Performance and characterization of a new crystalline SbRe2O6 catalyst for selective oxidation of methanol to methylal

Article abstract of DOI:10.1006/jcat.2000.2990

SbRe₂O₆, Sb₄Re₂O₁₃, and SbOReO₄·2H₂O, and a number of supported Re catalysts were used as catalysts for the selective oxidation of methanol to methylal (3CH₃OH + 1/2O₂ → CH₂(OCH₃)₂ + 2H₂O). At 573 K, a high selectivity of 92.5% was obtained on the novel crystalline catalyst SbRe₂O₆. No methylal formation or negligible activity was observed with the other catalysts. No structural change in the bulk and surface of the SbRe₂O₆ catalysts occurred after methanol oxidation < 600 K. The reaction rate increased with increases in methanol partial pressure, while the selectivity to methylal was independent of methanol partial pressure, as well as O₂ partial pressure (< 10 mol %). There were two kinds of active lattice oxygen species in TPD experiments on SbRe₂O₆, which were responsible for the methylal formation. The high performance of SbRe₂O₆ for the selective synthesis of methylal from methanol was attributed to the Re-oxide species stabilized by the specific connection with Sb oxides at the crystal surface.

Full text of DOI:10.1006/jcat.2000.2990

Journal of Catalysis 195, 51–61 (2000)
doi:10.1006/jcat.2000.2990, available online at http://www.idealibrary.com on
Performance and Characterization of a New Crystalline SbRe O
2
6
Catalyst for Selective Oxidation of Methanol to Methylal
1
2
Youzhu Yuan, Haichao Liu, Hideo Imoto, Takafumi Shido, and Yasuhiro Iwasawa
Department of Chemistry, Graduate School of Science, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
Received February 29, 2000; revised June 26, 2000; accepted July 14, 2000
diate for preparing high-concentration formaldehyde, and
Three well-defined compounds, SbRe2O6, Sb4Re2O13, and a reagent in organic synthesis, has not successfully been
SbOReO4 � 2H2O, and several supported Re catalysts were em- produced by catalytic methanol oxidation. The catalytic
ployed as catalysts for the selective oxidation of methanol to methy-
methylal synthesis from methanol has been reported on
V/TiO2 (2), V–Mo–O (5), PMoH-5.75/SiO2 (6), Mo/MCM-
lal (3CH3OH + 1/2O2 → CH2(OCH3)2 + 2H2O). High selectivity of
9
2.5% to methylal was obtained on the new crystalline catalyst
4
1 (7), a MoO3(100) plane (8), and electrocatalysts (9), but
SbRe2O6 at 573 K, while no methylal formation or negligible ac-
tivity was observed with the other catalysts. No structural change
in the bulk and surface of the SbRe2O6 catalyst occurred after
the methanol oxidation below 600 K as characterized by XRD,
Raman, XPS, and SEM. The reaction rate increased with increas-
ing methanol partial pressure, while the selectivity to methylal was
independent of methanol partial pressure as well as O2 partial pres-
the selectivities to methylal on those catalysts were low and
practically insignificant. Hence, discovery of a new selec-
tive oxidation catalyst is the key issue to realizing the direct
methylal synthesis from methanol, where three methanol
molecules are incorporated into a methylal molecule.
We have examined combinations of Sb and Re oxides as
sure (<10 mol%). There existed two types of active lattice oxygen selective oxidation catalysts (10, 11). The occurrence of rhe-
species in TPD experiments on SbRe2O6, both being responsible for nium in a wide range of oxidation states leads to a rich and
the methylal formation. The high performance of SbRe2O6 for the
selective synthesis of methylal from methanol may be ascribed to
the Re–oxide species stabilized by the specific connection with Sb
interesting chemistry that is illustrated in both binary and
ternary oxides (12–17). Supported rhenium oxide catalysts
are active for alkene metathesis (18, 19) and hydrodesulfu-
oxides at the crystal surface.
�c 2000 Academic Press
rization (HDS) of heavy fractions of crude oil (20, 21). For
these reactions supported rhenium oxide catalysts exhibit
activities superior to supported molybdenum and tungsten
oxide catalysts. Up to date, there are no rhenium oxide cata-
lysts that work efficiently in selective methanol oxidation
due to facile volatilization of rhenium oxide species formed
during catalyst pretreatments and under the reaction con-
ditions (12, 22–27). Meanwhile, antimony is a well-known
promoter element in selective oxidation catalysts, provid-
ing many mixed oxide formulations together with V, Sn,
Mo, Fe, and U (27–32). We reported good performances of
a Pt/SbOx catalyst for the selective oxidation of isobutane
and isobutylene to methacrolein (33, 34). On the Pt/SbOx
catalyst, a Sb6O13 phase formed in situ under the catalyticre-
action conditions contributes to the selective performance.
The Sb6O13 has a defect pyrochlore-type cubic structure,
Key Words: Sb–Re–O mixed oxide catalysts; SbRe2O6; selective
catalytic oxidation of methanol; methylal synthesis; XRD; XPS;
Raman; SEM; TPD; TPR.
1
. INTRODUCTION
Numerous efforts have been made with the development
of selective oxidation catalysts for methanol conversion
to formaldehyde, methyl formate, or dimethoxymethane
(
methylal) from both academic and industrial interests. The
methanol oxidation to formaldehyde has been extensively
studied and commercialized on silver and ferric molyb-
date catalysts (1). Methyl formate has also been produced
with high yields by the catalytic oxidation of methanol on
V–Ti oxides (2), Sn–Mo oxides (3), and Bi-based mixed ox-
ides (4). However, methylal, which is used as a gasoline
additive, a solvent in the perfume industry, a key interme-
consisting of a network of SbO6 octahedra with –Sb3 –O
+
2�
–
chains positioned inside vacancies formed by the network
of octahedra (35). A new crystalline compound, SbRe2O6,
has similar Sb3 –O chains that connect with [Re2O6]
+
2�
3�
1
Permanent address: Department of Chemistry, State Key Labora-
layers (14).
tory for Physical Chemistry of Solid Surface, Xiamen University, Xiamen
Recently, we have found that three Sb–Re–O crystal-
3
61005, China.
