Influence of molybdenum on ceria activity and CO2 selectivity in propene total oxidation

Article abstract of DOI:10.1023/B:KICA.0000023795.58615.b2

A total oxidation of propene into CO₂ is obtained on pure ceria at 673 K. However, in the presence of molybdenum, propene can be partially oxidized at room temperature. The Electron Paramagnetic Resonance (EPR) indicates changes in the oxidation state of molybdenum occurring upon interaction with propene. It has been found that the concentration of Mo(V) influences the propene conversion. The interaction between propene and molybdenum leads to the formation of surface species that, depending on the strength of their bonding to the surface, can be decomposed to ethene or coke. These results have been confirmed by infrared (FTIR) study. The oxidation reaction of propene is in competition with that of coke or ethene deposit on the catalyst surface, which can explain the decrease of the catalyst activity and selectivity in the presence of high molybdenum loadings.

Full text of DOI:10.1023/B:KICA.0000023795.58615.b2

Kinetics and Catalysis, Vol. 45, No. 2, 2004, pp. 219–226. From Kinetika i Kataliz, Vol. 45, No. 2, 2004, pp. 237–244.
Original English Text Copyright © 2004 by Flouty, Abi-Aad, Siffert, Aboukaïs.
CATALYTIC REACTION
MECHANISMS
Influence of Molybdenum on Ceria Activity and CO2 Selectivity
in Propene Total Oxidation1
R. Flouty, E. Abi-Aad, S. Siffert, and A. Aboukaïs
Laboratoire de Catalyse et Environnement E.A. 2598, Université du Littoral-Côte d’Opale, MREID,
145 avenue Maurice Schumann, 59140 Dunkerque, France
e-mail: siffert@univ-littoral.fr
Received December 23, 2002
Abstract—A total oxidation of propene into CO2 is obtained on pure ceria at 673 K. However, in the presence
of molybdenum, propene can be partially oxidized at room temperature. The Electron Paramagnetic Resonance
(EPR) indicates changes in the oxidation state of molybdenum occurring upon interaction with propene. It has
been found that the concentration of Mo(V) influences the propene conversion. The interaction between pro-
pene and molybdenum leads to the formation of surface species that, depending on the strength of their bonding
to the surface, can be decomposed to ethene or coke. These results have been confirmed by infrared (FTIR)
study. The oxidation reaction of propene is in competition with that of coke or ethene deposit on the catalyst
surface, which can explain the decrease of the catalyst activity and selectivity in the presence of high molybde-
num loadings.
1
1. INTRODUCTION
the formation of σ-allylic complex bonded to the
molybdenyl oxygen of MoO3. Otherwise, Carrazan
et al. [12] by means of their IR studies, mentioned that
the adsorption of propene on molybdenum based cata-
lysts gives rise to olefin/surface cation π-complex spe-
cies with the formation of unsaturated aldehydes. This
indicates the large selectivity of these species towards
acrolein formation. It was proposed, that the first step of
the reaction consists of the abstraction of α-hydrogen
and the formation of π-allylic species attached to the
metal centers. The next step is the insertion of lattice
oxygen into these species [12–15]. It has been shown
that molybdenum oxide contains active centers, which
are able to perform the insertion of lattice oxygen into
the propene molecules [13–15]. Anderson et al. [16]
suggested that activation of propene molecules occured
at the unsaturated Mo(VI) sites.
In the light of these results, concerning the nature of
the interaction between propene and molybdenum, it
seems of interest to investigate the catalytic oxidation
and the adsorption of propene on molybdenum cata-
lysts supported on ceria. Electron Paramagnetic Reso-
nance (EPR) has been used for this purpose. This tech-
nique being very sensitive to the electronic structure
and the symmetry of the surroundings of paramagnetic
species, it is most suitable to investigate the centers dur-
ing adsorption accompanied by electron transfer.
Catalysts based on pure or supported molybdenum
MoO3 are frequently used for a great number of differ-
ent reactions, especially of the redox type, such as oxi-
dation of hydrocarbons [1]. The condition for these
reactions to occur is the activation of oxygen or hydro-
carbon molecules [2]. It is known that if the reaction is
ignited by the activated oxygen species, the organic
molecule usually undergoes degradation, which leads
to the total oxidation of hydrocarbons. If the oxidation
of hydrocarbons starts with the activation of organic
molecules, followed by a nucleophilic addition of the
catalyst oxide ions, the formation of oxygenated com-
pounds occurs. The subsequent reoxidation of the cata-
lyst by gas phase oxygen completes the catalytic cycle.
