The role of temperature in copper electrocrystallization in ammonia-chloride solutions

Article abstract of DOI:10.1016/j.electacta.2006.06.022

This paper presents the analysis of temperature effect on the copper electrocrystallization process from the stainless steel/Cu(II)-Cu(I)-NH₄Cl-NH₃-H₂O system. Electrochemical techniques and scanning electron microscopy with energy dispersion spectroscopy were employed. An increment in temperature had a favorable effect in increasing the kinetic and nucleation parameters, favoring the copper reduction on the stainless steel substrate. From the chronoamperometric study, it was possible to find the transfer coefficient (α), which does not have a significant variation with temperature, and the exchange current density (i₀) for different temperatures, where a 50.63 kJ/mol value was estimated for the activation energy. The potentiostatic study suggested the presence of two processes involved: an electron transfer reaction and a 3D nucleation-growth process, under combined charge transfer and diffusion limitations. In addition, an important current contribution could be accounted for on the basis of the existence of a capacitive component in the system. This capacitive behavior was associated to the oxide layer (Cr₂O₃)-chloride interactions on the stainless steel surface. Current transients analyses at different potentials, based on the models of: Milchev (low overpotential), Scharifker-Mostany (SM) and Heermann-Tarallo (HT) (high overpotential) allowed to obtain the values of nucleation parameters, such as: the nucleation rate constant (A), the active nucleation sites number (N₀), the stationary nucleation rate (I_st = A × N₀) and the nuclei saturation number (N_s). Finally, the deposits obtained were analyzed by SEM, showing an acceptable correlation between the nucleation parameters and morphology of the deposits obtained. An increment in temperature favored the growth of the cluster before the coalescence occurred.

Full text of DOI:10.1016/j.electacta.2006.06.022

Electrochimica Acta 52 (2006) 892–903
The role of temperature in copper electrocrystallization in
ammonia–chloride solutions
Jorge Vazquez-Arenasa, Roel Cruza,∗, Luis H. Mendoza-Huizarb
a Facultad de Ingenier´ıa, Instituto de Metalurgia, Universidad Auto´noma de San Luis Potos´ı, Av. Sierra Leona 550,
Lomas 2a seccio´n, 78210 San Luis Potos´ı, S.L.P., Me´xico
b Centro de Investigaciones Qu´ımicas, Universidad Auto´noma del Estado de Hidalgo, Unidad Universitaria,
Km 4.5 Carretera Pachuca-Tulancingo, CP. 42184 Pachuca-Hidalgo, Me´xico
Received 2 March 2006; received in revised form 20 June 2006; accepted 21 June 2006
Available online 4 August 2006
Abstract
This paper presents the analysis of temperature effect on the copper electrocrystallization process from the stainless steel/Cu(II)–Cu(I)–
NH4Cl–NH3–H2O system. Electrochemical techniques and scanning electron microscopy with energy dispersion spectroscopy were employed. An
increment in temperature had a favorable effect in increasing the kinetic and nucleation parameters, favoring the copper reduction on the stainless
steel substrate. From the chronoamperometric study, it was possible to find the transfer coefficient (α), which does not have a significant variation
with temperature, and the exchange current density (i0) for different temperatures, where a 50.63 kJ/mol value was estimated for the activation
energy. The potentiostatic study suggested the presence of two processes involved: an electron transfer reaction and a 3D nucleation—growth
process, under combined charge transfer and diffusion limitations. In addition, an important current contribution could be accounted for on the
basis of the existence of a capacitive component in the system. This capacitive behavior was associated to the oxide layer (Cr2O3)–chloride
interactions on the stainless steel surface. Current transients analyses at different potentials, based on the models of: Milchev (low overpotential),
Scharifker–Mostany (SM) and Heermann–Tarallo (HT) (high overpotential) allowed to obtain the values of nucleation parameters, such as: the
nucleation rate constant (A), the active nucleation sites number (N0), the stationary nucleation rate (Ist = A × N0) and the nuclei saturation number
(Ns). Finally, the deposits obtained were analyzed by SEM, showing an acceptable correlation between the nucleation parameters and morphology
of the deposits obtained. An increment in temperature favored the growth of the cluster before the coalescence occurred.
© 2006 Elsevier Ltd. All rights reserved.
Keywords: Copper; Electrocrystallization; Kinetics; Temperature; Adsorption
1. Introduction
perature dependence of the nucleation rate is determined mainly
by the thermodynamic and kinetic barriers for nucleation. Since
a temperature decrease produces two effects: a decrease of the
thermodynamic barrier, due to an increase in the thermodynamic
driving force for crystallization, leading to a higher nucleation
rate and an increase of the kinetic barrier leading to a lower
nucleation rate [6].
However, a contrary behavior has been observed [1–5],
which is probably because the value of the surface energy
(σ) leads to a slight increase with increasing temperature
(dσ/dT ∼ (0.06–0.16) × 10−3 J/m2 K) regardless of the method
of estimating the kinetic barrier [7–11]. Moreover, if one take
in account a possible curvature (or nucleus size) dependence
of the surface tension (the specific surface energy, estimated
from nucleation rate data, refers to nuclei of critical size) [6].
Curvature corrections can be expected to lead to a reduction in
Electrocrystallizationisacomplexprocessaffectedbyfactors
such as: substrate, additives, electroactive species, bath compo-
sition, mass transport, temperature, etc. Although, the effect of
temperature is well known, only few studies have been reported
about this factor effect [1–5]. Such studies indicate that nucle-
ation and growth rate increase with the temperature, which could
intrinsically suggests a diminution of energy requirement for the
electrodeposition process. The Gibbs’s energies as well as the
activation energies are potential-dependent, but the temperature-
dependence of these quantities is at all not obvious. The tem-
∗
Corresponding author. Tel.: +52 444 825 43 26; fax: +52 444 825 43 26.
E-mail address: rcuz@uaslp.mx (R. Cruz).
0013-4686/$ – see front matter © 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2006.06.022

