Electronic and hydrogen bonding effects on the chain-breaking activity of sulfur-containing phenolic antioxidants

Article abstract of DOI:10.1021/jo060281e

A kinetic and thermodynamic investigation of phenols para-substituted with thiyl (SR), sulfinyl (SOR), and sulfonyl (SO₂R) groups and ortho-substituted with thiyl groups is reported. The effect of the sulfur substituents on the O-H bond dissoci

Full text of DOI:10.1021/jo060281e

Electronic and Hydrogen Bonding Effects on the Chain-Breaking
Activity of Sulfur-Containing Phenolic Antioxidants
Riccardo Amorati, Maria Grazia Fumo, Stefano Menichetti,*,† Veronica Mugnaini, and
Gian Franco Pedulli*
Dipartimento di Chimica Organica “A. Mangini”, UniVersita` di Bologna, Via S. Donato 11,
40126 Bologna, Italy, and Dipartimento di Chimica Organica, UniVersita` di Firenze,
Via della Lastruccia 13, 50019 Sesto Fiorentino (FI), Italy
gianfranco.pedulli@unibo.it
ReceiVed February 10, 2006
A kinetic and thermodynamic investigation of phenols para-substituted with thiyl (SR), sulfinyl (SOR),
and sulfonyl (SO2R) groups and ortho-substituted with thiyl groups is reported. The effect of the sulfur
substituents on the O-H bond dissociation enthalpy values, BDE(O-H), was measured by means of the
EPR radical equilibration technique and the reactivity toward peroxyl radicals, kinh, of these phenolic
antioxidants was determined by inhibited autoxidation studies. An inverse correlation between these two
parameters was found. A p-SMe substituent decreased the BDE(O-H) value to a lesser extent than a
p-OMe group (-3.6 vs -4.4 kcal/mol), whereas the effect of the same groups in an ortho position showed
an opposite trend (-0.85 vs -0.2 kcal/mol). The latter result is explained in terms of the different strength
of the intramolecular hydrogen bond between the OH proton and the sulfur or oxygen substituents in
ortho derivatives. ESI-MS analysis of the products formed by reacting the sulfides with peroxyl radicals
from the azoinitiator AIBN revealed the formation of a complex mixture of products, which may play an
important role in determining the overall antioxidant activity of the parent compounds.
Introduction
compounds might be better antioxidants than the analogue
phenols containing oxygen substituents.
Among inhibitors of free radical oxidation of hydrocarbon
substrates, sulfur-containing phenols are of particular interest
since they can behave both as chain-breaking antioxidants by
scavenging peroxyl radicals through hydrogen atom transfer
from the phenol group (eq 1) and as preventive antioxidants by
decomposing hydroperoxides to the corresponding alcohols by
a reaction involving nucleophilic attack by sulfur on oxygen.1
Moreover, since sulfur is considered to be more effective than
oxygen at stabilizing a neighboring radical center,2 these
kinh
ROO• + ArOH 8 ROOH + ArO•
ROO• + ArO• f products
(1)
(2)
A series of derivatives related to 1-thio-R-tocopherol were
investigated in the late 1980s by Ingold and co-workers and,
contrary to expectations, were less reactive toward peroxyl
radicals than the corresponding chromane derivatives.3 In
particular, the lower reactivity was found to be due both to the
smaller kinh (see eq 1) and to the fact that they trap fewer than
two peroxyl radicals per molecule. This also indicates that the
* To whom correspondence should be addressed. Tel: +39 51 2095681.
Fax: +39 51 2095688.
† Universita` di Firenze.
(1) Denisov, E. T.; Afanas’ev, I. B. Oxidation and Antioxidants in
Organic Chemistry and Biology; CRC Press: Taylor & Francis, Boca Raton,
2005.
(3) Zahalka, H. A.; Robillard, B.; Hughes, L.; Lusztyk, J.; Burton, G.
W.; Janzen, E. G.; Kotake, Y.; Ingold, K. U. J. Org. Chem. 1988, 53, 3739-
3745.
(2) Luedtke, A. E.; Timberlake, J. W. J. Org. Chem. 1985, 50, 268-
270.
10.1021/jo060281e CCC: $33.50 © 2006 American Chemical Society
Published on Web 07/25/2006
J. Org. Chem. 2006, 71, 6325-6332
6325

Amorati et al.
SCHEME 1
mechanism of inhibition is somewhat different from that one
simply consisting in the two reactions 1 and 2. On the other
hand, an investigation of 2,3-dihydrobenzo[b]furan-5-ol and 2,3-
dihydrobenzo[b]thiophene-5-ol, i.e., analogues of tocopherol
having a five-member condensed ring, suggested that the
antioxidant capacity of the sulfur derivative is better than that
of the oxygenated one.4
In a comparative study of the antioxidant properties of cyclic
aromatic amines,5 sulfur-containing derivatives are less reactive
than the corresponding oxygenated ones. Actually, substitution
of the heterocyclic oxygen atom of phenoxazine with sulfur
(phenothiazine) or selenium (phenoselenazine) increased the
dissociation enthalpy (BDE) of the N-H bond by 2.1 and 3.2
kcal/mol, respectively, thus making the hydrogen atom transfer
from the latter two amines to an attacking radical less exoer-
gonic. Accordingly, the rate constants for the reaction with both
TABLE 1. EPR Spectral Parameters for Phenoxyl Radicals from
Substituted Phenols Measured at Room Temperature in Benzene
Solution; Hyperfine Splitting Constants in gauss (1 G ) 0.1 mT)
phenol
a(2Hm)
a(3HMe
)
a(other)
g-factor
1
3
4
6
7
8
9
1.30
2.19
1.32
2.12
2.24
1.56
2.0055
2.0048
2.0054
2.0048
2.0054
2.0053
2.0050
ROO• and CH3 radicals decreased in the order O > S > Se.
•
In recent studies on the antioxidant effect of bis(4-hydroxy-
3,5-di-tert-butylphenyl) sulfides and polysulfides,6 it has been
reported that these compounds retard the hydrocarbon oxidation
both by terminating the oxidation chains by hydrogen transfer
to peroxyl radicals and by catalyzing the decomposition of
hydroperoxides. The inhibiting activity was found to increase
with the number of sulfur atoms. A similar investigation on
related sulfides also reached the conclusion that these consider-
ably surpass the antioxidative ability of the commercial 2,6-di-
tert-butyl-4-methylphenol (BHT).7
4-Thiaflavan derivatives, obtained by introducing a sulfur
atom on the C ring of flavan, have been synthesized and their
antioxidant activity evaluated.8 These heterocyclic sulfides
showed antioxidant properties similarly to the parent flavans,
while their oxidation to the corresponding sulfoxides gave rise
to an almost complete loss of activity.
To better clarify the behavior of sulfur derivatives and to
estimate on a quantitative basis the effect of sulfur substituents
on the antioxidant properties of phenols, we have undertaken a
thermodynamic and kinetic study of 2,6-di-tert-butylphenols
substituted at the 4 position with thiyl (SR), sulfinyl (SOR),
and sulfonyl (SO2R) groups. The last ones were investigated
because sulfides are easily oxidized to give sulfoxides or
sulfones under the experimental conditions employed to carry
out autoxidation studies. Therefore, the structure of the oxidation
inhibitor may change during the course of the reaction with a
consequent change in the antioxidant behavior.
We have also investigated a couple of phenols containing
thiyl substituents ortho to the hydroxyl group because these
structures are present in the active site of galactose oxidase,9
where the tyrosyl radical is connected to a neighboring cysteine
through a thioether bond.10 Knowledge of the thermodynamic
1.22, 1.38a
1.29, 1.59
1.13, 1.84
5.53
1.89
10.17
0.34 (9Ht-But
)
a From computer simulation.
and kinetic behavior of such modified amino acid and of the
effect of sulfur on the stability of o-phenoxyl radical centers is
essential to develop new artificial enzymes for the green
oxidation of alcohols.11
Results
The investigated phenols are shown in Scheme 1. These
structures were chosen in order to study the effect of sulfur
substituents in the ring positions conjugated with the hydroxyl
group. Ortho tert-butyl groups are helpful, when performing
EPR studies, for increasing the persistency of the corresponding
phenoxyl radicals due to the sterical protection of the radical
oxygen center.
EPR Spectra. The phenoxyl radicals from the examined
compounds were produced at room temperature inside the cavity
of an EPR spectrometer by reacting the phenols with alkoxyl
radicals generated photolytically from di-tert-butyl peroxide in
deoxygenated benzene solutions:
ArOH + RO• f ROH + ArO•
(3)
Highly persistent radicals showing easily interpretable EPR
spectra were always obtained with sulfides and sulfones; their
spectral parameters (hyperfine splitting constants and g-factors)
shown in Table 1 are consistent with those previously reported
for similar radicals.12 On the other hand, the sulfoxides 2 and
5, under the same experimental conditions, afforded only weak
EPR signals due to transient paramagnetic species. This is not
unexpected, since it has been reported by Gilbert et al. that
irradiation of 4-sulfinyl phenols, either in the presence or in
the absence of di-tert-butyl peroxide, affords sulfinyl radicals
formed by cleavage of the S(O)-R bond.12
(4) Malmstro¨m, J.; Jonsson, M.; Cotgreave, I. A.; Hammarstro¨m, L.;
Sjo¨din, M.; Engman, L. J. Am. Chem. Soc. 2001, 123, 3434-3440.
(5) Lucarini, M.; Pedrielli, P.; Pedulli, G. F.; Valgimigli, L.; Gigmes,
D.; Tordo, P. J. Am. Chem. Soc. 1999, 121, 11546-11553.
(6) Farzaliev, V. M.; Allakhverdiev, M. A.; Sattar-zade, R. I.; Rzaeva,
I. A. Russ. J. Appl. Chem. 2001, 74, 2083-2086.
(7) Prosenko, A. E.; Terakh, E. I.; Kandalintseva, N. V.; Pinko, P. I.;
Gorokh, E. A.; Tolstikov, G. A. Russ. J. Appl. Chem. 2001, 74, 1899-
1902.
(8) (a) Capozzi, G.; Lo Nostro, P.; Menichetti, S.; Nativi, C.; Sarri, P.
Chem. Comm. 2001, 551-552. (b) Menichetti, S.; Aversa, M. C.; Cimino,
F.; Contini, A.; Viglianisi, C.; Tomaino, A. Org. Biomol. Chem. 2005, 3,
3066-3072.
(9) Whittaker, J. W. Chem. ReV. 2003, 103, 2347-2363
(10) Ito, N.; Phillips, S. E. V.; Stevens, C.; Ogel, Z. B.; McPherson, M.
J.; Yadav, K. D. S.; Knowles. P. F. Nature 1991, 350, 87.
From the data of Table 1 it appears that the meta splittings
in the phenoxyl radicals from sulfides 1 and 4 (ca 1.30 G) are
(11) Pratt, R. C.; Stack, T. D. P. J. Am. Chem. Soc. 2003, 125, 8716-
8717.
(12) Gilbert, B. C.; Hanson, P.; Isham, W. J.; Whitwood, A. C. J. Chem.
Soc., Perkin Trans. 2 1988, 2077-2085.
6326 J. Org. Chem., Vol. 71, No. 17, 2006

Sulfur-Containing Phenolic Antioxidants
for the species ArOH were calculated, in the assumption that
the entropic term can be neglected,13 by means of eq 5 from Ke
and the BDE value of the reference phenol Ar′OH.
BDE(ArO-H) = BDE(Ar′O-H) - RT ln(Ke)
(5)
Rate Constants of Inhibition. The determination of the rate
constant for the reaction with peroxyl radicals, kinh (eq 10), of
the above derivatives was made by studying the inhibition of
the thermally initiated autoxidation of either cumene or styrene.
Cumene was preferentially used with weak inhibitors, because
its lower oxidizability with respect to styrene allows a better
detection of the induction period.
