Guaiazulene-based phenolic radical scavengers: Synthesis, properties, and EPR studies of their reaction with oxygen-centred radicals

Article abstract of DOI:10.1016/S0040-4020(03)00737-3

A variety of phenolic derivatives 4, carrying the guaiazulene moiety, were prepared starting from guaiazulene. Compounds 4 react with oxygen-centred radicals exhibiting chromotropic behaviour. The radical scavenging power of these compounds was evaluated

Full text of DOI:10.1016/S0040-4020(03)00737-3

Tetrahedron 59 (2003) 5003–5018
Guaiazulene-based phenolic radical scavengers: synthesis,
properties, and EPR studies of their reaction with
oxygen-centred radicals
a
b
,
Emanuela Franchi,a Giovanni Ingrosso,a Fabio Marchetti and Calogero Pinzino
,
*
*
a
`
Dipartimento di Chimica e Chimica Industriale, Universita di Pisa, Via Risorgimento 35, 56126 Pisa, Italy
bIstituto per i Processi Chimico-Fisici del C.N.R., Area della Ricerca di Pisa, Via G. Moruzzi 1, 56124 Pisa, Italy
Received 20 February 2003; revised 14 April 2003; accepted 8 May 2003
Abstract—A variety of phenolic derivatives 4, carrying the guaiazulene moiety, were prepared starting from guaiazulene. Compounds 4
react with oxygen-centred radicals exhibiting chromotropic behaviour. The radical scavenging power of these compounds was evaluated by
different methods. Compounds 4 are less efficient than some of the most common radical scavengers but show quite selective behaviour
towards different oxygen-centred radicals. A correlation is found between the antioxidant activity of the compounds 4 and the corresponding
phenolic O–H bond dissociation energy. Some aspects of the reaction of the compounds 4 with oxygen-centred radicals were elucidated by
EPR and DFT studies. q 2003 Elsevier Science Ltd. All rights reserved.
1. Introduction
corresponding phenoxyl radical. Thus, whenever the
chromophoric group absorbs in the visible region, the
phenol!phenoxyl transformation may be accompanied by a
colour change. Radical scavengers of this type should then
exhibit chromotropic behaviour on reaction with free
radicals and, at least in principle, can be used to detect the
presence of free radicals through a simple organoleptic or
colorimetric approach.
Recent years have witnessed a great level of interest in the
development of new compounds that allow us to prevent
damage that can be associated with oxidative chain
reactions induced by free radicals. Such processes can be
inhibited or retarded by the addition of certain compounds
known as ‘antioxidants’.1 The radical scavengers2 are a
particular class of antioxidants able to react with free
radicals giving rise to a new radical species less reactive
than its precursor, the original radical being transformed
into a non-radicalic species. A wide variety of natural and
synthetic compounds are known to exhibit powerful radical
scavenging activity and, within these, sterically hindered
phenols play a very important role.2 Typically, the reaction
of phenols with radicals implies a hydrogen transfer from
the phenol to the reactive radical with the formation of the
less reactive phenoxyl radical (Eq. (1)).3
We were attracted by the idea of synthesising new phenol-
based radical scavengers having the above properties. In
projecting the structure of the new compounds we were
inspired by an interesting paper dealing with the detection of
free radicals through their adducts with a guaiazulene-based
nitrone 1.4
ð1Þ
The electronic properties of a chromophoric group attached
to the phenol skeleton, and then its absorption spectrum,
may be markedly modified on passing from phenol to the
Guaiazulene, a deeply blue coloured alkyl derivative of
azulene, appeared particularly promising as the chromo-
phore to be introduced into the molecular skeleton of the
desired phenol derivatives. Azulene and its derivatives have
a remarkable dipole moments (m¼0.8–1.08 D) and unusual
Keywords: phenol derivatives; antioxidants; radical scavengers;
guaiazulene; EPR; DFT calculations.
*
Corresponding authors. vanni@dcci.unipi.it; rino@indigo.icqem.pi.cnr.it
0040–4020/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4020(03)00737-3

5004
E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
photo-physical properties, and, as such, have attracted great
interest in organic, inorganic, medicinal, theoretical, and
industrial chemistry.5
We report here the synthesis of several new phenol
derivatives of formula 4, as well as an evaluation of their
radical scavenging power, along with some mechanistic
aspects of their reaction with oxygen-centred radicals. Brief
reports on the preliminary results of this study have already
been communicated.6 A laboratory test for detecting the
presence of free radicals based upon the use of the
compounds 4 has been set up and patented.7
2. Results and discussion
2.1. Synthesis and properties of the compounds 4
The derivatives 4a–4j were prepared, according to
Scheme 1, by reacting the readily available acyl bromide
3, which derives from guaiazulene 2,4 with a variety of
dihydroxybenzene derivatives. All the compounds were
obtained in moderate yields (20–50%), the main side
product being in all cases the diesters of general formula 5
whose formation can be limited if the reaction is carried out
by adding slowly the acyl bromide 3 to an ethereal solution
of the appropriate dihydroxybenzene, at 2308C.
Compounds 4 and 5 were fully characterized by elemental
analysis, 1H NMR, and, in the case of 4j, also by 19F NMR.
The isomers 4d–4g were unequivocally identified by means
of NOE measurements.
With the aim of showing differences and analogies between
the role played by the guaiazulene moiety and that played by
other carbocyclic moieties in influencing the antioxidant
Scheme 1.

E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
5005
properties of the resulting phenolic derivatives, compounds
6 and 7, which contain the phenyl and the 1-naphthyl
moieties, respectively, were prepared.
with the hydroxyl radical. This is interesting at least within
some contexts as in the case of fibre bleaching by hydrogen
peroxide-based products. In such a case, in fact, while the
superoxide and its conjugate acid HOOz seems to be
essential for the bleaching process, the hydroxyl radical is
highly undesired, it being responsible for the degradation of
the fibre.11 In this connection, it is interesting to underline
that 2,6-di-tert-butyl-4-hydroxytoluene (BHT) exhibits only
a moderate selectivity on reacting with various oxygen-
centred radicals, under the same experimental conditions as
above. In fact, BHT is not totally inert towards KO2 in
DMSO.
On the basis of the above findings, a simple laboratory test
has been set up, based upon the use of the compounds 4, for
detecting the presence of oxygen-centred free radicals in
micellar media, simply by observing a colour change.7
Quite surprisingly, when dissolved in an aqueous solution of
n-decyl sulfate sodium salt (1 M; pH¼9), i.e. under the
same experimental conditions as above, compounds 6 and 7
undergo a slow hydrolytic reaction leading to the formation
of hydroquinone and of the conjugate bases of the
corresponding acids. Instead, 6 and 7 are stable in acidified
n-decyl sulfate sodium salt surfactant (pH¼4). Thus, it was
possible to test their reactivity towards oxygen-centred
radicals in such a medium. The results of these experiments
show that both 6 and 7 exhibit very poor scavenging power
towards the oxygen-centred radicals derived from the
Fenton reaction, if compared with the compounds 4. Indeed,
TLC of the reaction products revealed that both 6 and 7
remained unchanged within the adopted reaction time
(15 min). An interesting qualitative conclusion is that the
presence of the guaiazulenic moiety in the compounds 4 not
only confers chromotropic behaviour but also makes the
ester bond less prone to basic hydrolysis.
Like other phenol derivatives,2 compounds 4 react with
some oxygen-centred radicals (Scheme 2). Owing to their
insolubility in water, all the experiments were carried out in
an aqueous solution of n-decyl sulfate sodium salt (1 M).
The tests involving H2O2 as the oxidant were carried out at
pH<4 using 2 M H2O2 solutions. No reaction takes place
when the compounds 4 are dissolved in a O2-saturated
micellar medium, either in the presence or in the absence of
[Cu(en)2][ClO4]2 (0.01 M), or are reacted with H2O2 in the
absence of any metal redox couple. Indeed, no colour
variation occurred within 15 min. Moreover, TLC analysis
of the reaction mixtures revealed only the presence of the
starting guaiazulene derivative. Instead, a fast reaction takes
place when compounds 4 are reacted with the Fenton-like
system [Cu(en)2][ClO4]2/H2O2,8 that, as well known,9 gives
rise to the hydroperoxyl (HOOz), superoxide (O22z), and
hydroxyl (HOz) radicals: the blue colour of the reaction
mixtures disappears within 15 min, turning first red, then
pale orange, and, finally, pale yellow. The reactivity towards
the superoxide radical was tested by reacting the compounds
4 with a 0.1 M solution of KO2 in dimethylsulphoxide
(DMSO) in the presence of 18-crown-6 (0.1 M).10 No
colour variation was observed in any case. The observed
chemical inertness towards the superoxide radical seems to
indicate that, in the simultaneous presence of HOOz, O22z and
HOz, the guaiazulenic derivatives 4 are able to react fast only
The qualitative results concerning the reactivity of the
compounds 4 with some of the most common oxygen-
centred radicals prompted us to try to evaluate quanti-
tatively their radical scavenging power towards specific
radical species. To do this some of the published methods
were adopted as described in the literature or slightly
modified to take into account the properties (solubility and
spectroscopic features) of compounds 4. In particular, the
method of the inhibition of the autoxidation of linoleic
acid12,13 allowed us to evaluate their peroxyl radical
scavenging power in a micellar system. Moreover, some
of the compounds 4 were also tested with the oxygen up-
take method14 which allowed us to evaluate their peroxyl
radical scavenging power in a homogeneous organic
Scheme 2.

