Transformation of cis- and trans-2,3-dimethyloxiranes on a Pd/SiO2 catalyst

Article abstract of DOI:10.1006/jcat.1997.1543

The transformation of cis- and trans-2,3-dimethyloxiranes on a Pd/SiO₂ catalyst, leading to the formation of 2-butanone and 2-butanol, was studied in hydrogen and deuterium atmosphere. The effect of hydrogen pressure (1.3-100 kPa) and temperature (323-423 K) on the reaction rate was also measured. The transformation of the two stereoisomers involves different mechanisms. In the case of the cis-isomer, hydrogen participates in the cleavage of the C-O bond and different surface species belong to the two products. In the case of the trans-isomer, ring opening by hydrogen (the formation of 2-butanol) is less significant and the main reaction is intramolecular hydrogen migration leading to the formation of 2-butanone. Considering the geometry of the adsorbed species, cis-2,3-dimethyloxirane is most probably adsorbed on the surface of the Pd catalyst in a planar manner, while the adsorption of the trans-isomer is intermediate between edgewise and planar adsorption mode.

Full text of DOI:10.1006/jcat.1997.1543

JOURNAL OF CATALYSIS 167, 242–247 (1997)
ARTICLE NO. CA971543
Transformation of cis- and trans-2,3-Dimethyloxiranes
on a Pd/SiO2 Catalyst
Andra´s Fa´si, Ferenc Notheisz, and Miha´ly Barto´k1
Department of Organic Chemistry and Organic Catalysis Research Group, Hungarian Academy of Sciences, Jo´zsef Attila University,
Do´m te´r 8, H-6720 Szeged, Hungary
Received September 9, 1996; revised November 18, 1996; accepted November 21, 1996
much more favorable than in the case of the trans-isomer.
On nickel, however, an ionic insertion mechanism occurs,
accompanied by the formation of similar surface species
from both isomers.
The transformation of cis- and trans-2,3-dimethyloxiranes on a
Pd/SiO2 catalyst, leading to the formation of 2-butanone and 2-
butanol, was studied in hydrogen and deuterium atmosphere. The
effect of hydrogen pressure (1.3–100 kPa) and temperature (323–
423 K) on the reaction rate was also measured. The transformation
of the two stereoisomers involves different mechanisms. In the case
Much information has been accumulated by earlier stud-
ies on hydrogenative ring opening on Pd catalysts (1, 15).
Although the transformation rates of the two isomers have
been examined (1), and deuteration of the mixture of the
isomers has also yielded important knowledge for the inter-
pretation of the reaction mechanism (15), transformation
of the two isomers in the function of hydrogen pressure
and deuteration of separate isomers have not been studied
as yet. It is therefore important to examine how hydrogen
pressure affects transformation on Pd catalysts and whether
the picture thereby outlined agrees with mechanisms previ-
ously hypothesized. On the other hand, the study of deuter-
ation helps one to draw conclusions regarding the struc-
ture of adsorbed species on the metal surface, which yields
useful information about the reaction mechanism and the
types of active centers involved. The results of deuteration
may then be compared with reaction mechanisms proposed
earlier.
–
of the cis-isomer, hydrogen participates in the cleavage of the C O
bond and different surface species belong to the two products. In the
case of the trans-isomer, ring opening by hydrogen (the formation of
2-butanol) is less significant and the main reaction is intramolecular
hydrogen migration leadingtotheformation of2-butanone. Consid-
ering the geometry of the adsorbed species, cis-2,3-dimethyloxirane
is most probably adsorbed on the surface of the Pd catalyst in a
planar manner, while the adsorption of the trans-isomer is inter-
c
mediate between edgewise and planar adsorption mode.