Partial oxidation is preferred for systems with a low
thermodynamic stability of lattice oxygen, as MoO3
oxide [3]. For this reason, molybdenum based catalysts
are used in partial oxidation of propene [3, 4]. However,
the redox couple Ce4+/Ce3+ with the ability of ceria to
shift between CeO2 and Ce2O3, under oxidizing and
reducing conditions, respectively [5] and mainly the
ease of formation of labile oxygen vacancies [6, 7],
make ceria a promising material for use either as a sup-
port or active catalyst in total oxidation reactions of
hydrocarbons [8, 9]. A number of studies aimed to elu-
cidate the propene adsorption on pure molybdenum
oxide and molybdenum supported catalysts [10–16].
Davydov et al. [10, 11], on the ground of their IR and
EPR studies, argued that interaction of MoO3/Al2O3,
MoO3/MgO and gallium molybdate with propene led to
2. EXPERIMENTAL
2.1. Solid Synthesis
Cerium hydroxide Ce(OH)4 was prepared by pre-
cipitation from cerium (III) nitrate hexahydrated
1
This article was submitted by the authors in English.
0023-1584/04/4502-0219 © 2004 MAIK “Nauka /Interperiodica”

220
FLOUTY et al.
(3 × 10–5 mbar) for 1 h at 573 K and subjected to pro-
%
100
90
80
70
60
50
40
30
20
pene adsorption at different pressures for 30 min in the
temperature range of 298–573 K. The temperature was
raised with 2 K min–1 from 298 K to the desired temper-
ature and was maintained for 30 min. The EPR spectra
were recorded at liquid nitrogen temperature 77 K on a
EMX BRUKER spectrometer with a cavity operating at
a frequency of ~9.5 GHz (X band). The magnetic field
was modulated at 100 kHz. The g values were deter-
mined from precise frequency and magnetic field val-
ues. The relative concentration of paramagnetic sites Iint
was determined as normalized double integrals of EPR
spectra using Bruker’s WINEPR program.
– Conversion
– CO selectivity
2
10
0
CeO2
1Mo100Ce
2Mo10Ce
3Mo10Ce
1Mo1000Ce
1Mo10Ce
2.4. FTIR Analysis
Fig. 1. Propene conversion and CO selectivity of molybde-
The FTIR studies, were carried out using Perkin-
Elmer FTIR spectrum 2000. 20 mg of each catalyst was
pressed into self-supported wafers under a pressure of
8000 kg cm–2, and then placed with a sample holder
inside a Pyrex cell with NaCl windows, which allowed
pre-treatment under vacuum, introduction of reactants
and recording of the spectra. After pre-treatment under
vacuum (3 × 10–5 mbar) of 3 Mo10Ce catalyst at 573 K
for 1 h, 500 mbar of propene gas was introduced into
the cell which was cooled in nitrogen liquid for 30 min.
The spectra were recorded at room temperature 10 min
after the cell was taken out of the nitrogen liquid.
2
num-cerium catalysts in the oxidation of propene at 673 K.
Ce(NO3) · 6H2O 1 M with an alkali solution NaOH 2 M.
The resulting hydroxide Ce(OH)4 was filtered, washed,
and dried for about 20 h at 373 K. Ceria is obtained after
calcination of this latter at 773 K for 4 h under a flow of air
(75 mL min–1) with a temperature rate of 0.5 K min–1.
Molybdenum-cerium
catalysts
1Mo1000Ce,
1Mo100Ce, 1Mo10Ce, 2Mo10Ce, and 3Mo10Ce have
been prepared by impregnation of different solutions of
ammonium heptamolybdate Mo7O24(NH4)6 · 4H2O at
Ce(OH)4 support (S = 138 m2 g–1; Vporous = 0.44 mL g–1).
These solids have respectively the following atomic
ratios (Mo/Ce = 0.001, 0.01, 0.1, 0.2, 0.3). After drying
at 373 K for about 20 h, solids were stabilized by calci-
nation under dried air (4 L h–1) at 773 K for 4 h. The
temperature was raised at a rate of 0.5 K min–1.
