Photoinduced reactions of chloranil with 1,1-diarylethenes and product photochemistry-intramolecular [2 + 2] (ortho-)cycloadditions of excited enedione's C=C double bond with substituted benzene ring

Article abstract of DOI:10.1021/jo990831r

Photoinduced reactions of chloranil (CA) with 1,1-diarylethenes 1 [(p-X- Ph)₂C=CH₂, X = F, Cl, H, Me] in benzene afforded products 4-14, respectively, with the bicyclo[4.2.0]oct-3-ene-2,5-diones 4, the 6- diarylethenylcyclohexa-2,5-diene-1,4-diones 5, and 2,3,5,6- tetrachlorohydroquinone 13 as the major primary products. The cyclobutane products 4 are formed via a triplet diradical intermediate without involvement of single electron transfer (SET) between the two reactants, while 5 is derived from a reaction sequence with initial SET interaction between ³CA* and the alkene. The 9-arylphenanthrene-1,4-diones 6 and its 10-hydroxy-derivatives 7 are secondary photochemical products derived from 5. The isomeric cage products 9-11 are formed from 4 via intramolecular benzene- alkene [2 + 2] (ortho-)photocycloadditions induced by the triplet excited enedione moiety. The relative amount of the two groups of products (4 and its secondary products 9-11 via non-SET route vs 5 and its secondary products 6, 7, 8, 12, and 14 via SET route) shows a rather regular change, with the ratio of non-SET route products gradually increasing with the increase in oxidation potential of the alkenes and in the positive free energy change for electron transfer (ΔG(ET)) between ³CA* and the alkene, at the expense of the ratio of the products from the SET route. The competition between the SET and non- SET routes was also found to be drastically influenced by solvent polarity, with the SET pathways more favored in polar solvent. Photo-CIDNP investigations suggest the intermediacy of exciplexes or contact ion radical pairs in these reactions in benzene, while in acetonitrile, SET process led to the formation of CA·^- and cation radical of the alkene in the form of solvent separated ion radical pairs and free ions.

Full text of DOI:10.1021/jo990831r

30
J . Org. Chem. 2000, 65, 30-40
P h otoin d u ced Rea ction s of Ch lor a n il w ith 1,1-Dia r yleth en es a n d
P r od u ct P h otoch em istr ysIn tr a m olecu la r [2 + 2]
(Or th o-)Cycloa d d ition s of Excited En ed ion e’s CdC Dou ble Bon d
w ith Su bstitu ted Ben zen e Rin g
J ie Xue, J ian-Wei Xu, Li Yang,† and J ian-Hua Xu*
Department of Chemistry, Nanjing University, Nanjing 210093, China, and Department of Chemistry,
Lanzhou University, Lanzhou 730000, China
Received May 20, 1999
Photoinduced reactions of chloranil (CA) with 1,1-diarylethenes 1 [(p-X-Ph)2CdCH2, X ) F, Cl, H,
Me] in benzene afforded products 4-14, respectively, with the bicyclo[4.2.0]oct-3-ene-2,5-diones 4,
the 6-diarylethenylcyclohexa-2,5-diene-1,4-diones 5, and 2,3,5,6-tetrachlorohydroquinone 13 as the
major primary products. The cyclobutane products 4 are formed via a triplet diradical intermediate
without involvement of single electron transfer (SET) between the two reactants, while 5 is derived
from a reaction sequence with initial SET interaction between 3CA* and the alkene. The
9-arylphenanthrene-1,4-diones 6 and its 10-hydroxy-derivatives 7 are secondary photochemical
products derived from 5. The isomeric cage products 9-11 are formed from 4 via intramolecular
benzene-alkene [2 + 2] (ortho-)photocycloadditions induced by the triplet excited enedione moiety.
The relative amount of the two groups of products (4 and its secondary products 9-11 via non-
SET route vs 5 and its secondary products 6, 7, 8, 12, and 14 via SET route) shows a rather regular
change, with the ratio of non-SET route products gradually increasing with the increase in oxidation
potential of the alkenes and in the positive free energy change for electron transfer (∆GET) between
3CA* and the alkene, at the expense of the ratio of the products from the SET route. The competition
between the SET and non-SET routes was also found to be drastically influenced by solvent polarity,
with the SET pathways more favored in polar solvent. Photo-CIDNP investigations suggest the
intermediacy of exciplexes or contact ion radical pairs in these reactions in benzene, while in
acetonitrile, SET process led to the formation of CA•- and cation radical of the alkene in the form
of solvent separated ion radical pairs and free ions.
In tr od u ction
exceedingly strong electron acceptor ability of CA,14 which
causes a frequent involvement of single electron transfer
(SET) process in its reactions with alkenes. A brief survey
of reported results in this field has revealed some
common trends in the photoreactions of CA with alkenes
concerning the reaction site. For low oxidation potential
alkenes able to engage in SET interaction with triplet
CA (3CA*), as signified by a negative value of the free
energy change for electron transfer (∆GET), which can be
estimated by the Weller equation,17 photoinduced cyclo-
Photoinduced reactions of quinones and haloquinones
with alkenes are characterized by the diversity in reac-
tion modes and regioselectivities.1 The electronic config-
uration of the quinone reactive excited state; the possi-
bility of single electron transfer (SET) between photo-
excited quinone and the alkene in the reactions; special
structural features in the alkene moiety, e.g. the presence
of allylic hydrogens and steric effect toward cycloaddi-
tions; and such reaction conditions as solvent polarity and
irradiation wavelength all seem to play important roles
in deciding the reaction outcome. While the reactions of
the parent benzoquinone with alkenes take place at the
carbonyl group and yield spirooxetane products,1,2 the
reactions of the high potential 2,3,5,6-tetrachlorobenzo-
quinone (chloranil, CA; Chart 1) with alkenes, on the
other hand, display varied reaction sites and regioselec-
tivity, depending on the structure of the alkenes and
reaction conditions.3-13 This is partly the result of the
(6) Ogino, K.; Matsumato, T.; Kozuka, S. J . Chem. Soc., Chem.
Commun. 1979, 643.
(7) (a) Roth, H. D.; Albert, C. J . J . Am. Chem. Soc. 1986, 108, 2013.
(b) Miyashi, T.; Konno, A.; Takahashi, Y.; Kaneko, A.; Suzuki, T.;
Mukai, T. Tetrahedron Lett. 1989, 30, 5297.
(8) Christal, V. M.; Braun, M. Angew. Chem. 1989, 101, 636.
(9) J ones, G., II; Haney, W. A. J . Phys. Chem. 1986, 90, 5410.
(10) (a) Rentzepis, P. M.; Steyert, D. W.; Roth, H. D.; Albert, C. J .
J . Phys. Chem. 1985, 89, 3955. (b) Miyashi, T.; Takahashi, Y.; Mukai,
T.; Roth, H. D.; Schilling, M. L. M. J . Am. Chem. Soc. 1985, 107, 1079.
(11) (a) Oshima, T.; Nagai, T. Bull. Chem. Soc. J pn. 1980, 53, 726.
(b) Tanura, H.; Oshima, T.; Nagai, T.; Inaba, A.; Chihara, H. Acta
Crystallogr., Sect. 3, Struct. Commun. 1996, 46, 609.
(12) Eckert, G.; Goez, M. J . Am. Chem. Soc. 1994, 116, 11999.
(13) (a) Xu, J . H.; Wang, L. C.; Xu, J . W.; Yan, B. Z.; Yuan, H. C. J .
Chem. Soc., Perkin Trans. 1 1994, 571. (b) Xu, J . H.; Song, Y. L.; Zhang,
Z. G.; Wang, L. C.; Xu, J . W. Tetrahedron 1994, 50, 1199.
(14) CA has a half-wave reduction potential of 0.02 V (SCE)15 and
a triplet energy (ET) of 49 kcal/mol.16 The reduction potential of the
triplet excited state is 2.14 V (SCE).
† Lanzhou University.
(1) (a) Bruce, J . M. Photochemistry of Quinones. In The Chemistry
of Quinoid Compounds; Patai, S., Ed.; Wiley: New York, 1974; Part
1. (b) Maruyama, K.; Osuka, A. Recent Advances in the Photochemistry
of Quinones. In The Chemistry of Quinoid Compounds; Patai, S.,
Rappoport, Z., Ed.; Wiley: New York, 1988; Part 2.
(2) Bryth-Smith, D.; Evans, E. H.; Gilbert, A.; McNeill, H. S. J .
Chem. Soc., Perkin Trans. 2 1991, 1587; J . Chem. Soc., Perkin Trans.
1 1992, 485.
(15) (a) Peover, M. E. Nature (London) 1961, 191, 702. (b) Mann,
C. K.; Barnes, K. K. Electrochemical Reactions in Nonaqueous Systems;
Delaker: New York, 1970.
(3) Schenk, G. O. Z. Electrochem. 1960, 64, 997.
(4) Bryce-Smith, D.; Gilbert, A. Tetrahedron Lett. 1964, 3471.
(5) Barltrop, J . A.; Hesp, B. J . Chem. Soc., C 1967, 1625.
(16) Gschwind, R.; Haselback, E. Helv. Chim. Acta 1979, 62, 941.
(17) Weller, A. Z. Phys. Chem. (Wiesbaden) 1982, 133, 193.
