39.18
39.41
39.64
39.86
40.09
40.32
40.55
40.78
41.01
41.24
41.48
41.72
41.99
42.30
42.64
43.01
43.40

HCI.

Sp. Gr. determinations were made at 60 F., compared with water at 60 F.

From the Specific Gravities, the corresponding degrees Baume were calculated by the following formula: Baume = 145 -145/Sp. Gr.
Atomic weights from F. W. Clarke's table of 1901. 0 = 16.
ALLOWANCE FOR TEMPERATURE:
10-15 Be. 1/40 Be. or .0002 Sp. Gr. for 1 F.
15-2ZO Be~ - 1/30 Be. or .0003" " " 1 F.
22-25 Be. 1/28 Be. or .00035 "
" " 1 F.
AUTHORITY - W. C. FERGUSON.
This table has been approved and adopted as a Standard by the Manu_
facturing Chemists' Association of the United States.
W. H. BOWER,
JAS. L. MORGAN,
HENRY HOWARD,
ARTHUR WYMAN,
ROSENGARTEN,

A. G.
New York, May 14, 1903.

Executive Committu.

ACIDS

2208

HYDROCHLORIC ACID

Specific
Gravity.

Specific
I Liter
Specific
I Liter
,Gravity . Per Cent
Per Cent
cooGravity
Per Cent
cooRCI
15
tains
RCI
tains
RCI
IS
15
by Weight. Grams
by Weight. Grams
-4
4"
7" by Weight.
RCI. in Vacuo.
RCI. in Vacuo.
In Vacuo.
0

--1.000
1.005
1.010
1.015
1.020
1.025
1.030
1.035
1.040
1.045
1.050
1.055
1.060
1.065
1.070

0.16
1.15
2.14
3.12
4.13
5.15
6.15
7.15
8.16
9.16
10.17
11.18
12.19
13.19
14.17

1.6
12
22
32
42
53
64
74
85
96
107
118,
129
141
152

1.075
1.080
1.085
1.090
1.095
1.100
1.105
1.110
1.115
1.120
1.125
1.130
1.135
1.140
1.1425

15.16
16.15
17.13
18.11
19.06
20.01
20.97
21.92
22.86
23 82
24.78
25.75,
26.70
27.66
28.<14

163
174
186
197

f 09

220
232
,243
255
267
278
291
303
315
322

1.145
1.150
1.152
1.155
1.160
1.163
'1.165
1.170
1.171
1.175
1.180
1.185
1.190
1.195
1.200

28.61
29.57
29.95
30.55
31.52
32.10
32.49
33.46
33.65
34.42
35.39
36.31
37.23
38.16
39.11

Liter
coo- .
tains
Grams
HCI.
I

-328
340
345
353
366
373
379
392
394
404
418
430
443
456
469

COMPOSITION OF CONSTANT BOILING

HYDROCHLORIC ACID*

Pressure mm. of
Mercury.

770
760
750
740
730

rer cent of
HCl.

20.218
20.242
20.266
20.290
20.314

Grams' constant
bOiling distillate
for I mol. HCI.

180.390
180.170
179.96Q.
179.745 179.530

Temperature of constant boiling hydrochloric acid is 108.540 at 76a mIn.

Specific gravity 1.0962()25.
* Hulett and Bonner, Jour. Am. Cbem./Soc. xxxi,390.

ACIDS

2209

ANALYSIS OF HYDROFLUORIC ACID

\

The following constituents-hydrofluosilicic, sulfuric and sulfurous acidscommonly occurring with hydrofluoric acid, are determined in the analysis,
along with the hydrofluoric acid, by titration. Generally the acid contains a
slight residue upon ignition. The titrations are made in presence of KN0 3 ,
first ice cold, and then completed at 80c C. or more. The cold titration gives
the hydrofluoric, sulfuric and sulfurous acids and one-third of the hydrofluosilicic acid and upon heating the titration gives the remaining two-thirds of the
H 2SiF 6, the following reactions taking place. Titration cold, H 2SiF 6 +2KN0 3
=K 2 SiF 6 +2HNO a( =1 H 2 SiF 6). The liberated 2HNO a requires 2NaOH.
Titration hot, K zSiF 4 +4NaOH=4NaF+2KF+SiO z+2H zOCi H zSiF6 ).
The sulfuric acid is determined by titration with NaOH, upon expulsion of
the accompany.ing more volatile acids. Sulfurous acid is determined by titration with standard iodine.
Special Apparatus.-Chamber burette graduated from 75 to 100 ml. in 1/20
mf. as described under the determination of sulfuric acid, oleum, mixed acids,
etc.
Platinum Weighing Tube.-Length about 5 cm., diameter 1.4 cm. The
tube fitted with a platinum cap with a loop top to facilitate removal.
Brinton, Sarver and Stoppel found that the presence of silica in the NaOH
causes' an error in th~ titration, hence it is necessary to determine the SiO z in
the NaOH and apply corrective factor if SiO z is present. Keep reagent in
ceresin-lined bottles. If Si0 2 is present, the H 2SiF 6 will be too high and HF
too ,low. Calculate the SiO z in' one mi. of the reagent and the total Si0 2 in
the total titration (cold) may now be obtained. Multiply the total SiO z by
2.393 (factor HzSiO s divided by SiO z) and deduct result from H zSiF 6 obtained
for true H zSiF 6 Multiply the total SiO z by 1.991 (6HF divided by Si0 2)
and add to the HF results for true HF content. Ref. Brinton, Sarver and
Stoppel, Ind. Eng. Chern., IS, 1880 (1923).
DETAILS OF PROCEDURE. TOTAL ACIDITY AND
HYDROFLUOSILICIC ACID
A catch weight of the acid is taken by pouring the acid by means of the
thief or directly from the paraffin bottle into the platinum weighing bottle, such
a weight being taken as will require a titration of from 75 to 100 mi. of the
normal caustic solution. (This may be judged by a preliminary run if the
approximate value is not known.)
About 10 mi. of a saturated solution of KNO s is poured into a large platinum
dish (capacity about 125 mI.), and chipped ice added. About 50 ml. of Nil
NaOH solution is run in from a burette and three drops of the strong phenolphthalein added. The platinum weighing' bottle containing the sample is
inverted beneath the surface of the caustic, the cover cautiously removed from
the bottle ,by means of a heavy platinum wire, so as to allow the acid to mix
very gradually with the standard NaOH (rapid addition is apt to cause loss of
-acid by fumes). Standard Nil NaOH is added from the burette until the first
permanent pink color is obtained. (The end-point will be uncertain and fading

2210

ACIDS

unless the solution is kept cold-O C.)

The reading of the burette is noted-

total ":/1 fNaOHI

A. The dish is now placed on ~ hot plate ~nd the solution
t.o samp e
0
warmed to about 80 C. and the titration completed with 'the N/l NaOa
solution to a permanent pink. Additional rill. required divided by weigh,t of
sample=B.
SULFURIC ACID IN HYDROFLUORIC ACID
About 5 grams of the si'lmple are weighed in the platinum capsule and transferred to a large platinum dish, the capsule being rinsed out into the dish with
water. The solution is evaporated on the steam bath to small volume (the
evaporation is assisted by passing a hot current of pure dry air over the sample,
see method on page 2233), a few drops of water are added and the evaporation
repeated; no odor should be perceptible, all the hydrofluoric, hydrofluosilicic and
sulfurous acids being expelled. The sulfuric acid is cooled, taken up with
100 m!. of CO 2-free water, three drops of strong phenolphthalein added and the
acid titrated with N /1 NaOH solution in a 50-m!' burette. The m!. titration
divided by the weight of the sample is noted as C. (See calculations at the
close of .the procedure.)
.
SULFUROUS ACID IN 'HYDROFLUORIC ACID
Ten grams of the sample are weighed in a tared platinum capsule with cover
and washed into a large platinum dish with about 75 ml. of water. N /10 Iodine
solution is added to a faint yellow. Theend-point is made more distinct by
addition of a little starch solution 'near the end of the reaction.
One mI. of N/I0 1=0.0041 g. H 2S0 3

ml. N/I0 I"+wt. of sample=D.

Calculation of Results.

!
H 2S0 4 XOA904 = H2SiF~.
H 2S0 4 XOA080=HF. :
One m!. N /10 I = 0.004~ g. H 2SO S

Factors.

Symbols.

A=mI.
B=ml.
C=ml.
D=ml.

NaOH for total acidity (cold) +wt. of sample.

of NaOH addition for H 2SiF6 (Mt)+wt. of sample.
NaOH for H 2S0 4 +wt. of sample.
-N/IO iodine+wt. of sample.

Formulae for Calculation.-If E = value oLl ml. of the standard N /1 caustic

in terms of H 2S0 4 then
Per cent HF=
Per cent

(A-~-C) XExdA08XlOO-0.2D:

H2Si~6=~BXEXOA904XlOO;

Per cent H 2S0 4 =CXEXlOO;

Per cent H 2S0 3 =DXO.0041XlOO._":-

ACIDS

2211

Residue.-This is determined by evaporation of 15 to 20 grams of the acid

in a platinum dish, and gentle ignition of the dry residue.
NOTES AND PRECAUTIONs.-Weighings should be made quickly in covered platinum
weighing bottles.
It is advisable to weigh out the sample for the sulfuric acid determination first and
start the evaporation to facil~tate more rapid results.
Iodine is preferred to permanganate for titration of H 2SO a as the latter also titrates
organic matter that is apt to occur in the acid.

COMPLETE ANALYSIS OF NITRIC ACID

The acidity of nitric acid obtained by titration with standard caustic may be
due not only to HNO a but to impurities H 2S04, HCI and lower oxides of nitrogen, hence for extremely accurate analysis it is essential to look for these
impurities and make allowances accordingly if they are found to be present.
Nitric acid may be determined directly by titration with ferrous sulfate according to the procedure given in detail, page 2214; this titration will include
combined nitrates as well as the free acid, whereas the titration with caustic
includes only free acids. In addition to the above-mentioned impurities,
commercial nitric acid frequently contains free chlorine, chlorides, chlorates,
iodine, iodides, iodates, silica, and suspended solids; the last is reported as
insoluble residue. In an analysis of nitric acid the impurities, which are known
to be injurious to the art for which the acid is used, are looked for and determined if present.
DETERMINATION OF TOTAL ACIDITY

As in case of mixed acids and, in fact, all accurate determinations of acids

with caustic, such an amount of the sample should be taken as will require a
titration within the limits of the standard chamber burette-75 to 100 m!. For
normal caustic this would require 4.726 to 6.3 grams of 100% HNO a or a fifth or
tenth of this amount for N /5 or N /10 NaOH. From the specific gravity of the
acid its approximate strength can be obtained by referring to the table for
nitric acid and calculating the volume and approximate weight required for
analysis (see example under Methods of Weighing Acids-Dilute Acids-NonVolatile under Ordinary Conditions, page 2200).
The acid is weighed in a weighing bottle, or in the Dely tube or Blay-Burkhard pipette, if it is a fuming acid. The titration is made in a casserole, the acid
being mixed with 150 to 200 m!. of CO 2 free water and titrated in presence of
phenolphthalein indicator. (Methyl-orange is destroyed by nitrous acid.)
The total acidity is expressed in terms of H 2S0 4 if other acids are present.

2212

ACIDS

ml. Nil NaOHX.049043X100

Weight of the sample
1
= per cent H 2S0 4 equivalent.

H 2S0 4X1.285=HN0 3

.
.
ml. Nil NaOHXO.()63o'18XHio
W . h
h
:I
=per cent HN0 3
DIrect calculatIOn to HN0 3
eig t of t e sam:p e

DETERMINATION OF SULFURIC ACID IN NITRIC ACID

About 10 grams of the acid are evaporated to dryness on the steam ba}h.
The residue is taken up with about 10 ml. of water and the evaporation repeated
until free from nitric fumes, the residue finally diluted 'to 100 ml. and the sulfuric
acid titrated with N 15 NaOH, using phenolphthalein or methyl-orange indicator. Gravimetrically the acid may be precipitated from a hat solution as
BaS04 by addition of barium chloride reagent l1ccording to the method for
determining sulfur.
One ml. N/5 NI!OH=0.009809 gram H 2S0 4.
BaS04 X0.4202= H 2S0 4.
X H 2S04 obtained.

Per cent=100 divided by weight of sample

"

DETERMINATION OF HYDROCHLORIC ACID IN NITRIC

ACID
A 5- to 50-gram sample is taken, that is to say, a sufficient amount of the
acjd so that a weighable amount of AgCI may be obtained. The sample is
nearly neutralized with NH 40H (it should be slightly acid with HN0 3) ,and a
slight excess of silver nitrate reagent added to the hot solution; the mixture is
stirred thoroughly, then allowed to settle for one or two hours. The AgCI is
filtered thrqugh a weighed Gooch crucible containing an asbestos mat, then
washed, dried and ignited at 130-150 C. (See ~eneral method for. the determination of chlorine.)
Factors.

AgCl XO.2474 = C1. AgCliXO.2544 =HCl.

AgCI X 0.34212 = equivalent H 2S04.

Find the per cent HCI and the' per cent equivalent H 2SQ4'_'
LOWER OXIDES.

DETERMINED AS NITROUS ACID

For practical purposes the lower oxides of nitrogen that may be present in
nitric acid are calculated to N 20 3 or HN0 2 If it is desired to report these as
N 20 4 the conversion factor given below may be used .. The lower oxides may be
obtained by titration with standard permanganate, other reducing agents being
absent. In presence of organic matter titration with standard iodine solution
should be made. (See general procedure for determination,of nitrous acid, etc.)
It makes but little difference whether the pel'manganate is added to the
sample containing nitrite or the sample added:-to a measured amount of per-

ACIDS

2213

manganate, provided in the first method the titration be made as rapidly as

possible to prevent oxidation taking place due to dilution of the sample with
water. The end-point in the first procedure is quicker and sharper.
Potassium permanganate oxidizes nitrous acid to nitric according to the
reaction 2KMn04+5HN02+3H2S04=K2S04+2MnS04+5HN03+3H20.
Therefore 1 ml. of Nil KMn04=0.02351 gram HN0 2 or 0.019 gram N 20 3.
Twenty-five ml. of the acid are diluted in a casserole to about 300 ml. with
cold water, and 25 ml. of dilute H~S04, 1 : 4 added. The solution is titrated
immediately with N 15 KMn04, the reagent being added rapidly at first and
finally drop by drop as the end-point is approached. The reaction near the end
is apt to be slow, so that time must be allowed for complete oxidation. The,
titration is completed when a pink color is obtained, that persists for three
minutes.
ml. N/5 KMn04XO.004702=gram HN0 2
The result multiplied by
100 I per cent.
.
wt. samp e
HN0 2X 1.0431 = equivalent H 2S0 4,
NITRIC ACID
From the total acidity expressed as H 2S0 4 is subtracted .the acidity due to
HCI and HN0 2 (lower oxides of nitrogen) expressed in terms of H 2S0 4 The
remainder is due to nitric afid, in terms of sulfuric acid.
H 2S0 4X 1.285 = HN0 3.
DETERMINATION OF IODINE IN NITRIC ACID
Fifty ml. of the acid in an Erlenmeyer flask is neutralized with caustic, the
mixture being cooled in running water during the operation. The solution,
poured into a separatory funnel, is made acid with dilute H 2S0 4 and, a few
drops of 1% solution of KN0 2 added, follo\ved by about 25 ml. of CS 2 or CCI 4
The mixture is shaken to extract the free iodine and the CS 2 or CC1 4 drawn off
and the extraction repeated by addition of KN0 2 and CS 2 or CCl 4 until aU the
-iodine has been extracted. Iodine present as iodide is extracted by this method.
To obtain the iodine from iodate, H 2S water is added and the extraction with
addition of NaN0 2 and CS 2 repeated.
The combined extracts are washed in a separatory funnel until free of acid.
The iodine is now titrated with standard sodium thiosulfate by adding 25 to
30 ml. of water together with 5 ml. of 1 % sodium bicarbonate solution (10 grams
NaHC0 3 per liter+1 ml. HCI).
One ml. N 110 N a 2S203 = 0.01269 gram 1:

Reactions.
2HI+2KN02+H2S04=K2S04+2H20+I2+2NO;
2Na2S203+ I 2=2NaI+Na 2S406.

2214

ACIDS

DETERMINATION OF FREE CHLORINE IN NITRIC ACID

When a current of pure air is p~ssed into nitrid acid containing free chlorine
the air blows out the chlorine. If air aspirated th.rough a sample of nitric acid
is passed through a solution of potassium ~odide the free chloripe will displace
the iodine. The libera:ted iodine may now be tit.rated with standard sodium
;
thiosulfate and the equivalent chlorine calculated:
TOTAL 'NON-VOLATILE SOLIDS

These may be determined by evaporating a large sample of 100 to 200 m!. of

the nitric acid to dryness. The residue is heated gently to expel the last traces
of nitric acid and then washed into a platinum dish, again evaporated to ~ryness
and ignited to a dull red heat. The residue is due to non-volatile solids.

FERROUS SULFATE ~ETHOD FOR THE. DIRECT

DETERMINATION OF NITRIC ACID 11
Although the test for nitric acid by ferrous sulfate in presence of cone. sulfuric acid has long been known, the reagent has not been used for an accurate
quantitative method until F. C. Bowman and W. W. Scott devel~ped the
procedure herein given. Nitric acid may be determined quantitatively in
arsenic acid by titration with ferrous sulfate containing free sulfuric acid.
The method is also applicable to the determination of nitric acid in phosphoric
acid and in s~furic acid, including oleums and mixed acids. The reaction in
phosphoric acid and arsenic acid goes further thah it does in sulfuric acid. The
following equations represent the reactions; taking place:
Reaction in Arsenic or Phosphoric Aci4s:
6FeS04+2HNOa+3H2S04=3Fe2(S04)a+2NO+4H20.
Reaction in Sulfuric Acid:
4FeS04 +2HNOa+2H2S0 4= 2Fe2(S04)a+ N 20a+3H20.
Oxidizing agents such as chlorates, ioda-tes, bromates, etc., interfere, owing
to their oxidizing action on ferrous sulfate, hence these should be absent frorti
the sample or allowance made, if appreciable amounts are present. NaCI up to
0.002 gram does not interfere ; larger amounts tend to lower results. KI and KBr
react in a similar manner to NaCI, 0.002 gram causing no interference. KN02
present in amounts up to 50% of the HNO a does ~ot interfere. The sample
should not contain over 25% water, nor should the temperature exceed 60 C.
during titration. 0.1 to .0.8 gram HNO a are ~c:urately titrated, in sulfuric acid.
11 J. Ind. Eng. C~em., 7, 766 (1915).
.
.

ACIDS

2215

Special Reagents Required. Standard Ferrous Sulfate.

A. Reagent to be Used in Titration of Nitric Acid in Sulfuric Acid, Oleum,
etc.-176.5 grams of FeS04' 7H 20 are dissolved in about. 400 ml. of water,
and 500 ml. of about 60% H 2S04 (1 vol. 66 Be. acid per 1 vol. H 2 0) are added
with constant stirring, and the solution (cooled if necessary) made up to 1000 m!.
1 m!. wiII-be equivalent to 0.02gram HNO a, the exact value being determined
by standardization.
B. Reagent for Titration of Nitric in Phosphoric or Arsenic Acid.-Ferrous
sulfate to be used should be made up as follows: 264.7 grams of FeS04 7H 2 0
is dissolved in 500 m!. of water, 50 ml. of 66 Be. H 2S0 4 (93.2%), added-and the
solution made up to 1000 ml. 1 mI. will be equal to approximately 0.02 gram'
HNO a. The exact strength is ascertained by titrating a known amount of
nitric acid in phosphoric or arsenic acid upon warming to 40 or 50 C.
Standard Nitric Acid.-The acid should contain about 40 grams of HNO a
(100%) per liter of solution, e.g., 41 ml. of the desk reagent (sp.gr. 1.42) per
liter
give the strength desired, the exact value being determined by titration
of the acid against standard caustIC.
"
Potassium Dichromate, Nj2 K 2 Cr 2 07 Solution.-The exact value in terms
of iron should be known.

will

STANDARDIZATION OF FERROUS SULFATE REAGENT

1. Titration against Standard Nitric Acid.
A. Reagent to be Used for Determination of Nitric in Sulfuric Acid.-Ten
milliliters of the standard HNO a =' O.4g. (the exact amount having been ascertained) is run into 100 ml. of 66 Be. (93.2%) H 2S04, free from oxidizing agents
(HNO a, etc.) and the resulting mixture titrated with the standard FeS04 solution according to the directions given under the G.eneral Procedure for determining HNO a in H 2S0 4, page 2217.
Weight of HNO a taken diyided by mI. of FeS04 minus 0.2 mlP=grams
HNOa ml. FeS04.
B. Standardization of the Reagent Used in the Determination of Nitric in Arsenic and Phosphoric Acids.-Ten milliliters of the standard HNO a=O.4gram
(the exact amount having been ascertained) is run into 100 ml. of H 2 As0 4, or
H aP0 4, according to the" product to be titrated, the mixture warmed and
titrated according to directions given under Procedure for determination of
HNO a in arsenic or phosphoric acid.
Weight of HNO a taken, divided by ml. FeS04=grams HNO a per ml. FeS04.
2. Standardizing Ferrous Sulfate with Standard Potassium Dichromate.
. A. Reagent to be Used in Determination of Nitric in Sulfuric Acid.-Twentyfive milliliters of N /2 K 2 Cr 2 07 (or 125 milliliters N /10 K 2 Cr 2 07) are accurately
measured out into a 250-ml. beaker and the solution titrated with the FeSO,
reagent, until the first fraction of a drop of excess produces a blue color with
potassium ferricyanide indicator on a spot plate. Usually between 19 to 20
ml. are required. The iron value of the dichromate multiplied by 0.5643 = gram
HNO a for the total ml. of FeS04 required in the titration.
12 An excess of 0.2 rol. FeSO, is required to produce the desired color reaction in 100
ro1. of pure H 2S0 4

2216

ACIDS

Calculation.
Since N /2 K 2 Cr 207 =0.024517 gram salt, 1 gram K~Cr207= 1.13882 Fe, therefore 25 mI.=0.024517X1.13882XO.5643X25=0.3939 gram HNO s equivalent.
0.3939 divided by ml. FeS04 required in the titra~ion=grams.HNOs equivalent per mi.
.
B. Reagent to be Used in Determination of HN0 s in H'JAs04 or H S P04. Thirty-eight milliliters of K 2 Cr 2 07 solution are titrated with FeS04 according to
directions given in "A." The Fe value multiplied by 0.3762 = gram H:ijOa.
. Calculation.-Thirty-eight milliliters of K 2 Cr 207=0.024517X1.13882X
0:3762 X38 = 0.3991 gram HN0 3 equivalent. 0.3991 divided by ml. FeSj)4
I
required in the titration = grams H'N0 3 equivalent per ml.
Factors.
K 2 Cr 2 07 to Fe = 1.13882, reciprocal = 0.8781.
2Fe to. HNO s =0.5643, recip. = 1.7722. 3Fe to HN0 3 =0.3762, recip.
=2.6582.
HNO s to 2FeS047H 20=8.8235 recip. =0.1133.
HNOs'to 3FeS047H20 = 13.2348, recip. =0.07556.
K 2Cr 207 to HNO s =0.6426, recip. = 1.5562. (Titration of A reagent.)
K 2Cr 207 to HNO s = 0.4284, recip. = 2.3342. (Titration of B reagent.) .

..

GENERAL PROCEDURE. DETERMINATION OF NITRIC ACID

IN SULFURIC ACID .
The procedure is applicable to the determination of nitric acid, free or
combined as nitrate, sulfuric acid being used as the medium in which the
titration is made. Although 0.1 to 0.8 gram HNO s may be accurately titrated,
it is a general practice to have the nitric acid content of the .sample taken for
the titration about the same as the amount taken in standardization of the
FeS04 reagent. A preliminary run on the original material is made, if the
approximate nitric acid content is not known . Sqlids are dissolved in water
and made to the desired volume, strong HNO s is diJuted with water. In either
case the dilution should be such that 10 ml. of; the solution will contain approximately 0.4 gram HNO s. Mixed acids and oleum containing over 10% HNO s
should be mixed with additional 66 Be. (93+per cent H 2S0 4) and made to a
definite volume, an aliquot part being taken for titration.
EVALUATION OF NITRIC ACID OR NITRATES
If the nitric acid is known to be free of other acids it may be titrated directly
with caustic; combined nitrate cannot be titrated with caustic, but may be
accurately determined by the ferrous sulfate method. The approximate
strength of the HNO s or salt having been determ(ned on 1 ml. or 1 gram sample
(if the material is a solid), the requisite amount is weighed and made to volume,
10 ml. ot which should contain not more than 0.8 gram or less than 0.1 gram
HNO s, preferably about 0.4 gram.
Example.-Suppose 1 ml. required a titration of 43.18 ml. FeS04, 10 ml.
would require a titration of 438 mI., whereas_2l} rill. is desired. 438 divided
by 20 = approximately 22, e.g., the dilution sholild be to 22 volumes. 23 mr.

ACIDS

2217

of the solution diluted to 500' ml. will give a mixture of the desired strength.
23 m]. are accordingly weighed in a weighing bottle, the acid washed into a
beaker transferred then to the graduated 500-m!' flask and made to volume.
The preliminary run may be made in two or three minutes,
Titration.-A 250-mI. beaker containing 100 mI. of concentrated, nitric free,
H 2S0 4 (93 + %) is placed in a large casserole or deep porcelain dish containing
cold water. 10 m!. of the sample are measured out in an accurately marked
pipette, graduated to contain exactly 10 m!. The solution is' run under the
surface of the sulfuric acid, the delivery tip of the pipette being kept in constant
circular motion to prevent too much local heating. Since the sides of the
beaker are cooled, the tip of the pipette should be kept against the sides jn
the circular sweep during the delivery. By this procedure loss of nitric acid is
reduced to the minimum.
'
The ferrous sulfate solution is now added from a burette in a fine stream
until the yellow color that first forms takes on a faint brownish tinge (dirty
yellow). The pipette is now rinsed out by sucking up the mixture and draining
it back into the beaker. The titration is now completed, adding the FeS04
cautiously drop by drop until the yellowish brown color again appears, a drop
in excess producing an appreciable darkening of the solution. A larger excess
produces a brownish red color. With small amounts of HNO a a pink color
will be obtained, instead of the yellowish brown. The end-point 'once recognized is readily duplicated.
Calculation.-The mI. titration minus the blank 0.2 ml. multiplied by the
factor for FeS04=weight of HNO a. HNOax'IOO+wt.=per cent.
Example.-Suppose 10 m!. equivalent to 1/50 of a 42-g. sample weighed,
requires 22 ml. FeS04 whose value = 0.02 g. HNO a per mI., then (0.44 X'lOO) /0.84
=52.4% HNO a.
DETERMINATION OF NITRIC ACID IN OLEUM OR IN
MIXED ACIDS. FERROUS SULFATE METHOD
The rapidity and accuracy of the method for determining HNOa in sulfuric
acid makes it valuable for determining nitric~acid in ole urns and mixed acids.
Nitrated oleums may be weighed and titrated without diluting to definite
volume. Mixed acids containing large percentages of nitric acid, however,
require dilution with H 2S0 4 , as stated under General Procedure.
Procedure.-The sample may be weighed in a Dely weighing tube (see
analysis of oleum and mixed acids), or in a standard pipette (5 mI. generally
taken = 9.61 grams). If the latter is used, the sample is sucked into the pipette,
a rubber tube, with glass bead valve, being attached to the upper end, to which
suction is applied without danger of drawing SOa fumes into the mouth. A
little vaseline placed on the tip of the pipette prevents loss of acid during the
weighing. In routine analysis, where a large number of daily samples of oleum
are analyzed, and the specific gravity of the oleum does not vary appreciably,
5- to 10-mI. samples may be drawn out, by means of a pipette, and titrated'
without weighing, the weight being calculated from the' gravity.
The acid is run under cold concentrated HoSO. (930/,,). and titrated accord-

2218

ACIDS

ing to directions under General Procedure for Nitric Acid. A blank of 0.2 ml. 13
having been deducted, ml. FeS04XHNO a factor for FeS04XlOO divided by
wt. taken = per cent HNO a.
Correction Factor.-In making a number of ruAs with varying amounts of
HNO a, it was found that small quantities of nitric ~cid required a proportionally greater amount of FeS04 than larger quantitieS of HNO a. For example,
0.07392 gram HNO. required 3.9 ml. FeS04, four times the amount of HNO.
required 15 ml. FeS04, in place of 15.6 (3.9X4), and six times 0.07392 gram
HNO a required 22.5 ml. FeS04 in place of 23.4. It was observed that'even
traces of HNO a required a titration of over 0.2 ml. It isevident that a deduction of 0.2 ml. makes the titrations multiples of the lowest, e.g., 3.7, 14.1\ and
22.3. Again it was found that standardization of FeS04 with HNO a checked
the dichromate factor when 0.2 ml. was deducted from the first series of titrations. This led to the conclusion that a constant blank of 0.2 ml. should be
deducted from the ferrous sulfate titrations of nitric acid in presence of 100 ml.
of nitric free sulfuric acid (66 0 Be.).
Comparison of results:
FeS04 value by RN0 3 corrected = 0.02067 gram. Uncorrected = 0.02045' gram
RNO .
FeSO. value by K 2Cr 207 titration =0.02083 gram RN0 3
Accuracy of the Ferrous Sulfate Method.-Results obtained by the ferrous
sulfate method agree closely with those obtained by the nitrometer. ['he
following data were obtained by Mr. B. S. Clark, by the FeS04 method, on
nitrated oleums. The figures below the first row are checks obtained on these
samples by purchasers of the acid, the nitrometer methC!d being used.
FeSO. method.

2.40; 2.82; 3.23; 3.35; 3.52; 3.50; 3.48; 3.57; 3.53; 3.56.

Nitrometer method.

2.35; 2.79; 3.26; 3.39; 3.57; 3.53; 3.50; 3.58; 3.5'7; 3.56.

Difference.

0.05; 0.03; 0.03; 0.04; Op5; 0.03; 0.02; 0.01; 0.04; 0.00.

DETERMINATION OF NITRIC ACrD IN ARSENIC AND PHOSPHORIC ACIDS BY THE FERROUS SULFATE METHOD

A direct procedure for the determination of nitric acid in arsenic acid or

phosphoric acid has be~n sought on account of the inaccuracy of the evaporation
method, since it is difficult to completely expel HNO a from these"acids. Ferrous
sulfate, in, presence of sulfuric acid, quantitatively titrates nitric in arsenic acid,
the following reaction taking place:
6FeS04+2HNOa+3H2S04+xHaAs04=3Fe~(S04)3+2NO+4H20+xHaAsO4.

The procedure is applicable to the determination of nitric acid in phosphoric

acid, the end-point being sharper in this acid thim in arsenic. The procedure
gives very excellent results in either acid and is recommended for accuracy and
rapidity.
. 13 Back titrations of the.excess of FeSO. may be made with standard K 2Cr 207, using
I
the ferricyanide spot test for ferrous iron.
" Sp.gr. of twelve average sampl~s of oleum ha.!J,./o,"difference of only 0.01. 5 ml.
weighs 9.61 grams. This is found convenient for analysis. 10 ml. = 19.22 grams will
usually give a titration of about 20+ml. on the usual nitrated oleum.

ACIDS

2219

Standardization of Ferrous Sulfate has already been given under special

reagents. It must be remembered that the arsenic or phosphoric acid diluents
should be free from nitric acid or the blank on 100 ml. be ascertained and deducted from titrations made in' this diluent.
. Procedure.-The amount of the sample to be taken is governed by the nitric
acid present as an impurity. This may be quickly determined by a preliminary
run on a lO-ml. sample, the diluent being the same acid (HNO a free) as the acid
titrated.
Example.-Suppose 10 ml. require a titration of 4.5 ml. of FeS04, whereas a
titration of 20 ml. is desired, then 20XlO+4.5=44.44 m!. of the sample required.
" The required amount of the acid is measured out and weighed, if its sp.gr. is
not known. The acid is poured into a 4-in. casserole and diluted with ,100 m!.
of nitric free acid of the same kind as that being titrated. The mixture is
gently warmed to 40 to 50 C. and titrated with standard ferrous sulfate reagent
to a permanent yellowish brown. Towards the end of the titration the acid will
boil with each addition of the FeS04 and the characteristic reddish yellow fumes
will be given off. (This does not occur in titrations of HNQa in H 2S0 4.)
When very small amounts of nitric acid are present it is often necessary to
add a known amount of HNO a to start the reaction. The titration in excess of
that required by the added HNO a is due to the nitric acid in the sample. Very
small amounts of HNO a produce a pink color.
Calculation m!. FeS04XHNO a factorfor FeS04XI00 divided by weight taken
= % HNO a
'Factors.

Fe to HNO a=0.3762.

Reciprocal =2.6582.

NOTE.-In a 20-gram sample 1 ml. 0.02 reagent =0.1 % per ml.

DETERMINATION OF NITROUS ACID OR NITRITE

PERMANGANATE METHOD
Principle.-Potassium permanganate reacts with nitrous acid or a nitrite
as follows:
5N 20a+4KMn04+6H 2S0 4= 5N 205+2K 2S04+4MnS04+6H 20. ;
5HN0 2 +2KMn04+3H 2S0 4= 5HNO a+K2S0 4 +2MnS04+3H20.
Details of the genera1.procedure for nitrites is given on page 653, Vol. 1.
Nitrous acid in nitric acid may be determined as stated on page 2212. Lower
oxides of nitrogen in mixed acids are (l~termined as outlined on page 2234.

2220

ACIDS
NITRIC ACID
By W. C. FERGUSON

Degrees
Baume.

10 00
10.25
10.50
10.75
11.00
11.25
11.50
11.75
12.00
12.25
12.50
12.75
1300
13.25
13.50
13.75
14.00
14.25
14 50
14.75
15.00
15.25
15.50
15.75
16.00
16.25
16 50
16.75
17 00
17.25
17.50
17.75
18.00
18.25
18.50
18.75
19.00
19.25
19.50
, 19.75
20.00
20.25
.,20.50
20.75
21.00

Sp. Gr.
60
6o oF .

1 0741
1 07<31
1.0781
1.0801
1.0821
1.0841
1.0861
1.0881
1.0902
1.0922
1.0943
1.0964
1.0985
1.1006
1 1027
1.1048
1.1069
I 1090
1.1111
1 1132
1 1154
1.1176
1.1197
1.1219
1.1240
1.1262
1 1284
1.1306
1.1328
1.1350
1.1373
1.1395
1.1417
1.1440
1.1462
1.1485
1.1508
1.1531
1.1554
1.1577
1. 16CO
1.1624
1.1647
1.1671
1.1694

Degrees Per Cent

Twaddell. HNO

14.82
15.22
15.62
16.02
16.42
16.82
17.22
17.62
18.04
18.44
18 86
19 28
19 70
ZO 12
20 54
ZO 96
21.33
21.80
22.22
22 64
23.08
23.52
23.94
24.38
2480
25 24
25 68
20.12
2G.5B
27.CO
27.46
27.90
28.34
28.80
29.24
29.70
30.16
30.62
31.08
31.54
3200
32.48
32.94
33.42
33.88

Degrees
Baume.

12.86
13.18
13.49
13.81
14.13
14.44
14.76
1507
15 41
15.72
16.05
16.39
16.72
17.05
17 38
17.71
1304
18 37
18.70
19.02
19 36
19.70
20.02
20.36

21.25
21.50
21.75
22.00
22.25
22.50
22.75
23.00
23.25
23.50
23.75
24.00
24.25
24.f50
21,73
23 CO
25.25
25 . .';0
25.75
23.CO
26.25
26.[0
26 75
27.CO

::0 CD

27

21.03
21 36
21.70
22 04
22 38
22 74
23.08
23 42
23.77
24.11
24.47
24.82
25.18
25.53
25 88
26 24
26.61
26.96
27.33
27.67

27.50
27.75
28.00
28.25
28.50
28 75
29.00
29 25
29.50
29.75
30.CO
30.25
30.0
30 75
31 OQ
31.25
31.50
31 75
32.00
32.25

:::3

. Sp, Gr.
~F.
609

1.lp8

1,1741
1.1765
1.1789
1.1813
1.1837
1.1861
1.1885
1.1910
1.1934
1.1959
1.1983
1.2008
1.2033
1 2058
1.2083
1 2109
1 2134
1 2160
1 2185
1.2211
1.2236
1.2262
1 2288
1.2314
1 2340
1 2367
1.2393
1 2/420
1.;2446
1 '2473
1 2500
1.2527
1.2554
1.2582
I.Z00!>
1.2637
1.2664
1.2692
1.2719
1.2747
~.2775

1.2804
1 2832
1.2861

Degrees l'er.Cent
Twaddell. HNO

34 36
34 82
35.30
35.78
36 26
36.74
37.22
37.70
38.20
38.68
39.18
39.66
40.16
40 66
41.16
41.66
42.18
42 68
4320
43.70
44.22
44.72
45.24
45.76
46.28
46 80
47.34
47.86
48.40
48.92
4946
50.00
50 ..54
51.08
51.64
52.18
52 :14
53 ::3
53.84
54 28
54.04
55.[0
5603
56.64
57.22.

28.02
28.36
28.72
29.07
29.43
29 78
30.14
30.49
30 86
31 21
31.58
31.94
32.31
32 68
3305
33.42
33 80
34 17
34.56
34.94
35.33
35.70
3609
36.48
36 87
37 26
37.67
3806
3846
38.85
39.25
39 66
~0.06

<0.47
~0.89

41 30
41 72
42 14
42 58
4300
43A4
43 89
44 34
44.78
45.24

2221

ACIDS
NITRIC ACID (Continued)
Degre~s

BaulDe.

I Sp.
60" Gr.
600 F.

32.50 ,j 1.2889
32.75 1.2918
33.00 1.2946
33.25 1.2975
33.50 1.3004
33.75 1.3034
34.00 1,3063
34.25 1.3093
34.50 1.3122
34.75 1.3152
35.00 1.3182
35.25 1.3212
35.50 1.3242
35.75 1.3273
36.00 1.3303
36.25 1.3334
36.50 1.3364
36.75 1.3395
37.00 1.3426
37.25 1.3457
37.50 1.3488
37.75 1.3520
38.00 1.3551
38.25 1.3583
38.50 1.3615
38.'75 1.3647
39.00 1.3679
B9.25 1.3712
39.50 1.3744
39.75 1.3777
40.00 1.3810
40.25 1.3843
40.50 1.3876

per Cent
Degrees
Twaddell.
HNO,.

57.78
58.36
58.92
59.50
60.08
60.68
61.26
61.86
62,.44
63"04
63.64
64.24
64.84
65.46
66.06
66.68
67.28
67.90
68.52
69.14
69.76
70.40
71.02
71.66
72.30
72.94
73.58
74.24
74.88
75.54
76.20
76.86
77.52

45.68
46.14
46.58
47.04
. 47.49
47.95
48.42
48.90
49.35
49.83
50.32
50.81
51.30
51.80
52.30
52.81
53.32
53.84
54.36
54.89
55.43
55.97
56.52
57.08
'57.65
58.23
58.82
59.43
60.06
60.71
61.38
62.07
62.77

Degrees
Baumel.

40.75
41.00
41.25
41.50
41.75
42.00
42.25
42.50
42.75
43.00
43.25
43.50
43.75
44.00
44.25
44.50
44.75
45.00
45.25
45.50
45.75
46.00
46.25
46.50
46.75
47.00
47.25
47.50
47.75
48.00
48.25
48.50

Sp. Gr.
60 F"
60

Degrees
Twaddell.

1.3909
78.18
1.3942
78.84
1.3976
79.52
1.4010
80.20
1.4044
80.88
1.4078
81 :56
1.4112
82.24
1.4146
82.92
1.4181
83.62
1.4216
84.32
1.4251
85.02
1.4286
85.72
1.4321
86.42
1.4356
87.12
1.4392
87.84
1.4428
88.56
1.4464
89.28
1.4500
90.00
1.4536
90.72
1.4573
91.46
1.4610
92.20
1.4646
92.92
1.4684
93.68
1.4721
94.42
1.4758
95.16
1.4796
95.92
1.4834
96.68
1.4872
97.44
1.4910
98.20
1.4948 98.96
99.74
1.4987
1.5026 100.52

Per Cent
HNO,.

63.48
64.20
64.93
65.67
66.42
67.18
67.,95
68.73
69.52
70.33
71.15
71.98
72.82
73.67
74.53
75.40
76.28
77.17
78.07
79.03
80.04
81.08
82.18
83.33
84.48
85.70
86.98
88.32
89 76
91.35
93.13
95.11

Specific Gravity determinations were made at 60 F., compare(j with water at 60 F.

the Specific Gravities, the corresponding degrees Baume were calculated by the
following formula:
Baume=145-~.
Sp. Gr.
Baume Hydrometers for use with this table must be graduated 1)y the above formula,
which formula should always be printed on the scale.
Atomic weights from F. W. Clarke's table of 1901. 0= 16.
ALLOWANCE FOR TEMPERATURE:
At 10 - 20 Be. - 1/30" Be. or .00029 Sp. Gr.::::: l' F.
20 - 30 Be. -1/23 Be. or .00044"
"::::: 1 F.
30 - 40 Be. - 1/20 Be. or .00060" "::::: 1 F.
40 - 48.5 Be. _1/17 Be. or .00084"
"::::: 1 F.
Fl'OIn

AU1'HORITY- W. C. FERGUSON.

'I.'h!s table has been approved and adopted as a Standard by the Manufacturing Chemists'
.Association of the United States.
W. H. BOWER,
.JAS. I,. l\fORGAN.
HENRY HOWARD, ARTHUR WYMAN,
A. G. ROSENGARTEN,
Executive Committee
New York, May 14. 1903.

2222

ACIDS
NITRIC ACID
LUNGE AND RET

Specific
Gravity

100 parts by weight

contain

IS

in 7
vacuo

1.000
1.005
1.010
1.015
1.020
1.025
1.030
1.035
1.040
1.045
1.050
1.055
1.060'
1.065
1.070
1.075
1.080
1.085
1.090
1.095
1.100
1.105
1.110
1.115
1.120
1.125
1.130
1.135
1.140
1.145
1.150
1.155
1.160
1.165
1.170
1.175
1.180
1.185
1.190

%
l'I,O.

%
DO.

0.08
0.10
0.85
1.00
1.62
1.90
2.39
2.80
3.17
3.70
3.\:4
4.60
4.71
5.50
5.47
6.38
6.22
7.26
6.S7
8.13
7.71
8.99
8.43
9.84
9.15 10.68
9.87 11.51
10.57 12.33
11.27 13.15
11.96 13.95
12.64 14.74
13.31 15.53
13.99 16.32
14.67 17.11
15.34 17.89
16.00 18.67
16.67 1'J.45
17.34 20.23
18.00 21.00
18.66 21.77
-19.32 22.54
19.98 23.31
20.64 24.08
21.29 24.84
21.g4 25.60
22.60 26.36
23.25 '27.12
23 SO 27.88
24.54 28.63
25.18 29.38
25.83 30.13
26.47 30.88

tains grams

Specific
Gravity

100 parts

N,O.!HNO.

40
in vacuo

I liter con-

iSO

1
8
16
24
33
40
49
57
64
73
81
8)

by weight I liter concontain

tainsgrams
I

N,O.

1.195 27.10
1.200 27.74
1.205 28.36
1.210 28.99
1.215 29.61
1.220 30.24
1.225 30.88
1.230 31.53
1.235 32.17
1.240 32.82
1.245 33.47
1.250 34.13
1.255 34.78
1.260 35.44
1.265 36.09
1.270 36.75
1,275 37.41
/1.280 38.07
1.285 38.73
1.290 39.39
1.295 40.05
1.300 40.71
1.305 41.37
1.310 42.06
1.315 42.76
1.320 43.47
1.325 44.17,
1.330 ,44.81)"
1.3325 45.26
1.335 45.62
1.340 46.35
1.345 47.08
1.350 47.82
1.355 48.57
1.360 49.35
1.365 50.13
1.370 50.91
1.375 .51.69
1.380 52.52

1
10
19
28
38
47
57
66
75
85
94
104
!)7
113
105 123
113 132
121 141
129 151
137 160
145 169
153 179
161 188
170 198
177 207
186 217
195 227
202 236
211 246
219 256
228 266
237 276
245, 286
254 296
262 306
271 316
279 326
288 336
297 347
306 3157
315 367
1

H:O.

N,O.!HNO.

31..62
32.36
33.09
33.82
34.55
35.28
36.03
36.78
37.53
38.29
39.05
39.82
40.58
41.34
42.10
42.87
43.64
44.41.
45.18
45.95
46.72
47.49
48.26
49.07
49.89
50.71
51.53
52.37
52.80
53.22
54.07
54.93
55.79
56".66
57.57
58.48
59.39
60.30
61.27

324
333
342
351
360
369
378
387
397
407
417
427
437
447
457
467
$77
487
498
508
519
529
540
551
562
573
585
597
603
609
621
633
645
658
'671

684
698
711
725

378
g88
339
409
420
430
441
452
463
475
486
498
509
521,
533
544
556
568
581
593
605
617
630
643
656
669
683
697
704
710
725
739
7.53
768
783
798
814
829
846

2223

ACIDS
NITRIC ACro (Continued)
Specific
Gravity

100 parts by weight

IS

40

in vacuo

--1.3833
1.385
1.390
1.395
1.400
1.405
1.410
1.415
1.420
1.425
1.430
1.435
1.440
1.445
1.450
1.455
1.460
1.465
1.470
1.475
1.480
1.485
1.490

I liter contains grams

Specific
Gravity

R~O.

N,O.!RNO.

in vacuo

61.92
62.24
63.23
64.25
65.30
66.40
67.50
68.63
69.80
70.98
72.17
73.39
74.68
75.98
77.28
78.60
79.98
81.42
82.90
84.45
86.05
87.70
89.60

735
739
75.1
768
783
800
816
832
849
867
885
903
921
941
961
981
1001
1023
1045
1068'
1092
1116
1144

contain

~N,O.

53.08
53.35
54.20
5:>.07
55.97
56.92
57.86
58.83
59.83
60.84
61.86
62.91
64.01
65.13
66.24
67.38
68.56
69.79
71.06
72.39
73.76
75.18
76.80

JOO

parts by weight I liter concontain

tainsgrams

ISO

857
862
879
896
914
933
952
971
991
1011
1032
1053
1075
1098
1121
1i44
1168
1193
1219
1246
1274
1302
1335

40

--1.495
1.500
1.501
1.502
1.503
1.504
1.505
1.506
1.507
1.508
1.509
1.510
1.511

1.512
1.513
1.514
1.515
1.516
1.517
1.518
1.519
1.520

%
N,O.
!

78.52
80.65
81.09
81.50
81.91
82.29
82.63
82.94
83.26
83.58
83.87
84.09
84.28
84.46
84.63
84.78
84.92
85.04
85.15
85.26
85.35
85.44

R~Os

N,O, !RNO.

1174
1210
1217
1224
1231
1238
1244
1249
1255
1260
1265
1270
1274
1277
1280
1283
9~.07
1287
99.21 1289
99.34 1292
99.46- 1294
99.57 1296
99.67 1299

91.60
94.09
94.60
95.08
95.55
96.00
96.39
96.76
97.13
97.50
97 84
98.10
98.32
98.53
98.73
08.90

1369
1411
1420
1428
1436
1444
1451
1457
1464
14/0
1476
1481
1486
1490
1494
1497
1501
1504
1507
1510
1512
1515

2224

ACIDS

VOLUMETRIC DETERMINATION OF PHOSPHORIC ACID AND

ITS SALTS (METHOD OF S~ITH)

In titrating f~ee phosphoric acid it must be borrie in mind that NaH 2 P0 4

reacts acid to phenolphthalein and neutral to methyl orange, while Na 2HP0 4 is
neutral to phenolphthalein and basic to methyl orange. The reactions are
shown as follows:
HaP04+2NaOH=2H20+Na2HP04 (neutral to P. acid to M.O.).
\
HaP04=NaOH=NaH2P04=H20+NaH2P04 (neutral to M.O. acid to P).

J. H. Smith outlines the following process for analysis of phoSph9ric Ilcid

free and combined:
.
Procedure.-Take five or six grams of the material and dissolv.e in ,the
minimum quantity of distilled water. Add two dtops of phenolphthalein. If
the solution is not colored pink by the indicator the absence of trisodium
phosphate (NaaP0 4 ) and sodium carbonate (Na2COa) may be assumed. Heat
to 55 C. and titrate with N /1 NaOH to the characteristic pink en:d-point, being
plld}cul,u)r cllTelu) to keep t.M tAWPL'VlltuTe Ilt 55 C. when near the endpoint. This precaution should be observed in all the titrations as the temperature governs the accuracy of the titrations (see paper by Smith).
Record this titration as " A " and the ml. N aC)H required as (a).
The solution is now titrated back with N /1 HCI, using a few drops of methyl
orange as indicator, to the pink end-point.
Record this titration as " B" and the ml. acid required as (b).
Calculation of the Composition.-If the solution contains only Na 2HP0 4
+NaH 2 P0 4 or NaH 2 P0 4 +HaP0 4 or only one of these compounds we may
assume the following:
If (a) is greater than (b)
(a) - (b) = ml. equiv~Ient 9f N /f' HaPO 4,
(b) =ml. equivalent of N:;1 NftH 2P0 4

If (b) is greater than (a)

(b) -(a) =Na 2HP0 4,
(a) = NaH 2P0 4

If (a)

(b)

Each titration = N aH 2PO 4 plesent.

If (a) =2(b)

Only H aP0 4 is present .

To Ascertain the Presence of Other Salts'TAdd N /1 HCI (ml. HCI the
same as ml. in titration B). Call this (b l ) . Boil for at least 15 minutes. By
this procedure meta phosphoric acid or its salts is converted to the orthophosphate form, alkaline salts containing carbonlttes are destroyed and CO 2
evolved. Cool the solution to 55 C. and titrate back with N (I N aOH first to
" B " point in which (b fl ) ml. of alkali is used, an~finally lo third end-point ( C),
where the phenolphthalein pink color is reachetf._ ~rhe ml. N /1 NaOH required
is recorded as (c).

ACIDS

2225

In titrating back after boiling, the (b") amount of alkali required to reach
the methyl orange point where the pink coloration just appears should always
be the same amount as the (b') excess of acid added for boiling (except in the
case of polyphosphates being present-see paper). A slight loss of HCI during
boiling may cause (b") to be less than (b'). In any case it is preferable to
consider the point reached after addition of the (b") quantity of alkali as the
identical point with B, so that (b') and (b") are not used in the calculation, but
simply employed as a comparative check on each other.

If the" C" point coincides with the" A " point, i.e., if (c) = (b) metaphos,phoric acid and its salts as well as carbonates in the original substances are
excluded.
If (c) is greater than (b) the presence of meta phosphates is indicated.
If (c) is less than (b) carbonates are evidently present.
With the aid of the three fixed points" A," " B," " C " and the amounts of
acid or alkali required to reach them, i.e., (a), (b) and (c) it is possible to calculate the percentages of practically all the phosphoric acids and their salts, which
m~y exist together in a compound, including the carbonates and the free alkali,
which may be present with them.
CALCULATION 'OF CONSTITUENTS IN A MIXTURE OF
PHOSPHATES
If (a) is greater than (b)
(a) -,(b) =H aP0 4 ,
2(b) -'(a) = NaH 2 P0 4 ,
(c) -(b) = NaPOa (generally nil),
If (a) is greater than 2(b) ,
Then (a) -2(b) = measure of metaphosphoric acid or the foreign acid

present.
If (a) is less than (b)
(a) is the measure of N aH 2 P0 4 , (b) - (a) = N a 2 HP0 4 ,
(c)-(b)=NaP03 (usually nil),
With alkaline salts" C" coincides with" A," i.e., when no carbonatei:;
present.
.
If (a) is less than (b) then
(a) is a measure of Na 3P0 4 ,
(b)-(a)=Na 2 HP0 4
Where (c) is less than (b), i.e., when carbonate is present.
If (a) is less than (b)
(a)+(c) -(b) = Na 3P0 4 ,
(b) -(a) = Na 2 HP0 4 ,
(b) -(c) = Na 2 C0 3.
Where no carbonate is present and (a) is greater than (b).
(b) = Na 3P0 4 ,
tCa-b) =Na 20.
Where carbonate is present and (a) is greater than (b)
(b) -(c) =Na 2 CO a,
(c) =Na 3P0 4 ,
tea-b) =Na 2 0.
In calculating the weight of constituents from the above formulae, since
normal solutions are used in the titrations each ml. is equivalent to molecular
weights divided by 1/1000. In case 'of Na 2 C0 3 and Na 2 0 the absolute weights
are 1/2000 of the respective' molecular weights.
14

J. H. Smith, Soc. Chern. Ind., 36, 415 (1917).

ACIDS

2226
Formulae Mol. Wt.

H 3 P0 4 98.14
NaH 2P0 4 .120.05
Na 2HP0 4 .142.05
Na 3P0 4 164.04

1 mi. Equivalent of
Titration

Formulae Mol. Wt.

0.09814 g.
0.12005
0.14205
0.16404

NaP0 3 80.05
Na 2C0 3 106.01
Na 20 ...... 62.00

1 mi. Equivalent of
Titration

0.08005 g.
0.053005
0.031

With water of crystallization N aH 2 P0 4 H 20 = 138.07, N aH 2P0 4 2Hi)

156.09, N.a 2 HP0 4 12H 20 = 358.24, N aaP0 4 12H 20 0= 380.23, N a 2 CO a10H 20
=286:17, Na 2 CO aH 2 0=124.02.

NOTEs.-In presence of meta or pyro acids and their salts it is well to verify the
results with qualitative tests. When pyro acid is present point "A" 'is obscured and
lies much nearer "B," but the correct "A" point is "C" obtained after. boiling with
the excess acid.
In presence of meta phosphate it is advisable to repeat titrations employing N /1
H 2S0 4 and evaporating to near dryness to completely convert the meta to ortho compound.
.
Phosphates attack glass combining with the alkali present. The presence of NaCI
obtained.in the titrations counteracts this action.

2227

ACIDS
PHOSPHORIC ACID AT 17.5 0
Per
Specific
GrAvity. Cent.
P,O

-1.809
68.0
1.800
1.792
1.783
1. 775
1.766
1.758
1.750
1. 741
1.733
1.725
1.717
1.703
1.701
1.693
1.685
1.677
1.669
1.661
1.653
1.645
1.637
1.629
1.621
1.613
1.605
1.597
1.589
1.581
1.574
1.566
1.559
1.551
1.543
1.536
1.528
1.521
1.513
1.505
1.498
1.491
1.484
1.476
1.469

67.5
67.0
66.5
66.0
65.5
65.0
64.5
64.0
63.5
63.0
62.5
62.0
61.5
61.0
60.5
60.0
59.5
59.0
58.5
58.0
57.5
57.0
56.5
55.0
55.5
55.0
54.5
54.0
53.5
53.0
52.5
52.0
51.5
51.0
50.5
50.0
49.5
49.0
48.5
48.0
47.5
47.0
46.5

Per Cent.
H.t: 0 "

Specific
GrAvity.

93.67
92.99
92.30
91.61
90.92
90.23
89.54
88.85
88.16
87.48
86.79
86.10
85.41
84.72
84.03
83.34
82.65
81.97
81.28
80.59
79.90
79.21
78.52
77.83
77.14
76.45
75.77
75.08
74.39
'73.70
73.01
72.32
71.63
70.94
70.26
69.57
68.88
68.19
67.50
66.81
66.12
65.43
64.75
64.06

1.462
1.455
1.448
1.441
1.435
1.428
1.422
1.415
1.40;)
1.402
1.396
1.389
1.383
1.377
1.371
1.365
1.359
1.354
1.348
1.342
1.336
1.330
1.325
1.319
1.314
1.308
1.303
1.298
1.292
1.287
1.281
1.276
1.271
1.265
1.260
1.255
1.249
1.244
1.239
1.233
1.228
1.223
1.218
1.213

Per Cent.
H,PO,.

Per
Specific
GrAvity. Cent.
P.O.

Per cent.
H.PO,.

63.37
62.68
61.99
61.30
60.61
59.92
59.23
58.55
57.86
57.17
56.48
55.79
55.10
54.41
53.72
53.04
52.35
51.66
50.97
50.28
49.59
48.90
48.21
47.52
46.84
46.15
3~.0
45.45
44.77
32.5
44.08
32.0
43.39
31.5
31.0 42.70
30.5
42.01
41.33
30.0
40.64
29.5
29.0
39.95
39.26
28.5
28.0
38.57
27.5
37.88
37.19
27.0
26.5
36.50
26.0
35.82
25.5
35.13
25.0 34.44
24.5
33.75

1.208 24.0
1.203 23.5
1.198 23.0
1.193 22.5
1.188 22.0
1.183 21.5
1.178 21.0
1.174 20.5
1.169 20.0
1.164 19.5
1.159 19.0
1.155 18.5
1.150 18.0
1.145 17.5
1.140 17.0
1".135 16.5
1.130 16.0
1.126 15.5
1.122 15.0
1.118 14.5
1.113 14.0
1.109 13.5
1.104 13.0
1.100 12.5
1.096 12.0
1.091 11.5
1.087 11.0
1.083 10.5
1.079 10.0
9.5
1.074
9.0
1.070
8.5
1.066
8.0
1.062
7.5
1.058
7.0
1.053
6.5
1.049
6.0
1.045
1.041 5.5
5.0
1.037
4.5
1.033
4.0
1.029
3.5
1.025
3.0
1.021
2.5
1.017

33.06
32.37
31.68
30.99
30.31
29.62
28.93
28.24
27.55
26.86
26.17
25.48
24.80
24.11
23.42
22.73
22.04
21.35
20.66
19.97
19.28
18.60
17.91
17.22
16.53
15.84
15.15
14.46
13.77
13.09
12.40
11.71
11.02
10.33
9.64
8.95
8.26
7.57
6.89

Per
Cent.
P,O

46.0
45.5
45.0
44.5
44.0
43.5
43.0
42.5
42.0
41.5
41.0
40.5
40.0
39.5
39.0
38.5
38.0
37.5
37.0
30.5
30.0
35.5
35.0
34.5
34.0
33.5

6.20

5.51
4.82
4.13
3.44

ACIDS

2228

ANALYSIS OF SULFURIC ACID

I

Sulfuric acid made by the contact process is exceeqingly pure, the principal
impurity being iron, which causes turbidity in strong itcid. The' acid made by
this process is 99% strength. This is diluted to 66 'Be (93.19% H2S0~) and
is known as oil of vitriol. The acid is also commonly (marketed as 60 Be acid
(77.67% H 2S0 4) and 50 Be acid (62.18% H 2S0 4) obtained by further dilution
of the stronger acid. 50 0 Be acid is also obtained by the chamber process; this
acid is not as pure as that produced by the contact process. The impurities
occurring in sulfuric acid are iron, lead, copper, zinc, antimony, seleniufn,
arsenic, sulfur dioxide, hydrochloric acid, hydrofluoric acid, nitric acid.
Sulfuric acid up to 93% strength may be determined with a fair degree of
accuracy by ascertaining its specific gravity by means of a hydrometer and
referring to the tables on sulfuric acid.

1.84 4

/_
....

;;:
';J

""If)

1.83

vi

..........

~~

""

1,83593 . 19

!/

9
V ~ "-

1.'377

'\

II

,;..,
+
.;
~

1\

C)

...

1\

D.

94

95

96

97

98

9?

100

20

Percent H, 50 4

40

60

80

100

J.837
1.817

Percent free 50 3

FIG. 306.

FIG. 307.

Specific yravity Charts-Sulfuric Acid.

H,So.+7H,O H,50,.SH,0 H,5o.;411,0 I1,SO,+311,0 H,50,'~li,O H,~o.+IH,()
43.7

52.1

57.6

64.5

73.1

84.5

Percent HzSO.

10

20

30

40

50

100

90

70

60

60
40

,0

1/ \. I

-.....

u:
~ 0
g>
C>

20

Q.;

\.

1.1:',<'

11

/,

J},

_1

60

/,q,
IlL<.

_l

80
-100

308.

/'-.

'"

I'\.
\

/.,

L~

V'

- I fo,o

,
./< I

-/

") ,f>

..(

\I
,\I

po

C.'

, -550- 288

,-1\

If'<

u.: 40

FIG.

""

J.V

/
/

r\.

500" Z60

-450 =

..,
c

232~

'""
"

400= 204 ~
C>

f
'I

350' I77rD

.. 300' 149

...;

'c50= IZI

Chart Showing Freezing- and Boilirrg:j;oints of Sulfuric Acid

of Varying Concentration.

on

ACIDS

2229.

Sulfuric acid readily absorbs SOs so that its acidity may be considerably
over 100%. This acid, commonly known as oleum, fumes when exposed to
the air due to its low vapor tension, the SOs combining with the moisture of the
air with formation of H 2S0 4 mist.

SULFURIC ACID

METHOD OF ANALYSIS

15

Procedure for Titration with Standard NaOH.-Concentrated sulfuric acId

absorbs moisture rap"idly so that the weighing should be made quickly, the
acid having been placed in a special closed receptacle as recommended on
pages 2200 to 2203. Such an amount of acid is taken as will require a titration
that can be read on the chamber burette used. For example if normal caustic
is used and the chamber burette reads from 75 ml. to 100 ml. the acid should
contain from 3.7 grams to 4.9 grams of H 2 S0 4 The acid is added to CO 2 free
distilled water, preferably in a porcelain casserole, 1 m}. of 1 % phenolphthalein
added and the titration made to a faint pink color with standard NaOH. It is
a.dvisable to titrate the acid using methyl orange if the presence of CO 2 is
suspected in the water. Some chemists prefer the use of this indicator. It is a
common practice to standardize the acidimetric and alkalimetric reagents in
presence of methyl orange and phenolphthalein in separate titrations, recording
factors for each.
The temperature of the titrating alkali should be noted and a correction of
0.032 ml. be madwfor each degree deviation from the temperature at which the
reagent was standardized. For each 10 C. that the temperature is above that
at standardization subtract this correction, and add this for each degree the
temperature is below that at standardization.
(ml. titrationXH 2S0 4 value per ml.XIOO)+wt. of sample = % H 2S0 4

DETERMINATION OF IMPURITIES IN SULFURIC ACID

DETERMINATION OF RESIDUE ON HEATING

'

50 to 100 grams of the.acid are weighed into a platinum or silica dish and the
acid evaporated (Hood) by heating gently over a direct flame. 'When no more
fumes evolve, the dish is cooled and weighed. The increased weight of the dish
is due to the residue, composed of substances not volatile at red heat and
generally contains iron oxide. Chamber acid may have all of the substances
occurring in sulfur ores present in this residue.
16 Certain details of procedure are published by courtesy of the General Chemical
Company.
.

2230

ACIDS
DETERMINATION OF LEAn

Minute Amounts.-Determine by means of t:h4 colorimeter according to

the procedure described on page 517, Vol. 1.
High Lead Contamination.-Larger amounts of ~ead sulfate may be determined gravimetrically. The lead may be extracted from the residue, obtained
on evaporation of the acid and heating, by extractirtg with ammonium acetate
according to the standard procedure, and lead precipitated from the extract
made acid with acetic by addition of a small excess of K 2 Cr 207 solution.
If_ preferred, lead may be determined as follows: 100 grams of the acid are
diluted with an equal volume of distilled water and the solution stirred and
cooled. Twice the volume of alcohol is added, the lead sulfate allowed to
settle for two hours or more and the precipitate filtered on a Gooch crucible,
the residue washed with alcohol, dried and ignited at dull red heat and weighed
as PbS0 4. If contamination of the lead sulfate is suspected, this residue may
be extracted with ammonium acetate and lead precipitated from the extract
as PbCr04 according to the regular procedure.
PbCr04 X .641 = Pb.

PbS0 4X .6331 = Pb.

DETERMINATION OF IRON

Traces of lron.-The colorimetric procedure is used, iron being determined

by color comparison of the thiocyanate with standards. See procedure on
page 486, Vol. 1. Five grams of the acid are added to 10 ml. of water in a I?mall
beaker and the solution heated to dissolve any iron in suspension. The cooled
acid is poured into a 100 ml. Nessler tube with rinsing of the beaker. A few
drops of N /10 KMn04 are added in amount sufficient to obtain a faint pink
color, then 10 ml. of a 10% solution of NH 4CNS and the solution made up to
100 ml. . Comparison is now made as follows:
The standard is prepared by treating 5 ml. of iron-free sulfuric acid in
exactly the same manner as the tested sample, i,n a separate Nessler tube.
Standard iron solution is run in, stirring the solution with a glass plunger,
until the color matches that of the sample. It is advisable to make the standard
solution to contain 0.00005 g. Fe per 1 ml. Then each ml. will represent 0.001 %
Fe.
High lron.-If 0.02% of iron or over is present" dilute 25 to 50 grams of the
acid with distilled water. Add a few crystals of KCl0 3 aJ;!d boil. Cool and
add a slight excess of NH 40H. When the iron hydroxide hassettled, filter and
wash with hot water. Dissolve the precipitate with HCI (1 :1) catching the
solution in a casserole. Wash the iron out of the paper. Heat the combined
filtrate to boiling and titrate with standard Snt)12 solution. See page 484, Vol. 1.
If antimony or arsenic is present the solution turns dark during the titration. If this occurs treat a fresh sample, diluted, with H 2S to remove As and
Sb, and determine iron in the filtrate from the sulfide~.
DETERMINATION OF ARSENIC

Traces of Arsenic. Amounts less than .OOOp.s%.-Take 50 grams or more

of the acid and distill off the arsenic as As'Cla-according to the distillation

ACIDS

2231

method given on page 91. Determine arsenic in the distillate by the Cut zeit
method.
.
Amounts less than .005% may be determined directly by the Gutzeit
metnod. When the arsenic is greater than 0.05% it may be determined by
reducing the arsenic with tartaric acid and titrating with standard iodine solution, after neutralizing the acid with ammonia and adding sodium bicarbonate.
25-50 grams of the acid are taken and placed in a Kjeldahl flask and! gram of
tartaric acid added together with 2 grams of fused potassium bisulfate. The
acid is heated over the direct flame until the color, first becoming dark, changes
to a straw color. Fuming will cause loss of arsenic and should be avoided ..
The acid is washed into a beaker with 250 ml. of water, then neutralized with
NH 40H, and NaHC0 3 added. The titration with iodine is made in presence
of starch. The end-point is a blue color.
If antimony is present a separation of the arsenic must be made by distillation. 1 ml. NjlO I=.003748 g. As or .004948 g. As 20 a As 20 a Xl.11616
= As 2 0 .
DETERMINATION OF ANTIMONY

The arsenic is first removed by distillation as AsCla (page 91, Vol. J). The
antimony is now distilled by adding 20 grams of zinc dissolved in 20 grams of
concentrated HCl. The antimony in the distillate is determined by the modified Gutzeit method given in the chapter on Antimony, Vol. I.
DETER1~INATION OF ZINC

25 to 50 ml. of the acid are evaporated to dryness in a silica dish. The

. residue is treated with a strong solution of NH 4 CI and NH 4 0H solution. Mter
filtering off the ferric hydroxide, the filtrate containing the zinc is neutralized
with H 2S0 4 and 10 ml. added in excess (sp.gr. H ZS0 4 = 1.84). Potassium
ferrocyanide is now added in such amount that the solution will contain about
2.5%. Compare the turbidity with standards.
With larger amounts of zinc the zinc is deter:qlined by titrating with standard
potassium ferro cyanide. See chapter on Zinc, Vol. 1.
The zinc may be determined gravimetrically on a 200-gram sample. The
acid is fumed off until only 2-5 ml. remains in the dish. Water is added, the
solution neutralized with ammonia, sufficient dilute H 2S0 4 added to make a
2.5% free acid. Copper, lead, etc., are removed as sulfides by gassing with
H 2S. Zinc is determined in the filtrate, after oxidation of the iron with bromine
and removing it by adding NH 40H and NH 4 Cl. Zinc is precipitated as ZnS
from the filtrate made acid (2% free acid) with formic acid, by H 2S. The
washed sulfide is ignited and weighed as ZnO. ZnO X .8034 =Zn.
Copper.-This is determined colorimetrically. 25-50 ml. of the acid are
evaporated to dryness, the residue taken up with 100 ml. of water, 2 ml. HCI
(sp.gr. 1.19) added and copper precipitated with H 2S. The washed precipitate
is dissolved in HNO s and the solution made ammoniacal with ammonia. The
color IS now matched with standards containing known amounts of copper.

2232

ACIDS
DETERMINATION OF 'SELENIUM I

Traces.-l0 ml. of the acid diluted with an equal volume of water are placed \
in a test tube and 4 or 5 drops of 1 % KI solution added. ,Heat to boiling to
expel free iodine. The brick red color is compared with standards containing
known amounts of selenium made up in the same way as the sample. If
sufficient selenium in present this is thrown down as a brick red finely divided
preci pita te.
Standard selenium solution is prepared by dissolving .1 g. pure Se in 5 ml.
HNO a and 10 ml. HCl. The solution is evaporated to dryn~ss, then taken up
with water and a little dilute H 2S0 4 and made up to 10001)11. 1 m1.=.OOOl g. Se.
Larger Amounts.-500 grams of the. acid is diluted by pouring into a liter
of water, 100 ml. concentrated HCI added and the solution saturated with S02.
The selenium is allowed to settle 15 to 20 hours, then filtered through a filter of
fine texture. The precipitate is dissoh;ed by placing the filter and' precipitate in
sulfuric acid nitric acid mixture 5 ml. H 2S0 4 +50 mL HNO a. The solution
is evaporated to fumes, adding more HN0 3 if the solution is dark, and again
evaporated to expel HNO a. After diluting with water and filtering, the filtrate
is made up to about 100 mL and gassed boiling hot with H 2S. The precipitated
selenium is settled, filtered into a Gooch, washed with water, then alcohol and
ether and dried with a current of warm air to constant weight.
Sulfur Dioxide Determination in Sulfuric Acid.-1-10 ml. of N /10 iodine
solution are run into 200 ml. of water, starch solution is added and the acid to
be tested is allowed to flow from a burette into this reagent, until the blue color
disappears.
1 mL N /10 iodine = .0032 g. S02.
SULFURIC ACID
Hydrochloric Acid Determination in Sulfuric Acid.-50 to 100 ml. of the
acid is diluted to about 200 ml. by running the acid,'into chlorine-free water.
10 ml. of 1% silver nitrate solution is added and 'the ttirbidity compared against
a standard made by adding standard sodium chloride solution from a burette
into a solution containing an equal amount of sulfuric acid (free from chloride)
diluted with water and containing silver nitrate as in case of the sample. When
the turbidity is the same as the sample the nu~ber of ml. of NaCl solution is
noted. The turbidity is best seen by placing the two solutiollil. side by side on
a black paper. Standard 'chloride solution contains .16 g ..NaCl per liter.
1 ml. = .0001 g. HCl.
If considerable hydrochloric acid is present it may be deterrhineq. by Volhard's method given on page 271, Vol. 1. The sample of acid is diluted to about
20% strength. A measured excess of N /10 AgN9a added and the excess titrated with Nj10 NH 4CNS solution in presence o~ ferric ammonium alum.
1 ml. N/lO AgNO a=.003647 g. HCI.
DETERMINATION OF NITRIC ACID

Traces.-In absence of selenium traces {)PlINO a may be determined

colorimetrically. To 100 ml. of the acid! illl. saturated solution of FeS04 is

ACIDS

2233

added. Comparison is made in a Nessler tube with a standard by adding to

100 ml H 2S0 4,! ml. FeS04 solution and running in standard HNO a until the
color matches, the solution being stirred with a plunger during the addition.
Standard HNO a solution should contain .0001 g. HNO a per ml.
Larger Amounts.-See method on page 2216.
Determination of Fluorine.-This ,is determined by the etch test described
in Vol. 'I, page 421. The acid should not contain over 10% H 20. Strengthen
with oleum if necessary.

THE ANALYSES OF OLEUM OR FUMING SULFURIC ACID AND

OF MIXED ACID
The analyses of fuming sulfuric acid and mixed acid are placed under one
general scheme as the procedure for oleum is included in that of mixed acid.
The term oleum is given to concentrated sulfuric acid containing free SOa,

FIG. 309.

Method for Rapid Evaporation of Liquids.

the combined water in the product decreasing (along with sulfuric acid) with
the increase of free 80 a or sulfuric anhydride. Mixed acid is the technical
name for the mixture of concentrated sulfuric acid and nitric acid or of oleum
and nitric anhydride, the product being commonly used in nitrating glycerine,
cotton and other materials. The analysis includes the determinations of
H 2S0 4, HNO a, N 20 5, N 20a and in the case of oleum, the determination of SOa.
In the presence of the latter, HNO a is assumed to lose its combined water
according to the reaction:
2HN03+S0a=H2S04+N20S.
In absence of nitric acid S02 may be present. It is assumed that if HNO a is
present the S02 is oxidized to 80 3 with formation of H 280 4 and the anhydrides
SOa and N 20a according .to the reaction.
.
N205+H20+2S02=N20a+SOa+H2S04.
Some chemists prefer to express the reaction:
2HN0 3 +S0 2 = H 280 4 +N 20 4

2234

ACIDS

The analysis is carried out by three titrations-(a) determination of total

acidity, (b) determination of sulfuric acid including :SOdree in case of oleum,
(c) determination of lower oxides.
\
For economy of time the following order should:be observed: The sample
for the determination of sulfuric acid and fre'e SOa (oleum) !5hould be weighed,
diluted with water according to the detailed procedure and placed on the steam
bath for evaporation. During the evaporation the titration for total acidity of
the sample and the titration for the lower oxides are made and finally that of
oleum on the evaporated sample.
Special Reagents: Normal Sodium Hydroiide.-One mL = 0.04904 gram
H 2 S0 4
Tenth Normal Potassium Perm'anganate.-3.16 grams KMn04 per liter.
Standardized against N /10 solution of Sorensen's Sodium Oxalate. (See Preparation of Standard Reagents.) One mL=0.0019 gram N Z0 3, or 0.002351
gram HNO z.
PROCEDURE.

TOTAL ACIDS

The sample is accurately weighed by one.of the procedures recommended for

strong acids. The Dely or Blay-Burkhard tubes are best for this purpose.
Such a weight being taken as will require a titration between 75 and 100 mI.,
e.g., containing an equivalent of 3.675 grams to 4.9 grams H 2S0 4 The ,acid is
now run under cold CO 2 free water according to directions on pages 2201. and
2202, and the acid titrated with N /1 ~olution of NaOH.
m!. NaOHX.04904X100.
One m!. NaOH=0.04904 g. H 2S04. Calculate to % W' ht f 'd t k
elg 0 aCl a en
LOWER OXIDES
Ten m!. of the sample, weighed in a pipette with, capillary delivery tip, are
cautiously run into about 400 mL of cold water, k~~ping the delivery tip well
under the water and in rapid motion to prevent overheating through local.
action. The mixture is titrated with N /10 KMn04 until a pink tint is obtained
that does not fade in three minutes.
1 ml. N/10 KMn04=0.0019 g. N 2 0 a (N 2 0 a to N 2 0 4 fa~tor=2.42.)
1 ml. N/10 KMn04=0.0046 g. N 20 4
- ---1 ml. Nj10 KMn04=0.00235 g. HNO z. Equivalent H ZS0 4= 0.0049043 g. per mL
1 m!. N/10 KMn04=0.0032035 g. SOz.
1 mL N /10 KMn04 =0.0041043 g. H 2SO a.
NOTE.--With exactly N/10 KMn04 on a 19-9ram slfmple 1 ml. =0.01 % N 20 a.

SULFURIC ACID AND FREE 80 a

The sample is weighed in a Dely tube and run under c9ld water, as in case
of total acids, using in this case about 45 to 50 ml: of water in a 4-in. casserole.
The solution is evaporated on the steam bath to'eipel the volatile acids, lower

ACIDS

2235

oxides and nitric. The evaporation is hastened by blowing a current of hot, dry,
pure air over the sample, see Fig. 309. Instead of a cass'erole, a shallow glass
cell, 3 inches in diameter and 1! inches deep may be used. The air current in
this case is unnecessary. About 5 ml. of water are added and this again
evaporated off. The acid thus obtained is H zS0 4 +SO a.
The acid is taken up with water and titrated with standard caustic, using
CO 2 free water and phenolphthalein indicator.
One ml. Nil NaOH=O.004904 gram H 280 4
NITRIC ACID
This may be obtained from the above determinations according to calculations following. It may be obtained by direct titration with standard ferrous
sulfate, by running a weighed amount of mixed acid into 100 ml. of cold sulfuric
acid and titrating to a yellowish red tint. For the detailed procedure see Nitric
Acid: page 2216.
Calculation of Results. A. Nitric Acid and S02 being Absent.-The total
acidity is calculated to 80 3. Reference is made to the table for Oleum from
which the % H 280 4 and free 80 a are obtained.
Example.-8uppose the total acidity in terms of SOa was found to be 84.2.
The acid contains 86% H 2S0 4 and 14% free 80a.
B. Nitric Acid Absent, S02 is Present in the Mixture. Total SO a.-From
total acidity as 80 a subtract 80 2 X 1.25 (i.e., equivalent 80 a).
Combined Water.-100-(total 80 a+S0 2 ) =H 20.
Combined SOa.-HzO X 4.4445 = 80'3 equivalent or combined SOa.
Free SO a.-Total SOa-combined SOa=free SOa.
Sulfuric Acid.-Combined SOa+H 2 0.
Example.-If S~ was found to be 2% and the total acidity in terms of 80 3
=83.5%.
Total 80 3 = 83.5 - (2 X 1.25) = 81 %.
Water=100-(81+2) =17%.
Report
Combined 80 a = 17 X 4.4445 = 75.56%
8ulfuric acid = 92.56%.
Free SOa = 5.44.
Free 80 a = 81-75.56 = 5.44%.
Sulfuric acid =75.56+17 = 92.56%.
80 2 =2.00. Total 100.
C. Nitric Acid Present and SOz Absent.-Nitric acid in presence of free 80 a
is assumed to be the anhydride N Z05'
N z0 5.-From the total acidity is subtracted the acidity after evaporation,
both being calculated to equivalent 80 a. The difference multiplied by 1.349
=% N Z0 5

HzO.-Total SOa (after evaporation)+N 20 5 subtracted from 100=H 2 0.

Combined SOa.-H 2 0 X 4.4445 = SOa equivalent to H 2 0.
Free SO a.-Combined SOa subtracted from total 80 a=free 80 a.
Sulfuric Acid.-Combined SOa+H zO=H zS0 4
Example.-If total acidity=84% in terms of 80 a and the total 80 a (after
evaporation) =82%, then the difference 2X1.349=2.698% N 20 5
Water = 100-(82+2.698) = 15.302%.
Report

2236

ACIDS

Combined S03 = 15.302 X 4.4445 = 68.01 %.

Free Sb a =82-68.01 = 13.99%.
H 280 4 = 68.01 + 15.302 = 83.312%.

H 2S0 4 =83.312%.
80 3 ;= 13.980%.
N 20 5 =2.698%.
Total 100.

j Free

D. N 2 0ij Required to be Reported as HNO a, 96%.t6_HNO a, 96%=80a

equivalent (acid expelled by evaporation) X 1.64. 100 - % HNO a (96%)
'
I
100 Xtotal SOa
= 0 Ieum. T otaI SO a III 0 eum
t I
per cen 0 eum
Reference to the oleum table will give the per cent free 80 a in the oleum.
Example.-Suppose total acidity in terms of 80 a = 84% and the 80 3 after
"evaporation =82%, then HNO a, 96%=the difference=2X1.64=3.28% HNO a
(96%).
Oleum = 100 - 3.28 = 96.72%.
Total 80 a in oleum=100X82 divided by 96.72=84.78.
From the oleum table 84.78 gives 17.10% free 80 a.
Report

Nitric acid, 96% = 3.28%

Oleum
= 96.72%
=17.10%
Free 80 a
Total80 a
=84.78%
The nitric acid and oleum make a total of 100.
Formulae for Diluting or Strengthening Solutions.-To dilute a solution with
water:
DZ

~a) -y=X;

Y=Z-X;

or

(b)

To dilute a strong with a weaker liquid:

(a)

A-DZ
A-B =Y;

X-Z=Y;

To strengthen a weak with a

(a)

D-AZ
C-A =X;

strong~r

Y=Z-X;

or

A-DX
D

Y.

"

(b)

A-DX
D-B =Y.

(b)

D-AY , ,
C-D

liquid:
or

=X::

A = actual concentration of the solution that is to be corrected;

B = concentration. of the diluting solution;
C=concentration of the strengthening solution;
D=desired concentration;
,
X =amount of the stronger solution to be added, taken or prepared;
Y = amount of weaker solution to be added or taken;
Z = amount of solution desired, or given.
/
All data are in terms of weight of the con~t~q,ents.
16

This is the usual strength of the commercial. concentrated acid.

ACIDS

2237

DILUTION OF SOLUTIONS
If a=per cent of solution desired, b=specific gravity of solution to be
diluted and c its percentage then
Volume or ratio of the strong solu!ion is d
a
- - 1 - =d.
bX
c
100

. Volume or ratio of wate: required .

a
100--- o r 100-bXd.
c+100
Example.-To make 75% sulfuric acid from 87% solution having a specific
gravity of 1.8.
.
Substituting in above formula

7~

1.8X .8

7 = m!. or liters or cubic feet, etc. of

acid to be taken: i.e. ratio is 47.9 and 100 -

0~:7 = m!.

or liters or cubic feet of

water required for dilution, i.e. ratio is 13.79.

47.9 m!. of the strong acid diluted with 13.79 m!. of water will make the
acid 75% strength.
It is evident that the ratio being known large volumes of solution may be
made so that the procedure is appllcable for factory use.

ACIDS

2238

SULFURIC .ACID

By W. C.
SpecIl!c

Degrees
BaualC.

--

Gravity
60

6~O F.

Degrees
Twaddell.

FERGUSON AND

H. P.

-tALBOT

Weight of
Cu. Ft. iu

Per
cent
H,SO.

Lbs. A".

I. Per Cent
O. V

Pounds O. V.
in
x Cu bic Foot.

2
3
4

1.0000
1.0069
1.0140
1.0211
1.0284

0.0
1.4
2.8
4.2
5.7

0.00
1.02
2.08
3.13
4.21

62.37
62.80
63.24
63.69
64.14

0.00
1.09
2.23
3.36
4.52

0.00
0.68
1.41
2.14
2.90

5
6
7
8
9

1.0357
1.0432
1.0507
1.0584
1.0662

7.1
8.6
10.1
11. 7
13.2

5.28
6.37
7.45
8.55
9.66

64.60
65.06
65.53
66.01
66.50

5.67
6.84
7.99
9.17
10.37

3.66
4.45
5.24
6.06
6.89

10
11
12
13
14

1.0741
1.0821

14.8
16.4

10.77
11.89

66.99
67.49

1.0902

18.0

13.01

68.00

1.0985
1.1069

19.7
21.4

14.13
15.25

68.51
69.04

11.56
12.76
13.96
15.16
16.36

7.74
8.61
9.49
10.39
11.30

15
16
17
18
19

1.1154
1.1240
1.1328
1.1417
1.1508

23.1
24.8
26.6
28.3
30.2

16.38
17.53
18.71
19.89
21.07

69.57
70.10
70.65
71.21
71.78

17.58
18.81
20.08
21.34
22.61

12.23
13.19
14.18
15.20
16.23

20
21
22
23
24

1.1600
1.1694
1.1789
1.1885
1.1983

32.0
33.9
35.8
37.7
30.7

22.25
23.43
24.61
25.81
27.03

72.35
72.94
73.53
74.13
71'1. 74

23.87
25.14"
26.41
27.69
2\).00

17.27
18.34.
19.42
20.53
21.68

0
1

/'

Sp. Gr. determinations were made at GOo F., compared with water at 60 F.
From the Sp. Grs., the corresponding degrees Baume were calculated by
the following formula: Baume = 145-145/Sp. Gr.
Baume Hydrometers for use with this table must be. graduated by the
above formula, which !ormula should always be printed
the _scale.
* 66 Baume = Sp. Gr. 1.8354 = Oil of Vitriol (0. V.).
1 cu. ft. wat~r at 60 F. weighs 62.37 lbs. avo Atomic weights from F. W. Clarke's table of 1901. 0 = 16.
H 2S0 4 = 100 per 'cent.
% H 2S04
% O. V. I %600
O. V. = 93.19
100.00 =d 119.98
60 = 77.67
83.35 ~ 100.00
50 = 62.18
66.72 = 80.06

on

ACIDS

2239

SULFURIC ACID (Continued)

-Freezing
Degrees (Melting)

Baume.

~oint.

F.

--0
1
2
3
4

32.0
31.2
30.5
29.8
28.9

5
6
7
8
9

28.1
27.2
26 3
25.1
24.0

10
11
12
13
14

22.8
21.5
20 0
18.3
16.6

15
16
17
18
19

14.7
12.6
10.2
7.7
4.8

20
21
22
23
24

1.6
1.8
- 6.0
-11
-1.6

APPROXIMATE BOILING POINTS

50
60
61
62
63
64
65
66

B,
"
"
"
"
"
"
"

295 0
386
400 0
4150
432
451
4850
538

F.
"
"
"

"
"
"

"

FIXED POINTS
SpecifiC
Gravity.

1.0000
1.0048
1.0347
1.0649
1.0992
1.1353
1.1736
1.2105
1.2513
1.2951
. 1.3441
1.3947
1.4307
1:4667
1.4822

Per Cent
H.SO .

Specific
Gravity.

Per Cent
H,SO

.00
.71
5.14
9.48
14.22
19.04
23.94
28.55
33.49
38.64
44.15
49.52
53.17
56.68
58.14

1.5281
1.5440
1.5748
1.6272
1.6679
1.7044
1.7258
1.7472
1.7700
1.7959
1.8117
1.8194
1.8275
1.8354

62.34
63.79
66.51
71.00
74.46
77.54
79.40
81.32
83.47
86.36
88.53
89.75
91.32
93.19

Acids stronger than 66 Be. should have their percentage compositions

determined by chemical analysis.
* Calculated from Pickering's results, Jour. of Lon. Ch. Soc., vol. 57, p. 363.
AUTHORITIES -

W.

c.

FERGUSON; H.

P.

TALBOT.

This table has been approved and adopted as a standard by the Manufacturing Chemists' Association of the United States.
W. H. BOWER,
HENRY HOWARD,
JAS. L. MORGAN,
ARTHUR WYMAN,
A. G. ROSENGARTEN,

New York, June 23, 1904.

Jiucuti've Committe4.

2240

ACIDS
SULFURIC ACID (Continued)

Degrees
Baume.

Specific
Gravity
600
60 0F .

Degrees
Twaddell.

Per
Cent
H,SO,.

Weight of
[Cu. Ft. in
Lbs. Av.

,I

IPoundsin O. V.

Per Cent
O. V.

.[ Cubic Foot.

-25
26
27
28
29

1.2083
1.2185
1 .. 2288
1.2393
1.2500

41. 7
43.7
45.8
47.9
50.0

28.28
29.53
30.79
32.05
3;1.33

75.33
76.00
76.64
77.30
77.96

: 30.34
31.69
33.04
34.39
35.76

22.87
24.08
25.32
26.58
27.88

30
31
32
33
34

1.2609
1.2719
1.2832
1.2946
1.3063

52.2
54.4
56.6
58.9
61.3

34.63
35.93
37.26
38.58
3J.92

78.64
79.33
80.03
80.74
81.47

37.16
38.55
39.98
41.40
42.83

29.22
30.58
32.00
33.42
3L90

35
36
37
38
39

1.3182
1.3303
1.3426
1.3551
1.3679

63.6
66.1
68.5
71.0
73.6

41.27
42.63
43.99
45.35
46.72

82.22
82.97
83.74
84.52
85.32

44.28
45.74
47.20
48.66
50.13

36.41
37.95
39.53
41.13
42.77

40
41
42
43
44

1.3810
1.3942
1.4078
1.4216
1.4356

76.2
78.8

86.13
86.96
87.80
88.67
89.54

51.61
53.08

44.45
46.16

54.58

47.92

84.3
87.1

48.10
49.47
50.87
52.26
53.66

56.07
57.58

49.72
51.56

45
46
47
48
49

1.4500
1.4646
1.4796
1.4948
1.5104

0.0
92.9
95.9
99.0
102.1

55.07
56.48
57.90
59.32
60.75

90.44
91.35
92.28
93.23
94.20

59.09
60.60
62.13
63.65
65.18

53.44
55.36
57.33
59.34
61.40

50
51
52
53
54

1.5263
1.5426
1.5591
1.5761
1.5934

105 3
J08.5
111.8
115.2
118.7

62.18
63.66
65.13
66.63
68.13

95 .. 20
96.21
97,24
98.30
99.38

66.72
68.31
69.89
71.50
73"':"1"1

63.52
65.72
67.96
70.28
72.66

55
56
57
58
59 -

1.6111
1.6292
1.6477
1.6667
1.6860

122.2
125.8
129 ..5
133.3
137.2

69.65
71.17
72.75
74.36
75.99_

100.48
101.61
102.77
103.95
105.16 I

74.74
76.37
78.07
79.79
81.54

81.6

75.10
.77.60
80.23
'82.95
~5.75

2241

ACIDS
SULFURIC ACID (Continued)
Freezing
Degrees "(Melting)
B"ume.
Point.

-25
26
27
28
29

ALLOWANCE FOR TEMPERATURE

-23
-30
-39
-49
-61

At 10 Be .. 029 Be. or
" 20 " .036 .""
" 30 " .035
" 40 " .031
"
" 50 " .028
"
" 60 " .026
"
"
" 63 II" .026
.0235
," 66
"

.00023 Sp. Or.

.00034 "
.00039 "
.00041
"
.00045 "
.00053
"
.00057
"
.00054 "

= 1F.
= 1"
= 1"

=
=
=
=

-74
-82
-96
-97
-91

35
36
37
38
39

-81
-70
-60
-53
-47

Per Cent

40
41
42
43
44

-41
-35
-31
-27
-23

61.93
63.69
65.50
67.28
69.09

53.34
55.39
57.50
59.66
61.86

77.36
79.56
81.81
84.05
86.30

66.63
69.19
71.83
74.53
77.27

45
46
47
48
49

-20
-14
-15
-18
-22

70.90
72.72
74.55
76.37
78.22

64.12
66.43
68.79
71.20
73.68

88.56
90.83
93.12
95.40
97.70

80.10

50
51
52
53

-27
-33
-39
-49
-59

80.06
81.96
83.86
85.79
87.72

76.21
78.85
81.54
84.33
87.17

100.00
102.38
104.74
107.15
109.57

95:20
98.50
101.85
105.33
108.89

89.67
91.63
93.67
95.74
97.84

90.10
93.11
96.26
99.52
102.89

112.0l
114.46
117.00
119.59
122.21

112.55
116.30
120.24
124.31
128.52

55
56
57
58
59

~.

..
-

"

60

Baume.

Pounds
60v Bauma
in
I Cubic Foot.

Per Cent

Pounds
Baume
in
Cubic Foot.

500

50

Baume.

----.

}.

'"

1 "
1"
1"
1"
= 1"

30
31
32
33
34

54

of.

---

82.98
85.93
88.94
92.03

2242

ACIDS
SULFURIC ACID (Continued)
,

Baume.

Specific
Gravity
60 0 F.
60 0

Degrees
Twaddell.

Per Cent
H,SO .

Weight of
1 Cu. Ft. in
Lb. Av.

60
61
62
63
64

1.7059
1. 7262
1.7470
1.7683
1.7901

141.2
145.2
149.4
153.7
158.0

77.67
79.43
81.30
83.34
85.66

106.40
107.66
108.96
110.29
111.65

64i
64!
641
65
65!

1.7957
1.8012
1.8068
1.8125
1.8182

159.1
160.2
161.4
162.5
163.6

86.33
87.04
87.81
88.65
89.55

112.00
112.34
112.69
113.05
'113.40

92.64
93.40
94.23
95.13
96.10

103.75
104.93
106.19
107.51
108.97

65!
651
66

1.8239
1.8297
1.8354

164.8
165.9
167.1

90.60
91.80
03.19

113.76
114.12
'114.47

97.22
98.51
100.00

110.60
112.42
114.47

Degrees

Freezing
(Melting)
Point.

Per Cent
60 0 Baume.

Pound.

Baume.

GOO Baume in

Per Cent
50 0 13aumB.

Pounds
50 0 Baumein
Cubic FOQt.

60
61
62
63
64

+12.6
27.3
39.1
46.1
46.4

100.00
102.27
104:67
107.30
110.29

106.40
110.10
114.05
118.34
123.14

124.91
127.74
130.75
134.03
137.76

132.91
137.52
142.47
147.82
153.81

64!
64!
641
65
651

43.6
41.1
37.9
33.1
24.6

111.15
112.06
113.05
114.14
115.30

124.49
125.89
127.40
129.03
130.75

138.84
139.98
141. 22 I
142.57 1
144.02

155.50
157.25
159.14
161.17
163.32

65!
65!
66

13.4
.- 1
-29

116.65
118.19
119.98

132.70
134.88
137.34

145.71
1:i7.63
149.87

165.76
168.48

Degrees

Cubic Foot.

Per Cent
O. V.

Pounds O. V.
in
1 Cubic Foot.

83.35

88.68
91.76
95.06
98.63
102.63

, 85.23

87.24
89.43
91.92

~171.56

XI.-SULPHURIC ACID TABLE

94-100% H 2S0 4
By H. B. BISHOP

Be.

Sp. Gr. at 60 0 F.

Per Cent.
H2S0.

66
66.12
66.23
66.31
66.36
66.36
66.36
66.30
66.16

1.8354
1.8381
1.8407
1.8427
1.8437
1.8439
1.8437
1.8424
1.8391

93.19
94.00
95.00
96.00
97.00
97.50
98.00
99.00
100,00

Wt. 1 Cu. Ft.

114.47
114.64
114.80
114.93
114.99
114.99114.99
114.91
114,70
/'

Allowance for Temperature.

At 94% .00054 sp.gr. = 1 F.

" 96 .0053 ' , = 1 F.
" 97.5 .00052 " =1 F.
" 100 1:00052 " =1F.
0

_'-

r'-

-'

ACIDS

2243

FUMING SULFURIC ACID EQUIVALENTS

Total
SO.

Equivalent
H.BO.

81.63
81.82
82.00
82.18
82.37
82.55
82.73
82.92
83.10
83.29
83.47
83.65
83.84
84.02
84.20
84.39
84.57
84.75
84.94
85.12
85.31
85.49
85.67
85.86
86.04
86.22
86.41
86.59
86.78
86.96
87.14
87.33
87.51
87.69
87.88
88.06
88.24
88.43
88,61
88,80
88.98
89.16
89.35
89.53
89.71
89.90
90.08
90.27
90.45
90.63

. 100.00
100.23
100.45
100.67
100.90
101.13
101.35
101.58
101.80
102.03
102.25
102.47
102.70
102.92
103.15
103.28
103.60
103.82
]0405
104.27
104.50
104,73
104.95
105,18
105.40
105.62
105 85
106.07
106.30
106.53
106.75
106.98
107.20
107.42
107.65
107.87
108.10
108.33
108,55
108.78
109.00
109.22
109.45
109.67
109.90
110.13
110.35
110.58
110.80
111.02

Per Cent
H,SO.

100
99
98
97
96
95
94
93
92
91
90
89
88 '
87
86
85
84
83
82
81
80
79
78
77
76
75
74
73
72

71
70
69
68
67
66
65
64
63
62.
61
60
59
58
57
56
55
54
53
52
51

PerCent
Free SO.

0
1
2

3
4
5
6
7
8
9
1J.
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26'
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

Total
SO.

90.82
91.00
91.18
91.37
91.55
91.73
91.92
92.10
92.29
92.47
92.65
92.84
93.02
93,20
93.39
93.57
93.76
93.94
94.12
94.31
94.49
04.67
04.86
95.04
95.22
95.41
95.59
95.78
95.96
96.14
06.33
96.51
96.69
96.88
97.06
97.25
97.43
97,61
97.80
97,98
98.16
98.35
98.53
98.71
98.90
99.08
99.27
99.45
99.63
99.82
100,00

Equivalent
H,SO.

111.25
111.48
111.70
111.93
112.15
112.37
112.60
112.82
113.05
113.28
113.50
113.73
113.95
114.17
114.40
114.62
114,85
115.08
115.30
115.53
115.75
115,97
116.20
116.42
116.65
116.88
117.10
117.33
117.55
117,77
118.00
118.22
118.45
118.68
118,90
119.13
119.35
119,57
119.80
120.03
120.25
120.48
120.70
120.92
121.15
121.37
121.60
121.83
122.05
122.28
122.50

Per Cent
H.BO.

50
49
48
47
46
45
44
43
42
41
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11

10
9
8
7
6
5
4
3
2
1
0

I
,

\
\

Per Cent
Free SO.

50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100

Compiled from the table by H. B. Bishop, Van Nostrand's Chemical Annual, 1913;

2244

ACIDS
METHOD OF ANALYSIS OF

CHLORO~ULFO'NIC

ACID

Chlorosulfonic acid, 80 a HCl, decomposes to H 280 4 and HCI on addition

of water, the reaction being violent. Considerable ~are must be exercised to
prevent loss of acid during dilution with water for examination of the product.
The following method of analysis has been found satisfactory:
Total Acidity.-Three to four grams of the chlorosulfonic acid are weighed
in a Deli tube or small glass bulb. About 25 ml. of distilled (neutral) water and
about 10 ml. less NaOH (normal strength) than is necessary to neutralize the
.
weight sample -10
.
.
sample (I.e.,
0.04001
=ml. NaOH to be taken) are placed III a heavy
wall glass bottle (250-300 ml. capacity). If the sample is weighed in a Deli tube
it is run into the NaOH solution according to the procedure described on page
506. If the bulb is used, the bottle, with the sample inserted, is stoppered, .
wrapped in a towel and shaken vigorously until the bulb breaks and th~ acid
mixes with the water and NaOH. The excess of acid is now titrated with N
NaOH, using phenolphthalein or methyl red indicator. The total acidity is
calculated to 80 a and recorded as % 80 a.
Titration of Chloride.-The NaCI formed is titrated with N /3 AgNO a solution, using K 2Cr04 indicator. The ml. of the reagent are calculated to the
equivalent HCl. 8ince N aOH is apt to contain N aCI, the. blank is subtracted to
obtain the true HCI equivalent in the sample. The % HQI is calculated.
The Composition of the Acid is now determined as follows:
(a) Total acid as % 80 a= a.
(b) HCl obtained by titration of the neutralized solution (made faintly acid)
with N /3 AgNO a (the HCI blank for the N aOH used in (a) having been subtracted) =b. The HCI is converted to its equivalent 80 a by multiplying by
1.0978=b'.
(c) 80 a (combined and free). The 80 a equivalent of HCI obtained in b is
subtracted from the total acidity as 80 a of (a) is a...cq'=80 a total.
(d) 100 - (% 803+% HCI) = % H 2 0 in sample: This is combined with a
protion of the 80 a as H 280 4. Calculate to % ,H2804 by multiplying by 5.4444.
Allowance must be made for impurities if pref;1entY
(e) The 80 a combined with H 20 is subtracted from the total 80 a of (c):
The result is the 80 a of the chlorosulfonic acid and free~Oa (if any).
By inspection it is possible to ascertain whether the product. contains free
80 3 or free HCl since 80 aHCI are in the proportion 31.29% HCI and 68.71%
80 a, i.e., HCI : 80 a:: 1 : 2.2.
If 80 3 is in excess. The HCI is calculated to 80 3 HOI by multiplying by
3.1956, the result is the per cent chlorosulfonic acid in the sample.
Free 80 3 is obtained by subtracting % 80 3 H,CI+% H 280 4 from 100.
If Hel is in excess. The % 80 3 obtained in' (e) is multiplied by 1.4555;
the result is the per cent chlorosulfonic acid.
Free Hel is obtained by subtracting % 80 aHCl+% H 280 4 from 100.
Results are reported as % 80 3 HCI, H 280 4 , free 80 3 or free HCI.
17

(lOG-impurities non-titratable)-(%SOa+ o/oH.e1i= %H 20.

ACIDS

2245

,Factors:
HCIX1.0978=SOa, HCIX3.1956=SOa HCland HCI X2.1959 = SOainSOaHCl.
SOaXO.8998=HCI, SOaX 1.4555=SOaHCI and SOa XO.4554=HClin SOaHC!.
H 20 X 5.4444 = H 2 S0 4 , H 3S0 4 XO.1837 = H 20, H 2S0 4'XO.7436=HCl.
NaCIXO.6238=HCI, S03HCI=31.29% HCI and 68.71% SOa.

VOLUMETRIC ESTIMATION OF FREE ACID IN PRESENCE OF

IRON SALTS
The red precipitate formed ;hen solutions containing iron are titrated with
caustic makes it difficult to detect the end-point of neutralization; the method
suggested by C. A. Ahlum 18 takes advantage of the white compound formed by
precipitating the iron as a phosphate and the fact that monosodium phosphate
is neutral to methyl-orange indicator.
Reactions
Fe2(S04)s+2NaH 2P0 4 +X free acid=2FeP0 4 +Na2S04+x free acid+2H 2S0 4 ,
or 2FeCla + 2NaH 2 PO 4+x free acid = 2FePO 4 + 2N aCI +x free acid +4HCl.
The acid equivalent to ferric iron is deducted from the total acid found, the
(\xcess acid being due to the free ~cid in the solution.
Procedure.-To the solution c~mtaining the iron and free acid is added a,p.
excess of c.p. solution of monosodium phosphate (neutral to methyl-orange),
and then a few drops of the indicator. The acidity of the solution is now determined by titration with standard caustic in the usual way, the solution being
cold. From this titration the total free and combined acid are calculated.
Iron is now determined in a separate portion by titration with stannous
chloride or dichromate.
'
Calculation.

(A)

Fe 2 0a X 1.2285 = g. H 2S0 4 (combined).

/
Fe 20aXO.9135=g. HCI (combined).
One ml. N/5 NaOH=.00981 gram H 2S0 4 or .00729 gram HOI.
Total acid (B) minus combined acid (A) =free acid.

(B)

DETERMINATION OF BORIC ACID IN BORATES

The details of procedure for the determination of boric acid are given in the
chapter on Boron in Volume 1. The method takes advantage of. the fact that
boric acid is a weak acid having an end-point in an alkalimetric titration of
pH 10-11 hence cannot be titrated in presence of indicators having a lower pH
range, such as methyl red or methyl orange, pH 3.1-6.3. It is possible, therefore, to liberate boric acid from its combination by means of a mineral acid,
titrate or neutraliz.e the mineral acid with NaOH in presence of methyl red or
18

C. A. Ahlum, The Analyst, 31, 168 (1906).

2246

ACIDS

methyl orange and then titrate boric acid in- presence of phenolphthalein indicator after the addition of a polyhydric alcohol. such as 'glycerol or mannitol.
The borate is first converted to a soluble sodiu~ salt by transposition by
means of sodium carbonate and extraction w~th water pr the boric acid is set
free by boiling with a mineral acid, using a reflux condenser to prevent loss of
boric acid, which is volatile with steam. If sodium carbonate treatment is
used the CO 2 must be expelled by acidifying and heating fot: a few minutes in
a covered beaker avoiding vigorous boiling. The free mineral acid is now
neutralized with NaOH in presence of methyl orange or methyl red indicator.
Neutral glycerol is added in sufficient quantity (25 ml. is sufficient for 0.5 g.
B 2 0 s) and again the solution is neutralized if an acidity is evident from the
glycerol addition (commercial glycerol is apt to contain free acid). Phenolphthalein indicator is added and the boric acid titrated with standard NaOH
until a reddish pink color of the combined indicators is obtained.
The solution titrated and the reagents used must be free from CO 2, iron,
aluminum and products other than boric acid, that would react with the alkali
in the titration.
. .
1 ml. of 1 N NaOH is equivalent to 0.03482 g. B 20 a

ORGANIC ACIDS
Titrations of organic acids are made preferably with phenolphthalein.
Methyl orange cannot be used. The wate:.: used for diluting the organic acid
must be free of carbon dioxide.
ANALYSIS OF FORMIC ACID
Formic acid occurs in solutions of varying stre~gth, i.e., 30, 50, 75, 90%,
etc. The impurities that are frequently present are sulfuric and hydrochloric
acid. If formic acid alone is present it may be determined by direct titration.
1 ml. N/IO NaOH=0.004602 gram formic acid In presence of other acids
the following method is recommended.
Procedure.-About 10 grams of the sample weighed in a tared bottle
(Blay-Burkhard form is convenient) is run into about 200ml"o( distilled water
in a graduated flask of 500 ml. capacity and made to volume. 50 ml. of this
diluted sample pipetted into an Erlenmeyer flask is made alkaline with Na 2C0 3
solution. After warming a measured excess o( N /10 KMn04 is added. Formic
acid is oxidized to H 20+C0 2 and a precipitate of Mn02 is thrown down.
10 ml. of dilute H 2S04 are added and a measured/volume of N /10 oxalic acid is
added until all of the precipitate has dissolved 'and the permanganate color
has disappeared. The excess of oxalic acid is now titrated with N /10 permanganate reagent. From this the ml. KMn04 required by the formic acid
is obtained.
1 m!. N/IO KMn04=0.002301 gramformi~ acid.
--'The equivalent m}. values of the KMnO~ and H 2C 20 4 should be obtained
by titration.

ACIDS

'2247

Sulfuric Acid.-This is conveniently determined on about a 20 gram sample

diluted to 200 mI. by precipitation with BaCl 2 according to the standard
procedure.
BaS04 X 0.4202 = H 2S0 4
Hydrochloric Acid.-This may be determined on a 20 gram sample by
precipitation with AgNO. in presence of 5 mI. conc. HNO., the sample having
been diluted to about 200 mI.
.
AgCI XO.2545 = HCI.
ANALYSIS OF ACETIC ANHYDRIDE-ACIDITY METHOD
Acetic anhydride

g~:gg>o

is a volatile, colorless liquid, possessing a

characteristic sharp penetrating odor. The vapor is very irritating to the

nose and eyes. The pure anhydride boils at 1370 C. Mixed with water strong
acetic anhydride settles out into a distinct layer very similar to carbon disulfide .. It hydrolyzes slowly forming a solution of acetic acid. When below
50% in strength (the diluting liquid being acetic acid) the product mixes
readily with water at ordinary temperatures (20 0 C.). Distinct separation
from water takes place when the strength of the anhydride is over 55%, the
product being added to the water dropwise. The separation becomes more
decided with increase of the percentage of anhydride.
The analysis consists in titrating the acetic acid formed by the hydrolysis
of the anhydride. Since the product is volatile it is weighed in stoppered
bottles. Low results are obtained if sufficient time is not allowed for complete
hydrolysis.
.
Procedure.-A sample of approximately 5 to 5.5 grams is weighed in a
bottle 6 to 8 mI. in capacity. A small 2-d ram apothecaries' bottle with No. 00
rubber stopper is satisfactory. The stopper is fitted loosely and the bottle
containing the sample is immersed in 100 mI. of normal NaOH and 300 mI. of
distilled water in a 500-ml: " salt mouth" bottle with rubber stopper. The
bottle is closed tightly and shaken to free the small rubber stopper from the
weighing bottle, and permit the sample to mix with the reagent. After
hydrolyzing for at least one hour the sample is titrated with N /5 H 2S0 4 to
determine the excess of N /1 NaOH, using phenolphthalein indicator.
Calculation.
mI. NaOHXO.06003 X100 =% CH.COOH
wt. sample
(% CH.COOH-100)X5.665=% (CH.CO)zO.
Example.-If the sample titrated 101% CH.COOH then
(101-100) X5.665 = 5.665% (CH.CO)zO.
NOTEs.-lOO per cent (CH.CO).O=117.65% CH.COOH.
.
A-100
Per cent Anhydnde=117.65_100XlOO
= (A -100)5.665.

A = %CHaCOQH by titration.
1I3S0.X 1.2241 = CH.COOH.

2248

ACIDS

DETERMINATION OF ACETIC ANBYDRIDE-ANILINE METHOD

Procedure._:__The analysis should, be carried oue in duplicate runs, each
portion being determined as follows: Two ml. of anhydride are run from a
burette into a tared 5 ml. weighing bottle and carefully weighed, the bottle
being stoppered. The sample is now placed in an Erlenmeyer flask containing
50 ml. N NaOH and 50 ml. distilled water, the stdpper of the bottle being
loosened so that the anhydride mixes with the alkali. After forty mruutes
with occasional shaking of the flask to assist the mixing, the residual NaOH
is titrated with N H 2S0 4 , in prese'nce of phenolphthalein indicator, 2 or 3 ml.
of acid added in excess. After standing fifteen minutes longer the excess acid
is titrated back with N /10 NaOH. From the amount of N NaOH required
by the anhydride calculate the amount of N NaOH required by 100 grams of
the sample. Record this as " A."
Twenty ml. of recently redistilled, perfectly dry aniline are run into a 50
ml.. tared weighing bottle and 2 ml. of the anhydride added from a burette.
The sample is added slowly, swirling the aniline to get an equal distribution.
When the mixture has cooled to room temperature the weight of the anhydride
is determined (the aniline and bottle having previously been weighed). After
an hour the mixture is transferred to a 500 m}. volumetric flask and the solution
diluted to mark with neutral alcohol and distilled water (1 : 1). Fifty ml. of
this solution are titrated with N /10 NaOH. From the number of m!. required
calculate the ml. N NaOH corresponding to the residual acetic acid from 100
grams of the sample. Record this as " B."
Then A-B corresponds to one half the anhydride in 100 grams of the sample
This value multiplied by 0.10207=per cent acetic anhydride.
0.10207 = molecular weight of acetic anhydride divided by 1000.

Correction for Mineral Acids if Present.-In weighed portions (10 grams)

determine hydrochloric acid by precipitation with AgNO a and sulfuric acid
by precipitation with BaCl z in the usual way, the'samples having hydrolyzed
in 100 ml. portions of water. From the percentages of mineral acids present
calculate to ml. normal equivalents and deduct from the titrations above, or
calculate the anhydride equivalent and make deduction,
1 ml. N H 2S0 4 = .04904 g. 1 ml. N HCI = .03647 g. 1 ml. N CHaCOOH
= .06003 g. CHaCOOH X_0.85 = (CH aC0)20,
H 2S0 4 X 1.0404 ~ (CHaCO)zO,
HCI X L3889 = (CHllCO)zO.
ANALYSIS OF

ACET~C

ACID

The acidity of acetic acid may be determined by titration with standard

caustic, using phenolphthalein as indicator,I,
About 4 to 5 grams of glacial acetic acid or a corresponding,amount of dilute
acid are taken for analysis, being weighed out in a'weighing bottle or other suitable container used for strong and weak acids. The acid is mixed with about
250 ml. of water and titrated in the presence of phenolphthllJein indicator with
normal caustic.
-' _:_ ,
One ml. Nil NaOH=0.06005 gram- CHaCOOH.

ACIDS

2249

IMPURITIES IN ACETIC ACID

The more important impurities that are looked for in commercial acetic acid
are formic acid, furfurol, acetoJ;le, sulfmic acid, sulfurous acid, hydrochloric
acid, metals.
In the examination of the acid the physical appearance-turbidity and color
are noted.
FORMIC ACID IN ACETIC

Qualitative.-Ten ml. of t,he acid (glacial diluted 1 : 10) are heated with 1
gram of sodium acetate and 5 ml. of 5% mercuric chloride solution. A turbidity ~
indicates formic acid.
Quantitative.-Five grams of glacial acetic acid or corresponding quantity of
dilute acid are treated with 5 grams of sodium acetate and 40 ml. of mercuric
chloride solution (5%) and 30 ml. of water added. The mixture is heated for
two hours in a flask with a return condenser, the flask being surrounded by
steam.' The precipitated mercurous chloride, HgCI, is filtered off, dried and
weighed.
.
Weight of HgC1XO.0977 = formic acid equivalent.
FURFUROL IN ACETIC ACID

Qualitative.-Aniline dissolved in pure glacial acetic acid (5 ml. aniline in

2 m!. glacial acetic acid) and added to 100 m!. of the sample will produce a red
color in presence of. furfurol.
Quantitative.-The ttJst may Be made quantitative by comparing the color
produced with standard solutions containing known amounts of furfurol. One'
gram of redistilled furfurol is dissolved in 100 ml. of 95% alcohol. 1 ml. of
this solution is diluted to 100 ml. with 95% alcohol. 1 ml. =0.0001 gram of
the reagent.
Test for furfurol in vinegar. Fifty ml. of the vinegar is neutralized with
sodium hydroxide, and 15 to 20 ml. are distilled. Two ml. of colorless aniline
and 15 ml. of hydrochloric acid (1 : 12) added. The mixture is warmed to
about 15 C. for a few minutes and the color compared with standards prepared
in the same way.
GRA VIIIfETRIC METHOD WITH PHLOROGLUCID

Place a qjlantity of the material, chosen so that the weight of phloroglucid

obtained shall not exceed 0.300 gram, in a flask, together with 100 ml. of 12%
hydrochloric acid (specific gravity, 1.06), and several pieces of recently lle'ated
pumIce' stone. Place the flask on a wire gauze, connect with a condenser,
and heat, rather gently at first, and so regUlate as to distill over 30 ml. in about
ten minutes, the distillate passing through a small filter paper. Replace the
30 ml. driven over by a like quantity of the dilute acid added by means of a
separatory funnel in such a manner as to wash down the particles adhering to
the sides of the flask, and continue the process until the distillate amounts to
360 ml. To the completed distillate gradually add a quantity of phloroglucol
(purified if necessary) dissolved in 12% hydrochloric acid and thoroughly stir

2250

ACIDS

the resulting mixture. The amount of phloroglucol used should be about

double that of the furfural expected. The solutio!). first turns yellow, then
. green, and very soon an amorphous greeniAh precipitate appears, which grows
rapidly darker, till it finally becomes almost black. \ Make the solution up to
400 ml. with 12% hydrochloric acid, and allow to stand overnight.
Filter the amorphous black precipitate into a tare~ Gooch crucible through
an asbestos felt, wash carefully with 150 ml. of water in such a way th~t the
water is not entirely removed from the crucible until the very last, then dry for
four hours at the temperature of boiling water, cool and weigh, in a weighing
bottle, the increase in weight being reckoned as phloroglucid. To calculate the
furfural, pentose, or pentosan from the phloroglucid, use the following formulas
given by Krober:
(a) For weight of phloroglucid " a" under 0.03 gram.

Furfural = (a +0.0052) X 0.5170.

Pentoses = (a+0.0052) X 1.0170.
Pentosans = (a+0.0052) XO.8949.
(b) For weight of phloroglucid " a" over 0.300 gram.
Furfural = (a%0.0052) XO.5180.
Pentoses = (a+0.0052) X 1.0026.
Pentosans = (a +0.0(52) X 0.8824.
For weight of phloroglucid " a " from 0.03 to 0.300 gram use Krober's table
of the following formulas: .
Furfural = (a+0.0052) XO.51&5.
Pentoses = (a+0.0052) X 1.0075.
Pi:mtosans = (a+0.0052) XO.8866.
The phloroglucol is purified by recrystallization from hydrochlorib acid.
For details of the procedure see Bulletin 107, U. S. Dept. of Agriculture, Bureau
of Chemistry. (1912, .page 54.)
/
ACETONE IN ACETIC ACID

Fifteen grams of glacial acetic acid, or a corresponding amount of weak acid,

is treated.with 70 ml. potassium hydroxide (10% solution), or sufficient caustic
to make the solution slightly alkaline. The solution is eooled and 25 m!. N /5
iodine solution added and sufficient hydrochloric acid to make the mixture
faintly acid. The excess of iodine is titrated with N /5 sodium thiosulfate, using
starch indicator. The total iodine solution taken minus the equivalent ml. of
thiosulfate = the iodine combined with the acetone, then the weight of iodine in
grams multiplied by 0.07612 = grams acetone in the sample.
r

SULFURIC ACID IN

ACETIC~ ACID

This is best determined by the turbidity test. About 5 m!. of the sample are
taken and 1 drop of hydrochloric acid and half a m!. of 10/% barium chloride.
The turbidity is now compared with a standard pure acetic acid solution containing a known quantity of BaS04, the standar<1being added to a comparison

ACIDS

2251

cylinder until the turbidity is the same as that of the sample, which has been
diluted to a convenient volume in a Nessler tube or similar comparison cylinder.
The apparatus used in determining small amounts of titanium, lead, etc., is
suitable for this test. In this case the glowing wire or filament of an incandescent light is viewed through the solutions, the brightness of the wire acting as a
guide in matching the solutions.
SULFUROUS ACID IN ACETIC ACID

This is best detected by placing in a small flask about 20 m!. of the sample,
adding 5 m!. of strong hydrochloric acid and about 3 grams of zinc and covering
with a filter paper saturated with lead acetate. The blackening of the paper
indicates S02 in the sample (e.g., reduced to H 2S by the hydrogen generated by
the zinc).
The sulfurous acid is best titrated with N /10 iodine solution, using starch
indicator. 1 ml. N /10 I2 = .0032 gr. S02.
HYDROCHLORIC ACID IN ACETIC ACID

Determined by the turbidity test as in case of sulfuric acid, silver nitrate

solution being used to precipitate AgCI, and nitric acid substituted for hydrochloric acid.
METALS IN ACETIC ACID

Total Solids.-Ten to 100 grams of the acid is evapo~ated to dryness in a

platinum dish. The residue contains the non-volatile solids.
ACETATES
Two to 5 grams of the material is placed in a KjeJdahl flask connected by
means of a condenser to a receiving flask containing half normal caustic. About
20 m!. of 85% phosphoric acid are added and about 150 m!. of water. Gentle
heat is applied and gradually increased. About 100 m!. of the solution is
distilled into the caustic. Additional hot water is added to the residue in the
Kjeldahl flask and the distillation continued. This is repeated until about 800
m!. of solution has been distilled over. The CO 2 is boiled out of the distillate, a
reflux condenser being used to prevent loss of the acetic acid. If the solution
is alkaline, a known amount of acid is added and the CO 2 boiled out. The
excess acid is now titrated and the amount of acetic acid in the distillate
calculated.

One ml. N/2 NaOH=O.030015 gram CHaCOOH

ClfaCOOH X 1.3169 = Ca(CH aC0 2)2, or X 1.3663 = CH 2 COONa.
Acetates of the Alkalies and Alkaline Earths.-In absence of other organic
acids, nitrates, etc., a quick method is suggested by Sutton (Vol. Analy., X. Ed.,
p. 91). The salts are converted into carbonates by ignition and the residue
titrated with normal acid.
One m!. N /1 acid = 0.06003 gram CH 3COOH.

ACIDS

2252

ACETIC ACID AT 15
OUDEMANS

Sllecific
Gravity.

1:16'
utI[

Specific
Gravity.

"u

1 0007
1.0022
1.0037
1 0052
1.0067
1.0083
1.0098
l.01l3
1.0127
1.0142
1.0157
1.0171
1.0185
1.0200
1.0214
1.0228
1.0242
1.0256
1.0270
1.0284
1.0298
1.0311
1.0324
1.0337
1.0350

0
1
2
3
4
5
6
7
8
9
10
11

12
13
14
15
16
17
18
19
20
21
22
23
'"24
25

"u

tci'

"~
utll,

Specific
Gravity.

1i3~

-~=-

-~=
-

o 9992

ccr

~tll,

1.0363
1.0375
1.0388
1 0400
1.0412
1.0424
1.0436
1 0447
1.0459
1 0470
1 0481
1 0492
1.0502
1.0513
1.0523
1.0533
1.0543
1.0552
1.0562
1.0571
1.0580
1.0589
1.0598
1.0607
1.0615

26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

Specific
Gravity.

",:Q

51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71

1.0623
1 0631
1 0638
1 0646'
1.0653
1.0660
1.0666
1.0673
1.0679
1 0685
1.0691
1.0697
1.0702
1 . 0707
1.0712
'1.0717
1.0721
1.0725
1.0729
1.0733
1.0737
1.0740
1.0742
1.0744
1.0746

72

73
74
75
,

~&

~=I
tt.!

",:Q

1 0747
1.0748
1.0748
1.0748
1 0748
1.0747
1 0746
1 0744
1 0742
1.0739
1.0736
10731
1 0726
1.0720
1 0713
1 0705
1 0696
1.0686
1.0674
1.0660
1.0644
1.0625
1 06041.0580
1 0553

7G

77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100

CARBONIC ACID
FREE CARBONIC ACID IN AQUEOUS SOLUTION

The method is bas"ed on the reaction

HiCO a+Ba(OH)2=BaCO a+2H 20":
Procedure.-An excess of standard barium hydroxide solution is added to
the water containing the. carbonic acid. Barium carbon~e is precipitated as
shown in the reaction above. The excess of Ba(OH)2 is now titrated with
standard hydrochloric acid, using phenolphthalein indicator.
1 m!. N /10 Ba(OHh or 1 m!. N /10 HCI is equivalent to 0.0022 g. CO 2.
I

CARBONIC ACID PRESENT AS BICARBONATE

Reaction.-NaHCO a +HCI=NaCI+H 20+C0 2..

Procedure.-The solution is titrated with standard 'HCI, using methyl
-'orange indicator.
. 1 m!. N /10 HCI = 0.0044 g.-C02,

ACIDS

2253

CARBONIC ACID PRESENT J.S CARBONATE

Reaction.-N a 2 CO a + 2H CI = 2N aCI + H 20 + CO 2
Procedure.-The solution is titrated with standard HCI, using methyl
orange indicator. Carbonates of barium, strontium, calcium and magnesium
are titrated with an excess of hydrochloric acid and this excess determined with
standard alkali.
1 ml. N/10 HCI=0.0022 g. CO 2
CITRIC ACID
The free acid may be titrated with sodium hydroxide, using phenolphthalein

indicator. One ml. N /1 alkali = 0.07 gram crystallized citric acid.

VOLUMETRIC DETERMINATION OF OXALIC ACID
PERlIIANGANATE METHOD

About 3 grams of the oxalic acid or its salt are dissolved in 200 m!. of CO 2
free 'water and 50 m!. 2N sulfuric acid added. The solution is heated to
about 70 C. and titrated with standard, normal solution of potassium permanganate, to a faint pink, persisting for three minutes.
1 m!. N /KMn04 = 0.04501 g. HlJ 2 0 4 or 0.06302 g. H 2C 2 0 4 2H 20.
The acid may also be titrated with standard caustic solution.
made in a hot solution, using phenolphthalein indicator.

Titration is

DET:gRMINATION OF PHENOL (CARBOLIC ACID)

Bromine reacts with an aqueous solution of phenol decomposing it to the
water insoluble tribromophenol and forming hydrobromic acid as shown in the
reaction
C 6H 50H +3Br2 = C6H2Bra( 0 H) + 3HBr.
The excess of bromine is determined by adding potassium iodide solution
and determining the liberated iodine by titration with standard thiosulfate.
The method is applicable only to pure preparations of carbolic acid.
Reagent: Standard Bromine.-A solution of free bromine in water cannot.
be kept. The reagent may be prepared, however, by adding a definite amount
of bromate to a bromide solution, which on acidification, will liberate a definite
amount of bromine according to the reaction KBrOa+5KBr+6HCI=3Br2
+6KCI+3H 20. The acid is added only to the portions of the reagent used
for the tests. In preparing a N /10 solution 2.784 grams of pure dry potassium
bromate and 10 grams of potassium bromide are dissolved in a little water and
diluted to 1000 ml.
Procedure.-0.5 gram of phenol is dissolved in It little water and diluted to
1000 m!. To 100 m!. of this, equivalent to 0.05 g. of sample, are added (in a
stoppered bottle) 50 m!. of the bromate solution and the mixture shaken. 5
ml. of concentrated HCI are now added, the solution again shaken and after 15
minutes 2 gramsof potassium iodide are added. The liberated iodine is titrated
in presence of starch solution with N /10 thiosulfate.
1 ml. N /10 KBrOa = 0.001567 g. CeH60H.

,
-2254

ACIDS
TARTARIC ACID

Tartaric acid, cream of tartar, Rochelle salt, tartar ~metic, normal potas~
sium tartrate, iron tartrate and other salts of tartaric arid are obtained from
the residues of wine manufacture. The raw materials-consist of lees, tartars,
calcium tartrate. Tartaric acid is present in these re~idues in the form of
potassium hydrogen tartrate (vitartrate of potassium) br as normal calcium
tartrate. In the examination of the raw material for its evaluation total
tartaric acid and that present as bitartrate are determined.'
ESTIMATION OF ACID POTASSIUM T ARTRATE-OULMAN'S METHOD

Procedure.-3.76 grams of the powdered tartar is placed in a liter flask,

750 m!. of water added, the solution heated to boiling and boiled for 4 to 5
minutes. Prolonged boiling is avoided, as changes may occur which would
cause error in results. The flask is filled to the mark and allowed to cool.
After readjusting to exactly one liter the solution is filtered through a dry filter
,and 500 m!. of this evaporated to dryness on the water-bath in a porcelain
casserole. 5 m!. of water is added to moisten this residue and 100 mI. of 95%
alcohol added on cooling. After standing half an hour 'the alcohol is decanted
through a dry filter, allowed to drain and any potassium bitartrate on the filter
is washed back into the dish with hot water. The solution is made up to about
100 m!. and titrated hot with N /5 KOH. /A correction of 0.2 ml. is added for
the loss of bitartrate in the alcohol.
ESTIMATION OF TOTAL TARTARIC ACID.

GOLDENBERG METHOD

Procedure.-Six grams of the sample containing more than 45% of tartaric

acid or 12 grams if it contains less than 45% are added to 18 mI. of hydrochloric
acid (sp.gr. 1.1) and the mixture stirred 10 minutes. This is now transferred
to a 200-m!. measuring flask and distilled water added to the mark. The well
mixed solution is now filtered through a dry filter in~o a beaker. 100 ml. is
pipetted out and added to 10 m!. of a solution; of potassium carbonate con~ .
taining 66 grams of the anhydrous salt per 100 m!. in a 300 mI. beaker. The
beaker is covered by a clock glass and the solution boiled gently for twenty
minutes; the calcium carbonate precipitates in a crystalline form during this
heating. The solution is transferred to a 200 ml. measuring flask, made up to
mark after cooling, and filtered through a dry filter.
_
100 m!. of the filtrate is evaporated in a porcelain casserole (o?'Pyrex beaker)
on the water bath until the volume is about 15 ml. To the hot solution 3.5 m!.
of glacial.aetic acid are added gradually with constant stirring and the mixture
stirred for 5 minutes more after th'tl addition~ After . standing 10 minutes
100 m!. of 95% alcohol are added and the mixture again stirred 5 Ihinutes.
Upon settling 10 minutes the precipitated bitartratei'is filtered off, suction being
applied. The precipitate is washed with alcohol until, free of acid. (Test
30 m!. This should require the same titration with N /5 alkali with phenolphthalein indicator as 30 m!. of the alcohol used in the washing.) The precipitate on the filter is washed into a porcelain dish with 200 m!. of hot water
and titrated hot with N /5 potassium or sodium. hydroxide, using litmus paper
,as indicator.
- .

ALKALIES

2255

The alkali is standardized against pure potassium hydrogen tartrate ..

Corrections.-Deduct 0.3% for material containing less than 45% tartaric
acid, 0.3% for raw material containing 45-60% and 0.:2% for material containing 60-70%. No correction is made for better grades of material.
IMPURITIES IN TARTARIC ACIl)

Iron and Alumina (alum).-These are determined in the ash of the ignited
prodjlct, the ash being dissolved in hydrochloric acid. P 2 0 5 may also be determined in this ash.

Arsenic.-This may be determined by the Gutzeit Method. (See chapter

on Arsenic in Volume 1.)
Lead.-This is determined by the colorimetric prflcedure given on page 517,
Vol. 1.
Free Sulfuric Acid in Tartaric Acid Liquors.-The solution 10 ml. or more
is treated with ten times its volume of alcohol, and after settling overnight
the mixture is filtered . Sulfuric acid is determined in the alcohol filtrate in
the usual way.
.

ALKALIES
ANALYSIS OF SODIUM HYDROXIDE
Commercial caustic soda, purchased in blocks packed in iron drums, should
be sampled with care. The hydroxide sets first on the outside, so that the
impurities segregate towards the core of the block. In order to get a representative sample different sections of the block should be tapped. The material
takes up moisture and carbon dioxide from the air, sO that the surface of the
sample should be removed before weighing. The weighing and dissolving of
the sample should be done as rapidly as possible.
PROCEDURE

Ten grams of the hydroxide are dissolved in water and the solution made
up to exactly 500 ml. Aliquots of this solution are taken for the following
determinations.
Total Alkali.-FiftY mL of the caustic solution, equivalent to 1 gram of the
solid, are titrated with N /H 2S0 4 in presence of methyl orange indicator, until
the faint pink end-point is obtained.
1 mI.. N/H 2S0 4 =.031 grams of Na 20 (total alkali actual),.
NOTE.-In t4e ~ew York and Liverpool test N /H2S04 value is .032 g. NaOH per mi.

2256

ALKALIES

Sodium Hydroxide.-Fifty ml. of the sample is treated with 100 m!. of 10%
BaCh solution and the N aOH then titrated with lV/H 2S0 4 in presence of
phenolphthalein indicator.
I

1 mI. N/HzS04=.04 gram NaOH.

Sodium Carbonate.-Multiply the difference betwe~n the total alkali titration and the titration for NaOH by .. 053 the result is grams NuzCO s.
Alternative Methods. Sodium Hydroxide and Sodium Carbonate.-Carbon dioxide is readily picked up by NaOH so that it is invariably present in
caustic soda forming sodium carbonate. The carbonate and hydrate may be
determined in the same solution as follows .
.Fifty mI. of the sampl~equivalent to one gram of the solid, are titrated with
normal H 2S0 4, in presence of phenolphthalein, until the pink color just disappears. This occurs when all of the sodium hydroxide is neutralized and the
carbonate has been converted to bicarbonate. This titration may be recorded'
as mI. A.
Methyl orange is now added and the titration co;tinued until the yellow
'
color changes to pink.' This titration is recorded as ml, B.
Then NaOH=(A-B)XO.04 and Na 2 C0 3=2BXO.053.
DETERMINATION OF IMPURITIES
/'

Sodium Chloride.-This may be determined on 25 to 50 ml. (0.5-1 gram) of

the above solution by Volhard's method described in Vol. I, page 271, or the
method for NaCI in soda ash, page 2261, this volume.
1 m!. N/20 AgN0 3=0.002923 gram NaCl.
Sodium Sulfate.-Fifty ml. of the solution equivalent to 1 gram of the
solid are acidified with hydrochloric acid (H 2S0 4 free) and the sulfate precipitated from a boiling solution by addition of 10% ,BaC1 2 solution according
to the standard procedure.
<
BaS04XO.6086 = Na;2S0(.
The following determinations are Beldom required.
Sodium Silicate.-lO graml' of the sample are dissolved in about 100 mI. of
water and the solution acidified with hydrochloric acid. ':AfteJ: evaporation to
dryness the residue is taken up with water. Silica remains insoluble and is
filtered off, washerl, ignited and weighed. Si0 2 X2.0282 = NazSiO s
Sodium Aluminate.-The water extract from silica contains the alumina as
soluble chloride. AI(OH)a may be precipitated with ammonia according to
the standard procedure, then filtered off, washed, il1nited and weighed as Al;03<

AlzOs X 1.6067 = N a 2Alz0 4

Insoluble Matter.-lOD grams of the materi~l is dissolved in a liter of
water, phenolphthalein indicator is added and the solution almost neutralized
with hydrochloric acid, the solution should be slightly alkaline. The sample
is filtered and the residue of sand, Fe 20s, etc., weighed all insoluble matter.
Caustic sofutions attack filter paper so that it is_advisable to partly neutralize
the free alkali before filtration.
-

ALKALiES

2257

Water.-5 to 10 grams of the hydroxide are placed in a small Erlenmeyer

flask, whose weight is known, a funnel is placed in the neck of the flask to
prevent loss of the material and to prevent its absorbing carbon dioxide from
the air. After placing in a sand bath the material is kept at about 1500 C. for
about four hours. It is now allowed to cool, the funnel remaining in the flask,
and ~he
loss of weight determined.
.

DETERMINATION OF STRENGTH OF CAUSTIC LIQUORS BY THE HYDROMETER

The specific gravity of the liquor is taken by the hydrometer exactly as in

case of acids. The strength of the liquor is ascertained by reference to the
table on Sodium Hydroxide, page 2268.
.
POTASSIUM HYDROXIDE

Analysis of potassium hydroxide is similar to that of sodium hydroxide.

The following molecular weights will be of use in the calculations:
KOH=56.11;

K 2 CO a =138.21;
;(Cl=74.56;
K 2Si0 3 = 154.5.

VOLUMETRIC. DETERMINATION

OF

CARBONATE

IN

\",K 2 Ah04=196.4;

SODIUM

BICARBONATE

19

NaHC03 in presence of BaCl 2 gives insoluble BaC0 3 and NaCl, neither

subject to hydrolysis nor acting of phenolphthalein.
The NaHC0 3 to be analyzed is added to a solution of NaOH of known
strength and BaCl z is added to the NazCO a formed. By determining the
fr~e NaOH, that which is combined with NaHC0 3 . can be calculated and the'
equivalent NaHC0 3 thus determined.
ANALYSIS AND TESTING OF IMPURITIES IN CAUSTIC SODA 20
,~eagents and Indicators: Water.-The water used in the following methods should
be distilled water, neutral to phenolphthalein and methyl orange indicators.
Hydrochloric Acid (1-1).-Dilute 500 m!. of colorless hydrochloric acid (Hel) of 1.19
specific gravity with an equal volume of distilled water.
Sulfuric Acid (l-J).-Into 3 liters of distilled water cautiously add with stirring 1
liter of colorless sulfuric acid (H 2S0 4) of 1.84 specific gravity.
Nitric Acid (1-4).-Into 4 liters of distilled water add with stirring 1 liter of colorless
concentrated nitric acid (HNO,).
Ammonium Hydroxide.-The ammonium hydroxide (NH 40H) used should be a
pure, clear, colorless solution of 0.90 specific gravity.
Ammonium Hydroxide (Dilute).-Dilute 100 ml. of ammonium hydroxide (NH 40H)
specific gravity 0.90 with 900 mI. of distilled water.
Ammonium Sulphide Solution.-This solution is prepared by passing hydrogen
sulfide gas (H 2S) through a dilute ammonium hydroxide solution (NH.OH) until the
solution smells strongly of hydrogen sulfide.
Ammonium Oxalate Solution.-Dissolve 40 grams of ammonium oxalate, (NH 4)2C 20 4 H 20, in hot distilled water and dilute to one liter.
.
Disodium Phosphate Solution.-Dissolve. 100 grams of disodium phosphate (NazHPO . 12H 20) in distilled water and dilute to one liter.
Starch Solulion.-Dissolve 0.5 gram of soluble starch in 100 ml. of hot water. If
soluble starch is not available, macerate 0.5 gram of ordinary starch to a creamy paste in
a mortar with cold water, wash into 100 ml. of boiling water and boil for several minutes.
19
20

Ref. J. C. Chiarino, C. A., 18, 1799 (1924).

Standard Method of the Solvay Process Company.

By courtesy of the Company.

2258

ALKALIES

Allow to stand for a few minutes and then pour off. the supernatant liquor for use. Starch
Starch solutions do not keep well and for best results they should b~ prepared fresh every
few weeks.
.
Sodium Thiosulfate Solution.-Dissolve lOOgrams of sodium thiosulfate (Na 2S 20.5H 20) in distilled water and dilute to one liter.
I
Phenolphthalein.-Dissolve 3.5 grams of phenolphthalein 7n 650 m!. of grain alcohol
and make up to 1000 mi. with distilled water.
Potassium Chromate Solution.-Dissolve 50 grams of neutral potassium chromate
(K 2Cr0 4) in 400 mi. of distilled water. Add enough silver nitrate to produce a slight
red precipitate for the removal of chlorides. Filter and make up the filtrate to 1000 mi:
with distilled water.
Methyl Orange.-Dissolve one gram of methyl orange in 1000 mi. of distilled water.
Potassium Sulfocyanate Solution.-Dissolve 50 grams of C. P. potassium sulfocyanate (KCNS) in distilled water and dilute to one liter.
Standard Solutions: Tenth Normal Sulfuric Acid (N/l0 H 2 S0 4 ).-This solution
should be prepared from a stock solution of normal sulfuric acid (N /H2S04) by diluting
exactly 100 mi. to 1000 mi. with distilled water.
The stock solution of normal sulfuric acid is prepared as follows:-This solution
should contain 49 grams per liter of sulfuric acid (H 2S0 4) in distilled water. Mix 30 mi.
of pure sulfuric acid of 1.84 specific gravity with 150 mi. of distilled water in a 1000 mi.
measuring flask and allow to cool to room temperatur!'). Dilute to 1000 m!. with distilled
water and adiust the volume by titrating against 5.3 grams of chemically pure, dry
sodium carbonate (Na 2CO.) until 100 mi. of the acid will just neutralize the 5.3 grams 'of
sodium carbonate, using methyl orange as an indicator.
A watch glass should cover the beaker while' adding the acid to avo,id mechanical
loss. As the end-point is approached care should be taken to stir well, so that the color
change from yellow to pink will be sharp and easily noticeable, the acid being added a
/
drop at a time.
Tenth Normal Sodium Hydroxide (N/l0 NaOH).-This solution should be prepared
from a stock solution of normal sodium hydroxide (N /NaOH), by diluting exactly 100
mi. to 1000 mi. with distilled water.
The stock solution of normal sodium hydroxide is prepared as follows:-This solution
should contain 40 grams per liter of sodium hydroxide (NaOH). Dissolve 41 grams
of chemically pure sodium hydroxide in 250 mi. of distilled water in a 1000 ml. measuring
flask. Allow to cool to room temperature and dilute to 1000 mI. Adjust by adding
distilled water until 50 mI. are exactly neutralized by 50 ml. of normal sulfuric acid
(N /H ZS0 4), titrating cold with 5 drops of phenolphthalein indicator.
~
Twentieth Normal Silver Nitrate (N/20 AgNO.).-This I?olution should contain 8,495
grams per liter of silver nitrate (AgNO.). Dissolve 8.5 grams of chemically pure silver
nitrate in distilled water to which one drop of colorless nitric acid (HNO.) of 1.42 specific
gravity has been added and dilute to 1000 mI. with distilled water. This solution is
practically 1/20 normal. It may be standardized as follows:-Weigh out 0.8025 grams
of chemically pure, dry, ammonium chloride (NH.Cl). Dissolve in 250 mI. of distilled
water and titrate 25 mI. of this solution with the silver nitrate solution, using 3 or 4
drops of pota~sium chromatfil (K.Cr04) solution as an indicator,~l).til the red tint of the
silver chromate just appears. If the silver nitrate solution is exactly,.N /20, 30.0 mI.
will be consumed.,
,
Tenth Normal Potassium Permanganate (N /10 KMnO.).-This' solution should
contain 3.16 grams per liter of chemically pure potassium permanganate (KMn04).
Dissolve 3.2 grams of potassium permanganate crystals in 500 mI. of distilled water in 'a
1000 mI. measuring flask. Dilute to 1000 ml. Standardize as follows:-Add 0.67
grams of recrystallized, chemically pure sodium oxalate( (NazC 20.) and about 150 mi.
of distilled water to a liter boiling flask. Heat until dissolved, then add about 15 mi.
of (1-3) sulfuric acid (H 2S0 4). Heat nearly to boiling and titrate with the potassium
permanganate solution until a permanent pink color is obtained.
Adjust the solution to the required strength of calculate the proper factor from the
following data:
I
100 m!. N/10 KMnO. solution =0.67 grams of sodium-oxalate.
The potassium permanganate solution should be
in an amber colored or black
painted bottle.

!rea

ALKALIES

2259

One Hundredth Normal Potassium Permanganate (N/IOO KMnO.).- Dilute

exactly 100 ml. of the tenth normal potassium permanganate (N /10 KMnO.) solution to
one liter with distilled water. This solution should be kept in an amber colored or black
painted bottle.
One Hundredth Normal Sodium Thiosulfate (NjIOO Na2S20,).-This solution is
made by diluting exactly 100 mI. of tenth normal sodIUm thiosulfate (N /10 Na 2S 20,( to
1000 mI. with distilled water.
Tenth normal sodium thiosulfate is prepared as follows:-Dissolve 24.83 grams of
chemically pure sodium thiosulfate (Na 2S 20"5H 20) in distilled water and dilute to
one liter. Standardize as follows:-Place 40 mI. of tenth normal potassium permanganate (N /10 KMilO.) solution in a 500 mI. flask, dilute to about 100 mI. with distilled
water and add 15 mI. of )1-3( sulfuric acid (H 2SO.), and 10 mI. of 10% potassium iodide
(Kr) solution (100 grams per liter). Titrate the liberated iodine with the tenth normal '
sodium thiosulfate solution until the color becomes a very light yellow. Then add about
0.5 mI. of fresh starch solution and complete the titration until the blue color just disappears. Adjust the volume of the solution from the following data:
40 mI. N/lO KMnO.=40 mI. N/10 Na2B.O,.
.
This solution should be standardized once a month.
Standard Iron Solution.-Dissolve 0.4911 grams of chemically pure, crystallized,
ferrous ammonium sulfate (FeSO.(NH')2S0,6H 20) in 50 mI. of distilled water to
whicll20 mI. of (1-3) sulfuric acid (H 2SO.) have been added. Warm the solution slightly
and add tenth normal potassium permanganate (N /10 KMnO.) until the iron is completely oxidized. Dilute the solution to 1000 mI.
1 mI. of this standard solution =0.00007 grams Fe or=O.OOOl grams Fe.Oa.
Methods of Analysis
Preparation of the Sample.-This operation should be carried out without unnecessary exposure to the air, as caustic soda readily takes up moisture and carbon dioxide,
thus lowering the tests. A part of the sample to be tested is wrapped in a heavy cloth
and crushed with a hammer, the remainder being saved for later determinations. Weigh...
ings should be made as quickly as possible.
Silica (Si0 2).-Dissolve 50 grams of the prepared sample in 200-300 ml. of distilled
water in a 6-inch porcelain or Jena glass evaporating dish, add 2 drops of methyl orange
and acidify cautiously with concentrated hydrochloric acid (HCl), avoiding loss by
spattering. Evaporate to complete dryness on the steam bath, cover the dish and
dehydrate the silica by heating in an oven at 110 C. for one hour. Cool, drench the
residue with 10 mI. of concentrated hydrochloric acid, taking particular care that each
particle of the residue .is moistened with the acid. Then add 300 mI. of distilled water
and warm on the steam bath to dissolve the soluble salts. Filter and wash the residue
free from chlorides with cold water, adding only the first two washings to the filtrate.
Transfer the filter paper and silica to a weighed platinum crucible, ignite in the open
crucible over a low flame until the filter paper is burned off and then over the blast lamp
for 10 minutes. Cool in a desiccator and weigh.
Weight of Si0 2X2=% Si02.
Iron and Aluminum Oxides (Fe.O, AJ.Oa)-Dilute the filtrate from the determination
of silica to 500 ml. in a measuring flask and mix thoroughly. Place 250 ml. of this solution in a 400 ml. beaker, add 1 ml. of nitric acid (HNO,) and about 1 gram of ammonium
chloride (NH.Cl) and boil to oxidize the iron. Add slowly a slight excess of ammonium
hydroxide (NH.OH) and boil until the vapors have only a slight odor of ammonia.
Allow the precipitate to settle, filter and wash the precipitate free from chlorides with
hot water. Place the filter paper and precipitate in a weighed platinum crucible,
carefully burn off the filter paper in the open crucible over a low flame and finally ignite
with free access to the air for 10 minutes oVl\r the Meker burner or blast lamp. Cool in
a desiccator and weigh.

Weight Fe 20, & Al 20, X 4 = % Fe 20, & Al 20,.

Iron Oxide (Fe20,).-Transfer a 100 ml. aliquot portion of the 500 mI. filtrate from
the determinations of silica to a 250-ml. beaker, add about 1 mI. of nitric acid (HNO,)
and boil to oxidize the iron. Then add a slight excess of ammonium hydroxide (NH.OH)
and continue boiling for a few minutes. Allow the precipitate to settle, filter and wash
the precipitate about 5 times with hot water. Dissolve the precipitate off the filter in

ALKALIES
about 20 mI. of (1-3) sulfuric acid (H 2SO.) and "Wash into a :100 ml. graduated flask.
Dilute to the 100 ml. mark and mix thoroughly. Pipette off 25 ml. of this solution into
a 50 ml. Nessler tube, add 5 ml. of potassium sulfocyanate (KCNS) and dilute to the
50 ml. mark with distilled water.
I
To a second 50-ml. Nessler tube, add 5 ml. of (1-3) sulfuric acid and 5 m!. of potassium
sulfocyanate, dilute nearly to tfie 50 ml. mark with distilled water and titrate with
standard iron solution until the colors in the two Nessler tubes match.
Ml. Standard Iron SolutionXO.004= % Fe20s.
Aluminum Oxide (AlzO s).-% Fe 20s & AI 20 s-% Fe 203= % AI 20 s.
Calcium Carbonate (CaCOs).-Adjust the volume of the filtrate from the determination of iron and aluminum to about 300 ml. and acidify with (1-1) hydrochloric acid
(HCI), adding about 5 m!. in excess. Add 10 m!. of ammonium oxalate solution,
- (NH')2C 20., heat to boiling, make slightly alkaline by addin~ ammonium hydroxide
(NH.OH) drop by drop and boil for a few minutes. Allow tlie mixture to stand in a
warm place for 1~2 hours, filter and wash the precipitate free from chlorides with small
quantities of hot water. Usually four or five washings suffice. Place the filter paper
containing the precipitate in the beaker in which the precipitation was made and add
50 ml. of distilled water and 15 ml. of (1-3) sulfuric acid (H 2S04). Heat to 70-80 C.
and titrate while hot under constant stirring with N /10 potassium permanganate
(KMnO.) solution until a faint, permanent pink color is obtained.
M!. N/10 KMnO.XO.02=% CaCO.
Magnesium Carbonate (MgCO.).-Acidify the filtrate from the determination of calcium with (1-1) hydrochloric acid (HCl), adding about 5 ml. in excess. Heat to boiling
and add 10 ml. of disodium phosphate solution (Na2HPO,12H20). Then add ammonium hydroxide (NH.OH) drop by drop under constant stirring until the solution
becomes alkaline and a precipitate begins to form. If no precipitate forms the stirring
should be continued .for at least 5 minutes. ,Cool, add 30 ml. of strong ammonium
hydroxide and stir for 5 minutes more if no precipitate forms. Allow the mixture to
stand several hours or over night in a cool place. Filter and wash the precipitate with
dilute ammonium hydroxide until the washings acidified with nitric acid (HNO.) give
no test fot chlorides. Place the filter paper with the precipitate in an inverted position
in a 250-ml. beaker and dry at 60-80 C. When quite dry and free from ammonia, add
25 ml. of distilled water and a measured excess of N /10 sulfuric acid (H 2S0 4), Stir
until the paper is thoroughly diSintegrated and titrate the excess of acid with N'/10
sodium hydroxide (NaOH), using 2 drops of methyl orange as inaicator.
M!. N/10 H~S04-ml. N/IO NaOHXO.0169 = % MgC03.
Manganese (Mn).-Dissolve 40-50 grams of the sample in 300 m!. of distilled water,
add 2 drops of methyl orange indicator, acidify with hydrochloric acid (HCI) and make
slightly alkaline with ammonium hydroxide (NH.OH). Pass hydrogen sulfide gas (H 2S)
into the solution for Y:i hour and allow the sample to stand at about 100 C. until the
precipitate settles. Filter the precipitated sulfides and wash a few times with a weak
solution of ammonium sulfide, (NH.).S. Dry and ignite in a platinum crucible. To
the residue in the crucible add 5--6 times its weight of sodium carbonate (Na 2CO.),
thoroughly mix and fuse over a burner.
.
Dissolve the fusion in 10 oml. of concentrated nitric acid (HNO.f and heat for thirty
minutes on the steam bath until all of the nitrous oxide fumes are expelled. Wash into
a 250 ml. beaker with about 15 ml. of water, add 5 ml. of N /20 silver nitrate (AgNO.)
a~d warm to 70-80 C. Introduce 1 gram of ammonium persuIfate (NH.).S 20s, a little
at a time, and shake until a good color is developed. Cool, shake, transfer to a 50 m!.
Nessler tube and dilute to the 50 ml. mark. Should the color be too strong the solution
must be diluted to 100 ml. and an aliquot portion used.
Compare the color with N 1100 potassium permanganate (KMnO.) titrated into 50
m!. of cold (1-4) nitric acid which has been previously boiled for several minutes to expel
the nitrous oxide fumes. The (1-4) nitric acid is'free from nitrous fumes and ready for
use when it is colored by 1 drop of N /10() potassium permanganate.
Ml. N 1100 KMnO.X .011 o/t: M
Weight of Sample
0
n.
I .
Copper (Cu).--,Dissolve 100 grams of sample in aboutr400 ml. of distilled wat~r in a
one liter beaker, and acidify slightly with (1-3) suifurio itc1q(H 2SO.). Bring the solution
to a boil and add 20 ml. of strong sodium thiosulfate (Na 2S 20.) solution. Continue the
boiling for Yz hour, filter, using suction and wash several times with hot water. Ignite

ALKALIES

2261

the filter paper in a porcelain crucible, dissolve the residue with 6 to 8 drops of concentrated nitric acid (HNO s) and carefully evaporate to dryness over a low flame until
all of the nitric acid is expelled. Dissolve the residue by heating gently with 10 ml.
of N /10 sulfuric acid (H 2S0 4), wash into a 250 ml. Erlenmeyer flask and add a little
sodium carbonate (NaiCO.) to precipitate basic copper carbonate. Acidify with acetic
acid (C.H 40.), cool and add from 2 to 5 grams of potassium iodide (KI) crystals. Titrate
. the liberated iodine with N /100 sodium thiosulfate (N a 2S2 0.) until the color becomes
light yellow. Then add about 0.5 m!. of freshly prepared starch solution and complete
the titration with N /100 sodium thiosulfate until the blue color disappears.
M!. N /100 Na 2S20. X .063 'Ii C
.
0
opper..
Weight of Sample
ANALYSIS OF SODIUM CARBONATE
SODA

ASH

Sodium carbonate, Na 2CO. (soda ash), and two forms with water of
crystallization crystal carbonate, N a 2CO. H 20, and soda ~rystals or washing
soda, Na.CO.lOH.O, are commonly known. In the analysis of soda ash the
customary demand is for total alkali, sodium carbonate, sodium bicarbonate,
and sodium chloride. In a complete analysis, including the insoluble residue
iron, Fe20., sodium sulfate, sodium thio-sulfate, sodium sulfite, sodium sulfide,
sodium silicate, sodium hydroxide, alumina, and water may be required.
PROCEDURE

Total Alkali, Na 20.-Five grams of the soda ash are dissolved in 50 ml. of
distilled water, preferably in an Erlenmeyer flask, with a funnel, the stem
extending in the neck of the flaSK. 95 ml. of normal sulfuric acid are added
and the solution boiled gently to expel the CO 2, the funnel prevet'lts loss during
boiling. After cooling, methyl orange indicator is added and the titration
completed. The end-point is a faint pink color.
I ml. NjH 2S0 4 =.03I gram Na 2 0.
NOTE.-"New York and Liverpool test" the value of 1 m!. N H2S0 4 =.032 g. Na20.
Sodium Bicarbonate, NaHC0 3.-Fiv:e grams of the soda ash are dissolved
in 100 ml. of water and the solution titrated with normal sodium hydroxide
until a drop of the solution on a spot plate produces an immediate dark color
with drop of silver nitrate.

I ml. N/NaOH=.084 gram NaHCO a

Sodium Carbonate.-Deduct the ml. titration, of NaOH for NaHCO. from
the ml. H 2S0 4 titration for N a 2 0, the difference in ml. multiplied by .053
= gram N a 2 CO .
Sodium Chloride.-Two grams of the ash by the- ammonia-soda process or
5 $!ams by the Leblanc process are dissolved in 50 to 100 ml. of distilled water
and 5 mL of colorless HNO. (sp.gr., 1.42) added. 2 ml. of ferric ammonium
sulfate are used as indicator,21 followed by a few drops of N /20 KCNS solution,
the exact amount of this b~ing noted. The chloride is now titrated with N /20
21 Ferric ammonium sulfate 6% solution made by dissolving the salt in 50 parts of
water to 6 parts of salt and adding an equal volume of colorless nitric acid.

2262

ALKALIES

AgN0 3 solution 22 until the color is just dest:royed, then 1 ml. in excess is added.
The precipitate is filtered off and washed. The filtra;te and washings are now
titrated with N /20 KCNS to a permanent pink color ,23
The total mL KCNS is deducted from the ml. AgNO a added and the difference multiplied by 0.002923 = gram N aCI.
i
NOTE.-8ee Volhard's method for chloride in Vo!' I, page 271.

In addition to the above determin~tions the following may be desired:

Insoluble Matter.-Fifty grams of the soda ash are dissolved in about 500
m!. of water the insoluble matter allowed to settle, the clear solution decanted
through a double filter, which has been weighed, and finally the residue washed
onto the filter. The residue and filters are dried at 100 C. and weighed.
Weight minus tare of filters multiplied by 100 = per cent insoluble matter.
Iron, Alumina, Lime and Magnesia.-These are determined in the insoluble
residue by dissolving out by means of dilute hydrochloric acid. The residue
consists of sand and carbonaceous matter. Iron and alumina are precipitated out
together by addition of ammonia according to the standard procedur~, and
determined. as oxides, Fe 203 and AI 20 s Iron may be determined in this
residue l5y dissolving in HCI and titrating with stannous chloride according to
the procedure given in Vol. I, page 373. Alumina is obtained by difference.
Lime and magnesia will be found in the filtrates from iron and alumina precipitates. Calcium is thrown out as an oxalate and magnesium as a phosphate
. and determined as usual.
/
.
Sodium Sulfate.-Dissolve 5 to 10 grams in dilute HCI and add BaCI 2
The precipitate BaS04 is washed and ignited as usual.
BaS04XO.6086 = gram Na 2S04.
Sodium Sulfite.-5 grams are dissolved in water and the solution acidified
with acetic acid. Starch solution is added and the sul~te titrated with N /10
iodine until the blue color appears.
I

1 ml. N/IO 1=0.006303 gram Na 2SOa.

Sodium Silicate.-Ten grams of tJ:e soda ash are treated with an excess of
HCI and the solution evaporated to dryness,. the silica dehydrated at 110 C.
Mter leaching with water the insoluble Si0 2 is determined by filtering off and
.
igniting by the standar~ procedure.

-._

Si0 2X 2.0282 = Na2SiO a

Sodium Sulfide.-This may be estimated by titration with an ammoniacal

solution of silver nitrate (13.81 g. silver per liter = 1 ml.=.005 g. Na 2S). Add
the reagent until no further precipitation occurs of Ag 2S. To get a good endpoint the solution is filtered just before this point is reached and the titration
completed. See also chapter on Sulfur, Volume 'I.
Loss on Ignition.-The sample 10 grams is ignited at a temperature slightly
below 300 C. The loss is due largely to water and organic matter.
22

N /20 silver nitrate contains 8.495 g. AgNO s per. liter , St'andardize against pure

NaC!.

__ - -

N /20 potassium thiocyanate, KeNS, contaIns 4.86 grams of the salt per liter.
The solution should be standardized against the silver ~itrate ~olution.
23

2263

ALKALIES
MODIFIED SODAS
CAUSTICIZED ASH

The determination of total alkali, sodium hydroxide and sodium carbonate

are generally required. 20 grams of the sample are dissolved in water and the
solution made to one liter. Aliquot portions of this solution are taken for
analysis.
Total Alkali.-The determination is similar to that described for soda ash
on page 2261, the test being made on 50 ml. of the solution equivalent to 1 gram
of sample. It is advisable to add a slight excess of N jH2S0 4 and titrate back,
with N IN aOH using methyl orange indicator. The end-point is a faint pink.
1 ml. N/H 2S0 4 =0.031 gram Na 2 0.
Sodium Hydroxide.-Fifty ml. of the sample equivalent to 1 gram of the
solid is taken. The analysis is the same as that described for determining
NaOH in caustic soda on page 2255.
Sbdium Carbonate.-The difference between the acid titration for total
alkali and that for sodium hydroxide is multiplied by 0.053 = gram Na 2C0 3

WASHING SODA (NEUTRAL SODA)

The determination of total alkali, sodium carbonate, sodium bicarbonate,

and sodium chloride are generally required. The method of analysis is the
same as has been described under soda ash. See page 2261.
METHODS FOR TESTING SUPER ALKALIES 24

Preparation of Sample.-Weigh about 40 grams of the sample, using a weighing bottle and taking the weight of sample by difference. Wash into a standard
500 ml. graduated flask with distilled water, cool to room temperature, dilute
to the mark and mix thoroughly. Use aliquot portions of this solution for the
determinations.
All calculations below refer to the weight of this original sample
Total Alkali (Na 20).-Titrate 25 ml. of the prepared solution in a 6 inch
porcelain dish with N /1 H 2SO 4, using 3 drops of methyl orange as 'indicator.
ml. NIl H 2S0 4 X62.0 _01 N 0
Wt. of original sample - /0 a2 .

Sodium Hydroxide (NaOH).-To a 25 ml. aliquot of the prepared sample in

a 6 inch porcelain dish, add about 100 ml. of BaCb solution and titrate with
Nil HCI, using 5 or 6 drops of phenolphthalein as indicator, until the pink
color is discharged.
ml. Nil HCIXSQ.02
Wt. of original sample

% NaOH.

Sodium Carbonate (Na 2C0 3).-(% Na 2 0-% Na.OHXO.7748) X 1.7097

=% Na 2 C0 3
24

The Solvay Process Company's Standard Method by courtesy of the Company.

2264

ALKALIES
METHODS FOR TESTING NEUTRAL (MODIFIED) SODAS

Total Alkali (Na 20).-Weigh accurately 5 grams df the sample and dissolve
in about. 100 ml. of distilled water in a 250 ml. beaker. Titrate with Nil H 2S0 4,
using 3 drops of methyl orange as indicator.
\

Sodium Bicarbonate (NaHCOa).-Place 8.4 grams of the sample in a 250m!. beaker, dissolve in 50 m!. of distilled water and titrate with Nil NaOH ,
until a drop of the solution added to a drop of freshly prepared silver nitrate
indicator on a spot plate gives instantly a dark color.
'

m!. ~/1 NaOHX8.4=% NaHC0 3

WeIght of sample

Sodium Carbonate (Na 2 CO a).-(% Na2.0-% NaHCO aX.3690)X1.7097

=% Na 2 CO a.
Sodium Chloride (NaCI).-Dissolve 5 grams of the sample in-50 ml. of
distilled water in a 250-ml. beaker. Acidify slightly with iron indicator iwlution. Add a few drops of N 120 KCN~ ~rom a burette, .noting the quantity;
then titrate, while stirring constantly, with N /20 AgNO a to corpplete decolorization. Finally titrate back with N 120 KCNS until a faint reddish color
persists.

(m!. N/20 AgNOa-ml. Nj20 KCNS) X.2923

Weight of sample

o/c
0

N CI
a .

Moisture (H 20).-lOO%-(%Na 2C0 3 +%NaHC0 3 +%NaCI) =% H 20.

(By difference).
ANALYSIS OF SODIUM BICARBONATE 25

Place in a tared platinu{n crucible 4 g. of sodium bicarbonate and record

the total weight. (Weight A.) Dry the sample in a desiccator for 24 hours
over conc. H 2S0 4 and weigh. (Weight B.)
_..._

(1)

(A-B)

X 100

.
= % mOIsture.

Heat the platinum crucible for about 2 or 3 hours in a muffle at temperatures

between 500 and 600 0 F. Cool in a desiccator. :(Weight C.)
(2)

(B-C) X2.7097X100 ~% NaHCO .

a
4

25 By courtesy of the Diamond Alkali Compa?~t,biougl]. the kindness of J. D.

Mattern.
-

ALKALIES

2265

Dissolve the contents of the platinum crucible in water and titrate with
normal HCI using methyl orange as indicator.
m1.X.053XlOO
4

01
10

N
d as N a 2C 0 3 as f ound'm expreSSIOn
. 2.
az O
expresse

% Bicarbonate X84'
. 53
=

% Blcarbonate expressed as

Na 2 CO g

% total Na 2 C0 3 minus % Na 2 C0 3 from bicarbonate is equal to % Na 2 CO g

present in the sample.
.
SODIUM CARBONATE DETERMINATION IN SODIUM HYDROXIDE 26

Commercial sodium hydroxide is now manufactured in a high degree of

purity. The sodium carbonate contents may be found in some instances as
low as .4%. In order to determine the exact amount of Na 2 CO g present in
the sample the common methods using 1 or 2 g. of sample cannot be applied
because the difference 'in the two burette readings 'Will be as little as .1 ml.
(1/10 of 1 mI.).
Weigh out 220 g. of NaOH or if caustic liquor should be analyzed 440 g.
of iiquor, dissolve in H 2 0 and make volume up to 500 mi. Fill in a burette
50 mI. of this solution. Fill 500 ml. burette with normal HCI. Run out
exactly.500 ml. into a liter beaker (lower ~ark of the burette). Add 300 mi.
H 2 0. Add 3 or 4 drops of methyl orange and titrate the acid with the NaOH
solution in the 50 mI. burette until color changes to yellow. Note number of ml.
of caustic used. Titration A. Flace in another beaker exactly the same
amount of caustic solution out of the same burette. Add 300 mI. 10% BaCl z
solution and 3 or 4 drops phenolphthalein. Now titrate with normal HCI
filled in the 500 ml. burette, until pink color just disappears. Titration B.
Titration BX.04XlOO
22
(A-B) X.053XIOO
22

01
10
01
10

N OH

N CO
a2

DETERMINATION OF SODIUM CARBONATE AND HYDROXIDE IN A

MIXTURE.
WINKLER'S METHOD

Piinciple.-A weighed portion of the sample is tested for total alkalinity

with standard acid, using methyl orange indicator. A second weighed portion
is treated with an excess of barium chloride reagent, which precipitates the
carbonate leaving the hydroxide. This may now be titrated using phenolphthalein indicator. The reactions below show the steps of the procedure.
, A. Na 2CO a +H 2S0 4=Na 2S04 +H 20+C0 2 Methyl orange indicator.
2NaOH +H 2S0 4=Na 2S04 +H 2 0. Methyl orange indicator.
Be Na 2 CO a +BaC1 2 =BaC0 3 +2NaCI.
2NaOH +BaCI 2 = Ba(OHh+2NaCI.
Ba(OH)z+H zS0 4=BaS04 +2H 20. Phenolphthalein indicator.
26 By courtesy of the Diamond Alkali Company through the kindness ot J. D.
Mattern.

ALKALIES

2266

Procedure.-Dissolve in a beaker 5-6 grams of the powder (carefully

weighed) in water, transfer to a 500-ml. graduated v,olumetric flask, washing
out the beaker into the flask. Make the solution to 500 ml. by diluting to the
mark. For analysis take 100 ml. portion equivalent td 1/5 of the weight taken.
Portion A. To 100 ml. portion add methyl orange indicator and titrate
with normal sulfuric acid until an orange red color is obtained.. Record the ml.
titration and mark as "A." This titration is due to the total Na 2CO a and
NaOH.
Portion B. To 100 ml. portion 20-25 ml. of a 20% solution of BaC]2. Add
phenolphthalein indicator and titrate slowly with constant stirring with the
normal acid, and record the mI. under" B." This is due to N,aOH present in
the sample.
Calculation-" A" is total alkalinity.
" A "-" B " is Na2CO a Multiply ml. by 0.053.
" B " is NaOH. Multiply ml. by 0.04.
NOTEs.--Stirring is necessary in titration "B;' since the acid must not be allowed to
react with BaCO. as it would if there were an accumulation of the acid in one portion
of the solution.
Total alkalinity may be determined by phenolphthalein by the following procedure.
Titrate the solution in presence of this indicator. Now heat to boiling to expel C02,
cool and again titrate. Repeat this as long,as a color appears 'on boiling. When the
solution remains colorless on boiling the end-point has been reached.
ESTIMATION

OF

THE CARBONATES AND HYDROXIDES OF POTASSIUM

SODIUM WHEN TOGETHER IN SOLUTION 27

AND

Procedure.-A measured volume of the solution is titrated, using phenolphthalein as indicator. The acid used is equivalent to all of the hydrate and
half the carbonate; methyl orange is now added ancl/the titration completed;
the additional amount of the acid used is equivalent to half the carbonate,
therefore the amount of acid required for the carbonates and for the hydrates
can be calculated from these figures.
The fully neutralized solution is evaporated to dryness and the residue
weighed. The result is the weight of the mixed sulfates, due to the carbonates
----.......
and hydrates of potassium and soda in the solution.
Calculate the total acid required to its equivalent of potassium sulfate; subtract from this result the weight of the mixed sulfates, and the difference is due
to the sodium sulfate in the mixed sulfates, owing to the difference in the molecular weights of potassium sulfate and sodium sulfate. The whole of the acid
used has been calculated to potassium sulfate, and(as the acid was neutralized
by carbonates and hydrates, it is evident the proportion of total sulfate, due to
the carbonates and hydrates, is equivalent to the amount of the acid used for
each respectively; therefore the proportion of the above obtained difference due
to the carbonates and the hydrates respectively is also p,roportional to the
amount of acid used for each.
' -'
_,.27

W. A. Bradbury and F. Owen.

Chern. News, 107,-85,2778 (1913).

ALKALIES

2267

Example.-A solution of the mixed carbonates and hydrates of potassium

and sodium required:
(40 ml. NaOH : 40 ml. KOH) 80 ml. of acid to neutralize the hydrates.
(10 ml. Na 2 CO a : 10 ml. K 2 CO a) 20 ml. of acid to neutralIze the carbonates.
100 ml. total acid required to neutralize the solution.
Total acid 100 ml. calculated to K 2S0 4

=0.87 gram.

Total neutralized solution evaporated to dryness (K 2S0 4+Na 2S04)

=0.79 gram.
Difference due to NazSO. in weighed sulfates
=0.08 gram.
, (K 2S0 4-Na zS04) : Na 2S04 : : Diff. :
32
142 :: 0.08 : 0.355 Na 2 S04 present in the mixed sulfates.
And the mixed sulfates 0.79 gram-0.355 =0.435 K ZS0 4 present in the mixed
sulfates.
crherefore the .mixed sulff).tes consist of Na 2S04 0.355 gram, K ZS0 4 0.435
gram.
The proportion of the acid used for the hydrates is 80(100 and for the carbonates is 20(100.
Therefore the proportion of the difference (0.08) due to Na 2S04 from the
hydrate NaOH is 0.08X80j100=0.064. For the carbonates=0.08X20jI00
=0.016.
.
By the above ratios 32 : 142 : : 0.064: 0.284 Na 2S04 from NaOH=40 ml.
N /10 acid. 32: 142 : : 0.016 : 0.071 N a 2 S04 from N a 2 CO = 10 ml. acid.
80 ml. acid used for the hydrates =0.696 K 2S0 4
0.284 Na 2S04 from the NaOH
=0.348 K 2S0 4.
Difference = K 2S0 4 from KOH
= 0.348 = 40 ml. N /10 acid.
20 ml. acid used for the carbonates~0.174.
0.071 Na 2S04 from the Na 2 CO a
=0.087.
= 0.087 = 10 ~l. acid.
Difference = K 2 S0 4 from K 2 CO a
The figures correspond with the quantities taken.
80 ml. of acid to neutralize the hydrates.
20 ml. of acid to neutralize the carbonates.
NaOH 0.284=40 ml. acid}
.
Na 2S04 =0.355 { Na 2 CO a 0.071 = 10 ml. acid 80 ml. aCid.
KOH 0.348=40 ml. acid}
.
K 2S0 4= 0.435 { K 2 CO a 0.087 = 10 ml. acid 20 ml. aCid.
Totals

0.790

0.790(a)100.

Calculate these sulfates to the corresponding hydrates and carbonates.

OPTIONAL

METHOD FOR DETERMINATION OF SODIUM BICARBONATE

SODIUM CARBONATE IN PRESENCE OF ONE ANOTHER

AND

Five grams of the sample are dissolved in CO 2-free water, and the solution
made up to exactly 250 ml. in a measuring flask. Aliquot portions of 25 mI.,

ALKALIES

2268

SODIUM HYDROXIDE SOLUTION ,AT 15

LUNGE
j

Specific
Gravity.

Degrees
llaume.

Per Cent
N",O.

Degrees
Twaddell.

Per Ce1t
NaOH!

r Liter contains
Graws

Na20.

'1.007
1.014
1.Q;!2
1.0.29
1.036
1.0.45
1.0.52
1.060.
1.0.67
1.0.75
1.0.83
1.fl91
1.10.0.
1.10.8
1.116
1.12.3
1.134
1.142
1.152
1.162
1.171
1.]80.
1.190.
1.20.0.
1.210.
1 220.
1 231
1 241
] 252
1.263
1.274
1.285

1.297.
1.30.8
1.320.
1.332
1.345
1.357
1.370.
1.383
1.397
1.410.
1.424
1.438
1.453
1,468
1.483
1.498
1.514
1.530.

1.0
2.0
3.1
4.1
5.1
G.2
7.2
8.2
0.1
10..1
11.1
12.1
13.2
14.1
15.1
16.1
17.1
18.0.
10.1
20..2
21.2
22.1
23.1
24.2
25.2
26.1
27.2
28.2
29.2
30..2
31.2
32.2
33.2
34.1
35.2
36.1
37.2
38.1
39.2
40..2
41.2
42.2
43.2
44.2
45.2
4G.2
47.2
48.2
49.2
50..2

1.4
2.8
4.4
5.8
7.2
9.0.
10. .4
12.0.
13.4
15.0.
16.G
18.2
20..0.
21.6
23.2
25.0.
26.8
28.4
30..4
32.4
34.2
36.0.
38.0
40..0.
42.0. ,
44.0.
46.2
48.2
50..4
52.6
54.8
57.0.
59.4
61.6
64.0.
66.4
69.0.
71.4
74.0.
76.G
79.4
82.0
84.8
87.6
9D.G
93.G
96.G
99.0
102.8
106.0

.0 .47
0.93
1.55
2.10.
2.60.
3.10.
3.60.
4.10.
4.55
5.0.8
5.67
6.20.
G 73
7.30.
7 ..80.
8.50.
/9.18
9.80.
10..50.
11.14
11.73
12.33
13.0.0.
13.70.
14.40.
15.18
15.96
16.76
17.55
18.35
19.23
20..0.0.
20..80.
21.55
22.35
23.20.
24.20. ,
25.17
26.12
27.10.
28.10
29.05
30.0.8
31.00
32.10
33.20.. 34,40.
35.70
36.90
38.00
/"

0..61
1.20
2.00
2.70.
3.35
4.0.0.
4.64
5.29
5.87
6.55
7.31
8.0.0.
8.68
9.42
10..0.3 .
10.97
11.84
12.64'
13.55
14.37
15.13
15.91
16.77
17.67
18.58
19.58
20.59
21.42
22.64
23.67
24.81
25.80.
26.83___
27.80.
28.83
29.93
31.22
32.47
( 33.69
. 34.96
36.25
37.47
~8.80

39.99
.' 41,41

42.83
44.38
46.15
47.60.
49.0.2

4
9
16
22
27
32
38
43
49
55
61
68
74
81
87
96
104
112
121
129
137
146
155
164
174
185
196
20.8
220.
232
245
257
270.
282
295
309
326
342
359
375
392
410
428
446
466
487
510.
535
559
581

-_
NaOH,

6
12
21
28
35

42
49
56
63

70.
79
87
95
10.4
112
123
134
144
156
167
177
183
20.0.
212
225
239
253
266
283
299
316
332
348
.364
381
399
420
441
462
483
506
528
553
575
60.2
629
658
691
721
750.

