Deuterium kinetic isotope effects and mechanism of the thermal isomerization of bicyclo[4.2.0]oct-7-ene to 1,3-cyclooctadiene

Article abstract of DOI:10.1021/jo040220l

The thermal conversion of cis-bicyclo[4.2.0]oct-7-ene to cis,cis-1,3-cyclooctadiene might involve a direct disrotatory ring opening, or it might possibly take place by way of cis,trans-1,3-cyclooctadiene. This cis,trans-diene might possibly form the more stable cis,cis isomer through a [1,5] hydrogen shift or a trans-to-cis isomerization about the trans double bond. Deuterium kinetic isotope effect determinations for the isomerizations of 2,2,5,5-d₄-bicyclo[4.2.0]oct-7-ene and 7,8-d₂-bicyclo[4.2. 0]-oct-7-ene rule out these two alternatives because the observed effects are much smaller than would be anticipated for these mechanisms: k _H/k_D(d₄) at 250 °C is 1.17 (1.04 per D), and k_H/k_D(d₂) at 238 °C is 1.20 (1.10 per D). The direct disrotatory ring opening route remains the preferred mechanism.

Full text of DOI:10.1021/jo040220l

Deu ter iu m Kin etic Isotop e Effects a n d Mech a n ism of th e Th er m a l
Isom er iza tion of Bicyclo[4.2.0]oct-7-en e to 1,3-Cycloocta d ien e
,
†
‡
,‡
†
J ohn E. Baldwin,* Sarah S. Gallagher, Phyllis A. Leber,* Anuradha S. Raghavan, and
†
,§
Rajesh Shukla
Department of Chemistry, Syracuse University, Syracuse, New York 13244, and
Department of Chemistry, Franklin & Marshall College, Lancaster, Pennsylvania 17604
jbaldwin@syr.edu; phyllis.leber@fandm.edu
Received J une 23, 2004
The thermal conversion of cis-bicyclo[4.2.0]oct-7-ene to cis,cis-1,3-cyclooctadiene might involve a
direct disrotatory ring opening, or it might possibly take place by way of cis,trans-1,3-cyclooctadiene.
This cis,trans-diene might possibly form the more stable cis,cis isomer through a [1,5] hydrogen
shift or a trans-to-cis isomerization about the trans double bond. Deuterium kinetic isotope effect
determinations for the isomerizations of 2,2,5,5-d
oct-7-ene rule out these two alternatives because the observed effects are much smaller than would
be anticipated for these mechanisms: k /k (d ) at 250 °C is 1.17 (1.04 per D), and k /k (d ) at 238
C is 1.20 (1.10 per D). The direct disrotatory ring opening route remains the preferred mechanism.
4 2
-bicyclo[4.2.0]oct-7-ene and 7,8-d -bicyclo[4.2.0]-
H
D
4
H
D
2
°
In tr od u ction
SCHEME 1. In d ir ect Th er m a l Isom er iza tion of 1
to 2 by Wa y of 3
cis-Bicyclo[4.2.0]oct-7-ene (1) was first prepared by
Vogel in 1953 through a short route from cycloocta-
1
tetraene. It soon became more easily accessible through
2
the photoisomerization of cis,cis-1,3-cyclooctadiene (2),
a route more convenient than other options.3
6
favored by Branton and co-workers; however, other mecha-
nistic possibilities were recognized. Criegee and co-
4
7
workers pointed out that cis,trans-1,3-cyclooctadiene (3)
might be the primary product and reaction intermediate.
Bloomfield and McConaghy suggested that the reaction
path might involve the reversible formation of 3 as an
intermediate followed by a rate-determining isomeriza-
tion of 3 to 2 through a [1,5] hydrogen shift. Others
considered the possibility that intermediate 3 might lead
The thermal isomerization of 1 to 2 was reported in
4
1
965, and careful gas-phase kinetic work was published
5
the following year.
8
,9
to the final product 2 through rotation about the trans
olefinic bond.1
0,11
The reactions of Scheme 1 frame these
alternatives featuring cis,trans-diene 3 as an intermedi-
ate along the reaction coordinate. Here, k32 might be
The direct conversion of 1 to 2 could take place through
a disrotatory ring opening, a mechanistic formulation
(6) Branton, G. R.; Frey, H. M.; Montague, D. C.; Stevens, I. D. R.
Trans. Faraday Soc. 1966, 62, 659-663.
*
Author to whom correspondence should be addressed.
Syracuse University.
Franklin & Marshall College.
(7) (a) Cope, A. C.; Bumgardner, C. L. J . Am. Chem. Soc. 1956, 78,
2812-2815. (b) Shumate, K. M.; Neuman, P. N.; Fonken, G. J . J . Am.
Chem. Soc. 1965, 87, 3996. (c) Isaksson, R.; Roschester, J .; Sandstr o¨ m,
J .; Wistrand, L.-G. J . Am. Chem. Soc. 1985, 107, 4074-4075. (d)
Bouman, T. D.; Hansen, A. E. Croat. Chem. Acta 1989, 62, 227-243.
(e) Inoue, Y.; Tsuneishi, H.; Hakushi, T.; Tai, A. J . Am. Chem. Soc.
1997, 119, 472-478. (f) Yavari, I.; Kabiri-Fard, H.; Moradi, S. J . Mol.
Struct.: THEOCHEM 2003, 623, 237-244.
†
‡
§
Present address: Borregaard Synthesis, Inc., 9 Opportunity Way,
Newburyport, MA 01950.
1) (a) Vogel, E. Angew. Chem. 1953, 65, 346-347. (b) Allinger, N.
L.; Miller, M. A.; Tushaus, L. A. J . Org. Chem. 1963, 28, 2555-2557.
2) (a) Chappell, S. F.; Clark, R. F. Chem. Ind. (London) 1962, 1198.
b) Dauben, W. G.; Cargill, R. L. J . Org. Chem. 1962, 27, 1910. (c)
Srinivasan, R. J . Am. Chem. Soc. 1962, 84, 4141-4145.
3) (a) Ball, W. J .; Landor, S. R. Proc. Chem. Soc. 1961, 143-144.
b) Kirmse, W.; Pook, K. H. Chem. Ber. 1965, 98, 4022-4026.
4) Criegee, R.; Seebach, D.; Winter, R. E.; B o¨ rretzen, B.; Brune, H.
A. Chem. Ber. 1965, 98, 2339-2352.
5) Branton, G. R.; Frey, H. M.; Skinner, R. F. Trans. Faraday Soc.
966, 62, 1546-1552.
(
(
(
(8) McConaghy, J . S., J r.; Bloomfield, J . J . Tetrahedron Lett. 1969,
3719-3721.
