Study of cadmium electrochemical deposition in sulfate medium

Article abstract of DOI:10.1149/1.1393309

The cadmium electrochemical deposition process from sulfate medium was studied by means of different electrochemical techniques in both stationary and nonstationary diffusion regimes. The kinetics of the electrochemical reduction of cadmium on solid cadmium electrodes was examined and the kinetic parameters are presented, as well as the diffusion coefficient derived from the different techniques. Temperature has an important effect on the cadmium reduction kinetics, and the activation energy of the process was evaluated. The electrochemical deposition of cadmium is a complex process due to the coexistence of adsorption and nucleation processes; the adsorbed electroactive species appears to be Cd⁺², and we propose a mechanism for cadmium electrodeposition on solid cadmium electrodes.

Full text of DOI:10.1149/1.1393309

Journal of The Electrochemical Society, 147 (3) 1031-1037 (2000)
1031
S0013-4651(99)07-120-7 CCC: $7.00 © The Electrochemical Society, Inc.
Study of Cadmium Electrochemical Deposition in Sulfate Medium
T. Montiel,a,z O. Solorza,a,* and H. Sánchezb
aCINVESTAV-IPN, Departmento de Química, Col. San Pedro Zacatenco, Del. G. A. Madero, C.P. 07360, México D.F.
bUAM-Iztapalapa, Departmento, de Química, Area de Electroquímica, Col. Vicentino, Del. Iztapalapa, C.P. 09340, México, D.F.
The cadmium electrochemical deposition process from sulfate medium was studied by means of different electrochemical tech-
niques in both stationary and nonstationary diffusion regimes. The kinetics of the electrochemical reduction of cadmium on solid
cadmium electrodes was examined and the kinetic parameters are presented, as well as the diffusion coefficient derived from the
different techniques. Temperature has an important effect on the cadmium reduction kinetics, and the activation energy of the
process was evaluated. The electrochemical deposition of cadmium is a complex process due to the coexistence of adsorption and
nucleation processes; the adsorbed electroactive species appears to be Cdϩ2, and we propose a mechanism for cadmium elec-
trodeposition on solid cadmium electrodes.
© 2000 The Electrochemical Society. S0013-4651(99)07-120-7. All rights reserved.
Manuscript submitted July 29, 1999; revised manuscript received November 28, 1999.
sϪ1 have been obtained. For cadmium reduction on solid cadmium
electrodes, also a two-step mechanism involving the existence of
Cdϩ as an intermediate in the adsorbed state21 has been assumed.
However, a cadmium reduction mechanism involving one step has
also been proposed.22 On the other hand, Hedrich23 suggests that the
cadmium discharge mechanism depends on the applied current den-
sity, and he proposes that a one-step mechanism prevails at low cur-
rent density while a two-step mechanism takes place at high current
density. As is clear, there is no agreement about the cadmium reduc-
tion mechanism, and it remains an open question.
In this paper we report and discuss our results on cadmium depo-
sition, with the notion that this information will be helpful toward
baths for the preparation of CdTe. Specifically, the electrochemical
deposition of Cdϩ2 on solid cadmium electrodes from aqueous solu-
tion using different electrochemical techniques in both stationary
and nonstationary diffusional conditions is described. The investiga-
tion mainly focused on the kinetic and mechanistic aspects of the
cadmium deposition process and the associated phenomena coupled
to the charge-transfer reactions. In order to avoid electrode contam-
ination and uncertainty in the electrochemical data, this study was
carried out by using cadmium electrodes prepared in situ.
Cadmium is a chemical element of great importance because of
the wide variety of its applications such as in rechargeable batteries1
and solar energy capture devices.2 In the case of solar energy conver-
sion devices, cadmium is combined with other elements to obtain
semiconductor materials. These materials can be obtained from a
broad variety of physical chemical, and electrochemical techniques.3,4
Electrochemical techniques to prepare cadmium-based semicon-
ductor materials are principally of empirical character because of the
lack of precise information concerning the individual electrochemi-
cal behavior of the components in the electrochemical solution of
interest. Thus it is clear that a formal study about the alloy formation
process must be, first, focused on the study of the individual deposi-
tion process for each one of the alloy components in the same chem-
ical medium, just as it has been pointed out in Kroger’s5 and Lan-
dolt’s6 works.
Cadmium has been electrochemically studied for some time, and
the literature about the matter is quite broad. Most of the research
work has been accomplished by using either mercury or mercury
amalgam working electrodes and applying different techniques such
as ac polarography, potential step, current step, faradic impedance,
and others.7,8 In these kinds of electrodes the cadmium deposition
process, starting from either perchlorate or sulfate baths, is a fast
process (ko > 10Ϫ2 cm sϪ1). It is considered that values for the reac-
tion rate constant are abnormally high and, additionally, there is not
a clear explanation concerning the broad variety of values for the
charge-transfer coefficient. In some work9,10 the effect of the elec-
trode nature on the kinetic behavior of cadmium reduction has been
investigated and it has been demonstrated that mercury not always
behaves as an inert substrate, and in some cases reaction kinetics is
favored by the presence of mercury. Therefore the high values ob-
tained for ko are due to the use of noninert working electrodes and
consequently the polarographic results cannot be taken as a founda-
tion when solid electrodes are employed. In works carried out on
either mercury or mercury amalgam the cadmium reduction mecha-
nism has been analyzed.11-14 In this case, most of the authors assume
a two-step reduction mechanism involving the existence of an ab-
sorbed monovalent intermediate11.12; however, results do not always
support this assumption.
