Oxidation of lean methane over cobalt catalysts supported on ceria/alumina

Article abstract of DOI:10.1016/j.apcata.2019.117381

Cobalt (30 %wt.) oxide catalysts supported over ceria-modified (3–18 %wt.) alumina supports were examined for the combustion of lean methane. The prepared samples were characterised by wavelength dispersive X-ray fluorescence, N₂ physisorption, X-ray diffraction, Raman spectroscopy, X-ray photoelectron spectroscopy, transmission electron microscopy, temperature-programmed reduction with hydrogen and temperature-programmed reaction with methane. A significant enhancement of the activity was evidenced with respect to the catalyst supported over bare alumina, with an optimal cerium loading of 12 %wt. and a resultant T₅₀ value of 480 °C. The ceria modification was found to induce a dual positive effect on the performance of the cobalt catalysts. On one hand, it acted as a physical barrier between deposited cobalt and alumina, thus inhibiting the cobalt-alumina interaction and the subsequent cobalt aluminate formation. On the other hand, the insertion of cerium ions in the spinel lattice led to a distortion of the structure that in turn resulted in an enhanced mobility of the oxygen species. The optimised catalyst exhibited a relatively good thermal stability for prolonged reaction time intervals (150 h) under dry conditions. The presence of water vapour markedly affected the catalytic performance although this negative effect was partially reversible.

Full text of DOI:10.1016/j.apcata.2019.117381

AppliedCatalysisA,General591(2020)117381
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Oxidation of lean methane over cobalt catalysts supported on ceria/alumina
Andoni Choya, Beatriz de Rivas, Juan Ramón González-Velasco, Jose Ignacio Gutiérrez-Ortiz,
Rubén López-Fonseca*
Chemical Technologies for Environmental Sustainability Group, Department of Chemical Engineering, Faculty of Science and Technology, University of The Basque Country
UPV/EHU, P.O. Box 644, E-48080 Bilbao, Spain
A R T I C L E I N F O
A B S T R A C T
Keywords:
Cobalt (30 %wt.) oxide catalysts supported over ceria-modified (3–18 %wt.) alumina supports were examined
for the combustion of lean methane. The prepared samples were characterised by wavelength dispersive X-ray
fluorescence, N2 physisorption, X-ray diffraction, Raman spectroscopy, X-ray photoelectron spectroscopy,
transmission electron microscopy, temperature-programmed reduction with hydrogen and temperature-pro-
grammed reaction with methane. A significant enhancement of the activity was evidenced with respect to the
catalyst supported over bare alumina, with an optimal cerium loading of 12 %wt. and a resultant T50 value of
480 °C. The ceria modification was found to induce a dual positive effect on the performance of the cobalt
catalysts. On one hand, it acted as a physical barrier between deposited cobalt and alumina, thus inhibiting the
cobalt-alumina interaction and the subsequent cobalt aluminate formation. On the other hand, the insertion of
cerium ions in the spinel lattice led to a distortion of the structure that in turn resulted in an enhanced mobility
of the oxygen species. The optimised catalyst exhibited a relatively good thermal stability for prolonged reaction
time intervals (150 h) under dry conditions. The presence of water vapour markedly affected the catalytic
performance although this negative effect was partially reversible.
Methane oxidation
Cobalt oxide
Modified alumina
Ceria
Lattice distortion
1. Introduction
and redox properties of the spinel as intact as possible.
The most commonly applied enhancing strategy is to support the
Cobalt-based catalysts have been largely used as oxidation catalysts
in numerous environmental applications including VOC abatement and
combustion of soot [1,2]. In this context, the control of emissions from
natural gas vehicles (unburned methane) over this class of metal oxides
has been also explored [3–5]. The use of cobalt catalysts with Co3O4 as
the main active phase is frequently considered as a cheaper viable al-
ternative to expensive noble-metal based catalysts with a proven higher
specific activity [6]. The notable activity of this metal oxide is related to
its good redox properties, and in particular, the easiness that the cobalt
ions possess to switch between its two oxidation states (+3 and +2),
which provides oxygen species in the lattice with a high mobility [7].
Since the oxidation of methane by these type of catalysts is deemed to
occur via a Mars-van Krevelen mechanism, where the oxygen species
from the spinel lattice are responsible for the oxidation reaction [8,9],
these redox properties are crucial for the activity of Co3O4-based cat-
alysts, to the point that even catalysts with poor textural and structural
properties can exhibit a relatively good performance [10,11]. Conse-
quently, when proposing an efficient strategy for designing highly ac-
tive Co3O4 catalysts it is extremely important to maintain the structure
cobalt oxide over the surface of a porous media, in order to disperse it
and increase the amount of the available active surface area. Generally,
this option produces catalysts with high specific surface areas and a
small crystallite size of Co3O4, but it presents a major drawback in the
form of a strong cobalt-support interaction that often negatively affects
the redox properties of the cobalt oxide [12,13]. More specifically,
when the support is alumina, this cobalt-support interaction usually
provokes a partial reduction of Co3+ ions to Co2+ and their subsequent
fixation in the alumina lattice, which leads to the formation of cobalt
aluminate (CoAl2O4) [14]. The cobalt ions fixed in this phase lose al-
most all their mobility and cannot revert to the +3 oxidation state at
moderate temperatures (< 600 °C), and therefore are rendered inactive
for the oxidation reaction [15].
A possible solution to this problem is to modify the alumina support
with a metallic promoter, prior to the deposition of cobalt species, in
order to increase its stability and reduce its propensity to interact with
the cobalt oxide deposited over it. This promoter can be added during
the synthesis of the alumina support itself. Thus, Liotta et al. [16] found
that adding Ba during the synthesis of Al2O3 by a sol-gel method
⁎ Corresponding author.
E-mail address: ruben.lopez@ehu.eus (R. López-Fonseca).
https://doi.org/10.1016/j.apcata.2019.117381
Received 5 November 2019; Received in revised form 10 December 2019; Accepted 20 December 2019
Availableonline23December2019
0926-860X/©2019ElsevierB.V.Allrightsreserved.

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
inhibited solid state diffusion of the Co2+ ions into the alumina after Co
deposition. On the other hand, Cheng et al. [17] reported that the in-
corporation of a fourth element during the synthesis of an alumina
supported copper-cobalt catalyst improved the reducibility of both
metal cations, especially when that fourth element was either Mn or Fe.
Alternatively, the promoter can be deposited over the surface of a as-
synthesised alumina before the deposition of cobalt. In this sense, Park
et al. [18] and Park et al. [19], on different studies, observed that the
addition of P to Al2O3 resulted in the partial formation of AlPO4, which
suppressed the formation of CoAl2O4. This inhibition effect was also
found for other metallic promoters such as Mg or Zr [20,21]. In all
cases, the deposition of the promoter over the alumina favoured a co-
balt-promoter interaction at the cost of a cobalt-alumina interaction.
However, this interaction does not always work in favour of the redox
properties of the cobalt oxide. In this sense, previous investigations on
the effect of surface protection of Al2O3 with Mg revealed that the re-
sulting Co-Mg interaction led to the formation of a low reducibility
CoO-MgO solid solution, as reported by Ulla et al. [22] and Ji et al.
[23].
isothermal steps: an initial ramp from room temperature to 125 °C at
5 °C min−1, a second ramp up to 300 °C at 1 °C min−1, and a final ramp
at 5 °C min−1 up to 600 °C, temperature that was then maintained for
4 h.
2.2. Characterisation techniques
Textural properties were determined from the nitrogen-adsorption
isotherms at −196 °C obtained with a Micromeritics TriStar II appa-
ratus. The specific surface of the samples was obtained by the BET
method, while the pore volume and the pore size distributions were
estimated using the BJH method. All samples were degassed prior to
analysis on a Micromeritics SmartPrep apparatus at 300 °C for 10 h with
a N2 flow.