ꢀ 1997
Academic Press
INTRODUCTION
We have frequently studied the hydrogenolysis and iso-
merization of oxacycloalkanes on transition metal catalysts
(1–13). New information has been published on the behav-
ior of cis- and trans-dimethyloxiranes on Pt, Pd, and Ni cata-
lystsand on the stereochemistryofthese processes(1, 3, 6, 8,
13). The transformation of 2,3-dimethyloxiranes under H2
atmosphere on platinum, palladium, rhodium, and nickel
catalysts has also been studied by French authors (14, 15).
Recently, the two oxirane stereoisomers have been used as
suitable model compounds to find new regularities for the
isomerization reactions (16).
It has been established on the basis of earlier results that
the cis-isomer is transformed on Pt and Pd catalysts at a con-
siderably higher rate than the trans-isomer, while on nickel
the transformation rates of the two isomers are quite simi-
lar. The difference in reactivities indicates different mecha-
EXPERIMENTAL
Materials
cis- and trans-2,3-Dimethyloxiranes were purchased
from Aldrich. The purity of the isomers was 99 and 98% ,
respectively. The trans-isomer did not contain a detectable
amount of cis-isomer. Prior to use, they were purified by
several cycles of freezing–degassing–thawing.
The 3% Pd/SiO2 catalyst was prepared by impregnation
of Cab–O–Sil support by PdCl2 solution (17). After filtra-
tion the catalyst precursor was dried at 403 K for 16 h, it was
reduced in flowing hydrogen at 483 K for 3 h and at 673 K
for 4 h. The reduced catalyst was cooled in flowing nitro-
gen. The dispersion of the Pd/SiO2 catalyst was measured by
static hydrogen chemisorption at 340 K (dispersion, 0.15).
In the course of the pretreatment the catalyst was heated
to 423 K at 26.6 kPa hydrogen or deuterium pressure, kept
–
nisms. On platinum and palladium, the C O bond is cleaved
with the participation of hydrogen. In the case of the cis-
isomer surface adsorption for hydrogenolytic cleavage is
1 To whom correspondence should be addressed.
242
0021-9517/97 $25.00
c
Copyright ꢀ 1997 by Academic Press
All rights of reproduction in any form reserved.

Pd-CATALYZED RING OPENING OF OXIRANES
243
SCHEME 1. Reaction pathways in the transformations of cis- and trans-2,3-dimethyloxiranes.
at this temperature for 1 h, and cooled to the reaction tem- butanol and isomerization resulting in 2-butanone. Some
perature under continuous vacuum.
deoxygenation also occurs (Scheme 1).
Transformation rates of cis- and trans-oxiranes on
Pd/SiO2 catalyst were markedly different (Fig. 1). The cis-
isomer is already transformed to a considerable extent at
363 K, while transformation rate of the trans-isomer is one
order of magnitude lower even at 398 K. There is a signif-
icant difference between the product distributions as well:
in the case of the cis-isomer the main product is 2-butanol,
while in the case of the trans-isomer it is 2-butanone. Re-
garding the amount of hydrocarbons formed, non signifi-
cant difference is observed. 2-Butenes (less at higher hy-
drogen pressures) and butane (more at higher hydrogen
pressures) were also formed in the course of the transfor-
mation of both isomers.
Methods
Measurements were carried out in a circulation reac-
tor (3). The volume of the reactor was 69 cm3 and that of
the entire system was 170 cm3. The reactor was heated by
an air thermostat. When measuring the effect of hydrogen
pressure, analysis was carried out in a gas chromatograph
(HP 5890) equipped with a flame ionization detector. For
studies on deuterolysis, a gas chromatograph (HP 5890)
with a quadrupole mass selective detector was attached to
the system. Separation was done in a Carbowax 20M capil-
lary column (HP-20M). Data were processed by CSW and
HP5970 chromatogram analysis programs.
Hydrogen used for the measurements was generated in a
Matheson Type 8326 electrolysis apparatus equipped with
a Pd diffusion cell. Deuterium was obtained by a General
Electric 15EHG generator (the purity of D2O was 99.8% ).
