Photoenolization as a means to release alcohols

Article abstract of DOI:10.1021/jo0261193

We have designed molecules which release alcohols upon exposure to UV light independent of the reaction media, making it possible to liberate alcohols in a controlled manner in applications. Photolysis of 2-(2-isopropylbenzoyl)benzoate ester derivatives 4 in various solvents and in thin films results in the liberation of the alcohol moiety from the ester. The reaction mechanism for the release of the alcohol has been elucidated by time-resolved laser flash photolysis. Upon irradiation the triplet excited state of ketone, 4 is formed, and its lifetime can be estimated to be between 0.08 and 0.8 ns. The triplet excited state decays by efficient intramolecular H-atom abstraction to form a 1,4-biradical, 8, that has a lifetime of less than 17 ns and is trapped by molecular oxygen. In the absence of oxygen, biradical 8 intersystem crosses to form photoenols (Z)-9 and (E)-10 in a ratio of 5:2, respectively. Photoenol (Z)-9 has a lifetime of ～3000 ns in protic solvents and returns to the starting material through 1,5 intramolecular hydrogen transfer. The other isomer, (E)-10, is much longer lived (>1 ms) and releases the alcohol moiety through an intramolecular lactonization.

Full text of DOI:10.1021/jo0261193

P h otoen oliza tion a s a Mea n s To Relea se Alcoh ols
J ana Pika,† Armands Konosonoks,‡ Rachel M. Robinson,‡ Pradeep N. D. Singh,‡ and
Anna D. Gudmundsdottir*,‡
Department of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221-0172, and
Firmenich Inc., P.O. Box 5880, Princeton, New J ersey 08543
Anna.Gudmundsdottir@uc.edu
Received J une 28, 2002
We have designed molecules which release alcohols upon exposure to UV light independent of the
reaction media, making it possible to liberate alcohols in a controlled manner in applications.
Photolysis of 2-(2-isopropylbenzoyl)benzoate ester derivatives 4 in various solvents and in thin films
results in the liberation of the alcohol moiety from the ester. The reaction mechanism for the release
of the alcohol has been elucidated by time-resolved laser flash photolysis. Upon irradiation the
triplet excited state of ketone, 4 is formed, and its lifetime can be estimated to be between 0.08
and 0.8 ns. The triplet excited state decays by efficient intramolecular H-atom abstraction to form
a 1,4-biradical, 8, that has a lifetime of less than 17 ns and is trapped by molecular oxygen. In the
absence of oxygen, biradical 8 intersystem crosses to form photoenols (Z)-9 and (E)-10 in a ratio of
5:2, respectively. Photoenol (Z)-9 has a lifetime of ∼3000 ns in protic solvents and returns to the
starting material through 1,5 intramolecular hydrogen transfer. The other isomer, (E)-10, is much
longer lived (>1 ms) and releases the alcohol moiety through an intramolecular lactonization.
1. In tr od u ction
supported by the fact that chemical reduction of 2-
benzoylbenzophenone esters also yields lactone 2.8 The
reaction only takes place in the presence of electron
donors or H-donating solvents, which are necessary for
the initial H-atom abstraction by the ketone chromo-
phore, leading to the photoreduction of the benzoyl
benzoate esters.
Photoremovable protecting groups are of interest since
they have demonstrated applications in photolithogra-
phy, synthetic organic chemistry, and biochemistry.1-7
Recently, Porter et al. reported that photolysis of 2-
benzoylbenzoate esters 1 in 2-propanol yielded the cor-
responding alcohol and lactone 31 whereas irradiation of
1 in the presence of the electron donor cyclohexylamine
also released the alcohol but led to the formation of
lactone 2. Presumably these reactions take place by
photoreduction of 1 to form 2-(R-hydroxyphenylmethyl)-
benzoic ester, which then undergoes intramolecular
lactonization to release the alcohol. This mechanism is
We are interested in slow release of alcohols since they
are used as fragrances in applications such as body care
and household cleaning goods.9 One of the drawbacks of
using volatile alcohols in fragrances is that the desired
aroma is perceived for a relatively short time in applica-
tions. Thus, by forming an ester of a volatile alcohol with
2-benzoylbenzoic acid, it is possible to release the fra-
grance in a controlled manner over an extended time
period by exposure to light. The limitation of this method
is that the release of the alcohol takes place only in the
presence of an electron-donating molecule or in a solvent
that has abstractable H-atoms. Another problem is that
molecular oxygen quenches the excited state of the
benzophenone chromophore. Consequently we decided to
design a protecting group for alcohols that will release
the alcohol upon exposure to light, independent of the
reaction media, and thus make the photorelease of
alcohols feasible in film-type applications. The excited
state of the protecting group should be sufficiently short
lived that it is not quenched by oxygen, ensuring that
* To whom correspondence should be addressed.
† Firmenich Inc.
‡ University of Cincinnati.
(1) (a) J ones, P. B.; Porter, N. A. J . Am. Chem. Soc. 1999, 121, 2753.
(b) J ones, P. B.; Pollastri, M. P.; Porter, N. A. J . Org. Chem. 1996, 61,
9455.
(2) Pelliccioli, A. P.; Wirz, J . J . Photochem. Photobiol. Sci. 2002, 1,
441.
(3) (a) Zabadal, M.; Pelliccioli, A. P.; Klan, P.; Wirz, J . J . Phys. Chem.
A 2001, 105, 10329. (b) Pelliccioli, A. P.; Klan, P.; Zabadal, M.; Wirz,
J . J . Am. Chem. Soc. 2001, 123, 7931. (c) Klan, P.; Zabadal, M.; Heger,
D. Org. Lett. 2000, 2, 1569.
(4) (a) Lee, K.; Falvey, D. E. J . Am. Chem. Soc. 2000, 122, 9361. (b)
Banerjee, A.; Lee, K.; Yu, Q.; Fang, A. G.; Falvey, D. E. Tetrahedron
Lett. 1998, 39, 4635. (c) Banerjee, A.; Falvey, D. E. J . Am. Chem. Soc.
1998, 120, 2965.
(5) (a) Conrad, P. G.. II; Givens, R. S.; Hellrung, B.; Rajesh, C. S.;
Ramseier, M.; Wirz, J . J . Am. Chem. Soc. 2000, 122, 9346. (b) Rajesh,
C. S.; Givens, R. S.; Wirz, J . J . Am. Chem. Soc. 2000, 122, 611. (c)
Givens, R. S.; J ung, A.; Park, C.-H.; Weber, J .; Bartlett, W. J . Am.
