Mechanism of olefin epoxidation in the presence of a titanium-containing zeolite

Article abstract of DOI:10.1134/S003602441311006X

The effect of the nature of a solvent on the liquid-phase epoxidation of olefins with an aqueous solution of hydrogen peroxide over a titanium-containing zeolite is studied. Butanol-1, butanol-2, propanol-1, isopropanol, methanol, ethanol, water, acetone, methyl ethyl ketone, isobutanol, and tert-butanol are examined as solvents. A mechanism of olefin epoxidation with hydrogen peroxide in an alcohol medium over a titanium-containing zeolite is proposed. Epoxidation reactions involving hydrogen peroxide and different olefins are studied experimentally.

Full text of DOI:10.1134/S003602441311006X

ISSN 0036ꢀ0244, Russian Journal of Physical Chemistry A, 2013, Vol. 87, No. 11, pp. 1809–1812. © Pleiades Publishing, Ltd., 2013.
Original Russian Text © S.M. Danov, V.L. Krasnov, A.V. Sulimov, A.V. Ovcharova, 2013, published in Zhurnal Fizicheskoi Khimii, 2013, Vol. 87, No. 11, pp. 1841–1845.
CHEMICAL KINETICS
AND CATALYSIS
Mechanism of Olefin Epoxidation in the Presence
of a TitaniumꢀContaining Zeolite
S. M. Danov, V. L. Krasnov, A. V. Sulimov, and A. V. Ovcharova
Dzerzhinsk Polytechnic Institute, Nizhni Novgorod State Technical University, Dzerzhinsk,
Nizhni Novgorod oblast, 606000 Russia
eꢀmail: epoxide@mail.ru
Received January 14, 2013
Abstract—The effect of the nature of a solvent on the liquidꢀphase epoxidation of olefins with an aqueous
solution of hydrogen peroxide over a titaniumꢀcontaining zeolite is studied. Butanolꢀ1, butanolꢀ2, propanolꢀ1,
isopropanol, methanol, ethanol, water, acetone, methyl ethyl ketone, isobutanol, and tertꢀbutanol are examꢀ
ined as solvents. A mechanism of olefin epoxidation with hydrogen peroxide in an alcohol medium over a titaꢀ
niumꢀcontaining zeolite is proposed. Epoxidation reactions involving hydrogen peroxide and different oleꢀ
fins are studied experimentally.
Keywords: epoxidation, titaniumꢀcontaining zeolite, hydrogen peroxide, propylene oxide, epichlorohydrin,
glycidol.
DOI: 10.1134/S003602441311006X
INTRODUCTION
EXPERIMENTAL
The following solvents were used in this work:
butanolꢀ1, butanolꢀ2, propanolꢀ1, isopropanol,
methanol, ethanol, acetone, methyl ethyl ketone,
isobutanol, and tertꢀbutanol (highꢀpurity or analytical
grade), all of which were preliminarily dried and disꢀ
tilled; propylene oxide (highꢀpurity grade, GOST
Oxides of olefins (propylene oxide, epichlorohyꢀ
drin, glycidol, etc.) are valuable intermediates in comꢀ
mercial base organic and petrochemical synthesis.
Since these oxides exhibit high reactivity, they form
the basis for largeꢀscale processes of organic synthesis
of many useful materials, e.g., glycols, carbitols, glycꢀ
erin, cyclic carbonates, thioglycols, nonionic surfacꢀ
tants, etc. [1].
