Electrodeposition of iron in sulphate solutions

Article abstract of DOI:10.1016/j.electacta.2008.01.015

The kinetics of iron electrodeposition from acid sulphate solutions onto a platinum electrode was investigated by means of stationary polarisation curves and electrochemical impedance spectroscopy. Together with interfacial pH data previously obtained, the effect of pH was analysed. The formation of at least three adsorbed intermediates at the cathode surface was evidenced in all pH values. The relative rate of their formation and its surface concentration depend on the solution pH as well as on the electrode potential. It is suggested that two of these species catalyses the H⁺ reduction whereas the other one may have a blocking effect on this reaction.

Full text of DOI:10.1016/j.electacta.2008.01.015

Available online at www.sciencedirect.com
Electrochimica Acta 53 (2008) 7426–7435
Electrodeposition of iron in sulphate solutions
a,1
S.L. Dıaz , J.A. Calderonc,1, O.E. Barciaa,b,1, O.R. Mattosa,∗,1
´
´
a LNDC—Laborato´rio de Ensaios Na˜o-Destrutivos e Corrosa˜o Prof. Manoel de Castro, EE/PEMM/COPPE,
Universidade Federal do Rio de Janeiro, UFRJ, Cx. Postal 68505, CEP 21945-970, Rio de Janeiro, RJ, Brazil
b Dep. F´ısico-Qu´ımica, IQ/UFRJ, Rio de Janeiro, RJ, Brazil
c Grupo de Corrosio´n y Proteccio´n, Universidad de Antioquia, Medell´ın, Colombia
Received 18 July 2007; received in revised form 8 January 2008; accepted 8 January 2008
Available online 17 January 2008
Abstract
The kinetics of iron electrodeposition from acid sulphate solutions onto a platinum electrode was investigated by means of stationary polarisation
curves and electrochemical impedance spectroscopy. Together with interfacial pH data previously obtained, the effect of pH was analysed. The
formation of at least three adsorbed intermediates at the cathode surface was evidenced in all pH values. The relative rate of their formation and
its surface concentration depend on the solution pH as well as on the electrode potential. It is suggested that two of these species catalyses the H+
reduction whereas the other one may have a blocking effect on this reaction.
© 2008 Elsevier Ltd. All rights reserved.
Keywords: Fe electrodeposition; Adsorbed intermediates; Impedance; pH effects; Mechanism
1. Introduction
participation of Fe hydroxide species during iron electrodisso-
lution/electrodeposition. Some other more recent works [11,12]
have also considered the contribution of this species that result
from the increase on the pH of the solution adjacent to the
cathode. The presence of this species at the electrode/solution
interface might generate adsorption/desorption processes. By
investigating Fe nucleation kinetics on a vitreous carbon elec-
trode through electrochemical techniques and AFM microscopy,
Grujicic and Pesic [13] proposed a theoretical model for Fe elec-
trodeposition based on a progressive nucleation mechanism that
was associated with the degree of inhibition of the nucleation
sites.
Evidences of surface alkalination during Fe-group metal and
their alloy electrodeposition have been found by means of pH
measurements at the vicinity of the electrode using different
techniques [14,15]. In a previous work [16], interfacial pH
measurements were performed during Zn–Fe electrodeposition
using an in-situ non-intrusive setup. It was verified that the
interfacial pH increases during individual Fe electrodeposition,
meaning that it occurs with simultaneous H+ consumption.
Electrochemical impedance technique provides information
on electrodeposition processes by means of time constants asso-
ciated with the elementary steps of these systems. Epelboin
et al. [17] indicated that the charge transfer in multiple steps
might take place at the electrode surface through adsorbed
Investigations on iron electrodeposition have been basi-
cally restricted to the effects of deposition parameters on
the morphological, structural and magnetic properties of the
deposits [1–3]. Moreover, the fundamentals of the mechanism
that governs iron electrodeposition have received less atten-
tion if compared to the highly investigated Fe alloy systems.
Indeed, the nature of the anomalous electrodeposition of Fe
alloys, such as Zn–Fe, Co–Fe and Ni–Fe, has been the sub-
ject of many researches [4–6] but the mechanisms still require
improvements. In this particular case, to better understand
the codeposition processes, the knowledge on the individual
deposition mechanisms must be established. In addition, the
increasing interest on the development of Fe thin deposits for
magnetic applications [3,7] also calls for the comprehension of
the elementary steps that take place during the electrodeposi-
tion.
¨
Kabanov et al. [8], Bonhoeffer and Jena [9] and later Bock-
ris et al. [10] were the earliest to mention the formation and
∗
Corresponding author. Tel.: +55 21 2562 8550; fax: +55 21 2290 1544.
