Evidence for Concert in the Thermal Unimolecular Vinylcyclopropane to Cyclopentene Sigmatropic 1,3-Shift

Article abstract of DOI:10.1021/ja951578p

Gas phase pyrolysis of cis-2,3-dideuterio-trans-(1'-tert-butyl-2'-(Z)-deuteriovinyl)cyclopropane at 290 deg C gives trans,trans-3,4,5-trideuterio-1-tert-butylcyclopentene as the major 1,3-shift product with greater than 90percent stereospecificity in an orbital symmetry "allowed" suprafacial-inversion sense after correction for the geometric isomerization of starting material and the other materials present.The isotopic substitution at the critical sites of rearrangement eliminates steric of electronic influences of substitutents on a biradical pathway as a source of the suprafacial- inversion stereochemistry observed with more highly substituted drivatives.The stereochemical results coupled with a normal deuterium kinetic isotope effect (k^H/2k^D=1.14 at 311.6 deg C) at the exo-methylene carbon are best interpreted in terms of a concerted pathway for rearrangement.A less likely alternative is a stereospecific disrotatory ring opening to a biradical followed by rate-determining closure to a five-membered ring.Accompanying the rearrangement is geometric isomerization of starting material resulting from C-1-C-2 bond fission which favors either the single methylene rotation or a double rotation by a factor of 10 over single rotation at the vinyl-bearing carbon.

Full text of DOI:10.1021/ja951578p

VOLUME 118, NUMBER 2
J ANUARY 17, 1996
© Copyright 1996 by the
American Chemical Society
Evidence for Concert in the Thermal Unimolecular
Vinylcyclopropane to Cyclopentene Sigmatropic 1,3-Shift
Joseph J. Gajewski,† Leif P. Olson,† and M. Robert Willcott, III*,‡
Contribution from the Departments of Chemistry, Indiana UniVersity, Bloomington,
Indiana 47405, and Vanderbilt UniVersity, NashVille, Tennessee 37235
ReceiVed May 15, 1995. ReVised Manuscript ReceiVed October 23, 1995X
Abstract: Gas phase pyrolysis of cis-2,3-dideuterio-trans-(1′-tert-butyl-2′-(Z)-deuteriovinyl)cyclopropane at 290 °C
gives trans,trans-3,4,5-trideuterio-1-tert-butylcyclopentene as the major 1,3-shift product with greater than 90%
stereospecificity in an orbital symmetry “allowed” suprafacial-inversion sense after correction for the geometric
isomerization of starting material and the other materials present. The isotopic substitution at the critical sites of
rearrangement eliminates steric or electronic influences of substituents on a biradical pathway as a source of the
suprafacial-inversion stereochemistry observed with more highly substituted derivatives. The stereochemical results
coupled with a normal deuterium kinetic isotope effect (kH/2kD ) 1.14 at 311.6 °C) at the exo-methylene carbon are
best interpreted in terms of a concerted pathway for rearrangement. A less likely alternative is a stereospecific
disrotatory ring opening to a biradical followed by rate-determining closure to a five-membered ring. Accompanying
the rearrangement is geometric isomerization of starting material resulting from C-1-C-2 bond fission which favors
either the single methylene rotation or a double rotation by a factor of 10 over single rotation at the vinyl-bearing
carbon.
Vinylcyclopropane undergoes a gas phase, first-order allylic
rearrangement to cyclopentene1a with log k (s-1) ) 13.5 -
49700/2.3RT.1b Numerous studies reveal that geometric isomer-
ization of starting material occurs roughly 20 times faster than
allylic rearrangement.2 The mechanism of the geometric
isomerization is generally attributed to reversible formation of
a biradical in which rotation around non-allylic bonds occurs
faster than ring closure. Because of the orbital overlap in the
allylic moiety, it is generally recognized that both transoid and
cisoid species are possible and the only fate of the former is
reclosure to a vinylcyclopropane. The cisoid species may also
be responsible for geometric isomerization, and it may be
involved in the allylic rearrangement (Scheme 1). Alternatively,
a concerted sigmatropic 1,3-shift might be envisioned for the
ring expansion.
While the mechanism of the rearrangement is the focus of
this paper, it should be recognized that more than two decades
ago Willcott provided evidence for geometric isomerization
being a near-random process kinetically. That is, there was a
2:1 ratio of trans to cis-syn-2,3-dideuteriovinylcyclopropane
formed from cis-anti-2,3-dideuteriovinylcyclopropane even at
short reaction times.2c These data can be interpreted in terms
of a randomly rotating biradical (k1 ) k2 ) k12) or exclusive
single rotation around each stereocenter, with rotation around
the methylene being twice as fast as rotation around the vinyl-
bearing carbon (k2 ) 2k1), or combinations of these possibilities
(Scheme 2). Exclusive double rotation between carbons 1 and
2 (k12) is not permitted by the data. It is important to recognize
that C-2-C-3 bond fission is disfavored by ca. 12 kcal/mol
† Indiana University.
‡ Vanderbilt University.
X Abstract published in AdVance ACS Abstracts, January 1, 1996.
(1) (a) Overberger, C. G.; Borchert, A. E. J. Am. Chem. Soc. 1960, 82,
4896. Neureiter, N. P. J. Org. Chem. 1959, 24, 2044. (b) Wellington, C.
A. J. Phys. Chem. 1962, 66, 1761.
(2) (a) Marshall, D. C.; Frey, H. M. J. Chem. Soc. 1962, 3981. (b)
Willcott, M. R., III; Cargle, V. H. J. Am. Chem. Soc. 1967, 89, 273. (c)
Willcott, M. R., III; Cargle, V. H. J. Am. Chem. Soc. 1969, 91, 4310. The
vinyl group of the compounds described in Scheme 2 was also substituted
with deuterium which was omitted for purposes of clarity. (d) Doering, W.
von E.; Sachdev, K. J. Am. Chem. Soc. 1974, 96, 1168. (e) For a review
see: Gajewski, J. J. Hydrocarbon Thermal Isomerizations; Academic
Press: New York, 1980; pp 81-87.
0002-7863/96/1518-0299$12.00/0 © 1996 American Chemical Society

300 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Gajewski et al.
Scheme 1
Scheme 4
Scheme 5
Scheme 2
Scheme 3
Figure 1. 1H NMR spectrum (500 MHz) of the methylene protons
Ha-Hf of 1-tert-butyl-6-oxabicyclo[3.1.0]hexane in benzene solvent.
relative to C-1-C-2 bond fission due to the development of
allylic radical resonance energy by the latter cleavage.