The composition was determined by Wavelength Dispersive X-Ray
Fluorescence (WDXRF). From each sample in powder form, a boron
glass pearl was prepared by fusion in an induction micro-furnace, by
mixing the sample with the flux agent Spectromelt A12 (Merck) in an
approximate proportion of 20:1. Chemical analysis of each pearl was
Considering all the above-mentioned background, and based on our
previous results on the beneficial effect of cerium doping of Co3O4 bulk
catalysts [24], this work proposes the design of highly active Co/Al2O3
catalysts obtained by a previous incorporation of ceria onto the support
prior to cobalt precipitation. The premise supporting this hypothesis is
that the deposited CeO2 could have a twofold function as a physical
barrier between cobalt and alumina, thus inhibiting the formation of
CoAl2O4, and as a redox promoter for Co3O4, thereby enhancing the
intrinsic activity of the resulting catalyst. The specific objective of the
study will be to determine the amount of cerium (5–30 %wt.) to be
loaded on the alumina support for the optimal performance of a catalyst
with a 30 %wt.Co loading in the oxidation of methane in trace amounts
(< 1 %) under both dry and humid conditions.
performed under vacuum, using
WDXRF spectrometer, equipped with a Rh tube and three different
detectors (gas flow, scintillation and Xe sealed).
a PANalytical AXIOS sequential
Structural properties were determined by X-Ray diffraction, Raman
spectroscopy, and transmission electron microscopy. XRD analysis were
performed on a X’PERT-PRO X-Ray diffractometer using Cu Kα radia-
tion (λ = 1.5406 Å) and a Ni filter. The X-Ray tube was operated at
40 kV and 40 mA of current. The samples were scanned from an initial
value of 2θ = 5° to a final value of 2θ = 80°, with a step size of 0.026°
and a counting time of 2.0 s for each step. Phase identification was
performed by comparison of the obtained diffraction patterns with
JCPDS (Joint Committee on Powder Diffraction Standards) database
cards. A longer counting time (26.8 s) was applied to perform a detailed
XRD analysis over the xCe-Al supported cobalt catalysts. The cell size of
the cobalt spinel phase was obtained by profile matching of the detailed
XRD patterns using FullProf.2k software. Additionally, the analysis by
Raman spectroscopy was carried out by using a Renishaw InVia Raman
spectrometer, coupled to a Leica DMLM microscope. The excitation
wavelength was 514 nm (ion-argon laser, Modu-Laser). The spatial re-
solution was 2 microns. For each spectrum 20 s were employed and five
scans were accumulated with the 10 % of the maximum power of the
2. Experimental
2.1. Synthesis of the ceria-alumina supports and cobalt supported catalysts
The ceria-modified alumina supports were prepared by a basic
precipitation route of cerium (III) nitrate hexahydrate (Ce(NO3)3·6H2O,
Sigma Aldrich) on a commercial γ-Al2O3 (Saint Gobain) thermally
stabilized at 850 °C for 8 h. For each support, 5 g of γ-Al2O3 were mixed
with 100 ml of the cerium precursor with adjusted concentrations of Ce
and then a solution of Na2CO3 1.2 M was added dropwise until the pH
reached 9.5. The temperature was kept constant at 80 °C. The selected
cerium loadings were 5, 10, 15, 20 and 30 %wt. After an ageing step of
30 min at constant temperature (80 °C) and pH (9.5), the obtained
precipitates were filtered and thoroughly washed with at least 5 litres of
deionised water to remove all residual sodium ions that could remain in
the precursor, which have been proven to inhibit the activity of cobalt
oxide catalysts [25]. These modified supports were denoted as xCe-Al
where x stands for the nominal cerium loading. A support of pure ceria
(CeO2) was also synthesised by the same route starting from a cerium
(III) nitrate solution.
Cobalt catalysts supported on xCe-Al, CeO2 and Al2O3 were pre-
pared by following the same synthesis route as for the supports but
starting from a mixture of 5 g of each support (modified alumina, bare
ceria and bare alumina, respectively) and 100 ml of a cobalt (II) nitrate
hexahydrate (Co(NO3)2·6H2O, Fluka) solution. All catalysts were pre-
pared with a cobalt loading of 30 %wt. and denoted as Co/xCe-Al, Co/
CeO2 and Co/Al2O3, respectively. A reference sample of bulk Co3O4 was
also prepared as well by precipitation of a solution of cobalt (II) nitrate.
All supports and catalytic precursors were dried at 110 °C for 16 h
and then subjected to calcination in static air to produce the final
supports and catalysts. The calcination protocol was designed on the
basis of previous thermogravimetric analysis (Fig. S1, Supplementary
material) and consisted of three heating ramps separated by 30-min
514 nm laser in a spectral window of 150–1200 cm−1
.
Transmission electron microscopy (TEM) investigations were per-
formed using a Philips CM200 microscope equipped with LaB6 crystal
as electron source and operating at 200 kV. Bright field images were
acquired using a high resolution CCD camera. HRTEM measurements
were carried out with a FEI Titan Cubed G2 60-300 electron microscope
at 300 kV equipped with a high-brightness X-FEG Schottky field emis-
sion electron gun and a monochromator and CEOS GmbH spherical
aberration (Cs) corrector on the image side. The images were recorded
on a charge-coupled device (CCD) camera (2kx2k Gatan UltraScanTM
1000). X-Ray photoelectron spectroscopy (XPS) measurements were
carried out in a Kratos AXIS Supra spectrometer using a 225 W Al KAl
Kα radiation source with a pass energy of 160 eV for the general survey
and 20 eV for the specific spectra.
Redox properties and Co species distribution was investigated by
means of two temperature-programmed techniques on a Micromeritics
Autochem 2920 apparatus. Temperature-programmed reduction with
hydrogen (H2-TPR) was performed using a 5 %H2/Ar mixture as the
reducing gas. The analysis protocol involved an initial pre-treatment
step with a 5 %O2/He mixture at 300 °C for 30 min. After cooling down
to room temperature with flowing He, the TPR experiment was per-
formed with a heating rate of 10 °C min−1 up to 950 °C for all samples,
and that temperature was then maintained for 10 min. The water pro-
duced throughout the whole experiment was eliminated using a cold
trap, to avoid interference with the TCD detector. Additional informa-
tion regarding the activation of methane was obtained by means of
2

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
temperature programmed reaction with a 5 %CH4/He mixture in the
absence of oxygen (CH4-TPRe) coupled to mass spectrometry (MKS
Cirrus Quadrupole Mass Spectrometer). The experiments were carried
out up to 600 °C with a heating ramp of 10 °C min−1 followed by an
isothermal step at 600 °C for 30 min.
2.3. Catalytic activity determination
The activity of the prepared catalysts was tested in a bench-scale
fixed bed tubular reactor (PID Eng&Tech S.L.) in the 200−600 °C
temperature range. For each reaction experiment, 1 g of catalysts with a
particle size of 0.25-0.3 mm was diluted with 1 g of inert quartz with a
particle size of 0.5–0.8 mm, and the mixture was placed in a Hastelloy X
tube with a K type thermocouple located inside the catalytic bed. The
feedstream composition used was 1 %CH4, 10 %O2 and N2 as balance
gas, and it was fed to the reactor with a total flow of 500 ml min−1
,
which accounted for a space velocity of 300 ml CH4 h−1 g−1. The cor-
responding gas hourly space velocity calculated as a function of the
total flow rate was approximately 60,000 h−1. Additionally, stability
tests were conducted over a prolonged period of time (150 h) at 450
and 525 °C under dry and cycling dry/humid (10 %vol. H2O) condi-
tions. Methane conversion values were obtained from the methane
concentrations measured by a microGC equipped with a TCD. For each
temperature value or reaction time interval, the chromatographic
analysis were performed in triplicate to check the reproducibility. The
standard deviation found was less than 1 % in all cases. To ensure that
the obtained kinetic results were not affected by mass or heat transfer
limitations, the criteria for intra and extra-particle heat and mass
transfer, and temperature gradients were checked, according to the
Eurokin procedure (Table S1, Supplementary material).
Fig. 1. XRD profiles of the xCe-Al supports.
distribution (Fig. S2, Supplementary material) [26]. No significant
change in the hysteresis loops was apparently detected with the Ce
loading. Pore size distribution was estimated according to the BJH
method. The samples with low Ce loadings (5–15 %wt.) displayed a
bimodal distribution peaking at around 110 and 150 Å, similar to that
exhibited by the bare alumina, thus showing evidence of a good dis-
persion of the deposited Ce over the surface of the support. For the
20Ce-Al sample, however, the pore size switched to a unimodal dis-
tribution centred at 110 Å, probably due to the narrowing of the largest
pores (150 Å) caused by an increased cerium deposition. Finally, the
30Ce-Al sample exhibited the same maximum at 110 Å but with a lower
pore volume (0.25 cm3 g−1
) compared with the 20Ce-Al sample
(0.30 cm3 g−1). Besides, this sample presented a fraction of small pores
with a maximum around 35 Å derived from a marked pore narrowing of
the existing 110 Å pores because of the massive cerium incorporation.
