Solid-state photocyclization of 2,4,6-triisopropyl-4′-(methoxycarbonyl)benzophenone. Evidence for a narrow reaction cavity and a photoenol diradical intermediate

Article abstract of DOI:10.1021/jp973235e

The origin of the previously observed unusual photostability of 2,4,6-triisopropyl-4′-(methoxycarbonyl)-benzophenone (1-p-CO₂Me) in the solid state was investigated. 1-p-CO₂Me was found to photocyclize normally to produce the corresp

Full text of DOI:10.1021/jp973235e

J. Phys. Chem. A 1998, 102, 5415-5420
5415
Solid-State Photocyclization of 2,4,6-Triisopropyl-4′-(methoxycarbonyl)benzophenone.
Evidence for a Narrow Reaction Cavity and a Photoenol Diradical Intermediate†
Yoshikatsu Ito,*,‡ Satoshi Yasui,§ Jun Yamauchi,§ Shigeru Ohba,| and Gentaro Kano‡
Department of Synthetic Chemistry and Biological Chemistry, Faculty of Engineering, Kyoto UniVersity,
Kyoto 606, Japan, Graduate School of Human and EnVironmental Studies, Kyoto UniVersity, Kyoto 606, Japan,
and Department of Chemistry, Faculty of Science and Technology, Keio UniVersity, Yokohama 223, Japan
ReceiVed: October 3, 1997; In Final Form: December 4, 1997
The origin of the previously observed unusual photostability of 2,4,6-triisopropyl-4′-(methoxycarbonyl)-
benzophenone (1-p-CO2Me) in the solid state was investigated. 1-p-CO2Me was found to photocyclize normally
to produce the corresponding benzocyclobutenol 2-p-CO2Me when its solid-state photolysis was carried out
either (a) after thorough grinding, (b) after solid-solid mixing with 2,4,6-triisopropyl-4′-(ethoxycarbonyl)-
benzophenone (1-p-CO2Et), or (c) at elevated temperatures (an estimated energy barrier of 20 kcal/mol).
Furthermore, when the photolysis was performed under more carefully deoxygenated conditions (closed argon
atmosphere), formation of blue species that are persistent in the absence of oxygen was observed. On the
basis of oxygen trapping and ESR experiments, the blue species are regarded as a mixture of a diradical
intermediate DR and monoradicals derived thereof. The X-ray study of 1-p-CO2Me had revealed that the
distances between the carbonyl oxygen and the o-i-Pr methine hydrogens are within the critical limit for
hydrogen abstraction to occur, but a small reaction cavity or the compact crystal packing around both of the
o-i-Pr groups is interfering with the photocyclization. The present results are consistent with this X-ray
crystal structure; i.e., the photochemical hydrogen abstraction of 1-p-CO2Me to DR can take place, but DR
reketonizes back to 1-p-CO2Me under the usual photolysis conditions because there is a high topochemical
barrier to cyclization leading to 2-p-CO2Me.
Introduction
SCHEME 1
In contrast with efficient solid-state photocyclizations of
several 2,4,6-triisopropylbenzophenones 1-X (X ) p-OMe, p-t-
Bu, p-Me, H, p-Cl, p-CF3) into the corresponding benzocy-
clobutenols 2-X (89-100% conversion, ∼100% yield), 1-X f
DR f 2-X (Scheme 1), 2,4,6-triisopropyl-4′-(methoxycarbonyl)-
benzophenone (1-p-CO2Me) underwent virtually no photore-
action.1 This has been curious for us because we found later
that all the other carboxy derivatives 1-X (X ) m- and p-CO2H,
m-CO2Me, p-CO2Et, p-CO2Ca0.5‚0.5H2O, p-(R)-CO2CH*(Me)-
Et, m- and p-(S)-CONHCH*(CH2Ph)CO2Me, m-COCl), exclud-
ing 1-p-COCl, photocyclized with fair to good efficiencies in
the solid state (no. 1-10 in Table 1).2 Therefore, the X-ray
study for photoinert 1-p-CO2Me and 1-p-COCl was carried out.3
The result has revealed that the distances between the carbonyl
oxygen and the o-i-Pr methine hydrogens (2.95 and 2.98 Å for
1-p-CO2Me, 2.94 and 2.98 Å for 1-p-COCl) are within the
critical limit (3 Å)4 for hydrogen abstraction to occur, but a
narrow reaction cavity or the compact crystal packing around
both of the o-i-Pr groups is interfering with the photocyclization.