(
(9) Bloomfield, J . J .; McConaghy, J . S., J r. Tetrahedron Lett. 1969,
3723-3726.
(
(
(10) Padwa, A.; Koehn, W.; Masaracchia, J .; Osborn, C. L.; Trecker,
D. J . J . Am. Chem. Soc. 1971, 93, 3633-3638.
(
(11) Bramham, J .; Samuel, C. J . J . Chem. Soc., Chem. Commun.
1989, 29-30.
1
1
0.1021/jo040220l CCC: $27.50 © 2004 American Chemical Society
Published on Web 09/23/2004
7
212
J . Org. Chem. 2004, 69, 7212-7219

Thermal Isomerization of Bicyclo[4.2.0]oct-7-ene
associated with either a [1,5] hydrogen shift or a trans-
to-cis isomerization about the C3-C4 double bond of 3.
This mechanistic puzzle has been well recognized for
decades, and yet it has remained unclarified. It is
recalcitrant largely because of difficulties associated with
the relative magnitudes of rate constants that may be of
importance. Arrhenius parameters for the observable
conversion of 1 to 2 determined by Branton, Frey, and
SCHEME 2. Alter n a tives for
4
2,2,5,5-d -Bicyclo[4.2.0]oct-7-en e (1-d )
4
5
Skinner are given in eq 1. Bloomfield and McConaghy
29.0 kcal/mol) is 1.9 × 10 s-1, which is 160 times faster
-2
followed the isomerization of 3 to 1 over the temperature
range of 64-90 °C and derived the Arrhenius parameters
shown in eq 2.9
than k, the observed rate constant for the conversion of
15,16
1 to 2.
The mechanistic issues formalized in Scheme 1 have
long remained unresolved. While the indirect mechanism
from 1 to 2 by way of intermediate 3 “has much to
recommend it, the possibility that the isomerization of
log k ) 14.13 - 43180/(2.303RT)
log k31 ) 13.14 - 27850/(2.303RT)
(1)
(2)
1
2
[1 to 2] may proceed directly cannot be ruled out.” The
Temperature-dependent equilibrium constants for [1]/[3]
from 89.8 to 170.0 °C have been obtained, and together
hydrogen shift mechanism seems “permissible, if not
operative”, but it would require a small positive ∆S ,
hardly an expected characteristic.18
The most recent contribution toward a sure mechanis-
tic understanding of the isomerization of 1 to 2 reported
that cis,trans-1,3-cyclooctadiene labeled with one deute-
9
17
q
with the parameters of eq 2, they allow one to calculate
log k13 as a function of temperature (eq 3).12
log k13 ) 13.3 - 33450/(2.303RT)
(3)
rium at C5 or C8 (3-d) isomerized in a dilute CCl
4
The kinetic situation relevant to the mechanistic
problem is simple enough. Exact integrated mole fraction
versus time expressions corresponding to Scheme 1 are
well-known,13 and when, as here, a steady-state treat-
solution at reflux (bp 77 °C) under nitrogen to give 76%
of 1-d and 24% of 2-d. The deuterium label in 1-d was
found at C2, while the label in 2-d was exclusively at
C5. This finding seemed to rule out a mechanism for the
isomerization of 3 to 2 through a [1,5] hydrogen shift.
When a binary mechanistic choice was assumed, these
results led to the conclusion that a trans-to-cis rotation
ment is appropriate (k31 . k32), one has kobs ) k ) (k13
/
k
31)k32 (if k12 ) 0). If the mechanism of Scheme 1 is
correct, the Arrhenius equations of eqs 1-3 lead directly
to one for k32 (eq 4).12
11
must occur. Clearly, the deuterium-labeled cis,cis-1,3-
cyclooctadiene product did not form directly from 3-d
log k32 ) 14.0 - 37600/(2.303RT)
(4)
through a hydrogen shift.
The difficulties are posed not by kinetic complexity but
rather by the relative magnitudes of k13, k31, k12, and k32
.
At 250 °C, for example, the Arrhenius expressions in eqs
-
4
-1
1
2
-4 give k ) 1.2 × 10
s
for the isomerization of 1 to
, while the rate constant (k13 + k31) for the approach to
-
1
equilibrium between 1 and 3 is about 32 s , or 2.7 ×
The present work approached this long-standing mecha-
nistic problem anew.19 Distinctions among the three
alternatives (a direct disrotatory 1-to-2 process, a 3-to-2
reaction involving a [1,5] hydrogen shift, and a 3-to-2
path through rotation about the trans double bond) were
sought. These three mechanisms should be associated
with distinctive deuterium kinetic isotope effects, given
appropriate specific labeling of the bicyclooctene reactant.
Were the starting material labeled with deuterium
atoms at C2 and C5, as in Scheme 2, the k′32 reaction
would involve a primary deuterium kinetic isotope effect
and be expected, due to relevant precedents, to be
unmistakably primary. Relevant precedents are avail-
able; the [1,5] hydrogen shifts in 1,3,6-cyclooctatriene and
5
1
k
0 times faster than k ) k12 (if k32 ) 0) or (k13/k31)k32 (if
12 ) 0). Thus, whether 1 or 3 is the starting material,
the initial rate of formation of 2 will be identical, within
plausible analytical limits. The kinetic question of whether
2
is formed from 1 or from 3 cannot be answered readily
through kinetic approaches. Distinctions dependent on
detecting different initial rates of product formation
starting from different initial reactants14 would, in the
present instance, be extremely demanding and impracti-
cal.
In addition, simple isotopic labeling studies dependent
upon detecting just where a label that appears initially
in 1 appears in thermal product 2 cannot resolve the
mechanistic question. The [1,5] hydrogen shifts in diene
2
will redistribute label(s) at a rate orders of magnitude
(
15) Glass, D. S.; Boikess, R. S.; Winstein, S. Tetrahedron Lett. 1966,
faster than the rate at which 1 gives 2 through any
mechanism. At 250 °C, for example, the rate constant
for one [1,5] hydrogen migration in 2 (log A ) 10.4, E )
9
5
99-1008.
(16) Baldwin, J . E.; Leber, P. A.; Lee, T. W. J . Org. Chem. 2001, 66,
269-5271.
a
(17) Marvell, E. N. Thermal Electrocyclic Reactions; Academic
Press: New York, 1980; pp 171-173.
(
(
12) Cocks, A. T.; Frey, H. M. J . Chem. Soc. B 1970, 952-954.
13) For example, see: McDaniel, D. H.; Smoot, C. R. J . Phys. Chem.
(18) Gajewski, J . J . Hydrocarbon Thermal Isomerizations; Academic
Press: New York, 1981; pp 281-282.
