Thermal reactions of 8-methylbicyclo[4.2.0]oct-2-enes: Competitive diradical-mediated [1,3] sigmatropic, stereomutation, and fragmentation processes

Article abstract of DOI:10.1021/jo051505g

At 275 °C, 8-exo-methylbicyclo[4.2.0]oct-2-ene (1a) undergoes a [1,3] sigmatropic rearrangement to 5-methylbicyclo[2.2.2]oct-2-enes, of which the orbital symmetry-allowed si product is only marginally favored over the forbidden sr product; that is, silsr is 2.4. Accompanying the [1,3] shift are significant amounts of epimerization and fragmentation. The 8-endo epimer 1b, which yields no [1,3] product, experiences primarily direct fragmentation and secondarily epimerization. A diradical intermediate can account for all such observations.

Full text of DOI:10.1021/jo051505g

Thermal Reactions of 8-Methylbicyclo[4.2.0]oct-2-enes:
Competitive Diradical-Mediated [1,3] Sigmatropic,
Stereomutation, and Fragmentation Processes
Xavier S. Bogle, Phyllis A. Leber,* Lynne A. McCullough, and David C. Powers
Department of Chemistry, Franklin & Marshall College, Lancaster, Pennsylvania 17604-3003
phyllis.leber@fandm.edu
Received July 19, 2005
At 275 °C, 8-exo-methylbicyclo[4.2.0]oct-2-ene (1a) undergoes a [1,3] sigmatropic rearrangement
to 5-methylbicyclo[2.2.2]oct-2-enes, of which the orbital symmetry-allowed si product is only
marginally favored over the forbidden sr product; that is, si/sr is 2.4. Accompanying the [1,3] shift
are significant amounts of epimerization and fragmentation. The 8-endo epimer 1b, which yields
no [1,3] product, experiences primarily direct fragmentation and secondarily epimerization. A
diradical intermediate can account for all such observations.
Introduction
have been conducted, there is only one prior literature
reference to bicyclo[4.2.0]oct-2-ene thermal chemistry. In
1973, Berson and Holder reported8 the results of a gas-
phase thermal study of the 8-exo- and 8-endo-methyl
epimers of 7-endo-acetoxy-8-methylbicyclo[4.2.0]oct-2-
ene. Although Berson and Holder acknowledge a compet-
ing epimerization reaction, there is no mention of either
retro-Diels-Alder reactions of the bicyclo[2.2.2]octenes
or of Diels-Alder recombination reactions that are
known to complicate the analysis of the analogous
bicyclo[3.2.0] compounds.2b With respect to the significant
issue of [1,3] stereochemistry, Berson and Holder cite an
sr/si value of 0.45 (si/sr ) 2.2) for the 8-exo-methyl epimer
and sr/si of 12 (si/sr ) 0.08) for the 8-endo-methyl epimer.
Experimental9,10 and theoretical11,12 reappraisals of the
“not obviously concerted” vinylcyclobutane-to-cyclohexene
The vinylcyclobutane-to-cyclohexene rearrangement
has been formally classified as a [1,3] sigmatropic rear-
rangement. For a vinylcyclobutane appropriately labeled
at both the migrating carbon and migration terminus,
four discrete [1,3] products exist: si, sr, ai, and ar, where
s or a refer to suprafacial or antarafacial involvement of
the π system and where i or r, inversion or retention,
designate the stereochemical outcome with respect to the
migrating carbon as it moves across the π system.
According to the Woodward-Hoffmann selection rules,1
the si and ar products are symmetry-allowed; sr and ai,
symmetry-forbidden. Only si and sr products, however,
are geometrically accessible in structurally constrained
bicyclic vinylcyclobutanes. Thus, the si/sr ratio has been
viewed as a measure of the degree of orbital symmetry
control or concertedness of the [1,3] carbon migration of
bicyclic vinylcyclobutanes.
(6) Kla¨rner, G.-G.; Drewes, R.; Hasselmann, D. J. Am. Chem. Soc.
1988, 110, 297-298.
(7) Bender, J. D., Leber, P. A.; Lirio, R. R.; Smith, R. S. J. Org. Chem.
2000, 65, 5396-5402.
While numerous experimental studies2-7 of the [1,3]
sigmatropic rearrangement of bicyclo[3.2.0]hept-2-enes
(8) Berson, J. A.; Holder, R. W. J. Am. Chem. Soc. 1973, 95, 2037-
2038.
(9) (a) Baldwin, J. E.; Burrell, R. C. J. Am. Chem. Soc. 2001, 123,
6718-6719. (b) Baldwin, J. E.; Burrell, R. C. J. Am. Chem. Soc. 2003,
125, 15869-15877.
(10) (a) Doering, W. v. E.; Ekmanis, J. L.; Belfield, K. D.; Kla¨rner,
F.-G.; Krawczyk, B. J. Am. Chem. Soc. 2001, 123, 5532-554. (b)
Doering, W. v. E.; Cheng, X.; Lee, W.; Lin, Z. J. Am. Chem. Soc. 2002,
124, 11642-11652.
(11) (a) Carpenter, B. K. J. Org. Chem. 1992, 57, 4645-4648. (b)
Carpenter, B. K. J. Am. Chem. Soc. 1995, 117, 6336-6344.
