Selective oxidation of methane to formaldehyde over antimony oxide-loaded catalyst

Article abstract of DOI:10.1016/j.molcata.2006.01.043

A possibility of antimony oxide as a catalyst for the selective oxidation of methane with oxygen to formaldehyde was investigated. The activity measurement was carried out at an atmospheric pressure and at 873 K, where the homogeneous gas-phase reaction w

Full text of DOI:10.1016/j.molcata.2006.01.043

Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
Selective oxidation of methane to formaldehyde
over antimony oxide-loaded catalyst
Hiroyuki Matsumuraa, Kimito Okumuraa, Takahiro Shimamuraa,
Na-oki Ikenagaa, Takanori Miyakea,b, Toshimitsu Suzukia,b,∗
a Department of Chemical Engineering, Kansai University, Suita, Osaka 564-8680, Japan
b High Technology Research Center, Kansai University, Suita, Osaka 564-8680, Japan
Received 8 June 2005; received in revised form 23 January 2006; accepted 24 January 2006
Available online 28 February 2006
Abstract
A possibility of antimony oxide as a catalyst for the selective oxidation of methane with oxygen to formaldehyde was investigated. The activity
measurement was carried out at an atmospheric pressure and at 873 K, where the homogeneous gas-phase reaction was negligible. Oxidized
diamond (O-Dia)-supported antimony oxide catalyst produced 1.3 mmol h−1 g-cat−1 of formaldehyde with a formaldehyde selectivity of 23%. On
the other hand, SiO2 supported antimony oxide catalyst exhibited negligible catalytic activity. XRD and UV–vis analyses revealed that ␣-Sb2O4
was formed on the oxidized diamond while Sb6O13 was formed on SiO2. Selective oxidation of methane to formaldehyde seemed to proceed on
␣-Sb2O4 with moderate activity and selectivity to formaldehyde, via a redox cycle of ␣-Sb2O4 and Sb2O4−x. On the other hand, Sb6O13 on SiO2
was stable under the reaction conditions and the selective oxidation occurred only slightly.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Oxidation; Methane; Formaldehyde; Oxidized diamond; XRD; Redox cycle
1. Introduction
tion. Since the oxidation of methane requires a high reaction
temperature to activate relatively inert methane on catalysts,
deep oxidations of primary products to give COx occur more
easily than the formation of oxygenates because of the weaker
C–H bond in formaldehyde than that in methane [20]
Formaldehyde is widely used as a raw material for resins
and adhesives, and is industrially produced in a three-step pro-
cess. Therefore, direct conversion of methane with oxygen to
formaldehyde is one of the most challenging chemistry in the
field of catalytic science.
Selective oxidation of methane to formaldehyde has been
carried out with supported transition metal oxides, particu-
larly molybdenum oxide [1–8], iron oxide [9,10] and vanadium
oxides [1,2,11–14]. Vanadium oxide on MCM-41or -48 [15] and
vanadium and iron oxides on MCM-41 were reported to exhibit
high performance [16].
Tabata et al. reported that addition of NOx as an oxidant pro-
moted the selective oxidation of methane to formaldehyde and
methanol with MoO3 and MoO3/SiO2 catalysts [17–19].
Although a large number of catalysts have been reported,
none of them has satisfied the requirements for commercializa-
CH4 + O2 → HCHO + H2O, ꢀH◦298 = −319 kJ/mol
HCHO + O2 → CO2 + H2O, ꢀH◦298 = −571 kJ/mol
(1)
(2)
We have recently reported that oxidized diamond (O-Dia)-
supported catalysts exhibit high catalytic performance in various
reactions [21–23]. The surface of O-Dia is considered as a
pseudo-solid carbon oxide phase which does not exist above
ambient temperature.
Ni- and Co-loaded O-Dia catalysts showed significant cat-
alytic activity in the partial oxidation of methane to the synthesis
gas [22]
CH4 + 1/2O2 → CO + 2H2, ꢀH◦298 = −36 kJ/mol
(3)
∗
Furthermore, we have first reported that V2O5-loaded O-Dia
catalystproducedformaldehydeandacetaldehydefrommethane
Corresponding author. Fax: +81 6 6388 8869.
E-mail address: tsuzuki@ipcku.kansai-u.ac.jp (T. Suzuki).
1381-1169/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2006.01.043

H. Matsumura et al. / Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
123
and ethane with CO2 as a mild oxidant [23–25]
2.1.3. Preparation of catalysts
All the catalysts were prepared by a usual impregnation
method. Antimony oxide-loaded O-Dia catalysts were prepared
by impregnating O-Dia powder with an aqueous solution of anti-
monic acid and oxalic acid. The suspension was kept stirring and
refluxing for 3 h, followed by stirring at room temperature for
24 h. Excess water was removed under vacuum, and the obtained
catalyst precursor was calcined at 723 K for 3 h in the presence
of air.
The loading level of antimony oxide onto supports is
expressed by weight percentages of oxides; wt.% = (weight of
metal oxide)/(weight of metal oxide and support) × 100.
