You are on page 1of 13

This article was originally published in a journal published by

Elsevier, and the attached copy is provided by Elsevier for the


author’s benefit and for the benefit of the author’s institution, for
non-commercial research and educational use including without
limitation use in instruction at your institution, sending it to specific
colleagues that you know, and providing a copy to your institution’s
administrator.
All other uses, reproduction and distribution, including without
limitation commercial reprints, selling or licensing copies or access,
or posting on open internet sites, your personal or institution’s
website or repository, are prohibited. For exceptions, permission
may be sought for such use through Elsevier’s permissions site at:

http://www.elsevier.com/locate/permissionusematerial
Electrochimica Acta 52 (2006) 892–903

The role of temperature in copper electrocrystallization in

py
ammonia–chloride solutions
Jorge Vazquez-Arenas a , Roel Cruz a,∗ , Luis H. Mendoza-Huizar b

co
a Facultad de Ingenierı́a, Instituto de Metalurgia, Universidad Autónoma de San Luis Potosı́, Av. Sierra Leona 550,
Lomas 2a sección, 78210 San Luis Potosı́, S.L.P., México
b Centro de Investigaciones Quı́micas, Universidad Autónoma del Estado de Hidalgo, Unidad Universitaria,

Km 4.5 Carretera Pachuca-Tulancingo, CP. 42184 Pachuca-Hidalgo, México


Received 2 March 2006; received in revised form 20 June 2006; accepted 21 June 2006
Available online 4 August 2006

al
Abstract
This paper presents the analysis of temperature effect on the copper electrocrystallization process from the stainless steel/Cu(II)–Cu(I)–
on
NH4 Cl–NH3 –H2 O 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
rs

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 (Cr2 O3 )–chloride
interactions on the stainless steel surface. Current transients analyses at different potentials, based on the models of: Milchev (low overpotential),
pe

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


r's

1. Introduction perature dependence of the nucleation rate is determined mainly


by the thermodynamic and kinetic barriers for nucleation. Since
o

Electrocrystallization is a complex process affected by factors a temperature decrease produces two effects: a decrease of the
such as: substrate, additives, electroactive species, bath compo- thermodynamic barrier, due to an increase in the thermodynamic
th

sition, mass transport, temperature, etc. Although, the effect of driving force for crystallization, leading to a higher nucleation
temperature is well known, only few studies have been reported rate and an increase of the kinetic barrier leading to a lower
about this factor effect [1–5]. Such studies indicate that nucle- nucleation rate [6].
ation and growth rate increase with the temperature, which could However, a contrary behavior has been observed [1–5],
Au

intrinsically suggests a diminution of energy requirement for the which is probably because the value of the surface energy
electrodeposition process. The Gibbs’s energies as well as the (σ) leads to a slight increase with increasing temperature
activation energies are potential-dependent, but the temperature- (dσ/dT ∼ (0.06–0.16) × 10−3 J/m2 K) regardless of the method
dependence of these quantities is at all not obvious. The tem- 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
∗ Corresponding author. Tel.: +52 444 825 43 26; fax: +52 444 825 43 26. from nucleation rate data, refers to nuclei of critical size) [6].
E-mail address: rcuz@uaslp.mx (R. Cruz). Curvature corrections can be expected to lead to a reduction in

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

py
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)–NH4 Cl–NH3 –H2 O. For this study,

co
electrochemical techniques and scanning electron microscopy
with energy dispersion spectroscopy were utilized. This study
is part of an investigative work aimed at the understanding 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
of electrochemical systems based on Cu(II)–NH4 Cl–NH3 –H2 O
figure. Scan rate: 100 mV s−1 .
electrolytes [14,15].
ing bath was used to fix and to control temperature within

al
2. Methodology ±1 K.
A conventional three-electrode cell (100 mL) was used for
The experiments were carried out in 0.2 M Cu(II), 4 M the electrochemical experiments. A 304 stainless steel (∼19%
on
NH3 + NH4 + and 4 M Cl− electrolyte. All solutions were pre- Cr, ∼08% Ni and ∼2% Mn) rod was embedded in resin to
pared using ultrapure water (18.2 ␮ cm NanopureDiamond generate a rotating disk working electrode with a 0.125 cm2
TM
(Barnstead ) and CuCl2 ·2H2 O (Fisher), NH4 Cl (Fisher) and area. A graphite rod (Alfa AESAR, 99.999%) was utilized as
NH4 OH (Fisher) analytic grade reagents. In each case, the counter electrode and a saturated calomel electrode (SCE) was
rs

pH was controlled by addition of 4 M NaOH (Fisher) and the reference electrode; all reported potentials are referred to this
10 M H2 SO4 (J.T. Baker) solutions, and by a pH meter scale. Prior to each experiment, working electrode surface was
TM TM
Termo Orion 420A. A 10 L Polyscience heater circulat- polished to mirror finish using 0.05 ␮m Buehler alumina pow-
pe
o r's
th
Au

