Selective ammoxidation of isobutane on a crystalline SbRe2O6 catalyst

Article abstract of DOI:10.1039/b005868h

The catalytic ammoxidation of isobutane to methacrylonitrile at 673 K proceeds on a new class of Re mixed-oxide SbRe₂O₆ with alternate (Re₂O₆)^3- and (SbO)⁺ layers, where ammonia is prerequisite for the C-H bond scission of isobutane.

Full text of DOI:10.1039/b005868h

Selective ammoxidation of isobutane on a crystalline SbRe O catalyst
2
6
Haichao Liu, Takafumi Shido and Yasuhiro Iwasawa*
Department of Chemistry, Graduate School of Science, The University of Tokyo, Hongo-7-3-1, Bunkyo-ku, Tokyo
113-0033, Japan. E-mail: Iwasawa@chem.s.u-tokyo.ac.jp
Received (in Cambridge, UK) 20th July 2000, Accepted 31st August 2000
First published as an Advance Article on the web 18th September 2000
The catalytic ammoxidation of isobutane to methacrylo-
nitrile at 673 K proceeds on a new class of Re mixed-oxide
columns. The conversion of NH
under the present reaction conditions.
Table 1 presents the conversions, reaction rates and selectiv-
ities of the i-C 10 ammoxidation on the various Re–Sb–O
catalysts and bulk ReO and SbO at 673 K. SbOReO ·2H O,
Sb Re 13 and Sb oxides such as Sb and Sb showed no
activity. Bulk Re was active solely for i-C 10 combustion
to CO . Bulk ReO and ReO produced MAN, but the
selectivities were lower than 10% and the main product was
CO . Re /Sb or mechanically mixed Re ·Sb sam-
ples showed almost no activity for the i-C 10 ammoxidation
either. Copr.SbRe produced i-C (20.6% selectivity), but
no formation of MAN was observed. Isobutane combustion was
dominant also with this catalyst. Only the SbRe among these
samples was active for the ammoxidation of i-C 10 to MAN.
The selectivities to MAN and to the sum of MAN + i-C at
the steady-state conversion of 4.4% were 44.9 and 84.3%,
3 2 x
to N and NO was < 10%
32
+
SbRe
2
O
6
with alternate (Re
2
O
6
)
and (SbO) layers, where
ammonia is prerequisite for the C–H bond scission of
isobutane.
4
H
x
x
4
2
4
2
O
2
O
3
2 4
O
4
H
Much attention has been devoted to selective ammoxidation of
light alkanes from both fundamental and industrial interests. To
date, a large number of multicomponent metal-oxide catalysts
containing V, Mo, etc. have been explored to develop efficient
catalytic systems for selective ammoxidation. However, owing
to the inertness of light alkanes, very few catalysts have showed
good performances that are comparable to those for the
corresponding alkenes.1 There is thus a clear need to develop
new catalytic materials for the selective ammoxidation of light
alkanes. Except for the oxidation of methanol and ethanol,
compared with V, Mo and W there are limited uses of Re as a
key element in selective ammoxidation/oxidation, in spite of Re
having similar redox properties to those of V, Mo and W
oxides.5 Here, we report a first Re mixed-oxide catalyst
2
O
7
2
3
2
2
2
O
7
2
O
3
2
O
7
2 3
O
4
H
2
O
x
4 8
H
–4
2 6
O
4
H
4 8
H
2
1
respectively. A decrease in the GHSV from 5000 to 2500 h
increased the conversion to 7.5% while keeping good selectiv-
–7
ities to MAN and to MAN + i-C
Table 1.
4 8
H , as shown in parentheses in
(
crystalline SbRe
isobutane (i-C
where NH not only stabilizes the catalyst but also promotes the