2
line oxides, SbOReO � 2H O, Sb Re O , and particularly
To whom correspondence should be addressed. E-mail: iwasawa@
chem.s.u-tokyo.ac.jp. Fax: 81-3-5800-6892.
4
2
4
2
13
SbRe2O6, constitute a new family of promising catalysts
51
0021-9517/00 $35.00
Copyright �c 2000 by Academic Press
All rights of reproduction in any form reserved.

5
2
YUAN ET AL.
for the selective oxidation reactions of isobutane and
Pulse experiments were carried out in the same system by
isobutylene at 673–773 K (10, 11). The detailed characteri- using He as the carrier gas in a flow rate of 60 ml/min. The
zations of the crystalline oxides by XPS, Raman, SEM, and catalyst SbRe2O6 (200 mg) was pretreated at 573 K under
XRD, however, revealed that there was a significant decom- He for 1 h. The amount of ca. 1 ml of He/MeOH = 96.0/
position of the Sb–Re–O compounds during the catalytic 4.0 (mol% ) was pulsed into the catalyst bed at 573 K at an
selective oxidation of isobutane at 773 K (10, 11).
interval of about 15 min. The effluent gas from the reactor
More recently, we have discovered that the selective cata- was analyzed by the on-line gas chromatgraph.
lytic oxidation of methanol to methylal efficiently proceeds
2
.3. Characterization
on SbRe2O6 with a selectivity as high as 92.5% at 573 K,
in which there is no structural change in the SbRe2O6 crys-
tal (36). The aim of this paper is to discuss the key issues
relevant to the good performance of the new SbRe2O6 cata-
lyst for the selective oxidation of methanol to methylal by
means of flow and pulse experiments, XRD, XPS, Raman,
SEM, TPD, and TPR.
The Re–Sb–O catalysts were characterized ex situ before
and after selective methanol oxidation by powder X-ray
diffraction (XRD), X-ray photoelectron spectroscopy
(
XPS), scanning electron microscopy (SEM), and Mircro-
confocal laser Raman spectroscopy (Raman). In addition,
the samples were characterized by O2 temperature-
programmed desorption (TPD) and H2 temperature-prog-
rammed reduction (TPR).
2
. EXPERIMENTAL
XRD patterns were measured on a Rigaku Miniflex go-
niometer. The analysis was carried out in a continuous �ꢀ2�
scan reflection mode using Cu K� radiation (� = 0.15418).
The anode was operated at 30 kV and 15 mA. The 2� angles
were scanned from 5 to 60 at a rate of 2 /min.
Raman spectra were recorded under an ambient at-
mosphere by using a confocal microprobe Raman system
2
.1. Catalyst Preparation
Three well-defined Re–Sb–O compounds, SbOReO4 �
2
H2O, SbRe2O6, and Sb4Re2O13, were prepared accord-
�
�
�
ing to the procedures reported previously (10–14). Sb2O3
Soekawa, purity99.99% ) and SiO2 (Aerosil200) were used
(
as supports for Re oxides. The supported Re oxides were
prepared by an incipient wetness impregnation method
using an aqueous NH4ReO4 (Soekawa, purity 99.9% )
solution. A mechanical mixture catalyst of Sb2O3 with
NH4ReO4 was also prepared by the known method (37).
The decomposition of the ammonium perrhenate precur-
sor to Re2O7 in the supported samples and the mechanically
mixed sample was performed by temperature-programmed
calcination (4 K/min) up to 573 K in a flow of He/O2 =
(
LabRam I). A holographic notch filter was equipped to
� 1
filter the excitation line and an 1800 g mm holographic
grating was employed to disperse the scattered light. The
excitation wavelength was 632.8 nm with a power of 12 mW
from an internal He–Ne laser. The size focused on the sam-
ple surface was ca. 5 �m.
XPS spectra were measured on a Rigaku XPS-7000
spectrometer by using Mg K� radiation (1253.6 eV) with
X-ray power of 200 W (accelerating voltage, 20 kV; emis-
sion current, 10 mA). Samples were pressed into thin disks,
9
0.0/10.0 (mol% ) at atmospheric pressure and the samples
were further calcined at 573 K for 2 h. They are denoted
as Re2O7/Sb2O3, Re2O7/SiO2, and Re2O7 + Sb2O3, res-
pectively.
� 5
placed on holders, outgassed to less than 2.6 � 10 Pa in a
prechamber, and transferred to an analysis chamber. The
binding energy was referred to 284.6 eV for C 1s. The peak
intensity was normalized by the peak height of Sb 4d at
2
.2. Catalytic Methanol Oxidation
3
4.4 eV. The peak deconvolution and fitting was performed
The catalytic performances were examined in a conven- by using a software of SpXzeigR2.1 running with IgorPro
tional fixed-bed flow reactor by using 200 mg of catalyst and Gaussian–Lorentzian lineshape, fixing both spin–orbit
at 1 atm. All the catalysts were pretreated under He at splitting and the relative intensity of spin–orbit compo-
573 K for 1 h in situ before catalytic reaction. These sam- nents.
ples are denoted as “fresh” ones. Methanol (Wako, pu- SEM images were taken on a Hitachi S-4500 microscope
rity 99.8% ) was introduced to the flow reactor by bub- equipped with a field emission gun operated with an accel-
bling He gas through a glass saturator filled with methanol. eration voltage of 5 kV and an emission current of 10 �A.