CH4 + 2CO2 → HCHO + 2CO + H2O,
ꢀH◦298 = +246 kJ/mol
(4)
(5)
C2H6 + 2CO2 → CH3CHO + 2CO + H2O,
ꢀH◦298 = +195 kJ/mol
In these reactions, weak adsorption of produced oxygenates on
the catalyst surface decreased chances of deep oxidation of the
products.
Antimony oxides are widely used as a key promoter in mixed
oxide catalysts for the ammoxidation of propane to acryloni-
trile [26–36] and the partial oxidation of propane or propene
to acrylic acid [37–41]. Antimony oxides are considered to pro-
motetherate-limitingabstractionofhydrogenfromhydrocarbon
molecules, aswellasthesubsequentoxygeninsertionintoallylic
intermediate species [38]. However, only one paper dealt with
the antimony oxide single phase catalyst in the oxidation of
methane to formaldehyde [42]. Methane conversion of 0.4–1.0%
with formaldehyde selectivity less than 30% was obtained at a
low space velocity.
This paper deals with antimony oxide based catalysts for the
selective oxidation of methane to formaldehyde. We have found
that antimony oxide loaded on O-Dia exhibited moderate cat-
alytic activity for the oxidation of methane to formaldehyde with
a reasonable selectivity.
2.2. Activity measurement
A fixed-bed flow type reactor (quartz, i.d. 10 mm) was used
for measurement of the catalyst performance at an atmospheric
pressure. A 100 mg of catalyst was placed in the center of the
reactor tube with quartz wool plugs. A reaction gas mixture con-
sistings of CH4 and O2 was passed through the catalyst bed a
flow rate of 30 mL/min at an atmospheric pressure (space veloc-
ity = 18,000 mL h−1 g-cat−1). Oxygenates were analyzed by an
on-line high speed gas chromatograph (Chrompack CP-2002,
GLScienceCorp.)usingathermalconductivitydetectorwithCP
CIL 5CB and HYESEP A columns. CO and CO2 were analyzed
by a Shimadzu GC8AIT gas chromatograph, equipped with a
TCD detector using an activated carbon column (3 mm × 2 m).
2. Experimental
2.3. Characterization
2.1. Materials
UV–visabsorptionspectraofthecatalystswereobtainedwith
a diffuse reflectance mode by diluting the catalyst with SiO2
using a Jasco model V-550.
X-ray diffraction (XRD) measurements were performed with
aShimadzuXRD-6000withmonochromatizedCuK␣radiation.
X-ray photoelectron spectra were obtained with a Jeol model
JPS900 using Mg K␣ radiation.
As support materials, powdered diamond (General Electric
Co., surface area (SA) = 23 m2/g), SiO2 (Fuji Silicia Chem.
Co., SA = 110, 208 m2/g, and Fuso Chemicals, SA = 12 m2/g),
GeO2 (Kishida Chemicals, SA < 1 m2/g), Al2O3 (Sumitomo
Chemical Ind., SA = 159 m2/g), MgO (Ube Materials Ind.,
SA = 144 m2/g), TiO2 (Japan aerosil, SA = 50 m2/g) and ZrO2
(prepared by themal decomposition of Zr(OH)4, SA = 46 m2/g)
were used. Antimonic acid (H3SbO4) was used as the antimony
oxide source. Commercial SbCl5, Sb2O5, Sb2O4 and Sb2O3
were purchased from Wako Chemicals.
3. Results and discussion
3.1. Effect of metal oxides on the selective oxidation of
methane to formaldehyde
2.1.1. Preparation of antimonic acid
Table 1 shows the effect of metal oxides on the selective
oxidation of methane to formaldehyde. Although V2O5- and
Bi2O3-loaded O-Dia catalysts produced large amounts of COx
(runs 2, 4), MoO3- and antimony oxide- loaded O-Dia catalysts
afforded considerable amounts of formaldehyde (runs 1, 3). In
particular, antimony oxide-loaded O-Dia catalyst gave a large
amount of formaldehyde with a moderate selectivity.
Antimonic acid was prepared by dropwise addition of SbCl5
to a 30% aqueous solution of H2O2 at 273 K. The white pre-
cipitate of antimonic acid was filtered off after ageing for 6 h at
273 K, washed three times with distilled water and then dried at
383 K overnight.
2.1.2. Preparation of oxidized diamond
Man-made industrial diamond of fine particles less than
0.5 ␮m in diameter was hydrogenated at 1173 K for 1 h under
pure H2 stream. The hydrogenated diamond was partially oxi-
dized at 723 K for 5 h under air stream. Thus obtained O-Dia
is known to have functional groups of C–O–C and C O on the
carbon surface [24,25,43].