J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
893
the effective value of the surface energy. As the critical nucleus
size increases with temperature, the effect of curvature correc-
tions decreases, leading to higher effective values of the surface
energy [12]. Thus, the increase in the surface energy produces
an increment in the value of the rate constant for nucleation
favoring the electrodeposition process [12].
Another effect of temperature in the electrodeposition pro-
cess is the increase of the coefficient diffusion value as the
temperature increases [13], which generates a major flux to the
chemical species to electrode surface.
At present work, the temperature effect in the electrode-
position process of copper has been analyzed for the system
stainless steel (SS)/Cu(II)–NH4Cl–NH3–H2O. For this study,
electrochemical techniques and scanning electron microscopy
with energy dispersion spectroscopy were utilized. This study
is part of an investigative work aimed at the understanding
of electrochemical systems based on Cu(II)–NH4Cl–NH3–H2O
electrolytes [14,15].
Fig. 1. Typical cyclic voltammogram obtained in the SS/0.2 M Cu(II), 4 M Cl−
and NH4+ + NH3 (pH 7.5) system. The temperatures studied are indicated in
figure. Scan rate: 100 mV s−1
.
ing bath was used to fix and to control temperature within
1 K.
2. Methodology
A conventional three-electrode cell (100 mL) was used for
the electrochemical experiments. A 304 stainless steel (∼19%
Cr, ∼08% Ni and ∼2% Mn) rod was embedded in resin to
generate a rotating disk working electrode with a 0.125 cm2
area. A graphite rod (Alfa AESAR, 99.999%) was utilized as
counter electrode and a saturated calomel electrode (SCE) was
the reference electrode; all reported potentials are referred to this
scale. Prior to each experiment, working electrode surface was
polished to mirror finish using 0.05 ␮m BuehlerTM alumina pow-
The experiments were carried out in 0.2 M Cu(II), 4 M
+
NH3 + NH4 and 4 M Cl−electrolyte. All solutions were pre-
pared using ultrapure water (18.2 ␮ꢀ cm NanopureDiamond
(BarnsteadTM ) and CuCl2·2H2O (Fisher), NH4Cl (Fisher) and
NH4OH (Fisher) analytic grade reagents. In each case, the
pH was controlled by addition of 4 M NaOH (Fisher) and
10 M H2SO4 (J.T. Baker) solutions, and by a pH meter
Termo Orion 420A. A 10 L PolyscienceTM heater circulat-
Fig. 2. Typical cyclic voltammograms obtained in the SS/0.2 M Cu(II), 4 M Cl− and NH4+ + NH3 (pH 7.5) system, at different temperatures: (a) 298, (b) 313 and
(c) 323 K and at different scan potential rates indicated in figure.

894
J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
der, pre-washed for 5 min with pure water in an ultrasonic bath.
Inert (nitrogen) atmosphere was maintained during all experi-
ments. The working electrode potential was controlled by an
AUTOLABTM model PGSTAT 30 potenstiotat–galvanostat cou-
pled to a personal computer using GPESTM software.
Note that the current peak Ic increases as temperature,
which could be due to enhancement of kinetic and diffusive
processes, such as it is indicated by a better defined peak.
However, this current increment could also suggest a diminu-
tion of the Cr2O3–chloride interaction, which persist even at
323 K.
Fig. 2 shows a set of voltammograms obtained for different
potential scan rates and temperatures. Observe that for all cases,
a fast increase in cathodic current was obtained after the slow
current step (−E > 0.5 V). This result indicates that a favorable
kinetic condition for the copper electrodeposition (IIIc peak)
on the substrate surface has been reached. Furthermore, it can
be observed that the cathodic current associated with peak IIIc
increases with the temperature. Similar results were obtained
by Ramos et al. [16], where the electrocrystallization stage of
(Cu(NH3)2+/Cu(0)) was explained in terms of diffusional con-
trol.
The peak IIIc formation is certainly due to species deple-
tion in the electrode/solution interface. Observe that a gradual
shifting of peak (Ep) toward more negative values occurs as the
scan rate (v) increases. In order to determine the control type
limiting this process, the current value associated with the peak
(−Ip) was plotted as a function of v1/2 (Fig. 3). Note that a
reversible behavior is obtained approximately at scan rate lower
than 50 mV s−1 (Fig. 3a), 80 (Fig. 3b) and 50 mV s−1 (Fig. 3c)
for 298, 313 and 323 K, respectively. On the other hand, an irre-
versible behavior was observed at higher scan rates. This feature
is characteristic when a process passes through a region known
as quasi-reversible [17]. The linearity showed by the −Ip ver-
sus v1/2 at the different temperatures, suggests an irreversible
3. Results and discussion
3.1. Voltammetric study
Fig. 1 show the voltammetric direct scans from the 0.2 M
Cu(II), 4 M NH3 + NH4+ and 4 M Cl− solution at 298, 313, 323
and 343 K, at pH 7.5 on the SS substrate. At 298 K, it was
possible to observe three cathodic peaks: Ic, IIc and IIIc, at
−0.380, −0.550 and −0.650 V, respectively (see Fig. 1). In a
recent work, Vazquez-Arenas et al. have shown that, for this
system, the reduction process for predominant chemical species
(Cu(NH3)3Cl+ complex) involves the following steps [14,15]:
(a) Cu(NH3)3Cl+/Cu(NH3)2Cl, occurring in the region labeled
as Ic. The low current obtained in this step has been
attributed to a slow reduction process due to the barrier
formed by the oxide layer (Cr2O3)–chloride interactions in
the SS surface.
(b) Cu(NH3)2Cl/Cu(0), occurring in the region labeled as IIc.
This step is characterized by an important increase in the
cathodic current.
(c) A faster reduction Cu(NH3)3Cl+/Cu(0) labeled as peak IIIc,
which take place due to the higher overpotential.
Fig. 3. Experimental cathodic peak current (−Ip) as a function of scan rate (v1/2), obtained at temperatures indicated in figure.