The reactant mixture of He/O2/MeOH was adjusted to The samples were imaged without any metallic coating. The
8
6.3/9.7/4.0 (mol% ) or 79.3/9.7/10.0 (mol% ) by mass flow SEM micrographs were taken at many different places of
controllers. The typical performances were conducted at the sample.
�
1
� 1
GHSV = 10,000 ml h
g
. The outlet stream line from the
TPD spectra were measured in a fixed-bed reactor sys-
cat
reactor to the gas chromatgraph was heated at about 423 K tem equipped with a gas chromatograph. A dry-ice/acetone
to avoid condensation of reaction products. The products trap was used to eliminate the influence of water and hy-
were analyzed with an on-line gas chromatgraph using two drocarbons. SbRe2O6 was pretreated at 573 K for 1 h and
columns (3-m Porapak N and 3-m Unibeads C) at 423 K.
then cooled to 298 K in He. The O2 adsorption was carried

SbRe2O6 CATALYST FOR SELECTIVE OXIDATION
53
out at 298 K under He/O2 (90/10, molar ratio) in a flow
of 30 ml/min for 10 min. The sample was purged by a He
flow (40 ml/min) for 60 min. Then, O2-TPD spectra were
measured at a heating rate of 10 K/min.
TPR spectra were recorded in the same fixed-bed reac-
tion system as that for TPD. Again, a dry-ice/acetone trap
was used to eliminate the influences of water and hydrocar-
bons. The samples were exposed to Ar flow for 2 h at room
temperature and to a 5% H2/Ar flow of 30 ml/min. The TPR
spectra were recorded at a heating rate of 10 K/min.
3
. RESULTS
3
.1. Methanol Oxidation on Sb–Re–O Catalysts
The performances of several Sb–Re–O catalysts at 573 K
for the methanol conversion to methylal are shown in
Table 1. Sb oxides such as Sb2O3, Sb2O4, and Sb2O5, a me-
FIG. 1. Catalytic methanol oxidation on SbRe2O6 as a function of
�
1
� 1
reaction temperature; GHSV = 10,000 ml h
g
; He/O2/MeOH = 86.3/
-cat
9
.7/4.0(mol% );1 atm. ♦, methylal selectivity;� , dimethylether selectivity;
chanical mixture of Sb2O3 and Re2O7 (Re2O7 + Sb2O3), and 4, methyl formate selectivity; ᭢, reaction rate.
Re oxides supported on Sb2O3 and SiO2 (Re2O7/Sb2O3 and
Re2O7/SiO2, respectively) showed no or negligible activi-
ties for the methylal formation. The crystalline Sb–Re ox- 100% conversion of methanol corresponds to a reaction
�
3
� 1 � 1
ides, Sb4Re2O13 and SbOReO4 � 2H2O, produced almost rate of 16.4 � 10 mol h
gcat. The reaction rate and
no methylal either. Only SbRe2O6 among the crystalline methylal selectivity increased with increasing temperature
Sb–Re oxides synthesized thus far was active for selective up to 573 K, where the selectivity to methylal reached a
methanol oxidation to methylal. The methylal selectivity maximum of 92.5% . A main by-product was dimethyl ether,
was as high as 92.5% at a conversion of 6.5% at 573 K whose formation decreased with increasing temperatures.
(
Table 1).
The lower the GHSV, the higher the methanol conversion
Figure 1 shows the reaction rate and selectivity became, while selectivity to methylal was almost indepen-
for catalytic methylal synthesis on SbRe2O6 under dent of the GHSV.
�
1
� 1
GHSV = 10,000 mol h
g
and He/O2/MeOH = 86.3/9.7/
To check the catalytic performances of the three crys-
cat
4
.0 (mol% ) as a function of reaction temperature. The talline Sb–Re–O compounds at higher temperatures, we
TABLE 1
Methanol Oxidation over Several Sb–Re–O Catalysts at 573 K and on Other Typical Catalysts for Comparison
Selectivity (% )
Catalyst
SbRe2O6b
Conv. (% )
CH2(OCH3)2
HCHO
CH3OCH3
HCOOCH3
HCOOH
CO2
CO
Ref.
6.5
0
92.5
0
1.0
0
0
0
49.8
0
0
6.3
0
1.2
0
0
0
1.4
0
0
0
0
0
0
0
0
31.5
0
0
0
0
0
0
0
0
0
45.2
0
0
0
18.0
0
0
0
0
0
0
1.9
0
0
0
0
This work
This work
This work
This work
This work
This work
This work
This work
This work
6
b
Sb4Re2O13
b
SbOReO4 � 2H2O
Re2O7/Sb2O3
Re2O7/SiO2
Re2O7 + Sb2O3
Sb2O3
4.7
2.5
25.5
1.4
0
0
1.3
—
99.0
42.7
0.9
76.4
0
7.4
19.1
23.6
0
0
0
Sb2O4
0
0
Sb2O5
Trace
� 56
76.2
40
55.5
� 17
0
26.4
� 7
23.8
0
PMoH-5.75/SiO2c
� 20
0
0
0
d
2
mol% Mo/MCM-41
0.7
—
0
0
0
7
1
e
V/TiO2
60
0
0
a
Catalyst weight: 200 mg, pretreated under He at 573 K for 1 h before reaction. Reactant mixture: He/O2/MeOH = 86.3/9.7/4.0 (mol% ), GHSV =
�
1
� 1
g .
cat
1
0,000 ml h
-
b
2
� 3
� 1 � 1
g .
cat
Surface area = 1.0 m /g; reaction rate for SbRe2O6 = 1.06 � 10 mol h
-
c
d
e
� 1 � 1
g
Mo
Reaction temperature = 593 K; reactant mixture: He/O2/CH3OH = 85.2/10.3/4.5 (mol% ). Reaction rate = 0.52 mol h
.