3.2. Effect of various supports on the selective oxidation of
methane to formaldehyde
Since antimony oxide afforded the highest HCHO yield
among the oxides tested, the effect of various supports on the
selective oxidation of methane was examined. The results are

124
H. Matsumura et al. / Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
Table 1
Effect of loading metal on the selective oxidation of methane
Run
Loading metal
Conv. (%)
Yield (mmol h−1 g-cat−1
)
Selectivity (%)
HCHO
HCHO
CO
CO2
1.61
37.7
2.15
15.8
CO
CO2
1
2
3
4
Sb oxide
V2O5
MoO3
Bi2O3
0.88
7.91
0.84
2.65
1.31
0.16
0.36
0.05
2.59
23.9
0.3
7.1
0.3
47.0
33.1
51.0
2.2
29.2
66.6
41.9
97.5
18.8
2.61
0.36
Catalyst: 100 mg, loading loading: 3 wt.%, reaction temperature: 873 K, gas flow: 30 mL/min (CH4/O2 = 25/5), SV: 18,000 mL h−1 g-cat−1, reaction time: 90 min.
shown in Table 2. Group 14 oxides, SiO2 and GeO2, are selected
to compare with O-Dia, and Al2O3, MgO, TiO2 and ZrO2 are
selected as typical catalyst supports. As is shown, O-Dia sup-
ported catalyst exhibited the highest catalytic activity in the
selective oxidation of methane to formaldehyde among the sup-
port materials examined (run 5). SiO2 and GeO2 supported
catalysts afforded low activity (runs 6, 9), and Al2O3, MgO,
TiO2 and ZrO2 supported catalysts produced large amounts of
COx (runs 10–13). High surface area SiO2 was employed as sup-
ports in order to see the effect of surface area on the catalytic
performance of loaded SbOx. However, as seen in runs 7 and 8
in Table 2, no increases in HCHO yields were observed for the
high surface area SiO2
was observed with methane. However, ethane conversions were
slightly higher than those of methane, since dehydrogenation of
ethane to ethylene proceeded to some extent
C2H6 + 1/2O2 → C2H4 + H2O, ꢀH◦298 = −148 kJ/mol
(9)
O-Dia supported catalyst produced a considerable amount
of formaldehyde instead of the expected product, acetalde-
hyde (run 14). To study a possibility of decomposition of
acetaldehyde to formaldehyde, the reaction of acetaldehyde
was examined. The results are shown in Fig. 1. Without
catalyst, acetaldehyde decomposed to formaldehyde, CO and
CO2 (Fig. 1a). O-Dia supported catalyst exhibited a higher
acetaldehyde conversion than the thermal reaction without
catalyst at all the reaction temperatures. In the thermal reaction,
the formaldehyde selectivity decreased with increasing reaction
temperature, indicating that formaldehyde is decomposed to
CO at a high reaction temperature.
In the selective oxidation of lower alkanes to oxygenates, a
key step seems to be desorption of produced oxygenates from
the catalyst surface. When the desorption of oxygenates from the
catalyst is slow, the oxygenates tend to decompose to COx. Since
the O-Dia supported catalyst showed a moderate formaldehyde
selectivity even at a high reaction temperature, desorption of
formaldehyde from the catalyst surface would be rapid. Thus,
O-Dia supported catalyst could produce formaldehyde without
excessive decomposition into COx [23]
CH4 + 2O2 → CO2 + 2H2O, ꢀH◦298 = −801 kJ/mol
(6)
(7)
(8)
HCHO → CO + H2,
ꢀH◦298 = −1.87 kJ/mol
HCHO + O2 → CO2 + H2O, ꢀH◦298 = −571 kJ/mol
Basic supports, MgO and ZrO2, did not give even a trace
amount of formaldehyde (runs 12, 13). These results seem to
indicate that a support having weak interaction with antimony
oxide afforded a high activity for the selective oxidation of
methane.
3.3. Effect of various supports on the selective oxidation of
ethane to oxygenates
Table 3 shows the effect of various supports on the selec-
tive oxidation of ethane to oxygenates. The activity order of the
antimony oxide-loaded support materials was similar to what
CH3CHO → HCHO → CO, CO2
(10)
Table 2
Effect of various supports on the selective oxidation of methane
Run
Support
Conv. (%)
Yield (mmol h−1 g-cat−1
)
Selectivity (%)
HCHO
CO
CO2
1.61
0.26
0.04
0.17
0
3.98
5.19
1.20
HCHO
CO
CO2
5
6
7
8
9
10
11
12
13
O-Dia
SiO2
0.88
0.13
0.08
0.11
0.06
1.67
1.87
0.49
2.72
1.31
0.46
0.45
0.29
0.33
0.24
0
2.59
0.06
0.03
0.16
0
6.61
2.68
1.46
0.69
23.9
58.8
89.1
47.5
100.0
2.2
0
5.4
0
47.0
7.3
5.3
25.5
0
61.1
34.0
51.9
4.4
29.2
34.0
5.6
27.0
0
36.8
66.0
42.7
95.6
a
SiO2
SiO2
b
GeO2
Al2O3
MgO
TiO2
0.15
0
ZrO2
15.2
Catalyst: 100 mg, Sb2O4 loading: 3 wt.%, reaction temperature: 873 K, gas flow: 30 mL/min (CH4/O2 = 25/5), SV: 18,000 mL h−1 g-cat−1, reaction time: 90 min.
a
Surface area 101 m2/g.