10.1021/jo990831r CCC: $19.00 © 2000 American Chemical Society
Published on Web 12/16/1999

Photoinduced Reactions of Chloranil
J . Org. Chem., Vol. 65, No. 1, 2000 31
Ch a r t 1
additions take place at the carbonyl group of CA selec-
tively to give spirooxetane as typical products.6-8,10-12,13b
When the alkenes have a high oxidation potential and a
alkenes with allylic hydrogens and a negative ∆GET with
3CA*, formal hydrogen abstraction (via sequential elec-
tron transfer from the alkene to 3CA* and proton transfer
from the cation radicals of the alkene to CA•- in the
radical ion pairs) and subsequent semiquinone-allylic
radical pair coupling lead to the formation of phenol type
products such as 2.9,13b Since the energetics of the SET
process is dependent on solvent polarity,17 reaction modes
could also be influenced by solvent change, which may
affect the competition of different reaction pathways. As
an example, while the non-SET pathway prevails in
photoinduced reactions of CA with cyclohexene in ben-
3
significantly positive ∆GET value with CA*, the reactions
proceed without SET process involvement. In these cases,
structural features in the alkene that influence the steric
hindrance toward cycloadditions play important roles in
deciding the reaction site. Therefore, while terminal and
cyclic alkenes tend to react with CA’s ethene double bond
to give cyclobutane products,3-6,13b the reaction site is
moved to CA’s carbonyl group for 1,2-disubstituted
(internal) alkenes to yield spirooxetane products.13a For

32 J . Org. Chem., Vol. 65, No. 1, 2000
Ta ble 1. P h otoin d u ced Rea ction s of CA w ith Alk en esa
Xue et al.
ox b
c
E1/2
∆GET
irradiation
time (h)
conversion
of CA
products and
yields (%)d
alkene
(V, SCE)
solvent
(kcal/mol)
1a
1b
1c
2.04
1.96
1.91
C6H6
C6H6
C6H6
7.4
5.5
4.4
10
6.5
10
95.4%
100%
95.9%
4a (7), 5a (1.5), 6a (1.3), 7a (2.5), 8 (4.5), 9a (41)
4b (56), 7b (3), 9b (7.7), 10b (3.9), 11b (1.9)
4c (25), 5c (6.5), 6c (3.5), 7c (10), 9c (3.1),
10c (3.7), 11c (5.5)
1c
1d
1.91
1.79
CH2Cl2
C6H6
-2.8
1.6
15
40
51%
84.4%
4c (6), 5c (41), 6c (7), 7c (23), 13 (5), 14 (4), 15, 16
5d (19), 6d (6.7), 7d (1.7), 12 (18), 13 (9.6),
9d (0.9), 10d (2.7), 11d (12)
a
b
See Experimental Section. Measured by cyclic voltametry with platinum as working and auxiliary electrodes and a saturated calomel
electrode as a reference. c Calculated by eq 1. Yield of isolated product.
d
Ta ble 2. F or m a tion Con sta n ts a n d Mola r Absor p tion
zene to give the 2:1 cyclobutane product 3, the SET
Coefficien ts of Ch a r ge Tr a n sfer Com p lexes betw een CA
pathway predominates when the reaction proceeds in
acetonitrile, affording the phenol product 3. At the same
time, if the reaction is conducted in the medium polarity
solvent dichloromethane, the two pathways compete with
both 2 and 3 as products.13b
Therefore, in the photoinduced reactions of CA with
alkenes, of particular interest are the cases when the
alkenes have such an oxidation potential that their ∆GET
value with 3CA* is close to zero, because under these
circumstances, reaction pathways with and without SET
process involvement run competitively, and the results
would better reveal the full intricacies in reaction modes
and the factors influencing them.
a n d Alk en es
donor
solvent
λCT
KCT
ꢀCT
γa
Sb
1c
1d
CHCl3
CHCl3
442
497
0.67
1.18
697
886
0.9998
0.9993
0.20-0.53
0.07-0.27
a
Linear correlation coefficient in 1/[donor] vs [CA]/ACT plot.
Saturation S ) KCT[donor]/(1 + K[donor]). Deranleau, D. J . Am.
b
Chem. Soc. 1969, 91, 4044.
uncomplexed CA, as well as the molar extinction coef-
ficient of these two species in the wavelength region used
for photolysis. Since light of wavelength longer than 400
nm has been used in the following photolyses, it is
estimated that, in the photolysis of CA (50 mM) with 1c
(100 mM), CA absorbs >80% incident light, while in the
photoreactions of CA (50 mM) with 1d (100 mM), CA
absorbs ∼60% incident light.
In relation to our continuous interest in exploring the
scope and mechanism of photoinduced reactions of CA
with alkenes13 and other electron donors (such as imines18a
and heterocycles18b,c) and with the above-mentioned
considerations in mind, we report here the photoinduced
reactions of CA with 1,1-diarylethenes 1a -d . These
alkenes have oxidation potentials in the range ∼2 V
Irradiation of a benzene solution of CA (0.05 mol L-1
)
and 1a (0.10 mol L-1) with light of wavelength longer
than 400 nm for 10 h resulted in complete consumption
of CA and the formation of products 4a -7a , 8, and 9a
(Table 1). The structures of these products were deter-
mined by spectral and elemental analysis data. The most
characteristic spectral feature of these 1:1 cyclobutane
products of CA with alkenes is the enedione carbonyl Cd
O stretching band at 1700 cm-1, which is higher than
that in CA (∼1680 cm-1) but lower than that in the 2:1
cyclobutane adduct of CA with alkenes as in 3 (at ∼1720
cm-1). The cage compound 9a is derived from the cy-
clobutane product 4a by photoinduced intramolecular
benzene-alkene [2 + 2] (ortho-)cycloaddition, as is evi-
denced by a control experiment in which a pure sample
of 4a is irradiated in a benzene solution under same
conditions as for the photolysis of CA with 1a to give the
cage product in 94% yield. The CdO stretching absorp-
tion in 9a is at 1750 cm-1, indicative of the absence of
any R,â-unsaturation at the carbonyl group. In the 1H
NMR spectrum, the two methylene protons (H-1 and H-2,
Figure 1) appear as a pair of doublets at 3.08 and 3.28
ppm, respectively, with a coupling constant of 11.5 Hz.
The absorption of H-3 (which was originally an aromatic
proton in 4a ) is in the aliphatic region at ∼4 ppm with a
rather small coupling constant of 4.2 Hz with the olefinic
proton H-4 due to the unfavorable dihedral angle between
the planes containing the C-H(3) and C-H(4) bonds.
This is in sharp contrast with the large coupling constant
(9.5 Hz) between the remaining two olefinic protons (H-5
and H-6) in the two coplanar C-H bonds. The 3-vinyl-
2,5,6-trichlorobenzoquinone 5a is a formal substitution
product with one of the chlorine atoms in CA replaced
by the 2,2-diaryl-1-ethenyl group. The two carbonyl
groups in 5a absorb at 1672 and 1682 cm-1, respectively.
A control experiment showed that the phenanthrenedione
3
(SCE) and their SET interactions with CA* are nearly
isothermic in benzene. Furthermore, with the introduc-
tion of a substituent in the para-position of the benzene
ring, the E1/2ox of the alkenes and the corresponding ∆GET
3
value between the alkenes and CA* can be progressively
adjusted to provide an additional probe to test the effect
of the SET process involvement on the reaction mode.
Resu lts
All the alkenes 1a -d are good electron donors (Table
1) and form a ground-state charge-transfer complex
(CTC) with CA in solution, as visualized by the occur-
rence of a reddish color upon the mixing of the alkene
with CA in benzene, CHCl3, or CH2Cl2. The formation
constant (KCT) and extinction coefficient (ꢀCT) of the CTC
between CA and the alkenes 1c and 1d have been
determined with the Benesi-Hildebrand treatment19,20
by measuring the electronic spectra of a CHCl3 solution
of CA at a constant concentration (1.0 mM) with alkenes
of increasing concentrations. The results are in Table 2.
From these data, the degree of charge-transfer complex
excitation involvement in the photolysis can be estimated
by considering the concentrations of the CTC and of the
(18) (a) Chen, C. F.; Zhang, Z. G.; Yan, B. Z.; Xu, J . H. Tetrahedron
Lett. 1994, 35, 9221. (b) Yan, B. Z.; Zhang, Z. G.; Yuan, H. C.; Wang,
L. C.; Xu, J . H. J . Chem. Soc., Perkin Trans. 2 1994, 2545. (c) Xu, J .
H.; Xu, J . W.; Yan, B. Z.; Yuan, H. C. Chinese Chem. Lett. 1991, 2,
831.
(19) Benesi, H. G.; Hildebrand, J . H. J . Am. Chem. Soc. 1947, 71,
2703.
(20) (a) Foster, R. In Molecular Complexes; Foster, R., Ed.; Elek
Science: London, 1974; Vol. 2, Chapter 3. (b) Foster, R. Organic
Charge-Transfer Complexes; Academic Press: New York, 1969.

Photoinduced Reactions of Chloranil
J . Org. Chem., Vol. 65, No. 1, 2000 33
4b, 7b and 9b-11b. Product 4b constituted the main
part of the total products (Table 1). Compounds 9b-11b
came as a mixture of three isomers inseparable by
conventional preparative chromatography either on silica
gel or by thin-layer chromatography. The ratio of the
three isomers in the mixture as determined by high
resolution 1H NMR measurement is 9b:10b:11b ) 4.3:
2.3:1. A pure analytical sample of 9b was finally obtained
by stepwise crystallization of the mixture with benzene-
petroleum ether (60-90 °C) as solvent. The spectral data
of 9b are closely parallel to that of 9a . The structure of
10b and 11b is temporarily assigned mainly by the
1
analysis of their H NMR spectral data and by comparing
them with that of 9a and 9b. Therefore, an inspection of
the molecular model of the three isomers showed that,
in isomer 9b, H-6 is in the shielding area of the nearby
phenyl ring, and this shielding effect is decreased in
isomer 10b and vanished in isomer 11b as the distance
between the corresponding H-6 and the phenyl ring is
becoming larger and larger. As a result, while H-6 in 11b
resonates at 5.77 ppm, typical for olefinic protons in a
cyclohexadiene moiety (5.8-5.9 ppm),21 the absorptions
of H-6 in 10b and 9b occur at 5.42 and 5.22 ppm,
respectively, due to the increasing shielding effect of the
phenyl ring. Also, H-1 and H-2 in isomers 10b and 11b
are slightly deshielded by the flat carbonyl group in the
shallow boat conformation of the cyclohexanedione and
absorb at lower field (3.10 and 3.27 ppm for 10b and 3.18
and 3.47 ppm for 11b respectively) than the H-1 and H-2
in 9b, where the upward-directed carbonyl group in the
rather deeply concave boat conformation of the cyclo-
hexadione has no influence on the absorptions of H-1 and
H-2.
Molecular mechanics calculations of the structures of
the isomers 9b, 10b, and 11b have been carried out with
minimizer, using the DREIDING 2.21 force field method,
and the optimized structures are depicted in Figure 1.
These calculations indicate that all three isomers are
highly strained molecules, among which isomer 9b is the
most stable one with a total strain energy of 197 kcal
mol-1 and the other two are of similar thermodynamic
stability with total energy of 213 and 214 kcal mol-1
respectively.