1
956, 60, 966-969.
14) Baldwin, J . E.; Kaplan, M. S. J . Am. Chem. Soc. 1972, 94,
696-4699.
(19) A preliminary account of this work has appeared: Baldwin, J .
E.; Gallagher, S. S.; Leber, P. A.; Raghavan, A. Org. Lett. 2004, 6,
1457-1460.
(
4
J . Org. Chem, Vol. 69, No. 21, 2004 7213

Baldwin et al.
SCHEME 3. Alter n a tives for
2
,8-d -Bicyclo[4.2.0]oct-7-en e (1-d )
SCHEME 4. Con ver sion of Keton e 4 to
Bicyclo[4.2.0]oct-7-en e
7
2
SCHEME 5. P r ep a r a tion of
2
4
,2,5,5-d -Bicyclo[4.2.0]oct-7-en e
H D
in cis-1,3-pentadiene are characterized by k /k values
of 5.0 at 120 °C and of 5.1 at 200 °C, respectively. In
the alternative rate-determining disrotatory process
2
0
(k′12), modest â secondary deuterium kinetic isotope
effects might be seen. These kinetic isotope effects would
reflect structural changes at the transition-state region
of the potential energy surface; they would not be
obscured by subsequent rapid [1,5] hydrogen shifts
equilibrating all possible deuterium-labeled cis,cis-1,3-
cyclooctadienes.
2
SCHEME 6. 7,8-d -Bicyclo[4.2.0]oct-7-en e fr om
Keton e 4
Deuterium labels at C7 and C8 of the starting material,
as shown in Scheme 3, would become labels at C2 and
C3 of the monocyclic cis,trans-diene. The label on the
2
trans double bond in 3-d would be expected to show a
substantial kinetic isotope effect if the product were
formed through a k′′32 path with a trans-to-cis isomer-
ization about the C3-C4 double bond but not if the
reaction took place through the direct k′′12 route.
GC isolation, gave an authentic sample of 1.2,25 It was
also secured from the [2 + 2] cycloadduct of cyclohexene
and dichloroketene, 8,8-dichlorobicyclo[4.2.0]octan-7-one
The thermal isomerizations of trans-1-phenylcyclohex-
ene21 and trans-1-phenyl-2-d-cyclohexene to the respec-
2
6
(4), using the route outlined in Scheme 4 as a run-
tive cis isomers are characterized by k
H
/k
D
effects of 2.0
are
at room temperature.22 Certainly, dienes 3 and 3-d
through for analogous sequences affording labeled ana-
2
2
7
2
3
logues. When the methods of Greene and co-workers
and syntheses of various deuterium-labeled bicyclo[3.2.0]-
less strained than these trans-cyclohexenes, and a
possible cis,trans-to-cis,cis thermal isomerization of 3 and
2
8
hepta-2,6-dienes developed by Belfield were followed,
ketone 4 was reduced with NaBH and the epimeric
3
2
-d would take place with a higher activation energy.
4
There is precedent for large secondary deuterium kinetic
isotope effects on thermally activated rotations about the
double bond of alkenes, even at elevated temperatures.
mixture of alcohols produced was converted to the cor-
responding mesylates (5). Reduction of the mesylates
with sodium in liquid ammonia afforded the bicyclic olefin
For cis-2-butene (d
.58 (or 1.87 when corrected for tunneling) according to
MCSCF calculations; for cis-stilbene (d - and R,R′-d at
) ) 1.47 ( 0.13, as determined experi-
/k effect might be anticipated
0 2 H D 2
and 2,3-d at 300 °C), k /k (d ) )
1
, identical to the authentic reference sample.
Starting the reaction with 3,3,6,6-d -cyclohexene, the
cycloaddition with dichloroketene generated from trichlo-
1
4
0
2
2
87 °C), k
H
4
/k
D
(d
2
2
9
2
roacetyl chloride and zinc in ether (with sonication) gave
-d (Scheme 5). This intermediate was converted to 1-d
4
mentally. Thus, a large k
H
D
4
4
if k′′32 were rate limiting (Scheme 3). If k′′12 were rate
limiting, the experimentally observable â secondary
using the three-step sequence outlined in Scheme 5. After
isolation and purification of the mixture by preparative
k
H
/k
possible changes in geometry or hybridization or hyper-
conjugation in the transition structure leading from 1-d
D
effect would be expected to be small, reflecting
4
GC, the 1-d was characterized by NMR spectroscopy and
2
mass spectrometry. The H NMR spectrum showed the
2
two types of deuterium atoms as singlets at δ 1.67 and
2
to the cis,cis cyclic diene 2-d .
1
.48; the broad-band proton-decoupled carbon-13 spec-
trum had the expected resonances at δ 140.5, 41.3, and
8.6.
A modification of the reaction sequences used in
Schemes 4 and 5 led to the 7,8-d -labeled bicyclooctene
1-d (Scheme 6). Reduction of 4 with zinc in the presence
Resu lts
1
Syn th eses. Unlabeled cis-bicyclo[4.2.0]oct-7-ene (1)
was prepared through two routes. Photoisomerization of
cis,cis-1,3-cyclooctadiene (2) in heptane in the presence
of acetophenone as a sensitizer, followed by preparative
2
2
of O-deuterioacetic acid gave
a deuterium-labeled
(
20) (a) Kloosterziel, H.; ter Borg, A. P. Recl. Trav. Chim. Pays-Bas
(25) Adam, W.; Alt, C.; Braun, M.; Denninger, U.; Zang, G. J . Am.
Chem. Soc. 1991, 113, 4563-4571.
1
6
965, 84, 1305-1314. (b) Roth, W. R.; K o¨ nig, J . Liebigs Ann. 1966,
99, 24-32. (c) Roth, W. R. Chimia 1966, 20, 229-236.
(26) (a) Brady, W. T.; Waters, O. H. J . Org. Chem. 1967, 32, 3703-
3705. (b) Ghosez, L.; Montaigne, R.; Roussel, A.; Vanlierde, H.; Mollet,
P. Tetrahedron 1971, 27, 615-633. (c) Harding, K. E.; Trotter, J . W.;
May, L. M. J . Org. Chem. 1977, 42, 2715-2719.
(27) Greene, A. E.; Luche, M. J .; Serra, A. A. J . Org. Chem. 1985,
50, 3957-3962.
(21) Bonneau, R.; J oussot-Dubien, J .; Salem, L.; Yarwood, A. J . Am.
Chem. Soc. 1976, 98, 4329-4330.
22) Caldwell, R. A.; Misawa, H.; Healy, E. F.; Dewar, M. J . S. J .
Am. Chem. Soc. 1987, 109, 6869-6870.
