Metathesis of 1-Butene to Propene over Mo/Al2O3@SBA-15: Influence of Alumina Introduction Methods on Catalytic Performance

Article abstract of DOI:10.1002/asia.201500351

A series of Mo-based catalysts for 1-butene metathesis to propene were prepared by supporting Mo species on SBA-15 premodified with alumina. The effects of the method of introduction of the alumina guest to the host SBA-15 on the location of the Mo species and the corresponding metathesis activity were studied. As revealed by N₂ adsorption isotherms and TEM results, well-dispersed alumina was formed on the pore walls of SBA-15 if the ammonia/water vapor induced hydrolysis (NIH) method was employed. The Mo species preferentially interacted with alumina instead of SBA-15, as evidenced by X-ray photoelectron spectroscopy, time-of-flight secondary-ion mass spectrometry, and IR spectroscopy of adsorbed pyridine. Furthermore, new Bronsted acid sites favorable for the dispersion of the Mo species and lowerature metathesis activity were generated as a result of the effective synergy between the alumina and SBA-15. The Mo/Al₂O₃@SBA-15 catalyst prepared by the NIH method showed higher metathesis activity and stability under the conditions of 120°C, 0.1 MPa, and 1.5 h^-1 than catalysts prepared by other methods. Let me introduce myself: Mo-based catalysts for the metathesis of 1-butene to propene are prepared by supporting Mo species on SBA-15 premodified with alumina. The effects of the method of introduction of the alumina guest to the SBA-15 host on the location of the Mo species and the corresponding metathesis activity are discussed in detail. The Mo-based catalyst with uniformly coated alumina exhibits the best performance in the metathesis reaction.

Full text of DOI:10.1002/asia.201500351

DOI: 10.1002/asia.201500351
Full Paper
Metathesis
Metathesis of 1-Butene to Propene over Mo/Al O @SBA-15:
2
3
Influence of Alumina Introduction Methods on Catalytic
Performance
[a, b]
[a]
[a]
[a]
[a]
Dazhou Zhang,
Longya Xu*
Xiujie Li,* Shenglin Liu, Xiangxue Zhu, Fucun Chen, and
[a]
Abstract: A series of Mo-based catalysts for 1-butene meta-
thesis to propene were prepared by supporting Mo species
on SBA-15 premodified with alumina. The effects of the
method of introduction of the alumina guest to the host
SBA-15 on the location of the Mo species and the corre-
sponding metathesis activity were studied. As revealed by
na instead of SBA-15, as evidenced by X-ray photoelectron
spectroscopy, time-of-flight secondary-ion mass spectrome-
try, and IR spectroscopy of adsorbed pyridine. Furthermore,
new Brønsted acid sites favorable for the dispersion of the
Mo species and low-temperature metathesis activity were
generated as a result of the effective synergy between the
alumina and SBA-15. The Mo/Al O @SBA-15 catalyst pre-
N adsorption isotherms and TEM results, well-dispersed alu-
2
2
3
mina was formed on the pore walls of SBA-15 if the ammo-
nia/water vapor induced hydrolysis (NIH) method was em-
ployed. The Mo species preferentially interacted with alumi-
pared by the NIH method showed higher metathesis activity
and stability under the conditions of 1208C, 0.1 MPa, and
À1
1.5 h than catalysts prepared by other methods.
Introduction
thesis could easily regulate the stocks of light olefins on the
market demand, its industrial application between ethene and
[
8,9]
Recently, the interest in propene, as an important building
block material in the petrochemical industry, has increased
owing to the strongly growing demand for propene deriva-
tives. Traditional ways to produce propene are mainly based
on naphtha steam cracking and fluid catalytic cracking as
a coproduct in the petrochemical industry, and these methods
are currently not sufficient to meet the strong market demand
2-butene to propene was realized in the 1970s.
It has been
suggested that autometathesis of 1-butene, that is, isomeriza-
tion of 1-butene into 2-butene followed by cross-metathesis of
1-butene and 2-butene, should be a more attractive and prom-
ising way to enhance the productivity of propene as a result of
the abundance of butene as a feedstock in the petrochemical
[
10–15]
industry.
[
1]
around the world. Since the beginning of this century, the
feedstock of steam cracking has been shifting to lighter
ethane, especially in North America because of the emergence
of shale gas as an unconventional fossil resource, and this will
Usually, traditional porous alumina, silica, or their composites
are applied as supports during the preparation of metathesis
[
11]
catalysts. For example, Wang et al. studied the autometathe-
sis of 1-butene on a WO /SiO catalyst under the reaction con-
3
2
[
2]
further reduce the productivity of propene. Consequently,
ditions of 3508C, weight hourly space velocity (WHSV)=
[3]
À1
several other processes, such as methanol to olefin, propane
0.36 h , and 0.5 MPa, and an initial propene yield of 32.7%
[
4]
+
[5]
[15]
dehydrogenation, catalytic cracking of C
olefins, and
have been revived and devel-
oped for the production of propene. Given that olefin meta-
was obtained. Harmase et al. deposited W species on acidic
4
[1,6,7]
olefin metathesis reactions,
alumina–silica composite supports, and a maximum propene
yield of 31.6% was realized at 4508C and 0.085 MPa. Recently,
[
13]
Mazoyer et al. developed a catalyst with tungsten hydride
supported on alumina (WH /Al O ) through a surface organo-
[
a] D. Zhang, Dr. X. Li, Prof. S. Liu, Prof. X. Zhu, F. Chen, Prof. L. Xu
State Key Laboratory of Catalysis
Dalian Institute of Chemical Physics
Chinese Academy of Sciences
Dalian 116023 (P.R. China)
3
2
3
metallic chemistry method. This method allowed better molec-
ular design of the active site, and preferable propene produc-
tivity was obtained in the autometathesis of 1-butene. In addi-
tion, some easy and effective methods, for example, non-hy-
Fax: (+86)411-84379279
E-mail: xiujieli@dicp.ac.cn
[
7]
[12]
drolytic sol–gel and the one-pot aerosol route, have also
been developed, and the catalysts exhibit high activities at low
temperatures.
lyxu@dicp.ac.cn
[
b] D. Zhang
University of Chinese Academy of Sciences
SBA-15 is a mesoporous silica with a large surface area and
Beijing 100049 (P.R. China)
[
16]
an ordered pore structure; it has drawn great attention since
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/asia.201500351.
[
17,18]
its discovery.
For example, a W-substituted mesoporous
Chem. Asian J. 2015, 10, 1647 – 1659
1647
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
SBA-15 (W-SBA-15) metathesis catalyst has been synthesized,
and it showed excellent catalytic performance in the autome-
[
19–21]
tathesis of 1-butene to propene.
However, a reaction tem-
[
22–24]
perature higher than 3508C was still needed. Balcar et al.
prepared many W-, Mo-, and Re-based catalysts supported on
mesoporous silica and studied their catalytic performance in
olefin metathesis. However, only the pure SBA-15 component
was used as a support in their earlier reports.
The development of a highly efficient heterogeneous meta-
thesis catalyst has been attracting the interest of research-
[1,25–28]
ers.
It is known that alumina is an interesting and impor-
tant material owing to its tunable surface acidity/basicity and
its mechanical strength, and it has been widely used in the
petrochemical industry as a support and catalyst. However, to
our knowledge, there are no related reports outlining the use
of alumina-modified SBA-15 as a support for the autometathe-
sis of 1-butene. In this study, guest alumina was introduced
onto/into the SBA-15 host to combine the acidity of alumina
with the advantageous mesoporous structure of SBA-15. Thus,
a novel Mo-based metathesis catalyst was developed. In addi-
tion, the effects of the different introduction methods on the
physicochemical properties of the alumina–SBA-15 composite
support and the Mo-based catalysts are discussed in detail. Re-
lations between Mo dispersion on the support and the catalyt-
ic performance in the autometathesis of 1-butene were eluci-
dated.
Figure 1. a) N
2
adsorption/desorption isotherms and b) pore-diameter distri-
bution of the pristine SBA-15 and alumina premodified composite samples.
Plots are offset vertically for clarity.)