,
In a control experiment, an isolated pure sample of 4b
was photolyzed in benzene under same conditions as for
the photolysis of CA with 1b to give the mixture of cage
products 9b-11b in 95% yield.
F igu r e 1. Optimized structures of compounds 9b, 10b, and
11b.
The products isolated from the photolysis of CA with
1c in benzene upon the total conversion of CA under the
same conditions as mentioned above were 4c-7c and
9c-11c. Compounds 9c-11c also appear as an insepa-
rable mixture of three isomers with the ratio of 9c:10c:
11c being 1:1.2:1.8. The spectral data of the three isomers
closely resemble that of 9b-11b with the exception that
in the 1H NMR spectrum, isomer 9c has four olefinic
proton absorptions (5.24-6.06 ppm) and three alkanoic
proton absorptions (3.18-3.82 ppm), while isomers 10c
and 11c have only three olefinic proton absorptions but
four alkanoic proton absorptions in similar regions,
respectively. This spectral feature further supports the
structural assignment for the three isomers 9c-11c.
Since the absorptions in the alkanoic region (3-4 ppm)
6a is a secondary product derived from 5a on irradiation
by electrocyclic reactions. Product 7a is formed by the
oxidation of 6a during photolysis by the trace amount of
oxygen in the argon stream which was constantly purged
though the photolyzate during the irradiation. In product
5a , the olefinic proton resonates at 6.54 ppm as a singlet,
while in 6a , this singlet is moved to the aromatic region
at 8.15 ppm, and in 7a , this absorption is missing.
Another prominent aspect in the NMR spectrum of 6a
and 7a is that the absorption of H-5 occurs at an
unusually low field at δ 9.3-9.4 ppm due to the deshield-
ing effect of the nearby carbonyl. Compound 8 is a
hydration product of 5a . The methylene protons in 8
1
resonate at δ 3.74 ppm as a singlet in H NMR, while in
the IR spectrum, the two carbonyl groups absorb at 1683
and 1664 cm-1, respectively.
(21) Clerk, P.; Simon, S. Table of Spectral Data for Structure
Determination of Organic Compounds; Spring-Verlag: Berlin, 1983;
P H235.
Photolysis of a benzene solution of CA with 1b under
same conditions as mentioned above afforded products

34 J . Org. Chem., Vol. 65, No. 1, 2000
Xue et al.
and in the olefinic region (5-6 ppm) in the 1H NMR
spectrum of the isomer mixture are seriously overlapped
and the absorptions for different isomers are barely
discernible to be grouped, we were unable to measure
the coupling constants between the aliphatic protons for
the two isomers 10c and 11c. However, a pure analytical
sample of the main isomer 9c was obtained by stepwise
1
crystallization, and the H NMR, IR, and MS data were
measured on it.
In this case, independent irradiation of a pure sample
of 4c in benzene also afforded the mixture of the cage
products 9c-11c in 87% yield.
In the photoreaction of CA with 1d under same
conditions mentioned above, compounds 5d -7d , 12, 13,
and 9d -11d were obtained as products (Table 1). The
three isomeric cage products were also obtained as a
mixture in which the ratio of 9d :10d :11d is 1:3:13, as
1
F igu r e 2. 1H NMR of CA + 1a in CD3CN.
estimated from the high-resolution H NMR spectrum of
the mixture. Besides 6d and 7d , which are akin in their
origins to the corresponding products 6a , 6c, and 7a -c
in the reactions of 1a -c, another product (12) was found.
Solvent polarity has a profound effect on the reactions.
It is found that photoinduced reactions of CA with 1c in
the more polar dichloromethane took a much longer time
to reach similar conversions and resulted in very different
product distributions than in benzene. Therefore, pho-
tolysis of a dichloromethane solution of CA (0.025 mol
L-1) and 1c (0.050 mol L-1) for 15 h led to a 51%
conversion of CA with the 3-vinyl-2,5,6-trichlorobenzo-
quinone 5c and its secondary products (6c, 7c, and 14)
constituting the majority (76%) of the total products,
while the cyclobutane 4c accounting for only 6%, together
with 4% of tetrachlorohydronquinone 13. A small amount
of the mixture of the dimers of 1c (15 and 16 in the ratio
of 7:1) were also found in the products (see Table 1 and
Experimental Section).
Furthermore, photoinduced reactions of CA with 1c in
acetonitrile gave results entirely contrasting that in
benzene, where only the alkene’s dimers 15 and 16 and
the tetrachlorohydroquinone 13 are formed as products.
Although a complicated mixture of reaction products
is formed in these high-conversion photolyses, the reac-
tion products are found to be much simpler at low
conversions. In the photoreactions of CA with 1c, HPLC
monitoring of the reaction course showed that, at conver-
sions lower than ∼20% in benzene and lower than ∼10%
in dichloromethane, respectively, 4c and 5c are the only
products found and none of the secondary products could
be detected.
and the half-wave reduction potential of the ground state
of CA (+0.02 V, SCE), respectively. The ∆E* is the triplet
energy of CA (50 kcal/mol), and the values 0.06 and 0.38
take into consideration the solvation energy of the ion
radical pair in MeCN and of the exciplex in benzene,
ox
respectively. E1/2 of the olefins 1a -1d were measured
by cyclic voltametry in acetonitrile with a saturated
calomel electrode as reference. The ∆GET values for all
the alkenes investigated are listed in Table 1.
To get more insight into the reaction mechanisms,
photochemically induced dynamic nuclear polarization
(photo-CIDNP22,23) phenomena during the photoreactions
have been investigated. For all the four (CA-1) systems
investigated, CIDNP phenomena were not observed in
deuterated benzene, and no significant difference be-
tween the 1H NMR spectra in the dark and during
irradiation was found in these cases. On the contrary,
irradiation of the four (CA-1) systems in deuterated
acetonitrile led to the observation of distinct CIDNP
phenomena. The nuclear polarization detected came
mainly from the regenerated alkenes. In all cases, the
methylene protons in 1 show abnormally enhanced
absorption (A) as shown in Figure 2. For the (CA-1d )
system (Figure 3), an additional feature is the strong
emission peak (E) for the protons of the p-methyl group.
For all the CA-1 systems investigated, no prominent
polarizations were found in the aromatic region.
Discu ssion
Charge-transfer complex (CTC) formation has been a
common phenomenon when CA is mixed with electron
rich alkenes24a and other electron donors.24b In the case
of 1a -d , their respective addition to a CA solution in
benzene causes significant changes in the long wave-
length region of the UV spectrum. As is shown in Table
These diverse outcomes of the photoinduced reactions
of CA with 1a -d indicated the complexity in reaction
mechanisms. As is for all the reactions involving the
excited state of CA, the possibility of electron-transfer
3
interaction (SET) between CA* and the substrate must
be first considered in any discussion of the reaction
mechanisms. Therefore, free energy changes for electron
2, the absorption maxima (λCT
) of the CTC move
max
bathochromically from 1c to 1d , while the formation
constant of the CTC increases from 1c to 1d , in accord
with the increasing electron donor ability of the alkenes
and a more important contribution of the charge-transfer
3
transfer (∆GET) between CA* and the olefins have been
estimated by the Weller equation (eq 1), where E1/2ox and
ox
red
in benzene
∆GET ) 23.06 [E1/2 - E1/2
-
∆E* + 0.38] (kcal/mol)
(22) Adrian, F. L. In Chemically Induced Magnetic Polarization;
Muus, L. T., Ed.; Reidel: Dordrecht, Holland, 1977; Chapter 5.
(23) (a) Roth, H. D. Pure Appl. Chem. 1988, 60, 933. (b) Roth, H. D.
Acc. Chem. Res. 1987, 20, 343 and references therein.
(24) See for example, (a) Fukuzumi, S.; Kochi, J . A. J . Org. Chem.
1981, 43, 4116. (b) Andrews, L. J .; Keefer, R. M. J . Org. Chem. 1988,
53, 537.
ox
red
in acetonitrile ∆GET ) 23.06 [E1/2 - E1/2
-
∆E* - 0.06] (kcal/mol) (1)
E1/2red are the half-wave oxidation potential of the alkenes

Photoinduced Reactions of Chloranil
J . Org. Chem., Vol. 65, No. 1, 2000 35
Sch em e 1
F igu r e 3. 1H NMR of CA + 1d in CD3CN.
component in the complex. In the photolysis of CA with
1a -d , there are therefore two absorbing speciessthe
uncomplexed CA monomer and the CTC in equilibrium
with it. Under our reaction conditions, the excitation of
uncomplexed CA led to the population of triplet excited
state (3CA*) following fast intersystem crossing.25 From
the data in Table 2, it is estimated that, in the photolysis
of CA with 1a -c respectively under conditions mentioned
above, CA remains the principle absorbing species which
absorb >80% of incident light. For 1d , CA absorbs 60%
incident light. As has been shown in many recent
investigations,9,25a,26-29 the excitation of CTC of a donor-
acceptor pair in a nonpolar solvent such as benzene leads
to the formation of very short-lived singlet contact ion
radical pairs which undergo rapid back-electron transfer
to regenerate the two ground-state components. The
result is a diminished efficiency for the cycloaddition
reaction of the two addends. Therefore, excitation of CTC
constitutes a rather unproductive and energy wasting
route for product formation in the reactions. This could
explain in part the much longer irradiation time required
to reach similar conversions of the starting CA in the
photolysis of CA with 1d compared with that of 1a -c.
Although the results of these reactions of CA with the
alkenes 1 are rather complicated due to the presence of
parallel reaction pathways and extensive secondary
reactions, the products of all these reactions could be
divided into two groups in view of their origins, with 4,
5, and 13 as primary products, respectively. The cyclo-
butane products are formed via a triplet diradical mech-
anism without SET process involvement. This trend of
3CA* reacting at its CdC double bond with terminal or
cyclic alkenes of high oxidation potential with positive
cyclization to afford polycyclic products (as 6) has previ-
ously been found in photoinduced reactions of 2,3-
dichloronaphthoquinone (17) with 1,1-diphenylethenes.30
The mechanism for the formation of 18, by analogy with
that suggested by Maruyama for the photoreaction of 17
with 1, is shown in Scheme 1. Electron transfer from 1
to 3CA* led to the formation of highly polar exciplex and/
or contact ion radical pairs (CIP), from which reaction
sequences (a or b) ensued, resulting in concerted (in route
a) or stepwise (in route b) elimination of hydrogen
chloride and the formation of 5. Products 6 are derived
from
5 by photochemically allowed intramolecular
[π2s + π4a] electrocyclic addition followed by trans-
elimination of hydrogen chloride as was suggested by
Maruyama in the rationalization of similar product in
the reactions of 2,3-dichloronaphthoquinone 17 with 1,1-
diphenylethene.30 Compound 7, 8, 12, and 14 are also
secondary or tertiary products from a common ancestor
5.
