990

Anal. Chem. 1983, 55, 990-992

Electrogravimetric Determination of Cobalt, Nickel, and Copper in the Presence of

Chloride Ion
Judith F. Owen, C. Stuart Patterson,* and Gregory S. Rice
Department of Chemistry, Furman University, Greenville, South Carolina 296 13

Electrogravimetryis recommended as the quickest and one

of the most precise methods for determination of Co, Ni, and
Cu in aqueous solutions of their divalent salts. In the course
of isopiestic vapor pressure studies on various salts of these
cations, we have attempted to use the standard literature
methods ( I , 2). Our requirements are for analyses that yield
results that are consistently better than 0.1 % relative standard
deviation.
The recommended electrolytic bath solutions contain sulfate
ion and we have found the standard procedures to be equally
satisfactory when perchlorate is the counterion. However, we
have invariably obtained poor cathodic deposits and erratic
results when chloride ion is present. The poorly adhering
deposits are apparently related to the erratic results which
are variously attributed to the deposition of oxides with the
metal, to the deposition of oxidized species on the anode, or
to the codeposition on the cathode of platinum species produced by attack of chlorine atoms on the anode ( 3 , 4 ) . Some
of these problems are said to plague the cobalt and nickel
procedures even in the absence of chloride because of the high
ammonia concentration.
Our anodes did lose weight when electrolyzed in chloride
solutions and the results were high (based on analysis for
chloride and assumed 1:2 stoichiometry).
Two approaches have been recommended for ameliorating
the adverse effects of chloride-masking and removal.
1. Masking. Some standard procedures call for the addition of a small amount of an anodic depolarizer such as
hydrazine or hydroxylamine even when no chloride is present
if the plating bath is ammoniacal. Larger amounts are recommended when chloride is present. These compounds either
are oxidized at a lower potential than chloride or else react
with the nascent chlorine atoms thereby preventing their
attack on the anode.
2. Removal. Obviously if chloride can simply be removed
the problem is removed with it. However simple this may
appear in principle, it can be rather difficult in practice since
not only must removal be complete but the remaining sample
must be retained quantitatively for the subsequent electroanalysis. Some procedures call for preliminary electrodeposition of the metal followed by its dissolution in HN03,
heating with H2S04to drive off the HN03, and finally redeposition of the metal. Others suggest quantitative precipitation of the metal, e.g., by addition of carbonate followed by
filtration and dissolution of the precipitate in H2SO4 to form
the electrolytic bath.
We first attempted to use the masking approach and found
that these agents did significantly reduce the weight loss by
our anodes but the results remained unacceptably erratic and
the deposits poor. Of seven analyses attempted (three on a
CuClzsolution and two each on solutions of NiClz and CoCl2),
five gave precision poorer than 0.1% relative standard deviation and a sixth had to be aborted because of an obviously
nonadherent deposit.
Because of the complexity and time-consuming nature of
the various separation methods, we set about development
of an alternative approach for quantitative removal of chloride
prior to electrolysis which avoids the necessity for quantitative
transfers of precipitates, filtrates, or deposits. This report