Previous efforts to determine the stereochemistry of the
vinylcyclopropane to cyclopentene rearrangement3,4 demon-
strated dominant (ca. 55-70%) suprafacial-inversion stereo-
chemistry in this sigmatropic 1,3-shift, but all of these except
for those described in two recent papers4 have utilized methyl
substitution at the migrating carbon to reveal the pathway
(Scheme 3).
Results
Preparation and Pyrolysis of (1′-tert-Butylvinyl)cyclopro-
pane, 1H. All-protio vinylcyclopropane, 1H, was synthesized
from cyclopropylcarbinol by oxidation to cyclopropanecarbox-
aldehyde, Wittig-like dibromomethylenation, carbenoid rear-
rangement to ethynylcyclopropane by treatment with magnesium
in tetrahydrofuran, and then reaction with tert-butylmagnesium
cuprate with aqueous workup (Scheme 4).
There is concern that steric effects in the trans-2-methyl
compounds (the only geometric isomer studied)3 may be
responsible for the stereoselectivity by forcing outward rotation
of the trans-methyl group as the C-1-C-2 bond breaks which
might present the back side of the migrating carbon to the end
of the allylic moiety for a least motion closure. This prompted
examination of the pyrolysis of cis-2,3-dideuterio-trans-(1′-tert-
butyl-2′-(Z)-deuteriovinyl)cyclopropane, 1D3. The choice of
a tert-butyl-substituted material has been justified previously.3b
It reduces the ratio of geometric isomerization of starting
vinylcyclopropane to allylic rearrangement presumably by
destabilizing transition states leading to a transoid allyl moiety
which cannot give the allylic rearrangement but can only be
responsible for geometric isomerization in the starting material.
Our early work which was communicated previously4a revealed
that roughly 70% suprafacial-inversion stereospecificity was
observed, but more extensive work detailed below suggests still
higher stereospecificity. Since our initial paper, Baldwin has
communicated work on the parent vinylcyclopropane system
which indicates lower stereospecificity.4b
The synthetic sequence gives a 55:42:3 mixture of 1H, (trans-
2′-tert-butylvinyl)cyclopropane, 3, and apparently its cis isomer,
4, respectively. Separation was accomplished by preparative
GC on dibutyl tetrachlorophthalate after thoroughly cleaning
the injector and detector and alternating injections of the olefin
mixture and diisopropylamine. The (cis-2′-tert-butylvinyl)-
cyclopropane isomer, 4, could not be separated from the desired
(1′-tert-butylvinyl)cyclopropane, 1H, but subsequent experi-
ments revealed that it did not undergo the 1,3-shift under the
reaction conditions; however, it did undergo geometric isomer-
ization, a point that will be discussed below. The extent of
contamination of 1 by 4 was typically 7-8% in the experiments
described below. Fortunately, the presence of 4 provides an
internal reference for both the 1,3-shift of 1 and its geometric
isomerization.
As expected by analogy to all other vinylcyclopropane
thermolyses, pyrolysis of 1H at 290 °C gives 1-tert-butylcy-
clopentene, 2, which has been synthesized previously (Scheme
5).5 For purposes of examining the deuterium distribution in 2
when derived from deuterated 1, 2 was converted to the epoxide
where it was found that all ring protons were separately visible
(3) (a) Andrews, G. D.; Baldwin, J. E. J. Am. Chem. Soc. 1976, 98, 6705.
(b) Gajewski, J. J.; Squicciarini, M. P. J. Am. Chem. Soc. 1989, 111, 6717.
(c) Baldwin, J. E.; Gahtlia, N. D. J. Am. Chem. Soc. 1991, 113, 6273. (d)
Baldwin, J. E.; Bonacorsi, J., Jr. J. Am. Chem. Soc. 1993, 115, 10651.
(4) (a) Gajewski, J. J.; Olson, L. P. J. Am. Chem. Soc. 1991, 113, 7432.
This is a preliminary paper describing some of the results in the current
paper. (b) Baldwin, J. E.; Villarica, K. A.; Freedberg, D. I.; Anet, F. A. L.
J. Am. Chem. Soc. 1994, 116, 10845.
1
in its H NMR spectrum when taken in benzene-d6. Further,
all the coupling constants were different between the ring
protons. The proton assignments are consistent with previous
(5) (a) Brown, H. C.; Lynch, G. J.; Hammar, W. J.; Liu, L. C. J. Org.
Chem. 1979, 44, 1910. (b) Buechler, J. D. ibid. 1973, 38, 904.

Vinylcyclopropane to Cyclopentene Sigmatropic 1,3-Shift
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 301
Scheme 6
Figure 2. Proton magnetic resonance spectrum from 0.3 to 0.7 ppm
of the cyclopropane mixture from 290 °C pyrolysis for 5 h of 1D3
(85%) + 5D3 (2%) + 6 (10.5%) + 1D2 (2.4%).
work,6ab molecular mechanics calculations,6c and coupling
constants expected on the basis of a boat conformation7 (Figure
1).
Preparation of cis-2,3-Dideuterio-trans-(1′-tert-butyl-2′-(Z)-
deuteriovinyl)cyclopropane. Stereospecifically labeled tri-
deuteriovinylcyclopropane 1D3 was prepared from bis(trime-
thylsilyl)acetylene and methyl diazoacetate followed by hydrolysis
in D2O, lithium aluminum hydride reduction of the dideuterio-
cyclopropene ester at -78 °C, gradual warming to 0 °C, and
quenching with water which gives (cis-2,3-dideuterio-trans-
cyclopropyl)carbinol.8 Treatment of this material as described
above for the protio compound with the exception of a D2O
workup of the cuprate addition furnished 1D3 (Scheme 6). After
separation to remove 3D3, the 300 MHz 1H NMR spectrum of
the product revealed that 85% was 1D3, 2% was cis-2,3-
dideuterio-cis-(1′-tert-butyl-2′-(Z)-deuteriovinyl)cyclopro-
pane, 5D3, 10.5% was trans-2-deuterio-trans-(1′-tert-butyl-2′-
(Z)-deuteriovinyl)cyclopropane, 6, and 2.4% was cis-2,3-
dideuterio-trans-(1′-tert-butylvinyl)cyclopropane, 1D2; in addition,
roughly 7% of the sample appeared to be cis-2,3-dideuterio-1-
(1′-deuterio-cis-2′-tert-butylvinyl)cyclopropane, 4D3, from the
fact that the proton NMR spectrum had an additional doublet
at 0.27 ppm, a singlet at 1.18 ppm, and a singlet at 5.25 in the
appropriate ratio.
