Study of the effect of Gd-doping ceria on the performance of Pt/GdCeO2/Al2O3 catalysts for the dry reforming of methane

Article abstract of DOI:10.1016/j.cattod.2019.04.079

This work studied the performance of Pt/Al₂O₃ and Pt/Ce_xGd_1-xO₂/Al₂O₃ catalysts for the dry reforming of methane at 1073 K. Pt/Al₂O₃ catalyst underwent seve

Full text of DOI:10.1016/j.cattod.2019.04.079

Catalysis Today xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Review
Study of the effect of Gd-doping ceria on the performance of Pt/GdCeO2/
Al2O3 catalysts for the dry reforming of methane
a,b
a,b
b
b
c
c
R.O. da Fonseca , G.S. Garrido , R.C. Rabelo-Neto , E.B. Silveira , R.C.C. Simões , L.V. Mattos ,
a,b,⁎
F.B. Noronha
a
b
c
Military Institute of Engineering, Chemical Engineering Department, Praça Gal. Tiburcio 80, Rio de Janeiro, 22290-270, Brazil
National Institute of Technology, Catalysis Division, Av. Venezuela 82, Rio de Janeiro, 20081-312, Brazil
Fluminense Federal University, Chemical Engineering Department, Rua Passo da Pátria, 156, Niterói, 24210-240, Brazil
A R T I C L E I N F O
A B S T R A C T
Keywords:
This work studied the performance of Pt/Al
2
O
3
and Pt/Ce
x
Gd1-x
O
2
/Al
2
O catalysts for the dry reforming of methane
3
Dry reforming
Biogas
at 1073 K. Pt/Al O catalyst underwent severe deactivation at the beginning of reaction, whereas the deactivation of
2
3
Pt/Ce0.8Gd0.2/Al
2
O
3
and Pt/Ce0.5Gd0.5/Al
2
O
3
catalysts was less significant. Pt/36Ce0.8Gd0.2/Al
2
O catalyst remained
3
Pt catalysts
quite stable during reaction. in situ XRD and XANES experiments showed the formation of
Cerium-doped gadolinium
Hydrogen production
3+
oxygen vacancies and Ce species during the reduction of the mixed oxide, which promoted the mechanism
of carbon removal from the Pt particle. In the absence of mixed oxide, carbon deposition led to the deactivation
of Pt/Al
2
O catalyst. The cyclohexane dehydrogenation reaction revealed that Pt sintering is responsible for the
3
deactivation of Pt/Ce0.8Gd0.2/Al
2
O
3
and Pt/Ce0.5Gd0.5/Al
2
O
3
catalysts. The stability of Pt/36Ce0.8Gd0.2/Al
O
2 3
catalyst is likely due to the close contact between Pt and the CeGd mixed oxide because of the higher extent of
alumina coverage by the mixed oxide. This promotes the mechanism of carbon removal from the surface of Pt
particles and inhibits Pt particle size growth.
1
. Introduction
A promising alternative to the use of this gaseous mixture con-
taining high CO content would be its conversion into synthesis gas
2
With the discovery of large oil reserves located in the pre-salt area in
(mixture of H
2
and CO) by CO reforming of methane, the so-called dry
2
Brazil, the oil companies have faced new issues. The challenge for the
exploration of this area is related to the amount of associated gas
reforming of methane (DRM). Synthesis gas can be converted into
ammonia and methanol, as well as H and liquid fuels through Fischer-
2
produced and the high amount of CO
2
content in this gas. Due to the
Tropsch synthesis [3]. In addition, the synthesis gas can also be used to
directly feed the solid oxide fuel cells to generate energy [2].
restrictions of environmental laws, this gas cannot be vented or burned
on the platform because it would release large amounts of CO
2
into the
The dry reforming of methane has been widely studied for the
production of synthesis gas [4–7]. Among the advantages of this reac-
tion are the availability and low cost of the reagents and the use of two
of the greenhouse gases as reagents [8]. One of the major challenges of
this technology is related to the deactivation of the catalyst that occurs
mainly due to carbon formation during the process at high temperature
[9–11]. Therefore, the development of a catalyst resistant to carbon
deposition during dry reforming of methane is crucial.
atmosphere, contributing to the greenhouse effect and causing en-
vironmental damage. Furthermore, the costs associated with the se-
paration of CO from natural gas stream are high. Therefore, it is im-
2
portant to find an alternative to the use of this associated gas to
overcome these issues related to the oil exploration of the pre-salt re-
serves [1,2].
Like the associated gas, biogas, generated from the anaerobic di-
gestion of biomass, also presents high levels of CO
2
, resulting in emis-
The carbon deposition on this reaction depends on the type of metal
used. Some transition metals, such as Ru, Rh, Pt, Pd and Ni, are active
sions into the atmosphere that also contribute to the increase of the
greenhouse effect. As only a small part of this gas is used as a source of
heat and energy through combustion processes, the development of
alternative processes for the use of this gas would be of significant in-
terest [2].
for the CO reforming of methane and, according to the literature, the
2
extent of carbon formation on these metals is related to their activity for
CO dissociation [6]. For example, metals such as Rh and Ru have a
high activity for CO dissociation, with no significant carbon formation
2
2
⁎
Corresponding author at: National Institute of Technology, Catalysis Division, Av. Venezuela 82, Rio de Janeiro, 20081-312, Brazil.
E-mail address: fabio.bellot@int.gov.br (F.B. Noronha).
https://doi.org/10.1016/j.cattod.2019.04.079
Received 6 December 2018; Received in revised form 9 April 2019; Accepted 29 April 2019
0920-5861/ © 2019 Elsevier B.V. All rights reserved.
Please cite this article as: R.O. da Fonseca, et al., Catalysis Today, https://doi.org/10.1016/j.cattod.2019.04.079

R.O. da Fonseca, et al.
Catalysis Today xxx (xxxx) xxx–xxx
during the reaction. However, the cost of these metals is very high. Pt
The dehydrogenation of cyclohexane was used to determine the
and Pd metals are not as efficient at dissociating CO and, in this case,
2
metal dispersion of Pt-based catalysts. In this case, most commonly used
the use of a support is extremely important to avoid carbon deposition.
