Altering the allowed/forbidden gap in cyclobutene electrocyclic reactions: Experimental and theoretical evaluations of the effect of planarity constraints

Article abstract of DOI:10.1021/ja028963g

The allowed conrotatory cyclobutene ring-opening has a distinctly nonplanar carbon skeleton. Classic experiments by Brauman and Archie, and by Freedman et al., placed the allowed/forbidden gap at greater than 15 kcal/mol. Wolfgang Roth proposed that a system forced to planarity might have a smaller preference for the conrotatory mode than unconstrained systems. Such systems have now been studied theoretically and experimentally, and results that confirm Roth's postulate are presented here. The experiments were performed in Bochum, and the calculations were carried out in Osaka and Los Angeles. As the cyclobutene ring-opening transition structure approaches planarity, the energy gap between allowed conrotatory and the forbidden disrotatory pathways decreases. For the ring-opening of a cyclobutene fused to norbornene, the energy gap between the forbidden and the allowed transition state is only 4.1 kcal/mol by CASSCF and 8.0 kcal/mol by CAS-MP2 as compared to 13.4 and 19.2 kcal/mol, respectively, for the parent cyclobutene. Experimental studies of 3,4-dimethylcyclobutenes fused to various ring systems are reported, and a trend is found toward a reduced allowed/forbidden gap as the planarity of the cyclobutene is enforced.

Full text of DOI:10.1021/ja028963g

Published on Web 04/22/2003
Altering the Allowed/Forbidden Gap in Cyclobutene
Electrocyclic Reactions: Experimental and Theoretical
Evaluations of the Effect of Planarity Constraints
Patrick S. Lee,† Shogo Sakai,‡,| Peter Ho¨rstermann,§,| Wolfgang R. Roth,§,
E. Adam Kallel,† and K. N. Houk*,†
Contribution from the Department of Chemistry and Biochemistry, UniVersity of California,
Los Angeles, Los Angeles, California 90024-1569, Department of Information Systems
Engineering, Osaka Sangyo UniVersity, Daito Japan, 574-8530, and Abteilung fur Chemie,
Ruhr-UniVersitat Bochum, Germany
Received October 15, 2002; E-mail: houk@chem.ucla.edu
Abstract: The allowed conrotatory cyclobutene ring-opening has a distinctly nonplanar carbon skeleton.
Classic experiments by Brauman and Archie, and by Freedman et al., placed the allowed/forbidden gap at
greater than 15 kcal/mol. Wolfgang Roth proposed that a system forced to planarity might have a smaller
preference for the conrotatory mode than unconstrained systems. Such systems have now been studied
theoretically and experimentally, and results that confirm Roth’s postulate are presented here. The
experiments were performed in Bochum, and the calculations were carried out in Osaka and Los Angeles.
As the cyclobutene ring-opening transition structure approaches planarity, the energy gap between allowed
conrotatory and the forbidden disrotatory pathways decreases. For the ring-opening of a cyclobutene fused
to norbornene, the energy gap between the forbidden and the allowed transition state is only 4.1 kcal/mol
by CASSCF and 8.0 kcal/mol by CAS-MP2 as compared to 13.4 and 19.2 kcal/mol, respectively, for the
parent cyclobutene. Experimental studies of 3,4-dimethylcyclobutenes fused to various ring systems are
reported, and a trend is found toward a reduced allowed/forbidden gap as the planarity of the cyclobutene
is enforced.
The transition structures of the conrotatory (via 1)1-3 and
disrotatory processes (via 2)7 have been elucidated in recent
Introduction
The ring-opening of cyclobutene has been studied extensively
both theoretically and experimentally as the prototype of thermal
electrocyclic processes.1-3 The four-electron electrocyclic ring-
opening process is conrotatory under thermal conditions, where
the termini rotate in the same direction in accord with the
Woodward-Hoffmann rules (1).1 The experimental ∆Hq for the
electrocyclic opening of cyclobutene is 32.5 ( 0.5 kcal/mol,
and the reaction is exothermic by 10.8 kcal/mol.4 Cyclobutenes
have played a major role in the understanding of such processes
as torquoselectivity, the role of substituents on activation
energies of clockwise and counterclockwise conrotatory ring-
opening.5,6
years. The allowed conrotatory cyclobutene ring-opening has a
distinctly nonplanar carbon skeleton. Classic experiments by
Brauman and Archie, and by Freedman et al., placed the
allowed/forbidden gap at greater than 15 kcal/mol.
The late Wolfgang Roth proposed that a system forced to
planarity might have a smaller preference for the conrotatory
mode. The concept is as follows: as one destabilizes the
conrotatory transition structure by minimizing the overlap
between the σ orbital of the breaking bond and the π orbital
which exists in the cyclobutene, the allowed-forbidden gap
should decrease. This destabilization can be achieved by
introducing constraints to force either the H-C1-C2-H or the
C4-C1-C2-C3 angle to approach 0°. This planarity constraint
should play only a small role in the disrotatory process, because
this process occurs through virtually a planar process. The
primary overlap of the terminal orbitals of the conrotatory
transition state is stabilizing, whereas this stabilization is
decreased in the planar transition structure. Experimental work
was undertaken under Roth’s direction, and theoretical calcula-
tions were begun as a means to provide theoretical insights into
† University of California, Los Angeles.
‡ Osaka Sangyo University.
§ Abteilung fur Chemie, Ruhr-Universitat Boch.
| Corresponding authors for CASSCF and CiLC-IRC calculations,
experimental results, and overall interpretations, respectively.
Deceased, October 29, 1997.
(1) Lee, P. S.; Zhang, X.; Houk, K. N. J. Am. Chem. Soc., ASAP article.
(2) Zimmerman, H. E. In Pericyclic Reactions; Marchand, A. P., Lehr, E.,
Eds.; Academic Press: New York, 1977; Vol. I, pp 53-107.
(3) Houk, K. N. In Pericyclic Reactions; Marchand, A. P., Lehr, E., Eds.;
Academic Press: New York, 1977; Vol. II, pp 181-271.
(4) (a) Cooper, E.; Walters, W. D. J. Am. Chem. Soc. 1958, 80, 4220-4224.
(b) Wiberg, K. B.; Fenoglio, R. A. J. Am. Chem. Soc. 1968, 90, 3395-
3397. (c) Criegee, R.; Seebach, D.; Winter, R. E.; Bo¨rretzen, B.; Brune,
H. Chem. Ber. 1965, 98, 2339-2352.
(5) For a review: Dolbier, W. R., Jr.; Koroniak, H.; Houk, K. N.; Sheu, C.