(
Results and Discussion
aggregates formed inside the mesopores. In addition, the
degree of pore blocking was different between the two sam-
ples, as evidenced by their different pore-diameter distribu-
tions (Figure 1b). Interestingly, careful observation of the iso-
therm shape of the sample N-AS-30 indicated that the steep-
ness of the desorption branch in the capillary condensation
region improved. This change was also reflected by the nar-
rower pore-diameter distribution, as shown in Figure 1b. These
facts indicated that the introduced alumina was highly dis-
persed on the pore walls of SBA-15 through the ammonia/
water vapor-induced hydrolysis (NIH) method, which is in
Characterization of the supports
Figure 1a shows the N adsorption/desorption isotherm results
2
of the different composite supports. Pristine mesoporous SBA-
1
5 exhibits a typical type IV N isotherm with an H1-type hyste-
2
[
29]
resis loop according to the IUPAC classification. The amount
of adsorbed N decreased after introducing alumina, but yet
2
the amount was still higher than that of the pure g-Al O sup-
2
3
port. The type IV N isotherm is observed for all the alumina-
2
coated samples, which indicates that the mesoporous structure
of SBA-15 was preserved after introducing alumina. Notably,
the shape of the hysteresis loop was greatly altered depending
on the coating method. To be specific, M-AS-30 obtained by
mechanical grinding (MG) exhibited two-step capillary evapo-
ration in the desorption branch of the hysteresis loop, proba-
bly as a result of the presence of two different types of
[
30,33]
good agreement with previous reports.
The low-angle XRD patterns of pristine SBA-15 and the four
alumina-coated SBA-15 samples are shown in Figure 2a. The
characteristic (100), (110), and (200) reflections corresponding
to the hexagonal mesoporous structure can be clearly ob-
served in the SBA-15, M-AS-30, and N-AS-30 samples, respec-
tively, whereas the (110) and (200) reflections are barely visible
in the C-AS-30 and W-AS-30 samples. This may be explained
by the fact that pore blockage or improper coating on the
pore walls of SBA-15 could result in a broader distribution of
the electron density in the unit cell, which would lead to a de-
[
30]
pores.
According to the pore-diameter distribution results
(
Figure 1b), the peak value of M-AS-30 was close to that of
pristine SBA-15. Thus, we could tentatively speculate that most
of the introduced alumina accumulated outside the mesopores
of the SBA-15 host. The fact that small amounts of alumina
species migrated into the mesopores cannot be excluded be-
cause of the calcination treatment at 5008C. As to C-AS-30 and
W-AS-30, the type of hysteresis loop changed from H1 to H2.
Generally, the H2 hysteresis loop can be attributed to locally
[
30]
crease in the intensity at higher angle reflections; this im-
plies that the mesopores may be blocked by the loaded alumi-
na upon employing the conventional impregnation (CI) and
water vapor-induced hydrolysis (WIH) methods. From the
wide-angle XRD patterns (Figure 2b), no g-Al O phase was ob-
[31,32]
blocked or ink-bottle pores.
By applying the same view-
point to C-AS-30 and W-AS-30, variation of the isotherms may
be associated with pore blockage by the alumina particles or
2
3
served in the four samples.
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1648 ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
2
7
Figure 4. Al MAS NMR spectra of the alumina premodified composite sam-
ples.
tion are approximately 23–25% from the deconvolution analy-
sis. The presence of higher amounts of AlPd atoms indicates
the local arrangement of the aluminum atoms is quite different
from that in conventional g-Al O , as the spectrum of the latter
2
3
is mostly made up of AlTd and AlOh contributions (ꢀ30:70 ratio,
as shown in Figure S1, Supporting Information). Hence, it
[
36]
may be concluded that larger amounts of the amorphous alu-
mina-like phase was formed on the M-AS-30, C-AS-30, and W-
AS-30 samples, which is usually suggested to be located at the
interface between an aluminosilicate and an alumina
Figure 2. a) Low-angle and b) wide-angle XRD patterns of the pristine SBA-
15 and alumina premodified composite samples.
[
37–39]
phase.
Interestingly, as to the N-AS-30 sample, the intensi-
TEM is an important and direct technique to characterize
ty of the AlPd contribution dramatically decreased to 8%,
whereas the contribution of the AlTd atoms increased to 34%.
It is known that the formation of SiÀOÀAl linkages is closely
associated with resultant bridged hydroxy groups. Thus, the
nature of the surface hydroxy groups was evaluated after ther-
mal treatment at 3508C. As shown in Figure 5, the diffuse re-
flectance infrared Fourier transform (DRIFT) spectrum of SBA-
the mesopores of SBA-15. As shown in Figure 3a, some amor-
phous alumina aggregates are located outside the mesopores
of M-AS-30 besides the highly ordered mesopore channel of
SBA-15. This was not observed for N-AS-30 prepared by the
NIH method, which indicated that alumina was highly dis-
persed on SBA-15. Prepared N-AS-30 can also be denoted as
Al O @SBA-15, and consequently, most of the guest alumina is
À1
15 features a distinct band at n˜ =3747 cm resulting from the
2
3
uniformly and highly dispersed on the SBA-15 host.
isolated SiÀOH (silanol) groups. The intensity of this band de-
creases for the M-AS-30 sample, which is indicative of the con-
sumption of the silanol groups during alumina coating by the
The states of the Al species in the composite materials were
2
7
also investigated by Al magic angle spinning (MAS) NMR
spectroscopy. As shown in Figure 4, three distinct signals cen-
tered at d=61, 33, and 5 ppm are observed for M-AS-30, C-AS-
3
0, and W-AS-30, which can be assigned to tetrahedrally (AlTd),
pentahedrally (Al ), and octahedrally (Al ) coordinated Al spe-
Pd
34,35]
Oh
[
cies, respectively.
The Al species in tetrahedral coordina-
Figure 5. DRIFT spectra in the hydroxy region of the pristine SBA-15 and alu-
mina premodified composite samples. The spectra were measured at 308C
after pretreatment in Ar at 3508C for 0.5 h. For comparison, the spectrum of
HZSM-5 is also included.
Figure 3. TEM images of a) M-AS-30 and b) N-AS-30.
Chem. Asian J. 2015, 10, 1647 – 1659 www.chemasianj.org
1649
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
MG method, as already reported in Ref. [40]. The intensity of
this band decreased further for C-AS-30 and W-AS-30, and the
lowest intensity was observed for N-AS-30. The decreasing
trend of the band intensity is related to the consumption of si-
lanol groups as a result of their interaction with alumina. It is
well documented that bridging hydroxy groups in zeolites,
closely associated with Brønsted acid sites, are observed in the
may be generated after introducing alumina on SBA-15, espe-
cially over the N-AS-30 and W-AS-30 samples.
A Py-IR experiment was performed to distinguish the types
of acid sites. As shown in Figure 6, different signals are ob-
À1
served in the spectra in the region of n˜ =1700 to 1400 cm .
À1
Referring to Refs. [42,43], the bands at n˜ =1445 and 1595 cm
can be assigned to pyridine hydrogen bonded by surface hy-
À1
range of n˜ =3600 to 3640 cm (e.g., HZSM-5 zeolite, Figure 5).
droxy groups (e.g., SiÀOH), the bands at n˜ =1451 and
À1
No such remarkable characteristic band could be detected,
1620 cm can be assigned to pyridine adsorbed on strong
À1
and there was only a broad absorption band between n˜ =
Lewis acid sites, the band at n˜ =1577 cm can be assigned to
À1
3
700 and 3500 cm over all the alumina-coated samples. To
pyridine adsorbed on weak Lewis acid sites, and bands at n˜ =
À1
further confirm the formation of SiÀOÀAl linkages, which is
similar to the tetrahedral framework in the aluminosilicate zeo-
lite, temperature-programmed desorption of ammonia (NH3-
TPD) and infrared spectroscopy of adsorbed pyridine (Py-IR) ex-
periments were performed to understand the acidic properties
of the alumina-modified SBA-15 samples, especially the Brønst-
ed acid sites.
1545 and 1640 cm can be assigned to pyridinium-ion ring vi-
The NH -TPD results of pristine SBA-15 and the alumina-
3
modified SBA-15 supports are shown in Table 1 and Figure S2.
Clearly, the acid density on SBA-15 is low. The NH -TPD curve
3
of g-Al O features only one broad desorption peak at approxi-
2
3
mately T=2608C, and it can primarily be ascribed to ammonia
adsorbed on coordinatively unsaturated electron-deficient alu-
minum cations generated by surface dehydroxylation/deoxyge-
[
41]
nation. As to all of the alumina-modified SBA-15 samples,
the number of acid sites greatly increased relative to the
number of acid sites on pristine SBA-15, and the shapes of the
NH -TPD profiles are similar to that of g-Al O . According to
3
2
3
quantitative analysis from the NH -TPD experiment (Table 1),
3
the total acid amount decreased in the following sequence: N-
AS-30>W-AS-30>g-Al O >C-AS-30>M-AS-30>SBA-15. This
2
3
sequence implies that the acidity of the supports is strongly
dependent on the distribution of alumina, and new acid sites
3
Table 1. Acid density information from quantitative analysis of the NH -
TPD and Py-IR results.