XRD patterns of the xCe-Al supports calcined at 600 °C are shown in
Fig. 1. This graph also includes the diffractograms of the bare alumina
and bulk ceria. In all cases the diffraction signals could be indexed as γ-
alumina (2θ at 37.7, 45.8 and 67.3°, JCPDS 01-074-2206) and cerium
oxide (2θ at 28.5, 33.3, 47.5, 56.4 and 76.7°, JCPDS 00-004-0593).
Note that no peaks attributable to other possible crystalline phases such
as cerium aluminate were noticed. As an additional evidence of the
deposition of ceria on the surface of the alumina support, it was
checked that the intensity ratio of the characteristic signals of CeO2 (2θ
at 47.5°) and Al2O3 (2θ at 45.8°) increased with cerium loading.
Somewhat wide CeO2 peaks were visible, thereby suggesting that ceria
was present in the form of finely divided particles. An attempt was thus
made to estimate the mean crystallite size of ceria from the peak
broadening of the most intense reflection ((111) plane) using the
Scherrer equation. This size was close to 10 nm for the 5Ce-Al, 10Ce-Al
and 15Ce-Al samples, and 12 nm for the supports with a higher Ce
content (20Ce-Al and 30Ce-Al samples). These values were in contrast
with that found for the as-prepared bulk ceria (32 nm), thus evidencing
a good dispersion of ceria on the alumina support.
3. Results and discussion
3.1. Characterisation of xCe-Al supports
The composition and textural and structural properties of the ceria-
modified alumina supports were characterised by WDXRF, N2-physi-
sorption, XRD and Raman spectroscopy. Firstly, it should be pointed out
that the chemical analysis revealed that the amount of cerium species
deposited on the alumina support was very close to the nominal
loading, namely 5, 10, 15, 20 and 30 %wt.Ce (Table 1).
The specific surface area, total pore volume and pore size dis-
tribution maxima are summarised in Table 1. The textural properties of
the pure oxides (γ-Al2O3 and CeO2) are included for comparative pur-
poses. A roughly linear decrease in surface area was noticed with in-
creasing Ce content. Thus, this loss was around 25 % (with respect to
the bare alumina) for the sample with 30 %wt.Ce (139 vs 105 m2 g−1).
Judging from these results the impact of ceria coating was considered
moderate. Accordingly, the pore volume of the samples steadily de-
creased from 0.56 to 0.32 cm3 g−1
The samples showed type IV isotherms with H2 hysteresis loops,
commonly associated with pore blocking over
.
a wide pore size
The presence of segregated ceria was further corroborated by
Raman spectroscopy. Thus, all coated alumina supports showed a
strong peak assigned to the F2g Raman-active mode characteristic of the
fluorite-like lattice of CeO2 (Fig. S3, Supplementary material). When
compared with the Raman spectra of bulk ceria (464 cm−1), the signals
were slightly broader and shifted to lower values (462 cm−1) due to the
alumina-ceria interaction [27]. Moreover, two additional weak signals
at 260 and 595 cm−1 were observed. These bands corresponded to
nondegenerate longitudinal optical modes of CeO2, which were linked
to with oxygen vacancies from partially reduced CeO2-x species [28]. In
view of the higher intensity of these features for the 5Ce-Al and 10Ce-Al
supports, the formation of Ce3+ species, probably in the form of stable
CeAlO3-like species, was favoured with low Ce loadings. As
Table 1
Physico-chemical properties of the xCe-Al supports.
Sample
Ce, %wt. SBET
m2
,
Vpore
,
Pore Size
D
CeO2, nm H2 uptake,
cm3 g−1 Distribution
maxima, Å
mmol gCe
−1
g−1
Al2O3
5Ce-Al
10Ce-Al 9.8
15Ce-Al 15.3
20Ce-Al 20.6
30Ce-Al 30.3
–
139
128
126
122
117
105
8
0.56
0.51
0.51
0.49
0.42
0.32
0.03
110, 150
110, 150
110, 150
110, 150
110
–
10
9
10
12
12
32
–
5.2
1.9
2.1
2.6
3.3
3.1
1.8
35, 110
230
CeO2
81.4
3

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
favoured presence of Ce3+ species on these samples, in the form of
CeO2-x species with a strong interaction with the alumina, as evidenced
by Raman spectroscopy. For higher Ce concentrations, the predominant
cerium species was segregated CeO2.
3.2. Characterisation of Co/xCe-Al catalysts
The addition of cobalt oxide (in the 27.2–29.5 %wt. range as de-
termined by WDXRF) to the ceria-modified alumina supports (the ac-
tual Ce loading of the catalysts were in the 3.1–18.2 %wt. range as
determined by WDXRF as well) resulted in a notable, almost linear loss
of surface area of the resulting catalysts (73−102 m2 g−1) (Table 2).
With respect to the corresponding support (105−139 m2 g−1), this
accounted for a decline by 20–30 %. This negative effect on surface area
was comparable to that noticed for the Co/Al2O3 catalyst (22 %). As for
the pore volume, the Co/30Ce-Al sample presented the lowest value
(0.25 cm3 g−1) while no significant differences were found among the
other four samples (0.30–0.32 cm3 g−1).
Fig. 2. H2-TPR profiles of the xCe-Al supports.
The deposition of cobalt species preferentially occurred on the
wider pores (150 Å) of the modified alumina supports with a bimodal
pore size distribution, namely 5Ce-Al, 10Ce-Al and 15Ce-Al (Fig. S6,
Supplementary material). In fact, the profiles of these catalysts showed
a unimodal distribution peaking at 110 Å. In the case of the Co/20Ce-Al
sample the addition of cobalt to this modified support, characterised by
a unimodal distribution with a maximum at 110 Å, resulted in a shift of
the pore size distribution to slightly lower values (around 90 Å). This
shift was not observed over the Co/30Ce-Al catalyst with respect to its
counterpart alumina support (30Ce-Al), thus suggesting that cobalt
oxide was partially present on the external surface of this sample.
Fig. 3 shows the diffractograms of the Co/xCe-Al catalysts. For
comparative purposes, the patterns of the Co/Al2O3 and Co/CeO2
samples were included as well. Along with the characteristic features of
the ceria phase (2θ = 28.5, 33.3, 47.5, 56.4 and 76.7°), a set of signals
at 2θ = 31.3, 37.0, 45.1, 59.4 and 65.3° were clearly distinguished.
These were related to a spinel-like cobalt phase, namely Co3O4 (JCPDS
00-042-1467) and/or CoAl2O4 (JCPDS 00-044-0160). The formation of
cobalt aluminate was plausible due to the interaction of Co3O4 and
Al2O3 at high temperatures [31]. Unfortunately, it was not possible to
differentiate between these two oxides since both phases crystallise in
the cubic structure. Obviously, the formation of CoAl2O4 was only ruled
out in the case of the 30Co/CeO2 sample. After the addition of cobalt
the crystallite size of CeO2, which was in 9−14 nm range, was not
appreciably modified when compared with the corresponding xCe/Al
support (10−12 nm). As for the size of the spinel-like cobalt phase
estimated from the peak broadening of the signal at 37.1°, it remained
almost constant (23−24 nm) irrespective of the composition of the
support. Only for the Co/30Ce-Al catalyst the size was considerably
larger (31 nm), probably due to the high cerium loading.
aforementioned, these species were not detected by XRD analysis.
It seems clear that this varying abundance of Ce3+ and Ce4+ as a
function of the cerium content should be consistent with the specific
hydrogen uptake of the supports estimated by H2-TPR. In this way,
Fig. 2 shows the reduction profiles of bulk ceria and the various Ce-
coated alumina supports. The CeO2 sample exhibited a weak signal at
450−500 °C that corresponded to the surface reduction of the oxide
whereas the intense H2 uptake peaking at about 800 °C was related to
the reduction of the bulk [29]. For the xCe-Al supports, these two re-
duction events were also observed. However, while the surface reduc-
tion occurred at 450−500 °C as well, the bulk reduction only required
650−700 °C, probably due to the relatively small crystallite size of
deposited ceria (10−12 nm). In addition, a small shoulder above 850 °C
related to the formation of CeAlO3 as a result of the interaction of CeO2
and Al2O3 (2CeO2 + Al2O3 + H2 → 2CeAlO3 + H2O) was noted [30].
The generation of this perovskite phase during the reduction process
was further confirmed by subsequent XRD analysis of the samples after
the H2-TPR run. Hence, weak diffraction signals assignable to CeAlO3
were visible on all xCe-Al supports (2θ at 23.6, 33.5, 41.4 and 60.1°,
JCPDS 00-048-0051) (Fig. S4, Supplementary material). In view of the
similar shape of all H2-TPR traces, it could be assumed the redox
characteristics of deposited cerium species, with no marked shift in the
reduction temperatures, remained invariant although the degree of
reduction substantially varied as a function of the cerium loading.
The overall H2 consumption of the samples expectedly increased
with the cerium content, from 0.09 mmol H2 g−1 over the 5Ce-Al
sample to 0.94 mmol H2 g−1 over the 30Ce-Al sample (Table 1).
However, more notable differences in reducibility were noticed when
the specific uptake was examined. Hence, this intrinsic consumption
largely depended on the Ce content of the support (Fig. S5, Supple-
mentary material). Note that this property for the as-synthesised pure
bulk CeO2 was 1.8 mmol H2 gCe−1. The consumption progressively in-
The Raman spectra of the cobalt catalysts supported on pure alu-
mina (Co/Al2O3), pure ceria (Co/CeO2) and ceria/alumina (Co/xCe-Al)
are included in Fig. 4. As a reference, the spectrum of pure Co3O4 is
shown as well. Apart from a relatively weak band at 462 cm−1 (F2g
mode of CeO2), all supported catalysts displayed the five Raman actives
modes associated with Co3O4, namely three F2g modes located at 194,
−1
−1
creased from 1.9 mmol H2 gCe
(5Ce-Al) to 3.2–3.5 mmol H2 gCe
over the 20Ce-Al and 30Ce-Al samples. The low H2 consumption ob-
served for the 5Ce-Al and 10Ce-Al supports was coherent with a
Table 2
Physico-chemical properties of the Co/xCe-Al catalysts.
Sample
Ce, %wt.
Co, %wt.
SBET, m2 g−1
Vpore, cm3 g−1
Pore Size Distribution maxima, Å
DCeO2, nm
DCo-spinel, nm
Co/Al2O3
–
27.9
28.8
28.9
28.0
29.5
27.2
28.9
108
102
96
90
93
0.29
0.32
0.30
0.32
0.30
0.25
0.07
90
–
8
9
10
11
14
33
29
24
24
23
23
31
44
Co/5Ce-Al
Co/10Ce-Al
Co/15Ce-Al
Co/20Ce-Al
Co/30Ce-Al
Co/CeO2
3.1
5.9
9.2
12.4
18.2
48.8
110
110
110
90
35, 110
225
73
18
4

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
the Co3O4 signals could be attributed to the distortion of the spinel
lattice, possibly due to insertion of Ce ions [36]. Consistently, the
FWHM values of this band were higher for the Ce-rich cobalt catalysts
(25−26 cm−1 compared with 14−19 cm−1 for Co/5Ce-Al, Co/10Ce-Al
and Co/15Ce-Al).
The structural change of the Co3O4 lattice was further confirmed by
the estimation of the cell parameter of the Co spinel from XRD patterns.
As aforementioned, the detected diffraction signals could be assigned to
both Co3O4 and CoAl2O4. Consequently, the position shifts that denote
the change in the cell size could be initially attributed to distortion of
both phases. However, since only the Raman vibration modes of Co3O4
phase (and not those of CoAl2O4 phase) evidenced changes with the
addition of Ce, the shift in the diffraction signals was assumed to occur
only due to distortion of the cobalt oxide phase. Thus, the cell size of
Co3O4 in the Co/xCe-Al catalysts was found to be larger than that of the
catalyst supported over bare alumina (8.077 Å), as shown in Fig. S8
(Supplementary material). More importantly, for the Co/20Ce-Al and
Fig. 3. XRD profiles of the Co/xCe-Al catalysts.
Co/30Ce-Al samples, the cell size displayed
a maximum value
(8.092 Å), which evidenced the largest lattice distortion, in accord with
the results obtained from Raman spectroscopy. The enlargement of the
unit cell of Co3O4 has been previously reported by other authors
[37,38] and is associated with the larger ionic radius of Ce4+ ions
(101 pm), with respect to Co2+ and Co3+ ions (79 and 69 pm, respec-
tively). Note that this eventual insertion of Ce4+ ions into the spinel
lattice would be accompanied by an increase in the amount of Co3+ so
as to maintain the charge balance [39].
HAADF-STEM images along with the corresponding EDX maps of
the Co/xCe-Al catalysts are shown in Fig. 5. This analysis evidenced
that cobalt species were indeed located on the surface of the ceria-
modified alumina with a low presence of segregated entities. In addi-
tion, it was noteworthy that some regions of the support were not
massively covered by cobalt. The spinel active phase was present as
crystallites with average sizes around 20−25 nm, which was in good
agreement with XRD results (Table 2), that tended to aggregate into
larger patches of 100−150 nm. Only for the Co/30Ce-Al sample, seg-
regated Co3O4 crystallites with sizes up to 50−60 nm were marginally
observed. As for the cerium, it was homogeneously distributed on the
support in the form of quite small crystallites (smaller than 5 nm).
Additionally, CeO2 clusters of about 10 nm were also detected, in line
with XRD results (Table 2). For the Co/30Ce-Al sample, the size of the
clusters increased up to 20 nm, as also evidenced by XRD analysis.
Finally, Fig. S9 (Supplementary material) shows some selected
HRTEM images, where the lattice fringes of Co3O4 and CeO2 could be
resolved, along with their corresponding Fast Fourier Transform (FFT)
spot patterns. For the cobalt spinel, two lattice spaces, namely 0.29 and
0.24 nm, were identified, which corresponded to the {220} and the
{311} planes, respectively. On the other hand, for the ceria two lattice
spaces of 0.31 and 0.27 nm were observed, which corresponded to the
{111} and {200} planes, respectively.
The surface structure and composition was investigated by XPS. The
distribution of cobalt and oxygen species was determined from the
Co2p and O1s spectra of the samples, respectively (Fig. S10,
Supplementary material). In particular, the Co 2p3/2 signal could be
deconvoluted in five different contributions. The two signals located at
lower binding energies, namely 779.5 and 780.7 eV, were attributed to
the presence of Co3+ and Co2+ cations in a spinel phase, respectively,
while the signal centred at about 782.3 eV was attributed to the pre-
sence of CoO species [40]. The presence of this oxide was comparable
on all samples, and it accounted for around 5–8 % of the total detected
signal. Finally, the two signals located at higher binding energies (785.5
and 789.5 eV) were identified as the satellite signals from Co2+ and
Co3+ ions, respectively [41]. On the other hand, the O 1s spectra
showed three different signals located at 529.8, 530.8 and 532.4 eV,
respectively. The first signal was assigned to oxygen species from the
spinel lattice (Olatt), the second one was related to weakly adsorbed
oxygen species on the surface (Oads) and the last feature was attributed
Fig. 4. Raman spectra of the Co/xCe-Al catalysts.
519 and 617 cm−1, and the Eg and A1g modes at 479 and 687 cm−1
,
respectively [32,33]. Except from the Co/CeO2 catalyst, two shoulders
at 705 and 725 cm−1 attached to the A1g vibration mode were visible.
These two signals suggested the presence of cobalt aluminate [34,35]. It
was found out that the intensity of these features was more notable over
the Co/Al2O3 and Co/5Ce-Al samples, and tended to decrease when
ceria was deposited with a higher loading (> 10 %wt.Ce, more sig-
nificantly for 30 %wt.Ce). Although the formation of this undesired
cobalt phase was not completely inhibited, it seemed that incorporated
ceria acted as an efficient physical barrier to prevent the reaction be-
tween Co3O4 and Al2O3 to some extent. From a catalytic point of view,
the ceria coating of the alumina support would favour an increase in the
amount of highly active Co3O4 at the cost of CoAl2O4.
A closer inspection of the dependence of the A1g mode with the Ce
content of the catalysts could be helpful in determining a possible
distortion of the Co3O4 lattice due to the partial insertion of cerium
cations. This effect was analysed in terms of the shift and the full width
at half maximum (FWHM) of this signal (Fig. S7, Supplementary ma-
terial). On one hand, it was observed that the location of the band
varied from 689 cm−1 in the Co/Al2O3 and bulk Co3O4 samples to
681 cm−1 in the Co/CeO2 sample, thus pointing out that the lattice of
Co3O4 was much more affected by the presence of CeO2. As the Ce/Co
molar ratio of the Co/xCe-Al catalysts increased the band progressively
shifted to lower values, from 689 cm−1 for Co/Al2O3 to 688−685 cm−1
for Co/5Ce-Al, Co/10Ce-Al and Co/15Ce-Al, and to 682−680 cm−1 for
Co/20Ce-Al, Co/30Ce-Al and Co/CeO2. It should be noticed that the
other Raman active modes (F2g and Eg) shifted as well. This redshift of
5

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
Fig. 5. HAADF-STEM images of the Co/xCe-Al catalysts coupled with EDX elemental distribution of Co (red), Ce (green) and Al (blue). (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)
6

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
Table 3
Surface composition of the Co/xCe-Al catalysts as determined by XPS.
