Lewis acid property and catalytic performance of MoO3/SiO 2 for propylene epoxidation by CHP: Effects of precipitant pH value and rare earth additive

Article abstract of DOI:10.1016/j.molcata.2013.02.020

High dispersed 10% MoO₃/SiO₂ catalysts were prepared by the sol-gel method using a precipitant (ammonium hydroxide) with different pH values, and investigated by XRD, FT-IR spectroscopy of pyridine adsorbed and Raman spectroscopy techniques, and so on. The results show that the catalytic performance of MoO₃/SiO₂ for the epoxidation of propylene with cumene hydroperoxide (CHP) is affected by pH value of precipitant, and MoO₃/SiO₂ prepared with precipitant of pH 9 exhibits the highest yield of propylene oxide (PO). It has been found that the weak Lewis acidic sites on MoO₃/SiO₂ are the active sites of the propylene epoxidation with CHP, total amount of Lewis acid sites on the catalyst surface is related with the CHP conversion, and the weaker Lewis acid sites is in favor of the propylene epoxidation. When the amount of Lewis acid sites on the catalyst surface is more and their acid strength is higher, the CHP degradation and PO acid-catalytic hydrolysis would be speeded up, resulting in a reduction of the PO selectivity. The concentration and strength of the Lewis acid sites on MoO₃/SiO₂ are affected by pH of precipitant, and the catalyst prepared with precipitant of pH 9.0 possesses the most weakly Lewis acidic sites and the highest selectivity to PO (91.5%). Besides, the addition of certain amount of Nd can increase the weakly acidic sites to enhance CHP conversion and reduce the Lewis acidity of the catalyst thus suppress PO hydrolysis.

Full text of DOI:10.1016/j.molcata.2013.02.020

Journal of Molecular Catalysis A: Chemical 373 (2013) 78–84
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Lewis acid property and catalytic performance of MoO3/SiO2 for propylene
epoxidation by CHP: Effects of precipitant pH value and rare earth additive
Kai Shen1, Xiaohui Liu1, Guanzhong Lu∗, Yongxia Miao, Yanglong Guo, Yanqin Wang, Yun Guo
Key Laboratory for Advanced Materials, Research Institute of Industrial Catalysis, East China University of Science and Technology, Shanghai 200237, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
High dispersed 10% MoO3/SiO2 catalysts were prepared by the sol–gel method using a precipitant (ammo-
nium hydroxide) with different pH values, and investigated by XRD, FT-IR spectroscopy of pyridine
adsorbed and Raman spectroscopy techniques, and so on. The results show that the catalytic perfor-
mance of MoO3/SiO2 for the epoxidation of propylene with cumene hydroperoxide (CHP) is affected by
pH value of precipitant, and MoO3/SiO2 prepared with precipitant of pH 9 exhibits the highest yield of
propylene oxide (PO). It has been found that the weak Lewis acidic sites on MoO3/SiO2 are the active
sites of the propylene epoxidation with CHP, total amount of Lewis acid sites on the catalyst surface is
related with the CHP conversion, and the weaker Lewis acid sites is in favor of the propylene epoxidation.
When the amount of Lewis acid sites on the catalyst surface is more and their acid strength is higher,
the CHP degradation and PO acid-catalytic hydrolysis would be speeded up, resulting in a reduction of
the PO selectivity. The concentration and strength of the Lewis acid sites on MoO3/SiO2 are affected by
pH of precipitant, and the catalyst prepared with precipitant of pH 9.0 possesses the most weakly Lewis
acidic sites and the highest selectivity to PO (91.5%). Besides, the addition of certain amount of Nd can
increase the weakly acidic sites to enhance CHP conversion and reduce the Lewis acidity of the catalyst
thus suppress PO hydrolysis.
Received 4 December 2012
Received in revised form 22 February 2013
Accepted 22 February 2013
Available online 5 March 2013
Keywords:
Propylene epoxidation
Cumene hydroperoxide
MoO3/SiO2 catalyst
Lewis acid sites
Rare earth additive
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
amount of low value by-products. To be more environmentally
friendly, many studies have been made regarding the use of H2O2
Propylene oxide (PO) is an important chemical intermediate
[1]. It is used worldwide to produce versatile products including
polyether polyols, propylene glycols and propylene glycol ethers,
because it reacts readily with compounds containing active hydro-
gen atoms, such as alcohols, amines and acids. The Halcon method
is one of the main industrial process to manufacture propylene
oxide. However in the general Halcon process, the autoxidation
of ethylbenzene or isobutene, which is used to generate alkylhy-
droperoxide as the oxidant of the epoxidation, produces a large
[2,3] or O2 [4–6] as oxidants, which do not produce any by-product.
However, H2O2 is an expensive oxidant, and the catalytic epoxi-
dation of propylene with O2 usually gives poor selectivity for PO,
because of the high activity of the allylic C H in propylene. An
alternative route of propylene peroxidation, which does not involve
by-product, makes use of cumene hydroperoxide (CHP) as the oxi-
dant [7], in which the cumyl alcohol formed after epoxidation can
be recycled back to CHP after hydrogenolysis and oxidation. The
reaction route is shown in Eq. (1):
(1)
∗
Corresponding author. Tel.: +86 21 64253824; fax: +86 21 64253824.
E-mail address: gzhlu@ecust.edu.cn (G. Lu).
These authors contributed equally.
