Electrodeposition of nickel on carbon felt

Article abstract of DOI:10.1016/j.electacta.2004.06.004

In this work the study of nickel deposition on to carbon felt from chloride and sulphate electrolytes at pH 3, using electrochemical impedance spectroscopy, is presented. Results compare favourably to previous studies with the mechanism of nickel deposition dependent on the concentration and applied potential. All impedance spectra are characterised by the presence of a semicircle feature at high frequencies and by complex inductance features at low frequencies. For impedance studies performed in sulphate and chloride electrolytes, at potentials not corresponding to nickel deposition, an ion effect is observed. This effect is attributed to the formation of adsorbed pre-deposition intermediates; H_ads in sulphate and chloride electrolytes and Ni(Cl)_ads only in chloride electrolytes.

Full text of DOI:10.1016/j.electacta.2004.06.004

Electrochimica Acta 49 (2004) 4933–4938
www.elsevier.com/locate/electacta
Electrodeposition of Nickel on Carbon felt
a
Geoffroy R.P. Malpass , Maher Kalaji , Everaldo C. Venancio ,
b,1
a
a,1,
*
Artur J. Motheo
a
Instituto de Qu ´ı mica de S a˜ o Carlos, Universidade de S a˜ o Paulo, CP 780, SP 13560-970 S a˜ o Carlos- SP, Brazil
b
Department of Chemistry, University of Wales, Bangor LL57 2UW, UK
Received 26 January 2004; received in revised form 30 May 2004; accepted 6 June 2004
Available online 15 July 2004
Abstract
In this work the study of nickel deposition on to carbon felt from chloride and sulphate electrolytes at pH 3, using electrochemical
impedance spectroscopy, is presented. Results compare favourably to previous studies with the mechanism of nickel deposition dependent on
the concentration and applied potential. All impedance spectra are characterised by the presence of a semicircle feature at high frequencies
and by complex inductance features at low frequencies. For impedance studies performed in sulphate and chloride electrolytes, at potentials
not corresponding to nickel deposition, an ion effect is observed. This effect is attributed to the formation of adsorbed pre-deposition
intermediates; Hads in sulphate and chloride electrolytes and Ni(Cl)ads only in chloride electrolytes.
ß 2004 Elsevier Ltd. All rights reserved.
Keywords: Electrochemical impedance spectroscopy; Nickel; Electrodeposition; Carbon felt
1
. Introduction
electrolytes at low Ni concentrations. Fleischmann and
Saraby-Reintjes [5], suggest that deposition from a Watts
type bath proceeds via the chloride intermediate with no
Efficient recovery of metal ions from industrial waste is
+
both industrial and financially important. This is especially
true of nickel-bearing waste where both the quantity of
nickel produced and the high resale price make recovery
a commercially viable option. Examples of electrochemical
methods of retrieving metal ions from solution using carbon
felts or fibres as high-area electrodes can be found in the
literature [1–3]. In this context, the study of nickel deposi-
tion on carbon felt, resulting in an understanding of the
deposition mechanism, would be an important factor in
maximising the efficiency of any recovery process.
evidence of Ni(OH) .
The formation of two nickel phases containing included
hydrogen (H ), during electrodeposition has been
inc
reported [5,6]. These are known as a and b-nickel, both
being alloys of Ni and H with different Ni/H ratios in the
deposited structure. a-Nickel has a ratio of 0.03 and obeys
the laws of nucleation and growth. b-Nickel has a ratio of
between 0.6 and 1.0 and is its growth is dependent on the
migration of protons to the electrode surface. Formation of
b-nickel is accelerated at low pH and higher negative
potentials.
The mechanism of Ni deposition has been demonstrated
to depend on a number of factors, such as pH, concentration,
electrode material and bath additives. Proud and M u¨ ller [4]
studied the deposition of nickel on vitreous carbon and
Electrochemical impedance spectroscopy (EIS) is a
useful technique to study nickel deposition due to the
existence of previous studies with mechanistic models
elucidated using the technique [4,7–9], making interpreta-
tion easier. In this way the study of nickel deposition using
cyclic voltammetry and impedance spectroscopy is pre-
sented with a view to understanding the mechanistic pro-
cesses involved.
+
+
proposed the formation of Ni(OH) over NiCl in chloride
*
Corresponding author. Tel.: +55 16 273 9932; fax: +55 16 273 9952.
E-mail address: artur@iqsc.usp.br (A.J. Motheo).
1
ISE member.
0013-4686/$ – see front matter ß 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2004.06.004

