Oxetanes from [2+2] cycloaddition of stilbenes to quinone via photoinduced electron transfer

Article abstract of DOI:10.1021/jo981754n

The photochemical coupling of various stilbenes (S) and chloranil (Q) is effected by the specific charge-transfer (CT) activation of the precursor electron donor-acceptor (EDA) complex [S, Q], and the [2+2] cycloaddition is established by X-ray structure elucidation of the crystalline transoxetanes formed selectively in high yields. Time-resolved (fs/ps) spectroscopy reveals the (singlet) ion-radical pair ¹[S·⁺, Q·^-] to be the primary reaction intermediate and thus unambiguously establishes for the first time the electron-transfer pathway for this typical Paterno-Buchi transformation. The alternative cycloaddition via the specific activation of the carbonyl component (as a commonly applied procedure in Paterno-Buchi couplings) leads to the same oxetane regioisomers in identical molar ratios. As such, we conclude that a common electron-transfer mechanism applies via the quenching of the photoactivated quinone acceptor by the stilbene donor to afford triplet ion-radical pairs ³[S·⁺, Q·^-] which appear on the ns/μs time scale. The spin multiplicities of the critical ion-pair intermediate [S·⁺, Q·^-] in the two photoactivation methodologies determine the time scale of the reaction sequences (which are otherwise the same), and thus the efficiency of the relatively slow ion-pair collapses (k(C) ? 10⁸ s^-1) to the 1,4-biradical that ultimately leads to the oxetane product.

Full text of DOI:10.1021/jo981754n

2250
J . Org. Chem. 1999, 64, 2250-2258
Oxeta n es fr om [2+2] Cycloa d d ition of Stilben es to Qu in on e via
P h otoin d u ced Electr on Tr a n sfer †
Duoli Sun, Stephan M. Hubig, and J ay K. Kochi*
Department of Chemistry, University of Houston, Houston, Texas 77204-5641
Received August 26, 1998
The photochemical coupling of various stilbenes (S) and chloranil (Q) is effected by the specific
charge-transfer (CT) activation of the precursor electron donor-acceptor (EDA) complex [S, Q],
and the [2+2] cycloaddition is established by X-ray structure elucidation of the crystalline trans-
oxetanes formed selectively in high yields. Time-resolved (fs/ps) spectroscopy reveals the (singlet)
ion-radical pair 1[S•+, Q•-] to be the primary reaction intermediate and thus unambiguously
establishes for the first time the electron-transfer pathway for this typical Paterno-Bu¨chi
transformation. The alternative cycloaddition via the specific activation of the carbonyl component
(as a commonly applied procedure in Paterno-Bu¨chi couplings) leads to the same oxetane
regioisomers in identical molar ratios. As such, we conclude that a common electron-transfer
mechanism applies via the quenching of the photoactivated quinone acceptor by the stilbene donor
to afford triplet ion-radical pairs 3[S•+, Q•-] which appear on the ns/µs time scale. The spin
multiplicities of the critical ion-pair intermediate [S•+, Q•-] in the two photoactivation methodologies
determine the time scale of the reaction sequences (which are otherwise the same), and thus the
efficiency of the relatively slow ion-pair collapses (kC = 108 s-1) to the 1,4-biradical that ultimately
leads to the oxetane product.
In tr od u ction
and the mechanistic relevance of the transient species
have not been established in most studies, and the
general validity of an electron-transfer mechanism is
controversial. In fact, there is still no definitive answer
to the question as to whether (a) the 1,4-biradicals that
are observed in Paterno-Bu¨chi couplings11-14 result
directly from the quenching of photoexcited carbonyl by
the olefin, or whether (b) other reactive intermediates
(such as exciplexes7-9 or ion-radicals9-13) represent the
initial quenching products that are subsequently con-
verted to biradicals. These ambiguities in the identifica-
tion of the primary reactive intermediates are inherent
to the photoactivation methodology that is commonly
applied in Paterno-Bu¨chi couplings. Thus, the photo-
excitation of the carbonyl leads to its excited (singlet or
triplet) state which is known to be quenched via a variety
of pathways, including energy transfer, electron transfer,
bond formation, bond cleavage, etc. As a result, different
reactive species may be observed initially, but the
primary quenching product cannot be unambiguously
identified (experimentally) since the observation and
identification depend critically on its lifetime and spec-
troscopic accessibility.
The photoinduced [2+2] cycloaddition of an olefin to a
carbonyl center, also known as the Paterno-Bu¨chi reac-
tion,1 is a synthetically useful photochemical transforma-
tion since it provides a simple route to the preparation
of oxetanes with high regio- and stereoselectivity (eq 1).2
The scope of this photocycloaddition has been widely
expanded over the years to include (a) the preparation
of diastereochemically pure oxetanes3 and oxetane-
containing natural products,4,5 and (b) spectroscopic
studies6 leading to the discovery of various reactive
intermediates including exciplexes,7-9 ion-radical pairs,9-13
and 1,4-biradicals.2,11-14 However, the temporal sequence
† Dedicated to Professor George H. Bu¨chi (Aug 1, 1921 to Aug 28,
1998) in memoriam.
(1) (a) Paterno, E.; Chieffi, G. Gazz. Chim. Ital. 1909, 39, 341. (b)
Bu¨chi, G.; Inman, C. G.; Lipinsky, E. S. J . Am. Chem. Soc. 1954, 76,
4327.
(2) For reviews, see: (a) Demuth, M.; Mikail, G. Synthesis 1989,
145. (b) Porco, J . A.; Schreiber, S. L. In Comprehensive Organic
Synthesis; Trost, B. M., Fleming, I., Paquette, L. A., Eds.; Pergamon:
New York, 1991; Vol. 5, p 151. (c) Griesbeck, A. G. In CRC Handbook
of Organic Photochemistry and Photobiology; Horspool, W. M., Song,
P.-S., Eds.; CRC Press: Boca Raton, FL, 1995; p 522. (d) Arnold, D. R.
Adv. Photochem. 1968, 6, 301. (e) Creed, D. In ref 2c, p 737ff. (f) J ones,
G., II. In Organic Photochemistry; Padwa, A., Ed.; Dekker: New York,
1981; Vol. 5, p 1.
(3) (a) Bach, T.; J o¨dicke, K.; Kather, K.; Fro¨hlich, R. J . Am. Chem.
Soc. 1997, 119, 2437. (b) Bach, T.; Eilers, F.; Kather, K. Liebigs Ann.
Recl. 1997, 1529. (c) Bach, T. Liebigs Ann. Recl. 1997, 1627. (d) Bach,
T.; Schro¨der, J . Tetrahedron Lett. 1997, 3701.
(7) (a) Caldwell, R. A.; Hrncir, D. C.; Mun˜oz, T., J r.; Unett, D. J . J .
Am.Chem. Soc. 1996, 118, 8741. (b) Caldwell, R. A.; Sovocool, G. W.;
Gajewski, R. P. J . Am. Chem. Soc. 1973, 95, 2549.
(8) Schore, N. E.; Turro, N. J . J . Am. Chem. Soc. 1975, 97, 2482.
(9) Hu, S.; Neckers, D. C. J . Org. Chem. 1997, 62, 6820.
(10) Gersdorf, J .; Mattay, J .; Go¨rner, H. J . Am. Chem. Soc. 1987,
109, 1203.
(11) (a) Freilich, S. C.; Peters, K. S. J . Am. Chem. Soc. 1985, 107,
3819. (b) Freilich, S. C.; Peters, K. S. J . Am. Chem. Soc. 1981, 103,
6255.
(4) (a) Mattay, J .; Buchkremer, K. Helv. Chim. Acta 1988, 71, 981.
(b) Mattay, J .; Buchkremer, K. Heterocycles 1988, 27, 2153.
(5) (a) Schreiber, S. L. J . Am. Chem. Soc. 1983, 105, 660. (b)
Schreiber, S. L.; Satake, K. J . Am. Chem. Soc. 1984, 106, 4186.
(6) For a review that emphasizes the mechanistic considerations,
see ref 2f.
(12) Griesbeck, A. G.; Mauder, H.; Stadtmu¨ller, S. Acc. Chem. Res.
1994, 27, 70.
(13) Eckert, G.; Goez, M. J . Am. Chem. Soc. 1994, 116, 11999.
(14) (a) Schuster, D. I.; Lem, G.; Kaprinidis, N. A. Chem Rev.
(Washington, D.C.) 1993, 93, 3. (b) Hu, S.-H.; Neckers, D. C. J . Org.
Chem. 1997, 62, 564.
10.1021/jo981754n CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/17/1999

Oxetanes from [2+2] Cycloaddition
J . Org. Chem., Vol. 64, No. 7, 1999 2251
Ch a r t 1
To circumvent these mechanistic ambiguities which are
inherent to carbonyl-activated photocouplings, we now
take a different approach to explore the viability of an
initial electron transfer in Paterno-Bu¨chi reactions.
Thus, the charge-transfer (CT) activation of the electron
donor-acceptor (EDA) complex15,16 between carbonyls
and olefins represents the unambiguous method for
generating ion-radical pairs spontaneously and exclu-
sively as the primary reaction intermediates.17 As such,
we will demonstrate that the ion-radical pair is the
common precursor of the final (oxetane) products and any
other reactive (biradical) intermediate. To this end, we
choose the stilbenes (S) in Chart 1 as prototypical olefinic
donors and chloranil (Q) as the carbonyl acceptor, since
both components are readily involved in the rapid pre-
equilibrium formation of intermolecular EDA complexes,
i.e.,18
The subsequent (dark) coupling of this ion-radical pair
to the 1,4-biradical is then the critical first step toward
oxetane formation.23 Such an unambiguous demonstra-
tion of oxetane formation by photoinduced electron
transfer will then be related to the Paterno-Bu¨chi
cycloadditionsas usually carried out by specific carbonyl
activation24svia the detailed comparison of (a) the oxe-
tane-product mixtures and (b) the ion-pair kinetics for
charge-transfer versus carbonyl activation.25
Resu lts a n d Discu ssion
I. Oxeta n e F or m a tion via Ch a r ge-Tr a n sfer Acti-
va tion of th e Stilben e/Qu in on e EDA Com p lex. Vivid
colors appeared immediately upon the addition of (E)-
stilbene (S) to chloranil (Q) owing to the spontaneous
formation of the intermolecular EDA complex [S, Q] in
eq 2.18 [Note the EDA complex exhibited the character-
istic (charge-transfer) absorption bands in a wavelength
region where neither chloranil nor the stilbenes absorb
(see Figure 1).] Similar spectral changes were observed
with the other (substituted) stilbenes in Chart 1, and the
formation constants KEDA and the extinction coefficients
ꢀCT of the EDA complexes in Table 1 were evaluated by
the quantitative analysis of the spectral data according
to the Benesi-Hildebrand treatment29 (see the Experi-
mental Section).
S + Q {K } [S, Q]EDA
(2)
EDA
In each case, the CT absorption bands can be selec-
tively irradiated at wavelengths λexc > 480 nm19 (where
neither chloranil nor the stilbenes absorb).
Accordingly, our first task in this study is (a) to show
that the specific CT excitation of the EDA complex in eq
2 is a viable photochemical procedure and (b) to establish
the nature of the [2+2] cycloaddition by isolation and
X-ray crystallographic analysis of the crystalline trans-
oxetane photoadducts. Since stilbene cation radical20 and
chloranil anion radical21 both exhibit diagnostic absorp-
tion bands in the UV-vis wavelength region, we exploit
time-resolved (ps) spectroscopy to identify the singlet ion-
radical pair22 1[S•+, Q•-] as the primary reactive inter-
mediate upon photoexcitation of the chloranil/stilbene
EDA complex, i.e.,
Specific excitation of the charge-transfer absorption
bands (see Table 1) of the chloranil complexes with
(21) Andre´, J . J .; Weill, G. Mol. Phys. 1968, 15, 97.
(22) Ojima, S.; Miyasaka, H.; Mataga, N. J . Phys. Chem. 1990, 94,
4147.
(23) (a) According to the electron-transfer mechanism for the
Paterno-Bu¨chi photocyclization, the initial photoinduced electron
transfer to afford the ion-radical pair is followed by a fast (dark)
reaction involving ion-pair collapse to the 1,4-biradical and subsequent
ring closure, i.e.,
kC
hν
[S, Q]EDA
y\
z [S•+, Q•-
]
98 •SQ• f oxetane
hνCT
[S, Q]EDA
8 1[S•+, Q•-
]
(3)
(b) For the observation of 1,4-biradicals from the related benzoquinone
and styrene, see: (c) Maruyama, K.; Otsuki, T.; Tai, S. J . Org. Chem.
1985, 50, 52. (d) Maruyama, K.; Imahori, H. J . Org. Chem. 1989, 54,
2693.
(24) Xu, J .-H.; Wang, L.-C.; Xu, J .-W.; Yan, B.-Z.; Yuan, H.-C. J .
Chem. Soc., Perkin Trans.1 1994, 571.
(15) Foster, R. Organic Charge-Transfer Complexes; Academic
Press: New York, 1969.
(16) (a) Briegleb, G. Elektronen Donator-Acceptor Komplexe;
Springer: Berlin, 1961. (b) Andrew, L. J .; Keefer, R. M. Molecular
Complexes in Organic Chemistry; Holden-Day: San Francisco, 1964.
(17) Charge-transfer excitation of EDA complexes effects the trans-
fer of an electron from the donor to the acceptor in less than 500 fs.
See: Wynne, K.; Galli, C.; Hochstrasser, R. M. J . Chem. Phys. 1994,
100, 4797.
(18) See ref 15, p 40.
(19) The CT absorption band of the stilbene/chloranil EDA complex
is centered at 510 nm. See ref 18.
(25) In this regard, the present study draws upon our earlier
investigation of the photoinduced coupling of a series of diphenylacety-
lene donors with 2,6-dichlorobenzoquinone, also designated as
a
Paterno-Bu¨chi coupling,26 although the putative (highly unstable)
oxetenes28 were merely inferred as intermediates27 (direct observation
notwithstanding). Accordingly, we focus here on stilbene donors more
closely akin to the original Paterno-Bu¨chi substrates from which we
could actually isolate the [2+2] cycloadducts and verify their oxetane
structures by X-ray crystallography.
(20) (a) Lewis, F. D.; Bedell, A. M.; Dykstra, R. E.; Elbert, J . E.;
Gould, I. R.; Farid, S. J . Am. Chem. Soc. 1990, 112, 8055. (b)
J ungmann, H.; Gusten, H.; Schulte-Frohlinde, D. Chem. Rev. 1968,
68, 2690. (c) Shida, T. Electronic Absorption Spectra of Radical Ions;
Elsevier: New York, 1988.
(26) Bosch, E.; Hubig, S. M.; Kochi, J . K. J . Am. Chem. Soc. 1998,
120, 386.
(27) (a) Barltrop, J . A.; Hesp, B. J . Chem. Soc. (C) 1967, 1625. (b)
Farid, S.; Kothe, W.; Pfundt, G. Tetrahedron Lett. 1968, 4147. (c)
Pappas, S. P.; Portnoy, N. A. J . Org. Chem. 1968, 33, 2200.