Ri
9
initiator
8 R•
(6)
(7)
FIGURE 1. Experimental (upper) and simulated room-temperature
EPR spectrum obtained by photolyzing a mixture of BHT (0.055 M)
and 7 (0.19 M) in deoxygenated benzene containing 10% di-tert-butyl
peroxide. The 12 lines of the phenoxyl radical from BHT are marked
with a dot.
R• + O2 f ROO•
kp
9
ROO• + RH
8 ROOH + R•
(8)
(9)
lower than those measured (a(2Hm) ) 1.95G)13 in the corre-
sponding radical without 4-substituents (i.e., that obtained from
2,6-di-tert-butylphenol), whereas in the radicals from sulfones
3 and 6 the coupling at the meta hydrogens (ca 2.15 G) is larger.
This behavior is peculiar of electron donor and electron acceptor
substituents, respectively.14 Since also the g-factors of the
radicals from sulfides and sulfones differ considerably, being
larger for the former ones, there is little doubt that the observed
species result from the abstraction of the hydroxyl hydrogen
atom of the investigated phenols.
O-H Bond Dissociation Enthalpies (BDE). The O-H BDE
values for the title compounds were determined by using the
EPR radical equilibration technique that, among the various
experimental methods used for the determination of bond
strengths, seems to guarantee at present the best accuracy.13,14
To this purpose we measured the equilibrium constant, Ke, for
the hydrogen atom transfer reaction between a reference phenol
(Ar′OH) and one of the examined phenols (ArOH) and the
corresponding phenoxyl radicals (eq 4) generated under continu-
ous photolysis at room temperature (25 °C) in deoxygenated
benzene solutions containing 10% of di-tert-butyl peroxide.
2kt
ROO• + ROO• 98 products
kinh
ROO• + ArOH 8 ROOH + ArO•
(10)
(11)
ROO• + ArO• f products
The reaction was followed by monitoring the oxygen
consumption during the autoxidation with an automatic record-
ing gas absorption apparatus described previously,19 which uses
as detector a commercial differential pressure transducer. The
reactions, initiated by the thermal decomposition of AMVN
(2,2′-azobis(2,4-dimethyl-valeronitrile)), were carried out at 30
°C under controlled conditions in air-saturated solution of the
oxidizable substrate, both in the absence and in the presence of
each antioxidant. R-Tocopherol was used as reference chain-
breaking inhibitor.
The inhibition rate constants kinh of each compound were
determined by means of a kinetic treatment20 consisting in the
measure of the oxidation rate in the presence (-d[O2]/dt ) R)
and in the absence ((-d[O2]/dt)0 ) R0) of antioxidant as a
function of time. The kinh values were obtained from the slope
of the function F (eq 12) against time using the known rate
constants for the propagation reaction, kp (eq 8), of the oxidizable
substrates (RH).
ArOH + Ar′O• h ArO• + Ar′OH
(4)
2,6-Di-tert-butyl-4-methylphenol (BHT), whose revised BDE
value is 80.1 kcal/mol,15 was always used as reference phenol,
except in the case of 8, which was equilibrated with 2,4,6-
trimethylphenol (BDE ) 82.2 kcal/mol). As an example of the
quality of the EPR spectra obtained, Figure 1 shows that one
of the mixture of 7 and BHT.
Measurements were repeated under different light intensity
in order to check the constancy of Ke and, therefore, the
achievement of the equilibrium. In the calculation of Ke, the
initial concentration of ArOH and Ar′OH were used, and the
relative radical concentrations were obtained both by numerical
integration and by simulation of the EPR spectra. The BDEs
1 + R/R0 R0
kinhR0
F ) ln
-
)
t + constant (12)
1 - R/R0
R
kp[RH]
In the case of styrene, kp is 41 M-1 s-1 21
In the case of
cumene, its oxidation kinetics being dependent on substrate
.
(16) Mahoney, L. R.; Ferris, F. C.; DaRooge, M. A. J. Am. Chem. Soc.
1969, 91, 3883.
(17) Mulder, P.; Korth, H.-G.; Pratt, D. A.; DiLabio, G. A.; Valgimigli,
L.; Pedulli, G. F.; Ingold, K. U. J. Phys. Chem. A 2005, 109, 2647-2655.
(18) Burton, G. W.; Doba, T.; Gabe, E. J.; Hughes, L.; Lee, F. L.; Prasad,
L.; Ingold, K. U. J. Am. Chem. Soc. 1985, 107, 7053-7065.
(19) 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.
(13) Lucarini, M.; Pedrielli, P.; Pedulli, G. F.; Cabiddu, S.; Fattuoni, C.
J. Org. Chem. 1996, 61, 9259-9263.
(14) Brigati, G.; Lucarini, M.; Mugnaini, V.; Pedulli, G. F. J. Org. Chem.
2002, 67, 4828-4832.
(15) All the BDE values determined in benzene solution by means of
the EPR radical equilibration technique,13 based on the O-H BDE of 2,4,6-
tri-tert-butylphenol determined many years earlier by Mahoney et al.16 using
calorimetric measurements, must be downscaled due to the revision of the
BDE(OH) of unsubstituted phenol (86.7 kcal/mol).17
(20) (a) Roginsky, V.; Lissi, E. A. Food Chem. 2005, 92, 235-254. (b)
Roginsky, V. Arch. Biochem. Biophys. 2003, 414, 261-270. (c) Loshadkin,
D.; Roginsky, V.; Pliss, E. Int. J. Chem. Kinet. 2002, 34, 162-171.
(21) Howard, J. A. In Free Radicals; Kochi, J. K., Ed.; Wiley-
Interscience: New York, 1975; Vol. 2, Chapter 12.
J. Org. Chem, Vol. 71, No. 17, 2006 6327

Amorati et al.
FIGURE 2. Oxygen consumption observed at 30 °C during the AMVN
(5 × 10-3 M) initiated autoxidation of cumene (7.1 M) in the absence
(0) and in the presence of some sulfur-containing phenols (5.0 × 10-6
M).
