Understanding the differences in catalytic performance for hydrogen production of Ni and Co supported on mesoporous SBA-15

Article abstract of DOI:10.1016/j.cattod.2017.02.020

Three mono and bimetallic Ni_xCo_1?x/SBA-15 catalysts (x = 1, 0.5 and 0) with a total metallic content of 10 wt% have been prepared by a deposition–precipitation (DP) method. The catalytic performances on the dry reforming of methane reaction (DRM) have been determined and correlated with their physical and chemical state before and after the catalytic reaction. So, while the nickel monometallic system presents a high activity and stability in the DRM reaction, the Co/SBA-15 catalytic system turns out completely inactive. For its part, the Ni_0.5Co_0.5/SBA-15 has initially a catalytic performance similar to the Ni/SBA-15 monometallic system, but rapidly evolving to an inactive system, therefore resembling the behavior of the cobalt-based catalyst. The characterization by TEM and in situ XPS techniques has allowed us to ascribe these differences to the initial state of metallic particles after reduction and their different evolution under reaction conditions. So, while after reduction both nickel containing Ni_xCo_1?x/SBA-15 catalysts (x = 1 and 0.5) present a well dispersed metallic phase, the cobalt monometallic catalyst yields big metallic particles with a heterogeneous distribution of sizes. Additionally, unlike the Ni/SBA-15, the NiCo/SBA-15 system increases during reaction the metallic particle sizes. Besides indicating that the particle size is a major reason determining the catalytic performances, these results suggest that in the Ni–Co system both metals form after reduction a bimetallic phase mainly located inside the mesoporous channels of SBA-15 support. Under DRM reaction conditions, the cobalt is segregated to the surface of the bimetallic particles, which seems to determine the interaction with the support surface SBA-15. This feature gives rise to a much less stable metallic phase which suffers an important sintering process under DRM catalytic conditions.

Full text of DOI:10.1016/j.cattod.2017.02.020

G Model
CATTOD-10609; No. of Pages7
ARTICLE IN PRESS
Catalysis Today xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Understanding the differences in catalytic performance for hydrogen
production of Ni and Co supported on mesoporous SBA-15
Alberto Rodriguez-Gomez, Rosa Peren˜iguez, Alfonso Caballero∗
Instituto de Ciencia de Materiales de Sevilla (CSIC-University of Seville) and Departamento de Quimica Inorganica, University of Seville, Avda. Américo
Vespucio, 49, 41092 Seville, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Three mono and bimetallic NixCo1−x/SBA-15 catalysts (x = 1, 0.5 and 0) with a total metallic content of
10 wt% have been prepared by a deposition–precipitation (DP) method. The catalytic performances on the
dry reforming of methane reaction (DRM) have been determined and correlated with their physical and
chemical state before and after the catalytic reaction. So, while the nickel monometallic system presents
a high activity and stability in the DRM reaction, the Co/SBA-15 catalytic system turns out completely
inactive. For its part, the Ni0.5Co0.5/SBA-15 has initially a catalytic performance similar to the Ni/SBA-
15 monometallic system, but rapidly evolving to an inactive system, therefore resembling the behavior
of the cobalt-based catalyst. The characterization by TEM and in situ XPS techniques has allowed us
to ascribe these differences to the initial state of metallic particles after reduction and their different
evolution under reaction conditions. So, while after reduction both nickel containing NixCo1−x/SBA-15
catalysts (x = 1 and 0.5) present a well dispersed metallic phase, the cobalt monometallic catalyst yields
big metallic particles with a heterogeneous distribution of sizes. Additionally, unlike the Ni/SBA-15, the
NiCo/SBA-15 system increases during reaction the metallic particle sizes.
Received 23 November 2016
Received in revised form 13 February 2017
Accepted 16 February 2017
Available online xxx
Keywords:
Nickel cobalt catalysts
In situ XPS
Hydrogen production
DRM methane reforming
Besides indicating that the particle size is a major reason determining the catalytic performances, these
results suggest that in the Ni–Co system both metals form after reduction a bimetallic phase mainly
located inside the mesoporous channels of SBA-15 support. Under DRM reaction conditions, the cobalt
is segregated to the surface of the bimetallic particles, which seems to determine the interaction with
the support surface SBA-15. This feature gives rise to a much less stable metallic phase which suffers an
important sintering process under DRM catalytic conditions.
© 2017 Elsevier B.V. All rights reserved.