23) Goodman, J . L.; Peters, K. S.; Misawa, H.; Caldwell, R. A. J .
Am. Chem. Soc. 1986, 108, 6803-6805.
24) Olson, L. P.; Niwayama, S.; Yoo, H.-Y.; Houk, K. N.; Harris, N.
J .; Gajewski, J . J . J . Am. Chem. Soc. 1996, 118, 886-892.
(
(
(28) Baldwin, J . E.; Belfield, K. D. J . Org. Chem. 1987, 52, 4772-
(
4776.
(29) Mehta, G.; Rao, H. S. P. Synth. Lett. 1985, 15, 991-1000.
7
214 J . Org. Chem., Vol. 69, No. 21, 2004

Thermal Isomerization of Bicyclo[4.2.0]oct-7-ene
TABLE 1. Mole P er cen t Rela tive Con cen tr a tion Da ta
for Bicyclo[4.2.0]oct-7-en e (1) a n d
Least-squares best-fit calculations provided the unimo-
lecular rate constant for the isomerization of 1 at 238.2
7
,8-d 2-Bicyclo[4.2.0]oct-7-en e (1-d 2) a t 238.2 °C a s
-5 -1
(
0.8 °C (k ) 5.47 × 10
and 1-d
-d
(t)] ) 72 × exp(-kR t) + 28 × exp(-kRt). The least-
squares best-fit value of R ) (k /k per deuterium was
.912; k /k ) 1.10 per deuterium.
The imperfect temperature control experienced during
s
). The isomerizations of 1-d
2
F u n ction s of Rea ction Tim e
1
at 238.2 ( 0.1 °C were modeled with [1-d
2
(t) +
time (s)
0
1 (%)
time (s)
1-d2 (%)
2
1
1
-1
100.0
81.3
0
3600
6060
10800
14400
57600
100.0
84.4
75.1
59.9
50.6
8.8
D
)
H
3
3
7
9
600
600
200
000
0
H
D
a
a
81.7
68.6
60.1
50.3
43.1
38.6
27.8
8.7
the kinetic runs with 1 was annoying but not of telling
consequence, for its impact on the rate constant deter-
mination should be very small. Whatever reasonable
likely errors in the derived rate constants one may
postulate, the essential finding remains clear. The ob-
1
1
1
2
4
7
2600
4400
8000
5200
2000
2000
served k
short of the magnitude anticipated for a trans-to-cis
rotation transforming 3-d to 2-d . We conclude that the
H D 2
/k (d ) ratio, 1.20 (1.10 per deuterium), falls far
1.9
a
Duplicate kinetic run.
2
2
k′′32 path plays no significant role in the overall isomer-
ization. This conclusion follows from the observed
monochloride, which was reduced with NaBD
hols obtained were treated with methanesulfonyl chloride
to afford 5-d . Reduction with Na and NH /THF gave 7,8-
-bicyclo[4.2.0]oct-7-ene (1-d ).
The proton NMR spectrum of the monochloroketone
4
; the alco-
k
H
/k
D 2
(d ) and the rationale leading from reports in the
2
1-24
literature
to an anticipated isotope effect of 2 or so
2
3
for the k′′32 reaction. If that rationale were seriously off
the mark, the conclusion would be unwarranted. Further
experimental and theoretical effort would be required to
address this caveat.
The kinetic runs leading to an experimental definition
H D 4
of a k /k (d ) kinetic isotope effect were different in
several respects: a different bath temperature was
deliberately selected, 2,2-dimethylbutane was used as the
4
bath gas, and both 1 and 1-d samples were run simul-
taneously, thus eliminating any possibility that the
samples might be isomerized at different temperatures
or under different reaction conditions of any sort. The
original plan projected analyses based on both capillary
GC and GC/MS. Analyses by GC would have given
d
2
2
formed through the reduction of 4 with zinc and O-
deuterioacetic acid showed a distinctive doublet of dou-
blets from C8-exo-H at δ 4.95 (J ) 2.2 and 8.8 Hz),
revealing that the product included some unlabeled
3
0
8
-endo-chlorobicyclo[4.2.0]octan-7-one (6). Adventitious
moisture apparently contaminated the AcOD that was
utilized or reflected inadequately dried glassware. The
relative intensities of the strong MS ions at m/z 124
+
+
(C
8
H
10DO ) and 123 (C
8
H
1
11O ), 78:22, provided a quan-
:d product ratio.
titative estimate of the d
0
relative concentrations of [1 + 1-d
4
4
] and [2 + 2-d ] at
various reaction times, and the ratios of the isotopomers
of each structure at each reaction time would have been
followed from relative ion intensities at masses charac-
teristic of each species. In practice, GC proved to be
After the reduction with NaBD
the alcohols obtained the mesylate esters (5-d
esters were examined by mass spectrometry. The MS m/z
4
and the conversion of
2
), the
4
sufficient since the unlabeled and d -labeled starting
materials and products separated well enough to give
reliable integrated intensities through flame ionization
+
+
1
2
09:108 (C
8 9 2 8
H D :C H10D ) ion intensity ratio was 79:
detection.19 The retention times of the unlabeled and d
-
1, suggestive of essentially quantitative incorporation
4
of deuterium through the borodeuteride reduction. When
-d was in hand, another estimate of the isotopic
labeled bicyclooctene and cis,cis-diene isomers at 40 °C
differed by 10.5 and 12.2 s, respectively. In each pair of
1
2
integrity of the synthetic route was obtained. Integration
of the H NMR spectrum of a GC purified sample showed
δ 6.10 (vinyl H) and 2.83 (bridgehead H) relative intensi-
1 2
ties to be 0.14:1.00, consistent with 28% d and 72% d
4
isotopomers, the d -labeled sample had the shorter
1
19
retention time, as determined by GC/MS. Precedents
for separations of deuterium-labeled and unlabeled ver-
sions of the same structure have been reported in the
literature, and the labeled variants usually elute ear-
lier.32
labeling.
Kin etic Resu lts. Kinetic runs were conducted using
a gas-phase static reactor, an associated vacuum system,
and temperature-control and measurement instruments
that have been detailed elsewhere.31 The isomerizations
The kinetic data secured for runs starting from a 54:
46 mixture of 1 and 1-d at 249.9 ( 0.1 °C in the presence
4
of 2,2-dimethylbutane as a bath gas and cyclooctane as
an internal standard are recorded in Table 2.
of 1 and of 1-d
and decane as an internal standard; product mixtures
were analyzed by capillary GC. Cyclic dienes 2 and 2-d
2
were run with pentane as the bath gas
From these data, least-squares best-fit calculations
-
4
-1
2
provided overall rate constants of k ) 1.16 × 10
s
-
4
-1
were the only products observed. The mole percent
concentrations versus time data are given in Table 1.
and k(d ) ) 0.99 × 10
s . The ratio k /k (d ) ) 1.17;
k /k ) 1.04 per deuterium. The substantial primary k /
4
H
D
4
H
D
H
k
D
effect anticipated for a rate-determining [1,5] hydrogen
to 2-d was not observed.