-
� 3
� 1 � 1
Reaction temperature = 503 K; reactant mixture: He/O2/CH3OH = 76.0/16.3/7.7 (mol% ). Reaction rate = 0.74 � 10 ml h
g
.
cat
-
Reaction temperature = 473 K; reactant mixture: He/O2/CH3OH = 76/16.3/7.7 (mol% ). Reaction rate = 0.035 ml h� 1
g
� 1
.
-
cat

5
4
YUAN ET AL.
lower than 10 mol% in the reactant feed. When the O2 pres-
sure increased more, the selectivity decreased eventually to
4.0% at the 42 mol% O2 pressure, accompanied by an in-
8
crease in the dimethyl ether formation. The increase in the
concentration of methanol resulted in the increase in the
reaction rate. The selectivity to methylal was independent
of the concentration of methanol.
3
.3. Effects of Gaseous O2 on Methanol Oxidation
In methanol conversion on SbRe2O6 at 573 K in the
absence of O2, methylal was the only detectable product
with a selectivity of 100% in the effluent gas in the be-
ginning of the reaction and methanol conversion rapidly
decreased with time on stream. Then, methylal was not de-
tected as a main product, and instead, H2, CO2, CH4, and
HCOOCH3 were mainly produced. Note that the product
ratio of H2/CO2/CH4 after 2 h was close to 2/1/1 (molar
ratio).
Figure 3 shows methanol oxidation on SbRe2O6 at 573 K
in the absence of O2 for 70 min and then in the presence of
O2. After the methanol conversion dropped to nearly zero,
oxygen was introduced into the reactant feed, resulting in
the recovery of the activity and methylal selectivity to the
level similar to that at the steady-state reaction.
3
.4. Pulse Reaction
The pulses of ca. 1 ml of He/MeOH = 96.0/4.0 (mol% )
were introduced onto SbRe2O6 at 573 K at an interval of
FIG. 2. (a) Catalytic methanol oxidation on SbRe2O6 as a function
of oxygen partial pressure (MeOH = 15.0 mol% in the reactant feed).
�
1
� 1
g . ᭢, reaction
cat
Reaction temperature = 573 K; GHSV = 5000 ml h
-
rate; ♦, methylal selectivity; � , dimethyl ether selectivity. (b) Catalytic
methanol oxidation on SbRe2O6 as a function of methanol partial pres-
sure (O2 = 9.7 mol% in the reactant feed). Reaction temperature = 573 K;
�
1
� 1
cat
GHSV = 5000 ml h
g
. ᭢, reaction rate; ♦, methylal selectivity; � ,
-
dimethyl ether selectivity.
examined the conversion and selectivity at 673 K. It was
found that methanol conversion on SbRe2O6 at 673 K
increased to 86.2% , keeping a high selectivity of 85.4%
toward methylal. The performances of Sb4Re2O13 and
SbOReO4 � 2H2O were also improved to show 8.5% con-
version with 60.1% selectivity and 8.7% conversion with
5
3.0% selectivity, respectively.
3
.2. Effects of O2 and MeOH Pressures
Figures 2a and 2b show reaction rates and selectivities
for the methanol selective oxidation on SbRe2O6 at 573 K
as a function of O2 partial pressure and methanol partial
pressure, respectively. The reaction rate was almost inde-
pendent of O2 partial pressure. The selectivity to methylal
FIG. 3. Catalytic methanol oxidation on SbRe2O6 at 573 K as a
�
1
� 1
function of time on stream. GHSV = 5000 ml h
0.0 (mol% ) for 70 min and then He/O2/MeOH = 79.3/9.7/10.0 (mol% ). ᭜,
methanol conversion; ♦, methylal selectivity; � , dimethyl ether selectivity;
g
; He/MeOH = 90.0/
-cat
1
was kept constant at about 92–93% in the O2 pressure range 4, methyl formate selectivity.

SbRe2O6 CATALYST FOR SELECTIVE OXIDATION
55
FIG. 4. Methanol pulse reactions on SbRe2O6 at 573 K. ᭜, methanol
conversion; ♦, methylal selectivity.
about 15 min. Figure 4 is the plots of the methanol conver-
sion and the selectivity to methylal versus the number of
methanol pulses. The selectivity to methylal was as high as
FIG. 6. XPS spectra for fresh SbRe2O6 (a), SbRe2O6 after the
methanol oxidation at 573 K for 1 h (b), SbRe2O6 after the methanol
oxidation at 593 K for 1 h (c), and SbRe2O6 after the methanol oxidation
9
9.0% throughout the 19 pulses, which reproduces the re- at 673 K for 3 h (d).
sults of the catalytic oxidation of methanol in the absence of
oxygen in Fig. 3. The methanol conversion decreased with
an increase in the number of pulses, passed a plateau from
the fourth pulse to the seventh pulse, and decreased notably
with increasing pulses.
5
93, and 673 K. The XRD patterns of the fresh SbRe2O6 and
the SbRe2O6 after the catalytic reactions at 573 and 593 K
were similar to each other. But there existed an additional
�
broad diffraction peak at 2� � 31 for the SbRe2O6 catalyst
after the reaction at 673 K.
3
.5. Characterization by XRD, XPS, Raman, and SEM
Figure 5 depicts the XRD patterns of the SbRe2O6 cata-
lyst before and after the catalytic methanol oxidation at 573,
FIG. 5. XRD patterns for fresh SbRe2O6 (a), SbRe2O6 after the
methanol oxidation at 573 K for 1 h (b), SbRe2O6 after the methanol
oxidation at 593 K for 1 h (c), and SbRe2O6 after the methanol oxidation
at 673 K for 3 h (d).
FIG. 7. Confocallaser Raman spectra for fresh SbRe2O6 (a), SbRe2O6
after the methanol oxidation at 573 K for 1 h (b), SbRe2O6 after the
methanol oxidation at 593 K for 1 h (c), and SbRe2O6 after the methanol
oxidation at 673 K for 3 h (d); Sb2O3 (e) and Sb2O4 (f).