Experimental
Two solutions were prepared by using chemical reagents of ana-
lytic grade (Aldrich) and triply distilled/deionized water. A solution
of 0.2 M CdSO4и8H2O, in 0.5 M K2SO4 was used only for prepara-
tion of the working cadmium electrodes while for cadmium electro-
chemical reduction studies a solution of 0.01 M CdSO4и8H2O in
0.5 M K2SO4 was used. In both electrolytes H2SO4 was added until
the pH was 2.2.
Two conventional (80 and 250 mL) three-electrode electrochem-
ical glass (Pyrex) cells were used, the larger one being employed in
the electrochemical quartz crystal microbalance (EQCM) experi-
ments. The electrolytic solutions were purged with nitrogen for
30 min prior to each experimental series and then kept under flow-
ing nitrogen during the experiments.
Two kinds of working electrodes were used: a platinum rotating
disk electrode (RDE) (4 mm diam) embedded in a cylindrical Teflon
holder (13 mm diam) and for the microbalance sensor, a thin crystal
quartz disk (13.2 mm diam). Both electrodes were covered with a
cadmium thin film for immediate use as a working electrode. The
reference electrode was Hg/Hg2SO4/K2SO4 (saturated) (SSE, 0.64 V
vs. NHE at 25ЊC) and all potentials are referred to this electrode. As
the counter electrode a platinum gauze was employed. The working
temperature was controlled at 25 Ϯ 0.1ЊC. When cadmium reduction
was studied at different temperatures, the reference electrode was
placed out of the cell, maintained at constant temperature (25ЊC),
and connected to the cell by means of a glass tube (fritted glass in
one end) filled with supporting electrolyte.
For the cathodic deposition of Cd on solid electrodes only few
works are published.8,15-19 In this case, the cadmium deposition pro-
cess was studied by electrochemical techniques such as current step,
faradic impedance, and others. It has been established that the reduc-
tion process is affected by selective adsorption of anions, which
come from the electrolytic bath, and they cause that the deposition
process be kinetically less favorable than on mercury elec-
trodes.15,17,20 Using this, ko values in the range of 10Ϫ2 to 10Ϫ5 cm
* Electrochemical Society Active Member.
z E-mail: tmontiel@mail.cinvestav.mx
Downloaded on 2014-11-21 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

1032
Journal of The Electrochemical Society, 147 (3) 1031-1037 (2000)
S0013-4651(99)07-120-7 CCC: $7.00 © The Electrochemical Society, Inc.
works in the literature7,8 consider the cadmium reduction as a re-
versible process. The curves in Fig. 1 show a peak in the measured
current density, which is normal in the voltammetric technique under
nonstationary diffusion conditions. Such a peak is due to the elec-
troactive species depletion on the electrode surface, and the magni-
tude of the cathodic peak current density (jpc), was increased as v did
too. Besides this, a gradual displacement of the peak potential
toward more negative values was observed as v was increased. In this
sort of j-E curves hydrogen evolution is only observable at potentials
lower than Ϫ1.55 V.
Most measurements were performed using a standard electro-
chemical potentiostat (EG&G PAR, model 273 A) controlled by a
computer and coupled to a rotating disk electrode system (Pine,
model AFMSRX). For some experiments an EQCM (Maxtek Inc.,
model PM600) with crystals operating at 5 MHz was used. The
potential at the microbalance sensor was controlled by a potentiostat
(Voltalab model PGZ301).
Preparation of the working electrodes.—The platinum rotating
disk electrode was mechanically polished using alumina of 0.1 ␮m,
followed by cleaning with deionized water and ultrasound to remove
the alumina traces and finally the electrode was rinsed with deion-
ized water. The microbalance sensor was treated with 10% HCl solu-
tion for 10 min and then thoroughly rinsed with water. Both elec-
trodes were covered with a cadmium deposit (500 nm thickness) pre-
pared under the following galvanostatic conditions: cathodic current
density 153 ␮A cmϪ2, deposition time 81 min, temperature 25ЊC,
and a rotation rate of the electrode of 1500 rpm. As rotation was not
possible for the microbalance sensor, a magnetic stirrer was used to
impose the necessary hydrodynamic regime. This procedure was
used to obtain solid cadmium electrodes generated in situ and with
this, contamination problems (such as oxide formation) are avoided.
Once the working electrodes were prepared we proceeded to study
the electrochemical system Cdϩ2/Cd in the working solution by
means of different electrochemical techniques.