Sample
Co, %wt.
Ce, %wt.
Al, %wt.
Co/(Ce + Al) molar ratio
Co3+/Co2+ molar ratio
Oads/Olatt molar ratio
Co/Al2O3
22.6 (27.9)
26.7 (28.8)
30.4 (28.9)
30.1 (28.0)
32.9 (29.5)
29.7 (27.2)
35.6 (28.9)
0 (0)
32.1 (37.0)
23.6 (32.1)
20.1 (30.0)
19.4 (29.1)
17.2 (26.1)
18.4 (20.1)
0 (0)
0.32
0.51
0.68
0.69
0.84
0.70
3.26
0.67
0.91
0.99
1.15
1.38
1.17
0.95
1.41
1.30
0.98
0.82
0.77
0.88
1.07
Co/5Ce-Al
Co/10Ce-Al
Co/15Ce-Al
Co/20Ce-Al
Co/30Ce-Al
Co/CeO2
1.3 (3.1)
2.4 (5.9)
3.0 (9.2)
3.8 (12.4)
5.3 (18.2)
26.8 (48.8)
The values in brackets correspond to the bulk composition as determined by WDXRF.
to the presence of hydroxyl groups from adsorbed water [42].
Table 3 includes the surface molar composition of the investigated
catalysts. The bulk composition as determined by WDXRF is also in-
cluded for comparative purposes. Initially the properties of the re-
ference supported catalysts, namely Co/Al2O3 and Co/CeO2 samples,
will be discussed. The Co/Al2O3 sample showed a surface Co3+/Co2+
molar ratio of 0.67, which was noticeably lower compared with that of
the Co/CeO2 catalyst (0.95). Taking the latter value as representative of
the supported Co3O4 oxide phase, the low value of the alumina-sup-
ported catalyst was in agreement with the significant presence of
CoAl2O4, where cobalt cations are in the +2 oxidation state, as sug-
gested by Raman spectroscopy. As the formation of cobalt aluminate
involves a gradual diffusion of Co cations into the alumina structure,
this necessarily resulted in a lower presence of cobalt on the surface,
with a Co/Al molar ratio of 0.32. Furthermore, the relative abundance
of surface cobalt increased for the Ce-containing samples as revealed by
their higher Co/(Ce + Al) molar ratios (up to 0.84 for the Co/20Ce-Al
sample). This finding was coherent with a less favoured formation of
cobalt aluminate due to the barrier effect of ceria, in line with the re-
sults derived from Raman spectroscopy.
Fig. 6. Evolution of the surface molar composition (XPS) and O2 consumption
(CH4-TPRe) with the Ce loading of the Co/xCe-Al.
On the other hand, the Co3+/Co2+ molar ratio of the Co/xCe-Al
catalysts, except for the Co/5Ce-Al sample), was higher (0.99–1.38)
than that of the Co/CeO2 catalyst. This fact revealed that some cerium
cations were likely incorporated into the Co3O4 lattice, as previously
suggested by XRD. This partial insertion necessarily implied an increase
in the abundance of Co3+ cations at the cost of Co2+ cations to
maintain the charge balance. This higher abundance of Co3+ species on
the surface of the samples with increasing Ce loading was accompanied
by a concomitant more notable presence of lattice oxygen species in the
Co/xCe-Al catalysts (Table 3). These type of oxygen species are usually
involved in the oxidation of methane by a Mars–van Krevelen me-
chanism. As shown in Fig. 6, the relative presence of both species in-
creased with cerium loading and was optimised for the Co/20Ce-Al
catalyst.
Table 4
Results from the H2-TPR and CH4-TPRe analysis of the Co/xCe-Al catalysts.
Sample
H2-TPR
CH4-TPRe
H2 uptake at low
temperatures,
mmol g−1
H2 uptake at high
temperatures,
mmol g−1
O2 consumption at
low temperatures,
−1
mmol gCo
Co/Al2O3
2.74
2.01
2.83
3.02
3.29
3.68
6.77
2.84
3.86
3.27
3.22
3.19
2.61
0.82
0.28
0.45
0.68
0.72
0.88
0.77
0.70
Co/5Ce-Al
Co/10Ce-Al
Co/15Ce-Al
Co/20Ce-Al
Co/30Ce-Al
Co/CeO2
The nature of deposited cobalt species and their interaction with
underlying ceria-coated alumina was examined by H2-TPR analysis
(Table 4). Initially the behaviour of cobalt catalysts supported on pure
alumina or ceria (Co/Al2O3 and Co/CeO2, respectively) was compared.
Besides, the reduction profile of bulk ceria was taken into consideration
(Fig. S11, Supplementary material). Both cobalt catalysts displayed a
notable H2 uptake between 200−600 °C that was related to the re-
duction of precipitated Co3O4. This contribution showed two more or
less discernible peaks at 310 and 380 °C, which were indicative of the
sequential reduction to CoO and metallic Co, respectively [32,43].
Except for a small band observed at 800 °C (this peak corresponded to
the reduction of ceria support), the Co/CeO2 catalyst did not consume
hydrogen between 600−800 °C. The overall H2 uptake was 7.6 mmol
H2 g−1. Interestingly, it was found that when subtracting the amount of
H2 theoretically required for full reduction of deposited Co3O4
(equivalent to 22.6 mmol H2 gCo−1, according to following the equation
Co3O4 + 4H2 → 3Co + 4H2O) from the overall H2 consumption of the
sample, the H2 consumption related to the reduction of the support
could be estimated. This value was 1.8 mmol H2 gCe-1, that coincided
with the specific consumption measured for the pure CeO2 support.
These results led us to conclude that no mutual effects on the re-
ducibility of both cobalt oxide and ceria existed for this catalyst.
In the case of Co/Al2O3, an additional peak between 550−750 °C
was clearly ascertained. This was assigned to the presence of significant
quantities of CoAl2O4 derived from the strong interaction between
Co3O4 and Al2O3, as also evidenced by Raman spectroscopy. In fact, the
total measured amount of H2 (5.6 mmol H2 g−1) was rather far from the
stoichiometric uptake (6.8 mmol H2 g−1) when assuming that the co-
balt phase was exclusively Co3O4. These findings led to the conclusion
that the reducibility of free Co3O4 (cobalt species whose reduction oc-
curred at 200−600 °C) was much more favoured when cobalt was de-
posited over ceria instead of alumina, although the interaction between
cobalt and ceria did not seem to enhance the redox properties of the
Co3O4 deposited over pure ceria in comparison with pure Co3O4.
Simultaneously, the reducibility of ceria was also not improved.
7

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
Fig. 7. H2-TPR profiles of the Co/xCe-Al catalysts.
The H2-TPR profiles of the cobalt catalysts supported on ceria-alu-
mina also consisted of two H2 uptakes at 200−550 °C and 550−800 °C
(Fig. 7). The total H2 consumption of the samples is given in Table 4. In
all cases the H2 uptake was larger than that shown by the Co/Al2O3
sample. A significant increase by 5–16 % was found with the amount of
cerium in the support. Above 15 %wt.Ce the consumption was rather
similar (6.5–6.7 mmol H2 g−1). While no significant changes in the
reduction onset temperature (about 250 °C) were observed, the relative
contribution of these reduction events seemed to be shifted in favour of
the low-temperature band with the Ce loading (Table 4). The observed
increase in the H2 uptake at low temperature could be rationalised in
terms of a favoured reduction of Co3O4 with a distorted lattice due to
cerium doping as revealed by XRD and XPS analysis. Another plausible
phenomenon that would contribute to a higher reducibility between
200 and 500 °C could be a larger extent of ceria reduction owing to the
transfer of hydrogen by metallic cobalt onto the ceria surface [44,45].
On the other hand, and in agreement with Raman spectroscopy, the
interaction between Co3O4 and Al2O3 to give CoAl2O4 was somewhat
inhibited by deposited ceria, thereby resulting in a decreased H2 uptake
at high temperatures.
