A novel and selective photoisomerization of allylic benzoates

Article abstract of DOI:10.1021/ol016657m

(matrix presented) Ultraviolet irradiation of a series of polyunsaturated allylic benzoates results in highly selective E/Z interconversion of the allylic double bond.

Full text of DOI:10.1021/ol016657m

ORGANIC
LETTERS
2001
Vol. 3, No. 22
3547-3548
A Novel and Selective
Photoisomerization of Allylic Benzoates
Tao Hu and E. J. Corey*
Department of Chemistry and Chemical Biology HarVard UniVersity, 12 Oxford Street,
Cambridge, Massachusetts 02138
corey@chemistry.harVard.edu
Received August 27, 2001
ABSTRACT
Ultraviolet irradiation of a series of polyunsaturated allylic benzoates results in highly selective E/Z interconversion of the allylic double bond.
We describe herein an interesting and, to the best of our
knowledge, novel photoisomerization of the ethylenic linkage
in benzoate esters of allylic alcohols. For example, irradiation
of (E)-geranyl benzoate (1) in a methylcyclohexane solution
(3) was transformed cleanly in 40 min to a 48:52 mixture of
2,6,10-(E,E,E)- and 2,6,10-(Z,E,E)-geranylgeranyl benzoates.
Finally, irradiation of (all-E)-nonaprenyl (solanasyl) benzoate
(4) for 40 min under identical conditions resulted in clean
and highly selective photoisomerization of the 2E-olefinic
linkage to form a mixture of 2E- and 2Z-isomers in a ratio
of 48:52. Under these conditions, the more remote ethylenic
linkages in 2, 3, and 4 did not undergo detectable E/Z
photointerconversion.
Irradiation of (E)-geranyl 4-methoxybenzoate also resulted
in E/Z equilibration of the 2,3-olefinic linkage (E/Z ratio 45:
55 after 20 min and 42:58 after 40 min) as did irradiation of
(E)-geranyl 3-methoxybenzoate (E/Z ratio 74:26 after 20 min,
67:33 after 40 min and 62:38 after 70 min). In contrast,
irradiation of either (E)-geranyl 2-, 3-, or 4-nitrobenzoate
for 3.5 h under the same conditions led to no detectable
isomerization of the 2-olefinic linkage or other reaction as
(0.01-0.04 M under Ar or N2; UV max, 227, 273, 280 nm;
ꢀ, 7270, 600, 500) at 0 °C using a medium pressure mercury
vapor lamp in a Penn-Orr microreactor1 (available from Ace
Glass Co.) resulted in photoequilibration of the 2,3-ethylenic
linkage in 1-2 h with formation of a mixture of E- and
Z-isomers in a ratio of 41:59. The reaction was extremely
clean since no other products could be detected by careful
thin layer chromatographic and 500 MHz 1H NMR analysis.
Similarly, irradiation of 2,6-(E,E)-farnesyl benzoate (2) for
40 min under the same conditions afforded cleanly a 55:45
mixture of 2(E),6(E)- and 2(Z),6(E)-farnesyl benzoates by
selective stereomutation of the allylic olefinic linkage. Under
the same conditions 2,6,10-(E,E,E)-geranylgeranyl benzoate
1
determined by TLC and 500 MHz H NMR analysis.
We have also studied the photoisomerization reaction with
the 4-methoxybenzoate esters of (E)- and (Z)-2-octen-1-ol
(5 and 6, respectively). Irradiation of these esters in a 0.02
M solution in methylcyclohexane at 0 °C under Ar led
cleanly to the same 74:26 E/Z mixture of 5 and 6 after either
20 or 40 min. This photoisomerization was also carried out
under an atmosphere of O2 with no significant change in
the reaction as compared with that for an Ar atmosphere.
Indeed, it was also observed that the photoisomerizations of
substrates 1-4 as described above proceeded just as well
under O2 as with an Ar atmosphere. As a further example,
we investigated the irradation of the E,Z-triene benzoate 7.
(1) Penn, J. H.; Orr, R. D. J. Chem. Educ. 1989, 66, 86.
10.1021/ol016657m CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/29/2001

