A. Ouchi et al. / Tetrahedron Letters 46 (2005) 6713–6716
6715
Scheme 2 indicates the presence of a large steric interac-
tion between the ortho-methyl-substituted phenyl group
and C60 moiety in the case of fast reactions (1b,d), par-
ticularly when the phenyl group rotates around the N–
Ph bond, whereas smaller interactions are expected for
the slow reactions (1a,c). The Ar group without the
ortho-methyl-substituent may give the best angle with
respect to the N-lone pair that enables facile electron
transfer of the lone pair electron to the C60 moiety,
which is most probably due to the rupture of efficient
delocalization between the lone pair electrons and the
p-electrons of the phenyl ring. The generation of the po-
sitive charge on the nitrogen atom might be the reason
for the decrease in the rate of the rearrangement in the
case of charge-separated triplet states.
as triplet sensitizers, which is consistent with the fact
that rearrangement 1 ! 2 proceeded by a unimolecular
process. In contrast, it is reported that the triplet energy
1
1
of 2d is higher than that of 1d so that 2d can act as a
triplet sensitizer showing a large k value and nonexpo-
s
nential feature in the rearrangement. Although 2b has a
normal triplet character, the sensitization effect is not
apparently observed in the reaction; this is due to the
fact that the k value was 1/4.5 but the k value was
s
d
11.7-fold larger than that of 1d. This inefficient triplet
sensitization of 2b can be rationalized by 2b having sim-
ilar triplet energy to that of 1b.
Although the rearrangement rate is mostly controlled by
the steric effect, the electronic factor of the substituents
also affected the rate of the reaction. In the case of the
unimolecular process, kd values decreased with the
introduction of the electron-donating group to the phe-
nyl group; the rate decreased to 1/3.2 with para-methyl
(1c vs 1a) and to 1/11.7 with ortho-methyl (1b vs 1d) sub-
stitutions. The result on the para-methyl substitution
can be explained by the facilitation of charge transfer
from the electron-donating N-lone pair to the electron-
accepting C60 moiety due to the increase in the electron
density of the N-lone pair. In contrast, a 4.5-fold in-
Besides rearrangement rates, Figure 1 also demonstrates
a difference in the reaction mechanisms. The consump-
tion of 1a–c and the formation of 2a–c showed single
exponential-like decay and rise, which is an indication
of a unimolecular process. However, the reaction of 1d
shows nonexponential curves for the consumption of
1
d and the formation of 2d. This is explained by the trip-
3
let sensitization of the reaction by product 2d. To deter-
mine the contribution of the unimolecular and sensitized
processes, the rate constants of both processes were ob-
tained similar to the reported procedure with consider-
ation of the difference in e values of 1a–d and 2a–d.
crease in k value was observed with ortho-methyl substi-
s
3
b
tution (1b vs 1d), which is rationalized by the facilitation
of the triplet energy transfer from 2 to 1 due to the
increase in triplet energy of 2d.
4
Thus, the rate constants (normalized with e) obtained
for the unimolecular (k ) and the sensitized (k ) pro-
d
s
ꢀ
7
ꢀ1
ꢀ3
ꢀ1
cesses are 1.5 · 10 cm M s
and 7.3 · 10 cm s
In summary, the reaction rate of the photochemical
rearrangement 1 ! 2 differed ca. 3000-fold depending
on the position and number of the methyl substituents
on the N-phenyl group. The required time for the com-
pletion of the reaction decreased in the order
1d < 1b < 1a < 1c. The difference was mainly due to a
switching of the excited states between normal (fast
reactions) and charge-separated (slow reactions) triplet
states, which was caused by steric interactions between
the N-phenyl group and the C60 moiety. The large differ-
ence in the reaction rates can be considered as an ampli-
fication of small structural differences into large
chemical reactivities by the switching between the two
different excited states, which may be extended to the
development of molecular switches.
ꢀ
4
ꢀ1
ꢀ1
for 1a, 1.4 · 10 cm M s and 5.8 cm s for 1b, 4.7 ·
ꢀ
8
ꢀ1
ꢀ17
ꢀ1
ꢀ1
1
1
0
cm M s
and 6.3 · 10
cm s
for 1c, and
ꢀ
5
ꢀ1
4
.2 · 10 cm M s and 2.6 · 10 cm s for 1d.
The result of the sensitized reactions can be explained by
the relative triplet energies of 1a–d and 2a–d (Scheme 3).
The transient spectra of 2a–d are similar to those of 1a–d
so that those of 2a and 2c are assigned to the charge-sep-
arated triplet state and 2b and 2d to the normal triplet
state. As the triplet energy of charge-separated 2 is esti-
3
b
mated to be lower than that of 1, 2a and 2c cannot act
H
H(Me)
(Me)H
N
N
H
C H
H
H
Acknowledgements
We thank Mr. Shin-ichirou Kawabata and Mr. Akio
Tominaga, Shimadzu Co., for MALDI-TOFMS mea-
surements, Dr. Y. Xiao for the preparation of 1b–d
and 2b–d, and Dr. T. Oishi for the purification and spec-
troscopic measurements of 1 and 2.
slow reaction
fast reaction
Scheme 2. Steric interaction between N-aryl and C60 moieties.
normal triplet
Supplementary data
charge-separated
triplet
Experimental details for the synthesis, analytical data,
spectral data, steady-state photolyses, and flash photo-
lyses of 1 and 2. UV spectra of 1a–d and 2a–d, and deter-
mination of rate constants k and k . Supplementary
2d
1d
1b
2b
1a,c 2a,c
Scheme 3. Schematic diagram for the relative triplet energy of 1a–d
and 2a–d.
d
s