Photodimerization of 1-phenylcyclohexene. A novel transient-transient component

Article abstract of DOI:10.1021/ja9533860

Low-temperature irradiations of 1-phenylcyclohexene (PC6) yield two singlet state 2 + 2 photodimers and two 4 + 2 cycloadducts. The 4 + 2 adducts are not observed in low intensity lamp irradiations at room temperature but are produced in significant quant

Full text of DOI:10.1021/ja9533860

1682
J. Am. Chem. Soc. 1996, 118, 1682-1689
Photodimerization of 1-Phenylcyclohexene. A Novel
Transient-Transient Component
David J. Unett, Richard A. Caldwell,* and Duane C. Hrncir
Contribution from the Department of Chemistry, The UniVersity of Texas at Dallas,
Richardson, Texas 75083
ReceiVed October 5, 1995X
Abstract: Low-temperature irradiations of 1-phenylcyclohexene (PC6) yield two singlet state 2 + 2 photodimers
and two 4 + 2 cycloadducts. The 4 + 2 adducts are not observed in low intensity lamp irradiations at room temperature
but are produced in significant quantities when PC6 solutions are subjected to higher intensity laser irradiation. The
results indicate a reaction mechanism involving two trans-PC6 molecules. The findings are consistent with earlier
kinetic observations that the decay of trans-PC6 adopts a significant second-order component at lower temperatures.
The transient-transient reaction also occurs under triplet-sensitized low-temperature/low-intensity or room temperature/
high-intensity irradiation conditions. Triplet-sensitized irradiations also yield significant amounts of the 2 + 2
photodimers produced by attack of the PC6 triplet on its ground state. Quantum yield studies reveal that the rate of
addition of 3PC6* to its ground state is in the range 0.28 to 4.0 × 105 L mol-1 s-1 consistent with the 1,2-biradical
model for alkene triplet reactivity.
Photosensitized cis-trans isomerization of alkenes represents
particularly those where ring formation occurs with the produc-
tion of trans stereochemistry at the original location of the
cyclohexene double bond.
probably the most thoroughly documented organic photochemi-
cal reaction. The process is not confined to acyclic alkenes.
Irradiation of even small ring cycloalkenes, either direct or
sensitized, leads to the production of trans isomers which
become increasingly strained as the ring size is decreased.1
While trans-cyclooctenes have been demonstrated to be stable
at ambient temperatures,2,3 trans-cycloheptenes and cyclo-
hexenes are transient species with lifetimes for the latter
generally in the microsecond time regime.4-9 The high degree
of strain within a trans-cyclohexene structure represents a strong
potential driving force for novel chemical reactivity and trans-
cyclohexenes and cyclohexenones have been discussed as
possible intermediates in numerous chemical reactions,8,10-26
trans-Phenylcyclohexene (trans-PC6) is by far the most
widely studied example of a trans-cyclohexene in the literature
to date. In particular, the phenyl substituent on the double bond
allows for laser excitation of the cis isomer and spectroscopic
detection of the trans isomer in flash photolysis studies. trans-
PC6 was first characterized by Bonneau et al.,4 who reported
an absorption maximum around 380 nm and a transient lifetime
of 9 µs in methanol at room temperature. The authors also
reported that the species was quenched by the hydrogen ion
with a rate of 7.6 × 106 L mol-1 s-1. Photoacoustic studies of
the triplet-sensitized production of trans-PC67 reveal a lifetime
for the PC6 triplet of 65 ns and a relaxed triplet energy of 56.4
( 0.7 kcal mol-1 compared to the spectroscopic triplet energy
of 60.8 kcal mol-1. The quantum yield for the production of
trans-PC6 via this route is 0.36 and the energy of the trans
X Abstract published in AdVance ACS Abstracts, February 1, 1996.
(1) Strickland, A. D.; Caldwell, R. A. J. Phys. Chem. 1993, 97, 13394-
13402.
(2) Wallraff, G. M.; Michl, J. J. Org. Chem. 1986, 51, 1794-1800.
(3) Cope, A. C.; Kinnel, R. B. J. Am. Chem. Soc. 1966, 88, 752-761.
(4) Bonneau, R.; Joussot-Dubien, J.; Salem, L.; Yarwood, A. J. J. Am.
Chem. Soc. 1976, 98, 4329.
(5) Bonneau, R.; Herran, B. Laser Chem. 1984, 4, 151.
(6) Bonneau, R. J. Photochem. 1987, 36, 311-21.
(7) Goodman, J. L.; Peters, K. S.; Misawa, H.; Caldwell, R. A. J. Am.
Chem. Soc. 1986, 108, 6803-5.
(8) Saltiel, J.; Marchand, G. R.; Bonneau, R. J. Photochem. 1985, 28,
367-72.
(9) Caldwell, R. A.; Misawa, H.; Healy, E. F.; Dewar, M. J. S. J. Am.
Chem. Soc. 1987, 109, 6869-70.
(10) Evers, J. T. M.; Mackor, A. Tetrahedron Lett. 1978, 2321-2324.
(11) Evers, J. T. M.; Mackor, A. Tetrahedron Lett. 1978, 2317-2320.
(12) Evers, J. T. M.; Mackor, A. Tetrahedron Lett. 1980, 21, 415-418.
(13) Fujita, S.; Nomi, T.; Nozaki, H. Tetrahedron Lett. 1969, 3557.
(14) Fujita, S.; Hayashi, Y.; Nomi, T.; Nozaki, T. Tetrahedron 1971,
27, 1607.
isomer was found to be 44.7 ( 5 or 47.0 ( 3 kcal mol-1 27
,
relative to the ground state cis isomer. Varible-temperature
studies9 reveal an activation energy for trans-cis isomerization
of 12.1 ( 0.12 kcal mol-1. The isomerization transition state
is therefore 59.1 ( 3 kcal mol-1 above the cis isomer,
demonstrating a close approach of the ground and excited triplet
state potential energy surfaces at an approximately perpendicular
geometry.
The first stereochemical evidence for the existence of trans-
cyclohexenes was provided by Dauben et al. in a study of PC6
photochemistry.28,29 Prolonged irradiation of PC6 in methanol
generates the singlet state 2 + 2 photodimers (1 and 2) and a
(15) Kropp, P. J. J. Am. Chem. Soc. 1969, 91, 5783.