(12) (a) Beno, B. R.; Wilsey, S.; Houk. K. N. J. Am. Chem. Soc. 1999,
121, 4816-4826. (b) Wilsey, S.; Houk. K. N.; Zewail, A. H. J. Am. Chem.
Soc. 1999, 121, 5772-5786.
(1) Woodward, R. B.; Hoffmann, R. The Conservation of Orbital
Symmetry; Verlag Chemie: Weinheim, 1970; pp 119-120.
(2) (a) Berson, J. A.; Nelson, G. L. J. Am. Chem. Soc. 1967, 89, 5503-
5504. (b) Berson, J. A.; Nelson, G. L. J. Am. Chem. Soc. 1970, 92, 1096-
1097.
(3) Cocks, A. T.; Frey, H. M. J. Chem. Soc. A 1971, 2564-2566.
(4) Forman, M. A.; Leber, P. A. Tetrahedron Lett. 1986, 27, 4107-
4110.
(5) (a) Baldwin, J. E.; Belfield, K. D. J. Am. Chem. Soc. 1988, 110,
296-297. (b) Baldwin, J. E.; Belfield, K. D. J. Phys. Org. Chem. 1989,
2, 455-466.
10.1021/jo051505g CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/01/2005
J. Org. Chem. 2005, 70, 8913-8918
8913

Bogle et al.
SCHEME 1. Synthesis of 1
SCHEME 2. Synthesis of 2
rearrangement have proffered an alternative to a con-
certed orbital symmetry-controlled mechanism: a con-
formationally promiscuous diradical intermediate on a
relatively flat potential energy surface. A review13 of this
recent work has led to several important postulates for
bicyclic vinylcyclobutanes: (1) the si/sr ratio should
correlate with the conformational flexibility of the π
system, (2) a methyl substituent on the migrating carbon
slows down the rate of rotation relative to a deuterium
substituent in a short-lived diradical and thus results in
a higher si/sr ratio, and (3) fragmentation and stereo-
mutation will compete with [1,3] shifts in conformation-
ally labile diradicals.
We wish to report herein the results of a thorough
kinetic analysis of 8-exo-methylbicyclo[4.2.0]oct-2-ene
(1a) and 8-endo-methylbicyclo[4.2.0]oct-2-ene (1b) and,
as a corollary, to subject the aforementioned predictions
to experimental verification. A direct comparison of the
stereochemical and kinetic results with those of other
methyl-labeled vinylcyclobutanes will also be made.
the methyl ketone 4, primarily endo, is converted to
mostly 8-exo-methylbicyclo[4.2.0]oct-2-ene (1a), making
the purification and isolation of 1b from 1a difficult.
Thus, it was more convenient to acquire spectra of pure
1a as well as of a mixture of 1a and 1b.
The synthesis of 2 (Scheme 2), as an epimeric mixture,
is based on the application of the two-step low-temper-
ature Wolff-Kishner reduction15 to compound 6, the
Diels-Alder cycloadduct of 1,3-cyclohexadiene, and acro-
lein. The Lewis acid-catalyzed reaction, using either
aluminum trichloride17 or boron trifluoride,18 proceeds in
good to excellent yield, limited only by the gradual
evaporative loss of acrolein during the course of the
reaction. Spectral elucidation relies primarily on 13C
NMR values that closely match those reported in the
literature.19
Differentiation between the 5-exo- (2a) and 5-endo-
methylbicyclo[2.2.2]oct-2-ene (2b) epimers is based on 1H
and 13C NMR characterization.20-22 Although the endo-
methyl hydrogens experience a typical shielding environ-
ment,23 the 13C NMR chemical shift for the endo-methyl
in 2b is 23.1 ppm compared to 20.5 ppm for the exo-
methyl in 2a.21
Results
Syntheses. A mixture of the 8-methyl epimers 1a and
1b was prepared by subjecting 8-methylbicyclo[4.2.0]oct-
2-en-7-one14 (4) to a two-step cyclobutanone reduction
sequence15 based on low-temperature Wolff-Kishner
reduction of the hydrazone derivative 5. Because the
direct conversion of 1,3-cyclohexadiene and methylketene,
generated by treatment of propionyl chloride with tri-
ethylamine, resulted in compound 4 in only 4-5% yield,
we opted instead to initiate the synthesis of 1 from the
8-chloro-8-methylbicyclo[4.2.0]oct-2-en-7-one14 (3) precur-
sor (Scheme 1). As the latter ketene cycloaddition pro-
ceeded in 63% yield and the subsequent zinc reduction16
in 53% yield, the overall yield of 33% for the two-step
sequence was far superior to the one-step variation.