Several studies report the use of cerium oxide as catalyst support for
the dry reforming of methane due to its redox properties and the high
mobility of oxygen, which promotes the mechanism of carbon removal
from the metal surface [2,6,12]. The addition of dopants promotes the
oxygen mobility due to formation of a solid solution, improving the
carbon removal [2,13,14]. Zirconia has been largely studied as ceria
dopant on Ni-based catalysts for methane reforming reactions
techniques such as H or CO chemisorption are not recommended for
2
these catalysts due to the adsorption of both gases on ceria [20]. Fur-
thermore, determining Pt dispersion by transmission electron micro-
scopy is also challenging, because of the low contrast between metal
and support. Therefore, the Pt dispersion was determined by a corre-
lation between the rate of cyclohexane dehydrogenation and platinum
dispersion of a catalyst with known dispersion [21,22]. Cyclohexane
dehydrogenation was performed in a fixed bed reactor at 533 K and
atmospheric pressure. The reactants were fed into the reactor by bub-
[
7,13,15]. However, there are only a few papers investigating the effect
of the type of the ceria dopant on the performance of catalyst for me-
thane reforming reactions. Gaudillère et al. [16] compared the perfor-
mance of Ni supported on ceria doped with Zr, Pr, and Gd for the
combined reforming of methane at different temperatures. At 600 °C,
the Gd-doped catalyst exhibited higher activity than the catalyst con-
taining Zr. However, one of the main disadvantages of ceria and ceria-
mixed oxides is the low surface area, which leads to low metal dis-
persion and then favors the deposition of carbon. One approach to in-
crease the surface area of ceria and ceria-mixed oxides is to deposit
them over a high-surface area oxide such as alumina [17–19]. Faria
bling H
2
through a saturator containing cyclohexane at 285 K (H /
2
C H12 molar ratio of 13.6). At these conditions, no mass transfer or
6
equilibrium limitations were observed. The conversions were kept
below 10%. The exit gases were analyzed using a gas chromatograph
(Agilent 7890 A) equipped with a flame ionization detector and a HP-
INNOWAX column.
Three different experiments were carried out. In the first one, the
catalysts (45 mg) were reduced at 773 K for 1 h and then they were
cooled to 533 K under N . The cyclohexane dehydrogenation reaction
2
−1
was carried out at 533 K and WHSV = 120 h . For the second ex-
et al. [7] studied the performance of Ni/Al
2
O
3
and Ni/Ce
x
Zr1−x
O
2
2
/
/
periment, after reduction at 773 K for 1 h, the catalysts were heated to
Al
Al
2
O
2
O
3
3
(x = 0.5; 0.75; 1.0) catalysts for the DRM. Ni/Ce0.75Zr0.25
O
the reaction temperature (1073 K) under N . Then, the catalysts were
2
and Ni/Ce0.50Zr0.50
O
2
/Al
2
O
3
catalysts exhibited the lowest
cooled to 533 K under N and then, the cyclohexane dehydrogenation
2
amount of carbon formed, which was attributed to the higher oxygen
storage/release capacity of the ceria-zirconia mixed oxides. However,
there are no works in the literature about the performance of Gd-doped
ceria supported on alumina as catalyst support for the DRM.
reaction was carried out under the same conditions previously de-
scribed. The third experiment involved the same procedure previously
described but the dry reforming reaction was performed. After reaction
at 1073 K for 1 h, the reaction was stopped and H was passed through
2
Therefore, the aim of this work is to study the performance of Pt
supported on cerium oxide doped with gadolinium deposited on alu-
mina for DRM. The samples were characterized by in situ X-ray dif-
fraction (XRD), in situ X-Ray absorption spectroscopy (XAS), dehy-
drogenation of cyclohexane reaction and cyclohexanol dehydration.
The reasons for catalyst deactivation were investigated.
the catalyst at the reaction temperature for 30 min to remove any
carbon deposits present. Then, the reactor was cooled to 533 K under
H and the dispersion was measured by cyclohexane dehydrogenation
2
reaction.
Cyclohexanol dehydration reaction was used to measure the alu-
mina surface that is not covered by ceria mixed oxide [23]. The ex-
periments were performed using a fixed-bed quartz reactor at 1 atm and
523 K. The catalysts were first heated under air flow (30 mL/min) at
2
. Experimental
6
73 K for 1 h and maintained at this temperature for 30 min under He.
2.1. Catalysts preparation
Then, the samples were cooled to 523 K under He and the reactant
mixture was obtained by flowing He (100 mL/min) through a saturator
containing cyclohexanol, which was maintained at 340 K. The reaction
products were analyzed by GC (Agilent Technologies 7890 A) using an
HP-Innowax capillary column and aflame-ionization detector (FID).
The dehydration rate was calculated by the sum of cyclohexene, cy-
clohexane and benzene yields.
The alumina was prepared by calcination of bohemite (Puralox,
Condea) at 1073 K for 6 h in a muffle furnace. The samples were pre-
pared with 24 wt.% of ceria mixed oxides in order to cover the alumina
surface and Ce/Gd ratios of 4 and 1. One sample was synthesized with a
larger amount of ceria mixed oxide (36 wt.%) and this was indicated in
the nomenclature of the sample. For the supports, the alumina was co-
impregnated with an aqueous solution of cerium (IV) ammonium ni-
trate and the dopant precursor salt (nitrate of gadolinium). The samples
were dried at 373 K for 24 h and then calcined in air flow at 1073 K for
The samples were analyzed by X-ray Photoelectron Spectroscopy
(XPS) using a Hemispherical Energy Analyzer (Specs Phoibos 150),
equipped with an AlKα (1486.6 eV) unmonochromatic source at base
−10
pressures less than 10
mbar. All spectra were charge-compensated
5
h. The catalysts were synthesized by incipient wetness impregnation
of the supports with an aqueous solution of H PtCl (Sigma-Aldrich) to
obtain 1 wt.% platinum. The samples were dried at 373 K and calcined
under air (50 mL/min) at 673 K for 3 h. Pt/Al was also prepared as
, Pt/
by setting the binding energy of the C1s peak to 284.6 eV. Spectra were
fitted using CasaXPS Version 2.3.16 (Casa Software Ltd., Cheshire, UK).
In situ X-Ray powder diffraction (XRD) measurements were per-
formed at the beamline XRD1-D12 A [24] of the Brazilian Synchrotron
Light Laboratory (LNLS), Campinas, Brazil. The sample was loaded into
a quartz capillary of 1.0 mm diameter between two quartz wool beds.
The capillary was placed in a reaction cell connected to a 3-circle Heavy
2
6
2 3
O
reference. The following samples were synthesized: Pt/Al
2 3
O .