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. Note that the current peak Ic increases as temperature,
Inert (nitrogen) atmosphere was maintained during all experi- which could be due to enhancement of kinetic and diffusive
ments. The working electrode potential was controlled by an processes, such as it is indicated by a better defined peak.
TM
AUTOLAB model PGSTAT 30 potenstiotat–galvanostat cou- However, this current increment could also suggest a diminu-
TM
pled to a personal computer using GPES software. tion of the Cr2 O3 –chloride interaction, which persist even at
323 K.
Fig. 2 shows a set of voltammograms obtained for different
3. Results and discussion potential scan rates and temperatures. Observe that for all cases,
a fast increase in cathodic current was obtained after the slow

py
3.1. Voltammetric study current step (−E > 0.5 V). This result indicates that a favorable
kinetic condition for the copper electrodeposition (IIIc peak)
Fig. 1 show the voltammetric direct scans from the 0.2 M on the substrate surface has been reached. Furthermore, it can
Cu(II), 4 M NH3 + NH4 + and 4 M Cl− solution at 298, 313, 323 be observed that the cathodic current associated with peak IIIc

co
and 343 K, at pH 7.5 on the SS substrate. At 298 K, it was increases with the temperature. Similar results were obtained
possible to observe three cathodic peaks: Ic, IIc and IIIc, at by Ramos et al. [16], where the electrocrystallization stage of
−0.380, −0.550 and −0.650 V, respectively (see Fig. 1). In a (Cu(NH3 )2 + /Cu(0)) was explained in terms of diffusional con-
recent work, Vazquez-Arenas et al. have shown that, for this trol.
system, the reduction process for predominant chemical species The peak IIIc formation is certainly due to species deple-
(Cu(NH3 )3 Cl+ complex) involves the following steps [14,15]: tion in the electrode/solution interface. Observe that a gradual

al
shifting of peak (Ep ) toward more negative values occurs as the
(a) Cu(NH3 )3 Cl+ /Cu(NH3 )2 Cl, occurring in the region labeled scan rate (v) increases. In order to determine the control type
as Ic. The low current obtained in this step has been limiting this process, the current value associated with the peak
on
attributed to a slow reduction process due to the barrier (−Ip ) was plotted as a function of v1/2 (Fig. 3). Note that a
formed by the oxide layer (Cr2 O3 )–chloride interactions in reversible behavior is obtained approximately at scan rate lower
the SS surface. than 50 mV s−1 (Fig. 3a), 80 (Fig. 3b) and 50 mV s−1 (Fig. 3c)
(b) Cu(NH3 )2 Cl/Cu(0), occurring in the region labeled as IIc. for 298, 313 and 323 K, respectively. On the other hand, an irre-
rs

This step is characterized by an important increase in the versible behavior was observed at higher scan rates. This feature
cathodic current. is characteristic when a process passes through a region known
(c) A faster reduction Cu(NH3 )3 Cl+ /Cu(0) labeled as peak IIIc, as quasi-reversible [17]. The linearity showed by the −Ip ver-
which take place due to the higher overpotential. sus v1/2 at the different temperatures, suggests an irreversible
pe
o r's
th
Au

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

py
co
al
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
on
in figure. System temperature: (a) 298, (b) 313 and (c) 323 K.

reduction process occurring under diffusion control [17–19], ficient and C is the bulk concentration of the electroactive
according with the following equation: species.
rs

The absolute value between the cathodic potential peak (Ep )


Ip = 2.99 × 105 n(αnα )1/2 AD1/2 Cv1/2 (1) and the cathodic potential half-peak (Ep/2 ) was obtained from
the voltammograms shown in Fig. 2. The calculated values were
pe

where n is the number of the transferred electrons in the reduc- quite close to the expected for an irreversible system [18,19].
tion process, nα the number of electrons transferred up to, These results newly suggest that the process is changing, from
and including, the rate determining step, D the diffusion coef- reversible to irreversible system.
o r's
th
Au

