Bismuth molybdates prepared by solution combustion synthesis for the partial oxidation of propene

Article abstract of DOI:10.1016/j.cattod.2015.03.045

The solution combustion method (SCS) is demonstrated as an easy and fast alternative method allowing the synthesis of mixed oxides with thermally sensitive metals, still keeping a good control on their stoichiometry and phase composition. Bismuth molybdates with different theoretical Bi/Mo atomic ratios, namely α-Bi₂Mo₃O₁₂, β-Bi₂Mo₂O₉, and γ-Bi₂MoO₆ were synthesized by the SCS, fully characterized by XRD, Raman, SEM/EDX, BET and XPS analyses, and tested in the partial oxidation of propene to acrolein. The SCS method allowed obtaining crystalline catalysts with Bi/Mo atomic ratios close to the theoretical values and very good catalytic properties namely high propene conversion (from 13.4 to 23.4%) and acrolein selectivity (from 71.5 to 80.0%) at 425°C. A high purity of the SCS prepared bismuth molybdates was obtained by means of a subsequent calcination treatment. Such treatment did not alter the high catalytic activity of the catalysts, which slightly increased (from 12.3 to 25.5%), but induced a marked loss of acrolein selectivity (from 66.2 to 48.4%) at 425°C. This effect is due to a strong increase of the oxidation states of Mo and Bi and reduction of the specific surface area during the calcination.

Full text of DOI:10.1016/j.cattod.2015.03.045

G Model
CATTOD-9572; No. of Pages7
ARTICLE IN PRESS
Catalysis Today xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Bismuth molybdates prepared by solution combustion synthesis for
the partial oxidation of propene
Benjamin Farina, Alessandro H.A. Monteverde Videlab, Stefania Specchiab,∗
,
Eric M. Gaigneauxa,∗∗
a Institute of Condensed Matter and Nanosciences – MOlecules, Solids and reactiviTy (IMCN/MOST), Université catholique de Louvain,
Croix du Sud 2 box L7.05.17, 1348 Louvain-La-Neuve, Belgium
b Green Energy & Engineering (Gre.En2) Group – Department of Applied Science and Technology, Politecnico di Torino, Corso Duca degli Abruzzi 24,
10129 Torino, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
The solution combustion method (SCS) is demonstrated as an easy and fast alternative method allowing
the synthesis of mixed oxides with thermally sensitive metals, still keeping a good control on their stoi-
chiometry and phase composition. Bismuth molybdates with different theoretical Bi/Mo atomic ratios,
namely ␣-Bi2Mo3O12, ␤-Bi2Mo2O9, and ␥-Bi2MoO6 were synthesized by the SCS, fully characterized by
XRD, Raman, SEM/EDX, BET and XPS analyses, and tested in the partial oxidation of propene to acrolein.
The SCS method allowed obtaining crystalline catalysts with Bi/Mo atomic ratios close to the theoreti-
cal values and very good catalytic properties namely high propene conversion (from 13.4 to 23.4%) and
acrolein selectivity (from 71.5 to 80.0%) at 425 ◦C. A high purity of the SCS prepared bismuth molybdates
was obtained by means of a subsequent calcination treatment. Such treatment did not alter the high
catalytic activity of the catalysts, which slightly increased (from 12.3 to 25.5%), but induced a marked
loss of acrolein selectivity (from 66.2 to 48.4%) at 425 ◦C. This effect is due to a strong increase of the
oxidation states of Mo and Bi and reduction of the specific surface area during the calcination.
© 2015 Elsevier B.V. All rights reserved.
Received 28 October 2014
Received in revised form 16 March 2015
Accepted 18 March 2015
Available online xxx
Keywords:
Solution combustion synthesis
Bismuth molybdates
Propene
Partial oxidation
Acrolein
Selectivity
1. Introduction
added to the mixture to form complexes with the metal ions and
to facilitate the obtaining of a homogeneous solution. The organic
Co-precipitation, high temperature solid-state reactions, spray
drying or sol–gel syntheses are the most common methods used to
prepare mixed oxide catalysts [1–8]. These routes require severe
conditions and long calcination treatments at high temperatures
to prepare crystalline materials. This is why the morphology and
surface texture of the as-prepared catalysts and thus their resulting
catalytic activity are not easily controlled. On this point of view, the
solution combustion synthesis (SCS) appears as an elegant alter-
native protocol. This method is quite simple and presents several
advantages. It is based on highly exothermic redox chemical reac-
tions between metallic compounds and non-metallic ones.
molecule works as a fuel as its combustion enables to release a
huge amount of heat in a short period of time. Various organic com-
pounds can be used such as urea, glycine, hydrazine or precursors
containing a carboxylate anion. The two first compounds are the
most convenient ones since they are cheap and readily available
commercially. The second step of the SCS process thus consists in
heating the final solution until reaching temperatures in the range
of 300–450 ◦C, which brings the solution to ebullition. The resulting
mixture then becomes dry and in a matter of minutes ignites what
sets off highly exothermic, self-sustaining and fast redox reactions
that generate a dry, usually crystalline, fine powder.
The first step of the SCS involves the preparation of an aque-
ous solution containing suitable metal salts. Nitrates are chosen as
metals precursors because −of their high solubility in water and the
oxidizing potential of NO3 groups. An organic molecule is then
The exothermicity of these redox reactions allows reaching tem-
perature peaks that vary from 700 to 1500 ◦C [9–15]. The very short
residence time at high temperature enables synthesizing solid with
a minimized occurrence of sintering. Compounds having a rela-
tively high specific surface area can thus be prepared. The rapidity
of the method may also allow the formation of metastable phases
[9–12]. Finally, a rapid crystallization could favour defects which
sometimes enable improving the selectivity of catalysts towards
the wished product. The SCS method, in fact, has been favourably
used to produce cheaply a huge variety of nanomaterials, such as
∗
Corresponding author. Tel.: +39 0110904608.
