Modeling the Co-Antioxidant Behavior of Monofunctional Phenols. Applications to Some Relevant Compounds

Article abstract of DOI:10.1021/jo0351825

A study on the regeneration of α-tocopherol (vitamin E) by phenolic co-antioxidants in homogeneous hydrocarbon solution is reported. The behavior of some relevant phenols such as BHA, BHT, and trans-resveratrol appears to be nicely predicted by a model based on the knowledge of kinetic and thermochemical data concerning the various reactants. Despite its good reputation as an antioxidant, trans-resveratrol was found only moderately effective (k_inh = 2.0 × 10⁵ M^-1 s ^-1 in chlorobenzene at 303 K) and unable to recycle vitamin E.

Full text of DOI:10.1021/jo0351825

Mod elin g th e Co-An tioxid a n t Beh a vior of Mon ofu n ction a l P h en ols.
Ap p lica tion s to Som e Releva n t Com p ou n d s
Riccardo Amorati, Fiammetta Ferroni, Gian Franco Pedulli,* and Luca Valgimigli*
Dipartimento di Chimica Organica “A. Mangini”, Universita` di Bologna, Via San Donato 15,
I-40127 Bologna, Italy
pedulli@alma.unibo.it
Received August 12, 2003
A study on the regeneration of R-tocopherol (vitamin E) by phenolic co-antioxidants in homogeneous
hydrocarbon solution is reported. The behavior of some relevant phenols such as BHA, BHT, and
trans-resveratrol appears to be nicely predicted by a model based on the knowledge of kinetic and
thermochemical data concerning the various reactants. Despite its good reputation as an antioxidant,
trans-resveratrol was found only moderately effective (kinh ) 2.0 × 105 M-1 s-1 in chlorobenzene at
303 K) and unable to recycle vitamin E.
Resveratrol is a naturally occurring phenol, particu-
larly abundant in grapes1 and red wine.2 Because red
wine consumers in France have a lower than expected
incidence of cardiovascular diseases,3 the “French Para-
dox”, a considerable number of investigations have been
devoted to understanding the cardiovascular protective
activity of red wine. The general consensus is that this
is due to the antioxidant properties of the phenolic and
polyphenolic constituents present in grapes’ skin.4 Early
investigations suggested that trans-resveratrol was the
principal agent responsible for the French Paradox,
which has prompted massive examples of scientific
literature describing its antioxidant properties. Notably,
some of these reports attribute to trans-resveratrol an
antioxidant activity comparable to (or even higher than)
R-tocopherol (1),5 nature’s most effective lipid-soluble
chain-breaking antioxidant,6 while others indicate that
trans-resveratrol is the main contributor to the total
antioxidant power of red wine despite the fact that it may
not be the most abundant phenol in wine.7 trans-
Resveratrol (2) has also been found to protect blood
platelets from oxidative stress at concentrations 1000
times lower than ascorbic acid (vitamin C).8
no kinetic data have been reported regarding its reactiv-
ity toward peroxyl radicals, the key feature for chain-
breaking antioxidant activity. DFT calculations suggest
that the mechanism of reaction of 2 with chain-propagat-
ing peroxyl radicals is hydrogen-atom transfer,9 the rate
of which depends to some extent on the O-H bond
dissociation enthalpy (BDE).10 Due to delocalization of
the unpaired electron on both the A and B rings,9 the
phenoxyl radical arising from hydrogen abstraction from
the 4′-OH group will be significantly more stabilized than
the corresponding aryloxyl radicals obtained by abstrac-
tion from the hydroxyl groups in position 1 or 3, as also
predicted by DFT calculations.11 Since the BDE value for
the 4′ hydroxyl is expected to be significantly lower than
for 1- and 3-hydroxyl groups, trans-resveratrol can be
regarded, to a first approximation, as a monofunctional
phenol. Since the calculated 4′-OH BDE for trans-
resveratrol is about 3.1 kcal/mol higher than that of
R-tocopherol,9 its reported extraordinary antioxidant
activity is surprising. One possible explanation of res-
veratrol’s radical scavenging ability in vivo and in crude
plant extracts is that it acts synergistically with other
antioxidants, as recently claimed by Liu and co-workers.12
These authors reported that the strong antioxidant
activity of resveratrol and some of its derivatives in
(4) Renaud, S.; Lorgeril, M. Lancet 1992, 339, 1523-1526.
