D E I = A46 mV).

The D P P membrane was still highly perm-

selective to hydrogen ions. Response toward lead ion by Table V. Response to Various Cations of D P P and DEI
both the membranes was still appreciable. However, in the Membranes without Dithizone
absence of dithizone, the membranes showed more perm- DPP membrane DEI membrane
selectivity to K+, Na+, and Ca2+. potential, mV potential, mV
Acid-base and precipitation titrations were attempted with Cation 10e3M 10-2M AmV 10-3M 10-*M AmV
membranes without dithizone and the results are shown in Ag+ 23 59 36 64 102 38
Figure 18. In the titration of H N 0 3 os. KOH, the DPP mem- H+ 144 261 117 105 144 39
Na+ 22 21 -1 57 67 10
brane registered a sharp minimum at the end point. The K+ 30 73 43 62 78 16
DEI membrane yielded a less sharp break. The potential Pb2+ 43 79 36 120 140 20
span of the titrations was considerably increased when dithi- Ca 2+ 12 20 8 58 75 17
zone was present in the membrane matrix.
These results clearly demonstrate that dithizone contrib-
utes in some way to increase the response to several ions and
ions may make it valuable for measuring ions for which elec-
the response ultimately governs the general morphology of
trodes are not available, under well defined conditions. Ki-
the curves. The membrane with dithizone when precondi-
netic measurements involving changes in ion concentrations or
tioned by immersing in dilute Pb(NO& solution, becomes
valence might provide an example.
cherry red. Probably, the lead-dithizone chelate formed in
Varying the composition of the membrane changes its
the network of the plastic controls the mobility of various
relative response to different ions and it is apparent that both
ions and helps in attainment of an equilibrated diffusion
the plasticizer and the chelating agent play important roles
potential at the membrane-solution interface.
in this regard. The DDP membrane results in the best de-
In summary, the described DPP membrane is useful for
fined titration curves.
monitoring potential changes in cases where the absolute
magnitude of the potential is unimportant, as in potentio-
ACKNOWLEDGMENT
metric titrations. It functions well in all types of titrations,
regardless of the nature or valence of the ion and to a limited Gary L. Stucky provided valuable discussion.
extent in pure nonaqueous solvents. Because of the bi-ionic
nature of the membrane potential, its use in direct potentio- RECEIVED for review December 22,1969. Accepted December
metric measurements will probably be restricted to situations 18, 1970. A post doctoral research fellowship to S. L. from
in which calibration curves can be prepared under strictly Miles Laboratories, Inc., Ames Company Division, is grate-
defined conditions. The fact that it responds to nearly all fully acknowledged.

Use of Syringaldazine in a Photometric Method for

Estimating “Free” Chlorine in Water
Robert Bauer and Chauncey 0. Rupe
Ames Company, Dioision of’Miles Laboratories, Inc., Elkhart, Ind. 46514

A colorimetric method for estimating “free” chlorine precautions were taken. But only the a-tolidine test has
(hypochlorous acid, hypochlorite ion) in water has been
developed using syringaldazine buffered at pH 7 as been widely accepted as a convenient field test, even though
the chromophoric agent. The reagent is sensitive to it is known to react with various chloramines ( 2 ) . With
hypochlorite but i s insensitive to bound chlorine such most natural water, this presents a minor problem; but in
as chloramines, thus making it an important reagent water containing ammonia and other contaminants, the
for free chlorine in the presence of chloramines. a-tolidine test may indicate adequate chlorination where little
Syringaldazine reacts with chlorine on a mole for
mole basis. Under these conditions, the molar ab- or no unbound chlorine exists.
sorptivity ( E ) for the reaction i s about 65,000 at X = 530 The wide use of chlorine in bleaches and disinfectants has
nm. The developing color does not follow strictly further increased the need for new chlorine tests, especially
Beer’s law, but in dilute solutions of syringaldazine those tests which measure free chlorine.
and chlorine, the standard curve (absorbance vs.
chlorine concentration) is a straight line deviating only In searching for a convenient test for free chlorine, several
slightly from the origin. compounds which produce colored oxidation products were
studied. Since Bradley (3)and Gretton (4) have demonstrated
DIFFERENCES in the bactericidal properties of “free” (hy- the usefulness of syringaldazine and vanillinazine in the
pochlorous acid or hypochlorite ion) and “bound” chlorine estimation of hydrogen peroxide, it was decided to investigate
has focused attention on the need for a test for free chlorine in
water. Nicolson ( I ) has studied nine different methods seek-
( 2 ) “Chlorine, Its Manufacture, Properties, and Uses,” J. S.
ing an ideal colorimetric method. Several of these methods Sconce, Ed., Reinhold, New York, N. Y.,1962, pp 461-482.
were found to adequately measure free chlorine if certain (3) Wm. Bradley, Diagnostic Composition, U. S. Patent 3,233,974
(1966).
(4) N. B. C. Gretton and J. T. Rees, Test Composition and Device
(1) N. J. Nicolson, Anaiyst, 90, 187-198 (1965). for Detecting Glucose, U. S. Patent 3,290,228 (1966).