We have now found that the photocyclization of 1-p-CO2Me
occurs in the solid state when the photolysis was carried out
either (a) at elevated temperatures, (b) after thorough grinding,
or (c) after solid-solid mixing with 1-p-CO2Et. Furthermore,
when the solid-state irradiation of 1-p-CO2Me was performed
under more carefully deoxygenated conditions, formation of blue
radical species were observed, indicative of the existence of
the diradical intermediate DR. These findings are nicely
consistent with the above-mentioned X-ray crystal structure for
1-p-CO2Me.
Results and Discussion
The results that we have obtained so far from photolyses of
various carboxy derivatives of 2,4,6-triisopropylbenzophenone
in the solid state or in benzene solution are summarized in Table
1. Solid samples were prepared by gently pulverizing the
crystals with a mortar and pestle, and the powder was placed
in the Pyrex made apparatus for solid-state photolysis.5 Ir-
radiations were carried out with a 400-W high-pressure mercury
lamp for several hours under an argon stream. During the
irradiation, samples were cooled from the outside of the reaction
vessel either by circulation of cold water maintained at 4 °C
(solid) or by running water from a faucet (solution). However,
in the case of experiments 14 and 15, the vessel was mounted
on a hot plate.
* To whom correspondence should be addressed. Telephone: 075-753-
5670. Fax: 075-753-5676. E-mail: yito@sbchem.kyoto-u.ac.jp.
† This work is dedicated to Professor N. J. Turro on the occasion of his
60th birthday.
‡ Faculty of Engineering, Kyoto University.
Experiments 1-11 in Table 1 show that in solution, all the
carboxy derivatives 1-X can photocyclize to yield quantitatively
§ Human and Environmental Studies, Kyoto University.
| Keio University.
S1089-5639(97)03235-0 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/27/1998

5416 J. Phys. Chem. A, Vol. 102, No. 28, 1998
Ito et al.
TABLE 1: Photolyses of Carboxy Derivatives of 2,4,6-Triisopropylbenzophenone in the Solid State or in Benzene Solution (0.01
M)a
1-X
solid
solution
conv, %
no.
X
mp, °C
h
conv, %
h
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
p-CO2H
m-CO2H
p-CO2Ca0.5‚0.5H2O
p-(S)-CONHCH(CH2Ph)CO2Me
m-(S)-CONHCH(CH2Ph)CO2Me
p-COCl
m-COCl
p-(R)-CO2CH(Me)Et
p-CO2Et
221-223
185-186
4
2
4
5
4
10
4
5
4
2
10
10
8
10
10
28
100#
100
96c
100d
0
5
100
100
100
100d
100d
21
100
100
100
100
100
2
4b
5
4
4
4
4
4
2
4
>271
213-214
116-118
119-120
127-128
77-78
98-101
118-120
142-143
35
85
76
m-CO2Me
100#
0e
p-CO2Me
p-CO2Mef
4
p-CO2Meg
30h
54
p-CO2Mei
p-COClj
5
a Conversions were estimated by NMR and HPLC analyses. Yields of 2-X were nearly 100% in all cases unless marked by #, where minor
amounts of uncharacterized products were formed (<10% by NMR). b A suspension. c Diastereomer excess of 2, 87%. d Diastereomer excess of 2,
∼0%. e Reference 1. From closer examinations by HPLC, it became clear that a trace of 2-p-CO2Me (<1%) was always formed probably at crystal
defects. f The crystals were ground thoroughly by using an agate mortar and pestle. g An equimolar mixture of crystals of 1-p-CO2Me and 1-p-
CO2Et was ground together in an agate mortar and pestle. h Conversion for the 1-p-CO2Et component, 78%. i Irradiated at 84 °C. j Irradiated at 69
°C. Irradiation at 86 °C resulted in a higher conversion but side reactions occurred.