Deuterium was deoxygenated by a Model 1000 Oxygen
Trap (CRS, U.S.A.).
Hydrogen Pressure Dependence
Plotting turnover frequencies against partial pressures
of hydrogen in the case of cis-2,3-dimethyloxirane (Fig. 2)
reveals that the rates of both hydrogenolysis and isomer-
ization have maxima but these do not coincide. In the case
of 2-butanone the rate is highest at around 17 kPa and at
Procedure
Fresh catalyst (20 mg) and 1.33 kPa of cis- or trans-
oxirane were added for each measurement. When the effect
of hydrogen pressure was studied, hydrogen pressure was
varied within the range of 0–101.3 kPa. Reaction tempera-
tures were 363 K for the cis- and 398 K for the trans-isomer.
The effect of temperature (323–373 K for the cis- and 363–
423 K for the trans-isomer) was studied at a hydrogen pres-
sure of 20 kPa in the case of both isomers. Initial turnover
frequencies (minꢁ1) were calculated. Based on tempera-
ture dependence, apparent activation energies were deter-
mined.
Deuterolysis was studied at a deuterium pressure of
20 kPa and a reaction temperature of 363 K for cis- and
398 K for trans-isomer. The data were used for the calcula-
tion of deuterium distribution.
RESULTS AND DISCUSSION
FIG. 1. Transformation of cis-2,3-dimethyloxirane [363 K; catalyst =
20 mg; poxirane = 1.33 kPa; phydrogen = 13.3 kPa (4, 2-butanone; , 2-buta-
nol) and 40 kPa (᭺, 2-butanone; ♦, 2-butanol)] and trans-2,3-di-
Activity and Selectivity of Transformations
methyloxirane [398 K; catalyst, 20 mg; poxirane = 1.33 kPa; phydrogen
=
Two main reactions take place in the course of the trans-
formation: hydrogenolysis leading to the formation of 2-
13.3 kPa (᭡, 2-butanone; , 2-butanol) and 40 kPa (᭹, 2-butanone; ᭜,
2-butanol] over Pd/SiO2 catalyst.

´ ´
FASI, NOTHEISZ, AND BARTOK
244
FIG. 3. Turnover frequency of product formation vs. hydrogen
FIG. 2. Turnover frequency of product formation vs. hydrogen
pressure curves at trans-2,3-dimethyloxirane (catalyst = 20 mg Pd/SiO2;
pressure curves at cis-2,3-dimethyloxirane (catalyst = 20 mg Pd/SiO2;
398 K).
363 K).
2-butanol was formed from the cis-isomer at a consider-
ably higher rate than from the trans-isomer, it seems highly
probable that the “planar” adsorption of the oxirane is op-
higher pressures it decreases sharply, while maximal rate for
2-butanol is nearly constant within a wide pressure interval
(20–53 kPa) and declines only at high hydrogen pressures.
This observation suggests that the two products are formed
from different surface complexes, and the higher sensitivity
to hydrogen pressure may indicate that the surface species
leading to the ketone is more unsaturated than the one
leading to the alcohol.
The cis- and trans-isomers display significant differences
regarding not only the rates of transformation and product
distribution but also in the type of hydrogen dependence.
The effect of hydrogen pressure on the transformation of
the trans-isomer is shown in Fig. 3. Both isomerization and
hydrogenolysis have maxima which nearly coincide (at ca.
53 kPa), suggesting that in the case of the trans-isomer the
alcohol and the ketone are formed from similar surface
species and that this surface species are similar to ones re-
sulting in alcohol formation from the cis-isomer.
–
timal for hydrogenolysis of the C O bond. In the case of the
trans-isomer there is no possibility for planar adsorption of
a similar manner. Due to significant steric hindrance by the
methyl groups, only a distorted variant of planar adsorption
may occur.
