Mechanistic analysis of the photocycloaddition of silyl-tethered alkenes

Article abstract of DOI:10.1021/jo7014664

(Chemical Equation Presented) The photochemistry of substituted cinnamyloxy silanes has been examined in both cyclohexane and acetonitrile solvents. Alkene isomerization occurs in addition to cycloaddition. Fluorescence quantum yields and excited singlet state lifetimes have been determined for each compound. We have used the information in order to better understand the regio- and stereoselectivity of photocycloaddition between silyl-tethered cinnamyl groups. This study allows us to conclude that the 2 + 2 photocycloaddition between alkenes is not a Woodward-Hoffmann orbital symmetry controlled event. The most consistent explanation for the excellent regio- and stereoselectivity is that the photocycloaddition is conformationally controlled.

Full text of DOI:10.1021/jo7014664

Mechanistic Analysis of the Photocycloaddition of Silyl-Tethered
Alkenes
Steven A. Fleming,* Alexander A. Parent, Ephraim E. Parent, James A. Pincock, and
Lise Renault
Department of Chemistry & Biochemistry, Brigham Young UniVersity, ProVo, Utah 84602, and
Department of Chemistry, Dalhousie UniVersity, Halifax, NS B3H 4J3, Canada
steVe_fleming@byu.edu
ReceiVed July 5, 2007
The photochemistry of substituted cinnamyloxy silanes has been examined in both cyclohexane and
acetonitrile solvents. Alkene isomerization occurs in addition to cycloaddition. Fluorescence quantum
yields and excited singlet state lifetimes have been determined for each compound. We have used the
information in order to better understand the regio- and stereoselectivity of photocycloaddition between
silyl-tethered cinnamyl groups. This study allows us to conclude that the 2 + 2 photocycloaddition between
alkenes is not a Woodward-Hoffmann orbital symmetry controlled event. The most consistent explanation
for the excellent regio- and stereoselectivity is that the photocycloaddition is conformationally controlled.
SCHEME 1. Dicinnamyloxysilane Irradiation
Introduction
We have been interested in the synthetic utility of 2 + 2
photocycloaddition using silyl tethered alkenes. One of our first
observations was that irradiation of dicinnamyloxysilane (1)
results in nearly quantitative formation of the cis-substituted
cyclobutane (2) as shown in Scheme 1.1 We were encouraged
by the efficiency of this reaction and promptly explored the
generality of this carbon-carbon bond-forming methodology.
We have found that this cycloaddition is effective between a
styrenyl group and other extended π systems (e.g., styrenyl,
dienyl, furanyl, cyclopropyl, alkynyl).1,2 In all cases we found
that the π-containing groups are obtained very selectively cis
to each other on the cyclobutane. However, a number of issues
indicated that a more thorough mechanistic study was necessary
to expand the methodology to include all alkenes. For instance,
cycloaddition was not observed between a styrenyl and a non-
π-extended alkene (e.g., allyl, 3-methylallyl, and 3,3-dimethy-
lallyl). The fact that the non-extended π-substituted alkenes are
not found to undergo photocycloaddition has limited the scope
of the reaction.
Enone-alkene and carbonyl-alkene (Paterno-Buchi) pho-
tocycloadditions have been studied extensively and have been
found useful in numerous synthetic applications.3 Alkene-
alkene photocycloadditions have been studied less and present
challenges that the triplet pathway of enones and Paterno-Buchi
reactions are able to avoid. One of the most significant obstacles
in alkene photochemistry is the short excited singlet state
lifetime. However, we have found that tethering the alkenes is
a logical and effective way to deal with this issue. Interestingly,
the triplet excited-state alkene + alkene cycloadditions give
trans-substituted cyclobutanes as the major products.4
(3) (a) Griesbeck, A. Synth. Org. Photochem., Mol. Supramol. Photo-
chem. 2004, 12, 89-140. (b) Margaretha, P. Synth. Org. Photochem., Mol.
Supramol. Photochem. 2004, 12, 211-238. For an excellent review of the
stereocontrol in photochemical reactions see: Griesbeck, A. G.; Meierhen-
rich, U. J. Angew. Chem., Int. Ed. 2002, 41, 3147-3154.
(4) (a) DeBoer, C. D.; Wadsworth, D. H.; Perkins, W. C. J. Am. Chem.
Soc. 1973, 95, 861-869. (b) Vassilikogiannakis, G.; Hatzimarinaki, M.;
Orfanopoulos, M. J. Org. Chem. 2000, 65, 8180-8187. (c) Unett, D. J.;
Caldwell, R. A.; Hrncir, D. C. J. Am. Chem. Soc. 1996, 118, 1682-1689.