It is well-known that photoinduced cycloaddition reac-
tions between alkenes and benzenes take place in three
different ways: 1,2-(ortho-), 1,3-(meta-) and 1,4-(para-)
additions.31 With the 1,4-(para-)cycloadditions being
mainly seen only in photoreactions of benzene with
allenes and cyclobutenes, the 1,2-(ortho-) and 1,3-(meta-)
additions are by far the most commonly found reaction
modes in benzene-alkene photocycloadditions. The pref-
erential reaction mode can be predicted by the ∆Ip (∆GET
)
empirical rule,32,33 and the factors affecting the regio-
selectivity and stereoselectivity in these reaction modes
(26) (a) J ones, G., II; Haney, W. A.; Phan, X. T. J . Am. Chem. Soc.
1988, 110, 1922. (b) J ones, G., II.; Mouli, N. J . Phys. Chem. 1988, 92,
717. (c) J ones, G., II.; Mouli, N.; Haney, W. A.; Bergmark, W. R. J .
Am. Chem. Soc. 1997, 119, 8788.
(27) (a) Kolashi, H.; Funabashi, M.; Kondo, T.; Morita, T.; Ukada,
T.; Mataga, N. Bull. Chem. Soc. J pn. 1984, 57, 3557. (b) Ojima, S.;
Miyasaka, H.; Mataga, N. J . Phys. Chem. 1994, 100, 4797.
(28) (a) Maslak, P.; Chapman, W. M., J r. J . Chem. Soc., Chem.
Commun. 1989, 1809. (b) Maslak, P.; Chapman, W. H., J r. J . Org.
Chem. 1996, 61, 2647.
(29) (a) Bosch, E.; Hubig, S. M.; Kochi, J . K. J . Am. Chem. Soc. 1998,
120, 386. (b) Rathore, R.; Hubig, S. M.; Kochi, J . K. J . Am. Chem. Soc.
1997, 119, 11468. (c) Sun, D.; Hubig, S. M.; Kochi, J . K. J . Am. Org.
Chem. 1999, 64, 2250. (d) Hubig, S. M.; Sun, D.; Kochi, J . K. J . Chem.
Soc., Perkin Trans. 2 1999, 781.
(30) Maruyama, K.; Otsuki, T.; Tai, S. J . Org. Chem. 1985, 50, 52.
(31) For reviews, see (a) Gilbert, A. In Synthetic Organic Photo-
chemistry; Horspool, W. M., Ed.; Plenum: New York, 1984. (b)
Mccullough, J . J . Chem. Rev. 1987, 87, 811. (c) Cornelisse, J . Chem.
Rev. 1993, 93, 615.
3
∆GET values with CA* to afford cyclobutanes has been
noted in many examples, as in photoinduced reactions
of CA with isobutene,5 several 1,3-dienes,5 styrene,3,13b
R-chlorostyrene,13b allyl ethyl ether,13b cyclohexene,13b and
cyclooctene.4 Therefore, products 4 and the isomeric cage
products 9-11 derived from 4 can be regarded as
products from a non-SET route. On the other hand,
product 5 is formed via an SET mechanism. This type of
product, in which a chlorine atom in the haloquinone is
replaced by the alkenyl group, and their secondary
(25) (a) Hubig, S. M.; Beckman, T. M.; Kochi, J . K. J . Am. Chem.
Soc. 1997, 119, 2926. (b) Hilinski, E. F.; Milton, S. V.; Rentzepis, P.
M. J . Am. Chem. Soc. 1983, 105, 5193.

36 J . Org. Chem., Vol. 65, No. 1, 2000
Xue et al.
have been extensively investigated.32,33 While the meta-
cycloadditions are always the predominant reaction
pathways for benzenes and alkenes with similar ioniza-
tion potentials [when the free energy change for electron
transfer (∆GET) between the photoexcited benzene and
the alkene has a significantly large positive value, e.g.
>1.7 eV for the reactions of photoexcited benzene with
alkenes], a large ionization potential difference (∆Ip) and
the presence of charge-transfer interaction between the
two addends (when the ∆GET has a small positive value,
e.g. <1.4 eV for the reactions of photoexcited benzene
with alkenes) favor the 1,2-(ortho-)addition pathway.
Intramolecular version of the m-benzene-alkene photo-
cycloadditions have also been widely investigated and
have emerged as a powerful and versatile tool in the
syntheses of many natural products for their ability to
construct three rings and create up to six stereocenters
simultaneously in the adduct.34 In contrast with this, only
a few examples of intramolecular 1,2-(o-)benzene-alkene
photocycloadditions are known, and reported examples
mainly concern o- and p-butenoxyaceto- and valerophe-
nones,35 o- and p-cyano- or methoxycarbonyl-substituted
4-phenoxybut-1-ene36 systems. In these reactions, it is the
triplet excited benzene that adds in the 1,2-fashion to a
remote alkene moiety linked to the benzene ring by a
flexible chain consisting of three atoms. Photoinduced
1,2-cycloadditions of excited alkene moiety to a ground-
state benzene ring, either intermolecular37 or intramo-
lecular,38 are scarce.
The following considerations suggest that it is the
photoexcited enedione moiety that adds to the ground-
state benzene ring in these intramolecular ortho-cyclo-
additions of 4. (1) As a model compound of the enedione
moiety in 4, 2-cyclohexene-1,4-dione has a long wave-
length absorption band at 368 nm with a log ꢀmax of 1.78,39
while the substituted benzene ring has absorption maxima
in the region ∼260 nm and is transparent above 300
nm.40 (2) [2 + 2] (ortho-)cycloaddition of singlet excited
benzene ring to a ground-state alkene is symmetry
forbidden, while ortho-cycloaddition of a singlet or triplet
excited alkene to a ground-state benzene ring is sym-
metry allowed.41 Therefore, the transformation of the
cyclobutanes 4 to the cage products 9-11 not only
provided a new example of the rare intramolecular
benzene-alkene ortho-photocycloadditions, but to our
knowledge, is also the first reported intramolecular
benzene-alkene ortho-cycloaddition reaction induced by
a triplet excited alkene.
In intermolecular benzene-ethene photocycloaddi-
tions, the well-known ∆Ip-∆GET empirical rules require
a significant electron donor-acceptor interaction between
the two reactants, as reflected by a relatively large ∆Ip
or a slightly positive ∆GET value (e.g. ∼0.5-1.4 kcal/mol
for the reactions of benzene with alkenes), for the ortho-
cycloaddition to become the major reaction mode. Al-
though this correlation between the ∆Ip (∆GET) value and
the reaction mode has not been well-tested in intramo-
lecular benzene-alkene photocycloadditions for the lack
of enough examples, it has found some support in the
intramolecular photocycloadditions of o- and p-butenoxy-
aceto- and valerophenones35 and substituted 4-phenoxy-
but-1-ene systems,36 where the benzene ring and the
ethene double bond are linked by a flexible chain of three
atoms and are not constrained in a fixed geometry. In
these systems, the presence of a electron acceptor sub-
stituent (acyl, methoxycarboxyl, or cyano group) was
found necessary for the triplet excited benzene ring to
add to the remote ethene double bond in the ortho-mode
exclusively. However, steric factors should play a more
important role in intramolecular photocycloadditions
than in intermolecular ones. Also, FMO considerations
show that the ortho-mode is more favored by frontier
orbital overlap than meta- and para-cycloadditions.42
Therefore, with the two chromophores fixed in a geometry
suitable for [2 + 2] cycloadditions in a rigid molecular
framework, the ortho-mode could still be the main
reaction pathway even in the absence of apparent elec-
tron donor-acceptor relationship between the benzene
and ethene moieties. This situation is found in the
intramolecular ortho-cycloadditions of compound 19.43 In
the present case of intramolecular cycloadditions from
4, considerations of the molecular geometry and the
electron donor-acceptor relationship between the ben-
zene and enedione fragments show that both factors may
have come into play to bring the exclusive [2 + 2] (ortho-)
cycloadditions. In 4, the benzene ring and the enedione
CdC double bond are brought close to each other. An
inspection of molecular model shows that the benzene
ring is situated right above the CdC double bond at a
distance convenient for bond formation. However, this
molecular geometry is not as rigid as in 19; as a result,
three different regioisomeric cycloadducts could be formed.
At the same time, a comparison of the Ip values of the
two chromophores reveals that there is an unambiguous
electron donor-acceptor relationship between the two.
As a structural model of the dichloro-substituted enedi-
one moiety in 4, the trans-3,4-dichloro-3-hexene-2,5-dione
20 has an Ip of 9.77 eV,44 while as a model of the most
electron deficient benzene ring in 4, the fluorobenzene
has an Ip of 9.17 eV.45 This large ∆Ip also favors the
[2 + 2] cycloadditions.
(32) (a) Bryce-Smith, D.; Gilbert, A.; Orger, B. H.; Tyrrell, H. M. J .
Chem. Soc. Chem. Commun. 1974, 334. (b) Gilbert, A. Pure App. Chem.
1980, 52, 2669. (c) Bryth-Smith, D.; Gilbert, A.; Mattay, J . Tetrahedron
Lett. 1986, 21, 6011.
(33) (a) Mattay, J . Tetrahedron 1986, 41, 2393. (b) Mattay, J .
Tetrahedron 1986, 41, 2405. (c) Mattay, J . J . Photochem. 1987, 37,
167.
(34) See, for example Wender, P. A.; L. Siggel, Nuss, J . H. [3 + 2]
and [5 + 2] Arene-Alkene Photocycloadditions. In Comprehensive
Organic Syntheses; Trost, B. M., Fleming, I., Paquette, L. A., Eds.;
Pergamon: Elmsford, NY, 1991; Vol. 5, 645-673.
(35) (a) Wagner, P. J .; Nahm, K. J . Am. Chem. Soc. 1987, 109, 6528.
(b) Wagner, P. J .; Nahm, K. J . Am. Chem. Soc. 1987, 109, 4404. (c)
Wagner, P. J .; Cheng, K. L. Tetrahedron Lett. 1993, 34, 907. (d)
Wagner, P. J .; Alehashem, H. Tetrahedron Lett. 1993, 34, 914.
(36) Cosstick, K. B.; Dew, M. G. B.; Gilbert, A. J . Chem. Soc., Chem.
Commun. 1987, 1867.
It is interesting to note that, in these intramolecular
benzene-alkene ortho-cycloadditions, the reaction site on
the benzene ring shows a remarkable dependence on the
substituent, and the distribution of the isomeric cage
(37) Angus, H. J . F.; Bryce-Smith, D. Proc. Chem. Soc. 1957, 287.
(38) Aoyama, H.; Arata, Y.; Omote, Y. J . Chem. Soc., Chem.
Commun. 1990, 736.
(39) Catalani, L. H.; Toscano, V. G.; de Arruda Campos, I. P. J .
Chem. Soc., Perkin Trans. 2, 1995, 1863.
(42) Houk, K. N. Pure Appl. Chem. 1982, 54, 1633.
(43) (a) Fischer, G.; Beckman, E.; Prinzbach, H.; Rihs, G.; Wirz, J .
Tetrahedron Lett. 1986, 1273. (b) Sedelmeyer, G.; Fessner, W. D.;
Grund, C.; Spurr, P. R.; Fritz, H.; Prinzbach, H. Tetrahedron Lett. 1986,
1277 and references therein.
(40) Forbes, W. F. Can. J . Chem. 1959, 37, 7.
(41) (a) Bryce-Smith, D.; Gilbert, A. Tetrahedron 1976, 32, 1309.
(b) Bryce-Smith, D.; Gilbert, A. Tetrahedron 1977, 33, 2459.
(44) Gleiter, R.; J ahne, G.; Oda, M.; Iyoda, M. J . Org. Chem. 1985,
50, 678.
(45) Fischer-Hzalmars, I.; Meza, S. Acta Chem. Scand. 1972, 2991.

Photoinduced Reactions of Chloranil
J . Org. Chem., Vol. 65, No. 1, 2000 37
products shows a regular change as the substituent on
the benzene ring changes. As a rule of thumb, the
electron deficient CdC double bond in the dichloro-
substituted enedione moiety is inclined to add to that
CdC bond in the benzene ring which bears a electron-
donating substituent or else has no electron-withdrawing
substituent. Therefore, for the reaction of 4a , compound
9a is the only product in which the enedione olefinic bond
adds to the benzene ring at a CdC bond without the
fluorine substituent. In the reactions of 4b, the product
ratio is 3:2.3:1, with the product not involving the
chlorine-bearing CdC bond (9b) being the major one. For
the reaction of 4c in which the benzene ring are not
substituted, the selectivity in reaction site is not obvious,
and the three isomers are formed in comparable amounts
(9c:10c:11c is 1:1.2:1.8). In contrast, the product distri-
bution in the reaction of 4d is reversed from that of 4a
and 4b, and the ratio of 9d :10d :11d is 1:3:13. In this case,
the enedione olefinic bond adds to the methyl-bearing
CdC bond in the benzene ring predominately.
1c have been found in SET reactions of 1c in MeCN with
9,10-dicyanoanthracene (DCA), etc., as an acceptor sen-
sitizer.46
It is further noted that the products in photoinduced
reactions of CA with 1c in MeCN are totally different
3
then that in benzene. In MeCN, the ∆GET between CA*
and 1c turns to be more negative (-6 kcal mol-1).
Therefore, the non-SET reaction pathways are sup-
pressed and the corresponding products 4 and 9-11 are
fully eliminated. Furthermore, solvent polarity also affect
the form of the intermediate generated in SET process.
Therefore, not only the non-SET products (4, 9-11) are
no longer formed as the ∆GET becomes more negative,
but the products 5-8 which are formed via exciplex or
CIP intermediates in benzene and in dichloromethne
have also been eliminated as a result of the further
dissociation of the solvent separated ion radical pairs
formed in MeCN, and the only products formed are the
alkene dimers derived from attack of the escaped cation
radical to a neutral molecule of 1c.
To get more information concerning the reaction mech-
anisms, photo-CIDNP experiments have been conducted
for these CA-1 systems.
As mentioned above, the products from photoinduced
reactions of CA with 1a -d can be divided into two groups
according to their origin. The cyclobutane products 4 and
the isomeric cage products derived from 4 can be re-
garded as products via the non-SET route, while products
5 and their secondary products 6-8, 12, and 14 are
derived from the SET route. The results listed in Table
1 show that product distributions in the photoreactions
of CA with 1a -d respectively display a rather regular
change with the ∆GET values as the substituent on the
benzene ring changes. For 1a and 1b, which have the
As mentioned above, photo-CIDNP phenomena were
not observed in deuterated benzene solution for all the
(CA-1) systems investigated. Electron-transfer quench-
ing of the 3CA* by an electron donor in a nonpolar solvent
such as benzene resulted in the formation of triplet
contact ion radical pairs which can be viewed as excited
complexes with a large charge-transfer component con-
tribution.20 Since in the radical pair theory of the CIDNP
phenomena,22 the evolution of nuclear polarizations in
the in-cage geminate and out-of-cage trapping products
require a large distance (9 Å) between the cation and
anion radicals, which corresponds to a compactness of
solvent separated ion pairs (SSIP) or free ions, the
absence of nuclear polarizations in the regenerated
alkenes and in the product 5 formed via the SET route
further substantiated that products 5 were formed in
exciplex or CIP intermediate in benzene. In contrast with
the situation in benzene, strong nuclear polarizations
have been detected for all the (CA-1) systems in deu-
terated acetonitrile, and the polarization patterns are
compatible with an SET mechanism, in which the polar-
ized starting alkenes are regenerated by in cage back
electron transfer between CA•- and 1•+ in singlet ion
radical pairs (single exit). The other parameters that
decide the polarization direction by Kaptain’s rule47 for
net effect (eq 3) are as follows. The ion radical pairs
ox
highest E1/2 and comparatively larger positive ∆GET
values of the four alkenes, the non-SET route products
predominate. With the decrease in oxidation potential of
the alkenes, the SET route products increase at the cost
of the non-SET products. For 1d , the strongest donor of
the four, the SET products make up the majority of the
total products. This clear tendency that the ratio of SET
products rises with the decreasing ∆GET value strongly
supports our discussion on the mechanistic origin of the
two groups of products and provided an example to reveal
the subtle influence of the SET energetics as reflected
by e.g. Weller’s ∆GET value on the reaction outcome.
Since solvent polarity has a significant bearing on the
energetics of SET process, the competition of SET vs non-
SET reaction pathways are found to be remarkably
influenced by solvent change. Therefore, as can be seen
in Table 1, in sharp contrast with the reactions of CA
with 1c in benzene where the non-SET products (4c, 9c-
11c) are formed more preferably than the SET products
(5c-7c), the products from the SET route (5c-7c, and
14) constitute the overwhelming majority of the total
products in the reaction of CA with 1c in the more polar
CH2Cl2, where the ∆GET value between 3CA* and 1c is
-2.8 kcal mol-1, as is estimated by eq 2.17 At the same
Γnet ) µꢀ∆ga
(3)
generated in SET process are of triplet multiplicity
3
because they have a triplet precursor CA*. The alkene
cation radicals should have a smaller g value than CA•-
(the g value of CA•- is 2.004648), the ∆g value is therefore
negative. The methylene protons in 1•+ have a negative
hyperfine coupling constant with the unpaired electron.49
These lead to an abnormally enhanced absorption (A) for
ox
red
∆GET ) 23.06[E1/2 - E1/2 - ET + 2.6/ꢀ - 0.13]
(kcal/mol-1) (2)
(46) Matters, S. L.; Farid, S. J . Am. Chem. Soc. 1983, 105, 1386.
(47) Kaptain, K. J . Chem. Soc., Chem. Commun. 1971, 732.
(48) Roth, H. D.; Schilling, M. L. M. J . Am. Chem. Soc. 1980, 102,
4303.
(49) (a) Arnold, D. R.; Du, X. Y.; Chen, J . Can. J . Chem. 1995, 73,
307. (b) Helmstreit, W.; Hanschmann, G. J . Prakt. Chem. 1980, 322,
981. (c) Hanschmann, G.; Helmstreit, W.; Kochler, H. J . Z. Phys. Chem.
(Leipzig) 1980, 261, 81.
time, a small amount of dimers of 1c (15 and 16 in a
ratio of 7:1) has also been obtained in the reactions in
CH2Cl2, which are derived in a reaction sequence initi-
ated by attack of the escaped cation radical 1c•+ toward
neutral 1c. Similar formation of dimers 15 and 16 from

38 J . Org. Chem., Vol. 65, No. 1, 2000
Xue et al.
and the absorption coefficient (ꢀCT) of the CTC at its absorption
maxima were calculated by the Benesi-Hildebrand treatment.
Cyclic Volta m m etr ic Mea su r em en ts on 1,1-Dia r yleth -
ylen es. Oxidation potential of the alkenes were measured by
cyclic voltammetry in dry acetonitrile with tetrabutylammo-
nium perchlorate (0.1 M) as supporting electrolyte. Platinum
electrodes were used as working electrode and auxiliary
electrode, and the reference electrode was a saturated calomel
electrode (SCE). The scan speed was 100 mV s-1. The solutions
were 0.1-1 mM in the substrate and were purged with dry
nitrogen for 10 min before measurements to remove dissolved
oxygen.