ALKALIES

22Wl

AQUA AMMONIA
W. C. FERGUSON

ACCORDING TO

De,rees
Baume.

Sp. Gr.
60
_F.
6,

Per Cent Degrees

NH,.
BaumEl.

--- --- --- --- -----10.00 1.0000

10.25 .9982
10.50 .9964
10.75 .9947
11.00 .9929
11.25 .9912
11.50 .9894
11. 75 .9876
12.00 .9859
12.25 .9842
12.50 .9825
12.75 .9807
'13.00 .9790
13.25 .9773
13.50 .9756
13.75 .9739.
14.00 .9722
14.25 .9705
14.50 .9689
14.75 .9672
15.00 .9655
15.25

.00
.40
.80
1.21
1.62
2.04
2.46
2.88
3.30
3.73
4.16
4.59
5.02
5.45
5.88
6.31
6.74
7.17
7.61
8.05
8.49

.9639

15.50 .9622
15.75 .. 9605
16.00 .9589
16.25 .9573

8.93

Ii

Sp. Gr .

Sp. Gr.
60F. Per Cent Degrees
Baume.
NH3
60

60
_F.
60

Per Cent
NH..

- - - _-- - - 23 00 .9150 23.52

23.25 .9135 24.01
23.50 .9121 24.50
23.75 .9106 24.99
24.00 .9091 25.48
24.25 .9076 25.97
24.50 .9061 26.46
24.75 .9047 26.95
25.00 .9032 27.44
25.25 .9018 27.93
25.50 .9003 28.42
25.75 .8989 28.91
26.00 .8974 29.40
26.25 .8960 29.89
26.50 .8946 30.38
26.75 .8931 30.87
27.00 .8917 31.36
27.25
8903 31.85
27.50 .8889 32.34
27.75 .8875 32.83
28.00 .8861 33.32

16.50
16.75
17.00
17.25
17.50
17.75
18.00
18.25
18.50
18.75
19.00
19.25
19.50
19.75
20.00
20.25
20.50
20.75
21.00
21.25
21.50

.9556
.9540
.9524
.9508
.9492
.9475
.9159
.9444
.9428
.9412
.9396
.9380
.9365
.9349
.9333
.9318
.9302
.9287
.9272
.9256
.9241

11.18
11.64
12.10
12.56
13.02
13.49
13.96
14.43
14.90
15.37
15.84
16.32
16.80
17.28
17 ..76
18.24
18.72
19.20
19.68
20.16
20.64

21. 75

.9226

21.12

28.25

.8847

33.81

.9211
.9195
.9180
.9165

21.60
22.08
22.56
2304

28.50
28.75
29.00

.8833
.8819
.8805

34.30
34 79
35.28

9.38 .. 22.00
9.83 22.25
10.28 22.50
10.73 22.75

ALLOWANCE FOR TEMPERATURE

The coefficient of expansion for ammonia solutions, varying with the
temperature, correction must be applied according to the following table:
Corrections to be Added for Each
Degree Below 60F.
Degrees
Baume.

40F.

50F.

14 Be .015 Be .017 Be

16
18

20
22

26

.021
.027
.033
.039
.053

" .023 "

" .029 "
" .036 ",
" .042 "
" .057' "

Corrections to be Subtracted for Each Degree'

Above 60F.
70F.

.020 Be
.026 "
.031 "
.037. "
.043 "
;057 "

goOF.

80F.

.022
.028
.033
.038
.045
.059

De

"
"
"

"

,H

IooF.

.024 Be .026 Be
.030 " .032 "
035 " ,037 "
040 ,r .042 "
.047 "

ALKALIES

2270

equivalent to 0.5 gram, are taken for analysis. The sample taken is titrated
\vith N /5 hydrochloric acid in presence of phenolphthal~in indicator (2-3 drops).
The ml. titration recorded as " A," represents one half of the sodium carbonate
present. Methyl orange indicator is now added (2-3 ~rcips) and the titration
with the acid continued until the solution turns faintly pink. The remaining
carbonate and all of the bicarbonate are now titrate<,l. The ml. titration is
recorded as B.
2AXO.0106=Na 2CO a and B-AXO.0168=NaHCO a
The author desires to acknowledge his indebtedness to Dr. W. B. Hicks,
Chief of the Analytical Department, The Solvay Process Company, for his
review of the section on Alkalies.
Determination of Small Amounts of Acid in Ether.-L. P. Hall, Ind. Eng.
Chem., Anal. Ed., 2,244 (1930), recommends the use of sulfophthalein indicator
such as bromothymol blue for the titration. Care should be exercised against
CO 2 contamination. Color changes from green to blue.
ANALYSIS OF AQUA, AMMONIA
Details for the complete analysis of crude ammoniacal liquor will be found
in Volume I on pages 637-639.
The per cent ammonia in a solution free from other substances may be
determined' with a fair degree of accuracy by determining the specific gravity
of the solution by means of the hydrometer. Since the specific gravity of
aqua ammonia is less than 1, hydrometers graduated for measuring liquids
lighter than water are used. The formula for 1 degre~ Baume = (140jsp.gr.)
-130. Reference is made to the table on Aqua Ammonia.
.
Provided no other basic constituent is present, free ammonia in solution is
best determined by direct titration with an acid in presence of methyl orange
or methyl red as indicator.
!
Procedure.-About 10 grams of the solution .in a weighing bottle with glass
stopper is introduced into an 800-ml. Erlenmeyer flask containing about 200 ml.
of water and sufficient ! normal sulfuric acid to combine with the ammonia
and about 10 ml. in excess. The flask is stoppered and warmed gently. This
forces out the stopper in the weighing bottle, the ammonia combining with the
acid. Upon thorough mixing, the solution is cooled, and -the. excess of acid
is titrated with half normal caustic.
....... .
Factor.

One ml. ! N H 2S0 4 = 0.0085 gram NH g

H 2S0 4 XO.3473 = NH a

NOTE.-The aqua ammonia exposed to the air will lose ammonia, hence the sample
should be kept stoppered. This loss of ammonia is quite appreciable in strong ammoniacal solutions.
.
ZINC CHLORIDE METHOD FOR DETERMINING FREE .LIME IN
PRESENCE OF CALCIUM CARBONATE

The following procedure is of interest to the. suga~ chefuist. The method

developed by John C. Bailar, Great Western Sugar- Company, depends upon

ALKALIES

2271

masking by means of a green dye the faint pink color produced by calciu~
carbonate in water on phenolphthalein. The color of the dye does not interfere
in the titration of the alkalinity of calcium hydroxide. The cheapness of the
procedure as compared with the iodine method commends its use in the commerciallaboratory.
.
Solutions: Zinc Chloride Solution.-This is made by dissolving 25 grams of
the salt in a liter of water and filtering. 1 m!. of this solution is equivalent to
about 0.01 g. CaO.
Sodium Hydroxide Solution.-The solution is made of such strength that
1 ml. is equivalent to 1 m!. of the zinc chloride solution
Indicator.-5 grams of phenolphthalein are dissolved in 500 mt of 95%
alcohol. This is mixed with a solution containing 2 grams of Alkali Fast
Green E dissolved in 500 m!. of water.
Standardization.-T.he sodium hydroxide solution is standardized against
a standard acid. The zinc chloride solution is now standardized against the
alkali ll,s follows: Into a liter flask c~mtaining 500 rn!. of boiling hot distilled
water 25 to 50 ml. of the NaOH solution are accurately measured from a
burette .. About 0.5-1.0 m!. of the indicator is added and the zinc chloride is
run in until a green color is obtained that persists for half a minute. 10 ml.
excess of the ZnCl 2 solution are now added and this excess, after vigorous
shaking, is titrated with the NaOH solution until a lavender (not pink) color
is obtained.
Determination.-In a dry one-liter flask is placed 1 gram of the sample.
500 ml. of boiling, CO 2-free water are added. As soon as the lime has hydrated,
the solution is titrated with zinc' chloride solution until a green color appears,
an excess of zinc chloride is added and the excess determined with standard
NaOH solution as in case of the standardization above. The end-point is a
lavender color.
NOTEs.-The flask used with the lime determination must be dry, otherwise the
water in small amount reacting with the sample will cause difficulty in the subsequent

~~

The quantity of indicator in standardization and the determination must be the

sam!!, as the quantity influences results. The Alkali Fast Green E dye is made by the
National Aniline Company, Buffalo, N. Y.
.

ANALYSIS OF ALKALINE MIXTURES

SODIUM CARBONATE, -BICARBONATE, -BORATE

The determination of sodium carbonate, sodium borate and sodium bicarbonate in solids or in solution together with neutral salts can b~ carried out
by titration with standard acid and alkali.
Sampling.-In all determinations in which mixtures of carbonate and bicarbonate are evolved the contact of the sample with the air or moisture must
be avoided as far as possible. Solutions of carbonate and bicarbonate will be
altered rapidly until an equilibrium is established, which is independent of
the initiai composition of the sample.

* Courtesy of H. H. Chesny.

2272

ALKALIES

In the case of solids weigh out 0.5 to 2.0 g. and make up with carbon
dioxide free water to 100 ml. Weigh out a second sample for the bicarbonate
determination.
In the case of liquid samples the dilution of 10 ml.j to 100 ml. is convenient.
The bicarbonate determination requires an undiluted sample; It is convenient
to use 2 or 5 mI.
I
.
Procedure.-Total Alkali: Titrate a sample with O~l normal hydrochloric
acid, using as little methyl orange indicator as possible, to the dead ,neutral
point.
Let" a" =ml. of 0.1 n HCI used.
Bora'x: Use the same sample which has been titrated with the hydrochloric'
acid to the dead neutral point in the titration for total alkali, boil for 2 or 3
minutes, cool sufficiently, preferably below 20 C., add mannitol 28 (3. g. for
each gram of Na2B 40 7) or cerelose (30 g. for each gram borate), followed by
phenolphthalein indicator (excess not harmful) and titrate with O.l-normal
sodium hydroxide solution until a pink color is obtained. Add mqre of the
polyhydroxy compound and, if a fading of the pink color occurs, continue the
titration with the sodium hydroxide solution until on addition of ~annitol or
cerelose no more fading takes place.
Let" b "= m!. of 0.1 n NaOH used.
Bicarbonate: The separate samples (as described under" sampling") are
used for this determination. The samples are placed in a dry flask and at
once an excess of 0.1 normal sodium hydroxide solution is added, the solution
shaken gently for a few seconds (shaking violently too long will result in
absorption of carbon dioxide from the air) followed by a slight excess of neutral
barium chloride solution.
Let" c" =ml. of 0.1 n NaOH added.
The solution together with the white p~ecipit~te (consisting of barium
carbonate (and sulfate, is titrated with 0.1 normal hydrochloric acid using
phen~lphthalein as indicator until colorless.
Let" d" =ml. of 0.1 n,HCI used.
Determine the carbonate content of the sodium hydroxide solution' hy
measuring out 100.0 mI., adding barium chloride and phenolphthalein indicator
and titrating with 0.1 n HCI.
Let "e"=ml. of O.ln BCl used.

Calculation:
Total alkali N a 20: " a" times 0.0031 = g. of ]\fazO in sample taken.
Bicarbonate N aH C0 3 : [ ( " c "+" e " /100) - (Ii d " - t "b") X 0.0084J = g.
of sodium bicarbonate in sample taken. Note: The actual weights of the
flTiginal sampl~ used for the titration on total alkali and borax 29 must be equal

Mannitol as well as cerelose should be tested fO_E.oUeutraIity.

29 In the presence of large amounts of borax the-precipitate is filtered off and washed
carefully with CO 2-free water.
28

ALKALIES

2273

to the weight of the undiluted sample taken for the bicarbonate determination,
or due allowance for the ration between the number of ml. used in a," " b,"
" c" and" d" must be made, if a different weight is taken .
Borax N a2B407 IOH 20: "b" times 0.00955 = g. of sodium tetraborate
decahydrate, or " b" times 0.00505 = g. of anhydrous sodium tetra borate in
sample taken.
For more accurate analysis:
(I

[(" b "X" e "/100)

+ {" b "XO.02(IOO -" e ") 1]

times the above given factors equals the grams of borax.

Sodium carbonate Na 2 C0 3 : "a"-[("c"X"c"ilOO)-("d"-"b")]

XO.0053 = g. of sodium carbonate in sample taken.

DETERMINATION OF pH -HV-THEG0b0RIMETRI6-M-E-T:-H-G-])l
INTRODUCTORY
>

L.

Acidity and alkalinity have long been recognized as important factors in

practically all branches of research and industrial work. Sugar manufacturers
and refiners, electroplaters and electrotypers, paper manufacturers, sanitary
engineers, agriculturists, bacteriologists, biologists, pathologists, etc., have
therefore used various methods for detecting and controlling acidity and alkalinity, since they have learned by experience that this factor has a very
marked effect on the yield and quality of their products, the efficiency of manufacturing processes, corrosion of metals, adsorption of dyes by fabrics and" clays,
the stability of various materials, the growth of bacteria, the diagnosis of
diseases, etc.~
The colorImetric method of measuring hydrogen ions has many advant~ges
that appeal to practical workers. It is rapid and simple in operation and
compared to other methods very inexpensive. The entire pH range or any
portion of it can be covered by means of portable equipment for use in factory
or field. The difficulties are such as can invariably be remedied. For these
reasons the colorimetric method is now being used for investigation and control
in every field where the determination of hydrogen ions is important.
By mutual agreement the hydrogen electrode is regarded as the fundamental standard in all determinations of hydrogen ion. All other methods are
secondary whether potentiometric (quinhydrone electrode, glass electrode,
antimony electrode) or colorimetric. All such methods must be standardized
/.
against the hydrogen electrode. ~
MEANING OF

pH

Intensity of heat is expressed as degrees on the thermometer scale, and

density as degrees on the Baume scale. Similarly the pH scale is used to denote
intensity of acidity and alkalinity.
'"
The numbers from 0 to' 14 are used to express pH values, that-is,jntensity of
acidity and alkalinity. The value pH 7.0, halfway between 0 and 14, is the
neutral point; that i&, a solution having a pH of 7.0 is neither acid nor alkaline.
The numbers below 7.0 denote acidity, intensity of acidity increasing as the
numbers decrease. Thus a solution of pH 6.4 is wiry slightly acid, one of pH 6,0
is more intensely acid and one of 4,6 is still more intensely acid, etc.
On the other hand 'the numbers between 7.0 and 14.0 are used to denote
alkalinity, the intensity of alkalinity increasing as the numbers increase. Thus
a solution of pH 7.2 is very slightly alkaline, one of pH 8.6 is more intensely,
alkaline and one of 10.4 is still more intensely alkaline. These numbers represent absolutely definite degrees of acidity and alkalinity and therefore. a state_J-

Chapter by W. A, Taylor and F. R. Me Crumb; W.-A: Taylor & Co.

2274

COLORIMETRIC

2275

ment such as "acidify to pH 5.6" or "make alkaline to pH 9.4" has a very

definite meaning which cannot be expressed by the terms "slightly acid or
alkaline." .]
The values used in the pH scale are derived in the following manner. If
NjlO solutions of the .three acids, hy~rochloric, acetic and boric are titrated
under suitable conditions it is found that the total acidity is the same in all
cases. It is apparent, however, that these three acids differ radically in proper~
ties. Hydrochloric acid is a powerful acid whereas acetic and boric are relatively
weak."" Bases also show differences even when the total neutralizing ability is
the same. Solutions of sodium bicarbonate, sodium carbonate and sodium
hydroxide are entirely different even when the strengths are such that they'
will neutralize the same amount of acid per unit volume. These facts have
been known for years but it was not until the ionization theory was developed
.- that they could be explained. '.
. Q'he theory of ionization holds that certain types of chemical compounds are
_splIt apart or dissociated in aqueous solution, yielding electrically charged parti~
cles ,known as ions. These ions are responsible for many of the properties of
aqueous solutions. One general property is conduction of electric current.
When acids are dissociated they produce' hydrogen ions and negative ions,
HA::+H++ A-, and the acidic nature is determined by the extent to which such
dissociation occurs, acidity being due'to the hydrogen ions. ,Strictly speaking
free H .ions probably do not exist in aqueous solution as such but probably
combine with water to form HaO+. Practically this does not affect our assump~
tions. The differences among the three acids named above are due to the
concentrat)on or activity of hydrogen ions which they produce. In 0.1 N
'solutions these are roughly in the' proportions hydrochloric 15,000, acetic 200,
boric 1.
In a similar way bases produce hydroxyl ions, BOH::+B++OH-, and the
basic nature of an alkaline solution is determined by the extent to which such
dissociation occurs, alkalinity being due to the hydroxyl ions. For example
the ratios of the concentrations of hydroxyl ions in 0.1 N solutions of sodium hy~
droxide, sodium carbonate and sodium bicarbonate are roughly 16,667 : 667 : 1.
The fundamental reaction between acids and bases in aqueous solutions is
between hydrogen and hydroxyl ions to produce water, thus H++OH~HOH~
(H 20). Since pure water conducts a current to a very small degree this reaction
must be slightly reversible, i.e. even in pure water some hydrogen and hydroxyl
ions must exist. However this ionization is extremely slight only 1 molecule
of water in about 555,000,000 being dissociated. Using the symbols for
concentration, hydrogen ions [H+), hydroxyl ions [OH-) and undissociated
water [H 20), according to the law of chemical mass action, the product of
[H+) and [OH-] divided by [H 2 0] is constant, i.e. [H~~~-) . K. Since [H 2 0] is
extremely large as compared to [H+] and [OH-], it can be regarded as constant
and [H+] [OH-] = Kw, the ionization product of water, which at any definite
temperature is a constant. This constant, Kw, has been determined by a
variety of methods and found to be approximately 1/100,000,000,000,000 or
1/1014 or 10-14 at 25 C. The use of exponents such as 10-14 enables one to
readily handle extremely large or small numbers. This can be easily under-

2276

DETERMINATION OF pH

stood by recalling that

1
same as 103

lq X 10 X 10 is the same as

103 land

1~ X 1~ X l~

is the

By using a negative exponent simple f1actional terms can be

dIspensedWIt
h
1 = 10-3, etc.
, I.e. 10
3

Since in pure water the H ion and OH ion concentrations must be the same,
if [H+] [OH-] = 10-14 then [H+] = [OH-] = 10-7 Asolution in which [H+) = lOH-)
= 10-7 is neutral. Whenever [H+] is geater than [OH-] that is greatf'r than
10-7 at 25 C., the solution is acid; whenever [H+] is smaller than [OH-] that is
smaller than 10-7 at 25 C. the solution is alkaline. From the equation.
[H+] [OH-] = Kw it is seen that a simple relation always exists between [H+] ~
and [OH-], thus [H+] =

[~~-J.

It

i~ therefore unnecessary to refer to

OH ion

concentration at all and the term hydrogen ion concentration is used even when
referring to very alkaline solutions such as N sodium hydroxide, which contain
a vanishingly small number of hydrogen ions, the predominating ion being the
hydroxyl i;n'":""m
Expre~drogen ion concentrations as 1/10,000,000 or even as 10-7 is
extremely inconvenient. For this and other reasons, such 'as plotting curves,
etc., Sorensen introduced the method of expressing such values in terms of the
negative logarithm to the base 10. He called this the hydrogen exponent and
gave it the symbol pH. Thus pH is the negative logarithm of the hydrogen
ion concentration or the logarithm of the reciprocal of the hydrogen ion concentration, i.e. pH = -log [H+] = log [~+]' and [H+] = lO-pH Accordingly if in
a neutral solution [H+] = 10- 7, the pH will be 7.0.

Similarly pOH is the loga1

rithm of the .reciprocal of the hydroxyl ion concentration, pOH = log [OH-]
and pKw is the logarithm of Kl . If Kw= 10-14 ,
w

pR;~= 14.

The relationship

between hydrogen and hydroxyl ion is shown py recalling that if [H+] [OH-]
=Kw=10-14, then pH+pOH=pKw=I4, and pH=14-pOH. So in pure
water pH=pOH=7.0. In O.OIN Hel wh~re [H+] =10-2 , pH=2.0. In
O.01N NaOH where [OH-]=lO-2, pOH=2.0 and pH=14'_2=12.0.
The use of pH (negative logarithm) sometimes causes confusion at first. It
should always be remembered that a decrease in pH denotes an increase in
active acidity and an increase in pH denotes a decrease in active acidity (increase
in active alkalinity). In addition since pH is a logarithmic function, a solution
having pH 6.0 contains 10 times as many hydrogen ions as one with pH 7.0.
Similarly a solution with pH 5.0 contains 100 times, and a solution with pH 4.0,
1000 times as many hydrogen ions as one with pH 7.0.
A simple way to deal with hydrogen ion concentration is ip. terms of normality. If an acid solution contains 1 g. of available (ionizable) hydrogen
per liter it is said to be normal. Similarly a solution can be said to be normai
with respect to hydrogen ions when it contains 1 g. 9fionizeli' hydrogen per liter.
1
~Since [H+) is the reciprocal of the normality of liydrogen ions, pH can be defined

COLORIMETRIC
as the log:jlrithm of the denominator expressing such normality.

[H+] ~ 100' l?H =log 100 =2.0.

2277
Thus when

When [H+] = 10,000' pH =log 10,000 =4.0, etc.

The relationship between pH and hydrogen ion concentration is shown in the

table below. Expressing [H+] both as decimal fractions and as 10 with negative
exponents shows how pH is derived.
pH

H-Ion Concentration

1.0 .................................
0.1 ... ~ .............................
0.01 ................................
0.001 ............................ ~ ..
0.000,1. . . . . . . . . . . . . . . . . . . . . . . . . . . ..
0.000,01 ..........................' .. ,
0.000,001 ...................... : .... ,
0.000,000,1. ........................ ,
0.000,000,01. ....................... ,
0.000,000,001 ....................... ,
0.000,000,000,1 .......................
0.000,000,000,01. .................... ,
0.000,000,000,001. . . . . . . . . . . . . . . . . . . ..
0.000,000,000,000,1 ............. , .....
0.000,000,000,000,01 ................. ,

100
10-1
10-2
10-3
10-4
10-5
10-6
10-7
10-8
10- 9
10-10
10-11
10-11
10-13
10-14

0.0
1.0

2.0

3.0
4.0

5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0

Several concrete examples may serve to'make clear the exact meaning of pH.
Although a N /10 solution of hydrochloric acid contains 3.65 g. of HCI and 0.1 g.
of ionizable hydrogen per liter, only 84.0% (18 C.) is ionized (active). Recent
developments in the theory of ionization indicate that strong acids are com,pletely ionized and that the degree of ionization ordinarily assumed is only
apparent being due to interionic forces. Therefore the normality with respect
to ionized hydrogen is not N /10 but O.~~N or 11~0' Then pH = log 11.90 = 1.08.
. A N /10 solution of acetic acid contains the same amount of ionizable hydrogen as does N /10 HCI namely 0.1 g. but only 1.3% (18 C.) is ionized. The
. h respect to IOnIZe
. . d h y d rogen IS
. on Iy 0.013
N or 769
N an d pH IS
. Iog
-1-0normarIty WIt
769=2.89.
The apparent degree of dissociation of N /10 NaOH is 92.9% at 25 C. The
0:929
N = 10.76
N an d p OH'IS Iog
WIt
. h respect to OH"IOns IS t h erefore 1
norma1Ity
0
10.76 or 1.03.

pH=14-1.03=12.97.

Dissociation in

1~

ammonium hydroxide proceeds only to the extent of

1.33% at 25 C. Normality with respect to OH ions is therefore only

0.0133 N
N
10
or 752 and pOH=2.88. pH=14-2.88=11.12.
When certain salts are dissolved in water not only does ionization occur but
there is a tendency to react with the water as BA+HOH~HA+BOH. With
salts of strong acids and strong bases such as sodium chloride this tendency is
zero and such salts do not affect the pH of water. With salts of strong acids
and weak bases such as aluminum sulfate, ferric chloride, etc. the strong acid

2278

DETERMINATION OF.;pH

resulting dominates the weak base and the solution is acid. The reaction for
aluminum sulfate is Ab(S04)a+3H 2 0!::::j;3H 2S04+ AI(OH)a. With salts of weak
acids and strong bases, the strong base resulting dominates the weak acid
and the solution is' alkaline. The reaction for sodium carbonate is N a 2 CO a
+2H 20!::::j;H 2CO a+2NaOH. This general reaction is termed hydrolysis and
salts subject to it are for practical purposes either ac~ds or bases.
BUFFER ACTION

'"

~ In the case of a solution of a single pure acid or alkali, if the concentration

and degree of ionization (real or apparent) are known, the pH can be calculated
as shown above. In actual practice such a situation will rarely arise. Invariably solutions contain more th:;tn one substance. Frequently some 01
these substances form what is known as a buffer system and serve to modify
. and stabilize the pH.
When 1 ml. of 0.01 N Hel is added to 1 liter of pure water of pH 7.0, the pH
will drop to about pH 5.0. If however the same amount of .acid were added to a
solution containing a mixture of KH 2P0 4 and Na 2 HP0 4 of pH 7.0 the drop in
pH would be very small. This is illustrated by the curves on page 2279.
The phosphate mixture has the ability to resist change in pH on addition of acid
or alkali. It has what is called buffer acti6n and such a mixture constitutes a
buffer system or mixture. Such action is' encountered in numerous solutions
found in nature and in industrial processes. All mixtures of weak acids and
their salts or of weak bases and their salts are buffer mixtures. Some of the
most common buffer mixtures are acetic acid-acetate, carbonic acid (C0 2)bicarbonate, citric acid-citrate, mixed phosphates, etc. Other common buffering substances are proteins (gelatine, casein, etc.).
The mechanism of buffer action is most easily shown by means of the acetic
acid acetate system. In a solution of acetic acid dissociation occurs only to a
small extent as follows, HAc!::::j;H++Ac-. Sodium acetate however dissociates
practically completely, NaAc!::::j;Na++Ac- and yields. a large number of acetate
ions [Ac-]. A N /10 solution of acetic acid ha,s pH; 2.89 but a solution N /10
with respect to both acetic acid and sodium .acetate has pH 4.63. The pH
has been increased i.e. the H ion concentration has been lowered. The question arises as to what has become of the H ions. The mass law equation for
.
'd' [H+] [Ac-] K
h
K; h ,. . .
If
acetic aCl IS [HAc]
a, were a IS t e'lOmzatlOn~o~nstant.
anyone

:=

of the three concentrations be changed there must be a change -in the others
else Ka could not remain constant. In the case just cited the acetate ion concentration [Ac-] has been enormously increased so the H ion concentration
[H+} has to decrease. The unionized acetic acid will automatically bec'ome
greater so practically what occurs is that more inactive (unionized) acetic acid
is formed removing H ions from the solution. In tliis system the concentrations
of acetate ion [Ac-] and unionized acetic acid [HAd] are always extremely large
compared to the concentration of H.ions [H+] so that the ratio

[~l:i partially

determines the reaction or pH. This is important when dilution is necessary.

If small amounts of a strong acid (hydrochlotic}:be added ~o the system, the
large concentration of acetate ions quickly converts-most of the H ions added to

COLORIMETRIC

2279

unionized acetic acid and [H+] increases only. slightly. If a base (sodium
hydroxide) is added H ions are neutralized but more Ij-cetic acid is ionized
producing additional H ions to take the pla.ce of those neutr~lized and [H+J
decreases only slightly. Of course any buffer system has its limits and large
amounts of acid or base will overcome the buffer capacity. Each buffer
system has a rather narrow pH zone over which it functions. However, by
using different systems i.e. different mixtures, practiMlly the entire pH range
can be covered. This is extremely advantageous since it is possible to prepare
solutions of constant pH which serve as standards and it also serves as a means
of holding an industrial solution close to the d'ptimum pH. In the case of
fairly concentrated solutions of strong acids and bases, since the concentra,tiQn
of unionized substance is relatively small, there is no reserve supply of H or
OH ions: However there is such a large concentration of H or OH ions that
for practical purposes such systems function as buffer systems. For example
.when the concentration of either N /10 HCI or N /10 NaOH is reduced to NjlOO,
the pH is changed only about 1 unit.

II

10

a11

[7

114

___.--I---

12

Jr
10

,l I

ACID

.--

~V

[I

(~

J/:
r I clc i

&

ALKIALr

FIG. 311.

The stability of any buffer system will depend on total concentration, and
the ratio of salt to weak acid or base. Reasonable changes in concentration will
not materially change the pH as long as the ratio remains the same. In the
acetate system the ratio

[~l:l

does not change on moderate dilution and the

change in pH when the system N /10 acetic acid-N /10 sodium acetate is diluted
twenty fold is less than 0.1 pH. Many solutions met with in practice are
buffered. This is an advantage since colorimetric pH measurements can be
made on such solutions even when highly colored or turbid by diluting with
distilled water. The degree of dilution po~sible depends on the concentration
and the specific buffer system but the safest procedure is to dilute only as far
as necessary t? make sati~factory readings_;]
.
.
Buffer actIon can be Illustrated graphIcally by means of tItratIOn curves
(Fig. 311). Three such curves are shown in which various amounts of M/5
HCI and M/5 NaOH are added to 100 m!. samples of (A) distilled water, (B)
,

2280

DETERMINATION OF. pH

a buffer mixture composed of KH 2P0 4 and Na 2 HP()4 in which the total concentration is approximately 0.1 M, and (C) a similar buffer mixture in which
the total concentration is one half of B,all samples having an initial pH of 7.0.
The abscissa show the amounts of acid or alkali addea and the ordinates show
the pH of the respective mixtures.
I
The extremely steep curve (A) for distilled water shows that this substance is
devoid of buffer action since small amounts of acid or alkali change the pH
markedly. In the pronouncedly acid and alkaline ranges the curve flattens
out because one is then dealjng with relatively concentrated solutions of the
strong acid and base, where pH varies only moderately with concentration.
The curve (B) for the 0.1 M phosphate mixture differs radically from (A).
Addition of 11 mI. of acid only changes the :pH from 7.0 to 5.8, whereas th~
addition of 6 mI. of alkali only changes the pH from 7.0 to 7.8. Beyond these
points the change becomes more pronounced. Between 5.8 and 7.8 the curve
is relatively flat and in tliis zone the phosphate mixture shows strong buffer
action. As the curve indicates, buffer action is st.rongest near the middle of
the relatively fiat portion. It actually reaches a maximum at about 6.8 where
.
d
H
. [l'>ta2 H P0 4]
the concentratlOns of KH 2P0 4 an Na2 P0 4 are the same, I.e. [KH P0 ] = 1.
2

The steepness of the curve above and below the zone 5.8-7.8 irrdicates little
resistance to change in pH.
Curve (Q) resembles (B) in contour but the relatively fiat portion or zone of
effective buffer action is much shorter, extending only from about 6.5 to 7.5.
This shows the effect of concentration on buffer action.
These curves show that not all buffer mixtures have the same buffer capacity.
Van Slyke (J. BioI. Chern., 52,525 (1922)) .has formulated a method of expressjng the differences in terms of buffer capacity. A solution has a buffer capacity
of 1 when the'"reaction of a liter of it is changed one pH unit upon the addition of
one equivahmt of acid or base. A knowledge of the buffer capacity of a solution
is frequently of considerable value since it indicates ~ow easily the optimum pH
may be maintained in practice, and also what pfecautions are necessary in
determining the pH. It also provides a practical method of determining the
amount of acid or alkali required tb adjust the pH.
The requirements for buffer action should be clearly kept in mind. Mere
presence of salts does not ensure buffer action. Salts of ~trong acids and strong
bases impart no buffer action to a solution. Solutions which. contain only
salts of weak acids and weak bases may not be buffered. To ensure buffer
action a buffer system must be present i.e. both salt and weak acid or weak base
and the pH of the solution must be within the effective zone of the particular
system.
Tables showing the method of preparation fO.i" various buffer mixtures can
be found in Clark; The Determination of Hydro'gen Ions, 1928, Williams and
Wilkins, Baltimore, Md.
The following table gives some common buffer systems and the approximate
pH of maximum buffer capacity. The zone of effective buffer action will vary
with concentration but the general average wi!l:.be 1.0 pH from the value
given, for concentrations approximately 0.1 molar .

II

COLORIMETRIC

2281

Glycocoll-sodium chloride-hydrochloric acid. . . . . . . . . . . . . . . .. 2.0

Potassium acid phthalate-hydrochloric acid. . . . . . . . . . . . . . . . .. 2.8
Primary potassium citra,te ... " ......... '" . . . . . . . . . . . . . . .. 3.7
Acetic acid-sodium acetate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 4.6
Potassium Add Phthalate-Sodium Hydroxide. . . . . . . . . . . . . . .. 5.0
Secondary Sodium Citrate.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5.0
Carbonic Acid-Bicarbonate ...................... :. . . . . . . . . 6.5
Primary Phosphate-Secondary Phosphate. . . . . . . . . . . . . . . . . .. 6.8
Primary Phosphate-Sodium Hydroxide. '.' . . . . . . . . . . . . . . . . . .. 6.8
Boric Acid-Borax.. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . .. 8.5
Borax................ ...... .............................. 9.2
Boric Acid-Sodium Hydroxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 9.2
Bicarbonate-Carbonate .................................... 10.2
Secondary Phosphate-Sodium Hydroxide. . . . . . . . . . . . . . . . . . .. 11.5
THE USE OF INDICATORS

~ For making pH determinations substances kno,'n as indicators are required.

In indicator is a substance that gives different colors or different shades of color
at different pH values. Thus, phenol red changes from yellow at pH 6.8 to a
deep red at pH 8.4; that is, if phenol red is added to a solution having a pH of 6.8
it will give-this solution a yellow cvlor. If added to another solution of pH 7.0
it will give a somewhat reddeA,hade of yellow. If added to other solutions
having pH values of 7.2, 7.4, 7.6, etc. the amount of red color imparted to the
solution will increase progressively until at pH 8.4 the solution will have a deep
red color. Therefore if phenol red is added to any solution having a pH value
between 6.8 and 8.4, the pH of that solution is determined by observing the color
obtained. The method for. accurately reading the pH value is explained later
in this section.
Phenol red. covers only a limited part of the pH scale but other indicators are
available for other parts. Thus bromcresol green covers the range pH 3.8-5,4
and its color change is from yellow at pH 3.8 to blue at pH 5.4. Thymol blue
covers the range pH '8.0-9.6 and its color change is from yellow at pH 8.0 to
blue at pH 9.6. A complete list of indicators which cover the pH scale 0.2 to
13.6, with their pH ranges and color changes, is given in the table.

1:

Name

Acid cresol red ......................... .

Acid meta cresol purple ................. .
Thymol blue ........................... .
Benzo yellow .......................... .
Bromphenol blue ....................... .
Bromcresol green .........._............. .
Methyl red ............................ .
Chlorphenol red ........................ .
Bromcresol purple ...................... .
Bromthymol blue ...................... .
Phenol red ............................. .
Cresol red ............................. .
Meta cresol purple ..................... .
Thymol blue .............. , ............ .
Phthalein red .......................... .
Tolyl red .............................. .
Parazo orange ......................... .
Acyl blue ...............................

pH Range

Color Change

0.2- 1.8
1.2- 2.8
1.2- 2.8
2.4- 4.0
3.0- 4.6
3.8- 5.4
4.4- 6.0
5.2- 6.8
5.2- 6.8
6.0- 7.6
6.8- 8.4
7.2- 8.8
7.6- 9.2
8.0- 9.6
8.6-10.2
10.0-11.6
11.0-12.6
12.0-13.6

Red-yellow
Red-yellow
Red-yellow
Red-yellow
Yellow-blue
Yellow-blue
Red-yellow
Yellow-red
Yellow-purple
Yellow-blue
Yellow-red
Yellow-red
Yellow-purple
Yellow-blue
Yellow-red
Red-yellow
Yellow-orange
Red-blue

2282

DETERMINATION OF pH

Indicators are weak acids or bases in which. the ionized form differs in constitution and color from the normal form: An indicator acid will dissociate as
HIn=}H++In-, HIn being the acid form and In- the alkaline form. The mass
.
'11 b [H+] [In-] K
Th
[In-] . tKr
S
K'
Iaw expreSSIOn
WI e [HIn] = I
en [HIn] = [H+]' . mce I IS a constant the ratio

i~~~]

will depend on the value of [H+]; the hydrogen i(;m eon-

cent ration. This ratio determines the color that the indicator will show. For
example with phenol red the acid form [HIn] is yellow and the alkaline form
[In-] is red, consequently as [In-] increases the color changes from yellow
through orange to red. When HIn is half transformed [HIn] = [In-] and so
[H+j = K I. When the hydrogen ion concentration is the same as the ionization"
constant of the indicator, the concentrations of the acid and alkaline forms of
the indicator are equal. TP.e value of Kr varies for different indicators and
determines the range of hydrogen ion concentration over which each indicator
1
functions. If [H+]=KI then pH=pKI, where pKI=log KI' While this determines the theoretical midpoint of the pH range of each indicator,_ the actual
range used in practice depends on how readil~he colors can be distinguished
by the eye. Consequently the practical range is determined experimentally.
While pKI is_a function of pH, it is also affected to some extent by other factors,
such as salt concentration, proteins and temperature.
A satisfactory indicator for practical pH work must have certain well defined
qualifications. It must give well-defined color changes over a relatively short
range; it must not be unduly affected by substances other than hydrogen or
hydroxyl ions; it must immediately give colors that are stable for a reasonable
length of time, so that there will be no errors due to changes taking place during
the determination; its solution must be a stable reagent. An indicator that
fades rapidly, precipitates, undergoes change in color due to the presence of
salts, colloidal materials, etc., attains a stable color slowly, or is unstable as a
reagent, is certain to prove objectionable in practice. I
The essential requirements in applying the colorimetric method for the
determination of pH are (1) An accurately prepared solution of indicator,
(2) a set of color standards prepared from measured amounts of buffer solutions
having accurately adjusted pH values and containing accurately measured
amounts of the indicator solution. Standards in intervals "Of .0.2 pH are usually
employed. Such a set for phenol red will consist of standards with the pH
values 6.8,7.0,7.2,7.4,7.6,7.8,8.0,8.2 and 8.4. These standards. can be placed
in test tubes or ampoules having uniform bores. Ampoules are more satisfactory since they can be sealed. If a measured volume of clear sample is
treated with the same amouilt of indicator as is present in the standards, the.
resulting color will depend on the pH of the sample and by comparing this
color with the color standards the pH can be determined. It is essential that
all measurements be accurate and that the bore of the color standards and test
tube used for the sample be practically the same. The intensity of color and
depth of liquid in both cases should be uniform.
l
If the materials being tested are absolutely cle.a:r:.-and colorless accurate determinations can be lJlade by comparing the color of tne test sample with those of

COLORIMETRIC

2:.l83

the color standards, as outlined above. Most solutions encountered in actual

practice are however somewhat turbid or colored or both. Therefore when an
indicator solution is added to such a material the resulting color will not match
with the color standards which contain none of the color or turbidity of the
sample. For this reason some sort of comparator is required. In fact all
determinations should be made with a comparator since the presence of even
minute quantities of color or turbidity may cause considerable error or even
prevent a match all together when an attempt is made to match the sample
directly with the color standards. The procedure employed in all comparators
is essentially the same so the Slide Comparator (see page 2283) will be used to
illustrate the principle.
In making a determination three of the test tubes are filled to the mark
(5 m!.) with the material to be tested and 0.5 m!. of the indicator solution is
added to the middle tube. The color standards are then moved back and forth
in front of the test sample, always making sure that, when a comparison is made,
one of the ampoules of distilled water is in front.of the middle test tube containing. the indicator. Two consecutive color standards will then be directly in
front of the two tubes of test sample containing no indicator. The arrangement of the three tubes of test sample, the two color standards and the ampoule
of distilled water is illustrated in figure 312 below.
AA, BB, CC represent the three slots in the base and any three corresponding slots in the color standard slide. If we consider the color standards as
consisting of distilled water and indicator only, which is permissible since they
contain no turbidity and no color except that given them by the indicator, it is
clear that on looking through slots AA, BB, and CC, one is looking through
exactly the same materials-sample, indicator, distilled water-in each case.
This arrangement therefore eliminates any effect of color or turbidity in t4e
sample so that, when the proper color standard~ are in place, the color of the
sample will match one of the color standards or lie
between the colors of two consecutive standards.
ABC
Readings should never be made by matching
against either end standard in a given set. Thus
a sample may match the bromthymol blue color
standard marked pH 6.0 and actually have a pH
value of 5.0, 4.2, etc. If a match with the
standard marked 6.0 is obtained the determination should be repeated using an indicator coverA
6
C
ing a lower part of the pH scale, for example,
chlorphenol red (pH 5.2-6.8), bromcresol green
FIG. 312.
(pH 3.8-5.4), etc., until a 'match is obtained
with a standard other than the end one. Similarly if a sample matches the
bromthymol blue standard marked 7.6, the test should be repeated using phenol
red (pH 6.8-8.4) or a higher indicator if necessary.
.
The table on page 2281 shows that the ranges of the various indicators overlap. Thus the values pH 6.0-6.8 are common to both chlorphenol red and
bromthymol blue, the values pH 6.8-7.6 are common to both bromthymol blue
and phenol red, etc. It is,therefore usually possible to make a test on any given
sample with two different indicators and thus check results. This is one of the
outstanding advantages of using several indicators to cover a wide pH range

2284

DETERMINATION OF pH
I

instead of a single wide range indicator, such as the so c,alled universal or utility
indicators. "With the single test wide range indicator there is no way of checking results and marked errors may be made with no means of detec~ing them.
\

\
COLORIMETRIC DETERMINATION OF BUFFERED MATERIALS

One of the most iinportant factors to be considereCl when the colorimatric

method is applied is the degree to which the material being tested is buffered.
This is especially important when the pH of colored and turbid materials must
be determined. When a highly colored or turbid sample is buffered it can
frequently be diluted with distilled water to a point where satisfactory readings
can be made with the comparator method without introducing appreciabl!\
error. For example, sewage sludges have been accurately tested when a dilution of 1 : 49 was necessary.2 The presence of buffer systems in the sludge
makes this possible.
In applying the dilution procedure to any substance it is advisable to continue the dilution beyond that required for satisfactory readings so as to determine the point at which marked change starts to take place. This will furnish
information on the buffer capacity of the substance being tested. In all cases
the distilled water used should be of excellent quality and it should have a pH
of 6.4 to 6.8.
It is usually advisable to employ dilu~ion and make the tests on the supernatant liquid after any solids have partially settled. Filtration should be
avoided unless it is definitely known that changes in pH will not result. During filtration there may be a gain or loss in carbon dioxide or an action by the
filtering medium (paper, glass wool, etc.) which will introduce. errors.
UNBUFFERED OR SLIGHTLY BUFFERED MATERIALS

Some materials encountered in practice are not very highly buffered and a
few are unbuffered. In such cases care must be exerci,sed in determining the pH
since such solutions are very susceptible to change. Among such materials are
most waters, soil extracts, paper extracts, white 'water, pure sugar liquors, flotation feeds, laundry rinses, clay filtrates, etc.
Distilled water is devoid of buffer action and special precautions must be
observed in testing it. Absolutely pure water has pH 7.0....at 25 C. Ordinary
distilled water is usually 'acid, due to absorption of carbon -dioxide. Distilled
water in equilibrium'with the carbon dioxide of the air will have pH 5.7. It
may be supersaturated with carbon dioxide to give a pH as low as 5.0. Carbon
dioxide may be removed by boiling in Pyrex ve~sels to give water with a pH of
6.6 to 6.8 but superpure water with pH 7.0 can best be secured by means of a
special still. Distilled water should always be kept in Pyrex or tip.ned vessels
and contact with air held at a minimum. In making aqueous solutions or
extractions of slightly buffered materials, proportions of material and water
should be held constant so that results on different samples will be comparable.
In determining the pH of buffered solutions reasonable differences between
the pH of the sample and the indicator solution. will not' affect the results.
_.2 McCrumb, Sewage Works .:!ournal, 1, 534 (1929).
-

COLORIMETRIC

2285

However, in the case of unbuffered solutions the pH of the indicator solution

must be fairly close to the pH of the sample or the indicator will change the pH
(Acree and Fawcett, Ind. Eng. Chem. Anal. Ed., 2, 78 (1930)). For this reason
the pH of the indicator solution should always be at or near its own midpoint.
For example bromcresol green, 4.8, bromthymol blue, 6.8, cresol red, 8.0, etc.
This is why it is impossible to secure accurate results on slightly buffered materials when a single wide range indicator (universal, utility, etc.) is employed.
It has been shown that errors as high as 3.0 pH may result (McCrumb, Ind.
Eng. Chem. Anal. Ed., 3, 233 (1931). This is because the pH of the wide
range indicator may vary so :(Iluch from that of the sample.
Where extremely accurate results are desired on unbuffered materials, it'
may be advisable to use several solutions of the same indicator adjusted to
different points on its range, such as bromthymol blue at 6.2, 6.8, 7.4. If the
unknown is devoid of buffer action as in the case of distilled water, for example,
three values may result such, as 6.6, 6.8, and 7.0 respectively. Since the reading
6.8 coincides _with the pH of the indicator solution used for that test, the actual
pH~s 6.8.
An indicator solutions should be adjusted to their respective midpoints and
kept in Pyrex bottles sealed with rubber stoppers. Soft glass bottles and cork
stoppers should never be used or marked changes in pH may occur. Indicator
solutions should not be unduly exposed to the air (poor seals, etc.) since carbon
dioxide may lower the pH. In case changes occur the pH can be again adjusted
by adding dilute
/10) acid or alkali as required comparing the indicator
solution in a; special pair of tubes with the appropriate standard. If, however,
buffered solutions are being tested, only gross changes in the pH of indicator
solutions need be considered.
In making a pH determination on a slightly buffered solution the indicator
solution should be placed in the test tube and the solution being tested run in
from a pipette holding the tip of the pipette near the bottom of the tube.
Mixing is then secured by stirring either with the tip of the pipette or with a
thin glass rod. Mixing should never be done by placing finger or thumb over
the mouth of the tube. The recommended procedure prevents introduction of
impurities such as carbon dioxide, etc. and gives accurate results. This is
really the only way unbuffered materials can be tested since electro metric
methods are not reliable on very slightly buffered solutions.

eN

THE EFFECT OF SALTS

As stated previously the color of any indicator may be influenced by the

actual salt concentration as well as the pH. This effect is commonly known as
the" salt error."
By varying the concentration of a buffer mixture and determining the pH
both by a hydrogen electrode and the colorimetric method it will be found that
absolute agreement occurs only at the concentration of buffer employed in
making the standard. Above this concentration the Mlorimetric method will
give results that are slightly higher than those secured by the hydrogen electrode and below this concentration the colorimetric results will be slightly lower.
This is illustrated by the results given in the following table.

DETERMINATION OF pH,

2286

Effect of Varying the Concentration, of ,a Buffer Mixture

t
Molar Cone,

Hydrogen Electrode
pH

Colorimetric
pH

0.200
0.100
0.050
0.020
0.010
0.005

6.33
6.42
6.50
6.57
6.62
6.68

6.40
6.45
6.50
6.55
6.55
6.55

~H

I
I

Difference
Hyd. Elect. pH Color.

-0.07
-0.03
0.00
+0.02
+0.07
+0.13

It will be noted that absolute agreement occurs only at a molar con;entration of 0.05, the one employed in making the color standards. Of course none
of the difference are pronounced.
The effect of sodium chloride is shown by the data in the following table
secured by adding variable amounts of sodium chloride to a phosphate buffer
mixture having the same cpncentration as was employed in the preparation of
the color standa,rds.
Effect of Sodium Chloride
/

Sodium Chloride
Molar Conc.

Hydrogen Electrode
pH

Colori~etric
pH

Differen('e
pH Hyd. Elect. -pH Color.

0.00
0.05
0.10
0.20
0.50
1.00

6.40
6.25
6.23
6.15
6.00
5.90

6.40
6.40
6.40
6.35
6.25
6.15

0.00
-0.15
-0.17
-0.20
-0.25
-0.25
/

Similar data secured by adding magnesium chloride to the same phosphate

buffer mixture show the influence of a divalent salt.
Effect of Magnesium Chloride
Magnesium Chloride
Molar Conc.

Hydrogen Electrode
pH

0.00
0.05
0.10
0.20
0.50

6.40
5.70
5.40
5.20
4.70

6.40'
5.90
5.80
5.65
5.26

. Difference
pH Hyd. Elect. pH Color.

Colorimetric
pH

0.00
-0.20
-'0.,40
-0:45
-0.56

The indicators used in making the colorimetric lI).~asurem(mts given in the

three tables above were bromthymol blue, chlorphenol red, para nitrophenol,
and bromcresol green.
-

COLORIMETRIC

2287

Additional information on salt effect can be secured by consulting Clark,3

Kolthoff,4 Guntelberg and Schiodt,o Send roy and Hastings,6 McCrumb and
Kenny,7 and Blum and Bekkedah1. 8 The last paper deals exclusively with
nickel electroplating and electrotyping solutions.
The investigations of Guntelberg and Schiodt and Hastings and Sendroy
would indicate that in the case of the indicators bromphenol blue, bromcresol
green, bromcresol purple and phenol red, when the effects of salts are correlated
with ionic strengths, salts of different valence. types show at the same ionic
strengths approximately like effects on the apparent dissociation constants of
the indicators.
The ionic strength of a solution is obtained by multiplying the concentratiqn
of each ion by the square of that ion's valence number, summing these products
and dividing by two. In general the buffer solutions used as standards in the
colorimetric method have ionic strengths of 0.05 to 0.1.
In the majority of cases the differences encountered in industrial practice
may have little significance. The. correction can be determined and applied
for each case but frequently it can be disregarded. For example in the case
of a nickel plating bath it may make little difference to the plater whether the
bath is controlled at pH 5.8 colorimetric ally or 5.4 electrometrically. However, to avoid confusion when comparing procedures, the method employed
should be given.

THE EFFECT OF PROTEINS

Systematic calibration when proteins are encountered will prove to be much

more difficult than in the case of salts. Salts are at least substances of a
definite nature, whereas the term protein is used in a very loose manner. For
. this reason so-called "protein errors" of indicators should never be applied
unless it is known definitely that these variations were determined on solutions
that are practically identical with the one being tested.
For example many "protein errors" have been determined on materials
like beef broth or some type of bacteriological media. On many of these
materials the proteins have undergone considerable hydrolysis, or they are
rather dilute. Anyone taking the variations so determined and using them in
the measurement of the pH, of say gelatine or casein, might quickly get into
trouble.
The following table shows some variations encountered on a 1 % so.lution
of gelatine to which variable amounts of sulfuric acid and sodium hydroxide had
been added. The gelatine was ash free with a pH of 4.80 (isoelectric). Sulfuric acid was used because this acid is most likely to be encountered in practice.
Clark, "The Determination of Hydrogen Ions," 1928, Williams and Wilkins.
Kolthoff, "The Colorimetric and Potentiometric Determination of pH," 1931,
John Wiley and Sons, Inc.
Guntelberg and Schiodt, Z. Physik Chern., 135, 393 (1928).
6 Sendroy and,Hastings, J. BioI. Chern., 82,200 (1929).
, McCrumb and Kenny, J. Soc. Chern. Ind., 49, 425T, 427T (1930).
8 Blum and Bekkedahl, Trans. Electro Chern. Soc., 56, 291 (1929).
,
3

2288

'DETERMINATION'OF pH
I

Effect of Geiatille
'.

Hydrogen Electrode
pH

Colorimetric
I'll

Difference
pH Hyd, Elect. -pH Colo1

2.48
2.79
3.19
3.51
3.65
3.88
4.47
4.80
5.07
6.16
6.59
6.80
7.39
7.78

2.40
3.20
3.55
3.85
4.00
4.20
4.60
4.85
5.00
6.20
6.65
7.20
7.90
8.50

0.08
-0.41
-0.36
:'-0.34
-0.35
-0.32
-0.13
-0.05
0.07
-0.04
-0.06
-0.40
-0.51
-0.72

Indicator

I Meta cresol purple

I

Benzo yellow
Benzo yellow
Bromcresol green
Bromcresol green
Bromcresol green
Bromcresol green
Bromcresol green
Bromcresol green
Chlorphenol red
Bromthymol blue
Phenol red
Phenol red-Cresol red
Thymol blue

It will be noted that the differences encountered near the isoelectric point
are insignificant being within the sensitivity of the colorimetric method employed. When hydrochloric acid is used the, differences are greater in the acid
range.
The variations between pH measurements by the hydrogen electrode and the
colorimetric method in the precipitation of casein by hydrochloric acid have
been determined by Clark and others 9 using methyl red; and by Miss Benton 10
using bromcresol green.
_EFFECT OF TEMPERATURE

Since the pH of a solution depends on the degree of ionization of its constituents and this may be affected by temperature, the'influence of temperature
on pH is an important factor. Irrespective of the ~ethod of measurement
employed determinations must be made at a fixed temperature. In the majority of cases a temperature of 25 C. is the most satisfactory. although variations in temperature from 20 C. to 30 C. are u~ually insignificant for practical
purposes.
Certain buffer systems are more sensitive to temperature variations than
others. This is particularly true of the borate and carbonate.series. Solutions
which depend on a carbonate equilibrium are always susceIjltib1e-to changes in
temperature.
In the case of nickel solutions not only does the pH change with temperature
but the color of the solution itself may change. 'Consequently the blanks used
in the comparator method should be at the same temperature as the sample to
which indicator has been added..
(
The sulfone phthalein indicators are not influenced by temperature changes
to a marked degree but methyl orange and indicators resembling methyl orange
are very susceptible to such changes. The indicators LaMotte Yellow, Hellige
,
Orange and Benzo Yellow belong to this class.
9 Clark, Zoller, Dahlberg and Weimar, Ind. Eng. Ch~Iji., 12, 1163 '(1920).
10

Benton, Ind. Eng. Chern., 20, 15 (1928).

COLORIMETRIC

2289

SPECIFIC EFFECTS

In the case of certain sulfonephthaleins the shade of color depends on concentration and depth of layer. Dichromatism is most noticeable iIi practice
with such indicators as bromphenol blue and bromcresol purple. The transmitted light in such cases is primarily red and blue and the ratio of red and blue
will vary with the depth of layer. In testing turbid solutions much of the light
is reflected by the particles and this gives the same effect as a thin layer of
liquid and this tends to modify the color. In such cases it is advisable to
substitute another indicator.
Off color, while not exactly common to all indicators may occur in the case
of most two color indicators under certain conditions. Suspensions, emulsions~
and solutions of a colloidal nature may cause this difficulty. Concentrated
solutions of such salts as zinc chloride and sulfate frequently give off colors
with the sulfonephthaleins. Sometimes on standing a precipitate, presumably
of basic salts, will settle out and the colors then secured' on the supernatant
liquid may then correspond with the color standards. In such cases the original
solution may appear clear to the eye, but it must contain dispersed solids which
cause the off color of the indicator. When dealing with such solutions the
nitrophenols may be used. When tubes and standards are observed through a
blue glass satisfactory color distinctions can usually be secured.
The nitro phenols ordinarily employed for this purpose are 2.4 Dinitrophenol.
(Alpha) (pH 2.8-4.4), 2.5 Dinitrophenol (Gamma) (pH 4.2-5.8), para nitrophenol (pH 5.4-7.0) and meta nitro phenol (pH 7.2-8.8). The color changes
are all from practically colorless to varying shades of yellow.
For further details Michaelis 'and Gyemant 11 should be consulted.
Difficulty is sometimes encountered in testing soap solutions. In fairly
dilute solutions the results may be in good agreement with the hydrogen electrode but certain indicators tend to give low results as the concentration of soap
is increased. This is particularly true of thymol blue.
4 Some indicators are apparently influenced by certain ions in some way not
as yet explainable. Alizarine Yellow G (pH 10.0-12,0), Alizarine Yellow R,
(pH 10.0...,12.0), nitro yellow (pH 10.0-11.6) tend to give high results in the
presence of calcium salts.
Michaelis 12 has demonstrated the precipitation of bromthymol blue in a
solution of quinine hydrochloride.
However, not all the sulfonephthaleins are precipitated by quinine as
Michaelis appears to think. Chlorphenol red gives satisfactory readings in
such cases. None of the nitro phenols are precipitated by quinine. Precipitation of bromthymol blue also takes place in solutions used as local anaesthetics
in dentistry.
,
Ther~ are a few rare instances in which bromthymol blue gives results that
are slightly lower than those found by other sulfonephthaleins. This has been
pointed out by Sharp and McInerney'I3 in the testing of milk.
11 Michaelis and Gyemant, Biochem. Z., 109, 165 (1920) (see Clark, "The Determination of Hydrogen Ions,:' 1928, p. 128).
12 Michaelis, "Practical Physical and Colloid Chemistry," Trans. by Parsons, 1925,
p.41.
13 Sharp and McInerney, J. BioI. Chern., 70, 729 (1926).

2290

DETERMINATION OF p'H

Similar results are som~times encountered. in testing concentrated solutions

of sulfonated oils.
j
.

EFFECT OF CHLORtNE

The action of chlorine must sometimes be taken intb consideration in water

treatment plants, paper mills, textile mills, etc., whenevE;)r" the dosage of chlorine
is high enough to affect the indicators. When such action occurs the results
secured are influenced by the particular indicators in use, ~he concentration of
indicator employed, the buffer capacity of the solution being tested, the tsmperature and, of course, the concentration of available chlorine.
In the procedure employed with the Slide Comparator (Fig. 313) the danger
point with the sulfonephthaleins is about 2.0 parts per million of active chlorini"
Whenever lower concentrations of indicator are employed as in Nessler tubes,
etc., the action of chlorine may cause errors at lower concentrations of chlorine.
There appears to be two types of reaction, chlorination apd oxidation. In
the case of chlorphenol red and phenol red chlorination appears to predominate
at first and the chlorinated dyes which result have lower pH ranges than the
original dyes. This results in an increase in the alkaline color which leads. to
results that are higher than they should be. Of course on longer standing if
the concentration of chlorine is high enough fading will occur due to destruction
of the dye by oxidation. In the case of the other sulfone phthalein dyes the '
action is largely destructive and the reagings tend to be low or off colors are
secured.
The nitro phenols are more stable in the presence of active chlorine and can
sometimes be used when the concentration of chlorine is as high as 200 parts per
million. 14
One disadvantage of the nitro phenols should be pointed out here. They
are all fairly acidic and their solutions are difficult to stabilize at any definite pH.
Consequently they are more liable to give erroneous results in testing unbuffered samples than are the sulfonephthaleins . .
I

FIG. 313.
14

_,-

Lewis and Kukolich, Paper Trade Journal, 95, No. 11, p. 28 (Sept. 15,

193~).

COLORIMETRIC

2291

COLORIMETRIC pH EQUIPMENT
Colorimetric pH equipment can be
made by the worker if desired using buffer
mixtures in test tubes or ampoules but it is
more satisfactory to purchase such equipment. Several types are available and
the outstanding ones will now be described
in some detail. Fuller descriptions can
be secured from the manufacturers.
THE TAYLOR SLIDE COMPARATOR

The Slide Comparator consists of two

principal parts, the slide and the base, as
shown in the cut below. The Slide is a
Bakelite case 10 inches long, 2%, inches
high and % inch thick. It contains 17
vertical holes and 17 horizontal slots
whi~h pass through the exact centers of
the holes. . In these holes are placed the 9
color standards for any given indicator
and 8 ampoules of distilled water, the
color -standards alternating with the
ampoules of distilleq water. All these
ampoules are held in place by a lid which
is screwed On the top of the slide.
The base consists of two parts'. The
lower part contains a slot for the slide
fitted with a key to stop the slide at the
end standards, two holes containing vials
of indicator solution, with 0.5 m!. pipettes
and ~nipples, five holes containing test
tubes, and a closed compartment for a
ground glass plate. Horizontal slots run
through the three central holes in the base
holding the test tubes, these slots corresponding exactly with any three of the
slots in the slide. The upper part of the
base serves as a cover for the vials and test
tubes when the set is. not in use. It is
fastened to the lower part- by means of
spring catches. The complete set, including the slide, is 10 inches long, 2,Ys
inches wide and 4 inches high, and weighs
only 2 pounds.
Making a pH determination consists
in only three simple operations.
1. After removing the top of the base,
three of the test tubes are placed in the
holes back of the slots in the base and
filled to the mark (5 m!.) with the sample
to be tested.

FIG. 314.

FIG. 315.

FIG. 316.

2292

DETERMINATION OF pII

2. To the central tube 0.5 ml. of 'the indicator solution is added, by means of
the pipette and nipple, and the contents are thoroughly 'mixed.
3. The slide containing the color standards is now placed in position on the
base and, holding the instrument toward a window or other source of daylight,
the slide is moved back and forth in front of the test samples until a color match
is obtained. Tlie pH is then read off directly from the yalues on the front of
the slide.
Since all color standard slides are interchangeable, that is theyl can be used
with one ~ase, the range of a single slide Comparator can be increased by purchasing additional slides and a supply of the corresponding indicator solutions,
with pipettes and nipples. For, ranges longer than that of one indicator,
however, the Long Range and Dalite Slide Comparators are recommended.,
All color standards provided for the various models of slide comparators are
guaranteed to retain their accuracy for a period of five years.
THE LONG RANGE SLIDE COMPARATOR

This set is designed to meet the needs of workers who require a portaple
set covering a wider range. It is made in six models covering the range of any
3, 4, 5, 6, 7, or 8 indicators respectively between the limits pH
0.2 and pH 13.6. It is flexible
since its range can be increased
simply by. securing additional
color standard slides.
Each Long Range Slide Comparator contains one complete
Slide Comparator; 2 to 7 extra
color standard slides; vials of
the corresponding indicator solutions, ;rith 0.5 ml. pipettes and
nipples;) and 14-5 m!. test tubes.
All models are contained in a polished wooden case 11;\1 inches
long, 9incheshigh and 5%, inches
wide, with a handle for carrying.
FIG,317.
A
leqUlpm.en
l'
t necessary f or
~
making tests is contained in the case so that tltese sets can' be-used in field,
laboratory or plant. The total weight of the set varies from 7}'2 lbs. to 9;\1
Ibs. depending on the number of slides.
The method for making determinations is exactly the same as that described
for the single Slide Comparator.
I

DALITE SLIDE COMPARATdR

In many cases it is necessary to make pH determinations at night. Also in

many plants even during the day time the light is not suitable for this type of
work. The use of ordinary electric light bulbs and even t~e so-called dlilite
bulbs is not satisfactory, The Dalite Slide ComPEator enables the worker
not only to make determinations at night with just' as great accuracy as during
the day but also assures him that measurements made at night will check those
made on the same materials by ordinary daylight.

COLORIMETRIC

2293

The set is made in 6 models to cover the range of any 3, 4, 5, 6, 7 or 8 indicators respectively between the limits of pH 0.2 and pH 13.6 in intervals of 0.2
pH. The set is flexible since its range can be increased simply by purchasing
additional color standard slides and vials of the corresponding indicator solution.

FIG. 318.

The pH equipment supplied with this set consists of one complete Slide
Comparator; any 2 t o 7 additional color standard slides; vials of the corresponding indicator solutions with pipettes and nipples; and 10-5 ml. t est tubes.
Each slide contains 9 color standards for any given indicator and these slides
fit in slots in the front of the case as shown in the illustrll,tion. Each vial of
indicator solution is placed directly above the corresponding color standard
slide, thus eliminating error due to the use of the wrong indicator solutions.
A top protects the vials when the set is not in use.
The comparator base is supported by a shelf on the front of the box. Directly behind the slots in the base is a piece of Dalite glass and in the compartment behind this is a special 25 watt electric light bulb, wit h cord, switch and
socket. In the lower part of the box is a cupboard of sufficient size to hold four
1000 ml. Pyrex bottles. This cupboard serves not only as storage for stock
bottles of indicator solution but also to bring the comparator to a convenient
height for making readings. A rack on the inside of the door holds 10-5 ml.
t est tubes. A test tube brush is also supplied with the set.
In making determinations three of the 5 ml. test tubes are filled to the mark
with the solution to be tested, the proper color standard slide is placed on the
base and 0.5 ml. of the corresponding indicator solut ion is added to the middle
tube by means of the pipette and nipple. The light is then turned on and the

2294

DETERMINATION OF pH

slide is moved back and forth until a match is obtained with one of the color,
standards. The pH value is then read off directlyfrOIp. the values on the slide.
THE ROU~ETTE GO~PARATO'.
.
'
The Roulette Comparator consIsts essentIally of a statlOnary base and metal
band, and a wooden drum which revolves inside the metal band, on ball bearings,
A block, which contains three slots, and which is drilled to hold three graduated
test tubes, is fastened to the front of the metal band. A 40 watt lamp is fixed
in the center of the base and is connected with a standard plug by me:ms'of a
lamp cord, a piece of Dalite glass is placed in the back of the block between the
three test tubes and the color standards, and a piece of etched glass is placed

...

FIG, 319, LaMotte Roulette Comparator,

in a slot in the block directly in front of the three tst tubes. The block also
serves as a visor to cut-out light from the outside. With this arrangement the
observer always makes his readings under standard, diffused daylight conditions.
Any three sets of color standards' such as chlorphenol red (pH 5.2-6.8),
bromthymol blue (pH 6.0-7.6) and phenol red (pH 6.8-8.4) and 24 ampoules of
distilled water, are supplied with the set. An extra polished wooden case,
containing 50 ml. bottles of the corresponding indicator solutions, 1 empty
50 ml. bottle, 4-0.5 mL pipettes with nipples, and 18 marked. test tubes, is
included as part of the equipment. A fabricoid cover is provided for the
comparator.
The revolving drum contains three separate sets of holes, each set being
designed to hold 9 color standards and 8 ampouies of distilled water. The
color standards are placed in position in order alte~nating with t,he ampoules of
distilled water. To make a determination three test tubes are filled with the
solution to be teste!l and placed in the three holes in the block. Then 0,5 ml.
of indicator solution is added to the middle tube and mixed thoroughly. The
light is turned on and the drum revolved untiJ .the: appropriate standards are
behind the test sample. The standards are then snifted by turning the drum

COLORIMETRIC

2295

until the color through the middle tube exactly matches that Of one of the
standards on either side of it or lies between them. The pH value is then
read off directly from the color standards as the labels on the standards project
above the top of the drum. In making a reading the drum must be turned 130
that an ampoule of distilled water is always directly behind the central test
tube.
There is only room for three sets of color stf\lldards. Consequently, if the
pH of the samples falls outside of the range of these three indicators, another
set must be substituted for one of the sets present.
The individual color standards are contained in ampoules approximately
125 mm.long having an outside diamet\ilr of 15 mm. The test tubes are made to .
match the standards and are graduated at 10 ml. The color standards are
guaranteed for one year.

FIG. 320.
THE HEl;LIGE COMPARATOR

The lIellige Comparator (standard model) consists of a Bakelite housing

with hinged front and rear covers and a dust proof enclosed prism for bringing
the color fields into juxtaposition. Two acid-proof cemented cells with plane
and parallel walls are provided. One cell is to receivtl the unknown solution

2296

DETERMINATION OF pH,

with indicator, the other to receive the test solution without indicator. This
arrangement provides for color comparison with coloredlor turbid liquids as
~
well as clear solutions.
. In addition to the prism and cells the comparator is sup,l)lied with four round
measuring tubes with graduations from 5 to 10 ml. in 1/1 ml. for ,measuring the
test solution, 1 pipette with graduations 0.20,0.25,0.5 and ;t.O ml. for measuring
the indicator solution, 1 opal glass plate, 50 ml. of indicator solution and 1
color disc.
Each color disc consists of a molded resin circular frame with a numbe.r
(usually 9) of colored glass standards each one representing 0.2 pH so arranged
that they can be revolved in the comparator. The color glass standards are
made from variable layers of colored glass held together to simulate the"
colors resulting when buffer solutions of known pH are treated with the re- '
spective indicator. The value of each standard is placed on the disc so that
it is visible in a special opening when that particular standard is in the reading
position. Only one color standard is visible at a time. Separate color discs
are provided for the various indicators.
An illuminator with stand can also be secured to provide the proper artificial
illumination.

ELECTRO METRIC HYDROGEN

ION MEASUREMENTSl
Electrometric hydrogen ion (pH) measurements are made with a suitable
electrode system as the detecting unit and a potentiometer as the measuring
ul).it. The electrodes develop a potential (e.m~f.) proportional to the pH of
the solution being measured. The potentiometer measures e.m.f.
POTENTIOMETRIC PRINCIPLE

In order to determine the pH from the difference in potential of two electrodes, the voltage must be measured without allowing any appreciable current
of electr.i.city to flow from the cell, for with current flowing the voltage changes,
owing to polarization effects at the elecR
trodes. For this reason the me'asurement
cannot be made with a voltmeter, because
current is required to actuate such an instrument. The potentiometer is an ideal
instrument formeasuring voltage with no
etIITeiitfl
rom the source. To show
ow this instrument operates, a br~ef explanation of the potentiometer principle
is given here.
TM slidewire MN, Fig. 321, represents
an electrical conductor of uniform resistance, along which is plaged a scale graduated in uniform divisions. The battery
J3 causes a steady flow or" electricity in this
conductor. This is called the working current. The magnitude of this current can
be changed by adjusting the rheostat R.
With current flowing from M to N, M is
E E,
positive with respect to any point 0 between
M and N. Since the resistance is uniform,
FIG. 321.
the fall of potential is uniform per u'nit length.
If EEl represents the cell f6r pH measurements the calomel electrode is tho
positive pole, which is connected to M through the galvanometer G. This puts
like polarities in opposition. By adjusting the position of 0 the .potential difference between M and 0 can be made exactly equal to that between E and E 1
When the voltages are balanced the deflection of the galvanometer is zero, and
since this is It current indicating instrument, it proves that no current is flowing
between the electrodes~ Then if the fall of potential between M and 0 is
known, the voltage of EE is 'measured without any flow of current from the
electrodes.
1

W. N. Greer, Leeds and Northrup Co.

2297

2298

HYDROGEN ION

The fall of potential between M and 0 is determined when there is a known

potential difference between M and N: This is established by replacing the
pH measuring electrodes with a standard cell S, by means of the switch K. If
the voltage of the standard cell is, for example, ~.018 vdlts, the scale on MN is
graduated in 1018 uniform divisions, and the current i~ adjusted by means of
the rheostat R until the galvanometer shows no deflection. The potential
difference between M and N is then 1.018 volts, and the fall .of potential for
each scale division is .001 volt, or 1 millivolt. If the pH cell iR now substituted
for the standard cell, the potential difference between the electrodes wm be
shown directly in millivolts by the number of scale divisions between M and 0
when they are adjusted for voltage balance.
The accuracy of the measurement depends upon the standard cell voltage
and the uniformity of the slidewire (M N) resistance. Since this uniformity
can be determined' with practically any desired precision, the accuracy of
measurement depends principally on the constancy of the standard cell voltage,
and the method is therefore a primary method.
Potentiometers are available with various voltage ranges and accura,Qies,
some designed for portable use. Tables or equations for converting voltage
readings to pH values for all electrodes at various temperatures are obtainable.
Some forms of potentiometers are calibrat~d directly in pH, means being

provided to correct for solution temperatures.

In addition to the well-known standard types of potentiome~ers, there are
coming into greater use vacuum tube potentiometers. These have been designed primarily for use with the high resistance glass electrode circuits. The
resistances involved are too high to permit of the use of any except high sensitivity galvanometers, and only then with the low resistance glass electrode
(2-6 Megohms). Due to the high resistances involved it is necessary that
precautions be taken for properly shielding the potentiometer circuit. "This is
provided for in most of the commercially available instruments.
It is well to distinguish between accuracy and .sensitivity. It is possible
to have high sensitivity and low accuracy, and on/the other hand to have
conditions which would give high accuracy, if there were sufficient sensitivity.
The sensitivity depends principally on the gaivanometer used. If, with an
arrangement not capable of very accurate results, a very sensitive galvanometer
were used, its indications would be misleadingrif the sensitivity were not dis"tinguished clarly from a~curacy.
The most important factors affecting accuracy of results are the accuracy
of the measuring instrument used, the character of the solutions being measured
and the electrodes used, and the accuracy with which the referimce electrode
potential is known.
ELECTRODE SYSTEMS

A single electrode is not sufficient for pH measurements beca.use there is no

convenient method for measuring its potential, but the potentIal of tw.o elect.rodes, the absolute potential of one of which is known and is unaffected by pH,
can be compared. It is therefore customary to, use a ~~ed electrode as a
reference, and to use as a measuring electrode otteJ)est suited to the measurement at hand.

ELECTRO METRIC

2299'

For {h9se who are not familiar with the theory relating e.mJ. of electrodes
and pH, the rea~er is referred to the treatise on hydrogen-ion measurements by
, Clark, "The Determination of Hydrogen Ions," Williams & Wilkins.
Reference Electrodes: Calomel Electrode.-The calomel electrode is almost
universally used as a reference electrode, due to its constancy of potential and
ease of preparation. The calomel electrode is one in which mercury and calomel
are in contact with a definite concentration of potassium chloride. These are
contained in a vessel which may vary in design according to conditions of use. A common type of calomel
electrode is shown in Fig. 322; a more recent design is
shown in Fig. 323. The latter has the distinct advantage
that the electrode, immersed in the test solution, is at
solution temperature.
The potential of the mercury with respect to that of
the solution depends on the concentration of mercurous
ions from the calomel, but this is governed by that of
chlorine ~ns, which are produced in the solution by
potassium}hloride as well as by calomel. The proporFIG. 322.
tion of the total number of chlorine ions that is due to
the potassium chloride varies with the KCl concentration.
The voltage of the calomel electrode therefore depends on the concentration of
potassium chloride in the electrode solution. Saturated, normal and tenth
normal solutions are used.
The saturated calomel electrode is used more generally than
either of the others. l.t is more easily prepared, and has some particular advantages.
Speci5l1 precaution must be taken to prevent diffusion between the
saturated solution of a salt bridge with the less concentrated solutions of the normal and tenth normal electrodes, whereas with the
saturated electrode there is no tendency toward diffusion. When
diffusion is prevented the potentials of the normal and tenth normal electrodes are fairly stable at room temperatures, but at higher
temperatures 'it is diffi.cult to prevent diffusion, with its resulting
error in potential. The saturated electrode shows no change in
potential between 5 and 60 C., other than that due to temperature.
The temperature coefficient is approximately .8 millivolt per degree
C. Another advantage of the saturated calomel electrode is that its
FIG. 323. conductivity is high, which makes for increased sensitivity.
An objection to the saturated electrode is that the solution may creep over
the edge of the containing vessel with the resulting formation of masses of
crystals. This is avoided in the design shown in Fig. 323, if the electrode is immersed in a container of distilled water when not in use.
For accurate and reliable potentials, extremely pure chemicals mWlt be
used in calomel electrodes. Mercury specially purified and calomel produced
by electrolysis for the purpose can be commercially obtained in appropriate
quantities, and their use is recommended. Chemically pure potassium chloride
dissolved in distilled water is a suitable solution for use with them.
The saturated electrode is easily prepared. For a saturated KCI solution
put the salt into warm distilled water in a beaker and stir it 'thoroughly with

HYDROGEN ION

2300

repeated addition until no more of it will dissolve. Then stir in enough calomel
to saturate the solution. A small quantity is sufficient as it is not very soluble.
When the solution is cooled to room temperature, it sho~ld contain undissolved
potassium chloride and calomel. Prepare enough solut~on to fill a reservoir for
flushing the electrode from time to time. Add suffidient mercury (amount
depends upon construction of electrode) to make contact with the platinum
terminal, and sufficient calomel (HgCI) to assure saturation, an excess is
advisable. Moisten with saturated KCl solution, shake vigorously, and then
nearly fill with saturated KCI solution. If the electrode is not of the selfflushing type, care must be exercised in flushing the salt bridge occasionally.
Silver Chloride Electrode.-Comparable in principle to the calomel electrode
is the silver-silver chloride electrode. It has largely been used in studies on
activities of chlorides in solution, but recently has received increased attention
as a reference in pH measurements. This electrode is not as readily prepared
as is the calomel electrode, but it is highly reproducible, stable over extended
periods of time, and has a small temperature coefficient.
The potential of the silver-silver chloride electrode depends upon the method
of preparing the silver and the chloride. Methods have been described by
Brown,! Harned,2 Mclnnes,3 Noyes and others.4
Measuring Electrodes.-There are a number of electrodes which
have been used as pH measuring electrodes. Those which have
been best established and mQst largely used will be described.
Hydrogen Electrode.-The hydrogen electrode consists of a
noble metal coated with platinum black, immersed in the solution and under a definite partial pressure of hydrogen gas. Such
an electrode is illustrated in Fig. 324. Means are provided fer
introducing the hydrogen gas, and allowing the gas to escape wjthout forcing the solution entirely away from the metal.
Electrodes with large surfaces are generally not advisable as
they are more difficult to saturate with hydrogen and therefore
.'
come to equilibrium less quickly.
In work with ordinary chemicalrsolutfons, and when high accuracy is not essential, the solution being measured may be contained
FIG. 324. in an open vessel, but the electrode potential will be slightly influenced by oxygen absorbed from the air. For precise work the
vessel should be stoppered with provisions for carrying the electrodes through
the stopper, and for allOWing the gas to escape.
'"'
.. _
An electrode is said to be in equilibrium when its potential remains constant
if the solution does not vary. The potential of the hydrogen electrode is not
constant until the platinized surface is saturated with hydrogen, and any
reducing action due to hydrQgen is completed. '
Limitations.-In some solutions, notably biological fluids, the platinum
black may become clogged with substances which p~event absorptibn of hydrogen. The hydrogen electrode cannot be used in solutions containing strong
oxidants like ferric ion, dichromates, nitric acid, peroxides and chlorine, nor in
J.
J.
3 J.
4 J.

Am.
Am.
Am.
Am.

Chern.
Chern.
Chern.
Chern.

Soc.,
Soc.,
Soc.;
Soc.,

56,646 (1934).
55, 4849 (1933).
37, 1445 (1915).
39, 2532 (1917).

i?1~
ELECTRO METRIC

2301

those containing strong reductants, such as sulfurous acid (S02) and hydrogen
sulfide. Some organic compounds are very susceptible to reduction, particularly those of the aromatic series, such as azo benzene, aniline dyes and some
unsaturated acids. In a solution containing either oxidizing or reducing compounds, a noticeable drift in the voltage occurs and a balance is difficult to
obtain. The oxidizing or reducing action must be completed before the voltage
is steady. The measurement is not trustworthy, for the pH value of the solution
may be seriously changed in such reactions.
The hydrogen electrode cannot be successfully applied to solutions containing metal i~s--that fall below hydrogen in the electromotive series of metals
and, also lead, all of which are reduced on the platinum electrode.
P!atinizing the Electrodes.-The film of platinum black is deposited by
electrolysis in 3% platinic chloride solution containing 1/40% lead acetate.
The electrode should be thoroughly cleaned before platinizing. Using two
dry cells in series it usually requires from 30 seconds to one minute to deposit
the platinum black. Experience will determine the amount of plating neces. sary.. The electrodes should be washed with distilled water, and immersed in
distilled water when not in use. The electrodes should not be allowed to, dry
thoroughly.
Hydrogen Gas.-A very convenient source of hydrogen is the compressed
gas supplied commercially in metal cylinders. A cylinder containing 100 or 200
cubic feet is commonly used in the laboratory, while small cylinders are available
for portable use.
Compressed hydrogen produced by electrolysis may be sufficiently pure to
. use directly from the cylinder for measurements of ordinary accuracy, but it is
common practice to pass it through solutions of potassium hydroxide and
pyrogallol and then through water to remove traces of oxygen and other impurities that would affect the potential of the hydrogen electrode.
Quinhydrone Electrode.-Some of the difficulties encountered in the use of
the hydrogen electrode are avoided by using a quinhydrone electrode. The
basis of this electrode is a piece of platinum or gold the same as used for the
hydrogen electrode, but the surface of the metal is not platinized and it is not
supplied with gaseous hydrogen. Instead, a small quantity of quinhydrone
(benzoquinhydrone) is dissolved in the solution, and in certain conditions the
electrode in the solution acquires a potential that is definitely related to the
hydrogen ion concentration of the solution. The potential is measured against
t.hat of a calomel electrode, and the pH value is found from the measured
voltage in the same manner as with a hydrogen electrode.
.
Biilmann 5 was the first to describe the quinhydrone electrode for the pH
measurements. Morgan, Lammert and Campbell 6 have published a series of
papers covering detailed studies of this electrode. The quinhydrone electrode
is quickly prepared, develops its potential rapidly and is not readily poisoned.
Rather than describe the method of preparing quinhydrone, it' 'is suggested
that this material be obtained from chemical manufacturers or suppliers of
high grade chemicals.
Although the quinhydrone electrode can be applied successfully to many
solutions containing oxidizing and reducing substances in which the hydrogen
6 Ann. Chim. Phys., 15, 109 (1921).
6 J. Am. Chern. Soc., 53, 2154 (1931) and 54, 910 (1932).

2302

HYDROGEN ION

electrode would give erroneous results, still it is not applicable in all cases .
. Measurements have been made witll it on dilute nitric :acid, solutions, unsaturated organic acids, and on a variety of oxidizing systems that are of too low
an oxidizing intensity to disturb the electrode potentiall
Limitations of the Quinhydrone Electrode.-The qJinhydrone electrode is
not applicable over the entire pH range. In the acid range, 0 to 7 pH, it is
excellent. In alkaline solutions strongly buffered, results of fair accuracy can
be obtained to about 9 pH. In solutions poorly buffered, the upper limit is
approximately 8.0 pH. In the latter case it is essential not to add an excess ~f
quinhydrone, because it reacts with the alkali and changes the pH value of the
solution. About 6 to 8 drops of a saturated solution of quinhydrone in acetone
' .
is suffici~nt for 50 ml. of test solution.
If the pH value of a solution measured with a hydrogen electrode differs
from that with a quinhydrone electrode, the difference may be due to the" salt
error" in the quinhydrone value. The hydrogen electrode is assumed to have
no salt error. The magnitude and sign of the salt error (positive or negative
divergence from the correct pH value) of the quinhydrone electrode depep.ds
upon the type' and concentrations of the salts. present in the solution. The
salt error in nickel plating solutions of total salt concentration about 2.0 N is
below +0.05 pH. In a 10.Q N ammonium sulfate solution it is approximately
-0.2 pH, while in a 4.0 N sodium chloride solution it is approximately +0.2
pH. Since most solutions are far more dilute than these, the salt error is
generally negligible.
/'
In high purity (99%) sugar solutions the quinhydrone electrode gives
values which are approximately .2 pH low, while on low purity products, such
as molasses, the values are high. Over purity ranges of 80% to 97% the
quinhydrone electrode gives good results.
Antimony Electrode.-Of the various metal and metal oxide electrodes that
have been described, the antimony electrode hM proved to be the most satisfactory. It does not require gas or a catalytic surface as does the hydrogen
electrode; it does not require the addition of a reagent as does the quinhydrone
electrode. The antimony electrode is not readily prepared, but once made up
is useable for very extended periods: It may be obtained commercially ready
for use. It is a rugged electrode and in certain designs is used industrially for
continuous plant service. The inherent accuracy of the antimony electrode is
not as great as is the hydrogen or the quinhydrone electrode.
Its limitations have been described by Perley.7 Notable. among these is
that the e.m.f:-pH relation varies with the degree of buffering of the solution,
and that the degree of saturation with air, pltrticularly below 7 pH, affects the
potential-pH relationship. The antimony electrode is not applicable to solutions which are oxidizing or are strongly reducing, nor in solutions containing
more than a trace of copper, silver, or other metals below antimony in the
electrochemical series of the elements. However, uX;_der known conilitions this
electrode serves a very useful purpose in measuring pH over' the .range of 3 to
12 pH. It is being used quite extensively in sugar solutions (beet, raw cane,
and refinery), paper mills, water treatment, pigments, etc. For measurements
on a given sample extending over long periods of time, this electrode is particularly suitable.
7 Trans. Am. lnst. Chern. Engrs., 29 (1933).

ELECTRO METRIC

2303

, Glass Elecirode.-The work of Hughes,8 Kerridge,9 McInnes and Dole,IO. 11

and others have shown the possibilities of the glass electrode as a pH measuring
electrode. The glass electrode consists, primarily, of a thin membrane of glass,
of a definite composition (72% Si0 2, 22% N a 2 0, 6% CaO), one side of the
membrane being exposed to the solution being measured and the other side
exposed to a standard reference solution. The membrane is made up in the
form of a bulb, with a reference electrode and solution on the inside. The bulb,
along with a reference electrode, is immsered in the solution to be tested. The
glass electrode, which behaves like a hydrogen electrode, is thus interposed
between two fixed electrodes.
A high sensitivit7 galvanometer, or an electrometer potentiometer, is
required to me~sure pH with the glass electrode due to its high resistance.
Glass electrodes and suitable measuring devices are commercially obtainable.
The glass electrode is particularly suitable for pH measurements in the
presence of oxidizing and reducing solutions, where other types of electrodes
give erroneous results. It also is the best type of electrode to use in unbuffered
solutions, since the introduction of the glass bulb into an unbuffered solution
does not change the pH of the latter. The electrode is stable over rather long
periods, the life of the electrodes being largely determined by crystallization
of the glass. The characteristics of the internal reference element and solution
determine how often they must be replaced.
The glass electrode, in some forms, is rather fragile and must be handled
with care. In other forms, with heavier glass wall, the electrode often exhibits
an asymmetry potential; that is, it differs from the Nernst equation by more or
less constant potential. The E.M.F.-pH relationship of the glass electrode is
linear up to pH 9.6, above which it may be used with a special calibration to
12.5 pH. The greatest deviation from linearity is obtained in solutions containing sodium ion concentrations in excess of 0.1 molal.
The electrode can be made in small sizes, permitting of measurements on
minute samples. It has a.further advantage in that it does not contaminate the _ .
solution being measured, should only limited samples be available, and it is
necessary to make other determinations on the same material.
~

WORKING FORMULAE FOR

pH

ELECTRODES

1. Hydrogen-Saturated Calomel.
p

H- V-0.2648+.00076t
0.0001983T
'

at 25 C.,

H= V-.2458.
.' .0591

2. Quinhydrome-Saturated Calomel.

H_

Y+0.4529+ .00002t

0.00019832T
8 J. Am. Chern. Soc., 44, 2860 (1922).
9 Biochern. J., 19,611 (1935).
10 Ind. Eng. Chern. Anal. Ed., 1, 57 (1929).
p

11

J. Am. Chern. Soc., 53, 3315 (1931).

'

2304
at 25 C.;

HYDROGEN ION

H':_ V .4534_.
- .0591

3. Glass-Saturated Calomel.-When using quinhyd~one-l pH hydrochloric

acid as reference in glass electrode, use formula 'as applied to (2) Quinhydrone
above, up to 9.6 pH. Relation is non-linear above 9.6 pH in the presence of
sodium ions.
'
4. Antimony-Saturated Calomel.
Buffered solutions at 25 C.
Below pH 7,
Above pH 7,

H= V+.0157,
p

pH=

.0591

V -.0292
.0527 .

V = Observed potential in volts,

1= C.,
T=273.1+t.

'POTENTIOMETRIC METHODS

_Measurements of the electromotive forces.of electrolytic cells may be used in

analysis for two closely related purposes: (a) the determination of the concentration (or activity) of some kind of ion. The concentration of hydrogen ion,
or the pH of a solution, is frequently estimated by this means as has been
described in the preceding chapter. (b) The determination of end-points in
tit rations.
.
-krhe principle of the potentiometer has been explained in the preceding
chapter.i'"
Single Electrode Potentials.-The E.M.F. of a cell may be considered to be
made up of the following effects:
Electrode I

Total E.M.F.
Solution I II Solution II

I Electrode II.

abc
The total effect is: E.M.F.=(a-c)b. The sign of c may be different from
that of a. For ~xample, the values for copper and zinc are rllspectively +0.344
and -0.762 under certain conditions; in this case E.M.F.=0.344-(-0.762)
= 1.106 volts. The liquid junction effect b between the two solutions is due to
different mobilitie~ of the various ions; it is frequently small (1-3 millivolts),
and is ignored in most analytical determinations. In precise work the effect
may be measured, or taken into account by a calculated correction.
If one of the electrodes, for example, II, is a constant, or reference electrode,
the changes in total E.M.F. of the cell measure essentially the variations which
occur at electrode 1.
Suppose that electrode I consists of a metal in contact with a solution of its
ions, for example silver in contact with a solution of silver nitrate. In such an
instance Nernst found that the effect at a single electrode could be represented
by:

RT

Er =Eo+ nF In (Ion],
where R is the gas constant, T the absolute temperature, n the number of
Faradays concerned in. the transfer of 1 gram ion to one gram atom, oI vice
versa, and F the Faraday = 96,540 coulombs, In the natural logarithm, and [Ion]
the molecular concentration of the ion. Upon converting to common logarithms (base 10) and substituting for R its value 8316 volt-coulombs, we have:

E r = E 0 +0.000198
n T Iog [I on.]
For silver, for example, E Ag =0.798+0.0591 log [Ag+] (at 25 C.). The convention regarding the sign of the electromotive force here used is the opposite
1

N. Howell Furman, Professor of Chemistry, Princeton University, Princeton, N. J.

2305

2306

POTENTIOMETRIC METHODS
,

of that given by Nernst; we use the sign of the metal 'relative to the solution,
which is the convention adopted by foreign and American electrochemical
societies.
.

,
This electrode potential expression is not limited to substances which can
be formed into solid electrodes. For example, platinized pl(1tinum catalyzes
the reversible exchange of hydro.gen atoms to hydrog~n ions, and a hydrQgen
electrode may be formed by causlhg hydrogen gas to stream through a solution
,into which a lightly platinized platinum, gold, or iridium electrode is dipped.
At 1 atmosphere pressure of hydrogen gas:
EH = 0.0000+0.0591 log [H+] (at 25 C.).
The Eo value of the hydrogen electrode is arbitrarily chosen as the standard of
reference, and is set equal to zero. The Eo value of any other substance is then
found from the measured value of E (with refer~nce to the hydrogen standard)
when the concentration of the ion in question is equal to 1 molar.
If we are dealing with a negative ion, the sign of the second term in the
eleQtrode expression is negative. For the chlorine electrode: ECJ=Eo-0.0591
Jog [CI-] (at 25 C.).
The Eo value is characteristic for each substance and is a measure of the
relative oxidizing or reducing power of the substance relative to hydrogen.
It should be noted that Nernst derived the electrode potential formula on
the basis of ideally dilute solutions, and ~hat the electrod~ potential is really a
function of the activity rather than the concentration of the ion in question.
Hydrogen ion concentration measurements are standardized at present on
E.M.F. meas!lrements, or in other words on an empirical activity basis, since
the point of reference of the whole system is based on hydrogen-ion concentration values calculated from conductance measurements. Kolthoff gives the
relation between the exponents of the activity and ion concentration values as:
paH=pH+0.04. 2
As far as the potentiometric determination of ind-points of volumetric
reactions is concerned, activity relations need npt be taken into account.
Oxidation-Reduction Electrodes.-A burnished platinum electrode may,
when dipped into a solution containing an oxidizing agent and its reduction
product, assume a potential which is a measure of the reversible oxidationreduction power of the solution. Peters 3 found that for"*he reversible f~mic
ferrous ion system, the relations were expressed by the formula~, _..
[Fe+++]

E=Eo+0.0591 log [Fe++] ~at 25 C.).

In this case the Eo value is equal to the observed value when [Fe+++] = [FeH ] = 1
(the normal oxidation-reduction potential).
.'
In general, for a reversible oxidation-reduction equilibriu~:
Ox+n~Red,

in which Ox represents oxidizing agent, Red its redl!Qtion pr:6duct, F the Fara21. M. Kolthoff, .Rec. Trav. chim., 49, 401 (1930), ...1-:- _
3 R. Peters, Z. physik. Chern., 26, 193 (1898).
~

POTENTIOMETRIC METHODS

2307

day, and n the number of Faradays per gram ion of oxidizing agent, We have
if the electrode process is. reversible:
'
[Ox] ( t 25 C )
E =Eo+0.05911
n
og [Red] ~
..
The Eo value is an important characteristic constant for each system, and is
referred to that of the normal hydrogen electrode as zero.
If the partial oxidation-reduction reaction involves hydrogen or hydroxyl
ion, we obtain an electrode expression that includes the concentrations of these
ions. For example:
Mn04-+8H++5 e~Mn+++4H20,

E=E +.0591
o

5 1og

[Mn04-] [H+]8 .
[Mn++]

The concentration of water is practically constant for dilute aqueous solutions,

and hence its actual value need not be introduced in determining the Eo value
for the system.
Reference Electrodes.-The concentration .of a metallic i in co
.h
an electrode of the metal may e ep constant by covering the metal :with-a.
suspenSlOn of one of its slightly so1U6le salts in a solution containing an anion

Ke(

..

p:tt~.=:.::::I'''.4-.. H9~ C~

a.

,--...!.-_

_;r-H~

1.
~.
FIG. 325.-Calomel Electrode Vessels. (1) Modified Ostwald Type. The mercury
is overlain with calomel intimately mixed with mercury; the potassium chloride solution
is introduced through tube, T. The side tube at a ends in a capillary, or a porous plug
to prevent diffusion. Contact with the mercury is made by a platinum wire sealed
through a glass tube, and joined at b to a copper wire. (2) Represents the bottle type
electrode recommended by E. Muller. The siphon tube, H, is made of two glass
T-tubes, joined by rubber tubes, r. The bottom of each tube is plugged at c with agar
jelly made up in potassium chloride solution of the strength needed for the electrode
(0.1 N, N, etc.).

2308

POTENTIOMETRIC METHODS

in common with the precipitate. The calomel elec~rode consists of a mercury

electrode in contact with a potassium chloride solutiQn that is saturated with
calomel, Hg 2 Ch. The concentration of mercurous ion's is'therefore constant at
constant temperature, and hence a calomel half-cell \Serves as a constant, or
reference electrode. Some convenient types of half-cells are shown iI\ Fig. 325.
The concentrations of potassium chloride that are rhbst frequently used are:
(i) 0.1 N, (ii) N, (iii) 3.5 N, or (iv) a saturated solutio'n.
Values for Various Calomel Electrodes
(Referred to N hydrogen electrode as zero)
Concentration of
Kel

ECalomel in Volt

0.1 N ..........................
1 N ..........................
3.5 N'. , ........................
Saturated. ; . . . . . . . . . . . . . . . . . . . ..

0.3380-0.00006
0.2864-0.00024
0.2549-0,00039
0,2504 - 0.00065

(t-18)
(t-18)

(t-18)
(t-18)

Here t represents degrees Centigrade (applicable between 10 and 30).

Other electrodes that are frequently employed are the silver electrode in
contact with silver chloride and a potassium chloride solution of definite concentration. When the presence of chloride is undesirable a salt bridge of
sulfate or nitrate may be interposed between the reference electrode and the
solution. A mercury-mercurous sulfate~sulfuric acid electrode is convenient"
for reference purpos,ts in the study of oxidation-reduction processes in the
absence of chloride.POTENTIOMETRIC TITRATION'S

.Jf the E.M.F. of a ceil composed of two electrodes dipped ~nto an unknown
s01ution is measured during a titration, the plot of E.M.F. readings as ordinates
against ml. of standard solution as abscissas gives in general an abrupt change
in potential at the analytical end-point of the reactio.n at the end-point. One
electrode serves as an indicator of changes in ion concentration, and the other
electrode is a calomel, or other reference electrode. The indicator electrode
must be appropriate for the particular type of reaction involved. The solution
must be well stirred during the titration, preferably by a motor-driven stirrer.
If the standard solution is added in small equal incremen4l. near the end-j)oint,
say 0.05 ml. at a time, the end-point is in general marked by a maximum change
That is, the plot- of A E.M.F.
"
Am!.
against ml. of reagent gives a maximum at the end-point. In favorable cases
it is not even necessary to record the readings near the end-point, although this
is advisable until some experience has been had witp a particular type of titration. The outstanding advantages of the potentiometric method are (1)
applicability to colored solutiOIis; (2) precision of determination of end-point;
. (3) the possibility of obtaining a succession of end-points, and hence the determination of two or more substances by a single titration~ I
A type of assembly which is suitable for gene!].l study' of potentiometric
titration processes is shown in Fig. 326. If it is~ot desired to determine the
of E.M.F. per small increment of reagent.

2309

POTEN'PIOMETRIC METHODS

exact voltage range for a given process, the potentiometer may be replaced by a
simple slide wire potentiometer, and the use of the standard cell may be omitted.
Some still simpler forms of apparatus are described on p. 2318. One of the
types of titration cell whicJ;t may be used when air must be excIudci~ showl'l. in
.Fig. 327. The vessel shown in Fig. 328 may be used for hydr6~electrode
titrations if an opening is provided in the stopper for a burette.

N.E.

s.

---~-------I

I
0.1

I
I

I
I
I
I

I
I

I
I

1.4

11.31.2"

.9

I ___________________
Pot.. 1.0
I
L
~

FIG. 326.-Potentiometric Titration Circuit. Ba, battery; R, resistance; S, battery

switch; Pot., potentiometer; N.E. standard cell; S" reversing switch; S2, Sa, single-pole
double-throw switches; In.~ indicator electrode; Rej., reference electrode; B, beaker
containing solution to be titrated; p, porous plug, or capillary. In most cases a. motordriven stirrer is essentiaL (Furman, J. Chem. Educ., 3, 932 (1926).)

Neuhalizations.-A calomel electrode is in general suitable for reference.

The iudicator electrode may be anyone of the following:
Hydrogen Electrode. 4-The solution should be free from powerful oxidizing
and reducing agents, for example, permanganate, bichromate, ferric iroB, nitric
4 General literature references: Bottger, Z. physik. Chem., 24, 253 (1897); Hildebrand, J. Am. Chem. Soc., 35, 869 (1913); H. S. Britton, J. Chem. Soc. (London),
. numerous articles, 1924-1927.

2310

POTENTIOMETRIC METHODS

acid, ions of metals like Ag, eu which are precipitated, easily reducible organic
compounds, the alkaloids, certain dyestuffs.
I
Quinhydrone Electrode. 5-May be used in the titration of acid solutions
which are free from strong oxidizing and reducing, agfnts. May be used !Ol:
alcoholic solutions.
.I

FIG. 327.-Titration Vessel. A, inlet for inert gas, also used as electrode holder;
B, reductor, or other form of special apparatus used in introducing sample; C, motordriven stirrer; D, burette; E, tube leading to reference electrode; H is a sheltered platinum electrode to be inserted in place of E, when the differential method is used. The

stopper, G, is bored to give any desired number of openingS, through which glass tubes
are placed. The rubber tubes at F join the various pieces of apparatus to the glass
tubes carried by the stopper. The gas escapes around the shaft of the stirrer, C.
(Furman and Schoonover, J. Am. Chern. Soc., 53, 2561 (1931)).
I

u~ed; some antimony

trioxide, preferably of the'stable cubic form, should be prese~nt;-~l'he electrode
may be used for titrations where the hydrogen electrode applies, and also for
determining end-points when strong oxidizing agents' are present.
Oxygen, or Air Electrode.7-May be used for,all ordinary titrations and also
in the presenc~ of oxidizing agents.
5 Kolthoff, Rec. Trav. Chim., 42, 186 (1923); LaMer, and Parson, Proc. Expl. BioI.
Med., 20, 239 (1923); J. BioI. Chern., 57,613 (1923); Auerbach and Smolczyk, Z. physik.
Chern., 110, 65 (1924); Baggesgaard-Rasmussen and Schou, Z. Elektrochem., 21, 189
(1925); Wagener and McGill, J. Am. Pharm. Assoc., 14, 288 (1925); Klit, Z. physik.
Chern., 131, 61 (1927).
6 Ubi and Kestranek, Monatsh., 44, 29 (1923); Kolthoff and Hartong, Rec. Trav.
Chim., 44,113 (1925); Britton and Robinson, J. Chern. So<;:., p. 458 (1931).
7 Furman, J. Am. Chern. Soc., 44, 2685 (1922); TranS. Am.- Electrochem. Soc., 43,
79 (1923); Britton, J. Chern. Soc., 125, 1572 (1924); 127, -1896,2110 (1925).

Antimony Electrode. 6-A rod of cast antimony is

POTENTIOMETRIC METHODS

2311

Glass Electrode. 8-Suitable in the presence of oxidizing or reducing agents.

A special electrical measuring system including either a quadrant electrometer
or a vacuum tube voltmeter is usually required because of the high resistance
of the electrode. However with low-resistance glass films (2-5 megohms) a
simple potentiometer and a high resistance galvanometer may be used.

B,

(.

A
FIG. 328.-Cell for pH Measurement. Pure hydrogen enters a bubbler, B I , at a.
BI is filled with some of the solution whose pH is measured; the latter is put in vessel A,
and is saturated with hydrogen. The gas comes in contact with the lightly platinized
wire or foil b, which is of platinum, and is sealed into a glass tube, where contact is
made with the copper wire, E. The hydrogen escapes through the bubbler, B 2 The
tube, T, is filled with potassium chloride solution, and leads to the calomel electrode;
c is a porous plug, or capillary to prevent diffusion.

Additional metal-metal oxide electrodes which serve as hydrogen-ion indicators are known (mercury-mercuric oxide, copper-copper oxide, bismuthbismuth oxide, tellurium-tellurium oxide, etc.) but do not appear to offer any
decided advantages over those mentioned. 9
Some graphs of hydrogen electrode titration data are shown in Fig. 329
together with graphs of indicator transition intervals. Titration curves of
similar form are obtained when other indicator electrodes are used.
The electro metric method is superior to indicator methods when the solution
is highly colored; it also appears to offer decided advantages-in'estimating small
amounts of weak acids (see "differen~ial method," p. 2318).
8 Haber and Klemensiewicz, Z. physik. Chern., 67, 385 (1909); Hughes, J. Am. Chern.
Soc., 44, 2860 (1922); J. Chem. Soc., p. 481 (1928); see Kolthoff and Furman, "Potentiometric Titrations," 2nd. Ed., 1931, for further bibliography and description of apparatus.
9 See Kolthoff and Furman, "Potentiometric Titrations," for a review of the
literature.

2312

POTENTIOMETRIC METHODS

Potentiometric neutralizations may be made in many anhydrous solvents,

as, for example, ethyl alcohol,lo amyl alcohol,ll ,~laciallacetic aci9,12 etc.

PRECIPITATIONS AND COMPLEX FORlIATIONS

The reference electrode may be a calom,el half cell, )Jith a suitable salt bridge
interposed if the presence of chloride is harmful. Theoretically any sufficiently
pH

pH
12

12

Indicator
Intervals

Titration Curves

11

11

10

10

Neutral Point

----------

6
5

10

15
ml

20

25

Alkali~

30

35

40

25

50

% Alkaline

75

100

Form-~

FIG. 329.-,-Titration Curves and Changes of Indicators.-Curve 1, approx. 0.1 N

sulfuric acid titrated with equivalent alkali; (2) 40 cc. 0.0711 N phQsphoric acid titrated
with 0.2 N NaOH (Davis, Oakes, Salisbury, Ind. Eng. Chem., 15, 182_(1923. (3)
25 cc. 0.5 N acetic acid with 0.5 N NaOH. The dotted lines leading from the right of
the figure represent approximately the mid-points of the visible indicator changes.
unattacked metal or non-metal might serve as an indicator electrode. Practically, only a few have been used successfully, notably the silver, mercury or
iodine electrodes~ These electrodes serve for the titration of halides with silver
nitrate, or with mercury salts, or for the determination of silver or mercury with
10 Bishop, Kittredge and Hildebrand, J. Am. Chem. Soc., 44, 135 (1922).
11 Seltz and McKinney, Ind. Eng. Chem., 20, 542 (1928).
'
12 Hall and Conant, J. Am. Chem. Soc., 49, 3047 (1927); Conant and Hall, ibid., 49,
3062 (1927); Hall and Werner, ibid., 50, 2367 (1928); see-Clark, "Determination of
Hydrogen Ions," and Kolthoff and Furman, "PotentiomiiJric Titrations," 2nd Ed., for
further references.

POTENTIOMETRIC METHODS

2313

halides. Fig. 330 shows the graph of the data obtained in the titration of a
mixture of iodide and chloride with silver nitrate. This type of process is the
more accurate: (a) the more concentrated the solution, provided adsorption
phenomena can be dealt with satisfactorily; (b) the less soluble the precipitate.
+0.4
+0.3
+0.2

...

--

~lRt.
"

1\

+0.1

g +0.0

t--

...:

Aec/
-.;;;.::.:,

E-0.1
u.j

-0.2

-0.3

-0.4

1\

-0.5
5

10

15

20

25

30

35

40

45

50

ml AgN0 3

FIG. 330.-Titration of Mixture of Chloride and Iodide with Silver Nitrate. 25 cc.
of approx. O.l.N KI plus 25 CC. approx. 0.1 N KCI titrated with silver nitrate. The
dotted curve shows the graph obtained when iodide alone is present.

The silver electrode indicates the end-point of the important complex

formation process: Ag++ 2CN-~Ag(CNh-. When this process is complete
further addition, of silver ion gives a flat curve during the precipitation of silver
cyanide, followed by an abrupt change at the end of this second process. This
reaction may be used in the indirect determination of nickel, or in the estimation
of silver halide in a. photographic emulsionY
The titration of halide with silver may be carried out potentiometrically
with atomic weight precision. The potentiometric method offers especial
advantages in examining mixed halides (Clark, loco cit.),l4
The end-points of ferro cyanide precipitations may be determined with an
oxidation- potential indicator electrode (cf. next section). The ferrocyanide
solution is prepared with a small amount of ferricyanide present. As long as
ferrocyanide ions are removed by the formation of the precipitate, .the potential
See W. Clark, J. Chern. Soc. (London), p. 768 (1926).
References on halide titrations: Behrend, Z. physik. Chern., 11,466 (1893); Liebich,
Dissertation, Dresden, 1920; Treadwell, Janett and Blumenthal, Relv. Chim. Acta, 6,
513 (1923); E. MUller, Z. Elektrochern., 30, 420 (1924); Photo Ind., 18, 14 (1924); Clark,
J. Chern. Soc. (London), p. 768 (1926); Lange and Schwartz, Z. Elektrochem., 32, 240
(1926); Z. physik. Chern., 129, 111 (1929); MacInnes and Dole, J. Am. Chern. Soc., 51,
1119 (1929); Tornicek, Collect. Czechoslovak Chern. Commun., 1, 443, 585 (1929); 2,
1 (1930); Tornicek and Jansky, ibid., 1, 582 (1929).
13
14

POTENTIOMETRIC METHQDS

2314

remains in the oxidizing region; as soon ai'! the first, excess of ferro cyanide is
present there is a sharp drop in the E.M.F. of the titra,tion cell.
The titration of zinc with ferrocyanide has been !ltudied by a number of
investigators, and many conflicting statements and conclusions occur.i5 The
solution should be only slightly acid; the titration is.rri6re rapid at 65 than at
room temperature, but there is danger of overstepping the end-point even at the
higher temperature. Kolthoff ~nd Verzijl found that
in the titration of 100 ml. of 0.1 N zinc sulfate the
o
C.
error was as follows:
O

Temperature
o

C.

15 ......
65 .... "
15 ......
65 ... ~"
65... . ..
15 ......
65 .... "
15 ... '"
65. . . . ..

Substance Added

Error in Per Cent of ml. of

Reagent Required

(Neutral sol'n)
(Neutral sol'n)
0.12 N sulfuric acid
0.12 N H 2S0 4 , 3 g. K 2SO.
5 g. K 2SO.
10 g. ammonium. chloride
10 g. ammonium chloride
20 g. ammonium sulfate
20 gI ammonium sulfate

From Potentiometric, Titrations, 2nd Ed.,

Furman.
-

-1.04
-0.9
-0.77
-0.4
-0.45
-2.25
-0.8
-0.05
-1.15
Kolt~off

and

Large amounts oj calcium and magnesium ca~e

errors; addition of'"fluoride prevents the interference of
a.
ferric ion. Manganese must be removed before the
titration.
FIG. 331. - OxiWith potassium ferrocyanide the precipitate is
dation-Reduction Indicator Electrode. a K2Zna(Fe(CN)6)2, whereas with other alkali metal ferrorepresents about 5" of cyanides the precipitate has a different composition.
No. 28 B. and S. From all of the considerations which have been presented
platinum wire sealed it is evident that good results will be obtained only
through a glass tube
and fused at b to a under carefully standardized experimental conditions.
Cadmium and lead :r;nay be detm-mined by ferrocopper wire, which
leads to the binding cyanide precipitation, but silver and ,copper give erratic
post c. The rubber results. is
.
tube, r, supports the
Oxidation - Reduction Reactions. - The indicator
electrode on the bakelite plate, d, which electrode is of burnished' platinum, either wire, foil or
may be drilled to carry gauze. Fig. 331 represents the conStl'uctio,n of one satisa number of electrodes factory type of electrode. A calomel or - a mercuryof the construction
mercurous sulfate and sulfuric acid reference electrode
shown.
may be used.
The reagent may be any sufficiently stable oxidizing or reducing agent,
and each reagent may be use'd to determine one or more substances in a single
(

Knauth, Dissertation, Dresden, 1915; von Bischowsky, Ind. Eng. Chern., 9, 668
(1917); Hedrich, Dissertation, Dresden, 1919; Kolthoff, Rec. Trav. Chim., 41, 425
(1923); Kolthoff and Verzijl, ibid., 43, 380 (1924); Fr. Miiller, Z. anorg. allgem. Chern.,
128, 126 (1923); E. Miiller and Gabler, Z. analyt. Chern., 62, 29 (1923); Treadwell and
Chervet, Helv. Chim. Acta, 5, 633 (1922); 6, 550, 559 (1923).
I
16 See Kolthoff and Furman, "Potentiometric Titra~ions," pp. 331-333, 2nd Ed.,
1931, for determination of cerium, lanthanum, gallium;and indium by ferrocyanide
precipitation.
]5

2315 .

POTENTIOMETRIC METHODS

titration. In Fig. 332 is shown the determination of iron, vanadium, and uranium by a single titration of the reduced solution with permanganate. The
uses of potassium permanganate, bichromate, iodate, bromate or ceric sulfate
have been intensively studied, and the potentiometric method' offers the advantages of accuracy, speed, and usefulness in highly colored solutions. The
powerful reducing reagents, chroIllous, titanous, or stannous chloride, arsenious
acid, and ferrous sulfate have decided advantages in certain fields of industrial
analysis. We shall describe only those methods which are useful in steel
analysis.
,

+0.5

+0.4

r~ ~VV

+0.3

~ +0.2
/

E
L1.i

-II

Fe ---; Felli

+0.1

0.0

...J--/

;'VI_!! VIV

-0.1

-0.2

UJ.:', UVI

/~ 2

'a
4

I~ -c

b
10

12

14

16

18

20

-4

22

24

26

MI KMn0 4 -

FIG. 332.-8irnultaneous Determination of Reducing Agents. From data of Miiller

and Flath, Z. Elektro_chern., 29, 500 (1923). The cc. of perrnanganate given by (a-c)
correspond to oxidation of DIV to UVl; b cc. are equivalent to the iron, and c ce. to the
vanadium.

Determination of Chromium and Vanadium in Alloys.-The pioneer work

in this field was done by Kelley and his associates. 17 It was found possible to
balance the E.M.F. of the cell approximately during the titration. The endpoint was then marked by a single permanent galvanometer deflection. Kelley
17 Kelley and Conant, J. Am. Chern. Soc., 38, 341 (1916); Ind. Eng. Chern., 9,780
(1917); Keliey, Wiley, Bohn and Wright, Ind. Eng. Chern., 13,939 (1921) i 11,633 (1919);
Kelley and Wiley, Ind. Eng. Chern., 13, 1053 (1921).

. 2316

POTENTIOMETRIC METHODS

devised an electrotitration apparatus in which this principle of operation may

be most easily used. More recent work has led to improved and more selective
oxidation methods prior to titration. 1s The following procedures are those of
Willard and Young. They found that potassium brrmate oxidizes vanadium
selectively in the presence of chromium if the solutio:p. contains a definite COllcentration of hydrochloric acid.

2KBrOa+ lOVOS0 4 + 14H20

=
,

10HVOa+ Br2+ K 2S0 4 +9H 2S0 4

'

. Vanadium in Chrome-Vanadium Steels.-A sample of 4 or 5 g. is convenient

when the percentage of vanadium is low (0.15 to 0.25%). Place it in a 600-ml.
beaker, add 30 to 35 ml. of water and run in 1.5 ml. of sulfuric acid, sp.gr. 1.83,
for each gram of steel, and, 3 to 4 ml. of acid in addition, to speed the solution.
After the steel has completely decomposed, boil until a considerable quantity
of salts separates out, in order to assist in decomposing carbides. Dilute with
20 ml. of water and heat until the salts have dissolved. Add nitric acid, sp.gr.
1.42, drop by drop to the hot liquid until the violent oxidation of ferrous sulfate
is over (3 to 3.5 ml. of acid are sufficient). Boil the solution to destroy oxides
of nitrogen, dilute to 200 mI., and add to this solution which should be at room
temperature, 6 ml. of hydrochloric acid, sp.gr. 1 18 (if 4 or 5 gram sample is
used; with a smaller sample more acid mlist be added: 9, 12 and 14 ml respectively for 3, 2 and 1 g. of iron), 5 g. of ammonium sulfate, and.1.5 to 2 g.
of potassium bromate. Leave the mat~rial on the low temperature hot plate
for 15 minutes, by which time it should reach approximately 65 C., boil 10
minutes to destroy excess bromate, cool in ice to 5 C., add 25 ml. of ice-cold
sulfuric acid, sp.gr. 1.5, and titrate with 0.025 N ferrous sulfate.
Vanadium in Chrome-Vanadium-Tungsten Steels.-The sample may vary
conveniently from 1.5 g. for a steel containing 1 % or more of vanadium to 3 g.
for 0.2% of vanadium. Add 25 to 30 ml. of water to the weighed sample in a
400-ml. beaker. Use 1.5 ml. of sulfuric acid, sp.gr. 1.83, for e'ach gram of steel
and 3 to 6 ml. excess. Warm gently until there is no further evolution of
hydrogen. Boil over a free flame until salts begin to'separate in order to assist,
in breaking up carbides, dilute to about 50 ml ana boil until salts have dissolved. Add to the boiling hot solution nitric ltcid of sp.gr. 1.42, drop by drop,.
until the violent action is over, then 5 to 6 ml. in excess. Boil down until
tungstic acid separates out in considerable amount and becomes yellow;
dilute somewhat with hot water, let settle, and filter the tungstic acid, using
hot 1 % sulfuric acid to transfer the precipitate to the filter.' Place a 150-ml.
beaker under the funnel, puncture the paper, and wash through most of the
tungstic acid with water. Dissolve the remainder of the material on the filter
with hot 4% sodium hydroxide Add 5 to 10,ml. more to the original beaker
to dissolve any precipitate that adheres to the glass, using only about 15 ml. in
all. It is convenient to have on hand a stock solution of ferric ammonium alum
(345 g. of the salt per liter) of such strength that 25 ml. of the solution will
contain 1 g. of iron and 1 ml. of sulfuric acid, sp.gr. 1.83. To the main filtrate
from the tungstic acid, contained in a 600-ml. beaker, add 75 ml. of this stock
alum solution (equivalent to 3 g. of iron) and, stirring constantly, pour in the
sodium tungstate which will contain some ferric hydroxiq~ in suspension. A

18

Willard and YDung, Ind, Eng. Chern" 20, 764; 769 (1928).

POTENTIOMETRIC METHODS

2317

clear solution will result, and the volume of liquid at this point should be about
200 ml. Add 6 ml. of hydrochloric acid, sp.gr. 1.18, if a 1.5 g. sample of steel
was' used, or 4 ml. if a larger sample than 2 g. was taken, and 1.5 to 2 g. of
potassium bromate. From this point the procedure is as given in the preceding
section.
NOTE.-If undecomposed carbides are found in the foregoing procedure, the tungstic
acid may be fused 'with sodium peroxide in an iron crucible. After extracting with
water and decomposing excess peroxide, and addition of 0.1 g. of sodium sulfite to reduce
any chromate present, the solution may be returned to the main filtrate to which additional iron has been added.

Purity of Bromate.-Chlorate is frequently present; by blank tests it was

found that as much as 0.12% is not harmful; 0.25% of potassium chlorate, however, caused a blank correction of 0.12 to 0.15 mg. vanadium. Test of bromate:
add to 200 ml. of water in a 600-ml. beaker, 3 ml. concentrated sulfuric acid,
sp.gr. 1.83, 6 ml. hydrochloric acid, sp.gr. 1.18, 5 g. of ammonium sulfate, and
2 g. of the bromate. Boil 10 minutes and cool to rOom temperature. If not
more than 3 drops (0.22 llfl.) of methyl orange (0.02 g. indicator in 100 ml.)
will impart a permanent color to this solution. the material is satisfactory for
the bromate oxidation method.
Chromium plus Vanadium in Chrome-Vanadium-Tungsten Steels.-A 1
to 1.5 g. sample is convenient. Pla:ce it in a 400-ml. beaker, add 40 ml. of water
and 10 ml. sulfuric acid of sp.gr. 1.83. Warm gently until ther:e is nq further
evolution of hydrogen. Boil over a free flame until llalts begin to separate in
order to assist in breaking up carbides, dilute to about 50 ml. and heat until the
salts have dissolved. To the boiling hot solution add nitric acid, sp.gr. 1.42,
drop by drop, until the violent action is over, then 5 to 6 ml. in excess. Boil
down until the tungstic acid separates in considerable amount; dilute to 50 to 70
ml. with hot water, let settle, and filter off the tungstic acid, using 1% sulfuric
acid to transfer the precipitate to the filter. Place a 150-ml: beaker under the
funnel, puncture the paper, and wash through most of the tungstic acid with
water. Dissolve the remainder of the material on the filter with hot 4%
sodium hydroxide. Add 5 to 10 mi. more to the original beaker using only
about 15 ml. in all. To the main filtrate from the tungstic acid, contained in a
600-ml. beaker, add 1 g. of iron in the form of stock ferric alum solution (see
preceding procedure), and, stirring constantly, pour in the sodium tungstate
solution. A clear solution will result and the liquid should be diluted at this
point to 300 ml. Heat to boiling, add 10 ml. of silver nitrate containing 2.5 g.
of AgNO g per liter, and 5 g. of ammonium persulfate. If no permanga~ate
tinge appears in the solution on boiling, a little more persulfate must be added.
Boil the solution 10 minutes to decompose excess of persulfate. Add 5 ml. of
1 : 3 hydrochloric acid to reduce permanganate, and boil vigorously for 10
minutes to remove chlorine. Cool in icc t.o 5 C., add 2!'i ml. of icc-cold r.;nlflll'ii;
acid, sp.gr. 1.5, and tit-rat!' electrolllctrieally with 0.1 1\ ferrom; i'lulfatc.
Chromium plus Vanadium in Chrome-Vanadium Steels.-A 2 g. sample is
convenient when the steel contains not more than 2% chromium; for samples of
larger chromium content a 1 g. sample is suitable. Place it in a 500-ml. beaker,
and dissolve as directed in the preceding procedure. After completion of
solution and treatment with nitric acid, and expulsion of oxides of nitrogen, the

.'

2318

POTENTIOMETRIC METHODS

solution is diluted to 300 ml. and oxidized with persulfate in the presence of
silver nitrate, etc., exactly as described in the preceding section.
If no vanadium is present, the latter two procedures give chromium alofle.
If both chromium and vanadium are present, the vana~ium alone is determined
by the bromate oxidation procedure, and hence the amount of chromium can
be determined by difference after the determination of the sum of chromium
and vanadium by the appropriate one of the latter probedures.

---,,~SIMPLIFIED

TITRATION ME,THODS

Thus far we have considered in detail only the classical potentiometric

method, using indicator and reference electrodes. Mention has been made of
the Kelley single deRectlOn.end-pomt (p. 2315). Still simpler procedures are
p~~~.
.
Pinkhof-Treadwell Method. 19-The titration cell consists of the usual
indicator electrode, and a specially constructed reference electrode which has
the single potential that the indicator electrode will assume at the end-point.
For ,example, a silver-silver chloride electrode covered with a sodium nitrate
solution might serve in titrating chloride with silver, or vice versa. No
potentiometer or external source of current is needed.
The two electrodes (silver wire, and silver-silver chloride
half cell) are joined in series with a tapping key, resistance, and galvanQmeter. At the end-point tlie galvanometer shows zero deflection, and the direction of the
deflection reverses as the end-point is passed. Muller
made this procedure general by the device of opposing
an E.M.F. equal and opposite to that known to be
developed by the usual indicator-reference electrode
system at the end-point.
The Differential Method.-If one of two similar
metal electrodes is sheltered by some device that with:"
---holds a small part of the solution from mixture with the
FIG. 333.-Differen- remainder, then with s.rp.aU a'dditions of reagent near
tial Apparatus. Con- the end-point, followed by allowing the sheltered por-..
structed from a medi- tion of the solution to mix with the rest, the differences
cine dropper and two
platinum wires. After in E.M.F. between electrodes plotted against small
each reading near the increments of reagent gives a sharp maximum at the
end-point the sheltered end-point. 20 The extremely simpleelectrode system of
portion is forced into Hall, Jensen and Backstrom is shown in FIg. 333, and
the solution by squeezthe more elaborate ones of MacInnes and Dole, and of
ing the bulb.
Clarke and Wooten are represented' in Fig. 334. The
I- 19 J. Pinkhof, Dissertation, Amsterdam, 1919. Treadwell and Weiss, Helv. Chim.
Acta, 2, 680 (1919).
I
20 Cox, J. Am. Chern. Soc., 47, 2138 (1925); MacInnes and Jones, ibid" 48, 2831
(1926); MacInnes and Cowperthwaite, ibid., 53, 555 (1931); MacInnes and Dole, ibid"
51, 1119 (1929); Roth, Z. Elektrochem., 33, 127 (1927); Heckzo, Z. analyt. Chern., 73,
404 (1928); 74, 289 (1928); 75, 183 (1928); Kamienski, Bull. Acad. Polonaise des Sci.
et. Lettres, p. 33 (1928); Millier and Kogert, Z. physik. Chern., 136,448 (1928); Hall,
Jensen and Backstrom, J. Am. Chern, Soc" 50, 2217 (l92_8); Cltrke and Wooten, J.
Phys. Chern., 33, 1468 (1929); Shukoff and Gortikoff) Z;-Elektrochem., 35, 853 (1929);
Rabinovich, Z. Elektrochem., 34, 311 (1928).
,f.;:

POTENTIOMETRIC METHODS

2319

latter (Fig. 334, (2)) made possible the titration of 0.001 N acetic acid with
equivalent baryta solution with a mean deviation of 0.8%, using quinhydrone

1,
2
FIG. 334.-Differential Apparatus. (1) MacInnes and Dole's apparatus. Inert
gas passes through G, and operates the lift pump,. L. The bottom hole of tube A is of
capillary dimensions; opening, H, is a safety outlet which functions when the pump
operates too rapidly. (2) Apparatus of Clarke and Wooten. C is a sheltered electrode.
When tap A is closed, the solution about the sheltered electrode may be pumped into
'
the main solution by means of the rubber bulb, B.
()i..

oj-

FIG. 335.-A. MUller's Capillary Reference Electrode. So little liquid remains in

the capillary that no correction need be applied in ordinary work. B. Burette electrode
of Willard and Boldyreff.
.

POTENTIOMETRIC METHODS

2320

electrodes. MacInnes and Dole have shown that great precision (0.003%)
may be obtained in the differential method as applied to argentometric
titration.
\
Closely allied to the simplified differential method 3;1"e the capillary electrode
of Muller,21 Fig. 335A, and the burette electrode, Fig. 335B, of Willard and
Boldyreff. 22 These devices do not, however, give qifferential readings, I but
rather the cust'omary form of titration graph which has a maximum jump in
potential at the end-point.

Bimetallic Electrodes. (a) Dissimilar Electrodes.-A number of studies

have been made of the use of pairs of metallic electrodes in neutralizations. 23 A
large number of possible electrode pairs have been studied; the most satisfactory, according to Holt and Kahlenberg are tungsten-natural graphite,
antimony-natural graphite, and tungsten-silver. These pairs were applieq in
the titration of 0.1 N acids with 0.1 N bases.
The application of dissimilar metallic electrodes in the field of oxidationreduction reactions is largely the work of Willard and Fenwick. 24 They found
that a system composed of platinum, or platinum-rhodium, with tungsten gave
excellent results. There is a sharp break in the E.M.F. VB. ml. curve at the
end-point. After th,e break the two electrodes may stay at different potentials,
or come together, depending upon the nature of the reaction. Furman and
Wilson 25 found that the platinum-tungsten ,system if joined in series with a high
resistance and a galvanometer, gave a maximum galvanometer deflection at
, the end-points of many oxidation-reduction reactions (Fig. 336).
Polarized Systems.-If two platinum electrodes, as nearly as possible
identical, are polarized with a very small currel'lt, then at the end-points of
R

Pt.

FIG. 336.-Furman and Wilson's Apparatus. The less sensitive the galvanometer, G, the smaller the resistance, R,
which is a radio grid leak. For a galvanometer of 0.05 microamp. sensitivity, R
should be about 1 megohm.

FIG. 337.-Fo~k and Bawden Apparatus. E, E, electrodes;_B, dry cell, B,

switch, R, re~istance (the resistance between the electrodes must. be a. small portion of the total). G is a lamp and scale
galvano,meter of about 0.05 microampere
per m11l. sensitivity.

21 Muller, Z. physik. Chem., 135, 102 (1928).

Willard and Boldyreff, J. Am. Chem. Soc., 51, 471.(1929).

22

23 Brunnich, Ind. Eng. Chern., 17,631 (1925); Fuoss, Ind. Eng. Chern., Analyt. Ed.,
1, 125 (1929); Kahlenberg and Krueger, Trans. Am. Electrochem. Soc., 56, 201 (1929);
Holt and Kahlenberg, ibid., 57, 361 (1930).
24 Willard and Fenwick, J. Am. Chern. Soc., 44, 2504 (1922); Fenwick, Dissertation,
University of Michigan, 1922; Hostetter and Roberts, J ..Am. Chern. Soc., ~l. 1337
(1919); VanName and .Fenwick, J. Am. Chern. Soc., ~~\I(1925).
25 Furman and Wilson, J. Am. Chem. Soc., 50, 277' (1923).

POTENTIOMETRIC METHODS

2321

oxidation-reduction reactions there is a sudden depolarization phenomenon,

and the electrode pair suddenly develops a large difference in potential (loa to
200 millivolts) .26 This gives a sharp indication of the end-point which is
limited to a small volume of standard solution, and hence the end-point may be
over-stepped. Foulk.and Bawden 27 have devised a simple continuous-reading
8

GOODE (1922)

III H11---...,-----,

TREADWELL (1925)

IFIG. 338.-Thermionic Vacuum Tube Circuits. (1) Goode's circuit. Rs serves to

balance out the residual plate current, and resistance R2 serves as a shunt for the gal-vanometer. (2) Treadwell circuit. A small voltage is put in series with the titration
cell to operate on a favorable portion of the plate current-grid voltage line. The indicating instrument is ~ millivoltmeter of the kind used in the:mocouple work.

system with polarized platinum electrodes (Fig. 337). The end-point is a

sudden shift of the galvanometer needle either from zero to a .large deflection
or vice versa, after which it assumes a stationary position; it is, therefore, called
the" dead-stop" end-point.
Use of Vacuum-Tube Voltmeters.-The titration cell may be. bridged
between the negative end of the fiiament circuit in such a way that the negative
26 Willard and Fenwick, J. Am. Chern. Soc., 44, 2516 (1922); V:mName and Fenwicl< I
ibid., 47, 19 (1925).
.
27 Foulk and Bawden, J. Am. Chern. Soc., 48, 2044 (1926).

2322

POTENTIOMETRIC METHODS
I

end of the titration cell is attached to the gdd of the radio tube. A sensitive
milliammeter, millivoltmeter, or galvanometer in the plate circuit then gives
readings that are proportional to the E.M.F. of the titration cell, if the conditions are such that the plate current is a linear functiqn of the grid voltage. 28
Diagrams of two simple circuits are shown in Fig. 338.1 The circuit described
by Treadwell (Fig. 338 (2)), is easy to construct and ~perate. For industrial
use at intervals over long periods of time it is qesirable to have a rugged switchboard type of milliammeter, range 0 to 0.5 milliamp. as indicating instrument,
and a "B" battery of small lead storage units (cf. Ehrhardt, loco cit.). Many
commercial vacuum-tube voltmeters are now available. They have the advantage of compact construction, since batteries are eliminated by the use of a
"power-pack" which enables one to 'plug the instrument into the A.C.lighting
circuit and make continuous-reading titrations. 29
Micro-methods.-The potentiometric method offers certain advantages in
micro determinations, particularly in oxidation-reduction methods, and in 'certain precipitation processes. 30
28 Goode, J. Am. Chern. Soc.;.r44, 26 (1922); 47, 2483 (1925); J. Opt. Sci. Rev. Sci.
lnsts., 17,59 (1928); Calhane and Cushing, Ind. Eng. Chern., 15, 1118 (1923); Treadwell,
Helv. Chim. Acta, 8, 89 (1925); Bienfait, Rec. Trav. Chim., 45, 166 (1926); Ehrhardt,
"Chern. Fabrik," pp. 443, 455,' 463 (1929); Gelbach and Compton, Ind. Eng. Chern.,
Analyt. Ed., 2, 396 (1930).

29 Some typical circuits 'and further literature references are given by: Partridge,
J. Am. Chern. Soc., 51, 1 (1929); Fr. MUller, :Trans. Electrochem. Soc., 62, 335. (1932);
Compton and Haring, ibid., 62, 195 (1932); Willard and Hager, In,d. Eng. Chern., Anal.
Ed., 8, 144 (1936); Garman and Droz, ibid., 1,341 (1935).
.
30 Nussberger, Dissertation, Zurich, 1924, p. 40; Benedetti-Pichler, Z. analyt. Chern.,
73, 200 (1928); Nieder! and Muller, J. Am. Chern. Soc., 51, 1356 (1929); Zintl and Betz,
Z. analyt. Chern., 74, 330 (1928).

CONDUCTOMETRIC METHODS!
1.

THE PRINCIPLES OF CONDUCTOMETRIC TITRATIONS

Electrometric titrations can be divided into two groups: potentiometric

methods, discussed in a previous chapter, and conductometric procepures.
The theory underlying. these two groups is el!tirely different. The theory of
potentiometric titrations as has been pointed out is very similar to that 6f
ordinary titrations, insofar as the change of the potential of a suitable electrode
is a linear function of the change of the logarithm of the ion concentration or
the logarithm of the ratio of concentration of oxidant to reductant in the system
to be titrated. A large jump in potential at the equivalence point means a
sharp color change of a suitable indicator; and the titration to a definite potential is comparable with an ordinary titration in which the reagent is added until
the indicator has assumed a definite color. Therefore, the electrode in a
Rotentiometric titration can be more or less compared to a specific indicator for
the ion or redox system to be titrated:
In conductometric work, on the other hand, all ions present contribute to
the electrical conductivity of the solution. If an electrolyte is added tq a
solution of another electrolyte without changing the volume to any appreciable
extent, the conductivity increases insofar as the electrolytes do not react with
each other. If an ion of one electrolyte unites with an ion of the other to form
a slightly dissociated or slightly soluble substance or if it changes the total ion
concentration by an oxidation or reduction process, then the conductivity of
the solution may change in three different ways, before the equivalence point
has been reached: (1) The conductance decreases. (2) The conductance
remains unchanged. (3) The conductance increases.
The conductance of various ions for the electric current is different. It is
usually expressed by the mobility of the ion. The equivalent conductivity
A of an electrolyte BA is equal to the sum of the mobilities of both ions:

ABA=AB++AA-.
The equivalent conductance A is the conductance in reciprocal ohms of a solution containing one gram equivalent of solute when plac~d between electrodes
which are one cm. apart; hence, it is equafto the specific conductance divided
by the concentration, the latter being expressed in equivalents per m!.
K

A=C1000 .
denotes the specific conductance of the solution, c the concentrati9n expressed in equivalents per liter. The equivalent conductance in other 'words
is the specific conductance the solution would have if the electrolyte were
present in a concentration of one equivalent per m!. On account of the interionic effect the equivalent conductance or the mobility of the ions decreases
1 By 1. M. Kolthoff, Professor of Chemistry, University of Minnesota.
2323
K

CONDUCTOMETRIC MET,HODR

2324

with increasing electrolyte content of the solution. The values reach a maximum at infinite dilution. In dilute solutions the \following relation holds
between the equivalent conductance Ac at a concenttation c and A",:
I

A is a constant which is different for various electrolytes. The mobility of the

ions increases very much with the temperature, as their migration velocity
increases. The increase of the mobility of most ion" is 2 to 2.5% for 10 C.
increase in temperature, for hydrogen ions the change is only 1.5% for hydroxyl
ions 1.8% per one centigrade degree change in temperature.
In the following table the mobility of some ions at 25 0 and at infinite dilution
are reported. The data are taken from reliable studies reported in the
literature.
Mobility of Some Ions at 25 0 at Infinite Dilution
Li+
Na+
Ag+
K+
NH 4+
H+
!Ba++

!Ca++

!Mg++
!Pb++
Tl+
!Ni++
!Fe++
!Fe+++

41.7

193

50.8
63.4
74.8
74.9

75.8

70.9
47
39.6
40.8

35D

65.2
61
55.0

70

73.5 .
80.0

71.0

97.3

76.0
53.6
54

100.8

68.4

Let us now consider how ~he conductatic..e of a solution of a strong electrolyte

BA will change upon the addition of a reagent CD, assuming that the cation B+
reacts with the anion D- of the reagent. If the product of reaction is slightly
dissociated or in~oluble, the reaction may be expressEjd by the equation:

-+ C++ D--->BD+ A -+ C-

B++ A
Ion to be
determined

reagent

insoluble or
slightly
dissociated.

--

Therefore, by the reacti~n between B+ and D- ions, the B IOnS" during the
titration are replaced by C ions.
.
Case 1.-The mobility of the B ions (AB) is greater than Ac: the conductance
of the BA solution decreases upon' the addition-of the reagent CD. This case
generally occurs in the titration of strong acid with sttong bases or in the reverse
titration. The hydrogen and hydroxyl ions distinguish themselves from the
other ions by a much larger mobility. (See table, above.)
Case 2.-A8 and Ac are equal. The conductance remains unchanged by
the addition of CD until the equivalence point has been reached. This case is
met with in most precIpitation reactions. In the titration of,silver nitrate with
barium chloride, the barium ion takes the place of th~silver ion and as both ions
have about the same mobility, the conductivity{loes not change during the

CONDUCTOMETRIC METHODS
reaction. If instead. of barium chloride, sodium chloride is used as a reagent,
the conductivity decreases slightly as AAg+ is greater than ANa+. On the other
hand with potassium chloride as a reagent, the conductivity increases slightly
because AAg+ is smaller thauAK+.
Case 3.-The conductivity increases from the beginning of the titration if a
slightly dissociated substance is titrated and the reaction product is a strong
electrolyte. This case in general occurs in the neutralization of a weak acid.
with a strong base or a weak base with a strong acid. In any case the conductivity increases after the equivlJ,lence point has been reached, at least if the
reagent is a strong electrolyte ..
In a conductometric titration the conductance is measured after addition of .
successive amounts of reagent. The points thus obtained are plotted to give a
graph which as a rule consists of two straight lines intersecting at the equivalence
point. Therefore, the latter is found in a graphical way. In contrast to any
other type of titration method, measurements near the equivalence point have
no special significance. As a matter of fact the values fo(lnd near the equivalenlle point are often worthless in the construction of the two straight lines,
on account of the fact that the reaction product by its dissociation or solubility
contributes to the conductivity of the solution whereas we have to know the
data where the conductivity caused by the reaction product itself is negligibly
small. In reactions which can be made the basis of a conductometric titration,
the latter is always the case, when there is an excess of ion to be titrated or
reagent. Near the equivalence point the points often do not lie on one of the
two straight lines, but the conductivity found is higher than the corresponding
ones on the straight lines (titratiop of very weak acids and bases; hydrolysis;
precipitation reactions); compo Figs. 346, 347, and 348.
The point mentioned that a mar~ed hydrolysis, solubility or dissociation
of the reaction product does not affect the accuracy of the method very much
makes the application of conductomet~ic titrations possible to those cases where
other titration methods jdil to give results. This will be shown oy severa! examples
in the following review.
On the other hand, it is emphasized here that the conductometric method
can be much less generally applied than the ordinary or potentiometric one, on
account of the fact'that large amounts of foreign electrolytes which do not take
part in the reaction, affect the accuracy greatly. The relative change of the
conductivity during the reaction and upon addition of excess of reagent mainly
determines the a!lcuracy and this change is decreased by the presence of foreign
electrolytes. Of course one must not infer from this that the conductometric .
method is rendered impossible by the presence of foreign electrolytes; if precision methods are used in the measurement of the conductivity and the titration is carried out in a thermostat, the method can be applied in the presence of
relatively large amounts of indifferent electrolytes. This, however, is reached
at the cost of. the simplicity of the method.
2.

THE PERFORMANCE OF CONDUCTOMETRIC TITRATIONS

For details /tbout the measurement of electrical conductance and the

significance of the cell constant the reader is referred to textbooks on electrochemistry.
.

CONDUCTOMETRIC METHODS
An ordinary conductometric titration can be carried out in a relatively short
time (ten minutes or longer).
f
A titration cell as given in Figs. 339 or 340 can be, used. Figure 339 gives' .
the more or less classical model of Dutoit. The ~wo platinized platinum
electrodes are in a vertical position in order
to prevent the dep,osit of a precipitate in
precipitation reactions. The electrodes are
welded to a platinum wire, the latter being
sealed in the glass and making electrolytic
contact outside the vessel in a mercury
pocket. The cell is placed in a block of
wood or . paraffin. Successive portions of
the reagent are added from a microDuret
which can be connected in some way to a
reservoir containing a supply of the standard solution. It is r~commended to use a
reagent which is at least ten to twenty times
more conceiltra;ed than the solution to be
titrated. Th-e cell in Fig. 339 also contains
a thermometer divided into 0.1 degrees. A
change in temperature during the titration
will effect the result very much as the conductivity increases greatly with increasing
temperature (for most salts about 2 to 2.5%
for 1). In cases where the heat of reaction
... is fairly high, irregularities in the conductivity curve may occur on account of the
temperature effect. As a rule, however, the
heat effect during a titration is very small.
After the addition of reagent the cell is
shaken to secure /homogeneous mi~ing; in
this manipulati'on one has to be careful not
to warm the solution by taking the whole
FIG. 339.
cell in his hand. Convenient for the shaking is a glass handle attached to the cell
as in Fig. 340. If work is done with solution's of quite different conductance,
one should have a few titration cells with different cellconstants adapted to
the special work. The accuracy in the location of the end-point in conductometric titrations as a rule is not larger than 0.5 to 1 %. For precise work it is
recommended that the titration be performed in a thermostat.
It is not necessary to know the'absolute values of the specific conductivity
of the solution during the titration; the reciprocal value of the resistance can be
plotted as it is proportional to the conductance. '
The conduotance can be measured according to the classical method of the
Wheatstone bridge, using a telephone for the detection of the minimum.
Though this method is quite satisfactory, it has a special dis)tdvantage, namely,
that one needs a quiet room, undisturbed by nois~~- This requirement might
prevent a more general application of conductcrmetric titrations. Fortunately,

CONDUCTOMETRIC METHODS

2327

at the present time the telephone can be replaced by other instruments or

arrangements so that the minimum can be detected in a visual way.
In the first place the Leeds .and N orthrup alternating current galvanometer may
be mentioned. The equipment provides
means for utilizing current from a sixty
cycle 110 volt A.C. circuit and the galvanometer is especially recommended for
conductometric titration work. It is not
suitable for highly precise measurements
and even in titrations some difficulties
may arise by polarization of the electrodes and the heating of the solution
in the cell. If these difficulties are overcome, it seems that the A.C. galvanometer will furnish us with the simplest. apparatus for detection of the
point of balance on the Wheatstone
bridge. Of other methods recommended,
I
those advocated by G. Jander and O.
'-'--~_'
Pfundt 2 will be bri.efly discussed. They
'--_
_
--1
.
FIG. 340.
use a t h ermo-cross (Fig. 341) or a speclar
commutator (crystal detector) (Fig. 342 as rectifiers. In Fig. 341 the
thermo-cross, 1'h, and the transformers are placed in the Wheatstone bridge
instead of the telephone. The thermo-cross consists of a couple of constantan

FIG. 341.

and iron wire kept in a small box to protect it from atmospheric disturbances.
If a current passes through the bridge a thermo current is developed and the
2 G. Jander and O. Pfundt, "Die visuelle Leitfahigkeitstitration," 2d Ed., Stuttgart
1934, Verlag Enke, also in "Physikalische Methoden der Aualglischen Chemie," Akad.
Verlagsges. Leipzig, 1936, pp. 1-54.

2328

CONDUCTOMETRIC METHODS
,

galvanometer G gives a deflection. In order to prevent the direct current from

passing through the conductivity cell, the transforme.'r, T, is placed in the circuit. R is a resistance by means of which
the sensitivity of ~he system can be regulated. W is a \known resistance and
AB a slide wire u~ed as in the ordinary
method. The apparatus can be used in
different ways; many details have\ to be considered, for which the reader is ~eferred to
the work of Jander and Pfundt. Simpler
to manipulate is the system in Fig. 342
s
where U represents the rectifier. It. may
be mentioned that T. Callan and S. Hor3 a few years ago described a similar
robin
VJ
arrangement in which they used a carboQ
rundum detectol" (from the Carborundum
FIG. 342.
Company) as a rectifier, also W. D. Treadwell
and S. Janett 4 made use of a comm~tating device to avoid the use of the
telephone.
3.

APPLICATION OF CONDUCTOMETRIC TITRATIONS TO

AQID-BASE ~EACTIONS

In this paragraph a condensed review of the application of conductometric

titrations to acid-base reactions will be given. For details the reader is referred
to a monograph of 1. M. Kolthoff.5
Strong Acids with Strong Bases.-In
the following graphs the ordinate represents the conductivity and the abscissa the
volume of reagent added. In the titration
of a strong acid with a strong base (or vice
versa) a sharp break in the conductivity '~
s:
occurs at the equivalence point. Theoret- 't
ically the minimum should not occur exactly at the equivalence point where pH 8
is 7, but slightly at the alkaline side .as the
mobility of the hydrogen ions is much
larger than that of the hydroxyl ions.
From a simple differential equation it can
ml NaOH
be computed that the minimum occurs
at a hydroxyl ion concentration of
FIG. 343.
1.4 X 10- 7 The difference between this
and the neutral point is so small that it cannot be determined experimentally.
It is of interest to note that the shape of the conductivity line is independent
3 T. Callan and S. Horrobin, J. Soc. Chem. Ind., 4'7, T329 (1938).
Compare also
H. T. S. Britton, "Conductometric Analysis," D. Van Nostrand Company, Inc., New
York, 1934.
I
4 W. D. Treadwell and S. Janett, Helv. chim. act., 6 734 (1923).
51. M. Kolthoff, "Konduktometrische Titrationen;~' Dresden, 1923, Verlag Steinkopff.

.g

CONDUCTOMETIUC METHODS

2329

of the dilution. Extremely dilute solutions of strong acids or strong bases, of

the order of 0.0001 N, can be titrated with the same accuracy as more concentrated solutions if care is taken to exclude carbon dioxide.
On the whole, the application of the
o
conductometric titration in this case has
only practical significance when the solution has a dark color so that the indicator metho{1 fails to give results.
Intermediate Weak Acids and Bases.The shape of the neutralization curve
depends upon the concentration and the
ionization constant of the acid or base.
By neutralization of the dissociated part
of the acid, the. conductivity will drop
and on account of the formation of' its
salt which behaves as a strong electrolyte,
0.1 N
~ ml NaOH
the cpnductance increases. The practical
FIG. 344.
neutralization curve is obtained by adding the figures of the "acid-depression
curve" to those of the "salt line." The more strongly the acid is ionized,
the more its neutralization curve approaches that of a strong acid and the
less it is ionized the more it behaves like an extremely weak acid; the latter
case will be discussed later.
In Fig. 344 the neutralization
curves are given for the neutralization of 0.1 N, 0.001 Nand 0.0001 N
'acetic acid, respectively. From the
above it is evident that in manf'
cases a flat minimum will occur in
the neutralization curve. Its location can be calculated from the
ionization constant, the concentrab
tion and the mobilities of the ions
~o
present in the system. The minimum itself has no analytical signifi'C
"C
o
cance though it may give us an
U
indication of the magnitude of the
ionization constant of an acid which
_ _ ml
may be of importance in unknown
cases. If the acid is relatively
FIG.
highly ionized, the neutralization
ABE-Neutralization Curve.
line will give a curve up to the
ED-8odium Hydroxide Curve
equivalence
point and it is hard to
OEC-Salt Line.
find with any degree of accuracy the
AF-Acid Depression Curve.
point of intersection of the neutralization curve with the sodium hydroxide line, the latter being the straight line
found after addition of an excess of base.
Such a case occurs, for example, in the neutralization of 0.01 N salicylic acid
as is' shown in Fig. 345. An analysis of the curve shows that the equivalence
N~OH

2330

CONDUCTOMETRIC METHODS
J

point is found at the point of intersection of the salt, line and sodium hydroxide
line. The former can be determined in an experiment!l] way. Suppose 100 mI.,
0.01 N salicylic acid are titrated witli 0.5 N sodium hydroxide so that the sodium
hydroxide line can be drawn. In order to find the sha~e of the salt line a second
determination is made if? which 100 ml. of water instead of dilute acid is taken
'and successive amounts of 0.5 N sodium salicylate of exactly the same strength
as the sodium hydroxide are added. From the latter data, the salt line can be
constructed and the equivalence point can be found with an accuracy of at least

1%.

If the dissociation of the acid during the neutralization is less than in the
above case, the last part of the neutralization curve, before the equivalence
point will be a straight line. Under these conditions, the point of intersection
can be found by a single titration ..This is the case, if the ionization constant
of the acid is smaller than about 5 )10- 4 in the titration of 0.1 N solutions,
smaller than 5 X 10- for 0.01 N solutIOns and smaller than 5X10-6 for 0.001 N
solutions.
Very Weak Acids and Bases.-The ini~ial, conductivity is very small. and
during the neutralization, the conductivity i~creases according to the salt line.
On account of the hydrolysis of the salt formed, the experimental data near the
equivalence point are higher than the con'esponding points on the salt and
sodium hydroxide line; therefore, in the construction of the two lines, points
should be taken at .such a distance from/the equivalence point (at the hydrol-

FIG.

346.

ysis is negligibly small and the points can then be combined to give straight
lines. If the acid is extremely weak (e.g. hydrogen peroxide) the hydrolysis is
so large that the titration does not yield usefui results. For 'the titration
of 0.1 N solutions, the ionization constant should be larger than about 10-1,
for 0.01 N solutions larger than 10-9 , for 0.001 N solutions larger than 10-8
These figures have no exact significance; they only giv~ the order of magnitude
of the ionization constant at which useful results can still be obtained.
From the above we see that the conductoIDet.ric titration can be applied
where the potentiometric or ordinary method ~Qes'not give satisfactm-y.results.

CONDUCTOMETRIC METHODS

2331

In Fig. 346 neutralization curves of 0.1 Nand 0.01 N boric acid with sodium
hydroxide are given. Several applications can be made. Hydroxy benzenes,
like phenol and resorcinol can be titrated very nicely. A systematic study
has shown that resorcinol and hydroquinol behave like dibasic acids, pyrocatechol on the other hand as a monobasic acid. The trivalent phenols,
pyrogallol and phlorglucinol beha;ve like dibasic acids. The method has been
applied to the titration of nitro phenols, phenolphthalein and other weak acids
whose salts are colored. Phenolphthalein in .50% alcohol behaves like a
dibasic acid, a break occurring after the neutralization of the carboxyl and
phenol groups. The conductometric method should be very useful to organic
chemists in cases where they want to know quantitatively the acid or basic,
character of a colored substance. It can also be recommended for the deter'mination of the equivalent weight of amino acids and polypeptides and the acid
or base combining power of proteins. A practical application of the method
can be made in the determination of vanillin in vanillin sugar. The aromatic
substance is extracted with alcohol and titrated as a monobasic acid.
What has been said for very weak acids also holds for very weak bases.
Aniline, hexamethylenetetramine and pyridine can be titrated accurately with
hydrochloric acid.
Neutralization of Weak Acids with Weak Bases.-Though in analytical
work one always uses a strong acid or a strong base as a reagent, the' above
case has practical significance; e.g. when a pure am,monium salt of a weak acid
has to be prepared or when a weak acid must be titrated in the presence of an
ammonium salt or the salt of another weak base.
Figure 347 gives the titrationlines of 0.1 Nand 0.01 N acetic acid, respectively with ammonia. The neutralization curve up to the equivalence point is

____

ml NHa

FIG. 347.

about the same as found with sodium hydroxide since both sodium and ammonium acetate are strong electrolytes. After the equivalence point has been
reached, an excess of ammonia leaves the conductivity practically unchanged
as the dissociation of the weak base IS depressed by the presence of the ammonium salt. On account of the hydrolysis, the values .near the equivalence
point are somewhat lower than the corresponding points on the straight lines.
If the acid or Ibase to be titrated is extremely weak (e.g. boric acid with ammonia), no straight lines can be constructed on account of the strong hydrolysis.

2332

CONDUCTOMETRIC METHODS

Mixtures of a Strong and a Weak Acid.-Here again the conductometric

method can be advantageously applied whereas the'ortlinary or potentiometric
method does not yield satisfactory results. The above case is a combination
of two which have been discussed before. First, the ~trong acid is neutralized
and the conductivity drops following a straIght line. 'Near the first equivalence
point, the weaker acid will be neutralized and the conductivity will soon increase according to the salt line of the
weak acid. In Fig. 348, the change in
conductivity during the neutrdlization
of a mixture of 10 ml. 0.01 N \hydrochloric acid, 10 m!. 0.01 N acetic acid
and 10 ml. of water with 1 N sodium
hydroxide is given. The case has practical significance; e.g. for the determination of traces of mineral acids in vinegar.
Recently the author, in collaboration with
Mr. T. ;Kameda, has made another application for the determination of the
purity of sulfonphthaleins. The sulfonic
group behaves as a strong acid and'is
first. neutralized; after the first break, the
,
conductivity increases and after neutraliml NaOH
zation of the phenolic group, a second
FIG. 348.
break occurs.
The application of the conductometric method to the titration of a mixture
of two weak acids with quite different ionization constants (like acetic and
boric acid) 'does not have much practical significance, for a sharp break never
occurs after the neutralization of the stronger acid. The difference in mobility
of the anions of the two acids determines the sharpness of the break and as this
difference always is very small, the angle between the two salt lines will be very
obtuse. Therefore, as a rule. tlie first equivalence point cannot be found with
,
a high degree of accuracy.
Replacement Titrations.-When a salt of a weak acid is titrated with a
strong acid, the anion of the weak acid is replaced by that of the stronger one
and the weak acid itself is liberated in the undissociated form. Similarly, the
addition of a strong base replaces the weak base in a salt-of-the latter.
Many important applications of these replacement reactions can be made
to conductometric titrations; especially those cases are important which do IlOt
yield successful results with the indicator of potentiometric method.
If, for example, hydrochloric acid is added'to a solution of sodium acetate]
the acetate is replaced by the chloride ion. The cpnductivity increases slightly
on aceount of the fact that ;\Cl- is a little greater than ;\acetate. After all the
acetic acid has been liberated, continued addition of hydrochloric acid gives
:rise to a strong increase in conductance. Fig. 349 gives the conductance lines
in the titration of 0.01 N sodium acetate with N hydrochloric acid. On account of the dissociation of the acetic acid, the \):If]!erimehtal figures near the
equivalence point are somewhat higher than ~htr corresponding ones on the
straight lines. If the ionization constant of the liberated acid is smaller than

CONDUCTOMETRIC METHODS

2333

about 5 X 10-5, 0.01 N solutions of its salt can be titrated accurately. For
0.1 N solutions, the constant may even be as large as 5 X 10-4 Ethyl alcohol
decreases the ionization of weak acids and in the presence of enough alcohol
even salts of stronger acids than
those indicated can be titrated.
The method furnishes a simple
means for the evaluation of salts
of }veak acids, such as acetates,
benzoates, succinates, etc. Similarly, it can be applied to the ~
determination of ammonia in am- 1;
monium salts by titration ,vith .g
sodium hydroxide. A rapid determination of the ammonium content
of fertilizers can be made in this
way .
. In this review, a condensed
summary has been given Of the
application of conductometric ti~ ml H Cl
trations to acid-base reactions.
FIG. 349.
For details concerning the combination of different cases and specific cases (carbon dioxide, phosphoric
acid, alkaloids, phenols, salts of heavy metals, etc.) the reader is referred to the
monograph of Kolthoff (1.c., p. 2328).From all that has been said, it may be inferred that the significance of
conductometric titrations should be more generally recognized. They often
furnish us information which could otherwise only be obtained by elaborate
work.

4. APPLICATION OF CONDUCTOMETRIC TITRATIONS TO

PRECIPITATION AND COMPLEX FORMATION 'ANALYSIS
A reaction usually can be made the basis of a conductometric titration if
the reaction product is a slightly soluble substance or a stable complex. The
usefulness and accuracy are mainly dependent upon the following facts:
(a) Errors in the determination of the conductance data.

(b) Solubility of the precipitate or stability of the complex.

(c) Speed of formation of the precipitate.
(d) Constant composition. (purity) of the precipitate.
(a) The same error in the determination may affect the accuracy of the
titration in different ways. This depends upon the acuteness of the .angle
between the precipitation line (line combining the conductance data during thE)
precipitation) and the reagent line (line giving conductance data with excess of
reagent). The more. acute the angle is the more accurate the result. If the
angle is very obtuse, a small error in the conductance data can cause a large
deviation. Therefore, one always should endeavor to choose such experimental
conditions that the angle is as acute as possible. The following rules should be
borne in mind: (1) The smaller the mobility of the ion which replaces the react-

2334

CONDUCTOMETRIC METHODS

sal~

ing ion, the more accurate will be the result. If silver

is titrated with
lithium chloride, the conductivity decreases during the precipitation ;Z~d
increases after the equivalence point. If hydrochloric acid were used as a
reagent, the conductivity would increase from the peginning of the titration
and the angle between the precipitation and reagent line would be very obtuse.
Therefore, generally it is recommended to titrate qations with lithium salts
and anions with acetates. (2) The larger the mobility of the anion of the
reagent which reacts with the cations to be determined (or vice versa) the more
acute is the angle. For example, it is more advantageous to titrate silver salts
with sodium chloride than sodium nitro prusside, as the mobility of the chloride
ion is larger than that of the nitro prusside. (3) The titration of a slightly
ionized salt does not give good results since the conductivity increases relatively
much from the start of the determination. (4) As has been mentioned in the
beginning of this chapter, the accuracy of a conductometric titration always
suffers from the presence of electrolytes that do not take part in
the reaction.
.
(b) On account of the solubility of the precipitate, -the experimental figures near the equivalence point deviate from the
straight line as is shown in Fig.
350. A E D F C gives the
experimental curve. If the soluB
bility of the precipitate were
negligibly small, the conductivity
_
ml reagent at the equivalence point would
be equal to B G. By an excess
G
of the ion which. is precipitated
FIG. 350.
or an excess of reagent, the solubility is depressed and if the latter is not too large, i1f usually is possible to construct the precipitation and reagent line by joining points on A E and F C.
It can be shown that 0.1 N solutions can still be titrated if the solubility of
the precipitate formed is less than about 0.005 N (uni-univalent electrolyte);
for 0.01 N solutions the solubility should be smaller than about 0.0005 N.
(c) The formation of a microcrystalline precipitate is usually a time reaction. After addition of a small amount of reagent, the-conductance does not
immediately become constant; one must wait for some tir~e: Seeding the
solution with the precipitate itself has a favorable effect though in the titration
of very dilute solutions even this does not overcome the difficulty. Quite
generally it is much better to add enough alcohol to make its concentration
about 30-40%. The speed of formation of the precipitate then becomes much
larger and furthermore another advantage is g~ined, the solubility of most
slightly soluble substances is materially reduced by alcohol. Of course, one
should rea:iize that on addition of alcohol to a solution the temperature rises;
therefore, before the titration is started, the mixture should be cooled to room
I
temperature.
(d) If the precipitate has strong adsorben~ pr(}perties, the method does not
yield successful results. So, for example, the "titration of heavy metals with

CONDUCTOMETRIC METHODS

2335

sodium sulfide or ferro cyanide cannot be recommended on account of the

inconstant composition of the precipitate. If a microcrystalline precipitate
like barium sulfate or calcium oxalate is formed, occlusion phenomena may play
a part. Finally, it may be mentioned that under certain conditions, the surface
conductance of the solid precipitate may affect the results.
Many precipitation or complex reactions have been made the basis of a
conductometric titration. A short summary of most satisfactory cases is
-given below. For details, the reader is referred to Kolthoff's monograph
(I.c., p. 2328).
Silver nitrate as a reagent: chloride, bromide, iodide, cyanide, thiocyanate,
. chromate; molybdate,6 also iodide in presence of chloride in ammoniacal
medium.
Mercuric perchlorate as a reagent (complex formation): chloride, bromide,
iodide, cyanide, thiocyanate, formate, acetate and its homologues.
Lead nitrate as a reagent: iodide, ferro cyanide, ferricyanate, sulfate,
sulfite, thiosulfate, pyrophosphate, oxalate, tartrate, succinate, benzoate.
Bap,um acetate (or barium chloride) as reagent: sulfate, chromate, carbonate, pyrophosphate, oxalate, tartrate, citrate.
.
Sodium perchlorate as reagent: potassium 7 at 0; it is not accurate, according to the writer's experience.
Lithium sulfate as reagent: barium, strontium calcium, lead.
Sodium chromate as reagent: barium, lead, silver.
Lithium oxalate as reagent: silver, lead, copper, cadmium, nickel, cobalt,
manganese, zinc, calcium, barium, strontium, magnesium and strong acids.
Lead acetate as reagent: molybdate, tungstate. s
Bismuthylperchlorate as reagent: phosphate. 9
The author expresses his appreciation to Mr. H. E. Howe, editor of Ind.
Eng. Chem. for allowing him to use part of th~ original manuscript in the
preparation of this chapter. Part of this has been published in Ind. Eng. Chem.,
Anal. Ed., 2, 225 (1930), and in a monograph entitled "The Colorimetric and
Potentiometric Determination of pH. Electrometric Titrations," J. Wiley &
Sons, New York, 1931.
C. Candee and 1. G. Murgulescu, Bull. Soc. Chirn., Romania, 17, 103 (1935).
G. Jander and O. Pfundt, Z. anal. Chern., 71,417 (1927).
8 E. Rother and G. Jander, Angew. Chern., 43, 930 (1930).
9 J. Harms and G. Jander; Angew. Chern., 49, 106 (1936).

GAS ANALYSIS

SAMPLING
As i~ all analytical processes, the obtaining of a representative s~mple is of
cardinal importance. It is done by the insertion of a tube into the duct or flue
and the withdrawal of the gas into a container. Previous to this, the furnace,
duct or setting must be thoroughly inspected for air leaks and these carefully
stopped. This is all the more to be done if, as often happens, the boiler plant
has been built on filled or " made" land. Finally the fact must not be lost
sight of that the bricks themselves composing the setting are porous. The
search for leaks is conducted using a lighted candle, and the places where the
flame is sucked in are marked and plastered up: the joints in the cleanout doors
are often corroded or rusty and no longer fit tightly. This is also true of the
frames of the doors themselves in the brickwor.k If the bricks and mortar of
the setting leak, it should be given a thin coat 6f size and a couple of coa~s of
thin whitewash to close the pores. Few people have any idea of this loss from
air leakage, and by radiation of heat from the setting. A case cited 2 in a
western city in which the setting was covered with asbestos cement over which
canvas was pasted and the whole heavily painted, showed a saving of 12%.
If the sample be from a boiler it should be taken as soon after leaving it as
practicable-not over the grate where the gases might be dissociated, and not
at the end of a long duct where they may have been air-diluted. Bends in the
ducts should be avoided and a place chosen preferably four or five diameters of
the duct away from the bend, where the gases are flowing smoothly. The tube
should be inserted to a distance of one-fourth to one-third the radius of the duct
from the circumference. Multiple tubes, tubes with varying sized holes, are
not to be recommended, ,as they are of doubtful expediency. Insertion is
preferably made on the grate side of the damper, t9 avoia air leaks, by drilling a
hole in the flue and fitting a screwed nipple into the hole: the sampling tube is
cemented into this nipple with Portland cement in case of a permanent installation. When the connection is temporary the hole may be puttied or made tight
in ahY way-even with wet cotton waste. A similar hole is made for the
thermometer.,
"a. Tubes.-The best tube is of translucent quartz, about 74 i~ch'in internal
diameter and 3 feet long: this will withstand any temperature or gas likely to be
encountered in practice. Pyrex glass tubes will probabiy be found sufficient
for all practical purposes and may be used with-temperatures up to 600 0 C.
Uncooled metal tubes cannot be used aboye 250 0 as their oxides are reduced by
hydrogen and carbon monoxide even at 200 0 according to the equation Fe20a
+3CO = 2Fe+3C02, the sample will be defective and an utterly false idea
obtained of the process by which it was produced. Dust and smoke can be
removed by plugs of asbestos, or if the temperature permits, rolls of iron wire
gauze or steel wool. These must be replaced when c!o~ged. I
1 Chapter by the late Augustus H. Gill.
2 Electrical WorId.
2336

GAS ANALYSIS

2337

h. Pumps.-Any form of a mechanical pump may be used, of which the

ordinary rubber syringe bulb is the simplest. In using this it should be remembered that the valves do not shut tight, and in sampling it may have to be shut
off. by a pinch-cock. Where a head of water, steam or compressed air is available, the familiar Richards 3 jet pump (Figs. 351 and 352) may be employed.
This operates on the principle of
YQ\r R the boiler injector-small slugs or
pistbns of water filling the jet and
sucking in the gas after them.
Where a fall rather than a head
/.'R
w of water is available, the Bunsen
pump (Fig. 353) is used. This cona sists of a 7.4:' inch T connected to the
,
water reservoir and to
a 12 ft. piece of 74
inch rubber tubing b,
the stem of the T a
being connected to the
source of water. When
the water flows it sucks
the gas in after it
through c: 12 ft. of
tubing will usually be
found sufficient, 34 ft.
are nerded for the highest vacuum. If none of
I>
the foregoing be available, a pair of aspirator
FIG. 351.
FIG. 352. bottles described further on may be used.
c. Containers.-These should be of glass, or a
metal which is unacted upon by the gases to be
encountered. Rubber bags are unsuitable, for
rubber is by no means the inert substance which
we are in the habit of considering it. Rubber is
oxidized or attacked by various gases and allows
the lighter gases to transfuse through it, and dissolves the heavier hydrocarbons, as ethylene.
FIG. 353.
\cid gases like sulfur dioxide are dissolved per.ceptibly by passing through the usual white laboratory tubing, containing pO
to 60% of zinc oxide, giving low results. In consequence of this, rubber connections exposed to gases in all gas apparatus should be as. short ,as possible,
the glass tubes being butted together, leaving barely room for the pinchcock.
The ends of all glass tubes connected by rubber must be file-rounded, to remove the
sharp cutting edge of the glass tubing-a prolific source of leaks in stoppers as
well as rubber connectors.
The shape of the container should be such as to allow the water in it to
wholly run out; a glass bulb, pear-shaped at the bottom (Fig. 354), is excellent.

Am. J. Sci. (3), 8, 412 (1874).

2338

GAS ANALYSIS'

If this cannot be had a separatory funnel may 'be employed. It is used by

filling with water, attaching the sampling tube at ilthe pump opposite. Upon
pumping, a stream o( gas passes through the T
from which a samplel is taken by allowing the
water to run out completely: h serves as an attachment for water in :dispensing gas from i when
in the laboratory. The tubing shown in the
figure is of lead, which can be safely used for
chimney gases after it becomes attacked by them.
Its obvious advantage is found in the fact that
it bends rather than breaks.
A pair of aspirator bottles-Fig. 355-serving
both as pump and container will be found useful.
These are bottles or glass jugs, each fitted with
a right-angled long tube going nearly to the bottom and a similar short tube ending just under
the stopper. The long tube at its lower end carries a short piece of rubber tubing, extending to
the wall of the bottle. This enables it to be completely emptied of water when tipped. The long
tubes are qonnected by 4-6 feet of ~ inch rubber
tubing with a screw pinchcock upon it. The short
tubes carry rubber connectors and pinchcocks
completing the apparatus. Upon filling one
bottle with water, preferably saturated with the
FIG. 354.
gas, setting it 3 to 4 feet higher than the other
and blowing into the short tube, water flows into
the long rubber tube sucking air into
the bottle after it. The stoppers can
be wired in, taking the precaution to
put a piece of stiff leather or metal over
I
the top of the stopper: this prevents
the rubber stopper from being drawn
away from the glass tubing which would
give rise to leaks. Before using the
bottle into which the sample is to be
taken, it must naturally be completely
full of water. These can be used for
taking continuous samples.
For the collection and pr~servation
of small samples, glass tubes or bulbs
holding about 100 mI., glosed by extra
well-fitting stopcocks or which can be
sealed off, can advantageously be employed.
As a confining liquid, mercury should
be used for results of the highest degree
of accuracy-for our purpose it is a
FIG. 355.
super-refinement. Tap water saturated

GAS ANALYSIS

2339

with the gas sought, is the next best, or tap water alone, or strong brine. In
using the latter it should be remembered that it corrodes metal-pinchcocks,
etc., and salt is hygroscopic.
In this connection. the fact should not be overlooked that water does not
dissolve carbon dioxide instantly-but that time and the surf~ce exposed are
important considerations. In dispensing chimney gas to classes from the
sampling apparatus-Fig. 355--under'a pressure of 4 to 6 ft. of water, the writer
has repeatedly observed the fact that the last sample would often show more
carbon dioxide than the first, even after 15 minutes standing. The converse
of the proposition seems to hold true, as the writer has heated samples of water
10 above the saturation point for 24 to 48 hours without the water giving up

'
the excess oxygen. 4 5
The foregoing procedure applies to sampling gases from a boiler or flue.
To sample city gas, it should be allowed to burn until the old gas is certainly
out of the piping, or it can be blown out into the outside air.

MEASUREMENT AND EXAMINATION OF GASES

Ga~es are analyzed in two ways-by absorption of a certain constituent, and
noting the diminution either in the gas volume or the strength of the reagent
used; and by combustion, where the constituents do not lend themselves to
absorption. The absorption of oxygen from the air by potassium pyrogallate
is an example of the first method and the combustion of methane and subsequent absorption of the carbon dioxide, of the second.
The diminution in volume implies some method of measurement of gas
volumes, which brings us to the consideration of meters.
For the measurement of gas in large quantities meters are used. Several
types of instrument are available for this purpose,

Mes.surement

Permissible
Velocity. Feet
per Second

Volume

(100 r.p.m.)

Wet meter., ......... , ................ .

Dry meter-diaphragm type ............ .
Dry meter-Connersville type ........... .
Pitot tube ....... , .................... .
Rotameter or Thorp gauge ............. .
Thomas electric ....................... .

Velocity

1.6-71
Above 8

Weight
Velocity

Venturi tube .......................... .

Anemometer. . . . . . . . .................. .

"
"

3 to 500
Unlimited

Orifice ............................... .

"
"
"

"

1 to 40

Limit of
Accuracy,6
...

0.5
2.0
_7

0.5
2.0
1.0
1-5

Df all these instruments, the wet meter and Thomas meter are probably the
most accurate. They, however, like some of the others, cannot be used with
corrosive gases, being constructed of me~al; hence the Pitot tube and rotameter
are the ones to be employed in chemical works.
Technology Quarterly,S, 250 (1892) .
See also Ind, Eng, Chern., AnaL Ed., 7, 37 (1935).
6 Taken partly from Marks' "Mechanical Engineers' Handbook."
7 Said to be accurate.

4
. 6

2340

GAS ANALYSIS

The wet meter consists of a cylindrical drum divided into fom spiral compartments, suspended in a bath of water, surrounded !by a tight casing: the
pressure of the gas causes the drum to rotate, emptying a drumful of gas into
the casing and pipes. It must be set level, the water level carefully adjusted
to the mark on the glass gauge, with its top open; as wel~ as the, inlet and outlet
pipes of the meter. The higher the water level the faster the meter. The
dry meter consists of a pair of metal bellows, with sides of leather soaked in oil,
on either side of a diaphragm, and connected with slide valves so that a bellowsful of gas is alternately emptied into the upper part of the meter and piping.
The vibrations of the bellows produced by the gas pressure are transmitted to
clockwork and indexes. It is to be noted that the indexes apparently read-onetenth of the actual volume passed: the index must make a complete circuit to
register the amount stamped on the dial. A small index an~ial are usually
present for testing the meter, and a tolerance of 2% is allowed'by law. This
testing is done by meter-provers, carefully calibrated gas holders kept at
constant temperature and the rate determined at different speeds. The meterprovers, are, in their turn, calibrated by a cubic foot, standardized at the Bureau
of Standards. It should be noted that all this calibration is corrected, not to
standard conditions (0 C. and 760 mm.), but to the cubic foot as fixed by, lawgas saturated with moisture at 60 F. and 30 inches. 8
The Connersville meter is the pump of that name, or Roots blower, in which
the flow of gas operates the two figure 8 vanes connected with an index.
The Pitot Tube.-Fig. 356. This consists of two glass tubes, D, of about 7i
to 1 inch internal diameter, inserted in the gas stream: one is bent at right
angles and is set so that it receives the
o
impact of the gas movement; the other
merely registers the pressure of the gas
in the pipe. The impact tube should
have a straight portion parallel to the
air flow, of a length about 10 times the
internal diameter of the tube. The
open end I'lhould be fire polished. 9 The'
point of insertion of these tubes in the
chimney or duct should be in a long
straight run of pipe, so as to be as free
from eddies as possible. The tube when
inserted should be preceded by 10 to 20
FIG. 356.
diameters of straight duef 'Or pipe and
followed by about 5 diameters.9 Davis 10 says that authorities- differ as to
whether the tubes should be a third or one-sixth of the diameter from the circumference to show the mean speed: he states further that each flue or chimney
is a separate problem and as a result of hundreds of measurements there seems
to be "no settled proportionate distance correspondihg to the mean velocity."
The glass tubes are connected by rubber tubing,' C, through a reverser, B,.
with a U-tube_ A, which is -either set vertically, inclined 30, or one in ten-this
82 F. makes a difference of about 0.5% in the volume.
9 Marks' Handbook.
I
10 "Handbook of Chemical Engineering," 1, 197; also fd~the tables for its use.
Also
Walker, Lewis and McAdams "Principles of Chem1cal Engineering" or Marks'
Handbook.