J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
895
Fig. 4. A set of experimental current transients recorded in the SS/0.2 M Cu(II), 4 M Cl− and NH4+ + NH3 (pH 7.5) system, at different potential step (V) indicated
in figure. System temperature: (a) 298, (b) 313 and (c) 323 K.
reduction process occurring under diffusion control [17–19],
according with the following equation:
ficient and C is the bulk concentration of the electroactive
species.
The absolute value between the cathodic potential peak (Ep)
and the cathodic potential half-peak (Ep/2) was obtained from
the voltammograms shown in Fig. 2. The calculated values were
quite close to the expected for an irreversible system [18,19].
These results newly suggest that the process is changing, from
reversible to irreversible system.
Ip = 2.99 × 105n(αnα)1/2AD1/2Cv1/2
(1)
where n is the number of the transferred electrons in the reduc-
tion process, nα the number of electrons transferred up to,
and including, the rate determining step, D the diffusion coef-
Fig. 5. Comparison between experimental transient normalized through the coordinates of its respective local maximum (tm, Im), obtained during the copper
nucleation on SS at (a) −0.72 (298 K), (b) −0.71 (313 K) and (c) −0.717 V (323 K), with the theoretical non-dimensional curves corresponding to 3D instantaneous
and progressive nucleation.

896
J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
3.2. Potentiostatic study
This process is known to take place in two steps:
Cu2+ + e− → Cu+ (slow)
Cu+ + e− → Cu0 (fast)
The slow step (Eq. (6)) may lead to the accumulation of Cu+ at
the electrode surface if the nucleation and growth of copper clus-
ters is not sufficiently fast [25,26]. Accordingly with Milchev et
al., an electron transfer reaction takes place prior to and simulta-
neously with the process of the nucleus formation. The current
density associated to these processes, before the nuclei overlap-
ping, but when the crystal formation has already started, is given
by:
(6)
(7)
The charge transfer and nucleation kinetic of copper elec-
trocrystallization process on SS electrode, were studied from
+
a system 0.2 M Cu(II), 4 M NH3 + NH4 and 4 M Cl− at dif-
ferent temperatures by the potentiostatic technique. Fig. 4a–c
shows a set of current transients obtained from different over-
potentials and at differente temperatures. All transients shown
in the figure exhibited a falling current in the initial period of
time. After this falling current, the transients showed a typical
current maximum (Im), related with the coalescence of diffusion
fields with spherical symmetry. Once the current maximum was
reached, a decay of the current was obtained, which approaches
to a planar diffusion. The shape of these transients exhibited the
features of typical three dimensional nucleation process with
hemispherical diffusion control (3D-dc) of the growing crystal-
lites [20–22]. The nucleation model proposed by Scharifker and
Hills allows classifying the nucleation and growth mechanism
as instantaneous (Eq. (2)) or progressive (Eq. (3)), by compari-
son between the dimensionless curves of the experimental and
theoretical transients [20].
IM = aR exp(−bRt) + q[(1 + 2nt)1/2 − 1]3
(8)
where
ꢂ
ꢃ
sπF4(DC)3
i0VM
3αFη
RT
q =
Ist exp
−
(9)
vi0VMQ
F2DC
n =
(10)
ꢀ
ꢀ
ꢁ
ꢂ
ꢂ
ꢃꢄꢅ
ꢃꢄꢅ
2
with s = 32/3 and v = 1/4 and
I2
Im2
1.9542
t/tm
t
ꢂ
ꢃ
ꢁ
ꢄ
=
1 − exp −1.2564
(2)
(3)
2αFη
RT
−2(1 − α)Fη
tm
Q = exp
− exp
(11)
RT
ꢁ
2
I2
Im2
1.2254
t/tm
t
i0 is the exchange current density at the “copper–solution” inter-
face, α the charge transfer coefficient, η the overpotential, Ist the
stationary nucleation rate and Vm is the volume molar. All other
parameters have their conventional meanings. Since the values
obtained for the number of reduced species were higher than an
effectivemonolayer, firstterminEq. (8)hasbeenrelatedwiththe
electron transfer reaction, i.e. Iet = aR exp(−bRt), discarding an
ion transfer reaction [25,26]. The second term describes the cur-
rent of progressive nucleation and growth of the copper crystals
on the substrate, under combined “charge transfer” and “diffu-
sion limitations” [27]. Therefore, the region of current falling
observed at short times may be related with the discharge of the
copper complexes, in two monoelectronic steps.
In order to evaluate the existence of a similar behavior for the
system here analyzed, Milchev’s model was applied to poten-
tiostatic curves obtained at low overpotentials and short times,
ensuring that any overlapping and coalescence of the clusters do
not occur during the growth. In this analysis, the model proposed
by Milchev et al., could not account for the transient behavior at
shorter times than 0.4 s. The capacitive effect generated by the
oxide layer (Cr2O3)–chloride interactions, suggested by voltam-
metry, involves the contribution of two capacitances in series,
namely the oxide film thicknesses barrier, and the double layer
capacitance, which could be influenced by chloride adsorption
and/or its reaction with oxide layer. A detailed study of this two
capacitance contributions has been reported elsewhere [24,29].
Hence, taking into account the above described, the initial falling
cathodic currents observed in the experimental current transients
(IT) obtained during copper deposition on SS electrode, may be
described by the following equation:
=
1 − exp −2.3367
tm
Fig. 5 shows a comparison of the theoretical dimension-
less transients generated by Eqs. (4) and (5), with experimental
dimensionless current transients obtained at different tempera-
tures. It is clear that at 313 K (Fig. 5b), most of the experimental
data fall within the range of validity of the theory proposed by
Scharifker et al. [20,21]. Last result suggests that the experimen-
tal transients could be fitted using the following equation:
1
I3D−dc(SM)(t) = zFDC
(πDt)1/2
×[1 − exp(−αSN0(πDt)1/2 1/2
t
Θ)]
(4)
(5)
with
1 − exp(−At)
Θ = 1 −
At
where z is the number of exchanged electrons, F the Faraday
constant, N0 the number of active nucleation sites, A the nucle-
ation rate and αS = 2[2VmDC]1/2
.
Observe that the values (I/Im) at 298 and 323 K (Fig. 5a and
c) are always lower than those obtained from the experimental
data. This difference might be associated to the contribution, to
the total current, of another processes not considered by the non-
dimensional analysis [22–24]. Also, note that the falling current
cannot be accounted for the Scharifker’s model.
3.2.1. Analysis of current transients at low overpotentials
Recently, Milchev et al. [25] reported that, for the case of
copper in sulfate solution, falling currents observed in transients,
as shown in Fig. 4, must be related to the copper ions discharge.
IT = IC + IM
(12)