Fig. 8. CH4-TPRe profiles of the Co/xCe-Al catalysts (a). Close-up view of the
300−550 °C temperature range (b).
occurred at 465 °C over the Co/20Ce-Al sample and consumed the
largest amount of oxygen (0.88 mmol O2gCo−1). Moreover, the tem-
perature of the onset of reduction (marked by arrows in Fig. 8) was
significantly lower for the Co/20Ce-Al and Co/30Ce-Al samples
(around 350 °C) in comparison with the rest of the samples (around
400 °C). As stated above, the generation of CO2 was also visible at
higher temperatures owing to the oxidation of methane by oxygen
species associated with Co2+ ions [46]. This mass signal was detected
at 600 °C during the isothermal period. Exceptionally, CO2 was detected
at 580 °C over the Co/20Ce-Al catalyst, thereby suggesting a higher
mobility of the oxygen species in this sample. In all cases, the high-
temperature oxidation was accompanied by the generation of CO and
H2 to some extent that could be due to partial oxidation or cracking of
methane in the presence of metallic or oxygen-deficient cobalt species
[47]. In fact, the diffraction pattern of the samples after the CH4-TPRe
run evidenced the formation of graphitic carbon (signal at 2θ = 26.6°)
due to the occurrence of these two reactions [48].
In sum, the presence of ceria deposited on the alumina surface re-
sulted in remarkable changes in the redox properties by distorting the
Co3O4 crystalline lattice and/or promoting the ceria reducibility and/or
inhibiting the formation of CoAl2O4. These positive effects eventually
led to a larger abundance of highly active oxygen, in the form of lattice
oxygen species as evidenced by XPS analysis, over the Co/xCe-Al cat-
alysts in comparison with the Co/Al2O3 counterpart.
More useful insights on the influence of the catalyst composition on
the reactivity of available oxygen species for methane oxidation were
obtained by CH4-TPRe analysis coupled to mass spectrometry. The
analysis was performed between 50 and 600 °C with a subsequent iso-
thermal step at this temperature for 30 min. The evolution of CO2 (m/
z = 44) and CO (m/z = 28) was monitored (Fig. 8). CO2 formation
occurred at two clearly distinct temperature windows, namely
425−500 °C (only CO2 was detected) and above 580 °C (both CO and
CO2 were detected accompanied by H2 as well, not shown in Fig. 8). In
the low temperature range both Co/Al2O3 and Co/CeO2 catalysts led to
the generation of CO2 at about 485 °C, which was attributed to the
oxidation of methane by oxygen species associated with Co3+ ions.
Table 4 includes the O2 consumption corresponding to this generation
of CO2 for all examined catalysts. Note that the lowest value
(0.28 mmol O2gCo−1) corresponded to the Co/Al2O3 sample in agree-
ment with the marked presence of CoAl2O4. When cerium was added to
the alumina support, the CO2 formation was more evident ant took
place at significantly lower temperatures. Fig. 6 revealed that the extent
of CH4 oxidation was noticeably favoured with the presence of lattice
oxygen species. On the other hand, the reduction process interestingly
3.3. Catalytic behaviour of Co/xCe-Al catalysts
The performance of the synthesised cobalt catalysts was examined
by their corresponding light-off curves at 300 ml CH4 g−1 h−1
(30,000 ml g−1 h−1) in the 200−600 °C temperature range (Fig. 9).
Each catalyst was subjected to at least three consecutive runs. It was
found that the corresponding light-off curves were all comparable after
the second cycle. The representative light-off curve of each catalyst
corresponded to the third consecutive run. For the sake of comparison
8

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
then decrease significantly for the samples with a higher Ce con-
centration (2.9 mmolCH4 gCo−1 h−1). This trend was also consistent
with the intrinsic activity shown by the cobalt catalyst supported on
pure ceria (2.1 mmolCH4 gCo−1 h−1). The results evidenced a promoted
activity of Co3O4 when supported on ceria-coated alumina. The positive
effect of ceria coating was mainly ascribed to the enhanced reactivity of
the oxygen species of Co3O4 as also shown in Fig. 10. Hence, a marked
dependence of the intrinsic activity with the amount of O2 consumed at
low temperatures (< 500 °C) in the CH4-TPRe runs was evidenced. As
aforementioned, ceria addition was useful for partially reducing the
amount of inactive CoAl2O4 and simultaneously inducing a distortion of
the structure of Co3O4. Both phenomena ultimately led to catalysts with
an enhanced oxidation ability. The apparent activation energy of the
reaction over the investigated cobalt catalysts was evaluated by ap-
plying the integral method. A first pseudo-order for methane and a zero
pseudo-order for oxygen were assumed on the basis of a simplified
Mars–van Krevelen kinetics for this reaction studied with a high O2/
CH4 molar ratio [49,50]. The corresponding plots for the linearized
kinetic equation of the integral reactor are shown in Fig. S12, Supple-
mentary material. The values for the apparent activation energy were in
the 80−83 kJ mol−1 range for all catalysts, which were coherent with
the value obtained with a bulk Co3O4 catalyst (78 kJ mol−1) [24].
Finally, the stability of the most active sample (Co/20Ce-Al) was
examined under two different scenarios. On one hand, the evolution of
conversion with time on stream of a fresh sample was followed at
450 °C for 60 h. Then, the reaction temperature was increased up to
525 °C and the sample was maintained at this temperature during ad-
ditional 90 h. Results from this experiment carried out under dry con-
ditions are shown in Fig. 11. A relatively constant conversion level
around 30 % was observed at 450 °C with no clear evidences of deac-
tivation. After raising the temperature at 525 °C, a progressive loss of
conversion from 78 % to 64 % (25 h), and to 58 % (25 h) was noticed.
Then, conversion remained stable for the remaining 40 h of the stability
test.
Fig. 9. Light-off curves of the Co/xCe-Al catalysts.
the conversion-temperature profiles of the Co/Al2O3, Co/CeO2 and bulk
CeO2 catalysts are also included. It must be noticed that all catalysts
exhibited a 100 % selectivity towards CO2, irrespective of their cerium
loading. While a negligible activity in the whole temperature range was
noticed for bare ceria, the cobalt catalyst supported over uncoated
alumina (Co/Al2O3) gave a T50 value at around 550 °C. However, this
sample did not achieve a 100 % conversion at the highest investigated
temperature (only around 80 % at 600 °C). Interestingly, all Co/xCe-Al
catalysts evidenced a better behaviour irrespectively the ceria loading,
with T50 values in the 480−540 °C temperature range and a conversion
higher than 95 % at 600 °C. On the other hand, the observed higher
efficiency of the Co/CeO2 sample (T50 of 500 °C) with respect to its
counterpart supported on pure alumina indicated that a cobalt-ceria
interaction was catalytically more preferable. The following overall
trend was found: Co/20Ce-Al (480 °C) > Co/15Ce-Al (490 °C) > Co/
30Ce-Al (495 °C) > Co/CeO2 ≈ Co/10Ce-Al (505 °C) > Co/5Ce-Al
(525 °C) > Ce-Al2O3 (550 °C). Thus, an optimum cerium loading (20 %
wt.) deposited on alumina was defined.
In a second scenario a fresh sample was also evaluated during a
prolonged time span (150 h), in the absence or presence of water (10 %)
in the feed stream. Fig. 11 includes the evolution of methane conversion
during the stability test, where dry and humid conditions were cycled
each 25 h at constant temperature (525 °C). The results revealed a sig-
nificant deactivation for the first 25 h under dry conditions, with a
conversion decrease from 80 to 64 %. Note that the conversion profile
for this time interval was virtually identical to that observed for the
same conditions used in the first stability test. When water was added to
the gas stream, the impact on conversion was considerable as conver-
sion fell down to about 30 %. When dry conditions were re-established,
A clearer evidence of the superior performance of the Co/20Ce-Al
catalyst was given by the analysis of the specific reaction rate shown by
each Co/xCe-Al sample (Fig. 10). This reaction rate was calculated
under differential conditions (conversion < 15 %) at 400 °C. A volcano-
type relationship was distinctly seen, given that the reaction rate pro-
gressively increased with Ce loading between 0 and 20 %wt. (as re-
ferred to the alumina support) from 1.2–3.3 mmolCH4 gCo−1 h−1, to
Fig. 11. Evolution of conversion with time on stream of the Co/20Ce-Al cata-
lyst under dry conditions (450 °C/60 h and 525 °C/90 h) and cycling dry/humid
conditions (525 °C/150 h).
