Eco-friendly synthesis of m-tolunitrile by heterogeneously catalysed liquid phase ammoxidation

Article abstract of DOI:10.1246/cl.2005.646

The ammoxidation of m-xylene to m-tolunitrile over silica-supported Co-Mn-Ni catalyst was conducted for the first time in liquid phase without solvent in one-step procedure. Copyright

Full text of DOI:10.1246/cl.2005.646

646
Chemistry Letters Vol.34, No.5 (2005)
Eco-friendly Synthesis of m-Tolunitrile by Heterogeneously Catalysed Liquid
Phase Ammoxidation
Min Zhong, Yong-guang Liang, Yuan Liu, and Yu-long Maꢀ
Department of Chemistry, Wuhan University, Wuhan 430072, P. R. China
(Received January 20, 2005; CL-050088)
The ammoxidation of m-xylene to m-tolunitrile over silica-
supported Co–Mn–Ni catalyst was conducted for the first time in
liquid phase without solvent in one-step procedure.
Products were analyzed by a Shanghai 102G type gas chroma-
tography. Authenticated standard samples were used to deter-
mine the identity of the products. The total conversion and prod-
uct distribution were evaluated with calibration curves, which
were obtained by injecting known amount of authenticated
standard.
Aromatic nitriles are valuable intermediates and reactants in
the fine chemical sector used for the synthesis of several pharma-
ceuticals, dyestuffs and pesticides,1,2 However, their synthesis
can be problematic, hazardous or environmentally damaging.
For example: the yield of m-tolunitrile are sometimes high with
traditional synthetic methods but a number of hazardous re-
agents, expensive feed components and solvents are used. An-
other process for the preparation of m-tolunitrile in a single step
reaction is the vapour phase ammoxidation, which is ecological-
ly and economically profitable route,1,2 but the selectivity for m-
tolunitrile is very low.3–6 Therefore, in terms of simplicity, safe-
ty, and waste-minimization, the development of new and practi-
cal methodologies on the production of m-tolunitrile is welcome.
Vanadium and molybdenum oxides are well known as cata-
lysts for the oxidation of propane to acrylic acid.7 They can also
be successfully employed as highly active catalysts in the heter-
ogeneously catalysed ammoxidation of propane,8 and other sub-
stituted methyl aromatics.9,10 But it is diffcult for only one meth-
yl to be converted to cyano group, because two methyl groups in
m-xylene are identical. By comparison, cobalt oxides posses an
extremely high oxidative reactivity and manganese oxides has a
lower, although still high activation ability in the liquid phase
oxidation.11,12 They have similar redox properties to those of
Vanadium, molybdenum oxides. Then we chose silica-supported
Co–Mn–Ni oxides catalysts to improve the selectivity for
m-tolunitrile. To the best of our knowledge, there are no reports
on the heterogeneously catalyzed liquid phase ammoxidation of
m-xylene to m-tolunitrile.
The performance of catalysts for m-xylene ammoxidation
was studied under the same reaction conditions. For silica-sup-
ported Co–Mn–Ni catalyst, the selectivity is 99.3%, and the
yield is 16.3%. Two blank reactions were studied so as to dismiss
questions of support-catalyzed ammoxidation of m-xylene to
m-tolunitrile. In one blank reaction, no catalyst was used. In
the other blank reaction, only silica was used. Subsequent GC
analysis of the solutions indicated that no products were
presented. Catalyst without support exhibited lower activity
and selectivity than supported catalysts.
The powder X-ray diffraction (XRD) patterns of the samples
were recorded by a Shimadzu XRD-6000 diffractometer in a
scanning rate of 2ꢁ/min from 10 to 80ꢁ using Cu Kꢀ1 radiation
ꢀ
(ꢁ ¼ 1:54056 A). Figure 1 shows the XRD patterns of different
catalysts. For sample a, four sharp peaks, centered at about 29.2,
32.8, 36.3, and 60.7ꢁ are attributed to spinel-type composite
oxide (Co, Mn)(Co, Mn)2O4, which is corresponding with the
JCPDS card (number 18-408). For sample b and c, a broad peak
assigned to amorphous silica is observed at about 22.2ꢁ, which
indicates that the oxides are well dispersed on the surface of sili-
ca support within and without in the amorphous or microcrystal
(<4 nm) and small amount of spinel composite oxide emerged
on the surface of silica support. Furthermore, the intensity of
the peak at 36.3, 44.1, 57.2, and 64.5ꢁ in (c) increased compared
with (b), which is attributed to the better crystallized oxide after
the first run and enriched on the surface of silica. Hence, the
sinter was occured on the surface of catalyst.
The catalyst was prepared using a commercial silica (grain
diameter 125–425 mm) as support. The support was impregnated
with an aqueous solution of cobalt, manganese acetates and nick-
el chloride (n(Co):n(Mn):n(Ni) = 1:1.397:0.016) and the mix-
ture was dried at 393 K. The as-prepared precursor was calcined
for 3 h at 573 K and kept for 5 h at 773 K in a muffle to afford the
oxides. Weight per cent of Co + Mn + Ni are 9.32% on SiO2.
The catalytic ammoxidation reactions were carried out in a
300-mL capacity autoclave equipped with a magnetic-type stir-
rer (Dalian Instruments, china). m-xylene (180 g) and the cata-
lyst (0.5 g) were added to the rector; the mixture was stirred at
a constant speed of 500 rpm; it was heated to 363 K; and then
oxygen and ammonia with the ratio of 5:1 was injected into
the reactor and maintaining the reaction-system pressure of
0.5 MPa; finally the O2 concentration of tail gas was measured
by an automatic oxygen-measuring instrument on line. After
the reaction was finished, the catalyst was filtered and the resid-
ual mixture was distilled under the condition of decompression.
Figure 2 exhibits scanning electron micrographs of the fresh
catalyst and its sample after first run. As seen from the morphol-
a
b
c
10
20
30
40
50
60
70
80
2θ / degrees
Figure 1. XRD patterns of the samples: (a). catalyst without
support; (b). fresh supported catalyst; (c). first run.
Copyright ꢀ 2005 The Chemical Society of Japan