J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
897
Fig. 6. Experimental current transients at different overpotentials, with the theorical transients (ꢀ) generated by non-linear fitting of Eq. (12) to the experimental
data. The temperatures are given by figures.
where IC = k1 exp(−k2 × t) is the current due to the capacitive
effect of Cr2O3–Cl− interaction, being k1 = k2QC, where QC is
the charge density due to the overall capacitance effect.
(outer sphere reaction), in agreement with the reported in the
literature [30,31].
An average value of 0.0456, 0.0487 and 0.0509 was obtained
for the i0, at temperatures of 298, 313 and 323 K, respectively.
The value obtained at 298 K is similar to that reported in litera-
ture in a case of a copper reduction from sulfate solutions [25].
The (ln i0) was plotted as a function of (1/T), Fig. 7, exhibiting
a linear relationship, according with the Arrhenius’s equation
[19]. From the slope value and Eq. (13) it was possible to calcu-
late the activation energy (ꢁH=) of the system as 50.63 KJ/mol.
Last result compares favorably with the obtained in the case of
electrodeposition of copper from cupric pyrophosphate solu-
tions [32].
Fig. 6 shows the comparison of the experimental current
transients obtained at low overpotentials with the theoretically
generated by non-linear fitting of experimental data by Eq. (12).
It can be observed, that the model expressed by Eq. (12) ade-
quately accounts for the behavior of all experimental transients.
In Table 1, the best-fit kinetic and nucleation parameters at
different temperatures are shown. In this table, the reference
electrode potential (SCE) values for 313 and 323 K were cor-
rected according with the methodology reported in literature
[28]. It is important to mention, that the charge transfer coeffi-
cients obtained from the fitting parameters of Table 1 and Eq. (9),
do not vary significantly with temperature, and they are around
0.5. This fact is indicative that the reduction reaction involv-
ing Cu(NH3)3Cl+/Cu(NH3)2Cl is an electron transfer reaction
∂ ln i0
ꢁH=
R
= −
(13)
∂(1/T)
Table 1
Potential dependence from kinetic and nucleation best-fit parameters at different temperatures, obtained through the fitting process of the experimental transients
using Eq. (12)
−E (V) vs. SCE 298 K
aR (A) bR (s−1
−E (V) vs. SCE 313 K
aR (A) bR (s−1
−E (V) vs. SCE 323 K
aR (A) bR (s−1
)
Ist (cm2 s−1
)
)
Ist (cm2 s−1
)
)
Ist (cm2 s−1
)
0.367
0.417
0.447
0.467
0.487
0.497
0.517
0.0017 0.173
0.0025 0.226
0.0033 0.257
0.0037 0.294
0.0031 0.237
0.0043 0.534
0.0052 0.717
3835.095
4447.202
5124.677
5618.889
6590.043
8110.426
9064.838
0.4
0.45
0.5
0.55
0.6
0.62
0.0006 0.136
0.0010 0.155
0.0013 0.175
0.0018 0.189
0.0022 0.207
0.0023 0.243
1390.488
2492.439
2076.329
2412.502
2921.393
3573.664
0.41
0.46
0.51
0.55
0.56
0.58
0.6
0.0008 0.143
0.0013 0.165
0.0019 0.199
0.0029 0.232
0.0029 0.212
0.0045 0.729
0.0028 0.191
2315.287
2850.112
3411.408
4032.103
4081.080
8182.813
8453.429

898
J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
Note in Table 1, that for similar overpotentials, the Ist value
increases as a function of temperature and the overpotential
applied. Similar results have been obtained for copper reduc-
tion from sulfate solutions [25].
Under the methodology proposed by Scharifker for the satu-
ration nucleus densities (Ns) assess [33], Ns values were cal-
culated from the parameters reported in Table 1, using the
following equation:
ꢂ
ꢃ
1/2
Ist
2kꢂD
Ns =
(14)
where
kꢂ = [8πCVM]1/2
(15)
Results obtained for Ns are summarized in Table 2. These
results reveal that the Ns value increases with the applied overpo-
tentialandwithtemperatureatsameoverpotentialvalue, remark-
ing an important role of these parameters in the generation of a
high saturation nucleus density. The charge contribution of the
Fig. 7. Arrhenius equation linear plot (ln i0 vs. 1/T) for reaction
Cu(NH3)3Cl+/Cu(0).
Table 2
Potential dependence of the Ns, calculated from physical constants showed in Table 1 and Eq. (14)
−E (V) vs. SCE
Ns (×10−3 cm−2
)
−E (V) vs. SCE
Ns (×10−3 cm−2
)
−E (V) vs. SCE
Ns (×10−3 cm−2
)
0.367
0.417
0.447
0.467
0.487
0.497
0.517
36.327
39.119
41.993
43.971
47.620
52.828
55.850
0.4
0.45
0.5
0.55
0.6
0.62
21.874
29.285
26.729
28.812
31.705
35.067
0.41
0.46
0.51
0.55
0.56
0.58
0.6
28.225
31.316
34.261
37.248
37.474
53.063
53.933
Fig. 8. Comparison between an experimental current transient (−) recorded at (a) −0.66 (298 K), (b) −0.65 (313 K) and −0.657 V (323 K), with the theorical
transient (ꢀ) generated by non-linear fitting of Eq. (16) to the experimental data.

