Type II photoreaction of benzoyl- And (benzoylmethyl)cyclohexanones: Cyclohexanone ring inversion and other factors determining their photoreactivities

Article abstract of DOI:10.1002/poc.1008

X-ray crystallographic structure analysis showed that a molecule of 2-benzoylcyclohexanone (2), which undergoes an efficient Type II reaction in solution but not in the crystalline state, is in a keto and chair form with an equatorial benzoyl group. Irradiation of 3-benzoylcyclohexanone (3) gives 7-phenylhepten-4,7-dione (4, 61%) and a mixture of cyclization products 5 (25%). High photoreactivity of 2 and 3 is caused by the efficient cyclohexanone ring inversion. Upon irradiation, 2-(benzoylmethyl)cyclohexanone (6) gives acetophenone (7), 2-cyclohexenone (8) and the cyclobutanol 9 in 59, 59 and 35% yield, respectively, and 1-phenyl-3-propylpentan-l,4-dione (10) gives 7 and the cyclobutanol 11 in 73 and 4% yield, respectively. The quantum yields for the disappearance of 6 and 10 are 0.68 and 0.67, respectively. Irradiation of 2-(benzoylmethyl)-2-ethoxycarbonylcyclohexanone (12) gives 7 and the cyclobutanol 13 in 46 and 51 % yield, respectively. The quantum yield for the disappearance of 12 is 1.00, so that reverse hydrogen transfer is suppressed by intramolecular hydrogen bonding in the 1,4-biradical intermediate. When 6 is irradiated on a dry silica gel surface, the Type II reaction occurs to give 7, 8 and 9 in 38, 38 and 43% yield, respectively. The cyclization in the photoreaction of (benzoylmethyl)cyclohexane also increases from 68% in benzene to 81% on a dry silica gel surface. Restriction of conformational change in the 1,4-biradical is an important factor for the cyclization. Copyright

Full text of DOI:10.1002/poc.1008

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2006; 19: 122–128
Published online 9 December 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/poc.1008
Type II photoreaction of benzoyl- and
(benzoylmethyl)cyclohexanones: cyclohexanone
ring inversion and other factors determining
their photoreactivities
Tadashi Hasegawa,* Yiming Yang, Shuichi Kikuchi, Takayuki Nakamura and Yutaka Maeda
Department of Chemistry, Tokyo Gakugei University, Nukuikitamachi, Koganei, Tokyo 184-8501, Japan
Received 30 May 2005; revised 5 October 2005; accepted 11 October 2005
ABSTRACT: X-ray crystallographic structure analysis showed that a molecule of 2-benzoylcyclohexanone (2), which
undergoes an efficient Type II reaction in solution but not in the crystalline state, is in a keto and chair form with an
equatorial benzoyl group. Irradiation of 3-benzoylcyclohexanone (3) gives 7-phenylhepten-4,7-dione (4, 61%) and a
mixture of cyclization products 5 (25%). High photoreactivity of 2 and 3 is caused by the efficient cyclohexanone ring
inversion. Upon irradiation, 2-(benzoylmethyl)cyclohexanone (6) gives acetophenone (7), 2-cyclohexenone (8) and
the cyclobutanol 9 in 59, 59 and 35% yield, respectively, and 1-phenyl-3-propylpentan-1,4-dione (10) gives 7 and the
cyclobutanol 11 in 73 and 4% yield, respectively. The quantum yields for the disappearance of 6 and 10 are 0.68 and
0.67, respectively. Irradiation of 2-(benzoylmethyl)-2-ethoxycarbonylcyclohexanone (12) gives 7 and the cyclobu-
tanol 13 in 46 and 51% yield, respectively. The quantum yield for the disappearance of 12 is 1.00, so that reverse
hydrogen transfer is suppressed by intramolecular hydrogen bonding in the 1,4-biradical intermediate. When 6 is
irradiated on a dry silica gel surface, the Type II reaction occurs to give 7, 8 and 9 in 38, 38 and 43% yield,
respectively. The cyclization in the photoreaction of (benzoylmethyl)cyclohexane also increases from 68% in benzene
to 81% on a dry silica gel surface. Restriction of conformational change in the 1,4-biradical is an important factor for
the cyclization. Copyright # 2005 John Wiley & Sons, Ltd.
KEYWORDS: Type II reaction; phenyl ketones; photoreaction in solution; surface photoreaction; hydrogen abstraction;
ring inversion; conformational flexibility; reverse hydrogen transfer
INTRODUCTION
solution and even in the solid state.6 As ꢀ-hydrogen
abstraction in 1 occurs only from the conformer having
an axial benzoyl group, cyclohexane ring inversion must be
an important factor determining the photoreactivity.5 Upon
irradiation, 2-benzoylcyclohexanone (2) undergoes only
Type II elimination7,8 with a ‘true’ quantum yield of unity
for the disappearance of the keto form of 2; the enol forms
of 2 and the Type II elimination product act as internal
filters, so that the observed quantum yields for the disap-
pearance of 2 are less than unity and depend on the reaction
time.8 Introduction of a methyl group on the 2-position of 2
provides a remarkable changeover in the reaction course;
2-methyl-2-benzoylcyclohexanone undergoes Type II cy-
clization exclusively.9 We report here the factors determin-
ing photochemical behavior of phenyl ketones having the
cyclohexanone chromophore.
The Norrish Type II photoreaction of alkyl aryl ketones
having a ꢀ-hydrogen has been extensively studied. Inter-
vention of a 1,4-biradical intermediate, which undergoes
elimination and cyclization, is now well established.1 The
length of the alkyl chain hardly affects the cyclization
percentage in the Type II reaction of simple alkyl
aryl ketones, e.g. 12, 18, 20 and 20% in butyro-, valero-,
capro- and octyrophenone, respectively.2 (Aroylmethyl)-
cyclohexanes also undergo the Type II reaction.3 The
photoreactivity of the cyclohexanes is almost the same as
that of the simple ketones2,4 in solution. Although the
cyclohexanes undergo the Type II reaction even in the solid
state3c,d and in the polymer-bound form,3a benzoylcyclo-
hexane (1) shows fairly low photoreactivity.5 Introduction
of an isopropyl group on the 1-position of 1 increases the
Type II reactivity and allows the compound to cyclize in
RESULTS AND DISCUSSION
*Correspondence to: T. Hasegawa, Department of Chemistry, Tokyo
Gakugei University, Nukuikitamachi, Koganei, Tokyo 184-8501,
Japan. E-mail: tadashi@u-gakugei.ac.jp
Ground state molecular conformations influence photo-
chemical behavior in cases where excited state reactions
Contract/grant sponsor: Shinsei Foundation.
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 122–128

TYPE II PHOTOREACTION OF BENZOYL AND (BENZOYLMETHYL)CYCLOHEXANONES
123
Scheme 1
are more rapid than conformational transformation.10
High ground state population of the unreactive conformer
having an equatorial benzoyl group and fast radiationless
decay processes competing with ring inversion from the
equatorial to the photoreactive axial conformer should
result in low photoreactivity of 1 (Scheme 1).5c The high
photoreactivity of 2 may arise from a distorted chair
conformation which may be the most stable conformer
and allow the excited benzoyl carbonyl access to ꢀ-
hydrogen. If that is the case, the cyclohexanone 2 should
show high Type II photoreactivity in the crystalline state.
However, when crystals of 2 were irradiated with a 450 W
high-pressure mercury lamp, no reactions were observed.
This might suggest that the most stable conformer of 2 is
the equatorial conformer and the ring inversion is sup-
pressed in the crystalline state. Therefore, an x-ray
crystallographic structure analysis of 2 was performed.
The x-ray analysis showed that a molecule of 2 is in a
keto and chair form having an equatorial benzoyl group
as shown in Fig. 1. The distance between the migrating
hydrogen and carbonyl oxygen has been considered to be
the most important parameter for intramolecular hydro-
gen migration and the theoretically ideal value is
Scheme 2
cyclohexanone ring system. The activation energy for
ring inversion of cyclohexanone is known to be lower
than that of cyclohexane; the substitution of a methylene
group in cyclohexane to a carbonyl group leads to the
elimination of torsional strain during the ring inversion.12
The idea of the rapid cyclohexanone ring inversion as a
factor determining the photoreactivity of 2 is also sup-
ported by the following experiment.
Irradiation of a benzene solution of 3-benzoylcyclo-
hexanone (3) under nitrogen gave the Type II elimination
product, 7-phenylhepten-4,7-dione (4, 61%), and a
mixture of cyclization products 5 (25%) (Scheme 3).
The mixture could not be separated into its components.
The IR spectrum of the mixture showed the characteristic
absorption of the cyclohexanone carbonyl and hydroxyl
groups at 1740 and 3440 cmꢁ 1, respectively. Its mass
spectrum showed a molecular ion signal at m/z 202.