The measured jpc values showed a linear variation with regard to
v1/2, thus pointing out that the reduction process is under diffusion
control and therefore meets the Randles-Sevcik24 equation at 25ЊC
*
j
pc ϭ 2.69 ϫ 105n3/2C0D1/2v1/2
[1]
where n is the number of electrons involved in the reduction process.
*
Here n ϭ 2, Co is the bulk concentration of the electroactive specie,
and D is the diffusion coefficient of the same species. From the slope
of the jpc vs. v1/2 curve it was possible to estimate the apparent dif-
fusion coefficient for the electroactive species D ϭ 5.3 ϫ 10Ϫ5 cm2
s
Ϫ1. This value is larger than that reported in the literature (0.52 ϫ
10Ϫ5 cm2 sϪ1) for Cdϩ2 in a similar electrolytic medium.25
When the absolute value of the difference between the cathodic
peak potential and the cathodic half-peak potential, |Epc Ϫ Epc/2|,
was represented against log v, only for v values lower than 40 mV
sϪ1 such a difference was closer to that expected for a fast system
(29.5 mV at 25ЊC and n ϭ 2)24 and only in this condition our results
are in agreement with those normally observed in polarographic
studies,7,8 i.e., the system is reversible. However, for v larger than
100 mV sϪ1 the difference |Epc Ϫ Epc/2| was closer to the expected
value for an irreversible system (47.7 mV at 25ЊC, n ϭ 2 and ␣ ϭ
0.5).24 Then, in order to attain a better understanding about the elec-
trochemical behavior of Cdϩ2 reduction, a more complete kinetic
study was carried out.
In the case of electrochemical measurements performed at a slow
potential scan rate or in the case of pulsing techniques with long
periods, it was always necessary to use a new electrode to avoid the
influence of the electrode surface roughening. In the case of pulse
techniques with short periods, multiple experiments were performed
with the same electrode, because roughening was not appreciable.
Results and Discussion
The basic electrochemical characteristics.—The cadmium elec-
trochemical reduction process was first studied by using the linear
potential scan voltammetry technique in a nonstationary diffusion
regime. Potential scans in the cathodic direction were applied starting
from the equilibrium potential in the working solution (Ϫ1.141 V);
the study was made with potential scan rates (v) in the range of 5 to
The kinetics.—The cadmium electrochemical reduction in station-
ary diffusion regime was studied by using the RDE technique. A con-
stant potential scan rate (2 mV sϪ1) was imposed and the electrode
rotation rate (␻) was varied from 100 to 1200 rpm. Some j-E curves
obtained under these conditions are shown in Fig. 2. At low overpo-
tentials the j-E curves show a slight dependence with respect to the
electrode rotation rate, but such dependence is more evident at high
1000 mV sϪ1
.
Figure 1 shows voltammograms obtained for some potential scan
rates. In general, these curves show a quick growth of current at low
overpotentials, which is indicative of favorable kinetics for the
reduction process of Cdϩ2 to Cd0. Because of this behavior several
Figure 2. Potentiodynamical curves in stationary diffusion regime for cad-
mium reduction at 2 mV sϪ1; a, 200; b, 400; c, 600; d, 800; e, 1000; and f,
1200 rpm. Inset: Variation of the cathodic limiting current density with agi-
tation.
Figure 1. Linear sweep voltammograms for cadmium reduction in sulfate
medium: a, 30; b, 50; c, 70; d, 90; c, 125; and f, 200 mV sϪ1
.
Downloaded on 2014-11-21 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 147 (3) 1031-1037 (2000)
1033
S0013-4651(99)07-120-7 CCC: $7.00 © The Electrochemical Society, Inc.
overpotentials and different plateaus were observed which corre-
spond to the limiting current densities (jlim). The plateau width dimin-
ished as ␻ was increased. This is because the hydrogen evolution
reaction is favored. This observation can be explained as a result of
two effects: first, the Hϩ reduction is favored to produce H2 because
H2 is more rapidly displaced from the electrode surface when ␻ is
increased and, second, at the end of the j-E curve the electrode sur-
face has been modified by roughening (because of the slow potential
scan rate) thus increasing the number of active sites for Hϩ reduction.
These two effects lead to an apparent displacement of the hydrogen
reduction to less negative values of potential and, because of this, for
␻ > 1200 rpm, the limiting current density was difficult to define.
The limiting current densities showed a linear variation with
regards to ␻1/2, as can be appreciated in the inset of Fig. 2, which
means that the cadmium reduction process is limited by mass trans-
port of the electroactive species. Under these conditions the Levich26
equation is satisfied
rate equation related to a cathodic process under activation control,
Tafel28 equation
␣nF
[4]
jk ϭ j0 exp Ϫ
␩
RT
which is a function of the charge-transfer coefficient ␣, the overpo-
tential ␩, and the exchange current density defined28 as jo
nFkoCo(1Ϫ␣), where ko is the rate constant.