the corresponding 2-X. These compounds also photocyclized
in the solid state (28-100% conversions, >90% yields) except
1-p-CO2Me and 1-p-COCl, which were virtually photoinert.1,2
It was suggested on the basis of the crystal structure of 1-p-
CO2Me that a deficient free space around the o-i-Pr groups
might be responsible for the observed lack of its solid-state
photoreactivity.3 If this is true, any treatment that can disturb
the crystal lattice of this material may result in recovery of its
photoreactivity owing to decrease in the topochemical restriction
to the benzocyclobutenol formation. We carried out several
kinds of experiments, (i)-(iv), in order to prove this prediction.
(i) First, since there are in general more defects on the crystal
surface than in the bulk, pulverization of the crystal should
increase its defects. Therefore, the effect of complete grinding
was examined. The crystals of 1-p-CO2Me were ground into
fine powders with a mortar and pestle for a long time (30 min)
and then irradiated. As expected, 1-p-CO2Me was transformed
to the benzocyclobutenol 2-p-CO2Me, although the conversion
was low (4%) (Table 1, no. 12).
Figure 1. Temperature effects (a) on photolysis of 1-p-CO2Me in the
(ii) It is known that molecules of similar shape and size can
be easily substituted for each other in molecular crystals to form
mixed crystals (a solid solution).6 Hence, unreactive 1-p-CO2-
Me and reactive 1-p-CO2Et may mix freely, giving the former
a more vacant space. Thus, an equimolar mixture of crystals
of 1-p-CO2Me and 1-p-CO2Et was ground together in a mortar
and pestle, and this solid mixture was irradiated. Both
compounds were competitively converted to the corresponding
benzocyclobutenols 2-p-CO2Me and 2-p-CO2Et at 30% and 78%
conversion, respectively (Table 1, no. 13). It seems that 1-p-
CO2Me acquired the photoreactivity as a result of mixed crystal
formation with 1-p-CO2Et.
(iii) Third, inclusion of 1-p-CO2Me in the cavities of suitable
host crystals is expected to impart an increased free space to
the guest molecule.5 Therefore, cocrystallization of 1-p-CO2-
Me with (()-1-phenyl-1,2-ethanediol, deoxycholic acid, or
cholic acid was attempted. However, the resultant crystal
mixtures were photoinert. Probably 1-p-CO2Me failed to form
inclusion crystals with these hosts.
solid state (irradiated for 10 h) or (b) in benzene solution (0.01 M,
irradiated for 1 h) and (c) on photolysis of 1-p-CO2H in the solid state
(irradiated for 5 h).
1-p-CO2Me was converted to the benzocyclobutenol 2-p-CO2-
Me (Table 1, no. 14). Another photoinert crystal 1-p-COCl also
reacted at elevated temperatures (Table 1, no. 15). Figure 1a
shows that the conversion to 2-p-CO2Me dramatically increased
at temperatures higher than 55 °C. An energy term obtained
from the plot of ln(conversion) against 1/T (Figure 2) had as
large a value as 20 kcal/mol. This energy barrier might reflect
a high stiffness of the cavity wall as well as an unfavorable
size/shape of the reaction cavity.7 In contrast, a temperature
effect on the solution-state photolysis of 1-p-CO2Me was small
(Figure 1b). We have previously shown that the quantum yield
for photocyclization of 1-H to 2-H in benzene increased only
slightly with increasing temperatures: Φ ) 0.50, 0.52, 0.58,
and 0.58 at 11.0, 25.0, 56.0, and 68.0 °C, respectively.8
A
temperature effect on the solid-state photolysis of 1-p-CO2H
was also examined and turned out to be relatively small (Figure
1c).
(iv) Fourth, a temperature effect was examined. It is
remarkable that upon solid-state photolysis at 84 °C, 54% of

Solid-State Photocyclization
J. Phys. Chem. A, Vol. 102, No. 28, 1998 5417
Figure 2. Plot of ln(conversion) vs 1/T for photolysis of 1-p-CO2Me
in the solid state (the original data from Figure 1a).
The results described in (i), (ii), and (iv) led us to conclude
that the unusual photostability of crystalline 1-p-CO2Me is
simply due to its compact crystal packing, as was indicated by
the X-ray study.3 This conclusion is also consistent with the
calculated high crystal density for 1-p-CO2Me (1.14 g/cm3),
which is somewhat larger than that for 1-p-OMe (1.07), 1-H
(1.04), 1-p-CO2Et (1.10), 1-p-CO2H (1.11), or 1-m-CO2Me
(1.10).3 Since the reaction is sequential like 1-p-CO2Me f DR
f 2-p-CO2Me, the observed stability may be ascribed to a
topochemical hindrance to either of these steps. As we will
show later, however, photolysis of nonpulverized crystals of
1-p-CO2Me generated oxygen-sensitive blue radical species that
are likely to be a mixture of a diradical intermediate DR and
other radicals derived thereof. Therefore, it is probably the
process DR f 2-p-CO2Me that is hindered in the crystal lattice.
This inference is reasonable, since the carbonyl oxygen and the
o-i-Pr methine hydrogens are close enough for the hydrogen
abstraction to occur but the vacant space around the o-i-Pr
groups is too limited for the cyclization to take place.3
Figure 3. (A) Change in diffuse-reflectance spectra upon photolysis
of nonpulverized crystals of 1-p-CO2Me irradiated for 0, 2, and 6 h.
(B) Decay of photogenerated blue species in the air 0 h, 2 h, 1 day, 3
days, and 5 days after irradiation for 6 h.
e.g., their decay rate at room temperature was increased only
by 1.3 times at 83 °C.
When the blue crystals were dissolved in organic solvent,
the color disappeared immediately and a number of very minor
products were detected in the solution with virtually quantitative
recovery of 1-p-CO2Me (NMR and HPLC). Among these minor
products, two main products were identified as 2-p-CO2Me and
a cyclic peroxide 3 by means of HPLC analyses (Figure 4).
After repeated experiments, it was found that 3 was always
formed in a higher yield when the blue crystals were dissoved
in the presence of air (0.28-0.33% yield) than when dissoved
under argon (0.09-0.19% yield) (Scheme 2). This result
supports the existence of a small amount of DR in the blue
crystals because the peroxide 3 must have been produced as a
result of reaction between oxygen and the diradical DR.10 The
photoenol of 2,4,6-triisopropylbenzophenone (1-H) absorbs at
360 nm in hexane.11 Diradicals derived from o-alkylbenzophe-
nones exhibit a long-wavelength absorption band around 520-
550 nm12-15 and do not much differ from the blue species (λmax
624 nm). This finding also seems to indicate the presence of
some DR in the blue species.
Since the blue species were considerably quenched by contact
with air, they must be radicals. ESR spectra measured for a
single crystal showed one intense absorption with a g value of
2.002 87 and several paired lines on both sides of the central
line (Figure 5). The former is assigned mainly to monoradicals
produced in the crystals. However, there is a possibility of
overlapping triplet species with small fine structures. Variation
of its spectral intensity with irradiation time is shown in Figure
6. The time courses for the photolysis monitored by diffuse-
When the nonpulverized crystals of 1-p-CO2Me were irradi-
ated in a closed apparatus filled with the argon gas rather than
under an argon stream, their surfaces turned to deep blue,
although the inside remained colorless. In the aforementioned
photolysis of the powdered sample under an argon stream, such
blueing was not obvious.9 The photolysis of the nonpowdered
crystals was followed by diffuse-reflectance spectroscopy
(Figure 3A). The absorptions at λmax ) 624, 501, 480, 457,
and ∼360 nm continued to increase during irradiation for several
hours. This blue color did not fade visibly under argon. In
the air, however, the absorbance at 624 nm decayed rapidly at
first, then gradually more slowly in the course of a few days,
as displayed in Figure 3B. The absorbance at 624 nm decreased
more than that at 501 nm, indicating that the photogenerated
species are not a single substance. The decay rate became very
slow after several days. For instance, the remaining absorbance
at 624 nm relative to that immediately after irradiation was 38%
even after 40 days. This result can be understood by assuming
that more than one-third of the blue species are well protected
from the attack by oxygen owing to their presence in the crystal
bulk, which the oxygen molecule cannot easily penetrate into.
