Hydrogen evolution on porous Ni cathodes modified by spontaneous deposition of Ru or Ir

Article abstract of DOI:10.1016/j.electacta.2008.06.056

Porous Ni deposits, prepared by cathodic deposition, were modified by immersing them in acid deaerated solutions containing Ru(III) or Ir(IV) chloride complexes with which they readily reacted, without any activation procedure, giving rise to spontaneous deposition of either Ru or Ir. The obtained electrodes were investigated by cyclic voltammetry, impedance spectroscopy and scanning electron microscopy. All data showed that the initial large area of the Ni deposits further increased upon immersion in solutions of noble metal complexes. EDX analyses proved that the deposition of Ru reached a limiting situation in some hours, while that of Ir was slower and continued for a longer time. The persistence of intense peaks due to the Ni(II)/Ni(III) redox system showed that Ru and Ir did not form a continuous layer able to prevent the contact between Ni and electrolyte. Hydrogen evolution was studied in 1 M NaOH solutions. Spontaneous deposition of both noble metals markedly improved the performance of porous Ni. The best results were achieved with Ir-modified electrodes, after immersion in Ir(IV) solution for 6 h. Tafel slopes and overpotentials of Ru-modified electrodes were not as low as those of Ir-modified electrodes.

Full text of DOI:10.1016/j.electacta.2008.06.056

Electrochimica Acta 53 (2008) 8310–8318
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Hydrogen evolution on porous Ni cathodes modified by spontaneous
deposition of Ru or Ir
Lourdes Vázquez-Gómeza, Sandro Cattarina, Paolo Guerrierob, Marco Musiania,∗
a IENI CNR, Corso Stati Uniti 4, 35127 Padova, Italy
b ICIS CNR, Corso Stati Uniti 4, 35127 Padova, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 4 April 2008
Received in revised form 5 June 2008
Accepted 11 June 2008
Available online 2 July 2008
Porous Ni deposits, prepared by cathodic deposition, were modified by immersing them in acid deaerated
solutions containing Ru(III) or Ir(IV) chloride complexes with which they readily reacted, without any
activation procedure, giving rise to spontaneous deposition of either Ru or Ir. The obtained electrodes
were investigated by cyclic voltammetry, impedance spectroscopy and scanning electron microscopy. All
data showed that the initial large area of the Ni deposits further increased upon immersion in solutions
of noble metal complexes. EDX analyses proved that the deposition of Ru reached a limiting situation
in some hours, while that of Ir was slower and continued for a longer time. The persistence of intense
peaks due to the Ni(II)/Ni(III) redox system showed that Ru and Ir did not form a continuous layer able to
prevent the contact between Ni and electrolyte. Hydrogen evolution was studied in 1 M NaOH solutions.
Spontaneous deposition of both noble metals markedly improved the performance of porous Ni. The best
results were achieved with Ir-modified electrodes, after immersion in Ir(IV) solution for 6 h. Tafel slopes
and overpotentials of Ru-modified electrodes were not as low as those of Ir-modified electrodes.
© 2008 Elsevier Ltd. All rights reserved.
Keywords:
Electrocatalysis
Surface roughness
Impedance
Electrodeposition
Activation
1. Introduction
In many cases the surface roughness of Raney-Ni and similar elec-
trodes was comprised between 103 and 104, while the pore size
Ni electrodes of high specific surface area combine many desir-
able properties of an electrocatalyst for hydrogen evolution: large
surface roughness (defined as the ratio between true area and
geometric area), good – though not outstanding – intrinsic cat-
alytic activity, good stability and low cost. For these reasons, these
materials have been extensively investigated [1]. Most attention
has been devoted to Raney-Ni, obtained by leaching a non-noble
metal (Zn or Al in most cases) from variously prepared alloy or
composite precursors [1–15]. Owing to the multiplicity of the
preparation methods, variable overpotentials were reported for
hydrogen evolution at Raney-Ni electrodes. In warm concentrated
alkaline solutions, at a current density of 100 mA cm−2, overpoten-
tials were reported to be as low as 80–100 mV [2,9]. Similar values
were reported also in 1 M NaOH [6], but under these conditions
much larger values were also measured [7,8,10]. Materials compa-
rable to Raney-Ni were also prepared by sintering Ni particles [16],
but they were found to be less active as cathodes than as anodes
in alkaline water electrolysis. So-called macroporous Ni electrodes
were produced by cathodic electrodeposition, at large current den-
sities, using aqueous solutions containing NiCl2 and NH4Cl [17,18].
varied extensively, from some nm to some ␮m.