E-mail addresses: susana@metalmat.ufrj.br (S.L. D´ıaz),
omattos@metalmat.ufrj.br (O.R. Mattos).
1
ISE member.
0013-4686/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2008.01.015

S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
7427
intermediates, which can interact with other inhibiting species.
The coupling between these charge transfer reactions and
the adsorbates generates relaxation processes that strongly
depend on the electrolyte composition, pH, nature of the anion,
polarisation and hydrodynamic conditions [18]. Nickel elec-
trodeposition is an example of a process where strong surface
interactions between Ni2+ and H+ take place. These interactions,
together with the anion influence, were considered to account
for the electrochemical impedance results [19,20]. The interpre-
tation of the experimental results was done in terms of adsorbed
intermediates as well as of the potential dependence of growth
sites.
The objective of the present work was to study the kinet-
ics of Fe electrodeposition as a function of solution pH by
means of potentiostatic polarisation curves and electrochem-
ical impedance spectroscopy in sulphate medium. Taken into
account the results on interfacial pH presented elsewhere [16], it
is suggested the formation of three adsorbed intermediates at the
cathode surface. The presence of these intermediates validates
the participation of H+ and/or OH− on iron electrodeposition
process. The rate of their formation and their relative predomi-
nance depends on both potential and solution pH. The features
discussed in the present work are essential to the establishment
of a model for ZnFe alloy electrodeposition [16].
Fig. 1. Cathodic polarisation curves obtained in the 0.9 M FeSO4·7H2O solution
as a function of pH.
ohmic drop through the high frequency limit of electrochemical
impedance plots.
3. Results
2. Experimental
Fig. 1 presents the cathodic potentiostatic polarisation curves
obtained in the 0.9 M Fe sulphate solution in different pH values.
These curves can be divided into two distinct regions indepen-
dently of the pH. In the first one, at low polarisations, the current
varies slowly with cathodic polarisation. In this region, as pre-
viously discussed [16], the efficiency for metallic deposition is
extremely low and the H+ reduction reaction prevails. With fur-
ther polarisation increase, the current densities show a sudden
increase that confers an abrupt change in the slope of the curves,
which start to vary in a quite linear way with cathodic poten-
tial. High efficiency in metallic deposition was only detected at
this part of the curves, for all the solution pH investigated. Only
for reference, and following a general tendency in the literature,
hereafter in this text, the two parts of the curves will be called as
“non-linear”, wherein H+ reduction prevails, and a “linear” one,
mainly related to Fe electrodeposition. This was done to make
easier the discussion and to differentiate the two regions of the
curves where distinct cathodic process predominates.
With increasing pH, a decrease in current densities in the
non-linear region is observed in Fig. 1. For each pH value, a
current maximum is also detected in this potential range, which
depends on solution pH. After this maximum, all curves present
a slow decrease in current values and, subsequently, increase
abruptly with polarisation. As can be seen from Table 1, the
current maximum and the beginning of the linear part of the
curves are shifted to less cathodic potentials with increasing
pH.
The electrolyte used in this work consisted of 0.9 M
FeSO4·7H2O solution at different pH values. A sodium sulphate
solution in the same concentration was also employed as a blank
solution. The solutions were prepared from analytic grade chem-
icals and double distilled water, without addition of any other
substance such as additives, complex agents or supporting elec-
trolyte. The adjustment of the pH of solutions was done by 0.9 M
sulphuric acid additions.
Cathodic potentiostatic (steady state) polarisation curves and
electrochemical impedance measurements, in both potentio-
static or galvanostatic modes, were performed in a conventional
three-electrode cell, using a Pt rotating disc electrode (RDE)
with 0.2 cm2 as working electrode. All experiments were car-
ried out at 500 rpm of electrode rotation speed. The counter
electrode was a circular Pt grid (300 cm2) that was separated
from the working electrode by a sintered glass filter. A satu-
rated Hg/Hg2SO4 electrode (SSE) was used as reference. The
potential values given in this work were always versus SSE.
All experiments were performed at room temperature and in the
absence of O2 with iron powder at the bottom of the cell to
prevent Fe(II) oxidation.
An OMNIMETRA mod. PG19 potentiostat and a frequency
response analyser SOLARTRON SI1254, coupled with a PC
were used in the impedance measurements. The signals were
filtered by means of a KEMO mod. VBF-8 Filter in low-pass
mode. The rotating electrode was an Asservissement Eletron-
ique device. In most of the experiments the impedance diagrams,
in potentiostatic or galvanostatic modes, were obtained in
the frequency range of 40 kHz–1.3 mHz with a signal ampli-
tude of 7–15 mV. The polarisation curves were corrected from
Fig. 2 shows the cathodic polarisation curves obtained in
both 0.9 M Na2SO4 and 0.9 M Fe sulphate solutions at pH 3.0.
