Promoting effect of transition metal-doped Co-B alloy catalysts for hydrogen production by hydrolysis of alkaline NaBH4 solution

Article abstract of DOI:10.1016/j.jcat.2010.02.014

A systematic and comparative study is conducted on Co-B-based ternary alloy catalysts for H₂ generation by hydrolysis of NaBH₄. Various transition metals, namely Ni, Fe, Cu, Cr, Mo, and W, were added to Co-B catalyst by chemical reduction of the corresponding metal salts. All the transition metals in the Co-B compounds behave in a dissimilar manner while influencing the catalytic activity. Cr, W, Mo, and Cu impose significant promoting effects on the Co-B catalyst by increasing the H₂ generation rate by 3-4 times when compared to the undoped catalyst. On the contrary, Ni and Fe are only able to create a marginal increment in the catalytic performance of the Co-B catalyst. The metal/(Co + metal) molar ratio was varied in the catalyst in order to study the effect of metal doping on surface morphology, electronic interaction, and catalytic efficiency of the alloy catalyst. This systematic variation also contributed to clarify the role of each dopant metal in the electron exchange mechanism involved in NaBH₄ hydrolysis. The promoting effects of the dopant metals are mainly discussed in terms of large active surface area, ability to act as Lewis acid sites for better absorption of OH^- group, electronic interaction with Co active metal, and amorphous nature of the alloy catalyst.

Full text of DOI:10.1016/j.jcat.2010.02.014

Journal of Catalysis 271 (2010) 315–324
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Promoting effect of transition metal-doped Co–B alloy catalysts for hydrogen
production by hydrolysis of alkaline NaBH4 solution
*
N. Patel , R. Fernandes, A. Miotello
Dipartimento di Fisica, Università degli Studi di Trento, I-38100 Povo, Trento, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A systematic and comparative study is conducted on Co–B-based ternary alloy catalysts for H2 generation
by hydrolysis of NaBH4. Various transition metals, namely Ni, Fe, Cu, Cr, Mo, and W, were added to Co–B
catalyst by chemical reduction of the corresponding metal salts. All the transition metals in the Co–B
compounds behave in a dissimilar manner while influencing the catalytic activity. Cr, W, Mo, and Cu
impose significant promoting effects on the Co–B catalyst by increasing the H2 generation rate by 3–4
times when compared to the undoped catalyst. On the contrary, Ni and Fe are only able to create a mar-
ginal increment in the catalytic performance of the Co–B catalyst. The metal/(Co + metal) molar ratio was
varied in the catalyst in order to study the effect of metal doping on surface morphology, electronic inter-
action, and catalytic efficiency of the alloy catalyst. This systematic variation also contributed to clarify
the role of each dopant metal in the electron exchange mechanism involved in NaBH4 hydrolysis. The
promoting effects of the dopant metals are mainly discussed in terms of large active surface area, ability
to act as Lewis acid sites for better absorption of OHꢀ group, electronic interaction with Co active metal,
and amorphous nature of the alloy catalyst.
Received 17 December 2009
Revised 8 February 2010
Accepted 9 February 2010
Available online 19 March 2010
Keywords:
H2 generation
Hydrolysis
Sodium borohydride
Doped Co–B catalyst
Surface area
Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction
exothermic nature of the reduction reaction involves high surface
energy causing Co–B particles to agglomerate. This particle
Commercialization of fuel cell-based portable devices is hin-
dered by the lack of an efficient and safe medium for hydrogen
storage and generation. Chemical hydrides like sodium borohy-
dride (NaBH4) have attracted worldwide interest as a source to
supply pure hydrogen to fuel cells at room temperature, because
it is stable in alkaline solution, non-flammable, non-toxic in nature,
and with hydrogen storage capability of 10.8 wt.% [1]. Further-
more, hydrogen is generated by a water-based hydrolysis reaction
of NaBH4 with the advantage of producing half of the hydrogen
from the water solvent [2]. However, a catalyst is required to accel-
erate this hydrolysis reaction in a controllable manner. Therefore,
the catalyst is the key factor that affects the H2 generation rate
by hydrolysis of NaBH4. Solid-state catalysts such as precious met-
als (generally functionalized with support) [3–5] or transition met-
als [6,7] and their salts are found to be very efficient in accelerating
the hydrolysis reaction in a controllable manner.
agglomeration lowers the effective surface area of the catalyst
powder thus limiting its catalytic activity. In recent years, several
routes have been adopted to improve the catalytic activity of the
Co–B catalyst, mainly by increasing the surface area. This was
accomplished by depositing Co–B on a support material having
high specific surface area [9,10] or by using organic templates dur-
ing the reduction reaction [11]: the final surface area of these cat-
alyst powders is much higher than that of the ‘‘normal” Co–B
powder. Nanoparticle-assembled Co–B catalysts films synthesized
by pulsed laser deposition technique showed a performance simi-
lar to that of the noble metals because of the achieved high surface
area [7,12].
Another efficient route to avoid agglomeration of Co–B particles
is by introducing, between them, an atomic diffusion barrier in the
form of transition metals like Cr, Mo, or W [13]. These promoter
metals, mainly in form of oxides, are really efficient and even a
small atomic concentration is able to significantly increase the sur-
face area of the metal–borides catalyst powder [14] by avoiding
agglomeration. In addition, they also contribute to the overall cat-
alytic reaction by acting as acidic sites to improve the absorption of
the reactant species on the surface [15]. The second metal also acts
as an electron donor ligand, thus increasing the electron density on
the active metal atoms which favors the reaction kinetics [16].
These combined effects, along with the simple preparation meth-
od, make such alloy catalysts very attractive. Our previous reports
Amorphous cobalt–boride (Co–B) catalysts, prepared by reduc-
tion of metal salts with a reducing agent, have attracted great
attention in catalysis owing to their unique properties such as iso-
tropic structure, high concentration of coordinative unsaturated
sites, relevant chemical stability, and low cost [8]. However, the
* Corresponding author. Address: University of Trento, Department of Physics,
Via Sommarive, 14, I-38050 Povo, Trento, Italy.
E-mail address: patel@science.unitn.it (N. Patel).
0021-9517/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2010.02.014

316
N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
proved that the alloys exhibit superior catalytic activity in hydro-
gen production by hydrolysis of NaBH4 [14,16–18]. There are avail-
able literatures [14–16] on transition metal-doped Co–B catalyst
used for different reactions, but none of the published papers com-
pares the effect of the different dopants for a single specified reac-
tion. Thus, it is very important to make a comparative analysis on
the promoting effects of different dopants in Co–B catalyst, for a
single specified reaction, to gain insight into the involved
mechanisms.