In the case of cis-compound the different hydrogen pres-
sure dependences for the formation of 2-butanol and 2-
butanone allow the conclusion to be drawn that isomer-
ization and hydrogenolysis proceed according to different
mechanisms. In the literature, isomerization of oxiranes is
usually associated with a dissociative, while the hydrogenol-
–
ysis of oxirane C O bond with an associative mechanism
(14, 15, 17). As a matter of course, a mechanism of a
dissociative character has a higher hydrogen pressure de-
pendence than has an associative one. In the case of the
–
cis-isomer, planar adsorption allows the C H bond to be
–
cleaved simultaneously with the rupture of the C O bond
(Scheme 2).
Mechanisms of Transformations
The earlier studies, based mainly on stereochemical con-
siderations, pointed to the conclusion that hydrogen partic-
In the case of the trans-isomer, the rate-limiting steps of
isomerization and hydrogenolysis are most probably iden-
tical (Scheme 3).
–
ipates in the cleavage of the reactive C O bonds (18). Since
SCHEME 2. Mechanism of cis-2,3-dimethyloxirane transformation on Pd/SiO2 catalyst.

Pd-CATALYZED RING OPENING OF OXIRANES
245
SCHEME 3. Mechanisms of 2-butanone-d1, 2-butanol-d2, and 2-butanon-d0 formations in the case of trans-2,3-dimethyloxirane.
The cause of the low transformation rate observed may
In the course of experiments on deuterium, no deuterium
be that in the case of the trans-isomer the formation of the was found in unreacted oxirane, irrespective of the kind of
surface species necessary for ring opening with the partici- isomer studied, indicating that oxirane is irreversibly ad-
pation of hydrogen is inhibited by the steric strain between sorbed on the surface of Pd and is desorbed only as a prod-
the methyl group and the surface of the catalyst. At the uct. However, in 2-butene and butane formed in the course
same time, the hydrogen of the carbon atom of the forming of the reaction the maxima of deuterium distribution was
==
–
C
O group is located far from the surface, preventing the d8 and d10, respectively, showing that the C H bonds are
–
rupture of this C H bond simultaneously with the cleav- also cleaved, simultaneously with deoxygenation.
–
age of the C O bond. According to this mechanism, the
The results of measurements in deuterium are listed in
two products have common surface species. It is the further Tables 2 and 3. In the case of the cis-isomer, the observed
transformation of this intermediate that determines the se- deuterium distribution is in good agreement with the sup-
lectivity of the reaction and selectivity depends mainly on posed reaction mechanism. According to the associative
temperature and hydrogen pressure.
Apparent activation energies (Table 1) are nearly iden- uct, the Pd OR monoadsorbed species formed after the
mechanism, resulting in the formation of the alcohol prod-
–
–
tical for the two isomers and their values are low. This cleavage of the C O bond reacts with hydrogen and desorbs
allows one to conclude that the oxirane ring is opened relatively rapidly from the surface. In the case of the ketone
simultaneously with adsorption in the temperature range a multiple exchange of an extent considerably higher than
studied.
that for the alcohol was observed, supporting the assump-
tion that the rate-limiting steps of the two reactions are
not identical indeed in the case of the cis-isomer. Dissocia-
tive bond cleavage leading to ketone formation means that
the species formed is more strongly bound to the surface
of the catalyst, creating a higher probability for multiple
exchanges.