C–H bond scission of i-C
SbOReO ·2H O, SbRe
in the similar way to that reported previously.
surface areas of the three samples were approximately 1 m g
For comparison, Sb -supported Re catalyst (Re
Sb ; 10 wt% Re) was prepared by an impregnation method
using an aqueous solution of NH
SbRe catalyst (copr.SbRe ) was also prepared by a
coprecipitation method using an ethanol solution of ReCl and
SbCl , followed by washing with water to eliminate the residual
2
H
O
6
) active for the selective ammoxidation of
4
10) to methacrylonitrile (MAN) at 673 K,
The active SbRe
alternate octahedral (Re
thin plate-like crystals preferably exposing the (100) plane.
The square basal (100) faces of SbRe crystals remained
2 6
O compound consists of connected,
32
+
3
2
O
6
)
and (SbO) layers and grows as
1
2
4 10
H .
4
2
2
O
6
and Sb
4
Re
2
O
13 were synthesized
2 6
O
6–12
The specific
unchanged before and after the ammoxidation at 673 K as
imaged by scanning electron microscopy. Further, neither
change nor modification of the surface composition and
crystallinity were observed by means of X-ray diffraction, X-
ray photoelectron spectroscopy (XPS) and in-situ micro
confocal laser Raman spectroscopy. The results indicate that the
2
21
.
2
O
3
2
O
7
2 7
O /
2 3
O
9
4
4
ReO . A coprecipitated
2
O
x
2 x
O
3
crystalline SbRe
2
O
6
works as a promising catalyst for the
·2H O and Sb Re
possess Re7 species, but the Re species were reduced to two
3
ammoxidation. The fresh SbOReO
4
2
4
2 13
O
+
Cl ions from the sample. Ammoxidation reactions were carried
out in a continuous-flow, fixed-bed reactor at 673 K under the
6+
4+
valent states (possibly Re and Re ) which exhibit XPS
binding energies of 42.3 and 44.7 eV, and 45.1 and 47.5 eV for
Re 4f7/2 and Re 4f5/2, respectively (compared to C 1s = 284.6
eV) under the ammoxidation conditions. These binding ener-
conditions of 10% i-C
with He and a gas-hourly-space-velocity (GHSV) of 5000 h
at atmospheric pressure. Prior to each run, the catalysts
typically 0.3 g) were pretreated at 673 K under He for 1 h. The
4 3 2
H10, 15% NH and 25% O balanced
2
1
(
2 6
gies are similar to those for the SbRe O surface after
reactants and products were analyzed by two on-line gas
chromatographs with Unibeads C, Gaskuropack 54 and VZ-10
ammoxidation at 673 K. Thus, the difference in the catalytic
performances of the crystalline Re–Sb–O catalysts may not be
x x
Table 1 Isobutane ammoxidation on different Re–Sb–O catalysts and bulk ReO and SbO at 673 K
Selectivity (%)
mmol g-cat2 h21 MAN + i-C
Conversion
Reaction rate/
1
(%)
H
4 8
MAN
i-C H
4 8
3
CH CN
CO
2
SbRe
SbOReO
Sb Re
mix.Re
2
O
6
4.4 (7.5)a
0
0
0.5
0.4
0.1
11.6
2.1
5.6
0
785.6 (669.5)a
0
0
89.3
71.4
17.8
2071.3
374.9
999.9
0
84.3 (82.6)a
—
—
trace
20.6
0
44.9 (44.2)a
—
—
0
0
0
39.4 (38.4)a
—
—
trace
20.6
0
4.7 (5.6)a
—
—
0
0
0
0
25.7
32.1
—
—
10.2 (11.4)a
—
—
~ 100
79.4
100
100
42.6
36.4
—
4
·2H
2
O
4
2
O
13
7
2
O
·Sb
2
O
3
copr.SbRe
O
2 x
Re
Re
2
O
O
7
/Sb
7
O
2 3
2
0
0
0
ReO
ReO
3
31.5
31.0
—
7.7
9.1
—
—
23.8
21.9
—
2
Sb
Sb
2
O
O
3
2
4
0
0
—
—
—
a
The data listed in parentheses were obtained at a GHSV of 2500 h21
.
DOI: 10.1039/b005868h
Chem. Commun., 2000, 1881–1882
This journal is © The Royal Society of Chemistry 2000
1881

