Investigation of cobalt deposition using the electrochemical quartz crystal microbalance

Article abstract of DOI:10.1016/j.electacta.2005.07.003

Electrodeposition of cobalt from sulfate solutions at different pH values was investigated using the EQCM technique coupled with cyclic voltammetry. The results show that cobalt hydroxide is formed simultaneously with cobalt deposition during the early stages of reduction due to the pH variation near the electrode surface caused by the parallel hydrogen evolution reaction (HER). This result was confirmed using M/z values calculated using the Sauerbrey equation and Faraday's law, which showed the presence of cobalt hydroxide in the electrodeposits. A flux model was developed and it assumes a direct reduction of cobalt, simultaneous HER and the formation of cobalt hydroxide during the early stages of deposition at pH 4.10. When the solution pH is decreased to 3.33 only the direct cobalt reduction is observed without any hydroxylated species formation.

Full text of DOI:10.1016/j.electacta.2005.07.003

Electrochimica Acta 51 (2006) 1960–1966
Investigation of cobalt deposition using the electrochemical
quartz crystal microbalance
∗
J.T. Matsushima, F. Trivinho-Strixino, E.C. Pereira
Centro Multidisciplinar para o Desenvolvimento de Materiais Cer aˆ micos, Laborat o´ rio Interdisciplinar de Eletroqu ´ı mica e Cer aˆ mica,
Departamento de Qu ´ı mica, Universidade Federal de S a˜ o Carlos, Caixa Postal 676, S a˜ o Carlos 13565-905, SP, Brazil
Received 8 December 2004; received in revised form 21 March 2005; accepted 2 July 2005
Available online 3 August 2005
Abstract
Electrodeposition of cobalt from sulfate solutions at different pH values was investigated using the EQCM technique coupled with cyclic
voltammetry. The results show that cobalt hydroxide is formed simultaneously with cobalt deposition during the early stages of reduction due
to the pH variation near the electrode surface caused by the parallel hydrogen evolution reaction (HER). This result was confirmed using M/z
values calculated using the Sauerbrey equation and Faraday’s law, which showed the presence of cobalt hydroxide in the electrodeposits. A
flux model was developed and it assumes a direct reduction of cobalt, simultaneous HER and the formation of cobalt hydroxide during the
early stages of deposition at pH 4.10. When the solution pH is decreased to 3.33 only the direct cobalt reduction is observed without any
hydroxylated species formation.
©
2005 Elsevier Ltd. All rights reserved.
Keywords: Electrodeposition; Quartz crystal microbalance; pH effect; Cobalt; Cobalt hydroxide
1
. Introduction
also affects the pH at the electrode surface, modifying the
kinetics of metal reduction [12–14]. The increase of pH at
the electrode surface can result in the precipitation of cobalt
hydroxide, which interferes with the formation of the metallic
deposit and a porous metallic structure is produced [8–10].
Consequently, greater understanding of the HER mechanism
and control of its kinetics would allow the preparation of Co
deposits that are suitable for different applications.
Cobalt is an important metal due to its use in magnetic
alloys, cutting-wear resistant alloys and super alloys [1,2].
Recently, electrodeposited Co and Co alloys have been inves-
tigated with the aim of optimizing experimental conditions to
obtain films with magnetic properties and the usage in mag-
netic storage systems [3–7]. The magnetic properties depend
on the morphology and microstructure of the Co deposits
Several mechanisms for cobalt electrodeposition have
been proposed in the literature [9–11,15,16]. For some
authors [15,16] the cobalt electrodeposition mechanism is
[4,5]. As the morphology and microstructure of the films can
be controlled by the preparation parameters (e.g. solution
composition, applied potential, current density, additives and
pH [8–11]), the electrodeposition process can be used to pre-
pare thin films for magnetic applications.