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) (6)
The charge transfer and nucleation kinetic of copper elec-
trocrystallization process on SS electrode, were studied from Cu+ + e− → Cu0 (fast) (7)
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 The slow step (Eq. (6)) may lead to the accumulation of Cu+ at
shows a set of current transients obtained from different over- the electrode surface if the nucleation and growth of copper clus-
potentials and at differente temperatures. All transients shown ters is not sufficiently fast [25,26]. Accordingly with Milchev et
in the figure exhibited a falling current in the initial period of al., an electron transfer reaction takes place prior to and simulta-

py
time. After this falling current, the transients showed a typical neously with the process of the nucleus formation. The current
current maximum (Im ), related with the coalescence of diffusion density associated to these processes, before the nuclei overlap-
fields with spherical symmetry. Once the current maximum was ping, but when the crystal formation has already started, is given
reached, a decay of the current was obtained, which approaches by:

co
to a planar diffusion. The shape of these transients exhibited the 3
IM = aR exp(−bR t) + q[(1 + 2nt)1/2 − 1] (8)
features of typical three dimensional nucleation process with
hemispherical diffusion control (3D-dc) of the growing crystal- where
lites [20–22]. The nucleation model proposed by Scharifker and  
sπF4 (DC)3 3αFη
Hills allows classifying the nucleation and growth mechanism q= Ist exp − (9)
as instantaneous (Eq. (2)) or progressive (Eq. (3)), by compari- i0 VM RT
vi0 VM Q

al
son between the dimensionless curves of the experimental and
n= (10)
theoretical transients [20]. F 2 DC
   2 with s = 32/3 and v = 1/4 and
I2 t
1.9542
on
= 1 − exp −1.2564 (2)    
Im2 t/tm tm 2αFη −2(1 − α)Fη
Q = exp − exp (11)
   2 RT RT
I2 1.2254 t
= 1 − exp −2.3367 (3) i0 is the exchange current density at the “copper–solution” inter-
Im2 t/tm tm
rs

face, α the charge transfer coefficient, η the overpotential, Ist the


Fig. 5 shows a comparison of the theoretical dimension- stationary nucleation rate and Vm is the volume molar. All other
less transients generated by Eqs. (4) and (5), with experimental parameters have their conventional meanings. Since the values
pe

dimensionless current transients obtained at different tempera- obtained for the number of reduced species were higher than an
tures. It is clear that at 313 K (Fig. 5b), most of the experimental effective monolayer, first term in Eq. (8) has been related with the
data fall within the range of validity of the theory proposed by electron transfer reaction, i.e. Iet = aR exp(−bR t), discarding an
Scharifker et al. [20,21]. Last result suggests that the experimen- ion transfer reaction [25,26]. The second term describes the cur-
tal transients could be fitted using the following equation: rent of progressive nucleation and growth of the copper crystals
on the substrate, under combined “charge transfer” and “diffu-
1
I3D−dc(SM) (t) = zFDC sion limitations” [27]. Therefore, the region of current falling
(πDt)1/2
r's

observed at short times may be related with the discharge of the


×[1 − exp(−αS N0 (πDt)1/2 t 1/2 Θ)] (4) copper complexes, in two monoelectronic steps.
In order to evaluate the existence of a similar behavior for the
with system here analyzed, Milchev’s model was applied to poten-
o

1 − exp(−At) tiostatic curves obtained at low overpotentials and short times,


Θ=1− (5)
At ensuring that any overlapping and coalescence of the clusters do
th

not occur during the growth. In this analysis, the model proposed
where z is the number of exchanged electrons, F the Faraday
by Milchev et al., could not account for the transient behavior at
constant, N0 the number of active nucleation sites, A the nucle-
shorter times than 0.4 s. The capacitive effect generated by the
ation rate and αS = 2[2Vm DC]1/2 .
Au

oxide layer (Cr2 O3 )–chloride interactions, suggested by voltam-


Observe that the values (I/Im ) at 298 and 323 K (Fig. 5a and
metry, involves the contribution of two capacitances in series,
c) are always lower than those obtained from the experimental
namely the oxide film thicknesses barrier, and the double layer
data. This difference might be associated to the contribution, to
capacitance, which could be influenced by chloride adsorption
the total current, of another processes not considered by the non-
and/or its reaction with oxide layer. A detailed study of this two
dimensional analysis [22–24]. Also, note that the falling current
capacitance contributions has been reported elsewhere [24,29].
cannot be accounted for the Scharifker’s model.
Hence, taking into account the above described, the initial falling
cathodic currents observed in the experimental current transients
3.2.1. Analysis of current transients at low overpotentials (IT ) obtained during copper deposition on SS electrode, may be
Recently, Milchev et al. [25] reported that, for the case of described by the following equation:
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

py
co
al
on
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.
rs

where IC = k1 exp(−k2 × t) is the current due to the capacitive (outer sphere reaction), in agreement with the reported in the
effect of Cr2 O3 –Cl− interaction, being k1 = k2 QC , where QC is literature [30,31].
the charge density due to the overall capacitance effect. An average value of 0.0456, 0.0487 and 0.0509 was obtained
pe