(5) Fauconneau, B.; Waffo-Teguo, P.; Huguet, F.; Barrier, L.; De-
cendit, A.; Merillon, J .-M. Life Sci. 1997, 61, 2103-2110. Soares, D.
G.; Andreazza, A. C.; Salvador, M. J . Agric. Food Chem. 2003, 51,
1077-1080.
(6) Burton, G. W.; Ingold, K. U. Acc. Chem. Res. 1986, 19, 194-
Despite the great interest in resveratrol as an anti-
oxidant, the majority of investigations have relied on
indirect methods to assess its antioxidant activity, and
201.
(7) Alonso, A. M.; Dom`ınguez, C.; Guille´n, D. A.; Barroso, C. G. J .
Agric. Food. Chem. 2002, 50, 3112-3115.
(8) Olas, B.; Wachowicz, B. Thromb. Res. 2002, 106, 143-148.
(9) Wright, J . S.; J ohnson, E. R.; Di-Labio, G. A. J . Am. Chem. Soc.
2001, 123, 1173-1183.
* To whom correspondence should be addressed. Phone: +39-051-
243218.
(1) Langcake, P.; Pryce, R. J . Physiol. Plant. Pathol. 1976, 9,
(10) Lucarini, M.; Pedrielli, P.; Pedulli, G. F.; Cabiddu, S.; Fattuoni,
C. J . Org. Chem. 1996, 61, 9259-9263.
77-86.
(11) Cao, H.; Pan, X.; Li, C.; Zhou, C.; Deng, F.; Li, T. Bioorg. Med.
Chem. Lett. 2003, 13, 1869-1871.
(2) Siemann, E. H.; Creasy, L. L. Am. J . Enol. Vitic. 1992, 43,
49-52.
(12) Fang, J .-G.; Lu, M.; Chen, Z.-H.; Zhu, H.-H.; Li, Y.; Yang, L.;
Wu, L.-M.; Liu, Z.-L. Chem. Eur. J . 2002, 8, 4191-98.
(3) St Leger, A. S.; Cochrane, A. L.; Moore, F. Lancet 1979, 1, 1017.
10.1021/jo0351825 CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/15/2003
9654
J . Org. Chem. 2003, 68, 9654-9658

Co-Antioxidant Behavior of Monofunctional Phenols
SCHEME 1
micelles arises from its ability to regenerate R-toco-
pherol.
We have recently proposed a quantitative model ra-
tionalizing regeneration among antioxidants.13 This model,
which is based on kinetic and thermochemical data
concerning the reactants involved in the potential regen-
eration reactions, has been satisfactorily applied to the
regeneration of R-tocopherol in homogeneous solution.
It has been shown (see Scheme 1) that a co-antioxidant
(CoAH) may effectively recycle vitamin E (AH) when the
O-H bond dissociation enthalpy (BDE) for CoAH is
lower, or at least comparable to that of AH, provided the
rate constant for regeneration, kr, is at least 103 M-1 s-1
and that the rate constants for the reactions of peroxyl
radicals, ROO•, with A• and CoA• are similar.
Resveratrol constitutes an excellent compound to fur-
ther test this general model of regeneration among
monofunctional phenols. We have also extended the
investigation to 2,6-di-tert-butyl-4-styrylphenol (3, a syn-
thetic hydroxystilbene) and to three phenols commonly
used as antioxidants and co-antioxidants: 2,6-di-tert-
butyl-4-methylphenol (4, BHT), 2,6-di-tert-butyl-4-meth-
oxyphenol (5, BHA), and 2,4,6-trimethoxyphenol (6).
F IGURE 1. Oxygen consumption traces recorded at 30 °C
during the autoxidation of styrene 7.5 M initiated by AMVN
(5 × 10-2 M), in the absence of any antioxidant (a) and in the
presence of the following: (b) 5.0 × 10-5 BHA; (c) 5.0 × 10-5
M R-TOH; (d) 5.0 × 10-5 M R-TOH and 5.0 × 10-5 BHA; (e)
5.0 × 10-5 M R-TOH and 1.0 × 10-4 BHA. The inset reports
the length of the induction period as a function of the
concentration ratio of the two antioxidants.
