Redox reaction of benzyl radicals with aromatic radical ions photogenerated. The Marcus inverted region and the selective formation of carbocations or carbanions

Article abstract of DOI:10.1021/ja953689u

Efficient redox reactions of benzyl-type radicals were achieved by irradiating an aromatic donor/acceptor pair with substituted dibenzyl ketones as a radical precursor in MeOH-MeCN. In this system, the aromatic radical ion pair was generated by photoinduced electron transfer followed by one-electron oxidation and reduction of photogenerated benzyl radicals (R·) by the radical ions to afford benzyl cations (R⁺) and anions (R^-). The cations and anions were trapped by MeOH to yield ROMe and RH, respectively. The relative product ratios were determined for a variety of donor-acceptor pairs, reflecting the relative efficiencies of the redox reaction of benzyl radicals with a steady-state concentration of radical ions. The selective formation of carbocations or carbanions was achieved in some sets of donor/acceptor pairs. Assuming that the radical ions exist in a 1:1 ratio in the steady state, the product ratios are equal to the relative electron transfer rates, which are analyzed in terms of the free energy changes of the processes. The present results indicated that the rates became maximal at the energy gap of ca. -0.5 eV, representing a novel example of the Marcus inverted region. This interpretation is discussed in comparison with other related cases and in relation to recent theories on electron transfer rates.

Full text of DOI:10.1021/ja953689u

J. Am. Chem. Soc. 1996, 118, 7255-7264
7255
Redox Reaction of Benzyl Radicals with Aromatic Radical Ions
Photogenerated. The Marcus Inverted Region and the Selective
Formation of Carbocations or Carbanions
Katsuya Ishiguro, Tomoharu Nakano, Hiroki Shibata, and Yasuhiko Sawaki*
Contribution from the Department of Applied Chemistry, Faculty of Engineering, Nagoya
UniVersity, Chikusa-ku, Nagoya 464-01, Japan
X
ReceiVed NoVember 1, 1995
Abstract: Efficient redox reactions of benzyl-type radicals were achieved by irradiating an aromatic donor/acceptor
pair with substituted dibenzyl ketones as a radical precursor in MeOH-MeCN. In this system, the aromatic radical
ion pair was generated by photoinduced electron transfer followed by one-electron oxidation and reduction of
•
+
-
photogenerated benzyl radicals (R ) by the radical ions to afford benzyl cations (R ) and anions (R ). The cations
and anions were trapped by MeOH to yield ROMe and RH, respectively. The relative product ratios were determined
for a variety of donor-acceptor pairs, reflecting the relative efficiencies of the redox reaction of benzyl radicals
with a steady-state concentration of radical ions. The selective formation of carbocations or carbanions was achieved
in some sets of donor/acceptor pairs. Assuming that the radical ions exist in a 1:1 ratio in the steady state, the
product ratios are equal to the relative electron transfer rates, which are analyzed in terms of the free energy changes
of the processes. The present results indicated that the rates became maximal at the energy gap of ca. -0.5 eV,
representing a novel example of the Marcus inverted region. This interpretation is discussed in comparison with
other related cases and in relation to recent theories on electron transfer rates.
Organic electron transfer (ET) reactions proceed mostly via
an initial one-electron transfer (E) and the following reaction
(C) of radical ion intermediates. When the primary products
from radical ions still have an open-shell radical structure,
secondary electron transfers leading to non-radical products (eqs
1
and 2) are frequently involved. Such a sequence is known
as an ECE process and is very common in electrochemical
reactions.
of photosensitizers.5 However, little attention has been directed
1
toward the secondary redox processes because of masking by
the prior ET step or chemical reactions.
6
According to the Marcus theory, the electron-transfer rate
depends on the free energy change of the process (∆G).
Especially important is that the rate reaches maximum when
the energy gap (-∆G) equals the reorganization energy (λ) but
decreases with increasing exothermicity when -∆G > λ (the
Marcus inverted region). Since electron transfers between
radicals and radical ions are highly exothermic, consideration
of the inverted region might be very important.
Redox reactions of radical species are also involved in many
photosensitized ET reactions. Typically, a back electron
2
transfer between the radical and sensitizer radical ions, by which
the sensitizer is regenerated, is involved in many photocatalytic
Although actual examples for the Marcus inverted region have
3
ET processes. In the oxidative bond cleavage (eq 3) or in the
7
-19
been demonstrated recently,
many other cases in photo-
4
addition of nucleophiles to olefin radical cations (eq 4), the
efficiencies of the last steps are essential in determining
products. In these reactions, the redox process of organic
radicals (eqs 1 and 2) is of great importance and care must be
taken in selecting the applied electrode potential or on the choice
(6) (a) Marcus, R. A. J. Chem. Phys. 1956, 24, 966. (b) Marcus, R. A.
Annu. ReV. Phys. Chem. 1964, 15, 155. (c) Eberson, L. In AdVances in
Physical Organic Chemistry; Gold, V., Bethell, D., Eds.; Academic Press:
London, 1982; Vol. 18, p 10.
(7) Miller, J. R.; Calcaterra, L. T.; Closs, G. L. J. Am. Chem. Soc. 1984,
06, 3047.
1
X
Abstract published in AdVance ACS Abstracts, July 15, 1996.
(8) Wasielewski, M. R.; Niemczyk, M. P.; Svec, W. A.; Pewitt, E. B. J.
Am. Chem. Soc. 1985, 107, 1080.
(1) (a) Fry, A. J. In Synthetic Organic Electrochemistry; Wiley: New
York, 1989; p 60. (b) Amatore, C. In Organic Electrochemistry; Lund, H.,
Baizer, M. M., Eds.; Marcel Dekker: New York, 1991; p 76.
(9) (a) Beitz, J. V.; Miller, J. R. J. Chem. Phys. 1979, 71, 4579. (b)
Miller, J. R.; Beitz, J. V.; Huddleston, R. K. J. Am. Chem. Soc. 1984, 106,
5057.
(10) Levin, P. P.; Pluzhnikov, P. F.; Kuzmin, V. A. Chem. Phys. Lett.
1988, 147, 283.
(
2) (a) Kavarnos, G. J.; Turro, N. J. Chem. ReV. 1986, 86, 401. (b)
Photoinduced Electron Transfer; Fox, M. A., Chanon, M., Eds.; Elsevier:
Amsterdam, 1988.
(
3) (a) Arnold, D. R.; Maroulis, A. J. J. Am. Chem. Soc. 1976, 98, 5931.
(11) (a) Gould, I. R.; Ege, D.; Mattes, S. L.; Farid, S. J. Am. Chem. Soc.
1987, 109, 3794. (b) Gould, I. R.; Moser, J. E.; Ege, D.; Farid, S. J. Am.
Chem. Soc. 1988, 110, 1992. (c) Gould, I. R.; Moser, J. E.; Armitage, B.;
Farid, S.; Goodman, J. L.; Herman, M. S. J. Am. Chem. Soc. 1989, 111,
1917. (d) Gould, I. R.; Ege, D.; Moser, J. E.; Farid, S. J. Am. Chem. Soc.
1990, 112, 4290.
(
(
b) Okamoto, A.; Snow, M. S.; Arnold, D. A. Tetrahedron 1986, 42, 6175.
c) Arnold, D. R.; Lamont, L. Can. J. Chem. 1989, 67, 2119. (d) Popielarz,
R.; Arnold, D. R. J. Am. Chem. Soc. 1990, 112, 3068. (e) Albini, A.; Mella,
M.; Freccero, M. Tetrahedron 1994, 50, 575.
(4) (a) Shigemitsu, Y.; Arnold, D. R. J. Chem. Soc., Chem. Commun.
1
975, 407. (b) Gassman, P. G.; Bottorff, K. J. J. Am. Chem. Soc. 1987,
(12) (a) Ohno, T.; Yoshimura, A.; Mataga, N. J. Phys. Chem. 1986, 90,
3295. (b) Mataga, N.; Asahi, T.; Kanda, Y.; Okada, T.; Kakitani, T. Chem.
Phys. 1988, 127, 249.
1
09, 7547.
(5) Gassman, P. G.; De Silva, S. A. J. Am. Chem. Soc. 1991, 113, 9870.
S0002-7863(95)03689-4 CCC: $12.00 © 1996 American Chemical Society

7256 J. Am. Chem. Soc., Vol. 118, No. 31, 1996
Ishiguro et al.
Scheme 1
takes place without any chemical reactions. In such a case,
simultaneous generation of a transient radical (R ) may bring
•
about its redox reaction with the radical ions, leading to the
+
formation of the corresponding carbenium ion (R ) and car-
-
banion (R ). In the presence of MeOH, the cation and anion
are trapped by MeOH affording ROMe and RH, respectively,
the ratio of which may reflect the relative rate for the electron
transfer. We have employed diphenylmethyl and substituted
benzyl radicals, since they are stable enough to give the
28
corresponding coupling products in high yields and their redox
2
7
potentials are already known. The dibenzyl ketones seem to
be suitable as the precursor, since the photocleavage leading to
two benzyl radicals proceeds quite efficiently (Φ ≈ 0.7) and
nonsymmetrical dibenzyl ketones affords a statistical mixture
of coupling products, indicating that truly “free” radicals are
chemical processes do not show such a region, as exemplified
in the well-known fluorescence quenching study by Rehm and
2
0
Weller. In such cases, formation of radical ions in the low-
lying excited state has been suggested as the reason for the
2
9
•+
•-
21
generated.
If appropriate amounts of D and A are
absence of the inverted region. In fact, in all cases showing
the Marcus inverted region,7
-19
generated by adjusting initial concentrations of donor D and
acceptor A, the photochemical formation of oxidant and
the electron transfer occurs
between radical species affording closed-shell species where
excited products are unlikely to be formed. The inverted region
is not commonly observed for intermolecular electron transfers,
3
0
reductant ions may be achieved.
We wish to report here the results on the simultaneous
irradiation of donor, acceptor, and ketone. The redox reaction
of radicals with aromatic radical ions has been investigated from
the product ratios of ROMe/RH for a various combination of
radical/donor/acceptor. The selective formation of carbocations
and carbanions was achieved in some sets of donor/acceptor
pairs, and the relative rates were discussed on the basis of the
Marcus theory.
even in the ET’s between radical species.2
2-24
Recently, new
explanations have been proposed for the lack of the inverted
region, one of which deals with a dielectric saturation effect of
2
5
solvents. According to this, the ET rate depends on the type
of ET, i.e., charge separation, shift, or recombination. Another
theory has pointed out the dependence of donor-acceptor
distance on the reorganization energy,26 in which the ET rate is
essentially independent of the substrate character or the type of
ET. Thus, the systematic study of the electron transfer of radical
species is interesting and crucial for the consideration of such
theories.
Results and Discussion
Products of the Co-irradiation of sym-Tetraphenylacetone,
Naphthalene, and p-Dicyanobenzene. The irradiation of sym-
tetraphenylacetone (1) is known to generate effectively diphen-
ylmethyl radical 2 (eq 5), affording 1,1,2,2-tetraphenylethane
(5) as the radical-coupling product in high yield. However, the
irradiation (>290 nm) of 0.3 mM 1 in MeOH-MeCN (1:9) in
the presence of 10.0 mM each of naphthalene (N) and
p-dicyanobenzene (DCB) under Ar for 2 h resulted in the
formation of diphenylmethyl methyl ether (3) and diphenyl-
Recent development of the photomodulation voltammogram
allows the precise determination of redox potentials for various
radical intermediates.27 However, the direct observation of such
redox reactions of radicals with various oxidants and reductants
is quite difficult because of involvement of unstable species of
short lifetimes. Then, in order to study the ET of radicals with
aromatic radical ions, we have examined a redox system as
shown in Scheme 1. If photogenerated radical ions are stable
enough so that the radical cation/anion pair exists as free ions
in a 1:1 ratio during the irradiation, only a back electron transfer
methane (4) in 3.3 and 49% yields, respectively, in addition to
ox
a 22% yield of 5. In this system, naphthalene (E
1
/2
) +1.80
1
1d
31a
V vs SCE)
also absorbs light
and its fluorescence is
red
2
quenched by DCB (E1/2 ) -1.60 V vs SCE) yielding the
(
3.
13) Vauthey, E.; Suppan, P.; Haselbach, E. HelV. Chim. Acta 1988, 71,
•
+
•- 31b
corresponding radical ions N and DCB .
The formation
9
(
14) Kikuchi, K.; Takahashi, Y.; Koike, K.; Wakamatsu, K.; Ikeda, H.;
of 3 and 4 suggests the one-electron oxidation and reduction of
Miyashi, T. Z. Phys. Chem. (Munich) 1990, 167, 27.
ox
red
27a
2
(E1/2 ) +0.35 V and E1/2 ) -1.14 V vs SCE) to yield
(
(
15) Grampp, G.; Hetz, G. Ber. Busenges. Phys. Chem. 1992, 96, 198.
16) (a) Gould, I. R.; Moody, R.; Farid, S. J. Am. Chem. Soc. 1988,
carbenium ion 6 and carbanion 7 as shown in eq 6.
1
10, 7242. (b) Gould, I. R.; Young, R. H.; Moody, R. E.; Farid, S. J. Phys.
Chem. 1991, 95, 2068.
17) (a) Asahi, T.; Mataga, N. J. Phys. Chem. 1989, 93, 6575. (b) Asahi,
T.; Mataga, N. J. Phys. Chem. 1991, 95, 1956.
18) Zou, C.; Miers, J. B.; Ballew, R. M.; Dlott, D. D.; Schuster, G. B.
J. Am. Chem. Soc. 1991, 113, 7823.
19) (a) DeCosta, D. P.; Pincock, J. A. J. Am. Chem. Soc. 1989, 111,
948; 1993, 115, 2180. (b) Hilborn, J. W.; MacKnight, J. A.; Pincock, J.
A.; Wedge, P. J. J. Am. Chem. Soc. 1994, 116, 3337.
20) Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 259. (b) Nishikawa,
S.; Asahi, T.; Okada, T.; Mataga, N.; Kakitani, T. Chem. Phys. Lett. 1991,
85, 237.
21) (a) Suppan, P. Top. Curr. Chem. 1992, 163, 95. (b) Grampp, G.
Angew. Chem., Int. Ed. Engl. 1993, 32, 691.
22) Schomberg, H.; Staerk, H.; Weller, A. Chem. Phys. Lett. 1973, 21,
33; 1973, 22, 2735.
(
(
(
8
(
1
(
(
4
(
(
(
23) Gschwind, R.; Haselbach, E. HelV. Chim. Acta 1979, 62, 941.
24) Delcourt, M. O.; Rossi, M. J. J. Phys. Chem. 1982, 86, 3233.
25) Kakitani, T.; Mataga, N. Chem. Phys. 1985, 93, 381; J. Phys. Chem.
When the concentration of 1 was increased up to 1.0 mM,
the yield of dimer 5 increased (32%) and those of 3 and 4
1
985, 89, 4752; Chem. Phys. Lett. 1986, 124, 437.
26) (a) Kakitani, T.; Yoshimori, A.; Mataga, N. J. Phys. Chem. 1992,
6, 5385. (b) Tachiya, M.; Murata, S. J. Phys. Chem. 1992, 96, 8441.
27) (a) Wayner, D. D. M.; McPhee, D. J.; Griller, D. J. Am. Chem.
(
(28) Gould, I. R.; Baretz, B. H.; Turro, N. J. J. Phys. Chem. 1987, 91,
925.
(29) (a) Robbins, W. K.; Eastman, R. H. J. Am. Chem. Soc. 1970, 92,
6076. (b) Engel, P. S. J. Am. Chem. Soc. 1970, 92, 6074.
(30) Das, P. K. Chem. ReV. 1993, 93, 119 and references cited therein.
9
(
Soc. 1988, 110, 132. (b) Sim, B. A.; Milne, P. H.; Griller, D.; Wayner, D.
D. M. J. Am. Chem. Soc. 1990, 112, 6635.

Redox Reactions of Benzyl-Type Radicals
J. Am. Chem. Soc., Vol. 118, No. 31, 1996 7257
Table 1. Redox Potentials of Ketones and Radicals
decreased (1.9% and 41%, respectively), as the steady state
concentrations of radical ions may be decreased. However, the
product ratios 3/4 were almost unchanged (i.e., 0.06 ( 0.02),
indicating the constant relative efficiency for the ET processes
of 2 (eqs 6a and 6b). The free energy changes (∆G) in the
CdO (V vs SCE)a
•
R
2
R (V vs SCE)
ox
red
ox
red
R
E
1/2
E
1/2
E
1/2
E
1/2
b
c
c
c
c
b
c
c
c
c
(
4
C
4
C
6
H
5
)
6
2
CH
∼+2.3
∼+2.4
+2.29
∼+2.2
+1.59
<-2.2
<-2.2
<-2.2
<-2.2
<-2.2
+0.35
+0.80
+0.73
+0.51
+0.26
-1.14
-1.40
-1.43
-1.62
-1.82
-ClC
CH
-MeC
H
4
CH
2
•
electron transfer of radical R can be calculated from eqs 7 and
6
H
5
2
for the oxidation of 2 by N•+ and for the reduction by DCB ,
•-
8
6 4 2
H CH
respectively; for the case of 2, the oxidation (-∆Gox ) 1.45
eV) is much more exothermic than the reduction (-∆Gred )
6 4 2
4-MeOC H CH
a
Obtained by rotating ring-disk electrode voltammetry (see Experi-
0
.46 eV). Thus, the observation of predominant reduction of 2
mental Section). b Wayner, D. D. M.; McPhee, D. J.; Griller, D. J. Am.
Chem. Soc. 1988, 110, 132. Sim, B. A.; Milne, P. H.; Griller, D.;
was quite striking.
c
Wayner, D. D. M. J. Am. Chem. Soc. 1990, 112, 6635.
ox
•
red
•+
∆
G ) E
(R ) - E
(D ) )
(iv) Benzyl-type cations34ab and anions34c have been known
ox
1/2
1/2
ox
•
ox
to react with MeOH quite rapidly, the rate constants being 6-8
E1/2 (R ) - E
(D) (7)
1/2
7
8
-1 -1
×
10 and ca. 2 × 10 M s , respectively. Thus, in the
ox
•-
red
•
presence of >5% MeOH, the two ions react with MeOH within
0 ns, much faster than any other reactions with transient
∆
Gred ) E1/2 (A ) - E
(R ) )
1/2
1
red
red
•
E1/2 (A) - E1/2 (R ) (8)
•+
species. If a part of N is decomposed by the reaction with
MeOH, a decrease of cation formation is expected at higher
MeOH concentration. This is not the case, because the product
ratios were practically not affected by the amount of methanol
in the range from 5% to 20%. In contrast, at <1% methanol,
the ratio of 3/4 was decreased down to <0.03, and irradiation
in the absence of MeOH afforded 4 and a small amount of
benzophenone, which might be formed by reaction of 6 with
trace water followed by oxidation. The product ratios at the
lower MeOH concentration were not reproducible, probably
because of the variable amount of contaminating water. Thus,
while the electron transfer may be independent of methanol, a
substantial amount of methanol is required for the efficient
trapping of cation 6.
When a substituted naphthalene such as 2-methylnaphthalene
ox
11d
(
(
MN, E1/2 ) +1.68 V vs SCE) or 2,6-dimethylnaphthalene
ox
11d
DMN, E1/2 ) +1.59 V vs SCE) was employed, the yields
of 3, 4, and 5 for 1 mM 1 were 13%, 52%, and 12% for MN
and 21%, 43%, and 20% for DMN, respectively. The product
ratios of 3/4, averaged for the reaction with 1.0 to 5.0 mM 1,
increased up to 0.41 ( 0.17 for MN and 0.51 ( 0.02 for DMN.
If the ratios were directly related to the ET rates, the more
efficient oxidation of 2 by naphthalene radical cations with
decreasing oxidation potentials may be indicative of the Marcus
inverted region.
The redox reaction of radical 2 was supported by the
following facts:
(v) Since the fluorescence of N is unaffected by the addition
of ketone 1 but efficiently quenched by DCB 1 at nearly
(
i) The irradiation of N and DCB in the absence of 1 in
1
0
-1 -1 35
MeOH-MeCN for 2 h afforded practically no products, the
recoveries of N and DCB being more than 98%. This may lead
to an assumption that the radical ions N•+ and DCB are long-
diffusion controlled rate, k ) 1.8 × 10
M
s , the excited
q
singlet state of N undergoes exclusively ET with DCB. Under
these conditions, DCB slightly absorbs light, but irradiation of
1 and DCB does not yield 3 or 4. Moreover, the redox potential
•-
32
lived and stable until they react by the back ET between them.
ox
red
36
By prolonged irradiation for 30 h, both N and DCB were slightly
consumed (ca. 5%), and a trace amount of 1-methoxynaphtha-
lene was detected by GC-MS as a product, which was not found
among byproducts in the co-photolysis of 1, N, and DCB.
of 1 (E1/2 ) ∼+2.3 V and E
< -2.2 V vs SCE) indicates
1/2
•+
•-
that the electron transfer of 1 with N or DCB is energetically
unfavorable. Thus, the direct electron transfer of ketone 1 is
not involved.
(
ii) The irradiation of 1 in the absence of either N or DCB
(vi) Other transient species involved here are the excited state
•
did not yield 3 and 4 at all and dimer 5 was obtained in more
than 90% yield. This indicates that the radical 2 is unreactive
toward N, DCB, and MeOH, consistent with the well-known
inertness of benzyl-type radicals toward C-H bonds.
iii) When the solution was irradiated in 5% CH3OD (99.5
atom % D), monodeuterated 4 (i.e., Ph2CHD) was obtained with
5.7% isotopic purity, indicating clearly that 4 was formed by
ketone and an acyl radical (Ph CH-CO ). Although the lifetime
2
of the singlet excited state of 1 is not known, dibenzyl ketone
37
has only a few nanoseconds fluorescence lifetime. The triplet
state instantaneously undergoes the C-C bond cleavage gen-
erating benzyl and phenacetyl radicals, and the latter radical
immediately affords another molecule of benzyl radical by
3
3
(
3
8
•
9
decarbonylation. The acyl radical (Ph CH-CO ) has been
2
2
8
the protonation of carbanion 7, not by the hydrogen abstraction
of 2.
known to have a lifetime as short as ∼8 ns. Thus, such
transients are quite short lived and their intermolecular reactions
can be eliminated.
(
31) (a) Since the extinction coefficients of 1 and N at 313 nm are 220
and 90, respectively, about 30% of light is absorbed by N. The quantum
yields for the formation of D and A , however, are governed by the
back electron transfer and we have no information about the true yields for
free ions for various D-A pairs in the tables. (b) Grellmann, K. L.; Watkins,
A. R.; Weller, A. J. Phys. Chem. 1972, 76, 469.
(vii) As the alkylation of aromatic nitriles is known in
•
+
•-
39
photoinduced ET reactions, a radical-radical ion combination
(34) (a) Sujdak, R. J.; Jones, R. L.; Dorfman, L. M. J. Am. Chem. Soc.
1976, 98, 4875. (b) McClelland, R. A.; Kanagasabapathy, V. M.; Banait,
N.; Steenken, S. J. Am. Chem. Soc. 1989, 111, 3966. (c) Bockrath, B.;
Dorfman, L. M. J. Am. Chem. Soc. 1975, 97, 3307.
(35) Tsuchida, A.; Tsuji, Y.; Ito, S.; Yamamoto, M.; Wada, Y. J. Phys.
Chem. 1989, 93, 1244.
(36) Redox potentials of benzyl ketones were determined by rotating
ring-disk electrode voltammetry as listed in Table 1.
(37) Lipson, M.; Noh, T.; Doubleday, C.; Zaleski, J. M.; Turro, N. J. J.
Phys. Chem. 1994, 98, 8844.
(38) (a) Lunazi, L.; Ingold, K. U.; Scaiano, J. C. J. Phys. Chem. 1983,
87, 529. (b) Turro, N. J.; Gould, I. R.; Baretz, B. H. J. Phys. Chem. 1983,
87, 531.
(
32) (a) The rate constant for the reaction of naphthalene radical cation
with methanol is not known, although that for the disappearance of the
4
5 -1
radical cation in water have been reported to be 1.7 × 10 and ∼10 s by
3
2b
32c
the radiolysis or photolysis, respectively, in the presence of persulfate
•
+
ion. In the present redox reaction, the reaction of N with methanol seems
not to be significant since the resulting yields of oxidized product were
rather constant when changing the amount of MeOH from 3 to 20%. (b)
Zevos, N.; Sehested, K. J. Phys. Chem. 1978, 82, 138. (c) Steenken, S.;
Warren, J.; Gilbert, B. C. J. Chem. Soc., Perkin Trans. 2 1990, 335.
33) Lusztyk, J.; Kanabus-Kaminska, J. M. In Handbook of Organic
Photohemistry; Scaiano, J. C., Ed.; CRC Press: Boca Raton, 1989; Vol. 2,
(
Chapter 8.
(39) Fagnoni, M.; Mella, M.; Albini, A Tetrahedron 1994, 50, 6401.

7258 J. Am. Chem. Soc., Vol. 118, No. 31, 1996
Ishiguro et al.
0
.10 ( 0.01, 0.10 ( 0.01, 0.08 ( 0.01, and 0.10 ( 0.01,
respectively. This is another piece of evidence against the
formation of alcohol adducts.
(x) Dimer formation of aromatic radical ions may affect the
ET rates, as in the reported case of ET between naphthalene
dimer radical cation and 9,10-dicyanoanthracene radical anion
•
-
(
DCA ) which is faster than that in the solvent-separated ion
•- 43
pair of monomer and DCA . The laser flash photolysis study
has shown that a small portion of N exists as the dimer radical
cation at ca. 10 mM naphthalene.
•+
3
5
In the present case,
however, the contribution of dimer formation cannot account
•
+
for the slower ET for N , since the extrapolated product ratios
are not affected by the naphthalene concentrations, i.e., 3/4 )
0.18 ( 0.01, 0.13 ( 0.01, and 0.12 ( 0.02 at 5, 10, and 20
mM naphthalene, respectively.
Thus, all the facts described above suggest that the efficient
electron transfer between radical 2 and radical ions is achieved
in the present co-irradiated system, the reduction of 2 by DCB
anion radical being more efficient than the oxidation by
naphthalene radical cations.
The Free Energy Change (∆G) Dependence on the Redox
Reaction of 2. In order to examine the free energy gap
dependence on the ET rates, the relative efficiencies were
determined for a variety of donors and acceptors as listed in
Chart 1. Since it has been suggested that the electron transfer
rates between aromatic radical ion pairs depend on ring-size,11
donors and acceptors are classified as one-, two-, and three-
ring systems.
Figure 1. Product ratios of Ph
2 3 2 2
CHOCH /Ph CH on the irradiation of
1
.5 mM 1, 10 mM DCB, and 10 mM naphthalenes: naphthalene
(circle), 2-methylnaphthalene (spuare), or 2,6-dimethylnaphthalene
(triangle), in 5% MeOH-MeCN.
may lead to a coupling reaction rather than the electron transfer.
In the present case, such a reaction seems not to be a major
pathway, since the total yield of 3, 4, and 5 was usually as high
as 70-80% and the recoveries of N and DCB were more than
9
5%. Possible coupling products are adducts of radicals and
sensitizers (e.g., 8 and 9), only trace amounts (<1%) of which
The relative efficiencies for the oxidation of 2 with donor
radical cation vs the reduction with DCB anion radical were
obtained from the extrapolated product ratios of 3/4, as
summarized in Table 2 (runs 1-19). Those for the reduction
with substituted dicyanobenzenes, 2,5-dimethyl-1,4-dicyanoben-
were detected among the byproducts by GC-MS analyses. The
reported DCB-sensitized fragmentation of 5 (cf. eq 3, R ) R′
3
b
)
Ph2CH) in MeOH-MeCN afforded 3 and 4 in a 1:1 ratio,
red
•
-
zene (DMDCB, E1/2 ) -1.75 V vs SCE) or dicyanodurene
indicating that the reaction between 2 and DCB yields 7
quantitatively. While the reaction of 2 with N has never been
red
•+
(TeMDCB, E1/2 ) -1.90 V vs SCE), were also determined
with 1,3,5-trimethoxybenzene (135TrMOB) or 1,2-dimethoxy-
benzene (12DMOB) as reference donors. The resulting relative
rates were independent of the reference (Table 2, runs 20-23)
and were smaller than that for DCB, i.e., 0.30 for DMDCB and
examined, similar electron transfer of Me- or MeO-substituted
40
naphthalene radical cations with 1,1-diphenylethyl or bis(4-
methoxyphenyl)methyl radicals41 has been suggested to proceed
efficiently. Thus, the coupling reaction of radicals with radical
ions is not involved in the present system.
0
.35 for TeMDCB.
As shown in Figure 2, plots of the relative rates vs -∆G
(
viii) The observation of less efficient oxidation of radical 2
•
+
were rather scattered but seemed to be curved downward on
going more exothermic. When the theoretical curve was
calculated from the recent formulation of electron transfer theory
(eq 9) with suggested values for a vibrational reorganization
by N may be due to the formation of a side product that is
lower in oxidation potential than naphthalene or 2. If such a
product is formed, a secondary electron transfer will occur
•
+
between N and the byproduct with formation of its radical
cation, which cannot oxidize 2. In fact, the product ratio of
∞
3
/4 changed during the progress of irradiation, probably due to
3
2
1/2
2
S w
k ) (4π /h λ k T) |V|
(e- S /w!) exp{-[(λ + ∆G +
et
s
b
∑
s
some effect of byproducts. Therfore, the product ratios at zero
irradiation time which could be determined by extrapolation of
the values at the initial stage of photoreaction should be more
reliable (Figure 1). The resulting ratios of 3/4 were 0.10 (
w)0
2
whV) /4λ k T]} (9)
s
b
S ) λ /hν
v
0
.01, 0.12 ( 0.01, and 0.29 ( 0.01 with N, NM, and DMN,
respectively, which again indicated that the oxidation of 2
became more efficient with decreasing exothermicity. Repeated
experiments showed that the selectivities were reproducible
within (30%.
energy (λv) of 0.25 eV and an averaged frequency (ν) of 1500
-1 9b,11d
cm ,
an appropriate value for the solvent reorganization
energy (λs) was evaluated to be ca. 0.65 eV. This value is
significantly smaller than those for the return ET in geminate
(
ix) In the presence of nuleophiles such as alcohols, aromatic
radical ion pairs from the quenching of excited sensitizers (e.g.,
radical cations may be in reversible equilibrium with the alcohol
11-15
1
.5-2.0 eV),
but is rather close to reorganization energies
adducts.4 Therefore, the effect of alcohols was examined. With
2
for self-exchanging reactions of organic radical ions (e.g., 0.3-
5% MeOH, EtOH, i-PrOH, and t-BuOH, the product ratios
44
0
0
.6 eV) and for ET rates within contact radical pairs (i.e.,
extrapolated to initial time were almost constant, i.e., 3/4 )
.5, 0.2,19b and 0.55 eV ).
19a
16
(40) Arnold, D. R.; Maroulis, A. J. J. Am. Chem. Soc. 1977, 99, 7355.
(41) Johnston, L. J.; Kanigan, T. J. Am. Chem. Soc. 1990, 112, 1271.
(42) (a) Sehested, K.; Corfitzen, H.; Christiansen, H. C.; Hart, E. J. J.
The ∆G dependence was confirmed by the following experi-
ments:
Phys. Chem. 1975, 79, 310. (b) Sehested, K.; Holcman, J.; Hart, E. J. J.
Phys. Chem. 1977, 81, 1363.
(43) Gould, I. R.; Farid, S. J. Am. Chem. Soc. 1993, 115, 4814.
(44) See ref 6c, p 79.

Redox Reactions of Benzyl-Type Radicals
J. Am. Chem. Soc., Vol. 118, No. 31, 1996 7259
Chart 1
(
i) When donors such as 1235TeMB, 14DMOB, and D were
radical cations, H-abstraction of radical 2 from donors, and direct
oxidation of 1 by excited DCB could be eliminated.
irradiated in the presence of DCB in MeOH-MeCN for 2 h,
practically no reaction took place and more than 98% of donors
were recovered. The irradiation of 1 and methylated donors
such as 135TrMB, 1245TeMB, and 14DMOB in the absence
of DCB did not afford 3 and 4, the dimer 5 being obtained in
more than 90% yield. Since substituted benzenes and diphenyls
(ii) In the presence of a second donor (D ), the oxidation
2
potential of which is lower than primary donor (D ), the second
1
•+
•+
electron transfer from D to D will occur to yield D2 . Since
2
1
the second charge-shift reaction is expected to be as fast as the
diffusion controlled rate when the difference between oxidation
11d,45
(
runs 1-10 and 13) have only weak absorption at >290 nm,
potentials of D and D is more than 0.4 V,
the concentra-
1
2
DCB may also absorb light and oxidize 1 in such cases.
However, the yield of dimer 5 was not affected by the addition
of DCB. Thus, side reactions such as decomposition of donor
tion of D can be kept low enough so that it does not affect the
2
primary ET between D and A. Therefore, the addition of a
1
relatively low concentration of D is expected to change the
2
product ratio for D1/A to that for D2/A. This was examined
for naphthalene (N), mesitylene (135TrMB), and 1,2,3,5-
tetramethylbenzene (1235TeMB) as D1 and 1,4-dimethoxyben-
zene (14DMOB) and 1,2,4-trimethoxybenzene (124TrMOB) as
D2. As summarized in Table 3, the product ratio gradually
shifted to that for D2/A by the addition of D2, and it was found
that ca. 0.5 mM of D2 was sufficient to change the redox couple.
(iii) It has been shown that the rates of return electron transfer
in solvent-separated ion pairs decrease by introducing bulky
4
6
substituents such as tert-butyl by a factor of 2-4. Therefore,
steric effect on the redox reaction was examined using 1,3,5-
tri-tert-butylbenzene (1,3,5-t-Bu3B). Since the oxidation po-
ox
46
tential of 1,3,5-t-Bu3B (E1/2 ) 2.01 V vs SCE) is intermediate
ox
11d
between 135TrMB (E1/2 ) 2.11 V vs SCE) and 124TrMB
ox
11d
(
E1/2 ) 1.92 V vs SCE), the expected relative rate may be
Figure 2. Plots of relative ET rates (krel) vs free energy change (-∆G)
between diphenylmethyl radical and aromatic radical cations (open
circle) or radical anions of 1,4-dicyanobenzenes (filled circle). The
numbers denote the run number in Table 2. The curve is calculated by
using the values in the text.
ca. 0.2. However, a considerably lower value of 0.011 ( 0.02
(
45) Lewis, F. D.; Bedell, A. M.; Dykstra, R. E.; Elbert, J. E.; Gould, I.
R.; Farid, S. J. Am. Chem. Soc. 1990, 112, 8055.
(46) Gould, I. R.; Farid, S. J. Phys. Chem. 1993, 97, 13067.

7260 J. Am. Chem. Soc., Vol. 118, No. 31, 1996
Ishiguro et al.
Table 2. Product Ratios on the Co-irradiation of 1 and Donor-Acceptor Systemsa
g
k
rel
run no.
Db
-∆Gox (eV)c
Ad
DCB
-∆Gred (eV)e
3/4f
ox
red
1
2
3
4
5
6
7
8
9
135TrMB
124TrMB
1235TeMB
1245TeMB
PMB
135TrMOB
12DMOB
14DMOB
124TrMOB
D
1.76
1.57
1.48
1.43
1.36
1.14
1.10
1.00
0.77
1.61
1.45
1.33
1.32
1.24
1.17
1.38
0.74
0.70
0.64
0.46
0.16 ( 0.01
0.24 ( 0.03
0.31 ( 0.03
0.028 ( 0.001
0.017 ( 0.001
0.12 ( 0.01
0.17 ( 0.01
0.58 ( 0.02
0.87 ( 0.01
0.22 ( 0.05
0.10 ( 0.01
0.12 ( 0.01
0.14 ( 0.01
0.29 ( 0.01
0.44 ( 0.01
0.22 ( 0.01
0.040 ( 0.004
0.80 ( 0.03
1.30 ( 0.12
0.16
0.24
0.31
0.028
0.017
0.12
0.17
0.58
0.87
0.22
0.10
0.12
0.14
0.29
0.44
0.22
0.040
0.80
1.30
(1.0)
1
1
1
1
1
1
1
1
1
1
0
1
2
3
4
5
6
7
8
9
N
MN
DMD
DMN
MON
P
A
MOA
DMOA
20
21
22
23
135TrMOB
12DMOB
135TrMOB
12DMOB
1.14
1.10
1.14
1.10
DMDCB
TeMDCB
0.61
0.76
0.37 ( 0.01
0.63 ( 0.03
0.33 ( 0.01
0.50 ( 0.01
0.12h
0.31
0.28
0.35
0.35
i
0.17
h
0.12
i
0.17
a
A solution of 1.5 mM 1, donor, and acceptor in 5% MeOH-MeCN was irradiated (>290 nm) for 30-120 min under argon at 20 ( 3 °C.
b
c
Donors, 10 mM except for P, A, MOA, and DMOA (0.5 mM); abbreviations are shown in Chart 1. The free energy change on the electron
transfer from 2 to donor radical cation as calculated by eq 7. d Substituted dicyanobenzenes (Chart 1), 10 mM. The free energy change on the
e
f
electron transfer to 2 from acceptor radical anion as calculated by eq 8. Relative ET rates determined by extrapolation of the product ratios of 3/4
to zero irradiation time. The relative rates for electron transfer of 2 as compared with that for the reduction by DCB ; ox, the oxidation by donor
radical cation; red, the reduction by acceptor radical anion. Excerpted from run 6. Excerpted from run 7.
g
•-
h
i
a
Table 3. Effect of Second Donor, D
2
(Table 1), care must be taken to avoid the intervention of direct
reaction of the starting ketones. Thus, the donors were selected
so that the electron transfer between the ketone and the donor
radical cation was at least 0.2 eV endothermic.
b
c
3/4d
D
1
[D
1
] (mM)
10
D
2
2
[D ] (mM)
N
14DMOB
0.10 ( 0.01
0.12 ( 0.01
0.66 ( 0.01
0.88 ( 0.01
0.58 ( 0.02
0.16 ( 0.01
0.11 ( 0.01
0.21 ( 0.01
0.44 ( 0.02
0.82 ( 0.01
0.87 ( 0.01
0.16 ( 0.01
0.63 ( 0.04
0.58 ( 0.02
0.31 ( 0.03
0.78 ( 0.01
0.87 ( 0.01
10
10
10
0.005
0.05
0.5
In a similar way as the case of 1, relative ET efficiencies of
•
benzyl radicals (ArCH2 ; Ar ) p-X-C6H4) were examined with
1
0
substituted dibenzyl ketones, 11, (ArCH2)2CdO, and donor/
acceptor pairs as listed in Chart 1. For the semiquantitative
purpose, the ratios of the corresponding benzyl methyl ethers
1
35TrMB
10
124TrMOB
10
10
10
10
0.001
0.01
0.1
(11) and toluenes (12) were determined at only one irradiation
0.5
0
time and listed in Table 4. The listed values reflect the relative
•
1
1
1
ET rates of benzyl radicals (ArCH2 ) since the formation of
1
35TrMB
10
14DMOB
adducts of benzyl radicals and sensitizers was not significant
and the resulting ratios of 12/13 did not differ largely with the
irradiation time.
1
0
0.5
0
1
235TeMB
10
124TrMOB
1
0
0.5
0
a
2 2
A solution of donors, 1.5 mM (Ph CH) CO, and 10 mM DCB in
5
% MeOH-MeCN, was irradiated (>290 nm) for 30-120 min under
b
argon at 20 ( 3 °C. Primary donor: N ) naphthalene; 135TrMB )
1
,3,5-trimethylbenzene; 1235TeMB ) 1,2,3,5-tetramethylbenzene.
Second donor: 14DMOB ) 1,4-dimethoxybenzene; 124TrMOB )
c
d
1
,2,4-trimethoxybenzene. Relative ET rates for the oxidation of 2
determined by extrapolation of the product ratios of 3/4 to zero
irradiation time.
The present method allows the direct comparison for oxida-
tions of a benzyl radical with various radical cations by
employing the same accetor and radical precursor, and the
comparison for the reduction with different anion radicals can
be made similarly. However, the relative effeciencies for
different benzyl radicals cannot be evaluated directly. Then,
sets of radical-radical cation combinations with almost the same
-∆G values were assumed to have identical ET rates and were
employed as the references. Taking into account the ring-size
effect, the rates of the following three sets were selected: (i)
p-chlorobenzyl vs diphenyl (run 7, -∆G ) 1.16 eV), p-
methylbenzyl vs 4,4′-dimethyldiphenyl (run 28, -∆G ) 1.16
eV), and p-methylbenzyl vs 2-methylnaphthalene (run 33, -∆G
was obtained, indicating that the efficiency of electron transfer
decreases by a factor of ca. 20. The larger steric effect on the
present ET than the reported one in solvent-separated ion pairs
may suggest that the ET occurs at a closer distance, i.e., in
contact ion pair.
Products from Substituted Benzyl Radicals. In order to
look into a wider range of ∆G, the examination was extended
•
to substituted benzyl radicals, p-X-C6H4CH2 (10, X ) H, Me,
MeO, and Cl), with various redox potentials depending on the
substituents (Table 1).2
7b
For the methyl- and methoxy-
substituted cases which have relatively low oxidation potentials

Redox Reactions of Benzyl-Type Radicals
J. Am. Chem. Soc., Vol. 118, No. 31, 1996 7261
Table 4. Products on the Co-irradiation of 11 and Donor-Acceptor Systemsa
selectivity (%)f
12 13 14
1.6
k
g
rel
run
no.
[D]
(mM)
-∆Gox
[A]
(mM)
-∆Gred
conv
(%)e
c
Ad
CN
(eV)c
X
Db
(eV)
12/13
ox
2.6
red
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
Cl
135TrMB
1235TeMB
1245TeMB
12DMOB
14DMOB
124TrMOB
D
10
10
10
10
10
10
50
50
0.5
0.5
0.5
50
50
50
50
1.31
1.03
0.98
0.65
0.54
0.32
1.16
0.87
0.93
0.29
0.18
1.16
0.87
1.16
0.87
15
0.58
9
51
41
3
3
7
3
3
12
3
4
4.1
22
80
39
81
12
<4
6
2.6
3.1
7.3
2.1
5.5
4.6
0.17
3.4
(1.0)
7.2
0.7
26
7.3
12
3.1
7.3
2.1
5.5
4.6
0.17
3.4
5.1
55
40
55
4.5
12
3.4
10
7.5
2.6
43
4.4
14
26
3.5
57
20
23
2.8
2.1
1.6
51
27
19
42
55
90
51
93
DMD
P
A
0.060 0.060
0
1
2
3
4
5
0.50
0.33
0.93
0.50
0.33i
0.17
MOA
D
CA
0.5
0.5
0.18
38
68
70
42
0.19
0.17
0.063
j
DMD
20
2.8
>50
3.4
i
D
DCA
-0.49
0.17
j
DMD
<0.3 80
3.4
<0.07
1
1
1
6
7
8
H
N
MN
DMN
10
10
10
1.07
0.95
0.86
DCB
CN
10
15
0.17
0.36
55
37
22
1.0
2.9
3.5
13
11
12
53
50
43
0.077 0.90
0.26
0.29
3.1
3.4
j
12
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Me
124TrMB
1235TeMB
1245TeMB
PMB
135TrMOB
12DMOB
14DMOB
124TrMOB
D
DMD
A
MOA
DMOA
N
MN
DMN
MON
N
MN
DMN
10
10
10
10
10
10
10
10
50
50
0.5
0.5
0.5
10
10
10
10
10
10
10
1.41
1.32
1.27
1.20
0.98
0.94
0.84
0.81
1.45
1.16
0.58
0.54
0.47
1.29
1.17
1.08
1.01
1.29
1.17
1.08
h
h
h
h
h
27
27
19
3
h
66
h
(5:10:85)
(55:31:14)
(46:15:39)
(42:13:45)
(68:32:h)
7.7
0.50
1.8
3.1
0.63
2.3
3.9
3.2
2.1
4.1
2.7
58
49
71
2.0
36
39
1
7.5
4.9
9.6
6.2
10
12
19
5.9
0.11
0.14
7.5
8
0.13
(12:88:h)
66
(33:67:h)
(27:73:h)
15
0.17i
1.3
2.9
5
0.044 0.056
0.49
0.37
0.73
1.0
0.63
0.47
0.13
0.17i
0.23
0.082
h
DCB
10
-0.02
40
32
28
29
39
33
45
11
14
16
9.9
69
44
46
49
18
17
36
14
12
21
0.17
1.3
0.47
<0.07
<0.04
0.10
DMDCB
CN
10
15
0.13
0.16
<3.0 44
<1.7 39
3.2
32
0.23
2.3
3
4
4
4
4
4
4
9
0
1
2
3
4
5
MeO 135TrMOB
12DMOB
14DMOB
124TrMOB
A
10
10
10
10
0.5
0.5
0.5
1.23
1.19
1.09
0.89
0.83
0.79
0.72
27
27
20
36
53
37
36
65
73
63
84
15
29
29
22
8
3
8
5
8
9
11
3.0
9.2
13
1.3
7.9
5.0
6.5
70
52
48
4.1k
5.6
0.44
13
5.7
0.21
0.56
0.60
0.10
0.25
0.27
MOA
DMOA
a
6 4 2 2
A solution of 3.0 mM (p-X-C H CH ) CO (11), donor, and acceptor in 5% MeOH-MeCN was irradiated (>290 nm) for 90-120 min under
c
b
argon at 20 ( 3 °C. Donors: abbreviations are shown in Chart 1. The free energy change on the electron transfer between benzyl radicals and
radical ions as calculated by eq 7 or 8. Acceptors: abbreviations are shown in Chart 1. Conversion of 11. The product selectivities based on
d
e
f
g
1
1 reacted; the values in parentheses are product ratios. The relative rates of electron transfer of benzyls with radical ions as compared with that
h
•
•-
for the reduction of p-Cl-C
6 4
H
CH
2
by DCB ; ox, the oxidation by donor radical cation; red, the reduction by acceptor radical anion. Not determined.
i
j
k
Excerpted from run 7. Excerpted from run 8. Excerpted from run 22.
)
1.17 eV), (ii) p-chlorobenzyl vs 4,4′-dimethyldiphenyl (run
, -∆G ) 0.87 eV) and benzyl vs 2,6-dimethylnaphthalene
calculated since the absolute rate constants could not be
determined, but the relative values could be estimated, i.e.,
1:0.38:0.23 for one-, two-, and three-ring systems, respectively.
The higher λs and V values for smaller ring systems reflect a
more condensed and hence more effective charge density. This
tendency is qualitatively consistent with the reported result for
8
(run 18, -∆G ) 0.86 eV), and (iii) p-methylbenzyl vs
pentamethylbenzene (run 22, -∆G ) 1.20 eV) and p-meth-
oxybenzyl vs 1,2-dimethoxybenzene (run 40, -∆G ) 1.19 eV).
The relative ET rates for benzyl radicals were calculated and
are listed in the last two columns in Table 4.
1
1b,e
return ET within geminate radical ion pairs.
Despite such an arbitrary assumption, the plots of relative
ET rates vs ∆G clearly showed the Marcus inverted region again
as shown in Figure 3. The redox reaction seems to be more
efficient with the smaller-ring system, showing different curves
in terms of one-, two-, and three-ring donors/acceptors. The
fitting of the data with the theoretical curve (eq 9), with accepted
“Rehm-Weller” Behavior vs “Marcus” Behavior. All the
results indicated that efficient redox reactions of radicals with
aromatic radical ions were achieved by the co-irradiation of
donor, acceptor, and ketone. The product ratios of ROMe/RH
indicate the relative efficiency of the oxidation vs the reduction
with a steady-state concentration of the two radical ions. The
product ratios of ROMe/RH for various donor-acceptor pairs
generally decreased with increasing oxidation potentials. Since
-
1 9b,11d
values of λv ) 0.25 eV and ν ) 1500 cm ,
evaluation of the solvent reorganization energies (λs) of 0.41,
.37, and 0.24 eV for one-, two-, and three-ring systems,
respectively. The electronic coupling elements (V) were not
led to an
•
+
•-
0
the two radical ions, D and A , are produced in a 1:1 ratio,
the product ratios reflect the relative rates for the redox reaction,

7262 J. Am. Chem. Soc., Vol. 118, No. 31, 1996
Ishiguro et al.
rate at the distance at which the reorganization energy matches
the free energy gap. Such cases have been explored experi-
4
9-51
mentally.
In fact, it is common for the electron transfer between radical
species by diffusional collisions that the maximal rate is not
5
2-54
observed even in extremely exothermic cases.
However,
in a particular case, the inverted region has been reported for
the back ET of N-alkyl pyridinyl radicals with a radical cation
5
5
of an iridium complex.
Thus, if correct, the present ET
reaction seems to be the second example of Marcus-type
intermolecular electron transfer. Specific for these two case
may be the type of ET, i.e., charge shift between neutral and
charged molecules. However, it has been revealed theoreti-
5
6
11d,45,57,58
cally and experimentally
charge separation, shift, and recombination (a dielectric satura-
that the difference between
25
tion effect of solvents) has only a minor effect on the ET rates.
It has been indicated that the rate constant for the heteroge-
neous electron transfer of radical 2 on electrode (ca. 0.1 cm
-
1
s
for both oxidation and reduction) is similar to the values
5
9
for common hydrocarbon/ion radical couples. The kinetics
of ET between benzyl radicals and aromatic radical anions have
shown that the electron transfer proceeds with smaller reorga-
nization energy than the redox of alkyl radical/anion60 because
of a smaller change in the molecular structures by the redox
61
process. For the cation/radical pair, the vibrational reorganiza-
tion energy, λv, was estimated from molecular orbital calcu-
11d,62
lations
(see Experimental Section) to be ca. 0.10 eV, which
was still smaller than that calculated for radical/anion pair (0.13
eV). Thus, the vibrational reorganization energy is not the
critical factor for the present case.