Fig. 6 shows the comparison of the experimental current for the i0 , at temperatures of 298, 313 and 323 K, respectively.
transients obtained at low overpotentials with the theoretically The value obtained at 298 K is similar to that reported in litera-
generated by non-linear fitting of experimental data by Eq. (12). ture in a case of a copper reduction from sulfate solutions [25].
It can be observed, that the model expressed by Eq. (12) ade- The (ln i0 ) was plotted as a function of (1/T), Fig. 7, exhibiting
quately accounts for the behavior of all experimental transients. a linear relationship, according with the Arrhenius’s equation
In Table 1, the best-fit kinetic and nucleation parameters at [19]. From the slope value and Eq. (13) it was possible to calcu-
different temperatures are shown. In this table, the reference late the activation energy (H = ) of the system as 50.63 KJ/mol.
r's

electrode potential (SCE) values for 313 and 323 K were cor- Last result compares favorably with the obtained in the case of
rected according with the methodology reported in literature electrodeposition of copper from cupric pyrophosphate solu-
[28]. It is important to mention, that the charge transfer coeffi- tions [32].
cients obtained from the fitting parameters of Table 1 and Eq. (9),
o

do not vary significantly with temperature, and they are around


0.5. This fact is indicative that the reduction reaction involv- ∂ ln i0 H =
=− (13)
ing Cu(NH3 )3 Cl+ /Cu(NH3 )2 Cl is an electron transfer reaction ∂(1/T ) R
th

Table 1
Potential dependence from kinetic and nucleation best-fit parameters at different temperatures, obtained through the fitting process of the experimental transients
Au

using Eq. (12)


−E (V) vs. SCE 298 K −E (V) vs. SCE 313 K −E (V) vs. SCE 323 K

aR (A) bR (s−1 ) Ist (cm2 s−1 ) aR (A) bR (s−1 ) Ist (cm2 s−1 ) aR (A) bR (s−1 ) Ist (cm2 s−1 )

0.367 0.0017 0.173 3835.095


0.4 0.0006 0.136 1390.488 0.41 0.0008 0.143 2315.287 0.417 0.0025 0.226 4447.202
0.45 0.0010 0.155 2492.439 0.46 0.0013 0.165 2850.112 0.447 0.0033 0.257 5124.677
0.5 0.0013 0.175 2076.329 0.51 0.0019 0.199 3411.408 0.467 0.0037 0.294 5618.889
0.55 0.0018 0.189 2412.502 0.55 0.0029 0.232 4032.103 0.487 0.0031 0.237 6590.043
0.6 0.0022 0.207 2921.393 0.56 0.0029 0.212 4081.080 0.497 0.0043 0.534 8110.426
0.62 0.0023 0.243 3573.664 0.58 0.0045 0.729 8182.813 0.517 0.0052 0.717 9064.838
0.6 0.0028 0.191 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:
 

py
Ist 1/2
Ns = (14)
2k D
where

co
k = [8πCVM ]1/2 (15)
Results obtained for Ns are summarized in Table 2. These
Fig. 7. Arrhenius equation linear plot (ln i0 vs. 1/T) for reaction results reveal that the Ns value increases with the applied overpo-
Cu(NH3 )3 Cl+ /Cu(0).
tential and with temperature at same overpotential value, remark-
ing an important role of these parameters in the generation of a
high saturation nucleus density. The charge contribution of the

al
Table 2
Potential dependence of the Ns , calculated from physical constants showed in Table 1 and Eq. (14)
on
−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 36.327
0.4 21.874 0.41 28.225 0.417 39.119
0.45 29.285 0.46 31.316 0.447 41.993
rs

0.5 26.729 0.51 34.261 0.467 43.971


0.55 28.812 0.55 37.248 0.487 47.620
0.6 31.705 0.56 37.474 0.497 52.828
0.62 35.067 0.58 53.063 0.517 55.850
pe