Pyrolysis of cis-2,3-Dideuterio-trans-(1′-tert-butyl-2′-(Z)-
deuteriovinyl)cyclopropane. Geometric Isomerization. Vi-
nylcyclopropane 1D3 (as the mixture described above) under-
goes geometric isomerization at 290.0 °C as evidenced by an
increase in the relative area of the 0.65 ppm region. Using the
integrated ratios of the 0.65 and 0.39 ppm region and assuming
a 1:1 ratio of protons at equilibrium, the first-order, one-way
rate constant kgeo ) 1.43(0.04) × 10-5 s-1 could be determined
from data at t ) 0, 5.000 h, 8.000 h, and infinity. However,
analysis of this geometric isomerization is more complicated
(see below). Further, capillary GC analyses allowed determi-
nation of the first-order, irreversible rate constant for the 1,3-
shift: k[1,3] ) 2.34(0.17) × 10-6 s-1 in the same pyrolyses. Thus,
Figure 3. 1H{2H} NMR spectra (500 MHz) of the methylene protons
of the epoxide of 4D3 obtained from a 5.0 h pyrolysis of 1D3 at 290
°C.
geometric isomerization would appear to be roughly 6 times
faster than the 1,3-shift in this reaction; however, a kinetic
modeling study revealed (see the Discussion) that the geometric
isomerization is in fact 10 times faster than the 1,3-shift.
Examination of the 500 MHz deuterium-decoupled cyclo-
propane C-2 and C-3 ring proton resonance signals from 0.3 to
0.7 ppm of material recovered from the 5.00 h pyrolysis at 290
°C (Figure 2) reveals that conversion of 1D3 to a mixture of
geometric isomers proceeds to almost the extent of 50%; i.e.,
the 0.65 ppm proton area is ca. 1/3 that of the 0.39 ppm proton
area. Further, the 0.65 ppm signals indicate that the geometric
isomerization gives a mixture of diastereomers which favors
the trans-2,3-dideuterio isomer over the cis-2,3-dideuterio-cis-
1-vinyl material by a factor of 4.
Thus, the pattern2c consists of a doublet of doublets (8 and 6
Hz) for the one proton of the trans-2,3-diprotio material and
an 8 Hz doublet for the two protons of the cis-2,3-dideuterio-
cis-1-vinyl compound. The fact that all six of these peaks are
nearly equal in intensity requires a 4:1 ratio of the former to
the latter. Interestingly, from the spectrum of the cyclopropanes
recovered from the pyrolysis for 5 h, it is clear that complete
equilibration of the ring deuterium isomers of (cis-2′-tert-
butylvinyl)cyclopropane, 4D3, occurred since the signals at 0.27
ppm (H’s cis to vinyl) and 0.71 ppm (H’s trans to vinyl) are
nearly equal in intensity and consist of a random mixture of
the two diastereomers; i.e., the relative areas of the major doublet
and the doublet of doublets are 1:1 in each of the proton signals.
Stereochemistry of the Vinylcyclopropane Rearrange-
ment. In order to examine the stereochemistry of the rear-
rangement, the cyclopentene product from the 5.0 h pyrolysis
(6) (a) Lafferty, W. J. J. Mol. Spectrosc. 1970, 36, 84. (b) Steyn, R.;
Sable, H. Z. Tetrahedron 1971, 27, 4429. (c) Gajewski, J. J.; Gilbert, K.
E.; McKelvey, J. In AdVances in Molecular Modeling; Liotta, D., Ed.; JAI
Press: Greenwich, CT, 1990; Vol. 2. MMX is a force field similar to MM2
with updates from the current literature; it uses charge-charge rather than
dipole-dipole interactions for the electrostatics. PCMODEL is a program
containing the MMX force field and is distributed by Serena Software, P.O.
Box 3072, Bloomington, IN 47402 (K. E. Gilbert, President).
(7) Haasnoot, C. A. G.; deLeeuw, F. A. A. M.; Altona, C. Tetrahedron
1980, 36, 2783.
(8) Vidal, M.; Arnaud, P. Bull. Soc. Chim. Fr. 1972, 678. Domnin, I. N.
Dumon, K.; Vincens, M.; Vidal, M. IzV. Akad. Nauk SSSR 1985, 7, 1593,
1598. We thank Professor John E. Baldwin for suggesting these procedures.

302 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Gajewski et al.
deuterium kinetic isotope effect at the terminal methylene
position of trans-2-methyl- and trans-2-methyl-(1′-tert-butylvi-
nyl)cyclopropane in their rearrangements to the respective
cyclopentenes (kH/2kD ) 1.17 and 1.13, respectively, at 280 °C).
Further, Chickos and later Baldwin found a correspondingly
large DKIE at the terminal methylene of vinylcyclopropane itself
in its sigmatropic 1,3-shift.9 In the current system, the DKIE
at the terminal methylene of (1′-tert-butylvinyl)cyclopropane
1XD2 was found to be 1.14(0.02) for two deuteriums at 311.6
°C.
Discussion
Geometric Isomerization of 1D3. In order to determine the
stereospecificity of the vinylcyclopropane rearrangement of 1D3,
the direction and extent of loss of stereointegrity of starting
material must be known. To review the facts, starting 1D3 is
contaminated by ca. 2 mol % of the corresponding cis-2,3-
dideuterio-trans-1-vinyl material 5D3 and by ca. 10 mol % of
what appears to be trans-2-deuterio-1-(1′-tert-butyl-2′-(Z)-
deuteriovinyl)cyclopropane, 6. The latter determination follows
from the observation of a 5% proton signal at 0.66 ppm for
which there should be a corresponding 10% proton signal
enmeshed in the 0.40 ppm proton signal for 1D3. Table 1 shows
a breakdown of the contribution of each material at various
extents of pyrolysis. To model the relative proton signal
intensities of both the 0.65 and 0.39 ppm signals and the relative
amounts of 5D3 and the trans-2,3-dideuterio-1-(2′-deuteriovinyl)
isomer 7D3 at the indicated pyrolysis times, rate constants for
the three-way interconversion of the D3 isomers 1, 5, and 7
must be as indicated in Scheme 7 (in units of 10-5 s-1; all (20%
assuming that each undergoes the 1,3-shift equally fast and all
by unimolecular processes).