Acc. Chem. Res. 1996, 29, 471-477 and references therein.
(6) Lee, P. S.; Zhang, X.; Houk, K. N., submitted for publication.
(7) Sakai, S. THEOCHEM 1999, 461-462, 283-295.
9
10.1021/ja028963g CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 5839-5848
5839

A R T I C L E S
Lee et al.
Scheme 1
Scheme 2. Newman Projections of Cyclobutene Ring-Opening TS
and Several Constrained Geometries
Figure 1. Orbital correlation diagram of cyclobutene and conrotatory
transition state (HF/6-31G(d)). Correlations with butadiene π and π* orbitals
are also shown.
these processes. This work is the culmination of several decades
of study and establishes the effect of planarity on the allowed-
forbidden gap. Roth’s hypothesis has been verified.
Scheme 3
Background
Figure 1 illustrates an orbital correlation diagram of cy-
clobutene and the ring-opening transition state generated by HF/
6-31G(d) calculations. This figure is an elaboration of the orbital
correlations by Woodward and Hoffmann1 and also by Longuet-
Higgins and Abrahamson8 that stereospecificity may be rational-
ized by correlations of the π and π*, and the σ and σ* orbitals
of cyclobutene with the π and two π* orbitals of butadiene.
This correlation diagram has been also used to help explain
torquoselectivity.5,6,9 The conrotatory ring-opening, which retains
C2 symmetry throughout the reaction, correlates the σ bond of
cyclobutene with the HOMO of the transition state. The HOMO
of the transition state is essentially a distorted σ orbital of
cyclobutene. Similarly, the σ* orbital of the reactant correlates
with the LUMO of the transition state.10
The ring-opening transition state occurs at the point where
significant overlap of the breaking σ with the π* orbitals and
of the π with the σ* begins, as illustrated in Scheme 2. Once
the overlap is large, there is a rapid stabilization of the filled
orbitals to give the product. Roth noted that if the cyclobutene
were kept planar, this overlap would be quite small. If the
resulting destabilization were large enough, then the allowed
and forbidden transition states would have similar energies. The
overlap decreases from the unconstrained transition state
geometry to the H-planar geometry, which has constraints
mentioned previously, and the overlap is the smallest in the
all-planar geometry, as illustrated in Scheme 2.
0.005% of the disallowed product, which equates to an allowed-
forbidden gap of greater than 15 kcal/mol.11 Freedman et al.
found that cis-1,2,3,4-tetraphenylcyclobutene opens exclusively
to the cis,trans-1,2,3,4-tetraphenylbutadiene, and the corre-
sponding trans-cyclobutene opens exclusively to the cis,cis-
butadiene.12
These studies and the overlap factors illustrated in Scheme 2
led Roth to propose that it is possible to destabilize 1 by inducing
planarity on the dihedral angle of Ha-C1-C2-Hb (Scheme 1).
Closing the gap between the conrotatory and disrotatory
pathways could be achieved by distorting the transition structure
by inducing planarity on either Ha-C1-C2-Hb or C4-C1-C2-
C3. The cyclobutene ring-opening can, of course, be forced to
proceed through the disrotatory pathway by fusing the ring such
that a cis-bridge is introduced across C3-C4, as in 4 and 5,13
but this study involves systems 7-12, where the constraints
are at the double bond of cyclobutene instead (Scheme 3).
The high stereospecificity associated with the conrotatory
mode was confirmed by the pyrolysis of 3,4-dimethylcy-
clobutene by Brauman and Archie. They found approximately
(11) Brauman, J. I.; Archie, W. A., Jr. J. Am. Chem. Soc. 1972, 94, 4262-
4265.
(12) Freedman, H. H.; Doorakian, G. A.; Sandel, V. R. J. Am. Chem. Soc. 1965,
87, 3019-3020.
(13) (a) Baldwin, J. E.; Andrist, A. H. J. Chem. Soc., Chem. Commun. 1970,
1561-1562. (b) Brauman, J. I.; Golden, D. M. J. Am. Chem. Soc. 1968,
90, 1920-1921. (c) McDonald, R. N.; Reinecke, C. E. J. Org. Chem. 1967,
32, 1878-1887.
(8) Longuet-Higgins, H. C.; Abrahamson, E. W. J. Am. Chem. Soc. 1965, 87,
2045-2046.
(9) Niwayama, S.; Kallel, E. A.; Spellmeyer, D. C.; Sheu, C.; Houk, K. N. J.
Org. Chem. 1996, 61, 2813-2825.
(10) Rondan, N. G.; Houk, K. N. J. Am. Chem. Soc. 1985, 107, 2099-2111.
9
5840 J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003

Altering the Gap in Cyclobutene Reactions
A R T I C L E S
Scheme 4
There have been previous theoretical studies of the disrotatory
cyclobutene ring-opening. Breulet and Schaefer utilized a two-
configuration self-consistent field (TCSCF) ab initio calculation
to characterize conrotatory and disrotatory stationary points.14
They found that the conrotatory stationary point is a true
transition state, whereas the disrotatory stationary point has two
imaginary vibrational frequencies. This study was at the time
an improvement of initial MINDO/3 calculations by Dewar and
Kirschner who found a ground-state disrotatory transition state.15
Robb et al. presented a detailed excited-state potential energy
surface of butadiene using a multiconfigurational self-consistent
field (MCSCF) and proposed that a disrotatory process proceeds
through a conical intersection.16 Buenker et al. have attempted
to understand the role of ring torsion in the electrocyclic reaction
of cyclobutene to butadiene and have found that the rotational
phase of the process occurs over a narrow range of distance
between C3 and C4.17 Maier and Bothur also studied both
theoretically (B3LYP/6-311G(d)) and experimentally the thermal
and photochemical ring-openings of dichlorocyclobutenes.18
Only recently was the thermal disrotatory transition state for
cyclobutene ring-opening found by Sakai with CASSCF cal-
culations.7 The forbidden pathway has an activation energy 19.5
kcal/mol higher than that of the allowed conrotatory pathway,
according to MP2/CAS 6-311+G(d,p) calculations.7
The planarity effect on the allowed-forbidden gap was
originally investigated experimentally by Roth and Ho¨rstermann
in 1979.19 Houk and Kallel attempted theoretical studies of the
system in the late 1980’s, but they were unable to locate
disrotatory transition states.20 A collaborative manuscript be-
tween the Bochum and Los Angeles groups was begun at that
time. Interest in these reactions recently resurfaced due to
Sakai’s successful location of transition states for disrotatory
reactions of cyclobutenes and derivatives.7 We have now
combined modern computational work and the details of the
original experimental study of the effect of planarity on the
cyclobutene ring-opening transition states.
tions, a rapid equilibration to a ratio of 14:15 ) 5:1 occurred.