0.6 53.933
o r's
th
Au

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 0.61 0.004 0.715 1.225 11.61 2.828
0.62 0.63 0.004 1.070 1.947 17.18 2.907

py
0.64 0.002 0.198 0.926 2.202 2.428 0.65 0.006 1.321 2.641 21.67 2.833
0.66 0.003 0.238 1.511 3.235 2.339 0.67 0.007 1.032 2.898 25.87 2.794
0.68 0.003 0.226 1.785 3.759 2.665 0.69 0.001 0.496 2.862 50.83 2.688
0.70 0.004 0.358 2.305 4.552 2.460 0.71 0.002 1.178 8.873 57.30 2.703
0.72 0.005 0.488 2.754 8.661 2.383 0.73 0.004 1.569 11.675 65.88 2.534

co
0.74 0.005 1.187 3.510 33.95 2.356 0.75 0.005 2.292 14.158 88.79 2.506
0.76 0.005 1.279 3.472 55.20 2.126 0.77 0.004 2.368 12.675 150.3 2.446
0.78 0.006 1.652 4.849 79.15 2.042
0.80 0.008 2.269 5.029 88.01 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 )

al
0.597 0.004 0.407 1.495 4.613 3.647
0.617 0.004 0.705 3.341 8.606 3.322
0.637 0.005 1.044 4.510 12.95 3.477
0.657 0.007 1.536 8.194 20.55 3.333
on
0.677 0.008 3.177 9.373 49.33 3.144
0.697 0.010 4.753 10.098 82.93 3.052
0.717 0.009 0.890 27.477 6.194 4.116

Data were obtained from best-fit parameters found through the fitting process of the experimental transients using Eq. (16).
rs

capacitive effect generated by the oxide layer (Cr2 O3 )–chloride increment in temperature. It is interesting to observe that the
pe

interactions shows an increase as a function of the tempera- diffusion coefficient is greater at 323 K, suggesting that the dif-
ture (2.08 × 10−4 , 3.87 × 10−4 and 5.11 × 10−4 C, for 298, 313 fusion of the chemical species in solution is favored with the
and 323 K respectively). Even when charge increasing is almost temperature, similar to reported by others authors [1–3,5]. Like-
negligible, it could be indicative of a higher interaction between wise, this result agrees with the predicted by Hirschfelder et
Cr2 O3 and Cl− as temperature increases. However, deeper ana- al. [34] and Einstein–Stoke’s equations [13]. Thus, as diffusion
lyzes about this kind of interaction are still to be done. coefficient increases, more chemical species of copper reach the
SS electrode, favoring the nucleation process. Also, it is clear
r's

3.2.2. Analysis of current transients at higher that the same behavior is obtained in the case of the number of
overpotentials active nucleation sites. Additionally, at the higher temperature
From last section it is possible to observe that the falling cur- system, which probably involves a lower adsorption of the chlo-
rent in transients shown in Fig. 4 can be well described by Eq. ride ion, it facilitates the formation of the growing centers of
o

(12), while after this initial falling current, the data approaches copper phase on SS.
to the Scharifker’s et al. model behavior (Fig. 5b). Therefore, Heerman and Tarallo [35] have proposed a correction to the
th

we propose that the transients shown in Fig. 4 in higher overpo- model suggested by Scharifker et al. [21]. Thus the corrected
tentials could be described by following equation: equation it is expressed as follows:

IT = IC + Iet + I3D−dc(SM) 1 Φ
Au

(16) I3D−dc(HT) (t) = zFDC 1/2 Θ


(πDt)
Fig. 8 shows the comparison of the experimental current tran-
sients with the theoretically generated by non-linear fitting of ×(1 − exp[−αS N0 (πDt)1/2 t 1/2 Θ]) (17)
experimental data to Eq. (16). It can be observed, that the model
where
expressed by Eq. (16) adequately accounts for the behavior of
all experimental transients. Table 3 shows the best-fit nucle-  (At)1/2
exp(−At)
ation parameters at different temperatures, obtained through the Φ=1− exp(λ2 ) dλ (18)
(At)1/2 0
fitting process of the experimental transients using Eq. (16).
Note that in all cases the nucleation rate and the number of Thus, considering the model proposed by Hermann and Tarallo
active nucleation sites increases with the overpotential applied. and the contribution of the electron transfer and the capacitive
At similar overpotential, the nucleation rate increases with the 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 0.61 0.004 0.715 1.225 13.14 2.500
0.62 0.63 0.004 1.070 1.947 19.98 2.500

py
0.64 0.002 0.198 0.926 2.325 2.304 0.65 0.006 1.321 2.641 24.56 2.500
0.66 0.003 0.238 1.511 3.426 2.209 0.67 0.007 1.032 2.896 30.10 2.401
0.68 0.003 0.226 1.785 3.762 2.401 0.69 0.001 0.496 2.862 56.92 2.401
0.70 0.004 0.296 2.305 4.860 2.304 0.71 0.002 1.178 8.873 64.50 2.401
0.72 0.005 0.488 2.754 8.959 2.304 0.73 0.004 1.569 11.675 72.44 2.304