J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
899
Table 3
Potential dependence for the physical constants involved during copper reduction, at different temperatures
−E (V)
298 K
−E (V)
313 K
vs. SCE
vs. SCE
aR (A)
bR (s−1
)
A (s−1
)
N0 (×10−6
D (×107
aR (A)
bR (s−1
)
A (s−1
)
N0 (×10−6
D (×107
cm−2
)
cm2 s−1
)
cm−2
)
cm2 s−1
)
0.58
0.60
0.62
0.64
0.66
0.68
0.70
0.72
0.74
0.76
0.78
0.80
0.61
0.63
0.65
0.67
0.69
0.71
0.73
0.75
0.77
0.004
0.004
0.006
0.007
0.001
0.002
0.004
0.005
0.004
0.715
1.070
1.321
1.032
0.496
1.178
1.569
2.292
2.368
1.225
1.947
2.641
2.898
2.862
8.873
11.675
14.158
12.675
11.61
17.18
21.67
25.87
50.83
57.30
65.88
88.79
150.3
2.828
2.907
2.833
2.794
2.688
2.703
2.534
2.506
2.446
0.002
0.003
0.003
0.004
0.005
0.005
0.005
0.006
0.008
0.198
0.238
0.226
0.358
0.488
1.187
1.279
1.652
2.269
0.926
1.511
1.785
2.305
2.754
3.510
3.472
4.849
5.029
2.202
3.235
3.759
4.552
8.661
33.95
55.20
79.15
88.01
2.428
2.339
2.665
2.460
2.383
2.356
2.126
2.042
2.334
−E (V) vs. SCE
323 K
aR (A)
bR (s−1
)
A (s−1
)
N0 (×10−6 cm−2
)
D (×107 cm2 s−1
)
0.597
0.617
0.637
0.657
0.677
0.697
0.717
0.004
0.004
0.005
0.007
0.008
0.010
0.009
0.407
0.705
1.044
1.536
3.177
4.753
0.890
1.495
3.341
4.510
8.194
9.373
10.098
27.477
4.613
8.606
12.95
20.55
49.33
82.93
6.194
3.647
3.322
3.477
3.333
3.144
3.052
4.116
Data were obtained from best-fit parameters found through the fitting process of the experimental transients using Eq. (16).
capacitive effect generated by the oxide layer (Cr2O3)–chloride
interactions shows an increase as a function of the tempera-
ture (2.08 × 10−4, 3.87 × 10−4 and 5.11 × 10−4 C, for 298, 313
and 323 K respectively). Even when charge increasing is almost
negligible, it could be indicative of a higher interaction between
Cr2O3 and Cl− as temperature increases. However, deeper ana-
lyzes about this kind of interaction are still to be done.
increment in temperature. It is interesting to observe that the
diffusion coefficient is greater at 323 K, suggesting that the dif-
fusion of the chemical species in solution is favored with the
temperature, similar to reported by others authors [1–3,5]. Like-
wise, this result agrees with the predicted by Hirschfelder et
al. [34] and Einstein–Stoke’s equations [13]. Thus, as diffusion
coefficient increases, more chemical species of copper reach the
SS electrode, favoring the nucleation process. Also, it is clear
that the same behavior is obtained in the case of the number of
active nucleation sites. Additionally, at the higher temperature
system, which probably involves a lower adsorption of the chlo-
ride ion, it facilitates the formation of the growing centers of
copper phase on SS.
3.2.2. Analysis of current transients at higher
overpotentials
From last section it is possible to observe that the falling cur-
rent in transients shown in Fig. 4 can be well described by Eq.
(12), while after this initial falling current, the data approaches
to the Scharifker’s et al. model behavior (Fig. 5b). Therefore,
we propose that the transients shown in Fig. 4 in higher overpo-
tentials could be described by following equation:
Heerman and Tarallo [35] have proposed a correction to the
model suggested by Scharifker et al. [21]. Thus the corrected
equation it is expressed as follows:
1
Φ
1/2 Θ
IT = IC + Iet + I3D−dc(SM)
(16)
I3D−dc(HT)(t) = zFDC
(πDt)
Fig. 8 shows the comparison of the experimental current tran-
sients with the theoretically generated by non-linear fitting of
experimental data to Eq. (16). It can be observed, that the model
expressed by Eq. (16) adequately accounts for the behavior of
all experimental transients. Table 3 shows the best-fit nucle-
ation parameters at different temperatures, obtained through the
fitting process of the experimental transients using Eq. (16).
Note that in all cases the nucleation rate and the number of
active nucleation sites increases with the overpotential applied.
At similar overpotential, the nucleation rate increases with the
×(1 − exp[−αSN0(πDt)1/2 1/2
t
Θ])
(17)
(18)
where
ꢆ
(At)1/2
exp(−At)
(At)1/2
Φ = 1 −
exp(λ2) dλ
0
Thus, considering the model proposed by Hermann and Tarallo
and the contribution of the electron transfer and the capacitive
effect, the transients (at higher overpotentials) showed in Fig. 4