5
6
YUAN ET AL.
FIG. 8. Scanning electron micrographs of fresh SbRe2O6 (A), SbRe2O6 after the methanol oxidation at 573 K for 1 h (B), SbRe2O6 after the
methanol oxidation at 593 K for 1 h (C), and SbRe2O6 after methanol oxidation at 673 K for 3 h (D). Scale bars are shown in each SEM photograph.
Figure 6 shows the Re 4f XPS spectra for the fresh agrees with the chemical formula of SbRe2O6 containing
SbRe2O6, and the SbRe2O6 samples after methanol oxi- Re ions with an oxidation state of 4.5+. The SbRe2O6 sam-
dation at 573, 593, and 673 K. A small peak at 47.7 eV ples before and after methanol oxidation at 573 and 593 K
may be due to Re 4f5/2 for Re6 species, while a strong exhibited no significant difference in the Re 4f XPS spec-
+
peak at 42.3 eV may be referred to Re 4f7/2 for reduced Re tra (Figs. 6b and 6c). However, the SbRe2O6 sample af-
species, likely Re4
� 5+
(38, 39). The most intense peak at ter methanol oxidation at 673 K showed weakness in the
about 44.7 eV is then considered as the sum of a Re 4f7/2 Re 4f XPS peak intensities (Fig. 6d), indicating a change in
peak for Re6 species and a Re 4f5/2 peak for Re
The Sb 4d and Sb 3d3/2 bands appeared at the binding en-
ergies of 34.4 and 539.7 eV, respectively, which are the val- catalyst before exposure to methanol and the SbRe2O6
+
4� 5+
species. the catalyst surface.
Figure 7 shows the Raman spectra for the fresh SbRe2O6
ues typical of Sb3 . The observation of Re
+
4.5+
XPS peaks catalysts after methanol oxidation at 573, 593, and 673 K.

SbRe2O6 CATALYST FOR SELECTIVE OXIDATION
57
FIG. 8—Continued
The Raman spectrum for the fresh dehydrated SbRe2O6 3 �m in dimension and about 100 nm in thickness. The basal
catalyst showed bands at 152 sh, 171, 227 vs, 239 vs, 267, 282, (100) faces were smooth and had sharp and regular edges.
�
1
3
08, 347, 391, 416 sh, 525, and 756 cm . After methanol ox- Also, after selective methanol oxidation at 573 and 593 K,
idation at 573 and 593 K, there was no significant change in the SbRe2O6 crystals exhibited almost the same morphol-
the bands. The SbRe2O6 sample after the catalytic methanol ogy as the fresh sample. However, at 673 K the morphology
oxidation at 673 K, however, gave a new intensive Raman granulated a little, though the crystals still maintained their
�
1
band at 201 cm , accompanied by an increase in intensity regular shape.
�
1
of the broad band at 391 cm
.
Figures 8A–8D show the SEM micrographs for SbRe2O6
before and after the selective methanol oxidation at 573,
3
.6. TPD and TPR
5
93, and 673K, respectively. The fresh SbRe2O6 catalyst was
The O2 TPD spectra were obtained with the SbRe2O6
composed of crystals possessing square basal faces with 0.5– samples: (a) after pretreatment at 573 K in He and O2-

5
8
YUAN ET AL.
dation of methanol at 573 K in the absence of gaseous oxy-
gen for 60 min as shown in Fig. 9b. The two oxygen species
that were exhausted in methanol oxidation were regener-
ated during catalytic methanol oxidation as evidenced in
Fig. 9c.
Figure 10 shows the TPR spectra for SbRe2O6,
Sb4Re2O13, SbOReO4 � 2H2O, 10 wt% Re2O7/Sb2O3,
Sb2O3, and 10 wt% Re2O7/SiO2. The SbRe2O6 gave one
TPR peak at about 790 K, which was lower than the TPR
peak for Sb2O3 but higher than that for the SiO2-supported
Re2O7. Sb4Re2O13 and SbOReO4 � 2H2O showed totally
different TPR features from that of the SbRe2O6. The
reduction of Sb4Re2O13 occurred above 800 K and
SbOReO4 � 2H2O was reduced by two steps at about 610 K
and above 800 K.
4
. DISCUSSION
FIG. 9. O2 TPD spectra for fresh SbRe2O6 (a), SbRe2O6 after the
methanol oxidation at 573 K for 60 min in the absence of oxygen (b), and
SbRe2O6 after the methanol oxidation at 573 K for 70 min in the absence
of oxygen and then in the presence of oxygen for 10 min (c).
4
.1. Catalytic Performances of Three Crystalline
Sb–Re–O Compounds for the Selective Oxidation
of Methanol to Methylal
adsorption at 298 K, (b) after the methanol oxidation at
Many catalysts have been screened to improve the per-
5
73 K in the absence of oxygen for 60 min, followed by formance for catalytic methanol oxidation, but little is
cooling to 298 K in He, and (c) after the methanol oxi- known about the catalytic property of Sb–Re–O com-
dation at 573 K in the absence of oxygen for 60 min and pounds. The catalytic methylal synthesis by methanol ox-
next in the presence of oxygen (He/O2/MeOH = 79.3/9.7/ idation has been proposed on V/TiO (2), V–Mo–O (5),
2
1
0.0 (mol% )) for 10 min, followed by cooling to 298 K in PMoH-5.75/SiO (6), Mo/MCM-41 (7), and electrocatalysts
2
He. All the TPD experiments were carried out at a heating (9), but good achievements have not been reported so far
rate of 10 K/min. The results are shown in Fig. 9.
(Table 1). As listed in Table 1, SbRe O among the Sb–
2
6
The TPD spectrum (Fig. 9a) showed the desorption peaks Re–O catalysts employed is the only catalyst that has ex-
at about 650 and 830 K, which indicates the existence of cellent performance in the catalytic selective oxidation of
two types of lattice oxygen species. The intensity of the two methanol to methylal. The 2 mol% Mo/MCM-41 catalyst
peaks decreased in similar manners after the catalytic oxi- was reported to show high selectivity of 76.2% at a low
conversion of 0.7% at 543 K, but rapid deactivation was
serious due to a significant leaching of Mo species from
MCM-41 (7). We found that the new SbRe2O6 catalyst is
promising in the process for catalytic methylal synthesis
from methanol with good selectivity of 92.5% (Table 1).
The increase in the surface area of SbRe2O6 as well as the
coverage of adsorbed methoxy species/methanol may im-
prove the catalytic performance.
4
.2. Structure of SbRe2O6 before and after the Catalytic
Methanol Oxidation
The structure of SbRe2O6 consists of layers of the
composition [Re2O6]3 and antimony ions connecting the
layers, sharing all corners along the (100) axis, as shown in
Fig. 11. The (100) plane tends to grow (Fig. 8). The formal
oxidation states of rhenium and antimony in SbRe2O6 are
�
4
.5+ and 3+, respectively. The mixed valent state of the
rhenium makes Re–O bonds weaker and thus leads to a
softer lattice (14).
FIG. 10. TPR profiles for (a) fresh SbRe2O6, (b) fresh Sb4Re2O13,
c) fresh SbOReO4 � 2H2O, (d) 10 wt% Re2O7/Sb2O3, (e) Sb2O3, and
f) 10 wt% Re2O7/SiO2.
The results of XRD, XPS, Raman, and SEM demonstrate
that the structure of SbRe2O6 during the catalytic methanol
(
(

SbRe2O6 CATALYST FOR SELECTIVE OXIDATION
59
4
.3. Active Oxygen Species of SbRe2O6 for Selective
Methylal Formation
When SbRe2O6 was exposed to methanol in the absence
ofoxygen, methylalwasformed over 60min and then the re-
action products switched to H2, CO2, CH4, and HCOOCH3.
The molecular ratio of H2/CO2/CH4 was found to be ap-
proximately 2/1/1. The results suggest that the lattice oxy-
gen in the crystalline SbRe2O6 was active and selective for
the methanol oxidation to methylal:
3
CH3OH + [O] → CH2(OCH3)2 + 2H2O.
[2]
As shown in Fig. 4, a series of pulsed reactions of methanol
in the absence of O2 also indicates that lattice oxygen atoms
FIG. 11. A representative SbRe2O6 (100) plane; Re ions are located of the SbRe O catalyst contribute to the selective methylal
2
6
inside oxygen octahedra and Sb ions are shown with black scale. Small
white circles are O ions.
synthesis. Figure 4 shows a plateau with the following no-
table decrease in the methanol conversion, which suggests
that the surface lattice oxygen atoms consumed by the re-
2
�
oxidation at the temperatures below 600 K remained un- action with methanol can be replenished by bulk oxygen
changed. Another crystalline compound, Sb4Re2O13, was in SbRe O at 573 K, though the amount of replenished
2
6
also stable in the identical reaction conditions as charac- oxygen atoms is not so much.
terized by XRD, XPS, and SEM, but it showed no cata-
At the SbRe O surface where the active lattice oxy-
lytic activity for the selective methanol oxidation at 573 K gen atoms had been consumed, formaldehyde produced by
Table 1).
methanol dehydrogenation (CH OH → HCHO + H ) re-
Although the Re 4f XPS binding energies reported for acted with methanol to produce HCOOCH by the dehy-
2
6
(
3
2
3
Re species so far are scattered in the literature (38, 39), the drogenative condensation (Eq. [3]) rather than the dehy-
XPS spectra for SbRe2O6 in Fig. 6 indicate the presence of drative condensation to form methaylal:
4
� 5+
Re
species, which coincides with the chemical formula
6+
HCHO + CH3OH → HCOOCH3 + H2.
[3]
of SbRe2O6. The minor peaks for Re species in Fig. 6 sug-
gest that a part of Re4
� 5+
species at the catalyst surface is
Methyl formate was easily decomposed to CO2 and CH4 at
the reduced catalyst surface. Thus, the overall stoichiometry
of the surface reaction becomes H2/CO2/CH4 = 2/1/1.
When O2 was admitted into the reactant feed after the
active lattice oxygen was exhausted, the methylal synthesis
activity of SbRe2O6 was restored as shown in Fig. 3, which
demonstrates that the active lattice oxygen species were
regenerated by gaseous oxygen.
The TPD experiments in Fig. 9 reveal that there are two
types of lattice oxygen atoms; one desorbs at 663 K and the
other desorbs at 723 K. We may straightforwardly attribute
these two lattice oxygen species to a Re oxide site and Sb
oxide site in SbRe2O6 from the comparison with TPD of
Re2O7/SiO2 and Sb2O3 (not shown), respectively. These two
TPD peaks decreased in intensity similarly after methanol
oxidation in the absence of oxygen for 60 min (Fig. 9), which
indicates that both of the two lattice oxygen species are
responsible for selective methanol oxidation to methylal.
When oxygen was introduced into the reactant feed, those
decreased TPD peaks (Fig. 9b) developed again (Fig. 9c),
which coincides with the results in Fig. 3.
oxidized. It was observed that SbRe2O6 decomposed partly
under the catalytic methanol oxidation at 673 K. At this
reaction temperature a morphological change occurred as
imaged by SEM (Fig. 8D). The XPS spectrum for the sam-
ple after the reaction at 673 K in Fig. 6d also indicates a
surface change of the crystal, probably due to oxidation
7+
of Re species to Re species on the surface of SbRe2O6.