ANALYTICAL CHEMISTRY, VOL. 43, NO. 3, MARCH 1971 421

 "
0 2 4 6 8 IO
0' I
400 500 600 700 TIME (rnin.)
WAVELENGTH ( n m ) Figure 2. Rate of fading of color produced by the
syringaldazine-chlorine reaction in 0.025M phosphate
Figure 1. Absorption spectrum of color produced by the buffer
syringaldazine-chlorine reaction in 0.025M phosphate buffer,
PH 7 0-0 pH6
X-X pH 6.5
0-0 pH7
these compounds as to their use in determining chlorine in
water. A crude spot plate check of the two reagents showed
syringaldazine in phosphate buffer p H 3-8, to be quite sensitive The diluent was chlorine-free distilled water previously
to chlorine, while vanillinazine under the same conditions treated with chlorine and ultraviolet irradiation according
was rather insensitive even t o chlorine concentrations as high to Nicolson (1) to remove any preexisting chlorine demand.
as 5 ppm. Because of the low sensitivity of vanillinazine for The standards were prepared just before use because of the
instability of the more dilute chlorine solutions.
chlorine, it was dropped from further study.
Procedure. SPECTROPHOTOMETRIC STUDIES. The color re-
EXPERIMENTAL sponse of syringaldazine with chlorine was studied spectro-
photometrically in phosphate buffer a t p H 3, 5, 6, 7, 8, and 9.
Reagents. PHOSPHATE BUFFERS. A stock solution of In each setup, 1 ml of phosphate buffer was placed in each of
0.1M disodium phosphate (analytical grade) was prepared three small flasks, which had been repeatedly washed with
from which aliquots were taken and adjusted to the desired distilled water to remove any trace of detergents or material
p H (3-9) with either hydrochloric acid or sodium hydroxide that might react with chlorine. To one was added 10 ml
as needed. of chlorine water containing 1 ppm chlorine and 1 ml of
SYRINGALDAZINE SYNTHESIS.Syringaldazine was prepared 0.01 % syringaldazine; t o the second flask, 10 ml of chlorine
as follows: Syringaldehyde (Aldrich) in the amount of 3.6 free water and 1 ml of syringaldazine; and to the third, 10 ml
grams (0.02 mole) was dissolved in 50 ml of methanol. Two of water. Approximately 1 minute was given for color de-
hundred milliliters of 0.1N sodium hydroxide were added with velopment, and the spectrum was run using the solution in the
stirring. A pale yellow solution developed. Hydrazine third flask in the reference cell. At all p H levels, a n intense
dihydrochloride 1.05 gram (0.01 mole) dissolved in 50 ml violet color was produced. Both color development and
of distilled water was added with stirring. Within 5 minutes, stability were p H dependent. At p H 3, the color developed
yellow crystals began to appear. Gentle stirring was con- quickly and faded rapidly to a n almost colorless solution.
tinued at room temperature for another 2 hours. As the p H was increased from 3 to 6, both color development
The solution was then filtered by suction, and the precipi- and fading became slower. At p H 7, the color developed
tate washed on the filter with a small amount of distilled rapidly but fading remained slow. At p H 8, the color de-
water. veloped rapidly and fading also was rapid. At p H 9, the
The precipitate was redissolved in 400 ml of hot ethanol color was less intense and quickly faded to a bright yellow
and filtered. Upon cooling to room temperature, needle- solution. The spectrum obtained at p H 7 is shown in Figure 1.
like yellow crystals began to form. After 2 hours at room The violet color produced had a n absorption maximum
temperature, the solution was filtered. The precipitate was at X = 530 nm. Changes in p H did not influence the position
air-dried to remove the ethanol, and then placed in a vacuum of the absorption peak but did affect its height. These
oven at 50" overnight. changes possibly were related to color development and fade
The total yield of syringaldazine was 2.8 grams, 61 rate.
theoretical. Corrected melting point was found to be 210- The rate of color development and fading was studied more
211 "C. carefully over a p H range of 6-7 where the best color response
STOCKSYRINGALDAZINE SOLUTION. A 0.1 solution of occurred. Syringaldazine, 0.005 %, was prepared by placing
syringaldazine was prepared by dissolving, with the aid of 2.5 ml of 0.1 alcohol stock solution in a 50-ml volumetric
heat, 100 mg in reagent alcohol containing 90% ethanol, 5 Z flask, with 25 ml of 0.05M phosphate buffer of the desired
methanol, and 5 isopropanol and diluting to a total volume pH, and diluting to volume with reagent alcohol. One
of 100 ml with the same reagent alcohol. At room tempera- milliliter was added to 10 ml of water containing 0.5 ppm
ture, this is close to a saturated solution. A 0.01 % solution chlorine. A reagent blank was used as the reference solution.
of the reagent was prepared by diluting this solution 1 :10 Each solution was read at X = 530 nm a t 30- and 60-second
with reagent alcohol. and at 1-minute intervals for a total of 10 minutes.
CHLORINE STANDARD.Chlorine standards containing 0-2 The results are shown in Figure 2. At p H 6, three minutes
ppm expressed as chlorine (CLJ were prepared from either were required for maximum color development. At p H 6.5,
stock sodium hypochlorite solution or chlorine water, whose 60 seconds were required and 30 seconds at p H 7. The
chlorine concentration was established by the thiosulfate rate of fading was about the same for all three test solutions.
method. Fading appeared to set in almost immediately and to proceed

