Desymmetrization of meso epoxides with halides: A new catalytic reaction based on mechanistic insight [20]

Full text of DOI:10.1021/ja981031l

J. Am. Chem. Soc. 1998, 120, 7139-7140
7139
Scheme 1. Mechanism of Zr-Catalyzed Epoxide
Desymmetrization
Desymmetrization of Meso Epoxides with Halides: A
New Catalytic Reaction Based on Mechanistic Insight
William A. Nugent
The DuPont Company
Central Research and DeVelopment
P.O. Box 80328, Wilmington, Delaware 19880
ReceiVed March 27, 1998
The desymmetrization of meso epoxides via the enantioselec-
tive addition of nucleophiles is an efficient strategy for asymmetric
synthesis since it simultaneously establishes two contiguous
stereogenic centers.1 Several years ago we introduced2 precatalyst
1, a zirconium complex bearing homochiral tri-2-propanol amine3
ligands, to provide a highly selective catalyst for epoxide
desymmetrization. This catalyst promotes the addition of azido-
trialkylsilanes to meso epoxides (eq 2, X ) N3) in up to 93%
enantiomeric excess. Subsequently several other catalyst systems
have been reported which promote epoxide desymmetrization in
>90% enantiomeric excess.4-6
delineated. A simplified version of the catalytic cycle is shown
in Scheme 1; complete details will be reported elsewhere.10 In
common with other catalysts for epoxide desymmetrization,4,5 the
mechanism involves two metal centers, one of which activates
the azide nucleophile while the other activates the epoxide. Of
particular interest is the involvement of a discrete zirconium azide
intermediate in which the azide is transferred to the activated
epoxide in a relatively slow subsequent step. The implication is
that if the azide could be plucked from the zirconium atom and
replaced with a different nucleophile and proVided that the
replacement process is fast relatiVe to azide transfer, then the
alternative nucleophile might likewise undergo selective transfer
to the epoxide.
As a test of this hypothesis the reaction between cyclopentene
oxide and azidotrimethylsilane was carried out as usual but with
the addition of 2 equiv of allyl iodide.11 Under these conditions,
only 4% of the usual azide product 2 was observed; the remaining
96% of the observed product was the protected â-iodohydrin 3.
Moreover, chiral gas chromatographic analysis showed that 3 was
actually formed in significantly higher enantiomeric excess (95%)
than was 2 (79%):
For its success, the desymmetrization strategy requires that
epoxide opening occur by exclusive backside attack; however,
this also imposes a limitation, namely that the products will
necessarily be trans disubstituted.7 In principle, this limitation
could be circumvented by the introduction of a reactive nucleo-
phile such as a halide8,9 which could be displaced in a subsequent
step, thus inverting the stereochemistry at this carbon atom.
However, direct addition of trialkylsilyl halides to meso epoxides
in the presence of precatalyst 1 (e.g., eq 2, X ) Cl, Br, I)
invariably gave the products in low or negligible enantiomeric
excess.
Coproduction of allyl azide in an amount equal to that of 3 was
confirmed by gas chromatography and by NMR comparison with
an authentic sample.11
The mechanism of the enantioselective addition of azidotri-
As summarized in Table 1, the reaction could also be extended
to the synthesis of protected â-bromohydrins. Because allyl
bromide is a less reactive alkylating agent than allyl iodide, a
larger excess of allyl bromide was required to suppress formation
of the azide side-product. In all cases, 20 equiv of allyl bromide
were sufficient to keep the yield of azide <5%. For the
desymmetrization of the meso cycloalkene oxides summarized
in Table 1, the enantioselectivity appears to decrease monotoni-
cally as the ring size is increased from 5 to 8. Of interest from
the standpoint of organic synthesis, the reaction proceeds in useful
yield and enantioselectivity for functional epoxides containing
the ether or ester functional groups.
methylsilane to meso epoxides catalyzed by 1 has recently been
(1) Review: Hodgson, D. M.; Gibbs, A. R.; Lee, G. P. Tetrahedron 1996,
52, 14361.
(2) Nugent, W. A. J. Am. Chem. Soc. 1992, 114, 2768. Although the detailed
structure of precatalyst 1 is unknown, it reproducibly analyzes as L2(OH)(CF3-
CO2)Zr2, where L is the deprotonated tri-2-propanol amine ligand. Upon
exposure to azidotrimethylsilane 1 is converted to the discrete unsymmetrical
dimer L2(N3)(CF3CO2)Zr2.
(3) Nugent, W. A.; Harlow, R. L. J. Am. Chem. Soc. 1994, 116, 6142. See
also: Di Furia, F.; Licini, G.; Modena, G.; Motterle, R.; Nugent, W. A. J.
Org. Chem. 1996, 61, 5175.
(4) Martinez, L. E.; Leighton, J. L.; Carsten, D. H.; Jacobsen, E. N. J. Am.
Chem. Soc. 1995, 117, 5897. Larrow, J. F.; Schaus, S. E.; Jacobsen, E. N. J.
Am. Chem. Soc. 1996, 118, 7420. Martinez, L. E.; Nugent, W. A.; Jacobsen,
E. N. J. Org. Chem. 1996, 61, 7963.
(5) Iida, T.; Yamamoto, N.; Sasai, H.; Shibasaki, M. J. Am. Chem. Soc.
1997, 119, 4783.
(6) Cole, B. M.; Shimizu, K. D.; Krueger, C. A.; Harrity, J. P. A.; Snapper,
M. L.; Hoveyda, A. H. Angew. Chem., Int. Ed. Engl. 1996, 35, 1668.
(7) For an alternative strategy, see: Schaus, S. E.; Larrow, J. F.; Jacobsen,
E. N. J. Org. Chem. 1997, 62, 4197.
(8) The stoichiometric desymmetrization of meso epoxides with aluminum
or boron halides has been reported: Naruse, Y.; Esaki, T.; Yamamoto, H.
Tetrahedron 1988, 44, 4747. Joshi, N. N.; Srebnik, M.; Brown, H. C. J. Am.
Chem. Soc. 1988, 110, 6246. Srebnik, M.; Joshi, N. N.; Brown, H. C. Isr. J.
Chem. 1989, 29, 229.
(9) Very recently Denmark and co-workers have reported the catalytic
enantioselective ring-opening of epoxides with SiCl4 to afford optically active
chlorohydrins. Interestingly, in contrast with current work, the highest ee’s
using the Denmark catalyst are obtained with acyclic meso epoxides (e.g.,
87% ee for cis-stilbene oxide). Denmark, S. E.; Barsanti, P. A.; Wong, K.-T.;
Stavenger, R. A. J. Org. Chem. 1998, 63, 2428.
(10) McCleland, B. W.; Finn, M. G.; Nugent, W. A. Submitted for
publication.
(11) No reaction was observed in the absence of zirconium catalyst. Allyl
halides react directly with azidotrimethylmethane only in highly polar solvents
such as HMPA: Nishiyama, K.; Karigomi, H. Chem. Lett. 1982, 1477.
S0002-7863(98)01031-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/03/1998