The analysis of the recovered vinylcyclopropanes from
pyrolysis of 1D3 makes it clear that single rotation of the vinyl-
bearing carbon is distinctly slower, by a factor of 10, than either
single rotation at C-2 or C-3 or double rotation presumably after
breaking the C-1-C-2 or C-1-C-3 bond. This stands in
contrast to the observations in the parent case where all rotations
appear equally likely.2c The size of the tert-butyl substituent
on the vinyl appears responsible for the difference. Unfortu-
nately it is not possible to dissect the rate constants k1f7 and
k5f7 into single and double rotation components, but the fact
that these isomerizations occur with the rate constants indicated
provides an important correction to the extent of stereospecificity
observed in the 1,3-shift.
Stereochemistry of the 1,3-Shift in 1D3. Regardless of the
exact value of the extent of suprafacial-inversion stereochemistry
in the pyrolysis of 1D3, it is clear that the 55-70% suprafacial-
inversion stereochemistry observed with 2-trans-alkyl-substi-
tuted vinylcyclopropanes is not a result of steric effects.
Furthermore, the stereospecificity is higher with the minimally
perturbed system 1D3. Simulation of the kinetic scheme for
interconversion of the trideuterio isomers and irreversible
formation of a different cyclopentene isomer from each using
the rate constants described above and that determined directly
from GC for the 1,3-shift revealed that only 79% of the product
in the 5.0 h pyrolysis is formed from 1D3 while 16.4% of the
product is formed from 7D3 and 4.6% of the product should
arise from 5D3. Since suprafacial-inversion, si, is the major
stereopathway, the pyrolysis of 7D3 should give equal parts of
the suprafacial-retention (2sr) and antarafacial-retention (2ar)
products expected from 1D3 (Scheme 8); further, the suprafacial-
1
Figure 4. Simulation of the 500 MHz H{2H} NMR spectra of each
of the pairs of diastereomers of the epoxide of 4D3 derived by the si,
sr, ar, and ai modes of the 1,3-shift from 1D3 using the chemical shifts
and coupling constants which reproduce the spectrum of Figure 1.
of 1D3 at 290 °C was epoxidized and then purified by GC.
The unlocked deuterium-decoupled 500 MHz 1H NMR spectrum
above 1.80 ppm consisted of three singlets (full width at half-
height 4 Hz) (Figure 3) and of two doublets and a triplet (J )
9 Hz) which correspond to those expected for the suprafacial-
inversion pathway.
Using the assigned chemical shifts and coupling constants
for the pair of racemic epoxides which would result from each
of the four stereochemical pathways si, sr, ar, and ai for reaction
of achiral trideuterio compound 1D3, the four simulated spectra
of Figure 4 are obtained. The presence of the three singlets in
the actual spectrum is uniquely consistent with the dominant
presence of the deuterium isomer expected via the suprafacial-
inversion reaction.
For comparison, a sample of 1D3 was pyrolyzed for 5.5 half-
lives of geometric isomerization-59% conversion to 2, and the
1H NMR spectrum is that expected for randomized stereochem-
istry (Figure 5).
The presence of (nonzero) cis, Vicinal couplings in the mixture
causes the superimposition of 8 Hz doublets at both the endo-
C-2 and endo-C-4 chemical shifts and an 8 Hz doublet and triplet
at the exo-C-3 chemical shifts upon the singlets observed at
low pyrolysis time. There is a 0.007 ppm upfield isotope shift
of the exo-C-2 and exo-C-4 singlets which is probably due to a
gauche interaction with the endo-C-3 deuterium.
It is not possible to calculate the stereospecificity in the 1,3-
shift to high accuracy, but it appears to be greater than 70%
before correction for incomplete deuteration and loss of stere-
ointegrity of starting material. Indeed, the analysis in the
Discussion suggests nearly complete stereospecificity in the
suprafacial-inversion sense.
Secondary Deuterium Kinetic Isotope Effects (DKIE) at
the Terminal Methylene of 1. The stereospecificity of the 1,3-
shift prompted examination of the secondary deuterium kinetic
isotope effect at the terminal methylene position. Previous work
by Squicciarini3b revealed a substantial normal secondary
(9) Chickos, J. Abstracts of Papers, 187th National Meeting of the
American Chemical Society, St. Louis, MO, April 8-13, 1984; American
Chemical Society: Washington, DC, 1984; ORGN 228. Baldwin, J. E.
Villarica, K. A. Tetrahedron Lett. 1994, 7905.

Vinylcyclopropane to Cyclopentene Sigmatropic 1,3-Shift
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 303
Table 1. Analysis of the Upfield Proton Region in Recovered Vinylcyclopropane from Pyrolysis of 1D3 at 290.0 °C
mole fractions
5D3 (2)a
mole fractions
7D3 (1)a
rel area
(δ 0.65)
rel area
(δ 0.39)
time (h)
7D3 (1)a
6D2 (1)a
1D3 (2)a
6D2 (2)a
0.0b
0
0
0.0205
0.105
0.0525
0.105
0.065g
0.105
0.07g
0.07c
0.073
0.25c
0.26
0.875
0.875
0.557e
0.557
0.483e
0.483
0
0
0.105
0.105
0.105
0.085g
0.105
0.08g
0.93c
0.98
1H NMR:d
5.00
0.205
0.065e
0.065f
0.095e
0.095f
+
+
+
)
)
)
+
+
+
+
+
+
)
)
)
0.265e
0.132f
0.334e
0.167f
0.265e
0.1325
0.334e
0.167
0.75c
0.774
0.69c
0.73
1H NMR:d
8.00
+
+
0.31c
0.332
1H NMR:d
a The number in parentheses indicates the number of hydrogens in the signal. b At t ) 0, the ratio 1:7:5:6 is 0.875:0:0.02:0.105, and 1 is a 85:2.4
1
1
1
mixture of 1D3 and 1D2. c Relative H NMR area. d Contribution of mole fractions to the H NMR signal with summation below the relative H
NMR signal. e Relative mole fractions calculated using k1f7 ) k5f7 ) 2.4 × 10-5 s-1, k7f1 ) k7f5 ) 1.2 × 10-5 s-1, and k1f5 ) k5f1 ) 2.4 × 10-6
s-1
.
f The ratio 7:5 from the 0.65 ppm signal is 4 after 5.00 h of pyrolysis and 3.8 after 8.00 h of pyrolysis. g Estimated with the rate constants of
footnote e.
Scheme 8
1
Figure 5. Simulated (top) and experimental (bottom) H{2H} NMR
spectra (500 MHz) of the methylene protons of the epoxide of 4D3
obtained from a 59% conversion pyrolysis of 1D3 at 290 °C. The
simulated spectrum is of a 1:1:1:1 mixture of si, sr, ar, and ai product-
derived epoxides, respectively.