2252 J . Org. Chem., Vol. 64, No. 7, 1999
Sun et al.
Ta ble 1. CT Absor p tion Ma xim a (λCT), F or m a tion
Con sta n ts (KEDA), a n d Extin ction Coefficien ts (ECT) of th e
EDA Com p lexes of Ch lor a n il a n d Stilben e Don or sa
stilbene
donor
EP
λCT
[nm]
KEDA
ꢀCT
[M-1 cm-1
]
ox
[V vs SCE]
[M-1
]
2
1
5
4
1.59b
1.56b
1.46c
1.44b
507
510
533
540
1.1
1.6
2.6
3.1
592
680
286
189
a
b
In benzene solution at 23 °C. Oxidation (peak) potentials
taken from ref 20a and converted by subtracting 0.39 V from the
values vs Ag/AgI. c Determined by cyclic voltammetry.
Ta ble 2. Ch a r ge-Tr a n sfer -Activa ted [2+2] Cou p lin g of
Stilben es w ith Ch lor a n ila
F igu r e 1. UV-vis spectral changes upon the incremental
addition of stilbene to a benzene solution of 0.005 M chloranil
at [S] ) 0, 0.024, 0.032, 0.043, 0.051, 0.071, and 0.081 M
(bottom-to-top). Inset: Charge-transfer spectrum of the [S, Q]
complex by digital subtraction of chloranil.
various stilbenes was achieved by irradiation at wave-
lengths λexc > 480 nm where neither uncomplexed chlo-
ranil nor the stilbenes absorbed. For example, an equimo-
lar solution of chloranil and (E)-stilbene (1) in dioxane
was exposed to visible light for an extended period of
time. Periodic HPLC analysis of the photolyzate revealed
the simultaneous disappearance of (E)-stilbene and chlo-
ranil and the monotonic appearance of a single product
which was identified as the spirooxetane 1A in Table 2.
Charge-transfer activation of chloranil complexes with
the other stilbenes in Chart 1 yielded a pair of isomeric
oxetane products A and B described in eq 4 (where R1
and R2 represented substituted phenyl groups), i.e.,
Symmetrical stilbenes (R1 ) R2) stereoselectively af-
forded a single spirooxetane isomer. Unsymmetrical
stilbenes generated only two (out of four) possible stereo-
isomers (see eq 4), and the isomer ratios for the oxetane
products are reported in Table 2. In no case were other
isomeric adducts (such as cyclobutanes) found.30 X-ray
crystallographic analysis of the single crystals of 2A, 3B,
and 6B established the trans-configuration of the oxetane
products, as illustrated in Figure 2, and the trans-
configuration was also assigned to the other oxetane
a
In dioxane solution containing 0.05 M chloranil and 0.1 M
stilbene under argon at 25 °C irradiation at λexc > 480 nm.
Identified in Chart 1. c At ca. 2-5% conversion achieved after
b
d
ca. 50 h irradiation. Oxetane isomers obtained only by carbonyl
activation (see Experimental Section).
isomers (4, 5, 7, and 8) in Table 2. No (HPLC) evidence
for the formation of cis-isomers was found. The regio-
chemistry of the coupling resulting in the isomers A and
(28) Oxetenes are short-lived and exhibit lifetimes of up to several
hours only at low temperatures. See: (a) Friedrich, L. E.; Bower, J . D.
J . Am. Chem. Soc. 1973, 95, 6869. (b) Friedrich, L. E.; Lam, P.-S. J .
Org. Chem. 1981, 46, 306. (c) Friedrich, L. E.; Schuster, G. B. J . Am.
Chem. Soc. 1971, 93, 4602.
(29) (a) Benesi, H.; Hildebrand, J . H. J . Am. Chem. Soc. 1949, 71,
2703. (b) Person W. B. J . Am. Chem. Soc. 1965, 87, 167.
1
B in Table 2 was determined by H NMR analysis (see
the Experimental Section).31 Thus, the photocoupling of
chloranil with various substituted stilbenes always re-
sulted in the formation of oxetanes with high (trans)

Oxetanes from [2+2] Cycloaddition
J . Org. Chem., Vol. 64, No. 7, 1999 2253
F igu r e 2. PLUTO perspective of spirooxetanes 2A and 6B showing the trans-configuration and regiochemistry of the cycloadducts.
F igu r e 3. Transient spectrum obtained at 30, 40, 50, and 80 ps (top-to-bottom) upon the application of a 25 ps laser pulse at λCT
) 532 nm to a solution of 0.6 M stilbene and 0.005 M chloranil in dioxane.
stereoselectivity, but indifferent regioselectivity. How-
ever, regardless of the detailed selectivity, the important
result from the above-described photochemical experi-
ments is the fact that the Paterno-Bu¨chi coupling of
various stilbenes with quinone can be successfully achieved
by the specific charge-transfer excitation of their inter-
molecular EDA complexes to yield exclusively30 oxetane
adducts that are structurally verified by X-ray crystal-
lography.
Tr a n sfer Activa tion of Ch lor a n il/Stilben e Com -
p lexes. The successful preparation of oxetane adducts
by CT activation of chloranil/stilbene complexes neces-
sitated our next task to identify the primary reaction
intermediate in this photocoupling process. Accordingly,
the (ground-state) EDA complex of chloranil and stilbene
(1) was selectively excited by the deliberate irradiation
of its charge-transfer absorption band in Figure 1 with
the 25 ps laser pulse at 532 nm (from a mode-locked Nd:
YAG laser). First, a solution of chloranil (0.005 M) and
(E)-stilbene (0.6 M) in dioxane was exposed to the 532
nm laser pulse, and the resulting time-resolved (ps)
absorption spectra are shown in Figure 3. The principal
spectral feature that was immediately apparent upon the
laser excitation consisted of a strong absorption band at
λmax ) 480 nm and a weak, broad band around 760 nm,
which coincided with the composite absorptions of the
II. Dir ect Obser va tion of th e Ion -Ra d ica l P a ir
[S•+, Q•-] as th e P r im ar y In ter m ediate in th e Ch ar ge-
(30) (a) By contrast, C-C cycloadditions of photoactivated quinones
(leading to cyclobutane products) are observed with arylethenes such
as styrenes and 1,1-diphenylethenes,30b and in photocycloadditions of
either stilbene to the parent benzoquinone30c or phenylacetylenes to
methoxy-substituted quinones.30d (b) Covell, C.; Gilbert, A.; Richter,
C. J . Chem. Res. (S) 1998, 316. (c) Gilbert, A.; Kamonnawin, P.
Unpublished results (reported in ref 30b). (d) Pappas, S. P.; Pappas,
B. C.; Portnoy, N. A. J . Org. Chem. 1969, 34, 520.
(31) The regiochemistry of the coupling could not be established by
GC/MS analysis owing to complete decomposition of the spirooxetanes
(to the starting materials) upon their injection into the gas chromato-
graph.
chloranil anion radical (Q•- 21
) and the (E)-stilbene cation
radical (S•+).20 In other words, both ion radicals (Q•-) and
(S•+) were generated spontaneously within the 25 ps
laser-pulse duration as depicted in eq 3. Subsequently,

2254 J . Org. Chem., Vol. 64, No. 7, 1999
Sun et al.
Sch em e 1
for oxetane formation via CT activation were determined
using the 546 nm output of a mercury lamp as the
irradiation source and a potassium reineckate solution34
as the actinometer (see the Experimental Section).
Although the resulting quantum efficiencies listed in
Table 3 show that the photocoupling of chloranil and
stilbene by CT activation is rather inefficient, there is a
clear trend in ΦCT with the substitution pattern of the
stilbene, the highest yield (ΦCT ) 0.0012) being achieved
with the 4,4′-dichloro-substituted stilbene (7, see Table
3).
The low quantum efficiencies (ΦCT) of oxetane forma-
tion can be traced to a highly efficient back electron
transfer of the singlet ion-radical pair in Scheme 1.33 If
this pathway occurs with rate constants that are orders
of magnitude faster than the coupling, the changes in
the quantum yields ΦCT in Table 3 can be ascribed to
variations in back electron transfer as predicted by
Marcus theory for donors with different oxidation poten-
tials35 in Scheme 1.
According to Scheme 1, the efficiency (ΦCT) of oxetane
formation is a direct measure of the competition between
coupling (kC) and back electron transfer (kBET), i.e., ΦCT
) kC/ (kC + kBET). Accordingly, by taking ΦCT ) 3 × 10-4
(from Table 3) and (kC + kBET) ) 3 × 1011 s-1 (vide supra),
we calculate the rate constant of coupling between Q•-
and S•+ within the singlet ion-radical pair to be kC = 9 ×
107 s-1 (in dioxane). In other words, the low quantum
efficiencies for oxetane formation via CT activation of the
chloranil/stilbene EDA complexes are the direct result
of a very rapid back electron transfer (kBET = 3 × 1011
s-1), which is known to be characteristic for singlet ion-
radical pairs.33 Thus, the CT activation methodology
introduced in this study successfully establishes the
viability of the electron-transfer pathway for the [2+2]
cycloaddition of chloranil with stilbene. However, it is not
a synthetically relevant procedure (owing to its low
quantum efficiency), and we now turn to the commonly
applied irradiation methodology for Paterno-Bu¨chi reac-
tions and focus on the results of oxetane formation by
carbonyl (quinone) activation.
IV. Oxeta n e F or m a tion via Ca r bon yl (Qu in on e)
Activa tion . Selective photoactivation of the chloranil
was achieved by irradiation at wavelengths λexc > 370
nm where the stilbenes do not absorb.36 For example, an
equimolar (0.1 M) solution of (E)-stilbene (1) and chlo-
ranil was irradiated in dioxane until periodic HPLC
analysis revealed the complete consumption of the chlo-
ranil. Oxetane 1A was isolated as the single photoproduct
and shown to be identical to that obtained by charge-
transfer activation of the corresponding EDA complex
(vide supra). Most importantly, the carbonyl-activated
cycloadditions of the other (substituted) stilbenes in
Chart 1 consistently led to the same isomeric oxetanes
the quinone and the stilbene absorption bands decayed
simultaneously within the 25 ps pulse with decay rate
constants of k g 4 × 1010 s-1 (see Figure 3). Similar time-
resolved (ps) absorption spectra were obtained in aceto-
nitrile and benzene solution and also upon 532 nm
excitation of the chloranil complexes with the other
(substituted) stilbenes. In all cases, the decay rate
constants exceeded k > 4 × 1010 s-1, which corresponded
to the time resolution of the 25 ps laser spectrometer.
The chloranil/stilbene complex was also irradiated with
the 200 fs laser pulse at 406 nm of a Ti:sapphire laser to
obtain a more accurate rate constant for the ultrafast
decay of the ion radicals. Despite the fact that at λexc
)
406 nm chloranil also absorbed the laser light to some
extent, we clearly observed initially a strong transient
absorption band at λmax ) 480 nm with a shoulder at 450
nm which was readily ascribed to the stilbene cation
radical and the chloranil anion radical, respectively. This
initial transient spectrum subsequently decayed on the
early picosecond time scale, and a first-order decay rate
constant of k ) 3 × 1011 s-1 was determined from the
exponential fit of the absorbance/time profile.
The ultrafast generation and decay of the chloranil
anion radical and the stilbene cation radical upon laser
excitation is in accord with Mulliken theory, which
predicts a spontaneous and complete electron transfer
from the stilbene donor to the quinone acceptor upon CT
excitation of their mutual EDA complex.32 As a result, a
1
singlet ion-radical pair [S•+, Q•-] is formed22 (see Figure
3) which is identified by its ultrashort (τ = 3 ps) lifetime.
This very first intermediate is formed exclusively in the
photoexcitation of the EDA complex, and it is therefore
the precursor to the oxetane coupling products in Table 2
(which are preparatively obtained by essentially the same
irradiation method). In other words, the successful
preparation of the oxetane coupling products by charge-
transfer activation of the chloranil/stilbene complexes
combined with the unambiguous identification of ion-
radical pairs as the primary reactive intermediates
provides direct (irrefutable) evidence for our conclusion
that this typical Paterno-Bu¨chi coupling of chloranil and
stilbene occurs via an initial electron-transfer step. The
ion-radical pair is thus the critical intermediate which
undergoes the coupling ultimately to the oxetane product
via the formation of an intermediate 1,4-biradical.23
However, besides the coupling step (kC) the singlet ion
pair can also undergo back electron transfer33 (kBET) to
restore the original EDA complex (Scheme 1).
In fact, the latter pathway is known to be very efficient
in singlet ion-radical pairs,33 and thus the competition
between rapid back electron transfer and coupling in
Scheme 1 will limit the quantum efficiency of oxetane
formation significantly, as reported in the following
section.
(34) Wegner, E. E.; Adamson, A. W. J . Am. Chem. Soc. 1966, 88,
394. See also: Bunce, N. J . In Handbook of Organic Photochemistry;
Scaiano, J . C., Ed.; Vol. 1: Chapter 9, p 241.
(35) With increasing oxidation potentials of the stilbene, the back
electron transfer from stilbene cation radical to chloranil anion radical
becomes more exergonic, and thus its rate constant will decrease in
the Marcus-inverted region. See: (a) Marcus, R. A. Annu. Rev. Phys.
Chem. 1964, 15, 155. (b) Marcus, R. A. J . Chem. Phys. 1965, 43, 679.
(c) Marcus, R. A. Faraday Discuss. Chem. Soc. 1982, 74, 7. (d) Marcus,
R. A. J . Chem. Phys. 1956, 24, 966.
(36) Note that at these concentrations, the absorption of actinic light
by the chloranil/stilbene complex was negligible compared to that of
the chloranil component owing to the limited values of KEDA and ꢀCT
in Table 1.
III. Qu an tu m Efficien cies for th e Ch ar ge-Tr an sfer -
Activa ted Oxeta n e F or m a tion . The quantum yields
(32) (a) Mulliken, R. S. J . Am. Chem. Soc. 1950, 72, 600. (b)
Mulliken, R. S. J . Am. Chem. Soc. 1952, 74, 811.
(33) (a) Asahi, T.; Mataga, N. J . Phys. Chem. 1989, 93, 6575. (b)
Asahi, T.; Mataga, N. J . Phys. Chem. 1991, 95, 1956. (c) Peters, K. S.
Adv. Electron-transfer Chem. 1994, 4, 27. (d) Hilinski, E. F.; Masnovi,
J . M.; Kochi, J . K.; Rentzepis, P. M. J . Am. Chem. Soc. 1984, 106, 8071.
(e) Fox, M. A. Adv. Photochem. 1986, 13, 237.29.

Oxetanes from [2+2] Cycloaddition
J . Org. Chem., Vol. 64, No. 7, 1999 2255
Ta ble 3. Com p a r ison of CT-Activa ted a n d Ca r bon yl-Activa ted P h otocou p lin g of Stilben es w ith Ch lor a n il
d
e
stilbene
donor
CT activationa
ΦCT
carbonyl activationb
ΦQ
oxetane
ratioc (A:B)
[(0.0001]
oxetane
ratioc (A:B)
[(0.01]
1
2
3
4
5
6
7
8
1A
2A, 2B
3A, 3B
0.0003
0.0005
0.0007
1A
0.34
0.35
0.33
0.03
0.11
0.29
0.48
0.34
45:55
45:55
no reaction
no reaction
40:60
2A, 2B
3A, 3B
4A, 4B
5A, 5B
6A, 6B
7A
46:54
47:53
40:60
46:54
41:59
6A, 6B
7A
8A
0.0004
0.0012
0.0011
8A
a
b
Solution of 0.05 M chloranil and 0.1 M (E)-stilbene in dioxane irradiated at λexc ) 546 nm or λexc ) 436 nm. c The composition of the
photolysate analyzed by HPLC with biphenyl as an internal standard. Quantum yield by Reineckate salt actinometry. e Quantum yield
d
by ferrioxalate actinometry.
Sch em e 2
back-electron transfer to the (singlet) ground-state rep-
resents a rather slow (spin-forbidden) process. As a
consequence, the coupling process (kC) in Scheme 2 can
better compete with the back-electron transfer and leads
to the improved quantum efficiencies of oxetane forma-
tion in Table 3. In other words, the substantial difference
in the quantum efficiencies of carbonyl-activated and CT-
activated oxetane formation is readily accommodated by
a mere consideration of the ion-pair spin multiplicities
in an otherwise identical reaction sequence.
V. Dir ect Obser va tion of Tr ip let Ion -Ra d ica l
P a ir s u p on La ser Excita tion of Ch lor a n il in th e
P r esen ce of Stilben e. To directly observe the triplet
ion-radical pairs as the result of electron-transfer quench-
ing of triplet quinone, time-resolved absorption measure-
ments were carried out on the picosecond and early
nanosecond time scales as follows: A solution of chloranil
(0.008 M) and stilbene (0.1 M) was excited with the 200
fs laser pulse of a Ti:sapphire laser tuned at 406 nm. At
50 ps after laser excitation, a transient spectrum was
obtained which clearly showed the characteristic 510 nm
absorption of triplet chloranil.38a,42 This absorption de-
cayed over a time span of about 1.5 ns, and the concomi-
tant growth of the absorptions of chloranil anion radical21
and stilbene cation radical20 was observed at λmax ) 450,
480, and 760 nm (vide supra). Kinetics traces obtained
at all three wavelengths could be fitted to first-order
as those obtained by CT activation (with identical molar
ratios in Table 3).
To establish the photochemical efficiency of the quinone-
activated cycloaddition, a 0.05 M solution of chloranil
containing 0.01 M stilbene was irradiated with mono-
chromatic light at λexc ) 436 nm, and quantitative
measures of the light absorption were based on ferri-
oxalate actinometry.37 The results in Table 3 show that
the quantum efficiencies ΦQ exhibited the same trend
with the substitution pattern of the stilbenes as those
observed upon CT activation; that is, the highest ef-
ficiency (ΦQ ) 0.48) was obtained with the 4,4′-dichloro-
stilbene (7). Carbonyl activation resulted in a much more
efficient formation of oxetane relative to the correspond-
ing charge-transfer activation (compare columns 4 and
7 in Table 3). The difference can be explained by
considering the spin multiplicities of the two reaction
sequences in the following way.
Direct excitation of chloranil generates its triplet state
with unit efficiency owing to the ultrafast rate of inter-
system crossing.38 The sizable energy of triplet quinone
with ET = 50 kcal mol-1 39 is sufficient to effect the ready
oxidation of aromatic donors such as stilbenes at diffu-
sion-limited rates40,41 and to result in the formation of
triplet ion-radical pairs (Scheme 2).
kinetics with a single rate constant of k ) 2.5 × 109 s-1
.
The spectral changes were thus ascribed to the diffu-
sional electron-transfer quenching of excited chloranil by
stilbene with a second-order rate constant of kq ) k/[S]
) 2.5 × 1010 M-1 s-1. In other words, triplet chloranil
was quenched by stilbene at a diffusion-limited rate,43
in accord with the highly exergonic electron transfer.44
The resulting ion-radical absorptions subsequently de-
cayed by first-order kinetics with a relatively long lifetime
of τ ≈ 5 ns, and they were thus assigned to the triplet
ion-radical pair.
Let us now discuss the principal decay pathways of the
triplet ion-radical pair shown in Scheme 2: Since ionic
dissociation is unfavored in the nonpolar dioxane (kdiss
= 0), the decay of the ion pair can be entirely ascribed to
back-electron transfer (kBET) or coupling (kC) which
ultimately leads to the oxetane products. Thus, the
competition between these predominant decay routes
determines the quantum yields (ΦQ) of oxetane formation
Triplet ion-radical pairs exhibit rather long lifetimes
in comparison to the corresponding singlet ion pairs, since
(37) Hatchard, C. G.; Parker, C. A. Proc. R. Soc. 1956, 235A, 518.
See also: Calvert, J . G.; Pitts, J . N., J r. Photochemistry; Wiley: New
York, 1996; p 786.
(38) (a) Rathore, R.; Hubig, S. M.; Kochi, J . K. J . Am. Chem. Soc.
1997, 119, 11468, and references therein. (b) Hubig, S. M.; Bockman,
T. M.; Kochi, J . K. J . Am. Chem. Soc. 1997, 119, 2926.
(39) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photo-
chemistry, 2nd ed.; Dekker: New York, 1993.
(40) (a) Gschwind, R.; Haselbach, E. Helv. Chim. Acta 1979, 62, 941.
(b) J ohnson, L. J .; Schepp, N. P. J . Am. Chem. Soc. 1993, 115, 6564.
(c) Levin, P. P.; Kuz’min, V. A. Russ. Chem. Rev. 1987, 56, 307.
(41) (a) Levin, P. P.; Kuzmin, V. A. Bull. Russ. Acad. Div. Sci. Chem.
Sci. 1992, 41, 451. (b) Baciocchi, E.; Del Giacco, T.; Elisei, F.; Ioele, M.
J . Org. Chem. 1995, 60, 7974.
(42) On the early picosecond time scale, the singlet ion-radical pair
is observed due to CT excitation of the chloranil/stilbene EDA complex,
which also absorbed at 406 nm (vide supra).
(43) Diffusion-controlled reactions exhibit second-order rate con-
stants of k = 1010 M-1 s-1. See: Moore, J . W.; Pearson, R. G. Kinetics
and Mechanism; Wiley: New York, 1981; p 239.
(44) Rehm, D.; Weller, A. Isr. J . Chem. 1970, 8, 259.

2256 J . Org. Chem., Vol. 64, No. 7, 1999
Sun et al.
the microdynamics for ion-pair coupling to be rather
independent of the spin state, with a relatively invariant
value of kC ∼ 1 × 108 s-1 for both processes. As such, the
difference between CT and carbonyl activation lies solely
in the competitive rates of back electron transfer in
singlet and triplet ion-radical pairs with kBET ) 3 × 1011
and 1 × 108 s-1, respectively. Since the oxetane products
are derived only as a result of ion-pair coupling, it follows
from the identical values of kC that carbonyl activation
of the Paterno-Bu¨chi reaction must proceed via an ion-
radical pair that is essentially indistinguishable from
that derived via photoinduced electron transfer within
the intermolecular [S, Q] complex. It is noteworthy that
such a conclusion (based on kinetics) coincides with that
deduced from oxetane analysis (based on regioisomers)48
to confirm the generality of the electron-transfer mech-
anism.23
Su m m a r y a n d Con clu sion s
The Paterno-Bu¨chi coupling of various stilbenes (S)
with chloranil (Q) to yield trans-oxetane(s) is effectively
achieved by the specific charge-transfer photoactivation
of the electron donor-acceptor complexes [S, Q]. Time-
resolved (ps) spectroscopy reveals the singlet ion-radical
F igu r e 4. Spectral changes at 370, 430, 500, 620, 800, and
1000 ns following the application of a 10 ns laser pulse at λexc
) 355 nm to a solution of 0.0001 M stilbene and 0.004 M
chloranil in acetonitrile.