422 ANALYTICAL CHEMISTRY, VOL. 43, NO. 3, MARCH 1971

 I .o

0.8

w
0.6
a
m
a
0
9 0.4
i” 0.2

0.5 1.0 2 .o 3 .O
0
0 0.2 0.4 0.6 0.8 I0
MOLE RATIO OF SYRING / C H L O R I N E CHLORINE, pprn
Figure 3. Effect of increased mole ratio of syringaldazine
Figure 4. Standard curve for the chlorine-syring-
to chlorine (Cl,) upon color intensity
aldazine reaction
Lower curve: Chlorine held constant at 0.705 X 10-5M
Upper curve: Chlorine held constant at 1.41 X 10-5M 1 ml syringaldazine, 0.008 in 0.025M phosphate buffer,
pH 7; 10 ml chlorine water. X = 530 nm

at a slow but constant rate. By buffering at pH 7, advantage

can be taken of the rapid color development and the slow fade
rate. Table I. Comparison of the Syringaldazine Method of
Between pH 6 and 7, the syringaldazine reagent changed Estimating Chlorine in Water with the o-Tolidine (OT) and the
from pale yellow to an intense yellow solution having an o-Tolidine-Arsenite (0-TA) Methods
absorption peak at 410 nm. This peak was completely
absent at pH 6. The color change could be reversed by lower- Water
sample Chlorine content, ppm
ing the pH. This was taken to mean a change in the ioniza- No. 0-T 0-TA Syring.
tion of syringaldazine had taken place, possibly involving the
protonation of the azine group in the molecule. 1.12 0.98 1.01
STOICHIOMETRY OF THE CHLORINE-SYRINGALDAZINE RE- 0.45 0.395 0.40
0.97 0.72 0.87
ACTION, Syringaldazine reagents were prepared in phosphate
0.94 0.78 0.87
buffer from the stock solution as described above. Concen- 1.02 0.98 1.06
trations ranged from 2.75 X to 27.5 X lO+M. Each 1.12 0.91 0.95
solution was checked with chlorine (Cl,) concentrations of 0.99 ... 0.91
0.5 ppm (0.705 X 10-jM) and 1 ppm (1.41 X 10-jM). One 0.85 0.70 0.73
milliliter of the reagent was placed in a small Erlenmeyer
flask and 10 ml of chlorine water were added. After 1
minute, the absorbance was read in the spectrophotometer
at X = 530 nm against its appropriate reagent blank. These The method used in preparing the standard curve was
results are presented in Figure 3 with the mole ratio of syring- compared with the spectrophotometric procedures of the o-
aldazine to chlorine plotted against absorbance. tolidine and o-tolidine-arsenite methods. The comparisons
On each curve, the break occurred at a molar ratio of were made on laboratory tap water. The o-tolidine and the
approximately 1, indicating a definite stoichiometric relation- o-tolidine-arsenite methods were carried out at room tem-
ship in the reaction between chlorine and syringaldazine. perature according to “Standard Methods for the Examination
The intensity of the color resulting from the reaction was of Water and Wastewater” (5). The results are shown in
quite strong, having a molar absorptivity ( E ) of approximately Table I.
65,000 where the reactants were present in equal molar The o-tolidine-arsenite method measures free or unbound
amounts. chlorine, while o-tolidine measures total chlorine (5). The
DETERMINATION OF FREE CHLORINE IN WATER. A working differences between these values were quite small, indicating
chromogenic reagent containing 0.008 (22 X molar the amount of bound chlorine in the tap water was also small.
syringaldazine was prepared as follows: Eight ml of 0.1% Most of the syringaldazine test values fell between the two