4934
G.R.P. Malpass et al. / Electrochimica Acta 49 (2004) 4933–4938
2
. Experimental
analyser (Solartron S11250) linked to the potentiostat. The
amplitude of the alternated signal was 5 mVin the frequency
5
range of 10 –10 Hz.
ꢁ3
2
.1. Electrodes
In all cases the working electrode was a carbon felt disc
f = 1 cm, porosity > 90%) manufactured by Donac Co. Ltd.
(
from polyacrylnitrile (PAN) precursors. The counter elec-
4. Results and discussion
2
trodes were two platinum foils of area 1 cm each.
4.1. Cyclic Voltammetry
2
.2. Reagents and solutions
In Fig. 1 the cyclic voltammogram of carbon felt in
3
0 mmol/dm NiCl , at pH 3, repeated for two sweeps is
5
2
The following reagents were obtained from Fluka
purity 99.99%) and used without further purification—
shown. A cathodic current, commencing at ꢂꢁ0.3 V, is
observed when the potential is swept negatively. The catho-
dic current gives way to nickel deposition at ꢁ0.78 V. The
subsequent nickel-stripping peak is seen at ꢁ0.155 V. A
current loop, typical of a nucleation process, can be seen
between ꢁ0.78 and ꢁ0.90 V. The repeat sweep shows a
decrease in the amount of nickel deposited and stripped.
(
NiSO ꢀ6H O, NiCl ꢀ6H O and H BO . The salts Na SO
4
2
2
2
3
3
2
4
(
98%) and NaCl (>99.5%) were obtained from Avocado and
Merck, respectively and used without further purification.
3
Solutions of 50.0, 10.0 and 5.0 mmol/dm NiCl ꢀ6H O or
2
2
NiSO ꢀ6H O were prepared by dissolving the appropriate
4
2
salt in a previously prepared stock solution containing boric
This phenomenon might be due to the build-up of H in the
2
3
acid (8.1 mmol/dm H BO ). Sodium chloride (NaCl) or
inner areas of the carbon felt or to the presence of adsorbed
species at the electrode surface, as will be discussed at a later
stage in this paper.
3
3
sodium sulphate (Na SO ) was added to the solutions to
2
4
3
maintain the anion concentration at 1.0 and 0.5 mmol/dm ,
respectively. Chloride and sulphate solutions were acidified
to pH 3 using HCl or H SO , respectively.
In sulphate solutions at pH 3 deposition of nickel is
observed to proceed at lower overpotentials than for the
equivalent chloride systems, (Fig. 2). As in chloride systems
a cathodic current is observed and nickel deposition can be
seen to commence at ꢁ0.40 V. The stripping peak can be
seen at +0.145 V. Repetitive cycling results in a decrease in
the stripping peak, which is also shifted anodically (Fig. 2).
After 11 cycles the voltammetric response reaches a steady
state. The reasons for the peak decrease may be related to the
formation of different nickel phases, as will be described at a
later point.
2
4
3
. Procedure and equipment
In sulphate solutions the voltammetry and impedance
measurements were referred the mercury sulphate electrode
MSE), and to the saturated calomel electrode (SCE) in
(
chloride electrolytes. However, in the figures all potentials
are referred to the SCE to aid clarity.
All electrochemical measurements (CV and EIS) were
controlled using an EG & G/PAR potentiostat (model 273A)
and performed in the potential range of +0.6 to –1.250 V
versus SCE. All CVs were performed at 10 mV/s and
recorded using XY Recorder (Allen Datagraph). Impedance
measurements were performed using a frequency response
The formation of two phases of nickel (a- and b-nickel)
is observed when the potential is swept to increasingly
cathodic potentials for both chloride and sulfate systems.
3
In Fig. 3 the deposition of nickel from 50 mmol/dm NiCl is
2
observed when the potential is swept to ꢁ0.90 V showing a
single stripping peak at ꢁ0.156 V. When the potential is
3
Fig. 1. CV of carbon felt disc (f = 1 cm) in 50 mmol/dm NiCl
2
at pH 3: (solid line) 1st sweep; (dashed line) 2nd sweep; v = 10 mV/s.