3. RESULTS AND DISCUSSIONS
3.1. Catalytic Oxidation of Propene
Figure 1 shows the catalytic activity and the selec-
tivity in CO2 of molybdenum-cerium catalysts towards
the oxidation reaction of propene at 673 K (maximum
value reached in the exhaust steam). The propene con-
version ratio is about 62% for pure ceria catalyst, with
a high selectivity in CO2 (CO2/(CO + CO2) = 97%). Sim-
ilar activity and selectivity in CO2 were obtained when
the reaction is performed in the presence of catalysts
with small amounts of molybdenum (1Mo1000Ce,
1Mo100Ce). Whereas, for catalysts with higher molyb-
denum loadings, a lower activity is obtained with the
formation of a large quantity of CO during the propene
oxidation. Only 30% of propene is decomposed in the
presence of 3Mo10Ce with the formation of 63% of CO
(CO/(CO + CO2) = 63%). Thus, the presence of molyb-
denum decreases the catalytic activity and lowers its
selectivity in CO2 with respect to the propene oxidation
reaction.
2.2. Propene Oxidation
The propene (C3H6) oxidation was realized in a cat-
alytic microreactor coupled to a Varian 3600 gas chro-
matograph using a double detection FID and TCD. The
catalysts were reactivated at 723 K for 1 h under a flow
of dried air 4 L h–1 × 100 mg of these catalysts were
tested in the presence of a mixture (2.27% propene-
97.73% air). The propene conversion (%) at 673 K is
given by the following relation: [(CO + CO2)673 K/(CO +
CO2 + 3C3H6)673 K] × 100. The selectivity of the cata-
lysts towards the formation of CO2 was obtained from
the ratio [CO2/(CO + CO2)].
2.3. EPR Analysis
It is well-known that during oxidation reaction of
The EPR measurements were carried out on molyb- hydrocarbons, oxide catalysts are subjected to a reduc-
denum-cerium catalysts after treatment under a flow of tion after oxidation of hydrocarbon molecules. The
propene (2.27 ml min–1) from room temperature to more the catalyst is reduced, the more the hydrocarbon
373 K. Before treatment, catalysts have been activated molecules are oxidized. Otherwise, it is well-known
at 723 K for 1 h. EPR studies were carried out also that ceria is easily reduced at elevated temperatures in
on 3Mo10Ce catalyst initially evacuated in a vacuum non-stoichiometric oxides CeO2 – x (0 < x < 0.5) rich in
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

INFLUENCE OF MOLYBDENUM ON CERIA ACTIVITY AND CO2 SELECTIVITY
221
Gain
Iint, arb. units
g = 1.964
1.6
4.41
5
CeO2
*
*
*
*
*
*
1.4
1.2
1.0
0.8
0.6
0.4
0.2
g||(B)
g||(A)
1Mo1000Ce
g (1)
1Mo100Ce
1Mo10Ce
2Mo10Ce
5
g (2)
g||(1)
5
2
0
CeO2
1Mo100Ce
2Mo10Ce
3Mo10Ce
g||(2)
g = 2.022
g|| = 2.018
1Mo1000Ce
1Mo10Ce
3+
Fig. 3. Variation of superposed (Ce ) signal intensity with
the increase of molybdenum loading.
g (3)
g (4)
3Mo10Ce
3.34
g||(3)
g||(4)
tra recorded at 77 K of molybdenum-cerium catalysts
treated under vacuum (3 × 10–5 mbar) at 573 K for 1 h.
The EPR spectra recorded at 77 K for pure ceria,
1Mo1000Ce and 1Mo100Ce catalysts, present a super-
position of two signals (A) and (B) having the same
axial symmetry. These signals can be attributed to the
presence of two different sites of Ce3+ in the solid [elec-
tron configuration f1, ge > g > g||], or more exactly to the
presence of electrons trapped in the free orbital (4f) of
Ce4+ ions [5–7]. Moreover, according to the literature
[5–7], signal (A) characterized by g = 1.964; g|| =
1.939 and giso = 1.955 is due to Ce3+ ions in the bulk of
the solid, stabilized by some lattice defects and oxygen
vacancies favorable for oxidation reactions [18].
3000 3100 3200 3300 3400 3500 3600 3700 3800 3900
H, G
Fig. 2. EPR spectra, recorded at 77 K, of molybdenum–
cerium catalysts treated under vacuum at 573 K for 1 h.
oxygen vacancies [5–7]. The latter are responsible for
oxygen mobility on ceria surface, which is favorable for
the oxidation reactions [9]. It might be possible then,
that the reduction of ceria is more difficult in the pres-
ence of molybdenum, which can decrease the number
of oxygen vacancies on the surface and consequently its
activity with respect to propene oxidation. Otherwise,
the catalytic activity is related to the affinity of the
oxide towards the reactant [17]. The nature of the inter-
mediate products reversibly or irreversibly adsorbed on
the catalysts surface, plays a crucial role in the behavior
and the performance of the catalysts during the oxida-
tion reaction [17]. Thus, the oxidation ability of the
samples, has been checked by the redox behavior of the
catalysts under reduced atmospheres and through the
study of propene adsorption on the surface. The Elec-
tron Paramagnetic Resonance has been used for these
studies for further understanding of the catalytic
results.
