Rh-Ni/MgAl2O4 catalyst for steam reforming of methane: Effect of Rh doping, calcination temperature and its application on metal monoliths

Article abstract of DOI:10.1016/j.apcata.2018.11.021

The effect of adding small amounts Rh noble metal to 15 wt.% Ni/MgAl₂O₄ catalyst were studied for the steam reforming of methane. For this purpose, two series of catalysts were prepared and then the most active catalyst from this study was selected for metal monolith application. Both series of catalysts were tested at ambient and high pressures conditions. The first series of catalyst was used to determine the optimum Rh concentration by varying the Rh concentration from 0.1 to 1.0 wt.%. It was found that an optimum concentration of 0.5 wt.% Rh showed the highest activity and stability. The second series of catalysts containing 0.5 wt.%Rh-15 wt.% Ni/MgAl₂O₄ was used to study the effect of calcination temperature on the activity and stability. Studies on the second series of catalysts showed that the catalyst prepared by sequential impregnation and calcined at lower temperatures (600 °C after each Ni and Rh impregnation) showed the highest activity and stability. Among the several changes in factors detected due to Rh doping, the increase in catalytic activity for the catalyst calcined at lower temperatures was mainly attributed to its higher degree of reduction and higher abundance of Ni-Rh active sites. The most active Rh-Ni/MgAl₂O₄ catalyst was washcoated on FeCralloy metal monoliths and the performance compared with a packed bed and Ni/MgAl₂O₄ washcoated metal monolith. A significant increase of 24% in methane conversion was observed for Rh-Ni/MgAl₂O₄ washcoated metal monolith in comparison to NiMgAl₂O₄ washcoated metal monolith.

Full text of DOI:10.1016/j.apcata.2018.11.021

Accepted Manuscript
Title: Rh-Ni/MgAl2O4 catalyst for steam reforming of
methane: Effect of Rh doping, calcination temperature and its
application on metal monoliths
Authors: Sanjay Katheria, Goutam Deo, Deepak Kunzru
PII:
S0926-860X(18)30576-3
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2018.11.021
APCATA 16890
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
8 August 2018
15 October 2018
23 November 2018
Please cite this article as: Katheria S, Deo G, Kunzru D, Rh-Ni/MgAl2O4 catalyst
for steam reforming of methane: Effect of Rh doping, calcination temperature
and its application on metal monoliths, Applied Catalysis A, General (2018),
https://doi.org/10.1016/j.apcata.2018.11.021
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Rh-Ni/MgAl2O4 catalyst for steam reforming of methane: Effect
of Rh doping, calcination temperature and its application on
metal monoliths
Sanjay Katheria, Goutam Deo*, Deepak Kunzrua
Department of Chemical Engineering, Indian Institute of Technology Kanpur, Kanpur-208016, Uttar
Pradesh, India
*Corresponding author:
Goutam Deo
E-mail: goutam@iitk.ac.in
Telephone No.: +91 512-2507363
aPresent affiliation: School of Engineering & Applied Science
Ahmedabad University, Navrangpura, Ahmedabad-380009
Graphical abstract
1

Highlights
 An optimum amount of Rh to be added to 15 wt.% Ni/MgAl2O4 is 0.5 weight percent
 Optimum amount of Rh is based on activity and stability tests at ambient and high
pressure conditions
 Catalyst performance can be further improved by changing calcination temperatures
 Application of Rh-Ni/MgAl2O4 washcoated metal monolith further improves the process
efficiency.
Abstract
The effect of adding small amounts Rh noble metal to 15wt.% Ni/MgAl2O4 catalyst were studied
for the steam reforming of methane. For this purpose, two series of catalysts were prepared and
then the most active catalyst from this study was selected for metal monolith application. Both
series of catalysts were tested at ambient and high pressures conditions. The first series of
2

catalyst was used to determine the optimum Rh concentration by varying the Rh concentration
from 0.1 to 1.0 wt.%. It was found that an optimum concentration of 0.5 wt.% Rh showed the
highest activity and stability. The second series of catalysts containing 0.5wt.%Rh-15wt.%
Ni/MgAl2O4 was used to study the effect of calcination temperature on the activity and stability.
Studies on the second series of catalysts showed that the catalyst prepared by sequential
o
impregnation and calcined at lower temperatures (600 C after each Ni and Rh impregnation)
showed the highest activity and stability. Among the several changes in factors detected due to
Rh doping, the increase in catalytic activity for the catalyst calcined at lower temperatures was
mainly attributed to its higher degree of reduction and higher abundance of Ni-Rh active sites.
The most active Rh-Ni/MgAl2O4 catalyst was washcoated on FeCralloy metal monoliths and the
performance compared with a packed bed and Ni/MgAl2O4 washcoated metal monolith. A
significant increase of 24% in methane conversion was observed for Rh-Ni/MgAl2O4
washcoated metal monolith in comparison to NiMgAl2O4 washcoated metal monolith.
Keywords: Methane steam reforming; Rh-Ni/MgAl2O4 catalyst; Metal
Monoliths
3

1. Introduction
Hydrogen has the highest energy content in comparison to other conventional fuels and offers a
clean and environmentally friendly solution for future energy requirement [1]. Despite the
continuous improvements in technologies for producing hydrogen, steam reforming of natural
gas using conventional packed bed reactors is still widely used [2]. Several commercial catalysts
with different active metal compositions and supports are used for this large scale steam
reforming process and the details are available elsewhere [3]. Nickel, due to its low cost and
availability, and MgAl2O4 spinel, due to its stability at elevated temperatures, is extensively used
as the active metal and support, respectively. However, due to the propensity of nickel metal
towards carbon formation and sintering at high temperatures, its application for longer run-time
operations in conventional packed bed reactors, particularly at high pressures conditions, is a
challenging task [4]. Furthermore, the conventional packed bed steam reforming process also
suffers from other shortcomings, such as high pressure drop, lower catalyst utilization and heat
transfer limitations.
Several studies have reported the application of noble metals for steam reforming of methane to
address the shortcomings of conventional catalysts [5-7]. However, the high cost of noble metals
precludes their use as supported monometallic catalysts in industrial reforming. Fortunately,
several studies have also shown that introducing small amounts of noble metals, such as Pt, Rh,
Ag, and Ru can significantly improve the activity and stability of the nickel-based bimetallic
catalysts [8, 9]. In their work on reforming of n-butane, Ferrandon and co-workers found that for
the same metal loading, Rh doped nickel catalyst performed even better than nickel and Rh alone
[10]. Doping nickel catalysts with noble metals not only increases the resistance to carbon
formation but also decreases the affinity of nickel towards sulfur poisoning during steam
4