ϭ
The Tafel plot is shown in Fig. 4 and, as can be seen, there is an
overpotential range (Ϫ20 to Ϫ90 mV) where log jk vs. ␩ is linear,
just as predicted by the Tafel equation. From the slope of this curve
it was possible to obtain the charge-transfer coefficient ␣ ϭ 0.65,
and from the intercept the following values were estimated jo
ϭ
3.41 mA cmϪ2 and ko ϭ 8.98 ϫ 10Ϫ5 cm sϪ1. Therefore, the elec-
trochemical reduction process of Cdϩ2 on solid cadmium electrode
can be classified as quasi-reversible.24,28
Ϫ1/6
1/2
*
C0␻
j
lim ϭ 0.620nFD2/3
␯
[2]
As was expected, the cadmium electroreduction kinetics on solid
cadmium electrodes is slower than the electrochemical reduction on
a mercury drop or mercury amalgam electrode. In this work we have
obtained a rate constant which is three orders of magnitude lower
than in the case of mercury electrodes; however, the value for ko we
have obtained is similar to the ones reported from other electro-
chemical and nonelectrochemical experiments15,17,18 performed on
solid cadmium electrodes. As regards the charge-transfer coefficient,
it is difficult to establish a comparison because in each of these
works the values were different.
It is known that temperature plays an important role in the kinet-
ics of electrochemical reactions. Therefore, the temperature effect on
the cadmium electrochemical reduction process was investigated by
using the RDE technique. Potentiodynamic curves in stationary dif-
fusion regime were obtained for a well-defined hydrodynamic con-
dition ␻ ϭ 300 rpm and a slow potential scan rate (2 mV sϪ1). These
j-E curves are shown in Fig. 5. In these curves the equilibrium poten-
tial of the working electrode is displaced toward less negative values
when the temperature is increased. The direct comparison of the
curves shows that at low temperatures (curves a and b) an applied
overpotential of Ϫ30 mV, with respect to the start of each curve,
leads to a cathodic current of 2-3 mA cmϪ2, while the same applied
overpotential when the temperature is high (curves g and h) pro-
duces a reduction current approximately five times higher, thus re-
vealing that kinetics is enhanced with temperature increase.
where F is the Faraday constant, ␯ is the kinematic viscosity (0.0094
cm2 Ϫ1),27 and the other parameters have the same meaning as
s
defined in previous paragraphs. From the slope of the Levich-type
plot (inset, Fig. 2) the apparent diffusion coefficient for the elec-
troactive species was evaluated as D ϭ 2.19 ϫ 10Ϫ5 cm2 sϪ1. This
value is a better estimation for the diffusion coefficient and agrees
well with that reported in the literature from other techniques.17,27
Extrapolating the curve jlim vs. ␻1/2 to ␻1/2 ϭ 0 (Fig. 2, inset) leads
to a nonzero current density value, revealing the existence of other
phenomena coupled to the charge-transfer process which increase
the current density.26
In order to evaluate the kinetic parameters, the current densities
free from diffusional effects, jk, were evaluated by plotting jϪ1 vs.
Ϫ1/2
␻
␻
for several constant overpotentials and extrapolating to
Ϫ1/2 ϭ 0, according to the equation28
1/6
1
1
1.61␯
ϭ
ϩ
[3]
nFD2/3C0*␻
1/2
j
jk
Ϫ1/2
Some typical curves jϪ1 vs. ␻
are shown in Fig. 3 for some
fixed overpotentials and jk was obtained from the intercepts. From
the slope of each curve, the diffusion coefficient was evaluated, and
it was almost the same as that obtained by using the Levich equation.
By using the pure kinetic current densities jk for each overpoten-
tial, it was possible to estimate the kinetic parameters by using the
Ϫ1/2
Figure 3. Plot of jϪ1 vs. ␻
for some overpotentials: a, Ϫ40; b, Ϫ55; c,
Figure 4. Tafel plot for cadmium reduction at 25ЊC, jk in mA cmϪ2
.
Ϫ70; and d, Ϫ100 mV.
Downloaded on 2014-11-21 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

1034
Journal of The Electrochemical Society, 147 (3) 1031-1037 (2000)
S0013-4651(99)07-120-7 CCC: $7.00 © The Electrochemical Society, Inc.
regime is observed. However, at long times (not shown in the figure)
a slight increment of the current was obtained which is caused by the
nucleation phenomenon, as discussed below.
As potential steps are more negative (curves c and d, Fig. 7), a
slight mass-transfer dependence of the current is expected, that is to
say, a decay of the cathodic current density with time should be ob-
served. However, during the first 300 ms the current response in
Fig. 7 (curves c and d) shows that this is not the case. On the con-
trary, after 300 ms the cathodic current shows an increment charac-
teristic of a nucleation phenomena.25,32 At higher overpotentials a
maximum in the j-t response appears (curves e to j, Fig. 7) and such
a maximum increases and moves toward shorter times. In these
curves only after a certain time the measured current falls down with
t; this behavior being characteristic of systems with nucleation under
diffusional control as observed for other metallic systems.33-35 Con-
sequently, the shape of this set of curves indicates that the electro-
chemical cadmium reduction is initially (curves a-b in Fig. 7) sus-
tained at defects already present at the Cd surface. But at high over-
potentials new defects are needed to sustain the high rate of deposit
growth, and these are generated through nucleation, indicating a
transition from Stranski-Kossel to Erdey-Gruz and Volmer growth
modes as overpotential (or current density) is increased.36
Figure 5. Effect of temperature on cadmium reduction at 2 mV sϪ1 and ␻ ϭ
300 rpm: a, 5; b, 15; c, 25; d, 35; e, 45; f, 55; g, 65; and h, 75ЊC.