Whereas, signal (B) (g = 1.964; g|| = 1.946; giso
=
1.958) corresponds to Ce3+ ions with easily removable
ligands. As shown in Fig. 3, the intensity of these sig-
nals, calculated by double integration (DI) of the EPR
signals, decreases progressively with the increase of
molybdenum loading in the solid. It can be deduced
then, that the presence of Ce4+/Ce3+ redox couple which
is favorable for the oxidation reaction also decreases in
the presence of molybdenum. Thus, the presence of
molybdenum implies a decrease of the number of
anionic vacancies present on the surface. This can be
related to the low activity and selectivity in CO2 of the
catalysts containing higher quantity of molybdenum
with respect to propene oxidation reaction. Otherwise
an axial signal characterized by g
1.839 and giso = 1.879 is observed for 1Mo100Ce cata-
lyst (Fig. 2). In addition to this signal, another signal
with axial symmetry (g (2) = 1.915; g||(2) = 1.889; giso
= 1.899; g||(1) =
3.2. EPR Study of Molybdenum-Cerium Catalysts
(1)
The treatment under vacuum of the catalyst at high
temperatures can induce a partial reduction of the solid
comparable to that occurred during the combustion
cycle. The EPR technique has been used to study the
catalysts after thermal treatment under vacuum and
after interaction with propene at room temperature.
=
1.906) appeared in the EPR spectra of 1Mo10Ce,
2Mo10Ce, and 3Mo10Ce catalysts (Fig. 2). These sig-
nals are characteristic of the d1 configuration and can be
attributed to the presence of two different sites of
3.2.1. EPR study after thermal treatment under molybdenum (V) [Mo5+(1) and Mo5+(2)] stabilized
vacuum of the catalysts. Figure 2 shows the EPR spec- with oxygen ligands (g > g||) [19]. In fact, similar spec-
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

222
FLOUTY et al.
to hexacoordinated Mo5+ ions typical of MoO6 octahe-
dral in MoO3 lattice [24, 25]. Additionally, the value of
a ratio relative to these species is equal to 1.23
[Mo5+(2)] and 1.15 [Mo5+(4)]. This further indicates the
location of these species in octahedral sites with a very
weak tetragonal distortion. The presence of reduced
Mo5+ ions confirms that during treatment under vac-
uum, the catalysts can be partially reduced. It can be
deduced that, the Mo6+/Mo5+ redox couple is less active
with respect to propene oxidation than the Ce4+/Ce3+
redox couple, which can affect the catalytic activity of
the oxide. In addition to molybdenum V signals, an
axial signal with a very high intensity characterized by
g = 2.022; g|| = 2.010 and giso = 2.018 is observed for
2Mo10Ce and 3Mo10Ce treated under vacuum at
573 K (Fig. 2). This signal is characteristic of paramag-
netic oxygen species in interaction with the solid. A
similar signal has been observed in the literature and
attributed to O– species adsorbed on magnesium oxide
partially reduced [26, 27]. Such O– species are in inter-
action with molybdenum oxide, since this signal is not
observed for pure ceria. Furthermore, this signal char-
acteristic of O– species was not observed at room tem-
perature which can be related to spin-lattice relaxation
time that is too short at room temperature. Thus, it can
be deduced that a modification of oxygen lattice
arrangement can occur in the presence of molybdenum.
This modification can play an important role in the vari-
ation of the selectivity and the activity of catalytic sys-
tem containing a high amount of molybdenum, during
propene oxidation. The different signals observed after
thermal treatment of molybdenum-cerium catalysts are
reported in table.