Surface area 490 m2/g.
b

H. Matsumura et al. / Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
125
Table 3
Effect of various supports on the selective oxidation of ethane
Run
Support
Conv. (%)
Yield (mmol h−1 g-cat−1
)
Selectivity (%)
HCHO
HCHO
CH3CHO
CH4
C2H4
CO
CO2
1.72
0.05
0
11.5
15.8
CH3CHO
14
15
16
17
18
O-Dia
SiO2
GeO2
Al2O3
MgO
0.95
0.51
0.60
2.62
2.26
1.89
0.07
0.06
0.05
0
0.33
0.06
0.11
0.50
0.07
0.12
0.03
0
0.28
0.35
2.16
2.73
3.22
3.36
2.66
2.22
0.06
0.02
9.97
5.15
17.3
1.2
0.9
0.2
0
6.1
2.0
3.1
1.4
0.6
Catalyst: 100 mg, Sb2O4 loading level: 3 wt.%, reaction temperature: 823 K, gas flow: 30 mL/min (C2H6/O2 = 25/5), SV: 18,000 mL h−1 g-cat−1, reaction time:
90 min.
Fig. 1. Effect of reaction temperature on the decomposition of acetaldehyde (a) thermal reaction (b) oxidized diamond supported catalyst (Sb2O4 3%/O-Dia). Cata-
lyst = 100 mg, reaction temperature = 573–873 K, SV = 18,000 mL h−1 g-cat−1, gas flow = 30 mL/min (CH3CHO(0.95%-N2 balance)/O2/Ar = 3/5/22). (᭹) CH3CHO
conversion, (ꢀ) HCHO selectivity, (ꢀ) COx selectivity.
On the supports other than O-Dia, very small amounts of
oxygenates were obtained (runs 15–18). Again, Al2O3 and MgO
supported catalysts showed high ethane conversion with large
amounts of COx, indicating that complete oxidation of ethane
proceeded on these catalysts (runs 17, 18)
One possible reason for this seems to be combustion of O-Dia
support.
Fig. 2 shows TPO patterns of O-Dia and Sb2O4 loaded O-Dia
at different loading levels. Bare O-Dia was be oxidized at 773 K
C2H6 + 7/2O2 → 2CO2 + 3H2O, ꢀH◦298 = −1560 kJ/mol
(11)
3.4. Effect of antimonic acid-loading level on the selective
oxidation of methane to formaldehyde
Table 4 summarizes the effect of the antimonic acid-loading
level on the methane conversion, product yields and selectivities.
As described below, after calcianation of the as-prepared cata-
lyst, antimonic acid on O-Dia was transformed into Sb2O4. With
an increase in the loading level of Sb2O4, selective oxidation
proceeded and the yield of formaldehyde showed the maximum
at 3.0 wt.% (run 22). The CO and CO2 yields decreased with
an increase in the loading level. On the bare O-Dia support
without antimony oxide or catalysts with low loading of anti-
mony oxide, a large amount of CO2 was produced (runs 19, 20).
Fig. 2. TPO patterns of Sb2O4 loaded O-Dia sample: 100 mg, Ar: 25 mL/min,
O2: 5 mL/min, heting rate: 10 K/min to 873 K, holding time at 873 K: 20 min.
(a) O-Dia, (b) Sb2O4 (3 wt.%)/O-Dia, (c) Sb2O4 (7 wt.%)/O-Dia.

126
H. Matsumura et al. / Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
Table 4
Effect of loading level of Sb2O4 on the selective oxidation of methane
Run
Loading level (wt.%)
Conv. (%)
Yield (mmol h−1 g-cat−1
)
Selectivity (%)
HCHO
HCHO
CO
CO2
CO
0.2
4.3
36.3
47.0
45.0
43.3
41.6
CO2
19
20
21
22
23
24
25
0.0
1.0
2.0
3.0
3.3
–
–
0.00
0.33
0.82
1.32
1.10
1.02
0.93
0.11
0.98
3.06
2.59
1.79
1.67
1.76
48.00
22.70
4.56
1.61
1.10
1.17
1.55
0.0
1.4
9.6
23.9
27.5
26.3
21.9
99.8
94.5
54.1
29.2
27.5
30.4
36.6
1.31
0.88
0.66
0.63
0.69
7.0
10.0
Catalyst: 100 mg, Sb2O4 loading level: 3.3 wt.%, reaction temperature: 873 K (0 wt. % at 823 K), gas flow: 30 mL/min (CH4/O2 = 25/5), SV: 18,000 mL h−1 g-cat−1
reaction time: 90 min.