TABLE 2. O-H Bond Dissociation Enthalpies (BDE) of Substituted
Phenols 1-9 Measured at Room Temperature in Benzene
Containing 10% Di-tert-butyl Peroxide, Rate Constants kinh for
Their Reaction with Peroxyl Radicals in Cumene (or Styrene) at 30
°C, and Number of Radicals Trapped by Each Antioxidant Molecule
(n)g
FIGURE 3. FT-IR spectra of 0.05 M of 8 (dashed line) and 9
(continuous line) in CCl4 solution.
oxidizable
1, 2, and 3 at the same concentration, respectively, were
observed. This was unexpected since d[O2]/dt values twice as
small and stoichiometric coefficients twice as large as those of
the monophenols 1-3 were predictable in the bisphenols as a
result of the presence of two hydroxyl groups, as found with
4,4′-methylenebis(2,6-di-tert-butylphenol).23 It should, however
be emphasized that a similar result (n ) 2) was reported by
Farzaliev et al. when inhibiting with 4 the AIBN initiated
autoxidation of cumene at 60 °C.6
An anomalous behavior was also shown by bis(3-tert-butyl-
4-hydroxy-5-methylphenyl)sulfide (7), which differs from 4 for
bearing a methyl rather than a tert-butyl substituent at one of
the ring positions ortho to the hydroxyl group. Actually, 7 (see
Figure 2) terminates 3 radical chains instead of the 4 expected
for bis-phenols. A discussion about the possible reasons of these
differences is reported later in this paper.
compd BDE/kcal mol-1 substratea kinh × 10-4/M-1 s-1
n
1
78.3 ( 0.1
C
S
3.0 ( 0.2
4.0 ( 0.3
0.29 ( 0.05
0.22 ( 0.01
2.7 ( 0.1
0.34 ( 0.04
0.13 ( 0.02
61 ( 10
1.7 ( 0.1
2
3
4
5
6
7
8
81.6b
81.7 ( 0.4
78.5 ( 0.2
C
C
C
C
C
S
C
S
S
1.6 ( 0.2
1.3 ( 0.4
1.7 ( 0.1
2.2 ( 0.1
1.6 ( 0.1
3.1 ( 1
1.6 ( 0.2
2.2 ( 0.1
2
82.3 ( 0.3
79.6 ( 0.2
82.2 ( 0.2
81.0 ( 0.1
80.1 ( 0.1c
0.39 ( 0.05
21 ( 1
9
BHT
1.1d
BHAe 77.4 ( 0.1c
S
11.0f
2
a Measured in cumene (C) or in styrene (S). b Estimated from the plot
of Figure 5 by using the experimentally determined values of kinh.
c Revised
value, ref 13,15. d From ref 22. e 2,6-di-tert-butyl-4-methoxyphenol. f From
ref 18. g Reported data are the mean of three or more determinations with
the corresponding standard errors.
The antioxidant activity has also been studied for phenol 8
and bis-phenol 9 containing a divalent sulfur atom ortho to the
hydroxyl groups. The oxygen uptake traces recorded during the
inhibited oxidation of cumene, shown in Figure 2, clearly
indicate that the bis-phenol 9 is by far a much better oxidation
inhibitor than 8. The ratio of the inhibition rate constants,
experimentally determined as described above, is ca. 54 in favor
of 9. Also in this case, as with 4, the number of oxidation chains
n terminated by each molecule of inhibitor is about 2 despite
the presence of two phenolic units in the molecule.
FT-IR Measurements. In the two compounds 8 and 9,
containing ortho sulfur substituents, intramolecular hydrogen
bonding between the hydroxyl proton and the sulfur atom may
take place, as revealed by previous IR24 and NMR25 studies on
related derivatives. Since this interaction has a strong effect on
the antioxidant properties of phenols,23,26 we carried out FT-
infrared measurements in order to confirm these reports and to
identify the nature of the hydrogen bonded species.
concentration,21 the kp constant was calculated as 0.32 M-1 s-1
from our previously reported 2kt value of 4.6 × 104 M-1 s-1
,
measured by following the decay of the cumylperoxyl radicals
in neat cumene by EPR,22 and the oxidizability of neat cumene
(kp/2kt0.5 ) 1.5 × 10-3 M-0.5 s-0.5).21
The term n (eq 13) represents the stoichiometric coefficient,
i.e., the number of peroxyl radicals trapped by each antioxidant
molecule, and is determined from the length of the induction
period τ during which the rate of the oxygen consumption is
strongly reduced (see Figure 2). The values of τ were obtained
by following the procedure suggested in ref 20 and described
in details in the Experimental Section. The results of these
determinations are reported in Table 2.
Riτ
n )
(13)
[ArOH]
Figure 3 shows the FT-IR spectra of both phenols (0.05 M)
recorded at room temperature in carbon tetrachloride, in the
Figure 2 shows that the three sulfur derivatives 1-3 behave
as inhibitors of the initiated autoxidation of cumene, their
antioxidant activity decreasing with increasing degree of oxida-
tion of sulfur, being larger for the sulfide and very poor for the
sulfone. When using the bis-phenols 4, 5, and 6 as inhibitors,
oxygen uptake traces surprisingly similar to those obtained with
(23) Amorati, R.; Lucarini, M.; Mugnaini, V.; Pedulli, G. F. J. Org.
Chem. 2003, 68, 5198-5204.
(24) Kovac, S.; Solcaniova, E.; Baxa, J. Tetrahedron 1971, 27, 2826-
2830.
(25) Schaefer, T.; Salmam, R. S.; Wildman, T. A.; Clark, P. D. Can. J.
Chem. 1982, 60, 342-348.
(26) Lucarini, M.; Pedulli, G. F.; Valgimigli, L.; Amorati, R.; Minisci,
F. J. Org. Chem. 2001, 66, 5456-5462.
(22) Enes, R. F.; Tome´, A. C.; Cavaleiro, J. A. S.; Amorati, R.; Fumo,
M. G.; Valgimigli, L.; Pedulli, G. F. Chem.-Eur. J. 2006, 12, 4646-4653.
6328 J. Org. Chem., Vol. 71, No. 17, 2006

Sulfur-Containing Phenolic Antioxidants
SCHEME 2
2800-3800 cm-1 range. The spectrum of the mono-phenol 8
shows a single peak in the O-H absorption region, centered at
a frequency (3375 cm-1) characteristic of an intramolecularly
hydrogen bonded species. Since the absence of lines above 3600
cm-1 indicates that no free hydroxyl group is present, 8 must
exist in a single geometry presumably similar to that adopted
by 2-(methylthio)phenol, for which IR24 and NMR25 evidence
suggests that the preferred conformation is that one where the
OH‚‚‚S interaction forces the 3p orbital on sulfur into the
benzene plane and the methyl group completely out of the
molecular plane (see Scheme 2).