3
O
Ce0.8Gd0.2/Al
2
O
3
; Pt/36Ce0.8Gd0.2/Al
2
O
3
; Pt/Ce0.5Gd0.5/Al
2
®
2.2. Catalyst characterization
Duty diffractometer from Newport and oriented horizontally and
perpendicularly to the X-ray beam. A Yaskawa-Motoman robotic arm
was used to hold a hot air blower above the sample to control the
temperature during the experiment. The diffraction patterns in a 2θ
range between 10 and 120° with an acquisition time of 150 s were
The BET surface areas of the samples were measured using a
Micromeritics ASAP 2020 analyzer by nitrogen adsorption at 77 K.
Before the measurements, the samples (300 mg) were dried at 373 K for
®
2
4 h and degassed at 623 K for 1 h.
obtained using an array of 24 Mythen (Dectris ), installed in the delta
XRD data were collected on a Bruker ASX D8 diffractometer with
circle at 760 mm from the sample. The wavelength of λ = 1.0332 Å was
selected by a double-crystal Si (111) monochromator. λ and the dis-
tance of the sample to the detector were calibrated using Si (SRM 640d)
CuKa radiation (λ = 1.5406 Å, 40 kV, 40 mA) and a LynxEye uni-
directional detector. The 2θ range used was between 20 and 80°, using a
−
1
−1
scan rate of 0.05° step
and a scan time of 2 s step . The Scherrer
and Al
2
O (SRM 676a) powders NIST standards. The diffractograms
3
equation was used to estimate the crystallite size of CeO
2
.
were recorded while the sample was reduced under a flow of pure H
2
2

R.O. da Fonseca, et al.
Catalysis Today xxx (xxxx) xxx–xxx
(
8 mL/min) at from 298 to 773 K at a heating of 10 K/min, remaining at
diffractograms of all samples exhibited the lines typical of γ-alumina
phases at 2θ = 36.9°, 39.5°, 45.9° and 67.0° (PDF#10-0425). However,
the presence of θ and δ alumina phases could not be ruled out because
this temperature for 1 h.
in situ X-Ray absorption spectroscopy (XAS) experiments were per-
formed at the D08B-XAFS – 2 beamline of the Brazilian Syncrotron
Light Source (LNLS, Campinas). Typically, wafers containing pressed
mixtures of the sample with BN as a binder were then placed into a
sample holder located inside a quartz tubular reactor, which tempera-
ture was controlled by a furnace. XANES spectra were successively re-
their lines are too close to those of γ-Al
2
3
O (θ-alumina - PDF # 11-0517;
δ-alumina - PDF # 46–1131) [7]. In addition, the diffraction lines
corresponding to cerium oxide with a cubic structure (PDF # 34-0394)
were also observed for all samples containing the mixed oxide. The
diffraction lines of PtO
low metal content used.
The lattice parameter of ceria for the Pt/Ce
2
were not observed, which is likely due to the
corded around the Ce LIII edge (E = 5.723 keV), while the sample was
0
−1
reduced under a 5% H
2
/He mixture from 298 to 773 K (10 K.min ),
x
Gd1−xO
2
/Al
2
O samples
3
remaining at this temperature for 1 h; Data reducing was made using
the ATHENA software package. The background subtracted, normalized
XANES spectra were then trimmed at a suitable energy range (80 eV
ceria was calculated taking into account the line at 2θ = 55.74° (Table
S1). According to Mc Bride et al. [26], more accurate lattice constant
values are obtained using the high angular range. The lattice parameter
increased with the addition of dopant. For instance, the lattice para-
meter varied from 5.4112 Å for ceria to 5.4126 Å for the sample con-
taining 20% of Gd content. Further increase in the Gd content to 50%
increased the lattice parameter to 5.4245 Å. Borchert et al. [27] also
used the line at around 2θ = 57° to calculate the lattice parameter of
4
+
3+
above the absorption edge). The fraction of Ce and Ce during the
different treatments of the sample was estimated by linear combination
of normalized Ce LIII edge XANES spectra (LC-XANES) within the
5
.70–5.76 keV energy range, using CeO and cerium acetate as refer-
2
4
+
3+
ences for Ce
and Ce , respectively.
TPO experiments were carried out in a micro-reactor coupled to a
quadrupole mass spectrometer (Pffeifer, Prisma Plus). The TPO analyses
of carbonaceous deposits were used to determine the amount of carbon
that was deposited under reaction conditions. After reaction, the sam-
Gd
x
Ce1-x
O
2-x samples. The lattice parameter increased with the dopant
) to 5.435 Å (Gd0.5Ce0.5 2-x), indicating the
ions into the ceria lattice.
content from 5.411 Å (CeO
2
O
3
+
incorporation of the relatively large Gd
The same result was reported by [28]. Therefore, the lattice parameter
3
+
ples were cooled to room temperature in He. Then, the samples were
calculated in our work suggests the insertion of Gd
ceria lattice.
cations into the
3
heated at a rate of 10 K/min in a 20% O
2
:He mixture (65 cm /min) up
to 1173 K and the CO
mass spectrometer.
2
evolution was monitored by the quadrupole
The crystallite size of CeO
2
for all samples calculated by the
Scherrer equation is reported in Table 1. For the samples containing
4 wt.% of mixed oxide, the crystallite size of CeO was around 8–9 nm.
The increase in the mixed oxide content (36 wt.%) slightly increased
the crystallite size of CeO (12 nm), which is likely due to the lower
2
2
2
.3. Dry reforming of methane
2
CO
2
reforming of methane was performed in a quartz reactor at
surface area of this catalyst.
atmospheric pressure for all catalysts studied at 1073 K. The catalyst
samples were diluted with SiC (SiC/catalyst ratio of 1.5) in order to
The in situ diffractograms obtained during reduction under pure
hydrogen from 298 to 773 K are shown in Fig. 2. The diffractogram of
avoid temperature gradients. The samples were reduced under H
2
Pt/Ce0.8Gd0.2/Al
2
O sample at room temperature displays the char-
3
acteristics lines of CeO with cubic structure previously discussed.
2
(
30 mL/min) at 773 K for 1 h and then heated to the reaction tem-
perature (1073 K) under N . A reactant mixture with a CH :CO ratio of
2
4
2
These lines shifted toward lower angles when the temperature was in-
creased to 574 K. Further increase in the temperature to 773 K caused a
continuous shift of these lines. This is evidenced on the lattice para-
meter that varied from 5.4122 Å (298 K) to 5.5145 Å (773 K) (Table S1).