In order to investigate if the nucleation phenomenon was under
diffusional control we applied the Scharifker’s nondimensional
(j/jm)2 vs. t/tm analysis33 on our j-t transients (with jm and tm as de-
fined in Ref. 33). Figure 8 shows this type of plot for an applied
overpotential of Ϫ75 mV, and from this figure it is clear that the cad-
mium nucleation process does not fit to a diffusionally controlled 3D
nucleation and growth model (neither progressive nor instantaneous)
thus confirming that cadmium reduction is a complex process. In
Fig. 8, the experimental points are over the corresponding curves for
the nucleation models thus revealing the existence of a coupled phe-
nomenon which contributes to the measured current. The presence
of adsorbed species on the electrode surface gives rise to measured
currents higher than those expected (for a given nucleation model) in
the rising part of the plots (j/jm)2 vs. t/tm, as has been discussed in
some papers.33,37 Therefore, in our case, a phenomenon like this
might be possible, and this possibility was analyzed by means of
other techniques.
From the curves in Fig. 5 it was possible to evaluate the activa-
tion energy Ea for the reduction process by using the limiting cur-
rents29 and the Arrhenius equation in the form30
a
j
lim ϭ A exp(ϪE /RT)
[5]
where A is the frequency factor, R is the universal gas constant, and
T is the absolute temperature. The activation energy of the diffu-
sional process was evaluated from the slope of the plot ln jlim vs. TϪ1
(Fig. 6) as Ea ϭ 11.86 kJ molϪ1, which is a reasonable value for a
process under mass-transport control.31
Nucleation and adsorption processes.—Coupled phenomena to
the charge transfer were studied by using other electrochemical tech-
niques in nonstationary diffusional regime. The chronamperometric
behavior of the cadmium electrochemical reduction was studied by
imposing different overpotentials (Ϫ5 to Ϫ348 mV) to the working
electrode during different time intervals t (300-3000 ms). Figure 7
shows some j-t responses, and at low overpotentials (curves a and b)
a current response similar to that obtained in the stationary diffusion
For applied overpotentials larger than Ϫ285 mV the j-t response
does not any longer depend on the applied overpotential (curve k in
Fig. 7). Then, the typical decay of the current with time is observed,
and in this condition the Cottrell38 equation should be satisfied
Figure 7. Chronoamperometric curves for cadmium reduction after applica-
tion of several cathodic overpotentials: a, 19; b, 31; c, 35; d, 39; e, 43; f, 47;
g, 51; h, 55; i, 59; j, 63; and k, 285 mV.
Figure 6. Arrhenius plot for cadmium reduction, jlim in mA cmϪ2
.
Downloaded on 2014-11-21 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 147 (3) 1031-1037 (2000)
1035
S0013-4651(99)07-120-7 CCC: $7.00 © The Electrochemical Society, Inc.
Figure 8. Comparison between experimental points for cadmium deposition
on Cd (*) and the theoretical nondimensional plots for instantaneous (᭜) and
progressive (᭹) nucleation models.
Figure 9. Chronopotentiometric curves for different cathodic current pulses:
a, 39; b, 45; c, 51; d, 58; and e, 128 mA cmϪ2
.
cathodic current was applied during 1.5 s followed by a zero current
during 1 s; in these conditions, the electroactive species consumed
during the pulse are regenerated during the zero current condition.
Repeating this sequence produces the E-t response shown in Fig. 10.
The start of the experiment is at t ϭ 1 s; before this time the work-
ing electrode is only stabilized. When current is applied, the E-t
response shows transition times (␶) with the first one larger than the
second one and so on. This means that regeneration of the elec-
troactive species at surface level is not complete, even though the
electrode potential is completely recovered during j ϭ 0. On the
other hand, this behavior could be indicative of the existence of elec-
troactive species in the adsorbed state on the working electrode.
These adsorbed species are formed during the stabilization time of
the electrode at open circuit. Nevertheless, the most important point
in Fig. 10 is the observation of a minimum in the E-t curve (point A)
only the first transition time, such as that observed in the curve series
in Fig. 9. This means that the nucleation process certainly occurs but
once it takes place subsequent pulses simply result in deposition on
the previously formed nuclei. This behavior eliminates the possibil-
nFD1/2C0
*
[6]
j ϭ
␲
1/2t1/2
where all the parameters have been defined.