P h oto-CIDNP Mea su r em en ts. Photo-CIDNP measure-
ments were performed on a Bruker AC 80 MHz NMR spec-
trometer. A solution of CA (4 mM) and the substrate (8 mM)
in C6D6 or CD3CN was purged with dry argon for 5 min to
remove dissolved oxygen. The 1HNMR spectrum of the ex-
ample solution was then taken before, during, and after
irradiation. A glass filter which cuts off light with λ < 310 nm
was used during the irradiation.
the methylene protons in the regenerated 1, as is actually
observed in Figure 2. Since in the cation radical of 1 spin
density mainly concentrates at the methylene carbon
atom and the carbon atoms in the phenyl group share
only a small fraction of spin density,49 no significant
polarization was observed for the ring protons. In the case
of 1d , the protons of the p-methyl group on the benzene
ring have a positive hyperfine coupling constant with the
unpaired spin, while all the other parameters in eq 3 are
unchanged, as mentioned above for the methylene pro-
tons. This resulted in an emission peak (E), which is also
in accord with the experimental result. Therefore, these
photo-CIDNP phenomena in acetonitrile suggested the
formation of SSIP or free ions in this solvent by SET
3
between CA* and 1.
In summary, in photoinduced reactions of the high
potential quinone chloranil (CA) with substituted 1,1-
diarylethylenes 1a -d , reaction pathways with and with-
out electron transfer (SET) from the alkenes to 3CA* run
competitively as a result of the nearly isothermic nature
of the SET process. This competition is subtly influenced
by the change of substituent in the benzene ring of the
alkene, which causes a progressive change of the alkene
oxidation potential and a corresponding change in SET
energetics. Therefore, the relative amount of the non-
SET products vs the SET products increases with an
increasingly positive free energy change (∆GET) with
3CA*, whereas that of the non-SET products displays an
inverse trend. Photo-CIDNP investigations provided
further evidence for the intermediacy of polar exciplex
or contact ion radical pairs (CIP) in benzene and solvent
separated ion radical pairs (SSIP) or free ions in aceto-
nitrile in the reactions. The formation of the isomeric cage
products 9-11 from the cyclobutane products 4 provided
the first example of intramolecular [2 + 2] ortho-cyclo-
addition of the ethene CdC bond of a triplet excited
enedione moiety to the benzene ring. These diversified
reaction outcomes caused by parallel reaction pathways
and various secondary reactions of the products showed
that the long standing pursuit of new photochemical
reactions between haloquinones and alkenes is still far
from being exhausted and remains an area from which
both synthetic application and mechanistic explorations
can be rewardingly exploited.
Molecu la r Mech a n ics Ca lcu la tion s. Computational re-
sults were obtained using software programs Cerious2 3.5 from
Molecular Simulations Inc.; the structures of compounds 9b-
11b were drawn with a 3D-sketcher of visualizer and opti-
mized with minimizer, using DREIDING 2.21 force field. The
graphic displays were printed out from the Cerius2 molecular
modeling system. All the calculations were done on SGI Origin
200 server.
Gen er a l P r oced u r e of P r ep a r a tive P h otolysis of CA
w ith th e 1,1-Dia r yleth ylen es. The light source was a 500
W medium-pressure mercury lamp in a cooling water jacket
which was further surrounded by a 1 cm thick layer of solution
filter (20% aqueous sodium nitrite) to cut off light of wave-
length shorter than 400 nm. The solution of the respective 1,1-
diarylethylene and CA was placed in glass tubes (20 mL each)
around the light source and was photolyzed for the time
indicated in Table 1. A stream of dry argon was purged
through the solution 0.5 h prior to and throughout the
irradiation. The reaction course was monitored by TLC and
HPLC. At the end of reaction, the solvents were removed in
vacuo, and the photoproducts were separated by column
chromatography on silica gel column with petroleum ether (bp
60-90 °C)-ethyl acetate as eluents. For the isomeric cage
products (9-11) obtained as a mixture in the reactions of CA
with 1b-d , the relative ratio of the three isomers was
determined by 500 MHz 1H NMR measurements. Some of the
pure isomers (9a -c and 11d ) could be isolated from the
isomeric mixture by stepwise crystallization from chloroform
(or benzene)-petroleum ether (60-90 °C).
Ir r a d ia tion of CA w ith 1a . A solution of CA (984 mg, 4
mmol) and 1,1-bis(4,4-fluorophenyl)ethylene (1728 mg, 8 mmol)
in benzene (80 mL) was photolyzed for 10 h to lead to a 95.4%
conversion of CA. Work up as mentioned above afforded 4a
(120 mg, 7%), 5a (25 mg, 1.5%), 6a (20 mg, 1.3%), 7a (40 mg,
2.5%), 8 (80 mg, 4.5%), and 9a (720 mg, 41%).
Exp er im en ta l Section
Gen er a l. Melting points are uncorrected. The chemical
shifts are reported in δ (ppm) relative to tetramethylsilane as
internal standard. J values are given in hertz. Infrared spectra
were taken in KBr pellets. Acetonitrile (AR grade) was first
refluxed with phosphorus pentoxide and distilled and then
refluxed with anhydrous potassium carbonate and redistilled.
Benzene (AR grade) was dried with sodium and distilled before
use. Chloranil (CA) (CP grade) was recrystallized twice from
acetone. Methanol (AR grade) was refluxed with magnesium
and distilled. 1,1-Diarylethylenes were prepared by literature
procedures. Other reagents were CP or AR grade and were
used as received without further purification
CTC F or m a tion Con sta n ts (KCT) a n d ꢀCT Mea su r e-
m en ts. UV-vis spectroscopic measurements were carried out
with a matched pair of quartz cells. Solutions of CA (1 mM)
with different concentrations of the donor in HCCl3 were
purged with dry argon for 5 min to remove dissolved oxygen.
The spectrum of each of the solutions was then recorded
against a blank containing the same concentration of donor
in the same solvent. At least five different donor concentrations
1,3,4,6-Tetr ach lor o-7,7-bis(4-flu or oph en yl)bicyclo[4,2,0]-
oct-3-en e-2,5-d ion e (4a ). Recrystallization from petroleum
ether (bp 60-90 °C)-chloroform gave slightly yellow crys-
1
tals: mp 159-160 °C; H NMR (60 MHz, CDCl3) δ 3.49 (d, J
) 12.0 Hz, 1H), 4.29 (d, J ) 12.0 Hz, 1H), 6.69-7.53 (m, 8H);
IR (KBr) 1700, 1600, 844 cm-1; MS (EI) m/z 426 (M+ - Cl,
62), 390 (100). Anal. Calcd for C20H10Cl4F2O2: C, 51.98; H, 2.18.
Found: C, 51.82; H, 2.21.
2,3,5-Tr ich lor o-6-(2,2-bis(4-flu or op h en yl)eth en -1-yl)cy-
cloh exa -2,5-d ien e-1,4-d ion e (5a ). Recrystallization from
petroleum ether (bp 60-90 °C)-ethyl acetate gave red
needles: mp 196 °C; 1H NMR (60 MHz, CDCl3) δ 6.54 (s, 1H),
7.00-7.40 (m, 8H); IR (KBr) 1682, 1672, 1598, 822 cm-1; MS
(EI) m/z 424 (M+, 0.66), 87 (100). Anal. Calcd for C20H9-
Cl3F2O2: C, 56.40; H, 2.12. Found: C, 56.48; H, 2.09.
2,3-Dich lor o-6-flu or o-9-(4-flu or oph en yl)ph en an th r en e-
1,4-d ion e (6a ). Recrystallization from petroleum ether (bp
60-90 °C)-chloroform gave red needles: mp 218-220 °C; 1H
NMR (60 MHz, CD3COCD3) δ 7.50 (br, 5H), 7.90-8.10 (m, 1H),
were used for each donor. The CTC formation constants (KCT
)
8.15 (s, 1H), 9.40 (m, 1H); IR (KBr) 1675, 1600, 1575, 835 cm-1
;

Photoinduced Reactions of Chloranil
J . Org. Chem., Vol. 65, No. 1, 2000 39
MS (EI) m/z 388 (M+, 100). Anal. Calcd for C20H8Cl2F2O2: C,
61.70; H, 2.06. Found: C, 61.42; H, 2.20.
Hz, 1H), 5.77 (d, J ) 3.3 Hz, 1H), 6.01 (d, J ) 9.9 Hz, 1H),
6.95 (d, J ) 8.3 Hz, 2H), 7.42 (d, J ) 8.3 Hz, 2H).
2,3-Dich lor o-6-flu or o-9-(4-flu or op h en yl)-10-h yd r oxy-
p h en a n th r en e-1,4-d ion e (7a ). Recrystallization from petro-
leum ether (bp 60-90 °C)-chloroform gave red needles: mp
211-212 °C; 1H NMR (500 MHz, CDCl3) δ 6.13 (s, OH), 7.23-
7.33 (m, 4H), 7.47 (t, J ) 6.7 Hz, 1H), 8.40 (dd, J ) 6.7 Hz,
10.7 Hz, 1H), 9.30 (d, J ) 10.7 Hz, 1H); IR (KBr) 3400, 1678,
1597, 850 cm-1; MS (EI) m/z 404 (M+, 100). Anal. Calcd for
Ir r a d ia tion of CA w ith 1c. In ben zen e. A solution of CA
(1230 mg, 5.0 mmol) and 1,1-diphenyl ethylene (1800 mg, 10
mmol) in benzene (100 mL) was photolyzed for 10 h to lead to
a 95.9% conversion of CA. Workup as before afforded 4c (500
mg, 25%), 5c (120 mg, 6.5%), 6c (60 mg, 3.5%), 7c (180 mg,
10%) and a mixture of the isomeric cage products 9c, 10c, and
11c (250 mg, 12%) in a ratio of 1:1.2:1.8 as estimated from
1
C
20H8Cl2F2O3: C, 59.26; H, 1.98. Found: C, 59.50; H, 1.95.
the 500 MHz H NMR spectrum of the mixture.
