Ketoprofen
FULL PAPER
bonyl group. The activation energy barriers for proton trans-
fer from the carboxylate group to the carbonyl group in the
assistance of perchloric acid are À0.21 and À1.32 kcalmolÀ1,
respectively. Houk and co-workers[48] pointed out that a neg-
ative activation-energy barrier generally requires not only a
highly reversible initial step, but also that any subsequent
steps that lead to the product must have low kinetic barriers,
and this explanation is in good agreement with the results of
our calculations. For many chemical reactions, DFT methods
produce reliable activation barriers, but there are some sys-
tems for which almost all of the DFT methods generate neg-
ative reaction barriers. These reactions usually have low ac-
tivation reaction barriers.[48] Jursic and co-workers[49] report-
ed that they used DFT to perform computational studies for
the hydrogen abstraction from hydrochloric acid and found
the majority of the DFT methods to generate negative acti-
vation-energy barriers, which is consistent with the theoreti-
cal study of OH addition reaction to toluene with negative
activation-energy barriers.[50] Therefore, the negative activa-
tion-energy barriers for the ESIPT and decarboxylation of
3[KP] indicate that a very low activation reaction barrier
needs to be overcome when mediated by perchloric acid
(HClO4) and sulfuric acid (H2SO4). Furthermore, the very
low energy for final product complex (PC) relative to reac-
tant complex (RC) and transition state (TS) suggests the de-
carboxylation reaction of 3[KP] is an exothermic reaction.
Thus ESIPT is found to be energetically viable, which may
in turn be spontaneously companied by almost barrier-free
decarboxylation of 3[KP], and these predicted results are
consistent with the observation of ns-TR3 experiments.
tial step experiences a 3.95 kcalmolÀ1 activation-energy bar-
rier, and then the subsequent reaction overcomes a 6.26 kcal
molÀ1 activation-energy barrier (see Figure 11). The relative-
ly higher activation-energy barriers induced by hydrochloric
acid reasonably explain the experimental observations that
the decarboxylation process of 3[KP] occurs in about 1%
HClO4 or 1% H2SO4 ([H+]=0.1m) acetonitrile solutions,
whereas we cannot discern the 3BCH species in 1% HCl
([H+]=0.1m) acetonitrile solution. Figure S10 in the Sup-
porting Information displays the optimized structures of
RCn, TSn, and PCn catalyzed by HCl. The distance between
the carbonyl and the proton of hydrochloric acid is 1.32 ꢁ,
and the dihedral angle between C1, C2, C3, and C4 is 61.238
in TS1, relative to a corresponding distance of 2.02 ꢁ and di-
hedral angle of 64.138 in RC1 in the initial step. The main
changes are seen in the distance between the carboxylate
moiety and the chloric anion, which varies from 2.26 ꢁ in
RC2 to 1.60 ꢁ in TS2. The trend of the proton transfer leads
À
to an elongation of the bond length between C3 C4 from
1.55 ꢁ in TS2 to 1.75 ꢁ in RC2. Another obvious change is
that the dihedral angle between C1, C2, C3, and C4 is 48.978
in RC2, whereas this value jumps 75.808 in TS2. The larger
structural changes in the second step are accompanied by
the higher activation energy for the decarboxylation to pro-
ceed.
On the basis of the DFT investigations on the acid-cata-
lyzed decarboxylation process of 3[KP], although water-
mediated ESIPT may play a significant role in the water-
rich solution, the acid-mediated ESIPT and decarboxylation
reaction cannot be ruled out on account of the smaller rela-
tive energy barrier induced by the acid. Particularly in sub-
stantially lower water solution (0.86 mL 70% perchloric
acid or 0.54 mL 98% sulfuric acid was added into 100 mL
neat acetonitrile), acid-mediated ESIPT is likely to be pre-
dominant rather than the water-mediated ESIPT. Once the
proton of the carboxylate group is shuttled to the carbonyl
For the ESIPT and decarboxylation reaction catalyzed by
HSO4À1, the activation energy barriers are 2.51 kcalmolÀ1 in
the initial step and 12.95 kcalmolÀ1 in the second step (as
shown in Figure 11). Clearly, the second step is the rate-de-
termining reaction. Examination of Figure S9 in the Sup-
porting Information shows that the distance between the
À1
3
carbonyl group and the proton of HSO4 changes from
group, the decaboxylation reaction of [KP] could occur im-
2.08 ꢁ in RC1 to 1.40 ꢁ in TS1 and the dihedral angles be-
tween C1, C2, C3, and C4 are 62.578 in RC1 and 60.238 in
TS1 in the initial step. The slight structural changes between
RC1 and TS1 account for the relatively low activation-energy
barrier (2.51 kcalmolÀ1) in the proton-transfer process from
mediately within the time resolution of ns-TR3 experiments.
Scheme 2 as well as Schemes S1, S2, and S3 in the Support-
ing Information are the proposed schematic diagram of the
proposed mechanism with activation-energy profiles for
perchloric acid, sulfuric acid, and hydrochloric acid cata-
À1
3
HSO4 to the carbonyl group. However, the distance be-
lyzed decarboxylation reaction of [KP] in acidic solutions.
tween the carboxylate group and perchloric acid varies from
1.70 ꢁ in RC2 to 1.09 ꢁ in TS2 and the dihedral angle be-
tween C1, C2, C3, and C4 is 68.778 in TS2 and 53.588 in RC2
Conclusion
À
in the second step. The bond length between C3 C4 also in-
creases from 1.54 ꢁ in RC2 to 1.59 ꢁ in TS2. Therefore, the
significant variations of the dihedral angle and distance in
the structure RC2 and TS2 lead to a relatively higher activa-
tion-energy barrier (12.95 kcalmolÀ1). The above results
Scheme 3 summarizes the proposed mechanism for the de-
carboxylation reaction with the mediation of water mole-
cules or acid molecules. The decarboxylation reaction of
3[KP] will take place in water-rich solutions, which coincides
with the results of DFT calculations that more water mole-
cules involved in the reaction will decrease the activation-
energy barriers. The results of the DFT calculations demon-
strate that water molecules are able to assist excited-state
intramolecular proton transfer (ESIPT) from the carboxyl-
À1
demonstrate that although both H2SO4 and HSO4 can cat-
3
alyze the decarboxylation reaction of [KP], the pathway in-
duced by H2SO4 should be predominant.
As for HCl, although the structure of HCl is smaller than
HClO4 and H2SO4, it still can modulate the proton transfer
from the carboxylate moiety to the carbonyl group. The ini-
À
ate group to the carbonyl group, followed by C C bond
Chem. Eur. J. 2011, 17, 10935 – 10950
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10947