Scheme 7
Figure 6. Simulation of the spectrum of Figure 3 by assuming a 71:
11.3:6.2:11.5 ratio of si, sr, ar, and ai products.
by assuming only 71% suprafacial-inversion product which
implies about 90% stereospecificity.
Provided in Figure 7 is the upfield epoxide region from the
cyclopentene product from the 8 h pyrolysis of 1D3. Also
included is a simulated spectrum summing the contributions of
the four isomers assuming a 60:13:13:13 ratio of si, sr, ar, and
ai product-derived epoxides. Again the fit is remarkably good
which further confirms high stereospecificity in the reaction.
inversion product from 5D3 is the antarafacial-inversion product
(2ai) expected from 1D3 (Scheme 8).
The high stereospecificity in the rearrangement suggested by
the simulations is remarkable even before correction for the 10%
cyclopentene product derived from the dideuterio contaminant
6. If the si pathway were utilized by this material and there
was but a small kinetic isotope effect, which is expected at these
high temperatures, 6 should give equal amounts of cyclopentenes
8 and 9, which would give C-4 diprotio and C-3 diprotio
Addition of the four cyclopentene epoxide spectra of Figure
4 in the ratio described above assuming only suprafacial-
inversion stereochemistry for all vinylcyclopropane diastereo-
mers gives a spectrum which generally resembles that from the
5.0 h pyrolysis (Figure 6). A somewhat better fit is provided

304 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Gajewski et al.
prafacial-inversion stereochemistry is consistent with concert
in the rearrangement with a competing, entropically more
favorable, biradical pathway(s) which also results in loss of
stereochemistry of the starting material. The rearrangement
stereospecificity is also consistent with disrotatory ring opening
with inward rotation of the vinyl group to produce a biradical
which undergoes least motion closure faster than bond rotation.
The other biradical(s) resulting from outward rotation of the
vinyl group or random rotations about both bonds might be
responsible for the loss of stereointegrity of the starting material.
The stereospecific formation of a biradical and subsequent rapid
closure may not be unlike pathways attributed to dynamical
effects by Carpenter10 although the origin of the disrotatory ring
opening stereomode is not obvious. This suggestion runs
contrary to the expectation of conrotatory opening based on the
through-bond orbital interaction in a π-cyclopropane transition
state or intermediate first suggested by Hoffmann.11
Kinetic Isotope Effects. The observation of a substantial
normal deuterium kinetic isotope effect at the terminal meth-
ylene carbon in the rearrangement of vinylcyclopropane itself,9
of 2-methylvinylcyclopropanes,3b and of (1′-tert-butyl-2′,2′-
dideuteriovinyl)cyclopropane, 1XD2, which is most reasonably
attributed to a rotational isotope effect,12 is consistent with
concert. Alternatively, the DKIE may result from formation
of a biradical like that described above whose ring closure to
cyclopentene is rate determining. Rate-determining formation
of a biradical, even the one which must be generated stereospe-
cifically as described above, should result in a very small normal
DKIE at the terminal methylene (ca. 1.03) at the temperatures
of the reaction.13 In the concerted reaction, bond formation to
the terminal methylene cannot proceed to a significant extent,
or a normal DKIE should have been observed. Perhaps it is
not unexpected that extensive bond formation does not occur
in this reaction given that the activation energy is comparable
to the bond dissociation energy of the ring bond being ruptured.
Comparison to Previous Results. The current results
represent the first determination, albeit incomplete, of the
stereochemistry of the vinylcyclopropane rearrangement without
substitution at the migrating carbon and the migration terminus.
And, remarkably, it demonstrates that the orbital symmetry
controlled suprafacial-inversion pathway is utilized in a steric
unbiased case. In more recent work, Baldwin showed that the
parent vinylcyclopropane system with deuteriums at C-2, C-3,
and C-â also reveals a preference for the suprafacial-inversion
pathway; however, the preference is only 40% total relative to
all other pathways. Since the temperature is only 10 °C higher,
it is not clear what the difference is between the parent and the
tert-butyl case studied here.
1
Figure 7. Simulated (top) and experimental (bottom) H{2H} NMR
spectra (500 MHz) of the methylene protons of the epoxide of 4D3
obtained from an 8.0 h pyrolysis of 1D3 at 290 °C. The simulated
spectrum is of a 60:13:13:13 mixture of si, sr, ar, and ai product-derived
epoxides, respectively.
Scheme 9
cyclopentene epoxides, respectively, each one of which is a pair
of diastereomers (Scheme 9).
The C-4 diprotio epoxide results from C-1-C-2 fission
(cleavage to the methylene bearing deuterium) in 6. The C-4
diprotio epoxide should give a superimposed doublet (Jgem
)
13 Hz) and doublet of doublets (Jgem ) 13 Hz, JVic ) 9.7 Hz)
roughly 5 Hz downfield (due to no isotope effect) from the 1.75
ppm resonance and a similar pattern roughly 5 Hz downfield
from the 1.15 ppm resonance, each contributing roughly 10%
of the size of the peaks predicted for the si product alone; also
resulting from C-1-C-2 fission is an epoxide which should give
a 7.6 Hz doublet under the 1.67 ppm resonance signal, and its
intensity should be 5% of the main peak. The C-3 diprotio
epoxide should give a 7.8 Hz doublet under the 1.75 and 1.67
ppm resonances, contributing 5% of the main peak; in addition
a doublet of doublets (J ) 9.7 and 9.2 Hz), probably a triplet,
should be superimposed on the 1.15 ppm resonance, and it
should contribute 5% of the main peak. These additional
resonances have the effect of making the simulated spectra more
like that observed rather than less by broadening the signals at
the bottom. Exact quantification is ultimately foiled by the
signal to noise ratio in the spectra. However, 90% stereospeci-
ficity in a suprafacial-inversion sense for the vinylcyclopropane
rearrangement would appear to be a conservative estimate.
Mechanistic Possibilities. Demonstration of dominant su-
Conclusion. A stereospecific 1,3-shift of carbon is observed
in a thermal vinylcyclopropane rearrangement which has
deuterium as a stereochemical marker at both the migrating
carbon and migration terminus. This structural isomerization
occurs slower than geometric isomerization which occurs mostly
(10) Newman-Evans, R. H.; Simon, R. J.; Carpenter, B. K. J. Org. Chem.
1990, 55, 695. Carpenter, B. K. Acc. Chem. Res. 1992, 25, 520. Some
concern might be expressed about the experimental basis of the hypothesized
dynamical effect in the case of the rearrangement of the bicyclo[2.1.1]-
hexene system. While no temperature dependence of the stereospecificity
is observed in phenyl-substituted cases, implying that dynamical effects
are involved, the parent system shows normal temperature dependent
stereospecificity.
(11) Hoffmann, R. J. Am. Chem. Soc. 1968, 90, 1475.
(12) Dai, S. H.; Dolbier, J. J. Am. Chem. Soc. 1970, 92, 1774; 1972, 94,
3946. Crawford, R. J.; Chang, M. H. Tetrahedron 1982, 38, 837. Gajewski,
J. J. Benner, C. W.; Stahly, B. N.; Hall, R. F.; Sato, R. I. Tetrahedron
1982, 38 853.
(13) Gajewski, J. J.; Olson, L. P.; Tupper, K. J. J. Am. Chem. Soc. 1993,
115, 4548.

Vinylcyclopropane to Cyclopentene Sigmatropic 1,3-Shift
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 305
was taken during the addition so that the temperature did not rise above
5 °C. The resulting mixture was poured into a solution made by
dissolving 0.5 g of sodium cyanide in 50 mL of saturated aqueous
ammonium chloride. After 15 min of stirring, the mixture was extracted
with pentane (100 mL) and the organic layer was washed with water
(2 × 200 mL). The aqueous phases were extracted once with pentane
(100 mL), and the organic phases were combined, dried over
magnesium sulfate, and filtered. Slow distillation of the pentane
through an 18 cm Vigreux column provided a residue of 0.45 g (82%
yield) of three products in a 55:3:42 ratio. Preparative GC (10 DBTCP
on Chromosorb P, 10 ft × 1/4 in. column) allowed separation of the
latter two from the first peak to give 150 µL (0.8 mmol, 10% yield
from dibromide) of 1-(1′-tert-butylvinyl)cyclopropane. 1H NMR (300
MHz, CDCl3): δ 4.66 (d, J ) 1 Hz, 1H); 4.45 (d, J ) 1 Hz, 1H); 1.36
(m, 1H); 1.14 (s, 9H); 0.67 (m, 2H); 0.42 (m, 2H). The product is
contaminated with a small amount of what appears to be 1-(cis-2′-tert-
butylvinyl)cyclopropane since the first GC peak is almost certainly
1-(trans-2′-tert-butylvinyl)cyclopropane from its 1H NMR (300 MHz,
CDCl3): δ 5.55 (d, J ) 15 Hz, 1H); 4.83 (d,d, J ) 15, 8 Hz, 1H); 1.30
(m, 1H); 1.0 (s, 9H); 0.66 (m, 2H); 0.3 (m, 2H). This latter material
was not further characterized. 1H-1H NOE difference results for 1H:
Irradiation of 1.14 ppm (tert-butyl protons) resulted in a 17%
enhancement at 4.66 ppm and 0.5% enhancement at 4.45 ppm.
Irradiation of the multiplet at 1.36 ppm (cyclopropyl methine proton)
resulted in a 7% enhancement of the 0.67 ppm resonance and 0.1%
enhancement of the 0.42 ppm resonance. 13C{1H} NMR (125 MHz,
CDCl3): δ 159.7; 102.0; 31.2; 29.7; 12.9; 7.2. FTIR (CDCl3): 3050,
2950, 1650, 1380 cm-1. MS (EI, m/e): 124.
by methylene single rotation or by C-1-C-2 double rotation.
The stereosense of the 1,3-shift is that predicted for a concerted
reaction by the Woodward-Hoffmann rules. A large (at these
high temperatures) secondary deuterium kinetic isotope effect
at the exo-methylene group suggests its rotation in the rate-
determining transition state for the 1,3-shift. The overall
reaction can best be characterized as a competition between an
entropically demanding concerted 1,3-shift and a geometric
isomerization Via biradicals or concerted double rotation. A
nonconcerted path would require stereospecific disrotatory
opening with inward rotation of the vinyl group followed by
rapid bond formation, events which currently have no precedent.
However, it may be that the two radical sites may be coupled
in an orbital symmetry conserved sense and may be more stable
than the transition state for ring opening and that for ring closure.
Whether or not calculations will reveal this is not yet clear.
Experimental Section
1H and 13C NMR spectra were recorded either on a Bruker AM-500
or a Varian XL-300 spectrometer. 2H NMR spectra were obtained on
a Nicolet NT-360 spectrometer operating at 55 MHz. Deuterium-
1
decoupled H NMR experiments were performed using the AM-500
both at Indiana and at Vanderbilt. 1H-1H NOE difference spectroscopy
experiments were performed on degassed samples in deuteriochloroform
using the XL-300 instrument. IR spectra were obtained on a Perkin-
Elmer 298 or a Mattson Instruments Galaxy 4020 Fourier transform
spectrophotometer. Mass spectra were obtained on a Kratos MS-80
spectrometer or on a Hewlett-Packard 5980 GC/MS system with a
Model 5971 mass sensitive detector. Preparative GC was performed
using a Varian Model 90-P or a Varian 4700 gas chromatograph with
thermal conductivity detection. Capillary gas chromatography was
performed using a Varian 3700 gas chromatograph equipped with flame
ionization detection; the columns used were coated with DB-5 (50 m
× 0.25 mm) or with Supelcowax-10 (60 m × 0.25 mm). Peak area
ratios were determined by Varian 4240 or Hewlett-Packard 3390A
reporting integrators; the FID sensitivity factors were assumed to be
unity for isomers. Dry tetrahydrofuran was obtained by distillation
under nitrogen from sodium-benzophenone ketyl. All reactions were
carried out under an inert atmosphere of nitrogen.
1-(1′-tert-Butyl-2′-(Z)-deuteriovinyl)-trans-2,trans-3-dideuteriocy-
clopropane, 1D3. Trideuterio(tert-butylvinyl)cyclopropane 1D3 was
prepared from methyl bis(trimethylsilyl)cycloprop-2-enecarboxylate14,15
(8.7 g, 36 mmol) which was added dropwise over 5 min to 40 mL of
10% potassium deuteroxide in methanol-O-d1 at 0 °C with stirring.