Under the standard conditions which were applied for the
photoisomerizations of 1-6, benzoate 7 underwent clean Z/E
isomerization selectively at the 6-olefinic linkage of the
allylic benzoate subunit. The Z/E ratios were found to be
30:1 (initial composition) at zero time, 60:40 after 12 min,
46:54 after 30 min, and 44:56 after 55 min of irradiation.
Scheme 1
toaddition reaction with olefins to form oxetanes4 which is
probably a two-step cycloaddition process via 1,4-diradi-
cals.6,7
Upon prolonged irradiation of farnesyl benzoate (2), E f
Z isomerization can be observed not only in the 2-olefinic
linkage but also in the 6-olefinic bond. The much slower
isomerization about the more remote 6-olefinic bond of 2
would be consistent with either a photochemical pathway
such as that shown in Scheme 1 or a photophysical (energy
or electron transfer)3 process, both of which are expected to
be slower than for the allylic double bond. The remote
isomerization also is clearly intramolecular since it occurs
at the same rate at concentrations from 0.02 to 0.002 M.
Because of the facile photoisomerization of allylic ben-
zoates reported above, we have also examined the photo-
chemical stability of farnesyl acetate which shows an
apparent absorption max at 238 nm (ꢀ ) 280) in the
methylcyclohexane solution. Irradiation of farnesyl acetate
(0.02 M in methylcyclohexane) with the same apparatus as
used for the allylic benzoates 1-7 resulted in a slow but
definite photoisomerization of the 2-ethylenic linkage to give
an 88:12 E/Z mixture after 1 h and an 83:17 E/Z mixture
after 2 h. The low-intensity 238 nm absorption of farnesyl
acetate may correspond to an electronic transition which is
enhanced by the proximity of acetate and olefinic functions,
since it does not appear in the spectrum of ethyl acetate (end
absorption only at 238 nm, ꢀ ca. 1). In any event that
excitation could lead to photoisomerization by any of the
paths discussed above.
The exploratory studies described herein have revealed an
interesting new aspect of the photoisomerization of ethylenic
linkages. In the absence of extensive physical investigations,
the mechanistic details of these isomerizations must remain
a matter of conjecture. However, there would appear to be
two broad types of processes which may be involved. The
first is a photophysical mechanism involving electronic
excitation of the benzoyl moiety2 followed by energy transfer
to the nearby allylic double bond and subsequent rotation
about its σ-axis to effect E/Z interconversion. Such a process
could involve either singlet or triplet electronically excited
states. The fact that the photoisomerization of substrates 1-6
is not inhibited by an atmosphere of O2 somewhat favors
the involvement of singlet states (Fo¨rster energy transfer)3
rather than intramolecular triplet-triplet energy transfer,4
although it is certainly not decisive. An alternative photo-
physical pathway for isomerization involves electron transfer
from CdC to the electronically excited benzoyl group to
form a radical cation which can undergo facile C-C rotation
and electron return.5 The second type of process is a
photochemical mechanism, one version of which is shown
in Scheme 1. This reversible photochemical pathway gener-
ates a new chemical intermediate, such as the diradical shown
in Scheme 1 which can undergo facile rotation about the
σ-component of the bond undergoing stereomutation. Clearly,
the lifetime of any chemical intermediate must be short
compared to its rate of capture by O2, in view of the lack of
any photoreaction of substrates 1-6 with O2 under our
standard conditions. There is some precedent for the pathway
outlined in Scheme 1 in previously reported photoreactions
of methyl benzoate. The most relevant is the [2 + 2]-pho-
In summary, we have observed an intriguing new type of
stereomutation of nonconjugated olefinic linkages involving
photoexcitation of allylic benzoates or other esters. The
mechanistic aspects of this reaction invite further studies
which could deepen our understanding of organic photo-
chemistry of a little studied class of compounds.
Acknowledgment. We are grateful to the National
Science Foundation and Schering-Plough for financial support.
OL016657M
(2) Ultraviolet absorption of benzoate esters in MeOH at 280 nm (ꢀ ∼
1000) excites an ππ* transition (1A1g
f
1B2u) and that at 230 nm (ꢀ ∼
14 000) excites an intramolecular charge-transfer transition. Either or both
of these excitations may be involved in the photochemistry described herein.
See: (a) Tanaka, J.; Nagakura, S.; Kobayashi, M. J. Chem. Phys. 1956,
24, 311. (b) Tanaka, J.; Nagakura, S. J. Chem. Phys. 1956, 24, 1274. (c)
Nagakura, S.; Tanaka, J. J. Chem. Phys. 1954, 22, 236.
(3) For a discussion, see: Klessinger, M.; Michl, J. Excited States and
Photochemistry of Organic Molecules; VCH Publications: New York, 1995;
pp 291-295.
(4) (a) Cantrell, T. S. J. Chem. Soc., Chem. Commun. 1973, 468. (b)
Cantrell, T. S.; Allen, A. C. J. Org. Chem. 1989, 54, 135.
(5) Kavarnos, G. J.; Turro, N. J. Chem. ReV. 1986, 86, 401.
(6) See ref 3, pp 424-432.
(7) Another interesting photoreaction of methyl benzoate is the photo-
accelerated reduction by sodium borohydride. See: (a) Choi, J. H.; Kim,
D. W.; Shim, S. C. Tetrahedron Lett. 1986, 27, 1157. (b) Wigfield, D. C.;
Feiner, S.; Gowland, F. W. Tetrahedron Lett. 1976, 3377. For this reaction
Choi et al. report that the quantum yield for reduction of methyl benzoate
(0.50) is greater than that for methyl 4-methoxybenzoate (0.03) or methyl
4-nitrobenzoate (0.05).
3548
Org. Lett., Vol. 3, No. 22, 2001