EPR signals recorded at 77 K for molybdenum-cerium cata-
lysts after treatment under vacuum at 573 K
Signal
g
g||
giso
α
Ce3+(A)
Ce3+(B)
Mo5+(1)
Mo5+(2)
Mo5+(3)
Mo5+(4)
O– species
1.964
1.964
1.899
1.915
1.943
1.862
2.022
1.939
1.946
1.839
1.889
1.929
1.841
2.018
1.955
1.957
1.879
1.906
1.938
1.855
2.020
–
–
1.58
1.29
1.23
1.15
–
tra have been observed for these species (Mo5+) in dif-
ferent matrix [20–22]. According to the literature, the
Mo5+(1) ions can be attributed to the reduction of Mo6+
species dispersed on ceria surface. However, Mo5+(2)
ions observed in the presence of high molybdenum
loadings are due to the reduction of octahedral Mo6+
species polymerized on ceria. Moreover, the EPR spec-
tra of 1Mo10Ce, 2Mo10Ce, and 3Mo10Ce recorded at
room temperature have shown the presence of only the
signal characteristic of Mo5+(2) ions (g
= 1.915;
(2)
g
||(2) = 1.889 and giso = 1.906). Thus, these ions Mo5+(2)
are more probably located in octahedral sites with a
weak tetrahedral distortion. In fact, Shelimou et al.
[21], have attributed the absence of this EPR signal at
room temperature to Mo5+ ions with tetrahedral sym-
metry which is due to the fact that the spin-lattice relax-
ation is too short at room temperature. However, Mar-
tini [22] confirms that Mo5+ ions in tetrahedral sites
cannot be detected by EPR even at 77 K. In addition,
Che et al. [20] have shown that strong distortions of
Mo5+ ions can be observed in EPR spectra. Thus, it is
very probable that Mo5+(1) present a strong tetrahedral
distortion. Whereas Mo5+(2) ions are located in weakly
distorted octahedral sites. Nevertheless, for d1 ions
located in octahedral sites with tetragonal deformation,
the value of the g tensor corresponds to the α ratio [α =
(ge – g||)/(ge – g ) = 4 δ/∆]. This α ratio characterizes the
amplitude of the tetragonal distortion (δ) relatively to
that of the octahedral axial field Oh(∆) [23]. α is
respectively equal to 1.58 and 1.29 for Mo5+(1) and
3.2.2. EPR study of the catalysts after interaction
with propene. The molybdenum-cerium catalysts have
been studied in EPR after treatment under a flow of pro-
pene (2.27 ml min–1) from room temperature to 373 K.
The rate of temperature is 5 K min–1. As for propene
catalytic test, the catalysts were preliminarily reacti-
vated for 1 h at 773 K under a flow of dried air
(30 ml min–1). No change has been observed for pure ceria,
1Mo1000Ce, 1Mo100Ce, and 1Mo10Ce catalysts after
treatment under propene flow. However, for 2Mo10Ce
and particularly 3Mo10Ce, a new signal noted (S) with
isotropic symmetry (giso = 2.0034) and a very weak
intensity has been detected at room temperature. This
signal is associated with reduced molybdenum Mo5+
signals. The presence of this signal is due to propene
interaction with molybdenum oxide since it was not
Mo5+(2) ions. This indicates also that Mo5+(1) ions are observed for pure ceria. In order to confirm that, differ-
strongly distorted, whereas, Mo5+(2) ions are octahedral ent pressures of propene have been introduced on
with a weak tetrahedral distortion. At least two other 3Mo10Ce for 30 min at room temperature after being
axial signals characterized respectively with g
=
treated under vacuum at 573 K for 1 h. Vacuum treat-
(3)
1.943; g||(3) = 1.929 and giso = 1.938 and g (4) = 1.862; ment of the catalyst can induce more active vacancies
g||(4) = 1.841 and giso = 1.855 can be identified for the on the solid surface favorable for more interaction of
catalysts 2Mo10Ce and 3Mo10Ce. Similar spectra have propene with the solid. The resulting EPR spectra
been observed in the literature and have been attributed recorded at room temperature (Fig. 4a) and 77 K
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

INFLUENCE OF MOLYBDENUM ON CERIA ACTIVITY AND CO2 SELECTIVITY
223
(a)
Signal (S)
g = 2.0034
(b)
Signal (S)
g = 2.0034
Gain
Gain
15 mbar
15 mbar
30 mbar
1
0.5
0.5
30 mbar
90 mbar
1
0.5
90 mbar
0.5
0.5
0.5
120 mbar
210 mbar
0.5
0.5
120 mbar
210 mbar
0.5
0.5
330 mbar
450 mbar
0.34
0.34
330 mbar
450 mbar
3000
3200
3400 3600
H, G
3800
4000 3000 3200 3400 3600 3800 4000
H, G
Fig. 4. EPR spectra of 3Mo10Ce after adsorption of different pressures of propene at room temperature, (a) recorded at 298 K and
(b) at 77 K.
