BiBr3-catalysed formation of cyclic carbonates from epoxides and DMF: a new oxidation reaction with molecular oxygen

Article abstract of DOI:10.1039/a606592i

Cyclic carbonates are obtained in good yields from terminal epoxides and DMF in a BiBr3-catalysed reaction with molecular oxygen as the oxidant.

Full text of DOI:10.1039/a606592i

View Article Online / Journal Homepage / Table of Contents for this issue
BiBr3-catalysed formation of cyclic carbonates from epoxides and DMF: a new
oxidation reaction with molecular oxygen
Ve´ronique Le Boisselier, Miche`le Postel and Elisabet Duna˜ch*
Laboratoire de Chimie Mole´culaire, U.R.A. 426 du C.N.R.S., Universite´ de Nice-Sophia Antipolis, Parc Valrose, 06108 Nice
Cedex 2, France
Cyclic carbonates are obtained in good yields from terminal
epoxides and DMF in a BiBr3-catalysed reaction with
molecular oxygen as the oxidant.
corresponding bromo alcohols were not isolated in the ana-
logous reaction with BiBr3. Moreover, when BiIII mandelate
was used as the catalyst in DMF, no reaction occurred, and the
epoxide was quantitatively recovered. The reaction of the
epoxide with CO2 catalysed by BiBr3 in refluxing THF did not
give any carbonate.
When the reaction of 1,2-epoxydecane in DMF catalysed by
BiBr3 was carried out under an inert atmosphere (N2), no
carbonate was formed, although some polymerisation of the
epoxide occurred. Thus, the presence of molecular oxygen is
essential for carbonate formation. The consumption of O2
during the reaction was measured, and reached 1.1 moles of O2
per mole of epoxide at complete conversion (6 h reaction at
110 °C).
A series of epoxides were reacted in DMF under molecular
oxygen (1 atm) at 110 °C, with a 10 mol% of BiBr3 [eqn. (2)].
The results concerning the cyclic carbonate formation are
presented in Table 1. The reactions were clean and selective, the
only isolated products being the cyclic carbonates, the remain-
ing by-products being attributed to epoxide polymerisation.
Aliphatic as well as aromatic terminal epoxides (entries 1–3)
afforded the corresponding cyclic carbonates in good yields. An
o-chlorine substituent on the aromatic ring (entry 4) was
compatible with the reaction conditions, although it slowed the
reaction rate, probably due to steric hindrance. Benzylic
epoxides (entries 5, 6) and epibromohydrin (entry 7) led to the
corresponding cyclic carbonates without elimination. No addi-
tional intramolecular cyclisation or double bond modification
was observed in the reaction of 7,8-epoxyoct-1-ene (entry 8),
indicating that carbonate formation does not involve radical
intermediates. Several disubstituted epoxides were also tested
In the course of our ongoing research on the use of bismuth(iii)
derivatives as catalysts in oxidation reactions,1 we have recently
found a new BiBr3-catalysed reaction which provides cyclic
carbonates from terminal epoxides and DMF. Cyclic carbonates
present interesting applications as polar aprotic solvents and as
synthetic intermediates in organic synthesis,2 and they are also
important precursors to polycarbonates and other polymeric
materials.3
The synthesis of cyclic carbonates from epoxides is generally
carried out with carbon dioxide, and various catalysts have been
reported for this reaction.4 To the best of our knowledge, no
example in the literature reports the formation of carbonates
from epoxides, DMF and O2. Moreover, there has been no
report of the use of BiBr3 as the catalyst in oxidation
reactions.
We have recently explored new possibilities for BiIII catalysis
in the oxidation of epoxides5 and a-ketols6 to carboxylic acids
by molecular oxygen. With BiIII mandelate7 as the catalyst in
Me2SO, a C–C bond cleavage occurs on the oxirane ring or on
the ketol, the carboxylic acids being obtained in 40–65% yield.
When BiBr3 was taken as the catalyst (10 mol%) for the
oxidation of epoxides in Me2SO under O2 (1 atm), the
corresponding carboxylic acids were also formed. Thus, the
reaction of styrene oxide led to benzoic acid in 46% yield
[eqn. (1)].
BiBr3 (10 mol%)
O
PhCO2H
(1)
Ph
O2 (1 atm), 80 °C, Me2SO
Interestingly, when the same reaction was carried out in
DMF, no oxidative C–C bond cleavage occurred and no
carboxylic acid was obtained. The reaction selectively led to the
formation of the cyclic carbonate 4-phenyl-1,3-dioxolan-2-one
in 56% yield [eqn. (2)]. Thus, the nature of the solvent strongly
Table 1 Oxidation of terminal epoxides to cyclic carbonatesa
Reaction
time/h
Isolated yield of
carbonate (%)
Entry
R
1
2
3
4
Ph
7.5
6
8
56
70
64
68
C8H17
C6H13
o-ClC6H4
O
O
BiBr3 (10 mol%)
31
O
O
(2)
(conversion 88%)
O2 (1 atm), 110 °C, DMF
R
5
6
7
8
PhOCH2
p-OMeC6H4OCH2
BrCH2
24
30
7
76
23
64
32
R
influences the reaction pathway, determining the nature of the
reaction products.
CH2NCHC4H8
31
(conversion 93%)
The formation of cyclic carbonates from epoxides in DMF–
O2 catalysed by BiBr3 appears to be specific to this catalytic
system. Thus, when the experiment was carried out with
1,2-epoxydecane as the model epoxide, the corresponding
cyclic carbonate was obtained in 70% yield. In the absence of
the catalyst, no carbonate was formed. The replacement of
BiBr3 by BiCl3 did not afford any carbonate, but led to the
synthesis of 2-chloro alcohols (as a mixture of isomers) in 62%
yield when 30 mol% BiCl3 was used. It is worth noting that the
a
General reaction conditions: anhydrous DMF (5 ml) was stirred in the
presence of BiBr3 (134 mg, 0.3 mmol) at 110 °C for 30 min under molecular
oxygen (1 atm), followed by the addition of the epoxide (3 mmol). The
reaction was followed by GC until complete conversion of the substrate
(unless stated). Acidic hydrolysis with aqueous 0.1 m HCl, diethyl ether
extraction, drying of the organic layers with MgSO4 and solvent
evaporation led to the cyclic cabonates in the yields shown. The products
were analysed by GC, 1H, 13C NMR and mass spectroscopy, and their
spectral data compared with those of authentic samples.
Chem. Commun., 1997
95