Fig. 10. Specific reaction rate at 400 °C and O2 consumption at low tempera-
tures (CH4-TPRe) of the Co/xCe-Al catalysts.
9

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
methane conversion was partially recovered to give a stable value close
to 52 %. However, this was a little bit far from 58 % dictated by the run
carried out under exclusively dry conditions. Once again, the admission
of water led to an inhibition of activity as conversion was 25 %. When
water admission was stopped, conversion again increased (up to 48 %),
and finally decreased down to around 20 % in the presence of water.
These results evidenced that the suggested inhibiting effect of water
was not only due to competition effects with methane for surface ad-
sorption but it was also affecting the physico-chemical properties of the
catalyst [51,52].
CRediT authorship contribution statement
Andoni Choya: Investigation, Writing - original draft. Beatriz de
Rivas: Methodology, Validation. Juan Ramón González-Velasco:
Project administration, Funding acquisition. Jose Ignacio Gutiérrez-
Ortiz: Formal analysis. Rubén López-Fonseca: Conceptualization,
Writing - review & editing, Supervision.
Acknowledgements
The used samples from the two long-term stability tests conducted
under dry and dry/humid conditions, namely Co/20Ce-Al(d) and Co/
20Ce-Al(d/h) samples, were characterised by N2 physisorption, XRD
and CH4-TPRe. Attention was paid on the eventual changes on the
surface area, the crystallite size of oxide phases and the amount of
active oxygen species al low temperatures (< 550 °C). It was found that
the composition of the feed stream clearly affected the physico-che-
mical properties of the sample (Table S2, Supplementary material).
Thus, a decrease in the surface area was observed (from 93 to 85
The author wish to thank the financial support provided by the
Ministry of Economy and Competitiveness (CTQ2016-80253-R AEI/
FEDER, UE), Basque Government (IT1297-19) and the University of the
Basque Country UPV/EHU (PIF15/335), and the technical and human
support provided by SGIker (UPV/EHU).
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.apcata.2019.117381.
m2 g−1 over Co/20Ce-Al(d)), which was slightly larger (79 m2 g−1
)
when the sample operated in the presence of large amounts of steam.
However, both deposited ceria and cobalt spinel phase exhibited a high
thermal stability in view of their relatively invariable crystallite size (12
and 23 nm, respectively). The negative impact of water was reflected in
the amount and reactivity of active oxygen species as estimated by CH4-
TPRe analysis. The corresponding profiles are included in Fig. S13,
Supplementary material and revealed that the O2 consumption at low
temperatures decreased by 4 % for the Co/20Ce-Al(d) sample and 16 %
for the Co/20Ce-Al(d/h) sample. Moreover, a concomitant shift of the
peak temperatures for this uptake was noticed from 465 to 470 °C over
the Co/20Ce-Al(d) sample and 480 °C over the Co/20Ce-Al(d/h)
sample.
References
[1] G. Zhai, J. Wang, Z. Chen, W. An, Y. Men, Chem. Eng. J. 337 (2018) 488–498,
https://doi.org/10.1016/j.cej.2017.12.141.
[2] B. Bai, H. Arandiyan, J. Li, Appl. Catal. B: Environ. 142–143 (2013) 677–683,
https://doi.org/10.1016/j.apcatb.2013.05.056.
[3] G. Ercolino, P. Stelmachowski, G. Grzybek, A. Kotarba, S. Specchia, Appl. Catal. B:
Environ. 206 (2017) 712–725, https://doi.org/10.1016/j.apcatb.2017.01.055.
[4] K. Lehtoranta, T. Murtonen, H. Vesala, P. Koponen, J. Alanen, P. Simonen,
T. Rönkkö, H. Timonen, S. Saarikoski, T. Maunula, K. Kallinen, S. Korhonen, Emiss.
Control Sci. Technol. 3 (2017) 142–152, https://doi.org/10.1007/s40825-016-
0057-8.
[5] M. Keenan, R. Pickett, E. Tronconi, I. Nova, N. Kinnunen, M. Suvanto, T. Maunula,
K. Kallinen, R. Baert, Top. Catal. 62 (2019) 273–281, https://doi.org/10.1007/
s11244-018-1127-7.
[6] J. Chen, H. Arandiyan, X. Gao, J. Li, Catal. Surv. Asia 19 (2015) 140–171, https://
doi.org/10.1007/s10563-015-9191-5.
[7] F. Zasada, J. Janas, W. Piskorz, M. Gorczynska, Z. Sojka, ACS Catal. 7 (2017)
4. Conclusions
2853–2867, https://doi.org/10.1021/acscatal.6b03139.
[8] F. Zasada, W. Piskorz, J. Janas, J. Grybos, P. Indyka, Z. Sojka, ACS Catal. 5 (2015)
6879–6892, https://doi.org/10.1021/acscatal.5b01900.
[9] L.F. Liotta, G. Di Carlo, G. Pantaleo, A.M. Venezia, G. Deganello, Appl. Catal. B:
Environ. 66 (2006) 217–227, https://doi.org/10.1016/j.apcatb.2006.03.018.
[10] Z. Pu, H. Zhou, Y. Zheng, W. Huang, X. Li, Appl. Surf. Sci. 410 (2017) 14–21,
https://doi.org/10.1016/j.apsusc.2017.02.186.
The effect of cerium deposition on alumina-supported cobalt oxide
catalysts for the complete oxidation of lean methane was studied. The
addition of cerium (5–30 %wt.) to the alumina was carried out prior to
the deposition of cobalt (30 %wt.) by a precipitation synthesis route.
The textural properties of the modified supports were hardly modified
with respect to those of bare alumina, due to the good dispersion of
ceria over the surface of the alumina. This dispersion also was re-
sponsible for a relatively strong interaction between alumina and de-
posited ceria that resulted in an increased Ce3+ content, especially at
low Ce loadings (5–15 %wt.). For higher Ce loadings (20–30 %wt.) the
cerium-alumina interaction was weaker, thus resulting in a deposited
ceria with a higher Ce4+ content and better reducibility with respect to
pure bulk ceria.
When cobalt was deposited over the modified supports, a dual effect
of ceria was evidenced. On one hand, ceria was found to act as a
physical barrier between alumina and deposited cobalt, reducing the
cobalt-alumina interaction and the subsequent cobalt aluminate for-
mation, more remarkably with higher Ce loadings (20–30 %wt.). On the
other hand, the interaction between cobalt and ceria led to a partial
incorporation of cerium ions into the structure of the cobalt oxide. Both
findings ultimately led to a larger abundance of Co3+ and therefore to a
promoted mobility of the lattice oxygen species with ceria loading.
Consequently, cobalt catalysts supported over ceria-alumina were sig-
nificantly more active than that supported over bare alumina. The best
behaviour was found for the Co/20Ce-Al sample. Furthermore, this
sample exhibited a relatively good thermal stability for prolonged re-
action time intervals (150 h) under dry conditions. However, the pre-
sence of water vapour markedly affected the catalytic performance al-
though this negative effect was partially reversible.
[11] Z. Chen, S. Wang, W. Liu, X. Gao, D. Gao, M. Wang, S. Wang, Appl. Catal. A: Gen.
525 (2016) 94–102, https://doi.org/10.1016/j.apcata.2016.07.009.
[12] Q. Wang, Y. Peng, J. Fu, G.Z. Kyzas, S.M.R. Billah, S. An, Appl. Catal. B: Environ.
168–169 (2015) 42–50, https://doi.org/10.1016/j.apcatb.2014.12.016.
[13] G. Grzybek, K. Ciura, S. Wójcik, J. Gryboś, P. Indyka, M. Inger, K. Antoniak-Jurak,
P. Kowalik, A. Kotarba, Z. Sojka, Catal. Sci. Technol. 7 (2017) 5723–5732, https://
doi.org/10.1039/C7CY01575E.
[14] I. Zacharaki, C.G. Kontoyannis, S. Boghosian, A. Lycourghiotis, C. Kordulis, Catal.
Today 143 (2009) 38–44, https://doi.org/10.1016/j.cattod.2008.09.026.
[15] B. Solsona, T.E. Davies, T. Garcia, I. Vázquez, A. Dejoz, S.H. Taylor, Appl. Catal. B:
Environ. 84 (2008) 176–184, https://doi.org/10.1016/j.apcatb.2008.03.021.
[16] L.F. Liotta, G. Pantaleo, A. Macaluso, G. Di Carlo, G. Deganello, Appl. Catal. A: Gen.
245 (2003) 167–177, https://doi.org/10.1016/S0926-860X(02)00652-X.
