Optimization of the antioxidant activity of hydroxy-substituted 4-thiaflavanes: A proof-of-concept study

Article abstract of DOI:10.1002/chem.201101146

The design and the synthesis of a new family of hydroxy-4-thiaflavanes, in which the reactive phenolic OH is ortho to the sulfur atom of the benzofused oxathiin ring, allowed to prepare antioxidants that show rate constants for the reaction with peroxyl radicals (k_inh), and bond dissociation energies (BDE), of the ArO-H group identical to those of α-tocopherol, the main component of vitamin E and the most effective lipophilic antioxidant known in nature. The peculiar conformation of the six-membered heterocyclic ring prevents the formation of an intramolecular hydrogen bond between the OH group and the S atom, while ensuring a good stabilization by electron donation of the phenoxyl radical formed after the reaction with peroxyl radicals. The preparation of these compounds was achieved through an inverse electron demand hetero Diels-Alder reaction of styrenes with o-thioquinones, in turn prepared from accurately designed 1,3-dihydroxy arenes. Properly arranging the substitution pattern on the aromatic ring, as in derivatives 9 and 11, allowed to reach values of k_inh up to 4.0× 10⁶ M^-1 s ^-1 and BDE_(OH) of 77.2 kcal mol^-1. This approach represents an innovative way to obtain highly active antioxidants without using strongly electron donating alkylamino groups which are associated with adverse toxicological profiles. Sulfur makes the difference (again): The stereoelectronic features triggered by the sulfur atom in the benzo-fused six-membered ring of 4-thiaflavanes have been exploited to design a new family of phenolic antioxidants that show an ability to react with peroxyl radicals identical to that of α-tocopherol, the main component of vitamin E and the most potent natural lipophilic antioxidant (see scheme). Copyright

Full text of DOI:10.1002/chem.201101146

DOI: 10.1002/chem.201101146
Optimization of the Antioxidant Activity of Hydroxy-Substituted
4-Thiaflavanes: A Proof-of-Concept Study
Caterina Viglianisi,[a] Maria Grazia Bartolozzi,[a] Gian Franco Pedulli,[b]
Riccardo Amorati,*[b] and Stefano Menichetti*[a]
Abstract: The design and the synthesis
clic ring prevents the formation of an
intramolecular hydrogen bond between
the OH group and the S atom, while
ensuring a good stabilization by elec-
tron donation of the phenoxyl radical
formed after the reaction with peroxyl
radicals. The preparation of these com-
pounds was achieved through an in-
verse electron demand hetero Diels–
Alder reaction of styrenes with o-thio-
of a new family of hydroxy-4-thiafla-
vanes, in which the reactive phenolic
OH is ortho to the sulfur atom of the
benzofused oxathiin ring, allowed to
prepare antioxidants that show rate
constants for the reaction with peroxyl
radicals (kinh), and bond dissociation
quinones, in turn prepared from accu-
rately designed 1,3-dihydroxy arenes.
Properly arranging the substitution pat-
tern on the aromatic ring, as in deriva-
tives 9 and 11, allowed to reach values
of kinh up to 4.0ꢀ106 mꢀ1 sꢀ1 and
BDE(OH) of 77.2 kcalmolꢀ1. This ap-
proach represents an innovative way to
obtain highly active antioxidants with-
out using strongly electron donating al-
kylamino groups which are associated
with adverse toxicological profiles.
ꢀ
energies (BDE), of the ArO H group
identical to those of a-tocopherol, the
main component of vitamin E and the
most effective lipophilic antioxidant
known in nature. The peculiar confor-
mation of the six-membered heterocy-
Keywords: 4-thiaflavane · antioxi-
dants · cycloaddition · sulfur hetero-
cycles · tocopherols
Introduction
action between antioxidant and peroxyl radicals kinh
[Eq. (1)].
The autoxidation of unsaturated hydrocarbons and lipids is
one of the most studied reactions, both for technological
reasons and for its relevance in biological environments, be-
cause this process accounts for the damage caused by free
radicals to organic and bioorganic systems. This reaction can
be inhibited, or at least retarded, by chain-breaking antioxi-
dants capable of trapping free radicals without transforming
themselves in reactive intermediates. Phenols, which repre-
sent the main family of antioxidants, can donate the phenol-
ic hydrogen atom to the chain-carrying peroxyl radicals to
form a phenoxyl radical stabilized by resonance, hence they
are relatively unreactive toward oxygen and organic materi-
als.[1] The extent to which the autoxidation is retarded by
phenols depends upon the rate constant of the inhibition re-
kinh
C
C
ROO þ ArOH
ROOH þ ArO
ð1Þ
ƒ!
Over recent decades, much work has been done to clarify
the basic principles that determine the rates of Reaction (1)
and to synthesize compounds that show kinh values similar
(or higher) to those of a-tocopherol (a-TOH),[2] which is the
most effective lipid-soluble antioxidant present in nature.[2,3]
Actually, some of the synthetic phenols developed show ex-
cellent features, such as very high reactivity towards peroxyl
radicals and improved air stability.[4]
Another active area of research is the study of preventive
antioxidants,[5] compounds often containing divalent S, Se,
or Te atoms in the molecular skeleton,[6] which inhibit the
formation of initiating free radicals by decomposing hydro-
peroxides by a nonradical process. Chalcogen-substituted
phenols have been found to possess both chain-breaking
and preventive antioxidant activities.[7] Recently, an original
hetero Diels–Alder approach, consisting of the reaction of
electron-poor o-thioquinones with suitable alkenes
(Scheme 1),[8] has been used to prepare hydroxylated sulfur
heterocycles with a structure resembling natural polyphenols
present in plants (compounds 1–4).[9–11]
[a] Dr. C. Viglianisi, Dr. M. G. Bartolozzi, Prof. S. Menichetti
Dipartimento di Chimica “Ugo Schiff”
Polo Scientifico e Tecnologico Universitꢁ di Firenze
Via della Lastruccia 3–13, 50019 Sesto Fiorentino (Italy)
Fax : (+39)055-4573531
E-mail: stefano.menichetti@unifi.it
[b] Prof. G. F. Pedulli, Dr. R. Amorati
Dipartimento di Chimica Organica “A. Mangini”
Universitꢁ di Bologna
The main appeal of this new family of phenolic com-
pounds is that both the diene and the dienophile can be
easily modified to obtain a final benzoxathiine cycloadduct
with the required structural characteristics. In previous stud-
ies, several sulfur-containing analogues of flavonoids, like
Via S. Giacomo 11, 40126 Bologna (Italy)
Fax : (+39)051-2095688
E-mail: riccardo.amorati@unibo.it
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101146.
12396
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12396 – 12404

FULL PAPER
Scheme 2. 4-Thiaflavane antioxidants 6–11 prepared in this study.
was introduced to ensure the multidefence aptitude of these
compounds (Scheme 2).
To prepare derivatives 6–11 we adapted our general pro-
cedure based on an inverse electron demand hetero Diels–
Alder reaction of o-thioquinones with styrenes using as
starting phenols proper substituted resorcines. Compound 6
was obtained starting from 2,4-dihydroxymethyl benzoate
which was t-butylated and sulfenylated with phthalimidesul-
fenyl chloride 12 (PhtNSCl) without any protection at the
hydroxy groups but using a stoichiometry amount of 2,6-di-
t-butylpyridine to intercept the HCl formed during the reac-
tion which, otherwise, causes a dramatic decrease of the
final yield (Scheme 3). Generation of the thioquinone in the
Scheme 1. Hetero Diels–Alder approach used and selected lead 4-thiafla-
vane antioxidants 1–5 prepared prior to this study. EDG=electron-do-
nating group, Pht=phthaloyl.
1,[9] or tocopherols, like 2,[10] or compounds possessing a
structure related to both these families of powerful antioxi-
dant polyphenols, like 3 and 4,[11] were synthesized and
tested for their antiradical activity in model and/or in bio-
logical systems. Moreover, 4-thiaflavanes bearing an
OH group ortho to the heterocyclic sulfur, like 5, have been
used as suitable models to rationalize the mechanism of
action of galactose oxidase (GOase), a copper metalloen-
zyme that catalyses the two-electrons oxidation of primary
alcohols to aldehydes.[12] Herein, we report how the data col-
lected studying the relationship between stereoelectronic
features and chain-breaking activity of these 4-thiaflavanes
have been exploited to design and prepare a new class of ef-
ficient and air stable phenolic antioxidants.
Scheme 3. Reagents and conditions: a) tBuOH, H2SO4, 608C, 94 h, 91%.
b) PhtNSCl (12), 2,6-di-t-butylpyridine, CHCl3, 08C–RT, 45 h, 80%. c) p-
Methoxystyrene (13), Et3N, CHCl3, 608C, 4 h, 68%. d) LiAlH4, THF,
ꢀ108C, 30 min, 72%.
presence of p-methoxystyrene (13) gave benzoxatiine 6 iso-
lated as single regioisomer. Hence, only one of the free OH
groups participate to the formation of the dienic thioqui-
none, and we deduced that the intramolecular hydrogen
bond (IHB) between the ester group and the adjacent
OH group prevents its involvement into the cycloadditon.
Reduction of the ester moiety with LiAlH4 in dry THF, al-
lowed the isolation of hydroxymethyl derivative 7 as report-
ed in Scheme 3.
Results
Syntheses: The antioxidant activity of the several 4-thiafla-
vanes previously investigated by our research group suggest-
ed to focus our attention toward derivatives 6–11 that bear
as common feature an OH group on C5. Additional substi-
tutions on the A ring were chosen to further modulate their
antioxidant ability, while the catechol moiety on the B ring
Chem. Eur. J. 2011, 17, 12396 – 12404
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12397

S. Menichetti, R. Amorati et al.
suggested it was worthless to consume time in optimization
of the synthetic procedure required for the synthesis of 10.
On the contrary, the good antioxidant performances
showed by compound 9 prompted us to design a new syn-
thetic path which allowed its preparation, as well as that of
related 4-thiaflavanes possessing the same substitution pat-
tern on the A ring, as a single regioisomer in acceptable
overall yield. Thus, 2-methyl-4-methoxy phenol, obtained by
Wolf–Kishner reduction of the corresponding commercial
available benzaldehyde, was acetylated at the OH group
with Ac2O in TEA followed by ring acylation with AcCl/
AlCl3. A quantitative acetyl/benzyl protecting group switch-
ing, followed by Baeyer–Villiger oxidation and saponifica-
tion, allowed the preparation of the required phenol which
was sulfenylated with 12, and a stoichiometry amount of 2,6-
di-t-butylpyridine to obtain the precursor of the correspond-
ing o-thioquinone. Cycloaddition with 13 or with t-butyldi-
methylsilylether of 3,4-dihydroxystryrene (14) led to the for-
mation of the required cycloadducts which were debenzylat-
ed[15] (and desilylated) to isolate 4-thiaflavanes 9 and 11 as
reported in Scheme 6.
Scheme 4. Reagents and conditions: a) AlCl3, CH2Cl2, RT, 24 h, 71%.
b) NaCNBH3, THF, RT, 4 h, 76%. c) 12, CHCl3, 08C–RT, 17 h. d) 13,
CHCl3, 608, 18 h, 38% overall over two steps.
The possibility to run sulfenylation of 1,3-dihydroxy aryl
derivatives with 12 avoiding OH protection, suggested a
short path to thiaflavanes 8 and 9 as described in Scheme 4.
Selective 2,4-demethylation of 2,4,5-trimethoxybenzaldeyde
with AlCl3,[13] followed by NaCNBH3 reduction of the al-
[14]
dehyde to methyl group, allowed the isolation of 4-methoxy-
6-methyl resorcine which was sulfenylated with 12. All at-
tempts to isolate the corresponding N-thiophthalimide were
unsuccessful since it demonstrated a poor stability on stand-
ing and completely decomposed on silica gel, thus it was re-
acted, as crude material, with 13 in the presence of Et3N. In
this case both OH groups participate to the o-thioquinone
formation giving rise to a pair of dienes, and eventually to a
mixture of two regioisomeric cycloadducts 8 and 9 which
were obtained in moderate yields in roughly 1:1 ratio, and
separated by an accurate flash chromatography (Scheme 4).
The preparation of 6,7-dimethyl derivative 10 was ach-
ieved starting from 2,3-dimethyl hydroquinone which was
acetylated and then orthogonally protected at the
OH groups by a selective silylation of the phenolic OH not
engaged in an IHB with the acetyl group, followed by an ex-
haustive methylation. A Baeyer–Villiger oxidation allowed
the introduction of the third aryl–oxygen bond followed by,
in sequence, saponification, sulfenylation, and cycloaddition
(see the Supporting Information for details). Desilylation of
t-butyldimethylsilyl ether at position 5 allowed the isolation
of 4-thiaflavane 10 as described in Scheme 5. The overall
yield was modest, however, the results obtained by measur-
ing the antioxidant activity of derivatives 6–11, vide infra,
Scheme 6. Reagents and conditions: a) N2H4, KOH, glycol, RT, 1508C,
19 h,>98%. b) Ac2O, TEA, CH2Cl2, RT, 16 h,>98%. c) AcCl, AlCl3,
CH2Cl2, 08C–RT, 3 h, 64%. d) NaOH, H2O/THF, RT, 16 h,>98%.
e) BnBr, NaH, THF, 08C–RT, 24 h,>98%. f) m-CPBA, p-TsOH, CH2Cl2,
RT, 24 h. g) NaOH, H2O/CH3OH, RT, 24 h, 80% over two steps. h) 12,
2,6-di-t-butylpyridine, CHCl3, 08C–RT, 24 h, 75%. i) 13, CHCl3, 608C,
72 h, 74%. l) 14, CHCl3, 608C, 72 h, 83%. m) Pd, NH4COOH, EtOH/
.
AcOEt/H2O (7:2:1), 908C, 4 h, >98%. n) TBAF3H2O, AcOH, THF,
08C–RT, 1 h, 80%.
Antioxidant activity: The antioxidant activity of 4-thiafla-
vanes was evaluated by measuring the rate constant (kinh
)
C
for their reaction with peroxyl radicals (ROO ), that are re-
sponsible for the propagation step of peroxidation processes
in many natural and man-made materials. The values of kinh
were determined by studying the inhibition of the thermally
initiated autoxidation of either styrene or cumene [Eqs. (2)–
(7)] under controlled conditions, using chlorobenzene as a
solvent.[2,3,16] Being less oxidizable than styrene, cumene was
used to study weak antioxidants (see Table 1). In the case of
styrene, the propagation step [Eq. (4)] consists of the addi-
tion of peroxyl radicals to the olefinic double bond of sty-
rene.[2]
.
Scheme 5. Reagents and conditions: a) AcOH, BF3 2H2O, 1208C, 20 h,
65%. b) tBuMe2SiCl, IMI, DMF, RT, 18 h,>98%. c) CH3I, K2CO3, DMF,
1008C, 15 h, 75%. d) m-CPBA, p-TsOH, CH2Cl2, RT, 24 h, 68%.
e) KOH, DMF/CH3OH, RT, 18 h. f) 12, CHCl3, 08C–RT, 5 h, 45% over
.
two steps. g) 13, Et3N, CHCl3, 608C, 18 h. h) TBAF3H2O, THF, 08C–RT,
3 h, 67% over two steps.
12398
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12396 – 12404

Antioxidant Activity of Hydroxy Substituted 4-Thiaflavanes
FULL PAPER
Table 1. Kinetic rate constants for the reaction of the investigated com-
pounds with peroxyl radicals (kinh), and number of radicals trapped by
each antioxidant molecule (n).
easily obtained. The number of radicals trapped by each an-
tioxidant molecule (n) was obtained from the length of the
inhibition period, by comparison with the reference antioxi-
dant 2,2,5,7,8-pentamethyl-6-chromanol (PMHC), a model
for the physiological antioxidant a-tocopherol, for which n=
2.[2] For inhibitors which act by reactions 6 and 7, a n value
of 2 is expected.[2] The values of kinh and n, calculated from
the slope and the length of the inhibited period are reported
in Table 1.
From Table 1, it can be seen that the 4-thiaflavanes
having in ortho position strong H-bond accepting groups
(i.e., 6, 7, and 8) are weak inhibitors of the autoxidation re-
action.[18] Values of kinh comparable to that of PMHC were
obtained in the case of 9 and 11, while the presence of an
additional methyl group in 10 caused a small reactivity de-
crease. The stoichiometric coefficients smaller than 2 in the
case of 9, 10, and especially 8, indicate that the phenoxyl
[b]
Phenol
Substrate[a]
kinh [mꢀ1 sꢀ1
]
n[b]
[c]
6
7
8
9
C
C
C,S
S
S
S
<102
–
3.5ꢀ104
1.0ꢀ105
3.4ꢀ106
1.2ꢀ106
4.0ꢀ106
3.2ꢀ106
2.0
1.2
1.8
1.8
3.6
2.0
10
11
PMHC[d]
–
[a] C=cumene; S=styrene. [b] Error within 15%. [c] Not measurable.
[d] From Ref. [2].
Ri
C
Initiator
R
ð2Þ
ð3Þ
ð4Þ
ð5Þ
ð6Þ
ð7Þ
ƒ!
C
C
R þO2 ROO
!
kp
C
C
ROO þRH
ROOHþR
ƒ!
C
radicals from the antioxidants (ArO ) have some attitude to
propagate the oxidative chain, presumably by fragmentation
of the methoxyl group to form a quinone and a methyl radi-
cal, which in turn generates another peroxyl radical.[19] As
expected, the presence of two antioxidant moieties in com-
pound 11 (hydroxy 4-thiaflavane and catechol) afforded a
n value slightly smaller than 4 (see also Figure 1).[11]
2kt
C
C
ROO þROO
Nonradical products
ƒ!
kinh
C
C
ROO þArOH
ROOHþArO
ƒ!
C
C
ROO þArO
Nonradical products
!
These reactions were performed at 308C by using azobi-
s(isobutyronitrile) as initiator, and were followed by moni-
toring the oxygen consumption in an oxygen-uptake appara-
tus based on a differential-pressure transducer.[17] In the
presence of good antioxidants, oxidation of the substrate
and oxygen consumption are much slower than in their ab-
sence, and a clear inhibition period is observed (compare
EPR spectra and determination of the BDE(O
values:
ꢀ
H)
When photolyzing, inside the EPR cavity, oxygen-free solu-
tions of 7–10 in benzene containing di-t-butylperoxide (5%
v/v), EPR spectra characterized by g-factors and splitting
constants typical of aryloxyl radicals were observed (see
Table 2). In the case of 7, the coupling with one hydrogen of
traces c–e with trace a in Figure 1). On the other
Table 2. EPR spectral parameters of the aryloxyl radicals from 7–10.
hand, weak antioxidants only afford a retardation
of the oxygen consumption (trace b). As the rate of
O2 consumption during the inhibition period is in-
versely dependent on kinh and the concentration of
the antioxidant (see Experimental Section for ki-
netic details),[2,3] the rate constant for the reaction
Phenol
Hyperfine splittings (gauss)
g value
7
3.00 (1H); 1.12 (1H); 1.07 (1H); 0.99 (1H)
2.0048
2.0048
2.0048
2.0049
8
9
10
10.06 (3H); 1.50 (3H); 0.53 (1H); 1.70 (1H); 1.82 (1H)
6.84 (3H); 1.45 (3H); 1.59 (1H); 0.93 (1H); 0.35 (1H)
5.52 (3H); 1.08 (3H); 0.74 (3H); 2.02 (1H)
C
between ROO radicals and phenols 6–11 could be
the o-CH2OH group and three smaller couplings were de-
tected. The non-equivalence of the CH2 protons suggests
ꢀ
that there is slow rotation around the Ar CH2 bond, likely
because of the existence of an IHB between the alcoholic
OH group and the phenoxyl oxygen. The g values were simi-
lar to that measured previously for the phenoxyl radical
from 5 (g=2.0049).[12]
K8
0
0
ƒ!
C
C
ArOH þ Ar O
ArO þAr OH
ð8Þ
ð9Þ
ƒ
BDEðArOHÞ ¼ BDEðAr0OHÞꢀRTlnK8
ꢀ
The BDE of the O H bond in compounds 7–10 was de-
termined by using the electron paramagnetic resonance radi-
cal equilibration technique (Req-EPR).[16,20] The equilibrium
constant K8 of two equilibrating phenols and the corre-
sponding phenoxyl radicals [Eq. (8)] was measured in ben-
Figure 1. Oxygen consumption traces observed during the autoxidation of
styrene (4.3m) initiated by AIBN (0.05m) at 308C in a) chlorobenzene in
the absence of antioxidants or b) in the presence of 8 (1ꢀ10ꢀ5 m); c) 10
(8.5ꢀ10ꢀ6 m), d) 9 (8.5ꢀ10ꢀ6 m), and e) 11 (8.5ꢀ10ꢀ6 m).
Chem. Eur. J. 2011, 17, 12396 – 12404
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12399

S. Menichetti, R. Amorati et al.
zene by determining the relative concentrations of the phe-
noxyl radicals generated by photolyzing a deoxygenated
mixture of the investigated compound (ArOH) and a refer-
ence phenol (Ar’OH) (Figure 2). The reference was 2,6-di-t-
butyl-4-methoxyphenol (dBHA, BDE=77.2 kcalmolꢀ1)[20,21]
and 2,6-di-t-butyl-4-methylphenol (BHT, BDE=79.9 kcal
molꢀ1).[20,21] As the entropic term in H-atom transfer equili-
bria is negligible,[22] the BDE(OH) was determined by Equa-
tion (9) from the measured K8 values, as reported in Table 3.
Figure 3. Infrared spectra in the OH stretching region of diluted CCl4 sol-
utions of: a) 7, b) 8; c) 9; d) 10. Assignments of IR bands are indicated.
bonded species (syn conformation with respect to the sulfur
atom) and a sharp peak typical of the “free” OH (anti con-
formation), at about 3612 cmꢀ1.
As we pointed out in the case of 5, the inclusion of the
ortho S-alkyl substituent in a six-membered ring dramatical-
ly reduces the ability of the S atom to act as H-bond accept-
or because of the directionality of the sulfur lone pairs.[12]
Actually, a phenol having in ortho position a -SCH3 group
showed only one absorption band at 3375 cmꢀ1, indicating a
strong IHB with the sulfur atom.[23] The IR spectrum of
compound 8 showed a single peak at 3549 cmꢀ1, at a fre-
quency slightly lower than that of 2-methoxyphenol
(3558 cmꢀ1).[18b] Since in this molecule the S atom is a rather
weak H-bond acceptor, it can be supposed that in 8 the
OH group forms a strong IHB with the oxygen atom of the
ortho OMe group. The spectrum of 7 showed a band at
3418 cmꢀ1 due to the phenolic OH donating a IHB to the al-
coholic oxygen atom, and two overlapping bands at 3600–
3615 cmꢀ1 owing to different conformations of the free alco-
holic OH.[24]
Figure 2. a) Experimental and b) computer simulated EPR spectrum ob-
served when photolyzing a mixture of 9 and dBHA in a concentration
ratio 1:1.
Table 3. Req-EPR equilibrium constants and corresponding BDE(OH) for
4-thiaflavanes 7–10 at 298 K.
Phenol
Reference
K8
BDE [kcalmolꢀ1
]
7
8
9
10
BHT
BHT
dBHA
dBHA
–
0.4ꢁ0.1
0.99ꢁ0.08
0.9ꢁ0.1
0.06ꢁ0.01
–
80.4ꢁ0.02
79.9ꢁ0.02
77.2ꢁ0.02
78.8ꢁ0.02
77.2[a]
a-TOH
[a] From Refs. [20] and [21].
Discussion
The low solubility in benzene of compound 11 prevented
the determination of its BDE. However, kinetic studies sug-
gest that the hydroxy-4-thiaflavane moiety should have a
BDE(OH) almost coincident to that of 9, and the catecholic
portion should have a BDE similar to that of 4-methylcate-
chol (ca. 79.2 kcalmolꢀ1).[20–22]
Our ꢃfirst generationꢄ 4-thiaflavanes were built by consider-
ing Flavonoids as natural antioxidant models.[9] The struc-
ture of derivative 1 was inspired to that of catechin, a well-
known natural polyphenolic antioxidant bearing a 3’,4’-dihy-
droxy (i.e., catechol) fragment as B ring. Similarly to cate-
chin, 1 showed good chain-breaking antioxidant activity
(kinh =6ꢀ105 mꢀ1 sꢀ1)[9d] that was almost completely due to
the catechol moiety. It was also demonstrated that 4-thiafla-
vanes behaved as preventive antioxidants, thanks to their ac-
tivity as metal chelators, and as hydroperoxide quenchers,
due to the oxidation of the S-alkyl group to a sulfoxide.
These findings allowed us to consider them distinctive multi-
potent antioxidants.[9–11]
IR spectroscopy: To assess the role of the IHB on the kinet-
ics of the hydrogen atom transfer to peroxyl radicals, the
FT-IR spectra of phenols 7–10 were recorded in diluted
CCl4 solutions, taking care to avoid phenol autoassociation
(Figure 3). In the case of 9 and 10, two different absorptions
were observed in the OH stretching region: a broad band at
about 3540 cmꢀ1 attributed to a weakly intramolecularly H-
12400
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12396 – 12404

Antioxidant Activity of Hydroxy Substituted 4-Thiaflavanes
FULL PAPER
However, since the overall antioxidant activity of 1, in
terms of kinh, was roughly 5 times lower than that of vita-
min E, we designed a ꢃsecond generationꢄ of 4-thiaflavane
antioxidants possessing the A and C rings identical to the
trimethylchromanol ring of a-TOH. By exploiting the flexi-
bility of our hetero Diels–Alder based synthetic procedure,
we could prepare a-4-thiatocopherol 2,[10] with exactly the
same structure of natural vitamin E but with the sulfur atom
on position 4, as well as compounds like 3 and 4 with a mo-
lecular skeleton resembling both flavonoid and tocopherol
structures.[11] When a o-dihydroxy substituted B ring was
present, like in 4, the inhibition time and the number of in-
tercepted peroxyl radicals (n) were doubled.[11] By increas-
ing the similarities of our compounds with a-tocopherol, we
actually increased the antioxidant ability of 4-thiaflavanes
since the kinh values of 2–4 were a third of that of a-TOH,
and the BDE(OH) values were roughly 2 kcalmolꢀ1 higher.[11]
Despite their structural similarity to a-TOH, derivatives
Actually, the decrease in conjugative stabilization of the
phenoxyl radical as a result of the decreased overlap be-
tween the p orbitals of the aromatic ring and the 2pz orbital
of the heterocyclic oxygen atom is responsible for the larger
BDE(OH) value observed in 3 than in PMHC.[11]
The analysis of conformational preferences of 4-thiafla-
vanes suggested further important considerations. In fact,
while C2 is significantly out of the aromatic ring plane, C3 is
almost coplanar and, consequently, the C9-C10-S4-C3 dihe-
dral angle is very small (28 for 3, 88 for 5).[11,12] This is par-
ticularly important in the case of 5OH substituted 4-thiafla-
vanes, such as derivative 5. FT-IR studies demonstrated that
the strength of the intramolecular OH….S bond in 5 was
nearly zero, differently from o-alkylthio phenols in which
strong IHBs are formed if the substituent is free to rotate
about the aryl–sulfur bond (see Figure 5).
C
2–4 were less reactive toward ROO than a-TOH, because
of the substitution of a methylene bridge with a sulfide
bridge. A rational explanation for this result arose from the
X-ray analysis of 4-thiaflavanes, such as 3, when considering
the change in the dihedral C10-C9-O1-C2 angle (q) induced
by replacing the carbon atom with the sulfur atom in C ring
(Figure 4).[11] It is well known that the presence of the con-
Figure 5. IHB and conjugative phenoxyl radical stabilization in free to
rotate vs. conformationally constrained o-sulfanyl phenols.
Figure 4. Experimental and calculated dihedral angles q (C10-C9-O1-C2)
for 4-thiaflavane 3,[11] PMHC[3] and the corresponding phenoxyl radicals.
densed, saturated ring in chromanols is important for their
good antioxidant activity, as it forces the 2pz orbital of the
heterocyclic oxygen atom, which is sp2 hybridized and conju-
gates better with the aromatic ring.[3] The presence of the
sulfur atom in the saturated condensed ring affected the ge-
ometries of both phenol and phenoxyl radical, and the ex-
perimental (X-ray) and the computed [(B3LYP/6-31+G-
This behavior has been described previously by Schaefer
et al.[25] and has been explained by considering the poor ten-
dency of alkylthio substituents to conjugate with the phenol-
ic aromatic ring and their preference to form a stereospecif-
ic IHB using a 3p lone pair on sulfur that uses almost pure
3p orbitals to form its bonds.[26] As a matter of facts, com-
pound 5 showed a smaller BDE(OH) and a larger kinh with re-
spect to the corresponding o-alkylthio derivatives, because
of the absence of an IHB in the parent phenol and the large
conjugative stabilization of the phenoxyl radical by the
S atom (see Figure 5).[12] Such behavior is absolutely pecu-
liar of the sulfur atom since o-alkoxy substituents, despite
their superior electron-donating ability, give an overall
smaller BDE(OH) decreasing contribution, since the phenol
ACHTUNGTRENNUNG(2d,p)//B3LYP/6–31G(d)] q values of 3 were larger by 128
and 58 than those of PMHC[3] (Figure 4). This difference im-
plies that thiaflavanes, such as 3, have, in the preferred ge-
ometry, the C2 atom remarkably out of the aromatic plane,
hence the conjugation between the aromatic ring and the
lone pair of the oxygen atom is noticeably decreased with
respect to a-TOH in both phenol and phenoxyl radical.[11]
Chem. Eur. J. 2011, 17, 12396 – 12404
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12401

S. Menichetti, R. Amorati et al.
stabilization effect caused by the IHB is strong and unaffect-
Once determined the optimized A ring substitution pat-
tern we decided to prepare 6,7-dimethyl-8-methoxy deriva-
tive 10 to verify whether the introduction of an extra methyl
group in position 7 could further increase the antioxidant ac-
tivity. As described in Scheme 5, the preparation of 10 was
designed considering the selective protection of one of the
OH group, hence avoiding the formation of a mixture of re-
gioisomeric thiaflavanes. As reported in Table 1, despite the
extra ED methyl group, kinh and BDE(OH) of compound 10
are little worse than those of compound 9. This has been ra-
tionalized considering that the methyl group in position 7 is
likely to prevent the methoxy group in position 8 from
adopting the conformation required for an optimal stabiliza-
tion of the incipient phenoxyl radical (Figure 6), similarly to
what previously reported for methyl substituted p-methoxy-
phenols.[2,3]
[18]
ꢀ
ed by rotation around the Ar OR bond.
Initially, we utilized such observation to give a contribu-
tion on understanding the mechanism of action of the
enzyme galactose oxidase (GOase),[12] then we realized it
could be exploited to design a ꢃthird generationꢄ of 4-thiafla-
vane antioxidants. In fact, an OH group in position 5, an
alkyl group in position 6 and other appropriate electron-do-
nating groups on the A ring, could be considered as the sub-
stitution pattern of choice for an optimized 4-thiaflavane an-
tioxidant. Thus, we decided to focus our synthetic efforts on
4-thiaflavane derivatives similar to 5 using suitable 1,3-dihy-
droxyarenes (resorcines) as suitable starting phenols.
Since 4,6-disubstituted-1,3-dihydroxy benzenes can be sul-
fenylated with 12 avoiding protection of the phenolic OH,
we initially prepared derivatives 6 as reported in Scheme 3.
It is worth of mention that only the OH group not involved
in an IHB with the ester group participates to the formation
of the intermediate o-thioquinone, thus the cycloaddition
gave a single regioisomeric 4-thiaflavane (6). The antioxi-
dant ability of derivative 6 was, as expected, very low due to
the presence of the electron-withdrawing ester group in po-
sition 6 that, additionally, engages in a strong IHB the phe-
C
nolic OH responsible for the reaction with ROO radicals.
In fact, as reported in Table 1, the corresponding kinh was
under the limit of our measure (<102 mꢀ1 sꢀ1). Reduction of
the ester group, to give hydroxymethyl derivative 7, removes
almost entirely the electron-withdrawing character of the
substituent in position 6 and ensures a weaker IHB. In fact,
alcohol 7 was a much better inhibitor than 6 (BDE(OH)
=
80.4 kcalmolꢀ1; kinh 3.5ꢀ104 mꢀ1 sꢀ1) even if it is far to be
considered a good antioxidant. The step forward was ach-
ieved by preparing 4-methoxy-6-methyl-1,3-dihydroxy ben-
zene which was sulfenylated with 12 and directly used to
generate the corresponding o-thioquinone(s) (Scheme 4). In
this case, both OH groups participate to the 1,4-elimination
at the sulfur atom, giving rise to two dienic o-thioquinones
and, consequently, to a mixture of regioisomeric 4-thiafla-
vanes 8 and 9 formed in roughly 1:1 ratio. Having these two
derivatives in hand we could demonstrate our aforemen-
tioned deductions about the stereoelectronic influence of
the heterocyclic sulfur atom on the reaction of an o-OH
group with peroxyl radicals. In fact, the kinh and BDE of de-
rivative 8 are typical of a weak antioxidant (very similar to
7), since the 6-methoxy group causes the formation of an
IHB with the 5OH, while the methyl group in position 8
gives only a small contribution to phenoxyl radical stabiliza-
tion.[20] On the contrary in derivative 9 both the methyl
group on the 6 position and, above all, the 8-methoxy EDGs
strongly contribute to increase the stability of intermediate
phenoxyl radical, while the phenolic OH is not stabilized by
IHB with the heterocyclic sulfur atom in position 4 that is
structurally prevented to act as HB acceptor, as confirmed
by FT-IR experiments. In fact, with our great satisfaction,
compound 9 showed kinh and BDE(OH) better than any other
4-thiaflavane prepared so far and identical to those of a-
TOH.[2,3]
Figure 6. Structure of quinone 15; and effect of the C7 methyl group on
C
C
the conjugative stabilization of phenoxyl radicals 9 and 10 produced by
the C8 methoxy group.
Attempts to obtain crystals of 10 to verify the position of
methoxy group in the solid state were unsuccessful, as dark
red crystals of quinone 15 (Figure 6), were collected when a
diluted solution of 10 in CCl4 was slowly evaporated at
room temperature for 3 weeks. This reaction, which proba-
bly involves molecular oxygen as oxidizing agent, was not
observed for any of the other 5-hydroxy-4-thiaflavanes re-
ported in this paper.
On the light of these latter results, we redesigned a syn-
thetic procedure (see Scheme 6) that allowed the prepara-
tion of derivative 9 as single regioisomer, and of compound
11, which contains also a dihydroxy-functionalized B ring.
From the oxygen uptake plots reported in Figure 1, it can be
seen that in 11 both antioxidant moieties contributed to the
inhibition of styrene autoxidation, as the strongly inhibited
period due to the 5-hydroxy-4-thiaflavane is followed by a
weaker inhibition due to the catecholic group. Therefore,
12402
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12396 – 12404

Antioxidant Activity of Hydroxy Substituted 4-Thiaflavanes
FULL PAPER
system J=8.5 and 2.9 Hz, 1H), 5.71 (s, 1H, OH), 6.50 (s, 1H), 6.92–6.97
(m, 2H), 7.34–7.39 ppm (m, 2H); 13C NMR (50 MHz, CDCl3): d=16.0,
31.1, 55.3, 56.7, 75.7, 104.9, 109.7, 113.9 (2C) 117.6, 126.9 (2C) 132.8,
139.2, 139.5, 144.8, 159.3 ppm; IR (CCl4 0.01m): n˜(OH)=3550; (CCl4
0.001m): 3550 cmꢀ1; elemental analysis calcd (%) for C17H18O4S: C 64.13,
H 5.70; found: C 63.96, H 5.89.
the number of intercepted peroxyl radicals for each mole-
cule of 11 was doubled with respect to 9 (Table 1).
Conclusion
2,3-Dihydro-8-methoxy-2-(4-methoxyphenyl)-6-methylbenzo[b]-
AHCTUNGTERG[NNUN 1,4]oxathiin-5-ol (9): Directly after the cycloaddition (see Scheme 4), or
The unique stereoelectronic features of 5-hydroxy-4-thiafla-
vanes allowed the synthesis of a new class of phenolic anti-
oxidants having rate constants for the reaction with peroxyl
after cycloaddition and deprotection (see Scheme 6), a column flash chro-
matography on silica gel, using CH2Cl2/petroleum ether=8:1 as eluent,
allowed the isolation of compound 9 as a pale pink solid; m.p. 140–
1458C; 1H NMR (200 MHz, CDCl3): d=2.20 (s, 3H), 3.10–3.24 (m, AB
part of an ABX system, JAB =12.8 Hz, 2H), 3.79 (s, 3H), 3.81 (s, 3H),
4.47 (s, 1H, OH), 5.20 (dd, X part of an ABX system J=6.6 and 4.4 Hz,
1H), 6.49 (s, 1H), 6.89–6.94 (m, 2H), 7.32–7.37 ppm (m, 2H); 13C NMR
(50 MHz, CDCl3): d=15.7, 31.0, 55.3, 57.0, 75.9, 106.7, 111.3, 113.9 (2C),
114.6, 127.2 (2C), 132.1, 140.9, 143.0, 143.1, 159.4 ppm; IR (CCl4 0.01m):
n˜(OH)=3612, 3545; (CCl4 0.001m): 3436 cmꢀ1 (br); elemental analysis
calcd (%) for C17H18O4S: C 64.13, H 5.70; found: C 64.49, H 5.29.
ꢀ
radicals, kinh, and bond dissociation energy of the ArO H
group, similar or slightly better than those of a-tocopherol,
the main component of vitamin E and the more efficient lip-
ophilic antioxidant known in nature. In these compounds,
the cyclic o-alkylthio substituent stabilizes the phenoxyl rad-
C
ical formed on H-atom transfer to ROO , while having no
aptitude to accept intramolecular hydrogen bonding from
the reactive OH, which would decrease the reactivity. This
approach may represent an advantageous way to obtain
highly active antioxidants, not containing the chromane
moiety, without using strongly electron-donating alkylamino
groups that could cause toxicity problems.[27]
2,3-Dihydro-8-methoxy-2-(4-methoxyphenyl)-6,7-dimethylbenzo[b]-
AHCTUNGTERG[NNUN 1,4]oxathiin-5-ol (10): After cycloaddition and desilylation the crude
was purified by column flash chromatography on silica gel, using CH2Cl2
as eluent, to give derivative 10 as a glassy yellowish solid (67% yield
over two step); 1H NMR (400 MHz, CDCl3): d=2.12 (s, 3H), 2.18 (s,
3H), 3.10–3.15 (m, AB part of an ABX system, JAB =13.0 Hz, 2H), 3.72
(s, 3H), 3.83 (s, 3H), 4.57 (br s, 1H, OH), 5.18 (t, X part of an ABX
system, J=5.8 Hz, 1H), 6.92–6.95 (m, 2H), 7.34–7.38 ppm (m, 2H);
13C NMR (100 MHz, CDCl3): d=11.7, 12.2, 31.2, 55.3, 60.7, 76.2, 103.0,
114.0, 114.9, 127.1, 127.3, 132.5, 141.6, 144. 0, 145.0, 159.6 ppm; IR (CCl4
Experimental Section
0.01m): n˜(OH)=3614, 3549 cmꢀ1
; elemental analysis calcd (%) for
C18H20O4S: C 65.04, H 6.06; found: C 64.92, H 6.13.
See below for spectroscopic data of thiaflavanes 6–11, autoxidation stud-
ies, EPR spectroscopy, and IR measurements. All the other experimental
details are available in the Supporting Information.
4-(2,3-Dihydro-5-hydroxy-8-methoxy-6-methylbenzo[b]ACTHNUTRGNE[NUG 1,4]oxathiin-2-
yl)benzene-1,2-diol (11): After cycloaddition and deprotection the crude
reaction mixture was purified by flash chromatography over silica gel,
using CH2Cl2/EtOAc=2:1 as eluent, to give derivative 11 (80% yield) as
Methyl
8-t-butyl-2,3-dihydro-5-hydroxy-2-(4-methoxyphenyl)benzo[b]-
ACHTUNGTRENNUNG[1,4]oxathiine-6-carboxylate (6): After the cycloaddition the crude reac-
1
dark pink glassy solid; H NMR (400 MHz, [D6]aceton): d=2.21 (s, 3H),
tion mixture was purified by flash chromatography on silica gel, using pe-
troleum ether/EtOAc from 20:1 to 5:2 as eluent, to give derivative 6 as a
yellow oil (68% yield); 1H NMR (400 MHz, CDCl3): d=1.29 (s, 9H),
3.08 (dd, J=13.0 and 1.6 Hz, 1H), 3.27 (dd, J=13.0 and 9.8 Hz, 1H),
3.84 (s, 3H), 3.93 (s, 3H), 5.07 (dd, J=9.6 and 1.6 Hz), 6.95 (m, 2H), 7.26
(s, 1H), 7.35 (m, 2H), 7.52 (s, 1H), 11.30 ppm (s, 1H, OH); 13C NMR
(100 MHz, CDCl3): d=29.9 (3C), 30.7, 34.6, 52.1, 55.3, 76.5, 104.3, 106.6,
114.2 (2C), 122.8, 127.4 (2C), 131.3, 132.0, 157.0, 157.3, 159.8, 170.7 ppm;
elemental analysis calcd (%) for C21H24O5S: C 64.93, H 6.23; found: C
65.25, H 6.51.
3.12–3.20 (m, AB part of an ABX system, JAB =12.9 Hz, 2H), 3.73 (s,
3H), 4.96 (dd, X part of an ABX system, J=7.4 and 3.8 Hz, 1H), 6.55 (s,
1H), 6.88 (s, 2H), 7.02 (s, 1H), 7.21 (s, 1H, OH), 7.976 (s, 1H, OH),
7.984 ppm (s, 1H, OH); 13C NMR (100 MHz, [D6]aceton): d=15.2, 30.4,
56.3, 75.6, 111.4, 113.4, 115.1, 115.3, 117.8 (2C), 132.6, 141.6, 143.4, 144.2,
145.1 ppm (2C); elemental analysis calcd (%) for C16H16O5S: C 59.99, H
5.03; found: C 59.96, H 5.17.
Autoxidation studies: The chain-breaking antioxidant activity of the title
compounds was evaluated by studying the inhibition of the thermally ini-
tiated autoxidation of either styrene or cumene (RH) in chlorobenzene.
Autoxidation experiments were performed in a two-channel oxygen-
uptake apparatus, based on a Validyne DP 15 differential pressure trans-
ducer.[16] In a typical experiment, an air-saturated chlorobenzene solution
of styrene or cumene containing AIBN (0.05m) was equilibrated with an
identical reference solution containing also an excess of PMHC (1ꢀ
10ꢀ3 m) in the same solvent at 308C. When a constant oxygen consump-
tion was observed, the antioxidant was injected into the sample flask,
and the oxygen consumption was measured, after calibration of the appa-
ratus, from the differential pressure recorded with time between the two
channels. This experimental setting allowed us to subtract, from the au-
toxidation oxygen consumption, the N2 produced during AIBN decompo-
sition and the O2 consumed by the initiating radicals.[16] The value of kinh
was determined from the oxygen consumption during the inhibition
period by using Equation (10), which is the integrated form of Equa-
tion (11).[2] In Equations (10) and (11), kp is the propagation rate constant
of the oxidizable substrate (41mꢀ1 sꢀ1 for styrene[28] and 0.32mꢀ1 sꢀ1 for
cumene[23]) and t is the length of the induction period (an example of the
use of Equation (10) is reported in the Supporting Information). In the
case of compounds 7 and 8, which did not afford a clear induction
period, the kinh was determined using a different equation (see the Sup-
porting Information).[17] The coefficient n was determined experimentally
from t using Equation (12).[2]
8-t-Butyl-2,3-dihydro-6-(hydroxymethyl)-2-(4-methoxyphenyl)benzo[b]-
ACHTUNGTRENNUNG[1,4]oxathiin-5-ol (7): To a suspension of LiAlH4 (50 mg, 1.32 mmol) in
dry THF (1 mL) under a nitrogen atmosphere was added at ꢀ108C a so-
lution of 6 (64 mg, 0.15 mmol) in dry THF (2 mL) and stirred at this tem-
perature for 30 min. After this time the mixture was diluted with H2O
(30 mL) and acidified with a 10% solution of HCl in MeOH. The solu-
tion was extracted with Et2O (3ꢀ50 mL) and washed with H2O (3ꢀ
100 mL) and brine. The organic layer was dried over anhydrous Na2SO4
and concentrated under reduced pressure give derivative 7 as a colorless
oil (72% yield); 1H NMR (400 MHz, CDCl3): d=1.28 (s, 9H), 3.07 (dd,
J=12.8 and 1.8 Hz, 1H), 3.27 (dd, J=12.8 and 10.0 Hz, 1H), 3.84 (s,
3H), 4.82 (s, 2H), 4.98 (dd, J=10.0 and 1.8 Hz), 6.70 (s, 1H), 6.94–6.97
(m, 2H), 7.35–7.37 (m, 2H), 7.57 ppm (br s, 1H, OH); 13C NMR
(100 MHz, CDCl3): d=30.1 (3C), 31.4, 34.5, 55.3, 64.9, 75.9, 107.2, 114.0
(2C), 115.7, 121.0, 127.2 (2C), 130.9, 132.5, 150.2, 152.2, 159.3 ppm; ele-
mental analysis calcd (%) for C20H24O4S: C 66.64, H 6.71; found C 66.85,
H 6.59.
2,3-Dihydro-6-methoxy-2-(4-methoxyphenyl)-8-methylbenzo[b]-
ACHTUNGTRENNUNG[1,4]oxathiin-5-ol (8): After the cycloaddition and column flash chroma-
tography on silica gel using CH2Cl2/petroleum ether=8:1 as eluent, com-
pound 8 was isolated as a pale yellow solid; m.p. 138–1438C; 1H NMR
(200 MHz, CDCl3): d=2.16 (s, 3H), 3.04–3.23 (m, AB part of an ABX
system, JAB =13.2 Hz, 2H), 3.83 (s, 6H), 5.06 (dd, X part of an ABX
Chem. Eur. J. 2011, 17, 12396 – 12404
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12403

S. Menichetti, R. Amorati et al.
kp½RHꢂ
ð1 ꢀ tÞ
Viglianisi, Helv. Chim. Acta 2006, 89, 2462–2472; e) R. Amorati,
O. A. Attanasi, G. Favi, S. Menichetti, G. F. Pedulli, C. Viglianisi,
Org. Biomol. Chem. 2011, 9, 1352–1355.
ð10Þ
ð11Þ
ð12Þ
ꢀD½O2ꢂt ¼
d½O2ꢂ
ln
kinh
t
kp½RHꢂRi
n kinh½AHꢂ
[10] S. Menichetti, R. Amorati, M. G. Bartolozzi, G. F. Pedulli, A. Salvi-
ni, C. Viglianisi, Eur. J. Org. Chem. 2010, 2218–2225.
[11] R. Amorati, A. Cavalli, M. G. Fumo, M Masetti, S. Menichetti, C.
Pagliuca, G. F. Pedulli, C. Viglianisi, Chem. Eur. J. 2007, 13, 8223–
8230.
ꢀ
¼
dt
Ri t
n ¼
½AHꢂ
[12] R. Amorati, F. Catarzi, S. Menichetti, G. F. Pedulli, C. Viglianisi, J.
Am. Chem. Soc. 2008, 130, 237–244.
[13] J. Demyttenaere, K. Van Syngel, A. P. Markusse, S. Vervisch, S. De-
benedetti, N. De Kimpe, Tetrahedron 2002, 58, 2163–2166.
[14] L. Xie, Y. Takeuchi, L. M. Cosentino, A. T. McPhail, K.-H. Lee, J.
Med. Chem. 2001, 44, 664–671.
[15] T. A. Blizzard, F. DiNinno, J. D. Morgan II, H. Y. Chen, J. Y. Wu, C.
Gude, S. Kim, W. Chan, E. T. Birzin, Y. T. Yang, L.-Y. Pai, Z.
Zhang, E. C. Hayes, C. A. DaSilva, W. Tang, S. P. Rohrer, J. M.
Schaeffer, M. L. Hammond, Bioorg. Med. Chem. Lett. 2004, 14,
3861–3864.
[16] M. Lucarini, G. F. Pedulli, Chem. Soc. Rev. 2009, 38, 2106–2119 and
references cited therein.
[17] R. Amorati, G. F. Pedulli, L. Valgimigli, O. A. Attanasi, P. Filippone,
C. Fiorucci, R. Saladino, J. Chem. Soc. Perkin Trans. 2 2001, 2142–
2146.
EPR spectroscopy: Spectra were recorded in quartz tubes at 298 K on a
Bruker Elexsys 500 X-band spectrometer equipped with a Bruker VT-
1000 variable temperature unit. Spectral analysis was optimized by
means of computer simulations and subjected to a least-squares fitting
procedure based on the systematic application of the Monte Carlo
method, available in WinESR Commander V.1.0 software, developed by
Prof. Marco Lucarini (University of Bologna). Spectra were recorded in
deoxygenated benzene solutions containing 10% (v/v) tBuOOtBu, by ir-
radiating the samples with a 500W high-pressure Hg lamp and using cali-
brated metal sectors to modulate the intensity of irradiation. Measured
g-factors were corrected using those of reference compounds in benzene:
BHT, g=2.0046 and dBHA g=2.0047.[20]
IR measurements: FT-IR spectra of compounds 7–10, were measured in
diluted tetrachloromethane solutions (0.01–0.003m) in a sealed KBr cell
with a 0.5 mm optical path.
[18] a) M. I. de Heer, P. Mulder, H.-G. Korth, K. U. Ingold, J. Lusztyk, J.
Am. Chem. Soc. 2000, 122, 2355–2360; b) R. Amorati, S. Menichet-
ti, E. Mileo, G. F. Pedulli, C. Viglianisi, Chem. Eur. J. 2009, 15,
4402–4410.
Acknowledgements
[19] E. T. Denisov, Kinet. Catal. 2006, 47, 662–671.
[20] M. Lucarini, P. Pedrielli, G. F. Pedulli, S. Cabiddu, C. Fattuoni, J.
Org. Chem. 1996, 61, 9259–9263.
[21] P. Mulder, H.-G. Korth, D. A. Pratt, G. A. DiLabio, L. Valgimigli,
G. F. Pedulli, K. U. Ingold, J. Phys. Chem. A 2005, 109, 2647–2655.
[22] M. Lucarini, G. F. Pedulli, M. Guerra, Chem. Eur. J. 2004, 10, 933–
939.
[23] R. Amorati, M. G. Fumo, S. Menichetti, V. Mugnaini, G. F. Pedulli,
J. Org. Chem. 2006, 71, 6325–6332.
[24] M. H. Langoor, J. H. van der Maas, J. Mol. Struct. 1997, 403, 213–
229.
Financial support from Ministero dell’Universitꢁ e della Ricerca (MIUR)
(research project “Radicals and Radical Ions: Basic Aspects and Role in
Chemistry, Biology, and Material and Environmental Sciences”, contract
No.2006033539 is gratefully acknowledged. Fondazione CARIPLO
(Milano) and Consorzio Interuniversitario Nazionale, Metodologie
e
Processi Innovativi di Sintesi (CINMPIS) are acknowledged for grants to
C. V. and M. G. B., respectively. Dr. Cristina Faggi is acknowledged for
X-ray analysis and Prof. Marco Lucarini for access to the EPR simulation
software.
[25] a) T. Schaefer, T. A. Wildman, R. S. Salman, J. Am. Chem. Soc.
1980, 102, 107–110; b) T. Schaefer, R. S. Salman, T. A. Wildman,
P. D. Clark, Can. J. Chem. 1982, 60, 342–348; c) T. Schaefer, D. M.
McKinnon, R. Sebastian, J. Peeling, G. H. Penner, R. P. Veregin,
Can. J. Chem. 1987, 65, 908–914; d) T. Schaefer, G. H. Penner, Can.
J. Chem. 1988, 66, 1229–1238.
[26] As a referee suggested, the orbital situation around the O and
S atoms in 4-thiaflavanes and in the corresponding phenoxyl radicals
is more complex than those reported in Figure 4 and Figure 5. How-
ever, as it has already been done,[3,11,25] with the aim to better under-
stand and visualize the stereoelectronic contributions to phenoxyl
radicals stabilization we decided to indicate exclusively the p orbi-
tals in charge for delocalization.
[1] P. Mulder, H.-G. Korth, K. U. Ingold, Helv. Chim. Acta 2005, 88,
370–374.
[2] G. W. Burton, K. U. Ingold, J. Am. Chem. Soc. 1981, 103, 6472–
6477.
[3] G. W. Burton, T. Doba, E. J. Gabe, L. Hughes, F. L. Lee, L. Prasad,
K. U. Ingold, J. Am. Chem. Soc. 1985, 107, 7053–7065.
[4] M. Wijtmans, D. A. Pratt, L. Valgimigli, G. A. DiLabio, G. F. Pedulli,
N. A. Porter, Angew. Chem. 2003, 115, 4506–4509; Angew. Chem.
Int. Ed. 2003, 42, 4370–4373.
[5] E. T. Denisov, I. B. Afanasꢄev, Oxidation and Antioxidants in Or-
ganic Chemistry and Biology, CRC Press, Boca Raton, 2005.
[6] C. Jacob, G. I. Giles, N. M. Giles, H. Sies, Angew. Chem. 2003, 115,
4890–4907; Angew. Chem. Int. Ed. 2003, 42, 4742–4758.
[7] a) D. Shanks, R. Amorati, M. G. Fumo, G. F. Pedulli, L. Valgimigli,
L. Engman, J. Org. Chem. 2006, 71, 1033–1038; b) J. Malmstrçm,
M. Jonsson, I. A. Cotgreave, L. Hammarstrom, M. Sjodin, L.
Engman, J. Am. Chem. Soc. 2001, 123, 3434–3440.
[27] a) C. M. Burnett, E. I. Goldenthal, Food Chem. Toxicol. 1988, 26,
467–474; b) G. J. Nohynek, D. Duche, A. Garrigues P.-A. Meunier,
H. Toutain, J. Leclaire, Toxicol. Lett. 2005, 158, 196–212.
[28] J. A. Howard, K. U. Ingold, Can. J. Chem. 1965, 43, 2729–2736.
[8] G. Capozzi, C. Falciani, S. Menichetti, C. Nativi, J. Org. Chem. 1997,
62, 2611–2615.
Received: April 14, 2011
Revised: August 2, 2011
[9] a) G. Capozzi, P. Lo Nostro, S. Menichetti, C. Nativi, P. Sarri, Chem.
Commun. 2001, 551–552; b) S. Menichetti, M. C. Aversa, F. Cimino,
A. Contini, C. Viglianisi, A. Tomaino, Org. Biomol. Chem. 2005, 3,
3066–3072; c) M. Lodovici, S. Menichetti, C. Viglianisi, S. Caldini,
E. Giuliani, Bioorg. Med. Chem. Lett. 2006, 16, 1957–1960; d) R.
Amorati, M. G. Fumo, G. F. Pedulli, S. Menichetti, C. Pagliuca, C.
Please note: Minor changes have been made to this manuscript since its
publication in Chemistry—A European Journal Early View. The Editor.
Published online: September 29, 2011
12404
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12396 – 12404