The IR spectrum of the bis-phenol 9 shows, instead, two
absorptions: one centered at 3422 cm-1 and the other at 3503
cm-1. Both positions and intensities of these two bands are
independent of the phenol concentration and their frequencies
are characteristic of intramolecularly hydrogen bonded hydroxyl
groups. The former one, observed at a frequency close to that
one of the only band detected in 8, can be attributed to the OH
group H-bonded to the sulfur atom,25 whereas the latter
absorption (observed at a higher frequency and thus indicative
of a weaker H-bonding interaction) should be due to the second
hydroxyl group H-bonded either to the π-electrons of the
benzene ring or to other OH group. A possible structure for 9
is depicted in Scheme 2.
Product Studies. To clarify the reason for the anomalous n
values of compounds 4, 7, and 9, products formed when reacting
sulfides 1, 4, 7, 8, and 9 with peroxyl radicals were studied by
electrospray ionization mass spectrometry (ESI-MS),27 a tech-
nique that has been proven to give useful information about
intermediates occurring during the autoxidation reactions.23,26
An acetonitrile (ACN) solution of these sulfides (1 × 10-3 M),
containing a large excess of azo-bisisobutyronitrile (AIBN, 5
× 10-2 M), was kept at 60 °C in an open atmosphere for 1 h,
and the crude reaction mixture, diluted with methanol, was
injected in the spectrometer.
The results obtained using the negative ion detection mode,
reported in Figure 4, show that after 15-30 min, in addition to
the peak of the deprotonated parent phenol [M - 1]-, two peaks
with mass [M - 1 + 16]- and [M - 1 + 32]- are present, this
indicating that the starting phenol incorporated one and two
oxygen atoms, i.e., that it was oxidized to the corresponding
sulfoxide and sulfone, respectively. Only in the case of bis-
phenols 4, 7, and 9, a strong peak at m/z values lower than that
of the parent phenol was also present (m/z ) 285, 243, and
243, respectively), corresponding to the mass of a cleavage
product in which the sulfur atom is further oxidized to sulfate
(R-SO3H). In the case of 7 a strong peak at m/z ) 242 was also
observed, for which we have no reasonable explanation.
Using the positive ion detection mode, very complex spectra
were obtained (not shown), probably because of the tendency
of sulfur-containing molecules to give unstable radical cations
FIGURE 4. Negative ion mode ESI-MS spectra of the products formed
by reacting AIBN (5 × 10-2 M) in ACN under air at 60°, with the
substituted phenols 1 (A, sample reacted for 15 min) and 7 (B, sample
reacted for 30 min).
under electrospray conditions.28 Nevertheless, weak peaks
derived from the addition of peroxyl radicals from AIBN
((CH3)2CNCOO•; MW ) 100) on the phenoxyl radicals obtained
by hydrogen abstraction from the parent phenols were observed
as 23Na complexes, together with signals at +16 m/z and at +32
m/z.
Since sulfur-containing molecules are known to decompose
hydroperoxides via a nonradical mechanism,20 the peaks with
mass +16 and +32, observed using both the negative and the
positive ion mode, might derive from the reaction between
antioxidant and the hydroperoxides formed during the autoxi-
dation. To check this possibility, phenol 7 (1 × 10-3 M) was
incubated for 1 h at 60 °C in ACN with an amount of cumene
hydroperoxide similar to that produced during the first period
of the autoxidation reaction (1 × 10-3 M). Under these
conditions, no peaks derived from the oxidation of 7 to sulfoxide
or sulfone were observed, in line with the low rate constants
reported for such reactions.20
Discussion
Electronic Effect. The relative efficiency as oxidation
inhibitors of the investigated sulfides, sulfoxides, and sulfones
1-6 is in agreement with the well-known effect that para
substituents have on the BDE values of phenols, i.e., strong
O-H bonds are brought about by electron acceptors (SOR and
SO2R) and weak bonds by electron donors (SR).14 The additive
contribution of the p-thiomethyl group, obtained by subtracting
to the measured BDE(O-H) values of 1 the BDE(O-H) of
phenol (86.7 kcal/mol)15 and the known contribution of two
o-tert-butyl groups (-4.8 kcal/mol),14 is -3.6 kcal/mol, a value
between methyl (-1.7 kcal/mol) and methoxyl (-4.4 kcal/mol)
groups.14 This result shows that the sulfur atom induces a smaller
(O-H) bond strength than oxygen, similarly to what found in
phenothiazines for the N-H bond strength.5 It is somewhat
surprising that the BDE values of sulfones 3 (81.7 kcal/mol)
and 6 (82.3 kcal/mol) are very close to that of 2,6-di-tert-
butylphenol (81.9 kcal/mol),13 this implying that a p-SO2R
group, which is considered a strong electron-withdrawing
substituent, behaves similarly to a para hydrogen atom. This
does not seem to be due to some experimental artifact since
(27) Whitehouse, C. M.; Dreyer, R. N.; Yamashita, M.; Fenn, J. B. Anal.
Chem. 1985, 57, 675.
(28) Van Berkel, G. J.; McLuckey, S. A.; Glish, G. L. Anal. Chem. 1992,
64, 4, 1586-1593.
J. Org. Chem, Vol. 71, No. 17, 2006 6329

Amorati et al.