Chem. Soc. 1997, 119, 8369. (d) Park, C.-H.; Givens, R. S. J . Am. Chem.
Soc. 1997, 119, 2453.
(8) (a) Ramachandran, P. V.; Chen, G.-M.; Brown, H. C. Tetrahedron
Lett. 1996, 37, 2205. (b) Kamochi, Y. Kudo, T. Bull. Chem. Soc. J pn.
1992, 65, 3049. (c) Tobia, D.; Rickborn, B. J . Org. Chem. 1986, 51,
3849. (d) Yokoe, I.; Higuchi, K.; Shirataki, Y.; Komatsu, M. Chem.
Pharm. Bull. 1981, 29, 894. (e) Ullmann. J ustus Liebigs Ann. Chem.
1896, 24. (f) Rotering. J ahresber. Fortschr. Chem. Verw. Theile Anderer
Wiss. 1875, 596.
(6) Rochat, S.; Minardi, C.; de Saint Laumer, J .-Y., Herrmann, A.
Helv. Chim. Acta 2000, 83, 1645.
(7) Epling, G. A.; Provatas, A. A. Chem. Commun. 2002, 1036.
(9) Pika, J .; Herrmann, A.; Vial, C. U.S. Patent 6133228, 2000.
10.1021/jo0261193 CCC: $25.00 © 2003 American Chemical Society
Published on Web 01/31/2003
1964
J . Org. Chem. 2003, 68, 1964-1972

Photoenolization as a Means To Release Alcohols
the reaction will take place in air. We have therefore
synthesized derivatives of 2-benzoylbenzoate esters 4
with an isopropyl substituent ortho to the ketone that
allows for an intramolecular H-atom abstraction. For
synthetic reasons we also prepared compounds that
contained isopropyl groups in the para position. In this
paper we describe the photoreactivity of these com-
pounds, via intramolecular H-atom abstraction to form
the Z and E photoenols, which, theoretically, then can
undergo intramolecular lactonization to release the al-
cohol moiety. We are particularly concerned with how the
reactivity of the Z and E photoenols differs and whether
both isomers release the protected alcohol. We used time-
resolved laser flash photolysis experiments to elucidate
the reaction mechanism. For comparison we performed
the laser flash photolysis experiments using the analo-
gous 2-benzoylbenzoate geraniol ester, 1a , and 2-benzoyl-
benzoic acid, 1b, which cannot undergo intramolecular
H-atom abstractions.
5, whereas irradiations of 4a ,b with UV lamps or daylight
in air yielded the corresponding alcohol and peroxide 6
(see Figure 1). By comparison, photolysis of 1 in benzene
did not yield any detectable products, and irradiation of
1 in chloroform gave only a trace amount of products.
Irradiation of acid 1b in 2-propanol produced 3 in a
manner similar to that of the photolysis of 1a in 2-pro-
panol that gave geraniol and 3 (see Figure 2).1 Photolysis
of 1 in toluene resulted in 7 as the major product. In
Figure 3 are plotted the rates of conversion of 1a and 4a
to their respective photoproducts in toluene, 2-propanol,
benzene, and chloroform. The reactivity of 4a is not
affected significantly by the solvent, whereas 1a only
releases geraniol in appreciable amounts in solvents that
have abstractable H-atoms, such as toluene and 2-pro-
panol. Quantitative analysis of the photolysis of 4a in
toluene under argon showed that the photoproducts
geraniol and 5 are formed in similar amounts (see Figure
4). The depletion of the starting material 4a is quantita-
tive within experimental error to the formation of the
photoproducts, which indicates a clean reaction without
formation of polymeric tar.
Thin films of 4a were prepared in volumetric flasks
which were sealed under ambient air and exposed to
daylight. After 5 h the contents of the flask were diluted
with a solvent and analyzed by GC-FID and GC-MS,
which showed geraniol was released. In contrast, irradia-
tion of thin films of 1a did not release geraniol.
2.2. Qu a n tu m Yield s. We measured the quantum
yields (see Table 1) for the photoreactivities of 4a and
1a in 2-propanol using valerophenone as the actinom-
eter.10 The quantum yields for depletion of the starting
materials are similar to those for formation of products,
2. Resu lts a n d Discu ssion
2.1. P h otod ecom p osition of 1 a n d 4. Photolysis of
4a ,b under argon in toluene, 2-propanol, benzene, and
chloroform gave the corresponding alcohol and lactone
F IGURE 1. Photolysis of 4.
F IGURE 2. Photolysis of 1b.
J . Org. Chem, Vol. 68, No. 5, 2003 1965

Pika et al.
F IGURE 3. Conversions of 1a (broken lines) and 4a (solid lines) upon irradiation in various solvents under argon.
F IGURE 5. Phosphorescence spectra of 1a and 4a . The inset
shows the decay of the phosphorescence of 1a measured at
420 nm.
F IGURE 4. Quantitative measurement of formation of ge-
raniol from photolysis of 4a in toluene under argon.
TABLE 1. Qu a n tu m Yield s for F or m a tion of
P h otop r od u cts fr om a n d Dep letion s of 1 a n d 4
1a
4a
Φ for depletion
Φ for formation of products
0.62
0.17 (0.20)a
0.62b
0.14c (0.17)a,d
a
The number in parentheses is the quantum yield measured
b
under an O2 atmosphere. Quantum yield for formation of 3.
c Quantum yield for formation of 5. Quantum yield for formation
d
of 6.
indicating that the reaction is very clean without forma-
tion of high molecular weight polymers. In 2-propanol,
the quantum yields for 1a depletion are much higher
than for 4a . The quantum yield for product formation
from 4a increased slightly when the solution was bubbled
with oxygen prior to photolysis.
