934 ANALYTICAL CHEMISTRY

0.8% (absolute) lower than the hexahydrate standard. A fourth photometer is used, the range can also be extended upward, and
sample, dissolved in water, evaporated to dryness, and baked with some increase in accuracy, by the use of the 0.1 selector
for about 10 minutes, gave a very turbid suspension in ethanol,
and was therefore unsuitable. switch for measuring transmittancies below 11% ( 1 ) .
The rather high concentration range of the method for cobalt
The formation of the sparingly soluble basic salt when the is a good illustration of the fact that spectrophotometric methods
monohydrate is heated above 100” C. precludes this method of of analysis are not necessarily limited to the determination of
preparing standards, and indicates that care must be taken, in small amounts of constituent, but can apply to concentrations
analyses, to evaporate just to dryness, avoiding baking of the comparable to those used in gravimetric and titrimetric methods
residue. (1, 6 ) .
For estimrting water in ethanol by the use of cobalt chloride In comparison with gravimetric and titrimetric methods for
reagent, it would be immaterial whether hexahydrate or mono- cobalt, the proposed spectrophotometric method is somewhat
hydrate is used, inasmuch as the calibration and determination more rapid, and gives results of comparable precision and ac-
are made by the use of the same standard solution. curacy. The 1-nitroso-2-naphthol method requires filtration,
If desired, the determination of cobalt can be made using washing, and ignition to constant weight, all of which are time-
ordinary commercial (about 95%) ethanol instead of absolute consuming. The nitrite-permanganate method requires 12 to
ethanol; in this case the calibration curve, for measurements a t 24 hours’ standing for precipitation of the hexanitritocobaltiate,
655 mp, is nearly parallel to the curve of Figure 2, but covers a followed by filtration, washing, dissolving, and back-titrating.
cobalt concentration range about ten times as high. Spectro- The electrodeposition method requires fuming down with sulfuric
photometrically, solutions of cobalt chloride in 95% ethanol acid, and an electrolysis time of 2 hours. In contrast, the pro-
have considerable components of both blue and red; solutions posed colorimetric method, although requiring evaporation of
containing about 5000 p.p.m. (5.00 mg. per ml.) of cobalt are the solution just to dryness, is very rapid from that point on to
visually blue, and with decreasing cobalt concentration the the measurement of the desired constituent.
solutions show gradations through bluish purple to reddish
purple, Measured a t 655 mp (the absorption maximum of LITERATURE CITED
blue ethanol solutions), the solutions in 95% ethanol showed con-
siderable deviations from Beer’s law, but in such a way as to (1) Ayres, G. H., ANAL.CHEM.,21, 652 (1949).
(2) Brode, W. R., 2.physik. Chem., A187,211 (1940).
increase the analysis accuracy; at 515 mp (the absorption maxi- (3) Kolthoff, I. M., and Sandell, E. B., “Textbook of Quantitative
mum of pink dilute aqueous solutions; see Figure 3, curve 8) Inorganic Analysis,” rev. ed., p. 603, New York, Macmillan
the measurements followed Beei’s law, but a calibration curve Co., 1946.
based on these measurements is flat and if used for analysis would (4) Lundell, G. E. F., Hoffman, J. I., and Bright, H. A,, “Chemical
Analy5is of Iron and Steel,” pp. 339-42, New York, John Wiley
give larger relative error. & Sons, 1931.
The specifications of range and accuracy given herein for the ( 5 ) hlellor, J. W., “Comprehensive Treatise on Inorganic and
cobalt determination apply to the measurements made against a Theoretical Chemistry,” Vol. XIV, New York, Longmans,
Gieen and Co., 1935.
blank, using the Coleman Model 10-S spectrophotometer with (6) Ringbom, A., 2 . anal. Chem., 115, 332 (1939).
1.30-cm. absorption cells. As with other spectrophotometric (7) Toporescu, E., Compt. rend., 192, 280 (1931).
methods, for a given wave length and cell thickness the range of (8) Treadwell, F. P., and Hall, W. T., “Analytical Chemistry,”
the cobalt determination can be extended upward by measuring 9th English ed., Vol. 11, p. 197, New York, John Wiley &
against a standard solution of concentration somen hat lower Sons, 1942.
(9) Winkler, C., J . prakt. Chem., [ l ]91, 209 (1864).
than that of the solution measured; the standard is so chosen
that the transmittance ratio is near the optimum-theoretically RECEIVED Kovember 8, 1948. Condensed f r o m a thesis submitted by Betty
Vining Glanville t o the faculty of the Graduate School of the University of
37y0, although the analysis accuracy is almost as good at trans- Texas in partial fulfillment of the requirements of the degree of master of arts,
mittancies from about 20 to 60%. TT’hen a Beckman spectro- August 1948.