2
3001ꢀ88), epichlorohydrin (highꢀpurity grade,
GOST 12844ꢀ74), glycidol (highꢀpurity grade, TU 6ꢀ
9ꢀ14ꢀ2635ꢀ79), 33–34% hydrogen peroxide (special
0
purity grade, TU 2611ꢀ069ꢀ05807977ꢀ2006), propyꢀ
lene (GOST 25043ꢀ87), allyl chloride (TU 6ꢀ01ꢀ753ꢀ
A new method for producing olefin oxides is olefin
epoxidation with an aqueous solution of hydrogen
peroxide in the presence of heterogeneous catalysts
7
7), and allyl alcohol (TU 6ꢀ01ꢀ753ꢀ77). TCZ used as
a catalyst was prepared as described in [3]
Ti concentration in terms of TiO2, 3.16%; Si/Ti =
5). To characterize the TCZ, Xꢀray diffraction analyꢀ
[
2]. It is now generally thought that catalyst systems
(
2
based on titaniumꢀcontaining zeolites (TCZs) exhibit
the highest activity for these processes. They are
regarded as promising and owing the possibility of
exercising interrelated control over the composition
and microstructure of catalysts at the molecular level,
enabling us to prepare highly active samples with speꢀ
cific catalytic properties.
sis was performed, IR spectra were measured, and the
porous structure was examined. The powdered TCZ
sample prepared under laboratory conditions had a
2
specific surface area of 316.66 m /g, a pore volume of
3
0
4
.182 cm /g, and a characteristic diameter of 3.2–
.5 nm.
Despite the increasing importance of the direct
epoxidation of olefins, there have been very few literaꢀ oratory setup at a temperature of 35
°C, an olefin to
The experiments were conducted using a batch labꢀ
ture data on the mechanism and physicochemical laws hydrogen peroxide molar ratio of 3.0, and a solvent
of the process until recently. The aim of this work is, on concentration of 81.5 wt %.
the one hand, to determine the mechanism of the hetꢀ
erogeneous catalytic epoxidation of olefins and assess by GLC using a Khromos GKhꢀ1000 chromatograph
their reactivity in the presence of different solvents equipped with a flame ionization detector and a metal
The reaction mixture components were analyzed
and, on the other hand, to provide scientific foundaꢀ column (2 m
tions for the technology of synthesizing respective bowax 6000 supported on ChromatonꢀNꢀAW. Nitroꢀ
ꢀolefin oxides. gen was used as the carrier gas; its flow rate through the
×
3 mm); the sorbent was 15 wt % Carꢀ
α
1809

1810
DANOV et al.
Effect of the nature of the solvent on the yield (%) and initial rate of formation of epoxy compounds (mol/(s g cat))
Yield
Initial rate of formation
ECH
Solvent
PO
ХHP = 50%)
ECH
ХHP = 30%)
GD
ХHP = 30%)
PO
GD
(
(
(
Methanol
49.7
47.9
47.0
44.3
43.2
42.1
40.7
38.7
–
29.8
28.1
26.9
26.1
25.9
23.9
24.4
23.7
21.7
–
29.3
26.6
26.5
25.5
24.1
22.9
23.5
20.8
9.2
1.70
×
×
×
×
×
×
×
×
×
×
10–4
3.22
1.96
1.85
1.05
9.99
4.50
5.94
4.28
3.11
×
×
×
×
×
×
×
×
×
–
10–5
10–5
10–5
10–5
10–6
10–6
10–6
10–6
10–6
2.52
×
×
×
×
×
×
×
×
×
–
10–5
Ethanol
7.13
6.64
3.34
2.14
1.00
1.06
6.99
4.83
3.44
10–5
10–5
10–5
10–5
10–5
10–5
10–6
10–6
10–6
1.46
7.83
5.94
5.78
3.30
4.05
2.90
2.42
10–5
10–6
10–6
10–6
10–6
10–6
10–6
10–6
Isopropanol
Acetone
Propanolꢀ1
Butanolꢀ1
Methyl ethyl ketone
Butanolꢀ2
Isobutanol
tertꢀButanol
–
–
Note: PO is propylene oxide, ECH is epichlorohydrin, and GD is glycidol.
column was 50 mL/min. Hydrogen peroxide was
determined via iodometric titration.
Certain requirements are imposed on the selection
of the solvent: it must exhibit good solubility (with
respect to both the organic component and hydrogen
peroxide) and be fairly inert under the process condiꢀ
tions [4]. We selected the following solvents for this
work: methanol, ethanol, propanolꢀ1, butanolꢀ1,
butanolꢀ2, isobutanol, isopropanol, acetone, methyl
ethyl ketone, and tertꢀbutanol. The results from our
study are represented in the table.