Because compounds 3 and 4 have been reported in the
literature previously, only compound 1 was subjected to
rigorous structure proof. Using preparative GC, 1a was
obtained free of 1b. Base-catalyzed epimerization, which
has been noted previously,7 routinely occurs during the
low-temperature Wolff-Kishner reaction sequence. Thus,
Thermolysis Reactions. Gas-phase thermal reac-
tions of 1a, 1b (contaminated with 4-7% 1a), and a
mixture of 2a and 2b have been performed in a well-
conditioned kinetic bulb24 at 275.0 °C. As there was an
ample amount of pure 1a, obtained at greater than 99.5%
purity by preparative GC, eight kinetic points between
0.5 and 24 h, corresponding to between 7 and 68%
conversion, were determined. A representative capillary
GC chromatogram for the thermal reaction of 1a is shown
in Figure 1. Given the facile epimerization of 1b to 1a
and the relative dearth of 1b, only four kinetic runs of 1,
2, 3, and 4 h were done over ca. 1 half-life for 1b. The
thermal behavior of 1a was also monitored at a temper-
(17) Fringuelli, F.; Ming, G.; Minuti, L.; Pizzo, F.; Taticchi, A.;
Wenkert, E. J. Org. Chem. 1989, 54, 710-712.
(18) Masjedizadeh, M. R.; Dannecker-Doerig, I.; Little, R. D. J. Org.
Chem. 1990, 55, 2742-2752.
(19) Wenkert, E.; Cochran, D. W.; Gottlieb, H. E.; Hagaman, E. W.;
Filho, R. B.; Matos, F. J. A.; Madruga, M. I. L. M. Helv. Chim. Acta
1976, 59, 2437-2442.
(20) Smith, A. B., III; Agosta, W. C. J. Org. Chem. 1972, 37, 1259-
1262.
(21) Stothers, J. B.; Tan, C. T. Can. J. Chem. 1976, 54, 917-925.
(22) Grob, C. A.; Sawlewicz, P. Helv. Chim. Acta 1984, 67, 1859-
1867.
(23) Debande, G.; Huybrechts, G. . Int. J. Chem. Kinet. 1974, 6, 545-
552.
(24) Gajewski, J. J.; Chou, S. K. J. Am. Chem. Soc. 1977, 99, 5696-
5707.
(13) Leber, P. A.; Baldwin, J. E. Acc. Chem. Res. 2002, 35, 279-
287.
(14) Wu, H. Y.; Walker, K. A. M.; Nelson, J. T. J. Org. Chem. 1994,
59, 1389-1395.
(15) Burkey, J. D.; Leber, P. A.; Silverman, L. S. Synth. Commun.
1986, 16, 1363-1370.
(16) Koltun, E. S.; Kass, S. R. J. Org. Chem. 2000, 65, 3530-3537.
8914 J. Org. Chem., Vol. 70, No. 22, 2005

Thermal Reactions of 8-Methylbicyclo[4.2.0]oct-2-enes
SCHEME 3. Kinetic Schemea
a k1,1 ) rate of epimerization of 1a to 1b. k′1,1 ) rate of
epimerization of 1b to 1a. k1,3 ) rate of [1,3] sigmatropic rear-
rangement of 1a to 2a/2b. kf ) rate of direct fragmentation of 1a
to 1,3-cyclohexadiene and propene. k′f ) rate of direct fragmenta-
tion of 1b to 1,3-cyclohexadiene and propene.
TABLE 2. Rate Constants (×105 s) Derived Using
Runge-Kutta Numerical Integration
compd
T (°C)
kd (k′d)
k1,1 (k′1,1
)
k1,3 (si + sr)
kf (k′f)
1a
275
300
275
1.5
11.0
4.3
0.7
2.2
1.0
0.2
1.5
0.6
7.3
3.3
1b
FIGURE 1. Capillary GC chromatogram for thermolysis of
1a with elution order 2b, 2a, 1a, internal standard (propyl-
cyclohexane), and 1b. (Chromatogram traced for clarity; see
the Supporting Information for the original.)
Monitoring the disappearance of reactant 1a or 1b
versus time has resulted in the rate constant for overall
decomposition, kd or k′d, respectively. Approximate ex-
perimental values for k1,1 (or k′1,1), k1,3, and kf (or k′f) have
been obtained from the slopes of linear plots of product
TABLE 1. Rate Constants (×105 s) Derived Using
Simplified Parallel First-Order Reactions Model
mole fraction versus 1 - e-k t. This simplified rate model
d
compd T (°C)
kda (k′d)b
k1,1 (k′1,1) k1,3 (si) k1,3 (sr) kf (k′f)
proved satisfactory for obtaining accurate si (1a f 2a)
and sr (1a f 2b) rate constants (Scheme 3). A Runge-
Kutta four-step numerical integration26 (calculated using
an Excel spreadsheet) has afforded modestly different
rate constants, as reported in Table 2, presumably
yielding improved k1,1 and k′1,1 values.
1a
275
300
1.25 ( 0.04
8.0 ( 0.3
0.38
1.2
0.12
0.7
0.05
0.3
0.70
5.8
15
3.2
315 20 ( 1
275 4.1 ( 0.2
2.2
2.2
1.0
1b
0.9
a Rate of overall decomposition of 1a. b Rate of overall decom-
position of 1b.
