Effects of adding basic oxides of la and/or Ce to SiO2-supported Co catalysts for ethanol steam reforming

Article abstract of DOI:10.1039/c4ra04157g

In this study, cobalt catalysts supported on La₂O₃-CeO₂-SiO₂ were synthesized by the method of impregnation of the support with aqueous solution of the active phase precursor. These materials were tested in etha

Full text of DOI:10.1039/c4ra04157g

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Effects of adding basic oxides of La and/or Ce to
SiO2-supported Co catalysts for ethanol steam
reforming
Cite this: RSC Adv., 2014, 4, 43839
*
F. C. F. Marcos, A. F. Lucredio and E. M. Assaf
´
In this study, cobalt catalysts supported on La2O3–CeO2–SiO2 were synthesized by the method of
impregnation of the support with aqueous solution of the active phase precursor. These materials were
tested in ethanol steam reforming (ESR). The catalysts were characterized by in situ X-ray diffraction
(XRD), temperature-programmed reduction (H2-TPR), temperature-programmed hydrogen desorption
(H2-TPD), temperature-programmed oxidation (O2-TPO), adsorption–desorption of nitrogen (BET) and
energy-dispersive X-ray spectroscopy (EDS). After reaction, the spent materials were characterized by
XRD, SEM and elemental analysis. The in situ XRD analysis showed the formation of the solid solution
CexLa1ꢀxO2ꢀd and identified cobalt oxide in the spinel structure (Co3O4) and metallic cobalt (Co0). The
H2-TPR analysis showed the reduction of cobalt oxide, and it was observed that the peaks of cobalt
oxide shifted to lower temperatures as the content of lanthanum increased. The H2-TPD analysis
showed that metal surface area and metal dispersion decreased as the percentage of lanthanum oxide in
the supports increased. The O2-TPO analysis showed that La2%Ce48%Si50% consumed the most oxygen.
The SEM analysis identified filamentous carbon after reaction. The catalysts were active for H2 and CO2
production. The best catalytic performance was shown by Co10%La2%Ce48%Si50%, which gave the highest
H2 yield.
Received 5th May 2014
Accepted 26th August 2014
DOI: 10.1039/c4ra04157g
www.rsc.org/advances
This process is usually accompanied by parallel reactions,
which include: (a) dehydrogenation of ethanol to acetaldehyde,
Introduction
followed by acetaldehyde decomposition into methane and
carbon monoxide, (b) dehydration of ethanol to ethylene, (c)
methane steam reforming to produce carbon monoxide and
hydrogen and (d) water–gas shi reaction (WGSR) between
water and carbon monoxide, forming carbon dioxide and
hydrogen. Therefore, the conversion of ethanol and the product
distribution is directly related to the catalyst composition and
the operational conditions in the reaction system.3
In the literature, supported cobalt catalysts have shown
promising results ESR due to their ability to break the C–C bond
and also their low cost, compared to noble metals, like Rh
catalysts.2 Regarding the choice of support, CeO2 has advan-
tages over other supports, especially due to its redox properties,
the presence of oxygen vacancies and the activation of oxygen
compounds; however, its low thermal stability requires the use
of additives to prevent sintering with consequent catalytic
deactivation.4 Thus, when ceria is combined with other metal
oxide, a solid solution can be formed and the catalytic perfor-
mance is improved.5
Alternative energy sources that do not harm the environment
have great relevance and have attracted new study proposals.
Such relevance is justied by the increasing concern of envi-
ronmental degradation and the need to replace fossil fuels by
biofuels,1 which are produced from renewable sources. Bio-
fuels, such as for H2 production, can even be more efficient than
conventional fuels when consumed in fuel cells. However, an
effective and economical environmentally friendly biofuel is
still the subject of several studies. Ethanol is an attractive choice
of raw material for producing H2 because, besides its low eco-
toxicity, it can be produced from biomass and utilized without
net addition of carbon dioxide to the atmosphere.2
Hydrogen may be produced by catalytic steam reforming of
ethanol. The global reaction of hydrogen production by ethanol
steam reforming (ESR) leads to formation of 6 moles of H2 per
mole of ethanol (eqn (1)):
C2H5OH + 3H2O / 2CO2 + 6H2, DHꢁ298 K ¼ +174 kJ molꢀ1
(1)
It is known that carbon is formed on Co/CeO2 catalysts
during the ethanol steam reforming at 500 ꢁC, and the catalyst
particles are covered in coke, leading to severe deactivation.2
In this respect, the literature discusses the effect of incor-
porating certain metals into CeO2-based catalysts, such as Ca,
˜
˜
˜
Institute of Chemistry of Sao Carlos, University of Sao Paulo, 13560-970, Sao
Carlos-SP, Brazil. E-mail: eassaf@iqsc.usp.br; Fax: +55 16 33739952; Tel: +55 16
33739484
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 43839–43849 | 43839

View Article Online
RSC Advances
Paper
Pr or La,6,7 that cause unit cell expansion, resulting in more
The specic surface area, average pore size and total pore
oxygen storage capacity (OSC) and higher oxygen mobility. This volume of the catalysts were measured by N2 physisorption at
increases hydrogen yield, carbon and CO oxidation and catalyst liquid N2 temperature (77 K) in a Quantachrome Nova 1000 gas
stability. The OSC is related to the formation of oxygen vacan- adsorption analyzer. The samples were previously outgassed at
cies and the diffusion of oxygen ions between surface and 250 ꢁC for 2 h. Surface area, average pore size and total pore
bulk.8,9
volume were determined by Brunauer–Emmett–Teller (BET)
Thus, the incorporation of La3+ into the CeO2 lattice at Ce4+ and BJH methods.