,
and while the temperature was kept at 873 K for 20 min, almost
all O-Dia was burnt out
C (diamond) + O2 → CO2
(12)
As the loading level of antimony oxide increased, the amount
of CO and CO2 formed decreased; this may be correspond-
ing to the fact that antimony oxide is used as a flame retardant
[44]. Therefore, antimony oxides might have a role in protecting
the O-Dia surface from combustion. Oxidation of O-Dia above
500 ◦C markedly decreased with loading of Sb2O4 (3–7 wt.%)
as revealed by TPO in O2 and Ar mixed gas (Fig. 2b, c).
In order to improve catalytic activity and selectivity, binary
systems of Sb2O4 combined with V2O5, Bi2O3 or MoO3 were
examined. However, all these additives did not promote selective
oxidation but decreased methane conversion.
3.5. Effect of reaction temperature on the selective
oxidation of methane to formaldehyde
Fig. 3. Time-on-stream of formaldehyde yield in the selective oxidation of
methane. Catalyst = 100 mg, Sb2O4 loading level = 3 wt.%, reaction temper-
ature = 873 K, gas flow = 30 mL/min (CH4/O2 = 25/5), SV = 18,000 mL h−1 g-
cat−1. (᭹) Sb oxide/O-Dia, (ꢀ) O-Dia, (ꢁ) Sb oxide/SiO2, (ꢀ) SiO2.
Table 5 shows the effect of the reaction temperature on the
selective oxidation of methane over the O-Dia supported cata-
lyst. Below 773 K, the reaction did not proceed and the selective
oxidation of methane to formaldehyde occurred at above 823 K
(run 26). The yield of formaldehyde markedly increased with
increasing reaction temperature, and the highest formaldehyde
yield was obtained at 873 K (run 27). The formaldehyde selec-
tivity, however, decreased with increasing reaction temperature,
since the consecutive oxidation to COx occurred. The antimony
oxide began to sublime above the reaction temperature of 873 K.
However, at 873 K the yield of HCHO did not change for at least
12 h over the Sb2O4 loaded O-Dia catalyst.
3.6. Comparison between antimony oxide-loaded oxidized
diamond and SiO2 catalysts
Fig. 3 shows formaldehyde yield against a time-on-stream
in the selective oxidation of methane to formaldehyde over O-
Dia and SiO2 supported catalysts. Here we used a low surface
area spherical SiO2 (SA = 12 m2/g) in order to compare with
O-Dia. Both SiO2 alone and antimony oxide supported on SiO2
Table 5
Effect of reaction temperature on the selective oxidation of methane
Run
Reaction temp. (K)
Conv. (%)
Yield (mmol h−1 g-cat−1
)
Selectivity (%)
HCHO
HCHO
CO
CO2
CO
CO2
26
27
28
823
873
923
0.20
0.66
1.71
0.46
1.10
1.54
0.37
1.79
5.16
0.43
1.10
3.96
36.6
27.5
14.6
29.4
45.0
48.3
34.0
27.5
37.1
Catalyst: 100 mg, Sb2O4 loading level: 3.3 wt.%, gas flow: 30 mL/min (CH4/O2 = 25/5), SV: 18,000 mL h−1 g-cat−1, reaction time: 90 min.

H. Matsumura et al. / Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
127
Fig. 5. IR spectra of SbOx loaded SiO2. (a) SiO2 (Merck), (b) SbOx/SiO2, (c)
SbOx/SiO2 after the selective oxidation of methane for 2 h at 873 K.
Fig. 4. XRD patterns of Sb2O4 loaded catalysts. Loading level 5 wt.%. (a) Fresh
Sb2O4 loaded O-Dia catalyst calcined at 723 K for 3 h. (b) After reaction of CH4
with O2 over (a), (c) fresh Sb2O4 loaded SiO2 catalyst calcined at 773 K for 3 h.
(d) After reaction of CH4 with O2 over (c).
in Fig. 5. In these spectra, a broad absorption band at 820 cm−1
ascribed to Sb6O13 (see Fig. 10 trace b) was observed in both
fresh and used catalysts (see Fig. 5).
The selective oxidation of methane to formaldehyde over var-
ious un-supported antimony oxides was carried out in order to
clarify the active species. Results are shown in Fig. 6. Both
produced similar amount of formaldehyde. Therefore, antimony
oxide on the SiO2 seems not to have catalytic activity. On the
other hand, O-Dia support alone did not produce formaldehyde,
and O-Dia supported antimony oxide produced a significant
amount of formaldehyde, indicating that only antimony oxide
on the O-Dia exhibited catalytic activity.
Zhang et al. obtained considerable yield of HCHO on SiO2
supported Sb2O5 catalyst in the selective oxidation of methane
with oxygen [42]. Such differences may partly be ascribed to
the differences in the antimony sources. They used chlorides of
antimony and we used antimonic acid as a source for impregna-
tion.