(16) Luef, W.; Keese, R. Top. Stereochem. 1991, 20, 231-318.
(17) Marshall, J. A. Acc. Chem. Res. 1969, 2, 33.
(18) Miyamoto, N.; Isiyama, S.; Utimoto, K.; Nozaki, H. Tetrahedron
1973, 29, 2365-2371.
(19) Salomon, R. G.; Folting, K.; Streib, W. E.; Kochi, J. K. J. Am.
Chem. Soc. 1974, 96, 1145-1152.
(20) Schuster, D. I. In The Chemistry of Enones; Patai, S., Rappoport,
Z., Eds.; John Wiley and Sons Ltd.: New York, 1989; pp 623-756.
(21) Spee, T.; Mackor, A. J. Am. Chem. Soc. 1981, 103, 6901.
(22) Venkatraman, S. M. S. Thesis, University of Texas at Dallas, 1987.
(23) Warner, P. M. Chem. ReV. 1989, 89, 1067-1093.
(24) Dunkelblum, E.; Hart, H.; Jeffares, M. J. Org. Chem. 1978, 43,
3409.
(25) Mintas, M.; Schuster, D. I.; Williard, P. G. J. Am. Chem. Soc. 1988,
110, 2305.
(26) Mintas, M.; Schuster, D. I.; Williard, P. G. Tetrahedron 1988, 44,
6001.
(27) Ni, T.; Caldwell, R. A.; Melton, L. A. J. Am. Chem. Soc. 1989,
111, 457-64.
(28) Dauben, W. G.; van Riel, H. C. H. A.; Hauw, C.; Leroy, F.; Joussot-
Dubien, J.; Bonneau, R. J. Am. Chem. Soc. 1979, 101, 1901.
(29) Dauben, W. G.; Riel, H. C. H. A. v.; Robbins, J. D.; Wagner, G. J.
J. Am. Chem. Soc. 1979, 101, 6383.
0002-7863/96/1518-1682$12.00/0 © 1996 American Chemical Society

Photodimerization of 1-Phenylcyclohexene
J. Am. Chem. Soc., Vol. 118, No. 7, 1996 1683
distributions are GC peak areas normalized to 100% for the four major
products under investigation unless otherwise stated.
Chart 1
Laser flash photolysis experiments were performed as previously
described30 with a Q-switched Continuum model YG671C-10 Nd:YAG
laser, pulse width fwhm ) 5 ns after the doubling crystal, using a
repetition rate of 10 Hz. An Osram model XBO 150 W/I high-pressure
xenon lamp and CVI Digikrom 240 monochromator and photomultiplier
provided detection. Data were collected using a Tektronix DSA 602
digitizing signal analyzer and transferred to a DTK 486DX personal
computer for analysis. Laser and flash lamp control and kinetic analyses
were performed with the PC RAD and KS-01 software packages from
Kinetic Instruments (P.O. Box 49434, Austin, TX 78765).
Preparative irradiations were performed in a standard quartz
immersion well photochemical reactor with a Hanovia 679A 450W
medium-pressure mercury arc lamp. A Pyrex filter was added for
triplet-sensitized irradiations. Low-temperature irradiations were per-
formed at -78 °C and required the use of a triple-walled low-
temperature immersion well. The 2 + 2 photodimers were readily
prepared by room temperature irradiation of a 20 mM solution of PC6
in either cyclohexane or methanol. Use of cyclohexane avoids the
unwanted trans-PC6-methanol addition side reaction forming 1-meth-
oxy-1-phenylcyclohexane. The reaction was monitored by GC using
a tetradecane internal standard. Irradiations were generally run to >90%
conversion of starting material. The two 2 + 2 dimers were produced
in a 3:1 ratio determined by GC analysis of the final reaction mixture.
The two 4 + 2 adducts were isolated from a low-temperature irradiation
of PC6 (20 mM) in methanol. The final product distribution was 1
(20%), 2 (16.5%), 4 (26.5%), 5 (37%). The reaction was accompanied
by significant quantities of 1-methoxy-1-phenylcyclohexane (3). The
exact yield of this product was sensitive to the acidity of the solvent.
The GC peak areas of the four PC6 dimers, the methanol addition
product (3), and unreacted PC6 accounted for >90% of the total peak
area.
Isolation and Characterization of Photoproducts. Solvent was
removed from the irradiated mixtures which were then subjected to
repeated column chromatography on silica with hexane as the eluting
solvent. The Rf values for the 2 + 2 dimers were close and large
quantities of mixed fractions had to be discarded in order to obtain
enriched samples of each isomer. Final purification was effected by
multiple recrystallizations from hexane/ethyl acetate mixtures. The 4
+ 2 dimers were more difficult to isolate. Separation of 4 and 5 proved
impossible on silica gel. Little resolution was achieved even under
analytical HPLC conditions. A mixture of the two 4 + 2 adducts was
obtained free of other contaminants via column chromatography.
Recrystallizations employing a range of solvents eventually yielded a
pure sample of 4. Dimer 5 has a significantly lower melting point and
was eventually isolated, through numerous recrystallization procedures,
from the combined mother liquors from the 4 recrystallizations. This
process eventually yielded a crystal free from contamination by 4 and
suitable for X-ray analysis.
Melting points and 13C NMR data for the 2 + 2 dimers were identical
to literature values.29 Dimer 4 was recrystallized many times and an
eventual melting point of 207-208.5 °C was obtained. This is higher
than the previously reported value28,29 (198-201 °C) but was reproduc-
ible. Single-crystal X-ray analysis of this sample, as described below,
revealed unit cell dimensions identical to those reported by Dauben et
al.28 Dimer 5 has a significantly lower melting point at 131-3 °C. A
full single-crystal X-ray analysis was performed on this sample.