The IR spectra of the catalyst samples were meaꢀ
sured in air at room temperature in KBr pellets in the
–1
range of 400–4000 cm using a Shimadzu IRAffinityꢀ
FTIR spectrometer. The XRD patterns of powdered
samples of the compounds were recorded using a Shiꢀ
madzu LAB XRDꢀ6000 diffractometer (Cu radiaꢀ
1
K
α
tion, nickel filter, scintillation counter, a voltage of
3
2
0
0 kV, and a current of 30 mA) in an angle range of
= 10 –80 ; the scan rate was 2 /min in steps of
.02 . The compounds were identified using the
It is evident from the table that the yield of the
desired product (at a constant degree of ХHP converꢀ
sion) diminishes in the order methanol > ethanol >
isopropanol > acetone > propanolꢀ1 > methyl ethyl
ketone > butanolꢀ1 > butanolꢀ2. For the other solvents
θ
°
°
°
°
JCPDS database.
The specific surface area, total pore volume, and
pore size distribution of solid dispersed materials were
measured using a Micromeritics TriStar 3020 autoꢀ
mated gas adsorption analyzer. The parameters of the
porous structure of the catalyst samples were automatꢀ considerable effect on the initial rate of formation of
ically calculated from the acquired data.
(
isobutanol, tertꢀbutanol), the degree of conversion of
hydrogen peroxide remained less than 50% after the
80 min experiment. In addition, the solvent had a
1
epoxy compound.
This effect of the nature of the solvent on the proꢀ
cess can be attributed to the epoxidation of an unsatꢀ
urated substrate with hydrogen peroxide being mediꢀ
ated by tetracoordinated titanium atoms [5] and
occurring through the intermediate formation of a
transition complex with solvent molecules. It is known
from the literature [6] that when hydrogen peroxide
acts on titanium(IV) compounds, we observe the forꢀ
mation of peroxide bonds. When using TCZs in epoxiꢀ
dation, the action of hydrogen peroxide normally
causes reversible cleavage of the Ti–O–Si moiety to
RESULTS AND DISCUSSION
Since the mutual solubility of an organic substrate
propylene, allyl chloride) and an aqueous solution of
(
hydrogen peroxide is extremely low, epoxidation must
be conducted in a medium of an organic solvent that
homogenizes the unsaturated compound and hydroꢀ
gen peroxide and ensures their interaction on the surꢀ
face of the solid catalyst. It should be noted that allyl
alcohol dissolves aqueous hydrogen peroxide fairly
well, and it is not necessary to use a solvent. However, form Ti–OOH peroxytitanium bonds and Si–OH; in
our studies showed that for the selective production of addition, the solvent molecules are actively involved in
glycidol, it is better to epoxidate allyl alcohol in presꢀ stabilizing the resulting peroxytitanium bond. The
ence of a solvent. Using a solvent in the oxidation of subsequent action on the catalytic center by the olefin
allyl alcohol in an amount of 40% thus enables us to molecules results in the formation of
raise the selectivity for glycidol from 78.6 to 94.7%. the scheme in Fig. 1).