TABLE 3. Additive Group Effect on the Phenolic BDE(OH) of
Thiomethyl and Methoxyl Substituents
∆BDE(O-H) (kcal/mol)
SCH3
OCH3
ortho
para
-0.85
-3.6
-0.2
-4.4
group (-3.6 kcal/mol) because of the formation of an intramo-
lecular hydrogen bond in the parent phenol. Actually, the IR
spectrum of Figure 3 shows a single peak (at 3375 cm-1) in
the region above 3000 cm-1, indicating that the structure adopted
in solution at room temperature is that one where the OH is
pointing toward the SCH3 group.
It may be interesting to compare these numbers with the
corresponding ones reported for the methoxyl group (see Table
3) in ortho and para positions, i.e., -0.2 and -4.4.14 This
suggests that the intramolecular hydrogen bond of the phenolic
OH proton to the adjacent SMe group is weaker than that to an
OMe group.
FIGURE 5. Bond dissociation enthalpies (BDE) of 4-substituted 2,6-
di-tert-butylphenols against the logarithm of the rate constant for their
reaction with peroxyl radicals. For experimental data see the present
paper and ref 31.
even the inhibition rate constants kinh for 3 and 6 (2.2 and 1.3
× 103 M-1 s-1, respectively) are comparable to that of 2,6-di-
tert-butylphenol (3.1 × 103 M-1 s-1).18 This point deserves more
attention and seems worth to be further investigated by a
computational approach.
Under the present experimental conditions, no EPR spectra
of the phenoxyl radicals from the two sulfoxides 2 and 5 were
observed, and thus the corresponding O-H bond dissociation
enthalpies could not be measured. It was, however, possible to
estimate these values from the inhibition rate constant kinh. In
fact, it is known that BDEs vary linearly with the logarithm of
kinh for phenols having the same ortho substituents. Figure 5
shows the plot obtained by using all the reported BDE and kinh
literature data for 4-substituted 2,6-di-tert-butylphenols. From
this plot the BDE value of the sulfinyl derivative 2 can be
estimated as 81.6 kcal/mol.
The O-H BDEs of the sulfides 1 and 4, when compared to
that reported previously for probucol (2,2-di(4-hydroxy-3,5-di-
tert-butylphenylthio)propane) 81.1 kcal/mol,29,15 are lower by
ca. 1.8 kcal/mol despite the structural similarity of these
compounds. It should, however, be pointed out that the alkylthio
substituent para to the OH group of probucol, i.e., SC(Me)2SAr,
is much bulkier than those of both 1 and 4. Thus, it is
conceivable that the geometry adopted by the alkylthio group
in probucol and in the corresponding phenoxyl radical is out of
the aromatic plane, whereas in 1 (and presumably in 4) the SR
group lies on the molecular plane.12 Since in organic free radicals
the electronic character of alkylthio substituents conjugated with
the radical center changes with conformation,30 it seems
reasonable to explain the larger BDE value determined in
probucol, with respect to 1 and 4, in terms of the different
geometry adopted by these phenols and the corresponding
phenoxyl radicals.
A point worth of discussion is also the large difference
between the values of kinh and BDE of sulfides 8 and 9, both
having a thio substituent ortho to the hydroxyl group. Actually,
the bisphenol 9 is 54 times more reactive than 8, instead of
being only twice more reactive as expected from statistics. This
difference can be rationalized on the basis of the structure of
the two antioxidants as discussed above. In the intramolecularly
hydrogen bonded o-methylthio phenol 8, the bridge between
the hydroxyl hydrogen and the adjacent sulfur atom (Scheme
2) increases the O-H bond dissociation enthalpy by stabilizing
the starting phenol and makes more difficult for steric reasons
the approach of peroxyl radicals to the OH group. The result is
a low reactivity toward attacking radicals for both thermody-
namic and kinetic reasons. In the bisphenol 9 instead, the OH
group intramolecularly hydrogen bonded to sulfur should be
scarcely reactive as in 8, whereas the second one, giving a
weaker H-bond (presumably to the π-electrons of the other
benzene ring) is less embedded in the bulk of the molecule and
therefore more easily accessible to peroxyl radical. Thus, the
hydrogen atom of the second OH group can be abstracted more
easily because it is less sterically hindered and is characterized
by a lower bond strength (the BDE value is smaller by 1.2 kcal/
mol than that of 8).
Mechanism of Inhibition. In the case of sulfide 1, the shape
of the oxygen uptake trace and the measured induction period
seems to indicate that it behaves as a classical chain-breaking
phenol antioxidant capable of trapping ca. two peroxyl radicals
for each antioxidant molecule. This implies that the phenoxyl
radical formed in the first stage of the inhibition reaction reacts
with peroxyl radicals to give a quinolide peroxide similarly to
alkyl-substituted phenols such as BHT.1,20 From the ESI-MS
spectra of the reaction mixture of 1 with AIBN this adduct can
be identified as the Na+ complex by positive ion detection mode
at m/z ) 374. In the negative ion detection mode (see Figure
4A), besides the starting sulfides, peaks due to the corresponding
sulfoxide and sulfone were also observed. Since no oxidation
products could be detected by ESI-MS when reacting 1 with
cumene hydroperoxide, this means that the formation of
sulfoxides and sulfones is due to the reaction of the sulfide with
peroxyl radicals and not with peroxides.
Hydrogen Bond Effects. The additive contribution of an
o-SCH3 group to the phenolic BDE(O-H) can be determined
as -0.85 kcal/mol, by subtracting the contribution of phenol
and of two tert-butyl groups (o-CMe3 ) -1.75; p-CMe3 ) -1.9
kcal/mol)14 to the experimental BDE(O-H) value of 8. This
value is significantly smaller than the contribution of a p-SCH3
(29) Lucarini, M.; Pedulli, G. F.; Cipollone, M. J. Org. Chem. 1994,
59, 5063-5070.
(30) Alberti, A.; Guerra, M.; Martelli, G.; Bernardi, F.; Mangini, A.;
Pedulli, G. F. J. Am. Chem. Soc. 1979, 101, 4627-4631.