These results support the structure of 6-hydroxy-6-
phenylbicyclo[3.1.1]heptan-3-one. However, the 1H
NMR spectrum was complex and the 13C NMR spectrum
showed four signals attributable to carbonyl groups at ꢁ
197.3, 199.4, 200.2 and 210.1. As there are at least four
possible isomers for the Type II cyclization products,3 we
deduced that the mixture consists of four 6-hydroxy-6-
phenylbicyclo[3.1.1]heptan-3-one isomers. The Type II
reaction should occur from the axial conformer 3a. The
ꢀ 2.7 A.11 The distances between benzoyl carbonyl oxy-
˚
˚
gen and ꢀ-hydrogens in the crystals of 2 were 4.68 A
˚
[H(4) in Fig. 1] and 4.81 A (H5). These values deviated
greatly from the theoretically ideal value. These results
and the high quantum yield for the disappearance of 28
strongly suggest that high photoreactivity of 2 results
from rapid cyclohexanone ring inversion (Scheme 2). The
inversion should be much more faster than the ꢀ-hydro-
gen abstraction by the excited benzoyl carbonyl. The
difference between the photoreactivities of 1 and 2 can be
reasonably explained in terms of the difference between
the efficiency of the ring inversion in the cyclohexane and
Scheme 3
Figure 1. Structure of 2-benzoylcyclohexanone (2)
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 122–128

124
T. HASEGAWA ET AL.
quantum yields for the disappearance of 3 and the
formation of 4 were determined as 0.36 and 0.22,
respectively. Although the cyclization products could
not be isolated but were obtained as a mixture of four
isomers, the Type II reactivity of 3 is obviously much
higher than that of 1 (ꢂ < 0.005).5a This strongly supports
the idea of the rapid ring inversion as the factor. The
quantum yield for the disappearance of 3 was lower than
the ‘true’ quantum yield of the keto form of 2, which
shows keto—enol tautomerization. This indicates the
absence of intramolecular hydrogen bonding causing
suppression of reverse hydrogen transfer2,8 in the 1,4-
biradical intermediate from 3.
tional flexibility owing to the presence of the cyclohex-
anone ring. However, this idea is probably excluded for
the following reasons. The quantum yields of the phenyl
ketones increase to nearly unity in polar solvents and the
reproduction of the ground state ketone from the excited
phenyl ketones in benzene occurs only via reverse hydro-
gen transfer from the 1,4-biradical intermediates.4 These
indicate that conformational flexibility is not the factor
determining the quantum yield for the disappearance of
the starting ketones in the phenyl ketones and so the
decrease in the conformational flexibility in 6 is not the
reason for its higher quantum yield compared with that of
the phenyl ketones.
Insertion of a methylene group between the cyclohex-
ane ring and the benzoyl group of 1 has been reported to
increase the Type II reactivity3 compared with 1 and the
cyclization percentage compared with that in the Type II
reaction of the simple alkyl aryl ketones.13 Therefore,
we studied the photochemical behavior of 2-(benzoyl-
methyl)cyclohexanone (6) in order to compare it with
that of 2 and clarify the effect of the inserted methylene
group.
Irradiation of a benzene solution of 6 gave acetophe-
none (7), 2-cyclohexenone (8) and the cyclobutanol 9 in
59, 59 and 35% yield, respectively. The quantum yields
for the disappearance of 6 and for production of 7 and 9
were 0.68, 0.40 and 0.24, respectively (Scheme 4). The
quantum yield for the disappearance of 6 is lower than
that of the keto form of 2. This indicates that the insertion
of the methyl group prevents the formation of the
intramolecular hydrogen bonding which causes increase
of the Type II reactivity because of suppression of reverse
hydrogen transfer reproducing the starting ketone from
the biradical intermediate.2,8
Although the intramolecular hydrogen bond producing
a seven-membered ring in the 1,4-biradical intermediate
from 3 must be formed much less efficiently than that
producing a six-membered ring in the biradical from 2,
the hydrogen bonding in the former may cause the high
Type II reactivity of 6. The photoreaction of 1-phenyl-3-
propylpentan-1,4-dione (10) was studied in order to
clarify this possibility.
Irradiation of a benzene solution of 10 gave 7 and the
cyclobutanol 11 in 73 and 4% yield, respectively
(Scheme 5). The quantum yields for the disappearance
of 10 and formation of 7 were 0.67 and 0.49, respec-
tively. The quantum yield for the disappearance of 10 is
nearly equal to that of the cyclohexanone 6 (Table 1).
Therefore, the decrease in conformational flexibility due
to the presence of the cyclohexanone ring is not an
important reason for the more efficient Type II reaction
of 6 than the simple alkyl phenyl ketones. However, the
presence of the ring increases the cyclization percentage
in the Type II reaction, the values for 6 and 10 being 37
and 5%, respectively. In order for efficient cleavage to
occur, the radical-containing p-orbitals are required to be
parallel to the carbon–carbon bond undergoing scission
(0,0 geometry).1,3d,15 The 1,4-biradical from 6 is very
unlikely to rotate to the 0,0 geometry because of ring
constraints,3d so the probability of cyclization from the
biradical would be increased. That is, the increase in
cyclization may be due to the decrease in the conforma-
tional flexibility. As the quantum yields for the Type II
reaction of 6 and 10 were high, the presence of the
carbonyl group on the ꢃ-carbon of alkyl phenyl ketones
seems to increase the Type II photoreactivity. Therefore,
the photoreaction of 2-(benzoylmethyl)-2-ethoxycarbo-
nylcyclohexanone (12) was studied.