π
ꢀcomplex (see
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 87
No. 11 2013

MECHANISM OF OLEFIN EPOXIDATION IN THE PRESENCE
1811
R
H
R
H
R
H
O
Si
Si
O
Ti
O
O
Ti
O
Si
Si
Si
Si
Si
Si
Si
O
O
OH
Si
O
O
O
O
O
O
O
OH H2
ROH
H
2
O
2
C=CH–CH
3
+
Ti
+
Ti
–ROH
Si
O
O
O
O
O
R
O
O
R
O
H
Si
O
Si
Si
Si
Si
H
H C CH
2
H
H
CH3
R
H
Si
O
Si
Si
−
O
O
O
O
Ti
O
H
+
≠
≠
R
H
R
H
CH2 HC O O H
O
Si
O
Si
Si
CH3
Si
δ−
Si
O
O
R
R
H
O
O
(I)
O
δ−
Ti
H
Ti
O
O
Si
O
Rateꢀlimiting stage
O
δ+
O
Si
δ+ H
O
O
Si
H
CH2
O
H C
Si
2
O
Ti
O
H
Si
Si
Si
R
H
CH
H
CH
O
O
−
O
H C
(III)
3
CH3
H
H
+
O
R
CH2 HC O O
Si
CH3
H
(II)
R
H
O
Ti
O
Si
Si
O
O
O
+
+ H O
2
CH CH3
O
R
O
H2C
Si
Si
H
Fig. 1. Mechanism of olefin epoxidation with hydrogen peroxide in an alcohol medium over a TCZ.
An important step for catalysis is a subsequent electrophilic oxygen atom of the peroxide group,
change in the type of bonding from the initial ꢀinterꢀ which emerges upon the cleaving of the water moleꢀ
action to the Ti–C ꢀbond, which is accompanied by cule from form II. The possibility of H C–O bonds
π
σ
2
the effective displacement of the –OOH ligand from forming at this final stage of the process is ensured by
Ti to C. The mechanism of this interaction is similar to the 1,3ꢀcontact between the carbon and oxygen atoms
the action of Ziegler–Natta catalysts in the polymerꢀ that form an oxirane ring, the length of which can be
ization of ꢀolefins. A distinctive feature of an olefin’s calculated as the sum of the intramolecular radii of
α
incorporation into a peroxytitanium bond is the probꢀ carbon (125 pm) and oxygen (113 pm) [7]; i.e., it is as
ability of a fiveꢀmembered cyclic transition state formꢀ low as 2.38 Å. The breaking of the oxygen–oxygen and
ing at the rateꢀlimiting stage. This interaction leads to carbon–titanium bonds converts this 1,3ꢀcontact to a
the equilibrium formation of oxonium forms I and II C–O covalent bond (bond length, 1.43 Å).
of the
σ
ꢀcomplex on the surface of the catalyst. The
Kinetic studies of the epoxidation of propylene,
allyl chloride, and allyl alcohol have confirmed
assumptions about the mechanism of olefin epoxidaꢀ
tion. The results are discussed in detail in [8–10].
formation of reaction form II of the
moted owing to the presence of hydrogen bonds
between the silicate and peroxide groups.
σ
ꢀcomplex is proꢀ
The conversion of form II to the original complex
It should be noted that an increase in the size of the
of the catalyst with the solvent (which is accompanied alkyl substituent
by the formation of reaction products) occurs accordꢀ accordingly, in the sizes of the catalyst complex inhibꢀ
ing to a push–pull mechanism of intramolecular elecꢀ its the incorporation of the olefin molecule into the
R in the alcohol molecule and,
π
ꢀ
trophilic substitution on the carbon atoms; the strucꢀ complex and hinders the proposed mechanism of the
ture of transition state III shows that in this case, the heterogeneous catalytic oxidation of alkenes. This is
pulling effect on the outgoing moiety is due to the silꢀ responsible for the highest yield of the epoxy comꢀ
icate group, while the pushing is accomplished by the pound when using methanol as a solvent. Along with
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 87
No. 11 2013

1812
DANOV et al.
torily described by the Taft correlating equation
(Fig. 2).
The slope of the line yields a value of –0.47, the
7
+ log
k
2
1
1
0
.0
2
sign of which is characteristic of the reactions from
adding electrophilics to olefins; the low value of the
slope suggests that the electron density of the double
bond of the olefin has a relatively weak effect, indicatꢀ
ing a cyclic transition state at the rateꢀlimiting stage of
the reaction. An exception from the above set of unsatꢀ
urated substrates is allyl alcohol, which is poorly
described by the Taft correlation equation. The lower
rates of epoxidation of allyl alcohol relative to allyl
chloride (which is more electronꢀdeficient) probably
result from the formation of hydrogen bonds and the
association of the molecules of allyl alcohol in the
reaction medium as well as from its ability to interact
1
3
.5
.0
.5
4
6
5
*
σ
0
0.5
1.0
with the catalyst complex without involving
in complexation.
