Electrocatalytic activity of metal-loaded reticulated vitreous carbon electrodes for hydrogen evolution from flowing alkaline solutions

Article abstract of DOI:10.1246/bcsj.79.1711

Reticulated vitreous carbon (RVC) electrodes have been used as possible porous cathodes, operating in flow-through mode for the purpose of the electrolytic generation of hydrogen gas from flowing alkaline solutions. The electrocatalytic performance of the RVC cathode was tested under different operating conditions, such as electrolyte flow regimes and temperature. Also, the effect of electroplating the RVC with copper metal or Watts nickel in flowing electrolytes has been assessed. Potentiodynamic and galvanostatic techniques along with SEM images were used to investigate the performance of the different cathodes: bare (uncoated) RVC, copper-coated RVC (RVC/Cu), and Ni-coated RVC (RVC/Ni). The hydrogen evolution reaction (HER) was found to be enhanced by coating RVC either with Cu or Ni. Polarization behavior had higher rates for HER using RVC/Ni than using RVC/Cu or bare RVC. This trend was confirmed by using a planar glassy carbon electrode (GC) coated with Cu and Ni under the same conditions. The activation energies for HER using RVC and RVC/Ni were estimated to be 32 and 14 kJ mol^-1, respectively. Current transient and SEM micrographs showed good electrocatalytic and mechanical stability of the metal loadings. There was no indication of loss of the catalytic properties after operation for 30 h. Model calculations helped to extract kinetic parameters for the HER using different electrodes.

Full text of DOI:10.1246/bcsj.79.1711

ꢀ 2006 The Chemical Society of Japan
Bull. Chem. Soc. Jpn. Vol. 79, No. 11, 1711–1718 (2006) 1711
Electrocatalytic Activity of Metal-Loaded Reticulated Vitreous Carbon
Electrodes for Hydrogen Evolution from Flowing Alkaline Solutions
y
ꢀ;
Mahmoud M. Saleh, Maher H. El-Ankily, Mohamed S. El-Deab, and B. E. El-Anadouli
Department of Chemistry, Faculty of Science, Cairo University, Cairo, Egypt
Received December 7, 2005; E-mail: mahmoudsaleh90@yahoo.com
Reticulated vitreous carbon (RVC) electrodes have been used as possible porous cathodes, operating in flow-
through mode for the purpose of the electrolytic generation of hydrogen gas from flowing alkaline solutions. The elec-
trocatalytic performance of the RVC cathode was tested under different operating conditions, such as electrolyte flow
regimes and temperature. Also, the effect of electroplating the RVC with copper metal or Watts nickel in flowing elec-
trolytes has been assessed. Potentiodynamic and galvanostatic techniques along with SEM images were used to inves-
tigate the performance of the different cathodes: bare (uncoated) RVC, copper-coated RVC (RVC/Cu), and Ni-coated
RVC (RVC/Ni). The hydrogen evolution reaction (HER) was found to be enhanced by coating RVC either with Cu
or Ni. Polarization behavior had higher rates for HER using RVC/Ni than using RVC/Cu or bare RVC. This trend
was confirmed by using a planar glassy carbon electrode (GC) coated with Cu and Ni under the same conditions.
The activation energies for HER using RVC and RVC/Ni were estimated to be 32 and 14 kJ molꢁ1, respectively. Current
transient and SEM micrographs showed good electrocatalytic and mechanical stability of the metal loadings. There was
no indication of loss of the catalytic properties after operation for 30 h. Model calculations helped to extract kinetic
parameters for the HER using different electrodes.