In the present work, we investigated the effect of various do-
pant transition metals (such as Ni, Fe, Cu, Cr, Mo, and W) in Co–B
catalyst powder on hydrogen production by hydrolysis of NaBH4.
We discovered that these dopant metals behave in a specific man-
ner to enhance the catalytic performance of the Co–B catalyst. On
the basis of characterization results, the role of each metal species,
in the electron exchange mechanisms involved in the NaBH4
hydrolysis, is discussed.
course. A detailed description of the measurement apparatus is re-
ported in reference [20]. In all measurement runs, the catalyst was
placed on the insertion device inside the reaction chamber and the
system was sealed. Catalyst powder was added to 200 ml of the
above solution, at 298 K, under continuous stirring. For compari-
son, the stoichiometric cumulative hydrogen production yield (%)
versus time was plotted instead of the hydrogen volume (ml) ver-
sus time. The H2 generation rate was measured at different solu-
tion temperatures to determine the activation energy involved in
the catalytic hydrolysis reaction for all the catalysts.
To prove that metal cationic species, in Co–B catalyst, act as a
Lewis acid site in the absorption of OHꢀ group during catalytic
hydrolysis reaction, two types of hydride solution (0.025 M) were
prepared, namely: (1) NaBH4 solution stabilized with NaOH
(0.025 M) (designated as solution A) and (2) NaBH4 solution with-
out NaOH (designated as solution B). Hydrogen generation rate
was measured by hydrolysis of solutions A and B using all ternary
alloy catalysts and Co–B catalyst powders.
2. Experimental
3. Results and discussion
Ternary alloy catalysts in the form of Co–Ni–B, Co–Fe–B, Co–
Cu–B, Co–W–B, Co–Cr–B, or Co–Mo–B powder were synthesized
by mixing nickel chloride (NiCl2), iron chloride (FeCl2), copper ni-
trate (Cu(NO3)2), sodium tungstate dihydrate (Na2O4Wꢁ2H2O),
chromium nitrate (Cr(NO3)3), or molybdenum chloride (MoCl2)
salts, respectively, in the cobalt chloride (CoCl2) aqueous solution.
Further, these different mixture solutions were reduced by the so-
dium borohydride (NaBH4) under vigorous stirring. An excess
amount of borohydride was used in order to completely reduce
the metal cations. The black powder separated from the solution
during reaction course was filtered and then extensively washed
with distilled water and ethanol before drying at around 323 K un-
First of all, the effect of dopants concentration on the catalytic
activity of Co–B catalyst was studied by synthesizing the alloy cat-
alyst with different metal/(Co + metal) (vM) molar ratio. Fig. 1a–f
presents the hydrogen generation yield, as function of time, mea-
sured by hydrolysis of alkaline NaBH4 (0.025 M) solution using
Co–Ni–B, Co–Fe–B, Co–Cu–B, Co–Cr–B, Co–Mo–B, and Co–W–B cat-
alyst powders with different Ni, Fe, Cu, Cr, Mo, and W concentra-
tion, respectively. Inset of Fig. 1 presents the plot of Rmax value
(the maximum H2 generation rate, for all the catalyst powders,
which we established by a numerical procedure described else-
where [12]) as a function of
and W). The H2 generation rate increases with the increase in
M% and reaches the maximum with Ni = 15%, Fe = 35%, vCu
35%, Cr = 4%, Mo = 5%, and W = 5%. With further increase in vM,
vM% (where M = Ni, Fe, Cu, Cr, Mo,
der continuous N2 flow. The metal to cobalt molar ratio (vM = M/
v
v
v
=
(M + Co), where M = Ni, Fe, Cu, W, Cr or Mo) was adjusted in the fi-
nal ternary alloy catalysts by varying the molar concentration of
metal salts to cobalt salt in the aqueous solution. For comparison,
we also prepared Co–B powder with same procedure by reducing
only cobalt salt.
The surface morphology of all catalyst powders was studied by
scanning electron microscope (SEM-FEG, JSM 7001F, JEOL)
equipped with energy-dispersive spectroscopy analysis (EDS, INCA
PentaFET-x3) to determine the composition of the samples. Struc-
tural characterization of the catalyst powders was performed by
v
v
v
the activity of the powder decreases (each measurement is re-
peated at least three times). The promoting effect of Cr, Mo, and
W on catalytic activity is achieved at low concentration doping
(4–5 molar%), and it is so evident that even a small amount of dop-
ing (ꢂ1 molar%) is able to double the activity of the Co–B catalyst.
Much higher concentrations of Ni, Fe, and Cu have to be used in
Co–B catalyst to attain a pronounced promoting effect in H2 gener-
ation rate. No significant effect was observed on the catalytic activ-
ity of Co–B catalyst with low concentration (below 10 molar%)
doping of these metals (Ni, Fe, and Cu).
conventional X-ray diffraction (XRD) using the Cu K radiation
a
(k = 1.5414 Å) in Bragg–Brentano (h–2h) configuration. Surface
electronic states and the related atomic composition of the cata-
lysts were determined by using X-ray photoelectron spectroscopy
(XPS). X-ray photoelectron spectra were acquired using a SCIENTA
Comparative analysis of the alloy catalysts powders was carried
out with vM% values where Rmax is highest (see inset of Fig. 1), for
the hydrolysis of NaBH4. Hydrogen generation yield was measured,
as a function of time, during the hydrolysis of alkaline NaBH4
ESCA200 instrument equipped with
a monochromatic Al K
a
(0.025 M) solution in presence of Co–B, Co–Ni–B (
Fe–B ( Fe = 35%), Co–Cu–B ( Cu = 35%), Co–Cr–B (
Mo–B ( Mo = 5%), and Co–W–B ( W = 5%) powder catalysts at
vNi = 15%), Co–
(1486.6 eV) X-Ray source and a hemispherical analyzer. No electri-
cal charge compensation was required to perform XPS analysis.
The BET surface area of the powder catalysts was determined by
nitrogen absorption at 77 K (Micromeritics ASAP 2010) after
degassing.
For catalytic activity measurements, an alkaline-stabilized solu-
tion of sodium borohydride (pH 13, 0.025 0.001 M) (Rohm and
Haas) was prepared by the addition of NaOH. The titer of reagent
was independently measured through an iodometric method
[19]. The generated hydrogen quantity was measured through a
gas volumetric method in an appropriate reaction chamber with
thermostatic bath which maintains constant temperature within
accuracy of 0.1 K. The chamber was equipped with a pressure sen-
sor, stirrer system, catalyst insertion device, and also coupled with
an electronic precision balance to accurately measure the weight of
water displaced by the hydrogen produced during the reaction
v
v
vCr = 4%), Co–
v
v
298 K (Fig. 2). The expected total amount of H2 was measured irre-
spective of the type of catalyst used. The H2 generation yield val-
ues, reported in Fig. 2, were perfectly fitted by using a single
exponential function [17], as described by:
ꢀ
1t
1t
½H2ꢃðtÞ ¼ ½H2ꢃmax ꢄ ð1 ꢀ eꢀk Þ ¼ 4½½BH4 ꢃ0 ꢄ ð1 ꢀ eꢀk Þꢃ
ð1Þ
where ½BHꢀꢃ is the initial molar concentration of NaBH4 in the solu-
4
0
tion and k1 is the overall rate constant of the first-order reaction.
This indicates that hydrolysis reaction is first-order reaction with
respect to NaBH4. In the present case, a low hydride/catalyst ratio
was used which means that the first-order kinetics involving diffu-
sion of BHꢀ4 on the catalyst surface is the rate-limiting step during
the hydrolysis reaction (see discussion reported in Ref. [17]). Rmax
,

N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
317
100
(b)
(Co-B)
10 %
15 %
25 %
35 %
50 %
(a)
100
80
60
40
20
0
Co-B
10 %
25 %
80
35 %
50 %
Co-Fe-B
1250
1000
750
500
250
0
Co-Ni-B
1250
1000
750
500
250
0
60
40
20
0
10
20
30
40
50
60
0
10
20
Ni
30
40
50
χFe
%
χ
%
0
0
2000
4000
6000
0
2000
4000
6000
Time (sec)
Time (sec)
100
100
80
60
40
20
0
(c)
(d)
Co-B
9 %
5 %
4 %
2 %
1 %
Co-B
5 %
20 %
35 %
75 %
80
60
40
20
0
Co-Cu-B
Co-Cr-B
2000
1500
1000
500
0
3500
3000
2500
2000
1500
1000
500
0
10 20 30 40 50 60 70 80
0
0
2
6
8
10
χC4r
%
χCu
%
0
500
1000
1500
2000
2500
0
1000
2000
3000
Time (sec)
Time (sec)
100
80
60
40
20
0
Co-B
100
80
60
40
20
0
(e)
(f)
Co-B
1 %
4 %
5 %
7 %
1 %
3 %
4 %
5 %
Co-Mo-B
Co-W-B
3000
2500
2000
1500
1000
500
2500
2000
1500
1000
500
0
0
0
2
4
6
8
10
0
1
3
4
5
6
7
8
2 χMo
%
χW
%
0
1000
2000
3000
0
500
1000
1500
2000
2500
Time (sec)
Time (sec)
Fig. 1. Hydrogen generation yield as a function of reaction time obtained by hydrolysis of alkaline NaBH4 (0.025 M) with (a) Co–Ni–B, (b) Co–Fe–B, (c) Co–Cu–B, (d) Co–Cr–B,
(e) Co–Mo–B, and (f) Co–W–B with different vM values (where M = Ni, Fe, Cu, Cr, Mo, and W). Inset shows the maximum H2 generation rate (Rmax) as a function of vM
.
for all the catalyst powders, are summarized in Table 1. We observe
that these values decrease in the following order: Co–Cr–B > Co–
Mo–B > Co–W–B > Co–Cu–B > Co–Fe–B > Co–Ni–B > Co–B. The mea-
surements are repeated at least three times to verify the obtained
results. The Co–Cr–B catalyst powder (ꢂ3400 ml/min/g), with
have been studied to understand the promoting role of each tran-
sition metal.
In Fig. 3, we present the SEM images of all the alloy powder cat-
alysts. All samples show similar particle-like morphology with par-
ticles having spherical shape and size of a few nanometers. During
catalyst preparation, the use of NaBH4 as a reducing agent causes a
fast reduction of Co ions that inhibits particle growth above a few
nanometers. However, at higher SEM magnification, it is evident
that these particles are mostly in agglomerated state in the case
of Co–B, Co–Ni–B, and Co–Fe–B powders. On the contrary, a lower
degree of agglomeration is observed with Co–Cu–B, Co–W–B, Co–
Cr–B, and Co–Mo–B powders, and such a morphology is helpful
to enhance the active surface area of catalysts (as discussed later
v
Cr = 4%, shows about four times higher H2 generation rate when
compared to that achieved with Co–B catalyst powder (ꢂ850 ml/
min/g). The above results prove that doping Co–B powder with
whichever of the transition metal (Cr, Mo, W, Cu, Fe, and Ni), with
small or large concentration, enhances the catalytic efficiency of
the catalyst.
Morphological, structural, and electronic changes, occurring in
the Co–B powder with addition of the different transition metals,

318
N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
1.0–1.2 eV is evident when comparing the BE of pure boron
100
80
60
40
20
0
(187.1 eV) to that of boron in the catalyst. This shift indicates an
electron transfer from alloying B to vacant d-orbital of metallic
Co, which makes the B atom electron deficient and the Co atom
electron enriched: this is true for all the catalyst powders [24].
The lack of shift in Co peak in compound catalyst is due to the hea-
vy mass of the atom.
Co-B
Co-Cu-B
Co-Ni-B
Co-Fe-B
Co-Cr-B
Co-W-B
Co-Mo-B
Fig. 6a–f presents the XPS spectra of Ni2p, Fe2p, Cu2p, Cr2p, Mo3d
,
and W4f electronic levels in Co–Ni–B, Co–Fe–B, Co–Cu–B, Co–Cr–B,
Co–Mo–B, and Co–W–B, powders, respectively. The XPS result for
each catalyst powder is separately discussed below:
1. Co–Ni–B: Two peaks appear in Ni2p3/2 levels with binding ener-
gies of 852.6 and 856.3 eV (Fig. 6a), indicating that Ni metal
exists in both the elemental and oxidized states, respectively.
The oxide is mainly in the form of Ni(OH)2 which would have
been formed during the catalyst preparation reaction between
NaBH4 and metal salts, similarly to cobalt oxide. However, only
a few amount of Ni is present in oxidized form (15 at.%), while
most of it is in the form of metallic Ni (85 at.%). The BE of metal-
lic Co in Co–Ni–B is 0.2 eV lower than that in the Co–B, whereas
BE of metallic Ni in Co–Ni–B is also 0.2 eV lower than that in Ni–
B [16]. This proves that in Co–Ni–B catalyst, there is an elec-
tronic enrichment on Co and Ni metals when compared with
the Co–B and Ni–B catalysts: this is mainly attributed to the
B-enrichment in Co–Ni–B catalyst as seen by the compositional
analysis of XPS spectra (as reported in our previous work [16]
and also reported by Li et al. [25]). These authors also per-
formed theoretical calculations using DFT method to show that
the increase in the B content on the surface would contribute
more electrons to metallic Ni or Co in their corresponding
amorphous alloys [25].
2. Co–Fe–B: The elemental Fe in Co–Fe–B catalyst is present in
both metallic and oxidized state with binding energies of
707.1 and 712.7 eV, respectively (Fig. 6b). The oxidized Fe spe-
cies on the catalyst surface is mostly in the form of Fe2O3 state
(90 at.%) probably formed during the preparation. The metallic
Fe (10 at.%) shows a positive shift of about 0.3 eV in its corre-
sponding BE peak, which suggests electron transfer from Fe to
Co atoms. However, no shift is observed in the BE of the Co peak
probably because a few Fe atoms are present in metallic state to
transfer electrons to Co atoms.
0
1000
2000
3000
4000
Time (sec)
Fig. 2. Hydrogen generation yield as a function of reaction time obtained by
hydrolysis of alkaline NaBH4 (0.025 M) with Co–B, Co–Ni–B ( Ni = 15%), Co–Fe–B
Fe = 35%), Co–Cu–B ( Cu = 35%), Co–Cr–B ( Cr = 4%), Co–Mo–B ( Mo = 5%), and Co–
W–B ( W = 5%) powder catalysts at 298 K.
v
v
(v
v
v
v
on the basis of BET measurement). XRD patterns (Fig. 4) of all the
Co–B-based alloy powders show a broad peak at around 2h = 45°,
which is assigned to the amorphous state of Co–B alloy. Other
broad peaks at around 35° and 43°, attributed to the presence of
Cr2O3 and Cu metal, are also observed for Co–Cr–B and Co–Cu–B
catalyst powders, respectively. Finally, the diffraction spectra indi-
cate short-range order and long-range disorder, and both these fea-
tures might contribute to enhance the catalytic activity [21].
XPS was used to gain insight into the electronic states and sur-
face interaction between atoms in all catalyst compounds. Fig. 5a
and b shows the XPS spectra of Co2p3/2 and B1s electronic levels,
respectively, in all the catalyst powders. Two peaks appear, for
all the catalyst compounds, corresponding to the Co2p3/2 level at
the binding energies (BE) in a range of 778.2–778.4 and 781.6–
781.8 eV. This suggests that Co element exists in both elemental
and oxidized states, respectively. The +2 state related to the peak
of the oxidized cobalt is attributed to Co(OH)2: this molecule
would have been formed possibly during catalyst preparation
[22] or when it was exposed to ambient atmosphere before trans-
ferring to the reaction chamber. This cobalt hydroxide (Co(OH)2)
does not cause major effects on catalytic activity, because it can
be easily reduced by sodium borohydride in the course of the
hydrolysis forming Co–B active phase. Liu and Li [23] showed
experimentally that Co(OH)2 intermediate acts as nuclei for the
precipitation of Co–B active phase during the reaction with NaBH4.
There is no significant shift in the present Co BE peak when com-
pared to the standard binding energy of metallic Co for all the cat-
alyst powders, except for the Co–Ni–B catalyst. Two XPS peaks are
also observed at BE values of 188.1–188.3 eV and 192.1–192.3 eV,
which correspond to elemental and oxidized B1s levels, respec-
tively, for all catalyst powders [22]. A positive shift of around
3. Co–Cu–B: For Cu, only a single peak corresponding to a metallic
state is apparent at BE of 932.6 eV (Fig. 6c). This is consistent
with XRD result which signals indeed the presence of metallic
Cu. In addition, no shifts in the BE of Co and Cu peaks are
observed indicating the lack of electronic interaction between
these atoms. The electronic exchange between Co and B atoms
is also not affected by the Cu addition.
4. Co–Cr–B: The Cr2p spectrum for Co–Cr–B catalysts exhibits two
peaks at 575.7 and 585.3 eV (Fig. 6d), attributed to Cr2p3/2 and
2p1/2 peaks of Cr3+ species, respectively. This shows that Cr is
in completely oxidized state (Cr2O3), which is in agreement
Table 1
Molar ratio (%) of M/(M + Co) (
vM, where M stands for: Ni, Fe, Cu, Cr, Mo, and W), surface area, maximum H2 generation rate, ability to act as Lewis acid sites, and activation
energy of the as-prepared alloy catalyst powders.
Catalyst powders
M/(M + Co) molar
BET surface
area (m2/g)
Max H2 generation rate (Rmax) (ml/min/g catalyst)
Lewis acid
Activation energy
ratio (
v
M)(%)
(kJ molꢀ1
)
Alkaline NaBH4
Non-alkaline NaBH4
Co–B
–
20
22
33
115
40
43
850
1175
1300
2210
3400
2875
2570
860
1160
1270
2180
2250
2210
1940
No
No
No
No
Yes
Yes
Yes
45
34
31
30
37
39
41
Co–Ni–B
Co–Fe–B
Co–Cu–B
Co–Cr–B
Co–Mo–B
Co–W–B
15
35
35
4
5
5
54

N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
319
Fig. 3. SEM micrographs of Co–B, Co–Ni–B (
vNi = 15%), Co–Fe–B (
vFe = 35%), Co–Cu–B (vCu = 35%), Co–Cr–B (
vCr = 4%), Co–Mo–B (
v
Mo = 5%), and Co–W–B (
v
W = 5%) powder
catalysts.
with the XRD result. It is also noticed that Cr does not modify
the electronic interaction between Co and B, because their rel-
evant BEs remain unchanged after the Cr addition.
peaks assigned to MoO3 (BE of 233.7 and 235.8 eV are ascribed
to Mo 3d5/2 and Mo 3d3/2 levels of Mo6+, respectively). There is
no evidence of metallic Mo in the spectra indicating that Mo
atoms on the catalyst surface are in completely oxidized state.
The BE value for Co and B is unaffected by the addition of Mo
in Co–B catalyst.
