Electrodeposition of nano-structured nickel-21% tungsten alloy and evaluation of oxygen reduction reaction in a 1% sodium hydroxide solution

Article abstract of DOI:10.1016/j.electacta.2004.04.037

The nano-structured nickel-21at.% tungsten alloys were electrodeposited onto the copper substrates from unstirred sulfate-citrate-chloride-bromide-sodium tungstate electrolyte at 60°C. The maximum particle sizes of the deposits, as estimated from the atomic force microscopy (AFM), and scanning electron microscopy (SEM), respectively were 125, 75, and 100nm. The Tafel plot for oxygen reduction reaction in oxygenated unstirred 1% sodium hydroxide solution showed a Tafel slope of 130mV/decade. There were minor variations in the limiting current density with a change in the particle size.

Full text of DOI:10.1016/j.electacta.2004.04.037

Electrochimica Acta 49 (2004) 4411–4416
Electrodeposition of nano-structured nickel–21% tungsten
alloy and evaluation of oxygen reduction reaction in a
1% sodium hydroxide solution
a
a
b
c
a,∗
Miaomiao Ma , V.S. Donepudi , G. Sandi , Y.K. Sun , J. Prakash
a
Department of Chemical and Environmental Engineering, Center for Electrochemical Science and Engineering,
Illinois Institute of Technology, 10, W.33rd Street, Chicago, IL 60616, USA
b
Chemistry Division, Argonne National Laboratory, Argonne, IL 60616, USA
c
Department of Chemical Engineering, Hanyang University, South Korea
Received 12 November 2003; accepted 2 April 2004
Available online 9 June 2004
Abstract
The nano-structured nickel–21 at.% tungsten alloys were electrodeposited onto the copper substrates from unstirred sulfate–citrate–
◦
chloride–bromide–sodium tungstate electrolyte at 60 C. The maximum particle sizes of the deposits, as estimated from the atomic force
microscopy (AFM), and scanning electron microscopy (SEM), respectively were 125, 75, and 100 nm. The Tafel plot for oxygen reduction
reaction in oxygenated unstirred 1% sodium hydroxide solution showed a Tafel slope of 130 mV/decade. There were minor variations in the
limiting current density with a change in the particle size.
©
2004 Elsevier Ltd. All rights reserved.
Keywords: Electrodeposition; Nanostructured; Oxygen reduction; Atomic force microscopy; Tafel plot
1
. Introduction
loys such as nickel–tungsten. Numerous attempts have been
made to electrodeposit tungsten from various types of baths
[1]. Much of the early literature has been reviewed by Holt
and Kahlenberg [2]. It appeared that thin films of pure
tungsten could be electrodeposited by electrolyzing differ-
ent alkaline baths. However, this was proved incorrect be-
cause the deposits contained metallic iron as impurity [3].
A claim that pure tungsten can be electrodeposited from or-
ganic alkyl hydrocarbon polyamine [4] could not be con-
firmed. The electrodeposition of pure nickel from aqueous
electrolytes has been known for sometime and nickel plating
from sulfate–chlorideWatts type bath has been commercial-
ized for long time [1]. The anomalous nature of electrode-
position of tungsten alloys with transition metals such as
iron, nickel and cobalt is well known and the key features of
anomalous codeposition are: high overpolarization, poor ef-
ficiencies, and low tungsten content in the alloy deposit even
at high sodium tungstate concentrations in the electrolyte.
Some significant electrolyte systems used in electrodepo-
sition of nickel–tungsten alloys include the citrate baths
The applications of the electrodeposited metals and al-
loys are very well known and some of their applications can
be found in corrosion protection, electronics, and cataly-
sis. Hard microstructures are needed in the molding process
for LIGA applications and this is where electrodeposited
ductile and amorphous nickel–tungsten alloys are becom-
ing important. Similarly, low-cost and efficient electrocat-
alysts are becoming important for the fuel cell electrodes.
One of the major challenges in fuel cell electrocatalysis is
to replace platinum by inexpensive non-noble metals. The
nickel–tungsten alloy can be particularly attractive because
of its low-cost and ease of synthesis. However, the stud-
ies on the electrocatalytic properties of the nickel–tungsten
alloys are missing and such studies are important in prob-
ing new applications of electrodeposited nano-structured al-
∗
Corresponding author. Tel.: +1-312-567-3639;
fax: +1-312-567-8874.
[1,5,6], complex pyrophosphate baths [7], sulfamate baths
E-mail address: prakash@iit.edu (J. Prakash).
0
013-4686/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2004.04.037