1. Introduction
widely tested in the literature for this reforming reaction, even
though noble metal based catalysts such as Pt, Ru and Rh are much
Supported nickel catalytic systems are currently one of the
most important industrial heterogeneous catalysts because its
remarkable performance in a number of economically strategic
processes [1–5]. Among them, the steam reforming of methane
(SRM, CH4 + H2O ↔ 3H2 + CO) can be outlined as the main industrial
process for obtaining hydrogen and synthesis gas, used to syn-
thesize various important chemicals and fuels [6–9]. Although it
is not yet commercially exploited, the dry reforming of methane
(DRM, CH4 + CO2 ↔ 2H2 + 2CO) is an especially interesting reac-
tion that transforms two of the most harmful greenhouse gases,
methane and carbon dioxide, into a mixture of hydrogen and car-
bon monoxide [10–12]. Once again, Ni-based catalysts are the most
more performance toward methane conversion. The principal issue
comes from the fact that Ni typically undergoes severe deactivation
processes, mainly due to coke formation, but also due to sinter-
ing of the metallic phase, generating big metallic particles which
at the same time, favors the coke formation processes [13–17]. As
an alternative to overcome these issues, a number of publications
have shown as the use of bimetallic systems, as the combination of
nickel and cobalt modified the catalytic performance in steam and
dry reforming of methane [18–23]. But, depending on the support
or the preparation methods both effects, improvements and wors-
ening of the efficiency, have been reported. Main reasons explaining
these contradictory findings are probably related with differences
in the interaction of metals with support surface, which has been
recognized as an essential factor affecting the stability of metal
[24–26]. So, a strategy to avoid the growth of metallic particles is
the use of special supports, and in particular mesoporous supports.
∗
Corresponding author.
E-mail address: caballero@us.es (A. Caballero).
http://dx.doi.org/10.1016/j.cattod.2017.02.020
0920-5861/© 2017 Elsevier B.V. All rights reserved.
Please cite this article in press as: A. Rodriguez-Gomez, et al., Understanding the differences in catalytic performance for hydrogen
production of Ni and Co supported on mesoporous SBA-15, Catal. Today (2017), http://dx.doi.org/10.1016/j.cattod.2017.02.020

G Model
CATTOD-10609; No. of Pages7
ARTICLE IN PRESS
2
A. Rodriguez-Gomez et al. / Catalysis Today xxx (2017) xxx–xxx
These supports typically have high surface area and a tailored size
of the porosity, which can be used to control the size of the metallic
particles [27–29].
Typically, the experiment was carried out from room temperature
up to 1000 ◦C using a heating ramp of 10 ◦C min−1
.
In this work we have prepared three catalysts containing
nickel and/or cobalt supported on the mesoporous silica SBA-15.
The catalytic systems, with a total metal loading of 10 wt%, has
been prepared by means of a deposition–precipitation method
(NixCo1−x/SBA-15 catalysts, x = 1, 0.5 and 0). As the catalytic behav-
ior of these systems are pretty different, we have characterized the
catalytic systems after hydrogen reduction and after dry reform-
ing reaction conditions using among others, in situ XPS and TEM.
The obtained results have allowed us to correlate the catalytic
performances with both, the initial state of metallic phases after
reduction, and the final state of these metallic particles after DRM
reaction. The interaction of the metallic phase with the SBA-15
inner surface appears as a major factor determining the catalytic
stability of these NixCo1−x/SBA-15 catalysts.
2.4. Transmission electronic microscopy (TEM)
TEM images were obtained in a Philips CM200 microscope oper-
ating at 200 kV. Samples were dispersed on ethanol and deposited
onto a copper grid coated with a lacey carbon film. Histograms for
particle size distribution were obtained by sampling around 150
particles.
2.5. X-ray absorption spectroscopy (XAS)
XAS (EXAFS and XANES regions) were recorded at the BL22
beamline (CLAESS) of ALBA synchrotron and the BM25A beam-
line (SPLINE) of ESRF synchrotron facilities. The required amount of
sample to maximize the signal/noise ratio was pelletized and XAS
spectra collected at RT after calcination. Standards Ni and Co foils
were measured and used for energy calibration. The XAS spectra of
Ni K-edge and Co K-edge were recorded for a 1000 eV interval, with
a step of 0.5 eV step across the XANES region. Once extracted from
the XAS spectra, the EXAFS oscillations were Fourier transformed
2. Experimental
2.1. Catalysts preparation
(F.T.) in the range 2.4–11.0 A−1. Spectra were analyzed using the
˚
The mesoporous silica SBA-15 used as support was prepared
according to a method previously described in the literature
[30,31]. Summarizing, 18 g of P123 was dissolved in 270 mL of
distilled water and afterwards a solution of 675 mL of HCl 2.5 M
was added and heated up to 50 ◦C. Next, an amount of TEOS was
added and kept at 50 ◦C for 18 h. The product obtained was filtered,
washed with boiling distilled water, dried under vacuum at 70 ◦C
and calcined on for 3 h at 550 ◦C.
