the cellulose nitrate used to prepare the

propellant is available.
There are three semicritical steps in
the basic procedure. First, there must
be sufficient potassium hydroxide to
react with the acidic products of the
hydrolysis. Each gram of propellant
consumes about 1 gram of potassium
hydroxide. Consequently, a t least 60
ml. of the hydrolysis solution should be
used per gram of propellant. Whether
sufficient alkaline solution n-as taken is
checked when the solution is acidified
after the 15-hour reaction period. The
solution should turn red upon the addition of phenolphthalein. A large excess
of hydrolysis solution should also be
avoided. Two phases are present a t the
end of the reaction. The lower phase,
in contact with the cellulose, has a higher
concentration of base than the upper.
It may be possible, consequently, to
raise the hydroxide concentration in the
lower phase so as to attack the glucoside
chain of the cellulose if a high total base
content is present after hydrolysis.

The second semicritical step is in the

washing procedure. Washing must be
thorough if the solution being filtered
comes in contact with the portion of the
walls in contact with the Gooch rubber
tubing. Because this portion is above
the suction line, it is morc difficult to
wash adequately.
Obtaining and keeping dry cellulose
is the third operation requiring care,
Cellulose should be treated about the
same as phosphorus pentoxide in regard
to atmospheric exposure.
The basic procedure required about
1 /4 hours of operator time per sample
when performed in groups of six. The
total elapsed time for six samples is
about 26 hours. This procedure minimizes equipment costs, safety hazards.
working space, and operator time.
Safety
Certain precautions
should be taken as to safety. Because
diethyl ether forms explosive mixtures
with air, all operations involving it, or
the solvent containing it should be performed under a hood. This is partic-

ularly true when the solvent is acidified,

because the gases evolved are reported
to be poisonous ( 5 ) . Fifty per cent
acetic acid will cause blisters if i t remains on the skin; hence, precautions
should be taken when this reagent is
used. The alkaline solvent, after filtration may still contain glyceryl trinitrate
and should be disposed of properly.
LITERATURE CITED

(1) Fletcher, A. N., ANAL. CHEM. 29,

1387-8 (1957).
(2) Garetio, Giuseppi, Ruffoni, Alfredo,
Ibid., 27, 400-1 (1955).
(3) Joint Army-Navv Specification JANN-244 (July 31, 1925).
(4) Kenyon, W. O., Gray, H. L., J. Am.
Chem. SOC.58. 1422-7 11936).
(5) Will, Ber. 24, 400 (1891).

RECEIVED
for review October 27, 1958.
A4cceptedFebruary 2, 1059. Division of
Analytical Chemistry, 134th Meeting,
ACS, Chicago, Ill., September 1958.
Pacific Southvest Regional Meeting, XCS,
Redlands, Calif., October 1958.

Photometric Determination of Zinc with Zincon

Application to Water Containing Heavy Metals
J. A. PLATTE and

V. M. MARCY

Hagan Chemicals & Confrols, Inc., Piffsburgh, Pa.

b The colorimetric test for zinc with

Zincon
(2-carboxy-2-hydroxy-5sulfoformazylbenzene) is r a p i d and
fairly sensitive, but lacks specificity.
Many heavy metallic ions, particularly
copper, interfere. Zinc can b e separated from some interfering substances
with an anion exchange resin and
hydrochloric acid of various molarities,
but the method i s somewhat cumbersome and time consuming. In this
method, heavy metals including zinc
are complexed by adding cyanide to
the sample. Chloral hydrate is added
to free the zinc without destroying the
other metallic complexes. The blue
zinc-Zincon complex, formed on addition of Zincon to the sample buffered
to p H 9, is measured with a spectrophotometer.

accurate method is needed

for the determination of minute
amounts of zinc in potable and industrial waters. Zinc may be present as a n
impurity or added to a water with polyphosphate to minimize corrosion by
accelerating the formation of a protective film on metal surfaces ( 2 ) .
RAPID

1226

ANALYTICAL CHEMISTRY

Although other chromogenic agents

have been used, only methods employing
dithizone (diphenylthiocarbazone) and
Zincon (2-carboxy-2-hydroxy-5-sulfoforniazylbenzene) have been widely
accepted (1, i), and many heavy-metal
ions interfere with both methods.
The aim of this investigation TTas to
devclop a rapid direct test with Zincon
in which possible interfering substances
would be niinimizcd or eliminated.
Zincon forms a 1 to 1 blue complex
nith both zinc and copper in the p H
range 8.5 to 925. At a wave length of
620 mp, the absorbance of the zinc
comples follows Beers l a r in the range
of 0.1 to 2.4 p.p.ni. of zinc (5 to 120 y
of zinc for 50 nil. of final test solution).
According to Rush and Toe (T),aluminum, beryllium, bismuth, cadmium,
cobslt, chroniiuni. iron, ninnganese,
mol:-Ldenuni, titanium, and nickel interfere in addition to copper. Ton exchange resins and complexing agents
have been tried to remove or mask
5. 7 ) .
certain interfering metals (4,
PRINCIPLE OF MODlFlCATlON