sites leads to formation of a solid solution, CexLa1ꢀxO2ꢀd, which
X-ray diffraction (XRD) analysis was carried out to identify
contains oxygen vacancies that promote higher H2 yield, carbon the crystalline structures of the prepared samples and these
gasication and oxidation of CO to CO2.6,10 However, this solid experiments were performed in a Rigaku Multiex diffractom-
˚
solution has a low surface area, which is unsuitable for a good eter with a Cu Ka (1.5406 A) radiation source. The X-ray patterns
support catalyst. One way to solve this problem is to disperse were obtained for 2q values ranging from 20ꢁ to 80ꢁ. The
this solid solution over a high area support, such as silica. SiO2 diffraction patterns were identied by comparison with those of
has excellent support characteristics, such as a high surface known structure recorded in the JCPDS (Joint Committee of
area, low acidity and thermal and chemical stability, properties Powder Diffraction Standards) database.
that are essential to support the surface processes.11
X-ray diffraction patterns were also recorded in situ under
In light of the above, the goal of this study was the study of a activation and reaction conditions at the XPD-10B beam line of
catalyst containing active sites of cobalt on a CeO2–La2O3 solid the LNLS synchrotron radiation facility (Campinas, Brazil). The
solution, dispersed on a high surface area support, SiO2, to measurements were carried out in the reection mode, in a 2q
assess the effectiveness of the support vacancies in ethanol interval from 25ꢁ to 55ꢁ, using radiation of wavelength 1.54996
˚
steam reforming for H2 production.
A, selected with a Si (111) monochromator. Lattice parameters
were estimated from the (111) peak of the CeO2 phase. The
oxidized samples were heated, under H2 ow, from 25 ꢁC to 700
ꢁC, and held at 700 ꢁC for 1 hour. During this heating, a
diffraction patterns were collected and the average size of Co
metal crystallites on the catalysts was calculated from Scherrer's
equation: d ¼ kl/b cos q; where b is the full width at half
Experimental
Catalyst preparation
The supports, containing La2O3, CeO2 and SiO2 (Degussa
aerosil-200) in mass ratios of LaxCe50ꢀxSi50 (x ¼ 2; 4; 6), were
prepared by co-precipitation.12 An appropriate mixture of
aqueous solutions of Ce and La nitrates was co-precipitated on
the silica at pH 10 by addition of (NH4)2CO3 powder and 5 M
NaOH solution at room temperature. The resulting precipitate
˚
maximum of the (111) peak; k ¼ 0.90 and l ¼ 1.54996 A.
Temperature-programmed reduction (H2-TPR), was per-
formed in a U-shaped quartz tube reactor with a mixture of H2
(1.96% v/v)/Ar owing at 30 mL minꢀ1. The catalyst sample (100
ꢁ
mg) was heated from room temperature to 1000 C at a rate of
ꢁ
10 ꢁC minꢀ1. The water produced during the reduction was
removed by passing the effluent gas through a tube containing
dried silica gel. The outlet gas was analyzed with a thermal
conductivity detector (TCD) and the degree of reduction (%) was
calculated by comparing the corresponding peak area with that
produced by a standard CuO sample.
was ltered, washed with distilled water, dried at 80 C for 24
hours and nally calcined at 500 ꢁC for 4 hours under air
owing at 50 mL minꢀ1
.
Besides these mixed supports, pure CeO2 and La2O3
supports were prepared from cerium nitrate and lanthanum
nitrate that were precipitated separately by mixing with
(NH4)2CO3 powder and 5 M NaOH solution at room tempera-
ture. The oxides were obtained by washing, drying and calcining
in owing air under the same conditions as for the mixed
supports.
The catalysts were prepared by impregnation of the supports
with an aqueous solution of cobalt nitrate. Co loading was
adjusted to 10 wt% on all the supports. The resulting precursors
were dried at 80 ꢁC for 24 hours and calcined in air at 500 ꢁC for
3 hours.
The percent reduction of Co3O4 was found by subtracting the
H2 consumption of the supports alone.
H2-TPD measurements were performed in a Micromeritics
Pulse ChemSorb 2750 equipped with a (TCD). 50 mg of catalyst
ꢁ
was reduced for 1 hour at 700 C in pure H2. Next, the sample
was cooled to 25 ꢁC in N2 owing at 30 mL minꢀ1. At this
temperature, the sample was then subjected to a 30 mL minꢀ1
ow of pure hydrogen for 1 hour until saturated with adsorbed
H2. The hydrogen was then purged by N2 owing at
30 mL minꢀ1 for 2 hours, to expel accumulated hydrogen from
the pores, so that only chemisorbed hydrogen would remain in
the catalyst. Aer the purging process, the hydrogen desorption
measurement started in a 30 mL minꢀ1 ow of nitrogen, with a
heating rate of 10 ꢁC minꢀ1 up to 700 ꢁC, and the sample
remained at that temperature until the end of the
The prepared samples were designated Co10%La2%Ce48%Si50%
Co10%La4%Ce46%Si50% Co10%La6%Ce44%Si50% Co10%CeO2,
,
,
,
Co10%La2O3 and Co10%SiO2.
Characterization of samples
Energy-dispersive X-ray spectroscopy (EDS) was employed to measurement.
determine the chemical composition. The equipment used was
a Link QX 2000 EDS analyzer coupled to a LEO 440 electron eqn (2).
microscope linked to an Oxford detector.
The specic metal surface area (SM) was calculated by
43840 | RSC Adv., 2014, 4, 43839–43849
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
À
Á
unconverted ethanol, were collected aer the 6 hour reaction
and analyzed by gas chromatography (Hewlett Packard 5890),
with a HP-FFAP capillary column (25 m ꢂ 0.2 mm i.d) and FID
detector. The conversion of ethanol was taken as the difference
between the total volume of condensed ethanol aer the reac-
tion and the total volume of ethanol in the feed during the test.
The yield (Yi) of each product (i ¼ H2, CH4, CO2, CO or C2H4)
NH ðadsÞNAvnE
rmcat
2
SM
¼
(2)
NH (ads) ¼ amount H2 adsorbed (mol); NAv ¼ Avogadro number
2
(6.02 ꢂ 1023 atoms per mol); nE ¼ adsorption stoichiometry
factor (H2/Co3O4 ¼ 4/3); r ¼ density of metal sites (Co ¼ 1.46 ꢂ
1019 atoms per m );
m
cat
¼ catalyst mass (g).
2
13
The metal dispersion (D) was calculated by eqn (3).
moles of i produced
was calculated by eqn (4):Yi ¼
moles of ethanol fed
NH ðTPDÞ
2
Dð%Þ ¼
ꢂ 100
(3)
NH ðTPRÞ
2
NH (TPD) ¼ H2 released by TPD (mmol gcatꢀ1); NH (TPR) ¼ H2
2
2
Results and discussion
Characterization
consumed by TPR (mmol gcatꢀ1).
A similar procedure was used for temperature-programmed
oxidation (O2-TPO) analysis. Aer cooling to room tempera-
ture, the reduced supports were heated up to 500 ꢁC at a rate of 5
ꢁC minꢀ1 in the presence of 2 vol.% O2–He mixture gas owing
at 15 mL minꢀ1, using the apparatus described above.
The carbon formed during the reaction was determined by
an EA1110 CHNS-O Elemental Analyzer, equipped with a Por-
apak PQS column.
Table 1 shows the average mass percentage of each metal in the
catalysts, determined by EDS. It can be seen that the real
contents of individual elements are close to the nominal values.
Values of average pore size, total pore volume of catalysts
and specic surface area of the catalysts, calcined at 500 ꢁC and
reduced at 700 ꢁC, and supports are shown in Table 1. The
incorporation of 2% La2O3 into the CeO2 lattice increased its
average pore size and total pore volume, but when the La2O3
content increased to 4% and then 6%, the pore size values fell.
It can be seen that the pure CeO2 and La2O3 showed low surface
areas compared to samples supported on SiO2. The impregna-
tion of the supports with active phase (Co) reduced their surface
areas, except for the catalyst Co10%La2O3 (before reduction),
Morphology of the carbon was observed by scanning electron
microscopy (SEM) carried out in a LEO 440 microscope with an
Oxford detector, operating with a 20 kV electron beam.
Catalytic test
The catalytic tests were performed at atmospheric pressure, in a whose surface area increased. This increase may be due to the
xed-bed tubular quartz micro reactor. The product composi- direct addition of the surface area of the cobalt oxide species to
tion was recorded over 6 hours. Prior to the reaction, 150 mg of the low support area, leading to a higher total specic area. The
the catalyst was introduced into the reactor and reduced in situ decrease in surface area of the other catalysts can be attributed
in a ow of H2 (30 mL minꢀ1) at 700 ꢁC (10 ꢁC minꢀ1) for 1 hour, to the covering of the support by ne particles of cobalt oxide
to activate the catalyst. Next, the sample was brought down to deposited in the pores of the matrix, causing a surface area
the reaction temperature (500 ꢁC) in a ow of pure N2, purging reduction of the material and an increase in pore size, or may be
all H2 from the system. The temperature was controlled with a due to the aggregation of silica during heat treatments. Ana-
thermocouple introduced close to the catalytic bed. The cata- lysing the surface area of the reduced catalyst, it is observed that
lysts were tested in a hydrogen-free feed of molar ratio water– the reduction of the catalysts caused a loss of surface area in all
ethanol ¼ 3 : 1, owing at a constant rate of 3.5 mL hꢀ1. The the catalysts (Table 1), especially in Co10%CeO2 and
gaseous products were analyzed in-line by gas chromatography Co10%La2O3, which showed a considerable decrease in surface
in a CP-3800 Varian chromatograph with two parallel columns area aer reduction. This reduction is a consequence of the low
of Porapak N and Molecular Sieve each with a TCD used to thermal stability of these supports causing a sintering of these
analyze effluent. The condensed liquid products and materials. For the catalysts supported on the solid solution, a
Table 1 Average mass percentage of metal elements; average pore size, total pore volume; specific surface area (before and after activation) and
surface area of support
Average mass
percentages of metals in
the catalysts
Average pore Total pore volume
S
S
S
Catalysts
Co
La
Ce
size (A)
(cm3 gꢀ1
)
(m2 gꢀ1
)
(m2 gꢀ1) red Supports
(m2 gꢀ1
)
˚
Co10%SiO2
Co10%CeO2
Co10%La2O3
8 (10)
9 (10)
10 (10)
98
19
94
1.46
0.06
0.97
0.34
0.32
0.24
168
64
25
131
109
98
160
7
5
95
81
52
SiO2
CeO2
La2O3
La2%Ce48%Si50% 186
La4%Ce46%Si50% 143
La6%Ce44%Si50% 140
231
107
11
Co10%La2%Ce48%Si50% 10 (10) 2 (2) 44 (48) 90
Co10%La4%Ce46%Si50% 11 (10) 3 (4) 42 (46) 77
Co10%La6%Ce44%Si50% 11 (10) 5 (6) 41 (44) 54
Nominal percentage ( )
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 43839–43849 | 43841

View Article Online
RSC Advances
Paper
Fig. 1 XRD patterns of (A) supports and (B) catalysts: (a) Co10%La6%Ce44%Si50%, (b) Co10%La4%Ce46%Si50%, (c) Co10%La2%Ce48%Si50%, (d) Co10%CeO2,
(e) Co10%La2O3 and (f) Co10%SiO2.