During the electrodeposition of Co the hydrogen evolution
reaction (HER) occurs as a side reaction that consumes part
of the applied current and reduces the current efficiency and
+
thought to occur via the formation of CoOH and Co(OH)2
species. The formation of the hydroxide species depends
on the pH of the deposition bath. For solution pH lower
2
+
−
than 4.0, the authors [15,16] proposed that Co and OH
+
react producing CoOH “unstable complex”. The reaction
that follows is the reduction of this complex and its reaction
with adsorbed hydrogen to form metallic Co. For solution
pH between 4.0 and 4.5, the authors [15,16] proposed that
Co and OH reacts producing cobalt hydroxide. This com-
pound is reduced to produce metallic Co. According to the
∗
Corresponding author. Tel.: +55 16 3351 8214/213;
2+
−
fax: +55 16 3351 8214/208.
E-mail address: decp@power.ufscar.br (E.C. Pereira).
0
013-4686/$ – see front matter © 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2005.07.003

J.T. Matsushima et al. / Electrochimica Acta 51 (2006) 1960–1966
1961
authors, the last reaction involving the reduction of adsorbed
hydroxide causes the increase of OH concentration near
lytical grade reagents. The potential scan rate was set at
−
−1
10 mV s
.
the electrode which changes the surface pH. According to
−
the authors, since OH is formed during Cobalt reduction, at
pH 4.0 and 4.5, this mechanism does not result in hydrogen
evolution.
3. Results and discussion
On the other hand, Jeffrey et al. [11] proposed an alterna-
tive mechanism involving CoOH ions, whereas the authors
3.1. Theory
+
do not mention if these ions are free or adsorbed on the
electrode. Furthermore, the authors did not consider the dis-
sociation constant of those ions under their experimental
conditions, leaving their proposed mechanism very unpre-
dictable. For example, Jeffrey et al. [11] suggested that the
limiting step for their proposed mechanism is the reduction
of CoOH (an unpredictable species) to metallic Co. Further-
more, the authors did not mention how these particularly
unstable species were formed. It has also be shown [9,10]
different proposed mechanisms for cobalt electrodeposition
consideringdifferentreactionpathsthatinvolvetheformation
of cobalt hydroxides and oxides species, or unstable cobalt
complex ions, as prior products before effective deposition
of metallic cobalt.
The EQCM technique combined with cyclic voltamme-
try is a convenient method to investigate electrochemical
reactions by measuring the simultaneous current, charge and
related mass changes at the working electrode [17]. Accord-
ing to the Sauerbrey equation [18] the frequency variation
(ꢀf) of the quartz crystal is correlated with the mass varia-
tion of the electrode (ꢀm) and can be written as:
2
−
2f ꢀm
0
ꢀf =
√
= −K ꢀm
(1)
A µiρi
where f0 is the resonant frequency of the quartz crystal, A the
piezoelectric active area, µi the shear modulus of the quartz,
K the experimental mass coefficient and ρi is the density of
quartz.
Considering the mechanisms described above, sometimes
showing the presence of unstable species or unpredictable
ions, for the current experimental procedure, we present in
this paper a study aiming the usage of the electrochemical
quartz crystal microbalance (EQCM) technique simultane-
ously with cyclic voltammetry in order to compare past
results and to propose a practical and alternative mecha-
nism for cobalt electrodeposition at different pH unbuffered
solutions. It is proposed that only two different reactions,
i.e. direct cobalt reduction and cobalt hydroxide formation,
describe the current process without further intermediates.
The last reaction is a consequence of the parallel HER. A
flux model considering these proposals was made and fitted
the data.
When Eq. (1) is combined with the Faradays law, the num-
ber of electrons involved in the reaction can be calculated
[18–20].
The species involved in electrochemical reactions on the
substrate can be evaluated by plotting the quartz crystal fre-
quency variation (ꢀf) as a function of the charge consumed
during the reaction (Q) according to the equation:
ꢀ
ꢁ
KM
Fz
ꢀf = −
Q
(2)
where M is the molar mass of the deposit, F the Faraday
constant and z is the number of electrons. Rearranging Eq.