As can be seen in this figure, before −1200 mV, where the
curve for Fe(II) solution presents a maximum, the current den-
sities associated with this solution are higher than those for the

7428
S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
Table 1
Potential values at the current maximum and at the beginning of the linear part
of the polarisation curves in Fig. 1
Solution pH
Potential at current
Potential at the current
maximum (mVSSE
)
minimum (mVSSE)
2.5
3.0
3.5
4.0
5.0
−1223
−1201
−1174
−1165
−1160
−1322
−1309
−1295
−1255
−1230
Na2SO4 solution. After this potential, the current values for the
Na2SO4 solution become higher than those for the Fe(II) solu-
tion. Although not shown in this paper, a similar behaviour was
observed for all the pH values investigated. As already described
[16], at the potential range corresponding to the non-linear part
ofthecurvesin Fig. 1, themainelectrochemicalprocessistheH+
reduction, even though Fe electrodeposition can occur with very
low efficiency. Accordingly, the higher current values observed
until −1200 mV in the presence of Fe(II) can be considered as
a result of a catalytic effect on the H+ reduction reaction.
Fig. 3 displays the impedance diagrams obtained in the
Na2SO4 solution at pH 3.0 and in Fe(II) sulphate solution at pH
2.5, 3.0, 3.5, 4.0 and 5.0, all at potential values corresponding to
the region before the current maxima at the non-linear part of the
polarisation curves in Fig. 1. For each diagram, the capacitive
loop at high frequencies is related to the charge transfer resis-
tance in parallel with the double layer capacitance, whose values
Fig. 2. Typicalcathodicpolarisationcurvesobtainedin0.9 MNa2SO4 and0.9 M
FeSO4, at pH 3.0.
lie between 3 and 80 ␮F cm−2. From now on, only the features in
the low frequency domain related to the faradaic response will be
discussed. The diagram obtained in the Na2SO4 solution shows
two capacitive loops in the low frequency range that are associ-
ated with the H+ reduction as the single cathodic reaction. The
characteristic frequencies (the frequency at the maximum) of
these loops are 0.7 and 0.03 Hz. Analogous results were obtained
for Fe(II) solution in all pH values: two capacitive loops were
also detected in the low frequency domain, independently of the
Fig. 3. Impedance diagrams obtained in 0.9 M Na2SO4 (pH 3.0) and in 0.9 M Fe(II) solution at different pH and at the potential region before the current maximum
of polarisation curves in Fig. 1 (frequencies in Hz).

S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
7429
Fig. 4. Impedance diagrams obtained in 0.9 M Na2SO4 (pH 3.0) and in 0.9 M Fe(II) solution at different pH and at the potential region just after the corresponding
current maxima in the polarisation curves in Fig. 1 (frequencies in Hz).
pH. The similarity between the diagram obtained in the Fe(II)-
free solution and those for the Fe(II) solutions confirms that
the H+ reduction is the major cathodic reaction at this potential
region. This fact is in agreement with previous data related to
the interfacial pH [16] and confirms that the results in Fig. 2
correspond mainly to a catalytic effect due to the presence of
Fe(II) ions in the solution.
sation curves (Fig. 1) in the Fe(II)-containing solution at pH 2.5,
3.0, 3.5, 4.0 and 5.0, respectively. Independently of the pH value,
the diagrams show the same qualitative behaviour. At the very
beginning of the linear part of the polarisation curves (diagrams
(a) in Figs. 5–9) two inductive loops can be observed at the low
frequency domain. With further polarisation increase a new loop
appears between those two inductive loops in the diagrams for
all the pH values studied. This new loop is a capacitive feature
that increases in size with increasing polarisation and solution
pH.
In fact, from Figs. 5–9, it is noteworthy that this general
behaviour becomes more evident as the solution pH increases:
the polarisation level in which the capacitive loop at medium fre-
quencies appears decreases with increasing pH. For instance, at
pH 2.5 the capacitive loop emerges at a current density around
25 mA cm−2 (Fig. 5(e)), while at pH 5.0 this loop is already
present in the diagram obtained in 5.6 mA cm−2 (Fig. 9(c)).
Additionally, at pH 2.5, the dimension of this loop remains very
small even at high polarisation (Fig. 5(f)) if compared to the one
seen at pH 5.0 at lower current density (Fig. 9(c)).