The main by-products of this route are phenol, acetone,
dimethylphenylcarbinol (DMPC, here refers to those produced by
1
1381-1169/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcata.2013.02.020

K. Shen et al. / Journal of Molecular Catalysis A: Chemical 373 (2013) 78–84
79
the decomposition rather than the epoxidation of CHP) and ace-
tophenone due to the decomposition of CHP, as shown in Eqs.
(2)–(4):
(2)
(3)
(4)
Meanwhile, the hydrolysis of propylene oxide may also occur as
Eq. (5).
2. Experimental
(5)
All chemicals were purchased from commercial sources
(Sinopharm Chemical Reagent Co. Ltd., except CHP, which from
SINOPEC) and used for the reaction without further purification.
The epoxidation of olefins with hydroperoxides can take
place effectively over group IVB, VB and VIB metals catalysts.
Molybdenum-based catalysts are one of the most widely used
epoxidation catalysts [8]. The molybdenum complexes, ranging
from monomeric molybdenum to molybdenum clusters and from
(NH4)6Mo7O24·4H2O to metallic molybdenum, can catalyze the
epoxidation of different kinds of alkenes with various peroxides
as oxygen donors [9]. To overcome the difficulty in separating and
recycling a catalyst from the reaction mixture, in recent years the
molybdenum complexes are often heterogenized onto different
supports by an immobilized [10,11], impregnation [12], grafting
[13] or sol–gel method [14,15]. The supports may be polymers
[16], carbon materials [17], silica [18–20] and so on. Among the
molybdenum-based catalysts, molybdenum oxide supported on
silica has attracted much attention for its high commercialized
potential as solid acid catalyst [21]. The studies reported on the
MoO3/SiO2 catalyst are mainly focused on molybdenum specials,
such as Mo or Mo peroxide species [22–24], Mo valence [25,26] and
so on. However, the effect of preparation process on the property of
acid sites, especially a strength distribution of acid sites on the cata-
lyst surface due to the different interactions between molybdenum
and silica were rarely reported. As a solid acid catalyst, MoO3/SiO2
has been used wide in nitration [15,27,28], esterification [29],
transesterification [30], acetalization [31], partial oxidation [32,33],
condensation [34] and double bond migration [35] besides epox-
idation of olefins [24]. Therefore, it is necessary to study the
influence factors on the acid sites of MoO3/SiO2 catalysts during
the preparation procedure.
2.1. Catalyst preparation and pretreatment
The 10% MoO3/SiO2 catalysts were prepared by a sol–gel
method. In a typical preparation, tetraethylorthosilicate (TEOS,
88 mmol) was dissolved in ethanol (20 mL) followed by adding
ammonium hydroxide aqueous solution (40 mmol, pH = 8.0–10.5)
under mechanical stirring. After stirring at 353 K for 30 min,
(NH4)6Mo7O24·4H2O aqueous solution (0.65 mmol, 40 mL) was
added in above synthesis solution. The formed gel was aged at room
temperature for 2 days, dried at 383 K for 24 h, and finally calcined
at 823 K for 5 h. The modified catalyst samples were similarly pre-
pared while adding the mixed solutions of nitrates (W, Cr, La, Ce
and Nd, respectively) and ammonium heptamolybdate.
The MoO3/SiO2 catalyst was pretreated by degassing at 773 K
under 1.0 × 10−7 Torr for 4 h and then exposing it to 0.2 Torr pyri-
dine for 10 min at room temperature. This catalyst was marked as
MoO3/SiO2–Py.
3. Propylene epoxidation
A mixture of tert-butanol (40 mL), CHP (48%, industrial grade,
20 mL) and the catalyst (200 mg, 0.14 mmol Mo) was placed into
an autoclave and 0.6 MPa propylene gas was charged at room
temperature. The reaction mixture was heated to 353 K under stir-
ring and kept at 353 K for 4 h. The products were analyzed by
a Clarus 500 Gas Chromatograph (PerkinElmer) equipped with a
flame ionization detector (FID) and AE.PEG-20 M separation col-
umn (30 m × 0.32 mm × 0.33 ␮m), and toluene was used as an
internal standard. The conversion of CHP was determined by an
iodometric titration. The conversion of CHP (XCHP), the selectivity
to PO (SCHP) for CHP and the yield of PO (YPO) are calculated as
follows:
Herein, high dispersed 10% MoO3/SiO2 catalysts were prepared
by sol–gel method and the epoxidation of propylene with CHP on
the MoO3/SiO2 catalysts was investigated. The surface Lewis acidity
of MoO3/SiO2 catalyst and an effect of surface Lewis acidity on its
catalytic performance for the epoxidation of propylene with CHP
were investigated by changing pH value of precipitant and adding
rare earth additives in the preparation process, which were investi-
gated by XRD, pyridine FT-IR and Raman spectroscopies, and so on.
The relationship between the catalytic performance for the epox-
idation and the characters of acid sites of the MoO3/SiO2 catalyst
was discussed.
niCHP − ntCHP
XCHP (%)
× 100%,
niCHP

80
K. Shen et al. / Journal of Molecular Catalysis A: Chemical 373 (2013) 78–84
Table 1
The catalytic performances of 10%MoO3/SiO2 catalysts for the propylene epoxidation by CHP.
Catalyst (pH of precipitant)
XCHP (%)
Selectivity (%)
PO
YPO (%)
1,2-Propanediol
Acetone
DMPCa
Acetophenone
8.0
8.5
9.0
9.5
10.0
10.5
83.7
82.5
86.5
84.4
86.9
87.0
77.7
84.9
91.5
80.9
85.2
82.5
1.3
1.0
0.5
1.2
0.7
1.1
9.3
5.2
2.8
7.3
5.0
6.4
10.6
8.0
4.3
9.7
8.3
9.2
0.9
0.8
0.8
0.7
0.7
0.6
65.1
70.0
79.1
68.2
74.0
71.8
Reaction conditions: 200 mg catalyst, 20 mL CHP, 40 mL tert-butanol at 353 K for 4 h.
a
The DMPC amount only results from thermal decomposition of CHP and excludes those formed in propylene epoxidation.
nPO
niCHP − ntCHP
exhibits increasing selectivity and yield of PO. Using the catalyst
prepared with ammonium hydroxide of pH 9.0, the highest PO
selectivity and PO yield can be obtained. Further increasing pH
of precipitant, the PO selectivity and PO yield over the catalyst
prepared are decreased on the contrary.
SCHP (%) =
× 100%,
YCHP (%) = (SCHP × XCHP) × 100%
where niCHP and nCt HP are the initial and final mole number of CHP
in the reaction mixture respectively, and nPO is the mole number
of PO produced.
The by-products in the propylene epoxidation with CHP include
1,2-propanediol, acetone, DMPC (dimethylphenylcarbinol, exclud-
ing those formed in propylene epoxidation) and acetophenone,
in which acetone, DMPC and acetophenone result mainly from
decomposition of CHP. As shown in Table 1, the amount of formed
1,2-propanediol is very few, suggesting that the hydrolysis of PO to
1,2-propanediol barely occurs. The catalyzed decomposition of CHP
produces acetone and DMPC, and the thermal decomposition of
CHP generates acetophenone [37]. It can be noted that the amount
of produced acetophenone is very few and almost same for the all
catalysts; the amounts of formed acetone and DMPC are different
for different catalysts, and for acetone and DMPC resulted from
catalyzed decomposition of CHP, their concentrations are very low
over the catalyst prepared with ammonium hydroxide of pH 9.0,
comparing with other catalysts.
3.1. Characterization of catalyst
The powder X-ray diffraction (XRD) patterns of all samples were
recorded on a Rigaku D-MAX RB X-ray diffractometer operated at
40 kV and 40 mA with CuK␣ radiation (ꢀ = 1.541 A). The nitrogen
˚
adsorption at low temperature was performed on a Quantachrome
Nova 4200e surface area and pore size analyser using static adsorp-
tion procedures, and the surface areas of catalysts were calculated
by the BET method. The FT-IR spectra were recorded on a Nicolet
NEXUS 670 FT-IR spectrometer, and the sample to be measured
was ground with KBr and pressed into thin wafer. Laser Raman
spectra (LRS) of samples were collected at ambient condition on an
HR LabRaman 800 system equipped with a CCD detector. A laser
beam (ꢀ = 514 nm) was used for an excitation.
4.2. Characterization of MoO3/SiO2 catalysts
The FT-IR spectra of pyridine (Py) adsorbed on sample were also
recorded on a Nicolet Nexus 670 FT-IR spectrometer. The sam-
ple was pressed into 20 mg/cm2 self-supporting wafer, and then
this wafer was placed in a quartz infrared cell equipped with CaF2
windows, which contained a heating device. The IR chamber was
evacuated to 1.0 × 10−7 Torr at 773 K for 4 h, and then the wafer
was cooled down to room temperature and exposed to pyridine
of 1.5 × 10−2 Torr for 10 min. The spectra were recorded when the
pre-treated sample was desorbed at different temperature. The
acid concentration was quantitatively evaluated according to the
method of Emeis [36].
The BET surface area of SiO2 prepared by this sol–gel method
is 452 m2/g, and the MoO3/SiO2 catalysts prepared behave smaller
surface area than the pure silica (Table 2), which is due to a deposi-
tion of the molybdenum species on the silica surface. Furthermore,
the effect of precipitant pH on the surface area of these MoO3/SiO2
catalysts is unobvious, and their BET surface areas are ∼430 m2/g
except the catalyst (pH 10.5) of 408 m2/g, which indicates that the
difference of catalytic performance between catalysts is not mainly
resulted from the diversity of their surface areas.
As shown in Fig. 1S, the XRD spectra of the MoO3/SiO2 catalysts
exhibit the characteristic peak of amorphous SiO2, and no definite
diffraction peaks of molybdenum species can be observed for these
10 wt% MoO3/SiO2 catalysts, indicating good dispersion of molyb-
denum oxide on SiO2 support. In contrast, monolayer coverage by
impregnation on SiO2 was not achieved at above 5% molybdenum
oxide, due to the low concentration of surface OH group and the
low reactivity between adjacent silanol groups to form siloxane
bridges [38]. The high dispersion of MoO3 on SiO2 in this work has
been obtained due to the high concentration of OH group on silicon
gel in the sol–gel process.
4. Results and discussion
4.1. Catalytic performance of MoO3/SiO2 for the propylene
epoxidation
The catalytic performances of the MoO3/SiO2 catalysts prepared
with ammonium hydroxide (precipitant) of different pH values
were tested for the epoxidation reaction of propylene by cumene
hydroperoxide (CHP), and the results are presented in Table 1. It
can be seen that the catalyst prepared with ammonium hydrox-
ide of pH 9.0 exhibits the best catalytic performance, and 79.1% PO
yield, 91.5% PO selectivity and 87.5% CHP conversion are obtained,
respectively.