2.3. P h osp h or escen ce. The phosphorescence spectra
of 1 and 4 were obtained in ethanol glasses at 77 K. The
spectra for 1a and 4a are displayed in Figure 5. These
spectra have vibrational bands as typically observed for
triplet excited ketones with n,π* configuration. The 0,0
band can be estimated to be around 406 and 414 nm for
1a and 4a , respectively; thus, the triplet energy values
of the ketones are 70.4 and 69.1 kcal/mol, respectively,
similar to the triplet energy of benzophenone itself.11 The
yield of the phosphorescence emission from 4 is 5 times
F IGURE 6. Transient spectrum of 4a in 2-propanol. The inset
shows the kinetic trace at 390 nm.
less than for 1 because it has an additional channel for
decay via intramolecular H-atom abstraction that is not
available to 1. The lifetime of the phosphorescence of 4
was shorter than the time resolution of our phosphores-
cence meter (45 µs). The decay of 1 could be observed
but not fitted to a single-exponential function because
the compound was frozen in a wide variety of conforma-
tions, each having its own specific lifetime.
2.4. La ser F la sh P h otolysis. Laser flash photolysis
of 4a at 308 nm in dichloromethane or ethanol under
argon gave a broad transient absorption between 320 and
480 nm with a λmax at 390 nm (see Figure 6). On the basis
of the similarity of this spectrum to the transient spectra
of Z and E photoenols of triisopropylbenzophenone and
2-methylbenzophenone,12,13 we assigned this absorption
(10) Wagner, P. J .; Kochevar, I. E.; Kemppainen, A. E. J . Am. Chem.
Soc. 1972, 94, 7489.
(11) Murov, S. T.; Carmichael, I.; Hug, G. L. Handbook of Photo-
chemistry, 2nd ed.; Marcel Dekker: New York, 1993.
1966 J . Org. Chem., Vol. 68, No. 5, 2003

Photoenolization as a Means To Release Alcohols
SCHEME 1
to photoenols (Z)-9 and (E)-10 (see Scheme 1). The
absorption spectrum had two components. Spectra ob-
tained 100 and 5000 ns after the laser pulse were similar
except for slightly enhanced absorptions at lower wave-
lengths at the longer time. In the kinetic mode we
followed the decay of the shorter lived component and
found that it had a lifetime of ∼1300 ns in dichloro-
methane and accounted for approximately 75% of the
absorption. Its lifetime was prolonged to ∼3000 ns in
hydrogen-bonding solvents such as ethanol and 2-pro-
panol. The longer lived component did not decay signifi-
cantly over a 100 µs time scale. We proposed that the
short-lived component was the (Z)-9 photoenol that can
reketonize through 1,5 hydrogen transfer, whereas the
E enol was longer lived since it could only undergo
reketonization through proton transfer from the solvent.
We did not detect the absorption of the triplet excited
state of ketone 4a ; however, we did succeed in quenching
the triplet excited ketone with isoprene. By adding
isoprene, we reduced the yield of photoenols (Z)-9 and
(E)-10, whereas their lifetimes were not affected. Isoprene
F IGURE 7. Stern-Volmer graph for the quenching of the
triplet excited state of 4 with isoprene in 2-propanol.
did not affect the lifetime of photoenols (Z)-9 and (E)-10
since presumably their triplet energies are very low, but
reduced their yields by quenching of the triplet excited
state of the ketone precursor. Stern-Volmer treatment
of the quenching of the absorbance at 390 nm with
isoprene in dichloromethane gave a linear plot with a
slope of 0.8 (see Figure 7). By assuming that the energy
transfer from the triplet excited state of the ketone to
isoprene was diffusion controlled, or kq is between 1 ×
109 and 10 × 109 M-1 s-1, the lifetime of the triplet ketone
could be estimated to be between 0.08 and 0.8 ns.
(12) Haag, R.; Wirz, J .; Wagner, P. J . Helv. Chim. Acta 1977, 60,
2595.
(13) Guerin, B.; J ohnston, L. J . Can. J . Chem. 1989, 67, 473.
J . Org. Chem, Vol. 68, No. 5, 2003 1967

Pika et al.
F IGURE 8. Transient spectra of 1a in 2-propanol and
dichloromethane. The inset shows the kinetic trace at 530 nm.
SCHEME 2
F IGURE 9. Absolute minimal energy conformations of 4c and
biradical 8c.
obtained IR spectra. Upon irradiation the spectra did not
change. Presumably, in the matrix, only those conforma-
tions where the isopropyl group is located close to the
ketone react due to the restrictions imposed on molecular
motion by the matrix. Since o-alkyl-substituted aryl
ketones have been shown to undergo H-atom abstraction
at low temperatures via quantum mechanical tunnel-
ing,16 we expected 4c to react similarly to form biradical
8. The matrix must restrict the rotation of biradical 8c,
and thus, it could only intersystem cross to form (Z)-9c
but not (E)-10c.17 Since the intramolecular lactonization
was much slower than reketonization of (Z)-9c, we did
not observe any product formation in the argon matrices.
2.6. Ab In itio Ca lcu la tion s. We did ab initio calcula-
tions using Gaussian 98 and optimized the structures of
ester 4c, biradical 8c, (Z)-9c, and (E)-10c. For reasons
of economy we did these calculations only for ester 4c
and its photoproducts and presumed that the minimal
energy conformations of 4a ,b, 8a ,b, (Z)-9a ,b, and (E)-
10a ,b are very similar. We wanted to confirm that, in
the ground state conformation of ester 4, the o-isopropyl
group was in close proximity to the ketone chromophore
and to determine whether any low-energy conformations
of 4 where the o-isopropyl group faced away from the
ketone existed. In the optimized structure of 4 the phenyl
ring bearing the ester group was in conjugation with the
ester group, because the torsion angle between the CdO
in the ester group and the aromatic ring was ∼30° (see
Figure 9). The ketone CdO group was more conjugated
to the isopropyl-substituted phenyl, since the torsion
angle between the ketone CdO and the isopropyl-
substituted phenyl group was ∼30° whereas the torsion
angle between the ketone CdO and the ester-substituted
phenyl was ∼60°. The methine H-atom of the o-isopropyl
group was within 2.3 Å of the carbonyl ketone oxygen.
We looked for other minimal energy conformations of
Laser flash photolysis of 4 under oxygen gave reduced
yields of photoenols (Z)-9 and (E)-10 without shortening
the lifetime of (Z)-9, presumably because we trapped
biradical 8 with molecular oxygen. The triplet excited
state of 4 is too short lived to be quenched by molecular
oxygen, whereas biradicals are known to react with
oxygen with a rate constant in excess of 109 M-1 s-1 14
.