Further irradiation of the mixture gave three products which
were separated by gas chromatography and identified as the
expected cyclization products 7 and 8 and the diene 16 derived
from 14 via a 1,5-hydrogen shift (Scheme 4).
The NMR spectrum of cis-6,7-dimethylbicyclo[3.2.0]hept-
1(5)-ene (7) showed the doublet for the methyl groups at 1.0
ppm, whereas the corresponding signal of the trans-isomer 8 is
shifted to a lower field (1.15 ppm). For the tertiary protons of
the four-membered ring, the opposite effect is observed. The
cis-product showed a multiplet at 2.9 ppm for the allylic protons
separated from the other ring protons of the molecule. The
corresponding signal of the trans-isomer overlaps with the other
signals between 2.0 and 2.4 ppm. Similar effects have been
observed with other systems having vicinal substituents such
as 1,2-dimethylene-3,4-dimethylcyclobutane or 3,4-dimethyl-
cyclobutene.23 Furthermore, the stereochemistry can be properly
assigned from the configuration of the products of the pyrolysis
of 7 and 8, respectively. Because the mode of the thermal ring-
opening is clear (conrotatory), the main (allowed) product shows
whether the starting material was cis or trans. Product 14 is
derived from the cis-compound 7, and 15 is derived from the
trans-compound 8. The stereochemistry of 14 (Z,E) and 15 (E,E)
was assigned from the interpretation of their NMR spectra. The
NMR of the unsymmetrical (Z,E)-compound 14 is more complex
than the NMR of 15 (E,E).
1b. Thermal Ring-Opening of cis- and trans-6,7-Dimethyl-
bicyclo[3.2.0]hept-1(5)-ene (7 and 8). For the investigation of
the stereoselectivity of the ring-opening reaction, the isomeric
cyclobutenes 7 and 8 were thermolized in the gas phase. To
ensure the detection of the smallest amounts of the “forbidden”
products, the starting materials for the thermolysis were purified
by gas chromatography to an isomeric purity of >99.99%. The
ring-opening of 7 was investigated at three temperatures between
125 and 200 °C. Besides the expected 1,2-diethylidenecyclo-
pentanes 14 and 15, diene 16 could be detected in various
amounts depending on the reaction temperature. By thermolysis
of pure 14 and 15, it could be shown that only the Z,E-isomer
(14) undergoes the H-shift reaction, whereas the E,E-isomer (15)
was stable under the reaction conditions (Scheme 5). The results
of the thermolyses are listed in Table 1.
Results
Experimental. We summarize here the principal experimental
results. Full experimental details are given in the Supporting
Information.
1a. Synthesis of cis- and trans-6,7-Dimethylbicyclo[3.2.0]-
hept-1(5)-ene (7 and 8). The target molecules were synthesized
starting from ketone 13, which was prepared in two steps from
cyclopentanone according to Birkofer.21 The reaction of 2-eth-
ylidenecyclopentanone with the Wittig reagent prepared from
ethyltriphenylphosphonium bromide and n-butyllithium gave a
mixture of dienes 14 and 15 in a ratio of 1:7 in 15% yield. An
analytical sample of the mixture was separated by preparatory
gas chromatography, and the mixture of 14 and 15 was
photocyclized without separation.22 Under the reaction condi-
(14) Breulet, J.; Schaefer, H. F., III. J. Am. Chem. Soc. 1984, 106, 1221-1226.
(15) Dewar, M. J. S.; Kirchner, S. J. Am. Chem. Soc. 1974, 96, 6809-6810.
(16) (a) Olivucci, M.; Bernardi, F.; Ragazos, I. N.; Robb, M. A. J. Am. Chem.
Soc. 1993, 115, 3710-3721. (b) Bernardi, F.; De, S.; Olivucci, M.; Robb,
M. A. J. Am. Chem. Soc. 1990, 112, 1737-1744.
(17) Hsu, K.; Buenker, R. J.; Peyerimhoff, S. D. J. Am. Chem. Soc. 1972, 94,
5639-5644.
(22) The light source was a mercury low-pressure lamp of the type 100 W,
made by Gra¨ntzel.
(18) Maier, G.; Bothur, A. Eur. J. Org. Chem. 1998, 2063, 3-2072.
(19) Ho¨rstermann, P. Dissertation, Ruhr Universita¨t, Bochum, Germany, 1979.
(20) Kallel, E. A.; Houk, K. N., unpublished and now obsolete results.
(21) (a) Birkofer, L.; Barnikel, C. D. Chem. Ber. 1958, 91, 1996-1999. (b)
Birkofer, L.; Kim, S. M.; Engels, H. D. Chem. Ber. 1962, 95, 1495-1504.
(23) (a) Gajewski, J. J.; Shih, C. N. J. Am. Chem. Soc. 1969, 91, 5900-5901.
(b) Gajewski, J. J.; Black, W. A. Tetrahedron Lett. 1970, 899-902. (c)
Winter, R. E. K. Tetrahedron Lett. 1965, 1207-1212. (d) Criegee, R.;
Seebach, D.; Winter, R. E.; Borretzen, B.; Brune, H.-A. Chem. Ber. 1965,
98, 2339-2352.
9
J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003 5841

A R T I C L E S
Lee et al.
Table 1. Products of the Thermolysis of 7
unsubstituted cyclobutene. The activation parameters of the
electrocyclic ring-opening were determined to confirm this
assumption.
products
reaction conditions
conversion [%]
14 [%]
15 [%]
16 [%]
1c. Kinetics of the Gas-Phase Thermolysis of cis- and
trans-6,7-Dimethylbicyclo[3.2.0]hept-1(5)-ene (7 and 8). The
thermolysis of the substituted cyclobutenes 7 and 8 was carried
out in a hot air thermostat directly connected to a gas
chromatograph. The reaction was monitored without condensa-
tion of the products for each analysis. The reaction rates were
determined from the decrease of the starting material. The
isomerization of the cis-dimethyl derivative, 7, was monitored
at six temperatures between 100 and 150 °C at a pressure of 1
Torr. Under these conditions, the reaction was first order.
125 °C/120 min
160 °C/60 min
200 °C/60 min
45
100
100
98.76
93.67
33.15
0.27
0.42
0.61
0.97
5.91
66.25
Table 2. Products of the Thermolysis of 8
products
reaction conditions
conversion [%]
14 [%]
15 [%]
16 [%]
125 °C/30 min
160 °C/10 min
200 °C/10 min
100
100
100
0.05
0.06
0.03
99.95
99.94
99.93
0.04
The following activation parameters were obtained from the
Arrhenius or Eyring plots (available in the Supporting Informa-
tion): Ea ) 32.70 ( 0.08 kcal/mol, A ) (6.87 ( 0.77) × 1013/
s, ∆H130 °Cq ) 32.50 kcal/mol, ∆S130 °Cq ) 2.19 cal/mol‚K, and
Scheme 5
q
∆G130 °C ) 31.63 kcal/mol.
The corresponding kinetics of the trans-dimethyl derivative
8 were investigated at eight temperatures between 60 and 135
°C, and the following activation parameters were obtained from
the Arrhenius or Eyring plots: Ea ) 29.02 ( 0.04 kcal/mol,
q
A ) (4.32 ( 0.28) × 1013/s, ∆H97 °C ) 28.28 kcal/mol,
q
q
∆S97 °C ) 1.44 cal/mol‚K, and ∆G97 °C ) 27.72 kcal/mol.
The ring-opening of cis-6,7-dimethylbicyclo[3.2.0]hept-1(5)-
ene (7) has a 1.4 kcal/mol higher activation energy than the
1,2-dimethylcyclobutene ring-opening and 4.9 kcal/mol higher
activation energy than the parent cyclobutene ring-opening,
whereas the activation energy of the trans-6,7-dimethylbicy-
clo[3.2.0]hept-1(5)-ene (8) ring-opening is decreased by 2.4 kcal/
mol as compared to the 1,2-dimethylcyclobutene ring-opening.
These results are consistent with analogous systems shown in
Table 3. Trans substitution results in lower activation energies
than cis substitution. In general, cis substitution results in an
increase in activation energy as compared to the unsubstituted
systems, except in benzocyclobutene ring-opening. This series
is the only endothermic process of the systems listed in Table
3 due to the loss of aromaticity upon ring-opening. The methyl
substitution stabilizes the transition state that results in loss of
aromaticity, and therefore there is a drop in activation energy
for the cis and trans substituted benzocyclobutene ring-opening
as compared to the parent benzocyclobutene to o-xylylene
process.
Scheme 6
To eliminate any catalytic effects from the surface, the whole
apparatus was silylated multiple times. However, no influence
on the composition of the product mixture was observed in
comparison to the nonsilylated apparatus, and the results do not
appear to be influenced by secondary reactions. The average
difference between free energies of activation can be calculated
from the ratio of the products using eq 1, where k1 is the sum
of the products 14 and 15, and k2 is the percentage of 15.
2a. Strain and Stereochemistry of the Thermal Isomer-
ization of Tricyclo[4.2.1.02.5]non-2(5)-ene (12) and the cis-
and trans-3,4-Dimethyl Derivatives. The experiments described
above show that bridging the 1,2 positions in bicyclo[3.2.0]-
hept-1(5)-ene (9) leads to a significant decrease of the activation
barrier for the forbidden disrotatory ring-opening, but there is
only little effect of ring strain on the activation energy in the
case of the allowed conrotatory ring-opening. Another highly
strained cyclobutene is the tricyclo[4.2.1.02.5]non-2(5)-ene (12)
described by Aue28 which reacts to form 2,3-dimethylenebicyclo-
∆∆Gq ) RT ln(k1/k2)
(1)
∆∆G160 °Cq ) 4.7 kcal/mol
The thermolysis of the trans-olefin 8 was performed under
the same conditions used for the cis-isomer and gave identical
products (Scheme 6 and Table 2).
The difference of the free energies of activation for allowed
and forbidden paths was found to be ∆∆G160 °C ) 6.4 kcal/
q
mol. The difference of 1.7 kcal/mol between the ∆∆Gq values
can be explained by the steric hindrance of the methyl groups
which assists the forbidden disrotatory reaction of the cis-isomer.
A similar but opposite effect can be assumed for the allowed
ring-opening, leading to an increased activation energy of the
cis-diene 7 relative to the corresponding reaction of the
(24) Frey, H. M. Trans. Faraday Soc. 1963, 59, 1619-1622.
(25) Branton, G. R.; Frey, H. M.; Skinner, R. F. Trans. Faraday Soc. 1966, 62,
1546-1552.
(26) Roth, W. R.; Biermann, M.; Dekker, H.; Jochems, R.; Mosselman, C.;
Hermann, H. Chem. Ber. 1978, 111, 3892-3903.
(27) Roth, W. R.; Rekowski, V.; Boerner, S.; Quast, M. Liebigs Ann. 1996, 3,
409-430.
(28) Aue, D. H.; Reynolds, R. N. J. Am. Chem. Soc. 1973, 95, 2027-2028.
9
5842 J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003

Altering the Gap in Cyclobutene Reactions
A R T I C L E S
Table 4. Activation Energies of Strained Cyclobutenes
Table 3. Activation Energies of the Ring-Opening of
3,4-Disubstituted Cyclobutenes
alkene is 4.1 kcal/mol more strained than 9. Analogous to the
results found with the bicyclo[3.2.0] system, the activation
energy for the disrotatory ring-opening of 12 should be further
decreased. On the other hand, there should be no net effect of
the ring strain on the conrotatory ring-opening activation energy.
The activation energy is expected to be approximately 31 kcal/
mol. Aue had published an activation energy of only 26.6 kcal/
mol for this system, but this experiment was carried out in
solution and monitored by NMR spectroscopy. In solution,
compound 12 tends to polymerize, so the activation barrier might
not be reliable. Consequently, the ring-opening of 12 was
explored in the gas phase. Dimerization products were ruled
out because the gas chromatogram of the reaction mixture
showed no byproducts. Work by Aue and Reynolds reports
details of the problems with dimerization at the NMR concen-
trations used, and they conclude that the reaction has a barrier
closer to 30.5 kcal/mol.30
Compound 12 was synthesized from the diene 17 by
photochemical ring closure.28 Compound 17 can be prepared
easily by the Diels-Alder reaction of cyclopentadiene with (E)-
1,4-dichlorobutene.31 Because of the air-sensitivity of the olefin
12, the workup of the crude reaction mixture was carried out
under an argon atmosphere. The product was purified by
preparative gas chromatography. Thermolysis of 12 was carried
out at eight different temperatures between 92 and 165 °C in
the same equipment used for the rearrangement of the olefins
7 and 8.