TABLE 1
Comparison of Apparent Activation Energies
(Ea) (323–423 K; phydrogen = 20 kPa; poxirane
=
1.33 kPa; Catalyst = 20 mg Pd/SiO2)
In the case of hydrogenolysis leading to the formation of
alcohol, no significant difference was found between deu-
terium exchange by the two isomers. The main product of
deuterolysis was 2-butanol-d2 in both cases. However, a
closer examination of the data in the Tables 2 and 3 re-
veals that in the case of the cis-isomer 2-butanol-d3 appears
sooner than in the case of the trans-isomer. This observa-
tion strongly supports the assumption that the cis-isomer is
Ea (kJ/mol)
|
|
|
——
Q
Products
——
Q
ꢀ
ꢀ
|
ꢀ
ꢀ
Hydrocarbons
2-Butanone
2-Butanol
Total
5.6
14.3
20.4
16.6
18.5
20.4
14.9
15.4

´
´
246
FASI, NOTHEISZ, AND BARTOK
TABLE 2
Transformation of cis-2,3-Dimethyloxirane in Deuterium (Catalyst = 20 mg Pd/SiO2; pdeuterium = 20 kPa; poxirane = 1.33 kPa; 363 K)
Distribution of deuterium (% )
In 2-butanone
d3 d4 d5
In 2-butanol
t (min) Conversion (% ) S2-butanone S2-butanol Shydrocarbons d0
d1
d2
d6
d7
d8
d2
d3 d4 d5 d6
5
15
30
68
85
90
0.13
0.13
0.14
0.86
0.86
0.85
0.01
0.01
0.01
3.0 11.8 18.3 16.9 18.3 15.8
8.9
5.3
1.7 81.5 6.6 4.8 6.5 0.6
5.0 82.1 6.6 4.7 6.3 0.3
0
0
3.8 10.6 16.0 20.8 20.1 13.5 10.2
0
3.5 10.5 20.0 25.3 15.1 15.3 10.3 81.7 6.3 4.8 6.8 0.4
adsorbed in a planar manner, while in the case of the trans- is opened, resulting in the formation of 2-butanone-d0.
isomer only a deformed “planar-edgewise” attachment may Since this mechanism will only promote ketone forma-
occur, in the course of which the molecule cannot get close tion, the hypothesis offers an explanation why the main
enough to the surface and thereby deuterium exchange is product of the transformation of the trans-isomer is the
also inhibited.
In contrast to deuterolysis, however, there is a basic dif-
ketone.
However, a detailed analysis of the deuterium distribu-
ference between the two isomers regarding deuterium dis- tion of the ketone obtained from the trans-isomer also re-
tribution in the ketone formed. The trans-isomer yields veals that the d2-d7 deuterium distribution is highly similar
mainly 2-butanone-d0, while in the case of the cis-isomer to that observed in the case of the cis-isomer. This suggests
products in the range of d1–d5 with multiple deuteration that, in addition to hydrogenolysis leading to alcohol for-
are formed in a nearly equal ratio. The appearance of 2- mation, isomerization also takes place just like in the case of
butanone-d0 as a main product indicates that the isomer- the cis-isomer. The bulk of the ketone, however, is formed
ization of the trans-isomer proceeds mainly via intramolec- by the intramolecular mechanism. The mechanism shown
ular hydrogen migration. Transformations of this type are in Scheme 3 is basically of the associative type: therefore,
characteristic of electrophilic catalysis. The electrophilic a hydrogen dependence similar to that of the alcohol is to
sites necessary for isomerization may possibly by created be expected. At the same time it may also be established
by the effect of the deoxygenation reaction on the surface that multiple exchange takes place within the intermediate
of the catalyst (5, 11). Consequently, the isomerization of complex and not in the adsorbed ketone.
the trans-isomer necessitates a higher temperature because
It was observed in the course of the analysis of deuterium
the reaction mechanism is basically different. This conclu- distribution that the dissociative mechanism was associated
sion supports the assumption that hydrogen participates in with a higher extent of multiple exchanges than was the as-
–
the cleavage of the C O bond and it is favorable for the sociative mechanism; i.e., the stronger a species is bound
formation of the surface species necessary for this step if to the surface of the catalyst, the more extensive the ex-
the oxirane can be adsorbed in a planar manner. Adsorp- change is. This allows the conclusion to be drawn that d8
–
tion of the trans-isomer necessary for C O hydrogenol- and d10 maxima observed with butene and butane, respec-
ysis is inhibited, alcohol formation from this isomer is tively, indicate species very strongly bound to the surface;
therefore insignificant. At suitably high temperatures, how- i.e., it is species very strongly bound by multiple adsorption
ever, a possibility for intramolecular hydrogen migration that lead to deoxygenation.
