Highly effective synthesis of cyclohexanone oxime over a novel titanosilicate Ti-MWW

Article abstract of DOI:10.1246/cl.2005.1436

Liquid-phase ammoximation of cyclohexanone proceeds extremely efficiently, selectively, and stablely over a novel titanosilicate Ti-MWW in a clean solvent of water, which would lead to a potential process more environmentally benign for synthesizing cyclo

Full text of DOI:10.1246/cl.2005.1436

1436
Chemistry Letters Vol.34, No.10 (2005)
Highly Effective Synthesis of Cyclohexanone Oxime over a Novel Titanosilicate Ti-MWW
Fen Song, Yueming Liu, Haihong Wu, Mingyuan He, Peng Wu,ꢀ and Takashi Tatsumiy
Shanghai Key Laboratory of Green Chemistry and Chemical Processes, Department of Chemistry, East China Normal University,
North Zhongshan Rd. 3663, Shanghai 200062, P. R. China
yChemical Resources Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama 226-8503
(Received July 19, 2005; CL-050929)
Liquid-phase ammoximation of cyclohexanone proceeds
extremely efficiently, selectively, and stablely over a novel
titanosilicate Ti-MWW in a clean solvent of water, which would
lead to a potential process more environmentally benign for
synthesizing cyclohexanone oxime.
by Ti-MWW-HTS and Ti-MWW-PS, respectively. A rough es-
timation based on the prices of raw materials indicates that the
manufacture costs of these two Ti-MWW catalysts are approxi-
mately half and two third that of TS-1, respectively. Other
titanosilicates for control experiments, TS-1,11 Ti-MOR,5 and
Ti-ꢀ12 were prepared according to the procedures reported else-
where. All the catalysts have been characterized thoroughly by
various techniques (XRD, IR, SEM, UV, ICP, and N2-adsorp-
tion). The ammoximation runs were performed batchwise in a
25-mL flask equipped with a magnetic stirrer and a condenser.
In a typical run, 50 mg of catalyst, 10 mmol of cyclohexanone,
5 mL of solvent, and 12 mmol of NH3 aqueous solution (25%)
were charged into the flask and the mixture was heated to a de-
sirable temperature (303–353 K). The reaction was then started
by adding a diluted aqueous H2O2 (5%, 12 mmol) continuously
with a micro-pump for 1 h. After finishing the addition of H2O2,
the mixture was further stirred for 0.5 h. After the removal of
solid catalyst, the products were analyzed by a gas chromato-
graph (Shimadzu 14B, FID detector) equipped with a 30-m
DB-1 capillary column, using toluene as an internal standard.
The amounts of ammonia and H2O2 remaining in the mixture
were quantified by the acid–base titration method and iodomety,
respectively.
Cyclohexanone oxime is a key intermediate for the manu-
facture of "-caprolactam (CPL) with an annual world market
of 3.6 million tons in 1998.1 Over 90% of the world installed pro-
duction capacity of CPL is based on the non-catalytic oximation
of cyclohexanone with hydroxylamine derivatives, followed by
the liquid-phase Beckmann rearrangement in oleum. These con-
ventional commercial processes have been encountering serious
disadvantages, such as using poisonous hydroxylamine and cor-
rosive fuming sulfuric acid and producing a large quantity of by-
products. Particularly, the by-product of ammonium sulphate
counts high as 2.8 times as much as product CPL.2 A break-
