New insights into the antioxidant activity of hydroxycinnamic acids: Synthesis and physicochemical characterization of novel halogenated derivatives

Article abstract of DOI:10.1016/j.ejmech.2008.10.027

An interdisciplinary research project was developed combining the synthesis of a series of hydroxycinnamic acid derivatives and the evaluation of their physicochemical parameters (namely redox potentials and partition coefficients), along with the corresponding antioxidant activity. A structure-property-activity relationship (SPAR) approach was then applied aiming at establishing a putative relation between the physicochemical parameters of the compounds under study and their antioxidant activity. The results gathered allow concluding that the redox potentials could contribute to the understanding of the antioxidant activity and that the presence of an electron withdrawing group (EWG) of halogen type, namely a bromo atom, in an ortho position to a phenolic group of the cinnamic scaffold does not influence the antioxidant activity. On the other hand after the introduction of this type of substituent a significant increase on the lipophilicity of cinnamic derivatives was observed, which is a feature of extreme importance in the development of novel lipophilic antioxidants. The SPAR results revealed a relation between the redox potentials and the antioxidant activity of hydroxycinnamic acids and derivatives. The data obtained operate as a positive reinforce of the tendency to use redox properties as a guideline of the rational design of this type of compounds.

Full text of DOI:10.1016/j.ejmech.2008.10.027

European Journal of Medicinal Chemistry 44 (2009) 2092–2099
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
New insights into the antioxidant activity of hydroxycinnamic acids:
Synthesis and physicochemical characterization of novel
halogenated derivatives
a
b
a
a
c
´
Alexandra Gaspar , E. Manuela Garrido , Mario Esteves , Elias Quezada , Nuno Milhazes ,
Jorge Garrido b , Fernanda Borges
,
a,
**
*
a
ˆ
´
´
UQFM/Departamento de Qu´ımica Organica, Faculdade de Farmacia, Universidade do Porto, Rua Anıbal Cunha 164, 4050-030 Porto, Portugal
b UQFM/Departamento de Engenharia Qu´ımica, Instituto Superior de Engenharia do Porto, IPP, Rua Dr. Anto´nio Bernardino de Almeida 431, 4200-072 Porto, Portugal
c UQFM/Instituto Superior de Cieˆncias de Sau´de-Norte, Gandra, PRD, Portugal
a r t i c l e i n f o
a b s t r a c t
Article history:
An interdisciplinary research project was developed combining the synthesis of a series of hydrox-
ycinnamic acid derivatives and the evaluation of their physicochemical parameters (namely redox
potentials and partition coefficients), along with the corresponding antioxidant activity. A structure–
property–activity relationship (SPAR) approach was then applied aiming at establishing a putative
relation between the physicochemical parameters of the compounds under study and their antioxidant
activity.
Received 5 June 2008
Received in revised form
6 October 2008
Accepted 10 October 2008
Available online 5 November 2008
The results gathered allow concluding that the redox potentials could contribute to the understanding of
the antioxidant activity and that the presence of an electron withdrawing group (EWG) of halogen type,
namely a bromo atom, in an ortho position to a phenolic group of the cinnamic scaffold does not
influence the antioxidant activity. On the other hand after the introduction of this type of substituent
a significant increase on the lipophilicity of cinnamic derivatives was observed, which is a feature of
extreme importance in the development of novel lipophilic antioxidants.
Keywords:
Bromohydroxycinnamic acids
Ethyl bromohydroxycinnamates
Synthesis
Antioxidant activity
Voltammetry
The SPAR results revealed a relation between the redox potentials and the antioxidant activity of
hydroxycinnamic acids and derivatives. The data obtained operate as a positive reinforce of the tendency
to use redox properties as a guideline of the rational design of this type of compounds.
Ó 2008 Elsevier Masson SAS. All rights reserved.
1. Introduction
underlying their protective action are not completely understood
yet.
Antioxidants, used to prevent or inhibit the natural phenomena
of oxidation, have a broad application in diverse industrial fields as
they have a huge importance either as industrial additives or health
agents [1,2]. Research data have revealed that they could be suitable
for preventive and/or therapeutic purposes in several diseases
related with oxidative stress (e.g. atherosclerosis, inflammatory
injury, cancer and cardiovascular diseases) [3–7]. Despite the fact
that many natural and synthetic compounds display antioxidant
activity, the list of authorized antioxidant additives still correspond
to an incredibly restricted one, due mainly to solubility or toxicity
problems [8]. Although the quest for new phenolic antioxidants is
emergent many questions regarding primarily the mechanism
Phenolic acids, particularly hydroxycinnamic acids (HCA) such
as caffeic, ferulic and sinapic acids and flavonoids have been
considered as attractive potential antioxidants due to their natural
origin [9–11]. Some results evidence that their antioxidant activity
is strongly dependent on their structural features and intrinsically
related to the presence of hydroxyl function(s) in the aromatic
structure [12–16]. This type of phenolic compounds often acts
through its radical-scavenging activity that is linked to their
hydrogen- or electron-donating ability and to the stability of the
resulting phenoxyl radicals [17–19]. However, other mechanisms of
action have been suggested such as chelation of transition metals,
like copper or iron, which are well-known catalysts of oxidative
stress [20,21]. The knowledge gathered from this type of mecha-
nistic studies could lead in a near future to the rational design of
new and more effective antioxidants. Generally, phenolic acids
work well in aqueous media being their hydrophilic nature
a restriction for lipophilic systems protection (e.g. food or cosmetic
* Corresponding author.
** Corresponding author. Tel.: þ351 222078900; fax: þ351 222003977.
E-mail addresses: jjg@isep.ipp.pt (F. Garrido), fborges@ff.up.pt (F. Borges).
0223-5234/$ – see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2008.10.027

A. Gaspar et al. / European Journal of Medicinal Chemistry 44 (2009) 2092–2099
2093
matrix or biological membranes) due mainly to the difficulty of
their incorporation into fat and oil matrices [22].
and synthetic antioxidants mainly to get insight their mechanism
[27,28].