presents procedures for such an alternative method for preparing electrolytic bath solutions for Co, Ni, and Cu starting
with their divalent chlorides. The objectives of the new
procedure were to remove chloride completely and to produce
a resulting solution containing only electrolytes compatible
with the optimum electrolytic bath solutions. It was also
desirable to effect this exchange as quickly and easily as
possible with a minimum of transfer operations which threaten
recovery.
It is often recommended that excess HN03be removed from
electrolytic baths by fuming (5), and it has long been known
that similar exhaustive heating with nonvolatile acids will drive
out excess C1- as HC1 (6, 7). However, no procedure seems
to have been published which makes use of this reaction to
eliminate chloride interference in the analyses of interest. The
method we have devised involves addition of a measured
amount of concentrated HzS04directly to the weighed sample
(in our case, an aqueous solution of the divalent metal chloride) in the electrolytic beaker, volatilization of chloride as
HC1 gas by careful heating, and the addition of the components required to form bath solutions of the composition
recommended in the literature. The electrogravimetric
analyses then proceed by the standard methods ( I ) .
The method is stoichiometrically straightforward since the
recommended baths are comprised of aqueous mixtures of the
metal sulfate, H2S04,(NH4)2S04,and NH40H in varying
proportions for the different metals (excess acid for copper;
excess ammonia for cobalt and nickel). The relative concentration of each component and the total electrolyte concentration, which is critical for control of the current-voltage
relationship, are predetermined by selection of the sample size
and the quantity of HzS04used in displacing the chloride.
Procedures for Removal of Chloride. Before attempting
quantitative determinations, we carried out qualitative tests
to establish the minimum treatment required for quantitative
removal of chloride. It was also established that acceptable
deposits could be formed on electrolysis of the resulting solutions. The procedures developed on the basis of these testa
are given below.
1. Procedure for Electrogravimetric Determination
of Cu in Aqueous CuClZ. A sample containing approximately 0.3 g of Cu is weighed from a weight buret (to 0.1 mg)
into a 100-mL tall-form (electrolytic) beaker and 2.3 mL of
concentrated H2S04is added slowly to avoid spattering due
to overheating. Upon addition of HzS04,brown and white
precipitates form with the evolution of fumes. The mixture
is placed on a hot plate and heated to near boiling or until
vapor condensation can be seen on the beaker. The internal
condensation is then continuallyremoved by flaming the sides
of the beaker with a Bunsen burner to prevent return of
chloride to the beaker by reflux. When the mixture is initially
heated the brown and white solids dissolve leaving a green
solution. After about 2 min, the solution turns a pale green.
(Although all the chloride can be removed without the addition
of HN03, the addition of 1 mL of HN03 while the solution
is pale green helps remove HCl as an azeotrope, boils off any
nitrites, and improves plating.) In subsequent transitions, the
solution turns from pale green to blue green, blue green to
light blue, and finally to light blue with a white solid present.

0003-2700/83/0355-0990$01.50/00 1983 American Chemical Society

ANALYTICAL CHEMISTRY, VOL. 55, NO. 6, MAY 1983

Table I. Quantitative Tests of Procedures

Cu(I1) in CuCl, Solution
12 replicate analyses
1.2240 mmol of Cu(I1)
av concn =
g of solution
re1 std dev = 0.06%

sample
no.
1
2
3
4
5
6
7
8
9
10
11
12

g of

solution
3.596 7
4.9628
3.1324
4.0!377
3.3 08 3
3.3444
3.2710
3.3756
3.7 7 2 5
3.5496
3.4687
3.11.91

g of

mmol of Cu(II)/
g of solution

cuo

1.2229
1.2239
1.2229
1.2247
1.2244
1.2248
1.2244
1.2247
1.2244
1.2236
1.2236
1.2230

0.2795
0.3852
0.2473
0.3189
0.2574
0.2603
0.2545
0.2627
0.2396
0.2762
0.2697
0.2424

Ni(I1) in NiC1, Solution

10 replicate analyses

av concn =

1.9149 mmol of NilII)

\-

g of solution

re1 std dev = 0.04%

sample

g of

no.

solution

g of Nio

mmol of Ni(II)/
g of solution

1
2
3
4
5
6
7
8
9

2.3625
2.6052
2.6204
2.8191
2.4832
2.7186
2.3413
2.4821
2.4626
2.6911

0.2657
0.2929
0.2983
0.3168
0.2792
0.3056
0.2632
0.2791
0.2767
0.3024

1.9159
1.9153
1.9133
1.9 144
1.9154
1.9150
1.9161
1.9156
1.9142
1.93.43

10

Co(I1) in CoCl, Solution

4 replicate analyses
1.8531 mmol of Co(II\
g of solution
re1 std dev = 0.05%
-

av concn =

sample

g; of

no.