The following activation parameters were taken from the
Arrhenius and Eyring plots (available in the Supporting
Information): Ea ) 30.89 ( 0.08 kcal/mol, A ) (2.57 ( 0.30)
× 1013/s, ∆H128 °Cq ) 30.08 kcal/mol, ∆S128 °Cq ) 0.25 cal/mol‚
Scheme 7
q
K, and ∆G128 °C ) 29.98 kcal/mol.
The activation energy of 30.9 kcal/mol in the gas phase is
4.3 kcal/mol higher than the activation energy determined by
Aue in solution.28 The Ea value found for the ring-opening of
12 supports the hypothesis that the activation energy for the
conrotatory reaction depends very little on the ring strain of
the cyclobutene, which is obtained by the difference in heat of
hydrogenation of the compounds (Table 4).
3a. Synthesis of cis- and trans-3,4-Dimethyltricyclo-
[4.2.1.02.5]non-2(5)-ene (10 and 11). To determine the relative
activation energies of the allowed and forbidden disrotatory ring-
[2.2.1]heptane (17) (Scheme 7). We explored the strain in this
system and the stereochemistry of ring-opening in substituted
derivatives.
(29) Roth, W. R.; Adamczak, O.; Breukmann, R.; Lennartz, H. W.; Boese, R.
Chem. Ber. 1991, 124, 2499-2521.
(30) Reynolds, R. N. Ph.D. Thesis, University of California, Santa Barbara, 1977.
(31) Bowe, M. A. P.; Miller, R. G. J.; Rose, J. B.; Wood, D. G. M. J. Chem.
Soc. 1960, 1541-1547.
2b. Heat of Hydrogenation and Kinetics of the Gas-Phase
Thermolysis of Tricyclo[4.2.1.02.5]non-2(5)-ene (12). Lennartz
found a heat of hydrogenation of 44.9 kcal/mol for 12.29 This
9
J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003 5843

A R T I C L E S
Lee et al.
Scheme 8
Table 5. Thermolysis of the cis-Olefin 10
products
reaction conditions
conversion [%]
25 [%]
26 [%]
140 °C/60 min
150 °C/30 min
150 °C/100 min
83
91
>99
0.97
1.08
1.04
99.03
98.92
98.96
Table 6. Thermolysis of the trans-Olefin 11
products
reaction conditions
conversion [%]
25 [%]
26 [%]
93 °C/30 min
110 °C/25 min
125 °C/35 min
96
100
100
99.94
99.92
99.90
0.06
0.08
0.10
3b. Thermolysis of cis- and trans-3,4-Dimethyltricyclo-
[4.2.1.02.5]non-2(5)-ene (10 and 11). The dienes 10 and 11 were
purified by repeated gas chromatography to an isomeric purity
greater than 99.98%. The thermal isomerization of the com-
pounds was carried out at temperatures between 93 and 150
°C. The only products observed were the expected dienes 25
and 26. To exclude isomerization of these dienes, both 25 and
26 were separately subjected to the thermolysis conditions. Both
appear to be completely stable. The thermolysis of the cis-isomer
10 under different reaction conditions between 140 and 150 °C
gave the product ratios shown in Table 5.
From these results, the free energy of activation for reaction
of 10 is ∆∆G150 °C ) 3.8 kcal/mol. Thermolysis of the
corresponding trans-isomer 11 at temperatures between 93 and
125 °C gave the results shown in Table 6.
openings of the tricyclo[4.2.1.02.5]non-2(5)-ene system, the
dimethyl derivatives 10a, 10b, and 11 were synthesized.
Compounds 10 and 11 were synthesized in seven steps
starting from cyclopentadiene and (E)-hex-3-en-2,5-dione (19)
as shown in Scheme 8. The Diels-Alder reaction of cyclopen-
tadiene and 19 gave trans-5,6-diacetylbicyclo[2.2.1]hept-2-ene
(20) in almost quantitative yield as a single product. Subsequent
catalytic hydrogenation of 20 with Pd/C in glacial acetic acid
led to trans-2,3-diacetylbicyclo[2.2.1]heptane (21) in 96% yield.
Both the reduction of the diketone 21 using lithium aluminum
hydride and the esterification of the resulting diol 22 with
methanesulfonyl chloride to the corresponding bis-mesylate 23
gave yields higher than 95%. Elimination of 23 with potassium
tert-butylate in dimethyl sulfoxide at temperatures below 40 °C
gave the corresponding trans-2,3-divinylbicyclo[2.2.1]heptane
24 in ca. 80% yield. Treatment of 24 with potassium tert-
butylate in dimethyl sulfoxide at 60 °C led to isomerization and
gave a quantitative mixture of the dienes 25 and 26 in a ratio
of 5:1.
Separation of the mixture of 25 and 26 on the preparative
scale by gas chromatography was difficult, so a photocyclization
was carried out using the 5:1 mixture. A solution of the mixture
in pentane was irradiated at temperatures between -5 and 0
°C, and the reaction was monitored by gas chromatography.
The cyclization of the dienes 25 and 26 was preceded by fast
equilibration (25:26 ) 27:73) similar to the reaction of the 1,2-
diethylidenecyclopentanes, 14 and 15. The photoreaction gave
three major products separated by gas chromatography and
identified as compounds 10, 11, and 27 (Scheme 8).
q
In this case, the free energy of activation of 11 is
q
∆∆G110 °C ) 5.4 kcal/mol. Comparison of the ∆∆Gq values
of 10 and 11 shows that the disrotatory ring-opening of the cis-
isomer 10 is favored by 1.6 kcal/mol more than for the
analogous reaction of the trans-isomer 11. The effect is of the
same magnitude as observed at the bicyclo[3.2.0]hept-1(5)-ene
system and can be explained by steric interactions between the
methyl groups of the cis-olefin 10.
Theoretical. All of the systems studied experimentally were
also studied theoretically. All stationary points were initially
optimized with RHF with 3-21G and 6-31G(d) basis sets, then
optimized again at the CASSCF and MP2 levels with the 6-31G-
(d) basis set using the GAUSSIAN 9833 and GAMESS34
programs, respectively. For the CASSCF calculation, four active
orbitals corresponding to two π and two π* orbitals of the
butadienes were included. All configurations in the active space
were generated. The force-constant matrix and vibrational
frequencies were calculated with analytical second derivatives.
(32) Cocks, A. T.; Frey, H. M. J. Chem. Soc. B 1970, 952-954.
(33) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,
J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.7; Gaussian,
Inc.: Pittsburgh, PA, 1998.
The stereochemical assignment of the cis-product 10 to one
of the two possible structures 10a (exo) and 10b (endo) by
spectroscopic means was not possible. However, this is ir-
relevant for the following investigation as both compounds
would give the same product (26).
(34) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M.
S.; Jensen, J. J.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S.; Windus,
T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem. 1993, 14, 1347-
63.
9
5844 J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003

Altering the Gap in Cyclobutene Reactions
A R T I C L E S
Table 8. Computed Activation Energies (Relative Energies) of
Table 7. Products of Conrotatory and Disrotatory Ring-Opening of
3,4-Dimethylcyclobutenes and Derived Differences in Free
Energies of Activation
Conrotatory Cyclobutene Ring-Opening in kcal/mol
transition state RHF/3-21G RHF/6-31(d) MP2/6-31G(d) CAS/6-31G(d) CAS-MP2
q
% con
% dis
∆∆G (kcal/mol)
T (°C)
unconstrained 41.6 (0.0) 46.9 (0.0) 37.6 (0.0) 36.6 (0.0) 40.4 (0.0)
6
7
8
10
11
99.995
99.58
99.94
98.92
99.92
0.005
0.42
0.06
1.08
0.08
10.7
4.7
6.4
3.8
5.4
280
160
160
150
110
H-planar
C-planar
all planar
45.5 (3.9) 51.1 (4.2) 40.7 (3.1) 38.1 (1.5) 42.7 (2.3)
50.8 (9.2) 55.7 (8.8) 45.2 (7.6) 40.5 (3.9) 46.2 (5.8)
56.7 (15.1) 62.4 (15.5) 50.8 (13.2) 42.6 (6.0) 49.3 (8.8)
Parent System. Table 8 lists the computed activation energies
and relative energies for the unconstrained and several con-
strained cyclobutene ring-opening processes.37 The constraints
imposed are as follows. H-planar has a constrained Ha-C1-
C2-Hb dihedral angle of 0°. C-planar has a constrained C4-
C1-C2-C3 dihedral angle of 0°. Finally, all-planar has both of
the aforementioned dihedral angles constrained to 0°. The
H-planar structure is destabilized by 3.1 kcal/mol as compared
to the unconstrained system at MP2/6-31G(d) and 2.3 kcal/mol
at CAS-MP2. Inducing C-planarity increases the activation
energy by 7.6 kcal/mol as calculated by MP2/6-31G(d) and 5.8
kcal/mol by CAS-MP2. The all-planar activation barrier is
destabilized by 13.2 kcal/mol for the MP2/6-31G(d) calculation
and 8.8 for the CAS-MP2 calculation when compared to the
unconstrained transition state.
These were compared with the disrotatory transition state and
singlet diradical transition states. First, a singlet diradical was
formed by 90° rotation of one terminal methylene assuming Cs
symmetry and optimized with this constraint at the CAS
calculation level, which is found to be a stationary point. The
energy of this singlet diradical is only 1.7 or 7.2 kcal/mol above
the conrotatory transition state at the CASSCF and CAS-MP2
levels, respectively. Another singlet diradical was formed by
90° rotation of both terminal methylenes assuming C2V sym-
metry, at an angle of CCC ) 124.7°, which is equal to the
optimized angle of the above-mentioned singlet diradical with
one terminal methylene constrained to 90°. The energy of the
doubly constrained singlet diradical is 13.8 and 19.6 kcal/mol
at the CASSCF and CAS-MP2 levels, respectively, above the
conrotatory transition state, and is only 0.4 kcal/mol above the
disrotatory transition state at both the CASSCF and the CAS-
MP2 levels. Therefore, the disrotatory transition state is a
biradical state as examined in a previous paper by the CiLC-
IRC method with the configurational state function level.35a The
electronic state of this disrotatory transition state will be
discussed in a later section.
Additional calculations were performed to obtain improved
energy comparisons: calculations with the CASSCF optimized
structures with electron correlation incorporated through the
multiconfigurational second-order Møller-Plesset perturbation
theories (CAS-MP2) with the 6-31G(d) basis set. The intrinsic
reaction coordinate (IRC) was followed from the transition state
toward both reactants and products.
To interpret the reaction mechanisms of the conrotatory and
disrotatory pathways, a configuration interaction localized
molecular orbital CASSCF calculation along the intrinsic
reaction coordinate (CiLC-IRC) was performed following a
method described elsewhere:35 (1) A four-orbital and four-
electron CASSCF calculation is carried out to obtain a starting
set of orbitals for the localization procedure. (2) After the
CASSCF procedure is carried out, the CASSCF-optimized
orbitals are localized by the Boys localization procedure.36 The
calculated localized orbitals are atomic in nature. (3) By using
the localized MOs as a basis, we used a full CI with the
determinant level to generate electronic structures and their
weights in the atomic orbital-like wave functions. (4) The
calculation procedures from (1) to (3) are repeated along the
IRC path. The relative weights of the electronic configurations
are used for the analysis of the reaction mechanisms. The CiLC-
IRC calculations were also performed with the GAMESS
program package.34
Discussion
Table 7 summarizes the temperature of pyrolysis and the
ratios of products obtained for a series of 3,4-dimethylcy-
clobutenes (6-8, 10, and 11).11,19 The major product of ring-
opening is always the result of a conrotatory ring-opening, but
variable amounts, up to 1% of the forbidden product, were
observed. From the observed percentage of disrotatory product,
it is possible to calculate a ∆∆Gq between the allowed
conrotatory pathway and a pathway which gives the disrotatory
product. For such systems, the disrotatory product could come
from a concerted disrotatory path, or from a monorotatory
pathway, either of which could produce a mixture of conrotatory
and disrotatory stereochemistries. While the selectivities found
earlier for the unconstrained cyclobutene, 6, and the trans
dimethyl compounds, 8 and 11, are quite high in accord with
expectation, the stereoselectivities found with 7 and 10 are low.
Some other substituents may also tip the balance in favor of a
stepwise or disrotatory mechanism. For example, the cis-dichloro
analogue of 10 is estimated to have only a 2.5 kcal/mol
preference for the conrotatory pathway.10,39 To better understand
the mechanism of the forbidden pathway, theoretical explora-
tions of all of these systems were carried out.
The geometries of the constrained cyclobutene openings are
compared to the unconstrained ring-opening transition structure
in Figure 2. In the unconstrained conrotatory ring-opening
transition structure, the hydrogens at the C1 and C2 are out of
plane, allowing for the p orbitals of these carbons to overlap
with the breaking C3-C4 σ bond as illustrated previously in
Scheme 2. The destabilization of the H-planar transition state
is caused by the reduced overlap between these p and σ orbitals.
The outward and inward rotating hydrogens increase their
degrees of rotation ( HC4C1C2) from 131° and -76° with the
MP2 calculation (143° and -65° with the CASSCF calculation)
in the unconstrained transition structure to 133° and -70° (146°
(37) Spellmeyer, D. C.; Houk, K. N. J. Am. Chem. Soc. 1988, 110, 3412-
3416.
(38) Curry, M. J.; Stevens, I. D. R. J. Chem. Soc., Perkin Trans. 2 1980, 1391-
(35) (a) Sakai, S. J. Phys. Chem. A 1997, 101, 1140-1146. (b) Cundari, T. R.;
Gordon, M. S. J. Am. Chem. Soc. 1991, 113, 5231-5243.
(36) Foster, J. M.; Boys, S. F. ReV. Mod. Phys. 1960, 32, 296, 300-302.
1398.
(39) Pasto, D. J.; Krasnansky, R.; Zercher, C. J. Org. Chem. 1987, 52, 3062-
3072.
9
J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003 5845

A R T I C L E S
Lee et al.
Figure 2. Bond distances (Å) and angles of constrained cyclobutenes optimized by MP2 (CASSCF).
Table 9. Activation Energies of Conrotatory and Disrotatory
Electrocyclic Ring-Opening Reactions (kcal/mol)
and -61°) in the H-planar structure (Figure 2). In the C-planar
structure, a decrease in overlap of the breaking σ orbital with
the π bond is achieved by even greater bending of the H’s at
C1 and C2 away from planarity. The dihedral angle of the
hydrogens at C1 and C2 increases from 23° in the unconstrained
transition structure to 26° in the C-planar structure. The outer
and inner hydrogens now rotate to 136° and -71° at the MP2
(133° and -76° at CASSCF), respectively. The all-planar
transition structure has the least overlap of the orbitals and
exhibits the highest energy conrotatory ring-opening transition
state. The breaking C-C bond is elongated to 2.26 Å at the
MP2 and 2.41 Å with CASSCF. More rotation is seen in the
outward and inward rotating hydrogens (138° and -67° at the
MP2 (133° and -74° at the CASSCF), respectively) in the all-
planar transition structure.
Transition Structure of Substituted Cyclobutenes. Forcing
the bonds attached to alkene carbons C1 and C2 of the
cyclobutene to be coplanar should narrow the gap between the
biradical and the allowed transition state from 7.5-9.5 kcal/
mol to 4-7 kcal/mol according to these model calculations.
Substituents can also further narrow this gap. For example, Curry
and Stevens have shown that the electrocyclic opening of cis-
3,4-dimethylcyclobutene has a 1.5 kcal/mol higher activation
energy than cyclobutene as a result of the stabilization by
outward rotating methyl, but destabilization by inward rotation.38
Consequently, a cis-3,4-dimethyl and planar-constrained struc-
tures, such as 7 and 10, should have conrotatory transition
structures destabilized by a sum of 4.6 kcal/mol using the MP2
calculation and experimental values, which results from 3.1 kcal/
mol (for H-planar constraint) and 1.5 kcal/mol (for cis dimethyl
substitution). The stabilization resulting from substitution of a
methyl group on a primary radical center is 3.3 kcal/mol.39 The
other biradical center becomes a secondary terminus of an allyl
radical, which should be stabilized by about 1.7 kcal/mol.
Because of the similar destabilizing and stabilizing factors, this
conrotatory disrotatory energy gap
∆H
rxn
CAS CAS-MP2 CAS CAS-MP2 CAS CAS-MP2 CAS CAS-MP2
cyclobutene 36.6 40.4 50.0 59.7 13.4 19.2 16.0 8.9
9 (C2)
9 (C1)
12 (C1)
33.0 38.9 38.3 48.2
32.7 38.4
31.7 37.2 35.9 45.2
5.4
4.1
9.2 28.0 20.7
8.1 36.7 29.4
would make the conrotatory and monorotatory pathways about
the same in energy. Consequently, some “forbidden” product
could be produced through the biradical pathway. The trans
isomers 8 and 11 should still have a 4.0 kcal/mol allowed
mechanistic advantage.
The activation energies of ring-opening reactions of cy-
clobutene (1 and 2), bicyclo[3.2.0]hept-1(5)-ene (9), and
tricyclo[4.2.1.02,5]non-2(5)-ene (12) are listed in Table 9. The
geometries of stationary points were calculated by the CAS-
(4,4)/6-31G(d) level. The disrotatory transition states of these
compounds have two negative eigenvalues for their force-
constant matrix. The eigenvector of one negative eigenvalue
corresponds to the reaction coordinate, and the other corresponds
to an a′ symmetry coordinate. The second-order saddle point
of 9 with C2 symmetry has two negative eigenvalues in the force
constant matrix. By reducing the symmetry to C1, we obtained
the real transition state. The energy of the transition state with
C1 symmetry is only 0.5 kcal/mol lower that that with C2
symmetry.
The conrotatory stationary state of 9 with C2 symmetry is
1.5 kcal/mol more stable than 1 at the CAS-MP2 calculation
level. On the other hand, the disrotatory ring-opening of 9 has
an activation energy that is 11.5 kcal/mol more stable than that
for 2. The energy of reaction of 9 is -20.7 kcal/mol according
to the CAS-MP2 calculation and is 11.8 kcal/mol more
exothermic than that for cyclobutene. The similar activation
9
5846 J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003

Altering the Gap in Cyclobutene Reactions
A R T I C L E S
Figure 4. Square of CI coefficients of the CiLC-IRC along the disrotatory
Figure 3. Square of CI coefficients of the CiLC-IRC along the conrotatory
pathway of butadiene.
pathway of butadiene.
energy of the disrotatory pathway and the energy of reaction
between 9 and the parent system suggest that there is only a
small effect on the disrotatory pathway when planarity is
introduced to the system and that only the conrotatory pathway
is affected by this distortion.
The conrotatory pathway activation energy of 12 is 1.2 kcal/
mol more stable than that of 9 with C1 symmetry. The
disrotatory activation energy of 12 is 3.0 kcal/mol lower than
that of 9. The ring-opening process of 12 is 8.7 kcal/mol more
exothermic than that of 9. The energy difference of the
disrotatory pathway between 12 and 9 is smaller than the
difference between 2 and 9. This suggests that the mechanism
of the disrotatory pathway of 12 may be different from that of
9 and/or 2. These mechanisms will be discussed in a subsequent
section.
To investigate the constraints imposed on the system, we
compare the transition structures for the conrotatory electrocyclic
openings of 9 and 12. These transition structures are shown in
Figure 2. The C-C1-C2-C torsion angle of the cyclopentene
fused system is 7.2° in the transition structure for 9, as compared
to 23.4° in the unconstrained transition structures. Fusion to
norbornene reduces this dihedral angle to 3.6°. The 3.8 kcal/
mol ∆∆Gq conrotatory preference for 10 is somewhat lower
than expected, which may indicate that there is a small barrier
to formation of the biradical analogue to 3 and hence the
thermodynamically more stable, disrotatory product.
CiLC-IRC Analysis. To study the mechanisms for the
electrocyclic reactions of butadiene, 9 and 12, the CiLC-IRC
analysis was performed for the conrotatory and the disrotatory
pathways of three of the molecules studied here. This method
examines the nature of the electronic state along the path from
cyclobutene to butadiene. The relative weights of the electronic
configurations along the IRC pathways for the parent system
are shown in Figures 3 and 4. The results for 9 and 12 are given
in the Supporting Information. The most important electronic
configurations in these figures are depicted in Figure 5. In Figure
5, a dotted line denotes a triplet coupling (antibonding) between
the orbitals, and an ellipse denotes an ionic coupling (polariza-
tion). As shown previously,7,35 a chemical bond can be described
as the singlet coupling between two orbitals plus two polariza-
tion terms as shown in Figure 6.
Figure 5. Some electronic configurations of the reaction pathway.
Figure 6. Coupling and polarization representations.
C1dC4 and C2dC3 double bonds (Figure 5). As shown in Figure
3, the square of the coefficient for this configuration increases
dramatically as cyclobutene opens to butadiene. Each config-
uration set of 12 and 15 and 11 and 16 represent the polarization
parts of the π orbitals for C1-C4 and C2-C3, respectively.
Therefore, two sets of configurations, {19, 12, 15} and
{19, 11, 16}, can be described as the π bonds for C1-C4 and
C2-C3, respectively. Configuration 2 can be described as the
C1dC2 double bond and C3-C4 σ bond. Therefore, the
configuration sets {2, 5, 6} and {2, 3, 4} can be described as
the π bond for C1-C2 and the σ bond for C3-C4. For the ring-
opening of cyclobutene, Figures 3 and 4 show that the π bonds
for C1-C4 and C2-C3 (primarily configuration 19) increase
smoothly from the cyclobutene side to the transition state, and
the π bond for C1-C2 and the σ bond for C3-C4 (primarily
configuration 2) decreases from the cyclobutene to the butadiene.
The relative weights of these bonds cross at a point on the
butadiene side of the transition state. The “electronic transition
state”, reflected in this crossing, occurs after the “energetic
transition state”, which has the highest energy barrier on the
reaction pathway.
For the comparison of the conrotatory and disrotatory
pathways, the exchange of old π bonds of C1-C4 and C2-C3
with the new σ bond of C3-C4 occurs in a “concerted” fashion
Configuration 19 represents the singlet coupling part of the
π orbitals for C1-C4 and C2-C3, or could be described as the
9
J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003 5847

A R T I C L E S
Lee et al.
for the conrotatory pathway and a “stepwise” fashion for the
disrotatory pathway. For the conrotatory pathway, the relative
weights of configurations 3 and 4, which are the polarization
terms of σ bond for C3-C4, fade out after the appearance of
the C1-C4 and C2-C3 π-bonds of butadiene. On the other hand,
the relative weights of configurations of the π bonds for C1-
C4 and C2-C3 and the polarization terms of σ bond for C3-C4
are almost zero at the transition state for the disrotatory pathway.
The mechanisms for conrotatory and disrotatory pathways of 9
and 12 were analyzed in the same way (see the Supporting
Information for correlation diagrams).
unmethylated systems (4.1 kcal/mol for 12 and 5.4 kcal/mol
for 9). The decrease in the energy gap between the allowed
and forbidden pathway is attributed to the decrease in favorable
orbital interaction between both the σ and π* orbitals and the
π and σ* orbitals in the conrotatory transition state upon
introduction of planar constraints. The Roth hypothesis explains
how the large energy of concert in the electrocyclic ring-opening
of cyclobutene can be eroded by geometric constraints. In
general, skeletal structures that prevent the transition state from
achieving the cyclic delocalization that is intrinsic to pericyclic
transitions states40 will reduce, or even eliminate, the energy of
concert of a pericyclic reaction.
Conclusion
Acknowledgment. We are grateful to the National Science
Foundation for financial support of this research and to
Professors Dieter Hasselmann and Donald Aue for helpful
discussions.
This study supports both experimentally and theoretically that
the allowed-forbidden gap of cyclobutene ring-opening is
dependent on the planarity of the ring-opening transition
structure, as initially proposed by Roth and Ho¨rstermann in the
late 1970s. As planarity is introduced for the conrotatory
transition structure for cis-3,4-dimethylcyclobutene fused to
either norbornene or cyclopentene, the energy gap between the
allowed and forbidden pathway is reduced. The reduction in
energy difference between the allowed and forbidden processes
is evidenced by the experimental ∆∆Gq obtained from the
amount of disrotatory products formed: 3.8 kcal/mol for 10
and 4.7 kcal/mol for 7, a reduction from 10.7 kcal/mol from
the cis-3,4-dimethylcyclobutene studied by Brauman and Ar-
chie.11 These experimental values are further supported by the
activation energies obtained by CAS calculations for the
Supporting Information Available: Cartesian coordinates of
transition structures of the parent system, cyclopentyl, and
norbornyl fused systems optimized by CASSCF and MP2. List
of experimental techniques, NMR and IR data for intermediates
and products, Arrhenius plots, and CiLC-IRC analysis for
conrotatory and disrotatory ring-opening of cyclobutene fused
to cyclopentane and norbornene (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
JA028963G
(40) Houk, K. N.; Li, Y.; Evanseck, J. D. Angew. Chem., Int. Ed. Engl. 1992,
31, 682-708.
9
5848 J. AM. CHEM. SOC. VOL. 125, NO. 19, 2003