5006
E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
medium. Finally, by using an appropriately modified
version of the inhibition of the degradation of deoxyribose
method,15,16 the radical scavenging efficiency of 4a–4j
towards the hydroxyl radical17 was evaluated.
the radical scavengers 4h and 4i, which carry also an alkyl
group at the meta positions, show lower RAE values, and
the compounds 4e and 4g, which carry only one meta alkyl
group, exhibit very low scavenging power. Finally, the least
efficient radical scavengers are the compounds 4a and 4b,
both containing an unsubstituted phenolic rings which
undoubtedly have a lower electron density than all the
others because of the lack of the positive inductive effect
exerted by the alkyl groups. It is indeed well known18 that
the electron density on the phenolic ring is markedly
influenced by the position and the electron donor–acceptor
properties of its substituents, ortho and para substitution
being much more effective. Since radicals are electrophilic
species, a phenolic ring will be more susceptible to the
attack by such species, i.e. will be activated, if it carries
electron-donating substituents. On the other hand, the
presence of electron-withdrawing substituents on the
phenolic ring plays, expectedly, the opposite role.
The evaluation of the rate for the autoxidation of linoleic
acid to the corresponding conjugated diene hydroperoxide
(Eq. (2)), both in the absence and in the presence of an
antioxidant, constitutes a convenient method for measuring
the efficiency of a given compound in scavenging peroxyl
radicals in a micellar medium.12 This is based on the
recording the variation of the UV absorbance (l¼234 nm)
of the hydroperoxide vs time.
ð2Þ
Starting from the RAE value it is possible to calculate the
absolute value of the inhibition rate constant kinh of a given
radical scavenger.12 In fact, the multiplication of the RAE
value by the absolute rate constant for the reaction of
peroxyl radicals with a-tocopherol, converts the RAE value
into the absolute value of kinh for each inhibitor. The value
of kinh for a-tocopherol in this micellar system has been
measured by Barclay et al.19 and found to be
4£104 M21 s21. The calculated values for the rate constant
for the inhibition of the autoxidation of linoleic acid by
compounds 4d,4f,4h,4i and BHT are reported in Table 1.
Interestingly, the inhibition rate constant we find in the case
of BHT (1.04 M21 s21) agrees very well with the reported
value (1.1 M21 s21).20
Accordingly, an aqueous buffered (pH¼7.4) solution
containing linoleic acid (2.6£1023 M) and sodium
n-dodecyl sulfate (0.1 M) was reacted with 2,20-azobis(2-
amidinopropane)hydrochloride (ABAP) as radical initiator,
under air atmosphere at 378C. The mathematical treatment
of the experimental data was performed exactly as reported
in the literature.12 The resulting relative antioxidant
efficiency (RAE) of the compounds 4a–4i and of BHT are
summarized in Table 1.
The above results led to two main questions. Do compounds
4 have an intrinsically low peroxyl radical scavenging
activity? Does their activity depend markedly upon the
reaction medium? Thus, we were keen to evaluate the
antioxidant activity of some of the compounds 4 toward
peroxyl radicals in a non-micellar homogeneous solution.
This was done by adopting the oxygen up-take method that
is based on the direct recording of the dioxygen consump-
tion using a highly sensitive pressure transducer, during the
inhibition of an autoxidation process. This method, set up by
Ingold and co-workers14 to evaluate the styrene polymeris-
ation inhibition rate, is now largely employed in various
systems. Thus, we studied the inhibition of cumene
autoxidation to cumylhydroperoxide by some compounds
4 in cumene itself. The measurements were carried out at
508C monitoring the variation of oxygen concentration
during the inhibition of the autoxidation of cumene initiated
by 2,20-azo-bis(isobutyronitrile) (AIBN) (2.95£1023 M).
Four sets of experiments were run using the compounds
4a,4c,4i and 4j. Two types of kinetic profiles were observed:
the behaviour of 4i is quite similar to that observed for
a-tocopherol (Fig. 1), while those shown by the other
compounds look like that of BHT (Fig. 2). These results are
the consequence of the different efficiency of these
compounds in the peroxyl scavenging activity; in particular
in the case of the experiments reported in Figure 2, the
steady-state kinetic analysis could not be applied owing to
the existence of competitive reactions. For this reason, two
different kinetic mathematical treatments were carried out.
The RAE values show that the compounds 4a–4i are
markedly less efficient than BHT as far as their peroxyl
radical scavenging power in a micellar system is concerned.
In particular, three different behaviours can be envisaged
depending upon the position and the nature of the phenolic
substituents. The relative antioxidant efficiency of 4d,4f,4h
and 4i can be evaluated with a reasonable accuracy. The
compounds 4c,4e and 4g show, instead, too lower inhibition
power to be quantitatively measured. Finally, the other
radical scavengers are completely inert towards the peroxyl
radicals. Thus, the data show clearly that the most efficient
radical scavengers have at least one alkyl substituent at the
ortho positions (compounds 4d,4f,4h and 4i). Interestingly,
Table 1. Relative antioxidant efficiency of the compounds 4a–4i and BHT
and rate constants for the inhibition of the autoxidation of linoleic acid by
compounds 4d,4f,4h and BHT
Radical scavenger
RAE
Error (%)
kinh£1024 (M21 s21
)
BHT
4f
4d
4i
4h
4e
4g
4c
0.26
0.056
0.040
0.025
0.0077
Very low
Very low
Very low
No inhibition
No inhibition
10
20
2
8.6
1.04
0.22
0.16
0.10
0.03
–
–
–
–
–
3
–
–
–
–
–
4a
4b

E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
5007
Table 2. Rate constants for the inhibition of linoleic acid and cumene
autoxidation by a-tocopherol, BHT, and compounds 4a and 4i
Radical scavenger
kinh
(M21 s21
kinh
(M21 s21
)
)
(Homogeneous medium)
(508C)
Cumene autoxidation
(Micellar medium)
(378C)
Linoleic acid autoxidation
a-Tocopherol
n.d.
4£104
BHT
4i
4a
3.4£104
16£104
5.3£104
1.0£104
0.1£104
Too low to be evaluated
comparison between the kinh values in cumene and in SDS is
reported.
Figure 1. Dioxygen up-take (M) during the autoxidation of cumene
inhibited by a-tocopherol (—) and by 4i ( ) vs time; T, 508C;
concentration of the radical scavenger, 5£1023 mM; concentration of
the initiator, 3£1023 M.
Some interesting conclusions can be drawn from the
tabulated data. First, in all cases the peroxyl scavenging
efficiency decreases on going from homogeneous organic
medium to micellar medium, as already observed,23
although no dramatic variation of kinh is observed in the
case of BHT on passing from the micellar medium to
homogeneous organic medium, according to the literature.20
Instead, the guaiazulene derivative 4a, that is almost
inactive in micellar medium, shows a peroxyl scavenging
activity slightly higher than that of BHT in homogeneous
medium, and a much more marked activity increase is
observed in the case of 4i. This is 10 times less active than
BHT in micellar medium, and five times more active than
BHT in homogeneous medium. Finally, we would underline
that although the inhibition rate constants were calculated
from experiments carried out at different temperatures, the
observed differences cannot be accounted for only on this
basis.
Figure 2. Dioxygen up-take (M) vs time during the inhibition of cumene
autoxidation by BHT, 4a,4c and 4j; T, 508C; concentration of the radical
scavenger, 5£1023 mM; concentration of the initiator, 3£1023 M.
The HOz scavenging power of the compounds 4 was tested
by a very simple and largely employed method based upon
the inhibition of the deoxyribose degradation.15,16 Accord-
ingly, HOz radicals are generated by a classical Fenton
reaction and reacted with 2-deoxy-D-ribose which under-
goes an oxidative degradation reaction which results in the
formation of malondialdehyde (MDA). On reaction with
thiobarbituric acid (TBA), MDA gives rise to a pink adduct
In the first case (Fig. 1), the value of the inhibition rate
constant for the reaction of 4i with peroxyl radicals, i.e. kinh
(16£104 M21 s21), was calculated according to the pro-
cedure reported in literature.21 Furthermore, owing to the
fact that the experimental curves relative to 4i and to
a-tocopherol have quite similar shapes, one can conclude
that the stoichiometric factor (SF) for the reaction involving
4i, that was assumed to be 2 as verified for a-tocopherol, is
indeed not very distant from this value (1.98).12 From the
data reported in Figure 2, it was possible to calculate kinh
(5.3£104 M21 s21) only for the inhibition reaction by
compound 4a, using a computational program22 which
integrates all the kinetic equations involved in the inhibition
process. On the other hand, compounds 4c and 4j have very
poor autoxidation power. The different behaviour of
compound 4c, if compared with that of 4a, can possibly
be accounted for by considering that the hydroxyl group in
4c is too sterically hampered to be approached by the
cumyloxyl radical. On the other hand, the behaviour of 4j
confirms that the presence of electron-withdrawing groups
over the phenolic ring results in a marked decrease of the
radical scavenging efficiency.
which absorbs at 532 nm (1¼1.54£105 M21 cm21
)
(Eqs. (3)–(5)).
Fe2þ þ H2O2 ! OH2 þ OHz þ Fe3þ
OHz þ Deoxyribose ! Malonaldehyde
ð3Þ
ð4Þ
ð5Þ
On the basis of all the results obtained for the inhibition of
autoxidation process by compounds 4, some conclusive
considerations about the role played by the media reaction
in the scavenging activity may be reported. In Table 2 a
The extent of the formation of this adduct can be evaluated
through UV–vis measurements. When a radical scavenger
is added to the reactive system, it can compete with