reforming of hydrocarbons [11]. Studies have also reported that the introduction of small
amounts of noble metals increases the self- activation and self- regeneration of the catalyst,
which is particularly beneficial for operations involving frequent startup and shutdown [12, 13].
Addition of small amounts of these noble metals significantly changes the physicochemical
properties of the catalyst, such as metal dispersion, crystallite size of active metal, and metal
support interaction [14]. A study has shown that the enhanced properties of noble metal-doped
Ni catalysts could be also be due to alloy formation, such as Pt-Ni or Rh-Ni alloys, resulting in
enhanced electronic interaction between metal particles [15]. Furthermore, Garcia-Dieguez et al.
found that on alumina supported Pt-Ni bimetallic catalysts, the addition of Pt also decreased the
formation of nickel aluminate, which is a non-reducible inert species [16].
A few studies have been published on the effect of varying the amount of noble metal doped on
the nickel-based catalysts [17, 18]. However, most of these studies were performed at
atmospheric pressure and stability of the catalyst at high-pressure conditions was not reported. In
our earlier work on Pt doped Ni/MgAl2O4 catalyst, it was found out that addition of 0.1wt.% of
Pt to Ni/MgAl2O4 catalyst not only improved the activity of the catalyst but also its stability at
high-pressures [19]. Furthermore, it was observed that testing of the catalysts at high-pressures
was required for determining the stability of the catalyst.
In most of the earlier published studies on supported Rh-Ni bimetallic catalysts for steam
reforming of methane, MgAl2O4 was not used as the support [8]. As mentioned above, MgAl2O4
is widely used as a support due to its stability under high-temperature operations. Moreover,
most of the previous studies have been conducted at atmospheric pressure, whereas industrially
the steam reforming of methane is conducted at higher pressures. Furthermore, the application of
5

metal monolith washcoated with Rh-Ni/MgAl2O4 catalyst for the SRM process has not been
explored.
Besides the improvement in the catalyst effectiveness, application of noble metal doped
Ni/MgAl2O4 washcoated on metal monolith can significantly improve the overall process
efficiency in comparison to packed beds. Various studies have shown that due to their high
thermal conductivity, better catalyst utilization and lower pressure drop, metal monoliths reactors
can be an attractive alternative to conventional packed bed reactors for highly endothermic steam
reforming process [20, 21].
Thus, the objective of this study was to investigate the promotional effect of Rh on Ni/MgAl2O4,
determine the optimum doping of Rh, and examine the stability of the catalysts at ambient and
high-pressure operations. To further enhance the catalytic activity and stability of Rh promoted
nickel catalyst, the effect of calcination temperature for sequential impregnation was also
studied. For comparison, a catalyst prepared by co-impregnation method was also synthesized,
characterized and tested. After optimizing the amount of Rh and preparation method, the most
active Rh doped Ni/MgAl2O4 catalyst was washcoated over a FeCralloy metal monolith. This
washcoated metal monolith was then compared with a Ni/MgAl2O4 washcoated metal monolith
and packed bed reactor containing the Ni/MgAl2O4 and Rh doped Ni/MgAl2O4 catalysts.
2. Experimental
2.1 Catalyst preparation
Two series of Rh noble metal-doped Ni/MgAl2O4 were prepared using impregnation method.
The first series of catalysts were prepared to study the effect of Rh concentration, while the
second series of the catalysts were prepared to study the effect of calcination temperature at
6

different stages of preparation. The catalyst support was prepared by calcining aluminum
magnesium hydroxycarbonate powder, Mg2xAl2(OH)4x+4CO3.nH2O, (Pural MG30, supplied by
o
Sasol, Germany) at 900 C for 4 h to ensure that magnesium aluminate spinel was formed. The
MG30 sample contained about 30 wt.% MgO and 70 wt.% Al2O3.
Before making the two series of catalysts, a large batch of 15 wt.% Ni/MgAl2O4 catalyst was
prepared by wet impregnation. For this purpose, an aqueous solution of Ni(NO3)2.6H2O (Loba
Chemicals, India) was prepared. The previously prepared magnesium aluminate spinel powder
(catalyst support) was then added to the aqueous solution of nickel nitrate solution. After stirring
for 30 minutes, the prepared slurry was dried in a vacuum rotary evaporator at 90 oC. The nickel
impregnated support from the vacuum evaporator was further air dried in an oven at 120 oC for 3
h. The dried sample was then separated into two batches.
For preparing the first series of catalysts, one batch of the 15 wt.% Ni/MgAl2O4 catalyst was
calcined at 850 oC and then cooled to room temperature before sequential impregnation of the Rh
precursor solution. The calcined catalyst was mixed with the required amount of rhodium(III)
nitrate solution (10 wt.% Rh in HNO3 solution, Alfa Aesar) using the incipient impregnation
o
method. After Rh impregnation, the catalyst was dried in an oven at 120 C for 3 h followed by
o
calcination at 600 C for 3 h. These catalysts are denoted as xRh15Ni where x is the wt.%
loading of Rh. The Rh loading was varied from 0.1 to 1.0 wt.%, whereas the nominal Ni loading
was kept fixed at 15.0 wt.%.
A second series of catalysts containing 0.5 wt.% Rh doped 15wt.% Ni/MgAl2O4 were prepared
by calcining the catalysts at either 600 oC or 850 oC after sequential impregnation of Ni and Rh.
These catalysts were denoted as Cat T1T2, where T1 is the calcination temperature after Ni
7

impregnation and T2 is the calcination temperature after Rh impregnation. Thus, the catalyst
o
o
calcined at 600 C after Ni impregnation and 850 C after Rh impregnation is denoted as
Cat600850. In addition, one catalyst (0.5 wt.% Rh doped 15wt.% Ni/MgAl2O4) was also
prepared by co-impregnation of rhodium mixed with nickel precursor solutions and calcined at
600 oC. This catalyst is denoted as RhNicoimp. Calcination for all the catalysts was done with a
temperature ramp of 2 oC.min-1.
2.2 Catalyst characterization
Several techniques were used to characterize the fresh and used catalysts. The BET (Brunauer–
Emmett–Teller) surface area of the support and different catalysts were determined on
AutosorbiQ (Quantachrome). The temperature-programmed reduction (TPR) profiles of the
different catalysts were obtained on AMI200 instrument (Altamira). To remove the volatile
impurities from the catalysts before TPR analysis, the samples were degassed for 30 minutes at
250 oC in an argon flow of 30 ml.min-1. After degassing, the catalyst samples were reduced in a
flow of 10 mol% hydrogen/argon mixture in the temperature range of 50 to 1000 oC. During the
o
reduction step, the temperature was increased with a ramp of 10 C.min-1. The degree of
reduction of each catalyst was calculated as follows:
(1)
The amount of H2 consumed was calculated using the area obtained from the TPR profile of each
catalyst, while the H2 required for complete reduction was calculated from the total number of
moles of Ni and Rh present in the catalyst. The following reduction reactions were considered
for calculating the amount of H2 required for complete reduction of Ni and Rh oxides:
8