Molecular Modeling. Molecular modeling (AM1)
permitted an examination of the low energy conformers
of the following series of compounds: vinylcyclobutane,
bicyclo[3.2.0]hept-2-ene, and bicyclo[4.2.0]oct-2-ene. It is
apparent that the second ring in bicyclo[3.2.0]hept-2-ene
and bicyclo[4.2.0]oct-2-ene limits the distance between
the migrating carbon, C7 and C8, respectively, and the
migration terminus C3. The magnitude of this gap, not
surprisingly, increases in the following order: 3.15 Å for
the conformer of vinylcyclobutane that would afford an
allylic diradical with cisoid stereochemistry, 3.38 Å for
the single low-energy conformer of bicyclo[3.2.0]hept-2-
ene, and 3.57 Å, the average distance (of 3.47 and 3.67
Å) for the two low-energy conformers of bicyclo[4.2.0]oct-
2-ene. Whereas the cyclopentene ring in bicyclo[3.2.0]-
hept-2-ene is virtually planar, the cyclohexene ring in
bicyclo[4.2.0]oct-2-ene is decidedly puckered, giving rise
to two low energy conformers (see the Supporting Infor-
mation). This conformational flexibility in the six-
membered ring has the consequence of increasing the gap
between the migrating carbon and the migration termi-
nus.
ature of 300 °C to probe for any temperature dependence
in the si/sr ratio and to follow the kinetics for 5 half-lives.
Given the ca. 15 h half-life of 1a at 275 °C, the higher
temperature enabled us to examine 1a more conveniently
over an extended time frame. Thermal reactions of 1a
at 315 °C were also run at 1 h intervals for a total of 3 h
to allow determination of Arrhenius parameters.
A prior report of the activation parameters for the
retro-Diels-Alder reactions25 of 2a and 2b has allowed
us to predict that 2a and 2b are thermally inert at 275
°C. Nonetheless, we heated a mixture of 2a and 2b for
intervals up to 24 h to show experimentally that the
concentrations of 2a and 2b are invariant over 24 h at
275 °C.
The experimental first-order rate constants given in
Table 1 correspond to thermal interconversions depicted
in the kinetic scheme (Scheme 3). As shown in Scheme
3, 1a and 1b epimerize with each other. In addition, 1a
undergoes both [1,3] sigmatropic rearrangement to 2a
and 2b and direct fragmentation to 1,3-cyclohexadiene
and propene whereas 1b experiences only direct frag-
mentation. All experimental first-order rate constants are
given in Table 1 (k values refer to rate constants for 1a;
k′ values refer to those for 1b).
Spartan ′04 calculations have also been performed on
the compounds listed in Table 3, which provides a
compilation of stereochemical results for a series of
bicyclic and monocyclic vinylcyclobutanes. The only
(25) Huybrechts, G.; Ngoy, G. Int. J. Chem. Kinet. 1975, 7, 775-
784.
(26) Carpenter, B. K. Determination of Organic Reaction Mecha-
nisms; Wiley: New York, 1984; pp 76-81.
J. Org. Chem, Vol. 70, No. 22, 2005 8915

Bogle et al.
TABLE 3. Stereochemistry of [1,3] Shifts in Vinylcyclobutanes
TABLE 4. Kinetic Comparisona of
8-Methylbicyclo[4.2.0]oct-2-ene (1) and
7-Methylbicyclo[3.2.0]hept-2-ene (8), 275 °C
significant difference, compared to the nonmethylated
hydrocarbons, is that the average distance between the
migrating carbon C8 and the migration terminus in
compound 1a is 3.49 Å, 0.11 Å greater than the 3.38 Å
in 8a.
exo-CH3
endo-CH3
1a
8a
1b
8b
kd (×105 s)
1.5
15
47
38
1.5
94
5
4.3
1.0
40
38
12
k1,3 (%)
k1,1 (%)
kf (%)
Discussion and Conclusions
23
77
1
The crucial stereochemical probe resulting from this
study, the si/sr ratio of 1a, is robust because the retro-
Diels-Alder reactions of 2a and 2b are 3 orders of
magnitude slower than kd and at least 200 times slower
than k1,3. This conclusion has been supported experimen-
tally by showing that the concentrations of 2a and 2b
are constant over a 24 h thermal investigation at 275 °C
and by confirming invariant concentrations of 2a and 2b
using a Runge-Kutta numerical analysis based on
calculated retro-Diels-Alder rate constants (from litera-
ture activation parameters)24 as inputs. An experimental
si/sr ratio of 2.4 for 1a has been determined at both 275
and 300 °C; the value of si/sr at 315 °C is 2.3. Thus, the
si/sr value does not vary appreciably with temperature,
a fact consistent with our prior observation7 for 7-exo-
methylbicyclo[3.2.0]hept-2-ene (8a).
k1,1 > kf > k1,3 k1,3 . k1,1 > kf kf > k1,1 k1,3 ≈ k1,1 > kf
a Based on Runge-Kutta rate constants.
(si/sr ) 0.08). Compound 1b, the analogue without the
endo-acetoxy substituent, yields no observable 2a or 2b.
Conceivably, the small amount of [1,3] observed by
Berson and Holder was derived from the 8-exo-methyl
epimer formed by epimerization at C8. Alternatively, the
steric effect caused by the 7-endo-acetoxy and 8-endo-
methyl substituents could have favored a conformer in
which the migrating carbon C8 might have positioned
itself closer to the migration terminus.