In Fig. 4, the diagrams show the typical behaviour at the
potential region just after the current maximum at the non-linear
part of the polarisation curves (Fig. 1) obtained in Fe(II)-
containing solution. A diagram obtained in the Na2SO4 solution,
at a potential value corresponding to this region, is also dis-
played. As can be seen, this diagram is basically the same as the
one presented in Fig. 3 for the same solution, which is related to
the H+ reduction process. Conversely, for the Fe(II)-containing
solution at different pH values, the diagrams changed, if com-
pared to those in Fig. 3. At low pH values (2.5 and 3.0) the
diagrams show a negative polarisation resistance (Rp). In con-
trast, at higher pH values (3.5, 4.0 and 5.0), the negative Rp is
no longer verified but the capacitive loop at very low frequen-
cies increases significantly. These results are in conformity with
the decrease in current values observed in non-linear part of the
curves in Fig. 1. Since in this region the metallic electrodepo-
sition efficiency is still extremely low, a blocking effect on the
H+ reduction reaction might be assumed in the Fe(II)-containing
solutions [16].
Figs. 5–9 also show that the inductive loop at higher fre-
quencies (≥0.6 Hz) decreases in size and presents a great shift
of the characteristic frequency value with increasing polarisa-
tion, especially at low pH. The lower the solution pH, the more
remarkable is this trend. Correspondingly, for a given pH, the
inductive loop at lower frequencies (≤0.06 Hz) increases with
the polarisation and its characteristic frequency does not show
a considerable change. With pH increase, however, a decreas-
Figs. 5–9 present the impedance diagrams obtained in the
potential range corresponding to the linear region of the polari-

7430
S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
Fig. 5. Impedance diagrams obtained in the Fe(II) solution, pH 2.5, at the linear part of the polarisation curves in Fig. 1 (frequencies in Hz).
ing on the characteristic frequency of this inductive feature is
observed (from 0.06 Hz, at pH 2.5 to <0.01 Hz, at pH 5.0).
Consequently, it becomes hard to observe this inductive loop at
high polarisations. For this reason, and to help the visualization,
some of the diagrams have this loop extrapolated until the Rp
value associated with the slope of the corresponding polarisation
curves in Fig. 1.
with bi < 0 for the cathodic reactions (I < 0) and bi > 0 for anodic
reactions (I > 0), then the main global reactions for Fe electrode-
position mechanism can be written as:
K
1
Fe(II) + e−−→ Fe(I)ads
(θ1 ≡ Fe(I)ads
)
(I)
K
2
Fe(I)ads + e−−→ Fe
(II)
K
3
Fe(I)ads + H2O ꢀ Fe(I)∗ads + H+
(θ2 ≡ Fe(I)∗ads
)
)
(III)
4. Discussion
K
−3
K
4
From the results above presented, it is clear that the mech-
anism of iron electrodeposition from acid sulphate solutions
must include at least three adsorbed species. Thus, to account
for these experimental data, the following general kinetic model
can be considered:
Fe(I)ads + H2O ꢀ Fe(I)∗ad∗s + H+
(θ3 ≡ Fe(I)∗ad∗s
(IV)
(V)
K
−4
K5
Fe(I)∗ads + Fe(II) + 2e−−→ Fe(I)∗ads + Fe
K5∗
Fe(I)∗ads + Hads + e−−→ Fe + Hincluded
(Vꢀ)
(VI)
K6
∗∗
∗∗
Fe(I) + Fe(II) + 2e−−→ Fe(I) + Fe
ads
ads
where θi corresponds to the electrode coverage for the corre-
sponding adsorbates quoted in the above reactions.
To account for the stoichiometry of K3, K and K4, K
−3
−4
Taking into account that the corresponding rate constants, Ki,
of the reactions vary with the cathodic potential, E, in the form:
reversible reactions, the structure of both the intermediates,
∗∗
Fe(I)∗ads andFe(I) , mustcontainthe(OH)− anionintheirstruc-
ads
tures. As already proposed for Fe electrodissolution [21,22], the
adsorbed species associated with θ1–θ3 can be differentiated by
the presence of anions in their respective structure. Moreover,
Ki = K0i exp biE

S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
7431
Fig. 6. Impedance diagrams obtained in the Fe(II) solution, pH 3.0, at the linear part of the polarisation curves in Fig. 1. Rp represents the slope of the corresponding
polarisation curves at the polarisation conditions where each diagram was obtained (frequencies in Hz).
the rate of reaction (Vꢀ) is considered to be very small. Simi-
lar to what was proposed for nickel electrodeposition [19], this
postulationaccountsforthehydrogeninclusionintothedeposits.