syringaldazine in reagent alcohol were placed in a 100-ml o-tolidine test values, thus showing good agreement with the
volumetric flask containing 15-20 ml of alcohol; 50 ml of o-tolidine test.
0.05M phosphate buffer (pH 7) were added, and the total In order to determine whether the syringaldazine test
volume was brought to 100 ml with alcohol. A clear yellow was measuring total or free chlorine, samples of distilled water
solution resulted. This reagent contained enough syringalda- were prepared containing 1 ppm chlorine and varying amounts
zine in 1 ml to react with the chlorine in 10 ml of water con- of ammonium sulfate added to bind portions of the chlorine
taining 1 ppm chlorine. The working solution was prepared as chloramines.
fresh each day to ensure maximum color development. All samples were analyzed for free chlorine by the syring-
One milliliter of the reagent was placed in a series of small aldazine, o-tolidine-arsenite, and amperometric titration
Erlenmeyer flasks to which was added 10 ml of water contain- methods. They were also analyzed for total chlorine by the
ing 0, 0.1, 0.2, 0.4, 0.6, 0.8, and 1.0 ppm chlorine. One
minute was allowed for color development. Each solution
was read in a spectrophotometer at X = 530 nm using the (5) “Standard Methods for the Examination of Water and Waste-
chlorine free sample as the blank. The resulting standard water, 12th Ed.,” Amer. Public Health Assoc., Inc., 1790 Broad-
curve is shown in Figure 4. way, New York, N. Y . 10019, 1965, pp 93-102.

ANALYTICAL CHEMISTRY, VOL. 43, NO. 3, MARCH 1971 423

 samples encountered, the apparent quantity was small, 0.05
Table 11. Estimation of Chlorine in the Presence of Ammonia ppm or less.
by the Syringaldazine, o-Tolidine-Arsenite, a-Tolidine, and The agent responsible for the trace reaction with syring-
Free and Total Amperometric Methods (ppm) aldazine was not identified. The addition of Fe(II), Mn(II),
NHBadded Syring- 0-TA, Ampero, 0-T, Ampero, Cu(II), Ca(II), and Mg(I1) up to 10 ppm produced no color
as(NH&SO, aldazine free free total total response. Fe(II1) at 1 ppm gave nqcolor, but in a concen-
0.00
tration of 10 ppm a n absorbance reading equivalent to 0.01
1.03 1.03 1.03 1.03
0.05 0.70 0.91 0.63 1.03 0.97 ppm chlorine was obtained. It is unlikely that Fe(II1) was
0.10 0.41 0.73 0.35 1.00 0.95 the causative agent, since the water sample contained less
0.20 0.15 0.66 0.11 1.00 0.99 than 3 ppm iron.
0.50 0.00 0.15 0.01 0.94 1.00 The reproducibility of the syringaldazine test was checked
in tap water samples containing 1, 0.4, and 0.05 ppm free
chlorine. The results are given in Table IV.
Table 111. Estimation of Chlorine in Tap (T) and Swimming At all three chlorine levels, the new method showed a
Pool (SP) Waters by the Syringaldazine and Amperometric high degree of reproducibility.
Methods
Syring- Ampero, Ampero, DISCUSSION
aldazine, free, total,
Sample PPm ppm PPm The oxidation of syringaldazine by chlorine appears to be
1 (TI 0.00 0.00 0.00 on a mole to mole basis (Figure 3), possibly involving a shift
2 (TI 1.12 1.12 1.32 of two electrons similar to the oxidation of hydroquinone to
3 (TI 0.30 0.26 0.38 quinone. The reaction may be represented as follows:
4 (SP) 0.20 0.24 1.12
5 (SP) 2.48 2.50 3.52 CH,-0
6 (SP) 0.03 0.04 0.76 \
7 (SP) 1.14 0 .96a 1.25
End point erratic.