5. Co–Mo–B: The XPS spectra of Mo 3d levels (Fig. 6e) reveal two
peaks assigned to MoO2 (BE of 231.1 and 232.6 eV are ascribed
to Mo 3d5/2 and Mo 3d3/2 levels of Mo4+, respectively) and two

320
N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
The above XPS results prove that all the transition metal do-
Co-W-B
Co-Mo-B
Co-Cr-B
pants in Co–B catalyst exist in different states (elemental, oxidized,
oxidized in different proportion). This may be the prime reason to
explain the different promoting effects on Co–B catalyst powders
as we will discuss in the following sections.
The measured BET surface area of all the catalyst powders is re-
ported in Table 1. The specific surface area decreases in following
order: Co–Cu–B > Co–W–B > Co–Mo–B > Co–Cr–B > Co–Fe–B > Co–
Ni–B > Co–B. This shows that doping the Co–B catalyst with transi-
tion metals (Cr, Mo, W, Fe, Cu, and Ni) causes an increment in the
specific surface area irrespective of the type of dopant. However,
the trend obtained in the increment of the surface area is quite dif-
ferent from that of the catalytic activity (Fig. 2).
Co-Cu-B
Co-Fe-B
Co-Ni-B
Co-B
When small amount of Cr, Mo, or W (vCr,Mo,W = 4–5%) is in-
30
40
50
60
70
80
90
cluded as a dopant in Co–B, the specific surface area of the catalyst
increases by almost 2–2.5 times, a result which is due to the lim-
ited particle agglomeration process when compared to pure Co–B
(see Fig. 3). Surface characterization (Figs. 3, 5 and 6) of these cat-
alyst powders suggests that Cr, Mo, and W in the form of oxides are
located on the surface of the catalyst. These metal oxides act as a
kind of atomic barrier between the several Co–B nanoparticles,
thus avoiding agglomeration caused by their high surface energy.
As a result, the effective surface area increases significantly for
Co–Cr–B, Co–Mo–B, and Co–W–B catalyst. A remarkable increment
in the surface area, a factor of about 5.5, is observed for Co–Cu–B
catalyst when compared to undoped Co–B. As observed by XRD
and XPS, the Cu on the catalyst surface is in completely metallic
state which prevents the agglomeration of Co–B nanoparticles.
The highest surface area of Co–Cu–B is mainly due to the high con-
centration of Cu when compared to Cr, Mo, and W-dopant in Co–B
catalyst. Our previous work proved that when Cr is added at high
2Theta (deg.)
Fig. 4. XRD pattern of Co–B, Co–Ni–B (
Cu = 35%), Co–Cr–B
powder catalysts.
vNi = 15%), Co–Fe–B (
vFe = 35%), Co–Cu–B
(v
(vCr = 4%), Co–Mo–B (vMo = 5%), and Co–W–B (vW = 5%)
Co2p3/2
(a)
Co-Mo-B
Co-Fe-B
Co-W-B
Co-Cu-B
Co-Ni-B
Co-Cr-B
Co-B
concentration (vCr = 10%) in Co–B catalyst, the surface area in-
creases by one order of magnitude when compared to undoped
Co–B catalyst [14]. Similar result is reported by Chen et al. [26]
for Mo-promoted Co–B catalyst. However, a deleterious effect in
catalytic activity is detected with high concentration of the dopant
metals (Cr, Mo, and W) as we will discuss later. Co–Fe–B shows a
moderate enhancement in the surface area, when compared to
786
784
782
780
778
776
774
Binding Energy (eV)
B1S
(b)
Co-W-B
Co–B, even at high concentration of Fe (vFe = 35%), probably attrib-
uted to the formation of Fe2O3 on the surface. On the other hand,
negligible rise in surface area is monitored for Co–Ni–B catalyst
powder. For the moment, it is not clear why Fe and Ni are ineffec-
tive to significantly increase the surface area of Co–B powder.
In a binary metal alloy catalyst, the addition of a new metal, or
its oxide, may act as an electrophilic or Lewis acid site for the
absorption and activation of the reactants to increase the overall
reaction rate. To establish the possible role of the dopant metal
as a Lewis acid site in CoB catalyst, it is initially important to recall
the NaBH4 hydrolysis mechanism involving a solid-state catalyst.
The H2 production from the metal-catalyzed hydrolysis reaction
of NaBH4 takes place, according to Guella et al. [3], by the following
kinetics steps: (1) BHꢀ4 ions are chemisorbed on the metal (Co)
atoms; (2) Hꢀ is transferred from BHꢀ4 to an unoccupied adjacent
(Co) metal atom; (3) the hydrogen atom acquires an electron from
the metal and leaves the site in hydridic form (Hꢀ), while BH3ꢀ spe-
cies remains attached to the Co atom; (4) this Hꢀ reacts with water
molecule to produce H2 and OHꢀ; and (5) OHꢀ reacts with boron in
BHꢀ3 to produce the BH3(OH)ꢀ ion. Again, Hꢀ is transferred from
BH3(OH)ꢀ ion to an unoccupied adjacent (Co) metal atom. The cy-
cle of hydrogen absorption on metal sites continues till BH3(OH)ꢀ
forms BðOHÞꢀ4 and molecular hydrogen is released during each cy-
cle. However, during the last step, the reaction between OHꢀ and
BHꢀ3 ions is not favored by the catalyst which may cause lower
reaction rate. In present Co–B-based alloy catalyst, the dopant me-
tal or its oxides may act as the Lewis acid sites, which are readily
Co-Mo-B
Co-Cr-B
Co-Cu-B
Co-Fe-B
Co-Ni-B
Co-B
198
196
194
192
190
188
186
184
Binding Energy (eV)
Fig. 5. X-ray photoelectron spectra of Co2p3/2 and B1s levels for Co–B, Co–Ni–B
Ni = 15%), Co–Fe–B ( Fe = 35%), Co–Cu–B ( Cu = 35%), Co–Cr–B ( Cr = 4%), Co–Mo–
B ( Mo = 5%), and Co–W–B ( W = 5%) powder catalysts.
(v
v
v
v
v
v
6. Co–W–B: The W-4f spectrum (Fig. 6f) of Co–W–B catalyst shows
the presence of both oxidized and elemental state of W. The XPS
spectrum of W 4f7/2 (4f5/2) shows two peaks at BE values of 31.3
(33.4) eV and 35.5 (36.9) eV assigned to metallic W and WO3
species, respectively. About 33% of the XPS signal rises from ele-
mental W, whereas 67% is associated to the oxidized W. By
comparing with the undoped Co–B sample, no significant
change in XPS spectra is observed in connection with the pres-
ence of the W-dopant: this means that W does not affect the
electronic states of Co and B.