The simultaneous hydrogen reduction may occur by means
of the following reactions:
reduction. Therefore, Eq. (1) can be re-written as:
I
−
= K1(1 − θ1 − θ2 − θ3 − θ4) + K2θ1
F
IH
+ 2K5θ2 + 2K6θ3 −
F
K
7
Fe(I)ads + H+ + e−−→ Fe(I)ads + Hads
(θ4 ≡ Hads) (VII)
K
8
From Eq. (1), the admittance can be written as:
∗∗
∗∗
Fe(I) + H+ + e−−→ Fe(I) + Hads
(VIII)
(IX)
(X)
ads
ads
K
9
H+ + e−−→ Hads
˜
˜
I
1
θ1
=
+ F[K1 − K2 − K7[H+] + K9[H+]]
˜
E
˜
E
Rt
K
10
Hads + Hads−→ H2
˜
θ2
+F[K1 − 2K5 + K9[H+] − K5∗θ4]
˜
E
Considering that the Langmuir isotherm law is valid, the cur-
rent associated with the above set of reactions can be described
by the following equation:
˜
θ3
+F[K1 − 2K6 − K8[H+] + K9[H+]]
(2)
˜
E
˜
θ4
+F[K1 + K9[H+] − K5∗θ2]
−
I
˜
E
−
= K1(1 − θ1 − θ2 − θ3 − θ4) + K2θ1 + 2K5θ2 + 2K6θ3
+K7[H+]θ1 + K8[H+]θ3
+
F
˜
[H ]
F[K9(1 − θ1 − θ2 − θ3 − θ4) + K7θ1 + K8θ3]
˜
E
+K9[H+](1 − θ1 − θ2 − θ3 − θ4) + K5∗θ2θ4
(1)
where Rt is the charge transfer resistance and F is the Faraday’s
constant.
I
F
= K7[H+]θ1 + K8[H+]θ3 + K9[H+](1 − θ1 −
H
where
−
θ2 − θ3 − θ4) + K5∗θ2θ4 is the current related to the hydrogen
The mass balance leads to the following equations:

7432
S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
Fig. 7. Impedance diagrams obtained in the Fe(II) solution, pH 3.5, at the linear part of the polarisation curves in Fig. 1. Rp represents the slope of the corresponding
polarisation curves at the polarisation conditions where each diagram was obtained (frequencies in Hz).
dθ1
dt
ble if compared to IH. Therefore, the principal reaction at this
potential range should be the H+ reduction. This is in agreement
with the previous results [16] and coherent with what is shown
in Fig. 3.
β
= K1(1 − θ1 − θ2 − θ3 − θ4) − K2θ1 − K3[H2O]θ1
+K−3[H+]θ2 − K4[H2O]θ1 + K−4[H+]θ3
(3)
In this region, before the current maximum in the
curves for Fe(II)-containing solution, if the hydrogen
reduction reaction is considered to occur mostly through
K9[H+](1 − θ1 − θ2 − θ3 − θ4), then the current values in the
polarisation curves for both, Fe(II)-free and Fe(II)-containing
solutions, Fig. 2, would be the same. However, as seen in this
figure, the current densities associated with the curve for the
Fe(II) solution are higher than those for the Na2SO4 solution.
Consequently, one realize that the θ1 might play a catalytic role
on the H+ reduction as a supplementary source for Hads pro-
duction (reaction (VII)), through the K7[H+]θ1 contribution in
IH.
dθ2
dt
β
β
= K3[H2O]θ1 − K−3[H+]θ2 − K5∗θ2θ4
= K4[H2O]θ1 − K−4[H+]θ3
(4)
(5)
dθ3
dt
dθ4
dt
β
= K7[H+]θ1 + K8[H+]θ3 + K9[H+]
× (1 − θ1 − θ2 − θ3 − θ4) − K10θ42 − K5∗θ2θ4
(6)
With increasing polarisation, after the potential correspond-
ing to the current maximum in the curve for Fe(II)-containing
solution, the current values for the Na2SO4 solution become
higher than those for the Fe(II) solution (Fig. 2). Moreover, in
the same potential range, a decrease in the current densities for
the Fe(II) solution and a negative polarisation resistance in the
corresponding impedance diagrams (Fig. 4) are observed. This
drop in the current density can be explained by means of the
launching of θ2 participation. From the above reaction model, θ2
is produced by the consumption of θ1, previously formed (reac-
At the beginning of the non-linear part of the polar-
isation curves in Fig. 1, before the current maxima, the
major contribution to the current is considered to be given
by 2K2θ1 + IH, since only reactions (I), (II), (VII) and (IX)
have a significant input on the charge transfer in this
potential range. In this condition, IH consists essentially of
K7[H+]θ1 + K9[H+](1 − θ1 − θ2 − θ3 − θ4). In contrast, Fe elec-
trodeposition would be likely to occur through θ1, via reaction
(II). However, since the efficiency for metallic deposition is, in
this potential range, insignificant, then K2θ1 becomes negligi-

S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
7433
Fig. 8. Impedance diagrams obtained in the Fe(II) solution, pH 4.0, at the linear part of the polarisation curves in Fig. 1. Rp represents the slope of the corresponding
polarisation curves at the polarisation conditions where each diagram was obtained (frequencies in Hz).