900
J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
Table 4
Potential dependence for the physical constants involved during copper reduction, at different temperatures
−E (V)
298 K
−E (V)
313 K
vs. SCE
vs. SCE
aR (A)
bR (s−1
)
A (s−1
)
N0 (×10−6
D (×107
aR (A)
bR (s−1
)
A (s−1
)
N0 (×10−6
D (×107
cm−2
)
cm2 s−1
)
cm−2
)
cm2 s−1
)
0.58
0.60
0.62
0.64
0.66
0.68
0.70
0.72
0.74
0.76
0.78
0.80
0.61
0.63
0.65
0.67
0.69
0.71
0.73
0.75
0.77
0.004
0.004
0.006
0.007
0.001
0.002
0.004
0.005
0.004
0.715
1.070
1.321
1.032
0.496
1.178
1.569
2.292
2.368
1.225
1.947
2.641
2.896
2.862
8.873
11.675
14.158
12.675
13.14
19.98
24.56
30.10
56.92
64.50
72.44
96.57
159.6
2.500
2.500
2.500
2.401
2.401
2.401
2.304
2.304
2.304
0.002
0.003
0.003
0.004
0.005
0.005
0.005
0.006
0.007
0.198
0.238
0.226
0.296
0.488
1.187
1.279
1.652
2.324
0.926
1.511
1.785
2.305
2.754
3.511
3.472
4.849
5.029
2.325
3.426
3.762
4.860
8.959
34.73
53.13
73.17
87.45
2.304
2.209
2.401
2.304
2.304
2.304
2.209
2.209
2.401
−E (V) vs. SCE
323 K
aR (A)
bR (s−1
)
A (s−1
)
N0 (×10−6 cm−2
)
D (×10 cm2 s−1
)
0.597
0.617
0.637
0.657
0.677
0.697
0.717
0.004
0.004
0.005
0.007
0.008
0.010
0.010
0.407
0.705
1.044
1.536
3.177
4.753
0.935
1.495
3.341
4.510
8.194
9.373
5.989
10.57
16.65
25.33
59.64
97.32
8.542
2.809
2.704
2.704
2.704
2.601
2.601
3.025
10. 098
27.291
Data were obtained from best-fit parameters found through the fitting process of the experimental transients using Eq. (16ꢂ).
can be accounted for the following total current equation:
each temperature. This plot is similar to that obtained by using
the HT model (Figure not shown). In the framework of atomistic
theory of nucleation [36–40], it is possible to estimate the critical
size of the nuclei [40,41] with:
IT = IC + Iet + I3D−dc(HT)
(16ꢂ)
Similar fits to those showed in Fig. 8 were obtained employing
Eq. (16ꢂ). The model described by Eq. (16ꢂ) gives similar results
to that obtained by the Scharifker et al. model, see Table 4, and
only slight differences were registered in the N0 value indicating
a strong dependence of the kinetic and nucleation parameters
with temperature.
ꢂ
ꢃ ꢂ
ꢃ
kT
d ln A
dE
nc =
− α
(19)
ze0
where k is the Boltzmann constant, T the absolute temperature
and e0 is the elementary electric charge. By substituting the
values of d ln A/dE at 298, 313 and 323 K from Scharifker’s
model, in Eq. (19) and by considering the value of α found in
this study, it was obtained nc = 0 for every temperature. Same
results were obtained from the HT’s model. Within the atom-
istic theory of nucleation, the largest cluster for which one atom
attachment probability is less than one-half is defined as crit-
ical. Thus, the attachment of a new atom converts this cluster
into a stable one, for which the probability of attachment of
the next atom is already higher than one-half and henceforth is
able to grow spontaneously. Thus, in these cases every active
site onto the substrate surface acts as a critical nucleus [42,43],
suggesting that within the investigated potential interval, the
thermodynamic barrier for the nucleus formation are negligible
in comparison with the kinetics contribution [43]. Moreover,
accordingly with the Arrenihus equation the reaction rates are
function of temperature, suggesting an increase of the nucleation
rate with the temperature. Similar results have been reported for
cobalt and rhodium electrocrystallization [42,43].
Fig. 9 shows a plot of ln A versus Eappl, obtained from Eq.
(16) (see Table 3). Note that a linear relationship was found for
Fig. 9. ln A vs. −Eappl., obtained from data reported in Table 3, at temperatures
indicated in figure.