The blue species in the crystal bulk had a high thermal stability;

5418 J. Phys. Chem. A, Vol. 102, No. 28, 1998
Ito et al.
Figure 4. HPLC analysis of the blue crystals. In this experiment, only
the blue surface was washed off with MeOH in the presence of air and
was analyzed: (1) peroxide 3; (2) 2-p-CO2Me; (3) 1-p-CO2Me; x,
solvent; i, an impurity in 1-p-CO2Me.
SCHEME 2
Figure 5. ESR spectra of the photoirradiated (3 h) 1-p-CO2Me single
crystal in vacuo observed under amplitude of 320 with 0.2-mT field
modulation. The upper spectrum consists of the central intense line,
the paired two lines (designated by arrows), and the Mn2+ standard
signals of the third and fourth hyperfine components. The lower spectra
were taken under different amplitudes: 320 and 1000.
reflectance spectroscopy (Figure 3A) and ESR spectroscopy
(Figure 6) correlated with each other well, considering their
different irradiation conditions. Spin counting of the central
lines made after 13 h of irradiation showed the radical
concentration of ca. 0.3% on the basis of 4-hydroxy-2,2,6,6-
tetramethylpiperidine-1-oxyl as standard.
Several paired lines seen in Figure 5 exhibited a strong
angular dependence with respect to the magnetic field. Since
the ESR fine structure of a triplet species in a single crystal
comprises paired two lines and their separation changes depend-
ing on crystal orientation, these paired peaks indicate unambigu-
ously the zero-field splitting tensor for the triplet radical species.
Thus, we analyzed the angular dependence of the spectral
separation for the most dominant triplet species (designated by
arrows in Figure 5). The principal values of the zero-field
splitting gave |D| ) 5.962 and |E| ) 0.537 mT. This triplet
species (S ) 1) comes either from the intramolecular or
Figure 6. ESR spectral variation of the monoradical species depending
on the irradiation time. Intensities were normalized at ca. 800 min
irradiation.
intramolecular pair such as the diradical DR in Scheme 1.
Moreover, a crystallographic inspection has revealed that the
analyzed principal axes coincide not with the direction of the
two radical centers of DR but with the intermolecular direction
between the nearest-neighboring molecules, i.e., between mol-
ecule (x, y, z) and molecule (1-x, -y, 1 -z) as is shown in
Figure 7. Consequently, the most intense line pair in Figure 5
is assigned to an intermolecular radical pair, which was probably
1
intermolecular radical (S ) /2) pairs. The zero-field splitting
parameters obtained suggest the distance between the radicals
to be in the range 7-8 Å. This value is too large for an

Solid-State Photocyclization
J. Phys. Chem. A, Vol. 102, No. 28, 1998 5419
SCHEME 3
In summary, we have demonstrated by three simple means
(thorough grinding, solid-solid mixing, and elevated temper-
atures) that a narrow reaction cavity is responsible for the
previously found lack of solid-state photoreactivity for 1-p-CO2-
Me. Blue radical species, which are very stable in the absence
of oxygen, are regarded as a mixture of the photoenol-type
diradical DR and monoradicals derived thereof.
Figure 7. Crystal structure of 1-p-CO2Me.
Experimental Section
formed via DR through H abstraction from the nearest-
neighboring molecule. Judging from the yield of the O2-trapped
product 3 (0.28-0.33%), the DR species present in the crystal
is very small in quantity and hence may elude detection under
the present condition of ESR measurements.