Chemistry Letters Vol.34, No.5 (2005)
647
100
80
20
16
12
8
60
40
Selectivity
Yield
4
20
0
0
1
2
3
4
Run number
Figure 4. Reusage of CoMnNi/SiO2 catalyst.
enriched on the surface, in good agreement with the micrographs
of the used catalyst.
Finally, the catalyst was recycled for four times in order to
check the activity and stability. After reaction of each cycle, the
catalyst was filtered off without any processing. Figure 4 lists the
results of these experiments. However, after successive
recycling (cycles 2 to 4) of the catalyst, a minor decrease in
the catalytic activity was noticed, and by-product detected in
the reaction was m-tolualdehyde, but its content was small. when
m-tolualdehyde was contacted with ammonia on surface of
catalyst, m-tolunitrile was formed.13 No more runs were carried
out because of the decrease recovered each time. The silica
was splitted to pieces under solution-reaction conditions with
high-speed stirring. Future studies will concentrate on improving
the mechanical strength of catalyst by changing preparation
method toward large-scale applications.
Figure 2. SEM image of fresh catalyst (a) and used (b) catalyst
after first run.
0
2
4
6
8
10
12
14
In summary, the preparation of m-tolunitrile from m-xylene
was carried out directly by a heterogeneously catalytic method
over CoMnNi/SiO2 catalyst for the first time. Compared to orga-
nometallic catalysts11 and Cerium complexes with acetato-acyl-
bis(pyrazolinone)ligands,14 CoMnNi/SiO2 catalyst is less de-
manding, economical in preparation, and suitable for large-scale
production. Moreover, the route is environmentally friendly, so
it has a potential industrial application. Further, the catalyst
can be recycled and reused for four times. Now, a detailed inves-
tigation is under progress to prepare the optimized catalysts and
to increase the yield of m-tolunitrile.
Weight% Atomic% Rel.
Error
0.98
3.74
3.67
Si
83.42
7.35
9.23
91.09
Mn
Co
4.10
4.80
0
2
4
6
8
10
12
14
Weight% Atomic% Rel.
Error
1.22
2.10
2.73
This work was supported by National Natural Science
Foundation of China (project 20376064).
Si
62.36
21.48
16.15
76.95
Mn
Co
13.55
9.50
References
Figure 3. Superficial distribution of (a) fresh and (b) used
catalysts after first run.
1
2
3
4
A. Martin and B. Lucke, Catal. Today, 57, 61 (2000).
¨
A. Martin, V. N. Kalevaru, and B. Lucke, Catal. Today, 78, 311 (2003).
¨
H. Jie and Y. Ai-zheng, Shiyou Xuebao, 18, 42 (2002).
G.-y. Xie, Q. Zheng, C. Huang, and Y.-y. Chen, Synth. Lett., 33, 1104
(2003).
H. P. Angstadt, U. S. Patent 4110250 (1978).
S. Masao, T. Kengo, T. Noriko, and Y. Onda, Eur. Pat. Appl. EP 339,553
(1989); Chem. Abstr, 112, 157884v (1990).
ogy, particles were well distributed on porous silica in both (a)
and (b). The presence of loosely stacked grains (average size
250 nm) with a porous state is notable for a sample to be em-
ployed as a catalyst to provide the aisle of reactants and selectiv-
ity, which average size is less than 80 nm estimated in (a) from
SEM data. The particles in (b) demonstrated crystal phase well
and particle size grew larger after ammoxidation, which is coin-
cided with the XRD results.
5
6
7
8
9
K. Tomokazu, V. Damien, and U. Wataru, Chem. Lett., 32, 1028 (2003).
M. D. Lee and H. B. Chiang, Appl. Catal., A, 154, 55 (1997).
A. Martin, B. Lucke, H. Seeboth, and G. Ladwig, Appl. Catal., 49, 205 (1989).
¨
10 R. G. Rizayev, E. A. Mamedov, V. P. Vislovskii, and V. E. Sheinin, Appl.
Catal., A, 83, 103 (1992).
11 V. D. Sokolovskii, Catal. Rev.—Sci. Eng., 32, 1 (1990).
12 F. Wang, G.-y. Yang, W. Zhang, W.-h. Wu, and J. Xu, Chem. Commun., 2003,
1172.
13 M. Niwa, S. Inagaki, and Y. Murakami, J. Phys. Chem., 89, 3869 (1985).
14 Y. Tadatsugu, J. Mol. Catal. A: Chem., 187, 143 (2002).
Moreover, as superficial distribution of fresh (a) and used
(b) catalyst are shown in Figure 3, it is seen that the used catalyst
have higher superficial content of Mn, Co than the fresh one,
which indicated that the active components of catalyst are
Published on the web (Advance View) May 5, 2005; DOI 10.1246/cl.2005.646