GAS ANALYSIS

2341

carries a vernier reading to hundredths of an inch so that readings to thou~

'sandths are feasible. The U~tube is filled with solutions of different specific
gravity, although ether of 0.74 sp.gr. is the one commonly employed. The
difference between the arms of the U-tube represents the difference between the
kinetic and static pressures of the gas in the flue or chimney.
The formula for calculating the velocity recomlflended by W. W. Scott is
1'=1290

th(1

+ .002176t)
BM

where h = vertical differential of gauge reading in terms of inches of water.

the reading. is taken as the actual reading is double that due to flow pressure
alone. B = barometric pressure of the gas in inches of mercury (29.92/1).
M = specific gravity compared with H = 1. Air = 14.39. t = temperature of
the gas. V = velocity in feet ,per second. The
formula V =42~h gives fairly accurate results.
Davis' formula is

Vo= I h459+tO X28.55.

'\j

519

Marks says of the Pitot tubes that they ~' are the
most satisfactory devices for the measurement of

gas in large volumes."

The Rotameter, Fig. 357.-This is a German instrument depending upon the height to which a float
-.is carried in a glass tube by the velocity of the stream
of gas. A modification of it was used by some of
the gas-lighting companies under the name of the
Thorp Gauge:
It consists of a graduated glass tube fixed upon a
tripod and provided with a plumb linG so that it can
be set vertically: gas passes in at the lower end, raises
the clay or talc float to a certain height and passes
out at the top. The height to which the float is
raised is noted on the graduations of the tube. The
formula for its use is
I

Vl=V~,
V = volume gas as shown by reading of instruFIG. 357.
ment, M its specific gravity = 1.0.
V = volume desired, Ml =sp.gr. of gas.
It is made in all capacities from 0.2 cu. ft., per hour up.
The Thomas Electric Gas Meter.-This depends upon the principle that if
the specific heat of the gas be ~nown, and an amount of energy be put into it,
by means of a coil, sufficient to keep a certain difference of temperature between
two thermometerf1, one before and one after the coil, this energy is a direct
measure of the VOlUme of gas flowing. Two electrical thermometers are placed
in the stream of gas with a heating coil between them: 2 difference of tempera-

2342

GAS ANALYSIS

ture is automatically maintained between. the thermometers and 'the energy to

preserve this difference of temperature (.0127 watt hour per standard unit bf
gas) is read off on the meter as cubic feet of gas. It is independent of temperature or pressure changes in the gas, and is used up ~9 gas pressures of 180 lbs.
per square inch. This is used in a Western gas works measuring 200,000 ft.
of gas per hour.
I
The Orifice Meter.ll-In this the same principle is used as in measuring
water, by determining the diminution in pressure as
registered on delicate gauges before and after the gas
5"
has passed through a standard orifice. It is largely
a
used for measuring natural gas, the entrained sand
in which enlarges the' orifice.
The Anemometer is used ordinarily for measuring
60S Flow
currents of air leaving or entering a room, analogous
meier
to its employment in meteorology.
The Gas Flow Meter. l 2_This consists of a tube a,
into .which capillaries b of different sizes can be inserted by a rubber stopper (one at a time). On either
side of the capillary is a T, the stems of which are
joined together, making aU-gauge c. This, when
filled with water shows the difference)ll pressure of
the /Sas before and after passiI).g thrpugh the capillaries. The apparatus is calibrated by a wet or dry
FIG. 358.
meter. It has a capacity of from 0.5 to 500 liters
(0.0176 to 17.6 cu. ft.) per hour. See Fig. 358.
The Venturi Meter.-=-Fig. 359 is an apparatus consisting of a steeply conical
tube connected to one less steep, producing a constriction, which is inserted in
the pipe line conveying the gas. From the pressures each s~of the constric-

FIG. 359.
I

tion and its size the quantity of gas passing can be determined:- 'For its use see
Marks' " Handbook" or Walker, Lewis and McAdams, " Principles of Chemical Engineering."
MEASUREMENT OF GAS IN SMALL QUANTITIES.

GAS BURETTES

Here may be mentioned the Hempel gas burette, made for accurate work
with a compensation tube; the bulbed Orsat or Bunte burette; the separatory
11 See in this connection, "Measurement of Gas and Liquids by Orifice Meter',"
Westcott and Diehl, Metric Metal Works, Erie, Pa. Publishers. ,Also Bureau of Standards, Circular No. 309 (1926), "Gas Measuring InstrUl;nents."
l2 Benton, Ind. Eng. Chern., 11, 623 (1919) . . .Alstrarticle by Dodge, "Measurement of Gases and Vapors," Ind. Eng. Chern., 24, 361 (1932).

GAS ANALYSIS

2343

funnel and graduate. Fuwa and Shattuck 13 measure quantities of 30-200 m!.
per minute by bubbling through water or dilute sulfuric acid.
The Hempel Gas Burette, Fig. 366, consists of a lOO-m!. burette graduated
in fifths of a cubic centimeter, provided with a short capillary at the top and
closed with a rubber connector and pinchcock, and a wider tube at the bottom,
ov,er which a 3/16-in. rubber tube is drawn, which connects it with the leveling
tube of similar size and length to the burette. It is manipulated by filling the
leveling tube completely with water, opening the pinchcock on the top of the
burette and filling it with water. The gas to be analyzed is sucked in and
measured as with the Qrsat apparatus, Fig. 364. Both burettes must be
calibrated as is customary in volumetric analysis, and the water saturated with
the gas to be analyzed. The water may be acidified and a few drops of an
indicator added to turn it pink; this shows if any alkali accidentally runs over
into the burette.
'
Separatory Funnel and Graduate, Fig. 372.-From the water which has
flowed out, the quantity of gas can be determined.
St~ndards of Measurement.-The unit volume of gas for scientific purposes
is measured dry at 0 C. and 760 mm. pressure: the legal or industrial cubic
foot in English speaking countries is measured saturated with water vapor, at
60 F. and 30 inches barometric pressure. The gas is therefore actually under
a pressure equal to the height of the barometer diminished by the tension of
aqueous vapor (in inches) at the temperature at which the gas is confined:
this is given in Table 2 of the Appendix.
In ordinary gas analysis, it is unusual tJ correct a gas volume, as barometric
changes during the analysis are negligible, and the laboratory temperature is
commonly kept very constant. In case corrections are necessary, we correct
first for pressure (Law 'of Boyle, V : V' : : P' : P): the height of the barometer is
read, its temperature taken, and its height in millimeters reduced to 0 C. = B.
The tension pf aqueous vapor at the temperature II;t which the volume of the
gas is read is determined from tables, Table 2,-Aq and is deducted from 'the
barometric height: in case the gas be under pressure M, as of a pipe line, this
is changed to millimeters of mercury, and added to these figures. The formula

. V0 =

IS

v:'(obs.) X B-Aq+M
760
'

Vb'
h bserve d
0 emg t h e correcte d an d V (ohs.) teo

volume.
Correcting next for temperature (Law of Charles V : V' : : T : T') the formula is Vo : V(oh ) : : To : T(ohs.), T being in absolute degrees
273
Vo= V(ohs.) X273+To C.
.

1
1+0.00367 TO C.'
1

then the formula becomes Vo= V(obs.) X 1+0.00367 TO C.

Tables giving factors for 1 +0.00367 TO C. for a wide range of temperature
are to be found in the ordinary" Handbooks of Chemistry and Physics."
13 Ind. Eng. Chern., 15,230 (1923).
See also Page, "Constant Flow Orifice Meters,"
Ind. Eng. Chern., Anal. Ed. 7, 355 (1935).

2344

GAS ANALYSIS

For reduction to conditions prevailing in the gas industry, "Industrial

Standard Conditions," i.e., 30 inches, saturate,d wi* water, at 60 0 F., the
formula is
B-Aq+M
520 0 Ji'.
Vo= V(ob . ) X 30" -0.51" X'4bO+T? F.'

B, Aq and M are in inches or" mercury: 0.51 inch is the t1ension of aqueous vapor
at 60 0 F. In practice this reduction is obtained from Table 9 (Appendix).
Specific Gravity.-The specific gravity of a gas is usually referred to air as
unity: it can be readily calculated by dividing the molecular weight of the gas
by 28.95, the weight in grams of 22.4 liters of air (22.4 liters will be remembered
as the gram-molecular volume).
Specific gravity can be directly determined by Jenkins' apparatus, p. ~7, or
the Edwards gas balance. H , 15
(
This depends upon an indirect weighing of the gases. The apparatus
consists of a closed vessel counterpoised in a water-jacketed cylinder of the gas;
this is connected with a mercury gauge. Dry air is passed into the cylinder
and pressure applied to it until the vessel just balances at a predetermined
point and the gauge reading noted. TIle operation is repeated with the gas
to be tested. The densities of the gases must be the same at these pressures;
or the densities of the two gases at normal presflure are in the inverse ratio of
the pressures, when of equal buoyant foree. Results accurate to 0.2% are
obtained and no calibration is required. / It has given trouble- at times owing
apparently to leaks in the" closed vessel" mentioned.
With the effusion method results are dependent upon the orifice used, the
pressure and the confining medium. They may vary as much as 10% with
different observers. The orifice should be 0.18-0.3 mm. in diameter in a plate
0.04 mm. or less thick: the pressure should not be allowed to fall below 2 inches
of water: results to an accuracy of 2% may be expected.
ABSORPTION ApPARATUS, TUBES AND ?IPETTES

These are quite varied according to the purpose fof which they are intended.
A very efficient form is the Friedrichs Spiral Gas Washing bottle, Fig. 360;, here
the gas has to pass through a long spiral path. Dennis recommends this for
the absorption of sulfur dioxide. The gas is run through a solution until a
color change takes place. Were the reagent t~ be washed out and titrated it
would not seem so well 'adapted on account of the difllculty _of thoroughly
washing it.
The Varrentrapp and Will bulbs, Fig. 361, the use of which is evident from
the figure, are used for the absorption of ammonia in illuminating gas. The
Wolff absorption tube, Fig. 362, the empty arm 'of which is filled with beads or
broken glass, is used for the absorption of carbon monoxide by bloo.d. The
bulbed tube, provided with a small jet, is generally u'sed for containing standard
solutions of acid, alkali, or potassium permanganate or, in general, a solution
which is to be titrated after absorption. Where the presence of a gas as, for
example, water vapor is to be determined by the increase of weight of the
reagent used, Winkler's spiral, Fig. 363,16 may be employed/
14 Ind. Eng. Chern., 9, 790 (1917).
Bur. Stds. Tecl(n:Bull. No. 94.
16 Improvements by Warren, C. A., 30, 2425.

GAS ANALYSIS

2345

FIG. 361.

FIG. 360.

FIG. 363.

FIG. 362.

Pipettes.-Particularly for the Orsat Apparatus, Fig. 364. These are

ordinarily filled with glass tubes, but various modifications have been proposed:
these are the bubbling type of Hankus, the spiral bubbling variety of NowickiHeinz and a combination of the Friedrichs wash bottle proposed by Dennis.
All these suffer from the very serious disadvantage of a glass three-way stopcock at the top, which it is practically impossible to prevent from sticking,
unless the apparatus be used by one person and that one exceptionally careful
and painstaking. Dennis has shown that one variety is no more rapid than
the original and some of the others but little more so. When it is further
considered that they cost four times as much, their use would seem of doubtful
expediency.
It is interesting to note further that Anderson 17 has shown that with the
modified potassium pyrogallate which he uses, the original Orsat slightly
modified, to allow the precipitate to settle, is the best. Francis of the U. S.
Steel Corporation laboratory.has devised a bubbling pipette with automatic
valve, which is very effective: Shepherd of the Bureau of Standards has also
devised a bubbling pipette in which the gas is delivered into. the reagent through
a fine platinum sieve making a quantity of fine bubbles. It would seem likely
that the holes in the platinum might become clogged by the standing of the
reagent. Both of these are very expensive.
EXAMINATION OF THE GASES

The qualitative examination of a gaseous mixture is rarely resorted to in

technical work: a sufficiently close idea of the gases present can be obta'ined by
a consideration of the reactions involved in the various operations. It is,
however, not safe to rely upon this in matters of importance, as conditions may
change: for example if the gases be dry or dilute, hydrogen sulfide and sulfurous
anhydride can exist together. Similarly in sewage gases, the gas absorbed by
cuprous chloride IS probably not wholly carbon monoxide but higher paraffin
hydrocarbons~ The means of identifying the different gases will be found under
each gas.
16

17

Called "Winkler's Bulbs" in the apparatus catalogs.

Ind. Eng. Chern., 8, 131-3 (1916).

2346

GAS ANALYSIS.
I

Gases are sometimes tested by burning them arid investigating the products
of combustion. Often the gas refuses to burn, as in the case of wood gas, as it
contains so much incombustible gas (30% CO 2): ~his is removed with caustic
alkali and the burning test repeated. In performi~g this test, partic~with
electrolytic gases, it must be remembered that they\may contain enough oxygen
or chlorine to support their own combustion like ',' detonating gas." A glass
tube 18 inches long well packed with glass beads, broken glass or porcelain must
always be inserted between the gas holder and the jet: this will extinguish the
flame of the gas wave, should it start, after the manner of the gauze in the Davy
safety lamp, and prevent flame from reaching the explosive mixture in the gas
holder with serious results. Sufficient pressure should of course be maintained
on the holder itself.
DETECTION AND DETERMINATION OF THE VARIOUS GASES

Clemens Winkler 18 divided the gases into seven groups according to their
behavior with various solvents. These were contained in suitable absorption
tubes or vessels and the gases passed through them. His scheme was as
follows:
Gases absorbed by
1. H 2S0 4 1.7 sp.gr.; NHa, (N 20a), N 20 4 .1 9
II. KOH 1.3 sp.gr.; Ch, HCI, (CN)2, HCN, S02, H 2S, SiF 4, CO 2.
III. AgNOa; PHa, AsH a, SbHa. /
IV. Pyrogallol; 02, (Oa).
V. CuCI; CO.
VI. Acid FeS04 1 : 2; NO.
VII. Unabsorbed; H 2, CH 4 , C 2H 2, C 2H 4, N 20, N 2, CbS, and the noble gases.
The following tables give the specific gravity referred to air, the solubility in
water at 20, the qualitative tests and quantitative methods of determination
of these gases: additional means will be found mentioned under the several
gases themselves.
:
Group r
.
H 2S0 4 1.7

GASES ABSORBEP BY

SP.GR.

Nitrous anhydride
Name ............ . Ammonia
Gravity, air=l .... . 0.5971
Qualitative tests ... . Fumes w. HCI on a Acts like a mixture
rod. Nessler's reaof NO and NO"
gent
Quantitative det'n ... Absorption in standardacid.Or3NaBrO
+2NH s =N 2
+3NaBr+3H 20.
Or absorption in water and N esslerization
Solubility 20 C., 1
ml. water absorbs
ml.. ... .......... 702
I

Nitrogen tetroxide
1.5906
Absorb in KOH, test
-..._for nitrites
Absorption in N ItO
KMn04' 2KMn04
+ 10N0 2 +3H 2S04
+2H 20 "" 10HNOs
+K 2S04+2MnS04. Or in standard alkali
Forms HNO s, HN0 2

"Die Industrie Gase," 1876-1879.

.JSulfuric acid of this strength dissolves appreciable quantities of propane, propene,
butane and butene.
18

19

GAS ANALYSIS

2347

GroupU
GASES ABSORBED BY
J

KOH 1.3

SP.OR.

Hyd~ochloril: acid
1.268
Fumesw. NHa

N arne. . . . . . . . . . . .. Chlorine
Gravity, air=l ..... 2.4494
Qualitative test. . . .. KI starch paper

Cyanogen
1.7968
Pass through mixtureofFeSOd :10
+KOH 1 : 3, and
ppt. as Prussian
blue w. ferric alum
and H 2S0 4 20

Quantitative det'n ... 2KI+Cb=2KCI+I 2 Absorption in

Or absorption w.
standard alkali or
KOH
silver nitrate
Solubility 20 C., 1
rnl. water absorbs
444
mI ............... 2.29
4.5
Hydrocyanic acid

Sulfur dioxide

2.224 ..
0.9359
Absorption in KOR Fuchsine paper
bleached or
and FeS04 and
KIO, starch
FeCl 2 =Prussian
paper. SO.
blue. Absorption w. KOH or
+2I+2H 20
acid AgNO s 21
=H 2S0 4 +2HI
39.3
Very sol.

Hydrogen sul- Silicon fluoride

fide
1.1769
3.60
PbAc2 paper,
None.
absorption by 3SiFoI +4H.0
I solution.
=Si0 4H 4
H.S+21
+2H 2SiF s
=2HI+S
2.90

Decomposed

Carbon dioxide
1.5290
Ba02H2 on
black rod.
Absorption
w. KOH or
BaOzH 2 and
titration
0.874

Groupm
GASES ABSORBED BY AgNO a
Name ............ . Hydrogen phosphide,
phosphine
Gravity, air=l. ... . 1.1829
Qualitative test .... . Neutral HzO solution
KI + HgI 2 = cryst.
orange yei. ppt.
PHgsla 22
Quantitative det'n .. Pass through Br water and ppt. H 3P0 4
as usual,
Solubility 20 C., 1
mi. water absorbs
ml. ... , ......... 0.02

Hydrogen arsenide,
arsine
2.695
Blk. ppt. of AsAg s
w. AgNO,

Hydrogen antimonide, stibine

4.360
Blk. ppt. of SbAg s
w. AgNOa

Absorb w. NaCIO
cont. 3% C1 23

Decompose w. tartaric acid and det.

Sb

About 5

Slight

Nauss, J. Gasbeleuchtung, 43, 969 (1900).

Rhodes, J. Ind. Eng. Ch~m., 4, 652 (1912).
22 Lemoult, Compt. rend., 139, 478 (1904).
23 Reckleben, Z. angew. Chern., 19, 275 (1906).
20

21

2348

GAS ANALYSIS
Group IV

Group V

Group VI

Absorbed by

Absorbed by potassium pyrogallate

Name ..... ; .... Oxygen

~hr~rid:

Absorbed by FeSO.
1 : 2 acidulated w.
H2S0.

Ozone

Carbon :rilonNitric oxide

oxide I
1.624
Gravity, air=l.. 1.10.'52
0.9671
1.0367
Qualitative test. Darkening of MnCl. paper;
Absorb in blood, Oxidize, absorb in
and examine
light brown KI starch paKOH and test
per, N 20, and
"pyro"
w. spectrofor nitrites
H.O. being rescope
moved by
KMnO,
Quantitative
Absorption w.
det'n ......... By"Pyr!)."
Absorb in' FeSO,
CuCI in abCuCI
1 : 2 acidulated
sence of CO
w. H.SO, or with
KMnO, as N .0,
Solubility 20 0 C.,
1 ml. water ab0.6 at 0 0
0.470
sorbs m!. ..... 0.0317
0.0231

Group VII'
UNABSOH,BE.D

Name .............. Hydrogen

Methane

Ethylene
(ethene)
0.9684

Acetylene
(ethine)
0.5545
0.9056
Gravity, air = 1. ..... 0.0694
Red ppt. w.
Qualitative test ...... Dry PdCb gives None
am. CuCl
HCI
(explosive !)
Quantitative det'n ... By combustion By combustion Absorption w. As C.H" which
or explosion
Br water or
see
or explosion
H 2S 20 1
w. O.
w. O.
'"
0
Solubility 20 C., 1 ml.
0)'22
water absorbs ml. .. 0.0183
0.0330
1.03
Nitrous oxide
1.5301

Carbon oxysulfide
2.075

Nitroge~
0.9673

The noble gases

Helium, Neon,
Argon, Krypton,
Xenon

None
None
None
By absorption or com~
By explosion with H. Alcoholic KOH
bustion of all other
or combustion w.
1 : 3 in 66% algases and measuring
chol by weight
CuO
the residue which
also contains the noQle gases
.0159
0.670
0.568
"

.._

("

NOTEs.-Group I: NaBrO is made by saturating a 10% solution of .caustic soda with

bromine.
Group II: Chlorine can be removed from hydrochloric acid by passing the gases over
finely powdered antimony: hydrochloric acid can be removed from. chlorine by means of
manganese dioxide or zinc oxide.
Cyanogen is not absorbed by acid silver nitrate solUlJon, from which it can be separated by drawing air through it: hydrocyanic acid is precipitated under these conditions.

GAS ANALYSIS

2349

ANALYSIS OF GASES
The analysis of a gaseous mixture is effected by absorbing the various
constituents and observing the diminution in volume: in case the gas be unabsorbable, as for example methane (CH 4), it is burned and the carbon dioxide
and water determined.
t
(a) Analysis of mixtures for carbon dioxide, oxygen and carbon monoxide
(e.g., chimney gases, producer and blast furnace gas) can be done with any of
the apparatus to be described. The Orsat, or Elliott are the forms usually
employed.
(b) Analysis of mixtures as in (a) and also containing combustible gases as
hydrogen and methane, e.g., city gas.
The Orsat Apparatus. Description.-The apparatus,24 Fig. 364, is enclosed
in a case to permit of transportation from place to place; furthermore, the
measuring tube is jacketed with water
to prevent changes of temperature
affecting the gas volume. The appaTa tu's consists eS'sentially of the leveling bottle A, the burette B, the pipettes
pI, pll, pilI, and the connecting tube
T. Pipette P' is filled with potassium
(or sodium) hydroxide solution (see
Reagents) so that when it is <Jrawn
up into the front arm about half an
inch in depth is left in the rear arm.
Pipettes p" and pilI are similarly filled
with potassium (or sodium) pyrogallate and cuprous chloride solutions
respectively. These reagents require
\ to be protected from the oxygen of the
air by collapsible rubber bags of good
quality. As the oxygen in the air over
the reagent is absorbed, a diminution
in pressure takes place rendering it
difficult to bring the reagent to the
point on the stem: the obvious remedy
FIG. 364.
is to remove the bag temporarily and
adjust the reagent. When the apparatus is first set up, one or two blank
analyses should be madc) to saturate the water aqd reagents with the gases.
For example the potassIum hydroxide absorbs carbon dioxide, it also absorbs
about 3 m!. of oxygen, 2 Ill!. of carbon monoxide and 1.5 m!. of nitrogen, by
virtue of the 100 m!. of water which it contains. A change of temperature of
10 makes a change of 0.36 of the volume of the gas: a change of pressure of 1 mm.
produces 0.13 change in the volume.
Manipulation.-The reagents in the pipettes should be adjusted in the capillary tubes to a point on the stem about midway between the top of the pipette
24 The apparatus shown is one of the simplest and least expensive forms: it has proved
itself well adapted for occasional use. Hayes has constructed it in a very attractive and
compact form in metal.

GAS ANALYSIS
and the rubber connector. This is effected by opening wide the pinchcock
upon the connector, the bottle being on the table, and '[ery gradually lowering
the bottle until the reagent is brought to the point above indicated. Six
inches of the tubing used correspond to but 0.1 mI., so ~hat an error of half an
inch in adjustment of the reagent is without influence I upon the accuracy of
the result. The reagents having been thus adjusted, th!'l burette and connecting tube are completely filled with water by opening d and raising the leveling
bottle. The apparatus is now ready to receive a sample of gas (or air for
practice). In case a flue-gas is to be analyzed d is connected with i, Fig. 354,
A lowered and about 102 mI. of the gas forced over by opening h; or d may be
connected with a T-joint in the gas stream; the burette after filling is allowed
to drain one minute by the sand glass, c snapped upon its rubber tube, and the
bottle A raised to the top of the apparatus. By gradually opening c the water
is allowed to run into the burette until the lower meniscus stands upon the
100 or 0 mark (according to the graduation of the apparatus)'. The gas taken
is thus compressed into the space occupied by 100 mI. and by opening d the
excess escapes. Open c and bring the level of the water in the bottle to the same
level as the water in the burette and take the reading, which should be 100 ml.
Special attention is called to this method of reading: if the bottle be raised, the
gas is compressed; if lowered, it is expanded. ,
Determination of Carbon Dioxide.-The gas to be analyzed is invariably
passed first into pipette pI, containing potassium hydroxide for the absorption
of carbon dioxide, by opening e and raising~. The gas displaces the reagent
in the front part of the pipette, laying bare the tubes contained in it, which
being covered with the reagent present a large absorptive surface to the gas;
the reagent moves into the rear arm of the pipette, displacing the air over it into
the flexible rubber bag "\vhich prevents its diffusion into the air. The gas is
forced in and out of the pipette by raising and lowering A, the reagent finally
brought approximately to its initial point on the stem of the pipette, the burette
allowed to drain one minute and the reading taken. The.difference between
this and the initial reading represents the milliliters of. 'carbon dioxide preslmt
in the gas. To be certain that all the carbon dioxide is 'removed, the gas should
be passed a second time into pI and the reading t~ken as before; these readings
shouldtagree within 0.1 %, and are called" check readings."
Determination of Oxygen.-The residue frcim the absorption of carbon
dioxide is passed into the second pipette, pll, containing an,~lkaline solution of
potassium pyrogallate, until no further absorption will take'place.. The difference between the reading obtained and that after the absorption of carbon
dioxide, represents the number of milliliters of oxygen present. Determination of Carbon Monoxide.-The residue from the absorption of
oxygen is passed into the third pipette, pili, containing cuprous chloride, until
no further absorption takes place; that is, in this cafje until readings agreeing
exactly (not merely to 0.1) are obtained. The difference between the reading
thus obtained and that after the absorption of oxygen, represents the number
of milliliters of carbon monoxide present.
Determination of Hydrocarbons.-The residue left after all absorptions have
been made may consist, in additon to nitrogen, the:~principal constituent, of
. -' hydrocarbons and hydrogen: see p. 2356.

GAS ANALYSIS

2351

Accuracy.-The apparatus gives resuits accurate to 0.2 of 1 %, hence figures

obtained by division to 0.01 should not be reported.
Time Required.-About twenty minutes are required for an analysis; two
may be made in twenty-five minutes, using two apparatus.
NOTEs.-The method of adjusting the reagents is the only one which has been found
satisfactory: if the bottle be placed at a lower level. and an attempt made to shut the
pinchcock c upon the connector at thc proper time, it will almost invariably result in
failure.
The process of obtaining 100 m!. of gas is exactly analogous to filling a measure
heaping full of grain and striking off the excess with a straight edge; it saves arithmetical
work, as milliliters read off represent per cent directly.
It often happens when e is opened, c being closed, that the
reagent pi drops, due not to a leak, as is usually supposed, but to
the weight of the column of the reagent expanding the gas.
The object of the rubber bags is to prevent the access of air to
the reagents, those in p" and pili absorbing oxygen with great
avidity, and hence if. freely exposed to the air would soon become useless. These bags if well made will last two years and
am'ply protect the solution: or a "sealing pipette" can be used
for the pyro.
Carbon dioxide is always the first gas to be removed from a
gaseous mixture. In the case of air the percentage present is so
small, 0.08 to 0.1, as scarcely to be seen with this apparatus. It
is important to use the reagents in the order given; if by mistake
the g:;ts be passed into the second pipette, it will a-bsorb not only
oxygen, for which it is intended, but also carbon dioxide; similarly
if the gas be passed into the third pipette, it will absorb not only
carbonic oxide, but also oxygen afl. well.
'The use of pinchcocks and rubber tubes, original with the
author, although recommended by Naef, is considered by Fischer to be inaccurate. The experience of the author, however,
does not support this assertion, as they have been found to be fully
as accurate as glass stopcocks, and very much less troublesome
and expensive.
In case any potassium hydroxide or pyrogallate be sucked
over into the tube T or water in oA, the analysis is not spoiled,
but may be proceeded with by connecting on water at d, opening
this cock, and allowing the water to wash the tubes out thoroughly.
The addition of a little hydrochloric acid to the water in the bottle
A will. neutralize the hydroxide or pyrogallate, and the washing
,0
may be postponed until convenient: Methyl orange indicator
may be added to be sure the water is acid.
After each analysis the number of milliliters of oxygen and
carbonic oxide should be set down upon the ground-glass slip provided for the purpose. By adding these numbers and subtracting
their sum from the absorption capacity (see Reagents) of each
reagent, the condition of the apparatus is known at any time,
and the reagent can be renewed in season to prevent incorrect
analyses.
Elliott Apparatus. Description.---;The apparatus, 25 Fig.
365, consists of a burette holding 100 ml. graduated in
FIG. 365.
tenths of a milliliter and bulbed_ like the Bunte apparatus-the bulb holding about.30 ml.; it is connected with a leveling bottle similar
to the Orsat apparatus. The top of the burette ends in a capillary stopcock, the
stem of which is ground square to admit of close connection with the" labora20

The simplest form is here shown as being best

adapt~d

for teaching purposes.

2352

GAS ANALYSIS

tory vessel," an ungraduated tube similar to the burette, except of 125 m!.
capacity. The top of this" vessel" is also closed1with a capillary stopcock,
carrying by a ground-glass joint, or better a rubber stopper, a thistle-tube F,
for the introduction of the reagents. The lower end of this" vessel" is closed
by a rubber stopper carrying a three-way cock 0, and connected with a leveling
bottle D. The burette and vessel are held upon a block of wood-supporter!
by a ring stand-by fine copper wire tightened by violin keys.
Manipulation.-The ground-glass joints are lubricated with stopcock grease.
The leveling bottles are filled with water, the stopcocks opened, and the bottles
raised until the water flows through the stopcocks m and n. m is connected
with the source whence the gas to be analyzed is to be taken, n is closed, D
lowered and rather more than 100 ml. drawn in, and m closed. n is opened,
D raised and E lowered, nearly 100 ml. of gas introduced, and n closed; by
opening m and raising D the remainder of the gas is allowed to escape, the
tubes being filled with water and m closed. n is opened and the water brought
to the reference mark; the burette is allowed to drain one minute, the level of
the water in E is brought to the same level as in the burette, and the reading
taken.
Determination of Carbon Dioxide.-By raising E, opening n, and lowering
D, the gas is passed over into the laboratory vessel; F is filled within half an
inch of the top with potassium hydroxide, closed, m opened, and the reagent
allowed slowly to trickle in. A No. 3/evaporating dish is placed under 0, and
this turned to allow the liquid in the laboratory vessel to run into the dish.
At first this is mainly water, and may be thrown away; later it becomes diluted
reagent and may be returned to the thistle-tube. When the depth of the
reagent in the thistle-tube has lowered to half an inch, it should be refilled either
with fresh or the diluted reagent and allowed to run in until the absorption is
judged to be complete, and the gas passed back into the burette for measurement. To this end close and then m, raise E, open n, and force some pure
water into the laboratory vessel, thus rinsing out the capillary tube. Now raise
D and lower E, shutting n when the liquid has arrived at the reference mark.
The burette is allowed to drain a minute, the: level' of the water in the bottle E
brought to the same level as the water in the burette, and the reading taken.
Determination of Oxygen.-The manipulation is the same as in the preceding determination, potassium pyrogallate being substituted for potassium
hydroxide; the apparatus requiring no washing out. __
Determination of Carbon Monoxide.-Thc.laboratorY vess~l, thistle-tube,
and bottle if necessary, are washed free from potassium pyrogallate and the absorption made with acid cupr:ous chloride similarly to the 'determination of
carbon dioxide. The white precipitate of cuprous chloride may be dissolved by
hydrochloric acid.
Accuracy and Time Required.-The apparatus is as accurate for absorptions
as that of Orsat; it is stated to be much more rapid-a claim which the writer
cannot substantiate. It is not as portable, is more fragile, and more troublesome to manipulate, and as the burette is not jacketed, it is liable to be affected
by changes of temperature.
,
NOTEs.-In case at any time it is desired to stlp.lhe influx of reagent, 0 should be
closed first and then m; the reason being that the !tosorption may be so rapid as to suck
air in through 0, m being closed.

GAS ANALYSIS

2353

The stopcock should be so adjusted as to cause the reagent to spread itself as completely as possible over the sides of the burette.
By the addition of an explosion tube it is used for the analysis of city gas,' bromine
being used to absorb the "illuminants," Winkler '7. states that this absorption is incomplete; later work by Treadwell and Stokes, and also Korbuly,'s has shown that bromine
water, by a purely physical solution, does absorb the" illuminants" completely; Hempel 2g
states that explosions of hydrocarbons made over W!lter are inaccurate, so that the
apparatus can' be depended upon to give results upon methane and hydrogen only within
about 2%. It is, however, very rapid, a complete analysis of city gas can be made with
it in fifty-five minutes.

Hempel's Apparatus. Description.-The apparatus/a Figs. 366 and 367,

is very similar in principle to that of Orsat; the burette is longer, admitting of

FIG. 366.

the reading of small quantities of gas and the pipettes are separate and mounted
in brass clamps on iron stands. P shows a " simple" pipette 31 provided1with
a rubber bag; this form, after forty years of use, can be said satisfactorily to
take the place of the cumbersome" compound" pipette.
Mackintosh, Am. Chern. Jour., 9, 294 (1887).
Zeit. Anal. Chern., 28, 286 (1889).
28 Treadwell-Hall's "Quantita~ive Analysis." p. 569.
29 "Gasanalytische Methoden" p. 1O~
30 The simplest forms are here shown.
31 Gill. Am. Chern. J., 14, 231 (1892).
2.

27

2354

GAS ANALYSIS

The pipette for fuming sulfuric acid 32 _is shown at F, and differs from the
ordinary in that vertical tubes after the manner of tqose in the Orsat pipettes
replace the usual glass beads. This prevents the trapping of any gas by the
filling, which was so common with the beads and glas~ wool. E represents the
large explosion pipette,33 of about 250-ml. capacitY,lwith walls half an inch
thick; the explosion wires enter at the top and bottoml to prevent short-circuiting; mercury is the confining liquid. The small explosion pipette holds about
110 m!. and is of glass, the same thickness as the simple pipettes. Water is
here used as the confining liquid, and acidulated with Hel usually in the burette.

FIG. 367.

A " Ford" induction coil capable of giving a fourth-inch spark, with three
dry cells, four " simple" pipettes and a mercury filled burette, complete the
outfit.
The burette should be carefully calibrated and the corrections may very
well be etched upon it opposite the 10-m!. divisions.
..... .
In working with the apparatus the pipettes are placed up;Uthe adjustable
stand S and connection made with the doubly bent capillary tube.
Manipulation.-To acquire facility with the use of the apparatus before
proceeding to the analysis of city gas, it is well to make the following determinations, obtaining" check-readings" in every cas~: 1. Oxygen in air, by (1)
absorption with phosphorus; (2) absorption w;ith potassium (or sodium)
pyrogallate; 34 (3) by explosion with hydrogen.
32 Ibid.,_J. Am. Chem. Soc., 18,67 (1896).
33 Gill, J. Am. Chern. Soc., 17,771 (1895).
34 The writer finds after an experience of more than thirty-five years in the laboratory
with hundreds of students, that sodium pyrogallate can .be~used with practically the sameresults as the potassium compound. The absorption' is complete, as shown by subsequent treatment with cuprous chloride.

GAS
ANALYSIS.
.
1.

2355

DETERMINATION OF OXYGEN IN AIR

(1) By Phosphorus.-lOO m!. of air are measured out as with the Orsat
apparatus, the burette being allowed to drain two minutes. The rubber connectors upon the burette and pipette are filled with water, the capillary tube
inserted, as far as it will go, by a twisting motion, into the connector upon the
,burette, thus filling the capillary with water; the free end of the capillary is
inserted into the pipette connector, the latter pinched so as to form a channel
for the water contained in it to escape, and the capillary twisted and forced
down to the pinchcock. There should be as little free space as possible between the capillaries and the pinchcock. Before using a pipette, its connector
(and rubber bag) should be carefully examined for leaks, especially in the
former, and if any found the faulty piece replaced.
The pinchcocks on the burette and pipette are now opened, the air forced
over into the phosphorus, and the pinchcock on the pipette closed; action immediately ensues, shown by the white fumes; after allowing it to stand fifteen
minutes the residue is drawn back into the burette, the latter allowed to drain
and the reading taken. The absorption goes on best at 20 C., not at all below
15" C.; it is very much retarded by small amounts of ethane and ammonia.
It cannot be used to absorb commercial oxygen. No cognizance need be taken
of the fog of oxides of phosphorus.
(2) By Potassium PyrogaUate.-lOO m!. of air 35 are measured out as before,
the carbon dioxide absorbed with potassium hydroxide and the oxygen with
potassium pyrogallate, as with the Orsat apparatus; before setting aside the
pyrogallate pipette, the number of milliliters of oxygen absorbed should be
noted upon the slate s on the'stand. This must never be omitted with any
pipette save possibly that for potassium hydroxide, as failure to do this may
result in the ruin of an important analysis. The reason for the omission in this
case is found in the large absorption capacity-four to five liters of carbon
dioxide-of the reagent.
(3) By Explosion with Hydrogen.-Forty-three m!. of air and 57 m!. of
hydrogen are measured out, passed into the small explosion pipette, the capillary of the pipette filled with water, the pinchcocks and glass stopcock all
closed, a heavy glass or fine wire gauze screen placed between the pipette and
the operator, the spark passed between the spark wires, and the contraction in
volume noted. The screen should never be omitted, as serious accidents may occur
thereby. The oxygen is represented by one-third of the contraction. For
very accurate work the sum of the combustible gases should be but one-sixth
that of the non-combustible gases; otherwise some nitrogen will burn and high
results will be obtained; 36 that is, (H+O) : (N +H) : : 1 : 6.

II.

ANALYSIS OF MANUFACTURED OR CITY GAS

One hundred m!. of gas are measured from the bottle containing the sample
into the burette.
Determination of Carbon Dioxide.-The burette is connected with the
pipette containing potassium hydroxide and the gas passed into it with shaking
until no further diminution in volume takes place.
36
36

See Anderson's work, J. Ind. Chern., 7, 587 (1915).

This is shown in the work of Gill and Hunt, J. Am. Chern. Soc., 17,987 (1895).

GAS ANALYSIS

2356

Illuminants, Cn H 2n , Cn H 2n _ 6 Series.-The gas is passed into saturated

bromine water and thoroughly shaken, the bromine fumes removed, with the
potassium (or sodium) hydroxide, like the CO 2 , and the treatment with bromine
and hydroxide repeated uhtil check readings are Iobtained. The difference
betwe()n this last reading and that after the absoJ:1'ltion of the carbon dioxide
represents the volume of " illuminants " or " heavy) hydrocarbons" present.
Or fuming sulfuric acid can be employed instead 6f bromine water, removing
the sulfuric and sulfurous anhydrides by potassium hydroxide as in the case of
the bromine fumes. Fuming nitric acid is not recommended, as it is liable to
oxidize carbon monoxide.
Oxygen.-This is absorbed, as in the analysis of air, by potassium or sodium
pyrogallate.
Carbon Monoxide.-The gas is now passed into ammoniacal cuprous chloride, until the reading is constant to 0.2 ml.; it is then passed into a second
pipette, which is fresh, and absorption continued until constant readings are
obtained. The second pipette should not have absorbed more than 10 ml. of
CO.37
Gautier and Clausmann 38 have shown that some carbon monoxide escapes
solution in cuprous chloride, so that for very accurate work it may be necessary
to pass the gas through a U-tube containing iodic anhydrid~ heated to 70 C.
This is done by interposing this tube between the burette and a simple
pipette filled with potassium hydrate. / The reaction is 5CO + 12 0 5 = 5C0 2 +21.
The diminution in volume represents directly the volume of carbonic oxide
present.
The volume of air contained in the tube should be corrected for as follows:
One en.d of the tube is plugged tightly and the other end connected with the gas
burette partly filled with air. A bath of water at go C. is placed around the
U-tube and the reading of the air in the gas burette recorded when constant;
the bath is now heated to 100 and the burette reading again recorded,when
constant. The increase in reading represents one-third the volume of the
U-tube, 273 : 273 + (100 -9) : : 3 : 4.
...'
.
Methane and Hydrogen. (a) Hinman's Method. 39-The gas left trom the
absorption of carbon monoxide is passed into the large explosion pipette. About
half the requisite quantity of oxygen (40 mIl.) necessary to burn tile gas is now
added, mercury introduced through the T in the connector sufficient to seal
the capillary of the explosion pipette, all rubber connectors carefully wired,
the pinchcocks closed; and the pipette cautiously sh9:k:en.-... A screen of heavy
glass or fine wire gi'h~ze is interposed between the operator and the apparatus,
the explosion wires are connected with the induction coil, a spark passed between them and the pinchcocks opened, sucking in the remainder of the oxygen.
The capillary is again sealed with mercury, the stopcock opened and closed, to
bring the tontents of the pipette to atmospheric pressure, and the explosion
repeated as before, and the stopcock opened. .
It may be found expedient to increase the inflammability of the mixture, to
introduce 5 ml. of " detonating gas," the electrolytic mixture of hydrogen and
oxygen. The gas in the pipette containing carbon dioxide, oxygen, and
,

OIl cuprous sulfate (Reagents) may be used.

I '

._
38 Bull. Soc. Chim., 35, 513; Abstr. Analyst, 31';'349 (1906).
39 Gill and Hunt, J. Am. Chern. Soc., 17, 987 (1895).
37

GAS ANALYSIS

2357

nitrogen is transferred to the mercury burette and accurately measured. The

carbon dioxide resulting from the combustion of the marsh gas is determined by
absorption in potassium hydroxide; to show the presence of an excess of oxygen,
the amount remaining is determined by absorption with potassium pyrogallate.
The calculation is given below .. For very accurate work a second analysis
should be made, making successive explosions, using the percentages of methane
and hydrogen just found as a basis upon which to calculate the quantity of
oxygen to be added each time. The explosive mixture should be so proportioned that the ratio of combustible gas (i.e., CR., Rand 0) is to the gases
which do not burn (Le., N and the excess of CH 4 and H) as 100 is to about 50
(from 26 to 64) ; 40 otherwise the heat developed is so great as to produce oxides'
of nitrogen, which, being absorbed in the potassium hydroxide, would affect the
deterl1lination of both the methane and the hydrogen. The oxygen should
preferably be pure, although commercial oxygen, the purity of which is known,
can be used; the oxygen content of the latter should be tested from time to
time, especially with different samples.
(0) Hempel's Method. 41_From 12 to 15 m!. of the gas are measured off into
the burette (e.g., 13.2 ml.) and the residue is passed into the cuprous chloride
pipette for safe keeping. That in the burette is now passed into the small
explosion pipette; a volume of air more than sufficient to burn the gas, usually
about 85 mI., is accurately measured and also passed into the explosion pipette,
and in so doing water from the burette is allowed partially to fill the capillary
of the pipette and act as a seal. The rubber connectors upon the capillaries of
the ,burette and pipette are carefully wired on, both pinchcocks shut, and the
stopcock closed. The pipette is c,autiously shaken, the screen interposed, the
eJ:::plosion wires connecte.d with the induction coil, a spark passed between them,
and the stopcock immediately opened. The gas in the pipette, containing
carbon dioxide, oxygen, and nitrogen, is transferred to the burette, accurately
measured, by reading immediately, to prevent the absorption of carbon dioxide,
and carbon dioxide and oxygen determined in the usual way.
Calculation. Ca) Hinman's Method.-56.2 m!. of gas remained after the
absorption; 77.4 m!. of oxygen were introduced, giving a total volume of 133.6
ml.
Residue after explosion ...................... , ... , . , ... 46,9
Residue after CO 2 absorption, ......... ,. , .... , ........ , 28.2
Carbon dioxide formed ............ , ................... 18.7
Contraction ............. , ................. 133.6-46.9=86.7
Residue after 0 absorption ........................ , ., .. 25.6
Oxygen in excess ......... " " " " " " ....... 28.2-25.6= 2.6

m!.
"
"
"

"

The explosion of marsh gas or methane is represented by the equation

From this it is evident that the volume of carbon dioxide is equal to the volume
of methane present; therefore in the above example, in the 56.2 m!. of gas
burned, there were 18.7 m!. methane.
40

41

Bunsen, "Gasometrische Methoden," 2d ed., 73 (1877),

Hempel. "Gas Analytische Methoden," 3d ed., 245 (1901).

2358

GAS ANALYSIS

The total contraction is due (1) to the disapp~arance of oxygen in combining with the hydrogen of the methane, and (2) ;to the union of the free
hydrogen with oxygen. The volume of the methane having been found, contraction (1) as can be ascertained from the equation\ above, equals twice the
volume of the methane; hence .
I
86.7- (2 X 18.7) = 49.3 mI.,

contraction which is due to the combustion of hydrogen.

according to the following reaction: 42

This takes place

Hydrogen then requires for its combustion half its volume of oxygen; hence
this 49.3 ml. represents a volume of hydrogen with 72 its volume of oxygen, or
3/2 volumes; hence the volume of hydrogen is 32.9 ml.
(b) Hempel's Method.-Of the 82 mL of gas remaining after the "absorptions, 13.2 mL were used for the explosion; 86.4 mL air introduced, ~iving a
total volume of 99.6 ml.
Residue after explosion. . . . . . . . . . . ..... . . . . . . . . . . . . . . .. 78.0 ml.
Residue after CO 2 absorption ........ : . . . . . . . . . . . . . . . . .. 73.2 :~
Carbon dioxide formed ....... " . . . . . . . . . . . . . . . . . . . . . . . . 4.8
Contraction ................................ 99.6-78.0=21.6 "

Residue after 0 absorption . ...... ~ ................... .. 70.2 ".

Oxygen in excess ............... " ........... 73.2-70.2= 3.0 "

The carbon dioxide being equal to the methane present, in the 13.2 ml. of
gas burned there were 4.8 ml. of methane. The volume of methane is found
by the proportion 13.2 : 82 : : 4.8 : x, whence x = 29.8 ml.
,
The hydrogen is calculated similarly.
Another method for the estimatiop. of hydrogen is by absorption with
palladium sponge; 4a it, however, must be carefully prepared, and it is the
author's experience that one cannot be sure of its efficacy when it is desired to
make use of it. A still better absorbent of hydrdgen 44 is a i % solution of
palladous chloride at 50 C.; when fresh this WIll absorb 20-50 mL of hydrogen
in ninety minutes. A proportionately longer time is required if more hydrogen
be present or the solution nearly saturated. The methane could then be determined by explosion or by mixing with air and passing to and fro over a whitehot platinum spiral in a tubulated pipette called the grisoumeter 45 (grisou
= methane).
Nitrogen.-There being no direct and convenient method for its estimation
with this apparatus, the percentage is obtained l?y finding the difference between
the sum of all percentages of the gases determined and 100%.
New 46 determined nitrogen in illuminating gas/directly after tl).e method of
Dumas in organic substances; 150 ml. of gas are 'used, the hydrocarbons par42 H 20 being as steam at 100 C.
At ordinary temperatures this is condensed, giving
rise to "total contraction."
43 Hempel, Berichte, 12, 636 and 1006 (1879).
44 Campbell and Hart, Am. Chern. J., 18, 294 (1896)._
45 Winkler, Fres. Zeit., 28, 269 and 288.
~:__
46 J. Soc. Chern. Ind., 11, 415 (1892).

GAS ANALYSIS

2359

tially absorbed by fuming sulfuric acid and the remainder burned in a combustion tube with copper oxide; the carbon dioxide is a.bsorbed and the residual
nitrogen collected and measured.
Accuracy and Time Required.-For the absorptions the apparatus is
accurate to 0.1 ml.; for explosions by Hinman's method 47 the methane can be
determined within 0.2%, the hydrogen within 0.3%; by Hempel's method
within 1% for the methane and 7.5% for the hydrogen. The time required
by the student for the analysis of city gas is from three to three and one-half
hours; for air, from fifteen to twenty minutes. For very accurate work the apparatus of Burrell should be used.
NOTEs.-The object in filling the capillaries of the explosion pipettes with water or
mercury before the explosion is to prevent the bursting of the rubber connectors on them.
With mercury this is effected by introducing it through the T-joint in the connector.
After testing for oxygen with the pyrogallate a small quantity of dilute acetic acid is
sucked into the burette to neutralize any alkali which by any chance may have been
sucked over into it. The acid is rinsed out with water and this is forced out by mercury
before the burette is used again .
. For results of the highest degree of accuracy the following points must be borne in
~~

1. The burette must be calibrated.

2. The confining liquid in the burette must be saturated with gas.

3. The reagents must be saturated with the gases for which they are not the reagent:
a blank analysis should be run when the pipettes are first filled.
4. :rhe reagents must be capable of absorbing their intended gases, i.e., not be nearly
exhausted.
5. Check readings must be obtained after allowing the burette to drain 2 minutes.
6. Make sure connectors are tight, glass tubes close together, and that too great
difference~ in the leveling tube and burette are not suddenly applied.
7. The ratio of combustible to non-combustible gases should not be less than 1 : 6.
For safety the explosion pipette must be screened.
The method of analysis of the residue after the absorptions have been made
by explosion is open to two objections: 1st, the danger of burning nitrogen by
the violence of the explosion; and 2d, the danger of breakage of the apparatus
and possible injury to the operator. These may be obviated by employing
the apparatus of Dennis and Hopkins,48 which is practically a grisoumeter or
slow combustion pipette with mercury as the confining liquid; or that of
Jaeger,49 who burns the gases in a hard-glass or quartz tube filled with copper
oxide. By heating to -250 0 C. nothing but hydrogen is burned; if the temperature be carried to 3000 ,l0 and CO be present, both it and hydrogen can be
burned, and from the CO 2 formed, the amount of CO determined, thus saving
the time-consuming absorption; higher heating of the residue burns the methane. The air contained in the tube can be determined as in the case of the
tube containing iodic anhydride, page 2356.
To the method of explosion by the mixture of an aliquot part of the residue
with air, method (b), there is the objection that the carbon dioxide formed is
measured over water in a moist burette, giving abundant opportunities for its
absorption, and that the errors in analysis are multiplied by' about six, in the
example by 820/132.

Gill and Hunt, loco cit.

J. Am. Chern. Soc., 21, 398 (1899).
49 J. Gasbeleuchtung, 41, 764.
Abstr. J. Soc. Chern. Ind., 17, 1190 (1898).
47

48

2360

GAS ANALYSIS
,

THE PROPERTIES OF THE VARIOUS GASE$ AND'THE METHODS

OF THEIR DETERMINAirION
Group I-Gases soluble in H 2S0 4 , 1.7 sp.gr. A~monia, Nitrous anhydride,
Nitrogen tetroxide (also traces of propane, butane, propene and butene).
I

Ammonia was studied by Priestley in 1774, who named it " alcaline air."
Ammonia is a pungent, strong smelling gas, lighter than air, specific gravity
about 0.6 50 (air being 1), dissolving readily in water, giving an alkaline solution.
It is the most soluble gas, water dissolving about 700 volumes 50 at or<1inary
temperature (20). Alcohol dissolves 340 volumes, and ether dissolves 10
volumes at 15. Liquid ammonia has been an article of commerce for many
years, being used in the refrigeration industry. At ordinary temperature
liquid ammonia exerts a pressure of abbut 125 lbs. per sq. inch (8.5 atmpspheres) and it boils at -33.5. The capacity and sizes of the cylinders in
which it is sold are given in the Appendix, Table 8.
Cylinders containing liquefied gases must be stored in a cool place and
must not be handled roughly. They must be treated like a high explosiveas some of t)1em are-the writer knowing of several cases in which ~asualties
and deaths even have resulted from thoughtless handling. In handling ammonia the canister of the gas mask is filled with cupramite. 51 When inhaled
none of the gas is absorbed,s2 the eff~ct being entirely on the surface: 0.0.4% ,.,
irritates the throat and 0.3% is dangerous for short exposure.
Ammonia burns with a yellow flame with pure oxygen, it is explosive with
air, the explosive limits being from 16.5 to 26.8%.53 Accidents in refrigerating
plants have occurred from this cause. 54 From its high hydrogen content it .
behaves like hydrogen, reducing the oxides of the metals. It is decotrlposed
by the electric spark into its elements. The source of ammonia was th.e distillation of coal,55 but at present it is synthesized, using pressures of abottt 200
atmospheres and a temperature of 450 with a catalyst of iron and traces of
other metals. In nature it is formed by the decQmposition of protein, giving
the "albuminoid ammonia" and "free ammonia" of th~ water analyst.
These in turn are oxidized to nitrites and nitrates, forming the nitrogen cycle.
Qualitative detection is by the white fumes produced by strong hydrochloric
acid, and by the odor.
Quantitative determinliction: by (a) absorption in wl\.~er or standard acid and
subsequent titration, 'cochineal often being used- as an indicator, Table 9.
(b) In the case of tarry or highly colored solutions, an apparatus like the
nitrometer, p. 2418, may be used and the solution treated WIth sodium hypobromite, when a molecule of nitrogen is evolvt:\.d for every 2 mols of free aminonia
present.
2NH3+3NaBrO=N2+3Na~r+3H20.
For accurate figures see Appendix, Table 3 . '
\
51 Cupramite,is a mixture of 1.5 parts 8-14 mesh pumice with 1 part CUS04, 5H 20;
45 cu. in. will protect against 2% ammonia for 5 hours.
52 Henderson and Haggard, "Noxious Gases."
I
53 Refrigerating Engineering, 17, 65 (1929).
J. Gasbeleuchtung, 57, 941 (1914).
54 Power, 72, 128 (1930).
__;-:55 Ammonia is formed when very dilute nitric. acid acts upon zinc.
50

GAS ANALYSIS

2361

Or the gas may be strained through cotton, and passed into 10% sulfuric acid
in the familiar gas washing bottles, 5 to 10 cubic feet of gas being used.
The acid is transferred to an ammopia distillation apparatus, as that of
Kjeldahl, the contents made alkaline with sodium hydroxide, distilled into
standard acid and titrated with standard alkali. The distillation is continued
until the steam from the still does not change turmeric paper. (c) By Nessler's
solution. This is a solution of mercuric iodide in potassium iodide, made
strongly alkaline with potassium hydroxide which gives a yellow red coloration
~ith ammonium compounds: this is matched in tubes with colors produced by
standard solutions of ammonium chloride. The process is rapid and very
delicate, one part nitrogen as NHa being detected in 1 billion parts of water.
It is particularly used in water analysis.
Nitrous-Anhydride N 20a according to some authorities does not exist. It is
very unstable and behaves like a mixture of nitrogen dioxide and nitric oxide.
Nitrogen dioxide, Nitrogen peroxide, NO z,. or Nitrogen tetroxide, N 20 4
(cold).-Nitrogen dioxide is a reddish-brown acid gas heavier than air, of about
1.6 sp.gr. It is readily soluble in ,water forming nitric acid and nitric oxide,
H 20+3N0 2=2HNO a+NO (hot) and N204+HzO=HNOa+HN02 (cold). It
is prepared by the heating of lead nitrate 2Pb(NO a)2=4NO z+2PbO+0 2: also
by the action of concentrated nitric acid on copper or tin. At _9 it is a white
crystalline solid; at 26 a bright liquid; at 150 a deep brown gas.
On account of the high content of oxygen, nitrogen dioxide is an energetic
oxidizing agent, carbon monoxide and .phosphorus being oxidized by it. The
dry liquid gas mixed with benzene, C 6H6 or naphthalene C10HS forms a powerfur explosive when detonated, the pl;oducts being wholly gaseous. It was used
in depth bombs.
Physiological Action.-" Nitrous fumes" are intensely irritating to the
respiratory tract, and are the most insidious of the irritant gases. After inhaling the fumes from the burning of dynamite, celluloid 56 or from action of
nitric acid on metals or sawdust, one may be little inconvenienced and appear
perfectly well. Before noon of the next day or even before the morning he
may be dead, from the volume of liquid coming from the dropsy of the lungs;
.011 to .015% is dangerous and twice this quantity rapidly fatal.
Qualitative detection-by means of iodo starch pap_er; or absorb in alkali
and test for nitrates.

Quantitative determination-by absorption in standard potassium permanganate acidified with sulfuric acid.

Allison, Parker and Jones 51 determined small quantities of " nitrous fumes"
in the air by treating a known quantity of the gas with caustic soda solution,
evaporating, and treating the residue with phenol sulfonic acid after the procedure for nitrates in potable waters.
..
56 See Proceedings of a Board of the Chemical Warfare Service, etc. which investigated the disaster at the Cleveland Hospital Clinic, Cleveland, Ohio, May 15, 1929.
67 Bur. Mines Tech. Paper 249 (1921).

2362

GAS ANALYSIS

Group II-Gases soluble in KOH, 1.3 sp.gr. Chlorine, Hydrochloric acid,

Cyanogen, Hydrocyanic acid, Hydrogen sulfide, Sulfur dioxide, Silicon
fluoride, and Carbon dioxide.

~n

W~b~

P74,
named it " DephlogistiChlorine was discovered by Scheele
cated marine acid air;" .8 Davy called it chlorine ,in 1810. Chlorine is a
greenish yellow, stifling gas, about 2.5 times as heavy as air, dissolving readily
in water to the extent of about twice its volume at 20 .9 Liquid chlorine boils
at -33.6 and at 20 exerts a pressllre of 97 lbs. per sq. in. (6.6 ats.). It has
been an article of commerce for many years, and is now shipped in tank cars.
Chlorine is the most active gas except fluorine: it attacks the' metals' and
substances containing hydrogen, combining with it, decomposing them, destroying evil smells and colors-hence its use as a disinfectant and bleaching
agent. Brought in contact with water the following equa,tion takes place:
H 20+Cb=HOCl+HCl. The hypochlorous acid, HCIO decomposes, 2HCIO
= 2HCI +02 and the" nascent" 'oxygen really is the active agent, rather than
the chlorine. But this reaction also takes place, H 2S+Cb=2HCI+S. As a
bleaching agent, where it can be psed, chlorine is much more effective than some
others, as s~lfurous acid. Passed into caustic potash it forms hypochlorite or
chlorate according to the temperature" the equations being 2KbH +Ch
=KCIO+KCI+H 20 and 6KOH+3Ch=5KC1+KCIO a +3H 20. Chlorine
when passed into water at 0 forms a crystalline hydrate C1 2, 8H 20 which when
heated yields the pure (hydrous) gas. /
.
Chlorine finds many and important uses-in bleaching, and the preparation
of organic compounds, "as war gases, carbon tetrachloride (pyrene), chloroform,
and the chlorides of the elements generally.
Chlorine is made chiefly by the electrolysis of brine, caustic soda being an
important by-product. Where the Leblanc soda process is still employed, it is
made as did Scheele by the oxidation of hydrogen _in hydrochloric acid with
manganese dioxide, 4HCI+Mn02=MnCI2+CI 2 +2H 20.
Physiological Action.-Chlorine produces irritation of the tissues, probably
by their oxidation, as has already been explainea in the case of bleaching:
.0015% irritates the throat and .004% is dangerous; 0.1 % is fatal.
Chlorine can be removed from hydrochioric acid gas by treatment with
powder_ed antimony.
Hydrochloric acid (hydrogen chloride) was prepal'(~d by the alchemist Geber
in 1648, and in the anhydrous condition by Priestleyin 17J2, who called it
" marine acid air." It is a colorless gas, heavier than air (sp.gr.l.25), strongly
acid, and fumes when brought in contact with moisture. It is a remarkably
stable gas, requiring a temperature of 1600 to 1800 for its decomposition; at
this temperature it burns with oxygen. It is the second most soluble gas, water
at ordinary temperatures dissolving 450 volum,es:-an aqueous solution containing 20.24%, distils at a constant temperature of 110 at 760 mm. Alcohol
gives a 40% solution,. it is also soluble in ether and hydrocarbons. It has been
condensed to a liquid having a pressure of 42 atmospheres at 18 (620Ibs.).60
.8 I.e., "marine acid" (RCl) from which the phlogiston (R) had been removed.

Accurate figures in Table 3.

-The liquid gas is contained in a tube 5.5 mm: irtteynal diameter and 1 mm. thickness
of wall, which gives an idea of the pressure such tubes may be expected to withstand.
69

60

GAS ANALYSIS

2363

Qualitative detection: by the dense white fumes with ammonia.

QuaIrtitative determination: by absorption in standard alkali or silver
nitrate.
Hydrochloric acid can be removed from chlorine by passing the gas over
manganese dioxide or zinc oxide.
Cyanogen or dicyanogen, C 2 N 2, was discovered by Gay Lussac in 1815, and
is made by heating mercuric cyanide: Hg(CN)2=Hg+C 2 N 2. It is a poisonous,
colorless gas with an odor resembling prussic acid, heavier than air (sp.gr. 1.8):
it burns with a purplish flame, when liquefied it has a pressure of 73.5 lbs.
(5 atmospheres) at 20. At this temperature one volume of water dissolves
4.5 volumes of the gas. Alcohol dissolves 23 volumes and ether 5, in which .it
decomposes very slowly. The aqueous solution turns dark, precipitates
brown flakes, and contains a quantity of nitrogenous compounds, as oxamide,
prussic acid and urea.
.
Cyanogen is most readily prepared by heating the mixed concentrated.
solutions of potassiu,:rp. cyanide and copper sulfate .
. Qualitative Detection and Quantitative Determination.-By solution in
caustic soda and testing for cyanide. C 2 N 2 +2KOH=KCN+KCNO+H 20.
Cyanogen may be determined by Nauss' method 61 consisting in its absorption
in ferrous hydroxide or in presence of hydrocyanic acid by Rhodes method 62 by
absorption in silver nitrate: this precipitates the acid and the cyanogen can be
separated by drawing air through the solution. Another method consists in
absorbing in alkali hydroxide acidifying with acetic acid, adding a few drops of
copper sulfate (1 : 10), a drop or two of pyridine, and 5 ml. of chloroform. A
blue coloration in the chlorofoqn layer indicates a cyanate-which is not given
by a cyanide. The following equations serve to distinguish between the gases:

Hydrocyanic Acid or Prussic Acid ~CN was prepared by Scheele in 1782.

It is easily made by treating potassium or sodium cyanide with sulfuric acid
1 : 1. It is a colorless mobile liquid of a boiling point of 25.7, having a peculiar
odor resembling bitter almonds. The characteristic odor of potassium cyanide
is due to this gas, being such a weak acid that it is displaced by the carbonic
acid of th~ air. The gas is lighter than air (sp.gr. 0.93) and is one of the most
powerful and rapid poisons. It is readily soluble in water, alcohol and ether.
Both the liquid gas and the concentrated aqueous solution are readily inflammable, burning with a violet flame.
Hydrocyanic acid has of late years found important use as an insecticide,
particularly against the scale insects infesting citrus groves, and for fumigating
storage warehouses' for grains and furniture. 63 It is made by the action of
sulfuric acid upon sodium cyanide and in California the 97% acid is shipped in
J. Gasbeleuchtung, 43, 696 (1900).
Ind. Erig. Chern., 4, 652 (1912).
63 "The Modern Miller," Feb. 1927; The Furniture Warehouseman, 6, May (1926);
Pacific Marine Review, Nov. 1926.
61

62

2364

GAS ANALYSIS

tinned iron drums; 64 in 1926 Metzger 6y discovered that acid calcium cyanide
Ca(CNk 2HCN, gives off three-fourths of the acid in half a minut~ and the
whole of it in a minute. It is extremely sensitive to moisture: CaH 2(CN)
+2H 2 0= Ca(OHh+4HCN. It is known as "po,~dered hydrocyanic acid."
Physiological Action.-Cyanides or cyanogen artest oxidation in the tissues
and produce asphyxia. The person falls after a few breaths of the polluted
atmosphere and dies in 6 to 8 minutes. 0.012% is dangerous and 0.3% rapidly
fatal. Ammonia and chlorine are the antidotes. Like" nitrous fumes,"
it is evolved when celluloid burns.
Qualitative Detection and Quantitative Deterroination.-By absorp.tion in
water or standard alkali, and testing for a cyanide: or it may be absorbed in
standard silver nitrate, or converted to a ferricyanide. 66
Cyanogen compounds are formed whenever nitrogenous and carbonaceous
compounds cpme together at high temperatures, particularly if alkali be present.
They are ,found in blast furnace, coal and destructor gases.
Hydrogen Sulfide H 2S W5,S carefully investigated by Scheele, who synthesized it from its elements. It occurs in the natural gas of certain Texas localities to the extent of 20%,67 and in " sulfur' springs." It is a colorless gas of well
known odor, a little heavier than air (sp.gr. 1.18), boiling at -62: at 20 the
liquid gas exerts a pressure of about 150 lbs. per sq. in. (10 atmospheres). It is
readily decomposed-hence a good reducing agent; it is combustible, explosive
limits 4.3-46.0%, and soluble in water to the extent of 3 m!. per m!. of 'Yater
at 20,68 alcohol dissolves about 7 volumes. The usual method of preparation
from ferrous sulfide gives a gas containing hydrogen (40% in some cases) from
the excess iron. For the preparation of the pure gas, a pure sulfide, as Stibnite,
should be employed.
Physiological Action.-Sulfuretted hydrogen is nearly as toxic as prussic
acid, and its action may be as rapid. It is a non-cumulative poison, is 'rapidly
oxidized by the blood and its products are non-toxic; 0.01-0.015%, may produce
death in from 8 to 48 hours.59 .05 to .07% is dangerous and 0.1 to 0.3 rapidly
fatal. 70

/
Qualitative detection by " lead paper," i.e., filter paper moistened with a
solution of lead acetate giving a black or brown coloration.
Quantitative Determination.-It can b~ absorbed in a standard solution
of iodine in potassium iodide. Small quantities can be determined colorimetrically by p-amino dimethyl aniline sulfate. 71
Hydrogen sulfide 'can be separated from carbon-dioxi\!e, hydrogen and
hydrocarbons by powdered manganese dioxide.
Sulfur Dioxide S02-was first prepared in the pure state by Priestley in
1775, by the action of sulfuric acid on mercury, as the result of the breaking of a
thermometer in distilling the acid. It is a colorless gas of suffocating odQr, a
Ind. Eng. Chern., 25, 959 (1933).
Id., 18, 161 (1926).
66 Rhodes, Ind. Eng. Chern., 4,652 (1912).
67 Bur. Mines Reports of Investigations, No. 3128 (1931).
68 The solution saturated at 0 contains 4.3 rnl. per rnl. of water, and used as a gargle
or spray, is almost a. specific cure for tonsilitis.
,
69 Yant and Sayers, J. Chern. Ed., 4, 613 (1927).70 Haggard. Chern. Abstr., 19, 2095 (1925). . - ' 71 Sheppard and Hudson, Ind. Eng. Chern., Anal.-Ed. 2, 73 (1930).
64

66

GAS ANALYSIS

2365

little more than twice as heavy as air (sp.gr. 2.2), readily soluble in water, 1 ml.
dissolving 40 ml. at 20,72 alcohol dissolves 32%. At this temperature the
liquid exerts a pressure of 48 Ibs. per sq. in. (3}i atmospheres). It boils at
_10. It is an energetic reducing agent but a very stable gas. It is particularly harmful to vegetation, especially conifers. Three to 4 pll-rts per million
injure the leaves of plants after one hour exposure. 73 Therefore it is recommended that in works "here it is used, or in chemical works in general, that well
"authenticated, clear and dated photographs of the surroundings-particularly
with reference to arboreal or plant damage--be made from time to time.
Physiological Action.-It is said to cause decomposition of haemoglobin,
inflammation, and irritation of the respiratory tract may occur. Acute poisoning f~om the gas is rare, on account of the sensitivity of the eyes and throat to
it-it being the gas which is most easily perceptible. .0008% irritates the
'throat and .04% is dangerous.
Qualitative determination-by the bleaching of fuchsin paper or iodo-starch
paper .
.Quantitative determination-in standard iodine solution 74 S02+ 21 + 2H 20
= H 2S0 4 +2HI. This is the method gi ven on p. 2399. Sweeney, Outcault and
Withrow 75 advocate the use of .005 N permanganate; or by absorption in
standard alkali. 76 See also the report of the Selby Smelter, Commission Bur.
Mines. Bull. 98 (1915-).
As has' been stated, it is absorbed by zinc-filled rubber tubing. 77
Silicon fluoride or tetrafluoride was discovered by Scheele in 1771 and
investigated by Berzelius in 1823. It is a colorless gas fuming strongly in the
air, of odor resembling hydrochloric acid: It has a specific gravity of 3.6. It is
r'eadily decomposed by water, 'precipitating silicic hydroxide, 3SiF 4 +4H 20
=Si(OH)I+2H zSiF6 It has been liquefied-the liquid is said to have a
tension of 50 Ibs. (3.4 ats.) at 20. It is formed when hydrofluoric acid acts
upon silica or silicates:
Qualitative and quantitative determination-by the use of the decomposition with water.
Carbon Dioxide or Carbonic Acid, CO 2-was investigated by Bergman in
1774 who gave it the name of " acid of air." Layoisier showed its chemical
nature. It is a normal constituent of the air, present to the extent of 3 parts
per 10,000. It is a colorless and odorless gas, 50% heavier than air (sp.gr. 1.5);
but hot CO 2 at 150 for example, is lighter than air. Five to seven per cent
mixed with oxygen 78 serve as a sti,rnulant in artificial respiration, and a similar
effect is observed with carbonated waters and wines. Three or 4% in the air is
noticeable,19 6% produces panting, 7 to 8% more distress, increasing at 11 %,
15% extinguishes lights, and 50% causes death, the action being the exclusion
For accurate figures see Table 3.
Zimmerman and Crocker, J. Chern .. Ed., 8, 1344 (1931).
74 Bureau of Mines Bulletin 98.
7~ Ind. Eng. Chern., 9, 949 (1917).
76 Ind. Eng. Chern., 18, 747 (1926).
77 Power, 58, 507.
78 Henderson, Science, 83, 402 (1936).
79 Haldane, Am. Gas Lt. J., 65, 483.
72

73

2366

GAS ANALYSIS

of oxygen. It is a remarkably stable gas, requiring a very high temperature

-20000-for its decomposition. Water at ordinary temperature dissolves its
own volume of the gas; at a pressure of 3 to 4 atmospheres it becomes" soda
water" 80 of commerce. Liquid carbonic acid e'japorates at _79 and the
solid melts at _56. It has no boiling point. At120 it exerts a pressure of
870lbs. (59 atmospheres). From a cylinder of the c<?mpressed gas when o~ened
the gas escapes as a snow-like solid. This is compre~ed at a pressure of 500
to 800 pounds per sq. in., into cakes known as" dry ice," 81 dissolved in acetone
it is a very convenient method of producing low temperatures in the laboratory.
Qualitative detection by calcium or bari'um hydroxide on 'an ebonite rod.
Quantitative Determination.-By standard alkali or absorption in sodium or
potassium hydroxide and noting contraction in volume. Scheibler's calcimeter
is an ingenious apparatus, in which carbon dioxide evolved from bone charcoal
(containing calcium carbonate) and acid displaces an equal volume of air from.
a rubber bag inside a Woulff bottle, into a burette. The burette is graduated in
per cent of calcium carbonate, 1.7 grams of bone char being taken.
Carbon dioxide has of late found estended use: with a peculiar nozzle, the
liquefied gas serves as a fire extinguisher, as,a source of decentralized power for
spray painting, unplugging pipes, distilling tar 'and petroleum and displacing
flammable solvents and oil vapors in tankers. Used aboard ship as a refrigerant. 82 It may, if the percentage be high, cause the development of objectionable flavors 83 with fresh fruits and vegetables, for these carryon rellpiration. The solid gas compressed into oricks has replaced ice to a considerable
extent. It is said to be used also as a rat-killer and tire inflater.
Carbon dioxide in illuminating gas cuts down the i1luminati~g power almost
in the proportion to which it is present.
Group III-Gases Soluble in AgNO a Solution.-Phosphine PH a, Arsine AsH a,
Stibine SbH a
These gases are all poisonous, readily decompqsed with the liberation of
hydrogen, form chlorides) are all combustible ~nd decompose explosively.
They all combine with silver when passed into silver nitrate~ setting free nitric
acid. This acts upon the precipitated silver ,compound forming phosphoric or
arsenious acid; this does not take place with stibine, PH a +3AgNO a =Ag aP
+3HNO a They are not commonly met with except as impurities in other
ga~s.
_ ..
Phosphine PH a was' prepared by Gengembre in 1783 by heating phosphorus
with potassium hydroxide, giving at the same time a liquid. P 2H 4 , which is
spontaneously inflammable. 3KOH +4P+3H 2 0 = PH a + KH 2 P0 2 .As ordinarily encountered, phosphine smells like ,rotten fish and comes from the
decomposition of the metallic phosphides by acid. The pure gas inflames above
100 and it takes fire when treated with nitric acid or bromine. It is somewhat
80 It is interesting to note that perhaps Cleopatra made the first, carbonated drink,
when she dissolved a valuable pearl in vinegar. Dr. P. C. Physick of .Philadelphia prescribed soda water as a medicine in 1807, which was put up by Townsend Speakman. It
became very popular, and in 1928 the annual business in carbonated drinks amounted to
800 million dollars.
"._
81 Killeffer, Ind. Eng. Chem., 19, 192 (1927).
,_,.82 Power, 67, 805 (1928).
83 Thornton, Ind. Eng, Chern., 22, 1186 (1930).
Brooks, Power, 67, 805 (1928).

tiAS ANALYSIS

2367

heavier than air having a specific gravity of 1.18. It is very slightly soluble in
water, 100 ml. dissolving 2.6 m!. Alcohol dissolves about half its volume, and
ether twice its vqlume. Q'?lalitative detection-by the blackening of silver
nitrate solution in the absence of hydrogen sulfide and the other members of
this group.
Quantitative determination-by converting to phosphoric acid by burning,
after the manner of the determination of sulfur in city gas (p. 2395) and subsequent oxidation with bromine and precipitation of the phosphoric acid as
magnesium phosphate in the usual way. Dennis and O'Brien 84 recommend
the determination in acetylene by absorption in sodium hypochlorite and ultimate precipitation as magnesium phosphate: wt. Mg 2 P 20 7 XO.3059 =wt. PH a or
wt. Mg 2P 20 7 X 201.25 = m!. PHa.
Arsine, AsH a, was discovered by Scheele in 1775, who prepared it by heating
a solution of arsenic acid with zinc. It is evolved when acids act upon arsenical metals. It is a colorless gas of disagreeable-odor, extremely poisonous,85
easily decomposed by heat and readily inflammable. It is 2.7 times as heavy
as ~ir; water dissolves 1/5 its volume of the gas, it is insoluble in alcohol and
ether; it liquefies at - 55.
When inhaled it produces dizziness, faintness and headache, followed by
nausea and vomiting; it combines readily with hemoglobin, producing anemia.
In acute cases death ensues in from 2 to 6 days.
Qualitative detection-by passing into ammoniacal silver nitrate.
Quantitative determination-by the familiar Marsh test; also by the
Gutzeit 86 test: 2AsHa+12HgC12+3HzO=12HgCl+AszOa+12HC1; the calomel can be weighed, or potassium iodide solution is added sufficient to dissolve
it; 25% excess iodine solution added, and the excess iodine titrated with N /100
.
thiosulfate: 1 mg. As 20 a =8.08 ml. N7100.thiosulfate.
Stibine, SbH a, was prepared by Lewis Thompson in 1837; it is formed
whenever " nascent" hydrogen comes in contact with a soluble antimony s~lt.
It is a colorless gas, of odor faintly resembling hydrogen sulfide, inflammable,
easily decomposed by heat or even strong sulfuric acid. It is more than four
times as heavy as air, of boiling point -17, and is soluble in water to the extent
of 1/5 its volume, 1 m!. alcohol dissolves 15 ml.
Determination-by passing into silver nitrate, forming SbAg a, washing and
dissolving in tartaric acid and determining the antimony.
Group IV-Gases soluble in potassium pyrogallate.
Oxygen, the active constituent of the air, was known to the Chinese in the
eighth century; they even knew it could be made by heating saltpeter: 87
84 Ind. Eng. Chem., 4, 834 (1912).
85 It is stated that Gehlen lost his life in 1815 by the inhalation of a single bubble of
the gas.
86 Barnes and Murray, Ind. Eng. Chem., Anal. Ed., 2, 29 (1930).
87 This fact is made use of periodically, to foist upon an unsuspecting public the
various "coal savers" and "aids to combustion," to "burn up the noxious gases of
combustion, and safeguard the health of the family." They do contain a little Chili
saltpeter, sometimes permanganate, but their chief constituent is usually common salt.
This colors the slightly visible flames of carbon monoxide, so that the combustion appears to be energetic, and the furnace very hot. Two pounds are sufficient to treat a
ton of coal. 'When it is remembered that about 2 pounds of oxygen are required to burn
a single pound of coal, it will readily be seen how much influence this compound will have
.
in burning a ton of coal!

2368

GAS ANALYSIS

Mayow in 1669 found the percentage in the air, but'it remained for Scheele in
1771-2, and in 1774, Priestley, to prepare it in a pure state. They called it the
" vital spirit" or " fire air." It is prepared today,c.ommercially by the distillation of liquid air and the electrolysis of water. It is evolved by green
plants, particularly under the influence of sunlight, with the formation of
cellulose, 6C0 2+5H 20 = CSH1006+602. This is perhaps the most importjint
reaction taking place in the world-the carbonic dioxide evolved by combustion
and animals, being utilized by the plants for their sustenance, and they in turn
contribute oxygen for ours.
Oxygen is 10% heavier than air, and at ordinary temperatures a Ji.ter of
water dissolves 6.4 ml. of oxygen from the air, or 28.4 mI. from pure oxygen. It
dissolves to the extent of 10 volumes in molten silver, which when cooled suddenly evolves it, " the spitting of silver," causing losses sometimes in assaying.
It is colorless, odorless and tasteless. It was liquefied in 1877, becoming a steel
blue liquid, imU boiled at -182.5 (nitrogen -195.5). It comes in commerce
in 110 and 220 cu. ft. cylinders under a pressure of 1800 lbs. per sq. in. at ordinary temperature.
If the atmosphere contains les!; than 17% oxygen oil lamps refuse to burn,
at 16% gas does not burn, at 12% respiration becomes deeper, at 10%,respiration becomes more difficult; the lips become blue, 8% the face is leaden color,
and at 5 to 6%, panting and death ensue. 88
Compressed oxygen is one of the co~monest gases of today. The cylinders
are provided with right-hand threads, whereas all the combustible gas cylinders
have left-hand threads, and are served by oil lubricated pumps. "Never put
oxygen into a cylinder with a left-hand thread "-is one of the cardinal principles
to be observed in handling compressed gases. It is a " dangerous or hazardous
material," particularly in contact with oil, even in minute quantities, as ,little
as 0.6 mg. (1/50 drop) causing fire in the presence of high pressure oxygen. 89
Four drops of oil, in a 60 ml. bomb under a pressure of 2300 pounds causes an
explosion. 90 In a case which the writer investigated, 1800 pounds pressure
when exerted nearly 24 hours, blew a 220 ft. cylind~f of compressed oxygen to
pieces, caused the death of two persons, and wreckage of the laboratory; the
floor above was blown up with such violence, as to break the legs of persons
standing upon it; windows were broken two hundred feet away. The amount
of oil of course could not be determined, but other cylinders from the same lot
contained from 50 to lQO mgs. of mineral oil: they were._all intended for combustible gases, but as they were new and supposedly never"used, they were
temporarily used to hold oxygen, a special fitting being made for the purpose.
Cases are known in which people have been burned, and iron fitting& set fire
by incautious oiling, as firewood is kindled by kerosene. Water is the only
lubricant to be used on an oxygen compressor, with a trace of soap in it for the
fittings. Packings are of lead, asbestos or "nearly incombustible rubber."
Iron or steel should not be used for fittings for high pressure oxygen. A verdict
of $340,000 91 was awarded, due to a fire caused by oil in a steel manifold, even

Burrell and Oberiell, Bur. Mines Tech. Paper 122 (1915). I

News Edition of Ind. Eng. Chern., 2, 6 (Dec. 20,)924).
90 Hersey, C. A., 17, 3254 (1923).
Also Brooks, 0;-"1\:., 18, 905 (1924).
91 News Edition, Ind. Eng. Chern., 2, 6 (Dec. 20. 1924).
88

89

GAS ANALYSIS

2369

though it was made in conformity with current practice. Brass or bronze

fittings must be used-these melt but do not fire or burn. 92
Used with compressed hydrocarbon gases or gasoline it finds extensive
employment in cutting and welding iron and steel; this has developed an extensive technique and literature. It is to be noted that much depends upon
the purity of the oxygen, both as regard the time consumed and the quality
of cut produced. If 99.3% of oxygen or better, be regarded as standard, using
hydrogen, a drop of a single per cent in purity increases the amount used by
25%, and the time by 7%; a drop to 95% purity makes a ragged cut, increases
the time by 40% and the quantity of gas used by 60%.93 The temperatures
attained are 1800 with manufactured or illuminating gas, 2100 with hydrogen
and 2450 with acetylene. 94
Quatitative detection-by the familiar glowing splinter or by the passage of
the gas into nearly colorless pyrogallate of potassium. This can be prepared by
using well boiled water to dissolve the hydroxide and acid and mixing them out
of access of air .
. Quantitative Determination.-In case combustible gases are absent, by
explosion with hydrogen (p. 2355); otherwise absorbents as phosphorus or
potassium or sodium pyrogallate is used; this with much oxygen may evolve
CO. See Reagents. Phosphorus does not absorb pure oxygen; if the concentration be reduced to 75% they combine with explosive violence. Burrell
uses a stabilized solution of chromo us chloride called" oxsorbent " with success;
another excellent absorbent when carbon monoxide and acetylene are absent, is
ammoniacal cuprous carbonate or chloride. This is prepared by treating bits
of copper or copper gauze, cleaned with nitric acid, with 500 ml. of a saturated
solution of ammonium carbonat'e mixed with an equal quantity of ammonium
hydroxide, sp.gr. of 0.96. Badger 95 recommends treating the copper with a
solution of equal parts of concentrated ammonia and water, saturated with
ammonium chloride; this absorbs from 50 to 60 times its volume of oxygen.
Oxygen should not contain more than 4% of hydrogen and preferably none
at all, and for respiration be of the highest quality containing not more than 2%
impurities.
For the determination of oxygen dissolved in water the method of L. W.
Winkler 96 is employed; the quantity of dissolved oxygen is usually a measure
of the purity of the water.
Ozone 0 3 , van Marum observed in 1785 that when frictional electric
machines wen~ operated, a peculiar odor was developed at the same time; a
similar odor was observed by the passage of a lightning spark. Little attention
was paid to this until SchOnbein brought it to the attention of chemists in
1840. He and others investigated it and found it not to be an oxide of hydrogen
as was thought, but a form of oxygen. The investigation was a difficult one
as the quantities of it formed were small and being so active it disappeared.
The odor resembles dilute chlorine; it is faintly blue in layers of a meter
thick, and about 60% heavier than air; it is one of the most powerful oxidizing
Jakowsky and Butzler, C. A., 17,3920 (1923).
93 Circular International Oxygen Co., 1911; also J. Soc. Chern. Ind., 30, 779 (1911).
94 Drews, "Verdichtete und Verfiuessigte Gase" (1928).
95 Ind. Eng. Chern., 12, 161 (1920).
96 Ind. Eng. Chern., 5, 976 (1913); also C. A., 7, 2266 (1913).
92

GAS ANALYSIS

2370

agents; 1 ml. of water at 12 dissolves 0.5. ml. ozone., It forms a blue liquid
boiling at -119 and is an endothermic compound. I
Qualitative Detection.-Ozone browns manganese chloride paper and
liberates iodine from potassium iodide. It does not affect permanganate as do
hydrogen peroxide and" nitrous fumes." It of coursJ does not respond to the
nitrite test with sulfanilic acid and alpha naphthylamin~ as do " nitrous fumes"
(N 20 a and NO), nor is it changed by chromic acid as is hydrogen peroxide . ..It
turns silver foil black which neither of the others affectsY
Quantitative Determination.-By the action upon a neutral solution of
potassium iodide, and titration of iodine with N /10 thiosulfate. 98 2KI +0 3
+H 20=2KOH+2I+0 2 If an acid solution be used, hydrogen peroxide is
formed, which sets free an additional quantity of iodine. For the rapid determination of ozone in the air of health resorts use may be made of the Wtlrster 99
test, the quantitative oxidation of fresh p-tetramethyl phenylene diamine
which forms a blue color on strips of paper which may be matched with
standards.
Ozone is formed when oils, as turpentine, o,xidize, or during the operation of
blast lamps. It is used industrially for, " killing odors," oxidizing andJ'leaching oils, waxes, ivory, flour and starch; it has been used for purifying water, but
chlorine is said to be cheaper-that, however, produces a taste. Great hopes
were entertained for ozone when it was first made in quantity~ that it could be
used for improving the ventilation of h!_l>lls and schoolhouses; the fact that the
atmosphere was too dry for its efficient use was overlooked, and little effect
resulted other than the production of sore throats for those in the immediate
neighborhood of the machine. lOo In handling ozone, glass or metal apparatus
must be used as it oxidizes all other materials. Flexibility of glass apparatus
can be ensured by spirals or by the use of joints sealed with mercury cup,s.
0

Group V-Gas absorbed by Cuprous Chloride Solution.-Carbon Monoxide,

CO.
.
{

It was obtained in 1776 by Lassone by heating zinc oxide with charcoal; this
gave a gas which burned with a blue flame and which for a long time was mistaken for hydrogen. Cruikshank in 1800 fOlmd it could be made by heating
various metallic oxides with carbon and named it " gaseous oxide of carbone."
Its method of preparation industrially (with' hydrogen) is by the well known
water gas reaction; in the laboratory it can be made by the-removal of water by
sulfuric acid from formic acid or oxalic acid (with an equal :YP1-ame of CO 2).
HCOOH = CO + H 20;

COOH=H 2 0+CO+C0 2

COOH
If potassium ferrocyanide or cyanide be cauti~)Usly heated with 10 times its
weight of conc. sulfuric acid, pure carbonic oxide is evolved (dilute acid gives
Keiser and McMaster, Arn. Chern. J., 39, 96 (1908). .
.
Treadwell and Anneler, Z. Anorg. Chern., 48, 86 (1905); also Treadwell-Hall,
"Quantitative Analysis."
._ .
'
99 Berichte, 20, 921 (1888).
Obtainable frorn_T--,"~uchardt, Goerlitz, Germany_
100 Czaplewski, Chern. Met . .Eng., 12,254.
C. A., 8, 2017 (1914).
97

98

GAS ANALYSIS

2371

hydrocyanic acid); K4Fe(CN)6+6H20+6H2S04=6CO+2K2S04+FeS04+

3(NH 4)2S04.
Carbon Monoxide is colorless, possessing a slight odor and taste. It is a
trifle lighter than air, and has been liquefied, boiling point -190; 100 mi. of
water at 20 dissolve 2.3 mI., alcohol dissolves about 10 times as much. It is a
stable gas;-combining with chlorine to make the deadly phosgene COCh, with
potassium to make the explosive K SC 60 6, and with nickel (cobalt), and iron to
make the carbonyls Ni(CO)4 boiling point 43, and Fe(CO)o-boiling point
102.5. The latter occurs in city gas and deposits iron oxide on the gas mantles,
injuring their efficiency. It is combustible, the explosive limits with air being
from 12.5 to 75%. When heated with oxides of the metals it is a most powerful
redu.cing "agent, producing the metals. By its production from coal less than
% the heating value of the carbon is evolved, 4340 B.t.u. instead of 14544
B.t.u. per pound.
Physiological Action and Common Sources of Carbon Monoxide-Sewer
Gas.-When inhaled, carbon monoxide displaces oxygen from the blood by
combining with the hemoglobin, the affinity for which is 300 times that for
oxygen; consequently as the blood cannot be sufficiently oxygenated, the
patient dies from asphyxiation. It acts too, as a cumulative poison, a little
being taken into the system now and again until a toxic dose is attained. Coal
gas contains 6 to 8%, water gas 25, producer 17 to 30, blast furnace 24, and
automobile exhaust gases 7% carbon monoxide. No one would think of staying
in a room into which coal gas was issuing from a one-inch pipe, and yet people
stay in a garage with an automobile ~ngine running, delivering more carbon
monoxide than coal gas itself. A 3D-horsepower engine running 15 minutes in
a closed garage of 5000 cubic feet capacity will make the air decidedly dangerous. lOl
In the streets we find from 0.3 to 2.9 parts per 10,000; average 0.8 parts; 102
in auto repair shops from 0.1 to 11 parts, average 2.1 parts; in auto busses from
0.1 to 1.0 part. To get an idea of the danger, Henderson suggests that the
number of parts of carbon monoxide in 10,000 be multiplied by the hours of
exposure: if this product be 3 there is no effect, at 6 it is just perceptible, at 9,
headache and nausea result, and if the concentration be 15, it is dangerous. In
congested thoroughfares from 0.8 to 2.0 parts per 10,000 may be encountered,
with the latter a headache may be expected after several hours exposure.
0.1 % (10 parts per 10,000) produces loss of power over the lower limbs, 0.2%
complete helplessness, and 0.3% (30 parts per 10,000) death; 10 m!. per kilogram weight of an animal will cause death, i.e., 800 m!. per person. When
gassed " it is extremely important to hasten the elimination by inhalation of
oxygen plus 5-7% carbon dioxide, and terminate the asphyxia." "A person
who is allowed to recover spontaneously in fresh air may remain unconscious
for some time and be ill a day or two, whereas another equally gassed, is
resuscitated by inhalation, goes back to work within a half hour with no ill
effects." 103
101 "Vitiation of Garage air by Automobile Exhaust Gases," Bur. Mines Tech. Paper
216 (1919).

"Atmospheric Pcllution by Automobile Exhaust Gases," Ind. Eng. Chern., News

Edition, Aug. 20, 1928, p. 20.
1~Henderson and Haggard, "Noxious Gases" already mentioned.
102

2372

GAS ANALYSIS

Sewer Gas.-Another source of carbon monoxide is (( sewer gas," i.e., city

gas which leaks into sewers, not sewage gas, coming from the fermentation of
sewage in the sewers. This is unusual if not impossible, as the temperature of
the sewage is such as not to favor fermentation, assuming that sewage collected
in the sewers, which are inclined to ensure a sufficient) flow to prevent acc.lation. In long sewers of faulty construction and with' stale sewage a gas having
a high content of methane and no carbonic oxide 'might 'be produced. Gas
leakage from mains under asphalt pavement was recognized in New York-in
1897; 104 illuminating gas is found in electric manholes. l05 Cast iron pipes with
bell and spigot joints, made witli oakum and cement, are usually buried 30
inches, and present potential sources of leakage every 12 feet. The modern
gasket ted coupling and leak clamps have largely reduced this leakage. The
normal leakage of a 6-inch main is 225 thousand cubic feet of gas per mile
annually: a 12-inch main doubles this. This may mean from 7.1 to 25.9%
of the gas made. In Massachusetts it is about 202 thousand cubic feet. City
gas, therefore, may come into the house through defective plumbing or more
likely it may follow from the trench containing the gas main along the service
gas or water pipes or from electric ducts pr conduits through the foundation
walls into the house. Besides this source, carbon monoxide is formed wherever
carbonaceous substances burn incompletely; even in manufactured gas, and particularly from charcoal braziers and coke burning stoves.16 It is also formed
by the explosion of smokeless powder, and occurs in gun turrets.
Qualitative detection-by the blood test or palladous chloride, pp. 2405-7.
Quantitative determination-by absorption in acid or ammoniacal cuprous
chloride (see Reagents) each of which absorbs its own volume of the gas. The
compound formed-possibly 2CuClCOH 20-is very unstable, as carbonic oxide
may be freed from the solution .by boiling, or placing it in vacuo: even if the
solution containing it be simply shaken with an inert gas or one containing little
carbonic oxide it is given off. Where mercury is the confining liquid, amm(miacal cuprous chloride should be used, as the acid chloride causes a troublesome
flouring of the mercury. Cuprous chloride does ab~orb appreciable quantities
of the higher paraffine hydrocarbons.
.
The compound of carbon monoxide and cuprous chloride, being so unstable,
makes a very unsatisfactory absorbent for the gas. Two methods are in use
to remedy the difficulty, by the use of cuprous sulfate and combustion with
copper oxide. The compound formed with cuprous sulfate, CU 2S04CO, is
stable. It does not set free the gas, but acts more siOllTU' than the chloride,
particularly after it has become about half saturated. Whereas the chloride
absorbs about its own volume, the sulfate absorbs 18 tim~1;! as much. It
absorbs the unsaturated, but not the saturated hydrocarbons, nor hydrogen,
nitrogen, and only traces of oxygen.
By using first a cuprous chloride pipette until only a few cubic millimeters
are absorbed, and finishing to complete absorptiqn with a sulfate pipette, the'
analysis can be -rapidly and accurately executed. '
Report of the National Board of Fire Underwriters, No. 559.
.
Gases in Manholes." U. S. Bureau Mines R. 1. 3109 (1931); also R. 1. 2710
(1925) and Rept. Investig. 3192, C. A., 26, 6140 (1932); also White, C. A., 8,2914 (1914).
106 Hoffman, "Considerations on the Fatal Effect.a._:of the Vapour from Burning
Charcoal," 1716.
. .-/
104

105"

GAS ANALYSIS

2373

These absorption methods being slow, Burrell determines carbon monoxide

and hydrogen together, by burning with coppJr oxide in a U-tube heated to
300o 12: the CO 2 formed. equals the CO present.
These methods are those ordinarily used for the estimation of considerable
quantities of carbon monoxide: for the determination of small quantities using
iodine pentoxide, see Analysis of Air.
Katz and Bloomfield 107 find that as little as 0.07% CO oan be determined by
" Hoolamite," a mixture of iodine pentoxide and fuming sulfuric acid on granular pumice stone. The mixture is kept in sealed glass tubes the ends of which
are broken and connected up with a tube of activated charcoal through which
the air to be tested is drawn by means of It measured rubber bulb. The amount
of carbon monoxide is shown by comparing the color produced with a set of
standards.
Group VI-Gas Soluble in Ferrous Sulfate Solution, 1 : 2.
Nitric oxide, NO, was discovered by Van Helmont and fully investigated by
Priestley in 1772. It is a colorless gas, slightly heavier than air, has been
liquefied-boiling point -153.6, and very slightly soluble in water-1 ml.
dissolves 0.05 mI., forming nitrous and nitric acids-alcohol about five times as
much. It is the most stable of the oxides of nitrogen, decomposing slightly at
900; phosphorus, actively burning, continues to burn in it. Brought into the
air, it takes on oxygen, becoming N0 2 This makes it an important catalyst
in the chamber process for making sulfuric acid.
It is prepared by treating copper, silver or mercury with nitric acid, sp.gr.
1.2: the gas obtained is not pure but contains nitrous oxide. The pure gas may
be prepared by heating its solution in ferrous sulfate.
Qualitative Detection.-Absorb in water and test for nitrites. '
Quantitative Determination.-Absorh in ferrous sulfate 1: 2: slightly
acidified with sulfuric acid: or it can be passed into acidulated permanganate
like nitrogen tetroxide. The burning with hydrogen in a platinum capillary is
not satisfactory.
.
Group VII-Unabsorbable Gases. Hydrogen, Methane, Ethylene, Acetylene,
Nitrous Oxide, Nitrogen, Carbon Oxysulfide, and the Noble Gases.
Hydrogen was prepared in the sixteenth century by Paracelsus by the usual
method of the displacement from acids by metals. Cavendish about 200 years
later, in 1766, investigated it thoroughly and gave it the name of" inflammable
air." It occurs in' volcanic emanations and in the sun. It is produced in the
decomposition of organic bodies particularly when oxygen is absent. Hydrogen is the lightest substance, having a sp.gr. of 0.07, the liquid at its boiling point
-252.5 having a like density. It is one of the least soluble gases: 100 ml. of
water at 20 dissolve about 2 ml., alcohol dissolves about 3 times as much.
It is absorbed by some of the metals, hot iron taking up 19 volumes, powdered
platinum 50, and palladium black about 900 volumes: cocoanut charcoal
absorbs 4.4 volumes. Hydrogen is sometimes said to be spontaneously inflammable, many caes being known of the gas igniting when issuing from cylinders;
this has been shown by Nusselt 108 to be due to the dust it contains: the dust,
107 Ind. Eng. Chern., 14, 304 (1922); also 17,555 (1925).
108 Zeit. Ver. deut. lng., 66,203 (1922); C. A., 16, 1640 (1922).

2374

GAS ANALYSIS

iron rust, etc., issues with the hydrogen with such velocity through the jet as
to cause electric sparks, which ignite the hydrogen as lit comes in contact with
the air. Pails of water and wet rags are used to extinguish the flames from the
joints in the balloon filling pipes in the Zeppelin hanpar on Lake Constance.
The explosive limits with air are from 4.1 to 75%; if tl1e oxygen content be 5%
it is also explosive.
I
Qualitative Detection.-In the absence of olefines (C 2H 4, etc.) and carbQn
monoxide, by the evolution of hydrochloric acid when hydrogen is passed ove~
cold solid palladium chloride. lo9
Quantitative Determination.-If there be no other combustible gases
present, several methods are available:
(a) Combustion with a hot platinum spiral in the apparatus of Dennis and
Hopkins.no
.
(b) Combustion with copper oxide at a temperature of 250-300 according to

Jaeger 111 or in any other similar apparatus.

(c) Explosion with oxygen, the converse of the determination of oxygen in air.
I

Where combustible gases are present the procedure of Campbell and Hart 112
using a 1 % solution of palladous chloride at 900: when fresh, 125 ml. will absorb
25-50 ml. of hydrogen in 90 minutes. Car,bon monoxide and dioxide, unsaturated hydrocarbons, and oxygen must be absent. The palladium precipitated
can be recovered and a fresh solution ,made therefrom. Platinum black or
palladium sponge or asbestos treated with either has not, in the exp~rience of
the writer, given satisfactory results.
Hydrogen has of late assumed great industrial importance, in the synthesis
of ammonia, the hydrogenation of oils, changing them to fats of any degree of
solidity, the hydrogenation or Berginization of coal and petroleum, and 'prepara, tion of hydrocarbons resembling petroleum products from it, synthetic motor
fuels and for balloons: in this latter 'connection a thousand cubic feet of pure
hydrogen will lift 75 pounds, of the commercial gas, about 71 pounds. It is
also used for cooling synchronous condensers, increa~ing their capacity 25%.113
It is made commercially by the electrolysis of water or brine, by passing over
hot iron, the Lane process, by passing water gas over lime at 500. Hydrogen
is also made by passing natural gas and steam over a catalyst at 870: the CO
found is removed by another catalyst.1l4 In this connection it should be noted
that attemp~s are occasi9nally made to decompose steam.. ~y the" waste heat
of the boiler furnace" and get ,hydrogen as a fuel, and oxygEm~as,a by-product.
Cast iron retorts are set behind the bridge-wall, steam conductec! through them
and this passes under the grate and up through the coal. There are several
difficulties in the way here: first, at the temperature of 1600 0 no noticeable
(not to say cansiderable) decomposition of steam takes place whereas the
melting point of the cast iron retorts is 1500, or 100 lower; s~cond, when
Phillips, Am. Chern. J., 16, 259 (1894).
J. Am. Chern. Soc., 21, 398 (1899).
ill J. Gasbeleuchtung, 41, 764.
112 Am. Chern. J., 18,294 (1896).
109
110

113 Power, 68, 995.

Science, 72, X (1930). Power, ~,_902 (1930).
80, 362 (1936).
_,114 Byrne, Gohr and Haslam, Ind. Eng. Chern., 24, 1129 (1932).

Easley, Power,

GAS ANALYSIS

2375

the mixture of hydrogen and oxygen (supposing it to have been formed) is

conducted from the retort, if the temperature drops below 1600 they recombine making steam; third, if it were po13sible there would be no gain, the heat
evolved by the combustion of the gases (if formed) having been previously
abstracted from the system to produce any decomposition. A puzzling feature
was the fact that when the apparatus was first operated hydrogen was evolved.
This was of course due to the decomposition of the steam by the red hot
iron 3Fe+4H 20=4H 2 +Fe a04, and was overlooked.
Another alleged source of gain is that obtained in the so-called" water gas
process" of burning coal: Steam is used to produce draft and so cut down smoke,
and of course it is decomposed yielding, with the coal, hydrogen and carbon
monoxide.
Attention is called to the fact tJ:lat these are excellent fuels ready-formed
" made by waste heat," and the fact lost sight of that the reaction by which they
were produced abstracted heat, or cooled the furnace down, and that actually
there is no gain in the process. The thermal efficiency of hydrogen engines is
ve:t:y high (45-50%); they use about 21.2 cu. ft. per horsepower hour.ll5
Methane, CH 4, noted by Pliny, forms the fire damp of the coal mines and
some part of the fuel for the holy fires of the fire worshippers: when marshy
pools are stirred, the gas evolved is largely methane, and it is the chief constituent of natural gas. It too is formed by the decomposition of organic
matter by bacteria, septic tank gas being rich in methane. Volta in 1776
observed that it required twice as much oxygen for its combustion as did
hydrogen, Berthollet showed it contained hydrogen and carbon, but Henry
in 1805 determined its composition.
It has been made from its elements; a common method of preparation is by
heating an acetate with an hydroxide" CH aCOONa+HONa=CH 4+Na 2 CO a ;
this does not always furnish pure methane. Another method is by the action
of a zinc-copper couple on methyl iodide and alcohol;116 care must be taken to
wash and chill out the iodide vapors. Another method of preparation is by
the action of water upon aluminum carbide, Al 4C a +12H 2 0 =4Al(OH)a+3CH 4.
Methane is a colorless, odorless gas, about half as heavy as air (sp.gr. 0.55),
very slightly soluble in water, 1 ml. dissolving 0.035 ml., and being about 10
times as soluble in alcohol. It burns with a slightly luminous flame of about
5-candlepower (burned at the rate of 5 cubic feet per hour). The explosive
limits with air are from 4.9-15.4%. Its boiling point is -164. Cylinders of
the compressed gas are under about 1800 pounds pressure. It is a very stable
compound both as regards chemical reagents and heat.
Quantitative Determination.-Methane may be determined by 'explosion
with oxygen or air, as is hydrogen, the procedure being given on p. 2356.
Dennis 117 states that while the determination of hydrogen by the Jaeger copper
oxide method is very satisfactory, the determination of methane in the same
tube at a higher temperature cannot be recommended on aC90unt of the time
necessary for the oxidation of the methane. He recommends therefore combustion in the Dennis combustion pipette or grisoumeter.
Ethylene or Ethene, C 2 H 4 , was discovered by Becher and studied by four
Dutch chemists in 1795 who made it by dehydrating alcohol with sulfuric acid:
115 Power, 77, 90 (1933).
116 Gladstone and Tribe, J. Chern. Soc., 45, 154 (1884).
117 "Gas Analysis," 1929 edn., p. 163.

GAS ANALYSIS

2376

it gave an oily liquid (ethylene chloride) with chlorine and hence received the
mime olefiant (oil making) gas. Henry showed how it differed from methane
and gave it the name of " the heavy hydrocarbon" which still persists, being
applied to that por~ion of coal gas absorbecYby brorriine 'or fuming sulfuric acid
and also known as " illuminants."
\
Ethylene is a colorless gas, of peculiar ethereal odor, slightly lighter tha.
air and soluble in 4 times its volume of water at 20: alcohol dissolves about 10
times as much. Liquid ethylene is found in commerce in cylinders under a
pressure of 620 lbs. (42 atmospheres), and the boiling point is -104.3. It m
readily inflammable, giving a flame of about 70 candle power (burned at the
rate of 5 cubic feet per hour). Mixed with 3 times its volume of oxygen)' it is
powerfully explosive: the explosive limits with air are from 3.2 to 34%.
The principal source of ethylene at pre~ent is the gases from the stills used
in the cracking of petroleum. During the great war it was made by passing
alcohol over clay heated to 500-600. Ethylene is used as an anesthetic,
mixed with carbon dioxide as a stimulant; oxygen was formerly used for this
purpose instead of carbon dioxide, but gave rise to serious explosions in corpore.ll s It is used in a dilution of 1 part per million to color or " age" fruit;
one cubic foot per day introduced into a 5000 cubic foot room produces a
beautiful golden color in oranges-without, in some cases however, imparting
the flavor that goes with it.1l9
Quantitative Determination.-Ethylene is determined by absorption with
fuming sulfuric acid containing about 25% of the trioxide: Worst all 120 states
this may appreciably attack the saturated hydrocarbons if the time of absorption be more than 15 minutes. Inasmuch as the vapor tension of sulfuric
anhydride is higher than that of water, the gas should be passed into the potash
pipette to remove this vapor as well as any sulfur dioxide which may be formed.
Bromine water may equally well be employed, followed by treatment with
potash. In case of the determination of " illuminants " in ,coal or water gas,
bromine water is almost universally used, as it is more rapid: benzol and its
homologues while theoretically unattacked by bromine water are mechanically
dissolved.
I
Acetylene (Ethine), C 2H 2 , was discovered and analyzed by Edmund Davy in
1836, who prepared it from the black residue obtained in making potassium
.from cream tartar and charcoal. Berthelot made a thorough investigation of
it in 1859. It is formed by the incomplete ,combustion of many gases and
volatile liquids: it was the first hydrocarbon to be syntheslzedJr_~ its elements,
by passing an electric arc between carbon poles in an atmosphere of hydrogen;
it may also be made by treating ethylene bromide with alcoholic potash, or as
now commercially prOduced by the action of water upon calcium carbide: 121
118 "Boston Herald," Jan. 24, 1931.
119

C. A., 23, 218 (1929).

J. Am. Chern. Soc., 21, 245 (1899).

I
121 The early preparation of calcium carbide is interesting: Willson at Spray, N. C.,
was endeavoring to make metallic calcium by heating a mixture of tar and slaked lime
in the electric furnace. Instead of getting a bright beautiful metal, a clinker was obtained, which so disgusted him that he threw it into the rive1'. This treatment was
promptly followed by the boiling up of the water, bubbles of gas ,being evolved, which
burned when ignited, with a brilliantly luminous flame. :..Borne of the clinker was sent
to E. N. Dickerson, a patent lawyer of New York, witlr'the-request to ascertain what it
and the evolved gas were. The answer was returned that the clinker was calcium. carbide
and the gas acetylene. See also Wyatt, Eng. and Min. J., 58, 556 (1894).
120

GAS ANALYSIS

2377

generators in which the carbide is dropped into the water are the preferable form.
The yield is about 5 cubic feet to.the pound or about 11,000 cubic feet per ton.
Acetyle'ne when pure has a pleasant ethereal odor; as ordinarily generated,
the odor is unpleasant due to compounds of sulfur and phosphorus: air containing 9% can be breathed without ill effect but this mixture is very explosive;
the poisonous properties which it is said to possess are due to the sulfur, phosphorus and arsenic compounds contained in it.122 It is a trifle lighter than air
and dissolves in an equal volume of water at 20; alcohol dissolves 6 times as
much and it is quite insoluble in saturated brine. It is very soluble in acetone
(1 m!. at 15 dissolves 23 m!. and expands about 5%) and particularly under
pressure at 12 atmospheres (180 Ibs.) it holds 300 volumes. It burns in the air
with an intensely luminous flame of 240-250 candle power, at 5 cubic feet pEr
hour: the light approaches daylight the most closely of any gas light.
At a pressure of 903 Ibs. (63 atmospheres) and 10 acetylene condenses to a
liquid which solidifies at - 81 0, and which it was hoped could be shipped in
cylinders, and thus solve the problem' of the lighting of isolated houses, but
after serious explosions in Paris and Hartford, this was abandoned: the cause
of'the explosions could not be determined for lack of witnesses.l23 It may have
been due to the fact that acetylene being an endothermic compound (47,700
calories being absorbed when it is formed), this same amount of heat becomes
free when it is decomposed and expands the hydrogen left by the decomposition
(which is equal in volume to the acetylene originally present) producing a
pressure, which, coupled with that already existing in the cylinder, is sufficient
to burst it; a shock or the detonation of the acetylide of copper would be sufficient to have initiated the explosion.
The explosive limits of acetylene and air vary, according to different observers, from 2.5 to 73,124 or 1.5 to 80.5%,125 with the maximum effect at about
7.8. The ignition temperature is 480, the temperature of the oxyacetylene
flame is said to be the same as the electric arc (over 3000) and is consequently
extensively employed in welding and cutting metals even under water!
Qualitative Test.-The reddish-brown coloration produced by contact with
ammoniacal cuprous chloride ..
Quantitative Determination.-By absorption in bromine water, or fuming
sulfuric acid, and subsequent treatment with caustic potash: when passed
through ammoniacal cuprous chloride solution a reddish-brown precipitate of
acetylide of copper is precipitated. This is explosive if allowed to dry. Consequently it must always be kept moist. The precipitate is filtered off, washed
free of copper, and dissolved in hydrochloric acid: in case traces of the copper
acetylide remain undissolved, the filter is dried, burned, the ash treated with a
few drops of nitric acid to dissolve' the cODper oxide and added to the acid
solution above; the copper is here determined electrolytically or in any conChern. Ind., 20, 53.
A cylinder of compressed acetylene at about 15 atmospheres (without acetone)
was not detonated by the fall of a 50D-pound pile driver weight 20 feet, nor by a rifle
bullet shot through it: when put upon a wood fire-as in a burning car wreck-it detonated splendidly, and fairly promptly. When liquefied, it is explosive at 560 lb.
(38 atmospheres) and exerts, when exploded, a pressure of 40 tons per sq. in. (5500
atmospheres). Wolff, Z. angew. Chern., 11, 919 (1898).
124 Burrell and Oberfell, Bur. Mines Tech. Paper, 112.
12S International Critical Tables.
122

123

2378

GAS ANALYSIS

venient way. The copper in small quantities may be determined colorij

metrically.126
Acetylene dissolved in acetone, " Prestol~,,, finds, use for illuminating purposes (trucks), cutting and welding metals.127 The cylinders of mild steel ab01.l.t
1/8 inch thick are filled with discs of compressed a8b~stos, then with acetone,
into which the gas is pumped under a pressure of' a.bout 225 lbs. (15 atmos
pheres). Great care must be taken in handling these cylinders, not to droJl
them: the writer has read of a case in London and also known of a ca8e ill
Boston in which fatalities occurred by dropping a rsmall cylinder about a fool
upon a concrete floor. In the latter case a flaw in the wall of the cylinder left it
but half the usual thickness.
'
The fact that the action of moisture upon calcium carbide produces acetylene may be utilized to determine water in organic compounds or in other
cases.
Acetylene can be purified from arsine or phosphine by passing through
bleaching powder solution (Lunge) or chromic acid (IDlmann)_I28
Nitrous Oxide, N 20, was prepared by Priestley in 1772 by thereduction of
nitric oxide, perhaps with moist iron filings. It is at present made by heating
ammonium nitrate, NH 4 NO a = N 20+2H 20. It is noteworthy that it is
evolved when zinc is treated with very dilute nitric acid.
It is a colorless gas 50% heavier than air, with a pleasant smell and taste:
when inhaled, it produces a temporary anesthesia with hysterical effects, giving
rise to the name of " laughing gas": at/20 water dissolves 0.7 its volume, alcohol about 4 times as much; it boils at - 89.8. It is a supporter of combustion
but not of respiration. Liquid nitrous oxide has a pressure of 820 lbs. (55.4 ats.)
at 20 and is still sold by the gallon, cylinders of 100, 150 and 450 gallons (60
cu. ft.) being available (100 gallons weigh lIb. 9 oz.). It is used in minor surgical operations and as an anesthetic preliminary to ether or chloroform.
Determination.-By explosion with hydrogen, if the volume of the hydrogen
be 2 t.o 3 times t.he volume of the nitrous oxide, the combustion is aided by
mixing detonating gas (H2+0) so that there are between 26 and 64 volumes of
combustible gas to every 100 volumes of incombus'tible gases, which are the
Bunsen ratios.
It should be tested for chlorine and nitric oxide. 129 It may also have a bad
odor of rancid oil from the old petroleum oil used in lubrication of the compressor cylinders, which has passed through the filters.
Nitrogen was recognized as a definite body in 1772 by.R1!therford of Edinburgh, who called it " mephitic air" ; Priestley in the same year found that the
residue after carbon had been burned in air, and the gas thus generated absorbed, would support neither combustion nor life, and amounted to 4/5 of the
air originally used. It is prepared by burning the oxygen out of air with
phosphorus, or hot copper, and by the fractionation of liquid air. This nitrogen contains about 1 % of" the noble gases"; the p~re gas is obtain'ed by heating
ammonium nitrite solution, NH 4 N0 2= N 2+2H 20. The ammonium nitrite

126 Weaver, Sci. Paper Bur. Stds, No. 267 (1916).

See also Mathy, Petro World, 32, 59 (1935). Slottman, 9il and Gas J., 34, 52
(1936), says it is still the best and cheapest, as it useEl,the least oxygen.
128 Berichte, 12, 32 (1879).
.J129 C. A., 18, 2789 (1924).
127

GAS ANALYSIS

2379

. solution is best prepared by mixing strong solutions of ammonium chloride

and sodium nitrite; the oxidation of 'ammonia or reduction of nitric oxide also
yields the pure gas.

It is a colorless, tasteless and odorless gas, a trifle lighter than air: it is very
slightly soluble in water, 1 ml. dissolving at 20 0.015 mI., and alcohol about 10
times as much: it boils at -195.5 (13 lower than oxygen). It is shipped in
the usual 220 cu. ft. cylinder under a pressure of 1800 pounds per square inch.
It is a remarkably inert gas, combining directly with but few elements: lithium,
magnesium, barium and calcium form the nitrides, which wnen moistened yield
ammonia. Strutt 130 finds that when nitrogen containing a trace of oxygen
or other gases is subjected to an electric discharge at low pressure, this modified.
nitrogen acts upon mercury, phosphorus, acetylene and other organic compounds.
Determination.-NitJ:ogen is energetically absorbed by metallic calcium
heated to redness. Fischer 131 has shown that both nitrogen and oxygen may
be absorbed by calcium carbide at a temperature of about 800. Other gases
must of course be absent. Nitrogen can be determined in gases by combustion
with"hot copper oxide after the procedure for the determination of nitrogen in
organic compounds. 132 The tube is swept clear of air by carbon dioxide, the
gases burned, th.e tube again swept out, the carbon dioxide absorbed, and the
residual nitrogen measured.
.
Nitrogen is used industrially in making synthetic ammonia, which may
be oxidized to nitric acid. (This discovery prolonged the great war for years.)
Bradley and Lovejoy made nitric acid by passing an arcing current through the
air using copper terminals: it was not found to be economical. Were not heat
absorbed in the combination, which must be supplied from an external source~
the atmosphere would be set on fire and wholly consumed.
Carbon oxysulfide, COS, Carbonyl Sulfide, was discovered by Than 133 in
1867. It is probably contained in sulfur waters and illuminating gas. It is
made by the action of about 50% sulfuric acid upon poiassium or ammonium
sulfocyanate, KCNS+H20+2H2S04=COS+KHS04+NH4HS04. It is a
colorless gas, odorless, when pure, about twice as heavy as air, readily inflammable and poisonous: 0.8 ml. dissolves in 1 ml. of water at 13.5 in which it
slowly decomposes giving H 2S and CO 2, The gas could perhaps be determined
by alcoholic potash: 1 part potassium hydroxide in 2 parts water. and diluted
with an equal volume of alcohol: 1 mI. absorbs 72 ml. COS.
Gases of the Helium Group-the noble gases: Helium, Neon, Argon,
Krypton, Xenon (and Niton): sometimes also called" the noble gases," because, like the noble metals, they are unaffected by chemical agents.
Helium, Re.-Lockyer in 1868 observed in the spectrum of the sun a line
very close to! the double sodium line, and as it corresponded to nothing terrestrial, the element proaucing it received the name helium from the Greek
word elios, the sun. It has since been found in the air, In mineral waters, rare
minerals, and by Cady and McFarlan9. 134 of Kansas, in the natural gas of the
,

Proc. Royal Society~ 85 (1911)-93 (1917).

Ber., 41, 2017\(1908).
132 New, J. Soc. Chern. Ind., 11, 415 (1892).
133 Annalen Supp., 5, 236.
Hempel, F. Ang. Chern., 14, 865.
134 J. Am. Chern. Soc., 29/ 1523 (1907).

130
131

2380

GAS ANALYSIS

Mid-continent and Gulf regions and al~o Canada, to the extent of 1.84%.
Recently a gas containing about 7% has been discqvered. Before the war it
cost $1300 per cubic foot, but at present about $7 per thousand cu. ft.: 135 it is
obtained from the natural gas by compression and cooling. It is present in air
to the extent of 1 part in 250,000, and has been o~tained from liquid air by
fractionation. It is also evolved from an aqueous solution of radium bromide.
It is twice as heavy as hydrogen, much less soluble in water, 1 mI. at 20
dissolving 0.0138 mI.-the most insoluble gas, and boils at about -268.5.
It is detected by the spectrum, the double Da line being very characteristic,
along with seven others. It is used as balloon filling, for the yellow vacuum'
signs and for lamp filling. Mixed with 1/5 oxygen it is used on account of its
slight solubility in the blood to enable divers to come back to normal pressure.
Neon, Ne (Neon the new gas), occurs in the air to the extent of 1-2 parts per
100,000, and also in the gases from certain springs. It is prepared iby the distillation of the crude argon from liquid air. It is 10 times as heavy as hydrogen,
boils at -245.9 and is distinguished by the brilliant red spectrum. This finds
extensive industrial use for signs but the gas leaks out eventually through the
seals of the wires in the tubes. The blue signs are a mixture of mercury vapor
and neon, the green are the same as the blue using an amber tube. "
Argon, A (Argon idle).-Lord Rayleigh, observed that a liter of atmospheric
nitrogen weighed 7 'milligrams more than chemical nitrogen, showing heavy
gases in the air: Cavendish in 1785 had performed a similar experiment with a
like result. Rayleigh and Ramsay repeated the experiment, absorbing oxygen
with not copper, nitrogen with hot magnesium, introducing oxygen, sparking
the residue, absorbing the nitrous fumes in alkali, and removing oxygen, the
gases being circulated for about a week. About 1 % of a gas was obtained,
having a density 20 times that of hydrogen and an orange spectrum: it forms no
compounds, whence its name. The density of the pure gas is 19.0 and the boil. ing point -186.1.
Krypton, Kr (Krypton hidden), occurs in the air to the extent of 1 part in 20
million: its boiling point is -151.7, density 41.5, a,nd gives a yellowish-green
spectrum.
Xenon, X (Xenon stranger), is the least abundant gas, occurring in the air to
'the extent of 1 part in 170 million: its boiling point is -106.9, density 65.3 and
gives a blue-green spectrum. These last four gases are obtained by the careful
fractionation of liquid air.
I
Niton, Nt, is the emanation produced when an atom-of helium is expelled
from an atom of radium.
APPLICATIONS O:F GAS ANALYSIS 'ANI! INTERPRETATION
OF RESULTS
It is only within comparatively recent times 'that; Gas Analysis has assumed
great importance. The reasons are that the substances with ,vhich it deals are
so intangible, the apparatus is complicated and fragile, and until lately, competition has not compelled manufacturers to seek .every possible source of loss. '
136

Ind. Eng. Chern., t4, 1435 (1932).

2381

GAS ANALYSIS
Some of its applications are to:
I.
II.
III.
IV.
V.

Chimney and flue gases;

Producer and fuel gases;
City gas;
Sulfuric acid gases;
Mine gases;
I.

VI.
VII.
VIII.
IX.

Electrolytic gases;
Acetylene;
Atmospheric air;
Natural gas.

CHIMNEY AND FLUE GASES

Here the object is to keep the carbonic acid (C0 2 ) as high as possible, and to
avoid the formation of carbon monoxide: in large plants every additional,
per cent of carbonic acid means the saving of tons of coal. Savings of 20 to 33%
by the use of gas analysis alone have frequently come to the writer's notice.
A satisfactory procedure is to post in the fire-room the percentage of carbonic
a~id obtained by each stoker, and stimulate a rivalry among the men-a bonus
in the pay envelope is also effective. The determinations to be made are:
Analysis of Chimney Gases.-Determination of carbon dioxide, oxygen,
carbon monoxide, nitrogen, and in$ome cases hydrocarbons. For this purpose
the Orsat apparatus is widely employed: the hydrocarbons may be determined
by the Burrell apparatus. These need not be sought unless the carbon monoxide is more than 1 %.
Usually a few determinations of carbonic acid will suffice, but for regular
work the installation of some form of registering carbonic acid indicator is advisable.
Carbonic Acid Indicators. 13 6.-lhese usually depend upon the principle of
collecting 100 ml. of the gas, causing it to pass through a suitable absorber and
collecting the residue in a bell which floats to a greater or less height according
to the residual volume. The fluctuations of this bell are recorded after the usual
manner of self-registering barometers or thermometers: the usual time for this
analysis and record is five minutes.
By modifying this apparatus slightly, it can be applied to the determination
of any absorbable gas as, for example, sulfurous acid or chlorine. It has been
adapted to carbon monoxide absorption, but it is not usual or easy.
Haber 137 employs the refractive index of gases to determine the amount of
carbonic acid in chimney gas; it gives results within half of 1%; 138 it has also
been applied to other gaseous mixtures. The instrument is called the Interferometer or Gas Refractometer and is made by Zeiss of Jena.139
Another method for the determination of carbonic acid makes use of the
heat cond~ctivity of the'different gases; carbonic acid has but about 60% of
the heat conductivity of air. Consequently a wire will be hotter in it than in
air, all other conditions being the same. This increase in temperature' of the
wire increases its electrical resistance, from which the per cent of carbonic
:lei!! call hI! determined. The descriptioll woujd take UR beyond the lilllit,," of
136 ::lee Bureau of Minell Bulletin No. 91, "Inlltruments for Recording Carbon Dio:dde
in Flue Gases." Also Power, 51,784 and 1058.
137 Z. angew. Chem., 19, 1418 (1906); ibid., 23,.1393 (1910).
138 Mohr, ibid., 25, 1313 (1912).
139 Bureau of Mines Technical Paper 185, "Use of the Interferometer in Gas Analysis,"
1918.

2382

GAS

ANA~YSIS

this volume; according to Palmer and Weaver 140 it is by far the best method
for the determination, being accurate, sensit~'ve, easily operated and of moderate
cost.
The " Ranarex" recorder,141 an English instrument, consists of two fans
rotating each in its own chamber by the same electric motor: these fans cause
discs opposite them to rotate, one in air, and the other in the chimney gas.
Different torques are produced, depending upon the density of the gas in which
the discs rotate: this difference is transmitted by levers to a pointer which
indicates the per cent of carbonic acid.
The Determination of Temperature.-This is done by inserting a thermometer, mounted in a metal tube, on the chimney side of the gas sampling
tube. Care should be taken to protect it from radiation to and from the walls
of the flue. They resemble those used for determining steam temperatures or
for "running" varnish. It should register to 360 and, under certain circumstances, one showing 550 may be desirable. ,A chemical thermometer
with long stem may also be employed; it should never be inserted naked into
the flue-as a sudden hot blast may break it-but always in a tube of cottonseed oil or sea sand.142 These thermometers should be tested for accuracy by
comparison with a standard, in a carefully stirred oil bath. The standard
should be kept exclusively for the purpose and be allowed to stay in the bath
until cool. Sudden cooling of a thermometer changes the zero point. The
standard can be certified by the makers or the United States Bureau of
Standards.
Electric pyrometers are also of course available for these measurements.
An error of five degrees (5) in the reading of the thermometer affects the final
result by about 20 calories.
In case none of these appliances be at hand, the maximum 'temperature can
be determined by utilizing the melting-points of certain pure salts or metals;
as tin 232, bismuth 270, cadmium 302, lead 327, zinc 419, cadmium chloride
541, antimony 630, etc. These can be suspended in the chimney in srn"all
covered cast-iron boxes.
"
Composition of the Coal.-This is determined by the usual methods of
organic combustion and is required only for very accurate work.
Calculation:
'
a. Heat passing up chimney;
b. POl,lnds of air per pound of coal.
(a) Heat Passin'g up Chimney.-The accurate- calculation resolves it./Jelf
into finding what 'volume of gas of the compositi.gn determined by analysis
would be produced by a kilo of the coal used, the analysis of which is known.
The temperature of the escaping chimney gases being also known, and their
specific heat, the quantity of. heat they carry off can be calculated: this divided
by the calorific power of the coal gives' the per cent of heat lost in the chimney
gases. The calculation is rather long and will be found in detail inthe author's
book. l43
Bur. Stds. Tech. Paper 249; also Ind. Eng. Cp.em., 12, 359 and 894 (1919).
Grant, Industrial Chemist, 2, 435 (1926).
.
"
142 With rouruled grains, not river sand~as it would make scratches.
143 "Gas and F~el Analysis for Engineers," Wiley.
140
141

GAS ANALYSIS
The formula of Shields:

2383

144

Per cent carbon in coal

200+per cent CO 2
Per cent h eat Io s t = .
X-----'-,"-----__::_Heatmg value of coal
Per cent CO 2 +per cent CO
Xrise in temperature in

C. XO.2864,

gives results usually 0.5% low, as no cognizance has been taken of the water
vapor.
Another formula 145 in which only the carbon dioxide and its temperature
enters was proposed by Bunte and gives close results.
For every per cent of carbonic acid present 43.43 calories per cubic meter of
flue gases have been developed = W; C = specific heat of the flue gases per cubic
meter; then WIC represents the initial temperature (which is never attained)
the ratio of which to the actual exit temperature of the flue gases shows the
heat lost. If T = this initial temperature and t the rise of temperature of the
flue gases, then tiT represents the heat lost in the chimney gases.
The following table gives the data for the calculation for both pure carbon
and coal of average value.
If there were 11:5% carbonic acid, the initial temperature T would be
1762; the rise of temperature in the chimney gases is 250, the loss is 250/1762
or 14.2%. The accurate calculation gives 14.1.
Initial Temperature, WIC,
Per Cent of CO, in
Chimney Gas

1
2
3
4
5
6
7
8
9
10

11
12
13
14
15
16

C.

Specific Heat of
Chimney Gas

0.308
0.310
0.311
0.312
0.313
0.314
0.315
0.316
0.318
0.319
0.320
0.322
0.323
0.324
0.324
0.325

For Carbon = T

For Coal =T

Diff. for 0.1% CO,

141
280
419
557
694
830
962
1096
1229
1360
1490
1620
1750
1880
2005
2130

167
331
493
652
808
961
1112
1261
1407
1550
1692
1830
1968
2102
2237
2366

16
16
16
15
15
15
15
15
14
14
14
14
13
13
13

Finally, for very rapid work, Bunte's Fig. 368 may be used. The results
are within 2% for about 12% of carbonic acid. It is used by noting the point
where the diagonal line representing CO 2 cuts the ordinate of temperature-the abscissa corresponding to this point represents the per cent loss.
144

145

Power, 30, 1121 (1909).

J. f. Gasbeleuchtung, 43, 637 (1900).

GAS ANALYSIS

2384

The following table shows roughly the excess' of air, and per cent of heat lost
in the chimney gases, their temperature being 51~0 F.
I

4
3
5
6
7
8
Per cent CO 2 2
91 10 11 12 13 14 15
Vol. air more
than theory
=1.0 ....... 9.5 6.3 4.7 3.8 3.2 2.7 2.4 2.1 1.9 1.7 1.6 1.5 1.4 1.3
Per cent' loss of
heat", ..... 90 60 45 36 30 26 23 20 18 16 15 14 13 lZ

If the oxygen be from 1.5% to 2% with the temperature of escaping gases

at 400-500 0 F., the fires are too thick; if it be more than 8% they are too thin.
Per Cent Efficiency
70

80

~
-+--+--1300 '<C

50

40

20

30
Per Cent Loss

10

FIG. 368.

(bj Pounds of Air per Pound of Coal.-This can be dete;mined by calculating the ratio of ClJ,rbon to oxYgen in the carbonic acid and carbon monoxide
and oxygen of the chimney gase~, or by the formula of Shields.I 46

'

.
Per cent of carbon in coal
Pounds of atr per pound of coal = 2.31 p ,
CO
'
CO'
er cent
2+per cent
Lockwood 147 has shown how to calculate the excess of' air without the use
of at<?mic weights or other constants, from re.cently de~eloped formulae,
, 146
147

Already eited.
Power, 58, 461 (1923).

-'-

J. Soc. Automotive Eng., 12, 299 (1923).

GAS ANALYSIS

2385

Theoretical Air per Pound of Combustible.

18.3N 2+34.8CO - 69.60 2

O.52qN 2+ 5CO +4C0 2- 20 2

(1)

Substitute per cents of N 2, CO 2, etc., as found by the Orsat apparatus.

This formula as well as those including (9) applies to all fuels, flolid, liquid
and gaseous, except producer and blast furnace gases, where the amounts of
nitrogen and sulfur are negligible, and no free hydrogen or combustible but CO
is in the flue gas.
Actual Air Used per Pound of Combustible .

.l8.3N 2

(2)

Checking Orsat Gas Analysis.-(a) Compare theoretical aIr weight as

obtained by (1) with that calculated from the fuel analysis.
Table A shows the theoretical air required for perfect combustion for different fuels.
Table A
THEClRETICAL AIR REQUIRED FOR PERFECT COMBUSTION

Theoretical Air
per Pound of
Combustible

Vol. Matter

Fuel

11.6
11.82
12.22
12.65
12.85
15.21

Pure carbon ................... .'. . . . .. 0

Gas coke. . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5
Anthracite eoal . . . . . . . . . . . . . . . . . . . . .. 6.2
Semi-bituminous coal. ................ 21.0
Bituminous coal. . . . . . . . . . . . . . . . . . . .. 37.0
Gasoline (CciI,,) .................... .

(b) Equate the expressions for theoretical air from the gas analysis formula
(1) and from the fuel analysis.
(3)

Table B shows the different values for these constants for different fuels.
Table B
CONSTANTS TO CHECK ORSAT ANALYSIS

Fuel

Pure carbon .......................... _.

Gas cake ................................ .
Anthracite eoal .......................... .
Semi-bituminoes coal ..................... .
Bituminous coal. .......................... \
Gasoline ................................ .

1.00
0.98
0.95
0.91
0.88
0.70

0.60
0.614
0.63
0.645
0.652
0.71

20.9
20.35
19.52
18.7
18.4
14.5

GAS ANALYSIS

2386

Lockwood gives five other formulae wliich may b~ useful, as follows:

.
22COz+160z+14(Nz+CO)
Dry gas per pound of combU8t~ble=O.525N2+5CO+4C02-202
. I
. [
69.60 2-34.8CO
]
Percentage of excess atr= 18.3N2+34.8CO~69.60z XIOO.

(4)
(5)

\.
4.77N z -9CO-18(C02+02)
Water vapor per pound of combusttble= O.525N 2+5CO+4C02- 202
72

71

70

69
68
67

B%

6 56

6
63
6
61

.Leoz

9.~

-~

59
5B
57
56
5

CO 2

/
6

2%
. , CO2

51

-{.

5
6

"%~
_'c6~

4
47
46
45
4
43
42
41

14r_., CO2

3
4

~7---(i
32
31

..

:~

--,I

26

--,

3
3

!FI-co'
/--8

19
18
I
16

/'

~..p

_"i'

0,--'-'

CQ

,,11
9
8
7
6

~6 ~-*

16,2.'

Cor-'!

,-'l'-

:-:r-..
...l-_l_B;

02

CO

9
CO

,_~g~

2
r. I

2.

10

FIG. 369.

14
13
1

-0,

~~z ;~ .......

CO

,-1

22

-/

.~

..p

2-3-CO

st.-

CO 2

O'_'6%c '%--

--,

3
3'
33

3.

1%

53

lOj'b
CO2
6

11 12 13 14 15- :r6_17
Per Cent COz in Flue Gas

18

19

20

Brownlie "Combustion," April, 1923.

:rv==

(6)

238 7

GAS ANALYSIS.
.
.
[
6(C0 2+CO)
]
Percentage of carbon 1-n combusttble = 0.525N 2+5CO+4C0 -20 X 100.
2
2
"

(7)

Percentage of hydrogen in combustible

=

[0.525N2-CO-2(C02+02) ] X 100

0.525-N 2 +5CO+4C0 2 -20 2

.
. .
6(C0 2+CO)
Ratto of carbon to hydrogen w combusttble = 0.525N 2_ CO _ 2(C0 + O )
2
2

(8)
(9)

Loss Due to Carbon Monoxide.-For every gram of carbon burned to carbon monoxide there is a loss of 5.66 calories. This is shown graphically in
Fig. 369. Tag-Mono Duplex Loss Chart; vertical distances (ordinates) show
loss, horizontal distances (abscissae) per cent CO 2 in the chimney gas; the
numbers inside the circles the per cent of CO. For example the loss with 14%
CO 2 i~ 13, but if at the same time there be 1% CO, it becomes 20, found by
following the curve to the circle with 1 in it. Hence 12% CO 2 with no CO is
better than 13% CO 2 with 0.5% of CO.

COMBUSTION CHARACTERISTICS

From Gas Analysis

E,,:-Orsa, analysis equals 6.4%CO, and 9.6%0,

Connect CO 2 and O 2 with straight edge, Max. CO 2 equals lL8 %

Nitrogen equals 84", %; then follow down Nittogen line of 84 and
read 76 % of Excess Air. Then pivot straight edge at Max. C02
till it touches turning point and read 1100 B.T.U; heating value
of gas being burned, and 9.7 cubic feet of air required

FIG. 370.

Marx's Chart.

2388

GAS ANALYSIS

For gas, Marx's Chart, Fig. 370, is usefu1:148 Froin the Orsat Analysis it
shows the maximum CO 2, excess air, the B.t.u. in the gJs and the number of
I
cubic feet of air required for combustion.
Smoke.-For the determination of the amount of smoke in the chimney
gases, use may be made of the Ringelmann smoke scale. I This consists 149 of a
series of rectangles %: in. X Y2 in. filled with cross-hatching lines a Isreater or
less distance apart, with which the density of the smoke Can be compared.
Or the Eddy smoke recorder may be employed; this consists of a tube Of standard length through which the smoke gases are drawn. A standard electric
light is fixed at one end of the tube and viewed through the smoke; its density
is measured by the extent to which the light is obscured and by a photoelectric
cell.

II.

PRODUCER AND FUEL GASES.

BLAST-FURNACE GAS

Here the object is the reverse of that in, the chimney gases, to keep the
percentage of carbon monoxide as high as possible and, for gas-engine purposes;
the per cent of hydrogen constant.
The determinations made are the same as in chimney gas-C0 2 , 0, CO/50 N,
and oftentimes hydrogen and hydrocarbons; the quantity of dust is sometimes
important. The heating value is determined as in city gas, p. 2391. The
efficiency of conversion would be found by measuring the number of cubic feet
of gas made per ton of coal gasified; the calorific power of each' (gas and coal)
being known, their quotient represents the efficiency. The heat contained in
the gas due to its sensible heat, found after the manner of calculating the loss
in chimney gases (i.e., volume gas X weight X rise of temperature X specific heat),
is to be added to this for accurate work,
As showing producer gas practice, the following typical analyses are cited:

Anthracite 1~:I

CO ...........
27.0
H! ...........
12.0
1.2
CH ..........
C 2H .........
.. ..
CO 2 . .
2.5
57.0
N 2 .
O 2 . . e 0.3
B.t.u .......... 137

'Bittlminous 152

27.0
12.0
' 2.5
0.4
2.5
55.3
0.3
157

Blue
Water
Gas 153

Lignite.: U)3

45.0
45.0
2.0

22.0
9.G
l.G

0.7

., "

4.0
2.0
0.5
322

6.4
58.9
0.8,
.

..

Pea~ ""

30.6
6.1
5.1 '" ..
5.7
52.5
.,

..

."

Tan

Ui1 ,lil2

14.2
8.7
5.6
.. '\ ....

15.0
56:0
0.4
132

Wood'"

13.3
21.0
2.G
O.S
16.0
46.7
01
140

P. F. Marx, The Refiner and Natural Gas Manufacturer, 10, 121 (1931).
Power, 40, 66; also Carter, Power, 74, 678 (1931).
150 Tech. Paper 106, Bureau of Mines, "Asphyxiation from Blast Furnace Gas."
151 With 38.7% H 20, 3.2% ash.
.
152 From "Gas Producers and Producer Gas Power Plants," R. D. Wood & Co., 1906.
153 Richards, J. W., J. Frank. r_'st., 415 (1900),..J].oofed from V. Ihering, "Gas
Maschinen."
. .
148

149

GAS ANALYSIS

2389

Gas from Different Kinds of Producers

.
CO ...........
H 2 ..
CH ..........
C2 H .........
CO 2 . . . . .

N 2 .......

O2 . . . . . . .
B.t.u .. ........

Down
Draft 156

17.5
11.8
1.1
.04
9.2
60.1
0.2
110

Up

Draft 156

Suction

.18.3
12.9
3.1
0.2
9.8
55.6
.04

26.0
18.5
0.5

....

154

Pressure,
Tay-

Mond

154

101" 1.4, 155

....

8.0
47.0

..

145

"

22-30
15-7
3-1.5

. ...

6-1.5
54-60

....

138

16.0
24.0
2.2

. ...

12.4
45.4
0
146

Bla.st Fur- Siemens 1&7

nace 157

24
2

....
2
12
60

28
2

. ...
2
3
65

... .

. ...

106

122

Determination of Dust.-LiJdell 1s8 recommends the following: lump sugar

is crushed, and that which is retained by a 90-mesh sieve packed in a 2-in.
layer upon copper or brass gauze contained in a glass tube. The sugar is
slightly moistened and the gas sucked through it: it is then dissolved in water
and the dust collected upon a tared Gooch crucible and weighed. Another
procedure and apparatus recommended by the Sargent Steam Meter Co., of
Chicago,159 consists,in sucking the .gas through a diaphragm consisting of a'
weighed filter 4Yz ins. in diameter, .drying and noting the increase in weight.
The U. S. Steel Corporation uses the Brady dust filtE~r consisting of a Soxhlet
thimble 33 X75 mm. in a suitable case, which is inserted in the gas stream. For
details of the sampling and procedure which requires careful work, reference
may be had to an article by Brady and Touzalin. 16o The dust should not
exceed 0.01 grain per cu. ft. at 62 F'. and 30 inchE~s. For the methods of
cleaning gases see Mark's Handbook, also Prausnitz, J. Ind. Eng. Ch., 4, 430
(1932).

III.

CITY GAS 161

The determinations usually made are as follows:

a. Candle power (rarely done);
b. Calorific power;
c. Sulfur;
d. Ammonia;
e. Analysis;
f. Carbon dioxide;
g. Specific gravity;
h. Tar.

From "Gas Producers and Producer Gas Power Plants," R. D. 'Vood & Co., 1906.
With anthracite buckwheat.
.
156 "Resume of Producer Gas Investigations," Bureau oj" Mines Bulletin 13, Fernald
and Smith.
157 Richards, J. W., J. Frank. Inst., 415 (1900), quoted from V. Ihering, "Gar,
Maschinen."
158 Power, 38, 93.
159 Power, 27, 331.
160 J. Ind. Eng. Chern., 3, 662 (1911).
161 Circular 48, Bureau of Standards, "Standard MethOds of Gas Testing," 1916.
154

156

2390

GAS ANALYSIS
Technologic Paper No.36
Vent Tube

Inlet Water
rhermometer

Outret Water
Thermomet ,

To Drain

Wate, Inlet

t 1t

Combustion

_f.~aE~~L

Inlet Water
Valve

Sargent calorimeter (section)/

FIG.

371.

GAS ANALYSIS

2391

(a) Candle Power.-This can be very satisfactorily found using a 60-in.

open-bar photometer and Leeson contrast disc. The gas should be burned from
a burner commercially obtainable which gives the highest candle power; for
gas from 14 to 21 candle power, Sugg's London argand burner, sizes C to F,
should be used; for richer gases, Sugg's table top or the Bray slit burner. For a
standard of comparison, the sperm candle is
convenient, satisfactory, and very' extensively
used: the Elliott kerosene and Hefner amyl
acetate lamps are also employed.
For accurate work the Lummer-Brodhun
disc and electric standards, or the Hefn.er lamp
should be used. For the determination of candle power, reference may be had to Circular
No. 48 of the Bureau of Standards on "Standard Methods of Gas Testing," 1914, or Stone,
" Practical Testing of Gas and Gas Meters."
C:j,rburetted water gas shows from 20-28
candle power, coal gas 14-20, oil gas 45--60, oilair gas 30-35, gasoline 12-17, acetylene 17(}--200.
Inasmuch as 75% of the gas now made is
used for its heating power (mantle burners), the
candle power is infrequently determined.
(b) Calorific Power. (a) Direct Determination.-Thisis often determined by the Junkers
calorimeter, although others in use are the
Sargent, Doherty, and in England the Boys and
Simmance-Abady.
The Sargent calorimeter is shown in Fig.
372 and in section in Fig. 371. It consists of a
combustion chamber surrounded by a water
jacket, this being traversed by a great many
tubes. To prevent loss by radiation this water
jacket is surrounded by a closed annular air
FIG. 372.
space, in which the air cannot, circulate. The
whole apparatus is made of copper as thin as is compatible with strength.
The water enters the apparatus at the top (water inlet), passes by a thermometer and through the inlet valve, circulates around the apparatus, and
passes out by a thermometer to the measuring or weighing vessel: this can be
shunted in and out by electrical connections at the meter.
There is not only a very large surface of thin copper between the gases and
the water, but the two move in opposite directions, during which process all the
heat generated by the flame is transferred to the water, and the ~aste gases
leave the apparatus approximately at 'atmospheric temperature. The gas to be
burned is first passed, to insure constant pressure, through a pressure-regulator
and then through a meter, Fig. 373. The source of heat in relation to the unit
of heat is thus rendered stationary; and in order to make the absorbing quantity
of heat also stationary, two overflows are provided at the calorimeter, making

GAS ANALYSISi

2392

the head of water and overflow constant. The temperatures of the water enter_
ing and leaving the apparatuEl can be read by the thermometers as shown
before; the quantities of .heat and water passed through the apparatus are
constant. As soon as the flame is lighted, the m~rcury ip the out-going thermometer will rise to a certain point and will remain nearly constant.
Manipulation.-The calorimeter is placed as shown in Fig. 373 so that one
operator can simultaneously observe the two thermometers of the entering

FIG. 373.

Hinman-Junkers Calorimeter.

and escaping water, the index of the gas meter; and the mea swing or we~ghing
apparatus.
'
No draft of air must be permitted to strike the exhaust. of the spent gas.
The- water-supply tube is connected with the nipple in the center of the
upper container; the other nipple is provided with Ii wa)lte tube to carry away
the .overflow, which latter must be kept running while the readings are taken.
The nipple, through which the heated-'Witer leaves the calorimeter, is

GAS ANALYSIS

2393

connected by a rubber tube with the graduate or a weighed can. The conpensed water of the apparatus is caught in a small graduate .
. The thermometers being held in' position by rubber stoppers and the water
turned until it discharges, no water .must issue from the drain tube for condensate, Fig. 371, as this would indicate a leak in the calorimeter.
The inlet water valve is now set to allow about two liters of water to pass in
a minute and a half, and regulate the gas to flow at the rate of 4 to 7 cu. ft. per
hour as may be found to give the highest result with the gas tested. Adjust
the air so that the..f1ame shows a slight luminous tip, insert the burner as far as
the bracket permits and 09serve the condition of the flame in the mirror. The
excess of air should be as'low as possible, and is controlled by the damper in
the exhaust port: start with t:he damper closed, then open slightly, observing
the outlet thermometer wherrthis reads at a maximum, i.e., when the greatest
heat is given up to the water the damper is in about the correct position.
The flow of water should be regulated so that the temperature of the outgoing water and gas exhaust are the same or differ by no more than 3 C.: the
difference between the in-coming and out-going water should be about 10 C.
The temperature of the inlet water is noted, the reading of the gas meter taken,
and at this same time the outlet tube changed from the funnel to the graduate.
Ten successive readings of the out-flowing water are taken while the graduate
(2-liter) is being filled and the gas shut off.
A better procedure is to collect about 10 liters of water and weigh it. The
thermometer in the outlet can then be read every half minute. For the procedure of the Bureau of Standards, see pp. 2431-2433.
Calculations.-The volume of gas, V, is determined by the difference in the
m~ter readings: this must be reduced to operating conditions, i.e., 60 F., 30
inches and saturated with moisture. The height of the barometric column
must be reduced to 0: this is done by observing its temperature and referring
to the proper table in one of the usual" Handbooks of Physical Constants,"
and noting also whether the scale is brass or glass. The certification correction
is also to be applied. The pressure of the gas in inches of water at the meter is
changed to inches of mercury (1 in. =0.07 mm. mercury) and added to the
corrected barometric reading; this represents the total gas pressure: its temperature is given by the meter thermomete.r. Table 9 will then give a means
of reducing the volume.of gas shown by the meter to operating conditions.
To determine the temperature rise, T, subtract the temperature rise of the
inlet water from that of the outlet water: the corrections should have been
applied to these thermometer readings and the correction for stem exposure,
Table 10.'62 The amount of water, W, heated is determined by weighing, the
weights having been verified.
The observed heating value is
WXT
VXF

F is the reduction factor obtained from Table 9. This heating value is corrected for the effect of atmospheric humidity, Table 11; 163 this is additive to
.162

163

Bureau of Standards, Circular No. 48.

Circular Dept. Public Utilities of Massachusetts, Gas Inspection Division.

2394

GAS ANALYSIS

90% humidity. The heating value is further corrected for the loss of heat
from the calorimeter which under the usual operating conditions is 1 B.t.u.
Where the water temperature is quite different from room temperature, a
correction. must be applied, additive if the room is. warmer, and subtractive if
colder. For calculating total heating value, this' ~orrection amounts to 0.5
B.t.u. per degree difference F,at a room temperature of 50 F. and increases
by 0.1 B.t.u. for each increment of 10 F. room temperature: for calcula.ting the
net heating value it is 0.4 B.t.u. for 50 o-8ao F. room temperature and 0.5
B.t.u. for 90-100 F. The average of the different runs, which should agree
within 1 %, corrected for humidity and heat loss, is the total heating value of the
gas, and is the one to -be reported. The net heating value is obtained by subtracting the product of the volume of the condensate in ml. X2.3 B.t.u. (0.6 C.)
from the observed heating value and correcting for heat loss.
Page 8d from Circular 48, Bureau of Standards, shows how the record is to
be kept. (See p. 2433.)
Thomas Recording Meter.-In many gas works this apparatus is'in operation and agrees satisfactorily with the Junkers instrument. 164
NOTEs.-After setting up the apparatus the first thing to be done is to turn on the
water-(not the gas). Similarly, the water should be shut off last. All connections and
the meter should be tested for leaks before each test. The water level in the meter
should be checked daily.' Slight drafts caused by moving suddenly near the apparatus
will vary outlet readings and vitiate the test, The instrument should not be set up near
a window or heating apparatus where radiant heat might affect the readings.
If 0.2 cu. ft. of gas is burned, then an error of 0.1 F. in temperature of water means
an error of 4 B.t.u.; an error of 0.01 lb. water, 0.9 B.t.u.; 1 F. in gas temperature, 1.8
B.t.u.; 0.1 in. (barometer), 2 B.t.u.; 1 in. water pressure of gas, 1.5 B.t.U. 165
The total heating value of the gas is the heat given off when the gas is used for heating
water or in any operation where the products of combustion pass off below 100 C. The
net heating value represents the conditions in which by far the greater quantity qf gas is
consumed, for cooking, heating and gas engines, and is one which should be reported.
It should, howeyer, be corrected,!s6 to the legal cubic foot, that is; measured at 30 ins.
barometric pressure, and 60 F. saturated with moisture.
. The chief sources o~ error a~e,167 in adjusting the meter; in measur~ng the temperaturense of the water, and m changmg over the outflo~ wat~r to the welghed vessels.

(b) By Calculation,168-Let us suppose a city gas gave the following analysis:

Illuminants 15, carbon monoxide 25.3, methane 25.9, hydrogen 27.9%; the
heating value of these gases according to Ta;ble 4, page 2426, is as follows:

0.15 X 1900 = 285.0 B.t.u.

0.253X 323= 81.7
0.259XlO08=261.0
0.279X 325= 90.7

718.4 B.t.u.
I

164 Bur. of Standards J. Research, 10, 99. C. A., 27, 1239 (1933).
165 Rept. Joint Committee on Calorimetry, Public Service Commission and Gas
Corporations, in the Second Public Service District of New York State, p: 81, 1910.
166 A difference of 1 C. or of 3 rom. pressure makes a change of 0.3% in the volume.
Pfeiffe, J. Gasbeleucht., 50, 67 (1907).
rI
167 Technologic Papers of the Bureau of Standards~ No. ;36.
"Gas Calorimetry,"
Waidner and Mueller, page 100, 1914.
. _J- ~
168 U. S. Geol. Survey Paper No. :_lSj Part III, page 1005.

GAS ANALYSIS

2395

which is the gross heating value of the dry gas. The correction for the heat
lost is found, as with'liliimney gases, by multiplying the volume of the combustion gases by their weight X specific heat Xrise of temperature.
(c) Sulfur.-Sulfur, being present in gas in so many forms, is determined by
combustion and usually reported in grains of sulfur per hundr~d
cul)ic feet.
One of the most easily portable and satisfactory forms is that of
Hinman and Jenkins described as follows: 169 The upper vessel, Fig.
374, is a " bead glass" 300 mm. long and 60 mm. in diameter; this
is filled with large cut-glass beads, held up by a suitable fluted glass,
giving a large condensing surffl,ce without obstructing the draft.
To this bead glass is attached, by a rubber connector, the adapter,
410 mm. long and 50 mm. lower internal diameter. To the upper
adapter is attached, by means of the" connecting piece," the lower
adapter, 400 mm. long and 40 mm. lower diameter. The connecting piece projects 12 mm. above the top of a rubber stopper, fitting
the. upper adapter, and is surmounted by a watch-glass deflector
carried on platinum wires. An overflow tube carries the condensation to the Erlenmeyer flask hung on the stopper as shown; this
tube is so adjusted that some liquid remains on the stopper to keep
it cool and to absorb some of the ascending gases. The Bunsen
burner is fitted with a lava tip having a 5-mm. hole; surrounding the
burner is a glass tube 20 mm. in diameter, forming the inner wall of'
an annular chamber, of which the outer wall is a glass ring 50 mm.
in diameter. Into this chamber, which serves to contain 10% am- FIG. 374.
,monium hydroxide, the lower adapter dips 10 mm.
The lower adapter is joined to the" connecting piece" by a short cork-lined
metal tube. AlthOugh radically different in form, this apparatus is very similar
to the Referees' in general principle and in ~ethod of use, the principal difference being the use of ammonium hydroxide instead of dry ammonium carbonate
as a source of ammonia. About 10 m!. of concentrated ammonium hydroxide
is placed in the reservoir about the burner at the beginning of the test and about
5 m!. more added every fifteen or twenty minutes. The gas is consumed at
the rate of 0.4 to 0.6 cu. ft. per hour, and 2.5 to 3 ft. burn~d, if the sulfur is to
be estimated gravimetrically; otherwise 1 cu. ft. is enough. When the run is
coinpleted the apparatus is allowed to cool and is then flushed four times by
pouring 50 ml. portions of water in at the top of the bead tube. To the solutions
and washings are added 2-3 m!. bromine water, and it is evaporated to 30 or
40 m!.; an excess of a hydrochloric acid solution of barium chromate is added
to the hot solution, it is gently boiled, an excess of dilute ammonia added, again
boiled for a minute, filtered and washed. The ammonium chromate in the
filtrate (the chromic acid being equivalent to the sulfuric acid in the original
solution) after being boiled in a stout flask, with a Bunsen valve, to expel the
air, is cooled and titrated directly with stannous chloride (3.25 grams Sn per
Vter) using starch ~nd potassium iodide to accentuate the end-point.
169 Jenkins, J, Am. Chern. Soc" 28, 543 (1906); also Technologic Paper No. 20,
Bureau of Standards, McBride and Weaver, "Determination of Sulfur in Illuminating
Gas," 1913, also Stone, book cited.

2396

GAS ANALYSIS

The equations are:,

(NH4) 2S0 4+ BaCrO 4= BaS04 + (NH 4) 2Cr04,

2(NH 4)2Cr04 +2HCI = (NH4)2Cr207+2NH4CI + H 20,
.
1
3SnC1 2 + (NH4)2Cr207+ 14HCI = 3SnCl4+2NH 4CI +2CrCI 3+7H 20.
I

The strength of the stannous chloride should be determined at the same time
by standard bichromate of potassium.
Or the sulfuric acid can be determined with the turbidimeter as for sulfur
in coal. The amount of sulfur is usually from 20 to 30 grains per 100 cu. ft.
Hydrogen Sulfideyo-The test is made by hanging a strip of paper moistened with lead acetate solution (1 :20) in a bell-jar or tube through which
the gas is passing at about 5 cu. ft. per hour and allowing it to act for one minute. Usually several tests are made. The gas should be taken fresh from the
main and care should be taken not to confound any black tarry spots with lead
sulfide. A properly purified gas should give no test.
It is quantitatively determined by drawing a known volume of the gas
through standard iodine solution. Tutweiler 171 measures the gas in a modified
Bunte burette over starch solution, runs in a known quantity of standard iodine
. solution until it is in ~light excess. If 100 ml. of gas were taken, the number of
milliliters of solution gives the grains of H 2S per 100 cu. ft., 1 ml. iodine
= 0.0017076 gram iodine = 100 grains H 2S per 100 cu. ft.
(d) Ammonia. 172---This is determined by absorption in standard acid
colored with cochineal: 10 ml. of HCl173 are placed in the bulb, Fig. 361, 2--3
drops cochineal solution added, and the gas allowed to bubble through it until
the yellow calor changes to a deep purple; the meter is now read. The acid is
made by diluting 38.2 ml. N /10 HCI to 1 liter, 10 ml. = 0.01 grain of NH;; the
cochineal solution is made by treating 3 grams of the ground insect with 25Q ml.
20% alcohol, allowing to stand forty-eight hours and filtering. The bubble
tube is inserted in series with the gas supply to the sulfur apparatus, Fig. 374
(c), so that both determinations are run at one time: the gas is passed through
at the rate of 0.5 to 0.6 cu. ft. per hour. M~ssach'usetts law limits the amount
of ammonia to 10 grains per 100 cu. ft.
(e) Analysis.-The volumetric analysis is carried out according to p. 2355;
either bromine water or fuming sulfuric acid can be used to absorb the" illuminants." Besides ethylene, it may be desirable to determine benzol: this is
best done according to Dennis, O'Neill and McCarthJ-17_4.bJ' absorption in an
ammoniacal solution of nickel cyanide.
Naphthalene.-This is determined in purified gas by passing i\ through
N /20 picric acid solution. White 175 determines it in raw gas by precipitation
of the picrate and subsequent recovery of We naphthalene.
170 Tech. Paper, No. 41, Bureau of Standards, "Lead Acetate Test for Hydrogen
Sulfide in Gas."
,
"
171 J. Am. Chern. Soc., 23, 173 (1901).
172 Tech. Paper No. 34, Bureau of Standards, "Determination Of Ammonia in Illuminating Gas"; also Tech. Paper No. 48 (1916).
173 This is/used by Jenkins; Sulfuric Acid 1 mI. =0.01 grain of ammonia is recommended in T.P. No. 48.
.
,
174 J. Am. Chern. Soc., 30, 236 (1908).
. -' .176 Proc. Mich. Gas Association, 83, 1904 (1905).

2397

GAS ANALYSIS

(f) Carbon Dioxide.-This ma-y be determined by Rudorff's method 176

which consists in titrating about a liter of the gas with standard potassium
hydroxide. The arrangement and manipulation of this apparatus will be
evident from Fig. 375; the capacity of the Woulff bottle must be known and if

FIG. 375.

FIG. 376.

the gas contains hydrogen sulfide, it must be absorbed by passage over manganese dioxide. It can also be determined by the usual absorption method.
(g) Specific G~avity.177-The readiest method depends upon the time of
I

Composition of Commercial Gases 17S

CO,

Coal. ........
Carbo water...
Blue water ...
Pintsch ......
Blau .........
Oil-water .....
OiL .........
Gasoline .....
Acetylene ... ,

1.6
3.0
3.4
0.2
O.
2.6
0.3

NaturaL .....

0.3
0.4

..
..

Ill'ts

0,

co

cu,

II,

C,U,

- -- - - -- -- -- - - 4.0
13.3
0.0
30.0
51.9
7.0

31.3
U

96.0
0.3

... .

0.4
0.4
0.9
O.
O.
0.2
O.
18.5
0.8
0.3

.. ..

8.5
30.4
40.9
0.1
0.1
9.2
2.4

.. , .
....
0.5
'"

49.8
37.7
50.8
13.2
2.7
39.8
13.5
'"

3.2
3.2
O.
9.0
O.
.. ,3.9
. C6HI4.~ 10.3

.. ..

2.3

. ...

29.5
10.0
0.2
45.0
44.1
34.6
46.5
.,

..

92.6
87.0

..

"

....
9.6

Hempel, op. cit., 262.

See also Chapter II.
178 Fulweiler, Rogers, "Industrial Chemistry," 5th Ed., 729.

116

177

)[,

Candie
Power

B.t.u.

3.2 16.1
622
2.1 22.1
643
3.5
299
1276
1.6 43.0
1.2 48.2
1704
6.6 19.7
680
1320
1.1 38.0
69.7 16.0
514
13.~0
3.2 225.0
H 2S
3.5 ().2 840-1174
3.0 ....
1040
'0' .

GAS ANALYSIS

2398
efflux of the gas compared

~ith air; sp.gr'=AG2

2

G and A represent the times

j

of efflux in seconds of gas and air. The apparatus is that of Schilling and is
obtainable from the dealers, or may be constructed according to Jenkins 179, 180
as follows:
'\
" It consists, Fig. 376, of two large rubber stoppers, each having a brass tube,
projecting laterally near the large end, and connecting with the hole in the
stopper. A glass piece A in the form of a truncated cone fits tightly over one
stopper; it is 9 ins. long, lYz ins. diameter at the base and 1 in. at the top.
A similarly shaped piece B 9 ins. long by lYz ins. diameter at the lower end fits
over the second stopper; 2 ins. above the latter the tube has a constriction 1 in.
_in diameter, and at its upper part is narrowed to a neck 5/16 in. in diameter
which is ground on the inside to receive the end of a tube 7Yz ins. long aiid ~ in.
in diameter, in the upper end of which is fitted a platinum plate containing the
emission orifice. One and three-fourths inches below this plate is a three-way
glass stopcock, and 3 ins. below the latter a scratch surrounds the tube and
serves as the upper mark in the escape of the gas.
" Fitted into the hole in the stopper is a hollow cylinder of brass to which is
soldered a curved piece of brass wire pointed at the end, which rises lYz ins.
above the surface of the stopper. The tlVo brass tubes projecting from the
outside of the stoppers are joined by a piece of rubber tubing 15 to 18 ins. long.
" In using this instrument the larger tube B is filled with water, of the temperature of the room, nearly to the top, the stopcock being turned so that egress
of air from the smaller tube is prevented. The larger tube is placed on an
elevated surface just high enough so that its bottom is above the level of the
scratch on the narrow outlet tube, the cock is turned so that the air may escape
'through the orifice in the platinum plate, and on the second, when the point of
the bra,'ls wire breaks the surface of the rising water, a stop watch is started.
The latter is stopped when the water exactly reaches the scratch.
"
" The large tube is lowered, and the stopcock turned so that at( may enter
through its hollow point. When the water is again alnn the large cylinder, the
cock is turned to connect the small vessel with tlie outside air through the
platinum tip, the 'large cylinder is replaced on the elevation and the operation
repeated. Results should be obtained which ,check within one-fifth second.
" Now connect a rubber tube to the gas supply and to the tip of the stopcock, lower the large cylinder and force the water into the latter by means of
the gas pressure. ThOl;oughly saturate the water with-.the.t_gas to be tested;
this may be done by shaking gas and water together and by forcing the water
up and down in the small vessel in contact with the gas. Repeat the operation
with gas in the same manner as described for air. The calculation is made in
accordance with the formula.
" The advantages of this apparatus are its portability, its cheapness, its
rapidity and accuracy. When set up, the cylinders are inclined to be a trifle
unstable; this may be overcome by fastening a lead plate to the base of ea'llh
stopper. Four precautions in connection with its use should be emphasized:
179

Stone, book cited, 261.

J. S. Edward.

See also Bureau of Standards, Tech. Papers 89 and 94, by

I

Orton, Geol. of Ohio, VI, 137. See article by ~em:p~ Collins and Killin, Ind. Eng.
.
Chem., Anal. Ed. 7, 338 (1935).
180

GAS ANALYSIS

2399

(1) The water must be of the'room temperature; (2) the water must be thor~
oughly saturated with the gas; (3) the platinum tip, stopcock, and upper part
of the tube must be kept dry and clean; (4) the large cylinder must always, in
anyone determination, be placed at the same height,"
The use of the Edwards Gas Balance has already been explained on p, 2344.
Another method consists in the use of the Lux gas balance. This consists
of a balanced globe into which the gas previously filtered through cotton, passes
and its specific gravity is read off directly on a scale.
The knowledge of the specific gravity is important; as it is involved in the
formula for the calculation of the flow of gas in pipes; it also enables the gas
manager to ascertain the weight of gas produced from the coal, and to get anidea of the nature and amount of impurities in the gas, all these being heavier.
(h) Tar.-For the estimation of tar, Clemens Winkler 181 recommends the
procedure of Tieftrunk: This consists in passing the gas through 25% alcohol
and collecting and weighing the tar on a tared filter.
In this country the so-called" Tar Camera" is used: this consists of a brass
app.aratus 182 into which a 3-inch E and A No. 4970 filter paper is clamped,
through which a measured quantity of gas is passed. By the increase of weight
of the filter, the amount of tar is determined; the volume of gas may vary from
0.1 to 10 cubic feet.
Stone and Prince 183 use an ordinary Gooch asbestos filter for this purpose.
It is weighed after treating with dried and purified gas, subjected to the gas to
be tested and dried and weighed as before.

IV.

SU~FURIC ACID GASES,

the gases involved in the manufacture of sulfuric acid:

a. Burner gases;
b. Nitrogen gases;

c. Oxygen;
d. Gases'involved in the contact process.
(a) Burner Gases.

Sulfur Dioxide

This gas may be determined by the m~thod of Reich. It consists in aspirating the gas through standard iodine solution eN /10 is suitable) until it is decolorized. The amount of iodine used in the test and the volume of the
aspirated gas being known, the percentage of S02 can readily be calculated.
S02+2I +2H 2 0 = H 2S0 4 +2HI
Fig. 377 shows a Jorm of apparatus for making this determination. The
standard iodine, 5 to 25 mI. N /10 I, diluted to 150 to 200 mI., is placed in the'
bottle, about 400 mI. capacity, and starch indicator added. 184 The gas to be
tested is aspirated through the iodine until the color of the starch blue fades
completely. Water which flows out from the graduated cyllnder by lowering
the aspirating bottle, produces the suction, and the amount measures the
"Die Industrie Gase," page 52; also Hempel, op. cit., 239.
Made by the Steere Engineering Co., Detroit, Mich.
18.1 "Gas Age," pp. 531-3, 1920.
184 Starch indicator may be omitted if the light is good for observing the fading out
of the iodine color.
181
182

2400

GAS ANALYSIS

volume of the aspirated gas. From the quantity of iodine used and the volume
of the gas required to de colorize the solution the pel[ cent of sulfur dioxide i~
calculated.
Should the contact gas contain
80 3, this is: ~bsorbed by passing the
gas through: 50 to 100 ml. of. cone.
H 2S0 4 , to avoid the action of 80 3 on
the rubber tubing of the apparatus.
A rapid current of the gas is passed
through the acid to saturate it with
S02 before making the tests.
The Reich method is more applicable for determining small amounts of
S02. A 12-15-liter graduated aspirating bottle is used in works tests of
exit gases for measuring the gas. Since
these volumes are not under standard
conditions (760 mm. and 0 C.), it
will be necessary to convert the volumes obtained in the tests to these
conditions, using the formula .
.......

V =

FIG. 377.

Portable Reich

App~ratus.

po-w
V0 760 (1 +0.00367t) '

where VO = measured volume, po = observed barometric pressure, t = terp. perature of the gas, and w = aqueous ~apor
pressure at temperature of the test.

(b) Nitrogen Oxides

Nitrogen tetroxide, N 20, and nitrous acid ~ 20 3 , ,6an best be determined by
absorption in standard permanganate (acidulated with sulfuric acid).
Nitric oxide can be determined by passing the gases through soda lye, then
by adding air to the collected volume, converting it to nitrogen tetroxide and
determining it as above indicated.
Nitrous oxide is determined in the acid-free gases by explosion with hydrogen.
(c) Oxygen is usually determined by acid or ammoniac;r cuprous chloride-phosphorus is also employed. The percentage of oxygen should_not exceed six;
a larger amount means that heat is being lost from the chambers by the exhaust
gases. Knowing their temperature, the loss of heat can be calculated as with
chimney gases.
/
'
V. MINE GASES 186
The gases to be sought are those found in city gas and for most purposes the
procedure in p. 2355 can be followed. For small quantities of methane the
apparatus of Burrell 187 should be used.
186 See 'Technical Paper 14, Bureau of Mines, "Apparatus for 'Gas Analysis Labora~::_.
tories at Coal Mines."
181 Bull. 42, U. S. Bureau of Mines, 17,42, 1913; also Technical Paper 39, 13; also
Bulletin 331.

2401

GAS ANALYSIS

Table for Reich's Tests for Percentage 50 2, Using Tenth Normal Iodine Solution

"

Volume % SO,

12
11.5
11

,..

10.5
10
9.5
9
8.5
8
7.5
7
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
1.5
1.4
1.3
1.25

25 mr. NllO Iodine

205
215
226
238
252
265
283
300
321
344
371
402
'438

ml. gas

"
"
"
"
"
"
"
"
"
"
"

"

"
"
"

"
"

"

.. . .
,
. . . .
. . . .
.. . . .
. . .
. . . .
. . .
.. . . . .
. . . . .
. . . .
. . ., . .
. . . . .
'"

'

10 mi. NII0 Iodine

. . .

5 mr. N/lO Iodine

. . . .
.. . .
. . .. . .
. .. .
. . . .
. . . .
. . .. . .
. . .
.
148 ml. gas
161 " "
.

lUi

" "
"
"

192 "
212 "
237 "
268 "
308 "
361 "
436 "

.. . . ' '.' .
. .. ., . .
. .. . . .
. . . .
.. ., . .
'

1.54 ml.
181 "
218 "
274 "
367 "
393 "
424"
442 "

gas
"
"
"

Sulfur Dioxide in Exit Gases, Pages 1266 to 1269 inclusive.

For determination of methane alone, the apparatus of Shaw 188 may be

recommended. This determines first the per cent of city gas necessary to make
an explosion of definite strength with ordinary air; when this has been done,
mine air is used in place of the ordinary air and a smaller percentage of city gas
is required-smaller by the amount of combustible gas in the mine air. The
strength of the explosion is measured by noting by the ear the force with which
the plunger is driven out from the explosion cylinder against a bell.
In case this apparatus be not at hand, Brunck's 189 method can be employed. This consists in burning the methane in a 2-liter Erlenmeyer flask by
means of an electrically heated platinum spiral. The flask carrying the spiral
in the stopper is sunk inverted in a vessel of water and the current allowed to
pass for half an hour, which is sufficient to burn the methane. It is then cooled
and 25 ml. Ba02H 2 (1 mI. = 1 ml. CO 2 ) added, time allowed for absorption of
the carbon dioxide and the excess of Ba02H 2 determined, and the quantity of
methane calculated.
Clowes and Redwood 190. 191 have worked out a method for the detection of
inflammable gas in air, employing the" flame cap." When an inflammable
atmosphere is brought in contact with a candle or better, a hydrogen flame,
Berichte, 27, 692 (1894).
O. Brunck, "Die Chem. Unters. d. Grubenwetter," 1908.
190 "Detection and Estimation of Inflammable Gas and Vapor in the Air," 1896; also
Clowes, J. Soc. Arts, 41,307. Also McTrusty, "Mine Gases and Gas Testing," 1916.
191 "Flame Safety Lamps," Bur. Mines Cire. 227, 1924.
188
189

GAS ANALYSIS

2402

the gali burns, forming a " cap," like the colorless flame above the blue cone in
a Bunsen burner: the length of the flame is a measure df the percentage of gas,
and as little as 0.1 % is visible using the hydrogen flame.
. The ". J. W. Indicator" for the detection of inflar4mable greses or vapors,
depends for its use upon the fact that the resistance of a wire is increased by
the heat of combustion of the gas. The gas or vapor is sucked through a ~ube
containing a platinum wire heated to ,dull redness by a dry-battery current:
the burning of the gas increases the resistance of the wire and from this increase
the quantity of gas or vapor can be determined.
It is safe, easily portable, easily handled by the ordinarily intelligent and
gives direct readings. The Union Carbide Company's indicator is like it in
principle, but the heating element was dropped into the tank itself: this was not
.safe with hydrogen, acetylene or carbon bisulfide. The screens protecting the
element necessitated frequent cleansing. Obtainable from the Bullard Davis
Co., 67 Wall St., N. Y.
Carbon Monoxide.:"_Besides combustible gases or " fire damp," it is' sometimes necessary to get an idea of the amount of carbon monoxide (" white
damp ") in the mine air after an explosion or in the" after damp;" chemical
methods, being too slow, use is made of the behavior of birds and mice when
exposed to such an atmosphere. To this end they are carried in cages by the
rescuing party and their behavior noticed. Canaries show distress in an atmosphere containing 0.15% of CO in fi~e to twelve minutes, or with 0.20%
in half this time: Mice are less sensitive,I92 and men may display distress when
carbon monoxide is as little as 0.1 %, wherealil animals may be unaffected. In
case either is overcome by the gas, resuscitation can be effected by' bringing
them out into the open air again. Repeated exposure of the animals to the gas
would seem to be without influence.

VI.

ELECTROLYTIC GASES

Gases from electrolytic chlorine, hydrogen and oxygen generators. The

following are to be sought for:
/ .
(a) Chlorine, (b) oxygen, (l!) carbon dioxide, (d) carbon monoxide, (e) hydrogen.
'
(a) Chlorine.-HempeI 193 recommends measuring the gas quickly in his
burette over water and then sucking in 5 ml. of 50% potassium iodide solution
through the capillary and shaking; the diminution in volume gives the chlorine.
The other gases are determined in the usual way. l1.8 -phQsphorus' cannot
be used for pure oxygen, a specially prepared potassium pyrogallate, is employed; cuprous chloride or ammoniacal cuprous carbonate in -the absence of
carbon monoxide is very satisfactory.

VII.

ACETYLENE

Commercial acetylene may contain the followiIig gases:

1. Oxygen;
5. Nitrogen;
6. Sulfur-containing gases;
2. Carbon monoxide;
7. Phosphine.
3. Hydrogen;
I
4. Methane;
192 Burrell, Seibert and Robertson, Bureau of Mines-T-;cbnical Paper 62,1914, "Relative Effects of Carbon Monoxide on Small Animals." Also Tech. Paper 11.
198 "Gas-Analytische Methoden," 4th Ed., 1913, p. 278.

GAS ANALYSIS

2403

Oxyge:' is estimated after the absorption of the acetylene itself in bromine

water in the usual way with potassium pyrogallats. Methane and hydrogen
would be determined in this residue, after treatment with ammoniacal cuprous
chloride to remove the carbon monoxide, by the ordinary explosion methods.
Nitrogen would be left as a residue.
Sulfur-containing Gases.-These are most likely organic sulfides, as hydrogen sulfide is probably absent, since the solution from which the acetylene
escapes is strongly alkaline. They can be determined by combustion, as in
city gas, and best be reported as " total sulfur."
Phosphine is also similarly estimated and the phosphoric acid determined in
the usual way: the quantity of PHs rarely exceeds 0.05%.
Acetylene can be purified by passing over bleaching powder, through acid
cuprous chloride or chromic acid: The candle power is usually given as fifty
per cubic foot, or 180-200, when burned at the rate of 5 cu. ft. per hour. The
explosive limits are, according to Clowes 194 3 to 82%, to Burrell and Oberfell,
2.5 to 73%.190

VIII.