Colorimetric Determination of Capsaicin in

Oleoresin of Capsicum
HORACE NORTH, General Control Laboratory, Dodge & Olcott, Znc., Bayonne, N. J.

A method for the colorimetric determination of capsaicin in oleoresin of cap-

sicum has been developed in which the readily available vanillin is employed
for the standard solution in place of capsaicin.

T HE most important constituent of red pepper is the pungent

principle known as capsaicin, discovered by Thresh (8) in
1876. I n 1898 Micko (6) sholyed that the substance had the
method for the determination of capsaicin content. The organo-
leptic method formerly official in the United States Pharma.
copoeia and later in the Sational Formulary has now been dis-
properties of a weak phenol and contained one methoxyl group. carded entirely. Tice (9) brought out a colorimetric method
He found also that with an alcoholic solution of platinic chloride based on Fodor’s ( 3 ) reaction in which capsaicin gives a blue
an odor of vanilla was developed on standing. In 1919 the struc- color with vanadium oxytrichloride. -4study of this method by
ture of capsaicin was established by Kelson ( 7 ) ,who showed it to Hayden and Jordan ( 5 ) showed that the results were unreliable.
be the vanillyl amide of isodecenoic acid. However, some of Tice’s recommendations relative to the isolatio~
Because the pungency of different varieties of peppers varies of capsaicin, modified to meet the requirements of an analytical
enormously, there has long been a demand for an accurate procedure, have been incorporated in the method described here.
 V O L U M E 21, NO. 8, A U G U S T 1 9 4 9 935