4
412
M. Ma et al. / Electrochimica Acta 49 (2004) 4411–4416
◦
[8], and sulfate–citrate–chloride–bromide [9]. A recent study
Unstirred solutions at 60 C were used for the deposi-
reports the pulse plating effects during the nickel–tungsten
alloy electrodeposition [10,11] and suggest that pulse plating
can improve surface roughness. The objective of the present
study is to electrodeposit nano-structured nickel–tungsten
alloy using direct current and pulse plating procedures and
to investigate their electrocatalytic activity towards oxygen
reduction reaction as a function of particle size as controlled
by the plating current profile.
tion potential experiments as well alloy plating experiments.
Etched 1 cm × 1 cm copper foils were used as substrates.
The etching of the copper substrate was carried out in order
to improve the deposition and adhesion of the deposits.
The nickel–tungsten alloy electrodeposition was con-
ducted using a dc current and pulse plating methods. The
plating conditions are described in the legends in Fig. 2. A
CHI Electrochemical Workstation (Model CHI 660A), hav-
ing a wide variety of functional capabilities was used for the
polarization, dc and pulse plating experiments. A saturated
calomel electrode was used as a reference electrode and all
the measured potentials are referred to the same scale.
2
. Experimental
2
.1. Electrodeposition
2.2. Morphology of the deposits
The single metal and alloy deposition potentials for the
The morphology of the deposits was determined using the
electrodeposition of tungsten, nickel and nickel–tungsten al-
loy were determined from the current–potential curves in the
citrate–chloride–bromide based solutions (Aldrich Chemi-
cal Company) containing separately (a) sodium tungstate,
b) nickel sulfate and (c) a mixture of sodium tungstate and
nickel sulfate. The solution concentrations used in this study
are listed below:
atomic force microscopy (AFM) and the scanning electron
microscopy (SEM) techniques. The AFM on the deposits
was performed using the Digital Instruments Nanoscope.
The surfaces were scanned using tapping mode of scanning.
A canonical probe (ultrasharp hyperboloid at the extreme tip
used was silicon/doped n+ tip from nanosensor. The height
of the tip was about 10 ␮m. The cantilever had a spring
constant of 40–75 N/m (resonating frequency: ∼300 kHz).
Identical scanning parameters were used for all the
samples.
(
1
2
3
4
. Common constituents: sodium citrate (0.4 M), ammo-
nium chloride (0.5 M) and sodium bromide (0.15 M).
. Tungsten deposition potentials: solution (1) + sodium
tungstate dihydrate (0.14 M).
. Nickel deposition potentials: solution (1) + nickel sulfate
2.3. Oxygen reduction reaction
(0.06 M).
. Ni–W alloy deposition potentials: solution (1) + sodium
The electrocatalytic activities of the nickel–tungsten alloy
electrodeposits were compared by examining the potential
tungstate dihydrate (0.14 M) and nickel sulfate (0.06 M).
Fig. 1. Current–potential curves for single metal and alloy plating. Composition: (1) sodium citrate (0.4 M), ammonium chloride (0.5 M) and sodium
bromide (0.15 M). Curve (a): tungsten, solution composition (1) + sodium tungstate dihydrate (0.14 M); curve (b): nickel, solution composition (1)
+
nickel sulfate (0.06 M); curve (c): Ni–W alloy, solution composition (1) + sodium tungstate dihydrate and nickel sulfate (as for curves a and b).