When j values were plotted against tϪ1/2 for t > 40 ms, most of
the experimental points followed a linear behavior. But for t < 40 ms
the experimental points deviated from this trend, and deviation
increased as t was smaller. In this region (t < 40 ms) the measured
current was larger than that expected according to the Cottrell equa-
tion. As deviation is large, such behavior cannot only be attributed to
double-layer charging effects because these are not too high. Conse-
quently, nucleation can be viewed as the most important phenome-
non causing the observed deviation. Once this takes place, the reduc-
tion process is favored at short times thus abnormally increasing the
measured current. After this, the deposition rate is only limited by
the diffusional process, and the Cottrell behavior is observed. Then,
from the slope of the linear part of the j-tϪ1/2 plot, it was possible to
estimate the apparent diffusion coefficient as D ϭ 5.68 ϫ 10Ϫ5 cm2
s
Ϫ1, which is slightly larger than the value previously obtained from
voltammetry in nonstationary diffusion regime.
In order to clarify and obtain a better understanding of the pro-
cesses coupled to the charge-transfer reaction we continued the
study of the cadmium reduction process by the chronopotentiomet-
ric technique. In this case cathodic current steps (4 to 128 mA cmϪ2
)
were imposed to the working electrode during several time intervals.
Some E-t responses are shown in Fig. 9. In all cases the E-t response
has a minimum in the potential immediately after the current pulse.
This minimum grows and becomes narrow as the imposed current
density is increased. It has been suggested that such kind of mini-
mum in the E vs. t response may be caused by impurities in either
the working solution or on the working electrode surface,39 howev-
er, this is not the case because in this work special care was taken
about this. On the other hand, there are some papers40-42 in which
that behavior is associated with the discharge mechanism of metal-
lic divalent ions in two steps, that is to say, from Mϩ2 to Mϩ1 and
finally to M0. But in this study we found no evidence of such a
sequence. According to our results, the minimum in the E-t curves is
then clearly due to a nucleation process because nuclei generation
requires additional energy, but once these nuclei have been generat-
ed less energy is required to continue the reduction process and the
electrode potential goes toward less negative values. Milchev and
Montenegro43 have previously discussed this situation.
To reinforce this interpretation, the working electrode was sub-
Figure 10. Potential-time response after application of a cathodic pulsed cur-
jected to a pulsed current train without stirring. A 10.5 mA cmϪ2
rent: 10.5 mA cmϪ2 during 1.5 s and 0 mA cmϪ2 during 1 s.
Downloaded on 2014-11-21 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

1036
Journal of The Electrochemical Society, 147 (3) 1031-1037 (2000)
S0013-4651(99)07-120-7 CCC: $7.00 © The Electrochemical Society, Inc.
ity of impurities in solution as a cause for the observed effect and
besides this, the possible existence of Mϩ species as intermediates in
the reduction mechanism can be rejected because in such cases the
minimum in the E-t response should be observed in all the patterns
of successive galvanostatic pulses.
After these considerations, transition times showed in Fig. 9 were
corrected by subtracting the time associated to the nucleation step. If
the cadmium reduction process is limited by diffusion, the Sand38
equation must be satisfied
1/2
*
nFD1/2
␲
C0
[7]
j ϭ
1/2
2␶
Ϫ1/2
and this means that the plot j vs. ␶
must be linear. Figure 11
shows this kind of representation where a certain linear behavior is
obtained but only for relatively larger ␶ values (left side in the
graph), that is to say, at low j values. The deviation observed in the
experimental data (curve a, Fig. 11) at small ␶ is because of the high
contribution from the nucleation step and even though the ␶ values
were corrected (curve b, Fig. 11), it was not possible to get good
agreement with the Sand equation in such a region (curve c, Fig. 11).
From the corrected data (curve b, Fig. 11) at j < 70 mA cmϪ2 it
was possible to obtain the apparent diffusion coefficient of the cad-
Ϫ1 for cadmium reduction.
Figure 12. Plot of j␶ vs. j
mium species in the working solution, D ϭ 4.53 ϫ 10Ϫ5 cm2 sϪ1
,
which is similar to those obtained from the other nonstationary tech-
niques, and this is closer to the RDE value.
Our chronopotentiometric results allowed us to analyze the pos-
sible existence of adsorbed electroactive species on the electrode
surface. This was done by plotting j␶ vs. jϪ1 (Fig. 12) according to
the equation44
using two electrochemical techniques in nonstationary diffusion
regime: the linear potential scan voltammetry and the electrochemi-
cal quartz crystal microbalance.45,46 By coupling these two tech-
niques it was possible to obtain, simultaneously, both the electrical
response (voltammogram) and the corresponding deposit growth as
a function of time. The potential scan rates were varied from 2 to 50
mV sϪ1, and Fig. 13 shows the result for 2 mV sϪ1
.