N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
321
(a)
(b)
Ni2p
Fe2p
866 864 862 860 858 856 854 852 850
Binding Energy (eV)
720 718 716 714 712 710 708 706 704
Binding Energy (eV)
(c)
(d)
Cr2p
Cu2p
935
934
933
932
931
930
595
590
585
580
575
570
565
Binding Energy (eV)
Binding Energy (eV)
(e)
(f)
Mo3d
W4f
240
238
236
234
232
230
228
42
40
38
36
34
32
30
28
Binding Energy (eV)
Binding Energy (eV)
Fig. 6. X-ray photoelectron spectra of (a) Ni2p level for Co–Ni–B (vNi = 15%), (b) Fe2p level for Co–Fe–B (
vFe = 35%), (c) Cu2p level for Co–Cu–B (vCu = 35%), (d) Cr2p level for Co–
Cr–B ( Cr = 4%), (e) Mo3d level for Co–Mo–B ( Mo = 5%), and (f) W4f level for Co–W–B (v
v
v
W = 5%) powder catalysts.
available for the absorption of Lewis base such as OHꢀ ions. During
the hydrolysis reaction, the hydroxyl molecule in the solution can
be absorbed on these sites via donating the lone electron from OHꢀ
group. This bonding polarizes the hydroxyl group, which creates
favorable condition for the reaction with BHꢀ3 ions attached to
the neighboring active Co sites and thus increasing the overall
reaction rate.
A set of experiments was conducted to find out which dopant
transition metal in Co–B catalyst acts as Lewis acid site to assist
the absorption of the OHꢀ ions. For the catalytic hydrolysis reac-
tion, two types of hydride solution (0.025 M) were prepared,
namely: (1) NaBH4 solution stabilized with NaOH (0.025 M) (desig-
nated as solution A) and (2) NaBH4 solution without NaOH (desig-
nated as solution B). The hydrogen generation rate was measured
by hydrolysis of solutions A and B using all the alloy catalyst pow-
ders and reported in Table 1. No significant change in H2 genera-
tion rate was observed through hydrolysis of alkaline and non-
alkaline NaBH4 in the presence of Co–B, Co–Ni–B, Co–Fe–B, or
Co–Cu–B catalyst powders. On the contrary, the H2 generation rate
obtained in the presence of Co–Cr–B, Co–Mo–B, or Co–W–B for
solution A was much higher than that of solution B. This proves
that during the hydrolysis reaction, in the presence of hydroxyl
molecule in the solution, Cr, Mo, and W species, in the form of oxi-
des, act as the Lewis acid sites to absorb the OHꢀ group and cata-
lyze the reaction between OHꢀ and boron species (BH3ꢀ, BH2ðOHÞꢀ,
and BHðOHÞꢀ). However, it is observed that the increment in H2
2
generation rate, by addition of NaOH during the hydrolysis of
NaBH4, is far better for Co–Cr–B catalyst when compared to

322
N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
298 K
100
100
80
60
40
20
0
303 K
(a)
(d)
298 K
303 K
308 K
313 K
308 K
313 K
318 K
323 K
80
60
40
20
0
Co-Ni-B
Co-Cr-B
Actvation Energy = 34 kJ/mol
Actvation Energy = 37 kJ/mol
1/T (K )
1/T (K )
0
200
400
600
Time (Sec)
800
1000
1200
0
1000
2000
Time (sec)
3000
4000
100
80
60
40
20
0
100
80
60
40
20
0
298 K
303 K
308 K
313 K
318 K
(e)
(b)
298 K
303 K
308 K
313 K
318 K
Co-Fe-B
Co-Mo-B
Activation Energy = 38.7 kJ/mol
Actvation Energy = 31 kJ/mol
1/T (K )
1/T (K )
0
1000
2000
3000
4000
5000
0
500
1000
Time (sec)
1500
2000
Time (sec)
100
100
80
60
40
20
0
(c)
(f)
298 K
303 K
308 K
313 K
318 K
298 K
303 K
308 K
313 K
80
60
40
20
0
Co-Cu-B
Co-W-B
Activation Energy = 29.9 kJ/mol
Activation Energy = 41 kJ/mol
1/T (K )
1/T (K )
0
500
1000
1500
2000
0
500
1000
1500
Time (sec)
Time (sec)
Fig. 7. Hydrogen generation yield as a function of reaction time by hydrolysis of alkaline NaBH4 (0.025 M) solution measured at different solution temperatures with (a) Co–
Ni–B ( Ni = 15%), (b) Co–Fe–B ( Fe = 35%), (c) Co–Cu–B ( Cu = 35%), (d) Co–Cr–B ( Cr = 4%), (e) Co–Mo–B ( Mo = 5%), and (f) Co–W–B ( W = 5%) powder catalysts. Inset shows
the Arrhenius plot of the H2 generation rates.
v
v
v
v
v
v
Co–Mo–B or Co–W–B catalyst powders. The Cr3+ species might
have a stronger affinity for the oxygen of hydroxyl group than
W4+ or Mo4+ species, thus Cr-additives exhibit a stronger promot-
ing effect than Mo- or W-additives.
reaction by providing the electrons required by the hydrogen
atom during the hydrolysis reaction. Thus, the catalytic activity
is moderately higher for Co–Ni–B catalyst, when compared to
the Co–B, due to the electron-enriched active metal site: this
The role of each dopant transition metal in Co–B catalyst, in
enhancing the H2 generation rate, is now underlined on the basis
of the characterization results of the alloy powder catalysts.
is the case for the Ni-doping concentration with
15%. However, the H2 generation rate is maximum with
Ni = 15% and further increase in Ni concentration causes the
vNi = 10% and
v
activity to decrease. As observed in our previous work [16],
Co–B shows much higher catalytic activity when compared to
Ni–B for hydrolysis of NaBH4, thus indicating that Co metal is
a more active site than Ni for catalytic hydrolysis reaction.
The decrement in the activity by increasing the amount in Ni
is due to the fact that more active Co sites are replaced by less
active Ni sites.