π
ꢀbonds
Fig. 2. Reactivity of olefins according to Taft: (
, ( ) penteneꢀ1, ( ) buteneꢀ1, ( ) propylene, (
alcohol, and ( ) allyl chloride.
1
) hexeneꢀ
5
1
2
3
4
) allyl
6
CONCLUSIONS
According to the above reaction scheme, the deciꢀ
sive criterion for estimating the reactivity of olefins is
their ability to form ꢀcomplexes with catalytic cenꢀ
ters containing peroxytitanium bonds. One of the facꢀ
tors that contribute to the formation of these transition
states is the solvent.
steric hindrances, in the drop in the activity of the solꢀ
vents in the order methanol > ethanol > propanolꢀ2 >
butanolꢀ2 is a result of the lower electrophylicity of the
catalyst complex, which in turn also reduces the conꢀ
π
centration of the olefin’s ꢀcomplex and thus slows
π
the rate of the epoxidation reaction.
The rate of the reaction that occurs by this mechaꢀ
nism also depends on the structure of the olefin. Like
the process involving organic hydroperoxides or perꢀ
oxoacids in the presence of group IV–VI metals, the
epoxidation of unsaturated compounds with hydrogen
peroxide catalyzed by TCZs occurs through the interꢀ
mediate formation of a catalyst peroxocomplex with
oxidizing ability that acts as an electrophile. The rate
of these reactions depends on the presence of elecꢀ
tronꢀdonor or electronꢀacceptor substituents on the
double bond.
To better understand specific features of the proꢀ
cesses under discussion, we experimentally studied
epoxidation reactions involving hydrogen peroxide
and different olefins. Buteneꢀ1, penteneꢀ1, and hexꢀ
eneꢀ1 were selected for study in addition to propylene,
allyl chloride, and allyl alcohol. Our experiments conꢀ
firmed that the rate of the epoxidation reaction
depends on the nature of the olefin and the presence of
substituents. The initial rate of epoxidation slows in the
order hexeneꢀ1 > penteneꢀ1 > buteneꢀ1 > propylene >
allyl chloride > allyl alcohol. Analysis of the observed
relationships shows that with the exception of allyl
alcohol, the reactivity of the studied olefins is satisfacꢀ
REFERENCES
1
2
. http://marketpublishers.com
. H. Gao, G. Lu, J. Suo, and S. Li, Appl. Catal. A: Gen.
1
38, 27 (1996).
3
4
5
6
7
8
9
. M. Taramasso, G. Perego, and B. Notari, USA Patent
No. 4410501 (1983).
. C. Reichardt, Solvents and Solvent Effects in Organic
Chemistry (VCH, Weinheim, 1988).
. S. M. Danov, A. V. Sulimov, and A. E. Fedosov, Katal.
Promyshl., No. 6, 13 (2007).
. G. P. Luchinskii, The Chemistry of Titanium (Khimiya,
Moscow, 1971) [in Russian].
. I. Hargittai and M. Hargittai, Symmetry through the
Eyes of a Chemist (Springer, 1995; Mir, Moscow, 1989).
. S. M. Danov, A. V. Sulimov, V. A. Kolesnikov, and
A. V. Ovcharova, Kinet. Catal. 52, 793 (2011).
. S. M. Danov, A. V. Sulimov, V. A. Kolesnikov, and
A. A. Ovcharov, Khim. Promyshl. Segodnya, No. 9, 23
(2012).
1
0. S. M. Danov, A. A. Ovcharov, and A. V. Sulimov, Khim.
Interesah Ustoich. Razvit. 20, 693 (2012).
Translated by M. Timoshinina
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 87
No. 11 2013