Research efforts were also devoted to further activating Ni
electrodes of high specific surface, in a number of ways. Marked
reduction of the hydrogen evolution overpotential was achieved by
adsorption of Cd2+ cations which blocked the sites more strongly
adsorbing hydrogen atoms, thus favouring the electrochemical des-
orption step [19], by thermal deposition of RuO2 into the Raney-Ni
micropores [20] and by cathodic co-deposition of RuO2 particles
with a porous Ni matrix [21]. A recent paper [22] describes the
activation of bulk Ni, by spontaneous deposition of Ru, taking place
upon immersion of Ni electrodes, at open circuit, in acid deaerated
aqueous solutions of Ru(III) chloride complexes. In such a process,
Ni surface atoms are oxidized to Ni2+ ions, while Ru(III) complexes
are reduced to metallic Ru (and H+ is reduced to molecular hydro-
gen). Etching in boiling 20% HCl is recommended as a pre-treatment
of Ni supports and reported to produce a surface roughness of 1.7
(assuming to be 1 the surface roughness of Ni mirror finished by
means of alumina). The spontaneous deposition of Ru on Ni had
been previously exploited by Dumont et al. [23] as a step in the
preparation of electrodes for water electrolysis. In their work, Ru
was deposited on spiky filamentary Ni powder, then used for the
preparation of lanthanum phosphate-bonded composite materials.
In the present paperwedescribethe phenomena observed when
porous Ni electrodes prepared by electrodeposition, according to
∗
Corresponding author.
E-mail address: m.musiani@ieni.cnr.it (M. Musiani).
0013-4686/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2008.06.056

L. Vázquez-Gómez et al. / Electrochimica Acta 53 (2008) 8310–8318
8311
Refs. [17,18], and having an estimated surface roughness in the
range 10–50, were submitted to spontaneous deposition of either
Ru, according to Ref. [22], or Ir from H2IrCl6 solutions. The modified
porous Ni electrodes were characterized by SEM-EDX, electrochem-
ical impedance spectroscopy and cyclic voltammetry, and their
activity towards hydrogen evolution was assessed by recording
steady-state current potential curves, from which exchange cur-
rents, Tafel plots and overpotentials were evaluated.
As-deposited porous Ni, Ru-modified porous Ni and Ir-modified
porous Ni electrodes were submitted to the same sequence of
electrochemical tests, in 1 M NaOH aqueous solutions (not deaer-
ated and kept at 25 ◦C): (i) at first a steady-state current potential
curve was recorded from −0.5 to −1.4 V and back, to study the
2. Experimental
2.1. Equipment and materials
Porous Ni layers were deposited onto Ni rotating disc electrodes
(RDE) of 0.34 cm2. All current densities measured in electrochem-
ical experiments and henceforth reported refer to this geometric
area. Prior to use, these electrodes were polished with emery paper
1000, rinsed with water and dried in air. Ni electrodeposition was
performed under constant current conditions, in a cell consisting
of a main central cylindrical compartment, were the RDE was posi-
tioned, connected through glass frits to two lateral compartments
where two Ni wire counterelectrodes (overall area 2 cm2) were
placed. The working electrode was rotated at 2500 revolutions per
minute, so as to remove hydrogen bubbles formed in a side reaction.
A two-compartment cell was used in voltammetric and EIS
experiments and for recording steady-state current–potential
curves. The working electrode was positioned in the main compart-
ment, in such a way that the electrode/electrolyte interface was on
a vertical plane, so as to allow, when necessary, the free departure
of the hydrogen bubbles produced. A Pt wire in the same compart-
ment was used as counterelectrode. An Hg/HgO/5 M KOH electrode,
used as reference, was placed in the lateral compartment connected
to the main one through a Luggin capillary.
The electrochemical equipment consisted of an AMEL elec-
trochemical apparatus (Model 2053 potentiostat) used in Ni
deposition, and of an Autolab PGSTAT 100, used in the other elec-
trochemical experiments.
SEM analyses were performed with a Fei-Esem FEI Quanta 200
FEG instrument, equipped with a field emission gun, operating in
high vacuum condition at an accelerating voltage variable from 5 to
30 keV, depending on the observation needs. EDX analyses and X-
ray mapping (elemental mapping) were obtained by using an EDAX
Genesis energy-dispersive X-ray spectrometer at an accelerating
voltage of 25 keV.
2.2. Procedures
Porous Ni layers were deposited from 0.2 M NiCl2, 2 M NH4Cl
aqueous solutions, pH 4.5, kept at 25 ◦C [17,18], typically at a cur-
rent density of −1.0 A cm−2; their thickness was controlled through
the deposition charge density which, unless otherwise specified,
was fixed at 60 C cm−2. Taking into account the current efficiency
of Ni deposition [21], and neglecting the pores volume the Ni layer
thickness can be estimated at 20 ␮m. Once removed from the depo-
sition bath, the Ni deposits were carefully rinsed with water, dried
and transferred to the solutions containing noble metal complexes.
The spontaneous deposition of Ru was carried out according
to Bianchi et al. [22]: the porous Ni electrodes were immersed, at
open circuit in a 10−3 M Ru(III) chloride complex solution. A similar
procedure was followed for depositing Ir; in this case a deaerated
10−3 M H2IrCl6 solution was used. Both solutions, kept at 25 ◦C by
means of a thermostat, were thoroughly purged with an N2 flow
and their pH was 1. As recommended in Ref. [22], the Ru(III) solu-
tion was used after aging for a week. Such a precaution was not
necessary for the H2IrCl6 solution.
Fig. 1. SEM images of porous Ni deposits as-deposited (A) or after a 2-h immersion
in Ru(III) (B) or Ir(IV) (C) complex solutions.

8312
L. Vázquez-Gómez et al. / Electrochimica Acta 53 (2008) 8310–8318
10 mV s−1, to study the Ni(II)/Ni(III) reaction taking place at the
electrode surface [24,25]; (iv) a new steady-state current poten-
tial curve was recorded under the same conditions as in step (i).
The cyclic voltammograms shown below are always the last of each
series, corresponding to a reproducible behaviour. Tafel plots were
obtained from the steady-state current potential curves, after cor-
rection of the ohmic drop; the value of the electrolyte resistance
used for the correction was determined as the high frequency limit
of the impedance measured with the same electrode, electrolyte
and cell geometry.
hydrogen evolution reaction; (ii) the electrode was polarized at
−1.0 V for 10 min and three successive cyclic voltammograms were
recorded between −1.0 and −0.7 V at 10 mV s−1, to study hydro-
gen underpotential deposition and desorption; (iii) five successive
cyclic voltammograms were recorded between 0.2 and 0.6 V at
EIS experiments aimed at obtaining the electrode capacity were
carried out in 1 M NaNO3 solution thermostated at 25 ◦C, at the
open circuit potential, in order to be as close as possible to blocking
electrode conditions. The frequency range 100 mHz to 100 kHz was
covered with 8 points per decade. The amplitude of the modulated
potential was 10 mV.
3. Results and discussion
3.1. Electrode preparation—SEM-EDX characterization
Freshly deposited porous Ni layers were immersed in either
Ru(III) or Ir(IV) chloride complex solutions for durations varying
between 30 min and 18 h. The noble metal deposition process was
carried out in deaerated solutions since, according to Ref. [22], O2
and noble metal ions reductions competed with each other and
Fig. 2. SEM images of porous Ni deposits after immersion in Ru(III) (A) or Ir(IV) (B
Fig. 3. Comparison of the SEM image of a porous Ni deposits after an 18-h immersion
and C) solutions; immersion time 2 h (A and B) or 18 h (C).
in Ir(IV) solution (A) with the map obtained with the IrM radiation (B).

L. Vázquez-Gómez et al. / Electrochimica Acta 53 (2008) 8310–8318
8313
the presence of air severely hindered metal deposition. Many sam-
ples could be treated in the same solution, over several weeks,
without appreciable change in the electrochemical behaviour of
the resulting electrodes. A change in colour of the H2IrCl6 solution,
which became paler, was a warning that its effectiveness was being
lost.
factors; the darker areas in Fig. 3A are the recessed ones where
one can detect the marks due to the prolonged residence of hydro-
gen bubbles during porous Ni deposition [26]. The same pattern of
bright and dark areas can be recognized also in the Ir map, since the
X-rays emitted by protruding areas are more efficiently collected
than those emitted by recessed ones. The close similarity of both
images indicates that Ir deposits are fairly homogeneously present
on the whole electrode, and not preferentially localized on specific
sites, for instance on the dendrites’ top. Similar results, not shown,
were obtained by mapping Ru-modified electrodes.
Fig. 1 compares the SEM images obtained with an as-deposited
porous Ni layer (A), and with Ni layers modified by spontaneous
deposition of either Ru (B) or Ir (C), for 2 h. The unmodified Ni
deposit consists of an assembly of closely packed micro-dendrites,
with a typical lateral dimension of 2–5 ␮m, giving rise to irregu-
larly shaped pores of variable width, which according to previous
papers [21,26] extend throughout the whole Ni layer thickness.
Upon immersion in the noble metal solutions the deposit mor-
phology undergoes significant changes. The porous structure of
the Ni layer is retained, but the spacing between neighbouring
dendrites increases, as shown by the enhanced dark areas, espe-
cially in Fig. 1C. There is no indication that Ni dissolution and noble
metal deposition produce a smoothing of the deposits. A visible
increase in fine structure of the dendrites takes place, which is more
easily appreciated in Fig. 2A and B (four times larger magnifica-
tion than Fig. 1). These figures also highlight the different aspect
of Ru-modified and Ir-modified electrodes. In the former there is
no obvious evidence of deposited Ru crystals; in the latter the Ni
dendrites look fluffy, probably due to the deposition onto their sur-
face of very small Ir crystals. The changes induced by Ir deposition
become stronger after an 18-h treatment (Fig. 2C): the size of Ir
crystallites somewhat increases and the initial dendrite shape is
more deeply altered.
Fig. 4 compares the EDX spectra of electrodes obtained from
porous Ni layers after different times of immersion in either Ru(III)
(A) or Ir(IV) (B) solutions. The spectra have been normalized with
respect to the Ni K␣ peak (7.49 keV) and shifted along the ver-
tical axis. Fig. 4A shows that the intensity of the RuL and RuM
peaks (2.57 and 0.25 keV, respectively) progressively increases as
the immersion time increases from 30 min to 6 h, without further
significant increase for longer immersion times. Fig. 4B shows that
the peaks due to Ir (IrL and IrM peaks at 2.00 and 9.16 keV, respec-
tively) are hardly detectable after 30 min, progressively increase in
intensity for longer immersion times, up to 18 h, but never become
as intense as the Ru peaks. One can conclude that the deposition
of Ir is much slower than that of Ru, and that a limiting situa-
tion is not attained for Ir, in the explored immersion time range.