Fig. 3 shows the lattice parameter calculated as a function of reduction
−
1
1
and a flow rate of 100 mL min
was used. The exit gases were
analyzed using a gas chromatograph (Agilent 6890) equipped with a
thermal conductivity detector and a Carboxen 1010 column (Supelco).
temperature for all Pt/Ce
x
Gd1−xO
2
/Al
2
O samples. An experiment was
3
3
. Results and discussion
carried out with ceria under air to determine the effect of the thermal
expansion. The linear increase on the lattice parameter up to 773 K
observed for this sample is attributed to the thermal expansion of the
3
.1. Catalyst characterization
samples. However, the curves obtained for all Pt/CeGd/Al
2
O samples
3
Table 1 lists the values of BET surface areas obtained for the sup-
2
deviates from this linear increase on the lattice parameter after 400 K,
ports and catalysts. A large decrease in the alumina area (163 m /g)
was observed with the addition of mixed oxides, which is likely due to
the blockage of alumina pores by the deposition of doped cerium oxides
indicating the beginning of ceria reduction. Furthermore, a sharp in-
crease on the lattice parameter was observed for Pt/Ce0.8Gd0.2/Al
O
2 3
catalyst at around 723 K. The non-linear increase of the lattice para-
meter with the reduction temperature is likely due to the change in the
[
24,25]. The reduction of the surface area was more significant for the
support with higher content of CeGd (36 wt.%). However, no significant
changes were observed in the surface areas of the supports with the
addition of Pt.
4
+
3+
4
3+
oxidation state of ceria from Ce
to Ce . Ce
ions have a higher
+
ionic radius (1.034 Å) as compared to the Ce ions (0.92 Å). Then, the
3
+
introduction of Ce
ions into ceria causes a change in the Ce-O bond
The diffractograms of calcined samples are shown in Fig. 1. The
length, which produces a lattice expansion [29,30]. Crystallite size of
CeO at different reduction temperatures was also calculated and the
2
Table 1
results are reported in Table S2. Crystallite size of CeO
2
for all samples
BET surface area of supports and catalysts, crystallite size of CeO
2
calculated
3
+
through X-ray diffraction and Ce
molar fraction calculated by the linear
remained unchanged during reduction up to 773.
combination of Ce LIII-edge XANES spectra of the catalysts.
The X-ray photoemission spectra (XPS) for the Pt4d, Ce3d and 3d5/2
2
3+
peaks for Pt/Ce0.8Gd0.2/Al
2
O
3
and Pt/Ce0.5Gd0.5/Al
2
3
O are shown in
Catalysts
Surface Area (m /
d
CeO2
(nm)
Ce
molar fraction
Fig. S1. The Pt4f peaks were not used due to the superposition with the
g)
(%)
line characteristic of Al 2p. The peak at 316.9 and 315.8 eV, for Pt/
Pt/Al
2
O
3
157 (163)a
120 (113)
94 (93)
–
–
Ce0.8Gd0.2/Al
2
O
3
and Pt/Ce0.5Gd0.5/Al
2
O respectively, are assigned to
3
Pt/Ce0.8Gd0.2/Al
2
O
3
2
O
3
9
17
11
12
platinum oxide PtO (Fig. 4a) [31,32]. Shyu and Otto [31] reported a
x
Pt/36Ce0.8Gd0.2/Al
3
12
8
shift to lower binding energies in Pt-alumina catalysts containing ceria
Pt/Ce0.5Gd0.5/Al
2
O
114 (109)
indicating that Pt on the surface of alumina is present as PtO
2
, but the
a
+2
The values in the parenthesis corresponds to the BET surface area of the
interaction between Pt and CeO results in the formation of Pt , at-
2
supports.
tributed to PtO. The observed shift in binding energies for Pt/
3

R.O. da Fonseca, et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 1. X-ray diffraction patterns obtained for (a) Pt/Al
2
O
3
; (b) Pt/Ce0.8Gd0.2/Al
2
3
O ; (c) Pt/Ce0.5Gd0.5/Al
2
O
3
; (d) Pt/36Ce0.8Gd0.2/Al
2
3
O .
Ce0.5Gd0.5/Al
2
O catalyst may be the consequence of a greater interac-
The Pt dispersion was calculated using the dehydrogenation of cy-
clohexane (DHO) as a probe reaction and the results are reported in
tion with the ceria. The Ce3d spectra of both samples exhibits the in-
tense u”’ satellite at 917.0 eV, which indicates that the spectra are
Table 2. The Pt dispersion of the fresh Pt/Al
2
O catalyst (reduction at
3
4
+
3+
dominated by the Ce chemical state. The fraction of Ce species for
773 K) was 48% but it decreased to 42% after heating the sample to
1073 K and a strong reduction in the dispersion was observed during
DRM reaction (29%). A significant sintering of the Pt particles was
detected for all catalysts during heating the fresh catalysts to the re-
action temperature. The decrease in dispersion during reaction also
both samples was 8.4 and 6.9% for Pt/Ce0.8Gd0.2/Al
2
O
3
and Pt/
Ce0.5Gd0.5/Al
2 3
O
respectively. The Gd 3d5/2 spectrum of both samples
presents a peak at 1187.7 eV spin-orbit splitting of 32.0 eV assigned to
3
+
Gd
[33]. The Ce/Gd intensities ratio are reported in Table S3. The
Ce/Gd ratio was 4.89 and 1.79 for Pt/Ce0.8Gd0.2/Al
2
O
3
and Pt/
occurred for Pt/Ce0.8Gd0.2/Al
2
O
3
and Pt/Ce0.5Gd0.5/Al
2
O catalysts.
3
Ce0.5Gd0.5/Al
samples.
2
O
3
respectively, indicating a ceria segregation on both
The dispersion remained low and unchanged for Pt/36Ce0.8Gd0.2/Al
2 3
O
catalysts during DRM.
The XANES spectra at Ce LIII-edge of Pt/Ce0.8Gd0.2/Al
2
O
3
recorded
The cyclohexanol dehydration was used to evaluate the degree of
coverage of the alumina surface by taking into account that the number
during reduction under 5% H /He mixture from 293 to 773 K are
2
shown in Fig. 4. The XANES spectra of the other catalysts are presented
in the Supplementary information (Figs. S2,S3). The XANES spectra at
specific temperatures are also presented in Fig. 4b. In general, the
of acid sites of a Pt/CeO catalyst is very low and the cyclohexanol
2
dehydration reaction rate for ceria is negligible. Then, this reaction
measures only the acid sites on the exposed alumina surface that is not
covered by the mixed oxide. In our work, the reaction rate of dehy-
dration of cyclohexanol for the catalysts under different treatments is
4
+
XANES spectra of ceria exhibit four peaks attributed to Ce (A and B),
3
+
Ce (C) and a pre-edge peak (D) [34]. In our work, A and B were the
most intense peaks in the spectra of the calcined samples, suggesting
listed in Table 2. Pt/Al
2
O exhibited the highest rate of dehydration of
3
4
+
−1 −1
that they contain mainly Ce
species. After reduction at 773 K, the
cyclohexanol (4.1 mmol. min . gcat ), suggesting a higher density of
XANES spectrum exhibits the appearance of a shoulder (peak C), in-
acid sites. The samples showed the following order for the rate of re-
4
+
3+
dicating that ceria was reduced. The conversion of Ce to Ce species
during reduction of all samples can be better evaluated by the calcu-
lation of the linear combination of Ce LIII-edge XANES spectra of the
action of cyclohexanol dehydration: Pt/Al
2
O
3
> Pt/Ce0.8Gd0.2/
Al
2
O
3
> Pt/Ce0.5Gd0.5/Al
2
O
3
> Pt/36Ce0.8Gd0.2/Al
2
O . This result re-
3
vealed that Pt/36Ce0.8Gd0.2/Al
2
O catalyst exhibits the highest degree
3
4
+
catalysts using references as shown in Fig. 5. The reduction of Ce to
of coverage of alumina surface.