1
pair [S•+, Q•-] to be generated exclusively as the primary
reactive intermediate and thus establishes its unambigu-
ous role as the direct precursor of the oxetane and all
the (coupled) intermediates.23 Quantitative analysis of
in Table 3, i.e., ΦQ ) kC/(kC + kBET). Accordingly, the
(maximum) quantum yield of ΦQ ) 0.48 for stilbene 7 in
dioxane is the result of about equal rates of back-electron
transfer and coupling (i.e., kC ≈ kBET ) 1 × 108 s-1) in
the best case.
1
the singlet [S•+, Q•-] decay leads to kC ) 9 × 107 s-1 for
ion-pair collapse to the 1,4-biradical •SQ• and kBET ) 3 ×
1011 s-1 for back electron transfer to regenerate the initial
EDA complex. For comparison, carbonyl (quinone) acti-
vation (as usually employed in Paterno-Bu¨chi couplings)
leads to the same oxetane products (with identical isomer
ratios). Thus, an analogous mechanism is applied which
includes an initial electron-transfer quenching of the
photoactivated (triplet) quinone acceptor by the stilbene
donors resulting in triplet ion-radical pairs. Quantitative
In a polar solvent such as acetonitrile, the diffusional
quenching of excited chloranil (0.004 M) by stilbene
(0.0001 M) also resulted in the formation of ion radicals
as observed by time-resolved (ns/µs) spectroscopy (see
Figure 4). However, the ion radicals exhibited much
longer (µs) lifetimes as compared to those in dioxane, and
they slowly decayed by second-order kinetics (not shown
in Figure 4). We thus conclude that the ion radicals in
acetonitrile are not ion-paired, but exist as free, solvated
ions which recombine upon diffusional encounter on the
µs time scale. The observation of free ion radicals in
acetonitrile is due to efficient ion dissociation (kdiss in
Scheme 2), which has been previously established to
3
analysis of the triplet [S+•, Q-•] decay leads to kC ) 1 ×
108 s-1 for ion-pair collapse to the biradical and kBET ) 1
× 108 s-1 for the (spin-forbidden) back electron transfer.
The substantial differences in the quantum efficiencies
of CT-activated versus carbonyl-activated oxetane forma-
tion are thus readily explained by a simple consideration
of the difference in the spin multiplicity (i.e., singlet
versus triplet) in the ion-radical pairs in otherwise
identical mechanisms of ion-pair microdynamics. As such,
the electron-transfer mechanism established here (for the
first time) for a typical Paterno-Bu¨chi coupling of
chloranil and stilbene forms the basis for the quantitative
evaluation of various solvent and salt effects, which will
be reported separately.47
occur with rate constants of kdiss = 109 s-1 45,46
Such fast
.
ionic dissociation completely outruns the much slower
(spin-forbidden) back electron transfer as revealed by the
free-ion yields close to unity observed in acetonitrile,38a,40a
and it also renders the rather slow coupling process (kC)
completely suppressed (ΦQ ) 0) in acetonitrile.47
VI. Mech a n istic Im p lica tion s of th e Micr od yn a m -
ics for Sin glet a n d Tr ip let Ion -Ra d ica l P a ir s. Time-
resolved spectroscopy in two widely separated (ps and
ns) time regimes establishes the singlet and triplet ion-
Exp er im en ta l Section
1
3
radical pairs, i.e., [S•+, Q•-] and [S•+, Q•-], as the first
intermediates in charge-transfer activation and carbonyl
activation, respectively, of the stilbene/quinone pairs for
the Paterno-Bu¨chi photocycloadditions. Most strikingly,
the quantitative analysis of the decay kinetics reveals
Ma ter ia ls a n d Meth od s. Chloranil was sublimed in vacuo
and recrystallized from benzene. (E)-stilbene from Aldrich was
used as received. The chloro- and methyl-substituted stilbenes
2-8 were prepared via the Wittig coupling of the correspond-
(45) Knibbe, H.; Rehm, D.; Weller, A. Ber. Bunsen-Ges. Phys. Chem.
1968, 72, 257.
(46) Ojima, S.; Miyasaka, H.; Mataga, N. J . Phys. Chem. 1990, 94,
7534.
(47) A detailed study of the solvent effects on the back electron
transfer, the ion dissociation, and the coupling processes will be
published later.
(48) Regiochemistry in the electron-transfer mechanism23 is deter-
mined by the relative orientation of the ion-radical pair, particularly
as it is affected by the charge distribution. [As such, the spin
multiplicity of the singlet or triplet biradical state does not appear to
be a distinguishing factor.] Charge annihilation in the triplet ion-
radical pair may proceed either directly to the 1,4-biradical or via spin
inversion followed by ion-pair collapse to the singlet 1,4-biradical.

Oxetanes from [2+2] Cycloaddition
J . Org. Chem., Vol. 64, No. 7, 1999 2257
1
ing ylides and aldehydes.49 The trans isomers were isolated
by flash chromatography, and HPLC analyses showed that
they were not contaminated by the other (cis) isomers. Aceto-
nitrile and dioxane were purified according to published
procedures.50 Melting points are uncorrected. 1H NMR and 13C
NMR were recorded in CDCl3 on a 300 MHz NMR spectrom-
eter. UV-vis absorption and infrared spectra were recorded
separated from isomer 3B; H NMR (CDCl3) δ, 4.96 (d, 1H, J
9.0), 6.63 (d, 1H J 8.7), 7.39-7.65 (m, 8Η).
5,6,8,9-Tetr a ch lor o-2-(3,4-d ich lor op h en yl)-3-p h en yl-1-
oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 3B: mp 155-157 °C
(decomp); IR (cm-1) 1676, 1608, 1573, 1497, 1476, 1447, 1407,
1283, 1220, 1144, 1120, 1099, 1035, 980, 959, 919, 896, 872,
851, 835, 822, 780, 759, 743, 730, 714, 698, 679, 650, 598, 579,
561, 537, 476, 460, 445; 1H NMR (CDCl3) δ 4.92 (d, 1H, J 9.0),
on
a diode-array spectrometer and a Fourier transform
spectrometer, respectively. GC-MS analyses were carried out
on gas chromatograph interfaced to a mass spectrometer (EI,
70 eV). HPLC analyses were performed using a Hypersil BDS
C18 reverse-phase column (20 cm) with methanol/water
mixtures as eluent. All chemical analyses were carried out by
Atlantic Microlab Inc., Norcross, GA. The oxidation (peak)
potentials of the trans-stilbenes were determined by cyclic
voltammetry.
6.66 (d, 1H J 9.0), 7.39-7.65 (m, 8H). Anal. Calcd for C20H10-
Cl6O2: C, 48.51; H, 2.02. Found: C, 48.41; H, 2.01.
5,6,8,9-Te t r a ch lor o-2-(4-m e t h ylp h e n yl)-3-p h e n yl-1-
oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 4A: Could not be fully
separated from isomer 4B; 1H NMR (CDCl3) δ 2.41 (s, 3H),
5.02 (d, 1H, J 9.0), 6.71 (d, 1H J 8.7), 7.00-7.60 (m, 9H).
5,6,8,9-Tetr a ch lor o-2-(3,4-d ich lor op h en yl)-3-p h en yl-1-
oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 4B: Could not be fully
separated from isomer 4A; 1H NMR (CDCl3) δ 2.36 (s, 3H),
4.95 (d, 1H, J 9.0), 6.73 (d, 1H J 8.7), 7.00-7.60 (m, 9H).
5,6,8,9-Tetrachloro-2-(4-methylphenyl)-3-(4-cholorophen-
yl-1-oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 5A: Could not be
fully separated from isomer 5B; 1H NMR (CDCl3) δ 2.42 (s,
3H), 4.97 (d, 1H, J 9.0), 6.68, 2.42 (s, 3H), 4.97 (d, 1H, J 9.0),
6.68 (d, 1H J 8.7), 6.99-7.52 (m, 8H).
P h otoin d u ced Cou p lin g of Ch lor a n il w ith Stilben es.
Gen er a l P r oced u r e for th e P r ep a r a tive P h otolysis. A
solution of chloranil (1 mmol, 0.05 M) and stilbene51 (2 mmol,
0.1 M) in benzene was prepared under an argon atmosphere
and irradiated with a focused beam from a medium-pressure
mercury lamp (500 W) passed through an aqueous IR filter
and a 370 nm sharp cutoff filter. This ensured that only the
quinone (and not the stilbene) absorbed the actinic light. The
photoreaction was carried out until periodic HPLC analysis
showed that all chloranil was consumed. The solvent was
evaporated, and in the case of unsymmetrical stilbenes, the
relative ratio of the two regioisomers of spirooxetanes was
5,6,8,9-Tetr ach lor o-2-(4-ch lor oph en yl)-3-(4-m eth ylph en -
yl)-1-oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 5B: mp 132-134
°C (decomp); IR (cm-1) 1675, 1601, 1570, 1493, 1448, 1408,
1302, 1286, 1217, 1142, 1118, 1089, 1034, 980, 958, 917, 895,
871, 854, 834, 812, 780, 759, 744, 730, 714, 697, 679, 649, 597,
1
1
determined by H NMR.31 For example, the chemical shift of
579, 561, 536, 476, 461, 454; H NMR (CDCl3) δ 2.36 (s, 3H),
4.89 (d, 1H, J 9.0), 6.67 (d, 1H J 8.7), 6.99-7.52 (m, 8H). Anal.
Calcd for C21H13Cl5O2: C, 53.14; H, 2.74. Found: C, 53.14; H,
2.83.
5,6,8,9-Tet r a ch lor o-2-(4-m et h ylp h en yl)-3-(3,4-d ich lo-
r op h en yl)-1-oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 6A: Could
not be fully separated from isomer 6B; 1H NMR (CDCl3) δ 2.40
(s, 3H), 4.96 (d, 1H, J 8.7), 6.58 (d, 1H J 9.0), 6.92-7.61 (m,
7H).
the R proton in 6A (δ 6.58) was slightly shifted upfield as
compared to that in isomer 6B (δ 6.63), and the â proton in
6A was slightly shifted downfield (δ 4.96) compared to the
isomer 6B (δ 4.83), as confirmed in the X-ray structure
analyses. The crude product was washed with petroleum ether
and recrystallized from chloroform/petroleum ether. The spiro-
oxetane (containing two regioisomers for the unsymmetrical
stilbenes) was isolated in high yield (81-87%) based on the
chloranil conversion. All attempts to separate the regioisomers
by preparative TLC (alumina or silica gel) with numerous
solvent combinations and ratios (including ethyl acetate, ethyl
ether, dichloromethane, hexane, benzene, and petroleum ether)
remained unsuccessful. However, some of the isomers (2A, 3B,
5B, and 6B) could be separated from the isomeric mixture by
stepwise crystallization from chloroform/petroleum ether, and
the structures of 2A, 3B, and 6B were established by X-ray
crystallography.52 Characteristic physical data for the spiro-
oxetane products are as follows:
5,6,8,9-Tetr a ch lor o-2,3-d ip h en yl-1-oxa sp ir o[3,5]n on a -
5,8-d ien -7-on e, 1A: mp 150-151 °C (decomp) (lit. 150 °C);24
IR (cm-1) 1686, 1606, 1575, 1497, 1127, 1102, 968, 826, 773,
753, 742, 733, 723, 698, 682, 651, 551; 1H NMR (CDCl3) δ, 4.98
(d, 1H, J 8.7), 6.72 (d, 1H, J 9.0), 7.30-7.60 (m, 10H).
5,6,8,9-Te t r a ch lor o-2-(4-ch lor op h e n yl)-3-p h e n yl-1-
oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 2A: mp 143-144 °C (lit.
144-146 °C);24 IR (cm-1) 1671, 1600, 1570, 1490, 1279, 1120,
1101, 982, 956, 810, 784, 735, 724, 690, 682, 650, 553; 1H NMR
(CDCl3) δ, 4.95 (d, 1H, J 8.7), 6.68 (d, 1H, J 8.7), 7.27-7.58
(m, 9H).
5,6,8,9-Tet r a ch lor o-2-(3,4-d ich lor op h en yl)-3-(4-m et h -
ylph en yl)-1-oxaspir o[3,5]n on a-5,8-dien -7-on e, 6B: mp 141-
143 °C (decomp); IR (cm-1) 1674, 1601, 1571, 1493, 1449, 1407,
1301, 1216, 1140, 1119, 1088, 978, 960, 918, 896, 871, 780,
1
759, 742, 730, 697, 678, 649, 596, 578, 561, 536, 476, 461; H
NMR (CDCl3) δ 2.34 (s, 3H), 4.83 (d, 1H, J 7.8), 6.63 (d, 1H J
8.1), 7.30-7.60 (m, 7H). Anal. Calcd for C21H12Cl6O2: C, 49.54;
H, 2.36. Found: C, 49.52; H, 2.41.
5,6,8,9-Tetr ach lor o-2-(4-ch lor oph en yl)-3-(4-ch lor oph en -
yl)-1-oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 7A: mp 162-164
°C (decomp); IR (cm-1): 1682, 1603, 1568, 1494, 1119, 1106,
1100, 1012, 971, 955, 941, 914, 884, 829, 758, 745, 731, 523;
1H NMR (CDCl3) δ, 4.91 (d, 1H, J 8.7), 6.65 (d, 1H J 8.7), 7.04
(d, 2H, J 8.4), 7.37 (d, 2H, J 8.7), 7.43-7.57 (m, 4H). Anal.
Calcd for C20H10Cl6O2: C, 48.51; H, 2.02. Found: C, 48.57; H,
2.05.
5,6,8,9-Tetr ach lor o-2-(3,4-dich lor oph en yl)-3-(3,4-dich lo-
r op h en yl)-1-oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 8A: mp
169-171 °C (decomp); IR (cm-1) 1690, 1575, 1475, 1449, 1386,
1363, 1127, 1103, 1028, 983, 974, 8221, 769, 754, 733, 666,
650, 542; 1H NMR (CDCl3) δ 4.88 (d, 1H, J 8.7), 6.57 (d, 1H J
8.7), 6.87 (d, 1H, J 8.4), 7.34-7.38 (m, 2H), 7.47 (d, 1H J 8.4),
7.56 (d, 1H, J 8.4), 7.62 (d, 1H, J 1.8). Anal. Calcd for C20H8-
Cl8O2: C, 42.58; H, 1.42. Found: C, 42.51; H, 1.47.
Qu a n tu m Yield for th e P h otocou p lin g of Ch lor a n il
w ith Stilben es. Excita tion a t λexc > 480 n m . The quantum
yields were measured with the aid of a medium-pressure (500
W) mercury lamp focused through an aqueous IR filter
followed by a 480 nm cutoff filter. The intensity of the lamp
was determined with a freshly prepared potassium reineckate
solution.34 Equimolar (0.1 M) solutions of chloranil and (E)-
stilbene were irradiated, and the photoconversion was moni-
tored by HPLC. The product formation was quantified with
biphenyl as internal standard. Since the actinic light was not
completely absorbed by the reaction mixture, a correction for
transmitted light was made.
5,6,8,9- Tetr a ch lor o-2-p h en yl-3-(4-ch lor op h en yl)-1-oxa -
sp ir o[3,5]n on a -5,8-d ien -7-on e, 2B: Could not be fully sepa-
rated from isomer 2A; 1H NMR (CDCl3) δ, 4.93 (d, 1H, J 8.7),
6.66 (d, 1H J 8.7), 7.27-7.58 (m, 9H).
5,6,8,9-Tetr a ch lor o-2-(3,4-d ich lor op h en yl)-3-p h en yl-1-
oxa sp ir o[3,5]n on a -5,8-d ien -7-on e, 3A: Could not be fully
(49) McDonald, R. N.; Campbell, T. W. In Organic Syntheses;
Wiley: New York, 1973; Vol. 5; p 499.
(50) Perrin, D. D.; Armarego, W. L. F.; Perrin, D. R. Purification of
Laboratory Chemicals, 2nd ed.; Pergamon: New York, 1980.
(51) Since (Z)- and (E)-isomers generally yield the same trans-
spirooxetane(s),24 the isomeric mixture (E:Z 65-70:30-35) of the
stilbenes was used for preparative photocouplings. Quantum yields are
determined with pure (E)-isomers as starting materials, and the
photoreactivity of (Z)-stilbene is described separately.47
Excita tion a t λexc ) 436 n m . The quantum yields were
measured with the same lamp focused through an aqueous
(52) On deposit with the Cambridge Crystallographic Data Center,
12 Union Road, Cambridge, CB12 1EZ, U.K.

2258 J . Org. Chem., Vol. 64, No. 7, 1999
Sun et al.
IR filter followed by an aqueous NaNO2/CuSO4 solution filter
with a narrow-band-pass at 440 ( 30 nm. The intensity of the
lamp was determined with a freshly prepared potassium
ferrioxalate actinometer solution.37 The absorbance at 436 nm
of the solutions of chloranil (0.05 M) and (E)-stilbenes (0.1 M)
in dioxane remained above 1.5 throughout the irradiation, and
thus no correction for transmitted light was necessary.
Ch a r ge-Tr a n sfer Absor p tion Sp ectr a of EDA Com -
p lexes of Ch lor a n il w ith Stilben e Don or s. Ben esi-
Hild ebr a n d Eva lu a tion . The vivid colors that were observed
upon addition of stilbenes to a chloranil solution persisted
indefinitely when the solutions were kept in the dark, and
most importantly, the components S and Q could be quanti-
tatively recovered intact from mixtures that were stored for
several days at room temperature in the dark. The UV-vis
spectral changes shown in Figure 1 demonstrate that the
colored solutions resulted from a new (additional) absorption
band with λmax = 510 nm, the intensity of which increased with
incremental additions of stilbene. Such color changes are
diagnostic of the preequilibrium formation of electron donor-
acceptor (EDA) complexes (see eq 2),15,16 and they are ascribed
time-resolved absorption measurements were carried out with
a kinetic spectrometer including a Q-switched Nd3+:YAG laser
(10 ns pulse width).53 The third (355 nm) harmonic output was
used for the excitation of chloranil. The solutions of chloranil
and (E)-stilbenes were prepared under an argon atmosphere
in a 1 cm cuvette fitted with a Schlenk adapter. The concen-
trations of the components were adjusted for absorbances in
the range 0.5-0.8 at the excitation wavelength of λexc ) 355
nm.
Ch a r ge-Tr a n sfer Excita tion a t λexc ) 532 n m . The
picosecond transient absorption measurements following charge-
transfer excitation of the EDA complex between chloranil and
stilbene were performed with a mode-locked Nd3+:YAG laser
using the second-harmonic output at 532 nm (17 mJ per
pulse).53
La ser Excita tion a t 406 n m . To monitor the quenching
of photoexcited quinone by stilbene in dioxane on the picosec-
ond time scale, chloranil was excited at 406 nm (to avoid the
unintentional photoexcitation of stilbene). The excitation at
406 nm was achieved with a Ti:sapphire laser system consist-
ing of a Ti:sapphire oscillator coupled to an argon-ion laser
and two consecutive Ti:sapphire amplifiers pumped by a Nd:
YAG laser at 10 Hz.54
Electr och em ica l Mea su r em en ts. The cyclic voltammetry
(CV) measurements were carried out with a cell that was of
an airtight design with high-vacuum Teflon valves and Viton
O-ring seals to allow an inert atmosphere to be maintained
without contamination by grease. The working electrode
consisted of an adjustable platinum disk embedded in a glass
seal to allow periodic polishing (with a fine emery cloth)
without changing the surface area (of about 1 mm2) signifi-
cantly. The SCE reference electrode and the associated salt
bridge were separated from the catholyte by a sintered glass
frit. The counter electrode consisted of a platinum gauze that
was placed about 3 mm from the working electrode. The CV
measurements were carried out under an argon atmosphere
with 5 mM stilbene in dry acetonitrile containing 0.1 M tetra-
n-butylammonium hexafluorophosphate as supporting elec-
trolyte. All cyclic voltammograms were recorded at sweep rates
of 100 mV s-1 and were iR compensated. The oxidation (peak)
potentials of the stilbenes in Chart 1 were referenced to SCE,
which was calibrated with ferrocene (5 mM) as the internal
standard.
to charge-transfer absorptions. The absorbance changes (ACT
)
were monitored at the absorption maximum (λCT) and quan-
titatively evaluated with the aid of the Benesi-Hildebrand
relationship.29 Thus a benzene solution of 0.005 M chloranil
was prepared in a quartz cuvette, and the UV-vis absorption
spectrum of the yellow solution was recorded. Stilbene (44.9
mg, 0.0831 M) was added and the absorption spectrum
re-recorded. The absorption spectrum of chloranil was digitally
subtracted and the difference hereinafter referred to as the
charge-transfer (CT) absorption spectrum (see inset in Figure
1). To determine the formation constant of the CT complex,
the concentrations of chloranil and stilbene donors were chosen
such that the stilbene was added incrementally in large excess
(i.e., [Q] ) 0.005 M and [stilbene] ranging from 0.0242 to
0.0831 M). The charge-transfer absorptions at the spectral
maxima (ACT) were determined and quantitatively evaluated
applying the Benesi-Hildebrand relationship,29 i.e.,
[Q]
ACT
1
1
ꢀCT
)
+
(7)
K
EDAꢀCT[S]
where KEDA represents the formation constant and ꢀCT the
extinction coefficient of the EDA complex. The values of KEDA
and ꢀCT extracted from the linear correlation of [Q]/ACT versus
[S]-1 are reported in Table 1 together with the absorption
maxima of the EDA complexes. The absorption maxima λCT of
the EDA complexes were directly related to the donor strength
of the various (E)-stilbenes evaluated by their anodic peak
potentials (Epox) (see Table 1), and thus the progressive blue
shift of the absorption band for chloro substitution and red
shift for methyl substitution confirm the charge-transfer
character of the EDA complex in eq 2 according to Mulliken
theory.32
Ack n ow led gm en t. We thank S. V. Lindeman for
crystallographic assistance, and the National Science
Foundation and the Robert A. Welch Foundation for
financial support.
J O981754N
(53) Bockman, T. M.; Kochi, J . K. J . Chem. Soc., Perkin Trans. 2
1996, 1633.
(54) Hubig, S. M.; Bockman, T. M.; Kochi, J . K. J . Am. Chem. Soc.
1996, 118, 3842.
Tim e-Resolved Absor p tion Sp ectr oscop y. Dir ect Ex-
cita tion of Ch lor a n il a t λexc ) 355 n m . The nanosecond