On the other hand, the quantum yield of 6 is high
compared with those of alkyl phenyl ketones such as
butyrophenone (BP, 0.40),14,15 valerophenone (VP,
0.42)14,15 and ꢃ-methylbutyrophenone (0.30).15 This
may be explained in terms of a decrease in conforma-
Scheme 5
Scheme 4
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 122–128

TYPE II PHOTOREACTION OF BENZOYL AND (BENZOYLMETHYL)CYCLOHEXANONES
125
Table 1. Quantum and chemical yields in the Type II photo-
reactions
surface, the Type II reaction occurred as in solution to
give 7, 8 and 9 in 38, 38 and 43% yield, respectively, with
53% cyclization. The latter value is 1.4 times larger than
that in the photoreaction of 6 in benzene. Similarly, the
cyclization in the photoreaction of (benzoylmethyl)cy-
clohexane (14) increased from 68% in benzene to 81% on
a dry silica gel surface. Restriction of conformational
change in the 1,4-biradical must be an important factor
for the cyclization.
Irradiation
Ketone conditions
Cyclization
(%)
ꢂ
ꢂelim
ꢂcy
ꢁketone
6
In benzene
0.68
0.67
1.00
0.27
0.40
0.42
—
0.40
0.49
0.46
0.056
0.36
0.33
38c
0.24
0.03
0.51
0.12
0.042
0.091
37
5
53
10
12
14
BPb
VPb
6
68 (67)a
12
18
The absolute quantum yields for the Type II reaction on
the surface are difficult to determine because of scattering
of irradiated light. Therefore, the relative quantum yield
for the disappearance of 6 with respect to that of VP (ꢂ_6/
ꢂ_VP) was determined. The ratio was found to be 1.67,
which is smaller than that in benzene (1.73). The differ-
ence is not very large, but probably suggests a decrease in
the photoreactivity of 6 on the surface. This decrease in
reactivity may be due to the high polarity of the surface. It
is well known that energy levels of the electric config-
uration of ketones depend on solvent polarity and their
lowest triplet states may change from the n, ꢅ* to the ꢅ,
ꢅ* configuration in polar solvents.17 Weis et al. reported
that the polarity of silica gel is higher than that of
methanol.18 When 6 was irradiated in acetonitrile, etha-
nol or methanol, the Type II reaction products were also
formed almost quantitatively as in benzene. The ratios of
the Type II cyclization to elimination were nearly con-
stant, although the ratio of the Type II cyclization to
elimination in a polar solvent is slightly smaller than that
in a non-polar solvent [(ꢂ9/ꢂ9)MeOH/(ꢂ9/ꢂ7)PhH ¼ 0.88],
as known in the Type II reaction of phenyl ketones.2
Hence the quantum yields for the formation of 7 from 6 in
different solvents were determined in order to examine
the solvent effect on the Type II reaction of 6. The
relationship between reactivity and solvent polarity,
defined as the Z-value,19 is shown in Fig. 2. The results
indicate that the photoreactivity of 6 decreases with
increase in solvent polarity.
On SiO2
43c
55c
53
81
14
—
13c
a Ref. 13.
b Ref. 15.
c Chemical yield (%).
Irradiation of a benzene solution of 12 gave 7 and the
cyclobutanol 13 in 46 and 51% yield, respectively
(Scheme 6). The quantum yield for the disappearance of
12 was 1.00. Although ꢃ, ꢃ-dimethyl substitution in BP
has been known to decrease both the Type II photoreac-
tivity and cyclization percentage because of steric effect
in the 1,4-biradical intermediate,15 introduction of the 2-
ethoxycarbonyl group into the (benzoylmethyl)cyclohex-
anone 6 obviously increases the Type II photoreactivity
and the cyclization percentage. The percentage in the
Type II reaction of 12 (54%) is higher than those of
6 (37%) and simple alkyl phenyl ketones having no
ꢄ-substituents (ca 10–25%).2,15 The effect of the 2-ethox-
ycarbonyl group on the Type II photoreactivity can be
reasonably explained in terms of suppression of reverse
hydrogen transfer by intramolecular hydrogen bonding
between the hydroxyl and carbonyl groups in the 1,4-
biradical.2,8 The reverse hydrogen transfer in the 1,4-
biradical produced from 6 cannot be suppressed comple
tely because the quantum yield for the disappearance of
the starting ketone increased from 0.68 in 6 to 1.00 in 10.