General Procedures. 1H-NMR spectra (200 MHz) were
measured on a Varian Gemini-200 spectrometer in CDCl3. IR
and mass spectra were recorded on JASCO FT/IR-5M and JEOL
JMS-DX 300 spectrometers, respectively. Diffuse-reflectance
spectra were measured with a Shimadzu UV-2400PC spectrom-
eter equipped with a diffuse-reflectance attachment, and a BaSO4
powder was used as a standard for reflectivity. HPLC analyses
were performed with a Shimadzu LC-5A chromatograph and a
UV detector (fixed at 217 nm) by using a Cosmosil 5C18-AR
column (4.6 mm i.d. × 150 mm) and a mixture of methanol
and water (85:15 v/v) as eluent. ESR spectra were measured
at the X band using JEOL-FE1XG with a 100-kHz field
modulation.
Materials. 2,4,6-Triisopropyl-4′-(methoxycarbonyl)ben-
zophenone (1-p-CO2Me) and other derivatives (1-p-CO2Et, 1-p-
CO2H, 1-p-COCl, 1-H) employed in the present work were
available from our previous work.1,8,10,17-19 Preparations and
photolyses of 1-X (X ) p-CO2Ca0.5‚0.5H2O, p-(R)-CO2CH*(Me)-
Et, and m- and p-(S)-CONHCH*(CH2Ph)CO2Me) will be
reported elsewhere.2
Solid-State Photolysis. Crystals, which were ground into
powders in an agate mortar and pestle, were placed in our solid-
state photolysis vessel made of Pyrex glass.5 Then irradiation
was carried out with a 400-W high-pressure mercury lamp under
a slow stream of argon. During the irradiation, the vessel was
cooled from outside by circulation of cold water (4 °C) or heated
on a hot plate ((2 °C). After the photolysis, the reaction
mixture was dissolved in suitable solvents (usually MeOH and
CDCl3) and analyzed by NMR and HPLC.
Diffuse-Reflectance Spectra Measurements. Into a shallow
hollow for the diffuse-reflectance spectrum measurement, 3 g
of BaSO4 powders and subsequently 150 mg of nonpulverized
crystals of 1-p-CO2Me were pressed. This was placed in our
solid-state photolysis vessel5 and was degassed, then filled with
argon. This degassing-argon-charge cycle was repeated three
times and the vessel was closed. The sample was irradiated as
described above. The diffuse-reflectance spectra were measured
in the air at room temperature before and after irradiation.
ESR Measurements. An ESR tube containing a crystalline
2,4,6-triisopropylbenzophenone was sealed under high vacuum.
This was irradiated with a 450-W high-pressure mercury lamp
at ambient temperature, and the ESR spectra were measured.
Mn2+ solid solution in MgO (1/2000) and 4-hydroxy-2,2,6,6-
Figure 7 displays the crystal structure of 1-p-CO2Me as well
as the hydrogen atoms (H27B and H25B) that are nearest the
ketone carbonyl carbon (C10) or nearest the o-i-Pr methine
carbon (C23 or C17). Molecule (x, y, z) and molecule (1 - x,
-y, 1 - z), which are related by a center of symmetry, are
close to each other, and this characteristic dimeric packing was
considered to be a major driving force leading to its narrow
reaction cavity.3 The distances between the above carbon atoms
and hydrogen atoms are 3.3-3.5 Å. Although these distances
are considerably longer than the sum of van der Waals radii of
C and H (2.9 Å), the intermolecular H abstraction, unlike the
intramolecular one, may occur beyond this limit.16 Of course,
it is presumed here that the conformation of DR is not very
different from that of the original ketone and that the radical
centers are nearly localized. Indeed, the angular analysis of
the most intense ESR triplet peaks demonstrated that the
hydrogen-atom transfers occurred between DR and ketone in
the dimeric structure (vide supra).
Photogeneration of similar blue species was observed for none
of the other 2,4,6-triisopropylbenzophenones thus far studied.
ESR signals observed from photolysis of the crystals of 1-p-
COCl or 1-p-CO2H were by far the weaker compared with those
of 1-p-CO2Me. ESR signals were negligible in the case of 1-H.