co
0.74 0.005 1.187 3.511 34.73 2.304 0.75 0.005 2.292 14.158 96.57 2.304
0.76 0.005 1.279 3.472 53.13 2.209 0.77 0.004 2.368 12.675 159.6 2.304
0.78 0.006 1.652 4.849 73.17 2.209
0.80 0.007 2.324 5.029 87.45 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 )

al
0.597 0.004 0.407 1.495 5.989 2.809
0.617 0.004 0.705 3.341 10.57 2.704
0.637 0.005 1.044 4.510 16.65 2.704
0.657 0.007 1.536 8.194 25.33 2.704
on
0.677 0.008 3.177 9.373 59.64 2.601
0.697 0.010 4.753 10. 098 97.32 2.601
0.717 0.010 0.935 27.291 8.542 3.025

Data were obtained from best-fit parameters found through the fitting process of the experimental transients using Eq. (16 ).
rs

can be accounted for the following total current equation: each temperature. This plot is similar to that obtained by using
pe

the HT model (Figure not shown). In the framework of atomistic


IT = IC + Iet + I3D−dc(HT) (16 ) theory of nucleation [36–40], it is possible to estimate the critical
size of the nuclei [40,41] with:
Similar fits to those showed in Fig. 8 were obtained employing
Eq. (16 ). The model described by Eq. (16 ) gives similar results   
to that obtained by the Scharifker et al. model, see Table 4, and kT d ln A
nc = −α (19)
only slight differences were registered in the N0 value indicating ze0 dE
a strong dependence of the kinetic and nucleation parameters
r's

with temperature. where k is the Boltzmann constant, T the absolute temperature


Fig. 9 shows a plot of ln A versus Eappl , obtained from Eq. and e0 is the elementary electric charge. By substituting the
(16) (see Table 3). Note that a linear relationship was found for 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
o

this study, it was obtained nc = 0 for every temperature. Same


results were obtained from the HT’s model. Within the atom-
th

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
Au

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
Fig. 9. ln A vs. −Eappl. , obtained from data reported in Table 3, at temperatures rate with the temperature. Similar results have been reported for
indicated in figure. cobalt and rhodium electrocrystallization [42,43].
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 – – – – 0.597 1.541 0.334


0.60 – – 0.61 2.214 0.191 0.617 3.148 0.366
0.62 – – 0.63 3.390 0.198 0.637 4.486 0.346
0.64 0.838 0.381 0.65 4.441 0.205 0.657 7.617 0.371

py
0.66 1.298 0.401 0.67 5.082 0.196 0.677 12.620 0.256
0.68 1.520 0.404 0.69 7.080 0.139 0.697 16.990 0.205
0.70 1.901 0.418 0.71 13.240 0.231 0.717 7.658 1.236
0.72 2.867 0.331 0.73 16.280 0.247 – – –
0.74 6.408 0.189 0.75 20.810 0.234 – – –

co
0.76 8.126 0.147 0.77 25.620 0.170 – – –
0.78 11.500 0.145 – – – – – –
0.80 12.350 0.140 – – – – – –

Through the physical constants reported in Table 3 and 4, the within an exclusion zone, enhanced by the concentration deple-
saturation number of nuclei (Ns ) was obtained. This estimation tion around centers growing under mass-transfer control in the

al
was made using Eq. (20) [20]: vicinity of each nucleus [20].
 1/2
AN0
on
Ns = (20)
2k D 3.3. Scanning electron microscopy

The results obtained for Ns in this study, employing the data Fig. 10 shows the SEM micrographs of copper deposits
shown in Tables 3 and 4, are summarized in Tables 5 and 6 obtained at different temperatures, when approximately −1.0 V
rs

for the SM and HT models, respectively. Observe that the Ns was applied during 10 min to the SS/Cu(II)–NH4 Cl–NH3 –H2 O
values increase with the applied potential. It is important to men- system. In this figure it can be observed that granular deposits
tion that, due to the exclusion zones of the deposit, caused by were obtained for each temperature. The deposits with granular
pe

the hemispherical difusional gradients of 3D nucleus, the Ns morphology could be obtained due to the interaction between
will be always lower than the N0 values in the same applied the oxide layer (Cr2 O3 ) and Cl− , because of an inhibitory effect
potential, and both grow in accordance with a more negative of adsorbed chloride on the initially available active sites, as a
potential. consequence it would produce free zones of growth. An anal-
Analyzing the Ns /N0 ratio, it can be seen that this value is ysis of the grain density distribution was not possible due to
small (lower than 0.3 in most of the cases). This result suggests high grains agglomeration. Instead of, some characteristics of
that the copper electrodeposition is not favored in all initially the deposit could be used to conclude that the above approach
r's

available sites. The reason for this behavior could be the con- is acceptable. Grain size averages obtained were 1.04, 0.96 and
version of sites into growing nuclei, and because nucleation is 0.87 ␮m for 298, 313 and 323 K, respectively. If one perform the
confined to those active centers that have not been included profile analysis of the micrographs shown in Fig. 10, it is pos-
o