The cobalt and nickel metal phases were supported on SBA-15
by a deposition–precipitation method (DP) using urea as a pre-
cipitant agent according to the method described by Liu et al.
[32]. 1 g of calcined SBA-15 was dispersed in HNO3 0.01 M with
the required amount of metal precursors Ni(NO3)2·6H2O and/or
Co(NO3)3·6H2O. On stirring, an amount of urea was added and the
temperature increased up to 105 ◦C. After 2 h, the mix was cooled
down to room temperature and the powder filtered and dried at
110 ◦C for 24 h. Finally, it was submitted to a calcination treatment
on air at 550 ◦C. The resulting products were labeled as 10%Ni/SBA-
15, 10%Co/SBA-15 and 5%Ni–5%Co/SBA-15. In order to clarify the
effect of the preparation treatment, a similar procedure was accom-
plished without metals addition, and the sample was labeled as
SBA-15-DP.
software package IFEFFIT [34].
2.6. X-ray photoelectron spectroscopy (XPS)
XPS experiments were carried out in a VG-ESCALAB 210
equipment over pelletized samples. Samples were introduced
in a pre-chamber at 10−7 Torr. Acquisition was performed in
an appendant analysis chamber equipped with a SPECS Phoibos
100 hemispheric analyzer at 10−9 Torr using Mg K␣ radiation
(E = 1.5418 keV) with 20 mA of anode current and 12 kV of poten-
tial acceleration. Before acquisition, each sample was treated in situ
at different temperatures (RT, 350, 500 and 750 ◦C) in a flow of 5%
H2/Ar at atmospheric pressure using a cell chamber attached to the
abovementioned pre-chamber.
2.7. Catalytic activity tests
Dry reforming of methane (DRM) tests were performed using
20 mg of catalyst held in a tubular quartz reactor through wool
quartz. The catalytic systems were pre-treated in 5% H2/Ar at 750 ◦C
for 1 h with a heating ramp of 10 ◦C min−1. Reaction was carried out
with a not diluted mix (1:1) of 40 mL min−1 of CH4 and 40 mL min−1
of CO2 at 750 ◦C during 42 h. Reaction products were analyzed by
GC using an Agilent’s 490 microGC equipped with three micro-
columns and TCD detectors.
2.2. X-ray diffraction (XRD)
The XRD analysis of calcined and reduced samples were
recorded in a PANalytical X-Pert PRO diffractometer with a Cu
source (ꢀ=1.5418 A, Cu K␣), working in a Bragg-Brentano config-
uration and equipped with an X’Celerator Detector (active range
of 2 = 2.18◦). The data acquisition was carry out in a 2Â range of
10–80◦, a step of 0.05◦ and an acquisition time of 240 s.
3. Results and discussion
˚
3.1. Characterization of fresh calcined systems
The three catalytic systems were characterized by N2 adsorp-
tion analysis (BET, BJH), XRD, TEM, TPR, XAS and XPS. As expected
for mesoporous materials, all samples exhibited type-IV adsorp-
tion isotherms. However, as reflected in Table 1, they undergo a
significant change in BET surface area and mean porous size as
a consequence of the alkaline treatment during the DP prepara-
tion methodology. Every sample, including a blank SBA-15 support
submitted to the DP treatment without metal salts in solution (SBA-
15-DP), suffers a decrease in surface area of around 65%, from
738 m2 g−1 of the pristine SBA-15 to 250–280 m2 g−1. Simultane-
ously, the smaller porosity collapses (mean size increasing from 6.7
to 10–12 nm). Despite this loss of surface area, the analysis by TEM
(Fig. 1) shows that the channeled structure of the SBA-15 support
2.3. Temperature programmed reduction (TPR)
The temperature-programmed reduction profiles were
obtained using
a thermal conductivity detector based in a
Wheatstone bridge. The hydrogen consumption was previously
calibrated using commercial CuO. A 5% H2/Ar calibrated mix was
used as both carrier and reference gas, with a typical flow rate of
50 mL min−1. Following the recommendation to avoid peak coa-
lescence [33], the experimental conditions were chosen to assure
a total hydrogen consumption of approximately 100 ␮mol of H2.