(2), M/z can be determined by fitting the slope of the mass
versus charge curves:
ꢂ
ꢂ ꢀ
ꢁ
M
z
ꢂdꢀf ꢂ
F
ꢂ
ꢂ
ꢂ
ꢂ
2
. Experimental
=
(3)
dQ
K
The electrochemical measurements were carried out in a
The M/z value combines the ꢀf and Q data into a single
glass cell. The working electrode (WE) was a 9 MHz AT-
cut quartz crystal coated with a Pt film in contact with the
electrolyte (A = 0.2 cm ). The sensitivity of the EQCM used
was 770 Hz ␮g . The counter electrode (CE) was a Pt sheet
and all potentials are referred to saturated calomel electrode
factor that can be used to discuss the reaction mechanism in a
simple manner. The corresponding value can be easily calcu-
lated for any electrochemical reaction and it indicates if there
is a single reaction or several parallel reactions occurring at
the electrode surface.
2
−
1
(
SCE). The resonance frequency shift was measured with a
Many authors [21–23] have used the ion flux as a function
of potential calculated from EQCM experiments to investi-
gate reaction mechanisms. Following the same procedure,
the study of the electroreduction/electrodissolution mecha-
nism of Co using flux analysis is proposed here.
Seiko EG&G quartz crystal microbalance (model QCA 917).
The electrochemical measurements were conducted using an
EG&G PAR 263A potentiostat/galvanostat.
◦
Cobalt was deposited at T = 25 C under potentiodynamic
conditions on the Pt coated quartz crystal electrode using
As discussed in the introductory part, Jeffrey et al. [11],
Pradhan et al. [15] and Cui et al. [16] cited different mech-
anisms related to Co electrodeposition in which deposi-
tion proceeds through formation of CoOH or CoOH and
Co(OH)2. At low pH values it was proposed that formation
−1
−1
0
.1 mol L CoSO4·5H2O and 1.53 mol L Na2SO4 as sup-
porting electrolyte at different pH values (4.10 and 3.33).
−
1
+
+
The pH was adjusted by the addition of 0.1 mol L H2SO4.
All solutions were prepared with deionized water and ana-

1
962
J.T. Matsushima et al. / Electrochimica Acta 51 (2006) 1960–1966
of adsorbed hydrogen occurs. At pH values between 4 and
.5, the formation of unstable cobalt species, which may be
the rate-limiting step, is proposed. However, for the purpose
of elaborating a mathematical flux model, it is assumed that
only the direct reduction of cobalt takes place, as shown by
the equation below:
Table 2
M/z found by fitting the slope of the mass versus charge for cobalt dissolution
in the cobalt sulfate solution at pH = 4.10
4
Potential range (V)a
−1
M/z (g mol )
pH
4.10
−0.65 to −0.47
28.5 ± 0.5
27.8 ± 0.5
28.1 ± 0.5
29.0 ± 0.5
30.8 ± 0.5
33.4 ± 0.5
38.4 ± 0.5
42.4 ± 0.5
50.4 ± 0.5
39.7 ± 0.5
−
−
−
−
−
−
−
−
−
0.47 to −0.41
0.41 to −0.36
0.36 to −0.32
0.32 to −0.29
0.29 to −0.27
0.27 to −0.24
0.24 to −0.23
0.23 to −0.18
0.18 to +0.40
Co 2( + 2e → Co(s)
+
−
(4)
aq)
On the other hand, during metal electrodeposition from
aqueous solution at negative potentials, the parallel reaction
of water electrolysis occurs. Such a reaction can be related
to the formation of cobalt hydroxide as shown below:
a
−
−
Potential range used for calculate the M/z values obtained for cobalt
2
H2O + 2e → H2(g) + 2OH
(5)
(6)
(
aq)
dissolution region.
2
+
aq)
(aq)
Table 3
M/z found by fitting the slope of the mass versus charge for cobalt deposition
in the cobalt sulfate solution at pH = 3.33
Summarizing, Eqs. (5) and (6) are related to the formation
of Co(OH)2, while Eq. (4) is related to the direct deposition
of cobalt. Cobalt hydroxide reduction requires a much more
negative potential than direct cobalt reduction [9].