Three-dimensional electrodes made of reticulated vitreous
carbon (RVC) have been used in many applications, such as
flow reactors.1,2 This system offers several structural and op-
erational advantages including a high area/volume ratio, i.e.,
high specific surface area, high porosity, high conductivity,
and chemical and electrochemical stability. As well, they can
be machined easily and have reasonable electrocatalytic prop-
erties. RVC electrodes may also be useful for the destruction
of organic wastes3,4 and removal of heavy metal ions from
dilute streams.5–7 In order to improve the performance and
efficiency of the electrodes, the RVC is often modified and
in fact, RVC electrodes coated with conductive polymers
showed better electrocatalytic properties towards the removal
of metal ions from dilute streams.8
The production of pure hydrogen is important both for
potential industrial and energy storage applications. The devel-
opment of reliable and cost effective technologies for the pro-
duction of pure hydrogen is a challenging issue from electro-
chemical engineering point of view. For instance, for the
H2–O2 operating fuel cells, ultra pure hydrogen gas (as a fuel)
is a prerequisite. For hydrogen production from alkaline solu-
tions, the cathodes that are used must have high stability, high
specific surface area, and good electrocatalytic properties.9–11
Nickel metal is currently used as a cathode for water electro-
lyzers due to its high electrocatalytic activity towards the hy-
drogen evolution reaction (HER). Different procedures were
introduced to prepare nickel electrodes with high specific sur-
face areas. These include sintering micro porous nickel,12 pro-
duction of nanoporous Raney-Nickel,13 and incorporation of
Ni into PolyHIPE polymer (PHP) matrix.14 A nickel-loaded
RVC matrix should be a very good porous cathode for the
HER in flowing electrolytes. Porous electrodes have been used
in hydrogen production15,16 as well as in other industrial and
environmental applications.17–20 Three-phase (solid–liquid–
gas) porous electrodes could be a special category of porous
electrodes in which several operating parameters inherently
control the electrodes overall performance. Of particular im-
portance, the gas bubbles that are evolved accumulate inside
the porous matrix resulting in non-uniform distributions of
potential and current and, hence, in a lower effectiveness
factor of the porous electrode.21 Optimizing the electrolyte
flow rate by other structural and transport parameters plays
an important role in maximizing the performance of three-
phase porous electrodes. Mathematical modeling can help to
simulate the effects of different operating parameters on such
systems.
The present work aims to study the performance of RVC
and metal-coated RVC for electrocatalytic production of
hydrogen gas from flowing alkaline solutions. The effect
of the incorporation of Cu or Watts Ni (through electrode-
position) onto the RVC matrix towards the enhancement of
the electrocatalytic performance of the RVC cathodes to-
wards the HER was investigated. The results are compared
with those of planar glassy carbon electrode (GC) coated
with Cu and Ni in a stationary KOH electrolyte. A math-
ematical model is reviewed and applied to predict the be-
havior of the porous electrode and to extract kinetic param-
eters of HER for different candidate cathodes. The per-
formance and stability of the different cathodes are evalu-
ated.
y Present address: Department of Electronic Chemistry, Tokyo
Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama
226-8502
Published on the web November 9, 2006; doi:10.1246/bcsj.79.1711

1712 Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006)
Hydrogen Evolution on Modified RVC
Table 1. Bath Compositions and Operating Conditions for
Cu and Ni Electroplating on RVC and GC Electrodes
Bath composition
/g Lꢁ1
c.d.
/mA cmꢁ2
Duration
/h
Metal
RVC/Cu
CuSO4
H2SO4
NiCl2
NiSO4
H3BO3
150
50
25
90
100
1
RVC/Ni
(Watts Nickel)
100
1
60
(a)
Fig. 1. Cell arrangement and flow direction.
Experimental
Materials and Methods.
RVC blocks were supplied by
Electrosynthesis Co., Inc. (New York) and were used as received.
They had a specific surface area (S) equals 70 cm2 cmꢁ3 (i.e.,
cmꢁ1) and porosity (ꢀ) of 0.90. They were cut in cylinders to fit
the chamber of the packed bed electrode. The electrode had a
diameter of 1.0 cm and an ohmic resistance < 0:01 ohm. The
counter electrode was made of a platinum screen and was placed
on the down-stream side of the porous electrode (Fig. 1). The ref-
erence electrode was of the type Hg/HgO/1 M KOH (having an
equilibrium potential of 0.098 V vs NHE22). The potential report-
ed here is that at the exit face relative to the down-stream refer-
ence electrode (denoted as E). Current densities are reported on
the basis of the geometrical cross sectional area of the electrode
(0.80 cm2). A glassy carbon electrode (GCE) electrode of diame-
ter 3.0 mm (from BAS Inc, Japan) was used in the planar electrode
measurements. A conventional three-electrode cell was used in
this case. Prior to the measurements, the GCE was polished first
with emery paper (No. 2000), and then with aqueous slurries of
successively finer alumina powder (particle size down to 0.06 mm)
with the help of a polishing microcloth. The bare GCEs were then
sonicated for 10 min in Milli-Q water.
(b)
(c)
An EG&G potentiostat/galvanostat, model 273A controlled by
m352 electrochemical software was used in all measurements.
The solutions were prepared by dissolving the desired weights
of analytical grade KOH in doubly distilled water. The solution
was circulated using a variable speed pump, and the flow rate
was measured using volumetric measurements. The SEM micro-
graphs were taken on a JEOL JSM-20 Scanning Electron Micro-
scope. Further details of the experimental setup could be found
elsewhere.23
Electroplating Procedure. Electroplating of the RVC porous
matrix was performed in situ with copper or Watts nickel using
flowing electroplating solutions. The electrode arrangement was
described above (Fig. 1). The electroplating solutions and condi-
tions are listed in Table 1.24 The electroplating solutions were
circulated for 1 h at a linear flow rate of 1.0 cm sꢁ1. However, be-
cause of the non-uniform distributions of the metal depositions
within the porous matrix, we used RVC block of thickness equals
1.7 cm, and after the electroplating process, the last 0.7 cm at the
back of the electrode was disregarded. The remaining 1.0 cm (near
to the polarized face) had almost uniform distributions of the
metal loadings. It was used as is or cut as required to only
0.5 cm. The problem of non-uniform distributions of metal cur-
Fig. 2. SEM micrographs of the different electrodes a) bare
RVC, b) RVC/Cu, and c) RVC/Ni. The white zones are
the matrix. The arrows in Fig. 2b refer to a thread.
rents and hence metal depositions within porous electrodes is well
documented.25,26 After electroplating, the blocks were rinsed with
flowing water to remove all contaminants from the electrode
chamber and from the blocks. Measurements were performed im-
mediately after the cleaning process. An estimation of the extent
of metal loading onto the RVC was done using Faraday’s law. As-
suming 100% current efficiency, the loading of Cu and Ni was
found to be 95 and 88 mg cmꢁ3, respectively. Figures 2a–2c show

M. M. Saleh et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006) 1713
SEM micrographs of bare RVC, RVC/Cu, and RVC/Ni, respec-
tively. The bright zones are the matrix, and the black zones are the
voids. It is clear from the micrographs that the white zones are
more pronounced in the case of the RVC/Cu or RVC/Ni due to
metal depositions within the porous RVC matrix. The average
thread diameter (see arrows in Fig. 2b) of bare RVC, RVC/Cu,
and RVC/Ni were estimated to be 17.86, 30.95, and 31.55 mm,
respectively. This clearly reflects the effective electrodeposition
of Cu and Ni onto the fibrous texture of the RVC matrix. A similar
electroplating procedure for the planar GC electrode was conduct-
ed but from stationary solution and for a period of 30 s. This was
enough to cover the GC with Cu or Ni.
where JðxÞ is the reaction current per unit reaction volume
(A cmꢁ3). The other variables are listed in the notation section.
Ohm’s law relates the solution current travels through the gas-
electrolyte dispersion filing the pore space with the change in
polarization. Therefore,
dꢂðxÞ
dx
iðxÞ ¼ ꢁðxÞ
:
ð2Þ
The conductivity of the gas-electrolyte dispersion within the
pores, ꢁðxÞ, is a complex function that depends on composition
of the electrolyte, the porosity, and 2. The latter varies with the
distance within the electrode following the variation of iðxÞ
which generates the gas bubbles. Bruggeman’s equation is
used to correlate the conductivity of the gas-electrolyte disper-
sion which fills the pore space with the gas void fraction,27 i.e.,
Mathematical Model
The HER using porous RVC cathodes in flowing alkaline
solutions is a typical representative example of a three-phase
electrocatalytic reactor, in which gases are electrogenerated
within the solid porous matrix of the cathode, initially filled
with electrolyte. Hence, several disadvantages arose including
a decrease in the pore electrolyte conductivity, a decrease
in the effective real surface area of the porous matrix, and
thus, an overall decrease in the performance of the cathode.
In this section, a previously developed mathematical model
is reviewed.21 Electrocatalytic gas evolution is a significant
and complicated phenomenon in electrochemical systems from
both academic and engineering view points. The generation of
gas bubbles in the reaction medium strongly affects the pore
electrolyte resistivity and, hence, on the overall performance
of the system. Therefore, it is necessary to evaluate the effects
of the gas bubbles on the overall polarization behavior of the
electrode and on the current and potential distributions. When
gas is evolved within a porous electrode, the conducting medi-
um is heterogeneous mixture of three phases, i.e., solid porous
electrode, liquid aqueous electrolyte, and gas bubbles. The
geometry of the interfaces and the conducting medium within
the porous electrode is complicated and difficult to predict.
Since the rate of generation of gas bubbles is rather non uni-
form within the porous electrode, the gas void fraction, 2,
and the pore electrolyte conductivity, ꢁ, are also non uniform
within the electrode. The porosity of the packed bed is uniform
throughout the electrode, and therefore, the variation in pore
electrolyte conductivity is related to the variation in the gas
void fraction.