Table IV. Reproducibility of the Syringaldazine Test

Sample 0-T 0-TA
No. test test Syring. test t HCI
1 1.02 0.98 1.056 i 0.011 (10)"
2 0.45 0.395 0.399 i 0.013 (11) CH,-6 0-CH
3 0.16 0.07 0.055 i 0.011 (10)
a Number of trials with syringaldazine test. The reaction produces an intense color having a narrow
absorption maximum at 530 nm, with a small shoulder at
approximately 495 nm. The position and shape of the ab-
sorption curve is not altered by change of p H (3 to 9) nor by
o-tolidine and amperometric methods. The results are diluting the colored component.
given in Table 11. The amount of color produced depends upon the concen-
The syringaldazine values agreed well with those obtained tration of both syringaldazine and chlorine. By keeping the
with the amperometric titration method for free chlorine. concentration of syringaldazine low (10-5M),and the chlorine
Values by the o-tolidine-arsenite procedure were slightly concentration 1 ppm or less, the amount of color produced
higher. This might be expected since these samples were can be made chlorine dependent; although it does deviate
run at room temperature rather than at 1 "C. In general, slightly from Beer's law. The deviation is so slight that good
the data indicate that the syringaldazine method is responding quantitation can be made by using either a standard curve or
to free chlorine and not to chloramines.
a chlorine standard.
The syringaldazine method was compared with the ampero-
metric titration method, free and total, using chlorinated Syringaldazine does not react with chloramines which
tap and swimming pool waters. These results are summarized are found as common contaminants in swimming pool water.
in Table 111. Samples 1 through 3 represent three different The importance of this lies in the difference in the bactericidal
sources of tap water; samples 4 through 7 were swimming properties of free and bound chlorine. Free chlorine in
pool water samples. concentrations of 0.5-1 ppm has been shown to be an effective
Results obtained with the syringaldazine method agreed bactericidal agent, whereas chloramines are rather ineffective
well with those for free chlorine by amperometric titration, (6).
except for Sample 7 . Situations can and do arise where adequate chlorination
The titration of this sample for the free chlorine gave a n is indicated by the o-tolidine procedures, but little or n o
atypical end point, which may account for the observed dis- germicidal activity is present. The o-tolidine-arsenite pro-
crepancy. This was not encountered in any of the other titra-
tions. cedure, in addition to being cumbersome, may indicate free
In several samples of tap water from the same source, chlorine levels in excess of those actually present unless the
a substance other than chlorine was found which gave a trace entire reaction is run at or near 1 "C (5).
reaction with syringaldazine. The presence of such a com- In cases where little or no bound chlorine exists, the o -
ponent was first suspected when the values obtained with tolidine and syringaldazine methods are in close agreement
the syringaldazine method fell between those obtained with (Table I); but as the amount of chloramines increases, the
the o-tolidine and the o-tolidine-arsenite methods. This discrepancy between the two tests becomes greater (Table 11).
suspicion was substantiated in a sample of tap water rendered In this particular case (Table II), o-tolidine is reacting with
free of chlorine by heat and ultraviolet irradiation. On
this sample, syringaldazine gave a reading of 0.05 ppm ex-
pressed as chlorine, and the o-tolidine test read 0.0. In all (6) C. T. Butterfield, Pub. Health Rep., 63,934-940 (1948).

424 ANALYTICAL CHEMISTRY, VOL. 43, NO. 3, MARCH 1971

both free and bound chlorine and syringaldazine with only (Tables I1 and 111). The necessity of using a blocking agent,
the free, as indicated by the amperometric procedure. such as arsenite or controlled temperatures, is eliminated.
In some waters, where a large percentage of the chlorine The reagent itself should be prepared fresh each day from a
exists in the bound form, it is quite possible more free chlorine stock alcohol solution of 0.1 x
syringaldazine. Buffered
may be indicated by the o-tolidine-arsenite test than is actually solutions left standing at room temperature for 24 hours or
present. The effect of bound chlorine upon the free chlorine more showed a gradual decrease in color production. This
values can be minimized by chilling the water before adding does not represent a major problem, since the working reagent
the o-tolidine and arsenite reagents ( 2 ) . For reasons of con- can be prepared quickly from stock solutions of syringaldazine
venience, this restriction is rarely used in practice. Either and phosphate buffer.
the test is run at room temperature or no attempt is made at
estimating the free chlorine content.
With the syringaldazine procedure, free chlorine can be RECEIVED
for review September 14, 1970. Accepted Decem-
measured easily without interference from chloramines ber 21, 1970.