The amount of Re7 species must be very small due to the
+
volatile propertyofRe7 oxide species(25), so that the Re
+
7+
species was not definitely detected by XPS and Raman. The
surface decomposition of SbRe2O6 and the volatilization of
7+
Re oxide species caused a decrease in the intensity of the
XPS Re 4f peaks. By comparison with the Raman bands for
�
1
Sb2O4, the new Raman bands at about 201 and 391 cm for
the 673 K reacted sample are referred to as Sb2O4 forma-
tion, although Sb2O4 was not detected by XRD (Fig. 5d).
Thus, the decomposition at the SbRe2O6 surface at 673 K
may be postulated by Eq. [1]. The Re2O7 species in situ
formed under methanol oxidation at O2 partial pressures
above 10 mol% in the reactant feed or at 673 K resulted in
an increase in dimethyl ether formation due to the higher
acidity of Re2O7 compared to that of the lower valent Re
species.
The amount of the active lattice oxygen in SbRe2O6 can
be estimated from the TPD results. Also, according to the
reaction formula for the methylal formation from methanol
[1] (Eqs. [4] and 5), it is possible to calculate the amount of the
4SbRe2O6 + 6O2 → 4Re2O7 + 2Sb2O4.

6
0
YUAN ET AL.
TABLE 2
oxide species by the specific connection with the Sb oxide
chains may be the key issues relevant to selective oxida-
tion of methanol to methylal. The TPR results for SbRe2O6
in Fig. 10 showed that one reduction peak at 790 K, sug-
gesting the occurrence of a distinct connection between the
Re oxide and Sb oxide, which is different from the other
two Sb–Re–O compounds. The well-organized bifunctions
of the moderate redox and acidic property at the SbRe2O6
surface may be benefited from the specific valency and ar-
rangement.
The Amount of Active Lattice Oxygen Species for the Methanol
Oxidation in SbRe2O6 Estimated from Reaction and TPD Data
Active oxygen amount
atom/m2
mmol/g-cat
2
19
19
From the reaction (Fig. 3)
From the TPD (Fig. 9)
3.8 � 10�
2.4 � 10�
2.4 � 10
1.5 � 10
2
active lattice oxygen in Fig. 3 by curve simulation.
5
. CONCLUSIONS
CH3OH + [O] → HCHO + H2O
HCHO + 2CH3OH → CH2(OCH3)2 + H2O.
[4]
(
1) The novel compound SbRe2O6 showed good per-
[5] formance with high selectivity of 92.5% for the catalytic
methanol oxidation to methylal at 573 K. The reaction rate
increased with increasing methanol partial pressure but was
independent of oxygen partial pressure. The high selectivity
to methylal was independent of the methanol concentration
and the O2 concentration (<10 mol% ) in the reactant feed.
The calculated results from Figs. 3 and 9 are shown in
Table 2. The difference between the two experiments may
be due to the experimental errors since the amount of
the active lattice oxygen was substantially small. The small
amount of the active lattice oxygen in SbRe2O6 may be due
(
2) No structural change in the bulk and surface of
2
probably to its low surface area of 1 m /g. The amount of the
SbRe2O6 was observed during and after methanol oxida-
tion at 573 K by means of XRD, Raman, XPS, and SEM.
The SbRe2O6 surface benefited from the specific Re valency
and arrangement provided high activity and selectivity to
methylal.
19
2
active oxygen atoms corresponds to 1.5–2.4 � 10 atom/m ,
which falls into the range of oxygen concentration at solid
oxide surfaces (40).
4
.4. Redox and Acid–Base Catalysis of SbRe2O6
(
3) The pulse reactions showed that the lattice oxygen
From the mechanism of methanol oxidation reported on species in the SbRe2O6 contributed to the selective ox-
unsupported MoO3 (41) and V–Ti–O systems (42), the first idation of methanol to methylal. The TPD experiments
reaction step has been demonstrated to be the formation demonstrated that there existed two types of lattice oxy-
of methoxy species (42, 43) by dissociative adsorption of gen, both being responsible for the methylal synthesis. The
methanol on a dual acid–base site formed by accessible lattice oxygen atoms exhausted in the methylal formation
cation and surface oxygen ion. Further transformation of were regenerated by gaseous O2 during catalytic methanol
the adsorbed methoxy will depend on the redox property oxidation.
and acid–base strength of active centers on which it ad-
sorbs. Desorption of the formaldehyde produced by de-
hydrogenation of the methoxy species is favored on weak
ACKNOWLEDGMENT
acid sites more than on strong acid sites. If the acid sites
This work has been supported by Core Research for Evolutional
are strong, the residence time of formaldehyde at the sur- Science and Technology (CREST) of Japan Science and Technology Cor-
poration (JST).
face becomes long enough to form dioxymethylene species
(
42), and the dioxymethylene species can react with neigh-
boring methoxy species and/or adsorbed methanol to form
methylal (44). If both acidic and basic sites are too strong,
the dioxymethylene species are oxidized to formates, which
also react with methanol to form methyl formate molecules,
or are further oxidized to carbon dioxides (42). If strong
acid sites and very weak basic sites are present at catalyst
surfaces, only dimethyl ether is formed (45). The present re-
sultsindicate that SbRe2O6 can fit the requirementsfor both
REFERENCES
1. Tatibou e¨ t, J. M., Appl. Catal. A 148, 213 (1997).
. Busca, G., Elmi, A. S., and Forzatti, P., J. Phys. Chem. 91, 5236 (1987).
. Ai, M., J. Catal. 77, 279 (1982).
. Arora, N., Deo, G., Wachs, I. E., and Hirt, A. M., J. Catal. 159, 1 (1996).
. Tatibou e¨ t, J. M., and Germain, J. E., Bull. Soc. Chim. Fr. I 9–10, 343
2
3
4
5
(
1980).
6. (a) Fournier, M., Aouissi, A., and Rocchiccioli-Deltchff, C., J. Chem.
Soc. Chem. Commun. 307 (1994). (b) Rocchiccioli-Deltcheff, C.,
Aoussi, A., Launary, S., and Fournier, M., J. Mol. Catal. A: Chem.
“
redox” and “acid–base” catalyses for selective methanol
oxidation to methylal at 573 K. Only the SbRe2O6 among
the three crystalline Sb–Re–O compounds showed high
performance for catalytic methylal synthesis. The oxida-
tion state of Re oxide species and the stablization of the Re
1
14, 331 (1996).
7
. Shannon, I. J., Maschmeyer, T., Oldroyd, R. D., Sankar, G., Thomas,
J. M., Pernot, H., Baalikdjian, J. P., and Che, M., J. Chem. Soc. Faraday
Trans. 94, 1495 (1998).

SbRe2O6 CATALYST FOR SELECTIVE OXIDATION
61
8
. (a) Tatibou e¨ t, J. M., and Germain, J. E., J. Catal. 72, 375 (1981).
b) Tatibou e¨ t, J. M., and Germain, J. E., and Volta, J. C., J. Catal.
2, 240 (1983).
. For example, Otsuka, K., and Yamanaka, I., Appl. Catal. 26, 401(1986).
25. Wang, C.-B., Cau, Y., and Wachs, I. E., Langmuir 15, 1223 (1999).
26. Jehng, J.-M., Hu, H., Gao, X., and Wachs, I. E., Catal. Today 28, 335
(1996).
27. Albonetti, S., Cavani, F., and Trifiro, F., Catal. Rev.-Sci. Eng. 38, 413
(1996).
28. Nilsson, R., Lindblad, T., and Andersson, A., J. Catal. 148, 501 (1994).
29. Zanthoff, H.-W., and Buchhloz, S. A., Catal. Lett. 49, 213 (1997).
30. Grasselli, R. K., and Suresh, D. D., J. Catal. 25, 272 (1972).
31. Genti, G., and Trifiro, F., Catal. Rev.-Sci. Eng. 28, 165 (1986).
32. Berry, F. J., Adv. Catal. 30, 97 (1981).
(
8
9
1
1
1
0. Gaigneaux, E. C., Liu, H., Imoto, H., Shido, T., and Iwasawa, Y., Top.
Catal. 11–12, 185 (2000).
1. Liu, H., Gaigneaux, E. C., Imoto, H., Shido, T., and Iwasawa, Y.,
J. Phys. Chem. 104, 2033 (2000).
2. Harrison, W. T. A., McManus, A. V. P., Kaminsky, M. P., and Cheetham,
A. K., Chem. Mater. 3, 1631 (1993).
1
1
3. Watanabe, H., and Imoto, H., Inorg. Chem. 36, 4610 (1997).
4. Watanabe, H., Imoto, H., and Tanaka, H., J. Solid State Chem. 138, 245
(
33. Inoue, T., Asakura, K., and Iwasawa, Y., J. Catal. 171, 184 (1997).
34. Inoue, T., Asakura, K., and Iwasawa, Y., J. Catal. 171, 457 (1997).
35. Inoue, T., Oyama, S. T., Imoto, H., Asakura, K., and Iwasawa, Y., Appl.
Catal. A 191, 131 (2000).
36. Yuan, Y., Liu, H., Imoto, H., Shido, T., and Iwasawa, Y., Chem. Lett.
674 (2000).
1998).
5. Rae Smith, A. R., and Cheetham, A. K., J. Solid State Chem. 30, 345
1979).
1
(
1
1
6. Sleight, A. W., Inorg. Chem. 14, 597 (1975).
7. Butz, A., Miehe, G., Paulus, H., Strauss, P., and Fuess, H., J. Solid State
Chem. 138, 232 (1998).
37. Castillo, R., Dewaele, K., Ruiz, P., and Delmon, B., Appl. Catal. A 153,
L1 (1997).
1
8. Moulijn, J. A., and Mol, J. C., J. Mol. Catal. 46, 1 (1988). Mol, J. C., and
Moulijn, J. A., “Advances in Catalysis” (D. D. Eley, H. Pines, and P. B.
Weisz, Eds.), Vol. 24, p. 131. Academic Press, New York, 1975.
9. Olsthoorn, A. A., and Boelhouwer, C., J. Catal. 44, 207 (1976).
0. Pecoraro, T. A., and Chianelli, R. R., J. Catal. 67, 430 (1981).
1. Thomas, R., Van Oers, E. M., De Beer, V. H. J., Medema, J., and
Moulijn, J. A., J. Catal. 76, 241 (1982).
38. Cimino, A., De Angelis, B. A., Gazzoli, D., and Valigi, M., Z. Anorg.
Allg. Chem. 460, 86 (1980).
39. Shpiro, E. S., Avaev, V. I., Ryashentseva, M. A., and Minachev,
Kh. M., J. Catal. 55, 402 (1978).
40. Winter, E. R. S., J. Chem. Soc. A 2889 (1968).
41. Chung, J. S., Miranda, R., and Bennertt, C. O., J. Catal. 114, 354 (1988).
42. Busca, G., Catal. Today 27, 457 (1996).
1
2
2
2
2
2
2. Blom, R. H., Kollonitsch, and Kline, C. H., Ind. Eng. Chem. 54, 17
43. Liu, Y. C., Griffin, G. L., Chan, S. S., and Wachs, I. E., J. Catal. 94, 108
(1985).
44. Yang, T. J., and Lunsford, J. H., J. Catal. 103, 55 (1987).
45. Chauvin, C., Saussey, J., Lavalley, J. C., Idriss, H., Hindermann, J. P.,
Kienneman, A., Chaumette, P., and Courty, P., J. Catal. 121, 56 (1990).
(
1962).
3. Davenport, W. H., Kolonitsch, W., and Kline, C. H., Ind. Eng. Chem.
0, 10 (1968).
4. Kim, D. S., and Wachs, I. E., J. Catal. 141, 419 (1993).
6