decrease in surface area is also seen aer reduction, but it is not be due to solid solution formation arising from the partial
so pronounced, due to the thermal stability of the SiO2 support. replacement of Ce4+ (ionic radius 0.097 nm) with La3+ (ionic
XRD patterns are illustrated in Fig. 1. The CeO2 support radius 0.11 nm). This insertion of the larger La3+ ion into the
showed peaks referring to the uorite structure of face-centered CeO2 lattice causes expansion of the lattice (Fig. 1). As the La3+
cubic (fcc) (JCPDS 81-0792).
content increases, there is therefore a displacement of the main
The La2O3 support showed diffraction peaks corresponding peak of CeO2 to smaller angles.7 The analysis of XRD patterns
to the phases La2O3 (JCPDS 83-1344) and La(OH)3 (JCPDS 36- and calculation of the lattice parameter of CeO2 (Table 2)
1481), which may have been formed during the co-precipitation, conrmed the formation of a solid solution between the cerium
by reaction of NH4OH and La(NO3)3.
and lanthanum oxides, because the lattice parameters of the
For the SiO2 support, the broad peak centered at 2q ¼ 21.8ꢁ uorite structure of CeO2 increased linearly with La content.7 In
was assigned to the SiO2 structure (JCPDS 82-1576), which the diffractograms of the catalysts, there were peaks relating to
showed low crystallinity.
the supports and also peaks for the Co3O4 crystalline spinel
The XRD patterns of Ce–La–Si supports showed peaks structure (JCPDS 74-2120).
related to the uorite face-centered cubic CeO2 structure (JCPDS
Fig. 2 and 3 show the in situ XRD patterns collected at several
81-0792). The absence of diffraction peaks for La2O3 on mixed temperatures during the process of reduction with H2.
supports suggests that ceria is stabilized by lanthanum. In the
These temperatures correspond to the temperatures of
aligned XRD patterns, it is possible to observe displacement of reduction of cobalt species during TPR. Initially, peaks due to
the main CeO2 peak as La2O3 is added to the support. This may Co3O4 spinel reections were visible, at 2q ¼ 31.2ꢁ; 36.8ꢁ; 38.4ꢁ
Table 2 CeO2 lattice parameter (A˚) in supports; average crystallite sizes before and after reduction
˚
Aer reduction
dCo0 (111) (nm)
Lattice parameter (A) CeO2
Before reduction
Support
a (111)
Catalyst
dCeO2 (111)
dCo3O4 (311) (nm)
CeO2
5.4015
5.4216
5.4424
5.4632
Co10%SiO2
Co10%CeO2
Co10%La2O3
Co10%La2%Ce48%Si50%
Co10%La4%Ce46%Si50%
Co10%La6%Ce44%Si50%
CeO2
—
16.1
—
4.9
12.5
8.8
12.4
18.6
10.7
16.1
12.9
20.0
19.1
—
20.4
27.8
27.0
29.6
23.9
23.9
—
La2%Ce48%Si50%
La4%Ce46%Si50%
La6%Ce44%Si50%
CeO2(JCPDS-81-0792)
a ¼ 5.4124
43842 | RSC Adv., 2014, 4, 43839–43849
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Fig. 2 In situ XRD patterns taken during reduction of catalyst with H2: (A) Co10%SiO2 (B) Co10%CeO2 and (C) Co10%La2O3.
and 44.7ꢁ. As the temperature was increased, these peaks related to the reduction of O2ꢀ in the bulk of the material by
progressively disappeared and others related to CoO appeared removal of the anion from the lattice and subsequent formation
at 2q ¼ 36.6ꢁ and 42.4ꢁ. At 365 ꢁC, CoO started to be trans- of Ce2O3. The removal of this oxygen occurs at high tempera-
formed into fcc Co0, with reections at 2q ¼ 44.2ꢁ and 51.4ꢁ. tures, because it needs high energy for its transport to the
These progressive transformations are consistent with those surface. The reduction reaction is shown in eqn (5).
observed by O'Shea et al. and Bshish et al.7,14 and correspond to
the two Co3O4 reduction steps (Co3O4 / CoO / Coo).14
The average crystallite sizes of CeO2 and Co3O4 are shown in
Table 2. It can be observed that the average cobalt oxide crys-
tallite (before reduction) was smaller than the metallic cobalt
crystallite (aer reduction), suggesting that the reduction
process increased the crystallite size by aggregation or sintering
of the active phase. It can be said that these values are consis-
tent with the catalyst surface area: as the catalysts were reduced,
the surface area decreased and the crystallite size increased.