Kr can be estimated from the difference between the
bond dissociation enthalpies (BDE) of the two equilibrat-
ing antioxidants, ∆BDE ) BDE(CoAH) - BDE(R-TOH).
Because the entropy change, ∆S, of the hydrogen atom
exchange is negligible,14 eq 3 holds.
-RT ln Kr ) ∆G ≈ ∆BDE
(3)
An important condition for the recycling of R-TOH is
that the two antioxidants and the corresponding radicals
are in equilibrium, i.e., that the hydrogen exchange is
faster than the combination of R-TO• with peroxyl
radicals, this being true when eq 4 is satisfied.13
[CoAH] g (kA/kr)[ROO•]
(4)
The peroxyl radical concentration can be determined
by eq 5,15 where kinh is the rate constant for the reaction
of R-TOH with ROO•.
Resu lts a n d Discu ssion
[ROO•] ≈ Ri/(2kinh[R-TOH])
(5)
Con tr olled Au toxid a tion Stu d ies. The autoxidation
of a substrate initiated at a constant rate, Ri, by the
thermal decomposition of an azo compound and contain-
ing a specific concentration of R-tocopherol gives an
induction period whose duration is denoted as τ0. Under
the same conditions but with the simultaneous addition
of a co-antioxidant capable of regenerating R-TOH from
the corresponding R-tocopheroxyl radical, the reaction
will have an induction period, τ, as given by eq 113
In a typical autoxidation experiment,16 eq 4 holds if kr
g 3 × 103 M-1 s-1. It is therefore important to know
the value of kr since the experimental regeneration
coefficient will be lower than the value predicted by eq
2, if this constant is too small. The experimental value
of R can be obtained by plotting τ against the ratio
[CoAH]/[R-TOH]; intercept ) τ0, slope ) τ0R.
BHA (5) is a relatively poor antioxidant able only to
retard the autoxidation of styrene initiated by AMVN at
30 °C in chlorobenzene (Figure 1). The rate constant for
reaction of BHA with peroxyl radicals was determined
by analyzing the autoxidation trace by the equation
[CoAH]
τ ) τ0 + τ0R
R ) Kr
(1)
(2)
[R-TOH]
developed by Denisov.17 The value obtained, viz., kinh
)
kCoA
kA
1.2 × 105 M-1 s-1, is in excellent agreement with previous
(14) Lucarini M., Pedulli G. F., Cipollone M. J . Org. Chem. 1994,
59, 5063-5070.
where R is the regeneration coefficient (eq 2) whose value
may lie between 0 (no regeneration) and 1 (complete
regeneration).
(15) Equation 5 is the simplified form of a more complete equation
reported in ref 13 and is obtained assuming that the reaction of the
co-antioxidant with peroxyl radicals is negligible with respect to that
of R-TOH.
(13) Amorati, R.; Ferroni, F.; Lucarini, M.; Pedulli, G. F.; Valgimigli,
L. J . Org. Chem. 2002, 67, 9295-9303.
(16) Ri ) 5 × 10-9 M s-1; kinh
) 3.2 × 106 M-1 s-1; kA ) 1 ×
R-TOH
108 M-1 s-1
.
J . Org. Chem, Vol. 68, No. 25, 2003 9655

Amorati et al.
TABLE 1. Th er m och em ica l a n d Kin etic P a r a m eter s for
th e In vestiga ted P h en olic An tioxid a n ts Mea su r ed in
Ben zen e a t 298 K u n less Oth er w ise Noted
d
BDE
kinh
∆BDEe
R
antioxidant (kcal M-1
)
(M-1 s-1) (kcal M-1
)
Kr
exptlf
1
2
3
4
5
6
78.2 ( 0.3b 3.2 × 106
83.7a
2.0 × 105
+5.5
+0.7
+2.8
+0.1
+1.8
9 × 10-5
0.3
0
0.70
0
1
0
78.9 ( 0.2c 2.7 × 104
81.0 ( 0.1b 8.2 × 103
78.3 ( 0.1b 1.2 × 105
80.0 ( 0.1b 2.3 × 105
9 × 10-3
∼1
0.05
Estimated using the group additivity rule20 under the as-
a
sumption that the contribution of a p-3,5-dihydroxystyryl sub-
b
stituent is the same as that of a p-styryl substituent. Reference
d
10. c Reference 20. Measured in chlorobenzene, for these values
the standard error is approximately 10%. e BDE(CoAH) - BDE(R-
TOH). f The regeneration coefficient R refers to systems R-TOH/
co-antioxidant.