An interdisciplinary project was developed aiming a more reliable
understanding of the structure–property–activity relationships (SPAR)
that underlie the antioxidant activity of several hydroxycinnamic acids
(Table 1). With this aim, halogenated derivatives of hydroxycinnamic
acids, such as 5-bromoferulic acid (trans-3-(5-bromo-4-hydroxy-
3-methoxyphenyl)-2-propenoic acid) (2) and 5-bromocaffeic acid
(trans-3-(5-bromo-3,4-dihydroxyphenyl)-2-propenoic acid) (3) and
the corresponding ethyl esters, ethyl 5-bromoferulate (trans-ethyl
3-(5-bromo-4-hydroxy-3-methoxyphenyl)propenoate) (4); ethyl 5-
bromocaffeate (trans-ethyl 3-(5-bromo-3,4-dihydroxyphenyl)propenoate)
(5) were synthesized and some biological relevant physicochemical
parameters, such as redox potentials (Ep) and partition coefficients
(Log P), were evaluated.
In addition, the antioxidant activity of the synthesized
compounds was evaluated using total antioxidant capacity radical
assays (DPPHꢀ - 2,2’-diphenyl-1-picrylhydrazyl radical; ABTS - 2-
azino-bis(3-ethylbenzthiazoline)-6-sulfonic acid). Among the
different published methodologies for determining the antiradical
activity, both isolated compounds and complex mixtures, these
radical assays have been widely used due to their simplicity and
rapidity [23,24]. The antiradical properties obtained for the halo-
genated hydroxycinnamic derivatives were compared with those
acquired for the lead natural compounds (caffeic and ferulic acids)
and their corresponding ethyl esters.
Accordingly, in this work the redox potentials of the novel
synthesized compounds were measured and their electrochemical
behaviour compared with that obtained for the lead compounds
(caffeic and ferulic acids). The redox potentials of the synthesized
halogenated hydroxycinnamic derivatives at physiological pH 7.3
were obtained, at a glassy carbon working electrode, using differ-
ential pulse and cyclic voltammetry.
For caffeic and 5-bromocaffeic (3) acids and the corresponding
ethyl esters (dihydroxylated cinnamic series – see Table 1) one
well-defined anodic peak was observed at physiological pH using
differential pulse voltammetry (Fig. 1). The oxidation peaks of these
cinnamics compounds occurred at Ep ¼ þ0.183 V, Ep ¼ þ0.182 V for
the acids and Ep ¼ þ0.175 V and Ep ¼ þ0.174 V for the esters. These
waves are intrinsically related with the oxidation of catechol group.
Cyclic voltammograms were also recorded at different sweep
rates. The cyclic voltammograms obtained are characteristic of an
electrochemical reversible reaction showing only one anodic peak
and one cathodic peak on the reverse scan (Fig. 2). Linear plots of
peak current (Ip) as a function of the square root of scan rate (n)
were obtained indicating that the oxidation processes are diffusion
controlled [29]. The ratio of the anodic to cathodic peak heights
raises gradually with increasing potential scan rate from 10 to
100 mV sꢁ1 and the calculated current functions (Ipnꢁ1/2) diminish
gradually with the potential scan rate, showing the classical
behaviour of an oxidation process coupled with a slow subsequent
chemical reaction. All the results are in agreement with the liter-
ature which concerns the oxidative behaviour of the lead
compounds [30,31]. In fact, electrochemical studies on the caffeic
acid oxidation mechanism have shown that the first oxidation step
involves two electrons per molecule which likely correspond to the
formation of the caffeic acid o-quinone, an intermediate that is
quickly decomposed at pH higher than 7.4 [30–32].
The differential pulse voltammetric study of mono-
hydroxycinnamic acids, ferulic and 5-bromoferulic (2) acids
(Table 1), revealed the presence of two convolved anodic peaks at
physiological pH (Fig. 3). These oxidation peaks at Ep ¼ þ0.335 V,
E0p ¼ þ0.447 V and Ep ¼ þ0.335 V, E0p ¼ þ0.442 V, respectively, are
related with the oxidation of the phenolic group present in the
structures. The results may be interpreted by assuming that the
oxidation of cinnamic acids takes place by electron transfer for both
free and adsorbed forms. The free form corresponds to the first
peak, Ep, while the strongly adsorbed form, which is consequently
stabilized, is oxidized at a more anodic potential, E0p. The appear-
ance of an adsorption peak, E0p, is also pointed out by other authors,
namely in the electrochemical studies involving ferulic acid and
phenol in gold and platinum electrodes [33,34].
For ethyl ferulate and 5-bromoferulate (4) only one anodic wave
can be observed at physiological pH using differential pulse vol-
tammetry (Fig. 3). These oxidation peaks at Ep ¼ þ0.368 V and
Ep ¼ þ0.365 V, respectively, are related with the oxidation of the
phenolic group present in the structure. The large wave shape
observed could be a sign of a higher superimposition of the peaks
from both free and adsorbed forms or might indicate a lesser
adsorption propensity of this molecules. From the cyclic voltam-
metric studies we can conclude that the former is more likely.
Cyclic voltammograms were recorded at different sweep rates.
The cyclic voltammograms obtained for ferulic and 5-bromoferulic
(2) acids have also shown two convolved anodic peaks. For ethyl
ferulate and 5-bromoferulate (4) only one anodic wave is seen. All
these anodic peaks appear to correspond to irreversible processes,
as no wave is observed in the reverse scan (Fig. 2). Plots of peak
2. Results and discussion
2.1. Chemistry
Cinnamic acid derivatives were synthesized straightforward
either by a Knoevenagel-type condensation or by Fisher acid catal-
ysis esterification [6,9,12–14]. The hydroxycinnamic acids and esters
were synthesized by reaction between the corresponding benzal-
dehydes, with the same aromatic substitution pattern of the final
product, and malonic acid or monoethylmalonate, using aniline as
catalyst. The appropriate substituted benzaldehydes were obtained
by bromination of vaniline (Br2 solution in acetic acid) [25] followed
by O-demethylation with boron tribromide (BBr3) (Scheme 1) [26].
These reactions led to moderate/high yields of the desired
compounds, which were identified by spectroscopic techniques:
NMR (1H NMR,13C NMR), FT-IRand MS-EI. Thesynthetic strategiesof
the synthesized compounds are depicted in Scheme 1.
2.2. Electrochemical measurements
Voltammetric methods have been often applied to characterize
a diversity of natural phenolic antioxidants, including flavonoids,
Table 1
Hydroxycinnamic acids and their alkyl esters under study.