The difference in the [1,3] rate constants for the si and
sr pathways, the 1a f 2a or 1a f 2b conversions,
respectively, translates to a calculated difference in
transition state energy of less than 1 kcal/mol. This value
is remarkably close to one calculated previously by Houk
et al. for the energy difference between the si and sr
transition states of bicyclo[3.2.0]hept-2-ene.12b
The earlier prediction that the magnitude of the si/sr
ratio correlates with the relative flexibility of the π
system13 is borne out by the cumulative results in Table
3. The si/sr ratio for 1a is lower than that of 8a by a factor
of 3. More significantly, however, is the remarkable
similarity in the si/sr ratio of 2.4 for 1a and the si/sr value
of 1.8 reported earlier by Baldwin and Burrell for a
methyl-labeled monocyclic vinylcyclobutane (Table 3).
In the related study of 7-endo-acetoxy-8-exo-methylbi-
cyclo[4.2.0]oct-2-ene, Berson reports a similar stereo-
chemical ratio, si/sr ) 2.2. Given that we have studied
1a at three different temperatures, we have calculated
Arrhenius parameters (Ea ) 44.3 ( 1.5 kcal/mol; log A
) 12.8 ( 0.6) that are similar but not identical to those
cited by Berson27 for 7-endo-acetoxy-8-exo-methylbicyclo-
[4.2.0]oct-2-ene: Ea ) 46.7 kcal/mol; log A ) 13.2. In
addition, we have determined Arrhenius parameters
specifically for the 1a f 2a + 2b transformation: Ea )
46.3 ( 1.9 kcal/mol and log A ) 12.7 ( 0.7, values close
to those corresponding to the overall rate of disappear-
ance of 1a.
A kinetic comparison between 1a and 8a in Table 4
shows that the [1,3] shift, which was dominant for 8a,
only marginally competes with fragmentation and epimer-
ization in 1a. Molecular modeling of 1a compared to 8a
provides a rationale for the diminished [1,3] contribution
to the overall thermal manifold. The additional 0.1 Å gap
between the migrating carbon and the migration termi-
nus is apparently sufficient to advantage direct fragmen-
tation and epimerization over the [1,3] shift. That such
processes can indeed compete with the [1,3] sigmatropic
rearrangement further strengthens the presupposition
of an incipient diradical intermediate. There is no dif-
ference in the kd values for 1a and 8a, merely a difference
in the allocation of the various exit channels from their
corresponding diradical intermediates. Regardless of how
the resultant diradical intermediate distributes itself
among the various products, both 1a and 8a produce
diradicals with equivalent energy costs. This observation
For 7-endo-acetoxy-8-endo-methylbicyclo[4.2.0]oct-2-
ene Berson and Holder report8 an sr/si value of 12
(27) Berson, J. A. Acc. Chem. Res. 1972, 5, 406-414.
8916 J. Org. Chem., Vol. 70, No. 22, 2005

Thermal Reactions of 8-Methylbicyclo[4.2.0]oct-2-enes
TABLE 5. Kinetic Comparisona of 1a at 275° and 300 °C
tially; that is, in all cases kf > k1,1 g k1,3. A Runge-Kutta
numerical analysis has not been conducted on the
thermal data for 1a at 315 °C due to the limited number
of data points.
275 °C
300 °C
kd (×105 s)
k1,3 (%)
k1,1 (%)
kf (%)
1.5
15
47
38
11.
14
20
66
Whereas the [1,3] contribution remains relatively
constant at both 275 and 300 °C, an observation concep-
tually consistent with a nonstatistical11b diradical inter-
mediate, the amount of fragmentation relative to epimer-
ization increases directly with temperature. This phe-
nomenon suggests that both epimerization and fragmen-
tation might be linked to a statistical diradical interme-
diate that yields more fragmentation at higher temper-
ature due to the greater entropic favorability of fragmen-
tation over isomerization. If this is indeed the case, one
might conclude that 1b, given the complete absence of
2a and 2b at 275 °C, cannot generate a nonstatistical
diradical intermediate as a consequence of the extreme
steric crowding caused by the endo-methyl substituent.
Of the three postulates formulated in an Accounts of
Chemical Research review paper on the vinylcyclobutane-
to-cyclohexene rearrangement,13 two are unambiguously
supported by the results of the current study. The low
si/sr ratio of 2.4 for 1a is consistent with a model of a
conformationally flexible bicyclic vinylcyclobutane. As
shown dramatically in Table 3, the si/sr ratio for 1a is
closer to that of the corresponding monocyclic vinylcy-
clobutane than it is to that of the bicyclic vinylcyclobu-
tane 8a. The dominance of epimerization and fragmen-
tation in the thermal reactions accessible to 1a and the
exclusive presence of these two processes for 1b confirm
the prediction that the presence of stereomutation and
fragmentation implicates a diradical process. What re-
mains to be tested is the assumption of a lower si/sr ratio
when the migrating carbon is labeled with a deuterium
in lieu of a methyl, which due to its greater size is
expected to slow the rate of rotation in the incipient
diradical. The results of our investigation of 8-d1-bicyclo-
[4.2.0]oct-2-ene are forthcoming.
k1,1 > kf > k1,3
kf > k1,1 > k1,3
a Based on Runge-Kutta rate constants.
is in agreement with the third prediction (vide supra)
from the recent review13 of vinylcyclobutane thermal
chemistry.