Intrinsic End-Point Errors in Precipitation Titrations with Ion

Selective Electrodes
Peter W. Carr
Department of Chemistry, University of Georgia, Athens, Ga. 30601

Investigations on the effect of ionic interferences in liquid junction potentials, and spurious interferences which
potentiometric precipitation titrations with ion selec- change the activity of the analyte. The cost of this increased
tive electrodes are reported. When the inflection point
of the titration curve is used to locate the end point, accuracy and precision is a decrease in the speed and con-
serious errors (>l%) can result which are often outside venience of measurement.
the range of the ultimate analytical precision (0.1%) It is widely recognized that ion selective electrodes are not
achievable in such analyses. Results are presented in perfectly selective (7,8)and that in actual practice the absolute
terms of p, a dimensionless variable related to the solu- accuracy of direct potentiometric measurements with such
bility product and the initial analyte concentration, and
b, a parameter related to the concentration and selec- electrodes depends critically upon the presence of potential
tivity ratio of interfering ions. I n general, errors result determining foreign ions. It is generally considered that
from finite selectivity ratios and from the intrinsic the presence of these interfering ions in a potentiometric
absence of symmetry in the titration curve. The titration ultimately limits the magnitude of the “end point
errors are negligible only in the limit of very sparingly
soluble materials or in the absence of any interfering break” and, concomitantly, the precision of measurement.
ions. The fact that such interferences can seriously influence the
position of the inflection point of a titration curve and thus
THE PAST SEVERAL YEARS have witnessed a renaissance of its analytical accuracy has not yet been widely appreciated
potentiometry as an analytical tool (1-3). Most of this is (9). As will be shown, such errors often amount to several
due to the development of both anion and cation selective per cent, well outside the range of the precision (hO.1-0.2 x)
electrodes which measure the activity rather than the concen- routinely obtainable by potentiometric titration.
tration of the species of interest. For thermodynamic and Although there has been some discussion (10, 11) as to
kinetic studies, activity is of principal importance (4). Ana- evaluation of the extent of electrode selectivity, there seems
lytical measurements, however, are mainly concerned with to be fairly general agreement that if the liquid junction and
determining the concentration of analyte. Concentration reference electrode potentials are constant, then the potential
can be measured from a calibration curve, a plot of cell of a cell containing an ion selective electrode is well repre-
potential us. the decadic logarithm of concentration, from sented as follows:
the measured activity and knowledge of activity coefficients,
by precision null point potentiometry (5, 6), or by titration. E = const + RT In ( y x C x + B )
-
ZXF
(1)
Titration to the inflection point of the potential-volume
curve will generally be more precise and accurate than any In this equation (12), species X is the ion of interest (gen-
of the above direct measurement techniques. This method
minimizes errors due to variations in activity coefficients,
(7) J. W. Ross, Jr., “Solid State and Liquid Membrane Ion Selective
Electrodes,” in Ion Selective Electrodes, R. A. Durst, Ed.,
(1) G. A. Rechnitz, Accounts Chem. Res., 3,69 (1970). National Bureau of Standards, Special Publication 314 (1969).
(2) “Ion Selective Electrodes,” R. A. Durst, Ed., National Bureau (8) G. A. Rechnitz, Chem. Eng. News, 43 (25), 146 (1967).
of Standards, Special Publication 314 (1969). (9) M. Whitfield and J. V. Leyendekkers, Anal. Chim. Acta, 45,383
(3) “Glass Electrodes for Hydrogen and Other Cations: Principles (1969).
and Practices,” G. Eisenman, Ed., Marcel Dekker, New York, (10) M. S. Frant, ANAL.CHEM., 40,457 (1968).
N. Y . ,1967. (11) A. Shatkay, ibid., p 458.
(4) E. A. Guggenheim, J . Phys. Chem.. 34.1758 (1930). (12) G. A. Rechnitz, “Analytical Studies on Ion Selective Mem-
( 5 ) H. V. Malmstadt and J. D. Winefordner, Anal. Chim. Acta, 20, brane Electrodes,” in Ion Selective Electrodes, R. A. Durst,
283 (1959). Ed., National Bureau of Standards, Special Publication 314
(6) R. A. Durst, ANAL.CHEM., 40,931 (1968). (1969).

ANALYTICAL CHEMISTRY, VOL. 43, NO. 3, MARCH 1971 425