β
2
1
3
R1
COOR3
α
6
4
HO
5
R2
Hydroxycinnamic compounds
R1
R2
R3
Caffeic acid
OH
H
H
Ferulic acid
OCH3
OH
OCH3
OH
H
H
H
H
5-Bromocaffeic acid (3)
5-Bromoferulic acid (2)
Ethyl caffeate
Br
Br
H
CH2CH3
CH2CH3
CH2CH3
CH2CH3
H
Ethyl ferulate
OCH3
OH
OCH3
OCH3
H
Ethyl 5-bromocaffeate (5)
Ethyl 5-bromoferulate (4)
Sinapic acid
Br
Br
OCH3
current (Ip) as a function of the square root of scan rate (n) gave
a straight line indicating that the oxidation processes are diffusion
controlled [29]. The assembled data corroborate the results found

2094
A. Gaspar et al. / European Journal of Medicinal Chemistry 44 (2009) 2092–2099
H3CO
HO
CHO
H3CO
HO
CHO
HO
HO
CHO
1
Br
Br
c)
c)
COOCH2CH3
H3CO
HO
COOH
H3CO
HO
COOH
Br
4
HO
HO
Br
2
Br
3
COOCH2CH3
HO
HO
e)
a) Br 2, CH3COOH
b) BBr3, CH2Cl2
c) Malonic acid, Pyr, aniline
d) Ethylmalonate, Pyr, aniline
e) EtOH, H2SO4
Br
5
Scheme 1. Synthetic strategies used for the obtention of bromocinnamic acids and derivatives.
in literature regarding ferulic acid oxidative behaviour. The
proposed mechanism involves a one-electron transfer from the
phenolate ion followed by one irreversible dimerisation process
due to a radical–radical coupling reaction between two phenoxyl
radicals [30,33].
The voltammetric data obtained for hydroxycinnamic acids
under study appear to be in agreement with the general mecha-
nism proposed for the oxidation of a phenolic group [35]. The
results indicated that the higher the number of hydroxyl substit-
uents on the aromatic ring, the lower the electrochemical potential.
The shift of redox potential values indicates an increase in the
nucleophilicity of the compound and that its antioxidant activity is
thermodynamically favored.
The presence of a catechol moiety in cinnamic derivatives is
known to cause a decrease of the redox potential due to additional
resonance stabilization and o-quinone formation. This fact seems to
explain that redox potential of caffeic acid derivatives are lower
(and consequently antioxidative efficiency are greater) than those
of ferulic acid derivatives. Actually, the methoxylation of the meta-
phenolic group of the cinnamic derivatives (ferulic acid derivatives)
increases twice the peak potential.
The introduction of an alkyl ester group in the conjugated side
chain or a bromide group in position-5 did not significantly change
the oxidation potentials, when compared to the lead compounds.
The data obtained strongly suggest that alkyl ester group have
modest or even no effect on the electron density of the phenol or
catechol ring. Hence, the introduction of an alkyl group did not
significantly influence the energetic of the electron transfer as can
be ascribed from the similarity of the Ep values for these derivatives.
With regard to the aromatic substitution it’s generally accepted that
the introduction of groups that could have a positive influence on
the formation and lifetime, through a stabilizing effect, of the cor-
responding phenoxyl radical would significantly increase the
antioxidant activity of the compounds. Therefore, it would be
expected that the introduction of a bromide group in position-5 of
the phenol or catechol group could affect the formation and/or
stabilization of the radical intermediate causing a decrease of the
oxidation potential of the lead compounds. The results had showed
that this tendency had not occurred, either in free acids or their
corresponding esters.
250 nA
In order to verify this assumption, the voltammetric behaviour
of sinapic acid (3,5-dimethoxy-4-hydroxycinnamic acid) was also
studied at a glassy carbon working electrode using differential
pulse voltammetry. Sinapic acid presents two convolved anodic
peaks, Ep ¼ þ0.188 V, E0p ¼ þ0.295 V, at physiological pH, using
differential pulse voltammetry (data not shown). As expected, the
presence of a methoxyl group in position-5 causes a cathodic shift
on the oxidation potential, with respect to ferulic acid. This lower
potential value results from the increased stabilization of the
phenoxyl radical originated by the electron-donating effect of the
methoxyl group. So, the similarity of the anodic potentials between
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Ep / V vs Ag/AgCl
Fig. 1. Differential pulse voltammograms for 0.1 mM solutions of (d) caffeic acid, (——)
5-bromocaffeic acid, (.) ethyl caffeate and (-$-$-) ethyl 5-bromocaffeate, in physio-
logical pH 7.3 supporting electrolyte. Scan rate: 5 mV sꢁ1
.

A. Gaspar et al. / European Journal of Medicinal Chemistry 44 (2009) 2092–2099
2095
200 nA
200 nA
a
b
-0.2
0.0
0.2
0.2
0.2
0.2
0.4
0.4
0.4
0.4
0.6
0.6
0.6
0.6
0.8
0.8
0.8
0.8
1.0
1.0
1.0
1.0
1.2
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
200 nA
500 nA
c
d
-0.2
0.0
1.2 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
100 nA
250 nA
e
f
-0.2
0.0
1.2
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
200 nA
100 nA
g
h
-0.2
0.0
1.2
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Ep / V vs Ag/AgCl
Ep / V vs Ag/AgCl
Fig. 2. Cyclic voltammograms for 0.1 mM solutions of (a) caffeic acid, (b) 5-bromocaffeic acid, (c) ethyl caffeate, (d) ethyl 5-bromocaffeate, (e) ferulic acid, (f) 5-bromoferulic acid, (g)
ethyl ferulate and (h) ethyl 5-bromoferulate, in physiological pH 7.3 supporting electrolyte. Scan rate: 50 mV sꢁ1
.
5-bromoferulic acid (2) and ferulic acid could be interpreted
considering the electronic nature differences between the bromide
and methoxyl groups. In fact, although these two substituents are
group. This small change in substitution pattern could be respon-
sible for the diminishing of the stabilization of the phenoxyl radical
and as a result to attain comparable anodic potentials. Further
studies were designed using other halogen substituents in order to
s
-acceptors, bromide group is a weaker p-donor than methoxyl

2096
A. Gaspar et al. / European Journal of Medicinal Chemistry 44 (2009) 2092–2099
antioxidants. In this work the technique is based on the direct
þ
production of the blue/green ABTSꢀ chromophore through the
reaction between ABTS and potassium persulfate. Trolox, a water-
soluble vitamin E analogue, was used as reference standard in both
assays.
250 nA
Table 2 shows the scavenging effects of hydroxycinnamic acids
and derivatives on DPPH and ABTS radicals (expressed as Trolox
Equivalents (TE) at the concentration of 50 mM. The data obtained
with DPPH radical revealed that all the caffeates and ferulates
scavenged about 80 and 60% of DPPH radical, respectively.