3=2
ꢁðxÞ ¼ ꢁo½ꢀ ꢁ 2ðxÞꢂ
;
ð3Þ
where ꢁo is the bulk electrolyte conductivity (ohmꢁ1 cmꢁ1
)
and 2ðxÞ is the gas void fraction, which is given by19
ꢀiðxÞ
Qꢄ þ iðxÞ
2ðxÞ ¼
;
ð4Þ
where ꢄ is a conversion factor of the current to the volume of
the generated gas. Assuming ideal gas behavior, ꢄ is given by
2PF
ꢄ ¼
;
ð5Þ
RT
where ꢄ equals 7.87 C cmꢁ3 for hydrogen evolution reaction at
standard pressure and temperature.19
The boundary conditions are:
x ¼ 0; i ¼ icell
ð6Þ
x ¼ L; i ¼ 0 dꢂ=dx ¼ 0 ꢁ ¼ ꢁoꢀ1:5 and 2 ¼ 0
A system of four equations (Eqs. 1–4) describes the distri-
butions of four variables, i.e., i, ꢂ, 2, and ꢁ are obtained.
The above system of equations was solved using a numerical
technique developed by Newman.28 The present theoretical
framework was used to predict the polarization behavior of
the porous electrode, to fit the experimental data with the mod-
el predictions and to estimate some kinetic parameters such as
the exchange densities of the HER for different electrodes.
Results and Discussion
Figure 1 shows the arrangement of the porous electrode and
the direction of the electrolyte flow. In developing the model,
the following assumptions were made.21 Butler–Volmer equa-
tion controls the rate of charge-transfer reaction, which in-
volves a one-electron-transfer rate-determining step. The po-
rous electrode is assumed to be inert substrate and does not
anodically dissolve under the present conditions. The potential
gradient within the solid phase of the packed bed can be ne-
glected due to its high conductivity compared to the electro-
lyte. It is characterized by a uniform porosity, ꢀ, and a specific
surface area, S (cm2 cmꢁ3). Finally, ionic migration effects are
negligible.
Electrolyte Flow Rate. The significance of the bubble
formation within the porous matrix was studied by allowing
the bubbles to build up at stationary conditions (Q ¼ 0) at a
known generation rate, and the potential response was record-
ed. This was done by measuring potential–time relation for the
HER using bare RVC with a thickness (L) of 1.0 cm in 2 M
KOH. The current was fixed at 200 mA cmꢁ2, and the potential
measured under stationary conditions, (i.e., Q ¼ 0). As shown
in Fig. 3, the potential increases with time until it reaches a
plateau after about 30 min. The time period depends on the
transport and structural parameters of the system under inves-
tigation. When Q ¼ 1:0 cm sꢁ1 (point ‘‘a’’ in the figure), the
potential decreases sharply from ꢁ2:6 to ꢁ2:25 V with respect
to the reference electrode. When Q ¼ 0, bubbles accumulate in
the pores of the electrode, and they fully occupy the void frac-
tion.29 The experimental potential difference, ꢀEexp between
Butler–Volmer equation21 correlates the solution current
density, iðxÞ with the polarization by
diðxÞ
dx
ꢁio,HS½1 ꢁ expðꢂðxÞ=bÞꢂ
expðꢃꢂðxÞ=bÞ
¼ ꢁJðxÞ ¼
;
ð1Þ

1714 Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006)
Hydrogen Evolution on Modified RVC
-2.7
Table 2. Parameters Used in Production of the Theoretical
Results in Figs. 4 and 5 and in Fitting the Experimental
Data in Fig. 6
-2.6
(a) Flow start
S ¼ 70 cm2 cmꢁ3
L ¼ 1:0 cm
-2.5
-2.4
-2.3
ꢃ ¼ 0:35
½KOHꢂ ¼ 2 M
ꢁo ¼ 0:25 ꢂꢁ1 cmꢁ1
ꢀ ¼ 0:90
io,H ¼ 2 ꢆ 10ꢁ4 A cmꢁ2 (bare RVC)
-2.2
-2.1
-2.0
Flow regime
Stationary regime (no flow)
-2.4
-1
Q = 0.01 cm s
0
1000
2000
3000
4000
5000
-1
-2.0
0.1 cm s
t / s
-1
Fig. 3. Potential–time relations at different flow regimes for
HER on bare RVC at cell current = 0.2 A cmꢁ2 and 25 ꢅC.
1.0 cm s
-1.6
-1
10 cm s
50
45
-1.2
-0.8
40
35
30
25
20
15
10
5
−1
Q = 0.01 cm s
0.1
10
-2.0
-1.6
-1.2
log (i / A cm-2
-0.8
-0.4
)
1.0
Fig. 5. Theoretical results of the i–E relations of HER
on bare RVC at different flow rate, Q. L ¼ 1 cm and
½KOHꢂ ¼ 2 M.
JðxÞ=io,HS and the dimensionless distance, y was equal to
x=L. At lower Q values, i.e., Q ꢄ 0:01 cm sꢁ1, gas bubble for-
mation is significant and results in an increase in 2 and hence,
a decrease in ꢁ. This is in accordance with Eqs. 3 and 4. As
a result, the potential and, hence, the current becomes more
non-uniform.21,23 As Q increases, the reaction current becomes
more uniform. At higher values of Q, the bubbles are swept out
of the porous matrix and, consequently, the bubble effects are
minimum. Therefore, the polarization and current distributions
become more uniform.