Please cite this article in press as: A. Rodriguez-Gomez, et al., Understanding the differences in catalytic performance for hydrogen
production of Ni and Co supported on mesoporous SBA-15, Catal. Today (2017), http://dx.doi.org/10.1016/j.cattod.2017.02.020

G Model
CATTOD-10609; No. of Pages7
ARTICLE IN PRESS
A. Rodriguez-Gomez et al. / Catalysis Today xxx (2017) xxx–xxx
3
Table 1
Specific surface areas and average pore size and average crystallite size of Ni/Co after calcination, reduction and DRM treatment.
SBET (m2 g−1
[Calc/red]
)
Average pore size
(nm)a [Calc/red]
Average crystallite size
Average particle size
(XRDb) [Calc /Red]
(TEMc) [Red/DRM]
SBA-15
738
6.7
–
–
SBA-15-DP
249
9.1
–
10% Ni/SBA-15
5–5% Ni–Co/SBA-15
10% Co/SBA-15
283/242
261/232
280/220
10.3/10.1
12.0/11.1
11.9/11.0
5.7/4.1
7.0/4.5
5.1/13.2
4.3/4.8
5.9/7.7
14.4/13.9
a
Obtained by the BJH method.
Calculated from the Scherrer equation for reduced samples.
Obtained by sampling 150 particles.
b
c
Fig. 1. TEM images for calcined 10% Ni/SBA-15 (a), 5%–5% Ni–5% Co/SBA-15 (b) and 10% Co/SBA-15 (c).
Fig. 2. XRD patterns of catalytic systems calcined (a) and treated in 5% H2/Ar at 750 ◦C (b).
is preserved [35]. The calcined NixCo1−x/SBA-15 catalysts present
the characteristic “fibrous” structures of phyllosilicate phases, pre-
viously found in similar systems [36–40]. The ascription of these
fibrous shapes to phyllosilicate is confirmed by the XRD diagrams
included in Fig. 2a. According to the database and previous works
[38], the wide peaks appearing around 35◦ and 61◦ correspond
to nickel and/or cobalt silicate hydroxide. In no case, peaks from
Ni or Co oxides phases could be detected. It is worthy to note
the intermediate position at about 60.5◦ observed for the bimetal-
lic Ni0.5Co0.5/SBA-15 catalysts, which can be an indication of the
formation of a bimetallic nickel cobalt phyllosilicate phase. The for-
mation of these silicates in the calcined samples is also confirmed
by XAS spectroscopy. The XANES spectrum and Fourier Transforms
of EXAFS regions included in Fig. 3 clearly differ from those of
Please cite this article in press as: A. Rodriguez-Gomez, et al., Understanding the differences in catalytic performance for hydrogen
production of Ni and Co supported on mesoporous SBA-15, Catal. Today (2017), http://dx.doi.org/10.1016/j.cattod.2017.02.020

G Model
CATTOD-10609; No. of Pages7
ARTICLE IN PRESS
4
A. Rodriguez-Gomez et al. / Catalysis Today xxx (2017) xxx–xxx
Fig. 3. XANES spectra (left) and Fourier Transform functions of the EXAFS oscillations (right) of Ni K-edge (a) and Co K-edge (b) of calcined systems.
processes of the three catalytic systems. The Ni/SBA-15 sample
presents a main peak centered at 660 ◦C, previously assigned to a
nickel phyllosilicate phase [40,41], with a small shoulder at lower
temperature, around 400 ◦C. According to previous results obtained
by our group, these two reduction features can be respectively
ascribed to the reduction of oxidized nickel located inside and out-
side the mesoporous channels of SBA-15 support [40]. For its part,
the bimetallic NiCo/SBA-15 catalyst presents a similar profile, but
shifted to higher temperature (peaks maximum at 700 and 500 ◦C,
respectively). These similitudes suggest that also in this system
the metals could be located both in the inner and outer surface
of support. The formation of a bimetallic silicate with a stronger
interaction with the support can explain the higher reduction tem-
perature observed by TPR. Finally, as expected [43], the Co/SBA-15
catalyst reduces at higher temperature, presenting at least two
peaks centered at 750 and 850 ◦C.
Fig. 4. TPR profiles of the nickel and cobalt catalytic systems.
