alcohol 9a,b was obtained as a mixture of diastereomers
(1:1) in >90% yield from 8. While the separation of these
alcohols was fairly tedious, each diastereomer could be
obtained with >95% stereoisomeric purity through the use
of flash chromatography. While it is expected that stereo-
selective allylation of the cyclopropylaldehyde could be
accomplished using reagent-based methods,11 a stereorandom
allylation enabled us to test the stereospecificity of a second
ring closure.
Scheme 4a
The final sequence of steps is presented for each diaste-
reomer in Scheme 3. After silylation and ring-closing
metathesis, the fragmentation proceeded smoothly for each
isomer, yielding the cyclopropanation precursors 10a and
10b. Activation of either diastereomeric homoallylic alcohol
under our standard conditions provided high yields of the
bis-cyclopropane products. The only byproduct of this
reaction is the trimethylsilyl ether of 10a and 10b, which
can be isolated in as high as 20% yield. Apparently, the
cyclization step, which generates an equivalent of TMSOTf
is extremely fast relative to the generation of the triflate.
Although the formation of the bis-cyclopropanes was
extremely efficient, we were surprised to isolate an identical
1:1 (syn:anti) mixture of diastereomers from the reaction of
each precursor, 10a and 10b.12 This suggests that the second
ring closure is slower than ionization of the intermediate
triflate, formation of a cyclopropylcarbinyl cation, and loss
of the stereochemical integrity of the starting homoallylic
alcohol stereogenic center.
The lack of stereospecificity in the second cyclization
strongly contrasts our previously published work,13 which
contained a distal phenoxy group in place of the benzyl ether
(Scheme 1). A potential explanation is that benzyl ether
assists in the ionization of the secondary triflate and
stabilization of the homoallylic carbocation14 through pro-
posed intermediate A which then undergoes sequential
cyclizations to provide the mixture of diastereomeric bis-
cyclopropanes. On the basis of this hypothesis, it should be
possible to generate multiple cyclopropanes in a single step
from an acyclic skipped diene structure.
a (a) BuLi, THF, -78 °C; BF3‚OEt2; phenylglycidyl ether, 58%,
(b) pTsOH, CH3OH; 89%; (c) CBr4, Ph3P, CH2Cl2; 98%; (d)
Na2CO3, TBAI, CuI, DMF, propargyl-TMS; 67%; (e) H2, Lindlar,
pyr; 76%; (f) Tf2O, CH2Cl2, 2,6-lutidine; 69%.
mediately followed by acid-catalyzed deprotection. The
primary alcohol underwent selective conversion to the
propargylic bromide 12 in good yield. A skipped diyne was
then prepared by condensation of propargyltrimethylsilane
with the bromide under the conditions reported by Jeffery
et al.15 The diyne was then selectively reduced to the Z,Z-
diene 13 via Lindlar hydrogenation in the presence of
pyridine. Exposure of the 13 to the standard actiVation
conditions proVided the bis-cyclopropane 14 as a 1:1 mixture
of diastereomers in a remarkable 69% yield. Only minor
uncharacterizable byproduct formation was observed in the
1H NMR of the crude reaction mixture. Presumably, the bis-
cyclopropane forms via trapping of the intermediate cyclo-
propylcarbinyl cation.14 This appears to be faster than
trapping of the related homoallylic cation to form a ther-
modynamically more stable cyclohexene isomer through a
six-membered ring transition state. Alternatively, both carbon-
carbon bonds could be forming in a more concerted fashion
without the generation of a discrete cyclopropylcarbinyl
cation.
In summary, we have developed a practical method for
the conversion of readily available homoallylic alcohols to
trans-vinylcyclopropanes and oligocyclopropanes based upon
novel reactive intermediates. A ring-closing olefin metathesis
and an intramolecular displacement of a homoallylic triflate
with an allylsilane nucleophile highlight the efficient four-
step sequence. The route is general and applicable to the
stereocontrolled preparation of 1,2,3-trisubstituted cyclopro-
panes, which will be reported in a subsequent publication.
The stereospecificity of the second cyclization was shown
to be dependent upon the nature of functionality distal to
the cyclopropylcarbinyl triflate. The tandem cationic cy-
clization, reminiscent of the elegant work of W. S. Johnson16
is extraordinarily facile. We are continuing to explore many
aspects of this work and related synthetic and mechanistic
challenges.
Skipped diene 13 was efficiently prepared through the
sequence shown in Scheme 4. Epoxide fragmentation with
the alkynyl anion of O-THP-propargyl alcohol was im-
(10) For the preparation of olefinic diols via similar silyloxycycloalkenes,
see: Chang, S.; Grubbs, R. H. Tetrahedron Lett. 1997, 38, 4757-4760.
For the preparation of tetrahydrofurans and pyrans via similar silyloxycy-
cloalkenes, see: Meyer, C.; Cossy, J. Tetrahedron Lett. 1997, 38, 7861-
7864.
(11) (a) Keck, G. E.; Geraci, L. S. Tetrahedron Lett. 1993, 34, 7827-
7828. (b) Racherla, U. S.; Brown, H. C. J. Org. Chem. 1991, 56, 401-
404. (c) Roush, W. R.; Hoong, L. K.; Palmer, M. A. J.; Park, J. C. J. Org.
Chem. 1990, 55, 4109-4117.
(12) The relative stereochemistry of the diastereomeric homoallylic
alcohols 9a and 9b was determined by chemical correlation to previously
reported compounds. See: Mohapatra, D. K.; Datta, A. J. Org. Chem. 1998,
63, 642-646.
(13) (a) We have more recently repeated these experiments and have
shown that the phenoxy series is ∼80% stereospecific. (b) We do not believe
that the benzyl ether has an effect on the initial cyclopropanation reaction
(7 to 8, Scheme 2).
(14) The “classic” and “nonclassic” carbocationic behavior of homoallyl-
cyclopropyl carbinyl-cyclobutyl cations has recently been reviewed: Olah,
G. A.; Reddy, V. P.; Surya Prakash, G. K. Chem. ReV. 1992, 92, 69-95.
For a seminal discussion of these intermediates, see: Roberts J. D.; Mazur,
R. H. J. Am. Chem. Soc. 1951, 73, 2509.
Acknowledgment. We gratefully acknowledge support
from the National Science Foundation through an Early
Career Award (CHE97-33253). This work is also supported
(15) Jeffery, T.; Gueugnot, S.; Linstrumelle, G. Tetrahedron Lett. 1992,
33, 5757-5760.
(16) Johnson, W. S. Acc. Chem. Res. 1968, 54, 4731.
Org. Lett., Vol. 1, No. 8, 1999
1259