The other parameters governing the ET rates are the solvent
reorganization energy (λs) and the electronic coupling matrix
elements (V). The estimated λs values for benzyl radicals are
Figure 3. Plots of relative ET rates (krel) vs free energy change (-∆G)
between benzyl radicals and radical cations of one-ring (open circle),
two-ring (open spuare), and three-ring donors (open triangle) or radical
anions of one-ring (filled circle), two-ring (filled aquare), and three-
ring acceptors (filled triangle). The numbers in A denote the run number
in Table 4. The straight, dashed, and dotted curves in B are calculated
for one-, two-, and three-ring systems, respectively, by using the values
in the text.
0
.41, 0.37, and 0.24 eV for one-, two-, and three-ring systems,
respectively (see Figure 3). The λs and relative V values are in
the order of one- > two- > three-ring radical ions, which is
interpretable on the basis of the relative charge densities of the
three systems. That is, the charge density of one-ring aromatic
radical ions is higher than that of two-ring ions and hence the
ET rates are expected to be higher than two- or three-ring
systems. The higher charge density results also in the higher
solvent reorganization energy as observed. Thus, the relative
revealing the Marcus inverted region. Although actual examples
7-19
for the inverted region have been demonstrated,
the present
result seems to be a very specific case as discussed below.
The absence of the inverted region in the intermolecular
electron transfers between excited state molecules and quench-
ers, so-called the Rehm-Weller relationship,20 is well-known.
Among various explanations, the formation of electronically
(49) Kikuchi, K.; Niwa, T.; Takahashi, Y.; Ikeda, H.; Miyashi, T. J. Phys.
Chem. 1993, 97, 5070.
(50) Kikuchi, K.; Takahashi, Y.; Hoshi, M.; Niwa, T.; Katagiri, T.;
2
1
excited state products has gained the widest acceptance, and
proved to be true for some cases.47 Since radical ions generally
have much lower excited states than the neutral molecules, the
energy gap may become smaller by subtraction of the excitation
energy, so that the process cannot be sufficiently exergonic to
reach the inverted region. If the availability of the excited state
accounts for the crucial difference, the electron transfer between
two open-shell partners may show the inverted region so long
as the products are non-radical species with much higher
Miyashi, T. J. Phys. Chem. 1991, 95, 2378.
(51) Miyasaka, H.; Morita, K.; Kamada, K.; Nagata, T.; Kiri, M.; Mataga,
N. Bull. Chem. Soc. Jpn. 1991, 64, 3229.
(
52) Gschwind, R.; Haselbach, E. HelV. Chim. Acta 1979, 62, 941.
(53) Schomberg, H.; Staerk, H.; Weller, A. Chem. Phys. Lett. 1973, 21,
433; 1973, 22, 2735.
(54) Delcourt, M. O.; Rossi, M. J. J. Phys. Chem. 1982, 86, 3233.
(
55) McCleckey, T. M.; Winkler, J. R.; Gray, H. B. J. Am. Chem. Soc.
992, 114, 6935.
56) (a) Tachiya, M. J. Phys. Chem. 1989, 93, 7050. (b) Kakitani, T. In
1
(
Dynamics and Mechanism of Photoinduced Electron Transfer and Related
Phenomena; Mataga, N., Okada, T., Masuhara, H., Eds.; Elsevier: Am-
sterdam, 1992; p 71.
2
1
excitation energies. In fact, all the examples for the inverted
region have been demonstrated for electron transfer between
(
57) Legros, B.; Vandereecken, P.; Soumillion, J. J. Phys. Chem. 1991,
95, 4752.
(58) Eriksen, J.; Jørgensen, K. A.; Linderberg, J.; Lund, H. J. Am. Chem.
Soc. 1984, 106, 5083.
59) Nagaoka, T.; Griller, D.; Wayner, D. D. M. J. Phys. Chem. 1991,
95, 6264.
(60) (a) Fuhlendorff, R.; Occhialini, D.; Pedersen, S. U. Acta Chem.
7
-19
radical species affording closed-shell species.
Alternatively, Kakitani et al.26a and Tachiya et al.26b have
recently explained the dependence of donor-acceptor distance
(
on the solvent reorganization energy, which increases by
increasing the distance as indicated by Marcus.48 That is, for
Scand. 1989, 43, 803. (b) Occhialini, D.; Pedersen, S. U.; Lund, H. Acta
Chem. Scand. 1990, 44, 715. (c) Occhialini, D.; Kristensen, J. S.; Daasbjerg,
K.; Lund, H. Acta Chem. Scand. 1992, 46, 474.
a wide range of ∆G, the electron transfer between freely
diffusing substrates can take place at the diffusion-controlled
(61) Mikkelsen, K. V.; Pedersen, S. U.; Lund, H.; Swanstrom, P. J. Phys.
(
47) Kikuchi, K.; Katagiri, T.; Niwa, T.; Takahashi, Y.; Suzuki, T.; Ikeda,
H.; Miyashi, T. Chem. Phys. Lett. 1992, 193, 155.
48) Marcus, R. A.; Siders, P. J. Phys. Chem. 1982, 86, 622.
Chem. 1991, 95, 8892.
(62) Nelsen, S. F.; Blackstock, S. C.; Kim, Y. J. Am. Chem. Soc. 1987,
109, 677.
(

Redox Reactions of Benzyl-Type Radicals
J. Am. Chem. Soc., Vol. 118, No. 31, 1996 7263
λs and V values for these systems may be comprehensive on
the basis of their different charge densities.
14A gas chromatographs, using a 2.5 mm × 1 m column of Carbowax
3
00 M (2%) on Chromosorb WAW. A Shimadzu Chromatopac C-R3A
integrator was used for quantitative analyses. GC-MS analyses were
carried out with a JEOL D300 or a Shimadzu QP-5000 mass
spectrometer using a 2.5 mm × 1 m column of Carbowax 300 M (2%)
on Chromosorb WAW or a 0.2 mm × 25 m capillary column of Silicon
OV-1 (J&W Scientific, DB-1). Absorption spectra were recorded on
a Shimadzu UV-265 ultraviolet spectrometer.
Materials. Acetonitrile was distilled from phosphorus pentaoxide.
Methyl alcohol (Dotite Spectrozol), methyl alcohol-d (Aldrich, 99.5
atom % D), 2,6-dimethylnaphthalene (DMN; Aldrich), 4,4′-dimethyl-
diphenyl (DMD; Aldrich), and 1,2,4-trimethoxybenzene (124TrMOB;
Aldrich) were employed as received. Naphthalene (N), 2-methoxy-
naphthalene (MON), diphenyl (D), phenanthrene (P), anthracene (A),
We could not answer, however, why the inverted region was
observed so clearly for the intermolecular case of benzyl
radicals. Theoretical studies usually treat donor/acceptor mol-
ecules or solvents as “spheres”, which are far from the real
pictures, and always neglected are the characteristics of reacting
molecules such as the orientation of interacting orbitals or the
polarizabilities of molecules. A change of spin states might be
of some importance in the ET reaction of radicals as exemplified
as the magnetic field effect63 and spin statistical effect.64 Some
of these factors may be important in the ET reaction of benzyl
radicals, resulting in the slowdown of ET rates and hence in
the appearance of the inverted region. At present we cannot
clarify the origin of the inverted region. To clarify these points,
the direct determination of ET rates of benzyl radicals is desired
but it is not so easy to determine the absolute rates because of
1
-cyanonaphthalene (CN), and 9-cyanoanthracene (CA) were purified
by recrystalization. Mesitylene (135TrMB), 1,2,4-trimethylbenzene
124TrMB), 1,2,3,5-tetramethylbenzene (1235TeMB), pentamethyl-
(
benzene (PMB), 1,2-dimethoxybenzene (12DMOB), 1,4-dimethoxy-
benzene (14DMOB), 1,3,5-trimethoxybenzene (135TrMOB), and 2-
methylnaphhtalene (MN) were received from Tokyo Kasei and distilled.
•
+
•-
•
the involvement of so many intermediates, e.g., D , A , R ,
+
-
68
69
R , and R . Electrochemical methods are also not directly
accessible to very fast ET processes because of the instrumental
limitation.
sym-Tetraphenylacetone (1), substituted dibenzyl ketones (11),
9-methoxyanthracene (MOA),70 9,10-dimethoxyanthracene (DMOA),
,5-dimethyl-1,4-dicyanobenzene (DMDCB), dicyanodurene (Te-
MDCB), and 9,10-dicyanoanthracene (DCA) were prepared as
reported.
70
7
1
2
71
72
In the present paper, a fundamental assumption is that ion
•+
•-
pairs of D and A are formed and exist in ca. 1:1 ratio during
the irradiation. A case where this situation is not satisfied seems
Typical Procedure for Photolyses. Irradiations were done with a
3
00-W medium-pressure mercury lamp in a 25-mL Pyrex test tube,
•
+
also to be assumed. That is, if a portion of D is consumed
by some side reaction,65 an equal amount of A is to be
accumulated in the solution. In fact, when cation radicals are
i.e., >290 nm. A 10-mL MeOH-MeCN (1:9) solution of 0.3 mM
sym-tetraphenylacetone, 10 mM naphthalene, and 10 mM p-dicy-
anobezene in a test tube with a septum rubber cap was purged with
argon for 15 min and was irradiated at 20 ( 3 °C. The products were
identified by GC-MS, and the conversion of the ketone and the yields
of products were determined at appropriate intervals of time by GLC.
Electrochemical Measurements. The redox potentials of ketones
(1, 11) and acceptors (DMDCB and TeMDCB) were obtained in MeCN
•-
-
•
consumed, stable radical anions such as DCA have been
6
6,67
•+
shown to be accumulated.
In such cases, the ratios of D
•
-
and A become different from the initial 1:1 ratio and the
resulting product ratios ought to reflect the concentration ratios
•
+
•-
of D /A . However, control experiments seem to rule out
this possibilities. That is, redox product ratios were not altered
by changing light intensity and sensitizer concentrations. On
the other hand, the product ratio changed significantly by
substituents (see Table 4), reflecting the relative ET rates of
substituted benzyl radicals.
4 4
containing 0.1 M n-Bu NBF by the rotating ring-disk electrode
voltammetry73 using a Nikko Keisoku RRDE-1 electrode and a DPGS-6
potentiostat at 1000 rpm. The redox potentials of other acceptors (DCB,
2a
11d
74
CN, CA, and DCA) and methyl- or methoxy-substituted donors
were taken from literature. The potentials are summarized in Table 1
and Chart 1.
Conclusion. Efficient redox reaction of benzyl-type radicals
was achieved by the irradiation of aromatic donor/acceptor/
ketone in MeOH-MeCN. In this system, one-electron oxida-
tion and reduction of photogenerated benzyl radicals occurred
by the reaction with radical ions of donor and acceptor to yield
the corresponding benzyl cation and anion, which were trapped
by MeOH affording ROMe and RH, respectively. Relative
product ratios of products which indicate the relative ET rates
were determined for a variety of donor-acceptor pairs. The
apparent relative ET ratios were determined from the product
ratios and correlated with ∆G values of the redox reaction. The
relative ET rates became maximal at the energy gap of ca. -0.5
eV, representing a novel example of the Marcus inverted region.
Calculations of Theoretical Electron Transfer Rates. Calculation
of theoretical electron transfer rates according to eq 9 and the fitting
to the experimental data were carried out on an Apple Macintosh
computer using the program IGOR PRO (Wavemetrics). When the
v
values for a vibrational reorganization energy (λ ) of 0.25 eV and an
-
1
9b,11d
averaged frequency (ν) of 1500 cm were adapted,
reorganization energy (λ ) and an electronic coupling matrix element
V) were the variables to be optimized. Iterated calculation of the
a solvent
s
(
Franck-Condon weighted density was carried out for w ) 0 to 10,
since further iterations over 11 did not affect the rates.
Estimation of λ
v
by Molecular Orbital Calculations. The
vibrational reorganization energy, λ , was estimated by molecular orbital
v
6
2
calculations using a method similar to those described by Nelsen and
Gould et al.1 The semiempirical PM3 calculations were carried out
on Apple Macintosh and NEC PC-9801 computers using the MOPAC
package (versions 575 and 6 ). Calculations on benzyl cation or anion
1d
30
76
Experimental Section
(
69) (a) Kimura, Y.; Tomiya, Y.; Nakanishi, S.; Otsuji, Y. Chem. Lett.
General. 1H NMR spectra were recorded with Hitachi R24B (60
MHz) and Varian GEMINI-200 (200 MHz) NMR spectrometers. GLC
analyses were performed with Yanagimoto G180 and Shimadzu GC-
1
1
979, 321. (b) Turro, N. J.; Weed, G. C. J. Am. Chem. Soc. 1983, 105,
861.
(70) Meek, J. S.; Monroe, P. A.; Bouboulis, C. J. J. Org. Chem. 1963,
8, 2572.
2
(
(
(
63) Steiner, U. E.; Ulrich, T. Chem. ReV. 1988, 89, 51.
(71) Suzuki, H.; Hanafusa, T. Syntheses 1974, 1, 53.
(72) Dufraisse, C. Bull. Soc Chim. Fr. 1947, 302.
(73) Bard, A. J.; Faulkner, L. R. In Electrochemical Methods; Wiley:
64) Saltiel, J.; Atwater, B. W. AdV. Photochem. 1988, 14, 1.
65) Bard, A. J.; Ledwith, A.; Shine, H. J. AdV. Phys. Org. Chem. 1976,
1
2, 155.
New York, 1980.
(
66) (a) Schaap, A. P.; Zaklika, K. A.; Kaskar, B.; Fung, L. W.-M. J.
(74) (a) Zweig, A.; Hodgson, W. G.; Jura, W. H. J. Am. Chem. Soc.
1964, 86, 4124. (b) Zweig, A.; Hodgson, W. G.; Jura, W. H. J. Org. Chem.
1967, 32, 1322. (c) Hamada, T.; Nishida, A.; Yonemitsu, O. J. Am. Chem.
Soc. 1986, 108, 140.
Am. Chem. Soc. 1980, 102, 389. (b) Spada, L. T.; Foote, C. S. J. Am. Chem.
Soc. 1980, 102, 391.
(67) (a) Ci, X.; Kellett, M. A.; Whitten, D. G. J. Am. Chem. Soc. 1991,
1
13, 3893. (b) Freccero, M.; Mella, M.; Albini, A. Tetrahedron 1994, 50,
(75) MOPAC version 5.00 (QCPE No.445): Stewart, J. J. P. QCPE Bull.
1989, 9, 10. Hirano, T. JCPE Newsletter 1989, 1 (2), 36. Revised as version
5.01 for OS/2 Personal Computers (NEC PC-9801): Toyoda, J. JCPE
Newsletter 1990, 2 (1), 56.
2
115.
(68) Dean, D. O.; Dikinson, W. B.; Quayle, O. R.; Lester, C. T. J. Am.
Chem. Soc. 1950, 72, 1740.

7264 J. Am. Chem. Soc., Vol. 118, No. 31, 1996
Ishiguro et al.
and on radical were performed with the RHF and UHF methods,
respectively. The geometries were optimized within a C2V symmetry,
i.e., in planar geometry, with the keyword PRECISE to increase the
criteria for the terminating the optimization by a factor of 100. The
UHF heats of formation of benzyl radical in equilibrium geometries of
benzyl cation, radical, and anion were 40.62, 39.63, and 40.92 kcal/
mol, respectively. The RHF heats of formation of benzyl cation in
geometries optimized for benzyl cation and radical were 227.42 and
at radical and ion geometries. The resulting reorganization energies
were (40.62 - 39.63) + (228.63 - 227.42) ) 2.20 kcal/mol and (40.92
- 39.63) + (20.15 - 18.53) ) 2.91 kcal/mol (ca. 0.10 and 0.13 eV)
for benzyl cation/radical and radical/anion pairs, respectively.
Acknowledgment. This work was supported in part by a
Grant-in-Aid for Scientific Research from the Ministry of
Education, Science and Culture of Japan.
2
28.63 kcal/mol, and those of anion in radical and anion geometries
were 20.15 and 18.53 kcal/mol, respectively. The vibrational reorga-
nization energy, λ , can be estimated from the sum of the difference in
energy before and after electron transfer, i.e., the difference in energy
Supporting Information Available: Tables of the product
yields on the co-irradiation of 1 and donor-acceptor systems
v
(
3 pages). See any current masthead page for ordering and
(
76) MOPAC version 6.00 (QCPE No.445): Stewart, J. J. P. QCPE Bull.
990, 10, 86. Version 6.02: Hirano, T. JCPE Newsletter 1991, 2 (4), 39.
Macintosh version: Toyota, J., 1991.
Internet access instructions.
1
JA953689U