ATMOSPHERIC AIR

(a) Moisture; (b) Carbon dioxide; (c) Ozone; (d) Carbon monoxide; (e)
Bacteria. ,
(a) Moisture 196 by chemical means. The amount of moisture can be
determined by the sling psychrometer, or wet- and dry-bulb thermometer or
by the hair hygrometer. The sling psychrometer is the most rapid and accurate
-the wet- and dry-bulb thermometers are so arranged that they can be rapidly
whirled for fifteen or twenty seconds, stopped and quickly read, the wet bulb
first; this ,is repeated until closely agreeing readings are obtained. The
humidity is determined in the usual manner from the meteorological tables.
The amount of moisture in the air is probably indirectly responsible for our
sensations of comfort or discomfort, rather than the amount of carbon dioxide,
as was formerly thought. This moisture controls the heat loss from the body,
which loss must be normal-neither too high nor too low. The greater the
humidity the less the evaporation, consequently the less the cooling.
For comfort, the higher the temperature the less should be the humidity; the
following shows the relation between the two:
Temp., 0 F. . . . . . . . . . . . . . . . . . . . . . .. 60
Relative humidity, %.............. 67

68 197
40

70
49

80
31

90
16

(b) Carbon Dioxide.-One of the most satisfactory methods is that of

Hesse.l 98 This consists in absorbing the carbon dioxide from a definite volume
of air with standard 'barium hydroxide and determining its loss of strength.
The air is collected in heavy Erlenmeyer flasks of 100-500 or 1000 ml.
capacity, or clear glass bottles; these are stoppered with doubly perforated
rubber stoppers carrying glass plugs and the capacity of the flask or bottle
noted to the depth of the stopper. A IO-ml. pipette and a 15-ml. glass-stopBook cited.
Bureau of Mines Tech. Paper 112, "Explosibility of Acetylene."
196 Benedict, "The Composition of the Atmosphere with Special Reference to its
Oxygen Content," Carnegie Institution of Washington, 1912, Publication No. 166.
197 For indoor work.
198 Eulenberg's Vierteljahrschr. f. ger. Med. u. San., N. F., 31, 2.
19'

195

2404

GAS ANALYSIS

pered burette graduated in m!'/19 with an 8-cm. tip, complete the outfit; a
solution of 1.7 grams of barium hydroxide and chloripe (20: lrin a liter of
water, B, Fig. 378, and of oxalic acid 5.6325 grams ;per liter (1 ml. = 1 m!.
CO 2), with phenolphthalein 1: 250, are required. Th~ small bottle attached
to the large one is filled with soda lime.
The bottles are filled with steam by exposure for
,
three minutes and the vaselined stoppers inserted, or
they may be filled with distilled water and opened in
the place the air of which is to be examined.
In all this work, it should be remembered that the
exhaled breath contains about 400 parts carbon dioxide
per 10,000, consequently care should be taken not to
contaminate the samples, nor should they be warmed
A
with the hand.
The 10-m!. pipette is partly filled from the tube A,
Fig. 378, by means of the rubber connector-and sucking
the liquid into the pipette: it is rinsed, filled and inserted
through one of the holes in the rubber stopper of the
FIG. 378.
bottle, the other plug being momentarily opened. The
plugs are replaced and the bottles allowed. to rest on
their sides, with occasional rolling, for twenty minutes. Not more th;m onefifth of the solution should be used up by the carbon dioxide present.
During this time, the barium hydroxide should be standardized; to this end
a few drops of phenolphthalein and a quantity of the oxalic acid almost sufficient
to neutralize the hydroxide should be run into a 100-ml. Erlenmeyer flask from
the burette; this should be passed through the doubly perforated stopper; 10
ml. of the barium hydroxide solution are run int.o the flask as above described,
and also the oxalic acid until ,a pink color appears..
Phenolphthalein is added to the bottles containing the samples, the oxalic
acid burette inserted through the stopper and the excess of barium hydroxide
titrated.
The barometric pressure and temperature in the laboratory are noted and
the volume of the bottle less 10 ml. Ba(OHh calculatei to standard conditions;
the difference in the titer of the barium hydroxide solution gives the volume
of carbon dioxide in the bottle; this is calculated into parts per 10,000.
A more rapid method is that of Cohen and Appleyard.199 This uses a
standard solution of lime, water colored with phenolphthalein to absorb the
carbonic acid: it depends upon the principle that the quantity of air and lime
water remaining the same in each case, the rate of its decolorization will vary
inversely with the amount of carbon dioxide.
Half-liter bo'ttles are used for the purpose and either filled with water, or a
vacuum is produced in them by steaming out and stoppering; the sample is
taken by displacing the water or allowing the air td' fill the vacuum: 10 ml. of
the standard lime water (Reagents) measured with a Vanier overflow pipette
run into the bottle, the time noted, the rubber stopper replaced, and tht bottle
shaken vigorously and continuously, until the pink color disappears when seen
over a white paper, and the time again noted. From the ..time required for
decolorization, the amount of carbon dioxide mll:yjJ~dbund from the table.

m Chern. News, 70, 111 (1894).

GAS ANALYSIS
Time, Seconds

Parts CO, per 10,000

30
60
90
120

15.6
9.9
7.2
5.5

2405

Time, Seconds

Parts CO, per 10,000

150
180
210

4.4
3.8
3.6

Other methods for this determination are more strictly gasometric, measuring the diminution in volume by absorption: Benedict 200 used Sonden's, and
Anderson 201 recommends a shortened form of the Pettersson-Palmquist
apparatus.
This may be d~scribed as an Orsat apparatus using mercury instead of water
and with a 25-ml. burette the lower part of which is graduated to 0.0025 ml.;
this is connected to a pipette of potassium hydroxide, a delicate manometer
and compensating tube. The apparatus is delicate and rapid, but requires a
skilled operator to manipulate it. Any form of accurate apparatus may be
used.
The amount of carbon dioxide in the outdoor air of the city is about 3.1 parts
per 10,000, in the house, 3.7-3.9; with 6-7 parts in a room, the ventilation may
be considered as excellent, with 10 parts as about the upper limit. In some
theaters which were lighted bY, gas it approached 50 parts.
(c) Ozone.-Probably most conveniently determined by Wurster's method,
p.2370.
Determination of Ozone, Iodide Method of Schonbein. 202 The method
depends upon the reaction
2KI+Oa+H 2 0=I 2 +0 2 +2KOH.
Procedure.-A glass bulb of 300 ml. to 400 ml. capacity (the exact capacity
being determined by weighing empty, then full of water) with two tubes at
opposite sides, closed by stopcocks, one of which was a three-way cock, is
filled with water. The ozone test is introduced into the vessel, displacing the
water. The gas is brought to atmospheric pressure by quickly opening and
closing one of the cocks. A solution of 2 N potassium iodide is introduced in
excess of that required, through the three-way cock by ,means of a second bulb
connected by rubber tube to the vessel after displacing the air in the rubber
connection through the cock. The mixture is well shaken and.,allowed to stand
half an hour. The contents of the bulb are washed out with additional iodide
solution followed by ~istilled water and the liberated iodine titrated with
standard thiosulfate.
1 m!. N /10 N a 2S20a = 0.0024 g. Oa.
(d) Carbon Monoxide.-The qualitative detection is most certainly effected
by the blood test; to this end the gas is drawn through a solution of blood
contained in a Wolff, Fig. 361, or similar absorption tube and examined for its
2ao
2a1

Book cited, also Haldane's.

J. Am. Chem. Soc., 35, 162 (1913).

202 Further details of this method may be found in Treadwell and Hall, "Quantitative
Analysis."
.

GAS ANALYSIS

2406

absorption spectrum~ The blood solution is made by mixing ox blood which

has been defibrinated by whipping, with an equal quantity of a cold saturated
solution of borax; this can be kept (as a side-shelf reagent) in the laboratory
for months. This solution is diluted with 19 volumes of water, giving a solution
of blood of 1 in 40 whicli is pladed in the absorption tuHe. The air is drawn
through it at a r~te of 3 liters per hour,
F
bED
C B
requiring 10 liters in some cases; the
solution is put in a thin flat-sided
bottle and spectr~scopically examined.
Pure diluted blood, Fig. 379, shows
2
two d!1rk absorption bands, spectrum 2,
between the D and E line, these are
3
welded into one broad band by reducing
agents as NH 4 SH, spectrum 4; blood
which has absorbed carbon monoxide
shows two broader bands in the same
FIG. 379.
place, spectrum 3, which are unaffected
by reducing agents.
The quantitative determination depends upon the equation,

5CO+I 2 0s=5C0 2 +1 2
This has been studied by many, and recently by Vandaveer and Gregg 203
and by the writer. They pass the gas to/be tested through two scrubbing
Friedrichs (spiral) towers filled with chromic acid at the ordinary temperature,
and through solid potassium hydroxide and phosphorus pentoxide. The iodine
pentoxide is heated in a U-tube to 1500 in an electric air-bath. The highest
quality of the reagent should be used, as it is found that even this requires
two days' heating at 205.,..215 while air or nitrogen is slowly drawn through
the apparatus. When not in use, the apparatus should be kept tightly closed,
and if the parts of the apparatus be fused. together, instead of using rubber
connectors, a blank can be obtained in fifteen minutes. I Water' vapor seems
to liberate iodine.
I
When a blank is obtained, the gas to be tested is passed through the ap:paratus, and the iodine collected in KI solution; titrated with .001 normal
thiosulfate; every 2.27 mg. iodine corresponds to 1 ml. of CO under standard
conditions. A complete analysis with an accuracy of 0.002% can be made in
about twenty minutes.
Sayers and Yant 204 have worked out a method for the determi'ilation of the
gas by taking advantage of the carmine red c'olor produced by carbon monoxide
blood in tannic acid, and comparing the red produced with standards, as with
" Hoolamite," p. 2373.
For the determination of the gas in vehicular tunnels it is absorbed in
" Hopcalite," a mixture of the oxides of manganese, copper, cobalt and silver.
This produces heat whioh changes the conductivity of a' wire carrying a current;
this is converted into per cent carbon monoxide and automatically recorded;
the same current also turns on the ventilating fans. 20Ii
203 Ind. Eng. Chern., Anal. Ed., 1, 129 (1929).
204 Tech. Paper 373, Bur. Mines, 1925.
206 Bur. Mines Tech. Paper 3551926.
I

--

GAS ANALYSIS

2407

Palladium chloride solution absorbs carbon monoxide with the precipitation

of the metal as a black powder: attempts to make in quantitative have not
been successful. Other reducing gases must be carefully excluded. See page
2406.

IX.

NATURAL GAS

The determinations to be made are for carbon dioxide, oxygen, methane,

ethane, gasoline content and helium.
Burrell and Seibert 206 state that the accurate determination of the constituents of natural gas has proven a stumbling block to gas analysts not
familiar with the work. Technical forms of gas-analysis apparatus and
established rules for bringing a gas mixture in contact with the absorbents for
different constituents are not effective in all cases .. Many samples contain
absorbable constituents, such as carbon dioxide and oxygen, in extremely small
quantities. The fact that oxygen may be a constant constituent of natural gas
as it leaves a well has not been determined absolutely. The writers believe
that traces of oxygen reported in some samples were due to the contamination
of the samples with air. F. C. Phillips 207 detected only minute quantities of
oxygen in natural gas from western Pennsylvania after the gas had bubbled
continuously for many hours or days through reagents.
Olefin hydrocarbons and carbon monoxide have not .been identified in the
samples already received at the Bureau of Mines Pittsburgh laboratory. Because of the absorption of higher members of the paraffin series by fuming
sulfuric acid and cuprous chloride, a natural gas that does not contain olefin
hydrocarbons or carbon monoxide but does contain thelle higher members of
the paraffin series will, when treated with these solutions, undergo a reduction
in 'volume and lead the analyst to a wrong conclusion.
Natural gas as a rule contains a large proportion of paraffin hydrocarbons, in
some instances as much as 99%, so that if the paraffins are determined by
explosion methods, in which a small quantity (8 or 10 rn,l.) of gas is used, a
slight error of manipulattrm will be multiplied ten or twelve times in calculating
results to a percentage basis. The only hydrocarbon that some natural gases
contain is methane; but if a small quantity of sample is taken for the combustion
analysis, errors are magnified. The relation between the volume of carbon
dioxide and the contraction produced by the combustion may indicate hydrogen, although hydrogen is not present. Though the error in the observed data
may be small, yet by the calculation to a percentage basis it may amount to
several per. cent of hydrogen. Many published analyses of natural gas are
undoubtedly much in error owing to such causes.
The paraffin hydrocarbol(s that are gaseous at ordinary temperatures are
methane, ethane, propane, and butane. Since the last two are liquefied 208 at
ordinary temperatures by pressures below those Sound in most producing wells,
quantities other than very small proportions carried by the permanent gases
would scarcely be found in natural gases coming from wells under much pressure. On the other hand, gases drawn from wells by means of a partial vacuum
may contain these and higher paraffins in considerable quantity.
Bull. 42, Bureau Mines, 1913; also 197, 1926.
Phillips, F. C., "Composition of Natural Gas; Researches upon the Phenomena
and Chemical Properties of Gases," Am. Chern. Jour., 16, p. 411 (1894).
208 Tech. Paper 10, Bureau of Mines.
206

207

2408

GAS ANALYSIS
Determination of Moisture in Gases

The gas to be tested is passed through a dehydratihg agent such as phosphorus pentoxide, P 2 06, alumina, A1 20 3 , or lime CaO, contained in a weighed
U-tube, Fig. 380. The following facts should be borne.ih mind in selecting the
dehydrating agent: (a) It should
not absorb the gas; (b) it should
not react chemically with the gas.
For example-lime or alumina could
not be used for determining moisture
in sulfur dioxide, nor could phosphorus pentoxide be. used in determining moisture in ammonia. For
the former, phosphorus pentoxide is
best, and for the latter, lime. AlumFIG. 380.
ina that has been carefully heated to
1400 0 is useful for determining moisture in neutral gases. It should be remembered that gases
dried over calcium chloride will give up moisture to strong
sulfuric acid, and these in turn will give up moisture to
phosphorus pentoxid~: Professor Morley has even deter- .
mined the amount of moisture that is left after this latter
~~~~.
o'"
Procedure.-The volume of the gases required for the
test varies widely according to the percentage of moisture
1
in the gas, 1000 m!. to 10,000 ml. are generally required. C
For minute amounts of moisture it may be necessary to
lead the gas over the dehydrating agent for a given
length of time, using a manometer or differenoo. gauge, or
a gas meter. The absorption tube is weighed before and
after the test and the increase in weight taken as the
moisture content of the gas.
.'
Method of Determining Gasoline Vapor in ~asebus
Mixtures. 209-Fig. 381 shows the apparatus for the gasob
line-vapor determination. The bulb contains phpsphorus
pentoxide for removing water vapor. If the latter were not
removed it would also be retained at low temperatures
and would subsequently exert pressure when measurement was being made of the pressure exerted by the
gasoline vapor.
To start a determination the apparatus is connected to a vacuum pump and
its air exhausted. The mixtu,re of gasoline vapo'r and air is then introduced
at atmospheric pressure, the barometer is read, and th,e two bulbs are immersed
in liquid air contained in a Dewar flask. After about ten minutes, the air is
removed from the apparatus with a vacuum pump. The stopcock on the
apparatus is then closed, the liquid air in the Dewar flask removed, the gasoline
allowed to vaporize, and its pressure read on the mercury manometer attached
to the apparatus. The ratio of this pressure,to the pressure of the atmosphere
gives the percentage of gasoline vapor originally in-tl1_ll _air.
209 U. S. Bureau of Mines, Technical Paper No. 115, Burrell and Boyd.
/

GAS ANALYSIS

2409

Method of Sampling

Some samples of gas taken by the bureau's investigators are collected by

;tir displacement i.n bottles with ground glass stoppers. At the wells or supply
pipes a stopper".is removed and the gas is allowed to run into a bottle until
air has been entirely displaced. Usually a few minutes is sufficient. While
the bottle is still in place, the stopper is inserted. As a precaution a~ainst
leakage, melted paraffin is poured over the stopper. The bottles are then
well packed in suitable wooden boxes, and mailed to the bureau's Pittsburgh
laboratory. When received at the laboratory the stoppers are removed under
mercury and the gas is transferred to and analyzed in an apparatus that contains mercury as the confining fluid. Some samples have been received in
strong iron cylinders at the full pressure of the well. Other samples have been
collected by a method devised by G. A. Hulett, formerly chief chemist of the
bureau, and at present used for collecting
.
some of the bureau's natural gas samples.
The method is as follows:
Glass tubes about 17 cm. long and 6
cm. wide are provided. at" one end with ('.
a glass tube extension about 15 cm. long
and having a 6 mm. bore. The tube is
'left open. Natural gas is directed into.
the sampling tube through the orifice by
means of a long slender brass tube. When
the air originally in the tube has been
displaced, the brass tube is withdrawn
and the glass tube orifice closed by placing
the finger over it. The sampling tube is
then carried to a small alcohol flame and
r
the glass tube extension sealed off.

Simplified Apparatus Used in Analyzing

.
Natural Gas
The apparatus used to analyze natural gas is shown in Fig. 382. The pipette
a contains potassium hydroxide solution;
1
pipettes band d contain alkaline pyroI~
gallate solution. The pipette c is the slow
combustion pipette. The burette f has FIG. 382. Laboratory Apparatus for
a capacity of too ml. and is graduated to
Natural Gas Analysis.
0.1 ml.
Procedure of Analysis
Natural gas is analyzed at the bureau's Pittsburgh laboratory in the following manner:
.
Oxygen or other gas left in the horizontal capillary train is displaced by
drawing a few ml. of nitrogen from the pipette d (Fig. 380) into the burette.
This mixture is then allowed to escape into the atmosphere.
About 100 ml. of the gas sample is then drawn by mercury displacement
from the sample container into the burette. The sample is measured in the

2410

GAS ANALYSIS

burette against the pressure existing in the compensating tube f by bringing

the mercury in the manometer tube exactly to the mark g. The sample is
then }lltssed successively into the potassium hydroxide and alkaline pyrogallate
solutions for the removal of carbon dioxide and oxyger/J, burette measurements
being made in the same manner as with the original slimple.
The residual gas left after the carbon dioxide and Qxygen have been determined is discarded and a fr~sh part of the sample taken for the combustion
analysis. The capillary connections are cleared of combustible gas by dilution
with air, and about 100 m!. of oxygen is measured into the burette and passed
into the combustion pipette. About 35 m!. of the gas sample is then dtawn
into the burette from the sample container and measured. The platinum
wire in the combustion pipette is then heated to a white heat and the gas
sample passed at the rate of about 10 m!. per minute into the combustion
pipette containing the oxygen. The paraffins burn as fast as they enter, so
that an explosion consequent upon an accumulation of gas and oxygen.can not
follo'w. In analyzing natural gas the authors have obtained the be'st results by
passing the oxygen into the pipette first. When the natural gas is passed'in
first, the mixture does not always burn as satisfacto'rily.
After burning, which requires abo'ut four or five minutes, the combustion
pipette is allowed to cool, and the contractio'lf'in volume due to' the combustion
is measured. The carbo'n dio'xide pro'duced by the combustion is determined
by abso'rption in the potassium hydroxide-solution. Finally, the gas is passed
into the alkaline pyrogallate solution to make sure that sufficient o'xygen has
been present fo'r the complete oxidatio'n of the paraffins. Some samples may
contain such a large proportion of the higher paraffin hydrocarbo'ns that 100 m!.
of oxygen will no't be sufficient for the complete oxidation of 35 m!. of natural
gas. Fo'r such c;amples a smaller quantity of the gas must be used for..the
combustion. This statement has especial reference to' those natural gases
that are used for gasoline manufacture and contain a large percentage of the
higher paraffins.
,I
In no' case is the burette mercury raised abo'v~ the upper burette stopco'ck.
The gas remaining in the capillary tubing at any s'tage of the analysis is brought
in Co'ntact with the solutions by passing it back and forth several times between
the burette and the pipette. After combustion, for instance, some carbon
dioxide will remain in the capillary tubing between the combustion pipette
. and the burette when the, gas is drawn back into the bui'ette ..ts> reGord the
contraction in volume. After most of the carbon dioxide has been absorbed
by the passage o'f the gas into the potassium hydroxide pipette, the small
quantity of carbo'n dioxide in the capillary tubing is swept out of the tubing
into the potassium hydroxide pipette. This operation is repeated in order ~o
insure the complete remo'val of the carbo'n dioxide., If this precaution were'
not taken, an error of considerable magnitude would: result.
The apparatus may be used also' for the analysis of other industrial gl.tses,
such as producer gas and blast furnace gas, when all three of the comb-ustible
constituents, carbon monoxide, hydro'gen, and methane, canI be determined
simultaneously by triple, co'mbustion.

GAS ANALYSIS

2411

Calculations from Combustion Data

If the contraction and 'the volume of carbon dioxide from the combustion
of the gas indicate methane, only this constituent is reported. But if the data
indicate higher members of the paraffin series, a calculation is made that gives.
the two predominating constituents. The gaseous hydrocarbons react with
oxygen as follows:
CH 4 +20 2 =C0 2 +2H 20
C2H 6+3.50 2= 2C0 2+3H20
CaH s+502= 3C0 2+4H 20
C4 H lO +6.50 2= 4C0 2+5H 20

But since the fact that some gases deviate somewhat from the gas laws
has been shown by the experiments of Rayleigh, Leduc, Baume and Perrot,
and others, corrections must be made in the case of some gas analyses where
this deviation is greater than the experimental error.
The next table gives the theoretical and observed specific gravities of the
gases ip.volved in the calculations taken from Lannolt and Bornstein.2l0
BelQ.w are given the equations for the reactions of methane and ethane
with oxygen in which corrections have been made according to the specific
gravity determinations shown in this table.

0.999 CH4+2.000 O2=0.994 CO 2+2H 20

0.990 C2H6+3.500 O2= 1.988 CO 2+3H 20
THEORETICAL AND OBSERVED SPECIFIC GRAVITIES OF CERTAIN GASES 211

,
Gas

CH 4
C 2H 6
C 4H IO
CO ........
CO .......
CO 2
CO 2
N 2
N 2
O 2
O 2 :

SpeCific Gravity

Mol. Wt.

16.03
30.05
58.05
28.00
28.00
44.00
44.00
28.02
28.02
32.00
32.0Q

Theoretical

Observed

0.5539
1.0381
2.0065
.9673
.9673
1.5201
1.5201
.9680
.9680
1.1055
1.1051i

0.5544
1.0494
2.01
.96702
.96716
1.52874
1.52909
.96737
.96717
1.10535
1.10523

Theoretical
Observer
Observed
Baume & Perrot .....
do ..........
Frankland ..........
Leduc ..............
Rayleigh ...........
Leduc ..............
Rayleigh ...........
do .............
Leduc ..............
Rayleigh ...........
Leduc ..............

0.999
.990
.998
1.000
1.000
.994
.994
1.001
1.001
1.000
1.000

As the partial pressure/of gases in a mixture decreases, the gases more

nearly conform in behavior to the gas laws; consequently a table is herein
included showing the correct molecular volume to use in the case of carbon
dioxide for different partial pressures .. The partial pressure of the carbon
dioxide has reference to the ratio of the volume occupied by the carbon dioxide
210 Landolt, H., and Bornstein, R., Physikalisch-chemische Tabellen, 3d ed., 1905,
pp. 222-223; Jour. 'Chern. Phys., 7, p. 367 1909.
211 At 0 C. and 760 mm. pressure.

2412

GAS ANALYSIS

found after the combustion to the total volume of. residual gas found after the
combustion; that is, if the total volume after combustion is found to be 70 ml.
and the carbon dioxide is 40 mI., then the partial pressure of the carbon dioxide
40x760
I
.
will be
70
434 mm., and the proper molecular volum~, namely, 0.997,
will be found from the table.
I
The foregoing equations have reference to their use at 0' C. and 760 mm.
pressure. The following table shows the correct molecular volumes for carbon
dioxide at 20 C. and different partial pressures.
MOLECULAR VOLUME OF CARBON DIOXIDE CORRESPONDING TO DIFFERENT PARTIAL
PRESSURES AT 20 C.

Mm. of Mercury
Molecular Vol.
100, .. ' ., ............ , .. , ., .... " ' ..... ' .... , ........ , 0.9993
200 ...... " " ., , .... , .... , .. , .. , ... ' ., .. , .. , .' ... , .. .. .9986
300 ... , , .... , ....... ' .... ' ...... , . , . , . , ....... , . , .. , .' .9980
400, ..... , .. , .. , .... " " ., .' " ...... '. , .. , ......... , .. .9972
500 ....... , .. ' ,. , .... , .... , ........ ' " ., ., " '. , .. , '. .. .9965
600 .... " ., ., ., .,' .. " ., .' ..... , ,. , ........ , " , .. ' " " .9958
700 ... , .. " ... , .,' .. " ., ., ., " .. ,. ,. '. , .. , ,. ,. ,. " .. ,. .9951
760 ... ' ....... ' .. , ...... , .. ' , . ' , ......... ' , . " . . . . .. .. .9950

In compiling' this table advantage was taken of the work of Rayleigh,212

Leduc 212 and Chappius 212 having to do with the determination of the specific
gravity and coefficient of expansion of carbon "dioxide. The specific gravity
determina_tions were given by Rayleigh and Leduc for carbon dioxide at"Oo C.
and 760 mm. pressure. Values for 20 C. and 760 mm. pressure were determined from the coefficient of expansion of carbon dioxide between 0 C. and
200'C. A graph was plotted from two values, the deviation from the gas laws
at 760 mm. pressure and at 380 mm. pressure.
The coefficient of expansion of ethane between 0 and 20 C. has not been
determined, consequently the same molecular volume was used at 20 C., the
laboratory working temperature, as was reported by Bavme and Perrot 213
at 0 C. The error resulting from this usage can be disregarded without
introducing any appreciable error in the analyses, judging from the molecular volume of carbon dioxide which at 20 C. is only 0.001 different from the value
at 0 C.
Below are given the molecular volumes to be used in the case of ethane for
different partial pressures.
MOLECULAR VOLUME OF ETHANE CORRESPONDING TO DIFFERENT PARTIAL Pl!.ESSURES

Mm. of Mercury
0, ........ , ........... ' .......... ' .. , .....

Molecula'r Vol.
~,

.. , , ..... 1.000

100 .... , .. ' ...... , .., ............. ' .' " .... , ..... ,. .. .. ..

.999

~gg ::::::::::::::::::::::': :::::::::::::::::::::: : :::::: :~~~

400 ............... , .' ............ " , ... , ........ ' " " ..
500 ............... , ... , .' .... " ... , ., .. " ., ... ' ... , '. ..
600 ... , ., : .... , " ... ' ............. , ' .... , .. ' ......... ,.

.995
.993
.992

~gg: :::::::::::::~: :::::::::::::::::::::::::::::; ~~ :::: <~~

See Landolt and 13i:irnstein, Phys.-chem. Tabellen, 1912;'i1.14.
213 Sec Landolt and Bornstein, Physikalisch-chemische Tabellen, 1912, p. ).48.

?12

GAS ANA.4 YSIS

2413

Although the individual paraffins in a mixture of several can not be exactly

determined, one will know with a sufficient degree of accuracy which value
in the above column to use by accepting the value that corresponds to the
percentage of ethane determined from the combustion analysis. It is also
true that in natural gas from many places the partial pressure of ethane is so
low that only a small deviation from the gas laws occurs. A slight error arises
from the probable percentage of propane or butane in a mixture when the
combustion analysis indicates only methane and ethane. In most cases the
partial pressures of the propane and butane will also be so low that errors in
molecular volumes due to their presence can be disregarded'.
Methane conforms so closely to the gas laws that no deviation from the
given ,molecular volume need be made for different partial pressures.
The proper equations to use can be determined only from the partial pressures obtained from the analyses. For the purpose of determining the approximate percentage of ethane the theoretical equations can be used.
Application of the Use of Corrected Equations to the Analyses of Natural Gas
and Other Gas Mixtures
Although the combustion analysis does not show an accurate distribution
of hydrocarbons in a natural gas mixture, it does show the true total paraffin
content. The heating value calculated from such an analysis is also correct.
Determinations by the slow combustion method show that the natural gas
supplied to Pittsburgh from the Appalachian fields contains about 83% CH 4,
16% C 2H s, and 1% nitrogen. The proportions vary from time to time during
the year.
This gas is almost all methane but contains small amounts of ethane and
propane and some of the higher homologues.
A typical analysis with the calculation from the analytical data is given
herewith:
TYPICAL ANALYSIS A="D CALCl:LATION

Burette Readings (ml.)

Sample taken. , ............. , , , . , .... , ' , . , ' . , , . , , . , , .. ' 30.70
Volume after CO, absorption ..... , ..................... , 30.70
Portion taken for combustion ............................ ,30.70
Oxygen added.... . . . . . . . . . . . . .
. . . . . .. . . . . . . . . . . .. 74.85
Total volume. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 105.55
Volume after burning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 42.30
Contraction .......................................... , 63.~5
Volume after CO 2 absorption. . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.20
Carbon dioxide ..... , . '.' ..'. . . . . . . . . . . . . . . . . . . . . . . . . . . .. 35.10

CH 4 and C 2 Hs are calculated from the theoretical equations as follows:

Let x = methane,
and y = ethane.
Then 2x+2.5y = total contraction,
and x+2.0y = C02 produced.
C 2 Hs = 15.1 %.
. CH 4 =84.1%.
Total paraffins = 99.2%.

2414

GAS ANALYSIS

They are calculated from the corrected eql.!ation as follows:

I
Let x = methane,
and y=ethane.
Then 2.004x+2.5y=total pontraction,
and O.996x+2.0y=C0 2 produced.
C 2 H 6 = 15.7%.
I
CH 4 =83.1%.

Total paraffins = 98.8%.

Gasoline content is determined according to the method on p. 2408 .
. BEAM TEST FOR DETERMINING THE PRESENCE OF
SUSPENDED MATTER IN GASES 214

Tyndall demonstrated that dust particles are responsible for the visibility
of rays of light, such as the sun beam or the beam of the searchlight. He did
not make clear the best conditions enabling one to see the beam produced by
minute amounts of suspended matter in purified gases, nor did he show that a
gas free of dust may still produce a beam in presence of liquid mist. \The
opinion commonly current regarding the conditions necessary to see the beam
produced by light rays in purified gas, is that the rays s~ould be parallel and
Beam Test for Suspended Matters in Gases

Ughf_
(Inlef

FIG. 383.

should be viewed with the visual axis at right angles to this beam. Observing
these conditions the author 214 failed to see the bea'm using direct sunli~ht, bqt
discovered that a cone of light from an arc light gave a-distinct beam. This
led to the discovery that the beam is best viewed by glancing towards the source
.
of light. Accordingly the apparatlJs shown in Fig. 383 was designed.
For the examination of a corrosive gas, it is advisable to have the chamber
of the camera made of a material not affected bY-that gas. For example, in
214

Contributed by Wilfred W. Scott.

--

GAS ANALYSIS

2415

examining purified contact gas a lead chamber is recommended. The interior

of the chamber is blackened with dull paint that is a poor light reflector. In
the apparatus shown, circular glass discs close the ends of the tubes" a," b "
and c." The gas examined enters near the eyepiece" b " and passes out at
the opposite large end. The ray of light is directed through a small opening,
either circular (diam. 1 em.) or a narrow slit, located at " a " and passes through
the arm of the tube and out through a glass wintlow at "c." The idea is to
avoid light reflection in the chamber, as this would cause a glow in the tube
and greatly reduce the sensibility of the test.

FIG. 384.

Apparatus for Ascertaining the Pressure of Suspended Material in a Gas by

Means of a Beam of Light.

The beam caused by liquid mist alone appears as a uniform pencil or band
of light, whereas dust particles produce a sparkling effect 'br irregular flashes
of light: The beam is more intense with the gas in motion. Within certain
limits the intensity of the beam is directly proportional to amount of suspended'
matter in the gas.
If the gas is dry the beam is completely removed by passing the gas through
a three-inch mat of loosely carded purified asbestos. If the gas, thus purified
of solid matter, is passed through a liquid the beam again becomes apparent.
From this it is evident that the test must be applied to dry gaS'if it is desired to
ascertain the presence of solid matter in suspension.
Fig. 387 shows the apparatus set up for ascertaining the effectiveness of a'
scrubber in a quantitative determination of the solid and liquid suspended
matter in a gas. The apparatus is so arranged that comparisons may be made
of the unfiltered gas, by means of a bypass with that of the filtered or purified
gas. Advantage is taken of the difference in pressures at A and B, so that a
continuous sample may be taken of the gas without the necessity of an aspirator.
A hand arc light shown on the left furnishes the rays of light which enter at " a "
and pass out into the air at "c." The beam is viewed through the eyepiece
"b." The meter measures the vol~me of gas aspirated.
DETERMINATION 'OF SOLID AND LIQUID IMPURITIES IN A GAS

215

Solid or liquid impurities suspended in a gas may be removed quantitatively

by passing the gas through a filter of loosely carded dry asbestos. A filter two
216 Contributed by Wilfred W. Scott.

2416

GAS ANALYSIS

inches thick and lU inches in diameter is effective in removing suspended

matter completely with gas passing through it at the rate ot 2 to 3 liters per
second (5 to 7 cubic feet per minute). It is essential that the asbestos be free
from condensed moisturp- as a soggy filter is not effective for Igas filtration.
Determination of Sulfuric Acid Mist in Gas.-The gas is aspirated through
a filter of well-carded, blue fibre, purified asbestos, the asbestos extracted with
water and the extract titrated with standard alkali.
Apparatus.-An ordinary lU inch filter tube funnel (see Fig. 386) with
long stem, to permit a sample being drawn directly from the flue may be used.
If the gas is supersaturated with moisture it is advisable to use the form o'j
apparatus shown in Fig. 385, the bottle serving as a condensing chamber to
avoid water condensation on the asbestos. Occasionally it is necessary to
pass the gas through two such chambers to remove the condensed moisture
from warm moist gases.
Olltlef..

, FIG. 385. Apparatus for Determining Solids

and Liquid Mist in Wet Gases.

FIG. 386.

Filter for Gases.

The funnel is packed first with a layer of glass wool, followed by a-two-inch
layer of well-carded, purified, blue asbestos, and finally a second layer of glass
wool. The glass wool prevents the asbestos plugging the inlet and exit openings. Perforated disks may be used in place of the glas~ wool.
The asbestos is the blue African form. White asbes.tos will not give accurate results as this is attacked slightly by dilute sulfuric acid and decidedly
so by strong sulfuric. The blue form is not appreciably attacked so that the
acid may be extracted as free acid by water. The asbPJltos is purified by digesting with nitric acid, followed by a second digestion ,~it4 hydrochloric acid.
The material is now thoroughly washed to remove the..acids and is dried. The
dry asbestos is placed in wide mouth bottles until required. Before using it is

/'

GAS ANALYSIS

2417

carded or fluffed out by means of a pick and then placed loosely in the funnel
to be used in' the test.
Impurities in Gases
The Test.-The gas is aspirated through the filter at the rate of about five
cubic feet per minute, the gas being measured by a dry gas meter following the
apparatus. When a volume of 10 to 100 cubic feet of gas has been aspirated,
the amount 'depending upon the amount of acid mist present in the gas, the
asbestos filter is, placed in a beaker. (If 80 2 is present in the gas, as is the case
in contact sulfur gases, air is ,first passed through the asbestos to ;remove
the 80 2 .) The tube is rinsed out, the rinsings added to the asbestos. The.
contents of the beaker is now transferred to a two-inch Buechner funnel in
position on a half-liter capacity suction bottle. The ,vater is drawn out of the

FIG. 387. Apparatus for Determining Solid or Liquid Impurities in Gases.

'\

asbestos and the filtrate passed through'the mat, thus formed, to remove the
asbestos fibers. The mat is washed with small portions of pure water, until
free of acid and the extract, transferred to a beaker, is titrated with standard
alkali. N /20 alkali has been found to be satisfactory. . Methyl orange is used.
as an indicator. Determination of Arsenic in Gas.-The test is the same as the one described
for determining sulfuric acid. An aliquot portion of the water extract is taken
for a Gutzeit determination of arsenic. The Gutzeit method is described in
full in the chapter on arsenic, Volume 1. Arsenic in purified contact gas is
carried in the acid mist.
Determination of Chlorides.-A portion of the water extract, obtained by
the procedure described under the test for acid in gas and filtered through
paper if necessary, is examined for chloride by the turbidity test by addition of
nitric acid followed by silver nitrate reagent. Comparison is made with a
standard using either Nessler tubes or a calorimeter. If the amount of chloridc
is tOQ great for a turbidity test a gravimetric determination should be made,
or the aliquot portion taken should be sufficiently diluted with water to enable
the turbidity test to be made. The standard should contain the same amount
of sulfuric acid as the sample. The extract should be free of asbestos fiber.
Asbestos
Blue-African
White-Canadian
White-Italian
2

Si0 2
49.0
40.3
55.9

Composition of Asbestos
MgO
CaO
FeO
1.2
17.8

2.7
43.4
20.3

24.4
.9

AhG a
11.9
2.3
4.3

MnO
Ll

H 2O
13.7

2418

GAS ANALYSIS
DETERMINATION OF NITROGEN BY THE NITROMETER 216

The nitrometer, Fig. 388, consists of a gas-generating bklb fitted at the top
with a two-way cock leading to a dissolving cup and a gas-((xit tube, and which
has at the bottom a connection for a rubber tube leading tol a leveling tube, the
whole being filled with mercury to a level just below the uppJr cock; a cylindrical
glass measuring tube graduated from 0-100 m!. connected to a leveling tube

~
~

"
.g'"

Holder for
Gas Generating Bulb

Clamp
-{

-a

Clamp

"

~C/amp

CIamp
::t

.l2.

"

-a

.~

130

'"

-0

-~

go

"

-.I

.rr"

3>~ -FIG. 388.

through a T-tube leading to an equilibrium tube: The latter is shaped"like an

inverted 100-ml. pipette graduated downward below the bulb from 100-130 ml.
The whole system is filled with mercury so that the measuring tube may be
completely discharged by raising the leveling tube.
;'
Adjusting the Equilibrium Tube.-The volume of 100 ml. of dry air at
0 0 C. and 760 mm. under the temperature and barometric conditions prevailing
at the time is calculated, 3 drops of 98% sulfuric acid are drawn ihto the tube
and the level of the mercury fixed in accordance with,t~e calulation. The
_J-

216

Joyce and La Tourette, J. Ind. Eng. Chern., 5, 1017 (1913).

GAS ANALYSIS

2419

cock of the tube is then closed and sealed with melted paraffin. The volume
in this tube may be confirmed at any time by opening the measuring tube and
adjusting the leveling tube so that the level is the same in the three tubes. The
volume of gas in the equilibrium tube is then read and compared with the
calculated volume at the time, a correction in the subsequent gas readings in the
measuring tube being made accordingly.
Nitrogen in Potassium Nitrate.-Approximately 0.4 gram potassium nitrate
is placed in a weighing tube, dried two hours at 110 C., desiccated twenty-four
hours over sulfuric acid and weighed exactly by difference into the cup of the
gas-generating bulb. This is dissolved in 9 m!. 95% sulfuric acid, added through
a siphon thistle tube, entering the top of the dissolving cup through a tightfitting rubber stopper. When the salt is dissolved it is drawn into the gasgenerating bulb and followed by two washings 1.5 m!. each of 95% acid. The
bulb is then shaken with a motion normal to its long dimension until the volume
of gas determined by a rough paper s()ale pasted on the leveling tube becomes
constant, this operation taking from three to five minutes.
The gas is passed into the measuring tube, and after standing five minutes
the leveling and measuring tubes are so adjusted that the level in the equilibrium
tube reads 100 ml. and is the same as the level in the measuring tube. The
reading of the latter is then taken. As tJ'le temperature and barometric condi
tions, insofar as they affect the measured volume of the gas, are automatically
compensated by the equilibrium tube, and as the gas is washed with sulfuric
acid and is, therefore, dry, the percentage of nitrogen may be calculated directly,
correcting only for the calibration of the equilibrium and measuring tubes.
Seventeen determinations made when the room temperature ranged from
20-28 gave 13.71% nitrogen, the theoretical being 13.84%.
The nitrometer gives reliable results within 0.02% in nitrate nitrogen, using
0.4-0.5 gram sample. It is not available for the determination of nitrogen in
celluloid or other substances containing carbon ring compounds which appear
to prevent the complete liberation of nitric oxide in the presence of sulfuric
acid and mercury.
Besides the corrections for calibration and standardization of the nitrometer
in accordance with temperature and barometer, the gas readings should be
corrected for solubility of nitric oxide, which diminishes when the temperature
goes above 28 C., and the formation of other gases by the breaking up of the
'cellulose molecule which increases.
The following table gives the algebraic sum of these two corrections for
temperatures ranging from 20 to 35 C., for nitrocellulose.
Temp.

Ml.

Temp.

Ml.

20,0-27.5, ...........
28.0 ................
29.0 ................

+0.90
+0.74
+0.34

30.0
'30.5
31.0-35.0

-0.14
-0.70
-0.94

GAS ANALYSIS
COMPRESSED GASES AND THEIR HANDLING

The following gases are available commercIally:

Acetylene in acetone (prestolite)
Ammonia
Carbon dioxide
Chlorine
Ethylene
Hydrogen
Hydrogen Sulfide
Methane
Nitrogen
Nitrous oxide
Oxygen
Sulfur dioxide

217

218

Argon
. Helium
Krypton
Neon
Xenon
Butane, Isobutane
Butylene
Ethane
Propane, cyclo-propane
Propylene
Ethyl and methyl chlorides
Phosgene

They are shipped in cylinders of special steel, of which seven si~es are
available, fitted with valves and safety devices, and comply with the regulations
of the Interstate Commerce Commission and its Bureau of Explosives. The
valve stems used with corrosive gases are of Monel metal (68% nickel, 32%
copper) to minimize the gas attack: for ammonia, no copper or copper containing alloy is permissible, as it is attacked energetically. Cylinders containing
ammonia and chlorine are made often with siphon valves to enable the liquid
gases to be dispensed from them. To prevent the possible mixing of oxygen
or nitrous oxide with combustible "gases, the cylinders containing these (non
combustible) gases are fitted with right-hand' threads, whereas the ot~s are
fitted with left-hand threads. The table below shows this.
Table Showing the Cylinders for Different Gases
Right-hand

Ammonia
Carbon dioxide
Nitrogen
Nitrous oxide
Oxygen
Sulfur dioxide
Gases containing oxygen

Thread

Left-hand

Air
Blau gas
Chlorine
Hydrocarbons and their chlorides
Hydrogen
.'
Phosgene
Combustible gases

When used, these cylinders are usually connected with a gage giving the
original pressure, a reducing valve, another gage, giving the working pressure,
and then a needle valve. Gages showing the rate of flow a.v.d the capacity of
the different cylinders at different pressures can also be had. Before coupling
up this apparatus, the valve on the cylinder itself should always be tried by
opening it a little, and then closing quickly, to make sure it does not stick; any
foreign matter in the valve will be blown out. Cllre should be taken to insert
suitable washers where necessary. Except with oxygen, a little grease,.soap or
oil may be used on the fittings or washers. Great care must be used to exclude
oil with oxygen fittings, and new or repaired oxygen gages or fittings must be
washed out with carbon tetrachloride until the extract leaves no residue on
evaporation: dry air should tnen be drawn through them until. the solvent is
evaporated. Failure to do this has resulted in serious trolJble. To avoid
217 Booth, J. Chern. Education, 7, 1249 (1930).
.:..:218 Circular Ohio Chemical and Mfg. Co., Cleveland: -' _

GAS ANALYSIS

2421

explosions, a reducing regulator for any other gas must not be used with oxygen
or nitrous oxide, and the converse of the statement is equally true. Special
fittings to permit the use of apparatus with right-hand threads with left-hand
fittings and vice versa had best not be made or acquired, as this fitting has
caused fatal accidents.
As has been already stated, cylinders of compressed gases should be handled
with care: attention has been called to two fatalities caused by dropping
cylinders of compressed acetylene, a distance of about a foot.
APPENDIX
, REAGENTS AND TABLES

The reagents used in gas analysis, particularly in the absorption apparatus,

are comparatively few and easily prepared.
Hydrochloric Acid, sp.gr. 1.10. Dilute" muriatic acid" with an equal
volume of water. In addition to its use for preparing cuprous chloride, it finds
employment in neutralizing the caustic solutions which are unavoidably more or
less spilled during their use.
Fuming Sulfuric Acid.-Saturate "fuming sulfuric acid" with sulfuric
anhydride. Ordinary sulfuric 'acid may be used instead of the other; in this
case about an equal weight of sulfuric anhydride will be necessary. Absorption
capacity, 1 m!. absorbs 8 m!. of ethene (ethylene).
Acid Cuprous Chloride.-The directions given in the various textbooks
being troublesome to execute, the following method, which is simpler, has been
found to give equally good results: Cover the bottom of a two-liter bottle with
a layer of copper oxide or " scale" % in. deep, place in the bottle a number of
pieces of rather stout copper wire reaching from top to bottom, sufficient to make
a bundle an inch in diameter, and fill the bottle ,yith common hydrochloric acid
of 1.10 sp.gr. The bottle is occasionally shaken, and ,when the solution is
colorless, or nearly so, it is poured into the half-liter reagent bottles, containing
copper wire, ready for use. The space left in the stock bottle should be immediately filled with hydrochloric acid (1.10 sp.gr.).
.
By thus adding acid or copper wire and copper oxide when either is exhausted, a constant supply of this reagent may be kept on hand.
The absorption capacity of the reagent per m!. is, according to Winkler, 15
m!. CO; according to Hempel 4 ml. The author's experience with Orsat's
apparatus gave 1 ml.
Care should be taken that the copper wire does not become entirely dissolved and that it extend from the top to the bottom of the bottle; furthermore
the stopper should be kept thoroughly greased the more effectually to keep
out the air, which turns the solution brown and weakens it.
Ammoniacal Cuprous Chloride.-The acid cuprous chloride is treated with
ammonia until a faint odor of ammonia is perceptible; copper wire should be
kept in it similarly td the acid solution. This alkaline solution has the advantage that it can be used when traces of hydrochloric acid vapors might be
harmful to the subsequent determinations, as, for example, in the determination
of hydrogen by 'absorption with palladium. It has the further advantage of not
soiling mercury as does the acid reagent.
Absorption capacity, 1 m!. absorbs 1 m!. CO.

2422

GAS ANALYSIS

Cuprous c1-Iloride is at best a poor reagent for the absorption of carbon

monoxide; to obtain the greatest accuracy where the reagent has been much
used, the gas should be passed into a fresh pipette or cuprous sulfate pipette
for final absorption, and the operation continued until two consecutive read.
ings agree exactly.
Acid Cuprous Sulfate.-The chemists of the U. S. Steel Corporation make
the solution as follows: The cuprous oxide is prepared in the wet way, as that
obtained by the dry method gives a solution that acts too slowly. 100 grams
copper acetate are dissolved in a liter of water, filtered if necessary, boiled and
reduced with 60 grams of glucose in 400 ml. of water, which is added to the
boiling solution. The glucose is prepared by inverting 60 grams of cane sugar
in 400 m!. of water with 3 drops sulfuric acid and boiling 5 minutes. The
excess scid
l1.euks.liced wit'h clJ.leium CSrbOIlo.te, the precipitate allowed t(l
settle and filtered off. When the blue color has somewhat faded, the cuprous
oxide is allowed to settle, washed with water by decantation once with alcohol,
filtered and dried at 90 to 100 0 or better in a vacuum desiccator and kept in!:L
tightly stoppered bottle.
Two hundred mI. of c.p. sulfuric acid are-diluted with 25 ml. of water in the
usual way and cooled, 20 grams of the cuprous oxide are mixed with it in small
quantities by grinding in a porcelain mortar until the oxide is in suspension:
25 grams of beta naphthol are gradually added to the mixture in the same man
ner as the oxide. The mixture is filtered through glass wool kept above 150
to keep the naphthol in solution and well protected from the air. The solution
is best fitted for use with the pipette, rather than the burette type of apparatus
as the Elliott.
Bromine Water.-Dilute satul'ated bromine water with twice its volume of
water. This then contains about 1% of bromine:
Hydrogen.--A simple- and effective hydrogen generator can be made by
joining two 6-in. calcium chloride jars by their tubulatures. Pure zinc is filled
ill: as far as the constriction in one, and the mouth closed with a rubber stopper
carrying a capillarY,tube.and a pinchcock. The other jar is filled with sulfuric
acid 1: 5 which has been boiled and cooled out of excess of air. The mouth
of this jar is closed with a rubber stopper carrying one of the rubber bags used
on the simple pipettes.
Standard Lime Water is prepared by adding 6.3 mI. of limewater, saturated
at the ordinary temperature with Ca(OH)2, to a liter of distilled water prepared
as follows. To a liter of distilled water add 2.5 m!. phenolphthalein (0.7 gm.
in 100 ml: .50% alcohol) and titrate this with saturated lime water until a faint
pink colot/is obtained which is permanel'tt for one minute.
Mercury.-The mercury used in gas analysis should be of sufficient purity as
not to " drag a tail" when poured out from-a clean vessel. It may perhaps be
most conveniently cleaned, e,,"cept from gold and silver, by the. method of
J. M. Crafts, which consists in dl:'awing a moderate stream of air through the
mercury contained in a tube about 3 ft. long and 17i ins. internal diameter.
The tube is supported in a mercury-tight V-shaped trough, of size sufficient
Ito contain the,metal if the tube breaks, one end being about 3 ins. higher than
the other. Forty-eight. hours', pa~sage of air is sufficient to purify any ordinary
/ amalgam. The mercury may very well be kept in a large separatory funnel
u1li1er .1l layer of 8tro~~g sulfuric 3(Jia/

is

GAS ANALYSIS

2423

Or Meyer's method 219 may be used. A separatory funnel is used to hold

the mercury. The delivery tube of the funnel is slightly narrowed 0.5 cm. from
the lower end. Over this side is bound with twine a piece of rather closely
woven muslin. The mercury is allowed to flow through the cloth into a solution
of mercury nitrate contained in a tall cylinder, with stopcock at the lower end.
The tip of the funnel with the muslin dips under the surface of the cleaning
solution. The purified mercury is drained off from the bottom of the cylinder.
It can also be purified except from traces of zinc by distillation.
Mercury Poisoning. 2 2O_Very few are aware of the danger to which they are
exposed from the vapor of the metal when in ordinary use. This is undoubtedly
due to the efficient ventilation of educational laboratories. Air at ordinary
room'temperature contains when saturated 12 mg. mercury per cubic meter,
and 0.01 to 0.001 mg. are found in small closed rooms. This can be detected
by a special apparatus using selenium sulfide. 221
The symptoms of mercurial poisoning are intermittent headache, nervousness, catarrh, weakening of the senses of smell and hearing, inflammation of
the eyes,. sore spots on the tongue and throat, and reddening and soreness of
the gums with toothache. Further symptoms are mental weariness, shown in
loss of memory and acumen, and general lassitude and lack of "pep." No
medication appears to be of any use, only fresh air and time bring recovery. It
seems to act as a cumulative poison, most of the mercury being retained in
the lungs.
To avoid poisoning, no spilled mercury should be allowed to lie around in
trays, sinks, or on the floor, which should be tight-best covered with linoleum:
containers should be stoppered, and work with open mercury done under the
hood, unless forced ventilation be e~ployed. It is thought the sufferings of
Faraday in later years were due to mercurial poisoning.
Palladous Chloride.-Five grams palladium wire are dissolved in a mixture
of 30 m!. hydrochloric and 2 m!. nitric acid, this evaporated just to dryness on
a water bath, redissol ved in 5 m!. hydrochloric acid and 25 m!. water and
warmed until solution is complete. It is diluted to 750 m!. and contains about
, 1 %' of palladous chloride. It will absorb about two-thirds of its volume of
hydrogen.
Phosphorus.-Use the ordinary white phosphorus cast in sticks of a size
suitable to pass through the opening of the tubulated pipette.
Potassium Hydroxide.-(a) For carbon dioxide determination, 500 grams
of commercial hydroxide are dissolved in 1 liter of water.
Absorption Capacity.-One mI. absorbs 40 m!. CO 2
(b) For the preparation of potassium pyrogallate for special work, 120 grams
of the comme:z:cial hydroxide are dissolved in 100 m!. of water.
Potassium Pyrogallate.-Except for use with the Orsat or Hempel apparatus, this solution should be prepared only when wanted. The most convenient
method is to weigh out 5 grams of the solid acid upon a paper, pour it into a
funnel inserted in the reagent bottle, and pour upon it 100 m!. of potassium
'hydroxide (a) or (b). The acid dissolves at once, and the solution is ready for
use.
219 J. H. Hildebrand, J, Am. Chern. Soc., 31~ 934 (1909).

Dennis, Ind. Eng. Chern., 18, 1205 (1926).

Made by the General Electric Co. See Nordlander, Ind. Eng. Chern., 19, 518
(1927).
220

221

GAS ANALYSIS

2424

If the percentage of oxygen in the mixture does ndt exceed 28, solution (a)
may be used;222 if this amount be exceeded, (b) must be employed. Oth,rwise
carbon monoxide may be given off even to the extent. bf 6%.
Attention is called to the fact that the use of potassit:m hydroxide purified
by alcohol has given rise to erroneous results.
.
Absorption Capacity.-One ml. absorbs 2 ml. O2.
Sodium Hydroxide.-Dissolve the commercial hydroxide in thre~ times its
weight of water. This may be employed in all cases where solution (a) of
potassium hydroxide is used. The chief advantage in its use is its cht;apness.
Sodium pyrogallate is, however, a trifle slower in its actIOn than the cor- .
responding potassium salt.
Table 1 223
CONSTANTS OF THE GASll;S

Gas

I NH, ...
NO, ....
Ch .....
HC!. ...
(CN), ..
II HCN ...
SO, ....
H,S ....
SiF .....
CO2 ....
PH, ....
III AsH, ...
SbH, ...
02 .....
IV 0,
.....
v CO ....
VI NO ....
H, .....
CH .....
C,H, ...
C,H....
VII N'O
....
N, .....
COS ...
air .... .

Sp.Gr.
to
Air

Wt.(
Liter in
Grazus,
Calcu0,
lated 760mm.
0.5971
1.5906
2.4494
1.2684
1.7968
0.9359
2.224
1.1769
3.60
1.5290
1.1829
2.695
4.360
1.10523
1.624
.9671
1.0367
.0694
.5545
.9056
.9684
1.5301
.9673
2.075
1

Sp.Ht.(
Cu.Ft. Press.

at

Atmos.
Press.,
62 F.,
C.P.

----0.7710
.0234
2.0563
3.214
1.6392
2.3229
1.2096
2.9269
1.5392
4.684
1.9777
1.5293
3.4944
5.6
1.4290
3.0000
1.2504
1-3402
.08985
.7168
'1.173
1.2609
1.977
1.2505
2.727
1.293 I'

.0183
.0260

at

20 C.
in
Lbs.
125

.0243

.0182

1800

.0241

./

Iisoi. (If Gas in 1 Liter

of Water in ml.

B.Pt.

Gas in
Air

c.

At 0

702,000

505,000

2,229
444,000
4.500

4.3 --46.0

12.5 -75.0
4.15-75.0
4.9 -15.4
1.5 -80.5
3.2 -34.0

33.1
22
- 34.6
- 85
- 20.5
26
- 10
- 60.2
- 90
- 78.1;
- 87.4
- 54.8
- 17
-183
-112
-190
-150.2
-252.77
-161.4
- 83.6
-103.9
- 89.7
-195.8
- 48.0

79,789
39,374
4,370
2,905
Completely
8i4
713
1,
2,500?
2

49.61

31.71

35.37;'
73.81
21,53
55.63

23.19
47.06
18.37
33.08
103
122
67.00
- 15.98
568
~ 17.0

226
130.52
23.59
1,348
24.7

i\1.Pt.

C.

At 20 C.

16.1)'-26.8 1,176,000

97
620
73
48
150
50
870

.0180
.0183
.0181
.025
.024
.029
.0256
.0182

Explosivity

---

77.7

--1029

-111
- 34.4

-14

- 72.7
- 82.9
- 77
- 56
-132.5
-113.5
-83
-218.4
-251
-212.6
-160.9
-259.2
-184
- 81.8
-16\).4
-102.4
-209.8
-138

Clowes, Jour. Soc. Chern. Industry, 15, 170 (1896).

See also in this connection Bur. Mines Bulletin 279. Coward & Jones, "Limits of
Inflammability of Gases and Vapors." Also Bulletin 30, "Explosibility of Methane and
Natural Gas."
222

223

2425

GAS ANALYSIS

Table 2
TABLE SHOWING THE TENSION OF AQUEOUS VAPOR AND ALSO THE "VEIGHT IN GRAMS
CONTAINED IN A CUBIC METER OF AIR WHEN SATURATED FROM 5 TO 30 C.224

Temp.

Tension,
mm.

Grams

Temp.

Tension,
mm.

Grams

Temp.

Tension,
mm.

Grams

5
6
7
8
9
10
11
12
13

6.5
7.0
7.5
8.0
8..5
9.1
9.8
10.4
11.1

6.8
7.3
7.7
8.1
8.8
9.4
10.0
10.6
11.3

14
15
16
17
18
19
20
21
22

11.9
12.7
13.5
14.4
15.4
16.3
17.4
18.5
19.7

12.0
12.8
13.6
14.5
15.1
16.2
17.2
18.2
19.3

23
24
25
26
27
28
29
30

20.9
22.2
23.6
25.0
26.5
28.1
29.8
31.5

20.4
21.5
22.9
24.2
25.6
27.0
28.6
29.2

Table 3
SPECIFIC HEATS

B.t.u.

PER POUND OF GASES AT CONSTANT PRESSURE C.p.

Volumetric

Air ............................. .
Carbon dioxide ........................ .
Carbonic oxide. . . . . . .. . .............. .
Hydrogen ............................. .
"Illuminants" ........................ .
Methane .............................. .
Nitrogen .............................. .
Oxygen .............................. .
Aqueous vapor ........................ .

0.241
0.210
0.243
:).420
0.400 22.
0.593
0.247
0.217
0.480

225

0.019
0.027
0.019
0.019
0.040
0.027
0.019
0.019

The "volumetric" specific heat is the quantity of heat necessary to raise the temperature of 1 cu. ft. of gas from 32 F. to 33 F.
Gases are about 10 times as soluble in kerosene, Z. phYHik. Chern., 1, 70 (1887).
H. L. Payne, Jour. Anal. and Applied Chern., 7, 233.
226 Ethylene.
224

'26

GAS ANALYSIS

2426

Table 4
CALORIFIC POWER OF VARIOUI", GASES m IN BRITISH THERMAL UNITS PER CUBlc
AT 30 IN. PRESSURE
\

Name

Symbol

Hydrogen ...............
Carbonic oxide ..........
Methane ................
Illuminants 228
Ethane .................
Propane ................
Butane .................
Pentane ................
Hexane 229 ............ '"
Ethylene ........... , ....
Propylene .......... , ....
Benzene ................
Acetylene ...............

60 F. Initial 231

325
323
1008
1900 232
1778
2513
3269
4022
4744
1583
2340
,3741
1483

H2
CO
CH,
C 2H 6
C aH8
C.H10
C.H l2
C 6H14
C 2H,
CaH.
C 6H.
C 2H 2

32 F. Initial
32 F. Final

345.4
341.2
1065.0
2000.0
1861.0
2657.0
3441.0 .
4255.0
5017.0
1674.0
2509.0
4012.0 \
1477.0

FoOT

Ignition Point
of.

1085 230
1200 230
1230
....
1140
.1015
. ...
.0

1400

1010 23~
940
....

788 230

Table 5
ENGLISH AND METRIC EQUIVALENTS

1 pound
=453.59 grams
1 cubic inch
= 16.39 ml.
1 cubic foot
=28.315 liters
1 B. t. u. per cu. ft. = 8.89 calorie per cubic meter
1 in. H 20 pressure = 1.87 mm. Hg pressure

1 calorie
=3.969 B.t.u.
1 cubic meterl
=35.31 cu. ft;
1 calo~e perl~u. meter =0.1124.B.t.u.
1 mm i Hg pressure

227 H. L. Payne, already cited.

228 Where obtained by cracking oil.
229 The chief constituent of gasoline.
230 Dixon and Coward, Proc. Chem. Soc., 26, 67.
231 Gas Chemists' Handbook.
232 Calculated.

= 0.535 in. H 20
pressure

"

GAS ANALYSIS

2427

Table 6
The following table assembled from compilations by Henderson and Haggard
shows the:

233,

DANGEROUS CONCENTRATIONS OF THE VARIOUS GASES IN PARTS/MILLION

.
'"~

'"
::;"

"'"
s

.!l

"':"
s

.~

"
.a
.a.... ""
~
]
u
:
:t:
- - - - - - - - - - - ---3.5
Least detectible odor ....
53
Least am't irritating to
eyes."" ...........
Least am't irritating to
throat ......... " .. '
Least .mount causing
coughing ............

1l

"
~
u

~ ~

00

~
.a....
0

233

.~

--

.d

'~
0

'if:

is
13
~

00

3-5

..:
]

"

!f~

~~

I:Q

300

20

408

15.1

62

8-12

1720

30.2

100

20

100

100

'Maj. concentration for

Yo-I hr. exposure ..... 300-500
Dangerous for short exposure ............. . 2500Rapidly f.tal for short
exposure ............

!!l0

.S
"

698

Maj. concentration for

prolonged exposure ...

I'

600
50

4500
30 min.
5000250
10,000

15002000
4000

1.0
4-

10
50

40- I~Ol}60 2500

1000

3
50-60

10

120- S(}150 250

3000

100150

39

10
100- 50200 100

300

500- 117- 400- 400- 11,000700 154 600 500 22,000

3000

775

2000

31304700

24,300 19,000

"Noxious Gases and the Principles of Respiration Influencing their Action."

2428

GAS ANALYSIS

Table 7
VOLUMETRIC FACTORS

1 ml. of the Different NjlO Solutions Corresponds to the-Following Volumes and 'Veights
of the Different Gases
KMnO~

Hel

Nllnn

~!US~03

AgNO.

a"
lUI.

CI2 . . ~

)'1_!::.

Ml.

l\H.

Mg.

Mg.

1.12 3.54fJ

CO 2 .

HCN .. ~.
HC!. .. '. 2,24 3.646
If,S.
1.12
NO.
2.231
N0 2 . . . .
2.24
NUs ..... 2.24 1.703
80 2 .
1.118

0 3 ....

1.704
3.00
4.60

IHt::.

1.12
2.24
2.24

2.702

2.24

3.646

2.24

2.24

3.20

l\H.

1\11.

--

--

Mg.

::\1l

Mg

I-

2.20

2.702
3.646
1.12

1.704

1.118

~.20

4.60

1.118 3.20
2.4

(CK), ..

1.120 2.6008 1.120 2.6008

Table 8
DATA ON CYLINDER>S FOR COl't.fPRESSED GASES
~'t)'\I:: ...

. ...

'_'_'

..

1<
c
J'
"""2 ---"
-- - -9
3X 4)1 4)1 4)1 5)1 8>'2

Diameter incll<'s .
'"
.
Height inches . ..
..... 12
Capacity cu. in .. . ..
.. Capacity
ft .. \. . .. , " -' . ..
Capacity cu. em .... ......... . -

cu.

..

16
90
.052
1474

18
142
.OR2
2326

20
177
.102
2900

,55
29
M
55
288
930 2380 2660
1.54
.166 0.537 1.375
4723 152.52 39032 43624

The smaller cylinders from E to AAA have no threads on them to couple to fittings:
they are connected to apparatus by a shnple yoke with a hose connection.

GAS ANALYSIS

2430

CORRECTION' FACTORS FOR GAS VOLUME-Continued

[Multiply observed volume of saturated gas by' factor to cartect to v6lume.of saturated
gas at 30 inches of mercury pressure (32 F.) and 60 F.235]
~
I

Total Gas Pressure-Inches of Mercury-Continued

Temperature

29.3
of.
99
98
97
96
95

29.4

29.5

29.6

29.7

29.8

29.9

30.0

30.1

30.5

--

.917
.920

.911
.914
.917
.920
.923

.914
.917
.920
.923
.926

.917
.920
.923
.926
.930

.920
.923
.927
.930
.933

.923
.926
.929
.932
.935

.926
.929
.932
.935
.938

.930
.933
.936
.939
.942

.933
.936
.939
.942
.945

.936
.939
.942
.945
.948

.935
.938
.941
.944
.946

~.G&8

.941
.944
.947
.950

.941
.. 944
.947
.950
.953

.945
.948
.950
.953
.956

.948 .951
.954
.951
.954 .957
.957 .960
.960 .. 963

.946
.949
.952
.951
.957

.949
.952
.955
.958
.960

.952
.955
.958
.961
.964

.956
.959
.961
.964
.967

.959
.962
.965
.968
.970

.962
.965
.968
.971
.974

.957
.959
.962
.965
.967

.960
.963
.965
.968
.971

.963
.96!)
.969
.971
.974

.970
.972
.975
.978
.981

.973
.976
.978
.981
.984

.976
.979
.982
.985
.987

.967
.969
.972
.974
.977

.970
.973
.975
.978
.980

.973
.976
.979
.981
.984

.977
:.979
;.982
.984
,.987

.966
.969
.972
75
/7
.980
.983
.985
.988
.990

.983
.986
.989
.991
.994

.990 .993
.987
.993 .996
.989
.995 .999
.992
.998 1.001
.995
.997 1.001 1.004

.980
.982
.985
.987
.990

.983
.986
,988
.991
.993

.986
.990
.993
.996 1.000
.992
.989
.996 ~.999 1.002
.992
.995
.998 1.002 L005
.994 .997 1.001 1.004 . 1.008
.997 1.000 1.003 1.007 1.010

.901
.904

.904
.908

.907
.911

.910
.914

.914
.917

.914
.916
.920
.922
.925

.920
.917
.920 ' .923
.923
.926
.926
.929
.929
.932

.925
.928
.931
.934
.937

.928
.931
.934
.937
.940

.932
.934
.937
.940
.943

.936
.939
.942
.944
.947

.940
.942
.945
.948
.950

.943
.946
.948
.951
.954

.947
.949
.952
.955
.957

.950
.953
.955
.958
.961

.953
.956
.959
.961
.964

.957
.959
.962
.964
.967

.960
.962
.965
.968
.970

.963
.966
.968
.971
.974

.970
.972
.975
.977
.980

.973 .976
.975
.979
.978 ' .981
.980 .984
.983
.986

.882
.885
.888
.891
.894

.885
.888
.8111
.894
.898

.888
.892
.895
.898
.901

.892
.895
.898
.901
.904

.895
.898
.901
.904
.907

.898
.901
.904
.907
.910

.007" .911 . .914

83
83
87
83
85

.897
.900
.903
.906
.909

.901
.904
.907
.910
.912

.904
.907
.910
.913
.916

.907
.910
.913
.916
.919

.910
.913
.916
.919
.922

84
83
82
81
80

.912
.915
.918
.921
.924

.915
.918
.921
.924
.927

.919
.922
.924
.927
.930

.922
.925
.928
.930
.933

79
78
76
75

.926
.929
.932
.935
.937

.930
.932
.935
.938
.941

.933
.936
.938
.941
.944

74
73
72
71
70

.940
.943
.945
.948
.951

.943,
.946
.949
.951
.954

69
68
67
66
65

.953
.956
.958
.961
.964

64
63
62
61
60

.966
.969
.971
.974
.976

236

30.3 ,30.4

0.866 0.869 0.872 0.875 0.878 0.882 0.885 0.888 0.891 0.894 0.897 0.900 0.904
.869
.875
.879
.882
.904 .907
.888
.891
.872
.901
.885
.898
.894
.882
.891 .. 894
.907 .910
.872
.876
.879
.885
.904
.888
.901
.8iPt
.895
.910 .914
.876
.879
.882
.885
.888
.907
.898
.904
.9 1
.891
.914 .917
.879
.882
.885
.888
.892
.898 .901
.904
.907
.911
.895

94
93 .
92
91
90

1'7

30.2

-------- -- --------

Formula used: Correction factor

P = total gas pressure
E = vapor pressure of water at to
t=temperature of gas (0 F.).

(P-E) (60+459.4)

(t+459.4) (30-0.5170)

1.003
1.006
1.008
1.011
1.014

.966
.968
.97.1

~974

.977
.980
.982
.985
.988
.991

1.007
1.009
1.012
1.014
1.017

2431

GAS ANALYSIS
Table 10

EMERGENT STEM CORRECTIONS TO READINGS OF OUTLET-WATER THERMOMETERS FOR

DIFFERENT IMMERSIONS OF THERMOMETERS IN CALORIMETER

[Table applicable when temperature of inlet water is approximately equal to

room temperature]
Temperature
Rise of
Water

of.
0

Thermometer immersed to 30 F ..

Thermometer immersed to 40 0 F ..

Thermometer immersed to 50 0 F ..
Thermometer immersed to 60 0 F ..

10
15
20
10
15
20
10
15
20
10
15
20

Temperature of :Room

- - - - - - - - - ~I~ - - 50

60

70

100

+0.02 +0.03 +0.04 +0.05 +0.05 +0.06

+.04 +.05 +.06 +.07 +.09 +.10
+.06 +.07 +.09 +.11 +.13 +.15
+.01 +.02 +.03 +.03 +.04 +.05
+.03 +.04 +.05 +.06 +.08 +.09
+.04 +.05 +.07 +.09 +.11 +.12
+.01 +.01 +.02 +.03 +.04 +.05
+.02 +.03 +.04 +.05 +.07 +.08
+.02 +.04 +.06 +.07 +.09 +.11
+.00 +.01 +.02 +.02 +.03 +.04
+.00 +.01 +.03 +.04 +.05 +.06
+.00 +.02 +.04 +.05 +.07 +.09

This table is not applicable if the emergent portion of the stem includes an enlargement in the capillary.
Instead of using the above table, it will probably be somewhat more convenient to
make out a stem-correction table for the particular outlet-water thermometer that is to
be used with the calorimeter, the data for this separate stem-correction table being
interpolated from the above table.
Suppose, for example, the outlet-water thermometer to be used was one that was
immersed to the 300 F. mark on the scale, and a stem-correction table were wanted for
an 18 F. rise in.temperature, then from the above table we obtain the following stemcorrection table:
Stem Correction for Outlet-Water Thermometer No.[Table applicable when inlet water is approximately at room temperature, when
thermometer is immersed to the 300 F. mark, and when the temperature
rise is approximately 180 F.]
Inlet-water Temperature

Stem Correction

OF.
50
60
70

0.05
.06
.08

Inlet-water Temperature

OF.
80
90
100

Stem Correction
0

0.09

.11

.13

In the same way a table could be made out for any outlet-water thermometer' by
interpolation in the general table. The table so prepared would apply for the particular
point to which the thermometer was immersed and for the particular rise in temperature
with which the. observer had chosen to work.

CHEMICAL MICROSCOPY' 1
"

ADVANTAGES OF MICROSCOPICAL METHODS'

If the term "chemical analysis" is construed in a broad sense, it covers
determinations of properties and behavior as well as of composition. Most
purely chemical data must ultimately be translated into comparisons of purity,
costs, strength, or other aspects of the history' or future performance of the
material in question, but such interpretations are frequently indirect and
inconclusive. The great advantage of microscopical methods in analytical
chemistry lies in the directness of the information which they yield and the
rapidity with which it is obtainable. The ability to deal with minute amounts
of material is in most cases secondary to that of being able to observe the actual
physical condition of the sample and the phenomena which it exhibits, in a
vivid" close-up" rather than by inference.
A further advantage of the application of the mic.roscope to analytical
problems results from the fact that the sample itself is of necessity under direct
examination, so that in most instances unsuspected"and highly significant
additional information, which might other;vise have been missed, is brought
forth.

r
STUDIES OF SINGLE SUBSTANOES
As an aid to ordinary macroscopical analysis or: single substances the microscope is employed in various ways. Single substances can be examined for
purity and uniformity, and their physical state observed. Particle size and
shape of powdered materials may be ascertained, as factors .governing solubility, reactivity or suitability for pigments, fillers, abrasives, etc: Surface
character of coatings, such as paints, glazes, platings, or enamels, may explain
their failure better than data as to their constitution.
Identification of substances is rendered more positive by supplementing the
analytical information by a determination of characteristic structural or optical
properties, for which the microscope is the instrument par excellence. By
utilizing published descriptions or by comparison with authentic "knowns."
conclusive identification is frequently possible without the use of chemical
reactions and without the need of separating the material from its surroundings
or from a mixture.
I
1 E. M. Chamot, Professor of Chemistry, Emeritus, 9or~lf University, and C. W.
Mason, Professor, Chemical Microscopy, Cornell University. - .

2434

CHEMICAL MICROSCOPY

2435

Distinctive microscopical structure which serves as the basis for such

identifications is particularly advantageous in the case of materials that are of
highly complex chemical constitution, such as the plant and animal tissues of
foodstuffs, textiles and paper fibers. In such examinations the recognition is
bl!csed on characteristic morphological featurism supplemented in some cases
by staining reactions. Familiarity with the appearances of the substances
studied is essential; if in addition to identification the physical condition of the
fibers or the process used in their preparation is indicated, further valuable
information may be obtained.
Optical properties are a valuable means of identification, particularly in the
case of crystalline substances. The latter are not usually distinguishable by
form alone, but since some ten or twelve properties, about half of them I}umeri- '
cal constants and the rest" positive" (+) or "negative" ( - ), are determinable,
recognition is usually possible, provided the optical description of the substance
appears in the published tables or has previously been determined by the
analyst. These various optical tests serve for a classification which is analogous
to the groupings, subgroupip_gs, and successive reactions involved in qualitative
analytical schemes, and the identification based on them is no less conclusive
than that based on a series of chemical reactions.
Descriptions of all the common minerals are published in convenient tables 2
-an indispensable aid to the fullest analysis, for they permit the actual solid
phase to be determined and not merely the ions which it yields after a rather
violent" opening up" or solution process. In the case of inorganic salts the
published data are less complete and only partially tabulated,3 but the advantage of being able to ascertain the particular salt, hydrate, or allotropic
form in hand, rather than the anions and cations only, makes optical analysis
indispensable whenever this information is desired. Organic compounds are
as yet largely undescribed optically, but any properties available 4 are doubly
useful; in the absence of adequate data, checking the unknown against material
of known purity may prove time and labor saving, especially if the chemical
tests are inconclusive.
The actual examination of the optical properties of chemical substances is
preferably carried out on material which has been powdered, or better, recrystallized on a microscope slide.
,
The following properties of crystalline substances are determined if observable:5
2 Larsen and Berman, "Microscopical Determination of the Non-opaque Minerals,"
U. S. Geol. Sillvey Bull., 848 (1934); Winchell, "Elements of Optical Mineralogy,"
Part II (1927).
3 Winchell, "The Optic and Microscopic Characters of Artificial Inorganic Substances." (John Wiley and Sons, New York, 1932); Fry, "Tables for the Microscopic
Identification of Inorganic Salts," U. S. Dept. Agr. Bull., 1108 (1922); "International
Critical Tables," Vol. 1. (Refractive indices and crystal systems only.) Groth,
"Chemische Krystallographie," 5 Vols. (Data not tabulated, but very extensive.)
4 "International Critical Tables," Vol. I (1926), pp. 320-338.
Mayrhofer, "Mikrochemie der Arzneimittel und Gifte," 2 Vols. (1928). Groth, "Chemische Krystallographie."
.
5 For discussions of these properties and the methods of observing and describing
them, see Chamot and Mason, "Handbook of Chemical Microscopy," Vol. I (1938),
Chapter IX, especially pages 303 to 319. Winchell, "Elements of Optical Mineralogy,"
Part I (1928). Hartshorne and Stuart, "Crystals and the Polarising Microscope"
(Arnold, London, 1934).

2436

CHEMICAL MICROSCOPY

Phenomena of Crystallization.-Condition of crystals, as received. Action of

solvents. Effect of heating. Growth of crystals from solution OJ; from the melt. Evi1
dence of allotropic transformations or double salt formation.
Geometrical Form and Characteristic Angles.-Study and correlation 6 different
"views," of the simpler crystals, by means of drawings or modelsj of cork or wax.
Doubly Refractive Character.-Isotropic substances (cubic system, 01: basal views of
hexagonal and tetragonal systems) remain dark between crossed nicols, on rotating the
I
stage.
Anisotropic or d~ubly refrac~ive .substanc.es (all c.r~stal syste.mii\..except cubic) s~ow
"extinction" at 90? mtervals, with mtermedlate positIOns of bn~lhess. PleochrOism
is shown by some colored anisotropic substances, as a change in c or when the stage is
rotated, only one nicol being used. Strongly doubly refr<J,ctive subst nces show a change
of shading, with one nicol.
Positions of Extinction.-Parallel or symmetrical, if a prominent crystal edge or
diagonal is parallel to the cross hairs (which must represent the planes of vibration of the
polarizer and analyzer) when extinction occurs. Oblique extinction (possible only in the
monoclinic and triclinic systems) occurs if the position of extinction is oblique to the
cross hairs. The angle between the crystallographic edge or direction used for reference
and the nearest cross hair is called the" extinction angle."
.
Strength of Double Refraction.-The maximum difference of the values of refractivp
indices is a numerical measure of the strength of double refraction or "birefringence."
The polarization colors, which follow the same series asthose of thin films, give an approximate indication. For crystals of ordinary microscopic size and thickness, low order colors
(grays and straw-yellow to orange) indicate weak double refraction; brighter colors,
moderate to strong double refraction; and high order colors (pale pinks and greens, to
grayish white) strong double refraction.
Refractive Indices.-Two values are obtainable.for any anisotropic view of a crystal;
different views may show different pairs of values. After turning to the position of
extinction, the crystal is illuminated with polarized light (one nicol only) the plane of
vibration of which is known, and the tests given below are applied. The value for the
position of extinction 90 from the first is determined similarly. The highest and the
lowest values, for all possible views, are of particular interest.
Interference figures serve to indicate whether the substance is uniaxial (hexagonal
or tetragonal systems) or biaxial (orthorhombic, monoclinic, or triclinic_systems) ,
and whether its double refraction is positive (+) or negative (-). The plane of the
optic axes, and the angle between them, are also indicated. A discussion of these
phenomena and tests is beyond the scope of this chapter.
I

For the above determinations of properties a che'inical or petrographic

microscope is necessary; it should be provided with polarizer and analyzer nicol
prisms, graduated center able revolving stage, cro$shaired eyepieces, and compensators such as a" 1st order red" plate or a quartz wedge; a condenser for use
in the examination of interference figures is also,desirable.
Refractive index determinations 6 require a series of standard liquids of
known indices. The sample is compared with these by immersing-microscopic
crystals in the various liquids and applying tests to ascertain whether its index
of refraction is higher or lower, until one of the standards is found to match it.
In determining the refractive indices of doubly refractive substances illumination with polarized light is essential, and the orientation of the crystal with
respect to the direction of transmission and of vibration of the light must be
specified, especially if the values are to be compar~d with those commonly
designated as E and w for uniaxial, and ex, {3, and 'Y for biaxial substances.
The tests employed for comparing refractive indices are very simple and
rapid: Particles of the unknown are immersed in a dr~p of one of the series of
6

373.

See "Handbook of Chemical Microscopy," Vol. I, Cl1apter XI, especially pp. 366,

CHEMICAL 1\lIGIWSCOI'Y

2437

standard liquids on an object ~Iide, and c(wcrcd with a cover glass . \Vith the
diaphragm of the ~ub~tagc conden~r c'o~cd 8~ raf 8!-; prr.::sible, a naTrUW ~right
line (Il llecke line") appears to bound the graill.!:!; this moves toward the sub-

FlO. 389. The Becke Line. Sodium Chloride (n=1.544) in a Liquid of n=1.540.
A-on focusing upward slightly; B-on focusing downward slightly.

stance of higher refracth"e index as the microscope is focussed upward, and

toward the less refractive material as the focus is lower (sec Fig. 389). If the
condens~r, with diaphragm fully opened, is focussed to give strongly convergent
illumination and one edge of the diaphragm aperture is covered by the finger

'..-"II: ()
~

I'

~""'a
.~.

I
;

A.

-.

-FIG. 390. Axial and Oblique Transmitted Illumination. The upper crystal has
a higher re(ractive i.ndex tha.n the mQunting medium; the lower crystal, a,less rerra~tive
index.

or a piece of"paper, one side of the field will appear shaded. The boundary of
the particles will also be darker on G~e 8ide; if this shading is on the same side
as the shadow. in the neld, the particle has a higher refractive indexthan the
surrounding liqiJid,
, and !'ire 'versa (f'ee Fig. 390).

2438

CHEMICAL MICROSCOPY
STUDIES OF MIXTURES OF SUBSTAN.qES.

As an aid in the examination of mixtures, the microscope is part~cularly

useful; except in the case of gases and liquids or solid solutions; the constituents
are seen, under sufficient magnification, to be merely" side by side," and their
individual properties are available for purposes of identification, usually with~
out the need for more than the "optical separation" which the microscope
accom plishes ..
Whatever information can be gathered from the microscopical E)xamination
of heterogeneous mixtures is clear gain, for it supplements the chemical findings
and it helps to minimize the loss of evidence which is a consequence of the
usual methods of preparation for analysis. In other words, the maximum
amount of information should be obtaIned from the constituents while they are
still discrete, and before they are merged together by the processes of solution,
fusion, etc.
Furthermore, the mere recognition of the fact that ja ,mixture is present is a
guide to the interpretation of the analytical data, and the computation of the
results in terms of the different substances actually present as separate phases
is no longer a matter of conjecture or of arbitrary systems too often inconsisten,
with the facts. The conformity between calculated constituents and microscopically visible ones is an essential criterion in the analyses of rocks 'and ores
consisting of mixtures of silicates, sulfides and other minerals.
Even though it may not be possible to separate and identify individually
all the ingredients of a mixture, their nature may become apparent, in the light
of further quantitative or qualitative analyses of the mixture as a whole.
In connection with sampling of mixtures for macroscopic analysis, either
qualitative or quantitative, it is particularly important that any non-uniformity
should be recognized, in order that proper precautions may be taken to ins'lire a
representative sample. A sample is not representative, hoivever truly its
composition may summarize that of the specimen as a whole/if in obtaining it
the local variations. of composition have been" averaged together" without
knowledge or due appreciation of their existence or magnitude. Indeed, such
segregations of ingredients may be of more significance than the general analysis,
especially in the case of mixtures of powdered materials, or alloys, and they
are likely to be missed unless the microscope is employed.
Separation of powdered mixtures 7 is often necessary and always adv:i~able
as a preliminary to qualitative or quantitative analysis. Hand sorting, under a
low power magnifier or a Greenough binocular microscope, will be feasible for
all but the finest powders. Glass rods drawn to hair-like fineness, a thin
platinum wire, accurately pointed forceps, and capillary pipettes will be found
useful for picking up small grains or shoving them into heap/l. The rod or wire
may be moistened or rendered slightly sticky by a trace of oil, balsam, vaseline
or glycerine. Coherent aggregates such as rocks may need to be crushed fine
enough to yield non-composite grains, but it is also possible, .by means of a
hard needle, a silicon carbide or diamond splint, or a fine dental chisel; to scrape
off or dig out uncontaminated material sufficient for at least, a_microscopic or a
spectroscopic qualitative analysis.
-' 7 See "Handbook of Chemical Microscopy," Vol. I, pp. 141-148.

CHEMICAL MICROSCOPY

2439

For microanalysis, either iion-microscopical or by means of qualitative te~ts

carried out un<jer the microscope, still smaller samples are feasible. Streaks of
alloys or minerals, obtained by rubbing on a ground glass or silica plate, supply
material in finely powdered form, with minimum injury to the specimen.
"Chemical micrurgy" by means of the micromanipulator permits exceedingly
minute particles to be dealt with, and is invaluable. in the identification of
"specks." 8
Testing in situ is also a possibility, if separations are difficult; staining and
etching reactions, such as are employed in the microscopical study of ores or
alloys, enable the composition of the different grains of powdered or coherent
specimens to be recognized, and further serve to reveal localized variation in the
mixture.

HANDLING SMALL AMOUNTS OF MATERIAL

Solid particles can best be handled by a glass rod or platinum wire, as mentioned above. If suspended in liquid, sedimentation or centrifugal separation
will serve as a means of concentrating tl;tem. Fine capillary pipettes (Fig. 391)

FIG. 391.
made by drawing out glass tubing to about 1 mm. diameter, are useful for collecting sediments or transferrii:J.g small quantities of liquid.
The filtration 10 of small drops may be readily accomplished by the BehrensHemmes method (Fig. 392). A small square of
thick filter paper is placed close to the drop to
be filtered, against the paper is gently pressed a
capillary tube F whose lower end has been ground
true and smooth. The paper with the tube in a
vertical position is slowly pushed forward intolthe
drop. The liquid rises into the tube by capillarity. In general it is best to first attach a rubber
tube R to the tube F and draw up the liquid into
the tube by gentle suction applied by the mouth
to the mouth-piece M. The rubber tube is kept
slightly bent as shown in the figure. At the
m~ment of removal the tube is compressed beFIG. 392.
Titus and Gray, Ind. Eng. Chern. (Anal. Ed.), 2, 368-71 (1930).
"Handbook of Chemical Microscopy," Vol. I, pp. 141-5; Vol. II, p. 407.
10" Handbook of Chemical Microscopy," Vol. II, pp. 13-17.
8
9

2440

CHEMICAL MICROSCOPY,

tween the fingers and slightly stretched at the same time; this prevents the
liquid from flowing out of the tube as it is lifted off the fi:lter p~per. ,

When merely the preclPitate and not!

the liquid is required a narrow strip of
filter paper may be: Introduced il1to the
drop and the precillitate pushed into a
little heap.
Decantation 11 will usually accomplish a perfectly satisfactory separation
of liquid from a heavy precipitate. The
object slide on which the drop rests
must be scrupulously clean. If the end
of a drawn-out glass rod1or a platinum
wire is introduced into th'e drop near its
circumference and then moved outwards
the liquid will follow the rod and if at the
same time the slide is slightly tipped
the liquid Gim be decanted from the
FIG. 3~3.
precipitate.
Simple fractional distillations 12 can
be successfully performed by means of the apparatus shown in Fig; 393, 'the
successive fractions being removed from the receiver end of the distilling tube
by means of a capillary pipette.
/
'\.

:
MICROSCOPICAL QUALITATIVE ANALYSIS

13

Qualitative analyses are carried out under the microscope by means of

crystallization and precipitation reactions, many of )Vhich are used in ordinary
macroscopic analysis. Instead of basing the identification oj a precipitate
upon its color, upon the preceding steps in the analytical scheme, antI upon its
behavior toward the subsequent operations of the system, the micr:.Qscopical
form and optical properties of the precipitate are employed to characterize it.
This necessitates the choice of reactions which shaH yield practically specific
appearances, and hence largely eliminates the need for following any systematic
analytical scheme through a number of groups. Direc~ness and rapidity are
natural consequences of this procedure, for a precipitate when obtained is
recognized as such, and is not dependent for its significance upon the efficacy of
group separation~.
"Handbook of Chemical Microscopy," Vol. II, p. 10.
I
"Handbook of Chemical Microscopy," Vol. II, pp. 18-20: "
13 For further discussion, see Chamot and Mason, "HanclliQ9kof Chemical Microscopy," Vol. II (H)40), and the other books listed in the bibliography on page 2457.
II

12

CHEMICAL MICROSCOPY

2441

- Many of the reagents of microscopical qualitative analysis serve to show the

presence of several different iOI:\s, and are commonly applied to small separate
portions of the original solution. In special cases it may be desirable to separate or concentrate certain ingredients by means of some of the group reactions
of ordinary qualitative analysis, and then to make use of the microscopicaJ tests
as a means of final identification of the constituents of the group precipitates.
The minute amounts of material inciolved in microscopical analysis also
contribute to its rapidity; obviously, it is easier to dissolve a tiny fragment, or
to filter or evaporate a drop or two of liquid than to deal with the usual quantities. In cases where the supply of sample is limited, microscopical methods of
analysis are in order, and indeed may be indispensable. The sensitivity of the
tests, although striking when expressed in absolute amounts, is no greater in
terms of actual concentrations emplpyed than that of the same reactions as used
on a macroscopic scale. However, since it is possible to work with concentrates
obtained from a small amount of a relatively dilute solution, the effective sensitivity may be very great.
R.eagents are best kept as solids in small glass stoppered bottles, a drop of
solution being made up when needed.

METHODS OF APPLYING REAGENTS

There is usually a specific method for applying the various reagents, depending upon the relative concentrations required and the nature of the reaction
involved. In any case, however, actual precipitation is the criterion of a
positive test; crystalline products appearing only after considerable evaporation
has taken place are of little or no diagnostic value. Relatively dilute solutions
are preferable, in order that separation of the reagent, the unknown, or various
possible double salts may.not mask the precipitate obtained, and in order that
the latter may not be formed from such a highly supersaturated solution as to
appear "amorphous" or unresolvable by the microscope.
The most important methods of applying reagents are as follows:
a. The drop of r,!lagent is caused to flow into a drop of unknown solution on a microscope slide, by means of a small glass rod or platinum wire. The concentration gradient
in the zone of mixing affords a considerable range of conditions, so that various regions
of the preparation should be studied in order to observe a di~tinctive precipitate.
b. A solid fragment of the reagent is introduced into the test drop, either at the center,
or sometimes at or near the circumference. When introduced at the center, solution is
rapid, and if the reagent fragment is moderately large, excess of the reagent in high concentration is made effective. The reagent should be in small fragments-not finely
powdered.
c. A solution of the reagent is drawn in a narrow channel across a dry film of the
material to be tested. The unknown is dissolved in water or acid. A drop of the solution thus obtained is carefully evaporated to dryness upon a scrupulously clean object

2442

CHEMICAL MICROSCOPY
~

slide, using great care to avoid heating the dry film too hot. A concentrated drop of
the reagent is placed near the film and a narrow channel of. the reagent is drawn across
the dry film, the object slide being slightly tipped at the same time so as to feed reagent,
into the channel. Care must be taken that the reagent does not flood tp1l entire drop.
The reaction product is formed in the zone of contact between rea!?ent and solid unknown.
This method is used when the reaction product sought, is relatirely soluble, and when
high effective concentrations of reagent and of unknown are necessary for its precipitation.
d. The addition of another ion giving a less soluble compbimd can be resorted to
when an isomorphous series exists and the test as first performed has failed to yield a
decisive test. One of the best illustrations is that of testing for lead and (or) copper with
KNO.. If no precipitate is obtained aRb, Cs, or TI salt can be added; the solubility
of the compounds decreasing in the order named. The effect of such a procedure is to
increase the sensitivity of the reaction.
e. The reagent is incorporated into a fiber (viscose, silk, wo<;>l, collodion, etc.) and
the dry treated fiber is introduced into the test drop. This method is of great value in
many cases. One of the best examples is testing for boron with a viscose rayon fiber
'.
impregnated with turmeric (curcumin).
f. When HF or fluorides are employed as precipitating reagents they are added to
the test drop by either method a or method b, but special precautions are necessary.
The object slide must be of non-siliceous material such as celluloid, cellophane etc., or an
object slide may be protected by a coating of a hand impervious, transparent, colorless
varnish. All objectives save the one to be used should be removed from the nose-pieces
and this one protected by varnish, a protecting cell or drop of immersion oil and small
cover-glass.
g. In certain specific cases, it is found to be advantageous to cause the reaction tp
take place in capillary tubes.

..

Reactions involving gas evolution (NHa, ,or volatile acids)' are most simply
performed by placing the reacting substances in a tiny crucible (Fig. 394), over
which is placed a glass slide bearing
on its under surface a minute
hanging drop of an appropriate acid
.
v'or alkaline absorbent; the evolved
.
gas is then tested for in this drop.
Fusions are carried out in a tiny
platinum cup made from heavy foil;
other high timperature reactions
yielding; sublimates are brought
about in hard glass capillary tubes.
Reagents Used.-Some of the
most useful group reag!'lnts of
microscopical qualU;ative analysis
FIG. 394.
are as follows:

======

Potassium Mercuric Thiocyanate, K 2Hg(CNS) .-Zn-feathery white crosses; Cdcolorless prisms with one end bluntly pointed; Co-blue prisms in radiating clusters;
Cu-greenish yellow mossy dendrites or fusiform crystals;-Zn+Cu~Zn crystals colored
lavender, or brownish spherulites of radiating prisms; Zn+Cd, Ni, Mn-colorless
arrowhead crystals; Au-golden yellow mossy dendrites; Fe-,-pink color or pip.k Zn or
Cd crystals; Pb, Ag, Hg-very fine grained precipitate.
;
Cesium Chloride.-Sn-octahedra of Cs 2SnCI,; Ag-fine octahedra of AgCI, recrystallizable from NH.OH to give larger crystals; Pb-forked prisms of PbCI.; Sb-hexagonalplates of 3CsCI2SbCls; Bi-hexagonal plates of 3CsCI2BiCI 3 , or rhombic plates
of 3CsCIBiCI 3 ; Cd-hexagonal tablets or bipyramids of CsCICdCI 2 ; 9s, Ir, Pt, Pdyellow to orange octahedra of Cs 2 (Pt, etc.) C16 ; Cu-yellow or pra.nge pnsms and plates.

_.-

 CHEMICAL MICROSCOPY

2443

Cesium Sulfate.-Al--colodess octahedra of cesium alum) Cr-purplish octahedra

of cesium chrome alum; Ca-needles and oblique-ended prisms of CaS042H 20; Pb, Ba
and Sr-fine granules of BaS04 or SrS04 (on recrystallization from conc. H 2S0 4, Ba gives

featiiery crosses and Sr dagger-shaped crystals).

Potassium Nitrite.-Co-fine yellow cubes of 3KN0 2.Co(N0 2),.01 H 20; Cu or
Ni, +Pb-fine brown to black cubes of 2KN0 2 Cu(N02)2Pb(N0 2)..
Silver Nitrate.-Colorless precipitates insoluble in HNO,:
Cl- or Br--granular, or very tiny octahedra; I--fine granular; BrO,--tiny crosses
and prisms; CN--granular or fine prisms.
Colorless precipitates soluble in HNO,:
C 2H,02--thin plates and prisms; N0 2--1ong slender needles, pale yellowish green;
S04---rhombic tablets and prisms; CNO--fine needles and prisms on standing;
IO,--acicular and radiating crystals.
Colored precipitates:
As0 4----red granular, small tablets; Cr04-- or Cr 20,-red granules or pleochroic
plates and acute rhombs; P0 4----yellow granular or tiny stars and crosses.
Barium Chloride.-Precipitates insoluble in HNO,:
Cr04--, Cr 207---yellow granules, tiny crosses; IOa--granules, fine prisms and
crosses; S04----granules, fine crosses and platelets.
Precipitates soluble in HNO,:
BrO,--prisms; CNO--spherulites and needles; CO,---granules;
IonS yielding non-characteristic precipitates are not listed under th,; above reagents.

IDENTITY TESTS FOR ANIONS AND CATIONS

The most useful specific identity tests for the common anions and cations
are as follows:
Cations
Na-pale yellow tetrahedra with uranyl acetate; many of the tetrahedra appear to be
black because of total reflection. If the ratio Na : K is less thanl : 5, test usually
fails. Na+zinc uranyl acetate yields pseudo-octahedra, very characteristic;
presence of other alkalis or alkaline earths does not interfere with latter modification of test.
K-yellow or orange octahedra with H 2PtCl.; colorless prisms with HCI0 4. NH, must
be absent, as it gives identical reactions.
.
NH4-volatilization of NH,. Treat with NaOH in crucible (Fig. 394) covered with an
object slide carrying a hanging drop of very dilute HCI. Warm gently. Test
hanging drop with H 2PtCl.-yellow octahedra of (NH4)2PtCl .
Ca-fine needles, singly and in sheaves, oblique-ended prisms, and characteristic arrowhead twins, with dilute H 2S0 4. Prisms with tartaric acid. Very tiny octahedra
with oxalic acid. Sr and Ba interfere if present.
Sr-fine granular precipitate with H 2S0 4, which when recrystallized from hot conc.
H 2S04, yields dagger-shaped crystals. 'Ca and Ba must be absent. With oxalic
.
acid, .crystals similar to those of Ca.
Ba-feathery crosses on recrystallization from hot conc. H 2S0 4 Ca, Sr, Pb, Ag, must
be absent. In concentrated acetic acid K 2CrO, yields a very fine granular or
finely, crystalline precipitate, Ca gives none, and Sr rarely unless the solution
is made alkaline with NHiOH. With oxalic acid, mossy aggregates.

2444

CHEMICAL MICROSCOPY

Tests fall in mixtures of Ca, Sr and Ba. Separate Ca by extraction with boiling,
water, and Sr by dilute HN0 3 BaSO, remains insoluble; evaporl1tion of aqueous extract gives crystals of CaSO,2H 20. Evaporation of HNO a gives residue of SrS04,
which canbe recrystallized from conc. H 2SO,.
Mg-feathery stars developing into prisms, with P0 4--- and thetes~ drop made alkalil'\.e
with NH 40H. Fe, CO, Ni, Mn must be absent.
I,
,
Zn-feathery crosses with K 2Hg(SCN),. A very little Cd renders the crosses less
feathery; more Cd leads to the formation of arrow heads. Ni modifies the
crystals in a similar manner. Co colors the crystals blue, Cu lavender or brown,
Fe pink or reddish brown. Oxalic acid forms short prisms and tablets.
Hg-globules of metallic Hg by heating in a closed tube with dry Na<CO a Acidify a
solution with HCl, boil with a fragment of Cu wire or Cu foil. Hg amalgamates
with Cu. Heat coated Cu in closed tube-globules of Hg. Convert Hg into
HgI2 by iodine vapor-yellow and red crystals.
Hg(ous)-greenish "amorphous" ppt. with KI; white "amorphous" ppt. with HCl,
turning black with NH 40H; dark red granules or stars with K 2Cr.07 in drops
strongly acidified with HNOs.
Hg(ic)-rod HgI2 with KI, soluble in excess; colorless feathery crosses with Zn(C 2H s0 2).
and KSCN.
Al--octahedra with Cs 2SO,. Excess of H 2S0 4 must be avoided.
Sn-convert to SnCI 4; add CsCl-tiny colorless. octahedra. With oxalic acid, stannouSsalts give oblique ended prisms and skeletal forms; stannic salts give no precipitate.
.
Pb-yellow hexagonal plates with KI. Forked prismEl with HCI. Brown or black
cubes or tablets with KN0 2 and Cu(C,H S0 2)2 in.acetic acid solution.
Sb-hexagonal plates with CsCl; Bi gives hexagons unless the CsCI is in excess. Evolution as SbHs and blackening of AgNO s crystal; As must be proved absent by
evolution of AsHs with AI and KOH; H,S, HiP interfere.
\
Bi-hexagonal plates with CsCI; rhombic plates with excess. Small prisms with NIl-2S04
.
and H 2S0 4.
Cr-faintly purplish octahedra with CS 2S0 4, when Cr is present as Cr.(S04).. Conversion to Cr04-- by evap. with HNO s and KCIO a and subsequent testing With
AgNO a gives pleochroic Ag,Cr04' When Cr is present as a compound of CrO"
treat wi.th NaCl.and conc. H 2S0 4 in crucible with cover slide carrying a hanging
drop of NH 40H. Heat carefully; Cr02Cr. is evolved. Evaporate hanging drop;
test with AgNO s for Cr04--'
Mn-as Mn04- by oxidation with (NH 4)2S208( +AgNO a as catalyst). Or fuse with
Na,CO s and KNO s to give green Na 2Mn04, purple on acidi:(ication.
Mn++-long twinned prisms and six-armed radiates with H 2C.0 4.. Excess of NH4 salts
and mineral acids must be absent. Masked by many other cations.
Fe-blue color and precipitate with K,Fe(CN)6. Red color with KSCN if Fc+++.
Co-blue prisms with K,Hg(SCN)4. Excess of Fe or Ni interferes. When Zn is present,
blue feathery crosses or arrow-heads; organic acids interfere. KN0 2and HC2H sO.
-yellow cubes; Cu, Pb, Ni, Fe interfere.
Ni-fine pink needles with dimethylglyoxime. Tiny brown cubes.... ~ith KNO"
Pb(C,H s0 2). and HC 2H s0 2; Cu, Co must be absent; excess of Fe interferes.
Cu-greenish yellow mossy dendrites with K2Hg(SCN), (lavender or black spherulites if
Zn is present). Tiny black cubes with KN0 2, Pb(C 2H sO,). and HC 2H s0 2
(Ni, Co must be absent). Yellow prisms or orange plates with CsCI.
Ag-" amorphous" precipitate or tiny octahedra with HCl, larger on recrystl111ization
from NH 40H. Red pleochroic acute rhombic plates with K 2Cr 207. Red
granules and hexagonal tablets with AsO,---.

CHEMICAL MICROSCOPY

2445

SCHEME FOR THE QUALITATIVE ANALYSIS OF COMMON

NON-FERROUS ALLOYS
The following tabulation 14 may prove useful for the. systematic qualitative
analysis of alloys or mixtures containing salts of the metals indicated.
Dissolve in concentrated HN0 3. Evaporate to a syrupy consistency or to a moist
mass on water bath. Do not carry to complete dryness. Add water and again evaporate. Extract with warm water. Filter.
.
Residue A 'As; Sb; Bi; Sn-(Traces of Pb; Cu etc.) Remove a little of the residue and
test it for As by conversion in H3As.
Dissolve the remainder of Residue A in aqua regia. Divide into three parts.
1. Test with CsCI. Or reduce with metallic Fe to SnCb and test for Sn.
2. If Sn is absent test for Sb with CsCI; if Sn is present convert into potassium
antimonate and test with Na=Sodium antimonate.
3. Test with CsCI and confirm by conversion into sodium or potassium bismuth sulfate.
Filtrate I Add NH,CI and NH,OH Filter = Filtrate III.
Precipitate B.
Precipitate B Fe; Cr; Al; Pb; Hg-Dissolve in the least possible volume of dilute HCl.
Add N aOH in excess. Add N a 202, boil, filter = Filtrate II.
Precipitate C.
Precipitate C Hg, Fe-Dissolve in dilute Hel. Divide into two parts.
1. Test for Hg with KI. Or treat with metallic copper, sublime the Hg and
convert the globules into HgI 2 with iodine.
2. Test for Fe with KSCN or K,Fe(CN)6'
Filtrate II AI; Pb; Cr-Acidify with acetic acid. Divide into three parts.
1. Test for Pb with KI after adding HN0 3.
2. Test for CrO,-- with AgN0 3. (Silver acetate may be precipitated.)
3. Precipitate all the lead as PbSO,. Centrifuge. Concentrate. Test for
Al with Cs 2SO,.
Filtrate III Mn; Ni; Co; Zn-Add H 20 2, boil and filter = Filtrate IV. Precipitate D.
Precipitate D Mn; Co; Ni-Divide into three parts.
1. Fuse with Na 2CO.; green color=Mn. Or test for Mn with sodium
bismuthate.
2. '"Dissolve in HCI and test for Ni with dimethylglyoxime.
, 3, Dissolve in HCI and test for Co with K 2Hg(SCN),. Or test with sodium
nitrite.
Filtrate IV Cu; Zn; Cd; Ni; Mg-Acidify with HCl. Add Na 2S03 and NH,SCK, boil
and filter = Filtrate V, Precipitate E.
Precipitate E Cu-Dissolve in HCI, Test for Cu with K 2Hg(SCN),.
Filtrate V Zn; Cd; Ni; Mg-Add NH,OH in slight excess, Divide into three parts.
1. Acidify with HCl. Test for Zn, Cd, with K 2Hg(SCN),.
2. Test for Ni with dimethylglyoxime.
3. Add HNa 2 PO,. Decant, filter or centrifuge. Dissolve the precipitate in
the least amount of HCl. Reprecipitate the Mg as. NH,MgPO, -6H 20 with
NH,OH. Or add NaC 2H 30 2 and HC 2H.0 2 and test with U0 2(C 2H 30 2 )z;
Anions
B,-07--rose color on turmeric-viscose fiber, changing to blue with alkali.
C0 3---C0 2 evolved on acidification. Fine granular ppt. becoming rhombohedral,
with Ca(C 2H 30 2)..
14 Slightly modified from method of Whitmore and Schneider, Ind. Eng. Chern.
(Anal. Ed.), 2, 173 (1930).

2446

CHEMICAL MICROSCOPY

NOa--radiating needles with nitron sulfate.

N0 2--yellow cubes with K+ and Co++, and HC2H a0 2,
CN--HCN evolved on acidification. Tiny prisms with AgNO a.
SCN--red color with Fe+++. Feathery crosses with Zn++ and Hg++.
Fe(CN)6-----blue ppt. with Fe (-ous or -ic), brown with CU.l
Fe(CN 6)--.--blue ppt. with Fe++.
.
I .

PO,----tmy yellow 3-armed stars WIth AgNO a. Feathery rosettes and pnsms wIth
Mg++, NH,OH and NH,Cl. Tiny rounded octahedra :.vith ammonium molybdate.
AsO,----tiny red granules and hexagonal tablets with AgNO a. Reactions as for
PO,---, above.
As0 2---small yellow granular disks with AgNO a.
SO,---needles and oblique-ended prisms with Ca++. Rhombs with Ag+. Finegranules and stars with Ba++.
SOa----evolution of S02. Reducing action, with consequent oxidation to 80,--.
S20a---rhombic plates and sheaves of fine needles, with Pb(C 2H a0 2)..
S----evolution of H 2S. Oxidation to SO,-- with Br2.
CrO,--, Cr207---red pleochroic acute rhombic plates; imparts red or orange color and
pleochroism to co-precipitated Ag 2SO,.
Cl--tiny octahedra with AgNO a. Crosses with TINO a. Prisms with Pb(NO a)..
Cr02Cr. evolved on warming dry material with H 2SO, and K 2Cr 207, identified
as CrO,-- in hanging drop by AgNO a.
Br--reactions similar to Cl-. Br2, evolved with (NH')2S20S or KMnO" gives fine
needles in hanging drop of m-phenylenediamine HCI solution.
I--yellow hexagonal plates with Pb(NO a).. Red granules with Hg(NO a)2. 12,'
evolved with (NH,).S20S, gives blue cQlor wi~h starch in same or in hanging drop.
Cl0 3--reduction to Cl- by S03--' Yellow Cl0 2'evolved on moistening dry material
with conc. H 2SO,.
BrOa--crosses and prisms with AgNO a. Br2 on reduction with hydrazine or hydroxyl amine sulfate.
IOa--fine acicular crystals with AgN0 3. 12 on reduction.
ClO,---orthorhombic prisms with K+, colored purple if trace of Mn04- is present.
MnO,--purple color in solution, and in crys'tals of KCIO,.
The reactions listed above, while reliable in simple mixtures, may be interfered with
by the presence of certain other substances. In such an event alternative tests or preliminary separations may be necessary.

ANALYSIS OF MATERIALS OF UNKNOWN COMPOSITION

A crystallographic examination should always,be carried out. It is particularly important that the optical properties be determined even if the
examination be directed merely to ascertaining whether the refractive indices
are high or low and whether the material is isotropic or anisotropic.
When the material is isotropic its identification will, as a rule, be more
readily accomplished by first testing for anions. When the material is anisotropic no general rule can be formulated. Reactions which se,em most likely
to yield conclusive information should be tried first. :;Each test should be so
chosen and performed as to serve as a guide for the :6ext step to be taken.

2447

CHEMICAL MICROSCOPY

Formal schenles of analysis are not particularly advantageous, especially

Systematic procedures ate largely based on those of macroscopic qualitative
analysis.
'
,
Failure to obtain a positive test or one recognizable as due to an'y one of the
ions sought may be due to a number of causes.

if the presence or absence of certain elements only is to be ascertained.

I. The reagent may be yielding a precipitate with the ion of opposite sign.
II. The test drop or the reag~nt drop may be too concentrated.
III. The test drop may not have the correct pH.
IV. Ions other than that' being sought may interfere with the reaction. A
change in crystal habit may result.
V. The precipitate formed may be masked by the precipitation of other
compounds.
VI. The substance sought may be present in an amount below the sensitivity of the test .
.VII. The substance sought may be present in a complex ion, in which its
presence cannot be revealed by the reagent used. In order that the test may
succeed the complex must be destroyed.

MICROSCOPICAL QUANTITATIVE ANALYSES

15

In the analysis of heterogeneous mixtures the microscope is particularly

valuable when the ingredients are of a complex chemical character, and when
the ordinary methods of analysis yield data which do not directly indicate the
composition in terms of the constituents of which the mixture is made up.
Determinations of elements or ions by purely chemical means, although carried
to several decimal places, may not yield as many significant figures as the less
precisely quantitative microscopical methods, because of' the hair-splitting
assumptions often necessary before the results can _be computed to give the
information which is ultimately sought. This is particularly true in the case
of starches, meals, flours; adulterated spices, cocoa, drugs and other food
products; mixtures in which an undue proportion of naturally associated impurities is present; paper pulps, fabrics, felts; ore concentrates, refractorie~,
rocks, alloys, explosives, etc.
In order that microscopic quantitative analyses may be applicable to the
material, the following criteria should be satisfied: '
1. The components must be easily distinguishable in the mixture, or they
must be rendered so, by special preparation or illumination technique.
2. The observed components must bear a reasonably constant relationship
to the chemical or physical" properties sought by the ap.alysis.
15

"Handbook of Chemical Microscqpy," Vol. I, Chap. XIV.

2448

CHEMICAL MICROSCOPY

3. If weight percentage composition is sought, the specific gravities of the

components must be practically identical, oro-they -mu~t be known, or standard
mixtureR of known weight percentage compoRition must be available for
I '
comparison.
4. The sample must be representative, and satiSfactory checks must be
obtainable from duplicate determinations and from samples with known
composition.
I
Preparation methods must be appropriate to the character of the sample.
Powdered materials should be uniformly mixed and evenly distributed, without
segregation of the different ingredients. This is usually best accomplished by
suspending the material in a viscous liquid, such as glyc~rine or a heavy oil,
spreading a drop on the slide, and covering with a cover-glass. By choosing a
mounting medium with a refractive index nearer to one of the components than
to another, their outlines may be made to appear distinctly different and tllU~
aid rapid recognition. Examination between crossed nico!s may accentuate
differences between the ingredients. In food work ingredients, such as starch,
which are not essential to the analysis and which obscure the significant tissue
particles may be eliminated by chemical treatment. Paper fibers in mixtures
are commonly stained by Herzberg's reagent, which selectively colors pure
cellulose (rag) purplish red; purified cellulose (chemical wood, esparto) blue;
and lignified cellulose (ground wood) yellow. Alloys are polished and etched
to bring olit the different solid phases;' wineral mixtures may be examined in
powdered form, but are frequently surfaced or sectioned.
In any microscopical analysis a number of fields, usually taken from different preparations, should be studied. Few materials are microscopically
uniform, either in structure or composition, and this is particularly true the
higher the magnification, the smaller the field, or the coarser the grain of the
sample. Any data, such as counts or measurements, should be based on a
considerable number (usually hundreds) of particles; the average values will
then be consistent even when wide local variations occur. Individual counting
tallies for each constituent are invaluable for such work, as labor savers.
It should be borne in mind that the ingredients sought in the analysis need
not be these visible in the specimen, provided the relation between structure
and composition is known. For example, carb.on in steel may be estimated
microscopically to about 0.1 % (although no free carbon is present) by taking
advantage of the fact that pearlite, visible as a distinct constituent, has a known
carbon content of 0.8%, so that a steel containing 50% pearlite would have
0.4% C.

CHEMICAL MICROSCOPY

2449

ANALYSIS BY ESTIMATION
If a specimen is prepared so as to render its components indubitably recognizable, it is frequently possible to estimate its composition with reasonable
accuracy. Ideally, the constituents of the mixture should be of approximately
equal fineness and specific gravity, lest coarse' or bulky material be overestimated. An accuracy of 1O% of the proportio,n of each component is usually
possible with simple binary or ternary mixtures; practise on series of similar
constitution enables the observer to obtain results comparable with those based
on systematic cou:qting or measurements.
Comparisons with mixtures of known percentage composition lend further
accuracy to estimation analysis; if a series of standards is available it is usually
fairly easy for even the inexperienced observer to decide the value for the
unknown. Permanent preparations or photomicrographs are useful as reference standards, but a comparison microscope or comparison eyepiece, which
permits the unknown and the standard to be viewed simultaneously in the same
field, is a great aid.

ANALYSIS BY COUNTING PARTICLES

The great majority of microscopical quantitative analysis of powdered
mixtures are based on some form of counting procedure. Such numerical
methods are preferable to estimations (even though slower), if several components are I?resent, if there is an unavoidable tendency to over-estimate certain
ingredients, If the particles are not uniform in size, or if such high magnification
is necessary that the resulting small fields are not individually representative.
Ordinarily sever-al hundred particles are counted, in fields systematically
taken from more than one slide. If the particles vary in, size, they are
"weighted" in terms of their apparent volume.
Counting'one constituent only is a feasible means of microscopical analysis,
provided similar counts are carried out on the same quantity of material of
known composition, as a basis for numerical comparison. The number of
particles of a certain component in a definite amount of a mixture is frequently
used as a measure of its quality. In the food industries especially, such determinations serve for control and inspection, furnishing information unobtainable
by other means.
Seed hairs or bran particles in flour,16 hulls in rice bran,17 organisms in
water,1B molds in tomato products,19 are thus determined.
16 Keenan and Lyons, U. S. Dept. Agr. Bull., 938 (1922).
Keenan, U. S. Dept.
Agr. Bull., 1130 (1923).
17 A. O. A. C., "Methods of Analysis" (1925), p. 122.
18 Whipple, "Microscopy of Drinking Water" (1927), pp. 95, 123.
19 Howard, U. S. Dept. Agr. Bull., 581 (1917); A. O. A. C., "Methods of Analysis"
(1925), pp. 222-3; Jour. A. O. A. C., 6, 146 (1922).

2450

CHEMICAL MICROSCOP

The determination of cocoa shells in cocoa is a typical example. 20 A 2 mg.

sample is mounted in a clearing solution on a slide, an.~ after it becomes sufficiently transparent the total number of shell fragments is counted. 'By comparing with similar counts on mixtures having known shEfll conteJ).t, adulteration
up to 10% may be determined with an agreement of one or two %.
Various means are employed for insuring that the same quantity of material
shall be taken for each count: weighing a few milligrarris directly on tlie slide;
measuring the dry powdered material by volume; pipetting a definite volume of
a suspension carrying a known concentration of the sample; or using hollow
slides or cells of definite volume (see Fig. 395). The particles in the entire preparation must be counted, unless
the distribution is practically uniform and the area over which the
material is spread is kept the same
Depth
for
all pr(lparations by using cells
0.100 mm.
of definite area or definite thickness. Under these latter circumstances counts or small fractions of
the entire area of the preparation,
FIG. 395. Howard.Counting Cell.
as,defined by diaphragms or co-ordinate rulings in the eyepiece, or
markings on the slide, are representative of thE) composition as a whole.
Whether the entire area or several fields distributed over it are counted, a
mechanical stage for systematic movement of th(l preparation is almost
indispensable.
The average number of particles per field or the total for a number of fields,
preferably based on several different known mixtures, is plotted against the
percentage composition or incorporated into a formula which serves to give
the composition of any unknown mixture.
The ratio existing between the different constituellts of natural mixtures 'is
often characteristic, especially in food and drugs, any ,abnormal deficie-n.cy or
excess of a typical tissue element being suspect. An !uialytical method involving ratios 21 makes use of a reference substance (usually lycopodium powder)
which is added in definite amount to the unknown and in equal amount to a
sample having known percentage composition. Without any precautions as
regards the measurement of the preparation, counts of the particles in question
and of the reference substa:p.ce are made, the ratio betweeIi~them being dependent upon the composition of the sample being analyzed.
-Counting all constituents is preferable in most analyses of powdered materials, since known standards, measurements of the sample, and uniform size of
preparation or of field are thus rendered unnecessary. Such methods are
especially applicable to mixtures of several different components, such as papers,
or powdered minerals. The counts should take into ' consideration ,the size as
well as the number of the particles; the results will 1)e in terms of percentage
composition by volume unless the specific gravities of the ingredients are
'
SUbstantially equal, or are used to compute weight p(~r cent.
20 Silberberg, Jour. A. O. A. C., 6, 98 (1922); Pease, idem!.7, 141 (11:$23),8,176 (1924);
A. O. A. C., "Methods of Analysis" (1925), pp. 347-8. _ i - '
.
21 Wallis, Analyst, 41, 357 (1916); "Analytical Micioscc)py'" (1923), Chap. XI.
Mason, Analyst, 58, 440 (1933); Griffiths, ibid., 62, 510 (1937).

.CHEMICAL MICROSCOPY

2451

The analysis of a powdered mi);:ture of quartz and feldspars, as used in ceramics is

an example of this type of method. Various schemes have been devised to aid in differentiating the particles of quartz" orthoclase, and plagioclase.
a. Vitrification of ,the feldspars. by rapid fusion" without affecting the quartz
followed by mounting in a liquid of the same refractive index as that of the fused mate-:
rial, and comparispn with a series of standard samples.22
b. Mounting of the untreated mixture in a liquid of refractive index 1.540 which is
above feldspar and below quartz.23
c. Chemical diffetentiation, by exposure
to HF vapor, which roughens the feldspars
and leaves quartz clear, then treatment
with NasCo(N02)6 solution, which colors
the orthoclase yellow.24 . About 200 grains,
on each of three slides prepared by one of
the above methods, are counted for the
analysis.
Mixtures of gypsum and anhydrite are
analyzed similarly, by mounting in a Il).edium of refractive index 1.550, which is near
that of gypsum; this serves to render the
gypsum particles faint and transparent,
while those of anhydrite are boldly outlined 25 (see Fig. 396).
Ore concentrates, if of complex mineral composition, are advantageously analyzed by the microscope, particularly since
the actual amounts of the different minerals,
rather than of elerp.ents that may occur in
several, are sought. Composite grains,
Fw. 396. Mixture of Ground' Gypand the effectiveness of grinding and sepsum and Anhydrite, mounted in a liquid
aration, are also recognized. The material
of refractive index near that of Gypsum.
is spread over a surface that is marked off
in squares, and counted systematically, the
size of the grains being taken into account.26 A moderate magnification, such as is obtainable with a Greenough binocular microscope, is sufficient.
I
Paper fiber analysis is carried out by an analogous method, after pulping and staining with Herzberg's reagent. The fibers are tallied in terms of their length, measured
with reference to the diameter of the field. An alternative method ("dot-count") is
based on counts taken as the preparation is moved beneath the microscopEf~~' Every fiber
any portion of which passes beneath a given point (dot or cross haifs) 'In the field is
tallied, the longer ones thus being given proper, emphasis. Counts in different places
and directions on the preparation, dealing with at least 200 fibers on each of three slides,
are customary.
22 Booze and Klein, Jour. Amer. Ceram. Soc., 6, 698 (1!l23).
23 Insley, ibid., 10, 651 (1927).
.
:1 Gabriel and Cox, Amer. Mineral, 14,290 (1929).
25 Larsen, Proc. A. S. T. M., 23 I, 236 (1923); Gardner, idem, 26 I, 296 (1926).
26 Coghill and Bonardi, Bur. Mines. Tech. Paper 211 (1919).

2452

CHEMICAL MICROSCOPY
ANALYSES BASED ON MEASUREMEN1S OF AREA

Areal and linear analysis depends on the fact that if a section is taken
through an aggr~gate of heterogeneous material, the ~htercepts of the various
ingredients in the plane of the section, or in a line acros!, it, correspond to their
respective percentages by volume. The specimen should be in compact form,
so as to be surfaced or sectioned easily by cutting or grinding; alloys, rocks,
refractories, etc. are suitable for such preparation. If powdered mixtures,
such as ore concentrates, are to be dealt with they should be embedded in
sealing wax or other cementing material, and the aggregate surfaced for
measurement.
The areas of the constituents intersected by the plane of the specimen are
conveniently measured on a drawing traced on co-ordinate paper by means of a
drawing camera C" camera lucida "). A number of fields must be drawn,
especially if the material is coarse grained or not mixed uniformly. Photographs (provided they represent enough area) may be cut out along the boundaries of the various ingredients, and the respective areas totalled by weighing
the pieces of paper. Estimation or
comparison with a series of known samples is often feasible as a time-saver.
Linear analysis may be carried
but on drawings or photographs; the
intercepts are easily determined by
marking off those of each ingredient
on a separate strip of paper, the total
lengths finally being measured by a
ruler. Special mechanical stages' are
available, for moving the specimen
across the field of the microscope.
An individu~l screw is used for the
movement 6f each ingredient, the.
corresponding total movements being
shown on separate graduated scales.
The total distance measured should
be lOO_:'_ZOO times the diameter of the
FIG. 397. Cast Copper, Containing
about 0.25% OJ.:ygenas Cu- Cu 20 Eutectic average sized graIns. ~..___
(dark). Vertical illumination.
The volume percentage composition furnished by areal or linear analyses will represent the percentage composition by weight only in case the specific gravities of the ingredients are practically identical. Otherwise, the volume
per cent of each component must be multiplied by its specific gravity, and the
resulting ratios converted to weight per cent. SJch a calculation gives the
relative "amounts of the different solid ingredients, but not necessarily the
chemical analysis of the sample. If these ingr(1dients are of known chemical
composition, the analysis of the mixture may be computed in terms of elements
or radicals if desired, but this is not ordinarily of ,d irect importance in work
where areal analysis is used.
. _,In the analysis of aJloys it is sometimes desirable to ascertain the per cent
of a given element, such as oxygen or carbon, which is present in combination

CHEMICAL MICROSCOPY

2453

as an oxide, carbide, or eutectic of definite composition. In such a case, the

volume per cent of the compound is multiplied by the per cent of the required
'
element contained in it.
Examples of applications of areal methods of analysis are: oxygen in Copper 27
(see Fig. 397), composition of alloys,28 ores,z9 and rocks. ao

DETERMINATION OF PHYSICAL PROPERTIES

In addition to the optical constants mentioned above, various other properties may be studied microscopically on minute amounts of'sample. Melting
and transition points are.determined by means of a hot or cold stage,31 which
permits the temperature of the specimen to be controlled and measured while
it is being observed under the microscope.
If the substance studied is doubly refractive (as most solids are) its melting
may be observed with great sharpness by means of crossed nicol prisms, for
the solid will appear bright against the dark field, and will disappear on fusion;
a range of melting is also easily observed. For these reasons the melting points
obtained by a hot stage are likely to pe slightly lower than those by the capillary
tube method.

FIG. 398.

Section of Electrically Heated Stage.

Electrically heated stages are preferable; temperatures may be regulated by

a rheostat, over a range up to 400 0 or more. A thermometer (preferably a
small AnschUtz) or a thermocouple is used to measure the temperature; it is
essential that it be placed as near as possible to the object under observation,
and that the temperature within the stage show a minimum of locltl variations.
27 Bardwell, Bull. Amer. lnst. Min. Eng., 79, 1429; Tr. A. I. M. M. E., 46, 742 (1913).
28 Mathewson, Chern. Met. Eng., 11, 619 (1913); Logan, idem, 11, 668 (1913);
Polushkin, Tr. A. ,1. M. M. E., 71, 669 (1925).
29 Thompson, Jour. Geol., 38, 193 (1930).
30 Alling and Valentine, Amer. Jour. ScL, 14, 5Q-65 (1927).
31 "Handbook of Chemical Microscopy," Vol. I, pp. 198-209.

2454

CHEMICAL MICROSCOPY

It may be necessary to calibrate the stage or thermocouple by the use of substances having known melting points. A simple type Vf hot stage (see Fig.
398), that may be constructed by any competent mechanic, consists of a block
of aluminum about 70 X 100 X 12 mm.; perforated by an opening 20 mm. in
diameter. The top and bottom surfaces are recessed' Ito receive thin glass
windows, and a hole is drilled in one edge for the insertion of the thermometer.
About 2 meters of No. 20 (B. & S.) nichrome wire serves a~ the heating element;
it is wrapped on in several layers, each insulated by asbestos paper. In applying
the insulation, space is left for the insertion of the glass windows, which can be
covered with a perforated sheet of asbestos board. The sample to be studied is
mounted between two 18 mm. cover glasses, and supported on a tiny wire
tripod in the center of th~ heating chamber, practically in contact with the
thermometer.
Measurements with the microscope 32 are frequently employed in chemical
work, and are practically indispensable in dealing with fine or irregular structures, fibers, coatings, or particles below the range of testing sieves.
The,most widely used procedure employs a scale mounted within the eyepiece so as to appear superposed on the microscopic image. Since the value of
the divisions of such a scale will vary with the magnification employed, it must
be determined by comparison with a stage micrometer. The objective, eyepiece, and tube length used must be recorded, and maintained in any subsequent measurements, if the' calibration is to/be applicable.
Measurements of drawings, photomicrographs, or images projected on a
screen may be made by establishing the scale of the enlarge~ image; this is
done by replacing the specimen by a stage micrometer scale, and, tracing its
divisions enlarged to the same extent.
Particle-size determinations 33 are an essential part of the analytical examination of powdered materials. Especially in cases where the fine _state of subdivision is produced as an aid to chemical reactivity or as a result of chemical
treatment, such as precipitation or emulsification, size eXlj.minations fall within
the scope of chemical analysis. The behavior of a, pow~ered material is often
more 'affected by its fineness ~han by variations in ~ts chemical composition.

In dealing with powdered materials that lie within the range of testing
sieves, microscopical examination is a valuable supplement to the ordinary
methods. It reveals the shape of the particles, and hence their actual volume
and surface, rather than tho~e of an hypothetical sphere or cube ~~hose diameter is that of the sieve opening. Flaky -particles, of large diameter-but little
volume, or elongated grains of larger volume than their "sieve ,diameter"
would indicate, are likely to introduce serious discrepancies between screen
tests and performance, unless recognized microsc6pically. The character of
the grains, whether rough, angular, porous, aggregated, or composite, is further
revealed, and in mixtures the different ingredients m~y be followed 'through
the stages of the grinding, screening, and concentrating operations.
"Handbook- of Chemical Microscopy," VoL I, Chap. XII.
Handbook of Chemical Microscopy, VoL I, Chap. XIII.
I
See also the work of Green, Jour. Franklin Inst., 192,/637 (1921); Perrott and
Kinney, Jour. Amer. Ceram. Soc., 6, 417 (1923); Weigel, Bur. Mines Tech. Paper 296
(1924); Rept. of Comm. E-1, Proc. A. S. T. M., 19301,919-27.
32

33

CHEMICAL MICROSCOPY

2455

In almost all powdered materials a considerable range of sizes is present,

ranging, in the case of ground powders, from the maximum allowed down to
ultramicroscopic fineness. The fraction passing thtough a 325 mesh sieve
(opening 44 J.L), the finest standard size, contains the particles which contribute
almost all of the effectiveness of pigments, fillers, and most other finely divided
materials. This fraction cannot be analyzed by si(!ves, yet it may include
particles ranging from the diameter of ordinary fine pigments to those nearly
100 times as coarse.
The size and frequency of sub-sieve particles are best determined microscopically by one of the methods given below. Well dispersed preparations
are essential; the particles should lie substantially in Olle plane, and should be as
f!harply defined as possible. The mounting medium. should be of markedly
different refractive index in order to give good contrast, and the lenses used,
especially for the measurement of particles <5 J.L in diameter, should be of the
highest quality and resolving power.
Very fine and uniform particles are satisfactorily distributed by "rubbing
out" a suspension on a microscope slide, by means of a glass rod held flatwise.
The vehicle used should be volatile (xylene or turpentine) and the rubbing
SflOUfd continue until the film of liquid nas evaporated. This method is not
suitable for coarse particles, which may be crushed, ()r for non-uniform specimens, which may thereby be segregated. Materials of these latter types should
be stirred into a viscous mountant such as Canada balsam or glycerine, and the
suspension flattened to a thin layer by pressing bene~th a cover glass.
Interpretation of particle-size data in terms of other properties is dependent
on the validity of the methods of computation. Any ~ttempt to summarize the
dimensional properties of a powdere'd material by om) or two numerical terms
is-inadequate at best, and should be undertaken only after a thorough study of
the methods of expressing results and the significance of the various factors.
For exam pIe, the expression "average particle size" is almost meaningless unless it is clearly understood whether the average hats been computed on the
basis of weight, surface, or number. Furthermore, an average particle diameter calculated with respect to number is not nece!;sarily applicable to the
material.the surface properties of which are jmportant, nor can two materials
be properly compared except in terms of the same type of "average size."
The average size gives no indication as to whether the particles are all of like
dimensions, or vary over a wide range; in the latter Case averages are of little
value:
All particle size determinations, except mere estimations or comparisons
(which are very unreliable in the case of non-unifotm materials) should be
based on size-frequency measurements. The number of particles of each
size is counted, and serves for the calculation of av~rages and of uniformity
coefficients.
The methods of particle-size determination, while not standardized as yet,
fall into three classes:
Submicroscopic particles are measured, by counting the number present in a
known volume of a suspension of known concentrati(>n. 34 The volume of the
particles, and their diameter as spheres or cubes, is then computed. The
particles are not measured directly, nor is any account tl\ken of non-uniformity.
34

"Handbook of Chemical Microscopy," Vol. I, pp. 226, 425.

2456

CHEMICAL MICROSCOPY
,

The method gives no information as to size-distribution, and the average

particle size obtained is significant only in the case of 4ighly uniform particles;
it is suitable only for such material, or in case the fin~ness is beyond the limit
of resolution of the microscope. Fine colloids, "gas black," and the finest zinc
oxides are measured by this procedure.
i
I
Particles of microscopically measureable dimensions permit size-frequencies
to be determined. All the particles in a portion of the ~ample are counted, and
.their sizes recorded. A surface- or weight-distribution curve can then be
constructed, and averages computed on any desired basis. Such a procedure
is of wide application to all types of finely divided materials. 35
Material of highly non-uniform particle size is often inconvenient to count
and measure in all size ranges, and a preliminary fractionation, by sedimentation in liquid, is often advantageous. 36 The fractions may be measured for
size-distribution separately, or the average fineness of each may be estimated;
the amount of each fraction obtained must of course be taken into account in
computing the size-frequency of the original samplo.

APPARATUS FOR MICROSCOPICAL ANALYSIS

The apparatus required for microscopical analysis will vary depending on
the range of work to be undertaken. A good chemiqal microscope is the first
essential; it should be equipped for crystal studies with polarized light. Magnifications of 200 X or less are adequate for a large proportion of the examinations of materials, and for almost all qualitative and quantitative procedures.
For particle size determinations on fine particles, 2 mm. or 4 mm. objectives
are essential. Hand lenses, or a Greenough binocular microscope, are useful
for separating and handling material under moderate magnifications.
If any microscopic measurements are to be made, a calibrated micrometer
eyepiece is necessary; for counting of particles, a co-ordinate ruled micrometer
is desirable. For systematic movement of the preparation in analysis by
counting, a mechanical stage will be required: Counting tallies are a great
convenience if large numbers of particles are to be ,counted. For drawings to
:-;cale or areal analyses a drawing camera is essential. A comparison eyepiece
is advantageous, if two similar microscopes are available.
The study of opaque objects, such as alloys, will necessitate the use of a
vertical illuminator. An artificial light source of suitable intensity and uni-

Perrott and Kinney, loco cit.

36 Weigel, loco cit.
35

CHEMICAL MICROSCOPY

2457

formity, such as a 6' volt ribbon-filament lamp, is an important accessory for

examinations at high magnifications, and for photomicrography.
Refractive index determinations require a
series of suitable standard liquids covering the
,range of interest, at intervals of 0.01. A hot
stage, with thermometer or thermocouple,
must be provided if melting points are to be
determined.

A "micro" burner or other device giving a

tiny flame about 5 mm. high is a necessity.
When gas is available a bunsen burner with
" reserve" or "pilot" flame will be found to be
must satisfactory (See Fig. 399).
Reagent bottles (best contained in a perforated block), miniature crucibles and holders,
dIstilling and filtering tubes, dropping bottles
for liquid reagents, slides, cover-glasses, and
t\w, ~'cR}c'r,'r,'b,"''J "'~'b,~~~t'r, 'b,~~ w..~\\~t\~~ w..~~\'b,

are also indispensable to the "micro" laboratory. A collection of authentic samples, for
study and comparison, is eSRentiaL
The above list of equipment'may be somewhat curtailed, but for special purposes and
extensive investigations in the general field of
FIG. 399.
c)lemical microscopy numerous other pieces of
apparatus are required; the nature and uses of these are described in various
standard works.
BIBLIOGRAPHY
In addition to the works listed below, numerous references to others and to the
original literature are given in the authors' "Handbook of Chemical Microscopy,"
Vols. I and II.
,

General Principles, Methods, and Appsratus

Beck, "The Microscope," (R. and J. Beck, London, 1938).
Chamot and Mason, "Handbook of Chemical Microscopy," Vol. I (John Wiley and Sons,
New York, 1938).
Metzner, "Das Mikroskop" (Fr. Deuticke, Leipzig, 1928).
Barnard and Welch, "Practical Photomicrography" (Arnold, London, 1936).
Hind and Randles, "Handbook of Photomicrography" (G. Routledge and Sons, London,
1927).
Eastinan Kodak Co., "Photomicrography" (1935).

Microscopical Qualitative Analysis, and Microchemical Manipulations

Behrens-Kley, "Mikrochemische Analyse" (Voss, Leipzig, 1915).
Behrens-Kley, "Organische Mikrochemische Analyse" (Voss, Leipzig, 1922).
Benedetti-Pichler,-Spikes, "Microtechnique of Inorganic Qualitative Analysis"
(Microchemical Service, Douglaston, N. Y .. 1935).
Chamot and Mason, "Handbook of Chemical Microscopy," Vol. II (.John Wiley and
Sons, New York, 1940).

2458

CHEMICAL MICROSCOPY

Donau, "Arbeitsmethoden der Mikrochemie" (Franckh'sche Verlagshandlung, Stuttgart, 1913).

Emich, Methoden der Mikrochemie, in Abderhalden's "Handbuch der biologischen
Arbeitsmethoden," Abt. I, Tl. 3, Hit. I, pp. 45-:-324 (Urban & Schwarzenberg,
Berlin, 1921).
I
Emich, "Mikrochemisches Praktikum" (J. F. Bergmann, Munich, 1924).
Emich-Schneider, "Microchemical Laboratory Manual" (Wiley,' N. Y., 1932).
Emich, Umsetzungen sehr kleiner Stoffmengen (Mikrochemie), in Stahler's "Handhuch
der Arbeitsmethoden in der anorganischen Chemie," Bd. II, Tl. 2, pp. 655-882
(W. de Gruyter, Berlin, 1925).
Emich, "Lehrbuch der Mikrochemie" (J. F. Bergmann, Munich, 1926).
Geilmann, "BUder zur qUalitativen Mikroanalyse anorganischer stoffe" (Voss, Leipzig,
1934).
Klein and Strebinger, editors, "Fortschritte der Mikrochemie" (Fr. Deuticke, Leipzig,
1928). An extensive symposium covering the years 1915--1926.
Mayrhofer, "Mikrochemie der Arzneimittel und Gifte," Parts I and II (Urban &
Schwarzenberg, Berlin, 1928).
Short, "Microscopical Determination of Ore Minerals" (U. S. Geol. Survey, Bull. 825,
1931).
.
Foods, Drugs, and Sanitary C,hemistry
Barnstein, Il Anleitung zur mikroskopischen Prilfung und zur Begutachtung der Kraftfuttermittel" (Borntraeger, Berlin, '1920).
Clayton, "Compendium of Food Microscopy" (Wm. Wood & Co., New York, 1909).
Greenish, "Microscopical Examination of Foods and Drugs" (P. Blakiston's Son,
Philadelphia, 1910).
/ .
Greger, "Mikroskopie der landwirtschaftlichen Unkrautsamen" (P. Parey, Berlin,.
1927).
Kinzel, "Mikroskopische Futtermittelkontrolle" (E. Ulmer, Stuttgart, 1918).
Koch, "Die mikroskopische Analyse del' Drogenpulver" (Borntraeger, Leipzig, 1901).
Koch, "Einfiihrung in die mikroskopische Analyse der Drogenpulver" (Born traeger,
Berlin, 1906).
Mace, "Les Substances Alimentaires etudiees au Microscope" (Bailliere, Paris, 1891,.
Moeller-Griebel, "Mikroskopie die N ahrungs- und Genussmittel aus deri Pflanzenreiche"
(J. Springer, Berlin, 1929).
Morris, "Microscopic Analysis of Cattle-Foods" (Cambridge pniv. Press, 1917).
Schneider, "Microbiology and Microanalysis of Foods" (P. Blakiston's Son, Philadelphia"
1920).
'
,
Schneider, "Microanalysis of Powdered Vegetable Drugs" (P. Blakiston's Son, Philadelphia, 1920).
'
Scott, "The Microscope in the Mill" (Northern Pub. Co., Liverpool, 1920).
Ward and Whipple, "Fresh Water Biology" (John Wiley &; Sons, New York, 1918).
Whipple, "Microscopy of Drinking Water" (John Wiley & Sons, .New York, 1927).
Winton, "Microsc,opy of Vegetable Foods" (John Wiley & Sons, New Y.o_rk, 1916).
Winton, "Structure and Composition of Foods," 3 Vols. (John Wiley & Sons, N. Y.,
1935-37).
Textile and Paper Fibers
Beauverie, "Les Textiles Vegetaux" (Gauthier-Villars, Paris, 1913).
Herzberg, "Papierprilfung" (J. Springer, Berlin, 1927). !
Herzog, "Mikrophotographischer Atlas der technisch wichtigen Textilfasel'n," 1. Teil
(Obernetter, Munich, 1908).
Herzog, "Mikroskopische Untersuchung del' Seide und Kunstseide" (J. Springer,
Berlin, 1924).
I
Herzog, "Die Unterscheidung der Flachs- und Hanff'aser" (J. Springer, Berlin, 1926).
von H6hnel, "Mikroskopie der technisch verwendetenaserstoffe" (A. Hartleben,
Leipzig).
. _; _
Kronacher and Lodemann, "Technik der Haar- und Wolleuntersuchung" (Urban &
Schwarzenberg, Berlin, 1930).
,

C.HEMICAL MICROSCOPY

2459

Laurill, "Textile Microscopy" (E. Benn, London, 1928).

Mathews,. "Textile Fibers" (Jehn Wiley & Sons, New York, 1924).
Preston, Modern Textile Microscopy (Emmott, London, 1933).
Schwarz, Textiles and the Microscope (McGraw-Hill Book Co., New York, 1934).
Skinkle, "Elementary Textile Microscopy" (Howes Pub. Co., New York, 1930).
Sutermeister, "Chemistry.of Pulp and Paper Making" (John Wiley & Sons, New York,
1928).
Technical Association of Pulp and Paper Industry, "The Manufacture of Pulp and
Paper," Vol. V (McGraw-Hill Book Co., NeW"'York, 1929).

Metallography
Desch, "Metallography" (Longmans, Green and Co., New York, 1937).
Hoyt, "Metallography," 2 Vols. (McGraw-Hill Book Co., New York, 1920).
Sauveur, "Metallography and Heat Treatment of Iron and Steel" (McGraw-Hill Book
Co., New York, 1936).
Williams and Homerberg, "Principles of Metallography" (McGraw-Hill Book Co., New
York, 1935).

Miscellaneous Technical Applications

Erdmann-Konig, "Warenkunde," 2 Vols. (J. A. Barth, Leipzig, 1925).
Garner, "Industrial Microscopy" (Pitman & Sons, London, 1932).
Hager-Tobler, "Das Mikroskop und seine Anw.endung" (J. Springer, Berlin, 1925).
Hanausek-Winton, "Microscopy of Technical Products" (John Wiley & Sons, New York,
1916).
Lindsley; "Industrial Microscopy" (Wm. Byrd Press, Richmond, 1929).
Molisch, "Mikrochemie der Pflanze" (G. Fischer, Jena, 1921).
Poschl, "Technische Mikroskopie" (Stuttgart, 1927).
Sasserath, "Mikroskopische Warenprtifung" (S. Hirzel, Leipzig, 1910).
Schneider-Zimmermann, "Botanische Mikrotechnik" (G. Fischer, Jena, 1922).
Tunmann-Rosenthaler, "Pflanze mikrochemie" (Borntraeger, Berlin, 1931).
Wallis, "Analytical Microscopy" (Edw. Arnold and Co., London, 1923).
Wilson, "Chemistry of Leather Manufacture" (Chemical Catalog Co., oNew York, 1928;
1929).

QUANTITATIVE MICROCHEMICAL ANALYSIS 1

Introduction.-.l\1icrochemical analysis may be defined as the determiPation
of very small amounts of constituentR when using minute amounts of ID&terial
The weight of the sample used for an analysis ranges from 3 to 15 milli~rams.
\Vhere only small quantities of material are available micromethods mjlst, of
necessity, be used, but even where large 8smples could be used the speei:l with
which a microanalysis can be carried out, without sacrifice of accuracy, }:ias led
many men in the fields of both pure and applied chemistry to discard some of the
time-consuming macro methods. An analysis can be completed in many cases
in less than an hour and very few commme over an hour of the analyst's time.
In many cases, and especially, in organic analysis, it is necessary to coml)ietely
oxidize the material in order to obtain accurate results and the advantage here
of a small sample over a <large one is at once apparent. Anyone whO will
strictly adhere to the established rules and customs of good analytical tecllOique
can acquire within a very short time the skill required in microanalysis. 1rlany
chemists, especially in this country, are not using micromethods becaus~ they
have the false impression that skill can be acquired only after long practi~ and
that the procedures are for the analysis of substances of which only small
quantities arc available. It is hoped that this impression will be overcome
when chemists become familiar with the micromethods and reaHze that these
are authentic and becoming commonly used in all branches of ana~ytical
chemistry. .
Semimicro differ from micromethpds {n that a larger sample is used (15-30
mg.) and the weighing may be done on a less sensitive balance to 0.05 mg.
The apparatus used is essentially the same as in microwork except it way be
slightly larger. It is, however, the accuracy of the method which should determine the size of the sample in both methods. In many cases it is the method
",,<I. ""t, t,\\...
t,\\......."''''t.. ",\\.\."'\\. <1...1...,,,,,,,.... t,\\... "-M.t
Tho.
methods described in the foJIo'wing pages have been chosen from amopg the
great number pUblished, either because these have become standard after
many years of use, or are based on principles and'methods already well es
tablished in analytioal chemistry. They are also selected for their conve:nience
and accuracy in routine work. It is not possible, under ordinary laboratory
conditions, to weig~ to a 0.001 mg. For this reaSOD, sample weights should be
chosen that allow!the sixth place to vary and lIot affect the accuracy of the final
results. The analyst can, with due care and precaution, use 2 mg. ssm pies
and every routine microlaboratory :-lhould be prepared and know how to work
with such small quantities' \"'hen such problems arise. It if! not'felt, ho~\'ever,
that'micro8t)alysi3 will fulfill its greatest service-to pure or industrial cheJ11istry
unless anaI'ysts know that the methods are time saving and accurate for routine
work. ~he procedures are, therefore, selected and described with these f;:lcts in
mind.
I

"'' ' 1,\\.'' '1, '"

""''''''-'''''''j.

"-

I By L. T. Hallett, in charge of the depnrtment of quantitative microanalysis, East.

man Kodak Research Laboratories, Rochester, N. Y.
'. 2460

QUANTITATIVE MICROCHEMICAL ANALYSIS

246t

The Microlaboratory.-Tbe microlaboratory should have an abundanc<l

of north light and be free from direct sunlight. If air conditioning is ava.ilabl~
it should be used, but it is not essential. The humidity s)1ould be around 50%.
so that static charges on. apparatus are avoided. Microbalances are sensitiv(l
to sharp temperature changes so that it is well to try and keep the variation ill.
room temperature within 10 c.
Each man's apparatus (if electric furnaces are used) should be grouped
around his microbalance so that unnecessary steps are eliminated. Thiii\
arrangement is not possible where gas furnace:;> are employed because of th~
sharp temperature fluctuation . In such cases, the ba14nce must be locatect
far enough away to avoid this.
The Microchemical Balance.-Within recent years balances have beeI\
developed which 'veigh to 0.001 mg. carrying a maximum IO.!td of 20 grams,
These balances have without doubt done much towards making microchemistr~
a practical branch of analytical chemistry.

I
.~
(

i
I

FIG. 4.00.

The balanoe, Fig. -400, is best mounted on a concrete slab 3 inches thick
which is firmly bracketed-against_a wall at a height of 3 feet. Strips of lead or
rubber are pla._?ed between the bracket and the slab in order to absorb vibration.
Freedom from minor vibration is Dot as important as a uniform temperature in
the room and absence of air emrents. A small table 30 inches high, 12-18

2462

QUANTITATIVE MICROCHEMICAL , ANALYSIS

'

inches wide, and 30 inches long, mounted OIl casters, is used in front of the
balance for holding the notebook, etc. It is made low enough so that it may
be shoved under the balance table. The operator by such an arrangement does
not lean against or jar the table holding the balance. ~t is best to place the
balance in the same room where the microwork is done so that change in
temperature in passing from one room to another is aroided. In order to
shield the balance from dust and air currents, it is best 'placed within a large
plate glass balance case 36 inches long, 15 inches wide, and 18 to 24 inches high.
A light should not be hung close to the balance but indirect lighting used wherever possible or the light placed far enough away so that the heat given off does
not affect the weighing. Side doors are provided through which objects are
placed on, and removed from, the pans without raising the front door of the
balance. This prevents the breath of the operator from causing air currents
and changing the temperature within the balance case. Two hooks are provided on the left-hand stirrup in order that absorption tubes may be more conveniently weighed in a horizontal position.
Tlie beam is shorter than that used in the standard analytical balance and is
divided, beginning on the left-hand side, into 100 equal divisions. When a
5 mg. rider is placed in the first notch, the balance, when released, should be in
equilibrium. Each division corresponds to O.l.'of a milligram, The displace-.
ment of the rider one notch should cause a' ,change of the resting point 10
divisions. By means of the magnifying glass attached to the scale it is possible
to estimate one-tenth of a division so that/the weighing may be made to an
apparent accuracy of 0.001 of a milligram. The change in successive scale
readings of a good microchemical balance is very slight and very regular.
When the rider is moved the operator should make certain that it falls to the
lowest part of the notch and remains in an upright position. A second mag;nifying glass extends the length of the beam or moves with the rider-arm t,b
enable the operator to see that the rider is properly seated.
All objects which are frequently weighed such as crucibles, weighing bottles, and absorption tubes, should be
counterpoised, using a small glass flask, Fig. 401, and
glass beads for tare, making a final adjustment by adding
an aluminium wire filed the 'proper amount. For light
objects, such as boats, an' aluminium wire alone may be
used. By this means, most df the weighings may be performed by using only the rider. The 10'IDg., two 20 mg.,
and 50 mg. weights should be calibrated agairu,t the rider
and filed in order that corrections do not have--to be applied. All weights and counterpoises should be kept in
the balance-case, and forceps and spatulas should be kept
on a glass plate in the large glass balance case. The ..zero
point may be adjusted by mean's of the small vane at
one end of the beam. The sensitivity of the b,alance
F IG. 401.
should be tested with various loads in the same way
that the standard analytical balance is tested, except that the rider is moved
0.1 mg. instead of 1 mg.
I
The balance should be inspected regularly and th~.nife-edges and supports
kept clean by the use of dry chamois skin.
.

QUANTITATIVE MICROCHEMICAL ANALYSIS

2463

As far as possible all objects to be weighed should not be touched by the

hands because of the moisture and grease always present and also to prevent
change in temperature. The object should not be wiped immediately before
weighing because of the removal of the film of moisture which will be again
absorbed while weighing. The objj)ct should be placed on the balance and an
approximate weighing made, .the final exact'reading being taken after the lapse
of three minutes. All ~bjects weighed should be kept free from dust, and the
hands of the operator should be kept free from dirt and body oil by frequent
washing and careful drying. It is the adherence to details, such as these, that
makes for success in microanalysis.
A small block of copper, kept in the desiccator, is used to decrease the time
for the object to attain the temperature for weighing.
NOTES.-1. In reporting weights they are often expressed in terms of so many
(gamma) or (IL gm.) and')' = IL gm. = 0.001 mg.
2. The use of lead shot for filling counterpoise flasks is to be discouraged when the
period between weighings is of such length that conditions of temperature and pressure
have changed materially. Most of the, objects counterpoised are of glass which has a
different density than lead. The use of glass beads is, therefore, desirable. A tare of
the same size and shape should be used when possible. This cannot be easily done in
the case of weigh~ng absorption tubes with some types of balances.
When, {or example,2 five gms. of shot are used to balance a glass vessel, the difference in the volumes of air displaced by the lead and glass is about 1.5 ml. (sp.g. of glass
taken as 2.7). The weight of this excess air displaced by the glass vessel will vary 6')' per
degree of temperature variation, 23,), per cm. of barometric variation, and about 21')' for
the difference between dry and saturated air.
3. In order to simplify weighing in routine work without loss of accuracy, the scale
divisions of the microbalance are numbered beginning at the right 0, the midpoint 10,
and the extreme left 20. This avoids plus and minus signs. It has also been found that
the method of short swings used in macro work can be used with the microbalance. If
, the total deflection is not over 3 large divisions, tlien the mean of two successive readings
gives the rest point as
8.6+11.1=1;7 =9.8.

It is, of course, not necessar~ for the zero point of the balance to be exactly at the
midpoint 10 or that the balance be adjusted to a sensitivity of exactly 10 divisions for
0.1 mg. However, the sensitivity should be kept between 9 and 11 divisions. It is
important that the balance maintain a zero point, constant to 0.5 divisions during
a series of weighings and a properly working balance will do this if the room temperature
is kept within 0.5 0 C. Where initial and final weighings involve an interval of
time the zero point should be recorded in each case so that when a shift in zero point
greater than 0.5 divisions (0.005 mg.) takes place the appropriate correction can be
made. A typical example of recording and calculating weights is given on page 2464.
The zero point is assuJl1ed to be constant and a tare is used in weighing the boat.
4. A very good description of the various types of micro balances and their use is
given by G. Gorback. 3 Air or magnetic damping is used on some microbalances which
helps to speed up weighing. However, a sufficient period (usually 2-3 minutes) for
air currents to subside must still be allowed before the final reading is taken. Balances
which claim a greater sensitivity than 0.001 mg. are no1Fpractical under ordinary laboratory conditions. Such instruments must have every condition rigidly controlled.
2

Roger J. Williams, Ind. and Eng. Chem., Anal. Ed., 8,229 (1936).
Mikrochemie, 20, 254-337 (1936); also M. Furter, Ibid., 18, 1-10 (1935).

2464

QUANTITATIVE MICROCHEMICAL ANALYSIS

Example

Sensitivity of balance for 0.1 mg. = 9.2 divisfons or 0.1 division = 0.011 mg.
Beam
Reading
mg.

Rest
Point

Boat and Sample

Empty boat
Difference
Diff. R.P. 2.3 X 0.011
Weight sample

5.3
0.2
5.1.'
0.025
5.075

10.9
8.6
2.3

Boat and Sample

Empty boat
Difference
Diff. R.P. 2.3 X 0.011
Weight sample

5.3
0.2
5.1
0.025
5.125

6.3
8.6
2.3

When the rest point of the sample is

greater than the rest point of the boat, the
weight obtained by multiplying the difference in rest points by the sensitivity is
added to the difference of the beam readings in order to calculate the weight of
sample taken.

Boat a~d Sample

Empty boat
Difference
Diff. R.P. 0.0 X 0.011
Weight sample

5.3
0.2
5.1
0.000
5.100

8.6
8.6
0.0

When the rest points are the same, the

weight of the sample is found by taking
the difference of the beam readings.
.

Difference iiI rest points 2.3 X sensitivity 0.011 = 0.025 mg.

When the rest point of the sample is
greater than the rest point of the boat, the
weight obtained by multiplying the difference in rest points by the sensitivity is '
subtracted from the difference of the beam
readings in order to calculate the weight
of sample taken.

Sampling.-If the same care that is used in macro sampling is taken in

preparing samples in micro work, the error in microsampling
be no gr~ater
than when a larger sample is used. All samples should be finely ground for
the reason that a single crystal may weigh as much a,~ 2 mg.
For a mathematical study of the theory and errors in microsampling, the
,
following references are given: 4
Drying and Weighing Samples.-When a substance requires drying with~
o.ut vacuum, it has been found that a small sid~-outlet desiccator charged with
phosphorus pentoxide or other suitable drying agent, s1.1.h as "Drierite" or
"Dehydrite," and ,heated internally by a coil of resistance ~wire inserted
through the opening is useful. A vacuum desiccator may be used to hasten
drying at room temperature if a high temperature causes decomposition.
Substances for analysis should be dried to cQnstant weight. However, in
routine work, where a great number of samples must be analyzed each day and
speed is a factor, it has been found that for compounds with a high melting point
drying overnight at 100 C. in vacuum in the presence of a drying agent is
sufficient except in rare cases. A vacuum pistol dryer of the Abderhalden typ~
with electric- heating has been found very satisfactory. The design may be

will

4 A. Benedetti-Pichler, Z. AnaI. Chem., 61, 305-31 (1922); Biule and BenedettiPichler, Z. AnaI. Chem., 74, 442-56 (1928); Joseph Mi~Z: AnaI. Chem" 73, 257-64
(1928).
' . _,

QUANTITATIVE MICROCHEMICAL ANALYSIS

2465

seen in Fig. 402. The total length is 30 cm. and the diameter of the vacuum
drying chamber is 3.5 cm. Ten to fifteen samples, in weighed boats and
placed on a glass plate, may be dried at one time and for this reason it is
rgcommended for routine drying. After drying, the samples are placed in the

FIG. 402.

small side-outlet desiccator heated to 70 C. Before weighing a sample, it is

removed to an unheated desiccator to cool, then placed in the weighing bottle,
Fig. 403, and, after standing three minutes on the balance, its final weight is

FIG. 403.

taken. The long projections on the weighing bottle are to prevent heating of
the bottle by handling.
Drying and Weighing Hygroscopic Substances.-HygroScopic substances
which take up water slowly may be dried in a boat which is quickly placed in
the weighing bottle, Fig. 403, previously kept in the balance case. In the case
of substances which take up water rapidly and form hydrates, the method of
Milner and Sherman 5 is recommended.
5 Ind. and Eng. Chern., Anal. Ed., 8, 427 (1936).

2466

QUANTITATIVE MICROCHEMICAL ANALYSIS

A vapor bath, Fig. 404, maintains a constant temperature as in the ordinary

Abderhalden. The two-way stopcock can b-e turned to permit evacuation or
the drying of incoming air, the latter entering through a very fine capillary
cemented in the plug of the cock and then through the IU-tube containing the
desiccant. The other end of the tube carries two ground joints, a large one for
introducing the plate holding the pig and a small one to hold the brass rod used
to open and close the pig while the sample is in the drier. I A closed-end mercury
manometer is used to seal the small ground joint and to indicate leaks if present
when the apparatus is evacuated overnight. The weighing bottle or pig is
made without the long projections for handling. A phosphor-bronze spring
clip holds the pig firmly in place on an aluminum base with two depressions for

/
FIG. 404.

the front legs, Fig. 405. The base, whose edges are curved~t(), fit the drier, is
held fixed within the drier oy another spring clip. The solid glass stopper of
the pig has a small handle with projecting knob which is engaged_pya brass
sleeve comprising a bayonet joint and which enables the stopper to be withdrawn or firmly seated while the pig is in place. Before the sample is weighed
out, the pig and empty boat are dried and capped in the apparatus. The
sample is weighed and dried and the result recorded. ! Then the pig is opened
and the sample allowed to equilibrate with the moisture in the room. Usually,
after 30 to 60 minutes, equilibrium is so nearly approached that any further gain
in weight while transferring the sample to the combustion tube is negligible.
The pig is then closed and reweighed. The gain in weight upon exposure to
air is, subtracted from the weight of water found ll'iore the percentage of
hydroge~ is calculated.

QUANTITATIVE MICROCHEMICAL ANALYSIS

2467

The method (see Fig. 402) of heating electrically and having the drying
agent attached closer and with a wider opening to the sample is more convenient

FIG. 405.

than in the Milner and Sherman apparatus. Samples may then be dried at any
temperature by simply changing the rheostat setting.
Weighing Liquids.-When the substance is to be oxidized, as in the determination of carbon, hydrogen, sulfur or nitrogen, a capillary tube, Fig. 40(\ A-B,

_____

~>---------------------<~ B

-------+--ti(

;. E
I

I
I

FIG. 406.

is closed and drawn out. A small crystal of potas~ium chlorate, placed in the
capillary, Fig. 406 C, is melted and cooled. A hair capillary is formed by drawing out the open end of lrhe tube, this is broken at a suitable point, Fig. 406 D,
and the tube weighed. The tube is then carefully heated to expel the air but
not sufficiently to melt the potassium chlorate, the open end dipped into the

2468

QUANTITATIVE MICROCHEMICAL ANALYSIS

liquid to be analyzed and allowed to cool. The excess liquid is car~fully removed from the hair capillary by gently heating; after w~ich the tube is sealed
and weighed, Fig. 406 E.
.
The handle and tip are broken off and the tube is placed in the combustion
tube on a narrow strip of ignited platinum. It is then quickly placed in position
by the use of a platinum or nichrome wire.
I
For halogen determinations, the potassium chlorate is replaced by ammonium nitrate.
Another method 6 of filling capillary tubes, which is very simple, consists in
placing the open end of the capillary in the liquid contained in a small sample
bottle. Both ar'e placed in a small filter flask equipped with a rubber stopper
and stopcock, Fig. 407. Suction is applied and the flask evacuated, which also

FIG. 407.

evacuates the capillary. This takes only a few seconds. The stopcock is now
slowly opened and the difference in pressure between the outside and inside of
the capillary forces the liquid into it.
I
6 J. Linder, Mikromassanalytische Bestimung des Kohlensf;_OfIs und Wasserstoffs ~t
Grundlegender Behandlung der Fehlerquellen in der Elementar Analyse Verlag Chemle,
Berlin 1935.

QUANTITATIVE MICROCHEMICAL ANALYSIS

2469

Viscous liquids are weighed directly into boats.

Liquids with high vapor pressure,7 such ,as acetone, ether, etc., are weighed
into a capillary of 1 mm. bore, 30 mm.long wiJrh a hair fine capillary end 0.03 to
0.05 mm. bore, 20 to 30 mm. long. The liquid to be sampled is kept cold and
the main body of the capillary is warmed, e.g. with warm brass forceps, while
dipping in the liquid. It is allowed to cool aud the liquid rises in the capillary.
Before the capillary has completely cooled, it is lifted out of the liquid with
cooled forceps and a small amount of air allowed to enter in order to push the
liquid away from the hair fine capillary. The liquid may now be weighed as
the hair capillary prevents any weighable quantity of vapor from escaping in
Y2 hour. The sample is placed in the combustion tube and is surrounded by a
cooling mixture while the tube is being swept out. The analysis is then
carried out in the regular manner. By using the above method, the vapor
slowly escapes into the combustion tube from the open hair capillary as it is
carefully heated.
.
Gases.-Gases, such as those of the hydrocarbon series, may be introduced
into a combustion system by the method of Wm. F. Bruce 8 (see page 2480).
Treatment of Rubber Tubing.-All rubber tubing used for connections is
placed in a flask with molten vaseline while suction and pressure are alternately
appHed so that the pores of the tubing are completely filled. The special
tubing has an outside diameter of 8 mm. and a bore or' 2 mm. Tubing already
impregnated can also be purchased from supply houses.
, Micro Neubauer Crucible.-The micro Neubauer platinum crucible is used
to filter the precipitate in the determination of sulfur and arsenic. The filtering
medium is composed of a compressed
platinum iridium sponge.
The.
crucible is shown, Fig. 408, attached
to a small filtering flask by means
of glass tube G 10 mm. outside
'<iiameter and rubber sleeve M.9
The crucible is prepared for use by
washing with the aid of suction with
chromic-sulfuric acid mixture and
water .. The crucible is removed and
closed with a lid and lower cap,
ignited on a small square of sheet
platinum, cooled on a copper block,
and weighed.
The micro Neubauer crucible
may be profitably replaced for
routine work by porcelain micro
FIG. 408.
filtering crucibles with a filtering
medium of unglazed porcelain. In the case of porcelain crucibles,. they are
ignited in an electric mume, then, after cooling in the !tir for 2 to 3 minutes,
are placed in a desiccator until they comeio room temperature. They are then
G. Pirsch, Ber., 65, 865 (1932).
Mikrochemie, 18, 261 (1935).
9 The arrangement shown provides for the automatic transference of the precipitate
to the filter crucible.
7

2470

QUANTITATIVE MICROCHEMICAL ANALYSIS

weighed at the end of a definite time, say 5 or 10 minutes. If weighings are

made the same day, the change in moisture content -is :qot enough to cause
serious error when the crucible is ignited and weighed u~der the same conditions. However, when the analysis is not completed the same day or when the
highest accuracy is required, it is best to place the crucibleJ in weighing bottles.
The Filtering Tube.-The hard glass or quartz filtering tube 11 cm. in
length, Fig. 409 A, is first filled with a thin layer of dry Goobh-crucible asbestos,
a suspension of asbestos in water added and
then the asbestos pad is washed with sulfuricchromic acid mixture, water, hot' nitric acid,
water, finally with alcohol and dried at
120-125.
A micro Gooch-crucible may be used.
The filter tube of Pregl also permits the automatic transference, Fig. 409, of precipitates
such as silver halide on to the filtering surface. Hard glass filter tubes with fritted
glass filter beds of varying porosity are available and preferred for routine work.
Removal qf Traces of Precipitate.-A
small feather cemented to the end of a capillary tube cir a small rubber policeman is
used to remove from beakers and test tubes
the last traces of precipitate.
The Immersion Filter.-In the method
used by F. Emich JO the precipitation takes '
FIG. 409.
place in a micro beaker and the mother
liquor and wash water 'are removed by suction through an immersion filter
or "filterstick." 'l;'he filter remains in the beaker and both are weighed
together; thus, the complete transference of the precipitate to the filter is not
required. The method is simple and gives greater accuracy with less skill
than when the filter tube of Pregl is used.
The micro beaker has a capacity of 5-10 ml. and is easily fashioned from a
pyrex test tube. The immersion filter, Fig. 410, is a capillary glass tube, 2 mm.
outside diameter and 5 cm. in length, one end of which is expanded into a funnel
or head which may be varied in shape and size but is conveniently 8 mm. in
length and 6 mm. in greatest diameter. The total weight of-..filter and beaker
-is 6-10 gm.
The filter has a platinum spiral at P, Fig. 410, and on this is.formed an
asbestos mat which is washed and treated in the same way as the Pregl filter
tube. The filter and beaker, which have been thoroughly cleaned, are dried in
the oven at 105-120 C., cooled 'in a desiccator without drying agent, and the
final weight taken after standing 5 minutes on the balance. Care .must be
taken to keep the head of the filter up so that the drh;d asbestos mat will not
fall out.
The precipitation is now made in the beaker and after the precipitate has
coagulated, the asbestos mat of the filter is moistened with a drop-of water and
10 Stahler: "Handb. d. Arbeitsmethoden," V. 1, p. 6~Q.{~ipzig, 1913).
" Mikrochemisches Praktikum," p. 62 (Pub. J. F. Bergmann;-I93I).

F. Emich,

QUANTITATIVE MICROCHEMICAL ANALYSIS

2471

the filter attached to a suctiQn flask by means of a glass tube and rubber connection. The filter is placed in the beaker with the head just above the coagulated precipitate and the liquid withdrawn into the suction flask, Fig. 411. In

FIG. 410.

FIG. 411.

the same way the precipitate is wa~hed as prescribed and the wash water removed. The filter is disconnected and remains in the beaker and they are
both dried and weighed as before.
If the precipitate must be ignited then the filter and beaker are made of
platinum, porcelain or quartz and the filtering medium is
platinum sponge, porous porcelain or sintered quartz. A
thin porcelain crucible of 10 ml. capacity serves well as a
beaker. Where a volume of solution greater than 10 ml. must
be handled a micro beaker developed in this laboratory, Fig.
412, with a ground-glass joint, so that the top section may be
removed, is useful. The precipitation is carried out as usual
and then the filtrate removed by the immersion filter. The
bottom section with the filter stick, which has been previously weighed, is now detached, dried, and reweighed with the
pr_ecipitate. The total weight of the bottom section and filter
stick is 8-10 gms.
Where ignition is required it is best accomplished in an
electric muffle furnace heated to 1100-1200 C. and the beaker
with the filter is placed in a covered unglazed porcelain conFIG. 412.
tainer. After ignition, cooling takes place in the open air for
~ minutes, then in a desiccator with drying agent, and after 25 minutes they
are weighed.
In all the microgravimetric'determinations to be described the immersion

2472

QUANTITATIVE MICROCHEMICAL ANALYSIS

.filter method is applicable, even though it may not be specifically mentioned

in the procedure.
1
Burettes.-For general routine titration work, a 10 ml. burette graduated
in 0.05 ml. is suitable. The burette can be read to a.01 m!. and this is sufficiently accurate for most titrations. Where greater Accuracy is required, the
graduations should be in 0.01 ml. The graduations should extend around the
burette far enough so that parallax can be avoided in taking readings.,
Standard Solutions-Titration.-It is good practice to standardize accurately 0.1 N solutions, care being taken that the solutions are maintained at
20 0 C. 0.01 and 0.02 N solutions are freshly prepared as required by diluting
0.1 N solutions. 0.001 N solutions are not satisfactory as a rule for routine
work because of rapid deterioration. Because 0.1 N solutions yield such a
large volume on dilution to 0.01 N, much time and labor will be saved if the
reagent for the 0.1 N solution is purchased already weighed out and is then
diluted to a liter. For most work the accuracy of these reagents is quite
sufficient. It is desirable that the solutions be kept in a constant temperature
cabinet at 20 C. In summer the use of a constant temperature water jacket
on the burette is required for accurate titration, unless air conditioning is
available. Most 10 ml. burettes yield too large a drop to qbtain an accurate
end-point. It is best to wash off the tip of the burette just before commencing
the titration, and near the end-point to split the drops and succ.essively wash
0.01-0.02 of a ml. into the flask, until the end-point is reached.
Because of the precision required in'micro work all volumetric apparatu.s
should be standardized, preferably by the U. S. Bureau of Standards.
Colorimetry.-It is possible to match colors in the old style colorimetric
tubes. It is much better, however, from the standpoint of speed and accuracy
to use a microcolorimeter. These are available in 1 ml. capacity. Some special
~
.
forms have a capacity a" low as 0.1 ml.
The use of photoelectric colorimeters 11 is of advantage in routine work
where the instrument can be standardized for a particular procedure.
The Heating of Combustion Tubes.-Combustioh tubes may be heated
with gas or electricity. Gas has the advantage that it will heat the tube up to
temperature quickly. Gas burners are readily available in almost every laboratdry. However, it has the disadvantage of uneven heating and causing temperature changes in the room as it is turned off and on throughout the day.
Rapid variation in temperature and uneven heating are avoided by the use
of electric heating units., Electric heating requires special furnaces which are
so designed that if necessary they need not be shut off at the end of e~ch run
for cooling of the combustion tube. By this meanS there is no loss of time from
one run to another in getting the furnaces up to temperature:
Apparatus for determinations requiring combustion is usually equipped wit.h
gas heat. The author has used both gas and electric heating but favors the
latter for routine work. While the methods desc~ibed for carbon, hydrogen,
nitrogen, halogen, and sulfur use electric furnaces, gas furnaces may be substituted. In the case of carbon and' hydrogen, the lead dioxide is_heated with
0
lj, mortar filfed with cymene which boils around 180
C. Some laboratories use

Withrow, Shrewsbury, and Kraybill, Ind. Eng. Chern" Anal. 'Ed., 8, 214-19 (1936);
Ralph H. Miiller, Ind. Eng. Chern., Anal. Ed., 7, 223-6 (1935).
11

QUANTITATIVE MICROCHEMICAL, ANALYSIS

2473

electric heating for that section of the tube containing the catalyst or oxidizing
agent, but burn the sample with the gas burner.
The electric furnaces, Fig. 414, for the carbon and hydrogen apparatus are
constructed by winding nichrome wire on a 6 Y2 and two 3 inch sections of a
Y2 inch inside diameter porcelain tube. The 3 inch furnace used to heat the
lead dioxide' is wound with No. 30 wire in three layers separated by alundum
cement and asbestos paper. The other two furnaces have two layers of No. 25
wire. The wound cores are hardened by drying at 1500 C. The ends of each
furnace are made of circles of Ysft Transite, 3 inches in diameter. Each furnace
is covered with an aluminum or brass shell. Rheostats are used to adjust the
furnaces to proper temperature. Ammeters may be put in the circuit to see
that the proper amount of current is passing through. Thermocouples may be
embedded in each furnace and the temperature may be automatically regulated
if the voltage varies. It is best that the, lead dioxide furnace be regulated to
0.2 C. because variation will cause the material to give off or take up water.
The furnaces, Fig. 423, for Dumas nitrogen are of the split type. When
the.lower half,swings down the upper section may then be raised slightly and
the furnace pushed away from the combustion tube. The furnace may then
be closed and thus kept up to the proper temperature for the next determination.
In the drawing the furnace N is shown in the position away from the combustion
tube, and furnace 0 is shown in the position over the combustion tube. The
furnace N is provided with two notches, one of which shows at V, the other at
the same relative position, which fit on the rail X so that this furnace may be
pushed along the combustion tube.
The furnaces for the halogen-sulfur determination, Fig. 428, are also of the
split type, but cannot be shoved back from the combustion tube. The upper
half swings open and automatically locks in position. The combustion tube
may then be removed and placed on the supports on the front of the apparatus
to cool. By pushing the lever at the back the furnaces may be closed and
thereby kept up to temperature for the next run.
The construction is shown by the drawings. The furnaces are 3 inches and
7 inches in length and 3%, inches in diameter. The ends are made of transite.
The shell is made of aluminum. Fig. 428 with the top section of the small
furnace'removed shows the details of construction of the heating elements. The
heating element of each half consists of two coils of No. 25 nichrome wire supported on Ys inch alundum rods. This in turn is held loosely in place by
notched porcelain semicircles which are cemented to the alundum refractory.
This refractory consists of a stock alundum tube 1 inch inside diameter, sawed
in two. The top and bottom nichrome heating elements are connected in
series to binding posts on the transite e~ds. The heating elements are easily
replaced at any time. The furnace temperature is measured by a thermocouple
and, if required, suitable resistance is inserted.
For laboratories having a large number of routine organic combustions,
the split type electric furnace unit 12 shown in Fig. 413 can be used to advantage.
The apparatus is fabricated from aluminum and stainless steel. The drawing
shows the small burning furnace, tilted back corresponding to a position away
12Irul. Eng. Chem., Anal. Ed., 10, 110, 111 (1938). This paper gives details of the
application of the apparatus to the determination of carbon and hydrogen, nitroge~,
sulfur, and halogen.
.

2474

QUANTITATIVE MICROCHEMICAL ANALYSIS

from the combustion tube which is not shown, ana the large furnace in a position
over the combustion tube. A small electric motor with 'gear reduction driv{)s
a screw which moves the sample burning furnace along 'the cOII\bustion tube.
When the small furnace is brought in the burning positic;d' over the combustion
tube, it meshes with the screw, the motor starts and a brush contact, attached
to the bottom, passes over it series of insulated metal segments as the furnace
moves along. These segments, connected through suitable resistances, auto. matically vary the speed to give slow initial burning and accelerated burning
after the sample has carbonized. The furnace automatically stops when it has
reached a predetermined point and the sample has been burned. Any combination of speeds can be obtained by varying the number, positions: and length
of the segments. The heating elements are the same as in the hand operated
furnace.

FIG. 413.

The application of this type of burning furnace to the determination of

carbon and hydrogen, halogens, and nitrogen increases the number of determinations done per day by about 25%. There is less fatigue for the operator'
and more time available for weighing, calCUlating (and titrating while the
samples are burning.
Many laboratories run combustions only occasionally and a general utility,
combustion unit for all combustions is desirable. By const:r;ucting a three-unit
split type which can also be pushed away from the combustion tube, the needs
of the small laboratory are met. When running Dqmas nitrogen, sulfur, or
halogen, the small unit for heating the lead dioxide ill the carbon and hydrogen
determination may be removed.
'

QUANTITATIVE MICROCHEMICAL ANALYSIS

2475

Microchemical Equipment.-Practically all apparatus for the procedures

here de~cribed can be readily obtained from supply houses in the United States.

DETERMINATION OF CARBON AND HYDROGEN

IN ORGANIC MATERIAL *

The method 13 described gives a high degree of accuracy together with speed,
and can therefore be used to advantage in routine work.
The procedure is similar to the combustion method which is known to all
those. who have had occasion to determine carbon and hydrogen in organic
comp0'\lllds. The results obtained are as good as, and in many cases better
than, those obtained by using the macro method. Results averag91 O.l %
from theory for compounds containing carbon, hydrogen and oxygen. Those
-containing sulfur, halogen and nitrogen vary O.2% from theory.

FIG. 414.

Apparatus.-A general view of the apparatus is given in Fig. 414. The

pressure regulator A is half filled with concentrated sulfuric acid and is then
connected at the point B by a glass tube with rubber connection to a cylinder of
oxygen, preferably made from liquId air. The pressure regulator is attached
with a rubber connection to a small bubble counter C and drying tube D, made
of pyrex glass fitted with a ground-glass joint. The bubble counter is partly

* For an automatic combustion apparatus, see p. 2473 and ref. 12.

18

Kemmerer and Hallett, Ind. Eng. Chem., 19, 173 (1927).

2476

QUANTITATIVE MICROCHEMICAL ANALYSIS

filled with cohcentrated sulfuric acid by dipping the end into a beaker of the
acid and applying gentle suction on the drying' tube U. The excess acid is
then removed from the end by carefully heating over abuf!1er. The drying tube
is charged two-third! of its length with" Ascarite" (Trade name for sodium
hydroxide on asbestos), fine glass wool being placed a~ bo.th ends (Corning
standard wool No. 008); then a layer of phosphorus pentoxide mixed with
fine glass wool; and at the end another short layer of glas~ wool. The groundglass joint is then heated, the two parts coated with a small amount of glass
cement, and after the joint is firmly pressed together it is allowed to cool.
A rubber stopper is placed on the tapered capillary end of this joint. The
pyrex, or preferably quartz combustion tube, 53 cm. in length and 7 mm.
inside diameter, has a capillary opening 1.5 to 2.0 cm. in length. The combustion tube is charged, beginning at the capillary end, with a short wad
of ignited asbestos, then silver wool, 1.5 cm.; powdered lead dioxide mixed
with asbestos or granular lead dioxide, 4.5 cm.; silver wool, 3 cm.; finely broken
copper oxide wire impregnated with lead chromate (by igniting and sprinkling
with lead chromate which fuses on the copper oxide), 15 cm.; silver wool,
3 cm. jn ignited asbestos wad is placed between each of these charges
to prevent shifting. Care must be taken not ,to pack so tightly as to stop
the flow of gas through the combustion tube, Silver removes any halogen
and sulfur present, lead dioxide reacts with the oxides of nitrogen, and lead
chromate also removes oxides of sulfur. Lead dioxide is heated by the electric
furnace G to 1800 C., copper OXIde to 700..::750 0 C. by the furnace F, and the
third furnace E is employed to burn the material in the boat, the temperature of
which if necessary may attain 10000 C. The temperatures of the first two
furnaces are accurately measured by means of a thermocouple, and as each
furnace has a rheostat and a means of reading the current by a small ammeter,
it is possible to regulate the temperature accurately. The control of the temperature of the lead dioxide is very important as fluctuations in temperature
cause it to give up or absorb water vapor. The furnaces are placed on two rods
so that they may be moved along the combustion tl:lbe at will. At the point H
a small electric heater is placed in order to prevent condensation of moisture.
The heat applied should be regulated so as not to damage or destroy the rubber
connection. The combustion tube when freshly charged is kept at full heat for
not less than six hours while a slow stream of oxygen is passed through the tube
in order to burn organic matter and free the tube and filling from absorbed
gases, such as water vapor. '
Absorption Tubes.-Because the oxygen used in the combuStion is not
swept out with air and since phosphorus pentoxide is used for the absorption of
water, the absorption tubes must be capable of being sealed in order to prevent
the escape of oxygen or the entrance of moisture during weighing.
The absorption tube, Fig. 415, is made from a 6 mm. pyrex ground-glass
joint. The main body of the tube is enlarged until1t is 7 cm. in length and
1.3 cm. in diameter with thin walls to keep the weight as small as possible.
The mercury bulbs C at each of the tubes are identical, the internal diameter
being 3-4 mm. The internal' diameter at A and B is 0.2 mm. As an added
precaution bulbs 2.5-3.0 mm. are blown between the points Bland D to catch
any mercury which might be carrIed through. The..stPiiight part of the tube D
has an internal diameter of 1 mm. and for a length-of 1'.5' cm. an external diam-

QUANTITATIVE MICROCHEMICAL ANALYSIS

2477

eter of 3 mm., which fits the special micro tubing used. In Fig. 415, F is solid
glass to strengthen this part of the tube. The.mercury is shown sealing the
capillaries. When the tube is rotated 1800 and shaken slightly, the mercury
f~lls and the tube is open. The ground-glass joint is sealed with glass cement.

c
B
E

Mer,ury

FIG. 415.

The total length of the carbon dioxide absorption tube is about 12 cm. and
that of the water absorption tube slightly less. It is very important that these
tubes be washed in chromic-sulfuric acid mixture, and then with water, followed by alcohol and finally ether, so that the main body of the tube is clean
'arid also the capillaries on eit~er side of the mercury chamber.
A modified construction for these tubes has been described.1 4 The mercury
trap and seal are placed within two hollow glass stoppers which are ground to
fit the ends of a Pyrex tube 1 cm. in diameter, Fig. 416. The tube is shown

FIG. 416.

unsealed in the figure. When the tube is rotated 1800 the mercury drops seal
the capillaries.
The water absorption tube K, Fig. 414, is charged first with a wad of fine
glass wool (Pyre:lf.: Standard); then phosphorus pentoxide mixed with a small
amount of fine glass wool; and then a wad of glass wool. The end of the tube is
then heated and sealed with glass cement. A small drop of mercury is then
drawn into each end by J;tleans of gentle suction.
The carbon dioxide absorption tube L is charged first with a short layer of
glass wool; then three-fourths with 20-mesh "Ascarite"; a short layer of fine
glass wool; then a short layer of phosphorus pentoxide; and finally a wad of
fine glass wool. The tube is then sealed and mercury drawn in.
The ends of the tubes in all cases must be glass to glass in order to minimize
the diffusion of gases through the tubing. The charges in both tubes will last
for 30 to 40 determinations. The" Ascarite" changes color from a light brown
to a grayish white upon absorbing carbon dioxiae. A guard tube M, filled
witli phosphorus pentoxide or "Dehydrite," is placed at the end of the carbon
dioxide absorption tube and the guard tube is attached to the Mariotte flask.
The Mariotte flask or aspirator bottle S is used to measure the amount of gas
passed through the combustion tube and aids in keeping the system at atmos14

Cornwell, Ind. Eng. Chern., Anal. Ed., 3, 4 (1931).

2478

QUANTITATIVE MICROCHEMICAL ANALYSIS

through the rubber

pheric pressure, and thus minimizes diffusion of 'gases
connections.
By means of a three-way stopcock N between the flask and guard tube the
speed of the oxygen can be regulated. It is also possible by means of stopcock
N to open this line to the atmosphere before removing the absorption tubes
and thus create an outward pressure which prevents the absorption of moistur~
and carbon dioxide from the air. By opening stopcock 0 to the air, and attaching tube P to a ,water tap, it is possible to fill the bottle without removal from
the stand. Tube R runs to the sink and is constricted, or may have ~ stopcock,
so that while the flask is being filled, water will not readily escape..
Procedure.-:-The furnaces are turned on and also the small heater H; the
absorption tubes are wiped with a damp cloth, dried, and after unsealing (by
rotating 180 0 and tapping lightly), are attached to the combustibn tube. 'l'he
oxygen is turned on and the pressure in the ThIariotte flask regulated so that the
gas passes at the- rate 9f 10 ml. per minute, and is allowed to pass while a 3 to 7
mg. sample is being weighed in a cleaned and ignited porcelain or platinum
boat, suitable precautions, being taken if the substance is hygroscopic Or a
liquid. The heater II is turned off 5 minutes before the absorption tubes are
removed. The stopcock N on the oxygen line T of the ThIariotte flask is turned
so that the oxygen flows into the air. The.guard tube is removed while oxygen
continues to flow and then the" Ascarite" tube L is removed. The tube is
rotated 180 0 and is tapped lightly causing the mercury to seal the capillaries.
The ph.osphorus pentoxide tube is handled in a similar manner. In remoVing
the tube only the ends and not the main body are touched, and these are CEt,refully wiped each time. It is only necessary to wipe the main body of the
absorption tube at the beginning of the day's run to remove accumulated dust.
The phosphorus pentoxide tube is placed on the balance, its approximate
weight obtained, and is allowed to stand 10 minutes before the final reading
is taken. The other tube is ,then weighed in the same manner, after remaining
3 minutes in the balance. "\Vhile the first tube is on the balance the boat Containing the weighed sample is placed in the combustion tube by means of
platinum-tipped forceps and a clean platinum or nichrome wire. The flow of
oxygen now sweeps out any moisture and carbon dioxide which has entered.
The absorption tubes are attached, the flow of oxygen is regulated,_to 5-6 m!. per
minute by the stopcock N, the combustion furnace is turned on, and the sample
is burned by slowly moving the furnace over the boat. The small heating t::oil
at the capillary end of the combusti.on tube is turned on to prevent the Condensation of :~Yater vap9r._ The actual burning or the sample takes about 5--10
minutes and', another 20-30 minutes is allowed to sweep .out the products of
combust job., 100
of oxygen during the first 15-20 minutes, another 200 ml.
in 20 minut~s. In most cases a total of 200 ml. of oxygen is sufficient. 1:'he
correct amount sho!;Ild be checked by running standard samples. The absorp_
tion tubes are then removed, sealed, and weighed as' before.
}Vhile these are coming to equilibrium in the balance, the duplicate sample
is placed in the tube and after the tubes are weighed and attached, another
analysis is made in the same manner. From the weights of carbon dioxide B.nd
water the percentage .of c'arbon and of hydrogen in the sample is calculated.

ml:

QUANTITATIVE MICROCHEMICAL ANALYSIS

2479

NOTEs.-(l) The sample should not be oxidized so rapidly that the gases formed are
forced into the tube D. A capillary is placed in the tube to help prevent this.
.
(2) The current of oxygen should be carefully regulated so that the gases formed may
pass over the several fillings at such a rate that they are completely oxidized and
absorbed.
(3) The systeI;;,\ should be tested for leaks by turning off stopcock N and observing
the bubble counter C.
, (4) The blank test is very important. This is carried out in exactly the same manner
as in an ordinary combustion except that no sample is placed in the boat. If the absorption tubes show an increase of 0.1 mg. or IJ;lore, the apparatus should be inspected for the
following; leaks, el'cess glycerol on the rubber connections,
insufficient time allowed for burning and conditioning a
new tube filling, or oxygen containing impurities. Before
beginning a series of determinations a standard should
be run, preferably containing those elements present in the
samples to be analyzed.
(5) The absorption tubes and drying tube D are refilled
by heating the ground glass joint, which melts the glass
cement and allows the tubes to be emptied and refilled.
a
(6) Because phosphorus pentoxide and" Ascarite " are
used, oxygen may be passed through the apparatus at a
more rapid rate than that recommended by Pregl. For
compounds containing much nitrogen and sulfur the rate
should be slower, not more than 5 ml. per minute, so that
the oxides of nitrogen and sulfur will be reacted upon by
the reagents in the tube.
(7) The absorption tube next to the furnace should
always be placed with the ground glass joint away from the
furnace to avoid softening of the cement.
(8) The oxygen may contain organic impurities in
which case a preheater is used consisting of a combustion
chamber filled with platinized asbestos or CUO heated to
650-700 C. by an electric furnace or burner. The gas
b
is usually cooled by passing through a glass coil immersed
in water so that it will not decompose the rubber connections. The gas is then passed through" Ascarite " and
phosphorus pentoxide before entering the combustion tube.
A common type I. is shown in Fig. 417. Another simpler'
form made of Pyrex ,and without a ground_glass joint was
FIG. 417.
desigiled by White and Wright.I. "The apparatus is
constructed, Fig. 418, from 10 mm. Pyrex tubing, into which two 4 mm. tubes, AB
and BC, are joined by ring seals at B. The gas inlet and outlet being attached, the
bottom of the tube is sealed off, and, in operation, may be immersed in a beaker of water
to cool the burned gases. Into the annulus around BA is packed 1.5 em. lengths of
coarse fibre asbestos, and then fluffy platinized asbestos. The top of BA is then loosely
packed with coarse fibre asbestos and the top of the 10 mm. tube is sealed off. An electrical heating unit surrounds the top of the preheater. It should not fit too tightly
and should be closed at the top with a stopper of asbestos paper which can be removed
in order that the correct (dull red) temperature may be ascertained."
(9) Where compounds contain arsenic, antimony, tin, bismuth, or phosphorus, the
following procedure 17 is recommended which is a modification of the Falkov and Raiziss
procedure. '8 A portion of the copper oo:ide-Iead chromate filling of an ordinary Pregl
Universal filled combustion tube is removed and replaced by a 3 cm. cylinder made of
200 mesh platinum wire gauze filled with granulated lead and a 1 cm. plug made of
I. Nieder! and Roth, Ind. Eng. Chern., Anal. Ed., 6, 273 (1934).
Can. Jour. Res., 14, Section B, 427 (1936).
17 F. C. Silbert and W. R. Kirner, Ind. and Eng. Chern., Anal. Ed., 8, 353-5 (1936).
18 J. Am. Chern. Soc., 45, 998-1003 (1923).
16

2480

QUANTITATIVE MICROCHEMICAL ANALYSIS

platinum wire gauze. The red lead is prepared by igniting Merck's granular lead peroxide in a stream of oxygen in a microcombustion tube in an ele'ctric combustion furnace
at the normal combustion temperature. To prevent any red lead dust from sifting
through the platinum gauze a thin layer of asbestos lines the entire inner surface of the
,
cylinder, including the ~nds. This snug fitting
- - - - - - -....- cylinder enSUres more intimate contact of the
combustion gases with the red lead than if the
latter is merely placed in a boat as recommended
by Falkor and Raiziss. The details of the filling
are shown in Fig. 419.
(10) Analysis' of Coal for Carbon and
Hydrogen.
E
No special technique is required, but the
E
sample must be finely ground (about 100 mesh)
in an agate mortar. For accurate analysis, it is
necessary, of course, to dry the sample, weigh in
a weighing bottle, and then transfer it quickly to
the combustion tube before it takes up moisture.
Moisture cannot be determined by micro
methods because this figure should represent a
large mine fun sample. The sampling of the
coal is, likeWise, done in the usual macro way.
The final sample which is submitted to the
analyst is then taken and prepared for analysis .
. Kirner 19 gives the following example: 50 gms.of
German bituminous coal in fairly .large pieces
were crudely ground in a mortar, sample A.
Five grams of sample A were pulverized for 5
minutes, sample B. Sample B did not give concordant results but on taking 1 gram of sample B
and pulverizing in an agate mortar for 5 minutes,
it gave satisfactory checks. Lucas and Grassner 20
take the final 125 gm. sample, grind to 100 mesh,
and take 10 mg. of this for a determination.
(11) Solid compounds which are explosive
when heated are mixed in the boat with fine
copper oxide or some inert material such as
ignited finely ground ;iilica.21
(12) Wherel compounds are run which contain carbon, hydrogen and oxygen, the tube 'filling can be much simplified. Only platinized
asbestos, or better, platinum contacts are required to attain complete combustion. This
filling is heated by the furnace which heats the
lead chromate-copper oxide mixture. Because
nitrogen is absent, no lead dioXide is required.
c
(13) If the apparatus has been standing for
a week or so, it is best to burn in the combusFIG. 418.
tion tube a few crystals of sugar or, preferably,
a few milligrams of a compound containing the
elements present in the compounds to be analyzed. It.sometimes happens that the
first run is low in hydrogen due to the lead dioxide a\isorbing moisture. By first
burning an organic compound, the water-lead dioxide equilibrium is made constant.
(14) A side inlet combustion tube is often preferred, especially for ~arbon and hydrogen work. _ A solid rubber stopper is then used.
(15) Analysis of Gases for Carbon and Hydrogen.
The method essentially consists of measuring the gas accurately in a gas burette,
19
20
,21

Ind. Eng. Chern., Anal. Ed., S, 363-9 (1933). . _ . r Mikrochem., 6, 116-32 (1928).
M. Furter and J. L'Orange, Mikrochemie, 17,38-42 (1935).

'\

QUANTITATIVE MICROCHEMICAL ANALYSIS

2481

mixing it with oxygen, 'and then passing it into the combustion tube. A special tube
with side inlet, Fig. 420, filled with CO 2 absorbent, such as " Ascarite," is placed between
the pressure regulator and the drying tube with bubble counter. The measured gas
sample, plus an excess of oxygen contained in a t!;as burette, is attached to the inlet A
by a tight glass to glass connection of capillary tubing. The apparatus is swept out with
air or oxygen. The oxygen from the pressure regulator is cut off. The gas sample
ElectriC CombustIon Furnace

Ag Pt Pt.+ Pb 3 0 4

FIG. 419.
mixed with oxygen is now admitted from the gas burette at the same rate as in a regular
combustion. When all the gas is out, the burette cock is closed and the method of
sweeping out is the same as with the standard procedure.
(i6) Many absorption tubes have been designed which can be sealed. Those
which have a ground-glass joint, which must be turned and lubricated, may give difficulty due to loss of lubricant when wiped. The use of copper wires in the capillaries of

. FIG. 420.
the Pregl tube, Fig. 421, to help prevent entrance of moisture and carbon dioxide have
been suggested by Boetius 22 and used by Niederl and Roth.23 The tubes however are
fi!led with air when weighed.
The objection raised to a sealable tube is that changes in temperature and pressure
affe'<,t the constancy. The author has found, however, that during 45 minutes, which

FIG. 421.
is the time taken to make a carbon and hydrogen determination, it is rare that there is
so great a change in pressure as to affect the accuracy of the determination. The temperature of the laboratory'does not vary enough to make it a factor. For routine work,
a sealable tube which can be weighed while filled with oxygen is much more convenient.
The wiping at the end of each run of the main body of the absorption tube has not been
22 Uber die Fehlerquellen bei mikroanalytischen Bestimmung des Kohlen-und
Wasserstoffes nach der methode von Fritz Pregl, p. 27, verlag chemie Berlin 1927.
,
23 Ind. Eng. Chem., Anal. Ed., 6, 272 (1934).

2482

QUANTITATIVE MICROCHEMICAL ANALYSIS

found necessary. References to the different types of sealable absorption tubes are

given by A. Friedrich. 24
(17) The water absorption tube may be filled with a desiccant such as " Anhydrone "
(anhydrous magnesium perchlorate) and then ,finally a 72 inch layer of phosphorous
pentoxide, t? indicate when the tube m~t be refilled. Indi~ating" Drierite" (anhydrous calcIUm sulfate) may also be used m the water absorptlOn tube. It is impregnated with a cobalt salt which changes to a reddish color on taking up water. It can
be regenerated at least 12 times by heating to 235-250 C. for t to 2 hours.
(18) Good results have been obtained by replacing a part of the lead chromate-copper
oxide mixture with a 4 cm. section of platinum contact foil or a roll of platinum 80 mesh
gauze.
(19) The lead dioxide furnace should not be shut off at the end of a series of runs.
By always keeping the filling up to temperature the lead dioxide water equilibrium is
more readily maintained.
(20) One combustion tube filling will last for 200 determinations. If halogens are
frequently run the renewal of the silver wool nearest the boat after 20-30 such determinations is desirable.
(21) Compounds which are salts of alkalies or alkaline earths must be mixed with
dry pulverized potassium dichromate in a porcelain boat in order to prevent the formation of carbonates.

DETERMINATION OF CARBONATE CARBON

The carbonate is decomposed with hydrochloric acid and the evolved carbon
dioxide absorbed and weighed. The method 25 is accurate when samples containing not less than 1.5-2.0 mg. of carbon dioxide are used; with smaller
amounts of carbon dioxide, the error may reach 2%. The values for carbonate
,
carbon vary 0.3-0.4% from t h e o r y . '
Apparatus.-The apparatus, Fig. 422, is construcled of Pyrex glass. A
rubber stopper is used in the 50 m!. extraction flask A, so that flasks may be
interchanged. The water condenser is approximately 20 cm. long with the
central tube beveled on the lower end B, so that it will not be closed by drops
of condensed water. The trap P containing sulfuric acid, is sealed between the
condenser and drying tube,R, which latter contains the solid. drying agents to
absorb most of the water before it enters R. This trap also indicates. the rate of
flow of gas through the apparatus. It can be readily emptied by _Pleans of the
stopcock at the bottom and refilled through the same by applying ,suction to
the apparatus. The drying tube, R, 26 cm. long"is made of a 1.7 cm. groundglass joint, which is sealed with glass cement.
The drying tube is filled with magnesium perchlorate trihydrate, "Dehydrite" or anhydrous magnesium perchlorate, " Anhydrone.': This is powdered
so that about half passes through a 20-mesh sieve, the two portions being kept
separate. The tube is charged first with a wad of fine glass wool, then a 1 cm
l~yer of the coarser "Dehydrite"; fine glass wool; 1 cm. layer of anhydrous
24
25

Mikrochemie, 19, 23--37 (1935).

. -' ::.. Kemmerer and Hallett, Ind. Eng. Chem., 19, 1352 (1927).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2483

copper sulfate to show the. presence otwater; fin~ glass wool and alternate 1 cm.
layers of coarse and fine" Dehydrite" or "AnhydronEi." If sulfites 26 are present the carbonate is decomposed with an excess of potassium dichromate 5%
solution, boiled, and toward the end of the procedure, a small amount of dilute
sulfuric acid is added.

The acid bulb, W, supplied with an Ascarite-filled guard tube, T, to prevent

absorption of carbon dioxide from the air, is attached to the flask A through a

FIG. 422.

three-way stopcock. The air passed through the apI)aratus to remove carbon
dioxide is first drawn through a sulfuric acid trap S, which is charged with acid
through the funnel and may be emptied by disconnecting it at the rubber connection and blowing out through L. The air then Passes through the tube D
which is fil~ed with "Ascarite" to absorb any carbcm dioxide present. The
constriction N retards the diffusion of the carbon dio)(ide liberated in A. The
whole apparatus is easily attached to a single ringstand.
26

E. R. Marle, J. Chern. Soc., 95, 1491 (1909).

248.4

QUANTITATIVE MICROCHEMICAL ANALYSIS

Absorption Tube.-The carbon dioxide absorption tube' (see Fig. 415) is

.charged with" Ascarite" and a short layer of "DehYdrite". or" Anhydrone."
For very small amounts of carbon dioxide, 1 mg. or less, it is advisable to use
phosphorus pentoxide instead of "Dehydrite" in thJ drying tube R and the
absorption tube. The absorption tube is connected to the outlet of R. and a
guard tube charged with drying agent is attached.
I
Procedure.-When the apparatus is freshly filled, carbon dioxide is passed
through and allowed to stand for 2 hours or preferably overnight. The carbon
dioxide is then removed by passing air through the apparatus. This operation
is repeated whenever the trap P is recharged with sulfuric acid.
Five to ten mg. samples are placed on small watch glasses made by cutting
the bottom from a test tube, weighed, and transferred along with the watch
glass to the flask A. The flask is then attached to the apparatus and air
drawn through for 10 minutes to remove any carbon dioxide present. Approximately 2 N hydrochloric acid, which has been boiled to remove carbon
dioxide, is placed in the bulb W. The absorption tube, after careful wiping, is
attached to the apparatus.and air is drawn through. This fills the tube with.
dry air similar to that which fills it at the end of the determination.
As the apparatus is under a slight vacuum it is necessary, before removing the
the absorption tube, to bring the tube to atmospheric pressure by removing the
suction line from the guard tube or by opening a three-way stopcock placed in
the suction line beyond the guard tube. One minute should elapse before
removing the tube. The tube is then removed, handled and weighed as
described under carbon and hydrogen. It is then connected to the apparatu's
and the guard tu]Je attached. Air is aspirated through the apparatus at the
rate of 12 to 14 ml. per minute, and 5 ml. of hydrochloric acid allowed to drop
slowly into A. When the reaction is complete the current of air is passed for
5 minutes to remove most of the carbon dioxide. The flask A is then heated
with a micro burner and the solution boiled 2 to 3 minutes to compiete the
removal of gas. The air current is allowed to continue for 20 minutes or more
after the boiling is finished. The suction is then st,6pped for Iminute and the
tube is removed as before, and weighed. While the tube is coming to constant
weight on the balance, a second sample, which has been weighed and placed in
a clean flask, is attached to the apparatus which is then swept out with air as
before. The second determination is carried out exactly as the first. From
the weight of carbon diqxide the percentage of carbonate_ carbon in the sample
is calculated.
' -''__
NOTEs.-(l) Where oxides of nitrogen are given off 27 powdered lead dioxide mixed
with asbestos and heated in a glass tube with an electric furnace to 180 0 C. must be
inserted in the train followed by a tube containing the drying agent. The carbon dioxide
absorption tube is now attached as before.
27

Kenyon and Gray, J. Am. Chern. Soc., 58, 1422-7

(1936).
I

QUANTITATIVE MICROCHEMICAL, ANALYSIS

NITROGEN BY THE MICRO-DUMAS METHOD

2485

The micro-Dumas method 28 consists in decomposing the substance by heat

in a mixture of fine coyper oxide in the presence of carbon dioxide. The carbon
dioxide is absorbed in a ca:ustic potash solution and the nitrogen measured in a
precision nitrometer. Results average 0.1 % from theory.
Apparatus: The Carbon Dioxid_e Generator.-The Kipp generator shown
in Fig. ~23 is designed to q,eliver carbon dioxide of such purity that 100 ml. of

()

10

SCALECM.

FIG. 423.

gas contain only 0.001 ml. which is not absorbed by a 50% potassium hydroxide
solution. In order to keep the acid out of contact with air the upper and middle
chambers are connected with glass tubing containing ground glass joints. The
acid used is made by mixing equal parts of freshly boiled distilled water and
concentrated hydrochloric acid. The marble chips used are etched in dilute
hydrochloric acid and then washed with water before they are placed in the
generator, in order to remove absorbed gases.
When the generator is assembled and charged, the upper chamber is freed

* For an automatic combustion apparatus, see p. 2473 and ref. 12.

Trautz, Mikrochemie, 9, 300 (1931); Trautz and Niederl, Ind.Eng. Chern., Anal.
Ed., 3, 151 (1931).
28

2486

QUANTITATIVE MICROCHEMICAL ANALYSIS

from air by adding a few lumps of marble, the U tube A is put in place, and the
upper B and middle chamber C connected by a glass to ~lass rubber connection
at E. Carbon dioxide is allowed to flow from the middle C to the upper chamber
until there is a positive pressure shown by the mercury in the U tube which
seals the upper chamber from the atmosphere. An excess pressure should
exist within the upper chamber at all times and the U t',ube charged with me'i-cury allows the pressure or partial vacuum to be read. A column of mercury
5 cm. high in one arm is sufficient to take care of the variations in pressure.
Before an analysis and at other times when necessary, carbon dioxide is passed
from the middle to the upper chamber by turning stopcock F until gas escapes
through the valve. By removing the U tube spent acid may be siphoned off
and fresh acid added.
A gasometer K containing mercury is placed between the generator and
combustion tube in order to measure the volume of carbon dioxide used during
a determination. Specially treated and aged rubber tubing previously described (page 2469) is used for glass to glass rubber connections at the points
H, P and R. The diameter of the glass tubing at these points is 4 mm. The
gasometer is connected at M to the combustion tube by a capillary tube,
tapered at the end to accommodate the rubber stopper which fits into the
combustion tube.
Combustion Tube.-A Supremax or preferably quartz combustion tube,
50 cm. in length, and 7 mIll. inside diameter has a'l to 2 mm. capillary opening
1.5 to 2.0 cm. in length. This tube is the same as that used' for the determination of carbon and hydrogen.
The material used to charge the combustion tube is first ignited. The
tube is permanently charged, beginning at the capillary end, with a short wad
of asbestos, then broken copper oxide wire, 9 cm.; a short wad of asbestos, then
4 cm. of freshly prepared broken copper wire. This is made by passing hydrogen (washed with a solution of potassium permanganate acidified with sulfuric
acid) over heated broken copper oxide wire and cooling in a current of hydrogen.
A wad of asbestos is added, then 9 cm. of broken copper oxide wire. This is
held in place with a wad of asbestos. Before using, the charged tube is preheated in a current of carbon dioxide and then allowed to cool.
Nitrometer.-The precision micronitrometer :S, Fig. 423, has a capacity of
1.5 rol. and is read to the nearest 0.001 ml. It is charged by adding mercury
until it covers the gas inlet. Fifty per cmit caustic potash solution made by
adding 100 gm. of potassium hydroxide to 100 gm. of water, -is tlten added by
the funnel to fill the measuring tube.
Procedure.-A sample which will yield 0.3 to 0.5 ml. of nitrogen is placed
in a dry, weighed porcelain boat, dried, and then reweighed. If the material is
in large crystals, it should be ground in an agate mortar. Fine freshly ignited
copper oxide (prepared by grinding broken copper oxide wire until it passes a
60 mesh sieve) is added to the combustion tube until;a layer of 2 cm. is formed,
then the boat containing the sample is added, and the tube gently shaken and
rotated until the sample is mixed with a fine copper oxide. Anoth~r 2 cm. of
fine copper oxide is added and then a 4 cm. layer of coarse copper oxide. The
end of the tube is wiped with absorbent cotton to remove any' fine copper oxide
which would prevent the rubber stopper from making 3; tight fit.

QUANTITATIVE MICROCHEMICAL ANALYSIS

2487

A blank determination using pure cane sugar should be made when the
procedure or apparatus is modified, in order to determine the volume of unabsorbable gases which enter during each run. A stand~rd sample should be
run before making a" series of determinations in order to see that the apparatus
is working properly. Liquid materials are weighed in a sealed capillary, the
tip broken, and after adding 2 cm. of fine copper oxide, the capillary is dropped
into the tube (brol>:en tIP first). It is then covered with another 2 cm. of fine
copper oxide and finally with 4 cm. of coarse copper oxide.
. After introducing the sample, the combustion tube is attached to the
apparatus and the air is eliminated by passing through the system about 100 ml.
of carbon dioxide, while cold. At the end of this operation, which requires one
minute, the nitrometer is attached and small micro bubbles should appear in
the nitrometer. ,Bubbles of proper size are recognized by their uncertain rise
so that if the gas flow is sp~eded up the streaJ7l of bubbles which are no larger
than pin points have a tendency to turn over and go toward the bottom of the
nitrometer. If the bubbles are not of micro size, it shows that the air has not
been eliminated or that a leak in the system is present. In the 'next operation,
the actual eombustion stopcock L of the gasometer is closed, both furnaces are
brought over the tube, and the combustion tube is heated to red heat (700750 0 C.). Care must be taken that furnace N does not come too close to the
sample and cause premature combustion. After the bubbles have ceased to
come over, due to the heating of the tube, any gas which has collected in the
nitrometer is now released. The burning of the sample is started by gradually
moving furnace N nearer and flearer to the furnace O. The rate at which the
bubbles enter the nitrometer must not exceed 2 bubbles in 3 seconds and this
rate is carefully regulated by the proper movement of the furnace N.
During the combustion the gasometer is refilled with carbon dioxide. When
no more bubbles rise in the nitrometer, the combustion is at an end and the
gaseous combustion products are conveyed into the nitrometer by a current
of carbon dioxide; stopcock L is carefully op'ened and adjusted to give a rate
of one bubble per second for five minutes. The rate of carbon dioxide flow is
next increased to three bubbles per second and when the bubbles begin to
approach microsize both electric furnaces are opened and pushed away, and
the tube allowed to cool. Furnace N is shown in position away from the combustion tube while 0 is in the position over the combustion tube. Carbon
dioxide is passed through the system until micro bubbles of the same size as
at the beginning of the run are obtained. The levelling bulb is raised, attached
to a holder, and the nitrometer disconnected.
After 15 minutes the readings are taken of the nitrogen volume in the nitrom. eter with the aid of a magnifying lense to within 0.001 ml. with the potassium
hydroxide in the levelling bulb at the height of the nitrometer meniscus, and of
its temperature to within 0.10 C. with the bulb of the thermometer touching
the wall of the measuring capillary, and of the barometric pressure to within
0.1 mm. The time required from the filling of the combustion tube with the
sample and fine copper oxide to the completion of the run is 30 to 45 minutes.
Calculation According to Pregl.-A number of investigators have found
that a deduction of 2% from the total volume will take care of any unabsorbable
gases liberated from tube filling during the combustion, the vapor pressure of
potassium hydroxide and the film of liquid on the side of the nitrometer.

2488

QUANTITATIVE MICROCHEMICAL ANALYSIS

Example.-

1. ........ .

Acetanilide sample, mg ........ : ...................

Volume of nitrogen collected mI. at 29.5 C. and 753.3 mm.1ressure .. .
Calibration. correction for burette, mI.... ',' ............ 1 0.001
2% correctlOn, m!. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... 0.009

4.950
0.471

Total correction ................................... ". ~......... . 0.010

Net volume of nitrogen to be reduced to standard conditions, mI.... . 0.461
Nitrogen found, % ............................................ . 10.40
Nitrogen calcd., % ............................. : .............. . 10.37

Calculation According to Niederl and Trautz. 29-Niederl and Trautz after

careful research conclude that more accurate results are obtained by dealing
with each correction separately. It has been found that consistently high
results are largely due to a more than average amount of un absorbable gases
liberated by the tube filling, or allowed to enter at the time of filling the tube,
which amount varies according to individual procedure and set up. These
high values are corrected by running a blank before each series of runs and
subtracting the blank from the volume reading as illustrated below.
To the volume reading the following corrections are applied:
1. The calibration correction of the nitrometer (from the apparatus certificate).
2. The correction for the air and absorption errors obtained from the blank
analysis.
3. 0.5% for the adhesion of the potassium hydrs>xide solution to the wall; 0.3%
for the vapor pressure of >the potassium hydroxide solution; 0.3%
(approx.) for the temperature reduction of the barometer .reading (from,
18 to 0 C.) making a total for these three of 1.1 %.
The resulting nitrogen volume is reduced to normal conditions (760 mm.
and 0 C.) and the percentage of nitrogen calculated.
o~Toluamide sample, mg......... , ................... ',' .......... 3.475
Volume of nitrogen collected m!. at 24.5 C. and 767 mm.
pressure ................. '........................ : " 0.321
Calibration correction, mI. . . . . . . . . . . . . . . . . . . . . . .. 0.001
Air and adsorption correction m!. ........--. . . . . . . .. 0.008
1.1 per cent correction, m!. ......................'. 0.003

Total corrections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.012

--'-_
~

Net volume nitrogen to be reduced to normal conditions, mI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.309

Nitrogen found, %. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10.30
Nitrogen calcd., %......................... : ..... " 10.31
NOTES ON THE MICRO-DUMAS METHOD.-(l) The same quantity of fine copper oxide
should be used for each determination so that the volume of occluded gases is,the same
for each run.
(2) Approximately the same volume of carbon dioxide should be used 'for each
determination-when the gasometer is used the correct volume is readily measured out
for each run. When the analyst has had experience, the size of the bubbles will determine when enough carbon dioxide has been passed, rather than a fixedivolume. However, it will be found that approximately the same amount is used for each determination.

_,-

2g

Ind. Eng. Chern., Anal. Ed., 3, 151 (1931).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2489

(3) If there is a troublesome-adherence of the bubbles to the mercury potassium

hydroxide interface .it may be overcome .by the addit.ion of a few drops of lUercury
previously shaken with ether. The ether IS removed with filter paper and the lUercury
and sludge are washed with potassium,hydroxide solution and water, and both mercury
and sludge are introduced into the nitrometer through the lev-eIling bulb. The sludge is
'the essential part of the mixture.
(4) If the potassium hYdroxide solution has a tendency to foam in the nitrometer, it
is purified by treating with powdered barium hydroxide (5 g, to 200 mI. solution). After
slIaking, the solution is allowed to stand, and then is filtered.
(5) When standing, the apparatus should be connected to the generator so that a
positive pr~ssure of carbon dioxide is maintained, which help1lto prevent the absorption
of air.
(6) It sometimes happens that due to the needle-shaped crystals of certain organic
compounds and to static electricity, the quantitative transference of the sample from a
mixing tube to the' combustion tube is impossible. By weighing the sample into a small
porcelain boat as described, this difficulty is overcome.
(7) A good quality of_marble cannot always be obtained in which case the carbon
dioxide generator of Poth,30 Fig. 424, offers advantages and has been successfully used
by others. 31 A convenient design is shown in Fig. 424. A and B are made from 2- and

FIG. 424.

I-liter, ro~d-bottomed, Pyrex flasks, respectively. A is charged with 500 grams of acid
potassium carbonate dissolved in 1200 ml. of water, and B with 170 mI. of water followed by 17c1mll. of concentl'ated sulfuric acid. The heat of dilution raises the temperature of the acid almost to the boiling point and thus contributes to the subsequent
elimination of contained air, D and E are sealed, and the desired amount of water is
introduced into the bubble ()ounter G. Then the T-shaped, three-way stopcock J ~th
30 Edgar Poth, Ind. Eng, Chern., Anal. Ed., 3, 202 (1931).
31 Milner and Sherman, lnd. Eng. Chern" Anal. Ed., 8, <331 (1936).

2490

QUANTITATIVE MICROCHEMICAL ANALYSIS

its .perpendicular arm K, filled with and extending into mercUry, is temporarily connected through rubber tubing at I, and the entire system is eva;cuated with an oil pump
simultaneously through Hand L. The pumping is continuedfdr about 15 minutes when
the appropriate amount of mercury is run into F. Both during the pumping and for
spme time afterward, the bicarbonate solution in A will evolve. arbon dioxide. Besides
being a great help in sweeping out the last traces of dissolved ail.' from the solutions, this
property is used in starting off the generator. After closing the, T-shaped stopcock, it is
disconnected from the suction pump at L. The evolution of carbon dioxide in A drives
the gas into B to build up sufficient pressure to force acid into A upon evacuation through
H. If H is closed immediately following the introduction of the acid, the gas generated
will force more carbon dioxide into B. By repeating this procedure, sufficient pressure
can be built up in B to force the mercury seal F. During this process, the T-shaped
stopcock is cautiously opened from time to time to ascertain when the pressure in B is
equal to atmospheric pressure plus the height of the mercury column in F. When this is
attained, the T-shaped stopcock can be removed, and the generator will deliver a stream
of practically pure carbon dioxide whenever H is opened. As the apparatus cools, or
after it has stood for long periods of time, it may be necessary to apply suction at Hand
again force acid into A to reestablish the desired pressure in the system.
Where a supply of dry ice is available, the method of Hershberg and Wellwood 32
provides an excellent source of carbon dioxide and should be used in place of the Kipp
generator method if possible. A two-liter vacuum bottle fitted with a mercury U-shaped
safety valve and charged with dry ice will generate carbon dioxide for2-3 weeks. When
freshly filled, it should be connected to the apparatus and the carbon dioxide allowed to
. displace the air in the bottle. This usually requires about 12 hours but, in some cases,
it may require a slightly longer period before gas of sufficient purity is obtained.
(8) Ground-glass joints on the combustion tube instead of tubber connections have
been used. They are satisfactory but not necessary for accurate work.
(9) Some analysts evacuate the combustion tube after it has been connected for a
run in order to eliminate occluded gases. The apparatus requires a three-way stopcock
at Land P, Fig. 423. This procedure cuts down the blank, but where the procedure has
been standardized as to the correction to be applied, the evacuation procedure is superfluous when time is a factor.
(10) If the sample is difficult to oxidize, thus giving low results, a small amount, 30
mg., of finely divided potassium perchlorate is sometimes added to facilitate combustion.
The excess oxygen is removed by the metallic copper.
(11) Many analysts use a three-way stopcock between the nitrometer and combustion tube. The presence of a stopcock at this point is not required and often develops
leaks. The regulation of the flow of carbon dioxide with the stopcock between the
combustion tube and the gasometer is much more satisfactory.
(12) The heating of the sample twice, as advocatE)d by Pregl and others, has not
been found necessary.
(13) It is best to use two nitrometers so that while one is cofning to constant temperature the other can be used for the next run.
I
(14) The results of Milner: and Sherman and E. P. Clark for th~ vapor pressure of
50% potassium hydroxide show that the subtracting of 0.3% of the volume!?f gas is to.o
small. While these results are interesting and should be taken into consideration when
standardizing a new apparatus, it will be found that for ordinary routine work the
subtracting of a fixed percentage will prove more accurate. The blank is essentially a
constant for a particular apparatus and its method of operation, and thus will vary
according to the individual operator. The following table gives the results of Milner
and Sherman 33 together with those of Clark 34 and those from the international critical
' .
:
tables.
Ind. and Eng. Chern., Anal. Ed., 9, 303 (1937).
Ind. Eng. Chern., Anal. Ed., 8, 331 (1936).
34 Jr. &soc. Off. Agric. Chem., 16, 575 (1933).
32
33

QUANTITATIVE MICROCHEMICAL ANALYSIS

2491

Vapor Pressure of Potassium Hydroxide

Vapor pressure, in mm. of mercury, of solution containing approximately 71.5 grams
of potassium hydroxide per 100 grams of water

Temperature

15
20
25
30
35

C.

Fresh Solution

2.9
4.3
6.1
"8.4
11.4

Used Solution

3.5
5.1
7.1
9.6
12.9

Clark

I.C.T.

5.5
7.0
8.9
11.4
14.0

4.1
5.6
.7.4
9.6
12.7

(15) Compounds which cannot be analyzed by the Dumas method even with the'
use of potassium chlorate have been noted by A. Friederich,35 and Milner and Sherman.'"
No general rule so far has been formulated for those compounds which cannot be analyzed
by the Dumas method but there are relatively few such compounds reported.
Hayman and Adler 36 report success in the case of certain pyrimidines which gave
low resplts by mixing about 5 parts by weight of copper acetate to one part of sample
and carefully mixing with this about 6-7 grams of fine copper oxide. It was found that
decomposition was only complete when a very hot flame was used.
Spies and Harris 37report a modification of the perchlorate procedure. To overcome
the objectionable low decomposition point of the perchlorate, it is decomposed after the
combustion of the organic compound while the site of the unburned residue is heated to
about 600 0 C., thereby subjecting the nitrogenous residue to an atmosphere of oxygen
at high temperature. The potassium chlorate, 100--125 mg. contained in a porcelain
boat, is placed about 40 mm. from the regular filling which contains the sample. A 15
mm. section of copper is addetl, separated from the regular 40 mm. section of copper by a
3 mm. wad of asbestos. The 15 mm. of copper is removed after each analysis. They
report accurate results for compounds which were derivatives of pyrimidine and purine.
(16) For samples that contain only small amounts of nitrogen, less than 5%, and
for blank runs, a nitrometer of 0.2 ml. capacity graduated in 0.001 ml. is necessary.
Such nitrometers must bE,\made to order.
(17) In routine work the fine copper oxide takes care of the oxidation of the sample
so that the main filling of coarse copper oxide is but little reduced. Thus only the fine
coppe;r oxide is rejected after each DID. A filling normally lasts for several hundred
determinations.
(18) It sometimes happens that the liquid seeps from the funnel into the upper part
of the nitrometer, thus pushing the column of gas down and preventing a proper reading
of the volume. This column of liquid may be pushed back by gently opening the stopcock and raising the levelling bulb above the funnel. With a little practice the liquid
can easily be allowed to escape, and when the gas reaches the top of the nitrometer the
stopcock is quickly closed before any escapes.
Micro-Dumas Estimation of Nitrogen in Liquids Low in Nitrogen-Milk. 38_The
usual nllcro Kjeldahl method for estimating nitrogen in small quantities of a liquid such
as milk may be satisfactorily-replaced by a micro-Dumas process. A small copper bQat
9f approximately 0.25 ml. capacity enclosed in a small thin-walled glass dish covered
with a glass plate is accurately weighed, milk added and reweighed. The boat is placed
in a small vacuum desiccator and kept at 30--350 C. until the water has evaporated; the
boat and its contents are introduced into the combustion tube and covered with fine
copper oxide, and the combustion conducted in the standard manner.
3. "Die Praxis der Quantitativen Organischen Mikroanalyse," Leipzig and Vienna,
Franz Deuticke, 1933.
36 Ind. and Eng. Chem., Anal: Ed., 9, 197 (1937).
37 Ind. and Eng. Chem., Anal. Ed., 9, 304 (1937).
38 G. Kraemer, J. Prakt. Chem., 97, 59 (1918).

2492

QUANTITATIVE MICROCHEMICAL ANALYSIS

Nitrogen by Micro-Kjeldahl Method.-There is little that is new in this

procedure except the method of ammonia distillation.,9 The method can be
used for the determination of nitrogen in organic compounds, soils, leather
feeds, fertilizers, etc. A sample, 3-10 mgm., is weighed in a small watch glas~
cut from the end of a test tube and is placed into the' ~igestion flask. These
micro Kjeldahl flasks are thick walled test tubes with a bulb on the bottom.
The time of digestion for ~ny substance is not more than 2 hours, and usually
less.
.
Apparatus.-The apparatus illustrated, Fig. 425, is designed to prevent the
condensation of steam and make it possible to attach to a single ringstand.
It is built entirely of pyrex glass, eliminating all rubber connections. The
steam generator A is made from a 500 ml. Kjeldahl flask sealed to the chamber
E (5.5 cm. wide and 10 cm. high), with openings near the t?P of E to allow the

s
FIG. 425.

steam to pass into E, then M to the distilling flask N. The lower bulb N is
6 cm. in diameter and the total height from lower to top of upper bulb is 22 cm.
The use of a vacuum jacket around the bulb N and partially coverjng the upper
bulb is recommended for routine distillation as it eliminates the burner at this
point. The condenser T is supplied with running water and burners are placed
under A and N. The steam generated in A passes to N, liberates and carries
ammonia and steam through the tube H, and on passing through T is condensed and collected in a 50 ml. graduated flask containing 25 ml. of ammoniafree water. The condenser tube is placed below the surface of the liquid and
25 ml. of distillate are collected.
The efficiency of the apparatus is shown by an experiment jn which a sample
of ammonium chloride containing 0.070 mg. nit~o5ej}.. was introduced and the
39

Kemmerer and Hallett, Ind.Eng. Chern., 19, 1295 (1927).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2493

amount recovered was determined by nesslerizatio~. The average deviation

.
from theory of five determinations was O.OOI mg.
Proc~dure.-The weighed sample is placed in the flask and the dige~tion
,mixture added. In general this consists of 1 ml. ()oncentrated sulfuric acid ,
0.5 g. to 1 g. of potassium sulfate and a small crystal of copper sulfate; 3 drops
of alcohol may be added to insure complete dec~lIiposition, and to continue
time of digestion. For nitro, azo and similar compounds 4.0 a mixture of 50-100
mg. glucose or dextrose, 1 g. potassium sulfate, a small crystal of copper sulfate
and 3 ml. of concentrated sulfuric acid is !lsed. 'the digestion is continued.
until the liquid is perfectly clear after which the distillation is performed as
described below.
Ammonia-free water is added to A through the tunnel D, and a blank run
to make sure that the water or apparatus is free frolIi ammonia. The digested
sample and an excess of potassium hydroxide are added to N through the funnel
P. Copper sulfate in the digested sample indicates when an excess of alkali
has been added.
The apparatus is prepared for the next sample by removing the burners and
JI}}DW.\>;lg jJ;w .m.I,).1j.w..l) i>;l N jD .be .cb..ll.w.>;l il)jD !'.b.B.roru-r Jj! by the portia} vacuum
caused by the condensation of the steam. A small depression at the bottom
of flask N facilitates complete removal of the distillec[ sample. The distillation
flask is rinsed by pouring water into N through P, abd heating A for a minute.
The apparatus is now ready for a second determiIlation. Five minutes are
required fQr a complete distillation.
An aliquot of the distillate, usually one-half, is diluted to 50 ml. in a Nessler
tube, nesslerized, and compared with standards tr&'ated similarly and at the
same temperature.
.
Results that equal those found by nesslerization are obtained by collecting
the distillate in an exc{\'Ss of N /100 sulfuric acid and titrating back with N /100 .
sodium hydroxide using methyl red as the indicator. If the solution contains
carbon dioxide, it must be boiled and cooled just before titrating. Where the
sample contains a large percentage of nitrogen, titration is preferable to nesslerization.
NOTEs.-(l) Lauro 41 suggested the use of seleniuJ? as a catalyst in macro Kjeldahl
digestions in place of copper or mercury. It w~s clalmeq 42 that more rapid oxidation
took place when selenium or selenium oxychlorIde was used, but from later investigations 43 it would seem that in most cases the time of digestion is not cut down to any great
extent. Selenium or selenium oxychloride has ~ sli~ht a<lvantage over copper sulfate,
but no advantage over mercuric oxide. A combmatlOn of selenium with mercuric oxide
is. better than any of these catalysts used singly. The Proportions used are 100 parts
sodium sulfate (Na 2SO.7H 20), 7 parts HgO, 1.5 parts selenium. The above work has
been done with macro digestions.
40 Elek anq Sobotka, J. Am, Chern. Soc., 48, 501 (1926); also Robert A. Harte,
Ind. Eng. Chern., Anal. Ed., 7, 432 (1935).
41 M. F. Lauro, Ind. Eng. Chern., Anal. Ed., 3, 401 (1\)31) .
2 Tennant, Harrell, and Stull, Ind. Eng. Chern., Anal. Ed, 4, 410 (1932) and O. E.
Rich, Cereal Chemistry, 9, 118 (1932) and R. M. SandstEJdt, Cereal Chemistry, 9, 156
(1932').
'
43 R. A. Osborn and A. J. Krasnitz, J. Assoc. Official Agr. Chern., 16, 110 (1933);
. Sandstedt, Cereal Cliemistry, 9,118 (1932), and L. V. Tayl()r,.Jr., Ind. Eng. Chern., Anal.'
Ed., 5, 263 (1933).

2494

QUANTITATIVE MICROCHEMICAL IANALYSIS

I

West and Brandon report that rapid oxidation is obtaiIled by the addition of 0.2%
selenium to the micro Kjeldahl digestion mixture. Unlike mercury-it does not have to
be precipitated before distillation. They give two micro Kjeldahl digestion mixtures.
For biological fluids, add,l mI. of selenium oxychloride to 500 mI. of 1 to '1 sulfuric acid;
usually 1 ml. is added to the sample as the digestion mixture. IFor organic compounds,
mix one volume of concentrated sulfuric acid to one volume bf a saturated solution of
potassium sulfate and add 1 mI. of selenium oxychloride. Use 11 m!. of this mixture for
.
I
digesting the sample.
(2) The distilling of ammonia into boric acid and titrating directly with mineral
acid has been used successfully in several laboratories and was first proposed by Winkler.""
This eliminates one standard solution. The boric acid retains the ammonia but does
not interfere with its titration with acid. A definite quantity of boric acid should be
used each time because the color of the indicator varies with the cOl1Centration of boric
acid. Stover and Sandin 46 used 4 m!. of a 2% solution which is sufficient to retain 0.1
to 21.1 mg. of nitrogen. Distillation is continued until a. volume of 50 ml. is obtained.
A freshly distilled sample may be titrated directly, but if CO 2 is present the solution must
be heated to drive off the carbon dioxide and cooled before titrating. The following
indicators may be used: 1. Methyl red (concentration 0.1 gm./100 ml. in 95% alcohol)
(use 2 drops). 2. Methyl red and tetrabromphenol blue, made by mixing equal volumes
of solutions of the separate indicators. Concentration of methyl red is given above.
Tetrabromphenol blue has a concentration of 0.1 gm. in 20 m!. of warm alcohol and
diluted to 100 mI. The mixture of methyl red and tetrabromphenol blue changes from
green to a gray color when almost at the end-point, and to pale gray-violet with slightly
more acid. (Use 2--4 drops.)
,

DETERMINATION OF HALOGENS IN ORGANIC COMPOUNDS *;

1. Acc.ording to Pregl.-The method consists in burning the substance in
oxygen, passing the products of combustion over red hot platinum contacts and
then over a glass spiral moistened with a solution of sodium carbonate containing sodium bisulfite. The sodium carbonate solution absorbs the hydrogen
halide and the sodium bisulfite reduces any halogenate or hypohalogenate to
sodium halide. Perhydrol is used to oxidize sulfites to sulfates, and silver
nitrate to precipitate the halogen.
'

FIG. 426.

Apparatus.-A combustion tube, preferably quar.tz, Fig. 426, 60 cIll. in

length and 7 mm. in diameter, is filled to a length of 20 cm. with a hard glass
.. Ind. Eng. Chem., Anal. Ed., 4, 314 (1932).
46 Z. Angew. -Chern., 26, 231 (1913).
46 Ind. Eng. Chem., Anal. Ed., 3, 240 (1931).
,
* For an automatic combustion apparatus, see p. 247q and ref.- 12. For a more
convenient absorption apparatus for routine work, se~ .D.-Hallett, Ind. Eng. Chem.
Anal. Ed., 10, 111 (1938).
-

QUANTITATIVE MICROCHEMICAL ANALYSIS

2495

spiral. The spirals have a pitch of 5 mm. and through them runs a glass rod
which is sealed to the spiral at each end. The combustion tube is drawn out
toa capillary end 1.5-2.0 cm. ,in length with an opening of 1 mm.. A large test
tube about 25 mm. in diameter is carefully cleaned with sulfuric chromic acid
mixtu~e and water, and charged with a mixture of 2 ml. ()f pure saturated sodium
carbonate solution and three drops of sodium bisulfite solution. This solution
is aspirated into that' portion of the combustion tube containing the spiral. Then two freshly ignited platinum contacts are
introduced. If the contacts become" poisoned" they are etched
with hot aqua regia. These contacts are made of platinum foil
0.05 mm. thi~k. Each contact is 75 mm. in length and is formed
in the shape of the letter Z. Contacts having greater surface,
shaped with six fins (Fig. 427) are on the market and are very
satisfactory, or a 15 cm. roll of 80 mesh platinum gauze may be
used. The contacts are placed so that they are 8-9 cm. distant
from the glass spiral. .
.Procedure.-5-8 mg. of the substance to be analyzed are
weighed in a platinum boat and introduced into the combustion tube. (When iodine is determined gravimetrically 5-10
mg. are used.) Oxygen from a cylinder ,enters the tube at a rate
of 2 bubbles per second, through a capillary tube inserted in a
rubber stopper. The platinum contacts are heated to 700750 0 C. by the furnace A, Fig. 428, and then the material in the
boat is slowly burneq by furnace B. (When iodine is determined
it sometimes condenses beyond the heated part and must be
driven by heating into the spiral filled portion.) When the combustion is finished, both furnaces are opened and the tube removed to the supports;1t C and D on the front of the apparatus
where it is allowed to cool in a current of oxygen. The tube and
spiral should be washed three times with water free from halogen,
FIG. 427.
the total volume of solution being 30 m!. Two more drops
of bisulfite are added and the solution is filtered to remove any 'suspended
matter. In the determination of chlorine and bromine 2 drops of "Perhydrol"
are added to the solution, and it is then warmed on the water bath 3-5 minutes.
When iodine is present in the solution, heating causes the formation of iodate.
This is avoided by adding 4-5 drops of "Perhydrol" and allowing the solution
to stand at room temperature for ten minutes. After cooling, 1 ml. of concentrated nitric acid is added and double the quantity of silver nitrate necessary
to precipitate the halide. This solution is heated on the water bath 10-15
minutes to coagulate the precipitate, is cooled, and then filtered. The filter
tube prepared and dried as previously described under general directions, page
2470, is hung on the balance pan by suspending it on the hook, and weighed.
The tube is placed in a filtering flask shown in Fig. 409 and the precipitate is
transferred by means of a syphon tube not more than 4 mm. in diameter. The
syphon tube is placed just ab'bve the precipitate, suction is applied, and the
sup'ernatant liquid filtered off. The precipitate is washed by adding a few m!.
of nitric acid 1 : 100, then the long arm of the syphon is lowered, and the precipitate transferred to the filter
, tube by suction. The test tube is rinsed twice,

2496

QUANTITATIVE MICROCHEMICAL ANALYSIS

alternately with dilute nitric acid and alcohol. The syphon is removed, the
precipitate washed once with alcohol, the tube wiped and dried at 1250 C.,
cooled, and weighed. The percentage of halogen prese* in the sample ,is
calculated:
t
Weighed

Sought

AgCl. ... , .............. Cl

AgBr ................... Br

AgI .................... I

Factor

0.2474
0.4255
0.5406

TOP REMOVED

@
o

10
I

SCALE eM.

FIG. 428.
NOTES.-(1) When substances contain halogen free from sulfur and other elements
which combine with metallic silver, the simple method of Dennstedt 7 is often used.
The combustion tube without a bead mling is charged first with a platinum boat containing the weighed sample, then two platinum contactsand finally a boat containing a
known weight of molecular f;ilver. The contacts are heated to redness and the molecular
silver to a temperature of 500-600 C. The sample is slo:rly burned in a current of
oxygen and the.products of combustion passed over the platinum contacts, the halogen
being absorbed by the metallic silver. The percentage of halogen in the sample is
calculated from the increase in weight of the boat containing the silver.
(2) Burning -the organic material in a platinum lined micro Parr bomb has been done
successfully.48 A stainless steel bomb is not satisfactory due to corrosion and subsequent
erratic results .. Beamish .9 in a macro procedure reports th~t a nickel'bomb gives good
.7 C. Funk and Kon, J. Am. Chern. Soc., 48, 1667 (1926),:48
49

Elek and Hill, ibid., 55, 2550-4 (1933).

Ind. Eng. Chem., Anal. Ed., 6,352 (1934).

QUANTITAT~VE

MICR9CHEMICAL ANALYSIS

2497

results even ~ith prolonged heating with bromine compounds, but a stainless steel bomb
gives low results under such conditions. The explosion mixture consists of 1.5 gm. of
pure sodium peroxide, 300 mg. of a mixture of potassium nitrate and sugar in a ratio of
3 : 1. The lid is clamped on, the mixture well shaken, and exploded by heating the
bomb with a small, flame. The fusion is dissolved in 1 : 1 nitric acid and filtered, and
the halogen determined gravimetrically as silveFhalide.
(3) Occasionally, a compound will be found which will give low results by the combustion method due to incomplete burning. In such a case the micro Carius method
should be used which is, of course, the same as the macroprocedure except for the
quantities of material used.
A tube closed at one end 20 cm. long, bore 8 mm., wall 2 mm., is charged with
approximately 0.5 ml. fuming nitric acid, 15-25 mg. of silver nitrate, and the sample,
which is contained in a capillary if a liquid, or in a glass capsule, if a solid. The tube
is sealed, placed in a bomb furnace, and heated at 3000 C. for 5-8 hours, cooled (usually
overnight), and then after carefully breaking the tube, the silver halide is washed on
to a filter, dried, and weighed.

The Volumetric Determination of Iodine.-The method 50 described, is both

rapid l111d accurate. It is recommended in place of the gravimetric procedure
for-routine work. With samples containing 2-3 mg. iodine, the results vary
0.1 % from theory.
Reagents.-N/50 sodium thiosulfate (made by diluting N/l0 thiosulfate
with freshly boiled out distilled water).
Normal sulfuric acid solution.
5% solution phenol (quinone-free).
Saturated bromine water.
Potassium iodide (iodate-free).
Starch indicator solution.
Methyl orange indicator solution.
procedure.-A sample containing 2-3 mg. iodine is burned in oxygen as
previously described on page 2494 except that no bisulfite is added to the sodium
carbonate solution.
After combustion, the sodium carbonate solution containing the absorbed
iodine is washed into a 125 ml. Erlenmeyer flask, neutralized (indicator-methyl
orange) with normal H 2S0 4 and 0.1 ml. of acid added in excess. The total
volume of solution should not exceed 50 m!. One m!. of saturated brominewater is added, the solution is shaken and allowed to stand 3 minutes. The
excess brom,ine is removed by the addition of 1 m!. of 5% phenol solution .
After shaking and allowing it to stand for 3-5 minutes, solid potassium iodide
(Approx. 0.5 gms.) is added in excess, and the liberated iodine titrated with
thiosulfate. The iodine present in the original sample is equal to one-sixth
of that which is titrated.
9.997
7.55
1.258
2.538 mg. iodine
1.258X2.538
1.258X2.538X100 =3194
Percentage of iodine In sample .......... '" .... . 9.997
.

Weight of sample .................... " " .. ,., ,

mI. of N /50 thiosulfate ....................... .
One-sixth .. , .... , ........................... .
1 m!. 0.02 N thiosulfate ...................... .
Weight iodine in sample ....... " ............. .

50

Joseph L, Goldberg, Michrochemie, 14, 161 (1934).

2498

QUANTITATIVE MICROCHEMICAL ANALYSIS

'

NOTEs.-(l) The volume of excess acid (0.1 mI.) added to the bromine solution
may be exceeded by 0.1 to 0.2 ml. but more than 'this is. inadvIsable.
.
(2) The phenol should not be added dropwise but in one portion. The dropwise
addition of phenol often yields a turbid solution consisting of tri-bromophenol and tribromophenol bromide. The latter compound liberates iodinelfrom KI and thus high
results are obtained.
/
'I
(3) The distilled water used should be freshly boiled to ren;tove oxygen.
(4) For micro work, N/100 thio~ulfate is often recommended, but at this dilution
rapid deterioration of the solution takes place and the time lost in frequent restandardization makes its use in routine work unsatisfactory. N /50 thiosulfate has a constant titer
for at least one week. For routine work it is best to make up every week the required
amount of thiosulfate by diluting the N /10 solution. The addition of 1 % of amyl
alcohol helps to prevent the deterioration of thiosulfate solutions. 51 Another method
used is by adjusting the pH between 9-9.5 by the addition of borax (3.8 gm. per liter of
thiosulfate solution). It is recommended that one of these two methods be used for
preserving the N /10 solution.

Volumetric Determination of Chlorine and Bromine.-A micro Volhard

method can be used with satisfactory results.
Solutions.
0.02 N silver nitrate.
0.01 N ammonium thiocyanate.
Indicator-saturated solution of ferric alum.
Solution of sodium sulfite.
Concentrated nitric acid (freed from _oxides of oxygen by bubbling air
through the solution).
Procedure.-The sample containing 1 to 3 mg. of halogen is burned in a
combustion tube with oxygen, and the halogen absorbed in sodium carbonate
or alkaline hydroxide as previously described. The combustion products are
washed into a 50 ml. centrifuge tube in the case of chlorine and into a 125 ml.
Erlenmeyer in the case of bromine. The total volume should not exceed 30
to 40 ml. Three drops of sodium sulfite are added and then nitric acid until
there is 1 ml. in excess. The presence of sulfur dioxide changes any halogenate
to halide. The addition of sulfur dioxide is not necessary if the halogens are
in the halide form. An excess of silver nitrate is now added, usually 3 to 5 ml.
and in the case of the chloride the precipitated silver chloride is centrifuged off
and the excess of silver nitrate is titrated with ammonium thiocyanate, using
1 ml. of indicator solution until a reddish-brown dolor appears and remains for
3-5 minutes.
._
NOTES.-(1) It is best to check the end-point by adding another 0.1 ml. of silver'
nitrate and then titrating again to the appearance of the faint red color.
(2) If the silver chloride is left in the solution, it produces a fading end-point, so it
must be removed before titrating. The presence of silver bromide or iodide does not
interfere with the titration so they need not be removed. The use of ether,52 nitrobenzene/a benzyl alcohol,54 toluene,55 has been suggested fOr titrating without filtering
off the silver chloride. These organic compounds have the effect of coagulatin'g the
silver chloride in large particles so that relatively little surface is exp'?_Sed to cause fading
51 Mayr and Kerschbaum, Z. anal. Chern., 73, 321 (1928).
52 P. B. Rethberg, Biochem. J., 20, 483-5 (1926).

53 Caldwell and Moyer, Ind. Eng. Ghem., Anal. Ed., 7, ,38-9 (1935).
54 W. Kimura, J. Soc. Chern. In9., Japan, 37, suppl, binding 590-1 (1934).
55 Vaille and Hauteville, J. Pharm. Chim., 22, 61-7 (1935).

QUANTITAT~VE

MICROCHEMICAL ANALYSIS

2499

of the end-point. Ilowever, it is best to filter or ceptrifuge off the precipitate for accurate work.
.
(3) The presencll of sulfur dioxide in the solution also tends to proc1uce a fading
end-point and shoul(i be removed by heating.

1itrating B.a1og~n.l> \5,?':..n.g k\1l>~ipt':..~n. In.\1':..t.at~"1l>.-\','r.t\I)n.u~, !;:ll'VW,:'-;I.", VI

iodide may be titrated dIrectly with silver nitrate using dichlorofluorescein.
For micro work t}1e halide solution must be neutral. This may be adjusted
by adding acid or alkali until the characteristic green fluorescence of the
indicator appears. Near the end-point a brownish tinge appears; at the endpoint the precipitate assumes a reddish-orange color. As the precipitate is
very finely divided in dilute solution, the solution itself takes on tIlis color.
NOTEs.-(I) The Volhard method for bromide is satisfactory because the precipitate
of silver bromide need not be removed, and the acid concentration is not critical. When
titrating, using dichlorofluorescein, the solution must be carefully neutralized; thus, time
is not saved by using this method for bromine. With chlorine, where centrifuging is
required, the use of the dichlorofluorescein titration may save some time. However,
the end-point is not strikingly sharp with 0.01 N silver nitrate. Bulloc)!: and Kirk 56
found the only way to use the method was by adding an equal volume of acetone to
the solution which sharpened the end-point. They also suggest titrating in a partially
darkened room by the light of a daylight lamp placed at the side of the observer. A
black background is preferred to a white one. Back titrating with sta]"ldard 0.01 N
chloride is possible SO that the end-point may be checked. This procedure is also suggested. for dilute solutions by Kolthoff, Lauer, and Sunde 57 but the reverse titration
must be made immediately, as the indicator reacts slowly if allowed to stand. The
correct amount of illdicator is 2-4 drops of 0.1 % per 50 ml. of solutio]"l If smaller
volumes are Ui?ed, it is better to employ a concentration of 0.01 %, 2 drops beirig sufficient
for 4.-5 ml. Under the best conditions, the titration error is 0.5 to 1 %.
Calculation.-Vsing 0.020 N silver nitrate.
_1._5_98__:_,X:...,m;;cI:=-:. .:::si::..lv:;-e:_r_:n=i~tr..::a.:::te:....:X....:...:.lO:..:0:!,
.
PerceIl.t age b romme=
Wt. of sample
'
0.70914Xml. silver nitrateXI00.
Percell t age chi onne=
W
f
I
t.o samp e

THE lVIICRODETERMINATION OF PHOSPHORlJS

\;
IN ORGANIC SUBSTANCES
Preparation. of Solutions: 58 1. Sulfate-Molybdate Reagent.~50 gm. of'
ammonium sulfate are dissolved in 500 ml. of nitric acid (sp.gr. 1.36) in a liter
flask. 150 gm. of powdered ammonium molybdate are treated with 400 ml. of.
boiling water in a porcelain dish and stirred until solution is complete. The
~ solution is rinsed into a flask with a little water, cooled to the telllperature of
.6 Ind. Eng. Chern., Anal. Ed., 7, 178-180 (1935).
67 J. Arner. Chelll. Soc., 51, 3273 (1929).
68 Neubauet and LUcker, Z. Anal. Chern., 51, 161-175 (1912).

2500

QUANTITATIVE MICROCHEMICAL ANALYSIS

the room, and poured in a thin stream with stirring into the nitric acid solution
of ammonium sulfate. The resultant liquid is diluted tb 1 liter, allowed to
stand for two days, filtered and kept in a well~stoppered bottle of brown glass
in a dark, cool place.
2. Nitric Acid Containing Sulfuric Acid.-30 ml. of sulfuric acid Csp.gr.
1.84) are poured into 1 liter of nitric acid Csp.gr. 1.19 to 1.21). The latter acid
is obtained by mixing 357 ml. of nitric acid Csp.gr. 1.40) with 500 ml. of water.
3. A 2% Aqueous Solution of Pure Ammonium Nitrate.-The solution is
acidified with a few drops of nitric acid.
Methods of Oxidizing the Organic Material.-I. According to H. Lieb:
The material, 2 to 5 mg., is weighed into a small platinum boat. An excess of
finely powdered pure calcined sodium carbonate and sodium nitrate mixed in
equal portions is added, the whole is mixed with a short looped platinum wire
which is finally laid on the boat, and then carefully covered with the oxidizing
mixture. The boat is placed in a pyrex combustion tube about 15 cm. long,
one end of which is drawn out to a capillary and bent upwards at right angles.
The material is burned in a slow stream of oxygen. As soon as the reaction is
finished the boat is heated for some minutes to the full temperature of a Bunsen
burner and is then allowed to cool in a stream of oxygen. The material in the
boat is then boiled out with dilute nitric acid in a small test tube and poured
through a filter into a precipitation vessel which consists of a large thick walled
test tube which has previously been washed in sulfuric-chromic acid mixture.
If any of the melt has passed into the combustion tube during t~e reaction this
is washed out with dilute nitric acid. To the clear filtrate, 2 ml. of nitric acid
containing sulfuric acid are added, and also water if necessary to dilute the
total volume to 15 ml.
II. According to H. Lieb and O. Wintersteiner: 59 The following .method
requires less attention and is simpler than the other, and can be used also to
determine arsenic as well. It is reliable in most cases. 60 The sample, 3-6 mg.,
is digested in a micro Kjeldahl flask with 0.5 ml. of concl3ntrated sulfuric acid
and 0.5 ml. of perhydrol until fumes of sulfur trioxide /tre evolved, cooled, a
few drops of perhydrol added and the evaporati<;m is repeated until a clear
solution is obtained. To this are added 2 ml. of nitric acid containing sulfuric
acid, and water if necessary to make a total volume of 15 ml.

NOTE.-(l) Elek and Hill 61 employ a Parr microbomb using sodium peroxide to
destroy the organic matter.
.
-_...__

Treatment with Molybdate of Solution Obtained Above.-According to

H. Lieb: The liquid, 15 ml., is warmed on the boiling water bath. The hot
solution is well shaken, 15 ml. of the ~ulfate-moly-bdate reagent added. The
vessel then is allowed to stand for three minutes, and again shaken for half a
minute. The precipitate is allowed to settle for 30-,60 minutes, and filtered
with the aid of suction. The filter tube, or filter crucible, is washed WIth water,'
hot dilute nitric 'acid, and again with water, finally displaci.\lg the water by
alcohol and ether or by acetone. The filter tube is then wiped with a dry
Mikrochemie, 2, 78 (1924).
A. Elek, J. Am. Chern. Soc., 50, 1213 (1928), reports..a low result when thymus
neuclic acid was analyzed by this method.
. -"
61 J. Am. Chern. Soc., 55, 3479 (1933).
59

~o

QUANTITATIVE MICfWCHEMICAL ANALYSIS

2501

chamois leather, and left for at least half an hour in an evacuated desiccatorcontaining no drying agent. It is removed from the desiccator immediately
before use, placed on the balance, and the time carefully noted which has
elapsed between removal from the desiccator and the final weighing, which
may conveniently be 5 minutes. These conditions must be exactly repeated
when the' precipitate is weighed.
The precipitate is drawn into the filter tube by means of the same automatic
arrangement that is used for the silver halide precipitate, page 2470. After the
supernatant liquid has been syphoned from the precipitate it is washed in the
test tube with 2% ammonium nitrate solution, and afterwards transferred to
the filter; in order to remove the last traces of precipitate from the walls of the
precipitating vessel, 95% alcohol and ammonium nitrate solution are used
alt~nately. The tube is finally filled oD,ce with alcohol, and then twice with
ether or acetone. After drying half an hour in vacuo the tube is weighed
under exactly the same conditions as those of the initial weighing. The factors
are:
Phosphorus pentoxide. . . . . . . . . . . . . . .. 0.33260
Phosphoric acid. . . . . . . . . . . . . . . . . . . . .. 0.03326
Phosphorus ......................... 0.014524

. Volumetric Method. 62-Instead of weighing the yellow precipitate it may

be transferred to the original beaker and approximately twice the amount of
N /50 sodium hydroxide necessary for dissolving the precipitate added, and the
solution heated to boiling for 5 minutes to drive off the ammonia formed when
the caustic is added. The volume of solution should be reduced to 15-25 m!.
To the solution after cooling,' 5 drops of Y2% phenolphthalein indicator is
added and 3 to 5 ml. of N /50 hydrochloric acid in excess. The solution is
boiled for ten to fifteen seconds, cooled and titrated back with N /50 sodium
hydroxide. From the amount of alkali neutralized, the weight .of phosphorus
can be calculated. According to the equation for the reaction, 28 equivalents
of alkali are required per atom of phosphorus:
1 m!. N /50 sodium hydroxide= 0.02214 mg. phosphorus.

Accurate results are obtained with quantities of phosphorus 0.01-10 mg.

Procedure when Arsenic is Present (Kuhn, loco cit.).-Phosphorus may be
determined in material also containing arsenic by the addition to 5 ml. of the
phosphorus-arsenic solution of 10 ml. concentrated hydrochloric acid) 0.2 g.
potassium bromide and 0.3 g. hydrazine sulfate and then.. distilling. The
phosphorlls is then determined nephelometrically or otherwise after previous
destructitin of the excess hydrazine with 0.5-1.0 m!. concentrated nitric acid.
Colorimetric Determination of Phosphorus and Arsenic.-When the amount
of ammoniUI!l phosphomolybdate is so small that it cannot be weighed or
titrated with accuracy the following Deniges 63 colorimetric method is used.
If ferric iron is present in concent:ration greater than 4 to 6 parts per million,
it interferes with the determination. If phosphorus and arsenic are both
62 Richard Kuhn, Z. Physiol. Chern., 129, 64 (1923); R. H. A. Plirnrner, Biochern. J.,
27, 181'0--13 (1933).
63 Deniges, Cornpt. rend., 171, 802 (1920); Troug and Meyer, Ind. Eng. Chern.,
Anal. Ed., 1, 136 (1929).

2502

QUANTITATIVE MICROCHEMICAL ANALYSIS

present the two are first determined together. In a second sample the arsenic
is reduced with hydrogen sulfide or a little sodium sulfide in acid solution.
After boiling to drive off the excess of hydrogen sulfide, free sulfur is removed
by adding pure filter paper pulp, shaking vigorously and ~ltering. The color
due to the phosphorus alone is then developed. By difference the amount' ~f
arsenic may be obtained.
Preparation of Reagents: 1. Ammonium Molybdate-Sulfuric Acid Solution.
-Dissolve 25 grams of ammonium molybdate in 200 m!. of water heated to
60 C. and filter. Dilute 280 ml. of arsenic and phosphorus-free concentrated
sulfuric acid to 800 mI. After both solutions have cooled, the ammonium
molybdate solution is added slowly, with shaking, to the sulfuric acid solution.
After the combined solution has cooled to room temperature it is diluted with
water to exactly 1000 ml. This is a 10 N sulfuric acid solution containing 2.5
grams.of ammonium molybdate per 100 mI.
2. Stannous Chloride Solution.-25 grams of. stannous chloride crystals
are dissolved in 1000 mI. of hydrochloric acid (sp.gr. 1.05). Filter if necessary.The solution should be protected from the air by floa.ting a layer of white mineral
oil about 5 mm. thick over the surface.
3. Standard Phosphate Solution.-Dissolve 1).2195 gram of potassium dihydrogen phosphate and dilute to 1000 ml. A stock solution is made by taking
50 ml. of this and diluting to 500 ml. This stock solution contains 5 p.p.m.
and is used for making the standard solution for comparison. To make this
standard solution 5 m!. of the stock solution are taken, diluted to 95 m!. with
distilled water, 4 ml. of the ammo~ium molybdate-sulfuric acid solution are
added and thoroughly mixed by shaking in an Erlenmeyer flask. Six drops of
stannous chloride are added and shaken. It is diluted to exactly 100 mI.,
shaken, and the solution is then ready for use. It contains 0.25 p.p.m. of
phosphorus. For very dilute solutions 2 ml. of the stock solution are used, but
the same amount of reagents, giving a standard which contains 0.1 p.p.m. of
phosphorus. After standing 10 to 12 minutes, the standll'rd will start to fade,
and a drop more of stannous chloride should then, be arTded to bring the full
color back, which will again be p'ermanent for 10 to 12 minutes.
Procedure.-In the analysis of water, water ,extracts of soils, minerals,
fertilizers, etc., it is permissable to add the reagents directly to these unless they
are colored, turbid, or decidedly acid or alkaline. I Turbidity and color should
be removed by appropriate. means. A decidedly acid or ~alkaline reaction
should'be neutralized before adding the reagents. Organic mater'ials may be
ignited with magnesium nitrate. In this case it is best to keep the 10 N sulfuric
acid and 2.5% ammonium molybdate as separate solutions, in order that the
ignited residue may be dissolved with the sulfuric acid and the ammonium
molybdate added after proper dilution. In all cases the reagents should be
present after final dilution, in the proportion of 4 mI. of'lO N sulfuric acid, 4 mI.'
of 2.5% ammonium molybdate, and 6 drops of stannous chloride per 100 m!.
The reagents should be thoroughly mixed with the test solution before the
stannous chloride is added, after which thorough mixing should again be
accomplished. Comparison with the standard should be clade within 10
minutes after adding the stannous chloride. Blan..Ltests should be run.

QUANTITATIVE MICROCHEMICAL ANALYSIS

2503

NOTEs.-In work where maximum accuracy is required, the presence of much ferric
iron will interfere with the results by decreasing the color intensity and giving troublesome greenish tints. By reducing the ferric iron in a Jones reductor, using metallic
cadmium, these troubles are eliminated. The reduction should be made just previous
to the addition of the ammonium molybdate.
Aluminum and manganous salts may be present in considerable quantity without
influencing the color. Nitrate, as potassium nitrate, equivalent to 100 p.p.m. of nitrogen
does not affect the color, while 200 p.p.m. reduces the intensity about 10%. Calcium
oxide or magnesium oxide may be present up to 1000 p.p.m. without influence on the
color.

THE MICROANALYTICAL DETERMINATION OF ARSENIC

64

The substance is oxidized with sulfuric acid and "Perhydrol" and the
arsenic precipitated with magnesia mixture.
Procedure.-5-10 mgm. of organic substance are placed in a micro-Kjeldahl
flask with 4-5 drops of dilute sulfuric acid, heated, then cooled, 5 ml. of "Perhydrol" added, and again heateq until fumes of sulfur trioxide are evolved. The
oxidation is continued in the same way until Mter evaporation a clear solution
is obtained. The solution is transferred to a glass dish of 30 to 40 ml. capacity
and placed on a water bath, evaporated to dryness, then the residue is dissolved
'in 3 to 4 mI. of ammonia and treated while hot with 1 ml. of magnesia mixture
(this is prepared by dissolving 5.5 gm. magnesium chloride crystals and 10.5
gm. ammonium chloride in 100 ml. of water). The magnesium ammonium
arsenate which is precipitated is allowed to stand for 6 or preferably 12 hours
at room temperature until it becomes crystalline.
The precipitate is then filtered through an ignited and weighed micro
Neubauer or porcelain crucible, and alternately washed with a 3% solution
of ammonia and alcohol with the aid of gentle suction. The crucible is removed and ignited. The residue of magnesium pyroarsenate still occludes
magnesium salts, and must therefore again be washed repeatedly with a very
dilute solution of ammonium hydroxide. After igniting strongly, the crucible
is placed on a copper block and may be weighed after 10 minutes. The percentage of arsenic in the sample is calculated.
2As
Factor MAO
g2

S2

0.4826.

b. Volumetric Method for the Determination of Arsenic in Organic Substances.-The method 65 depends upon the fact that during the destruction
of tlrtJ organic substance arsenic is oxidized to arsenate; iodine is liberated from
potassium iodide which can be titrated with sodium thiosulfate. The method
Abderhalden Handbuch d. Biochemischen Arbeitsmethoden," p. 727 (1919).
Wintersteiner and Hannel, Mikrochemie, 4, 155 (1926); also, Beamish and Collins,
Ind. Eng. Chern., Anal. Ed., 6, 379 (1934).
64"
66

2504

QUANTITATIVE MICROCHEMICAL ANALYSIS

gives reliable results when halogens are absent in. the organic substance and is
particularly serviceable in elementary organic analysis. With halogens absent
the average deviatIon from theory is 0.2%, when titrating 2 to .4 mg. of
.
I
arsemc.
I
Procedure.-7-12 mg. of substance are boiled with 1 In!. of 30% sulfuric
acid and a few drops of concentrated nitric acid, more nitric acid is added and
boiling continued. It is evaporated several times with the addition of a few
drops of "Perhydrol," the residue is dissolved in 1 ml. of water, evaporated
until sulfur trioxide is evolved, this process is repeated, then 1 m!. of water is
added.
Beamish and Collins use a nickel microbomb for oxidizing the material.
It is charged with a sample containing 1.5-3.0 mg. arsenic, 20-25 mg. (approx.)
sucrose, and 1 gram (approx.) of sodium peroxide. The lid is clamped on, the
contents thoroughly shaken, and tapped on the table to insure that all material
is in the bottom of the bomb. Ignite by heating in the tip of a small hot flame
for 35-40 seconds. Allow the bomb to stand in air for 5 seconds and then cool
an'd rinse by immersing in cold distilled water. Remove the lid, wash it off
into a pyrex tube, 35 mm. diameter and 155 mm. length, then add the cup and
enough water to cover the cup. Heat to bring about solution of fusion and
remove the cup with_a nickel wire, and rinse ,thoroughly. The volume of
solution should be 20-25 mI. and is evaporated to 10 mI., which also removes
the hydrogen peroxide.
,
The procedure is now the slllme for 'both methods. To the cold solution
12 m!. of 12 N freshly boiled hydrochloric acid are added. To the solution at
room temperature is added 1 ml. of freshly prepared 10% potassium iodide
solution. Allow to stand for exactly 3 minutes and titrate the liberated iodine
with 0.01 sodium thiosulfate. Near the end-point add starch indicator and
titrate to the end-point. The color should not return for at least 5 minutes.
A blank should be run for the reagent used and subtracted.
0.01 m!. of 0.01 N thiosulfate equals 0.00375 mg. of arsenic.
Arsenic in Cadavers.-The above method of O. Wintersteiner and H.
Hannel has been modified for arsenic in cadavers.66 'The whole analysis
should not take more than one and one-half hours and gives satisfactory results,
with amounts of arsenic as low as 0.1 mg.
,
Procedure.-2 g. of material are employed and treated as described in the
original method. The iron invariably present is determined in...a portion of the
solution colorimetrically. A 'correction for the amount of iron is subtracted
from the titration value. Copper, if present, must be removed by the micro
electrolytic method before determining the arsenic.
'
NOTE.-The results reported by Wintersteiner and Hannel are 0.02 mg. high when
using samples containing 0.15--0.75 mg. arsenic trioxide, so that a sample should be
taken which has an amount of arsenic not below 0.1 mg. and preferably twice this 'amount.
This method should be of value because of the time saved in carrying out a determination.

Gutzeit Method.-Where small amounts of arsenic (0.001 to 0.04 mg.) are

estimated, the Gutzeit method is recommended. There are a ,great number
of modifications but for routine work, especially for ~~e:oic in organic material,
66

Szendro and Fleischer, Mikrochemie, Pregl Festschr., 321 (1929).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2505

the method developed by C. E. Lachele 67 is recommended as being simple

and giving reproducible results. It is applicable in the presence of impurities
such as iron, tin, antimony, or reducible sulfur and in the presence of phosphorus
comp0l'!lds. Not over 30 mg. of antimony can be present. With modification,
it is possible to analyze many products for arsenic using an undigested sample.
Reagents: Hydrochloric Acid.-Make up 50 ml. of 35% arsenic-free hydrochloric acid to 100 m!. with water.
Stannous Chloride Solution.-Dissolve 40 grams of arsenic-free SnCl2 H 20
in concentrated hydrochloric acid and make up to 100 m!. with the same strength
'acid.
Mercuric Bromide Paper.-Carefully select filter sheets, similar to Sand
'S No. 589 black ribbon paper, for weight and texture, and soak one hour in a
saturated alcoholic solution of mercuric bromide. After sensitizing, remove
sheets and dry by means of an air blast, and cut in disks of the same diameter
as the diaphragm tube. Ayoid touching the sensitized disks with the hands
as far as possible. Disks should not be used when more than 5 or 6 days old.
, . Zinc.-Cut arsenic-free zinc in pieces about 1 cm. in length (approximately
5 grams). Activate the pieces by covering with hydrochloric acid (1 : 3) containing about 2 m!. of special stannous chloride solution for each 100 m!. of
acid. Allow the action to continue for 15 minutes and wash well with distilled
water.
Potassium Iodide Solution.-Dissolve 15 grams of potassium iodide in
water and dilute to 100 m!.
Ferrous Ammonium Sulfate.-Use this substance or ferrous sulfate crystals
free from arsenic.
, Cuprous Chloride Solution.-Dissolve 15 grams of cuprous chloride in
'100 m!. of 1 to 1 hydrochloric acid.
Cadmium Iodide Solution.- Dissolve 15 g'rams of cuprous chloride In
100 m!. of 1 to 1 hydrochloric acid.
Standard Arsenic Solution.- Dissolve 1 gram of arsenic trioxide in 25 m!.
of 20% sodium hydroxide. Saturate the solution with carbon dioxide and
dilute to 1 liter with recently boiled water. One ml. of this solution contains 1
mg. of arsenic trioxide. Dilute 40 m!. of this solution to 1 liter, diluting 50 m!.
of the resulting solution to 1 liter. One m!. of this latter solution contains
0.002 mg. of arsenic trioxide and is used to prepare standard stains. Prepare
fresh dilute solutions at frequent intervals.
Apparatus and Procedure.-Use an Erlenmeyer flask connected by a twohole stopper to a Liebig or Allihn condenser. .A gas such as nitrogen or hydrogen is introduced through a tube extending below the surface of the contents
of the flask. The upper end of the condenser is packed with absorbent cotton,
the lower half of which is saturated with the cuprous chloride solution which
acts as a scrubber to remove impurities, such as phosphine, stibine, and hydrogen sulfide, from the evolved arsine. A chamber having a diameter that will
ac~ommodate a sensitized diaphragm of a size sufficient to combine with all the
arsine is fitted to the condenser. The chamber consists of two thick-walled
tubes,of the same diameter, with ground-glass joints held together by an outside
67

Ind. Eng. Chern., Anal. Ed., 6, 256 (1934).

2506

QUANTITATIVE MICROCHEMICAL 4NALYSIS

sleeve of Gooch rubber tubing. The impregnated, paper is fitted between the
two ground joints as pictured in Fig. 429.
I
An aliquot obtained from a sample digested with sulfjuric and nitric acid
in the usual manner (or the entire digestion in case of small amounts) is placed
in a 1000 ml. Erlenmeyer flask with enough distilled wAter to make about
200 ml. of solution. : Then 2 to 3 grams
Gooch
of solid ferrous ammonium sulfate or
Rubber
ferrous sulfate, 10 to 15 drops of stannous
Tubing
chloride solution and 50 ml. of 1 to 1
Ground
hydrochloric
acid are added. Just be,
Glass
fore connecting to the condenser, introJoint
duce two or three pieces of activated
Impregnated
zinc. A continual stream of nitrogen is
Nitrogen
D,sc
or
kept flowing through the system to carry
Hydrogen
all traces of evolved arsine through the
Supply
scrubber cotton to the paper diaphragm
impregnated with mercuric bromide.
The contents of the flask are boiled until
all the arsine is evolved, which usually
requires 3:bout 15 minutes. When the
reaction is complete, the disk is removed
and developed in cadmium iodide solution. This fixes the stain more permanently to light and water than does
potassium iodide and changes the color
from yellow to brown. The disk lis
FIG. 429.
allowed to remain in the iodide solution
until all traces of red mercuric iodide, which immediately forms, have been
dissolved. After prolonged use, the cadmium iodide solution turns yellow
from the accumulation of complex cadmium mercuric iodides and should be
replaced by fresh reagent. The excess reagent is removed from the disk by a
water wash followed by a rinse with alcohol, and the disk i~ then dried between
blotters. When dry, the stains are compared with standards. The standards
are prepared in a similar manner to the unknowns in steps of 0.0025 mg. of
arsenic trioxide, keeping them free from excess' light exposure. The stains
have a tendency to bleach slightly with age, and it is desirable. t_o prepare fresh
standards frequently, especially in the approximate range of the iIDknown.
With some modifications, the method is applicable to the undigested material, thus shortening the total time of an arsenic determination to 15 to 30
minutes and eliminating the costly procedure of digestion with nitric and sulfuric acids.
From 5 to 100 grams of the undigested sample may be placed directly in the
1000 ml. Erlenmeyer flask together with 100 ml. of 1 to 1 hydrochloric acid;
1 to 2 ml. of ~tannous chloride solution, 15 ml. of potassium iodide solution,
and 1 to 2 grams of solid ferrous sulfate. The activated zinc (3 or 4 pieces) is
added, and the flask immediately connected to the condenser unit. The
reaction is usually complete after 15 minutes of boilin~ except that, wheJt using
large amounts of sample, a somewhat longer time may be necessary to evolve

QUANTITATIVE MICROCHEMICAL ANALYSIS

2507

all the arsine. The scrubber cotton should be saturated with a 15% solution
of cuprous chloride in 1 to 1 hydrochloric acid to remove hydrogen sulfide,
phosphine, and stibine, if present. 6. In, case these gases have not been efficiently removed, the yellow stain imparted to the disk is not changed to brown
when imm@rsed in the" cadmium iodide solution. When using large samples of
some products that contain considerable reducible sulfur or phosphorus compounds, the scrubber cotton may not remove all impurities, and a preliminary
treatment is necessary. The sample is heated a short time with an alkaline
solution, after which it is slightly acidified with hydrochloric acid and the excess
sulfide removed by boiling. The remaining sulfur and reduced phosphorus
compounds are oxidized with bromine water, and then the bromine is removed
by boiling before adding the sample to the reaction flask.

THE MICRODETERMINATION OF ORGANIC SULFUR

The material is burned in oxygen, the oxides of sulfur absorbed in "Perhydrol," and the sulfur precipitated as barium sulfate.
Apparatus.-The same apparatus is used. as was described for the halogen
determination.
Procedure.-The procedure is the same except that the spiral tube is filled
with" Perhydrol," which has been diluted five times its volume with water.
A 5-6 mg. sample is weighed and placed in the combustion tube-, and the material slowly burned while the oxygen flows at the rate of 2 to 3 bubbles in 2
seconds. After combustion, the spiral filling is rinsed with 30 ml. of water into
a pyrex beaker of 50 ml. capacity. One ml. freshly prepared 10% barium
chloride solution added and 5 to 10 drops dilute hydrochloric acid. The dish is
covered with a watch glass, evaporated to 5-10 ml. on a water bath, cooled,
and the barium sulfate allowed to precipitate and settle for at least one hour,
preferably overnight. The precipitate is transferred to a previously ignited
and weighed micro porcelain filter crucible (or filter stick method may be used),
gentle suction applied and alternately washed with alcohol and water. Water
is used for the final washing. The precipitate is ignited, cooled, again washed,
and one drop of dilute sulfuric acid added, and washed three times with pure
water. It is again ignited, cooled on a copper block, and weighed to 0.005 mg.
Determination of Sulfu;t by Fusion.-A modification of the macro method
employing potassium nitrate as the oxidizing agent is used. 69
A fusion mixture which can be kept indefinitely is prepared by mixing four
J1arts by weight of sodium carbonate with three parts of potassium nitrate
and grinding to a fine powder.
68 Where more than 30 mg. of antimony are present, a Milligan absorption bulb containing 100 m!. of special cuprous chloride solution should supplement the cotton scrubber,
69 Harold Emerson, J. Am. Chern. Soc., 52, 1291-2 (1930).

2508

QUANTITATIVE MICROCHEMICAL ANALYSIS

Procedure.-A finely powdered sample containing apprqximately 0.5 mg.

of sulfur is weighed and mixed with 100 times its weight of the fusion mixture
in a 20 mL nickel crucible. A thin layer of the mixture is spdnkled on the top
to prevent escape of sulfur fumes. A covel' is placed on theicrucible and it is
set in a porcelain crucible of convenient size. Low heat with a Bunsen burner
is applied for 10 minutes and full heat for 15 minutes. Th1e contents of the
crucible are cooled and dissolved in 5 ml. of warm water, filtered, the crucible
and filter washed with 10 ml. of water and the sulfur precipitated with a 10%
barium chloride solution as previously described on p. 2507.
NOTEs.-(I) Elek and Hill 70 use a micro Parr bomb and use an explosion mixture
as given on page 2496 note 2 for the determination of halogens. The fusion mixture is
dissolved in 15 ml. of hot water. When cold 5 m!. concentrated hydrochloric acid are
added to the solution. It is then filtered and the sulfur precipitated as barium sulfate in
the usual manner.
(2) For routine work evaporation to 1(}--20 m!. is sufficient, and double ignition and
trestroent with sulfuric acid is not required.
(3) Funk and Kon 71 suggest using a 1 % solution of barium chloride for precipitating
the sulfur as barium sulfate in order to cut down occlusion,/noted by Pregl in his method.
This eliminates igniting the precipitate twice. They also describe a modification by
which the determination of sulfur and halogen can Qe effected on 80 single sa.mple of the
substance.
The procedure in this case is somewhat modified. The halogen is absorbed in alkaline formaldehyde solution instead of sodium bisulfite. After the combustion is completed, the tube is washed out as usual, the solution acidified with acetic acid and the
formaldehyde oxidized with "Perhydrol" by heating the mixture on a water bath.
After the reaction is completed, the sulfuric acid formed is precipitated with a 1 % barium
acetate solution. The precipitation is completed by heating and the barium sulfate
transferred to, ignited in, and weighed in a Neubauer or porcelain filter crucible; the
test tube is washed out three or four times and the filtrate then removed and the silver
chloride precipitated with silver nitrate as before.
(4) By adding picric acid 72 the particle size is increased and filtering can take
place immediately, so that no time is lost in digestion and standing in order to complete
precipitation. About 0.4 ml' of a saturated solution of picric acid is added per 15 ml.
of solution. The solution is heated to near boiling, the picric acid added, then the
barium chloride. After waiting a few minutes for the solution td cool, it is filtered and
the precipitate ignited.
,
(5) Where the highest accuracy is not required, the barium sulfate, when picric acid
0
is not used, can be dried at 150 C. and weighed. Some occlusion of mother liquor, of
course, is present, and the results will tend to be high. With 2 mg. barium sulfate it is
about 0.01 to 0.02 mg. high, when no other interfering elements are present except the
original barium chloride. Those ,who wish to eliminate ignition mu!t~ determine their
error under the conditions with which they are working. It often happe~ that the
accuracy is sufficient for the problem in question and the ignition is eliminated. The
use ?f picric acid causes considerable occlusion so that ignition of the precipitate is
reqUIred.
(6) Crucibles must be dried in an air oven at 1200 C. before ignition, because f3pattering of the precipitate and cracking of the "crucibles ofteI1 results, if they are placed
directly into the muffle.

Volumetric Determination of Sulfur.-In the absence of nitrogen or halogen

in the original ~ubstance, the sulfuric acid instead of being precipitated as
barium sulfate, is titrated with N /100 sodium hydroxide, using methyl red as an
indicator.
J. Am. Chem. Soc., 55, 3479 (1933).
J. Am. Chem. Soc., 48, 1667 (1926).
72 See Herbert Lindsly, Ind. Eng. Chem., Anal. Ed., 8, 176-8 (1936).

70
71

QUANTITATIVE MICROCHEMICAL ANALYSIS

2509

If the "Perhydrol" used is not neutral it must be neutralized with sodium

hydroxide using methyl red as the indicator.
The ter Meulen Hydrogenation Method (Volumetric).-The method 73
consists in burning the sample in hydrogen and absorbing the hydrogen sulfide.
which is formed in a solution of sodium hydroxide.
An acidified standardized solution of iodine is added
and the iodine remaining after the reaction with the
liberated hydrogen sulfide is titrated with thiosulfate. The method has not been used by a sufficient
number of investigators to evaluate it as a standard
procedure. It is, therefore, described here in order
to bring it to the attention of those who may wish
to tryout the method. From the published results
and the author's experience with a limited number
of compounds, it seems to offer a simple rapid volumetric method where a high degree of accuracy is not
required. Other micro volumetric procedures for
sulfur which have been published have not proven
satisfactory. Where a high degree of accuracy is
re'quired, the gravimetric procedure of weighing the
sulfur as barium sulfate must be employed.
Results by the ter Meulen method usually vary
0.2% from theory when using samples of organic
material containing 1-3 mg. of sulfur.
Solutions.
0.02 N Iodine.
0.02 N Thiosulfate.
0.2 N Hydrochloric Acid (approximately).
Starch Indicator Solution.
0.05 N Sodium Hydroxide Solution (approximately).
Apparatus.-A glass or preferably silica combustion tube, such as is used for carbon and
hydrogen, is filled for a distance of 10 cm. with 5%
platinized asbestos . A simple gas absorption bulb
for the hydrogen sulfide, Fig. 430, and a purifying
train for the hydrogen, complete the setup. The
hydrogen may be generated from hydrochloric acid
and zinc or be taken from a cylinder.
Procedure.-A boat with a sample containing
1- 3 mg. of sulfur is placed in the combustion tube.
The platinum catalyst is heated to 700-750 C.
The air in the combustion tube is swept out with
hydrogen. The hydrogen passes through the tube
Fla. 430.
during the actual combustion at the rate of 500 m!.
in 30 minutes. The sample is slow.ly burned by moving the furnace or burner
over the sample. The total time of burning should take 20 minutes. The tube
73 H. t er Meulen, Rec. tray. chirn., 41, 112 (1922); Chern. weekblad, 23, 348 (1926)
'
19 (1930) and Goswami and Sarkar, J. Ind. Chern. Soc., 10, 611 (1933).

2510

QUANTITATIVE MICROCHEMICAL, ANALYSIS

is then swept out for another 25 minutes while the boat is kept at red heat (700750 0 C.). The hydrogen sulfide is absorbed in 4 nil. of sbdium hydroxide and,
at the completion of the run, this is washed with boiled water into a 125 ml.
. Erlenmeyer flask containing 6-7 ml. of iodine and 4 mq of hydrochloric acid,
and diluted with 10 mI. of water. The total volume of the solution should not
be over 30 to 40 nil. The excess iodine is titrated with 0.02 N thiosulfate and
the percentage of sulfur is calculated.
.
1 ml. of 0.02 N.iodine= 0.3206 mg. sulfur.
The iodine solution is standardized against the thiosulfate.
NOTEs.-(l) The sulfur in most organic compounds is quantitatively reduced to
hydrogen sulfide with platinized asbestos, but occasionally some, according to ter Meulen,
require a catalyst of platinum foil or a 6 inch roll of 80 mesh platinum gauze for rapid
conversion. If platinized asbestos is used with such compounds, the sample must be
burned very slowly and with extreme care. H If metallic platinum is used, the presence
of nitrogen in the sample causes the formation of hydrogen cyanide which acts the same as
hydrogen sulfide thus giving high results. In order, therefore, to prevent the absorption
of hydrogen cyanide, a solution of dilute zinc sulfate, approximately 0.05 N, to which is
added sodium acetate, apprmdmately 0.05 N, and a drop of acetic acid must be used.
As before, iodine together with dilute hydrochloric acid, is added, and when all the zinc
sulfide has disappeared, the excess iodine is titrated with thiosulfate.
(2) ter Meulen also states that platinum black (prepared by reducing at a low temperature ammonium chloroplatinate with hydrogen) must be mixed with petroleum in
order to get complete conversion of sulfur to hydrogen sulfide. This catalyst must be
freshly prepared, I1S heating in air causes a diminishing of its activity.
(3) When a zinc sulfate solution is used and zinc sulfide is precipitated, the spiral
type of absorber shown in Fig. 430 cannot be used because of the" difficulty in washing
out the precipitate. In such a case it is best to omit the spiral and make the bore of the
stopcock at the bottom sufficiently large, so that the precipitate can be easily rinsed out.
The Pregl type of absorption tube used with the methoxyl ethoxy apparatus, Fig. 438, is
also satisfactory.
(4) After 5-6 runs, oxygen must be passed through the tube in order to burn out
the organic matter and regenerate the catalyst. To prevent the danger of an explosion,
carbon dioxide should be passed through first to sweep out tJle hydrogen.

MICROELECTROANALYSIS
The Microelectrolytic Cell of Clarke and Hermance. 75_A cell for micro
electroanalysis has recently been constructed wh,ich permits more effective
agitation of the solution during electrolysis than can be accomplished by the
Pregl method of boiling, or allowing the solution to cool during the period of
electrolysis.. The cell is constructeq so that the solution is agitated by a jet of .
air and the solution maintained at a suitable temperature by a water jacket
heated with a micro burner. While the cell is more difficult to construct than
74

75

Rec. Trav. Chim., 53, 121 (1934).

J. Am. Chem. Soc., 54, 877 (1932).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2511

that of Pregl, it yields accurate results even when the deposit on the electrode
weighs 1 mg. The average results reported are 0.1 % hightlr than theory. The
individual determinations had an average deviation of 0.2-0.3% from the
amount recovered.
Apparatus.-The large outer vessel (Fig. 431) sealed to the electrode compartment at its base is a water-jacket which can be heated by a micro-flame.
Figure 432-1 shows the glass frame which in the assembled cell fits inside the
electrode vessel and serves to support both the electrodes and pumping device.

FIG. 431.

Its lower end is flared to forin a conical bell having a diameter about 4 mm. less
than that of the container. While the two electrodes are held very close
togother, they cannot make contact or become fouled by touching the glass wall
of the container. The support permits a free circulation of the electrolyte.
This circulation is accomplished by the tube shown in Fig. 432-II, which is
held erect in the axis of the cell. A stream of bubbles of air or inert gas is
introduced by means of this tube into the space under the bell and causes a
ontinuous overflow of liquid over the top of the tube A that supports the irner .

2512

QUANTITATIVE

MICROCHEMICA~

ANALYSIS

electrode. At the same time, the suction caused by the removal of the liquid
from under the bell further assists circulation in this direction by drawing upon
that portion of the electrolyte in the space between the outer electrode and the
wall of the containing vessel. A rapid agitation is thu1s maintained whereby
the whole of the electrolyte is forced through the inter-~lectrode space several
times a minute.
. "The electrodes are two platinum gauze cylinders 15 mm. high, the one
having an inside diameter of 9 mm., the other 18 mm., weighing 0.8 g., and
1.5 g., respectively."

II

III

AJR TUBE /
o

3 eft
FIG. 432.

Procedure.-The eleptrodes are cleaned with hot concentrated nitric acid,

rinsed with distilled water, dried by heating to oull redness, over the Bunsen
flame, then cooled and weighed. The electrodes' are assembled (Fig. 432-1)
a'nd the unit lowered into the cell vessel containing the solution to be electrolyzed, the volume of which (about 5 ml.) is then so adjusted that its level
coincides with the upper end of the tube A in Fig. 432-1. The .air tube is now
introduced as indicated, and air admitted at such a rate that the liquid may be
seen to flow regularly over the top of the tube A. The electrical connections
are made and the electrolysis begins. The depositln, is completed when the
current falls to a negligible value. The re-solution of the deposit is overcome

QUANTITATIVE MICROCHEMICAL ANALYSIS

2513

by washing<a!J.d withdrawing the electrodes without interrupting the current.

The electrolyte is drained from the cell by opening the stopcock at the bottom
and at .the same time distilled water is admitted at the top of the cell.
If the electrolyzed solution is to be used for the determination of other
constituents, excessive dilution with wash water is objectionable and the device
shown in Fig. 432-III is used, which minimizes the quantity of wash water when
only the 'inner~lectrode is to be weighed.
At the completion of the electrolysis, the air tube is withdrawn and this
perforated tube is inserted in the tube A (Fig. 432-1) and connected to a
reservoir containing distilled water. The cell is slowly drained as before, and
wash water admitted through the tube at the same rate. At the same time
and without interrupting the current, the electrode is drawn up over the tube A
until it occupies a position level with the perforations on the washing tube. The
electrode held in this position is withdrawn together with the washing tube
from the electrolyte. At the time it is' pulled thr~mgh the liq.uid surface it is
surrounded by nearly pure wash water. If the holes in the washing tube are
sufficiently fine and the operations executed rapidly and skillfully the electrolyte suffers a minimum amount of dilution. After removal from the cell the
deposit is further washed with water, alcohol and ether, and then dried in an
oven at a temperature of 90 C. unless different directions are given. The
electrode is protected from dust while cooling, and is finally weighed.
Determination of Metals Using the Cell of Clarke and Hermance: Determination of Nickel.-To a 5 mI. solution containing 1 mg. of nickel 0.2 g.
ammonium sulfate and 1.7 mI. concentrated ammonia solution are added.
The air used for stirring is passed through a concentrated ammonia solution,
which serves to prevent loss of ammonia from the electrolyte. The conditions
for deposition are: e.mJ. 2.4-2.5 volts (60-70 Milliamperes); temperature
60 C.; time 15 to 20 minutes.
Determination of Zinc.-The deposition is made from a potassium hydroxide
solution. The deposit assumes a granular form if the temperature is too high,
the ~alkali content too low, or if hydrogen is evolved at the cathode. The
conditions which must be rigidly maintained are: potassium hydroxide concentration 10%; e.mJ. 3 volts (150 Milliamperes); temperature 25 C.; time
15 minutes.

NOTES. -The cathodic evolution of hydrogen is prevented by giving the platinum

gauze a thin preliminary coat of cop~er. Since the zinc deposit is very soluble in 10%
caustic solution, special care muse be taken to wash out all traces of electrolyte before
interrupting the current.

Determination of Tin ....J..As recommended by A. Fischer 76 the electrode is

first prepared by depositing a thin coat of copper, followed by one of tin from
an oxalate solution. The electrolyte consists of ammonium mono sulfide solution'containing sodium sulfite. The best conditions for electrolysis are: e.mJ.
1.2 volts (1200 Milliamperes); temperature 60 C.; time 20 to 30 minutes.
Afte:r:. electrolysis the electrode is washed with water, alcohol, carbon bisulfide,
alcohol, and ether, in the order given.
NOTEs.-The copper plate should contain about two tenths of a milligram of the
Thetal, plated from dilute nitric acid solution containing sulfuric acid. The electrolyte
76 Z. anorg. Chern., 42, 367 (1904).

2514

QUANTITATIVE MICROCHEMXCAL ANALYSIS

should contain 2% sulfuric acid and 1% nitric acid. Temperature 70-75 C.; e.mJ.
3.5 volts; time to deposit 1 mg. about 20 minutes.
J
The thin plate of tin is next deposited. The electrolyte consists of 5 mI. of a solution
made up to contain 3.6 gm. of ammonium oxalate, 3.6 gms. ofjoxalic acid and 100 mg.
of tin as stannous chloride, in 100 m!. of water. Tl1e solution is electrolyzed at an
e.m.f. of 3.0-3.8 volts; temperature 25 C. The electrolysis is !continued until a bright
coating of tin is obtained over the copper. The electrOde is then removed, washed with
water, acetone, dried at 70-80 C. and weighed.
In the determination of the unknown the tin is dissolved in hydrochloric acid, the
solution evaporated to remove the excess acid, and the moist residue redissolved in 1 mI.
of water. Four mI. of the ammonium monosulfide solution are then added, and the
electrolysis carried out. The monosulfide solution is Prepared as follows: Concentrated
aqueous ammonia is saturated with hydrogen sulfide glis, followed by the addition of an
equal volume of ammonia. Sixteen mI. of the resulting solution are diluted to 120 mI.,
and 8 gms. of sodium sulfite added. The solution shOuld always be freshly prepared.

The Determination of Traces of Metal Impurities by the Method of Clarke

and Hermance.-It often happens that the Volume of a solution is large,
relative to the metal concentration and cannot be reduced owing to high salt
concentration. The apparatus described below IDay be uelld to effect a separation of the trace of metal present in the large 'volume, and the deposit, if

FIG. 433.

contaminated with other material, is redissolved and the final quantitative

determination carried out in the small micro cell preyiouslyldescribed.
The apparatus,77 consists of a combination Qf.-alflift arid micro electrolysis
77

Clarke and Hermance, Mikrochemie, 20, 126-8 (1936).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2515

chamber constructed in a single assembly. This may be used in beakers, tanks,

or vessels of almost any description regardless of size, provided there is a
sufficient depth of liquid to accommodate the pump tube. The liquid continuously raised to a height several cm. above the level of that in the containeI:,
flows from a small reservoir into the cell containing the electrode assembly.
In this chamber a platinum gauze electrode cylinder rests directly above the
inlet tube and is held in position by four glass prongs, spaced at equal intervals
around it. A solid glass ball is placed in the upper open end of this electrode
to divert the inflowing liquid through its meshes. The outer electrode, also
cylindrical, is held concentrically about the inner one by the 4 prongs over.
which it is slipped into position. The liquid flows out through small holes in
the cell wall near the top. By having the electrolysiil chamber independent of
the air pump and the containing chamber, the electrode assembly is readily
accessible for .adjustment, observation, or removal of either electrode. Washing is effectively carried out by simply stopping the pump and flushing out the
chamber with a stream of water introduced below tlle electrode level, from a
suitable nozzle. With this apparatus, 0.1 mg. of copper was quantitatively
recovered from a volume of 20,000 ml. This solution was 0.1 molar sodium
Buliate and ~.l M Bulimic acid.
Determination of Traces of Zinc in Aluminum Solutions.-To 250 ml. of
solution is added enough potassium hydroxide to redissolve the aluminum and
furnish 5%-in excess. The solution is electrolyzed for four hours at room
temperature, at 3.5 volts. If iron is present, it will deposit with the zinc.
The impure deposit is redissolved and the iron precipitated as the hydroxide
between pH 4 and 5. The zinc is then replate.d from the filtrate on a coppered
.
.
electrode using the small cell.
Determination of Lead in the Presence of Zinc.-To 250 ml. of solution
2% of nitric acid, and 25 mg. of copper sulfate are added. The latter functions
as a depolarizer which increases the hydrogen overvoltage and prevents cathodic
reduction of the quadrivalent lead. The conditions for electrolysis are: e.m.f.
2.-8 volts; temperature 6{}--70 C.; time 3 hours. The final determination of
the plated lead is made volumetrically as recommended by Lucas and Grassner. 78
The deposited le~d is brought into solution with 1 m!. of 10% potassium iodide
solution and a drop of concentrated acetic acid. Several crystals of sodium
acetate are then added and the solution rinsed into a emaIl flask with a stopper.
The iodine that is liberated is titrated with 0.001 N sodium thiosulfate solution
(titrate back with N /1000 iodine solution). A micro burette of 2 m!. capacity
graduated in 0.01 m!. is used. 0.100 ml. N /1000 sodium thiosulfate solution
correspondll to 10.36 gamma lead (0.01036 mg.).
NOTEs.-(I) Weighing amounts of lead as low as 0.5 mg. directly as lead dioxide may
be accomplished with the following procedure: 79 The electrodes are formed of 1 inch sq.
platinum foil, 0.09 mm. in thickness, to which platinum wire is welded. They are corrugated and etched with aqua regia. The lead is deposited under the following conditions: Two drops of sulfuric acid, 4.5 ml. of 15-20% nitric acid in 30 ml. of solution,
temperature of 85-90 C., electrolysis at 5.5 amperes per square decimeter for 2 hours.
'The above authors found that even when using large electrodes (8.75 sq. in.) by first
coating with about 0.05 gm. of lead dioxide, they could successfully determine small
78 Mikrochemie, " Emich Festschr.," p. 199 (1930).
79 Day, Delano, and Schrenk, School of Mines and Metallurgy, Univ. of Missouri
Bull., 12, N~. 2, pages 37-41 (1935).

2516

QUANTITATIVE MICROCHEMICAL ,ANALYSIS,

amounts of lead. Conditions are as follows: 25 mI. oUree nitric acid, 0.25 mI. of sulfuric
acid in 140 mI. of solution. This was heated up to 85-90 C.!and electrolyzed at 5.5
amperes per square decimeter for 2 hours. The difference in weight before and after
electrolysis is the weight of the newdeposit. 0.0010 mg. of leadjdioxide was determined
with a maximum deviation of 0.0001 mg. Another method which was successfully used,
was adding known amounts of lead dioxide to the unknown sample; for example, 10 mg.,
which amount can then be accurately determined. The amount of lead dioxide over
th!\t added, gives the amount present in the sample.
A. J. Lindsey 80 deposits from a solution containing not over 6 mg. of lead using 2 mI.
concentrated nitric acid and total volume of solution of 12 mI. The solution is heated
nearly to boiling and stirred with a gas stream of nitrogen. It is electrolyzed at 1.0 volt
(current 120-200 milliamperes) for 7 minutes. The sides of the cell are washed down
. and electrolysis is continued for 3 minutes.

Determination of Traces of Copper in Nickel Solutions.-To 250 ml. of

solution containing 1 % nitric acid and 2% sulfuric acid, a potential of 3 volts is
applied for 4 hours with the solution at a temperature of 60-70 C. The
copper plate is dissolved and redeposited in the small cell and weighed.
NOTES.-Okac 81 has also improved the Pregl cell along the same lines as have Clarke
and Hermance. Carbon dioxide is passed into the solution through an anode tube
during electrolysis and after deposition is completed, the cathode is washed by admitting
water through the same tube.
The electrolyte for all depositions is an ammonium hydroxide solution with a volume
,of 5~ mI. containing 0.5-1.5 mI. of concentrated ammonia. The electrolysis takes place
at room temperature. The results reported have an average deviation from theory of
0.01 mg. when the amount of metal deposited is 0.5 to 5 mg., which is not equal to the
accuracy when using the other procedures described. The method is illustrated by the
following two determinations:

. Cadmium.-Six mI. of a cadmium sulfate solution

containing 1 to 3 mg., of cadmium and 1.5 ml. concen~
trated ammonia are electrolyzed. Conditions for elec~
trolysis are: volts 3; Milliamperes 20-,30 (0.1-0.15
Amp./q.d.m.); time 10 minutes; temperature 20 C.
Mercury.-Six mI. of a solution containing 0.5 to 15
mg. of mercury as the nitrate or, chloride, and 1 ml. of
concentrated 'ammonia are electrolyzed. A platinum
gauze cathode is used. Conditions for electrolysis are:
volts 2.4-2.5; Milliamperes 20-100 (0.1-0.5 Amp./
q.d.m.); time 10 minutes; temperature 20 C. The
cathode is washed with water, alcohol, and ether, and
dried over calcium chloride for one hour. .
The Pregl Cell.-This cell is readily _constructed
from materials in the laboratory and for \hose who
only occasionally need an electrolytic cell, it is adequate.
Typical determinations using this cell are given, but of
course the methods may be used also with the Clarke
and Hermance apparatus and vice versa.
A general view of the apparatus is shown in Fig. 434.
FIG. 434~
The platinum gauze cathode, Fig. 435, is 10 mm. in
diameter and 30 mm. in length, with at stout platinum
Wlfe 100 mm. in length welded to it. Three small.glass beads are sealed to
_.
80 Analyst, 60, 74H (1935).
81 Z. Anal. Chem., 88, 108, 189, 431 (1932); 89, 106 (1932).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2517

the upper and three to the lower edge of the gauze to prevent contact with the
sides of the cell during withdrawal.
The anode is a platinum wire 130 mm. in length (Fig. 435), to which two
glass extensions are sealed. so that it may be more easily kept in the center of
the cell.
The cell consists of a test tube, 16 mm. in diameter, and 105 mm. in length.
The terminals of the electrodes are bent
so as to make contact by dipping into
mercury (Fig. 434).
Procedure.-The gauze cathode is successively dipped into hot concentrated
nitric acid, water, alcohol, and ether, then
dried over the Bunsen flame or in an electric oven, cooled, and weighed. '
Determination of Metals Using the
Pregl Cell: Determination of Copper 'in
Alloys.8L-The material is brought into
solution with concentrated nitric acid.
The amount of copper present for each
determination should be about 10 mg. and
the volume of the solution 5 to 6 ml. If
the lUoy i~ high in copper, an aliquot of the
original soluj;ion is taken for electrolysis.
The cell is assembled, the solution
added, and heated to boiling. The current is turned on, and electrolysis takes
place with an e.m.f. of 2.7 to 3.1 volts.
The flame is removed and the solution
allowed to cool during electrolysis. If the
c~ncentration of nitric acid is too high,
FIG. 435.
deposition will not take place, and in this
case a crystal or so of potassium sulfate is added; and if this is not sufficient,
ammonia solution is added drop by drop until deposition begins. A drop
of alcohol added to the solution at the commencement of the determination
will prevent spraying due to evolution of gas at the anode. A few minutes
after electrolysis begins, the walls of the cell should be rinsed free of any traces
of electrolyte with a few drops of 1% nitric acid solution.
After 25 minutes the deposition should be complete, and without interrupting the current, remove first the anode, then the cathode, and wash the copper
deposit with distilled water. If lead has been present it will be deposited on the
anode and may be removed with nitric and oxalic acid. Benedetti-Pichler
recommends that only water should be used to wash the deposited copper if it
appears spongy, as alcohol tends to cause a flaking off. It is then dried at
90 C., cooled and weighed.

..

NOTEs.-(I) In order to keep the volume of the solution constant during electrolysis,
water must be added or a water condenser (A in Fig. 434) used.
(2) Where electrolysis is carried out in sulfuric acid solution, the time is shortened
by 10 to 15 minutes and in many determinations there is nothing to prevent the use of
82 Benedetti-Pichler, Z. Anal. Chern., 62, 321 (1923).

2518

QUANTITATIVE MICROCHEMICAL ANf\._LYSIS

sulfuric acid. One drop of concentrated acid in 5 ml. of solution is sufficient and an
e.m.f. of 2 volts is maintained. Where organic matter is 'present i~ the electrolyte, the
metal is deposited in the regular manner, but in order to free the deppsit from impurities,
the electrodes are removed, fresh electrolyte placed in the cell, the 'electrodes again assembled, and the current reversed so that the metal goes ba<;k into solution. The current
is again changed and the metal redeposited.
\.

The Determination of Mercury.83-The method will determine 0.1 to 0.9

mg. of mercury with an average deviation from theory of 0.003 mg.
Dissolve the metal in 0.5 to 2 ml. of concentrated nitric acid, transfer to a
measuring flask, and take an aliquot of the diluted solution for electroly'sis.
The cell is a small narrow vessel larger than the Pregl cell and holds 8 to 20 ml.
The cathode is a gold wire 8 cm. in length and 0.3 mm. in diameter and the
anode a platinum wire formed in a spiral. Conditions for electrolysis are:
acidity 0.1 N; current 0.015-0.08 amperes; e.mJ. 3 volts; time 15-20 hours;
temperature 20 C. The electrodes are removed without interrupting the
current, and the cathode carefully washed with water, dried over phosphorus
pentoxide for 3 hours, and then weighed. To remove the mercury, the cathode
is heated carefully over a small flame .
. NOTEs.-Verdino 84 uses the Pregl celI and a gold plated platinum gauze cathode prepared as follows: 50 mg. of gold are dissolved in aqua regia and the solution evaporated
to dryness several times with water to remove the excess acid. The residue is finally
dissolved in 5 ml. of water containing 0.65 g. of pure potassium cyanide. The gold is
deposited at an e.m.. of 3.5 volts; temperature 55 C.; time 2 hours.
The organic material containing 2 to 5 mg. of mercury is placed in a small hard glass
tube, and 10 drops of fuming nitric acid added. The tube is sealed and placed in a bomb
for 2 hours at 270-280 C. After cooling, the tube is broken and the contents rinsed
into the electrolytic cell with 5 ml. of water. The conditions for electrolysis are: e.m.f.
3.5 volts; temperature 40 C.; time 40 minutes. Acidity in the ratio of 1 ml. concentrated nitric acid to 9 ml. of water-or 5 ml. of electrolyte should contain 0.5 m!. concentrated nitric acid. Before removing the electrodes the cell is cooled for 5 minutes
with cold water. The cathode is washed with water, alcohol, and ether. It is dried
by waving in the air and weighed. If the deposit is contaminated with organic matter,
it should be redeposited in fresh electrolyte.

The Determination of Goldy'-It is possible to obtlj-in exlellent results with

a solution containing less than 0.5 g. gold per liter, using a volume as small as
0.5 ml. and electrolyzing from a potassium cyanide solution which is the same
procedure as given for gold plating the platinum cathode in the determination
of mercury outlined above.
Fuchs suggests covering the solution with a little toluene to...protect the
anode from possible attack by' chlorine. The temperature of thesolution
should be close to boiling and a current of 3 amperes at 3-3.5 volts is sui~able.
The Determination of Silver. 8lL_The total volume of aqueous solution is
7 m!. to which are added a few drops of concentrated sulfuric acid, and 1 ml. of
20% tartaric acid. If a wet oxidation to recover the silver has been carried out,
the acid concentration should be reduced by evaporation to less than 1 ml.
Conditions for electrolysis are: 1.3-1.4 volts for 15 minutes, then 1.8 volts for
30 minutes. A sample containing 1-4 mg. of silver is best, but samples containing as low as 32 'Yare reported as being accurately analyzed.
83 F. Patat, Mikrochemie, New Series, 5, 16 (1932).
84 Mikrochemie, 6, 5 (1928).
85 Fuchs, Mikrochemie, 1, 86 (1923).
._
86 Friederick and Rap6port, Mikrochemie, 18, 227-34 (1935).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2519

I Copper in Organic and BioZogicaZMateriaZ.-In many cases. the amount of

copper is too small to be determined by weighing, so that a colorimetric method
is required The following methods have been critically studied and are
accurate within the limits_licated.
Ashing.-A sample containing 0.02 to 0.05' mg. of copper is ashed in a
porcelain, or preferably platinum or silica dish in an electric furnace at a temperature of 510-565 C. for 3 to 4 hours or until ~ clean whitish-grey ash is
obtained. In the case of a liquid it must be evaporated before ashing. The
addition of 5 drops of glacial acetic acid added to 250 ml. of milk speeds up
the process of evaporation and prevents frothing.
Treatment of Ash.-(l) Copper may be separated' from the ash by micro
electrolytic deposition. From 0.02 to 0.05 mg. of copper should be present.
The ash is dissolved in 1 : 1 nitric acid and added to the electrolytic cell as
described under microelectrolytic analysis. 'The voltage should be' kept
between 2.7 to 2.1. If the copper deposit is contaminated with other metals,
it is best to reverse the current in fresh 1% nitric acid and redeposit. If the
amount Df copper is 0.5 mg. or more, it may be weighed directly. If not, it
can be dissolved in acid and treated by any of the colorimetric methods d'escribed below.
.
(2) The ash may be treated directly by the Biazzo method if impurities,
such as iron, which interfere with the colorimetric procedure are absent. If
such impurities are present, then the copper must be precipitated as the sulfide.
(3) Without copper sulfide precipitation.
The ash is taken up in 15 ml. of 1 : 1 hydrochloric acid. The solution is
evaporated in order to render the silica insoluble, the residue moistened with
5 ml. of hydrochloric acid and 5 ml. Of water added. It is then warmed on a
sand bath for 72 hour, filtered, the residue washed thoroughly with water to a
volume of 100 ml., evaporated to 10 ml., cooled and transferred to a 25 ml.
volumetric flask.
(4) With copper sulfide precipitation.
Ferric iron when present in large amounts gives a brownish tinge to the
chloroform layer. The iron is removed by taking up the ash with 10 ml. of
1 : 1 hydrochloric acid, diluted and filtered. The iron is reduced by adding
granulated zinc, copper free, and the solution boiled until it is colorless. It is
then filtered, the filtrate made up to a volume of 200 ml. which gives a solution of
about 1% acid concentration. Hydrogen sulfide is passed into the warm solution for 15 minutes and the precipitation vessel kept stoppered until the copper
sulfide settles. The precipitate may be allowed to settle out overnight. The
copper sulfide is now filtered off or it may be centrifuged. The filter and
precipitate are disintegrated in 5 ml. of nitric acid. The solution is diluted,
brought to boiling, the paper filtered off, and evaporated at 10 ml., and is then
treated by one of the procedures outlined below.
The Biazzb Method. 87-The method is based upon the fact that a' neutral
solution of a copper salt when treated with afew drops of pyridine and potassium
thiocyanide gives a green precipitate of the composition Cu(C 5H 5 N)2(CNS)2
which IS soluble in chloroform. ltJ gives accurate results with amounts of
50 to 150 'Y.
87 R. Biazzo, Ann. Chirn. AppI., 16, 2 (1926); C. A. Elvehjem and W. C. Lindow,
J. BioI. Chern., 81, 435(1929).

2520

QUANTITATIVE MICROCHEMICAL ANALYSIS

Preparation of Standards.-A copper sulfate solution is prepared containing

0.1 mg. of copper in 1 ml. of solution by dissolving 0.3928 gm.1pure CuS045H 2 0
in water and diluting to a liter. The crystals must not show signs of efflorescence. Then 0.5, 1, and 2 ml. of this standard solution lare treated with
acetic acid, potassium thiocyanate, pyridine, and chIorofodn in exactly the
same manner as the sample solution and made up to 25 ml. A standard'is
chosen which has approximately the same intensity as the sample.
One N sodium hydroxide is added to make solution just alkaline to phenolphthalein, 1 ml. glacial acetic acid added, 1 ml. 10% potassium thiocyanate
solution, 10 drops pyridine, 5 ml. chloroform, accurately measured, and made up
to volume. The solution is shaken, the chloroform layer allowed to settle to
the bottom of the flask, and most of the aqueous layer is removed. The
remaining chloroform and water layer is transferred to a Bausch and Lomb
colorimeter and the color is compared to a standard similarly prepared.
NOTEs.-(I) In a critical study of the micro methods proposed for the determination
of copper in organic material, Ansbacher, Remington, and Culp 88 found that better
results were obtained if the chloroform in the colorimetric 'determination was replaced
by bromo benzene as proposed by Kleinmann and Klinke. 89
An aliquot of the neutral solution is placed in separatory funnel, add 10 mI. of water,
25 drops of glacial acetic acid, 30 drops of pyridine, 3 mI. (10%) of potassium thiocyanate,
and 2 mI. of bromobenzene (b.p. 154-155 C.), and the original procedure is followed.
These workers report that the most satisfactory results were obtained when the amount
of copper in the bromobenzene layer was 50 to 150 '),-,(0.050 to 0.150 mg.).

Carbamate Colorimetric Method.-This method 90 gives accurate results

with 1'Y to 5'Y copper with an accuracy of 0.2 'Y.
After obtaining the electrolytic copper or copper sulfide as described before,
it is dissolved in a few drops of concentrated nitric acid and heated on a water
bath for a few minutes. The sample is cooled, diluted to 5 mI., and 1 ml. of
20% ammonium hydroxide added or enough to give a distinct odor. The
solution is cooled, made up to exactly 10 ml. and thoroughly mixed. It is then
centrifuged for 10 minutes to remove any turbidity which :ay have appeared
with the addition of ammonia and poured into another'tube. An aliquot
containing 0.001 to 0.005 mg. of copper is taken for the colorimetric determination. Two ml. of ammonia are added and the volume is brought to 10 mI.,
1 ml. of 0.1% aqueous solution of sodium diethyldithiocarbamate is added
and the solution mixed well. The yellow color is compared .immediately with
the color in a set of standards prepared in the same manner from a cORper sultion containing 1 mg. of copper per ml. The standard should be freshly
prepared for each set of determinations from an accurately standardized stock
solution containing 1 mg. of copper per ml. Quantities of copper below 0.001
mg. cannot be accurately determined.
'
NOTEs.-(l) Ignition vessels, especially new-porcelain, shouid be treated by adding
an alcoholic solution of sodium acetate equivalent to 1 grn. of tlie salt, fonowed by ignition in an electric furnace with subsequent extraction in 1 : 1 hydrochloric add for a
period of several days. All reagents, including filter paper, should be carefully checked
to see .that they are not a source of contamination.

Ind. Eng. Chern., Ana1. Ed., 3, 314 (1931).

Arch. path. Anat., 275, 422 (1930).
'. .
90 Conn and co-workers, Ind. Eng. Chern., Anal. Ed., 7, 15 (1935).

88

89

QUANTITATIVE MICROCHEMICAL ANALYSIS

2521

(2) Both the electrolyti!l and sulfide method are effective in separating the copper
from its impurities. Conn and co-workers recovered 97% by the precipitation method
and 98% by the microelectrolytic method, but the copper sulfide method will recover
smaller amounts of coppttr-as low as 0.002 gm.
(3) If undesQ;Qyed carbon remains after the period of ashing, it is filtered off,
returned to the qUrtz dish and again ashed.
.

GRAVIMETRIC DETERMINATION OF NICKEL AND COPPER

Brunck's dimethyl glyo.xime metho.d is fo.IIDwed. 91 The nickel co.ntent
o.f the sample shDuld be fro.m 0.1-4.0 mg. in. o.rder to. o.btain accurate results.
The o.riginal sample co.ntaining nickel is bro.ught into. so.lutiDn by any o.f the
well-kno.wn pro.cedures. Benedetti-Pichler using 1-1.5 mg. o.f nickel ammDnium sulfate, Dbtained an avetage deviatiDn o.f -0.05 frDm the theo.retical
nickel percentage o.f 14.86.
Pro.cedure.-2-3 ml. Df the sDlutio.n to. be analyzed are transferred to. a
precipitating flask which has been tho.ro.ughly cleansed with chro.mic acid.
A few dro.Ps o.f ammo.nium hydro.xide, so. that the so.lutio.n turns blue, and a
slight excess o.f 1 % dimethylglyo.xime in alco.ho.l are added; usually 4 to. 5 dro.Ps
are sufficient. The precipitate is digested fo.r abo.ut 30 minutes at 70 and then
filtered Dff with the aid o.f the filter tube o.r filter crucible used fo.r halo.gen determinatio.n. The precipitate is washed 2-3 times with a little ho.t water. If
CDpper is to. be determined this filtra;te is saved. The washing is co.ntinued,
using hDt 3% ammo.nia so.lutio.n in o.rder to. remo.ve all the so.luble salts fro.m
the precipitate, finally washed with alco.hDI, and dried to. CDnstant weight at
105 C. If iro.n is present it sho.uld be reduced with sulfur dio.xide.
Copper.-Co.Ppet, if present, may be determined fro.m the first filtrate by'
adding 3-4 drDps o.f 1 % benzo.in o.xime so.lutio.n in alcDho.l. The precipitate is
digested o.n the water bath fDr 30 minutes, filtered and the precipitate washed
with 3% ammo.nia and finally with 50% alco.ho.l, then dried to. CDnstant weight
at 105 C. The precipitate (C 14H l1 0 2N)Cu co.ntains 22.02% cDpper. 'The
sDlutio.n sho.uld co.ntain little if any ammo.nium salt.
AnDther excellent methDd fDr the determinatiDn Df co.Pper is by the use Df
5,7-dibro.m-8-hydro.xyquino.line (C 9H 4 Br/)N)2Cu. 92 The test so.lutio.n is
evapDrated to. dryness in a weigMd micro. beaker with filter stick, and taken up
in 0.1-0.2 mI. o.f dilute (1 : 10) nitric acid, and a few tenths Df a ml. o.f ho.t water
until the vDlume is arDund 2 ml. Then 0.1 ml. o.f aceto.ne is added and the
beaker warmed to. 50 C. A saturated aceto.ne sDlution o.f dibrDmo.-8-hydro.xyquino.line (abo.ut 0.3%) is added drDp by drDp. Every 0.1 mg. Df co.Pper o.xide
requires abDut 1 ml. o.f reagent to. give the usual 3 to. 4 times excess. After 10
91 0. Brunck, Z. angew. Chern., 20, 834; Strebinger and Pollak, Mikrochemie, 2,
'
125 (1924); Benedetti-Pichler, Mettalu u. Erz., 25, 206 (1928).
92 F. Hecht and R. Reissner, Mikrochem, 17, 127-34 (1935); R. Berg, " Mikrochem. '
Emich festschrift," 26 (1930).

2522

QUANTITATIVE MICROCHEMICAL ANALYSIS

minutes on a gently boiling water bath the mixture is filtered warm, and the
precipitate is washed three times with wash liquid (0.4 m!. of 6.5% nitric acid
and 15 ml. of acetone diluted to 20 ml. with water), aoo dried for an hour at
110-1150 C. The method is suitable for amounts up to' 1 mg. of copper
(C,II.Br,ON),Cu contains 9.53% copper.

VOLUMETRIC DETERMINATION OF IRON"

The material containing iron is brought into solution, preferably with
hydrochloric acid, reduced, and titrated in the presence of carbon dioxide with
potassium dichromate using diphenylamine as the indicator. Because the
oxidation of diphenylamine is catalyzed by the presence of ferric ions, a small
amount of ferric chloride must always be added if traces of ferrous iron are to
be titrated. The average deviation from theory when titrating 0.2--0.5 mg. of
ir9n is O.002 mg.
Reagents.-Solution 1-5 g. ferric chloride plus 6 mI. concentrated sulfuric
acid with water to make 100 m!.
Solulion 2-7 m!. phosphoric acid (sp.gr. 1.76) with water to make 100 rn!.
A stannous chloride solution containing 8 mg. stannous chloride per ml.
A saturated 'mercuric chloride solution. I
An N /25.0 potassium dichromate solution.
Indicator-O.1 g. diphenylamine in 1UO ml. of concentrated sulfuric acid.
Procedure.-A sample of material containing 0.2-0.5 mg. of iron is weighed
and brought into solution with hydrochloric acid. The vessel used ior solution
and titration is a test tube 16 mm. in diameter and 10 cm. high. The solution
is heated to boiling and stannous chloride is added drop by drop until the
solution is colorless and then one drop in excess. The sides of the vessel are
rinsed with 4-{i drops of water-the total volume of the solution should be 1-2
mi. Carbon dioxide free from air is now passed through the solution for 5
minutes using a capillary tube. \Vithout interrupting the gas stream, 0.2 ml.
of a cold saturated solution of mercuric chloride is added. The precipitate
forme,:! should have a white silky appearance. Ii a black precipitate results,
too,much stannous chloride has been added .and, metallic mercury is present,
and the determination is spoiled. With the gas stream bubbling slowly
through the solution 1.5 ml. solution 1 are added, then 1 ml. solution 2, and
3 drops of indicator, and the titration commenced using N /250 potassiu.m
dichromate and a micro burette. 'The appearance of the qnal color blue violet
marks the end-point. In ft, blank experiment without the ferrous salt, the
amount of dichromate solution necessary for the color change is determined.
This aino~nt is ded~cted as a correction and the percentage of iron in the sample
!t3 J. Knop, J. Am. Chern. Soc., 46, 263 (1924); Benedetti-Pichler, Z. Anal. Chem.,
73, 200 (1928).

Q\]ANTIJATrYE MICROCHEMICAL ANALYSIS

2523

calculated.J' One ml. of N /250 potassium dichromate solution equals 0.223

mg. of iron.
NOTEs.-Knop and Kubelkova 94 iIi a.study of indicators suitable for titrating small
amounts of iron with potassium permanganate hav~ found that by using suitable indicators 1 mg. of iron can be titrated with an accuracy of 0.1 to 0.5%. Among the suitable
indicators are Erioglaucin A and Eriogrun B. No indicator correction is required.
When chloride is present in the solution it is necessary to have about 0.2 g. of manganous salt present to prevent destruction of the dyestuff because of induced oxidation.
An excess of indicator up to four times the amount neceSS~ry to impart a distinct color
only causes an error of 0.3% when 1 mg. of iron is titrated with 0.005 N potassium permanganate solution.

DET~RMINATION

OF ACETYL

The methods for the determination of acetyl genel'slly involve s preliminary

hydrolysis, followed by a determination of the liberated acetic acid. All the.
methods proposed have their limitations when applied to the analysis of all
types of compounds. The following two methods, h{)wever, are given together
with the types of compounds for which they are best, suited.
The Method of Elek and Harte. 95_The hydrolysis is perf9rmed with
aqueous p-toluene sulfonic acid, followed by distillation in vacuo. The sulfur
dioxide and acetic acid are measured by iodometric titrations. The" method is
applicable to o-acetyl and n-acetyl compounds, except where more than one
volatile acid is produced by hydrolysis, in which case the results are excessively
_high.
Apparatus.-"The apparatus, as shown. in Fig. 4:36, is constructed wholly
of Pyrex glass, with the joints especially ground to be vacuum tight with only
water as lubricant. The connection between the side arm of the flask and the
condenser is glass to glll;ss, sheathed in steamed rubber tubing in order to provide sufficient flexibility to permit tapping the flask. It is important, therefore,
that these parts be made of the same diameter tubing.
"The whole apparatus, with the exception of the rubber connection which
is well steamed, is thoroughly cleaned with chromic acid cleaning mixture, well
rinsed, and dried before using. Between successive determinations thorough
rinsing with distilled water is sufficient and only the flask need be dried."
Solutions:
1. 25% aqueous solution of p-toluene sulfonic acid.
2. 0.01 N sodium thiosulfate.
3. 0.01 N iodine.
4. 4% potassium iodate.
5. Starch indicator.
Procedure.-"A well-ground and thoroughly dried sample of from 4 to
10 mg., accurately weighed by difference on a micf()balance, is placed on the
94

95

Z. Anal. Chern., 85, 401 (1931).

Ind. Eng. Chern., Anal. Ed., 8, 267 (1936).

2524

QUANTITATIVE MICROCHEMICAL ,ANALYSIS

bottom of the hydrolyzing flask, preferably using a long-handled charging tube.

The flask is then filled about three-quarters full with sh6~t pieces (4 to 5 mm.
each) of 4 mm. Pyrex rod. The funnel, wet on the ground joint with distilled
water, is inserted and fastened .with short wire springs. [rwo milliliters of the

,,!,. I I,' I I

oeM.

10

FIG. 436.

sulfonic acid reagent are then added thorough the funnel and the closed stopcock
is sealed with 2 to 3 drops of distilled water added through the funnel. In
the analysis of halogen-containing compounds, a J~w milligrams of silver
sulfate should be added.
. ..,..."The side arm of the flask is connected to the c-ondenser by means of the

QUANTITAT~VE.

MICROCHEMICAL ANALYSIS

2525

freshly rinsed short section of rubber tubing, a glass-to-glass joint being easily
effected. The receiving flask, previously charged with 5 ml. of the 0.01 N
iodine solution with 1 to 1.5 ,grams of potassium iodide (conveniently measured
out as a powder by means of a small marked test tube) dissolved therein, is
attached to the rubber stopper on the lower end of the condenser and adjusted
so that the sintered plate is aboqt 1 cm. above the surface of the liquid. The
side arm of the receiver is capped with a well-washed rubber nipple. The
receiver, up to the side arm, is cooled by immersion in a beaker of finely cracked
ice, renewed from time to time during the hydrolysis after removal of the water
formed by the melting. A stream of cold water, is circulated through the
jacket of the condenser. The flask is tapped sufficiently to mix the sample
intimately with the acid.
"An asbestos board is placed as a shield between the burner and the
condenser. The mixture is heated by bringing the water in the surrounding
beaker to boiling. The level of water in the beaker is maintained during the
period of heating by using an inverted bottle containing water, held well
above the apparatus, and having a 1 cm. tube reaching just to the desired level
in the beaker.
"During the course of the heating, the flask is vigorously tapped at intervals
in order to agitate its contents. Hydrolysis is continued for 1 hour for o-acetyl
and 2.5 to 3 hours for n-acetyl compounds. Where the nature of the linkage is
unknown the longer period is, of course, to be employed. At the end of the
hydrolysis, the hot water is siphoned out of the beaker and replaced by ice cold
water. Mter leaving the apparatus to cool for about 5 minutes, an aspirator is
connected through a manometer and stopcock to the side arm of the receiver
and the system is evacuated to 'a pressure of 50 to 60 mm. The receiver is
lifted, together with its cooling beaker, until the sintered plate is about 2 mm.
from the bottom of the receiver and the water surrounding the hydrolyzing
flask is 'heated. At first a few very small bubbles are seen to pass through
the iodine solution, but the bubbling stops in a few seconds as the pressure is
~qualized throughout the system. When the fluid in the flask has completely
distilled over, aided by occasional tapping, about 1.5 mI. of water are run in
through the funnel without either breaking the vacuum or stopping the heating.
After the contents of the flask have become dry, a second, and finally a third
portion of water is introduced in a similar manner. Heating is then continued
until the contents of the flask are completely dry and for 5 to 10 minutes
thereafter. The receiver is then lowered, using a gentle rotatory motion, until
the sintered plate is 2 to 3 cm. above the fluid. The stopcock between the
receiver and the gage is closed; the connection between the stopcock and the
gage is broken, and then by slowly opening the stopcock the system is brought
back to atmospheric pressure, and the burner is extinguished.
"The condenser is disconnected from the flask and washed through three
times with small volumes of distilled water, flow across the sintered plate being
aided by the application of suction through the side arm of the receiver. The
outside of the lower part of the condenser-is washed into the receiver with
distilled water; the total volume should be about half its capacity. The solution
in the receiver whose side arm is capped with the rubber nipple, is titrated with
the standard thiosulfate, using'2 drops of the starch indicator added toward
the end of the titration. To the now colorless solution about 2 ml. of the

2526

QUANTITATIVE MICROCHEMICAL ANALYSIS

potassium iodate solution are added, and the receiver is immediately stoppered
with a soda-lime tube, placed in a beaker of water at about 35" C., and allowed
to .stand for 20 ~nutes. At the ~nd of ~hat time the soluti<)n is titrated with the
thIOsulfate solutIOn, the end-pomt bemg taken where the blue color which
I
develops during the titration just disappears.
"The value from the initial titration is to be subtra~ted from the value
obtained by titration of 5 m!. of the iodine solution in which 1 to 1.5 grams of
the potassium iodide are dissolved and 2 to 3 drops of dilute acetic acid added.
(It is especially important when measuring the iodine solution ,from a microburet to wait 1 to 2 minutes before reading the buret; the drainage error may
otherwise be considerable.) This latter titration should not be made until the
mixture has stood for about 5 minutes at room temperature. The difference
so calculated is a correction for the sulfur dioxide and varies from 0.00 to 0.05
m!. Where it exceeds the upper limit the analysis should be rejected. The
difference, doubled to account for the dibasicity of the sulfuric acid formed, is
deducted from the final titration as a correction for the acid introduced as
sulfurdioxide."
Calculation: The acetyl content is given by the formula
01 CH CO
10
3

= lOX

volume of thiosulfate (corr.,)'Xfactor of solution X 0.4302

.
weIght of sample

Method of Robinson and Bailey.-This simple method 96 is an adaptation

of the well-known macro method. It is applicable to compounds which do not
possess acidic functions either in themselves or as a result of alkali splitting,
Procedure.-A 5-10 mg. sample is weighed into a small porcelain boat and
introduced into a 20 m!. Pyrex flask. Five m!. of N /25 aqueous sodium hydroxide are added to the flask, and a micro reflux condenser attached. The
solution is boiled gently for 12 to 35 hours. The time required for complete
hydrolysis depends largely upon the character of its acetyl linkage. It has
been found that the acetyl N linkage possesses grea~er stability than the acetyl-O
linkage. After hydrolysis is complete, any alkali adhering to the condenser
and rubber stopper is washed into the flask and the excess alkali titrated with
N /100 hydrochloric acid using phenolphthalein as the indicator. A blank
correction as determined by a control experiment: is necessary. The accuracy
of the method is equal or better than that of Pregl-Soltys--Qr Kuhn-Roth 97
methods. The results are usually low, 0.06 to 0.20, from tIi!:)" theoretical
percentages.
.
96

97

Mikrochemie, 15,233 (1934).

Ber., 66, 1274 (1933).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2527

THE DETERMINATION OF METHOXYL-ETHOXYL

The micromethod uses a modified Zeisel procedure. The RI may be collected in alcoholic silver nitrate and weighed as silver iodide or it may be titrated according to Viebock and Schwappach. 98 The titration method may
be used in the presence of sulfur. The method is based upon the following
reaction: R10R+HI---.R10H+RI. The volumetric procedure is based upon
. the following reactions:
CH aI+Br 2---.CH aBr+IBr.
IBr+2Br 2+3H20---.HIO a +5HBr.
The accuracy of the method is O.l % from the theoretical percentage.
2

Hooks to be on each
side of tube and flask
for tension springs
14mm.

FIG. 437.

Apparatus.-The apparatus described 99 is typical of most and may be used

for either micro or macro determinations. It is constructed of Pyrex glass and
the details are given by Fig. 437.
98 Ber., 63, 2818 (1930); Viebock and Brecher, Ber., 63, 3207 (1930).
99 E. P. Clark, J. A. O. A. C., 15, 136 (1932).

2528

QUANTITATIVE MICROCHEMICAL ANALYSIS

Reagents.-Formic acid, 90-100%.

Analytically pure sodium acetate.
Iodate free potassium iodide.
Red phosphorus.
0.02 N thiosulfate.
Constant boiling hydriodic acid, 57% (store in the dark).
Indicators-starch, methyl red.
'
Procedure.-Five to ten mg. of dried sample or an amount equal to 3-5 m!.
of thiosulfate are weighed into a small glass capsule which will readily slip into
the decomposition flask A. A boiling rod is, placed in the flask. 2.5 m!. of
melted analytical quality phenol and 5 m!. of constant boiling hydriodic acid
are added. The flask is then connected with the remainder of the apparatus
which consists of scrubber B, containing a little water and the receivers, C and
D. The receivers contain 10 m!. of a 10% glacial acetic acid solution of potassium acetate to which 6 drops of bromine have been added. Approximately
6 ml. of the solution is placed in C and 4 ml. in D. (See Note (14).)
A slow uniform stream of carbon clioxide is passed through the capillary
sidearm of the boiling flask and the liquid is gently heated by means of a
mantled micro burner at such a rate that the vapors of the boiling liquid rise
half-way up the condenser. For most substances, 30 minutes are sufficient
to complete the reaction and sweep out the apparatus. The contents of both
receivers are washed into a 250 m!. Erlenmeyer flask which contains 5 ml. of a
25% aqueous sodium acetate solution. The volume of the liquid is then
adjusted to approximately 125 m!. and 6 drops of 90% formic acid added.
The flask is rotated until the color due to the bromine is discharged, then 12
more drops of formic acid are added, and the solution is allowed to stand for
1 to 2 minutes. One gm. of potassium iodide and a few ml. of 10% sulfuric acid
are next added and the free iodine titrated with 0.02 N thiosulfate. A blank
should be run on the reagents, as phenol contains some substance which gives
a small blank.
.'
Calculation.-31.03 parts of ,methoxyl or 45.05 parts of ethoxyl yields 6
atoms or 761.5 parts of iodine.
'
One ml. 0.02 N thiosulfate=0.1034 mg. methoxyl or 0.1501 mg. ethoxyI.
NOTEs.-(l) The boiling rod is a glass tube approximately 60 mm. long and 3.5 mm.
outside diameter with 1 mm. bore. It is sealed at one end and also closed about 10 mm.
from the other. The open end is fire polished. When this is placecrln th~ flask with
the open end down it will induce uniform boiling as long as sufficient heat is constantly
applied to the'l.iquid.
. (2) The hydriodic acid employed usually contains more or less free iodine resulting
from the decomposition of the hydriodic acid. A permanently colorless solution of
hydriodic acid may be obtained if the liquid heated to approximately 100 C. is treated
with slightly more than the necessary quantity of 50% hypophosphorous acid. When
a hydriodic acid solution thus treated is used for al)wxyl determination, no free iodine
will be evolved and consequently no phosphorus need be employed in the scrubber B.
(3) Gibson and Caulfield 100 have studied the Viebock procedure and advocate the
following absorbing solution. It should be freshly prepared by adding 6 drops (0.05 mI.)
of bromine to 18 ml. of 20% aqueous potassium bromide. One-half of ,this serves 'as a
blank and \s kept beside the.absorption tubes during the estimation, being titrated under
the same conditions at the end. The remaining absorp~i~:.solution 'is filled into two
100 J. Chem. Soc., 1935, 1419-20.

QUANTITATIVE MICROCHEMICAL ANALYSIS

2529

absorption tubes, 7.5 ml. into the first, and 1.5 ml. into the second. The absorption
tubes and the blank are immersed in ice water during the estimation.
(4) White and Wright 101 studied the modification of Gibson and Caulfield and find
that the largest error is the action of .light on the bromine in the absorbing brominepotassium Jjtomide solution. They use the same procedure as Gibson and Caulfield, but simplify the absorp.
tion train by using only one receiver. The fig. 438
shows the regular Pregl type absorption tube but modified as follows: Before the outlet tube from the apparatus
is inserted into one of the receivers, a glass sleeve H is
placed at I. When the assembly is complete, the sleeve H
is allowed to fall so that it rests on the constriction of the
receiver as shown. This sleeve, about 1.5 cm. long, is cut
from glass tubing aoout 1 ml. larger than the tube forming the bottom of the receiver and is notched 4 times
around the periphery of the edges with a wet file. When
the sleeve is correctly made, it permits free circulation
between the top and the bot~om of the receiver and yet
it will retain the ascending bubble of gas long enough to
facilitate absorption. Tufts of cotton are placed in the
mouths of the receiver.
Another method is to use a glass spiral which fits
snugly against the wall of the absorption tube and the
tube I, so that the ascending bubbles follow the spiral.
A stopcock may be sealed to the bottom of the absorption tube and the contents washed into the flask without
removing the tube from the apparatus.
(5) If on the addition of formic acid the bromine
does not disappear, it indicates that insufficient sodium
acetate was used.
FIG. 438.
(6) If several determinations are to be run at once,
the same charge of hydriodic acid and phenol may be used
for all of them, but a fresh boiling rod should be used for each determination and the
blank due to the phenol should be subtracted only from the first run as the material
responsible for the blank is destroyed during the first run. This matter, however,
should be checked to make certain that such is the case.
(7) If the hydriodic acid contains free iodine, this must be removed by putting a
water suspension of finely divided red phosphorus in the scrubber B. The apparatus
should be so constructed and operated that no red phosphorus goes over into the absorption tubes and thus ruins the determination.
(8) The time of heating varies with different compounds but there are few that take
over one hour.
(9) The procedure should be standardized as to time and volume of solutions, so
that a constant blank may be obtained for any particular apparatus.
(10) The carbon dioxide from cylinders may be pure enough for use without a purifying train. If halides are present, they should be removed by a solution of silver nitrate
and the gas dried by passing through a dehydrating agent.
(11) An oil bath can be used to heat the flask A, but for rapid routine work the use
of a micro burner is preferable.

(12) The determination of N linked to CH. methylimide can be determined after

0
the ethoxyl methoxyl has distilled over at 135 C. The method is tedious and not satisfactory as a routine determination. The method as given by Slotta and Haberland 102
_ uses the same apparatus as for the alkyloxyl determiriation, except two bulbs as traps,
Fig. 439, are inserted in the distilling tube. The flask is. charged with 3-5 mg. sample,
2 grll.nules req. phosphorus, 1.5 ml. hydriodic acid, 30-100 mg. ammonium iodide, and a
small amount of phenol. The gas washer is filled with a 1.5 % sodium thiosulfate solution
containing 0.5% sodium carbonate. The procedure is then carried out as described
previously in order to distill off any alkoxyl. The scrubber and the receivels are re101
102

Can. Journ. Res., 14, 427 (1936).

Ber., 65B, No.1, 127-9 (1932).

2530

QUANTITATIVE MICROCHEMICAL ANALYSIS

charged and the flask heated to 3500 C. The. hydriodic acid distills over. into the trap
and the methylimide comes off and is carried over by the stream of carbon dioxide.
After heating for one hour with carbon dioxide flowing, the flask is cooled to' 200 0 C.
the hydriodic acid is sucked back into the flask and
.
the flask is again heated to 350~ 'C. The solution in
the receiver is titrated as for alkoxyl. A third heating should not yield a titratable amount in the receivers.
(13) Where a compound yields both methyl and
ethyl iodide, a separation may be made:
A. By the combustion method of A. Friederich.1 02a
The total percentage of methoxy and' ethoxy is first
determined. Then, from another sample, the distilled methyl and ethyl iodide is led into a microcombustion furnace and the percentage of carbon determined. The proportion of oxygen atoms to carbon
atoms can then be calculated. If a methoxy group
is present, the ratio of _oxygen to carbon is 1 : 1; an
FIG. 439.
ethoxy group, if 1 : 2; and if 2 : 3, the proportion of
.
methoxy to ethoxy is 1 : 1.
B. By absorbing the alkyl iodides in an alcoholic solution of tetramethy!amine. 102b
The receivers are charged with approximately 5 ml. each of a 10% solution (by weight)
of anhydrous trimethylamine in absolute ethanol and 10 ml. of absolute ethanol. After
the run, the absorption tubes are stoppered and allowed to stand for 24 hours. Separation is effected by taking advantage of the relative insolubility of tetramethylammonium
iodide in absolute alcohol (the trimethylethylammonium iodide is 100 times more
soluble). For accurate work a correction is made for the loss by solubility ot tetramethylammonium iodide and the amount of loss is best determined by each operator
under the experimental conditions used. The precipitate is dissolved in hot water,
3-5 drops conc. nitric acid are added and the methoxy is determined by adding silver
nitrate and weighing the silver iodide. In the same way the ethoxy in the filtrate
from the trimethylamine separation is determined.
.
(14) 0.3-0.5 m!. acetic anhydride may also be added to the reaction mixture in
flask A.

,
THE DETERMIN:ATION OF MOLECULAR

~IGHT
--._

For Solids-By Rise of the Boiling Point.-The following is a micro modification of the method developed by Menzies and Wright.103 There are two
forms of the apparatus, the chief difference being that in one the solvent is
boned by a small coil of platinum wire placed directly in the solvent 104 while in
the other a small gas or alcohol flame is used.I. Fi~. 440 shows the bottorp
of the boiling tube with platinum coil, and Fig. 441 shows the Smith and Milner
apparatus. B9th are the same' except for the lower section. By using a
A. Friederich, Mikrochemie, 7, 185 (1929).

Willstatter and Utzinger, Ann. Chem., 382, 148 (1911). Kjister and Maag,
Z. Physiol. Chem., 127, 190 (1923). Phillips and Goss, J. A~.o. A. C., 20, 292 (1937).
103 J. Am. Chem. Soc., 43, 2314 (1921).
.J10; A. F. Colson, Analyst, 57, 757 (1932).
105 Smith and Milner, Mikrochemie, 9, 117 (1931).
102.
l02b

QUANTITATIVE MICROCHEMICAL ANALYSIS

2531

differential ther~ometer the observed change in boiling point is much larger

than wlten a micro Beckmann is used, and changes in pressure do not affect
the determination. The volume of solvent is determined,
whichis not done with other microforms of apparatus using
the rise of the boiling point method. Therefore, more accurate
results are possible.
Apparatus.-Fig._ 441-1. T is the boiling tube; S, vacuum
jacket; F, the boiling bulb, 3.0 ml. capacity to large mark.
The neck G is graduated in 0.1 ml.; W, tungsten wire; b1 and
b2 , enlargements to accommodate thermometer bulbs. Fig.
441-II Cu, copper spiral; L, etched millimeter scales; Bl and
B 2 , thermometer bulbs Fig. 441-II1 P, Cottrel pump. The
differential thermometer may be made in the laboratory or
FIG. 440.
purchased.

Procedure.-The boiling tube is clamped in a" vertical position and an

asbestos board 10 cm. square with a 5 mm. hole in its center, is placed under
.
the boiling tube so that the tungsten
!!
T
wire, W, protrudes through the hole.
U
This wire is sealed through the bottom
of the boiling tube and enables the
liquid to boil without bumping and
directs the bubbles of vapor through
C
the pump. Water is then started
Cu
through the condenser andf3.00 mJ. of"
solvent are pipetted into the bulb F.
A micro burner or alcohol lamp is
placed directly under the tungsten
wire and so adjusted that the liquid
will reflux gently from the condenser.
I
A standard thermometer is then inserted into the vapors and the true
II
boiling point of the solvent determined to the nearest 0.1. The
amount of solvent actually present in
s
the bulb F is then determined by reScale
o
moving the burner, quickly checking
1
the ebullition by touching the bottom
2
of F with a beaker of water and im3
mediately reading the volume of the
4
solvent from the graduation G, which
5
b,
has been previously calibrated.
B,
o
The standard thermometer is then
7
removed and the apparatus assembled
8
for the determination of the rise of the
G ;;;
F
9
boiling point. The pump is lowered
10 em.
over the tungsten wire. The differW
ential thermometer is then clamped
FIG. 441.
in such a position that its lower bulb
is directly under the opening of the pump, but not in contact with it. The
thermometer should not touch any portion of the boiling tube. The burner

2532

QUANTITATIVE MICROCHEMICAL ,ANALYSIS

is then returned to its original position and the .liquid boiled briskly enough
to pump it in an almost continuous stream over th~ lower bulb of the
thermometer. Then at intervals of 1 or 2 minutes, readings ,are made on
the millimeter scales etched on the thermometer: Whe? the readings have
become constant, they are recorded. The sample, 10-15 mg., made into the
form of a pellet in a pellet press, is accurately weighed and is then introduced
into the boiling tube. Again the thermometer liquid is allowed to become
constant and the reading recorded. This may require 5-10 minutes. Additional pellets may then be introduced and the corresponding records made.
Calculation.-For the conversion table of the differential thermometer the
original paper should be consulted. lo6 The ebullioscopic constant, K 1 , is for
one mole of solute in 100 gm. of solvent. This requires also knowing the
density of the solvent at its boiling point. The ebullioscopic constant, K 2 ,
is for one mole of solute in 100 ml. ot'"solvent at its boiling point.
Kl = Rise B.P. Xmolecular wt. of solute Xwt. of solvent,
100Xwt. of solute

Rise B.P. X molecular wt. of solute Xml. of solvent

100 Xwt. of solute
Constants Useful in EbuJ.lioscopy

J. Am. Chern. Soc., 43,2320 (1921)

Solvent

B.P. C.

dpldt

dDldt

KJ

K,

Acetone .. ' .... .' .......

Benzene ...............
Carbon disulfide ........
Carbon tetrachloride ....
Chloroform ............
Ethyl acetate ..........
Ethyl alcohol. .........
E-thyl ether ............
Methyl acetate ....... , .
Methyl alcohol .........

57.0
80.1
46.3
76.8
61.2
77.2
78.4
34.6
57.1
64.7

26.4
23.4
24.7
23.3
25.2
25.1
30.3
26.9
26.8
29.6

0.7522
0.8149
1.2223
1.4801
1.4101
0.8302
0.7389
0.6968
0.8802
0.7468

0.0011
0.0010
0.0016
0.0021
0.0018
0.0010

17.2
26.1
23.7
48.0
39.0
27.9
11.7
21.6
20.6
8.8

22.9
32.0
19.4
32.4
27.7
33.6
15.8
31.1
23.4
11.8

o.ooil

0.0011
0.0014
0.0010

NOTEs.-The following precautions should be taken:

.... , __ ,
(1) An almost continuous stream of liquid must be pumped over the lower 'bulb of
the differential thermometer, and most of the ebullition takes place through the pump.
(2) That the flame be shielded from draughts to prevent fluctuations of the thermometer readings.
,
(3) That the upper bulb of the differential thermometer be well laved by the condensed solvent flowing down the copper spiral. Colson (loco cit.) has at the top two
copper wires touching the wall of the boiling tube near the edge of the condenser, which
directs the liquid onto the copper coil on the.thermometer stem.
(4) Water should not be allowed to condense from the atmosphere onto the inner
surface of the boiling tube. This is best accomplished by circulating water at room
temperature through the condenser at a constant rate of about 4 mi. per minute.
With the above precautions the differential thermometer can be read with an accuracy of 0.1 mm. and the volume of- the solvent in the b~b,::E-to 0.03 mi.
106

Menzies, J. Am. Chem. Soc., 43, 2309 (1921).

-'

QUANTITATIVE MICROCHEMICAL ANALYSIS

2533

(5) If the molecular weight of a non volatile viscous liquid is to be obtained, the
liquid may be introduced into the apparatus in a small glass tube, 2 X 15. mm., constricted to " pin point opening at one end. When introduced into the boiling tube the
solvent will quickly remove the solute from such a tube, yet the tube will retain the
sample during the wei~ing.
'
(6) ",hen the solvent is boiled, by the method of Colson (loc. cit.) the problem of
superheating, bumping and shielding the burner from air currents is avoided. The
reading of the volume is taken the moment the current is shut off. No momentary cooling with a beaker of water is necessary, thus giving greater accuracy in the volume
re9<ding.
"
(7) The apparatus of Sucharda. and Bobranski 107 and of Rieche 107 use the micro
Beckmann thermometer and an attempt is made to prevent superheating. However,
for accurate work two set-ups must be used, so that one, with solvent only, will give the
change of boiling point with any change in pressure. No attempt is made to measure the
actual volume of solvent during the experiment. Both methods require 5 ml. of solvent
and 20-30 mg: of material.

Determination of Molecular Weight by the Method of Rast.-The method

of Rast 108 is very simple. Camphor possesses exceptional solvent power for
compounds and gives a high molecular depression of the meltIng point. One
mole of substance in I kg. of camphor as solvent gives a melting point depression of 40 0 C. The method cl!-nnot, of course, be used for those substances
which are insoluble in camphor or decompose below
1700 C. The method really only involves the taking
of a melting point. The melting point of the camphor
is first taken and then a substance of known molecular
weight is added and the depression of the melting point
noted. By this means the camphor is standardized for
further determinations.
A conical capillary, Fig. 442-1, 40 mm. long, 3 to 5
mm. at the top, and 2 to 4 mm. at the bottom, is weighed
exactly on the balance. The sample finely ground is
picked up by pressing a small capillary into the material.
This small capillary with a bore of 2 mm. and 50 mm.
long is wiped free of material adhering to the outside
and placed in a large capillary. A fine solid rod, which
fits into the small capillary, is inserted and forces out
the sample into the bottom of the large capillary.
When a sample 0.2 to 0.3 mg. 'has been added, the large
capillary is again weighed which gives the exact weight of
the sample. The camphor is introduced in the same
manner except that a larger capillary may be used for
2
8
introducing the camphor. When. about 10 times the
weight of sample has been added, the large capillary is
FIG. 442.
weighed the third time and this gives the weight of
the camphor added. The capillary, Fig. 442-2, is now sealed 20 mm. from
the bottom and drawn out to a fine thread 4 cm. long. The capillary, Fig.
442-3, is next dipped into a heating bath, such as concentrated sulfuric acid
so that the material is melted and mixed together. The capillary is now
attached to a melting point thermometer and placed in a melting point
19 7
108

Chem. Ztg., 51, 568 (1927) and A. Rieche, Mikrochemie, 12, 129 (1932).
Ber., 55, 1051 and 3727 (1922).

2534

QUANTITATIVE MICROCHEMICAL ANALYSIS

apparatus. The melting point is then determined with the aid of a lens.
capillary is cooled and a second reading taken to check the first.
Calculation:

At
S=
l( =
L=
Ll =

The

1000 S.K.
Lfl
weight of substance,
moleculitr freezing point depression of the camphor used,
weight Of camphor in mg.,
observed depressing of the freezing poin.t.

o~en

a~s

NOTES.-(l) It
happens that a larger sample may be used
a iarger
capillary Can be employed. It will be found much easier to charge a large rather than
small one !Such as is described above. It is also much easi~r to add camphor to a larger
capillary.
'
(2) For those who find that they have a great number of determinations to.carry out
by tIns method, the apparatus described by Carl Tiedcke lO~ may be of interest. It gives
greater cOhvenience in holding the thermometer and capillary, together with a novel
container for the sulfuric acid. It consists of a 250 ml: fia~k with ground-glass top into
which is inserted a thermometer. The flask is provided with a stirrer and pla.tinum hook
to hold th~ capillary.
(3) JO~eph Pirsch uo suggests Bornylamine (m.p. 1640 C.; mol. m.p. depression
E=40.6) :).nd Camphene dibromide (E=115), as also suitable for the determination of
molecular weight.

For Liquids-Micro Victor Meyer Method.-The principle of the method

consists in vaporizing the liquid under controlled conditions and indirectly
measuring the volume by observing the change in pressure. The method 111
is simple, and 'the entire determination takes about 20 minutes. Another form
of the micro Victor l\1:eyer apparatus was developed by takshrninarayan and
Nayak. l12 The average error is about 2.0%.
l\Ieth()ds based on the Victor )'\feyer principle,lla but which vaporize in a
system initially filled with mercury, .are not as simpl~ to carry out as the above
method.. The volume is determined by weighing the mercury displaced by the
vaporized sample.
Apparatus.-The construction of the apparatus III is simple. The details are
given in Fig. 443. It is essentially an inner jacket A, fitted with a mercury
manometer; the inner jacket is surrounded by an Ollter jacket B, to maintain
constant temperature during the vaporization. A and B are constructed from
a 15 X 150 mm. and a 25 X200 mm. Pyrex test tlIbe respectively. On the
bottom of A is sealed a tube 40 mm. long and 4 mill. in diameter, and three
short lenrsths of rod are sealed at C equidistant arQund the circumference so
that A is centered in the jacket B.
The Pyrex stopcock at D should have a bore nQt less than 2 mm. E is a
capillary tube of 0.5 mm. bore. A 100 mm. length of 5 mm. tubing is sealed
to the capillary,iubing"'at F. G ,is a trap to catch any air rising from the
manometl;lr. J is a water condenser. A reference mark for the mercury
Mik:rochemie, 1; 223 (1935).
...
Rer., 6SB, 1227-9 (1932).
III Bra1,ton and Lochte, Ind. Eng. Chern., Anal. Ed., 4, 365-7 (1932).
112 J~ iridian Chern. Soc., 8, 599-603 (1931).
113 Nieqerl, Trautz and Plenti, Ind. Eng. Chern., Anal. :Ed., 8, 252 (1936).
109
110

QUANTITATIVE MICROCHEMICAL ANAL'YSIS

2535

manometer is etched just below the seal F. Before use the inner jacker A is
filled with water to the reference mark to determine its volume. N is an asl1tstos board with a hole~to admit the jacket B. A is supported in the outer
jacket B by a split stopper. H is a 300 mm. reference scale. Determinations
may be carried out under reduced pressure by attaching th'e rubber connection
K to A and applying suction.
Determination.-The liquid, 2-10 mg., is weighed in the' usu'al manner into
a capillary of such diameter that it will readily pass through the bore of the
stopcock D. The form is shown in Fig. 444-A. The upper end of the constric-

FIG.

443.

tion is closed and a bead formed, Fig. 444-B, which is larger than the bore of the
stopcock. A liquid with a boiling point 15-20 C. above that of the sample i8
poured into B to a depth 10 mm. below the end of the inner jacket A. The
manometer is filled with mercury and A tilted to fill trap, G. It is then placed
inside of B, stopcock D is closed, and the burner under B is lighted. The
liquid is boiled just vigorously enough so that B is filled with vapor at all times.
The level of the mercury in the right hand manometer is brought to the reference mark and D is closed. When no change in mercury level is noted when D
is closed; the determination is begun.
Stopcock D is opened and the capillary ip.serted, the bead of which rests on
the bore of the stopcock. The mercury is brought to the reference mark and
the level of the mercury in I is read on the scale H to 0.2 mm. D is closed, thus
breaking the capillary and it falls to the bottom of A. The levelling bulb is
gradually raised and when no further inc~ease in pressure in A is noted, the

2536

QUANTITATIVE MICROCHEMICAL ANALYSIS

mercury is adjusted to the reference mark and the level read. The difference
in the initial and final readings gives the change in pressure in A. I is lowered,
the plug from the broken capillary r~mqve.d from D and a
long capillary inserted into A, and the :vapor removed by
suction. By making the long capillary slightly smaller than
the sample capillary the latter may be retained on the end of
the long capillary, when suction is applied, and thus is removed from the apparatus.
Those compounds which tend to decompose when heated
to their boiling points at atmospheric can be vaporized at
reduced pressure as follows: The sample tube is inserted in
D and L is connected to D by the rubber pressure tubing
K. L is closed, and M is connected to a vacuum line. I is
lowered so that the mercury at F falls to a point just above
G. L is slowly opened and if necessary I is again lowered.
L is then closed and the mercury brought to the reference
mark. If the mercury level remains constant, the level
of the mercury is recorded in mm., the sample tube broken
by turning D and the 'determination completed as previously described.
:
Calculation:
B
A
MW =22410X760XTXW
..
lOOOX273XVXLlp'
where T = absolute temperature of the vapor bath,
W = weight of sample taken inmg.,
V = volume of inner jacket in mI.,
FIG. 444.
Llp = change of pressure within A in mm.
Collecting Constants.
TXW
M.W. = KXt;:p'
f
I

The constant K for any apparatus can be calculated when the volume of the
inner jacket is known.
N OTES.-(l) The following minor changes will be found to make the apparatus more
convenient:
'
(a) The tube 'I should be ~ade of tubing 6-7 mm. inside diaxneter with a funnel
made of tubing 15 mm. X 100 mm. sealed on. It is easier to read the mer.cury' on the
scale when it is contained in a smaller tube.
.
(b) The tube beginning at F should not have as steep an angle, as there is considerable strain on this tube 'with the weight of mercury.
(c) A vacuum jacket beginning at the point C and extending up to the bottom of the
water condenser will prevent excessive condensation and maintain a more uniform
temperature in A.
(d) If the apparatus is in constant use, a thermometer welt is useful so that the tube A
need not be removed to get the temperature of the boiling liquid
.
(2) If air causes decomposition of the sample, jacket A may be filled with an inert
gas.
.
(3) Bath liquids which can be used are: water, B.P. 1000 C., p-cymerie, B.P. 1800 C.,
alpha-naphthyl methyl ether, B.P. 2690 C., benzyl benzoate,_ B_:P. 3199 C.

,
QUANTITATIVE MICROCHEMICAL ANALYSIS

2537

GAS ANALYSIS

n
=

The apparatus described is designed for

carrying out a gas analysis using 25 to 100 cu.
mm. The degree of precision is of the same order
of magnitude as that ordinarily attained in macro
gas analysis. ll4.
Apparatus.-Fig. 445 is a diagram (Blacet
et al.) drawn to scale of the apparatus. A is
a water jacketed microburette. This is a capillary tube of approximately 0.5' mm. in diameter
and graduated to read in cu. mm. with the zero
mark at the top of the scale. The total capacity
is about 112 cu. mm. The upper end is bent as
shown and ends in a capillary of small external
diameter to facilitate the intake and discharge of
gases. The tip of this capillary is ground and fire
polished to a radius of curvature somewhat less
than that of the top of the, gas holders Hand J.
The lower end of the burette is sealed to a larger
tube containing a trap to insure the exclusion
from the capillary of impurities and chance solid
material which may be introduced in the sealing
process. The device represented by P, R, S, and
B is used to move the mercury and gas sample up
and down the microburette. 'p is an iron
cylinder into which the bell-shaped lower end of
the burette is cemented by the sealing wax N.
The wax is also used here to close the water
jacket of the burette. R represents an elastic
rubber cap. The lower flange of this cap is
firmly secured against the lower edge of P in a
manner which may be deduced from a study of
the drawing. S is a metal plunger which is
attached by a swivel to the screw B. ThtJ space
above R is filled with mercury. By manipulating
B the plunger is forced either up or down causing
a flow of mercury or of gas through the capillary
burette A. Since P is the only one of the metal
parts which comes in contact with the mercury,
it is the only one which must be made of some'
non-analgamating material. By means of telescoped brass tubes C and D and the machined
screw E, everytliing on the table F can be moved
up or down with perfect control. G is a mercury
reservoir of 7 cm. diameter. Hand J are the
containers for the gas sa~ples and are also used

~-..

-M

--, L

0.. rlt.... -- J

FJIf:-:

----H

---- . -" G

--....-c

-- --A

--...... D

tyt;,

u -- ...... E
-~

-. N

,... -_._p
, -'. R

rut:':
FIG. 445.

114 Blacet et aI., Ind. Eng. Chern., Anal. Ed., 3, 266 (1931); 5, 272 (1933); 6, 334
(1934); 9, 44 (1937). Swearingen, Gerbes and Ellis, Ind. Eng. Chern., Anal. Ed., 5,
369 (1933); Seevers and Stormont, Ind. Eng. Chern., Anal. Ed., 9,39 (1937).

2538

QUANTITATIV~ MICROCHEMICAL ANALYSIS

in the processes of absorption and explosion as will be explained later. They

have a capacity of approximately 2 ml. each. In the diagram only two are
shown but in fact 4 of them are symetrically attached by mbans of circular steel
springs to ,the revolving table M (see ,Fig. 445), K, tile absorbent holder
guides, are attached directly on the shaft which supports th,e gas holders Hand
J and are made of a brass tube which acts as a guide for placing and supporting
the absorbent holder L. Since there are 4 absorbent holders, several separate
analyses may be in progress at one time. A rubber band made by cutting a
short length from a suitable rubber tube prevents the absorbent holder L from
extending too far into the gas sample. In Fig. 445, K and J are shown projected on a vertical plane. Their relative positions are
shown more clearly in Fig. 446, in which the spid~
K
supporting the 4 guides is shown properly oriented
above the one which supports the gas holders. The
gas holders are held in position by steel ribbon springs. '
The apparatus may be purchased from Eimer and
Amend, New York City.
'
Collection and Storage of Samples.-It is not always
possible to collect the gas sample directly in the analytical
apparatus. For transportation of such samples, the following apparatus is u~ed. A circular iron pep.umatic
FIG. 446.
trough, 2 cm. in diameter and 1 cm. in depth, which can
be lowered into the mercury reservoir of the apparatus, Fig. 445, is welded to
an adjustable iron arm of a small ringstand. A second adjustable arm carries
a steel spring clamp into ,which the gas holders can be slipped by a hq,rizontal ,
movement. This device makes it possible to take a small
sample of gas to or from the main apparatus without
danger of loss or contamination.
Storage of samples may be accomplished by having a
metal clip, Fig. 447, made to fit around the gas holde~'
and this in turn fit on the rim of a mercury reservoir.
The apparatus described above has been modified by
Swearingen, Gerbes and Ellis, and later these modifications incorporated into an apparatus by Seevers and
Stormont. The design is primarily for rapid'and ac'curate
work on a large number of samples. The use of a horizontal gas burette eliminates the tedious raising and lowering of the mercury column. The apparatus as described
by Seevers and Stormont follows:
The details of the microbm:et, drawn to scale, are
FIG. 447.
shown in Fig. 448. The whole unit is rotated around tpe
horizontal axis of the capillary on the bearings, NN, by the lever, P. The unit
may be easily removed from the bearings, for cleaning the buret, by release of
the thumbscrews. The extent (If rotation of the unit is limiten by the adjustable stop H to prevent damage to the capillary arm. This arm i,B further protected by the support I.
/' '_
'
The microburet is of O.(j6 mm. capillary Pyrex .tjib!rig, fused to a 2 mm.
capillary at the manipulator end to prevent the accidental aspiration of the
sample into. the manipulator reservoir. The portion of the capillary used for

QUANTITATIVE MICROCHEMICAL ANALYSIS

2539

volume measurements has a capacity of 0.18 ml. and is enclosed in an ordinary

50 ml. buret with scale divisions which completely encircle the buret to prevent
parallax. Both capillary and reading buret are further enclosed in a heavy
glass water jacket which fits snugly into a rigid brass tube, threaded at both
ends for the brass end caps and cut away on both sides, as indicated, to allow
Tor reading of the buret. The glass pai:t~ are thus held rigidly in place between
two rubber gaskets when the end plate nearest the buret tip is screwed into
place. Manipulation of the thumbscrew R results in pressure on the 18-gage
Allegheny metal diaphragm L and activation of the mercury, thread. The
metal diaphragm is much e.uperior to any of the nume~ous contrivances which
the authors have utilized to activate the mercury thread, since there are no
parts to replace and hysteresis is reduced to a minimum.
The manipulator reservoir is machined fro~ a steel block, open at the end
to receive the plate T which may be conveniently removed to allow for cleaning
of the capillary. A thin cardpaper gasket, saturated with de Khotinsky wax,

FIG. 448.

and sealed into place after warming the metal parts, effects a mercury~tight
union of .both the end plate and the diaphragm with the reservoir. In as':
sembling, tlie capillary tube is inserted through the unit and spaces VV are
filled with sealing wax over a rubber gasket. Mercury is introduced into the
apparatus through the opening at the top of the manipulator reservoir and
sealed with a'tight-fitting needle valve. As indicated by the arrows in the
diagram, water from a constant-temperature bath circulates by gravity feed
between the capillary and th'e reading buret and returns to the outlet through
the space between the reading buret and the outer jacket. It empties into a
reservoir from which it is elevated to the bath by a water pump. Forced
pumping through the apparatus results in sO'muchbubbling as to make reading
difficult.
The relation of the microburet to the absorption reservoir is shown in
Fig. 449.
By rotation, of the buret unit around its horizontal axis, the buret tip may be
introduced into any capsule of either the sample carrier N or the capsule holder
Q, since the latter two have not only a vertical axis of rotation, but are free to
describe an arc around their respective supporting rods.

2540

QUANTITATIVE MICROCHEMICAL

A,~ALYSIS

Absorption of the various constituents is accomplished in the four capsules'

in the holder Q, which is shaped like a segment of a circle to allow vertical passage of the bead holders when the capsules are turned toward the buret tip.
The four capsules are held in the same horizontal pJane byl the stops A. The
capsules in this holder, as in the carrier, are held firmly in place by a coiled
spring and ball bearing W. The authors have found this to be a most satisfactory method of holding the capsules firmly and stil( allowing for easy

FIG. 449.

removal. Steel forceps made from a sponge clamp are used to handle the
capsules, which are filled by slowly inverting under mercury before insertion
into the holder. Care must be exercised to prevent .small air bubbles from
adhering to the walls of the capsule during this procedure.
The interchangeable bead holders are of glass tubing fastened into the metal
shank with wax, or preferably an adjustable rod of stainless steel fastened to
the bead-holder arm by means of a set screw. The platinum till is of sufficient
length that the glass portion of the holder does not ,enler the'sample. The
bead holders are of uniform size and shape and hel<fin fixed position by two
pins upon the cross arm T (Fig. 449) which is raised or lowered by the rack and

QlJANTITATIVE MICROCHEMICAL ANALYSIS

2541

pIllIOn. It is thu~ possible to insert four beads into the absorption capsules
without any bead coming in contact with the inner surface of tlle capsules.
As reported repeatedly, such an occurrence ruins the analysis. The level of the
mercury in the reservoir may be changed by raising or lowering the Bakelite
block K. This provision is necessary in order to maintain a constant level of
mercury in the reservoir, 'and is shown by the ivory tipped indicator at M.
Preparation of Samples.-All samples must be dried before analysis and the
burette must also be kept free from moisture. It has been found that the
normal vapor pressure of water does not exist in such small gas volumes. The
dry wall of the burette absorbs moisture and erratic results are thus obtained
in the analysis of samples containing moisture.
The Preparation of Beads.-(l) Beads of solid reagent are prepared unless
otherwise stated, by melting the material and taking it up in a small loop of
platinum wire. The platinum wire is sealed into a glass holder as shown in
Fig. 445.
(2) Beads for liquid reagents are made by taking soft glass 100-150 mesh
and heating it carefully until it just begins to fuse. It is shaped to a sphere of
approximately 2 mm. diameter. This is then attached to a straight platinum
wire by heating the wire and letting the end of it become embedded in the
bead. The pores of the bead are filled when it is dipped into a liquid reagent.
Swearingen, Gerbes and Ellis made a bead for liquid reagents by taking a
mixture of 70% grog and 30% kaolin. The grog, ground porous earthenware,
is sifted to a uniform size, passing 120 and retained by 140 mesh screen. The
kaolin should pass-200 mesh. The consistency of the mixture Ilhould be such
as to allow it to be dipped up witj:J. the platinum wire loop in order to get a
smooth surface. The bead is baked in a flame for a few minutes and is then
ready for use.
Operation of Apparatus.-The operation of the apparatus, Fig. 445, as
described by the authors is given below taking the determination of oxygen in
dry air as an example.
Snecific Determinations: Oxygen.-First the containers Hand J are filled
with mercury. This is done by placing a glass tube, which has been drawn out
to a capillary and properly bent, so that its end will go down into the reservoir
G and up to the top on the inside of the containers. If the containers are clean
,and have a top of smooth curvature all the air can be drawn out with ease. The
hi~ surface tension of mercury facilitates this. Several hundred cubic millimeters of the air are introduced into H. This constitutes the sample for analysis upon which several determinations presumably are to be made. If a
Toepler' pump is used for the purpose of introducing samples, its outlet is bent
in the same manner as the tip of the buret and placed permanently in position
in the reservoir and at the same height as the buret tip. The rotating table
M is mounted on ll.ij, adjustable arm so that the containers can not only rotate
completely around the central shaft but may be moved to any desired position
within the re~ervoir. Accordingly, the outlet for the Toepler pump may be at
the back side..of the reservoir and a container can be moved over its tip and
filled.
The next step is to obtain a known volume for analysis. By means of the
screw B all of the air in the buret is replaced by mercury. The tip must be
under the surface of the mercury in G while this is done, otherwise when it is

2542

QUANTITA~rrVE

MICROCHEMICA~

ANALYSIS

submerged it will invariably carry a small aIr 'bubble w.ith it. H is 'now placed
directly over the tip, and by lowering the table F, the tip is brought into the
gas sample. By means of B a portion of the sample IS drawn into t~e buret.
The volume ordinarily taken varies from 25 to 100 cu. :&tm. depending upon the
total amount available and the subsequent treatment to which it is to be sub
jected. The table is then raised so that. the level of the1mercury in the reservoi~
is on the level with the uppermost calibration mark in the water-jacketed part
of the buret. By further unscrewing B, the gas, followed by a thread of mercury, is brought into the calibrated portion for measurement of its volume.
When its upper level is at the same height as the surface of the mercury in the
reservoir, it is then theoretically under atmospheric pressure. However, the
apparent pressure in the buret may be con~iderably in error, owing to the fact
that mercury has a tendency to stick to clean dry glass. If no precautions are
taken two subsequent volumes on the same gas sample may be obtained which
vary as much as 1% from one another. To reduce this source of errorJ.wo expedients are used. First; the buret is gent~y tapped with the finger as the gas
is being brought to the desired position. Second, five independent readings
are always recorded and the average taken in calculating the volume. These
. precautions reduce the probable error in all/cases below 0.2%.
After the volume readings for the run have been recorded along with t.emperature and barometric readings, container, J is brought over 'the buret tip and
the gas in the buret expelled into it. Tlie gas is now ready to have the oxytoo
absorber introduced. The removable holder L has a platinum loop sealed in
its tip. In th\s loop is fused a small bead of yellow phosphorus. The fusion
may ordinarily be done by placing a fresh piece of phosphorus in the loop by
means of forceps and then holding it over a warm resistance coil or even in. the
sunlight. By quickly removing when fusion occurs and placing it under the
mercury, spontaneous combustion may be avoided. In any case it should not
be allowed to remain in the air any length of time, for the moisture of the air
along with oxygen causes a film of phosphoric acid to coat the bead, thus making
it a much slower absorber. As shown in Fig./445,A;he holder is placed in the
guide K. The container J is brought into position and the phosphorus introduced into it. In this operation care must be taken not to allow the glass of
the holder L to touch the walls of the container. It has been found that a cold
tube introduced into mercury. in this way is coated' with a film of air, part of
which will be imparted to the walls of J if contact is made-between the two ~lass
surfaces. On this account the guide is necessary for the introduction of the
absorber up the center of the container. The height of K is-,also arranged so
that only the phosphorus bead and a little platinum wire enters the gas bubble.
Fifteen minutes has been found sufficient time to allow for complete absorption in practically all cases. However, absorption to constant volume is
practiced. After absorption has occurred and th'e absorbent removed, the -tip'
of the buret is brought to the top of the container (more accurately, the top of
the container is brought to the buret tip) and the residual gas taken into the
buret. With the proper manipulation of B, E and M, this operation presents
no practical difficulties. If the entire gas bubble is not drawn into the buret
intact the first time, it need only be driven b~ck:-into the container and the
process repeated. The experimenter soon acquires skill in this regard. The
volume of the residual gas is now measured and the temperature and the

QUANTITATIVE MICROCHEMICAL' ANALYSIS

2543

barometric pressure again recorded. The absorption process is repeated for

5 minutes more, and if no further reduction in volume occurs, all data necessary
for the simple standard method of calculation of the determination are complete.
NOTEs.-(l) The bead of yellow phosphorus as prepared by Swearingen, Gerbes
and Ellis 115 gives very rapid absorption. The presence of moisture gives a film of
phosphoric acid. To remove this, the bead is washed in water, then alcohol and then
dried in a stream of natural gas
(2) The operation ot the apparatus of Seevers and Stormont differs only slightly
from that outlined above, and such variation may be seen from a study of the drawings,
Fig. 448 and Fig. 449.

Water Vapor.-A bead of fused phosphorus pentoxide is used. It is prepared over a glowing electric coil. A flame should not be used because of the
moisture' generated by it.
.
Carbon Monoxide.-Specially prepared dry silver oxide is used. It is prepared by means of a strong base and thoroughly washed by decantation.
W~ile still somewhat moist the solid is compressed into pellets by applying 'a
pressure of 6006 lbs. per square inch. The pellets are allowed to dry at room
temperature: They should not be unduly exposed to air as they take up
carbon dioxide. The tip of a straight platinum wire is dipped in a concentrated
solution of sodium silicate and then touched to a piece of regular shaped silver
oxide 1.5 mm. in diameter. The cement is dry in 10 minutes. Absorption
of the carbon fIlonoxide is complete in 10 minutes. No volatile products are
formed, so no follow-up absorbent is required.
'Ethyl en e.-The sintered glass bead is dipped slowly into fuming sulfuric
acid. The space in the bead becomes filled with liquid and the air is forced out.
The bead is wiped with a piece of filter paper to remove excess acid from the
exterior surface, and then introduced in the usual way into the gas sample.
After 2 minutes the acid is removed and a slightly moist potassium hydroxide
bead is added to take up the sulfur trioxide vapor. This absorption is also
very rapid. After another 2 minutes the bead is removed and the decrease
in volume of the sample due to the presence of unsaturated hydrocarbons is
measured.
Acetylene.-A stiff paste ip made by moistening cuprous chloride with a
dilute potassium hydroxide solution. This is molded in the platinum loop and
heated gently until dry, care being taken not to heat the solid until it turns
dark in color. This bead 'gives complete absorption of the acetylene contained
in an average sample in 5 minutes. The copper acetylide formed becomes an
integral part of the .solid bead and does not contaminate the mercury as it is
being removed from the reacting system. The potassium hydroxide takes up
all of the hydrogen chloride or water vapot formed, so that the decrease in
volume gives directly the amount of acetylene which is-present.
Hydrogen Chloride, Sulfur Dioxide, Sulfur Trioiide, Hydrogen Cyanide,
Carbon Dioxide.-(l) May b~ absorbed by using a bead of fused potassium
hydroxide -'which, before use, is allowed to absorb moisture from the air.
(2) Or, by using a sintered glass bead and a concentrated solution of
potassium hydroxide. The water vapor pressure over this absorbent is negli'
gible. 2 minutes are required for either method.
115

Ind. Eng. Chem., Anal. Ed.,

~,

369-70 (1933).

2544

QUANTITATIVE MICROCHEMICAU ANALYSIS

Ammonia.-(l) By means of a moistened phosphoru:s pentoxide bead.,

(2) By a sintered glass bead using concentrated sulf~ric acid. 'The absorption is rapid. The results are high by 0.25% from theo'ry.
(3) By means of a potassium bisulfate bead. After 4ipping the loop in the
molten bisulfate, the bead is moistened with water and dried in order to leave
a surface coated with many fine crystals.
'
Nitrogen-Inert Gases.-All such gases are determined by difference.
Acetaldehyde and Vapors of Similar Compounds.__:_A solid potassium, 'hydroxide bead is used. This polymerizes the aldehyde to a resin-like substance
which adheres to the bead.

.
The Determination of Combustible ,Gases.-Only: faIr results with smal~
samples are obtained by analyzing such gases using the familiar explosion
method. For this method to operate~ the mixture must be adjusted rather
critically. Blacet and Volman 116 have
developed a method wherebJ; the reaction is carried out on a hot surface
placed in the system. This was accomplished by a platinum loop 'shown in the
operating' position, Fig. 450.
The mounting for the platinum is
made from a piece of soft glass tubing of
the same length and diameter as an
absorbent holder. A 7 cm. length of 2,*gage platinum wire was placed in the
tube in such a manner that about 1 cm.
of it extends beyond one end of the
tube. Starting at this end, a narrow
portion of the tubing is heated carefully;
so that a fusion of glass to platinum
occurred. By slo,fly moving the tube'
through the flame' the platinum wire is
encased in glass for' a length of about 5
cm. It is important that no air bubbles
be trappe'd i,n this process. The glass
and wire are' then bent to assume the
shape shown in the diagram~ '- ,
The exposed platinum is bent so
that the top of the loop is about 1 mm ..
above the end of the glass and the tip of
FIG. 450.
the loop can be immersed in the mercury
within the gas holder . The top of the loop is worked down for a distance of 1
mm. by careful filing and the use of Crocus cloth, uritil the cross section is
approximately one-third of the original. By placing a small amount of mercury
in the unconstricted portion of the mounting, electrical contacts can be made, as
indicated. The large mercury reservoir serves to complete ,the qircuit.
In the use of this combustion coil, it is found conve:t;lie,nt to have available
an electrical potential of about 10 volts and to regulate<the glow of the coil by
means of a slide wire rheostat placed in the circuit.
-

116

Ind, Eng. Chern., Anal. Ed., 9, 44-5 (1937),

QUANTITATIVE MICROCHEMICAL, ANALYSIS

2545

Nitrous Oxide.-A measured volume of pure hydrogen is added to the

sample and the combustion 'coil placed in position as shown in Fig. 450. After
adjusting the coil so that a small amount of the glass insulation appears above
the mercury surface, the current is turned on and regulated until the thin portion of the platinum wire appears reddish orange in color. Complete reduction
. takes pliwe in 4 minutes. The coil is removed and a bead of fused potassium
,hydroxide introduced to remove the water vapor. The volume decrease is a
direct measure of the amount of nitrous oxide originally present.
Methane.-The procedure is the same as above except oxygen or carbon
dioxide free air is introduced. After the water vapor has been removed the
carbon 4ioxide produced is determined in the usual way and represents the
amount of methane in the original sample .
. Hydr()gen-Oxygen'l-Hydrogen may be determined by the use of oxygen
I!nd vice versa, by using the combustion method described above.
A bead made by fusing together cupric oxide and potassium hydroxide can
also be used for the rapid and complete absorption of hydrogen. The reagent
is prepared by first fusing the oxide in the platinum loop and then touching the
molten mass to an approximately equal amount of the solid hydroxide. The
two substances .coalesce and, afte~ cooling, a smooth dark blue bead results.
By l;teating the gas sample ip. the presence of this solid, hydrogen is rapidly
oxidized and the water formed in the reaction is absorbed at the same time.
'l'he decrease in volume represents directly the amount of hydrogen in the
sample. The total absorption tim~ will not exceed 15 minutes.
N OTE.-The method of surface combustion cannot be used if the time for the reaction
to go to completion causes oxidation of the mercury. There is some evidence for this
at ~h~ end of 15 minutes.

Key References to Additional Applications and Methods

References are given below to micromethods and determinations which
will be of value to those working on special problems. Bibliographies, together
'with brief details of the procedures on specific determinations, appear frequently
in the volumes of Mikrochemie and Mikrochimica Acta.
Applications of Microanalysis
1. Applications of Microanalysis to Agricultural Industries.

de Saint-Rat IV Congr.
intern. tech. clum. indo agr. Bruxelles, 2, 80-101 (1935).
2. The application of microanalysis to the analysis of commercial products is carefully reviewed by Edmund Grundsteidl (Mikrochemie, 16,247-320 (1934. Papers are
revie~ed which appeared from 1918 to 1928.
Mikrochenlie beginning volume 14 (1933-34) reviews the microchenlicalliterature
for the preceding year and has a section on technical analysis. This supplement appears
regularly with each volume.
3. Mineral Analyses. Microanalysis of Silicates.-Determination of Iron, Aluminum, Magnesium, Calcium, Titanium and Manganese (Preliminary announcement).
Karl Schoklitsch, Mikrochemie, 20, 247-53 (1936).
Microanalysis of Monazite.-A method for the analysis of small quantities of
monazite. Hecht and Kroupa, Z. anal. Chem., 102, 81-99 (1935).
4. Microchemical Analysis of Steel. Determination of Manganese. Andre Leroy,
Bull. Soc. Chim. (5), 3, 1125-6 (1936).
'5. Photometric Microanalysis of Water.-C. Urbach, Mikrochemie, 10, 483-504
(1931-32); 11,37--60 (1932); 13,31-54,201-224 (1933).

2546

QUANTITATIVE MICROCHEMICAL ANALYSIS

J'

Special Methods of Analysis

1. Quantitative Analysis of Unweighable Amounts of Materilll. F. Emich, Mikrochemie, 13, 283-288 (1933). Suggestions are made as to the use of an "enlargement"
process whereby the final factor for calculation may be increaied. Several examples
I ,
together with possible methods are given.

2. The Heyrovsky Polarograph. The principle is not new but the design of the
apparatus and method of analysis was developed by J. Heyrovsky! of Charles University,
Prague.
By using a dropping mercury cathode and by constantly increasing the voltage to a
sample in solution, a point is reached, characteristic for each reducible compound, where
the intensity of the current suddenly increases. The intensity varies with the concentration of the substance. Only a few drops of solution are, required and the sample is
not destroyed by the analysis. While the method is not recommended as a routine
procedure of general application it has been found useful, by workers in this cOlrntry and
abroad, in the study of special problems dealing with the quantitative analysis of traces
of metals, and some inorganic and organic radicals. The usual limit of sensitivity is
0.0001 mg. A study of the method has been made at the University of Wisconsin, and
numerous references to the method appear in the literature. A booklet, "Chemische
Analysen mit demo Polaragraphen" by Dr. Hans Hohn, published by Julius Springer,
Berlin (1937) deals with the polarograph and its practical applications in a very thorough
manner.

3. The Use of Fluorescent Analysis in Michrochemistry. Max Haitinger, Mikrochemie, 16, 321 (1934-35).
'
Physical Methods'
1. Microcalorimetry. Brian Whipp, Phil. Mag., 18, 745-59 (1934). Determina-tion of Heat of Combustion with the Microbomb. J. J. Vrigling, Chern. Weekblad., 32,
20-2 (1935). Researches on Calorimetry. Generalization of the Method of Electrical.
Compensation. Microcalorimetry. A. Tian, J. chim. phys., 30, 665-708 (1933). A
Semimicrocalorimeter for Measuring Heat Capacities at Low Temperatures. Daniel
R. Stull, J. Am. Chern. Soc., 59, 2726-33 (1937).
2. Density. Micropycnometer Method for Density Determinations. G. R. Clemo'
and A. McQuillen, J. Chern. Soc., 1220 (1935). Micro Density Deter,mination of Solids
and Liquids. Eugene W. Blank and Mary L. Willard, J. Chern. Ed., 10, 109-12 (1933).
3. Distillation,. Microanalytical notes: 1. Some Improvements in Methods of
Dealing with Small Quantities of Liquids and Precipitates. B. L. Clarke and H. W.
Hermance, Mikrochemie, 18, 289-298 (1935). This deals with improved methods of
filtration, evaporation and distillation. Fractional Distillation of Extremely Small
Volumes of Liquids. Benedetti-Pichler and Rachele, Mikrochemie, 19, 1-5 (1935).
Separates 10 cu. mm. into fractions of 2 cu. mm. Microdistillation of Liquids. A
Microdistillation Apparatus. Lyman C. Craig, Ind. Eng. Chern., Anal. Ed., 8, 219-20
(1936). Isolation from Human Tissues of Easily Volatile Organic Liquids and their
Identification. A. O. Gettler and H. Siegel, Arch. Path., 19, 208-1.2--(.lQ_35). Microfractionating Column for Liquids Having Low Heat of Vaporization. PaillEl-. Weston,
Ind. Eng. Chern., Anal. Ed., 5, 179 (1933).
4. Extraction. Micro Fat Extraction. Microextractor. L. Titus and V. W.
Meloche, Ind. Eng. Chern., Anal. Ed., 5, 286-88 (1933). '
5. Microrefractometer. A microrefractometer (for liquids) and its use in chemical
microscopy. E. E. Jelley, J. Roy. Microscop. Soc., 54, 23H5 (1934).
6. Viscosity. Microviscometer. F. M. Lidstone, J. Soc.' Chern. Ind., 54, 189-90 T
(1935). Viscosity Determinations with Small Amounts of Material. Otto Merz,
Farben. Ztg., 37, 1192-3 (1932). A Microsaybolt-type Viscometer. Stuart M. Rogers
and Linden R. Adkins, Ind. Eng. Chern., 20, 742 (1928).
/
Detennination of Organic Radicals' - ~.1. Carbonyl Group. Microdetermination of Carbonyl Groups. 'F. Frhr. v. Falkenhausen, Z. Anal. Chern., 99, 241-57 (1935).

QUANTITATIVE MICROCHEMICAL ANALYSIS

2547

2. Carboxyl. A Micro Method for the Determination of Carboxyi Groups in

Organic Acids. Tsurumi and Sasaki, Science Repts. Tokohu Imp. Univ., 1st Ser., 19,
681-8 (1930). They employ a micromodification of the macro procedure of Hurter
and Edward, J.A.C.S., 35, 452-61 (1913).
3. Active Hydrogen. A. Soltys, Mikrochemie, 20, 107-125 (1936) (Mikro-Zerewitinoff). P. M. and G. F. Marrian, Biochem. Jr. London, 24, 746 (1930). H. Roth,
Microchemie, 11, 140 (1932). Micro Determination of Active Hydrogen with Deuteridm Oxide. R. J. Williams, J.A.C.S., 58, 1819-21 (1936).
Fats:
4. Microdetermination of the Reichert-Meissl-Wollny, Polenkske and New Kirschner
.
Numbers, H. Vasbinder, Pharm. Weekblad., 71, 1193-7 (1934).
Methods of Analysis for Specific Compounds or Elements
A. Benedetti-Pichler, Mikrochem. Pregl Festschr., 9-13 (1929)
(gravimetric determination).
2. Benzene. A Microcolorimetric Method for the Determination of Benzene.
Pearce, Schrenk and Yant, Bur. of Mines, Report of Invest. No. 3287 (1935).
3. Microcolorimetric Determination of Benzene in. Blood and Urine. Pearce, Schrenk
and Yant, Bur. of Mines, Rept. of Invest. No. 3302 (1936) ..
4. Bismuth. The Micro-electrolytic Determination of Bismuth and Lead and their
Separation by Graded PotentionaI. A. J. Lindsey, Analyst, 60, 744-6 (1935). MicrodetermiI~ation of Bismuth as Bismuthyl Iodide. Hecht and Reissner, Z. Anal. Chern.,
103,283-8 (1935).
5. Calcium. Microdetermination of Calcium in Sea Water. Kirk and Moberg,
Ind. Eng. Chern., Anal. Ed.; 5, 95-8 (1933).
6. Microdetermination of Carbohydrates. Collected References. 1. Z. Dische,
Mikrochemie, 10, 127-187 (1931). Gollected References. II. A. Wasitzky, Mikrochemie, 16, 87-114 (1934-35). The above references bring the subject up to 1934 and
,also outline the various procedures.
7. Palladium. Microanalytical Determination of Palladium by Means of Dimethylglyoxime, Methyl Benzoylglyoxime and Salicylaldioxime. H. Holzer, Z. anal. chern.,
95,392-400 (1933).
8. Platinum and Iridium. Microdetermination of Platinum and Iridium and of
Associated Chloride and Potassium. Drew, Treso and Wyatt, J. Chern. Soc., 1787-90
1. Aluminum.

~934).

9. Toluene. A Microcolorimetric Method for the Determination of Toluene. Yant,

Pearce, Schrenk, Bur. of Mines, Rept. of Invest. No. 3323; 311 (1936).
10. Zinc. Micromethods for the Determination of Zinc. P. L. Hibbard, Ind. Eng .
Chem., Anal. Ed., 6, 423-4 (1934).

METALLOGRAPHY 1
Metallography is that branch of the science of metallurgJ:" which treats of
the constitution and internal structures of metals and 'their relations to thO.
physical properties. The success of 'its application to the examination of
metals or alloys is dependent to a great extent not only upon a knowledge
of their chemical composition, method of manufacture, and mechanical or
thermal treatment, but also 'on the fundamental laws which relate to their
fusion and solidification and the internal changes which occur while. whole or
partly s~lid.

THE MICROSCOPE
The application of the compound micro;cope to the study of opaque
objects examined by reflected light introduces conditions somewhat different
from those where thin transparent sectionS' are used. The fundamental laws
pertaining to the manipulation and use of the microscope are, however, applicable in either case.
There are two methods of illumination: (1) normal or vertical illumination'
in which the light is directed on to the specimen at right angles to its surface;
(2) oblique illumination in which the light is directed obliquely to the spe~men
from a source outside the lenses of the instrument. Fig. 451 is a sketch showil!g
the two methods.

FIG. 451.

Oblique illumination is the simplest because no special attachments to the

microscope are necessary, but it is valuable only when low magnifications
are used and where it is a question of examining small differences in level
on the polished and etched surface. Vertical illumination
is the method more
_r
1 Chapter by Joseph Winlock, Chief Metallurgist, Edward' G. Budd Manufactur
ing Co.
2548

METALLOGRAPHY

2549

commonly used. In this system the light eI,lters the instrument through an
aperture in the side of the tube and strikes a disc of glass set at an angle of
45 degrees which reflects the light downward to the specimen. The light
reflected by the surface of the specimen then passes upward through the disc
and forms the'image. If desired, a totally reflecting prism may be used instead
of the plain glass disc,2
"
.
The choice of lenses depends upon the character of the metal and upon
the magnification desired, but for ordinary work the following types 3 are
suggested as being suitable:
(a) For magnifications from 0 to 10: a photographic lens of approximately'
72 mm. focal length to be used without ocular.
(b) For magnifications from 10 to 30: a photographic lens of approximately
35 mm. focal length to be used without ocular.
(c) For magnifications from 30 to 75: an achromatic 4 objective of approximately 32 mm. focal length with Huyghens ocular (approximatelyX5).
(d) For magnifications from 75 to 150: an achromatic objective of approximately'16 mm. focal length with Huyghens ocular (approximatelYX5) .
. (e) For magnifications from 150 to 1000: an apochromatic objective of
approximately 4 mm. focal length with an ocular compensated for apochromatic
objectives.

The magnifications in microP!lOtography should be accurately determined

by measuring the image of a stage micrometer scale and not by calculations
from lens combinations and projection distances.
In oraier to overcome slight degrees of chromatic aberration even in apochromatic objectives, the use of monochromatic light is preferable and for
this purppse the light before reaching the I'lpecimen should be allowed to pass
through 1\ colored glass screen. Filters having a dominant wave-length of
5500 A. U. (i.e., yellowish green) are very satisfactory for general use. In
very high magnification violet
blue filter is recommended.

or

2 A slight degree of oblique illumination may be obtained by rotating the reflector

of the vertical illuminator.
3 It should be borne in mind that increasing the magnification by increasing the
bellows length of the camera adds nothing to the detail of micrograph. This depends
for the most part upon the resolving power of the objective used. See also Patterson,
W. L" Trans. A. S. S. T., Vol. 11, No.2, Sept., 1921.
.
4 Achromatic lenses are those corrected for two colors.
6 Apochromatic lenses are those corrected for three colors.

2550

METALLOGRAPHY
PHOTOG~HIC MATE~S

A medium speed, moderately fine-grained orthoch:r:omatic photographic

plate such as Eastman's Wratten and Wainwright l\ietallographic Plates'
should be used. In order to bring out the details of the i{:ltructure, a printing
paper wit~ a glossy surface should be used, e.g., Eastm~n's Regular Glossy
Velox. The maximum degree of gloss may be obtained by drying the p'rints
on "ferrotype" plates which have been previously wiped with a solution of
paraffin in benzol and then polished with- a clean dry cloth.
Complete directions regarding the process of developing and printing are
furnished with each box of plates and paper and for the. best results should'
be rigorously adhered to.
Full information regarding magnification, type and kind' of ocular and
objective used should accompany each photomicrograph.

PREPARATION OF SPECIMENS FOR MICROSCOPICAL

EXAMINATION
The sample chosen for microscopical examination should be as characteristic
of the specimen it represents as is a sample for chemical analysis-the size
and location depending upon the nature of the article to be examined. Samples
of rolled objects, for example, are usually taken so that the surface examined
is longitudinal to the direction of work. For convenience in, handling, a
sample from! to ! inch square and i of an inch thick is best.
Soft metals may be cut with a hand or power hacks,w and hard material,
by means of a thin emery disc or broken with a hammElr. Great care should
be taken at all times to prevent the metal from becQming unduly heated during
the cutting operation on account of the change! in structure which might
result. If the sample is small, such as wire or thin sheets, some sort of mounting
device such 'as a clamp or small container filled with a low melting alloy 6 in
which the sampie may be imbedded is desirable. If the edge_(_lf the sample is_
to be examined, it is sometimes convenient-particularly with iron-and steelto plate the sample with a thin coating of copper to prevent the 'dges from
becoming rounded during the grinding and polishing operations.
The coarse scratches caused by' the saw or emery disc are removed by
grinding on a fine grade emery wheel such as alundum grade 80 P revolving
at a speed of approximately 1000 r.p.m. From the erpery wheel 7 the sample
is taken to a canvas-covered disc revolving at a speed of about 400 r.p.m.
.
arm,ed with flour emery suspended in water.
0
6 A very satisfactory alloy melting at about 50 , C. can be made up as follows:
Lead 30 grams, bismuth 50 grams, tin 25 grams, and zinc 3 grams.
I
7 Between each grinding or polishing operation the saIIlllle should be thoroughly
washed in water. Subsequent polishing scratches should""'always be at 900 to those'
immediately preceding.

METALLOGRAPHY

2551

Mechanical polishing is generally much 'more convenient and satisfactory

than hand polishing. For mechanical polishing a disc of any convenient size
covered with a fine-grade broadcloth and revolving at about 400 r.p.m. has
been found to be suitable. Jeweler's rouge or Ievigated alumina are the
principal powders used. The 6-hour alumina is a rapid polishing powder and
is used on steel and hard materials. Two ounces of the powder are taken for
100 ounces of distilled water. The 12-hour alumina is more comparable to
jeweler's rouge and is' generally preferred. It is particularly useful for the
non-ferrous alloys such as brass, bronze, German silver, etc. One ounce of
the powder is taken for 100 ounces of distilled water. The 24-hour alumina
is very fine and is therefo,\e used only in special cases such as preparing sections
for examination under high magnifications or for polishing very soft alloys.
Six-tenths of an ounce of the powder are taken for 100 ounces of distilled water.
Certain alloys, such as those rich in lead, become tarnished if water is used for
polishing in which case the water should be replaced by alcohol or a thin oil.
With hand grinding and polishing, the sample is rubbed on each of the
following emery papers in the order given:
1. Rough commercial No. O.

2.
3.
4.
5.

Smooth commercial
French-Hubert
French-Hubert
French-Hubert

No. 1)0.
No. O.
No. 00.
No. 000.

, Final polishing may be conducted on a block of wood covered with broadcloth armed with jeweler'S rouge or.levigated alumina.
" After grinding and polishing, the samples should be carefully washed in
alcohol and dried in an air 'blast or with a soft cloth. They should be kept
in a desiccator until ready for microscopical examination ..

FIG . .452.

Equipment for Photomicrograph.

(Courtesy of Bausch & Lomb.)

2552

METALLOGRAPHV

ETCHING SAMPLE FOR EXAMINATION

I
,
Ther.e are occasions when the sample is exaZUined without etching, e~g,:
if the metal is thought to contain an excessive a.tnount bf dirt, slag, or other
non-metallic inclusions. The graphite in gray calilt iron lor malleable castings
or manganese sulfide in steel, etc., may also be seen b.:lfore the sample has
been etched.
In the majority of cases, however, some suitable etching medium is employed
to reveal the internal structure by differentially attacking the constituents
or crystalline grains of the metal. A few of the more important' etching
reagents are given in the following list.
.
Nitric Acid and AIcohol.-A 3% alcoholIc sOlution of nitric acid is used
for etching plain carbon, alloy steels and cast irob.. The sample is dipped in
the solution from two to ten seconds, washed in a.lcohol and dried in air blast
or with soft cloth.
Picric Acid and AIcohol.-A 5% alcoholic sOlution of picric aCid is also
used for steels and has been found especially valuable in revealing the structure'
of pearlite and sorbite.
'
Concentrated Nitric Acid (sp.gr. 1.42) (after ~auveur).-Used for etching
iron and steel.' Sample is dipped in acid where it assumes the passive state.
Then held in stream of running water causing.it t<) be momentarily vigorousl:J\
attacked. It is then washed in alcohol and dried.
,
Sodium Picrate.-Solution made by adding ~ parts of picric acid to 98
parts of a solution containing 25% of caustic soda. Sample is immersed from
4 to 6 minutes in solution heated to boiling, then 'washed in alcohol and dried. '
Colors cementite black, leaving ferrite unaffected.
Yatsevitch's Reagent.-Ten milliliters of h)rdrogen peroxide added to
20 milliliters of ,a 10% solution of sodiuJll hydra'te in water. Sample is immersed from 10 to 20 minutes in solution, washeq in alcohol and dried. Reagent should be prepared fresh every day. Used f'or high-speed steels, coloring,
the special carbides brown or black.
"
Le Chatelier No. 1.-200 ml. methyl alcohol, 20 ml~ distilled water, 4 ml.'
hydrochloric acid (sp.gr. 1.18), 2 grams cupric c1iloride, 8 grams magnesium
chloride and 1 gram of picric acid. This solution is used to reveal the macrostructure (structure visible to naked eye) and local segregations, particularly
phosphorous. Sample is immersed in reagent until a thin film of copper has
, been deposited on entire surf:;tce. Copper then washed fromspeciI!1en with a'
piece of absorbent cotton dipped in ammonia.. Sttmple then washed'in water,
alcohol and dried. Operation is repeated untIl stI'ucture is plainly visible.
Ferric Chloride (Moore and Gilligan).-40 grams of ferric chloride, 3
grams cupric chloride, 40 ml. hydrochloric acid, arta 500 m!. of distilled water.
The modus operandi is the same as with Le Chateher No. 1.
.
Ammonia.-Used for etching copper alloys, gun :metal, bell metal, cast
bronze;. etc. To 5 ml. of ammonia are added a fe", drops of hydrogen peroxide
and applied to sample with a small wad of absorb~nt cotton. ,
.

METALLOGRAPHY

2553

EQUILmRIUM DIAGRAMS
The equilibrium diagrams are based on thermal and microscopical data
and show the changes observed in many alloys of the same series 8 when
passing from the liquid to the solid state (or vice versa) and, if any, those
taking place after the metal has completely solidified.
There are three main types of equilibrium diagrams relating to the solidification o~ metals and alloys, viz., those binary alloys (1) whose component
metals are completely soluble in each other when solid; (2) whose component metals are partly soluble in each other when solid; and (3) whose component metals form a definite compound.

FIG. 453.

Figure 453 9 shows the equilibrium diagram obtained by plotting several

cooling curves of alloys whose component metals are completely soluble in the
solid state. It will be noticed that the pure metals at each end of the series
solidify at definite temperatures,I and not over a temperature range as'in the
case of an alloy (the portion of curve 2 from A to B).
S With the few exceptions where a pure metal is used, the metals or alloys used in
commercial practice are composed chiefly of two elements, The other elements often
I1dded to impart special properties and those whose presence is due to the prohibitive
cost of removl11 may be considered separately.
9 Mter Sauveur.
10 "Surfusion" or undercooling may take place if the cooling is slow and undisturbed.
Surfusion or superfusion is the unstable condition of a body which, under certain
conditions, e.g., slow and undisturbed cooling, may remain liquid after the temperature
has fallen below the freezing point.

METALLOGRAPHY

2554

The appearance under the microscope of alloys of this series is similar to

that of a pure metal as in Fig. 454 which shows the str;u~ture of almost pure
iron magnified 100 diameters. The photomicrograph sh~ws a polygonal network, indicating that the metal itself is composed of irregular polyhedral
crystalline grains, each polygon representing a section th*ough a polyhedron.
The dissimilarity in shade of the grains is due to the vary~ilg orientation of thp.
different grains.
I
II

FIG. 454.

Photomicrograph of Almost Pure Iron Magnified 100 Diameters,

Etched in a 4% Alcoholic Solution of Nitric Acid.,
,

The alloys of which this diagram is typical I1re'those of gold and platinum, copper
and nickel, gold and silver, etc,
The equilibrium diagram of alloys whose component metals are partly soluble in
the solid state is shown in Fig, 455.

Composition

FIG. 455.
I

The portions at each end of the diagram appear similar to that 'shown in
Fig. 453 of alloys whose component metals are completely soluble in the solid
state whereas the middle portions show the diagram typical of those alloys
whose component m~tals form a eutectic. A photomicrogra,.vh taken at C
would appear as in Fig. 456,11 Here the metal in exceEjs_Qf the ,e utectic ratio has
_r11

After. Sauveur.

METALLOGRAPHY

2555

solidified first, forming the major portion of the alloy with the matrix composed
of the eutectic 12 of the two metals which solidifies along the line AB. At E

the structure would be composed entirely of eutectic as shown in Fig. 458.

At D the reverse of the conditions appearing at C would. have taken place
and the structure would appear as' in Fig. 457.
The alloys of which tliis diagram is typical are those of silver and copper,
lead and antimony, etc.

Composition

FIG. 459.
(from the Greek EV'f'1K'fO$, easily melting), the alloy of a certain series
possessing the lowest temperature of solidification. Their structures are made up of
alternate plates or "lamellae" of the different constituents.
12 EUTECTIC

METALLOGRAPHY

2556

The equilibrium diagram of alloys whose component metals form a definite

compound is shown in Fig. 459-the definite compounU being formed at C.
If this diagram is divided into two parts at the line AB, it is noticed that.
each of the parts resembles Fig. 455. The'diagram, t4Jn, resolves itself into
two systems, viz., alloys of one pure metal and the eutedtic of that metal and
the definite compound of the two metals, and alloys of ithe 'other pure metal
and the eutectic of that metal and the definite compound.
The alloys of magnesium and tin form the definite compound of Mg 2 Sn.
In addition to the phenomena observed in the transition from the liquid
to the solid state, further changes occur in many alloys and upon which to a
great extent their usefulness depends. These changes, usually accompanied
by allotropic 13 modifications an,d consequently accompanied by changes in
physical properties, are also marked by evolutions of heat on cooling (absorptions of heat on heating) and are attended by variations in structure often
readily discernible by means of a metallographic examination. It is due to t
knowledge of the laws governing these modifications that it is possible to
regulate and control the physical properties of an alloy by means of heat
treatment.
'
1100

1000

.... ~~~d
. . . .Its

~
............. i'..
.

900

,/

1\ ~~ ~Zn3
-~

\\ 1\

.. 800
."

l!!

.~

LJ

\\'Y+~
, ,

700

\ a +11\

1!

600

.{,a + Li

fJ+'Y
I

500

"'"

'Y

,,
\

.'\

- ., - '

r\

VHf .. \+
,,--- ..... i~.

,,

400

a+'Y

"Io Z n.

% Cu.

10

20

90

80

30
70

40
60

50
50

60
40

70

30

+'1]

'Y+

300

80

90

20

10

'I]

100

FIG. 460.
13 ALLOTROPY. A change in the properties of an element withoutl,change of state.
Probably due to a change in the number or in the arrangement of the atoms in the
molecule. Best defined by the example of the three aUotropig forms of carbon, viz.,
diamond, graphite, coke (amorphous).

METALLOGRAPHY

2557

COPPER-ZINC ALLOYS
The application of the principles of metallography in connection with
only one of the commercial non-ferrous alloys will be discussed and in a general
way due to the somewhat greater importance of iron and steel in the industrial
worl~ and because the metallography of iron and steel embodies in a greater
degree all the principle~ found in the metallography of the non-ferrous alloys.
Figure 460 14 shows the equilibrium diagram of the brasses, an examination
of which will show that the shape of the liquidus and solidus curves is of the,
type indicating the formation of a solid solution, except that in this case
there are several solid solutions occurring instead of one. Below the solidus
the existence of several different constituents can be noted which are formed
after the metal has solidified and known as alpha, beta and gamma, etc.
There are, however, practically no commercial alloys containing more than
60% zinc and so it is only necessary that attention be given to the alloys
containing less than that amount.

FIG.

461. PhotOmicrograph of Brass Cold Rolled and Annealed.. Magnified 100

diame~ers. Etched with ammonia and hydrogen peroxide.

Here there are three solid solutions, alpha, beta and gamma. The color
of the alpha brasses varies from a pale yellow to a yellowish red, the color
becoming redder as the copper content increases.' These brasses have moderate
tensile strength, high ductility and are suitable for cold rolling. Figure 461 is
a photomicrograph of a section of alpha brass which has been cold rolled and
annealed. Alloys containing the beta solution are somewhat harder and when
cold are much less malleable \nd ductile. Beta brass has a greenish-red color
and is darkened more quickly by the etching reagents than the alpha variety.
Brasses containing the gamma solution being very brittel are without mechanical value and are only employed for castings of a purely ornamental character.
By regulating the amounts of copper and zinc and the temperature to
which the alloy is heated and the rate of cooling from that temperature, the
most advantageous properties of each of the constituents can be imparted
to the ;metal. Mechanical working also improves the 'physical properties.
Other metals are often added to copper-zinc alloys to make the alloy more
.suitable for the purpose for which it is to be used, e.g., iron, manganese, and
aluminum are added in small quantities and increase the tenacity of the
alloys considerably.
14

Mter Hofman.

METALLOGRAPHY

2558

IRON-CARBON ALLOYS
The equilibrium diagram of the iron-carbon alloys is sh\)wn in Fig. 462.1 5
By comparing the shape of the curves of the liquidus and sql~dus with those in
Figs. 453 and 455, it may be seen that in the region embracing the. steels, the
alloys solidify as solid solutions; and in the region containi'ng from about. 2%
to 5% carbon and in which fall the cast irons, a eutectic is formed. Is Imme-

500

9(,Co

%Fe3C

1.0
1.5'

1.7 2.0
25.5 30

3.0
45

FIG. 462.

4.0 .3
bo 64.5

5.0

7~
I

6.0

6.67

90

100

diately after solidification has taken place, the steel/? consist of a solid solution
of carbon in gamma ironP Gamma iron, then, ,vill dissolve carbon until a
total content of approximately 1.7% is reached at which point the.sQlid t3olution
becomes saturated and further increase in the amount of carbon res:D:lts in the
precipitation of the carbon in the form of the aarbide of iron (Fe 3 G) (called
cementite) which forms a eutectic with the saturated austenite. At the pqint
E the alloys solidify as 100% eutectic. With more than 4.3% carbon the
constituent in excess of that necessary to form the eutectfc is the carbide of iron ...
15

After Sauveur.

16 The division between the steel series and the cast iron series is a purely arbitrary
one, but the fact that there are very few commercial products containfng between
1.7% 'and 2.5% carbon makes it a particularly convenient one.
.
.
17 Many authorities define austenite as a solid solution of FeaC jn gamma iron.
Recent research, however, on crystal structure founded on X-ray examinations shows
that independent molecules of FeaC do not exist in austEfrl'ite._ Z. Jeffries .and R. S.
Archer, Chern. and Met. Eng., 24, 1057 (1921); 26, 249 (1922).

METALLOGRAPHY

2559

In Fig. 463 is shown on a larger scale the transformations which take

place in the steels after the metal has solidified. The shaded portions indicate
the intensity of the critical points. It may be seen that pure iron has two
critical points on cooling to atmospheric temperature, one occurring at about
900 0 C. and another at about 760 0 C. This indicates that iron exists in three
allotropic forms-alpha, beta,18 and .gamma. The point marking the division
between gamma and beta iron is known as the Ara point 19 and the point marking
the division between the beta and alpha iron is known as the Ar2 point. As
soon as any carbon is present, another point appears at about 690 0 C. (1274 F.)
know~ as the Arl point. When the carbon content reaches 0.4%, the two upper

0/0 Carbon
0/0 Fe3C

0.0

0.2

0.4

0.6

0.8
12

1.0

1.2

15

18

1.4
21

1.6
24

1.8

2.0

27

30

FIG. 463.

points merge forming a single point, Ara.2, which in turn merges at 0.85%
carbon content with. the Arl point forming the Ara.2.1 point. From 0.85%
carbon to about 2.0% carbon there are two points-the new point being
known as the Arcm point.
It may be seen that the shape of these curves closely resembles that shown
in the diagram illustrating the general type of curve representing .the formation
18 There is much controversy as to the presence of more than two allotropic forms.
Some authorities contend that beta iron does not exist. Recently another allotropic
modification has been noted as occurring at 1400 C. (2552 0 F.), denoting the presence
of delta iron.
19 The "r" in "Ar," etc. is the first letter of the French word "refroidissement."
The points obtained on heating are "Aca" etc., the "e" being the first letter of the
French word "chauffage." Due to hysteresis the transformation points occur at
somewhat IO'Yer temperatures on cooling than on heating.

2560

METALLOGRAPHY

of a eutectic. On account of its analogy to this diagram as evidenced by the

mechanism of the formation of the different constit~ents, .and the thermal
phenomena noted, the laminated structur.e formed on cooli~g 0.85% carbon
steel has been named eutectoid.

FIG. 464.
.

Photomicrograph of 0.40% Carbon Steel (Cast). Magnified 100 diameters.

Etched in a 4% alcoholicsolution of nitric aCid.
.,

The four photomicrographs shown in Figs. 464, 465, 466, and 4'67 illustrate
the structure of ~ast steel slowly cooled containing varying amounts of carbon.
The black constituent in the above photomicrograph is the eutectoid of
iron and the carbide of iron called "pearlite" because of its resemblance in
color with the naked eye to mother-of-pearl; the white constituent is made

FIG. 465.

Photomicrograph of 0.85% Carbon Steel (Cast). Magnified 100 diametern.

Etched in a 4% alcoholic solution of nitric acid.

up of the crystals of iron (in excess of the eutectoid ratio) a1)d is called "ferrite."
Pearlite is relatively strong and hard, whereas ~!'.rite is )V'eak and soft.
Figure 465 shows the structure of slowly cooled steel which is made up of
the single constituent pearlite, and Fig. 466 snows the same steel under a higher

METALLOGRAPHY

2561

magnification which brings out the characteristic laminations. In this photomicrograph the white constituent is the cementite and the dark constituent
is the ferrite. (The ferrite is dark here because of its greater solubility in
the etching reagent than'the cementite and is consequently thrown in shadow.)
Cementite is hard and strong, but very brittle.

FIG. 466.

Photomicrograph of 0.85% Carbon Steel. Magnified 500 diameters.

Etched with a 4% alcoholic solution of nitric acid.

The excess constituent is cementite-the needle-like forms in the grains of

pearlite and the white substance at the grain boundaries.

FIG. 467.
,

Photomicrograph of 1.25% Carbon Steel (Cast). Magnified 100 diameters.

Etched in a 4% alcoholic solution of nitric acid.

The carbide of iron is an unstable compound and in the case of the cast
irons at high temperatures is .readily decomposed into graphite and iron.
The result of this decomposition is that when the iron has cooled to atmospheric
temperature, part of the carbon is present as FeaC, and part as free graphite.
Certain impurities as well as the rate of cooling influence the breaking up of
the carbide, e.g., silicon and slow cooling promote the formation of graphite,
and manganese and fast cooling tend to keep the carbon in the combined
J

METALLOGRAPHY

2562

condition. Figure 468 shows the structure of cast iron anB explains why this
metal is weak, lacks ductility and cannot be forged or hammered. The
black plates of graphite break up the continuity of, the mAss rendering it soft
and weak.
.
I

FIG. 468. Magnified 100 diameters.

Etched wjth a 4% alcoholic solution of
nitric acid.

FIG. ,469. Magnified 100 diameters.

Etched, with a 4% alcoholic solution of
nitric acid.
/'

If the percentage of silicon is low and the percentage of manganese is high,

the carbon, in spite of a fast rate of cooling, will remain in the combined con~
dition and the resultant product is "white" cast iron (so called from the
appearance of its fracture as contrasted with that of "gray" cast iron).
This metal although too brittle for commercial purposes is used for the production of the so~called malleablEl castings. Castings of this material are
heated in large containers together with an oxidizing packing material to a

FIG. 470. Magnified 100 diameters.

Etched with a 4% alcoholic solution of
njtric acid.

FIG. 471. Magnified 100 diameters.

Etched with a 4% alcoholic solution of
nitric acid.

high temperature for forty~eight hours or more, ca3-~iDg the .(arbide of iron to
decompose into iron and graphite and a small amount of carbon to be removed
from the surface. The graphite particles resulting from this decomposition
due to their size and shape are less d'e trimental to the physical properties than
I

' .

2563

METALLOGRAPHY

t}J.e plates present in gray cast iron. This may readily be seen in Figs. 469
and 471. The G~!'k constituent in Fig. 469 is pearlite and the white is the carbide of iron. in Fig. 471 the black constituent is the graphite resulting from the
breaking up of the Fe 3 C and the white the ferrite.
Figure 470 shows the process only partly completed. The black areas are
graphite, the lighter areas pearlite, ,and the white areas ferrite.
INFLUENCE OF MECHANICAL WORK
The purpose of hot work 20 such as forging, hammering, etc., is to shap'ethe metal into useful articles and to decrease the size of the grains which
greatly improves the physical properties, increasing the tensile strength and
elastic limit. The diagram shown in Fig. 472 21 depicts the influence of hot
work on the structure of steeL While this is not' strictly true for alloys other
.han steel, tlie same principles are involved.

+
+
+
+
+
+
+
+

~s
Very coarse
.>tructure

Coarse
structure

Less
coarse

Fine
strl)cture

FIG. 472.

As the steel solidifies at A (or is reheated to TV after the steel has cooled)
the crystalline grains of the solid solution increase in size as the temperature
falls to the critical range. In order to reduce this growth to a minimum, work
should be stopped just above the critical range. If the work is stopped at
for 1', a larger grain growth will resulUhan if the work is stopped atf". Work
performed below the critical range when the metal is composed of an aggregate
distorts the grains into elqngated masses producing harmful internal strains.
In :Figs. 473, 474, 475 are shown the structure resulting from hot working
steels of different carbon content.
20 The term ,. Hot Work" as used here denotes work performed above the critical
range and "Cold Work" that performed below the critical range.
21 After Sauveur.

2564

METALLOGRAPHY

Cold work is performed in wire drawing, rolling sheJt steel, etc. It increases the elastic limit, tensile strength and
and decreases the
ductility as represented by the elongation and reduction

FIG. 473. Photomicrograph Showing

Structure of Hot Worked .030% Carbon
Steel. Magnified 100 diameters. Etched
with a 4% alcoholic solution of nitric acid.
Pearlite dark; ferrite white.

FIG. 474. Photomicrograph Showing

Structure of Hot Worked Eutectoid Steel.
Magnified 100 diameters. Etched with a '
4% alcoholic solution of nitric acid.

The permanent deformation of metals results in the occurrence of certain

characteristic changes in structure some of which are discernible under the
microscope. The most important of these are: (1) Slip bands; 22 (2) Neumann
lines; and so-called critical or exaggerated grain growth.23

FIG. 475. Photomicrograph Showing

the Structure of Hot Worked 1.25 % Carbon
Steel. Magnified 100 diameters. Etched
with a 4% alcoholic solution of nitric
acid. Pearlite dark; cementite white.

FIG. 476. Photomicrograph Showing

the Structure Produced by Cold Work in a
Steel Con$ining about , 0.20% Carbon
Magnified 100 Diameters.. Etched with
a 4% alcoholic solution. of nitric acid.
Pearlite dark; ferrite light.

Slip Bands.- Deformation under a slowly applied load is accomplished by

a process of slip, i,e'., the crystalline fragments of-the grains slide on each other
in much the same manner as a pile of books evenly stacked will slide on
22
23

Z. Jeffries and R. S. Archer, Chem. and Met. Eng., Vol. 27, No. 18, p .. 882.
Sauveur, "Met. of Iron and Steel," p. 265.

METALLOGRAPHY

2565

one another if a force is exerted which disturbs them. The slip bands appear
on polished and etched surfaces after the metal has been deformed as small
hair lines (actually differences in level) for the most part running in one direction in the same grain, but in different directions in different grains. In
some metals such as brass (see Fig. 461) annealing after cold work results in the

FIG. 477. Photomicrograph Showing Small Strip of Steel Slightly Bent after Polishing and Etching. Slip bands appear as black lines in those grains where the strain
was greatest. Magnified 100 diameters. Etched with a 4% alcoholic solution of
nitric acid.

formation of twins produced by part of the crystal having rotated through an

angle of 180 0
I
Neumann Lines.-The appearance of these under the microscope is much
the same as slip bands. They are, however, still present after repolishing

FIG. 478. Magnified 100 diameters.

Etched with a 4% alcoholic solution of
nitric acid.

FIG. 479. Magnified 100 diameters.

Etched with a 4% alcoholic solution of
nitric acid.

and etching whereas slip bands disappear. Their exact nature is not known
but they are probably caused by a suddenly produced deformation, which
mayor may not have been accompanied by fracture.

!566

METALLOGRAPHY

Critical or Exaggerated Grain Growth.-This condItion bccurs in low

carbon steel (.08% to 0.12%) strained by cold wor.k corresponding to a 5 to
15% reduction and . subsequently annealed at a temRerature! varying from
625 0 C. to 775~. Steels composed of very large grains resulting from this
treatment give poor results under both shock and. fatigue stre~ses. Fig. 478
shows a low carbon steel in the normal condition and Fig. 479 after critical
.
grain growth has taken place.

HEAT TREATMENT OF STEEL

Heat treatment as defined by Tiemann is "the change, or series of changes,
in temperature, also the rate of cooling from one temperature to another
brought about to secure certain desired conditions or properties in a metal
or alloy." Heat treatment, in general, may be conveniently divided into
three classes: (1) softening treatment; (2) hardening treatment; and (3)
strengthening treatment.
There are three structural transformation changes that take place in steel
as the metal cools through the critical range, each of which ,has definite
physical, properties and characteristic appearanqe under the microscope.
Either one of these constituents or any combination of them may be retained
in the metal when cold by cooling the steel through the critical range at
different speeds, or by hardening followed by tempering. The names of the
transformation constituents are martensite, troostite, and sorbite.

FIG. 480. Photomicrograph of 1.25%

Carbon Stefl: Quenched in Water from
Above the Critical Range. Magnified 100
diameters. Etched with a 4% alcoholic
solution of nitric acid. Martensite light;
troostite dark.

FIG. 481. . Photomicrograph of 0.75%

Carbon Steel Quenched in Oil freim Above
Its Critical Range and Reheated to 525 0
C. Magnified 100 diameters. Etched in
a 4% alcoholic solution of nitric acid.
I

In order to pr{)duce maximum softness a very slow rate of cooling through

the ~ange is used which produces pearlite. To produce maximum hardness,
the steel should be in the martensitic condition. This is obtain7d by cooling
through the range at a fast rate, i.e., by quenching i]l::"water or oil. Such
cooijng should, theoretically, prevent any transformii'ion from taking place,
but in plain carbon steels partial transformation always occurs. If austenite
is desired, certain other metals must be added which act as retarding agents
on the transformation.

METALLOGRAPHY

2567

Steel containing 12% manganese or 25% nickel, for example, will remain
austenitic even after slow cooling. Austenite is mineralogically softer than
martensite, but is more resistant to abrasion .. Under the microscope it has
the general characteristics of .a solid solution. Troostite is often present
along with martensite in hardened steel and possesses similar physical properties, but in a lesser degree. Figure 480 shows the appearance of martensite
and troostite.
Maximum strength combined with maximum ductility is obtained by
producing the constituent sotbite. This can be done by cooling through the
range at the proper speed, but more exactly by hardening followed by heating
to a temperature between 400. and 600 0 C. according to the composition
the steel and the properties desired. Tro(')stite may also be formed in such a
llJII!flner if a lower reheating or "drawing" temperature is used. The appearance of sorbite when examined under the microscope is shown in Fig. 481.

of

CASE HARDENING
Case hardening, as the name implies, is the production of a hard case or
shell on the, surface of a soft steel.in order to make it more resistant to wear
witho~t materially reducing its toughness and resistance to shock. Case
haljdening is made possible by the ability of iron when in the gamma condition
to absorb car-bon. The article to be so treated is packed in a suitable carbomiceous material and heated to a high temperature (about 950 0 C.) for a time
depending upon the depth of case desired ..

FIG. 482. Photomicrograph Showing the Edge of a Case-hardened Sampie of Steel.

Magnified 100 diameters. Etched with a 4% alcoholic solution of nitric acid.
The core and case may then be heat treated to further improve the physical
properties. Oase-h!1rdened objects appear under the microscope as in Fig. 482.

2568

METALLOGRAPHY

INFLUENCE OF SOME ELEMENTS ON IRON lAND STE:E:L

Manganese has a greater affinity for sulfur than . ironl and, therefore, if
sulfur is present, manganese sulfide. is formed. lJnder the microscope this
appears in castings as small round areas, pale gray in coI6r,' and in forgings
as elongated streak~ of the Bame ~l)ll)r. If there is any mangane'ile remaining
after the sulfide has formed, the ~arbide I)f manganese (MnaC) is produced,
which forms a solid solution with the iron.
Sulfur, when present in sll).all amounts (e.g., less than 0.08%), combines
wjth the manganese as explained, but if in excess of this amount, the sulfide
of iron (FeS) is produced which forms around the grains. Steel containing
iron sulfide is weak and brittle at high temperatures. FeS appears a reddish
brown under the microscope.
Phosphorus _is present in solid solution as the phosphide of iron (Fea P ).
In cast irons, due to the large amount of carbon present, a eutectic of iron and
the phosphide of iron is formed. Iron phosphide makes the steel brittle,
especially when cold.

FIG. 483. Photomicrograph Showing Appearance of -Manganese Sulfide _(Small

Round Areas Pale Gray in Color) in Low Carbon Cast Steel. Magnified 100 diameters.
,
Etched with a 4% alcoholic solution of nitric acid.

Silicon is present as FeSi, ',\'hich enters into solid solution ,with the iron.
When in small am.ounts, silicon produces no marked influence on the-physical
properties.

METALLOGRAPHY

2569

Other elements such '3 nickel, chromium, vanadium, tungsten,' and

molybdenum are added to steel usually in amounts less than 5.0% which

.FIG. 484. Photomicrograph Showing

Appearance of Manganese Sulfide in
Forged Low Carbon Steel (over 1.0%
Sulfur). Pale gray colored areas drawn
out in the direction of rolling. Magnified
100 diameters. Etched with a 4% alco~
holic solution of nitric acid.

FIG.

FIG. 485. . Photomicrograph Showing

Structure of Cast Manganese Steel (Man-
ganese 12%) Austenitic. Small spots are
undissolved carbides. Magnified 100 diam~ters. Etched with a 1()% alcoholic
sollltion of nitric acid.

486. Photomicrograph Showing Structure of Cast High Speed Steel.

Magnified 500 diameters. Etched with Yatsevitch's Reagent.

impart to the metal specially desired properties such as increase in hardness,

without decrease in ductility, machine tools capable of cutting at a high speed,
and for greater shock and fatigue-resisting properties.

METHODS FOR THE DETERMIN:A~ION OF

, SOLUBiLITY 1
I
A quantitative determination of a solubility consists essentially of two
operations; the preparation of the saturated solution and its subsequent
analysis. In those cases where thesa steps are p~rformed separately the
method may, in general, be designated as the analytical and in those where
they are combined, as the synthetic. In both cases, however, the consideration'
of first importance is the assurance that final equilibrium between solvent and
solute has been reached. Since this point is that at which no further change
occurs in the relation between the amount of the compound in solution and
that remaining undissolv,ed, the only criterion of saturation is the evidence'
that the concentration of the solution has not changed during a longer or
shorter interval of time, during which those conditions which \\"'ould tend to
promote such a change have been allowed to operate.
Of the conditions which promote most effectively the attainment of equilibrium between a solute and a solvent, the provision for the intimate contact
of the two is most important. In other words, only by the' thorough mixing
which agitation or effective stirring provides can the point of saturation be
reached with certainty. In the case of the reciprocal solubility of liquids, the
point of equilibrium is usually attained within a much shorter period than in
the case of solids .dissolved in liquids. In the later case, the necessary disintegration of the solid, incident to its'solution in the liijuid, is a process wh~ch
is restricted to the surface layers of the solid, and, t~refore, unless a large
area, such as a finely divided state provides, is available, and unless that portion
of the solvent which has acted upon a given surface area is repeatedly replaced
by fresh solvent, the process of solution will be greatly/retarded. It is quite
evident that, although a solution in contact with ehn very finely divided solid
may promptly become saturated in the immediate vicinity ofj the solid without
stirring, the distribution of the dissolved material to the remainder of the
s9lvent would depend upon diffusion, and since the rat~' at \,>:hich this proceeds
would diminish as the concel]tration differences became eciualized~he process
would take place at a gradually diminishing rate. If the point of equilibrium is
approached from supersaturation, the above remarks apply with equal effect,
since only at the surface of the solid can the excess of salt leave the solution
and, without other provision than diffusion for successively bringing the entire
amount of the solution in contact with the solid, the deposition of the excess of
dissolved material can occur only at a very slow rate. The importance of
activ,( and continuous agitation of the solid and solution, in effecting saturation, cannot, therefore, be ,too strongly emphasized. It may in fact be assumed
that determination;; of the solubility of solids, made without continuous agitation, are always open to the suspicion that the results do nClt represent the
! final equilibrium which such data are required to s~Q.)V'..:~.
'
1

Reprinte~ from 2d Ed. of "Solubilities."

2570

Contributed by Atherton Seidell.

DETERMINATION OF SOLUBILITY

257~

Since solubility is a function of temperp.ture, the accurate control of the

temperature in making a solubility determination is another one of the indispensible requisites of accuracy. In general, it may be stated therefore, that
every pracedure designed for preparing a saturated solution must inclUde
provision for -the accurate control of the temperature and for active and continuous agitation or stirring of the solution. In the case of the solubility of
gases, which will be considered in a separate section, provision for the control
of the pressure must also be made.
It is obvious that since the solubilities of various compounds differ, and
that'oof one compound is affected by the presence of another, the accurate
determination of this constant for a particular molecular species presupposes
that only this one substance is present in the pure solvent. That is, accuracy
of results demand that only pure compounds be involved in a given determination, consequently, no effort should. be spared to make it certain that the
highest possible purity of both solute and solvent has been attained.
, Apparatus for the Determination of the Solubility of Solids by the Analytical
Method.-The types of apparatus which have been developed for the preparation of saturated solutions of solids in liquids differ principally in respect to
wb.etb.e! de",igned l~! ll\\lltiple O! ",ingle detennin~tion'" ~t ~ given tempe!atme.
Examples of the first type are illustrated by Figs. 487 and 488.

FIG. 487.

Determination of Solubility of Solids.

It will be noted that in one case (Fig. 487)' the bottles containing the solu~
tions are stationary and the liquid in each and in the constant temperature
bath is kept in motion by means of revolving stirrers. This form of apparatus
was used by Moody and Leyson (1908) for the determination of the solubility
of lime in water and is particularly adapted for relatively slightly soluble COm-

DETERMINATION OF SOLUBILITY

2572

pounds for which rather large quantities of the' satura~ed solution are needed
for accurate analysis. There is also shown in the figure~ the provision for withdrawing the saturated solution through a filter within th,e inverted ,thistle tube.
The stirrers in the bottles are fitted with mercury seals to prevent access of
air containing carbon dioxide. Other features of the a:pparatus will be readily
understood from the drawing.
I

FIG. 488.

Noyes Apparatus for Determining the Solubility of Solids.

/ .

A more common type of apparatus, designed for the simultaneous saturation

of several solutions at the same temperature, is that illustrated by Fig. 488,
in which the bottles containing the solutions are slowly rotated in the constant
temperature bath. The form shown is that described by Noyes (1892).
This type of apparatus h!}s the advantage that the solid is,. to a large extent,
kept in suspension in the liquid and, therefore, offers the most ra-vorable opportunity forcontinuous and uniform contact with the solution. Many examples
of this form of apparatus, differing prin-cipally in size and in the direction of
movement of the containers, are described in the literature.
Of the second type of apparatus, designed for a single determination at a
given temperature, many varieties have been developed for particular conditions. Of these, the following examples have been selected as typical of this
class and, it is hoped, will illustrate most of their desirable features. They
are, in general, adaptations of earlier designs and it is not intended that the
name given in connection with each is that of the investigator who deserves
the credit for originating the type. The drawings~ill, for the most part, be
readily understood without detailed explanatiolls.- The dimensions are not
stated, since they can usually be varied to suit the needs of almost any problem.

DETERMINATION OF SOLUBILITY

2573

In Fig.A89 is shown the apparatus used by the Earl of Berkeley (1904) for
the very careful determinations of the solubility of inorganic salts in water.
The features of particular interest in connection with it are, that the water
bath itself is made to serve as the temperature regulating device, and the
apparatus for withdrawing and simultaneously filtering the saturated solution
is a combination of pipet and pycnometer. This was provided with ground
glass caps for each end and the stem was accurately graduated. It was, of
course, carefully standardized before use. The flexible iron plate shown was
made of a disc from the receiver of a telephone. The apparatus was used for

Flexible

Solution

('==r==J::=-=t~~II::
Gas

Inlet

Salt

~]r
FIG..489.

Berkeley's Apparatus for Determining Solubility.

determinations at temperatures between 30 and 90 and the range of variations from the set temperature of the bath was, for 2-3 hour periods, within
about 0.2. For the inner vessel containing the salt, the range was about
0.05. At each temperature two determinations of density and solubility were
made; one on the solution obtained by stirring a supersaturated solution in
contact with solid salt, and the other on the solution obtained by stirring an
unsaturated solution in contact with an excess of salt.
In the case of determinations at the boiling point a special apparatus was
required. Two forms, described by the Earl of Berkeley (1904), are shown in

2574

DETERMINATION OF SOLUBILITY

Figs. 490 and 491. The first was used for the less soluble salts and consisted of
an outer tube A containing water and an inner tube B'containing salt and solution: By boiling the water vigorously and closing the .side tube C, steam
passing through the tube '[) stirred the solution thorollghly arl.d the temperature
rose to the boiling point of the saturated solution and remaihed C'Onstant when
saturation was attained. The second form of apparatus (Fig. 491) was devised

Thermometers

/
Filter
Salt
Stirrer

FIG. 490.
FIG. 491..
Berkeley's Apparatus for Determining Solubility.
!

for use with extremely soluble salts. In these cases it was found that the
larger quantity of steam required for thorough stirring dissolved so much salt
that it was necessary to have a very large excess present. In this apparatus
the steam was generated in a boiler- A and conducted through $he tube B to
the bottom of the large test tube C containing the excess- of salt and solution.
The test tube was immersed in the oil bath D which wn:s-:Vigorously stirred and
maintained ,at a temperature close to that of the boiling point of. the saturated

DETERMINATION OF SOLUBILITY
~

2575

solution. When the temperature of the oil bath was below the boiling point,
salt dissolved; when above, salt was thrown out of solution. Considerable
difficulty was experienced in filling the pycnometer with the saturated solution
without introducing errors due to steam bubbles causea by the suction which
was applied.
K

FIG. 492.

FIG. 493.

Victor Meyer Type of Apparatus for Solubility Determination.

Walton and Judd Apparatus for

Solubilities.

A comparatively simple form of the type of apparatus used by Victqr

Meyer in. 18.75 and modified by Reicher and van Deventer (1890) and by
Goldschmidt (1895), is described by Hicks (1915) and shown in. the accom-

DETERMINATION OF SOLUBILITY

2576

panying Fig. 492. A glass cylinder A is closed at each- end jwith large one-DOle
rubber stoppers. The mixture of salt and solution is contained In this cylinder
and is stirred by the rotation of the tube E which is provided with an enlargement at its lower end in which there are two small holes at H and I. The
stirrer rotates in the bearing formed by the hollow woode!l cylinder J. The
glass rod K carries a rubber stopper L which closes the filtering tube M, in
which a platinum cone N supports an asbestos filter O. The siphon P connects the filtering tube with the flask R which is provided with an outlet through
the small tube S. The apparatus is immersed in a constant temperature water
bath W, to about the level shown. After stirring the mixture of salt and

FIG. 494.

Donnan and White Apparatus for DeterminingSolubilities.

solution a sufficient length of time for attainment of saturation, the-undissolved

salt is allowed to settle and the rubber stopper is withdrawn from the filter tube
by means of the glass rod K. Suction is applied through the tube S to hasten
the filtering and the clear solution collected, at the temperature of the bath, in
the previously weighed flask R.
'
A similar apparatus was used by Walton and Judd (1911), for determination
of the solubility of lead nitrate in pyridine. This is shown in Fig. 493 and
consists of a glass test tube fitted with a stirrer which turns iI) a mercury seal,
thus preventing loss of solvent by evaporation or the a,dmission of moisture from
the air. To take a sample of the saturated solutio'i,-the weighing tube A was
introduced into the larger tube through a hole in the stopper. After reaching

DETERMINATION OF SOLUBILITY

2577

.the temperature of the bath the stirrer was stopped, the end of the small tube B,
which lYas covered with a piece of closely-woven musHn, was dipped below the
surfac~ of'the solution and the liquid drawn into A l)y applying suction at C.
The tube A was then removed, weighed and the contents analyzed.
An apparatus which was used by Donnan and W}lite (1911), for the determination of equilibrium in the system palmitic acid and sodium palmitate is
shown in Fig. 494. The stirring in this case was accomplished by means of a
current of dry air, free of carbon dioxide. The appar!l-tus consists of two parts,
namely, an ip.ner chamber E, where equilibrium walS attained, and an outer
case A, designed for isothermal filtration. The whole was immersed in a
thermostat to the level W. A side tube B permitte(i connection with a filter
pump. C is a weighing bottle to receive the filtered saturated solution and D
a Gooch crucible provided with a paper filter. The cork, closing A, was covered with a plastic layer to render it airtight. The tllbe at the lower end of E
was closed with a ground glass plug F, the stem of which was enlarged to a
small bulb at G and then drawn out to ,pass easily through H, leaving an air
free outlet' around it. The small cork I was used to &upport the stopper when
lifted to allow the contents of E to flow down for filtration. The dry air .by
which the mixture was stirred was drawn through K 1)y applying suction ~t H.
'tb..~ \lr~b..~atit\.'b Qf_ tb..i" air wa" a~el)m\llis.b..~d b-s drawin'b it thrullJbQ a.. thin.
spiral immersed in the thermostat. The connection between the equilibrium
apparatus and preheater was made through a mercury seal, which permitted
lifting the apparatus easily without damage to the fragile preheater permanently
mounted in the bath. This apparatus provided for the recovery, separately, of
the saturated solution and undissolved, solid. These authors also des.cribe an
improved l!1ectrically heated and controlled constant temperature bath.
Determinations' at lower temperatures that can l>e constantly maintained
with the aid of a water bath require special forms of apparatus which permit
of temperature control under more or less restricted conditions. ' An apparatus .
of this type, which was used by Cohen and Inouye (1910), for determination
of the solubility of phosphorus in carbon disulfide, is shown in Fig. 495, and is
intended for the range of temperature between -10" and 10. The saturating vessel D consists of a glass cylinder to the upper end of wh~ch is cemented
a steel collar E, containing a deep cha9nel. A mixture of litharge and glycerol
was used as the cementing material for this purpose. The inverted steel
cover F fits into the channel of this collar and the seltl of the joint is effected,
in the usual way, by means of a layer of mercury. The cover F is provided
with a brass tube K, to which the pulley M is attached, and is also pierced by
the tightly cemented-in glass tube I. The glass rod (J, containing on its lower
end the three stirring wings H H H, is cemented into the brass tube K. The
saturating vessel is, for stability, tightly fastened in It hole in a block of lead,
S, contained in the Dewar cylinder A. An atmosphere of CO 2 in the saturating
vesse~ is provided by introducing CO 2 under pressure through I and allowing
the excess to escape through the mercury seal in E. After charging the'
apparatus, I is closed with a rubber tube and plug and the stirrers H H H set
in motion. A Witt stirrer, 0, keeps the contents of the bath in rapid circulation, Water is used in the bath for temperatures about 0, and alcohol for
those below 0. The regulation of the'temperature is Itccomplished by addition

2578

DETERMINATION OF SOLUBILITY

of ice or solid CO 2 as found necessary and, therefore, requjres very close attention on the part of the experimenter.
A novel and simple form of apparatus, which was u~ed by Ba,hr (1911),
for the determination of the solubility of thallium hydroxide at temperatures
up to 40 is shown in Fig. 496. As will be seen, this consist.s of a gas washing
flask to the arms of which a Y tube provided with two stop-cocks is sealed:"
The .inside walls of the apparatus were coated with hard paraffin and the

FIG. 495.

Cohen and Inouye Apparatus

for Solubilities.

FIG. 496. Bahr's Apparatus for De'termining Solubilities.

required amounts of thallium hydroxide and water introduced:'It was then

immersed in a water bath and the contents stirred by means of-a current of
hydrogen, which entered as shown and \vith A and E closed, passed through
D aIid out at B. When it was desired to remove a sample of the solution for
analysis, Band D were closed and the liquid forced through A into the pycnoFor temperatures above
meter by means of gas pressure entering through
40, the form of apparatus shown in Fig. 497 was used. In this case K represents a copper cylinder with double walls, of which the inner comp(1rtment G,
contains concentrated salt solution which is stirred by ,a stream of air (not
shown), and the outer compartment contains a laye~ of heating liquid H. The
glass tube L contains the mixture of thallium hdroxide. and water which is
stirred by means of a current of hydrogen (not shown). When saturation is
attained the tube AI of small bore and thick walls and provided with a small

Ef

DETERMINATION OF SOLUBIL1TY

2579

asbestos filter, is introduced and the saturated solution forced over into the
receptacle B by pressure of hydrogen which enters at C. The heating liquid
in B is the same as used in H. The following he~ting liquids with the boiling
points shown were used: Allyl chloride, 46; Ethylene chloride, 55; Chloroform, 61; Methyl alcohol, 66; Benzene,
80; Benzene-Toluene mixture, 91;
Water, 100.
A somewhat more elaborate apparatus, in which the constant temperature
is maintained by means of the vapor of
a boiling liquid, is shown in Fig. 498.
This apparatus was developed by Tyrer
(1910) for the very accurate determination of the solubilities of anthraquinone, anthracene and phenanthra-quinone in single and mixed organic solvents.
The solvent with excess .of the solute
was placed in A and kept in constant
agitation by means of the vertically
acting stirrer shown. The tube A is
surrounded by a bath of vapor which
B
circulates through the cylinder B, condenses in C, and returns to the boiling
flask M. When the solution is saturated it is allowed to settle, and 'the
clear solution run out (by raising the
tube D) into a small graduated flask E,
H
which is maintained at the same temperature as the solution A. The temperature of the vapor bath is varied by
changing the pressure under which the
FIG. 497.
liquid in the flask M is boiling. For
this purpose, the manostat P is provided. The temperature can, with care, be maintained constant to O,O1~. For
this purpose the apparatus must be airtight, the liquid in the boiling flask
must not bump (which is entirely prevented by placing a layer of mercury in
the flask) and a pure boiling liquid must be used.
Although illustrations of special forms of apparatus designed for securing
equilibrium in solubility determinations could be extended far beyond the
number given, it is believed that the principal features have been made clear
and it will no doubt be possible to adapt the devices here shown to ma~y other
cases for which accurate determinations of sohibility may be desired.
Separation of Saturated Solution from Undissolved Solid.-The next point,
after the establishment of equilibrium between the solvent and solution, is the
matter of successfully separating the saturated solution from the undissolved.
solid, preparatory to its analysis. There are, undoubtedly, many cases where
this is a very serious problem. This is especially so for extremely soluble compounds, which yield viscous solutions as well as for those which do not readily
settle out of the solution or cannot be removed by ordinary filtration. It is,

2580

DETERMINATION OF SOLUBILITY

of course, necessltry to maintain the mixture at-the temperature at which satqration -was obtained until tbe complete sepa;rat,-,:m ot tbe ~olution and solid has
been effected. The operation should, therefore, as a g~neral thing,\ be conducted in the same bath used for preparing the saturated solution. Several
forms of apparatus designed for this purpose are shown iIi the diagrams given
in the pteceding pages. Fat solutionB which can be readily separated from the
undissolved solid, a graduated pipet to which a stem with a plug Ii filtering
;naterial can be attached and which is adapted to being easily weighed, is the
most convenient.
To Pump

R1
",,

I I
I

~I tI
I I

\I \I
11

: ;

: I

I,

~...

\ .....~:::.,~,.

FIG. 498.

Tyrer's Apparatus for Solubilities.

Analysis of the Saturated Solution.-The weight of a known volume of the

perfectly clear solution, that is, its specific gravity, should always be determined. This weighed quantity of solution, or a known dilution of it, furnishes
11 very convenient sample for the determination of the amount of dissolved
compound.
~:_ In regard to the analysis, the procedure must be 'selected entirely on the
basis of the number and character of the constituents present. In cases of the

DETERMINATION OF SOLUBILITY

2581

solubility of single non-volatile compounds, in solvents which can be more or

less easily removed by volatilization, the plan in most general use is the vaporation of a known amount of the solution to dryness and weighing the residue.
Special forms of apparatus to be used for this purpose have been proposed
from time to time. These are, usually, vessels with tubular openings, arranged
so that a current of dry air can be drawn over the surface of the heated sample.
In the case of solubility determinations in which the saturated solution
contains more than one dissolved compound, the application of the usual
gravimetric or volumetric procedures will, of course, be necessary.
In certain cases, where the direct determination of the amount of the dissolved compound present in the solution would be very difficult or impossible,
an indirect method can sometimes be used. For this purpose, a carefully
weighed amount of the compound must be used, and, after the period of saturation, the undissolved residue is filtered off under conditions which reduce
losses to a minimum and, after drying to its original condition, it is weighed,
and the amount which has been dissolved found by subtracting the weight of
the undissolved residue from the quantity originally present.
Identification of the Solid Phase.-The solubility of a compound, which is
capable of existing in several forms, depends upon the particular form in which
it is present in contact with the saturated solution. The question of the composition of the solid phase is, therefore, of considerable importance for the
accurate determination of solubility. Although the identification of the solid
phase presents little difficulty in the majority of cases, it sometimes happens
that it can be made only by a more or less indirect method. The principal
re'ason for this is that adhering solution can usually not be completely removed
from the solid phase and the analysis, consequently, does not give direct information of the required accuracy.
A method which has been used considerably for identifying the solid phase
is that known as the residue method of Schreinemakers (1893). It is based
on the principal that if an analysis is made of both the saturated solution
and of a mixture of the saturated solution and the solid phase of unknown
composition, the two points so obtained, when plotted on a coordinate system,
lie on a line connecting the point representing the composition of the solid
phase and the solubility curve of the system. Similar analyses of another
saturated solution of the system and of its mixture with the solid phase, locate
another such line. Since all lines so determined when extended, pass through
the point representing the composition of the solid phase, their intersection
locates this point definitely.
Although the original description of this method by Schreinemakers was
illustrated by an example drawn on the rectangular system of coordinates, it
has been used much more extensively, in a practical way, in connection with
the later developed equilateral triangular diagram. In this case, each apex of
the triangle represents one of the three components of the system" each point
on a leg, a mixture of two, and each point within the triangle a mixture of all
three components. When a number of saturated solutions are analyzed, the
results correspond to points on the solubility curve of the system. If now some
of the solid phase with adhering solution is removed from each mixture and
analyzed, it is evident that the results thus obtained, being for samples made
up of both the saturated solution a~d the solid phase, give points which lie on

2582

DETERMINATION OF SOLUBILITY
,

lines connecting the two. Tile points on th{l curve for the pure saturated
solutions being known, it is necessary only to connect them with the points
for the corresponding mixtures of solid phase and saturated solution, and to'
prolong the lines to their common intersection. This "iill necessarily be at
the point representing the composition of the pure solid phase.
In applying the residue method of Schreinemakers, if the interse'cting lines
which fix the point corresponding to the solid phase meet at a very narrow angle,
definite information as to its composition may not be secured'. For cases
such as these, a procedure to which the name" tell-tale", method was given by
Kenrick (1908) and which is described in detail by Cameron and Bell (1910),
has been developed. This method consists in adding to the mixture a small
amount of an entirely different compound which remains wholly in the solVtion.
After equilibrium has been reached, a portion of the saturated solution and of
the solid phase with adhering solution are analyzed, and th:e quantity of the
added "tell-tale" compound in each determined. From. the result, showing
the concentration of the added compound in the saturated solution, and the
amount of it found in the mixture of solid and solution, the quantity of solution
in contact with the solid can be calculated. Since the composition of the
solution is also known, the difference between the composition of the solid
plus solution and of the amount of solution known to be present, is the composition of the pure solid.
Transition Temperatures can frequently be accurately determined by relatively simple means, and since such data are useful in establishing fixed points
on solubility curves they are valuable adjuncts to directly determined solubility
data.
Synthetic Method.-The procedures which have, so far, been mentioned
are all classed as analytical methods 9f solubility, determination. In contra,
distinction to these is the equally useful reverse process, by which the solvent
and solute are brought together in previously measured quantities and the
temperature ascertained at which the solution is saturated. To this procedure the designation synthetic method of solubility detFrmination has been
applied. One of the earliest investigators to use ihis method extensively was
Alexejeff (1886) and it is, therefore, frequently referred to as the Alexejeff
,.
.
. synthetic method of solubility determination.
The synthetic method can, of course, be used 'both for .the solubility 6f
solids in liquids and for liquids in liquids, but it is in the latter case that it is
of greatest service. Its points of superiority, particularly in'the-CJ!:.se of the
reciprocal solubility of liquids, are that the upper limits of the determinations
can be extended far beyond the boiling point temperature and are; in fact,
limited only by the resistance .of' the glass to pressure or to the action of the
liquid. Only small quantities' of the solute and solvent are required for a
determination. It is applicable to compounds for which quantitative methods
of analysis are not available or are of a tedious character! The mixtures, being
contained in sealed tubes, are not subject to the action' of constituents of the
air, nor are losses, due to volatilization, to be feared. Although, in the case
of solids, difficulties incident to the supersaturation, resulting from failure of
the crystals t.9 separate on cooling, are encountered, wit~ liquids the point of
saturatidn is made instantly and strikingly evident by_tIle -beginning of opalescence or clouding which occurs, and errors due to supersaturation are rarely

DETERMINATION OF SOLUBILITY

2583

encountered. A su'fe criterion that supersaturation does not occur rests on the
observation of the temperature at which the cloudy solution again clears. If
this temperature coincides with the temperature of the beginning of opalescence,
it is certain that supersaturation has not occurred. The observation of the
temperature of saturation can be repeated as often as desired, and the accuracy
of the determination is ordinarily limited only by the care taken in making it.
The limitations of the method, aside from the supersaturation which may
occur in the case of solids, are principally those resulting from the low temperature coefficients of solubility possessed by certain compounds, and which
usually occur in the vicinity of maxima or minima of solubility curves. Although a "critical clouding" occurs in the vicinity of the so-called critigal
solution point, this possesses a characteristic appearance which is easily distinguishable from the clouding observed at the saturation point, and errors
of observation due to it are not to be apprehended. In fact, it has been pointed
out that supersaturation disappears at the critical point, and the synthetic
method is ordinarily very accurate in the vicinity of the critical solution
temperature.

Since, by the synthetic method the results are necessarily obtained under
different pressures, this question has been given consideration from the theoretical and the practical side. Although it is possible that extremely high
pressures would exert an influence, the conclusion appears justified that under
ordinary conditions, in which pressures of 10 atmospheres are not exceeded,
no notable effect would be produced. The solubility curves obtained by this
method do not show any abnormalities due to this cause.
In the case of the determination of the solubility of solids by the synthetic
method, the operation consists in preparing a mixture of a carefully determined amount of the solvent 'and of the 'solid, and subjecting it to gradually
increasing temperature and to constant agitation, while a continual observation of the changes taking place in the solid is made. When all but a few
small crystals have dissolved, the change in temperature is regulated much
more carefully and note is taken of the point at which th~ edges of these final
crystals begin to change from sharp to rounded, or vice versa, or where the
sizes of the particles visibly increase or diminish. Care must, of course, be
taken not to allow the last portions of the solid to dissolve; otherwise, on
cooling, considerable supersaturation may occur before the solid begins to
separate from solution. The method is, naturally, most serviceable where the
change in solubility with temperature is considerable, and where convenient
methods for the direct analysis of the solution are not available.
The procedure of a determination in the case of the reciprocal solubility of
liquids consists in introducing by means of capillary funnels weighed amounts
of the two liquids into small glass tubes and sealing the ends. The amount
of air space in the tubes should be kept low. Many convenient devices for
weighing and introducing the liquids have been described. In the case of very
volatile liquids it may be necessary to introduce them in thin walled bulbs,
which can be broken after the tube containing the mixture has been sealed.
The tube is then placed in a large beaker of water, or higher bOiling liquid if
necessary, and heat applied until the contents of the tube, on being shaken,
become homogeneous. The temperature is then allowed to fall very slowly
and an observation made, while the tube is constantly agitated, of the tempera-

2584

DETERMINATION OF SOLUBILITY

ture of first appearance of opalescence. This. obs~rvation can be repeated as

many times as desired and the temperatures of aPl:>earanpe and disappearance
of the clouding, which usually differ by only a few tenths of a'degree, can be
ascertained with certainty.
Since, by the synthetic method the data ar~ for .Aregular intervals ~f
temperature, in order to obtain results for a particlliar teI]1perature it is necessary to plot the several determinations on coordinate paper and from the
solubility curve so obtained, read the value for th.e temperature in question.
Freezing-point Method.-A modification of the synthetic method, which is
applicable particularly to solutions which contain relatively large amounts of
the dissolved compound, is that which consists in a determination of the freez~
ing-point of the mixture. This point is, in fact, the temperature at which the
separating solid compound isin equilibrium with,the solution.
The difference between the freezing-point dete:rmination and the observation of the point of growth or diminution of a crystal in a liquid is that, in the
former, the establishment of equilibrium is. rec()gnized exclusively by the
change of the thermometer. The solution is cooled gradually, during which
the thermometer sinks slowly to' a point below tht\ freezing temperature .. As
soon as the first crystal appears, either spontaneottsly or by intentional introduction (seeding), the thermometer rises suddenly to the freezing-point and
remains stationary (or some time.
This method can, of course, be used in a lar~e number .of cases for the
determination of solubility. Those portions of the solubility curves of salts
in water for which ice is the solid phase, are pract~cally always determined in
this way and it may be said, in general, <that for c1eterminations made at low
temperatures, the freezing-point method is to be tselected whenever possible.
For the practical execution of the method the "very well known apparatus
of Beckmann is most convenient and satisfactory. The determinations must,
of course, be made with all the refinements whiCh have been developed for
accurate freezing-point measurements.
The method has been used extensively for the discovery of addition compounds. Its use for this purpose is based upon th\l: princ'iple that if to.a pure
compound, A, a second, B, is added, the freezing-bbint of A is lowered; similarly the freezing-point of B is lowered by A, and the two descending curves
thus obtained intersect at the eutectic. If, how{~ver, a compound, A",BlI is
formed, this also acts as a pure substance and its freezing-point is lowered by
either A or B. Hence the freezing-point lines do n.ot meet at luingi_e eutectic
but exhibit in this case a maximum, the position {)f which indicates the composition of the compound.
Volume Change Method.-Still another methoc1, which is a modification of
the synthetic, is that designed to indicate the recipJoocal solubility of liquids by
a determination of the volume changes which OCCUr when two relatively sparingly miscible liquids are shaken together in a closed ~essel. The apparatus
consists usually of a cylindrical receptacle which is provided with a constricted
graduated section either at one end or near the middle. Such volumes of
liquids are chosen that the meniscus separating them lies in the constricted
graduated tube. The determination consists iT! sup~rimposihg measured
volumes of each liquid and noting the position Qf the-"ffieniscus before and
after a period of shaking at constant temperature. From the increase or

DETERMINATION OF SOLUBILITY

or

2585

decrease of volume
the two layers, as estimated from the change in position
of the meniscus, the reciprocal solubility of the two liquids is calculated. It is
to be noted, however, that the solubility of liquids is in practically all cases
reciprocal, and without an analysis of the two layers the true solubility can
not usually be deduced.
Titration Method.-A special case of the reciprocal solubility of liquids is
that representing equilibrium in ternary systems yielding two liquid layers.
Such equilibria are usually determined by relatively simple titration procedures, but for the interpretation and description of the results, special terms
have been developed and these require more or less detailed explanation.
When a third liquid is added to a mixture of two others which are miscible
to only a slight extent, the added liquid, if soluble in each of the others, will
distribute itself between the two and an equilibrium will be reached. If the
two layers are then analyzed and the results plotted on coordinate paper, two
points, corresponding to the two layers, will be obtained. If more of the third
liquid is added, equilibrium will again be established after a short period of
shaking and the analysis of the two layers, to which the designation conjugate
layers has been given, will fix two more points when plotted on the coordinate
paper. The process may be repeated until a considerable number of points
have been obtained. When this has been done, it will always be found that
these points are the locus of a smooth curve, to which the designation binodal
curve has been given. If the pairs of points corresponding to the conjugate
layers are. connected, the lines so obtained are defined as tie lines. Since it is
evident that with the continued addition of the third or consolute liquid, a
point .\Dust finally be reached at ,which the resulting mixture will no longer
s'eparate into two conjugate layers, the tie lines successively determined as
above described will become shorter and shorter until finally the last one is
reduced to the point corresponding to the homogeneous mixture of the three
components. To this is given the name plait point.
Although for the above example a ternary system made up of three liquids
has been taken, there are a large number of salts and other solid compounds
which, when dissolved in mixtures of liquids of certain concentrations, cause
the latter to separate into conjugate liquid layers. These systems have aroused
much interest from time to time and considerable data for them are given in
the literature.
Since~t is usually difficult and frequently impossible to analyze directly a
homogeneous mixture of liquids, and thus determine the points on a binodal
curve, a simple titration method for this purpose has come into general use.
By means of this a homogeneous mixture of known amounts of two of the
components is titrated with the third just to the point of initial separation of
the second layer, which is usually very sharply indicated by the appearance of
clouding or opalescence. The procedure may also be reversed and the consolute liquid added just to tile point of clearing of the clQudy mixture of the
other two. By this plan the synthetically derived composition of one of the
two conjugate layers and thus of one point on the binodal curve is known.
The determination of the tie line and therefore, the identification of the corresponding point on the curve for the conjugate liquid, requires an additional
experiment for its location. Several procedures for this purpose have been
developed. They usually depend upon the determination of one or more

2586

DETERMINATION OF SOLUBILITY

constants of specially prepared pairs of conjugated liquids, such as their specific

refractiv~ indices. ~n the c.ase. of mix~ures 9f which one member
gravities
can be easIly determmed analytIcally, tIe hnes call be located by the quantitative determination of this member in pairs of COlljugated liquids.
In general, the titration method for the deterIll_inatio~ of the solubility of
liquids is applicable to many cases. The' facts, that eqhilibrium is attained
so promptly in liquids and that the evidence of the appearance Of a second
insoluble layer is usually so striking, make it of great value. Refinements
have been introduced such as the I;tddition of liquid or solid dyes to the mixture in order to facilitate the detection of the end l:loint, and the development
of particular forms of apparatus for measuring and weighing 'the liquids.
The constituents of the mixtures are usually weighlld but the volume relations
and, therefore, the specific gravities can also be aDproximately estimated, by
using graduated vessels for making the titrations, and measuring in them the
volumes of the final mixtures.
As a usual thing the. temperature coefficients are not very great in the
case of liquid mixtures and the very accurate contrl)l of the temperature is not
imperative. When such control is necessary, however, the use of a thermostat
does not seriously complicate the determination.
Distribution Coefficients.-As mentioned ab,ovll, when a third compound
j8 .2{){}er} to .2 mjxtll;re of two ).iq~ll{k whje.~ .2;r& ;relatively immiscible, it- will
dissolve to a certain extent in each and the composition of the two' layers
represent conjugate points on the binodal curve for the system. The results
are, however, of interest from another point of view, namely that of the distribution of the compound between the two solvents. This distribution coefficient is, in many cases, of considerable interest in connection with analytical
methods based on shaking out procedures and also in connection with such
problems as the molecular state of compounds in solution, their dissociation
and other points of theoretical interest. Distributi.on coefficients have, therefore, been studied to a large extent and much da,ta for them are available.
In general, the determinations are made by relatively simple methods: The
amount of the compound present in a definite amour;tt of ach layer, after equilibrium has been established by adequate agitation, 'is determined in any manner most convenient. If the total amount of solut~ is known, and that found
in one layer, the amount in the other can, of course, be calculated by difference.,
The results are usually expressed on the volume bilBis, since it is the ratio of
the amounts pres.ent in the s~me molecular state irt equal volumes of the two
layers which is a constant, independent of temperature and concentration.
It is evident that when the concentration at the saturation point is con-..
sidered, the amount of the compound which enter!> each layer depends upon
its solubility in the liquid, consequently the distribution coefficient is the
relation of the solubilities of the dissolved substance in the two solvents.
Variations from this, aside from changes in molecular state, etc., in one or the
other solvent are due to such causes as the reciprocftl solubility of the so-called
immiscible solvents, which will, of course, be influerlCed by the presence of the
dissolved compound, especially at the higher concentrations. Variatio~,'; of the
coefficient with temperature would result in cases where the solubilities of the
compound in the two solvents do not change at the sam~llte with temperature.
Electrolytic Conductivity Method.-Of the physical- properties which can
be used for the determination of the concentrati.on of a solution, such as

0:

DETERINATION OF SOLUBILITY

2587

specific gravity, refractive index, etc~, the electrolytic conductivity is of particular value in the case of those very sparingly soluble compounds which yield
solutions too dilute to be analyzed by gravimetric or volumetric methods.
By its use the progress of the saturation can be followed without separating the
undissolved solid from the solution, or even removing the portion used for the
determination. The special electrical equipment which is required, however,
and the need for water of exceptional purity and of vessels of particular qualities, restrict its general use.
The method of calculating the concentration from the conductivity is based
on the assumption that at the very great dilutions involved, complete dissociation occurs. Therefore, the limiting value to which the equivalent conductivity approaches at infinite dilution is, for practical purposes, attained,
and A=Aoo=la+lk, where la and lk are the ionic conductivities of the anions
and kations. These values'are known for all the principally occurring ions.
The observed specific conductivity K is, however, connected with the equivalent
conductivity and the concentration TJ by the equation A =K/TJ, in which TJ rep.resent~ the concentration in gram-equivalents per cubic centimeter. Rearrangement and substitution giveTJ =K!Cla+lk). From this equation the solubility of the
substance under investigation is calculated by substituting the measured specific
conductivity of the solution and the known values of the ionic conductivities.
The Solubility of Gases in Liquids.-When a gas and a liquid are intimately
mixed by shaking, a definite amount of the gas will be dissolved by the liquid
and, simultaneously, the vapor of the liquid will mix with the gas in the space
~bove the liquid. The partial pressure of the liquid in the gas space is almost
exactly the same as that of the pure liquid at the solution temperature, since
the influence of the relatively slight amount of dissolved gas is insignificant
in by far the most cases. The amount of gas which is dissolved depends both
on tlie nature of the gas and of the liquid and is, furthermore, a function of the
temperature, and pressure.
In regard to the influence of pressure, the absorption ll\w of Henry holds for
the most part, when the gas solubility is not too great. According to it, the
amount of pure gas, which is taken up at constant temperature by a given
a1nount of liquid is proportional to the pressure of the gas.
The temperature acts almost always in the sense that the solubility decreases
as the temperature rises.
The solubilities of gases are usually expressed either in terms of the Bunsen
"Absorption Coefficient" {j,t or the Ostwald" Solubility Expression" [.2
~~=the Bunsen Absorption Coefficient which signifies the volume (v) of the gas (reduc& to 0 and 760 mm.) taken up by unit volume (V) of the liquid when the pressure
of the gas itself minus the vapor tension of the solvent is 760 mm.
v
f:J=V(1+0.00367 t)
2 l=the Ostwald Solubility Expression which represents the ratio of the volume (v)
of gas absorbed at any pressure and temp~rature, to the volume (V) of the absorbing

liquid, i.e. 1=.p.

This expression differs from the Bunsen Absorption Coefficient, ~, in

that the volume (v) of the dissolved gas is not reduced to 0 and 760 mm. The solubility
1 is therefore the volume of gas dissolved by unit volume ofthe solvent at the temperature of the experiment. The two expressions are related thus:
1
l=,8(1+0.00367 t),
,8 (1+0.003670

2588

DETERMINATION OF SOLUBILITY

The experimental methods for the determination of tl;le solubility of gases

vary according to the nature of the gas. For tliose ~hich Idissolve in relatively
large amounts and can be analytically determined with accuracy, the saturated
solution may be analyzed by ordinary quantitative methods. Thus, in the
case of the solubility of sulfur dioxide in aqueous, $olutiohs of salts, the solutions were saturated by passing a stream of the gas througl\. 'them at atmospheric
pressure and, ,when equilibrium was attained, a measured p6rtion of the solution
was withdrawn, transferred to an excess of standardized iodine solution and the
excess of the latter titrated with thiosulfate. A gravimetric procedure was
used by Christoff (1905) for the determination of the solubility of carbon dioxide
in aqueous salt solutions. In this case the solutions were weighed before and
after the passage of the gas through them and the increase in weight, after
applying necessary corrections, taken to represent the solubility at the temperature of the experiment and at atmospheric pressure. The absorption flasks were
of special shape and the gas was previously passed through a series of U tubes,
containing the same aqueous solution, in order to prevent loss of water from the
experimental solution which, otherwise, would have occurred.
In the great majority of cases, however, gas solubility is determined by a
method based upon the measurement of the volume of the gas absorbed. The
apparatus consists essentially of an absorption flask for the liquid, connected
by means of a tube of small bore to a graduated buret in which the gas is
measured above mercury, the level of which can be altered by raising or lowering
a container connected with the buret by means of a rubber tube. Many forms
of this apparatus have been described and the disadvantages of the earlier
forms have gradually been remedied. A relatively simple form of this apparatus, but one which embodies the essential features required for accuracy, is
that described by McDaniel (1911) for the determination of the solubility of
methane, ethane and ethylene in a large number of organic solverfts at various
temperatures.
This apparatus is shown in Fig. 499. A is an ordinary gas buret and B an
absorption pipet of the form first used by Ostwald. '''.J!he buret and pipet
are connected by means of the glass capillary M sealed directly onto each, so
that the whole forms one solid piece of glass apparat:us without rubber or cement
connections of any kind; thus any possibility of leaks from these extremely
troublesome sources is entirely avoided. The whole apparatus is clamped
solidly to a rigid support so that it can be taken up in the hands and shaken
for the purpose of bringing the, gas into intimate contact with theJiguid. The
pipet and buret are each provided with a three-way stopcock, C and D.- These
can be turned in such a way as to allow the gas to sweep out the air-from the
connecting capillary. By the same means the two vessels may also be connected directly with each other as well as separatefy with the outside air or
source of gas supply. The pipet and buret are elLch provided with a water
jacket, P and Q. The temperature of each is regulated by means of the
electrically heated coils K and L." These coils are of rtJ.anganin wire' and are
connected in series. The rate of evolution of heat in the jackets was' adjusted
in the first place by varying the length of the manganin wire, until the tempfrature was the same in each jacket. Stirring was accomplished by blowing air
through the tubes I and J. The differences in temper.!!-,tiue between the pipet
and buret were never greater than 0.1.
.

DETERMINATION OF SOLUBILITY

2589

In carrying out a determination by this method it is, of course, necessary

that the solvent be completely free of dissolved air or other gas. This is perhaps the most important part of the determination and a special form of apparatus for the purpose is described by McDaniel (1911) and is shown in Fig. 500.
"The liquid was boiled under diminished pressure in the flask C attached
directly to the lower opening of the pipet by means of the rubber stopper as

FIG. 499.

McDaniel's apparatus for determining

gas solubilities.

FIG. 500.

shown in the figure. Connection with the air pump is made at D. During the
boiling the lower opening of the inlet tube E is above the surface of the liquid
in C, the stopcock B being closed. When the air.has been completely expelled,
the screw pinchcock F is closed while the air pump is still in operation. The
flask C is now raised until the lower end of E reaches nearly to the bottom of the
flask. The air pump is now connected at G and the cock A opened so as to
make connection with the pipet. B is now opened .and the inflow of air
through D regulated by gradually opening F in such a manner that the liquid
is very slowly forced up into the pipet. In this manner the liquid never comes
into contact with the air under full atmospheric pressure but only under greatly
diminished pressure. The absorption of air under these conditions can only
be inappreciable, especially since the liquid in the flask remains perfectly quiet,
and only the lower portion is used."

2590

DETERMINATION OF SOLUBILITY
,

Having filled the pipet B, Fig. 499, with the air-free solvent as just described,
"T is connected with the source of gas supply and the cocks C and D are turned
in such a way as to allow the gas to sweep out th~ air from the capillary, M.
The buret is then filled in the usual manner by lo\yering the leveling tube F,
the cock D having been turned so as to connect T \with E. Care is taken to
keep the entering gas under a slight pressure by keening the mercury level in F
slightly above that in A. This prevents air from ehterink through any leaks
in the train connecting the gas generator with the buret." The gas must be
completely saturated with the vapor of the solvent and this, with other than
aqueous solvents, may require, in addition to drawing it through some of the
solvent in H, that a thin layer be placed in the buret and time allowed for it to
saturate the gas sample.
'; After again allowing the current of gas to flow through the capillary M
for a short time the buret and pipet are connected with each other by turning
the three-way cocks D and C in the proper direction. The determination of
the amount of absorption is then made as follows: A portion of the- gas is
passed into the pipet by raising F and opening G, the displaced liquid being,
caught in a graduated cylinder. The cock q .is closed and the gas and liquid
in the pipet brought into intimate contact with each other by shaking the whole
apparatus. C is now opened to allow gas .to enter from the buret to replace
that absorbed. This process is repeated until, on ope~ing
there is no further
decrease in the volume of gas in A. The volume absorbed is found by subtracting from the original volume of gas, the volume remaining in the buret
plus the volume in the pipet. The volume of gas in the pipet is equal to the
volume of liquid drawn off. The volume of liquid remaining is easily calculated fromth!l known volume of the pipet. The absorption coefficient or
'solubility' is the ratio of the volume of gas absorbed, measured at th~ temperature of the experiment, to the volume of the saturated liquid. It may be
reduced to the coefficient used by Bunsen by dividing by (l+dt)."
In the case of the majority of investigators who have used this method,
particularly for determinations at high or low temp~ratures, the absorption
pipet has been kept at the temperature of the experim~nt and the gas measuring
buret at room temperature, the two being conhected by mean.s of a flexible.
capillary which permits the absorption pipet to be independently shaken.
This arrangement makes it necessary, in calculating the absorption coefficients,
to apply the usual corrections for temperatu~e and vapor pressure to the
volume pf gas in the buret~ This is a complication which in some cases causes
uncertainties in regard to the accuracy of the results as finally calculated.
An apparatus designed for determinatio~s at very high presirures, using a
Caillet compression tube, is described by Sander. (1911-12). It was used for
determination of the solubility of carbon dioxide in water, alcohols, and other
organic solvents. The principle involved is that the pure gas is first compressed above mercury in a graduated tube an~ the volumes correspondiNg to
given pressures noted. Similar readings are then taken for the same gas after a
i:\maU accurately measured amount of solv~nt has been introduced into the
graduated tube. The difference between the two'volumes at the same temperature and pressure, reduced to 1 kg. per sq. cm:_...aRd-1 ml: of liquid, represents the solubility of the gas in the given solvent.
-

c,

DETERMINATION OF SOLUBILITY

2591

Finally, attention should be called to the method of determination of gas

solubility based on the principle that, for volatile solutes which obey the laws
of Dalton and Henry, the amount which is carried away by an inert gas when
known volumes are bubbled through solutions of known strength of volatile
solute, can be used to measure the comparative solubilities in solvents of different concentrations. An example of this method is the determination of the
solubility of ammonia in aqueous salt solutions by Abegg and Riesenfeld (1902).
The very ingenious apparatus consists of a generator for developing a stream
of H 2+0 2 from aqueous NaOH, by means of an electric current measured with
the aid of a copper voltmeter, and the volume of gas thus determined. This
was passed through a spiral in the vessel containing the ammonia solution or
known concentration. The mixed gasses passing out of thIs were received in a
third vessel containing 5 m!. of 0.01 n HC!. Electrodes were provided in this
vessel and, by means of conductivity measurements, the point determined at
which all of the HCI became saturated with NH a. Since the volume of the
H 2 O2 required for this purpose was known, the partial pressure of the NHa
in the mixture could be directly ascertained. Comparative determinations of
the'vapor pressure of the ammonia in water and a series of salt solutions made
in this way were calculated to ammonia solubilities on the basis of the relation
that, for two solutions of equal ammonia content, the ammonia pressure is
reciprocally proportional to the solubility of the ammonia in them.

SPECTROGRAPHIC ANALYSIS
1
,
,

Introduction.-The analysis of materials by a spJctrographic method is

dependent on the atomic excitation of the substance and the observation of
the radiant e:q.ergy produced by means of a spectroscope or spectrograph which
separates this energy into an ordered arrangement of increasing wavelength or
frequency of vibration. The success of the method is dependent on the ability
to excite the molecule or atom and the intensity and character of the radiant
energy produced. One might divide the subject of spectrographic analysis
into qualitative and quantitative analysis, and also into the fields of. emission
and absorption analysis. From the standpoint of elemental analysis the division of Qualitative and Quantitative emission analysis is most important,
although some mention will also be made of the application of absorption
spectrographic analysis. In general it might, be said that qual~tative and
quantitative spectrographic analyses are usually applied to the detection and
estimation of minute amounts or traces of ,substances. The' accuracy of a
spectrogra.phM quantit~tive analysis is seldom greater than 5% of the amount
of substance present and, as a 'rule, this represents a rather large error compared
with the accuracy of chemical methods, except when the amount of substance
or element present is very small.
The excitation of an atom by means of. high temperature or electrical discharge results in the forcing of electrons from certain inner levels to outer
levels or even completely out of the sphere of influence of the atom. The retwn
of the electron to its original level or some other level and then from that level
to another or its original level results in the emission of radiant energy which is
definitely characterized or we might say quantized, in that it constitutes a
definite voltage energy change which may be measur~d by the frequency or
wavelength of the radiated energy. This radiat()d energy when observed in a
spectroscope or spectrograph appears as a spectrllm line and indicates not only,
the existence of the element but also the state of the atom (Le. the number of
electrons completely removed or the state of ionization). The number and
intensity of the spectral lines are influenced by many factors. In some elements the excitation energy is concentrated in a limited 'numbei' of possible
electron jumps with the production of a rather simple spectrum (such as K or
Hg), Fig. 501, whereas, in other elements the conditions in the atom may
produce a scattering of the energy over a large number of possible electron
jumps with the production of a very complicated'tpectrum (such as Fe or ~~).
The theoretical spectroscopist often indicates these differences in energy
levels in volts, but for purposes of analysis it is more convenient to record them
in the frequency or wavelength of radiant energy ~hich is prod~ced by this
voltage change. Th.e un~t o.f wavelength (X) in ~pectros~opy is usually the
angstrom (At (sometImes mdlCated by A, A. U.. Or I.A.) whlCh may be defined
as 1 A = 1 meter X 10-1 . Because of the importance of this uhit and the large
1 Chapter by Wallace R. Brode, Associate Profess<lr-of"'Chemisfry, The Ohio State
University.

- '

2592

SPECTROGRAPHIC ANALYSIS

2593

number of accurate observations which are based on it, the actual unit of
measure is not derived from subdivision of the'meter but rather by the assignment to the red cadmium line the value 6438.4696 as the wavelength primary
standard. The various units for the indication of wavelength or frequency
values bear the following indicated relationships.

..
FIG.5010 Energy Level Diagram for Potassium (Numbers on lines between
levels indicate the wavelength of the emitted energy.)

Wavelength (h) (length of wave in a complete cycle).

1 A (angstrom) = 1O-1Q meters = 1{6438.4696 of the wavelength

of the cadmium red line.
1 m,u = 10 A = 10-9 meters.
1 ,u = 1000 m,u =f 10,000 A = 10-'-:6 meters.
Wavenumber (v') oscillations per unit distance.

cm-1 = oscillations per cm.

Frequency (v): Vibrations per unit time.

f (fresnel) = vibrations/seconds X 1012

Relation between systems
l{X=v' =v{3 (for example, 5000 A=500 m,u =20,000 cm-1 =600 f).

2594

SPECTROGRAPHIC ANALYSIS

The photographic range of radiant energy .include~ the spectral range of

2000 A to 14,000 A. This includes portions of the infra red, the visible region
and portions. of the ultraviolet.
Infra red above
Red
Yellow
Green
BJU8

mtraviolet below

7500 A
7500 to
6000 to
5500 to
5DDD to
4000 A

6000
5500
5000
4DDD
\

It is interesting to note that the wavelength region of the visible in n\JL

units includes the numerical range from 400 to 750, while the frequency range
for the same spectral range goes from 750 to 400 J. For the recording of spectral
data, especially in connection with absorption and band spectra, it is more
advisable to record in frequency or wavenumber" in that the data is usually,
capable of analysis into multiple band systems on a frequency basis. If a
quartz spectrograph is 1,lsed the dispersion is nearly linear for the spectral region
when plotted on a frequency scale so that the accuracy of plotting or reading
is approximately the same in all regions of the spectrum. With a grating
spectrograph the data is observed on a linear or nearly linear dispersion based
on the wavelength scale. For all spectrographic data involving line or emission
spectra the data are recorded in wavelength in angstrom units, and for nearly
all types of qualitative and quantitative analysis this is the only system' with
which one need be familiar.
4

APPARATUS
Spectrographic analysis involves rather complicated apparatus and requires
a considerable capital outlay with regard to the cost of the equipment. The
upkeep, however, is slight and once the apparatus has been properly adjusted
and calibrated the technical knowledge for the preparing of the photographic
data and the interpretation of the data is such that much of the work can bEt1
done by technical assistallce in the laboratory., The .type of spectrograph
_ necessary for analytical work w~ll depend on the (lomposition of the samples to
be analyzed, the degree of accuracy expected and the speed with ;Which the
analysis must be made. For many of the non ferrous metals giving simple
spectra, such as Zn, Cd, Hg"Ag, Au and the like a high dispersionspectrograph
is not necessary. The usual type of Cornu prism instrument (Fig. 502) with
a dispersion of about 10 inches from 7000 to 2000 A is quite satjsfactory, and in
some cas!,!s a slightly lower dispersion such as given .!>l'--a small model Littrow
or the Cornu instruments may be sufficient for puqfoses of analysis (Figs. 503
and 504).

SPECTROGRAPHIC ANALYSIS

2595

For the analysis of ferrous metals, W, Mo, U, Pt and elements giving large
numbers of lines it may be necessary to use an instrument capable of giving
sufficient dispersion to separate the spectrum lines. For this purpose one
usually employs a grating instrument or a prism instrument of the Littrow
. typ~ (Figs. 505a, band 506).

PRODUCTION OF A SPECTRUM BY MEANS OF A PRISM

,~

.J

SOURCE

SLIT

... 1

/1

LENS

LENS

L3
OPTICAL DIAGRAM OF A
SPECTROGRAPH

FIG. 502.

Production of a Spectrum by a Prism and an Optical Diagram of

a Prism Spectrograph (drawing is not to scale).

Both of these instruments (C~rnu and Littrow) employ the same optical
principles as indicated in the illustrating figures (Figs. 502 and 505). The
prism spectrograph or spectroscope utilizes the optical principles of refraction
and dispersion, i.e., the bending of a light beam by a prism and the fact that

FIG. 503.

Bauscl;t and Lomb Spectrograph Medium Size (4X10 in spectrum plate).

the degree of refraction is influenced by the wavelength or frequency of the

incident light or radiant energy. For most spectrographs the prism and lens
systems are made from quartz because of its transmission of radiant energy

2596

SPECTROGRAPHIC ANAU)fSIS

ove'r the entire spectral region to which the phot"ographic plate is sensitive.
The'light which is focused on the slit of the instrument is collected by a collimating lens so as to project a parallel beam through the i~rism, After refraction

FIG, 504. Hilger Spectrograph

(Sma~

Size instrument

No. E 484),

by the prisms and the dispersion of the beam the light is focused by a second
lens on the plate or eyepiece, It should be noted that in the Lit.t row instrument a single lens serves both as focusing and collimating, and the prism 1s only
30 with a mirro~ back so that the light beam travels. half 'lay through !"hat

FIG. 505a.

Small Littrow Spectrograph.

(Bausch and Lomb.)

would be a 60 prism if it were continued, and then is reflected 'b'ck so that the
beam passes through the prism twice, In view of.ltlle economy of .quartz
pieces, as well as size of the instrument, the Littrow construction is particularly
" well suited for .the larger instruments with high dispersion.

SPECTROGRAPHIC ANALYSIS

'2597

The illumination of the slit of the spectrograph may be varied to suit the
conditions of the analysis. While the common practice is to focus the image of

FIG. 505b.

Illumination System for Small Littrow Spectrograph.

(Bausch and Lomb.)

the source on the slit, and so diaphragm the slit with a wedge or Hartmann
diaphragm (Fig. 507) as to cut off undesirable illumination from certain portions

FIG. 506.

Hilger Littrow Spectrograph.

of the source, there may be occasions where it is desirable to use a large

amount of the sample so as to obtain an average analysis of the sample. In this
latter case the condensing lens may be entirely eliminated. If it is still desirable

2598

SPECTROGRAPHIC ANALYSIS

to diaphragm part of the source one.may use screen absorbers near the focusing
lens to reduce the total intensity. For certain types of ~nalysis it is desir'\ble
to have the slit illuminated uniformly along its entire length; this is especially
true in the use of the logarithmic sector method or the step slit, method of
analysis. This can be easily accomplished by the 'use of a.long focu::, tens placed
as near the slit as IS po5'sible and placing
the light source irl. such a position...that
the image of the source is focused on
the collimating lens of the spectrograph
(L1-Fig. 502). If one attaches a recFIG. 507. Wedge and Hartmann Diatan gular diaphragm to this lens it is
possible to eliillinate,the undesira,ble porphragm for Spectrograph Slit.
,tions of the spectrum. This system
of illumination does not make full use of the resolving power of the instrument,
but in most cases this is not necessary and the reduction in the light intensity
is not a serious handicap in most analytical work since the increased e~posure
provides for a more thorough sampling of the material to be analysed.
>

LIGHT SO,tJRCES

Flame.-Flame sources are useful for rough qualitative tests which may
be applied in addition to visual color tests of the flame, for, such elements av
Cu, Na, Li, K, Ca, Sr;Ba, etc. The temperature a.nd ,constancy of the flame
are not very satisfactory for the spectrographic analysis and with the exception
of the alkali metals, those elements which give colored flames, usually do so as
compounds with certain anions (i.e. QaCI+, CuCI"", etc.), the color being caused
by the band spectrum of the diatomic or triatomic molecule rather than by the
simple element' A quantitative method has been described, however, for the
estimation of boron by the color of the green flame produced-by the burning of
methyl borate, by which relatively small amounts of boron may b;aetected and
estimated quantitatively.
Arc.-The two most important methods, of atomic excitatiqn for spectrographic analysis are by means of the electric arc and the electric spark. It is
difficult,to draw a sharp dividing line between these methods especially when
high voltage arcs of 2 to 3 thousand volts are comtnonly used. In general,
however, the arc sources are confined to the use of current densities of 3 to 5
amps. and 50 to 200 volts (Fig. 508). The arc constancy can be improved by
the inclusion in the circuit of a reactance (such as the primary coil of a 1 KVA
transformer in which the secondary circuit has been shorted). While the
current source may be from 120 to 220 volts (DCl.,tli.e-actmtl drop across the
arc gap may be only 40 to 60 volts, due to the use of a resistance in the circuit.

SPECTROGRAPHIC ANALYSIS

25~9

One usually uses rods of the sample about t inch in diameter. Metals are
usually worked by using pairs of rods of the material. For non-conducting
substances and small quantities, as well as solutions, carbon rods are usually
used. In such cases the,lower carbon is usually cupped so as to take the sample.
The lower carbon may be the positive or negative electrode, depending on the
substance for which the analysis is being made and the spectral characteristics
desired. The commercially available carbon corries in several grades, of which
the spectrographic and high purity types are suitable for most analyses. The
highest purity material is usually free from nearly all impurities with the excep-

~
~
D.C.
Supply

o_ _ _ _

Spark

Arc