G.R.P. Malpass et al. / Electrochimica Acta 49 (2004) 4933–4938
4935
3
Fig. 2. CV of carbon felt disc (f = 1 cm) in 50 mmol/dm NiSO
4
at pH 3: (solid line) 1st sweep; (dashed line) 11th sweep (after reaching steady state); v =
10 mV/s.
swept to a more cathodic potential at ꢁ1.00 V, a second
stripping peak is observed at ꢁ0.215 V, with each of the two
peaks representing a different nickel/hydrogen alloy [5,6]. In
addition, the stripping peak observed at –0.156 V (for a limit
potential of –0.90 V) shifts anodically to ꢁ0.145 V (Fig. 3).
The peaks at ꢁ0.145 and ꢁ0.215 V can be attributed to the
stripping of a-and b-nickel, respectively [5,6]. The fact that
there is a peak shift as the cathodic limit potential is
increased, indicates that phase separation of a- and b-nickel
occurs at more negative potentials. Subsequently this would
indicate that the stripping peaks observed in Figs. 1 and 2 are
formed by combination of a- and b-nickel with separation of
phases only being seen at more negative potentials. Such a
picture of mixed deposition in the initial stages of nickel
deposition would be in agreement with previous studies by
Fleischmann and Saraby-Reintjes [5].
potential required to deposit nickel. In Fig. 4 the voltammo-
3
gram of nickel deposition from 5 mmol/dm NiCl at pH 3 is
2
shown. The cathodic current is present and the nickel-
stripping peak is shifted anodically to ꢁ0.210 Vand is much
broader and less well defined. The current loop behaviour
3
observed in 50 mmol/dm is not seen. Repeated sweeps
demonstrate that the stripping peak does not diminish as
seen previously. In the cathodic region of Fig. 4, a decrease
in the current response with repeated cycling can be
observed. It is possible that this decrease is related to the
formation of intermediates involved in H formation or to
2
adsorbed Ni intermediates, both of which block the electrode
surface (see later in paper).
In the literature, various authors [4,7–9] suggest that
nickel deposition occurs via the initial formation an
adsorbed Ni(I) intermediate. Ni(OH)ads and NiClads have
been proposed. The formation of species of this nature has
been reported to occur at more anodic potentials than nickel
The effect of decreasing the concentration is manifested
in chloride and sulfate systems by an increase in the over-
3
Fig. 3. CV of carbon felt disc (f = 1 cm) in 50 mmol/dm NiCl
2
at pH 3. Cathodic limit: ꢁ0.9 V (solid line); ꢁ1.0 V (dashed line); v = 10 mV/s.

4936
G.R.P. Malpass et al. / Electrochimica Acta 49 (2004) 4933–4938
3
Fig. 4. CV of carbon felt disc (f = 1 cm) in 5 mmol/dm NiCl
2
at pH 3: (solid line) 1st sweep; (dashed line) 2nd sweep; v = 10 mV/s.
deposition itself [4]. With this in mind, it can be suggested
that the cathodic current observed prior to Ni deposition
formation of Hads and H gas in the inner pores of the carbon
2
felt is a possible reason for this passivation, However, H2
evolution was not observed. The formation of NiClads has
been suggested by Chassaing et al. [7] as a passivating factor
in nickel deposition from chloride electrolytes. Secondly, in
this study the phenomenon is only observed in chloride
electrolytes. The formation of a passivating NiClads inter-
mediate is suggested to be the cause of the inhibition of
nickel deposition here. As stated the deposition and stripping
of nickel does not disappear completely with repeated
sweeps. It was found [7] that the presence of boric acid
in chloride electrolytes broke or inhibited the passivation of
the electrode surface at higher overpotentials. Hoare [10]
found that boric acid acted as a catalyst in nickel deposition.
It is probable that the small quantity of boric acid present in
the electrolyte prevents the complete passivation of the
electrode towards nickel deposition. In sulphate solutions
there is no analogous effect probably due to the weaker
sulphate ion not interacting in the system to the same extent
as the chloride ion [4].
(Fig. 1) is due to the formation of such an intermediate at the
electrode surface. As the feature is observed in both systems,
the intermediate formed would most probably be the Ni(O-
H)ads species.
In chloride media, continued cycling shows that the Ni
stripping peak diminishes until reaching a steady state, but
does not disappear completely. The diminishing of the peak
current could be due to a number of factors. For example, the
3000
2500
2000
1500
1000
1
00 mHz
For sulphate electrolytes the effect of repeated sweeps is
clearly demonstrated as the 11th sweep is diminished and
clearly shifted anodically (Fig. 2). A substantial quantity of
hydrogen gas was observed to form over this time, but the
peak does not diminish at the same rate as in chloride
systems, this again supports the hypothesis of a passivating
NiClads species in chloride electrolytes. At pH 3 nickel
deposition is accompanied by high-levels of hydrogen
evolution [6]. It is reported by Proud et al. [6] that the
1
00 mHz
500
1
Hz
formation of a b-nickel enhances the formation of H and
2
10 MHz
0
this in turn promotes the production of a-nickel. The
production of a-nickel is manifested in Fig. 2 by the peak
shift or in effect phase separation of the two nickel forms,
0
750
1500
Z/
2250
3000
3
i.e. over repeated sweeps the H , incorporated in the b-
2
Fig. 5. Impedance plots of carbon felt disc (f = 1 cm) in 50 mmol/dm
(open square) at ꢁ0.2 V vs.
3
(full square) and 50 mmol/dm NiSO
NiCl
SCE.
2
4
nickel phase, evolves and leaves only a-nickel resulting in
an anodic peak shift.