2,3,5-Tr ich lor o-6-(2-h yd r oxy-2,2-b is(4-flu or op h en yl)-
In d ich lor om eth a n e. A solution of CA (738 mg, 3.0 mmol)
and 1,1-diphenylethylene (1080 mg, 6.0 mmol) in dichloro-
methane (120 mL) was photolyzed for 15 h to lead to a 51%
conversion of CA. Workup as before afforded 4c (40 mg, 6%),
5c (260 mg, 41%), 6c (40 mg, 7%), 7c (130 mg, 23%), 13 (20
mg, 5%), 14 (25 mg, 4%), and a mixture of products 15 and 16
(25 mg) in a ratio of 7:1 as estimated from the 300 MHz 1H
NMR spectrum of the mixture.
eth yl)cycloh exa -2,5-d ien e-1,4-d ion e (8). Recrystallization
from petroleum ether (bp 60-90 °C)-ethyl acetate gave yellow
1
needles: mp 182-183 °C; H NMR (500 MHz, CDCl3) δ 3.77
(s, 2H), 7.00 (t, J ) 8.5 Hz, 4H), 7.34 (dd, J ) 6.2, 8.5 Hz, 4
H); IR (KBr) 3550, 1683, 1664, 1600, 1578, 824 cm-1; MS (EI)
m/z 424 (M+ - H2O, 0.37), 123 (100). Anal. Calcd for C20H11
Cl3F2O3: C, 54.11; H, 2.48. Found: C, 54.01; H, 2.57.
-
11-F lu or o-4-(4-flu or op h en yl)-1,3,6,8-tetr a ch lor op en ta -
cyclo[6,6,0,0,3,60,4,50,4,1409,14]tetr adeca-10,12-dien e-2,7-dion e
(9a ). Recrystallization from benzene gave a white solid: mp
1,3,4,6-Tetr ach lor o-7,7-diph en ylbicyclo[4,2,0]oct-3-en e-
2,5-d ion e (4c). Recrystallization from petroleum ether (bp 60-
90 °C)-chloroform gave slightly yellow crystals: mp 143-145
°C; 1H NMR (60 MHz, CDCl3) δ 3.46 (d, J ) 12.0 Hz, 1H),
4.28 (d, J ) 12.0 Hz, 1H), 6.90-7.60 (m, 10H); IR (KBr) 1701,
1596, 750, 710 cm-1; MS (EI) m/z 424 (M+, 0.9), 390 (100). Anal.
Calcd for C20H12Cl4O2: C, 56.37; H, 2.84. Found: C, 55.96; H,
2.70.
1
181-184 °C; H NMR (500 MHz, CD3COCD3) δ 3.08 (d, J )
11.5 Hz, 1H), 3.28 (d, J ) 11.5 Hz, 1H), 4.08 (dd, J ) 4.2, 6.5
Hz, 1H), 5.19 (d, J ) 6.5 Hz, 1H), 5.74 (dd, J ) 4.2, 9.5 Hz,
1H), 5.97 (t, J ) 9.5 Hz, 1H), 7.18-7.22 (m, 2H), 7.42 (br, 2H);
IR (KBr) 1750, 1673, 1604, 1514, 1409, 1163, 1110, 840, 817,
771, 710, 675 cm-1; MS (EI) m/z 425 (M+ - Cl, 7.4), 83 (100);
HRMS (m/z) calcd for C20H9Cl3F2O2 (M+ - HCl) 423.9637,
found 423.9620 (mmu 1.6).
2,3,5-Tr ich lor o-6-(2,2-d ip h en ylet h en -1-yl)cycloh exa -
2,5-d ien e-1,4-d ion e (5c). Recrystallization from petroleum
ether (bp 60-90 °C)-chloroform gave red needles: mp 171
1
Ir r a d ia tion of CA w ith 1b. A solution of CA (1476 mg, 6
mmol) and 1,1-bis(4-chlorophenyl)ethylene (2988 mg, 12 mmol)
in benzene (120 mL) was photolyzed for 6.5 h to lead to the
total conversion of CA and afforded 4b (1650 mg, 56%), 7b
(80 mg, 3%), and a mixture of the isomeric cage products 9b,
10b, and 11b (400 mg, 14%) in a ratio of 4:2:1 as estimated
°C; H NMR (60 MHz, CDCl3) δ 6.60 (s, 1H), 7.00-7.40 (m,
10H); IR (KBr) 1672, 1600, 742, 705 cm-1; MS (EI) m/z 388
(M+, 66), 255 (100). Anal. Calcd for C20H11Cl3O2: C, 61.62; H,
2.82. Found: C, 61.74; H, 2.79.
2,3-Dich lor o-9-p h en ylp h en a n th r en e-1,4-d ion e (6c). Re-
crystallization from ethyl acetate gave red needles: mp 170-
171 °C; 1H NMR (60 MHz, CD3COCD3) δ 7.50 (br, 5H), 7.30-
8.20 (m, 3H), 8.17 (s, 1H), 9.58 (d, J ) 9.7 Hz, 1H); IR (KBr)
1672, 1593, 764, 702 cm-1; MS (EI) m/z 352 (M+, 100), 317
(23), 289 (28), 226 (69). Anal. Calcd for C20H10Cl2O2: C, 68.01;
H, 2.85. Found: C, 67.72; H, 2.66.
2,3-Dich lor o-9-p h en yl-10-h yd r oxyp h en a n t h r en e-1,4-
d ion e (7c). Recrystallization from petroleum ether (bp 60-
90 °C)-chloroform gave a red solid: mp 243 °C; 1H NMR (500
MHz, CD3Cl3) δ 6.16 (s, OH), 7.26 (s, 3H), 7.56 (d, J ) 7.0 Hz,
1H), 7.60 (t, J ) 7.0 Hz, 2H), 7.68 (t, J ) 7.0 Hz, 1H), 7.79 (t,
J ) 7.8 Hz, 1H), 9.35 (d, J ) 7.8 Hz, 1H); IR (KBr) 3480, 1680,
1650, 1600, 760, 710 cm-1; MS (EI) m/z 368 (M+, 100). Anal.
Calcd for C20H10Cl2O3: C, 65.22; H, 2.72. Found: C, 65.16; H,
2.72.
1
from the 500 MHz H NMR spectrum of the mixture.
1,3,4,6-Tetr ach lor o-7,7-bis(4-ch lor oph en yl)bicyclo[4,2,0]-
oct-3-en e-2,5-d ion e (4b). Recrystallization from petroleum
ether (bp 60-90 °C)-chloroform gave a slightly yellow crys-
tal: mp 163-164 °C; 1H NMR (60 MHz, CDCl3) δ 3.46 (d, J )
12.0 Hz, 1H), 4.26 (d, J ) 12.0 Hz, 1H), 7.10-7.60 (m, 8H); IR
(KBr) 1717, 1702, 1590, 837 cm-1; MS (FAB) m/z 457 (M+
Cl, 5.2), 178 (100). Anal. Calcd for C20H10Cl6O2: C, 48.89; H,
1.89. Found: C, 48.53; H, 2.04.
-
2,3-Dich lor o-6-ch lor o-9-(4-ch lor op h en yl)-10-h yd r oxy-
p h en a n th r en e-1,4-d ion e (7b). Recrystallization from petro-
leum ether (bp 60-90 °C)-chloroform gave red needles: mp
192 °C; 1H NMR (500 MHz, CDCl3) δ 6.11 (s, OH), 7.21(d, J )
8.3, 2H), 7.45 (m, 1H), 7.58 (d, J ) 8.3 Hz, 1H), 7.64 (dd, J )
8.5, 1.6 Hz, 1H), 8.31 (d, J ) 8.5 Hz, 1H), 9.62 (d, J ) 1.6 Hz,
1H); IR (KBr) 3460, 1675, 1650, 1600, 865 cm-1; MS (EI) m/z
436 (M+, 42), 406 (100). Anal. Calcd for C20H8Cl4O3: C, 54.79;
H, 1.83. Found: C, 54.87; H, 1.95.
4-(4-Ch r olop h en yl)-1,3,6,8,11-p en ta ch lor op en ta cyclo-
[6,6,0,0,3,60,4,50,4,1409,14]tetr adeca-10,12-dien e-2,7-dion e (9b).
This isomer 9b could be isolated from the mixture of 9b, 10b,
and 11b by stepwise crystallization from benzene-petroleum
ether (60-90 °C) to give colorless crystals: mp 192-194 °C;
1H NMR (500 MHz, CDCl3) δ 2.96 (d, J ) 11.2 Hz, 1H), 3.11
(d, J ) 11.2 Hz, 1H), 3.92 (d, J ) 3.4 Hz, 1H), 5.22 (d, J )
10.2 Hz, 1H), 5.80 (d, J ) 3.4 Hz, 1H), 5.94 (d, J ) 10.2 Hz,
1H), 7.09 (br, 2H), 7.40 (d, J ) 8.9 Hz, 2H); IR (KBr) 1755,
1596, 826 cm-1; MS (FAB) m/z 457 (M+ - Cl, 0.3), 91 (100).
HRMS (m/z) calcd for C20H9Cl5O2 (M+ - HCl) 455.9047, found
455.9042 (mmu 0.4).
4-(4-Ch r olop h en yl)-1,3,6,8,14-p en ta ch lor op en ta cyclo-
[6,6,0,0,3,60,4,50,4,1109,14]tetr adeca-10,12-dien e-2,7-dion e (10b):
1H NMR (500 MHz, CDCl3) δ 3.10 (d, J ) 11.6 Hz, 1H), 3.27
(d, J ) 11.6 Hz, 1H), 3.58 (d, J ) 3.3 Hz, 1H), 5.42 (d, J )
10.2 Hz, 1H), 5.74 (d, J ) 3.3 Hz, 1H), 5.96 (d, J ) 10.2 Hz,
1H), 6.88 (d, J ) 8.2 Hz, 2H), 7.37 (d, J ) 8.2 Hz, 2H).
4-(4-Ch r olop h en yl)-1,3,6,8,9-p en t a ch lor op en t a cyclo-
[6,6,0,0,3,60,4,50,4,1209,14]tetr adeca-10,12-dien e-2,7-dion e (11b):
1H NMR (500 MHz, CDCl3) δ 3.18 (d, J ) 11.7 Hz, 1H), 3.47
(d, J ) 11.7 Hz, 1H), 3.82 (d, J ) 3.3 Hz, 1H), 5.34 (d, J ) 9.9
4-P h e n yl-1,3,6,8-t e t r a ch lor op e n t a cyclo[6,6,0,0,3,60,
4,50,4,1409,14]tetr a d eca -10,12-d ien e-2,7-d ion e (9c). This
isomer 9c could be isolated from the mixture of 9c-11c by
stepwise crystallization from chloroform-ethyl ether to give
colorless crystals: mp 196-198 °C; 1H NMR (500 MHz, CDCl3)
δ 3.18 (d, J ) 11.5 Hz, 1H), 3.49 (d, J ) 11.5 Hz, 1H), 3.82 (br,
1H), 5.24 (d, J ) 9.6 Hz, 1H), 5.72 (d, J ) 9.6 Hz, 1H), 5.89 (d,
J ) 6.8 Hz, 1H), 6.06 (dd, J ) 6.8, 9.6 Hz, 1H), 7.02 (d, J )
7.4, 2H), 7.36-7.43 (m, 3H); IR (KBr) 1750, 760, 710 cm-1
MS (FAB) m/z 425 (M+ + 1, 17), 91 (100). HRMS (m/z) calcd
for
;
C
20H11Cl3O2 (M+ - HCl): 387.9826 found: 387.9800
(mmu: 2.5).