Laser synthesis of PC6 photodimers was conducted in a 5 × 10
mm quartz cuvette placed at the exit of the Nd:YAG laser described
above. The laser output was controlled by variations in the oscillator
and/or amplifier flash lamp voltages. Irradiations were performed at
266 nm (10 Hz) for direct excitation of PC6 or 355 nm for sensitized
experiments. The laser beam impinged on the 10 mm face of the
cuvette and the irradiated sample was subjected to a continuous nitrogen
sparge. The laser power was monitored periodically during irradiations,
which generally took 30 min to 1 h to perform. The diameter of the
laser beam for all irradiations was 10 mm. Product distributions were
again monitored by gas chromatography.
methanol addition product (3) resulting from protonation of the
trans-PC6 isomer either by the solvent or traces of acid. The
yield of this product is increased upon addition of sulfuric acid
to the reaction mixtures. In addition, the authors were able to
isolate and characterize a new dimer (4) resulting from a 4 +
2 cycloaddition reaction. The product retained a trans ring
junction at the site of one of the original cyclohexene double
bonds and was assigned to the Woodward-Hoffmann allowed
concerted [π4s + π2s] addition of ground state cis-PC6 across
the strained double bond of a trans-PC6 molecule. Rearoma-
tization via a 1,3-hydrogen shift affords the isolated product.
The new 4 + 2 adduct was isolated only from irradiations
conducted at low temperatures in which the lifetime of the trans-
PC6 was expected to be dramatically lengthened.
During our Arrhenius studies9 of trans-PC6 isomerization,
we noted that the transient decay of the trans isomer adopted a
significant second-order component even at temperatures as high
as 10 °C and speculated that we might be observing a reaction
in which the 4 + 2 adduct was generated. A second-order
component in the transient decay obviously suggested a process
involving reaction between two trans isomers and not trans +
cis as previously assumed. We have therefore conducted a
detailed investigation of the origin of the unusual 4 + 2 adduct
and find that it is indeed generated, along with a second,
previously unreported stereoisomer (5), in a reaction involving
two trans-cyclohexene molecules. In addition, we have been
able to study the triplet-sensitized photolysis of PC6 in greater
detail and find that the reported 2 + 2 photodimers (1 and 2)
are also produced via the photoaddition of the excited PC6 triplet
to a ground state PC6 molecule. This reaction is discussed in
the context of the 1,2-biradical model for alkene triplets27,30-32
with reference to a number of other recently published results
reported by our group.
Experimental Section
General. 1-Phenylcyclohexene was obtained from Aldrich and
purified by column chromatography (silica/hexane) and kugelrohr
distillation prior to use. Spectroscopic grade solvents were used as
received. Sensitizers were purified previously in our laboratories by
recrystallization. 13C NMR spectra were run on a JOEL-FX200 (200
MHz) for solutions in CDCl3 containing tetramethylsilane as internal
standard. Gas chromatography was performed on a Shimadzu GC-
14A gas chromatograph using a Restek Rtx-20 (crossbonded 80%
dimethyl 20% diphenyl polysiloxane) megabore capillary column (0.53
mm i.d. × 30 m) with flame ionization detection. Reported product
(30) Caldwell, R. A. In Kinetics and Spectroscopy of Carbenes and
Biradicals; Platz, M. S., Ed.; Plenum Publishing Corp.: New York, 1990;
pp 77-116.
(31) Caldwell, R. A.; D´ıaz, J. F.; Hrncir, D. C.; Unett, D. J. J. Am. Chem.
Soc. 1994, 116, 8138-8145.
For variable-temperature work, the 5 × 10 mm cuvette was replaced
by a 10 × 10 mm jacketed quartz cuvette. The sample path length of
the cuvette was approximately 4 mm. Heating and cooling of the
(32) Caldwell, R. A.; Zhou, L. J. Am. Chem. Soc. 1994, 116, 2271-
2275.

1684 J. Am. Chem. Soc., Vol. 118, No. 7, 1996
Unett et al.
Table 1. Summary of the Crystallographic Data for the Structure
of Dimer 5
compd
mol wt
C24 H28
316.49
space group
P1 (No. 2)
cell constants
a, Å
b, Å
c, Å
10.070(2)
14.180(4)
6.715(2)
94.35(3)
107.93(2)
75.41(2)
887.6(7.9)
2
1.184
Mo KR (0.71073)
3094
328
0.0467
R, deg
â, deg
γ, deg
cell vol, Å3
Z
density (calcd), g/cm3
radiation (λ, Å)
no. of reflcns obs
no. of params refined
R
Rw
0.0463
sample was effected by continuous circulation of water through the
cell jacket from a thermostatically controlled bath. The temperature
of the sample was monitored directly using a thermocouple.
Figure 1. Thermal ellipsoid plot of dimer 5, showing the atom-labeling
scheme. The trans configuration across the original double bond of
the dienophile (C1C-C6C) is retained.
X-ray Structure Determination of the Second 4 + 2 Adduct (5).
Preliminary examination and data collection were accomplished using
an Enraf-Nonius CAD4 diffractometer. Crystallographic and data
collection parameters are given in Table 1. The space group of the
cell was determined to be P1 (No. 2). The structure was solved by
direct methods and refined on the basis of 3094 observed reflections.
Hydrogen atoms were calculated at idealized positions and included
in the calculations but not refined. Least-squares refinement based upon
3094 reflections converged at R ) 0.0467 and Rw ) 0.0463.
(4) which has been shown to arise from chemistry involving
the trans isomer of PC6. Melting points and 13C NMR data
for 1 and 2 were identical to those previously reported. The
melting point of 4 was slightly higher (Vide infra) than expected
but was sharp and reproducible. The unit cell dimensions of a
crystalline sample of 4 determined by a preliminary X-ray
analysis were found to be identical to literature values.29
In addition we were able to isolate and characterize a second
4 + 2 adduct (5), a diastereomer of 4. This new dimer is
generally produced in slightly higher yields than 4 but proved
much more difficult to isolate in pure form. Repeated recrys-
tallization of mixtures of 4 and 5 isolated by column chroma-
tography yielded relatively pure samples of 4. Careful GC
analysis revealed traces of 5 in these crystals and multiple
recrystallizations were required to produce a pure sample. This
may account for the slightly higher melting point for 4 reported
here. Dimer 5 possesses a melting point some 76 degrees below
its stereoisomer and tends to co-crystallize with dimer 4. These
dimers were also inseparable by HPLC. Exhaustive recrystal-
lization efforts did eventually provide a pure sample, the
structure of which was obtained crystallographically (Figure 1).
Low-Intensity (Lamp) Irradiations. The product distribu-
tion upon irradiation of PC6 in solution is known to be
temperature dependent, favoring the formation of 4 + 2 adducts
at lower temperatures. In our hands, irradiation of a 0.1 M
solution of PC6 in either methanol or cyclohexane at room
temperature yielded exclusively the 2 + 2 singlet state photo-
adducts 1 and 2. An upper limit for the formation of the 4 +
2 products is set at 0.1% based on the sensitivity of our GC
analysis conditions and the ratio of 2 + 2 products 1:2 was
3:1. Irradiations in methanol also yielded significant quantities
of the trans-PC6 methanol addition product 1-methoxy-1-
phenylcyclohexane.
Quantum yields for the triplet-sensitized production of dimer 1 were
determined by comparison with the rate of isomerization of trans-â-
methylstyrene, which is known to have a trans-cis isomerization
quantum yield of 0.5.33 Dimer 1 was chosen over dimer 2 as its yield
is greater by a factor of 5 under these conditions. Irradiations were
performed in 10-mm jacketed quartz cuvettes arranged to be equidistant
from the light source, a medium pressure mercury arc lamp. The cells
were constructed with a path length of approximately 4 mm while the
path length of the jacket is approximately 2 mm. Solutions of the two
alkenes of identical concentration (either 0.1 or 1.0 M) were made up
in the same stock solution of benzophenone (OD366 ) 1.0) in methanol
containing a known concentration of tetradecane as an internal standard.
The cell jackets were filled with 1.0 M solutions of the appropriate
alkene in methanol. The lamp was allowed to warm up for 30 min
before irradiation of the samples to ensure a uniform irradiation intensity
for the duration of the experiment. Samples were subjected to a
continuous gentle nitrogen sparge and sampled at time intervals
appropriate to the rate of product formation. Product concentrations
were determined relative to the internal standard following determi-
nation of their relative GC detector response factors. Plots of
concentration versus time give the relative quantum yields for product
formation. The reaction was performed in the absence and presence
of sulfuric acid (50 mM).
Results
Preparation and Characterization of Photodimers. The
2 + 2 photodimers 1 and 2 were isolated from room temperature
irradiations of PC6 in either methanol or cyclohexane. The 4
+ 2 photodimers were isolated and characterized from a reaction
performed in a low-temperature quartz irradiation vessel at -78
°C. The reaction solvent was methanol. In our hands, the low-
temperature reaction yielded four major photodimers. Two of
these (1 and 2) were the 2 + 2 dimers resulting from the excited
singlet state photocycloaddition reaction reported by Dauben
et al.29 We were also able to isolate the reported 4 + 2 adduct
Lamp irradiation of PC6 in methanol was also performed at
-78 °C. The concentration of the starting material was lowered
somewhat to 0.02 M due to solubility problems at lower
temperatures. The expected change in product distribution was
observed with 4 + 2 adducts accounting for 63.5% of the total
dimer yield. The relative product yields for all the irradiations
performed here, determined by GC, are given in Table 2.
The irradiation was performed under triplet-sensitized condi-
tions. Irradiation, at room temperature, of a 0.1 M cyclohexane
(33) Caldwell, R. A.; Sovocool, G. W.; Peresie, R. J. J. Am. Chem. Soc.
1973, 95, 1490.

Photodimerization of 1-Phenylcyclohexene
J. Am. Chem. Soc., Vol. 118, No. 7, 1996 1685
Table 2. Experimental Conditions, PC6 Concentrations, and
Normalized Product Distributions for Irradiations of PC6a
irradiation conditions
[1-PC6] (M) % 1 % 2 % 4
% 5
<0.1 <0.1
20 16.5 26.5 37
RT,b lamp
0.1
76 24
LT,c lamp
0.02
0.1
0.1
0.02
0.1
RT, laser (266 nm)
RT, sens,d lamp
LT, sens, lamp
RT, sens, laser (355 nm)
60 25
80 20
7
8
<0.5 <0.5
43
36
2
2
53
32
23 10
a For full experimental conditions, refer to text. b Room temperature.
c Low temperature (-78 °C). d Triplet sensitized.
solution of PC6 containing p-methoxyacetophenone (OD366
)
0.3), was performed through a Pyrex filter. The reaction,
somewhat surprisingly, yielded large quantities of 2 + 2
photoadducts with a 1:2 ratio of 4:1 (Table 2, entry 4). The 4
+ 2 adducts were not detected in the reaction mixture and a
reasonable upper limit for their formation would be 0.5% for
the particular GC analysis conditions employed here.
The sensitized reaction was also studied at -78 °C. The PC6
concentration was again lowered to 0.02 M and the reaction
was run in toluene due to additional solubility problems
encountered with the sensitizer at low temperatures. The effect
of temperature upon the triplet-sensitized reaction is dramatic
and 4 + 2 adducts are almost the exclusive products. In this
respect, low-temperature sensitized irradiations appear to be the
preferred method for the synthesis of 4 + 2 photoadducts. The
2 + 2 dimers were each present at the 2% level. The observed
4:5 ratio of 1:1.23 was essentially within experimental error of
the value 1:1.40 noted for direct irradiation at low temperatures.
High-Intensity (Laser) Irradiations. Irradiation of a 0.1
M solution of PC6 in cyclohexane with the 266-nm output from
a Nd:YAG laser (8 mJ/pulse), operated at a repetition rate of
10 Hz, for 15 min, led to the production of easily detectable
concentrations of photodimers. Although, under these condi-
tions, 2 + 2 photodimers are the dominant product, significant
quantities of the 4 + 2 addition products were present.
The triplet-sensitized reaction was performed under laser
irradiation conditions. Laser irradiation (10 mJ/pulse, 15 min,
355 nm) of a solution of PC6 (0.1 M) in cyclohexane containing
p-methoxyacetophenone (OD355 ) 1.0) produced large quantities
of 4 + 2 dimers (Table 2, entry 6). The 2 + 2 dimers were
also present as a significant minor component. We had initially
been concerned that the formation of 2 + 2 adducts in the triplet-
sensitized lamp irradiation might be due to incomplete filtering
of the shorter wavelength emissions from the mercury lamp.
Clearly, under 355 nm laser irradiation, these dimers do not
result from direct photon absorption by PC6. Having established
that, in addition to their appearance in low-temperature lamp
irradiations, 4 + 2 adducts are present as a significant component
of reaction mixtures resulting from higher intensity laser
irradiations, we sought to illustrate this variation in the product
distribution more clearly. Irradiations of 0.1 M solutions of
PC6 in cyclohexane were performed with the 266-nm output
from a Nd:YAG laser in the 2-30 mJ/pulse range. We observe
a smooth change in the product distribution upon changes in
laser power (Figure 2) with the formation of 4 + 2 adducts
favored at higher irradiation intensities. At 2 mJ/pulse the
concentrations of 4 + 2 dimers dropped below the sensitivity
of the GC analysis (<0.1% of the dimer mixture).
Figure 2. Plot of normalized total dimer percentages as a function of
incident laser power for 266 nm laser irradiation of a solution of PC6
(0.1 M) in cyclohexane. Key: 2 + 2 dimers (b), 4 + 2 dimers (O).
Figure 3. Plot of normalized total dimer percentages as a function of
incident laser power for 355 nm laser irradiation of a solution of PC6
(0.2 M) in benzene containing benzophenone (OD355 ) 1.5). Key: 2
+ 2 dimers (b), 4 + 2 dimers (O).
The reaction temperature was also varied under 266 nm direct
laser irradiation conditions (8 mJ/pulse, 15 min) between 5 and
50 °C for a solution of PC6 (0.1 M) in cyclohexane. The
expected change in product distribution to favor the 4 + 2
dimers at lower temperatures was observed (Figure 4).
Triplet-Sensitized Photoadditions. The appearance of 2 +
2 photoadducts during triplet-sensitized irradiations might at first
seem surprising. Dauben et al.29 reported trace quantities of
these products in their sensitized experiments but their presence
was simply assigned to a possible residual absorbance by PC6
at longer wavelengths. We had initially considered that
incomplete screening of the shorter-wavelength emissions from
the mercury lamp used in our experiments might be a factor.
However, sensitized experiments performed under 355 nm laser
irradiation also yield 2 + 2 adducts. It is highly improbable
A power dependence study was also performed for the
sensitized reaction (benzophenone OD355 ) 1.5, PC6 (0.2 M),
laser power 5-20 mJ/pulse, in benzene). The product distribu-
tion (Figure 3) again shows a smooth change to favor the
formation of 4 + 2 adducts at higher irradiation intensities. GC
analysis suggested photoreduction of PC6 in this system was
negligible.

1686 J. Am. Chem. Soc., Vol. 118, No. 7, 1996
Unett et al.
a
b
Figure 4. Plot of normalized total dimer percentages as a function of
temperature for 266 nm laser irradiation of PC6 (0.1 M) in cyclohexane.
Key: 2 + 2 dimers (b), 4 + 2 dimers (O).
that there is any residual absorption by PC6 at such long
wavelengths.
The formation of cyclobutanes via a triplet state mechanism
is not so surprising in the light of recent work published by
Caldwell et al.31 in which the photoaddition of p-acetylstyrene
triplets to styrene was documented. In addition, the change in
the 1:2 ratio from 3:1 for direct irradiations to 4:1 for the triplet-
sensitized reaction may be indicative of a different reaction
mechanism under sensitized conditions. With this in mind,
further investigation of the mechanism of this reaction was
warranted.
In order to further rule out the possibility of residual long-
wavelength absorptions by PC6, an irradiation was performed
using a PC6 filter solution. Irradiation of a 0.1 M solution of
PC6 in cyclohexane containing p-MAP (OD366 ) 1.0) was
performed in a jacketed quartz cuvette. The lamp emissions
were filtered through Pyrex and the cell jacket was filled with
a 0.5 M solution of PC6. The effective path length of the cell
jacket was approximately 2 mm. Good yields of 2 + 2 adducts
were again obtained.
Having established that 2 + 2 dimers are indeed produced
when the photolysis is triplet sensitized, one must now consider
the mechanism of this new reaction. 2 + 2 dimer formation
could, in principle, result from either a triplet state addition
reaction or possibly the addition of trans-PC6 to a ground state
cis-PC6 molecule. Since the lifetime of trans-PC6 is known
to be extremely sensitive to the presence of acids,4,29 a simple
test for the intermediacy of this species is to measure the
quantum yield for dimer formation in both the presence and
absence of acid.
trans-PC6 is quenched by sulfuric acid with a rate constant
of 9 × 106 L mol-1 s-1. The lifetime of trans-PC6 at room
temperature in methanol is 9 µs.4 The addition of 50 mM
sulfuric acid to a sensitized photolysis reaction should therefore
result in the interception of 78% of any trans-PC6 molecules
formed and the quantum yield for any process involving this
intermediate should decrease accordingly.
The quantum yield for 2 + 2 dimer formation was determined
by comparison with the rate of isomerization of trans-â-
methylstyrene.33 Solutions of the two alkenes in methanol
containing benzophenone (OD366 ) 1.0) were irradiated simul-
Figure 5. (a) Concentration of cis-â-methylstyrene as a function of
time upon irradiation of trans-â-methylstyrene (1.0 M) in methanol
containing benzophenone (OD366 ) 1.0). (Curve fit: f(x) ) (2.23 ×
10-3)x + 1.89 × 10-4; R2 ) 0.9978.) (b) Concentration of dimer 1 as
a function of time upon irradiation of PC6 (1.0 M) in methanol
containing benzophenone (OD366 ) 1.0). (Curve fit: f(x) ) (8.01 ×
10-6)x - 3.35 × 10-5; R2 ) 0.9984.)
taneously. Benzophenone triplet is also quenched by sulfuric
acid with a rate constant of 1.7 × 106 L mol-1 s-1. The alkene
concentration of 1.0 M ensured efficient quenching of the
sensitizer even in the presence of 50 mM acid. The irradiations
were sampled at regular intervals and the concentrations of cis-
â-methylstyrene and 1 obtained by GC analysis against a
tetradecane internal standard. Though not investigated in detail,
solutions appeared to be stable in the dark on the time scale of
our experiments. There was no evidence for styryl radical cation
mediated dimerization.34,35 Plots of product concentration versus
time were constructed (Figure 5a,b). The relative gradients lead
T
to Φ1 , the quantum yield for triplet sensitized formation of 1.