J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
901
Table 5
Potential dependence of the Ns, calculated from physical constants showed in Table 3 and Eq. (14)
−E (V) vs. SCE
298 K
−E (V) vs. SCE
313 K
−E (V) vs. SCE
323 K
N0 (×10−6 cm−2
)
Ns/N0
Ns (×10−6 cm−2
)
Ns/N0
Ns (×10−6 cm−2
)
Ns/N0
0.58
0.60
0.62
0.64
0.66
0.68
0.70
0.72
0.74
0.76
0.78
0.80
–
–
–
–
–
–
–
–
0.597
0.617
0.637
0.657
0.677
0.697
0.717
–
–
–
–
–
1.541
3.148
4.486
7.617
12.620
16.990
7.658
–
–
–
–
–
0.334
0.366
0.346
0.371
0.256
0.205
1.236
–
–
–
–
–
0.61
0.63
0.65
0.67
0.69
0.71
0.73
0.75
0.77
–
2.214
3.390
4.441
5.082
7.080
13.240
16.280
20.810
25.620
–
0.191
0.198
0.205
0.196
0.139
0.231
0.247
0.234
0.170
–
0.838
1.298
1.520
1.901
2.867
6.408
8.126
11.500
12.350
0.381
0.401
0.404
0.418
0.331
0.189
0.147
0.145
0.140
–
–
–
Through the physical constants reported in Table 3 and 4, the
saturation number of nuclei (Ns) was obtained. This estimation
was made using Eq. (20) [20]:
within an exclusion zone, enhanced by the concentration deple-
tion around centers growing under mass-transfer control in the
vicinity of each nucleus [20].
ꢂ
ꢃ
1/2
AN0
2kꢂD
Ns =
(20)
3.3. Scanning electron microscopy
The results obtained for Ns in this study, employing the data
shown in Tables 3 and 4, are summarized in Tables 5 and 6
for the SM and HT models, respectively. Observe that the Ns
values increase with the applied potential. It is important to men-
tion that, due to the exclusion zones of the deposit, caused by
the hemispherical difusional gradients of 3D nucleus, the Ns
will be always lower than the N0 values in the same applied
potential, and both grow in accordance with a more negative
potential.
Analyzing the Ns/N0 ratio, it can be seen that this value is
small (lower than 0.3 in most of the cases). This result suggests
that the copper electrodeposition is not favored in all initially
available sites. The reason for this behavior could be the con-
version of sites into growing nuclei, and because nucleation is
confined to those active centers that have not been included
Fig. 10 shows the SEM micrographs of copper deposits
obtained at different temperatures, when approximately −1.0 V
was applied during 10 min to the SS/Cu(II)–NH4Cl–NH3–H2O
system. In this figure it can be observed that granular deposits
were obtained for each temperature. The deposits with granular
morphology could be obtained due to the interaction between
the oxide layer (Cr2O3) and Cl−, because of an inhibitory effect
of adsorbed chloride on the initially available active sites, as a
consequence it would produce free zones of growth. An anal-
ysis of the grain density distribution was not possible due to
high grains agglomeration. Instead of, some characteristics of
the deposit could be used to conclude that the above approach
is acceptable. Grain size averages obtained were 1.04, 0.96 and
0.87 ␮m for 298, 313 and 323 K, respectively. If one perform the
profile analysis of the micrographs shown in Fig. 10, it is pos-
Table 6
Potential dependence of the Ns, calculated from physical constants showed in Table 4 and Eq. (14)
−E (V) vs. SCE
298 K
−E (V) vs. SCE
313 K
−E (V) vs. SCE
323 K
Ns (×10−6 cm−2
)
Ns/N0
Ns (×10−6 cm−2
)
Ns/N0
Ns (×10−6 cm−2
)
Ns/N0
0.58
0.60
0.62
0.64
0.66
0.68
0.70
0.72
0.74
0.76
0.78
0.80
–
–
–
–
–
–
–
–
0.597
0.617
0.637
0.657
0.677
0.697
0.717
–
–
–
–
–
1.756
3.488
5.087
8.457
13.880
18.400
8.962
–
–
–
–
–
0.293
0.330
0.305
0.334
0.233
0.189
1.049
–
–
–
–
–
0.61
0.63
0.65
0.67
0.69
0.71
0.73
0.75
0.77
–
2.355
3.661
4.727
5.480
7.492
14.040
17.070
21.700
26.400
–
0.179
0.183
0.192
0.182
0.132
0.218
0.236
0.225
0.165
–
0.861
1.336
1.521
1.965
2.916
6.482
7.972
11.060
12.310
0.370
0.390
0.404
0.404
0.325
0.187
0.150
0.151
0.141
–
–
–

902
J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
Fig. 11. Average profiles of the deposits showed in Fig. 10.
occurs. Note that the deposit obtained at 323 K, exhibits sharp-
ness. Thick and sharp deposits are certainly due to a kinetical
and diffusionally favored system.
4. Conclusions
Temperature effect on electrocrystallization process of cop-
per from the SS/Cu(II)–NH4Cl–NH3–H2O system was ana-
lyzed. From the chronoamperometric study, it was possible to
evaluate the transfer coefficient (α), and the exchange current
density (i0) at different temperatures. The potentiostatic study
suggested that at least three processes are involved: an elec-
tron transfer reaction, a 3D nucleation and growth process and
a capacitive behavior because of the interaction between oxide
layer (Cr2O3)–Cl−, i.e. the combined contribution of the double
layer and the oxide layer barrier on the stainless steel. Analysis
of the transients allowed the obtaining of the nucleation param-
eters values, such as: the nucleation rate constant, the number of
active nucleation sites, the stationary nucleation rate and the sat-
uration number of nuclei, which are favored by the temperature
increment. SEMimagesofthedepositsobtainedatdifferenttem-
peratures suggest that Cl− adsorption plays an important role,
blocking the surface active sites, which inhibits the electrocrys-
tallization process in some zones of stainless steel electrode.
The copper electrocrystallization in zones no inhibited by Cl−
adsorptionfavorstheformationofgranulardepositsandrevealed
that the number of grains and the deposit thickness increases
with temperature. The nucleation kinetic enhancement could be
related with a decrease in adsorption at higher temperature, on
the basis from that revealed with the SEM micrographs. The
increase in temperature favors the growth of the cluster before
the coalescence occurs.
Fig. 10. SEM images recorded for SS/0.2 M Cu(II), 4 M Cl− and NH4+ + NH3
(pH 7.5) system, at: (a) −1.0 (298 K), (b) −1.01 (313 K) and (c) −1.017 V
(323 K) during 10 min.
The nature of oxide layer–Cl− interactions could evolve by
the temperature, from adsorption to chloride-oxide reaction.
Indeed, further experimentation is needed to reveal the temper-
ature effect on this interaction.
sible to get further conclusions. Thus, by employing this skill
we plot the average profiles of the deposits showed in Fig. 10
(see Fig. 11), by the software ImageDig. V.2TM. In this skill the
white zone is considered as closer to viewer while the black zone
has the opposite mean. From Fig. 11, it is clear that at 298 K,
the grains are width and small, while at 323 K the grains are
thick and sharp, an intermediate situation is obtained at 313 K.
Moreover the cluster’s size increases in height and decreases
in width with the temperature increment. Thus, a raise in tem-
perature favors the growth of the cluster before the coalescence
Acknowledgements
This research has been financially supported by CONACyT
– Mexico (project No. 39585-Y). J. Vazquez-Arenas thanks
CONACyT for his postgraduate scholarship. L.H. Mendoza-
Huizar wishes to thank SNI (Mexico) for the stipends received.