(
30) Wakamatsu, T.; Miyachi, N.; Ozaki, F. Heterocycles 1987, 26,
shift converting 3 to 2 and 3-d
4
4
1
445-1448.
(
31) (a) Baldwin, J . E.; Burrell, R. C. J . Org. Chem. 1999, 64, 3567-
3
3
571. (b) Baldwin, J . E.; Burrell, R. C. J . Org. Chem. 2002, 67, 3249-
256.
(32) Angelis, Y. S.; Orfanopoulos, M. J . Org. Chem. 1997, 62, 6083-
6085.
J . Org. Chem, Vol. 69, No. 21, 2004 7215

Baldwin et al.
TABLE 2. Mole P er cen t Rela tive Con cen tr a tion Da ta
for Bicyclo[4.2.0]oct-7-en e (1) a n d
2
,2,5,5-d 4-Bicyclo[4.2.0]oct-7-en e (1-d 4) a t 249.9 °C a s
F u n ction s of Rea ction Tim e
time (s)
0
1 (%)
1-d4 (%)
100.0
77.5
53.2
36.1
16.8
100.0
83.9
60.5
41.0
19.0
2
5
9
192
112
000
1
6041
Discu ssion
From the present work, it appears that the cis,trans
isomer of cyclooctadiene (3) is not an intermediate on the
mechanistic path leading from cis-bicyclo[4.2.0]oct-2-ene
(
1) to cis,cis-1,3-cyclooctadiene (3), for neither a direct
F IGURE 1. Correlation of ∆H
cis-3,4-(CH
) 1), bicyclo[2.2.0]hex-2-ene (n ) 2), bicyclo[3.2.0]hept-6-ene
n ) 3), and cis-bicyclo[4.2.0]oct-7-ene (1; n ) 4), with E for
r
for thermal isomerizations of
) -bridged cyclobutenes, bicyclo[2.1.0]pent-2-ene (n
2 n
trans-to-cis rotation taking 3-d
deuterium shift leading from 3-d
with the kinetic isotope effects observed. The k /k values
H D
found are completely consistent with expectations for â
secondary deuterium kinetic isotope effects and direct
2
to 2-d
to 2-d
2
nor a [1,5]
4
is consistent
4
(
a
the reactions. For n ) 1-3 isomerizations (9), the linear
relationship shows that more exothermic reactions have lower
activation energies. For the n ) 4 case (1-2; b), the extra-
4 4 2 2
reactions converting 1 to 2, 1-d to 2-d , and 1-d to 2-d .
polated E
determined E value (43.18 kcal/mol)
a
value (46.9 kcal/mol) and the experimentally
are noticeably distinct.
Among the roster of postulated mechanisms advanced
in the literature, only the direct disrotatory electrocyclic
reaction remains viable. How does it fit with related
reactions shown by other bicyclic cyclobutenes?
This question played an interesting role in the early
developments of orbital symmetry theory. The thermal
isomerizations of monocyclic hydrocarbons based on
cyclobutene substructures take place at relatively low
temperatures and invariably prefer to proceed in a
conrotatory fashion. However, geometrically restricted
bicyclic systems, such as 6,7-dimethylbicyclo[3.2.0]hept-
5
a
hydrocarbons rearrange through conrotatory ring open-
ings. The trend was summarized as follows, giving the
temperature (°C) at which the half-life for the electrocy-
clic isomerization is about 2 h, for n ) 1 (bicyclo[2.1.0]-
pent-2-ene), n ) 2 (bicyclo[2.2.0]hex-2-ene), n ) 3 (bicyclo-
[3.2.0]hept-6-ene), n ) 4 (bicyclo[4.2.0]oct-7-ene), n ) 5
(bicyclo[5.2.0]non-8-ene), and n ) 6 (bicyclo[6.2.0]dec-9-
ene).35
n
1
2
3
4
5
6
6
-ene, isomerize much more slowly.33 These experimental
T (°C)
<100
195
>380
350
335
180
observations were an instrumental stimulus for the
intellectual and pattern-recognition insights associated
with Woodward and Hoffmann’s breakthrough first com-
munication on the stereochemistry of electrocyclic reac-
tions.34 That paper includes a clear perspective on the
relevance of the reactivity differences noted for bicyclic
systems restricted to disrotatory isomerizations or some
other less-direct mechanism: “It should be emphasized
that our hypothesis specifies in any case which of two
types of geometrical displacements will represent a
favored process, but does not exclude the operation of the
other under very energetic conditions. ... In [6,7-
dimethylbicyclo[3.2.0]hept-6-ene], the presence of the
five-membered ring makes a conrotatory process impos-
sible, and the disrotatory opening is observed, but only
slowly at 400°.” In the later review article and book on
The Conservation of Orbital Symmetry, these authors
emphasized that cis-fused bicyclo[n.2.0]alkenes similar
to bicyclo[4.2.0]oct-7-ene react to give monocyclic cis,cis-
dienes at rates dependent on reaction thermochemistry.35
The less-strained bicyclic olefins isomerize to give cis,cis-
diene products more slowly than similar monocyclic
Careful gas-phase kinetic studies on the n ) 1-4
systems have now been published, and this qualitative
presentation of the unmistakable and clearly important
influence of structure on reactivity can be revised as
5,6,9,36
follows:
n
1
2
3
4
5
6
T (°C)
50
130
270
250
235
225
When n ) 5 or 6, the conrotatory electrocyclic process
leading to an equilibrium between bicyclic cyclobutene
and monocyclic cis,trans-diene dominates, yet the overall
isomerization rates do not reflect a comparable advan-
36,37
tage.
Attempts to correlate overall reaction rates with the
heats of reaction for these isomerizations have been made
3
4
in several ways, including tabulations of E
a
and ∆H
r
38
values. Using thermochemical and kinetic data better
than those available to Lupton in 1968, the correlation
shown in Figure 1 is obtained. The activation energies
for the n ) 1-4 bicyclo[n.2.0] systems are 26.9, 33.0, 45.5,
(
36) (a) Schumate, K. M. Ph.D. Dissertation, University of Texas,
(
33) (a) Criegee, R.; Zirngibl, U.; Furrer, H.; Seebach, D.; Freund,
Austin, TX, 1966. Schumate, K. M. Electrocyclization of Some Conju-
gated Medium Ring Dienes. Diss. Abstr. B 1967, 27, 3466-3467. (b)
Brauman, J . I.; Golden, D. M. J . Am. Chem. Soc. 1968, 90, 1920-1921.