Fig. 4 shows the X-ray diffraction (XRD) patterns of O-Dia
and SiO2 supported catalysts before and after the reaction at
873 K. Powdered diamond has no pore structure; therefore all
antimony oxide species seems to be dispersed on the diamond
surface. O-Dia supported catalyst exhibited strong diffraction
peaks of ␣-Sb2O4 and weak diffraction peaks of Sb6O13. On the
other hand, SiO2 supported catalyst exhibited diffraction peaks
of Sb6O13. Both of the catalysts did not change their diffraction
patterns before and after the reaction. Cody et al reported that
heating of antimonic acid above 573 K afforded Sb6O13 whereas
␣-Sb2O4 phase appeared above 1000 K [45]. The formation of
Sb6O13 on SiO2 was further confirmed by the IR spectra shown
Fig. 6. Time-on-stream of formaldehyde yield in the selective oxidation of
methane. Catalyst = 100 mg, unsupported oxides were used, reaction temper-
ature = 873 K, gas flow = 30 mL/min (CH4/O2 = 25/5), SV = 18,000 mL h−1 g-
cat−1. (ꢀ) Sb2O3, (ꢀ) Sb6O13, (♦) Sb2O5. Sb2O4 indicates crystalline form of
the catalysts after the reaction.

128
H. Matsumura et al. / Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
Fig. 7. XRD patterns of unsupported antimony oxides before and after selective oxidation of methane. (a) Sb2O3, (b) after reaction over Sb2O3, (c)Sb6O13, (d) after
reaction over Sb6O13, (e) commercial Sb2O5, (f) after reaction over Sb2O5. Catalyst = 100 mg, reaction temperature = 873 K, gas flow = 30 mL/min (CH4/O2 = 25/5),
SV = 18,000 mL h−1 g-cat−1
.
commercial Sb2O3 and Sb2O5 produced about 0.3 mmol h−1 g-
cat−1 of formaldehyde, and the formaldehyde yield gradu-
ally increased in the initial 60 min over Sb2O3. On the other
hand, the formaldehyde yield on Sb6O13 decreased from 0.4 to
0.1 mmol h−1 g-cat−1 in the initial 15 min.
Fig. 7 shows the XRD patterns of un-supported antimony
oxides before and after the selective oxidation of methane at
873 K. As seen in Fig. 5, Sb2O3 was transformed into ␣-Sb2O4
during the reaction. Commercial Sb2O5 was identified as a
mixture of Sb6O13, ␣-Sb2O4 and Sb2O3. After the reaction,
the mixture was predominantly transformed into the ␣-Sb2O4
phase with several unidentified peaks. Powder X-ray diffrac-
tion patterns of antimony oxides are variable depending on the
morphology of crystal [46]. Commercial Sb6O13 was not trans-
formed into ␣-Sb2O4 and remained in its original form only
with growth in the crystallite size (sharper diffraction peaks).
These results suggest that the active form of antimony oxide is
␣-Sb2O4. On the other hand, Sb6O13 is stable under the reaction
conditions and the selective oxidation proceeded only to a small
extent. Therefore, O-Dia could be a good support keeping anti-
mony oxide as Sb2O4, while SiO2 was not a favorable support
for keeping antimony oxide as Sb2O4.

H. Matsumura et al. / Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
129
Fig. 8. Influence of atmosphere on formaldehyde yield. Ar and CH4 flow was
switched to O2 and CH4 flow after the run for 60 min. Catalyst = 100 mg,
Sb2O4 (3 wt.%) on O-Dia, reaction temperature = 873 K, gas flow = 30 mL/min
Fig. 9. Diffuse reflectance UV–vis spectra of Sb2O4 loaded catalysts. (a) O-Dia,
(b) Sb2O4/O-Dia, (c) after the selective oxidation of methane on Sb2O4/O-Dia
in CH4/O2 = 25/5. Loading level 3 wt.% Samples was measure by diluting with
SiO2 (sample:SiO2 = 1:2).
(CH4/Ar or O2 = 25/5), SV = 18,000 mL h−1 g-cat−1
.
The average turnover frequency (TOF) to formaldehyde
during the reaction from 30 to 90 min was calculated as
13.5 mol formaldehyde h−1/mol Sb2O4 on O-Dia. Although the
number of active centers in the oxide-based catalyst cannot accu-
rately be evaluated, it can be expected that the TOF per active
center would be much higher.
reaction, an overlapping peak appeared at 541.0 eV (Sb2O3 Sb
3d3/2).
IR spectra of SbOx loaded on O-Dia are shown in Fig. 10. In
the IR absorption, broad bands at 780–690 cm−1 on ␣-Sb2O4/O-
3.7. Valence state of antimony oxide on oxidized diamond
supported catalyst
Fig. 8 shows the change in the formaldehyde yield by switch-
ing Ar atmosphere to O2 during the reaction over the SbOx
loadedO-Diacatalyst. UnderArflow, theformationofformalde-
hyde was observed only in the initial short period, and the yield
of formaldehyde rapidly decreased by the consumption of active
oxygen in the antimony oxide lattice on the O-Dia surface. When
the flowing gas was switched from Ar to O2, oxygen defect on
the antimony oxide on the O-Dia was rapidly recovered and the
yield of formaldehyde rapidly increased.