Folin and Denis (5)devised a solution of phosphotungstic- Make up to the mark with acetone, mix thoroughly, and
filter through a dry double filter. The filtrate should be per-
phosphomolybdic acid which gives a blue color with phenols, and fectly clear. .
applied this reagent to the determination of vanillin ( 4 )in vanilla Pipet 50 ml. of the clear filtrate into a 250-ml. beaker marked at
extracts. This is now an official method ( 1 ) of the Association of 20 ml. and evaporate on top of a steam bath (not directly over the
Official Agricultural Chemists. steam) a t a temperature not over 65" C., using a small thermom-
A colorimetric method of analysis of capsaicin using the pure eter as a stirring rod, until the volume of liquid is reduced to 20 ml.
By this treatment the acetone is removed from the solution and
drug as a standard would be unsatisfactory, as the preparation the crude capsaicin separates as an oily sediment. Solutions of
of pure capsaicin is a difficult, tedious, and very unpleasant task. capsaicin should be heated as little as possible and a t as low a tem-
This isolation of pure capsaicin has been circumvented by the dis- perature as possible.
covery that vanillin, which like capsaicin contains a phenolic Cool the liquid to room temperature, add 10 ml. of 0.5 N
sodium hydroxide, and stir until the oily sedimen't has dissolved.
hydroxy grouR in the same relative position, serves just as well as Pour the solution into a 250-ml. Squibb separatory funnel, and
capsaicin for the standard solution. The molecular weight of wash the beaker with two further 5-ml. portions of 0.5 N sodium
vanillin is 152, while that of capsaicin is 305. For practical pur- hydroxide and finally with two 5-ml. portions of water, pouring the
washings into the separatory funnel. Now add to the funnel 5.0
poses the latter may be considered double the former, so that 5 grams of sodium hicarbonate and 150 ml. of petroleum ether,
ml. of a solution containing 0.5 mg. of vanillin are equivalent to shake moderately 15 minutes, and let stand until the layers sepa-
1.0 mg. of capsaicin. Although this relationship is assumed, we rate sharply (overnight, if necessary). The amount of petroleum
have in hand a practical means for comparing the pungencies of ether is sufficient for 1.0 gram of a normal oleoresin. I n special
cases it may be necessary to use a larger quantity of solvent.
different oleoresins of capsicum. Draw off and reject the lower layer and carefully filter the
Before applying the colorimetric test, however, it is necessary upper layer into a clean 250-ml. Squibb separatory funnel, wash-
to isolate the capsaicin present in the sample to be tested, in a ing the separatory funnel and the filter with small portions of
sufficient degree of purity, in order to eliminate other substances of petroleum ether. It is essential that the yellow substance which
' separates a t this point be carefully excluded from the filtrate.
a phenolic nature which also give a blue color with the phospho- Shake the petroleum ether solution with 10 ml. of 0.5 N sodium
tungstic-phosphomolybdic acid reagent. I t is believed that the hydroxide, add 10 drops of 957, ethyl alcohol, and without further
number of steps necessary to accomplish this purpose has been shaking let stand until the layers separate sharply. Filter the
reduced to the minimum possible under the circumstances. lower layer into a 50-ml. volumetric flask and extract the petro-
leum ether further with three IO-ml. portions of water, passing the
Duplicate results obtained by the application of this method are extractions successively through the filter into the flask. Fill up
in excellent agreement, as evidenced by the following table: the flask with water to the 50-ml. mark and mix thoroughly. This
solution should be nearly colorless. The concentration remains
Sample S o . % Capsaicin the same as the 50 ml. of clear filtrate originally taken for evapo-
1 ration.
2 Pipet 5 ml. of the solution into a 50-ml. volumetric flask and
3 into another 50-ml. volumetric flask pipet 5 ml. of standard vanil-
lin solution. To each flask add from a pipet 5 ml. of the p h 0 5
photungstic-phosphomolybdic acfd reagent, allowing it to flow
down the neck of the flask in such a way as to wash down
the solution that may be on the sides of the flask. RIix con-
The substitution of vanillin for capsaicin in the present analyti- tents of flasks by rotating and after 5 minutes dilute contents to
cal procedure suggests that similar procedures may be applicable 50 ml. with saturated sodium carbonate solution. Mix thor-
to some other colorimetric determinations in which the substances oughly by inverting the flasks several times and shaking and then
place the flasks in a shaking machine until 30 minutes have
to be determined are not readily obtainable in a pure state and in elapsed since the phosphotungstic-phosphomolybdic acid reagent
which other substances of suitable composition are so obtainable was first added to the solutions. This thorough shaking is neces-
and can be used for the preparation of the standard solutions. sary in order to precipitate the sodium phosphate completely and
prevent the filtrate from becoming hazy while the solution is being
read in the colorimeter. Filter the solutions through dry double
ANALYTICAL PROCEDURE filters and compare the blue colors of the clear solutions without
Special Reagents. Ultrasene. This is purified, deodorized delay in a colorimeter.
kerosene much more suitable for analytical work than kerosene
itself. Kerosene can be used, if it is treated with sulfuric acid and With samples poor in capsaicin, there may be a slight hue dif-
redistilled ference between the standard solution and the test solution be-
Acetone, 607, by volume.
Phosphotungstic-Phosphomolybdic Acid. To 100 grams of cause then the traces of color carried through from the oleoresin
pure sodium tungstate and 20 grams of phosphomolybdic acid have a greater influence on the total color. This does not inter-
(free from nitrates and ammonium salts) add 100 grams of sirupy fere in any way with the usefulness of the method. In this labora-
phosphoric acid (containing 857, H3POI) and 700 ml. of water; tory it is customary for two observers to read the color and their
boil over a free flame for 1.5 to 2 hours; then cool, filter if neces-
sary, and make up with water to a volume of 1 liter. An equiva- results uniformly agree within one or two tenths of the color-
lent amount of pure molybdic acid may be substituted for the imeter scale. It is essential that the blue solutions be perfectly
phosphomolybdic acid. clear. Ordinarily the standard blue color is set a t 20, but if the
Standard T'anillin Solution. Dissolve 0.1 gram of vanillin in test solution is pale it may be necessary to set the standard a t 10
sufficient distilled water to make 1000 ml. This solution must be
freshly prepared each day. or even 5. After a reading is made, the positions of the cups
Procedure. JTeigh 1.0 gram of oleoresin red pepper in a small should be reversed and another reading made. The average of
beaker and transfer to a 125-m1. Squibb separatory funnel by these two readings is used for the calculation. The zero points on
solution in 20 ml. of ultrasene, using the ultrasene in portions. the colorimeter should be checked and corrected if necessary be-
Dissolve 1.0 gram of sodium chloride in 80 ml. of 60% acetone (by
volume) and nash out the beaker Lvith 20 ml. of this solution, in fore the instrument is used.
portions, transferring the washings to the separatory funnel. If it is a question of determining capsaicin in the spice, 5 to 10
Shake the funnel sufficiently to keep the liquids well mixed and grams of the ground material are extracted with acetone or
continue this gentle shaking for about 5 minutes. On standing, ether in a Soshlet extraction apparatus and the extract is tested as
the mixture wparates within 2 or 3 minutes into two sharply de-
fined layers but the lower layer is always cloudy. Draw the above described.
lower layer into a 125-ml. Squibb separatory funni.1 and continue
the extraction of the solution of oleoresin in like manner, using the ACKNOW LEDGM E S T
balance of the acetone solution in 20-ml. portions. T o the com-
bined extractions add 5 ml. of ultrasene and shake gently for a few Acknowledgments are due to V. H. Fischer, vice president of
minutes Let stand 1 hour to separate. Draw off the still hazy
lower layer into a 100-nil. volumetric flask containing 0.5 gram of Dodge & Olcott, Inc., who assigned this problem for study, t o
Filter-Cel, cork the flask, and shake 0.5 hour in a machine. Herman Wachs, director of research, who supervised the prepara-
936 ANALYTICAL CHEMISTRY