The metallic particles obtained after the reduction treatment
can be visualized in Fig. 5. Both nickel containing samples,
Ni/SBA-15 (Fig. 5a) and NiCo/SBA-15 (Fig. 5b) generate small and
typical nickel and cobalt oxides, and resemble those previously
identified as silicates [3,41,42].
homogeneous nickel particles of about 4–6 nm (Table 1), slightly
biggerfor the bimetalliccatalyst. From the visual inspectionof these
3.2. Characterization of reduced catalytic systems
Fig. also includes the diffraction patterns of the three
two TEM images, it is evident that many metal particles appear
aligned with the mesoporous channels of support, once again sug-
gesting that they could be located inside the mesoporous [40]. On
the contrary, the TEM image of Co/SBA-15 (Fig. 5c) shows bigger
heterogeneous particles with a mean diameter around 13–15 nm
(Table 1), necessarily located outside the mesopores of support.
Additionally, the image shows the presence of much smaller parti-
cles.
The reduction processes of these three SBA-15 supported cata-
lysts have been studied by means of in situ XPS spectroscopy under
a hydrogen atmosphere at 350, 500 and 750 ◦C. Fig. 6 includes the
XPS obtained for the monometallic Ni/SBA-15 (Ni 2p region) and
2
NixCo1−x/SBA-15 catalysts after a reduction treatment in hydro-
gen at 750 ◦C. In all cases, the diagrams show new diffraction peaks
around 44.5◦ and 52◦ that can be ascribed to reduced nickel and/or
cobalt phases. As detailed in Fig. 2b, the position and intensity
of peak around 44–45◦ corresponding to (111) reflection of the
cubic Ni and/or Co phases, slightly differs between samples: 44.3◦
for reduced Co/SBA-15, 44.6◦ for Ni/SBA-15, and an intermediate
value for the NiCo/SBA-15 catalyst, which once again suggests the
formation of a Ni–Co bimetallic phase. The TPR profiles included
in Fig. 4 show also interesting differences between the reduction
Please cite this article in press as: A. Rodriguez-Gomez, et al., Understanding the differences in catalytic performance for hydrogen
production of Ni and Co supported on mesoporous SBA-15, Catal. Today (2017), http://dx.doi.org/10.1016/j.cattod.2017.02.020

G Model
CATTOD-10609; No. of Pages7
ARTICLE IN PRESS
A. Rodriguez-Gomez et al. / Catalysis Today xxx (2017) xxx–xxx
5
Fig. 5. TEM images and particle size histograms of 10% Ni/SBA-15 (a), 5%–5% Ni–5% Co/SBA-15 (b) and 10% Co/SBA-15 (c), reduced in H2 at 750 ◦C.
Fig. 6. XPS spectra for calcined and in situ reduced 10% Co/SBA-15 (a) and 10% Ni/SBA-15 (b).
Fig. 7. XPS spectra for calcined and in situ reduced 5%–5% Ni–5% Co/SBA-15.
Please cite this article in press as: A. Rodriguez-Gomez, et al., Understanding the differences in catalytic performance for hydrogen
production of Ni and Co supported on mesoporous SBA-15, Catal. Today (2017), http://dx.doi.org/10.1016/j.cattod.2017.02.020

G Model
CATTOD-10609; No. of Pages7
ARTICLE IN PRESS
6
A. Rodriguez-Gomez et al. / Catalysis Today xxx (2017) xxx–xxx
Co/SBA-15 (Co 2p region), while the Fig. 7 gathers the respective
regions of the bimetallic NiCo/SBA-15 catalysts (Ni 2p and Co 2p
regions). In all cases, the reduction treatment at 350 ◦C does not
provoke noticeable changes while the use of the higher reduc-
tion temperature (750 ◦C) gives rise in all catalysts to well reduced
metallic nickel and cobalt phases. However, interesting differences
of behavior can be observed after reduction in hydrogen at 500 ◦C.
At this temperature nickel appears completely reduced in Ni/SBA-
15 and NiCo/SBA-15 catalysts. As the TPR profiles in Fig. 4 indicated
that after this treatment most of metal should remain oxidized, the
undetected oxidized nickel must be located inside the mesoporous
channels in both catalysts, which are invisible for XPS, a surface sen-
sitive technique. So, as found previously for different preparation
methods [40], the deposition–precipitation methodology seems
to favor the confinement of nickel inside the porous of SBA-15.