Fig. 9. Impedance diagrams obtained in the Fe(II) solution, pH 5.0, at the linear part of the polarisation curves in Fig. 1. Rp represents the slope of the corresponding
polarisation curves at the polarisation conditions where each diagram was obtained (frequencies in Hz).

7434
S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
tion (III) and Eq. (4)). Therefore, as the polarisation increases,
θ2 increases, replacing θ1 at the electrode surface and, as a con-
sequence, K7[H+]θ1 contribution decreases. Assuming that θ2
doesnotcatalysetheH+ reduction, thisprocessbecomestooccur
simply via K9[H+](1 − θ1 − θ2 − θ3 − θ4), whose rate should be
lower than that of K7[H+]θ1. In addition, the available surface
area for the reduction of H+ through K7[H+]θ1 also starts to
decrease with the increase of θ2. Consequently, one can expect
that, at a given potential value, θ2 may be so high that the global
current for H+ reduction starts to decrease. Since in this potential
range, which corresponds to the non-linear part of the curves in
Fig. 1, the efficiency of the metallic deposition is still extremely
low, a decrease in current density values is observed in Fig. 1.
Accordingly, this current diminution is accompanied by a drop
in the interfacial pH, previously reported [16].
take place at lower potentials with increasing solution pH, as
shown in Table 1.
The impedance diagrams obtained at the very beginning of
the linear part of the polarisation curves for the Fe(II) solu-
tions (diagrams (a) in Figs. 5–9) are characterized by two
inductive loops at th+e low frequency domain. At this potential
˜
˜
E
˜
4
E
˜
˜
E
θ
3
θ
˜
[H
]
range,
,
and
are considered negligible and the main
˜
˜
θ
˜
E
θ
˜
1
relaxation occurs via
and 2 , both considered as positive.
Moreover, the assumptions ofEK1 + K9[H+] > K2 + K7[H+] and
K1 + K9[H+] > 2K5 + K5∗θ4 are necessary to assure that the
relaxation of both θ1 and θ2 can generate inductive behaviours.
However, with additional polarisation increase, a θ3 enrichment
occurs, giving rise to the emergence of a capacitive loop between
the two inductive ones previously formed. From Eqs. (3)–(5) the
˜
˜
E
˜
θ
˜
θ
3
θ
1
2
main relaxation frequencies associated with
,
and (fθ1,
˜ ˜
E E
Taking into account the effect of the solution pH on the
Fe electrodeposition kinetics, the formation of θ2 from θ1 is
considered to be pH dependent (reaction (III)). This means
that K3[H2O]θ1 corresponds to the conversion of θ1 into θ2
whereas K−3[H+]θ2 corresponds to the conversion θ2 into
θ1. According to this reaction, the lower the solution pH
the slower will be the formation of θ2. Consequently, the
production of θ2 will be speeded up by the pH increase.
As above discussed, the maximum of the current in this
potential range is a result of the transition on H+ reduction
rate: from K7[H+]θ1 + K9[H+](1 − θ1 − θ2 − θ3 − θ4) to only
K9[H+](1 − θ1 − θ2 − θ3 − θ4), when θ2 begins to replace θ1 at
the electrode surface. Actually, the position of current maxima
in the non-linear part of the polarisation curves (Fig. 1) depends
on the ratio between θ1 and θ2 (Eq. (4)) and will occur at lower
potentials with increasing solution pH, as observed in Fig. 1 and
detailed in Table 1.