Laser flash photolysis of 1a and 1b in dichloromethane
under argon yielded transient absorptions with maxima
at λ 330 and 530 nm (see Figure 8). We assigned these
absorptions to the triplet excited ketones because of their
similarity to the triplet excited state of benzophenone.15
In 2-propanol the absorption maximum was shifted to
590 nm instead of 550 nm, due to the absorption of the
ketyl radical that formed upon hydrogen atom abstrac-
tion from the solvent (see Scheme 2). In dichloromethane
the lifetimes of the triplet excited states of 1a and 1b
were 100 and 130 ns, respectively. In 2-propanol these
lifetimes were reduced to 35 and 50 ns, for 1a and 1b,
respectively, due to intermolecular H-atom abstraction
by the triplet excited ketones. In the kinetic mode the
absorption did not decay fully to zero absorbance in
2-propanol because the ketyl radical absorbed in the same
region as the triplet excited state of 1.
2.5. Ma tr ix Isola tion of P h otolysis of 4c. We
synthesized ester 4c because it is more volatile than ester
4a or 4b. We deposited ester 4c in argon matrices and
(16) (a) J ohnson, B. A.; Garcia-Garibay, M. A. J . Am. Chem. Soc.
1999, 121, 8114. (b) Garcia-Garibay, M. A. Gamarnik, A.; Bise, R.;
J enks. W. W. J . Am. Chem. Soc. 1998, 102, 5491. (d) Al-Soufi, W.;
Eychmuller, A.; Grellemann, K. H. J . Phys. Chem. 1994, 95, 3033.
(17) Gebicki, J .; Krantz, A. J . Chem. Soc., Perkin. Trans 2 1984,
1623.
(14) N. J . Turro. Modern Molecular Photochemistry; Benjamin/
Cummings: New York, 1978; Chapter 14.
(15) Beckett, A.; Porter, G. Trans. Faraday Soc. 1963, 59, 2051
1968 J . Org. Chem., Vol. 68, No. 5, 2003

Photoenolization as a Means To Release Alcohols
because it was short lived due to efficient intramolecular
H-atom abstraction, but the lifetime was determined to
be between 0.08 and 0.8 ns by quenching with isoprene.
Since we observed a linear Stern-Volmer plot for the
quenching of the triplet excited state of 4 with isoprene,
the triplet excited state must be able to reach rotational
equilibrium before it reacts. Therefore, any populated
conformations of 4 where the o-isopropyl group is not in
close proximity to the ketone chromophore must be able
to rotate in the excited state before decaying. At low
temperatures the phosphorescence yields for 4 were 5
times less than for 1. This is consistent with our calcula-
tions which demonstrated that 4 existed mainly in a
lowest energy conformation at 77 K that favored efficient
intramolecular H-atom abstraction. The intramolecular
H-atom abstraction in 4 resulted in biradical 8, which
was trapped by oxygen. Since we did not observe biradical
8 with the laser flash apparatus, we can assume that it
has a lifetime less than the time resolution of the laser
flash apparatus or 17 ns. The calculated minimal energy
conformation of biradical 8 was very similar to the
minimal energy conformations of 4 and photoenol (Z)-9.
Biradical 8 was, however, long lived enough to inter-
system cross, forming both (Z)-9 and (E)-10. It is reason-
able to assume that the ratio of the E/Z photoenols must
be a reflection of the populations of the various biradical
8 conformations that decay into either the Z or E
photoenol depending on their structure or conformational
alignment.
F IGURE 10. Minimal energy conformations of (Z)-9c and (E)-
10c.
ketone 4c where the isopropyl group was facing away
from the carbonyl group and found one that was 1.5 kcal/
mol higher in energy than the absolute minimal energy
conformation of 4c.
We optimized the structure of the triplet biradical 8c.
Its minimal energy conformation was very similar to the
one for ester 4c (see Figure 9). The hydroxyl proton was
localized 2.85 Å from the isopropyl carbon radical, which
was twisted out of conjugation with the phenyl ring. Both
the phenyl groups were rotated approximately 30° out
of the plane of the ketyl radical.
We optimized the structures of the photoenols since
we were particularly interested in determining whether
the minimal energy conformations of (Z)-9 and (E)-10
have intramolecular H-atom bonding and the energy
difference between these two isomers. In the optimized
structure of photoenol (Z)-9c, there was H-atom bonding
between the hydroxyl group and the carbonyl group of
the ester (see Figure 10). The distance between the
hydroxyl proton and the ester carbonyl oxygen was 1.96
Å, and the angle O-H‚‚‚O was 135°. The hydroxyl proton
was located 3.66 Å away from the isopropyl vinyl carbon
atom, making 1,5 hydrogen transfer feasible with some
minor conformational changes of the molecule. By com-
parison, the optimized structure of (E)-10c predicted a
much lower probability for the 1,5 hydrogen transfer
since the distance between the O-H group and the
isopropyl vinyl carbon atom was 5 Å. As in the case of
(Z)-9c, the hydroxyl proton and the ester group CdO of
(E)-10c were hydrogen bonded since the distance between
them was only 1.90 Å with an O-H‚‚‚O angle of 137°.
Our calculations showed that (Z)-9c was only 0.6 kcal/
mol more stable than (E)-10c.
(Z)-9 had lifetimes of ∼3000 and 1300 ns in 2-propanol
and dichloromethane, respectively. This was comparable
to the lifetime of the Z photoenol of methylacetophenone
in protic solvents.3 The main difference between (Z)-9 and
the methylacetophenone Z photoenol was that the latter
was much shorter lived in nonprotic solvents where
hydrogen bonding between the solvent and the hydroxy
group did not retard reketonization. According to our
calculations, (Z)-9 had intramolecular H-atom bonding
between the hydroxyl group and the ester carbonyl group
which delayed reverse H-atom transfer in nonprotic
solvents. The involvement of the hydroxyl group in the
intramolecular hydrogen bond was, nonetheless, not
sufficient to impede the reverse H-atom transfer long
enough to release the alcohol moiety of (Z)-9. By com-
parison, (E)-10 was much longer lived than (Z)-9 because
it could not revert to 4 through intramolecular 1,5
hydrogen transfer. Presumably, (E)-10 reacted with
molecular oxygen, but since this was a ground-state
reactivity, it was much slower than the trapping of
biradical 8 with oxygen, and we did not observe it under
our experimental conditions in the laser flash apparatus.
The rate of trapping of the E photoenol of 2,4-dimethyl-
benzophenone with dimethyl acetylenedicarboxylate has
been measured to be 102 M-1 s-1 18
.
Irradiation of 4 in various solvents led to efficient
release of the alcohol moiety of the molecule. Similarly,
exposure of 4 in thin films to sunlight also released the
alcohol. In the absence of air the major photoproduct was
5, whereas in the presence of oxygen the major product
was 6. Peroxide 6 must be formed by trapping both
biradical 8 and (E)-10 with oxygen. The resulting per-
oxide could then release the alcohol moiety through
3. Discu ssion
The primary photoreaction of 4 is photoenolization, and
the proposed mechanism for the photorelease is shown
in Scheme 1. We did not observe the triplet state of
ketone 4 with the laser flash photolysis apparatus
(18) Porter, G.; Tchir, M. F. J . Chem. Soc. A 1971, 3772.
J . Org. Chem, Vol. 68, No. 5, 2003 1969

Pika et al.
arc lamp was filtered through a basic solution in potassium
chromate to isolate the 313 nm emission band. The disappear-
ance of the starting material and the formation of photo-
products were analyzed by GC using hexadecane as an internal
standard and valerophenone as an actinometer.8
Ca lcu la tion s. The geometries of all species were optimized
by the Hartree-Fock method as implemented in the Gaussian
98 programs, using the 6-31G basis set.20
P h osp h or escen ce. The phosphorescence spectra were
obtained on a phosphorometer that has a time resolution of
45 µs.
La ser F la sh P h otolysis. Laser flash photolysis was carried
out with an excimer laser (308 nm, 17 ns). The system has
been described in detailed elswhere.21
intramolecular lactonization. It is also possible that the
oxygen trapping took place after the intramolecular
lactonization of (E)-10. In the absence of oxygen (E)-10
underwent conrotatory electrocyclization and released
the alcohol to form cyclobutene 5.19 The quantum yields
for product formation upon photolysis of 4 were low
because the major photoproduct (Z)-9 reversed back to
the starting material without releasing the alcohol.
Quantum yields for formation of products from 4a in the
presence of O2 were somewhat higher than in the absence
of O2 because biradical 8 was trapped before it could
convert into (Z)-9, which did not yield any products.
Similarly, 4b did not show any reactivity in argon
matrices although the ground-state conformation of this
molecule favored H-atom abstraction, presumably be-
cause in these matrices only (Z)-9 was formed.
By comparison, irradiation of 1 in dichloromethane or
in thin films did not result in significant product forma-
tion, but photolysis of 1 in hydrogen-abstractable solvents
such as 2-propanol led to release of the alcohol or water.
The lifetime of the triplet excited state of 1 was, as
expected, much longer than that of ketone 4. In solvents
such as toluene and 2-propanol the triplet excited state
of 1 decayed by intermolecular H-atom abstraction to
form a ketyl intermediate. In toluene the ketyl radical
was intercepted by the benzyl radical formed by inter-
molecular H-atom abstraction to form 7, whereas in
2-propanol two ketyl radicals combined to form dimer 3.
In 2-propanol the quantum yield for formation of products
from 1 was much higher than from 4.
Ma tr ix Isola tion s. Matrix isolation studies were done
using conventional equipment.22
P r ep a r a tion of Sta r tin g Ma ter ia ls. 2-Benzoylbenzoic
acid is commercially available.
Ger a n yl 2-Ben zoylben zoa te (1a ). The geraniol ester of
2-benzoylbenzoic acid was prepared according to the litera-
1
ture.1 The H NMR, IR, and MS spectra of 1a fit with the ones
in the literature.1 IR (neat): 2925, 1720, 1675, 1450, 1280,
1125 cm-1. 1H NMR (360 MHz, CDCl3): δ 8.06 (dd, 6 Hz, 1H),
7.74 (d, 7 Hz, 2H), 7.56 (m, 3H), 7.42 (m, 3H), 5.07 (m, 2H),
4.53 (d, 9 Hz, 2H), 1.97 (m, 4H), 1.67 (s, 3H), 1.59 (s, 3H), 1.55
(s, 3H) ppm. CILR-MS (NH3): m/z 380 [M + 18]+.
2-(2′,4′-Diisop r op ylben zoyl)ben zoic Acid . Phthalic an-
hydride (19.3 g, 0.13 mol) was placed in a flame-dried three-
neck round-bottom flask under nitrogen. 1,2-Dibromoethane
(100 mL) and aluminum chloride (36.0 g, 0.27 mol) were added,
and the resulting solution was stirred at room temperature
while 1,3-diisopropylbenzene (20.4 g, 0.126 mol) was added
dropwise. The reaction mixture was stirred at 100 °C for 2 h
and then poured over ice/hydrochloric acid (1:1). The solution
was extracted twice with dichloromethane. The organic extract
was washed with a saturated aqueous sodium chloride solu-
tion, dried over anhydrous sodium sulfate, and concentrated
under vacuum to yield an oil which was characterized as being
2-(2′,4′-diisopropylbenzoyl)benzoic acid (isolated yield 36 g,
74%).
4. Con clu sion
We have made protective groups that release alcohols
upon exposure to UV light independent of the reaction
media and are therefore ideal for slow fragrance release
in applications. The release takes place through intra-
molecular H-atom abstraction and formation of biradical
8 that intersystem crosses to form photoenols (Z)-9 and
(E)-10. The shorter lived photoenol (Z)-9 did not result
in product formation but reverted back to the starting
material through 1,5 hydrogen transfer. In contrast, (E)-
10 released the alcohol through an intramolecular lac-
tonization. The quantum yields for release of the alcohol
from 4 were low since the major photoenol (Z)-9 did not
yield any products. Interestingly, we can trap the 1,4-
biradical 8 with molecular oxygen. The trapped biradical
released the alcohol, and thus, the photorelease was not
decreased in atmospheric air. By comparison the photo-
release from 1 was not possible unless the reaction took
place in solvents that can undergo hydrogen abstraction.
In a solvent such as 2-propanol or toluene the reaction
was very efficient. We are currently designing systems
that release alcohols through photoenolization but where
the intramolecular lactonization is more efficient with
the expectation that the Z photoenol will yield products
as well as the E photoenol.