(Fig. 4b) for 3Mo10Ce catalyst, show the presence of related to this signal, experiments on the reduction of
this signal with a very high intensity for a low propene pure molybdenum oxide (MoO3) with hydrogen or pro-
pressure (15 mbar). It is important to notice that the pene were performed. This signal (S) appeared only
very low intensity observed for signal (S) after treat- during interaction of MoO3 with propene and disap-
ment of 3Mo10Ce under a flow of propene is due to the peared after prolonged oxidation at a higher tempera-
fact that the catalyst was activated under air before ture; it was not observed on interaction with hydrogen
treatment under propene and the surface is not com- at the same conditions. Such behavior indicates that the
pletely free of water molecules and other residual spe- signal (S) is more probably due to paramagnetic hydro-
cies that can delay the interaction of propene with carbon species attached to molybdenum centers of the
molybdenum oxide. In addition, an increase of signal catalysts surface. In fact, Carrazan et al. [12] have evi-
(S) intensity has been observed for 3Mo10Ce with the denced a partial oxidation of propene to acrolein after
increase of propene pressure (Fig. 5). This variation its adsorption on the surface of molybdenum supported
indicates that the signal (S) is produced after interaction catalysts at room temperature. Moreover, a lot of stud-
of propene with molybdenum oxide. Furthermore, the ies were aimed at elucidating the mechanism of the
isotropic symmetry and the g value permit one to redox processes occurring during selective oxidation of
attribute this signal (S) either to trapped electron in the propene to acrolein using molybdenum based catalysts
lattice MoO3 or to organic unsatured hydrocarbon spe- [1–4, 10–16]. However, no acrolein was found in the
cies or coke. In order to explain the nature of the species products for our catalytic test. The single electron is
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

224
Iint, rel. units
FLOUTY et al.
pene is recorded at 77 K, the intensity of signal (S)
increases slightly and does not follow the Curie law,
while the intensity of the reduced molybdenum signal
Mo5+ increases considerably, particularly that of
Mo5+(3) species (Fig. 6). Such behavior indicates, that
at a low temperature, the single electron is more polar-
ized and shifted towards molybdenum (VI) centers,
which are further reduced to Mo5+ ions. Based on these
results, it can be deduced then, mat after adsorption,
propene molecules are activated on the surface of
molybdenum oxides leading to the formation of surface
species which, depending on the strength of its bonding
to the surface, may be decomposed to ethene or coke.
Nevertheless, it seems important to notice that the axial
signal (g = 2.022, g|| = 2.010, giso = 2.018) characteris-
tic of O– oxygen species, which has been observed after
treatment of 3Mo10Ce at 573 K (Fig. 2), disappears
completely in the presence of propene (Fig. 6). This can
be related to the modification of lattice and surface oxy-
gen arrangement of 3Mo10Ce. Moreover, the EPR
study of 3Mo10Ce catalyst after interaction with pro-
pene at higher temperatures, have shown a modification
of signal (S) intensity with temperature. A higher inten-
sity of this signal is observed at 373 K (Fig. 7). The
more the temperature increases, the easier the activa-
tion of propene is. When the temperature reaches
473 K, the intensity of this signal decreases consider-
ably, which implies a further oxidation of propene. Oth-
erwise, the intensity of the signal (S) decreases in the
presence of oxygen (Fig. 7). In fact, this signal was not
observed at room temperature when the 3Mo10Ce cat-
alyst is treated under a mixture of propene (450 mbar)
and oxygen (50 mbar). Thus, the presence of oxygen
enhances the decomposition of propene to oxygenated
species at room temperature. Finally, this signal is due
to propene interaction with molybdenum oxide and
could be attributed to the presence of organic unsatured
hydrocarbon species or coke.
6
5
4
3
2
1
0
50 100 150 200 250 300 350 400 450
Propene pressure, mbar
Fig. 5. Variation of signal (S) intensity for 3Mo10Ce cata-
lyst with the increase of propene pressure.
Signal (S)
g = 2.0034
(a)
Mo5+(3)
Mo5+(2)
Mo5+(1)
(b)
Mo5+(4)
3000
3200
3400
3600
3800
4000
H, G
Fig. 6. EPR spectra of 3Mol0Ce catalyst treated under
450 mbar of propene recorded at (a) 298 and (b) 77 K. DI/N:
double integral/normalized.
In order to confirm these results, the structures
formed during adsorption of propene on 3Mo10Ce cat-
alyst were studied by means of IR spectroscopy.