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
y = x/L
Fig. 4. Theoretical results of the distributions of the dimen-
sionless reaction current of the HER on bare RVC at dif-
ferent Q. L ¼ 1 cm and ½KOHꢂ ¼ 2 M.
the stationary and the flowing regimes, which are shown in
Fig. 3, at this current density is:
Figure 5 shows the polarization curves simulated from the
model calculations at the same Q used in Fig. 4. As Q increas-
es, the polarization lines shifted to higher currents at the same
values of polarization, and the degree of 2ðxÞ decreases while
ꢁðxÞ increases. Accordingly, the polarization required to obtain
definite reaction rates (current densities) decreases. Also, as
the conductivity increases, the polarization becomes more uni-
form, and a higher utility of the bed was obtained. There are
ꢀEexp ¼ EQ¼0 ꢁ EQ¼1 ꢃ 0:35 V;
ð7Þ
where EQ¼0 is the electrode potential at stationary regime and
EQ¼1 is the potential at Q ¼ 1 cm sꢁ1 and current density =
200 mA cmꢁ2. The ꢀEexp is a function of the system parame-
ters, and mathematical modeling must be used to account for
the effects of gas bubble formation on the behavior of porous
electrode.
It has been shown before that Q has a dramatic effects on
the distributions of 2ðxÞ and ꢁðxÞ.21,23 As the gas void fraction
increases, the pore electrolyte conductivity decreases accord-
ing to Eq. 3. This causes non-uniform polarization and current
distributions within the porous electrode. Figure 4 shows the-
oretical results of the effects of Q on the distributions of the
no differences in polarization between Q ¼ 1:0 and 10 cm sꢁ1
.
Under these conditions, Q ¼ 1:0 cm sꢁ1 is enough to reduce
the effects of the gas bubbles using the present system param-
eters. Similar curves were calculated at KOH concentrations
of 1 and 3 M, and the same results were obtained. The above
results showed that Q ¼ 1:0 cm sꢁ1 is good for operating the
present gas evolving flow-through porous electrode.
ꢁ
dimensionless reaction current, JðyÞ of HER using the param-
Metal Loadings. Metal loading onto RVC were carried
out as described in the experimental section. Figure 6 shows
ꢁ
eters shown in Table 2. The values of JðyÞ were taken as

M. M. Saleh et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006) 1715
-1.8
-1.6
-1.4
-1.2
-1.0
-0.8
RVC
(exp), --- (model)
-2.8
-2.4
-2.0
-1.6
-1.2
-0.8
RVC/Cu (exp), --- (model)
RVC/Ni (exp), --- (model)
bare GC
GC/Cu
GC/Ni
-4
-3
-2
-1
0
log (i / A cm-2
)
Fig. 7. i–E relations for HER in 2 M KOH with planar GC,
GC/Cu, and GC/Ni electrodes at 25 ꢅC.
-0.4
-2.4
-2.0
-1.6
-1.2
-0.8
-0.4
0.0
log (i /A cm-2
)
Fig. 6. i–E relations for HER from 2 M KOH on different
electrodes at Q ¼ 1 cm sꢁ1, L ¼ 0:5 cm at 25 ꢅC. The
dashed lines are the model calculations and the symbols
are the experimental data.
25°C
40°C
60°C
-2.4
-2.0
i–E relations for the HER on bare RVC, RVC/Cu, and RVC/
Ni in 2 M KOH at Q ¼ 1 cm sꢁ1 at 25 ꢅC. The symbols repre-
sent the experimental data, and the lines represent predicted
values. First, the experimental data were fitted with the model
calculations for the HER on bare RVC in 2 M KOH using a
value of ꢃ equals 0.35 and exchange current density io,H equals
2 ꢆ 10ꢁ4 A cmꢁ2. The experimental data of RVC/Cu and
RVC/Ni were then fitted only by using io,H as the fitting pa-
rameter. The rest of parameters used in the model calculations
are the same as those listed in Table 2 but with L ¼ 0:5 cm.
Comparison of the model predictions satisfactorily agrees with
the experimental results at low and moderate rates of gas evo-
lution. However, at higher current densities, the model predicts
lower polarizations. This was attributed to the trapped gas bub-
bles inside the pores of the matrix. Also, the model should be
modified in the future to account for the surface coverage by
gas bubbles. The fitted values of io,H on RVC, RVC/Cu, and
Bare RVC
-1.6
-1.2
RVC/Ni
-0.8
-2.0
-1.6
-1.2
log (i /A cm-2
-0.8
-0.4
0.0
)
Fig. 8. Experimental data of i–E relations for HER in 2 M
KOH with bare RVC (solid lines) and RVC/Ni (dashed
lines) at Q ¼ 1 cm sꢁ1, L ¼ 0:5 cm at different tempera-
tures.