Fig. 4A shows the H2-TPR proles for the supports and it can
be observed in the SiO2 prole, as already known that this oxide
did not suffer reduction in the temperature range examined.
The CeO2 H2-TPR prole showed two reduction peaks. The
2CeO2 + H2 / Ce2O3 + H2O
(5)
In the H2-TPR prole for La2O3, it can be seen that this
support has a single reduction peak at 690 ꢁC. According to the
literature, lanthanum oxides are reduced with difficulty because
of a lack of oxygen species on the surface,17 but according to
Fleys et al.18 this reduction can be promoted by hydroxide
groups and carbonate that may be le in the support by the
precursors used in the preparation; according to the XRD
analysis, this reduction step may be related to the presence of
La(OH)3.17
The proles of the mixed supports (Ce–La–Si) showed two
regions of reduction at temperatures below 650 ꢁC and one at a
higher temperature. The peaks at temperatures below 650 ꢁC
are related to the simultaneous reduction of cerium and
lanthanum species. The addition of La3+ to the CeO2 lattice
ꢁ
peak at low temperature (<600 C) is attributed to the surface
reduction of CeO2, which corresponds to non-stoichiometric
reduction of the cerium oxide.15,16 The peak at 800 ꢁC is
Fig.
3 In situ XRD patterns taken during reduction of catalyst with H2: (A) Co10%La2%Ce48%Si50%, (B) Co10%La4%Ce46%Si50% and (C)
Co10%La6%Ce44%Si50%
.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 43839–43849 | 43843

View Article Online
RSC Advances
Paper
Fig. 4 H2-TPR profiles of (A) supports and (B) catalysts.
caused an increase in the O2 on the CeO2 surface, leading to
Table 3 shows the degree of reduction, metal area and
more intense peaks.10,19 The second region of reduction refers to dispersion of the catalysts supported on La2O3–CeO2–SiO2. The
the reduction of oxygen in the bulk CeO2.
Co10%La6%Ce44%Si50% has a smaller metallic area than Co10%-
Fig. 4B shows the H2-TPR proles for the cobalt catalysts. La4%Ce46%Si50%. This decrease might be due to the aggregation
The prole of the catalyst Co10%SiO2 has two peaks around 333 of cobalt into larger particles aer reduction, facilitated by
ꢁ
ꢁC and 360 C, which correspond to the reduction of Co3O4 to La2O3–CeO2–SiO2 interaction with the metal sites. The Co10%
-
CoO and then to Co0;20 these steps were also monitored by in La2%Ce48%Si50% catalyst showed the largest metallic area and
situ XRD, as presented in Fig. 2. the La2O3–CeO2–SiO2 interaction favoured the growth of
In the H2-TPR prole of the Co10%CeO2 catalyst, two reduc- smaller cobalt crystals. However, addition of lanthanum and
tion regions can be seen; the rst region, around 410 ꢁC, can be decrease of cerium in supports causes a decrease in the cobalt
related to the reduction of cobalt species and the CeO2 surface dispersion. This decrease is attributed to the coating of cobalt
oxygen, which also has reduction peaks in this temperature metal particles by lanthanum oxide.21 According to in situ XRD
ꢁ
range. The reduction peak at 650 C corresponds to the reduc- (Fig. 3), Co10%La4%Ce46%Si50% and Co10%La6%Ce44%Si50%
tion of bulk CeO2. This decrease in the temperature of CeO2 exhibited the same crystallite size and degree of reduction of
bulk reduction was due to the presence of Co3O4 interacting Co3O4.
with the oxygen of the support, which enhanced oxygen
mobility, leading to its reduction at a lower temperature.4
Fig. 5 shows the O2-TPO proles of the supports aer they
had been reduced at the temperature of activation of the cata-
In the H2-TPR prole of the Co10%La2O3 catalyst, two lyst (700 ꢁC). A broad peak appears at about 100 ꢁC for all
ꢁ
reduction peaks around 350 C and 650 ꢁC were observed. The
rst peak represents the direct reduction of cobalt oxide to
metal cobalt (Co3O4 / CoO / Co0) on the surface of the La2O3
matrix and the second peak represents reduction of hydroxyl
and carbonate species on the surface of La2O3, as also observed
in the La2O3 support.
The mixed catalysts Co/Ce–La–Si showed an increase in H2
consumption related to CeO2 reduction coupled with reduction
of Co3O4. The reduction temperature of cobalt interacting with
cerium decreased with increasing substitution of the samples
by the addition of La3+. This fall in temperature is the result of
the dissolution of La3+ ions in the lattice of cerium oxide,
causing distortion and generating oxygen vacancies in the
lattice with high mobility, thereby facilitating cobalt
reduction.6,8
Table 3 Degree of reduction (%) of the catalysts during TPR; metal
surface area and dispersion estimated by TPD
Co3O4
Metal areab
Dispersionb
(%)
ꢀ1
Catalysts
reductiona (%)
(m2 gcat
)
Co10%SiO2
Co10%CeO2
Co10%La2O3
Co10%La2%Ce48%Si50%
Co10%La4%Ce46%Si50%
Co10%La6%Ce44%Si50%
99
98
23
55
42
41
—
—
—
13.6
12.1
11.5
—
—
—
23.2
22.2
19.7
a
Tred ¼ 1000 ꢁC. b Tred ¼ 700 ꢁC.
43844 | RSC Adv., 2014, 4, 43839–43849
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Fig. 5 O2-TPO profiles of the supports.
supports, except the La2O3 sample, but with differences
between the peak areas. This temperature is consistent with the
values found in the literature.22 The areas are proportional the
consumption of oxygen.
The CeO2 support showed a small peak area and the La2O3
support did not show any oxygen-consumption peak. The
incorporation of La2O3 into the ceria lattice boosts the redox
activity of the oxide support by creating oxygen vacancies,
resulting in higher oxygen consumption in the O2-TPO analysis.
According to the literature, this also occurs when other metals
are incorporated into the ceria lattice.8,23
The largest oxygen-consumption area was observed for the
La2%Ce48%Si50% support, indicating that this support had the
highest number of mobile oxygen vacancies. Aer cobalt oxide
impregnation, this support showed the highest metal area, and
smaller average crystallite size for (Co3O4) and (CeO2). It was
found that the order of the oxygen-consumption area for the
Catalytic test
Table 4 and Fig. 6 show the catalytic test results for ethanol
steam reforming (ESR) at 500 ꢁC: the ethanol conversion, liquid
products yield and carbon formation aer 6 hours of ESR (Table
4) and gaseous product yields (Fig. 6).
It can be observed that the conversion of ethanol (Table 4)
was quite varied among the catalysts. Analysing the mixed
catalysts (Co/La–Ce–Si), we see that the catalyst Co10%La2%
Ce48%Si50% achieved the highest ethanol conversion and H2
yield and the lowest carbon deposition. The catalysts Co10%
La4%Ce46%Si50% and Co10%La6%Ce44%Si50% showed lower
values of ethanol conversion and H2 yield, than
-
-
Co10%La2%Ce48%Si50%
.
In general, the catalysts exhibited low formation of CO,
together with higher CO2 formation (CO2/CO > 2) (Fig. 6). Thus,
it can be said that the water–gas shi reaction was favoured
(eqn (6)) by catalysts supported on CeO2.14
mixed supports was La2%Ce48%Si50% > La4%Ce46%Si50% > La6%
-
Ce44%Si50%, the same as the order of reducibility observed in the
H2-TPR analysis.
CO + H2O(steam) / CO2 + H2
(6)
According to Reddy et al.,8 the introduction of La into the
CeO2 lattice decreases the activation energy of oxygen release by
the spontaneous generation of both localized and delocalized
oxygen vacancies around the dopant cations and, with an
increasing amount of La in the lattice, there is an enhancing of
the concentration of oxygen vacancies.8 The OSC is a function of
oxygen mobility between surface and bulk, and oxygen diffusion
is a function of the concentration and mobility of the oxygen
vacancies (delocalized vacancies). Now, according to Efstathiou
et al.,24 as the amount of La increases, more localized (or stable)
oxygen vacancies are formed, limiting the number of delo-
calized (mobile) vacancies.24 Thus, the results shown in Fig. 5
(O2-TPO) are consistent with this hypothesis: as the amount of
La increased, the O2 consumption decreased.
Co10%SiO2 (Fig. 6) showed a low yield for H2. This result and
its high catalytic conversion of ethanol suggest that the yield
was increased for liquid products such as acetone and acetal-
dehyde (Table 4).
Co10%CeO2 showed high ethanol conversion and carbon
deposition (Table 4). This high conversion was directed to the
formation of liquid products such as acetone and acetaldehyde
in moderate amounts, and acetic acid in very small amounts
(Table 4). Besides, during the reaction, this catalyst exhibited
deactivation (Fig. 6), which may have been due to oxidation of
metal cobalt sites and the formation of C by the Boudouard
reaction (eqn (7)).
2CO / CO2 + C
(7)
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 43839–43849 | 43845

View Article Online
RSC Advances
Paper
Table 4 Ethanol conversion (%); mean yields (mole i produced/mole ethanol fed) and carbon deposited (mmol hꢀ1) after 6 hours of ESR at 500
ꢁC
Yields of liquid products
Sample
Conversion (%)
Acetaldehyde
Acetone
Acetic acid
Ether
Carbon
Co10%SiO2
Co10%CeO2
Co10%La2O3
Co10%La2%Ce48%Si50%
Co10%La4%Ce46%Si50%
Co10%La6%Ce44%Si50%
La2%Ce48%Si50%
87
93
57
97
86
73
31
2.0 ꢂ 10ꢀ3
2.6 ꢂ 10ꢀ2
1.1 ꢂ 10ꢀ1
1.2 ꢂ 10ꢀ1
1.9 ꢂ 10ꢀ1
2.2 ꢂ 10ꢀ1
6.6 ꢂ 10ꢀ2
2.0 ꢂ 10ꢀ4
1.2 ꢂ 10ꢀ2
3.5 ꢂ 10ꢀ1
—
—
—
—
—
1.2
1.1
0.5
0.5
0.7
1.3
—
2.63 ꢂ 10ꢀ5
3.43 ꢂ 10ꢀ5
1.5 ꢂ 10ꢀ3
1.4 ꢂ 10ꢀ3
3.0 ꢂ 10ꢀ3
—
3.1 ꢂ 10ꢀ2
3.1 ꢂ 10ꢀ2
—
—
1.8 ꢂ 10ꢀ3
4.3 ꢂ 10ꢀ2
1.8 ꢂ 10ꢀ4
Co10%La2O3 catalyst (Table 4) showed low conversion and H2 oxidize CO and hydrocarbons arising from the decomposition
yield (Fig. 6), higher acetone and acetaldehyde yields, but little of ethanol to CO2.4 Furthermore, its yield was directed to other
deposition of C. This can be explained by the low surface area of liquid products such as acetaldehyde, acetone and ether.