~~~_ _ _ _~~t
.[

FIG. 508. Electrical Diagram for Spark and Arc Sources. T=Transformer 10,000
to 20,000 V.; L = Self Inductance 0.25 millihenry; C = Capacity, 0.005 microfarads:
D.C. Source 110-220 volts; R=Resistance; .Arc current 2 to 5 amps., 40-70 volts;
I = Reactance.
.

tion of a trace of Si. Less expensive rods usually contain minute traces of
Fe, AI, Ca, Cu, Ti, Mg, and V. These may not interfere with the analysis
or they can be removed by purification. One method of purification involves
the soaking of the carbon in HCI and HNO a A recently described method of
purification which yields excellent results is effected by heating the carbon as a
resistance rod to a temperature of about 2400 in a vacuum furnace. This
treatment seems to be effective in the removal of nearly all of the impurities
from carbon rods containing a considerable amount of impurity. In some cases
it is pQssible to use pure metal electrodes, such as copper as the holder for the
substance to be examined. Duffendack and Wolfe have recommended the use
of a high voltage arc using 2.5 amps. at '2500 volts. It is claimed,that this procedure reduces if not entirely eliminates background and that in the examination of salts the thin layer of salt solution on the carbon rods, formed by the
evaporation of a drop of solution on the rods is sufficient to prevent the production of the carbon-cyanogen bands. The temperature of the electrodes remains
very low so that low-melting alloys may be conveniently analysed by this
method.
.
Spark.-The change in electrical excitation from arc to spark results in
an alteration in the spectrum with regard to intensities of lin(Os, production of

SPECTROGRAPHIC ANALYSIS
new lines and the elimination of arc lines. :rhe usual electrical system is
indicated in Fig. 508. The use of a self inductance changes the spark charaderistics with a reduction in the intensity of air lines (from 0 and N) and in some
c.ases may so alter the characteri~tics of the spark so thaH~t. may appear. I?-0re
hke an arc spectrum. The self mductance does help stabIhze the condItIOns,
and it is to be recommended in spectrographic analysis,) especially where it
may be desirable to have reproducible conditions. There appears to be a satisfactory medium between the condensed and oscillatory spark since the factor
determining the number of oscillations is a function of tIre product of the
capacity of the condenser system and the inductance of the self inductance.
Since the equipment available in each laboratory will differ in some detail
and the' conditions of excitation must be vaHed for different elemental analysis,
it is not practical to give in this limited discussion any more detailed descriptions of arc and spark procedures. Neariy all articles on spectrographic
analysis indicate with some detail the method of excitation and in attempting
to duplicate such methods it is advisable to follow th~ excitation proced)lre as
closely as possible. In the spectrographic analysis of solutions the sample
may be applied to a carbon electrode and arced or sparked or a spark d~scharge
may be effected by using the solutio.n as one electrode and a carbon, gold' 01:
copper rod as the other electrode.
Other Light Sources.-For the analysis of &ases one usually uses a gas
discharge tube with a high frequency high voltage current. The presence of
impure gases may often interfere with, or eve~ completely eliminate, the lines
of certain substances so that the method may not be entirely satisfactory.
For the purpose of providing a b,Jtckground of nearly continuous radiation for
the observation of absorption spectra one may use ,the incandescent tungsten
filament for the visible and near ultraviolet. The ultraviolet range may be
extended a little by the use of a special bulb of correx glass or a quartz window
in the glass bulb. For an extension in the ultraviolet to the limit ofthe gelatin
emulsion one may use an under water spark from a tesla coil or a hydrogen
discharge, both of which give a continuous spectrum, without spectral lines over
For some types of
nearly the entire photographic range from 2000 to 8000
absorption work, especially where the bands are broad, satisfactory results can
be obtained by using an arc or spark between tungsteiJ., iron or nickel electrodes.
Spectrographic Standards.-In the definition of dispersion measurement
terms it was indicated that rather than defining wav~length. in angstrom units
as a ten thousand millionth of a meter it was defined as 1/6438.4696-part of the
wavelength of the cadmium red line, and this line is known as the primary
standard.' As further guides in the measurement of the wavelength of unknown lines a number of secondary standards have been designated and their
wavelength .accurately measured. Tliese secondary'standards include some
217 iron, 9 copper, 3 silicon, 2 nickel and 25 neon lines. I The almost universal
use of the iron spectrum as a means of identification and comparison of other
spectra has resulted in the accurate determination of the wavelength values
of" some three hundred and eighty-four additional lines which are known as
tertiary standards. The wavelength values of unknown lines can he accurately
determined by measurement of the spacing between the Jlnknowti line and selected standard lines and the application of the Hartmfnn formula to these
data .

A:

SPECTROGRAPHIC ANALYS!S

2601

Hor the identification of the unknown element in a spectrographic analysis,

a number of special tables and photographic charts have been prepared. These
tables and charts may be subdivided and classified in accordance with their
.
method of preparation and use.
1. Spectrum lines in accordance with the element. The most complete
table (in so far-as it has been finished) is the Kayser-Konen Handbuch (Vol.
7 and 8), in which the elements up to Nd in an alphabetical arrangement have
been covered. It is of course impossible to cover adequately or report all of the
possible lines of an element that might be produced by extremely long exposures
or by increasing the ionization state. The Kayser collection presents, however,
the most complete assembly of spectral line values for these elements that has
as yet been published. It is hoped that within a few years a more complete
, set of tables can be published from data which is being collected by G. R.
Hairisop of the Massachusetts Institute of Technology, and for which over two
million line measurements have been made with an accuracy of within .001 A.
Kayse.r published in 1910-1912 two volumes (Vol. 5 and 6 of the Kayser
Handbuch der Spectroscopie) covering available data on the elements at that
"time. Other available lists included the data published by Exner and Haschek,
the Tables Annuelles des Qonstantes and miscellaneous original research.
publications. For the aid of the qualitative spectroscopist condensed table-:
of the more intense lines have been published, the most convenient of these is to
be found in the International Critical Tables. (This data is also reprinted in
the Handbook of Chemistry and Physics.) It has been shown that upon
dilution of elemental concentration .of a known substance in an arc or spark,
that the intensities of the spectral lines due to that element will be reduced and
that the weaker lines will gradually fade out through lack of sufficient intensity
to affect the photographic emulsion. The last lines to disappear are known as
the persistent lines or the Raies Ultimes. These lines have been determined
for nearly all of the known elements, together with their spectral class (Table I).
In the examination of a spectrogram of an unknown sample for certain elements,
the absence of the Raies Ultimes of that element in the spectrogram is usually
sufficient proof for the absence of the element.
2. A second type of spectral tabular data involves the arrangement of lines
in accordance with wavelength values rather than by elements. This type of
arrangement is particularly useful in the qualitative identification of unknown
elements present in the sample. The best available table of this type is that of
Kayser (Hauptlinien). There are a number of more condensed tables such as
the 10,000 line table in the Lange Handbook, Twyman (second edition), Lowe,
Scheibe, etc. A similar table to some of these has been prepared directly from
more extended tables (Table II). This table is confined to the relatively more
intense lines and especially those of the arc spectrum. It is of course difficult
to know where to limit a table of this type since one may.not be sure in the
identification of a single line as to whether it is a weak uncharted line of the
element present in greatest amount or the strong line of the element present in
weak amounts. For this reason it is usually customary to verify the presence
of elements by the identification of several lines of the element in question.
In Fig. 509, which shows the spectra of brass and copper samples, one
can note the procedure in the identification of impurity or alloying elements.
The method of taking photographs of comparison or pure samples adjacent

2602

SPECTROGRAPHIC ANALYStS
TABLE

PERSISTENT ARC LINES OF THE E:J.,EMENTS

Ag I
Ag II

All

. Al II

AI III
AsI

Au I

BI
BII

Ba I'

Ba II
Be I

Be II

BiI

CI

CII

Ca I

(I, II etc. indicate atomic states; 1 ind\cates the most 'persistent lin~s)
,
13280.7A
'Ca I
4456.6
'Er I
3692.7 I \
La I
-5455.1
3382.9
Ca II 13933.7
3906.3 I
15930.6
Eu I
4129.7
6249.9
2246.4
3968.5
2437.8
Cb I
14059.0
4205.0
La II 13949.1
3082.2
4079.7
Fe I
13719.9
4077.4
3092.7
4101.0
3737.1
4123.2
13232.7
3092.9
4123.9
3745.6
Li I
3944.0
4137.1
3748.3
16701.9
13961.5
Cb II 13094.2
3745.9
Lu I
4518.5
Fe II 12382.0
Lu II
2894.9
1671.0
3130.8
1856.0
3163.4
2395.6
2911.4
1858.1
3195.0
2404.9
3391.0
3225.5.r:J
2410.5
3472.5
1862.5
Cd I
12288.0
2413.3
3554.4
1854.7
1862.9
3403.7
Ga I
4033.0
Mg I
12852.1'
346'6.2
14172.0
3829.4
1889.9
1936.9
3510.5
Gd I
3646.2
3832.3
Cd II 12144.4
3768.4
3838.3
1972.0
Ge I
2651.2
Mg II 12795.5
2288.1
2265.0
2349.8
Ce II
4012.4
i 2651.6
2802.7
2780.2
4040.8
3039.1
Mn I
14030.8
2860.5
4165.6
3269.5
4033.1
12427.9
4186.6
4226.6
4034.5
Hf I
2898.2
Mn II 12576.1
2675.9
Co I
13453.5
2496.8
3465.8
2904.4
2593.7
12497.7
3529.8
2916:5
2605.7
2940.8
Mo I
13798.3
1362.5
Co II 12286.2
3452.3
2307.8
3072.9
3864.1
5424.6
2378.6
4093.2
3903.0
Hf II
2513.0
Mo I I 12816.2
5519.1
2388.9
15535.5
Cr I
14254.3
2516.9
2848.2
5777.7
4274.8
2641.4
2871.5
14554.0
. 4289.7
2773.4 /
2891,0
4934.1
5204.5
2820.2 ,
2909.1
520&.0
3134.7
Na I
3302.3
12348.6
Hg I
11849.6
3302.9
3321.0
5208.4
Cr II 12835.6
2536.5
5890.0
3321.1
3321.4
2843.3
3650.2
5895.9
Nd I
3951.2
13130.4
2849.8
3654.8
3131.1
2855.7
3662.9
__ _
4177.3
Hg II
1649.8
-----4303.6
2061.7
2860.9
2276.6
Cs I
18521.2
1942.3
Ni I
13414.8
Ho I
3748.2
-3493.0
2780.5
8943.6
2809.6
4555.3
3891.0
3515.1
Ho II
2936.8
3524.5
2898.0
4593.2
Cu I
13247.5
In I
4101.8
Ni II
2253.9'
2938.3
3274.014511.3
2264.5
2989.0
Ir I
2849.7
2270.2
13067.7
Cu II 12136.0
12287.1
12478.6
2192.3
2924.8
1334.5
2247.8
13220.8
Os I
3262.3
Dy I
4000.5
3437.1
3267.9
1335.7
4267.0
4046.0
3513.7
3301.6
4267.3
4078.0
K I
4044.2
I
3752.5
14226.7
4168.0
4047.2 ~
3782.2
4454.8
4211.7
17664,(1' - . PI
1774.8
Er I
3499.1
7699.0
1782.7
4455.9

SPECTROGRAPHIC ANALYSIS

2603

TABLE I. Continued

,PI
Pb I
C>

PI) II
Pd I

Pd II

Pr I

Pt I

Ra I
Ra II

Rb I

Rh I

1787.5
2136.8
2149.8
2536.4
2554.0
12170.0
2833.1
3639.6
3683.5
4057.8
1682.4
12203.6
13404.6
3421.2
3517.0
3609.6
3634.7
2488.9
2498.8
2505.7
2658.7
2854.6
4062.8
4179.4
4189.5
4225.3
2659.4
2830.3'
2929.8
2998.0
i3064.7
14825.9
13814.4
4682.2
4201.8
4215.6
17800.3
7947.6
3323.1
3396.8
134.34.9
3658.0
3692.4

Ru I
Ru II

Sa I
Sb I

Se I
Se II
Se I

Si I

Si II
Sn I

3436.7
13499.0
3596.2
2678.7
2692.1
2712.4
2945.7
2965.6
2976.6
4390.9
4424.4
4434.3
2068.4
2175.9
2311.5
2528.5
2598.1
3232.5
3267.5
3907.5
13911.8
13613.8
3630.8
3642.8
1960.2
2039.7
2062.6
14730.9
4739.1
4742.3
2506.9
2516.1
2528.5
12881.6
3905.5
1526.8
11533.6
2840.0
2863.3
3009.1
3034.1
3175.0
3263.3

Sn I
SrI

Sr II
Ta I
Tb I

Te I

Th I
Th II
Til

Till

TIl
Tu I

UI

4524.7
14607.3
4832.1
4872.5
4962.3
14077.7
4215.5
13311.1
3318.9
3406.7
3509.2
3561.8
3848.8
3874.2
2142.8
2383.3
2385.8
2530.7
2769.7
3538.8
3601.1
4019.1
3290.6
3635.5
3642.7
13653.5
4981.7
4991.1
4999.5
5007.2
5014.3
13349.0
3361.2
3372.8
3383.8
3775.7
15350.5
3462.2
3761.3
3761.9
355'2.2
3672.6
4241.7

VI

VII

WI
WI!
YI
Yt I!
Yb I
Zn I

Zn II
Zr I

Zr II

3183.4
3184.0
3184.0
13185.4
13093.1
3102.3
3110.7
3118.4
3125.3
14008.8
4294.6
4302.1
2397.1
2589.2
3613.8
4643.7
14674.8
13710.3
3774.3
3788.7
3289.4
3694.2
3988.0
12138.5
3282.3
3302.6
3344.5
12025.5
2061.9
13519.6
3547.7
3601.2
4687.8
4710.1
4739.5
4772.3
4815.6
13392.0
3438.2
3496.2
357'2.5

to the unknown sample provides a rapid procedure for the identification of

impurity lines and the approximate identification of the wave length value,
especially if the adjacent spectrum be iron or some other element which gives
a large number of lines and for which accurate photographic charts are available. In connection with the spectrographic program at, the Massachusetts
Institute of Technology it is hoped that a table of 100,000 or 200,000 lines
arranged in accordance with the wavelength of the line will be published.
, The photographic charts available are usualfy of the iron spectrum although
there are charts for some other elements such as eu and Ne. A number of
elements are available in extremely high states of purity for the purpose of
taking comparison spectra. The most complete photographic set of data avail-

2604

. SPECTROGRAPHIC ANALYf;lIS
TABLE II
ARC LINES OF THE ELEMENTS

Zn II
8e I
BiI
Zn 11
8e I
8b I
Cu II
PI
Zn I
Te I
Cd II
PI
Pb I
8b I
Pb II
Cu
Cu
Cu
Cu
Cu II
Ag
NiII
Te
Te
All
NiII
Cd II
Te

All
NiII
BiI
Co II
NiII
Cd I

AsI

Ni
Pd
C
La
Ba II
Co II
Ag
8b I
Co I
Cd II
Ni
Ag
Ag
Ag
Ag
Fe II
Ba II
Fe II
Ba II
Fe II
Be I

(arranged according to wavelength values, I, II indicate atomic states)

2025.5A
As I
2349.8
Pd II 2505.7' \
Pt
~50.9'
2039.7
Pd
2351.4
Ag
2506.7
Be
2650.9
2061.7
8n
2354.8
8i
2506.9 I
Ge I.
2651.2
21)1)1.9
Co
2'30'3.S
Ni
2511).9
Ge 1
2051:\,
2062.6
Al
2367.1
Hf II 2513.0
:AI I
2652.5
8i I
2514.3
Pt I
2659.4
2068.4
As
2369.7
2\35.9
Cu I
2369.9
8i I
2516.1
Ag
2660.4
2136.8
As
2370.8
Hf II 2516.9
8n
2661.3
2138.5
Al I
2373.1
8i I
2519.2:
Pb
2663.2
2142.8
Co II 2378.6
Co
2519.8
Cr
2666.0
Fe II 2382.0
Fe
2522.9
Al II 2669.2
2144.4
2149.8
Te I
2383.3
8i I
. 2524.1
Cr II 2671.8
Te I
2385.8
Fe I
2524.3
Mo
2672.8
2170.0
2175.9
Co II 2388.9
Cs
2525.6
Au I
2676.0
2203.6
Pb
2393.8
Fe I
2527.4
Cr
2677.2
2227.8
Ni
2394.6
Ba II 2528.5
V
2678.6
2230.1
Fe II 2395.6
8i I
2528.5
Ru II 2678.7
W II
2397.1
8b I
2528.5
Cr
2678.8
2242.6
2247.0
Fe II . 2404.9
Co I
2529.0
Mo
2684.1
2247.8
Fe
2406.7
Te I
2530.7
Pt
2702.4
PI
2535.6
Pt
2705.9
2248.7
Co I
2407.3
2253.9 .
Fe II 2410.5
Bg I
2536.5
V
2706.2
2255.5
Ag
2411.4
Mo
2538.5
8n
2706.5
2259.0
Ag
2413.2
Ni
2545.9
Ge
2709.6
2263.5
Fe II 2413.3
P
2553.3
V
2715.7
Ni
2416.1
Zn II 2558.0
Pt
2719.0
2264.5
2265.0
Pd
2424.5
Fe II
2562.5
Fe
2719.0
Pd
2426.9
Al II
2568.0
Fe I
2721.0
2265.5
2269~1
Mn II 2427.8
Zr
2571.4
Fe II 2739.6
2270.2
Mn II 2427.9
W
25l2.3
Fe II 2743.2
2276.6
Au I
2427.9
Cd II 2573.0
As 12745.0
2286.2
8n
2429.5
Al I . 2575.1
Fe II 2746.5
2287.1
Pd
2430.9
Mn II 2576.1
Au
2748.3
2288.0
Pd
2433.1
Pb
2577.3 1
Cd II 2748.7
8i I
2435.2
Co
2580.3 i
Fe I
2749.3
2288.1
2290.0
Pd
2435.3
Fe
2585.9
Cr II 2750.7
2296.5
Mn II 2437.5
Ge
2593.6
W
2764.3
Mn II 2593.7
Cr II 2766.5
2296.8
Ag II 2437.8
2297.8
Ni
2437.9
8b I
2598.1
Ag
2767.5
Fe II
2598.4
Fe I
2767.5
2304.2
Ag
2447.9
Mn II .2452.5
Fe II
2599.4
--Tl I
2767.9
2307.9
2309.5
As
2456.5
Mn II 2605.7
Te - ___2769.7
Fe II 2607.1
8b I' 2769.9
2311.5
Ag
2473.9
Pb
2476.4
Fe II
2611.9
Hf II- 2773.4
.2311.6
2312.9
C II
2478.6
Pb
2613.7
Mg I
2776.7
2316.0
Fe
2483.3
Fe II 2613.8
Mg I
2778.3
Fe I
2484.2
Pb
2614.2
Mg I
2779.9
2317.0
W
2488:8
Pt
2628.0;
As I
2780.2
2320.2
Pd II 2488.9
Fe II 2628.3
Bi I
2780.5
2324.6
2331.4
As
2492.9
Cs
2630.6
Mg I
2781.4
Be
2494.4
Fe II 2631.1
Mg I
2783.0
2332.8
Be
2494.9
Al II
2631.6
Mg II 2790.8
2335.3
2343.5
B I
2496.8
Ba
2634.8
Mp II 2795.5
2347.6
BI
2497.7
Hf II 2641.4. _
Mg II 2798.0
Pd II 2498.8
Au
264L5- Mn I
2798.3
2348.1
2348.6
Zn II 2502.0
Be
2650.3-'
Mn I
2801.1
I

SPEG"TROGRAPiliC ANALYSIS
TABLE II.

Au

~tg

Bi I

II

Mo II

All!

HIll

Ph
Ptl
Phi
OIl
Cdl

Sn I
0
011
Moll
Ir II
Cr II
MgI
~Io

I'd II

on
AlII

Oll
Snl

Moll
Ph
SbI

2802.2
2802.7
2800.6
2816.2
2816.2
2820.2
2823.2

2S3O.3
2833.1

2835.6
2836.9
2840.0

2S4O.0
2843.3
2848.2
2849.7
2849.8
2852.1

2853.2

2&>4.6

2855.7
2S6O.5
2S6O.9
2863.3
2871.5

2873.3

2877.9
288[,6
Btl
MoIl 2891.0
V
2891.7
2892.7
V
V
2893.3
LuIl 2S!>I.9
Hi I
2898.0
2898.3
HrI
AlII
2898.7
Lull 2900.3
HII
200U

Os

290SS
2009.1
2009.1

Hfl

2311.4
2316.5

Moll
LuI!

V
Irl

2924.0
2324.8

Mgll 2928.1
2929.8
Ptl

MDII 2933.1

Mgll 2936.5

Bi I

Ptl
~i

Ni I

Snl

Ni I
0
Fel
Fe I
Bi

Sbl

So I

Nil

Ge I
In I
C<> I
Fe I
Xi I
1\1 I
Ni I

Osl
Ptl
Till
V

Bi I.

Till
HI

Tin
lol
All
Tin
Mgl
All
AI
M~I

VIl
Cbll

Mt l

V I
VII
VII
Ma

.Au

VII
lIg r
BelT
Cb II

toll
gI

Ho II
Bil

2936.S
2938Z

Rull

2!>I0.8
2945.7

Irl

296.'1.6
2967.3
2976.6

Ca II
CblI

Hfl

Mnll mg.2

Ru I!

~h

CAli

POI
Bi I

2980.6
2983.6
2989.0

~\

Hfll

Ti II'
~Io

Snl
Call

2993.3

2998.0

3002.5

3002.6
30091
3012.0
3014.9
3020.7
3021.1
3024.6
3029.8

303t.1
3031.9
303\1.1
303\1.4
30.0
3047.6

3050.8
3().;.I.3
3051.7
305S.7
3064.7

3066.4
3Of.6.4
3057.1
3067.7
3072.1
3072.9
3075.2
3075.9
:lOIl2.2
3088.0
309Ll
3092.1
3092.9
3093.1
3093.1

30!>1.2
3096.9
3102.3
3110.7
3118.4
3122.0
3122.8
3125.3
3125.6
3130.4
3130.8
3131.1
3131.6
313L8
3132.6
3133.3
3134.7

3158.9
3163.4
3168.5
3170.3
3175.0

3179Z

2605

e"""",..<I
VI
VI

VI
Mol
Cbll
YII
Ti I
Ir I
Cb II

3183.4
3184.0
31&>.4
311>1.0
3195.0
3195.6
3199.9
3220.8

3225.5

Sid

3232.5

Till

3232.7
3234.5

Lil

Till
Till
Till

. Pd I
Cui
Cd II
Cd I
In 1
Cd I

Osl
Sn I

Sbl
Os!

Ge I
Zr II
Cu I
Zr II
AgI

ZnI

SA

Zr I
Tbll
Tbll

OsI
I'd

1\"8 I

Zn I
X. I
T .. I
Tb
T.r
&1

Rei

3236.6
3239.0
3242.0
32-12.7
3247.6
3'l5O.3
3252.5
3256.1
326Ll
3262.3
3'.Itl2.3
3267.5
3267.9
3269.a

3273.1
3274.0
3279.3
3280.7
3282.3 .
328.1.5
3284.1
3290.6

3:0).5

3301.6
3302.1
3302.3
3301.6
3302.9
3311.1
3313.7
3318.9
=1.0
=1.1
3321.3
3322.9
3323.1

&1
Till
Rh I
. 3330.0
Sn
La II 3337.5

Till
Zn I
Zn I

Till
Till

So
Ti 1
Zr II
~Io I
CbI

3341.9
3345.0
3345.5
3349.0
3349.4
3352.3
3354.6
3357.3
33.".\

3358.4

Till
Til
Till
Til
:Sir
Ag I

Till
Zr
Co I
Zr
Zr 1I
Xi I
Co I
Rh I
IAI II

Rei

III
Cd I
Pd I
Co
Ta I
C<> I
Zr II
C" I

3361.2
3371.5
3372.8

3377.6
3380.6
3382.9

3383.8
33S7.1l
3383.2
3383.3
3392.0
3393.0

3395.4
3396.8
33\l7.0

3399.3
3399.8
3403.1

340t.6
3-105.1

3406.9
3409.2
3410Z
3U2Z
!'\i I
3414.8
3421.2
Pd I
I'd I
343.1.4
Ni I
3433.6
Rh I
3434.9
Bu I 3436.7
Ir I
3437.0
Ni I
3437.3
Zrll 3438.2
3440.4
K
F. I
3440.6
3441.0
F. I
3441.4
I'd
Mnll 3442.0
Ni I
3446.3
Mo
3441.1
3447.4
K
3452.3
HIl
3453.5
Co I
Mnll 3460.3
3460.8
Pd
3462.2
Tu 1
3462.8
Co r
341">4.5
Srll

Rei
Co r

Cdl

Cd I
Lull
Co I
Mn I!

F. I

Zr
I'd I

3460
3465.8

34&1.2
3467.7
3472.5
3474.0
3474.0
3475.5
3481.2
3481.2

~In II 34!l2.9
Mnli 3488.7

SPECTROGRAPHIC ANALYSIS

2606
Fe I
Ni I
Zr II
Ru I
Er
Ba I
Co I
Rh I
Tb I
IrI
Nil
Pd I
TIl
Zr I
Nil
Co I
Rh I
Co I
Th
Zr I

UI

Pd I
Lu II
Ce II
Tb I
Fe I
Zr II
Se II
Pb
Cr I
Nb I
Se II
Fe I
Ru I
CrI
RuI
Rh
YII
Th I
Zr I
Cr I
Pd I
Cd I
W
Se II
Ni I
Se II
Pd I
Til
Pb I
Til
Se II
Gd I
Fe I
HgI
Se II
Til
HgI
B,h I
Ru I

3490.6
3493.0
3496.2
3499.0
3499.1
3501.1
3502.3
3502.5
3509.2
3513.7
3515.1
3517.0
3519.2
3519.6
3524.5
3526.9
3528.0
3529.8
3538.8
3547.7
3552.2
3553.1
3554.4
3560.8
3561.8
3570.1
3572.5
3572.6
3572.7
3578.7
3580.3
3581.0
3581.2
3593.0
3593.5
3596.2
3596.2
3600.7
3601.1
3601.2
3605.3
3609.6
3610.5
3613.8
3613.8
3619.4
3630.8
3634.7
3635.5
3639.6
3642.7
3642.8
3646.2
3647.9
3650.2
3651.8
3653.3
3654.8
3658.0
366l.4

HgI
HgI
D
Pb
DI
Pb I
Ti II
Pd
Rh I
ErI
YbI
Zr
RhI
YII
Fe I
Fe I
Ru I
RuI
Fe I
Fe I
Pb
Th
Fe I
Fe I
HoI
Fe I
Fe
OsI
Fe I
Ti II
Tu I
Tu I
Fe I
- Gd I
YII
TIl
Hf
Os I
YII
Mol
-RuI
RhI
Sn
Ce II
Ra II
Fe I
Fe I
MgI
MgI
MgI
Co I
Tb I
RhI
D
Fe I
Mol
WI
Co I
Co I
Tb I

TABLE II. Continued

I
3662.9
Fe I
3886.3
Th
411118
Ho I ,3891.0
3663.3
4118.8
\
Co
I
Ba II 389~.8
3670.1
' Co I
4121.3
3671.5
Ti II
3900.5
La II 4123.2
3672.6
Mol 3903.0
leb I
4123.9
3683.5
Si I
3905.5
VI
4128.1
3685.2
Er I
3906.3
Eu II 4129.7
3690.4
Se I
3907.5
Ba II 4130.7
3911.1:)
3692.4
Se I
4132.0
VI
3692.7
Pt
3923.0
Li L
4132.4
3694.2
Ca II 3933.7
.Ce
4133.8
3698.2
All
3944:0
VI
4134.5
3700.9
La II 3949.1
4137.1
Cb I
3710.3
NdI
3951.2
4137.6
Ce
3719.9
Til
3956.3
Ce .
4165.6
3722.6
Til
3958.2
4168.0
DyI
3726.9
RhI
3958.9
4168.0
Pb
3728.0
Al I
3961.5
4172.1
Ga I
3734.9
Til
3964.3
4174.3
Hf
3737.1
Pt
3966.4
4177.3
Nd I
3740.0
Ca II 3968.5
4177.5
YII
3741.2
Til
3981.8
4179.4
Pr I
3745.6
Yb I
3988.0
Ce II 4186.6
3745.9
La II , 3988.5
Pr I
4189.5
3748.2
Ti I
3~89.8
4201.8
RbJ
3748.3
3995.3
Co I
4205.0
Eu I
3749.5
Til
3998.6
4211.1
Rh I
3752.5
Dy I
4000.5
4211.7
Dy I
3758.2
Ag I ' 4212.0
WI
4008.8
3759.3
4012.4
.
Ce II
Sr II
4215.5
3761.3
Th I
4019.1,
Rb I' 4215.6
3761.9
4020.4
Se I
4225.3
PrI
3767.2
4023.7
Se I
-4226.6
GeI
3768.4
MnI 4030.8
4226.7
Ca I
4033.()
3774.3
Ga I
4241.7
DI
3775.7
MnI 4033.1
4246.9
/ Se II
3777.6
MnI 4034.5
I Cr I
4254.3
3782.2
Ce II 4040.8
4260.9
Os I
3788.7
4044.2
KI
4267.1
CII
3798.3
~Fe I
4045.8
4267.3
CII
3798.9
Dy I
4046.~)
4271.8
Fe I
4046.(j
3799.3
HgI
4274.8
CrI
4047.2
380~.0
KI
Th""_ 4282.0
3801.5
4057.R
Pb I
Ca I 4283.1
4059.()
3814.4
Cb I
4289.4
Ca I
3820.4
Pr I
4062.8
4289.7
Cr I
4063.(j
3827.8
Fe I
4294.6
WI
4074.1
3829.4
WI
4302.1
WI
3832.3 ,
La II 4077.4
4302.5
Ca I
3838.3
4077.1'
Sr II
4302.1
Nd
4078.()
3842.1
DyI
4305.9
Ti I
3848.8
4090.8
D
4307.9
Fe I
3856.5
HfI
4093.2
4314.1
Se II
3859.6
VI
4099.8
4318.7
Ca I
410U)
3859.9
Cb I
Ce I 4320.7
__ Se II
3864.1
In I
410l.8
4320.8
3868.0
VI
4105.2 ~
4325.0
Sen
- Fe I
3873.1
Co I
4110.5
4325.8
3873.9
VI
4111.8
La
4333.8
4116.(
3874.2
VI
La
4334.0

SPECTROGRAPHIC ANALYSIS
U

HgI
Se II
Rh I
YII
VI'
Mo
Th
Ce
Fe I
VI
VI
Sa I
.Th
VI
Se II
VI
Fe II
VI
VI
VI

Mo
Os I
Sa I
Ca I
La
Sa I
. Mo
Ca. I
Ca'I
Nd
Nd
Ca I
Ca I
Ca I
VI
Nd
U

Ba I
In I
Re I
Lu I
Ce
Sn I
Ce
Ce
Rh
Til
Ti I
Ti I
Ti I
Ti I
Ce
Ba II
Ru
Cs I
Ce
MgI
Rb
Ce

4341.7
43"58.3
,4374.5
4374.8
4375.0
4379.2
4381.7
4381.9
4382.2
4383.5
4384.7
4390.0
4390.9
4391.1
4395.2
4400.4
' 4400.6
4404.8
4406.7
4408.2
44-08.5
4411.7
4420.5
4424.4
4425.4
4429.9
4434.4
4435.0
4435.0
4435.7
4446.4
4451.6
4454.8
4455.9
4456.6
4460.6
4463.0
4472.3
4489.0
4511.3
4513.3
4518.5
4523.1
4524.7
4528.4
4528.5
4528.7
4533.3
4534,8
4535.6

!~~~:i

4539.7
4554.0
4554.5
4555.3
4562.4
4571.1
4571.8
4572.3

TABLE II.
Be I
Cs I
Ce
VI
Li I
Sr I
Ce
YI
Ag I
YI
Ag I
YI
Cd I
Zn I
RaIl
Zr I
SI
SI
SI
MgI
MnI

Zr I
Ni I
Zn I
BiI
MnI
- Se I
Se I
Zr I
Se I
MnI
MnI
Zr I
MnI
Pd
Au
Co I
Os
Cd I
Zn I
Co I
Zr I
Pd
MnI
Ra I
Sr I
VI
Th
VI
Co I
Ru
Sr I
Pd
VI
VI
Re I
Th
La
Ba II
Sr I

4572.7
4593.2
4593.2
4594.1
4603.2
4607.3
4628.2
4643.7
4668.5
4643.7
4668.5
4674.8
4678.2
4680.1
4682.2
',1687.8
4694.2
4695.5
4696.3
4703.1
47V9.7
4710.1
4714.4
4722.2
4722.5
4727.5
4730.9
4739.1
4739.5
4742.3
4754.1
4762.4
4772.3
4783.4
4788.2
4792.6
4792.9
4794.0
4799.9
4810.5
4813.5
4815.6
4817.5
4823.5
4825.9
4832.1
4851.5
4863.2
4864.8
4867.9
4869.2
4872.5
4876.4
4875.5
4881.6
4889.2
4919.8
4921.8
4934.1
4934.1

2607

Continued

Li
NaI
Til
Til
Til
Til
Til
Ti I
Th
U

Se
Th
W

Ce
Ni
Ni
Se I
Se I
Cd I
Cu

C-u I

Pd
MgI
Ru
MgI
MgI
Cr I
Cr I
Cr I
Ag I
Cu I
W

Fe
Pt
Fe I
Fe I
Pd
Pt
Fe I
TIl
Pt
Fe I
Pt
Cr I
Ba 1
La I
HgI
Ag I
Ag I
Pt
Nil
Pt
W
U.

Mol
Ce

Ba I
Se

4971.9
4979.0
4981.7
4991.1
4999.5
5007.2
5014.3
5016.2
5017.2
5027.4
5031.0
5049.8
5053.3
5079.7
5080.5
5081.1
5081.6
5083.7
5085.8
5105.6
51.S3.2
5163.8
5167.3
5171.0
5172.7
518H
5204.5
5206.0
5208.4
5209.0
5218.2
5224.7
5227.2
5227.6
5232.9
5269.5
5295.6
5301.0
5328.0
5350.5
5369.0
5371.5
5390.8
5409.8
5424.6
5455.1
5460.7
5465.4
5471.5
5475.8
5476.9
5478.5
5492.3
5492.9
5506.5
5512.1
5514.7
5519.1
5526.8
5527.8

MgI
Mol
BaI
Pd
Pd
Bi
Mol
Ca I
Pb
Se
NaI
Se
NaI
Cu I
HgI
Ba I
Cu I
KI
HgI
K

5528.5
5533.0
5535.5
5542.8
5547.0
5552.2
5570.5
5590.1
5608.8
5671.8
5682.8
5686.9
5688.3
5700.3
5769.6
5777.7
5782.1
5782.6
5790.7
5801.9
Ta
b81l.1
5812.4
K
Au
5837.4
Pt
5840.1
Mol 5858.3
Mol 5888.3
NaI 5890.0
Na I 5895.9
.5915.4
U
LaI
5930.6
Pb
6002.0
MnI 6013.5
MnI 6016.6
. MnI 6021.8
Mol 6030.7
LiI
6103.6
Ca I 6122.2
Ba II 6141.8
Ca I
6162.2
AlII 6231.8
Al II 6243.4
La I
6249.9
Se I
6305.7
. Zn I
6362.4
Ta
6430.8
Cd I 6438.5
6439.1
Ca I
6449.2
U
6462.6
Ca I
Ta
6485.4
6493.8
Ca I
Ba II 6496.9
Ta
6516.1
6707.9
Li I
KI
6911.3
6939.0
KI
KI
7664.9
KI
7699.0
Rb I 7800.3
Rb I 7947.6

2608

SPECTROGRAPHIC ANALYSIS.
I

FIG. 509.

Arc Spectra of Copper and 13ras&.Samples

.

_r-

SPECTROdRAPHIC ANALYSIS

2609

ableait the present time is the Eder and Valenta atlas, which includes arc, spark
and flame spectra of nearly all of the available elements, taken with glass prism,
quartz prism and grating spe<;trographs. The dispersion, h~wever, is not
satisfactory for the accurate identification of many elements, especially where
the spectrum is complicated. Adam Hilger have published several photographic tables, one of which gives the persistent lines of a large number of
elements. They also have published iron, copper and neon spectra in which
the dispersion is high and a large number of the lines identified. To further
aid in the identification of other elements, tables have been prepared in which,
in addition to the iron spectrum, there is indicated the position at which a
strong line of some other element would appear if it were present. Tables of
this type have been published by Bardet, Crook and others. From the tables
in this' chapter and with the use of an accurately calibrated scale on the instrument or with an iron spectrum in contact with the unknown spectrum and a
photographic identification of the iron spectrum, one can satisfy nearly all the
requirements of the qualitative analysis of elements 'by the spectrographic
metho.d.

QUANTITATIVE ANALYSIS

de Gramont has proposed that the.ultimate lines or the Raies Ultimes of the
elements are to a certain d~gree the basic lines of the element and as such
their intensity will be proportional to the amount of the element present. The
first practical application of this theory to quantitative analysis was made by
Meggers Kiess and Stimson in the analysis of platinum metals. In all of the
suggested' methods of quantitative spectrographic analysis the results are
dependent on the known chemical composition of comparative standards.
One might subdivide the me_thods into those in which the comparative standard
is observed under similar conditions to that of the unknown observation, and
those in which the standard substance is added to the unknown so that both
photographs are made simultaneously.
.
Comparison with Standards.-tJertain of the tentative standard and
recommended methods of analysis of materials by the spectrographic method
involve the comparison of line intensities between standard photographs and
unknown photographs taken under similar conditions. If the conditions of
exposure, development, slit, current density, and type of plate are maintained
it may not be necessary to repeat the photograph of the known standard bu~'
merely to compare with a previously prepared standard negative. A certai:'l
amount of possible error can be avoided by the taking of a photograph of the
standard sample adjacent to the unknown sample on the same negative. In
many cases this method of comparative analysis is satisfactory. On account
of the large number of variable factors which exist it is not possible to supply

SPECTROGRAPHIC ANAL Y;SIS

2610'

standard spectra or spectrum plates for the comparison measurements, so that

each worker must provide his own comparative standar8s.
Internal Standards.-The satisfactory results obtained in the analysis of
substances by comparison photographs has led to thJ improvement of the
,I
-3.0

-5,0
c

Vl
I

co
L9G SECTOR METHOD

-7.0

%
0.50
1.00
SJ Log %
0.75 -0.00
-0.50
-1.00
-1.50
FIG. 510. Working Curve for the Analysis of Bismuth, Using Tin as the In
ternal Standard. Determinations made by the Log Sector Method. Bi=2989.04 A
Sn=3009.14 A.

~o

-.04

gjl..,08

,_J

PHOTOMETRIC METHOD
-:12

Bi

%
log % -1.50

0.25

-1.00

0.50

-0.50

0.75 I -000
.

,-_,'

1.00

FIG. 511. Working Curve for the Analysis of. Bi~ifnuth, Using Tin as the Internal Standard. Determinations made by the PhotrQnic Cell Densitometer Method.
Bi=2989.04A, Sn=3009.14A.

SPECTROGRAPHIC ANALYSIS

2611

method by the incorporation of a standardizing substance with the material

which is being analysed. This procedure eliminates many of the possible
sources of error, since the two exposures are made simultaneously. The method
of internal standards requires the preparation and study of a series of known
mixtures and the measurement of the relative intensities of known and unknown lines. The logarithm of the ratio of line densities, plotted against the
log of the concentration of the element to be analysed for, should yield a working
diagram from which the concentration in an unknown sample could be determined. In Figs. 510 and 511 and Table III are given sample working curves.
TABLE III
REPORT ON WORKING CURVE DATA

Spectrographic Analysis for Bismuth

Internal Standard = Tin
Log Sector Method
Sample

%Bi

Log %
Bi

%
Sn

Len(!th
Bl
2\\8\\.04

Length
Sn

3()()\\.H

Bi-Sn

Photometer Method

Bi

2989.04

Sn
3009.14

Log Bi/Sn
Bi/Sn

---- ---_--------- - -----1

0.000
2
10.05
0.000
1.0
11.5
3
4
5
6
7
8

0.031
0.063
0.125
0.250
0.500
1.000

-1.502
-1.204
-0.903
-0.602
-0.301'
0.000

1.0
1.0
1.0
1.0
1.0
1.0

3.6
4.0
5.0
5.8
7.3
8.5

11.5
11.5
11.4
10.7
11.3
11.3

-7.9
-7.5
-6,4
-4.9
-4.0
-2.8

7.5
7.85
8,40
8.88
9.3
9.65

10.18 0.736
10.03 0.783
10.10 0,832
10.0
0.888
10.0 +0.930
10.0
0.965

-0.1331
-0.1062
-0.0797
-0.0516
-0.0315
~0.0155

The choice of line pairs for comparison will of course determine the concentration range over which the data will best apply. Over the major portion of the
working curve an accuracy to within about 5% of the amount of 'unknown
element can be determined. For high concentration the. method may be used,
although this same percentage error will still hold and is usually much greater
than that obtained by other analytical methods. For extremely small amounts,
however, an error of 5 or even 10% of the amount present may be quite permissible.
Measurement of Intensities of Lines.-The determination of the blackening
or line densities on the. photographic plate may be photometrically measured
by visual or electrical means. A number of recording and semi-automatic
densitometers for the measurement of plate and line density are available for
this purpose. These instruments may have certain constant errors such as the
failure to obtain a linear response from the photo cell over a range of varied
light intensities. A combination of densitometer and spectrograph can be
effected by placing directly in front' of the spectrograph slit a rotating sector:
cut either smoothly or in steps and usually cut in a logarithm spiral or in
logarithmic steps, or by the use of a step slit such as the Hansen or Duffendack
type. This procedure produces a spectrogram in which the line tapers in its
degree of blackening until it fades completely out, or is reduced by steps to

2612

SPECTROGRAPHIC ANALYSIS

complete extinction. In the step system a photometer may!be applied to the

lines at different densities. The continuous line (Fig. 512) produced by the
logarithmic spiral sector may be measured as to length with a micrometer
eyepiece. The difficulty of deciding on the point of absolute extinction or

..

FIG. 512.

Typical Spectra by Log Sector and Ordinary Spectrum Photographs

_ Used in the Preparation of Working Curves.

termination of the line may be eliminated in part by the }l of a reversed line

. in the eyepiece and matching the two lines to a uniform,crensity.
, In methods where comparison is to be made between spectrograms, especially where the spectrograms are on different plates, it is somewhat important
to determine the plate characteristics and many workers feel that this procedure
should be carried out when the data are all on the)lame plate. The variation

SPECTROGRAPHIC ANALYSIS

2613

of the plate characteristics (i.e. the d/log I curve) with different spectral regions,
nature of exposure, development time and temperature, etc., may require that
for accurate work the plate characteristics should be determined through the
use of a standard lamp for illumination of the spectrograph and that correction
be made so as to indicate the absolute intensities of the lines. In routine
analysis where the same plate treatment is used, the same type of plates and
photometric methods, it is possible to prepare working curves which will give
accurate quantitative results.
Control of Conditions.-An important working condition in quantitative
spectrographic analysis which must be carefully controlled involves the light
source. From a study of the cause and nature of spectrum lines it will be
noted that there are successive stages of ionization of the element, which are
dependent on the applied voltage and temperature. The spectrum produced
under 10':' io:t;lization conditions may be quite different in the lines present and
their intensities as compared with the spectrum from the highly ionized
element. This change in relative intensities of the emission lines with a change
in exciting conditions requires a careful duplication of conditions when making
a series of observations, and requires a rather exact definition of the exciting
conditions if one is to apply the working curve or analytical data prepared and
published by others. Gerlach has attempted to define the working conditions
by which repeated observations could be made, through the selection of lines
of the known, element from different states of ionization, which at the given
conditions of observation are equal in intensity. When they are equal the
excitation can be said to he at the standard or specified condition. Gerlach
has then varied the amount of "unknown" dement with respect to the internal
standard (the internal standard may be the element present in largest amount
or an added standard of known amount) and selected for each concentration
pairs of lines of the standard known and unknown which are of equal intensity.
These pairs are known as homologous pairs. In view of their equal intensity
it can be seen that these results will not be appreciably influenced by plate
characteristics and other errors which have been mentioned. If the internal
standard is to be some element other than the substance present in largest
amount (the pure substance containing the impurity for which the analysis
is being made) it will be necessary to control the amount of internal standard
which is added. This is accomplished by choosing a pair of lines, one from the
pure or base substance and the other from the internal standard, 'which are
equal in intensity under the excitation and concentration conditions which are
used for the observation. This method may be better understood from Table
IV whieh is a portion of an analyses table published by Scheibe.
TABLE IV
Quantitative analysis for Titanium in an Aluminum sample. Control of excitation
conditions Al II 2616.18=AI I 2575.11 A. Internal standard control or coupling,
pair Al I 2652.48=Cu 2618.38 A.
Wavelength
Titanium

Ti II
Ti II
Ti II
Ti II
Ti II

Wavelength
Copper

3361.22 .............. 3307.95

3361.22 .............. 3337.85
3335.19 ..... , ........ 3349.26
3361.22 ..... , ....... .4022.70
3361.22 ............. :3365.36

Percent of Ti in Sample when

Lines Are of Equal Density

0.26
0.204
0.188
0.166
0.109

2614

SPECTROGRAPHIC ANALYSIS

It would seem that while this method may have eliminated a number of the
possible errors of other methods it involves the setting up ofidefi:tlite conditions
of excitation, internal standard concentration and line intensity mealffirement,
which in the end are required of nearly all methodFl whichl give reproducible
results. In the method described by Gerlach the spectrum of the so-called
internal standard may be taken separately, either on top lof or adjacent to
the sample to be analysed and need not be that of a mixture of the sample and
internal standard, just so long as the homologous and coupling or control pairs
are equivalent in density.
,
In order to control more carefully and standardize the spectrographic data
in an analysis, Duffendack and his coworkers have developed the use of plate
calibration through the use of a step slit so as to remove possible errors which
might 'result in the failure of the negative to yield a linear d-Iog I curve and
of the same slope for all wavelength values,
.
It would seem that while the method of quantitative spectrographic analysis
has been rather carefully studied and applied to a large number of prQblems
with considerable success, it is not possible to give such specific data as is available for qualitative spectrographic analysis in the w!1velength tables. Because
of.the nature of the method it is necessary in nearly all cases to prepare standard
negatives, mixtures or working curves which apply to the particular spectrograph, photographic plate, photometer or other influencing factors. The value
of published work in the field of quantitative spectrographic methods lies,
largely in the development of technique and the experirnental determination of )
If
satisfactory groups or pairs of lines for specific analytical purposes.
Plates.-For the usual type of qualitative or quantitative spectrographic
analysis a medium speed photographic plate such as the Eastman 33 or 40 is
quite satisfactory. Development can be by elon-hydroquinone (metol-hydroquinone) or other standardized developer. For spectral regions 'beyond 5000 A
panchromatic or special red sensitive negatives are available to about 1,4,000 A.