(c) Brauman, J . I.; Golden, D. M. Trans. Faraday Soc. 1969, 65, 464-
469. (d) Goldstein, M. J .; Leight, R. S.; Lipton, M. S. J . Am. Chem.
Soc. 1976, 98, 5717-5718.
G. Chem. Ber. 1964, 97, 2942-2948. (b) Criegee, R.; Furrer, H. Chem.
Ber. 1964, 97, 2949-2952.
(34) Woodward, R. B.; Hoffmann, R. J . Am. Chem. Soc. 1965, 87,
3
95-397.
(35) (a) Woodward, R. B.; Hoffmann, R. Angew. Chem., Int. Ed. Engl.
1
969, 8, 781-853. (b) Woodward, R. B.; Hoffmann, R. The Conservation
of Orbital Symmetry; Verlag Chemie: Weinheim, Germany, 1971.
(37) Radlick, P.; Fenical, W. Tetrahedron Lett. 1967, 4901-4904.
(38) Lupton, E. C. Tetrahedron Lett. 1968, 4209-4212.
7
216 J . Org. Chem., Vol. 69, No. 21, 2004

Thermal Isomerization of Bicyclo[4.2.0]oct-7-ene
and 43.2 kcal/mol, respectively;36 the corresponding ∆H
values are -47.8, -36.0, -10.8, and -8.7 kcal/mol.39
quantitatively (or nearly quantitatively).7b,9,41 An excep-
tion was noted by Padwa and co-workers in the course
of their study of the thermal bimolecular combination of
r
The linear correlation for the bicyclics with n ) 1-3
would extrapolate to an E value of 46.9 kcal/mol for 1,
1
0
3
to form mixtures of three isomeric [2 + 2] dimers.
a
5
Under the reaction conditions (concentrated mixture of
cis,trans-diene and diethyl phthalate as an internal
standard), both bicyclooctene 1 and diene 2 were detected
as minor, nondimeric products. However, the proportions
of 1 and 2 were not constant over reaction time, nor were
they consistent from run to run. The cis,cis-diene might
well have derived from a bimolecular path involving the
formation of a transient diradical derived from two
molecules of 3.10
whereas the experimental value is 43.18 kcal/mol. This
difference may signal that another influence begins to
contribute to the overall trend when the isomerizations
become less dramatically exothermic as the starting
materials become less strained (from left to right in
a
Figure 1). However, the difference in E values between
bicyclo[3.2.0]hept-6-ene and 1 is only 2.3 kcal/mol, and
5
both probably occur through a common mechanism. This
plausibility consideration, combined with the present
In Bramham and Samuel’s work,11 the 5 (or 8)-d-
cis,trans-1,3-cyclooctadiene was isolated by an aqueous
silver nitrate extraction following the work of Cope and
Bumgardner;7 the complex, in a mixture of ice and
water, was treated with ammonia and extracted with
carbon tetrachloride. The organic solution was either
passed through a short column of activity I alumina or
dried over magnesium sulfate and filtered to remove the
residual silver complex.42
demonstration of a k
H D 4
/k (d ) ratio inconsistent with a
[1,5] hydrogen shift, brings one back to the break in the
linear correlation of Figure 1. One possibility is that the
additional molecular flexibility in the C8 homologue
relative to the smaller bicyclic cyclobutenes facilitates the
disrotatory ring-opening reaction by allowing access to
more geometrically favorable transition structures. This
hypothesis implies that the transition structure for the
a
s
disrotatory isomerization lacks C symmetry, a testable
proposition.
Were even extremely minor traces of the silver ion-
cis,trans-diene complex to survive these precautions, the
formation of a cis,cis product at 77 °C could be easily
rationalized and the observed labeling outcome would be
consistent with expectations. It is of course notoriously
difficult to obviate, with certainty, the intervention of
With regard to the activation parameters derived for
the rate constant k32, are they really plausible for a [1,5]
hydrogen shift reaction? The E value, 37.6 kcal/mol, is
a
clearly within a reasonable range, but the preexponential
factor is much too high. The log A term (14.0) translates
q
43
to ∆S ) +2.4 eu at 250 °C, corresponding to a “loose”
catalytically active trace impurities. The absence of a
transition structure, whereas [1,5] hydrogen shifts typi-
cis,cis product from 3 in kinetically monitored first-order
q
7b,9,41
cally have substantial negative ∆S values. For cis-1,3-
reactions
is strong evidence that the direct isomer-
q
q
pentadiene, ∆S ) -12 eu, and for diene 2, ∆S ) -10
eu. The symmetry-allowed suprafacial [1,5] hydrogen
shift has a relatively low activation energy but very
demanding geometric preferences in the transition re-
gion.40
ization of 3 to 2 is not kinetically competitive with the
well-established 3-to-1 isomerization. Small amounts of
catalytically active impurities could not account for the
clean kinetic results for the 3-to-1 reaction reported from
three laboratories. Given the possibility of silver ion
catalysis as a contributing factor in the isomerizations
In hindsight, another strike against the 3-to-2 [1,5]
hydrogen shift mechanism for the overall 1-to-2 isomer-
ization follows from a careful consideration of the geo-
metrical limitations inherent in the cis,trans-diene 3. The
most favored conformation of this highly skewed diene
locates the end of the trans double bond (C4) closest to
the hydrogens bonded to C8 that might be available for
the shift in a disposition limiting the shift to an ant-
arafacial outcome. This limitation is evident from a close
examination of the stereoview along the C4-C3 bond of
reported,11 and the present k
H D 2
/k (d ) determination, the
case for a direct trans-to-cis isomerization from 3 to 2
does not seem persuasive.
Con clu sion s
The deuterium kinetic isotope effect determinations
reported here are inconsistent with mechanistic formula-
tions of the thermal isomerization of cis-bicyclo[4.2.0]oct-
7-ene (1) to cis,cis-1,3-cyclooctadiene (2) through an
indirect route involving cis,trans-1,3-cyclooctadiene (3).
The mechanism consistent with the isotope effects that
were seen, one dependent on a disrotatory ring-opening
process leading from 1 to 2 directly, takes place in
competition with the symmetry-allowed conrotatory
isomerization leading reversibly to cis,trans-diene 3. The
activation energies for the disrotatory and conrotatory
electrocyclic reactions shown by cis-bicyclo[4.2.0]oct-7-ene
7
c
3
, provided by Isaksson and co-workers, or through a
direct consideration of molecular models. The [1,5] hy-
drogen shift hypothesis does not provide an alluring
dodge for avoiding a “forbidden” reaction in a mechanistic
formulation; it merely transposes the forbidden event to
another topology.