2
2
2
*
n F ␲DC0
4 j
[8]
j␶ ϭ
ϩ nF⌫o
Curve 1 in Fig. 13 shows a fast current increasing at the begin-
ning of the potential scan but, in spite of the high current density, the
deposit thickness (␦) is not modified (zone A, curve 2). This behav-
ior is due to the electrochemical reduction of the adsorbed species,47
already quantified by the quartz crystal microbalance sensor. The B
region in curve 2 corresponds to reduction of the electroactive
species existing in the electrode vicinity, and, once the process is
limited by diffusion, a peak appears in curve 1. The C region in curve
2 corresponds to the diffusion-limiting step, which implies a slight-
ly lower deposition rate. For potential scan rates up to 30 mV sϪ1, it
is still possible to observe the three zones in the thickness-time curve
but at higher potential scan rates the A region is shortened and the B
and C regions are superimposed.
where ⌫o is the amount of adsorbed species. Figure 12 shows a lin-
ear dependence between j␶ and jϪ1 which means that both dissolved
and adsorbed electroactive species are reduced; however, the ad-
sorbed electroactive species are not totally reduced to Cd before
reduction of the dissolved (Cdϩ2) species starts.28,44 From the inter-
cept in Fig. 12 the amount of adsorbed electroactive species was
evaluated as ⌫o ϭ 1.31 ϫ 10Ϫ9 mol cmϪ2, which is less than the
necessary quantity to form a monolayer. This value is small, and we
decided to apply another technique in order to confirm such a result.
Another evidence for the existence of absorbed electroactive
species on the electrode surface was obtained by simultaneously
Figure 13. Linear sweep voltammetry for cadmium reduction at 2 mV sϪ1
(curve 1) and the corresponding growth of the deposit thickness, ␦ (curve 2).
Figure 11. Sand plot for cadmium reduction: a, experimental transition
times; b, nucleation-corrected transition times; c, Sand equation.
Downloaded on 2014-11-21 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 147 (3) 1031-1037 (2000)
1037
S0013-4651(99)07-120-7 CCC: $7.00 © The Electrochemical Society, Inc.
CINVES TAV-IPN assisted in meeting the publication costs of this article.
The necessary time to observe a variation in the deposit thickness
in Fig. 13 (curve 2) is approximately 5 s. For this time interval, inte-
gration under curve 1 (same figure) gives the charge associated to
the adsorbed species reduction (0.37 mC cmϪ2), and with this value
References
1. K. Iwai and T. Kojima, Prog. Batt. Sol. Cells, 4, 291 (1982).
2. R. H. Bube and K. W. Mitchell, J. Electrochem. Mater., 22, 17 (1993).
3. D. R. Biswas, J. Mater. Sci., 21, 2217 (1986).
4. P. C. Searson, Sol. Energy Mater. Sol. Cells, 27, 377 (1992).
5. F. A. Kröger, J. Electrochem. Soc., 125, 2028 (1978).
6. D. Landolt, Electrochim. Acta, 39, 1075 (1994).
7. N. Tanaka and R. Tamamushi, Electrochim. Acta, 9, 963 (1964).
8. N. A. Hampson and R. S. Lathan, in Encyclopedia of Electrochemistry of the Ele-
ments, Vol. 1, Marcel Dekker, New York (1974).
9. J. A. Harrison and D. R. Sandbach, J. Electroanal. Chem., 85, 125 (1977).
10. I. R. Burrows, K. L. Dick, and J. A. Harrison, Electrochim. Acta, 21, 81 (1976).
11. B. Lovrecek and N. Marincic, Electrochim. Acta, 11, 237 (1966).
12. A. R. Despic, D. R. Jovanovic, and S. P. Bingulac, Electrochim. Acta, 15, 459
(1970).
13. M. Anbar, Q. Rev. Chem. Soc., 22, 578 (1968).
14. T. Biegler, E. R. Gonbzalez, and R. Parsons, Collect. Czech. Chem. Commun., 36,
414 (1971).
15. N. A. Hampson, R. J. Latham, and D. Larkin, J. Electroanal. Chem., 23, 211
(1969).
16. N. A. Hampson and R. J. Latham, J. Electroanal. Chem., 32, 175 (1971).
17. H. P. Agarwal and P. Jain, Electroanal. Acta, 26, 621 (1981).
18. G. L. Statulyavichius, Y. P. Butkyavichius, and Y. P. Shivitskis, Russ. J. Elec-
trochem., 32, 594 (1996).
it was possible to obtain the amount of adsorbed species, ⌫o
ϭ
1.92 ϫ 10Ϫ9 mol cmϪ2 (coverage degree ␪ ϭ 0.87). This value is
close to that obtained in chronopotentiometry and also corresponds
to an amount of adsorbed species less than that necessary to form a
monolayer, which confirms our previous result. As a reference, the
quantity of adsorbed Cdϩ2 has been reported as 10Ϫ12 mol cmϪ2 in
a perchlorate medium with this value being almost independent of
the Cdϩ2 concentration.16
The existence of adsorbed electroactive species on the working
electrode gives rise to an apparent higher concentration of Cdϩ2 in
the vicinity of the electrode surface. This fact might be an explana-
tion for the relatively high D values obtained from electrochemical
techniques in nonstationary diffusion regime. Consequently, our D
values are called here “apparent” and only when Cdϩ2 reduction
takes place on electrodes of different nature and the process is sim-
pler (without adsorption) is it possible to evaluate D in the order of
10Ϫ6 cm sϪ1, as reported in some papers.25,27
In this work the presence of small quantities of adsorbed elec-
troactive species on the electrode surface has been shown by using
different electrochemical techniques. None of these techniques sup-
port the presence of a monovalent intermediate as was proposed in
some papers. Therefore the Cdϩ2 seems to be the only adsorbed elec-
troactive species, then leading to a reduction mechanism of the way
19. R. D. Armstrong, M. Todd, J. W. Atkinson, and K. Scott, J. Appl. Electrochem., 26,
379 (1996).