,
ABSORPTION SPECTRA
When an obj~ct or solution absorbs a portion of the homogenous spectral
radiation which passes through it we describe the J;Ilaterial as being colored.
The absorption of homogenous spectral radiation by molecules is usually quite
selective and limited to definite spectral regions, or bands. These absorption
bands are usually diffuse, with partial absorption ove~ a considerable region
and a gradual increase in absorption to a maximum of absorption near the
center of the absorption band. The intensity and position of the absorption
maximum, as well as the rate of increase in absorption on eag}l side of the
maximum are definite properties of the compound being_exammed and may
serve to identify or characterize the substance. WithIn. the visible spectral
'region one might indicate the production of color by absorption bands as follows:

SPECTROGRAPHIC ANALYSIS

2615

TABLE V
'RELATION BETWEEN ABSORPTION BAND ,POSITION AND OBSERVED COLOR

Position of Absorption Band

Color'Absorbed

'450 filt. . . . . . . . , .. Blue

500 ............... Blue-Green
550 ........ , , , , . , . Blue-Green-Yellow
600 ..... , , . , . , , ... Green-Yellow
650 ..... , .. , .. , . , . Yellow-Orange
700 .... ,., .... ,., . Red
400 and 750 ... , ... Blue and Red

Color Transmitted

Yellow
Orange
Red
Red and Blue
Blue
Blue-Green
Green

Since absorption bands are usually not well defined like emission lines it is
the practice to record wave length values in millimicrons (mtL=metersXlO-9),
For work in the ultraviolet region the data is often recorded in frequency or
wavenumber, since absorption bands represent in many cases harmonic
vibrations which give regularly spaced bands on a frequency scale.
The intensity of absorption is an important value for the quantitative
estimation of the absorbing material. Where one is interested in a description
of the transmitted color one usually records the per cent of light transmitted
~"Ot variobs wavelength values. If one is interested in the amount or character
of the absorbing substance one usually records the extinction coefficient. The
extinction coefficient may be defined as the Log of the incident intensity divided
by the transmitted intensity (log 1 0 /1). By recording in extinction coefficient
values, the height of the absorption band is directly proportional to the con-
centration or cell thickness (assuming that Beer's law holds for the material).
,From the extinction coefficient of the maximum of the absorption band of a
substance one may obtain a much more accurate value for the concentration
... than through a colorimetric measure of the transmitted light. In addition, the
position and shape of the curve may enable one to identify the molecule producing the absorption. The production of absorption is dependent on ,the
I
presence in the molecule of an absorbing or chromophoric group. Not all
'substances absorb radiation within the visible or photographic ultraviolet
region, and hence it is possible to detect and identify traces of absorbing
materials in otherwise transparent materials. For example, it is possible to
detect and quantitatively estimate the amount of benzene present in ethyl
alcohol, with an accuracy of about 10% of the amount present, when the
concentration may be as low as 0.001 %. The spectrophotometer has been
used in the quantitative estimation of such colored inorganic substances as
cobalt, iodine, copper, and the rare earths (Fig. 513). For the purpose of
estimation of impuriti~s andtheir detection one may use the'method of subtractive absorption in which one subtracts the absorption band of the known
pure ,substance from the observed band, the difference representing the impurity
absorption (Fig. 514). Commercial dyestuffs are standardized and their
concentration determined by absorption spectra methods which give rapid and
accurate quantitative results. Much of the synthetical work in the fields of
vitamins, hormones and natural pigments is based on the absorption spectra
data obtained by the study of these and related compounds.
The data in the field of absorption spectra measurements is so varied that
it is not possible to give methods of procedure and lnterpretation of data. For
a more detailed discussion of these fields one should consult the references in the
bibliography which follows.
I

2616

SPECTROGRAPHIC ANALYSIS

~
::: 1.00

!.,-----+---,_,J-+--i-_.,.

700
Wavl>

j&ngth, m p.

800

FIG. 513. The Absorption Spectra of Cobaltous Halides in their Corresponding

Halogen Acids - - - CoCh in HCI; ----- CoBr2 in HBr; - - - CoI2 in HI.
(Brode,.J. Am. Chern. Soc., 53, 2457 (1931).)
/
000

o.20

-_ -_

o.40

/.

"

",

"

o.60

1.20

1.40

.60

0.90

o.so

0.70
0.60
0.50

,f

...

<7

0.40

1/

0.30

//

.so
.00

'f

1 .00

,:/ l7

0.20
0.15

1.

I/
/

~\

~
.~

0.100 E

~:g~g ~

/'

\ \ __ .I 1/

0.070 ..:
0.060
0.050

0;040_

\J /

0.030
0.020

1.S0
00

0.015

400 420 440 460 480 500 520 540 560 580 600 620 640 660 ~80 700 :720 740 760

Wave

0.010

Length

FIG. 514. Identification of Impurities in Commercial Dyestuffs. (1) Mother

liquor absorption from a 16% dyeing of a commercial sample of Agalma Black lOB
whose shade was rated as RRR by dyeing tests. (2) Absorpton spectru~' of the pure
dye. (3) Difference curve or. absorption spectrum of impurity; - (Note. Data are
recorded in -log Transmittancy which is the same as extinctiorf'f<)efficient but plotted
as'a negative value.) (Brode, Ind. Eng. Chern., 18, 708 (1926).)

SPECTROGRAPHIC ANALYSIS

2617

Special Note-Patent _Situation.-Users of spectrographic methods of

analysis, including both for non-profit research as well as commercial applications, should be cognizant of the existence of U. S. Patent No. 1,979,964
(application made March 14, 1932 and granted Nov. 6, 1934) by Duffendack
and Wolfe. The claims include the following:
"1. The method of quantitative spectroscopic analysis which comprises the following steps: establishing an arc current between electrodes comprising the unknown
material of value such that the intensity of spectral lines does not vary materially with
unavoidable fluctuations in arc currentl recording the spectrum of the arc, measuring the
ratio of intensities of a pair of lines in the spectrum by comparison of the strengths of the
corresponding recorded lines, repeating the process with a series of electrodes containing
the same kind of material but of known analysis, making use of the same spectral lines,
evaluating the ratio of intensities obtained from the unknown material in terms of
percentage of an element producing one of the lines by comparing the ratio of intensity of
the unknown with ,the ratio of intensity of said known specimens."

In following claims the use of a spark and luminous discharge are covered
as well as the recording of the spectrum on a photographic plate. In view of the
large amount of previously published work in the field of spectrographic quantitative analysis, much of which is referred to in the patent, it seems unfortunate
that a patent claim should be granted which rather effectively covers the
fundamental principles involved in this analytical field. It seems unfortunate
that basic claims in patents should be granted for analytical methods, especially
if the methods have been widely used and published in the literature.
References
Comprehensive discussions of Spectrographic AnalysiS with an extended list of
Bibliographical references are to be found in the following:
1. "Chemical Spectroscopy," W. R. Brode, John Wiley and Sons, New York (1939).
2. "Spektroskopische und Radiometrische Analyse," Scheibe, Akad. Verlag., Leipzig
(1933).
3. "Metallurgical Analysis by the Spectrograph," D. M. Smith, British Non-Ferrous
Metals Research Association, Monograph No.2. (1933).
4. "Die Chemische Emissions Spektralanalyse," Gerlach et al. Verlag Voss, Leipzig,
3 Vol. (1929--1933-1936), English Translations of Vol. 1 and 2 published by Adam
Hilger, London.

Analysis Tables:
Kayser, "Handbuch Der Spectroscopie," 8 Vol., Hirzel, Leipzig.
"International Critical Tables," McGraw Hill.
Kayser, "Hauptlinien der Linienspektra aller Elemente," Springer, Berlin (1926).
Smith, "Visual Lines," Hilger, London (1928).
Twyman and Smith, "Wavelength Tables for Spectrum Analysis," Hilger (1931).
~General:

Bardet, "Atlas de Spectres d'arc," Doin, Paris (1921).

Baly, "Spectroscopy," 3 VoL, Longmans Green (1927-29).
Lewis, "Spectroscopy in Science and Industry," Biackie, London (1933).
Lundegardh, "Die Quantitative Spektralanalyse der Elemente," 2 Vol., Fischer,
Jena (1929--1934).
Twyman and Allsopp, "Spectrophotometry," Hilger (1934).
Crook, "Metallurgical Spectrum Analysis," StanfoI:d Univ. Press (1934) .
.Spectrographic catalogs from Adam Hilger, 24 Eochester Place, Camden Road,
London, N.W. 1, England; The Bausch and Lomb Optical Company,
Rochester, N. Y.

INDEX TO AUTHORS *
A
Abady, J., 2391
Abbe, B., 17M, 17\)5
Abderhalden, E., 2503
Abegg, R., 2591
Abraham, H. A., iii, v, 1510
Abria, J. J. B., 1186
Acker, C. E., 961, 963
Acree, S. F., 1919, 2285
_<\.dams, R., 952, 1102
Addie, R. H., 874
Adkins, L. R., 2546
Adler, S., 2491
Adolph~ W. H., 409
Agricola, G., 399
4gulhon, H., 163, 185
Ahlum, C. A., 2245
Aisinman, S., 1737
Alexander, P., 1975
Alexejeff, W., 2582
Allart, A., 966
Alldredge, S. M., 368, 503, 504, 527
Allen, A. H., 656, 1764, 1771, 1777, 1781,
,1787, 1876, 2147
t Allen, l!l. T., 179, 180, 905, 917
Allen, r. C., 1717
Allen, N., 4Q6, 2198
Allen, W. S., v, 640, 909
Alling, H. L., 2453
Allison, V. C., 2361
Allsopp, C. B., 2617
Alsberg, C. L., 226, 635
Alten, F., 1217
Aluminum Company of America, iii, 1
American Brass Company, iv
American Chemical Society, v
American Cyanamid COIJ1pany, iv
American Society for Testing Materials,
iii, iv, v, 1011, 1071, 1421, 1485, 1597,
1620, 1705, 1713, 1719, 1727, 1732,
1737, 1743, 1750, 1761, 1807, 1814,
, 1818, 1823, 1830, 1839, 1850
Amsbary, F. C., 2115
Amsel, H., 1784
Anderson, R. P., 686, 1818, 2345, 2355,
2405
Andrew, R. L., 460
Andrews, L. W., 194, 269, 455
Andrews, T. M., 1185, 1880
Angstrom, A., 2592
Anneler, E., 2370

Ansbacher, S., 2520

Appelyard, G., 2404
Archbutt, L., 1726, 1780, 1784
Archer, R. S., 2558, 2564
Aristotle, 574
Arkel, A. E. van, 1104
Arnold, C., 632
Arnold, F. W., 1216
Arnold, J. 0., 52
Arrhenius, S., 1191
Association of Official and Agricultural
Chemists, v
Astin, S., 1218
Atack, F. Vi., 2, 311, 617, 618
Atkinson, A. J., 1642
Atwater, C. G., 1654, 1834
Aubert, A. B., 1789
Auerbach, F., 2310
Austen, L., 435
Austin, M., 560
Autenrieth, W., 1945, 1950
Avogadro, A., 1191
B
Bach, 0., 1781
Backstrom, S. A., 2318
Badger, W. L., 2369
Baggesgaard-Rasmussen, H., 2310
Bahlmann, C., 2081
Bahr, F., 2578
Bailar, J. C., 213, 2270
Bailey, A. J., 2526 .
Bailey, E. G., 1318
Bailey, H. S., 1780
Bailey, R. K., 872
Bainbridge, 1186
Baker, H. A., 957, 964, 971
Baldsiefen, W. D., 1780
Ball, T. R., 984
Ballantyne, H., 1767, 1769
Baly, E. C. C., 2617
Bamberger, M., 1893
Banks, R. M., 97Barbieri, G., 320
Bardet, J., 259, 2609, 2617
Bardwell, E. S., 2453
Bardy, C., 2147
Barker, J. F., 242
Barlow, H. C., xiii, 305
Barnard, J. E., 2457
Barneby, O. L., 978

* Page numbers gretlter than 1300 are in Volume II.

Small Roman numerals in italics refer to the list of authors in Volume II.
1

INDEX TO AUTHORS

. Berthelot, M. P' 1186, 1188, 1190,

Barnes, J. W., 2367
\.
2137, 2 3 7 6 ,l
Barnes, O. A., 2111, 2115
Berthollet, C. L., 2,375
Barnes, S. K., 140, 148
Bertrand, E., 309 i
Barneveld, C. E. van, 393
Bertrand, G., 163,' '185
Barnitt, J. B., v, 1301
Beshgetoor, A. W.,l iv, 539, 2187
Barnstein, F., 2458

Berzelius, J. J., 775, 785, 794, 803, 2365

Barton, L. E., 656, 991, 992
Bettel, W., 439
Battle, J. R., 1789
Betts, A. W., 650, 652
Baubigny, H., 276
Betz, K., 2322
Baudisch, 0., 6, 7, 468, 469
Betz, L. D., 2092
Baudoin, 1781
Betz, W. H., 2092
Baughman, W. F., 194, 459
Bevan, E. J., 1913
Baule, B., 2464
Bezzenberger, F. K., 50
Baume, A., 2411
.
Biazzo, R., 2519
Bawden, A. T., 2320, 2321
Bidtel, E., 419
Bayer, A., 874
Bidwell, G. L., 1340, 1342, 1343, 1345
Beadle, C., 1973
Bienfait, H., 2322
Beamish, F. E., 2496, 2503, 2504
Bierer, J. M., 1996
Bearwood, J. P., 310
Biginelli, Po, 1805
Beauverie, J., 2458
Biilmann, E., 2301
Becher, J. J., 2375
Biltz, H., 69
Bechi, E., 1777, 17-78
Biltz, W., 1094
,
Beck, C., 2457
Bingham; E. C., 1725
Beck, L. C., 1619
Bird, C. D., 56
Becker, F., 1670
Bisbee, H., 360
Beckstrom, R.' C., v, 1806
Bischowsky, F: R. von, 2314
Beckurts, H., 1950
Bishop, E. R., 2312
Bedford, C. W., 1998
Bishop, H. B., 909, 2242, 2243
Beeson, F. M., 2073
Blacet, F. E., 2537, 2544
Behrend, R., 2313
Blacher, C., 2082
Behrens, H., 2439, 2457
Black, J. W., 618
Bekk, J., 276
Blair, A. A., 36, 232, 244, 560, 996, 1407
Bekkedahl, N., 2287
Blake, J. C., 193
Belasio, L., 3, 53
Blake, J. T., 1998
Bell, J. M., 2582
Blalock, T. L., 1223
Bellier, J., 1780
Blank, E. W., 2456
Benedetti-Pichler, A. A., 2322, 2,1,57,
Blasdale, W. C., 360
2464, 2517, 2521, 2522, 2546, 2547
Blattner, N.; 274, 275
Benedict, F. E., 2403, 2405
Blay, V. L.,12.202
Benedict, R., 1893
Bloomfield, J. J., 2373
Benedict, S. R., 1908
Bloxam, A. ,G., 1957
Beneker, J. C., 234
Blum, W., 3, 60, 1360, 2287
Benner, R. C., 824
Blumenthal, 'M., 2313
Bennett, H. C., 2031
Bobranski, B., 2533
Bennett, J. F., 100
Bodansky, M., 1068, 1084, 1685; 1392
Benton, A. F., 234;2'
Bomer, A., 1775
Benton, A. G., Miss, 2288
Boer, J. H. de, 1104
Berg, P. von, 203
Boerner, E. G., 1341, 1345
Berg, R., 12, 157, 455, 531, 535, 866,
Bornstein, R:; 2411, 2412
2521
Boetius, M., 2481
Bergey, D. H., 2333
Bott'ger, W., 273.; 2309
Bergman, T., 2365
Bogojawlenski, A., 1189
Beringer, C., 374
Bogue, R. H., 1618
Beringer, J. J., 374
Bohn, R. T., 2315
Berk, L. H. van, 457
Boldyril:ff, A. W., 211, 2319, 2320
Bockeley, Earl. of, 2573, 2574
Bolliger, A., 1217
'I
Berl, E., 2198
Bonardi, J. P., 2451
Berman, H., 2435
Bond, F. C., _827- .
Bone, A. J., 360Bernard, E., 633, 2077
Bonner, W. D., 1144, 2196, 2197, 2208
'Bernini, A., 1188

INDEX TO AUTHORS
Booth, H. S., 242Q
Booze, M. C., 2451
Bosch, C., 442
Bosworth, R. S., 823, 832
}3ower, W. H., 1143, 1146, 1151, 2207,
2239
Bowman, F. C., 2214
Bowser, L. T., 874
Boyd, H. T., 2408
Boyer, W. J., v, 1480
Boyle, R., 1191, 2243
Boys, C. V., 2391
Brackett, J. H., 931
Bradbury, W. A., 2266
Bradley, C. S., 2379
Brady, W., 2389
Braidy, H., 1914
BraiIIier, P. S., 481
Brand, H., 689
Brandonl A. L., 2494
Br~ndt, G., 305
Brannt, W. T., 1741
Brasseuer, J., 274, 275
Bratton, A. C., 2534
Bratton, W. M., 527
Bray, 2391
Bray, M. W., 1880, 1893
Bray, W. C., 429, 430, 434, 46~, 647,
893
Brearley, H., 589
Bre~her, C., 2527
Brennecke, E., 1205, 1206
Brewer, F. M., 429
Brewster, R Q., 266
Breyer, F. G., 1054, 1057, 1062
Brickwedde, F. G., 444
Bridges, R. W., iii, xiii, 1, 5, 8, 37, 50
Bright, H. A.,,297, 626, 1209, 1421, 1615
Brinton, P. H. M-P., iii, xiii, 245, 249,
252, 591, 946, 947, 1093, 1097
Britton, H. T. So, 148, 2309, 2328
Broch, O. J., 1160
Brode, W. Ro, iv, V, 1914, 2592, 2616,
2617
Brodhun, 2391
Briinsted, J. N., 1189
Brooks, C., 2366
Brooks; S. Ho, 2368
Brophy, 'D. Ho, 953 (see also Hall, D.)
Brown, A. S., 2300
Brown, A. W., 59
Brown, E., 1341, 1345
Brown, M. H., 892
Brown, W. A., 112
Brown, W. J., iii, v, V, 502, 504, 527, 974,
1348
Browne, F., 1847, 1848
Browning, P. E., 148, 252, 401, 481, 767,
786
Brownlie, D., 2386
Bruce, W. F., 2469

Briinni~h,

J. C., 2320
Brukl, A., 427, 943
Brunck, 0., 619, 2401, 2521
Brush, W. W., 2115
Buch, N. W., 1644
Budde, T., 1975
Buehrer, T. F., 481
Biilow, B. F. von, 108
Bugarsky, S., 194
Bugbee, E. E., 857
Bull, I. C., 508
Bullock, B., 2499
Bunsen, R, 1189, 2357, 2587, 2590
Bunte, H., 2342, 2383, 2396
Burford, M. G., 270, 577, 960, 1335
Burgess, G. K., 975, 1140
Burgess, R., 314
Burkhard, W. E., 2202
Burnes, J. W., 108
Burrell, G. A., 2359, 2368, 2377, 2400,
2402, 2403, 2407, 2408
Burton, J. 1., 532, 696
Burton, J. 0., 1913, 1914, 1919
Busch, M., 635
Bushnell, L. S., v, 939
Bussy, A. A. B., 137
Busvold, N., 653
Butzler, E. W., 2369
Byrd, R M., 882
Byrne, P. J., 2374

o
Cady, H. P., 2379
Cady, L. C., 543, 1065
Cagliotti, V., 148
Cahen, E., 976
CaiIletet, 1769
Cailletet, L. P., 2590
Cain, J. R, 231
Cake, W. E., 802, 1479, 1615
Calbeck, J. H., 527, 1870, 1874
Caldwell, J. R., 272, 311, 2498
Caley, E. R., v, 270, 538, 577, 875, 879,.
892, 960, 1335
Calhane, D. F., 2322
California Portland Cement Co., V
Callan, T., 376, 2017, 2338
Cameron, F. K., 874, 2098, 2582
Camoin, 1781
Camp, J. M., 1318
Campbell, E. D., 7, 518, 606, 607, 949,
2358, 2374
Campbell, M. A., 2301
.Campbell, W., 1348
Candee, C., 2335
Canneri, G., 248
Carius, L., '265, 450, 581, 693, 1986
Carney, R. J., 949
Carpenter Steel Company, 1480
Carrero, J. Go, 156
Carter, F. Wo, 2388

INDEX TO AUTHORS

Carter, J. A., 3
Carter, P. H., 634
Cartledge, G. H., 950
Cascoriolus, V., 117
Caspari, R., 871
Oaspari, W. A., 1973, 1997
Cassel, H. R., 439
Cassius, L., 432
Catlin, C. N., 237
Caulfield, 'T. II., 2528, 2529
Cavendish, H., 442, 2373, 2380
Chamot, E. M., iv, v, 2434, 2435, 2438,
2439, 2440, 2447, 245a, 2454, 2457
Chancel, A. M., 934, 935, 936
Chandlee, G. C., 1216
Chapin, W. n., v, 165, 176, 178, 179,
180, 181, 185, 187
Chapman, H. D., 210
Chapman, J. E., 479
Chapman, R. :P., 1207
Chappuis, J., 2412
Charabot, E., 1777
Charles, J. A. C., 1070, 1191, 2343
Charles on, J. T., 1995
Chatard, T. M., 589
Chatelain, J. B., v, 937
Chelle, L., 188, 193
Chervet, E., 2314
Chesney, H. n., 180, 2271
Chester, F. D., 2133
Chiang, M. C., 266
Chiarno, J. C., 2257
Chiddey, A., 856
Chirnside, R. C., 626
Christoff, A., 2588
Chugaev, L., 749
Church, A. K., 2028
Churchill, H. V., iii, xiii, 1, 5; 8, 37, 50,
141
Claessens, J., 961
Clark, A. R., 396
Clark, A. W., 862, 874
Clark, B. S., 83, 964, 974
Clark, E. P., 225, 2490, 2527
Clark, H. S., 939
Clark, P. C., 1345
Clark, T., 2081
Clark, W., 2313
Clark, W. M., 1919, 2073, 2074, 2133,
2281, 2287, 2288, 2289, 2312
Clarke, B. L., 2318, 2319, 2510, 2513,
2514, 2516, 2546
Clark, F. W., 68, 861, 875, 1150, 2238
Clark, S. G., 350
Classen, A., 73, 435, 436
Claude, G. G., 687
Clausman, P., 2356
Clayton, H. D., 517
Clayton, J. E. L., 2457
,Cleaves, H. E., 231, 436
Clennell, J. E., 831

Cleopatra, 236

Clibbens, D. A., 1888,: 1914

Clifford, C. W., 266
Clifford, P. A., 521
Cline, G. E., 181, 187:
Clowes, F., 265, 2401, 2403, 2424
Cockburn, T., 618 I
Coghill, W. H., 2451
Cohen, E., 1190, 2577, 2578
Conen, J. B., 2404
Colburn, Z., 1185
Coleman, D. A., 1341, 1345
Coleman, J. B., 265
Coleman, W. C., 1481
Collins, H. L., 2503, 2504
Collins, J. F., 2398
Collins, W. D., 891
Colson, A. F., 2530
Coltman, R. W., 563
Commercial Solvents Corporation, 2134,
2147
Compton, K. G., 2322
Conant, J. B., 2312, 2315
Condon, L., A., 3
Cone, W. H., 543, 1065
Conn, L. W., 2520, 2521
Connell, G. A., 174
Conradson, P. H., 1737
Cope, W. C., 1671
Copthorne, H. N., 1842
Core, C. M., 817
Cornu, M. A., 2594, 2595
Cornwell, R. T. K., 2477
Coste, J. H., 1873
Cottle, 1969, 1970
Coulomb, C. A., 1192
Courtois, B., 448
Coward, H. F., 2424J 2426
Cowperthwait~) 1. A.) 2318
Cox, D. C., 2318
Cox, E. P., 2(51
Cox, G. J., 56, 1213
Coye, J. S., 1219
Crafts, J. M.; 2422
Craig, A, 1393
Craig, L. C., 2546
Craig, T. J. I., 3, 1,6
Craighead, C. M., 1394
Crawford, W. G., 976, 994
Cremer, F., 395
Crocker, W., 2365
Cronstedt, A. F., p14
Crook, W. J., 133,7, 2609, 2617
Crookes, W., 738, 739, 750
Cross, C. F., 1913
Crossley, H. E., 108
Crowell, W. R., 372
I
Cruess, W., 360
Cruikshank, W~2i70
Culp, F. B., 2'520 - Cummings, T~ C., 1171

INDE.X TO AUTHORS
Cunningham, T. R., iv, v, xiii, 285, 296,
304, 563, 567, 568, 597, 610, 613, 706,
808, 983, 1005, 1011, 1016, 1018; 1053
Curtis, H. L., 1998
D

Dahlbergj A. 0o, 2287

Dains, Fo Bo, 266
Dalton, J" 1191, 2S91
Daman, 1318
Dammer, 0o, 744
Danziger, Jo Lo, 1761
Darroch, Jo, 957
Da Silva, see Silva
Davies, H., 862, 1874
Davies, J., 486
Davies, J. Lo, 374
Davies, W., 862, 1874
Davis, 435
D<1vis, 1654, 1834
Davis, A. L., 2154
Davis, O. C., 1805, 1996, 1998
Davis, C. H., iv
Davis,,JJ. P., iv
Davis, G. E., 2340
Davisson, B. So, 646, 647
Davy, E., 2376
Davy, H., 117, 163, 528, 899
Day, T. G., 2515
Dean, E. W., 1340, 1342, 1345, 1717, 1841,
1889
Deeley, R. M., 1741
Deering, E. C., 981
De Gramont, A., 2609
Delano, P. Ho, 2515
Del Rio, M., 1030
De Luce, R., 435
Dely, J. G., 687, 2201
Demorest, D. Jo, 85, 374, 1384
Demorest, N. Bo, 355
Deniges, G., 118, 188, 193, 2501
Dennis, L. M., 430, 2344, 2345, 2359,
2367, 2374, 2375, 2396,2423
Dennison, C. Ho, 1778
Dennsb3dt, M., 490, 2496
Derby, W. Go, 398, 441, 834
Desains, P., 1189
Desch; C. Ho, 2459
Desha:yes, V., 491
Devarda, Ao, 634, 640, 641, 642, 646
Deventer, C. M. Van, 2575
Deville, H. St. C., 716, 728, 733, 735, 754,
976, 1186
Dewey, F. P., 433

Diamond Alkali Company, 2264

Dick, J., 320, 582
Dickens, Ho P., 628
Dickers.on, E. N., 2376
Diehl, J. C., 2342
Diehl, Wo, 524, 1870
Dienert, F., 832

Dieterich, Eo, 1769, 1770

Dietrich, E., 244
Dietz, H., 454
Dingley, W. F., 174
Di Nola, E., 432, 819
Dinsmore, R. P., 2016, 2017
Dische, Z., 2547
Ditmar, R., 1997
Ditt, M., 203
Dittmar, Po, 1760
Dittrich, Mo, 289, 980
Dixon, B. E., 148
Dixon, Ho B., 2426
Dobbins, J. T., 320, 882
Dodge, B. Fo, 2342
Dodson, R. Wo, 465
Doherty, H. L., 2391
Dole, Mo, 2303, 2313, 2318, 2319
Dole, R. B., 2085, 2088
Donath, E., 958
Donau, J., 2457
Donnan, F. G., 2576, 2577
Dow Chemical Company, iv, 2181
Dowd, J. E., 2106
Dowsett, 439
Drawe, P., 2084
Drechsel, E., 195
Dressler, R. Go, 179
Drew, H. D. K., 2541
Drew, S. K., 2369
Drinkwater, 2142
Droz, M. E., 2322
Drnshel, W. Ao, 874, 936
Dnbosc, A., 1998
Dubsky, J. V., 1214
Duffendack, O. S., 2599, 2611, 2614, 2617
Dufty, L., 330
Dulong, P. L., 1185, 1186, 1187, 1191
Dumas, J. B. A., 634, 2358, 2474, 2485,
2491
Duncan, J. B., 74
Dunlap, F. G., 1771
Dunnington, F. P., 988, 992
Dupre, A., 272
Dupre, F., 456
DuPont, G., 1190
DuPont, J., 1777
Duquenois, Po, 64
Dutoit, P., 2326
DuveI, J. W. To, 1341, 1345

E
Easley, 2374
Ebelman, M., 977
Eckardt, M., 922
Eckel, E. Co, 1619
Eddy, R. L., 2388
Eder, J. Mo, 2609
Edgar, G., 596, 1039, 1041
Edwards, J. Do, 1, 2344, 2398, 2399, 2547
Eggertz, V., 234

INDEX TO AUT_lIORS

Egloff, G., 1742

Erhardt, 0., 1190
Erhardt, U., 2322
Ehrmann, W., 1217
Eibner, A., 1779
Eisenschiml, 0., 1842
Elderdice, H. L., Jr., 707
Electrometallurgical Company, iv, 1005,
1047
Elek, A., 2493, 2496, 2500, 2508, 2523
Elliott, 2391
Elliott, A. H., 2349, 2351
Ellis, C. J., 1767
Ellis, E. W., 2537, 2538, 2543
Ellis, L., 521
Elston, O. M., 936
Elvehjem, O. A., 2519
Elving, P. J., 538
Emerson, 1654, 1834
Emerson, H., 2507
Emich, F., 2458, 2470, 2546
Emmerton, F; A., 706
Engel, E. W., 529
Engle, W. D., 318
Engler, C., 1518, 1718
Engles, M., 7
Ensign, J. R., 190
Ephraim, F., 1218
Epperson, A. W., 532
Epstein, S. W., 1986, 1987
Erdenbrecher, A. H., 212
Erdmann, H., 2459
Erdmann, O. L., 578
Erlenmeyer, H., 444
Esch, W., 1998
Eschka, A., 578, 906, 1640, 1641
Essex, J. L., 1764
Estreicher, T., 1188
Evans, B. S., 148, 626
Evans, P. S., Jr., 793
Evans, R. D., 953
Evenson, R. F., 372
Evers, N" 376
Exner, F., 2601
Exner, F. F., 824
Faber, H. B., 51
Faber, P., 988
Fahrion, W., 1769
Failyer, G. H., 874
Fair, G. M., 2103
Fairbanks, 0., 593
Fairlie, A. M., 360
Fajans, K., 273, 457, 830
Faleev, P. V., 320
Fales, H. A., 593, 1057
Falk, M. J., 1779
Falkenhausen, F. Fr. VOll, 2546
Falkov, M., 2479
Faraday, M., 1191

Faragher, W. P., 1764

Farkas, A., 444 I
Favre, P, A., U8?, 1186
Fawcett, E. H., 1919, 2285
Fay, H., 785
I
Fehling, H., 1908
Feigl, F., 819, 1212, 1213, 1214, 1218
Feldstein, P., 875
Fellenberg, T. von, 460
Fenner, C. N., 953
Fenwick, F., 2320, 2321
Ferguson, W. C., 78, 158, 365, 390, 1142,
1143, U45, 1146, 1150, 1151, 1157,
2206, 2207, 2220, 2221, 2238, 2239,
2269
Fernald, R. H., 2389
Fernandes, L., 248
Fesefeldt, H., 148
Fevert, H. L., 230
Fieldner, A. C., 1628, 1629, 1633, 1635,
1642, 1643
Finney; G. G., 953
Fischer, A., 2513
Fischer, Franz, 2379
Fischer, H., 148, 521,' 820
Fischer, K., 1778
Fish, F. H., 7
Fisher, G. E., 2164, 2169
Fisher, H., 138
Fisher, H. L., 2016
Fitzpatrick, E., 78
Fitzpatrick, E. C., iv, xiii, 394, 398
Flagg, 'J. F., V
Flath, A., 2315
Fleischer, G., 250i
Fleming, W.R.,!221, 222, 235, 236
Flexner, S., 2132, 2133
Flint, W. R., 786
Fogg, H. C., 952, 1102, 1216
Fo-lin, 0., H17
Foote, H. 'W., 372, 1190
Forcrand,' R. H. de, 1189
Ford, Henry, 2354 "-"-.
Ford, S. A., 310, 571, 1412
Forney, F. J., 465
Forrest, L. R., 1774
Foster, M. D., 187
Foulk, C. W., 879, 1144, 2196, 2320, 2321
Fowler, R. M., 297, 1209
Francis, C. B.,' 2345
Frank, F., 197:3
Frank, G., 1204
Frankland, P. F., 2411
Frary, F. C., 360
Freeport Sulfur Compavy, 937, 941
Freiberger, F" 628
French, D. K.;-iy, v, 2046
Fresenius, C. R, 166, 265, 581
Fresenius, L., 148
Fresenius, R., 123, 454
Freund, S., 980
'
/'

INDEX TO AUTHORS
Friederick, A., 2518
Friedheim, C., 921, 1004, 2064
Friedman, H. B., 2202
Friedrich, A., 129, 2482, 2491, 2530
Friedrichs, F., 2344, 2345
Friend, J. N., 246, 740, 746, 753
Froehde, A., 1959, 1960, 1961, 1965
Frommes, M., 148,407
Fry, W. H., 2435
Fuchs, K., 2518
Fudge, T., 628
Fulton, C. H., 440, 857
Fulton, J. W., 225
Funk, C., 2496, 2508
Furman, C., Miss, v
Furman, N. H., iii, vi, xiii, v, 12, 71, 78,
99, 254, 406, 456, 474, 539, 582, 596,
.662, 697, 882, 958, 961, 1027, 1065,
L193, 1203, 1205, 1206, 1421, 2134,
2191, 2305, 2310, 2311, 2312, 2314,
2320
Furter, M., 2463, 2480
Fuwa, T., 2343
Gabriel, A., 2451
Giibler, K., 2314
Gaillard, D. P;, 659
Galletly, J. Co, 1975
Gannter, F., 1769
Gardiner, A. D., 618
Gardner, H. A., v, 1839, 1854, 1864, 1876
Gardner, Ho F., 2451
Garman, R. L., 2322
Garner, W., 2459
Garratt, Do Co, 521
Garrigues, W. E., 376
Gatenby, R., 3
Gates, L. K.) 180, 181
Gautier, A., 2356
GayoLussac, Jo L., 100, 448, 826, 827,
829, 1191, 2363
Geake, A., 1888, 1914
GJber (or Jabir), A. Mo, 2362
Geer, W. Co, 1996
Gehlen, A. Fo, 2367
((eiger, Ho, 767
Geilmann, Wo, 2457
Geissler, Ho, 221
Gelbach, R. Wo, 2322
Genberg, G. Po, 1908
Gerber, Eo, 1782
Gerbes, 0., 2537, 2538, 2543
Gerhardt, Ko, 221
Gerlach, W., 2613, 2614,2617
Gettler, A. 0o, 2546
Gibbs, W., 560
Giblin, J. C., 129
Gibson, Do To, 2528, 2529
Gibson, 1186
Gibson, R. C., 1477
0

Gilchrist, R., 716, 719, 724, 733, 736, 741,

747, 752
Giles, Do Jo, 329
Giles, Wo Eo, 333
Gill, Ao Ho, iii, v, v, 686, 687, 1701, 1703,
1737, 1768, 1773, 1774, 1778, 1782,
1783, 1786, 1789, 2336, 2353, 2354,
2355, 2356, 2359
Gilligan, F. Po, 2552
Gilpin, 2142
Ginsberg, Ho, 988
Gladding, T. So, 871
Gladstone, J. H., 2375
Glaser, Co, 6
Gleu, Ko, 100, 1204, 1210
Glockler, Go, 682
Gmelin, L., 246
Gohr, Eo J., 2374
Goldberg, Jo Lo, 2497
Goldenberg (Coo), 2254
Goldschmidt, Ho, 2575
Gooch, F. A., 12, 127, 166, 168, 185, 190,
193, 269, 401, 437, 481, 560, 593, 596,
647, 648, 785, 793, 823, 832, 874, 888,
891, 926, 981, 982, 1039, 1041, 1498
Goode, K o H., 2321, 2322
Goodwin, H. To, 1190
Goodwin, L o Fo, 2136
Gorback, Go, 2463
Gortikoff, V. Mo, 2318
Goss, M. Jo, 2530
Gosse, P. Ho, 2115
Goswami, Ho C., 2509
Gottlieb, Eo, 1186
Gottlob, Ko, 1998
Grassner, F., 2480, 2515
Gray, H. L., 2439, 2484
Greathouse, L. Ho, 573
Green, H., 2454
Greenish, H. G., 2457
Greenwood, Ho D., 788
Greger, E., 2458
Gregg, R. C., 2406
Gregor, Wo, 975
Gregory, C. H., 455
Grew, No, 528
Grey, T. T., 1801
Griebel, C., 2458
Griffin, R. Co, 1884, 1910, 1922, 2031
Griffing, Eo P., 226, 635
Griffith, J. W., 2115
Griffiths, J. G. A., 2450
Groger, M., 285
Groschuff, Eo, 1717
Groth, A., 2435
Gruenberg, Po, 2082
Gruener, Ho W., 646, 647
Gruner, Lo Eo, 1188
Grunsteidl, E., 2545
Giintelberg, Eo, 2287
Guel'!,s, G. A., 375

INDEX TO AUTHORS

Guinchant, M., 1190

Guire, O. D., 1619
Guldberg, C. M., 1192
Gunning, J. W., 632, 663, 1652, 1653,
1873, 1906
Gustavson, R. G., 318
Gutzeit, G., 64, 87, 93, 101, 102, 103, 105,
106, 107, 108, 185, 1919, 1920, 2255,
2367, 2418
Gyemant, A., 2289
Gyiiry, S., 74
H
Haber, F., 442, 687, 2311
Haberland, G., 2529
Hackl, 0., 407
Haddock, L. A., 376
Hager, H., 1149, 2459
Hager, O. B., 2322
Haggard, H. N., 2360, 2364, 2427
Hahn, F. L., 12, 866
Hahn, R. M., 56, 1213
Haitinger, M., 2546
Haldane, J. S., 2365, 2405
Hale, F. E., iv, v, 2079, 2083, 2086, 2099,
2106,2115
Hall, C. M., 1
Hall, D., 311, 1217 (see also Brophy, D.
H.)
Hall, G., 1911
Hall, L. 'P., 2270
Hall, N. F., 2312, 2318
Hall, W. T., 454, 635, 639, 908, 919, 926,
1036, 2353
Hallett, L. T., iv, v, 2460, 2473, 2475,
2482, 2492, 2494
Halpern, K:, 1774
Halphen, G., 1'176, 1778, 1787
Hammed, H., 1190
Hammett, L. P., 2, 1207, 1213
Hammond, L. D., 405
Hammond, W. H., 1055
Hampe, W., 310
Hanausek, T. F., 2459
Handy, J. 0., 534
Hankus, E., 2345
Hanna, W. C., v, vi, 1597, 1618
Hannel, H., 115, 2563
Hansen, G., 2609
Hanus, J., 1767, 1800
Harcourt, A. G. V., 1200
Haring, H. E., 2322.
Harms, J., 2335
Harned, H. S., 2300
Harper, L. E., v, 1032, 1053
Harrell, H. L., 2493
Harries, C., 1975
Harris, J. A., 256, 257
Harris, T. H., 2491
Harrison, G. R., 2601
'Harrison, J. B. P., 66, 165

Hart, E. B., 2358, 2374

Hart, T. S., 874
l
Harte, R. A., 2493, 2523
Hartley, 0., 634
I
Hartmann, J., 2598, 2600
Hartmann, 'vi., 69 1
Hartong, B. D., 2310
Hartshorne, N. A., 2435
Haschek, E., 2601
Haslam, R. T., 2374
Hastings, A. B., 2287
Hauser, E. A., 1968, 1998
Hauser, 0., 248, 947
Hauteville, P., 2498
Havens, F. S., 12, 1498
Hawley, F. G., 406, 413, 418, 806, 80'1
Hawley, L. F., 944
Hawse, V. P., 1672
Hawson, J. W., 435
Hayman, D. F., 2491
Heath, G. L., 360, 362, 378
Hecht, F., 1217, 1218, 2521, 2545, 2547
Heczko, T., 626, 2318
Hedrich, ,G., 2314
Hefner, A. von, 2391
Hefter, G., 1785
Hehner, 0., 2149
Heil, 'A., 1998
Heinz, C., 2345
Heller, K., 64
Helmick, H. H., 947
Hemmes, 2439
Hempel, W., 444, 686, 1166, 2342, 2343,
2353, 2357, 2358, 2359, 2397, 2399, 2402
Henderson, J. A. R., 376
Henderson, J. M., 460
Henderson, Y., 2360, 2365, 2427
Henfrey, A., 2115 .'
Henne, A. L.; 407 I
Henriques, ~., 1737, 1978, 1986
Henry, W., 1191, 2375, 2587, 2591
Henz, F., 73:
Herig, H. W., 2199
Hermance, H. W., 2510, 2513, 2514, 2516,
2546
,
Hermann, K. S. L., 197
Herner, E. P., v, 405, 418
Heroult, P. L. T., 1
Herr, W., 1318
Hersey, M. D., 2368
Herzberg, W., 2458
Herzfeld, E., 456;
Herzog, A., 2458
Hess, G. H.', 1192
Hess, W. H., 7
Hesse, E., 2123
Hesse,.W., 2403
,
Hesslinga, J., 225,_226, 227
Hevesy, G. von,.,7'67, 1104, 1189, 1190
Heyrovsky, t., 2546
Hibbard, P. L., 1215, 2547

INDEX TO AUTHORS
Hibbert, E., 479
Hickman, R. E., xiii, 710, 730
Hicks,. W. B., iii, xiii, 861, 870, 872, 2575
Hickson, E. F., 984
.
Hildebrand, J. H., 2309, 2312, 2423
Hilger, A., 2609
Hill, D. W., 2496, 2500, 2508
Hill, L. A., 874
Hillebrand, W. F., 30, 176, 179, 534, 575,
576, 626, 799, 801, 803, 882, 884, 955,
981, 983, 988, 996, noo, n02, 1922,
2035
Hillis, T. E., 372
Hind, H. L., 2457
Hinman, B. C., 2356, 2357, 2395
Hinman, C. W., 919
Htnrichsen, F. W., 1997, 1998
Hintz, E., 908
Hippensteel, C. L., 1998
Hiss, P. H., 2132,2133
Hohne, V., 2458
Hoffman, E. J., 1672
Hoffman, H. 0., 2557
Hoffmann, A: W., 356
Hoffmann, F. L., 2372
Hoffmann, J. 1., 5, 312, 405, 408, 538,
626, 997, 1421
Hohn, H., 2546
Holde, D., 1556, 1703, 1736, 1769,.1776,
1778
Holladay, J., A., 1018
Hollingl;worth, M., 1144, 2196
Holloway, G. T., 578
Holstein, L. S., 197, 1054, 1079
Holt, D. A., 429
.
Holt, M. L., 2320
Holzer, H., 2547
Homberger, A. W., 781
Homerberg, V. 0., 2459
Honig, M., 1773
Hooke, R., 1996
Hooper, H. E., 66
Hoover, C. R., 219
Hoover, T. J., 1318
Hopkins, B. S., 148, 256, 257
Hopkins, C. G., 646, 2359, 2374
Horrobin, S., 2328
Hostetter, J. C., 487, 2320
Howard, B. J., 2449
Howard, C. D., 400
Howard, D. L., 66
Howard, H., 1143, 1146, 1151, 2207,2221,
2239
Howe, H. E., 2335
Hoyt, H. E., 2335
Hoyt, L. F., 1345, 2039
Hoyt, S. L., 2459
Hradecky, II., 749
Hu., M., 266
Hub, A., 402
Hubbard, J. M., 407

Hudson, C. T., 2115

Hudson, J. II., 2364
Hiibl, B. von, 1766, 1767, 1768,1769
Hughes, W. S., 2303, 2310
Huismann, J., 1782
Hulett, G. A., 1144, 2197, 2208, 2409
Humphreys, T. C., 360
Hunt, 395
Hunt, S. P., 687, 2355, 2356, 2359
Hunter, M. A., 975, 977
Hunter, W. II., 683, 2547
Hurd, L. C., iii, xiii, 396, 768
Hurley, W. II., 518, 519, 606
Husband, A. E., 460
Hyde, A. L., 1671, 1675
I
Ibbotson, F., 52, 589, 1421
Ihering, V., 2388, 2389
Ilinsk;, M., 306
IlJarionov, V., 634
Illingworth, C. B., 1028
Ingle, II., 1769
Inouye, K., 2577, 2578
Insley, H., 2451 .
International Nickel Co., iv
Isham, R. M., 978
Ishibashi, K., 1218
Ivanov, I., 3, 11, 53

J
.Tacks on, D. D., 2051
Jackson, R. F., 1725
Jackson, V. T., 949
Jacobs, W. A., 1717
Jacoby, K., 174
Jaeger, E., 2359, 2374, 2375
Jaeger, F., 725
Jahn, C., 334, 339, 341, 1004
Jakowsky, J. J., 2369
James, C., 245, 252, 256, 947, 952, 1096,
n01, n02, 1216
Jamieson, G. S., 99, 274, 455, 1209, 2180
Jander, G., 2327, 2328, 2335
J anett, S., 2313, 2328
Jannasch, P., 155, 799, 802
Jansky, A., 2313
Jantzen, V. T., 1104
Jarmus, J. M., 989
Jeffries, Z., 2558, 2564
Jelley, E. E., 2546
Jenkins, E. II., 2344, 2395, 2396
Jentgen, H., 2318
Jewell, P. W., 238
Jilek, A., 964
Joannis, J. A., n88
.Tohnson, C. M., 703, 1409, 1421
.Tohnson, E. B., 430
Johnson, O. C., 262, 743
Johnston, J., 905, 917
.Toly, A., 1190

10

INDEX TO AUTHORS

I'

King, W .T., 606 ,

.Joly, .T., 767
Kingman, W. A., 1340, 1343, 134411 1345
.Iones, B., 350
Jones, C., 469, 475, 476, 495, 594, 595,
Kingscott, P. C. R, 265'
.'
Kingzett, C. T., 21$0
612, 699, 1925
Kinney, S. P., 2454,2456
Jones, D. E., 1991
Kinzel, W., 2458 '\
Jones, G. W., 1963, 2361, 24024
Kinzie, C. J., 1864 ,
Jones, L. C., 166, 168,185
Kirk, P. L., 2499, 2547
Jones, P. T., 2318
Kirner, W. R, 2479, 2480
Joule, J. P., 1192
Kishi, H., 1218
Joyce, C. M., 2418
li.issa, M.., 2\)'i'>2
Juuu,1:\. C., 2515, 2511)
Kissling, 0., 1716
Junkers, H., 1802, 2391, 2392
Kittredge, E. B., 2312
K
Kjehldahl, .T., 632, 634, 693, 1652, 1653,
1671, 1873, 1905, 1906, 1973, 1974,
Kabulov, I., 1189
1987, 2007, 2491, 2492, 2494
Kahlenberg, L., 892, 2320
Klaproth, M. H., 975, 1017
Kahler, H. L., 2092
Klein, A. A., 2451
Kahler, W. H., 919
Klein, G., 2458
Kalmus, H. T., 1190
Kleinmann, H., 2520
Kameda, T., 1097, 2332
Klemenc, A., 1893
Kamienski, B., 2318
Klemensiewicz,' Z., 2310
Kamm, 0., 1908
Klemp, G., 50
Kao, C. H., 434, 782, 783
.. Kley, P. D. E., 2457
Kar, H. A., 1482
Kling, A., 958, 961, 2148
Kato, S., 429
Klinger, P., 1102, 1216
Katz, S. H., 2373
Klinke, J., 2520
Kaye, F., 1984
Klit( A., 2310
Kayser, H., 259, 2601
Knauth, 2314
Kearney, T. H., 2098
Knecht, E., 479, 2158
Keenan, G. L., 2449
Kneeland, E., 402
Keffer, R., 379 '
Knietsch, R., 1190
Kehrmann, F., 1003
Knight, R. S. G., 265
Keiser, E. H., 2370
Knop, J., 291, 473,1205,1206,1478,2522, \
Keller, E., 433, 779, 787, 789, 1306
2523
'
Kellermann, K. F., 2111, 2112, 2114
Knorr, A. E., 1203
Kelley, G. L., 1028, 2315, 2318
Knorr, E. S., 1975
Kelly, F. C., 460
Knorr, K., 93, 235,
Kelly, W . .T., 2014
Knorre, G. yon, 2,53, 306, 310, 2063
Kemmerer, G., 103, 2475, 2482, 2492
Knowles, D., C. Ji., 434
Kern\!, L. C., 2398
Knowles, H. B., 8, 14, 531, 587, 591, 955, .
Kendall, E. C., 460
980, 981,-~ 984, 1027, 1094, 1097, 1100,
Kennicott, 606
1102, 1213, 1216, 1461,
Kenny, W. R, 2287
Koch, A.
403
Kenrick, F. R, 2582
Koch, W., 2458
Kent, W. S., 2115
Konig, J., 2459
Kenyon, W.O., 2484
Koettstorfer, .I., 1771
Kerr, C. H., 1318
Kogert, H., 2318
Kerr, R H., 1787
Kohler, S., 1911
Kerridge, P. M. T., Mrs., 2303
Kohman, G, T., 1998
Kerschbaum, P., 2498
Kestranek, W., 2310
Kohn, C. A., 7
Kohn-Abrest,.E., 50
Kiely, H. U., 1908
Kohnstamm,
1767
Kiess, ,C. C., 2609
Kolokolov, N., 116
Killeffer, D. H., 2366
Kolthoff, I. M., iv, v, vi, 8, 12, 143, 193,
Kilpatrick, M. L., 990
273, 296, 310, 350, 351, 423, 456, 457,
Kimberly, A. E., 1916
479, 529, 531, 535, 559, 571, 582, 640,
Kimura, K., 1104
644, 662, 819, 878, 879, 882, 1027,
Kimura, W., 2498
1065, 1203, :].206, 1214, 1471, 1478,
King, E .T., 803
1925, 2191;r2287, 2306, 2307, 2308,
King, .T. F., 481
2310, 2311, 2312, 2314, 2323, 2328,
, King, J. T., 838, 839
2333, 2335, 2499
King, V. L., 7,468

A.,

L.,

INDEX TO AUTHORS
Kon, S., 2496, 2508
Konen, H. M., 2601
Koninck, L. L. de, 532, 533
Koppeschaar, W., 1950
Korbuly, M., 2353
Koref, F., 1188
Koremnann, 1. M., 64
Kraemer, E. 0., 521
Kraemer, G., 2491
Krasnitz, ~, 634
Kra.tz, G. D., 1973
Krauskopf, F. C., 892
Kraut, K., 246
Kraybill, H. R., 2472
Kreis, H., 1782
Kretzschmar, M. J., 7
Krieger, K. A., 2035
Kriesel, F. W., 426
Kronacher, C., 2458
Kroupa, E., 2545
Krueger, A. C., 2320
Krass, G., 486
Krug, W. H., 3
Krumholz, P., 1~14
Kubelkova, A. 0., 2523
KiVm, W. E., 2398
Kuenen, J. P., 1189
Kaster, W., 2530
Kuhn, R., 2501, 2526
Kukolich, S. 1., .2290
Kundert, A., 582
Kuras, M., 203
~rtenacker, A., 285
Kurtz, F. E., 634
Kyle, T. D., 157

L
Lachele, C. E., 108, 2505
'Ladd, R. M., 655 .
Lakshminarayan, S., 2354
Lamar, M. 0., iv, 5, 32, 181, 538, 813
Lamb, A. B., 219
Lamb, F. W., 521
LaMer, V. If:-., 2202, 2310
Lammert, O. M., 2301
Landholt, H., 2411, 2412
Landrum, R. W., 329, 330
Lang, R.,' 455, 1210
Langbein, G., 824
Lange, E., 2313
Lange, N. A., 2601
Lange, W. E., 1986, 1987
Langley,IJ. W., 656
Langmuir, A. C., 534
Laplace, P. S., 1192
Larison, E. L., 360
Larsen, E. S., 2435, 2451
Larson, W. D., 582
Lassieur, A., 958, 961, 2HS
Lassone, J. F. M., 2370
Lathrop, E. C., 1178

11

La Tourette, H., 2418

Laufis, A., 1782
Laurie, L. G., 2459
Lauro, M. F., 633, 249,3
Lavoisier, A. L., 442, 629, 1192, 2365
Lea, C., 738
Leach, A. E., 1787, 2149, 2158
Leaver, E. S., 393
LeBlanc, M., 922
LeChatelier, H., 1140, 2552
Ledue, A., 1189, 2411, 2412
Lee, M. F., 8
Leech, W. D., iv, 460
Leeds and Northrup Co., iii
Leidy, J., 2115
Leighton, P. A., 2537
Leitch, I., 460
Lenher, V., v, xiii, 434, 438, 779, 781,
782, '783, 793
Lenssen, E., 964
Leopoldi, G., 521, 820
Lepper, W." 945
Lerch, W., 1618
Leroy, A., 2545
Levenson, H., 1699
Levol, A., 96
Levy, L., 979

Levy, L. H., 374, 1190

Lewis, E. J., 626
Lewis, H. F., 2290
Lewis, S. J., 2617
Lewis, W. H., 2340 .
Lewkowitsch, .T., 1773, 1775, 1777, 1779,
1787, 1789
Lichtin, J. J., 296
Liddell, D. M., 833, 2389
Lidstone, F. M., 2546
Lieb, H., 2500
Lieben, A., 1948
Liebermann, C., 1774) 1781) 1979
Liebich, C., 2313
Liebig, J. von, 221, 277, 661, 1650
Lind, S. C., 759, 167
Linder, J., 2468
Lindgren, J. M., 622
Lindo, D., 871
Lindow, W. C., 2519
Lindsey, A. J., 2515, 2547
Lindsley, L. C., 2459
Lindsly, H., 2508
Lingane, J. J., 882, 1027, 1471
Little, H. V. F., 246
Littrow, M. de, 2594, 2595, 2596, 2597
Lochte, H. L., 2534
Lockemann, G., 108
Lockhart, L. B., 1789
Lockwood, E. H., 2384, 2386
Lockyer, N., 1789
Lodemann, G., 2458
Losch, J., 252
Liiwe, F., 2601

12

INDEX TO AUTHORS

McGill, A., 237

Lowe, J., 154
McGill, W. J., 2310
/
Logan, W. N., 2453
McIlheny, P. G., i767, 1770, 1773, 1800,
Long, C. P., iii, vi, 1340, 2026
2043
L'Orange, J., 2480
McInerney, T. J., ~289
Lord, N. W., 525
MacInnes,. D. A., I ~300, 2313, 2318
Lovejoy, D. R., 2379
MacIntosh, J. B., 2353
Low, A. H., 65, 75, 91, 375, 402, 405,
MacKay, W., 225 \
509, 535, 580, 957, 1022, 1064, 1202,
McKenzie, A. E., 647
1856
MacKey, W. McD., 1782
Low, G. W. Jr., 2198
McKinney, D. C., 2312
Lowenthal, R., 21G8
McMaster, L., 2370
Lowrey, T. M., 311
MacMichael, R. F., 1724
Lowry, H. H., 1998
McNabb, W. M., 77, 696
Lubs, H. A., 20n
McNeill, C. L., 379
Lucas, C. C., 803
McNeill, T. R., 285, 296
Lucas, R., 2480, 2515
McPherson, A~ T., 1998
Lucko"iy, C., 73
McTrusty, J. W., 2401
Lucker, F., 2499
Maczowski, E. E., 1614
Luff, B. D., 1998
Maes, P., 225
Luiginin, W. F., 1190
. Mahin, E. H., 'v
Lukens, H. S., 2035
Mahler, .P., 1187, 1654, 1834
Lummer, 0., 2391
Mahr, C., .156, 157
Lundegardh, H., 2617
Maier, C, G., 396
Lundell, G. E. F., 5, 8, 14, 30, 405, 408,
Makower, W., 767
531, 534, 538, 575, 576, 626, 955, 980,
Malatesta, G., 432, 819
981, 983, 984, 997, 1027, 1094, 1097,
Manning, R. J., 1893
1100, 1102, 1216, 1421, 1922
March, F., 1777
Lunge, G., 244, 486, 935, 956, 967, 1144,
Marchand, R. F., 578, 1650
1147, 1158, 1167, 2147, 2198, 2201,
Marchlewski, L., 244, 1144
2222, 2268, 2378
Marcusson, J., 1343, 1345, 1548
Lungwitz, 1736, 1769
Margosches, B. M., 273, 454
Luttringer, A., 1998
Marignac, C., 332
Lux, F., 1736
Mariotte, E., 1650, 1652
Lyons, E., 487
Marks, L. S., 2339, 2340, 2389
Lyons, M. A., 2449
Marks, M. E., 226
Lyson, L. T., 2571
MarIe, E. R., 2483
M
Marme, M., 1959, 1p61, 1965
Maag, O. L., 606, 613, 1481
Marr~an, G. ,F., 2qii7
Marnan, P.: M., 2547
Maag, W., 2530
Marsh, J., 87, 109, 2367
Maas, A. R., vi, 1943
Marsh, W. -T., 620
Maas, H., 1215
Martin, 222
Maasen, G., 628
Martin, G., 11981, 2017
McAdams, W. A., 2340
Marvin, G. G., 1095, 1481
McBride, R. S., 1209, 2395
Marx, P. F., 2387, 2388"-- _____ _
McCabe, C. R., 994
Mascazzini, A., 73
McCandless, J. M., 532, 696
Mason, C. W., iv, vi, 2434, 2435, 2428,
McCarthy, E. S., 2396
2439, 2440, ~447, 2453, 2454, 2457
McCay, L. W., 71, 78, 95, 590, 862, 958
Mason, H. M., 2450
McClendon, J. F., 460
Mason, S. N.~ 1774
McCollam, C. H., 606
Massol, G., 1189
McCoy, H. N., 255
Matheson, H., 1919
McCroskey, C. R., 1481
Mathews, J. H.,' 2459
McCrumb, F. R., iv, vi, 2274, 2284, 2285,
Mathewson, C. H., 2453
2287
Mathy, E. L., 2378
McDaniel, A. S., 2588, 2589
Mattern, J. D., 2264
McDonald, R. G., 1884, 1910, 1913, 1916
Maumene, E., 1701, 1766, 1.1.813
McDonnell, C. C., 91
Maxymowicz, W..!..:_152
/
McDowell, J., 660
Mayer, J. L.; f'965 Mace, J., 2458
Mayow, J., 2368
,McFarland, D. F., 2379

INDEX TO AUTHORS
Mayr, c., 2498
Mayrhofer, A., 2435, 2458
Mazzotto, D., 1188
Muzzuchelli, A., 753
Meade, R. K., v, 374, 1384, 1619
Mecklenberg, W., 1094
Meggers, W. F., 2609
Meiklejohn, C., 957
:lVIeiklejohn, R. M., 1161, 1219
Meineke, K., 1441
Melaven, A. D., 47, 1458
Mellor, J. W., 148, 244, 246, 410, 721,
729, 733, 956, 968, 1036, 1141
Meloche, Y .. W., 2546
Memmler, K., 1997, 1998
Mendeleef, D., 429, 1192
Mene, C., 966
Mmmick, H., 276
Menzies, A. W. C., 2530, 2532
Merwin, H. E., 410, 413, 988
Merz, 0., 2546
Messinger, J., 2136
Metzel, A., 72
Metzger, F. J., 254, 2364
Metzner, P., 2457
Meulen, H. Ter, 225, 226, 227, 635, 682,
2509, 2510
Meyer, A. H., 2501
Meyer, Y., 2423, 2534, 2575
Michaelis, L., 2289
Middleton, A. R., 2
Mika, J., 2464
Miller, A. L., 5
Miller, C. 0., 00
Miller, E. J., 521
Miller, L. F., 749, 751
Miller, S. P., 435
Milliau, E., 1766
Milligan, L. H., 2507
Millon, N. A. E., 1950, 1964
Mills, E. J., 1769
Mills, P. L., 108
Milner, H. W., 2530
Milner, R. T., 2465, 2489, 2490, 2491
Miner, H. S., 948, 950
Mitchell, C. A., 1767
Mitscherlich, A., 646
Moberg, E. G., 2547
Moeller, J., 2458
Mohler, E., 1185, 1186
Mohr, F., 75, 76, 238, 271, 272, 456, 1541
Moir; .T., 439
Moisaan, H., 399
Molisch, H., 2459
Mollow, G., 823
Monfort, W. F., 2111, 2115
Montequi, R., 156
Moody, L. T., 2571
Moore, G. T., 2111, 2112, 2114
Moore, H., 2552
Moore, R. L., 1987

13

Moraht, H., 486

Morley, E. W., 437, 1347, 2408
Morgan, J. L., 1143, 1146, 1151, 2207,
2221, 2339
Morgan, J. L. R., 2301
Morgan, S., 1998
Morrell, J. C., 1764
Morris, G. W., 2458
Morse, H. N., 1223
Moser, L., 152, 427, 445, 446, 896, 897,
943
Moskovitz, B., 644
Muehlberg, W. F., 1095
Muhlsteph, W., 108
Muelken, Mrs. A. Z., v
Mueller, C. W., 2394
Muller, E., 2307, 2313, 2314, 2315, 2318,
2319, 2320
Muller, F., 988
Muller, Fr., 2314, 2322
Muller, R. H., 2322, 2472
Muer, H. F., 2065
Muggenthaler, H., 1179
Muir, M. M., 157
Muller, J. H., 429, 430
Mulligan, J. J., xiii, 159, 160
Munroe, C. R, 1767
Munson, L. S., 1768, 2039
Muraour, H., 1698
l"furgelescu, I. G., 2335
Murphy, G. M., 444
.
Murray, O. W., 108, 2367
Murray, W. M. Jr., 596
Myers, F. B., 1028
Mylius, F., 434, 753

N
/
National Lead Company, iii, iv, 995,
1387 ff.
Nauss, A. 0., 2363
Nayak, U. M., 2534
Neal, J. L., 56
Neller, J. R., 108
N elsmann, H., 966
Nemst, W., 2305, 2306
Nessler, J., 630, 656, 657, 658, 1974,
2019
Neubauer, H., 2469, 2499
Neuberg, C., 1776
Neuman, C., 1182
Neumann, B., 2565
Neumann, E. VV., 905, 908
New, C. H., 2358
Newhall, C. A., 934
New Jersey Zinc Co. (of Pa.) iii, 1971,
1202
Newton, H. D., 984
Nichols, M. S., 461
Nicklay, A. J., iv, 527
Nicloux, M., 1948
Nicolardot, P., 1102

14

INDEX TO AUTHORS

Niederl, .T. B., 2322, 2479, 2481, 2485,

2488, 2534
Nierenstein, M., 1893
Nikolow, C., 320
Nissensohn, H., 74
Nitchie, C. C., 1089
Nola, see Di Nola, E.
Noponen, G. E., 640, 1214
Nordlander, B. W., 2423
Norris, J. F., 785
Norris, W. V., 410
Norton Company, iv
Norton, J., 1786
Norton, R. H., 229, 230, 231
Noyes, A. A., 208, 429, 430, 434, 465,
893, 979, 2300, 2572
Nussberger, 2322
Nusselt, W., 2373
Nydegger, 0., 921, 2064

Papish, .T., 429, 430, 894, 896

Paracelsus, 149, 1054
Park, B., 372; 626 '
Park, W. F., 2132, 2133
Park, W. H., 2133, \
Parker, J. C., 2164, 2169
Parker, W. L., 2361
Parr, S. W., v, 622, 1643, 1654, 1834,
2508
Parrodi, G., 71 "
Parry, E. J., 1873, 1981
Parsons, C' L., 140, 148 .
Parsons, J. L., 1913
Parsons, T.. R., 2289, 2310
Partridge, H. M., 2322
Pass, A., 1217
Patat, F., 2518
Patterson, W. L., 2549
Pattinson, H. S., 95
Pattinson, J., 95, 560
o
Paul, .T., 72
Obach, E., 1993, 1994
, Payne, H. ,L., 2425, 2426
Oberfell, G. G., 2368, 2377, 2403
Peachey, S. J., 1977
O'Brien, W. J., 2367
Pearce, :EJ. V., 100
Ode, W. H., 803
Pearce, S. J., 2547
Oersted, H. C., 1
Pearson, E. A., 1925
Oesper, R. E., 455, 1205, 1206
PearsOli, H. C., 1998
Offerman, H., 409
.
Pease; V. A., 2450
Ogburn, S. C., 738, 749, 751
Pechard, E., 1004
Ohm, G. S., 1192
Peligot, M., 1021
Oilar, R. D., 1778, 1779
Pemberton, H. V., 2039
Okac, A., 1214, 2516
Penfield, S. L., 409, 1,340
Okuma, K., 529
Perkin, F. ~., 436
O'Leary, W. J., 894, 896
Perley, G. A., 23(l2
Olsen, J. C., 401, 1136
Perrot, F. L., 2411
Olson, H. W., 180
Perrot, G. St. J., 2454, 2456
Opfermann, E., 1884
Person, G. G., 1188, 1189, 1190
Orsatt, M. H., 2342, 2345, 2349, 2385,
Peters, 1654, 1834/
Peters, F. P., 627
2388
Orton, E., 2398
Peters, R., 2306
Osborn, R. A., 634, 2493
Petit, T. A., '1191
Ostwald, Wm., 2307, 2587, 2588
Petkow, N., 1778
Ostwald, W 0., 1977
Petterson, 0:, 2405
Otto, C., 2
Petterssoli, 0., 1188
Oudemans, A. C., 1156, 2252
Peyau, H., 1778
-....:
Oulman, 2254
Pfiitzer, G., 120
--.....
Outcault, H. S., 2365
PfUlldt, 0., 2327, 2328, 2335
Outerbridge, A. E., 1701
Phelps, E. B., 2081
Owen, F., 2266
Phillips, F. C., 2407
Phillips, M., .;l530
p
Physick, P. C., 2366
Pacific Coast Borax Co. iv, 174
Pickering, P. S. V., 1190
Page, R. T., 2343
Pickering, S. W j 11g'1, 2239
Pagel, H. A., 249
Pickles, S. S., 2u12
Pagenstecher, A., 1946
Pied, H., 340, 343
Palkin, S., 890
Pierce, A. W., 785
I
Pallet, K., 966
Pierle, C. A., 1005, 1021
Palmer, C. S., 952, 1102
Pierson, R. H., v, 1804
I
Palmer, I. A., v, xiii, 835, 2025
Pinkus, A., 96I_,':" ~
Palmquist, A., 2405
Pirsch, .T., 25-34, 2469
J>aneth, F., 767
Pisani, F., 832

INDEX TO AUTHORS
Pitkin, Lucius Co., v
Pitman, J. R., 650
Pitot, 1173, 2340
Plato, W., 69, 1189, 1190
Plechner, W. W., 989
Plent1, A., 2534
Plimmer, R. n. A., 2501
Pliny, 954, 2375
Polenske, E., 1786
Pollak, J., 2521
Pollak, L., 582
Pollitzer, F., 1188
Polushkin, E. P., 2453
Polyakov, A., 116
Ponomarev, V. D., 108
Pope, G. S., 1318
Popoff, S., v, 905, 908
Popp, G., 166
Porter, L. W., 1903
Poschl, V., 2459
Poth, .E., 2489
Powell, A. R., 333, 338, 339, 346, 445,
773, 942, 1004, 1099
Pozzi-Escot, E., 433
Prandtl, W., 252, 1104
Pratt, L. A., 252
Prausnitz, P. n., 2389
Prenle, W.O., 436 '
Pregl, F., 2481, 2487, 2490, 2494, 2510,
2511, 2516, 2517, 2518, 2526, 2528,
2547
Prescott, A. B., 262, 743
Prescott, S. 0., 2133
Preston, J. M., 2459
Price, T. S., 360
Price, W. B., 355, 374, 1384
Priestly, J., 2360, 2362, 2364, 2368,
" 2373, 2378
Prince, N. F., 2399
Pringsheim, P., 267
Prister, "A., 437
Procter and Gamble 00., iii
Proctor, H. R., 2158
Pulman, O. S., 593, 1027
~

Quill, L., 978

R
Rab.inovich, A. I., 2318
Rachele, J. R., 2546
Rafter, G. W., 2102, 2103, 2104
Raikow, P. N., 1778
Raiziss~ G. W., 2479
Rajmann, E., 1214
Rammelsberg, 0., 889
Ramsay, W., 683, 1190, 2380
Randall, D. L., 594
Randles, W. B., 2457
Rapoport, S., 129, 2518
Rasch, R. S., 1913, 1914, 1916

15

Raschig, F., 647, 920, 926

Rast, K., 2533
Rathgen, F., 3
Rathsburg, H., 78
Rauchswerger, D., 1776
Rawlins, n. J. B., 966
Rawson, 0., 2158
Ray, P. R~ 1217, 1218
Rayleigh, Lord (J. Strutt) 2380, 2411
Rebiere, G., 832
Redmond, J. C., 1607
Redwood, B., 1741, 2401
Reed, R. M., 1345
Reedy, J. H., 892
Regester, S. n., 1642, 1656
Reglade, A., 1102
Regnault, H. V., 1160, 1188, 1189
Reich, F., 2399, 2400, 2401
Reicher, L. T., 2575
Reid, E. E., 1986
Reif, W., 943, 1218
Reiff, O. M., 990
Reinhardt, C., 299, 476, 482
Reiss, M. A., von, 73
Reissner, R., 1218, 2521, 2547
Reith, J. F., 460
Remington, R. E., 460
Renard, A., 1780
Rethberg, P. B., 2498
Rey, n., 1147, 2201, 2222
Rhodes, F. n., 2364 .
Rhodin, J. G. A., 50
Rice, A. C., 952, 1102, 1216
Rice, F. 0., 990
Rich, C. E., 2493
Richards, J. W., 1188, 1783, 2388, 2389
Richards, T. W., 360, 831, 832 2337
Richardson, 0., 1619
Riche, A., 2147
Richmond, H. D., 165, 1767, 1801
Rieche, A., 2533
Riesenfeld, E. n., 120, 2591
Rigg, W. L., xiii, 305
Riley, H. L., 1218
Ringelmann, 2388
Rinkenbach, W. R., 1660, 1686
.
Risdale, N. D., 997
Ritschel, E., 896, 897
Rittenberg, G. S. C., 372
Roark, R. 0., 91
Robbs, O. E., 157
Roberts, E. J., 252
Roberts, H. S., 2320
Roberts, L. D., iii, xiv, 682, 758
Roberts, R. T., 392
Robertson, D. W., 1864
Robertson, 1. W., 2402
Robertson, P. W., 1188, 1190
Robinson, R. A., 2310
Robinson, R. J., 2526

16

INDEX '1'0 AUTHORS

Robitschek, J., 531

Robson, W. G., 1189
Roe, R.. B., 1880
Rohre, K., 69
Rossler, L., 432
Roger, R., 225
Rogers, A., 1789
Hogers, L. A., 1343, 1345
Hogers, S. M., 2546
Roscoe, H. E., 730, 742, 744, 748, 1036
Rose, H., 122, 156, 314, 403, 1036
Hose, T. K., 436, 440
Roseman, R., 479, 989
Rosengarten, A. G., 1143, 1146, 1151,
2207, 2221, 2239
Rosimthaler, L., 2459
Ross, E. M., 2092
Ross" J. F., 872, 884
Ross, J. H., 1913
Ross, W. H., 824
Rossem, A. van, 1998
Roth, H., 918, 2526, 2547
Roth, R. T., 2479, 2481
Roth, W. A., 1189, 2318
Rothaug, C., 286
Rothe, J., 1983
Rothe, J. W., 312,317
Rother, E., 2335
Roux, L., 1185
Rowe, F. M., 529
Rowell, H. W., 74
Rudberg, F., 1188
Ruder, W. E., 811
Rudorff, Fr., 1156, 2397
Rudy, R. B., 1615
Ruggeri, R., 1778
Ruigh, W. L., 529
Rupert, F. F., 2179
Russ, F., 346
Russell, W., 225, 226
Russell, W. J., 2123, 2132
Rutherford, D., 2378 .
Rutherford, E., 767
S
Saint-Rat, L. de, 1245
Saint-Sernin, A., 216
Salas, L. E., iv, vi, 1966
Sand, H. J. S., 71
Sandell, E. B., 8, 296, 310, 531, 559, 571,
644, 1471, 1478
Sander, W., 2590
Sanders, J. McC., 931, 932
Sanders, M. T., 1677
Sanders, R. B., 2493
Sandin, R. B., 2493
Sandstedt, R. M., 2493
Sargent, G. W., 2390, 2391,
Sarker, P. B., 2509
Sarrau, E., 1185
Sarver, L. A., 249, 319, 474, 628, 1206

Sauveur, A., 2457, 2552,2553, 2554, 2558,

2563,2564 _
Savell, W. h, 6281
Saybolt, G. M., 1~18, 1717, 17,19
Sayers, R. R., 1963, 2364, 2406
Schaeffer, ,J. A., vi, 525, 527, 1839, 1870,
I
1874, 1876
Scheele, C. W., 2611 554, 689, 1001, 2362,
2363, 2364, 2365, 2368
Scheibe, G., 2601, 2617
Scheibler, C., 244, 2366
Scherrer, J. A., 70
Schicht, H., 1774
Schidrowitz, P., 1996, 1998
Schiff, H., 1198, 2148
Schiodt, E., 2287
Schloklitsch, K., 2545
Schmidt, G., 1973
Schmitz, W., 82
Schneider, A., 2458
Schneider, F., 2445, 2458
Schneidewind, R., 917
Schoeller, W. R., iii, v, xiv, 331, 332, 33:
334, 338, 339, 340, 34-1, 342, 343, 34'
345, 3;16, 348, 426, 445, 773, 793, 94i
981, 1004, 1099
Schonbein, C. F., 1946, 2369, 2405
Scho}l, C. Eo, 1026
School, A . T., 210
,
Schoonover, I. C., 820, 882, 1027, 1214
Schopper, L., 1995
Schor lemmer, C., 730, 742, 744, 748
Schotz, S. P., 1998
Schou, S. A., 2310 Schreiber, H., 1737
Schreinemakers, 2581
Schrenk, H. H., 103, 2547
Schrenk, W. T., 80~, 2515
Schroeder, W. C., 129, 919
Schroetter, A. R. :Von, 238, 1541
Schuchardt, T., 2370
Schiitzenberger, R., ~76
Schulek, E., 180, 662
Schultz, G., p29
'"
Schulze, F., 50, 2310
__
Schum, E., 3, 53
Schumb, W. C., 1095, 1101, 14:81Schupp, O. E., 481
SchwackhOfer, 1185
Schwaibold, J. von, 460
Schwalbe, C. 'G., 1891, 1914
Sc~wappach, A., 2527
Schwarz, E. R., ~459
Schwartz, 1948 ,
Schwartz, E., 2313
Schwartz, R., 976
Schweitzer, E., 2617
Schweitzer, H., 1736, 17691
Scott, A. H., 199,8__
I
Scott, A. W.,.121~ _
Scott, J., 2458 -

INDEX TO AUTHORS

Scott, W. W., iii, iv, v, vi, viii, ix, xiv, vi,

16,97,180,185, 233,238,368,413,478,
488, 503, 821, 824, 827, 913, 970, 1024;
1421, 2020, 2180, 2191, 2203, 2214,
2414, 2415
Scribner, B. W., 1914
Seamon, W. H., 580, 917
Sears, G. W., 978
Sedgwick, 2102, 2103, 2104
Seeker, A. F., 517
Seeman, L., 432
Seevers, M. H., 2537, 2538, 2543
Seibert, F. M., 2402, 2407
Seidell, A., vi, 2570
Selenyi, P., 953
Selfstrom, N. G., 1030
Sellars, W. S., 966, 967, 968
Seltz, H., 2312
Selvig, W. f>._., 1628, '1629, 1633, 1635,
1642, 1643
S'elwoQd, P. W., 255, 259
Sendroy, J., 2287
Sharp, R. S., 1984
Sharp, T. F., 2289
Sharwood, W. J., 578, 583
Shattuck, G. A., 2343
Shaw, E. H., 1986
Shaw, H. G., 1179
Shaw, T., 2401
Sheen, R. T., 919, 2092
Shelberg, E. F., 67
Shelton, W., q8
Shennan, R. J., 203
Shepherd, M., 2345
Sheppard, J. R., iv, xiv, 527
Sheppard, S. E., 2364
Sherer, .T., 1945
Sherman, H. C., 1767, 1770
Sherman, M. S., 2465, 2489, 2490
'Shields, J.; 2383, 2384
Shimer, E. B., 985
Shimer, G., 231
Shimer, P. W., 985
Short, M. N., 2458
Shrewsbury, C. L., 2472'
Shukoff, J. J., 2318
Sickman, D. Y., 879
Siedler, P., 74
Siegmann, F., 445, 446
Silberberg, B. H., 2450
Silber mann, J. T., 1185, 1186
Silberstein, J., 958, 968
Silbert, F. C., 2479
Silvn, A. J. F. da, 1782
Silverman, L., 1481
Simmance, J. F., 2391
Simms, H. S., 1768, 1773
Simpson, E. S., 331, 345
. Simpson, M. E., 647
Simpson, S. G., 1095, 1101
Singleton, W., 953

17

Skinkle, J. H., 2459

Skinner, L. B., 11, 482
Skinner, W. W., 194, 459, 891
Skowronski, S., v, xiv, 383, 384, 386, 398,
727
Skraup, Z. H., 531
SIegel, H., 2546
Slossen, 1185
Slotta, K. H., 2529
Slottman, G. Y., 2378
Slyke, D. D. van, 2280
Smith, A. W., 1189
Smith, E. A., 437, 440, 831
Smith, E. F., 202, 720
Smith, E. H., 371
Smith, D. M., 2617
Smith, D. P., .779
Smith, F. L., 2nd, 521
Smith, G. Frederick, v, xiv, 861, 872, 884,
1204, 1205, 1335
Smith, G. McPhail, 984
Smith, Gilbert B. L., 434
Smith, Gilbert, M., 2115
Smith, J. H., 2224, 2225
Smith, J. H. C., 2530
Smith, J. H. F., 203
Smith, J. Lawrence, 6, 36, 344, 534, 799,
881, 882, 1612
Smith, J. M., 7
Smith, M. M., 1096, 1101
Smith, W. H., 2174
Smolczyk, E., 2310
Smoot, A. M., 857, 1046
Snell, C. T., 989
Snell, F. D., 989
Snelling, W.O., 1660
Snodgrass, J., 1769
Sobotka, H., 2493
Soddy, F., 767
Sorensen, S. P. L., 2276
Soloveva, N. A., 634
Soltsien, P., 1778, 1779
Soltys, A., 2526, 2547
,
Solvay Process Company, iii, 240, 2257
Someya, K., 255
Sonden, K., 2405
Sottery, C. T., 2, 1213
Soxlet, F., 196(l, 2029
Spacu, G., l-57, 203, 320, 582
Speakman, T., 2366
Spear, E. B., 465
Spence, D., 1973, 1974, 1975, 1997
Spies, J. R., 2491
Spikes~ W. F., 2457
Spitz, G., 1773
Spring, W., 1188
Squibb, 2142
Stabler, H., 2085, 2098
Stabler, 1i. J., v
Stahler, A" 2470
Staniewski, M., 1188

18

INDEX TO AUTHORS

Stansby, M. E., 423

Stanton, F. M., 1629, 1633
Stapleton, P. L., 811
Starck, G., 405
Stark, D. D., 1340, 1341, 1345, 1841, 1889
Starke, K., 634
State, H. M., 582, 697, 1205
Stead, J. E., 91
Steele, F. A., 1925
Steiger, B., 1218
Steiger, G., 410
Steinbeck, D., 374
Stelling, E., 963, 970
Stenger, V. A., 193
Stephanov, A., 266
SterbaBiihm, J., 588
. Sterling, C., iv, xiv, 619, 625, 628
Sterling, W. F., 1340, 1342, 1345
Stetser, J. B., 221, 229, 230, 231
Stevens, H. P., 1973, 1983, 1985, 1998
Stevens, W. H., 1968
Stief, F. A., 114, 1384
Stimson, F. J., 2609
Stoddard, J. T., 360
Stoddard, W. B., 51
Stohman, F. C. A., 1185, 1186
Stokes, H. N., 2353
Stone, C. H., 2395, 2398, 2399
Stoppel, A. F., 591
Storch, L., 1774, 1781, 1979
Storm, C. G., iv, vi, 1660, 1664, 1671,
1675,,1677, 1693
Stormont, R. T., 2537, 2538, 2543
Stortenbeker, W., 1190
Stover, N. M., 1215, 2494
Strafford, N., 2017
Straub, F. G., 2089, 2090
Straus, J. V., 1053
Strebinger, R., 682, 2521, 2548
Stromeyer, F., 122, 197
Strutt, R. J., 2379
Stuart, A., 2435
Stull, A., 2493
Stull, D. R., 2547
Sucharda, E., 2533
Suggs, 2391
Suitsu, K., 529
Sullivan, E. C., 2166
Sullivan, T. J., 2205
Sullivan, V. R., 1204
Sun, C. H., 225
Sunde, C. J., 273, 2499
Sutermeister, E., 1902, 1903, 2459
Sutton, F. A., 272, 374, 437, 456, 918,
926, 2187
Swanger, W. H., 754 .
Sward, G. G., 1839
Swearingen, J. S., 2537, 2538, 2543
Sweeney, O. R., 2365
Sweetham, 1737

Swift, E. H., 427, 455, 465, 980

Szebelleay, 'L., 474 .1
Szendrii, P., 2504

T
Tabern, D. L., 67 '\
Taufel, K., 634
L
Talbot, H. P., 422, Lt50, 1151, 2238, 223&
Tananaev, N. A., 108
Tang, N. G., 11
Taylor, B. S., v
Taylor, C. A., 1660, 1686
Taylor, G. B., 1671
Taylor, L. V., 2493
Taylor, W. A;, iv, vi, 1919, 2274, 2291
Taylor, W. C., 2166
Technical Association of Pulp and Paper
Industries, v, 1?78
Tennant, J., 2493
Ter Meulen, see Meulen
Thaler, H., 634
Thenard, ,L., 163
Thiollet, R., 2017
Thomas, -T. S., ,2394
Thompson, A., 486
Thompson, C. L., 2453
Thompson, G. W., 85, 957, 1381, 1860
Thompson, H. W., 2025

Thompson, J. J., 1470

Thompson, L., 2367
Thompson, R. T., 165
Thomsen, J., 1167, 1185, 1186, 1190
Thomson, R., 1767
'Thornton, N. C., 2366
Thornton, W. M., Jr., iv, v, 479, 481, 707,
975, 980, 981, 982, 983, 989, 992, 996
Thorp, G., v, 494
Thorpe, T ..E., 691, /17'3, 781, 2205
Thresh, T. C., 159 I
Thurston, R. V., 395
Tickell, F. G., 2093
Tieftrunk, F:, 2399
Tiemann, H. 'P., .2566
Titanium Alloy Manufacturing Company,
991
Titus, L., 2546
Titus, R., 2435
Tobler, F., 2459
Tolloczko, S., 1189
Tomicek, 0., _?65, 479, 628, 2313
Topf, G., 1870
Topf, W., 524
Tortelli, M., 1766; 1778
Tougarinoff, B., 1215, 1218
Touzalin, L. A., ,2389
Town, G. G., 2081
Trautz, O. R., 2485, 2488, 2534
Treadwell, F. P., 307, 403/454, 6,,5, 639,
889, 900, 908, 919,921, 926, 1021', 1036,
2353, 2370 .,:.J __

tJ

INDEX TO AUTHORS
Treadwell, W. D., 2313, 2314, 2318, 2321,
2322, 2328
Tremain, B. E., 281
Trentinaglia, A. von, 1190
Tress, B. J., 2547
Treubert, F., 156
Tribe, B., 2375
rruesdale, E. C., 904
rruog, E., 801, 2501
Tschirch, A., ;1973
Tschugaeff, L., 616
Tseng, C. L., 266
Tsurumi, S., 2547
Tunman, 0., 2459
'rurner, J. L., iv, xiv, 1000
Turner, R. G., 460
Tuttle, J. B., 1975, 1984, 1997, 2174
Tutwei!er, C. C., 2396
Twiss, D. F., 1981
Twyman, F., 2601, 2617
Tyndall; J., 2414
Tyrer, 0., 2579, 2580
U
Ubbelohde, L., 1785
Uhl, A., 2310
Ullmann, B. M., 1644, 2378
Unangst, R. B., 56, 1213
Urbach, C., 2545
Urbasch, S., 98
Urey, B. C., 444
Uslar, B. von, 820
Utz~ F., 1779
Utzinger, M., 2530

v
Vaille, C., 2498
Valenta, E., 1765, 2609
Val!9ntine, W. G., 2453
Vamari, 646
Van Brunt, C., 953
Vance, J. E., 372
Vandaveer, F. E., 2406
Van Belmont, J. B., 2373
Vanier, G. P., 221, 235, 1652
Vanino, L., 156, 432 ,
Van Marum, M., 2369
Van Name, R. G., 2320, 2321
Varrentrapp, F., 2344 '
Vasbi-ttder, B., 2547
Vastagn, G., 180
Vauquelin, L. N., 137, 282
Venturi, 2342
Verdino, A., 2518
Verneui!, G., 252
Verjijl, E. J. A. B., 2314
Viebock, F., 2527, 2528
Vieweg, K, 866

19

Villavecchia, 1782
Violle, J., 1188
Virgili, J. F., 96
Vitali, D., 1949, 1958.
Vogel, A., 968
Vogel, B., 832
Vogt, W. W., 2017
Voit, C., 1041
Volhard, J., 100, 192, 194, 271, 278, 376,
561,577,661,662,824,926, 1201, 2262,
2498
Volman, D. B., 2544
Volta, A., 2375
Voorhees, V., 1908
Vorontzov, R. V., 307
Vortmann, G., 72
Vostrebel, J., 588
Vries, O. de, 1998
Vrigling, J. J., 2546

w
Waage, P., 1192
Wada, 1., 429 '
Wagener, L. R., 2310
Wagner, E. G., 77
Wagner, P., 209
Waidner, C. W., 1726, 2394
Walden, G. B. J,r., 1207
Walker, P. B., 524
Walker, W. B., 2340
Wallace, J. B. Jr., 474, 1206
Wallace, R. E., 2196, 2197
Waller, F., 1769
Wallis, T. E., 2450, 2459
Walt enberg, R. G., 975
Walton, J. B., 991, 2575, 2576
Ward, A. M., 203, 875, 1217
Ward, H. B., 2115, 2458
Ware, G. M., 1057
Warren, W. B., '2344
Warren, W. H., 1950
Warwick, A. E., 157
Washburne, R. N., 481
Wasitzky, A., 2547
Waterhouse, E. F., 342, 343, 344
Waters, C. E., 1984
Wauters, J., 1778
Weatherby, L. S., 180
Weaver, E. R., 2378, 2395
Webb, H. W., 338, 340, 344, 348
Webb, S. K., Mrs., iv, 185
Weber, H., 486, 802
Weber, L. E., 1998
Weber, R., 1978, 1997
Wehmer, P. F., 1911
Wehmhoff, B. L., 1911, 1919
Weibe, H. F.; ~160

20

INDEX TO AUTHORS

Weibl, R., 64
Weigel, W. M., 2454, 2456
Weimar, A. C., 2288
Weinig, A. J., 676
Weiss, G., 1057, 1062
Weiss, R. L., 2318
Welch, F. V., 2457
Weld, F. C., 1318
Weller, A., 75
Wels, R. L., 374
Wels, R. C., 249, 831, 832
Wense, W., 871
'Verner, T. R., 2312
Wesson, D., 1777
Wesson, L. G., 1975
West, E. S., 2494
Westcott, R. P., 2342
Weston, P. E., 2546
Wheatstone, C., 2326
Whip, B., 2546
Whipple, G. C., 2100, 2103, 2105, 2109,
2112, 2114, 2115, 2449, 2458
Whitby, A., 310
Whitby, G. S., 1995, 1998
White, 353
White, A. R., 2396 '
White, A. S., 2576, 2577
White, B. S., 527, 1870, 1874
White, E. V., 2479, 2529
White, L., 527
White, W. E., 521, 1213
Whitfield, J. E., 996
Whitmore, W. F., 2445
Whitton, 578
Wiborg, J., 244
Wichers, E., 719, 724, 736, 744, 747, 752
Widal, F., 2132
.
Wigand, A., 1188
Wijs, J. J. A., 1768, 1846, 2043
Wilcox, L. V.,'176, 187
'Vi.Idenstein, R., 918
Wiley, J. A., 2315
Wilfarth, R., 1974
.1
Wilkie, J. M., 51,6
Wilkins, E. S., 521
Will, 2344
Willard, R. R., v, xiv, 11, 211, 253, 254,
311, 407, 474, 573, 582, 802, 884, 917,
944, 1205, 1206, 1207, 1208, 1217, 1470,
1473, 1474, 1477, 1478, 1479, 1615,
2316, 2319, 2320, 2321, 2322
Willard, M.L., 2546
Williams, 1654, 1834
Williams, A. W., 2133
Williams, F. W., 310, 571, 1412
Williams, R. J., 2463, 2547
Williams, R. S., 2459
Willits, C. 0., 862, 874
Willoughby, C. E., 521

Willson, T. L., 213.76

Willson, R. M., ,816, 817
WiIIstatter, R., 21530
Wilson, E. B. Jr'" 2320
Wilson, J. A., 1778, 2459
Wilson, J. R., 2028
Wilson, J. M., 2161
Wilson, L. A., iii, xiv, 197, 1054, 1079
Wilson, N. F., 1185 .
Wilson, R. E., 1744
Wilson, S. D., 225
Winchell, A. N., 2435
Winchell, N. R., 2435
Winkel, A., 1215
Winkleman, R. A., 1998
.
Winkler, C., 429, 2265, 2344, 2346,' 2353,
2358, 2359
Winkler, L. W., 456, 633, 892, 2077,
2079, 2082, 2369, 2494
Winlock, 'J., vi, 2548
Winslow, C.-E. A., 2133
Winter, O. B., 56, 407, 521
Winter,stein, E., 456
Wintersteiner, 0., 115, 2500, 2503
Winton, A. L., 2458, 2459
W~se, L. E., 683
WIthrow, J. R., 2365
Withrow, R. B., 2472
Witthaus, R, A., 1945
Wohler, F., 137
Wolbing, R., 1218
Wolf, A. G., 1318
Wolfe, R. A., 2599, 2617
Wolff, C. A., 2344
Wolff, Ranllo 2044
Wolff, R., 273, 457, 830
Wolff, P., 21317 /
Wolle, .F., 2115 1
Wo;>d, A. :E., 479, 481
Wood, R. p., 2388, 2389
Wood, T. LB., 874
Woodbury, 1741
Woodgate; J., 7
Woodman, A. G., 422';"'1786
Wooten, L. A., 2318, 2319
Wooton, W.O., 976
Worst all, R. A., 1849, 2376 '
Wright, C. R., 1764
Wright, G:=- F., 2479, 2529
Wright, S. L., 2530
Wright, T. A., iv, xiv, vi, 835,1301
Wright, W. C.; 2315
Wurm, 0., 157
Wurster, C., 2370, 2405
Wyatt, F., 2376
Wyatt, G. R., 2547
I
Wyman, A., 1.143, ::;146, 1151, 2207,
2221, 223!T - '
Wyrobouff, G., 252

INDEX TO AUTHORS
y
Yagoda, H., 593
Yant, W. P., 1963, 2364, 2405, 2547
Yatsevitch, M., 2552, 2569
Yntema, L. F., 255
i"oe, J. H., 487, 1213
Yonng, P., Miss, 253, 254, 474, 582,
.'144, 1205, 1206, 1207, 1208, 1473,
~474, 1478, 1479, 2316
Young, W. G., 272
7nrow, L., 1975

21
Z

Zakrzewski, 1., 1189

Zeisel, S., 1893, 2527
Zeismer, A., 180
Ziez, E. G., 179, 180
Zimmerman, P. W., 2365
Zimmermann, A., 2459
Zimmermann, C., 299, 476, 482
ZintI, E., 2322
.
Zoller, H. F., 2288
Zombory, L. von, 582

SUBJECT INDEX *
A
Abbe refractometer, 1765
Absorption bulbs:
carbon dioxide, Fleming, Geissler,
Gerhardt, Liebig, Vanier, 221
gas analysis, Friedrich, Hankus, Nowicki-Heinz, Varrentrapp, Winkler,
Wolff, 2344, 2345
Absorption
, spectra, 2614 ff.
of the rare earths, Table, 257
spectrum,
carbon monoxide in air, 219
Accelerators, in rubber, 1981, 2016
Acetates, analysis, 2251
Acetic acid,
analysis for acetone, formic acid,
furfurol, hydrochloric acid, metals, sulfuric acid, sulfurous acid,
2248-2251
est. in white lead, 1860
method for nitrite, 631
method for separation U from V,
1021
specific gravity table, 1156, 2252
Acetic anhydride,
analysis of,
acidity method, 2247
aniline method, 2248
method for silica, 816
Acetin method, for determining glycerol,
2150
Aeetone,
aldehydes in, 2136
analysis of, 2134
,
Messinger's method, 2136
extraction of rubber, 1969, 2002'
in acetic acid, 2250
Potassium permanganate test, 2136
residue, 2135
specific gravity, 2135
water in, 2135
Acetyl,
determination, micro, 2523
value for oils, 1777
Acetylene,
properties of, 2348, 2376
test for Pd, 722
Acid, free
in ether, 2270
in pressure of iron salts, 2245
in soap, 2032

Acidimetric and' alkalimetric methods

for metabisulfites, sulfites, sulfurous
acid, 925
phosphorus, 69
Acidimetry and alkalimetry, 2191 fl'.
see' also conductometric methods' Potentiometric methods.
'
Acidity in explosives, 1182
~n paper, total, 1910
III rubber,
acetone, 2135
acetone extract, 1971
alcohol, 2139
in starch, 2157
in sulfur, 934
crude, 941
in water, 2072
pH scale, of, 2274
Acid number,
Chinese wood oil (Tung oil), 1847
"linseed oil, 1844
varnish, 1852
Acid value, of soap, 2044
Acids,
analysis of,
acetic, 2248
boric in borates, 2245'
carbolic, 2253
carbonic, 2252
chlorosulfonic, 2244
citric, 2253 /
formic, 2246,
hydrochloric,
determination of total acidity, ar-,
senic, barium, chloride, chlorine,
nitric acid, snlfuric acid silica,
total solids, 2203
hydrofluoric,
"
of
d~terminations
acidity, hydrofluosilicic acid, sulfuric acid, sulfurous acid, 2209
nitric,
determination of acidity, fIte chlorine, hydrochloric acid, iodine,
nitric a~d nitrous acids, ,non-volatile solids, sulfuric acid, 2211
in arsenic acid,
ferrous sulfate method, 2218
in oleum and mixed acids, 2217
in phosphoric acid, 2218
ih sulfuric'acid, 2216
.../'-

* Pages with numbers greater than 1300 are in Volume' II.

22

SUBJECT INDEX
Acids, nitrous,
permanganate titratiol). of, 653,
2219
oleum and mixed acids,
complete analysis, determination of
total acidity, lower oxides, nitric
acid, sulfuric acid and free SO.
calculations, table, 2233-2243
organic, 2246 ff.
oxalic, 2253
phosphoric acid,
analysis of, 2224
specific gravity, see tables
sulfuric acid,
analysis of, 2228
determination of antimony in, 83
determination of lead, iron, arsenic,
zinc, selenium, hydrochloric and
nitric acids, 2229-2233
tartaric acid, 2254
Acide,
fatty, from oils, 1794
formulae for diluting or strengthening
of, 2236
.
indicators for determination 0:1'\ 2192
number in oil analysis. See oils.
reactions-tables of, 1128-1135
tests for, 1112
reagent solutions, 1193
standards, preparation of benzoic, hydrochloric, sulfuric, 2194-2197
'titration of, 2191 ff.
See also under individual acids as
acetic, boric, etc.
weighing,
concentrated acids, 2200
Blay-Burkhard
Burette,
2202
(Fig. 3041
Dely Tube, 2201 (Fig. 302)
glass bulb, 2203
Lunge-Ray pipette, 2201 (Fig.
301)
snake tube, 2202 (Fig. 303)
dilute acids, 2200
Adsorption Indicators,
brom phenol blue, 582
dichlorfluorescein, 273
eosin, 192, 457
fluorescein, 273
for halide titration,
bromide, 192, 2499
chloride, 173, 2499
iodide, 457, 2499
for silver, 830
rhodamine, 6 G. 830
tetrahydroxyquinone, 919
Agar,
eosine-methylene blue, 2121
Hesse, 2123
nutrient, 2121
stock, 2121

23

Air,
composition of, 629
examination of. See Gas Analysis.
oxygen in,
hydrogen
combustion
method,687
phosphorus method, 685
pyrogallate method, 686
pound per cubic foot coal burned,
calculation of, 2384 .
Albuminoid ammonia, determination of,
in water, 2050
Alcohol,
ethyl,.
acidity, 2139
aldehydes in, 2139
detection of, 2137
Berthelot's test, 2131
ethyl acetate test, 2137
iodoform test, 2137
methanol in, 2148
determination of, distillation and
evaporation methods, 2137, 2138
purification of, 1771
specific gravity tables, 2142-2146
test for Pd, 722
methyl, determination of, 2147
in ethyl alcohol, 2148
Alcohol-ether method, separation Li.
from K, N a, 890
Alcoholic-potash, extraction of rubber,
1982, 2003.
Alcohol-insoluble matter in soap, 2037
Aldehyde, determination of, 2139
in alcohol, 2139
Alizarin red 407
Alizarine S,
test for aluminum, 2
Alkali arsenates, arsenic determination
in, 96
Alkali, combined, in soap, 2029
free, in soap, 2032
ill water, 2097
standard for acidimetry and alkalimetry, 2197
Alkali, fluorides, determination of fluorine, 417
Alkali, xanthate, separation cobalt from
nickel, 310
Alkalies, analysis of,. see soda, sodium
hydroxide, etc.
Alkalies, chapter on, 861
detection of caesium, lithium. potassium, 861-862
rubidium, sodium, 875, 876
estimation in,
.
alundum, 36
clay, 2163
glass, 2166
Portland cement, 1612
preparation of the sample,
brines, 864

24,

SUBJECT INDEX

Alkalies, fertilizers, 863

organic compounds, 864
plants, ashes of, 864
potash salts, 863
rocks and insoluble mineral products, 863
saline residues, soluble salts, 864
soils, 863
water, 864
separation from,
barium, calcium, strontium, sulfuric
acid, 121, 865
each other, 884
heavy metals, 864
magnesium, 866
test for,
irridium, 730
palladium, 721
Alkalimeter,
Mohr's, 238
Schroetter"s, 238
Alkalimetric method for determining,
persulfates, 922
phosphorus, 697
. strontium, 902
Alkaline earths. See Barium.
separation from one another, 119
silica tes in soap analysis, 2034
in water, 2096
Alkalinity determination,
in sodium sulfide, 2183'
in starch, 2157
in water, 2066
total, trisodium phosphate, 705
Alkaloids,
detection, 1951 ff.
Allotropy, 2556
Alloy steel. See Iron and Steel.
Alloys, see also Metallography.
decomposition of, 1348
determination of aluminum in, 1359,
1389, 1394
antimony in, 83, 1369, 1382, 1385,
1390, 1401
arsenic in, 1371, 1384, 138\5, 1402
beryllium in, 146, 148
bismuth in, 160, 1376, 1387, 1399
cadmium in, 1371, 1396, 1400, 1403
carbon in, 1380, 1468
chromium in, 301, 1406, 1411
cobalt in, 1409
copper in, 1352, 1362, 1376, 1378,
1382, 1384, 1393, 1394, 1400, 1402
insoluble residue in, 1408
iron in, 1357, 1365, 1373, 1377, 1379,
1387, 1396, 1401, 1404, 1407, 1411
lead in, 505, 1352, 1364, 1375, 1381,
1384, 1387, 1396, 1399, 1404
magnesium in, 1395
manganese in, 1358, 1374, 1380, 1398,
1407, 1412

Alloys, mQlybdenum in, 1411

nickel' :in, 136/), 1374, 1377, 1378,
1390, 1404, 1405
phosphorus in, 698, 1366, 1368
silicon in, 802, 11379, 1408, 1412
silver in, 1402,
sulfur in, 1370,c 1379, 1409
.
tin in, 1356, 1362, 1363, 1375, 1383,
1385, 1393, 1396, 1399, 1404
titanium in, 42
zinc in, 1366, 1375, 1377, 1392, 1401,
1403
non-ferrous (See also Pt, Au, Ag)
accuracy of analysis, 1350
analysis of,
,
alloys of .antimony, copper, lead,
and tin, 1381
impurities in, 1387
aluminUJIl alloys, 37-52
,
determination Si, Cu, Fe, Ti, N~
Pb, Zu, Mg, Mn, Cr, Sn
babbitt metal,
Cu in, 398
,Sn in, 968 .
battery plate, 1393
beryllium-copper alloys
determination of AI, Be, Au,
/
Fe, Ni, Si, 146-148
brass and sand castings,
Cu and Pb, 392
Zn in, 1084
bronze and bearing metal, A. S.
T. M., 1362-1372
copper, nickel, zinc alloys, 1373 ff.
fusible metals, 1399,
German silver, 1373
gold, dental, 754-757
magnesium/alloys, 539-553
determination of AI, Mn, Zn,
; Si, Cu, Cd, Fe, Ni, Sn, Ca,
Pb, Mg, pp. 539-553
manganese bronze, 1352-1362
monel metal, 1378
nichrome, 1405
silver alloys, 1402.
solder, 1390
silver, 1402
steIlite, 325
white metals,
'antimony in, 84, 85, 114
arsenic in, 112, 114
tin ,in, 973
Woods metal, 1399
Zinc base alloys, 1394,
classification of, 1348
examination of, ,1348
general procedurElf 1349
chart of;.::1349
sampling, 1350
pyrophoric, 246

SUBJECT INDEX
}upha benzyldioxime method,
for determining nickel, 617
separation of nickel from cobalt,
310
test for nickel, 615
Alpha cellulose. See Cellulose.
Alpha ray, method for radium, 759
instrument (Fig. 88), 759
Altitude, effect on solubility of oxygen,
- 677
Alumina, calcined, analysis of, 26
Alumina,
in bauxite, 25
in cement, 1602, 1614
in chrome green, 1876
in glass; 2166
in .nitrate of soda, 644
in its ores, 19, 20
in Prussian blue, 1873
in ultramarine blue, 1872
in sand, 806
in silicate of soda, 805
in silicon carbide, 815
in slag, 2022
in titaniferous ores, 986
in water, 2059
combined in aluminum sulfate and
aluminum salts, 14
Alumina and iron oxide. See also Iron
oxide and Alumina
in blanc fixe, 136
in composite white paint, 1868
in glass, 2166
in ores, 20
in Portland cement, 1614
Alumina ores,
.. arsenic determination in, 106
commercial valuation, 17, 19
determinatiotf of
alumina, 20
calcium oxide, 21
iron, ferrous, 20
total, 20
loss in ignition,' 19, 23, 26, 29
magnesium oxide, 22
manganous oxide, 22
moisture, '19
phosphorus pentoxide, 22
silica, 19
sodium oxide, 21
titanium oxide, 21
vanadium, 23
zirconium oxide, 2.3
Aluminon,
colorimetric method for AI, 55
reagent, 1213
test for AI, 2
Aluminum
detectio.n of, 1
alizarin S test, ';l
aluminon test, 2

25

Aluminum
cobalt nitrate test, 2
1, 2, 5, 8-, hydroxyanthraquinone
test, 2
estimation, 3
general procedures,
colorimetric,
aluminon method, 55
in animal tissues and foods, 56
lake development, 58
reagents, 56
separation from other consti-
tuents, 57
gravimetric,
by hydrolysis,
with ammonium hydroxide,
8
with sodium thiosulfate, 10
with urea, 11
by precipitation,
as aluminum chloride, 12
as oxinate, 12
as phosphate, 13
volumetric,
determination of combined
alumina, 14
.
free alumina or free acid, 16
of aluminum, oxine method, 12
estimation, s,pecial procedures of aluminum metallic, 50
hydrogen evolution method, 51
in alloys, 1359, 1389, 1394
in bauxite, calcined alumina, aluminum hydrate, 19-29
reagents and standard solution,
29-32
in bismuth bronze, 1378
in bronze, 55, 59, 1359
in ferrovanadium, 1498
in iron and steel, 52, 53, 1482
in magnesium alloys, 539
gravimetric, 540
potentiometric, 540
in manganese bronze, A: S. 'T. M.,
1359, 1360
in metallic aluminum dust, 50, 51
in nickel, 627
in presence of iron, phenylhydrazine
method, 60
in tinbase metal, 1389
in zinc base alloy, 1394
impurities in metallic aluminum, 38
industrial application, 1
occurrence, mineral and ores, 1
preparation and solution of the sample,
4

general procedure for ores, "4

fusion method (sodium carbonate),
4
metallic aluminum and its alloys, 5

26

SUBJECT INDEX

Aluminum
properties, 5
aluminum phosphate, 14
sampling aluminum and its alloys, 31,
38
separation
from beryllium, 1
hydrochloric acid method, 139
oxine method, 139, 141
from chromium, 1
from iron, 6
from iron, manganese, nickel, phosphoric acid, silica, titanium, uranium, zinc, copper, mercury, bismuth, rare earths, beryllium, 6
from manganese, basic acetate, 558
from manganese, cobalt, nickel, zinc,
alkaline earths and alkalies, 1
from nickel, 8
from phosphates, arsenates, fluorides,
borates, Mo, Cb, Ta, Ti, V, Be,
U, oxine method, 8, 12
from phosphoric acid, 1
from silica, 6
from titanium, 1
from uranium, 1
from zinc, 1
from zirconium, rare earths, manganese iron, titanium, 8
solubilities, metal and its alloys, 5
oxide, 4
special procedures: for small quantities, 59
traces, detection and estimation of,
with alizarine, 2
Aluminum Company of America
methods for alloys, 31-52
methods for aluminum, 31-52
methods for bauxite, calcined
alumina, aluminum hydrate, 1929
methods for beryllium, 141
Aluminum hydrate, analysis of, 28-29
Aluminum precipitation, of Cu, 355, 358
Alum~num reduction method for nitrates,
640, 641
I
Aluminum sulfate, analysis of, 14-16
Alundum, analysis of, 32-31
separations in, 33
Amalgamated zinc, determination of
mercury in, 582
Amalgams, 515
American Brass Co.
analysis of berylliumcopper alloys, 146-148
American Cyanamid Co. methods
for converter efficiency, 659
for simple and complex cyanides,
663-668
for trisodium phosphate, 703705

American Society for Testing Materials

(A. S. T. M.)
methods for testing I of
alloys (nonferrous), 1348 ff.
cement, 1597 ff.1
coal and coke, :~620 ff.
ferro.alloys, 1011, 1485 ff.
oils, 1705, 1713; 1719, 1727, 1732,
1737, 1743, 1744, 1750, 1752,
1758, 1761
petroleum, 1807, 1814, 1818, 1823,
1830, 1839
steel, 1421 ff.
turpentine, 1850
zinc, 1071
Amino-nitrosophenyl-hydroxylamine
method for iron, 468
Ammonia
albuminoid,
in water, estimation of, 2050
aqua, volumetric method for, 2270
conversion to RNO., .
apparatus (Fig. 75),659
efficiency of, 659
fixed, in ammoniacal liquor, 638
free in water, estimation of, 2049, 2077
gaseous,
absorption of, 2346
determination of, 2360
in illuminating gas, 2396
properties of, 2360
gravimetric method for,
determining as platinochloride,
638
in fused zinc chloride, dct 'no of, 1083
in latex, 1968
Nessler's test for, 630
specific gravity table, 1157, 2269
test for palladium; 721
total, in ammonia'cal liquor, 637
traces, determination of, in gas, 2396
volatile, in ammoniacal liquor, 637
volumetric method, 636, 637
Ammoniadynamite, analysis of, 1667 ff.
Ammonia method for copper, 378
Ammonia precipitate, fotal, _in silicon
carbide, 814
-~
Ammonium chloride,
standard solution, 1199
test for iridium, 730
test for palladium, 722
test for platinum, 712
hydroxide metliod,
for determining traces of copper, 378
test for iron, 463
test for osmium, 748
test for palladium, 721
test for rhodium, 742 I
iodide method .10r- tin oxide, 960
iridium chlo~rge, determination. of
iridium as, 732

SUBJECT INDEX
,Ammonium,
magnesium phosphate,
acid titration of, fi34
direct precipitation of, 696
properties of, 534
manganese phosphate, 566
molybdate, standard solution for lea'd,
1201
nitrate washing solution, for aluminum
hydroxide, 32
nitrate, pyridine in, 655
apparatuB for (Fig. 74), 655
persulfate method for manganese, 572
phosphate method,
for separating magnesium from
alkalies, 531, 867
for manganese, 566
for zinc, 1058
for zirconium, 1102
phosphomolybdate
method for phosphorus, 694
precipitation of, 694
weighing of, 695
picrate, analysis of, 1683
platinochloride method for determining
ammonia, 636
platinum, 719
salts,
effect on magnesium precipitation,
533
effect on sulfur determination, 908
detection 'of ammonia in, 630
.
determination of ammonia in, 637
mixtures, ammonia in, 634
salicylate, test for silver, 819
sulfide group, separation of, see Separations under Element in question
sulfide method for mercury, 577
test for ruthenium, 737
test for vanadium, 1031
table, sp. gr., 1157, 2269
thiocyanate test for cobalt, 306
test for 'iron, 462
Ammonium zinc phosphate, 1058
Atnorphous sulfur, determination of, 934
Ampere, definition of, 1174
Amyl alcohol. See Fusel Oil.
Anemometer, 2342
Aniline method, for determining acetic
anhydride, 2248
Animal Dand vegetable oils, 1764. See
Oils.
analysisof, 1701 ff.
test for, 1775
Anim~l tissues, aluminum in, 56
Anthraquinone,
1, 2, dihydroxy-see alizarin
1, 2, 4 trihydroxy-see purpurin
test of bituminous substances, 1562
Antifiuorescents, tests for, 1776
Antilogarithms, 1181-1182

27

Antimonous and antimonic salts, distinction between, 64

Antimony
detection of,
by hydrogen sulfide test, 63
by hydrolysis, 64
in minerals, 64
traces, 64
estimation, general procedure, 64
gravimetric,
as metallic antimony, electrolytic,
73
as oxide, 72
as trisulfide, 72
volumetric,
bromate method, 74
ceric sulfate, 78
indirect evolution method, 77
oxidation with iodine, 76
permanganate method, 76
potassium iodate, 78
potassium iodide method, 75
estimation, special procedures:
determination,
in alloys with Sn and Pb, 83
in ash of rubber, 2006
in battery plate, 1393
in brass and bronze, 76
in bronze, A. S. T. M., 1369
in commercial "arsenic," 111
in copper, 387
in ferro tungsten ana tungsten
metal, 1015, 1500
in metallic lead (pig lead), 1416
in rubber goods, 2006
in soft solder, 83, 1390
in stibnite, 77
in sulfuric acid, 83, 2231
in tartar emetic, 82
in tin and lead alloys, 83, 1382,
1385
in white metal, 84
rapid method, 85
in Wood's metal, 1401
occurrence, I)linerals, ores and alloys,
63
preparation and solution of the sample: 64
alloys
of antimony, tin and lead, 66
of hard lead, 67
of low-grade oxides, 65
of mattes, slags and speisses, 66
of ores, fusion, 66
of sulfides, 65
of organic compounds, 67
properties, antimony trisulftde, 72
reduction to trivalent state, 74, 78
separation
.
from arsenic and tin, 68
distillation procedure, 69, 1385

28

SUBJECT INDEX

Antimony
Arsenic,
from bismuth, cadmium, copper, lead
permanganate tmethod, 1210
and mercury 67
silver method, 100
from members of the- subsequent
estimation, special :procedures:
groups, aluminum, chromium, codetermination I
balt, iron, manganese, nickel, zinc,
in alloys, brass and bronze, 1371,.
alkaline earths and alkalies, 67
1384, 1385, 1402
from molybdenum, 71
in brimstone, ~33
from silicon, 71
in bronze, A. S, T. M., 1371
from tin in alloys, 68, 71
in commercial "arsenic," 109
solubility of the element and its oxin copper, 112, 387
ides, 63, 64
,in gases, 2417
solution, for standard strains, 80, 1199
in hydrochloric acid, 89, 2204
traces of, in refined copper, 387
in iron, :).12
,
traces, determination of, 78
in lead-base alloys, 1384, 1385
distillation, 81
in metallic lead, 1414
generator, 79
in molybdenite and wulfenite, 5~l
method, 80
in nickel, 625
{J
reagents, 79
in organic substances, 115
test paper, 79
in pig lead, 1414
Antimony chloride-ferric chloride, method
in presence of vanadium, 1034
for Os, 897
'
in steel, 112
Antimony electrode, 2302
in sulfur, 933
Antioxidants, for rubber, 2016
in sulfuric acid, 89, 2230
Apothecaries' weight table, 1170
in tin-base alloys, 1384, 1385
Apparatus. See method or process in
in tungsten ores, 1003
question, e.g. distillation, reductors,
in white metal, 112
etc.
/iil. Wood's metal, 1402
volumetric, calibration of, 1219-1227
impurities in "arsenic" (arsenous oxAqua ammonia, 2269
;
ide), det'n of moisture, antimony,
Arc, spectra, 2598
arsenic, calcium, cobalt, copper, iron,
Tables of lines, 2602-2603; 2604-2607
lead, nickel and zinc oxides, silica,
Areas,
sulfuric acid, 109-111
, ..
analysis, by measuring, 2452
industrial application of methods, 88
formulae for determining, 1170
occurence, minerals, ores, 87
triangle, rectangle, parallelogram, trapreparation and solution of 'the sample,
pezoid, circle, 1170
88
Argon in the atmosphere, 2348
alkali arsenates, 91
properties of, 2380
arsenic aci,d, 91
" Armco" iron, 808
arsenious oxide, 89
Arsenates, alkali, 91
copper, 91
Arsenates, distinction from arsenites, 87 '
hydrOchloric and sulfuric acids,
Arsenic,
89 1
detection of,
iron,' 91
- by Gutzeit test" 87 ,
lead arsenate, 9.Q
by hydrogen sulfide, 87
organic matter, 89~ 91.....
volatility test, 87
pyrites ore, 88
estimation, general procedures:
zinc arsenite insecticides, 90
gravimetric,
propertie9, 87
determination as arsenic trisulseparation, fide, 96
from antimony and tin, 68, 95
determination as magnesium pyrofrom other el,ements,.91 ff.
arsenate, 96
by distilla~ion, 69, 91
volumetric,
apparatus for, 92, 93
bromate method, 100
hydrazine method, 94
ceric sulfate metliod, 100
solubility of oxides, sulfides, salts, 88
iodate method, 99
solution, standard, 1199 I
iodine method, 99
Arsenic traces, / __
in presence of mercuric salts,
colorimetric. -determination (arseno99
molybdate) 116, 2503
Mohr's method, 99

SUBJECT INDEX
Arsenic traces,
determination of, by modified Gutzeit
method, 101-108
apparatus, 105
micro metho"d, 2504
procedure, 107
reagents, 103-104
stains, 103
in acids, hydrochloric, nitric, sulfuric,
105, 106
in baking powders, canned goods, meat,
organic matter, etc., 107
in ores, alumina bearing, bauxite, cinders, pyrites, 106
\in organic matter, 107
in paper, 1919
in refined copper, 387
in salts, sodium chloride, magnesium
sulfate, etc., 106
in tartaric acid, 2255
Marsh test for, 87, 109
Arsenic acid, nitric acid in, 2218
Arsenious,
. acid, determination of, with iodine, 99
chloride, volatility, 88
oxide, commercial, determination of
moisture, SO., As2 0" non-volatile,
SiO, Pb, Cu, Fe, Ni, Co, Zn, Sb,
CaO, 109-11:!for determining bromates, 193

primary standard, 1203

Arsenious method for determining peroxides, 2180
Arsenite reagent 0.1 N, 453, 1203
Arsine, 2347
properties of, 2367
Asbestine, paint pigment, 1866
Asbestos, composition of, 2417
Assay, fire_ See Fire Assay.
Ash,
in black paint pigments, 1877
in 'coal and coke, determination of,
f637
fusibility of, 1645 ff.
in explosives, 1661, 1667, 1678
in grease, 1824
in linseed oil, 1845
in paper, 1905
in paper pulp, 1896
in starch, 2157
in sulfur, 941
i{l varnish, 1852
of rubber, 1971, 2014
sulfur in, 2006
analysis of, 2015
phosphorus in, 1644
solution of, 864
Asphalt, 1853: See Bituminous substances

29

Assaying of gold and silver. See Fire

Assay of Gold and Silver, 835 ff.
wet assay for gold, 435 .
Astringency, of tannic acid, 2159
A. S. T. M. See American Society for
Testing Materials
Atmosphere, composition of, 629
Atomic weights, international table of.
See front cover
of rare earths, average, 259
Atropine, 1958
Aurin tricarbo!Kylic acid. See Aluminon.
Available chlorine, in bleaching powder,
278
Available lime, 212
Available oxygen, 675
apparatus (Fig. 77), 675
Available sulfur, in brimstone; 934
Avoirdupois weight, table of! 1169
B

Babbitt metal,
analysis of, 968
copper in, 398
Bacteria,
coli, 2125
snbclassification of, 2125
confirmed test, 2127
presumptive test, 2126
reaction classification, aerogenes
group, 2126
total in water, 2123
typhi, 2131
subclasses, 2132
Widal test, 2132
test 2132
Bacteriological exam. of water, 2166
Baking powder,
arsenic determination in, 107
carbon dioxide in, available, 237
'residue, 237
Balances,
Ainsworth (Fig. 132), 1227
for fire assay, 841, 844
micro, 2461
Westphal (lJ'ig. 239), 1704
Balata, 1966, 1992
Ball mill, 1630
Barite,
analysis of, 129
determination of,
barium sulfate, .barite, free from
strontium, 130
strontium present, 131
carbon dioxide, 135
fluorine, 135
iron, 134
iron and alumina, 134 .
lime, 134
loss on ignition, 136

30'

SUBJECT INDEX

Barite,
magnesia, 134
strontium sulfate, 131
sulfur, as S03, 133
solutions for, 130
Barium,
detection of,
by flame test, 117
spectrum, 117, 120 (see plate fac-,
ing p. 119)
with saturated solution of calcium
or strontinm sulfates, 117
with soluble chromate or fluosilicate,
117
estimation, general procedures, ,124
gravimetric, determination as carbonate, 126
as chromate, 125
as sulfate, 127
volumetric, acid titration of carbonate, 129
dichromate method, 128
permanganate method, '128
potassium iodide method, 128
potassium sulfate method, 129
indicators for, 129
estimation, special procedures:
determination in insoluble residue,
125
in glass, 2168
in ores, 125
in silicates, 125
ore, valuation of. See Barite.
rubber, 1989
industrial application, 118
occurrence, ores, commercial products;
117, 118
preliminary tests, 120
preparation and solution of the
sample,
carbonates, insoluble residue,
sulfates, 118
sulfides, soluble salts, organic
matter, 118-119
properties, barium sulfate, 127
barium chromate, 126
'
separations, general considerations,
119
from alkalies and magnesium by the
oxalate and sulfate methods, 121
from calcium, 2168
from calcium and strontium, 122,
123, 2172
from lead, 124
from molybdenum and from P.O"
124
from strontium, 124, 126
of the alkaline earths, 119
solubilities, barium compounds, 117,
118, 121
traces, detection of,

Barium,
by :flame and spectrum, 117
by microscope, 118
Barium acetate, 2335
carbonate,
,:
analysis of, 2172
in rubber, 2010 I
method, removal' of Fe, AI, etc.; 174
precipitation of uranium, 1018
chloride in hydrochloric acid, 2264
method for sulfate (volumetric),
919
-potassium chromate method for sulfur determination, 918
test for chrqmate, 282
for Se and Te, 776
'chromate, determination of chromium
as, 125, 287
properties of, 126, 287
solubility of, 126
-thiosulfate method for sulfur, 919
hydroxide method for separating magnesium from alkalies, 867
ores, analysis of, 129
sulfate, apparatus for filtering of,
(Fig. 110), 911
decomposition for sulfur det'n., 907
in barit, 130, 131
in blanc fixe, 136, 1867
in Iithopone, 1863
precipitation of (Fig. 109a), 910
properties and solubility of, 127
in rubber, 2006
in titanium pigments, 1863
Barker '8 hydrometer, for CO, determination in carbonates, 242
B. coli, examinatio)l of, 2125
B. typhi, examination of, 2131
Barometric ,pressure and distillation
tests, 1711
Barytes. See Barite.
Bases, in bleaching powder, 2160
Basic acetate method, for precipitating
aluminum and iron ~(s.eparation from
manganese etc.), 558
-''-Basic alumina in aluminum salts, determination of, 16
Basic carbonate of bismuth, determination, 154 '
,
Basic carbonate of lead, determination,
524
(
Basic nitrate, precipitation, of bismuth
as, 154
Basicity, of aluminum sulfate, 16
Battery plate, 1393
Baume, conversion degr,ees to specifie
gravity, 1168 r -Bauxite,
- -'analysis of,

SUBJECT INDEX
Bauxite,
moisture, loss on ignition, silica,
iron oxides.
alumina,
titanic
oxide, calcium and magnesium
oxides, MnO, P,O., V,O" 23-26
arsenic determination in, 106
Bead test, borax,
for cobalt, 306
for manganese, 555
for titanium, 976
. table, 1109
Bead test, mic~ocosmic salt,
for 'silica, 795
table of, 1109
Beam test, for matter suspended in ga~,
2414
.
Bearing metal,
Bronze, A. S. T. M., 1362
Bechi's test 'for cottonseed oil, 1777
BellJar filter (Fig, 60), 563
'Benze:pe,
micro determination, 2547
Benzidine acetate, test for gold, 432
hydrochloride method, for determining sulfates, 920
reagent, 1213
a- Benzil dioxime,
method for nickel, 617
test for nickel, 615
Benzoic acid standard for acidimetry,
2197
a,Benzoin oxime,
determination of molybdenum,
591
separations with, 587
reagent, 1213
Be.-yl, 137
Beryllium,
detection of, 137
determination, general, 138
colorimetric,
cur cumin, 144
1,2,5,8,-oxyanthraquinone method,
143
gravimetric,
hydroxide method, 139
oxine method, 141
phosphate method, 141
tannic acid method, 142
volumetric,
quinalizarin, 145
determination, special methods,
beryllium-copper alloys, 146-148
determination of
aluminum, 147
beryllium, 146, 147
copper, 146
iron, 146,
nickel, 147
silicon, 147

31

BeryUium,
separations from,
aluminum, 139, 140, 141
RCl method, 139
oxine method, 139
chromium, 143
hydrogen sulfide group, 139
iron, 139, 140, 141, 142
rare earths and chromium, 139,
143
silica, 139
thorium, 143
titanium, 139, 143
tungsten, 143
zirconium, 139, 143
Bicarbonate, determination of, 2252
in water, 2095
Bichloride of tin, hot water precipitation
of tin in, 962
Bichromate method for determining glycerol, 2153
Bimetallic electrode pairs, 2320
Bismuth,
detection of, general procedure, 149
blowpipe test, 150
reducing agents, 149
estimation, -general procedures:
colorimetric,
bismuth iodide method, 159
cinchonine-potassium iodide
method, 158
gravimetric,
determination as basic chloride,
153
as chromic thiocyanate, 156
as complex salts, 157
as metal, by cyanide reduction,
156
as metal, by electrolysis, 156
micro, 2547
as oxide, Bi,O" 154
as sulfide, Bi,S" 155
volumetric,
permanganate titration of oxalate.
, 157
estimation, special procedures:
in alloys and metals, 160
in bismuth bronze, 1375
in .lead bum on, 159
.
in metallic copper, 382, 384
in metaUic lead, 159
in ores, mattes, 161
in pig lead, 1413
in tin-base metal, 1388
in Wood's metal, 1399
industrial application of methods, 150
occurrence, 149
organic precipitants for, 157
preparation and solution of the sample,
alloys, lead bullion, refined lead, ores,
150-151

32

SUBJECT INDEX

Bismuth,
properties, lim ff.
basic carbonate, 155
basic nitrate, 154
hydroxide, 155
oxide, 154
oXY'chloride, 153
sulfide, 156.
separations, aluminum, chromium, iron,
cobalt, manganese, magnesium,
nickel, zinc, alkaline earths, alkalies, 152
from arsenic, antimony, tin, molybdenum, selenium, tellurium, 152
from cadmium and copper, 152
from lead, copper alid cadmium, 154
from lead and mercury, 152
oxybromide method, 152
solubility, metal and salts, 150
traces, determination of, 158, 159
standard solution for, 158, 1199
Bismuth bronze, 1375
Bismuthate method,
for determining cerium, 254
for determining manganese, 562
application to large amounts
of manganese, 563
reagents for, 567
in bronze, 1358
in ferromanganese and man
ganese metal, 565
in manganese ores, 563
electro metric, 577
in nickel, 627
in silico manganese, 568
in steel, 491
in water, 2063
Bismuthyl perchlorate, precipitant, 2335
Bisulfate fusion. See chapter "Decomposition of Sample," Vol. 2
See. under individual substances, e.g.
columbium and tantalum, etc.
Bisulfite. See Metabisulfite.
Bisulfite liquor, paper industry, 1929
Bituminous substances, asphalts, tars,
pitches, 1510 ff.
'
bituminous solvents,
examination of,
physical tests, 1586
separation into component
parts, 1586
base, 1589
chart for analysis of, 1590
pigment and filler, 1589
steam distillation, 1586
app. for, 1587 (Fig. 221)
carbenes, 1546
carbon bisulfidl> extract, 1542
evaporation of, 1543
apparatu~ (Fig. 204), 1544

Bituminous sul1stances, evaporation ru

CO 2 atmosphere (Fig. 205)
1545 1
\
characteristiCf\ of table, 1510, 1511.
1512
chemical tests,
anthraquinone reaction, 1562
diazo reaction, 1560
oxygen "in non-mineral matter,
1554
saponifiable and unsaponifiable
matter, 1559
solid paraffins, 1555
sulfonation residue, 1557
water, high and low distillation,
1551
apparatus (Figs. 208, 209),
1551, (Fig. 210), 1553
distillation test, fiasK and retort
method, 1533
apparatus (Fig. 200), 1534,
(Figs. 201, 202), 1535, (Fig.
203), 1537
ductility test of, 1521
" apparatus (Fig. 192), 1522
evaporation test, 1530
air bath for (Fig. 198), 1531
/
apparatus for- (Fig. 199~ 1532
fabrics, examination of,

physical tests of, 1574

resistance to heat, 1577
sampling, 1575 (Fig. 219)
thickness, 1577
weight, 1574
fabrics, separation, 1578
flash-point, 1533
filler and pigment in, 1589 ,
finished product, tests of, 159i
settlement test, 1591 .
demulsiO'ility test, 1591
.
freezing, effects of, 1592
separation into component parts,
1592
,distillation residue, 1593
apparatus for (Figs. 222,
223), 1'593 ~
water and volatileoils, 1594
fixed carbon, 1533
fracture of, 1513
"free carbon " 1548
extractio~ apparatus (Figs. 206,
207), 1549
hardness or consistency test, 1520
by needle penetrometer (Fig.
195), 1520
heat tests, softening or fusion,
cube method,
in air (Fig,,195), 1527
inrwater, 1p25
. ..r1tpparatus for (Fig. 194),
- 1524

'

SUBJECT INDEX
Bitum'lnou& substances, ball and ring
method, 1523
apparatus (Fig. 193),
1524
insoluble residue, examination, 1541
physical characteristics, 1513
physical tests
compression strength,
1565
apparatus (Fig. 213),
1566
displacement
apparatus (Fig. 211),
1563
tensile strength, 1564
apparatus (Fig. 212),
1565
separation bituminous from discrete
aggregate,
centrifugal method, 1570
apparatus (Figs. 217, 218),
1571
'
cold extraction, 1569
apparatus (Figs. 215, 216),
1570
hot extraction method, 1567
ext r act ion apparatus
(Fig. 214), 1568
separation, bituminous mineral and
fibrous matter, 1578
analysis of saturated fabrics, 1578
ash, 1579
'
moisture, 1578
analysis of saturated and coated
fabJ;ics, 1580
ash, 1582
bituminous saturation, 1582
felt, 1581
mineral matter, 1581, 1584
results, 1585
tDp and bottom coatings,
1583
solubility tests
by carbon bisulfide, 1539
examination of residue, 1541
by carbon tetrachloride, 1546
by petroleum naphtha, 1547
solvent. See bituminous. solvents.
specific gravity by hydrometer, 1513
by pycnometer, 1515
by Westphal balance, 1517
streak on porcelain, 1513
table of, 1510-1512
thermal tests, 1523 . See heat tests.
-viscosity,
Engler, 1518
float tists, 1518 (Fig. 190)
volatile matter, 1527
oven for (Fig. 196), 1528
shelf for (Fig. 197), 1529
Black filter paper test, for fluorine, 401

33

Black pigments. See Pigments.

Black powder, analysis of, 1660
Blanc fixe, analysis of, 136
Blank determinations,
alundum analysis, 36
boron determination, 182
Blast-furnace gas, analysis of, 2388
slag analysis. of, 2021
Blasting caps, analysis of, 1686
Bleach liquid,
determination of available chlorine,
~a.O, NaOH, Na.C0 31 279, 280
Bleachmg powder, evaluation of, 278
sampling, available chlorine, chIorates, bases and silica, free lime,
coarse impurities, 2159-2161
Blister copper, see Copper.
Blood, carbon monoxide determination,
by means of 219, 1964
"Bloom" in oils 1776
Blowpipe test fo; bismuth, 150
test for cadmium, 197
test for zinc, 1055
Blowpipe and flame tests, table of, 11071108
Blue lead, sublimed, analysis of, 1874
Blue pigments, analysis of, 1872
Boat and holder, carbon determination,
232
Boiler-room control, 2088
Bomb calorimeter
sulfur in washings, 1642
Borates, boric acid in, 2245
Borax,
bead, boron test, 162. See also Bead
test.
estimation in soap, 2035 '
Boric acid, titration of, 170, 173, 178
Boron,
Chapin's method, 176
distillation apparatus for, 167
detection of,
borax bead ~~st, flame ~est, tumeric
test, 162-163'
estimation, general procedures:
colorimetric (turmeric), 185
reagents for, 185
gravimetric, lime fixation method,
of Gooch and Jones, 166
volumetric determination, 168
estimation, special procedure:
determination in boron carbide, 181
in butter, 165
in canned goods, 166
in crude borates,
water insoluble, 170
water soluble, 169
Pacific C 0 a 8 t B 0 r a x Co.
method, 174-175
in glass, 2167
in milk, 165

34

SUBJECT INDEX

.. ~
in mineral water, 164
irrigation water, 186
natural water, 187
in ores, 171-173
acid extraction method, 171
in plants, 180
in silicates, enamels, etc., 164
industrial application of methods, 163
occurrence, ores, alloys, sundry products, 162
preparation and solution of the sample,
boric acid, boric oxide, boracite, boro
calcite, boronatrocalcite, calcium borate, enamels, silicates, etc., 163-165
separations, distillation method, 165
solubility of boron, boric acid, borax,
163, 164
traces, determination of, 185, 186, 187
detection of, 162, 163
Boron carbide,
analysis of,
methods of the Norton company,
181
boron, 181
other constituents, 184
silicon, 183
Brass and bronze, analysis of, A. S. T.
M., 1351-1372
determination of antimony in, 1369
determination of cadmium in, 1371
determination of copper in, 392
determination of lead in, 392
determination of vanadium in, 1051
determination of zinc in, 1084
Brimstone, analysis of, determining moisture, ash, arsenic, chlorine and available sulfur in, 933
Brines, determination of bromine in, 195
determination of iodine in, 459
British thermal unit (B.t.u.), calculation
of, 1657, 1658
determination. of heat value of coal,
1654
Brom-thymol blue, titrations with, 2191
Bromate method,
,
for determining antimony, 74
for oxidizing manganese to the dioxide, 554
Bromates,
detection of, 188
determination of, by arsenious acid reduction, 193
Bromide paper for arsenic determination,
1197
Bromine, detection of, tests with carbon
disulfide, carbon tetrachloride, barium
chloride, macenta test, silyer nitrate,
188
estimation, general procedures:
Boron,

Bromine, gravimetric, determination as

silver bromide, 191
volumetric, free: 'bromine by potas-,
sium iodide method, 191
chlorine method for soluble bro
mides, 192 \
silver nitrate method, 192
Eosin indicator
silver thiocyanate method of Volhard, 192
micro, 2498
estimation, special procedures:
crude potassium bromide, analysis of,
194
,mixtures of halogens, 276
impurities in commercial bromine,
chlorine in, 194
in brines, 195
apparatus for, 196
in mineral waters, 195
occurrence, 188
preparation and solution of the sample,
organic matter, salts, etc., 189, 190
properties, 189
separations from chlorine and iodine;
190
from heavy metals, cyanides, and
./ silVer, 190
from iodine, 195
traces, determination of,
magenta method, 192
phenol red method, 193
Bromine-ammonia method for separating
manganese from zinc, 558
Bromine method for HON, 662
Bromine number in oil analysi9, 1770
Bromphenol blue, absorption indicator,
582
'
Bronze, analysis of/l351, 1362. See also
Alloys; Brass.
bismuth, 1375
manganese, 1352
Brown pigments. See pigments.
Broths for, bacteriological examination,
nutrient, 2120
sugar, 2120
Brucine, 1957
Buffer action, 2278
materials, pH of, 2284
mixtures, 2281
Bulb, weighing acids, 2203
Bullion, assay of. See Fire Assay.
Buoyancy constants, of water, 1226
correction, in weighing, 1233
Burettes,
calibration of, 1224
apparatus for (Fig. 131), 1225
chamber, iron determination, 472
micro, 2472 -', _
.I
standard fQl-:~1\cidimetry and alkalimetry (Fig. 300), 2199

SUB.TECT INDEX
Burettes,
weighing type (Fig. 304), 2202
Burning oils. See Oils.
.
B. typhi, examination of, 2131
Butter, boric acid determination in, 165
method of analysis. See Oils, Fats,
Waxes
, nButyl alcohol method, for separation
and determination of alkalies, 884
C

Cadmium,
detection of,
blowpipe test, 197
general procedure, 197
spectrum, 198

tube test, apparatus, 198

estimation, general procedure:
gravimetric, determination as cadmium sulfate, 201
as cadmium sulfide, 200
as metallic cadmium, by electrolysis,
202
micro, 2516
volumet~ic, iodine titration of cad. mium sulfide, 203
other methods, 203
estimation, special procedures:
determination in alloys and metals:
\ in brass, A. S. T. M., 1371
in magnesium alloys, 546, 547
in metallic cadmium, 203
in metallic lead, pig, 1417
in ores, 200
in silver solder, 1403
in slab-zinc, 1074
in Wood's metal, 1400
in zinc (spectrochemical), 1089
in zincbase alloy, 1396
industrial application, 197, 199
occurrence, 197
preparation and solution of the sample!
alloys, carbonates, in presence of
lead, metal, ores, sulfides, 199
separation,
from arsenic, antimony, and tin in
presence and in absence of copper,
200
from bismuth and lead, 199, 200
from copper in alloys, 200
from members of the ammonium
sulfide group, alkaline earth and
,the alkalies, 199
from mercury, 200
from silica, 199
from zinc, 200
Cadmium metal, zinc in, 1078
pyridine thiocyanate, 203
sulfate, determination of, 201
sulfide, gravimetric determination of
cadmium as, 200

35

Cadmium sulfide, volumetric determination of cadmium as, 203

Caesium, detection of, 893
estimation of, 898
separation,
from lithium and sodium, 894
from potassium, 894, 896
from rubidium, 897
Caesium bismuth nitrate,
method for sodium, 882
test for sodium, 876
Caesium perchlorate, test for rllen i11111 ,
769
Caffeine, 1959
Calcined alumina, analysis of, 26
Calcium,
detection of,
flame test, 205
general procedure, 205
spectrum, 205
estimation, general procedures:
gravimetric,
other methods, 216
oxalate method,
ignition to the carbonate, 211
ignition to the oxide, 210
volumetric,
iodine method for, 213
permanganate titration of the oxalate, 211
estimation special procedures (see also
Lime)
in barium carbonate, 2172
in glass, 2166
in green paint pigments, 1876
in gypsum, 214
in magnesite, 536, 538
in magnesium alloys, 549
potentiometric method, 549
volumetric, 550, 551
in orange and yellow paint pigments,
1875
in Portland cements as CaO (lime),
1603
in sand, 806
in sea-water, micro, 2547
in water, 217, 2061, 2097
in water as calcium sulfate, 2084
in white paints, 1868
in whiting, etc., 1866
industrial application of methods, 206
ocurrence, ores, minerals, etc., 205
preparation and solution of the sample,
cements, dolomites, limestone,
magnesite, gypsum, plaster of
Paris, silicates, sulfate, suI:fides, pyrites, salts, decompo
sition of refractory ores, 206,
207
rapid iodide method, 213

36

SUBJECT INDEX

Calcium,
separation,
from alkalies, barium, strontium,
magnesium, 121, 122, 123, 209
from aluminum, iron, copper, cobalt,
nickel, manganese, zinc and members of the hydrogen sulfide group,
208
from fluoride, 208
from iron and aluminum, 209
from magnesium and the alkalies,
209
from phosphate, 208
from silica, 208
solubilities, 206
traces, detection of, 205
small amounts in magnesium alloys,
551
Calcium acetate method for fluorine, 413
calcium and equivalent fluorine,
415
fluorine, 416
Calcium carbiue, determination in cyanamid, 663
Calcium carbonate, in fluorspar, 419
fluoride decomposition of, 402
in fluorspar, 420
method for fluorine, 404
Calcium hypochlorite, see Bleaching
Powder; see Chloride of Lime; see
H.T.H.
Calcium oxalate,
determination of calcium as, 219
precipitation in presence of iron,
aluminum, etc., 209
Calcium oxide,
determination in,
aluminum ores, 21
bauxite, 25
cement, 1604, 1614
commercial "arsenic," III
presence of CaC03 , 213
titanox pigment, 1000
Calcium sulfate,
determination iii,
titan:ic pigment, 1863
water, 2084
precipitation from methanol solution, 538
solubility, 121
Calculations,
gas analysis, 2357, 2358
from combustion data, 2411
moist to dry basis, 1345
coal, 1634
rnbber analysis, 2011, 2015
Calibration of volumetric apparatus,
1219-1227
flasks, 1223
Calomel electrode, 2299
Calorie, definition of, 1174

Calorific p'ower of illpminating gas, 239li)

of fuel, 1764
value of inuustrial gases, 2426
of coal and co!re, 1654 ff.
Calorimeter,
\
Junkers' for gas, 1?argent (Fig. 371),
2390; (Fig. 372), 2391; (Fig.
3'(3), 2392
for liquid fuels, 1862
Calorimetry, micro, 2546
semi micro, 2547
C. and A. Copper Co. methods, 372
Candle-power, definition and method of
determination, 1174
of illuminating gas, 2391 .
Canned goods, arsenic determination in,
107
tin determination in, 971
Caoutchouc; 1966
in lubricating oil, 1736
Capacity, unit table of, 1169
Capsule, moisture determination (Fig.
231), 163.6
Car benes, 1546
Carbohydrates,
micro
drtermination,
2547
Carbolic acid (phenol),
detection of, 1949
determination of, 2253
Carbon,
activated, 2154
detection of, 218
carbon dioxide in gas, 218
carbonates, 218
_
carbonic acid in water, 218
carbon monoxide in blood, 219
estimation, general, procedures:
gravimetrill, drYi combustion method,
weighing CO"
macro ni"ethod, 220
micro m~thod, 2475
semi micro, 225
wet oxidatioll process, weiglIing
CO,,226
volumetric,
barium hydroxide method, 243
measurement of vqIume of CO"
formed, 244
titration 'of carbonate formed, 243
estimation, special procedures:
in bitumens "f_ree," 1548
in coal, 1649 if.
in coal as "fixed ~arbon," 1640
in ferro chromium, 1493
in ferromanganese, 1488
in ferro molybdenum, 1596
in ferrotungst~n,.J.500 '
in ferrovan.a4iUm~ 1050, 1495
in graphitic carbon, 232
boat for, 232 495

SUBJECT INDEX
Carbon,
ill iron and steel, 2il9, 488, 495,
1426, 1428
colorimetric method, 232
combined, 231
in minerals! 224
ill mOllel metal, 1380
in nichrome, 1408
in organic substances, 222
nitrogenous, 224
tube for, 223
in pigments, 1874, 187 7
in rubber goods, 2007
in silicon carbide, 815
.free and total, 816
in soils, 224
in stellite, 329
occurrence, 218
In:eparation of the samp1c, .alloys, iron
and steel, 219
general, 219
organic matter, cafbonates, etc.,
219
residue test of lubricat~ng oils, Con, rtlllson method, 1737
.
separation from iron and steel, CUpl'lC
potassium chloride method, 220
Carbonate method,
for barium, 126
"for calcium, 216
for strontium, 902
("nrhonate, removal from alkali, 2197,
2198
Carbonate,
titration of, 2253
in mixture ,vith hydroJ{ides, 2266
Carbonates,
cletermina tion of,
in ccmE',llt materia1s, 1616
in presence of other combined acids,
929
in soap, 2033
in sodium sulfate, 645
in ores. See Prep. of Samples under
various elements.
precipitation of palladiulll by, 722
Carhon hlack, in rubber, 2018
Carbon flioxide combined as carbonatc:
egtimation, general procedure for
determination, Z35
apparatus for, 236
micro, 2482
gravimetric determination in CItTbonates,
.
hydrometric method, 242
Joss of weight method, 237, 239
apparatus for, 238, 239
volumetric, measuri!lg gas evolved,
240
apparatus for, 24 0
estimation, special procedures:

37

Carbon dioxide combined as carbonate:

in alloys. See Iron and Steel; see
ferroalloys.
in ammoniacal gas liquors, 638
III baking powder,
available CO2 , 237
residual CO" 237
'in 'barIum ores, l'3'5
in blanc fixe, 136, 1867
in composite white paint, 1867
in gypsum, 214, 1866
in pigments, 1866
in soap, 2034
in water, 2070
in white lead, 1860
free in gaseous mixtures,
a bsorption of, 2347
determination of, 2350, 2352,
in air, 2403
in illuminating gas, 2355, 2397
indicators for, 2381
properties of, 2365
generator for (microanalysis) (Fig.
424), 2489
Carbon disulfide,
detection of bromine with, 188
determination of iodine with, 448
extract of bituminous substances,
1539
purification of (Fig. 16), 155
monoxide,
absorption of, 2348
and heat loss, 2387
in air, 1963, 2405
in blood, detection of, 1964
in gaseous mixtures, determination
,()f_~ \1350, \135\1

in illuminating gas, determination

of, 2356
in mine gases, 2402.
properties of, 2348
oxysulfide, 2348, 2379
tetrachloride, bromine test' with, 188
extract of carbenes, 1547
tubes, for colorimetric determination
of carbon in steel, 232
Carbonic acid,
in water, 2075
free, 2252
as bicarbona~e, 2252
Carbon steel, metallography of, 2558 ff.
Carbonyl,
group, micro determination, 2546
method, separation' of cobalt and
nickel, 311
Carborundum. See Silicon Carbide.
Carboxyl group, micro determination of,
2547
Cadus method for halogens in organic
matter, 265

38

SUBJECT INDEX

Carnotite,
uranium in, 1024, 1026
vanadium in, 1042
Case hardening, 2567
Casein, analysis of, 1931
Caustic,
liquors, hydrometer test~, 2257
standard solution, 2197
Caustic soo:a,
impurities in, 2259
-sodium tartrate method for distinguishing copper compounds, 395
Causticized ash, analysis of, 2263
Cellulose,
.
alpha,
in paper, 1911
in pnlp, 1881'
isolation, chlorination J1lethod, '1878
in rubber, 1987, 2009
Cement, specifications anJ chemical analysis, A. S. T. M., 1597
composition of, table, 1619
lime, free in, 1618
Portland cement,
analysis, A. S. T. :M., 1599
~alcium oxide, 1603
ferric oxide, 160~
and alumina, 1602
insoluble residue, 1601
loss on ignition, 1600
magnesia, 1604
silica, 1601
sulfuric anhydride, 1601
rapid methods,
calcium oxide, 1614
ferric oxide and alumina,
1614
magnesia, 1615
silica, 1614
Portland cement
tentative methods
alkalies, 1612
magnesia
acid -alkali nlethod, 1601)
oxyquinolate method, 1607
manganese, 1611
phosphorus, 1609
raw materials
carbonates in, 16Hi
composition of, table, 1619
limestone, lime rock, raw mix,
1617
sampling, standa'rd methods, 1599
silica in, 816
specifications, 1597
Centipoises (viscosity), 1725
Centrifuge, for extractions (Figs. 217,
218), 1571
Ceric sulfate, 246
applications in volmnetric analysis,
1205

Cerie sulfate, methoq ffJ? determining:

antimony, in presence of As 78
arsenic, 100

'
iodide, 454 I
iron; 474 . \
mercurous lon, 582
molybdenum; 596
nitrite, 655
uranIum, 1027
preparation of standard solution of,
1204
standardization, 1204
Cerium
det.ec~i~n, spectroscopic, 252
estImatIOn, general methods
gravimetric methods, 252
volumetric methods
persulfate method, 253
sodium bis111utlia te method 254
estimation, special cases
' ,
in presence of rare earths 253
industriaL application of methods 246
preparat,ion of the sample, 247 '
fus,ion method, 247
separations
fr9111 rare earth metals, 252
from scandium, 249
from thorium, 249
rare earths from other elements,
247-250
Cerium group, approximate estimation,
250
Chancel degree, for specific volume of
sulfur, 934
sulfurimeter for (Fig. 116) 935
Chapin's boric acid method 176'
Characteristics of oils fat~ waxes. Sec
Oils, Fats and Wa;'es, 1701 if.
Chatcoal, in explosives, 1661
Chemical)aws, 1191-1193
Chemical mieroscopy. See Mieroscopy
Chimney ;fiue :gases, see Gases
China clay
_
as glass raw material, 2170
as paint pigment, 1866
Chinese blue." See Pigments,
9hinese wood oil. See Tung oil under
Paint Vehi(lles, 1847
Chloral hydrate, detection of, 1951
Chlorate
determination ~f, 274
in bleaching powder, 2160
in explosives., 1677
ill presence of perchlorates, 275
and HCl, 276
reduction to chloride ion; 274
removal of,. in' '!fUlfur estimations, 907
test for, 262
-'
.
Chloric acid, 274

SUBJECT INbEX
Chloride
and cyanide, determination of, 277
chlorate, perchlorate, examination of,
276
cyanide and thiocyanide, .determihation of, 277
detection of, 261
in bleaching powder, determination of,
2160
in gases, 2417
in potash, 2169
in soap, 2030
in water, 2075
method, for silver, 821
test for silver, 818
Chloride of lime, 278 (see also bleaching
powder i. bleach liquor)
Chlorine '
detection
combined
as chlorate, chlorite, hypochlorite,
perchlorate 262
as chloride, 261
as hydrochloric acid, 261
free, 261
in presence of bromide and iodide,
262
cyanate, cyanide, thiocyanate,
261
estimation
chlorine, combined as chloride
gravimetric, as silver chloride, 269
volnmetric
adsorption indicator method,
273
silver chromate method (Mohr),
272,
silver thiocyanate method (Volhard), 271

micro techniqne, 2498

estimation special cases
in bleaching powder, 278,
2160
in obromine, 194
in chlorine water, 273
in cement-copper and copper ores, .391
in chlorosnlfonic acid,
2244
in gases, 268
in liquid bleach, 279
in organic compounds
Carius method, 265
lime method, 266
sodium-alcohot method,
266
sodium-liqnid ammonia
method,266
sodinm peroxide
method,267

Chlorine, estimatioll
in presence of free bromine and iodine, 276
in presence of other acids,
2i6
in rnbber, 1986
in soap, 2030
in sulfur, 933
in water, 280, 2054
in zinc chloride, 1080
chlorine, free, 264 (see also hypochlorites, chI orates, etc.)
in gases, 268, 2347
electrolytic, 2402
in hydrochloric acid, 2204
in nitric acid, 2214
properties of, 2362
industrial application of methods, 263
occurrep.ce, 261
preparation and solntion of the sample
ores, cinders, rocks, water soluble and insoluble chlorilles,
silver chloride, 263, 264
organic substances, 265, 266
separation
from bromine, 268
cyanides, thiocyanates, bromides
and iodides, 268
iodine, 268
of halides from
heavy metals, 268
one another, 268
silver and silver cyanide, 268
Chlorine consumption, by pulp, 1880
Chloroform, determination of, 1948
elO.traction of rubber, 1982, 2002
Chloronitrotoluene sulfonate, test for potassium, 862
Chloroplatinate method,
for ammonium, 636
for potassium, detection, 862
determination, 869
ChI oro sulfonic acid, analysis of, 2244
Cholesterol, 1775
Chromate
method for barium, 125
volumetric) 128
for chlorine, 272
for lead, 507
reduction of, 286
test for barium, 117
Chromate-ferrous sulfate method, for
lead, 513
iodide method, for lead, 512
Chrome alum, light filter, 862
Chrome green, see Pigments
Chrome-iron ores, fusion, 284
Chrome ores, analysis of, 296
Chromic acid
determination in presence of vanadie,
293

40

SUBJECT INDEX

Chromic acid
.
removal of, from nitric acid 635
Chromic hydroxide, preeipitatiori of, 286
oxide, properties, 286
Chromite, chromium in, 295
Chromium
detection
tests with barium aeetate, ether, lead
..3l'JlwfJ). _ml.lT.!mrflJzy ZE}tZO.'}t&, llJ-i}2-ogen peroxIde, reducing agents
diphenyl carbazide 282 283
'
distinction, chromic and' chro~ous salts
~3
. ,
chrominm and vanadium salts 1031
estimation, general prllcedures '
colorimetric
'
as the chromate 289
with diphenyl c~rbazide 290
gravimetric
'
as barium chl'omate, 287
as mercurous chromate 287
'
as' oxi~e, Cr2 (Js, 286
volumetrIc
ferrous sulf;tte method diphenylamille indicator,' 291
ferrous sUUate-permanganate method 288
potentiolQetric 2315
iodide methoil 288'
persulfate m(lthod 301
estimation, special prOeedu;es
in aluminum alloys 48
in alundum, 35
'
..
in chrome-iron alloys 301
in chrome-nickel alloys 301
in chromite, 29(;
,
~n ~errochromiuln, 1490
~n I~0!l ores ~~Q alloys, 291
In lllcnrome, .l4\]6
in ores, 296, 321}
in presence of vanadium 293
in steel and iro~ 293 495 1454
1473, 1477, Li79 ,
,
,
in stellite, 328, 14H
traces, 290
industrial application of meth~ds, 283
metal, 302
occurrence, 282
oxidation of, 296
preparation and solutil)n of the -ample
283
~,
refractory materials materials
high in silica, ores' 283 284
separations, 285
"
solubility of the metal 283
traces, detection; 282, 283. See colorimetric estimation.
r'hromium plating baths, sulfate in, 917
('inehonine method, for tungsten 1011
Cinchonine-potassium i()dide method for
bismuth, 158
'

Cinchonine..reagent, for tungsten, 1011,

1095
.
j
Cinders
arsenic in, 106
I
chlorine in, 264
Citrates, test for palladium, 722
Citric acid, analysis of, 2253
reagent for calcium, 119fi
{)).'}J" .,?));-a)}es ).'2,

2]f}3

analysis of, 6, 2162 f.

silica, state of, 2164
titanium, 2163
Cleveland open cup (Fig. 247), 1733
(Fig. 248), 1734
Clinker, silica in, 816
Cloud and pour points, of oils, 1727
Coal and coke, A. S. T. M. methods,
1620 f.
ash, determination of, 1637
fusibility of, 1645
apparatus
(Fig. 233)
]645
- (Fig. 234), 1646
fused cones (Fig. 236), 1649
mold (Fig. 235), 1647
ph.osphorus in, 1644
calculation, . air dried to "as received, " 1634
calorific value, determination, 1654
results, 1657
fixed carb~n, 1640
fusibility, 1637
hydrogen in, 1649
moisture
in coal, 1636 (Figs. 230, 231)
in coke, '1628, 1635, 1636
nitrogen iIi, 1652
oxygen in, 1653.
preparation of the laboratory sample,
II6ZfJ r

ball-mill method, 163;1'

drier for, 1629 (Fig. 227)
riffle sampler, 1631 (Figs. 228l....
229)
/
sampling
coal, standard fuethQ..ds, 1620, 1622
(Fig. 225)
---.
coke, 1624
sampling points,. 1626 (Fig.
226)
sulfur, bomb washing method, 1643
Esehka's method, 906, 1640
sodium pe 70xide method, 1~43
volatile comljustible matter, 1638
apparatus, 1639 (Fig. 232)
Cobalt
detection, bead test, 306
general procedure, 305, 306
thiocyanate jes_i, 306
estimation, general procedures
gravimetiic
.
as metal, by electrolysis, 315

SUBJECT INDEX
Cobalt
nitroso-beianaphthol method, 314
_potassium nitrite method, 312
pyrophosphate method, 330
otller methods, 320
volumetric methods
perborate method, 318
potassium cyanide method, 319
potassium ferricyanide method,
628
estimation, special procedures
in "arsenic," 110
in cobalt oxide, 320, 330
in copper, 385
in ferro-cobalt, 322
in lead (pig), 1419
in metallic cobalt, 321
nickel, 322, 628
in ores, 315
separation from iron, 317
in steel, 323, 325, 1466
in stellite, 325, 327, 1409
industrial importance of methods, 307
minerals of, 305
QCC"\l-n:~;\:\.c~, ~(\()

preparation and solution of the sample

general procedure, 307
metallic cobalt, nickel, cobalt
alloys, cobalt oxides, ores, 308
separations
. ammonium sulfide group from alka~
line earths, 309
hydrogen sulfide group, 309
cobalt and nickel from chromium,
311
iron, 312
manganese, 309, 310
zinc, 311
cobalt froIn nickel
alkali xanthate method, 310
alpha benzildioxime m'ethod, 311
carbonyl method, 311
dimethyl glyoxjme method, 310
nitro so betanaphthol method,
310
phenylthiohydantoic acid method,
310
potassium nitrite method, 310
Cobalt nitrate, test for aluminum, 2
for zinc, 1055
Cobalti nitrite test, for potassium, 862
Cobalt uranyl acetate test, for sodium,
875
Co~aine, 1958
Codeine, 1958
Coin ingots, 829
Coke_ See Coal.
Cold test. See Oils.
Colemanite, 164
Colophony, in rubber, 1981
ColoI' comparator, 234 (Fig. 33)

41

Color standafds, for copper, 380

Color test, in turpentine, 1850
of waterl 2048
Colorimeter, 292 (Fig. 38), 989 (Fig.
.124). see also Colorimetry, Hydrogen-ion Concentration, pH.
Campbell-FJurley, 519 (Fig. 59)
Duboscq 988 (Fig. 123)
photoele~tric, 2472
Colorimeter plock, 992 (Fig. 992). See
also pH Measurement..
,Colorimetric Determinations. See also
:l\EcrOllnalysis.
aluminuJ1l, 55
in anirnal tissues and fooiJ.s, 5, 5658
ammonia, 630
in water, 2077
bismuth, 158, 159
carbon, in. steel, 232, 234
comrarator for, 233
racks, 233
tubeS, 234
cIllo due, 280
chromiuJ1l, as chromate, 289
diphenylcarbazide method, 290,
292
copper, :)76, 381, 2519, 2520
fluorine, Steiger-Merwin method, 410
gold, 438, 439
hydroge~-ion concentration, 2281
iron, 486, 487
lead, as sulfide, 517
dithizone method, 521
magnesiOm, 529
manganese, 572, 573
molybdeIlUm, 606, 609, 610
nitrate, 640
pH, 228:1
rhenium, 772
silicon, g03
sulfur, 9 36
titanium, 987
thp~ol method, 994
Columbates, ))46
Colum)Jite, 331
Columbium and Tantalum
analytical teehnique
destruction of organic matter,
335
filtering and washing, 335
fusion, 334
ignition, 336
leaclling, 335
"quarter gram" analysis, 336
detection, 332, 333
Giles' test, 333
in mil(ed oxides, 334
Marig~ac 's test, 332
Powell and Schoeller '8 test, 333

42

SUBJECT IN,DEX

Columbium and Tantalum

Schoeller's test, 332
small amounts, 333
estimation, 347
isolation of earth acids
alkali fusion, method, 345
hydrofluoric acid method, 344
tartaric acid method, 344
occurrence, 331
preparation of the solution, 337
minerals, 338
separations
columbium and tantalum,. 346
columbium, tantalum, titanium and
zirconium from tungsten, 339
earth acids from iron, 340
silica, 338
tho ria and rare earths, 343
tin, 338
titanium, 340
tungsten, 338
'small amounts, 339
zirconium and hafnium,' 342
solubilities, 337
tannin, as reagent, 336
volumetric methods, 347
Combinations, hypothetical in water
analysis, 2087
Combined carbon, in steel, 231
colorimetric determination, 232
Combined sulfuric acid, in aluminum
salts, 15
Combustion,
air requirement, 2384, 2386
furnaces. See Carbon; Fire Assay;
Steel, etc.
heat of, Table, 1185, 1186
method, for carbon, 221
train, 222, 229, 489, 1428
Comparators, for pH. measurement, 2290,
2291, 2292, 2293, 2294, 2295
Compressive strength, of bitumens, 1565
Conductance,
equivalent, 2323
specific, 2323
Conductivity method,
silicon in steel, 811
apparatus for, 811, 812 (Figs.
97, 98)
chart, resist. vs. % Si, 812
(Fig. 99)
Conductometric methods, 2323 .
apparatus, 2326 (Fig. 339); 2327
(Figs. 340, 341)
mobilities of ions, table, 2324
principles of, 2323
titrations, 2328
neutralizations,
intermediate acids and bases,
2329
mixtures of acids or bases, 233:]

Conductometric me,thods, strong acid9 and

bases, '2328
replaceme)1t methods, 2332
weak acids and bases, 2330, 2331
precipitatio~s and complex-formations, 2333
summary of practical cases,
2335
Constant-boiling acids. See Hydrochloric acid, etc.
Constant temperature bath, 365
Constants. See Tables; see Gases; Oils;
Acids, etc., etc.
Conversion factors,
Baume to specific gravity, 1168
temperatures, Centigrade to Fahenheit, 1165
volume, weight and energy, 11761178
weights and measures, metric and
ot,her systems, 1169-1170
.
Converter' efficiency, ,ammonia to oxides
of nitrogen, 659
apparatus, 659 (Fig. 75)
Copper,
"
detection,
general' procedure, hydrogen sulfide
test, flame test, reduction test,
ferocyanide test, rhodanine test,
dimethyl glyoxime test, 349, 350
estimation, general procedures,
colorimetric,
ammonia method, 378
eolor standards, 380
Biazzo method, 2519
.
diethyldithiocarba'mate method,
376, 2520 I
/
ferrocyanide method, 377
hydrogen srtlfide method, 380
potassium ethyl xanthate met!lOd,
376
'
standard solution fot, 1199
gravimetric,
alphabenzoinoxime method, 2521
cuprous thiocyanate.m.!lthod, 357.
electrolytic methods, 359"11'.
electrolysis,
__
apparatus, 359, 361, 362,
366
'effect of impurities, 367
precautions, 366
reqioval of deposit, ,368
introduction,' 359-361
rapid methods, 362, 363
slow methods, 364,
large portion method, 365
small portipn method, 364
hydrogen ,reduction method, 390
fl,lrRill!e .for, 391 (Fig. 46)
volumetric, Demorest '8 method, 375

SUBJECT INDEX
Copper,
Garrigue's method, 376
permanganate method, 375
potasdum cyanide method, 373
potassium iodate method, 372, 374
potassium iodide method, 368
shorn. modification, 371
. Park's procedure, 372
Vol hard 's method, 376
estimation, special procedures,
in aluminum and its alloys,
electrolytic method, 41
iodide method,- 40
in "arsenic," 110
in Babbitt metal, 398, 1382
in be~yllium-copper alloys, 146
in bismuth bronze, 1376
in blister copper, 364, 390
in blue vitriol, 392
in brass, 392, 398
. in bronze beating metals,.A. S. o:r:. M.,
1362
in ferromolybdenum, A. S. T. M.,
1509
in ferro tungsten, A.' S. T. M., 1015,
1504
in lead (pig), 1417
'" in lead-base alloys, 1382, 1384, 1385
in magnesium alloys, 544, 545
in manganese bronze, A. S. T. M.,
1352
.
in mill tailings, 379
in molybdenite, 601
in monel metal, 1378
in ores, 369
in refined copper, 390
in reverberatory slag, 379
in rubber, 1971
in selenium, commercial, 7.88
in silver solder, 1402
in slag, 2024, blast furnace, 379
in slime tailings, 379
in solder, 1393
in steel, 1446, 1448
in sulfuric acid, 2231
in tin-base alloys, 1382, 1384, 1385
in tungsten, ,metallic, 1015
in water, 2080
in Wood's metal, 1400
in zinc (slab), 1077
in zinc-base alloy, 1394
in zinc chloride, 1082
im puri ties in blist~r and refined copper~
determina tion of:
.
antimony, 387
arsenic, 91, 112, 387
bismuth, 382, 384
cobalt, 385
iron, 383
lead, 386
nickel, 385

43

Copper, impurities in blister and refined

copper,
oxygen, 389
phosphorus, 390
.
selenium, 388, 789, 791
sulfur, 389
tellurium, 388, 789, 791
zinc, 385
separations,
copper, 382, 384, 386
hydrogen sulfide group, 385
iron, 385
iron and bismuth, 382
lead, 385
zinc, nickel and cobalt, 384
industrial application of methods, 351
occurrence, 349
preparation and solution of the sample,
,
353
alloys, cast iron, steel, matte
slag, iron ores, briquettes, etc.,
352-354
separations,
as metal, 355, 358
as thiocyanate, 355, 357
from bismuth, 358
from cadmium, 356
from hydrogen sulfide group metals,
356
,
from members of subsequent groups,
355
.
from selenium and tellurium, 356
state of combination in ores, 393, 395,
396
solubility, 352
Copper bullion, assay of; 853
Copper-copper oxide mixtures, 394, 395
Copper method, oxygen in gas, 670
Copper number:, of paper, 1913
Copper slimes, platinum and palladium
in, 726
selenium and tellurium in, 789
Copper sulfate, method -for cyanide, 660
treatment of water with, 2111, 2112
Copper-zinc alloys, metallography of,.
2556, 2557
Corn oil constants. See Oils ..
Corning glass works, method for Sele
nium, 791
Corrosion, by acid waters, calculation of,
2087
by gasoline, 1742
Cottonseed oil, 1777, 1778
Coulomb, definition of, 1174
Counting methods, microscopical, 2449
Crank case dilution, 1820
Crushers and grinders, 1309, 1313
Cystalline state, 2435
optical properties of, 2436
refractive index, 2436

44

SUBJEOT INDEX

Culture media. See Water, Bacteriologi

cal examination.
Cupel, in fire assay, 837
Cup ferron (betanitrosophenylhydroxyla
mine)
elements precipitated by, 468
method for iron, 6
for tin, 961
reagent, 469
and its applications, 1223
separation of Fe, Ti, V, Zr from Al,
Cd, Co, Cr, Mn, Ni, Zn, 466
of U and V, 1021
of U in carnotite, 1026
of vanadium from other elements,
1035
of zirconium, 1099
Cuprammonium reagent, for paper pulp,
. 1886
I
CUll ric-potassium chloride reagent, 2~Q
Cupric oxide. See Copper-copper oxide
mixtures.
Cuprous chloride, ammoniacal and acid
for gas analysis, 1195
for reduction of arsenic, 94
Cuprous thiocyanate method, for copper,
357
test for paladium, 722
Cupro-vanadium, 1051
Curcumin, method, for Be, 144
Cyanamide, analysis for N, CaC" and oil,
663
.
Cyanide and chloride, 277
Cyanide and thiocyanate, 661
Cyanide, analysis of commercial, 665, 666
Cyanide, copper method, 660
Liebig's method, 661
method for copper, 373
for nickel, 622
for silver, 831
precipitates, mercury in, 584
process, for tin ores, 955
silver determination, 823
solution,
mercury in, 583
oxygen in, 676
-thiocyanate separatio~, 277
Cyanogen, -absorption of, 2347
detection, 2363
Cylinder oil, constants. See Oils; Tables.
D
Decomposition of the sample, chapter on,
1334
reagents for, 1335, 1337
Demulsibility test, bituminous substs.,
1591
oils, 1740
Density_ See also Specific Gravity.
of glass, 2165
of micro samples, 2546

Dental -gold alloys, 754

Detonators, ()lectricl analysis of, 1688
Deuterium, 444
Devarda's nitrate apparatus, 643, Fig. 68
nitrate method, 6~0

reagents, 641
weighing bottle,! 643
Developer, photographic, 1091
Diazzo reaction, bituminous _substnnePR,
1560
Dichlol'iluoresceill indicator, 273
Dichromate method for iron, 471
titration of, chromium determination,
288, 291
Dicyandiamidine test, for cobalt, 306
Die-casting alloy, zinc, 1394
Diethyldithiocarbamate, test for copper,
376
.
Differential method, potentiometric titration, 2318
Dilution, crankcase oils, 1820
Dilution of ,mixtures, formulae for, 2236
Dimercaptothiodiazol, 1214
p-Dimethy lamino benzylidene-rhodanin
reagent, 1214
test for copper, 350
test for silver, 819
Dimetliylglyoxime
method for Ni, 619
method for Pd, 724
reagen t, 1214
separation, Ni and Co, 310
tesf for Cu, 350, 351
test for Ni, 614
Diphenylamine
indicator, iil chromium determination,
291
in iron determinittions, 473
with permailganate, 478
in zinc determinations, 1065
solution or, 1205
in explosives; 1696
test for nitrates, 631
Diphenylamine' sulfonate, indicator, 397,
1206
preparation, 1214
,.
Diphenylbellzidine, indicator, ZI1 titratration, 543, 1206
Diphenylcarba:tide, determination of Cr, I
290
preparation of, 293
test for Cr, 290 '
Diphenylendoanilohydrotriazole (nitrOll)
method for nitrates, 635
Diphenylthiocarbazone. See Dithizone
Dip!crylamine, 1215
Displacement test, bitum. s\lbsts., 1563
... ' _
.
DistilIa tion
method" arsen~<69; 91, 92, 93
boron, 167, 176ethyl alcohol, 2137

SUBJEC'r INDEX
DistiVa tion
germanium, 430
mercury, 578
osmium, 750
rhenium, 770
selenium, 779
tellurium, 779
water, 1340. See Water
micro, 2546
residue, bitumin. substs., 1593
silicon tetrafluoride and hydrofluo
silicic acid, 408, 423
steam, bitumin. substs., 1586
test, bitumin. substs., 1533
oils, A. S. T. M., 1705-1713
turpelltihe, 1851
Distribution coefficients, 2586
DithizQue, method for lead, 521
reagents, 522
te~t for silver, 820
Dividintr pipette, 365 (Pig. 43)
Dotltor test, 1742
Dolomites, calcium in, 210
solution of, 207
Dow Chemical Co. standard methods of,
. 539-553, 2181-2187
Drier, paint oils, 1845
see Ovens.
Drying oils, list Of, 1849
test, of oils, 1784, 1846
'Duboscq colorimeter, 292
Ductility, bituminous substances, 1521
mold for, 1522 (Fig. 192)
See Metallography
Du Pont de Nemours, E. 1. and Co.
Methods of, 649, 653
nitrometer, 651 (Pig. 73)
1)nst, in gases, 2389
Dynamite, analysis oi, 1662
ammonia, 1667
.
qualitative tests, 1662
J1yne, definition, 1175
Dysprosium, 245

E
"Earth Acid~" (oxides of" Cb and Ta)
332 detection in mixed oxides, 334
l;bonite in rubber, 1991
]~bullioscopy, 2532
]<~dible fats, see Oils
Ekasilicon. See Germanium
Bliadin test, for oils, 1701, 1765
Electrical conductivity, see Conductivity;
// Conductometric Methods
Electricity, units, 1174
Electroanalysis
apparatus, 359
beaker, 390 (Fig. 45)
Braun cabinet, 390

45

Electroanalysis
hydrometer jar, 366 (Fig. 44)
solenoid, 362
methods, :rapid and slow, 360
micro, 2510
Electrode,
air, 2310
antimony, 2302, 2310
bimetallic, 2320
calomel, 2299, 2303
E. M. P. values, 2308
glass,. 2303
. graphite, spectrographic, 1089, 1091
hydrogen, 2300 (Fig. 324); 2311
limitations, 2300
titrations, 2309
metal-metal oxide, 2311
oxidation-reduction, 2306
apparatus, 2314 (Fig. 331)
oxygen, 2310
polarized, 2320
potentials, 2305
quinhydrone, 2301, 2310
limitations, 2302
reference, 2307 (Pig. 325)
.silver-silver chloride, 2300
systems, for pH measurement, 2298
. working formulae, 2303
"Electrolyte, IJ copper in, after electrolysis, 366
'
Electrolytic conductivity method, solubility, 2586
.
Electrolytic determination (electrodeposition)
antimony, 73
bismuth, 156
cadmium, 202, micro method, 2516
cobalt, 315
separation from manganese, 310
copper, 359 ff. micro method, 2517
gold, 435, micro method, 2518
lead, 507, micro method, 2515
mercury, 579, micro method, 2516,
2518
nickel, 620
micro method, 2513
separation from manganese, 310
platinum, 719
silver, 824, micro method, 2518
tin, micro method" 2513
zinc, 1059, micro method, 2513, 2515
J!]lectrolytic separations
applied to aluminum alloys, 47
Melaven cell, 47
applied to alloys, aluminum in, 1389
applied in steel analysis, 1482
Electrometallurgical Company, standard
methods
.
for chromium, 293, 296, 301,
302
for ferrosilicon, 808-810

46

SUBJECT INDEX

Electrometallurgical Company, for manganese, 567, 568

for molybdenum, O1'es and COIlI]entrates, 597-606
steel, 610
for phosphorus, presence of
vanadium, 706-707
Electrometric determinations (see also
Potentiometric
determinations;
Potentiometric titrations.)
of pH or Hydrogen-ion concentration
cell for, 2311
electrode systems, 2298
antimony, 2302
glass, 2303
hydrogen, 2300
quinhydrone, 2301
of paper, 1916
apparatus, 1916 (Fig. 272)
potentiometri() principle, 2297
(Fig. 321)
reference
electrodes, 2299,
2300 (Figs. 322, 323)
Electromotive force, definition, 1174
series, table, 1141
Eliadin test, for oils, 1765
Emulsions, bituminous, 1591
Enlargement methods, 2546
Eosin, adsorption indicator, 192, 457
solution, 1201
Erg, de:finition, 1175
Erbium, 245
Erioglaucine, indicator, 474, 1206
Eriogrecn, indicato1', 1206
Errors, see under individual determinations, for sources.
Esters, in alcohol, 213l)
Etching test, for fluor~ne, 399 (Fig. 47),
400
quantitative, 421, 422
apparatus, 422 (Fig. 53)
Ether, acid in, 2270'
chromium test, 282
method, extraction of iron, etc., 312,
465
isopropyl ether in, 465
Ethine, see Acetylene; Gases
Eth6xy, micro determination of, 2527
Ethyl acetate, test for alcohol, 1948, 2137
alcohol, analysis, 2137 if.
detection, 1948, 1949, 2137
pnrification, 1771
speci:fic gravity tables, 2142-2146
Ethylene, properties ,of, 2348, 2375
Europium, 245
iodometric method for, 255
Eutectic mixture, 2554, 2555
Euxenite, 246

Eva pora ti,oll

apparatus for, 910! (Fig. 109)
rapid, 2233 (Fig. 309)
test, bituminous s~bstances, 1530
Evolution method, splfur, 1442
apparatus, 1443 I(Fig. 1,85)
Explosion method, ,gas analysis, 687
(Fig. 85)
.
Explosives, analysis of, 1660 ff.
ammonia dynamite, 1667
ammonium picrate, 1683
black powder, 1660
asb, charcoal, moisture, nitrates,
sulfur, 1661,
sampling, 166Q
blasting caps, 1686
detona tors, 1686
dynamite, 1662
gelatin dynamite, 1668
lead azide, 1685
low-freezing, 1669
mercury :t;ulminate, 1685, 1687
nitrocellulose, 1678, 1690
guncotton, 1690
nitrogen in, 1690
preparation of sample, 1690
pyrocellulose in, 1691
stability test, 1692
test paper for, 1694
nitrocellulose-smokeless powders, 1695
nitrochlorhydrin, 1672
nitroglycerin dynamite, 1662
nitroglycerin smokeless powders, 1698
nitroglycol, 1673
nitropolyglyeerin, 1672
nitrostarch, 1678, 167Q
nitro sugars, 1672
"permissible" e~plosives, 1673
list of ingre,ilients, 1674
qualita,tive examination, 1674
picric acid, 1682
soluble lead in, 1683
primers, 1688
composition, 1688, 1689
qualitative analysis...Qf, 1689
sampling, 1689
smokeless powder, ]695.
nitrocellulose, 1696
nitroglycerin, 1698
" straight' t dynamite, 1662
tetryl, 1684
trinitrotoluene" 1681
Explosives, ingreflients in
acidity, test, 1182
ash, 1661, 1667, 1678, 1681, 1682, 1697
centralite, 1699
cereal products, 1680
charcoal, 1661,_1680
i
chlorates, 161)5,-1677
chlorides, T6 75 - .
gum arabic, 1675, 1677

SUB.TEeT INDEX
E;xplosives, gums,. 1680
insoluble matter, 1667, 1678, 1680,
1682
nitrates, 1661, 1666, 1677, 1680
nitrocellulose, 1669
nitrocompounds, 1687
separation from nitroglycerin, 1671
nitrogen, 1671, 1682, 1697
nitroglycerin, 1665
oils, 1665
perchlorates, 1675, 1677
resins, 1665
sMids, 1675
. specific gravities of, 1676
starch, 1667
sugar, 1674, 1678
sulfur, 1661, 1665
sulfuric acid, 1682
Explosives, tests and sundries
extraction with acetone, 1678, 1692,
1697
with acid, 1666, 1678
with ether, 1663, 1664, 1676, 1690,
1697
evaporation of extract, 1664
with petroleum ether, 1677
with water, 1666, 1677, 1680
moisture in, 1661, 1663, 1670, 1676,
1679, 1681, 1682, 1687, 1696
,apparatus, 1670 (Fig. 237)
potassium iodide-starch test, 1692
sampling of, 1660, 1662, 1668, 1686
solidification point, 1681, 1682
sta,bility tests, 1692
at 134.5 C., 1693
surveillance test, 1697
typical compositions, 1700
viscosity, 1694
Extraction, see Ferric Chloride; Rubber;
Bituminous Substances; Explosives.
micro, 2546
F

"

Fabrics, see Bituminized fabrics

Factors, see Tables; conversion factors,
etc.
Fats, 1785
in grease, 1828
Ji'atty acids, table of, 1794
in soap, 2029
Fatty oils, in lubricants, 1736
Feedstuffs, iodine in, 460
Fehling's solution, 1206
Feldspar, 2170
J Fergusonite, 246
Ferric alum, indicator, 1201
Ferric chloride,
extraction with ether, 465
reagent for tin, 967
test for phenol, 1950
Ferric iron, see Iron

47

Ferric oxide, determination of iron as,

467
in cement, 1602, 1614
in sand, 805
in silicon carbide, 815
see also Iron; estimation, special
cases; Iron and Alumina.
salts, distinction from ferrous, 463
oxidation of iodides by, 454
titanium by, 985
Ferricyanic acid, 663
l'erricyanide, test for iron, 463
Ferro-alloys, analysis or 1485 ff, A. S .
T. M. methods
Ferrocarbon-titanium, determination of
Ti in, 983, 987
Ferrochromium, chromium in, 302, 1490
Ferrocyanic acid, 663
Ferrocyanide
method for copper, 377
for lead, 508
for zinc, 1060
standard solution, 1060, 1065, 1202
test for copper, 350
for iron, 463
Ferromanganese, manganese in,. 565,
1486
Ferromolybdenum, 1506
Ferron, reagent for iron, 487
Ferrosilicon, analysis, 808; A.' S. T. M.,
1485
Ferrotungsten, Analysis, A. S. T. M.,
1011 ff; 1503 ff.
Ferrovanadium, 1494
Ferro-uranium, 1028
Ferrous ammonium sulfate, reagent, 571
Ferrous iron (see also Iron)
reduction of chlorates; 274
of chromates, 286
test for palladinm, 722
for platinum, 712
Ferrous salts, as reducing agents, 288
distinction from ferric, 463
Ferrous sulfate
method for nitrates, 644, 2214
reagents, 1206, 2215
standardization, 2215
method for peroxides, 675
for per sulfates, 922
reagent solution, 2206
test for cyanide, 660
for nitrates, 630
for nitrites, 631
Ferrous sulfide, in sodium sulfide, 2182
separation method, 469
Fertilizers, solution of, 863 (see Nitrogen; Potash; Phosphate). ,
Fiber, see Bituminous Substances.
Field assay, of water, 2,987

48

SUBJECT INDEX

Fillers, in grease, 1825

in paper, 1909
in rubber, 1982, 1988, 2018
Filtration apparatus,
bell-jar, 563 (Fig. 60)
microanalysis, 2470 (Fig. 409), 2471
(Figs. 410, 411)
immersion technique, 2470
sulfur determination, 911 (Fig. 110)
Fineness, of silver, 828
of sulfur, 934
Fire assay, of gold and silver 835 ff.
apparatus, special for, 841 (Fig.
105); 842 (Fig. 106)
assay slags, 845
assay ton, 840
balances, 841; 844 (Fig. 107)
borax, amount of, 845
bullion, 854
copper bullion, 853
crucible assa.y, 841
crucible charges, 847
cupellation, 849, 850
cupels, 837
cyanide solutions, 856
definitions, 835
Dore bullion, assay, 855
furnaces and equipment, 837
muffle furnace, 839 (Fig. 103)
oil-fired furnace, 838 (Fig.
102)
two-unit furnace, section,
840 (Fig. 104)
fusing the charge, 846
general outline, 835
gold.and silver ,bunion, 852
sampling, 852
gold bullion assay, 855
U. S. mint method, 856
lead bullion, 853
lead reduction, oxidized ores, 842
sulfide ores, 842
litharge, amount, 843
parting, 851, 852
platinum, palladium, gold and silver,
857
reagents, 836, 837
sampling, 840
scorification assElY, 847, 848
silver, or Dare bullion, assay, 855
silver in ores and concentrates, 860
sodium carQonate, amount, 844
special methods of assay, 857
weighing and mixing, 846
weights, 841
Fire test, of lnbricating oils, 1732
of oils, 1703
.
Fixed carbon, see Coal; Bituminous Substances.
Fixed oils, etc. see Oils, Fats, Waxes.
in varnish, ~852

Flame, coloration ~y. (See 9.lso Spectra.)

barium, 117
boron, 162
\
caesium; 893 '
calcium, 205
copper, 349
lithium, 887
potassium, 862
rubidium, 893
sodium, 876
strontium, 899
thallium, 943
tests, table of, 1108
Flash test or point,
of bitumin. substs., 1533
burning oils, 1702, A. S. T. M.
fuel oils,
Pensky-Marteris test,
1744, A. S. T_ M.
Tag Test, 1752, A. S_
T.M.
.lubricating oils, 1732, A. S.