[17] J. Cheng, J. Yu, X. Wang, L. Li, J. Li, Z. Hao, Energy Fuels 22 (2008) 2131–2137,
https://doi.org/10.1021/ef8000168.
[18] S. Park, G. Kwak, Y. Lee, K. Jun, Y.T. Kim, Bull. Korean Chem. Soc. 39 (2018)
540–547, https://doi.org/10.1002/bkcs.11431.
[19] J. Park, S. Yeo, T. Kang, I. Heo, K. Lee, T. Chang, Fuel 212 (2018) 77–87, https://
doi.org/10.1016/j.fuel.2017.09.090.
[20] Y. Zhang, H. Xiong, K. Liew, J. Li, J. Mol. Catal. A: Chem. 237 (2005) 172–181,
https://doi.org/10.1016/j.molcata.2005.04.057.
[21] V. Garcilaso, J. Barrientos, L.F. Bobadilla, O.H. Laguna, M. Boutonnet,
M.A. Centeno, J.A. Odriozola, Renew. Energy 132 (2019) 1141–1150, https://doi.
org/10.1016/j.renene.2018.08.080.
[22] M.A. Ulla, R. Spretz, E. Lombardo, W. Daniell, H. Knözinger, Appl. Catal. B:
Environ. 29 (2001) 217–229, https://doi.org/10.1016/S0926-3373(00)00204-6.
[23] S. Ji, S. Ji, H. Wang, E. Flahaut, K.S. Coleman, M.L.H. Green, Catal. Lett. 75 (2001)
65–71, https://doi.org/10.1023/A:1016711207912.
[24] A. Choya, B. de Rivas, J.R. González-Velasco, J.I. Gutiérrez-Ortiz, R. López-Fonseca,
Appl. Catal. B: Environ. 237 (2018) 844–854, https://doi.org/10.1016/j.apcatb.
2018.06.050.
[25] A. Choya, B. de Rivas, J.I. Gutiérrez-Ortiz, R. López-Fonseca, Catalysts 8 (2018)
427, https://doi.org/10.3390/catal8100427.
10

A. Choya, et al.
AppliedCatalysisA,General591(2020)117381
[26] K.M.S. Khalil, J. Colloid Interface Sci. 307 (2007) 172–180, https://doi.org/10.
1016/j.jcis.2006.11.039.
S. Górecka, M. Molenda, Catal. Today 333 (2019) 196–207, https://doi.org/10.
1016/j.cattod.2018.03.042.
[27] Q. Yu, X. Wu, C. Tang, L. Qi, B. Liu, F. Gao, K. Sun, L. Dong, Y. Chen, J. Colloid
Interface Sci. 354 (2011) 341–352, https://doi.org/10.1016/j.jcis.2010.10.043.
[28] A. Martínez-Arias, M. Fernández-García, L.N. Salamanca, R.X. Valenzuela,
J.C. Conesa, J. Soria, J. Phys. Chem. B 104 (2000) 4038–4046, https://doi.org/10.
1021/jp992796y.
[29] B. de Rivas, C. Sampedro, E.V. Ramos-Fernández, R. López-Fonseca, J. Gascon,
M. Makkee, J.I. Gutiérrez-Ortiz, Appl. Catal. A: Gen. 456 (2013) 96–104, https://
doi.org/10.1016/j.apcata.2013.02.026.
[40] J. González-Prior, R. López-Fonseca, J.I. Gutiérrez-Ortiz, B. de Rivas, Appl. Catal. B:
Environ. 222 (2018) 9–17, https://doi.org/10.1016/j.apcatb.2017.09.050.
[41] M. Singh, N.K. Ralhan, S. Singh, Bull. Mater. Sci. 38 (2015) 297–301, https://doi.
org/10.1007/s12034-015-0872-5.
[42] M.C. Biesinger, B.P. Payne, A.P. Grosvenor, L.W.M. Lau, A.R. Gerson, R.S.C. Smart,
Appl. Surf. Sci. 257 (2011) 2717–2730, https://doi.org/10.1016/j.apsusc.2010.10.
051.
[43] C. Tang, C. Wang, S. Chien, Thermochim. Acta 473 (2008) 68–73, https://doi.org/
10.1016/j.tca.2008.04.015.
[30] M. Gil-Calvo, C. Jiménez-González, B. de Rivas, J.I. Gutiérrez-Ortiz, R. López-
Fonseca, Ind. Eng. Chem. Res. 56 (2017) 6186–6197, https://doi.org/10.1021/acs.
iecr.7b00986.
[44] Ming Meng, Y.Q. Zha, J.Y. Luo, T.D. Hu, Y.N. Xie, T. Liu, J. Zhang, Appl. Catal. A:
Gen. 301 (2006) 145–151, https://doi.org/10.1016/j.apcata.2005.06.035.
[45] H. Sohn, G. Celik, S. Gunduz, D. Dogu, S. Zhang, J. Shan, F. Feng Tao, U.S. Ozkan,
Catal. Lett. 147 (2017) 2863–2876, https://doi.org/10.1007/s10562-017-2176-4.
[46] L. Hu, Q. Peng, Y. Li, J. Am. Chem. Soc. 130 (2008) 16136–16137, https://doi.org/
10.1021/ja806400e.
[31] Y. Ji, Z. Zhao, A. Duan, G. Jiang, J. Liu, J. Phys. Chem. C 113 (2009) 7186–7199,
https://doi.org/10.1021/jp8107057.
[32] Q. Liu, L. Wang, M. Chen, Y. Cao, H. He, K. Fan, J. Catal. 263 (2009) 104–113,
https://doi.org/10.1016/j.jcat.2009.01.018.
[33] D.K. Chlebda, R.J. Jędrzejczyk, P.J. Jodłowski, J. Łojewska, J. Raman Spectrosc. 48
[47] A.W. Budiman, S. Song, T. Chang, C. Shin, M. Choi, Catal. Surv. Asia 16 (2012)
(2017) 1871–1880, https://doi.org/10.1002/jrs.5261.
183–197, https://doi.org/10.1007/s10563-012-9143-2.
[34] X. Jiang, Y. Ma, Y. Chen, Y. Li, Q. Ma, Z. Zhang, C. Wang, Y. Yang, Spectrochim.
Acta Part A: Mol. Biomol. Spectrosc. 190 (2018) 61–67, https://doi.org/10.1016/j.
saa.2017.08.076.
[48] M. Gil-Calvo, C. Jiménez-González, B. De Rivas, J.I. Gutiérrez-Ortiz, R. López-
Fonseca, Chem. Eng. Trans. 57 (2017) 901–906, https://doi.org/10.3303/
CET1757151.
[35] V. D’Ippolito, G.B. Andreozzi, D. Bersani, P.P. Lottici, J. Raman Spectrosc. 46
[49] N. Bahlawane, Appl. Catal. B: Environ. 67 (2006) 168–176, https://doi.org/10.
(2015) 1255–1264, https://doi.org/10.1002/jrs.4764.
1016/j.apcatb.2006.03.024.
[36] Q. Zhang, S. Mo, B. Chen, W. Zhang, C. Huang, D. Ye, Mol. Catal. 454 (2018) 12–20,
https://doi.org/10.1016/j.mcat.2018.05.006.
[37] D. Li, L. Wang, M. Koike, Y. Nakagawa, K. Tomishige, Appl. Catal. B: Environ. 102
(2011) 528–538, https://doi.org/10.1016/j.apcatb.2010.12.035.
[38] G. Zou, Y. Xu, S. Wang, M. Chen, W. Shangguan, Catal. Sci. Technol. 5 (2015)
1084–1092, https://doi.org/10.1039/c4cy01141d.
[50] P. Stefanov, S. Todorova, A. Naydenov, B. Tzaneva, H. Kolev, G. Atanasova,
D. Stoyanova, Y. Karakirova, K. Aleksieva, Chem. Eng. J. 266 (2015) 329–338,
https://doi.org/10.1016/j.cej.2014.12.099.
[51] A.K. Dalai, B.H. Davis, Appl. Catal. A: Gen. 348 (2008) 1–15, https://doi.org/10.
1016/j.apcata.2008.06.021.
[52] Ø Borg, Z. Yu, D. Chen, E.A. Blekkan, E. Rytter, A. Holmen, Top. Catal. 57 (2014)
491–499, https://doi.org/10.1007/s11244-013-0205-0.
[39] R. Dziembaj, A. Chojnacka, Z. Piwowarska, M. Gajewska, M. Swietoslawski,
11