Thus, considering that Eqs. (4)–(6) occur simultaneously,
the mass balance as a function of the applied potential for
cobalt electrodeposition at the working electrode is:
ꢃ
pH
.33
Potential range (V)a
−1
M/z (g mol )
3
(a) −0.89 to −0.90
28.3 ± 0.5
30.6 ± 0.5
29.7 ± 0.5
29.0 ± 0.5
(a) −0.87 to −0.88
(a) −0.89 to −0.90
(b) −0.86 to −0.84
a
Potential range used for calculate the M/z values obtained for cobalt
deposition region: (a) scan toward negative potential and (b) scan toward
positive potential after potential inversion at −0.90 V.
ꢀ
m(E) =
niMi(E) = nCo2+
M
2+
Co (E)
i
+
nCo(OH) MCo(OH) (E)
(7)
counter ions intercalate (or desintercalate) in the polymer
matrix. Besides, for double layer structure investigation [27]
or monolayers adsorption [28], the contribution of H2O
molecules is important. In these case, the adsorbed solvent
molecules must be displaced to accommodate the adsorbed
layer. Then, the calculated M/z change account for both
processes: the mass increase due to the adsorption of the
molecules minus the mass contribution from the displaced
adsorbed solvent molecules. For metals deposition, consid-
ering that the quantity of atoms deposited is much more than
one monolayer, the contribution of the displacement of the
solvent (water) molecules from the interface, when the first
deposition layer occurs, is negligible compared to the overall
mass change.
2
2
The inclusion of water molecules (solvent) to the total
mass balance is not suitable in this case, considering the
experimental conditions described here. None of the cal-
culated data (Tables 1–4) showed M/z values related to
water molecules separately (M/z(H O) = 9, considering 2
2
electrons). Furthermore, the overall Faradaic reaction is due
to reactions described in Eqs. (4)–(6), and therefore the
overall mass variation masks, if it exists, a variation of
mass related only to water molecules. The participation of
solvent molecules is important in conducting polymers or
polyelectrolytes redox processes [24–26] once the solvated
The charge balance as a function of the applied potential in
this process is related to the number of moles of ionic species
Table 1
M/z found by fitting the slope of the mass vs. charge for cobalt deposition
in the cobalt sulfate solution at pH = 4.10
and can be written as:
ꢃ
Potential range (V)a
−1
M/z (g mol )
pH
.10
Q(E) = −
ziFni(E) = −2nCo2+ F(E) + 2nOH
−
F(E)
(8)
4
(a) −0.80 to −0.84
47.6 ± 0.5
47.5 ± 0.5
43.0 ± 0.5
40.7 ± 0.5
38.2 ± 0.5
35.6 ± 0.5
32.5 ± 0.5
29.0 ± 0.5
29.3 ± 0.5
29.7 ± 0.5
i
(a) −0.84 to −0.85
(a) −0.85 to −0.86
(a) −0.86 to −0.87
(a) −0.87 to −0.88
(a) −0.88 to −0.89
(a) −0.89 to −0.90
(b) −0.90 to −0.85
(b) −0.85 to −0.80
(b) −0.80 to −0.65
where F is the Faraday constant. The charge balance is in
accordance with the reduction/dissolution of 1 mol of Co and
Table 4
M/z found by fitting the slope of the mass versus charge for cobalt dissolution
in the cobalt sulfate solution at pH = 3.33
Potential range (V)a
M/z (g mol
29.6 ± 0.5
−1
)
pH
a
3.33
−0.52 to −0.21
Potential range used for calculate the M/z values obtained for cobalt
deposition region: (a) scan toward negative potential and (b) scan toward
a
Potential range used for calculate the M/z values obtained for cobalt
positive potential after potential inversion at −0.90 V.
dissolution region.

J.T. Matsushima et al. / Electrochimica Acta 51 (2006) 1960–1966
1963
with the electrolysis of 1 mol of H2O. It is assumed that only
Faradaic electrochemical processes occur.
