Solvent effect on epoxidation of allyl chloride with hydrogen peroxide on titanium-containing silicalite

Article abstract of DOI:10.1134/S1070427208110189

The solvent effect on liquid-phase epoxidation of allyl chloride with an aqueous solution of hydrogen peroxide on TS-1 titanium-containing silicalite was examined. 1-Butanol, 2-butanol, 1-propanol, isopropanol, methanol, ethanol, water, acetone, methyl ethyl ketone, and 1-pentanol were tested as solvents.

Full text of DOI:10.1134/S1070427208110189

ISSN 1070-4272, Russian Journal of Applied Chemistry, 2008, Vol. 81, No. 11, pp. 1963–1966. © Pleiades Publishing, Ltd., 2008.
Original Russian Text © S.M. Danov, A.V. Sulimov, A.V. Sulimova , 2008, published in Zhurnal Prikladnoi Khimii, 2008, Vol. 81, No. 11, pp. 1847–1850.
ORGANIC SYNTHESIS
AND INDUSTRIAL ORGANIC CHEMISTRY
Solvent Effect on Epoxidation of Allyl Chloride with Hydrogen
Peroxide on Titanium-Containing Silicalite
S. M. Danov, A. V. Sulimov, and A. V. Sulimova
Dzerzhinsk Polytechnic Institute, Branch of Nizhni Novgorod State University, Dzerzhinsk, Nizhni Novgorod oblast, Russia
Received May 27, 2008
Abstract—The solvent effect on liquid-phase epoxidation of allyl chloride with an aqueous solution of hydro-
gen peroxide on TS-1 titanium-containing silicalite was examined. 1-Butanol, 2-butanol, 1-propanol, isopropa-
nol, methanol, ethanol, water, acetone, methyl ethyl ketone, and 1-pentanol were tested as solvents.
DOI: 10.1134/S1070427208110189
Epichlorohydrin is an important product of basic
organic synthesis. It is highly reactive owing to the
presence of an epoxy group and a labile chlorine atom.
Epichlorohydrin readily reacts with compounds of
various classes, which allows preparation of many
valuable products from it (epoxy resins, lacquers,
paints, adhesives, synthetic fibers, ion-exchange resins,
rubbers characterized by high oil and heat resistance
and impermeability to gases). However, despite diver-
sity of products prepared using epichlorohydrin, the
major area of its application is production of epoxy
resins (68% of the whole amount of epichlorohydrin
produced) [1].
an aqueous solution of hydrogen peroxide on TS-1
titanium-containing catalyst.
In our study we used the following chemicals. Sol-
vents: 1-butanol, 2-butanol, 1-propanol, isopropanol,
methanol, ethanol, water, acetone, methyl ethyl ketone
of pure or analytically pure grade, distilled and dehy-
drated before use; epichlorohydrin of pure grade,
GOST (State Standard) 12844–74; 33–34% hydrogen
peroxide of ultrapure grade [TU (Technical Specifica-
tion) 2611-069-05 807 977–2006]; and allyl chloride
(
TU 6-01-753–77). The catalyst, titanium-containing
zeolite, was prepared by the procedure described in [3]
Ti content counting on TiO 3.16%). Experiments
(
2
Today epichlorohydrin is produced on the large-
tonnage scale (the annual world production exceeds
were performed on a batch laboratory installation at
30°C, allyl chloride : hydrogen peroxide molar ratio of
3.0, and solvent : allyl chloride weight ratio of 9.1. The
content of the titanium-containing zeolite remained the
1
.3 million tons [1]). In Russia, the demand for prod-
ucts based on epichlorohydrin increases by 4–5% an-
nually, and the major fraction of epichlorohydrin is
imported.
–
1
same in all the experiments (8.0 g l ).
The reaction mixture components were analyzed by
GLC on a Tsvet 500 chromatograph under the follow-
ing conditions: 2000 × 3-mm metallic column; station-
ary phase Carbowax 6000 (15 wt % on Chromaton
N-AW); flame ionization detector; carrier gas nitrogen,
With the steadily growing demand for epichloro-
hydrin, it becomes necessary to put into operation new
facilities capable of producing a competitive product.
Our analysis of procedures for epichlorohydrin produc-
tion [2] showed that the most promising procedure is
liquid-phase oxidation of allyl chloride with hydrogen
peroxide in an organic solvent on a heterogeneous
catalyst, titanium-containing zeolite TS-1 [3]. How-
ever, published data on the solvent effect on the proc-
ess are virtually lacking.
–
1
flow rate 50 ml min ; vaporizer and column tempera-
tures 180 and 130°C, respectively. Hydrogen peroxide
was determined by iodometric titration.
Epoxidation of allyl chloride is usually performed
with aqueous solutions of hydrogen peroxide in an or-
ganic solvent. In the process, the solvent acts as a ho-
mogenizer of allyl chloride and hydrogen peroxide,
providing their reaction on the surface of a solid cata-
Here we report the results of studying the solvent
effect on liquid-phase oxidation of allyl chloride with
1
963

