ReactiVity of Substituted Phenols Toward Alkyl Radicals
J. Am. Chem. Soc., Vol. 121, No. 3, 1999 513
and, by definition,
Scheme 2
kSH[ArOH] ) k0H[ArOH]
total
free
the measured rate constant can be expressed (eq 10) as a function
of KS and of the concentration of the HBA solvent. This equation
implies that KS can be calculated from kH0 and kSH.
radicals is more exothermic by ca. 10 kcal/mol than the reaction
with peroxyl radicals.33 The higher reactivity of peroxyl radicals
with phenols can be nicely rationalized on the basis of the
magnitude of the triplet repulsion terms in the transition state
for the H-transfer which will be lower for transfer between two
oxygen atoms rather than between oxygen and carbon.34
A special comment is deserved for 2,3,5,6-tetramethyl-4-
methoxyphenol (3b) which, despite having a structure very
similar to that of 2,3,6-trimethyl-4-methoxyphenol (3a), is
characterized by a kH value that is lower by a factor of 2.5.
These anomalous data parallel the BDE’s for these two
compounds and can be attributed to the fact that in 3b the
methoxy group is approximately perpendicular to the plane of
the aromatic ring for steric reasons and hence its oxygen atom’s
2p lone pair cannot conjugate with the π system.4,5,29 In this
geometry both the stabilization of the phenoxyl radical and the
destabilization of the phenol by the para MeO substituent are
substantially lost with a consequent increase of the activation
energy for H-transfer with respect to 3a where the OMe group
is coplanar with the aromatic ring.
As far as the effect of solvents on the rate constants for
hydrogen abstraction from phenols is concerned, Table 4 reports
the experimental values for the reaction of R-tocopherol with
alkyl radicals obtained in the present study together with those
reported for the reaction of the same substrate with tert-butoxyl
radicals. It can be seen that direction and magnitude of the
solvent effect is the same for the two reactions despite the fact
that the absolute kH values differ by 4 orders of magnitude. In
fact, a plot (not shown) of log kH(RO•) versus log kH(R•) is linear
and has the expected11 slope of 1. This provides additional
evidence that kinetic solvent effects for hydrogen atom abstrac-
tion from phenols are totally independent of the nature of the
attacking radical and are due solely to hydrogen bond formation
between solvent and the phenolic substrate.
kH0
kSH )
(10)
1 + KS[S]
The value assumed for k0H was that measured in pure
isooctane (1.15 × 106 M-1 s-1). The kSH values, plotted against
the molar concentrations of the HBA, were fitted to eq 10 to
obtain KS (see Figure 5). The resulting room temperature values
of KS for the formation of the hydrogen-bonded complex
between R-TOH and the HBA were 3.7 M-1 for tert-butyl
alcohol and 5.2 M-1 for γ-valerolactone (see Table 4).
For γ-valerolactone the equilibrium constant, KS, was deter-
mined by FT-IR by fitting the experimental concentrations of
free R-TOH to eq 11. The resulting value of the equilibrium
constant, KS ) 4.7 M-1, is in excellent agreement with that
obtained from the kinetic data (5.2 M-1).
[ArOH]tot
1 + KS[S]
[ArOH]free
)
(11)
Since the kinetic and the IR methods give comparable results,
we conclude that measurements of the rate constants for
hydrogen atom abstraction using radical clocks represents a new
and reliable technique for the determination of equilibrium
constants for hydrogen bonding by phenols and, presumably,
by other hydrogen bond donating substrates. This technique is
especially useful with those solvents, such as alcohols and
amines, for which the IR method cannot be used because of
the strong absorption in the 3300-3600 cm-1 spectral region.
The kinetic data obtained in toluene and γ-valerolactone for
2,4,6-trimethylphenol (1a) and 2,6-di-tert-butylphenol (2c)
(Table 4) indicate that the kinetic solvent effect for 1a is large
and comparable to that for R-TOH whereas it is negligible for
2c. This implies that solvation of the hydroxylic hydrogen atom
is about the same in the two ortho dimethyl substituted phenols
but is unimportant in the di-tert-butyl substituted phenol because
of the large steric hindrance due to these bulky groups.35 IR
spectroscopic data support this view; in fact, the vibrational
spectra of 2c show in the region of the free O-H absorption a
sharp peak at 3640 cm-1 having the same intensity in pure
isooctane and pure γ-valerolactone, indicating that no hydrogen
bonding occurs in either solvent. Analogous IR experiments
carried out with 1a in γ-valerolactone give KS ) 7.80 M-1, a
value only slightly higher than that of R-TOH, due to the higher
acidity of 1a.
The dependence of the measured values of kH on solvent
composition in mixtures of isooctane and tert-butyl alcohol or
γ-valerolactone (Figure 5) can be analyzed in terms of the simple
model proposed by Ingold and co-workers. This model is based
on the following assumptions: (i) solvent, S, and substrate,
ArOH, can give rise only to 1:1 hydrogen-bonded complexes,
ArOH‚‚‚S, and the equilibrium constant for complexation, KS,
is independent of the nature of the surrounding medium; (ii)
the attacking radicals react with the free phenol with a rate
constant k0H, but not with the complexed phenol (see Scheme
2) since, for steric reasons, they cannot approach the solvent-
bonded hydroxylic hydrogen atom. The rate constant, k0H, for
reaction with free phenol is the value in a hypothetical solvent
where no substrate-solvent interaction occurred.
Experimental Section
Materials. Solvents were of the highest purity grade commercially
available and were used as received. Phenol, 4-chlorophenol, 4-bro-
mophenol, 4-fluorophenol, 4-cyanophenol, 4-trifluoromethylphenol,
4-methoxyphenol, 4-methylphenol, and 4-tert-butylphenol were pur-
chased from Aldrich and were used as received. 2R,4′R,8′R (d)-R-
tocopherol (Aldrich) was purified by column chromatography on silica
gel according to a previously described method.11 All other phenols
were available from previous studies29 and were purified by crystal-
The equilibrium constant is given by:
[ArOH‚‚‚S]
[ArOH]free[S]
KS )
(9)
(33) Kerr, J. A. In Handbook of Chemistry and Physics, 74th ed.; CRC
Press: Boca Raton, 1993, p 9-123.
(34) Zavitsas, A. A. J. Am. Chem. Soc. 1972, 94, 2779-2789. Zavitsas,
A. J. Am. Chem. Soc. 1991, 113, 4755.
(35) The failure of 2,6-di-tert-butylphenols to form hydrogen bonds was
originally recognized in the early sixties. See for instance: Ingold, K. U.
Can. J. Chem. 1960, 38, 1092. Ingold, K. U.; Taylor, D. R. Can. J. Chem.
1961, 39, 471. Ingold, K. U. Can. J. Chem. 1962, 40, 111.