For the CHP conversions, the MoO3/SiO2 catalysts prepared with
precipitant of different pH values are very similar to each other,
but their selectivity and yield of PO are various. With an increase in
pH of the ammonium hydroxide, the prepared MoO3/SiO2 catalyst
To further explore the structure of MoO3 on the SiO2 surface,
the FT-IR spectroscopy was used to test the surface properties of
SiO2, MoO3 and MoO3/SiO2 samples. As shown in Fig. 1, the FT-IR
spectra of all MoO3/SiO2 samples exhibit the absorption bands at
1300–1020, 800 and 467 cm−1 which are corresponding to asym-
metric stretching, symmetric stretching and bending modes of bulk
Si
O Si, respectively [39]. These IR spectra are similar to the IR
spectra of SiO2 (Fig. 1a) and indicate that the MoO3/SiO2 samples
have the similar surface chemical structures with SiO2.

K. Shen et al. / Journal of Molecular Catalysis A: Chemical 373 (2013) 78–84
81
Table 2
BET surface areas (SBET) and amounts of Lewis acid sites of MoO3/SiO2 catalysts.
Catalyst (pH of precipitant)
SBET (m2/g)
Amount of Lewis acid sitesa (mmol/g-cat)
373 K
573 K
673 K
8.0
8.5
9.0
9.5
10.0
10.5
439
424
437
432
425
408
0.411
0.404
0.423
0.412
0.415
0.434
0. 061
0.048
0.035
0.056
0.045
0.050
0.042
0.031
0.022
0.039
0.029
0.035
a
The amount of Lewis acid sites was calculated by the method of Emeis [36] according to the IR absorption peak area of pyridine adsorbed on MoO3/SiO2 catalyst after
desorption at different temperatures.
a
873
956
467
800
a
b
c
b
d
e
c
915
960
d
e
994
875
620
600
400
600
800
1000
1200
1400
1400
1200
1000
800
400
Raman Shift (cm-1
)
Wavenumbers
(
cm-1
)
Fig. 2. Raman spectra of MoO3/SiO2 catalysts prepared with ammonium hydroxide
of (a) pH = 8.0, (b) pH = 8.5, (c) pH = 9.0, (d) pH = 9.5, and (e) pH = 10.0.
Fig. 1. FT-IR spectra of SiO2 (a) (prepared by sol–gel method), MoO3/SiO2 prepared
with ammonium hydroxide of (b) pH = 8.0, (c) pH = 9.0, (d) pH = 10.0, and (e) MoO3
(calcined by ammonium heptamolybdate).
octahedrally coordinated molybdenum species exist on the sur-
face of MoO3/SiO2 catalysts rather than tetrahedral coordinated
molybdenum species.
In the IR spectrum of pure bulk MoO3, there are a band at
994 cm−1 and two broad bands at 875 and 620 cm−1 corresponding
to the Mo
ric stretching vibration and Mo
O
group of MoO3 phase, Mo
O Mo asymmet-
4.3. FT-IR spectroscopy of adsorbed pyridine
O
Mo symmetric stretching
vibration, respectively. These three IR absorption bands do not
appear in the IR spectra of 10%MoO3/SiO2 catalysts, which indicates
that bulk MoO3 phase is not present in the prepared MoO3/SiO2
catalyst, which is consistent with the XRD results.
To evaluate the type and strength of acid sites in MoO3/SiO2 cat-
alysts, the FT-IR spectra of pyridine adsorbed on the catalysts were
measured and the results are shown in Fig. 3. As it has been reported
The bands at 915 and 960 cm−1 in IR spectra of the MoO3/SiO2
samples are attributed to Mo
groups of surface molybdate phase [40]. The intensity of the band
at 915 cm−1 (Mo
Si) is clearly varied with the pH value of
O Si vibration and terminal Mo O
pH 8.0
O
B
1540
L
1451
L
1611
ammonium hydroxide, indicating that the interaction between Mo
species and silica surface is affected by the pH value of synthesis
solution in the sol–gel process. The IR absorption band at 915 cm−1
on the surface of the catalyst (pH 9.0) has the highest intensity, that
is, there is the strongest interaction between MoO3 and SiO2 on the
catalyst (pH 9.0).
pH 8.5
pH 9.0
pH 9.5
As shown in Fig. 2, the bands at 996, 821 and 667 cm−1 of
crystalline MoO3 are absent in the Raman spectra of MoO3/SiO2
catalysts, and the bands at ∼873 and ∼956 cm−1 of surface
polymolybdate species are observed remarkably. The broad
pH 10.0
pH 10.5
band at 873 cm−1 is assigned to the Mo
O
Mo stretching mode
of polymerized three-dimensional surface molybdenum oxide
4−
6−
[29], and the band at
species, such as Mo8O26
or Mo7O24
2000
1800
1600
1400
956 cm−1 is attributed to the terminal Mo O Raman stretching
Wavenumbers ( )
cm-1
vibration of various octahedrally coordinated molybdenum oxide
4−
species (such as Mo8O26
tetrahedral molybdenum species [41–43]. This indicates that only
or Mo7O246−) rather than isolated
Fig. 3. FT-IR spectra of pyridine adsorbed on MoO3/SiO2 catalysts after desorption
at 373 K. L: Lewis acid sites; B: Brønsted acid sites.

82
K. Shen et al. / Journal of Molecular Catalysis A: Chemical 373 (2013) 78–84
a
L
1451
b
B
1540
L
1451
B
1540
L
1611
L
1611
373K desorption
373K desorption
573K desorption
673K desorption
573K desorption
673K desorption
2000
1800
1600
1400
2000
1800
1600
1400
cm-1
)
Wavenumbers
(
cm-1
)
Wavenumbers
(
Fig. 4. FT-IR spectra of pyridine adsorbed on MoO3/SiO2 catalysts prepared with ammonium hydroxide of (a) pH 8.0 and (b) pH 9.0 after desorption at 373–673 K. L: Lewis
acid sites; B: Brønsted acid sites.
[44,45], the absorption bands at about 1451 and 1611 cm−1 should
be attributed to the absorption frequencies of pyridine interac-
ting with the Lewis acid sites, and the bands at about 1540 and
1634 cm−1 should be attributed to the absorption frequencies of
pyridine coordinated in the Brønsted acid sites. The results in Fig. 3
show that only a band at 1451 cm−1 appears on the MoO3/SiO2
catalysts after desorption at 373 K, suggesting that only Lewis acid
sites exist on the MoO3/SiO2 catalysts and there are no Brønsted
acid sites or its acid property is very weak.
After pyridine adsorbed on the catalyst (pH = 8.0 and 9.0) was
desorbed at 373 K, 573 K and 673 K, their FT-IR spectra are shown
in Fig. 4, and the concentrations of Lewis acid sites on the all cata-
lysts were calculated by the Emeis method based on the bands at
1451 cm−1 [36] and are listed in Table 2. With increasing desorp-
tion temperature, the amount of pyridine adsorbed on Lewis acid
sites would decrease, because pyridine adsorbed on weaker Lewis
acid sites would be desorbed at higher temperature; only pyri-
dine adsorbed on relatively stronger Lewis acid sites can be kept.
The results in Table 2 show that, pH of ammonium hydroxide has
an influence on the strength distribution of the acid sites on the
MoO3/SiO2 catalysts.
The results in Table 2 show that, with an increase in pH of ammo-
nium hydroxide, the initial (or total) amount of Lewis acid sites
on the MoO3/SiO2 catalyst is almost alike, but the acid amount
has been varied after desorption at different temperatures, which
shows that the precipitant pH value can affect the acid strength of
surface acid sites of the MoO3/SiO2 catalyst. Among all catalysts, the
catalyst (pH 9.0) possesses the lowest acid amount (0.022 mmol/g-
cat) after desorption at 673 K, that is, its Lewis acid strength is
the weakest. The results in Table 1 show that the catalyst (pH 9.0)
behaves the best performance, 91.5% PO selectivity and 79.1% PO
yield. Relating the catalytic performance of the MoO3/SiO2 catalyst
(Table 1) and its Lewis acid property (Table 2), it can be concluded
that weaker Lewis acid sites on MoO3/SiO2 catalyst is in favor of
the propylene epoxidation with CHP.
The data in Table 1 show also that, for the products (acetone
and DMPC) resulted from the catalyzed decomposition of CHP, the
concentrations over the catalyst (pH 9) are the lowest among all
catalysts, which is agreement with the situation that the catalyst
(pH 9) possesses the weakest Lewis acid sites (Table 2). It was
reported that both Lewis acid strength and Lewis acid site density
are important properties for the catalyst used in the CHP decom-
position [46]. Because CHP is a Lewis base, the MoO3/SiO2 catalyst
as a Lewis acid catalyze the CHP decomposition. Stronger acidity of
the Lewis acid sites can accelerate CHP decomposition [47]. The PO
selectivity can be improved when the stronger Lewis acidic sites
are obviously reduced, such as the catalyst (pH 9). The increase in
the concentration of strong Lewis acid sites promotes the catalyzed
decomposition of CHP, which forms more acetone and DMPC by-
products, and also promotes the acid-catalyzed hydrolysis of PO to
form the 1,2-propanediol.
The results above show that, the MoO3/SiO2 catalysts prepared
with ammonium hydroxide of different pH values behave different
concentration distribution and strength of their Lewis acid sites
on the surface, resulting in different catalytic performances for the
propylene epoxidation with CHP.
4.4. Effect of additive on the performance of MoO3/SiO2 catalyst
To modify the surface Lewis acidity to improve the catalytic
performance of the MoO3/SiO2 catalyst, VIB elements (W, Cr) and
rare earth elements (e.g., La, Ce, Nd) were used as a promoter, and
the catalytic properties of the modified catalysts are listed in Table
1S. The results show that the catalytic property of the MoO3/SiO2
catalyst (pH 9) can be improved after adding Nd. Using other
elements as an additive to modify this catalyst, the selectivity to PO
is reduced on the contrary. The effect of Nd loading on the catalytic
performance of this catalyst was studied. It has been found that the
optimal loading of Nd is ∼0.5 wt%. Using 0.5 wt%Nd–MoO3/SiO2
(pH 9) catalyst, 91.4% CHP conversion and 91.1% PO selectivity can
be obtained. When the Nd amount exceeds 1 wt% in this catalyst,
its catalytic property is severely deteriorate (Table 3).
The results in Table 3 show that, the moderate addition of Nd in
MoO3/SiO2 catalyst affects hardly the PO selectivity, can increase
the conversion of CHP and inhibit the PO hydrolysis, that is, 1,2-
propanediol cannot be found in the products; and that acetone and
DMPC from the CHP decomposition are almost unchanged. If the
Nd amount in the catalyst is >0.7 wt%, the concentrations of ace-
tone and DMPC in products are increased, which indicates that the
surface Nd species may promote the CHP decomposition, result-
ing in no increase in PO selectivity. For acetophenone by-product,
the addition of Nd has no obvious effect on its amount, because it
results from thermal decomposition of CHP.
The Lewis acid concentrations of the Nd–MoO3/SiO2 catalysts
were also measured by in situ FT-IR of adsorbed pyridine, and the
results are shown in Table 3. As seen in Table 3, the presence of Nd in
the MoO3/SiO2 catalyst does not vary the Lewis acid type on the sur-
face of catalyst, but can affect the acid amount and acid strength of
the catalyst. Comparing with unmodified MoO3/SiO2, the stronger
Lewis acidic sites of Nd–MoO3/SiO2 is obviously reduced. When the
Nd content is <0.7 wt%, the stronger Lewis acidic sites is gradually
decreased with an increase in the Nd amount; when the Nd content

K. Shen et al. / Journal of Molecular Catalysis A: Chemical 373 (2013) 78–84
83
Table 3
Amounts of Lewis acid sites and catalytic performances of Nd–MoO3/SiO2 catalysts for the propylene epoxidation.
Nd (wt%)
XCHP (%)
Selectivity (%)
YPO (%)
AmountofL-acidsites(mmol/g-cat)b
PO
1,2-Propanediol
Acetone
DMPCa
Acetophenone
373 Kc
573 K
673 K
–
86.5
89.8
91.4
86.7
72.5
66.2
91.5
91.0
91.1
87.0
90.1
84.4
0.5
–
–
–
–
2.8
2.8
2.6
3.9
2.9
4.2
4.3
5.1
5.3
7.8
5.9
0.8
0.9
0.8
1.1
1.0
1.1
79.1
81.7
83.3
75.4
65.3
55.9
0.423
0.440
0.457
0.425
0.387
0.342
0.035
0.026
0.023
0.029
0.019
0.016
0.022
0.013
0.011
0.015
0.010
0.008
0.3
0.5
0.7
1.0
1.5
–
10.2
Reaction conditions: 200 mg catalyst, 20 mL CHP, 40 mL tert-butanol, at 353 K for 4 h.
a
The DMPC amount only results from thermal decomposition of CHP and excludes those formed in propylene epoxidation.
The amount of Lewis acid sites was calculated by the method of Emeis [36] according to the IR absorption peak area of pyridine adsorbed on Nd–MoO3/SiO2 catalyst after
b
desorption at different temperatures.
is >0.7 wt%, total acid sites and the stronger Lewis acidic sites are
all decreased with an increase of Nd amount.
Relating the catalytic performance and Lewis acidity of
Nd–MoO3/SiO2 catalyst, we can see, the presence of Nd in the
catalyst decreases the acid strength of the catalyst and increases
amount of weak acid sites on the catalyst surface, inhibiting the PO
hydrolysis thus largely eliminate 1,2-propanediol. When the Nd
content exceeds 0.7 wt%, a decrease in total acid sites makes the
CHP conversion notably fall, and a decrease in stronger Lewis acidic
sites should improve the PO selectivity, but in fact the PO selectivity
still decrease instead, which is possible that the presence of higher
Nd loading promote CHP degradation to produce more acetone and
DMPC.
Fig. 5. Possible molecular structure of MoOx cluster models: distorted tetrahedral
coordination (I), distorted tetragonal pyramidal coordination (II) and distorted octa-
hedral coordination (III and IV) [63].
4.5. Discussion of the active sites and effect of precipitant pH
value
and the pH value of the hydrolyzing environment (as precipitant)
directly affects the silica structure. It is supposed that more highly
cross-linked gels and short silica chain are obtained when ammo-
nium hydroxide of pH 9.0 was used. Therefore, the MoO3/SiO2
catalyst prepared with ammonium hydroxide of pH 9.0 may contain
more short chain molybdenum species (such as Mo (a) in model III)
than other catalysts in the preparation of the MoO3/SiO2 catalyst.
The FT-IR spectra (Fig. 2) show that there is the strongest interac-
tion between MoO3 and SiO2 on the catalyst (pH 9.0), resulting in
a formation of more short chain molybdenum species like Mo (a)
after calcination. This is the reason that the 10wt%MoO3/SiO2 cat-
alyst (pH 9.0) behaves the excellent catalytic performance for the
epoxidation of propylene with CHP.
It has been proved that Mo-based catalysts possess Lewis acid
sites in their highest oxidation state d0. These Lewis acid sites
have a low redox potential and are labile to accept ligands. Olefin
epoxidation by d0 metal peroxo complexes is generally considered
to proceed through an electrophilic attack of the oxygen on the
electron-rich olefin.
In order to confirm that the Lewis acid sites serve as the active
sites in the epoxidation of propylene, the catalytic activity of
MoO3/SiO2–Py pretreated with pyridine was studied. As it can be
seen from Table 2S, after 2 mL pyridine was added in the reaction
the conversion of CHP over MoO3/SiO2 is only decreased from 86%
to 72%, but using MoO3/SiO2–Py catalyst the conversion of CHP falls
down to 4%, which is similar to the catalytic performance of SiO2.
This result indicates that the Lewis acid sites on the catalyst are the
active sites to catalyze propylene epoxidation. The decrease in per-
formance of MoO3/SiO2–Py catalyst is due to pyridine on adsorbed
on the catalyst blocking the Lewis acid active sites.