through has been made in the industrial production of CPL in
2003 by Sumitomo, which started the operation of a new CPL
process by combining together the liquid-phase ammoximation
of cyclohexanone with aqueous ammonia and hydrogen perox-
ide over titanosilicate catalyst and the heterogeneous vapour-
phase Beckmann rearrangement over a highly siliceous ZSM-5
(silicalite-1). The process is extremely environmentally friendly
from the viewpoint of zero emission. The catalytic ammoxima-
tion technique in the first step of this novel process is initially de-
veloped by Enichem using the MFI type titanosilicate, TS-1.1
However, this process still has some drawbacks, such as a high
cost for TS-1 manufacture due to using expensive structure-di-
recting agent of tetrapropylammonium hydroxide and organic
silicon source, necessity of a high TS-1 amount relative to sub-
strate and a volatile solvent of tert-BuOH.3 To develop more
efficient titanosilicates, TS-2,4 Ti-MOR,5 and Ti-ꢀ6 have been
developed. However, none of them shows potential capability
of substituting for TS-1 in the liquid-phase ammoximation.
Ti-MWW with the MWW structure (generally known as
MCM-22 for its alluminosilicate analogue), a novel titanosilicate
recently developed by us, shows superior activity and product
selectivity to conventional titanosilicates in the epoxidation
of alkenes with H2O2.7 The unique catalytic properties of
Ti-MWW has been closely related to its pore system consisting
of 12-membered ring (MR) side cups and two independent inter-
layer and intralayer 10-MR channels.8
Table 1 shows the results of ammoximation on various cat-
alysts. The product was predominately cyclohexanone oxime
when the reaction proceeded to high levels, while peroxydicy-
clohexyl amine, and the products due to the aldol condensation
of cyclohexanone were also co-produced particularly at a lower
conversion of cyclohexanone. The solvent effect investigated
on Ti-MWW-PS showed a catalytic activity order of H2O >
MeOH > t-BuOH > MeCN (Table 1, Nos. 1–4). Both the con-
Table 1. The results of cyclohexanone ammoximationa
No.
Cat.
Si/Ti
Solvent
Conv.b/% Oxime sel.c/%
1
2
3
4
5
6
7
8
9
Ti-MWW-PS
Ti-MWW-PS
Ti-MWW-PS
Ti-MWW-PS
55 H2O
99.4
45.7
75.4
74.1
97.0
97.0
16.2
60.0
15.0
99.0
99.9
82.4
99.9
99.9
99.9
99.9
72.8
95.0
4.0
55 MeCN
55 MeOH
55 t-BuOH
Ti-MWW-HTSd 50 H2O
TS-1e
TS-1e
Ti-MOR
Ti-ꢀ
51 H2O-t-BuOH
51 H2O
90 H2O
76 H2O
60 H2O
10 Ti-MWW-PSf
99.9
We report here that Ti-MWW is a more effective catalyst for
the liquid-phase ammoximation of cyclohexanone in the pres-
ence of water, which may bring out a much greener process
for the oxime synthesis than TS-1.
Ti-MWW catalysts were synthesized by hydrothermal
synthesis9 and postsynthesis methods,10 which were denoted
aReaction conditions: cat., 50 mg; cyclohexanone, 10 mmol; solvent, 5 mL;
NH3 (25%), 12 mmol; H2O2 (5%), 12 mmol; temp., 338 K; time, 1.5 h.
H2O2 was added dropwise at a constant rate within 1 h. bThe conversion
of cyclohexanone. cBy-products were mainly peroxydicyclohexyl amine
etc. dCat, 100 mg. eCat., 200 mg; reaction time, 5 h. The reaction conditions
were similar to those reported previously.13 fNo. 1 was reused for 5 times.
Copyright Ó 2005 The Chemical Society of Japan