Regarding the state of cobalt, in the monometallic Co/SBA-15 cat-
alysts it remains basically unreduced after hydrogen treatment at
500 ◦C, as can be expected from the TPR of Fig. 4. However, from the
Co 2p XPS signal, it can be estimated that a small fraction of about
10% is reduced to the metallic state. Taking into account that XPS
only detects the metal located in the external surface of SBA-15,
the reduced fraction at 500 ◦C must be located in the outer surface
of SBA-15. More interesting is the fact that cobalt is much more
reducible in the bimetallic NiCo/SBA-15 catalyst. A similar estima-
tion of Co 2p XPS signal of Fig. 7, allows estimating that around
65% of cobalt is reduced to the metallic state at 500 ◦C. According
to TPR in Fig. 4, only the shoulder at low temperature is affected
by this treatment at 500 ◦C, which represents less than 5% of total
cobalt content. Once again, considering the surface sensitivity of
XPS, it can be concluded that in the bimetallic NiCo/SBA-15 cata-
lyst most of the cobalt particles must be located into the channels
of the SBA-15 support, which remain oxidized but are invisible for
XPS.
Fig. 8. CH4 and CO2 conversions in the reaction of DRM at 750 ◦C of the nickel and
cobalt catalytic systems.
3.3. Catalytic studies
The catalytic activity of the three catalytic systems was stud-
ied under dry reforming reaction of methane (DRM) conditions. As
depicted in Fig. 8, the catalytic performances are completely differ-
ent depending on the metals presented in the catalysts. Face to the
outstanding activity and stability of Ni/SBA-15, which maintains its
catalytic properties during the 48 h period of DRM reaction, Co/SBA-
15 is completely inactive with absolutely no conversion of methane
from the very beginning of reaction.
In a certain way, the bimetallic NiCo/SBA-15 presents a dual cat-
alytic behavior. The initial activity is not much less than half of the
initial Ni/SBA-15 activity (30% vs. 73% conversion of methane), as
could be expected considering the relative nickel content of these
two catalysts (10% and 5%, respectively) and the slightly higher
particle size in the bimetallic NiCo/SBA-15 catalyst after reduction
(Fig. 5). But the bimetallic system loses completely the activity after
Fig. 9. TEM images and particle size histograms of 10% Ni/SBA-15 (a), 5%–5% Ni–5% Co/SBA-15 (b) and 10% Co/SBA-15 (c), after DRM reaction at 750 ◦C.
Please cite this article in press as: A. Rodriguez-Gomez, et al., Understanding the differences in catalytic performance for hydrogen
production of Ni and Co supported on mesoporous SBA-15, Catal. Today (2017), http://dx.doi.org/10.1016/j.cattod.2017.02.020

G Model
CATTOD-10609; No. of Pages7
ARTICLE IN PRESS
A. Rodriguez-Gomez et al. / Catalysis Today xxx (2017) xxx–xxx
7
5 h on stream, resembling now the catalytic behavior of Co/SBA-15
monometallic catalyst.
Competitividad” of Spain (Projects ENE2011-24412 and CTQ2014-
60524-R) for funding this research. Also, we thank the ESRF facility
and staff (BM25 SPLINE beamline) and ALBA facility and staff (BL22
CLAESS beamline) for their experimental support. A.R.G thanks
also the Spanish Government for the PhD fellowship (BES-2012-
061744).
After the DRM reaction, the catalysts present the appearance
showed in the TEM images of Fig. 9. The metallic particles essen-
tially maintain their size and distribution of the freshly reduced
Ni/SBA-15 (4–6 nm) and Co/SBA-15 (13–15 nm) catalysts (see
Table 1, Figs. 5a, 5c and 9a, 9c, respectively), which can be related
with the stable catalytic performance of both catalysts over time
(Fig. 8). In particular, the high stability of nickel agrees with that
previously observed by other authors in similar systems [37]. On
the contrary the image in Fig. 9b clearly shows that metallic par-
ticles of bimetallic NiCo/SBA-15 have increased their mean size,
from 6 nm of the freshly reduced catalyst to about 8 nm after
DRM reaction. Once again this behavior seems to be correlated
with the observed evolution in the catalytic performance, initially
similar to that of monometallic nickel one, and finally reaching
that of the cobalt monometallic system. This increase in size indi-
cates that the interaction with support of bimetallic NiCo/SBA-15
and monometallic Ni/SBA–15 differs, strong in the monometallic,
which maintains the nickel particle size, and much less intense in
the bimetallic system. This different behavior suggests that after
reduction of the bimetallic catalyst the nickel is responsible for the
interaction with support (which assures tiny metallic particles in
both catalyst), but under reaction conditions cobalt is preferentially
segregated to the surface of bimetallic particles, which weakens the
interaction with support, favoring the sintering processes.