Interestingly, the intensity of the peak at 0.86 keV, (NiL) is not inde-
pendent of immersion time, i.e. proportional to that of the Ni K␣
peak (as it would be if the samples had a random distribution of
Ni and noble metal atoms) but has a dependence on immersion
time opposite to that of the Ru and Ir peaks. This dependence is
in agreement with the formation of the noble metal deposits on
top of the Ni layer. With such a geometry, the low energy radiation
emitted by underlying Ni is more strongly absorbed by the more
homogeneous and thicker Ru and Ir layers formed during longer
treatments. The attenuation of the NiL peak is stronger for Ru than
for Ir, suggesting a more extensive deposition of the former noble
metal.
In Fig. 3, a SEM image of an Ir-modified electrode (A), recorded
with a low magnification, is compared with a map of the same
portion of the electrode obtained by selectively detecting the IrM
radiation (B). The presence of Ir is witnessed by the light spots.
Although the SEM image obtained with backscattered electrons is
sensitive to elemental composition, differences in brightness of dif-
ferent areas of very rough samples are mainly due to geometric
Fig. 4. EDX spectra of porous Ni deposits after immersion in Ru(III) (A) or Ir(IV) (B) solutions for different durations indicated in the figure. Spectra are normalized with
respect to the Ni K␣ radiation, and shifted with respect to the intensity axis.

8314
L. Vázquez-Gómez et al. / Electrochimica Acta 53 (2008) 8310–8318
Fig. 5. Cyclic voltammetries (in the potential range of hydrogen underpotential
deposition and desorption) recorded at as-deposited, Ru-modified and Ir-modified
porous Ni electrodes, in 1 M NaOH. Scan rate: 10 mV s−1
.
Although the EDX spectra in Fig. 4 cannot be exploited to obtain
a quantitative measure of the amount of noble metals deposited
onto porous Ni, owing to the fact that Ni and noble metals are not
present as an alloy but as overlapping layers, they provide some
interesting, though qualitative hints on the kinetics of Ru and Ir
spontaneous deposition.
Fig. 6. Cyclic voltammetries (in the potential range of the Ni(II)/Ni(III) couple)
recorded at Ru-modified (A) and Ir-modified (B) porous Ni electrodes, in 1 M NaOH.
Scan rate: 10 mV s−1
.
Fig. 6 shows cyclic voltammetries recorded in the potential range
where the oxidation of Ni(OH)2 to NiOOH and the reverse reduc-
tion take place [24,25], for Ru-modified electrodes (part A) and
Ir-modified electrodes (part B). The voltammetry obtained with an
as-deposited Ni layer is shown for comparison in both parts. Immer-
sion times of 30 min and 2 h in Ru(III) solution induce a progressive
and marked increase in the peak heights; after a 6-h immersion no
further increase in the peak size is seen, and the voltammetry is dis-
torted (Fig. 6A). Table 1 shows that the oxidation/reduction charges
of the NiOOH/Ni(OH)2 system decrease from 2- to 6-h treatment.
Instead, the size of the voltammetric peaks monotonically increases
3.2. Electrochemical characterization
3.2.1. Cyclic voltammetry
Fig. 5 shows cyclic voltammograms recorded at as-deposited
and modified porous Ni electrodes, previously used as cathodes
for hydrogen evolution, in the potential range where the hydrogen
underpotential deposition and its desorption are expected to take
place on noble metals. Very little current is observed at porous Ni,
and no sign of the onset of hydrogen evolution. At the Ru-modified
electrodes, broad peaks are detected, not as well-defined as those
shown by Bianchi et al. [22] who reported that the charge associated
to hydrogen UPD/desorption could be used to quantify the extent
of Ru deposition onto bulk Ni electrodes. Comparable peak current
densities were measured in our work and in Ref. [22] for compara-
ble times of immersion in Ru(III) solution, although the sweep rates
were 10 and 100 mV s−1, respectively. This result suggests the for-
mation of Ru deposits of very large surface area, as a consequence
of the large roughness of the Ni deposits. The voltammetry relevant
to the Ir-modified electrode is featureless, as compared to that of
the Ru-modified electrode and to results in literature [27], but the
current densities measured at Ru- and Ir-modified electrodes are
very similar. Owing to the difficulties in measuring unambiguously
the redox charges associated with hydrogen UPD/desorption on our
electrode materials, we did not attempt to derive information on
noble metal surface area from these voltammetries.
Table 1
Oxidation/reduction charge of the surface NiOOH/NiO couple measured at as-
deposited and noble metal modified porous Ni electrodes in 1 M NaOH
Treatment of porous Ni
Anodic charge
Cathodic charge
(mC cm−2
)
(mC cm−2
)
Noble metal ion
Immersion time
(min)
None
Ru(III)
Ru(III)
Ru(III)
Ir(IV)
Ir(IV)
Ir(IV)
Nonea
0
30
120
360
30
120
360
360
7.24
15.7
44.6
24.9
12.4
21.7
53.5
7.66
7.48
19.0
64.7
34.9
13.3
25.5
66.3
8.17
a
Porous Ni was exposed to aerated 0.1 M HCl.

L. Vázquez-Gómez et al. / Electrochimica Acta 53 (2008) 8310–8318
8315
Fig. 7. Nyquist plots of the impedance of as-deposited (left), Ru-modified (centre) and Ir-modified (right) porous Ni electrodes, recorded at the open circuit potential in 1 M
NaNO3. Immersion time in either Ru(III) or Ir(IV) solutions: 2 h.
with the immersion time in Ir(IV) solution (Fig. 6B), as do the oxi-
dation/reduction charges reported in Table 1. For immersion times
up to 2 h, the peak height is larger for the Ru-modified than for
Ir-modified electrodes: the increase in oxidation/reduction charge
measured after a 2-h treatment is of a factor of about 6 for Ru and
about 3 for Ir.
The observed increase in the peak size of the NiOOH/Ni(OH)2
system testifies that the noble metal induced dissolution of Ni pro-
duces a marked increase in the deposit roughness, as SEM images
suggested. It is also interesting that noble metal deposition does
not inhibit the NiOOH/Ni(OH)2 interconversion, suggesting that
noble metal deposits are not homogeneously covering the porous
Ni surface. Indeed, a continuous film should prevent both the redox
reactions of the Ni surface and further Ni oxidation/dissolution. The
decline in the reduction charge observed after a 6-h treatment in
Ru(III) solution, which can be related to the EDX data in Fig. 4A
showing that Ru deposition had attained, after 6 h, a limiting value,
is probably the result of an extensive, though not complete surface
blocking.
results agree with the indication provided by the SEM images in
Fig. 2 that the porous structure of Ni deposits is maintained upon
immersion in noble metal ions solutions. What is markedly changed
by the immersion treatments is the frequency range where the
quasi-vertical line is observed: deposition of either Ru or Ir shifts to
lower frequency the capacitive response. Such a result is explained
by an increase in the electrodes effective areas, and therefore in
their capacity. Table 2 collects capacity values measured for elec-
trodes submitted to immersion treatments of variable durations,
estimated from the low frequency impedance, where the response
of porous electrodes becomes identical to that of planar electrodes
with an area equal to the developed inner area of the pores [29]. The
capacity values were determined in two ways: (i) a rough estimate
was made by assuming ideal capacitive behaviour and calculating
the capacity at a single frequency according to
−1
C =
(1)
jω Im(Z)
(the arbitrarily selected frequency was one corresponding to an
imaginary part of the impedance close to 10 ꢀ cm2, and thus varied
from sample to sample); (ii) in a more rigorous approach, recog-
nizing that the low frequency line suggested a CPE behaviour, the
parameters Q and ˛ which appear in the CPE impedance
The effect of immersion in acid solutions, pH 1, devoid of
noble metal ions on the morphology of porous Ni deposits was
checked. After 6 h, no sizeable change in the size and shape of the
NiOOH/Ni(OH)2 voltammetric system was detected, and the redox
charges were of the same order as those of as-deposited Ni (Table 1).
This result was obtained using aerated solutions in which oxygen
might have acted as oxidant, instead of noble metal ions.
1
ZCPE
=
(2)
Q(jω)˛
3.2.2. Impedance
Table 2
Capacity of as-deposited and noble metal modified porous Ni electrodes, measured
in NaNO3 at open circuit potential
Fig. 7 compares the impedance of an as-deposited porous
Ni electrode to those of Ru-modified and Ir-modified electrodes,
recorded at the open circuit potential in 1 M NaNO3. The open cir-
cuit potential was always in the range −0.1 to −0.2 V, i.e. in a domain
where neither hydrogen UPD/desorption nor the NiOOH/Ni(OH)2
system are involved. Thus, the electrode capacity measured at the
open circuit potential is expected to correspond to double layer
capacity, free from contributions by redox pseudo-capacities.
All plots in Fig. 7 consist of two linear regions: the one at high
frequency forms a 45◦ angle with the real axis, the one at low fre-
quency is a quasi-vertical line. These plots conform to the expected
response of porous blocking electrodes [26,28–30], and the fact
that the low frequency line forms an angle lower than 90◦ can be
explainedby adistributionin porediameters [31]. Theseimpedance
Treatment of porous Ni
Capacitya
Capacityb
(␮F cm−2
)
(␮F cm−2
)
Noble metal ion
Immersion time (min)
None
Ru(III)
Ru(III)
Ru(III)
Ir(IV)
Ir(IV)
Ir(IV)
Ir(IV)
–
30
120
360
30
120
360
1080
400
1,930
7,150
9,600
970
1,810
4,200
14,700
355
1,840
6,560
8,830
925
1,740
4,280
16,000
a
Measured at a single frequency.