1
964
DANOV et al.
–1
Solvent effect on the initial rate of hydrogen peroxide conversion and on the yield of epichlorohydrin. 30°C, 8.0 g l catalyst,
catalyst : hydrogen peroxide molar ratio 3.0, solvent : allyl chloride weight ratio 9.1
Initial rate of H
2
O
2
conversion,
Yield of propylene oxide
at 50% conversion of H O , %
2 2
Solvent
–1
–1
2
mol l min , ×10
Methanol
Ethanol
1.61
0.66
0.47
0.37
0.33
0.29
0.16
0.7
49.6
45.7
43.4
42.4
42.9
25.9
–
Isopropanol
1
-Propanol
Acetone
-Butanol
Methyl ethyl ketone
-Butanol
2
1
–
Water
0.5
–
lyst. Therefore, the solvent should meet certain re-
quirements. It should exhibit good solvency (toward
both allyl chloride and hydrogen peroxide) and be
sufficiently inert under the reaction conditions. Among
diverse available solvents, we chose 1-butanol,
decreases in the order methanol, ethanol, 2-propanol,
acetone, 1-propanol. In the other solvents (methyl
ethyl ketone, 1-butanol, 1-pentanol, water), 50% con-
version of hydrogen peroxide was not attained in 180
min. Also, the solvent strongly affects the initial rate of
hydrogen peroxide conversion. This solvent effect can
be understood taking into account the mechanism of
allyl chloride epoxidation with hydrogen peroxide in a
proton-donor organic solvent on titanium-containing
silicalite [4]:
2
-butanol, 1-propanol, isopropanol, methanol, ethanol,
water, acetone, methyl ethyl ketone, and 1-pentanol.
The results are given in the table.
The table shows that the yield of epichlorohydrin
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 81 No. 11 2008

SOLVENT EFFECT ON EPOXIDATION OF ALLYL CHLORIDE
1965
Epoxidation of allyl chloride with hydrogen perox-
ide occurs on titanium atoms in the four-coordinate
state [5] via formation of an active five-membered cy-
clic complex I in which the hydrogen peroxide mole-
cule contains an active oxygen atom attacking the allyl
chloride molecule [6].
the order 2-butanol, 1-propanol, acetone, 2-propanol,
ethanol, methanol. This leads to a decrease in the reac-
tivity of the five-membered complex I and, as a result,
to a decrease in the rate of oxygen transfer from the
hydrogen peroxide molecule to allyl chloride. How-
ever, this hypothesis has not yet been confirmed ex-
perimentally.
Thus, the solvent not only acts as a homogenizer of
the mixture of the starting reactants but also actively
participates in the stabilization of the Ti peroxo com-
plex I formed by a hydrogen peroxide molecule on a
titanium center of the catalyst. Such stabilization is
possible owing to coordination of a solvent molecule
with a titanium center and formation of a hydrogen
bond with a hydrogen peroxide molecule.
Furthermore, owing to an increase in the volume of
the solvent molecule, the rate of diffusion of the sol-
vent and reaction products in catalyst pores decreases,
which leads to a decrease in the process selectivity.
Numerous studies were aimed to examine the struc-
ture of the five-membered active complex, but its
structure is still unclear because of the instability of the
complex consisting of a titanium atom incorporated in
the framework of titanium-containing silicalite, a sol-
vent molecule, and hydrogen peroxide [8, 9]. Attempts
are made [10, 11] to elucidate the structure of the five-
membered complex by quantum-chemical calculations.
It should be noted that an increase in the size of the
alkyl radical R in an alcohol molecule and, corre-
spondingly, in the size of the transition complex I
complicates the approach of an olefin molecule to the
activated complex and thus hinders the oxygen trans-
fer. Therefore, the yield of epichlorohydrin is the high-
est when methanol is used as solvent. Certain authors
believe [7] that, along with the steric factor, a decrease
in the yield of epichlorohydrin in the series methanol,
ethanol, 2-propanol, acetone, 1-propanol, 2-butanol
may be due to the fact that the binding energy of the
solvent molecule in the transition complex increases in
It is important that complex I is formed only in pro-
ton-donor solvents and is not formed in aprotic sol-
vents. However, the Ti–OOH species exists in both
systems [12]. The mechanism of epoxidation of allyl
chloride with hydrogen peroxide in an aprotic organic
solvent on a titanium-containing zeolite is as follows:
In this case, the effect of the aprotic solvent on the
epoxidation of allyl chloride can be interpreted from
the standpoint of the solvent polarity. Transition states
observed in the course of the reaction are characterized
by certain charge separation, i.e., they are more polar
than the starting reactants. This should lead to varia-
tion of the epichlorohydrin yield with variation of the
solvent permittivity. The higher the solvent permittiv-
ity, the lesser the extent to which the polar transition
state is stabilized by solvation and the weaker the ef-
fect of internal factors on the reaction rate. Thus, an
increase in the solvent polarity facilitates the cleavage
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 81 No. 11 2008