IR (neat): 2965, 1695, 1670, 1605 cm-1. 1H NMR (360 MHz,
CDCl3): δ 7.98 (dd, 1 and 8 Hz, 1H), 7.59 (m, 1H), 7.52 (m,
1H), 7.37 (dd, 1 and 8 Hz, 1H), 7.31 (d, 1 Hz, 1H), 7.09 (d, 8
Hz, 1H), 6.94 (dd, 2, 8 Hz, 1H), 3.82 (m, 1H), 2.91 (m, 1H),
1.25 (m, 12H) ppm. 13C NMR (90 MHz, CDCl3): δ 198.6 (s),
170.9 (s), 153.2 (s), 151.0 (s) 143.8 (s), 133.9 (s), 132.3 (d), 131.7
(d), 130.6 (d), 129.8 (d), 128.9 (s), 128.7 (d), 125.0 (d), 122.7
(d), 34.3 (d), 29.0 (d), 24.1 (q), 23.7 (q) ppm. LREIMS: m/z
(relative abundance) 310 (5, M+), 265 (43), 249 (45), 221 (100),
149 (32), 84 (41), 49 (35).
Ger a n yl 2-(2′,4′-Diisop r op ylb en zoyl)b en zoa t e (4a ).
2-(2′,4′-Diisopropylbenzoyl)benzoic acid (1.15 g, 3.7 mmol) was
dissolved in dry pyridine (10 mL) in a flame-dried three-neck
round-bottom flask. To the solution were added geraniol (0.5
g, 3.6 mmol), 4-(dimethylamino)pyridine (0.10 g, 0.8 mmol),
and 1,3-dicyclohexylcarbodiimide (0.76 g, 3.7 mmol). The
resulting solution was stirred overnight, poured onto a mixture
(20) Gaussian 98, Revision A.9: Frisch, M. J .; Trucks, G. W.;
Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J . R.;
Zakrzewski, V. G.; Montgomery, J . A., J r.; 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.; Baboul, A. G.; 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.; Challacombe, M.; Gill,
P. M. W.; J ohnson, B.; Chen, W.; Wong, M. W.; Andres, J . L.; Gonzalez,
C.; Head-Gordon, M.; Replogle, E. S.; Pople, J . A., Gaussian, Inc.,
Pittsburgh, PA, 1998
5. Exp er im en ta l Section
Qu a n tu m Yield s. Solutions of ester 1 or 4 (∼10-2 M) and
valerophenone in 2-propanol were irradiated simultaneously
in a “merry-go-round” apparatus. The light from a mercury
(21) Gritsan, N. P.; Zhai, H. B.; Yuzawa, T.; Karweik, D.; Brooke,
J .; Platz, M. S. J . Phys. Chem. A 1997, 101, 2833.
(22) Ault, B. S. J . Am. Chem. Soc. 1978, 100, 2426.
(19) Wagner, P. J .; Subrahmanyam, D.; Park, B. S. J . Am. Chem.
Soc. 1991, 113, 709.
1970 J . Org. Chem., Vol. 68, No. 5, 2003

Photoenolization as a Means To Release Alcohols
of shaved ice (20 g), hydrochloride acid (32%, 24 g), and ethyl
acetate (30 mL), and stirred vigorously for 10 min. This
mixture was extracted twice with diethyl ether and the organic
layer washed twice with a saturated aqueous sodium bicar-
bonate solution, twice with water, dried over anhydrous
sodium sulfate, and concentrated under vacuum. The residual
oil was purified by silica gel column chromatography eluted
with 20% diethyl ether in heptane to yield geranyl 2-(2′,4′-
diisopropylbenzoyl)benzoate as a pale yellow oil (isolated yield
1.08 g, 75%).
tion of 4b (40 mg, 0.12 mmol) in toluene (5 mL) was saturated
with oxygen and irradiated with a mercury arc lamp through
a Pyrex filter. GC analysis of the reaction mixture showed
formation of product 6 and remaining starting material in a
ratio of 1:10, respectively. The solvent was removed under
vacuum and the resulting mixture purified on silica gel eluted
with ethyl acetate/hexane (1:10). Column chromatography
yielded a pure fraction of 6 which was crystallized from ethanol
to yield colorless needles (23 mg, 68% yield).
Mp: 142-143 °C. IR (neat, gas phase): 2970, 1810, 1290,
940 cm-1 1H NMR (360 MHz, CDCl3): δ 7.98 (m, 1H), 7.67
.
IR (neat): 2965, 1725, 1670 cm-1 1H NMR (360 MHz,
.
(m, 2H), 7.28 (m, 1H), 7.12 (d, 2 Hz, 1H), 7.03 (dd, 2 and 8 Hz,
1H), 6.63 (d, 8 Hz, 1H), 2.92 (m, 1H), 1.81 (s, 3H), 1.63 (s, 3H),
1.25 (d, 7 Hz, 6H) ppm. 13C NMR (90 MHz, CDCl3): δ 168.0
(s), 150.8 (s), 145.2 (s), 141.6 (s), 134.7 (d), 131.6 (d), 128.0 (s),
126.6 (d), 125.6 (d), 125.4 (s), 125.3 (d), 124.0 (d), 121.7 (d),
106.9 (s), 81.3 (s), 34.2 (d), 27.7 (q), 25.1 (q), 23.9 (q), 23.8 (q)
ppm. LREIMS: m/z (relative abundance): 324 (M+, 3), 292
(38), 264 (18), 249 (45), 221 (100), 189 (39). HRMS: m/z calcd
for C20H21O4, [M + H]+, 325.1440, found 325.1447.