Corresponding author. Tel.: +32 010473665.
E-mail addresses: stefania.specchia@polito.it (S. Specchia),
∗∗
eric.gaigneaux@uclouvain.be (E.M. Gaigneaux).
http://dx.doi.org/10.1016/j.cattod.2015.03.045
0920-5861/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: B. Farin, et al., Bismuth molybdates prepared by solution combustion synthesis for the partial oxidation
of propene, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.045

G Model
CATTOD-9572; No. of Pages7
ARTICLE IN PRESS
2
B. Farin et al. / Catalysis Today xxx (2015) xxx–xxx
catalysts, phosphors, pigments and refractories, for many different
applications [16–22].
80 ◦C to ensure the complete dissolution of all reagents. It was then
transferred in a ceramic dish and placed into an electric oven set at
430 ◦C. After water evaporation and a significant increase in the
system viscosity, the heat released in the fast reaction allowed
the formation of the catalytic powders. The total time necessary
to produce the powder was approximately 20 min. Subsequently,
after grinding in an agate mortar, half of the prepared powders
were calcined in an oven at 425 ◦C for 4 h in static air, to favour the
decomposition of the eventually unreacted nitrate precursors and
improve the formation of the desired ␣, ␤, or ␥ crystalline phases.
The samples obtained right after the SCS are noted with F at the
end of their label (F for fresh), whereas the calcined ones are noted
with C at the end of their label (C for calcined).
A reference sample (Bi2Mo2O9-Cop) having a Bi/Mo ratio of 1
was also prepared via the co-precipitation method, as described
by Carrazan et al. [29]. Bi(NO3)3·5H2O (0.04 mol) was dissolved in
a distilled water solution (1 L) heated at 50 ◦C by using concen-
trated HNO3 (0.05 L). (NH4)6Mo7O24·4H2O (0.006 mol) dissolved in
distilled water (0.35 L) was added to the bismuth solution before
adjusting the pH at 5 with diluted NH3 (5 mol L−1). The final solu-
tion was stirred for 2 h, aged 24 h and filtered. The recovered solid
was dried overnight at 110 ◦C and calcined at 550 ◦C, for 4 h, in static
air and in a muffle oven. A 5 ◦C min−1 ramp was used to reach the
desired calcination temperature.
However, severe conditions as the high-temperatures that are
reached during the SCS method appear conflicting with the syn-
thesis of materials made of temperature sensitive metals like
molybdenum, rhenium and ruthenium. Among these, MoO3 indeed
sublimates starting from 800 ◦C [23]. The opportunity to generate
temperature sensitive oxides with relatively high specific surface
area via the SCS method is then here broached through the syn-
thesis of bismuth molybdates. Three main crystallographic phases
of BiMo-mixed oxides are distinguished at atmospheric pressure:
Bi2Mo3O12 (the ␣-phase), Bi2Mo2O9 (the ␤-phase) and Bi2MoO6
(the ␥-phase) [24–26]. These phases are generally prepared via a
solid co-precipitation within an aqueous solution having a Bi/Mo
molar atomic ratio equal to 2/3, 1/1 and 2/1, respectively. In the
present study, these materials are prepared via the SCS method.
A first key question is to evaluate to what extent the SCS method
allows preparing bismuth molybdates keeping the control of the
stoichiometry and homogeneity of the prepared solids. After their
synthesis, the bismuth molybdates are used as catalysts in the
propene partial oxidation to acrolein. The second key question is
to check whether the simple SCS method is able to produce highly
efficient catalysts. This reaction was selected because of the well-
known efficiency of bismuth molybdates for the allylic oxidation
of olefins. In fact, the reaction’s products are important intermedi-
ates involved in numerous downstream chemical processes [6,27].
In a world where sustainability and the development of environ-
mentally friendly chemical processes has become a major concern,
this work is thus also in line with the efforts trying to improve the
valorization of light olefins [28].
2.3. Characterization
X-ray diffraction measurements (XRD) were performed on
a Siemens D5000 diffractometer using the K␣ radiation of Cu
(ꢁ = 0.15418 nm). The 2ꢂ diffractograms were recorded at a rate of
1.2◦ min−1 between 5 and 75◦. The ICDD-JCPDS database was used
to identify the detected crystalline phases.
2. Experimental
Nitrogen physisorption was performed at −196 ◦C on
a
Micromeritics ASAP 2020 instrument. Before the measurement,
each sample (about 100 mg) was outgassed overnight at 150 ◦C
in vacuum (7 Pa). The specific surface area was evaluated by
the BET method between 0.05 and 0.30 p/p◦. The pore diameter
distribution was evaluated by the Barrett–Joyner–Halenda (BJH)
method, calibrated for cylindrical pores according to the improved
Kruk–Jaroniec–Sayari (KJS) method, with the corrected form of the
Kelvin equation, from the desorption branches of the isotherms.
Confocal Raman spectroscopy was performed on an InVia
2.1. Chemicals
Bismuth nitrate pentahydrate (Bi(NO3)3·5H2O, >98% purity),
ammonium molybdate tetrahydrate ((NH4)6Mo7O24·4H2O, >99%
purity), and urea (NH2CONH2, > 99% purity), were purchased from
Sigma–Aldrich. Nitric acid (HNO3, 65 wt.%) was purchased from
Merck. All aqueous solutions were prepared using ultrapure
water obtained from a Millipore Milli-Q system with a resis-
tivity > 18 Mꢀ cm−1. Technical grade propene (99.5%), research
oxygen and nitrogen (99.999% purity) gases were provided by Prax-
air and used as received.
Raman microscope (Renishaw) equipped with
a diode light
(785 nm). The spectra were recorded between 100 and 3500 cm−1
with a 4 cm−1 resolution. 10 scans were recorded and averaged for
each sample. Moreover, an acquisition time of 10 s was selected, a
laser power was set to 10 mW and the 50× objective was used to
focus the apparatus.