Iint, rel. units
35
30
25
450 mbar of propene
450 mbar of propene
3.3. IR Study of 3Mo10Ce Catalyst
after Interaction with Propene
20
15
10
5
and 50 mbar O
2
The IR spectra, recorded at room temperature, of
3Mo10Ce catalyst treated under vacuum at 573 K after
adsorption and desorption of propene (500 mbar at
77 K) are presented in Fig. 8. Propene adsorption on
3Mo10Ce catalyst produces two series of bands in the
3100 to 2700 cm–1 and in the 1800–1200 cm–1 regions
(Fig. 8, spectrum 2). The presence of IR band at 3092 cm–1
indicates that a certain portion of propene molecules
are adsorbed on the surface without breaking the dou-
ble bond. This band is characteristic of ν=CH stretching
vibration of physisorbed propene [10, 12]. The bands
0
298
373
473 T, K
Fig. 7. Variation of signal (S) intensity with the increase of
the temperature for the catalyst 3Mo10Ce.
polarized and in interaction with Mo6+ centers, since no
hyperfine structure was observed for the EPR signal (S)
recorded at room temperature. Moreover, when the
spectrum of 3Mo10Ce treated under 450 mbar of pro- observed at 2972, 2932, and 2896 cm–1 can be assigned
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

INFLUENCE OF MOLYBDENUM ON CERIA ACTIVITY AND CO2 SELECTIVITY
225
lytic system in the presence of molybdenum. Moreover,
the oxidation reaction of propene is in competition with
that of coke or ethene deposit on the catalyst surface,
which can explain the decrease of the catalyst activity
and selectivity in the presence of high molybdenum
loadings.
1
1624
2960
3080
2328
1488
1408
2
2968
2908
2856
1444
1368
1252
1452
1384
1816
4. CONCLUSION
3
This work has shown the relation between the cata-
lytic properties of the catalysts with respect to the pro-
pene oxidation reaction both with its redox behavior
and its interaction with propene molecules. The
decrease of the catalyst activity in the presence of
molybdenum can be explained by the decrease of the
number of anionic vacancies on the surface, which can
be due to the reduction of ceria that might be difficult in
the presence of molybdenum. Otherwise, the strong
interaction of propene with molybdenum and its partial
oxidation at room temperature can affect the catalysts
activity and selectivity during the oxidation of propene.
Thus, a competition of partial oxidation of propene on
MoO3 can be produced with a complete oxidation of
propene on ceria, during the catalytic oxidation of pro-
pene in the presence of catalysts with high molybde-
num loadings.
3200
2700
2200
1700
1200
Wave numbers, cm–1
Fig. 8. IR spectra, recorded at room temperature, of
3Mo10Ce catalyst (1) treated under vacuum for 1 h at 573 K
(2) after adsorption of propene (500 mbar at 77 K) and (3)
after propene desorption at room temperature.
to stretching and deformation vibration of C−H bond in
CH, CH2, or CH3 groups. In fact, it has been shown by
spectral analysis of individual propene complexes [28]
that it is difficult to elucidate differences between vari-
ous complexes via ν=CH stretching and deformation
vibrations because of the overlap of the frequency
ranges. According to the literature [10, 28], the band
observed at about 1625 cm–1 should be attributed to
stretching vibration of C=C bond. However, the band
observed at about 1624 cm–1 for the untreated sample
with propene (Fig. 8, spectrum 1) should be due to phy-
sisorbed H2O and the treatment with propene should
remove with water. The frequencies at about 1400 cm–1
are characteristic of C–H deformation vibrations. The
other bands observed at 1828 and 1264 cm–1 are due to
the vibrations of C–O bond in CO2 and CO3 carbonate
species. In addition, after propene desorption the bands
due to propene chemisorbed on the surface remain sta-
ble, whereas that of physisorbed molecules of propene
disappears (3092 cm–1) (Fig. 8, spectrum 3). Thus, the
IR results seem to evidence the formation of organic
unsatured hydrocarbon species or coke during propene
chemisorption with 3Mo10Ce catalyst.
REFERENCES
1. Haber, J., Molybdenum: An Outline of Its Chemistry and
Uses, Brathwaite, E.R. and Haber, J., Eds., Amsterdam:
Elsevier, 1994, p. 477.
2. Haber, J., Solid State Chemistry in Catalysis, Grasselli, R.K.,
and Brazdil, J.F., Eds., Washington, DC: ACS, 1985, p. 3.