3
+
Ce
73 K), regardless of the preparation method, which agrees very well
with the XRD results. However, the extent of reduction was low for all
species began approximately at the same temperature (around
4
3.2. Catalytic tests
3
+
catalysts, showing
a
fraction of Ce
species between 11–17%
The methane and CO2 conversions and H /CO molar ratio as a
2
(
Table 1), which is likely due to the low reduction temperature (773 K).
function of time on stream (TOS) obtained for the CO reforming of CH
2
4
3
+
The highest fraction of Ce species is obtained for Pt/Ce0.8Gd0.2/Al
2
O
3
over Pt/Al O and Pt/Ce Gd O /Al O catalysts are shown in Fig. 6.
2
3
x
1-x
2
2 3
catalyst, which was also observed on the in situ XRD experiments. The
The doped-ceria/alumina supports (without Pt) did not exhibit any
same result was reported by Sadykov et al. [28] for Ce1-xGd
x
O
2-y sam-
activity under this reaction conditions. For Pt/36Ce Gd /Al O
0
.8
0.2
2 3
ples (x = 0.1; 0.2; 0.3; 0.4; 0.5). The oxygen mobility at temperatures
up to 700 °C was determined by using the isotope exchange under dy-
namic conditions. The maximum oxygen mobility was observed for the
sample containing 20% of Gd due to clustering of Gd cations at higher
doping content [28,35].
catalyst, the methane and CO conversions remained constant during
2
24 h of TOS, whereas they strongly decreased for Pt/Al O catalyst at
2 3
the beginning of reaction. Pt/Ce Gd /Al O and Pt/Ce Gd /Al O
0
.8
0.2
2
3
0,5
0,5
2 3
catalysts underwent deactivation after 24 h of TOS. The H /CO molar
2
ratio was in the range between 0.9 – 0.96 during the first 2 h of reaction
4

R.O. da Fonseca, et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 3. Lattice parameter of CeO
different temperatures for: (■) Pt/Ce0.8Gd0.2/Al
◄) Pt/36Ce0.8Gd0.2/Al and (▼) CeO
2
calculated during reduction under pure H
2
at
2
O
3
, (▲) Pt/Ce0.5Gd0.5/Al
2
O ,
3
(
2
O
3
2
.
Fig. 4. (a) in situ XANES spectra of Pt/Ce0.8Gd0.2/Al
2
O during reduction at
3
7
73 K under 5% H2/He mixture; (b) XANES spectra at specific temperatures:
fresh sample(blue curve); after reduction at 773 K for 1 h (green curve); and
references compounds.
Fig. 2. X-ray diffraction patterns obtained during reduction from room tem-
perature to 773 K for (a) Pt/Ce0.8Gd0.2/Al
2
O
3
; (b) Pt/36Ce0.8Gd0.2/Al
2
3
O ; (c)
Pt/Ce0.5Gd0.5/Al
2 3
O .
5

R.O. da Fonseca, et al.
Catalysis Today xxx (xxxx) xxx–xxx
3.3. Characterization of post-reaction catalysts
The decrease in the activity during DRM reaction is generally as-
sociated with catalyst deactivation mainly due to carbon deposition or
metal sintering. In order to investigate the formation of carbon during
the reaction, the post-reaction catalysts were characterized by TPO
analysis.
The TPO profiles of the used catalysts after DRM for 24 h of TOS are
shown in Fig. 7. The TPO profile of all catalysts exhibited a peak at high
temperature between 700–1000 K. However, the intensity of this peak
was higher for Pt/Al
Al and Pt/Ce0.5Gd0.5/Al
tween 550–650 K. The amount of carbon deposited during the DRM for
4 h of TOS measured by TPO are reported in Table 3. The highest
amount of carbon deposited was observed for Pt/Al . All Pt/CeGd/
Al catalysts showed lower carbon formation during the DRM.
2
O
3
catalyst. Furthermore, the TPO profile of Pt/
2
O
3
2
O catalysts also show a small peak be-
3
2
O
2 3
O
2 3
3.4. Catalyst deactivation during DRM
Pt/Al catalyst significantly deactivated during the first hour of
2 3
O
reaction while the Pt/Ce0.8Gd0.2/Al and Pt/Ce0.5Gd0.5/Al ex-
2
O
3
2
O
3
hibited a moderate drop in activity during 24 h of TOS. Increasing the
mixed oxide content to 36 wt% improved catalyst stability. The TPO
profiles revealed the formation of carbon over all catalysts. Therefore,
the decrease in methane conversion after 24 of TOS could be associated
with the deposition of carbon. However, the amount of carbon de-
posited on these catalysts was low (Table 3). Bitter et al. [36] reported
similar amount of carbon deposited on Pt/Al
2
O
3
, Pt/TiO
2
and Pt/ZrO
2
catalysts during DRM reaction. According to the authors, carbon de-
position on the metal particles was responsible by catalyst deactivation.
In our work, the cyclohexane dehydrogenation probe reaction de-
monstrated that Pt sintering occurred for all catalysts, except for Pt/
3
6Ce0.8Gd0.2/Al
2
O catalyst (Table 2). Then, Pt sintering also con-
3
tributes to catalyst deactivation for DRM. Bitter et al. [36] proposed
that the deactivation of Pt/Al
2
O
3
, Pt/TiO
2
and Pt/ZrO catalysts during
2
DRM was not caused by Pt sintering by carrying out hydrogen chemi-
sorption of the fresh, used and regenerated catalysts. In order to vali-
date our methodology of Pt dispersion measurement, the DRM was
performed over Pt/Ce0.5Gd0.5/Al
2
O catalyst during 1h and then, the
3
reaction was stopped and the same procedure used for the dispersion
measurement was carried out. Instead of performing the cyclohexane
dehydrogenation reaction, the reactor was cooled to 298 K and a TPO
was performed. The TPO profile did not reveal the presence of any
peak, indicating that carbon deposits were not formed during the re-
action for 1 h (Fig. S4). Then, the drop in the reaction rate of cyclo-
hexane dehydrogenation for Pt/Ce0.5Gd0.5/Al
2
O catalyst is not due to
3
the blockage of the surface of Pt particles by carbon deposits. This result
confirm that Pt sintering is also responsible for the deactivation of Pt-
based catalysts in our work.
However, the extent of Pt sintering of Pt/Al
2
O
3
catalyst was similar
catalysts,
to that one for Pt/Ce0.8Gd0.2/Al
2
O
3
and Pt/Ce0.5Gd0.5/Al
2 3
O
but Pt supported on alumina strongly deactivated. The TPO profiles of
the catalysts may provide the explanation for these results. The TPO
profile of Pt/Al
2
O catalyst showed two peaks, whereas the other cat-
3
alysts exhibited only one peak at high temperature region.
Fig. 5. Linear combination of XANES spectra of catalysts considering CeO and
2
Silva et al. [19] also observed the presence of two peaks at 870 and
cerium nitrate as references: (a) Pt/Ce0.8Gd0.2/Al
2
O ; (b) Pt/36Ce0.8Gd0.2/
3
1
003 K in the TPO profile of a Pt/Al
2
O catalyst after the partial oxi-
3
Al
2
O
3
; (c) Pt/Ce0.5Gd0.5/Al
2 3
O .
dation of methane at 1073 K. However, TPO analysis did not reveal the
deposition of carbon over a Pt/CeZrO /Al catalyst. The absence of
2
2 3
O
but it strongly decreased to 0.55 (Pt/Al
2
O
3
), 0.77 (Pt/Ce0.5Gd0.5/Al
after 24 h of TOS. For the Pt/
catalyst, the H /CO molar ratio remained un-
/CO molar
2
O
3
)
carbon deposits over Pt/CeZrO
2
/Al
2
3
O catalyst was attributed to the
and 0.80 (Pt/Ce0.8Gd0.2/Al
O )
2 3
mechanism of carbon removal promoted by the ceria-zirconia mixed
oxide.
3
6Ce0.8Gd0.2/Al
2
O
3
2
changed at around 0.95 during 24 h of TOS. The values of H
2
The TPO profile of Pt/Al
2
O catalyst after DRM during 1, 12 and
3
ratio lower than 1.0 indicates the occurrence of the reverse of the
water-gas shift reaction. Then, the decrease in the H /CO molar ratio
corresponds to the decrease in the extent of the DRM reaction.
2
4 h exhibited one peak above 840 K and a shoulder at 750 K [37].
2
Increasing the reaction time led to an increase in the intensity of both
peaks. On the other hand, no peaks were observed in the TPO profile of
6

R.O. da Fonseca, et al.
Catalysis Today xxx (xxxx) xxx–xxx
Table 2
Reaction rate of dehydrogenation of cyclohexane (DHO) at 533 K, Pt dispersion of fresh and used catalysts and reaction rate of dehydration of cyclohenaxol (DHA) at
5
23 K.
Rate of DHO (mmol. gPt1. min−1
−
)
D (%)a
Rate of DHA (mmol.gcat−1. min−1
)
Catalysts
Fresh
0.44
0.17
0.11
0.18
Used
0.30
0.13
0.11
0.10
Fresh
42
Used
29
Pt/Al
2
O
3
4.0
2.3
0.9
1.1
Pt/Ce0.8Gd0.2/Al
2
O
3
2
O
3
16
12
Pt/36Ce0.8Gd0.2/Al
3
11
11
Pt/Ce0.5Gd0.5/Al
2
O
17
10
a
Pt dispersion of the fresh catalysts calculated using the reaction rate of cyclohexane dehydrogenation.
Pt/ZrO
of the catalysts after the methane decomposition reaction. The TPO
profile of Pt/Al catalyst was quite similar to the one obtained after
2
catalyst after DRM. The authors also carried out TPO analysis
mixed oxide. Then, the close contact between Pt and the CeGd mixed
oxide promotes the mechanism of carbon removal from the surface of Pt
particles [6] as well as inhibit metal particle growth [38]. Further proof
for the role of Pt sintering on catalyst deactivation was obtained by
2 3
O
DRM, whereas a single peak was observed at 615 K for Pt/ZrO . Only
2
one peak was also observed in the TPO profile of unsupported metallic
Pt (620 K) and alumina (930 K). The authors attributed the low-tem-
perature peak to carbon deposited on Pt particles and the high-tem-
perature peak was associated with carbon on the support.
performing a series of experiments with the Pt/36Ce0.8Gd0.2/Al
2 3
O
catalyst, which did not show any Pt particle growth. In this case, the
DRM of methane was carried out at 1073 K for 24 h and then the re-
action was stopped and the reactor was cooled down to room tem-
perature. Then, the TPO of the used catalyst was performed to remove
all carbon residues from the catalyst surface. Finally, the catalyst was
reduced again at the same previous conditions and a new reaction was
started. The initial methane conversion of the second reaction was ex-
actly the same as the first one (65.4 and 65.7%, respectively). This
result confirms that Pt sintering plays a key role on catalysts deacti-
vation.
Bitter et al. [36] studied the deactivation of Pt/Al
2
O
3
, Pt/TiO and
2
Pt/ZrO catalysts during DRM reaction. They proposed that carbon is
2
deposited over both metal and support but deactivation is caused by
carbon formation on the metal particle. The carbon deposited on the
metal is oxidized at low temperatures, whereas high temperatures are
necessary to remove carbon from the support.
Noronha et al. [6] also observed the presence of two peaks in the
TPO profile of Pt/Ce
x
Zr
1
−
x
O
2
catalysts after CO
2
reforming of me-
In order to confirm that the first peak corresponds to carbon de-
posited on the Pt particles, the DRM reaction was carried out over Pt/
thane, with a more intense peak at low-temperature region. The TPO
profile of Pt/Al
2
O
3
catalyst showed only one peak in the high-tem-
Ce0.5Gd0.5/Al
2
O
3
catalyst under more severe conditions (CH
4
/CO
2
perature region. In situ XPS analysis of the used catalysts showed that
the different peaks in the TPO were not due to different forms of
carbon, but rather to different locations on the catalyst surface. The
peak at low temperature was attributed to the oxidation of carbon de-
posits close to the metallic particle that is easily oxidized, while the
high temperature peak was assigned to the oxidation of carbon on the
molar ratio of 3.0). Catalyst rapidly deactivated at the beginning of
reaction and the total amount of carbon increased (0.11 mg
CH converted.h) (Table 3). Increasing the CH /CO molar ratio to 3.0
increased 4-fold the amount of carbon deposited in comparison to the
C
/(gcat.mols
4
4
2
reaction with CH
4
/CO molar ratio of 1.0. In addition, the intensity of
2
the peak at around 550–600 K increased and a new intense peak ap-
peared at 750 K (Fig. S5). The increase of the intensity of the peaks in
the low temperature region corresponds to an increase in the amount of
support. The higher stability of Pt/Ce
x
Zr1 − x
2
O catalysts was attributed
to the higher oxygen mobility of ceria and ceria-zirconia mixed oxide,
which promoted the oxidation of carbon deposits at the metal-support
interface. For alumina supported catalyst, carbon is not continuously
removed during the reaction in the absence of this oxygen transfer
mechanism, and then it accumulates on the surface, resulting in catalyst
deactivation.
carbon deposits in the surface of Pt particles. Increasing the CH
4
/CO
2
molar ratio led to an unbalance between the reaction rate of methane
decomposition to H and C and the rate of reaction of carbon with
oxygen from the support, which leads to carbon accumulation on the
surface of metallic particle. This result demonstrate that catalyst de-
activation is likely due to carbon deposition on the Pt particle.
In our work, the TPO profile of used catalysts after DRM showed the
presence of the peak corresponding to carbon deposited on the Pt
particle for Pt/Al
2
O
3
catalyst. The absence of this peak for Pt/
4
. Conclusions
Ce
x
Gd1−x
O
2
/Al
2
3
O catalysts suggests that the mechanism of carbon
removal is operational. In fact, the in situ XRD and XANES experiments
showed that the reduction of CeGd mixed oxide at 773 K created oxygen
This work studied the deactivation of Pt/Al
2
O
3
and Pt/Ce
x
Gd1-x
O /
2
Al
2
O
3
catalysts during DRM reaction. Pt/Al
2
3
O catalyst strongly deac-
3
+
vacancies and Ce
species. Therefore, carbon formed by the dis-
tivated at the beginning of DRM reaction, whereas Pt/Ce0.8Gd0.2/Al
and Pt/Ce0.5Gd0.5/Al underwent less severe loss of activity during
4 h of TOS. Pt/36Ce0.8Gd0.2/Al catalyst remained quite stable
catalyst
2
O
3
sociation of methane on the surface of the metallic particle reacts with
the oxygen from the support at the metal-support interface, which re-
duces or inhibit the accumulation of carbon [6]. However, the peak
attributed to carbon deposited on the support is observed in the TPO
profile of all catalysts. In this case, the mechanism of carbon removal is
2 3
O
2
2 3
O
under this reaction conditions. TPO profile of used Pt/Al
O
2 3
after DRM revealed the formation of carbon deposits on the metal
particle and support. The addition of metal oxide suppressed the de-
position of carbon on the surface of Pt particles. This is likely due to the
mechanism of carbon removal promoted by the oxygen vacancies and
not effective even on the Pt/Ce
x
Gd1−x
O
2
/Al
2
3
O catalysts because it
takes place at the metal-support interface and then, carbon accumulates
on the support. But this carbon is not responsible for the continuous
decrease in the methane conversion during the reaction for all Pt/
3
+
Ce species created during the reduction of the mixed oxide as shown
by in situ XRD and XANES experiments. In this mechanism, the highly
mobile oxygen from the mixed oxide reacts with carbon formed by the
dissociation of methane on the surface of the metallic particle, in-
hibiting carbon accumulation. Therefore, the deactivation of Pt/
Ce
x
Gd1−x
O
2
/Al
2
3
O catalysts. For these catalysts, the deactivation is
likely due to the sintering of metallic particles during reaction. For Pt/
3
6Ce0.8Gd0.2/Al
2
3
O catalyst, there is no accumulation of carbon on the
metal particle and Pt sintering does not occur and then, catalyst does
not deactivate. This is likely due to the higher fraction of Pt located on
doped-ceria because of the larger coverage of alumina surface by the
Ce0.8Gd0.2/Al
2
O
3
and Pt/Ce0.5Gd0.5/Al
2
O catalysts was caused by Pt
3
sintering as demonstrated by the probe reaction of cyclohexane dehy-
drogenation. Pt/36Ce0.8Gd0.2/Al
2
3
O catalyst did not deactivate because
7

R.O. da Fonseca, et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 7. TPO profiles obtained after DRM at 1073 K.
Table 3
The amount of carbon deposited during the DRM at 1073 K for 24 h of TOS.
Catalysts
mg
C
/gcat.h
C 4
mg /gcat.mols CH converted.h
Pt/Al
2
O
3
0.45
0.20
0.22
0.07
0.39
0.07
0.07
0.03
Pt/Ce0.8Gd0.2/Al
Pt/36Ce0.8Gd0.2/Al
Pt/Ce0.5Gd0.5/Al
2
O
3
2
O
3
O
2 3
Acknowledgements
The authors acknowledge the scholarship and the financial support
received from CNPq (314018/2017-4) and CAPES (001). The group
thanks the LNLS for the assigned time at D08B-XAFS – 2 (20171082)
and XRD1-D12A (20171087) beamlines.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.cattod.2019.04.079.
Fig. 6. Methane and CO
2
conversion and H
2
/CO ratio versus TOS for CO
2
re-
References
forming of methane at 1073 K for the catalysts (♦) Pt/Al
2
O
3
, (■) Pt/Ce0.8Gd0.2/
Al
2
O
3
, (◄) Pt/36Ce0.8Gd0.2/Al
2
O
3
(▲) Pt/Ce0.5Gd0.5/Al
2
O .
3
[
[
1] L.A. Rodrigues, I.L. Sauer, Energy Policy 87 (2015) 486–495.
2] R.O. Fonseca, A.A.A. Silva, M.R.M. Signorelli, R.C.R. Neto, F.B. Noronha,
R.C. Colman, L.V. Mattos, J. Brazilian Chem. Soc. 25 (2014) 2356–2363.
3] H. Ay, D. Uner, Appl. Catal. B 179 (2015) 128–138.
of the absence of carbon deposition and Pt sintering. The higher cov-
erage of alumina surface by the mixed oxide favors the intimate contact
between Pt and the CeGd mixed oxide, promoting the mechanism of
carbon removal from the surface of Pt particles and stabilizing the Pt
particles.
[
[4] J.H. Edwards, A.M. Maitra, Fuel Processing Technol. 42 (1995) 269–289.
[
[
5] M.C.J. Bradford, M.A. Vannice, J. Catal. 173 (1998) 157–171.
6] F.B. Noronha, E.C. Fendley, R.R. Soares, W.E. Alvarez, D.E. Resasco, Chem. Eng. J.
8
2 (2001) 21–31.
[7] E.C. Faria, R.C.R. Neto, R.C. Colman, F.B. Noronha, Catal. Today 228 (2014)
38–144.
1
8

R.O. da Fonseca, et al.
Catalysis Today xxx (xxxx) xxx–xxx
[
[
8] A.N.J. Vankeuken, M.E.S. Hegarty, J.R.H. Ross, P.F. Vandenoosterkamp, Studies
Surf. Sci. Catal. 107 (1997) 537–546.
[24] A. Vazquez, T. Lopez, R. Gomez, X. Bokhimi, J. Mol. Catal. A: Chemistry 167 (2001)
91–99.
9] K. Seshan, H.W. Tenbarge, W. Hally, A.N.J. Vankeulen, J.R.H. Boss, Studies Surf.
Sci. Catal. 81 (1994) 285–290.
[25] N.D. Charisiou, G. Siakavelas, K.N. Papageridis, A. Baklavaridis, L. Tzounis, J.
Natural Gas Sci. Eng. 31 (2016) 164–183.
[
10] J.A. Lercher, J.H. Bitter, W. Hally, W. Niessen, K. Seshan, Studies Surf. Sci. Catal.
[26] J.R. McBride, K.C. Hass, B.D. Poindexter, W.H.J. Weber, Appl. Phys. 76 (1994)
2435.
1
01 (1996) 463–472.
[
[
11] J.R. Rostrup-Nielsen, B. Hansen, J. Catal. 144 (1993) 38–49.
[27] H. Borchert, Y. Borchert, V.V. Kaichev, I.P. Prosvirin, G.M. Alikina,
A.I. Lukashevich, V.I. Zaikovskii, E.M. Moroz, E.A. Paukshtis, V.I. Bukhtiyarov,
V.A. Sadykov, J. Phys. Chem. B 109 (2005) 20077–20086.
12] S.M. Stagg-Williams, F.B. Noronha, G. Fendley, D.E. Resasco, J. Catal. 194 (2000)
2
40–249.
[
[
13] J. Kaspar, P. Fornasiero, M. Graziani, Catal. Today (1999) 285–298.
14] B.L. Augusto, L.O. Costa, F.B. Noronha, R.C. Colman, L.V. Mattos, Int. J. Hydrogen
Energy 37 (2012) 12258–12270.
[28] V.A. Sadykov, Y.V. Frolova, G.M. Alikina, A.I. Lukashevich, V.S. Muzykantov,
V.A. Rogov, E.M. Moroz, D.A. Zyuzin, V.P. Ivanov, H. Borchert, E.A. Paukshtis,
V.I. Bukhtiyarov, V.V. Kaichev, S. Neophytides, E. Kemnitz, K. Scheurell, React
Kinet. Catal. Lett. 85 (2005) 367–374.
[
[
[
[
[
15] A. Kambolis, H. Matralis, A. Trovarelli, C.H. Papadopoulou, Appl. Catal. A: General
3
77 (2010) 16–26.
[29] A.A.A. Da Silva, N. Bion, F. Epron, S. Baraka, F.C. Fonseca, R.C. Rabelo-Neto,
L.V. Mattos, F.B. Noronha, Appl. Catal. B: Environmental 206 (2017) 626–641.
[30] A. Bonk, A. Remhof, A.C. Maier, M. Trottmann, M.V.F. Schlupp, C. Battaglia,
U.F. Vogt, J. Phys. Chem. C 120 (2015) 118–125.
16] C. Gaudillère, P. Vernoux, C. Mirodatos, G. Caboche, D. Farrusseng, Catal. Today
57 (2010) 263–269.
17] P.P. Silva, F.A. Silva, L.S. Portela, L.V. Mattos, F.B. Noronha, C.E. Hori, Catal. Today
07–108 (2005) 734–740.
18] L.V. Mattos, E.R. Oliveira, P.D. Resende, F.B. Noronha, F.B. Passos, Catal. Today 77
1
1
[31] J.Z. Shyu, K. Otto, J. Catal. 115 (1989) 16–23.
[32] M.J. Tiernan, O.E. Finlayson, Appl. Catal. B: Environmental 19 (1998) 23-3.
[33] D. Raiser, J.P. Deville, J. Electron Spectros. Relat. Phenomena 57 (1991) 91–97.
[34] M.C. Ribeiro, R.C. Rabelo-Neto, L.V. Mattos, G. Jacobs, B.H. Davis, F.B. Noronha, J.
Phys. Chem. C 118 (2014) 28007–28016.
(
2002) 245–256.
19] F.A. Silva, D.S. Martinez, J.A.C. Ruiz, L.V. Mattos, C.E. Hori, F.B. Noronha, Appl.
Catal. A: General 335 (2008) 145–152.
[
[
20] P. Pantu, G.R. Gavalas, Appl. Catal. A: General 223 (2002) 253–260.
21] F.B. Passos, E.R. Oliveira, L.V. Mattos, F.B. Noronha, Catal. Today 101 (2005)
[35] T. Ohashi, S. Yamazaki, T. Tokunaga, Y. Arita, T. Matsui, T. Harami, K. Kobayashi,
Solid State Ion. 113–115 (1998) 559–564.
2
3–3030.
[36] J.H. Bitter, K. Seshan, J.A. Lercher, J. Catal. 183 (1999) 336–343.
[37] K. Nagaoka, K. Seshan, K. Aika, J.A. Lercher, J. Catal. 197 (2001) 34–42.
[38] R.C. Rabelo-Neto, H.B.E. Sales, C.V.M. Inocêncio, E. Varga, A. Oszko, A. Erdohelyi,
F.B. Noronha, L.V. Mattos, Appl. Catal. B: Environmental 221 (2018) 349–361.
[
[
22] P.P. Silva, F.A. Silva, H.P. Souza, A.G. Lobo, L.V. Mattos, F.B. Noronha, C.E. Hori,
Catal. Today 101 (2005) 31–37.
23] D. Martin, D. Duprez, J. Mol. Catal. A: Chemistry 118 (1997) 113–128.
9