In the case of bisphenol 4, the length of the induction period
is one-half of the expected value since only two peroxyl radicals
per molecule are trapped, despite the presence of two phenolic
6330 J. Org. Chem., Vol. 71, No. 17, 2006

Sulfur-Containing Phenolic Antioxidants
SCHEME 3
radicals. The enhanced activity as chain-breaking antioxidants
of phenols para-substituted with a SR group is, however,
substantially smaller than that of the corresponding p-methoxy
phenols. This result, together with recent findings36 on seleno-
tocopherol indicating that the replacement of the chromanolic
oxygen with selenium does not improve neither the BDE(O-
H) nor the kinh value with respect to R-tocopherol, suggest that
the chain-breaking antioxidant efficacy of phenols para-
substituted with XR groups decreases in the order X ) O > S
> Se. However, it should be emphasized that this holds for
initiated oxidations at low temperature, whereas spontaneous
oxidations proceeding more slowly or at higher temperatures
might be inhibited efficiently also by sulfur- or selenium-
containing phenols as a result of their ability to behave as
preventive antioxidants by decomposing hydroperoxides to
alcohols.
moieties. The same result was previously found by other authors
who performed similar measurements in cumene at different
temperatures.6 The presence in the ESI-MS spectra of the
products of reaction between 4 and AIBN of an intense peak at
m/z ) 285 (not shown) corresponding to a cleavage product in
which sulfur is oxidized to sulfate seems to suggest that, after
trapping two peroxyl radicals, the bisphenol 4 splits in two
fragments not capable of further inhibiting the autoxidation
reaction.
It is possible that the complex behavior of the investigated
sulfides may be due to the instability of the quinolide peroxide
formed by addition of a second peroxyl radical to the carbon
linked to the sulfur substituent (10, Scheme 3). This adduct,
characterized by the presence of two reactive moieties (-OO-
and -S-) close together, might be cleaved thermally giving
sulfoxides and sulfones and other products devoid of any
antioxidant activity.34
It should be emphasized that the n ) 3 value observed with
bisphenol 7 might be consistent with the above suggestion.
Actually, due to the reduced size of the methyl group in one of
the ortho positions, attack of the second peroxyl radical to the
primarily formed phenoxyl radical from 7 may occur not only
at C-4 but also at C-6, thus giving the analogue of 10 as well
as 11, where sulfur and peroxide groups are isolated. In fact,
11 could be identified by ESI-MS at 456 m/z by negative ion
detection mode.
The formation of sulfoxides and sulfones might arise, beside
from decomposition of quinolide peroxides such as 10,34 also
from direct reaction of peroxyl radicals with the starting sulfides,
as has been reported previously by different authors.35 Thus,
competition for peroxyl radicals between the phenolic OH and
the sulfur atom can give rise to stoichiometric factors shorter
than expected, as, for instance, the n ) 1.7 value found for
phenol 1 and n < 2 observed by Ingold and co-workers3 for
sulfur analogues of vitamin E. However, this competition is not
believed to be responsible for the abnormally small n coefficients
of bisphenols 4, 7, and 9, which are very similar despite the
large difference between the kinh values.
By investigating phenols 8 and 9, we have also been able to
measure the contribution of an ortho SR group to the antioxidant
activity of phenols and to evaluate the importance of the
intramolecular O-H‚‚‚ SR hydrogen bonding in determining
the properties of such molecules. The present data imply that
the intramolecular hydrogen bond of the phenolic OH proton
to the adjacent SMe group is weaker than that to an OMe group,
in contrast with DFT calculations carried out on a number of
ortho-substituted phenols.37
At variance with what is observed with SR groups, substitu-
tion of the para position with oxygenated sulfur groups does
not lead to any improvement (SO2R substituents) of the
antioxidant activity of phenols or only to a very small improve-
ment (SOR groups).
Some important points concerning the radical chemistry of
sulfur-containing phenols remain to be better clarified, in
particular the fate of oxidized products deriving from the
reaction with peroxyl radicals. It is worth to be noted that
product studies are only seldom reported in the characterization
of new antioxidants, although this piece of information may be
crucial for practical purposes.
Experimental Section
Matherials. Sulfide 1 was prepared by methylation 2,6-di-tert-
butyl-4-mercaptophenol with methyl iodide.38 The mercaptophenol
was obtained by LiAlH4 reduction of the corresponding N-
thiophthalimide.39
Sulfide 4 was prepared reacting 2,6-di-tert-butylphenol with
sulfur and potassium hydroxide in refluxing ethanol as previously
reported.38 Sulfoxides 2 (racemic mixture) and 5 and sulfones 3
and 6 were prepared from the corresponding sulfides by oxidation
with m-CPBA in dichloromethane.
2,4-Di-tert-butyl-6-methylthiophenol (8) was prepared by reacting
2,4-di-tert-butylphenol with sulfur monochloride to obtain the
isolable intermediate 2,2′-trithiobis(2,6-di-tert-butylphenol), which
was reduced with zinc in acidic conditions to give 3,5-di-tert-butyl-
2-hydroxybenzenethiol.40 This compound was treated with io-
Conclusions
The kinetic and thermochemical data collected in this work
demonstrate that the effect of p-thiyl substituents (SR) on the
antioxidant properties of phenol derivatives is that of decreasing
the bond dissociation enthalpy of the phenolic O-H bond and
of increasing the rate constant for the reaction with peroxyl
(31) BDE and kinh values have been taken from refs 13 and 18 for X )
H, Me, and OMe and from ref 33 for X ) CHdCHC6H3(OMe)2. The
equation of the regression line is BDE ) 91.34-2.80 Log kinh
.
(36) Shanks, D.; Amorati, R.; Fumo, M. G.; Pedulli, G. F.; Valgimigli,
L.; Engman, L. J. Org. Chem. 2006, 71, 1033-1038.
(37) Himo, F.; Eriksson, L. A.; Blomberg, M. R. A.; Siegbahn, P. E. M.
Int. J. Quantum Chem. 2000, 76, 714-723.
(38) Neuworth, M. B.; Laufer, R. J.; Barnhart, J. W.; Sefranka, J. A.;
McIntosh, D. D. J. Med. Chem. 1970, 13, 722-725.
(39) Cantini, B.; Capozzi, G.; Menichetti, S.; Nativi, C. Synthesis 1999,
1046-1050.