Although few noteworthy differences among each series were
observed, the activity decreased in the following order: trolox 4
caffeic acid (4) z 5-bromocaffeic acid (3) > ethyl 5-bromocaffeate
(5) z ethyl caffeate > ferulic acid z 5-bromoferulic acid (2) > ethyl
5-bromoferulate (4) z ethyl ferulate. No significant differences
among halogenated and non-halogenated counterparts were
observed. The scavenging activity order was consistent with the
antioxidant efficacy described in the literature for the parent
compounds [9,40,41]. From the results it can be also concluded that
ethyl bromocinnamates and ethyl cinnamates present almost the
same scavenging activity that is, in both series, lower than that
displayed by their parent compounds (bromocaffeic and bromo-
ferulic acids and their non-halogenated counterparts). This
tendency is also described in the literature for related esters and
could be due to steric hindrance caused by the bulkiness of alkyl
groups [42].
The radical-scavenging ability data obtained for hydroxycin-
namic acids and derivatives against DPPH radical were in good
agreement with the expected activities of this type of phenolic
systems: higher when a catechol group is present (caffeic series)
and with a decrease of activity when the meta-hydroxyl function is
substituted by a methoxyl group (ferulic series). Important to
notice that the introduction of another methoxyl group in a ortho
position to a hydroxyl group (e.g. sinapic acid) lead to an increase of
the antioxidant activity relatively to ferulic acid (see Table 1).
On the other hand, it was demonstrated that the presence of an
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Ep / V vs Ag/AgCl
Fig. 3. Differential pulse voltammograms for 0.1 mM solutions of (d) ferulic acid,
(——) 5-bromoferulic acid, (.) ethyl ferulate and (-$-$-) ethyl 5-bromoferulate, in
physiological pH 7.3 supporting electrolyte. Scan rate: 5 mV sꢁ1
.
evaluate their ability to stabilize the phenoxyl radical. Nevertheless,
Diniz et al. [36] have shown through quantum chemistry calcula-
tions at the B3LYP theory level, together with the 6-31G* basis set,
that the O–H bond dissociation energies (BDEO–H) for homolytic
cleavage of an hydroxyl group are not significantly affected by the
presence in an ortho position of any type of halogen (chloro, bromo
or fluor). Really, this BDEO–H energy is similar to that one obtained
for a phenolic group without substituents. Hence, the same results
would be expected for other halogen derivatives. Nitration of
aromatic compounds has been extensively reported in recent years
due to its valuable biological activity [37]. Although the introduc-
tion of a nitro group on the cinnamic scaffold was a plausible
hypothesis, literature data show that the O–H bond dissociation
energy found for nitro substituted phenols is comparable to that
obtained for halogen derivatives [38]. Moreover the Log P value
obtained for the nitro cinnamic analogues is quite similar to that
found for the parent compounds. Hence, little valuable contribution
in terms of lipophilicity would be expected for this type of
compounds.
electron withdrawing group (EWG) of halogen type (
and a weak -donor) in a ortho position to a hydroxyl group does
s-acceptor
p
not influence appreciably the activity. This type of substituent
appears not to disturb the stability of the phenoxyl radical formed
as well as its stabilization through intramolecular interactions.
In which concerns to the ABTSꢀþ radical-scavenging activity one
can conclude that all the phenolic compounds were found to be
active toward ABTSꢀþ and better antioxidants than trolox. However,
structure–activity relationships (SAR) could not be established
since they exhibited the same antioxidant profile. The results are in
From the electrochemical results one can conclude that the
structural principles governing the redox potentials of the cinnamic
acids and derivatives under study were found to be the presence of
a phenolic group, preferentially a catechol, and additional reso-
nance-effective substituents in the aromatic ring.
þ
accordance with Nenadis et al. [43] that concluded that ABTSꢀ
radical assay may give an indication for the presence of antioxi-
dants in a certain system but SARs cannot be readily inferred.
2.3. Antioxidant activity
In order to evaluate the radical-scavenging ability of the
phenolic acids and its ester derivatives (Table 1) total antioxidant
capacity assays (DPPHꢀ - 2,2’-diphenyl-1-picrylhydrazyl radical;
ABTS - 2-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) were
used. The DPPH method was carried out in a homogeneous phase
and has the advantage of establish a real ranking hierarchy of
antioxidant activity of electron- or H-donating agents, since it was
not affected by some factors which interfere in other model
systems, such as metal chelation or partitioning abilities [9,11]. The
generation of the ABTS radical cation form is the basis of other
spectrophotometric method that is currently used for the
measurement of the total antioxidant activity of solutions of pure
substances, aqueous mixtures and beverages [39]. This assay is
based in a decolorization technique in which the radical is gener-
ated directly in a stable form prior to reaction with putative
Table 2
Antioxidant activity (DPPH and ABTS), redox potentials (Ep) and partition coeffi-
cients (Log P) values acquired for the hydroxycinnamic acids and their derivatives.
Compounds
DPPH (TE)a
ABTS (TE)a
Ep (V)
Log P
Caffeic acid
Ferulic acid
5-Bromocaffeic acid (3)
5-Bromoferulic acid (2)
Ethyl caffeate
0.978
0.781
1.007
0.808
0.877
0.593
0.938
0.558
0.862
1.000
2.538
2.534
2.496
2.554
2.558
2.538
2.527
2.558
2.538
1.000
þ0.183
1.15
1.42
1.98
2.25
1.75
2.02
2.58
2.85
–
þ0.335; þ0.447
þ0.182
þ0.335; þ0.442
þ0.175
Ethyl ferulate
þ0.368
Ethyl 5-bromocaffeate (5)
Ethyl 5-bromoferulate (4)
Sinapic acid
þ0.174
þ0.365
þ0.188; þ0.295
–
Trolox
–
a
TE - Trolox Equivalents (procedure is given in Section 4).

A. Gaspar et al. / European Journal of Medicinal Chemistry 44 (2009) 2092–2099
2097
2.4. Estimation of partition coefficients (Log P)
(distortionless enhancement by polarization transfer) (underlined
values). Electron impact mass spectra (EI-MS) were carried out on
a VG AutoSpec instrument; the data are reported as m/z (% of
relative intensity of the most important fragments). Melting points
were obtained on a Ko¨fler microscope (Reichert Thermovar) and
are uncorrected.