1. Co–Ni–B catalyst: Nickel on the catalyst surface is mostly in a
metallic state, which neither increases significantly the surface
area of Co–B powder nor acts as a Lewis acid site for OHꢀ ions
absorption. However, Ni is able to increase the density of elec-
trons on the Co active sites due to boron enrichment. These
electron-enriched metal active sites will favor the catalysis

N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
323
2. Co–Fe–B catalyst: By increasing
alyst first increases and then decreases. The maximum activity
is obtained with Fe = 35%. By knowing that Fe–B alloy catalyst
v
Fe, the activity of Co–Fe–B cat-
the surface covering of the active Co metal by Cr, Mo, or W
species.
v
is less active than Co–B for the hydrolysis of NaBH4, it is easily
understood that only metallic Co acts as a main active site in the
Co–Fe–B alloy and that Fe-dopant behaves as a promoter. At vFe
values from 10% to 35%, the promoting effect of Fe-dopant is
mainly attributed to the increased surface area of the Co–Fe–B
catalyst due to the existence of Fe2O3 species, on the catalyst
surface, which act as an inhibitor against agglomeration of the
Co–B particles. The presence of a little amount of metallic Fe
could also make a minor contribution to the enhancement of
catalytic activity by electron transfer to the active Co sites as
observed by XPS (Fig. 6b). Electron-enriched Co active sites will
facilitate the catalysis reaction as explained for previous Co–Ni–
B catalyst. At larger concentration of Fe-dopant, the catalytic
activity decreases because too many surface Co active sites
would be covered by inactive Fe species.
To summarize, the transition metals Ni, Fe, Cu, Cr, Mo, and W,
when added to the Co–B, promote the catalysis efficiency through
three effects, namely by:
1. Creating higher electron density on to the active Co sites.
2. Increasing the specific surface area of the final catalyst powder
by inhibiting Co–B particle agglomeration and thus favouring
better particle dispersion.
3. Acting, in the form of oxides on the catalyst surface, as the
Lewis acid sites for the absorption of OHꢀ group ions and cata-
lyzing its reaction with boron species.
The investigated transition metals are able to enhance the cat-
alytic efficiency of the Co–B catalyst by only one of the three above
effects (Co–Ni–B and Co–Cu–B catalyst) or a combination of two ef-
fects (Co–Cr–B, Co–Mo–B, Co–W–B, and Co–Fe–B catalyst).
To conclude the analysis of the doped Co–B catalysts, the H2
generation yield as a function of time was measured at different
solution temperatures by hydrolysis of alkaline NaBH4 (0.025 M)
3. Co–Cu–B catalyst: As with the above two catalysts, Co–Cu–B cat-
alyst also requires high dopant (Cu) concentration to have the
promoting effect. The catalytic activity increases with the
amount of Cu-dopant in Co–B alloy catalyst, it reaches maxi-
mum at the optimum concentration of
v
Cu = 35%, then it
solution using Co–Ni–B (
Cu–B ( Cu = 35%), Co–Cr–B (
Co–W–B ( W = 5%) powder catalysts as reported in Fig. 7a–f,
v
Ni = 15%), Co–Fe–B (
v
Fe = 35%), Co–
decreases. Copper, in fully metallic state, is unable to either
cause any electron transfer to active Co site or act as Lewis acid
site for OHꢀ ions absorption. However, Cu on the catalyst sur-
face acts as a promoter by increasing the specific surface area
by more than five times, when compared to Co–B catalyst, by
preventing particle agglomeration. According to the above
reported hydrolysis reaction mechanism [3], the initial reaction
occurs between BHꢀ4 ions and metal sites, which means that the
reaction rate is proportional to the number of available metal
sites for the absorption of BH4ꢀ ions in the solution. In this reac-
tion scheme, the Co–Cu–B catalyst powder, having a surface
area almost five times higher than that of Co–B, may provide
an ideal condition for BHꢀ4 ions absorption for catalytic hydroly-
sis reaction. However, a concentration of the Cu promoters
v
vCr = 4%), Co–Mo–B (
vMo = 5%), and
v
respectively. As expected, H2 generation rate increases with the
temperature. Using the Arrhenius plot (inset of Fig. 7) for the
hydrogen production rate, the activation energies are obtained
and summarized in Table 1. The results clearly show that in
Co–B catalyst doped with transition metal, the activation energy
is always lower when compared to undoped Co–B catalyst powder
(45 kJ molꢀ1) [16]. In general, the obtained activation energies are
lower than that obtained with Raney Co (53.7 kJ molꢀ1) [6], carbon
supported Co–B (57.8 kJ molꢀ1
)
and Co–B nanoparticles
(42.7 kJ molꢀ1) [27]. The values are also lower than that found by
Amendola (47 kJ molꢀ1) [4] by using Ru catalyst. Kaufman and
Sen [28], by using different bulk metal catalysts, obtained
75 kJ molꢀ1 for cobalt, 71 kJ molꢀ1 for nickel, and 63 kJ molꢀ1 for
Raney nickel. The activation energy values obtained in the present
case are comparable to that obtained with nanoparticle-assembled
Co–B thin film (30 kJ molꢀ1) [12], Pd/C powder (28 kJ molꢀ1) [29],
exceeding
vCu = 35% is deleterious to the hydrolysis reaction,
because major portion of the surface Co active sites would be
covered by the Cu metallic species.
4. Co–Cr–B, Co–Mo–B, and Co–W–B catalysts: These three powders
seem to be acting as a catalyst with a similar mechanism,
because in all the three cases, a small amount of dopant (Cr,
Mo, or W) is enough to create the promoting effect in the
Co–B catalyst. Even in these cases, by increasing the dopant
concentration, the activity first increases and then decreases.
The optimum contents of Cr-, Mo-, and W-promoters are:
Co supported on
c ) [9], Ru nanoclusters
-Al2O3 (33 kJ molꢀ1
(29 kJ molꢀ1) [5], and Ru-C (37 kJ molꢀ1) [30]. The favorable activa-
tion energy values obtained in the present work are attributed to
promoting effects, caused by the different transition metals, to en-
hance the catalytic hydrolysis reaction.