References
[1] D. Pakhare, J. Spivey, Chem. Soc. Rev. 43 (2014) 7813–7837.
[2] V.M. Gonzalez-Delacruz, R. Peren˜iguez, F. Ternero, J.P. Holgado, A. Caballero,
ACS Catal. 1 (2011) 82–88.
[3] J.P. Espinos, A.R. Gonzalez-Elipe, A. Caballero, J. Garcia, G. Munuera, J. Catal.
136 (1992) 415–422.
[4] C. Batiot-Dupeyrat, G. Sierra Gallego, F. Mondragon, J. Barrault, J.M. Tatibouet,
Catal. Today 107-108 (2005) 474–480.
[5] J.R. Rostrup-Nielsen, R. Nielsen, Catal. Rev. 46 (2004) 247–270.
[6] J.R. Rostrup-Nielsen, J. Sehested, J.K. Norskov, Adv. Catal. 47 (2002) 65–139.
[7] Y. Matsumura, T. Nakamori, Appl. Catal. A: Gen. 258 (2004) 107–114.
[8] W. Dong, H. Roh, K. Jun, S. Park, Y. Oh, Appl. Catal. A: Gen. 226 (2002) 63–72.
[9] J.N. Armor, Appl. Catal. A: Gen. 176 (1999) 159–176.
[10] C. Batiot-Dupeyrat, G. Valderrama, A. Meneses, F. Martinez, J. Barrault, J.M.
Tatibouet, Appl. Catal. A: Gen. 248 (2003) 143–151.
[11] Y.H. Hu, E. Ruckenstein, Adv. Catal. 48 (2004) 297–345.
[12] M.O. Moctar, V.M. Gonzalez-DelaCruz, G. Colón, S. Sebti, A. Caballero, Appl.
Catal. B: Environ. 150 (2014) 459–465.
[13] X. Du, D. Zhang, L. Shi, R. Gao, J. Zhang, Nanoscale 5 (2013) 2659–2663.
[14] L. Smolakova, M. Kout, L. Capek, A. Rodriguez-Gomez, V.M. Gonzalez-Delacruz,
L. Hromadko, A. Caballero, Int. J. Hydrog. Energy 41 (2016) 8459–8469.
[15] D. Chen, K.O. Christensen, E. Ochoa-Fernandez, Z.X. Yu, B. Totdal, N. Latorre, A.
Monzon, A. Holmen, J. Catal. 229 (2005) 82–96.
[16] S. Tang, L. Ji, J. Lin, H.C. Zeng, K.L. Tan, K. Li, J. Catal. 194 (2000) 424–430.
[17] D.P. Liu, X.-Y. Quek, H.H.A. Wah, G.M. Zeng, Y.D. Li, Y.H. Yang, Catal. Today 148
(2009) 243–250.
4. Conclusions
[18] V.M. Gonzalez-Delacruz, R. Peren˜iguez, F. Ternero, J.P. Holgado, A. Caballero, J.
Phys. Chem. C 116 (2012) 2919–2926.
In this work we have studied the state and catalytic per-
[19] J. Zhang, H. Wang, A.K. Dalai, J. Catal. 249 (2007) 300–310.
[20] K. Takanabe, K. Nagaoka, K. Nariai, K.I. Aika, J. Catal. 232 (2005) 268–276.
[21] L. Wang, D. Li, M. Koike, H. Watanabe, Y. Xu, Y. Nakagawa, K. Tomishige, Fuel
112 (2013) 654–661.
[22] C. Yu, W. Weng, Q. Shu, X. Meng, B. Zhang, X. Chen, X. Zhou, J. Nat. Gas Chem.
20 (2011) 135–139.
formance of three NixCo1−x/SBA-15 catalysts prepared by
a
deposition–precipitation method. Combining the physical and
chemical state of these systems with the catalytic behavior, it has
been established a clear correlation between the outstanding cat-
alytic performance of nickel with the strong interaction of 5 nm
metallic particles mainly located into the mesoporous channels of
SBA-15, which remain unchanged during the 48 h period under
DRM reaction conditions. On the contrary, upon hydrogen reduc-
tion the monometallic cobalt catalyst generates bigger metallic
particles, mainly located on the outer surface of the SBA-15 support
and inactive for DRM reaction.