(40) Pastor, S. D.; Denny, D. Z. J. Heterocycl. Chem. 1988, 25, 681-
683.
(32) Amorati, R.; Ferroni, F.; Pedulli, G. F.; Valgimigli, L. J. Org. Chem.
2003, 68, 9654-9658.
(33) Amorati, R.; Lucarini, M.; Mugnaini, V.; Pedulli, G. F.; Valgimigli,
L.; Roberti, M.; Pizzirani, D. J. Org. Chem. 2004, 69, 7101-7107.
(34) Koenig. In Kochi, J. K., Ed. Free Radicals; Wiley: New York,
1973; Vol. I, pp 113-155
(35) (a) Fukuzumi, S.; Shimoosako, K.; Suenobu, T.; Watanabe, Y. J.
Am. Chem. Soc. 2003, 125, 9074-9082. (b) Schoneich, C.; Aced, A.;
Asmus, K. D. J. Am. Chem. Soc. 1991, 113, 375-376.
J. Org. Chem, Vol. 71, No. 17, 2006 6331

Amorati et al.
domethane in 2-propanol/NaOH, and after chromatography on silica
temperature in the cavity of an EPR spectrometer and photolyzed
with the unfiltered light from a 500 W high-pressure mercury lamp.
The temperature was controlled with a standard variable temperature
accessory and was monitored before and after each run with a
copper-constantan thermocouple.
The EPR spectra were recorded on a spectrometer equipped with
a microwave frequency counter for the determination of the
g-factors, which were corrected with respect to that of perylene
radical cation in concentrated H2SO4 (g ) 2.00258).
When using mixtures of BHT and one of the investigated
phenols, the molar ratio of the two equilibrating radicals was
obtained from the EPR spectra and used to determine the equilib-
rium constant, Ke. Different concentration ratios of starting phenols
were used in order to check if the equilibrium is reached. Spectra
were recorded a few seconds after starting to irradiate in order to
avoid significant consumption of the phenols during the course of
the experiment.
Relative radical concentrations were determined by comparison
of the digitized experimental spectra with computer simulated ones
as previously described.
FT-IR Measurements. The FT-IR spectra were measured from
4000 to 600 cm-1 using a spectrometer having a resolution of 1.0
cm-1. Tetrachloromethane solutions (0.05-0.01 M) of phenols 8
and 9 were examined in a sealed KBr cell with 0.5 mm optical
path.
MS Analysis. An acetonitrile solution of sulfides 1, 4, 7, 8, and
9 (1 × 10-3 M) was stirred under air at 60 °C in the presence of
AIBN (5 × 10-2 M). Aliquots of the reaction mixtures were diluted
1:10 with MeOH, cooled to 0 °C, and analyzed by mass spectrom-
etry using electrospray ionization (ESI) by direct liquid injection
at a flow rate of 10-20 µL/min. The spectrometer employed was
equipped with a single quadrupole analyzer. The most appropriate
instrumental settings were as follows: ESI type, positive and
negative ions; desolvatation gas (N2), 242 L/h; cone gas (skimmer),
33 L/h; desolvatation temperature, 120 °C, source block temper-
ature, 80 °C; capillary voltage, 3.00 kV; cone voltage, 22 V;
hexapole extractor, 3 V.
gel (petroleum ether/ethyl acetate 8/2) 8 was obtained.41
All other compound used in the present investigation were
commercially available. Solvents of the highest purity grade were
used as received. Styrene was percolated on alumina before each
experiment to remove traces of inhibitor.
Kinetic Measurements. The rate constants for the reaction of
the title compounds with peroxyl radicals were measured by
following the autoxidation of either cumene or styrene at 30 °C
using as initiator AMVN (5 × 10-3 M). The reaction was been
performed in a oxygen uptake apparatus built in our laboratory and
based on a differential pressure transducer. The entire apparatus
was immersed in a thermostated bath that ensured a constant
temperature within (0.1 °C. In a typical experiment, an air-saturated
cumene (7.1 M) or styrene (4.2 M) solution in chlorobenzene
containing the antioxidant was equilibrated with the reference
solution containing an excess of R-tocopherol (1 × 10-3 to 1 ×
10-2 M) in the same solvent at 30 °C. After equilibration, a
concentrated chlorobenzene solution of AMVN was injected in both
the reference and sample flasks, and the oxygen consumption in
the sample was measured, after calibration of the apparatus, from
the differential pressure recorded with time between the two
channels. This instrumental setting allowed us to have the N2
production and the oxygen consumption derived from the azo-
initiator decomposition already subtracted from the measured
reaction rates. Induction period lengths (τ) were determined (eq
14) using an integration procedure suggested by Roginsky et al.20
τ ) 0 {1 - (R/R0)2} dt
(14)
∫
∞
Initiation rates, Ri, were determined for each condition in prelimi-
nary experiments by the inhibitor method18 using R-tocopherol as
reference antioxidant: Ri ) 2[R - TOH]/τ. Because reaching the
steady state conditions, after injecting the azo-initiator in the reaction
vessel, requires some time during which the reaction is slightly
slowed, the zero time of the reaction was extrapolated using the
trace obtained in the absence of antioxidant.
Acknowledgment. Financial support from MIUR (Research
project “Radical Processes in Chemistry and Biology: Synthesis,
Mechanism, Application”, contract 2004038243) is gratefully
acknowledged by G.F.P.
EPR and Thermochemical Measurements. Deoxygenated
benzene solutions containing the phenols (0.01-0.001 M) and di-
tert-butyl peroxide (10% v/v) were sealed under nitrogen in a
suprasil quartz EPR tube. The sample was inserted at room
Supporting Information Available: Detailed kinetic results for
the determination of kinh and n values. This material is available
free of charge via the Internet at http://pubs.acs.org.
(41) Kuliev, A. M.; Farzaliev, V. M.; Allakhverdiev, M. A.; Mamedov,
C. I. J. Gen. Chem. USSR 1982, 52, 1889-1893.
JO060281E
6332 J. Org. Chem., Vol. 71, No. 17, 2006