Chemistry Letters Vol.34, No.10 (2005)
1437
peated reaction, and XRD patterns and UV spectra showed that
the used Ti-MWW had the same crystallinity and coordination
state of Ti as the fresh one. Thus, Ti-MWW possesses excellent
stability and reusability in the ammoximation of cyclohexanone.
The activity of Ti-MWW greatly depends on the adding
mode of reactants. In particular, adding H2O2 into the reaction
system rapidly decreased the oxime yield greatly. For example,
when a desirable amount of H2O2 was added all at once, the re-
action resulted in only ca. 3% conversion of cyclohexanone,
while the conversion of H2O2 and NH3 reached 99 and 81%, re-
spectively. This dramatic behavior is presumed to be related to
the reaction mechanism of ammoximation. The titanosilicate-
catalyzed ammoximation plausibly involves the formation of
an intermediate of hydroxylamine formed through the catalytic
oxidation of NH3 with H2O2 on Ti active sites and a subsequent
non-catalytic oximation of cyclohexanone with NH2OH to
oxime.5,14 The latter oximation should compete with the consec-
utive oxidation of NH2OH by H2O2. Ti-MWW with superior
oxidation capability would accelerate the oxidation of NH2OH
before reacting with cyclohexanone to oxime particularly when
an excess amount of H2O2 exists. Thus, slow addition of H2O2
would reduce its concentration in the reaction system, and hence
avoid a deep oxidation of NH2OH, which is then in favor of
oxime formation.
100
Ti-MWW
80
60
40
TS-1
20
0
0
10
20
Ti/(Ti+Si) molar ratio (*1000)
Figure 1. Dependence of cyclohexanone conversion on Ti
content. Reaction conditions: cat., 50 mg; 338 K; time, 1.5 h;
others, see Table 1.
version of cyclohexanone and the oxime selectivity reached as
high as 99% in water solvent. Simultaneously, the utilization
of both NH3 and H2O2 was reasonably high (>95%) as the reac-
tion was performed under the conditions of almost stoichiomet-
ric molar ratios for the substrates. Despite being somewhat low
in the conversion of cyclohexanone, a high performance was
also achieved on Ti-MWW-HTS prepared by a hydrothermal
synthesis (No. 5).
As a result we have found that Ti-MWW deserves to be a
promising catalyst for the greener synthesis of cyclohexanone
oxime actively, selectively, and regenerably.
When the ammoximation was carried out on TS-1 under the
same conditions as reported,1,13 TS-1 catalyzed the ammoxima-
tion very actively and selectively in a cosolvent of water and
t-BuOH (No. 6). However, in a single solvent of water the reac-
tion retarded greatly (No. 7). It should be noted that not only in
the solvent, the optimized conditions of TS-1 also differed in the
amount catalyst used and the reaction time. To give a conversion
of 97%, TS-1 required a catalyst amount of 20 wt % relative to
cyclohexanone and a longer reaction time of 5 h. Other titanosi-
licates, Ti-MOR and Ti-ꢀ were much less active (Nos. 8 and 9).
Thus, considering the fact that Ti-MWW is capable of giving a
conversion >99% at 5 wt % of catalyst relative to cyclohexa-
none, we claim that we have estabilished a new ammoximation
catalyst more effective than TS-1.
To further confirm this issue, a series of Ti-MWW and TS-1
both with various Si/Ti ratios have been synthesized, and they
have been applied to the ammoximation. In order to fairly eval-
uate the catalytic activity of these two titanosilicates, the reac-
tions have been carried out under optimum conditions for each
catalyst but with the same catalyst amount relative to cyclohex-
anone substrate (Figure 1). The oxime selectively was compara-
bly high for Ti-MWW and TS-1 (>98%). The conversion of
cyclohexanone increased with increasing amount of Ti for both
Ti-MWW and TS-1. Obviously, the former is more effective
from the viewpoint of catalytic activity.
We thank the financial supports by NSFC (Nos. 20473027
and 20233030) and STCSM (03DJ14005). P.W. thanks Program
for New Century Excellent Talents in University (NCET). T.T.
thanks the support by Core Research for Evolutional Science
and Technology (CREST) of JST Corporation.
References
1
2
H. Ichihashi and H. Sato, Appl. Catal., A, 221, 359 (2001).
G. Bellussi and M. S. Rigguto, Stud. Surf. Sci. Catal., 137, 911
(2001).
3
4
P. Roffia, M. Padovan, and G. Leofanti, U.S. Patent, 4794198
(1988).
J. S. Reddy, S. Sivasanker, and P. Ratnasamy, J. Mol. Catal., 69,
383 (1991).
5
6
P. Wu, T. Komatsu, and T. Yashima, J. Catal., 168, 400 (1997).
J. Le Bars, J. Dakka, and R. A. Sheldon, Appl. Catal., A, 136, 69
(1996).
7
P. Wu, T. Tatsumi, T. Komatsu, and T. Yashima, Chem. Lett., 2000,
774; P. Wu, T. Tatsumi, T. Komatsu, and T. Yashima, J. Catal.,
202, 245 (2001); P. Wu and T. Tatsumi, Chem. Commun., 2001,
897.
8
9
M. E. Leonowicz, J. A. Lawton, and S. L. Lawton, Science, 264,
1910 (1994); S. L. Lawton, M. E. Leonowicz, and P. D. Partridge,
Microporous Mesoporous Mater., 23, 109 (1998).
P. Wu, T. Tatsumi, T. Komatsu, and T. Yashima, J. Phys. Chem. B,
105, 2897 (2001).
10 P. Wu and T. Tatsumi, Chem. Commun., 2002, 1026; W. Fan,
P. Wu, S. Namba, and T. Tatsumi, Angew. Chem., Int. Ed., 43,
The stability and reusability of Ti-MWW in the ammoxima-
tion have been investigated. The used Ti-MWW catalyst was re-
generated by washing with acetone and drying at 393 K, and was
subsequently subjected to repeated ammoximation at a constant
ratio of catalyst-substrates-solvent. After five recycles of reac-
tion-regeneration, the Ti-MWW catalyst maintained the activity
and oxime selectivity (Table 1, No. 10). The quantification of Ti
by ICP indicated that the Ti leaching was within 5% after the re-
236 (2004).
11 T. Taramasso, G. Perego, and B. Notari, U.S. Patent, 441050
(1983).
12 M. A. Camblor, A. Corma, and J. Perez-Pariente, Zeolites, 13, 82
(1993).
13 A. Thangaraj, S. Sivasanker, and P. Ratnasamy, J. Catal., 131, 394
(1991).
14 L. Dal Pozzo, G. Fornasari, and T. Monti, Catal. Commun., 3, 369
(2002).
Published on the web (Advance View) September 24, 2005; DOI 10.1246/cl.2005.1436