Determined according to the formula proposed by Brug et al. [33].
b

8316
L. Vázquez-Gómez et al. / Electrochimica Acta 53 (2008) 8310–8318
were estimated using the graphical method proposed by Orazem
et al. [32] as
d log[−Im(Z)]
˛ =
(3)
d log(f )
and
ꢀ
ꢁ
˛ꢁ
−1
Q = sin
(4)
Im(Z)(2ꢁf )˛
2
and then used in the calculation of the impedance according to the
formula proposed by Brug et al. [33]
1−˛
Q = C˛(Re−1
)
(5)
Both sets of capacity values were reported in Table 2; for each
sample their difference is never larger that 15%. Therefore, the dis-
cussion of these data, aimed at elucidating the marked effect of the
treatment of porous Ni with noble metal ion solutions on electrode
capacity, may be based on either set.
Clearly, upon prolonging the immersion, the capacity of the elec-
trodes increases and the factor by which the initial capacity of
porous Ni is multiplied is definitely larger than the corresponding
factor relevant to redox charges, i.e. (Ct/C0) > (qt/q0), where Ct and qt
represent capacity and redox charge after a treatment time t, and C0
and q0 represent capacity and redox charge for as-deposited porous
Ni. At the longest immersion time the capacity of the Ru-modified
electrode keeps increasing, though slowly, whereas its redox charge
declines. A quantitative correspondence between the evolution of
the redox charge and of the electrode capacity with the duration of
the immersion in noble metal ion solutions should not be expected,
because the former is only sensitive to the area of Ni in contact with
the electrolyte, while the latter is sensitive to the whole electrode
area (Ni and noble metals).
An overall analysis of EDX, voltammetric and impedance results
suggests the following general picture: (i) in the initial stages, under
the oxidizing effect of noble metal ions, Ni undergoes dissolution
with increase of its surface area; (ii) noble metals are simultane-
ously deposited, forming crystals of very small size which further
contribute to enhance the electrode roughness; (iii) as the noble
metal deposition continues, the Ni area accessible to the electrolyte
decreases, and the deposition process slows down and eventually
stops; (iv) the time scale of these events is shorter for Ru than for
Ir, possibly due to different morphology/crystallinity of the noble
metal deposits.
Fig. 8. Tafel plots of the current–potential curves recoded at Ru-modified (A) and
Ir-modified (B) porous Ni electrodes, in 1 M NaOH, for different immersion times in
noble metal ion solutions. As-deposited porous Ni is shown for comparison.
(Ir-modified electrodes) shows a monotonic performance improve-
ment with immersion time.
Table 3 collects the numerical values of exchange current,
Tafel slope and overpotential (at an arbitrary current density of
100 mA cm−2) obtained from the analysis of the data in Fig. 8.
When two linear regions may be detected (porous Ni) only the
data corresponding to the better defined one are reported. In the
case of Ir-modified porous Ni, the (ohmic drop corrected) poten-
tial range is fairly narrow and thus the values in Table 3 are roughly
approximated. However, differences between the various electrode
materials are large enough to allow a sound discussion of their
activity in the hydrogen evolution reaction. The data in Fig. 8 and
Table 3 suggest that 2 h is a convenient duration of the immersion
treatment for both noble metals, as most of the improvement in
electrocatalytic activity is already achieved in this time. Further
minor progress can be obtained in the case of Ir-modified elec-
trodes by prolonging the immersion in Ir(IV) solutions, whereas an
extension of the treatment is detrimental to the performance of
Ru-modified electrodes. Clearly, Ir-modified electrodes are supe-
rior to Ru-modified electrodes; since exchange currents are of the
same order, the difference is due to the markedly different Tafel
slopes.
3.2.3. Hydrogen evolution—steady-state current–potential curves
Tafel plots for the hydrogen evolution reaction were recorded
for pristine and modified porous Ni electrodes, both immediately
after deposition or immersion treatments and after submitting the
electrodes to the cyclic voltammetries described above. With an
exception mentioned below, the electrodes became definitely more
active as cathodes after the voltammetric study; in the potential
range explored in this study both the Ni oxidation, already reported
by Lasia and Rami [34] to lead to Ni activation, and the growth
of hydrous surface oxides on both Ru and Ir [27,35–37] should
be expected. Only the Tafel plots obtained after the voltammetric
study are shown in Fig. 8.