Considering the simultaneous reaction of water electrol-
ysis and Co2 reduction, it can be proposed that the amount
+
−
of generated OH is directly proportional to the amount of
Co(OH)2 formed, according to Eqs. (5) and (6). Now, if it is
considered that Co(OH)2 formation is generated from water
electrolysis and proceeds at 100% efficiency in the early
stages of electrodeposition, the charge balance as a function
of the applied potential can be written as:
ꢃ
Q(E) = −
ziFni(E) = −2nCo2+ F(E) + 2nCo(OH) F(E)
2
i
(9)
Rearrangement of Eqs. (7) and (9) has an analytical solu-
tion, since there are two equations and two variables. Con-
sidering this, from Eqs. (7) and (9), the number of nCo2+ and
nCo(OH) species can be calculated as a function of the applied
2
potential (E):
ꢀ
ꢁ
ꢁ
Q(E)MCo(OH)2
n
2+
=
ꢀm(E) −
ꢀ
β
β
(10)
(11)
Co (E)
2
F
Q(E)M 2+
Co
nCo(OH) (E) = ꢀm(E) +
2
2
F
where β is:
1
Fig. 1. (A) Cyclic voltammograms and (B) mass variation as a function of
potential measured in a solution containing 0.1 M CoSO ; 1.53 M Na SO at
pH 4.10 for cobalt electrodeposition. (C) Cyclic voltammograms at different
β = MCo
2+
+ MCo(OH)2
4
2
4
By differentiation of Eqs. (10) and (11) with respect to
time, the species flux as a function of the current density and
mass flux, is obtained:
polarization times for cobalt electrodeposition: 15 min (- - -), 60 min (. . . .).
ꢀ
ꢁ
which the parameters of the nucleation-growth mechanism
aretaken[29]. Thesecondcrossoverwasobservedat−0.66 V
and corresponds to the crossover potential. At potentials more
negativethan−0.85 V, thecurrentforthescantowardpositive
potential is lower than that before the inversion of potential
dnCo2+
d(ꢀm(E)
)
i
M
(E) Co(OH)2
(E)
=
−
β
β
(12)
(13)
dt
dt
2F
ꢀ
ꢁ
dnCo(OH) (E)
d(ꢀm(E)
dt
)
i
M
(E) Co
2+
2
=
+
dt
2F
(
−0.90 V), probably due to the corresponding change in the
2+
As a convention, positive flux values indicate deposition
and negative values indicate dissolution or ejection of the
ionic species.
concentration of Co(H O)
ions at the electrode surface
2
6
[29], due to the deposition process. On the other hand, in the
range between −0.85 and −0.66 V, the current toward pos-
itive potential scan was higher than before the inversion of
potential. According to this, as expected, it can be concluded
that the energy required for cobalt deposition on the cobalt
film, during the forward sweep, is lower than that required
for cobalt deposition on the bare Pt electrode.
At more anodic potentials the cobalt dissolution process
begins at −0.60 V which is characterized by both an anodic
current peak and a mass decrease on the electrode. Two cur-
rent peaks were observed at −0.45 V (process I) and −0.27 V
(process II) in Fig. 1A. These processes can be related to the
HER that occurs simultaneously with cobalt reduction (Eq.
(4)). According to Soto et al. [29], processes I and II could be
related either to the dissolution of cobalt from electrode into
the solution in form of different ionic species or to the disso-
3
.2. Deposition and dissolution of cobalt
Fig. 1 shows the cyclic voltammogram and the mass vari-
ation as a function of potential for cobalt electrodeposition
at pH 4.10. The cobalt reduction begins at −0.8 V and it is
accompanied by a mass increase that indicates Co deposi-
tion on the electrode. After reversing the potential at −0.9 V,
the cobalt reduction proceeds up to −0.66 V. Between this
potential and −0.60 V the mass of the electrode remains
constant where no Faradaic reaction occurs. Two crossovers
between cathodic current branches are observed during the
scan toward positive potential. The more cathodic crossover
(
−0.85 V) corresponds to the nucleation overpotential from

1
964
J.T. Matsushima et al. / Electrochimica Acta 51 (2006) 1960–1966
lution of different cobalt phases that were previously formed
during the cathodic sweep.
In order to understand the results shown in Fig. 1 and to
evaluate the mechanism of cobalt deposition in these exper-
iments an analysis of the pH of the deposition bath and
working electrode polarization time was carried out.
Different polarization times at −0.64 V (Fig. 1C) were
investigated to characterize the dissolution peaks observed in
Fig. 1A. As can be seen in the inset, the potential peak related
to process I vanishes as the polarization time increases, while
displacement toward positive potentials of the potential peak
related to process II is observed, which can now be associated
with the dissolution of cobalt without adsorbed hydrogen.
Since the greater part of the chemisorbed hydrogen is formed
during the HER side reactions, this hydrogen rich phase is
formed during cobalt reduction, as [12]:
+
−
H3O + e + Co → CoH
+ H2O
(14)
(15)
(16)
(ads)
+
−
H3O + CoH(
+ e → H
+ H2O
ads)
2(ads)
H2(ads) → H2(g)
+
Eq. (14) refers to the discharge of H3O ions initially
located in the Outer Helmhotz Plane, while Eqs. (15) and
(16) are related to the electrochemical dessorption and dif-
fusion of adsorbed hydrogen molecules from the electrode
surface. Although we did not observe bubbles formation over
the electrode and considering that process I disappear when
the electrode is polarized, we believe that this process can
be associated with the dissolution of hydrogen rich cobalt
phase. For Ni deposits, it is proposed [30–32] the formation
of a nickel hydride phase. In the case of Co, to our knowledge,
there is not any experimental evidence of the formation of a
CoH phase. Our results indicate that at least a hydrogen rich
cobalt phase is formed under the investigated conditions.
Using the M/z values determined by the slope of the fitting
mass data as a function of charge, all species involved during
Co electrodeposition process, studied at different values of
solution pH, can be visualized. Table 1 shows the M/z val-
ues calculated from experiments carried out at pH 4.10 for
deposition and dissolution of Cobalt. In the initial stages of
Fig. 2. (A) Flux of Co and Co(OH)2 as a function of potential and (B) cyclic
voltammogram measured in a solution containing 0.1 M CoSO4; 1.53 M
Na SO at pH 4.10 for cobalt electrodeposition.
2
4
Consequently, the remaining current and mass variation after
this potential can be related to the dissolution of cobalt and
cobalt hydroxide that had been previously deposited simul-
taneously on the electrode surface.
The flux of Co and Co(OH)2 during the cobalt electrode-
position at pH 4.10 can be calculated from the EQCM and
voltammetry data, according to Eqs. (12) and (13) (Fig. 2).
The Co reduction begins at a −0.8 V, the same potential range
where Co(OH)2 was detected (Table 1), indicating a simul-
taneous deposition of these species. Nakahara and Mahajan
[8] reported the same behavior for cobalt electrodeposition
on a sheet of annealed copper. At −0.86 V, Co(OH)2 reaches
the maximum flux value and coincides with the same poten-
tial where the experimental M/z value is compatible with
−
1
cobalt reduction, the calculated M/z value is 46.6 g mol
,
which is compatible with Co(OH)2 species and in accor-
dance with Eqs. (5) and (6) where two electrons are involved
−
(
MWCo(OH) /2e ). As the potential is displaced toward more
2
−
1
negative values, M/z decreases close to the theoretical value
Co(OH)2 species (46.6 g mol ) (Table 1). At this point it
is assumed that the formation of Cobalt hydroxide in the
vicinity of the electrode ceases or its rate decreases and the
−
1
predicted for direct cobalt reduction (MW = 29.5 g mol ) at
0.86 V, just after the inversion of potential. In this potential
−
2+
2+
range it is suggested that direct Co reduction occurs (Eq.
remaining current is used in a direct deposition of Co . After
(4)), as mentioned earlier.
this potential, the flux of cobalt hydroxide begins to decrease
For Co dissolution, M/z values correspond only to pure
until reaching zero (Fig. 2), where at −0.9 V only Co species
−
1
Cobalt dissolution (Table 2). Co(OH)2 species were also
detected for potentials more positive than 0.25 V during a
short range of potentials (Table 2). We can compare the same
potential range in Tables 1 and 2 with the data presented in
Fig. 1, where the current density decreases after process II.