tion of this material for publication, and to Thaddeus Ptaszynski, (3) Folin and Denis, J . Bid. Chem., 12, 239 (1912).
(4) and Denis, J . Ind. Chem.*4, 670 (1912).
Philip Catanzaro, and Thomas hledwick who performed a large
(5) Hayden and Jordan, J . Am. Pharm. Assoc., 30, 107 (1941).
part of the experimental work. (6) Micko, 2..Vahr. Genussm., 1 , 818 (1898); 2, 411 (1899).
(7) Nelson, J . Am. Chem. Soc., 41, 1115 (1919).
LITERATURE CITED (8) Thresh, Pharm. J . Trans. ( 3 ) , 7, 21, 259, 473 (1876-77); 8, 187
/ 1 Q77-7Qj
,LU, I I Y,.

(1) Aasoc. Offic. Agr. Chemista, “Official and Tentative Methods of (9) ~ i A ~ J ,. pharm.,
~ , 105, 320 (1933).
Analysis,” 6th ed., p. 366, 1945.
(2) Fodor, 2. Cntersuch. Lebensm., 61, 94 (1931). RECEIVEDOctobrr 9, 1948.

Rapid Identification of Manganese Dioxide, Ores

GLENN A. MARSH’ AND HUGH J. McDON..ILD2
Illinois Institute of Technology, Chicago, I l l .