The

complexing

of

metals

cyanide, follon ed by selective demashing of zinc and cadmium cyanide coniplexes, has been used for the (ethylenedinitri1o)tetraacetic acid (EDTA) titrntion of zinc and cadmium in the presence of heavy metals ( 3 ) . Zinc can be
determined photometrically with Zincon
by the selective deniasking of the cyanide complex with chloral hydrate.
Many metals form cyanide complexes. On the addition of chloral
hydratc (6) or formaldehyde (3) to the
solution, however, zinc and cadmium
cyanide complexes are destroyed r i t h
the excess cyanide and these metallic
ions are liberated according to the
reaction:
[Zn(CN)d]-- t 4 CCll CH(OHh --c
(chloral hydrate)
Zn++
4 CCla CH(0H)CN
4 OH(3-trichloro-2-hydrosypropionitrile)
Although formaldehyde reacts more
rapidly than chloral hydrate with the
zinc complex, it liberates other metals
from cyanide complexes fast enough to
cause interferences.

APPARATUS AND REAGENTS

with

-4.11 absorbancc measurements ivere

Table 1. Change in Intensity of Zincon

Color Complex with Time

(Solution contains 10 y zinc)

Absorbance -Rending
Cysnide-

Figure 1. Effect of varying volume of cyanide

solution on test samples
containing 10 y of zinc

Chloral

Usual Zincon
Nethod

Hydrate
Rfodificntion

All tests run with 3 rnl. of chloral

hydrate solution

0.165
0.165
0.165
0.165
0.165
0.164

0.2
1
5

120
180

0.050

05

IO

15

20

25

30

300

1.80 0.155 0.145

1 . 8 0 0.155 0.154
1.80 0.155 0.155
1.80 0.155 0.156
1.80 0.153 0.360
1.80 0.151 1 . 7 0

CYANIDE SOLUTION, Ml

Table II.

made with either a Becknian Model

DU spectrophotometer, using matched
1.00-cm. cells, or a Bausch &- Lomb
Optical Co. Spectroiiic 20 colorimeter,
using '/pinch tubes. Any photometer
having a band pass of 20 mp or less a t a
wave length of 620 nip may be used.
d Becliman pH meter, Alodel G, was
used for all pH measurements.
Reagent grade chemicals were used.
Cyanide Solution. Dissolve 1.00
gram of potassium cyanide in distilled
n-ater and dilute to 100 ml.
Buffer Solution, pH 9.0. Dilute 213
nil. of 1N sodium hydroxide to about
600 i d . with distilled water. Dissolve
37.3 grams of potassium chloride and
31.0 grams of boric acid in the solution
and dilute to 1 liter.
Zincon Solution. Prepare b y dissolving completely 0.130 gram of
polvdered Zincon in 2 ml. of 114' sodium
hydroxide and dilute t o 100 ml. T h e
solution is deep red in color and is stable
for about 1 week. Zincon may be obtained from T h e LnMotte Chemical
Products Co., Chestertown, Md.
Chloral Hydrate Solution. Dissolve
10 grams of chloral hydratc in distilled
n-ater and dilute to 100 nil.
Sodium Ascorbate.
Standard Zinc Solution. Dissolve
0.2745 gram of zinc sulfatr monohydrate
in distillrd water and dilute t o 1 liter.
Dilute 10 ml. of stock solution to 100
ml. to prepare a standard containing
10 y of zinc per ml.
Solutions of Diverse Ion,i. For the
study of interferences, use solutions that
contain either 10 or I00 y of metallic ion
pcr mi
PRELIMINARY STUDY

Sufficient cyanide solution must be

added to complex all the metallic ions
that form cyanide complexes. Also,
sufficient chloral hydrate must be used
to destroy the excess cyanide and to
free the zinc from its cyanide complex.
-4s seen in Figure 1, under conditions of
the test 3 ml. of chloral hydrate solution
is adequate for the reaction with somewhat more than 1 nil. of cyanide solution. It is dcsirahle to use only a
minimum of chloral hydrate so as not to
increase the color intensity of the treated
blank.