In addition to the diminution of the value of the cathodic
potential where the current maxima take place, the increase in
solution pH causes the decrease of current densities in the non-
linear part of the curves in Fig. 1. This is due to a decrease of
the H+ concentration available for IH. Therefore, the negative
Rp observed in the impedance diagrams in Fig. 4 (pH 2.5 and
3.0) can be attributed to a blocking effect on the H+ reduction at
the electrode surface brought about by the effect of a high θ2. At
higher pH values (3.5, 4.0 and 5.0), the blocking effect on the
H+ reduction still exists, as confirmed from the decrease in the
interfacial pH, already reported [16]. However, the negative Rp
in the corresponding diagrams in Fig. 4 is no longer observed
owing to the lower H+ concentration in these solutions.
fθ2 and fθ3, respectively) can be calculated by [21,22]:
βfθ1 = K1 + K2 + K3[H2O] + K4[H2O]
βfθ2 = K−3[H+] + K5∗θ4
(7)
(8)
(9)
βfθ3 = K−4[H+]
Furthermore, the following kinetic conditions are required:
K1 + K9[H+] < 2K6 + K8[H+], to guarantee the capacitive relax-
˜
θ
ation of 3 ; and K−3 + K∗θ4 < K−4, to oblige the value of the
˜
5
E
˜
˜
˜
˜
˜
θ
1
θ
θ
3
2
characteristic frequency of to lie between those of and .
˜
From the results in FigsE. 5–9, the potential valueEin whiEch
the capacitive loop associated with the relaxation of θ3 appears
depends on the pH of the solution. To take into account this
result, the formation of θ3 from θ1 via reaction (IV) is pro-
posed. According to this reaction, the lower the solution pH,
the more difficult becomes the formation of θ3. Therefore, the
lower the pH, the higher will be the potential in which the capac-
itive loop might appear. Indeed, in the solution at pH 2.5, only
at 25 mA cm2 (Fig. 5(e)) a very small capacitive loop appears,
while at pH 5.0 the diagram obtained at 5.6 mA cm2 (Fig. 9(c))
already presents a big capacitive loop.
For even higher polarisations, the capacitive loop becomes
bigger, indicating a continuous increase on θ3, as seen in
Figs. 5–9. Simultaneously, theinductiveloopathigherfrequency
considerably decreases in size. From the reaction model, since
θ3 is produced at the expense of θ1, the mentioned inductive
feature is then associated with the relaxation of θ1, which is
progressively consumed.
Going further on cathodic polarisation, θ2 is enhanced and
the corresponding current densities continue going down until
reaching a minimum. As can be seen in Fig. 1, after the cur-
rent minima, for all polarisation curves, the current starts to rise
sharply and vary in a quite linear way with cathodic potential
increase. As already mentioned, high efficient metallic deposi-
tionwasonlyobservedinthisregion. Accordingly, thebeginning
of the linear part of the curves can thus be associated with
the rate of θ2 formation. If the production of this intermedi-
ate is favoured, the potential value where the linear region starts
becomes lower. Therefore, the minimum of current densities in
the non-linear part of all the polarisation curves in Fig. 1 will
Correspondingly, the inductive loop in very low frequencies
can be attributed to the relaxation of θ2. Another hypothesis is
that this inductive loop at very low frequencies could be related
to the coupling of the θ2 with θ4 relaxations via reaction (Vꢀ),
as an analogy with what was proposed by Epelboin et al. [19]
for nickel electrodeposition from acid sulphate solution. In this
˜
θ
4
case, to describe this loop at very low frequencies,
should
˜
be considered as negative, which is most probably toEoccur in
this potential range. The decision between the correlation of this
˜
˜
θ
4
θ
˜
E
2
inductive loop simply with
or coupled with
will only be
˜
E
possible during the simulation, already underway.

S.L. D´ıaz et al. / Electrochimica Acta 53 (2008) 7426–7435
7435
To summarize the above considerations, taking into account
F
Faraday’s constant, 96,500 C/eq
the data in Figs. 5–9, one should conclude that the Fe deposition
with high efficiency begins to occur essentially through θ1 and
θ2, via reactions (II) and (V), respectively. However, with fur-
ther polarisation increase, a θ2 and θ3 enrichment takes place at
the expense of θ1, whose concentration progressively decreases.
Despitethevanishingofθ1, whichisacatalyserforH+ reduction,
an increase in the interfacial pH was observed in the potential
range related to the linear branch of the polarisation curve for the
Fe(II) solution [16]. This behaviour can be attributed to the pres-
ence of high θ3 concentrations that also catalyses H+ reduction
through reaction (VIII). As a consequence, at very high polar-
isations, Fe electrodeposition takes place mainly by means of
reactions (V), (Vꢀ) and (VI). This means that, at this polarisation
condition, the electrodeposition becomes to occur through the
relaxation of both, θ3 and θ2 that give rise to big loops at the low
frequency domain, a capacitive and an inductive one, respec-
tively, due to their corresponding enrichment at the electrode
surface.