* Address correspondence to this author at Brigham Young University.
(1) Ward, S. C.; Fleming. S. A. J. Org. Chem. 1994, 59, 6476-6479.
(2) Bradford, C. L.; Fleming, S. A.; Ward, S. C. Tetrahedron Lett. 1995,
36, 4189-4192. For a recent summary of cyclobutane biomolecules, see:
Ortun˜o, R. M.; Moglioni, A. G.; Moltrasio, G. Y. Curr. Org. Chem. 2005,
9, 237-259.
10.1021/jo7014664 CCC: $37.00 © 2007 American Chemical Society
Published on Web 11/13/2007
9464
J. Org. Chem. 2007, 72, 9464-9470

Photocycloaddition of Silyl-Tethered Alkenes
SCHEME 2. Mechanistic Pathways Available for Alkene-Alkene 2 + 2 Photocycloaddition
Copper coordination has also been used to bring alkenes
together for photocycloaddition. Steric interactions and het-
eroatom coordination control this useful reaction.5 Ramamurthy
and co-workers6 have reported the selectivity of photocycload-
ditions in caged systems. This very useful methodology,
however, does not easily extend to photocycloaddition between
two differing alkenes. Solid-state photochemistry7 has also been
used for providing stereocontrol of cycloadducts. Once again,
this approach is difficult to apply when cycloaddition between
two different alkenes is desired.
Silicon is not the only tethering moiety available for alkene
photocycloadditions. Carbon-tethered styryl groups also give
high yields of stereoselectively formed cyclobutanes,8 although
this tether unit is more difficult to remove. It is worth noting
that these examples, and the earlier work by many others,9
consistently show a preference for formation of the cis-
substituted cyclobutanes. Also worth noting is that the use of
cinnamyl groups for photo-cross-linking in polymer systems has
been reported by Nakayama and Matsuda.10 A better under-
standing of the reaction would improve the utility of polymer
cross-linking.
Possible Mechanisms
There are at least three mechanistic pathways (see Scheme
2) that one might consider for the silyl-tethered photocycload-
dition. Path A is the concerted, symmetry allowed, 2 + 2
process. This should, however, lead to trans-diphenylcyclobu-
tanes or trans-fused bicyclo[5.2.0] ring systems. Path B is a
concerted π-stacking pathway. It is possible that the cis
selectivity might arise from a π-stacking phenomenon. Good
evidence has been presented for this type of interaction in the
vinyl anthracene case reported by Nishimura and co-workers.11
If one assumes that the π interaction is necessary, either in the
ground or the excited state, the lack of reactivity of non-extended
π-substituted alkenes would be explained. Path C that should
be considered, is a stepwise radical mechanism proceeding
through a 1,4-diradical.12 This mechanism is consistent with the
need for radical-stabilizing extended π-groups on the reacting
alkenes and fits particularly well with the cross 2 + 2 results
from the irradiation of silyl enol ether 3 shown in Scheme 3.1
To further elucidate the mechanism of this reaction, we have
synthesized a number of silyl-tethered alkenes and, on the basis
of their photophysical properties and photochemical outcome,
are prepared to draw conclusions concerning the reaction
pathway.
(5) (a) Avasthi, K.; Raychaudhuri, S. R.; Salomon, R. G. J. Org. Chem.
1984, 49, 4322-4324. (b) Ghosh, S.; Banerjee, S.; Chowdhury, K.;
Mukherjee, M.; Howard, J. A. K. Tetrahedron Lett. 2001, 42, 5997-6000.
(c) Sarkar, N.; Ghosh, S. Indian J. Chem. 2006, 45B, 2474-2484.
(6) Kaanumalle, L. S.; Natarajan, A.; Ramamurthy, V. Synth. Org.
Photochem., Mol. Supramol. Photochem. 2004, 12, 553-618. See also:
Noh, T.; Kwon, H.; Choi, K.; Choi, K. Bull. Korean Chem. Soc. 1999, 20,
76-80.
(7) Amirsakis, D. G.; Elizarov, A. M.; Garcia-Garibay, M. A.; Glick, P.
T.; Stoddart, J. F.; White, A. J. P.; Williams, D. J. Angew. Chem., Int. Ed.
2003, 42, 1126-1132.
(8) For a review of the alkene-alkene photocycloaddition, see: Fleming,
S. A. Synth. Org. Photochem., Mol. Supramol. Photochem. 2004, 12, 141-
160. For a review of regio- and stereocontrol, see: Fleming, S. A.; Bradford,
C. L.; Gao, J. J. Org. Photochem.; Mol. Supramol. Photochem. 1997, 1,
187-244. For a recent example of carbon tethered 2 + 2 see: Vedernikov,
A. I.; Lobova, N. A.; Ushakov, E. N.; Alfimov, M. V.; Gromov, S. P.
MendeleeV Commun. 2005, 173-175.
Results and Discussion
1. Bis-Silyl Enol Ether (Z,Z-4). The bis-silyl enol ether (Z,Z-
4, R ) CH3) was reported to undergo bond formation through
(11) Nakamura, Y.; Fujii, T.; Nishimuri, J. Chem. Lett. 2001, 30, 970-
971. See also: (a) Nishimuri, J.; Horikoshi, Y.; Wada, Y.; Takahashi, H.;
Sato, M. J. Am. Chem. Soc. 1991, 113, 3485-3489. (b) Ricks, H. L.;
Shimizu, L. S.; Smith, M. D.; Bunz, U. H. F.; Shimizu, K. D. Tetrahedron
Lett. 2004, 45, 3229-3232.
(12) Calculations suggest that the ethylene + ethylene pathway “involves
an asynchronous two-step process involving anti attack.” See: Palmer, I.
J.; Olivucci, M.; Bernardi, F.; Robb, M. A. J. Org. Chem. 1992, 57, 5081-
5087.
(9) (a) Ito, Y.; Fujita, H. Chem. Lett. 2000, 288-289. (b) Lewis, F. D.;
Hirsch, R. H. J. Am. Chem. Soc. 1976, 98, 5914-5924. (c) Greiving, H.;
Hopf, H.; Jones, P. G.; Bubenitschek, P.; Desvergne, P.; Bouas-Laurent,
H. J. Chem. Soc., Chem. Commun. 1994, 1075-1076.
(10) Nakayama, Y.; Matsuda, T. J. Polym. Sci. A: Polym. Chem. 2005,
43, 3324-3336. See also: Nagarathinam, M.; Vittal, J. J. Macromol. Rapid
Commun. 2006, 27, 1091-1099.
J. Org. Chem, Vol. 72, No. 25, 2007 9465

Fleming et al.
SCHEME 3. Cinnamyloxysilyl Enol Ether and the Cross 2 + 2 Cycloaddition Product
SCHEME 4. Bis-Silyl Enol Ether Z,Z-4 SET Results
SCHEME 5. Bis-Silyl Enol Ether Photochemical Results
an intermedate radical cation in the presence of ceric ammonium
nitrate (CAN) in a regio- and stereoselective fashion under
thermal conditions (see Scheme 4) by Schmittel and co-
workers.13 They found that the product in this reaction resulted
from an avoidance of the steric interactions between the phenyl
groups. They argued that the favored ground state conformation
of this molecule has the phenyl groups not π stacking. On the
basis of this work, we felt that a photochemical study of this
compound would be an excellent way to establish the role of π
stacking in alkene-alkene tethered systems. If cycloaddition
occurs and π stacking is required, then π stacking would
necessarily be an excited state phenomenon and one would
expect the phenyl groups to be cis to each other in the resulting
cyclobutane. If cycloaddition occurs and π stacking is not
required, then the Schmittel-favored conformation would yield
trans phenyl groups in the resulting cyclobutane. If the cyclobu-
tane product has the phenyl groups trans to each other, then we
would conclude that the photocycloaddition requires only the
radical stabilizing ability of the extended π system.
We synthesized compound Z,Z-4 (R ) iPr) using the literature
procedure13 and irradiated the resulting tethered diene. Unfor-
tunately, the cyclobutane products that were obtained included
both cis- and trans-1,2-diphenylcyclobutanes in nearly equal
amounts as shown in Scheme 5. These ratios were determined
by NMR analysis of the crude mixture of hydrolyzed material.
Diols A and C both have two methyl doublets, one of which is
significantly shielded (0.40 and 0.71 ppm, respectively). The
symmetrical cyclobutanes B and D have only one methyl
doublet (1.24 and 1.18 ppm, respectively). We believe that the
cis-diphenyl arrangement in B results in less symmetry for the
cyclobutyl ring hydrogens which are observed at 3.10 ppm
compared to the clean quartet for the ring hydrogens of D that
appear at 3.07 ppm.
isomerization during the reaction. For instance, as determined
by NMR at 65% conversion of Z,Z-4, both E,Z-4 and E,E-4
were observed as 60% of the photomixture and less than 5% of
the cyclobutanes were formed. It is obvious that this lack of
stereochemical integrity complicates what was intended to be
an easy answer to the mechanistic question. Our study of
compound 4 does not elucidate the mechanism of the silyl-
tethered photocycloaddition.