What of the evidence derived from experiments with
a deuterium-labeled version of 3 reported by Bramham
and Samuel?11 Three groups have previously reported
(43.2 and 33.5 kcal/mol, respectively) reflect the substan-
that 3 isomerizes at 80 °C to the bicyclic product 1
tial energetic advantage of the allowed over the forbidden
reaction. Yet the forbidden isomerization provides the
(39) (a) Conn, J . B.; Kistiakowsky, G. B.; Smith, E. A. J . Am. Chem.
Soc. 1939, 61, 1868-1876. (b) Turner, R. B.; Mallon, B. J .; Tichy, M.;
Doering, W. von E.; Roth, W. R.; Schr o¨ der, G. J . Am. Chem. Soc. 1973,
(41) Liu, R. S. H. J . Am. Chem. Soc. 1967, 89, 112-114.
(42) Professor Christopher J . Samuel, private communication,
August 19, 1999.
(43) Compare: (a) Liu, Y.; Lindner, P. E.; Lemal, D. M. J . Am. Chem.
Soc. 1999, 121, 10626-10627. (b) Liu, Y.; Lemal, D. M. Tetrahedron
Lett. 2000, 41, 599-602.
9
5, 8605-8610. (c) Roth, W. R.; Kl a¨ rner, F.-G.; Lennartz, W.-W. Chem.
Ber. 1980, 113, 1818-1829. (d) Roth, W. R.; Adamczak, O.; Breuck-
mann, R.; Lennartz, H.-W.; Boese, R. Chem. Ber. 1991, 124, 2499-
2
521.
(40) Hess, B. A.; Baldwin, J . E. J . Org. Chem. 2002, 67, 6025-6033.
J . Org. Chem, Vol. 69, No. 21, 2004 7217

Baldwin et al.
mechanistically viable route for the overall reaction,
while the allowed process simply forms a short-lived
intermediate which does not lie along the reaction
coordinate. The reversible interconversion of 1 with 3
takes place but does not lead to the reaction product,
cis,cis-diene 2.
No serious theoretical work on the isomerization of 1
to 2 appears to have been undertaken. Such an effort
would be quite valuable but would require work using
more than single-determinant methods. The geometry,
energy, and vibrational frequencies of the transition
structure would provide insights into the reaction not
accessible directly through the mechanistic definition
obtained in the present work. Similar calculations for the
disrotatory isomerization of cis-bicyclo[3.2.0]hept-6-ene
to cis,cis-1,3-cycloheptadiene could well provide a ratio-
0.60 g of 3,3,6,6-d
with zinc and trichloroacetyl chloride in ether under the
sonication conditions detailed above for the unlabeled cycload-
dition to provide 1-d
4
-cyclohexene (98 atom % D) was combined
1
4
:
H NMR δ 6.12 (s, 2H), 2.83 (s, 2H),
13
1
.56 (m, 2H), 1.35 (m, 2H); C NMR δ 140.5 (C7, C8), 41.3
2
(
C1, C6), 18.6 (C3, C4); H NMR δ 1.67 (br s, 2D), 1.49 (br s,
D).
-en d o-Ch lor o-8-d -bicyclo[4.2.0]octa n -7-on e. Ketone 4
2
8
(768 mg, 4 mmol), dissolved in 5 mL of CH COOD (99% D),
3
was added to 262 mg of zinc dust (4 mmol). The suspension
was stirred at room temperature periodically monitored by GC.
After 3 h, 32 mg of zinc (0.5 mmol) was added. After another
3
0 min, the starting material was consumed. The reaction
mixture was filtered, and the solid was washed with several
small portions of ether. The filtrate was washed with 1 N
NaHCO , water, and brine; it was then dried over MgSO and
3 4
filtered. Concentration provided 445 mg of the 8-endo-chloro
compound (70% yield) as a pale-yellow liquid. Analysis by
1
nale for the 2.3 kcal/mol difference in E
disrotatory ring-opening isomerizations of 1 and bicyclo-
3.2.0]hept-6-ene or the 3.7 kcal/mol energetic preference
a
values for the
capillary GC revealed one isomer. The H NMR spectrum of
the ketone showed residual C8-H absorptions at δ 4.95 as a
3
0
doublet of doublets (J ) 2.2 and 8.8 Hz). The intensities of
[
+
+
8 8
the strong MS ions at m/z 124 (C H10DO ) and 123 (C H11O )
relative to the extrapolated value obtained from the
correlation of Figure 1. One may reasonably imagine that
differences in geometrical flexibility along the reaction
coordinate enjoyed by the larger bicyclic system provide
the advantage, but the findings of appropriate compu-
tational work will be necessary before the issues may be
resolved.
were in the ratio of 78:22.
7
5-d
2
-Met h ylsu lfon yl-8-ch lor o-7,8-d
2
-b icyclo[4.2.0]oct a n e
). The ketone prepared immediately above (400 mg, 2.5
OD were placed in a 25-mL three-
necked flask fitted with a magnetic stirrer and solid addition
device. The solution was cooled to 0 °C; NaBD (925 mg, 25
(
mmol) and 5 mL of CH
3
4
mmol) was added in small portions over 1 h. The reaction
mixture was allowed to warm to room temperature and was
stirred for 21 h; it was then cooled to 0 °C and diluted with a
cold mixture of 10 mL of 1 N HCl and 20 mL of ether. The
layers were separated, and the aqueous phase was extracted
three times with ether. The combined ethereal material was
Exp er im en ta l Section
Bicyclo[4.2.0]oct -7-en e (1) was prepared by the photo-
isomerization of cis,cis-1,3-cyclooctadiene using acetophenone
as the sensitizer, heptane as the solvent, a 450-W medium-
washed with water, 1 N HCl, water, saturated NaHCO
and brine; it was then dried over MgSO and filtered. Con-
centration afforded 327 mg (80% yield) of crude 7-hydroxy-8-
chloro-7,8-d -bicyclo[4.2.0]octanes as a colorless liquid.