The inward dynamic trajectory that both Carpenter11b
and Houk12 have predicted for bicyclo[3.2.0]hept-2-ene
and its derivatives is favored in 1a as well, if both [1,3]
and [1,1] sigmatropic processes are linked to this motion.
The increased separation between the migrating carbon
and the migration terminus for the [1.3] rearrangement
might well permit a greater extent of rotation before
closure to the bicyclo[2.2.2] molecular framework, thus
accounting for the lower si/sr ratio of 1a compared to
that of 8a (Table 3). The inward trajectory, however, is
completely ineffective for 1b due to the extreme van der
Waals repulsion between the endo-methyl substituent on
the migrating carbon and either an endo-H on C4 or an
endo-H on C5 (see the Supporting Information). By
default, outward rotation of C8, the migrating carbon,
during the bond-breaking event favors fragmentation
over epimerization by a factor of 3 to 1. Fragmentation,
which is virtually nonexistent in 8a and only a minor
contributor to the thermal profile of 7-endo-methylbicyclo-
[3.2.0]hept-2-ene (8b), is actually competitive with epimer-
ization for 1a and becomes the major thermal event for
1b. As fragmentation becomes more prominent, there is
correspondingly less [1,3] migration. The following k1,3
/
kf ratios provide a compelling comparison between the
thermal behavior of the corresponding bicyclo[3.2.0] and
[4.2.0] compounds: 8a, 150; 8b, 3; 1a, 0.4; 1b, 0.
Doering has provided an apt characterization of “non-
interactive singlet diradicals”10b with excited vibrational
states accessible only when all residual bonding between
the migrating carbon and the migration origin is de-
stroyed by internal rotation. The consequence of this
internal rotation is that “orbital overlap vanishes and
with it rule by Woodward and Hoffmann.” Doering also
enunciates “the principle of least rotational involvement”
and its corollary, which is that a larger number of
internal rotations for a given diradical species diminishes
its contribution to the product distribution. The diradical
species that forms from 1a and closes to 2a and 2b
exhibits only a modest selectivity toward 2a, indicative
of a greater propensity toward rotation. This also pro-
vides a strong rationale for the minor amount of 2a and
2b, the [1,3] migration products, in the thermal manifold
of 1a and the complete absence of 2a and 2b in the
thermal profile of 1b.
Experimental Section
Detailed general experimental procedures can be found in
ref 7.
8-Chloro-8-methylbicyclo[4.2.0]oct-2-en-7-one (3). To a
solution of 50 mL of 1,3-cyclohexadiene (0.52 mol) and 31 mL
of 2-chloropropionyl chloride (0.32 mol) in 75 mL of chloroform
was added a solution of 35 mL of triethylamine (0.25 mol) in
25 mL of chloroform dropwise over ca. 6 h. The reaction was
then stirred overnight at room temperature. After removal of
the chloroform via simple distillation and the triethylamine
hydrochloride by vacuum filtration, the filtrate was washed
successively with water and brine and dried over MgSO4.
Vacuum distillation afforded 26.9 g (63%) of compound 3 as
an isomeric mixture (54:46 ratio by GC), which was isolated
as a viscous oil: IR (cm-1) 3010 (w), 1780 (s), 710 (m), 700
(m); MS (m/z) 172 (M + 2, 2), 170 (M, 6), 135 (14), 80 (38), 79
(54), 55 (100).
8-Methylbicyclo[4.2.0]oct-2-en-7-one (4). To a mixture
of 11.6 g (177 mmol) of zinc dust in 23.5 mL (151 mmol) of
TMEDA and 59 mL of absolute ethanol was added 12 mL (0.21
mol) of glacial acetic acid over ca. 10 min. After addition of a
solution of 5.55 g (32.5 mmol) of compound 3 in 13 mL of acetic
acid over ca. 10 min, the reaction mixture was warmed to 40
°C and stirred overnight. Exercising care to minimize exposure
of the activated zinc to air, the mixture was first filtered
through a sintered-glass funnel, and the zinc residue was then
washed with water to deactivate the zinc. The filtrate was
A comparison of Runge-Kutta rate constants for 1a
at 275 and 300 °C can be found in Table 5. For the
resultant Runge-Kutta rate constants at 300 °C, there
is greater uncertainty because compound 1b has not been
studied at 300 °C due to its limited quantity. The
reported Runge-Kutta rate constants represent an aver-
age of three reasonable “fits” that do not differ substan-
J. Org. Chem, Vol. 70, No. 22, 2005 8917

Bogle et al.
washed successively with 1 N HCl, NaHCO3, and brine, dried
over MgSO4, and concentrated under reduced pressure to
formed on a sample consisting of a mixture of isomers that
was subjected to chromatographic purification by elution from
silica gel with 90:10 hexane/ethyl acetate: IR (cm-1) 3040 (w),
1725 (s), 700 (s). Bicyclo[2.2.2]oct-5-en-2-endo-carboxaldehyde
(major epimer): MS (m/z) 136 (12), 108 (14), 80 (53), 79 (100);
1H NMR (500 MHz, CDCl3) δ 9.43 (d, 1H); 13C NMR (125 MHz,
CDCl3) δ 204.2, 136.3, 130.8, 51.1, 30.9, 29.4, 26.9, 25.3, 24.9.