vCr = 4%, vMo = 5%, and vW = 5%, respectively. The promoting
effect is mainly caused by the corresponding oxide species of
the dopant transition metals (Cr, Mo, or W) which, on the cata-
lyst surface, inhibit the particle agglomeration thus increasing
the surface area of the Co–B catalyst by favoring a better disper-
sion of the particles. In addition, the Cr3+, Mo4+, and W4+ species
also act as Lewis acid sites for OHꢀ absorption due to the strong
affinity of these sites toward oxygen of the OHꢀ group. A polar-
ization of the O–H bond occurs which favors the reaction
between hydroxyl group and boron species attached to the
neighboring Co active sites. In conclusion, the oxide species pro-
vide high surface area and act as Lewis absorption sites for OHꢀ
ions: these are the two main reasons involved in the significant
enhancement of Rmax observed for Co–Cr–B, Co–Mo–B, and Co–
W–B catalysts with respect to that observed with the Co–B cat-
alyst. Cr-doped Co–B shows the highest Rmax when compared to
W- and Mo-doping: this is attributed to the superior capability
of Cr3+ species to absorb OHꢀ group ions. The deleterious effect
on catalytic activity at higher dopant content is attributed to
4. Conclusions
We have investigated the role of the dopant transition metals:
Ni, Fe, Cu, Cr, Mo, and W on the catalytic efficiency of the Co–B
powder for H2 generation by hydrolysis of NaBH4. Transition met-
als added to the Co–B catalyst behave in a different way while
influencing the catalytic activity. Cr, W, Mo, and Cu impose signif-
icant catalytic effects by increasing 3–4 times the H2 generation
rate when compared to the Co–B catalyst. On the contrary, Ni
and Fe cause only marginal increment in the catalytic performance
of the Co–B catalyst. The metal/(Co + metal) molar ratio was varied
in order to study the effect of each metal dopant concentration on
the catalytic efficiency. We observed that the best promoting ef-
fects are obtained with specific molar concentration of the transi-
tion metals. The promoting effects of the dopants are related to: (1)
large active surface area of the catalysts (Co–Fe–B, Co–Cu–B, Co–
Cr–B, Co–Mo–B, Co–W–B), (2) ability to act as Lewis acid sites for
better absorption of OHꢀ groups (Co–Cr–B, Co–Mo–B, Co–W–B),

324
N. Patel et al. / Journal of Catalysis 271 (2010) 315–324
[8] Y. Pei, P. Guo, M. Qiao, H. Li, S. Wei, H. He, K. Fan, J. Catal. 248 (2007)
303.
[9] W. Ye, H. Zhang, D. Xu, L. Ma, B. Yi, J. Power Sour. 164 (2007) 544.
[10] P. Krishnan, S. Advani, A.K. Prasad, Appl. Catal. B 86 (2009) 137.
[11] Dong-ge Tong, Wei Chu, Yong-yue Luo, Hong Chen, Xiao-yang Ji, J. Mol. Catal.
A: Chem. 269 (2007) 149.
(3) electronic interaction with Co active metal (Co–Ni–B, Co–Fe–B),
and (4) amorphous nature of the alloy catalyst (all investigated cat-
alysts). Finally, all the activation energies involved in the catalytic
hydrolysis of NaBH4 are comparable to the lowest values of the re-
ported literature catalysts.
[12] N. Patel, R. Fernandes, G. Guella, A. Kale, A. Miotello, B. Patton, C. Zanchetta, J.
Phys. Chem. C 112 (2008) 6968.
[13] H. Li, J.F. Deng, J. Chem. Technol. Biotechnol. 76 (2001) 985.
[14] R. Fernandes, N. Patel, A. Miotello, Appl. Catal. B: Environ. 92 (2009) 68.
[15] J. Fang, X. Chen, B. Liu, S. Yan, M. Qiao, H. Li, H. He, K. Fan, J. Catal. 229 (2005)
97.
[16] R. Fernandes, N. Patel, A. Miotello, M. Filippi, J. Mol. Catal. A: Chem. 298 (2009)
1.
[17] N. Patel, R. Fernandes, A. Miotello, J. Power Sour. 188 (2009) 411.
[18] R. Fernandes, N. Patel, A. Miotello, Int. J. Hydrogen Energy 34 (2009)
2893.
Acknowledgements
We thank N. Bazzanella for SEM–EDS analysis, C. Armellini for
XRD analysis, S. Dirè for BET measurements, and S. Torrengo for
XPS analysis. The research activity is financially supported by the
Hydrogen-FISR Italian project.
[19] D.A. Lyttle, E.H. Jensen, W.A. Struck, Anal. Chem. 24 (1952) 1843.
[20] C. Zanchetta, B. Patton, G. Guella, A. Miotello, Meas. Sci. Technol. 18 (2007)
N21.
References
[1] S.C. Amendola, S.L. Sharp-Goldman, M.S. Janjua, M.T. Kelly, P.J. Petillo, M.
Binder, J. Power Sour. 85 (2000) 186.
[2] Jaeyoung lee, Kyung Yong Kong, Chang Ryul Jung, Eunae Cho, Sung Pil Yoon,
Jonghee Han, Tai-Gyu Lee, Suk Woo Nam, Catal. Today 120 (2007) 305.
[3] G. Guella, C. Zanchetta, B. Patton, A. Miotello, J. Phys. Chem. B 110 (2006)
17024.
[4] S.C. Amendola, S.L. Sharp-Goldman, M.S. Janjua, N.C. Spencer, M.T. Kelly, P.J.
Petillo, M. Binder, Int. J. Hydrogen Energy 25 (2000) 969.
[5] S. Özkar, M. Zahmakiran, J. Alloys Compd. 404–406 (2005) 728.
[6] B.H. Liu, Z.P. Li, S. Suda, J. Alloys Compd. 415 (2006) 288.
[7] N. Patel, G. Guella, A. Kale, A. Miotello, B. Patton, C. Zanchetta, L. Mirenghi, P.
Rotolo, Appl. Catal. A: Gen. 323 (2007) 18.
[21] A. Baiker, Faraday Discuss. Chem. Soc. 87 (1989) 239.
[22] W.L. Dai, M.H. Qiao, J.F. Deng, Appl. Surf. Sci. 120 (1997) 119.
[23] B.H. Liu, Q. Li, Int. J. Hydrogen Energy 33 (2009) 7385.
[24] H. Li, W. Wang, H. Li, J.F. Deng, J. Catal. 194 (2000) 211.
[25] H. Li, Y. Wu, J. Zhang, W. Dai, M. Qiao, Appl. Catal. A: Gen. 275 (2004) 199.
[26] X. Chen, H. Li, H. Luo, M. Qiao, Appl. Catal. A: Gen. 233 (2002) 13.
[27] A. Garron, D. Swierczynski, S. Bennici, A. Auroux, Int. J. Hydrogen Energy 34
(2009) 1185.
[28] C.M. Kaufman, B. Sen, J. Chem. Soc., Dalton Trans. 2 (1985) 307.
[29] N. Patel, B. Patton, C. Zanchetta, R. Fernandes, G. Guella, A. Kale, A. Miotello, Int.
J. Hydrogen Energy 33 (2008) 287.
[30] Y. Shang, R. Chen, Energy Fuels 20 (2006) 2149.