The increase of the cyclization percentage might be also
explained in terms of the hydrogen bonding in the
biradical that causes a decrease in conformational flex-
ibility and in the possibility of rotation to the 0,0 geome-
try.
Conformational transformation may be restricted on a
silica gel surface16 and this restriction may facilitate the
cyclization of 6. When 6 was irradiated on a dry silica gel
Figure 2. Relationship between reactivity and polarity of
solvent
Scheme 6
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 122–128

126
T. HASEGAWA ET AL.
Introduction of the ethoxycarbonyl group to the 2-
position of 6 increases the Type II photoreactivity be-
cause of possible intramolecular hydrogen bonding in the
1,4-biradical intermediate. A silica gel surface provides a
polar media increasing the ꢅ, ꢅ* character in the excited
state of ketones. The surface also restricts conformational
change of the 1,4-biradical and increases the cyclization
percentage in the Type II reaction.
EXPERIMENTAL
IR, UV, mass and phosphorescence spectra were recorded
with a JASCO IR Report-100, Shimadzu UV-3150,
Varian SATURN 2000R and Hitachi F-4500 spectrometer,
respectively. 1H and 13C NMR spectra were measured with
a JEOL FX-90Q and Bruker AVANCE 300 spectrometer
using tetramethylsilane as an internal standard. The dike-
tones were prepared according to reported methods.8,20–22
Ushio 450W and Taika 100 W high-pressure mercury
lamp were used as the irradiation sources. A potassium
chromate filter solution was used for isolation of the
313 nm line.23 Gas–liquid chromatographic (GLC) analy-
sis was performed with a Shimadzu GC-8A gas chroma-
tograph equipped with a flame ionization detector which
was connected to a Shimadzu C-R6A Chromatopac inte-
grator, using a 2 m column containing 15% propylene
glycol succinate or SE-30 on Uniport B. Hexadecane,
octadecane or tetradecanol was used as calibrant for the
GLC analysis. Silica gel (Merck Kieselgel 60, Art. 7734)
was used as received for surface photoreactions.
Figure 3. UV absorption spectra of 6 in (a) hexane,
(b) cyclohexane, and (c) EtOH
Figure 3 shows the UV absorption spectra of 6.
The broad absorption peak at around 320 nm in hexane
underwent the blue shift and moved into the ꢅ, ꢅ* end
absorption in ethanol. The lifetimes and triplet energies
of 6 in cyclohexane and EPA were determined as 1.9 and
88 ms and 74.4 and 73.9 kcal molꢁ1, respectively, from its
phosphorescence spectra (Table 2). These results indicate
that the lowest triplet state of 6 in non-polar solvents is
the n,ꢅ* state and that in polar solvents is the ꢅ, ꢅ* state.
The decrease in the photoreactivity of 6 with increase in
solvent polarity can be rationalized to the increase in the
ꢅ, ꢅ* character in its excited state. The decrease in the
photoreactivity of 6 on the silica gel surface can also be
explained in terms of the increase in the ꢅ, ꢅ* character
by the highly polar surface.
In conclusion, the much higher photoreactivity of 2
than 1 can be reasonably explained in terms of rapid ring
inversion. Insertion of a methylene group between the
benzoyl group and the cyclohexanone ring of 2 increases
the Type II reactivity; the quantum yield for the disap-
pearance of 6 is higher than those of simple alkyl phenyl
ketones. Reverse hydrogen transfer should be partially
suppressed by intramolecular hydrogen bonding between
the cyclohexanone carbonyl and hydroxyl hydrogen in
the 1,4-biradical intermediate from 6. Although the
decrease in conformational flexibility due to the presence
of the cyclohexanone ring is not an important reason for
the efficient Type II reaction of 6, the presence of the ring
increases the cyclization percentage in the reaction.
X-ray structure determination of 2
A
Rigaku AFC6S diffractometer with graphite-
˚
monochromated Mo Kꢄ radiation (0.71069 A) was
used. The unit-cell parameters were determined from
25 reflections with 25.09 ꢂ ꢀ 2ꢆ ꢀ 27.12 ꢂ. Intensity data
with 2ꢆ ꢀ 55.1 ꢂ were collected with the ! ꢁ 2ꢆ scan
technique (scan speed 4.0 ꢂ/min) at 2864 reflections.
The intensities of three representative reflections which
were measured after every 150 reflections remained
constant throughout data collection indicating crystal
and electronic stability. No decay correction was applied.