Absorbance measurements were made

at intervals on treated samples containing zinc, with and without copper, t o
determine optimum test conditions and
t o ascertain the stability of the zincZincon color complex. The results
(Table I) show that test color in the
absence of copper is completely developed in 12 seconds and is stable for
at least 2 hours whether or not the
cyanide-chloral hydrate modification
of the Zincon test is employed. The
intensity of the color with the cyanidechloral hydrate modification is slightly
less than that with the usual Zincon
method [Rush and Yoe's Procedure A
(7) I.
When the sample contains copper,
the usual Zincon method measures all
the copper in addition to zinc. The
color reaction for both Zincon complexes
is complete mithin 1 minute. However, with the cyanide-chloral hydrate
modification the reaction M ith zinc is
complete in 1 minute and the copper is
effectively tied up as the nonreactive
cyanide complex for 2 hours. l f t e r
2 hours a n increasing amount of copper
is liberated by the chloral hydrate.
A coloi development time of 2 to 5
minutes is suggested in this procedure
to ensure that the zinc color is conipletely developed and that the other
cyanide complexes are not affected.
The use of formaldehyde in place of
chloral hydrate as a demasking agent
was unsatisfactory because it destroyed
cyanide complexes much faster than
chloral hydrate.
PROCEDURE

Transfer to a 50-nil. Erlenmeyer

flask a 10-ml. aliquot of the sample
(approximately neutral) containing 0
to 50 y of zinc. Add reagents to the
sample in the following order with
mixing between additions: 1.0 nil. of
cyanide solution, 5.0 of buffer, 3.0 of
Zincon, and 3.0 of chloral hydrate.
Measure absorbance 2 to 5 minutes
after adding the last reagent. A.s reference solution, use a blank obtained by
cftrryiiig distilled water through the
complrte procedure. Estimate zinc from

Effect of Diverse Metallic

Ions

(Solutions contain 0 and 10 y of zinc)

Apparent Zinc
Content, y
Cvanideihloral
Zinc hydrate Usual
Metallic Present, modifi- Zincon
cation method
Ion, Y
7
10.0
cu++, 10
0
0.0
30.0
20
10.0
10
...
10.0
300
10
10.0
c o + + , 30
0.0
0
19.9
9.9
30
10
...
10.0
300
10
Ni++, 30
0.0
8.4
0
7.0
10.0
30
10
...
200
10.1
10

M+++, 50

50
Fe+++, 10
20
70

Fe+T,

10
20

90
CP+++, 100
100
Cd++, 20
10
50

10
0

10
10
0

10
10
0
10
0

10
10

0.0

0.0

10.3

10.3

0.0

<o.o

10.0
0.0
9.8
10.3
0.0
10.0
0.9

...

10.0

10.8
14.2

8.5

2.3
11.4
...
0.0
0.0
2.0

14.0
24.2

a calibration curve obtained with

known amounts of zinc treated as above.
During the latter part of the investigation, the preferred reference
solution for photometric measurements
was one prepared by adding the first
three reagents t o the unknown sample,
follorred by 3 ml. of distilled water in
place of 3 ml. of chloral hydrate solution. By adding 0.5 gram of sodium
ascorbate to both sample and reference
solution, before addition of the other
reagents, interference from manganese
was greatly decreased.
EFFECT O F METALLIC IONS

Tests viere ~ u nby using the above

procedure and also the usual Zincon
method on samples containing diverse
metallic ions and 0 or 10 y of zinc.
Except for manganese (discussed later),
results are listed in Table XI. The
values show that of the metals tested,
VOL. 31, NO. 7, JULY 1959

1227

Table Ill. Effect of Manganese

Zinc Found,
Cyanide-

Metallic Ions
LTsual
Added, y
hydrate
Zincon
lln+
Zn++ modification method
10
0
4 2
12 4
50
10
34 0
34 6
50
0
3 8a
18 4
50
10
13
50
0
1 Ob
18 4
50
10
10 4b
..
0 5 gram of sodium ascorbate added to
sample.
b 0.5 gram of sodium ascorbate added
to sanmle. Reference solution for uhotometric *measurements prepared sime as
sample but 3 ml. of n-ater was added in
place of the 3 ml. of chloral hydrate solution.

Table IV.

Recovery Tests on Various

Types of Water

Zinc, P.P.M.
ReSample
Present Added covrred
1 00
1 46
Cooling water 0 50
0 16
1 00
1 18
Potable water 0 05
1 00
1 02
Effluent from 0 70
1 00
1 70
metal finishing plant
0 10
0 10
Plating waste 0 00
vater
0 00
1 00
1 00
Chromate0 02
treatedproc- 0 02
ess nater

