Full Paper
log ðkS=mÀ1 sÀ1Þ ¼ log ðk0=mÀ1 sÀ1ÞÀ8:3a2Hb2H
ð13Þ
On the basis of Equation (13), our rate constants in acidic
water should be close to the values recorded in acetonitrile,
which has an almost identical HBA ability (i.e., bH2 =0.38 for
water[45] versus b2H =0.39 for MeCN[38]). However, inspection of
the data in Table 3 shows that they are generally lower. Clearly
the relevant concentration of THF (bH2 =0.51[45]) in the reaction
mixture (25% v/v; 3.1m) contributes to the HBA ability of the
medium. To confirm this point, we studied the reaction of
phenol 1 in a 3:1 PhCl/THF solvent mixture and obtained kinh
=
2.8105 mÀ1 À1, that is, significantly smaller than the value in
s
PhCl and MeCN and similar to the value recorded in water/
THF. Taken together, these data suggest that, at least under
acidic conditions, in which the SPLET mechanism is sup-
pressed, the reaction in water is subjected to an identical sol-
vent effect as recorded in organic solution, hence it conceiva-
bly occurs by the same mechanism, one in which the formal
transfer of hydrogen is rate-determining.
Figure 6. UV/Vis spectra of Trolox (2) in phosphate buffer at different pH
(2.1–12) and in aqueous NaOH at pH 13 and 14.
Indeed, as predicted by Equation (13), a semi-logarithmic
plot of the ratio of the rate constants measured for each
phenol in PhCl and in water versus aH2 gave a perfect straight
line (r2 =0.995, see the Supporting Information). From this cor-
relation it is possible to estimate the HBA ability of our reac-
tion medium (3:1 water/THF containing 0.1m phosphate
buffer) by means of Equation (14), which is derived from Equa-
tion (13).[46]
that, in the case of autoxidations in water solution, although
the redox potential of the phenol decreases upon increasing
the pH due to progressive dissociation, which increases its re-
activity, the peroxyl radical becomes progressively less oxidis-
ing, thereby partly compensating the effect. Indeed, the redox
C
potential of CH3OO decreases by about 0.6 V upon passing
from pH 7 to pH 12,[23] which is a result of the deprotonation
of the hydroperoxide product. The pKa values of several organ-
ic hydroperoxides in water fall in the range 11.5–12.5[37] and
we found that tert-butyl hydroperoxide is partly dissociated in
phosphate buffer at pH 12 (see the Supporting Information).
Indeed, when we turned our investigation to the more acidic
4-methoxyphenol (5, pKa =10.1[8]), the overall rate enhance-
ment recorded upon passing from pH 2.1 to pH 12 was signifi-
cantly larger (50-fold). UV/Vis spectroscopy confirmed that the
dissociated fraction at pH 7.4 was <1%, which accounts for
the limited increase in kinh, whereas it was nearly completely
dissociated at pH 12 (see the Supporting Information).
log ðkS1=kS2Þ ¼ 8:3aH2 ðbH2 Àb2H
Þ
S2
S1
ð14Þ
The resulting bH2 value is 0.49,[47] sensibly higher than the
value for neat water, but slightly lower than that for THF. For
comparison, the bH2 values obtained by using Equation (14)
(using the data of phenol 1) for unbuffered water/THF (3:1)
and PhCl/THF (3:1) mixtures are, respectively, 0.44 and 0.43.
Therefore, the presence of phosphate buffer makes a minor
contribution to the HBA ability of the medium.
Compared with known kinetics in organic solution, such as
chlorobenzene, the rate constants measured in water at acidic
to neutral pH are generally smaller, the difference being more
marked for phenols with greater hydrogen-bond-donating
H(D) kinetic isotope effect
To gain a deeper insight into the mechanisms underlying the
reactions with peroxyl radicals in water, we measured the H(D)
kinetic isotope effect (KIE) by performing matching autoxida-
tion studies in H2O and D2O as solvent, exploiting the dynamic
OH!OD exchange in acidic H(D) atom donors.[48,49]
(HBD) ability. Indeed, the kinetic solvent effect (KSE) kPhCl
/
kH2O(2.1) is, respectively, 183, 20 and 3 for phenols 6, 1 and 4,
characterised by Abraham’s aH2 values of 0.73,[41] 0.37[42] and
0.18,[43] respectively. This KSE is typical for the EPT (or HAT) re-
The results collected in Table 4 show that both PMHC (1)
and Trolox (2) give the H(D) KIE as kH/kD ꢀ2 at pH 2.1, at which
they exist in the neutral form. Although this value is lower
than typical values in organic solvents (kH/kD was reported to
be 4.0 and 5.1 for a-tocopherol and PMHC at 308C in sty-
rene[49]), it is still consistent with the proton (or hydrogen
atom) being transferred in the rate-determining step. At
pH 7.4, the H(D) KIE for Trolox decreases (kH/kD =1.2), which
suggests a change in mechanism, such as a stepwise PT–ET, in
which the ET is rate-determining. However, it should be noted
that the value measured at pH 7.4 is not significantly different
from that recorded at pH 2.1 when the experimental errors in
C
C
action XÀH+Y !X +YÀH, in which X is an electronegative
C
heteroatom (e.g., O, N) and Y is any radical species, and arises
from hydrogen bonding of XÀH to the solvent, which hampers
its reactivity.[44] This KSE implies that the transfer of the hydro-
gen atom (or proton) is rate-determining and it is described
quantitatively by Ingold’s equation (Eq. 13),[42] in which kS and
k0 are the rate constants, respectively, in the solvent of interest
and in a non-HBA solvent (e.g., CCl4), and aH2 and b2H are Abra-
ham’s solvatochromic parameters (range 0–1) describing the
HBD ability of the reactant XÀH and the hydrogen-bond-ac-
cepting (HBA) ability of the solvent, respectively.[45]
Chem. Eur. J. 2016, 22, 7924 – 7934
7930
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim