Contrastive photoreduction pathways of benzophenones governed by regiospecific deprotonation of imidazoline radical cations and additive effects

Article abstract of DOI:10.1021/jo0514220

In the photoreaction of benzophenones with 1,3-dimethyl-2- phenylbenzimidazoline (DMPBI), benzhydrols were major products. Addition of H₂O accelerated the reaction with no change in the product distribution, while AcOH, PhOH, and metal salts such as LiClO₄ and Mg(ClO₄)₂ were effective additives to produce benzpinacols. In contrast, benzpinacols were exclusively formed regardless of the solvent and the additive in the reactions with 2-(o-hydroxyphenyl)-1,3- dimethylbenzimidazoline (o-HPDMBI). These observations are consistent with the hypothesis that DMPBI^.+ donates a proton at the C₂ position to the benzophenone ketyl radicals while o-HPDMBI^.+ donates a phenol proton.

Full text of DOI:10.1021/jo0514220

CHART 1
Contrastive Photoreduction Pathways of
Benzophenones Governed by Regiospecific
Deprotonation of Imidazoline Radical
Cations and Additive Effects
Eietsu Hasegawa,*,† Takayuki Seida,† Naoki Chiba,†
Tomoya Takahashi,† and Hiroshi Ikeda‡
Department of Chemistry, Faculty of Science, Niigata
University, Ikarashi-2 8050, Niigata 950-2181, Japan, and
Department of Chemistry, Graduate School of Science,
Tohoku University, Sendai 980-8578, Japan
ehase@chem.sc.niigata-u.ac.jp
Received July 10, 2005
1,3-dimethylbenzimidazolines (HPDMBI), shown in Chart
1, act as effective reagents to promote the photoinduced
reduction of ketones. The synthetic utilities of these
benzimidazolines are thus far recognized; however, their
reaction behaviors in PET systems has been less ex-
plored. Therefore, we investigated photoreactions of
benzophenone (1a), a simple but useful mechanistic probe
substrate in various PET reactions,7 and m-methylben-
zophenone (1b) with DMPBI and 2-(o-hydroxyphenyl)-
1,3-dimethylbenzimidazoline (o-HPDMBI). Here, we re-
port contrastive photoreduction pathways of 1 governed
by regiospecific deprotonation of the radical cations of
DMPBI and o-HPDMBI and by properties of the proton
donors and metal salts.
In the photoreaction of benzophenones with 1,3-dimethyl-
2-phenylbenzimidazoline (DMPBI), benzhydrols were major
products. Addition of H2O accelerated the reaction with no
change in the product distribution, while AcOH, PhOH, and
metal salts such as LiClO4 and Mg(ClO4)2 were effective
additives to produce benzpinacols. In contrast, benzpinacols
were exclusively formed regardless of the solvent and the
additive in the reactions with 2-(o-hydroxyphenyl)-1,3-
dimethylbenzimidazoline (o-HPDMBI). These observations
are consistent with the hypothesis that DMPBI•+ donates a
proton at the C2 position to the benzophenone ketyl radicals
while o-HPDMBI•+ donates a phenol proton.
(2) (a) Yoon, U. C.; Marinao, P. S.; Givens, R. S.; Atwater, B. W.,
III. In Advances in Electron Transfer Chemistry; Mariano, P. S., Ed.;
JAI Press: Greenwich, CT, 1994; Vol. 4, p 117. (b) Lewis, F. D. In
Advances in Electron Transfer Chemistry; Mariano, P. S., Ed.; JAI
Press: Greenwich, CT, 1996; Vol. 5, p 1. (c) Das, S.; Suresh, V. In
Electron Transfer in Chemistry; Balzani, V., Ed.; Wiley-VCH: Wein-
heim, Germany, 2001; Vol. 2, p 379.
(3) (a) Berger, D. J.; Tanko, J. M. In The Chemistry of Double-
Bonded Functional Groups; Patai, S., Ed.; Wiley: New York, 1997; p
1281. (b) Griesbeck, A. G.; Shieffer, S. In Electron Transfer in
Chemistry; Balzani, V., Ed.; Wiley-VCH: Weinheim, Germany, 2001;
Vol. 2, p 457. (c) Hasegawa, E. J. Photosci. 2003, 10, 61.
(4) (a) Hasegawa, E.; Kato, T.; Kitazume, T.; Yanagi, K.; Hasegawa,
K.; Horaguchi, T. Tetrahedron Lett. 1996, 37, 7079. (b) Hasegawa, E.;
Tamura, Y.; Tosaka, E. J. Chem. Soc., Chem. Commun. 1997, 1895.
(c) Hasegawa, E.; Yoneoka, A.; Suzuki, K.; Kato, T.; Kitazume, T.;
Yanagi, K. Tetrahedron 1999, 55, 12957. (d) Hasegawa, E.; Chiba, N.;
Nakajima, A.; Suzuki, K.; Yoneoka, A.; Iwaya, K. Synthesis 2001, 1248.
(e) Hasegawa, E.; Takizawa, S.; Iwaya, K.; Kurokawa, M.; Chiba N.;
Yamamichi, K. J. Chem. Soc., Chem. Commun. 2002, 1966.
Photoinduced electron-transfer (PET) processes provide
useful methods to generate organic radical ions in solu-
tion, and the reaction pathways of radical ions are
principally governed by not only their inherent property
but also the surrounding medium including solvents and
additives.1 Because pairs of radical ions are uniquely
generated with the PET methods in contrast to other
nonphotochemical methods, specific interactions between
the radical ion pairs are essentially important in the PET
reaction systems.1 Among the compounds that have been
frequently subjected to PET reactions are amines2 and
ketones.3 Previously, we reported4,5 that 1,3-dimethyl-2-
phenylbenzimidazoline (DMPBI)6 and 2-hydroxyphenyl-
(5) Hasegawa, E.; Chiba, N.; Takahashi, T.; Takizawa, S.; Kitayama,
T.; Suzuki, T. Chem. Lett. 2004, 33, 18.
(6) (a) Chikashita, H.; Itoh, K. Bull. Chem. Soc. Jpn. 1986, 59, 1747.
(b) Chikashita, H.; Ide, H.; Itoh, K. J. Org. Chem. 1986, 51, 5400. (c)
Chen, J.; Tanner, D. D. J. Org. Chem. 1988, 53, 3897. (d) Tanner, D.
D.; Chen, J. J. J. Org. Chem. 1989, 54, 3842. (e) Tanner, D. D.; Chen,
J. J.; Chen, L.; Luelo, C. J. Am. Chem. Soc. 1991, 113, 8074.
(7) (a) Cohen, S. G.; Palora, A.; Parsons, G. H., Jr. Chem. Rev. 1973,
73, 141 and references therein. (b) Hoshino, M.; Shizuka, H. In
Photoinduced Electron Transfer; Fox, M. A., Chanon, M., Eds.;
Elsevier: Amsterdam, 1988; Part C, p 313. (c) Davidson, R. S. J. Chem.
Soc., Chem. Commun. 1966, 575. (d) Pac, C.; Sakurai, H.; Tosa, T. J.
Chem. Soc., Chem. Commun. 1970, 1311. (e) Arimitsu, S.; Masuhara,
H.; Mataga, N.; Tsubomura, H. J. Phys. Chem. 1975, 79, 1255. (f) Inbar,
S.; Linshitz, H.; Cohen, S. G. J. Am. Chem. Soc. 1981, 103, 1048. (g)
Simon, J. D.; Peters, K. S. J. Am. Chem. Soc. 1981, 103, 6403. (h)
Simon, J. D.; Peters, K. S. J. Am. Chem. Soc. 1982, 104, 6542. (i) Simon,
J. D.; Peters, K. S. J. Am. Chem. Soc. 1983, 105, 4875. (j) Devadoss,
C.; Fessenden, R. W. J. Phys. Chem. 1991, 95, 7253. (k) Jones, P. B.;
Pollastri, M. P.; Porter, N. A. J. Org. Chem. 1996, 61, 9455. (l)
Reynolds, J. L.; Erdner, K. R.; Jones, P. B. Org. Lett. 2002, 4, 917.
† Niigata University.
‡ Tohoku University.
(1) (a) Eberson, L. Electron Transfer Reactions in Organic Chemistry;
Springer-Verlag: Berlin, 1987. (b) Photoinduced Electron Transfer; Fox,
M. A., Chanon, M., Eds.; Elsevier: Amsterdam, 1988; Parts A-D. (c)
Kavarnos, G. J. Fundamental of Photoinduced Electron Transfer; VCH
Press: New York, 1993. (d) Advances in Electron Transfer Chemistry;
Mariano, P. S., Ed.; JAI Press: Greenwich, CT, 1991-1999; Vols. 1-6.
(e) Electron Transfer in Chemistry; Balzani, V., Ed.; Wiley-VCH:
Weinheim, Germany, 2001; Vols. 1-5.
10.1021/jo0514220 CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/18/2005
9632
J. Org. Chem. 2005, 70, 9632-9635

TABLE 1. Photoreaction of 1b with DMPBI
SCHEME 1
yieldsc,d (%)
additive
(equiv vs 1b)
conv of 1bc
entry solvent
(%)
2b
3b
1a
2a
3a
4a
5a
6a
7b
8b
9b
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
DMF
-
37
88
88
88
8
0
H2O (13.9)
PhOH (6.0)
AcOH (2.4)
LiClO4 (1.2)
Mg(ClO4)2 (1.2)
-
0
92
100
86
91
0
100
100
100
89
0
8
7
4
0
DMF
H2O (13.9)
AcOH (6.6)
14
100
6
0
86
DMF
56
a 1b (0.20 mmol), MeCN (4 mL), λ > 280 nm for 2 h for entries
1-6. b 1b (0.20 mmol), BDMAP (0.010 mmol), DMF (2 mL), λ >
360 nm for 4 h for entries 7-9. c Determined with 1H NMR.
d Based on the conversion of 1b.
We first conducted a preparative photoreaction (λ >
280 nm) of 1a and DMPBI (1.2 equiv versus 1a) in THF
and found low conversion of 1a (21%) and predominant
formation of benzhydrol (2a) (78% yield).8,9 Although
addition of H2O (13.9 equiv versus 1a) increased the
conversion of 1a to 48%, 2a was still the sole product
(85%). When H2O was replaced with AcOH (6.6 equiv
versus 1a), however, a significant amount of benzpinacol
(3a, 47%) was obtained along with 2a (53%) at 65%
conversion of 1a. To investigate in detail the photoreac-
tion of benzophenone-DMPBI systems, we then con-
ducted photoreactions of 1b and DMPBI under various
conditions, as shown in Table 1, in which the quantities
of 1b, 2b, and 3b in the reaction mixtures were easily
[DMPBI•+ 1•-].14 Subsequent proton transfer (PT) pro-
duces a radical pair [5• and 4•] that successively under-
goes SET to give a DMPBI-derived cation (5+) and a
carbanion (4-).15 Protonation to 4- gives 2 in the presence
of H2O. Feasibility of the SET process between 5• and 4•
could be consistent with the fact that the formation of 2
was predominant in the case of DMPBI and not in the
cases of simple tertiary amines.16 In the presence of
AcOH and PhOH, these proton donors intercept 1•- with
protonation to give 4•.18 Because the potential reducing
agent 5• is not generated in this case, 4• survives to
undergo dimerization with another 4• to 3. Ion pair
exchange7i between [DMPBI•+ 1•-] and the metal salt,
represented by LiClO4 in Scheme 1, also prevents PT
from DMPBI•+ to 1•- to give the metal-coordinated ketyl
radical (Li+1•-) that dimerizes to 6, which finally converts
to 3.
1
determined with H NMR based on the signals of the
methyl protons of 1b, 2b, and 3b. In the reactions with
the proton donors, the conversion of 1b was significantly
increased in the presence or absence of 1,6-bis(N,N′-
dimethylamino)pyrene (BDMAP)4b-e,5,10 as a sensitizer
(entries 2-4 or 8 and 9).11 Addition of the proton donors
also changed the yields of 2b and 3b in varying extent,
apparently depending on the acidity of proton donors
12
(pKa in water: H2O 15.75; PhOH 10.0; AcOH 4.75)
(entries 2-4 and 8 and 9). Note that addition of metal
salts such as LiClO4 and Mg(ClO4)2 markedly changed
the product distribution (entries 5 and 6).13 Interestingly,
photoreactions of 1b with triethylamine, N,N-diethylani-
line, N,N-dimethylaniline, or tribenzylamine in aqueous
MeCN predominantly produced 3b (59-95%) along with
a small amount of 2b (2-11%).
A plausible reaction mechanism for the reaction of 1
with DMPBI is proposed in Scheme 1. Light-induced
single electron transfer (SET) produces a radical ion pair
(14) Feasibility of electron transfers between 1 and amines was
confirmed by discussion using their redox potential data. For details,
see the Supporting Information.
(15) One reviewer suggested another possible pathway in which the
(8) A preliminary observation that the irradiation of 1a with DMPBI
predominantly produced 2a in aqueous MeCN was reported.4c
(9) Yields of the products reported in this Note represent those based
on the conversion of the corresponding starting materials unless
otherwise stated.
(10) Okada, K.; Okamoto, K.; Oda, M. J. Am. Chem. Soc. 1988, 110,
8736.
(11) Similar marked change of the product distribution depending
on the proton donors was also observed in the BDMAP-sensitized
photoreaction of cyclohexylphenyl ketone with DMPBI in DMF (alcohol/
pinacol ) 100:0 for H2O and 10:90 for AcOH).
(12) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456.
(13) Although the influence of alkali metal salts on the PET
processes of 1a and amines has been thoroughly investigated,7i
surprisingly, the influence of the salts on the product formation has
not been explored well.
ion radical pair [DMPBI•+ •-] is converted to an ion pair [5+ PhArCH-
1
O-] by a hydrogen atom transfer. This pathway must be a minor one,
even if it exists, because a hydrogen atom transfer from DMPBI•+ to
the sterically more crowded ketyl carbon of 1•- should be less likely
than a proton transfer to the oxy anion part of 1•-
.
(16) Electron-donating ability of 5• is considered to be greater than
those of R-amino radicals derived from simple tertiary amines on the
basis of the oxidation potentials of the benzthiazolyl radical that
possesses a structure similar to that of 5• (-1.3 V versus SCE)17a and
those of simple R-amino radicals (ca. -1.0 V versus SCE).17b
(17) (a) Shukla, D.; Liu, G.; Dinnocenzo, J. P.; Farid, S. Can. J.
Chem. 2003, 81, 744. (b) Wayner, D. D. M.; McPhee, D. J.; Griller, D.
J. Am. Chem. Soc. 1988, 110, 132.
(18) Because the pKa value of the protonated ketyl radical of 1a in
water has been reported to be 9.3,19 the protonation to 1•- by AcOH or
PhOH must be a reasonably efficient process.
J. Org. Chem, Vol. 70, No. 23, 2005 9633

TABLE 2. Photoreaction of 1b with o-HPDMBI
SCHEME 2
yieldsc,d (%)
additive
(equiv vs 1b)
conv of 1bc
entry solvent
(%)
2b
3b
1a
2a
3a
4a
5a
6a
7a
8a
9b
10b
MeCN
MeCN
MeCN
MeCN
MeCN
THF
-
52
47
0
0
0
0
0
0
0
8
0
0
100
100
100
100
100
100
96
H2O (13.9)
AcOH (6.6)
LiClO4 (1.2)
Mg(ClO4)2 (1.2)
-
100
100
100
86
THF
AcOH (6.6)
-
85
DMF
54
91
DMF
-
34
100
93
DMF
AcOH (6.6)
79
a 1b (0.20 mmol), solvent (4 mL), λ > 280 nm for 2 h for entries
1-8. b 1b (0.20 mmol), BDMAP (0.010 mmol), DMF (2 mL), λ >
1
360 nm for 4 h for entries 9 and 10. c Determined with H NMR.
d Based on the conversion of 1b.
We next conducted photoreactions of 1b with o-HPD-
MBI and found a marked change in product distributions
from those with DMPBI. As shown in Table 2, 3b was
exclusively produced in high yield regardless of the
reaction conditions. Particularly, the exclusive formation
of 3b was contrastive to the predominant formation of
2b with DMPBI in the presence of H2O. Significant
conversion of 1b, even in the absence of proton donors,
was also notable (entries 1, 6, and 8). Interestingly,
addition of AcOH significantly increased the conversion
of 1b in MeCN and DMF (compare entries 1 and 9 with
3 and 10, respectively); however, the conversion of 1b did
not change very much with or without AcOH in THF
(entries 6 and 7). Again, both LiClO4 and Mg(ClO4)2
increased the conversion of 1b and led to the exclusive
formation of 3b.
To gain further insight into the effects of the hydroxy
substituent on the phenyl group at the C2 position of
DMPBI, we conducted photoreactions of 1b with p-
HPDMBI or 2-(p-methoxyphenyl)-1,3-dimethylbenzimid-
azoline (ADMBI) in MeCN. The reaction with p-HPDMBI
produced 3b (94%) at 52% conversion of 1b, whereas the
reaction with ADMBI in the presence of H2O (13.9 equiv
versus 1b) produced 2b (98%) at 100% conversion of 1b.
The fact that p-HPDMBI behaves similarly to o-HPDMBI
rather than ADMBI suggests that electron-donating
properties of the hydroxy substituents of o- and p-
HPDMBIs are not crucial factors, but their proton-
donating abilities are responsible for the exclusive for-
mation of 3b.
proton donor and from o-HPDMBI•+ are competitive. This
situation should be more plausible in polar solvents than
in nonpolar solvents because a dissociation of [o-HPD-
MBI•+ 1•-] is most likely to occur in the former solvents.
Note that the addition of AcOH significantly increased
the conversion of 1b in both MeCN and DMF, but not in
THF (compare entries 1, 9, and 6 to 3, 10, and 7,
respectively, in Table 2). The above hypothesis of the
deprotonation regioselectivity of o-HPDMBI•+ is particu-
larly noteworthy because the proton at the C2 position
of o-HPDMBI•+ (Hb) is considered to be more acidic than
the phenol proton (Ha) judging from their expected
acidities.22 Therefore, it must be assumed that the order
of their kinetic acidities25 is reverse to that of the
corresponding thermodynamic acidities in these reaction
systems.
In conclusion, we have found that the photoproduct
distributions from 1 markedly changed depending on
DMPBI or o-HPDMBI and also were influenced by the
property of the additives. The contrastive observations
between DMPBI and o-HPDMBI are consistent with the
hypothesis that DMPBI•+ donates a proton at the C2
On the basis of these observations, we propose a
plausible mechanism for the reaction of o-HPDMBI and
1, as shown in Scheme 2. The initially generated o-
HPDMBI•+ releases either a phenol proton (Ha) or a
(20) Reducing abilities of the phenoxide ions should be much lower
than those of R-amino radicals,16 judging form their oxidation poten-
tials. The oxidation potential of the phenoxide ion was reported to be
ca. 0 V versus SCE.21
(21) Yousefzadeh, P.; Man, C. K. J. Org. Chem. 1968, 33, 2716.
(22) Acidities of some amine radical cations were reported.23 For
example, the pKa of the radical cation of N,N-dimethylaniline was
estimated to be 9 in MeCN.23a The pKa value of phenol in MeCN was
estimated to be 27.12,24
(23) (a) Nelsen, S. F. J. Am. Chem. Soc. 1988, 108, 4879. (b)
Dinnocenzo, J. P.; Banach, T. E. J. Am. Chem. Soc. 1989, 111, 8646.
(c) Parker, V. D.; Tilset, M. J. Am. Chem. Soc. 1991, 113, 8778.
(24) Liu, W. Z.; Bordwell, F. G. J. Org. Chem. 1996, 61, 4778.
(25) The concept of kinetic acidity has been often applied to
rationalize the reaction of radical ion pairs generated by PET pro-
cesses.26 Then it was suggested that the kinetic acidity of a certain
radical cation is influenced by the nature of a radical anion generated
simultaneously.