data.18 When R-TOH was added to the system, BHA
became a very good inhibitor (Figure 1, traces d-e), that
completely recycled R-tocopherol; the regeneration coef-
ficient R was ca. 1 (Figure 1, inset).
Because the O-H BDE values for BHA and R-TOH are
almost identical (see Table 1), the equilibrium constant
for the regeneration reaction Kr is equal to 1. Since the
experimental condition requiring fast equilibration for
the hydrogen exchange reaction is guaranteed by the kr
value of 5.9 × 103 M-1 s-1, determined in the separate
F IGURE 2. (A) Oxygen consumption traces recorded during
the autoxidation of styrene 7.5 M at 30 °C initiated by AMVN
(5 × 10-3 M) in the absence of any antioxidant (a) and in the
presence of the following: (b) 2,4,6-trimethoxyphenol 5.0 ×
10-6 M; (c) R-TOH 5.0 × 10-6 M; (d) R-TOH and 2,4,6-
trimethoxyphenol both 5.0 × 10-6 M. (B) Oxygen consumption
traces recorded during the autoxidation of styrene 7.5 M at
30 °C initiated by AMVN (5 × 10-3 M) in the absence of any
antioxidant (a) and in the presence of the following: (b) BHT
5.0 × 10-6 M; (c) R-TOH 5.0 × 10-6 M; (d) R-TOH and BHT
both 5.0 × 10-6 M.
experiments reported below, the rate constant ratio kCoA
/
kA is also close to 1.
Both BHT (4) and 2,4,6-trimethoxyphenol (6) are
moderately active chain-breaking antioxidants able to
retard thermally initiated (AMVN) styrene autoxidation
but not giving a sharply defined inhibition period. Their
rate constants for peroxyl radical trapping, measured by
the method reported by Darley-Usmar and co-workers,19
are 8.2 × 103 M-1 s-1 (4) and 2.3 × 105 M-1 s-1 (6), and
their O-H BDE values are 81.0 and 80.0 kcal/mol,
respectively. Unlike BHA (5), neither 4 nor 6 was able
to regenerate R-tocopherol (1) when they were mixed with
equimolar quantities of R-TOH (R ) 0). At the end of the
inhibited period, when R-tocopherol was completely
consumed, the oxygen uptake was still mildly retarded
due to presence of the co-antioxidant. This behavior
can be simply explained in terms of eq 2 by considering
that the Kr values of ca. 0 and 0.05 (estimated for 4 and
6, respectively, from the ∆BDE differences reported in
Table 1) are too small to give regeneration of R-TOH
(Figure 2).
however, is not completely unexpected in view of its O-H
BDE value, which is predicted by DFT calculations9 to
be 3.1 kcal/mol higher than that of R-TOH. An experi-
mental BDE value for 2 is not yet available but can be
estimated with reasonable accuracy by using the group
additivity rule.10,20 With the assumption that the contri-
bution of a p-3,5-dihydroxystyryl substituent is the same
as that of a p-styryl substituent, a BDE ) 83.7 kcal/mol
is obtained for the 4′ hydroxyl group. Considering the
large BDE difference between 2 and R-tocopherol (+5.5
kcal/mol), Kr is predicted to be 9.2 × 10-5, and assuming
that the ratio kCoA/kA is nearly 1 (as is observed for 5,
vide supra), the resulting regeneration coefficient R would
be negligible, i.e., no regeneration of R-tocopherol from
2 should be expected. This is, indeed, the case since
Figure 3A reveals that the inhibition of styrene autoxi-
dation by mixtures of 2 and R-TOH (1) is simply the sum
of individual effects and synergism does not occur.