0 10
l 00

0 11
l 00

Treated com- 0.00

bined plat- 0 00
ing and sewage n-aste

0 10
1 00

0.09
1 00

only cadmium interferes in the cyanidechloral hydrate modification. Zinc can

be determined accurately by the cyanide
-chloral hydrate modification in the
presence of many heavy metal ions even
though the concentration of these ions
is many times grcater than that of zinc.
I n contrast, serious errors are produced
with most metals in the usual Zincon
method. I n the presence of ferric ion
the color intensity of the sample is less
than that of the blank, producing lorn
results. Aluminum in the concentration tested did not interfere with either
procedure. Cadmium produces high
results Tvvith either procedure but the
error is less with the cyanide-chloral
hydrate modification.
Table I11 shows that manganese interferes seriously with both the usual
Zincon method and its modification.
Hydroxylamine hydrochloride and salts
of polyphosphates, ferrocyanide, ferricyanide, fluoride, persulfate, iodine, and
ascorbate !?-ere tried in an attempt to
eliminate the effect of manganese.
Only sodium ascorbate decreased the
error caused by nianganesc. The influence of manganese can be decreased
further if the reference solution for
photometric nieasurement is prepared
by treating another aliquot of the sample with 0.5 gram of soclium ascorbate,
followed bj. the reagents in the procedure, but replacing the 3 ml. of chloral
hydrate solution with 3 ml. of distilled
water. This technique compensates for
interfering substances because the value
of the reference solution not only includes
the color of the Zincon reagent but also
the effect of interfering metals not complexed with cyanide. Conditions of the
test are the same for both sample and
reference solution up to the point at

which chloral hydrate is added to the

sample to release the zinc which then
combines lyith Zincon. This special
reference solution is particularly recommended for color and turbidity in the
sample. All steps in the procedure can
be completed within 5 minutes.
APPLICATION OF METHOD
Recovery tests were run OII various
types of m t e r samples to check the
accuracy of this method.
Sodium
ascorbate and the special reference
solution discussed previously were used.
Table IV shows a maximum recovery
error of 4% for a 1 p.p.in. zinc addition
and 10% for a 0.1 p.p.m. zinc addition.
Zinc can he determined rapidly and
accurately in various types of water by
the Zincon method using the proposed ascorbate-cyanide-chloral hydrate
modification. The method should also
find use for the analysis of materials
other than water.
LITERATURE CITED

(1) Ani. Public Health Assoc., New \-o;k,

Standard Methods for the Esaminatlon
of Water and Sewage, 10th ed., p. 212,
1955.
(2) Hatch: G. B., Corrosion 11, 461t
(1955).
(3) Kinnunen, J., Merikanto, B., Chemist
Analyst 41,76 (1952).
(4) hicCal1, J. T., Davis, G. K., Stears,
T. W.,
AS-AL.CHEM.30, 1345 (1958).
(5) Maier, R. H., Kuykendall, J. R.,
Chemist Analyst 47,4(1958).
(6) Pribil, R.?Collection Czechoslou. ('bent
Communs. 18, 783 (1953).
( 7 ) Rush, R. M.,
Yoe, J. H., ANAL.CHEM.
26, 1345 (1984).
RECEIVEDfor review October 14, 1958.
hccepted March 4, 1959. Division of
Water, Sexage, and Sanitation Chemistry,
134th Meeting, .4CS, Chicago, Ill.,
September 1958.

Titration of Ketimines in Glacial Acetic Acid

P. L. PlCKARDl and F. A. IDDINGS
University of Oklahoma, Norman, Okla.

The titration of ketimines in acetic

acid solution b y perchloric acid has
been developed to replace the more
tedious and time-consuming Dumas
and Kjeldahl methods. Accuracy and
precision are within 1% when either a
potentiometric or visual end point i s
used for individual samples containing
as little as 0.06 meq. of ketimine.
The potentiometric end point used a
platinum-glass electrode system in
conjunction with a Beckman Model H
pH meter. The familiar S-shaped
titration curves are obtained. The
violet to blue color change of crystal
violet occurred a t the potentiometric

1228

ANALYTICAL CHEMISTRY

end point in the titrations. Few impurities other than amines or ammonia
interfered.

have been analyzed for

yield and purity by micro-Kjeldah1 and micro-Dumas determinations
of nitrogen content (6,6-8). Although
some modifications n ere made (6), the
analyses were inconvenient for occasional samples and time-consuming for
even a fen samples. A more convenient method requiring less time per
sample and less exhaustive techniques
TTBS desired. This paper is thr result
LTIJIINES

CJf a study of the feasibility of titration

of ketimines by perchloric acid in

glacial acetic acid solutions (referred to
here as acetic acid).
Because of the instability of many of
t,he ketimines in water solutioiis, a
nonaqueous solvent for the titrations
was desirable. Acetic acid was chosen
for its leveling &cct upon weak organic
bases and esccllcnt solvent, properties
(9)

1 Present address, Chemical Division,

Celanese Corp. of America, Clarkwood,

Tes.