Sample Total acid amount from
Acid amount from Py-IR
À1
NH
3
-TPD [mmolg
]
Brønsted acid Lewis acid B/L
sites [au]
sites [au]
ratio
SBA-15 0.01
M-AS-30 0.17
C-AS-30 0.30
W-AS-30 0.37
N-AS-30 0.40
0
122
170
181
362
0
0
554
581
659
654
645
786
–
0.22
0.29
0.27
0.55
–
g-Al
Mo/
SBA-15
Mo/M- 0.32
AS-30
Mo/C- 0.38
AS-30
Mo/W- 0.37
AS-30
Mo/N- 0.40
AS-30
Mo/g- 0.30
Al
2
O
3
0.32
6
0.04
0
–
6
224
283
292
411
173
423
449
497
583
632
0.53
0.63
0.59
0.70
0.27
6
6
6
Figure 6. IR spectra of pyridine adsorbed on a) alumina premodified compo-
site samples and b) the corresponding Mo-based catalysts after evacuation
at 1508C for 0.5 h.
6
2
O
3
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1650
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
brations resulting from proton transfer from the Brønsted acid
À1
sites to pyridine. The band at n˜ =1492 cm is ascribed to pyri-
dine associated with both Lewis and Brønsted acid sites. As to
pristine SBA-15, only two intense bands at n˜ =1445 and
À1
1
595 cm are detected, which indicates that neither Lewis nor
Brønsted acid sites exist on pristine SBA-15. As far as the g-
Al O support is concerned, the presence of bands at n˜ =1448,
2
3
À1
1
492, 1575, and 1614 cm demonstrates the existence of
Lewis acid sites. Furthermore, no band at approximately n˜ =
À1
1
545 cm corresponding to the Brønsted acid sites is ob-
served, which indicates that the surface hydroxy groups on g-
Al O are not strong enough to protonate the pyridine mole-
2
3
[
44]
cules. In the spectra of the alumina-modified SBA-15 sam-
À1
ples, news bands at n˜ =1545 and 1640 cm corresponding to
Brønsted acid sites are observed, which may confirm the for-
mation SiÀO(H)ÀAl linkages. On the basis of quantitative analy-
sis of the Py-IR results (listed in Table 1), introducing alumina
on SBA-15 could generate both new Brønsted and Lewis acid
sites, and the number of Brønsted acid sites decreases in the
following order: N-AS-30>W-AS-30>C-AS-30>M-AS-30.
Moreover, the ratio of Brønsted acid sites to Lewis acid sites
(
B/L ratio) decreases in the same order. The higher B/L ratio of
N-AS-30 demonstrates that more Al atoms are transformed
into the SiÀOÀAl framework structures. This order is also in
good agreement with the band intensity of the silanol groups
in the hydroxy region (Figure 5), that is, the higher the number
of Brønsted acid sites, the lower the band intensity of the sila-
nol groups. On the basis of these results, we could confirm
that the relative interaction between the alumina guest and
the SBA-15 host is strongly dependent on the method used to
introduce the alumina. Introducing alumina by the NIH
method was more favorable for the interaction of the oxygen
atoms of the silanol groups with the neighboring Al atoms,
Figure 7. a) Low-angle and b) wide-angle XRD patterns of Mo-based cata-
lysts supported on the pristine SBA-15, g-Al O , and alumina premodified
2 3
composite samples.
crease in the specific surface area for the four alumina-modi-
fied catalysts was more pronounced, and among these, Mo/N-
2
À1
AS-30 possessed the highest value of 296 m g (Figure S3).
Great difference was observed in the wide-angle XRD patterns
and this left fewer free silanol groups. As a result, more SiÀ of the Mo-based catalysts (Figure 7b). Diffraction peaks of crys-
O(H)ÀAl framework structures and Brønsted acid sites were
talline MoO were observed for Mo/SBA-15 owing to the weak
3
formed over N-AS-30. Careful investigation of the band at n˜ =
interaction between the SBA-15 parent and the Mo species,
À1
1
595 cm of pristine SBA-15 to that of the alumina-modified
whereas no diffraction peaks of crystalline MoO could be ob-
3
SBA-15 samples also revealed some critical information. Specifi-
cally, the intensity of this band for M-AS-30 decreased only
slightly relative to that of pristine SBA-15, and moreover only
served for Mo/C-AS-30, Mo/W-AS-30 and Mo/N-AS-30; this indi-
cated that the loaded Mo species were highly dispersed on
these composite supports as amorphous Mo species or existed
À1
[45]
a weak band at n˜ =1545 cm was observed. With respect to
as MoO crystallites too small to be detected by XRD. On the
3
À1
N-AS-30, the intensity of the band at n˜ =1595 cm greatly de-
contrary, the diffraction peaks for crystalline MoO were still
3
creased, and much more Brønsted acid sites were generated.
visible for Mo/M-AS-30, and this implies a lower dispersion of
Mo species over the M-AS-30 support. This conclusion was fur-
ther proved by Raman spectroscopy through the presence of
Combined with the N adsorption/desorption isotherms and
2
the TEM results, it is evident that most of the alumina accumu-
lates outside the mesopores if the MG method is used.
À1
characteristic bands (e.g. n˜ =994, 818, and 665 cm ) resulting
[
46]
from crystalline MoO3 for Mo/SBA-15 and Mo/M-AS-30 (Fig-
ure S4).
Characterization of the Mo-based catalysts
The chemical compositions of the Mo-based catalysts are
summarized in Table 2. According to the X-ray fluorescence
spectrometry (XRF) results, the overall Mo contents of these
samples were approximately 6 wt%, except for Mo/SBA-15.
The bulk Si/Al ratios were in good agreement with the nominal
values. On the basis of the XPS results, the surface Si/Al ratios
exhibited great differences depending on the alumina coating
method. The lowest value was observed on the well-coated
Mo/N-AS-30 sample, whereas Mo/M-AS-30 possessed the high-
The pore structure of the Mo-based catalysts was investigated
by N2 adsorption/desorption isotherms and low-angle XRD
measurements (Figure S3 and Figure 7a). All the catalysts
(
except for Mo/g-Al O ) exhibit a typical mesoporous structure
2 3
similar to that of the corresponding supports, which indicates
that loading the Mo species does not destroy the mesostruc-
ture of the supports. Furthermore, the specific surface area de-
creased as expected. Relative to that of Mo/SBA-15, the de-
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1651
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Chemical compositions of the Mo-based catalysts supported on
the pristine SBA-15, g-Al
2
O
3
, and alumina premodified composite sam-
ples.
Sample
Bulk (XRF)
Mo loading Si/Al
wt%] [mol/mol] [mol/mol] [mol/mol] [mol/mol]
Surface (XPS)
[
a]
[b]
[c]
Si/Al
Mo/Si
Mo/(Si+Al)
[
Mo/g-Al
2
O
3
5.6
–
–
0.039
0.040
0.037
0.030
0.030
–
Mo/N-AS-30 5.8
Mo/W-AS-30 6.3
Mo/C-AS-30 5.9
Mo/M-AS-30 5.5
Mo/SBA-15 4.8
2.1
2.0
2.1
2.7
–
1.2
1.9
2.5
4.4
–
0.072
0.057
0.042
0.037
0.020
[
a] The surface atomic ratio of Si to Al. [b] The surface atomic ratio of Mo
to Si. [c] The surface atomic ratio of Mo to (Si+Al).
est value. This phenomenon was supposed to be related to
the distribution of aluminum, and the alkaline environment
may assist migration and dispersion of the Al species over pris-
tine SBA-15. In addition, the surface Mo/Si and Mo/(Si+Al)
ratios of the catalysts increased in the reverse order of the sur-
face Si/Al ratio, and Mo/N-AS-30 possessed the highest value,
which demonstrates that the distribution of Mo species paral-
lels that of the Al species on the support. Consequently, inter-
action of the Mo species with the alumina component was
preferred to that with SBA-15. Good dispersion of alumina
inside the SBA-15 channel may guarantee the good distribu-
tion of Mo species on the catalyst. Li et al. also observed the
preferential distribution of Mo on alumina on the Hb-Al O
Figure 8. Partial negative TOF-SIMS spectra of the Mo-based catalysts a) in
the m/z=120–170 range corresponding to monomeric Mo clusters and b) in
the m/z=275–320 range corresponding to dimeric Mo clusters. The frag-
ments with one Mo atom are donated as “mono-Mo” clusters, whereas
those containing more than one Mo atoms are denoted as “poly-Mo” clus-
[26,50]
ters.