CDCl3): δ 7.76 (dd, 3, 6 Hz, 1H), 7.51 (m, 2H), 7.37 (dd, 3, 6
Hz, 1H), 7.31 (d, 2 Hz, 1H), 7.18 (d, 8 Hz, 1H), 6.97 (dd, 2, 8
Hz, 1H), 5.22 (m, 1H), 5.04 (m, 1H), 4.64 (d, 7 Hz, 2H), 3.79
(m, 1H), 2.92 (m, 1H), 2.1-1.9 (m, 4H), 1.74 (br s, 3H), 1.62
(br s, 3H), 1.58 (br s, 3H), 1.28 (d, 7 Hz, 6H), 1.25 (d, 7 Hz,
6H) ppm. 13C NMR (90 MHz, CDCl3): δ 198.5 (s), 167.2 (s),
152.9 (s), 150.6 (s), 142.7 (s), 142.4 (s), 134.1 (s), 131.7 (s), 131.2
(s), 131.6 (d), 131.1 (d), 129.9 (d), 129.6 (d), 128.7 (d), 124.7
(d), 123.8 (d), 122.8 (d), 117.9 (d), 62.3 (t), 39.5 (t), 34.3 (d),
29.2 (d), 26.3 (t), 25.7 (q), 24.1 (q), 24.1 (q), 23.7 (q), 23.7 (q),
17.7 (q), 16.4 (q) ppm. LREIMS: m/z (relative abundance) 446
(M+, <0.5), 309 (100), 265 (29), 249 (52), 231 (28), 221 (49),
149 (52), 93 (34), 69 (55), 41 (53). HRMS: m/z calcd for
Sp ir o(3′-isop r op yl-8′,8′′-d im e t h ylb icyclo[4,2,0]oct a -
1(b),2,4-tr ien e)-3-isoben zofu r a n -1-on e (5). A degassed so-
lution of 4b (33 mg, 0.10 mmol) in 2-proponal (3 mL) was
irradiated using a medium-pressure mercury arc containing
a Pyrex filter at 25 °C for 16 h. The reaction was monitored
using thin-layer chromatography (10% ethyl acetate in hex-
ane). After completion of the reaction, the solvent was evapo-
rated and the product was purified using column chromatog-
raphy. The resulting solid was recrystallized from pentane to
yield needles of 5 (yield after recrystallization 18 mg, 62%).
Mp: 104-105 °C. IR (neat): 2961, 2869, 1760, 1466, 1230
C
30H38O3Na, [M + Na]+, 469.2756, found 469.2719.
Meth yl 2-(2′,4′-d iisop r op ylben zoyl)ben zoa te (4b) was
formed by refluxing 2-(2′,4′-diisopropylbenzoyl)benzoic acid (11
g, 0.03 mol) in methanol. The resulting oil was purified by
silica gel column chromatography eluted with 20% ethyl
acetate in hexane (2.2 g, 22%).
1
IR (neat): 2962, 1728, 1671, 1605, 1288, 943, 769 cm-1. H
1
cm-1. H NMR (250 MHz, CDCl3): δ 7.9 (d, 1H), 7.5 (m, 2H),
NMR (250 MHz, CDCl3): δ 7.54 (m, 2H), 7.42 (m, 1H), 7.33
(s, 1H), 7.16 (d, 8 Hz, 1H), 6.98 (d, 8 Hz, 1H), 3.78 (m, 7 Hz,
1H), 3.69 (s, 3H), 2.93 (m, 1H), 7.86 (m, 1H) 1.30 (d, 7 Hz,
6H), 1.26 (d, 7 Hz, 6H) ppm. 13C NMR (60 MHz, CDCl3): δ
198.5, 167.8, 153.2, 150.7, 142.7, 131.6, 131.5, 134.2, 131.0,
130.2, 129.8, 129.1, 124.9, 123.0, 52.5, 34.5, 29.4, 24.3, 23.9
ppm. MS (EI): m/z (relative intensity) 324 (M+, 6), 309 (14),
292 (11), 265 (33), 249 (58), 231 (20), 221 (100), 193 (16), 178
(18), 163 (52), 115 (13), 91 (14), 77 (22). HRMS: m/z calcd for
7.1 (m, 3H,), 6.9 (d, 1H), 3.0 (septet, 1H), 1.29 (s, 6H), 1.26 (d,
6H) ppm. 13C NMR (60 MHz, CDCl3): δ 170.5, 152.5, 149.0,
138.3, 133.2, 129.1, 127.2, 126.4, 125.2, 124.0, 122.8, 118.2,
92.1, 55.9, 35.0, 24.1, 23.7 ppm. GC-MS: m/z (relative
intensity) 292 (M+, 5), 249 (50), 221 (100). HRMS: m/z calcd
for C20H21O2, [M + H]+, 293.1542, found 293.1536.
3-Ben zyl-3-p h en yl-3H-isoben zofu r a n -1-on e (7). A de-
gassed solution of 1b (100 mg, 0.44 mmol) in toluene (20 mL)
was irradiated with a medium-pressure arc lamp through a
Pyrex filter. GC analysis of the reaction mixture showed
formation of product 7 and remaining starting material in a
ratio of 95:5, respectively. The solvent was removed under
vacuum and the resulting oil purified on a silica column eluted
with 20% ethyl acetate in hexane. The resulting solid was
recrystallized from ethanol (80 mg, 61% yield).
C
21H24O3Na, [M + Na]+, 347.1618, found 347.1664.
Meth yl 2-(2′-Isop r op ylben zoyl) ben zoa te (4c). To a
stirred solution of phthalic anhydride (1.69 g, 11.4 mmol) in
anhydrous benzene (80 mL) was added 2-isopropylphenyl-
magnesium iodide, prepared from 2-isopropyliodobenzene (2.80
g, 11.4 mmol), magnesium turnings (0.35 g, 14.4 mmol), and
a crystal of iodine in anhydrous ether (50 mL). Over a 15 min
period at room temperature, a yellow solid formed. The
mixture was refluxed for 12 h and poured onto ice/hydrochloric
acid and the resulting mixture extracted with ether (2×50 mL).
The organic phase was washed two times with brine, dried
over magnesium sulfate, and evaporated to give 2-(2′-iso-
propylbenzoyl)benzoic acid as a dark oil. Without further
purification, the material was dissolved in methanol (50 mL)
and titrated with diazomethane in ether and the resulting
solution evaporated. The residue was purified by silica gel
column chromatography to yield methyl 2-(2′-isopropyl-
benzoyl)benzoate as a yellow oil which eventually solidified
and was recrystallized from hexane (yield after recrystalliza-
tion 220 mg, 7%).
Mp: 95-6 °C (lit.23 97-8 °C). IR (CHCl3): 1770, 1287 cm-1
.
1H NMR (250 MHz, CDCl3): δ 7.7-7.5 (m, 5H), 7.4-7.2 (m,
4H), 7.1-7.0 (m, 3H), 7.0-6.9 (m, 2H), 3.60 (d, 15 Hz, 1H),
3.40 (d, 15 Hz, 1H) ppm. HRMS: m/z calcd for C21H16O2Na,
[M + Na]+, 323.1048, found 323.1017.