The intensities were corrected for Lorentz and polariza-
tion factors, but not for absorption. The structure
was solved by direct methods24 and expanded using
Fourier techniques.25 The full-matrix least-squares
refinement for non-hydrogen atoms was carried out for
Table 2. Lifetimes (ꢈ), triplet energies (ET) and electronic
configurations of the lowest triplet state of 6
2
2
ÆwðjFoj ꢁ jFcjÞ , where the weight w ¼ 4Fo2=ꢇ2ðFoÞ ,
for 2594 independent reflections with I > ꢁ 10.00 ꢇ(I ).
The final discrepancy factors were R ¼ 0.051 and
Rw ¼ 0.062. The maximum and minimum peaks on the
final difference Fourier map corresponded to 0.23 and
Electronic
configuration
Solvent
ꢈ (ms)
ET (kcal molꢁ1 a
)
Cyclohexane
EPA
1.9
88
74.4
73.9
n,ꢅ*
ꢅ,ꢅ*
ꢁ0.22 e Aꢁ3, respectively. Neutral atom-scattering fac-
˚
tors were taken from Cromer and Waber.26 Anomalous
a 0–0 band of phosphorescence.
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 122–128

TYPE II PHOTOREACTION OF BENZOYL AND (BENZOYLMETHYL)CYCLOHEXANONES
127
dispersion effects were included in Fc;27 the values for
Áf 0 and Áf 0 were those of Creagh and McAuley.28 The
values for the mass attenuation coefficients are those of
Creagh and Hubbell.29 All calculations were performed
using the CrystalStructure30,31 crystallographic software
package.
2-Hydroxy-2-phꢂenylbicyclo[4.2.0]octan-5-one (9). 35%;
m.p. 103.5–105 C; ꢊmax(KBr) 1690 and 3370 cmꢁ 1; ꢁH
(CDCl3) 1.7–2.1 (2H, m, CH2), 2.19 (1H, s, OH), 2.3–2.6
(3H, m, CH2 þ CH), 2.6–2.9 (2H, m, CH2), 3.0–3.3 (1H,
m, CH) and 7.3–7.5 (5H, m, aromatic); ꢁC (CDCl3) 21.4
(t), 23.2 (t), 38.3 (t), 39.2 (d), 39.7 (t), 47.3 (d), 77.5 (s),
124.9 (d, 2C), 127.4 (d), 128.6 (d, 2C), 146.4 (s) and
213.5 (s). Found: C, 77.90; H, 7.47. C14H16O2 requires C,
77.75; H, 7.46%.
Crystal data for 2
3-Acetyl-2-ethyl-1-phenylcyclobutanol (11). 4%,
max(neat) 1700, 1720 and 3450 cmꢁ1; ꢁH (CDCl3) 0.73
A colorless needle crystal, C13H14O2, M ¼ 202.25, mono-
ꢊ
˚
clinic, a ¼ 12.577(3), b ¼ 5.567(2), c ¼ 15.618(2) A,
(3H, t, J 7.0, CH3), 1.3–1.5 (2H, m, CH2), 2.27 (3H, s,
COCH3), 2.3–2.7 (2H, m, CH2), 2.94 (1H, q, J 7.8, CH),
3.50 (1H, tt, J 7.9 and 4.0, CH), 4.53 (1H, s, OH) and 7.1–
7.5 (5H, m, aromatic). Complete purification of this
compound could not be achieved.
3
ꢂ
˚
ꢃ ¼ 100.05(1) , V ¼ 1076.8(5) A , T ¼ 293 ꢃ 1 K, space
group P21/n, Z ¼ 4, ꢉ(Mo Kꢄ) ¼ 0.8 cmꢁ1, 2864 reflec-
tions measured, 2594 unique (Rint ¼ 0.052).
General procedure for determination of
quantum yields in solution
4-Ethoxycarbonyl-2-hydroxy-2-phenylbicyclo[4.2.0]oc-
tan-5-one (13). 51%; m.p. 100 – 102 ꢂC; ꢊmax(KBr)
1700, 1720 and 3475 cmꢁ1; ꢁH (CDCl3) 1.24 (3H, t, J 7.0,
CH3), 1.7–2.6 (6H, m, 3 ꢄ CH2), 2.19 (1H, s, OH), 2.93
(2H, ABq, J 17.1, CH2), 3.41 (1H, dd, J 3.8 and 7.7, CH),
4.15 (2H, q, J 7.0, CH2) and 7.3–7.5 (5H, m, aromatic);
ꢁC (CDCl3) 14.1 (q), 20.6 (t), 3.96 (t), 42.7 (t), 52.9 (d),
53.2 (s), 61.7 (t), 76.9 (s), 124.8 (d, 2C), 127.6 (d), 128.6
(d, 2C), 145.5 (s), 171.4 (s) and 209.5 (s). Found: C,
70.90; H, 6.92. C17H20O4 requires C, 70.81; H, 6.99%.
