Novel Catalytic Kinetic Resolution of Racemic Epoxides to Allylic Alcohols

Article abstract of DOI:10.1021/ol025983e

(Matrix Presented) The kinetic resolution of racemic epoxides via catalytic enantioselective rearrangement to allylic alcohols was investigated. Using the Li-salt of (1S,3R,4R)-3-(pyrrolidinyl)methyl-2-azabicyclo [2.2.1] heptane 1 as catalyst allowed both

Full text of DOI:10.1021/ol025983e

ORGANIC
LETTERS
2002
Vol. 4, No. 22
3777-3779
Novel Catalytic Kinetic Resolution of
Racemic Epoxides to Allylic Alcohols
Arnaud Gayet, Sophie Bertilsson, and Pher G. Andersson*
Department of Organic Chemistry, Institute of Chemistry, Uppsala UniVersity,
Box 531, S-751 21 Uppsala, Sweden
phera@kemi.uu.se
Received April 8, 2002 (Revised Manuscript Received July 17, 2002)
ABSTRACT
The kinetic resolution of racemic epoxides via catalytic enantioselective rearrangement to allylic alcohols was investigated. Using the Li-salt
of (1S,3R,4R)-3-(pyrrolidinyl)methyl-2-azabicyclo [2.2.1] heptane 1 as catalyst allowed both epoxides and allylic alcohols to be obtained in an
enantioenriched form.
Kinetic resolution involves a chiral reagent that promotes
the selective reaction of one enantiomer over the other to
give both the starting material and product in an enantio-
merically enriched form.1 This methodology is a well-
established approach to the generation of optically active
compounds.2
(S)-2-(pyrrolidin-1-yl)methylpyrrolidide, employing two dif-
ferent cis-3-alkylcyclohexene oxides as substrates. However
a stoichiometric amount of this lithium base was needed in
order to produce allylic alcohol in an enantiomeric excess
over 60%. Lithium amide Li-1 (Scheme 1) was introduced
We have investigated the asymmetric â-elimination of
racemic epoxides as a route to produce the optically active
epoxides and allylic alcohols.3 The epoxide deprotonation
reaction has been widely employed for the synthesis of
enantioenriched allylic alcohols from meso-epoxides. A
number of chiral, nonracemic bases have been developed
and used in the reaction.4,5 However, only two of the bases
reported have proven to be useful in the catalytic version of
the reaction.6,7 With regards to kinetic resolution, Asami8
and co-workers used a chiral lithium amide base, lithium
Scheme 1. Kinetic Resolution of cis-ss-Methylstyrene Oxide
Promoted by Li-1
(1) (a) Kagan, H. B.; Fiaud, J. C. Topics in Stereochemistry; Eliel, E.
L., Fiaud, J. C., Eds.; Wiley: New York, 1988; Vol. 18, pp 249-330. (b)
Kagan, H. B. Tetrahedron 2001, 57, 2449-2468.
by our research group in 1998 and has proven to be an
excellent catalyst for a large group of meso substrates.6
However, the use of 1 in the kinetic resolution of racemic
(2) Keith, J. M.; Larrow, J. F.; Jacobsen, E. N. AdV. Synth. Catal. 2001,
343, 5-26, and references therein.
(3) For earlier studies in this field, see: (a) Mori, K.; Hazra, B. G.;
Pfeiffer, R. J.; Gupta, A. K.; Lindgren, B. S. Tetrahedron 1987, 43, 2249-
2254. (b) Asami, M.; Kanemaki, N. Tetrahedron Lett. 1989, 30, 2125-
2128. (c) Bigi, A.; Mordini, A.; Thurner, A.; Faigl, F.; Poli, G.; To˜ke, L.
Tetrahedron: Asymmetry 1998, 9, 2293-2299.
(4) For reviews, see: (a) O’Brien, P. J. Chem. Soc., Perkin Trans. 1
1998, 1439-1457. (b) Asami, M. J. Synth. Org. Chem. Jpn. 1996, 54, 188-
199. (c) Hodgson, D. M.; Gibbs, A. R.; Lee, G. P. Tetrahedron 1996, 52,
14361-14384. (d) Cox, P. J.; Simpkins, N. S. Tetrahedron: Asymmetry
1991, 2, 1-26.
(5) For the first reports on the different groups of lithium amide bases,
see: (a) Asami, M. Chem. Lett. 1984, 829-832. (b) Bhuniya, D.; Singh,
V. K. Synth. Commun. 1994, 24, 1475-1481. (c) Milne, D.; Murphy, P. J.
J. Chem. Soc., Chem. Commun. 1993, 884-886. (d) Tierney, J. P.; Alexakis,
A.; Mangeney, P. Tetrahedron: Asymmetry 1997, 8, 1019-1022.
(6) (a) So¨dergren, M. J.; Andersson, P. G. J. Am. Chem. Soc. 1998, 120,
10760-10761. (b) So¨dergren, M. J.; Bertilsson, S. K.; Andersson, P. G. J.
Am. Chem. Soc. 2000, 122, 6610-6618.
10.1021/ol025983e CCC: $22.00 © 2002 American Chemical Society
Published on Web 10/05/2002