this catalyst, which did not favour good dispersion of the cobalt
In general, the catalysts supported on ceria studied here,
on the surface. Moreover, the catalyst preparation method may showed good ethanol conversion, compared to other catalysts
not have formed enough active sites for chemisorption and reported the in literature.25,4
product formation.
Examining the mixed catalysts Co/Ce–La–Si, with varied La (H2, CH4, CO2, C2H2) and of the main liquid products (acetal-
and Ce contents (Fig. 6), it can be seen that Co10%La2%Ce48% dehyde, acetic acid, ether and acetone), it can be inferred that
From the distribution of the main gaseous products formed
-
Si50% gave the best yield for H2 of all the catalysts. This higher the predominant reactions are: the decomposition of ethanol
yield is related to smaller Co3O4 crystallite size (Table 2), greater (eqn (8)), followed by the WGSR (eqn (6)) and the dehydroge-
metal area and dispersion (Table 3), which means more active nation of ethanol to acetaldehyde (eqn (9)). However, the
sites available to react. Moreover, the formation of the solid temperature used did not favour the steam reforming of
solution (CexLa1ꢀxO2ꢀd),conrmed by XRD, increased the methane (reaction (10)), nor steam reforming of acetaldehyde
number of mobile O2 vacancies in the support, as shown by O2- (eqn (11)), resulting in a lower yield for H2 than predicted by the
TPO, which probably promoted the oxidation of CO to CO2 by overall stoichiometry of the reaction (eqn (1)).
the WGSR, contributing to H2 production.6
C2H5OH / CH4 + CO + H2
(8)
In assessing the related support La2%Ce48%Si50% (Table 4), it
was shown that this support also contribute to ethanol
conversion. This activity may be due to the ability of CeO2 to
C2H5OH / C2H4O + H2
(9)
Fig. 6 Yields of gaseous products during ESR at 500 ꢁC for 6 hours on the catalysts: (A) Co10%SiO2, (B) Co10%CeO2, (C) Co10%La2O3, (D)
Co10%La2%Ce48%Si50%, (E) Co10%La4%Ce46%Si50%, (F) Co10%La6%Ce44%Si50%
.
43846 | RSC Adv., 2014, 4, 43839–43849
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Fig. 7 XRD patterns of the catalysts (A) after ESR reaction: (a) Co10%SiO2, (b) Co10%La2O3, (c) Co10%CeO2, (d) Co10%La2%Ce48%Si50%, (e) Co10%
-
La4%Ce46%Si50%, (f) Co10%La6%Ce44%Si50% and (B): Co10%La2O3: (a) before and (b) after ESR.
due to the reaction of La2O3 with CO2 present in the reaction
products, forming La2O2CO3, which can react with some of the
surface carbon deposited during the ESR reaction and prevent
catalyst deactivation.28 Observing the diffractograms of catalysts
supported on La–Ce–Si, there was no formation of the species
La2O2CO3 possibly because the La2O3 was inside the lattice of
cerium oxide, forming the solid solution, and thus unavailable
to react with the CO2 produced in the reaction to form the
reactive species. Another possibility is that the low content of
La2O3 present in the catalysts could be insufficient to react with
the CO2 produced by the reaction.
CH4 + 2H2O / CO2 + 4H2
C2H4O + H2O / 2CO + 3H2
(10)
(11)
Fig. 7A shows the XRD patterns of the catalysts aer the ESR
reactions and Fig. 7B shows the catalyst Co10%La2O3 before and
aer the reaction, in order to show the formation of the
La2O2CO3 species, which can reduce carbon formation. Ana-
lysing the XRD patterns of catalysts aer 6 hours of ESR reaction
(Fig. 7A), a peak was observed at 2q ꢃ 26ꢁ which was attributed
to a deposit of graphite-type carbon,19,26,27 and the catalyst
supported on La2O3 produced peaks related to the lanthanum
oxycarbonate species, aer the ESR reaction. The low carbon
formation on Co10%La2O3 (Table 4), according to Sun et al.,28 is
Fig. 8 and 9 show the scanning electron micrographs (SEM)
of catalysts Co10%SiO2 and Co10%La2%Ce48%Si50% before and
aer 6 hours of ESR reaction. A change is noted in the
Fig. 8 Scanning electron micrographs (SEM) of Co10%SiO2 (A) before and (B) after ESR the reaction.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 43839–43849 | 43847

View Article Online
RSC Advances
Paper
Fig. 9 SEM of Co10%La2%Ce48%Si50% (A) before and (B) after the ESR reaction.
morphology of the catalysts with the addition of Ce and La
oxides on the silica. The images of the catalysts aer the 6 hours
of reaction conrm the formation of carbon laments. It is
known that the type of carbon formed is inuenced by the
particle size.29 The type of carbonaceous deposits formed may
decrease the performance of the catalyst in the steam reforming
reaction of ethanol. Carbon bre laments allow the metallic
cobalt to remain exposed to the reactants for a longer period of
time, causing the catalyst to remain active throughout the
reaction, which explains its good performance and non-
deactivation, despite the presence of carbon (Table 4).2
3 S. A. A. Haryanto, S. Fernando and N. Murali, Energy Fuels,
2005, 19, 2098–2106.