plays a crucial role in enhancing and/or generating the activity
of SbRe for the dehydrogenation (C–H bond breaking) of i-
10 to i-C . Upon pulsing NH on SbRe , N and H
were produced indicating the reaction of NH with the lattice
oxygen atoms of SbRe to form oxygen vacancies that may
also be responsible for the C–H bond breaking. To our
knowledge, SbRe is the first case where NH changes an
inactive catalyst to an active one for light alkane activation,
2 6
O
C
4
H
4
H
8
3
2
O
6
2
2
O
3
2 6
O
2
O
6
3
1
4,15
3
although promoting effects of NH have been documented.
Further study is necessary for depicting a detailed mechanism of
3 2 6
the promoting effect of NH on the SbRe O selective
catalysis.
In conclusion, a new class of Re mixed-oxide SbRe
catalyzes the selective ammoxidation of i-C 10 to MAN. The
activity of SbRe may be relevant partly to its specific crystal
2 6
O
4
H
2 6
O
structure. The presence of ammonia is considered to be
prerequisite, not only for maintaining the stable crystal structure
Fig. 1 Yields of MAN (Î, 8) and i-C
of pulses of i-C 10 alone (8, 2) and an i-C
NH -preadsorbed SbRe catalyst at 673 K.
4
H
8
(5, 2) as a function of the number
of SbRe
bond breaking in i-C
2
O
6
, but also for promoting SbRe
2 6
O activity for C–H
4
H
4
H
10–O mixture (Î, 5) on an
2
4
H10 under the ammoxidation conditions.
3
2 6
O
This work has been supported by Core Research for
Evolutional Science and Technology (CREST) of the Japan
Science and Technology Corporation (JST). The authors
acknowledge Professor H. Imoto, Utsunomiya University, for
his helpful discussion on the preparation and structure of
due to the difference in their surface Re oxidation states, but to
the difference in their surface structures. This is entirely
different from the finding in the selective oxidation of i-C
773 K) and i-C (673 K) to methacrolein (MAL) that the
activities of the three crystalline Re–Sb–O compounds are
ascribed to a cooperation between Re and Sb Re 13, both
being formed by decomposition of the compounds under the
4 10
H
(
4
H
8
2 6
SbRe O .
2
O
7
4
2
O
Notes and references
1
2
R. K. Grasselli, Catal. Today, 1999, 49, 141.
S. Albonetti, F. Cavani and F. Trifiro, Catal. Rev.-Sci. Eng., 1996, 38,
oxidation conditions.8
observed in the ammoxidation of i-C
Sb–O compounds are more or less active at 673 K.
It is to be noted that SbRe was inactive for i-C
selective oxidation as well as for the total oxidation at 673 K in
the absence of NH , whereas it exhibited a good performance
for i-C 10 ammoxidation at 673 K (Table 1). These results
may indicate a promoting effect of NH on the C–H activation
in i-C 10. To examine the role of NH , a series of pulse
experiments were conducted on SbRe at 673 K in Fig. 1. No
products were produced by pulsing i-C 10 alone or an
i-C mixture on to the SbRe catalyst, which indicates
that no C–H bond breaking in i-C 10 molecules occurs on the
catalyst. However, when the catalyst was pretreated with an
NH pulse (the catalyst surface was saturated with NH ),
i-C 10 was converted to MAN and i-C . The promotion
effect of NH pretreatment on the formation of MAN and i-
was more remarkable with the i-C pulse reaction
as shown in Fig. 1. The formation of MAN and i-C
decreased with the number of the i-C pulses. The pulse
experiments show that adsorbed NH species are incorporated
to the ammoxidation of i-C 10 to form MAN. These results
demonstrate that NH not only behaves as a reactant but also
–10
It is also different from the feature
where the three Re–
4 8
H
4
13.
13
3
C. J. Pereira, Science, 1999, 285, 670.
2
O
6
4
H
10
4 T. Inoue, S. T. Oyama, H. Imoto, K. Asakura and Y. Iwasawa, Appl.
Catal., A, 2000, 191, 131.
5 J.-M. Jehng, H. Hu, X. Gao and I. E. Wachs, Catal. Today, 1996, 28,
3
3
35.
4
H
6
7
8
W. T. A. Harrison, A. V. P. Mcmanus, A. P. Kaminsky and A. K.
Cheetham, Chem. Mater., 1993, 5, 1631.
Y. Yuan, H. Liu, H. Imoto, T. Shido and Y. Iwasawa, Chem. Lett., 2000,
3
4
H
3
2 6
O
6
74.
4
H
H. Liu, E. M. Gaigneaux, H. Imoto, T. Shido and Y. Iwasawa, J. Phys.
Chem. B, 2000, 104, 2033.
4
H
10–O
2
2 6
O
4
H
9 E. M. Gaigneaux, H. Liu, H. Imoto, T. Shido and Y. Iwasawa, Top.
Catal., 2000, 11–12, 185.
10 H. Liu, E. M. Gaigneaux, H. Imoto, T. Shido and Y. Iwasawa, Appl.
Catal. A: General., 2000, 202, 251.
3
H
x
4
4 8
H
1
1
1 H. Watanabe and H. Imoto, Inorg. Chem., 1997, 36, 4610.
2 H. Watanabe, H. Imoto and H. Tanaka, J. Solid State Chem., 1998, 138,
3
C
4
H
8
4
H10–O
2
2
45.
4 8
H
1
1
3 H. Liu, H. Imoto, T. Shido and Y. Iwasawa, submitted for publica-
tion.
4 V. D. Sokolovskii, A. A. Davydov and O. Yu. Ovsitser, Catal. Rev.-Sci.
Eng., 1995, 37, 425.
4
H
10–O
2
x
4
H
3
15 G. Centi and S. Perathoner, Catal. Rev.-Sci. Eng., 1998, 40, 175.
1882
Chem. Commun., 2000, 1881–1882