TABLE 3
Transformation of trans-2,3-Dimethyloxirane in Deuterium (Catalyst = 20 mg Pd/SiO2; pdeuterium = 20 kPa; poxirane = 1.33 kPa; 398 K)
Distribution of deuterium (% )
In 2-butanone
d3 d4 d5
In 2-butanol
t (min) Conversion (% ) S2-butanone S2-butanol Shydrocarbons
d0
d1
d2
d6
d7
d8
d2
d3
d4
d5
5
15
30
6
10
11
0.62
0.64
0.63
0.24
0.17
0.18
0.14
0.19
0.19
58.7 6.9 4.1 5.2 7.5 9.4 5.2 3.0
60.4 6.5 5.0 4.8 7.7 9.1 4.1 2.1 0.3
58.2 6.8 5.0 5.4 8.4 9.4 4.4 1.5 0.9
0
100
72.2
78.1 6.5
0
0
0
0
14.5 13.3
7.4 8.0

Pd-CATALYZED RING OPENING OF OXIRANES
247
´
ACKNOWLEDGMENTS
8. Barto´k, M., Notheisz, F., and Zsigmond, A. G., J. Mol. Catal. 44, 337
(1988).
´
9. Ostgard, D., Notheisz, F., Zsigmond, A. G., Smith, G. V., and Barto´k,
Funding provided by the Hungarian National Science Foundation via
Grant OTKA T016109 is gratefully acknowledged.
M., J. Catal. 129, 519 (1991).
10. Notheisz, F., Zsigmond, A. G., Barto´k, M., Ostgard, D., and Smith,
´
G. V., Stud. Surf. Sci. Catal. 59, 153 (1991).
11. Barto´k, M., Notheisz, F., and Kiss, J., J. Catal. 68, 249 (1981).
12. Barto´k, M., and Notheisz, F., J. Chem. Soc., Chem. Commun., 744
(1979).
REFERENCES
´
1. Barto´k, M., Notheisz, F., and Zsigmond, A. G., J. Catal. 63, 364 (1980).
2. Notheisz, F., and Barto´k, M., J. Catal. 71, 331 (1981).
3. Barto´k, M., Notheisz, F., Zsigmond, A. G., and Smith, G. V., J. Catal.
13. Barto´k, M., and Notheisz, F., J. Chem. Soc., Chem. Commun., 667
(1980).
´
100, 39 (1986).
14. Cornet, D., Gault, Y., and Gault, F. G., in “Proceedings, 3rd Interna-
tional Congress on Catalysis, Amsterdam, 1964,” p. 1184. Wiley, New
York, 1965.
´
´
4. Notheisz, F., Molna´r, A., Zsigmond, A. G., and Barto´k, M., J. Catal.
98, 131 (1986).
5. Notheisz, F., Barto´k, M., and Zsigmond, A. G., React. Kinet. Catal. 15. Se´ne´chal, G., and Cornet, D., Bull. Soc. Chim. Fr., 773 (1971).
Lett. 29, 339 (1985). 16. Lifshitz, A., and Tamburu, C., J. Phys. Chem. 99, 10251 (1995).
6. Barto´k, M., Notheisz, F., Zsigmond, A. G., and Smith, G. V., J. Catal. 17. Pa´linko´, I., Notheisz, F., and Barto´k, M., J. Mol. Catal. 63, 43
´
´
102, 471 (1986).
(1990).
´
7. Notheisz, F., Zsigmond, A. G., Barto´k, M., and Smith, G. V., J. Chem.
Soc., Faraday Trans. 1 83, 2359 (1987).
18. Barto´k, M., and Notheisz, F., in “Stereochemistry of Heterogeneous
Metal Catalysis,” chapter 9, p. 435. Wiley, New York, 1985.