Table 6
th

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


Au

0.58 – – – – 0.597 1.756 0.293


0.60 – – 0.61 2.355 0.179 0.617 3.488 0.330
0.62 – – 0.63 3.661 0.183 0.637 5.087 0.305
0.64 0.861 0.370 0.65 4.727 0.192 0.657 8.457 0.334
0.66 1.336 0.390 0.67 5.480 0.182 0.677 13.880 0.233
0.68 1.521 0.404 0.69 7.492 0.132 0.697 18.400 0.189
0.70 1.965 0.404 0.71 14.040 0.218 0.717 8.962 1.049
0.72 2.916 0.325 0.73 17.070 0.236 – – –
0.74 6.482 0.187 0.75 21.700 0.225 – – –
0.76 7.972 0.150 0.77 26.400 0.165 – – –
0.78 11.060 0.151 – – – – – –
0.80 12.310 0.141 – – – – – –
902 J. Vazquez-Arenas et al. / Electrochimica Acta 52 (2006) 892–903

py
Fig. 11. Average profiles of the deposits showed in Fig. 10.

co
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

al
Temperature effect on electrocrystallization process of cop-
per from the SS/Cu(II)–NH4 Cl–NH3 –H2 O system was ana-
on
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
rs

a capacitive behavior because of the interaction between oxide


layer (Cr2 O3 )–Cl− , i.e. the combined contribution of the double
layer and the oxide layer barrier on the stainless steel. Analysis
pe

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. SEM images of the deposits obtained at different tem-
peratures suggest that Cl− adsorption plays an important role,
blocking the surface active sites, which inhibits the electrocrys-
r's

tallization process in some zones of stainless steel electrode.


The copper electrocrystallization in zones no inhibited by Cl−
adsorption favors the formation of granular deposits and revealed
that the number of grains and the deposit thickness increases
o

with temperature. The nucleation kinetic enhancement could be


related with a decrease in adsorption at higher temperature, on
th

Fig. 10. SEM images recorded for SS/0.2 M Cu(II), 4 M Cl− and NH4 + + NH3 the basis from that revealed with the SEM micrographs. The
(pH 7.5) system, at: (a) −1.0 (298 K), (b) −1.01 (313 K) and (c) −1.017 V increase in temperature favors the growth of the cluster before
(323 K) during 10 min.
the coalescence occurs.
Au

The nature of oxide layer–Cl− interactions could evolve by


sible to get further conclusions. Thus, by employing this skill the temperature, from adsorption to chloride-oxide reaction.
we plot the average profiles of the deposits showed in Fig. 10 Indeed, further experimentation is needed to reveal the temper-
(see Fig. 11), by the software ImageDig. V.2TM . In this skill the ature effect on this interaction.
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, Acknowledgements
the grains are width and small, while at 323 K the grains are
thick and sharp, an intermediate situation is obtained at 313 K. This research has been financially supported by CONACyT
Moreover the cluster’s size increases in height and decreases – Mexico (project No. 39585-Y). J. Vazquez-Arenas thanks
in width with the temperature increment. Thus, a raise in tem- CONACyT for his postgraduate scholarship. L.H. Mendoza-
perature favors the growth of the cluster before the coalescence 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 [18] A.J. Bard, L.R. Faulkner, Electrochemical Methods. Fundamentals and
discussions. Applications, John Wiley & Sons, New York, 2001.
[19] C.H. Hamann, A. Hamnett, Electrochemistry, Wiley-Wch, New York,
1998.
References [20] B. Scharifker, G. Hills, Electrochim. Acta 28 (1983) 879.
[21] B.R. Scharifker, J. Mostany, J. Electroanal. Chem. 177 (1984) 13.
[1] J. Yu, L. Wang, L. Su, X. Ai, H. Yang, J. Electrochem. Soc. 150 (2003) [22] M. Palomar-Pardavé, M. Miranda-Hernández, N. Batina, I. González,
C19. Recent Res. Dev. Electrochem. 1 (1998) 15.
[2] T. Montiel, O. Solorza, H. Sánchez, J. Electrochem. Soc. 147 (2000) 1031. [23] L.H. Mendoza-Huizar, J. Robles, M. Palomar-Pardavé, J. Electroanal.
[3] G.R. Pattanaik, D.K. Pandya, S.C. Kashyap, J. Electrochem. Soc. 149 Chem. 545 (2003) 39.