We have also studied the antioxidant behavior of
mixtures of resveratrol (2) and BHT (4) because 4 has
an O-H BDE value (81.0 kcal/mol)14 which is 2.7 kcal/
mol below that estimated for 2 (vide supra), correspond-
ing to a Kr ≈ 96. This would imply that BHT (4), although
significantly less reactive than resveratrol (2) toward
peroxyl radicals (see Table 1) could quantitatively regen-
erate 2 from the corresponding phenoxyl radical. Figure
Despite being considered a very good antioxidant,
resveratrol (2) behaves in homogeneous solution as a mild
retarder of the thermally initiated autoxidation of styrene
(Figure 3). Its rate constant for the reaction with peroxyl
radical in chlorobenzene at 30 °C, kinh ) 2.0 × 105 M-1
s-1, obtained as already described,19 is about 16 times
lower than that of R-TOH (Table 1). Its lower reactivity,
(17) Denisov, E. T.; Khudyakov, I. V. Chem. Rev. 1987, 87, 1313.
Denisov, E. T. Liquid-Phase Reaction Rate Constants; Plenum: New
York, 1974. Amorati, R.; Pedulli, G. F.; Valgimigli, L.; Attanasi, O. A.;
Filippone, P.; Fiorucci, C.; Saladino, R. J . Chem. Soc., Perkin Trans. 2
2001, 2142-2146.
(18) Burton, G. W.; Doba, T.; Gabe, E. J .; Hughes, L.; Lee, F.; Prasad,
L.; Ingold, K. U. J . Am. Chem. Soc. 1985, 107, 7053.
(19) Darley-Usmar, V. M.; Hersey, A.; Garland, L. G. Biochem.
Pharm. 1989, 38, 1465-1469.
(20) Brigati, G.; Lucarini, M.; Mugnaini, V.; Pedulli, G. F. J . Org.
Chem. 2002, 67, 4828-4832.
9656 J . Org. Chem., Vol. 68, No. 25, 2003

Co-Antioxidant Behavior of Monofunctional Phenols
peroxyl radicals, the higher reactivity of CoA• with
respect of the R-tocopheroxyl radical may possibly be due
to the fact that the unpaired electron is delocalized into
the styryl substituent of CoA•, so that additional modes
of attack by peroxyl radicals are possible.
Deter m in a tion of th e Regen er a tion Ra te Con -
sta n ts, kr . The rate constant for regeneration of R-toco-
pherol by BHA (5) (eq 6) has been obtained in benzene
solution at room temperature, by the kinetic EPR method
previously described.13 For convenience, we measured the
reverse rate constant k-r, i.e., the rate for hydrogen atom
transfer from R-tocopherol to the phenoxyl radical of 5,
from which kr was obtained using the Kr value estimated
from the known ∆BDE difference (Table 1). The method
consisted in monitoring the decay of a spectral line of the
very persistent 2,6-di-tert-butyl-4-methoxyphenoxyl radi-
cal, not superimposed to those of R-TO•, after a fast
injection of a concentrated stock solution of R-TOH (final
concentration ca. 1 mM) into the sample tube, while
continuously bubbling with nitrogen to induce rapid
mixing. Under the experimental conditions employed, the
EPR signal decayed completely in approximately 5-10
s after the R-tocopherol injection, following good pseudo-
first-order kinetics. The measured rate constant, k-r, was
7.0 × 103 M-1 s-1 from which the value kr ) 5.9 × 103
M-1 s-1 in benzene at 298 K was calculated using the
difference of 0.1 kcal/mol between the BDE’s of 5 and 1.