were detected (29 g mol ) as can be seen in Table 1. The
dissolution of the cobalt hydroxide is observed at the end of
2
+
process II, in conjunction with the dissolution of Co . Sim-
ilar results were observed by Soto et al. [29] using the same
voltammetric technique and the deposition and dissolution of

J.T. Matsushima et al. / Electrochimica Acta 51 (2006) 1960–1966
1965
not change during the experiments [27]. We believe, that this
parameter, surface roughness, indeed does not change due
to: (a) in the present case, the cobalt deposit thickness is only
−3
6
7 nm (deposited mass = 12 ␮g, Cobalt density = 8.9 g cm
,
2
electrode area = 0.2 cm ). Therefore, it is not expected that
a change in the roughness factor occurs during the experi-
ment. (b) Besides, a signature of a change in the electrode
area is a loop during the reduction process, i.e. at the same
potential, the current observed during the sweep towards the
negative direction is lower than those observed during the
sweep towards positive potentials. In Figs. 1 and 3, it is clear
that the loop is absent for the experiment done at both pH
investigated. Then, we conclude that the roughness factor is
constant during our experiments.
4. Conclusions
The analysis of the EQCM data during cobalt electrode-
position points to the formation of cobalt hydroxide during
deposition. The absence of cobalt hydroxide in the exper-
iments carried out at pH 3.33 was also confirmed. This
result was confirmed by both the M/z values and flux model
analyses. The deposition of cobalt was accompanied by the
HER and the results indicates that at least an hydrogen rich
cobalt phase is formed under the investigated conditions, as
confirmed by polarization experiments. Cobalt hydroxide is
formed due to variation of solution pH (>4.10) and the forma-
tion of hydroxide ions at the electrode surface caused by the
HER. The EQCM technique combined with cyclic voltam-
metry was found to be a useful in situ tool for investigating the
effects of pH and polarization time on the electrode surface
during metal deposition.
Fig. 3. (A) Cyclic voltammograms and (B) mass variations as a function of
potential measured in a solution containing 0.1 M CoSO4; 1.53 M Na2SO4
at pH 4.10 (—) and 3.33 (. . . .) adjusted with H2SO4. (C) Inset: flux of Co
and Co(OH)2 as a function of potential measured in a solution containing
0
.1 M CoSO4; 1.53 M Na2SO4 at pH 3.33.
Co and Co(OH)2 occur simultaneously [8] at the beginning
and end of the process, respectively.
Acknowledgement
For experiments carried out at pH 3.33 (Fig. 3), a decrease
in the mass change is observed probably due to the increas-
ing effect of the HER, which decreases the Co deposition
efficiency[8]. Afterthis, themassvariationattheendofdepo-
sition returns to the same value as at the beginning, while for
a solution of pH 4.10 the mass variation was not reversible.
However, cobalt hydroxide was not detected as illustrated by
the data in Tables 3 and 4, where no M/z values for Co(OH)2
are observed. The absence of hydroxide species can be related
The authors are grateful to FAPESP and CNPq (Brazil)
for the financial support.
References
[
1] F.R. Morral, in: M.G. Howe (Ed.), Othmer-Kirk-Encyclopedia of
Chemical Technology, vol. 5, John Wiley & Sons, New York, 1964,
p. 716.
−
2+
to the decreasing amount of free OH ions reacting with Co
in the vicinity of the electrode surface, leading to direct Co
deposition (Eq. (4)).
[
[
2] D. Landolt, Eletrochim. Acta 39 (8–9) (1994) 1075.
3] B. Brozzini, D. De Vita, A. Sportoreletti, G. Zangari, P.L. Cavalloti,
J. Magn. Magn. Mater. 120 (1993) 300.
The flux curve, in this case, indicates only the existence of
Co2+ instead of Co and Co(OH)2 during the electrodeposi-
tion process (Fig. 3C). The deviation for small flux values is
inside the error expected considering the procedure of calcu-
lation and the sensibility of the quartz crystal microbalance
during electrodeposition at pH 3.33 (inset in Fig. 3C).