G.R.P. Malpass et al. / Electrochimica Acta 49 (2004) 4933–4938
4937
-20
-16
-12
10 Hz
-8
1
Hz
-4
Inductance loops
10 MHz
0
10 mHz
2
1
0
4
8
12
16
20
Z/Ω
3
Fig. 6. Impedance plots of carbon felt (f = 1 cm) in 50 mmol/dm NiCl
2
at ꢁ0.8 V vs. SCE.
4
.2. Impedance studies
species was thought to be a chloride containing complex
7] (e.g. NiClads), Eqs. (1) and (2).
[
The overpotential for the deposition of nickel on carbon
felt was observed to be concentration dependent. The non-
faradaic region (were no electron transfer takes place) varied
þ
ꢁ
NiX þ e $ NiX
(1)
(2)
ads
ꢁ
ꢁ
NiXads þ e $ Ni þ X
where X is HO or Cl . Proud and M u¨ ller [4] suggested
that the predominant species at the low concentrations was
Ni(OH)ads. However, Chassaing et al. [7] and Epelboin et al.
from ꢁ0.78 V in 50 mM NiCl and ꢁ0.40 V in 50 mM
2
ꢁ
ꢁ
ꢁ
NiSO to more negative potentials for lower Ni(II) concen-
4
trations. In all impedance spectra the presence of a high
frequency semicircle feature, typical of porous electrodes, is
observed.
[8] found no evidence of a hydroxylated nickel intermediate.
It is also difficult to imagine that a Ni(OH) species at pH 3
+
An interesting feature of the impedance spectra taken in
both electrolytes is their non-linearity in regions of poten-
tials where nickel deposition does not occur (Fig. 5). At
potentials prior to Ni deposition, an ion effect can be seen in
both electrolytes. From Fig. 4 it can be observed that the
spectra obtained in chloride solutions are less linear than
those obtained in sulphate media. This behaviour was also
observed by Proud and M u¨ ller [4] using vitreous carbon as
the electrode material. Proud and M u¨ ller attributed the non-
would be long-lived. In this way the inductance feature is
ꢁ
probably not attributable to desorption of HO .
The second inductive feature is ascribed to strongly
bonded Hads on the electrode surface and is observed to
disappear at more negative potentials (Fig. 7) being replaced
by a capacitive feature. The variation of coverage by Hads
induces the capacitive loop.
-
11
10
+
linear behaviour to the adsorption of Ni(OH) in both
-
chloride and sulphate electrolytes [4]. The accentuated
nature of the non-linearity in chloride electrolytes indicates
the additional adsorption of NiClads [4].
-
9
8
-
-7
In chloride electrolytes at potentials corresponding to the
deposition of nickel two low frequency inductive features
can be seen (Fig. 6). The presence of two inductive features
is in agreement with observations made by Chassaing et al.
-
6
-5
-4
-3
-2
10 Hz
1
0 mHz
1
Hz
[
7] during impedance studies of nickel deposition on brass or
-
1
0
1
10 MHz
nickel electrodes. Proud and M u¨ ller [4] only observed one
inductive feature in chloride electrolytes on vitreous carbon.
In chloride solutions the first of the inductive loops is
ascribed to the slow desorption of an anionic species and
the second to the presence of Hads on the electrode surface
Capacitive feature
1
2
3
4
5
6
7
8
9
10
11
Z/
3
Fig. 7. Impedance plots of carbon felt (f = 1 cm) in mmol/dm NiCl
[7]. It was found that the first feature was not dependent on
the chloride concentration and as a result the adsorbed
2
at –
1
.0 V vs. SCE.

4938
G.R.P. Malpass et al. / Electrochimica Acta 49 (2004) 4933–4938
In sulphate solutions only one low frequency inductive
subsequent conversion to H2 (capacitance). In chloride
electrolytes an additional inductance feature is observed
that can be attributed to the formation of a passivating
NiClads intermediate, agreeing with the conclusions drawn
from cyclic voltammetry.
feature is observed. This phenomenon was observed [4] to
depend on the pH of the solution and is ascribed to Hads on
the electrode surface. This feature is seen to give way to a
capacitive loop due to the variation of electrode coverage, as
seen in chloride systems previously.
5.1. Mechanism
5
. Conclusions
A global mechanism can be suggested when considering
the literature in conjunction with the present results:
The study of the deposition of nickel from sulphate and
Ni2 þ X $ NiX
þ
ꢁ
ꢁ
ꢁ
þ
(3)
(1)
(2)
chloride electrolytes on carbon felt presents similar results
to those previously shown in the literature. Cyclic voltam-
metry and impedance results suggest that a number of
intermediate complexes and adsorbed hydrogen, Hads, are
involved. The behaviour in chloride electrolytes is charac-
terised by a rapid decrease in the peak associated with Ni-
stripping. In sulphate electrolytes the Ni-stripping peak also
decreases, but to a lesser degree than in chloride systems.
Considering the literature, the more accentuated decrease is
probably due to the additional formation of a passivating
NiClads intermediate.
þ
NiX þ e $ NiX
ads
ꢁ
NiX þ e $ Ni þ X
ads
ꢁ
ꢁ
In the case of X = HO the reaction in steps 1 and 2
would be very fast and no inductance feature is seen at low
ꢁ
ꢁ
frequencies. However, for X = Cl the discharge is much
slower and results in the inductance feature described.
References
In both electrolytes, the impedance spectra are charac-
terised by the presence of a semicircle feature at high
frequencies and by complex inductance features at low
frequencies. For impedance studies performed in sulphate
and chloride electrolytes, at potentials not corresponding to
nickel deposition, an ion effect is observed, which is
attributed to the formation of adsorbed pre-deposition
intermediates.
[
[
1] Y. Oren, A. Soffer, Electrochim. Acta 130 (9) (1983) 1814.
2] E.O. Vilar, E.B. Calvalcanti, H.R. Carvalho, F.B. Sousa, Braz. J.
Chem. Eng. 20 (3) (2003) 291.
[
[
[
3] B. Delange, S. Telier, M. Astruc, Environ. Tech. 11 (1990) 999.
4] W.G. Proud, C. Muller, Electrochim. Acta 38 (2–3) (1993) 405.
5] M. Fleischmann, Saraby-Reintjes, Electrochim. Acta 29 (4) (1984)
5
57.
[
[
[
[
6] W.G. Proud, E. Gomez, E. Sarret, E. Valles, C. Muller, J. Appl.
Electrochem. 25 (1995) 770.
7] E. Chassaing, M. Joussellin, R. Wiart, J. Electroanal. Chem. 157
Both electrolytes present an inductance feature at low
frequencies and at low potentials. At higher potentials, in the
region of the hydrogen evolution reaction, this inductance
feature gives way to a capacitance feature. This phenomenon
is attributable to the formation of Hads (inductance) and its
(
1983) 75.
8] Epelboin, M. Joussellin, R. Wiart, J. Electroanal. Chem. 119 (1981)
1.
9] M. Holm, T.J. O’Keefe, J. Appl. Electrochem. 30 (2000) 1125.
6
[10] J.P. Hoare, J. Electrochem. Soc. 133 (12) (1986) 2491.