T. M.
solvents or diluents, 1756
varnish, 1851
Flasks, volumetric,
calibratio.n, 1223 (Fig. 130)
Fleming absorption bulb, 222
Float te~t, viscosity, 1518
Flow meter, 2342
Flowers of sulfur, 935
appearance, 935 (Fig. 117)
Flue dust, selenium and tellurium in, 790
Flue 'gases. See gases.
Fluorescein,
indicator, 273
I
dichlor derivati;ve as, 1201
solution, !1201 '
Fluorescence, 'analysis by, 2546
Pluoride, method for calcium, 216
test for silica, 794
Pluorides, deVu. of silica in presence of,
799
__
Pluorine,
detection,
etching, black filter and ]langingdrop tests, 399-401
estimation,
colorimetric,
Steiger-Merwin method, 423
effect off H,SO" 411 (Fig. 51)
fiuoride-titration, 412 (Pig. 52)
stand. solutions, 411
Zirconium-purpurin method, 423
fluoride equivalent' to Zr, Table, 424
I
interferences, 425
micr-e1l:pplication, 424
gravimetric,
calcium fluoride method, 404

SUBJECT INDEX
Fluorine,
.
lead chlol'oflnoride method, 405
triphenyltinfluoride method, 406
voJ,nmetric,
calcium acetate metho(i, 413
lead chlorofluoride method, 418,
silicon tetrafluoride nlethod (Offerman), 409
Adolpn's apptlratus, !Jo\)l)
(Fig. 50)
zirconium-alizarin-thorl.um method
(Willard and Winter), 407
occurrence, 399, 401
preparation of the sample
calcium fluoride, hydrafluor~c. acid,
Q r g ani c substances, slhceous
ores, slags, soluble fluorides,
fluorspar, 402, 403
removal, before determining calcium,
_208
separation by distillation, 408, 423
from boric acid, 404
hydrochloric acid, -404
.. phosphoric acid, 403
silica, 403
solubility of salts, 402
.
traces, determination of, 4:10, 421, 423
estimation, special procedures
in alkali fluorides, 417
in barium ores, 135
in minerals, calcium and equiv.
fluorine, 415
in organic compound~, 407
in rocks, '407
in sulfuric acid, 2233
in water, 407
Fluorspar, analysis for c~lcium carbonate,
calcium fluoride and silica, 419, 420
l)erchloric acid method, 403
silica in presence of, 802
Foaming, of water, 2086
Foods, aluminum'in, 56-58
arsenic in, 107
boron in, 165, 166
_ tin in, 971
Poot-candle, definition, 1174
Poot-pound, definition, .1175
Foots, in linseed oil, 1843
Pard-Williams method, for manganese, 571
Foreign matter, in sodium ferro cyanide,
668
Formaldehyde, detect. and estimation of,
2149
peroxide method, 2140
test for silyer, 819
Formic acid, analysis of, 2246
in acetic acid, 2249
test. for iridium, 730
osmium, 748
palladium, 722

49

Formic acid, platinum, 713

rhodium, 742
Formulae for:
areas, 1170
diluting mixtur\lS, 2236
gas velocities, 1173
heat value, of coal, 1657, 2384, 2386
pH, electrodes, 2303, 2304
vo)umes, reauctlon to"N. 'L "P., 11'13
Free acid. See Acidity.
Freeport Sulfur Company, methods for
sulfur, 937-941
.
Freezing-point curves for sulfuric acid,
2228
method, solubility, 2584
Fuel gases. See Gases.
Fuel oils. See Oils.
Fuels. See CQal and Coke.
mixed, heat of combustion, 1187 - .
Fuming sulfuric acid, analysis of, 2233
equivalent, table, 2243
Furfural, in acetic acid, 2249
in alcohol, 2140

Furnace, Denver Fire Clay, 1645 (Fig .

233), 1646 (Fig. 234)
electric. See Iron and Steel; Carbon;
Coal and Coke, etc.
for microanalysis, 2472
See also muffle furnaces; ovens; driers.
Fusel oil, in ethanol, 2140
Fusibility, of ash, 1645 .
Fusible metals, analysis, 1399
Fusion, heat of, table, 1188
methods, general, 1336
for chrome ores, 284
for iron ores
_ bisulfate, 464, 477
carbonate, 465, 477
for ni~kel ores, 616
for silica, 797, 799
for tin ores, 956 ff.
for titanium ores, 978
for tungsten minerals,
for vanadium ores, 1033, 1047
See also Prepartion of Sample under
elment or substance in question.
tests, bitumin, substauces, 1523
G

Gadolinite, 246
Gadolinium, 245
Gallium
behavior in solution, 426
detection
. spectroscopic, 426
estimation,
gravimetric, as the oxide, 428
in ores, 428
occurrence, 426
preparation and solution of the sample,
426

50

SUBJECT INDEX

Gallium
separations,
from aluminum, chromium, indium,
uranium, cerium, 427
aluminum, indium and iron, 427
from bivalent metals, 427
iron, 427
uses, 426
Galvanized sheets, zinc determination,
1087
wire, zinc determination, 1088
Gas, .analysis, 2336, 2349
accuracy, 2351, 2352, 2359
apparatus, absorption bulbs, pipettes,
tubes, 2344
measurement of large quantities of
gas, 2339
anemometer, flow meter, orifice meter, Pitot tube, rotameter, Thomas electric
meter, wet meter,
measurement of small quantities of
gas, 2343
Rempel's gas burette (Fig.
366), 2353, separatory
funnel and graduate,
2400
apparatus, analytical, Elliot's apparatus, 2351
Rempel's (Fig. 84), 686 (Fig_ 366),
2353 (Fig. 367), 2354
Orsat's, 2349
pipette, explosion, (Fig. 85), 687
sampling tubes, pumps, containers,
2336
special,
for traces of oxygen (Fig. 86),
688
application and interpretation of results, 2380 ff.
combustion data and c~lculations,
2411-2413
compressed, and handling, 2420
cylinders, 2360, 2420
examination of gases, 294'5, 2346
absorbed by group reagents:
cuprous chloride, 2348
ferrous sulfate, 2348
potas~um hydroxide, 2347
pyrogallol, 2348
silver nitrate, 2347
sulfuric acid, 2346
unabsorbed, 2348
acetylene, det 'n' oxygen, hydrogen,
methane, nitrogen, sulfur-containing gases, phosphine, 2402
air, moisture, carbon, monoxide,
ozone, sulfur dioxide in, 2403
chimney and flue gasel'l, carbon dioxide, indicators, temperature,
heat loss, smoke, det 'n of, 2403

Gas, coke-oven, 2i54

electrolytic gas; chlorine and other
gases, 2402
helium, group, 2348
illuminating gas,
ammonia ini 2396
calorific value of, 2391
apparatus (Fig. 371), 21l90;
(Fig., 372), 2391; (1<'ig.
373), 2392
calculations; 2393, 2394
record, 2433
'
candle-power of, 2391
carbon dioxide in, 2355, 2397
carbon monoxide in, 2356
hydrogen sulfide in, 2396
illuminants, oxygen in, 2356
methane and hydrogen in,
Rempel's method, 2357
Rinman's method, 2356
naphthalene in, 2396
nitrog(\n in, 2358.
specific gravity of, 2397
sulfur in, 2395
gasoline, in natural gas, 2408
industrial gases, analysis of, 2380
liquias, 2414
m i c r 0 methods_ See Microanalysis,
2537 ff.
mine gases, 2400
carbon monoxide in, 2401
" methane in, 2401
moisture in gases, ] 347, 2408
apparatus (Figs. 380, 381), 2408
natural gases, 2407
sampling, 2409 (Fig. 382), 2409
nitrogen in gases" 'nitrometer method,
2418
noble (helium etc.), 2348
oxygen, in gases
.
by explosion, 687, apparatus (Fig.
85),
by phosphorus, 685; apparatus (Fig.
84), 686
__ _ .
by pyrogallate, 686
traces of, 687
,Dely apparatus (Fig. 86), 688
producer, fuel and blast furnace gases,
2388
properties of gases and their estimation, 2360-238p ,
reagents, 2421-2424
sampling of (see also Sampling), 2336
solubility, det 'n of, 2587, 2589 (Figs.
489, 500)
specific gravity of, 2344
standard conditiqns, 2340,'
sulfuric acid g,S.IlSburner gases; 2399
nitrogen oxides in, 2400
oxygen, 2400

SUBJECT INDEX
Gas, sulfuric acid gases, solids (suspended), 2414
sulfuric acid mist, 2416, apparatus (Figs. 385, 386)
sulfur dioxide, 2399
time required for, 2351, 2352, 2359
Tyndall test of, 2414, apparatus (Figs.
383, 384)
,
procedurei'\ with special apparatus:
Elliott apparatus for CO, CO,
and 0" 2351
Hempel apparatus, determination of oxygen,
by combustion with hydrogen, 2353
by phosphorus, 2355
by pyrogallolate, 2355
Junker's calorimeter, 2392
Oi'satt apparatus, determination
of CO, CO" hydrocarbons, 0"
2349 (Fig. 364)
tables
absorption, 2346-2348
aqueous vapor pressure, ,2425
calorific power, 2426
constants, 1166-1167, 1171-1172,
2424
correction factors, 2429-2430
.eylinder dimensions, 2428
dangerous concentrations, 2427
emergent stem corrections, 2431
humidity corrections, 2432
specific gravities, 2411
specific heats, 2425
volumetric factors, 2428
water, determination in, 1347
Gas, velocity of flow (Pitot), 1173
Gas volumes, conversion formulae, 1173
Gasoline; see Oils
in natural gas, 2408
specifications, table', 1791
test of, 1742
Gay-Lussac Method, for silver, 826 (Fig.
100)
pipette, 829 (Fig. 101)
Geissler absorption bulb, for. CO" 221
Gelatin dynamite, analysis oi', 1668
Gelatin, effect on pH measurement, 2288
nutrient, 2120
Gelation, of tung oil, 1847
Generator, traces of antimony, 79
Gerhardt absorption bulb, for COo, 221
Gerplanium,
detection, 429
estimation
gravimetric
.
as magnesium orthogermanate, 430
'its oxide, 430
pcurrence, 429
preparation and solution of the sample,
429

51

Germanium,
separa tions
by vola tiliza tion, 430
from the copper group, 429
volatility (chloride), 429
Glass, analysis of

alkalies, 2166
alumina, ferric oxide, lime, magnesia, lead oxide
and silica, 2165, 2166
boric oxide, 2167
calcium and barium separations, 2168
factors, 2169
.
Glass electrode, 2303
Glass-making materials, analysis
barium carbonate, 2172
feldspar, kaolin, clay, 2170
lime, 2173
potash, 2169
sand, iron in, 482, 483
Glucose, in boric acid titration, 180
Glucosides, detection of, 1951 ff.
Glue, det'n in rubber, 1986, 1987, 2007
Glycerine (or glycerol)
detection, in grease, 1827
estimation, 2150, 2153
in soap, 2038, 2039
in boric acid titrations, 170, 171
Gold
detection
benzidine acetate test, 432
in alloys, 431
minerals, 432
phenylhydrazine acetate test, 433
estimation
colorimetric
Cassell, Moir and Priester methods, 438, 439
gravimetric
electrolytic, 435, micro method,
2518
fire assay, 835 ff.
wet assay, 435, 436
Volumetric
iodide method, 437
Lenher's method, 438
permanganate method, 437
occurrence, 431
precipitation as metal, 434
preparation and solution of the sample, 433, 436
proof gold, preparation, 440
separations
by displacement, 433
from platinum metals, 434
from tellurium, 434
solubility, 433
Gold bar, platinum and palladium in,
725, 727
Gold bullion, fire 'assay, 855, 856
G"rain, water in, 1340

52

SUBJECT INDEX

Grain-growth in steel, see Metallography Hexabromide test,\for linseed oil, 1779

Graphite, carbon in, 235
yield, from various oils, 1780 in iron and steel, 232, 1430 (see also
l
Holmium, 245
Metallography)
Horse-power, defi11ltion, 1175
Gravimetric methods, see individual ele- H. T.H. = nigh-Te~t Hypochlorite
ments.
.
Hydrazine sulfate,1 test for Se and Te,
Gravity, see Specific gravity
776
Grease analysis of, 1823
Hydrazine-sulfur dioxide method, for
Greases, 1791
tellurium, 785
Green pigments, see Pigments
Hydriodic acid, reagent for "insolGrinders, 1310, 1311
ubles," 1335
Guanidines, ill rubber, 2017'
removal from nitric acid, 635
Gum arabic, in explosives, 1677
solubility of mercuric sulfide in, 577
test for, 1675
- Hydrobrolnic acid, removal from nitl'ic
Gum content, of gasoline, 1814
acid, 635
Gumming test, of oils, 1737
Hydrocarbons
Guttapercha, 1966, 1992
in gas analysis, 2350
Gutzeit apparatus, 107 (Fig. 14)
in illuminating gas, 2356, 2357, 2396
Gutzeit method, for arsenic, 101-108
in natural gas, 2407
micro, 2504
in rubber, 1966, 2008, 2009
applied to antimony, 78-82
in soap, 2040
purification of acid for, 104
Hydrochloric acid (see also Adds)
Gypsum, analysis, A. S. T. M.
analysis of, 2203
calculations, 215, 216
detection, 261
carbon dioxide,' 214
estimation, gravimetric, 269
iron and alumina, 214
free, determination of, 277
lime, 214
volumetric, 2209
magnesia, 215
in acetic acid, 22:51
sulfur trioxide, 215
in formic acid, 2247
water, combine~, 214
in liquors, ammoniacal, 638
free, 214
in nitric acid, 2212
in grease, 1825
in presence of chloric and perIn paints, 1866
chloric acids, 276
1I
in sulfuric acid, 2232
impurities in
Hafnium, see Zirconium, 1093 ff.
arsenic, 89, 106, ,2204
Halogens, determination in organic compounds, 265 ff.
barium chloride" 2204
chlorine, free, 2204
micromethods, 2494
silica and :soli.ds, 2205
separation and determination in their
sulfuric acid and sulfates, 2204
mixtures, 276
method for separating potassium alld
test for, para-aminodimethylaniline,
sodium, 868
281
preparation, arsenic-frellj-l_04, 1196
Halphen test, cottonseed .oil, 1778
Hanging-drop test, for fluorine, 400, ap-specific gravity, 2205
tables, 2206-2208
paratus (Fig. 48)
constant-boiling acid, 2f96
Hard lead, see Lead
standard solution, 2195
rubber, see Rubber
gravimetric standardization, 2196
Hardness, determination, of water, 2082
test for iron, 462
of bitumens, 1520
Heat, definitions, 1174
I
lead, 500
mercury, 562 I
of combustion, table 1:l85, 1186
mixed fuels, 1187
silver, 818
of fnsion, eleibents and compounds, Hydrocyanic acid, commercial, analysis
tables, 1188, 1189, 1190
of, 665
detection, 660, 1946
passing up chimney, calculation, 2382,
2383
gaseous,-2347, 2q!i.3
loss due to CO, 2387
volumetric estimation
bromine- method (Schulek), !i62
test, for oils, 1847
treatment, alloys, see Metallography
copper method, 660
silver method (Liebig), 661
Helium, 2348, 2379

SUBJECT INDEX
Hydroferricyanic acid, 663
Hydroferrocyani~ acid, 663
Hydrofluoric acid,
analysis of, 2209
hydrofluosilicic acid in, 2209
method for alkalies, 883
,sulfuric acid in, 2210
sulfurous acid in, 2210
Hydrofluosilic acid, in hydrofluoric, 2210
Hydrogen
active, micro method, 2547
detection, 442 _
estimation
in coal, 1649, ff ..
in gases, 2356, 2357, 2374
in organic compounds, 223, 224
micromethods, 2475 ff.
semi micro, 225
in' steel, 443 (Fig. 54)
isotopes, 444
occurrence, 442
properties, 442, 2348, 2373
reduction method for
copper, 389
osmium, 748
rhodium, 742
tin ores, 957
Hydrogen antimonide, 2347
arsenide, 2347
Hydrogen chloride, gaseous
absorption, 2347
determinatjon, 2363
generation of, 13
properties, 2362
Hydrogen electrode, 2300 (Fig. 324);
2309
platinizing, 2301
Hydrogenion concentration; see also pH
ana the pH scale, 2277
determination of
colorif;letric, 2274 ff.
electrometric, 2297 ff.
of paper extracts, 1916
of wllter, 2073
. buffer solutions, 2074 ~
indicators, 2073
Hydrogen peroxide, detection, 904
determination
arsenious acid method, 2180
_ ceric sulfate method, 1205
iodide method, 2180
permanganate method, 2181
method for chromium, detect., 282
formaldehyde, 2149 .
iodide, 456
titanium, 987
_
vanadium, deteCt., 1031
-phosphoric acid method for iodides,
456
test for titanium, 976
Hydrogen phosphide, 2347

53

Hydrogen suInde
detection, 904
estimation, 924
evolution method, 911, 914
(see also iron and steel)
generator for,
Scott's, 97
Urbasch's 98
group, see Qualitative analysis,
1111 ff.
see Separations under various eJements.
in ammoniacal liquors', 638
in gases, 2347, 2364
illuminating, 2396
in water, 2078
method for iron, 469
precipitation of molybdenum, 592
platinum metals, 710
reduction of chromates, 286
test for copper, 350
colorimetric determination, 380
for iridium, 730
for lead, 500
for mercury, 574
for nickel, 6'14
for osmium, 748
for palladium, 722
for platinum, 712
for rhenium, 768
for rhodium, 742
for ruthenium, 737
for selenium, 776
for tellurium, 776
for tin, 954, 960
for valladium, 1031
for zinc, 1054
Hydrometer, 242, 1513, 1514 (Fig. 187),
See also Specific gravity.
jar, 366 (Fig. 44)
Hydrosulfite, sodium, 930
Hydroxide see under the element in question
ion, in water, 2066, 2095
see Alkalies; Sodium hydroxide; Potassinm hydroxide, etc.
1,2,5,8Hydroxyanthraquinone, test forAI, 2
for Be, 138
colorhnetric method, 143
volumetric method, 145
Hydroxylamine, determination, 647
8-Hydroxyquinoline, see Oxine.
Hypochlorite; see also Bleaching powder
reagent for phenol, 1950
test for, 262
in presence of chloride, 274
Hypochlorous acid, detection, 262
Hypophosphorus acid,
reagent for Se and Te, 776
test for, 689
Hypotheses, 1191-1193

SUBJECT INDEX

54
.I

Identification, by the microscope, 2434

Ignition loss, see Loss on Ignition
Illinium, 245
Illmenite, analysis, 495-499
Illuminants, in gases, 2356
Illuminating gas, analysis of, 2355
Impurities, see under individual substances.
Indicators (see also under Reagents)
acid-alkali, 2191
stock solutions, 2192
adsorption, for precipitation titrations
for halides, 192, 273" 45~, 2499
for mercury, 582
for silver, 830
for sulfates, 919
oxidation-reduction, 1205, 1206, 1207
pH measurement, 2281
Indigo carmin, 2158
Indigo solution, standard, 930
Indigotin disulfonate, indicator, 679
Indium
behavior in solution, 445
detection, 445

estimation, as oxide, 446

as sulfide, 447
in zinc blende, 447
preparation of the solution, 446
separation'
from alnminum, 446
gallium, 446
iron, 446
manganese, 446
zinc, 446
Industrial gases, see Gases
Inks, printing, analysis, 2174
determination of oil, 2174
separation from pigment, 2.174
pigment analysis
black, 2175
blue, 2177
green, 2178
red, 2177
Insecticides, water soluble arsenic in, 90
.Insoluble matter
in asbestine, china clay, silica, sjlex,
1866
in cement, 1601
in composite paint, white, 1867
in explosives, 1667, 1678
in green pigments, 1875
in gypsum, 214
,
in nichrome, 1408
in pigments, 1875
in potash, 2169
in red lead or orange mineral, 1869
in rubber, 1973
in soap! 2037, 2038
in soda ash, 2262

Insoluble .matter
iJ1 sodium fludride, '420
in sodium hYdroxide, 2256
in sodium nitr,ate, 645
in sodium sulfide, 2182
in trisodium phosphate, 705
International Niekel Company, standal'il'
nlethods, 619, 624-628
Interpretation. of results of analysis, see
Gases; Oil; -Water.
Iodate, determination, 458
and periodate in mixtures, 458
method for arsenic, 100
potassium, 454 .
methods, 99, 372, 455, 582, 1209
Iodide method
for antimony, 75
chromillm, 288
topper, 368
'gold,437
manganese, 570
oxidation by chlorine, 456
to io~ate, 456
to iodine mono chloride, 455
Iodine
detection, 448
,
estimation, general procedures
gravimetric
as palla dons iodide, 452
as silver iodide, 451
volumetric
hydriodic acid, solnble iodides, 452
liberation of iodine'
by chlorine (Mohr-Dupre),
456
ferric salts, 454
hydrogen peroxide, 456
nitrous aci", (Fresenins),
, 454', 455 (Fig_ 55)
,,potassium iodate, 454
oxidation to iodate,. 456
iodine monochloride, 455
Volhard's method, 457
with silver nitrate, 457
adsorption iirdicator, 457
.
'.
speCial procednres
in brines, 459
in feedstuffs, etc., 460
in halogen mixtures, 276
in nitric acid, 2213
in organic substances, 450
in water f 459
occnrrence, 448.
preparation of the sample
iodides, iodine (commercial), iodates, minerals, organic substanc~s,. phosphates, water, 449,
450
_
reagent, 0.1_ N;'preparation, 965, 1207
separation, frolli heavy metals, other
halogens, 449

SUBJECT INDEX
'iodine
solubility, element and its salts, 449
traces, estimation of, 460, 461
Iodine method
for antimony, 76
for arsenic, 99
for hydrogen sulfide, 924
for metabisulfites, sulfites, sulfurous
acid, thiosulfate, 926
for molybdenum, 594
for tin, 964
Iodine monochloride, 455
reagent, 1209
. Iodine number, see also Oils; Paints;
. Soap
. determination
Hanus, 1767
Hiib1, 1768
of Unseeu oil, 184:i
value, of soap, 2043
Iodine pentoxiue, method for CO, 2406
Iodine test, for Pd, 722
Iodoform test, for alcohol, 213T
7 Iodo:8hydroxyquinoline (ferron), iron
reagent, 487
.
Iodometric mGthods. See also Iodine and
Potassium Iodide methods.
for available oxygen, 675
for manganese, 570
for molybdenum, 593
for nitrates (Gooch and Gruener),
647
for selenium, 793
for tellurium, 793
Ionic mobilities, table, 2324
Iridium
detection, 730
estimation
gravimetric
ammonium "hlol"id(', method, 732
DevilleStasGilchrist method, 733
micromethod, 2547
precipitation as dioxide, 735
residue method, 732
zinc reduction method, 732
occurrence, 730
preparation and solution of the salllple,
731
separations, from osmium, 732
from platinum, 731
solubilities, 730
test for nitrite, 655
Iron, "Armco," 808
detection
hydrochloric acid, ferrocyanide, saHcylic acid, sodium peroxide and
thiocyanate tests, 462, 463
estimation, general procedures
colorimetric
ferron method, 487
salicylic acid method, 487

55 .

J ron, standard solutions, 1199

thiocyanate method, 486
thioglycollic acid method, 487
gra vimetric
as tHe oxide, Fe2 0., 467
cup ferron method, 468
separation as FeS, 467
volumetric
oxidation methods, 469
eerie SUlfate,' 474
potassium dichromate, 471
diphenylamine in dica tor,
473
microapplication, 2522
potassium permanganate, 474
reduction, prior to titration
by hydrogen sulfide, 469
by metals, 470
test lead, 470
by potassium iodide, 470
by stannous chloride, 470
by sulfurous acid, 470
by zinc, {70
reduction methods,
with potassium iodide, 482
,yith stannous chloride, 484
with titanous salt, 479
estimation, SIJecial procedures (sec also
. Iron oxide and Alumina)
in alumina, 27
.
in aluminum, and its alloys, 42
in aluminum ores, 20
salts, 15
in alundum, 34
in "arsenic," 110
in barium ores, 134
in bauxite, 24, 997
in bisllluth bronze, A. S. T. 1\1., ];) 77
in bronze, A. S. T. M., ] 36,)
in copper, 382, 383
in German silver, A. S. T. ]\f., 1373
in glass, 2166
'
in glass sands, 482
in illmenite, 497
in lead (pig), 1418
in magnesiulll alloys, 548
in manganese bronze, A. S. T. :M.,
1357
in monel metal, 1379
in nichrome, 1407
in nickel, 627
in ores, 469, ff.
_ optional method, 477
in paper, 1915
in pigments (paint), 1872, 1873,
1876
ill sand, 482
in SlIver solder, 1404
in slag, 2022
in stellite, 327, 1411
in sulfuric acid, 2230

56

SUBJEOT INDEX

Iron, in tin-base metal, 1389

in titaniferous ores, 995
in water, 2059
ill zinc, spectrochemical method, 1089
ill zinc-base alloy, 1396
in zilte ehtoride, 1080
occurrence, 462
preparation ahd solution of the sample,
463
iron and steel, 142~ ff.
ores, salts, silicates, 464
fusion methods, 464, 465
separations
ether method, 465
from aluminum, sulfirle'method, 467
beryllium, 141, 142
manganese, basic acetate, 558
general method, 465
iron group from Al (U, V, Ob, Ta,
P, Zl'), 466
with cupferron, 466
solubilities, 464
traces, 486, 487 .
,
Iron and Steel. See also Metallography.
analysis, A. S. T. M. methods,
1421 ff.
mierolllethods, 2545
determination of:
aluminum, 52, 53, 1482
arsenic, 91, 112
carbon, combined, 232
graphitic, 232, 495, 1430
total, 229, 488, 1426
apparatus, 1428 (Fig. 184)
details, 489
chromium, 291, 293, 495,' 1454,
1473, 1477 .
potentiometric, 2315
copper, 355, 1446, 144$
cobalt, 323, 1466
graphite, 232, 491l, 1431l
hydrogen, 443
manganese
bisn:uthate method, 491" 494,
1431
lead oxide method, 491
periodate methoa, 1470
persulfate method, 490, 1434,
1471
molybdenulll, ~olorimetric, 606;
609, 610, 1464, 1481
gravimetric, 608, 1461, 146.2
nickel, 495
cyanide method, 624, 1452
electrolytic, 1451
glyoJ(ilIle method, 1450
nitrogen, 656
oxygen, 684 (Fig. 83)
phosphorus, 492, 494, 699, 700,
1435, 1438

Iron and steel, in\ alloy steels, 701, 703,

1437
in cast iron, 702

selenium, l'!8Q
silicon, 493', ',494, SOG, 1444
condudivity methl)d, 811
perehlol'ic aciu method, 1446
rapid foullury method, 493
sulfur, graVimetric, 492, 911, 1440
volumetric, 492, 494, 913, 1442
apparatus, 1443 (Fig. 18i)
titanium in, 1468, 1482
tungsten in, 1465
uranium, 1028
vanadium, 293, 1458, 1459, 1473,
1477
detection, 1031
potentiometric method, 2315
zirconium, 1470, 1482 ,
preparation of sample, 488
precautions, 1423
reagents, 697, 699
sampling, 1422, 1423 (Fig. 183)
Iron are, chrominm in, 291 '
11'011 ore briquettes, copper in, 354.
reduction for sulfur determination,
913
Iron oxide and alumina
iLl barium carbonate, 2172
in barium ores, 134
in blanc fixe, 136
in caustic soda, 2259
in chrome ores, 297
in gypsum, 214
in magnesia, 538
.in magnesite, 5,36, 538
in sand, 806 I
in soda Il.sh, 2262
in sodium nitrate, 645
in sodium silicate, 805
in tartaric acid, 2255
in zinc chloride, lOSO
Iron oxidea, as pigments, 1.87_~
Iron salts, free acid \n, 2245
---Iron sulftde, available H,S, 914
Irrigating waters, examination, 2098
Isopropyl etller, extraction of ferric cJllorkle, 465
J
I
Jones reductol', 476 ,(Fig. 56)
apparatus, pl'epamtion and blank,
476
method f01: iron, 469, 475
molybdenum, 594
I
apparatus, 1595- (Figs. ~4, u5)
pho13phoru~, ~99 _
.
titanium, 495, >984
vanadium, 1039
J oulo, definition, 1175

SUBJEOT' INDEX
K
Kaolin, 2170
Kilowatt, 1175
Kjehldahl digestion
apparatus, 633 (Fig. 633)
selenium catalyst, 633
Koettstorfer number, 1771
Krypton, 2348
properties; 2380
L
Laboratory, for micronalysis, 2461
Lamp metllQd, for sulfur, A. S. T. M.,
1713
\
Lanthanum, 245
Lard, see Oils: fats, etc.
Latent heat, definition, 1174
Latex, 1968
Llws, ehemical, 1191-1192
I"ead acetate, precipitant, 2335
~ead,
.
detectIOil, general procedure, 501
tests with HCI, H 2 S, chromate, 500
estimation, general procedures
colorimetric
dithiione method, 521
reagents, 522
stand:n-d lead solution, 1200
sulfide method, 517, apparatus, 518
.
reagents, 519
gravimetric
as chromate, 507
as molybdate, 506
as peroxide, PbO" by electrolysis,
507
micromethod, 2515
as sulfate, 504
volumetric
clnomateiouide method, 512
ferrocyanide .method, 508
.molybdate method, 511
permanganate method (A. H.
Low), 509
estimation, special procedures
in alloys and metals, 505
in aluminum and its alloys, 43
in "arsenic," 110
in basic lead carbonate, 524, 1859
iri basic lead sulfate, 1856
in bearing metals, A. S. T. 1\-1., 1364
in bismuth bronze, 1375
in blanc fixe, 136
in. brass, 392
in bronze, A. S. T. M., 1354, 1355
in copper, 386
in corroded white lead, 524
in glass, 2165
in leadbase alloys, A. S. T. M., 1381,
1384
in lead ore, 523
in magnesium alloys, 552

57

Lead,
in orange mineral, 524;, 1869
in pigments
chrome green and yellow, 1856,
1875, 1876
composite white paint, 1868
see also basic carbonate, sulfate,
red lead, etc.
in picric acid, soluble, 1683
in red lead, 524, 525, 1869
in selenium, commercial, 788
in silver solder, 1404
in sublimed blue lead, 1858, 1874
in sulfuric acid, 2230
in tartaric acid, 2255
in tin-base alloys, A. S. T. M., 1381
1384, 1387
in water, 2080
in Wood's metal, 1399
in zinc (slab), 1071
spectrochem. method, 1089
in zinc-base alloy, 1396
in zinc lead and leaded zinc, 1861
in zinc ores, 527
impurities in metallic lead,
bismuth in lead bullion, 159
complete analysis of pig lead,
A. S. T. M., 1413-1420 (de
termining Bi, Ag, As, Sb, Sn,
Fe, Co, Ni, Mn)
zinc, 1083
industrial application of met)lOd, 501
ocurrence, 500
pi,g lead, 1413
preparation and solution of the samplE
minerals, ores, alloys, 501, 502
reducing agent, 470
separations
as lead sulf~te, 502
extraction with ammonium ace
tate, 503
from barium, 124
barium, columbium, tantalum, am
monium sulfide and ammonium
carbonate groups, 503
calcium, 503
solubilities, the metal and its salts,
501,502
traces, See also, estima,tion, colorimet
ric, 517":522 .
gravimetric, 514
acetate extraction, 515
dithizone extraction', 521
occlusion' by hydroxide of anotl1er
metal, 516
Seeker Clayton method, modified,
517
Lead acetate method
for molybdenum, 589
for vanadium, 1036
test, for iridium, 730

58

SUBJECT ,INDEX

Lead arsenate, arsenic in, 90

Lead azide, 1685
Lead-base alloys, analysis, 1381
Lead basic carbonate, analysis, 1859
lead in, 1860
Lead basic sulfate, 1856
Lead bullion, assay, 853
Lead carbonate, in blue lead, 1874
Lead chloro-fluoride method, for fluorine,
405
Lead chromate, test for Cr, 282
Lead nitrate, as precipitant, 2335
Lead oxide, fusion method for silicates,
799
in blue lead, 1874
in glass, 2165
Lead peroxide, electrodeposition, 507
in red lead, 524, 525, 1870
method for Mn, 491
Lead slimes, Se and Te in, 789
Lead sulfate, decomposition, 907
in basic lead sulfate, 1857
in blue lead, sublimed, 1874
Lead sulfite, in sublimed blue lead, 1874
Lead titanate, in pigments, 1864
Liebig bulb, 221
cyanide method, 661
Light absorption, by glass, 1648 .
Light filter, chrome alum, 862
Lime, see also Calcium Oxide
analysis, for glass, 2173
and magnesia

in barium ores, 134

available, 212
chloride of, 278
free
in bleaching powder, 2161
in calcium carbonate, 2270
in cement, 1618
from silicon carbide, 813
in cement, 1617
in chrome ores, 299
in gypsum, 214
in magnesia, 538
in magnesite, crude or burned,,536, 538
in sand, 806
in slag, 2021
in soda ash, 2262
in sodium nitrate, 645
in sodium silicate, 805
in zinc chloride, 1081
method for
fixation of methyl borate, 166
halogens in organic matter, 266
value, in water analysis, 2084
Limestone
analysis of, 1617
in cement, 1617
silica in, 802
Linde Air Products Co, Standard method
for oxygen, 670-674

Lindo-Gladding method, for potassium,

871
1
Linseed oil, analysis, 1843 ff.
hexabromide tes~ for, 1779
Liquid bleach, 279 I
Liquid fuels, analysis of, 1742 ff.
Liquids, evaporation of, 2233, Fig, 309
in gases, 2414
sampling, see Sampling
Litharge, see Fire Assay
Lithium,
detection, 887
estimation, general
as lithium <1,hloride, 888
alcohol-ether procedure, 890
as lithium sulfate, 888
Gooch's method, 888
other methods, 892
Rammelsberg's method, 887
spectroscopic method, 891
estimation, special case
in water, 2065
occurrence, 887
separations
from sodium and potassium, 890, 891
from sodium, potassium, rubidium
{ and caesium, 884-886
LithIUm oxalate, as reagent, 2335
Lithium sulfate, as. reagent, 2335
Lithopone, analysis of, 1862
Logarithms, tables, 1779-1780
Loss
on heating, linseed oil, 1843
on ignition
aluminum hydrate, 28
ores, 19
barium ores, 136
bauxite, 23 /
blanc fixe, 136
calcined alumina, 26
. cement, 1600
magnesia, 538
magnesite, 536
soda ash, 2262
Low-freezing dynamite;"analy_sis, 1669
~
Lower oxides
in nitric acid, 2212
in oleum and mixed acids, 223,4
Lutecium, 245
Lux, definition, 1174
M

Magenta test, f6r bromine, 188

quantitative application, 193
Magnesia, see also Magnesium, estimation; Lime and Magnesia:
analysis of, 5~8
'
in glass, 21!i.6in slags, 2023 mixture, preparation, 695, 696

SUBJEC'l' INDEX
Magnesium
detection, 528
p-nitrobenzeneazoresorcinol test, 529
titan yellow test, 529
estimation, general
gravimetric
,
as magnesium pyrophosphate,
532
as the oxinate, 535
volumetric
ammonium magnesium phosphate
method, 534
oxinate method, 47, 535
estilpation, special procedures
in alloys, 552
in aluminum alloys,
oxine method, 47
pyrophosphate method, 45
in aluminum ores, 22
in alundum, 35
in barium cal'bonate, 2172
in barium ores, 134
in bauxite, 25
in cement, 1604, 1615
acid alkali method, 1605
oxine method, 1607
in gypsum, 215
in magnesia, 538
in magnesite, 537
in paint and pigments
composite white paint, 1868
green pigments, 1876
orange and yellow pigments,
1875
in sand, 806
in soda ash, 2262
in sodium nitrate, 645
in sodium silicate, 805
in water, 2062, 2096
in whiting, 1866
in zinc-base alloy, A. S. T. M.,
1395
in zinc chloride, 1081
occurrence, 528
preparation and solution of the sample,
530
separation
as magnesium oxinate, 531
from the alkalies, 531
from the alkaline earths, 121, 531
. from hydrogen sulfide group metals,
531
'
from iran, aluminum, manganese and
zinc, 531
solubility, 530
Magnesium alloys, 539-552
Magnesium ammonium phosphate
method for magnesium, 532
method for phosphorus, 695, 696
properties, 534
:Magnesium chloride, in water, 2083

59

.Magnesium pyroarsenate, determination

of As, 96
Magnesium pyrophosphate, method for
phosphorus, 695
'
Magnesium uraeyl acetate,
method for sodium, 879
test for sodium, 875
Manganese,
detection, bead test, 555
general procedure, .554
in soils, minerals; plants, etc., 554
estimation, general methods
colorimetric
periodate method, 5,73
persulfate method, 572
red lead method, 573
gravimetric
as the dioxide, 559
bromate oxidation, 559
as the pyrophosphate, 560
volumetric
bismuthate method, 562
conditions for, 566
iodometric method, 570
persulfate, 572
arsenite, 572
ferrous sulfate, 571
Vol h a r d's permanganate
method, 561
estimation, special procedures
in aluminum alloys, 48
in alundum; 35
in bismuth bronze, 1378
in bronze, A. S. T. M., 1358
ill cast iron, 1433
in cement, 1611
in ferromanganese, 565, A. S.
T. M., 1486
'
typical data, 566
in German silver, 1374
in glass, 2166
in magnesium alloys, 542
in manganese bronze, A. S.
T. M., 1358
in metallic manganese, 565
in monel metal, 1380
in nichrome, 1407
in nickel, 627
in ores, 563
nnal estimation, 565
reagents, 563
solution of ore, 564
trial determination, 564
typical analysis, 565
in rubber, 1972
in slag, 2023
in steel, 490-492
bismuthate method, 1431
period ate method, 573, 14,70
persulfate method, 1434, 1471
in stellite, 327, 1412

60

SUBJECT INDEX

Manganese, estimation,
Mercurous iOll, titration with halide, 582
in water, 2063
Mercurous nitrate
. in zinc-base alloy, 1398
method f6r chromium, 287
in zinc chloride, 1081
molybd~num, 590
industrial importance_ of methods, 555
vanadium, 1036
occurrence, 554

test for cluomate,. 282, 287

preparation and solution of the samruthenium, 737
ple, 555-5:>7
Mercury, cleaning, 2422
alloys, ferro aluminum, metallic
detection, 574
chromium,
ferrotitanium,
estimation
mangaliese
bronze, molybgravimetric, electrolytic n;ethod, 579
denum and tungsten alloys,
micro, 2516, 2518
silicon alloys, iron and steel,
Halloway-Eschka
method,
578
. ores, sulfides, slags; silico(Fig. 63)
manganese; ferrochrome, fersulfide method, 577
rotungsten.
volumetric, Seamon's process, 580
separations
thiocyanate method, 580
as manganese dioxide, 558
other methods, 582
from hydrogen sulfide group, alkaestimation, special cases
line
in cyanide precipitates, 584
earths and alkalies, 557
in cyanide solutions, 583
from iron and aluminum, basic acein organic matter, 581
tate method, 558
in rubber, 1989
from nickel and cobalt, 558
. in zinc amalgam, 582
solubility, metal and oxides, 555
occurrence, 574
traces, 572, 573
/ poisonous properties, 2423
:lIfanganese dioxide, precipitation, 559.
preparation and solution of the sam
Manganous oxide, in aluminum ores, 22
pIe, 575
.
bauxite, 25
'
'reductions by:
Mannitol, in boric acid titration, 170
Marsh gas, see Gases; see Methane
antimonic acid, .78
molybdenum, 596
Marsh test, for arsenic, 87, 109
separation
'
Masses, table, 1176
by distillation, 576
Matte, copper, 352
from As, Sb, Sn, Pb, Bi, Cu, Cd, Se,
slag, 353
Te,576
Mauruene test, oils, 1701, 1766
from metals of subsequent groups,
Measuring apparatus, volumetric, ~219576
1227
frpm organic substances, 576
pipette, automatic, 563 (Fig. 61)
solubility, 575
Meat, arsenic in, 107
uses; 574
boron in, 166
volatility, 575
Melaven cell, electrolysis, 47
Mercury fulminate, analysis, 1685
Melting points, under the microscope,
Mercury test, for osmium, 749
.
2453
Metallic ions, qu'lilitatiye analysis, 1111
temperatures, elements, table, 1140
tables of reactions, 1114-1127

Mercaptobenzothiazole, in rubber, 20] 7

Metal, see individual subjects, AlumiMercuric bromide and chloride paper, 79,
num, Barium, etc.; see AllOYS, Iron
103, 1197
and ,Steel, Ferroalloys.
lHercuric chloride, test for tin, 954 .
Metallography, 2548
Mercuric cyanide, detection, 1947
test for Pd, 710, 721;
brass, cold rolled, 2557 (Fig. 461)
case har(lening, 2567
.
Mercuric oxide method, separation of
copper-zinc alloys, metallography of,
magnesia, 867
2556 (Fig. 459), 2557
, test for HCN, 660
equilibrium diagrams, 2553 (Fig.
Mercuric perchlorate, reagent, 2335
455), 2554 (Fig. 459), 2555
Mercuric pyridine bichromate, 582
equipment, 2551 ,(,Fig. 452)
Mercuric sulfide, extraction of S from,
etchin]J :Jor micro examination, 2552
577 (Fig. 62)
reagen.ts. for, 2552
precipitation, 577
eutectic, 2555 (Figs. 458, 459)
solubility in hydriodic acid, 577
llCat treatment, of steel? 2566
Mercuric thiocyanate test, for zinc, 1055

SUBJECT INDEX
:Metallography, influence of various elements in iron and steel, 2568
iron-carbon alloys, 2558 (Figs. 462,
463)
photomicrographs
(Figs.
4644 i1), 2560-2563
iron, pure, photomicrograph, 2554
microscope, directions, 2548
X eumann lines, 2565
I hotographic materials, 2550
slip bands, 2564
specimens, preparation, 2550
Metaphosphoric acid, test for, 690
Meters, see Gases.
Methane, determination, 2356, 2357
properties, 2348, 2375
Methoxy groups in pulp, 1891
microdetermination, 2527
Methyl acetate, mixtures with methanol,
2147
Methyl alcohol (Methanol), detection anll
llet'n, 2147 ff.
Methyl borate, distillation, 166
apparatus, 167
Methylene blue, method for phosphorus,
707, 708 (Fig. 87)
Methyl orange indicator, 2192
red indicator, 2192
Methyl violet, test paper, 1694
Metric aud other units, tables, 1169,
1170
Microanalysis, quantitative, or
Microchemical analysis, 2460 ff.
acetyl determination, 2523, 25~6
app., 2524 (Fig. 436)
applications, references, 2545, 2546
arsenic determination, 2503
.
micro Gutzeit, 2504
apparatus,
2505
(Fig.
426), 2506
reagents, 2505
burettes, 2472
carbonate, 2482, 2483 '(Fig. 422)
carbon-hydrogen determination, 2475
absorption tubes, 2476, 2477
(Figs. 415, 416)
apparatus, 2475 (Fig. 414)
notes, 2479
procedure, 2478
colorimetry, 2472
copper, gravimetric, 2520
in organic material
Biazzo method, 2519
carbamate method, 2520
drying, 2464, 2465 (Fig. 402), 2466
(Fig. 404)
furnaces and heaters, 2472
automatic, 2474 (Fig. 413)
gas analysis, 2537
acetylene, 2543
ammonia, 2544

01

Micl'ogas analysis, apparatus, 2537 (Fig,

445), 2538, (Figs. 446, 447),
2538 (Fig. 448), 2539 (Fig.
449), 2540

beads, 2541
carbon monoxide, 2543
ethylene, 2543
hydrocarbons, 2544
hydrogen chloride, 2543
nitrous oxide, 2545
oxygen, 2541
samples, 2541
water vapor, 2543
halogens
Pregl's method, 2494
volumetric method, Cl and Br, 24.8
I, 2497
use of adsorption indica tors,
2499
Iron, volumetric, 2522
manipulations, 2469
filtering, 2469, 2470
immersion method, 2470
Neubauer crucible, 2469
methoxy-ethoxy groups, 2527 (Fig.
437)
microbalance, 2461 (Fig. 400)
microclectroanalysis, 2,510
Clark and Hermance apparatus, 2511
(Fig. 431), 2512 (Fig. 432)
estimation of nickel, zinc, tin,
2513
traces of metals, 2514
zinc, lead, copper, cadmium, mercury, 25142516
Pregl cell, 2516 (Fig. 434), 2517
(Fig. 435)
. procedure, 2517
estimation of
mercury, 2518
gold, 2518
silver,' 2518
microlaboratory, 2461
mineral analysis, 2545
molecular weight, 2530
rise of boiling-point, 2530
Rast method, 2533
Victor Meyer method, 2534
apparatus, 2535 (Fig. 443)
calculations, 2536
notes, 2536
nickel, gravimetric, 2521
nitrogen
micro-Dumas method, 2485
calculations, Niederl and Trautz,
2488
Pregl, 2487
carbon dioxide generator, 2489
liquids low in N, 2491

62

SUBJECT INDEX

MicroDumas method, notes, 2488

Microscopy, crystalline substances, 2435
vapor pro of KOll sol 'ns, 2491
optical properties, 2435
. micro-Kjehldahl method, 2492 (Fig.
refractive index, 2436
425)
technique~, 2439
organic radicles, 2546
decantation, 2440
phosphorus, 2499
distillation, 2440
colorimetric method, 2501
filtration, 2439
physical methods, 2546
Microcosmic salt, table of reactions, 1109
polarographic method, 2546
test for silica, 795
potentiometric methods, 2322
test for titanium, 876
sampling, 2464
Milk, boron in, 165
specific applications, elements and
Milliequivalents, of inorganic compoundR,
compounds, 2547
1161-1164
standard solutions, 2472
Mine gases, see gases
steel, 2545
Mineral acids, see Acids
Mineral filler, in paper, 1909
sulfur
.
Mineral matter, in sulfur, 933
combustIon method, 2507
fusion method, 2507
Mineral residue, in water, 2077
Meulen method, 2509
Minerals, list of common, 1183-1184
Parr bomb method, 2508
microanalysis, 2545
water, 2545
~est for gold, 432
weighing, example, 2464
water in, 1340
balance, 2461
see under Occurreuce nnder the dam i, nating element, Yol. I
bottles, 2465 (Fig. 403), 2467 (Fig.
405)
Mint, method for silver, 827
counterpoise, 2462 (Fig. 401)
MischmetalI, 246
liquids, 2467 (Fig. 406)
"Miscellaneous analyses, of various inorfilling capillaries, 2468 (Fig. 407)
ganic and organic substances, 21342187
notes, 2463
Microscope, 2105 (Fig. 292), 2548 (Fig.
Mixed acid, analysis, 2233
451)
_
Mixed fuels, heat of combustion, 1187
. lenses, choice of; 2549
Modified soda, see Soda ash
ocular micrometer, 2105 (Fig. 293)
Mohr's alkalimeter, 238
method for chloride, 272
Microscopical examination, of water, 2099
organisms, contl;ol of, 2111
Moisture, see 'also Water.
cal.culation, moist to dry basis, 1345 .
identification of, 2109
determimnation, 1338 ff.
illustrations of, 2107, 2108
odors of, 2101, 2102
in air, 2403
.
in aluminum ores, 19
sampling for, 2102
in "arsenic," commercial, 109
Sedgwick-Rafter method, 210'2
Microscopy, chemical, 2434 ff.
in barytes, 1867
alloys, non-ferrous, analysis, 2445
in bauxite, 23
iden tity tests
in blanc fixe, 136, 1837
in brimstone, 933
anions, 2445,' 2446
in coal, 1636-_
cations, 2443
in coke, 1636
"mixtures of substances, 2438
in explosives, see "Explosives
physical properties, determination,
in gases, 2408
2453
melting points, 2453
in oils, see Oils, water and sediment
in paints and pigments, 1869, 1872,
particle size, 2454
1873, 1875, 1876, 1877
qualitative analysis, 2440
reagents, 2442
in paper, 1889
in paper pulp, 1889
methods of applying, 2441
substances of unknown composiin potash, 2169
_tion, 2446
in rubber, 1969
quantitative analysis
in soap, 2028
in sodium ferrocyani(le, 666
area, measure of, 2452
counting methods, 2449, 2450
fl"goride, 4~:(
..nitrate, 644
rri teria, 2447
estimation, 2449
silicate, 805
single substances, identification, 2434
in starch, 2157

~UBJECT

Moisture, in sulfur, 933, 940

in trisodium phosphate, 704
in zinc pulp, 1056
see also under special headings, as
impurities in various substances,
etc.
Mold, for cones, 1647
Molecular weights, determination, micro,
2533, 2534
of inorganic compounds, tables, 1161,
1164
1.folybdate method,. for lead
gravimetric, 506
volumetric, 511
Molybdenite, 597
Molybdenum, detection
general procedure, 585
.
tests with thiosulfate, sulfur dioxide,
disodium phosphate, sulfuric acid,
585, 586
. estimation, general procedures
colorimetric, see estimation in steel
gravimetric
alphabenzoinoxime method, 591
lead. molybdate method, 589
, mercurous nitrate precipitation,
590
molybdenum oxide method, 590
molybdenum sulfide method, 592
silver nitrate method, 590
volumetric
ceric sulfate method, 596
mercury reduction, 596
iodometric method, 593
permanganate method, 51t4
zinc reduction, 594
estimation, special procedures
. in ferromolybdenum, 1507
in molybdenite, 597
in molybdic-vanadic acid mixtures,
596
in ores wulfenite, molybdenite, 597
in steel
colorimetric, 608, 609; 1464
Electrometallurgical Co. method,
610-613
gravimetric, 608
alphabenzoinoxime
separation,
1461
sulfide separation, 1462
rapid. methods, 1481
in stellite, 328, 1411
in tnngsten ores, 1007
industrial importance of methods, 586
occurrence, 585
ores, copper in, 601, 603
phosphorus in, 604
.
preparation and soluqon of the sample,
587
ores, 587
iron and steel, 587

INDEX

63

Molybdenum, separation
ether extraction method, 589
from alkalies, alkaline earths (124),
iron and zinc .groups, bismuth,
copper, cadmium, lead, arsenic,
titanium,
vanadium,
tungsten,
phosphoric acid, 588
fro m i ron, alphabenzoinoxime
method, 587
from tungsten; phosphoric acid, 588
from tungsten, 593
volatilization method, 1004
solubilities, 586
Monazite, decomposition, 247
microanalysis, 2545
occurrence, 246
Monel metal, analysis, A. S. T. M., 1378
Morphine, detection, 1960
test for titanium, 976
~Iuriatic acid, analysis of, 2203
N

Narcotine, 1960
Naphthalene, in gas, 2396
Naphthol yellow, test for potassium, 862
,a-Naphthylamine test, for osmillm, 749
Narcotine, 1960
National Lead Co. Methods
for antimony, .in white metals, 83
arsenic in white metals, 112
tin in white metals, 973
illmenite, analysis, 495-499
lead ores, 523
see also non-ferrous alloys.
Natural gas, see Gases
Neodymium, 245
Neon, 2348, properties, 2380
Nephelometric method, for silver, 831
832
Nessler metho_d, ammonia, 639
nitrogen in steel, 656
water, 2077
reagent, 1197
test for ammonia, 630
tubes, 292 (Fig. 38)
Neumann lines, metallography, 2565
~eutralization, see Acidimetry and alkalimetry;
Conductometric methods;
Potentiometric methods; Soda ash;
Acids, Alkalies; Oils; Soap, etc.
number, petroleum products, 1818
New Jersey Zinc Co. (of Pa.) Standard
methods for
barium ores, 129
barite, ]29
witherite, 129
zinc, chapter, 1054 if.
Nichrome, analysis, A. S. T. M., 1405

64

SUBJECT INDEX

Nickel, detection
alphabenzildioxime, 615
dimethylglyoxime, 614
general procedure, 614
hydrogen sulfide test, 614
estimation, general p'rocedures
gravimetric
alphabenzildioxime, me th 0 d,
617
dimethylglyoxime method, 619
micro, 2521
electrolysis, 620
micro, 2521
volumetric
method of Parr and Lindgren,
622
potassium cyanide method,
622
estimation, special procedures
in aluminum and its alloys, 43
in "arsenic," 110
in bismnth bronze, 1377
in bronze, A. S. T. M., 1365
in co baIt and co baIt oxide, 621
in copper, 385
in German silver, 1374
in iron and steel, 495, 1450
cyanide method, 624, 1452
in magnesium alloys, 548, 549
in metallic lead (pig), 1419
in metallic nickel, 621, 627
in monel metal, 1378
in nichrome, 1405
in plating solutions, 621
in silver solder, 1405
in tin-base metal, 1390
impurities, metallic nickel, International Nickel Co. methods, C, Si, S,
eu, Mn, Fe, AI, Co, 624-628
industrial application of methods, 615
occurrence, 614
preparation and solution,of the sample,
general procedure, 615
metal and alloys, 616 I
separation
from alkalies, alkaline earths, ]lYdrogen sulfide group, 616
from aluminum; 617
aluminum, chromium, cobalt, irOll.,
manganese, zinc, 616
chromium, iron, 617
in the analysis of nickel, 626, 627
solubilities, 615, 625
Nickel plating solution, nickel, in, 621
Nickel uranyl acetate test, for sodium,
875
Nicotine, 1955
Niopium, see Columbium
Niter slag, selenium and tellurium in, 790
Niton, 2379_ See Radon

Nitrate; of soda, analysis of (moisture,

insolubility, sodium ,sulfate, iron and
aluminum oxides, lime magnesia, so
dium chloride, arbon dioxide), 644,
645
standard solution of, 64i
Nitrates, ferrous sulfate method for, 644
2216
iodide method, 647
apparatus, Gooch and Gruener,
648 (Fig. 71)
Nitrates in
alcohol, 2141
black powder, 1661
explosives, 1666, 1677
hydrochloric acid, 2204
ni trites, 655
water, 2076
Nitrato pentammine cobaltic reagent, for
phosphate, 696
Nitric acid,
arsenic in, 106
complete analysis, 2211
conyersion efficiency, ammonia to,
659
determination
<"
by the Devarda method, 640-644
ferrous sulfate method, 644, 221.4,
2216
free, ~53
nitron method, 635
in arsenic acid, FeSO., method, 2218
in hydrochloric acid, 2204
in mixed acid, 221.7
in oleum
ferrous sulfate method, 2235
nitrometer ,method, 652
in phosphori!l' acid, 2218.
in sulfuric acid, "2216, 2232
specific gravity, tables, 1145, 1146,
1147, 1148, 2220-24
standard solution, ferrous sulfate
method, 1207
see :also Acids; Reagents
Nitric oxide, 2348. __ "
determination, 2400 .... ~
properties, 2373
Ni trite (see also Nitrous acid)
detection, 631
estimation, general
gravimetric, Busvold '8 method, 653
volumetric
.
eerie sulfate method, 655
permanganate method, 653, 2219
nitrate in, 655
test for phenol, 1950
see Water
Nitro benzene,rill oils, ;{805
p-Nitrobe!lZilneazoresorcinol, reagent for
Mg,529
-"
'
Nitrocellulose, in explosives, 1669, 1678
r'

SUBJEOT INDEX

65

Nitrogen, estimation,
Nitrochlorhydrin, 1672
in paper, 1905
Nitrogen,
detection
in potassium nitrate 041
in rubber, 1982
'~9
combined as ammonia, Nessler's test,
630
in soil extracts, 646 app
(Fig. 61.l)
,
aratus
ammonium salts, 630
nitric acid, copper test, 631
.
available nitrates, 634
diphenylamine test, 631
In steel, 656
ferr:ms sulfate test, 630
free, See also Gas analysis
phenolsulfonic acid test 631
in air, 629
nitrous acid, acetic acid 'test, 631
in illu~inating gas, 2358.
permanganate test, 631
propertIes, 2348
organic nitrogen, 630
importance of methods 631
free, in gases, 630
'
occurrence, 629
estimation, all forms, 629
preparl_ltion of t,he sample,
('oJnbined, and free, volumetric estiaCId, and mlXed acids 634
mation, 6'36, 637
ammonium salts and ~ixtures 63{
combined, Il,; ammonia, gravimetric
organic substances, 632
'
method, 636
soils, 634
as nitrate
separations,
colorimetric estimation, 640
ammonia, 635
'Devarda method, 634
nitric acid, isolation of, 635
app. 642 (Fig. 67)
removal of impurities, 635
reagents, 641
solubilities, 632
iodometric method, 647
Nitrogen dioxide, 2361, 2400
Gooch-Gruener apparatus,
Nitrogen tetroxide, 2346
648 (Fig. 73)
Nitroglycerin, determination, 1665
water, nitrate in, 2052
dynamite, analysis of, 1662
as nitric acid, gravimetric method,
Nitroglycol, 1673
Nitrometer, 651 (Fig. 72)
635
DuPont, 651
volumetric
Devard:t method, 640-6,14
gas analysis, 2418 (Fig. 388)
Ferrous sulfate method, 644
micro, 2485 (Fig. 423); 2486
Standardizing, 650
See Acidimetry -alkalimetry
'as nitrite or nitrate,
. Nitron acetate, method for rhenium, 771
nitrometer method, 649
Nitron method, for nitric acid, 635, 639
reagent, 1216
DuPont nitrometer, 650,
651 (Figs. 72, 73)
Nitropolyglycerin, in explosi"es, 1672
Nitrositc method, for rubber, 1975
in oleum, 652
Nitroso-betanaphthol, detection of cobalt,
as organic nitrogen,
306
Dumas method, 634
precipitation of cobalt, 314 .
micro, 2485
reagent, 1216
semimicro, 635
separation, cobalt from nickel, 311
Kjeldahl method, 632
micro, 2492
Nitrostarch, in explosives, 1678
Meulen, method, 635
test for, 1675
estimation, special procedures
Nitrosugars in explosives, 1678
in ammonium salts, 634
Nitrotoluenes, in dynamite, 1670
in coal, 1652
Nitrous acid, and lower oxides, in nitric
in cyanamide, 663
acid, 2212
in oleum and mixed acids, 2234
in mixed acid, 634
in nitrates, 634
decomposition of iodides with, 454
in nitrites, 634
permanganate method for, 653, 2219
in nitrocellulose, 1690
removal from nitric acid, 635
in organic matter,
Nitrous anhydride, 2346
nitrates absent, 632
Nitrous oxide, in gas, 2348, 2400
properties, 2378
apparatus, !.i33 (Fig.
(6)
Normal solutions. See Reagents
catalysts, 633
Northwest Magnesite Co., Methods, 535
nitrates present, 633
Norton Company, standard methods for
in pain~ pIgments, 1873, 1876
alundum, 32-37

66

SUBJECT INDl~X ,

Oils f::tts edible

,_ j:mtter, !1785
fat, 1786
preservatives, 1786
1ard, 1787
o
fatty acids, \from oils, 1787
Occurrences, see Element in question
floor oils, n~trobenzene in, 1805
OLlor, organisms in water, 2100
hardilned oils, 1787
test of water, 2048
miscilllaneous oils and lubricants, 178&
Ohm, definition, 1175
painf oils, see under Paints
Oil, in cok~-oven gas, 2154
Chinese wood oil, tung oil, see
in cyanamid, 663
Paints
in explosives, 1665
petrdleum prod\lcts (principally A. R.
in grease, 1828
'1'. M. methods)
in inks, 2174
tmrning oils, 1702, 1816
in pigments, 1877
acidity, detection, 1716
,in ~ulfur, crude, 939
color, 1717
in water, 2078
distillation, A. S. T. M.,
Oils, Fats, Waxes, 1701 ff.
1705-1713
examination of
app., 1706 (Fig. 240);
animal and vegetable oils, 1764
1707 (Fig. 241)
general test, 1764
fire test, 1703
special tests
flash test, 1702
'd\;_\o'l,J\ "'d\'O.\o, \1'1'5
appara'tus, )_jlr6 tl"lg.
antifiuorescents, 1776
238)
bromine number, 1770
mineral salts in, 1717
eliadin test, 1701, 1765
specific gravity, 1704 (Fig.
iodine number, 1767
239); 1816
Hanus' method, 1767
sulfuric acid test, 1717
Hiibl '8 method, 1768
sulfur, 1712
Maumene test, 1701, 1766
lamp method, A. S. T. M.,
refractive index, 1764
1713, 1714 (Fig. 242)
refractometer, 1765
water in, 1717
(Fig. 256)
Dean and Stark methoil,
saponification value, 1771 .
1340

table, 1773
crank case oil, dilution test, A. S.
specific gravity, 1764
T. M., 1820
unsaponifiable oils, detec;tpparatus, 1821 (Fig. 263)
tion, 1773
crude petroleum
identification, 1774
distillation test, A. S. T. M.,
.
Valenta test, 1765
1807
drying, semi-drying, non-drying oils
physical tests, 1807
special tests for certain oils
app., 1808 (Fig. 259);
cottonseed oil, Bechi test, 1777
1809 (Fig. 260); 1810
Halphcll test, 1778
-.(l:ig:_ 261)
drying' on glass, 1784
fuel oils
free acid in, 1782
calorific power, 1764; A. S.
T. M., 1834
linseed oil, analysis, 1843
flash point, A. S. T. M., 1744
hexabromide test, 1779
Pensky Martens tester,
peanut on, Bellier's test, 1780
1745, 1746 (Figs. 250,
Renard's test, 178Q
2(1)
rapeseed oil, Bach's test, 1781
Tag tester, A. S. T. M.,
rosin oil, Lieberman-Storch test,
1752, 1753 (Fig. 252)
1781
sediment, 1761 (Fig. 254)
sesame oil, Baudoin, or Camoin
sulfur, A. S. T. M., 1750
test, 1781
water and sediment, centri.
spontaneous combustion test, 1782
fuge method, A. S.
:Mackey's apparatus, 1782
T: M., 1761
(Fig. 257)
tubes for, 1763 (Fig,
titer test, 1783
255)
Villavecchia test, 1782
Norton Company, boron carbide, 181-184
magnesia, 538
silicon carbide, 813-816

SUBJECT INDEX
Oils, grease, analysis, A. S. T. M. meth
ods, 1823
alkali, acid, 1828
ash, 1824
chart of analysis, 1826 (Fig.
264)
fat, 1828
fillers, etc., 1825
glycerin, 1827
gypsum, 1825
insoluble carbonates, 1829
petroleum oil, 1828
soap, 1825
water, 1829
waxes, 1830
heavy oils, 1833
liquid fuels
calorimeter for, 1802
gasoline, 1742
Bur. Mines test, 1813
corrosion test, 1742
doctor test, 1742
gum content, 1814, 1815
(Fig. 262)
specific gravity, 1812
Bulfur, A. S. T. M., 1743
tetraethyl lead in, 1804
lubricating oils, 1717, 1817
, carbon residue, A. S. T. M.
, test, 1737, 1738 ,(Fig~ 249)
caoutchouc, test for, 1736
cloud' and pour points, A. S.
T. M., 1727
app., 1729 (Fig. 246)
demnlsibility, 1740
. evaporation, 17126
fatty oils, 1736
fire test, A. S. 'l:. M., 1732
flash test, A. S. T. M., 1732,
1733 (Fig. 247);\ 1734
(Fig. 248)
friction test, 1741
gasoline test, 1739
gumming test, 1737
heat test, 1740
microscopical test, 1740
neutralization number, A. 8.
T. M., 1818
soap, detection, 1736
specific gra vi ty, 1726
viscosity
.
conversion eh'art, 1798
table, 1799
Engler viscosimeter, 1718'
(Fig. 243)
Saybolt viscosimeter, A. 8.
T. M., 1719
oil tubes, 1720
MacMichael's viscosimeter,
1724
'
table, constants, 1725

67

Oils, reagents, 1799

road oils, 1837
shale oils, 1838
tables
fatty acids, characteristics and
constants, 1794
mineral oils, properties, 1788
vegetable and animal oils, 17951796
viscosity conversion tables, 1799
waxes, 1797
turpentine, see also Paints, 1850
U. S. Gov't Specifications
fuel oils, greases, 1791
gasoline, burning oils, 1790
lubricating oils, 1792, 1793
waxes, 1788
paraffin, melt. point, A. S.
T. M., 1830, 1831 (Fig. 264)
table of, 1797
Oleum, acidity, lower oxides, etc. in, 2234
analysis of, 2233
nitric acid in, ferrous sulfate method,
2217
nitrometer method, 652
table of equivalents, 2243
Opium, 1961
Optical properties, microscopy, 2435
Orange pigments, see pigments
Ores. See under Preparation and solution of the sample under individual
elements.
Organic acids, see individual acids,
Acetic add, Formic acid, etc.
Organic matter
decomposition
by persulfate, 576
for determination of
alkalies, 864
antimony, 67
arsenic, 89, 91
barium, 119
boron, 165,'166
bromide, 189
mercury, 576
nitrogen, 630
phosphorus, 693
determina tion, arsenic in, 107, 115
carbon and hydrogen in, 222, 224
chlorine and halogens in
Carius method, 265
lime method, 266
sodium-alcohol method, 266
sodium-liquid ammonia method, 266
sodium peroxide method, 267
fluorine, 402
menQury, 581
nitrogen, 632, 634, 684
oxygen, 682
see also Microanalysis

68

SUBJECT INDEX ,

Organic matter
Oxine (8- Hydroxyquinoline) method for:
in sulfur, 933
aluminudJ, 12
in ,~ater, see Water, oxygen consumed
in alloys, 47
Orgamc reagents, for inorganic ions,
acidi titration, 48
1212
peryllimrl, 141
Organisms, in water
separation from Al 139
classitication, 2100
magl1esiu'm, 535
'
control of, 2111
separation of Mg, 531
chemicals required 2113 2114
separation, alkalies from Mg 866'
identification of, 2106 2'109
separation, Mo from Re 771'
Orifice meter, 2342
'
Oxya.nthraqnillone, see Hy:lroxyauthraOrsat appar~tus, 2349 (Fig. 364), See,
qumone
Gas analysIs
Oxygen, See also Gases.
Ortho nitro see nitro
detection, 669
-~ompounds, see under name of com
estimation
ponnd.
combined
Osmium,
available in higher oxides, 675
det.ecti~n, 748, 749
.
. ap.paratus, 675 (Fig. 77)
esilmatlOn, 749, 750
III bitumm. substs., 1554
occurrence, 748
in copper, 389
pl'eparation and solution of sample,
~n organic compounds, 682 ..
749
11l steel, 684 (Fig. 82)
properties, 748
free, in gases, 685, 2350, 2352, 2400
separation, by ilistillation 750
copper method, 670-674 2369
tetroxide, solution 1204 '
rocking pipette, 671 (Fig. 76)'
Osmiridium, soluti01{ of 731
672 (Fig. 76a)
,
Ovens, for drying, 1316: 1317 (Figs. 163, ~
explo~ioll 'with
hydrogen, 687,
164, 165, 166), coal and coke 1629
(FIg. 85); 688
(Fig. 227); 1636 (Fig. 230) ,
phosphorus method, 685, 2355
vacuum, 1317 (Fig. 167)
pyrogallate method, 686 (Fig. 84)
volatile matter, 1528 (Fig. 196)
free, in solution
Oxalic acid, determination, 2253
cyanide solution, 676' 678 (Fig.
method for persulfates 922
79); 680 (Figs. 80,' 81)
Oxidation methods, See' under Estima
saturation curve, 677 (Fig. 78)
tion, volumetric under the element or
'vater, 2079, 2369
compound in question; see also indio
occurrence, 669
vidual oxidizing ~gents, as Iodine,
propertie~~ 669, 2348, 2367
Permanganate, CerlC sulfate etc.
trace~ o~', in gases
Oxidation test, for methyl alcoilOl 2148
apparatus, 688 (Fig. 86)
'
Oxide method for:
~olorimetric estimation, 687
aluminum, 8 10
)xygen consumed, organic matter in
aIitimony, 72
miter, 2053
beryllium, 139
)xygcn cylinders, 2368
bismuth, 154
)zone, propertilQ 2348, 2369
chromi um, 286
in air, 2405
columbium, 347
iron, 467
p lead, 507
manganese, 559
"aints
and
Paint
vehicles,
analysis of,
molybdenum, 590
1839 ft.
strontinm, 902 .
pigments, classificatiOl~, 1854
tantalum, 347
bla<{k pigments, analysis of, 1876
thorium, 953
blue pigments, analysis of, 1872
tin, 959
Prussian blue, Chinese blue,
titanium, 981 '
Antwerp blue, 1873
tungsten, 1002
sublimed blue lead, 1858, 1874
uranium, 1022
_!!ltramarfne blue, 1872
vanadium, 1035
grtliln 'pigments, 1875
zinc, 1058
-red -and br01vn pigments, 1869
zirconium, 1099
iron oxides, 1872
Oxidized oils, see Oils

::mB.JEC'l' INDEX
l'aints and Paint vehicles, red lead and
orange mineral, 1869
vermillion, 1871
yellow and Allange pigments,
chrome yellow, American vermillion, 1874
basic lead chromate, 1874
white pigments, 1856
antimony- oxide, 1864
basic carbonate of lead, 1859
basic lead sulfate, 1856
calc. of comp., 1857
barytes and blanc fixe, 1867
composite white paint, 1868
gypsum, 1866
lithopone, 1862
silica, silex, China clay, asbestine, 1866
titanium pigments, 1863
titano", 986
whiting, Paris white, 1866
white lead, 1859
zinc lead and leaded zinc, 1861
zinc oxide, 1862
zirconium pigme11ts, 1862
results of analysis, 1839
vehicles,
composition of, 1840
oils, detection of, 1842
determination and analysis, 1842
linseed oil
acid number, 1844
ash; 1845
driers, 1845
drying time, 1846
foots, gen. test, 1843
iodine nnmber, 18415
loss on heating, 1843
reagents, 1846
saponification number, 1844
specific gravity, 1844
unsaponifiable matter, 1844
tung oil, (Chinese wood oil)
gelatin test, 1847
heat test
Browne, 1847, 1848 (Fig.
2(:i5)
W orstall, 1849
resina,tes, 1842
separation of components, 1840
thinners, detect. of, 1842
turpentine, ex.amination of, 1850
vo la tiles, 1841
water, 1841
Palladium,. _
detection, n1'; 722
~stimation, general
gravimetric, as metal, 723
dimethylglyoxime method, 724
micro, 2547
mercuric cyanide method, 724

69

Palladium,
estimation, special methods
in gold bar, 725
in refined gold, 727
in silver, 726
in slimes (copper), 726
occurrence, 721
preparation and solution of the sample, 722
properties, 721
separation from Pt, Ir, Ag, Au, etc.,
723, 857
solubilities, 721
PaUadous chloride, reagent, 1197
iodide, method for iodine, 452
sep. from Br and Cl, 451
Paper and Paper-making Materials.
Methods of the Technical Association of the Pulp and Paper Industries (TAPPI), 1878-1942
bisulfite liquor, analysis of, 1929
gravimetric methods, 1930
casein, moisture, fat, ash, 1931
cellulose
chlorination method, 1878; 1879
(Fig. 266)
paper, acidity, total, 1910
alpha cellulose, 1911
arsenic, 1919, 1920 (Fig. 273)
ash, 1905
coating, 1902
conditioning, 1900
copper number, 1913
iron, acid-soluble, 1915
mineral fillers, 1909
moiBtnre, 1904
llitrogeh. (proteinaceous), 1905
paraffin, 1901
resin, 1903
sampling, 1899
starch, 1906, 1907
sulfur, active, 1901
titanium pigments, 1925 (Fig. 274)
zinc pigments, 1922
paper extracts, pH of, 1916 (Fig. 272)
pulp
alpha cellulose, 1881
ash, 1896
chlorine consumption of, 1880
cuprammonum disperse visco8ity,
1884
apparatus, 1885 (Fig. 267,
268); 1886 (Fig. 269)
methoxy groups, 1891
apparatus, 1892 (Fig. 271)
moisture, 1889
apparatus (Fig. 270)
test for, 1893
pitch, 1883
sampling, 1893
solubility in 1 % NaOH, 1897

70

SUBJECT INDEX'

Papcr and Paper'making Materials.

Permangauate, arsenic, 1210
weighing, 1893
barium). 128
rosin, 1940
bismuth, 157
acid number, 1941
calcium, 211
ash, 1942
chromium, 288\
color grade, 1940
I.
copper, 375
ester number, 1941
ferricyanide, 6~3
insol., in toluene, 1942
ferrocyauide, 663
sampling, 1940
fluorine, 413
saponification, number, 1941
formic, 2246
un saponifiable, matter, 1941
gold, 437
salt cake, complete analysis, 1935
hydroxylamine, 647
(moisture, insol., acidity, chloride,
iodide (oxidized to IOl), 455
ferrous iron, ferric iron, alumiiron, 474
num, lime, magnesia, chromium
lead, 509
trioxide, total soda, sulfur trilead peroxide, 525
oxide)
mercurous salts, 582
molybdenum, 594
starch and starch products, 2156-2158
nitrite, 653
Paraffin, in paper, 1901
nitrogen dioxide, 2361
solid, in bituminous substances, 1555
nitrous oxides, 653
refractive index, 1556
oxalic acid, 2253
solution, 1197
oxygen (available),675
wax, meltingpoint determination, 1830
percarbonate, 243
Para-aminodimethyl aniline, test for
peroxide, 2181
halogens, 281
phosphorus, 699
Paris white, analysis, 1866
Particle size, microscopic determination,
staudard solution, 475
24M
1!tandardization, 1209
tannic acid, 2158
Parting, see Fire assay
Peanut oil, tests for, 1780
uraninm, 1027
Perborate, in soap, 2036
vanadium, 1037
method for cobalt, 318
water, oxygen consumption, 2053
Percarbonate, determination, 243
test of acetone, 2136
Percentage, calc. moist and dry basis,
Permissible explosives, analysis of, 1673
1345
qualitative tests, 1674
Perchlorate,
Peroxide, available oxygen in, 675
detection, 262
determination, 2180
determination, 275'
method for chromilfm, 2180
formaldehyde, 2,149
in explosives, 1677
in presence of chlorate an<i chloride,
lead, 507 :
titanium, 987
276
Persulfate, determinatio'n
method for potassium, 871
alkali method, 922
detection, 862
ferrous sulfate method, 922
separation of K, Rb, Os from Na and
oxalic acid method, 922~
Li,885
method for cerium, 253
Perchloric acid method
manganese, 572
fluorine determination (distillasee Iron and steel, 490
tion), 401
vanadium, 1048
,
fluorspar, 403
Pctroleum products, see uuder Oils, Fats,
silica, 802
Periodate, determination, 458
Waxes
...pH. See also Hygrogen-ion conc~ntrain presence of iodate, 458
'fcrtion.
'
method for manganese, 573
buffering, 2278 '
in steel, 1470
Permanent standards, see Colorimetric
colorimetric method
methods; pH, Determination; Water,
comparators for
analysis.
block, 2283
Hellige, 2295
!
Permanganate, see also Potassium permanganate
La Motte,}'~ulette fOrm, 2294
method for:
Taylor, -slide comparators, 2291,
2292, 2293
antimony, 76

SUBJECT INDEX
/PH, colorimetric method
determination
effect of proteins on, 2287
salts on, 2.2.5
temperature on, 2288 '
in solutions for bacterial examination, 2119
electrometric method, 2297
electrodes for
antimony, 2302
glass, 2303 .
hydrogen, 2300
quinhydrone, 2301
of paper extracts, 1916
working formulae, 2303, 2304
hydrogen-ion concentration, relation to
pH, 2277
indicators for, 2281
meaning of, 2274
oPhenanthroline-Ferrous sulfate indicator, 474, 1207
p-Phenetidine indicatol', 474
Phenol, detection, 1949
determination, 2253
Phenol phthalein indicator, preparation,
32, 176

Phenol red, method for bromine, 193

Phenolsulfonic acid, method for nitrates
(in water), 631
l'henyl arsonic acid, method for thorium,
952 ~
zirconium, 1102
reagent, 1216
Phe.nyl hydrazine, test for gold, 433
Pheriylthiohydantoic acid, method for cobalt, 311
reagent, 1216 ,
test for cobalt, 306
Ph~nyltrimethylammonium iodide, Cd re. agent, 1217
Phoroglucid method, for furfural; 2249
Phosphate
'arsenic in, 106
.
determination, in blanc fixe, 126
in mixtures, 2225
in soap, 2036
in watoo, 2061, 2090
effect on alkaline earth determinations,
119
methods, see ammonium magnesium
phosphate, etc.
Bee Pyrophosphate
trisodium, analysis of, 703-705
typic,al analyses, phosphate ores, table,
691~.
Phosphil' 2347, 23-66
test f~'r ~, 722
Phospho!~t;~bdate method, for phos
pho\;tr!l, 694
separation, Rb /ind Cs from K, 894
test for molybdate, 586

71

Phosphoric acid, and its salts, analysis,

2224. See Phosphorus, estimation.
comparison, ortho, meta and pyro
acids, 690
detection, 690
effect on aluminum determination, 7
separations, alkaline earths, 124
impurities in, arsenic, 106
nitric acid, 2218
specific gravity, Tables, 1149, 2227
Phosphoric acidammonium chI a rid e
method, copper compounds, 395
Phosphorous acid, detection, 689
precipitation of Se and Te, 776
Phosphorus
'comparison of acids, 690
detection, 689, 690 (in various states
of oxidation and combination)
estimation, general (see also phosphate; phosphorus pentoxide)
gravimetric,
ammonium magnesium phose
phate method, 696
phosphomolybdate method, 694
micro, 2499
pyrophosphate method, 695
other methods, 696
volumetric
alkali titration (phosphomolybdate), 697
micro, 2501
methylene blue method, 707, 708
(Fig. 87)
permanganate method, 699
estimation, special procedures
in alloy steel, 701
in bronze, A. S. T. M., 1363,
1366, 1368
in cast iron, 702
in cement, 1609
in coal ash, 1644
in copper alloys, 698
in ferromanganese, 1488
in ferromolybdenum, 1509
in ferro titanium, 702
in -ferrotungsten, 1012, 1500
in ferrovanadium, 1495
in iron, 492, 494
in molybdenite, 604
in steel, 697, 699, 700, 1435,
1438
in tungsten metal, 1012
in vanadium steels, 703
in wulfenite, 604
with vanadium present, zirconium method, 706
importance of methods, 690, 691
occurrence, 689
preliminary remarks, 691

72

SUBJECrl' INDEX

Phosphorus
preparation and solution of sample,
692-694 (ores, alloys, minerals, iron
and steel soluble phosphates, baking
powder)
separation, as ammonium phosphomlybdate, 69~
traces, see Volumetric estimation .
see Microanalysis, 2499 ff.
in poisoning cases, 1945
Phosphorus method, oxygen in gases, 685
reduction of osmium, 749
Phosphorus pent oxide (see also Moisture,
determinatiou; Water, determination)
in aluminum ores, 22
in alundum, 35
in bauxite, 25
Phosphotungstate test, for potassium,
862
Phosphuretted hydrogen, see Phosphine
Photographic developer, . spectrochemical
analysis, 1091
materials, for metallography, 2550
Photometry, definitions, 1174
Physiological effects, of gases, see Gases
Phytosterol, in oils, 1775, 1776
Picnometer, see P.ycnometer
Picric acid, examination of, 1682
test for potassium, 862
Picrolonic acid, reagent, 1217
Pig iron, sampling, 1424
see Iron
Pig lead, analysis, 1413 ff.
Pigments, in inks, 2175
in rubber, 2019
of paint, see Paint
with tatanic oxide, 998
Pipettes. See also Gases
aut61natic, for liquids, 563 (Fig. 60)
dividing, for liquids, 365 (Fig. 43).
for oxygen, special rocking type, 671,
672 (Figs. 76, 76a)
standard, calibration, 1224
Pitch. See Bituminous Substances
in pulp, 1883
Pitot tube, 2340 (Fig. 356)
Plants, iodine in, 460
Plaster of Paris, 1866
Platinum, attack by fluxes, 284
detection, 712, 713
estimation, gravimetric
as metal, 717-719
as ammonium salt, 7] 9
electrolytic method, 7] 9
micro, 2547
preparation and solution of the sample,
714
scrap, 714
small amounts, in presence of
Fe, MgO, etc., 715

Platinum, _pr_operties , 712

separa tion, .from j
gold, 715, 728
\
iriilium, J16
osmium" 716
palladium, 716 ,.
I
rhodium, 716
ruthenium, 716, 728
solubility, 712, 714
Platinum group, 710
estimation in gold alloys, 754
rarer members (Ir, Ru, Rh, Os)
730 ff.
separation, 710
solution, ores, etc., -710
Table, of reactions, 753
Tabular outline, quantitative separations, 729
Platinum, Palladium
fire assay, 857
in copper anoile slimes, 726
in gold, refined, 727
in silver, refined, 726
Poisons, clllssification, 1944
detection', 1943
estimation, 1943
general procedure, 1944 _
individual poisons, detection, etc.
acetanilide, 1953
atropine, 1958
brucine, 1957
caffeine, 1959
carbolic acid, 1949, 1950
carbon monoxide, in air, 1963
in blood, 1964
chloral hydrate, 1951
chloroform, 1948
cocaine, 1958 /
codeine, ;1958, 1959
ethyl alcohol, 1948, 1949
hydrocyanic acid, 1945
estimation, 1947
in presence of Prussian blue,
1947
. metallic poisons, -1962..
arsenic, 1963
-mercury, 2423
methyl alcohol (methanol; wooil
alcohol), 1949
morphine, 1960
narcotine, 1960
nicotine, 1959
opium, 1961 ,
phosphorus (yellow), 1945
ptomaines, 1962
quinine, 1956
jltrychnine, 1956
'thallium, in r,-aJ; _poisOI){ 945
veronal, 19;l-reagents, 1964
.
Plumbago, see Graphite

SUBJECT INDEX
Polarized electrodes, 2320
Polarograph, ref., 2546
Polymerization, turpentine analysis, 1851
Polysulfide, in sodium sulfide, 2183
Portland cement, see Cement.
Potash, analysis of, 2169
Potassium (See also .Alkalies, estimation)
(\etection, bpectrum, etc., 861, 862 '
estimation, general methods'
gravimetric, chloroplatinate
method, 869
modified, 870
Lindo-Gladding method, 871
other methods, 874
perchlorate method, 871
weighing as the chloride or
sulfate, 869
estimation, special procedures
in alundum, 36
sJlOrt method for silicates, 872
occurrence! 861
preparation of the sample, 863, 864
separations, from
from aluminum, chromium, iron,
alkaline earths, phosphoric and
sulfuric acids, 866
ammonium salts, 868
barium, calcium, strontium, 865
boric aciJ, 867
heavy metals, 864
magnesium
ammonium phosphate method,
867
barium hydroxide method, 867
8-hydroxyquinoline method, 866
mercuric oxide method, 867
., rubidium and caesium, 868
sodium, large amount, 868
sodium and lithium, 868 .
sulfates, 865
uranium, phosphoric acid, 866
Potassium acid phthalate, 2197
Potassium acid sulfate (bisulfate), see
Fusion. Action on Pd, 722
Potassium bromate, 0.1 N, 1207
method for antimony, 74
Potassium carbonate and hydroxide, 2253,
2~66

Potassium carbonate, analysis of, 2169

fusion with, see Fusion; Decomposition of Sample.
,Potassium chloride, primary stallllard,
1202
test for iridium, 730
platinum, 712
Potas8ium chromate, test for lead, 500
Potassium cyanide, method for
cobalt, 319
copper, 3U
nickel, 622
test for Se and Te, 777

73

Potassium dichromate, method for barium, 128

iron, 471
solution, 0.1 N, 471
standardization, 1208
Potassium ethyl xanthate, method for
copper, 376
Potassium ferrieyanide; test for ferrous
iron, 463, 1197
fe~rocyanide" reagent, 1202
test for ferric iron, 463
uranium, 1017
zinc determination, 1054
fluoride, alumina in aluminum salts, 16
reagent for, 16, 1197
Potassium hydroxide, analysis of, 2257
test for iron, "463
osmium, 748
rhodium, 742
ruthenium, 737
Potassium iodate
method for arsenic, 99
copper, 372, 374, 375
iodide, 455
mercury, 582
standard solution, 1209
Potassium iodide
method for antimony, 75
barium, 128
bismuth, 159
chromium, 288
copper, 368
gold, 437
iron, 482 ,
lead, 512
mercury, 580
selenium, 784
reduction of ferric salts, 474
Potassium iodide-starch test
for iodine, 448
for nitrocellulose, 1692, 1694
test for palladium, 721
platinum, 712
ruthenium, 737
selenium and telluriUm, 776
Potassium nitrate, standard solution,
1200
nitrogen in, 2419
Potassium nitrite method, for cobalt, \l12
separation from Ni, 311
test for cobalt, 306
osmium, 748
palladium, 721
rhodium, 742
Potassium periodate, method for Mn, 573
Potassium permanganate. See permanganate methods.
solution, 0.1 N, 475
standardization, 478
Bureau of Standards procedure,
1209

74

SUBJECT INDEX

Potassium permanganate,
calculations, 478
other. procedures, 1210
test for nitrite, 631
Potassium pyrogallate, ~ee Oxygen
reagent, 1197
Potassium thiocyanate, test for cobalt,
306
.
iron, 737
ruthenium, 737
,
/
standard solution, 1203
VPotentiometric determinations
aluminum in magnesium alloys, 540
calcium in magnesium alloys, 550
see Potentiometric titrations
see pH, Electrometric method
V'methods, 2305

"-'i1rinciple, 2297 (Fig. 321)

titrations, 2308 ff.
apparatus, 2309 (Fig. 326)
electrodes for, 2309~2311
micro methods, 2322
neutralizations, 2309
graphs, 2312 '(Fig. 329)
in non-aqueous solutions, 2312
oxidation-reduction reactions
chromium in alloys, 2315
vanadiulll in alloys, 2315
,precipitations, complex reactions,
2312
simplified methods, 2318
bimetallic electrodes, 2320
differential method, 2318
polarized systems, 2320
reference method, 2318
vacuum-tube voltmeters, 2321
Pour point, oils, 1727
Powder, see Explosives
Praeseodymium,' 245
Precipitation and complex-forlPation titrations, 2312
see Conductometric titrations; Potentiometric titrations; Chloride, Bromide, Iodide, Silver, Mercury, etc.
Preparation and solution of the, sample,
see the Substance in question
Preserva ti v,es, in butter, 1786
Pressure-solubility curves, Oxygen in
water, 677 (Fig. 78)
Primary standard substances, 1200, 1201,
see Standard Subs~,nces, also ultimate
standards
Primers, table of, 1688, 1689
Priming, of water, 2086
Printing ink, 2174
Producer and fuel- gases, see Gases
Proof gold, preparation of, 440
Properties of compounds, Vol. I, Part II,
tables, pp. 1161-1164; of elements, see
Chapter on element in question
Propylarsonic acid, 1216

Protein, effect 0;11 pH determinations,

2287
in rubber, 1973
~russian blue test) for HCN, 1946
Psychrometer, sling type, 2403
Ptomaines, 1962
Pulp, see Paper; \see Zinc.
Purpurin test, for fluoride, 423
Pycnometer, see also Specific gravity.
for bitumin.' substances, 1515 (Fig.
188)
for specific gravity, 2137 (Fig. 297);
2148 (Fig. 298)
Pyridine, in ammonium nitrate, 655
apparatus, 655 (Fig. 74)
Pyrite ores, arsenic in, 88, 106
Pyroantimonate, test for sodium, 876
Pyrogallate method, for oxygen, 686
Pyrophoric alloys, 246
Pyrophosphate method
for cobalt, 330
magnesium, 532
manganese, 560
thorium, 949
zinc, 1058
Pyrophosphoric ac,d, see Phosphorus;
Phosphoric acid.

Q
Qualitative analysis, tables, 1107 ff. See
also Microscopy; Spectrochemical
analysis.
acids, 1112
tables of reactions, 1128-1135
bases, reactions of, 1114-1127
bead tests, 1109
blow-pipe test!!" 1107-1108
flame tests, 1108
groups of metals, 1111
microsc<;lpical (chemical microscopy),
2440 ff.
alloys, 2445
manipulations, 2439, 2441
reagents, 2442 .... _
tests, 2443, 2445
Quinaldinic acid, reagent, 12Ji7
Quinalizarine, reagent, 1217
test for Be, 138
'
Quinhydrone,electrode, 2301
limitations, 2302
Quinine, detectio,n, 1956
Qui~oline, 8-Hyqroxy, see Oxine
7-lOdo-8-hydroxy, or ferron, 487
I
R
Radium,
detection, 758 .
estimation _ -' ,__
alpha ray method, 759
instrument, 759 (Fig. 88)

SUBJECT INDEX
Radium,
emanation (radQn) method, 761
ltPparatus, 760 (Fig. 89)
bisulfa te fusion, 761
carbonate fusion, 763
boat f.or, 764 (Fig. 91)
gamma ray method, 766
apparatus, 766 (Fig. 93)
Radon, 761
Raies ultimes, 2601, 2609
Table of, 2602, 2603
Rapeseed oil, 1781
Rapid methods, soo Cement; Copper, etc.
Rare earths
atomic weight, average, 259
colors, of solutions, 260
detection, 258
arc and spark spectra, 257, 258
estimation, gravimetric, 247
cerium group, 250
composition of groups, 256
yttrium group, 250
occurrence, 246
preparation of the sample, 247
propert~es, 245
separatlOn
from other elements, 247-250
phosphoric acid, 248
scandium and thorium, 249
spectra, absorption, table, 257
Rarer elements, allied to platinum, 730 ff.
Rasorite, 164
Rat poison, thallium in, 945
Reaction coefficient, water analysis, 2086
Reactions, qualitative, 1114-1135
Reagents. See also Chapter on, Vol. I,
pp. 1193-1218
acetaldehyde, 2139
acetic acid, 1193
acetic anhydride, 2150
'acid mixture, for iron titration, Knop 's,
473
Reinhardt-Zimmermann, 476
for silica determination, 29
alcohol, 1193
for oil analysis, 1771
free of aldehyde, 2139
-methyl chloride, 29
alizarine S, aluminum reagent, 2
al uminon, 1213,
aluminous materials, analysis, 29-32
amino-nitrosophenylhydroxylamine, for
iron, 469
ammonium acetate, lead extraction,
512, 119~
ammonium chloride, standard, water
analysis, 1199
wash solution, 32
molybdate, for lead, 511, 1201
for phosphorus, 32, 494, 1194
in water, 2090

75

Reagents,
.
nitrate, wash solutIon, 32
per sulfate, for Mn, 563
phosphate, 1194
sulfide, 1194
thiocyanate, ~71
antimonyl chlonde, 119~
antimony, traces, Gutzelt, 79-80
arsenic Gutzeit, 102-104
standard solution, 1199
arsenite, 0.1 N, 453
arsenous acid, for manganese, 31
iodometry, 1203
various oxidants, 1204
barium chloride, 1194
hydroxide, 1194
benzidine hydrochloride, for sulfates,
1213
benzoic acid, standard, 2197
a-benzoinoxime, for molybdenum, 591
reagent, 1213
a-benzyldioxime, for Ni, 618
bismuth, standard solution, 158, 1199
borate, standard solution of, 185
brilliant-green lactose peptone bile,
2122
Brom cresol purple, 2192
bromine-potassium bromide, sulfiile
oxidation, 1194
bromine, 1194, 2253
for gas analysis, 2422
saturated-HCI, 74
buffer reagents, 2074
cadmium, standard solution, spectrographic, 1092
cadmium chloride, sulfur reagent, 912
calcium acetate, for fluorine, 413
calcium chloride, hardness of water,
2082
calcium hydroxide (lime water), 1195
carbon dioxide, micro generator, 2489
eerie sulfate, standard solution, 1204
chlorine water, for bromine determination, 192
citric acid, for calcium determination,
209
cinchonine, for tungsten, 1011, 1195
-potassium iodide, for Bi, 158, 1195
color standards, permanent
liitrite in water, 2052
copper, standard solution, 1199
copper sulfate, for HCN determina
tion, 660
cuprammonium solution, for paper
pulp, 1886
cupric potassium chloride, C in steel,
220
cuprous chloride, acid for CO in gas,
1195,2421
ammoniacal for CO in gas, 1195,
2421

76

SUBJEc'r INDEX

Reagents,
sulfate, 2422
cur cumin for Be, 144
Devarda alloy; nitrate reduction, 641,
1196
dichlorfluorescein, adsorption indicator, 273
dichromate, potassium, Nj5 and. Nj10
for iron, 471
diethyldithiocarbamate, for copper, 376
dimereaptotliiodiazol, 1214
p-dimethy laminoazo benzy lidener hodanine, for Ag, 1214
dimethylglyoiime, for nickel, 619
diphenylamine indicator solution, 32,
291
for lead, 514, 1205
zinc, 1065
diphenylamine sulfonic acid, 1206;
for nitrate, 1214
diphenylbenzidine, indicator, 543, 1206
diphenylcarbazide, for chromium, 283,
290, 1215 _
dipicrylamine, for potassium, 1215
dithizone, for lead, etc_, 522, 1215
Endo's medium, 2121
eosin, indicator, 192, 1201
-methylene blue agar, 2121
erioglaucine, indi(,ator, 1206
eriogreen, indicator, 1206
etching reagents, for steel, 2552
Fehling's solution, 1206
1!'erric alum, Volhard titration, 271,
1201
ferric chloride, for tin titration, 967
nitrate,
for Volhard titration, 825
zinc analysis, 1196
ferricyanide, potas:ium, 472
ferrocyanide, potassium, 1060
ferron, 487
ferrous ammonium sulfate, for glycerol, 2153
for steel analysis, 295, 304, 1207
ferrous sulfate, for chromium, 291
for nitric acid, 1206,' 1225
for persulfate, 922 .
fire assay, 836-837
fluorescein, adsorption ~ndicator, 273,
1201
fluoride, standard for fluorine determination, 411
Frolihde's reagent, 1965_
fuch'3in-sulfite, 2139
fusion mixture, aluminous materials,
29
gas analysis, list of reagents, 24212424
glucose, for bodc acid titration, 180
glycerol, for boric' acid. determination,
171

Reagents,
IIessa's agat, 2123
hydrochloric acid, free of arsenic, 104,
.
1196
0.1 N, ~talldard, 176, 2195 :
1-2-5-8-hydroxyanthraquinone, for Be,
143
I

8-hydroxyquinoIine, see oxine

hydrogen peroxide, 1196
indicators, acid-alkali, 2192
mixed, 2192
for pH, 2073
oxidation-reduction, see Reagellh,
1203 ff.
precipitation, see Adsorption inJicators
indigo carmin, 2158
indigotindisulfonate, indicator, 679
iodate see potassium iodate
iodide solution, oxygen in water, 2079
iodine, 0.1 N reagent, 1207
'fI;Jagent for lime, 213
for sulfur, 912
for tin, 965
ioaine mono chloride, 0.005 M solution,
1209
iodo-potassium iodide, 1965
/' iron, standard solntion, spectrographic
work, 1092
,
spelter, 1199
stannous chloride titration, 485
traces, 486
lead, standard solution, 518, 1200
for spectrographic work, 1092
lead acetate, cotton,' Gutzeit method,
104, 1196
lead acid, lead analysis, 504
lead subacet;ite, 2153
magenta, 198
magnesia mixture, for phosphorus, 695,
119p
magnesium uranyl acetate, for sodium,

880'
manganous sulfate, oxygln in water,
2079
___ .
mannitol, boric acid; - determination,
176
Marme's reagent, 1965 .
Mayer's reagent, 1965
mercuric ammonium I thiocyana te, for
zinc, 45
mercuric <;hloride paper, for antimony,

'79

for arsenic, 103, 1197

see also microanalysis
mercury, 2422
methyl orange, 2192
red, 32, .2192
!
yello\!, s.2, 2192
violet, paper, 1694
microscopy, list of reagents, 2442

SUBJECT INDEX
Reagents,
Millon's reagent', 1964
molybdenum solution, colorimetric determination, 609, 612, 1009
llaphthylamine reagent, nitrites in
water, 2052
Nessler's solution, ammonia detetmination, 630, 657, 1197, 2049
nitrate, standard solution, 641
for ferrous sulfate method,
1207
nitric acid, 0_1 N solution, 698
p-nitrobenzeneazoresorcinol, for Mg,
529
nit ron, 639, 1216
a-nitroso-,e-naphthol, 1216
nutrient agar, 2121
broth, 2120
gelatin, 2120
oils, fats waxes, reagents for, 17991801
for paint vehicle, 1846
osmium tetroxide, 1204
oxalic acid, 212, 922
oxine (8-hydroxyquinoline), 1216
palladous chloride, 1197, 2423
paraffin, 1197
paranitrophenol indicator, 181
permanganate, see Potassium permanganate
,
peroxide solution, for Ti det'n_, 1197
o-phenanthroline-fcrrous sulfate, oxid.red. indicator, 1207
phenolphthalein, indicator, 32, 176,
2192
phenolsulfonic acid, nitrates in water,
2076
phenylarsonic acid, 1216
phenylthiohydantoic acid, 1216
phenyltrimethylammonium iodide, for
Cd, 1217
phosphate, standard solution, water
analysis, 2090
9-phosphomolybdic acid, for Rb and
Cs, 894
picrolonic acid, 1217
potassium antimonyl tirtrate, 80
bromate, 0.1 N, 74, 1207.
rhromatc, indicator, 272, 2094
cyanide solntion, for copper, 374
for nickel, 623
dichromate, 0.05 N, 29; 0.2 and 0.1
N, 471, 1208
ethyl xanthate, for cqpper, 377
ferricyanide, indicator, 472, 1197
ferrocyanide, for copper, 378
for zinc, 1060, 1065, 1202
flnoride, for aluminum salts, 16,
1197
hydroxide, for carbon det'n., 1197
iodate, general purposes, 1209

77

Reagents,
for copper, 373
iodide, for bismuth, 1197
mercury, 580
nitra1le, standard, 1200
periodate, 573
permanganate, 0.018 N, 31; 0.1 N,
31, 85, 295, 304
for lead det 'n., 526
standardization, U. S. Bur. Stds.
meth., 1209
water, oxygen consumed, 2050
thiocyanate, 0.1 N, 271, 1203,
for iodate method, 373
proof gold, preparation, 440
propylarsonic acid, 1216
pyrogallate, for oxygen, 1197, 2423
qUinaldinic acid, 1217
,quinalizarine, 1217
reducing agents, for sulfates, 912
rubber analysis, reagents, 2001
rubeanic acid, 1217
Russell's media, 2123
salicylaldoxime, 1218
salicylic acid, for iron, 487
Schiff's fuchsin-bisulfite, 1198, 2139,
2148
selenious-sulfuric acid, 1965
silver carbonate, 2153
silver nitrate, 0.1 N, 271, 1202
for nickel, 622; silver, 825; water,
2094
soap, standard solution, water analysis, 2082
sodium arsenite, see arsenious a'eid
bismuthate, 1198
bismuth nitrite, Rb and Cs, 896
carbonate, pure, 2193
chloride, primary standard, 1203
hydrosulfite, for dissolved oxyge.n,
67!l
hydroxide, 0.1 N, 176, 181; 0.2 N,
181; 0.5 N, 176
plumbite, 1742
sulfide, 936, 1198
thiosulfate, 213, see Thiosulfate
sofnol red indicator, ] 76
stains, standard, plate facing p. ] 07
(arsenic)
for antimony, 80, 81
for arsenic, 102, 103
for sulfur, 936
stannous chloride, 471, 485, ] 212
starch solution, 453, 1198, 2079
for sulfur determination, 912 _
steel analysis, reagents, see Chapter
011; also individual
elements, C, Mn, P, S, Si, etc.
sugar broths, 2120
sulfanilic acid, for nitrites in water,
2052

78

SUBJECT INDEX

Reagents,
sulfuric acid, standard solution, 2194
free ammonia determination, 641
tartrate solution, lead determination,
1198
tetrahydroxyanthraquinone indicator,
for sulfates, 919
thiocyanate, ammonium or potassium,
0.1 N, 825
thioglycollic acid, for iron, 487
thionalide, reagent, 1218
thiosulfate, sodium, 0.1 N solution,
452,1211
for copper, 308, 1211
for water, 0.01 N, 2079
thymol blue, 2192
thymol, for titanium determin3;tion,
994, 1199
_
thorium nitrate, for fluoride, 407
tin, standard solution, 965
titanium, standard solution, for fluorine, 411
for titanium, 30, 989
titan yellow G, for Mg, 529
triphenyltin chloride, 1218
turmeric indicator, 185
uranyl zinc acetate, 32
vanadium sulfate, solution, 31
Wagner's sol 'n, for Ca, 209 .
Zimmermann-Reinhardt solution, 476
zinc uranyl acetate, for sodium, 879
zirconium-alizarin indicator, for fluorine, 407
zirconium-purpurin, reagent for fluorine, 423
zinc, amalgamated, 1199
Red lead, 1869
Jead in, 1869
Jead peroxide in, 524, 525
method for Mn, 573
organic color in, 1869
true red lead, 1870
Red pigments, 1869
Reducing agents, for chlorate, 274
for chromate, 282, 286
for vanadium, 1031
Reduction of iron compounds, 469-471
ni tra tes, 644
Reduct():r, Jones', 476 (Fig. 56); 595
(Figs. 64, 65); 986 (Fig. 122)
method for iron, 469, 475
. for. molybdenum, 594
for titanium, 984
for uranium, 1027
for vanadium, 1039
Reference electrode, 2307

References, selected, are given to literature at the end of many special chapters, e.g. Chemical microscopy, Microanalysis, etC. etc.

Refractive index, I see also chapter 011

Chemical Microscopy
animal and vJgetable, oils, 1764
Chinese wood or tung oil, 1847
turpentine, 1851
Refractometer, Abbe, 1765 (Fig. 256)
micro, 2546
Refractory
materials,
decomposition,
1334 ff.
for chromium determination, 284
for iron determination, 477
Residue, total, in sulfuric acid, 2229
in water, 2077
Resin, in explosives, 1665
in paper, 1903
in rubber, 1969, 1981
in varnish, 1852
separation from polymerized oils, 1852
Resinates, in paint vehicles, 1842
Reverberatory slag, analysis, 2021
Rhenium) detection, 768, 769
estimation, 769
colorimetric, 772
gravimetric, 771
occurrence, 768
preparation and solution of sample,
769
separation, by distillation, 769, 770
from mOlybdenum; 770
oxine method, 771
Rodamine 6 G, adsorption indicator for
Ag, 830
Rhodium, detection, 742
estimation, graviJlletric, 746, 747
occurrence, 742 :
prep. and sol. of the sample, 743
properties,: 742
separation; from iridium, 744
from .palladium, platinum, 743, 745
from Tuthenium, 745
I
Rhombic sulfur, 935 (Big. 118)
Riffle samplel', for coal,"'1630 _
copper, 364
see also Sampling
Road oils, 1837
Rocks, see D~composition of Sample; see
Silica ; Silicates; Minerals.
chloride in, 264
Rosin, analysis, 1940. See also Resin
in oil, 1781
in paper, 1903
in soap, 2044
Rotameter, 2341 (Fig. 357)
R 20" see Iron and alumina!
Rubber, analysis 1966 ff.
accelerator~ fl5r, 1981, 2016
acetone extraction of, raw 1969
apparatus, 1970 (Fig. 285)
value, acid of extract, 1971
vulcanized, 1979

SUBJECT INDEX
Rubber, detailed examination of extract, 1980
alcoholic potash extraction, 1982, 2000
2003
'
ammonia, in latex, 1968
analysis, methods of
Findler's; 1975
nitrosite, 1975
precipitation, 1974
Spence's method, 1974
tetrabromide method, 1975
Wesson and Knorr's method, 1975
antimony in, 2006
ash, 1971.
analysis of, 2015
antimony in, 2006
sulfur in, '2006
barium carbonate in, 2010 ,
barium compounds in, 1989
barium ~ulfate in, 2006
ealculations, 2011
carbon in, 2007
ear bon black in, 2018
cellulose in, 1987, 2009
chloride in, 1986
chloroform extract, 1982, 2002
colophony, 1981
color pigments, 2019
copper, in crude rubber, 1971
crude, analysis of, 1967, 1968
ebOnite, 1991
fillers, 'inorganic, 1988, 2018
total, 1982
glue in, 1986, 1987, 2007
guanidines, 2017
gutta percha und balata, 1992

separation from rubber, 1995

hydrocarbons in, 1966, 1974, 1994,
2008, 2009
insoluble matter, organic, 1973
latex, 1968
manganese, in crude rubber, 1972
mercaptobenbothiazole, 2017
mercury, 1989
mineral matter, 1989, 1990
moisture, in raw rubber, ~967'
in vulcanized rubber, 1978
nitrogen in, 1973
in alcoholic potash extract, 1982
outline of analysis, vulcanized rubber,
1979
physical tests, 1995-1997
preparation of sample, 2000
protein in, 1973
raw, 1966, 1968, 2016
analysis, 1969
reclaimed, 1992, 2016
resin (rosin) in, 1969, 1981
rubber determination, 2000
as compounded, 2011
by volume, 2012

79

Rubber, hyiirocarbons, 1974

saponifiable matter in, 1980
selenium :in, 1986
softeners, 2017
solutions, analysis, 1976
solid~, solvents, 1976
solvents, ;lxamination, 2019
specific gravity of, 1992, 2002, 2012
sulfur, combined, 1985
free, 2000, 2003
in alco]::lolic potash extr., 1982
of vulcllllization, 1985
total, 2000, 2003, 2005
Kay aild Sharpe method, 1984
nitric acid method, 1984
Roth(l's method, 1983
Stevens' method, 1983
zinc oxide-nitric acid method,
19$4
unsaponifiable matter in, 1980, 2008
vermillion in, 1991
vulcanization coefficient, 2014
vulcanized, examination, 1977
scheme of analysis, 1979
water soluble components, 1978
zinc oxide, 2019
Rubber goods, standard methods of analysis, 1999 ff.
acetone extract, 2002
alcoholic-alkali extract, 2003
antimoI).y, 2006
ash, 2005, 2014, 2015
antimony in, 2006
sulfur in, 2006
barium carbonate, 2010
barium sulfate, 2006
calculations, 2011
composition of stock, 2015
carbon, 2007
,
cellulose, 2009
chloroform extract, 2002
extracts, 2012
glue, 2007
hydrocarbons, 2008, 2009
preparation of sample, 2000
reagent@, 2001
results, 2012
rubber itS compounded, 2011
by volume, 2012
specific gravity, 2002, 2012
sulfur, free, 2003
total, 2003, 2005
unsaponifiable, 2008
vulcanization coefficient, 2014
zinc, 1085
oxide, 2019
Rubeanic acid, reagent, 1217
Rubidium, detection, 893
estimation, 897
separations, 894

80

SUBJECT INDEX

Rubidium, from caesium

antimony chloride-ferric chloride
method, 897
silicotungstic acid method, 897
from potassium
9-phosphomolybdic acid method,
894
sodium bismuth nitrite method,
896
Ruthenium, detection, 737, 738
estimation, gravimetric, 740
occurrence, 737
preparation and solution of the sample,
738
separations, from iridium, 738
osmium, ~39
platinum, 739
rhodium, 739
solubilities, 737; 738
S
Salicylaldoxime, reagent, 1218
Salicylic acid method, for iron, 487
test for iron, 463
for silver, 819
Saline residues, alkalies in, 864
Salt cake, analysis, 1935
Samarium, 245
Sample, decomposition of, 1334 ff.
See laso Fusion; Preparation and solution of sample under each element,
Vol. 1.
Sampling, general methods, 1301 ff.
Introductory, 1301
methods for
gases, apparatus, 1325 (Fig.
174); 1326 (Fig. 175) ;
1327 (Fig. 177); 1328
(Figs. 178, 179)
general considerations, 1324
grab sample, 1326, 1327 (Fig.
176)
liquids, apparatus, 1319 (Fig.
168); 1320 (Fig. 169)
general outline, 1318
in motion, sampling, 1320
apparatus, 1321 (Fig.
170); 1322 (Figs. 171,
172)
special, 1323, 1324 (Fig. 173)
mill solutions, 680 (Fig.
81)
solids, apparatus for reduction
and preparation
crusher, 1312 (Fig. 159)
crusher and grinder, 1310
(Fig. 154)
crusher and sampler, 1313
(Fig. 160)
disc grinder, 1312 (Figs.
157, 158)

1:3am pIing, jaw and toggle crusher, 1300

. I (Fig. 153)
roll grinder, 1311 (Figs.
I 155, 156)
apparatus for sampling, 1304
I
(Fig. 151)
Iriffle sampler, "364, 1314
(Figs. 161, "162)
collection of gross sample,
1302-1307
cone and quartering "proress,
1309
drying, for moisture detenlli
nation, 1316
ovens, 1316 (Figs. 163, 164,
165), 1317 (Fig. 166)
long pile and shovel metho(l,
1307, 1308 (Fig. 152)
mixing and di~iding, 1314
reduction and preparation,
1307
referenccs on sampling, 1318
sieves, standard, 1314
dimensions, 1315
specifications, A. S. T. l\f.,
1329-1333
/'
unit of, 1301
vacuum ovens, 1301
Sampling, special methods and materials
for microanalysis, 2464
ccment, 1600
coal, 1620
paper pulp, 1896
water, 2102, 2116
see also individual materials, Alloys;
Aluminum, etc.
Sand, iron in, 482, 483
silica in, S05
Sanita,ry analysis of water, 2047
Saponifiable matte'r, in bituminous, substances, 1559
in rubber, 1980
number, in oil analysis, lS44
value, in oils;-ln~
table of, 1773-'
in soap, 2043
_
Saturated solution, analysis, solubility
determination, 2580
Scale, "in water, 2086
Scandium, detection, 773
estimatipn, 774
occurrence, 773
preparation and solution of the sample,
774
separations, 774
Schroetter's alkalimeter, 238
Sediment, in "oils, 1'761
Selenate;-lo1:" selenic acid, 785
Selenious arid, method for Zr, 11 01

SUBJEC'l' INDEX

81

Silica, in boron carbide, 183

Selenium,
in ctlment and clinker, 816, 1601, 1614
catalyst, Kjahldahl method, 633, 634
table of results, 817
commercial, analysis of, 787 (estimain fluorspar, 419
tion Se, Te, Pb, Cu, Fe)
in ghu;s, 21\)5
detection, 775, 77\)
estimation, gravimetric
in gypsum, 214
by reduction with KI, 784
in hydrochloric acid, 2205
lvith SO" 781, 784
in illmenite, 498
volumetric, thiosulfate method, 785
in pigments (paint), 1866
procedure, 793
in portland cement, 1601, 1614
special methods
in sllnd, 805
in commercial selenium; 787
in s~licon carbide, total, 814
in complex mixtures, 777
in sodium fluoride, 421
in'copper
in sodium silicate, 805
blister or pig, 789
in Htaniferous orcs, 996
electrolytic slimes, 789
in ultramarine blue, 1872
refined, 388, 791
in water, 803, 2059
in flue dust, 790
Silicates, alkalies in, hydrofluoric acid
in glass, 791
method, 883
.
in lead slimes, 789
J. Lawrence Smith's method, 882
in rubber, 1986
barium in, 125
in sodium selenite, 791
boron in, 164
in steel, 1480
decomposition, for fluorine determinain sulfuric acid, 2232
t\"''1., '\
'm 'tc).'mr'rnm, '1'1140
list, acid-soluble and insoluble, 797
occurrence, 775
mic),oanalysis of, 2545
preparation and solution of sample, 778
pot:.tsh in, short method, 872
separation, selenium and tellurium
See als.o Fusion methods; Silicon, prepfrom:
a),atlOn and solution of sample
iron group, zinc group,
Silicate of soda, analysis, 804, 805
alkaline earths and alkalies, 778,
Silico-fluoride test, for potassium, 862
779
Silicomanganese, manganese in, 568
selenium from tellurium
Silicomolybdic acid method, for silicon,
by distillation, 779
803
apparatus, 780, 782 (Figs. 94,
Silicon, detection, 794
95)
bead test, 795
recommended procedure, 782
lelld cup apparatus for, 794 (Fig.
Keller's method, 779
96)
Selenium steel, see Iron and steel
estimation, general gravimetri~ method,
Semi drying oils; see Oils
800
Separation, see Element, or Substance in
errors, 796; 806-807
Question
perchloric acid dehydration
solid phase, solubility measure, 2579
802
'
Sesame oil, 1781
special procedures
Retting time, of cement, 1597
in alloys, 802
Settlement basis, borate concentrates, 174
rapid method, 804
Settlement test, bitulllblQus substances,
in aluminum and its alloys, ~9
~5!H
in ferro chromium, 1493, A. S. T.
Sewer gas, 2372
M.
Shale oil, 1838
in ferromanganese, 1490, A. S. T .
. Rhort iodide methofl, for copper, 371
M.
Sieves, standard, 1314
in ferro silicon, 808, A. S. T. }\L
specifications, A. S. T. M., 1329-1333
1485
'
Silica, determination,
in ferrotungsten, 1497, A. S. T. M.
in aluminum hydrate, 28
in ferrovanadium, 1051, A. S. T.
ores, 19
M.,1497
in alundum, 33
in iron and steel, 493, 494, 806,
in "arsenic" (arsenious oxide), 110
A. S. T. M. methods, 1444, 1446
in barium orcs, 133
conductivity method, 811
in bauxite, 23
in fluorspar, 802
in bleaching powder, 2160

82

SUBJECT INDEX

Silicon, in limestone and soluble silicates,

Silver, inaustrial ilfPortance of methods,
802
820
in magnesium alloys, 543
occurrence, 818
in monel metal, 1379
preparation of p~re silver, 833
in nichrome, 1408
separations, 821 i
in nickel, 627
solubilities, 820; $21, 823
in refined silicon, 808
traces, rhodanine method, 820
in stellite, 326, 1412
Silver bullion, fire assay, 855
importance of methods, 795
coin ingots, 829
occurrence, 794
. fine silver (wet assay), 828
preparation and solution of the sample,
SlIver refined, Pt and Pd in 726
795
Silver residues, recovery of Ag from, 830
decomposition by aci"ds, 797
Silver arsenate, method for As". 100
fusion methods, 797
bromate, method for nitrite, 653
Hawley's procedure, 807
broinide, precipitation of, 191
methods for decomposition, alSilver carbonate, reagent, 2153
loys, minerals, ores, steels,
Silver chloride, electrode, 2300
e,tc., 798, 799'
method for chloride in, 264
separations, general, 800
precipitation of, 269, 281
lead and boron, 800
Silver <3hromate, indicator, 272
traces, colorimetric estimation, 803
. Silver cyanide, method for silver, 823
Silicon alloys, analysis of, 568
test for HCN, 1947
ferrosilicon and refined silicon, 808Silver iodide, method for iodine, 451
810
Silver 'molybdate, method for molybSilicon carbide (umpire analysis), total
denum, 590
silica, R,O., lime, magnesia, free carSilv!)r nitrate, detection of bromine, 188
bon, total carbon, calculations, 813of chromate, 282
816
.
method for bromide, 192
Silicon tetra:fluoride
chloride, 271, 273
in detection of Si, 794
cyanide, 661
method for :flourine, 409
iOdide, 457
properties of, 2347, 2365.
precipitant, 2335
volatility of, 401
standard solution, 0.1 N, 272, 120::1
Silicotungstic acid method, for caesium,
test for chloride, 261
897
hydrogen cyanide, 1947
Silver, detection
.
ruthenium; ;137
.
dry method, 818
SlIver solder, 1402
silver chlocide, 818
SiLver sulfate-sulfuric acid method for Cu
other tests, 819, 820
comp,ounds, 394
estimation, general procedures
Silver thiocyanate, method for bromid~,
gravimetric
192
electrolytic method, 824; micro,
chloride, 271
2518
Sizing strength, of starch, 2158
fire assay, 835 fl.; 860
Slag, analysis. 2020 -"'.
silver chloride method, 821
decomposition, 2021
- ~
,Silver cyanide method, 823
for Cu determination, 35:\
other methods, 823
for fluorine determiniiUon, 402
volumetric
for titanium determination, 978
adsorption indicator method (Faniter, for Se and Te, 790
jan's), 830 I
Slime, copper, Se and Te in, 789; Pt and
Deniges' method, 831
Pd in, 721)
Gay-Lussac's method, 826
lead, Se and Te in, 789
miscellaneous methods, 832
Sling pyschrometer, 2403
nephelometric method, 831, 832
Slip bands, metallography, 2564
U. S. Mint methods, 827
Slow method, for copper (electrolytic),
tables, 829
364
'
V olhard 's thiocyanate method, 824
Smith alkali method, 6,. 36, 882
estimation, special methods
Smoke, in .cltirifni:lY gases, 2388
in ores and concentrates, 860
Smokeless powder, analysis, 1695
in pig lead, 1413
Snake weighing tube, 2202 (Fig. 303)
in silver soldier, 1402

SUBJECT INDEX
Soap, detection in lubricating oil, 1736
determination, in grease, 1825
standard methods for analysis of
alkali (combined) fatty acid, fatty
anhydride, anhydrous soap, 2029
alkali (free); acid (free) 2032
borax in, 2035
chloride in, 2030
glycerine in, 2038
insoluble (in alcohol), analysis of,
2037
iodine value" 2043
moisture and volatile matter, distillation method, 2028
oven method, 2028
perborates in, 2036
phosphates in, 2036
rosin, 2044
sampling of, 2027
saponification value, 2043
'silica in, 2034
sodium carbonate, 2033
sugar in, 203.8
sulfatQs in, 2037 .
titer test, 2042
unsaponified and unsaponifiable matter, 2031, 2032
volatile hydrocarbons in, 2040, apparatus (Fig_ 287)
water soluble and insoluble matter,
2038
Soda-asbestos (" ascarite"), 1429
Soda ash, analysis of, 2261
modified, 2263, 2264
Sodium anthranilate, test for Cd, 203
Sodium bicarbonate, analysis of, 2264
in soda ash, 2261
mixtures with sodium carbonate,
2267
Sodium, detection
flame test; triple acetates; pyroantimonate; caesium bismuth nitrate
test, 875, 876
.
estimation, general procedures
gravimetric, difference method, 878
indirect method, .883
other methods, 882
sodium chloride method, 877
magnesium uranyl ace tat e
method, 879
sulfate method, 878
zinc uranyl acetate method, 878
volumetric, 882
estimation, special procedures
after removal of potassium, 8.81
in alundum, 36
in presence of K, Li, Rb and Cs, 884
in silicates,
.
hydrofluoric acid method, 883
J. Lawrence Smith method, 881,
882

83

Sodium, occurrence, 875

preparation and solution of the sample,
876
separations
from all constituents except alkalies,
876
from potassium, lithium, rubidium
and caesium, 876, 884

from potassium (large amounts),

877
Sodium-alcohol method, for halogen, 266
Sodium-ammonia (liq.) method, for
halogen, 266
Sodium alizarin sulfonate, 407
Sodium aluminate, ill NaDH, 2256
Sodium arsenite, 453
Sodium bismuth nitrite, method for Rb
and Cs, 896
Sodium carbonate and hydroxide in mixtures, 2265
Sodium carbonate, analysis of, 2261
fu~ion of silicates, 797
fusion test, for manganese, 555
fusion with, 4, 465, 477, 616, 797,
956, 1344 :II.
in mixtures, with bicarbonate and
borate, 2271
in presence of sulfur acids, 929
in sodium fluoride, 421
in sodium hydroxide, 2256
in sodium sulfide, 2184
preparation, pure, 2193
Sodium chlorate, method for ni.anganese
rlioxide, 310
Sodium chloride, determination
in gypsum, 215
in soap, 2033
in soda ash, 2261
in sodium ferrocyanide, 667
in sodium fluoride, 421
in sodium hydroxide, 2256
in sodium nitrate, 645
in sodium silicate, 805
in trisodium phosphate, 705
standard, ultimate, 1203
Sodium chromate, as precipitant, 2335
Sodium ferro cyanide, sampling, moisture,
total ferrocyanide, sodium chloride, sodium sulfate, foreign matter, 665~668
reagents for analYSis, 666, 667
Sodium fluoride, analysis, insoluble NaF,
Na,SD., sodium sulfate, carbonate and
chloride, silica, in, 420, 421
Sodium hydrosulfite, estimation of, 930
(Fig. 114)
method for oxygen, 681
standard solution, 679
Sodium hydroxide, alcoholic solution,
1198
analysis of, 2255

SUBJECT INDEX

84

Sodium hydroxide, determination in

liqnid bleach, 280
fnsion method for ores, 957
specific gravity, tables, 1158 1159,
2268'
'
standard solntion, 2197
removal of carbonate, 2197,
2198
test for Pt, 713
Sodium metabisuliite, alkali titration of,
927
\
determination in presence of carbonates, chlorides, sulfates, suI.
:fites, thiosulfates, 929
reductant for ferric salts, 471
Sodium nitrate, analysis of, 644, 645
Sodium oxide, determination
in aluminous materials, 21 27, 28
in liquid bleach, 279
'
in sodium silicate, 804
in ultramarine blue, 1873
see also, Alkalies, determination of
Sodium peroxide, fusion method chromite, 284
'
tin ores, 957
method, halogens in organic matter,
~67

test, for iron, 463

. see also peroxide
Sodium phosphate, test for u~anium, 1018
Sodium plumbite, 1742
Sodium p),roantimonate, test for Na, 875
Sodium rhodizinate, indicator, 129
Sodium, selenite, analysis of, 790
Sodium silicate, analysis of, 804, 805
in soap, Zl'l34
in soda ash, 2262
in sodium hydroxide, 2256
Sodium sulfate (salt cake), analysis of,
1935
determination of sodium: as, 878
in soda ash, 2262
in sodium ferrocyanide, 668
hydroxide, 2256
nitrate, 645
sulfide, 2183
in trisodium phosphate, 705
with other sulfur salts present, 929
Sodium sul:fide, analysis of, 2181-2187
(insoluble, FeS, sulfide, thiosulfate, sulfite, alkalinity, polysulfide,
sulfate, carbonate, calculations)
in soda ash, 2262
Sodium sulfite} acid titration of, 928
determination, in presence Of carbonates, chlorides, metabisul:fites,
sulfates and thiosulfa tes, 929
in soda ash, 2262
reduction of ferric salts by, 471
test for osmium,' 748

Sodium thios1jlfate, _determination in

presence of metabisul:fites, sulfates, lhiosulfates, carbonates,
chlorides l 929
standard solution, 0.1 N, 452, 1211
for copper, 1211
test for molybdenum, 585
ruthenium, 738
Sofnol red, 176
Softeners for Rubber, 2017
Softening, of water, 2086
range, bituminous substances, 1523
Soils, iodine in, 460
nitrogen in, 634, 646
Solder, antimony in, 83, 1390
copper in, 1393
silver, 1402
soft, and white metal, 84
zinc in, 1085, 1392
Sole~oid apparatus, electrolysis, 362
Solidification points, of explosives 168:j,.
1682
'
,
'Solid phase, identification of, 2581
So!ids, in explosives, separation of, 1675
m gas, suspended, 2414
,,'in hydrochloric acid, 2205
in water, 2055
sampling of, 1301-1318
Solubilities, elements, oxides and salts,
see under Estimation, Prep. and
solution of sample, solubilities, under each element, Vol. 1.
Table, inorganic componnds, 11611164; also inside cover
table, qualitative, 1136
Solnbility, dete,rmil1ation of, 2570 ff.
apparatus,' 2571-2579 (Figs. 4874Q7) ,
,
distribution coefficients, 2586
electrolytic
conductivity
method,
2586

free:zing point method, 2584

of gases in -liquids, 2587, 2589
(Figs. 499, 500) -~ .
saturated solntion, analysis of, 2580
separation, saturated -solution from
solid, 2579
solid' phase, identi:fication, 2581
synthetic method, 2582
titrationl method, 2585
transition temperatures, 2582
volume change method, 2584
tests, bitumin. substs., 1539
Solvay Process Co" methods, 240
Solvents, for bitumino,us substs., 1586
for inorganic- substances, 1334-1338
for rullbei,-:--1976, 2019
Soundness, of cement, 1597
Spark spectra, 2599 (Fig. 508)

SUBJECT INDEX
Spceific gravity, see also under tables
bitumin. ~ubsts., determination
analytical balance method,
1517 (Fig. 189)
hydrometer
method,
1514
(Fig. 187)
pycnometer
method,
1515
(Fig. 188)
comparison with degrees 'Baume,
1168
definition and methods of determi
nation, 1175
animal and vegetable oils,
1764
burning oils, 1704
linseed oil, 1844
turpentine, 1851
gases, 2344, 2397; table, 2411
hydrometer method, 1704, 2205
(Fig. 305)
of acids, see individual acids; see
Tables
.
of glass, 2165
pycnometer method, 1764
rnbber, 2002, 2012
Westphal balance method, 1704
(Fig. 239)
Spe~ific heat, definition, 1174
of gases, 2425
Specificaticns, for cement, 1597
lil\Seed oil, 1843
Spectral lines, for Cd, Pb, Fe in zinc,
1092
Spectroch~mical analysis and Spectro
graphic analysis, 2592 if.
apparatus, 2594 ff.
light sources, 2598
are, 2598, 2599 (Fig.
508)
flame, ~598
other, 2600
bpark, 2599 (Fig. 508)
spectrographs, 1091, 2594 if.
(Figs. 502-506)
Hartmann. diaphragm, 2598
(Fig. 507)
electrodes, 1089, 1091, 2599
introduction, 2592
patent situation, 2617
photographic materials
developer, 1091, 2614
plates, 1091, 11614
quantitative methods, 2609
comparison, 2609
control of conditions, 2613
intensity
measuren~ents,
2611
.
internal standards, 2610
working curve, data, 2610,
2611, 2613
references, 2617

85

Spectrochemical analysis and Spectrographic analysis, spectra

absorptioh, 2614
emission, 2592 ff.
standards, 2600
tables, spectral lines, 26022603, 2604-2607
zinc, examination for Pb, Fe,
Cd, A. S. 'D. M., 1089
standard solutions, 1092
Spectroscopic method, for Li, 891
i:lpectra, absorption, 2614
of carbon monoxide in blood, 219
are, tables of lines, 2602-2603; 26042607
Spectrum, method for detection of:
barium, 117, 119, 120
cadmium, 198
caesiUm, 893
calcium, 205
galliUm, 426
lithium, 887
rhenium, 769
rubidium, 893
scandium, 773
sodium, 875
strontium, 899
thallium, 943
thorium, 947
Plate, II, alkaline earths, facing, p.
119
Spelter, see zinc, slab
Spent oxide, evaluation of, 932
Spontaneous combustion test, of oils,
1782
Standard acids; 2194 .ff., see hydrochloric, sulfuric acid, etc.
effect of temperature changes on
normality, 2198.
standardization, see Acids, Bases,
etc.
approx. by sp. gr., 2198
S~andard
substances (primary standards; ultimate standards)
for volumetric analysis, 1200-1201
acidimetry and alkalimetry
benzoic acid, 2197
hydrochloric acid, 2196, 2197
potassium
acid
phthalate,
2197
sodium carbonate, 2193 (Fig.
299)
oxidation-reduction rea c t ion s ,
1203 ff. (See also individual
substances as: Arsenious oxide,
Sodium
oxalate;
Potassium
iodate; Potassium dichromate;
etc.)
precipitations and complex reactiolls, 1200 ff.

86

SUBJEC'r INDEX

Stannous chloride, analysis of, 962

method for iron, 484
apparatus, 485 (Fig. 58)
reduction of ferric solutions, 470
standard solution of, 1212
test for mercury, 574
osmium, 749
palladium,' 722
platinum, 713
selenium, 776
tellurium, 776
, Starch, analysis, 2156
converted starches, 2157
in explosives, 1667
in paper, 1906, 1907
in soap, 2039
solution, preparation, 453, 1198
prllservation of, 453, 1198
Steam distillation, see Distillation
Steel, see Iron and Steel, especially 1421

Sulfates, detection, 904estimationl

in barytes, 1867
in chromium pJating baths, 917
)n gypstim, 215
in hydrdchloric \acid, 2204
in leaded zinc or zinc lead, 1861
in potash, 2169, 2170
in rubber, 1985
in soap, 2037
in sublimed white lead, 1858
in water, 2064
volumetric method, 2091, 2096
in whiting, 1866
in zinc oxide, 1862
soluble, estimation of, 917
in blanc fixe, 136, 1867
Sulfates and sulfides, .estimation in mhtures, 925
Sulfhydrate in presence of sulfide, 924
Sulfides, detection, 904
,
iI.
estimation, in ores, by combustion, 916
Stellite, analysis of, 325, 1409
in paint (composite white), 1868
Stibine, 2347
in sodium sulfide, 2183
properties, 2367
in sulfates, 925
Stibnite, antimony in, 77
stains, 936
Stirrer, 680 (Fig. 80)
/' Sulfide and sulfhydrate, ,in mixtures, 924
Stopcock grease, 1219, 1220
Sulfide ores, decomposition of, 906
Streak t,est, of bitumens, 1513
Sulfide sulfur
Strengthening, of solutions, formulae,
in 'rubber, 1985
2236
traces, 936
Strontium, detention, gen. proc.; flame
Sulfites, detection, 904
test; sulfate test; spectrum, 899
estimation, iodine titration method,
estimation, gravimetric
926
as car bOlla te, 902
in pres. of other acids of sulfur, 925
as oxide, 902
in rubber, 1985
as sulfate, 901
in sodillm sulfide, 2183
volumetric, alkalimetric method, 902
Sulfo~yanate, see thiocyanate
silver nitrate method, 902
Sulfonation test, bituminous substan('cs,
occurrence, ores "and minerals, 899
1557
preparation and solution of the samSulfur, amorphous, determination' oj',
ple, 900
934
separation, from alkalies and magin crude sulfur, 938
nesium, 900
detection, 'elellJ~nt, 903
from barium, 122, 123, 901
. compounds, 904 -~
from calcium, 123, 901
estimation, general procedures
solubilities, 900
graVimetric, as bal'ium sulfate, 908
sulfate, determination in barite; 131
from coId diluted solutions,
solubility, 121
909
Strychnine, test for, 1956, 19.57
oxidation and precipitation,
Sugar, in explosives, 1674, 1678
apparatus" 910 (Figs. 109,'
I
109a)
in soap, 2039
combustion method, 916
Sulfate method, for barium, 1.27
cadmium, 201
volumetric
calcium, 216
barium chloride method, tetrahylead, 504
droxyquinone indicator, 919
strontium, 901
barium chloride-potassium dichrozinc, 1058
mate method (Wildenstein), 918
-15'arium 'chromate-iodometric
separation, alkaline earths from aImethod (Hinman), 919
kali,es, earths and magnesium, 121

SUBJECT INDEX
Sulfur, benzidine hydrochloride method,
920
evolution as hydrogen sulfide, 911
in iron and steel, 911, A. S. T.
M., 1442
apparatus, 912 (Fig. 111),
915 (Fig. 113), 1443
(J'~ig. 185)
reagents, 911-912
in iron ore briquettes, sodium
sulfide, 913
procedure, 914
reduction of sulfates, 913,
apparatus (Fig. 112)
estimation, special procedures
available, in brimstone, 934 '
free, in mixtures, 931 (Fig.
115)
"in alcohol, 2141
in alloys, see Alloys; sec
Bronze; Steel; Ferroalloys,
etc.
in afundum, 36
in ammoniacal liquor, 638
in barium ores, 133
in black powder, 1661
in bronze, A. S. T. M., 1370
in cement, 1601
in chromium plating baths, 917
in coal, alld coke, 906, 1640,
1642, 1643
in copper, 389
in crude sulfur, 937
in cyanide, 664
in ferromanganese, 1490
in ferro molybdenum, 1508
in ferro tungsten, 1014, 1502
in f\lrrovanadium, 1497
in fuel oils, bomb method, 1756
in gas, 2395
in gasoline, A. S. T. M., 1743
in glass, 2168
in iron and steel, 492, 494, 911,
1442 ff.
'
in monel metal, 1379
in nichrome, 1409
in nickel, metallic, 626
in oils, 1712; lamp method, 1713
and liquid fuels, 1743, 1750
in organic substances, 2507
(micro)
in paper, active, 1901
in pigments, 1874
in refined sulfur, 934
in rocks, silicates, insoluble sulfates, 906
in rubber, 1982, 2003; total,
1983, 1984, 2003, 2005
in spent Oxide, 932

87

Sulfur, in steel, evolution method, A. S.

T. M., 1442, 1433 (Fig.
185)
gravimetric, 1440
in sulfide materials, available
H 2 S, 923
orcs, 916
in thiocyanic acid and its salts,
925
in ultramarine blue, 1873
in' zinc oxide, 1862
forms of, 903, 935 (Figs. 116, 117)
importance of methods, 904
occurrence, 903
preparation and solution of the sample,
905
ores, coal, rocks, silicates, insoluble sulfates, 906, 907
sulfide, sulfate, thiosulfate, 905
separation from
ammonium and alkali salts, 908
/ chl'\)rates, 907
iroil-bearing materials, 907
metals having insoluble sulfates, 907
nitrates, 907
silica, 907
total, in crude brimstone, 937
in refined sulfur, 933
Sulfur dioxide, in bisulfite liquor, 1929
in burner gas, 2399
in zinc oxide, 1862
method for selenium, 784
separation from Te, 779
properties of, 2364, 2365
gas analysis, 2347
reduction method for iron, 976
test for molybdenum, 585
test for selenium and tellurium, 775
Sulfur extraction lI-pparatus, 577 (Fig.
62)
Sulfuric acid, analysis of, 2228; see also
Oleum, 2233
antimony in, 2231
arsenic in, 2230
arsenious acid in, 89
Gutzeit method, 105
boiling-point-composition c h a I' t ,
2228
combined, in soluble sulfates, 917
copper in, 2231
fluoride in, 2233
free, estimation, 921
-freezing-point-composition c hal' t.,
2228 (Fig. 308)
gases, 2399
hydrochloric acid in, 2232
iron in, 2230
lead in, 2230
mist in gases, determination, 2416
(Fig. 387)
nitric acid in, 2216, 2232

88

SUBJEC'l' INDEX

Sulfuric acid, normal solution of, 2194

gravimetric
standardization,
2195
volumetric standization, 2194
residue, non-volatile, 2229
selenium in, 2232
specitic gravity, 1150---1154, 22382243
zinc in, 2231
Sulfuric acid, determination in
,. acetic acid, 2250
aluminum salts, 15
ammoniacal liquor, 638
arsenious acid, 109
formic acid, 2247
hydrochloric acid, 2204
hydrofluoric acid, 2210
nitric ac,id, 2212
oleum and mixed acids, 2233
paint pigments, 1873, 1875, 1876
tartaric acid, 2255
test for lead, 500
molybdenum, 586
selenium, 776
tellurium, 776
Sulfuric acid-mercury method, copper
compounds, 396
Sulfurimeter, 935 (Fig. 116)
Sulfurous acid (see also Sulfur dioxide)
free or combined, in sulfites, bisul-
fites, metabisulfites and
thiosulfa tes
acidimetric-alkaIimetric, 927
gravimetric method, 925
iodometric method, 926
in lleetic acid, 2251
in hydrofluoric acid, 2210
method for copper compounds, 393
gold, 438
reduction of chlorates, 274
iron, 471
test for iridiulI\, 730,.,
Sulfur trioxide (see also Sulfur; Sulfuric acid)'
in gypsum) 215
in sodium silicate, 80;;
in zinc chloride, 1080
Surveillance test, explosives, 1697
Synthetic method, solutary determination; 2582
T
'rabIes,
. absorption spectra, rare. earths, 257
acids, qualitative separations, 1112,
1128-1135
specific gravity
acetic acid (Oudemans') table,
1156, 2252
melting points (Rudorff),
1156

Tablell, specitic gravity

hydrochloric acid, Ferguson's
table, 1142.:.1143; 22062207
cOll-stant-boiling, 1144, 2196,
2208
nitric a.cid, Ferguson '8 table,
1145-1146; 2220-2221
nitric acrd, Lunge, Rey, 11471148; 2222-2223
phosphoric acid, Hager table,
1149, 1150, 2127
sulfuric acid, Ferguson and Talbot's tables, 1150-11&1;
1152-1154; 2238-2242.
Bishop table, 1154; 2242
approximate boiling points,
1151
fuming, 1155" 2243
specific gravity charts and
boiling points, 2228
alcQhql, ethyl, specific gravity, 21422146
alkalies
aqua anlnl011ia, Ferguson '8 table,
1157, 2269
coefficient o~ expansion, 1157
/ sodium hydroxide, Lunge's table,
1158-1159; 2268
arc spectra, 2602-2603; 2604-2607
atomic weights, front cover.
bases, reactions of, 1114-1127
Baume degrees, comparison with specific gravity, 1168
bead tests, 1109
bituminous substances, 1510-1512
buoyancy constants, of water, 1226
cement analysi's,. 817, 1619
centigrade anti. Fahrenheit temperature
scale~, 1165
compounds, inorganic, useful data
(Meiklejohn), 1161-1164
conversion tables, volume, weight, energy; 1176-1178
density, 'of water, .J.g28
ebullioscopic constants,~25;lj.
electromotive series of elements, 1141
energy, conversion tables, 1177
fatty acids, characteristics, 1794
flame tests, 11 08
gases
aqueous yap or pressure vs. temperature, 2,425
calorific power, 2426
chimney or flue, calculations, 2383
commercial, compositions (typical),
2397
constants,of, 11661:1167; 1171-1172;
2424 ._correctiOn factors, 2429-2430
cylinders, data, 2428

SUBJECT INDEX
'fables, gases
dangerous concentrations (toxic),
2427
efficiency and heat loss, 2383
emergent stem data, calorimetry,
2431
humidity corrections, 2432
producer gas, composition, 2388
specific gravity, 2411
specific heat,' 2425
s.ulfur dioxide, iodine values, 242ii
volumetric factors, 2428
hal'uness, in water; 2082
heat of combustion, -compounds, 11851186
.
fusion, elements and compounus,
1188-1190
j.ndicators, for pH measurement, 2281
for titrations, 2191
ionic mobilities, 2324
logarithms and antilogarithms, 11791182
melting points, of elements, Table I,
1140
metric and other units, 1169-1170,
2426
microorganisms, in water, 2100; identification, 2106; c9ntrol, 2113, 2114
milliequivalents, inorganic substances,
1161-1164
minerals, common, list of, 1183-1184
oils, characteristics of, 1795, 1796
drying oils, 1849
phosphate, typical compositions, 691
platinum metals reactions, 753
qualitative tests, 1107-1135
reaction coefficients, water analysis,
2086'

salt effects, on pH, 2286

saponification numbers, 1773
separations, platinum metals, 729
silver tables, U. S. Mint, 829
Simpson's table, columbite-tantalite
specific gravity, 331
solubility, qualitative, 1136
quantitative, back cover
specifi~ations

fuel .oils, greases, 1791

gasoline, burning oils, 1790
lubi'icating oils, 1792, 1793
specific gravity, and degrees Baume,
1168
('see Acids; Alkalies; Alcohol, etc.)
spectra, absorption, rare earths, 257
arc, persistent lines, by- elements,
2602-2603
by wavelengths, 2604-2607
"temperature, conversion, centigrade to
Fahrenheit, 1165
from colors of heated metals, 1141
of flames, 1140

S9

Tables, temperature
standards, 1141
vapor pressure of water, Regnault,
Broch and Weibe, 1160
vapors, constants of, 1166-1167
viscosity, 1725
conversion data, 1799
chart, 1798
water, buoyancy constants, 1226
density, 1228
weights for various volumes, 122fl
waxes, characteristics of, 1797
weights and measures, conversions,
1176
precision and tolerances, analytical, 1234
zinc, slab, tolerances, 1071
Tallow, see oils, fats, waxes, tables.
Tannic acid, 2158
Tannins, 2159
in columbium-tantalum separation, 346
Tantalite, 331
Tantalum (see also Columbium and Tan
talum)
detection, 332, 333
traces, 333
estimation, 347
separations, 344-347
specific gravities, minerals (Simpson),
331
uses, 331
Tar, see Bituminous substances.
Tartar emetic, antimony in, 82
Tartaric acid, determination, Goldenberg
Co. method, 2254
Oulman method, 2254
Tartrate test, for potassium, 862
Tellurates, telluric acid, estimation of,
786
Techniques, special, Vol. II, part II, 2189
and ff.
Tellurium, detection, general procedure,
775, 776
in complex mixtures, 777
estimation, general procedures
gravimetric, as dioxide, 786
hydrazine-sulfur d i 0 x ide
method, 785
sulfur dioxide method, 781
volumetric, 793
special procedures
analysis of metallic Te, 788
and selenium, in complex mix
tures, 777
in copper, 388, 789, 792
in flue dust, 790
in lead slimes, 789
separation, 780 (Fig. 94)
and
determination,
782
(Fig. 95)
occurrence, 775

90

SUBJECT INDEX

Tellurium, preparation arid solution of

sample, 778
separations, see Selenium
Tellurium and selenium, in selenium, 787
in tellurium, 788
refinery methods, 787
Temperature, effect on normalities of
solutions, 2198
on pH measurements, 2288
estimation, by optical method, 1141
Fahrenheit-centigrade conversions, 1165
of flames, 1140
transition, and solubility, 2582
Tensile strength, of bituminous substances, 1564
of cement, 1597
Terbium, 245
Test lead, reduction of ferric salts by, 470
Test papers, see Antimony; Arsenic;
Lead; Sulfur
Tetrabromide, in rubber analysis, 19.75
Tetraethyl lead, in gasoline, 1804
1, 2, 5, 8-Tetrahydroxyanthraquinone, test
for Be, 138
Tetrahydroxyquinone, indicator, 129
for sulfate, 919
Tetryl, analysis of, 1684
Thallium, detection, 942
estimation, gravimetric, 944
volumetric, 944
preparation of solution, 943
separations, 943
Thallous oxide test, for platinum, 713
Thermometer, for disti1l~ions, oils, 1707
see A_ S. T. M. methods, calorific
value; :flash points, etc.
Thiocyanate, determination in presen~
of chloride and cyanide, 277
cyanide, 661
method for determination of
copper, 357
iron, 486
mercury, 580
rhenium, 772
silver (Volhard), 824
test for hydrocyanic acid, 1946
silver, 819
Thiocyanic acid, sulfur in, 925
Thioglycollic acid, method for iron, 487
Thionalide, reagent, 1218
Thiosulfate, detection, 904
determination in presence of sulfi(le
and sulfhydrate, 925
in sodium sulfide, 2183
method for thorium separation, 950
titrating iodine, 925
standard solution, for copper, 308,
1211
for water analysis, 2079
general purposes, 452, 1211

Thiosull'ate'iodide. lIlethod, lead peroxide

estimation, 524 '
Thiourea, test for osmium, 749
Thorium, detection, ;946, 947
estimation, 953 . I
occurrence, 946
preparation and sMution of the sample,
948
separations, phenylarsonic acid method,
952
pyrophosphate method, 949
thiosulfate method, 950
Thulium, 245
Time-viscosity data, conversion Say bolt
to Engler, 1799
Time required in, gas analysis, 2351,
2352, 2359
Tin, detection, 954, 955
estimation, general procedures,
gravimetric
;electrolytic, micro, 2513
hydrolysis; ignition to oxide,
959
precipitation by cupferron,
961
purification of the oxide, 960
sulfide methOd, 960
volumetric, ferric chI 0 rid e
method,966
iodine method (Lenssen), 964
Baker's modification, 964
Sellars' reduction appapar a t us, 966, 967
(Figs. 119, 120)
estimation, special procedures
in alloys, brass and bronze, 969
Fig. 121 .
in aluminum alloys, 49
in Babbitts, 968
in bearing metals, 1362, 1363
in 'bichloride of tin, 962, 963
in bismuth bronze, 1375
in, bronze, A. 8. T. M., 1356
in canned food products, 971
in ferrotungsten ll!ld tungsten
powder, 1015, 15Q4
in lead-base alloys, l383, 1385
in lead (metallic) pig, 1416
in, magnesium alloys, 549
in silver solder, 1404
in tin-base alloys, 1383, 1385
in water 2080
in white metal, 973
in Wood's metal, 1379
in zinc-base alloys, 1396
I importance of methods, 955
occurrence, 954
I
preparation ~:pd solution of the sample,
-955
cyanide process, 955
hydrogen reduction, 957

SUBJECT INDEX
Tin, sodium carbonate method, 956 .
hydroxide method, 957
peroxide method, 957, 972
separation, from antimony, 69, 71
in alloys, 957
from arsenic and antimony, distillation method, 69, 1385
from copper; lead, 957
from iron, etc., 958
from phosphorus; tungstic acid, 958
from silicon, 958
from silver, etc., 958
general method, 957
Tin-base alloys, 1381
Tincal, 164
- Tissue, iodine in, 460
Titaniferous materials, analysis of, 995
slags, decomposition of, 978
Titanium Alloy Manufacturing Co.,
methad for N in steel, 656-658
Titanium, detection, 976
estimation, general procedures
colorimetric, hydrogen peroxide
method, 987; app., 988, 989
(Figs. 123, 124)
thymol method, 994
gravimetric, Gooch-Thornton
method, 981
volumetric, after reduction with
zinc
apparatus, 986 (Fig. 122)
ferric salt titration, 985
permanganate titration, 984
estimation, special procedures
.
in aluminous materials, 21, 25,
27
in aluminum and its alloys, 42
in alundum, 34
in bauxite, 997
in clay, 2163
in ferro-carbol1-titanium, 983,

98i
in illmenite, 496
in p.ilper, 1925
. in pigments, mixed, 998
titanium oxide, 986
titmlOx, 999
in silicon carbide, 815
iii steel, 1468, 1482
treated with Fe-C-Ti, 991-993
importance of methods, 976
occurrence, 975
_
preparation and solution of the sample,
977-979
separations, from alkaline earths, 980
aluminum, cupfel'ron method, 980
bivalent metals (Mn, Ni, .co, Zn,
etc.), 980
copper, zinc, aluminum, iron, etc.,
979
hydrogen sulfide group, 979

91

Titanium, iron, ether method, 980

other elements, 980, 981
standard solution of, 989
Titaninm oxide, in pigments, 1863
Titanium pigments, analysis of, 986, 99a,
. 1863
Titanous salt, estimation of iron, 479
apparatus, 480 (Fig. 57)
prepm-ation of solution, 479
standardization, 480, 481
Titanox, analysis of, 999
Titan yellow G, test for Mg, 529
Titer, see Milliequivalents, table of,
1161-1164
Titer test, for oils, 1784
for soap, 2042
Titration method, solubility determina(ion, 2585
of acids and alkalies, 2200
Titrations, see Acids; Alkalies; various
oxidizing and reducing agents, etc.
see also Conductometric methods; Potentiometric methods; Microanalysis
TNT, see Trinitrotoluene
Tolerances, for weights, 1234
Toluene, microdetermination of, 2547
use in moisture distillation, 1342
o-Toluidine, 280
Traces of: (see also under element in
question; see Poisons; see Microscopy)
aluminum, detection with alizarine S, 2
aluminon, 2
determination, with aluminon, 55
general method, 59
rapid method, 61
ammonia, 630
antimony, detection, 64
determination, 78-81
arsenic, detection, 87
determination, 105-107
barium, detection, flame and spectrum,
117, 120
beryllium, 143, 145
bismuth, colorimetric determination,
158, 159
boron, tests, 162, 16il
quantitative, 185', 186
bromine, detection, 188
estimation, 192, 193
cadmium, 202
caesium, 893
calcium, flame and spectrum, 120, 205
carbon, colorimetric est., 232, 234
cerium and other rare earths, detection,
257, 258
chlorine, detectiol!., 261, 281
chromium, detection, 282, 283
determination, 289, 290
cobalt, detection, 305, 306
estimation, 314

92

SUBJECT INDEX

Traces of, columbium, 332, 333

copper, detection, 349, 350
determination, 376-381
fluorine, detection, 399-401
estimation, etching method, 421, 422
Steiger-Merwin method, 410
zirconium-purpurin nlethod, 423
gallium, detection, 426
gold, detection, 431-433
determination, 438, 439; Fire assay,

835 ff.
halogens, detection, 281
indium, 445
iodine, detection, 448
iron, detection, 462, 463
determination, 482, 486, 487
lead, detection, 500
determination, colorimetric, 517, 521
gravimetric, 514
lithium, detection, 887
estimation, spectroscopic, '891
magnesium, detection, 529
manganese, detection, 554, 555
determination, 572, 573
_mercury, detection, 574
metals, microelectrolytic estimation,
2514
molybdenum, detection, 585, 586
estimation, colorimetric, 606, 608,
609, 610, '612
nickel, detection, 614, 615, 617
nitrogen, .as an\monia, Nessler's test,
630
as nitric acid, detection~ 630, 631
det!)rmination, 640
as nitrous acid, detection, 631
oxygen, estimation, 687, 688
phosphqrus, detection, 689, 1945
platinum, detection, 712, 713
metals, iridium, 730
osmium, 748, 749,
palladium, 721, 722
rhodium, 742
ruthenium, 737, 738
potassium, detection, 862
radium, detection, 758
rhenium, detection, 768, 769
()stimation, 77'it
rubidium, detection, 893
scandium, detection, 773
selenium, detection, 775, 776
silicOll, detection, 794, 795
estimation, colorimetric, 803
silver, detection, 818-820
estimation, 820; fire assay, 835 ff.
sodium, detection, 875, 876
strontium, detectio~, 120, 899
sulfur, combined and free, detection
903,904
estimation, ~36
tantalum, detection, 332, 333

Traces of}tellurium, detection, 775, 776

thalliurri, detection, 942
thorium) detection, 946, 947
tin, detection, 954, 955
titanium, detection, 976
estimation, colorimetric, 987, 994
tungsten, detection, 1001
uranium, detection, 1017
vanadium, detection, 1031
estimation, 1051, 1052
zinc, detection, 1054
estimation, 1068
zirconium, detection, 1093
Transition temperature and solubility,
2582
'
Trinitrotoluene, examination of 1681
Ti-iphenyltin chloride, 1218
'
method for fluoride, 406
Trisodium phosphate, analysis of', 703~

105
Trisulfide, antimony, 72
arsenic, 96 '
Tropeoliu 00, adsorption indicator, 281
Tung oil (Chinese wood oil) e::mmination of, 1847
'.
specifications, A. S. T. M., 1847
Tungstate method, for calcium, 216
Tungsten, detection, in alloys and steel,
1001
4n minerals, 1001, 1002
estimation, general
grl:!vimetric, 1002, 1005
speda]
in' ferrotungstell, 1011, 1499
in stellite, 326
in steel, A. S. T. M., 1465
ores and concentrates, analysis,
1005
acid attack, 1005
fusitill-atta.c_k, 1010
iron in, 1007'
molybdenum, in, 1007
tantalum and columbium in,
1007
occurrence, 1001
prePfll'ation and solution of the sa Illpl,e,.1002
reagent, cinchonine, 1006
separation fro.m, arsenic and phosphorus, 1003
iron, 1005
lead 1005 I
_n1eiy'bdenum,' 1004
ililie,((, 10'0&
tin, 1003
titanium, 1004
uraniulll, 1005
vanadium, 1004
solubilities of acids and oxides, 1002

SUBJECT INDEX
Tungsten,

tungsten metal and ferrotungsten, A. S. T. M. methods,

1011, 1499 ff.
determination of
antimony, 1015
copper, 1015
nh0sphorus, 1012
silicon, 1014
sulfur, 1014
tin, 1015, 1016 (Fig. 126)
tungsten, 1011, 1499
Tungstie acid, method for tnngsten, 1002
Turbi<1ity tests, of water, 2047
(rnrmeric indicator, 185
test for boron, 163
colorimetric estimation, 185
Turpentine, analysis of, 1850 (color; sp.
gr.; ref. index; dist. test; polymerization; standards)
'.\

u
Ulexite, 164
Ultimate standards, 1200, 1201
Ultrl1marille blue, analysis of, 1872
Unbuffered solutions, pH of, 2284
United States Government, etc., see U. S.
Units, metric and other, tables, 1169,
. 1170
physical, 1174, 1175
Unsaponifiable matter, in linseed oil,
1844
detection of, 1773
in oil, identification 'of, 1774
in rubber, 1980, 2008
in soap, 2031, 2032
Uranium, detection, 1017
urano'us and uranyl salts, 1017
estimation, general
gravimetric, as oxide, 1022
volumetric, 1027
special
.
ill carnotite, 1024
in ferro uranium, 1028
hl iron and steel, 1028
ill ores, 1022
occilrrence, '1017
preparation and solution of the sample, 1019
separation, cuprerron method, 1021
from hydrogen sulfide group metals,
1019
from iron, and metals having insoluble carbonates, 1020
from molybdenum, tungsten, vanadium, 1021
from vanadium, 1020, 1021
solubilities, 1019
uranous snIts, detection, 1017
uranyl salts, detection, 1017
uses, 1018

93

Uranyl :reetates of sodium, with magnesium, zinc, cobalt or nickel, 875

Urea method (Willard and Tang) for
aluminum, 11
'
Useful data, inorganic compounds, table,
1161-1164
U. S. Government specifications, petroleum products, 1790-1793
U. S. Mint, tables, silver, 829
V

Vacuum correction, for weighings, 1223

Vacuum oven, 1317
Vacuum-tube voltmeter, 2321
Valenta test, for oils, 1765
Valuation, of aluminum ores, 17, 19
of fluorspar, 419, 420
Vanadium, detection, 1031
contrast with Cr salts, 1031
estimation, general procedures, 1035
gravimetric,
lead
acetate
method, 1036
mercurous nitrate metholl,
1036
volumetric
ether extraction, 1052
phosphomolybdate
method,
1051
reduction
by Hel, 1046
by H 2 S; S02) 1037
by zinc, 1037
titration by permanganate,
1037, 1039
potentiometric, 2315
estimation, special procedures
differential reduction and titration:
vanadium, arsenic, antimony,
1039
chromium, 1.041 (Fig. 127)
iron, 1040
molybdenum, 1039
in aluminous materials, 23, 26 .
in alundum, 35
in antimony, arsenic, vanadium
mixtures, 1039
in brasses, bronzes,' cuprovana
dium, 1051
in ferrovanadium, 1050; A. S.
T. M., 1494
in molybdic acid, 597
in ores, 1042
fusion method, 1047
persulfate method, 1048
in steel, electrolytic, 1458
potentiometric method" 2315
ff.
with chromium absent, 1051,
1459

94

SUBJECT INDE.X

Vanadium, with chromium present, 1473,

1477, 1479
importance of methods, 1030
occurrence, 1030
phosphorus in presence of vanadium,
zirconium method, 706
preparation and solution of the sam
pIe, 1032
alloys, 1034
ores, general procednre, 1033
separation, cup ferron method, 1035
from arsenic, 1034; chrominm, 1035;
molybdennm, 1034; phosphoric
acid, 1034
solubilities, element, 1032
oxides and salts, 1033
Vanier absorption bulb, 221
Vapor pressure of water, table, 1160
Vapors, table of constants, 1166-1167
Varnish, analysis (acid 110.; ash; fixed
oils; res.; flash-pt.; sep'l1 of polymerized oils and resins) 1851
Varrentrapp-WilJ absorption bulbs, 2344
Ve'betable oils, "'ee Gil"'> fats" waxes,
Vehicles, paint
analysis, 1839
liquid, per cent of, 1840
separation of components, 1841 '
Velocity, gas flow, Pitot formulas, 1173
Venturi meter, 2342
Vermillion, analysis of, 1871
determination, in red ink, 2177
in rubber, 1991
Veronal, determination, 1953
Villavecchia test, oils, 1782
Viscometer, conversion chart, 1798 (Fig.
258)
tables, 1799
Viscosimeters, 1718, 1719, 1724; micro,
2546
Viscosity, in centipoises, sugar solution,
1725
of bituminous substances, 1518 (Fig.
190 (float test apparatus))
of nitrocellulose, 1694
of oil, determination, 1718
Engler apparatus, 1718 (Fig. 243)
l\facMiehael, 1724
Saybolt, 1721 (Figs. 244, 245)
of pulp (cuprammonium), 1884 ff.
of starch, 2157
Volatile combustible matter, in coal
apparatus, 1639
method of determination, 1638
Volatile hydrocarbons, in soap, 2040
matter, in bituminous substances, It27,
1528 (Fig. 196)
in paint, distillation of, 1841
in pigments, 1874
in sodium fluoride, 421

Volhard titration
for ~romide, 192
c]lloride, 271
C()pper, 376
cyanide, 661
iudide, 457
manganese, 561
silver, 824
Volt, depnition, 1175
VoltmetElr, vacuum-tube, 2321
Volume-i3hange, and solubility, 2584
Volumes, formulae for, 1170
unit (lOnversion table, 1176
VOlum'etric apparatus, calibration, 1219
ff.
bf direct measurement, 1220
fi'om weight of water, 1221
of burettes, 1224, 1225 (Fig, 131)
of flasks, 1223 (Fig. 130)
of pipettes, 1224
cleaning of, 1219
illu~trations of, 1219 (Fig. 129)
Volumetric methods, see Element or suhstance in Q.uestion,
Vulcani~ation coefficient, of rubber, 2014
Vulcani~ed rubber, see Rubber
W

Wagner'S solution, 209

Washing soda, analysis of, 2263
Waste and reclaim, see Rubber
Water IInalysis, general consiaeration~,
2046
micfomethods, .2545
min eral analysis
abstract of general scheme,
205'7>, 2058
lacidity, determination of, 2072
alkalinity,
determination
of,
2066
alumina, determination of, 2059
amrrionia
determination
of,
2077
bicarb.Qnate, 209q
boiler-rooill-cQntrol, 2088
boron, 164, 186
bromine, 195
calcium, determination of, 216,
2061, 2096
calcium sulfate, determination
of, 2084
carbonate, 2095
carbon dioxide, apparatus fAl"
determination,2070 (Fig.
289)
fraction as pH, 2072 (Fig.
2(0)
total, 2070
carbonic acid, determination of,
2075

SUBJECT INDEX
Water analysis, chlorine, determination
of, 280, 2075, 2095
combinations in, hypothetical,
2087
corrosion, acid waters, determination of, 2087
field assay, 2087
foaming, 2086
hardness in, 2081,
hydrogen-ion concentration, 2073
hydrogen sulfide in, 2078
hydroxide-ion
concentration,
2067, 2095
standard method, 2068
interpretation of analysis, 2085
iodide or iodine in, 459
iron, determination, 2059, 2060
irrigating waters, 2098
lead, zinc, copper, tin, 2080
lime value, 2084
lithium determination, 2065
magnesium in, 2062, 2096
ehlm:ide, 2083
manganese, determination, bismuthate method, 2063
persulfate method, 2063
mineral residue, 2077
nitrates, determination of, 2076
oil, determination of, 2078
oxygen, dissolved, est. of, 2079
phosphates, determination, 2061,
2090
preparation of solutions for,
2094
priming, 2086
rapid methods, 2093
reaction coefficients, 2086
residue, total mineral, 2077
scale, determination of, 2087
silica, determination of, 2059
colorimetric, 803
soda value, 2084
sodium and potassium, determination, 2065
softening, 2086
sulfates, determination, as barium sulfate, 2064
benzidine method, 2064
volumetric method, 2091,
2096
. traces, of heavy metals, 2080
anitary analysis, 2047, ff.
ammonia, albuminoid in, 2050
free, 2049; 2050 (Fig. 288)
chlorine, as chloride, 2054
color test, 2048
interpretation of analysis, 2055
nitrogen as nitrate, 2053
as nitrite, 2052
organic, 2051
odor test, 2048

95

Water analysis, oxygen consumed, 2053

physieal tests, 2047
residue, total solids in, 2055
taste, 2049
turbidity, 2047
Water, bacteriological examination
apparatus and materials, 2117,
2118 (Fig. 296)
bacteria, total in, 2124
brilliant-green, lactose, peptone
bile, 2122
b. coli, examination for, 2125
reaction classification, coli aerogenes group, 2126
selection of tests, 2129
subclassification of, 2125
tests, completed, 2128
confirmed, 2127
presumptive, 2126; 2127
b. typhi, examination of, 2131
subclassifications of, 2132
test, Widal, 2132
Endo's medium, 2121
eosin-methylene blue agar, 2121
Hesse agar, 2123
nutrient agar, 2120
broth, 2120
,gelatin, 2120
pH of media and solutions, 2119
Russell '8 media, 2123
sampling, for examination, 2116
(Fig. 295)
sterilization procedure, 2120
sugar broths, 2120
Widal test, b. typhi, 2132
,
examination by microscope, 2099
control of organisms, 2111
chemicals needed, 2113, 2114
importance of e,xaminatioll, 2099
microscope, 2105 (Fig. 292)
ocular micrometer (Fig. 293)
microscopical apparatus, 2112
(Fig. 294)
,
organisms, identification, 2109
odors, 2100
plates, 2107, 2108
tabular outline, 2106
-references, 2114-2115
sampling, 2102 (Fig. 291)
Sedgwick-Rafter method, 2102
Water, determination, 1338-1345. See
also Moisture.
by absorption .and weighing, 1339
(Fig. 180)
by loss in weight (indirect), 1338
centrifugal method, A. S. T. M.,
for oils,' 1761
distillation method, 1340
Bidwell-Sterling, 1342
Brown-Duvel, 1341
Dean and Stark, 1340

96

SUBJECT INDEX

Water, Kingman, 1343, apparatus (Fig.

181), 1344
Marcusson, 1343
in acetone, 2135
.in bituminous substances, 1551
(Figs. 208, 209); 1553 (Fig.
210)
finished product, 1595 (Fig.
224)
in gases, 1347
in grease, 1829.
in oils (fuel) A. S. T. M., 1758
in paint vehicles, 1841
in soap, 2028
in sodium silicate, 805
solubility of oxygen in, 677
tables, buoyancy constants, 1226
density, 1228
vapor pressure, 1160
weight, apparent, in air, 1229
Water gas, see under Gases
composition, 2397
Watt, definition, 1175
Watt-hour, 1175
Waxes, 1788
in grease, 1830
ta91e of characteristics, 1797
Weighing, in microanalysis, 2464 ff.
Weighing bottle, for acids, 643 (Fig. 68)
burette, 2202 (Fig. 304)
tubes, . for acids, 2201 (Figs. 301,
302), 2202 (Figs. 303, 304)
platinum, 2209
\Veights, correction for buoyant effects,
1233
for fire assay, 841
precision and tolerances, 1233
standa~dization of, 1230
Westphal balance, 1517
Wet assay, gold in minerals, 434
combustion, carbon determination,
226
tests, for silver, 818
White lead, corroded, analysis, /i24
sublimed, analysis, 1858
White metals, analysis .(see also Alloys;
Tin-base alloys; lead-base alloys)
determination of antimony, 84, 85,
1M

determination of arsenic, 112, 114

tin, 973
Whiting, 186,6

\Vire-bar copper, iron in, 383

Witherite, analysis of, 129
commercial valuation, 130
Wolframite, see Tungsten, ,occurrence
Wood alcohol, see Alcohol methyl; methanol
Wood's metal, analysis of, 1399
Wulfenite, molybdenum in, 597

X
Xanthate separationj cobalt and nickel,
310
Xenon, 2348; properties, 2380
Xenotime, 246
\
Xylene, use in' moisture determination,
1343
,
Y

Yellow pigments, analysis, 1874

Ytterbium, 245
Yttrium, 245
group, 245
approximate estimation, 250

z
Zimmermann-Reinhardt solution, 476
Zinc, detection, 1054
. hydrogen sulfide; ferro cyanide tests,
1054
mercuric thiocyanate, blow-pipe and
cobalt nitrate tests, 1055
estimation, general procedures, 1055,
'105"'gravimetric, electrolytic, 1059;
micro, 2513, ,2315
oxide method, 1058

phosphate, pyrophosphate,
1058
sulfate, 1058
see also micromethod, 2547
volumetric
ferro cyanide titration
diphenylamine indicator for,
1065
~ for small amounts, 1068
. in acid solution, 1060
in. alkaIjne solution, 1066
potentiometric method, 2314
rapid method, 1064
estimation, special procedures
in aluminum alloys
mercury thiocyanate. method,
45
....
oxide method, 44
in "arsenic," 110
in bismuth bronze, 1377
in brass and bronze, 1084; A.
S. T, M., 1366
in cadmium, 1078
in copper, /385
,
in galvanized sheets (coating),
1087
in German silver, 1375
in lead, pig, 1083, 1418
in lead sulfate, basic, 1856
in magnesium alloY~r 542
in ores, _!5J.6.7 .
in paint, cmnposite white, 1,868
in paper, 1922

ANDHRA PRADESH AGRICULTURAL UNIVERSITY

Regional Library. BAPATLA-522101.

Accn.
~
Call ....fJ!2
..........~.... .
_.-t.__
Tile book should be, returned on 0;

NO.S..Z ..

oewre-

the last date stamped below. Otherwise fines a~

per the library rules will be charged.