RVC/Ni were 2 ꢆ 10ꢁ4, 4 ꢆ 10ꢁ4, and 1:5 ꢆ 10ꢁ3 A cmꢁ2
,
Cu coatings, which increase the true surface area of the porous
RVC matrix.
respectively. These values are comparable with the literature
values.30 The model and the experimental results conclude that
the potentials required to obtain certain current densities (i.e.,
rates of HER) decrease in the order:
Effects of Temperature. Figure 8 illustrates the effects of
temperature on the current–potential relations of HER in 2 M
KOH with RVC and RVC/Ni. The results of RVC/Ni elec-
trode are introduced only because they are interesting, and
no model studies are introduced here since the effects of tem-
perature on such systems require that the model be modified.
As the temperature increases, the cathodic potential, i.e., the
power consumption, required to attain specific rates of HER
decreases. At a given potential, an increase in temperature
leads to an increase in the current density. The effects of tem-
perature on the i–E relations can be attributed to two primary
factors. The conductivity of the electrolyte and the rate of
charge transfer at the electrode/electrolyte interface increase
with temperature. A secondary factor may arise due to the
effects of temperature on the surface tension and viscosity of
the electrolyte and, hence, on the bubble size, which, in turn,
affects the pore electrolyte resistively. While an increase in
the conductivity of the electrolyte leads to a decrease in the
ohmic potential drop in the pore electrolyte, the increase in
RVC/Ni < RVC/Cu < RVC:
ð8Þ
The above trend was confirmed by using a planar GC electrode
instead of the porous RVC electrode. Data for the HER in non-
flowing 2 M KOH using GC, GC/Cu, and GC/Ni electrodes
were recorded as shown in Fig. 7. Better electrocatalytic activ-
ity toward the HER was obtained with GC/Ni than with GC/
Cu or bare GC. Note that the values of the reversible electrode
potential, EH, in both cases, i.e., RVC (Fig. 6) and GC (Fig. 7)
electrodes, are closer, and both are comparable with the theo-
retical value. The theoretical value of EH can be calculated
from the relation: EH ¼ ꢁ0:0591xpH as ꢁ0:83 V (NHE). The
above results imply that the RVC/Ni and RVC/Cu electrodes
have enhanced electrocatalytic properties for the HER than
that of RVC. These were attributed to the electrocatalytic
properties of Ni and Cu as well as the roughness of Ni and

1716 Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006)
Hydrogen Evolution on Modified RVC
35
30
25
20
15
10
5
2.4
RVC/Cu
RVC/Ni
2.2
RVC/Ni
2.0
1.8
1.6
1.4
1.2
1.0
0.8
Bare RVC
0
-2.2
3.0
3.2
3.4
-2.0
-1.8
-1.6
-1.4
-1.2
-1.0
-0.8
T
−1/10−3 K−1
log (i / A cm-2
)
Fig. 9. Arrhenius plots of the HER with RVC and RVC/Ni.
The same parameters as in Fig. 7 are used.
Fig. 10. Power savings for the HER on both RVC/Cu and
RVC/Ni. The same parameters as in Fig. 6 are used.
temperature leads to an increase in io,H of the hydrogen evolu-
tion reaction and hence to a decrease in the activation polari-
zation. Consequently, at relatively high temperatures the acti-
vation and ohmic polarization associated with the HER are
lower, and thus, power consumption is lower.
The apparent total exchange current density per geometrical
surface area, Io, can be estimated by extrapolating cathodic
Tafel line to the equilibrium potential.31,32 This can be done
with the i–E relations shown in Fig. 8 for the HER with bare
RVC and RVC/Ni. Using this method the values of Io are
estimated for the two electrodes at different temperatures.
The values of Io are only apparent values and may have some
error, since the current and the polarization are not completely
uniform.33,34 The apparent activation energy, Ea for the HER
reaction can then be estimated from the temperature depen-
dence of Io using Arrhenius equation;31,35
HER in 2 M KOH using L ¼ 0:5 cm. The data were drived
from Fig. 6. This was done by subtracting the potentials for
RVC from that of RVC/Cu or RVC/Ni at the same current
densities. The amount of hydrogen gas produced in gram
equivalent is given by Faraday’s law as it=F. Consequently,
the energy savings, !, at the cathode in kW h per kilogram
of hydrogen gas is given by,
iꢀEðt=3600Þ ꢀEF
! ¼
¼
:
ð11Þ
iðt=FÞ
3600
The values of ! were calculated at different current densi-
ties and plotted against the current density as shown in
Fig. 10. ! per kilogram of hydrogen gas increases with the
increase in the operating cell current, and its values are more
pronounced for the RVC/Ni electrode. Significant differences
begin to appear at high operating cell currents which are of
practical interest for water electrolysis.