A solution of a ketone (ca 0.05 mol dmꢁ3) containing an
appropriate amount of a calibrant was placed in
150 ꢄ 15 mm Pyrex tubes. The tubes were degassed by
three freeze–pump–thaw cycles and then sealed. Irradia-
tion was performed on a ‘merry-go-round’ apparatus with
an Ushio 450 W high-pressure mercury lamp. A potas-
sium chromate filter solution was used to isolate the
313 nm line. Analyses of products and the unreacted
starting ketone were performed with a gas chromato-
graph. Valerophenone was used as an actinometer.15
General procedure for preparative
irradiation on silica gel
Five grams of silica gel and ca 1.5 mmol of 6 or 14 in
20 cm3 of dichloromethane were placed in a 50 cm3 of
round-bottomed flask. The mixture was sonicated for
5 min and then the solvent was evaporated under reduced
pressure. The coated silica gel was separated into five
nearly equal portions, each of which was placed in a
18 ꢄ 180 mm Pyrex culture tube. The tubes were rotated
and irradiated with a 100 W high-pressure mercury lamp.
The irradiated silica gel was collected and placed in a
50 cm3 round-bottomed flask. Acetone (20 cm3) was
added and the mixture was sonicated for 10 min. The
silica gel was separated by filtration and washed with
10 cm3 of acetone. The filtrate and washings were col-
lected and then the solvent was removed under reduced
pressure. The residue was chromatographed on a silica
gel column. Elution with hexane–ethyl acetate gave
unreacted starting ketone and photoproducts.
General procedure for preparative
irradiation in solution
A solution of a ketone (ca 0.05 mol dmꢁ3) was irradiated
with the 450 W high-pressure mercury lamp. After re-
moval of the solvent, the residue was chromatographed
on a silica gel column with hexane—ethyl acetate as
eluent.
7-Phenylhepten-4,7-dione (4). 61%; ꢊmax(neat) 1685
and 1720 cmꢁ1; ꢁH (CDCl3) 2.78 (2H, t, J 6.6, CH2), 3.17
—
—
2
(2H, t, J 6.6, CH COPh), 3.20 (2H, d, J 6.9, CH C C),
2
5.08 (1H, dd, J 17.1 and 1.6, olefinic), 5.10 (1H, dd, J 10.2
and 1.6, olefinic), 5.87 (1H, ddt, J 17.1, 10.2 and 6.9,
olefinic), 7.3–7.5 (3H, m, aromatic) and 7.8–7.9 (2H, m,
aromatic). Found: C, 77.33; H, 7.13. C13H14O2 requires
C, 77.20; H, 6.98%.
Determination of relative quantum yield for
production of 7 on a silica gel surface
6-Hydroxy-6-phenylbicyclo[3.1.1]heptan-3-ones (5).
25%; ꢊmax(neat) 1740 and 3440 cmꢁ1; ꢁH (CDCl3)
1.8–2.9 (8H, m), 7.1–7.7 (5H, m, aromatic); m/z 202
(Mþ). The isomers of the bicycloheptanones could not be
isolated.
Each of a 0.3 mmol amount of 6, 14 and valerophenone
was loaded on 1.0 g of silica gel (the surface coverage
was <60%). The coated silica gel was placed in
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 122–128

128
T. HASEGAWA ET AL.
9. Hasegawa T, Nishimura M, Kodama Y, Yoshioka M. Bull. Chem.
Soc. Jpn. 1990; 63: 935–937.
10. Lewis FD, Johnson RW. J. Am. Chem. Soc. 1972; 94: 8914–
8916.
11. Scheffer JR, Garbibay M, Nalamasu O. In Organic Photochem-
istry, vol. 8, Padwa A (ed). Marcel Dekker: New York, 1988; 249–
347.
12. Juaristi E. Introduction to Stereochemistry and Conformational
Analysis. Wiley: New York, 1991; 255–258.
13. Goswami PC, de Mayo P, Ramhath N, Bernard G, Omkaram N,
Scheffer J, Wong YF. Can. J. Chem. 1985; 63: 2719–2725.
14. Wagner PJ. Tetrahedron Lett. 1967; 1753–1756.
15. Lewis FD, Hilliard TA J. Am. Chem. Soc. 1970; 22: 6672–6674.
16. Hasegawa T. Recent Res. Dev. Photochem. Photobiol. 1997; 1:
19–39.