Finally, it is important to stress that the interpretation
given above in correct only if the surface roughness does
[
[
4] T. Osaka, Electrochim. Acta 42 (1997) 3015.
5] S. Armyanov, Electrochim. Acta 45 (2000) 3323.
2+
[6] P.L. Cavalloti, N. Lecis, H. Fauster, A. Ziclonka, J.P. Celis, G. Wan-
tas, J. Machado da Silva, J.M. Brochado Oliveira, M.A. S a´ , Surf.
Coat. Technol. 105 (1998) 232.
[7] J.L. Su, M.M. Chen, J. Lo, R.E. Lee, J. Appl. Phys. 63 (1988) 4022.
[8] S. Nakahara, S. Mahajan, J. Electrochem. Soc. 127 (1980) 283.
[9] S.P. Jiang, Y.Z. Chen, J.K. You, T.X. Chen, A.C.C. Tseung, J. Elec-
trochem. Soc. 137 (1990) 3374.

1
966
J.T. Matsushima et al. / Electrochimica Acta 51 (2006) 1960–1966
[
10] S.P. Jiang, A.C.C. Tseung, J. Electrochem. Soc. 137 (11) (1990)
[20] A. Marlot, J. Vedel, J. Electrochem. Soc. 146 (1999) 177.
[21] A.R. Hillman, Solid State Ionics 94 (1997) 151.
[22] S.L.A. de Maranh a˜ o, R.M. Torresi, Electrochim. Acta 44 (1999)
1879.
[23] H. Varela, M.R. Torresi, J. Electrochem. Soc. 147 (2000) 665.
[24] H. Varela, M. Malta, R.M. Torresi, J. Power Source 92 (2001) 50.
[25] H. Varela, S.L.A. de Maranh a˜ o, R.M.Q. Mello, E.A. Ticianelli, R.M.
Torresi, Synth. Met. 122 (2001) 321.
3387.
[
[
[
11] M.I. Jeffrey, W.L. Choo, P.L. Breuer, Miner. Eng. 13 (2000) 1231.
12] D.R. Gabe, J. Appl. Electrochem. 27 (1997) 908.
13] Ph. Vermeiren, R. Leysen, H. Vandendorre, Electrochim. Acta 30
(1985) 1253.
[14] A. Zhou, N. Xie, J. Colloid Interface Sci. 220 (1999) 281.
[15] N. Pradhan, T. Subbaiah, S.C. Das, J. Appl. Electrochem. 27 (1997)
7
13.
16] C.Q. Cui, S.P. Jiang, A.C.C. Tseung, J. Electrochem. Soc. 137 (1990)
418.
[26] A. Jackson, A.R. Hillman, S. Bruckenstein, I. Jureviciute, J. Elec-
troanal. Chem. 524–525 (2002) 90.
[27] V. Tsionsky, G. Katz, E. Gileadi, L. Daikhin, J. Electroanal. Chem.
524–525 (2002) 110.
[28] Z. Jusys, S. Bruckenstein, Electrochem. Commun. 2 (2000) 412.
[29] A.B. Soto, E.M. Arce, M. Palomar-Pardave, I. Gonz a´ lez, Elec-
trochim. Acta 41 (1996) 2647.
[30] M. Fleishmann, A. Saraby, Electrochim. Acta 29 (1984) 69.
[31] Y.P. Lin, J.R. Selman, J. Electrochem. Soc. 140 (1993) 1299.
[32] H. Hayashi, S. Matsuda, Electrochemistry 70 (2002) 768.
[
[
3
17] D.A. Buttry, in: H.D.A. Armario (Ed.), Electrochemical Interfaces:
Modern Techniques for in situ Interface Characterization, VHC, New
York, 1991 (Chapter 10).
18] H. Saloniemi, M. Kemel, M. Ritala, M. Leskel a¨ , J. Electroanal.
Chem. 482 (2000) 139.
[
[
19] J.G.N. Matias, J.F. Juli a˜ o, D.M. Soares, A. Goresnstein, J. Elec-
troanal. Chem. 431 (1997) 163.