The NH -TPD profiles of the Mo-based catalysts are present-
3
ed in Figure 9. It was found that the Mo loading did not clearly
alter the overall shape of the TPD peak, but the total amount
of acid changed greatly depending on the corresponding sup-
ports (Table 1). Mo/SBA-15 showed a small NH3 desorption
peak, which indicated that new acid sites were formed after
loading Mo species on SBA-15. An increase in the total amount
of acid was also observed for Mo/M-AS-30 and Mo/C-AS-30,
whereas the total amount of acid for Mo/W-AS-30 and Mo/N-
AS-30 was almost the same as that of their corresponding sup-
ports. It is well known that the reactivity of transition-metal
ions with supports depends on the type of surface hydroxy
groups or acid sites. For example, as far as Re O /g-Al O cata-
2
3
[
47]
support. In addition, the formed bridging hydroxy groups
may be beneficial to the dispersion of Mo species on the com-
posite support.
Time-of-flight secondary-ion mass spectrometry (TOF-SIMS)
was employed to further investigate the surface states of the
Mo species. This technique is particularly helpful in under-
standing the mutual interactions between the interatoms and
analyzing the outermost surface compositions of the heteroge-
[
25,26,48,49]
neous catalysts.
Here, more attention was focused on
the Mo dispersion in terms of fragments containing one or
2
7
2
3
À
À
À
x
more Mo atoms (i.e., MoO , Mo O , and Mo O ). As shown
lysts are concerned, the rhenium ions predominately react
with basic hydroxy groups at low Re contents, whereas they
also react with neutral and more acidic hydroxy groups at
x
2
x
3
in Figure 8, the fragments containing one or more Mo atoms
were all observed on the prepared Mo-based catalysts, and
this is indicative of the presence of both isolated MoO species
x
and polymeric molybdates. It is clear that Mo/SBA-15 exhibited
largely polymeric Mo species, and the ratio of “poly-Mo/(poly-
Mo+mono-Mo)” was as high as 49.3%. The value decreased
to only 2.5% for Mo/g-Al O , which was due to the stronger in-
2
3
teraction of the Mo species with the g-Al O support. As
2
3
a result, the majority of Mo species was highly dispersed as
monomeric Mo species. With respect to the Mo-based catalysts
supported on alumina-modified SBA-15, the ratio was 6.2 and
9
.1% for Mo/N-AS-30 and Mo/M-AS-30, respectively, and this
indicated that the degree of dispersion of the Mo species was
much higher if the composite support was prepared by the
NIH method. In addition, these ratios are only slightly higher
than that of Mo/g-Al O but are remarkably lower than that of
2
3
Mo/SBA-15. This result revealed the advantageous function of
alumina for enhancing the dispersion of Mo species.
Figure 9. NH
tine SBA-15, g-Al
3
-TPD profiles of the Mo-based catalysts supported on the pris-
2
O
3
, and alumina premodified composite samples.
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1652
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
51]
higher Re loadings. From quantitative analysis of the Py-IR
results (Table 1), loading of the Mo species decreased the
number of Lewis acid sites, whereas it increased the number
of Brønsted acid sites (except for Mo/SBA-15), which indicated
that new Brønsted acid sites related to Mo species were gener-
ated. As a result, the B/L ratio of the catalysts increased relative
to that of the corresponding supports. Particularly, no Brønsted
acid sites could be observed on the pure g-Al O support,
we speculate that the broad peak at T=720–7808C is primarily
related to the Brønsted acid sites on the composite support.
This assumption is also supported by our previous study con-
cerning the H -TPR profiles of the Mo/b-Al O metathesis cata-
2
2
3
[
53]
lyst. As to the lower temperature reduction peak between
T=430 and 6008C, the peak temperature increased in the
order: Mo/g-Al O <Mo/N-AS-30<Mo/W-AS-30<Mo/C-AS-30<
2
3
Mo/M-AS-30<Mo/SBA-15. Thus, it is clear that introduction of
alumina into pristine SBA-15 changes the interaction between
the Mo species and the supports. The newly formed bridging
hydroxy groups are favorable for the reduction of highly dis-
persed Mo species.
2
3
whereas they appeared after loading the Mo species, and this
[52]
is in agreement with recently published work. The formed
surface MoÀOH and MoÀO(H)ÀSi/Al groups may be the origin
of the newly generated Brønsted acid sites. On the whole, the
amount of Brønsted acid sites decreased in the order: Mo/N-
AS-30>Mo/W-AS-30>Mo/C-AS-30>Mo/M-AS-30>Mo/g-
Al O . This is in full conformity with order of the supports.
Catalytic performance of the Mo-based catalysts
2
3
The effect of different reaction temperatures
The effects of the way in which alumina was introduced on
the reducibility of the Mo species were investigated by tem-
First, the 1-butene metathesis reaction at different tempera-
perature-programmed reduction of H2 (H -TPR) experiments.
tures over Mo/SBA-15, Mo/g-Al O , Mo/W-AS-30, and Mo/N-AS-
2
2
3
As shown in Figure 10, Mo/SBA-15 started to be reduced at ap-
proximately 5008C, and it progressively continued until 9008C.
The broadening signal contains three reduction peaks centered
at T=592, 778 and 8958C. Two characteristic reduction peaks
centered at T=436 and 9008C were observed on Mo/g-Al O .
30 was evaluated, and the results are presented in Figure 11.
The catalyst activity and product yield were strongly depen-
dent on the reaction temperature. Notably, the Mo/SBA-15 cat-
alyst exhibited poor metathesis activity and propene yield at
temperatures below 2508C. Both butene conversion and pro-
pene yield increased with increasing temperature. Notably, the
molar ratio of 2-butene to 1-butene was already higher than
1.0 in the product at 2508C (Figure 11 a). This was an indication
that a high temperature was a requisite for metathesis I
(Scheme 1) over Mo/SBA-15. The yield of ethene also slightly
increased with increasing temperature, but the yield was lower
than 3% in the temperature region investigated in this work.
Only a trace amount of isobutene was observed even at
4408C, which is indicative of the low acidity of the Mo/SBA-15
2
3
Relative to that observed for Mo/SBA-15 and Mo/g-Al O , there
2
3
was a great change in the reduction behavior of the Mo-based
catalysts supported on the alumina-modified SBA-15 sample as
a result of changes in the properties of the support. The most
significant change in the H -TPR profile was the emergence of
2
a broad peak at T=720–7808C; this change was accompanied
by the formation of a shoulder peak at approximately T=
9
008C, which was distinct for Mo/SBA-15 and Mo/g-Al O . As
2 3
[43]
reported by Klimova et al., a new reduction peak appeared
for Mo-based catalysts supported on Al-containing SBA-15
after aluminum incorporation into SBA-15, whereas the higher
temperature peak was still legible. The disparity between the
present work and that of Klimova et al. may result from differ-
ent alumina contents in modified SBA-15. With respect to the
zirconia-modified SBA-15 support, the intensity of the higher
temperature reduction peak of the Mo species also markedly
decreased, and this was accompanied by an increase in the
catalyst, and this is in agreement with the above NH -TPD
3
result and results reported in Ref. [54].
Mo/g-Al O showed better performance than Mo/SBA-15 at
2
3
low temperatures. At 1208C, the yield of ethene was 11.4%,
whereas the yield of propylene was only 10.2%; this indicated
that the self-metathesis of 1-butene dominated in the meta-
thesis reactions and that metathesis I was suppressed owing to
the low molar ratio of 2-butene to 1-butene in the product
[18]
=
=
[55]
lower temperature peak. Combined with the Py-IR results,
[n(2ÀC4 )/n(1ÀC4 )=0.1].
The conversion of butene
reached a maximum of 65.6% at 1808C, whereas the highest
yield of propene of 27.6% was obtained at a temperature of
2
108C (Figure 11 b). In addition, the yield of ethene gradually
diminished upon increasing the reaction temperature. This
may be due to the increased isomerization activity of 1-butene
=
to 2-butene [as confirmed by the increased value of n(2ÀC
4
=
)
/n(1ÀC ) with temperature], which was not beneficial for
4
[
14,56]
metathesis II (Scheme 1).
The introduction of alumina species into pristine SBA-15
during the preparation of the supports showed a strong
impact on the catalytic performance of the prepared Mo-based
catalysts. As shown in Figure 11c, the Mo/W-AS-30 catalyst al-
ready exhibited certain metathesis activity at 608C, and the
conversion of butene and the yield of propene were 36.3 and
1
2.2%, respectively. The yield of propene increased with in-
Figure 10. H
tine SBA-15, g-Al
2
-TPR profiles of the Mo-based catalysts supported on the pris-
, and alumina premodified composite samples.
2
O
3
creasing temperature, and it reached a maximum of 27.3% at
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1653
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
lar catalytic behavior was observed on Mo/N-AS-30. To be spe-
cific, two volcano-shaped patterns were found with respect to
the conversion of butene and the yield of propene as a func-
tion of reaction temperature. The maximum yield of propene
of 29.8% was also found at 1208C, and this is slightly higher
than that found on Mo/W-AS-30. Notably, the metathesis activ-
ity of Mo/N-AS-30 was better than that of Mo/W-AS-30 at
lower temperature (i.e., 60–1208C), whereas as the activity de-
creased with increasing temperature (120–2108C), and the de-
creasing trend was clearer for Mo/N-AS-30. Particularly, the
conversion of butene and the yield of propene on Mo/N-AS-30
decreased to 25.0 and 11.4%, respectively, at 2108C, and these
values are lower than those on Mo/W-AS-30. According to the
Py-IR results and H -TPR measurements, the number of Brønst-
2
ed acid sites on Mo/N-AS-30 was higher than that on Mo/W-
AS-30. This would facilitate reduction of the Mo species to
metal–carbene intermediates by the olefin at a relatively lower
[
57]
temperature. Recently, Amakawa et al. confirmed the impor-
tance of Brønsted acidity (originating from the molybdenol
groups) on Mo-based catalysts supported on SBA-15. They
postulated that the Brønsted hydroxy groups protonate the
olefin reactant to form an intermediate species, which is fur-
ther converted into the metathesis-active Mo–alkylidene spe-
cies. Therefore, it is reasonable that preferable metathesis ac-
tivity was observed on Mo/N-AS-30, whereas deep reduction
of Mo species may occur under the influence of Brønsted acid
sites if the reaction temperature is higher, and this is not bene-
[
58,59]
ficial for the olefin metathesis reaction.
It is generally ac-
cepted that highly isolated Mo species are the active precur-
[
26,60]
sors for olefin metathesis.
Although Mo/g-Al O exhibits
2 3
numerous isolated Mo species, poor metathesis activity was
observed at 1208C owing to a low number of acid sites.
Higher metathesis activity and product yields were obtained
over Mo/N-AS-30 and Mo/W-AS-30, which benefit from a suffi-
cient number of acid sites and well-dispersed Mo species on
these catalysts. On the whole, introducing alumina onto/into
pristine SBA-15 take advantages of both mesoporous SBA-15
and traditional alumina, and synergetic effects occur inside the
composite supports in term of Mo dispersion and the low-tem-
perature metathesis reaction. Indeed, this method can be used
for the preparation of effective metathesis catalysts. Therefore,
we subsequently investigated the effects of the method of in-
troduction of alumina, that is, NIH, WIH, CI, and MG, on the ac-
tivity of the final catalysts in the metathesis of 1-butene.
The effect of the method of introduction of alumina
The catalytic performance of 1-butene metathesis over the as-
prepared Mo-based catalysts with different methods for the in-
troduction of alumina are presented in Table 3 and Figure S5.
Similar metathesis activity and product distributions were ob-
served at 1 h on stream. The butene conversions were in the
range of 56 to 62%, and Mo/N-AS-30 exhibited the highest
conversion. The main metathesis products were propene (26–
30%) and pentene (18–20%), with ethene (2–4%) and hexene
(6–7%) in relatively lower proportions. These results indicate
that metathesis I was the primary metathesis reaction, whereas
Figure 11. Temperature dependence of the catalytic performance on the
a) Mo/SBA-15, b) Mo/g-Al
2
O
3
, c) Mo/W-AS-30, and d) Mo/N-AS-30 catalysts.
À1
(
Reaction pressure: 0.1 MPa, WHSV: 1.5 h , time on stream: 1 h).
1
208C. Both the conversion of butene and the yield of pro-
pene decreased upon increasing the temperature further, and
they decreased to 27.3 and 12.6%, respectively, at 2108C. Simi-
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1654
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the self-metathesis of 1-butene (metathesis II) contributed less
to the overall reaction. Generally, the double bond isomeriza-
tion step of 1-butene to 2-butene is fast relative to the meta-
[54]
thesis steps over more acidic catalysts, and this reaction may
be limited by its thermodynamic equilibrium and is favorable
for the formation of 2-butene with a relatively higher ratio of
2
-butene to 1-butene at 1208C (in the range of 3.2 to 5.7, as
[
13]
listed in Table 3).
Consequently, it is understandable that
metathesis I should take place predominately over the cata-
lysts. Moreover, if propene and pentene were mainly produced
=
=
through metathesis I, the stoichiometric ratio of n(C3 )/n(C5
)
should be equal to1.0. In fact, it was above 1.0, that is, the
yield of propene was higher than that of its co-product pen-
tene, because of the contribution of metathesis III (Scheme 1).
This was also verified by the fact that the yield of ethene was
Figure 12. Deactivation rate of Mo-based catalysts supported on alumina
premodified composite samples in terms of propene yield. (Reaction tem-
À1
perature: 1208C, reaction pressure: 0.1 MPa, WHSV: 1.5 h
)
[56]
lower than that of hexene. Meyer et al. performed thermody-
namic calculations, and they also found that the theoretical
yield of propene was higher than that of pentene by using
a mixture of 1-butene and 2-butene as feed.
during the reaction period (Figure 12), and this value is much
lower than that on Mo/M-AS-30. Propene yields of 12.3 and
14.2% were obtained over Mo/C-AS-30 and Mo/W-AS-30, re-
spectively; these yields are higher than the yield over Mo/M-
AS-30 but lower than the yield over Mo/N-AS-30. For all cata-
lysts, the yields of ethene and hexene from metathesis II also
decreased with time on stream, and still, the Mo/N-AS-30 cata-
lyst showed the highest yields of ethene and hexene at 28 h
on stream, which implies that both metathesis I and metathesi-
s II were highly active on Mo/N-AS-30. As discussed above, alu-
mina introduced by the NIH method is highly dispersed on the
pore walls, and the well-defined mesopores of the SBA-15 host
are preserved. Moreover, many more Brønsted acid and active
metathesis Mo species were formed on Mo/N-AS-30. As
a result, a much lower deactivation rate was observed. Conse-
quently, the effective synergy of alumina and SBA-15 in light of
the metathesis stability was realized only if the NIH method
was employed. As to Mo/M-AS-30, the majority of alumina ac-
cumulated outside the channels, and this was detrimental to
host–guest interactions and to the dispersion of the Mo spe-
cies. Thus, it is reasonable that Mo/M-AS-30 exhibited poor
metathesis stability. The interaction of the alumina guest and
The conversion of butene and the product distribution after
2
8 h on stream are also included in Table 3. It can be seen that
the metathesis stability is highly dependent on the method of
introduction of alumina. The conversion of butene and the
yield of propene on Mo/M-AS-30 dropped from 57.6 to 17.7%
and from 26.9 to 8.5%, respectively, after 28 h on stream. To
be specific, the Mo/M-AS-30 catalyst underwent a faster deacti-
À1
vation rate of 1.23%h within the first 10 h (Figure 12), and
after this period, the deactivation rate slowed down. Anyway,
this catalyst showed the lowest yield of propene after 28 h on
stream, and the corresponding percent yield of propene de-
creased by 68.4% relative to that obtained after 1 h (Table 3).
The Mo/N-AS-30 catalyst exhibited the best stability among
the tested Mo-based catalysts, and it was even more stable
than the previously reported Mo/HMOR-Al O catalyst under
2
3
[61]
the same reaction conditions. The yield of propene was still
as high as 23.1% after 28 h on stream, and the percent yield
decreased by 22.5% relative to that obtained after 1 h. More-
over, the profile of the yield of propene on Mo/N-AS-30 was
À1
approximately linear with a deactivation rate of 0.25%h
[a]
Table 3. Product distribution over the Mo-based catalysts at various stages in the 1-butene metathesis reaction.
Catalyst
Conversion [mol%]
Yield [mol%]
Mole ratio of the product [mol/mol]
=
=
=
=
+
=
=
=
=
=
=
C
2
C
3
C
5
C
6
C
7
n(2ÀC
4
)/n(1ÀC
4
)
n(C
3
)/n(C
5
)
2 6
n(C )/n(C )
Time on stream: 1 h
Mo/M-AS-30
Mo/C-AS-30
Mo/W-AS-30
Mo/N-AS-30
57.6
56.1
59.2
61.8
2.6
2.4
3.1
3.8
26.9
26.4
27.3
29.8
19.8
19.0
19.5
18.4
6.2
6.0
6.8
6.9
2.0
2.1
2.4
2.7
3.9
5.7
4.7
3.2
1.4
1.4
1.4
1.6
0.42
0.40
0.45
0.55
Time on stream: 28 h
[
[
[
[
b]
Mo/M-AS-30
Mo/C-AS-30
Mo/W-AS-30
Mo/N-AS-30
17.7
25.7
30.0
47.0
0.4
0.6
0.8
1.4
8.5 (68.4)
12.3 (53.4)
14.2 (48.0)
23.1 (22.5)
7.6
0.9
1.5
2.1
3.9
0.4
0.5
0.5
1.0
4.7
6.1
5.7
8.1
1.1
1.1
1.2
1.3
0.44
0.40
0.38
0.36
b]
b]
b]
10.7
12.3
17.6
À1
[
a] Reaction temperature: 1208C, reaction pressure: 0.1 MPa, WHSV: 1.5 h . [b] Number in parentheses is the decrease in the percent yield of propene
after 28 h on stream relative to that after 1 h.