4-P h e n yl-1,3,6,8-t e t r a ch lor op e n t a cyclo[6,6,0,0,3,60,
4,50,4,1109,14]tetr a d eca -10,12-d ien e-2,7-d ion e (10c): H NMR
1
(500 MHz, CDCl3) δ 3.12 (d, J ) 11.4 Hz, 1H), 3.30 (d, J )
11.4 Hz, 1H), 3.76 (br, 1H), 3.60 (br, 1H), 5.35 (d, J ) 9.7 Hz,
1H), 5.68-5.83 (m, 1H), 5.98-6.04 (m, 1H), 6.95 (d, J ) 7.5,
2H), 7.31-7.44 (m, 3H).
4-P h e n yl-1,3,6,8-t e t r a ch lor op e n t a cyclo[6,6,0,0,3,60,
1
4,50,4,1209,14]tetr a d eca -10,12-d ien e-2,7-d ion e (11c): H NMR
(500 MHz, CDCl3) δ 2.98 (d, J ) 11.5 Hz, 1H), 3.14 (d, J )
11.5 Hz, 1H), 3.93 (br, 1H), 3.65 (br, 1H), 5.14 (d, J ) 9.7 Hz,
1H), 5.68-5.83 (m, 1H), 5.98-6.04 (m, 1H), 7.15 (br, 2H),
7.31-7.44 (m, 3H).
2,3-Dih yd r o-2,2-d ip h en yl-4,6,7-t r ich lor ob en zofu r a n -
3,5-d iol (14). Recrystallization from chloroform gave colorless

40 J . Org. Chem., Vol. 65, No. 1, 2000
Xue et al.
1
needles: mp 146-147 °C; H NMR (300 MHz, CDCl3) δ 5.89
14-Meth yl-4-(4-m eth ylph en yl)-1,3,6,8-tetr ach lor open ta-
cyclo[6,6,0,0,3,60,4,50,4,1109,14]tetr adeca-10,12-dien e-2,7-dion e
(10d ): 1H NMR (500 MHz, CDCl3) δ 1.69 (s, CH3), 2.33 (s, CH3),
3.09 (d, J ) 11.2 Hz, 1H), 3.26 (d, J ) 11.2 Hz, 1H), 3.72 (br,
1H), 5.32 (d, J ) 9.7 Hz, 1H), 5.33 (br, 1H), 5.86 (d, J ) 9.7
Hz, 1H), 6.83 (d, J ) 7.8 Hz, 2H), 7.16 (d, J ) 7.8 Hz, 2H).
9-Meth yl-4-(4-m eth ylp h en yl)-1,3,6,8-tetr a ch lor op en ta -
cyclo[6,6,0,0,3,60,4,50,4,1209,14]tetr adeca-10,12-dien e-2,7-dion e
(11d ). This isomer 11d could be isolated from the mixture of
9d -11d by stepwise crystallization from chloroform-ether to
give a white solid: mp 208-210 °C; 1H NMR (500 MHz, CDCl3)
δ 1.71 (s, CH3), 2.36 (s, CH3), 3.13 (d, J ) 11.2 Hz, 1H), 3.44
(d, J ) 11.2 Hz, 1H), 3.76 (br, 1H), 5.36 (br, 1H), 5.22 (d, J )
9.8 Hz, 1H), 5.90 (d, J ) 9.8 Hz, 1H), 6.89 (d, J ) 8.0 Hz, 2H),
7.21 (d, J ) 8.0 Hz, 2H); IR (KBr) 1749, 1612, 820 cm-1; MS
(s, 1H), 7.33-7.43 (m, 6H), 7.55-7.64 (m, 4H); IR (KBr) 3350,
1600, 890, 760, 705 cm-1; MS (EI) m/z 406 (M+, 100), 239 (45),
167 (100). Anal. Calcd for C20H13Cl3O3: C, 58.90; H, 3.19.
Found: C, 59.02; H, 3.16.
Ir r a d ia tion of CA w ith 1d . A solution of CA (1476 mg, 6
mmol) and 1,1-bis(4-methylphenyl)ethylene (2496 mg, 12
mmol) in benzene (120 mL) was photolyzed for 40 h to lead to
a 84.4% conversion of CA and afforded 5d (400 mg, 19%), 6d
(130 mg, 6.7%), 7d (35 mg, 1.7%), 12 (415 mg, 18%), 13 (230
mg, 9.6%), and a mixture of 9d , 10d , and 11d (350 mg, 15%)
in a ratio of 1:3:13 as estimated from the 1H NMR spectrum
of the mixture.
2,3,5-Tr ich lor o-6-(2,2-bis(4-m eth ylp h en yl)eth en -1-yl)-
cycloh exa -2,5-d ien e-1,4-d ion e (5d ). Recrystallization from
petroleum ether (bp 60-90 °C)-chloroform gave red needles:
mp 167-169 °C; 1H NMR (60 MHz, CDCl3) δ 1.95 (s, CH3),
2.00 (s, CH3), 6.49 (s, 1H), 6.90-7.40 (m, 8H); IR (KBr) 1681,
1671, 1611, 1593, 825 cm-1; MS (EI) m/z 416 (M+, 11), 57 (100).
Anal. Calcd for C22H15Cl3O2: C, 63.23; H, 3.59. Found: C,
63.26; H, 3.62
2,3-Dich lor o-6-m e t h yl-9-(4-m e t h ylp h e n yl)p h e n a n -
th r en e-1,4-d ion e (6d ). Recrystallization from benzene gave
red needles: mp 264-266 °C; 1H NMR (60 MHz, CDCl3) δ 2.49
(s, CH3), 2.62 (s, CH3), 7.20-8.10 (m, 6H), 8.13 (s, 1H), 9.42
(m, 1H); IR (KBr) 1672, 1615, 1594, 851 cm-1; MS (EI) m/z
380 (M+, 100). Anal. Calcd for C22H14Cl2O2: C, 69.31; H, 3.70.
Found: C, 69.44; H, 3.92.
(EI) m/z 452 (M+, 1.2), 418 (100). HRMS (m/z) calcd for C22H15
Cl3O2 (M+ - HCl) 416.0139, found 416.0080 (mmu 5.8).
-
2,3-Bis(4-m eth ylp h en yl)-4,6,7-tr ich lor oben zofu r a n -5-
ol (12). Recrystallization from petroleum ether (bp 60-90 °C)-
benzene gave colorless needles: mp 245-247 °C; 1H NMR (500
MHz, CDCl3) δ 2.31 (s, CH3), 2.45 (s, CH3), 5.76 (s, OH), 7.08
(d, J ) 8.1, 2H), 7.25 (d, J ) 8.1, 2H), 7.28 (d, J ) 8.1 Hz, 2H),
7.41 (d, J ) 8.1 Hz, 2H); 13C NMR (125 MHz, CDCl3) δ 153.7,
145.0, 144.4, 139.1, 138.1, 130.8, 129.6, 129.3, 129.2, 128.5,
126.7, 117.1, 116.7, 114.8, 109.9, 21.5, 21.4; IR (KBr) 3500,
1610, 820 cm-1; MS (EI) m/z 416 (M+, 100). Anal. Calcd for
C
22H15Cl3O2: C, 63.23; H, 3.59. Found: C, 63.46; H, 3.64.
Ack n ow led gm en t. This project was supported by
2,3-Dich lor o-6-m eth yl-9-(4-m eth ylp h en yl)-10-h yd r oxy-
p h en a n th r en e-1,4-d ion e (7d ). Recrystallization from petro-
leum ether (bp 60-90 °C)-ethyl acetate gave red needles: mp
264 °C; 1H NMR (500 MHz, CDCl3) δ 2.49 (s, CH3), 2.61 (s,
CH3), 6.14 (s, OH), 7.14 (d, J ) 6.8, 2H), 7.39 (d, J ) 6.8, 2H),
7.49 (d, J ) 8.0 Hz, 1H), 8.22 (d, J ) 8.0 Hz, 1H), 9.31 (s, 1H);
IR (KBr) 3495, 1680, 1658, 1608, 1560, 830 cm-1; MS (EI) m/z
396 (M+, 100), 381 (69). Anal. Calcd for C22H14Cl2O3: C, 66.50;
H, 3.53. Found: C, 66.30; H, 3.56.
11-Meth yl-4-(4-m eth ylph en yl)-1,3,6,8-tetr ach lor open ta-
cyclo[6,6,0,0,3,60,4,50,4,1409,14]tetr adeca-10,12-dien e-2,7-dion e
(9d ): 1H NMR (500 MHz, CDCl3) δ 1.67 (s, CH3), 2.63 (s, CH3),
2.93 (d, J ) 11.2 Hz, 1H), 3.10 (d, J ) 11.2 Hz, 1H), 3.87 (br,
1H), 5.41 (br, 1H), 5.12 (d, J ) 9.7 Hz, 1H), 5.84 (d, J ) 9.7
Hz, 1H), 7.15-7.18 (m, 2H), 7.04 (br, 2H).
the National Natural Science Foundation of China
(29772016) and the Natural Science Foundation of
J iangsu Province (BK97017). The molecular mechanics
calculations are partly supported by project 211 in the
Department of Chemistry at Nanjing University.
Su p p or tin g In for m a tion Ava ila ble: Results of molecular
mechanics calculations on compounds 9b, 10b and 11b (total
strain energy and energy decomposition) and copies of 1H NMR
for compounds 9a -c and the mixture of compounds 15 and
16. This material is available free of charge via the Internet
at http://pubs.acs.org.
J O990831R