NiO + H2 → Nio + H2O
Rh3+ + 1.5H2 → Rho + 3H+
(2)
(3)
H2 chemisorption was measured using the same AutosorbiQ instrument (Quantachrome) as
o
mentioned above. Before H2 chemisorption, the samples were degassed at 350 C for 3 h
followed by reduction at 850 oC for 3 h. After reduction, the catalyst sample was cooled to 30 oC
in a flow of helium. Two cycles of adsorption isotherm in the pressure range of 0.10 to 1.06 bar
were measured to differentiate between the weak, strong and combined isotherms of adsorbed
hydrogen over the metal. Based on the hydrogen uptake, the dispersion (D) was calculated, while
the effective dispersion (Deff) of the catalyst samples was calculated using the following
equation:
(4)
where f is the degree of reduction (%) of the catalyst as determined by temperature-programmed
reduction.
X-ray diffraction (XRD) patterns of the magnesium aluminate spinel support and reduced
catalyst samples were obtained from PAN analytical X’Pert instrument using Ni-filtered Kα
radiation from a Cu target (λ = 1.541841Å
). XRD data was recorded from 10o to 90o with a
4omin-1 scanning rate. Crystallite sizes of the metal particle of the different catalysts were
calculated using the Scherrer equation. For this purpose, the peak at 51.7o, due to the (200)
crystal plane family of Ni, was used [22].
Transmission electron microscopy (TEM) analysis of 0Rh15Ni and Cat600600 was carried out
using an FEI Titan G2 60-300 (300 kV) high-resolution transmission electron microscope to
9

determine the particle size change due to the addition of Rh. Before TEM imaging, the catalyst
samples were reduced at 850 oC for 3 h, and the reduced catalyst was dispersed in ethanol. About
5 μl of this dispersion was loaded on a carbon-coated copper grid, and then the loaded grid was
dried at ambient conditions using a table lamp.
A Cary500 UV-vis (Agilent) instrument was used to obtain UV-vis spectra of the different
samples. Spectra were recorded from 200 nm to 1800 nm under ambient conditions. Thermo-
gravimetric analysis (TGA) of the as-received MG30 powder was obtained on a SDT Q600, TA
o
o
o
Instruments. During TGA, the sample was heated from 25 C to 1000 C with a ramp of 10 C/
min in a flow of air.
2.3 Washcoating of metal monolith
To compare the performance of metal monoliths with a packed bed reactor, 33 mm o.d.
FeCralloy metal monoliths (supplied by Emitec Emission Control Tech. Pvt. Ltd, Pune) were cut
into pieces of 30 mm length and the catalyst was deposited on these by washcoating..
Washcoating of metal monoliths using Ni/MgAl2O4 catalyst was done according to the procedure
described in our earlier work [23]. Similar washcoating procedure was applied for Rh doped
Ni/MgAl2O4 catalyst and Cat600600. Well-adhered homogeneous washcoat layer was obtained
over FeCralloy metal monoliths for both the catalysts.
2.4 Activity tests
The activity tests of the two series of catalysts and metal monoliths were performed in two
different Inconel600 reactor tubes. The effect of adding Rh to Ni/MgAl2O4 catalysts (Series I and
II) was studied in a reactor tube of length 40 cm and diameter of 3.4 mm at 600 oC, and 1 or 10
bar pressure. In this study, internal mass transfer limitations were eliminated by keeping catalyst
10

particle size in the range of 0.25-0.30 mm. 100 mg of the catalyst was mixed with 100 mg of
quartz particles of the same average size and placed at the centre of the tubular reactor. The
temperature of the catalyst bed was measured with a Type K thermocouple placed just above the
catalyst bed. The upper and lower parts of the reactor tube were filled with quartz chips for
efficient mixing of reactants, preheating and supporting the catalyst bed.
As metal monoliths had a diameter of 33 mm, a different Inconel600 reactor tube of internal
diameter 36mm and length 40 mm was used for the metal monoliths experiments. To avoid the
bypassing, ceramic paper of thickness 1 mm was wrapped around the metal monolith before
sliding it into the reactor tube. To investigate the SRM process near industrial conditions and to
compare the performance of metal monoliths with a corresponding packed bed catalyst, packed
bed experiments with average catalyst particle size of 1.5 mm were also performed in this reactor
tube. Two Type K thermocouples were used to measure the tube outer wall temperature (Twall
)
and the temperature at the centre of the catalyst bed (Tcenter). During these experiments, the total
pressure was kept at 1 bar and the Twall was maintained at 600 oC.
For both reactor configurations, methane and nitrogen (used as internal standard for GC analysis)
were fed through mass flow controllers (Bronkhorst, Netherlands) while water was fed through a
high pressure metering pump (Laballiance, USA). The reactant mixture was preheated to 400 oC.
Excess water in the reactor effluent was condensed using a refrigerated circulator. Condensed
water from the product gases was separated in a gas-liquid separator. To pressurize the reactor
system at 1 or 10 bar, a Tescom back pressure regulator was installed after the gas-liquid
separator. The reactor pressure was measured by a pressure gauge installed at the reactor inlet.
Analysis of product gases was achieved in a gas chromatograph (GC) equipped with a thermal
conductivity detector. Separation of product gases in GC was performed using a Carbosphere
11

column (3 m long). Before the activity test, the catalyst samples and washcoated monoliths were
reduced in-situ at 850 oC for 3 h in an H2 flow of 60 ml.min-1. The reduction temperature of 850
o
oC was reached with a ramp rate of 3 C.min-1. Activity tests were conducted for 8 h with a
W/FA0 (ratio of weight of the catalyst to inlet molar flow rate of methane) of 0.34 gcat.h.mol-1
and CH4:H2O:N2 molar ratio of 1:5:1. Several runs were taken, and the reported values for the
variation of CH4 conversion with time-on-stream (TOS) were the average of these runs. Methane
conversion and CO selectivity for each run were calculated as follows:
(5)
(6)
where
,
and
are the outlet molar flow rates of CO, CO2, and CH4, respectively.
Hydrogen yield for each catalyst was calculated by dividing the molar fraction of hydrogen
produced to the molar methane fed in the reactor.
3. Results and discussion
3.1 Catalyst characterization
Table 1 shows the surface area of the as-received MG30 powder, calcined MG30 and the
different catalysts that were prepared. The surface area of the catalyst support decreased from
o
250.1 to 73.6 m2.g-1 after calcination at 900 C. A TGA analysis showed a weight loss of 40%
during calcination of the as-received MG30 powder. This decrease in surface area and weight
loss was attributed to the change in pore structure and evolution of volatile substances present in
the as-received MG30 powder during its transformation to the magnesium aluminate spinel
12