The mixture turned yellow upon addition. After 4 h of stirring at room
temperature, the mixture was cooled to 0 °C, and 10% deuterium
chloride in deuterium oxide was added until the mixture was neutral
to pH paper. At pH 7 the yellow color of the mixture changed to dark
green. Methanol was removed by careful rotary evaporation. The
resulting aqueous mixture was extracted with diethyl ether (3 × 50
mL portions). The organic phase was washed with water (20 mL),
dried with magnesium sulfate, and filtered. The solution was added
dropwise to a stirred solution of lithium aluminum hydride in diethyl
ether (1 M, 190 mL, 190 mmol) at -78 °C. The mixture was warmed
to 0 °C over 3/4 h, and quenched with water (5 mL), 15% aqueous
sodium hydroxide solution (5 mL), and then water (15 mL) again.
Filtration of the precipitated salts, drying of the filtrate with magnesium
sulfate, filtration of the drying agent, and distillation of the solvent
through an 18 cm Vigreux column gave a residue which was subjected
to short-path distillation (100-140 °C) to give 0.6 g (25% yield) of
cis-2,3-dideuteriocyclopropane-trans-1-methanol. 1H NMR (300 MHz,
CDCl3): δ 3.43 (d, J ) 8 Hz, 2H); 2.35 (s, 1H); 1.09 (m, 1H), 0.18 (d,
J ) 5 Hz, 2H).
Collected samples of this alcohol were oxidized to the corresponding
aldehyde under Swern conditions: To a mechanically-stirred solution
of oxalyl chloride (2.8 g, 22 mmol) in methylene choride (20 mL) at
-60 °C under a nitrogen atmosphere was added dropwise anhydrous
dimethyl sulfoxide (2.4 g, 30 mmol) in methylene chloride (5 mL).
After stirring for 10 min, cis-2,3-dideuteriocyclopropane-trans-1-
carbinol (2.0 g, 27 mmol) in methylene chloride was added over 5
min, and stirring was continued for another 10 min. Triethylamine
(12 mL) was added dropwise, and then the cooling bath was removed
and 1 M aqueous hydrochloric acid (100 mL) was added. The aqueous
phase was extracted with methylene chloride (30 mL), and the combined
organic phases were washed with water (40 mL), 10% aqueous sodium
carbonate solution (15 mL), water (10 mL), and saturated brine (10
1-(1′-tert-Butylvinyl)cyclopropane, 1H. To a stirred solution of
triphenylphosphine (16.8 g, 64 mmol) in 35 mL of methylene chloride
at 0 °C was added 10.6 g (32 mmol) of carbon tetrabromide in 20 mL
of methylene chloride. To the resulting yellow-brown suspension was
added over a period of 5 min 1.1 g (16 mmol) of cyclopropanecar-
boxaldehyde in 10 mL of methylene chloride. After 5 min of stirring,
the mixture was poured into 200 mL of pentane. Filtration, followed
by reworking the precipitate by two more cycles of methylene chloride
dissolution and pentane precipitation, gave a filtrate which was
concentrated by rotary evaporation and distilled bulb to bulb (90 °C, 1
Torr) to give 2.3 g of (dibromovinyl)cyclopropane (11 mmol, 70%
yield). 1H NMR (300 MHz, CDCl3): δ 5.78 (d, J ) 1 Hz); 1.64 (m,
1H); 0.87 (m, 2H); 0.55 (m, 2H). FTIR (CDCl3): 3050, 2950, 1600
cm-1. MS (EI, m/e): 226.
A 1.1 g (4.4 mmol) sample of (dibromovinyl)cyclopropane in 10
mL of tetrahydrofuran was added dropwise to a refluxing mixture of
magnesium turnings (0.14 g, 5.8 mmol) in 3 mL of tetrahydrofuran
which had been previously treated with a small crystal of iodine and
heated to reflux until the brown color was discharged. After 30 min
at reflux the flask was fitted for distillation, and cyclopropylacetylene
was collected along with tetrahydrofuran. Copper(I) bromide (0.726
g, 5.1 mmol) was placed in a flame-dried flask equipped with a stir
bar and placed under vacuum (1 Torr) for 1 h. Under nitrogen, 15 mL
of dry tetrahydrofuran was added, and then the mixture was cooled to
-60 °C and 5.0 mL of tert-butylmagnesium chloride (2.0 M in
tetrahydrofuran, 10 mmol) was added dropwise with stirring so that
the temperature did not rise above -55 °C. The mixture was then
stirred at -60 °C for 1 h, after which the cyclopropylacetylene/
tetrahydrofuran solution was added dropwise. The mixture was warmed
to 0 °C and stirred for 20 min. The reaction was then quenched by
dropwise addition of 10% aqueous sodium hydroxide (4.4 mL). Care
(14) Dolgii, I. E.; Okonnischnikova, G. P.; Nefedor, O. M. Bull. Acad.
Sci. USSR, DiV. Chem. Sci. (Engl. Transl.) 1970, 204. Dolgii, I. E.;
Okonnischnikova, G. P.; Nefedor, O. M.; Schwedova, I. B. Angew. Chem.,
Int. Ed. Engl. 1972, 11, 929.
(15) Maier, G.; Hoppe, M.; Reisenauer, H. P.; Kruger, C. Angew. Chem.,
Int. Ed. Engl. 1982, 21, 437; Angew. Chem. Suppl. 1992, 1061. Maier, G.;
Wolf, B. Synthesis 1985, 571.

306 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Gajewski et al.
mL). This solution was dried over magnesium sulfate, filtered, and
concentrated by distillation of the solvent through an 18 cm Vigreux
column, and the residue was subjected to short-path distillation to give
1.0 g of cis-2,3-dideuterio-trans-cyclopropanecarboxaldehyde (50%
yield based on starting alcohol). 1H NMR (300 MHz in CDCl3): δ
8.89 (d, J ) 4 Hz, 1H); 1.84 (m, 1H); 1.09 (s, 0.20H); 1.08 (d, J ) 7
Hz, 2H).
mmol, 85% yield) was purified by GC. 1H NMR (300 MHz, CDCl3):
δ 4.65 (s, 0.04H); 4.45 (s, 0.04H); 1.36 (m, 1H); 1.14 (s, 9H); 0.67
(m, 2H); 0.42 (m, 2H).
Pyrolyses. Most of the pyrolyses were carried out in a 2 L Pyrex
bulb immersed in a mechanically stirred bath of molten potassium
nitrate/sodium nitrite (10:7 v/v). The majority of heat was supplied to
the bath by a 500 W Vycor immersion resistance heater connected to
a variable transformer; fine temperature control was provided by a 125
W knife blade heater controlled by a Bayley Model 76-8 precision
temperature controller with a temperature feedback probe immersed
in the bath. The temperature of the bath was monitored with an
electronic digital thermometer equipped with a platinum resistance
probe. The precision of the thermometer was (0.1 °C, and the observed
temperature variation of the bath during pyrolysis runs was no more
than (0.1 °C.