5008
E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
Table 3. OH-Scavenging activity of derivatives 4a–4j and BHT
parameters that influence the reaction.24–26 As shown in the
next section, we have acquired some evidence that the
derivatives 4 exert their radical scavenging activity giving
rise to the corresponding phenoxyl radicals. Thus, we tried
to verify if a correlation exists between the antioxidant
power of compounds 4 and their O–H bond dissociation
energy, D(O–H). The DFT calculation of the D(O–H) for
the compounds 4a,4c,4i and 4j was carried out with Spartan
5.1 program,27 assuming that the difference in the solvating
free energy between the phenolic derivative and the
corresponding phenoxyl radical is independent of substi-
tution. The energies E, ER and EH, from the optimised
geometries at pBN/DNp p level, were obtained for phenol
derivatives, phenoxyl radicals, and hydrogen radicals,
respectively. The phenolic O–H bond dissociation
energy values were thus obtained from the equation
D(O–H)¼E2ER2EH (Table 4).
Radical scavenger
OH-SAa (%)
Error (%)
BHT
4i
4b
4d
4a
4
4h
4f
4e
4c
4j
47.7
46.2
46.2
38.4
37.0
36.8
34.7
34.3
34.1
33.7
4.0
5.5
2.2
8.5
5.7
11.0
8.3
2.9
2.3
4.2
3.0
2.5
a
Each figure represents the average over three runs.
deoxyribose in the reaction with the hydroxyl radical, thus
limiting the degradation process of the sugar and then
limiting the formation of the MDA–TBA adduct. UV–vis
measurements allowed us to evaluate the entity of the
inhibition process by recording the decrease of the
absorbance at 532 nm vs the absorbance of a reaction
mixture not containing any radical scavenger. It was
necessary to modify adequately the published experimental
procedure16 in consequence of the insolubility of the
compounds 4 in water as well as of their spectroscopic
peculiarities. Thus, the reactions were carried out in sodium
n-dodecyl sulfate (0.1 M). Moreover, in order to avoid the
interference of the absorption by the guaiazulenic moiety at
532 nm, the measurement of the absorption of a blank,
containing only the reactive system and the radical scavenger,
was necessary. The OH-Scavenging Activity (OH-SA)
(Table 3) could be then determined through the Eq. (6).
It is apparent from the data of Table 4 that the bond
dissociation energies of the examined guaiazulene deriva-
tives decrease in the order 4i,4a,4c,4j. Interestingly, the
hydroxyl radical scavenging activity of these compounds
(see Table 3) decreases in the same order. The fluorinated
derivative 4j exhibits the highest bond dissociation energy
and, accordingly, a very poor radical scavenging activity.
On the other hand, compound 4i exhibits the lowest value of
bond dissociation energy and is the most efficient scavenger
of peroxyl radicals both in micellar and organic media,
being almost five times more efficient that BHT in the latter
medium (see Table 2).
Table 4. Bond dissociation energy of some phenolic derivatives
(
)
A
ðRþSþTBAþTCAAÞ 2 AðRþSÞ
Radical scavenger
O–H Dissociation energy
(kcal mol21
OH-SA ¼ 1 2
£ 100
ð6Þ
)
AðRþTBAþTCAAÞ
4a
4c
4i
84.9
86.5
80.4
90.2
where R indicates the reactive system: FeSO4·7H2O, EDTA,
2-deoxyribose, phosphate buffer, H2O2; S refers to the radical
scavenger; TBA is thiobarbituric acid; TCAA is trichloroace-
tic acid; A(RþSþTBAþTCAA) is the absorbance of the inhibited
system at 532 nm; A(RþS) is the absorbance at 532 nm of the
inhibited system not containing either TBA or TCCA, which
correlates to the unreacted scavenger; A(RþTBAþTCAA) is the
absorbance of the uninhibited system at 532 nm.
4j
87.632
The data in Table 3 show clearly that the compounds 4, with
the only exception of the fluorinated compound 4j, are
moderately less efficient than BHT in scavenging the
hydroxyl radical. All of them exhibit indeed a radical
scavenging power not higher than 46%. A main conclusion
is then that the substitution of the phenolic ring has a little
effect on the reactivity of the compounds 4 towards the
hydroxyl radical. Only the fluorine derivative 4j shows a
very low radical scavenging power, which confirms the
already observed effect caused by the presence of electron-
withdrawing groups on the phenolic ring.
81.132
78.732
It is well known3 that phenol derivatives react with free
radicals undergoing the hydroxylic hydrogen transfer
process, thus giving rise to the corresponding phenoxyl
radical. It is thus expected that a low phenolic O–H bond
dissociation energy (OH-BDE) should favour the hydrogen
transfer process, although this is only one of the several

E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
5009
˚
Table 5. Selected experimental and (calculated) bond lengths (A) of 4a and
4b
Finally, a comparison of the O–H bond dissociation
energies of the guaiazulenic derivatives with those reported
in the literature for phenol, BHT, and a-tocopherol (see
Table 4) points out clearly that BHT and a-tocopherol are
indeed characterised by a lower bond dissociation energy
value, while phenol, which is not a good antioxidant, has an
O–H bond dissociation energy higher than those exhibited
by the compounds 4a,4c and 4i.
Compound 4a
Compound 4ba
Bond
Bond
Length
Length
O(3)–C(20)
C(17)–O(1)
O(1)–C(1)
C(1)–C(2)
1.369(6)
(1.3687)
1.417(6)
(1.3958)
1.368(6)
(1.3777)
1.456(7)
(1.4752)
O(3)–C(19)
O(6)–C(41)
C(17)–O(1)
C(39)–O(4)
O(1)–C(1)
O(4)–C(23)
C(1)–C(2)
1.355(9)
(1.3683)
1.404(8)
(1.4090)
1.349(8)
(1.3921)
1.456(9)
(1.4748)
2.2. Mechanistic aspects of the reaction of compounds 4
with oxygen-centred radicals: EPR and DFT studies
C(23)–C(24)
First of all, our efforts were directed to get evidence for the
hypothesized formation of a transient phenoxyl radical on
reaction of the compounds 4 with oxygen-centred radicals.
At least in one case, i.e. by reacting 4i with the H2O2/Cu(II)
system, we succeeded in this. Indeed, operating at very low
copper(II) concentrations (ca. 1027 M) so as to avoid the
appearance of the signal due to paramagnetic copper(II)
centres, the EPR spectrum reported in Figure 3 was
obtained, which shows a double doublet (giso¼2.0042).
This was nicely simulated28 by best fitting to the
experimental spectrum, with the optimisation of the
hyperfine coupling constants (hfcc) and the line-widths.
Thus, it was associated with the phenoxyl radical 8 (Eq. (7))
whose spin delocalisation extends almost exclusively over
the phenolic ring, preferably over the ortho positions rather
than over the meta ones, as indicated by the hyperfine
coupling constant with ortho- and meta-hydrogen nuclei:
aH(ortho)¼5.8 G; aH(meta)¼1.5 G.
a
The data refer to the two independent molecules present in the cell, and
are given as the average over the parameters observed for the two
independent molecules.
The above hfcc assignments were substantiated by DFT
calculations of the electron spin densities starting from the
optimised molecular geometry of the phenoxyl radicals. The
molecular geometries of some derivatives of 4 and of the
corresponding phenoxyl radicals were obtained from DFT
data. Interestingly, the data reported in Table 5 show that a
very high accord exists between the bond lengths calculated
for 4a and 4b and those obtained by single-crystal X-ray
analysis (Fig. 4).
Thus, the electron spin densities at the hydrogen atoms were
calculated, and the isotropic hfcc were obtained through the
Fermi contact interaction. The spin density map for the
ð7Þ
radical 8 is shown in Figure 5. The hyperfine coupling
constants calculated for the radical 8 (aH(ortho)¼5.5 G;
aH(meta)¼1.5 G) are in a remarkable accordance with those
obtained by computer simulation of the experimental
spectrum.
The above EPR data strongly support the hypothesis that the
reaction of the compounds 4 with oxygen-centred radicals
implies first the abstraction of the hydroxylic hydrogen atom
by the reacting radical, with the consequent formation of the
corresponding phenoxyl radicals. A theoretical insight into
the nature of the reaction between compounds 4 and free
radicals came from DFT calculation of the electrostatic
potential over the entire molecule as a consequence of the
approaching of a positive charge, being the molecule
considered to be isolated and in the gas phase. This
appeared to us to be one of the best models to simulate the
reaction of compounds 4 with an electrophilic species, like a
free radical. The maps of the calculated electrostatic
potential surfaces show clearly that all the guaiazulenic
derivatives 4 exhibit very similar electronic distribution
Figure 3. Experimental (a) and computer simulated (b) X-band EPR
spectrum of the radical 8, at 58C.

5010
E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
Figure 4. Molecular geometry of 4a (a) and 4b (b).
Figure 5. Calculated spin density surface (0.002 e au23) of the radical 8.

E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
5011
Figure 6. Calculated electrostatic potential surface of 4a (the electrostatic potential decreases on going from blue to red lines).
Figure 7. Representation of the calculated transition state geometry for the reaction between 4a and the hydroxyl radical.
almost independently of the substitution on the phenolic
ring. In particular, the most positive electrostatic potential
(blue coloured iso-potential lines) is over the phenol
hydroxylic hydrogen while the most negative one (red
coloured iso-potential lines) is near the ester bond (Fig. 6).
which the phenoxyl radical 8 derived from 4i could release
2,5-di-tert-butylquinone 9, in addition to unidentified
species. Indeed, TLC of the organic extracts of the reaction
of 4i with the H2O2/Cu(II) system revealed the presence of 9
along with other unidentified products.
The DFT energy profile of the reaction between 4a and the
hydroxyl radical was calculated and the activation energy
value (Ea¼9.5 kcal mol21) and the transition state were thus
obtained. The calculated data confirms that the position of
the attack by the radical species to the phenolic group
(Fig. 7) is in a good agreement with the transition state
reported in the literature30 for the hydrogen abstraction
reaction from a generic phenol derivative by a peroxyl
radical.
According to the reactions outlined in Scheme 3, we are
strongly inclined to conclude that the transient red colour
observed when compounds 4 react with oxygen-centred
radicals might be associated with the phenoxyl radical,
although we did not succeed in obtaining a UV–vis
spectrum of any of these radicals owing to their short
lifetime. Analogously, we did not succeed in identifying the
products resulting from the radicalic degradation of
compounds 4.
As far as the evolution of the phenoxyl radicals derived from
the compounds 4 is concerned, significant information
comes from some theoretical and experimental data.
According to the calculated geometries of 4i and 8, it can
be observed that the C(1)–O(2) bond length increases and
the O(2)–C(3) bond length decreases, on going from 4i to
the corresponding phenoxyl radical 8 (Fig. 8). This, in
conjunction with the EPR data, seems to suggest the radical
fragmentation pathway shown in Scheme 3, according to
3. Conclusions
As hypothesised, the phenol derivatives 4 indeed exhibit
chromotropic behaviour on reaction with oxygen-centred
radicals, the observed colour variations being due to the
oxidative degradation of the starting compounds. Their
peroxyl radical scavenging activity is markedly influenced
by the nature of the substituents present on the phenolic ring

5012
E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
Figure 8. Comparison of some calculated bond lengths of 4i and 8.
Scheme 3.
as well as by the reaction medium and the nature of the
radical partner. In particular, the compounds 4 carrying
unsubstituted phenolic rings are less active than those
containing alkyl-substituted phenolic rings, and those
carrying at least one ortho-alkyl group result to be the
most active. As stated above, the reaction medium plays a
very important role in influencing the scavenging activity of
the compounds 4, which decreases on going from organic
solvents to micellar media. For instance, 4a is a very poor
radical scavenger in SDS micelles, while it is as active as
BHT in cumene solution. This is quite a general trend in this
field.20,23,31
it must be underlined that some of the compounds 4a–4j are
as powerful as BHT on reaction with the hydroxyl radical.
The observed antioxidant activity, at least for 4a,4c,4i and
4j, is quite interestingly in perfect accordance with the
corresponding phenolic O–H bond dissociation energy.
Indeed, this results to be higher in the case of 4a,4c and 4j,
which show moderate to low antioxidant activities, and
lower in the case of the most efficient one, i.e. 4i.
Finally, some aspects of the reaction of the compounds 4
with oxygen-centred radicals have been elucidated. In
particular, it has been shown that such compounds transfer
their hydroxylic hydrogen to the reacting oxygen-centred
radical, as already observed for most of the phenol-based
antioxidants, which results in the formation of the
corresponding phenoxyl radicals. These last, in turn,
undergo a fragmentation reaction which produces the
corresponding quinone derivative. The DFT calculation of
the transition state of the reaction between compound 4a
and the hydroxyl radical not only gives further grounding to
the hydroxylic hydrogen transfer process but also allows an
evaluation of the activation energy for such a process.
It must be underlined that, although the efficiency of
compounds 4 has resulted to be lower than that of BHT and
a-tocopherol in scavenging the typical oxygen-centred
radicals occurring in the oxidation of organic substrates,
they exhibit a markedly selective behaviour. Indeed, they
react much fast with the hydroxyl radical than with the
superoxide radical, this difference being much more marked
than that observed in the case of commonly employed
antioxidants. As for the reaction of 4a–4j with the hydroxyl
radical, the rate is scarcely influenced by structure of the
phenolic ring, with the only exception being the reaction
with the fluoro-substituted compound 4j. In this connection,
The comparison of the calculated electrostatic potential

E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
5013
surfaces of the compounds 4 and of two non-guaiazulenic
derivatives, 6 and 7, which carry a phenolic or naphthylic
moiety, respectively, instead of the guaizulenic one,
allowed us to rationalize the different behaviour exhibited
by the two classes of compounds. Interestingly, this means
that the presence of the guaiazulenic moiety not only
confers chromotropic behaviour to the derivatives 4 but also
makes their ester bond less prone to hydrolysis, according to
the experimental data. As a final remark we would underline
that several aspects of the oxidative degradation of the
compounds 4 are still obscure. Thus, our research efforts are
now directed to try to clarify these aspects.
through a 1 m long Todd column and then saturated with
argon, at 208C for 2 h.
4.2. Computational details
Full optimisation of the geometry of the radical species and
calculation of the spin density surfaces, electrostatic
potential surfaces and energies were carried out using the
UNIX version of Spartan 5.1 computer program,27 running
on a Digital/Compaq workstation, by adopting the p-BP86
scheme that uses the functional proposed by Becke29 and the
Perdew’s correlation functional, adopting the DNp p base
functions set, appropriate for isotropic hfcc calculations of
split-valence-polarisation quality.
4. Experimental
4.1. General
4.3. General procedure for the preparation of
compounds 4a–4j
A solution of oxalyl bromide (0.6 g, 2.7 mmol) in dry
diethyl ether (5 ml) was added to a solution of guaiazulene
(0.5 g, 2.5 mmol) in diethyl ether (25 ml), at 08C, under
dinitrogen atmosphere. After 10 min, the mixture was
allowed to warm at room temperature. After 2 h, 70 ml of
diethyl ether were added to the above mixture and the new
solution was added drop wise to a solution of the appropriate
dihydrobenzene derivative (5 mmol) in diethyl ether
(170 ml), at 2208C. The resulting mixture was stirred at
room temperature for 12 h. Afterwards, it was treated with
30 ml of a 15% NaHCO3 solution and with solid NaHCO3
until neutral. The aqueous phase was extracted with benzene
(2£50 ml) and the extracts were combined with the ethereal
solution. The new mixture was dried over Na2SO4 and dried
under reduced pressure. A microcrystalline solid was
obtained in all cases. The purification of crude products
was carried out by column (internal diameter, 25 mm;
length, 500 mm) chromatography of silica gel 60 (230–400
mesh) (Merck) or of aluminium oxide 90 (70–230 mesh)
(Merck) by using the appropriate eluant as specified in the
sections below.
The elemental analysis were performed by the Micro-
`
analysis Laboratory (Facolta di Farmacia, Pisa, Italy).
Melting points were determined by MEL-TEMP II
1
apparatus. H NMR spectra were run at 200 MHz by a
Varian Gemini 200 instrument or by a Varian VXR 300
instrument, using TMS as internal standard. The Ionspray
mass spectra (IS-MS) were recorded by using a mass
spectrometer Perkin–Elmer SCIEX API III plus (SCIEX
Co., Thornhill, ONT, Canada). Spectrophotometric UV–vis
measurements for the inhibition of the autoxidation of
linoleic acid were recorded on a Perkin–Elmer Lamba 9
UV/VIS spectrophotometer. Spectrophotometric VIS
measurements for the deoxyribose method were recorded
on a Jasco Uvidec 710 spectrophotometer UV/VIS. The
measurements based upon the Oxygen Up-take Method
were carried out at the Department of Organic Chemistry
‘A. Mangini’, University of Bologna (Italy), by using the
apparatus built up in that laboratory.32
All the solvents were dried or freshly distilled following
the standard procedures. All the purification procedures,
which were done by column chromatography, were
executed analysing all the fractions by TLC. All chemicals
were used as received from Aldrich unless otherwise stated.
Bis(ethylenediamine)copper(II) perchlorate was prepared as
described.8 The ethanol and the methanol used in the
spectrophotometric measurements were spectrophotometric
grade.
4.3.1. 1,4-Dimethyl-3-(4-hydroxyl-1-carbophenoxy)-7-
isopropylazulene (4a). The crude reaction product was
purified by column chromatography on silica gel using a 7/3
(v/v) diethyl ether/n-hexane mixture as the eluant. Pure 4a
(0.34 g, 40%) and the diester 5a (0.04 g, 6%) were obtained
as blue-violet solids.
1H NMR of 4a (R1¼R2¼R3¼R4¼H) (200 MHz, CDCl3,
Me4Si): d 1.37 (6H, d, J¼6.8 Hz, Ha); 2.61 (3H, s, Hd); 3.05
(3H, s, Hm); 3.12 (1H, m, J¼6.8 Hz, Hb); 4.92 (1H, s, OH);
6.95 (4H, m, R1–R4); 7.33 (1H, d, J¼11.0 Hz, Hn); 7.58
(1H, dd, J¼11.0, 2.1 Hz, Ho); 8.17 (1H, s, He); 8.27 (1H, d,
J¼2.1 Hz, Hc). Anal. calcd for C22H22O3 4a (344.41): C,
79.0; H, 6.6%. Found: C, 78.9; H, 6.6%. Mp 175–1768C.
The X-band EPR spectra were obtained on a Varian E112
spectrometer. The temperature of the samples was con-
trolled by a Varian E257 temperature control unit. The EPR
spectrometer was interfaced to a IPC 610/P566C industrial
grade Advantech computer by means of a home made data-
acquisition system consisting of an acquisition board
capable of acquiring up to 500000 12-bit samples per
second including 32-bit add to memory, thus giving on-line
signal averaging,33 and a software package specially
designed for EPR experiments.34 The EPR spectra were
run by placing the sample into quartz tubes fitted with a
quartz-Pyrex joint and a Bibby Quickfit Rotaflo PTFE tap
(Disa, Milan) or Standard TE102 Cavity Aqueous Cells
(Wilmad) and were recorded according to already
published procedures.35 Water used in the EPR experiments
(MILLIPORE MilliQ grade) was first distilled from KMnO4
nmax (KBr) 1711 (CO), 3415 (OH) cm21
.
1H NMR of 5a (R1¼R2¼R3¼R4¼H) (200 MHz, CDCl3,
Me4Si): d 1.37 (12H, d, J¼6.8 Hz, Ha,); 2.61 (6H, s, Hd);
3.05 (6H, s, Hm); 3.13 (2H, m, J¼6.8 Hz, Hb); 7.30 (4H, s,
R1–R4); 7.36 (2H, s, Hn); 7.57 (2H, dd, J¼11.0, 2.1 Hz, Ho);
8.19 (2H, s, He); 8.28 (2H, d, J¼2.1 Hz, Hc) ppm. Anal. calcd
for C38H38O4 5a (558.72): C, 81.7; H, 6.9%. Found: C, 81.8;
H, 6.9%. Mp.2508C (dec.). nmax (KBr) 1720 (CO) cm21
.

5014
E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
4.3.2. 1,4-Dimethyl-3-(3-hydroxyl-1-carbophenoxy)-7-
isopropylazulene (4b). The crude product was purified by
column chromatography on silica gel using a 9/1 (v/v)
benzene/ethyl acetate mixture as the eluant. Pure 4b (0.36 g,
43%) and the diester 5b (ca. 4%) were obtained as blue-
violet solids.
H, 6.8%. Found: C, 79.4; H, 6.7%. Mp 146–1478C. nmax
(KBr) 1657 and 1729 (CO), 3361 (OH) cm21
.
1H NMR of 4e (R1¼CH3, R2¼R3¼R4¼H) (200 MHz,
CDCl3, Me4Si): d 1.12 (6H, d, J¼7.0 Hz, Ha); 2.19 (3H, s,
R1); 2.42 (3H, s, Hd); 2.69 (1H, m, J¼7.0 Hz, Hb); 3.12 (3H,
s, Hm); 4.82 (1H, s, OH); 6.47 (1H, d, J¼11 Hz, Ho); 6.82–
7.13 (4H, m, Hn, R2–R4); 8.11 (1H, s, Hc); 8.40 (1H, s, He)
ppm. Anal. calcd for C23H24O3 4e (348.44): C, 79.31; H,
6.8%. Found: C, 79.4; H, 6.8%. Mp 136–1378C. nmax (KBr)
1H NMR of 4b (200 MHz, CDCl3, Me4Si): d 1.10 (6H, d,
J¼6.8 Hz, Ha); 2.38 (3H, s, Hd); 2.72 (1H, m, J¼6.8 Hz,
Hb); 3.07 (3H, s, Hm); 4.79 (1H, s, OH); 6.50 (1H, m, Hl);
6.78–7.10 (5H, m, Hf, Hg, Hh, Hn, Ho); 8.11 (1H, d, J¼2 Hz,
Hc); 8.30 (1H, s, He) ppm. Anal. calcd for C22H22O3 4b
(334.41): C, 78.8; H, 6.6%. Found: C, 79.0; H, 6.6%. Mp
126–1278C. nmax (KBr) 1678 and 1712 (CO), 3384 (OH)
1682 (CO), 3385 (OH) cm21
.
4.3.5. 1,4-Dimethyl-3-(4-hydroxyl-3-tert-butyl-1-carbo-
phenoxy)-7-isopropyl-azulene (4f) and 1,4-dimethyl-3-
(4-hydroxyl-2-tert-butyl-1-carbophenoxy)-7-isopropyl-
azulene (4g). Both these compounds were purified by
column chromatography on silica gel using a 9/1 (v/v)
benzene/ethyl acetate mixture as the eluant. The desired
monoesters 4f (0.3 g, 30%) and 4 g (0.1 g, 10%) and the
diester 5f (ca. 5%) were obtained as blue-violet solids.
cm21
.
1H NMR of 5b (200 MHz, CDCl3, Me4Si): d 1.15 (12H, d,
J¼6.7 Hz, Ha); 2.42 (6H, s, Hd); 2.72 (2H, m, J¼6.7 Hz,
Hb); 3.10 (6H, s, Hm); 6.82–7.25 (8H, m, Hf, Hg, Hh, Hn,
Ho); 8.13 (2H, d, J¼2.1 Hz, Hc); 8.32 (2H, s, He) ppm. Anal.
calcd for C38H38O4 5b (558.72): C, 81.7; H, 6.9%. Found:
C, 81.7; H, 6.8%. Mp.2508C (dec.). nmax (KBr) 1716 (CO)
1H NMR of 4f (R1¼R3¼R4¼H, R2¼But) (200 MHz,
CDCl3, Me4Si): d 1.16 (6H, d, J¼7 Hz, Ha); 1.49 (9H, s,
R2); 2.45 (3H, s, Hd); 2.75 (1H, m, J¼7 Hz, Hb); 3.17 (3H, s,
Hm); 4.17 (1H, s, OH); 6.10 (1H, d, J¼8.4 Hz, R3); 6.90
(1H, d, J¼11 Hz, Hn); 7.12 (1H, dd, J¼8.4, 2.6 Hz, R4);
7.40 (1H, d, J¼2.6 Hz, R1); 8.18 (1H, d, J¼2 Hz, Hc); 8.42
(1H, s, He) ppm. Anal. calcd for C26H30O3 4f (390.52): C,
79.9; H, 7.7%. Found: C, 80.0; H, 7.7%. Mp 158–1598C.
cm21
.
4.3.3. 1,4-Dimethyl-3-(2-hydroxyl-1-carbophenoxy)-7-
isopropylazulene (4c). The crude product was purified
by column chromatography on silica gel using a 9/1
(v/v) benzene/ethyl acetate mixture. 4c (0.4 g, 48%) and
the diester 5c (ca. 4%) were obtained as a blue-violet
solids.
nmax (KBr) 1697 (CO), 3439 (OH) cm21
.
1H NMR of 4c (200 MHz, CDCl3, Me4Si): d 1.11 (6H, d,
J¼6.8 Hz, Ha); 2.35 (3H, s, Hd); 2.70 (1H, m, J¼6.8 Hz,
Hb); 3.05 (3H, s, Hm); 6.05 (1H, s, Hl); 6.7–7.3 (6H, m, Hf,
Hg, Hh, Hi, Hn, Ho); 8.10 (1H, d, J¼2.1 Hz, Hc); 8.23 (1H, s,
He) ppm. Anal. calcd for C22H22O3 4c (344.41): C, 79.0; H,
6.6%. Found: C, 78.8; H, 6.6%. Mp 105–1068C. nmax (KBr)
1H NMR of 4g (R1¼But, R2¼R3¼R4¼H) (200 MHz,
CDCl3, Me4Si): d 1.16 (6H, d, J¼7 Hz, Ha); 1.50 (9H, s,
R1); 2.40 (3H, s, Hd); 2.70 (1H, m, J¼7 Hz, Hb); 3.15 (3H, s,
Hm); 4.32 (1H, s, OH); 6.40 (1H, dd, J¼8.7, 3.0 Hz, R2);
6.80 (1H, d, J¼3 Hz, R3); 6.90 (1H, d, J¼11 Hz, Hn); 8.10
(1H, d, J¼2 Hz, Hc); 8.45 (1H, s, He) ppm. Anal. calcd for
C26H30O3 4g (390.52): C, 79.9; H, 7.7%. Found: C, 80.1; H,
7.7%. Mp 172–1738C. nmax (KBr) 1698 (CO), 3447 (OH)
1685 (CO), 3379 (OH) cm21
.
1H NMR of 5c (200 MHz, CDCl3, Me4Si): d 1.24 (12H, d,
J¼6.9 Hz, Ha); 2.54 (6H, s, Hd); 2.71 (6H, s, Hm); 2.95 (2H,
m, J¼6.9 Hz, Hb); 6.68 (4H, m, Hf, Hg); 6.90 (2H, d,
J¼11.0 Hz, Hn); 7.30 (2H, dd, J¼11.0, 2.1 Hz, Ho); 7.5
(2H, s, He); 8.08 (2H, d, J¼2.1 Hz, Hc) ppm. Anal. calcd
for C38H38O4 5c (558.72): C, 81.7; H, 6.9%. Found: C,
81.8; H, 6.9%. Mp.2508C (dec.). nmax (KBr) 1712 (CO)
cm21
.
1H NMR of 5f (R1¼R3¼R4¼H, R2¼But) (200 MHz,
CDCl3, Me4Si): d 1.12 (12H, d, J¼7 Hz, Ha); 1.47 (9H, s,
R2); 2.42 (6H, s, Hd); 2.70 (2H, m, J¼7 Hz, Hb); 3.16 (6H, s,
Hm); 6.93 (2H, d, J¼11 Hz, Hn), 6.95–7.60 (4H, m, R1, R3,
R4, Ho); 8.13 (2H, d, J¼2 Hz, Hc); 8.45 (2H, s, He) ppm.
Anal. calcd for C42H46O4 5f (614.82): C, 82.0; H, 7.5%.
Found: C, 81.8; H, 7.6%. Mp.2508C (dec.). nmax (KBr)
cm21
.
4.3.4. 1,4-Dimethyl-3-(3-methyl-4-hydroxyl-1-carbo-
phenoxy)-7-isopropyl-azulene (4d) and 1,4-dimethyl-3-
(2-methyl-4-hydroxyl-1-carbophenoxy)-7-isopropyl-
azulene (4e). Both these compounds were purified by
column chromatography on silica gel using a 9/1 (v/v)
benzene/ethyl acetate mixture as the eluant. 4d (0.27 g,
31%) and 4e (0.24 g, 28%) were obtained as blue-violet
solids.
1715 (CO) cm21
.
4.3.6. 1,4-Dimethyl-3-(2,3-dimethyl-4-hydroxyl-1-carbo-
phenoxy)-7-iso-propylazulene (4h). The crude product
was dissolved in the minimum amount of n-hexane and
purified by column chromatography on Alumina. The
column was eluted first with n-hexane, which allowed the
separation of the unreacted guaiazulene, then with a 1/1
(v/v) benzene/ethyl acetate mixture, which yielded the
diester 5 h as blue crystals (0.160 g, 22%), then with
benzene, which allowed to separate the unreacted hydro-
quinone and, finally, with ethyl formate, which yielded 4 h
as a crystalline blue solid (0.362 g, 27%).
1H NMR of 4d (R1¼R3¼R4¼H, R2¼CH3) (200 MHz,
CDCl3, Me4Si): d 1.11 (6H, d, J¼7.0 Hz, Ha); 2.03 (3H, s,
R2); 2.42 (3H, s, Hd); 2.70 (1H, m, J¼7.0 Hz, Hb); 3.12 (3H,
s, Hm); 4.39 (1H, s, OH); 6.37 (1H, d, J¼11 Hz, Ho); 6.85–
7.15 (4H, m, Hn, R1, R3, and R4); 8.12 (1H, s, Hc); 8.39 (1H,
s, He) ppm. Anal. calcd for C23H24O3 4d (348.44): C, 79.31;
1H NMR of 4h (R1¼R2¼CH3, R3¼R4¼H) (200 MHz,

E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
5015
CDCl3, Me4Si): d 1.37 (6H, d, J¼6.8 Hz, Ha); 2.17 (3H, s,
R2); 2.18 (3H, s, R1); 2.62 (3H, s, Hd); 3.10 (3H, s, Hm); 3.12
(1H, m, J¼6.8 Hz, Hb); 4.69 (1H, s, OH); 6.65 (1H, d,
J¼8.9 Hz, R3); 6.95 (1H, d, J¼8.9 Hz, R4); 7.33 (1H, d,
J¼11.0 Hz, Hn); 7.57 (1H, dd, J¼11.0, 2.1 Hz, Ho); 8.23
(1H, s, He); 8.28 (1H, d, J¼2.1 Hz, Hc) ppm. Anal. calcd for
C24H26O3 4h (362.47): C, 79.5; H, 7.2%. Found: C, 79.5; H,
7.2%. Mp 163–1648C. nmax (KBr) 1690 (CO), 3424 (OH)
Ho); 8.24 (1H, s, He); 8.29 (1H, d, J¼2.0 Hz, Hc). 19F NMR
of 4j (C6D6, 188.135 MHz, d from CFCl3): 2164.3 (2F, d,
J¼17 Hz, R2 and R3); 2156.4 (2F, d, J¼17 Hz, R1 and R4)
ppm. Anal. calcd for C22H18F4O3 4j (406.38): C, 65.0; H,
4.4%. Found: C, 65.1; H, 4.4%. Mp 140–1418C. nmax (KBr)
1698 (CO), 3291 (OH) cm21
.
1H NMR of 5j (R1¼R2¼R3¼R4¼F) (200 MHz, CDCl3,
Me4Si): d 1.10 (12H, d, J¼6.9 Hz, Ha); 2.33 (6H, s, Hd);
2.69 (2H, m, J¼6.9 Hz, Hb); 3.05 (6H, s, Hm); 6.87 (2H, d,
J¼11.0 Hz; Hn); 7.08 (2H, dd, J¼11.0, 2.0 Hz, Ho); 8.09
(2H, d, J¼2.0 Hz, Hc); 8.40 (2H, s, He) ppm. Anal. calcd for
C38H34F4O4 5j (630.68): C, 72.3; H, 5.4%. Found: C, 72.2;
cm21
.
1H NMR of 5h (R1¼R2¼CH3, R3¼R4¼H) (200 MHz,
CDCl3, Me4Si): d 1.38 (12H, d, J¼6.8 Hz, Ha); 2.24 (6H, s,
R1 and R2); 2.63 (6H, s, Hd); 3.06 (6H, s, Hm); 3.13 (2H, m,
J¼6.8 Hz, Hb); 7.10 (2H, s, R3 and R4); 7.33 (2H, d,
J¼11.0 Hz, Hn); 7.58 (2H, dd, J¼11.0, 2.1 Hz, Ho); 8.23
(2H, s, He); 8.28 (2H, d, J¼2.1 Hz, Hc) ppm. Anal. calcd for
C40H42O4 5h (586.77): C, 81.8; H, 7.2%. Found: C, 81.9; H,
H, 5.4%. Mp 135–1368C. nmax (KBr) 1698 (CO) cm21
.
4.3.9. 4-Hydroxyphenylbenzoate (6). Following the
reported procedure,36 1.72 ml (15 mmol) of benzoyl
chloride were added to a magnetically stirred solution of
hydroquinone (1.5 g, 13.6 mmol), under a dry dinitrogen
atmosphere. The pH of the mixture was maintained at 8.6 by
a 10% NaOH solution while adding benzoyl chloride, in ca.
40 min. The mixture was then extracted with benzene
(3£40 ml). The combined extracts were dried over Na2SO4
and then evaporated to dryness under reduced pressure. The
resulting solid residue was dissolved CHCl3 (10 ml) and
purified by column chromatography on Silica gel, using
CHCl3 as the eluant. The desired monoester 6 (0.306 g,
10%) was obtained as a white microcrystalline solid from
the third band eluted. Mp 162–1638C.36
7.2%. Mp.2508C (dec.). nmax (KBr) 1715 (CO) cm21
.
4.3.7. 1,4-Dimethyl-3-(2,5-di-tert-butyl-4-hydroxyl-1-
carbophenoxy)-7-iso-propylazulene (4i). The crude
reaction product was dissolved in the minimum amount of
n-hexane and purified by column chromatography on
aluminium oxide. The column was eluted first with
n-hexane, which allowed the separation of unreacted
guaiazulene, then with a 1/1 (v/v) n-hexane/diethyl ether
mixture, which yielded the diester 5i as microcrystalline
blue solid (0.192 g, 23%) and, finally, with a 1/1 (v/v) ethyl
acetate/diethyl ether mixture, which yielded 4i, as a
microcrystalline blue solid (0.383 g, 35%).
4.3.10. 4-Hydroxyphenyl-1-naphthoate (7). A solution of
1-naphthoyl chloride (0.377 ml, 2.5 mmol) in 100 ml of
diethyl ether was added dropwise to a magnetically stirred
solution of hydroquinone (0.55 g, 5 mmol) in diethyl ether
(200 ml). After 10 min, the mixture was allowed to warm to
room temperature. After 1 h, the reaction mixture was
treated with triethylamine (0.348 ml, 2.5 mmol) mixture
and maintained at room temperature for 48 h. The mixture
was then extracted with benzene (100 ml). The organic
extract was washed with water (3£100 ml), dried over
Na2SO4 and then kept to dryness under reduced pressure.
A microcrystalline solid was obtained which was dissolved
in 8 ml of a 6/4 (v/v) CHCl3/ethyl acetate mixture.
Chromatography on Silica gel, using a 6/4 (v/v) CHCl3/
ethyl acetate mixture as the eluant, furnished 7 (0.24 g,
36%) (first band).
1H NMR of 4i (R1¼R3¼H; R2¼R4¼But) (200 MHz,
CDCl3, Me4Si): d 1.30–1.33 (24H, m, Ha, R2, R4); 2.56
(3H, s, Hd); 2.97 (3H, s, Hm); 3.08 (1H, m, J¼6.8 Hz, Hb);
4.72 (1H, s, OH); 6.63 (1H, s, R3); 6.84 (1H, s, R1); 7.25
(1H, d, J¼11.0 Hz, Hn); 7.51 (1H, dd, J¼11.0, 2.1 Hz, Ho);
8.17 (1H, s, He); 8.23 (1H, d, J¼2.1 Hz, Hc) ppm. Anal.
calcd for C30H38O3 4i (446.63): C, 80.8; H, 8.5%. Found: C,
80.7; H, 8.5%. EI-MS m/z: 446 [M]þ. Mp 109–1108C. nmax
(KBr) 1687 (CO), 3426 (OH) cm21
.
1H NMR of 5i (R1¼R3¼H; R2¼R4¼But) (200 MHz,
CDCl3, Me4Si): d 1.42–1.58 (30H, m, Ha, R2, and R4);
2.73 (6H, s, Hm); 3.15 (6H, s, Hd); 3.22 (2H, m, J¼6.8 Hz,
Hb); 7.33 (2H, s, R1 and R3); 7.42 (2H, d, J¼11.0 Hz, Hn);
7.67 (2H, dd, J¼11.0, 2.1 Hz, Ho); 8.33 (2H, s, He); 8.39
(2H, d, J¼2.1 Hz, Hc) ppm. Anal. calcd for C46H54O4 5i
(670.93): C, 82.3; H, 8.1%. Found: C, 82.3; H, 8.1%. IS-MS
m/z: 688 [MþNH4]þ and 671 [MþH]þ. Mp.2508C (dec.).
1H NMR of 7 (200 MHz, CDCl3, Me4Si): d 5.06 (1H, s, Hl);
0
0
6.80–7.18 (4H, m, Hh, Hh , Hi, Hi ); 7.48–7.65 (3H, m, Hb,
He, Hf); 7.90 (1H, d, J¼8.6 Hz, Hd); 8.09 (1H, d, J¼7.4 Hz,
Hc); 8.44 (1H, dd, J¼7.3, 1.5 Hz, Ha); 8.99 (1H, d,
J¼8.4 Hz, Hg) ppm. Anal. calcd for C17H12O3 7 (264.28):
C, 77.3; H, 4.5%. Found: C, 77.3; H, 4.4%. Mp 143–1448C.
nmax (KBr) 1720 (CO) cm21
.
4.3.8. 1,4-Dimethyl-3-(2,3,5,6-tetrafluoro-4-hydroxyl-1-
carbophenoxy)-7-iso-propylazulene (4j). The crude
reaction product was dissolved in the minimum amount of
n-hexane and purified by column chromatography on
aluminium oxide, using a 9/1 (v/v) benzene/ethyl formate
mixture. 4j (0.520 g, 51%) and the diester 5j (ca. 5%) were
obtained as blue-violet solids.
nmax (KBr) 1697 (CO), 3411 (OH) cm21
.
4.4. Reaction of compounds 4a–4j with H2O2 in the
presence of (Cu(en)2((ClO4)2
0.5 ml of H2O2 (2 M, pH 4) were introduced in a test-tube
and treated with a solution of the appropriate compound 4
(60 ml) of a 0.01 M solution in sodium decyl sulfate (1 M,
pH 9) and then with 50 ml of a 0.01 M solution of
(Cu(en)2((ClO4(2 in water, at room temperature. In all
1H NMR of 4j (R1¼R2¼R3¼R4¼F) (200 MHz, CDCl3,
Me4Si): d 1.35 (6H, d, J¼7 Hz, Ha); 2.61 (3H, s, Hd); 3.04
(3H, s, Hm); 3.15 (1H, m, J¼7 Hz, Hb); 6.10 (1H, s, OH);
7.42 (1H, d, J¼11.0 Hz, Hn); 7.63 (1H, dd, J¼11.0, 2.0 Hz,

5016
E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
cases the mixture became yellow in colour within ca.
15 min. All the mixtures were treated with ethyl formate
(0.5 ml) and shaken for a few seconds. The organic layers
were then tested by TLC (SiO2, using benzene/ethyl acetate
mixtures in different ratios, depending upon the reactive
system).
4.8. Inhibition of the degradation of deoxyribose
0.2 ml of a solution of FeSO4·7H2O (1022 M) and EDTA
(1022 M) in sodium n-dodecyl sulfate (0.1 M), 0.2 ml of
2-deoxyribose solution (1022 M) in sodium n-n-dodecyl
sulfate (0.1 M), 0.4 ml of phosphate buffer solution (0.1 M,
pH 7.4) in sodium n-dodecyl sulfate (0.1 M) and 1 ml of
antioxidant solution (1023 M) in methanol were added to a
screw-capped test tube. Finally, 0.2 ml of a H2O2 water
solution (1022 M) were added and the whole was incubated
at 378C for 4 h. After this incubation, 1 ml of a
trichloroacetic acid solution (2.8%) in sodium n-dodecyl
sulfate (0.1 M) and 1 ml of thiobarbituric acid solution (1%)
in NaOH (5£1022 M) in sodium n-dodecyl sulfate (0.1 M)
were added and the resulting mixture was boiled for 45 min,
then cooled in ice and its absorbance measured at 532 nm.
Each run was repeated three times.
4.5. Reaction of compounds 4a–4j with KO2
These reactions were carried out under a dry dinitrogen
atmosphere. 0.5 ml of a solution of KO2 in DMSO (0.1 M)
containing 18-crown-6 (0.1 M) were introduced into a
test-tube and then treated with 10 ml of a 0.007 M solution
of the appropriate compound 4 in DMSO. After 5 min, the
reaction mixture was treated with 3 ml of H2O, 1 ml of ethyl
formate and shaken for a few seconds. The organic layer
was then tested by TLC (SiO2, using benzene/ethyl acetate
mixtures in different ratios, depending upon the reactive
system).
4.9. Single crystal X-ray structure determination of 1,4-
dimethyl-3-(4-hydroxyl-1-carbophenoxy)-7-isopropyl-
azulene (4a) and 1,4-dimethyl-3-(3-hydroxyl-1-carbo-
phenoxy)-7-isopropylazulene (4b)
4.6. Inhibition of the autoxidation of linoleic acid
The experiments were carried out according the procedure
reported in literature.12
The diffractometric measurements were carried out at room
temperature by means of a Bruker P4 diffractometer
equipped with a graphite-monochromated Mo Ka radiation
Preparation of the solutions. Stock solutions of phosphate
buffer (0.3 M, pH 7.4) were prepared and stored in
plastic bottle0s for no more than 2 weeks. The solution of
0.5 M 2,2 -azobis(2-amidinopropane)dihydrochloride
(ABAP) was freshly prepared in 0.05 M phosphate
buffer (pH 7.4). The stock solution of linoleic acid in
sodium n-dodecyl sulfate was prepared as follows: 0.692 g
of sodium n-dodecyl sulfate (SDS) and 20 ml of linoleic
acid were dissolved in 20 ml of deareated water under
nitrogen atmosphere. The solution was stored in a Carius
tube at 2188C. The stock solutions of antioxidants were
freshly prepared in absolute ethanol at concentrations of
1023 M.
˚
(l¼0.71073 A). All data were collected using an v/2u scan
mode, and three standard reflections were monitored every
97 measurements for checking the crystal decay and
equipment stability. Data reduction was done by the
SHELXTL program. All refinement cycles were done by
full-matrix least-squares procedures based on F2.
Single crystals of 4a are violet-blue monoclinic tables. One
of them with dimensions 0.68£0.48£0.06 mm3 was glued at
the end of a glass fibre. The cell parameters listed in Table 1
were obtained from the setting angles of 42 strong
reflections. The upper u limit of collection was restricted
to 21.28 because behind this value the fraction of observed
reflections was very low. 2848 reflections, having
21#h#5, 237#k#1, and 28#l#8, were collected.
After correction for Lorentz and polarization effects the
equivalent reflections were merged giving a set of 2018
Autoxidation procedure. In a typical run, the UV cuvette
containing 2 ml of linoleic acid in SDS solution and 0.4 ml
of phosphate buffer (0.3 M; pH 7.4) was thermostated at
37^0.28C and stirred continuously in the sample compart-
ment of the spectrophotometer. After about 20 min, when
the thermal equilibration in the spectrophotometer was
completed, an aliquot (10^0.1 ml) of ABAP 0.25 M was
added and incubated for an additional 40 min in order to
allow the rate of autoxidation to become constant. An
aliquot (10^0.1 ml) of antioxidant solution was then added
and the change in absorbance was monitored with time at
234 nm.
P
independent
reflection
(Rint¼[ lF2o2Fo2(mean)l/
P
(F2o)]¼0.0497). The structure solution was obtained by
means of direct methods by using SHELXS-97 program.
The hydrogen atoms were introduced in calculated positions
and let to ride on the connected carbon atoms. The final
refinement cycle gave the reliability factors listed in Table 6.
A crystal with dimensions 0.48£0.20£0.05 mm of a deep
red crystal of 4b was glued at the end a glass fibre. The unit
cell parameters together with the space group and some
structural details are listed in Table 6. The intensity data of
5769 reflections having 2.28,u,22.50 were collected.
After correction for Lorentz, polarization the equivalent
reflections were merged giving a total of 4640 independent
reflection (Rint¼0.0972). The structure was solved by
standard Patterson and Fourier methods by means of
SHELXS-97 program. The final refinement cycles gave
the reliability factors listed in Table 6.
4.7. Inhibition of the autoxidation of cumene
4 ml of cumene and 10 ml of a 2£1023 M benzene solution
of the appropriate radical scavenger were added to the
sample flask; 4 ml of cumene and 20 ml of a 1022 M
benzene solution of a-tocopherol were added to the
reference flask. Both the flasks were thermostated at 508C.
When the thermal equilibration was completed, an aliquot
(40^0.1 ml) of a 0.34 M solution of AIBN in acetonitrile
was added to both the flasks and the oxygen up-take was
recorded.
Complete crystallographic data for 4a and 4b been

E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
5017
PS 24. (c) Franchi, E.; Ingrosso, G.; Pinzino, C. 18 Sigma
Aldrich Young Chemists Symposium, SAYCS Riccione, 18–19
Ottobre 2001, Abstracts p PO 49.
Table 6. Crystal data and structure refinement
Compound
4a
4b
Empirical formula
Formula weight
Temperature (K)
Crystal system
Space group
C22H22O3
334.40
293(2)
Monoclinic
P21/n (No.14)
C22H22O3
334.40
293(2)
Monoclinic
P21/c (No.14)
7. Chiaradonna, G.; Cicogna, F.; Ingrosso, G.; Pinzino, C.;
Franchi, E.; Del Duca, V.; Scialla, S. European Patent EP
1094057 A1, Date of publication 25.04.2001; International
Patent Classification: C07C 69/753; International Publication
Number: WO 01/28973 A1 (International Publication Date
26.04.2001).
Unit cell dimensions
˚
a (A)
˚
5.873(1)
37.216(9)
8.458(1)
94.28(1)
1843.5(6)
4
27.275(5)
11.214(3)
11.826(2)
101.33(1)
3546.6(13)
8
1.253
0.082
4640/0/453
0.0803
0.1739
b (A)
˚
c (A)
b (8)
Volume (A )
8. Ozawa, T.; Hanaki, A. J. Chem. Soc., Chem. Commun. 1991,
330.
3
˚
9. Catalytic Oxidation with Hydrogen Peroxide as Oxidant;
Strukul, G., Ed.; Kluwer Academic: Dordrecht, 1992.
10. Kim, S.; Di Cosimo, R.; San Filippo, J. J. Anal. Chem. 1979,
51, 679–681.
Z
r
calc (Mg m23
)
1.205
0.079
m (mm21
)
Data/restraints/parameters
R(Fo) [I.2s(I)]
2018/0/235
0.0760
0.1888
11. Dannacher, J.; Schlenker, W. Text. Chem. Color. 1996, 28,
24–28.
Rw (F2o) [I.2s(I)]
P
P
P
P
R(Fo)¼ llFol2lFcll/ lFol; Rw(F2o)¼[ [w(Fo22Fc2)2]/ [w(Fo2)2]]1/2; w¼
12. Pryor, W. A.; Cornicelli, J. A.; Devall, L. J.; Tait, B.; Trivedi,
B. K.; Witiak, D. T.; Wu, M. J. Org. Chem. 1993, 58,
3521–3532.
¼1/[s 2(F2o)þ(AQ)2þBQ] where Q¼[MAX(Fo2,0)þ2F2c]/3.
13. Dangles, O.; Dufour, C.; Fargeix, G. J. Chem. Soc., Perkin
Trans. 2 2000, 1215–1222.
deposited, in the form of CIF files, with the Cambridge
Crystallographic Data Centre (Dep. No. CCCD 188035 and
CCCD 188036, respectively for 4a and 4b).
14. Howard, J. L.; Ingold, K. U. Can. J. Chem. 1969, 47,
3809–3815.
15. Halliwell, B.; Gutteridge, J. M. C.; Aruoma, O. I. Anal.
Biochem. 1987, 165, 215–219.
Acknowledgements
16. Chung, S.; Osawa, T.; Kawakishi, S. Biosci. Biotechnol.
Biochem. 1997, 61, 118–123.
We want to express our gratitude to Dr G. Chiaradonna
(Procter & Gamble fellowship, University of Pisa, Italy) for
the contribution she gave to this work through the
preliminary investigation of the synthesis of some guaiazu-
lenic derivatives and the study of their reactivity towards
free radicals. We are grateful to Professor G. F. Pedulli
(University of Bologna, Italy) for the kind permission to use
the oxygen up-take apparatus and Dr R. Amorati for his
precious assistance in carrying the oxygen up-take measure-
ments. Moreover, we thank Procter & Gamble Italia S.p.A.
(Rome, Italy) and University of Pisa (Italy) for financial
support.
17. Molecular Biology of Free Radicals in Human Diseases;
Aruoma, O. I., Halliwell, B., Eds.; OICA International: Saint
Lucia, London, 1998.
18. (a) Bordwell, F. G.; Cheng, J. J. Am. Chem. Soc. 1991, 113,
1736–1743. (b) 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. (c) Brinck, T.; Haeberlein, M.;
Jonsson, M. J. Am. Chem. Soc. 1997, 119, 4239–4244.
(d) Franchi, P.; Lucarini, M.; Pedulli, G. F.; Valgimigli, L.;
Lunelli, B. J. Am. Chem. Soc. 1999, 121, 507–514.
19. Barclay, L. R. C.; Baskin, K. A.; Locke, S. J.; Vinqvist, M. R.
Can. J. Chem. 1989, 67, 1366–1369.
20. Pryor, W. A.; Strickland, T.; Church, D. F. J. Am. Chem. Soc.
1988, 110, 2224–2229.
21. (a) Burton, G. W.; Ingold, K. U. J. Am. Chem. Soc. 1981, 103,
6472–6477. (b) Denisov, E. T.; Azatin, V. V. Inhibition of
Chain Reactions; Gordon & Breach: London, 2000.
22. Amorati, R. PhD Thesis, University of Bologna, Italy, 1998–
1999.
References
1. Handbook of Synthetic Antioxidants; Packer, L., Cadenas, E.,
Eds.; Marcel Dekker: New York, 1997.
2. (a) Ingold, K. U. Chem. Rev. 1961, 61, 562–589. (b) Neri, C.
Chim. Ind. 1997, 79, 1223–1232.
23. Valgimigli, L.; Ingold, K. U.; Lusztyk, J. J. Am. Chem. Soc.
1996, 118, 3545–3549.
3. (a) Denisov, E. T.; Khudyakov, I. V. Chem. Rev. 1987, 87,
1313–1357. (b) Foti, M.; Ingold, K. U.; Lusztyk, J. J. Am.
Chem. Soc. 1994, 116, 9440–9447.
24. (a) Lucarini, M.; Pedulli, G. F.; Cipollone, M. J. Org. Chem.
1994, 59, 5063–5070. (b) Yamamura, T.; Suzuki, K.;
Yamaguchi, T.; Nishiyama, T. Bull. Chem. Soc. Jpn 1997,
70, 413–419. (c) Bordwell, F. G.; Zhang, X.; Satish, A. V.;
Cheng, J. J. Am. Chem. Soc. 1994, 116, 6605–6610. (d) Brink,
T.; Haeberlein, M.; Jonsson, M. J. Am. Chem. Soc. 1997, 119,
4239–4244.
4. Becker, D. A. J. Am. Chem. Soc. 1996, 118, 905–906.
5. (a) Wang, B.; Lin, Y.; Chang, J.; Wang, P. Can. J. Chem.
2000, 78, 224–232. (b) Liu, R. S. H.; Muthyala, R. S.; Wang,
X.; Asato, A. E. Org. Lett. 2000, 2, 269–271. (c) Razus, A. C.
J. Chem. Soc., Perkin Trans. 1 2000, 981–988.
25. (a) Migliavacca, E.; Carrupt, P.; Testa, B. Helv. Chim. Acta
1997, 80, 1613–1626. (b) Migliavacca, E.; Ancerewicz, J.;
Carrupt, P.; Testa, B. Helv. Chim. Acta 1998, 81, 1337–1348.
26. Denisov, E. T.; Denisova, T. G. Handbook of Antioxidants;
CRC: Cleveland, OH, 2000.
6. (a) Franchi, E.; Ingrosso, G.; Pinzino, C.; Chiaradonna, G.;
Cicogna, F.; Del Duca, V.; Scialla, S. Sixth International
Symposium on Spin Trapping, Marseille, France, 27–31
August 2000, Abstracts, p 135. (b) Franchi, E.; Ingrosso, G.;
Pinzino, C.; Chiaradonna, G. 68 Convegno Nazionale del
Gruppo Italiano di Risonanza di Spin Elettronico, GIRSE,
Urbino-Sogesta, 17–20 Settembre 2001, Atti del Convegno
27. Spartan version 5.1; Wavefunction, Inc.: 18401 Von Karman
Avenue, Suite 370 Irvine, CA 92612, USA.

5018
E. Franchi et al. / Tetrahedron 59 (2003) 5003–5018
28. Duling, D. R. J. Magn. Reson. B 1994, 104, 105–110.
29. Becke, A. D. J. Chem. Phys. 1993, 98, 1372–1377.
30. (a) Tanaka, K.; Sakai, S.; Tomiyama, S.; Nishiyama, T.;
Yamada, F. Bull. Chem. Soc. Jpn 1991, 64, 2677–2680.
(b) Tomiyama, S.; Sakai, S.; Nishiyama, T.; Yamada, F. Bull.
Chem. Soc. Jpn 1993, 66, 299–304.
33. Ambrosetti, R.; Ricci, D. Rev. Sci. Instrum. 1991, 62,
2281–2287.
34. Pinzino, C.; Forte, C. EPR-ENDOR, ICQEM-CNR Rome,
Italy, Copyright 1992–2000.
35. (a) Diversi, P.; Forte, C.; Franceschi, M.; Ingrosso, G.;
Lucherini, A.; Petri, M.; Pinzino, C. J. Chem. Soc., Chem.
Commun. 1992, 1345–1347. (b) Bruni, M.; Diversi, P.;
Ingrosso, G.; Lucherini, A.; Pinzino, C.; Raffaelli, A.
J. Chem. Soc., Dalton Trans. 1995, 1035–1041. (c) Bruni,
M.; Diversi, P.; Ingrosso, G.; Lucherini, A.; Pinzino, C. Gazz.
Chim. Ital. 1996, 126, 239–250. (d) Carano, M.; Careri, M.;
Cicogna, F.; D’Ambra, I.; Houben, J. L.; Ingrosso, G.;
Marcaccio, M.; Paolucci, F.; Pinzino, C.; Roffia, S. Organo-
metallics 2001, 20, 3478–3490. (e) Carano, M.; Cicogna, F.;
Houben, J. L.; Ingrosso, G.; M.archetti, F.; Mottier, L.;
Paolucci, F.; Pinzino, C.; Roffia, S. Inorg. Chem. 2002, 41,
3396–3409. (f) Carano, M.; Cicogna, F.; D’Ambra, I.; Gaddi,
B.; Ingrosso, G.; Marcaccio, M.; Paolucci, D.; Paolucci, F.;
Pinzino, C.; Soffia, S. Organometallics 2002, 21, 5583–5593.
36. Bredereck, H.; Heck, H. Chem. Ber. 1958, 91, 1314–1318.
31. (a) Barclay, L. R. C.; Edwards, C. D.; Mukai, K.; Egawa, Y.;
Nishi, T. J. Org. Chem. 1995, 60, 2739–2744. (b) Castle, L.;
Perkins, M. J. J. Am. Chem. Soc. 1986, 108, 6381–6382.
(c) Barclay, L. R. C.; Baskin, K. A.; Dakin, K. A.; Locke, S. J.;
Vinqvist, M. R. Can. J. Chem. 1990, 68, 2258–2269.
(d) MacPaul, P. A.; Ingold, K. U.; Lusztyk, J. J. Org. Chem.
1996, 61, 1316–1321. (e) Valgimigli, L.; Banks, J. T.; Ingold,
K. U.; Lusztyk, J. J. Am. Chem. Soc. 1995, 117, 9966–9971.
(f) Pryor, W. A.; Kaufman, M. J.; Church, D. F. J. Org. Chem.
1985, 50, 281–283. (g) Barclay, L. R. C.; Edwards, C. D.;
Vinquist, M. R. J. Am. Chem. Soc. 1999, 121, 6226–6231.
32. Pedulli, G. F.; Lucarini, M.; Pedrielli, P. In Free Radicals in
Biology and Environment; Minisci, F., Ed.; Kluwer Academic:
Dordrecht, 1997; pp 169–179.