20. R. N. Adams, Electrochemistry at Solid Electrodes, Marcel Dekker, New York
(1970).
21. K. E. Heusler and L. Gaiser, J. Electrochem. Soc., 117, 762 (1970).
22. W. Lorenz, Z. Elektrochem., 58, 912 (1954).
23. H. D. Hedrich and E. Raub, Metalloberflaeche, 29, H.4, 160 (1975).
24. Southampton Electrochemistry Group, Instrumental Methods in Electrochemistry,
John Wiley & Sons, Inc., New York (1985).
ϩ2
Cdϩ2 r Cd
25. G. Gunawardena, G. Hills, and I. Montenegro, J. Electroanal. Chem., 184, 371
(1985).
ads
26. Y. V. Pleskov and V. Y. Filinovskii, The Rotating Disc Electrode, Consultants
Bureau, New York (1976).
Cdϩad2s ϩ 2eϪ r Cd
27. V. M. M. Lobo, Handbook of Electrolyte Solutions, Elsevier, New York (1989).
28. A. J. Bard and L. R. Faulkner, Electrochemical Methods, Fundamentals and Appli-
cations, John Wiley & Sons, Inc., New York (1980).
29. H. Sánchez, E. Chainet, B. Nguyen, P. Ozil, andY. Meas, J. Electrochem. Soc., 143,
2799 (1996).
This reaction sequence is taking place principally at temperatures
below 35ЊC but the reduction process is complicated by a nucleation
step, and at high temperatures cadmium reduction is kinetically
more favorable.
30. E. Gileadi, Electrode Kinetics for Chemist, Chemical Engineers and Materials Sci-
entists, VCH Publishers, Inc., New York (1993).
31. E. A. M. Hughes, The Chemical Statics and Kinetics of Solutions, Academic Press,
New York (1971).
32. A. I. Danilov, E. B. Molodkina, andY. M. Polukarov, Russ. J. Electrochem., 33, 288
(1997).
33. B. R. Scharifker and G. Hills, Electrochim. Acta, 28, 879 (1983).
34. L. Heerman and A. Tarallo, J. Electroanal. Chem., 470, 70 (1999).
35. A. Serruya, J. Mostany, and B. R. Scharifker, J. Electroanal. Chem., 464, 39
(1999).
36. K. J. Vetter, Electrochemical Kinetics Theoretical Aspects, Academic Press, New
York (1967).
37. E. Bosco and S. K. Rangarajan, J. Chem. Soc., Faraday Trans. 1, 77, 1673 (1981).
38. D. P. Macdonald, Transient Techniques in Electrochemistry, Plenum Press, New
York (1981).
39. G. Razumney and J. O’M Bockris, J. Electranal. Chem., 46, 185 (1973).
40. I. H. Plonski, J. Electrochem. Soc., 116, 1688 (1969).
41. Z. Nagy, N. C. Hung, K. C. Liddell, M. Minkoff, and G. K. Leaf, J. Electroanal.
Chem., 421, 33 (1997).
42. L. H. Jenkins, J. Electrochem. Soc., 117, 630 (1970).
43. A. Milchev and M. I. Montenegro, J. Electroanal. Chem., 333, 93 (1992).
44. S. V. Tatwawadi and A. J. Bard, Anal. Chem., 367, 2 (1964).
45. D. A. Buttry and M. D. Ward, Chem. Rev., 92, 1355 (1992).
46. M. R. Deakin and D. A. Buttry, Anal. Chem., 61, 1147A (1989).
47. G. S. Ostrom and D. A. Buttry, J. Electroanal. Chem., 256, 411 (1988).
Conclusions
The cadmium electrochemical reduction in sulfate medium was
studied by means of different electrochemical techniques in both sta-
tionary and nonstationary diffusion regimes on solid cadmium elec-
trodes generated in situ. Our experiments led to the following kinet-
ic parameters: ␣ ϭ 0.65, jo ϭ 3.41 mA cmϪ2, and ko ϭ 8.98 ϫ
10Ϫ5 cm sϪ1, so demonstrating that the electrochemical process is
quasi-reversible at 25ЊC. The diffusion coefficient of the electroac-
tive species was almost the same for all the applied techniques, being
the value from RDE closer to the literature’s value. The diffusional
step of the reduction process is characterized by an apparent activa-
tion energy of 11.86 kJ molϪ1, and temperature increase favors the
deposition rate. The reduction process is preceded by an adsorption
step of electroactive species, whose amount is less than the corre-
sponding to form a monolayer on the working electrode surface.
Besides, the deposition process is complicated by a nucleation step
detected by impulsional techniques. The mechanism seems to be a
two-step mechanism involving Cdϩ2 species in adsorbed state.
Downloaded on 2014-11-21 to IP 138.251.14.35 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).