The results (Table 3) clearly demonstrate that Φ1T is independent
of the presence of acid. We therefore conclude that the 2 + 2
dimers are formed via a triplet state addition reaction. The
T
quantum yield, Φ1 , was also determined for a PC6 concentra-
tion of 0.1 M and, as expected for a reaction which is first order
T
in [PC6], the value of Φ1 was lower by a factor of 10.
(34) Schepp, N. P.; Johnson, L. J. J. Am. Chem. Soc. 1994, 116, 10330.
(35) Schepp, N. P.; Johnson, L. J. J. Am. Chem. Soc. 1994, 116, 6895.

Photodimerization of 1-Phenylcyclohexene
J. Am. Chem. Soc., Vol. 118, No. 7, 1996 1687
Table 3. Quantum yields for the triplet-sensitized production of
T
dimer 1 (Φ1 )
T
[PC6]/M
[H2SO4]/mM
φ1
1.0
1.0
0.1
0
50
0
1.80 × 10-3
1.95 × 10-3
2.00 × 10-4
Discussion
The [4 + 2] Dimerization. We have been able to isolate
and characterize two 4 + 2 cycloaddition products from low-
temperature irradiations of phenylcyclohexene in solution. One
of these products was previously unknown and is a diastereoi-
somer of the 4 + 2 adduct originally characterized by Dauben
et al.28,29 Both products appear to arise from cycloadditions
involving transition states in which π-interactions between the
two phenyl groups are maximized. GC analyses of the reaction
mixtures suggest that other possible isomers, if present at all,
are produced in yields not exceeding a few percent of those
observed for the four characterized photodimers.
Figure 6. Transient decay of trans-1-phenylcyclohexene, with a
competing first- and second-order kinetic fit, monitored at 370 nm at
a temperature of -16.2 °C. The second-order fit overlays the raw data
while the first-order component lies below it. trans-Phenylcyclohexene
was generated by 266 nm laser excitation of a heptane solution of the
cis isomer (OD266 ) 5).
The formation of a 4 + 2 cycloaddition product of PC6 (4)
in which one ring junction retains a trans configuration has been
presented as strong evidence for the involvement of trans-PC6
as the dienophile in this reaction. The configuration of the new
4 + 2 adduct (5) is in agreement with this conclusion. The
conformation of the cyclohexene double bond in the PC6 diene
moiety prior to the cycloaddition step is chemically ambiguous
as a result of the 1,3-hydrogen shift, leading to rearomatization,
which is expected to follow the initial cycloaddition event. The
reaction was at first presumed to proceed via the addition of a
trans dienophile to a ground state cis-PC6 diene.28,29 The
possibility that formation of the 4 + 2 adducts might arise from
a 2-photon reaction in which trans-PC6 behaves as both diene
and dienophile was raised in our laboratory during variable-
temperature studies of trans-PC6 isomerization.9 The lifetimes
of trans-cyclohexenes have long been known to be strongly
temperature dependent. At room temperature, the decay of
trans-PC6 is almost cleanly first order, although a small second-
order component is sometimes observed possibly as a result of
high laser intensities and/or higher PC6 concentrations. Its
lifetime is in the 10-µs range depending on the solvent. The
decay of trans-PC6 at lower temperatures was found to include
a significant second-order component (Figure 6) and we
speculated at the time that this might be a manifestation of the
reaction forming the 4 + 2 dimer isolated by Dauben et al.
only at low temperatures.
Scheme 1
The original observation that the product distribution upon
irradiation of PC6 in solution is strongly temperature dependent,
favoring the production of 4 + 2 adducts at lower temperatures,
is clearly consistent with both a “trans + cis” and a “trans +
trans” cycloaddition mechanism. In both cases, lengthening
of the trans-PC6 lifetime at lower temperatures favors any
reaction involving this intermediate.
The crucial observation made in this study is that the product
distribution is also strongly dependent on the intensity of the
irradiating light source and therefore the concentration of
transient intermediates in solution. An intensity-dependent
product distribution is not consistent with a reaction scheme in
which all products are dependent upon the transient concentra-
tion in a first-order manner. One would expect that the quantum
yield for production of trans-PC6 upon irradiation of a solution
of cis-PC6 of fixed concentration at room temperature would
be independent of light intensity. The resulting product
distribution should therefore also remain independent of this
variable. The results presented in Table 2 and Figures 2 and 3
clearly demonstrate that this is not the case.
The four possible triene intermediates which may result from
4 + 2 cycloadditions involving a trans dienophile and both trans
and cis dienes are shown in Scheme 1. The “trans + cis”
adducts (6 and 7) will lead to the two isolated 4 + 2 dimers
following a 1,3-antarafacial hydrogen shift. Though theoreti-
cally predicted, 1,3-antarafacial hydrogen shifts remain experi-
mentally undocumented. In this particular instance, the rigid
carbon framework of the reaction intermediate suggests attain-
ment of the highly strained transition state geometry required
for a concerted 1,3-antarafacial shift would be impossible.
Rearomatization of these intermediates would therefore almost
certainly have to be a non-concerted process. The corresponding
hydrogen shift required for rearomatization of the “trans +
trans” reaction intermediates (8 and 9) to yield the observed
products would be 1,3-suprafacial. While a non-concerted
process may again be involved, the requirement for a 1,3-
suprafacial hydrogen shift also raises the possibility that
rearomatization might occur photochemically, although, to date,
we have not been able to establish this experimentally.
Although observed 4 + 2 dimer yields under direct laser
irradiation conditions are not as high as those observed in the
low-temperature irradiation, the data clearly demonstrate that
significant quantities of these products are available through
high-intensity irradiations. The power dependence experiments
were ultimately limited to powers below 30 mJ cm-2 per pulse.
At higher powers, a solid deposit was observed on the front

1688 J. Am. Chem. Soc., Vol. 118, No. 7, 1996
Unett et al.
Scheme 2
the fraction of 1,4-biradicals which undergoes ring closure to
give the final 2 + 2 adduct cyclobutane, and τ is the PC6 triplet
lifetime, known to be 65 ns at infinite dilution. For high
concentrations of PC6, energy transfer from the sensitizer is
expected to occur with unit efficiency. One can now estimate
T
kr from the measured value of Φ1 by substituting reasonable
values of R. The absolute lower limit for kr in this work is
obtained by substituting R ) 1 and leads to kr ) 2.8 × 104 L
mol-1 s-1
.
We attempted to measure the rate of quenching of the PC6
triplet by its ground state using laser flash photolysis. PC6 was
sensitized by benzophenone (OD355 ) 1.2) in cyclohexane. The
lifetime of the PC6 triplet was slightly longer than the previously
published value at 71 ns and was insensitive to the ground state
concentration up to 1.68 M. Given that we routinely expect to
be able to detect 5% changes in triplet lifetimes, we estimate a
maximum value for kr of 4.0 × 105 L mol-1 s-1 which, using
the above equation, corresponds to R ) 0.07.
Recently published studies27,31,32 have demonstrated that the
thermochemical and kinetic behavior of alkene triplets may be
predicted with good accuracy by consideration of the relaxed
perpendicular triplet as a 1,2-biradical in which there is little
or no interaction between the two radical termini. The study
most relevant to the present investigation is that in which the
addition of the p-acetylstyrene triplet to styrene was monitored.31
We reported a rate of addition in this process of 3.2 × 106 L
mol-1 s-1 and a 1,4-biradical ring closure efficiency of 0.38.
The results were compared to the published rate of addition of
face of the cuvette. This deposit eventually carbonized and
prevented further irradiation. However, one would expect the
proportion of 4 + 2 adducts would continue to increase at even
higher light intensities.
In an effort to extend these studies to higher laser powers,
preliminary experiments have been performed using a laser-
drop apparatus similar to that developed by Scaiano et al.36
4
+ 2 photodimers were again detected in these experiments but
the reaction mixtures proved more difficult to analyze. The
low overall quantum yield for dimer formation required multiple
passes through the apparatus and significant decomposition of
the starting material appeared to occur.
We had initially hoped to obtain more quantitative data from
our laser photolysis studies which would discriminate between
single- and multi-photon processes. To date this has proven
difficult. The relatively complex distribution of excited states
in the irradiated sample, both in terms of the depth of penetration
of the exciting laser pulse and as a function of time following
the pulse, added to the competing photoreactions in the system
of interest, have so far led to rather complex behavior as
variables such as laser power and temperature are changed. The
data presented in Figure 2 appear to suggest that the proportion
of 4 + 2 photodimers in the reaction mixture may level off to
a constant value as the laser power is increased. Ultimately,
the product distribution in the power dependence experiments
will reach a limiting value at the point at which complete
bleaching of the reaction sample occurs. This process is unlikely
to occur in our experiments. The apparent leveling off of the
distribution is an experimental artifact. Future investigations
may encompass studies of excitation in optically thin samples
where complete bleaching may be possible. In this case the
limiting value of the product distribution should be obtained.
Triplet-sensitized experiments have proven extremely useful
in the synthesis of the 4 + 2 photodimers. These dimers account
for almost 70% of the characterized products under laser
irradiation at 355 nm in comparison to only 15% under direct
irradiation at 266 nm. Low-temperature sensitized irradiations,
in our hands, gave 4 + 2 photodimers almost exclusively.
Triplet State Photoadditions. The quantum yield studies
of the triplet-sensitized formation of dimer 1 provide strong
evidence that the formation of 2 + 2 adducts under these
conditions is a result of a triplet state addition process analogous
to that outlined by Caldwell et al.31 for the addition of
p-acetylstyrene triplets to ground state styrene. The reaction is
expected to proceed via a 1,4-biradical as depicted in Scheme
2. The quantum yield for the reaction is given by
the 5-hexenyl radical to styrene37 of 1.4 × 105 L mol-1 s-1
.
We concluded that the order of magnitude discrepancy between
the two rates was due to the involvement of an exciplex
intermediate evidenced by the order of magnitude increase in
the rate of addition when styrene is replaced by its p-methoxy
derivative.
The present study has revealed an order of magnitude range
of possible rate constants for the photoaddition of the PC6 triplet
to its ground state. Published values for rates of radical addition
to ethylene,38 including methyl, isopropyl, and tert-butyl, are
essentially independent of the primary, secondary, or tertiary
nature of the attacking radical. One would expect steric factors
for additions to ethylene to be relatively low. The rate of
addition of the tert-butyl radical to styrene,37 at 1.3 × 105 L
mol-1 s-1, is essentially identical to that for primary radical
addition. Steric factors for addition to the styrene methylene
group again appear to be negligible. Our upper limit for the
rate of attack of the PC6 triplet on its ground state at 4 × 105
L mol-1 s-1 is therefore entirely in line with what one would
expect for an unhindered alkyl radical addition process while
the possible lowering of the rate constant by up to an order of
magnitude would be indicative of some degree of steric
hindrance for the addition of a secondary alkyl radical to the
methine group of a more highly substituted styrene derivative.
It is also interesting to consider the stereochemistry of the 2
+ 2 adducts arising from triplet-sensitized photoreactions.39 It
is, at first, surprising to note that the triplet-sensitized reaction
yields only the cis-fused 2 + 2 adducts previously isolated from
direct irradiations of PC6. One might imagine that the involve-
ment of rotationally equilibrated triplet 1,4-biradicals might
result in the production of trans-fused adducts as is often the
Φ1T ) kr[PC6]R(kr[PC6] + τ-1)-1
(37) Lusztyk, J.; Kanabus-Kaminska, J. M. In Handbook of Organic
Photochemistry; Scaiano, J. C., Ed.; CRC Press, Inc.: Boca Raton, 1989;
Vol. 2, pp 177-209.
(38) Kerr, J. A.; Trotman-Dickenson, A. F. In Progress in Reaction
Kinetics; Pergamon Press: New York, 1961; Vol. 1, pp 107-127.
(39) We would like to thank one of the referees for raising the question
of the selectivity of the ring junction stereochemistry in this reaction.
where kr is the rate of addition of the PC6 triplet to a ground
state PC6 molecule to form the 1,4-biradical intermediate, R is
(36) Banks, J. T.; Scaiano, J. C. J. Am. Chem. Soc. 1993, 115, 6409-
13.