2.2. Synthesis of materials
The bismuth molybdates were prepared by the SCS method
according to the following stoichiometric reactions:
Scanning electron microscopy coupled with energy-dispersive
X-ray spectroscopy (SEM-EXD FEI-QuantaTM Inspect 200 with
EDAX PV 9900 instrument working at 15 kV) was performed to
analyze the morphology and evaluate the average chemical com-
position of the prepared samples. EDX spectra were analyzed with
the Genesis Spectrum v. 6.04 (EDAX Inc.) software.
X-ray photoelectron spectroscopy (XPS) was performed using
a Physical Electronics PHI 5800 Versa Probe electron spectrome-
ter system with monochromated Al K␣ X-ray source at 1486.60 eV
operated at 25 W, 15 kV, with 100 micron X-rays spot. To reduce
any possible charging effects of X-rays, a dual beam charge neu-
tralization method was applied, combining both low energy ions
and electrons. The samples were previously outgassed in an ultra-
high vacuum chamber at 2.5 × 10−6 Pa for 12 h. Survey scans were
recorded from 0 to 1200 eV. The narrow Bi 4f spectra were col-
lected from 148 to 170 eV, the narrow Mo 3d spectra from 218
to 240 eV, and the narrow O 1s spectra from 524 to 536 eV. The
samples were analyzed under identical conditions and corrections
3
7
26
␣-phase : 2Bi(NO3)3
+
(NH4)6Mo7O24
26
+
+
CH4N2O
H2O
7
88
→ Bi2Mo3O12
+
CO2 + 8N2
7
7
29
7
2
7
+
␤-phase : 2Bi(NO3)3
+
(NH4)6Mo7O24
29
+
CH4N2O
H2O
82
7
→ Bi2Mo2O9
CO2 + 8N2 +
7
1
7
+
32
7
␥-phase : 2Bi(NO3)3
+
(NH4)6Mo7O24
32
7
+
CH4N2O
76
→ Bi2MoO6
CO2 + 8N2
+
H2O
7
To prepare 500 mg of each type of catalyst, Bi and Mo precur-
sors, and urea were used in a stoichiometric amount, according to
the ␣-, ␤-, or ␥-phase, and dissolved in ultrapure distilled water.
For each catalyst, the obtained solution was thoroughly stirred at
Please cite this article in press as: B. Farin, et al., Bismuth molybdates prepared by solution combustion synthesis for the partial oxidation
of propene, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.045

G Model
CATTOD-9572; No. of Pages7
ARTICLE IN PRESS
B. Farin et al. / Catalysis Today xxx (2015) xxx–xxx
3
referred to C 1s at 284.5 eV, for electrostatic charging. The Multipak
9.0 software was used for obtaining semi-quantitative atomic per-
centage compositions. The peak position and areas were evaluated
using symmetrical Gaussian–Lorentzian equations (in the fraction
of 70% and 30%, respectively) with Shirley-type background, and a
standard deviation in locating the peaks equal to 0.3 eV.
2.4. Catalytic tests
Bismuth molybdates (0.1 g) were tested in the propene par-
tial oxidation to acrolein at 425 ◦C and in
a
0.04 nL min−1
total flow (C3H6/O2/N2 = 12.5/25/62.5 vol.%). The catalyst particles
(200 ␮m < Dp ≤ 315 ␮m) were mixed with 0.4 g of quartz spheres
(Dp < 200 ␮m) previously checked to be inactive and put inside a
U-shaped fixed bed quartz reactor with an internal diameter of
4 mm. Thereby, a plug flow was warranted and heat transfer limita-
tions were avoided. A hot-box maintained the gas lines at 100 ◦C in
order to preheat the flow upstream the catalytic reactor. The reac-
tion mixture was analyzed by using an on line Varian GC CP3800. A
Hayesep column coupled with a Molecular Sieve column and a TCD
detector were used to separate and quantify O2, N2, CO and CO2.
The oxygenates (acrolein, propanal, acetone, acetaldehyde and pro-
pylene oxide) and propene were detected and quantified by means
of an EC-Wax column coupled with a FID detector. Traces of acrylic
acid and acetic acid were sometimes detected but not quantified.
Carbon balances superior or equal to 90% were calculated from
the catalytic results. Here below, the catalytic performances are
expressed in terms of propene conversion, product selectivity and
acrolein yield.
Fig. 2. Raman spectra of fresh and calcined bismuth molybdates prepared by the
SCS method. The one of the co-precipitated sample (Bi2Mo2O9-Cop) calcined for
4 h at 550 ◦C is also shown. ␣-Bi2Mo3O12 (␣), ␤-Bi2Mo2O9 (␤), ␥-Bi2MoO6 (␥) were
identified.
prepared by co-precipitation, Bi2Mo2O9-Cop, is also presented. The
phases composition of the fresh materials prepared via the SCS
method, BixMoyOz-SCS F, is always improved by the 4 h calcination
treatment at 425 ◦C. The materials having a Bi/Mo theoretical ratio
of 1/1 (Bi2Mo2O9-Cop, Bi2Mo2O9-SCS F, and Bi2Mo2O9-SCS C) dis-
play a mix of ␣- and ␤-phase regardless the preparation method
and the calcination conditions. However, the ␤-phase systemati-
cally remains the dominating phase. The Bi2Mo3O12-SCS F sample
presents a mix of ␣-phase and MoO2. The latter is easily detected
due to the presence of a characteristic sharp peak at 2ꢂ = 26◦ [30]. A
calcination at 425 ◦C for 4 h of this sample improves the purity of the
␣-phase as no more MoO2 is then detected in the Bi2Mo3O12-SCS C.
The main reflections of the diffractogram of the Bi2MoO6-SCS F are
assigned to a mix of ␤-Bi2O3 and the ␣- or ␤-phase. Only small
reflections reveal also the minor presence of the desired ␥-phase.