The bulb was connected to a vacuum line that kept the pressure at
<10-3 Torr. Samples could be transferred into and out of the bulb
with high efficiency. The bulb was conditioned by transferring in 1
mL of dichlorodimethylsilane. The silane was kept in the bulb at 280
°C for 15 min and transferred out. After identical treatment with
diisopropylamine, the bulb was evacuated for 3 h. Cyclohexene (0.25
mL) was then used similarly to further condition the bulb overnight. A
system of high-vacuum valves was used to ensure that no air was
allowed to enter the bulb at any time after conditioning.
The aldehyde was immediately converted to the corresponding
acetylene, and treated with tert-butylmagnesium cuprate as described
above for the synthesis of 1H; however, the vinylcopper intermediate
was quenched with 10% sodium deuteroxide in deuterium oxide. GC
1
purification gave 200 µL of 1D3 (10% yield from the dibromide). H
NMR (300 MHz, CDCl3): δ 4.66 (d, J ) 1 Hz, 0.024H) (1D2); 4.43
(s, 1H); 1.35 (m, 1H); 1.14 (s, 9H); 0.66 (m, 0.14H) (30% 5D3, 70%
6); 0.39 (d, J ) 5 Hz, 2H) plus a downfield shoulder from 6. 2H NMR
(55 MHz, CDCl3) δ 4.66 (br s, 1D); 0.61 (br s, 2D). FTIR (CDCl3):
3100, 2967, 2875, 2871, 2361, 2357, 1606, 1362, 912 cm-1. MS (EI,
m/e):127.
Also present in this sample was ca. 7% of what appears to be cis-
2,3-dideuterio-1-(1′-deuterio-cis-2′-tert-butylvinyl)cyclopropane since
a doublet appeared at 0.27 ppm and a singlet at 5.26 ppm in addition
to what appears to be a tert-butyl singlet at 1.18 ppmsall in the
appropriate ratios.
1-tert-Butyl-6-oxabicyclo[3.1.0]hexane. To a magnetically stirred
slurry of m-chloroperbenzoic acid (0.8 g, 10 mmol) in methylene
chloride (50 mL) under a nitrogen atmosphere was added dropwise
1-tert-butylcyclopentene4 (1.2 g, 10 mmol) with cooling. The mixture
was stirred at room temperature for 28 h; then it was filtered and washed
with 5% aqueous sodium bicarbonate solution (2 × 15 mL), water (10
mL), and saturated brine (5 mL). After drying with magnesium sulfate,
filtration, and concentration under aspirator vacuum, the residue was
distilled (BP 75-90 °C, 25 Torr) to give 1.0 g of epoxide which was
further purified by preparative GC (OV-101, 20% on Chromosorb P,
12 ft × 1/4 in. column). 1H NMR (500 MHz, benzene-d6): δ 3.07 (s,
1H); 1.77 (dd, J ) 13.6, 7.8 Hz, 1H); 1.68 (dd, J ) 12.6, 7.5 Hz, 1H);
1.44 (ddddd, J ) 11.6, 9.7, 9.2, 7.8, 7.5 Hz, 1H); 1.39 (ddd, J ) 12.6,
9.7, 8.4 Hz, 1H); 1.29 (ddd, J ) 11.6, 9.2, 8.4, 1H); 1.17 (ddd, J )
13.6, 9.2, 9.2, 1H); 0.91 (s, 9H) (see figure in the Results for methylene
proton region). IR (neat): 3050, 2950, 1370, 1250 cm-1. MS (EI,
m/e): 140.
(1′-tert-Butyl-2′,2′-dideuteriovinyl)cyclopropane, 1XD2. (1′-tert-
Butyl-2′,2′-dideuteriovinyl)cyclopropane, 1XD2, was prepared from
tert-butyl cyclopropyl ketone16 (0.8 g, 6.4 mmol) in 2 mL of dimethyl
sulfoxide-d6 which was added dropwise to a stirred solution of
(dideuteriomethylene)triphenylphosphorane (8.9 mmol, prepared ac-
cording to literature procedures17 ) in 20 mL of dimethyl sulfoxide-d6
under a nitrogen atmosphere. After 5 h of stirring, the reaction mixture
was poured into water (75 mL), and extracted with pentane (2 × 75
mL). The combined organic phases were washed with water (2 × 10
mL), dried over magnesium sulfate, filtered, and concentrated by solvent
removal through a 18 cm Vigreux column. The residue (0.7 g, 5.2
Samples were purified by preparative GC prior to pyrolysis. They
were degassed by two freeze-pump-thaw cycles, and then distilled
into the pyrolysis bulb.
For rate studies of rearrangements of (1′-tert-butylvinyl)cyclopro-
panes, 1.2 µL samples were pyrolyzed; after recovery and dilution with
0.5 mL of pentane, capillary GC analysis was performed. For studies
of the rate of geometric isomerization of 1D3, 5 µL samples were
pyrolyzed; after recovery, dilution with 0.5 mL of deuteriochloroform
1
gave a solution suitable for H NMR and capillary GC analysis. For
stereochemical studies of the rearrangement of 1D3, 30-70 µL samples
were pyrolyzed for varying times; the recovered pyrolysate was diluted
with methylene chloride and the percent conversion to cyclopentene
product determined by capillary GC. This solution was then treated
with m-CPBA, and the residue after workup and concentration was
subjected to preparative GC. The separated [3.1.0] compound was then
diluted with benzene-d6 (0.5 mL) and analyzed by capillary GC and
1H{2H} NMR to determine the purity and to analyze the stereochemistry
of the product. A sample of unpyrolyzed 1 was subjected to similar
epoxidation, and no [3.1.0] product was observed by capillary GC or
1
by H NMR, demonstrating that cyclopentene is formed exclusively
by thermal rearrangement.
Acknowledgment. We thank the National Science Founda-
tion for financial support, Professor Christopher Samuel (Uni-
versity of Warwick) who provided the initial NMR data and
spectral analysis of the product epoxide to warrant pursuit of
this work, and Dr. Kevin Gilbert for his continued interest in
this problem.
(16) Hrubiec, R. T.; Smith, M. B. J. Org. Chem. 1984, 49, 431.
(17) Greenwald, R.; Chaykovsky, M.; Corey, E. J. J. Org. Chem. 1963,
28, 1128.
JA951578P