P r od u ct Stu d ies fr om P h otolysis of 1 a n d 4. P h otolysis
of 4a in Va r iou s Solven ts u n d er Ar gon . Solutions of 4a in
chloroform (0.015 M), benzene (0.015 M), and toluene (0.015
M) were placed in test tubes, degassed with argon, and
photolyzed simultaneously for 4 h with a mercury arc lamp
through a Pyrex filter. The contents of the reaction mixtures
were analyzed with a GC chromatograph, and in all instances
formation of geraniol and 5 were detected along with some
unreacted 5a . The test tubes were degassed again with argon
and photolyzed for an additional 4 h and their contents
analyzed with a GC chromatograph to reveal further formation
of geraniol and 5.
P h otolysis of 4a in Va r iou s Solven ts in Air . Solutions
of 4a in toluene (0.01 M) and benzene (0.01 M) were photolyzed
using an 8 W 366 nm lamp. After 10 h of irradiation GC
analysis of the reaction mixture showed 40% conversion and
formation of 6 and geraniol.
Mp: 71-72.5 °C. IR (neat): 2965, 1728, 1671, 1287, 931,
1
762 cm-1. H NMR (250 MHz, CDCl3): δ 7.84 (m, 1H), 7.41-
7.56 (m, 5H), 7.13-7.26 (m, 2H), 3.70 (m, 7 Hz, 1H), 3.70 (s,
3H), 1.29 (d, 7 Hz, 6H) ppm. 13C NMR (60 MHz, CDCl3): δ
198.7, 167.9, 150.2, 142.1, 136.8, 131.9, 131.4, 130.7, 130.6,
129.8, 129.4, 126.7, 125.2, 52.6, 29.5, 24.3 ppm. MS (EI): m/z
(relative intensity) 282 (M+, 3), 267 (12), 250 (66), 231 (33),
222 (44), 207 (67), 193 (68), 181 (78), 178 (58), 163 (100), 152
(33), 103 (30), 91 (37), 77 (71). HRMS: m/z calcd for C18H19O3,
[M + H]+, 283.1334, found 283.1339.
Isola tion of P h otop r od u cts. Sp ir o (3′-isop r op yl-8′,8′′-
d im eth ylben zod ioxin )-3-isoben zofu r a n -1-on e (6). A solu-
(23) Ismail, M. F.; Enayat, E. I.; El-Bassiouny, F. A. A.; Younes, H.
A. Gazz. Chim. Ital. 1990, 120, 677.
J . Org. Chem, Vol. 68, No. 5, 2003 1971

Pika et al.
Qu a n tita tive An a lysis for F or m a tion of P h otop r od -
u cts fr om P h otolysis of 4a in Solu tion s. To solutions of
4a in benzene (0.0067 M) and toluene (0.011 M) was added
an external standard, hexadecane. These solutions were
degassed with argon and photolyzed with a mercury arc lamp
through a Pyrex filter. The quantities of geraniol, 5, and 4a
were determined after 4, 10, and 12 h of photolysis by using
GC analysis.
flasks were diluted quantitatively with solvent and analyzed
for release of geraniol by GC-FID. The quantity of geraniol
released was determined by using GC analysis relative to an
external standard of geraniol. Films of 4a that were exposed
to daylight for 5 h released 5 mol % geraniol, whereas films
kept in the dark did not yield any geraniol.
P h otolysis of Th in F ilm s of 1a . Films of 1a were prepared
as described for 4a . Prolonged photolysis of these films did
not lead to any detectable release of geraniol.
P h otolysis of 1a in Va r iou s Solven ts. Solutions of 1a in
chloroform (0.017 M), benzene (0.017 M), 2-propanol (0.017 M),
and toluene (0.017 M) were placed in test tubes, degassed with
argon, and photolyzed simultaneously for 6 h with a mercury
arc lamp through a Pyrex filter. The reaction mixtures were
analyzed with a GC chromatograph, which revealed that the
reaction mixture of 1a in toluene contained 7, geraniol, and
unreacted starting material. The photolyzed sample of 1a in
2-propanol showed a small amount of remaining starting
material and formation of geraniol and 3. The irradiated
solution of 1a in benzene did not show any product formation,
only unreacted starting material. GC analysis of the photo-
lyzed solution of 1a in chloroform showed mainly unreacted
starting material and formation of dimer 3 and geraniol in
trace amounts. The test tubes were degassed with argon and
photolyzed for an additional 2 h and their contents analyzed
by GC chromatography. The reaction mixtures in 2-propanol,
chloroform, and benzene remained similar to those at the
earlier time point, whereas photolysis of 1a in toluene went
to further completion.
Ack n ow led gm en t. We thank the University of
Cincinnati Childcare and the Child Development Cen-
ter, Bristol-Myers Squibb, for their support. Thanks are
due to Christian Vial and Herve´ Pamingle for their
assistance with the synthesis of the esters. We give our
appreciation to Walter Thommen and Robert Brauchli
for help with the characterization of the photoproduct
of geranyl 2-(2′,4′-diisopropylbenzoyl)benzoate. We also
owe thanks to Danielle Crammer, Shannon Fenton,
Raphael Grenno, and J acques Sulzer for their help with
compound preparation and the photolysis assays. We
are grateful to Professor Matt Platz at The Ohio State
University for allowing us to use some of his instru-
mentation. We are indebted to Professor Bruce Ault at
the University of Cincinnati for his help with the matrix
isolation studies.
P h otolysis of Th in F ilm s of 4a . Films of 4a were prepared
by adding a pentane solution of 4a (20 mg, 0.05 mmol) to
volumetric flasks and removing the solvent under a stream of
nitrogen while turning the flasks slowly to distribute the ester
evenly on the surface of the glass. The flasks were then sealed
under ambient atmosphere, half of them exposed to daylight
for 5 h, and the others kept in the dark. The contents of the
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
4a , 4b, 4c, 5, and 6 and Cartesian coordinates, number of
imaginary frequencies, and total energy of 4c, 8c, (Z)-9c and
(E)-10c. This material is available free of charge via the
Internet at htpp://pubs.acs.org.
J O0261193
1972 J . Org. Chem., Vol. 68, No. 5, 2003