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1655
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the SBA-15 host is stronger in Mo/W-AS-30 and Mo/C-AS-30
than in Mo/M-AS-30, whereas blockage of the mesopore chan-
nel was found by the accumulated alumina species. This would
also influence molecular diffusion and catalytic stability to
some extent.
great consistency between the mechanisms of both the meta-
thesis reaction of 1-butene and its deactivation over these cat-
alysts, irrespective of the method used to introduce alumina.
Carbonaceous deposits on active metathesis sites were criti-
cal factors leading to catalyst deactivation; their formation was
mainly caused by olefin oligomerization, which was generally
Unlike the varied trends of the metathesis products, the
molar ratio of 2-butene to 1-butene increased, especially for
[
58,61]
associated with the acid sites.
To analyze the carbonaceous
[
14]
Mo/N-AS-30. One possible reason is the gradual decay of
metathesis III; as a result, the content of 2-butene increases to
a higher level with time on stream. As far as Mo/N-AS-30 is
concerned, the molar ratio of 2-butene to 1-butene was the
highest after 28 h on stream, and thus, the lowest deactivation
rate of metathesis I on Mo/N-AS-30 could not be excluded. On
the whole, the ratio of 2-butene to 1-butene remained above
species deposited on the spent catalysts, an oxygen tempera-
ture-programmed oxidation (O -TPO) experiment was per-
2
formed. As shown in Figure 14, all of the spent catalysts exhib-
ited mainly two burning peaks at the same temperatures (358
and 3878C), which indicated that the deposited carbonaceous
species possessed similar properties. Interestingly, the integral
area of all peaks increased in the order: Mo/N-AS-30>Mo/W-
AS-30>Mo/C-AS-30>Mo/M-AS-30, which correlates well with
the catalyst stability. Thus, the amount of carbonaceous depos-
its on the catalysts was not directly related to the metathesis
stability, as Mo was the metathesis-active site. On basis of the
above results, it was concluded that the high dispersion of Mo
species and the well-defined mesostructure of N-AS-30 guaran-
tee high accommodation coke capacity of Mo/N-AS-30 and
good metathesis stability in the reaction.
3
.0 during the reaction period, which indicated that double-
bond isomerization was not the critical factor leading to cata-
lyst deactivation.
Variations in the product yield with butene conversion are
presented in Figure 13. The yield of each product seems to
depend on butene conversion following a similar trend. The
yield of propene increased linearly with the conversion of
butene, irrespective of the catalyst, and this is indicative of
constant propene selectivity during the reaction. Additionally,
the most selective product was propene, especially at higher
butene conversions. The steady yield of propene upon Mo/N-
AS-30 was due to its higher catalytic activity. Interestingly, the
yield of pentene initially increased with the conversion of
butene, similar to that observed for propene. Nevertheless, it
exhibited a maximum if the conversion was higher than 45%.
As shown in Scheme 1, pentene was primarily produced from
metathesis I, with propene as a coproduct. Propene could also
be produced through metathesis III followed by metathesis II.
Therefore, it is speculated that metatheses I, II, and III exhibit
high reactivity at high conversions of butene. This could be
further proved by the observation that the yields of both
ethene and hexene presented an approximately exponential
increase if the conversion of butene was higher than 40%. Al-
though the deactivation rates of the three Mo-based catalysts
were different, the product yields of ethene, propene, pentene,
and hexene as a function of the conversion of butene followed
similar trends. This gave us some indication that there existed
2
Figure 14. O -TPO profiles of the spent Mo/M-AS-30, Mo/C-AS-30, Mo/W-AS-
30, and Mo/N-AS-30 catalysts.
Conclusions
In this study, four Al O –SBA-15 composite supports with dif-
2
3
ferent alumina location states were prepared through postsyn-
thesis methods, that is, ammonia/water vapor induced hydroly-
sis, water vapor induced hydrolysis, conventional impregna-
tion, and mechanical grinding. The corresponding Mo-based
catalysts were evaluated in the autometathesis of 1-butene. It
was found that the dispersion of the Mo species and the meta-
thesis activity and stability were closely associated with the lo-
cation states of guest alumina on the SBA-15 host, which was
strongly dependent on the modification methods of SBA-15
with alumina.
Upon using the mechanical grinding method, most of the
alumina accumulated outside the mesopore channels. Water
vapor induced hydrolysis and the conventional impregnation
method led to blockage of the mesopores to some extent
Figure 13. Variation in the product yields with butene conversion over the
Mo/M-AS-30, Mo/W-AS-30, and Mo/N-AS-30 catalysts.
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1656
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
with the introduction of alumina. Well-dispersed alumina on
the mesopore walls of SBA-15 was realized without pore-chan-
nel blocking by using the ammonia/water vapor induced hy-
drolysis (NIH) method. In fact, a better combination of guest
alumina and host SBA-15 was realized by the NIH method
through the formation of SiÀO(H)ÀAl hydroxy groups.
resultant support is denoted C-AS-30. As to the solvent-free grind-
ing method, Al(NO ) ·9H O (3.16 g) was manually ground together
3 3
2
with SBA-15 (1.0 g) in a mortar at room temperature for 0.5 h. The
obtained powder was calcined at 5008C for 5 h in air and is denot-
ed M-AS-30.
Catalyst preparation
It was found that the Mo species preferentially interacted
with alumina and not siliceous SBA-15. A uniform distribution
of alumina around the SBA-15 channel guaranteed the high
dispersion of the Mo species on the support. Moreover, the
generation of new Brønsted acid sites was also favorable for
dispersion of the Mo species and low-temperature butene
metathesis activity. The optimal reaction temperature for the
metathesis of butene was related to the number of Brønsted
Catalysts containing approximately 6.0 wt% Mo were prepared by
incipient wetness impregnation of the composite supports with an
aqueous solution of (NH ) Mo O ·4H O and left at room tempera-
4 6
7
24
2
ture for 24 h. The concentration of ammonium molybdate depend-
ed on the quantity and water absorbility of the support. The sam-
ples were then dried at 1208C for 4 h and finally calcined at 6008C
for 2 h. The final catalysts are denoted Mo/N-AS-30, Mo/W-AS-30,
Mo/C-AS-30, and Mo/M-AS-30 for Mo species supported on N-AS-
acid sites. In general,
a novel Mo-based catalyst with
3
0, W-AS-30, C-AS-30, and M-AS-30, respectively. For comparison,
a Al O @SBA-15 composite support prepared by the NIH
Mo-based catalysts supported on g-Al O and SBA-15, respectively,
2
3
2
3
method exhibited the best metathesis activity and stability
were also prepared.
among the tested catalysts under the reaction conditions of
À1
1
208C, 0.1 MPa, and weight hourly space velocity=1.5 h as
Catalysts characterization
a result of its good Mo dispersion and high Brønsted acid den-
The N adsorption/desorption isotherms measurements were per-
2
sity.
formed at À1968C with a Micromeritics ASAP 2020 instrument. X-
ray diffraction (XRD) patterns were collected with an X’Pert PRO
diffractometer. Transmission electron microscopy (TEM) images
were obtained with an FEI Tecnai G2 Spirit microscope with an ac-
Experimental Section
27
celeration voltage of 100 kV. Al NMR spectra were measured with
a Bruker AVANCE 500 spectrometer. Diffuse reflectance infrared
Fourier transform spectroscopy (DRIFTS) experiments in the hy-
droxy region were performed with a Bruker VERTEX 70 spectrome-
ter equipped with a Harrick Praying Mantis optical accessory. Infra-
red spectroscopy of adsorbed pyridine (Py-IR) was used to differen-
tiate between Lewis and Brønsted acid sites. Temperature-pro-
Coating alumina on SBA-15
Commercial mesoporous SBA-15 was supplied by Novel Chemical
Technology Corporation and was calcined at 5008C before use.