Fig. 9 shows the diffuse reflectance UV–vis spectra of the O-
Dia supported catalyst before and after the selective oxidation
of methane in O2. The profile of fresh catalyst contained various
antimony oxides that are assigned to Sb3+ (195–230 nm), Sb4+
and Sb4.33+ (230–280 nm and 480 nm) and Sb5+ (300–340 nm).
Although these values are not consistent with those reported
in the literature [47], XRD data of these samples support our
present UV–vis assignments. After the selective oxidation of
methane with O2 at 873 K, the absorption assigned to Sb3+
species increased. This indicates that slight loss of surface oxy-
gen from Sb2O4 proceeded during the oxidation.
Fig. 10. Diffuse reflectance IR spectra of Sb2O4/O-Dia before and after the
selective oxidation of methane. (a) O-Dia, (b) Sb6O13 bulk sample, (c) fresh
catalyst after calcinations at 723 K for 3 h, (d) after the selective oxidation of
methane at 873 K for 1.5 h.
X-ray photoelectron spectra revealed the same features as
observed with UV–vis spectra, where the fresh Sb2O4/O-Dia
catalyst exhibited a broad peak at 543.2 eV (Sb 3d3/2). After the

130
H. Matsumura et al. / Journal of Molecular Catalysis A: Chemical 250 (2006) 122–130
[12] C.-B. Wang, R.G. Herman, C. Shi, Q. Sun, J.E. Roberts, Appl. Catal.
A 247 (2003) 321–333.
[13] V. Fornes, C. Lopez, H.H. Lopez, A. Martinez, Appl. Catal. A 249
(2003) 345–354.
[14] B. Lin, X. Wang, Q. Guo, W. Yang, Q. Zhang, Y. Wang, Chem. Lett.
32 (2003) 860–861.
[15] H. Berndt, A. Martin, A. Bru¨ckner, E. Schreier, M. Mu¨ller, H. Kosslick,
G.-U. Wolf, B. Lu¨cke, J. Catal. 191 (2000) 384–400.
[16] Q. Zhang, W. Yang, X. Wang, Y. Wang, T. Shishido, K. Takehira, Micro-
por. Mesopor. Mater. 77 (2005) 223–234.
[17] Y. Teng, F. Ouyang, L. Dai, T. Karasuda, H. Sakurai, K. Tabata, E.
Suzuki, Chem. Lett. 28 (1999) 991–992.
[18] T. Takemoto, K. Tabata, Y. Teng, A. Nakayama, E. Suzuki, Appl. Catal.
A 205 (2001) 51–59.
[19] K. Tabata, Y. Teng, T. Takemoto, E. Suzuki, M.A. Banares, M.A. Pena,
J.L.G. Fierro, Catal. Rev. 44 (2002) 1–58.
[20] C. Batiot, B.K. Hodnett, Appl. Catal. A 137 (1996) 179–191.
[21] K. Nakagawa, C. Kajita, N. Ikenaga, T. Kobayashi, M.N. -Gamo, T.
Ando, T. Suzuki, Chem. Lett. 29 (2000) 1100–1101.
[22] H. Nishimoto, K. Nakagawa, N. Ikenaga, M.N. -Gamo, T. Ando, T.
Suzuki, Appl. Catal. A 264 (2004) 65–72.
[23] K. Nakagawa, K. Okumura, T. Shimamura, N. Ikenaga, T. Suzuki, T.
Kobayashi, M.N. -Gamo, T. Ando, Chem. Lett. 32 (2003) 866–867.
[24] K. Okumura, K. Nakagawa, T. Shimamura, N. Ikenaga, M.N. -Gamo,
T. Ando, T. Kobayashi, T. Suzuki, J. Phys. Chem. B 107 (2003)
13419–13424.
Scheme 1. Catalytic cycle of selective oxidation.
Dia catalyst were observed throughout the reaction, indicating
thatSb2O4 wasthepredominantphaseonO-Dia [45]. Duringthe
selective oxidation, peaks at 780, 690 and 480 cm−1 increased.
Slight increase in peaks around 1100 cm−1 was seen and this
seems to be ascribed to the Sb3+ species. Increase in the absorp-
tion at 480 cm−1 could not be interpreted at present.
These results seem to indicate that a certain portion of lattice
oxygen of ␣-Sb2O4 on the O-Dia was consumed in reacting
with methane leaving Sb2O4−x. The reduced Sb2O4−x on the
O-Dia could be re-oxidized to ␣-Sb2O4 by O2 in the gas phase.
Therefore, it is strongly suggested that the redox cycle existed
between ␣-Sb2O4 and Sb2O4−x (Scheme 1).
4. Conclusions
[25] T. Shimamura, K. Okumura, K. Nakagawa, T. Ando, N. Ikenaga, T.
Suzuki, J. Mol. Catal. A 211 (2004) 97–102.
[26] M.D. Allen, S. Poulston, E.G. Bithell, M.J. Goringe, M. Bowker, J.
Catal. 163 (1996) 204–214.
