Discussion
The results obtained in the absence of base clearly point out
to KPÏs behavior as a conventional substituted benzophe-
none; i.e., intersystem crossing yields are high and lead to a
triplet state with virtually identical spectroscopic properties to
those of benzophenone itself.12,13 This triplet can be readily
quenched by energy acceptors, such as 1-methylnaphthalene,
1,3-cyclohexadiene and sorbic acid. Further, while the KP
triplet is moderately long lived in waterÈacetonitrile mixtures;
in methanol it decays by hydrogen abstraction to yield the
corresponding ketyl radical.
In contrast with the behavior outlined above, experiments
in the presence of base in either methanol or water-rich ace-
tonitrile (with water content [5% v/v) lead to the character-
istic carbanion absorbance in the 600 nm region. This
transient signals shift with the polarity of the solvent from 580
nm in water to 620 nm in 5% waterÈacetonitrile. This suggests
that the ground state carbanion is considerably more stabil-
ized than its excited state in hydrogen bonding media.
The question arises concerning the origin of the 600 nm
transient. SpeciÐcally, is its precursor the excited singlet or
triplet state of KP? The former would be an unusual behavior
for a benzophenone chromophore. Benzophenones rarely
show any excited singlet chemistry; normally singlet processes
are dominated by intersystem crossing to the triplet state,
although some processes, such as delayed Ñuorescence have
been characterized.14 Whatever the precursor state it must be
very short lived, given that the D600 nm signal grows-in
within the duration (6 ns) of the short laser pulse. Fig. 7 shows
Fig. 7 Transient absorption changes observed in 2.5 ] 10~4 M KP
in waterÈacetonitrile solution, x
\ 0.13, 7.0 ] 10~3 M TMAOH,
water
Laser 308 nm. N bubbled. (a) In the absence of 1,3-cyclohexadiene.
2
(b) 1.04 ] 10~2
M
1,3-cyclohexadiene used as quencher. Time
windows: (…) 12 ns; (K) 45 ns; (]) 153 ns; (|) 390 ns after the laser
pulse.
that the 600 nm and a long lived triplet (j
530 nm) can be
max
detected simultaneously. The long lived triplet KP may ini-
tially be the undissociated carboxylic acid form of KP, which
we already know (from the base-free experiments) does not
yield any signiÐcant amounts of carbanion. However, acidÈ
base equilibria are usually quite fast, and in the presence of 8
mM base, one would expect the deprotonation to be estab-
lished in the excited triplet state of KP within 10 ns if the
process was near di†usion-controlled. Even if this was two
orders of magnitude slower, it should take less than 1 ls to
establish acidÈbase equilibration; this would be expected to
shorten the triplet lifetime. Clearly, added base does not lead
to accelerated decay of the triplet signal (see Fig. 8); in fact
there is no indication that this equilibrium plays any signiÐ-
cant role in the formation of the carbanion, or in controlling
the KP triplet lifetime. Thus, we conclude that when the
triplet KP is produced it does not lead to the formation of the
600 nm carbanion, even when the conditions are controlled
(with base) in such a way as to promote the formation of the
carboxylate form of the KP triplet.
If the triplet state of KP, in either its acid or dissociated
carboxylate form, cannot yield the carbanion, then one must
conclude that the process of decarboxylation must be singlet-
mediated. While this may be counter-intuitive in the context
of benzophenone photochemistry, it does have the appeal of
being able to produce a ground state carbanion in a spin-
allowed process. A few examples of triplet mediated carbanion
formation have been reported in the literature, but they gener-
ally involve adiabatic processes leading to the formation of
triplet carbanions,15,16 thus also meeting spin conservation
requirements. This does not appear to be the case for KP,
given that neither the lifetime nor the yield of the 600 nm
species are inÑuenced by the addition of triplet quenchers.
In view of the analysis presented above, we conclude that
the mechanism of Scheme 2 accommodates all the experimen-
tal data available. In particular, we note that in this mecha-
nism the triplet state (in either phototropic form) and the
carbanion are not connected.
troscopically similar to that obtained in aqueous bu†ered
solutions, while the longer-lived component resembles that
observed in the absence of base in acetonitrile-rich solutions
(vide supra). Addition of 10 mM 1,3-cyclohexadiene (an effi-
cient triplet quencher) completely quenched the slow decaying
component (see Fig. 7b) with a bimolecular rate constant of
9.9 ] 109 M~1 s~1. We also determined the rate constant for
quenching by biphenyl, which led to k \ 1.1 ] 109 M~1 s~1.
q
Neither the rate of decay nor the signal intensity of the fast
620 nm component were a†ected by these typical triplet quen-
chers.
The Arrhenius parameters from the decay of the 620 nm
transient were also determined for the 5% v/v waterÈ
acetonitrile mixture used above under basic conditions. The
activation energy was 9.6 kJ mol~1 and the pre-exponential
factor 2.3 ] 109 s~1. The corresponding Ðt has also been
included in Fig. 2. We note that in these experiments the base
concentration was reduced to 2.5 mM to avoid its precipi-
tation at the lower temperatures.
Product analysis following laser irradiation of KP in 5%
waterÈacetonitrile in the presence of base (8 mM) showed 3-
ethylbenzophenone as the only product.1 In order to accumu-
late sufficient amounts of products, ten separate 2 mL samples
were exposed to 200 laser shots and then combined.
Finally, experiments were carried in fresh waterÈacetone
mixtures as solvent in the presence of base irradiating with a
355 nm laser. Once more, the absorption of the transient pro-
duced after irradiation of KP resembles those observed in
base-rich solutions in water, methanol and waterÈacetonitrile
mixtures. As already observed in waterÈacetonitrile, the tran-
sient decays faster as the water content decreases showing a
lifetime of 180 and 20 ns for x
0.5 and 0.18 in acetone,
water
respectively. A red shift of the maximum on the visible part of
the spectrum is also evident, thus qualitatively supporting the
results found in waterÈacetonitrile mixtures. The product
studies in this system of waterÈacetone as solvent reveal once
more that only 3-ethylbenzophenone is formed upon irradia-
tion.
Another important piece of evidence consistent with this
mechanism, is that the pre-exponential factors obtained for
3536
Phys. Chem. Chem. Phys., 1999, 1, 3533È3537