+
proton at the C2 position (Hb). If Hb is transferred to
1•-, the resulting radical 7• would reduce 4• to 4-, which
finally becomes 2 similarly to the reaction with DMPBI
(see Scheme 1). However, this was not witnessed. In-
stead, o-HPDMBI•+ donates Ha+ to 1•-, and the phenoxide
part20 of the resulting zwitter ion radical 8•+- does not
reduce 4• to 4- so that 4• has an opportunity to undergo
dimerization to 3. When a proper proton donor is added
to the reaction system, protonations to 1•- from the
(19) Lilie, J.; Henglein, A. Ber. Bunsen-Ges. Phys. Chem. 1969, 73,
170.
9634 J. Org. Chem., Vol. 70, No. 23, 2005

position to 1•-, whereas o-HPDMBI•+ donates a phenol
proton to 1•- rather than a proton at the C2 position.
Investigation of the factors to control the regioselectivity
of deprotonation of o-HPDMBI•+ is the next subject to
be undertaken.
Quantities of 1b, 2b, and 3b were determined on the basis of
integrations of the methyl peaks of these compounds (2.42 ppm
for 1b, 2.34 ppm for 2b, 2.19 and 2.20 ppm for the two
diastereomers of 3b) and the methine peak of triphenylmethane
(5.56 ppm).
Acknowledgment. E.H. and H.I. thank the Minis-
try of Education, Culture, Sports, Science, and Technol-
ogy of Japan for financial support in the form of Grants-
in-Aid for Scientific Research C (No. 15550028) and for
Scientific Research in Priority Areas (Area No. 417),
respectively. E.H. acknowledges financial support from
the Uchida Energy Science Promotion Foundation. E.H.
thanks Professor J. P. Dinnocenzo (University of Roch-
ester) for his valuable comments. Both authors thank
Professors T. Horaguchi (Niigata University) and M.
Ueda (Tohoku Univeristy) for their generous assistance.
Experimental Section
General Procedure for Photoreaction of 1b with Amines.
A solution of 1b (0.20 mmol) and amine (0.24 mmol) with or
without BDMAP (0.010 mmol) in a solvent (MeCN, THF, and
DMF 4 mL; DMF 2 mL for the BDMAP sensitized reactions) in
the presence or absence of H2O, PhOH, AcOH, LiClO4, or Mg-
(ClO ) was purged with N2 for 5 min prior to irradiation. The
2
4
irradiation and workup procedures were essentially the same
as those of the reactions of 1a described in the Supporting
Information. The photolyzates containing PhOH, LiClO4, and
Mg(ClO4)2 were worked up in the same manner to that contain-
ing AcOH. Each photoproduct mixture was analyzed with 1H
NMR (200 MHz) using triphenylmethane as an internal stan-
dard. Benzhydrol 2b27 and benzpinacol 3b28 were identified by
comparison of their spectral data with those of the products
obtained by the NaBH4 and SmI2 reductions of 1b, respectively.
Supporting Information Available: General experimen-
tal procedures, preparations of o-HPDMBI, p-HPDMBI, and
ADMBI, their IR and NMR spectra and cyclic voltammogram,
photoreaction procedure for 1a, and additional discussion. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(26) (a) Lewis, F. D.; Ho, T. I.; Simpson, J. T. J. Org. Chem. 1981,
46, 1077. (b) Lewis, F. D.; Ho, T. I.; Simpson, J. T. J. Am. Chem. Soc.
1982, 104, 1924. (c) Xu, W.; Zhang, X. M.; Mariano, P. S. J. Am. Chem.
Soc. 1991, 113, 8863. (d) Zhang, X.; Yeh, S. R.; Hong, S.; Freccero, M.;
Albini, A.; Falvey, D. A.; Mariano, P. S. J. Am. Chem. Soc. 1994, 116,
4211.
JO0514220
(27) Guijarro, D.; Yus, M. Tetrahedron 2000, 56, 1135.
(28) Yamamoto, J.; Matsuo, A.; Okamoto, Y. Nippon Kagaku Kaishi
1988, 1709.
J. Org. Chem, Vol. 70, No. 23, 2005 9635