Fig. 8A (Ru-modified electrodes) shows that immersion treat-
ments lasting 30 min or 2 h led to a marked progressive increase of
the electrocatalytic activity, with an evident reduction of overpo-
tential and Tafel slope, with respect to as-deposited Ni. By further
prolonging the immersion in Ru(III) the electrode performance
became very unstable (being better before than after the voltam-
metric study) and much worse than those achieved with shorter
treatments. Inspection of electrodes modified by a 6-h treatment
showed significant loss of deposit at the end of the test. Fig. 8B
A comparison of the charge and capacity data (Tables 1 and 2,
respectively) with the exchange current data (Table 3), and of
their dependence on the duration of the treatment in noble metal
ion solutions, allows to estimate to what extent the increase in
effective electrode area is responsible for the improved cathode

L. Vázquez-Gómez et al. / Electrochimica Acta 53 (2008) 8310–8318
8317
Table 3
Kinetic parameters for the hydrogen evolution reaction measured at as-deposited and noble metal modified porous Ni electrodes in 1 M NaOH (Ni deposition charge 60 C cm−2
)
Treatment of porous Ni
Noble metal ion
Kinetic parameters
Immersion time (min)
Exchange current (mA cm−2
)
Tafel slope (mV)
Overpotential at 100 mA cm−2 (mV)
None
Ru(III)
Ru(III)
Ru(III)
Ir(IV)
Ir(IV)
Ir(IV)
0
30
120
360
30
0.32
8.5
10.7
2.23
12.2
10
138
118
87
166
125
50
298
126
84
274
114
50
120
360
7.2
42
48
Table 4
Effect of the thickness of porous Ni deposit on the kinetic parameters for the hydrogen evolution reaction in 1 M NaOH (treatment duration 2 h)
Modified porous Ni electrode
Kinetic parameters
Noble metal ion
Ni deposition charge (C cm−2
)
Exchange current (mA cm−2
)
Tafel slope (mV)
Overpotential at 100 mA cm−2 (mV)
Ru(III)
Ru(III)
Ir(IV)
Ir(IV)
30
120
30
6.8
6.1
8.7
8
105
80
90
119
87
89
57
120
52
performance. If one considers the electrodes obtained with 2-h
immersion treatments, the exchange current increases by a factor
of about 30, with respect to as-deposited Ni, for both noble met-
als. For Ir-modified electrodes, redox charge and capacity increase
by factors of about 3 and 4.5, respectively; for Ru-modified elec-
trodes, the same factors are about 7 and 18, respectively. These
data suggest that, for both noble metals the geometric effects are
not negligible. However, the change in Tafel slopes is a final evi-
dence that a true electrocatalytic effect is induced by noble metal
deposition. Furthermore, it should be considered that, under gas
evolution condition, part of the electrode area which is active in
voltammetric and impedance experiments might become inac-
tive.
For the sake of comparison, Ru- and Ir-modified electrodes were
prepared by immersing in the appropriate noble metal ion solu-
tions quite smooth Ni layers obtained from the usual NiCl2, NH4Cl
bath [17,18] at low current density (0.1 A cm−2) [26]. Even in the
most favourable case, i.e. Ir-modified electrode tested after the
voltammetric study, their performance were not good: exchange
current was 3.3 mA cm−2, Tafel slope 180 mV and overpotential at
100 mA cm−2 280 mV. These tests further showed that the use of
Ni deposits of a high surface roughness was beneficial to achieve
active cathodes.
4. Conclusions
Two procedures described in the literature (cathodic deposition
of porous Ni layers [17,18] and spontaneous deposition of ruthe-
nium [22]) have been combined to obtain cathodes for hydrogen
evolution. The same approach has been extended to spontaneous
deposition of iridium, and found to lead to cathodes of higher cat-
alytic activity than that achieved with ruthenium. The advantages
of porous Ni layers, as compared to bulk Ni electrodes are (i) their
large effective area and (ii) their strong reactivity with Ru(III) or
Ir(IV) solutions, which allows noble metal deposition without Ni
etching in hot HCl solutions, described in Ref. [22] as a necessary
activation procedure. The noble metal deposition is paralleled by
an increase in the electrode effective area caused by both the fur-
ther roughening of Ni (as testified by the cyclic voltammograms
recorded in the Ni(II)/Ni(III) potential region) and the deposition
of noble metal crystallites of very low dimension, as clearly shown
by SEM pictures of Ir-modified electrodes. Significant differences
exist in the kinetics of Ru and Ir deposition: under the adopted
experimental conditions the former forms thicker deposits than
the latter, on a shorter time scale, up to reach a limiting situation
(not observed for iridium with deposition times up to 18 h). The
catalytic activity of the noble metal modified electrodes generally
improves upon cycling at potentials likely to cause, besides the oxi-
dation of Ni [34], the growth of hydrous surface oxides on both Ru
and Ir [27,35–37].
3.2.4. Effect of Ni layer thickness
All the results reported so far refer to Ni deposits prepared with
a deposition charge of 60 C cm−2. In order to test the effect of the Ni
layer thickness, other electrodes were prepared with Ni deposition
charges of either 30 or 120 C cm−2 and a constant 2-h immersion
time in the noble metal ion solutions; these electrodes were then
submitted to the same sequence of electrochemical tests. For both
noble metals, the voltammetric current measured in the poten-
tial range of the hydrogen UPD/desorption increased with the Ni
deposition charge in a roughly linear way. The redox charge of
the Ni(II)/Ni(III) system measured at Ir-modified electrodes was
also proportional to the Ni deposition charge, whereas that of
Ru-modified electrodes increased only slightly with the Ni layer
thickness. The kinetic parameters measured with electrodes of vari-
able thickness are shown in Table 4. Comparison of Tables 3 and 4
shows that increasing the Ni deposition charge from 30 to 60 C cm−2
induces a significant improvement in the cathodes performance,
whereas a further increase to 120 C cm−2 is ineffective.
Acknowledgments
The authors are indebted to FILA INDUSTRIA CHIMICA SPA, San
Martino di Lupari, Padova, Italy, owner of the Fei-ESem FEI Quanta
200 FEG instrument, for allowing its use for the research work
described in this article.
References
[1] S. Trasatti, in: H. Gerischer, C.W. Tobias (Eds.), Advances in Electrochemical
Science and Engineering, vol. 2, VCH, Weinheim, 1992, p. 1.
[2] E. Endoh, H. Otouma, T. Marimoto, Y. Oda, Int. J. Hydrogen Energy 12 (1987) 473.
[3] Y. Ogata, H. Hori, M. Yasuda, F. Hine, J. Electrochem. Soc. 135 (1988) 76.
[4] T. Kenjo, Electrochim. Acta 33 (1988) 41.
[5] H. Wendt, H. Hofmann, W. Plzak, Mater. Chem. Phys. 22 (1989) 27.
[6] Y. Choquette, L. Brossard, A. Lasia, H. Menard, J. Electrochem. Soc. 137 (1990)
1723.

8318
L. Vázquez-Gómez et al. / Electrochimica Acta 53 (2008) 8310–8318
[7] L. Chen, A. Lasia, J. Electrochem. Soc. 138 (1991) 3321.
[8] L. Chen, A. Lasia, J. Electrochem. Soc. 139 (1992) 3214.
[9] M. Okido, J.K. Depo, G.A. Captano, J. Electrochem. Soc. 140 (1993) 127.
[10] A.K. Cheong, A. Lasia, J. Lessard, J. Electrochem. Soc. 140 (1993) 2721.
[11] S.-I. Tanaka, N. Hirose, T. Tanaki, J. Electrochem. Soc. 146 (1999) 2477.
[12] S.-I. Tanaka, N. Hirose, T. Tanaki, Y.H. Ogata, J. Electrochem. Soc. 147 (2000)
2242.
[13] C. Hitz, A. Lasia, J. Electroanal. Chem. 500 (2001) 213.
[14] G. Sheela, M. Pushpavanam, S. Pushpavanam, Int. J. Hydrogen Energy 27 (2002)
627.
[24] M.A. Sattar, B.E. Conway, Electrochim. Acta 14 (1969) 695.
[25] R.S. Schrebler Guzman, J.R. Vilche, A.J. Arvia, J. Electrochem. Soc. 125 (1978)
1578.
[26] F. Huet, M. Musiani, R.P. Nogueira, J. Solid State Electrochem. 8 (2004) 786.
[27] J. Mozota, B.E. Conway, Electrochim. Acta 28 (1983) 1.
[28] R. de Levie, in: P. Delahay (Ed.), Advances in Electrochemistry and Electrochem-
ical Engineering, vol. VI, Interscience, New York, 1967, p. 329.
[29] J.-P. Candy, P. Fouilloux, M. Keddam, H. Takenouti, Electrochim. Acta 26 (1981)
1029.
[30] J.-P. Candy, P. Fouilloux, M. Keddam, H. Takenouti, Electrochim. Acta 27 (1982)
1585.
[31] H.-K. Song, Y.-H. Jung, K.H. Lee, L.H. Dao, Electrochim. Acta 44 (1999) 3513.
[32] M.E. Orazem, N. Pébère, B. Tribollet, J. Electrochem. Soc. 153 (2006) B129.
[33] G.J. Brug, A.L.G. Van Den Eeden, M. Sluyters-Rebach, J. Sluyters, J. Electroanal.
Chem. 176 (1984) 275.
[15] I. Birry, A. Lasia, J. Appl. Electrochem. 34 (2004) 735.
[16] D.E. Hall, J. Electrochem. Soc. 128 (1981) 740.
[17] C.A. Marozzi, A.C. Chialvo, Electrochim. Acta 45 (2000) 2111.
[18] C.A. Marozzi, A.C. Chialvo, Electrochim. Acta 46 (2001) 861.
[19] N.V. Korovin, E.Ya. Udris, O.N. Savel’eva, Sov. Electrochem. 23 (1986) 330.
[20] R.K. Shervedani, S.H. Kazemi, A. Lasia, H.R.Z. Mehrjerdi, J. New Mater. Elec-
trochem. Syst. 8 (2005) 213.
[34] A. Lasia, A. Rami, J. Electroanal. Chem. 294 (1980) 123.
[35] B.E. Conway, J. Mozota, Electrochim. Acta 28 (1983) 9.
[36] S. Hadzi-Jordanov, H. Angerstein-Kozlowska, B.E. Conway, J. Electroanal. Chem.
60 (1975) 359.
[21] L. Vázquez-Gómez, S. Cattarin, P. Guerriero, M. Musiani, Electrochim. Acta 52
(2007) 8055.
[22] I. Bianchi, E. Guerrini, S. Trasatti, Chem. Phys. 319 (2005) 192.
[23] H. Dumont, P. Los, A. Lasia, H. Menard, L. Brossard, J. Appl. Electrochem. 23
(1993) 684.
[37] L.D. Burke, J.K. Mulcahy, J. Electroanal. Chem. 73 (1976) 207.