solution

1
2
3
4

2.4673
2. !57 7 9
2.5320
2.5 46 8

g of

coo

0.2690
0.2817
0.2764
0.2780

mmol of Co(II)/
g of solution
1.8538
1.8542
1.8523
1.8522

Upon the formation of white solid (enough to loosely cover

the bottom of the beaker) most of the C1- is removed and
heating should be continued for 5-10 min to assure the complete removal of chloride.
When removal is complete, the solution is diluted by addition of 100 mL, of distilled water and the electrolysis is
carried out by the standard procedure (8).
2. Procedure for Electrogravimetric Determination
of Ni in Aqueous NiC1,. A sample equivalent to -0.3 g of
Ni is weighed to 0.1 mg into a 100-mL electrolytic beaker from
a weight buret and 4.1 mL of concentrated H2S04is added
slowly down the sides of the beaker. Upon the addition of
the concentrated H2S04,fumes are evolved and a yellow
precipitate forms. Then the mixture is placed on a hot plate
and heated to boilling or near boiling until the solid changes
from yellow to white, and finally to light green. Any condensation is continiually removed from the sides of the beaker

991

by flaming the sides with a Bunsen burner. Heating should

be continued for 5 min after the indicated color changes have
taken place.
After chloride is removed, the solution is diluted with 65
mL of distilled water; then 35 mL of concentrated NH,QH
is added, the solution is heated to -70 OC, and the electrolysis
is carried out by the standard procedure (9) using a copperplated platinum cathode. (Some standard procedures call for
addition of hydrazine but none was used in these nickel
analyses.)
3. Procedure for Electrogravimetric Determination
of Co in Aqueous CoCl,. A sample containing -0.3 g of
cobalt is weighed to 0.1 mg into an electrolytic beaker from
a weight buret and 4.1 mL of concentrated H2S04is poured
slowly down the sides of the beaker. This addition should be
made very carefully since a violent reaction can result. Fumes
are evolved and the solution changes from red to blue and blue
to purple or blue with a pink solid. The mixture is placed
on a hot plate and heated to near boiling. Any condensate
in the beaker should be continuously removed by flaming the
sides of the beaker with a Bunsen burner. After being heated,
the mixture changes from a light purple solution with a pink
solid to a dark, bluer solution with a pink solid. The solution
is heated for 10 min or until no more fumes are visible and
the boiling appears to have ended.
When the C1- is removed, the mixture is diluted with 65
mL of distilled water, and 36 mL of concentrated NHIOH and
0.5 mL of 85% hydrazine hydrate are added. The solution
is heated to -70 *@and the electrolysis carried out by the
standard procedure (IO).

RESULTS
Multiple replicate analyses were carried out on a solution
of each of the three metal chlorides using the procedures given
above. Twenty-six determinations were made; the results of
these analyses are tabulated in Table I to demonstrate the
consistency of the methods.
It was intended to run 12 replicates of each solution but
two nickel samples were lost to a known mechanical error while
the initial samples ofthe cobalt solution gave poor deposits
and somewhat poorer precision (although still superior to the
standard masking procedure) than the four samples reported. It was discovered that these poor deposits were
evidently caused by the presence of residual traces of chloride.
The procedure used for the four samples reported is that given
above and differed from the earlier version in the addition
of the prescribed 0.5 mL of hydrazine and extra ammonia
when deposition seemed abnormally slow as indicated by the
solutions color.
We conclude from these results that the depolarizer (hydrazine) works quite well for traces of chloride although, as
reported above, it is ineffective when stoichiometric concentrations of chloride ion are present. It may be that residual
chloride is not the problem here at all and that the depolarizes
serves only to prevent anodic corrosion by the NH, solution.
Hydrazine is often a recommended component of the conventional baths for both nickel and cobalt analyses (but not
copper which is plated from acidic medium) although our
results seem to indicate that it is not necessary in the case
of nickel if chloride is absent.
That the precision obtained is a reasonable indication of
accuracy is supported by the following data: chloride analyses
(gravimetric as AgC1) gave chloride-to-metalratios of 2.O0Ob
and 2.002, for the NiCl, and CuCl, solutions, respectively. A
similar analysis of the CoCl, solution gave a chloride-to-cobalt
ratio of 1.993j. We do not believe that this less satisfactory
result is a consequence of the chloride problem but rather
simply confirms that the standard electrogravimetric method