[27] G. Centi, P. Mazzoli, S. Perathoner, Appl. Catal. A 165 (1997) 273–290.
[28] M.O. Guerrero-Perez, M.A. Banares, Chem. Commn. (2002) 1292–1293.
[29] G. Centi, F. Guarnieri, S. Perathoner, J. Chem. Soc., Faraday Trans. 93
(1997) 3391–3402.
Antimonyoxidesupportedontheoxidizeddiamondshoweda
moderate catalytic activity for the selective oxidation of methane
to formaldehyde. It was suggested that the reaction proceeded by
a redox mechanism between Sb2O4 and Sb2O4−x, while stable
Sb6O13 was almost inactive.
[30] A. Wickman, L.R. Wallenberg, A. Andersson, J. Catal. 194 (2000)
153–166.
Acknowledgment
[31] N. Ballarini, F. Cavani, M. Cimini, T. Trifiro, R. Catani, U. Cornaro, D.
Ghisletti, Appl. Catal. A 251 (2003) 49–59.
[32] J. Nilsson, A.R. Landa-Canovas, S. Hansen, A. Andersson, J. Catal. 186
(1999) 442–457.
[33] L.C. Brazdil, A.M. Ebner, J.F. Brazdil, J. Catal. 163 (1996) 117–121.
[34] J. Nilsson, A.R. Landa-Canovas, S. Hansen, A. Andersson, J. Catal. 160
(1996) 244–260.
This work was partly supported “Hi-Tech Research Center”
Project for Private Universities: by matching fund subsidy from
Ministry of Education, Culture, Sports, Science and Technology
(2003-2007).
References
[35] H.W. Zanthoff, S. Schaefer, G.-U. Wolf, Appl. Catal. A 164 (1997)
105–117.
[36] G. Centi, F. Marchi, S. Perathoner, Appl. Catal. A 149 (1997) 225–244.
[37] Y.A. Saleh-Alhamed, R.R. Hudgins, P.L. Silveston, J. Catal. 161 (1996)
430–440.
[1] A. Parmaliana, F. Arena, J. Catal. 167 (1997) 57–65.
[2] F. Arena, N. Giordano, A. Parmaliana, J. Catal. 167 (1997) 66–76.
[3] H.-F. Lin, R.-S. Liu, K.Y. Liew, R.E. Johnson, J.H. Lunsford, J. Am.
Chem. Soc. 106 (1984) 4117–4121.
[38] F.J. Berry, Adv. Catal. 30 (1981) 97–131.
[39] E.V. Steen, M. Schnobel, R. Walsh, T. Riedel, Appl. Catal. A 165 (1997)
349–356.
[40] P. Botella, P. Concepcion, J.M.L. Nieto, B. Solsona, Catal. Lett. 89
(2003) 249–253.
[41] W. Ueda, K. Oshihara, Appl. Catal. A 200 (2000) 135–143.
[42] H. Zhang, P. Ying, J. Zhang, C. Liang, Z. Feng, C. Li, Stud. Surf. Sci.
Catal. 147 (2004) 547–552.
[4] N.D. Spencer, J. Catal. 109 (1988) 187–197.
[5] T. Suzuki, K. Wada, M. Shima, Y. Watanabe, J. Chem. Soc., Chem.
Commun. (1990) 1159–1160.
[6] L.-X. Dai, Y.-H. Teng, K. Tabata, E. Suzuki, T. Tatsumi, Chem. Lett.
29 (2000) 794–795.
[7] A. de Lucas, J.L. Valverde, L. Rodriguez, P. Sanchez, M.T. Garcia, Appl.
Catal. A 203 (2000) 81–90.
[43] T. Ando, K. Yamamoto, M. Ishii, M. Kamo, Y. Sato, J. Chem. Soc.
Faraday Trans. 89 (1993) 3635–3640.
[8] X. Zhang, D.-H. He, Q.-J. Zhang, Q. Ye, B.-Q. Xu, Q.-M. Zhu, Appl.
Catal. A 249 (2003) 107–117.
[44] C.P. Fenimore, F.J. Martin, Combust. Flame 10 (1966) 135–139.
[45] C.A. Cody, L. Dicarlo, R.K. Darlington, Inorg. Chem. 18 (1979)
1572–1576.
[46] K.A.L. Abdelouabad, U. Roullet, M. Brun, A. Burrows, C.J. Kiely, J.C.
Volta, M. Abon, Appl. Catal. A 210 (2001) 12–136.
[47] F. Sala, F. Trifro, J. Catal. 34 (1974) 68–78.
[9] Y. Wang, K. Otsuka, J. Chem. Soc., Chem. Commun. (1994) 2209–
2210.
[10] A. Parmaliana, F. Arena, F. Frusteri, A. Martinez-Arias, M.L. Granados,
J.L.G. Fierro, Appl. Catal. A 226 (2002) 163–174.
[11] M.M. Koranne, J.G. Goodwin Jr., G. Marcelin, J. Catal. 148 (1994)
378–387.