Alumina was introduced into the SBA-15 host by four different
methods, namely, ammonia/water vapor induced hydrolysis
[30,62]
(
NIH),
water vapor induced hydrolysis (WIH), conventional im-
pregnation (CI), and mechanical grinding (MG), in which
grammed desorption of ammonia (NH -TPD) and temperature-pro-
3
[63]
Al(NO ) ·9H O was used as the source of alumina. Iengo et al. re-
grammed reduction of H (H -TPR) experiments were performed in
3
3
2
2
2
ported that the amount of aluminum required to create an Al
a quartz microreactor (i.d. 4 mm) that was connected to an online
gas chromatograph (Shimadzu GC-8A) equipped with a thermal
conductivity detector. The chemical compositions of all the sam-
ples were analyzed by using a PANalytical Axios X-ray fluorescence
spectrometer. X-ray photoelectron spectra were recorded by using
a VG ESCALAB MK-II spectrometer with a sample transfer system.
TOF-SIMS investigations were performed by using an IONTOF in-
2
À1
monolayer coverage on silica with a surface area of 280 m g was
À1
2
7
5
.1 mmolg , which corresponds to a surface concentration of
À2
.5 mmolm . The surface area of SBA-15 used herein was
2
À1
32 m g . Therefore, approximately 30 wt% of alumina coating
was employed on SBA-15 in this work by different methods to
obtain an Al coverage higher than its monolayer.
+
strument equipped with a Bi primary ion source in the static
Alumina coating onto/into SBA-15 by the NIH method was per-
mode. Temperature-programmed oxidation (TPO) experiments
were performed in a U-shaped quartz microreactor with an online
Omnistar mass spectrometer. (Detailed descriptions are provided in
the Supporting Information.)
[30]
formed by following a reported procedure. Typically, mesopo-
rous SBA-15 (1.0 g) was dispersed in a 10 mL aqueous solution of
Al(NO ) ·9H O (1.58 g). The mixture was then evaporated at 608C
3
3
2
until dryness, which was followed by drying at 1008C for 6 h. To
improve the dispersion of alumina inside the pores, this impregna-
tion step was repeated once more, which allowed an alumina load-
ing of 30 wt% to be obtained. Subsequently, the Al-precursor-
loaded SBA-15 was put in an open container made of Teflon, and
then kept in an autoclave containing a NH /H O solution with no
direct contact between the solid and solution. The tightly closed
autoclave was then heated to 1008C for 7 h. After treatment, the
solid was dried at room temperature for 6 h and then at 1008C for
Catalyst evaluation
The catalytic test was performed in a fixed-bed flow microreactor
with an inner diameter of 7 mm. The reaction temperature was
measured with a coaxial thermocouple, the end of which was fixed
in the middle of the catalyst bed. In each experiment, the catalyst
(0.5 g) was charged at the center of the reactor. Conventionally,
the catalyst was pretreated in situ at 5508C with inert N2
3
2
À1
1
2 h. Finally, the material was calcined at 5008C for 5 h in air, and
(20 mLmin ) for 2 h, and then the system was cooled down to
the obtained composite support is denoted N-AS-30. The proce-
dure for the WIH method was similar to that for the NIH method,
except that the liquid in the autoclave was deionized water, and
the resultant material is denoted W-AS-30. For the conventional
impregnation method, alumina-loaded SBA-15 was directly cal-
cined at 5008C for 5 h without NH /H O vapor treatment, and the
the desired reaction temperature under N2 flow. Afterwards, 1-
butene feed with a weight hourly space velocity of 1.5 h was in-
troduced into the reactor. The reaction products were analyzed by
an online Varian CP 3800 gas chromatograph equipped with
a flame ionization detector. The connecting lines between the reac-
tor outlet and the sampling valve were heated to prevent conden-
À1
3
2
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1657
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
sation of the products with high molecular weights. Quantitative
analyses of the products were calculated by using the corrected
area normalization method, and carbonaceous deposits on cata-
lysts were not taken into account for the calculation.
[2] C. Angelici, B. M. Weckhuysen, P. C. A. Bruijnincx, ChemSusChem 2013, 6,
1595–1614.
[
[
[
3] M. Stçcker, Microporous Mesoporous Mater. 1999, 29, 3–48.
4] F. Cavani, N. Ballarini, A. Cericola, Catal. Today 2007, 127, 113–131.
5] X. X. Zhu, S. L. Liu, Y. Q. Song, S. J. Xie, L. Y. Xu, Appl. Catal. A 2005, 290,
1
91–199.
Isomers of 1-butene , that is, cis-2-butene and trans-2-butene pro-
duced from the isomerization of 1-butene, were considered to be
unconverted feedstock. The conversion (X) of 1-butene and the
yield (Y) to a particular product were calculated by using Equa-
tions (1) and (2):
[
6] C. Lin, K. Tao, H. B. Yu, D. Y. Hua, S. H. Zhou, Catal. Sci. Technol. 2014, 4,
4010–4019.
[7] D. P. Debecker, K. Bouchmella, M. Stoyanova, U. Rodemerck, E. M.
Gaigneaux, P. H. Mutin, Catal. Sci. Technol. 2012, 2, 1157–1164.
8] R. L. Banks, G. C. Bailey, I&EC Prod. Res. Dev. 1964, 3, 170–173.
9] J. C. Mol, J. Mol. Catal. A 2004, 213, 39–45.
[
[
[
[
10] J. Engelhardt, D. Kallo, I. Zsinka, J. Catal. 1984, 88, 317–324.
11] Y. D. Wang, Q. L. Chen, W. M. Yang, Z. K. Xie, X. Xu, D. Huang, Appl.
Catal. A 2003, 250, 25–37.
1ÀC4¼
Š À½1ÀC Š À½2ÀC ¼Š
F
4
¼
p
4
p
½
Xð1ÀC4¼Þ ¼
 100 %
ð1Þ
ð2Þ
¼
½
1ÀC
Š
[12] D. P. Debecker, M. Stoyanova, F. Colbeau-Justin, U. Rodemerck, C. Bois-
siere, E. M. Gaigneaux, C. Sanchez, Angew. Chem. Int. Ed. 2012, 51,
2129–2131; Angew. Chem. 2012, 124, 2171–2173.
[13] E. Mazoyer, K. C. Szeto, S. Norsic, A. Garron, J. M. Basset, C. P. Nicholas,
M. Taoufik, ACS Catal. 2011, 1, 1643–1646.
4
F
¼
¼
p
Xð1ÀC Þ Â ½C
Š
YðC ¼Þ ¼
4
n
 100 %
p
n
C2¼Šp þ ½C3¼Š þ ½C Š þ ½C6¼Šp þ ½C7þŠ
¼
½
p
5
p
[14] X. J. Li, D. Z. Zhang, X. X. Zhu, F. C. Chen, S. L. Liu, S. J. Xie, L. Y. Xu, J.
Mol. Catal. A 2013, 372, 121–127.
[15] L. Harmse, C. van Schalkwyk, E. van Steen, Catal. Lett. 2010, 137, 123–
131.
[16] D. Y. Zhao, Q. S. Huo, J. L. Feng, B. F. Chmelka, G. D. Stucky, J. Am. Chem.
Soc. 1998, 120, 6024–6036.
[17] X. Li, W. Zheng, H. Pan, Y. Yu, L. Chen, P. Wu, J. Catal. 2013, 300, 9–19.
=
in which [1ÀC ] represents the weight percent of 1-butene in
4
F
=
=
=
=
+
the feed; [C ]p, [C3 ] , [2ÀC ]p, [C5 ] , [C ] , and [C ] are the
2
p
4
p
6 p
7
weight percentages of each component in the products, respec-
=
tively; Y(C ) is the product yield with the carbon number n. In ad-
dition, C7 denotes all the heavy products having seven or more
n
+
[
[
[
18] O. Y. Gutierrez, D. Valencia, G. A. Fuentes, T. Klimova, J. Catal. 2007, 249,
40–153.
carbon atoms.
1
19] J. C. Hu, Y. D. Wang, L. F. Chen, R. Richards, W. M. Yang, Z. C. Liu, W. Xu,
Microporous Mesoporous Mater. 2006, 93, 158–163.
20] T. I. Bhuiyan, P. Arudra, M. N. Akhtar, A. M. Aitani, R. H. Abudawoud,
M. A. Al-Yami, S. S. Al-Khattaf, Appl. Catal. A 2013, 467, 224–234.
Scheme 1 illustrates the main possible reaction pathways for the
metathesis of 1-butene over Mo-based catalysts.
pene could come from metathesis of 2-butene and 1-butene (met-
athesis I) and metathesis of ethene and 2-butene (metathesis III). 1-
Butene isomerization to 2-butene as the prerequisite step of meta-
thesis I and self-metathesis of 1-butene to ethene and hexene
[14,61]
Desired pro-
[21] B. Hu, H. Liu, K. Tao, C. R. Xiong, S. H. Zhou, J. Phys. Chem. C 2013, 117,
26385–26395.