In view of better understanding the overall properties of the
compounds the lipophilicity, expressed as the octanol–water
partition coefficient and herein called Log P, was calculated
according to Crippen’s fragmentation method [44].
The examination of the data obtained allow to conclude that the
introduction of a halogen group in the position-5, that is in an ortho
position to a phenolic group, lead in both series to a significant
increase of the Log P of the cinnamic acids and derivatives (Table 2).
It must be stressed that the partition coefficients calculated for the
hydroxycinnamic derivatives are related with their structural
features.
Higher lipophilicity is often required in this research field since
it could modify the ADME (absorption, distribution, metabolism,
and excretion) properties of hydroxycinnamic antioxidants,
increasing their ability to interact with the polar head groups of the
membrane and local concentration at the water–lipid interface.
They can prevent the initial reaction between aqueous radicals and
membrane phospholipids.
Voltammetric studies were performed using an Autolab PGSTAT
12 potentiostat/galvanostat (Eco-Chemie, Netherlands) and a one-
compartment glass electrochemical cell. Voltammetric curves were
recorded at room temperature using a three-electrode system. A
glassy carbon working electrode (GCE) (d ¼ 2 mm), a platinum wire
counter electrode and an Ag/AgCl saturated KCl reference electrode
were used. A Crison pH-meter with glass electrode was used for the
pH measurements (Crison, Spain).
Spectrophotometric measurements were carried out with
a Varian Cary 1E spectrophotometer and on a Bio-Tek Synergy HT
multiplate reader.
4.3. Synthesis
4.3.1. Synthesis of 3-bromo-4,5-dihydroxybenzaldehyde (1)
3. Concluding remarks
5-Bromovanillin(3-bromo-4-hydroxy-5-methoxybenzaldehyde)
(8.7 mmol, 2 g) was suspended in anhydrous dichloromethane
(50 mL) in an inert atmosphere (N2). To this suspension, at ꢁ60 ꢂC,
boron tribromide (15 mL of 1 M solution in dichloromethane) was
added and the reactionwas then allowed to reach room temperature
with additional stirring for 6 h. After this period, the reaction was
quenched bycautious addition of methanol (15 mL). The solvent was
partially evaporated and the compound was extracted between
dicloromethane/water. The organic layer was then dried over
Na2SO4, filtered, and the solvent was evaporated to a residue which
was recrystallized from water as a white solid.
Taken together, the results obtained in the present work reveal
the existence of a relation between redox potentials and antioxi-
dant activity either for hydroxycinnamic acids or their alkyl ester
derivatives. The antioxidant efficacy is higher when the redox
potential is lower (as seen in the caffeic series). The presence of
a catechol group leads to an increase of the activity due mainly to
resonance stabilization of the phenoxyl radical intermediate with
subsequent ortho-quinone formation [45]. On the other hand the
methoxylation of the meta-phenolic group (ferulic series) lead to an
increase of redox potential and a decrease of the antioxidant
activity, as shown throughout the TAC assay. In addition, it was
shown that the introduction of a substituent of halogen type,
bromo atom, in an ortho position to a phenolic group of hydrox-
ycinnamic acids or their linear monoalkyl esters does not affect
neither the redox potential nor the antioxidant activity. SPAR’s
were found to rule the antioxidant activities of these compounds,
being an effective approach for the design of new antioxidant
agents as well as for the understanding of their mechanism of
action [46,47].
Yield 86%. IR: 3426, 3060, 2974, 1659, 1579, 1441, 1404, 1361,
1307, 1252, 1180, 862, 685, 631, 588, 543, 401. 1H NMR
d
: 7.26 (1H, d,
J ¼ 1.6, H(2)), 7.58 (1H, d, J ¼ 1.6, H(6)), 9.71 (1H, s, CHO), 10.45 (1H,
s, 3-OH), 10.55 (1H, s, 4-OH). 13C NMR
: 109.4 (C5), 112.7 (C2), 127.4
d
(C6), 129.0 (C1), 146.5 (C–OH), 149.3 (C–OH), 190.6 (C]O). mp 219–
221 ꢂC.
4.4. Synthesis of bromocinnamic acids
The synthetic procedure was a modification of the process
described by Freudenberg and Hubner [48]: the corresponding
benzaldehyde (13 mmol, 3 g) and malonic acid (35 mmol, 3.6 g)
were dissolved in pyridine (15 mL) using aniline (1 mL) as catalyst.
The reaction occurred at room temperature during 24 h and was
followed by thin layer chromatography (hexane/ethyl acetate 6:4).
The reaction was then diluted with HCl 5 M and extracted with
3 ꢃ 25 mL of dichloromethane. The organic layer was dried over
anhydrous Na2SO4, filtered and concentrated until a residue was
obtained. Compound 2 was purified by flash chromatography (from
ethyl acetate/petroleum ether 5:5 until ethyl acetate), while
compound 3 was recrystallized from diethyl ether.
In conclusion, these results confirmed the importance of
exploring the phenolic cinnamic systems as templates to obtain
new antioxidant candidates.
4. Experimental section
4.1. Chemicals
Caffeic, ferulic and sinapic acids, 4-hydroxy-3-methox-
ybenzaldehyde, malonic acid and ethylmalonate were purchased
from Sigma-Aldrich Quı´mica S.A. (Sintra, Portugal). All other
reagents and solvents were pro analysis grade and were acquired
from Merck (Lisbon, Portugal) and used without additional purifi-
4.4.1. trans-3-(5-Bromo-4-hydroxy-3-methoxyphenyl)-2-
propenoic acid (2)
cation. Deionised water (conductivity < 0.1
m
S cmꢁ1) was used
throughout all the experiments.
Yield 50%. IR: 3378, 2967, 2932, 1655, 1619, 1499, 1423, 1384,
1274,1172,1038, 970, 836, 589, 541. 1H NMR
d
: 3.94 (3H, t, CH3), 6.37
)), 7.19 (1H, d, J ¼ 1.8 H(2)), 7.34 (1H, d, J ¼ 1.8
H(6)), 7.55 (1H, d, J ¼ 15.9, H( : 56.5 (OCH3), 109.5 (C5),
)). 13C NMR
110.0 (C2), 117.6 (C ), 125.8 (C6), 126.8 (C1), 143.3 (C ), 146.0 (C–
4.2. Apparatus
(1H, d, J ¼ 15.8, H(
b
a
d
1H and 13C NMR data were acquired, at room temperature, on
Bru¨ker AMX 300 spectrometer operating at 300.13 and
a
b
a
OH), 148.7 (C–OCH3), 168.0 (COOH). EI-MS m/z (%): 274 (M þ 2, 97),
272 (Mþꢀ,100),178 (33),150 (22),133 (25),105 (28), 97 (22), 85 (23),
76 (23), 75 (28), 71 (38), 65 (27), 63 (22), 58 (72), 57 (67). mp
233–235 ꢂC (lit. 246 ꢂC (dec.). [49]; 243–244 ꢂC [50]; 257–258 ꢂC
(dec.) [51]).
75.47 MHz, respectively. Dimethylsulfoxide-d6 was used as
a solvent; chemical shifts are expressed in (ppm) values relative to
tetramethylsilane (TMS) as internal reference; coupling constants
d
(J) are given in Hz. Assignments were also made from DEPT

2098
A. Gaspar et al. / European Journal of Medicinal Chemistry 44 (2009) 2092–2099
4.4.2. trans-3-(5-Bromo-3,4-dihydroxyphenyl)-2-propenoic
acid (3)
ethanol. The voltammetric working solutions were prepared, in the
electrochemical cell, by diluting 0.1 mL of the stock solution in
10 mL of supporting electrolyte to get a final concentration of
0.1 mM.
The pH 7.3 supporting electrolyte used in the voltammetric
determinations was prepared by dilution to 100 mL of 6.2 mL of
0.2 M dipotassium hydrogen phosphate and 43.8 mL of 0.2 M
potassium dihydrogen phosphate.
Yield 90%. IR: 3409, 2928, 1678, 1626, 1497, 1430, 1365, 1283,
1182, 1127, 1006, 979, 841, 591. 1H NMR
d
: 6.21 (1H, d, J ¼ 15.9, H(
7.01 (1H, d, J ¼ 1.9 H(2)), 7.30 (1H, d, J ¼ 1.9 H(6)), 7.38 (1H, d,
J ¼ 15.9, H( )), 9.71 (1H, s, 3-OH), 10.06 (1H, s, 4-OH), 12.25 (1H, s,
COOH). 13C NMR
: 110.0 (C5), 113.4 (C2), 116.8 (C ), 123.9 (C6),
126.5 (C1), 143.1 (C
b)),
a
d
a
b
), 145.4 (C–OH), 146.3 (C–OH), 167.7 (COOH). EI-
MS m/z (%): 260 (M þ 2, 83), 258 (Mþꢀ, 86), 214 (31), 162 (70), 134
(53), 133 (39), 105 (41), 79 (30), 77 (67), 71 (54), 69 (56), 62 (31), 58
(100), 57 (91), 55 (69), 53 (60), 51 (78). mp 177–178 ꢂC.
4.8. Total antioxidant capacity (TAC) assays
4.8.1. DPPH radical assay
4.5. Synthesis of trans-ethyl 3-(5-bromo-4-hydroxy-3-
methoxyphenyl)propenoate (4)
A total antioxidant capacity assay was carried out using DPPH as
radical. The experimental procedure was adapted from the litera-
ture, only with slight modifications [53,54].
Compound 4 was synthesized by a Knoevenagel-type conden-
sation between the corresponding benzaldehyde (4.6 mmol, 1 g)
and monoethylmalonate (9.0 mmol, 1.2 g) in pyridine (5 mL) and
using aniline (1 mL) as catalyst. The reaction occurred at 50 ꢂC
during 20 h and was followed by thin layer chromatography
(chloroform/methanol 9:1). The solvent was then partially evapo-
rated, diluted with diethyl ether and washed twice with HCl 2 M
and water. The organic layer was dried over Na2SO4, filtered and
concentrated under reduced pressure. The remaining residue was
recrystallized from dichloromethane/n-hexane.
Briefly, 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical in ethanol
(250
test compounds. The final concentration of the test compounds in
the reaction mixtures was 50 M. Each mixture was then shaken
mM, 2 mL) was added to 2 mL of an ethanolic solution of the
m
vigorously and held for 40 min at room temperature in the dark.
The decrease in absorbance of DPPH at 517 nm was then measured.
Ethanol was used as a blank and a DPPH solution (2 mL) in ethanol
(2 mL) as the control solution. All tests were performed in triplicate.
4.8.2. ABTS radical cation assay
þ
The spectrophotometric analysis of ABTSꢀ radical-scavenging
Yield 70%. IR: 3293, 3098, 3069, 3006, 2970, 2938, 2845, 2740,
1673, 1585, 1499, 1461, 1424, 1401, 1351, 1289, 1155, 1042, 968, 851,
activity was determined according to the method of Re et al. [55].
826, 789, 676. 1H NMR
d
: 1.24 (3H, t, CH3), 3.87 (3H, s, OCH3) 4.16
(2H, m, CH2), 6.59 (1H, d, J ¼ 16.0, H( )), 7.39 (1H, d, J ¼ 1.4, H(2)),
)), 10.05 (1H, s, 4-
: 14.2 (CH3), 56.4 (OCH3), 59.9 (CH2),109.3 (C5),110.0
), 125.9 (C6), 126.51 (C1), 143.6 (C ), 146.0 (C–OH),
þ
The ABTSꢀ cation radical was produced by the reaction between
a
7 mM ABTS in H2O and 2.45 mM potassium persulfate, stored in thþe
dark at room temperature for 12–16 h. Before usage, the ABTSꢀ
solution was diluted to get an absorbance of 0.450 ꢄ 0.001 at
734 nm with water.
7.47 (1H, d, J ¼ 1.4, H(6)), 7.53 (1H, d, J ¼ 16.0, H(
b
OH). 13C NMR
(C2), 116.4 (C
d
a
b
148.5 (C–OCH3), 166.4 (COOH). EI-MS m/z (%): 302 (M þ 2, 57), 300
(Mþꢀ, 58), 257 (32), 252 (31), 230 (26), 228 (32), 176 (58), 133 (32),
105 (34), 77 (27), 58 (100), 53 (27), 51 (31). mp 134–136 ꢂC.
