Torquoselectivity of Electrocyclic Ring Opening
J . Org. Chem., Vol. 61, No. 2, 1996 645
Sch em e 8
Sch em e 9
and the solution turned brown, presumably due to the
formation of HI and I2.
diene:Z-diene) to increase the proportion of the more
stable E-diene 20. This ratio is the result of thermody-
namic equilibration of the initial kinetically determined
ratio. After about 6 h, the E-diene 20 was the only
detectable product. The similar thermodynamic control
of stereochemistry was described before for the phenyl
ketone 12.9
The influence of other Lewis acids was also investi-
gated. Aluminum trichloride and trimethylsilyl triflate
immediately afforded complex mixtures upon heating. In
contrast, catalysis with tin(II) triflate gave exclusively
the E-diene while zinc chloride or silica gel did not affect
the product ratio. Finally treatment with zinc bromide
increased the ratio of stereoisomers to approximately 1:1.
Hence experimentally, zinc iodide turned out to be the
most effective Lewis acid at reversing the torquoselec-
tivity.24
As pointed out earlier, these results are most easily
interpreted in light of the practical difficulty in detecting
the thermodynamically less stable Z-diene 21 which
isomerizes to the E-diene 20 quite rapidly with a catalytic
amount of acid, which is inevitable even at room tem-
perature.
An authentic sample of the minor product, (Z)-3,5-
hexadien-2-one, 21, was prepared for comparison in a
separate experiment as shown in Scheme 8.17-19 Al-
though the procedure for preparation of the Z-diene 21
has already been reported,20 an easier method was
developed. The 3-butyn-2-ol was coupled with vinyl
bromide catalyzed by copper(I), according to Sonogash-
ira’s procedure.17,18 The resultant enyne 22 was submit-
ted to stereoselective reduction of the triple bond,19
followed by oxidation of the resultant allylic alcohol 23
1
as depicted in Scheme 8. This H-NMR spectrum of 21
1
was identical to the reported H-NMR data.20
Con clu sion
The influence of Lewis acids on the stereochemistry of
the ring opening was also studied. The Lewis acid BF3‚
OEt2, which is soluble in organic solvents, gave only the
E-diene smoothly at room temperature on completion of
the reaction after 6 h. However, this ring opening is
easily influenced by acid isomerization to afford more
stable E-diene 20. In fact, in a separate experiment,
rapid isomerization of Z-diene 21 to E-diene 20 was
observed at room temperature upon addition of BF3‚OEt2
to 21 in C6D6 solution. Therefore two-phase reaction with
1 equiv of a solid Lewis acid, ZnI2, was performed in the
presence of a slight molar excess of Na2CO3 in deuterated
benzene. A thermolysis experiment was carried out in
the same manner as the thermolysis of 3-acetylcy-
clobutene. Upon heating this mixture for 30 min at 80
°C in an NMR probe, a reversal of torquoselectivity was
observed. The product ratio was 17:83((5) throughout
this period in favor of the Z-diene 21 (Scheme 9).22
Therefore for the same reason stated above, the kinetic
ratio is concluded to be 17:83 (E-diene:Z-diene).23 With-
out Na2CO3, rapid isomerization to the E-diene 20 occurs
The rotational selectivity on the thermal ring opening
of 3-acetylcyclobutene (1) was predicted by ab initio
calculations. Calculations correctly predicted that Lewis
acids can reverse the selectivity. All of these predictions
were verified by the experiments. This is the first
unambiguous experimental evidence that Lewis acid,
ZnI2, reverses the rotational selectivity (“torquoselectiv-
ity”) from outward rotation to inward rotation, during
the electrocyclic ring opening of a substituted cy-
clobutene.
Exp er im en ta l Section
Gen er al. 1H-NMR spectra were recorded on Bruker AM360,
AM500, or AF200 instruments. The residual proton peak of
the deuterated solvent was used as the chemical shift stan-
dard. Mass spectra were obtained on an Associated Electrical
Industries Double Focusing Mass Spectrometer Model MS-902.
All the thermolyses were carried out by variable tempera-
ture NMR in sealed tubes in an NMR probe in benzene-d6
(dried over 3 Å molecular sieves). The NMR tubes were
washed with 10% aqueous ammonium, dried in an oven, and
sealed under a nitrogen atmosphere prior to thermolyses. The
product ratios were determined directly by integration of the
singlet signals from the acetyl protons of the resulting E-diene
20 and Z-diene 21 in the 1H-NMR spectra of the reaction
mixtures. The dienes 20 and 21 were identified on the basis
(17) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett.
1975, 4467.
(18) Mushegyan, A. V.; Kinoyan, F. S. Arm. Khim. Zh. 1972, 25(1),
3.
(19) Avignon-Tropics, M.; Pougny, J . R. Tetrahedron Lett. 1989, 30,
4951.
(20) Wolfhugel, J .; Maujean, A.; Chuche, J . Tetrahedron Lett. 1973,
18, 1635.
(21) (a) LePage, T. J .; Wiberg, K. B. J . Am Chem. Soc. 1988, 110,
6642. (b) Reetz, M. T.; Hu¨llmann, M.; Massa, W.; Berger, S.; Radema-
cher, P.; Heymanns, P. J . Am Chem. Soc. 1986, 108, 2405.
(22) During this period of reaction, remarkable reaction rate ac-
celeration was not observed.
1
of the reported H-NMR spectrum data.16,20 Thermolyses were
conducted in duplicate, and the results were observed to be
reproducible.
P r epar ation of 3-Acetylcyclobu ten e (1). 2-(3-cyclobu te-
n yl)-2-m eth yl-1,3-d ith ia n e (19). 3-chlorocyclobutanecar-
boxylic acid (18) was prepared as a mixture of two stereoiso-
(23) These product ratios were determined on the basis of the
relative intensities of the singlet signals from the proton of acetyl
groups of E-diene 20 and Z-diene 21 at regular time intervals. The
plot of the ratios of 20 to 21 versus reaction time gave nearly flat lines
for both uncatalyzed and catalyzed thermolyses during the initial stage
of the ring openings. The kinetic ratios were determined by extrapolat-
ing these plots to time zero.
(24) Considering the fact that only zinc Lewis acids induced inward
rotation effectively, alternative interpretation might be possible on the
basis of the “softness” of these acids. ZnI2, ZnBr2, and BH3 are all
classified as soft Lewis acids.25 However, this argument would require
more systematic theoretical study.
(25) Pearson, R. G. Struct. Bonding (Berlin) 1993, 80, 1.