fθi
relaxation frequency associated with electrode cover-
age by the adsorbed species i ( s−1
)
+
[H ]
fluctuation of H+ concentration (mol)
current density (A cm−2
fluctuation of current density (A cm−2
current density for hydrogen evolution (A cm−2
˜
I
)
˜
I
)
IH
Im
)
imaginary part of the electrode impedance in the
Nyquist diagram (ꢀ cm2)
Ki
Re
rate constants of the reaction (I) (mol s−1
)
real part of the electrode impedance in the Nyquist
diagram (ꢀ cm2)
Rp
Rt
polarisation resistance (ꢀ cm2)
charge transfer resistance (ꢀ cm2)
Greek symbols
β
surface concentration of atoms (mol cm−2
)
electrode coverage by the adsorbed species i
fluctuation of the electrode coverage by the adsorbed
species i
θi
˜
θi
5. Conclusions
References
The experimental results presented in this paper show that
iron electrodeposition in acid sulphate medium occurs by means
of at least three adsorbed intermediates. At the non-linear part
of the polarisation curves, iron electrodeposition takes place at
very low efficiency. In this region, the main cathodic reaction is
the hydrogen reduction and the low efficient iron electrodepo-
sition proceeds by means of two intermediates. One of this two
adsorbed species catalyses the hydrogen reduction, whereas the
other blocks this process. The third intermediate species appears
only at the potential range corresponding to the linear part of the
polarisation curves. The surface concentration of this species
shows a strong dependence on the electrode potential.
[1] S. Yoshimura, S. Yoshihara, T. Shirakashim, E. Sato, Electrochim. Acta 39
(1994) 589.
[2] W. Schindler, O. Schneider, J. Kirschner, J. Appl. Phys. 81 (1997)
3915.
[3] E. Jartych, M. Jalochowsski, M. Budzynski, Appl. Surf. Sci. 193 (2002)
210.
[4] D. Landolt, E.J. Podlaha, N. Zech, Z. Phy. Chem. 208 (1999) 167.
[5] N. Zech, E.J. Podlaha, D. Landolt, J. Electrochem. Soc. 146 (1999)
2886.
[6] N. Zech, E.J. Podlaha, D. Landolt, J. Electrochem. Soc. 146 (1999) 2892.
[7] E. Jartych, J.K. Zurawicz, E. Maczka, J. Borc, Mat. Chem. Phys. 72 (2001)
356.
[8] B. Kabanov, R. Burtein, A. Frumkin, Discuss. Faraday Soc. 1 (1947) 259.
[9] K.F. Bonho¨effer, W. Jena, Z. Elektrochem. 55 (1951) 180.
[10] J.O’M. Bockris, D. Drazic, A.R. Despic, Electrochim. Acta 4 (1961) 325.
[11] K.-M. Yin, B.-T. Lin, Surf. Coat Technol. 78 (1996) 205.
[12] S. Gada, T.M. Harris, J. Electrochem. Soc. 145 (1998) 3699.
[13] D. Grujicic, B. Pesic, Electrochim. Acta 50 (2005) 4405.
[14] S. Hessami, C.W. Tobias, AIChE J. 39 (1993) 149.
[15] H. Deligianni, L.T. Romankiw, IBM J. Res. Dev. 37 (1993) 85.
[16] S.L. D´ıaz, O.R. Mattos, O.E. Barcia, F.J. Fabri Miranda, Electrochim. Acta
47 (2002) 4091.
Acknowledgements
The authors are grateful to the support given by the Brazil-
ian agencies: CNPq, FAPERJ, CAPES, FINEP and FUJB. J.A
´
Calderon also thanks RICICOP II—Alfa II project Europe
Union for the postdoctoral scholarship.
[17] I. Epelboin, M. Ksouri, R. Wiart, J. Electroanal. Chem. 65 (1975) 373.
[18] R. Wiart, Electrochim. Acta 35 (1990) 1587.
[19] I. Epelboin, M. Joussellin, R. Wiart, J. Electroanal. Chem. 119 (1981) 61.
[20] E. Chassaing, M. Joussellin, R. Wiart, J. Electroanal. Chem. 157 (1983)
75.
[21] M. Keddam, O.R. Mattos, H. Takenouti, J. Electrochem. Soc. 128 (1981)
257.
[22] M. Keddam, O.R. Mattos, H. Takenouti, J. Electrochem. Soc. 128 (1981)
267.
Appendix A. Nomenclature
bi
E
activation coefficient for electrochemical reactions
(V−1
potential (V)
)
˜
E
fluctuation of the potential (V)