phase. The surface area of the support decreased further on addition of the metals. The surface
area of the catalyst 0Rh15Ni decreased to 43.2 m2.g-1 after the calcined MG30 was impregnated
o
with nickel metal and calcined at 850 C. The surface area decreased once again for the Rh
doped catalysts, which were prepared by sequential impregnation (Series 1 catalysts). Such a
decrease in the surface area occurred probably due to two factors: re-calcination after metal
impregnation, and blocking of support pores by metal particles [24]. Effect of re-calcination was
confirmed by measuring the BET surface area of the Cat850 re-calcined at 600 oC. It was found
that the surface area of this unpromoted re-calcined catalyst decreased from 43.2 m2/gm to 42.1
m2/g. This observation indicates that re-calcination marginally affects the catalyst surface area
and the decrease in surface area is mainly due to the addition of Rh metal. However, an increase
in Rh concentration does not change the surface area significantly. For Series 2 catalysts,
RhNicoimp had the highest surface area (60.7 m2.g-1) followed by Cat600600. It was observed
o
that catalysts that had experienced a calcination temperature of 850 C either after Ni or Rh
impregnation had a lower surface area. The surface areas of these catalysts of Series 2 were
similar to those catalysts of Series 1 since both series of catalysts experienced a calcination
temperature of 850 oC. These results indicate that calcination temperature is the dominant factor
for the decrease in surface area. Similar results have also been reported by Miyata and coworkers
[25].
UV-vis spectroscopy was used to detect the presence of the nickel aluminate (NiAl2O4) phase.
Studies have shown that nickel with alumina can form either tetrahedral or octahedral spinel
structures in the oxygen lattice. Absorption bands for octahedral Ni(II) in a NiO lattice are
observed at about 380, 430 and 720 nm, whereas the absorption bands for tetrahedral Ni(II) in
the nickel aluminate lattice are observed at 550 nm and in the 600-650 nm region [26-29]. The
13

UV-vis spectra of catalysts of Series 1 and 2 are shown in Figs. 1(i) and 1(ii), respectively. As
can be seen from Fig. 1(i) the spectra of 0Rh15Ni revealed the presence of bands corresponding
to NiAl2O4 in addition to those due to NiO. The other Series 1 catalysts revealed no significant
change in the peak due to NiAl2O4. It appears that NiAl2O4 was formed when the nickel
o
impregnated catalyst was calcined at 850 C. Furthermore, the presence of Rh did not affect the
formation of NiAl2O4. For Series 2 catalysts shown in Fig. 1(ii), the formation of NiAl2O4 could
be identified for the three catalysts (Cat600850, Cat850600, and Cat850850) that had
o
experienced a calcination temperature of 850 C during their synthesis. These results confirm
that increasing the calcination temperature to 850 oC facilitates the formation of NiAl2O4 and that
the presence of Rh does not have an effect on the formation of NiAl2O4. Furthermore, the
presence of bands at 380, 430 and 720 nm may be due to the NiO species that diffused into the
support as previously reported [4, 26].
The TPR profiles of Series 1 and 2 catalysts are shown in Figs. 2(i) and 2(ii), respectively. The
o
maximum of the peak in the TPR profiles of these samples (Tmax) ranges from 745 to 806 C.
The presence of Tmax at these high temperatures has usually been attributed to the reduction of
NiAl2O4 [30]. Earlier studies have shown that metal oxide solid solutions are formed by NiO and
MgO, and these solid solutions also reduce at high temperature [31, 32]. As suggested previously
excess MgO is present on the surface of the MgAl2O4 spinel support [30]. The Ni impregnated
can interact with this excess MgO to form NiO-MgO solid solutions, in addition to forming
NiAl2O4. Thus, the Tmax at these high temperatures is due to the reduction of NiAl2O4 and/or
NiO-MgO solid solutions.
For the Series 1 catalysts, shown in Fig. 2(i), it can be observed that all Rh doped Ni/MgAl2O4
catalyst samples showed a slightly higher Tmax temperature relative to Ni/MgAl2O4. This increase
14

in Tmax can be attributed to the sequential impregnation method adopted during catalyst
preparation. The sequential impregnation method adopted in this study involved two calcination
steps, which facilitated the formation of strongly bonded Ni sites and NiAl2O4. The TPR profiles
in Fig. 2(i) also revealed that with an increase in Rh content, Tmax decreased. A peak attributed to
o
the reduction of Rh-oxide was also observed between 300 and 400 C, which became more
intense with an increase in Rh content. Furthermore, the degree of reduction increased with Rh
content as shown in Table 1. Earlier studies have also reported that addition of noble metals
facilitates hydrogen spillover, which results in a decrease in reduction temperature and an
increase in degree of reduction [33, 34]. It appears that with an increase in Rh content, the
hydrogen spillover increases, which facilitates the reduction of the catalyst to occur at lower
temperatures and increases the degree of reduction of the NiAl2O4 species and/or NiO-MgO
solid solution.
The TPR profiles for the Series 2 catalysts and RhNicoimp are shown in Fig. 2(ii). In three of
these catalysts (Cat600850, Cat850600, and Cat850850) NiAl2O4 was detected in the UV-vis
spectra, whereas NiAl2O4 was not detected in RhNicoimp and Cat600600. As mentioned above,
it appears that the lower Tmax of Cat600600 and RhNicoimp catalysts was primarily due to the
reduction of NiO-MgO solid solution [35]. Furthermore, the higher Tmax temperature for all the
Series 1 catalysts and Cat600850, Cat850600, and Cat850850 was due to the reduction of
NiAl2O4 in addition to the reduction of the NiO-MgO solid solution. The change in Tmax for
Series 1 catalysts was related to the amount of Rh present on the catalyst surface, and a similar
phenomenon might occur for the Series 2 catalysts. Furthermore, the presence of NiAl2O4 and
NiO-MgO solid solution for Cat600850, Cat850600, and Cat850850 may also play a role in the
change in Tmax
.
15

The degree of reduction of three catalysts of Series 2 (Cat600850, Cat850600, and Cat850850)
o
and Series 1 catalysts that had experienced a calcination temperature of 850 C was lower than
the two catalysts (RhNicoimp and Cat600600) that were calcined at 600 oC. The lower degree of
reduction for Cat600850, Cat850600, and Cat850850 can be readily attributed to the detectable
amount of NiAl2O4 formed in these catalysts. Furthermore, the degree of reduction of NiO-MgO
solid solution also decreases with increasing calcination temperature [32].
As active metal size and dispersion can significantly affect the catalytic activity, three
characterization techniques (XRD, H2 chemisorption and TEM) were used to determine these
properties. Figs. 3(i) and 3(ii) show the XRD profiles of the Series 1 and 2 catalysts,
respectively. In our earlier work, we have demonstrated that MG30 was converted to magnesium
o
aluminate spinel after calcination at 900 C for 4 h [4]. Peaks corresponding to MgAl2O4 spinel
and Ni metal can be observed in both Figs. 3(i) and 3(ii). Peaks of Rh(111) and Rh(200) metal
are expected at 41o and 47o [36]. However, such peaks were not detected in the XRD profiles
either due to small amounts of Rh or the high dispersion of Rh or both. The peaks corresponding
to NiAl2O4 could not be readily identified because these peaks overlap with those of MgAl2O4
spinel [37, 38]. It is interesting to note that even after reduction of the catalyst at 1000 oC, a small
peak at 43.8o could be identified, which could be due to NiO and/or MgO [39]. In our earlier
studies, this peak was attributed to NiO and the formation of metal oxide solid solutions. Some
part of this metal oxide solid solution can migrate into the bulk support and remain inaccessible
for reduction [4]. Presence of non reducible species or partial reduction of Series I and II
catalysts also support this observation.
Ni metal peaks were identified at 51.8o and 76.3o, and the peak located at 51.8o was used to
estimate the crystallite size using Scherrer equation. Table 1 shows the crystallite size (dNi)
16