4 T. A. Maia, J. M. Assaf and E. M. Assaf, Chem. Mater., 2012,
132, 1029–1034.
5 A. J. Jacobson, Chem. Mater., 2010, 22, 660–674.
6 K. Lakshimi, G. Thrimurthulu, B. M. Reddy, M. Muhler and
W. Grunert, Catal. Sci. Technol., 2011, 9, 1645–1652.
7 P. Fleming, S. Ramirez, J. D. Holmes and M. A. Morris, Chem.
Phys. Lett., 2011, 509, 51–57.
8 B. M. Reddy, G. Thrimurthulu, L. Katta, Y. Yamada and
S. Park, J. Phys. Chem., 2009, 113, 15882–15890.
9 H. Song and U. S. Ozkan, J. Phys. Chem. A, 2010, 114, 3796–
3801.
Conclusions
10 B. M. Reddy, L. Katta and G. Thrimurthulu, Chem. Mater.,
2010, 22, 467–475.
The XRD patterns showed that a solid solution was formed
between CeO2 and La2O3 in the mixed catalysts and a crystalline 11 B. Ernst, L. Hilaire and A. Kiennemann, Catal. Today, 1999,
phase of Co3O4 on all catalysts. 50, 413–427.
The adding of basic oxides La and Ce to SiO2 led to active 12 F. C. F. Marcos, Study of cobalt catalysts supported on CeO2–
catalysts for H2 production and low carbon formation. The low
formation of carbon was due to oxygen vacancies of the support
that favoured the oxidation of carbon.
La2O3–SiO2 for the production of hydrogen from steam
reforming of ethanol, Master's dissertation, University of
˜
˜
Sao Paulo, Sao Carlos, 2012, http://www.teses.usp.br/teses/
disponiveis/75/75134/tde-10102012-142141/en.php.
The best catalyst for ethanol steam reforming and H2
production was Co10%La2%Ce48%Si50%. Its high performance 13 R. C. Reuel and C. H. Bartholomev, J. Catal., 1984, 85, 63–77.
was related to the oxygen vacancies, greater metal surface area 14 A. Bshish, Z. Yaakob, B. Narayanan, R. Ramakrishnan and
and low carbon formation.
A. Ebshish, Chem. Pap., 2011, 65, 251–266.
15 Q. Guo and Y. Liu, Appl. Catal., B, 2008, 82, 19–26.
16 R. Rao. and B. G. Mishra, Bull. Catal. Soc. India, 2000, 2, 122–
134.
Acknowledgements
The authors thank CNPq and Fapesp for nancial support, the 17 B. Zhang, D. Li and X. Wang, Catal. Today, 2010, 158, 348–
Electrochemistry Group at IQSC for XRD analyses and the LNLS
353.
(Campinas, Brazil) for in situ XRD analyses at XPD beamline.
18 M. Fleys, Y. Simon, D. Swierczynski, A. Kiennemann and
P. M. Marquaire, Energy Fuels, 2006, 20, 2321–2329.
´
19 A. Bueno-lopez, K. Krishna, M. Makkee and J. A. Moulijn, J.
References
Catal., 2005, 230, 237–248.
1 I. A. C. Ramos, T. Montini, B. Lorenzut, H. Troiani, 20 H. Wang, J. L. Ye, Y. Liu, Y. D. Li and Y. N. Qin, Catal. Today,
F. C. Gennari, M. Graziani and P. Fornasiero, Catal. Today,
2010, 10, 96–104.
2 Y. N. Q. H. Whang, Y. Liu and L. Wang, Chem. Eng. J., 2008,
145, 25–31.
2007, 129, 305–312.
21 T. F. De Silva, A. C. Dias, C. G. Marciel and J. M. Assaf, Catal.
Sci. Technol., 2012, 3, 635–643.
43848 | RSC Adv., 2014, 4, 43839–43849
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
22 G. A. C. S. Bernal, G. Blanco and M. A. Caurqui, Catal. Lett., 26 K. Y. Koo, H. Roh, Y. T. Seo, D. J. Seo, W. L. Yoon and
1998, 53, 51–57. S. B. Park, Appl. Catal., A, 2008, 340, 183–190.
23 W. S. F. Wang, W. Cai, H. Provendier, Y. Schuurman, 27 S. Therdthianwong, C. Siangchin and A. Therdthianwong,
C. Descorne and C. Mirrodatos, Int. J. Hydrogen Energy,
2011, 36, 13566–13574.
24 A. M. E. K. C. Petallidou, Appl. Catal., B, 2013, 140, 333–347.
Fuel Process. Technol., 2008, 89, 160–168.
28 J. Sun, X. P. Qiu, F. Wu and W. T. Zhu, Int. J. Hydrogen Energy,
2005, 30, 437–445.
25 W. G. A. Machocki, A. Denis and W. Grzegorczyk, Appl. Surf. 29 W. J. Lee and C. Z. Li, Carbon, 2008, 46, 1208–1217.
Sci., 2010, 256, 5551–5558.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 43839–43849 | 43849