Attempts to measure the rate of reaction of resveratrol
with R-tocopheroxyl radicals were unsuccessful due to
limited solubility of 2 in benzene. Similarly, we could not
measure the rate of regeneration of resveratrol by BHT
(4). However, this value could be reliably estimated by
combining our previous thermodynamic data20 with the
kinetic data reported by Mahoney and DaRooge in their
pioneering work.21 The rates for reaction of some 4-sub-
stituted phenols (XC6H4OH), with X) CH3O, (CH3)3C,
and H) with the phenoxyl radical from 2,4,6-tri-tert-
butylphenol (7) (eq 7) were measured by stopped flow-
kinetic EPR as 6.0 × 103, 1.1 × 102, 8.0 M-1 s-1
respectively.21
F IGURE 3. (A) Oxygen consumption traces recorded during
the autoxidation of styrene 4.3 M at 30 °C initiated by AMVN
(5.0 × 10-3 M) in the absence of any antioxidant (a) and in
the presence of the following: (b) resveratrol 5.0 × 10-6 M; (c)
R-TOH 5.0 × 10-6 M; (d) R-TOH and resveratrol both 5.0 ×
10-6 M. (B) Oxygen consumption traces recorded during the
autoxidation of styrene 4.3 M at 30 °C initiated by AMVN (5.0
× 10-3 M) in the presence of the following: (a) BHT 5.0 × 10-6
M; (b) resveratrol 5.0 × 10-6 M; (c) resveratrol and BHT both
5.0 × 10-6 M; (d) resveratrol 5.0 × 10-6 M and BHT 1.0 ×
10-5 M.
3B shows that resveratrol (plot b) is more effective in
inhibiting the autoxidation of styrene than BHT alone
(plot a) although neither of these antioxidants gave well-
defined inhibition periods. With 1:1 and 1:2 resveratrol/
BHT mixtures the oxygen uptake traces were the same
as those produced by doubling and tripling the concen-
tration of (2) (plots c and d). This indicates that BHT is
regenerating resveratrol. The efficiency of regeneration,
obtained by simulating the experimental autoxidation
plots by a stochastic method previously described,13
provides an R value of 1 (100% regeneration), as would
be predicted by eq 3.
To stress the importance of ∆BDE as the primary
factor governing regeneration among monophenolic an-
tioxidants, we have investigated the antioxidant behavior
of mixtures of R-TOH (1) and 2,6-di-tert-butyl-4-styrylphe-
nol (3). The latter is structurally related to resveratrol
and has an experimental O-H BDE of 78.9 kcal/mol,20
i.e., 4.8 kcal/mol lower than 2. The rate constant for the
trapping of peroxyl radicals by 3 was determined19 as 2.7
× 104 M-1 s-1. As shown in Figure S1, compound 3 regen-
erates R-TOH with good efficiency (R ) 0.7, see inset),
despite the fact that its antioxidant activity is lower than
that for 6. This value is higher than the efficiency
predicted by eq 3 considering the ∆BDE between 1 and
3 (+0.7 kcal/mol, corresponding to Kr ) 0.3) and assum-
ing a ratio kCoA/kA ) 1. Indeed, according to eq 2, a value
R ) 0.7 requires that kCoA/kA ) 2.3, i.e., that the phenoxyl
radical from 3 reacts with peroxyl radicals about twice
as fast as the tocopheroxyl radical. Although no experi-
mental data are available on the rate of radical-radical
combination between the phenoxyl radical from 3 and
(21) Mahoney, L. R.; DaRooge, M. A. J . Am. Chem. Soc. 1975, 97,
4722-4731.
J . Org. Chem, Vol. 68, No. 25, 2003 9657

Amorati et al.
kept higher than 30 °C to accelerate the decomposition
of the azo initiator. Matched autoxidation experiments
performed in the presence of 4.3 M styrene (see Figure
4) showed induction periods identical to those obtained
from the disappearance of R-TOH. The plots obtained in
the presence of both antioxidants indicate that the
lengthening of the induction period is due to a slower
consumption of R-tocopherol. No retarding of R-TOH
consumption was observed when resveratrol was used
as the co-antioxidant and induction periods recorded
during autoxidation in the presence or absence of revera-
trol are almost superimposable (plot b). Simulations of
the time dependence of the R-tocopherol consumption
under the experimental conditions employed (see the
Experimental Section) matched the experimental obser-
vations.
F IGURE 4. R-TOH consumptions observed from a 1 × 10-4
M chlorobenzene solution at 60 °C during the thermal decom-
position of 5 × 10-3 M AIBN in the absence of co-antioxidant
(O) and in the presence of 1 × 10-4 M resveratrol (2) and 1 ×
10-4 M BHA (b). The plot also shows the oxygen uptake traces
obtained under similar conditions except for the presence of
4.3 M styrene (traces a, b, and c refer to the experiments
carried out in the presence of only R-TOH, the mixture (1:1)
R-TOH/resveratrol, and the mixture (1:1) R-TOH/BHA, respec-
tively). Dotted lines show computer simulation of R-TOH
consumption.
Con clu sion
The general model we have recently proposed13 can
explain and predict synergistic behavior among mono-
functional phenolic antioxidants. In general, the regen-
eration of an antioxidant AH by a co-antioxidant CoAH
is possible when the rate constant, kr, is 103 M-1 s-1 or
higher and CoAH has an O-H BDE value lower or at
least close to that of AH. When the ratio between the
rate constants for the reaction of peroxyl radicals with
the corresponding phenoxyl radicals kCoA/kA is ∼1, the
above conditions are verified when the ∆BDE difference
BDE(CoAH) - BDE(AH) is less than 1 kcal/mol.
Contrary to general belief, resveratrol in homogeneous
solution is neither an outstanding antioxidant nor ca-
pable of effectively regenerating R-tocopherol. Indeed,
other phenolic derivatives present in red wine might be
more likely responsible for its antioxidant activity.
Among them gallic acid and the numerous flavonoids
containing the catechol ring which are characterized by
low O-H BDE values due to intramolecular hydrogen
bonding.23 Whether the reported antioxidant and co-
antioxidant virtues of resveratrol would manifest them-
selves only in heterogeneous systems and whether our
model would be suited to describe autoxidations in such
media will be a matter for further work.
When the logarithm of the rate constants is plotted
against the ∆BDE difference BDE(XC6H4OH) - BDE(7)
an excellent linear correlation was obtained (Figure S2),
similar to our previously reported plot in a closely related
system.13 The values for the BDE(XC6H4OH) are 82.8,
85.3, and 88.3 kcal/mol for X ) MeO, Me3C, and H,
respectively,20 and 81.2 kcal/mol for 7,22 affording the
empirical correlation log k7 ) 4.45 - 0.52 (∆BDE). By
substituting in this equation the ∆BDE values for res-
veratrol, the value k7 ) 1.5 × 103 M-1 s-1 is obtained
which provides, together with ∆BDE, the rate constant
for the reverse reaction k-7 ) 9.6 × 104 M-1 s-1. Since
2,4,6-tri-tert-butylphenol (7) and BHT (4) have almost
identical O-H BDE’s and the same substituents ortho
to the OH group, this value also represents a good
estimate for the rate of regeneration of resveratrol (2)
by BHT (4), i.e., kr ∼ 9.6 × 104 M-1 s-1, a result in
agreement with the observed synergistic behavior of
resveratrol/BHT co-antioxidant mixtures.
P r od u ct Stu d ies. The complete regeneration of R-to-
copherol by BHA (5) was confirmed by studying the time
dependence of the concentrations of the two antioxidants,
measured by HPLC-MS using electrospray ionization
(ESI-MS), during the thermal decomposition of AIBN at
60 °C, i.e., under experimental conditions similar to those
used when carrying out the autoxidation reactions. In
these experiments no styrene was introduced to avoid
interference with the phenol analytes in the ESI-MS
analysis due to matrix effects, and the temperature was
Ack n ow led gm en t. Financial support from the Uni-
versity of Bologna and MIUR (Research project “Free
Radical Processes in Chemistry and Biology: funda-
mental aspects and applications in environment and
material sciences”) is gratefully acknowledged. We also
thank a reviewer for substantially improving the En-
glish of the original manuscript.
Su p p or tin g In for m a tion Ava ila ble: Figures S1 and S2
and the Experimental Section. This material is available free
of charge via the Internet at http://pubs.acs.org.
J O0351825
(22) Mahoney, L. R.; Ferris, F. C.; DaRooge, M. A. J . Am. Chem.
Soc. 1969, 91, 3883.
(23) Lucarini, M.; Mugnaini, V.; Pedulli, G. F. J . Org. Chem. 2002,
67, 928-931.
9658 J . Org. Chem., Vol. 68, No. 25, 2003