M. Ma et al. / Electrochimica Acta 49 (2004) 4411–4416
4413
versus logarithm current density relationships for the oxy-
3. Results and discussion
gen reduction reaction. The slow sweep polarization curves
were obtained using a three electrode system consisting
of the nickel–tungsten alloy working electrode, a saturated
calomel reference electrode, and a platinum counter elec-
trode. All the measured potentials are referred to the same
reference electrode scale. An all-glass-three compartments,
hollow cylindrical cell having a 150 ml volume working
electrode compartment was used. Oxygen saturated, un-
stirred 1% NaOH solutions (pH at room temperature = 12.7)
The current–potential curves for the deposition of tung-
sten, nickel and the nickel–tungsten are shown in Fig. 1. The
deposition potentials, as estimated from the points of inflex-
ion in these curves are: tungsten, −1.3 V (curve a); nickel,
−1.25 V (curve b); nickel–tungsten alloy, −1.20 V (curve
c). There is a 100 mV depolarization for the alloy deposi-
tion potential in comparison to that of tungsten, a metal that
cannot be easily deposited from the aqueous solutions.
The atomic force micrographs (AFMs) of the direct cur-
rent and pulse electrodeposited nickel–tungsten alloys are
shown in Fig. 2. The morphology of the copper substrate was
◦
at room temperature (∼23 C) were used for the slow
sweep polarization experiments at scan rates of 0.004 and
0
.05 mV/s.
2
Fig. 2. AFMs of electrodeposited Ni–21 at.% W alloy samples. Alloy plating solution composition see Fig. 1. Conditions: (a) dc current = 0.1 A/cm ;
(
b) differential pulse voltammetry, initial E = 0.0 V, final E = −1.6 V, increment = 0.004 V, pulse width = 0.05 s, sampling width = 0.0167 s and pulse
period = 0.1 s; (c) conditions as in (a), pulse period = 0.2 s, quiet time = 2 s, sensitivity = 0.1 A/V; (d) as in (a), pulse period = 0.5 s; (e) square wave
voltammetry, initial E = −1.2 V, final E = −1.4 V, increment = 0.004 V, amplitude = 0.025 V and frequency = 15 Hz.

4
414
M. Ma et al. / Electrochimica Acta 49 (2004) 4411–4416
similar for all experiments and we observed that the prede-
position morphology of the copper substrate did not have any
significant effect on the deposited morphology. A banded
structure with a maximum particle size of 100 nm can be
seen for the sample that has been plated using a direct current
of 100 mA/cm for 300 s (Fig. 2a). On the other hand, use of
a differential pulse produced deposits of different morpholo-
gies depending upon the pulse duration (conditions: initial E
approximately complement with those observed in the AFM
images. In addition, the SEM image of the plated sample us-
ing a direct current of 100 mA for 300 s (Fig. 3a) shows the
presence of distinct grain boundaries and of a second phase
within the grains. These results are consistent with the pre-
vious studies [10,11] carried out under similar experimental
conditions and suggest that the electrodeposition produces
a Ni–W alloy on the copper substrate. Superimposition of
differential pulse leads to possible dissolution of the second
phase particles when the potential reaches a value of 0.0 V.
Here again, the superimposition of 15 Hz results in a de-
posit having well rounded particles (Fig. 3b). Contrary to
the AFM, the columnar structures are obvious and this may
be due to the magnification of the SEMs.
2
=
0.0 V, final E = −1.6 V, increment = 0.004 V, pulse width
=
0.05 s and sampling width = 0.0167 s; variable = pulse
duration). Deposits with a maximum particle size of 75 nm
resulted for pulse durations of 0.1 and 0.2 s. On the other
hand, 125 nm sized particles resulted for a pulse duration of
0.5 s. Again a banded structure resulted in this case. Appli-
cation of square wave of 15 Hz (initial E = −1.2 V, final E
Fig. 4 shows the logarithm current density versus poten-
tial curves for the oxygen reduction reaction at electrode-
posited Ni–21 at.% W alloy in unstirred, oxygenated 1%
=
−1.4 V, increment = 0.004 V and amplitude = 0.025 V)
gave rise to a columnar deposit having a maximum particle
size of 100 nm.
The scanning electron micrographs (SEMs) of the direct
current and pulse electrodeposited nickel–tungsten alloys are
shown in Fig. 3. The energy dispersive scan showed that the
tungsten content in the alloy is 21 at.%. The particle sizes
◦
sodium hydroxide solution at room temperature (∼23 C).
These curves were obtained at a scan rate: 0.05 mV/s. It can
be seen that the currents in the potential −0.2 to −0.4 V for
the dc current plated sample having a maximum particle size
= 125 nm (curve 1) are lower than those for the differential
Fig. 3. SEMs of electrodeposited Ni–21 at.% W alloy samples. Conditions as in Fig. 2: (a) Direct current for 300 s, (b) pulse deposition.

M. Ma et al. / Electrochimica Acta 49 (2004) 4411–4416
4415
electron transfer step [11,12]. On the basis of the present
evidence we think that the dissocialize adsorption of oxygen
is the most likely pathway for oxygen reduction consistent
with previous studies [11,12] on metal surfaces. In order to
be reduced the oxygen molecule has to reach on the surface
of the electrode and interact with the electrode surface, and
this is very much dependent on the electrode/electrolyte in-
terface. The interaction of these species with the electrode
surface may determine the effectiveness of the catalyst for
oxygen reduction. Unlike noble metals/electrolyte interface
in which H2O molecules are adsorbed by weak interactions
with the metal surface, the oxide covered metal/electrolyte
interface consists of a hydroxyl Ted surface. This is be-
cause when a metal oxide comes in contact with the water,
the metal ions on the surface get coordinately saturated
Fig. 4. Logarithm current density vs. potential curves for oxygen reduction
reaction at electrodeposited Ni–21 at.% W alloy in unstirred, oxygenated
◦
1
0
% sodium hydroxide solution at room temperature (∼23 C). Scan rate:
.05 mV/s. Conditions—curve 1: dc current plated sample, maximum
particle size = 125 nm; curve 2: differential pulse plated sample using
conditions stated in Fig. 2b, maximum particle size = 75 nm; curve 3:
square wave plated sample using the conditions stated in Fig. 2e, maximum
particle size = 100 nm.
−
by binding OH and this in turn hydroxylates the surface
[
13,14]. The metal atoms in this hydroxylated surface are
−
2−
chemically bonded to OH or O . For oxygen reduction
to occur on this surface, O2 has to displace these groups
from the surface before reaching the surface of the oxide.
Oxygen is a non-polar molecule and therefore the proba-
pulse plated (curve 2) or square wave plated samples (curve
3
) having the maximum particle sizes of 75 and 100 nm,
respectively. On the other hand, there are only minor differ-
ences in the current densities in the limiting current region.
Marginal differences in the limiting current densities as a
function of particle size for differently electrodeposited the
nickel–tungsten alloy can be attributed to the narrow dif-
ferences in the maximum particle sizes, i.e., 125 nm for dc
plated sample versus 75 nm for differential pulse plated sam-
ple. Despite the differences in the particle size, the over-all
surface area differences, based on a range of particle sizes
for the given samples may be marginal, thereby producing
marginal changes in the electrochemical effects for the oxy-
gen reduction reaction.
−
2−
bility of O2 to replace chemically bonded OH or O is
negligible. Therefore, the first step probably involves an
outer sphere electron transfer to form a superoxide ion,
which is polar in nature and hence can find its way to reach
metal sites for subsequent reaction. We also observed only
minor variations in the Tafel slopes and the limiting current
density with a change in the particle size of the electrode-
posits. This result is important since the corrosion rates in
oxygen containing alkaline solutions are proportional to the
limiting current densities for oxygen diffusion [8]. In this
particular study, there were only minor changes in the lim-
iting current densities for oxygen reduction for the deposits
having the particle sizes between 75 and 125 nm, thereby
meaning that the corrosion rates do not vary substantially
when the particle size differences are marginal.
Fig. 5 shows the logarithmic current density versus poten-
tial curves for oxygen reduction at differential pulse plated
Ni–21 at.% W alloy sample in unstirred, oxygenated 1%
◦
sodium hydroxide solution at room temperature (∼23 C).
A linear relationship is obtained with a Tafel slope of
1
30 mV/decade and is suggestive of a rate-determining first
4. Conclusions
Electrodeposition of nickel–21 at.% tungsten alloys was
possible from the citrate–bromide electrolyte. Variation in
the direct and pulse plating parameters and use of differ-
ential pulse voltammetry and square wave voltammetry re-
sulted in the electrodeposits with particle sizes from 75 to
125 nm. A linear relationship is obtained with a Tafel slope
of 130 mV/decade was observed for the oxygen reduction
reaction at the electrodeposits in unstirred, oxygen satu-
rated, 1% sodium hydroxide solution at room temperature
and is suggestive of a rate-determining first electron trans-
fer step. There were marginal variations in the limiting cur-
rent densities for the oxygen reduction as the particle size
decreased from 125 to 75 nm suggesting that the corrosion
rates do not vary substantially when the particle size differ-
ences are marginal.
Fig. 5. Logarithm current density vs. potential curves for oxygen reduc-
tion reaction at a scan rate of 0.04 mV/s on Ni–21 at.% W alloy elec-
trodeposited using dc current.

4
416
M. Ma et al. / Electrochimica Acta 49 (2004) 4411–4416
[6] M.L. Holt, L.E. Vaaler, Trans. Electrochem. Soc. 90 (1946)
Acknowledgements
43.
[
[
[
7] K.I. Vasu, T.L. Ramachar, J. Sci. Ind. Res. (India) 19B (413) (1960)
This work is supported in part by the University IT Re-
search Center Project (Hanyang University, South Korea).
510.
8] N. Atanassov, K. Gencheva, M. Bratoeva, Plat. Surf. Finish. 84
1997) 67.
(
9] T. Yamasaki, R. Tomohira, Y. Ogino, P. Schlossmacher, K. Ehrlich,
Plat. Surf. Finish. 148 (2000) 87.
References
[
10] S.-L. Lee, Y.-F. Lee, M.-H. Chang, J.-C. Lin, Corros. Prevent. Control
[
1] M.L. Holt, Uncommon metals, in: F.A. Lowenheim (Ed.), Modern
Electroplating, 3rd ed., Wiley, 1974, (Chapter 17) p. 467.
2] M.L. Holt, L. Kahlenberg, Met. Ind. 31 (1933) 94.
3] M.L. Holt, Trans. Electrochem. Soc. 66 (1934) 453.
4] C.C. Hartford, US Patent 2,384,301 (1945).
71 (1999).
[11] J. Prakash, H. Joachin, Electrochim. Acta 45 (2000) 2289.
[12] R.E. Meyer, J. Electrochem. Soc. 107 (1960) 847.
[13] R.G. Edgell, J.B. Goodenough, A. Hammet, C.C. Naish, J. Chem.
Soc., Farad. Trans. I 79 (1983) 893.
[14] J. Prakash, D. Tryk, E.B. Yeager, J. Electrochem. Soc. 146 (1999)
4145.
[
[
[
[
5] A. Brenner, P. Burkhead, E. Seegmuller, J. Res. Natl. Bur. Standards
39 (1947) 351.