Angewandte
Chemie
zene. For reproducible results, the addition of 1% water to
rigorously dried acetonitrile was necessary. The addition of
less water resulted in lower rate accelerations, whereas no
further improvement was observed when water was added in
larger amounts. Water itself (in the absence of acids) did not
affect the rate of reaction.
Some structure–activity data can be extracted from the
data in Table 1, which lists the relative rate constants along-
side some common thermodynamic properties relevant to
phenolic-hydrogen-atom-transfer activity. At first glance, it is
clear that the extent of acid catalysis is not simply related to
the properties which normally influence the reactivity of
Figure 1. Left: oxygen consumption at 303 K during the azobisisobutyr-
onitrile-initiated autooxidation of styrene (4.3m) in “wet” CH3CN
(dotted line), and when the reaction is inhibited by pentamethylchro-
manol (2; 1.25 mm) in the absence (dashed line) or presence (solid
line) of AcOH (88 mm); right: dependence of k1 on the concentration
of the acid.
ꢀ
phenols with peroxyl radicals, that is, in particular, the O H
BDE and/or the oxidation potential (Eox) of the phenol.
Instead, the dependence of the rate constant on the concen-
tration of AcOH roughly parallels the hydrogen-bond acidity
2
Whereas the autooxidation of styrene in neat CH3CN is only
retarded by 2, the addition of AcOH (88 mm) results in a fully
inhibited autoxidation up to a reaction time of approximately
2000 s. This effect corresponds to a more than 16-fold increase
in the k1 value, from 6.8 ꢀ 105 to 1.1 ꢀ 107 mꢀ1 sꢀ1. This result is
particularly striking when you consider that the rate constant
determined for the reaction of 1 with peroxyl radicals by flash
photolysis in neat AcOH (k1 = 8.8 ꢀ 105 mꢀ1 sꢀ1) is only roughly
twice as large as that measured in CH3CN (k1 = 3.8 ꢀ
105 mꢀ1 sꢀ1).[10]
The rate constants determined for 2–6 as a function of
added AcOH are summarized in Table 1 (see the Supporting
Information for all data). The substantial rate accelerations
observed were proportional to the concentration of the acid
and not to the strength of the acid: AcOH produced the
largest effect for the series of acids surveyed (which also
included HCl, p-toluenesulfonic acid, and trichloroacetic
acid; see the Supporting Information for further details). No
acceleration was observed in the apolar solvent chloroben-
of the phenols, as quantified by the Abraham aH parame-
ter.[26] This result suggests that the mechanism of the reaction
of peroxyl radicals with phenols is different in the presence of
acids from the mechanism in the absence of acids, as the
ꢀ
acidity of the phenolic O H group generally hampers the
reaction (see above). To investigate this hypothesis further,
we examined the effect of added AcOH on the autooxidation
of styrene (again, in CH3CN containing 1% H2O) in the
presence of the pyrimidinol 7, a highly acidic phenol-like
ꢀ
antioxidant with a relatively high oxidation potential and O
H BDE,[24] as the inhibitor in place of phenols 2–6. Owing to
the extent of H bonding to the solvent, the presence of 7 led to
almost no inhibition of autooxidation under our experimental
conditions in the absence of an acid; however, the addition of
AcOH led to a dramatic increase in the k1 value in a dose-
dependent fashion. It was possible to increase the k1 value by
almost three orders of magnitude in this way (see the
Supporting Information for further details).
To obtain direct kinetic evidence for the acid-accelerated
reactions of peroxyl radicals with phenols, we turned to EPR
spectroscopy. Although EPR is unsuitable for monitoring
these reactions at room temperature, Fukuzumi et al. showed
that the kinetics of reactions of cumylperoxyl radicals with
substituted dimethylaniline derivatives and phenols can be
studied in propionitrile at 193 K.[27,11] Through a slightly
modified procedure, we were able to measure the k1 values
directly for the reactions of cumylperoxyl radicals with
phenols 2, 4, and 6 by monitoring the signal decay of the
cumylperoxyl radical formed upon photolysis of a mixture of
cumene (2m) and di-tert-butylperoxide (2m) in oxygenated
propionitrile in the presence of the phenols between 193–
213 K. In the absence of phenols, the acid did not affect the
lifetime of the cumylperoxyl radicals, which decayed with
clean bimolecular kinetics with an apparent self-termination
rate coefficient, 2kt, of 57 ꢂ 5mꢀ1 sꢀ1 at 193 K. However, when
one of the phenols 2, 4, or 6 was present in excess, the decay
was first-order, and the addition of an acid (HCl (100 mm),
AcOH (150–300 mm), CCl3CO2H (150 mm)) accelerated the
rate of decay in all cases (see the example in Figure 2).
Interestingly, the relative rates were smaller at 193 K than
those observed in the experiments on autooxidation kinetics
at 303 K, and increased when the temperature was increased
(see the Supporting Information). This observation suggests
Table 1: Rate constants for the reactions of phenols 2–6 and pyrimidinol
7 with peroxyl radicals, as determined at 303 K from inhibited
autooxidation reactions of styrene in CH3CN containing 1% H2O, and
the relative rate constants as a function of added AcOH.[a] Some relevant
properties of 2–7 are included for comparison.
k1acid=k1
k1 [mꢀ1 sꢀ1
]
BDE[c]
Epa
aH
[d]
2[e]
½bꢃ
½acidꢃ
[kcalmolꢀ1
]
[V vs. NHE]
2
3
4
5
6
7
(6.8ꢂ0.6)ꢀ105
(2.2ꢂ0.2)ꢀ104
(5.0ꢂ1.0)ꢀ103
(8.0ꢂ0.5)ꢀ104
(2.0ꢂ0.4)ꢀ104
(7.9ꢂ5.2)ꢀ102
170
3.9
110
81
1200
740
77.1[21]
77.2[21]
81.7[21]
79.7[h]
+1.13[19]
+1.31[20]
+1.50[g]
+1.26[20]
+1.45[23]
+1.58[23]
0.37
0.18[f]
0.57
0.55[f]
0.69
78.2[22]
81.4[24]
0.64[f]
[a] Errors correspond to ꢂ2 standard deviations. [b] Concentration of the
ꢀ
acid in m. [c] Experimental O H bond-dissociation energies in benzene
at 298 K downscaled by 1.1 kcalmolꢀ1 on the basis of the revised value
for phenol.[25] [d] Experimental Epa (anodic peak potential) in CH3CN at
298 K. NHE=normal hydrogen electrode. [e] aH2 =Abraham hydrogen-
bond-donor parameter.[26] [f] Estimated from kinetic measurements in
different solvents and Equation (2). [g] The value from Ref. [27] was
adjusted since the values reported therein are larger by approximately
0.2V than those from Refs. [23] and [19] for common compounds.
[h] Value for 2,5-di-tert-amylhydroquinone.[22]
Angew. Chem. Int. Ed. 2009, 48, 8348 –8351
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim