Concentration and temperature dependency of regio- and stereoselectivity in a photochemical [2 + 2] cycloaddition reaction (the Paterno-Buechi reaction): Origin of the hydroxy-group directivity

Article abstract of DOI:10.1021/ja1088524

A set of photochemical oxetane formation reactions, i.e., the Paterno-Buechi (PB) reactions, of tetrahydrobenzofuranol derivatives 1a-d with benzophenone (BP) was investigated to examine poorly understood hydroxy-group directivity on regio- and stereoselectivity. The selectivities of the PB reactions for allylic alcohols 1a,d were found to be largely dependent upon the concentration of the allylic alcohols and the reaction temperature. The temperature-dependent change of the regioselectivity at high concentrations of allylic alcohols was similar to that of the hydroxy-protected methyl ether 1b and tetrahydrobenzofuran (1c). The effect of concentration on regioselectivity can be explained by the hydrogen-bonded aggregates, which mimic the selectivities observed during the PB reaction of 1b,c. The hydroxy-directed cis-selectivity of the higher-substituted oxetane 3a,d formed at low concentration of 1a,d was found to be larger than that at the higher concentration of 1a,d. The cis-selectivity of 3a,d was found to be higher than that of the lower-substituted oxetane 2a,d. The regioselectivity of the cis-configured oxetanes was higher than that of the trans-configured oxetanes. These experimental results strongly suggest that hydroxy-group directivity, induced by hydrogen-bonding stabilization, plays a role in controlling the regio- and stereoselectivity of the PB reactions. The steric effect was also responsible for the diastereoselectivity, as shown by the fact that the cis selectivity in 3d was higher than that in 3a. Computational studies at the (U)MP2 and (U)DFT level of theory revealed that hydrogen-bonding stabilization becomes important only in the excited complex (exciplex) between the triplet excited state of carbonyls and alkenes, in which the charge transfer occurs from the alkene to the excited carbonyl to make the carbonyl oxygen nucleophilic. No significant stabilization energy was found in the intermediary triplet state of biradicals. The combined experimental and computational studies have clarified the origin of the poorly understood hydroxy-group effect on a high degree of regio- and stereoselectivity, i.e., the cooperative effect of hydrogen-bonding stabilization in exciplexes and the steric bulk of the substituents.

Full text of DOI:10.1021/ja1088524

ARTICLE
pubs.acs.org/JACS
Concentration and Temperature Dependency of Regio- and
Stereoselectivity in a Photochemical [2 þ 2] Cycloaddition Reaction
ꢀ
€
(the Paterno-Buchi Reaction): Origin of the Hydroxy-Group
Directivity
Youhei Yabuno,† Yoshikazu Hiraga,† Ryukichi Takagi,† and Manabu Abe*,†,‡
†Department of Chemistry, Graduate School of Science, Hiroshima University (HIRODAI), 1-3-1 Kagamiyama, Higashi-Hiroshima,
Hiroshima 739-8526, Japan, and
‡Japan Science and Technology Agency, CREST, 5 Sanbancho, Chiyodaku, Tokyo, 102-0075, Japan
S Supporting Information
b
ABSTRACT: A set of photochemical oxetane formation reactions, i.e., the
Paternꢀo-B€uchi (PB) reactions, of tetrahydrobenzofuranol derivatives 1a-
d with benzophenone (BP) was investigated to examine poorly understood
hydroxy-group directivity on regio- and stereoselectivity. The selectivities of
the PB reactions for allylic alcohols 1a,d were found to be largely dependent
upon the concentration of the allylic alcohols and the reaction temperature.
The temperature-dependent change of the regioselectivity at high concen-
trations of allylic alcohols was similar to that of the hydroxy-protected
methyl ether 1b and tetrahydrobenzofuran (1c). The effect of concentration on regioselectivity can be explained by the hydrogen-
bonded aggregates, which mimic the selectivities observed during the PB reaction of 1b,c. The hydroxy-directed cis-selectivity of the
higher-substituted oxetane 3a,d formed at low concentration of 1a,d was found to be larger than that at the higher concentration of
1a,d. The cis-selectivity of 3a,d was found to be higher than that of the lower-substituted oxetane 2a,d. The regioselectivity of the cis-
configured oxetanes was higher than that of the trans-configured oxetanes. These experimental results strongly suggest that hydroxy-
group directivity, induced by hydrogen-bonding stabilization, plays a role in controlling the regio- and stereoselectivity of the PB
reactions. The steric effect was also responsible for the diastereoselectivity, as shown by the fact that the cis selectivity in 3d was
higher than that in 3a. Computational studies at the (U)MP2 and (U)DFT level of theory revealed that hydrogen-bonding
stabilization becomes important only in the excited complex (exciplex) between the triplet excited state of carbonyls and alkenes, in
which the charge transfer occurs from the alkene to the excited carbonyl to make the carbonyl oxygen nucleophilic. No significant
stabilization energy was found in the intermediary triplet state of biradicals. The combined experimental and computational studies
have clarified the origin of the poorly understood hydroxy-group effect on a high degree of regio- and stereoselectivity, i.e., the
cooperative effect of hydrogen-bonding stabilization in exciplexes and the steric bulk of the substituents.
’ INTRODUCTION
disfavored process. Thus, the temperature and the concentration
of substrates, e.g. [ROH], largely affect the equilibrium con-
stants, K1, K2, and K3, and the reactive species of monomers (k1),
dimers (k2), and/or trimers (k3) (Scheme 1). When the reaction
partners, i.e., reagents, also have a potentially hydrogen-bonding
nature, the binding constants also play a crucial role in the
chemistry.
In organic synthesis, the energetic stabilization from hydrogen
bonds in reaction intermediates and/or transition states is a well-
known tool for the control of regioselectivity, chemoselectivity,
diastereoselectivity, and enantioselectivity.3 In photochemical
cycloaddition reactions, the structure of hydrogen-bonded com-
plexes of two reacting compounds in the ground state was found to
play crucial roles in controlling stereoselectivity, as exemplified in
The combined experimental and computational studies re-
ported here were intended to reveal the origin of the poorly
understood hydroxy-group directivity of regio- and diastereose-
lective [2 þ 2] photocycloadditions [the Paternꢀo-B€uchi (PB)
reactions]. The present study is a timely contribution to the
current interest in the hydrogen-bonding interaction of open-
shell molecules.1
Hydrogen bonds have a significant influence on the structure
of molecules/aggregates, on the function of materials, and on the
selectivities of many chemical and biological systems.2 The
stabilization effects mediated by the hydrogen atom, i.e., hydrogen-
bond energies, are largely affected by the acidity of donor groups
and the basicity of acceptor groups. For the formation of a water
or alcohol dimer, the enthalpy change of stabilization was found
to be ca. 4-6 kcal mol-1. Aggregation is an entropically
Received: September 30, 2010
Published: February 9, 2011
r
2011 American Chemical Society
2592
dx.doi.org/10.1021/ja1088524 J. Am. Chem. Soc. 2011, 133, 2592–2604
|

Journal of the American Chemical Society
ARTICLE
Scheme 2.4 The regio- and diastereoselectivity was found to
increase in nonpolar solvents and at low temperature, under
which conditions the entropically disfavored factor for the
formation of the host-guest complex should be minimized. As
for an intramolecular example, Snapper and co-workers reported
a notable hydrogen-bonding effect on stereoselectivity in intra-
molecular [2 þ 2] photocycloaddition reactions.5 In 2000, two
research groups independently reported hydroxy-group-directed
regio- and stereoselectivity during the formation of oxetanes in
the Paternꢀo-B€uchi (PB) reaction6 of allylic alcohols with
benzophenone (Scheme 3).7 Hydrogen-bonded stabilization in
the excited states, i.e., the interaction of the n,π* excited state of
benzophenone (BP3*) and alcohols, was proposed to play a role
in controlling regio- and diastereoselectivity to afford the inter-
mediary triplet 1,4-biradicals, followed by intersystem crossing
(ISC) to stereoselectively produce oxetanes. A significant in-
crease in diastereoselectivity was also found with an increase in
the steric bulk of the substituent R and R0 groups. The question
then becomes how much energetic stabilization is expected to
occur in the interaction between the excited state molecule (BP3*),
which possesses an excitation energy of ∼70 kcal mol-1 and the
lifetime of ∼7 μs,8 and the hydroxyl group (-OH). The
characteristics of an excited-state carbonyl can be seen in the
resonance structures depicted in Scheme 3b.9 The oxygen atom
of the n,π* excited-state carbonyl possesses a radical chara-
cteristic10 and an electron deficiency,11 which compares to that
of a ground-state carbonyl.
state should not cause the regio- and diastereoselectivity ob-
served in the PB reactions. Thus, it is not clear whether the regio-
and diastereoselectivity in the oxetane formation reaction is
controlled by the hydrogen-bonding stabilization between the
excited-state carbonyl and the hydroxyl group of the allylic
alcohols. Griesbeck et al. have also pointed out a similar funda-
mental question concerning the hydroxyl-group directivity,15 but
the origin of the directivity is still unclear. In recent reports, the
stereoselectivity observed in the PB reaction of cyclic allylic
alcohol, e.g., 2,3-dihydrofuran-3-ol, cannot be simply explained
and predicted by the hydroxy-group directivity.16
As shown in Scheme 1, the concentration of substrates and the
reaction temperature should mostly affect the reactive species, e.
g., k1/k2/k3, and thus, the product selectivity is expected to be
significantly dependent on the reaction conditions. It is some-
what surprising that the effects of the reaction conditions have
thus far not been examined for product selectivity in photo-
chemical oxetane formations. In consideration of this back-
ground, we decided to conduct the PB reaction of BP with
furan derivatives 1a,d with the hydroxyl-group configurationally
locked, thus giving clarification to the hydroxy group and its
steric effect on product selectivity. First, the effects of concentra-
tion and temperature17 were thoroughly examined on regio- and
stereoselectivity in the PB reaction of tetrahydrobenzofuranol
derivative 1a with the triplet n,π* excited state benzophenone
(BP3*).18 Photocycloaddition reactions with 1b,c were also
done to clarify the role of the hydroxyl group on regio- and
stereoselectivity in a photochemical reaction. The steric effect,
which should be considered as one of the important factors in
In fact, Iwata and Morokuma reported that almost no stabi-
lization energy was computed in the interaction of the n,π*
singlet/triplet state of formaldehyde with H2O, although
the stabilization energy of the hydrogen bond of ground-
state formaldehyde (H2CO) with H2O was found to be ca.
Scheme 3
3.5 kcal mol .
-1 12 The computational results are consistent with
the well-known blue shift of the n,π* transition of carbonyls.13
Thus, the transition in protic solvents needs more energy than
that in aprotic solvents. The ground state stabilization retards the
process of the n,π* excitation of the carbonyl, which is necessary
for the PB reaction. It is well-known that triplet state benzophe-
none abstracts a hydrogen atom from the CH group, rather than
from the OH group, in the reaction with 2-propanol to give a
ketyl radical.14 The hydrogen-bonded structure in the ground
Scheme 1
Scheme 2
2593
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
controlling selectivity, was examined in the PB reaction of 1d.
Computational studies have provided insight into the interaction
of a hydroxyl group with a triplet n,π* excited-state carbonyl
during the formation of the intermediary triplet 1,4-diradicals
and also have revealed the conformational stability of the
intermediary triplet biradicals.19 Combined experimental and
theoretical study has revealed that hydroxy-group selectivity in
the PB reaction could be reasonably explained by the cooperative
effect of hydrogen-bonded stabilization in the intermediary
exciplex and by the steric bulk of substituents, leading to the
preferred formation of oxetanes. Finally, the optimized condi-
tions found in the present study were applied to improve the
stereoselectivity of the PB reaction of allylic alcohol 4.
(m, 1 H), 5.16 (d, 1 H, J = 2.9 Hz), 6.33 (d, 1 H, J =
2.9 Hz), 7.10-7.40 (m, 8 H), 7.56-7.65 (m, 2 H). 13C NMR
(125 MHz, CDCl3): δ 19.30 (CH2), 28.97 (CH2), 30.04 (CH2), 61.12
(C), 72.26 (CH), 94.43 (C), 111.27 (C), 111.90 (CH), 125.23 (CH),
126.72 (CH), 126.86 (CH), 127.39 (CH), 128.14 (CH), 128.61 (CH),
143.05 (C), 145.00 (CH), 145.23 (C). HRMS (EI): calcd for C21H20O3
320.1413, found 320.1420. Anal. Calcd for C21H20O3: C, 78.73; H, 6.29.
Found: C, 78.71; H, 6.33. IR (KBr): 3529, 3023, 2991, 2948, 2893, 1955,
1699 cm-1
.
(3aR*,7R*,7aR*)-9,9-Diphenyl-4,5,6,7-tetrahydro-7a,3a-
(epoxymethano)benzofuran-7-ol (cis-3a). 1H NMR (500 MHz,
CDCl3): δ 1.18-1.31 (m, 1 H), 1.31-1.44 (m, 1 H), 1.47-1.73 (m,
3 H), 2.29-2.43 (m, 1 H), 2.60-2.65 (m, 1 H), 3.86-3.95(m, 1 H),
4.89 (d, 1 H, J = 2.9 Hz), 6.27 (d, 1 H, J = 2.9 Hz), 7.10-7.17 (m, 1 H),
7.17-7.23 (m, 1 H), 7.23-7.30 (m, 2 H), 7.30-7.37 (m, 2 H), 7.37-
7.43 (m, 2 H), 7,45-7.52 (m, 2 H). 13C NMR (125 MHz, CDCl3): δ
16.13 (CH2), 27.07 (CH2), 27.29 (CH2), 59.21 (C), 67.61 (CH), 97.40
(C), 109.24 (CH), 112.08 (C), 124.09 (CH), 125.01 (CH), 126.59
(CH), 127.20 (CH), 127.96 (CH), 128.57 (CH), 143.41 (C), 145.07
(C), 146.29 (CH). Mp: 132 °C. HRMS (EI): calcd for C21H20O3
320.1413, found 320.1420. Anal. Calcd for C21H20O3: C, 78.73; H, 6.29.
Found: C, 78.67; H, 6.29. IR (KBr): 3528, 3061, 2940, 2643, 1968, 1820,
’ EXPERIMENTAL METHODS
1764, 1613 cm-1
.
(2aR*,6S*,7aR*)-2,2-Diphenyl-2a,3,4,5,6,7a-hexahydro-
2H-oxeto[2,3-b]benzofuran-6-ol (trans-2b). 1H NMR (C6D6,
500 MHz): δ 7.56-7.17 (m, 10 H), 6.26 (d, 1 H, J = 4.4 Hz), 4.34 (d, 1
H, J = 4.4 Hz), 3.99-3.93 (m, 1 H), 3.47 (s, 3 H), 1.85-1.68 (m, 2 H),
1.58-1.45 (m, 1 H), 1.45-1.32 (m, 1 H), 1.24-1.13 (m, 1 H), 0.95-
0.83 (m, 1 H). 13C NMR (CDCl3, 125 MHz): 154.34 (C), 144.79 (C),
141.70 (C), 128.23 (CH), 127.57 (CH), 127.42 (CH), 127.25 (CH),
126.14 (CH), 125.87 (CH), 112.96 (C), 104.47 (CH), 94.05 (C) 71.52
(CH), 58.20 (CH), 57.45 (C), 28.03 (CH2), 22.66 (CH2), 18.65 (CH2).
General Procedure for the Photoreactions of Benzophe-
none with Furan Derivatives. A dried and degassed toluene
solution of freshly prepared furan derivative 120 and benzophenone
was irradiated for 6 h with a high-pressure Hg lamp through a Pyrex filter
at various temperatures and concentrations. After the solvent was
removed under reduced pressure using a vacuum pump (0.2 mmHg,
<10 °C), the photolysate was directly analyzed by 1H NMR (500 MHz)
spectroscopy. The product yields were determined on the basis of the 1H
NMR (500 MHz) peak areas (error (3%). Triphenylmethane (Ph3CH)
was used as an internal standard. The product ratios were determined by
comparisons of the peak areas. The photoproducts were isolated using
silica gel column chromatography. All of the configurational determina-
tions, i.e., trans- and cis-configuration, were unequivocally determined by
measuring the NOE enhancements in compounds 2a,b,d and 3a,b,d. This
spectroscopic evidence is summarized in the Supporting Information.
(2aR*,6S*,7aR*)-2,2-Diphenyl-2a,3,4,5,6,7a-hexahydro-
2H-oxeto[2,3-b]benzofuran-6-ol(trans-2a). 1H NMR (500 MHz,
CDCl3): δ0.88-1.01 (m, 1 H), 1.10-1.21 (m, 1 H), 1.45-1.82 (m, 4 H),
1.99-2.09 (m, 1 H), 4.37 (d, 1 H, J = 4.1 Hz), 4.39-4.47 (m, 1 H), 6.24
(d, 1 H, J = 4.1 Hz), 7.15-7.37 (m, 6 H), 7.38-7.51 (m, 4 H). 13C NMR
(125 MHz, CDCl3): δ 19.32 (CH2), 22.94 (CH2), 31.74 (CH2), 58.58
(CH), 63.24 (CH), 94.52 (C), 104.94 (CH), 112.33 (C), 126.14 (CH),
126.35 (CH), 127.56 (CH), 127.75 (CH), 127.86 (CH), 128.55 (CH),
141.93 (C), 145.00 (C), 155.51 (C). Mp: 157 °C. HRMS (EI): calcd for
C21H20O3 320.1413, found 320.1424. Anal. Calcd for C21H20O3: C, 78.73;
H, 6.29. Found: C, 78.72; H, 6.39. IR (KBr): 3529, 3023, 2991, 2948, 2893,
Mp: 97 °C. IR (KBr): 2944, 1958, 1694, 1598 1490 cm-1
.
(3aR*,7S*,7aR*)-9,9-Diphenyl-4,5,6,7-tetrahydro-
7a,3a-(epoxymethano)benzofuran-7-ol (trans-3b). 1H NMR
(C6D6, 500 MHz): δ 7.66-7.59 (m, 2 H), 7.38-7.11 (m, 8 H), 6.38 (d,
1 H, J = 2.8 Hz), 5.13 (d, 1H, J = 2.8 Hz), 3.92-3.87 (m, 1 H), 3.65 (s, 3
H), 2.31-2.16 (m, 1 H), 1.81-1.66 (m, 1 H), 1.51-1.34 (m, 2 H),
1.21-1.06 (m, 1 H), 0.77-0.57 (m, 1 H). 13C NMR (CDCl3, 125
MHz): 145.07 (CH), 145.03 (C), 142.89 (C), 128.14 (CH), 127.75
(CH), 126.88 (CH), 126.37 (CH), 125.36 (CH), 124.64 (CH), 111.31
(CH), 110.82 (C), 93.39 (C) 81.08 (CH), 62.28 (C), 58.51 (CH), 28.72
(CH2), 26.69 (CH2), 19.86 (CH2). Mp: 140 °C. IR (KBr): 2942, 1770,
1614, 1449 cm-1
.