1
966
DANOV et al.
the use of concentrated hydrogen peroxide solutions.
Methanol
It is important that, in the synthesis of epichloro-
hydrin, the desired reaction can be accompanied by
side reactions, in particular, by oxidation of the solvent
(of methanol to formaldehyde, of ethanol to acetalde-
hyde, of isopropanol to acetone). It was found that the
reactivity of lower alcohols increases with an increase
in their molecular weight. When the experiments were
performed under the same conditions, formaldehyde
was not detected (trace amounts), whereas in isopropa-
nol the yield of acetone reached several percents.
Aprotic solvents (acetone, methyl ethyl ketone) do not
undergo such side transformations under the reaction
conditions. However, the use of acetone or other ke-
tones as solvents is undesirable, because in the pres-
ence of hydrogen peroxide they can form dangerously
explosive organic peroxides [14].
Ethanol
-Propanol
1
1
-Butanol
1
-Pentanol
(
ε – 1)/(2ε + 1)
0
Correlation of the initial reaction rate W with the
permittivity function (ε – 1)/(2ε + 1) in oxidation of allyl
–
1
chloride in various solvents. 30°C, catalyst content 8.0 g l ,
allyl chloride : hydrogen peroxide molar ratio 3.0, solvent :
allyl chloride weight ratio 9.1.
of the peroxy bond in complex II and the oxygen
transfer to allyl chloride. In particular, the yield of
epichlorohydrin increases as the solvent polarity grows
in going from acetone to methyl ethyl ketone. Similar
trends are observed in proton-donor solvents. The fig-
ure shows how the initial rate of allyl chloride epoxi-
dation correlates with the dielectric constant of the me-
dium within the same homologous series of alcohols
CONCLUSION
Methanol is the most suitable as solvent for epoxi-
dation of allyl chloride with hydrogen peroxide on the
commercial scale.
REFERENCES
1
2
. http://marketpublishers.com/
. Huanxin Gao, Gongxuan Lu, Jishuan Suo, and Shuben
Li, Appl. Catal. A: General, 1996, vol. 138, pp. 27–38.
(
methanol, ethanol, 1-propanol, 1-butanol, 1-pentanol).
For solvents of different classes, it is impossible to
construct a correlation between the initial rate and sol-
vent permittivity because of the growing role of steric
factors.
3
4
. US Patent 4 410 501.
. Clerici, M.G. and Inoallina, P., J. Catal., 1993, vol. 140,
pp. 71–83.
In view of the trends revealed for organic solvents,
it could be expected that the use of water as solvent
would lead to an increase in the epoxidation rate. How-
ever, despite the absence of steric hindrance caused by
large molecular volume of organic solvents and high
polarity (ε = 78.5) of water, it is a poor solvent for per-
forming the process. This may be caused by the fact
that the concentration of the five-membered active
complex in catalyst pores will be very low because of
hydrophobicity of the titanium-containing silicalite [5].
This conclusion is consistent with the fact that in other
processes the catalyst also shows lower activity with
water as solvent [4].
5
6
7
. Danov, S.M., Sulimov, A.V., and Fedosov, A.E., Katal.
Prom–sti., 2007, no. 6, pp. 13–17.
. Guo Xinwen, Wang Xiangsheng, Liu Min, et al., Catal.
Lett., 2002, vol. 81, pp. 125–131.
. Hayashi, H., Kikawa, K., Murai, Y., et al., Catal. Lett.,
1
996, vol. 36, pp. 99–102.
8
9
. Notari, B., Adv. Catal., 1996, no. 41, pp. 253–334.
. Bellussi, G. and Rigutto, M., Stud. Surf. Sci. Catal.,
1
994, vol. 85, pp. 177–183.
1
1
1
0. Neurock, M. and Manzer, L.E., Chem. Commun., 1996,
no. 10, pp. 1133–1134.
1. Vayssilov, G.N. and Van Santen, R.A., J. Catal., 1998,
no. 175, pp. 170–174.
Vankelecom et al. [13] suggested to perform oxida-
tion on titanium-containing silicalite without a solvent
in a three-phase system. Unfortunately, it is difficult to
judge the efficiency of such a procedure because of the
lack of experimental data on epoxidation of olefins.
However, it is undesirable to perform the process un-
der such conditions because of hazard associated with
2. Gang Li, Jiwen Meng, Xiangsheng Wang, and Xinwen
Guo, React. Kinet. Catal. Lett., 2004, vol. 82, pp. 73–
80.
13. Vankelecom, I., Vercruysse, K., Moens, N. et al., Chem.
Commun., 1997, no. 1, pp. 137–138.
14. Karnojitzki, V., Les peroxides organiques, Paris:
Hermann, 1958.
RUSSIAN JOURNAL OF APPLIED CHEMISTRY Vol. 81 No. 11 2008