[23] C.R.B. Silva, L. Conceicao, N.F.P. Ribeiro, M.M.V.M. Souza, Catal. Commun. 12
(2011) 665–668.
[24] V.M. Gonzalez-Delacruz, J.P. Holgado, R. Peren˜iguez, A. Caballero, J. Catal. 257
(2008) 307–314.
[25] J.A.C. Dias, J.M. Assaf, Catal. Today 85 (2003) 59–68.
[26] N. Wang, Z.X. Xu, J. Deng, K. Shen, X.P. Yu, W.Z. Qian, W. Chu, F. Wei,
ChemCatChem 6 (2014) 1470–1480.
[27] A.J. Vizcaíno, A. Carrero, J.A. Calles, Fuel Process. Technol. 146 (2016) 99–109.
[28] U. Oemar, Y. Kathiraser, L. Mo, X.K. Hoa, S. Kawi, Catal. Sci. Technol. 6 (2016)
1173–1186.
Interestingly, the bimetallic NiCo/SBA-15 catalyst does not
present an intermediate catalytic performance but a dual behavior.
Presenting an initial catalytic activity similar to that of monometal-
lic nickel system, it quickly evolves to an inactive system, as is the
case for the monometallic Co/SBA-15. This can be explained accept-
ing that the bimetallic cobalt nickel silicate formed after calcination
mostly behaves as the monometallic nickel silicate, in the sense that
it is mainly placed in the inner porous surface of the SBA-15 sup-
port, which yields after reduction small bimetallic particles into the
mesoporous channels. However, it sinters during reaction show-
ing that, unlike the nickel monometallic system, these bimetallic
particles present a weak interaction with the SBA-15 support, and
consequently, this interaction is mainly determined by cobalt. As
a general conclusion, these findings suggest that cobalt should be
segregated to the surface of these bimetallic entities, which inter-
act weakly with the support, increasing their particle size under the
harsh conditions of DRM reaction.
[29] J.P. Thielemann, F. Girgsdies, R. Schlögl, C. Hess, Beilstein, J. Nanotechnol. 2
(2011) 110–118.
[30] A.I. Acatrinei, M.A. Hartl, J. Eckert, E.H.L. Falcao, G. Chertkov, L.L. Daemen, J.
Phys. Chem. C 113 (2009) 15634–15638.
[31] D.Y. Zhao, J.L. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D.
Stucky, Science 279 (1998) 548–552.
[32] H. Liu, H. Wang, J. Shen, Y. Sun, Z. Liu, Appl. Catal. A: Gen. 337 (2008) 138–147.
[33] P. Malet, A. Caballero, Chem. Soc. Faraday Trans. 84 (1988) 2369–2375.
[34] M. Newville, J. Synchrotron Radiat. 8 (2001) 322.
[35] J.F. Li, C. Xia, C.T. Au, B.S. Liu, Int. J. Hydrog. Energy 39 (2014) 10927–10940.
[36] X. Kong, Y. Zhu, H. Zheng, X. Li, Y. Zhu, Y.-W. Li, ACS Catal. 5 (2015) 5914–5920.
[37] A. Albarazi, M.E. Gálvez, P. Da Costa, Catal. Commun. 59 (2015) 108–112.
[38] M.E. Gálvez, A. Albarazi, P. Da Costa, Appl. Catal. A: Gen. 504 (2015) 143–150.
[39] C. Zhang, H. Yue, Z. Huang, S. Li, G. Wu, X. Ma, J. Gong, ACS Sustain. Chem. Eng.
1 (2012) 161–173.
[40] A. Rodriguez-Gomez, A. Caballero, ChemNanoMat 3 (2017) 94–97.
[41] P. Burattin, M. Che, C. Louis, J. Phys. Chem. B 101 (1997) 7060–7074.
[42] I. Arcon, N. Novak Tusar, A. Ristic, A. Kodre, A. Kaucic, Phys. Scr. T115 (2005)
810–812.
[43] J. Grams, A. Ura, W. Kwapinski, Fuel 122 (2014) 301–309.
Acknowledgments
We thank the FEDER program and the “Ministerio de Economia y
Please cite this article in press as: A. Rodriguez-Gomez, et al., Understanding the differences in catalytic performance for hydrogen
production of Ni and Co supported on mesoporous SBA-15, Catal. Today (2017), http://dx.doi.org/10.1016/j.cattod.2017.02.020