J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903
903
We also acknowledge to Professor Ignacio Gonzalez for fruitful
discussions.
[18] A.J. Bard, L.R. Faulkner, Electrochemical Methods. Fundamentals and
Applications, John Wiley & Sons, New York, 2001.
[19] C.H. Hamann, A. Hamnett, Electrochemistry, Wiley-Wch, New York,
1998.
[20] B. Scharifker, G. Hills, Electrochim. Acta 28 (1983) 879.
[21] B.R. Scharifker, J. Mostany, J. Electroanal. Chem. 177 (1984) 13.
[22] M. Palomar-Pardave´, M. Miranda-Herna´ndez, N. Batina, I. Gonza´lez,
Recent Res. Dev. Electrochem. 1 (1998) 15.
References
[1] J. Yu, L. Wang, L. Su, X. Ai, H. Yang, J. Electrochem. Soc. 150 (2003)
C19.
[2] T. Montiel, O. Solorza, H. Sa´nchez, J. Electrochem. Soc. 147 (2000) 1031.
[3] G.R. Pattanaik, D.K. Pandya, S.C. Kashyap, J. Electrochem. Soc. 149
(2002) C363.
[4] S. Peulon, D. Lincot, J. Electrochem. Soc. 145 (1998) 864.
[5] G. Trejo, A.F. Gil, I. Gonza´lez, J. Appl. Electrochem. 26 (1996) 1287.
[6] J.W.P. Schmelzer, Nucleation Theory and Applications, WILEY-VCH Ver-
lag GmbH & Co. KGaA, Weinheim, 2005.
[7] V.M. Fokin, E.D. Zanotto, J. Non-Cryst. Solids 265 (2000) 105.
[8] D. Turnbull, Physics of Non-Crystalline Solids, North-Holland, Amster-
dam, 1964.
[9] I. Gutzow, D. Kashchiev, I. Avramov, J. Non-Cryst. Solids 73 (1985) 477.
[10] A.I. Rusanov, Phasengleichgewichte und Grenzfla¨chenerscheinungen,
Akademie-Verlag, Berlin, 1978.
[11] V.P. Skripov, M.Z. Faizullin, in: J.W.P. Schmelzer (Ed.), Nucleation Theory
and Applications, Wiley-VCH, Berlin, 2005, pp. 4–38.
[12] M.Y. Abyaneh. The electrocrystallization of Niquel. Ph.D. Thesis, Univer-
sity of Southampton, Department of Chemistry, London, 1980.
[13] J.R. Welty, C.E. Wicks, Robert E. Wilson, Fundamentos de Transferencia
de momento calor y masa, Editorial Limusa. Noriega editores, Me´xico,
1991, p. 550.
[14] J. Vazquez-Arenas, Study of the involved processes in the electrowinning
of copper from NH4Cl-NH3-H2O solutions. M. Sc. Thesis, Universidad
Auto´noma de San Luis Potos´ı, Instituto de Metalurgia, San Luis Potos´ı,
Mexico, 2006.
[23] L.H. Mendoza-Huizar, J. Robles, M. Palomar-Pardave´, J. Electroanal.
Chem. 545 (2003) 39.
[24] M.H. Ho¨lzle, U. Retter, D.M. Kolb, J. Electroanal. Chem. 371 (1994) 101.
[25] A. Milchev, T. Zapryanova, Electrochim. Acta 51 (2006) 2926.
[26] A. Milchev, T. Zapryanova, Electrochim. Acta 51 (2006) 4916.
[27] A. Milchev, Electrocrystallization: Fundamentals of Nucleation and
Growth, Kluwer Academic Publishers, Boston/Dordrecht/London, 2002.
[28] D.J.G. Ives, G.J. Janz, Reference Electrodes Theory and Practice, Aca-
demic Press, New York, 1961.
[29] T. Pauporte´, J. Finne, J. Appl. Electrochem. 36 (2006) 33.
[30] W. Schmickler, Electrochim. Acta 41 (1996) 2329.
[31] L.M. Torres, A. Gil, L. Galicia, I. Gonza´lez, J. Chem. Educ. 73 (1996) 808.
[32] H. Konno, M. Nagayama, Electrochim. Acta 22 (1977) 353.
[33] B.R. Scharifker, Acta Sci. Venez. 35 (1984) 211.
[34] J.O. Hirschfelder, R.B. Bird, E.L. Spotz, Chem. Rev. 44 (1949) 205–
231.
[35] L. Heerman, A. Tarallo, J. Electroanal. Chem. 470 (1999) 70.
[36] A. Milchev, S. Stoyanov, R. Kaischev, Thin Solid Films 22 (1974)
255.
[37] A. Milchev, S. Stoyanov, R. Kaischev, Thin Solid Films 22 (1974)
267.
[38] A. Milchev, S. Stoyanov, J. Electroanal. Chem. 72 (1976) 33.
[39] A. Milchev, E. Vassileva, J. Electroanal. Chem. 107 (1980) 323.
[40] A. Milchev, Contemp. Phys. 32 (1991) 321.
[15] J. Vazquez, R. Cruz, ECS Trans. 2 (3) (2006) 355.
[16] A. Ramos, M. Miranda-Herna´ndez, I. Gonza´lez, J. Electrochem. Soc. 148
(2001) C315.
[41] D. Kaschiev, J. Chem. Phys. 76 (1982) 5098.
[42] L.H. Mendoza-Huizar, J. Robles, M. Palomar-Pardave´, J. Electroanal.
Chem. 521 (2002) 95.
[17] D. Pletcher, R. Greef, R. Peat, L.M. Peter, J. Robinson, Instrumental Meth-
ods in Electrochemistry, Horwood Publishing, Chichester, 2001.
[43] M. Arbib, B. Zhang, V. Lazarov, D. Stoychev, A. Milchev, C. Buess-
Herman, J. Electroanal. Chem. 510 (2001) 67.