18 ꢄ 180 mm Pyrex culture tubes. The tubes were rotated
and irradiated with 313 nm radiation from a 100 W high-
pressure mercury lamp. After the irradiation, 4 cm3 of
acetone containing a known amount of hexadecane as a
calibrant were added and the mixture was sonicated for
10 min. The supernatant solutions were analyzed to
determine the amounts of 7.
Acknowledgement
Financial support from The Shinsei Foundation is grate-
fully acknowledged.
17. Wagner PJ, Kemppainen AE, Schott HN. J. Am. Chem. Soc. 1972;
95: 5604–5614.
18. Weis LD, Evans TR, Leermarkers PA. J. Am. Chem. Soc. 1968;
90: 6109–6118.
19. Kosower E.M. J. Am. Chem. Soc. 1958; 80: 3253–3260.
20. Mitani M, Tamada M, Uehara S, Koyama K. Tetrahedron Lett.
1984; 25: 2805–2808.
REFERENCES
21. Baumgarten HF, Creger PL, Villars CE. J. Am. Chem. Soc. 1958;
80: 6609–6612.
22. Raffaello R, Trave R. Gazz. Chim. Ital. 1950; 80: 496–501.
23. Murov SL, Handbook of Photochemistry. Marcel Dekker: New
York, 1973; 99.
1. Wagner PJ. Acc. Chem. Res. 1971; 4: 168–177; Wagner PJ. Acc.
Chem. Res. 1983; 16: 461–467; Wagner PJ. In Rearrangements in
Ground and Excited States, vol. 3, de Mayo P (ed). Academic
Press: New York, 1980; 381–444; Scaiano JC, Lissi EA, Encina
MV. Rev. Chem. Intermed. 1978; 2: 139–196.
24. Altomare Cascarano AG, Giacovazzo C, Guagliardi A, Burla M,
Polodori G, Camali M. J. Appl. Crystallogr. 1994; 27: 435–
436.
25. Beurskens PT, Admiraal G, Eurskens G, Bosman WP, de Gelder
R, Israel R, Smits JMM. The DIRDIF-99 Program System,
Technical Report of the Crystallography Laboratory. University
of Nijmegen: Nijmigen, 1999.
26. Cromer DT, Waber JT. In International Tables for X-Ray Crystal-
lography, vol. IV. Ibero JA, Hamiltone WC (eds). Kynoch Press:
Birmingham, 1974; Table 2.2A.
27. Ibers JA, Hamilton WC. Acta Crystallogr. 1964; 17: 781–782.
28. Creagh DC, McAuley WJ. In International Tables for Crystal-
lography, vol. C. Wilson AJC (ed). Kluwer: Boston, 1992; Table
4.2.6.8, 219–222.
2. Wagner PJ, Kelso PA, Kemppainen AE, McGrath JM, Schott HN,
Zepp RG. J. Am. Chem. Soc. 1972; 94: 7506–7512.
3. (a) Wamser CC, Wagner WR. J. Am. Chem. Soc. 1981; 103:
7232–7234; (b) Ariel S, Evans S, Omkaram N, Scheffer JR,
Trotter J. J. Chem. Soc., Chem. Com. 1986; 375–377; (c) Ariel
S, Ramamrthy V, Scheffer JR, Trotter J. J. Am. Chem. Soc. 1983;
105: 6959–6960; (d) Ariel S, Evans S, G-Garibay M, Harkness
BR, Omkaram N, Scheffer JR, Trotter J. J. Am. Chem. Soc. 1988;
110: 5591–5592.
4. Wagner PJ, Kelso PA, Kemppainen AE, Zepp RG. J. Am. Chem.
Soc. 1972; 94: 7500–7506.
5. (a) Padwa A, Eastman D. J. Am. Chem. Soc. 1969; 91: 462–467;
(b) Lewis FD, Johnson RW. Tetrahedron Lett. 1973; 2557–2560;
(c) Lewis FD, Johnson RW, Johnson DE. J. Am. Chem. Soc. 1974;
96: 6090–6099.
29. Creagh DC, Hubbell JH. In International Tables for Crystal-
lography, vol. C. Wilson AJC (ed). Kluwer: Boston, 1992; Table
4.2.4.3, 200–206.
6. Xia W, Scheffer JS, Botoshansky M, Kaftory M. Org. Lett. 2005;
7: 1315–1318.
30. CrystalStructure 2.00. Rigaku/MSC: The Woodland, TX, 2001.
31. Watkin DJ, Prout CK, Carruthers JR, Betteridge PW. CRYSTALS,
Issue 10. Chemical Crystallography Laboratory: Oxford, 1996.
7. McIntosh CL. Can. J. Chem. 1967; 45: 2267–2270.
8. Hasegawa T, Nishimura M, Yoshioka M. J. Phys. Org. Chem.
1990; 3: 230–234.
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 122–128