Photodimerization of 1-Phenylcyclohexene
J. Am. Chem. Soc., Vol. 118, No. 7, 1996 1689
case in enone-alkene photoannelations.20 Molecular modeling
of the biradicals, however, clearly demonstrates that orbital
overlap leading to cyclobutane ring-closure and a trans-fused
adduct is severely inhibited unless significant pyramidalization
of one or both phenyl-substituted radical centers is invoked. In
contrast, ring closure to afford cis-fused adducts involves
biradical conformations in which orbital overlap is easily
attained and steric interactions between the two phenyl groups
appear to be minimized.
Scheme 3
It appears that orbital overlap leading to cis-fused adducts
takes place only from 1,4-biradicals in which the initially formed
linkage between the cyclohexane rings is diaxial. When this
bond is formed in an equatorial-axial or diequatorial manner,
orbital overlap is difficult to obtain and the two orbitals generally
lie perpendicular to each other. There are two distinct diaxially
linked 1,4-biradicals. Ring closure of one of these biradicals
leads to the isolated cis-anti-cis cyclobutane 1. Ring closure
of the alternative diaxially linked biradical leads to the cis-syn-
cis cyclobutane 2. However, orbital overlap in the latter case
involves significant steric interactions between the axial hy-
drogens on the two cyclohexane rings. This may account for
the observation that formation of 1 over 2 in the triplet-sensitized
system is favored in a 4:1 ratio.
How Reactive Are trans-Cyclohexenes? Our original
overarching objective was to evaluate the reactivity of trans-
cyclohexenes and their potential for providing novel chemistry.
The present case has, alas, afforded a novel mechanism but not
new reaction. The [4 + 2] dimerization of styrene under thermal
conditions is well-known.40 Of course, the obvious manifesta-
tion of the enorous strain (94 ( 6 kcal mol-1) in two molecules
of trans-PC6 is a far higher rate. Without further work we
cannot provide a rate constant but it clearly is very high. The
strong temperature dependence of the 4 + 2 dimerization
reaction requires that the activation energy for dimerization be
much lower than the 12.1 kcal mol-1 observed for trans-PC6
to cis-PC6 isomerization.9 A photochemical 4 + 2 cycloaddition
between two styrene chromophores has been outlined by Hart
et al.41 in a study of benzocycloheptadienone photochemistry.
The dimerization mechanism was suggested to involve attack
of a photogenerated trans double bond on a ground state cis
styrene moiety. Hart has also raised the possibility that the
observed 2 + 2 cycloaddition reactions of these and other
cycloheptenones might originate from either a trans + cis
interaction or one involving two trans intermediates.41,42 Bon-
neau43 has also suggested that the dimerization of cyclohep-
tenones might involve two trans intermediates based on kinetic
studies.
starting materials. This hypothesis is, however, inconsistent with
our previous observation9 that the lifetime of trans-PC6 is
essentially independent of the concentration of the ground state
cis compound. Studies of the chemistry of strained cyclo-
hexenes performed by House et al.44,45 reveal that steric
interactions may also be a factor in the observed reactivity.
Phenyl substitution at the 2-position of a bicyclo[3.3.1]non-1-
en-3-one effectively blocks the rapid 2 + 2 dimerization reaction
encountered for the parent molecule. That the only observed
dimerization process involving trans-PC6 appears to be the 4
+ 2 cycloaddition involving two trans intermediates may
therefore reflect a complex balance of both thermodynamic and
steric factors.
Summary
The photoreactions of PC6 as revealed in this study are
summarized in Scheme 3. Direct irradiation of PC6 leads to 2
+ 2 singlet state photodimers and the generation of trans-PC6.
Under conditions where transient-transient reactions are likely,
i.e., low reaction temperatures or high irradiation intensities,
reaction between two trans-PC6 molecules occurs giving rise
to two isomeric 4 + 2 adducts. When the reaction is performed
in methanol, these dimers are accompanied by significant
amounts of the trans-PC6 methanol addition product. Under
triplet-sensitized conditions trans-PC6 is again generated and
the above arguments apply. In addition, a triplet state photo-
dimerization occurs to produce the same pair of 2 + 2 adducts
but in a ratio which increasingly favors formation of dimer 1
over dimer 2. The rate of addition of the PC6 triplet to its
ground state is in the range (0.28-4.0) × 105 L mol-1 s-1
,
consistent with the modeling of alkene triplets as 1,2-biradicals.
Acknowledgment. This work was supported by the Robert
A. Welch foundation (Grant AT-0532) and the National Science
Foundation (Grant CHE-9121313). The trans-PC6 decay trace
(Figure 6) was obtained at the Center for Fast Kinetics Research
at the University of Texas at Austin.
That the 47 ( 3 kcal mol-1 present in one trans-PC6 molecule
is insufficient to allow an observable trans + cis 4 + 2
dimerization process is interesting. Presumably, the loss of
aromaticity in one ring, which would be mechanistically required
from either configuration of reactant, significantly counterbal-
ances the strain of the reactant and the thermodynamic driving
force is lost. An alternative possibility is that a trans + cis 4
+ 2 cycloaddition occurs but the resulting triene intermediate
is unable to undergo the required 1,3-antarafacial shift to afford
the Diels-Alder products and therefore fragments to yield
Supporting Information Available: Tables of positional
and thermal parameters and complete bond distances and angles
(8 pages). This material is contained in many libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, can be ordered from the ACS, and can
be downloaded from the Internet; see any current masthead page
for ordering information and Internet access instructions.
(40) Pryor, W. A.; Lasswell, L. D. In AdVances in Free Radical
Chemistry; Williams, G. H., Ed.; Elek Science: London, 1975; Vol. 5, pp
27-99.
(41) Dunkelblum, E.; Hart, H.; Suzuki, M. J. Am. Chem. Soc. 1977, 99,
5074.
(42) Hart, H.; Dunkelblum, E. J. Org. Chem. 1979, 44, 4752.
(43) Bonneau, R.; Violet, P. F. d.; Joussot-Bubien, J. NouV. J. Chim.
1977, 1, 31.
JA9533860
(44) House, H. O.; DeTar, M. B.; VanDerveer, D. J. Org. Chem. 1979,
44, 3793.
(45) House, H. O.; Outcalt, R. J.; Haack, J. L.; VanDerveer, D. J. Org.
Chem. 1983, 48, 1654.