ꢀ
ꢁ
@ ln Io
Ea ¼ ꢁR
;
ð9Þ
Stability of Cu and Ni Coatings. The results discussed so
far showed that Cu and Ni coatings enhanced the rate of HER
both at lower and relatively higher temperatures. This was
attributed to the good electrocatalytic properties due to metal
loading. However, the stability of these coatings is of primary
importance. The stability was determined by holding the po-
tential at ꢁ2:0 V and measuring current–time relations for both
coatings. It was found that the current dropped in the first few
minutes to lower values and remained constant for a period of
about 30 h for both electrodes metal loaded. These results im-
plied that there was no surface failure for either RVC/Cu or
RVC/Ni. This period might be too short; however, it is reason-
able for the purpose of the present study. The stability of the
metal coatings was also examined by SEM micrographs before
and after the above period of operation. The micrographs are
shown in Figs. 11a and 11b for Cu and Ni coatings, respective-
ly. A comparison between the micrographs of RVC/Cu and
RVC/Ni before (Figs. 2b and 2c) and after (Figs. 11a and
11b), operation for this period indicates that the metal disper-
sions are mechanically stable. A lower degree of metal load-
ings on the used electrodes (Figs. 11a and 11b) could be seen
from the micrographs, and this was attributed to the removal of
@ð1=TÞ
where R is the gas constant. A plot of log Io vs 1=T gives a
straight line from which the slope can be used to determine
Ea. Arrhenius plots for the bare RVC and RVC/Ni are shown
in Fig. 9. The estimated values of Ea are found to be 32 and
14 kJ molꢁ1, for bare RVC and RVC/Ni, respectively. The
results showed that modification of RVC cathodes by coating
RVC with Ni reduces the apparent activation energy of the
HER. The reduction in Ea for RVC/Ni is pronounced which
indicates that the HER with RVC/Ni is more facile than on
RVC.
Energy Savings. The decrease in the polarization when
RVC was coated with Cu or Ni, as shown in Fig. 6, was further
analyzed. The energy saving in watt hours, ꢀꢅ, at the cathode
at a particular current, i is given by
iꢀEt
ꢀꢅ ¼
;
ð10Þ
3600
where t is the electrolysis time in seconds, ꢀE is the lowering
in the cathodic polarization (in Volt) at a particular current, i
(in Ampere) at the cathode due to Cu or Ni loading. The values
of ꢀE were calculated from the collected i–E relations for the

M. M. Saleh et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006) 1717
Ni electrodes are stable. The power savings due to metal load-
ing were evaluated. The model must be further modified to
account for surface coverage by gas bubbles.
(a)
References
1
J. M. Friedrich, C. P. de-Leon, W. Reads, F. C. Walsh,
J. Electroanal. Chem. 2004, 561, 203.
E. L. Gyenge, C. W. Oloman, J. Appl. Electrochem. 2003,
33, 665.
2
3
679.
4
2483.
5
1850.
6
1814.
7
R. Hozle, A. M. L. Castro, J. Appl. Electrochem. 1988, 18,
A. A.-Gallegos, D. Pletcher, Electrochim. Acta 1999, 44,
M. Maltoz, J. Newman, J. Elecrochem. Soc. 1986, 133,
J. Wang, H. D. Dewald, J. Elecrochem. Soc. 1983, 130,
R. C. Windner, M. F. B. Sousa, R. Bertazzoli, J. Appl.
(b)
Electrochem. 1998, 28, 201.
C. M. S. Piatnicki, D. S. Azambuja, E. E. S. Hasse, K. R. L.
Castagno, S. B. Guterres, Sep. Sci. Technol. 2002, 37, 2459.
I. A. Raj, Appl. Surf. Sci. 1992, 59, 245.
8
9
ˇ
´
ˇ
´
10 D. L. Stojic, M. P. Marceta, S. P. Sovilj, S. S. Miljanic,
J. Power Sources 2003, 118, 315.
11 M. Treuba, S. Trasatti, S. Trasatti, Mater. Chem. Phys.
2005, in press.
12 O. Bohme, F. U. Leidich, H. J. Salge, H. Wendt, Int. J.
¨
Hydrogen Energy 1994, 19, 349.
13 S. Rausch, H. Wendt, J. Electrochem. Soc. 1996, 143,
2852.