Different ethanolic solutions of each cinnamic acids and deriv-
atives (with concentration of 50
of each were added to 180 L of radical solution (in triplicate) and
absorbances were recorded for ABTSꢀ
a 20 min period. The absorbance of a blank control (20
and 180 L of radical) was set as 100% of radical (0% bleaching).
mM) were prepared. A total of 20 mL
m
þ
,
every 5 min during
L ethanol
4.6. Synthesis of trans-ethyl 3-(5-bromo-3,4-
dihydroxyphenyl)propenoate (5)
m
m
The radical-scavenging activity of the samples was expressed as
Trolox Equivalents (TE) and calculated according to the following
equation:
The ethyl ester was synthesized by a Fischer esterification,
following the procedure described by Borges and Pinto [52]: 5-
bromocaffeic acid (3.9 mmol, 1.0 g) was heated under reflux for ca.
5 h, in ethanol (150 mL) containing H2SO4 (2 mL). The reaction was
followed by thin layer chromatography (chloroform/methanol 9:1)
until no 5-bromocaffeic acid was detected. After cooling, the
solvent was partially evaporated under reduced pressure and the
solution was neutralised with 10% Na2CO3. The mixture was then
extracted with diethyl ether. The organic phases were combined,
washed twice with water, dried over Na2SO4 and concentrated
under reduced pressure. The residue was purified by column
chromatography (silica gel, petroleum ether/ethyl ether).
TE ¼ ðAcontrol ꢁ AtestÞ=ðAcontrol ꢁ Atrolox
Þ
where Acontrol is the absorbance of the control solution (DPPH or
ABTS solution without test sample) and Atest is the absorbance of
the test sample (DPPH or ABTS solution plus compound). Trolox
was used as reference compound in both assays.
4.9. Calculation of partition coefficient (Log P)
Calculation of the Log P values, simulating partitioning of
phenols in an n-octanol/water (1:1, v/v) system, was based on
Crippen’s fragmentation method [44] and was accomplished using
the ChemDraw software (ChemDraw Ultra 11.0, Cambridge Soft
Corp.).
Yield 70%. IR: 3694, 3303, 3079, 2977, 2929, 2860, 1685, 1630,
1595, 1519, 1429, 1370, 1351, 1303, 1281, 1177, 1117, 1030, 973, 875,
846, 808, 722, 619, 587, 508, 446, 466. 1H NMR
d
: 1.23 (3H, t, CH3),
)), 7.05 (1H, s, H(6)), 7.36
)), 9.77 (1H, s, 3-OH), 10.10 (1H,
: 14.3 (CH3), 59.9 (CH2), 110.0 (C5), 113.7 (C2),
4.15 (2H, m, CH2), 6.32 (1H, d, J ¼ 16.0, H(
(1H, s, H(2)), 7.45 (1H, d, J ¼ 16.0, H(
s, 4-OH). 13C NMR
115.8 (C ), 124.1 (C6), 126.3 (C1), 143.6 (Cb), 145.6 (C–OH), 146.3 (C–
a
b
d
Acknowledgments.
a
OH), 166.3 (COOH). EI-MS m/z (%): 288 (M þ 2, 43), 286 (Mþꢀ, 45),
243 (34), 241(35), 216 (22), 214 (32) 162 (100) 134 (49) 133 (18),
105(30), 77 (35), 53 (30), 51 (50). mp 159–161 ꢂC (dec.).
The authors thank financial support from FCT (Portugal) –
Project POCTI/QUI/55631/2004 (co-financed by the European
Community Fund FEDER).
4.7. Electrochemical measurements
References
The stock solutions of the cinnamic acids and derivatives
(10 mM) were prepared by dissolving an appropriate amount in
[1] O.I. Aruoma, B. Halliwell (Eds.), Free Radicals and Food Additives, Taylor &
Francis, London, 1991.

A. Gaspar et al. / European Journal of Medicinal Chemistry 44 (2009) 2092–2099
2099
[2] B. Halliwell, M.A. Murcia, S. Chirico, O.I. Aruoma, Crit. Rev. Food Sci. Nutr. 35
(1995) 7–20.
[25] A. Ault, Techniques and Experiments for Organic Chemistry, sixth ed.
University Science Books, 1998.
[3] F.A.M. Silva, F. Borges, M.A. Ferreira, J. Agric. Food Chem. 49 (2001) 3936–3941.
[4] W.-Y. Chung, Y.-J. Jung, Y.-J. Surh, S.-S. Lee, K.-K. Park, Mutat. Res. 496 (2001)
199–206.
[5] C.A. Gomes, T.G. Cruz, J.L. Andrade, N. Milhazes, F. Borges, M.P.M. Marques, J.
Med. Chem. 46 (2003) 5395–5401.
[6] S.M. Fiuza, C. Gomes, L.J. Teixeira, M.T.G. Cruz, M.N.D.S. Cordeiro, N. Milhazes,
F. Borges, M.P.M. Marques, Bioorg. Med. Chem. 12 (2004) 3581–3589.
[7] P. Fresco, F. Borges, C. Diniz, M.P.M. Marques, Med. Res. Rev. 26 (2006)
747–766.
[26] N. Milhazes, P. Martins, E. Uriarte, J. Garrido, R. Calheiros, M.P.M. Marques,
F. Borges, Anal. Chim. Acta 596 (2007) 231–241.
[27] C.A. Collins, F.H. Fry, A.L. Holme, A. Yiakouvaki, A. Al-Qenaei, C. Pourzand,
C. Jacob, Org. Biomol. Chem. 3 (2005) 1541–1546.
[28] B. Yang, A. Kotani, K. Arai, F. Kusu, Chem. Pharm. Bull. 49 (2001) 747–751.
[29] A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamentals and Appli-
cations, second ed. Wiley, New York, 2001.
[30] P. Hapiot, A. Neudeck, J. Pinson, H. Fulcrand, P. Neta, C. Rolando, J. Electroanal.
Chem. 405 (1996) 169–176.
[8] D.L. Madhavi, S.S. Deshpande, D.K. Salunkhe (Eds.), Food Antioxidants: Tech-
nological, Toxicological, and Health Perspectives, Dekker, New York, 1996.
[9] F.A.M. Silva, F. Borges, C. Guimara˜es, J.L.F.C. Lima, C. Matos, S. Reis, J. Agric. Food
Chem. 48 (2000) 2122–2126.
[10] S. Teixeira, C. Siquet, C. Alves, I. Boal, M.P. Marques, F. Borges, J.L.F.C. Lima,
S. Reis, Free Radical Biol. Med. 39 (2005) 1099–1108.
[11] C. Siquet, F. Paiva-Martins, J.L.F.C. Lima, S. Reis, F. Borges, Free Radical. Res. 40
(2006) 433–442.