To a solution of 300 mg of these alcohols (1.85 mmol) in 20
mL of CH Cl in a 50-mL flask cooled to 0 °C was added 1.0
3
, water,
2
4
pressure mercury lamp, and an immersion apparatus. After
an unexceptional workup and isolation was completed, a
sample of 1 was purified and isolated through preparative GC
on a 1-m 10% SE-30 column at 60 °C. Sample homogeneity
was confirmed using a fused silica 0.2-mm i.d., 25-m cross-
linked 5% phenyl methyl siloxane capillary GC column. The
recorded NMR spectra matched those reported in the literature:
2
2
2
mL of triethylamine (7.4 mmol). Methanesulfonyl chloride (0.5
mL, 6.13 mmol) was placed in an addition funnel and added
slowly over 1 h. The reaction mixture was allowed to warm
slowly to room temperature, stirred for 7 h, and then cooled
to 0 °C. A mixture of 5 mL of cold water and 5 mL of CH Cl
2 2
was added. The mixture was stirred for 30 min; cold 1 N HCl
was added, and the layers were separated. The aqueous phase
2
1
4
3
1
H NMR δ 6.10 (s, 2H), 2.83 (m, 2H), 1.30-1.75 (m, 8H);
C NMR δ 140.5 (C7, C8), 41.5 (C1, C6), 24.9 (C2, C5), 18.8
(
C3, C4).
,8-Dich lor obicyclo[4.2.0]octa n -7-on e (4).26 To a mixture
8
of zinc (6.53 g, 0.1 mol), cyclohexene (5.0 mL, 0.05 mol), and
1
25 mL of anhydrous ether placed in a sonication bath29
was extracted three times with CH Cl . The combined organic
2
2
maintained at 15-20 °C was added over a 90-min period a
solution of 13 g (0.072 mol) of trichloroacetyl chloride in 63
mL of ether. Sonication of the reaction mixture at 15 °C was
continued for another 6.5 h. The mixture was quenched with
wet ether and filtered through a sintered glass funnel; the zinc
was rinsed with wet ether, and the total filtrate was washed
material was washed with cold 1 N HCl, cold water, saturated
Na CO , cold water, and brine; it was dried (MgSO ), filtered,
and concentrated to afford 260 mg (60% yield) of crude
mesylates 5-d as a dark-orange liquid. The MS m/z 109:108
2
3
4
2
+
+
(C
8
H
9
D
2
/C
8
H10D ) ion intensity ratio was 79:21.
7,8-d -Bicyclo[4.2.0]oct-7-en e (1-d ). Ammonia (100 mL)
2
2
with water (2 × 20 mL), NaHCO
20 mL). The ethereal solution was dried (Na
and concentrated to yield 7.33 g (78%) of dichloroketone 4:
3
(aq) (5 × 20 mL), and brine
was condensed at -78 °C in a 250-mL three-necked flask fitted
with a magnetic stirrer, a dry ice/acetone condenser, and an
addition funnel. Sodium (1.0 g, 43 mmol) was added in small
pieces, producing a deep-blue color, and the solution temper-
ature was maintained at -78 °C. A solution of 400 mg of crude
monochloromesylates 5-d2 (260 mg, 1.66 mmol) in 20 mL of
dry THF was placed in the addition funnel and added to the
Na/NH3 reaction mixture over 30 min. The reaction mixture
was allowed to warm to -35 °C and stirred for 3.5 h. Then,
(
2
SO ), filtered,
4
1
H
NMR δ 3.86 (m, 1H), 2.90 (m, 1H), 2.03 (m, 2H), 1.54 (m, 3H),
1
4
1
1
.33 (m, 1H), 1.18 (m, 2H); 13C NMR δ 196.7, 87.86, 52.67,
3.2, 25.56, 21.54, 21.21, 20.67; MS (m/z) 68 (base), 122, weak
+
35
92 (M , C
8
H
10 Cl
2
O). The 192:194:196 intensity ratio was
00:65:10.
Bicyclo[4.2.0]oct-7-en e (1) fr om 8,8-Dich lor obicyclo-
2
7,28
[
4.2.0]octa n -7-on e (4).
roketone 4 to bicyclooctene 1 (reduction with NaBH
mixture of alcohols, reaction of these alcohols with MsCl and
Et N in CH Cl to afford the corresponding mixture of mesylate
esters, and finally reduction with Na in NH /THF, following
the detailed protocols used by Belfield) gave olefin 1, identical
A three-step route from dichlo-
4
NH Cl was added until the blue color was gone. The dry ice/
4
to give a
acetone condenser and the addition funnel were replaced with
two cold-water condensers with attached oil bubblers. The
reaction mixture was allowed to warm to 0 °C. Water (10 mL)
was added to dissolve the remaining salts, and then 30 mL of
pentane was added. The layers were separated, and the
aqueous phase was extracted three times with pentane. The
combined organic material was washed with water, 1 N HCl,
3
2
2
3
1
3
with an authentic sample: C NMR δ 140.5, 41.4, 24.8, 18.8.
,2,5,5-d -Bicyclo[4.2.0]oct-7-en e (1-d ) fr om 2,2,5,5-d
). A sample of
2
4
4
4
-
8
7
,8-Dich lor obicyclo[4.2.0]octa n -7-on e (4-d
4
3
water, NaHCO (aq), water, and brine. It was dried over
218 J . Org. Chem., Vol. 69, No. 21, 2004

Thermal Isomerization of Bicyclo[4.2.0]oct-7-ene
MgSO
-d . Integration of the H NMR spectrum of a GC-purified
sample showed relative intensities for absorptions at δ 6.10
vinyl H) and 2.83 ppm (bridgehead H) of 0.14:1.00 (86%
average d labeling or 28% d and 72% d ).
Th er m olysis r ea ction s in the gas phase were conducted
in a static reactor with pentane or 2,2-dimethylbutane as a
bath gas and cyclooctane as an internal standard. The ap-
paratus, associated instrumentation, and methods used for the
kinetic runs have been detailed elsewhere. The time-dependent
decays of bicyclooctenes as cyclooctadienes were formed are
summarized in Tables 1 and 2.
4
, filtered, and concentrated to provide 20 mg (11%) of
sity and CHE-0075097 at Franklin & Marshall College,
and acknowledge the donors of the Petroleum Research
Fund, administered by the ACS, for partial support of
this research at Franklin & Marshall College. S.S.G.
received partial support through the Annie and Ernest
Weibrecht Endowed Scholarship for 2003-2004. Fur-
ther special thanks are due to Professor David K. Lewis,
Connecticut College, for a generous gift of 3,3,6,6-d4-
cyclohexene and Professor Christopher J . Samuel, Uni-
versity of Warwick, for his most forthcoming and
thoughtfully shared considerations of experimental
details and mechanistic implications of reactions bear-
ing on the present work.
1
1
2
(
2
1
2
Ack n ow led gm en t. We thank the National Science
Foundation for support of this work through Grants
CHE-9902184 and CHE-0211120 at Syracuse Univer-
J O040220L
J . Org. Chem, Vol. 69, No. 21, 2004 7219