Nevertheless, the calcination at 425 ◦C for 4 h favours the crys-
tallization of the ␥-phase, which then becomes the dominating
crystalline phase in the Bi2MoO6-SCS C catalyst. These observa-
tions are confirmed by Raman results whose spectra were analyzed
according to data gathered by Li et al. [3], as shown in Fig. 2. The
␣-Bi2Mo3O12 phase is identified by means of six bands at 960,
928, 906, 862, 845 and 818 cm−1, respectively. They correspond
to the stretching modes of each tetrahedral molybdate species.
The ␤-Bi2Mo2O9 phase is rather detected through a strong band
at 886 cm−1 and two weaker bands at 786 and 754 cm−1 whereas
3. Result
The XRD patterns of the fresh and calcined bismuth molybdates
made from the SCS method are shown in Fig. 1 (hereafter named
BixMoyOz-SCS F or BixMoyOz-SCS C, respectively). The catalyst
the ␥-Bi2MoO6 phase displays 3 bands at 852, 796 and 718 cm−1
,
respectively.
The content of each crystalline phase detected within the sam-
ples characterized was calculated by using the following equation
[31]:
I␣
ꢀ
˛(%) =
(I␣ + I␤ + I␥ + . . .)
where ␣, ␤, ␥ represent a given phase and I the intensity of
the strongest XRD reflection. The obtained results are reported in
Table 1. The reflections at 2ꢂ = 27.95, 27.86, 28.31, 27.95, and 25.81◦
are the strongest ones of ␣-Bi2Mo3O12, ␤-Bi2Mo2O9, ␥-Bi2MoO6,
␤-Bi2O3 and MoO2, respectively. Since the strongest XRD reflec-
tion of the ␣ and ␤ phases are overlapping, the strongest Raman
band at 906 and 886 cm−1 were respectively used to determine the
composition of samples (Bi/Mo = 1/1) containing both phases.
The Bi/Mo atomic ratio of each catalyst was calculated by EDX
and shown in Table 1. Several zones of the samples examined (from
Fig. 1. Diffractograms of fresh and calcined bismuth molybdates prepared by the
SCS method. The one of the co-precipitated sample (Bi2Mo2O9-Cop) calcined for 4 h
at 550 ◦C is also shown. ␣-Bi2Mo3O12 (␣), ␤-Bi2Mo2O9 (␤), ␥-Bi2MoO6 (␥) and MoO2
(0) were identified. The symbol “*” refers to a mix of ␤-Bi2Mo3 and ␤-Bi2Mo2O9 or
␣-Bi2Mo3O12
.
Please cite this article in press as: B. Farin, et al., Bismuth molybdates prepared by solution combustion synthesis for the partial oxidation
of propene, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.045

G Model
CATTOD-9572; No. of Pages7
ARTICLE IN PRESS
4
B. Farin et al. / Catalysis Today xxx (2015) xxx–xxx
Table 1
with pores within 1.7 and 300 nm), with an average pore size
of 2.2 nm.
Phase composition, specific surface area (BET s.s.a.) and Bi/Mo atomic ratio of BiMo-
based materials prepared through the SCS and co-precipitation methods.
In agreement with the literature, the bismuth molybdate pre-
pared by co-precipitation has a very low specific surface area
(1 m2 g−1), much lower than all the SCS-calcined catalysts [32]. On
the other hand, bismuth molybdates prepared by SCS have higher
specific surface area (6.5–9 m2 g−1), compared to bismuth molyb-
dates prepared by spray drying (<2 m2 g−1, [23]).
Table 2 gathers the catalytic performances of the bismuth
molybdates in the partial oxidation of propene at 425 ◦C. The mate-
rials having a theoretical Bi/Mo ratio of 1/1 are the most active ones
(Cpropene between 23.4 and 25.5%), whereas those having a theo-
retical Bi/Mo ratio of 2/3 are the less active (Cpropene around 13%).
Intermediate propene conversions (around 18%) are obtained with
Bi2MoO6-SCS samples. Besides, the activity of the catalysts made
from the SCS method is not significantly affected by the calcination
treatment. Blank tests performed on all of the prepared catalysts at
the same temperature of 425 ◦C showed that there is a few auto-
oxidation of propene into CO2, equivalent to a propene conversion
of maximum 3.5%.
Bi2Mo2O9-Cop rather deeply oxidizes propene as its CO2
selectivity (66.2%) is superior to its acrolein selectivity (21.3%)
(Table 2). At the opposite, the SCS method generates materials
being much more selective in acrolein, especially when being
fresh (not calcined: Sacrolein > 71.5%). These catalysts, in fact, pref-
erentially produce acrolein to the detriment of CO and CO2. CO
and oxygenates (besides acrolein) remain secondary by-products
as their selectivity never exceeds 11.1% and 4.2%, respectively.
As in the case of the specific surface area, bismuth molybdates
prepared by SCS have higher acrolein selectivity (71.5–80%), com-
pared to bismuth molybdates prepared by spray drying (65–75%,
[23]). The calcination of the SCS catalysts at 425 ◦C strongly
modifies their selectivity profile. This treatment indeed simulta-
neously increases the CO2 production and decreases the acrolein
selectivity. This effect is particularly pronounced for the Bi2Mo2O9-
SCS C catalyst whose acrolein selectivity is roughly decreased of
about 40% compared to the Bi2Mo2O9-SCS F catalyst (from 80.0
to 48.4%).
Catalysts
Phase
composition
Bi/Mo atomic
ratio from EDX
analysis
BET
s.s.a.
(m2 g−1
)
Bi2Mo2O9-Cop
Bi2MoO6-SCS F
␤ (79%) + ␣ (21%)
␤ or ␣ + ␤-Bi2O3
(81%) + ␥ (19%)
␥ (84%) + ␤ and
␤-Bi2O3 (16%)
␤ (52%) + ␣ (48%)
␤ (63%) + ␣ (37%)
␣ (76%) + MoO2
(24%)
–
1.0
6.7
1.29–1.37
Bi2MoO6-SCS C
1.44–1.47
4.3
Bi2Mo2O9-SCS F
Bi2Mo2O9-SCS C
Bi2Mo3O12-SCS F
0.64–0.71
0.86–0.88
0.53–0.58
9.3
4.0
6.8
Bi2Mo3O12-SCS C
␣ (100%)
0.65–0.72
2.8
3 to 5 per sample) were screened to check the homogeneity of
the samples. These analyses highlight that the Bi/Mo ratios of the
samples after the calcination are higher than the ratios for the cor-
responding fresh samples.
The SEM analysis on the prepared SCS catalysts is shown in
Fig. 3, at two different magnification values. Images allow appreci-
ating a spongy-like structure of the fresh catalysts, obtained thanks
to the specific technique used to prepare them, because of the
very fast reaction combined with the release of the gases from
the explosive-like process. BET measurements, listed in Table 1,
confirm the relatively high specific surface area of the fresh cat-
alysts, ranging from a maximum value of 9.3 m2 g−1 exhibited by
the Bi2Mo2O9-SCS F to a minimum value of 6.7 m2 g−1 exhibited
by the Bi2MoO6-SCS F. However, the calcination of these oxides
at 425 ◦C induces a decrease of the specific surface area, ranging
from 4.3 (the Bi2MoO6-SCS C) to 2.8 m2 g−1 (Bi2Mo3O12-SCS C),
because of possible sintering effects. Despite a higher specific
area, the fresh bismuth molybdates prepared by SCS were roughly
non porous. The measured total pore volume was indeed quite
small and ranged between 0.0040 and 0.0045 cm3 g−1 (coinci-
dent with the BJH adsorption cumulative volume of mesopores
Fig. 3. SEM images at 10,000× and 80,000× of fresh bismuth molybdates prepared by the SCS method.
Please cite this article in press as: B. Farin, et al., Bismuth molybdates prepared by solution combustion synthesis for the partial oxidation
of propene, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.045

G Model
CATTOD-9572; No. of Pages7
ARTICLE IN PRESS
B. Farin et al. / Catalysis Today xxx (2015) xxx–xxx
5
Table 2
Propene conversion (Cpropene), propene reaction rate (RRropene), intrinsic activity (Aint), product selectivities (Sproduct), and acrolein yield (Yacrolein) of bismuth molybdates made
via the SCS and co-precipitation methods.
b
b
a
Catalyst
Cpropene
(%)
RRpropene
Aint
Sacrolein
(%)
SCO
SCO
(%)
Soxygenates
(%)
Yacrolein
(%)
(mol m−2 s−1
)
(% m−2
)
(%)2
Bi2Mo2O9-Cop
Bi2MoO6-SCS F
Bi2MoO6-SCS C
Bi2Mo2O9-SCS F
Bi2Mo2O9-SCS C
Bi2Mo3O12-SCS F
Bi2Mo3O12-SCS C
20.6
17.7
18.1
23.4
25.5
13.4
12.3
7.659 × 10−6
9.822 × 10−7
1.565 × 10−6
9.355 × 10−7
2.370 × 10−6
7.326 × 10−7
1.633 × 10−6
206
21.3
77.3
66.2
80.0
48.4
71.5
65.3
63.4
17.3
26.9
14.8
36.7
21.4
29.8
12.1
2.6
4.0
1.0
11.1
4.7
3.2
2.8
2.9
4.2
3.8
2.4
2.7
3.8
14.0
11.7
19.0
11.7
9.3
26.4
42.1
25.2
63.7
19.7
43.9
2.2
7.4
a
Soxygenates regroups the selectivity of acetaldehyde, propanal, propylene oxide and acetone.
The reaction rate (RRpropene) and the intrinsic activity (Aint) refer to the propene reacted or propene conversion, respectively, normalized on the specific surface area and
b
the mass used of the respective catalysts.
Table 3
At this stage, the acrolein yield of each catalyst can be calcu-
lated (Table 2). Compared to Bi2Mo2O9-Cop, the catalysts made
from the SCS method are even overall active and much more selec-
tive in acrolein. This is why they also display a superior acrolein
yield. Bi2Mo2O9-SCS F and Bi2MoO6-SCS F are the most efficient
catalysts as they display an acrolein yield equal to 17.3 and 14.8%,
respectively. For these catalysts, the calcination treatment led to a
reduction of the acrolein yield down to 11.7% for both of them. Such
a value is anyway much higher compared to the acrolein yield of
the Bi2Mo2O9-Cop catalyst.
XPS analyses: binding energies from the decomposition of the high resolution Bi 4f,
Mo 3d, and O 1s spectra for the Bi2Mo2O9-SCS in the fresh, calcined and used status.
Catalyst
Bi 4f7/2
Mo 3d5/2
O 1s
Bi3+
Bi0
Mo6+
Mo04+
228.8
Bi2Mo2O9-SCS F
Bi2Mo2O9-SCS C
Bi2Mo2O9-SCS U
159.9
159.8
160.1
156.8
–
–
231.8
232.2
232.6
530.3
530.3
530.7
On the most critical catalyst, namely the Bi2Mo2O9-SCS, which
presented the highest propene conversion and acrolein selectiv-
ity at the fresh status, but suffered the highest acrolein selectivity
reduction because of the calcination treatment, the XPS analysis
was performed. The obtained results are shown in Fig. 4 and in
Table 3, as binding energies for Bi 4f, Mo 3d, and O 1s from the
decomposition of the high resolution spectra. The Bi/Mo atomic
ratio calculated from the XPS shifted from 0.99 for the fresh cata-
lyst to 1.13 for the calcined one, which is the sign of a superficial
enrichment of Bi. These values are in agreement with the Bi/Mo
atomic ratios calculated by EDX analysis (Table 1). Observing the
high resolution spectra of Bi 4f, Mo 3d, and O 1s, their shapes
changed markedly from the fresh status to the calcined one. This is
the sign of the existence of different chemical species on the sur-
face. Specifically, for Bi 4f spectra, peaks at 156.8 and 162.1 eV on
the fresh catalyst disappeared with the calcination. The main top
peak of Bi 4f7/2 at 159.9 eV can be assigned to Bi3+ and, considering
the standard deviation affecting the measurements, this peak did
not shift with the calcination treatment [33,34]. The presence of
the peak at 156.8 eV on the fresh spectrum can be assigned to Bi0
[33,34], which disappeared with calcination. For Mo 3d spectra,
the change in the shape indicates the transition from Mo4+ to Mo6+
(MoO3) [24,30,34,35], with a small shift in the binding energies to
higher values. Considering the O 1s, the spectrum of the fresh cata-
lyst appeared as an asymmetrical peak with a shoulder at lower
binding energy. The shoulder reduced drastically in intensity in
the calcined catalyst, but the main peak remained centred at the
same binding energy. This is the sign that complex oxides are dis-
tributed in different amounts [33]. These results are in agreement
with the XRD analysis (Table 1 and Fig. 1): the calcined Bi2Mo2O9-
SCS resulted with a higher content of the ␤-phase respect to the
␣-phase compared to the fresh Bi2Mo2O9-SCS catalyst.
Furthermore, on the most interesting catalyst, namely the
Bi2Mo2O9-SCS F, the selective oxidation of propene to acrolein was
Fig. 4. XPS Bi 4f, Mo 3d, and O 1s high resolution spectra measured on the fresh, calcined, and used Bi2Mo2O9 prepared by the SCS method.
Please cite this article in press as: B. Farin, et al., Bismuth molybdates prepared by solution combustion synthesis for the partial oxidation
of propene, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.045

G Model
CATTOD-9572; No. of Pages7
ARTICLE IN PRESS
6
B. Farin et al. / Catalysis Today xxx (2015) xxx–xxx
conducted at the same conditions previously described for a period
of time equal to 20 h, for a first check on the stability. At the end of
this time on stream, the propene conversion slightly increased and
stabilized at around 25%, and acrolein selectivity dropped down to
47%, with the consequent increase of the carbon dioxide selectiv-
ity. The performance results obtained for this catalyst, that is the
Bi2Mo2O9-SCS U where “U” stands for “used” catalyst, are very sim-
ilar to the results obtained for the calcined catalyst Bi2Mo2O9-SCS C
(Table 2). To document this, BET and XPS measurements were per-
formed on the Bi2Mo2O9-SCS U catalyst. The BET specific surface
area dropped down to 3.6 m2 g−1, namely a value very similar to the
BET value of the calcined catalyst Bi2Mo2O9-SCS C (Table 1). XPS
high resolution spectra of the Bi2Mo2O9-SCS U catalyst are shown
in Fig. 4, together with the binding energies of the various high
resolution spectra in Table 3. These spectra are very similar to the
spectra of the calcined catalyst Bi2Mo2O9-SCS C. Specifically, the
spectrum of the Bi 4f overlapped the spectrum of the calcined cat-
alyst. On the other hand, the spectra of the Mo 3d and O 1s became
sharper and more symmetrical, with a slight shift of the bind-
ing energies towards higher binding energies (Table 3). The Bi/Mo
atomic ratio calculated from the XPS was 1.22, showing an increase
compared to value of the calcined catalyst (1.13), which was already
increased compared to value of the fresh catalyst (0.99). Thus, the
Bi2Mo2O9-SCS U catalyst, after 20 h of time on stream, appears in
a more oxidized state [34], and richer in bismuth.
on their acrolein selectivity. This tendency is confirmed regarding
the Bi2Mo2O9-Cop sample which presents an acrolein selectiv-
ity lower than 20%. A calcination treatment is known to be able
to affect both the structure and the catalytic behaviour of cata-
lysts. For example, the sintering of isolated vanadium species in
V2O5 crystalline particles induces a decrease of the propene selec-
tivity of VOx/SiO2 materials in the oxidative dehydrogenation of
propane to propene [36]. Here, the calcination causes a further
crystallization of the bismuth molybdates and likely favours the
formation of sites which overoxidize propene into COx. At the
opposite, the fresh materials made via the SCS method present an
anarchical crystallinity due to the rapid crystallization promoted by
the exothermic combustion of an organic fuel. Such materials then
rather possess defects and punctual sites being more efficient to
selectively oxidize propene to acrolein. The calcination treatment,
in fact, clearly cancels the potential of the SCS method to prepare
highly acrolein selective materials.
Considering specifically bismuth molybdates, very similar
results were obtained by Schuh et al. [2] studying the selective
oxidation of propylene to acrolein with catalysts prepared by
hydrothermal method. In this case, the catalytic activity of the fresh
hydrothermally prepared catalysts was significantly higher than
after their calcination (4 h at 550 ◦C). Furthermore, the obtained
acrolein yield, that is proportional to the selectivity, decreased
regardless of the phase composition. According to the literature,
in fact, calcination time and temperature strongly influence the
activity of the bismuth molybdenum catalysts, causing surface
enrichment in bismuth after calcination at high temperature or
long time exposure [37]. This is exactly what happened to our
fresh catalysts after calcination (Tables 1 and 2), and after 20 h of
time on stream: with the exposure at high temperature (because
of the calcination treatment, or because of the prolonged time on
stream in the reactive mixture) the activity remained constant,
or slightly increase, whereas the selectivity decreased and the
intrinsic activity increased because of the decrease of the specific
surface area. Moreover, the surface of the catalyst got enriched in
bismuth.
This is in fact demonstrated by the BET and XPS analyses con-
ducted on the fresh, calcined and used Bi2Mo2O9-SCS catalyst. The
calcination treatment, in fact, induced a decrease of the specific
surface area (Table 1) and a strong variation of the oxidation states
of Mo and Bi (Fig. 4). The spectrum of the Bi2Mo2O9-SCS F, whose
acrolein selectivity was the highest, pointed out the presence of
Bi2O3 and Bi0. After the calcination process, metallic Bi disappeared,
and simple Bi and Mo oxides appeared. Moreover, a readjustment
of superficial oxides gave a Bi rich surface after calcination. Prac-
tically the same happened with used catalyst: the only difference
between the calcined and the used catalysts lies in the exposure
time at high temperature (of course longer for the used catalyst).
In agreement with the literature, these results highlight the pre-
disposition of the bismuth molybdates’ structure to change. The
best catalytic performances were obtained when the surface of
the oxide was maintained in a slightly reduced state, as it is the
case for the fresh materials prepared by SCS [33,38]. At the oppo-
site, the formation of more stable oxides that results from the
calcination treatment, or from the time on stream under the reac-
tive flow, is responsible for a higher intrinsic activity, and higher
propene reaction rate, but a lower acrolein selectivity. The result-
ing effect was thus the decrease of performance in terms of acrolein
yield.
4. Discussion
The SCS method is suitable to prepare crystalline bismuth
molybdates being effective in the selective oxidation of propene
into acrolein. According to XRD and Raman results, the desired
phase has been successfully obtained (Figs. 1 and 2). However, its
purity can be clearly improved by means of a calcination treatment.
The increase of the purity corresponds to an increase of the reaction
rate expressed as the number of moles of reacted propene normal-
ized on the specific surface area and the mass used of the respective
catalysts, as listed in Table 2. These values are better compared to
the data calculated for bismuth molybdates synthesized by spray
drying (values ranging from 1.8 × 10−7 to 2.8 × 10−7 [23]). Accord-
ing to the SEM pictures (Fig. 3), the SCS method also promotes
the rapid crystallization of finely divided materials presenting a
higher specific surface area with respect to the co-precipitation
method.
The catalytic behaviour of bismuth molybdates can be under-
stood by means of a careful observation of their specific surface
area and composition phase. On the one hand, the crystallinity of a
bismuth molybdate is necessary to obtain active sites for the con-
version of propene into acrolein (i.e., Bi-O-Mo-O) [23]. On the other
hand, a high specific surface area is required in order to maximize
the number of accessible sites on the catalyst surface. Here, the cal-
cination treatment favours a further crystallization of the samples
made by the SCS method to the detriment of their specific surface
area (Table 1). This further crystallization then makes the calcined
materials more intrinsically active as it is clearly demonstrated in
Table 2. The propene conversion drop linked to a specific surface
area decrease is then balanced by a propene conversion raise linked
to an improvement of the crystallinity. This phenomenon makes the
activity of the bismuth molybdates not really sensitive to the cal-
cination treatment. This statement can be strengthened regarding
the Bi2Mo2O9-Cop sample which was calcined for 4 h at 550 ◦C.
The very low specific surface area of this sample is balanced by an
important crystallinity that maintains its activity at a same level as
the Bi2Mo2O9-SCS sample.
Considering the results obtained on the used catalyst Bi2Mo2O9-
SCS U, which confirmed that after 20 h of time on stream the
performance was practically the same of the calcined catalyst
Bi2Mo2O9-SCS C (Tables 1–3 and Fig. 4), the calcination treatment
performed on the fresh catalyst can be considered as a stabiliz-
ing treatment, and the performance of the calcined catalyst can be
While calcination does not affect the activity of bismuth molyb-
dates made from the SCS method, it has got a marked adverse effect
Please cite this article in press as: B. Farin, et al., Bismuth molybdates prepared by solution combustion synthesis for the partial oxidation
of propene, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.045

G Model
CATTOD-9572; No. of Pages7
ARTICLE IN PRESS
B. Farin et al. / Catalysis Today xxx (2015) xxx–xxx
7
considered stable in terms of propene conversion. On the other
hand, the acrolein selectivity decreased, whereas the intrinsic activ-
ity and the resulting propene reaction rate increased.
[2] K. Schuh, W. Kleist, M. Høj, V. Trouillet, P. Beato, A. Degn Jensen, G.R. Patzke,
J.-D. Grunwaldt, Appl. Catal. A: Gen. 482 (2014) 145–156.
[3] H.H. Li, K.W. Li, H. Wang, Mater. Chem. Phys. 116 (2009) 134–142.
[4] L.M. Thang, L.H. Bac, I. Van Driessche, S. Hoste, W.J.M. Van Well, Catal. Today
131 (2008) 566–571.
[5] E. Vila, J.E. Iglesias, J. Galy, A. Castro, Solid State Sci. 7 (2005) 1369–1376.
[6] M.D. Wildberger, J.-D. Grunwaldt, M. Maciejewski, T. Mallat, A. Baiker, Appl.
Catal. A: Gen. 175 (1998) 11–19.
[7] R.P. Rastogi, A.K. Singh, C.S. Shukla, J. Solid State Chem. 42 (1982) 136–148.
[8] S. Yue, M. Tian, Z. Mao, L. Chen, Y. Zhang, J. Cryst. Growth 222 (2001)
747–750.
[9] S. Specchia, C. Galletti, V. Specchia, Stud. Surf. Sci. Catal. 175 (2010) 59–67.
[10] S. Specchia, E. Finocchio, G. Busca, V. Specchia, Combustion synthesis, in: M.
Lackner, F. Winter, A.K. Agarwal (Eds.), Handbook of Combustion, Wiley-VCH
Verlag GmbH & Co. KGaA, Weinheim, 2010, pp. 439–472.
[11] J.J. Moore, H.J. Feng, Prog. Mater. Sci. 39 (1995) 243–273.
[12] J.J. Moore, H.J. Feng, Prog. Mater. Sci. 39 (1995) 275–316.
[13] K.C. Patil, S.T. Aruna, S. Ekambaran, Curr. Opin. Solid State Mater. Sci. 2 (1997)
158–165.
[14] A. Varma, A.S. Rogachev, A.S. Mukasyan, S. Hwang, Adv. Chem. Eng. 24 (1998)
79–226.
[15] K.C. Patil, S.T. Aruna, T. Mimani, Curr. Opin. Solid State Mater. Sci. 6 (2002)
507–512.
[16] D. Pereira Tarragó, C. de Fraga Malfatti, V. Caldas de Sousa, Powder Technol.
269 (2015) 481–487.
[17] S.L. González-Cortés, F.E. Imbert, Appl. Catal. A: Gen. 452 (2013) 117–131.
[18] G. Liu, J. Li, K. Chen, Int. J. Refract. Met. Hard Mater. 39 (2013) 90–102.
[19] S.T. Aruna, A.S. Mukasyan, Curr. Opin. Solid State Mater. Sci. 12 (2008)
44–50.
5. Conclusion
Bismuth molybdates with different theoretical Bi/Mo atomic
ratios, namely ␣-Bi2Mo3O12, ␤-Bi2Mo2O9, and ␥-Bi2MoO6 were
successfully synthesized by the solution combustion method. SCS
thus appears as a suitable technique to prepare highly efficient
and selective catalysts, as bismuth molybdates, keeping the con-
trol of the stoichiometry and homogeneity of the prepared solids,
despite the presence of temperature sensitive metals like molybde-
num. SCS catalysts showed very good catalytic results in the partial
oxidation of propene to acrolein, particularly in terms of acrolein
selectivity. However, after a calcination treatment, which improved
their crystallinity, the propene conversion slightly increased but
the acrolein selectivity decreased. The calcination treatment, in
fact, induced a marked variation of the oxidation states of Mo and
Bi, leading to the formation of more stable oxides, with an enrich-
ment of Bi. Efforts must be put to better refining the SCS method,
investigating whether the amount of organic fuel used during the
SCS synthesis can be optimized, to promote a better crystallization
of the material without needing a calcination step. Highly selective
materials could then be prepared without affecting their activity
level and thus their acrolein yield. SCS also enables to synthesize
mixed oxides made of thermally sensitive metals, still keeping a
good control on their stoichiometry and phase composition.
[20] S. Ekambaram, K.C. Patil, M. Maaza, J. Alloys Compd. 393 (2005) 81–92.
[21] S. Specchia, A. Civera, G. Saracco, Chem. Eng. Sci. 59 (2004) 5091–5100.
[22] A. Varma, J.-P. Lebrat, Chem. Eng. Sci. 47 (1992) 2179–2180.
[23] M.T. Le, J. Van Craenenbroeck, I. Van Driessche, S. Hoste, Appl. Catal. A: Gen.
249 (2003) 355–364.
[24] B. Grzybowska, J. Haber, J. Komorek, J. Catal. 25 (1972) 25–32.
[25] M. Egashira, K. Matsuo, S. Kagawa, T. Seiyama, J. Catal. 58 (1979) 409–418.
[26] R.K. Grasselli, Top. Catal. 21 (2002) 79–88.
[27] M. Massa, A. Andersson, E. Finocchio, G. Busca, F. Lenrick, L.R. Wallenberg, J.
Catal. 297 (2013) 93–109.
[28] D.P. Debecker, V. Hulea, P. Hubert Mutin, Appl. Catal. A: Gen. 451 (2013)
192–206.
Acknowledgments
The authors gratefully acknowledge the ‘Fonds pour la forma-
tion à la Recherche dans l’Industrie et dans l’Agriculture (FRIA)’ of
Belgium for financial support and B. Farin’s PhD fellowship. Finally,
the institute is indebted to the ‘Communauté Franc¸ aise de Belgique’
for financial support through the ARC programme (08/13-009). The
authors thank A.G. Goiri Virto for the assistance in the lab during
the SCS of bismuth molybdates. The present research has been con-
ducted under the GDRI Catalyse network “Catalysis for polluting
emissions after treatment and production of renewables energies”
(2010–2014).
[29] S.R.G. Carrazán, C. Martín, V. Rives, R. Vidal, Appl. Catal. A: Gen. 135 (1996)
95–123.
[30] J. Wu, H. Yang, Y. Fan, B. Xu, Y. Chen, J. Fuel Chem. Technol. 35 (2007) 684–690.
[31] M.T. Le, W.J.M. Van Well, P. Stoltze, I. Van Driessche, S. Hoste, Appl. Catal. A:
Gen. 282 (2005) 189–194.
[32] J. Chul Jung, H. Kim, A.S. Choi, Y.-M. Chung, T.J. Kim, S.J. Lee, S.-H. Oh, I.K. Song,
J. Mol. Catal. A: Chem. 259 (2006) 166–170.
[33] K. Uchida, A. Ayame, Surf. Sci. 357–358 (1996) 170–175.
[34] A. Ayame, K. Uchida, M. Iwataya, M. Miyamoto, Appl. Catal. A: Gen. 227 (2002)
7–17.
[35] E. Godard, E.M. Gaigneaux, P. Ruiz, B. Delmon, Catal. Today 61 (2000)
279–285.
[36] I. Rossetti, L. Fabbrini, N. Ballarini, C. Oliva, F. Cavani, A. Cericola, B. Bonelli,
M. Piumetti, E. Garrone, H. Dyrbeck, E.A. Blekkan, L. Forni, J. Catal. 256 (2008)
45–61.
References
[37] W.J.M. van Well, M.T. Le, N.C. Schiodt, S. Hoste, P. Stoltze, J. Mol. Catal. A: Chem.
256 (2006) 1–8.
[1] M.T. Le, W.J.M. Van Well, I. Van Driessche, S. Hoste, Appl. Catal. A: Gen. 267
(2004) 227–234.
[38] Y. Cui, Y. Xia, J. Zhao, L. Li, T. Fu, N. Xue, L. Peng, X. Guo, W. Ding, Appl. Catal. A:
Gen. 482 (2014) 179–188.
Please cite this article in press as: B. Farin, et al., Bismuth molybdates prepared by solution combustion synthesis for the partial oxidation
of propene, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.03.045