2. Cyclohexenyl Cinnamyl Ether (6). Because the bis-silyl
enol ether did not provide an increase in mechanistic under-
standing, we turned our attention to a tethered silyl enol ether
in which one substituent was lacking a conjugated phenyl group.
The synthesis of the silyl enol ether of cyclohexanone with use
of chlorocinnamyloxydiisopropylsilane (5) resulted in formation
of the diene 6 (see Scheme 6). We reasoned that this compound
would not undergo photocycloaddition if π stacking (Path B)
is required. We also felt that the diradical-stabilized pathway
(Path C) would lead to formation of the cross 2 + 2 product,
compound 7.
Irradiation of 6 leads to two major cyclobutanyl products 8
and 9 (1:1) and two diastereomeric cyclohexanones 10 (see
Scheme 7). Identification of the photoproducts was accom-
plished by several NMR methods including COSY, HETCOR,
and NOE. The fact that the cross 2 + 2 product (7) is not formed
establishes that the most stable diradical intermediate is not
involved (Path C). Interestingly these products implicate the
need for a minimum energy conformational arrangement of the
alkenes in order for the cycloaddition to occur. Thus, a π
stacking of the alkenes is needed, but not necessarily a stacking
of aromatic rings.
The cyclohexanone products are thought to arise from
hydrolysis and ring-opening of the initially formed cyclobutane
products. We have found that the cyclobutane products are
photochemically stable and do not appear to be undergoing
secondary photochemistry to give the 2-(1-phenylallyl)cyclo-
hexanone.
In addition to the nearly equal distribution of cyclobutanes,
we observed that the original silyl enol ether undergoes E-Z
(13) Schmittel, M.; Burghart, A.; Malisch, W.; Reising, J.; So¨llner, R.
J. Org. Chem. 1998, 63, 396-400.
9466 J. Org. Chem., Vol. 72, No. 25, 2007

Photocycloaddition of Silyl-Tethered Alkenes
SCHEME 6. Synthesis and Predicted Photochemistry of Cinnamyloxysilyl Enol Ether (6)
SCHEME 7. Photochemical Results of Dialkoxysilane 6
3. Cinnamyloxy Silanes 11a,b and 12a-e. To probe for a
possible interaction between the rings in diaryl derivatives, we
have examined the photochemistry and photophysics of two
ring-substituted dicinnamyloxysilanes, 11a and 11b. These were
designed with an electron donor on one ring (methoxy group)
and an electron-withdrawing group on the other (trifluoromethyl
group) to test for the possibility of π stacking with this push-
pull type system.14 We also speculated that the alkenes would
more efficiently undergo cycloaddition in this type of tethered
system. The synthesis of the dialkoxy silanes was performed
as with previous tethering of alkenes proceeding through
reaction of dichlorodiisopropylsilane with 1 equiv of the
appropriate cinnamyl alcohol to give the alkoxy chlorodiiso-
propylsilane, which was then distilled and finally reacted with
a second alcohol. In this fashion, (4-methoxycinnamyloxy)(4-
trifluoromethylcinnamyloxy)diisopropylsilane (11a) and (3-
methoxycinnamyloxy)(3-trifluoromethylcinnamyloxy)diiso-
propylsilane (11b) were synthesized (see Figure 1).
Although we thought we might see evidence of a charge-
transfer interaction in the ground state between the two arms
of the tethered silane, none was found. For instance, the UV
spectrum in acetonitrile (AN) for (4-methoxycinnamyloxy)(4-
trifluoromethylcinnamyloxy)diisopropylsilane 11a (ꢀ at the λmax
of 258 nm ) 3.95 × 104 M-1 cm-1) was superimposable on
the sum of the normalized UV spectra of 4-trifluoromethylcin-
To compare the photophysical properties of the methoxy- and
trifluoromethyl-substituted compounds, we also synthesized and
studied the TMS ethers 12a-d (Figure 1) of the para- and meta-
substituted cinnamyl alcohols.
(14) For a recent example of photophysical evidence of π stacking see:
Hisamatsu, Y.; Takami, H.; Shirai, N.; Ikeda, S.; Odashima, K. Tetrahedron
Lett. 2007, 48, 617-621.
FIGURE 1. Alkenes 11a,b and 12a-d synthesized for photophysics.
J. Org. Chem, Vol. 72, No. 25, 2007 9467

Fleming et al.
TABLE 1. Photophysics of Silanes in Acetonitrile (AN) and Cyclohexane (CX) at 25 °C
quantum yield of
fluorescence
lifetime (ns)a
compd
AN
CX
AN
CX
4-CF3 cinnamyloxy silane, 12a
3-CF3 cinnamyloxy silane, 12b
4-MeO cinnamyloxy silane, 12c
3-MeO cinnamyloxy silane, 12d
0.17
0.31
0.36
0.50
0.26
0.22
0.50
0.44
3.0
5.4
6.3
6.5
3.7
4.2
6.5
5.8
4-H, 4-H dicinnamyloxy silane, 1
4-MeO, 4-CF3 dialkoxy silane, 11a
3-MeO, 3-CF3 dialkoxy silane, 11b
0.01
0.035
0.12
0.47 (98); 8.4 (2)
0.31 (96); 5.7 (4)
0.95 (98); 7.7 (2)
0.043
0.064
0.35 (97); 4.4 (3)
0.72 (98); 5.1 (2)
a The fluorescence decay of the dicinnamyl derivatives was biexponential and both lifetimes are given. The numbers in parentheses give the relative
contribution of each decay to the total decay.
TABLE 2. Photoproducts (% Yield) from Irradiation of TMS Ethers for 1 h at 254 nm
E-starting
material
Z-starting
material
E-1,3
H-
Z-1,3
H-
compd
solvent
unknown
4-CF3 TMS, 12a
4-CF3 TMS, 12a
3-CF3 TMS, 12b
3-CF3 TMS, 12b
4-methoxy TMS, 12c
4-methoxy TMS, 12c
3-methoxy TMS, 12d
3-methoxy TMS, 12d
cinnamyl TMS, 12e
cinnamyl TMS, 12e
CX
AN
CX
AN
CX
AN
CX
AN
CX
AN
51
< 5
37
17
3
trace
24
46
38
45
40
30
35
22
4
trace
8
12
40
29
9
20 (ald)
25 (ald)
55 (ald)
trace
6
trace
59
60
28
21
34
40
25
21
9
15
12
13
14
namyloxy trimethylsilyl ether 12a (λmax ) 252 nm, ꢀ at 250
nm ) 1.96 × 104 M-1 cm-1) and 4-methoxycinnamyloxy
trimethylsilyl ether 12c (λmax ) 261 nm, ꢀ at 250 nm ) 1.90 ×
104 M-1 cm-1). In fact the UV spectra were all very similar to
that expected for styrene derivatives with a lower intensity, long
wavelength band exhibiting vibrational fine structure between
270 and 310 nm and a broad, higher intensity short wavelength
band around 250 nm; the relative intensity of these two bands
was about 10:1.
However, the fluorescence spectra did give evidence for a
π-stacking interaction in the excited state. Table 1 shows the
fluorescence quantum yields for each of the cinnamyl com-
pounds in both acetonitrile (AN) and cyclohexane (CX). The
mono-cinnamyloxy silanes (12a-d) are similar in fluorescence
quantum yield in both solvents although the meta-substituted
compounds 12b and 12d fluoresce somewhat more efficiently
than the para analogues 12a and 12c in AN but less efficiently
in CX. In addition, the methoxy cinnamyl derivatives 12c and
12d are more efficient fluorophores than the trifluoromethyl
compounds 12a and 12b. This is a consequence of the extended
π system available to the alkoxy-substituted rings.
As indicated by their decreased fluorescence yields, the
dialkoxy-substituted silanes clearly have a pathway of deactiva-
tion that is not available for the monomers, which is clear
evidence of π-π interaction between the chromophores.
Interestingly, the meta-disubstituted silane 11b is more fluo-
rescent than the para-substituted one (11a). We suspect this is
due to increased overlap between the two styrenyl chromophores
in the para-substituted isomer. This overlap may enhance the
alternative, nonradiative pathway for deactivation of the excited
state.
sponding mono-ene, where determined. This observation is
perhaps indicative of two conformers, a major one that is short-
lived because of π-stacking and a stretched one that is long-
lived. This implies that the dienes have a rapid decay route that
is unavailable to the monomers, presumably a rapid and efficient
cycloaddition process with a rate constant on the order of 109
s-1 that competes with fluorescence. Although the differences
are small, this decay pathway is most efficient for the case that
has the tethered cinnamyloxy groups substituted in the para
positions (lifetime of 0.31 ns for 11a) compared to the isomeric
meta compound (lifetime of 0.95 ns for 11b). This decay route
is also faster than that of the unsubstituted dicinnamyloxy silane
(lifetime of 0.47 ns for 1), although the difference is not as
significant. The data for the dialkoxy-tethered silanes determined
in cyclohexane shows similar lifetime behavior.
The photoproducts from each of the trimethylsilyl ethers are
provided in Table 2 from irradiations in both cyclohexane and
acetonitrile. Along with rapid photochemical E to Z isomeriza-
tion, the major product formed upon irradiation of the unsub-
stituted cinnamyl TMS ether 12e in both cyclohexane and
acetonitrile was the 1,3-hydride shift product, 3-phenyl-1-
propenyl TMS ethers (13E and 13Z, see Scheme 8). This
compound was independently synthesized to verify its identity.
SCHEME 8. Photoisomerization of Cinnamyloxy
Trimethylsilyl Ether
The excited state lifetime data provide additional evidence
of π-stacking. Each of the tethered dienes exhibited a biexpo-
nential fluorescence decay. In all cases, the short-lived com-
ponent is the major contributor to the decay (>95%). Moreover,
the long-lived component is similar in lifetime to the corre-
The 3-methoxy-substituted derivative 12d behaves similarly
to the unsubstituted cinnamyl silyl ether, providing evidence
of poor communication between the chromophore, which is the
9468 J. Org. Chem., Vol. 72, No. 25, 2007

Photocycloaddition of Silyl-Tethered Alkenes
styrene moiety, and the meta substitutent.15 In contrast, the
trifluoromethyl analogues (12a and 12b) were much more
photoactive than the unsubstituted TMS ether (12e), both
undergoing essentially complete conversion to the 1,3-hydride
shift product in the 1 h of irradiation. Presumably this is a
reflection of the significant inductive effect that the CF3 group
provides from both the 3- and 4-positions. The aldehyde that is
observed in these reactions is detected by NMR and likely arises
from hydrolysis of the moisture-sensitive silyl enol ethers.
Irradiation of the 4-methoxy analogue, compound 12c, results
in cis/trans isomerization as the predominant pathway in either
solvent, with only low yields of the 1,3-shift product. A low
yield of the comparable product was also observed for the
photochemistry of 4-methoxycinnamyl acetate in both cyclo-
hexane and methanol.15
The photochemistry of the tethered alkenes provides strong
evidence for the importance of the interaction between the two
chromophores. The dicinnamyl silane, as previously reported,1
undergoes a very efficient cycloaddition. In fact, we were forced
to reduce the rayonet photoreactor light intensity by reducing
the number of lamps from 16 to 4 in order to see partial
conversion (less than 40%) after 5 min of irradiation. Even
compound 11b, with the m-trifluoromethylcinnamyl group
tethered to the m-methoxycinnamyl, reacted with a similar
efficiency to compound 1. In contrast, the 4-methoxycinnamyl-
4-trifluoromethylcinnamyl silane, 11a, was significantly more
photoactive. In the same photolysis setup, a 75% conversion
resulted after only 5 min. We interpret this as evidence of
increased π stacking and more conformational control between
the styrenyl groups in the excited state for the para-substituted
analogue (Path B). Both tethered systems 11a and 11b give the
corresponding cyclobutanes as single products with the same
regio- and stereochemical outcome as the unsubstituted com-
pound 1 (see Scheme 1). The yield for the photocycloaddition
is >95% based on NMR analysis. Prolonged irradiation does
result in loss of the original cyclobutane.
of a catalytic amount of p-toluenesulfonic acid for 24-36 h. The
isolated ester was then treated with 3 equiv of diisobutyl aluminum
hydride (DIBAL) in toluene in an ice bath. The reaction was
allowed to stir for 2 h. The reaction was quenched with aqueous
5% HCl and extracted with ether. The organic layer was washed
with aqueous 5% HCl, water, and brine, and then dried. The
resulting alcohol was purified by recrystallization or chromatog-
raphy. Treatment of the alcohol with excess chlorotrimethylsilane
in methylene chloride, then purification by chromatography, yielded
the substituted cinnamyl silyl ether. The dialkoxysilanes were
synthesized by treating an alcohol with dichlorodiisopropylsilane
(4 equiv) and triethylamine (1 equiv) in methylene chloride in an
ice bath. The reaction was stirred overnight at rt. The monochlo-
rosilane was isolated by concentration and pentane extraction.
Kugelrohr distillation (approximately 120 °C, 0.2 mmHg) gave the
pure monochloride, which was treated immediately with 1 equiv
of the second alcohol and triethylamine in methylene chloride. The
mixture was stirred for 18 h at rt, then concentrated and washed
with pentane. Concentration and chromatography (silica gel, 5%
ethyl acetate:hexane) gave the desired dialkoxysilane.
(Z,Z)-Diisopropyldi(1-phenyl-1-propenoxy)silane (4).13 Syn-
thesis of the bis-silyl enol ether was carried out as indicated in the
literature with use of dichlorodiisopropylsilane rather than dimeth-
yldichlorosilane. The crude mixture (41% yield) was purified by
chromatography (silica gel, hexane) to give the desired compound
in 30% yield.
Diisopropylcinnamyloxysilyl Enol Ether of Cyclohexanone
(6). The synthesis of the cinnamyloxysilyl enol ether of cyclohex-
anone was accomplished by using a modified literature procedure.16
Sodium iodide (0.60 g, 4.0 mmol, dried in a 200 °C oven for a
week) was dissolved in 5 mL of acetonitrile and added to 0.37 g
(3.8 mmol) of cyclohexanone in 10 mL of ACN. To this mixture
was added 0.55 g (5.4 mmol) of Et3N, followed by dropwise
addition of 1.07 g (3.8 mmol) of diisopropylcinnamyloxysilyl
chloride. The resulting solution was diluted to 25 mL with
acetonitrile and heated at reflux with stirring for 4 h. The solution
was allowed to cool to room temperature, diluted with ap-
proximately 20 mL of distilled water, and extracted with pentane
three times. The combined organic layers were then washed three
times with distilled water, dried over anhydrous sodium sulfate,
and concentrated to give a crude yield of 1.16 g (3.4 mmol, 89%
yield). Chromatography gave 0.383 g (1.1 mmol, 29% yield) of
pure product.
Conclusion
Our results indicate that singlet 2 + 2 photocycloaddition
between alkenes is dependent on a close relationship between
the alkene groups. We have observed that the alkenes can be
held sufficiently close as a result of either π stacking or a
conformational preference. This is consistent with a short-lived
excited state or exciplex, which has the ability to stabilize polar
groups. Most importantly, these results indicate that the regio-
and stereochemical outcome of the 2 + 2 cycloaddition between
tethered alkenes is not dictated by the rules of orbital symmetry
nor by stability of radical intermediates.
We are continuing to study the utility of the alkene + alkene
photocycloaddition. We expect that a polar group on the para
position will produce sufficient π-π interactions between a
styrenyl group and a non-aryl-substituted alkene (e.g., allyl,
3-methylallyl, or 3,3-dimethylallyl) to result in efficient pho-
tocycloaddition.
General Photochemical Procedure. The silyl ethers were
irradiated in nonpolar (cyclohexane) or polar (acetonitrile) solvents.
Irradiations were performed on a rayonet apparatus, using with use
of sixteen 254 nm lamps with a quartz well in the center of the
irradiation chamber unless otherwise indicated. The samples were
approximately 100 mg in 100 mL of solvent. All irradiations were
deoxygenated by bubbling with nitrogen for at least 30 min prior
to and during the irradiation. In the rayonet setup the solutions were
stirred throughout the experiment and a temperature probe main-
tained the temperature at 25 °C. The solutions were monitored by
tlc or GC/MS during the experiment and concentrated after 1 h
unless otherwise indicated. NMR analysis was used to determine
the components present in each photomixture. For cycloaddition
reactions the conversion to cyclobutane was confirmed by GC and
NMR. The amount of conversion was based on NMR integration
of the -CH2O- group.
Irradiations in Cyclohexane: (a) (Z,Z)-Diisopropyldi(1-phen-
yl-1-propenoxy)silane (4). The bis-silyl enol ether (300 mg, 0.79
mmol) was added to 350 mL of cyclohexane and the mixture was
irradiated with a 450 W immersion well apparatus with a quartz
jacket for 10 h. The solution was concentrated and transferred to
50 mL of methanol to which NH4F (105 mg, 2.84 mmol) was added
and refluxed. A water/chloroform extraction yielded 180 mg of
crude product. The mass lost after the extraction roughly corre-
Experimental Section
General Synthetic Procedure for Substituted Silanes. The
commercially available substituted cinnamic acid was converted
to its respective ester by refluxing in 100% ethanol in the presence
(15) Fleming, S. A.; Grundy, E.; Renault, L.; Pincock, J. A. Can. J. Chem.
2006, 84, 1146-1154.
(16) Manis, P. A.; Rathke, M. W. J. Org. Chem. 1981, 46, 5348-5351.
J. Org. Chem, Vol. 72, No. 25, 2007 9469

Fleming et al.
sponds to the theoretical amount lost upon cleavage of the silyl
tether (120 mg, 1.05 mmol). Separation by chromatotron, using an
increasing concentration of an ethyl acetate:hexane solvent, yielded
8 UV active bands. These bands included the four 3,4-dimethyl-
1,2-diphenyl-1,2-cyclobutanediol products (42 mg, 0.15 mmol,
19%), propiophenone (57 mg, 0.43 mmol, 27%), polymeric
material, and two other minor fractions (<5%) that were unidenti-
fied.
(b) (Z,Z)-Diisopropyldi(1-phenyl-1-propenoxy)silane (4), Low
Conversion. The bis-silyl enol ether (110 mg) was added to 100
mL of cyclohexane and the mixture was irradiated with a 450 W
immersion well apparatus with a quartz jacket for 1 h. The solution
was concentrated and an NMR was taken of the crude material.
The presence of two major methyl signals in the 1.8 region establish
the E-Z isomerization of the alkene. There is more E isomer (1.75
ppm) than the initial Z (1.82 ppm) and only a small amount of
cycloaddition (methyl doublet at 0.5 and 0.8 ppm) after this short
time of irradiation.
Irradiations in Acetonitrile: (a) (Z,Z)-Diisopropyldi(1-phenyl-
1-propenoxy)silane (4). The bis-silyl enol ether (488 mg, 1.28
mmol) was added to 510 mL of acetonitrile and the mixture was
irradiated with a 450 W immersion well apparatus with a quartz
jacket for 11 h. The solution was concentrated and transferred to
50 mL of methanol to which NH4F (190 mg, 5.14 mmol) was added.
This solution was refluxed for 10 h and then concentrated under
reduced pressure. A water/chloroform extraction yielded 331 mg
of crude product. The mass lost after the extraction corresponds to
the theoretical amount lost upon cleavage of the silyl tether (149
mg, 1.30 mmol). Separation by chromatotron, using increasing
concentrations of an ethyl acetate:hexane solvent, yielded 13 UV
active bands. These bands included the four 3,4-dimethyl-1,2-
diphenyl-1,2-cyclobutanediol products (116 mg, 0.44 mmol, 34%),
propiophenone (93 mg, 0.70 mmol, 27%), polymer (85 mg, 26 mass
percent of crude product), and seven other fractions that were
unidentified.
(b) Diisopropylcinnamyloxysilyl Enol Ether of Cyclohexanone
(6). Irradiation of the cinnamyloxysilyl enol ether of cyclohexanone
(0.242 mg in 250 mL) was carried out with use of a 450 W
immersion well with a quartz jacket. Complete conversion of
starting material was noted by 15 min. Four major products were
obtained. Two are cyclobutanes and two are diastereomeric isomers
of desilylated photoproduct (ca. 30%). The diastereomers are known
compounds and are very difficult to separate.22 The percent yield
of the cyclobutane products is 50%.
(c) Dicinnamyloxydiisopropylsilane (1). Irradiation of 0.099 g
of dicinnamyloxy silane1 in 100 mL of acetonitrile with the standard
rayonet apparatus (16 lamps) for 20 min resulted in complete
consumption of starting material. Concentration of solvent provided
0.101 g of cyclobutane (>80%). Irradiation of 0.105 g of dicin-
namyloxy silane in 100 mL of acetonitrile under the same conditions
for 10 min resulted in 0.110 g of material that contained no starting
material and ∼85% cyclobutane. Irradiation of 0.095 g of dicin-
namyloxy silane in 95 mL of acetonitrile for 5 min resulted in
complete conversion to ∼95% cyclobutane. Irradiation of 0.097 g
of dicinnamyloxy silane in 99 mL of acetonitrile with four lamps
in the rayonet apparatus for 5 min gave a clean incomplete
conversion to cyclobutane. The ratio of dialkenylsilane to cyclobu-
tane was 61.9:38.1. Irradiation of 0.055 g of dicinnamyloxy silane
in 75 mL of acetonitrile for 10 min with the four-lamp arrangement
gave 0.060 g of cyclobutane (>95% conversion).
Supporting Information Available: Experimental procedures
for this work and the spectral data for 4, 6, 8, 9, 10, 11a,b, 12a-e,
and photoproducts of 11a,b and 12a-e. This material is available
free of charge via the Internet at http://pubs.acs.org.
JO7014664
9470 J. Org. Chem., Vol. 72, No. 25, 2007