There may be three kinds of molybdate species in the
MoO3-based catalyst, i.e., tetrahedral species, tetragonal species,
octahedral species and polymeric MoOx species. The models of
the possible surface molybdenum species can be demonstrated in
Fig. 5. The models I and II represent the tetrahedral and tetragonal
coordinated molybdenum species, respectively. The models III and
IV represent the distorted octahedral configurations with penta-
coordinated (MoO5) species [48]. The Raman spectra reveal that
only octahedral molybdenum species exist on the MoO3/SiO2 cat-
alysts, which can be represented by model III and IV. Based on the
structural models and Lewis acidity properties of the MoO3/SiO2
catalysts, we can conclude that the Lewis acid sites in the octahedral
molybdenum species are the active catalytic sites.
5. Conclusions
In summary, the high effective 10 wt% MoO3/SiO2 catalysts for
the propylene epoxidation with CHP were successfully prepared
by the sol–gel method using ammonium hydroxide as a precipi-
tant. The results show that there are mainly Lewis acid sites on
the surface of the MoO3/SiO2 catalyst, and these Lewis acid sites
consisted of the octahedral molybdenum species are the active
sites of propylene epoxidation. The concentration distribution and
strength of Lewis acid sites are related to the catalyst performance
of MoO3/SiO2 catalyst, and the total amount of Lewis acid on the
catalyst surface is related with the CHP conversion and the weaker
Lewis acid sites on MoO3/SiO2 catalyst is in favor of the propylene
epoxidation with CHP. When the amount of Lewis acid sites on the
catalyst surface is more and their acid strength is higher, the CHP
degradation and PO acid-catalytic hydrolysis would be speeded up,
resulting in a reduction of the selectivity to PO.
The concentration and strength of the Lewis acid sites on the
MoO3/SiO2 catalyst are affected by pH of precipitant, and the cata-
lyst prepared with precipitant of pH 9.0 possesses the most weakly
Lewis acidic sites on the surface, resulting in the highest selectiv-
ity to PO (91.5%). The addition of Nd in the MoO3/SiO2 catalyst can
In the octahedral molybdenum species in Fig. 5, the Mo (b) has
stronger Lewis acidity than Mo (a) because of its lower electronic
density. The hydrolysis and condensation of TEOS produce three-
dimensional structure network of SiO4 at alkaline condition [49],

84
K. Shen et al. / Journal of Molecular Catalysis A: Chemical 373 (2013) 78–84
increase the weakly acidic sites to enhance CHP conversion and
reduce strongly acidic sites to inhibit PO acid-catalytic hydrolysis.
Using 0.5 wt% Nd-10 wt%MoO3/SiO2 catalyst, 91.4% CHP conversion
and 91.1% PO selectivity can be obtained.
[15] X.M. Ma, B.D. Li, C.X. Lv, M. Lu, J. Wu, L.J. Liang, Catal. Lett. 141 (2011) 1814–1820.
[16] M. Afsharpour, A.R. Mahjoub, M.M. Amini, Appl. Catal. A: Gen. 327 (2007)
205–210.
[17] M. Moghadam, S. Tangestaninejad, V. Mirkhani, J. Mol. Catal. A: Chem. 329
(2010) 44–49.
[18] N. Ohler, A.T. Bell, J. Catal. 231 (2005) 115–130.
[19] M. Masteri-Farahani, F. Farzaneh, M. Ghandi, J. Mol. Catal. A: Chem. 243 (2006)
170–175.
Author contributions
[20] P. Ferreira, I.S. Goncalves, F.E. Kühn, A.D. Lopes, M.A. Martins, M. Pillinger, A.
Pina, J. Rocha, C.C. Romao, A.M. Santos, T.M. Santos, A.A. Valente, Eur. J. Inorg.
Chem. 10 (2000) 2263–2270.
[21] A.P. Amrute, S. Sahoo, A. Bordoloi, Y.K. Hwang, J.S. Hwang, S.B. Halligudi, Catal.
Commun. 10 (2009) 1404–1409.
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of the
manuscript.
[22] J.P. Thielemann, C. Hess, J. Catal. 288 (2012) 124–126.
[23] U. Neuenschwander, E. Meier, I. Hermans, Chem. Eur. J. 18 (2012) 6776–6780.
[24] Y.X. Miao, G.Z. Lu, X.H. Liu, Y.L. Guo, Y.Q. Wang, Y. Guo, J. Mol. Catal. A: Chem.
306 (2009) 17–22.
Acknowledgements
[25] F. Adam, A. Iqbal, Microporous Mesoporous Mater. 141 (2011) 119–127.
[26] N.U. Silva, C.I. Fernandes, T.G. Nunes, M.S. Saraiva, C.D. Nunes, P.D. Vaz, Appl.
Catal. A: Gen. 408 (2011) 105–116.
[27] S.B. Umbarkar, A.V. Biradar, S.M. Mathew, S.B. Shelke, K.M. Malshe, P.T. Patil,
S.P. Dagde, S.P. Niphadkar, M.K. Dongare, Green Chem. 8 (2006) 488–493.
[28] S.M. Mathew, A.V. Biradar, S.B. Umbarkar, M.K. Dongare, Catal. Commun. 7
(2006) 394–398.
[29] A.E.A. Said, M.M.M. Abd El-Wahab, J. Chem. Technol. Biot. 81 (2006) 329–335.
[30] Y.T. Kim, E.D. Park, Appl. Catal. A: Gen. 356 (2009) 211–215.
[31] S.B. Umbarkar, T.V. Kotbagi, A.V. Biradar, R. Pasricha, J. Chanale, M.K. Don-
gare, A.S. Mamede, C. Lancelot, E. Payen, J. Mol. Catal. A: Chem. 310 (2009)
150–158.
This project was supported financially by the National Basic
Research Program of China (2010CB732300), the Fundamental
Research Funds for the Central Universities, the “Shu Guang” Project
(10GG23) and Leading Academic Discipline Project (J51503) of
Shanghai Municipal Education Commission and Shanghai Educa-
tion Development Foundation.
Appendix A. Supplementary data
[32] G.S. Li, D.H. Hu, G. Xia, Z.C. Zhang, Top Catal. 52 (2009) 1381–1387.
[33] H.Y. Zhao, S. Bennici, J.Y. Shen, A. Auroux, Appl. Catal. A: Gen. 385 (2010)
224–231.
[34] S.V. Bhosale, M.B. Kalyankar, S.V. Nalage, D.S. Bhosale, S.L. Pandhare, T.V. Kot-
bagi, S.B. Umbarkar, M.K. Dongare, Synthetic Commun. 41 (2011) 762–769.
[35] Y.K. Park, S.J. Kim, N. You, J. Cho, S.J. Lee, J.H. Lee, J.K. Jeon, Ind. Eng. Chem. 17
(2011) 186–190.
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.molcata.
2013.02.020.
References
[36] C.A. Emeis, J. Catal. 141 (1993) 347–354.
[1] T.A. Nijhuis, M. Makkee, J.A. Moulijn, B.M. Weckhuysen, Ind. Eng. Chem. Res. 45
(2006) 3447–3459.
[2] Z.W. Xi, N. Zhou, Y. Sun, K.L. Li, Science 292 (2001) 1139–1141.
[3] Y. Zuo, X.S. Wang, X.W. Guo, Ind. Eng. Chem. Res. 50 (2011) 8485–8491.
[4] G.Z. Lu, X.B. Zuo, Catal. Lett. 58 (1999) 67–70.
[5] Y. Lei, F. Mehmood, S. Lee, J. Greeley, B. Lee, S. Seifert, R.E. Winans, J.W. Elam,
R.J. Meyer, P.C. Redfern, D. Teschner, R. Schlogl, M.J. Pellin, L.A. Curtiss, S. Vajda,
Science 328 (2010) 224–228.
[6] A. Seubsai, M. Kahn, S. Senkan, ChemCatChem 3 (2011) 174–179.
[7] K.T. Li, P.H. Lin, S.W. Lin, Appl. Catal. A: Gen. 301 (2006) 59–65.
[8] K.A. Jørgensen, Chem. Rev. 89 (1989) 431–458.
[9] F.E. Kuhn, A.M. Santos, M. Abrantes, Chem. Rev. 106 (2006) 2455–2475.
[10] M. Farias, M. Martinelli, G.K. Rolim, Appl. Catal. A: Gen. 403 (2011) 119–127.
[11] M. Masteri-Farahani, N.J. Tayyebi, Mol. Catal. A: Chem. 348 (2011) 83–87.
[12] P.C. Bakala, E. Briot, J.Y. Piquemal, J.M. Bregeault, P. Beaunier, Catal. Commun.
8 (2007) 1447–1451.
[13] P.C. Bakala, E. Briot, L. Salles, J.M. Bregeault, Appl. Catal. A: Gen. 300 (2006)
91–99.
[14] H. Vrubel, K.J. Ciuffi, G.P. Ricci, F.S. Nunes, S. Nakagaki, Appl. Catal. A: Gen. 368
(2009) 139–145.
[37] M.S. Kharasch, A. Fono, W. Nudenberg, J. Org. Chem. 16 (1951) 113–127.
[38] H.C. Hu, I.E. Wachs, J. Phys. Chem. 99 (1995) 10897–10910.
[39] Y. Liu, X.B. Ma, S.P. Wang, J.L. Gong, Appl. Catal. B: Environ. 77 (2007) 125–134.
[40] M.A. Vuurman, I.E. Wachs, J. Phys. Chem. 96 (1992) 5008–5016.
[41] H.Y. Zhu, M.M. Shen, Y. Wu, X.W. Li, J.M. Hong, B. Liu, X.L. Wu, L. Dong, Y. Chen,
J. Phys Chem. B 109 (2005) 11720–11726.
[42] Z.X. Song, N. Mimura, J.J. Bravo-Suárez, T. Akita, S. Tsubota, S.T. Oyama, Appl.
Catal. A: Gen. 316 (2007) 142–151.
[43] M.A. Ban˜ares, I.E. Wachs, J. Raman. Spectrosc. 33 (2002) 359–380.
[44] T. Barzetti, E. Selli, D. Moscotti, L. Forni, J. Chem. Soc., Faraday Trans. 92 (1996)
1401–1407.
[45] T.R. Hughes, H.M. White, J. Phys. Chem. 71 (1967) 2192–2201.
[46] M.E. Levin, N.O. Gozales, L.W. Zimmerman, J. Yang, J. Hazard. Mater. 130 (2006)
88–106.
[47] K.Y. Koltunov, V.I. Sobolev, Appl. Catal. A: Gen. 336 (2008) 29–34.
[48] G. Tsilomelekis, A. Christodoulakis, S. Boghosian, Catal. Today 127 (2007)
139–147.
[49] J.D. Wright, N.A.J.M. Sommerdijk, Sol–Gel Materials Chemistry and Applica-
tions, Taylor & Francis, London, 2002, pp. 19–22.