Bicyclo[2.2.2]oct-5-en-2-exo-carboxaldehyde (minor epimer):
MS (m/z) 136 (1), 80 (52), 79 (100), 58 (55). 1H NMR (500 MHz,
CDCl3) δ 9.74 (d, 1H); 13C NMR (125 MHz, CDCl3) δ 204.6,
135.8, 133.6, 50.4, 30.5, 29.7, 25.5, 25.0, 21.3.
Bicyclo[2.2.2]oct-5-ene-2-carboxaldehyde Hydrazone
(7). To a solution of 6.5 g (44 mmol) of hydrazine sulfate in 12
mL of hydrazine hydrate was added 3.6 g (26 mmol) of
compound 6. After the reaction mixture was heated overnight
at 65 °C, it was extracted three times with ether, washed with
water and brine, dried over MgSO4, and concentrated under
reduced pressure. Compound 7 (3.14 g, 79%) was used without
further purification: IR (cm-1) 3370 and 3200 (m, NH2), 3020
(w), 1630 (m, CdN), 700 (s).
afford 2.36 g (53%) of compound 4, 88% endo by GC. IR (cm-1
)
3020 (w), 1780 (s), 1640 (w), 700 (m). The exo- and endo-
epimers were separated by flash chromatography on silica gel
by elution with 95:5 pentane/ether. 8-exo-Methylbicyclo[4.2.0]-
oct-2-en-7-one (4a as a mixture with 4b): MS (m/z) 136 (10),
107 (100), 91 (78), 79 (87), 77 (48); 1H NMR (300 MHz, CDCl3)
δ 1.26 (d, 3H); 13C NMR (75 MHz, CDCl3) δ 214.0, 128.2, 128.0,
60.3, 54.2, 31.5, 21.4, 20.0, 14.6 (exo-CH3). 8-endo-Methylbicyclo-
[4.2.0]oct-2-en-7-one (4b): MS (m/z) 136 (42), 107 (59), 91 (100),
1
79 (100), 77 (77). H NMR (300 MHz, CDCl3) δ 5.94 (m, 1H),
5.75 (m, 1H), 3.59 (br m, 1H), 3.41 (ddq, 1H), 3.00 (br m, 1H),
1.98 (m, 3H), 1.50 (m, 1H), 1.00 (d, 3H); 13C NMR (75 MHz,
CDCl3) δ 214.1, 130.1, 125.5, 55.4, 55.2, 27.6, 21.3, 18.5, 8.8
(endo-CH3).
Using the same methodology as in the preparation of 3 but
substituting 26 mL (0.30 mol) of propionyl chloride for 2-chlo-
ropropionyl chloride resulted in only 1.7 g (5%) of compound
4, 97% endo by GC.
8-Methylbicyclo[4.2.0]oct-2-en-7-one Hydrazone (5). To
a solution of 2.4 g (18 mmol) of hydrazine sulfate in 7.0 mL of
hydrazine hydrate was added 1.2 g (8.9 mmol) of compound
4. After the reaction mixture was heated overnight at 65 °C,
it was extracted three times with ether, washed with water
and brine, dried over MgSO4, and concentrated under reduced
pressure. Compound 5 (1.16 g, 85%) was used without further
purification: IR (cm-1) 3370 and 3260 (m, NH2), 3020 (w), 1610
(m, CdN), 720 (s).
8-Methylbicyclo[4.2.0]oct-2-ene (1). To a solution of 2.21
g (19.7 mmol) of potassium tert-butoxide in 25 mL of anhydrous
DMSO was added a solution of 1.16 g (7.6 mmol) of compound
5 via syringe over 4-5 h. After being stirred overnight, the
reaction mixture was quenched with water and extracted with
pentane. The organic extract was then washed seven times
with water, dried over MgSO4, and concentrated by simple
distillation. The crude product (0.57 g, 61%) consisted of a 93:7
(as determined by GC) mixture of 1a/1b, respectively: IR
(cm-1) 3015 (w), 700 (s). 8-exo-methylbicyclo[4.2.0]oct-2-ene
(1a): MS (m/z) 122 (1), 81 (9), 80 (100), 79 (58), 77 (14). 1H
NMR (500 MHz, CDCl3) δ 5.75 (d, 2H), 2.50 (sextet, 1H), 2.20
(br t, 1H), 2.01 (dt, 2H), 1.93 (m, 1H), 1.82 (m, 1H), 1.59 (m,
1H), 1.53 (M, 2H), 1.12 (d, 3H); 13C NMR (125 MHz, CDCl3) δ
130.1 (HCd), 126.8 (HCd), 40.4 (HC), 36.8 (HC), 30.5 (H2C),
28.9 (HC), 25.3 (H2C), 22.1 (H2C), 21.3 (H3C). 8-endo-
methylbicyclo[4.2.0]oct-2-ene (1b): MS (m/z) 122 (0.5), 81 (9),
80 (100%), 79 (54), 77 (12); 1H NMR (300 MHz, CDCl3) δ 5.90
(m, 1H), 5. 66 (br d, 1H), 0.84 (d, 3H); 13C NMR (75 MHz,
CDCl3) δ 128.4, 127.8, 36.3, 33.6, 29.9, 29.6, 22.6, 21.2, 16.7
(H3C).
Bicyclo[2.2.2]oct-5-ene-2-carboxaldehyde (6). To a solu-
tion of 14.5 mL (120 mmol) of 1,3-cyclohexadiene in 100 mL
of toluene was added 1.66 mL (12.6 mmol) of anhydrous AlCl3
followed by a solution of 1.67 mL (90% Aldrich, 22.5 mmol) of
acrolein in 100 mL of toluene. After being stirred at 40 °C for
2 days, an additional increment of 1.67 mL of acrolein was
added. The next day, the cooled reaction mixture was poured
into ice-water. The aqueous layer was then extracted with
ether, which was washed with NaHCO3 and dried over Na2-
SO4. Distillation of the volatiles resulted in 3.7 g (60%) of crude
6, which was used without further purification.
Alternatively, to a solution of 5.5 mL (90%, 74 mmol) of
acrolein, 15.0 mL (97%, 153 mmol) of cyclohexadiene, and 82.0
mL of anhydrous ether was added by syringe over a period of
10 min 4.0 mL (32 mmol) of boron trifluoride diethyl etherate.
After being stirred overnight, the reaction mixture was
quenched with water and then extracted three times with
ether, washed with brine, dried over MgSO4, and concentrated
under reduced pressure to afford 9.4 g (95%) of crude 6, 80:20
endo/exo by GC analysis. Spectral characterization was per-
5-Methylbicyclo[2.2.2]oct-2-ene (2). To a solution of 2.6
g (24 mmol) of potassium tert-butoxide in 15 mL of anhydrous
DMSO was added a solution of 3.1 g (21 mmol) of compound
7 in 5 mL of anhydrous DMSO via syringe over 5 h. After being
stirred overnight, the reaction mixture was quenched with
water and extracted with pentane. The organic extract was
then washed seven times with water, dried over MgSO4, and
concentrated by simple distillation to afford 1.3 g (52%) of
compound 2, which consisted of a 23:77 mixture of 2a/2b,
respectively. 5-exo-methylbicyclo[2.2.2]oct-2-ene (2a): MS (m/
1
z) 122 (7), 80 (100), 79 (46); H NMR (300 MHz, CDCl3) δ 6.4
(t, 1H), 6.2 (t, 1H), 2.4 (br s, 1H), 2.2 (br s, 1H), 1.8 (m, 1H),
1.5 (m, 2H), 1.4 (m, 1H), 1.2 (m, 1H), 1.1 (m, 1H), 1.0 (d, 3H),
0.9 (m, 1H); 13C NMR (75 MHz, CDCl3) δ 136.6, 133.1, 35.6,
34.8, 30.6, 30.3, 26.2, 20.5 (CH3), 18.8. 5-endo-Methylbicyclo-
1
[2.2.2]oct-2-ene (2b): MS (m/z) 122 (7), 80 (100), 79 (43). H
NMR (300 MHz, CDCl3) δ 6.2 (t, 1H), 6.1 (t, 1H), 2.4 (br s,
1H), 2.2 (br s, 1H), 1.7 (m, 2H), 1.5 (m, 1H), 1.4 (m, 1H), 1.2
(m, 2H), 0.8 (d, 3H), 0.7 (m, 1H); 13C NMR (75 MHz, CDCl3) δ
134.2, 132.2, 36.5, 35.8, 32.1, 30.3, 26.5, 24.2, 23.1 (CH3).
Gas-Phase Reactions. Thermal reactions were conducted
in a previously described apparatus.7 Thermolysis samples
were analyzed by GC on an HP cross-linked methyl silicone
column (50 m × 0.2 mm i.d. × 0.10 µm film thickness)
operating at an initial temperature of 60 °C held for 1 min
followed by a temperature ramp of 0.6 °C/min to a maximum
temperature of 100 °C. Retention times (min) were as follows:
2b (12.92), 2a (13.50), 1a (13.86), the internal standard
propylcyclohexane (14.24), 1b (15.35), as in Figure 1. Concen-
trations of fragments 1,3-cyclohexadiene and propene were
determined by difference compared to a time zero sample for
each kinetic run.
Acknowledgment is made to the donors of the
American Chemical Society Petroleum Research Fund
and the Franklin & Marshall College Hackman Scholars
Program for support of this research. D.C.P. received
support through a Schappell Research Scholarship in
the summer of 2005. We are grateful to Prof. John E.
Baldwin of Syracuse University for performing a critical
reading of the manuscript.
Supporting Information Available: Copies of first-order
kinetic rate plots for 1a and 1b; NMR spectra for 1a and 1b;
tables of mole fractions for thermal runs of 1a, 1b, and a
mixture of 2a and 2b; low energy conformations for 1a and
1b generated from Spartan ′04 for Macintosh. This material
is available free of charge via the Internet at http://pubs.acs.org.
JO051505G
8918 J. Org. Chem., Vol. 70, No. 22, 2005