The reason for this conspicuously high radical yield from 1-p-
CO2Me is not clear at this moment. Possibly DR is longer-
lived in the rigid crystal lattice of 1-p-CO2Me owing to
inhibition of the cyclization process DR f 2-p-CO2Me. Hence,
DR may have chances to be trapped by oxygen to yield 3 and
to further abstract hydrogen atoms of neighboring 1-p-CO2Me
molecules, giving a mixture of monoradical and diradical species
(Scheme 3). Although DR has been assumed to exist in
photocyclization of 2,4,6-triisopropylbenzophenones to benzo-
cyclobutenols in solution,8,10,17,18 they were unable to be detected
by laser flash photolysis probably because of their short lifetimes
(,100 ns).11,12 The present results suggest that the lifetime of
DR photogenerated from crystalline 1-p-CO2Me is lengthened
as a consequence of its compact crystal packing around the o-i-
Pr groups, and hence, DR may be directly detectable. Time-
resolved ESR spectroscopy is in progress along this line.20

5420 J. Phys. Chem. A, Vol. 102, No. 28, 1998
Ito et al.
(6) Kitaigorodsky, A. I. Molecular Crystals and Molecules; Academic
Press: New York, 1973; Chapter 1.
tetramethylpiperidine-1-oxyl benzene solution were used for the
calibration of the magnetic field and for spin-counting, respec-
tively.
(7) (a) Ramamurthy, V.; Weiss, R. G.; Hammond, G. S. AdV.
Photochem. 1993, 18, 67-234. (b) Weiss, R. G.; Ramamurthy, V.;
Hammond, G. S. Acc. Chem. Res 1993, 26, 530-536. (c) Hashizume, D.;
Uekusa, H.; Ohashi, Y.; Matsugawa, R.; Miyamoto, H.; Toda, F. Bull. Chem.
Soc. Jpn. 1994, 67, 985-993.
(8) (a) Ito, Y.; Nishimura, H.; Umehara, Y.; Yamada, Y.; Tone, M.;
Matsuura, T. J. Am. Chem. Soc. 1983, 105, 1590-1597. (b) Ito, Y.;
Umehara, Y.; Hijiya, T.; Yamada, Y.; Matsuura, T. J. Am. Chem. Soc. 1980,
102, 5917-5919.
(9) Quenching of the blue species by the residual oxygen in the argon
gas, which should be more efficient in the powdered state, can explain this
phenomenon. In fact, when the powdered 1-p-CO2Me was irradiated under
a closed argon atmosphere, the powder became considerably blue. On
exposure to air, the color faded in only 1 h.
(10) (a) Ito, Y.; Matsuura, T. J. Am. Chem. Soc. 1983, 105, 5237-
5244. (b) Ito, Y.; Nishimura, H.; Shimizu, H.; Matsuura, T. J. Chem. Soc.,
Chem. Commun. 1983, 1110-1111.
Preparation of Cyclic Peroxide 3. An authentic sample of
a cyclic peroxide 3 was prepared by solution photolysis of 1-p-
CO2Me in the presence of oxygen.10 A 0.01 M solution of 1-p-
CO2Me (226 mg) in benzene (60 mL) was irradiated through
Pyrex with a 400-W high-pressure mercury lamp for 3 h under
bubbling of oxygen. The reaction mixture was evaporated under
reduced pressure to remove the solvent. The pale-yellow
residue, which was a mixture consisting of 2-p-CO2Me and a
small amount of 3 (NMR), was recrystallized from hexane (20
mL) to afford 108 mg of 2-p-CO2Me as colorless needles. The
mother liquor was concentrated and then was subjected to
preparative TLC on silica gel (Merck Kieselgel 60 F254, benzene)
followed by recrystallization with pentane (2 mL) to give 14
mg (6%) of 3 as white crystals, mp 143-147 °C. 1H NMR
(CDCl3, 200 MHz): δ 0.47 (3H, d, J ) 7.0 Hz, CHMe), 1.06
(3H, d, J ) 7.0 Hz, CHMe), 1.27 (6H, d, J ) 7.0 Hz, CHMe2),
1.55 (3H, s, Me), 1.75 (3H, s, Me), 2.91 (2H, finely split septet,
J ) 7 Hz, 2-CHMe2), 3.91 (3H, s, CO2Me), 4.12 (1H, s, OH),
6.92 (1H, finely split s, aromatic), 7.04 (1H, finely split s,
aromatic), 7.53 (2H, broad s, phenylene), 7.99 (2H, d, J ) 8.4
Hz, phenylene). MS m/z (rel intensity): 398 (2, M+), 367 (91),
366 (72), 365 (100), 351 (57), 307 (71), 231 (71). IR (KBr):
(11) Guerin, B.; Johnston, L. J. Can. J. Chem. 1989, 67, 473-480.