(2aR*,7aR*)-2,2-Diphenyl-2a,3,4,5,6,7a-hexahydro-2H-
oxeto[2,3-b]benzofuran (2c). 1H NMR (500 MHz, CDCl3): δ
0.90-1.04 (m, 1 H), 1.03-1.74 (m, 3 H), 2.10-2.38 (m, 4 H), 4.28
(m, 1 H), 6.19 (d, 1 H, J = 4.2 Hz), 7.01-7.74 (m, 10 H). 13C NMR
(125 MHz, CDCl3): δ 23.25 (CH2), 23.30 (CH2), 23.62 (CH2),
24.02 (CH2), 59.36 (CH), 94.69 (C), 105.45 (CH), 108.33 (C),
127.08 (CH), 127.52 (CH), 128.07 (CH), 128.30 (CH), 129.01
(CH), 129.83 (CH), 143.41 (C), 145.79 (C), 146.48 (C). Mp: 102 °C.
HRMS (EI): calcd for C21H20O2 304.1463, found 304.1466. IR (KBr):
1955, 1699 cm-1
.
(2aS*,6S*,7aS*)-2,2-Diphenyl-2a,3,4,5,6,7a-hexahydro-
2H-oxeto[2,3-b]benzofuran-6-ol (cis-2a). 1H NMR (500 MHz,
CDCl3): 0.84-0.95 (m, 1 H), 1.04-1.21 (m, 1 H), 1.24-1.40 (m, 2 H),
1.54-1.81 (m, 2 H), 1.89-2.00 (m, 1 H), 4.27-4.39 (m, 1 H), 6.26 (d,
1 H, J = 4.2 Hz), 7.20-7.53 (m, 10 H). 13C NMR (125 MHz, CDCl3):
δ19.18 (CH2), 22.87 (CH2), 31.54 (CH2), 58.40 (CH), 63.09 (CH),
94.28 (C), 105.20 (CH), 111.84 (C), 126.31 (CH), 126.72 (CH),
127.83 (CH), 127.84 (CH), 128.01 (CH), 128.62 (CH), 141.74 (C),
145.12 (C), 155.84 (C). HRMS (EI): calcd for C21H20O3 320.1413,
found 320.1402.
(3aR*,7S*,7aR*)-9,9-Diphenyl-4,5,6,7-tetrahydro-7a,3a-
(epoxymethano)benzofuran-7-ol (trans-3a). 1H NMR (500 MHz,
CDCl3): δ 0.60-0.83 (m, 1 H), 1.13-1.36 (m, 1 H), 1.37-1.56 (m, 1 H),
1.64-1.84 (m, 2 H), 2.14-2.29 (m, 1 H), 2.29-2.48 (m, 1 H), 4.14-4.30
3063, 2967, 2887, 1966, 1898, 1656 cm-1
.
(3aR*,7aR*)-9,9-Diphenyl-4,5,6,7-tetrahydro-7a,3a-(epo-
xymethano)benzofuran (3c). 1H NMR (CDCl3, 500 MHz): δ
0.47-0.66 (m, 1 H), 1.05-1.21 (m, 1 H), 1.22-1.34 (m, 1 H), 1.37-
1.61 (m, 2 H), 1.98-2.11 (m, 1 H), 2.17-2.33(m, 2 H), 5.10 (d, 1 H, J =
2.9Hz), 6.32(d, 1H,J= 2.9 Hz), 7.08-7.43 (m, 8 H), 7.54-7.67 (m, 2 H).
13C NMR (CDCl3, 125 MHz): 20.81 (CH2), 23.86 (CH2), 29.78
(CH2), 33.09 (CH2), 58.64 (C), 94.07 (C), 111.45 (CH), 112.07
(C), 125.46 (CH), 126.14 (CH), 126.89 (CH), 127.34 (CH), 128.32
(CH), 128.68 (CH), 143.74 (C), 145.74 (CH), 146.06 (C); IR (KBr):
3278, 3058, 2975, 1956, 1887, 1662 cm-1. Mp: 118 °C. HRMS (EI):
2594
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
calcd for C21H20O2 304.1463, found 304.1454. IR (KBr): 3060, 2979,
BP was investigated using a high-pressure Hg lamp (150 W)
through a Pyrex filter (hν > 300 nm) in a dried and degassed
toluene solution (eq 1 and Table 1). In accordance with hydroxy-
group directivity, a cis-configured oxetane 3a is the preferred
result of the PB reaction. As shown in entries 1-5 ([1a] =
340 mM), at a higher temperature (60 °C) the trans-selective
formation was observed in the lower-substituted oxetane 2a
(entry 1). However, a slight, but nonetheless apparent, cis-
selective formation was detected in the higher-substituted
oxetane 3a. Compounds 5-8 were observed as minor products
in the photochemical reactions. The trans-selectivity of 2a
increased very slightly with a decrease in the reaction tempera-
ture, e.g., 70/30 at 60 °C (entry 1) and 82/18 at -75 °C (entry 5).
By contrast, the stereoselectivity of 3a, in which the oxetane ring
is closely located to the hydroxyl group, was drastically changed
from cis-selectivity at high-temperatures to trans-selectivity at
low temperatures at [1a] = 340 mM, i.e., from a ratio of 42/58 at
60 °C (entry 1) to a ratio of 72/28 at -75 °C (entry 5). The
regioselectivity 2a/3a was also found to be largely temperature-
dependent. Thus, the formation of 3a increased with an increase
in the reaction temperature, e.g., from a ratio of trans-2a/3a = 48/
52 and cis-2a/3a = 35/65 at -75 °C (entry 5) to a ratio of trans-
2a/3a = 20/80 and cis-2a/3a = 8/92 at 60 °C (entry 1). The
regioselectivity of the higher-substituted oxetane 3a was signifi-
cantly higher in the cis-isomer than that in the trans-isomer. The
effect of temperature on regioselectivity in the trans-2a/3a was
very similar to our previous observation in the PB reaction with
2-methylfurane, i.e., a lower-substituted/higher-substituted ox-
etane = 42/58 at -77 °C and 19/81 at 61 °C.24
To obtain further insight into the hydroxy-group effect on
stereoselectivity and regioselectivity, the effect of concentration
on selectivity in the formation of compounds 2a and 3a was
examined (entries 6-11, Figures 1 and 2). First, the effects of
concentration on the regioselectivity of trans- and cis-2a/3a were
examined at 60 °C (entries 6-8) and -75 °C (entries 9-11). At
60 °C (Figure 1a,b), concentration had no effect on regioselec-
tivity for either trans- or cis-isomers in a concentration range of
10 mM < [1a] < 1000 mM; for trans-isomers, trans-2a/3a ∼ 22/
78 (Figure 1a); for cis-isomers, cis-2a/3a ∼ 8/92 (Figure 1b).
These results are consistent with the general phenomena that
entropically disfavored aggregations are no longer possible at
such high temperatures. However, at low temperatures, such
as -75 °C (entries 9-11), a significanteffect of concentrationwas
observed on regioselectivity in cis-isomers cis-2a/3a (Figure 1d).
The ratio was changed from 47/53 at [1a] = 1000 mM to 19/81
at [1a] = 10 mM. An almost negligible effect was observed in the
trans-isomer trans-2a/3a (Figure 1c), i.e., from 53/47 at [1a] =
1000 mM to 45/55 at [1a] = 10 mM. Thus, the higher-
substituted cis-configured oxetane cis-3a was preferably formed
at a low concentration of 1a.
2863, 1958, 1660 cm-1
.
(2aR*,6S*,7aR*)-6-Methyl-2,2-diphenyl-2a,3,4,5,6,7a-hex-
ahydro-2H-oxeto[2,3-b]benzofuran-6-ol (trans-2d). 1H NMR
(500 MHz, C6D6): δ 0.80-0.97 (m, 1 H), 0.97-1.09 (m, 1 H), 1.24-
1.40 (m, 2 H), 1.40-1.55 (m, 5 H), 1.67-1.71 (m, 1 H), 3.91 (d, 1 H, J =
4.2 Hz), 6.03 (d, 1 H, J = 4.2 Hz), 6.90-7.20 (m, 6 H), 7.35-7.41 (m,
2 H), 7.44-7.75 (m, 2 H). 13C NMR (125 MHz, C6D6): δ 20.55 (CH2),
23.58 (CH2), 26.87 (CH3), 39.02 (CH2), 58.92 (CH), 67.35 (C), 94.42
(C), 105.05 (CH), 109.94 (C), 126.59 (CH), 126.88 (CH), 127.61
(CH), 127.74 (CH), 127.93 (CH), 128.65 (CH), 142.85 (C), 145.82
(C), 158.46 (C). HRMS (EI): calcd for C22H22O3 334.1569, found
334.1567.
(2aR*,6R*,7aR*)-6-Methyl-2,2-diphenyl-2a,3,4,5,6,7a-hex-
ahydro-2H-oxeto[2,3-b]benzofuran-6-ol (cis-2d). 1H NMR
(500 MHz, C6D6): δ 1.00-1.08 (m, 1 H), 1.10-1.19 (m, 1 H),
1.24-1.41 (m, 4 H), 1.41-1.50 (m, 1 H), 1.63-1.71 (m, 1 H), 1.82-
1.93 (m, 1 H), 3.92 (d, 1 H, J = 4.2 Hz), 6.05 (d, 1 H, J = 4.2 Hz),
6.90-7.21 (m, 6 H), 7.34-7.40 (m, 2 H), 7.53-7.58 (m, 2 H). 13C
NMR (125 MHz, C6D6): δ 20.28 (CH2), 23.33 (CH2), 26.06 (CH3),
38.93 (CH2), 58.54 (CH), 67.04 (C), 94.33 (C), 105.12 (CH), 109.62
(C), 126.70 (CH), 127.25 (CH), 127.73 (CH), 127.86 (CH), 128.09
(CH), 128.65 (CH), 142.47 (C), 145.83 (C), 158.72 (C). HRMS (EI):
calcd for C22H22O3 334.1569, found 334.1567.
(3aS*,7S*,7aS*)-7-Methyl-9,9-diphenyl-4,5,6,7-tetrahydro-
7a,3a-(epoxymethano)benzofuran-7-ol (trans-3d). 1H NMR
(500 MHz, C6D6): δ 0.79-0.89 (m, 1 H), 1.49 (s, 3 H), 1.58-
1.65 (m, 2 H), 1.66-1.75 (m, 1 H), 1.75-1.84 (m, 1 H), 1.95-1.99
(m, 1 H), 1.99-2.07 (m, 1 H), 4.62 (d, 1 H, J = 2.9 Hz), 5.82 (d, 1 H, J =
2.9 Hz), 6.91-6.98 (m, 2 H), 7.05-7.12 (m, 4 H), 7.40-7.44 (m, 2 H),
7.45-7.50 (m, 2 H). 13C NMR (125 MHz, C6D6): δ 16.44 (CH2), 23.06
(CH3), 25.94 (CH2), 32.62 (CH2), 58.99 (C), 70.78 (C), 95.70 (C),
110.35 (CH), 114.26 (C), 125.40 (CH), 125.89 (CH), 126.68 (CH),
127.14 (CH), 128.52 (CH), 128.79 (CH), 144.35 (C), 146.25 (C),
146.85 (CH). Rf: 0.625. HRMS(EI): calcdforC22H22O3 334.1569, found
334.1551.
(3aS*,7R*,7aS*)-7-Methyl-9,9-diphenyl-4,5,6,7-tetrahydro-
7a,3a-(epoxymethano)benzofuran-7-ol (cis-3d). 1H NMR
(500 MHz, C6D6): δ 1.24-1.48 (m, 6 H), 1.48-1.59 (m, 1 H), 1.68-
1.79 (m, 1 H), 2.43-2.49 (m, 1 H), 2.59-2.69 (m, 1 H), 4.42 (dd, 1 H,
J1 = 1.1 Hz, J2 = 2.8 Hz), 5.89 (dd, 1 H, J1 = 1.1 Hz, J2 = 2.8 Hz), 6.88-
7.21 (m, 6 H), 7.30-7.35 (m, 2 H), 7.44-7.50 (m, 2 H). 13C NMR (125
MHz, C6D6): δ 17.24 (CH2), 23.62 (CH3), 25.31 (CH2), 31.33 (CH2),
60.15 (C), 69.21 (C), 98.73 (C), 108.17 (CH), 114.58 (C), 125.03
(CH), 125.42 (CH), 126.75 (CH), 127.14 (CH), 128.52 (CH), 128.90
(CH), 144.68 (C), 146.01 (C), 147.31 (CH). HRMS (EI): calcd for
C22H22O3 334.1569, found 334.1551.
’ COMPUTATIONAL METHODS
All of the structures calculated for the hydrogen-bonded interaction
between H2CO and CH3OH in Figure 4 were optimized using
(U)MP2/6-31þG(d,p).21 The PB reactions of 1a,c with BP were
calculated at the (U)B3LYP/6-31þG(d)22 level of theory. All these
calculations were performed using the Gaussian 03 suite programs.23
The energies and optimized Cartesian coordinates are summarized in
the Supporting Information.
A similar effect of concentration was also observed in the
stereoselectivity of trans/cis-2a and -3a (Figure 2). At high
temperature, 60 °C, concentration had no effect on the stereo-
selectivity of 2a and 3a (entries 6-8, Figure 2a,b). However, a
notable effect of concentration at -75 °C was observed in the
formation of 3a (entries 9-11, Figure 2d). Thus, the formation
of cis-3a increased remarkably (from 19% to 49%) with a
decrease in the concentration of 1a. A very small increase
(from 15% to 21%) in the formation of cis-2a was observed
(Figure 2c). The effects of concentration shown in Figures 1 and
2 clearly indicate that the intermolecular hydrogen-bonding
interaction, e.g., the aggregation of 1a (Scheme 1), plays a crucial
’ RESULTS AND DISCUSSION
Effects of Temperature and Concentration on Product
Selectivity in a Photocycloaddition Reaction of 1a with
Benzophenone (BP). First, the temperature and concentration
effect on regio- and stereoselectivity in the PB reaction of 1a with
2595
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
Table 1. Effects of Temperature and Concentration on the Product Ratios in the PB Reaction of 1a with BPa
products and yieldsb/%
stereoselectivityc
regioselectivityd
entry
temp/°C
[1a]/mM
trans-2a
cis-2a
trans-3a
cis-3a
trans/cis-2a
trans/cis-3a
trans-2a/3a
cis-2a/3a
1
2
60
20
340
340
340
340
340
1000
100
10
7
3
4
4
7
7
3
3
4
7
7
6
0
28
30
32
39
33
25
22
30
34
34
27
44
37
36
34
23
13
36
32
42
8
70/30
73/27
79/21
80/20
82/18
70/30
67/33
67/33
85/15
81/19
79/21
>97/3
42/58
45/55
48/52
63/37
72/28
41/59
41/59
42/58
81/19
63/37
51/49
>97/3
20/80
28/72
33/67
42/58
48/52
22/78
21/79
21/79
53/47
47/53
45/55
31/69
8/92
11/89
11/89
23/77
35/65
8/92
12
15
28
31
7
3
0
4
-45
-75
60
5
6
7
60
6
7/93
8
60
8
8/92
9
-75
-75
-75
20
1000
100
10
39
30
22
20
47/53
26/74
19/81
-
10
11
12e
20
26
0
340
a The photoreaction of 1a with benzophenone (BP) was performed in degassed toluene with a high-pressure Hg-lamp through a Pyrex filter. The
recovered BP was less than 3% after 6 h of irradiation. Small amounts of compounds 5-8 (∼10%) were detected as minor products in the photolysate.
The mass balances were >83%. b The product yields were determined on the basis of 1H NMR (500 MHz) peak areas (error (3%). Triphenylmethane
(Ph3CH) was used as an internal standard. The ratios in parentheses are the normalized trans/cis stereoselectivity of compounds 2a and 3a. c The
e
stereoselectivities were normalized to 100%. d The regioselectivities were normalized to 100%. The reaction was performed in a dry and degassed
DMSO solution.
Figure 2. The effects of the concentration of 1a on stereoselectivity in
the formation of oxetane 2a and 3a, which were normalized to 100%:
(a) trans/cis-2a at 60 °C, (b) trans/cis-3a at 60 °C, (c) trans/cis-2a at
-75 °C, and (d) trans/cis-3a at -75 °C.
Figure 1. Concentration effect of 1a on regioselectivity in the formation
of oxetanes 2a and 3a, which were normalized to 100%: (a) trans-2a/3a
at 60 °C, (b) cis-2a/3a at 60 °C, (c) trans-2a/3a at -75 °C, and (d) cis-
2a/3a at -75 °C.
role in controlling both regio- and stereoselectivity in the
formation of oxetanes. In fact, the H NMR resonance of the
hydroxyl proton of 1a was temperature and concentration
sensitive (Figures 3a ,b). Thus, a downfield shift of the OH
1
2596
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
Scheme 4
mass balance of 84% (entry 8 in Table 1). The ratio was normal-
ized to 100%. Of note, trans-selectivity was found for 2a, but cis-
selectivity was found for 3a, although the steric bulk of a hydroxy
group is much larger than that of a hydrogen atom.25 These results
strongly suggest that a hydrogen-bonding interaction in the PB
reaction of 1a plays an important role in controlling regio- and
stereoselectivity at a low concentration of 1a in a nonpolar solvent,
although negligible hydrogen-bonding stabilization is expected in
the interaction of the n,π* excited carbonyl with alcohol (vide
infra).12 Under conditions of low temperature and high concen-
tration, the trans-configured oxetanes were selectively formed, i.e.,
trans-2a/cis-2a/trans-3a/cis-3a = 44/8/37/11(entry 9 inTable 1).
These results clearly indicate that the aggregation plays a crucial
role in controlling the face selectivity. Therefore, only the trans-
isomers of 2a and 3a were detected in DMSO, which is a good
hydrogen-bondacceptor (entry 12inTable 1). Aggregation would
suppress the approach of BP3* from the side of the hydroxyl group
in 1a to afford the trans-selective formation of the intermediary
triplet 1,4-diyls t-T-BR2a and t-T-BR3a (Scheme 4).