Alanganese dioxide ores obtained from different method for testing ores would be of great interest.
geographic locations and as a result of different Through the use of the pulse polarizer, it was pos-
methods of preparation differ in their depolarizing sible to differentiate in a few minutes between the
ability when used in the common Leclanchk t >pe of poor and good ores in a set of samples. The ores
dry cell. Ores are commonly tested for their quality were rated independently on the basis of test cells,
by constructing an actual cell and making suitable and the comparisons between predicted and actual
measurements on its current capacity and shelf life. depolarizing ability were, in the majority of cases,
Because such tests are time-consuming, a rapid found to be good.

I S A dry cell of the common Leclanche type, a depolarizer or

oxidizing agent is used to provide the cathodic reaction.
Manganese dioxide is ordinarily used for this purpose, but the de-
platinum cylinder 0.5 inch in diaiiieter is filled with ore and
tamped until tightly packed. I n the top of the cylinder an elec-
trolyte is added, in which an “inert” electrode is immersed. This
polarizing characteristics of different commercial batches vary electrode, which merely completes the electrical circuit, is a fine
considerably. Ordinary chemical and x-ray analytical method? iron wire. A thin calomel half-cell is then lorered into contact
have not been used extensively in the evaluation of manganese with the ore surface. The experimental setup is shoivn i n
dioxide ores. The method presented here shows promise of pro- Figure 1.
viding a rapid and reliable means of identifying good and poor Either the cathodic or anodic polarization may be studied with
battery depolarizer ores. the apparatus, but the cathodic polarization is of interest in this
The depolarizing characteristics of manganese dioxide ores are parti ular case. By cathodic polarizat’ion is meant the change in
ordinarily evaluated on the basis of dry cell tests, carried out ill the electrode potential of the manganese dioxide when electrons
special test cells, which are rather complicated and time-consum- are forced into it from the external circuit.
ing to construct. Rapid evaluation of the characteristics of man- The voltage applied is constant for each pulse, and the time
ganese dioxide ores, from the standpoint of not only immediate during which the ore surface is polarized is between 0.05 and 0.1
capacity but also shelf life, is highly desirable from the standpoint second. Experiments showed that the curves obtained with
of battery makers and ore suppliers. although these character- platium alone are long but very narrow, radically unlike those ob-
istics have not been studied, the work described here indicates tained with the ores. On this basis, the results must be attributed
that i t may be possible to establish rapidly the depolarizing almost entirely to the depolarizing characteristics of the ore.
characteristics of an ore as measured by the initial capacity of the The electrical discharge obtained from the pulse polarizer is
(Bell. standardized to the extent that curves obtained are, for the most
THE PULSE POLARIZER part, reproducible in minute detail, and curves obtained with a
The method described here involved use of the pulse polarizer,
developed during 1947. The instrument employs a system of
electronic circuits to polarize an electrode over a brief time inter-
val. The polarization and depolarization a t the surface of the ELECTRICAL
electrode are recorded continuously on a high-speed strip chart I
MECHANISM
(Brown Instruments Division, Minneapolis-Honeywell Regulator I
Co., Philadelphia, Pa.; single record; full scale, 5 mv.) and
show up as a curve which is distinctive for each set of conditions.
The polarizing circuit consists of a condenser (125 mfd.) which is
charged to 310 volts. Discharge of this condenser brings about CALOMEL
the polarization a t the electrode surface. The polarization poten-
tial is amplified by means of an electronic direct current voltage
amplifier, and then recorded. The pulse polarizer has been sur- n
f HALF- CELL

cessfully employed in the field of corrosion research (%,$).

bill I I
EXPERIMENTAL TECHNIQUE
A technique has been developed which permits rapid study of
the depolarization characteristics of manganese dioxide ores. -4 AMPLIFIER
1Present address, The Pure 011 Company, Research and Development
Laboratories, Northfield, Ill. LPLATINUM
2 Prpsent address, Loyola C-nirersity, Stritch School of Medicine, Chicago,
Ill. Figure 1. Experimental Setup