@92 ANALYTICAL CHEMISTRY, VOL. 55, NO. 6, MAY 1983

tends to give slightly high results for cobalt unless the current
density is kept very low (11). Apparently this problem can
be overcome by addition of a measured amount of nickel to
be codeposited with the cobalt (12) but we discovered this fact
too late to conduct an adequate test of that approach. Preliminary results seem to confirm this effect.
CONCLUSIONS
The results reported justify the conclusion that the method
presented shows six distinct advantages over the literature
methods.
I. Time. Prior to the beginning of the electrodeposition
no procedure requires more than 20 min of preliminary
treatment of the bath solutions.
2. Chloride Removal Is Complete. After the electrolytic
deposition of the metal, each bath was tested for the presence
of chloride, and the tests were either entirely negative or, at
most, resulted in a trace indication of chloride. None of the
deposits formed from solutions prepared by the procedures
prescribed displayed any of the undesirable characteristics
that result from the presence of chloride.
3. Ease of Bath Preparation. During pretreatment, no
extraneous solutes or solvents are needed which must later
be removed. Thus, the bath is made directly without the use
of any separation processes such as deposition, redissolution,
filtration, or elution.
4. Minimum Sample Handling Required. Transfer
operations invariably threaten recovery, but since the sample
is weighed directly into the beaker and the entire procedure
carried out in the same container, this source of error is removed.
5. Flexibility. Other acids of low volatility may be substituted for the HzS04if the anion is a more desirable counterion than S042- (e.g., HC104 or H,PO,) for other metals.

6. Self-Indicating. An unanticipated series of diagnostic

color changes occurred upon the addition of concentrated
H2S04and during the subsequent heating and gradual removal
of chloride. Consequently, the procedure is self-indicating so
that after brief experience,easily obtained by qualitative trial
runs, the pretreatment can be carried out with little concern
for careful timing or temperature control.
Finally, the ultimate success of the procedure is indicated
by the precision of the 26 analyses of the three solutions which
yielded consistent precision with a pooled relative standard
deviation of 0.05%. In addition, the experimental ratios of
anion to cation indicate that this precision is a reliable indicator of the accuracy of the results.

Registry No. CuCl,, 7447-39-4; NiClZ,7718-54-9; CoC12,

7646-79-9;Co, 7440-48-4;Ni, 7440-02-0;Cu, 7440-50-8;sulfuric
acid, 7664-93-9.
LITERATURE CITED
Erdey, Laszlo "Gravlmetric Analysis", Part 11, Belcher, R., Gordon, L.,
Eds.; Pergamon: London, 1965; Internatlonal Series of Monographs
on Analytlcal Chemistry, Vol. 7.
Bassett, J.; Denney, R . C.; Jeffery, G. H.; and Mendham, J.; "Vogel's
Textbook of Quantitative Inorganic Analysis", 4th Ed.; Longman: New
York, 1978.
Reference 1, p 76.
Reference 2, p 533.
Reference 2, p 537.
Ephraim, F. "Inorganic Chemistry", 5th ed.;Thorne, P. C. L., Roberts,
E. R., Eds.; Interscience: New York, 1948.
Reference 1, p 77.
Reference 1, p 77.
Reference 1, p 390.
Reference 1, p 406.
Reference 1, p 407.
Reference 1, p 391.

RECEIVED
for review August 30, 1982. Resubmitted January
19, 1983. Accepted February 17, 1983.