View Article Online
but did not afford cyclic carbonates under the same conditions.
trans-5,6-Epoxydecane led to decane-5,6-dione in 56% yield.
The reaction presumably involves an epoxide ring opening
with the incorporation and oxidation of the carbonyl moiety of
DMF, the oxidant being O2. 1,2-Diketones are not intermediates
in this reaction; the reaction of hexane-3,4-dione did not afford
any carbonate. In order to quench the amine residue eliminated
from DMF, an equivalent amount of benzoyl chloride was
added after the reaction with 1,2-epoxydecane. However, no
N,N-dimethylbenzamide was isolated. In a different experi-
ment, the reaction was carried out in N-benzylformamide. The
epoxide consumption was slow (72 h) and no cyclic carbonate
was formed. We could however isolate the carbamate 1, arising
from the oxidation of the benzylamine residue.†
V. L. B. is grateful to the French MESR for a research
grant.
Footnote
† Selected data for compound 1: 1H NMR (CDCl3): d 7.65–7.15 (5 H, m),
6.20 (NH, s), 4.50 (2 H, s), 3.85–3.60 (1 H, m), 3.12 (1 H, dd, J 15.0, 6.0
Hz), 3.08 (1 H, dd, J 15.0, 7.0 Hz), 1.40–0.78 (17 H, m). 13C NMR (CDCl3):
d 164.1, 129.0–127.6, 71.2, 69.1, 53.1, 39.7–14.3; m/z (70 eV) 342 and 340
(1:1) (M+ 2 29, 5), 291(1), 262(3), 149(68), 91(100).
References
1 M. Postel and E. Dun˜ach, Coord. Chem. Rev., 1996, in the press.
2 M. Brunner, L. Mussmann and D. Vogt, Synlett, 1993, 893.
3 S. Oi, K. Nemoto, S. Matsuno and Y. Inove, Macromol. Rapid Commun.,
1994, 15, 133; Organic and Bio-Organic Chemistry of Carbon Dioxide,
ed. S. Inoue and N. Yamazahi, Kodanska Ltd, Tokyo, 1982.
4 N. Kihara, N. Hara and T. Endon, J. Org. Chem., 1984, 115, 205;
G. Rokicki, W. Kuran and B. P. Marcinick, Monatsch. Chem., 1984, 115,
205; T. Nishikubo, A. Kameyama, J. Yamashita and T. Tomoi, J. Polym.
Sci., 1993, 31, 947; J. E. Ba¨ckvall and O. Karlson, Tetrahedron Lett.,
1980, 21, 4985; R. L. De Pasquale, J. Chem. Soc., Chem. Commun., 1973,
157; M. Brunner, L. Mussmann and D. Vogt, Synlett, 1993, 893;
P. Tascedda and E. Dun˜ach, J. Chem. Soc., Chem. Commun., 1995, 43.
5 T. Zevaco, E. Dun˜ach and M. Postel, Tetrahedron Lett., 1993, 34, 2601;
V. Le Boisselier, E. Dun˜ach and M. Postel, J. Organomet. Chem., 1994,
482, 119.
O
O
NHBn
Br
C8H17
1
Compound 1, formed in 40% yield and purified by column
chromatography, clearly establishes the epoxide–formamide
coupling and the oxidation of the formamide group. The
presence of a bromide on the less hindered side of the starting
epoxide also shows the role of the catalyst in the initial oxirane
ring opening. We can assume that an analogue of 1 is likely to
be the intermediate in the carbonate formation in DMF. The
reaction would be followed by oxygen transfer to the carbamate
and ring closure with bromide elimination.
6 V. Le Boisselier, C. Coin, M. Postel and E. Dun˜ach, Tetrahedron, 1995,
51, 4991.
7 T. Zevaco, M. Postel and N. Benali-Cherif, Main Group Metal
Chemistry, 1992, 5, 217.
Received, 25th September 1996; Com. 6/06592I
96
Chem. Commun., 1997