py
(2002) C363. [24] M.H. Hölzle, U. Retter, D.M. Kolb, J. Electroanal. Chem. 371 (1994) 101.
[4] S. Peulon, D. Lincot, J. Electrochem. Soc. 145 (1998) 864. [25] A. Milchev, T. Zapryanova, Electrochim. Acta 51 (2006) 2926.
[5] G. Trejo, A.F. Gil, I. González, J. Appl. Electrochem. 26 (1996) 1287. [26] A. Milchev, T. Zapryanova, Electrochim. Acta 51 (2006) 4916.
[6] J.W.P. Schmelzer, Nucleation Theory and Applications, WILEY-VCH Ver- [27] A. Milchev, Electrocrystallization: Fundamentals of Nucleation and
lag GmbH & Co. KGaA, Weinheim, 2005. Growth, Kluwer Academic Publishers, Boston/Dordrecht/London, 2002.

co
[7] V.M. Fokin, E.D. Zanotto, J. Non-Cryst. Solids 265 (2000) 105. [28] D.J.G. Ives, G.J. Janz, Reference Electrodes Theory and Practice, Aca-
[8] D. Turnbull, Physics of Non-Crystalline Solids, North-Holland, Amster- demic Press, New York, 1961.
dam, 1964. [29] T. Pauporté, J. Finne, J. Appl. Electrochem. 36 (2006) 33.
[9] I. Gutzow, D. Kashchiev, I. Avramov, J. Non-Cryst. Solids 73 (1985) 477. [30] W. Schmickler, Electrochim. Acta 41 (1996) 2329.
[10] A.I. Rusanov, Phasengleichgewichte und Grenzflächenerscheinungen, [31] L.M. Torres, A. Gil, L. Galicia, I. González, J. Chem. Educ. 73 (1996) 808.
Akademie-Verlag, Berlin, 1978. [32] H. Konno, M. Nagayama, Electrochim. Acta 22 (1977) 353.
[11] V.P. Skripov, M.Z. Faizullin, in: J.W.P. Schmelzer (Ed.), Nucleation Theory [33] B.R. Scharifker, Acta Sci. Venez. 35 (1984) 211.
and Applications, Wiley-VCH, Berlin, 2005, pp. 4–38. [34] J.O. Hirschfelder, R.B. Bird, E.L. Spotz, Chem. Rev. 44 (1949) 205–

al
[12] M.Y. Abyaneh. The electrocrystallization of Niquel. Ph.D. Thesis, Univer- 231.
sity of Southampton, Department of Chemistry, London, 1980. [35] L. Heerman, A. Tarallo, J. Electroanal. Chem. 470 (1999) 70.
[13] J.R. Welty, C.E. Wicks, Robert E. Wilson, Fundamentos de Transferencia [36] A. Milchev, S. Stoyanov, R. Kaischev, Thin Solid Films 22 (1974)
de momento calor y masa, Editorial Limusa. Noriega editores, México,
on 255.
1991, p. 550. [37] A. Milchev, S. Stoyanov, R. Kaischev, Thin Solid Films 22 (1974)
[14] J. Vazquez-Arenas, Study of the involved processes in the electrowinning 267.
of copper from NH4 Cl-NH3 -H2 O solutions. M. Sc. Thesis, Universidad [38] A. Milchev, S. Stoyanov, J. Electroanal. Chem. 72 (1976) 33.
Autónoma de San Luis Potosı́, Instituto de Metalurgia, San Luis Potosı́, [39] A. Milchev, E. Vassileva, J. Electroanal. Chem. 107 (1980) 323.
Mexico, 2006.
rs
[40] A. Milchev, Contemp. Phys. 32 (1991) 321.
[15] J. Vazquez, R. Cruz, ECS Trans. 2 (3) (2006) 355. [41] D. Kaschiev, J. Chem. Phys. 76 (1982) 5098.
[16] A. Ramos, M. Miranda-Hernández, I. González, J. Electrochem. Soc. 148 [42] L.H. Mendoza-Huizar, J. Robles, M. Palomar-Pardavé, J. Electroanal.
(2001) C315. Chem. 521 (2002) 95.
[17] D. Pletcher, R. Greef, R. Peat, L.M. Peter, J. Robinson, Instrumental Meth-
pe

[43] M. Arbib, B. Zhang, V. Lazarov, D. Stoychev, A. Milchev, C. Buess-


ods in Electrochemistry, Horwood Publishing, Chichester, 2001. Herman, J. Electroanal. Chem. 510 (2001) 67.
o r's
th
Au

You might also like