14 I. J. Brown, S. Sotiropoulos, J. Appl. Electrochem. 2000,
30, 107.
Fig. 11. SEM micrographs of the different electrodes: (a)
RVC/Cu and (b) RVC/Ni after an operation period of
30 h of HER at ꢁ2:0 V. The white zones are the matrix.
excess deposited metal (weak adhesion) which probably flaked
off in the electrolyte streams. This may occur the first few
minutes concurrent with the decrease in the initial current
decrease as stated above. A slight thinning of the average wire
diameter down to 20.83 and 21.43 mm is observed after pro-
longed operation of the RVC/Cu and RVC/Ni electrodes,
Figs. 11a and 11b, respectively. This fact reflects the loss of
some of the Cu and Ni from the fibrous matrix of the RVC
electrodes, while having a relatively thicker average wire di-
ameter compared to the bare RVC. The coatings remain ad-
hered for long time thus have the required performance and
stability. Electroplating for less than one hour is enough for
the present investigation.
15 M. S. El-Deab, M. M. Saleh, Int. J. Hydrogen Energy
2003, 28, 1199.
16 V. D. Stankovic, R. Grujic, A. A. Wragg, J. Appl. Electro-
chem. 1998, 28, 321.
17 M. R. V. Lanza, R. Bertazzoli, J. Appl. Electrochem. 2000,
30, 61.
18 F. Coeuret, V. E. Oliveira, C. E. Bezerra, J. Appl. Electro-
chem. 2002, 32, 1175.
19 J. Newman, W. Tiedman, Flow-through Porous Elec-
trodes, ed. by H. Gerischer, C. W. Tobias, in Advances in Electro-
chemistry and Electrochemical Engineering, ed. by H. Gerischer,
C. W. Tobias, Wiley, New York, 1978, Vol. 11, pp. 352–438.
20 R. E. Sioda, K. B. Keating, in Electroanalytical Chemistry,
ed. by A. J. Bard, Marcel Dekker Inc., New York, 1982, Vol. 12.
21 M. M. Saleh, J. Phys. Chem. B 2004, 108, 13419.
22 D. Dobos, Electrochemical Data, Elsevier, New York,
1975, p. 242.
23 B. G. Ateya, B. E. El-Anadouli, J. W. Weidner, M. M.
Saleh, J. Elecrochem. Soc. 1995, 142, 4122.
24 O. N. Tandon, V. K. Aggarwal, K. C. Dhinga, Handbook of
Electroplating and Anodization, Delhi, 1981, p. 181.
25 J. A. Trainham, J. Newman, J. Electrochem. Soc. 1977,
124, 1528.
Summary and Conclusion
The present study evaluated the possibility of modifying the
RVC electrodes by loading metal dispersions onto the porous
matrix. A mathematical model helped to understand the effects
of Q on the overall behavior of the porous electrode and on the
potential and current distributions of the HER within the elec-
trode. It was also possible to extract the exchange current den-
sity of the HER using different electrodes. Disagreements
between the calculated values and the experimental data occur-
red at higher current densities. RVC/Cu and RVC/Ni have
better electrocatalytic properties towards the HER than the
bare RVC, and this trend was confirmed by comparison with
a planar GC electrode instead of the RVC. Current-transient
and SEM micrographs showed that the RVC/Cu and RVC/
26 R. Alkire, B. Gracon, J. Electrochem. Soc. 1975, 122,
1594.
27 R. E. Merdith, C. W. Tobias, Conduction in Heterogeneous
Systems, in Advances in Electrochemistry and Electrochemical
Engineering, ed. by H. Gerischer, C. W. Tobias, Wiley, New
York, 1962, Vol. 2, pp. 17–48.

1718 Bull. Chem. Soc. Jpn. Vol. 79, No. 11 (2006)
Hydrogen Evolution on Modified RVC
28 J. Newman, Electrochemical Systems, 2nd ed., Prentice-
Hall, Englood Cliffs, NJ, 1991, p. 552.
29 M. M. Saleh, Electrochim. Acta 1999, 45, 959.
30 A. J. Appleby, M. Chelma, H. Kita, G. Bronoel, Encyclo-
pedia of the Electrochemistry of Elements, ed. by A. J. Bard,
Marcel Dekker, NY, 1982, p. 556.
31 W.-X. Chen, Int. J. Hydrogen Energy 2001, 26, 603.
32 C. Hitz, A. Lasia, J. Electroanal. Chem. 2000, 500, 213.
33 B. G. Ateya, L. Austin, J. Electrochem. Soc. 1977, 124,
1528.
34 B. G. Ateya, J. Electroanal. Chem. 1977, 75, 183.
35 H. J. Miao, D. L. Piron, Electrochim. Acta 1993, 38, 1079.