7140 J. Am. Chem. Soc., Vol. 120, No. 28, 1998
Communications to the Editor
Table 1. Synthesis of Protected â-Bromohydrins via
to convert racemic bromohydrins to the corresponding amino
alcohols; of these, the most efficient proved to be condensation/
cyclization with benzoyl isocyanate as reported by Knapp and
co-workers.14 This method has now been applied to nonracemic
bromohydrins as exemplified by eqs 4 and 5. The silylated
Desymmetrization of Meso Epoxides with Precatalyst (S,S,S)-1a
bromohydrins were first desilyated under acidic conditions
(Dowex 30/methanol). The deprotected bromohydrins were
converted to the corresponding carbamates by treatment with
benzoyl isocyanate and were then cyclized (KH/THF/reflux). The
bicyclic carbamates were formed with no loss of optical activity.
Moreover, in each case a single crystallization15 from hot toluene
was sufficient to increase the enantiomeric excess of the product
to >99%. The overall yield of enantiopure product was 78%
for eq 4 and 73% for eq 5.
The novel molecular “bait and switch” strategy that was used
to redirect Scheme 1 has opened the door to a useful new catalytic
transformation. We are currently investigating other catalytic
reactions where this strategy may be applied.
Acknowledgment. The author thanks Professors Douglass F. Taber,
M. G. Finn, Eric A. Maatta, and Barry M. Trost for helpful discussions.
Supporting Information Available: Experimental procedures for the
synthesis, isolation, and characterization of the products from Table 1
and eqs 3-5 (5 pages, print/PDF). See any current masthead page for
ordering information and Web access instructions.
a All runs contain epoxide (2.4 mmol), azidotrimethylsilane (3.0
mmol), and precatalyst 1 (0.12 mg‚atom Zr) in chlorobenzene (4 mL)
and allyl bromide (4 mL), 25 °C, 48 h. b c ) 1 (CHCl3). c Isolated
yield after flash chromatography except where indicated. d By GLC
JA981031L
f
analysis except as indicated. e By NMR with chiral shift reagent. GLC
yield (product not isolated).
(13) Wohl, R. A. J. Org. Chem. 1973, 38, 3858. Das, J. Synth. Commun.
1988, 18, 907. Samano, M. C.; Robins, M. J. Tetrahedron Lett. 1989, 30,
2329.
(14) Knapp, S.; Kukkola, P. J.; Sharma, S.; Dhar, T. G. M.; Naughton, A.
B. J. J. Org. Chem. 1990, 55, 5700. See also: McCombie, S. W.; Nagabhushan,
T. L. Tetrahedron Lett. 1987, 28, 5399. Larsen, R. D.; Davis, P.; Corley, E.
G.; Reider, P. J.; Lamanec, T. R.; Grabowski, E. J. J. J. Org. Chem. 1990,
55, 299.
(15) This useful situation undoubtedly reflects the significantly higher
melting points of the enantiopure products versus the corresponding racemates.
This difference is 29 °C for the product of eq 4 and 23 °C for the product of
eq 5.
We anticipate that the enantiopure â-bromohydrins in Table 1
will serve as broadly useful chiral building blocks.12 One
application that we have begun to explore is the conversion of
the bromohydrin products to enantiopure cis â-amino alcohols.
We examined several literature procedures13 which were reported
(12) However, reactions requiring strongly basic conditions must be avoided
because of the potential reversion of these products to the starting epoxides.