3. Aso, I., Nakao, M., Yamazoe, N., et al., J. Catal., 1979,
vol. 57, p. 287.
4. Haber, J., Proc. 8th Int. Congr. Catal., Berlin: Dechema
Chemie Verlag, 1984, vol. 1, p. 85.
5. Abiaad, E., Bennani, A., Bonnelle, J.P., et al., J. Chem.
Soc., Faraday Trans. I, 1995, vol. 91, p. 99.
6. Aboukaïs, A., Bennani, A., Aïssi, C.F., et al., J. Chem.
Soc., Faraday Trans., 1992, vol. 88, p. 615.
7. Abiaad, E., Bechara, R., Grimblot, J., et al., Chem.
Mater., 1993, vol. 5, p. 793.
8. Zwinkels, M.F.M., Jaras, S.G., and Menon, P.G., Catal.
Rev. Sci. Eng., 1993, vol. 35, p. 319.
According to these observations and taking into
account that the catalytic oxidation of propene was
realized in the presence of oxygen (2.27% of propene
in air) at 673 K, it can be deduced then, that at the
beginning of the catalytic test and at low temperatures
(<373 K), a partial oxidation of propene could occurred
on molybdenum oxide MoO3 for 1Mo10Ce, 2Mo10Ce,
and 3Mo10Ce catalysts. This reaction leads to the for-
mation of surface species that, depending on the
strength of its bonding to the surface, can be decom-
posed to ethene or coke [12]. This process is easier in
the presence of oxygen and at higher temperatures. It
can induce a modification of catalyst surface structure
and a modification of oxygen arrangement in the solid
that can affect the activity and selectivity of the cata-
9. Faber, J., Seitz, M., and Mueller, M., J. Phys. Chem. Solids.,
1976, vol. 37, p. 909.
10. Davydov, A.A., Mikhaltchenko, V.G., Sokolovski, V.D.,
et al., J. Catal., 1978, vol. 55, p. 299.
11. Maksimov, N.G., Ismailov, E.G., Anufenko, W.F., et al.,
Teor. Eskp. Khim., 1975, vol. 11, p. 260.
12. Carrazan, S.R.G., Martin, C., Rives, V., et al., Spectro-
chim. Acta A., 1996, vol. 52, p. 1107.
13. Adams, C.R. and Jennings, T.J., J. Catal., 1963, vol. 2,
p. 63.
14. Adams, C.R. and Jennings, T.J., J. Catal., 1964, vol. 3,
p. 549.
KINETICS AND CATALYSIS Vol. 45 No. 2 2004

226
FLOUTY et al.
15. Grabowski, R., Sloczynski, J., Dyreck, K., et al., App. 21. Shelimou, B.N., Pershin, A.N., and Kazanski, V.B.,
Catal., 1987, vol. 32, p. 103.
J. Catal., 1980, vol. 64, p. 426.
22. Martini, J., J. Magn. Reson., 1974, vol. 15, p. 262.
16. Grybowska, B., Mazurkieewicz, A., and Stoczynki, J.,
23. Dufaux, M., Che, M., and Naccache, C., J. Chem. Phys.,
Appl. Catal., 1985, vol. 13, p. 223.
1970, p. 527.
17. Spivey, J.J., Complete Oxidation of Volatile Organics:
Catalysis, Anderson, J.R. and Boudart, M., Eds., Cam-
bridge: Royal Society of Chemistry, 1989, vol. 8, p. 158.
24. Labanowska, M., Phys. Chem. Chem. Phys., 2000,
vol. 2, p. 3307.
25. Dyrek, K. and Labanowska, M., J. Chem. Soc., Faraday
Trans. 1, 1991, vol. 87, p. 100325.
26. Giamello, E., Theobald, F., Naccache, C., et al., J. Chem.
Phys., 1978, vol. 75, p. 305.
18. Fierro, J.L.G., Soria, J., Sanz, J., et al., Solid State
Chem., 1987, vol. 66, p. 154.
19. Courcot, D., Abi-Aad, E., Capelle, S., et al., Stud. Surf.
Sci. Catal., 1998, vol. 116, p. 625.
27. Smith, D.R. and Tench, A.J., Chem. Comm., 1968, p. 1113.
20. Che, M., Louis, C., and Sojka, Z., J. Chem. Soc. Faraday 28. Fisher, E. andVerner, G., Complexes of Metals, Moscow:
Trans., 1989, vol. 85, p. 3939.
Mir, 1968.
KINETICS AND CATALYSIS Vol. 45 No. 2 2004