Photocycloaddition Reaction of 1b,c with Benzophenone
(BP). To obtain more information about the hydroxyl group
effect on regio- and stereoselectivity, the PB reactions of the
hydroxyl-protected methyl ether 1b26 and tetrahydrobenzofuran
(1c) with BP were investigated (eq 2, Table 2). As shown in
entries 1-6, the clean formation of the tricyclic oxetanes 2b,c
and 3b,c were observed with high mass balances, and thus, the
determination of the product ratios was feasible. The temperature-
dependent change of the regioselectivity observed in the PB
reactions of both 1b and 1c was quite similar to that detected in
trans-2a/3a; compare entries 1-5 with entries 1-5 in Table 1.
In sharp contrast to the stereoselectivity observed in the reaction
with 1a, i.e., trans-selectivity in 2a and cis-selectivity in 3a, only
trans-configured oxetanes 2b and 3b were detected. We could
not observe the cis-isomers using 1H NMR (500 MHz) spectro-
scopic analyses. No effect of concentration was observed for
product selectivity (entries 5 and 6). These results clearly indi-
cate that the notable concentration-dependent change of regios-
electivity in cis-2a/3a (Figure 1) and the stereoselectivity in
trans/cis-3a (Figure 2) were due to the aggregation around the
hydroxyl group in 1a.
The PB Reaction of 1d with Benzophenone (BP). To gain
insight into the steric effect on diastereoselectivity in oxetane
formation, the methyl-substituted alcohol 1d was used for a PB
reaction with benzophenone (BP3*), as shown by eq 3 and in
Table 3. Cis-stereoselectivity was apparent in the formation of
3d, i.e., trans/cis-3d = 24/76 at 60 °C (entry 1), which was higher
than that obtained in 3a, trans/cis-3a = 42/58 (entry 1 in
Table 1). As observed for stereoselectivity in 2a and 3a, a
significant decrease in the cis-selectivities of 2d and 3d was
obtained with a decrease in the reaction temperature (entries
1-5). At a low concentration of 1d, the cis-selectivity in 3d was
higher than that at higher concentration of 1d, as shown by
1
Figure 3. Variable temperature H NMR spectrum of 1a (500 MHz,
C7D8) at (a) higher concentration of [1a], i.e., [1a] = 100 mM and [BP]
= 50 mM, and (b) lower concentration of [1a], i.e., [1a] = 6.8 mM and
[BP] = 3.4 mM.
proton resonance was observed from δ 1.86 (0 °C) to 4.02
(-60 °C) with a decrease in temperature when compound 1a
(100 mM) was dissolved in dried toluene-d8 (Figure 3a). A similar
change in the hydroxyl proton resonance was also observed in the
absence of benzophenone. No change was observed in the
resonance of the ortho-protons of benzophenone (δ 7.88 ppm).
Of note, a small, but significant, shift was detected in the resonance
of the proton Hb, i.e., δ 4.70 (0 °C) and 4.80 (-60 °C). No
movement was observed in the resonance of the Ha proton, δ 6.15
(0 °C) and 6.14 (-60 °C). This observation supports that the
aggregation is the intermolecular hydrogen bonding around the
hydroxyl group. A small shift of the resonance of the OH proton,
i.e., δ 1.47 (0 °C) and 1.70 (-60 °C), was observed at the lower
concentrationsof 1a(6.8 mM) and BP (3.4 mM) in driedtoluene-
d8 (Figure 3b). Thus, the entropically disfavored intermolecular
interaction will be diminished under the conditions of low
concentration and high temperature. Under these conditions
(10 mM and 60 °C), the cis-configured 3a was found to be a
major product, trans-2a/cis-2a/trans-3a/cis-3a = 10/4/35/51 at a
2597
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
Table 2. Effects of Temperature and Concentration on the Product Ratios in the PB Reaction of 1b,c with BPa
products and yieldsb/%
for 1b þ BP
for 1c þ BP
3c
regioselectivityc (mass balance/%)d
entry
temp/°C
concn /mM
trans-2b
trans-3b
2c
trans-2b/3b
2c/3c
1
2
3
4
5
6
60
20
340
340
340
340
340
6.8
20
15
16
30
34
27
52
32
33
50
51
41
13
16
16
26
28
23
66
61
61
59
52
46
28/72 (82)
31/69 (80)
32/68 (84)
37/63 (81)
40/60 (87)
40/60 (96)
16/84 (87)
21/79 (83)
21/79 (83)
31/69 (92)
35/65 (87)
34/66 (91)
0
-45
-75
-75
a The photoreaction of 1b or 1c with benzophenone (BP) was performed in degassed toluene with a high-pressure Hg-lamp through a Pyrex filter. The
recovered BP was less than 3% after 6 h of irradiation. Small amounts of compounds 5-7 (<10%) were detected as a minor product in the photolysate.
b The product yields were determined on the basis of 1H NMR (500 MHz) peak areas (error (3%). Triphenylmethane (Ph3CH) was used as an internal
standard. c The regioselectivities were normalized to 100%. d The vales represent the total yields of 2, 3, 5, and the recovered BP.
Table 3. Product Ratios in the PB Reaction of 1d with BPa
products and yieldsb/%
stereoselectivityc
regioselectivityd
entry
temp/°C
[1d]/mM
trans-2d
cis-2d
trans-3d
cis-3d
trans/cis-2d
trans/cis-3d
trans-2d/3d
cis-2d/3d
1
2
3
4
5
6
7
8
60
20
340
340
340
340
340
6.8
3
4
4
5
12
18
23
30
29
7
39
55
58
38
22
18
30
34
44/56
42/58
53/47
62/38
70/30
40/60
48/52
51/49
24/76
25/75
29/71
44/56
57/43
26/74
26/74
32/68
19/81
17/83
22/78
34/66
42/58
23/77
25/75
33/67
9/91
8/92
0
7
6
9/91
-46
-75
60
16
21
2
10
9
20/80
30/70
14/86
12/88
19/81
3
0
6.8
4
4
11
-75
6.8
8
8
17
a The photoreaction of 1d with benzophenone (BP) was performed in degassed toluene with a high-pressure Hg-lamp through a Pyrex filter. The
recovered BP was less than 3% after 6 h irradiation. Small amounts of compounds 5-7 (∼10%) were detected as minor products in the photolysate. The
mass balances were >83%. b The product yields were determined on the basis of 1H NMR (500 MHz) peak areas (error (3%). Triphenylmethane
(Ph3CH) was used as an internal standard. The ratios in parentheses are the normalized trans/cis-stereoselectivity of compounds 2d and 3d. c The
stereoselectivities were normalized to 100%. d The regioselectivity were normalized to 100%.
entries 5 and 8. The notable temperature and concentration
effects can be reasonably explained by the existence of the
hydrogen-bonded aggregation, which decreases the cis-selec-
tivity in 3d. Thus, a hydrogen-bonded aggregation at low
temperature occurs in some cases even with a low concentration
of the substrate. The cis-selectivity in 3d was higher than that in
3a; therefore, the steric effect was also responsible for increasing
cis-selectivity. The regioselectivity observed in the PB reaction of
1d, i.e., 2d/3d, was quite similar to that detected in the PB
reaction of 1a.
Hydrogen-Bonding Interaction of the Ground-State and
the n,π* Triplet Excited-State Formaldehyde with Methanol.
As mentioned in the Introduction, Morokuma and Iwata reported
a very small energetic stabilization of 0.2-0.4 kcal mol-1 for the
2598
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
triplet state were significantly smaller than those calculated for the
ground state complex H-O: 4.5, 6.0, and 6.0 kcal/mol, respec-
tively. Even with the low-cost calculations, the hydrogen-bonded
stabilization energy in the triplet state was estimated to be
significantly smaller than that in the ground state. However, our
experimental studies clearly indicate that the hydroxy group
indeed directs the regio- and stereochemical outcomes in the PB
reaction of 1a with BP, i.e., the selective formation of cis-3a,d at a
low concentration of 1a,d.
Computational Studies on the PB Reaction of 1c. To
elucidate the role of the hydroxy group, the Mulliken-charge
change on the oxygen atom of H2CO was calculated in a model
PB reaction of the triplet excited formaldehyde H2CO3* with 1c
at the UB3LYP/6-31þG(d) level of theory, leading to the
formation of the intermediary triplet biradical T-BR3-in
(Figure 5).22,28 As Figure 5b (red line) clearly shows, the negative
charge density on the oxygen atom of H2CO3* increased from a
value of -0.14 to -0.25 with a decrease in the distance, r, of
C(8)-O, i.e., from ca. 350 to 230 pm within the sum of the van
der Waals radii of carbon (170 pm) and oxygen (152 pm). The
scan-mode calculations clearly showed that the charge transfer
occurred from the alkene part (furan) to the electrophilic oxygen
of the n,π* excited H2CO during a decrease in the atom distance
of C(8)-O. At the local energy minimum structure (Figure 5c),
i.e., exciplex, at a distance, r, of 253 pm, the negative charge of the
carbonyl oxygen was calculated to be -0.25, which is the
maximum value at the same level of theory. The Mulliken charge
on the carbonyl oxygen of the ground-state H2CO was found to
be -0.30 at the same level of theory. Thus, the charge transfer29
during the formation of the exciplex occurred to make the oxygen
of H2CO3* nucleophilic, i.e, a good hydrogen-bonding acceptor.
During the bond formation between C(8) and the carbonyl
oxygen with a small energy barrier of ca. 0.9 kcal/mol, the charge
on the oxygen atom was found to decrease to ca. -0.10
(Figure 5b), which was smaller than the value of H2CO3*
(-0.15). The decrease in the negative charge is rationalized by
the resonance structure of T-BR3-in in Figure 5, i.e., the
character of Oþ•-C-.30 The electron delocalization from the
2p AO of the oxygen to the radical p-orbital is responsible for the
resonance structure and the perpendicular orientation of the
carbonyl moiety with a furan ring in the optimized structure of
T-BR3-in (Figure 5). Thus, the basicity of the oxygen atom
derived from H2CO in T-BR3-in is weaker than that of the
oxygen atom in H2CO3*.
As shown in Figure 5, the transition-state energy for the
generation of T-BR3-in from the intermediary exciplex was
calculated to be lower than the energy of the reactants (1c þ
H2CO3*). Thus, the process is not the selectivity-differentiating
step, i.e., the rate-determining step, in the PB reaction of the small
molecule H2CO3*. However, the transition state for generating
the intermediary triplet biradicals becomes important for diaster-
eoselectivity in the PB reaction of the more-steric bulk benzo-
phenone triplet state BP3* (Figure 6 and Scheme 5). As shown in
Figure 6, the basicity of the benzophenone oxygen, i.e., the
Mulliken charge on the oxygen atom, increases with a decrease in
the bond distance C(8)-O, which is quite similar to the charge-
change calculated for H2CO3* (Figure 5). Of note, the transition
state TS was calculated to be higher in energy than that of the
reactants 1c and BP3* (Scheme 5). The negligible basicity was
also calculated for the biradical T-BR3c-in. Thus, the charge of
the oxygen atom in T-BR3c-in was found to be -0.13, although
the charge of the triplet benzophenone (BP3*) was found to
Figure 4. Hydrogen-bonded structures and stabilization energies
(ΔEbsse) at the (U)MP2/6-31þG(d,p) level of theory after BSSE
corrections between a and c for the n,π* excited state H2CO and
CH3OH and (d) for the ground state H2CO and CH3OH.
interaction between the n,π* triplet excited state formaldehyde
(H2CO) and H2O, which was found to be about 3.5 kcal mol-1
smaller than the stabilization energy for the hydrogen-bonding
interaction of ground-state H2CO with H2O.12 To clarify the
structures of the hydrogen-bonded complexes and stabilization
energies of a similar interaction with alcohols, we first calculated
the equilibrium structures and the stabilization energies of ground-
state and n,π* triplet excited-state H2CO with methanol (CH3-
OH) at the (U)MP2/6-31þG(d,p) level of theory (Figure 4). As
reported for the H2CO-H2O interaction, the structures of the
vertical approach of the hydroxyl group to the H2CO plane, i.e.,
V-O1, V-O2, V-C, were optimized as equilibrium struc-
tures in the interaction of the triplet excited H2CO and CH3OH,
in which the stabilization energies were calculated to be 0.70, 0.65,
and 0.50 kcal mol-1, after the counterpoise (BSSE) corrections,
respectively. In the ground-state calculations, the structure of the
horizontal approach of the hydroxyl group, H-O, was optimized
at the same level of theory. The interaction energy was found to be
4.68 kcal mol-1 after the BSSE correction. Such horizontal
complexes were not found to be equilibrium structures for the
triplet-state calculations. This result is quite reasonable, since the
direction of the original n-orbital was changed to electrophilic after
the n,π* electronic excitation (Scheme 3b). The ab initio calcula-
tions clearly indicate that an almost negligible stabilization effect is
expected between the hydroxyl group and the n,π* triplet-state
carbonyl (Figure 4). The stabilization energies for V-O1 and
H-O were also estimated by more conventional low-cost
(U)DFT methods at the UB3LYP, UM05, and UM06 with the
6-31þG(d) basis set, although the DFT methods are known to
overestimate dispersion forces.27 Low-cost calculations are needed
for the PB reactions of 1 with BP. The stabilization energies were
established at 2.1, 2.6, and 3.2 kcal/mol for V-O1, respectively,
which were higher values than those obtained using the UMP2
method. However, the stabilization energies calculated for the
2599
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
Figure 5. The UB3LYP/6-31þG(d) analysis of (a) the energy profile and (b) the distance-dependent change of the Mulliken charge of the oxygen
atom of H2CO during the bond-formation of the triplet n,π* excited formaldehyde (H2CO) with the furan derivative 1c. (c) The structure at the distance
of r = 253 pm.
Scheme 5. Optimized Structures and Energies (kcal/mol) of
the Transition State TS and the inside Conformer of the
Biradicals T-BR3c-in Derived from the Reaction of BP3* with
1c at the UB3LYP/6-31þG(d) Level of Theory
Figure 6. The UB3LYP/6-31þG(d) analysis of (a) the energy profile
and (b) the distance-dependent change of the Mulliken charge of the
oxygen atom of Ph2CO (BP) during the bond formation of the triplet
n,π* excited benzophenone (Ph2CO) with the furan derivative 1c.
biradicals, the cis-configured biradical c-T-BR3a-in was calcu-
lated to be only 0.2 kcal/mol more stable than the trans-
configured biradical t-T-BR3a-in. The hydrogen-bond distance
of 218 pm was longer than that found in the transition state cis-
TS (188 pm). Thus, the energetic stabilization in the exciplex was
found to be significantly larger than that in the intermediary
triplet biradicals. The preferred formation of cis-3a in our experi-
ment under a low concentration of 1a, where the aggregation
would be suppressed, can be reasonably explained by the ener-
getic preference of the formation of the cis-configured biradical
c-T-BR3a-in. The inside conformer was calculated to be more
stable than the anticonformer c-T-BR3a-anti by 3.8 kcal/mol,
which would produce the starting compounds 1a and benzo-
phenone after the ISC process and followed by the C-O bond
breaking. Indeed, the optimization of the singlet state of the anti
biradical c-S-BR3a-anti produced 1a and BP at the broken-
symmetry (U)B3LYP/6-31þG(d) level of theory, which clearly
indicated that the C-O bond cleavage is a process without
barrier and is faster than the conformational change to the
productive inside conformer. The anticonformer of the trans-
configured biradical t-T-BR3a-anti was also calculated to be less
stable by comparison with the inside conformer t-T-BR3a-in,
be -0.20 at the same level of theory. This clearly indicates that
the oxygen atom derived from BP in the triplet biradicals is
not the hydrogen-bond acceptor. The charge density in TS
was calculated to be -0.25, which was larger than those
in either the biradical or BP3*, and therefore, it should be
nucleophilic.
Computational Studies on the PB Reaction of 1a with
BP3*. As mentioned above in the computational studies, hydro-
gen-bonding stabilization is possible only during the formation of
the charge-transferred exciplex. To prove the stabilization effect,
the energy profile of the PB reaction of 1a with the triplet state of
benzophenone BP3* was calculated at the UB3LYP/6-31þG(d)
level of theory (Scheme 6). Indeed, the transition state cis-TS
(ΔE = 11.4 kcal/mol) with an imaginary frequency of νi -140.4
cm-1, which produces cis-3a via the triplet biradical c-T-BR3a-in
after the intersystem crossing (ISC) process, was calculated to be
lower in energy by 1.4 kcal/mol than the transition state trans-TS
(ΔEZPE = 12.8 kcal/mol) with an imaginary frequency of νi -
99.9 cm-1, which affords trans-3a via t-T-BR3a-in. As expected
from the negligible basicity of the oxygen atom in the triplet
2600
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
Scheme 6. Energy Profiles of the PB Reaction of 1a with BP3* at the UB3LYP/6-31þG(d) Level of Theorya
a The energies (ΔEZPE in kcal mol-1) were relative to the energy for c-T-BR3a-in after the zero-point energy corrections.
although the energy difference (ΔΔEZPE = 1.5 kcal mol-1
)
biradicals BR3a, however, the productive conformers, T-BR3a-
in, were found to be more stable than the unproductive con-
formers, T-BR3a-anti. The cis-configured T-BR3a-in was found
to be a much more stable conformer than the anticonformer
(T-BR3a-anti), since the steric repulsion between the hydroxy
group (R = OH, R0 = H) and the diphenylmethyl moiety is more
severe than that between the hydrogen and the diphenylmethyl
group (R = H, R0 = OH) in the trans-configured isomer.
Thus, the higher population of the inside conformer in the cis-
configured biradical, by comparison with the trans-configured
one, is a reason for the higher regioselectivity in cis-configured
oxetane.
was smaller than the cis-isomer c-T-BR3a (ΔΔEZPE = 3.8 kcal
mol-1). The energetic preference of the inside conformers can
be reasonably explained by the steric repulsion between the
phenyl ring with the cyclohexane ring system in the antic-
onformer. Although the outside conformers t-T-BR3a-out and
c-T-BR3a-out are also the possible structures of the intermediary
triplet biradicals, the conformers were not calculated to be the
equilibrium structures at the same level of theory. The energetic
preference of the inside and anti conformers can be reasonably
explained by the stereoelectronic effect, i.e., the gauche effect, on
the conformational stability of the ketal moieties.22,31
Regioselectivity in Oxetanes 2 and 3. As found in our
previous study on the PB reaction with 2-methylfurane,22a the
higher-substituted oxetanes 3 were selectively formed at higher
temperatures (Tables 1 and 2). The selective formation of 3 can
be understood by the equilibrium constants between the pro-
ductive inside conformers and the unproductive anticonformers.
As exemplified in Scheme 7, the lower-substituted biradicals
BR2a and the anticonformers T-BR2a-anti for both the cis and
trans-configured biradicals were calculated to be more stable than
the productive conformers, T-BR2a-in, by 1.4 kcal/mol at the
UB3LYP/6-31þG(d) level of theory. In the higher-substituted
The PB reaction of Allylic Alcohol 4 with BP, Revisited. As
found in the PB reaction of 1a,d with BP, the temperature and
concentration largely affect the degree of the regioselectivity and
diastereoselectivity. The diastereoselectivity of the oxetane for-
mation of 4 with BP also was influenced by the reaction
conditions. Thus, the PB reaction was revisited to find the opti-
mal conditions for stereoselectively obtaining oxetane (eq 4),
and the results are summarized in Table 4. Actually, diastereos-
electivity was largely dependent on the reaction temperature and
the concentration of 4. Thus, the diastereoselectivity, threo/
erythro, at high temperatures was found to be higher than that at
2601
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
Scheme 7
Table 4. Temperature and Concentration Effect on the
low temperature, e.g., 90/10 (entry 2, at 20 °C) and 78/22 (entry
4, -75 °C) at [4] = 340 mM. This clearly indicates that the
hydrogen-bonded aggregation around the hydroxy group de-
creased diastereoselectivity. Indeed, when the reaction was
performed at a higher concentration of 4, substantial decreases
in diastereoselectivity were observed, 54/46 (entry 5) and 65/35
(entry 6). A significant increase in diastereoselectivity was
obtained at a lower concentration and reaction temperature
(entry 8).
Product Ratios in the PB Reaction of 4 with BPa
products and
yieldsb/%
Summary and Conclusions. In the present study, we thor-
oughly investigated the temperature and concentration effect on
regio- and stereoselectivity in the PB reaction of allylic alcohols
1a and 1d to clarify the origin of the hydroxy-group directivity in
oxetane formation and to provide a reasonable mechanism for a
photochemical reaction. In the experimental studies, we found
that the selectivities were largely dependent on the reaction
conditions due to the existence of the aggregates. At a low
concentration of substrate and under high temperature condi-
tions, a cis-selective formation for oxetanes 3a and 3d was
found. Thus, a hydrogen-bonded interaction together with
the steric effect was strongly proposed in the origin of the
stereochemical outcome, although hydrogen-bonding stabili-
zation was found to be negligible in the triplet state. The
contradiction was solved when the energy profile of the PB
reaction of the triplet H2CO or benzophenone (BP) with 1c
was calculated, in which a clear charge transfer was found from
the alkene part to the electron-deficient carbonyl oxygen,
leading to the exciplex. The charge transfer makes the carbonyl
oxygen nucleophilic, and thus, the oxygen atom becomes the
hydrogen-acceptor. The hydrogen-bonded stabilization was
actually found in the transition state that produced the cis-
configured triplet biradicals. The regioselective formation of
the higher-substituted oxetane can be reasonably explained by
the equilibrium constant between the productive conformers
and the unproductive conformers. The present study gave
diastereo-selectivity:c mass
entry temp/°C [4]/mM threo erythro
threo/erythro
balanced
1
2
3
4
5
6
7
8
60
340
340
340
340
5000
1000
100
6.8
54
67
70
60
37
50
62
72
4
88/12
90/10
88/12
78/22
54/46
65/35
84/16
96/4
92
88
92
88
90
86
74
77
20
7
-25
-75
-75
-75
-75
-75
14
17
31
27
10
3
a The photoreaction of 4 with benzophenone (BP) was performed in
degassed toluene with a high-pressure Hg-lamp through a Pyrex filter.
The recovered BP was less than 3% after 6 h of irradiation. Small
amounts of compounds 5-7 (∼10%) were detected as minor products
in the photolysate. b The product yields were determined on the basis of
1H NMR (500 MHz) peak areas (error (3%). Triphenylmethane
(Ph3CH) was used as an internal standard. c The ratios are the normal-
ized threo/erythro stereoselectivity. d Thevalues are totalyields of4, 5-7,
and the recovered BP.
clarity to the poorly understood hydroxy-group directivity of
the PB reaction and showed that it is cooperative hydrogen-
bonding stabilization and the steric effect that determine
selectivity.
2602
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
’ ASSOCIATED CONTENT
Photobiology; Horspool, W. M., Song, P.-S., Eds.; CRC Press: Boca
Raton, FL, 1994; pp 522-535 and 536-549. (g) Griesbeck, A. G.,
Bondock, S. CRC Handbook of Organic Photochemistry and Photobiology,
2nd ed.; Horspool, W. M.; Lenci, F., Eds.; CRC Press: Boca Raton, FL,
2004; Chapters 59 and 60. (h) Abe, M. CRC Handbook of Organic
Photochemistry and Photobiology, 2nd ed.; Horspool, W. M.; Lenci, F.,
Eds.; CRC Press: Boca Raton, FL, Chapter 62. (i) Porco, J. A.;
Schreiber, S. L. Comprehensive Organic Synthesis; Trost, B. M., Ed.;
Pergamon Press: Oxford, 1991; Vol. 5, pp 151-192. (j) Bach, T. Liebigs
Ann. 1997, 1627–1634. (k) Bach, T. Synthesis 1998, 683–703.
(7) (a) Adam, W.; Peters, K.; Peters, E. M.; Stegman, V. R. J. Am.
Chem. Soc. 2000, 122, 2958–2959. (b) Adam, W.; Stegmann, V. R.
Synthesis 2001, 1203–1214. (c) D’Auria, M.; Racioppi, R.; Romaniello,
G. Eur. J. Org. Chem. 2000, 3265–3272. (d) D’Auria, M.; Emanuele, L.;
Poggi, G.; Racioppi, R.; Romaniello, G. Tetrahedron 2002, 58, 5045–
5051. (e) D’Auria, M.; Emanuele, L.; Racioppi, R. Photochem. Photobiol.
Sci. 2004, 3, 927–932. (f) D’Auria, M.; Emanuele, L.; Racioppi, R.;
Valente, A. Photochem. Photobiol. Sci. 2008, 7, 98–103.
(8) Wolf, M. W.; Brown, R. E.; Singer, L. A. J. Am. Chem. Soc. 1977,
99, 526–531.
(9) Zimmerman, H. E. Angew. Chem., Int. Ed. 1969, 8, 1–11.
(10) Paul, H.; Small, R. D., Jr.; Scaiano, J. C. J. Am. Chem. Soc. 1978,
100, 4520–4527.
(11) (a) Scaiano, J. C. J. Photochem. 1973, 2, 81–118. (b) Rao,
C. N. R.; Murthy, A. S. N. Theoret. Chim. Acta 1971, 22, 392–395. (c)
Del Bene, J. E. J. Am. Chem. Soc. 1973, 95, 6517–6522. (d) Del Bene,
J. E.; Gwaltney, S. R.; Bartlett, R. J. J. Phys. Chem. A 1998, 102, 5124–
5127. (e) Del Bene, J. E.; Watts, J. D. Int. J. Quan. Chem. 2000, 77, 187–
191. (f) Pischel, U.; Nau, W. M. Photochem. Photobiol. Sci. 2002, 1, 141–
147.
(12) (a) Morokuma, K. J. Chem. Phys. 1971, 55, 1236–1244. (b)
Iwata, S.; Morokuma, K. Chem. Phys. Lett. 1973, 19, 94–98. (c) Iwata, S.;
Morokuma, K. J. Am. Chem. Soc. 1973, 95, 7563–7575.
(13) (a) Liao, D.-W.; Mebel, A. M.; Chen, Y.-T.; Lin, S.-H. J. Phys.
Chem. A 1997, 101, 9925–9934. (b) Wagner, P. J.; Truman, R. J.;
Scaiano, J. C. J. Am. Chem. Soc. 1985, 107, 7093–7097.(c) Wagner, P. J.;
Park, B.-S. In Organic Photochemistry; Padwa, A., Ed.; Macel Dekker:
New York, 1991; Vol. 11, Chapter 4.
(14) (a) Ciamician, G.; Silber, P. Chem. Ber. 1900, 33, 2911–2913.
(b) Jacques, P. J. Photochem. Photobiol. A: Chem. 1991, 56, 159–163.
(15) Griesbeck, A. G.; Bondock, S. J. Am. Chem. Soc. 2001, 123,
6191–6192.
(16) (a) Abe, M.; Terazawa, M.; Nozaki, K.; Masuyama, A.; Hayashi,
T. Tetrahedron Lett. 2006, 47, 2527–2530. (b) Yabuno, Y.; Hiraga, Y.;
Abe, M. Chem. Lett. 2008, 822–823.
(17) Chiral Photochemistry; Inoue, Y., Ramamurthy, V., Eds.; Marcel
Dekker: New York, 2004.
(18) In our previous study (ref 16b), only the temperature effect on
selectivity was examined in the PB reaction of 1a with benzophenone.
The concentration effect was also investigated in the present study to
clarify the effect of the aggregation on the regio- and stereoselectivity
(Scheme 1).
(19) The conformation of the intermediary triplet biradical is
important in controlling the stereoselectivity in the PB reactions, i.e.,
Griesbeck model:(a) Griesbeck, A. G.; Stadtm€uller, S. J. Am. Chem. Soc.
1990, 112, 1281–1288. (b) Griesbeck, A. G.; Stadtm€uller, S. J. Am.
Chem. Soc. 1991, 113, 6923–6928. (c) Griesbeck, A. G.; Mauder, H.;
Stadtm€uller, S. Acc. Chem. Res. 1994, 27, 70–75. (d) Abe, M.; Torii, E.;
Nojima, M. J. Org. Chem. 2000, 65, 3426–3431. (e) Abe, M.; Fujimoto,
K.; Nojima, M. J. Am. Chem. Soc. 2000, 122, 4005–4010. (f) Abe, M.;
Taniguchi, K. ARKIVOC 2007, 58–65. (g) Hei, X. M.; Song, Q. H.; Li,
X. B.; Tang, W. J.; Wang, H. B.; Guo, Q. X. J. Org. Chem. 2005, 70, 2522–
2527.
S
Supporting Information. Materials list, complete cita-
b
tion for ref 23, NMR spectra for compounds 2 and 3, NMR
spectra for temperature and concentration effects on the product
ratios, absolute energies, and optimized Cartesian coordinates.
This information is available free of charge via the Internet at
http://pubs.acs.org/.
’ AUTHOR INFORMATION
Corresponding Author
mabe@hiroshima-u.ac.jp
’ ACKNOWLEDGMENT
NMR and MS measurements were made using JEOL JMN-LA500
and Thermo Fisher Scientific LTD Orbitrap XL spectrometers,
respectively, at the Natural Science Center for Basic Research
and Development (N-BARD), Hiroshima University. M.A.
acknowledges financial support in the form of a Grant-in-Aid
for Scientific Research on Innovative Areas “π-Space” (No
21108516), the Scientific Research (B) (No. 19350021), and
Tokuyama Science Foundation.
’ REFERENCES
(1) (a) Aloisio, S.; Francisco, J. S. Acc. Chem. Res. 2000, 33, 825–830.
(b) Qi, X.-J.; Liu, L.; Fu, Y.; Guo, Q.-X. Struct. Chem. 2005, 16, 347–353.
(c) DeMatteo, M. P.; Poole, J. S.; Shi, X.; Sachdeva, R.; Hatcher, P. G.;
Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2005, 127, 7094–7109.
(d) Mardyukov, A.; Sanchez-Garcia, E.; Crespo-Otero, R.; Sander, W.
Angew. Chem., Int. Ed. 2009, 48, 4804–4807. (e) Hammeru, S. J. Am.
Chem. Soc. 2009, 131, 8627–8635. (f) Mitroka, S.; Zimmeck, S.; Troya,
D.; Tanko, J. M. J. Am. Chem. Soc. 2010, 132, 2907–2913.
(g) Mardyukov, A.; Crespo-Otero, R.; Sanchez-Garcia, E.; Sander, W.
Chem.—Eur. J. 2010, 16, 8679–8689. (h) Muthukrishman, S.; Sankar-
anarayanan, J.; Pace, T. C. S.; Konosonoks, A.; DeMichiei, M. E.; Meese,
M. J.; Bohne, C.; Gudmundsdottir, A. D. J. Org. Chem. 2010, 75,
1393–1401.
(2) (a) Jeffrey, G. A. In An Introduction to Hydrogen-Bonding; Oxford
University Press: New York, 1997. (b) Steiner, T. Angew. Chem., Int. Ed.
2002, 41, 48–76.
(3) (a) Johnson, R. A.; Sharpless, K. B. In Catalytic Asymmetric
Synthesis; Ojima, I., Ed.; VCH: New York, 1993; Chapter 4.1 (b) Adam,
W.; Wirth, T. Acc. Chem. Res. 1999, 32, 703–710. (c) Tripathy, R.;
Carroll, P. J.; Thornton, E. R. J. Am. Chem. Soc. 1990, 112, 6743–6744.
(d) Pearson, R. J.; Kassianidis, E.; Philp, D. Tetrahedron Lett. 2004, 45,
4777–4780. (e) Narayan, S.; Muldoon, J.; Finn, M. G.; Fokin, V. V.;
Kolb, H. C.; Sharpless, K. B. Angew. Chem., Int. Ed. 2005, 44, 3275–3279.
(4) The importance of hydrogen-bonding in photochemical reac-
tions, see:(a) Bauer, A.; Westkaemper, F.; Grimme, S.; Bach, T. Nature
2005, 436, 1139–1140. (b) Yokoyama, A.; Mizuno, K. Org. Lett. 2000, 2,
3457–3459. (c) Bach, T.; Bergmann, H.; Harms, K. J. Am. Chem. Soc.
1999, 121, 10650–10651. (d) Sieburth, S. McN.; McGee, K. F., Jr. Org.
Lett. 1999, 1, 1775–1777. (e) Sydnes, L. K.; Hansen, K. I.; Oldroyd,
D. L.; Weedon, A. C.; Jorgensen, E. Acta Chem. Scand. 1993, 47, 916–
924. (f) Hoffmann, N. Chem. Rev. 2008, 108, 1052–1103.
(5) Ng, S. M.; Bader, S. J.; Snapper, M. L. J. Am. Chem. Soc. 2006,
128, 7315–7319.
(6) (a) Paternꢀo, E.; Chieffi, G. Gazz. Chim. Ital. 1909, 39 (I), 341–
361. (b) B€uchi, G.; Inman, C. G.; Lipinsky, E. S. J. Am. Chem. Soc. 1954,
76, 4327–4331. (c) Yang, N. C.; Nussim, M.; Jorgenson, M. J.; Murov, S.
Tetrahedron Lett. 1964, 5, 3657–3664. (d) Arnold, D. R. Adv. Photochem.
1968, 6, 301–349.(e) Carless, H. A. Synthetic Organic Photochemistry;
Horspool, W. M., Ed.; Plenum Press: New York, 1984; p 425-487.
(f) Griesbeck, A. G. CRC Handbook of Organic Photochemistry and
(20) (a) Lee, M.; Moritomo, H.; Kanematsu, K. Tetrahedron 1996,
52, 8169–8180. (b) Lautens, M.; Fillion, E. J. Org. Chem. 1997, 62, 4418–
4427.
(21) (a) Head-Gordon, M.; Pople, J. A.; Frisch, M. J. Chem. Phys.
Lett. 1988, 153, 503–506. (b) Saebø, S.; Alml€of, A. Chem. Phys. Lett.
2603
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604

Journal of the American Chemical Society
ARTICLE
1989, 154, 83–89. (c) Frisch, M. J.; Head-Gordon, M.; Pople, J. A. Chem.
Phys. Lett. 1990, 166, 275–280. (d) Frisch, M. J.; Head-Gordon, M.;
Pople, J. A. Chem. Phys. Lett. 1990, 166, 281–289. (e) Head-Gordon, M.;
Head-Gordon, T. Chem. Phys. Lett. 1994, 220, 122–128.
(22) (a) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28,
213–222. (b) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (c) Lee,
C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789. (d) Foresman,
J. B.; Head-Gordon, M.; Pople, J. A.; Frisch, M. J. J. Phys. Chem. 1992, 96,
135–149.
(23) Frisch, M. J. et al. Gaussian 03, Revision C.02, Gaussian Inc.:
Wallingford, CT, 2004.
(24) (a) Abe, M.; Kawakami, T.; Ohata, S.; Nozaki, K.; Nojima, M.
J. Am. Chem. Soc. 2004, 126, 2838–2846. (b) Griesbeck, A. G.; Abe, M.;
Bondock, S. Acc. Chem. Res. 2004, 37, 919–928. (c) Abe, M. J. Chin.
Chem. Soc. (Taipei) 2008, 55, 479–486.(d) Abe, M. In Handbook of
Synthetic Photochemistry; Albini, A., Fagnoni, M., Eds.; Wiley-VCH,
Weinheim, 2010, 217-239.
(25) A-factor of OH (solvent), 0.60 (C6H12) and 0.95
(CH3CHOHCH3): Eliel, E. L.; Wilen, S. H. In Stereochemistry of Organic
Compounds; John Wiley & Sons: New York, 1994; pp 696-697.
(26) A PB reaction of benzophenone with the methyl ether of
2-furylmethanol was reported to produce no oxetane formation; see
ref 7c.
(27) Zimmerli, U.; Parrinello, M.; Koumoutsako, P. J. Chem. Phys.
2004, 120, 2693–2699.
(28) Triplet 1,4-biradicals are the common intermediate in the PB
reaction of the electron rich alkenes: (a) Freilich, S. C.; Peters, K. S.
J. Am. Chem. Soc. 1981, 103, 6255–6257. (b) Freilich, S. C.; Peters, K. S.
J. Am. Chem. Soc. 1985, 107, 3819–3822. (c) Palmer, I. J.; Ragazos, I. N.;
Berbardi, F.; Olivucci, M.; Robb, M. A. J. Am. Chem. Soc. 1994, 116,
2121–2132.
(29) Matsumura, K.; Mori, T.; Inoue, Y. J. Org. Chem. 2010, 75,
5461–5469.
(30) Kutateladze, A. G. J. Am. Chem. Soc. 2001, 123, 9279–9282.
(31) In The Anomeric Effect and Associated Stereoelectronic Effects;
ACS Symposium Series 539; Thatcher, G. R. J., Ed.; American Chemical
Society: Washington, DC, 1993.
2604
dx.doi.org/10.1021/ja1088524 |J. Am. Chem. Soc. 2011, 133, 2592–2604