Table 1. Kinetic Resolution of Racemic Epoxides Using Li-1
a All reactions were set up using the conditions described in Scheme 1. Conversions were determined by CSP (chiral stationary phase, chirasil Dex-CB)
on the basis of epoxide consumption relative to an internal standard (n-dodecane). b Determined by CSP. c Isolated yields. d Absolute configuration tentatively
assigned on the basis of the stereochemical model published earlier.6b
epoxides has only been briefly investigated.6b In this initial
study, cis-â-methylstyrene oxide was resolved to produce
both the epoxide and the allylic alcohol in enantioselectivities
of 77-94% (Scheme 1). Also, 1-methylcyclohexene oxide
was rearranged with high enantioselectivity (see Table 1,
entries 1-2).
Encouraged by our preliminary results, we started inves-
tigating the possibility of broadening the scope of the
reaction. A group of racemic cyclic epoxides were prepared
and resolved via the preferential rearrangement of one of
the enantiomers to an allylic alcohol when treated with Li-1
(10 mol %). All reactions were performed at 0 °C using an
excess of DBU9 and LDA as the stoichiometric base. To
produce both the unreacted epoxide and the allylic alcohol
in high ees, the reaction was quenched either shortly before
or after 50% conversion (Table 1).
The 1-alkyl-substituted cyclohexene oxides investigated
produced enantioenriched epoxides and tertiary alcohols in
up to 99% ee at conversions close to 50%. The best results
were obtained for the rearrangement of 1-tert-butylcyclo-
hexene oxide, where both the epoxide 10 and the allylic
alcohol 11 were formed with excellent asymmetric induction
at 58% conversion10 (Table 1, entry 8). The rearrangement
of 1-methyl cyclohexene oxide (entries 1 and 2) produces
the highly enantioenriched allylic alcohol 5. This is particu-
larly interesting since 1-methylcyclohex-2-en-1-ol is an
aggregation pheromone of the beetle Dendroctonus pseu-
dosugae Hopkins. Ethyl-substituted epoxide was resolved
giving 6 with moderate enantioselectivity (e70% ee), while
the enantiomeric excess of the corresponding allylic alcohol
obtained was high (90%) (entries 3 and 4). When the alkyl
substituent in the 1-position was changed to the more bulky
isopropyl, the epoxide was enriched in up to 92% ee (entries
5 and 6). Unfortunately, the two enantiomers of the allylic
alcohol product could not be separated either by chiral GC
or derivatization (Mosher ester). From the results, it is
obvious that the selectivity of the reaction is related to the
size of the substituent in the 1-position. The more bulky this
substituent is, the more selective the reaction is.
Other 1-tert-butyl-substituted cyclic epoxides were treated
and gave encouraging results, but they were inferior to the
six-membered ring analogue. The 1-tert-butylcyclopentene
oxide 16 gave moderate ees, 79% for the allylic alcohol 17
at 41% conversion (entry 12), while the 1-tert-cycloheptene
oxide 18 leads to the corresponding allylic alcohol 19 (41%
conversion) in 94% ee (entry 14). Unfortunately, we were
(9) The beneficiary effect of DBU (1,8-diazabicyclo[5.4.0]undec-7-ene)
in the catalytic rearrangement reaction has been shown earlier; see: Asami,
M.; Ishizaki, T.; Inoue, S. Tetrahedron: Asymmetry 1994, 5, 793-796 and
refs 6 and 3c therein.
(10) Also, about 10% yield of 2-tert-butylcyclohex-2-en-1-ol was
obtained in the reaction.
(7) Asami, M.; Suga, T.; Honda, K.; Inoue, S. Tetrahedron Lett. 1997,
38, 6425-6428.
(8) Asami, M.; Sato, S.; Honda, K.; Inoue, S. Heterocycle 2000, 3, 1029-
1032.
3778
Org. Lett., Vol. 4, No. 22, 2002

not able determine the enantiomeric purity of the epoxide
using chiral GC.
Scheme 2. Kinetic Resolution of trans-4,5-Dimethyl
In earlier investigations, a competing â-hydrogen abstrac-
tion in the exo-cyclic position was detected for 1-alkyl-
substituted epoxides.11 This feature was also observed for
some of the substrates investigated. However, the byproduct
was formed in small amounts (<15%), except in the case of
1-methyl-cycloheptene oxide where the abstraction occurs
exclusively on the exo-cyclic position and leads to a racemic
allylic alcohol.
Cyclohexene Oxide
Furthermore, we were interested in 2,2-disubstituted cy-
clohexene oxides as substrates because of the fact that they
only have one â-hydrogen cis to the epoxide oxygen
accessible to the chiral base. This should make these
substrates good candidates for the kinetic resolution. Two
substrates were synthesized and tested. As expected, we
obtained very good to excellent ees for both candidates. The
2,2-dimethyl-cyclohexene oxide 12 was tried first; the
reaction was stopped at 52% conversion, and the allylic
alcohol 13 was obtained in 94% ee and the epoxide 12 in
99% ee (entry 9). When the methyl substituents were
replaced by a spiro ring, the allylic alcohol 15 was obtained
in 90% ee after 56% conversion, while the enantiomeric
excess of the epoxide was 93% (entry 10).
a kinetic resolution (Scheme 2). However, the enantioselec-
tivity dropped dramatically when the center of chirality was
moved to the 4- and 5-positions of the epoxide ring.
The allylic alcohols were easily separated from the epoxide
substrates by column chromatography after the resolution.
This makes the procedure highly useful for the preparation
of both enantioenriched epoxides and allylic alcohols.
Acknowledgment. This work was supported by the
Swedish Natural Research Council (NFR), the Swedish
Research Council for Engineering Sciences (TFR), and the
Swedish Council for Strategic Research (SSF).
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Racemic trans-4,5-dimethylcyclohexene oxide 20 was also
subjected to the rearrangement reaction in order to achieve
(11) Crandall, J. K.; Lin, L.-H. C. J. Org. Chem. 1968, 33, 2375-2378.
See also ref 3a.
OL025983E
Org. Lett., Vol. 4, No. 22, 2002
3779