[12] M.P.M. Marques, F. Borges, J. Sousa, R. Calheiros, J. Garrido, A. Gaspar,
F. Antunes, C. Diniz, P. Fresco, Lett. Drug Design Discov. 3 (2006) 316–320.
[13] N.F.L. Machado, R. Calheiros, S.M. Fiuza, F. Borges, A. Gaspar, J. Garrido,
M.P. Marques, J. Mol. Model 13 (2007) 865–877.
[14] R. Calheiros, N.F.L. Machado, S.M. Fiuza, A. Gaspar, J. Garrido, N. Milhazes,
F. Borges, M.P.M. Marques, J. Raman Spectrosc. 39 (2008) 95–107.
[15] S.M. Fiuza, E. Van Besien, N. Milhazes, F. Borges, M.P.M. Marques, J. Mol. Struct.
693 (2004) 103–118.
[16] P. Cos, P. Rajan, I. Vedernikova, M. Calomme, L. Pieters, A.J. Vlietinck,
K. Augustyns, A. Haemers, D.V. Berghe, Free Radical. Res. 36 (2002) 711–716.
[17] S. Son, B.A. Lewis, J. Agric. Food Chem. 50 (2002) 468–472.
[18] M.P.M. Marques, F. Borges, A.M. Amorim da Costa, L.A.E. Batista de Carvalho,
in: K. Kneipp, R. Aroca, H. Kneipp, E. Wentrup-Byrne (Eds.), New Approaches
in Biomedical Spectroscopy, ACS Symposium Series 963, Oxford University
Press, Washington, DC, 2007 cap. 21.
[31] C. Giacomelli, K. Ckless, D. Galato, F.S. Miranda, A. Spinelli, J. Braz. Chem. Soc.
13 (2002) 332–338.
[32] S.K. Trabelsi, N.B. Tahar, R. Abdelhedi, Electrochim. Acta 49 (2004) 1647–1654.
[33] S.K. Trabelsi, N.B. Tahar, B. Trabelsi, R. Abdelhedi, J. Appl. Electrochem. 35
(2005) 967–973.
[34] M. Gattrell, D.W. Kirk, J. Electrochem. Soc. 140 (1993) 1534–1540.
[35] M.M. Baizer, H. Lund, Organic Electrochemistry, Harper&Row, New York, 1972.
[36] J.E.M. Diniz, R.S. Borges, C.N. Alves, J. Mol. Struct. (Theochem) 673 (2004) 93–97.
[37] J.-L. Grenier, N. Cotelle, P. Cotelle, J.-P. Catteau, Bioorg. Med. Chem. Lett. 6
(1996) 431–434.
[38] A.K. Chandra, T. Uchimaru, Int. J. Mol. Sci. 3 (2002) 407–422.
[39] N.J. Miller, C.A. Rice-Evans, Free Radical. Res. 26 (1997) 195–199.
[40] K.L. Wolfe, R.H. Liu, J. Agric. Food Chem. 55 (2007) 8896–8907.
[41] J.A. Go´mez-Ruiz, D.S. Leake, J.M. Ames, J. Agric. Food Chem. 55 (2007)
6962–6969.
[42] H. Kikuzaki, M. Hisamoto, K. Hirose, K. Akiyama, H. Taniguchi, J. Agric. Food
Chem. 50 (2002) 2161–2168.
[43] N. Nenadis, L.-F. Wang, M. Tsimidou, H.-Y. Zhang, J. Agric. Food Chem. 52
(2004) 4669–4674.
[44] A.K. Ghose, G.M. Crippen, J. Chem. Inf. Comput. Sci. 27 (1987) 21–35.
[45] M. Antolovich, D.R. Bedgood Jr., A.G. Bishop, D. Jardine, P.D. Prenzler,
K. Robards, J. Agric. Food Chem. 52 (2004) 962–971.
[46] O. Firuzi, A. Lacanna, R. Petrucci, G. Marrosu, L. Saso, Biochim. Biophys. Acta
1721 (2005) 174–184.
[19] A.M.S. Silva, C.M.M. Santos, J.A.S. Cavaleiro, H.R. Tavares, F. Borges, F.A.M. Silva,
in: G.A. Webb, P.S. Belton, A.M. Gil, I. Delgadillo (Eds.), Magnetic Resonance in
Food Science: A View to the Future, Royal Society of Chemistry, Cambridge,
2001, pp. 110–116.
[47] N. Milhazes, R. Calheiros, M.P.M. Marques, J. Garrido, M.N.D.S. Cordeiro,
C. Rodrigues, S. Quinteira, C. Novais, L. Peixe, F. Borges, Bioorg. Med. Chem. 14
(2006) 4078–4088.
[20] F. Borges, J.L.F.C. Lima, I. Pinto, S. Reis, C. Siquet, Helv. Chim. Acta 86 (2003)
3081–3087.
[21] J. Vaya, M. Aviram, Curr. Med. Chem.–Imm. Endocr. Metab. Agents 1 (2001)
99–117.
[48] K. Freudenberg, H.H. Hubner, Chem. Ber. 85 (1952) 1181.
[49] J. Haworth, D. Robert, J. Chem. Soc. (1944) 535–537.
[50] V.S. Webster, J. Pharm. 112 (1940) 291–296.
[51] W.M. Whaley, M. Meadow, W.L. Dean, J. Org. Chem. 19 (1954) 1022–1026.
[52] M.F.M. Borges, M.M.M. Pinto, J. Liq. Chromatogr. 17 (1994) 1125–1139.
[53] N. Nenadis, M. Tsimidou, J. Am. Oil Chem. Soc. 79 (2002) 1191–1195.
[54] A. Torres de Pinedo, P. Pen˜alver, J.C. Morales, Food Chem. 103 (2007) 55–61.
[55] R. Re, N. Pellegrini, A. Proteggente, A. Pannala, M. Yang, C. Rice-Evans, Free
Radical Biol. Med. 26 (1999) 1231–1237.
´
´
[22] A. Torres de Pinedo, P. Pen˜alver, I. Perez-Victoria, D. Rondon, J.C. Morales, Food
Chem. 105 (2007) 657–665.
´
˜
´
´
[23] M.D. Rivero-Perez, P. Muniz, M.L. Gonzalez-Sanjose, J. Agric. Food Chem. 55
(2007) 5476–5483.
[24] M.C. Foti, C. Daquino, C. Geraci, J. Org. Chem. 69 (2004) 2309–2314.