[
22] P. Topka, H. Balcar, J. Rathousky, N. Zilkova, F. Verpoort, J. Cejka, Micro-
porous Mesoporous Mater. 2006, 96, 44–54.
[23] H. Balcar, D. Mishra, E. Marceau, X. Carrier, N. Zilkova, Z. Bastl, Appl.
Catal. A 2009, 359, 129–135.
(
metathesis II) is an alternative route for the production of propene
through metathesis III.
[14]
[
[
24] H. Balcar, J. Cejka, Coord. Chem. Rev. 2013, 257, 3107–3124.
25] D. P. Debecker, K. Bouchmella, C. Poleunis, P. Eloy, P. Bertrand, E. M.
Gaigneaux, P. H. Mutin, Chem. Mater. 2009, 21, 2817–2824.
[
[
[
[
[
26] D. P. Debecker, B. Schimmoeller, M. Stoyanova, C. Poleunis, P. Bertrand,
U. Rodemerck, E. M. Gaigneaux, J. Catal. 2011, 277, 154–163.
27] K. Bouchmella, P. H. Mutin, M. Stoyanova, C. Poleunis, P. Eloy, U. Rode-
merck, E. M. Gaigneaux, D. P. Debecker, J. Catal. 2013, 301, 233–241.
28] N. Liu, S. L. Ding, Y. M. Cui, N. H. Xue, L. M. Peng, X. F. Guo, W. P. Ding,
Chem. Eng. Res. Des. 2013, 91, 573–580.
29] D. Y. Zhao, J. L. Feng, Q. S. Huo, N. Melosh, G. H. Fredrickson, B. F.
Chmelka, G. D. Stucky, Science 1998, 279, 548–552.
30] K. K. Cheralathan, T. Hayashi, M. Ogura, Microporous Mesoporous Mater.
2
008, 116, 406–415.
Scheme 1. Reaction pathways for the autometathesis of 1-butene over Mo-
based catalysts, as proposed in Ref. [64].
[
[
31] A. García-Trenco, A. Martínez, Catal. Today 2013, 215, 152–161.
32] M. Shakeri, R. Gebbink, P. E. de Jongh, K. P. de Jong, Microporous Meso-
porous Mater. 2013, 170, 340–345.
[
[
[
[
[
33] X. Y. Liu, L. B. Sun, F. Lu, T. T. Li, X. Q. Liu, J. Mater. Chem. A 2013, 1,
Acknowledgements
1623–1631.
34] X. Y. She, J. H. Kwak, J. M. Sun, J. Z. Hu, M. Y. Hu, C. M. Wang, C. H. F.
Peden, Y. Wang, ACS Catal. 2012, 2, 1020–1026.
35] P. C. Bakala, E. Briot, Y. Millot, J. Y. Piquemal, J. M. Bregeault, J. Catal.
We are grateful for financial support from the National Natural
Science Foundation of China (Grant Nos. 20903088 and
2008, 258, 61–70.
2
1376235).
36] C. Pecharromµn, I. Sobrados, J. E. Iglesias, T. Gonzalez-Carreno, J. Sanz, J.
Phys. Chem. B 1999, 103, 6160–6170.
37] B. M. De Witte, P. J. Grobet, J. B. Uytterhoeven, J. Phys. Chem. 1995, 99,
Keywords: alumina · metathesis · molybdenum · propene ·
6961–6965.
sba-15
[38] A. H. Tavakoli, P. S. Maram, S. J. Widgeon, J. Rufner, K. van Benthem, S.
Ushakov, S. Sen, A. Navrotsky, J. Phys. Chem. C 2013, 117, 17123–17130.
[39] A. Goldbourt, M. V. Landau, S. Vega, J. Phys. Chem. B 2003, 107, 724–
731.
[
1] E. Mazoyer, K. C. Szeto, N. Merle, S. Norsic, O. Boyron, J.-M. Basset, M.
[40] H. N. Wang, X. H. Li, Y. M. Wang, P. Wu, ChemCatChem 2010, 2, 1303–
Taoufik, C. P. Nicholas, J. Catal. 2013, 301, 1–7.
1311.
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1658
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
[
41] R. Vidruk, M. V. Landau, M. Herskowitz, V. Ezersky, A. Goldbourt, J. Catal.
011, 282, 215–227.
42] J. M. R. Gallo, C. Bisio, G. Gatti, L. Marchese, H. O. Pastore, Langmuir
010, 26, 5791–5800.
43] T. Klimova, J. Reyes, O. Gutierrez, L. Lizama, Appl. Catal. A 2008, 335,
59–171.
[54] L. Alvarado Perea, T. Wolff, P. Veit, L. Hilfert, F. T. Edelmann, C. Hamel, A.
Seidel-Morgenstern, J. Catal. 2013, 305, 154–168.
[55] H. J. Liu, S. J. Huang, L. Zhang, S. L. Liu, W. Wang, W. J. Xin, S. J. Xie, L. Y.
Xu, Chin. J. Catal. 2008, 29, 513—518.
[56] W. H. Meyer, M. M. D. Radebe, D. W. Serfontein, U. Ramdhani, M. du Toit,
C. P. Nicolaides, Appl. Catal. A 2008, 340, 236–241.
2
2
1
[
[
44] X. S. Liu, R. E. Truitt, J. Am. Chem. Soc. 1997, 119, 9856–9860.
45] Y. C. Lou, H. C. Wang, Q. H. Zhang, Y. Wang, J. Catal. 2007, 247, 245–
[57] K. Amakawa, S. Wrabetz, J. Krohnert, G. Tzolova-Muller, R. Schlçgl, A.
Trunschke, J. Am. Chem. Soc. 2012, 134, 11462–11473.
[58] X. J. Li, W. P. Zhang, X. Li, S. L. Liu, H. J. Huang, X. W. Han, L. Y. Xu, X. H.
Bao, J. Phys. Chem. C 2009, 113, 8228–8233.
[59] H. Aritani, O. Fukuda, T. Yamamoto, T. Tanaka, S. Imamura, Chem. Lett.
2000, 66–67.
2
55.
[
46] K. Amakawa, L. L. Sun, C. S. Guo, M. Havecker, P. Kube, I. E. Wachs, S.
Lwin, A. I. Frenkel, A. Patlolla, K. Hermann, R. Schlogl, A. Trunschke,
Angew. Chem. Int. Ed. 2013, 52, 13553–13557; Angew. Chem. 2013, 125,
1
3796–13800.
47] X. J. Li, W. P. Zhang, S. L. Liu, L. Y. Xu, X. W. Han, X. H. Bao, J. Catal. 2007,
50, 55–66.
[60] J. Handzlik, P. Sautet, J. Catal. 2008, 256, 1–14.
[61] S. J. Huang, H. J. Liu, L. Zhang, S. L. Liu, W. J. Xin, X. J. Li, S. J. Xie, L. Y.
Xu, Appl. Catal. A 2011, 404, 113–119.
[
[
2
48] F. De Smet, M. Devillers, C. Poleunis, P. Bertrand, J. Chem. Soc. Faraday
Trans. 1998, 94, 941–947.
49] L.-T. Weng, Appl. Catal. A 2014, 474, 203–210.
50] S. Rondon, W. R. Wilkinson, A. Proctor, M. Houalla, D. M. Hercules, J.
Phys. Chem. 1995, 99, 16709–16713.
[62] C. K. Krishnan, T. Hayashi, M. Ogura, Adv. Mater. 2008, 20, 2131–2136.
[63] P. Iengo, M. Di Serio, A. Sorrentino, V. Solinas, E. Santacesaria, Appl.
Catal. A 1998, 167, 85–101.
[64] X. J. Li, X. X. Zhu, D. Z. Zhang, F. C. Chen, P. Zeng, S. L. Liu, S. J. Xie, L. Y.
Xu, J. Energy Chem. 2013, 22, 145–150.
[
[
[
[
[
51] M. Sibeijn, R. Spronk, J. A. R. Vanveen, J. C. Mol, Catal. Lett. 1991, 8,
2
01–208.
52] T. Kitano, S. Okazaki, T. Shishido, K. Teramura, T. Tanaka, J. Mol. Catal. A
013, 371, 21–28.
Manuscript received: April 5, 2015
Accepted article published: May 22, 2015
Final article published: June 25, 2015
2
53] S. L. Liu, S. J. Huang, W. J. Xin, J. Bai, S. J. Xie, L. Y. Xu, Chem. Res. Chin.
Univ. 2005, 21, 213–216.
Chem. Asian J. 2015, 10, 1647 – 1659
www.chemasianj.org
1659
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim