2
Y. He et al. / Journal of Molecular Catalysis A: Chemical 331 (2010) 1–6
commercially and used without further purification. SiO2 sup-
ported MoBiOx, MoTeOx, TeOx, and MoOx catalysts, which had
the same molar ratio (6 mol%) of the loading element to sili-
con (Si), were prepared via the impregnating method. Taking
the MoBiOx/SiO2 ([nMo + nBi]/nSi = 6 mol%) catalysts as an exam-
ple, solutions of (NH4)6Mo7O24 and Bi(NO3)3 were first mixed. An
appropriate amount of SiO2 was then added, and the resulting mix-
ture was allowed to impregnate at room temperature for 5 h. The
water in the solution was evaporated at a temperature of 80 ◦C.
After drying at 100 ◦C for 12 h, the sample was calcined at 600 ◦C for
4 h and then cooled to room temperature. The catalyst was obtained
as the final product.
SiO2 supported MoBiTeOx mixed oxide catalysts were prepared
via a two-step impregnating method. First, stoichiometric amounts
of bismuth and molybdenum ([nMo + nBi + nTe]/nSi = 6 mol%) were
loaded onto SiO2. After calcination at 600 ◦C for 2 h in air, the SiO2
mixture was then impregnated with stoichiometric amounts of Te,
dried at 100 ◦C, and calcined at 600 ◦C for 4 h. The catalyst was
obtained as the final product.
3. Results and discussion
3.1. Catalytic performance
The catalytic performance of the catalysts in selective oxidation
of propane to acrolein is shown in Table 1. No reaction occurred
over quartz sand, indicating that the contribution of homogeneous
reaction could be ignored. The MoOx/SiO2 catalyst showed the
highest activity, but the propane was mainly converted to COx, and
the selectivity toward acrolein was low. Bismuth and Te doping to
MoOx/SiO2 catalyst promoted acrolein selectivity at the expense of
propylene and COx selectivity. By two-step Bi and Te optimization,
Mo1Bi0.05Te0.05Ox/SiO2 and Mo1Bi0.05Te0.1Ox/SiO2 catalysts were
observed to show the highest catalytic performance. About 5.0%
acrolein yield was obtained at 550 ◦C, C3H8/O2/N2 = 1.2/1/4. How-
ever, MoBiTeOx/SiO2 catalyst activity was not high, which might
be due to the limited ability of molybdenum oxide to activate
propane. To obtain high performance, propane homogeneous
reaction is needed for the MoBiTeOx/SiO2 catalyst. Under optimum
reaction conditions (530 ◦C, C3H8/O2/N2 = 1.2/1/4, and bigger
quartz tube reactor [i.d. = 10 mm]), the Mo1Bi0.05Te0.05Ox/SiO2
2.2. Catalytic testing
and Mo1Bi0.05Te0.1Ox/SiO2
alytic performance. Moreover, about 14.0% acrolein yield
(STYACR = 487 g kgCat−1 h−1) was obtained with the help of homo-
geneous reaction. The catalytic performance of MoTeOx/SiO2
catalysts is also listed in Table 1. These catalysts were found
to show lower acrolein selectivity and yield compared with
MoBiTeOx/SiO2 catalyst. The Bi component is clearly needed to
obtain higher catalytic performance, although Bi concentration
in the MoBiTeOx/SiO2 catalyst is very low. Numerous researchers
reported that Te plays the same role as Bi in propane partial
oxidation [16]. However, in MoBiTeOx/SiO2 catalysts, Bi and Te
appeared to play different roles. The role of the Bi component is
thus worth studying.
Catalytic testing was performed in a tubular fixed bed flow
quartz reactor (inside diameter [i.d.] = 6 mm) with 100 mg cata-
lyst under atmospheric pressure, gaseous hourly space velocity
(GHSV) = 7200 mL (g-cat)−1 h−1, and C3H8/O2/N2 = 1.2:1:4. Propane
oxidation reaction was performed at 550 ◦C. The reactants and
products were analyzed using two on-line gas chromatography
detectors with three columns. A TDX-601 column and an Al2O3
column impregnated with squalane (GC-950, TCD) were used for
the separation of C3H8, C3H6, C2H4, CO, and CO2. A GDX-103 col-
umn (GC-950, FID) was used for the acrolein, acetone, and propanal
separation. Exitgases wereheatedto120 ◦Ctoprevent productcon-
densation. The catalysts were allowed to equilibrate under reaction
conditions for at least 30 min.
3.2. Effect of Bi component on the catalyst specific surface area
and structure
The oxidation of propylene was performed with pulse method
at 400 ◦C. Pulse of 100 L propylene was introduced into a stream
of N2 carrier gas passing through the catalyst bed at a flow rate of
20 mL/min. The products were analyzed by gas chromatography.
The BET surface areas of catalysts are shown in Table 2. The spe-
cific surface area of the amorphous silica sample equals 277 m2 g−1
as shown by N2 adsorption. The loading of Mo, Bi and Te decreased
the BET surface area. With the increase of Te content, the spe-
decreased, which might be one of the reasons for their decreased
activities. The MoBiTeOx/SiO2 catalyst has larger specific surface
areas than MoTeOx/SiO2 catalyst, indicating that Bi doping could
2.3. Characterization of catalysts
N2 adsorption and desorption isotherms were recorded on an
automated Micromeritics Tri-Star3000 apparatus at LNT. X-ray
diffraction (XRD) analysis of the catalysts was carried out on a
PANalytical X’Pert Pro diffractometer using Cu Ka (ꢀ = 0.15406 nm)
radiation (40 kV/40 mA).
Raman spectra were recorded on a Renishaw-UV–vis Raman
System 1000 spectrometer with a CCD detector. Raman excitation
at 325 nm provided by the He–Cd laser source was used for exci-
tation, and the laser power applied on samples was 7 mW. In situ
Raman experiments were also carried out on the same Raman sys-
tem using a home built high temperature in situ Raman cell [15].
Powder samples were pressed into the sample holder equipped
with a thermocouple placed underneath for temperature control.
In order to avoid the temperature effect all in situ Raman spectra
were collected after the cell temperature was cooled down to room
temperature.
H2-TPR experiment was carried out using a temperature-
programmed reaction-TCD instrument. The sample (ca. 20 mg) was
exposed to a 20 mL/min 5% H2/Ar flow, and heated at a rate of
10 ◦C/min.
X-ray photoelectron spectroscopy (XPS) measurements were
performed with a Quantum 2000 Scanning ESCA Microprobe
instrument using AlK␣. The C 1s signal was set to a position of
284.6 eV. The measurement error is 0.2 eV.
Table 2 also provides the XPS results of catalysts. The binding
BE of MoO3 was 232.6 eV, while Mo4+ cations were observed at
229.6 eV [17]. These changes in BE indicate that Mo6+ was present
on the surface of these catalysts. In the same way, the Bi and Te
valence state of the catalysts was determined to be +3 and +4,
respectively, based on the literature reported [18,19]. By calcu-
lating the peak area of their core-level spectra, the molar ratios
of Mo/Si, Bi/Si, and Te/Si can be obtained on the catalyst surface.
the MoBiTeOx/SiO2 catalyst is very low. These results are consistent
with their theoretical value.
The catalyst powder X-ray diffraction patterns are shown in
Fig. 1. The wide hump between 15◦ and 35◦ could be assigned to
amorphous SiO2. In most of the catalysts, the presence of MoO3
was characterized by reflections at 2ꢁ = 12.7◦, 23.3◦, 25.7◦, 27.3◦,
and 38.9◦ (JCPDS 05-0508). With the increase of Te content, peaks
of the MoO3 phase weakened, indicating that the Te component