(12) Johnston, L. J.; Scaiano, J. C. Chem. ReV. (Washington, D.C.) 1989,
89, 521-547.
(13) Porter, G.; Tchir, M. F. J. Chem. Soc. A 1971, 3772-3777.
(14) Haag, R.; Wirz, J.; Wagner, P. J. HelV. Chim. Acta 1977, 60, 2595-
2607.
(15) Nakayama, T.; Hamanoue, K.; Hidaka, T.; Okamoto, M.; Teranishi,
H. J. Photochem. 1984, 24, 71-78.
(16) Ito, Y. In Photochemistry on Solid Surfaces; Anpo, M., Matsuura,
T., Eds.; Elsevier: Amsterdam, 1989; pp 469-480.
(17) Ito, Y.; Giri, B. P.; Nakasuji, M.; Hagiwara, T.; Matsuura, T. J.
Am. Chem. Soc. 1983, 105, 1117-1122.
(18) Ito, Y.; Kawatsuki, N.; Giri, B. P.; Yoshida, M.; Matsuura, T. J.
Org. Chem. 1985, 50, 2893-2904.
3467, 2962, 1721, 1711, 1611, 1282, 1188, 1114, 771 cm-1
.
(19) Ito, Y.; Umehara, Y.; Nakamura, K.; Yamada, Y.; Matsuura, T.;
Imashiro, F. J. Org. Chem. 1981, 46, 4359-4362.
Acknowledgment. This work was supported by the Grant-
in-Aid for Scientific Research on Priority Area from the
Japanese Government. A part of this work was carried out at
the laboratory of Professor J. R. Scheffer, University of British
Columbia, Canada.
(20) There is a dispute about which is an immediate precursor to
benzocyclobutenol, diradical ()enol triplet) or enol.18,21 This paper cannot
answer this question. Certainly, however, the blue species are radical, not
enol.22
(21) Wagner, P. J.; Subrahmanyam, D.; Park, B.-S. J. Am. Chem. Soc.
1991, 113, 709-710.
(22) We have recently reported a single-crystal-to-single-crystal dias-
tereoselective photocyclization of N-(p-(2,4,6-triisopropylbenzoyl)benzoyl)-
L-phenylalanine methyl ester (1-p-(S)-CONHCH*(CH2Ph)CO2Me).23 The
X-ray crystal-structure analysis of this transformation has revealed that the
carbonyl carbon and the nearer o-i-Pr methine carbon approached each other
with nearly no rotation around the single bonds linking these carbons to
the triisopropylphenyl ring. This result seems to eliminate enol as an
intermediate to benzocyclobutenol. More detailed discussion will be reported
elsewhere.
References and Notes
(1) Ito, Y.; Matsuura, T.; Fukuyama, K. Tetrahedron Lett. 1988, 29,
3087-3090.
(2) Ito, Y.; Kano, G.; Nakamura, N. To be published.
(3) Fukushima, S.; Ito, Y.; Hosomi, H.; Ohba, S. Acta Crystallogr.,
Sect. B, submitted for publication.
(4) Scheffer, J. R. Organic Solid State Chemistry; Desiraju, G. R., Ed.;
Elsevier: New York, 1987; Chapter 1.
(23) Hosomi, H.; Ito, Y.; Ohba, S. Acta Crystallogr., Sect. B, submitted
(5) Ito, Y. Synthesis, in press.
for publication.