calculated for both series of catalysts. For Series 1 catalysts, on increasing the Rh loading from
0.0 to 1.0 wt.%, the crystallite size of Ni metal marginally decreased from 10.9 nm to 9.9 nm.
Simultaneously, the dispersion of active metals increased from 5.4 to 6.3%. This marginal
o
decrease in crystallite size was attributed to the high temperature (850 C) calcination of the
Ni/MgAl2O4 catalyst after nickel impregnation. Others have also reported that doping of small
amount of noble metal increases the dispersion of nickel [40]. It appears that an inverse-
correlation exists between the crystallite size and dispersion . Our earlier study had shown that
for the Ni/MgAl2O4 catalyst calcined at 850 oC, the nickel active metal present on the surface of
the support showed anchoring effect [4]. This anchoring of nickel appears to inhibit further
dispersion of active metal on the support surface. Furthermore, the anchoring of nickel may be
associated with the MgO present on the surface of the spinel support as discussed above.
Table 1 also lists the crystallite sizes of Series 2 catalysts. As can be seen from Table 1, the
crystallite size of the catalyst prepared by co-impregnation (RhNicoimp) was smaller than
Cat600600, which was prepared by sequential impregnation. This effect of preparation method
on crystallite size can be attributed to two factors. Firstly, during the preparation of RhNicoimp,
the spreading of NiO on the catalyst surface is more likely as both Rh2O3 and NiO are present
o
during calcination of the catalyst. Secondly, Cat600600 was calcined twice at 600 C instead of
once for RhNicoimp. Similar observation have also been reported by Chantaravitoon and co-
workers [41]. In their work, it was observed that co-impregnation of noble metal facilitates the
dispersion of active metals on the support surface. The smallest crystallite size in Cat850850 can
be attributed to the presence of rhodium oxide during the second calcination step, which further
synergizes the dispersion of active metals on the support surface. Nagaoka et al. [40] suggested
that the presence of Rh also enhances the anchoring of Ni particles, which results in a smaller
17

crystallite size. The nickel crystallite size of Cat600850 was smaller than that of Cat850600,
which indicates that high calcination temperatures after addition of Rh give rise to a better
dispersion of nickel.
To further investigate the effect of Rh on the particle size of the active metal, TEM imaging of
reduced Cat600600 and 0Rh15Ni catalysts was carried out. These catalysts were selected for
imaging because the highest activity was obtained with Cat600600 and the lowest with 0Rh15Ni
(discussed later). The TEM images along with the particle size distribution are shown in Fig. 4.
Addition of Rh resulted in a change in particle size distribution with a significant increase in the
number of particles in the size range of 0-10 nm and a decrease in the number in the size range of
10-20 nm. The average particle size calculated from TEM analysis for 0Rh15Ni and Cat600600
catalysts was 11.7 and 8.6 nm, respectively. This observation further confirms the effect of Rh in
decreasing the average particle size of the active metal, thereby increasing the dispersion of
nickel on the support.
The dispersion data of Series 1 and 2 catalysts are also tabulated in Table 1. As discussed earlier,
calcination at high temperature resulted in stronger metal oxide support interaction, which
decreases the availability of nickel. To account for the loss of available nickel during catalyst
calcination, the effective dispersion, Deff, of the active metal particles was calculated using eqn.
(1). However, the calcination temperatures used during catalyst preparation were different, and
an appropriate degree of reduction needs to be considered. The Deff of 0.1Rh15Ni, 0.5Rh15Ni
and 1.0Rh15Ni catalysts was calculated using the degree of reduction of 0Rh15Ni catalyst,
which was 46.3%. Such a choice was justified since these catalysts were calcined at 600 oC after
Rh impregnation and no additional formation of NiAl2O4 relative to 0Rh15Ni was detected in the
UV-vis spectra of these Rh doped catalysts as shown in Fig. 1(i). For the Series 1 catalysts, there
18

was a noticeable increase in Deff when small amounts of Rh was added (comparing Deff of
0Rh15Ni and 0.1Rh15Ni). This high value of Deff was relatively constant for the 0.1Rh15Ni and
0.5Rh15Ni catalyst, whereas a marginal increase in Deff was observed for the 1.0Rh15Ni
catalyst. Various researchers have attributed this increase in the dispersion to the synergetic
effect between Ni and the noble metal resulting in alloy formation [42, 43]. Earlier studies have
also shown that although the addition of noble metal increases the dispersion, higher surface
concentrations can result in the agglomeration of active metals [44].
For two of the Series 2 catalysts, Cat600600 and RhNicoimp, the Deff was calculated using the
o
degree of reduction of the catalyst that was calcined at 600 C after nickel impregnation. The
degree of reduction of such a catalyst was 56.4%, which is consistent with the data reported
previously [39]. Based on this information, the Deff for RhNicoimp was slightly more than
Cat600600. Tomishige has also reported similar results for his work on sequential and co-
impregnated Ni-Pt/γ-Al2O3 catalyst [45]. The Deff values for Cat600850, Cat850600 and
Cat850850 were difficult to determine since the surface concentration of the Ni-Rh alloy, the
formation of metal oxide support solid solution, and the formation of NiAl2O4 would have an
effect on the degree of reduction and consequently Deff.
3.2 Activity test
3.2.1 Promotional effect of Rh metal
The variation of methane conversion with time-on-stream for catalysts of Series 1 at a reactor
pressure of 1 and 10 bar is shown in Figs. 5(i) and 5(ii), respectively. Table 2 shows the initial
methane conversion and hydrogen yield (after 1h of time-on-stream) of all the catalyst. At both
pressures, small amounts of rhodium (0.1 wt.% Rh) in the catalyst significantly increased the
initial methane conversion and hydrogen yield. However, with a further increase in Rh content of
19

the catalyst, the methane conversion and hydrogen yield did not increase. Li and co-workers
have also reported similar results for steam reforming of methane on Rh-Ni/Mg(Al)O catalyst
[18]. Furthermore, at 1 bar pressure, the stability of the catalysts was high, and the presence of
Rh resulted in only a marginal improvement in the stability of the catalyst of Series 1 (Table 2).
Since the stability of catalyst is strongly affected by the reactor pressure, the stability of the
catalyst was also checked at a reactor pressure of 10 bar. At 10 bar reactor pressure, the initial
conversions achieved with 0.1Rh15Ni, 0.5Rh15Ni and 1.0Rh15Ni were similar and greater than
the conversion achieved with 0Rh15Ni. However, the stability of the catalysts was different. As
can be observed from Table 2 and Fig. 5(ii), compared to 0Rh15Ni the stability of 0.1Rh15Ni
was lower, while the stability of 0.5Rh15Ni and 1.0Rh15Ni was similar.
Fig.6 shows the change in CO selectivity for Series 1 catalysts at 1 and 10 bar pressure. On
increasing the Rh doping in the catalyst the CO selectivity increased at both reactor pressures.
These results indicate that CO selectivity is directly related to the Rh concentration in the
catalyst. To confirm these results, another experiment was conducted with 0.1%Rh/MgAl2O4 at
similar conditions. A CO selectivity of 0.24 was observed for the 0.1%Rh/MgAl2O4 catalyst,
which was significantly higher than the selectivity obtained with either 0Rh15Ni or 1.0Rh15Ni.
Thus, the increase in CO selectivity can be attributed to the presence of Rh metal on the catalyst
surface. In addition to the presence of Rh, the crystallite size of the active metals can also affect
the CO selectivity [46]. Catalysts having a smaller crystallite size tend to form CO during
reforming. As shown above, the addition of noble metal caused a decrease in the crystallite size
of active metals. Thus, an increase in CO selectivity can also be attributed to this phenomenon.
Our earlier studies have shown that the calcination temperature can have a significantly effect on
the activity and stability of the catalyst [4]. Consequently, the effect of calcination temperature
20

was investigated for 0.5wt.% Rh doped 15wt.%Ni/MgAl2O4 catalyst. For the Rh doped catalysts
prepared by sequential impregnation there were two independent calcination temperatures. The
two independent calcination temperatures gave rise to four catalysts, namely Cat600600,
Cat600850, Cat850600 and Cat850850. A co-impregnated catalyst, RhNicoimp, that was
o
calcined at 600 C was also included in the comparison to see the effect of preparation method.
Figs. 7(i) and 7(ii) show the variation of conversion with time-on-stream for Series 2 catalysts at
1 and 10 bar pressure, respectively. As can be seen from these figures Cat600600 showed the
highest activity among the Series 2 catalysts, while Cat850600 had the lowest activity. As
observed for the activity of the Series 2 catalyst, the hydrogen yield also followed a similar trend.
At both pressures (1 and 10 bar), the trend in activity was as follows:
Cat600600 > RhNicoimp > Cat600850 ~ Cat850850 > Cat850600.
The higher activity of Cat600600 and RhNicoimp catalyst calcined at a relatively lower
o
temperature of 600 C can be attributed to the relative abundance of surface Ni and Ni-Rh sites.
Moreover, the catalytic activity of Cat600600 was higher than RhNicoimp catalyst. This increase
in catalytic activity of sequentially impregnated Cat600600 catalyst in comparison to co-
impregnated RhNicoimp catalyst was attributed to a higher surface concentration of Rh species
in the sequentially impregnated catalyst. A higher surface concentration of Rh would facilitate
the formation of Ni-Rh alloy. Such a proposal is consistent with the work on oxidative reforming
of methane on Pt-doped Ni/Al2O3 by Li et al.[47]. In their study the FTIR spectra of adsorbed
CO revealed that the catalyst prepared by sequential impregnation had a higher surface
concentration of Pt metal than the co-impregnated catalyst, which resulted in a higher catalytic
activity.
21

The activity of Series 2 catalysts shown in Figs. 7(i) and 7(ii) reveals that the intermediate and
final calcination temperature used during synthesis also affects the activity and stability of the
catalyst. Comparing the activity of Cat600600 with the catalysts that have experienced a
calcination temperature of 850 oC during their synthesis suggests that the decrease in activity of
the latter was due to the loss of Ni metal caused by the formation of NiAl2O4 and/or NiO-MgO
solid solution. An increase in calcination temperature may also cause a decrease in activity of the
NiO-MgO solid solution as reported previously [31, 32]. However, the trend in the catalytic
activity of Cat600850, Cat850850, and Cat850600 is difficult to reconcile and may be related to
the surface abundance of the Ni-Rh sites, the amount of NiAl2O4 formed and the availability of
Ni in the NiO-MgO solid solution. Comparing the catalytic activity of Cat850600 and Cat850850
o
suggests that an increase in the surface concentration of Rh is facilitated at 850 C. For these
catalysts attempts to correlate the catalytic activity with Deff were ineffective since, as discussed
above, the basis for the calculation of Deff was difficult to ascertain.
Table 2 shows the stability data for Series 2 catalyst at 1 and 10 bar pressure. Stability of Series 2
catalysts varied in the order of
Cat600600 > RhNicoimp > Cat600850 ~ Cat850850 > Cat850600.
In our earlier work, we have shown that the main cause of deactivation of catalyst in steam
reforming of methane was carbon formation on the catalyst surface [4]. This carbon formation
over the active metal surface is affected by several factors. These factors include the crystallite
size of active metal particles, the formation of metal support complexes such as NiAl2O4, metal
oxide support interaction, nature and surface concentration of active metal species present on the
catalyst surface, e.g., Ni-Ni, Ni-Rh and Rh-Rh [40, 46, 48, 49]. The higher stability of
22

Cat600600 and RhNicoimp catalyst was attributed to the higher abundance of Ni-Rh sites on the
catalyst surface, which also gave rise to a higher activity of these catalysts. Among the catalysts
that experienced high calcination temperatures, Cat 600850 and Cat850850 showed higher
stability relative to Cat850600. Higher stability of these two catalysts was attributed to the high
surface concentration of Rh, which is known to improve the stability [50, 51].
Fig.8 shows the CO selectivity of Series 2 catalysts at 1 and 10 bar pressure. Similar to the
observation for Series 1 catalysts, a decrease in CO selectivity was observed when the pressure
was increased to 10 bar pressure. Similar to the trend observed for catalytic activity, Cat600600
had the highest CO selectivity while Cat850600 had the lowest. It would appear that the increase
in conversion is directly related to the increase in CO selectivity. To further confirm the
observation of an increase in CO selectivity due to the presence of Rh metal, CO selectivity of
Cat600600 and 0Rh15Ni were compared at iso-conversions, which can be achieved by changing
the contact time. It was found that the CO selectivity of Cat600600 was 0.12 and of 0Rh15Ni
was 0.10 at ~49% methane conversion, which confirms that Rh gives rise to these changes.
Similar observations have also been by reported by Kusakabe and co-workers [52]. As discussed
earlier for the CO selectivity of Series 1 catalysts, the availability of Rh metal on the catalyst
surface, as well as the particle size of active metals, can significantly affect the CO selectivity of
the catalyst. Due to its highest degree of reduction, Cat600600 showed the highest selectivity.
Similar CO selectivities were observed for RhNicoimp, Cat600850, and Cat850850, while the
CO selectivity was lower for Cat850600. As discussed earlier, lower CO selectivity of
Cat850600 can be attributed to lower active metal dispersion on the catalyst surface.
3.2.2 Application of washcoated metal monolith
23

To further improve the performance of Rh-Ni/MgAl2O4 catalyst for the steam reforming
reaction, the most active catalyst and base catalyst, Cat600600 and 0Rh15Ni, were washcoated
on separate metal monoliths and compared with the corresponding catalysts in packed bed
reactors. Fig. 9 shows the comparison of metal monoliths and packed bed reactor for both
o
Ni/MgAl2O4 and Rh-Ni/MgAl2O4 catalysts at 600 C reactor wall temperature and atmospheric
pressure. Negligible pressure drop was detected across both metal monolith reactor and packed
bed reactor. Figure 9 reveals that Rh-Ni/MgAl2O4 catalyst was more active than Ni/MgAl2O4
catalyst in both packed bed and metal monolith reactors. Moreover, for both catalysts, the
performance of the metal monolith was significantly better than that of the packed bed reactor.
Lower methane conversion in the packed bed reactor was attributed to the presence of heat and
mass transfer limitations. Our previous studies have shown that catalyst with an average particle
size above 0.275 mm show internal mass transfer limitations [23]. Since in this study the average
size of catalyst particle was kept as 1.5 mm, the presence of strong internal mass transfer
limitations was expected. In contrast, the washcoat thickness in the metal monolith reactor was
about 17 μm and the internal mass transfer limitation was eliminated. Furthermore, the Tcenter in
o
o
the monolith reactor was 570 C due to higher thermal conductivity, in comparison to 530 C
observed in the packed bed reactors, which suggest the presence of heat transfer limitations.
Apart from heat and mass transfer limitations, our previous study revealed that the activity of the
catalyst also increased after the washcoat preparation step [22]. To deposit a well- adhered and
homogeneous washcoat on the monoliths, the washcoating preparation step required the particle
size of the catalyst slurry to be decreased in a ball mill for 48 h. This milling process not only
reduces the catalyst particle size but also redistributes the active metal present on the catalyst
surface. This active metal redistribution increased the active metal dispersion which also
24

contributed to the higher conversion obtained in the monolith reactor. Furthermore, it was also
noticed from Fig. 9 that there was an increase of ~9% in methane conversion for the two metal
monoliths compared to ~ 4% for the two packed bed reactors. Higher increase in methane
conversion in metal monoliths in comparison to packed bed reactor for both catalysts was
attributed to higher geometrical surface area in case of metal monolith reactors. Earlier studies
have shown that an increase in geometrical surface area significantly increases the availability of
active metal over the catalyst surface, which results in a higher catalytic activity of the
washcoated metal monolith in comparison to catalyst pellets [41].
The change in product distribution due to change in reactor system for both Ni/MgAl2O4 and Rh-
Ni/MgAl2O4 catalysts is presented in Table 3. CO selectivity of both catalysts was calculated
using equation (6). It was found that for both catalysts the H2/CO ratio was higher in case of
packed bed reactors compared to monolith reactor. However, higher H2/CO2 ratio and CO
selectivity was observed in metal monolith reactors for both the catalysts. Decrease in H2/CO
ratio, and increase in H2/CO2 ratio and CO selectivity was attributed to higher methane
conversion in metal monolith reactors.
Conclusions
Addition of Rh noble metal has a significant effect on the physicochemical properties of the
Ni/MgAl2O4 catalyst, and these changes result in an increase in catalytic activity, hydrogen yield
and stability. Addition of 0.1wt.% Rh to Ni/MgAl2O4 was sufficient to significantly increase the
catalytic activity. Further addition of Rh did not have an appreciable effect on the initial catalytic
activity. However, the stability of the catalyst increased till the addition of 0.5 wt.% Rh and
further Rh addition did not affect catalyst stability. The increase in activity and stability of the
25

Rh doped catalyst was attributed to an increase in degree of reduction, increase in dispersion of
active metals and formation of Ni-Rh alloy. A change in intermediate or final calcination
temperature after impregnation of either Ni or Rh also had an effect on the physicochemical
properties of the catalyst. The catalysts calcined at 600 oC had a smaller active metal particle size
o
compared to catalysts calcined at 850 C after either Ni or Rh impregnation. TPR results also
revealed that the catalysts calcined at 600 oC have a higher degree of reduction. Activity test also
shows that the 0.5wt.%Rh modified Ni/MgAl2O4 catalyst calcined at 600 oC both after Ni and Rh
impregnation, i.e. Cat600600, had the highest activity, hydrogen yield and stability amongst
those studied. Furthermore, the sequentially impregnated catalyst Cat600600 showed better
activity than co-impregnated method catalyst calcined at 600 oC. Washcoating of this Cat600600
catalyst on metal monoliths further improves its performance. Furthermore, the higher activity of
this Rh-Ni/MgAl2O4 washcoated metal monolith was attributed to the absence of heat and mass
transfer limitations and increase in active metal dispersion of washcoat catalyst.
Acknowledgment
The financial support provided for this study by Science and Engineering Research Board
(SERB), Government of India under Grant No. SB/S3/CE/066/2013 is gratefully acknowledged.
26

References:
[1] F. Suleman, I. Dincer and M. Agelin-Chaab, Int. J. Hydrogen Energy 40 (2015) 6976-6987.
[2] S. E. Hosseini and M. A. Wahid, Renew. Sustainable Energy Rev. 57 (2016) 850-866.
[3] R.J. Farrauto and C.H. Bartholomew, Fundamentals of Industrial Catalytic Processes, John
Wiley & Sons Inc., 2010.
[4] S. Katheria, A. Gupta, G. Deo and D. Kunzru, Int. J. Hydrogen Energy. 41 (2016) 14123-
14132.
[5] D. Mei, V.A. Glezakou, V. Lebarbier, L. Kovarik, H. Wan, K. O. Albrecht, M. Gerber, R.
Rousseau and R. A. Dagle, J. Catal. 316 (2014) 11-23.
[6] E. Kikuchi, S. Tanaka, Y. Yamazaki and Y. Morita, Bulletin of Jpn. Pet. Inst. 16 (1974) 95-
98.
[7] F. Watanabe, I. Kaburaki, N. Shimoda, S. Satokawa, Fuel Process. Technol. 152 (2016) 15-
21.
[8] D.L. Li, Y. Nakagawa, K. Tomishige, Appl. Catal. A: Gen. 408 (2011) 1-24.
[9] L.P.R. Profeti, E.A. Ticianelli, E.M. Assaf, Fuel, 87 (2008) 2076-2081.
[10] M. Ferrandon, A.J. Kropf, T. Krause, Appl. Catal. A: Gen. 379 (2010) 121-128.
[11] C. Hulteberg, Int. J. Hydrogen Energy. 37 (2012) 3978-3992.
[12] T. Miyata, M. Shiraga, D. Li, I. Atake, T. Shishido, Y. Oumi, T. Sano, K. Takehira, Catal.
Commun. 8 (2007) 447-451.
[13] D. Li, K. Nishida, Y. Zhan, T. Shishido, Y. Oumi, T. Sano, K. Takehira, Appl. Catal. A:
Gen. 350 (2008) 225-236.
[14] E. Finocchio, G. Busca, P. Forzatti, G. Groppi, A. Beretta, Langmuir 23 (2007) 10419-
10428.
[15] Z.H. Wei, J.M. Sun, Y. Li, A.K. Datye, Y. Wang, Chem. Soc. Rev. 41 (2012) 7994-8008.
[16] M. Garcia-Dieguez, I.S. Pieta, M.C. Herrera, M.A. Larrubia, L.J. Alemany, J. Catal. 270
(2010) 136-145.
[17] E.L. Foletto, R.W. Alves, S.L. Jahn, J Power Sources, 161 (2006) 531-534.
[18] D.L. Li, T. Shishido, Y. Oumi, T. Sano, K. Takehira, Appl. Catal. A: Gen. 332 (2007) 98-
109.
[19] V.K. Jaiswar, S. Katheria, G. Deo, D. Kunzru, Int. J. Hydrogen Energy 42 (2017) 18968-
18976.
27

[20] H.S. Roh, D.K. Lee, K.Y. Koo, U.H. Jung, W.L. Yoon, Int. J. Hydrogen Energy 35 (2010)
1613-1619.
[21] J.H. Ryu, K.-Y. Lee, H. La, H.-J. Kim, J.-I. Yang, H. Jung, J. Power Sources 171 (2007)
499-505.
[22] H. Özdemir, M.A.F. Öksüzömer, M.A. Gürkaynak, Fuel 116 (2014) 63-70.
[23] S. Katheria, G. Deo, D. Kunzru, Energy Fuels 31 (2017) 3143-3153.
[24] A. Zhao, W. Ying, H. Zhang, M. Hongfang, D. Fang, J Nat. Gas Chem. 21 (2012) 170-177.
[25] T. Miyata, D. Li, M. Shiraga, T. Shishido, Y. Oumi, T. Sano, K. Takehira, Appl. Catal. A:
Gen. 310 (2006) 97-104.
[26] R. López-Fonseca, C. Jiménez-González, B. de Rivas, J.I. Gutiérrez-Ortiz, Appl. Catal. A:
Gen. 437–438 (2012) 53-62.
[27] M. Jitianu, A. Jitianu, M. Zaharescu, D. Crisan, R. Marchidan, Vib. Spectrosc. 22 (2000)
75-86.
[28] G. Garbarino, S. Campodonico, A.R. Perez, M.M. Carnasciali, P. Riani, E. Finocchio, G.
Busca, Appl. Catal. A: Gen., 452 (2013) 163-173.
[29] G. Garbarino, C. Wang, I. Valsamakis, S. Chitsazan, P. Riani, E. Finocchio, M. Flytzani-
Stephanopoulos, G. Busca, Appl. Catal. B: Environ., 174-175 (2015) 21-34.
[30] L. Smolakova, M. Kout, E. Koudelkova, L. Capek, Ind. Eng. Chem. Res. 54 (2015) 12730-
12740.
[31] M. Jafarbegloo, A. Tarlani, A.W. Mesbah, J. Muzart, S. Sahebdelfar, Catal. Lett. 146 (2016)
238-248.
[32] Y. H. Wang, H. M. Liu and B. Q. Xu, J. Mol. Catal. A: Chem. 299 (2009) 44-52.
[33] M.A. Ali, T. Kimura, Y. Suzuki, M.A. Al-Saleh, H. Hamid, T. Inui, Appl. Catal. A: Gen.
227 (2002) 63-72.
[34] Y. Mukainakano, B.T. Li, S. Kado, T. Miyazawa, K. Okumura, T. Miyao, S. Naito, K.
Kunimori, K. Tomishige, Appl. Catal. A: Gen. 318 (2007) 252-264.
[35] L.P.R. Profeti, J.A.C. Dias, J.M. Assaf, E.M. Assaf, J Power Sources, 190 (2009) 525-533.
[36] C. Jin, W. Xia, T.C. Nagaiah, J. Guo, X. Chen, N. Li, M. Bron, W. Schuhmann, M. Muhler,
J. Mater. Chem. 20 (2010) 736-742.
[37] E. Heracleous, A.F. Lee, K. Wilson, A.A. Lemonidou, J. Catal. 231 (2005) 159-171.
28

[38] Q. Wei, X. Gao, G. Liu, R. Yang, H. Zhang, G. Yang, Y. Yoneyama, N. Tsubaki, Fuel 211
(2018) 1-10.
[39] A. Aguinaga, J.C. De La Cal, J.M. Asúa, M. Montes, Appl. Catal. 51 (1989) 1-11.
[40] K. Nagaoka, A. Jentys, J.A. Lercher, J. Catal., 229 (2005) 185-196.
[41] P. Chantaravitoon, S. Chavadej, J. Schwank, Chem. Eng. J. 98 (2004) 99-104.
[42] C. Crisafulli, S. Scire, R. Maggiore, S. Minico, S. Galvagno, Catal. Lett. 59 (1999) 21-26.
[43] E.C. Luna, A.M. Becerra, M.I. Dimitrijewits, React. Kinet. Catal. Lett., 67 (1999) 247-252.
[44] R.V. Malyala, C.V. Rode, M. Arai, S.G. Hegde, R.V. Chaudhari, Appl. Catal. A: Gen. 193
(2000) 71-86.
[45] K. Tomishige, J Jpn. Pet. Inst. 50 (2007) 287-298.
[46] A. Karim, T. Conant, A. Datye, J. Catal. 243 (2006) 420-427.
[47] B.T. Li, S. Kado, Y. Mukainakano, M. Nurunnabi, T. Miyao, S. Naito, K. Kunimori, K.
Tomishige, Appl. Catal. A: Gen. 304 (2006) 62-71.
[48] K.O. Christensen, D. Chen, R. Lodeng, A. Holmen, Appl. Catal. A: Gen. 314 (2006) 9-22.
[49] A.J. de Abreu, A.F. Lucrédio, E.M. Assaf, Fuel Process. Technol, 102 (2012) 140-145.
[50] M. Ocsachoque, F. Pompeo, G. Gonzalez, Catal. Today, 172 (2011) 226-231.
[51] H. Arbag, S. Yasyerli, N. Yasyerli, G. Dogu, Int. J. Hydrogen Energy 35 (2010) 2296-2304.
[52] K. Kusakabe, K.I. Sotowa, T. Eda, Y. Iwamoto, Fuel Process. Technol, 86 (2004) 319-326.
29