6816 J . Org. Chem., Vol. 61, No. 20, 1996
Cevasco and Thea
F igu r e 4. Plot of log ka vs log Kredox. Kredox values were
calculated (see text) from the following E0 values (in mV)
reported in increasing order of reactivity: 748, 736, 711, 681,
656, 597.
F igu r e 3. Hypothetical effect of SIS on the energy profiles
(lines shown are merely notional).
by reinforcing such a hydrogen bond. It is therefore
possible for this ester that, different from the 3,5-
disubstituted ones considered before, the expected in-
crease in reactivity is matched by a nearly equal corre-
sponding increase in the pKa, due partly to SIS and partly
to the intramolecular hydrogen bond, so that no overall
deviation from the regression line is observed.
As an alternative to these speculations, one might
postulate that reactivity-acidity correlations could not
be quite satisfactory in this case simply because the
chosen standard chemical equilibrium (substrate dis-
sociation) is not well suited to the purpose. However, in
a somewhat similar system, i.e., the dissociative hydroly-
sis of 2′,4′-dinitrophenyl esters of substituted 4-hydroxy-
benzoic acids, such a Brønsted-type relationship was
suggested to hold.8 In light of the present findings we
now believe that resort to additional suitably substituted
4-hydroxybenzoates might reconcile these seemingly
conflicting behaviors.
reaction) will be significantly higher, other things being
equal, than that in the absence of SIS. In contrast, it is
reasonable to suppose that the differences in the energies
of the E1 transition states, as due to SIS, will be
negligible, since the negative charge is almost completely
transferred onto the leaving group, as suggested by the
very large value of the âLG (-1.51) found for the alkaline
hydrolysis of meta- and para-substituted aryl 3,5-di-
methyl-4-hydroxybenzenesulfonates.1c Therefore, SIS
will give rise to a decrease in the free energy of activation
for the E1 reaction, matched by an increase in reactivity.
However, since SIS is also responsible for an increase in
the pKa values of the 2,6-disubstituted phenols, positive
deviation of the three esters from the regression line is
probably due to the fact that phenol ionization is affected
by SIS less than the E1 step. As a matter of fact, SIS
will act also on hindered neutral phenols, raising their
free energy although to a minor extent with respect to
the corresponding phenoxide ions since solvation of the
negative charge of the anion is more important energeti-
cally than solvation of the neutral species. In other words
it is likely that the effect of SIS on the free energy content
is only moderate for the undissociated substrates, rather
strong for their conjugate bases, and negligible for the
E1 transition states. This is tentatively illustrated in
Figure 3, where hypothetical energy profiles in the
presence (dashed line) or in the absence (solid line) of
SIS are roughly represented.
As Figure 2 shows, two other 3,5-disubstituted esters
(namely, the 3,5-difluoro- and the 3-methyl-5-nitro-
substituted ones) nicely fit the relationship for the
monosubstituted derivatives. In the former case, SIS is
expected not to be very effective owing to the small size
of the fluorine atoms; as for the latter compound, which
is a 2-methyl-6-nitrophenol further substituted in posi-
tion 4, it is well-known6 that the acid strength of
o-nitrophenols is diminished by the presence of a strong
intramolecular hydrogen bond between the two groups.
Furthermore, it has been suggested7 that the presence
of an additional alkyl substituent adjacent to the OH
group further decreases the acidity of the phenol group
Therefore, we suggest that a better standard for the
reaction could be the redox equilibrium shown in eq 5.
(5)
The relevant equilibrium constants Kredox were obtained
from the relationship ∆G° ) -RT ln Kredox ) -nFE0
making use of the E0 values (in ethanol) reported in the
literature for substituted quinones.9
As shown in Figure 4, the log ka vs log Kredox plot is
remarkably linear, thus suggesting that the transition
state of the rate-determining step has an electronic
structure resembling that of p-quinones10 and, by infer-
ence, that it is similar to the intermediary sulfoquinone.
These results not only confirm that the reaction
pathway related to ka is dissociative in nature but also
successfully introduces, to our knowledge for the first
time, such a relationship between reactivity and Kredox
.
(8) Thea, S.; Cevasco, G.; Guanti, G.; Kashefi-Naini, N.; Williams,
A. J . Org. Chem. 1985, 50, 1867
(9) Conant, J . B.; Fieser, L. F. J . Am. Chem. Soc. 1923, 45, 2194.
Scrocco, E.; Marmani, G. Ann. Chim. 1951, 41, 716.
(10) Thea, S.; Williams, A. Chem. Soc. Rev. 1986, 16, 125.
(6) Arnold, R. T.; Sprung, J . J . Am. Chem. Soc. 1939, 61, 2475.
(7) Dearden, J . C.; Forbes, W. F. Can. J . Chem. 1960, 38, 1852.