Solvent and pH effects on the antioxidant activity of caffeic and other phenolic acids

Article abstract of DOI:10.1021/jf053159+

The antioxidant activity of several phenolic acids and esters has been investigated both in organic solutions and in large unilamellar phosphatidylcholine vesicles. In solution these compounds behaved as good antioxidants, with the exception of protocatechuic acid, due to the presence of the catechol moiety. Because their antioxidant activity followed an inverse dependence on the magnitude of their O-H bond dissociation enthalpies (BDE), the key mechanism of the chain-breaking action was attributed to hydrogen atom transfer (HAT) from the phenolic OH to peroxyl radicals. In unilamellar vesicles the antioxidant activity was strongly dependent on the pH of the buffer solution. In acid media (pH 4) all of the examined phenolic acids or esters behaved as weak inhibitors of peroxidation, whereas, with increasing pH, their antioxidant activity increased substantially, becoming comparable to or even better than that of Trolox. At pH 8 they also gave rise to lag phases 2-3 times longer than that of Trolox. The increased activity being observed in proximity of the pK_a value corresponding to the ionization of one of the catecholic hydroxyl groups, this effect has been attributed to the high antioxidant activity of the phenolate anion.

Full text of DOI:10.1021/jf053159+

2932 J. Agric. Food Chem. 2006, 54, 2932−2937
Solvent and pH Effects on the Antioxidant Activity of Caffeic
and Other Phenolic Acids
RICCARDO AMORATI,† GIAN FRANCO PEDULLI,*,† LUCIANA CABRINI,‡
‡
LAURA ZAMBONIN,‡ AND LAURA LANDI
Dipartimento di Chimica Organica “A. Mangini”, Universita` di Bologna, Via San Giacomo 11,
I-40126 Bologna, Italy; and Dipartimento di Biochimica “G. Moruzzi”, Universita` di Bologna,
Via Irnerio 48, I-40126 Bologna, Italy
The antioxidant activity of several phenolic acids and esters has been investigated both in organic
solutions and in large unilamellar phosphatidylcholine vesicles. In solution these compounds behaved
as good antioxidants, with the exception of protocatechuic acid, due to the presence of the catechol
moiety. Because their antioxidant activity followed an inverse dependence on the magnitude of their
O-H bond dissociation enthalpies (BDE), the key mechanism of the chain-breaking action was
attributed to hydrogen atom transfer (HAT) from the phenolic OH to peroxyl radicals. In unilamellar
vesicles the antioxidant activity was strongly dependent on the pH of the buffer solution. In acid
media (pH 4) all of the examined phenolic acids or esters behaved as weak inhibitors of peroxidation,
whereas, with increasing pH, their antioxidant activity increased substantially, becoming comparable
to or even better than that of Trolox. At pH 8 they also gave rise to lag phases 2-3 times longer than
that of Trolox. The increased activity being observed in proximity of the pKa value corresponding to
the ionization of one of the catecholic hydroxyl groups, this effect has been attributed to the high
antioxidant activity of the phenolate anion.
KEYWORDS: Antioxidants; catechol; liposomes; peroxyl radicals; pH effect; phenolic acids
INTRODUCTION
is important because it may contribute to the resonance
stabilization of the phenoxyl radical formed, whereas others
claim that the conjugated double bond is not required for
antioxidant efficacy (9-13). Despite many studies, important
points regarding the structure-activity of cinnamic derivatives
are still poorly clarified, especially those concerning the relative
antioxidant activity in organic and aqueous environments (5)
and the effect of pH on the antioxidant behavior in water systems
(14). Actually, this information is of great importance in
predicting many biological aspects of practical interest, given
the large pH range experienced by food during its way through
the digestive tract.
The present study was undertaken to obtain quantitative data
on the peroxyl radical scavenging activity of caffeic acid and
related derivatives and to better understand the mechanism of
their chain-breaking activity in different experimental systems.
These studies were carried out in a homogeneous nonpolar
medium, such as the mixture styrene/chlorobenzene, and in a
heterogeneous system consisting of large unilamellar vesicles
of egg yolk lecithin, where interfacial interactions, molecular
packing, and dynamics of the lipid phase can be envisaged to
be similar to those of natural membranes.
Polyphenols are found ubiquitously in both edible and
nonedible plants and are important for normal plant growth and
defense against infections and injury (1). There is currently an
intense research interest regarding the phenolic constituents of
the diet and their role in contributing to the maintenance of
good health (2-4). It is generally assumed that their beneficial
effects are due, at least partially, to their antioxidant activity,
which may be brought about at three distinct levels: scavenging
of reactive oxygen and nitrogen species; chelating of prooxidant
transition metal ions such as copper and iron; and inhibition of
prooxidant enzymes [lipoxygenase, myeloperoxidase, xanthine
oxidase, NAD(P)H oxidase, cytochrome P-450 enzymes] (3, 5).
Phenolic acids have been extensively investigated for their
antioxidant activity in vitro and in vivo in humans and rats,
and particularly as agents able to enhance the resistance to
oxidation of low-density lipoproteins (6-8). Structure-activity
relationship studies on cinnamic acids have pointed out the
importance of the catechol group for the radical scavenging
efficacy (5), although the role of the ethylenic side chain remains
controversial. Some authors suggest that this structural feature
MATERIALS AND METHODS
* Author to whom correspondence should be addressed (telephone
++39.051.2095681; fax ++39.051.2095688; e-mail gianfranco.pedulli@
unibo.it).
Chemicals. Solvents were of the highest grade commercially
available and were used as received. Egg yolk phosphatidylcholine (PC)
was purchased as chloroform solution from Lipid Products (Redhill,
† Dipartimento di Chimica Organica “A. Mangini”.
‡ Dipartimento di Biochimica “G. Moruzzi”.
10.1021/jf053159+ CCC: $33.50
© 2006 American Chemical Society
Published on Web 03/22/2006

Antioxidant Activity of Phenolic Acids
J. Agric. Food Chem., Vol. 54, No. 8, 2006 2933
× 10-3 M Na2EDTA to obtain a PC dispersion mainly constituted by
multilamellar vesicles. The milky suspension was then transferred into
LiposoFast (Avestin, Ottawa, Canada) and extruded 21 times back and
forth through two polycarbonate filters with 100 nm pore size
(Nucleopore Corp., Pleasanton, CA) to obtain large unilamellar vesicles
(LUVET). The total volume was then adjusted to give a final
concentration of 1.5 × 10-2 M PC.
Vesicle Peroxidation. Autoxidation experiments in the presence and
in the absence of antioxidants were carried out by monitoring the
oxygen concentration with a miniaturized Clark-type electrode (Instech,
Plymouth Meeting, PA) provided with an automatic data recorder
(World Precision Instruments, Sarasota, FL). The measurement chamber
(internal volume of 0.6 mL) was kept at constant temperature by
circulating water and was protected from room light to avoid initiator
photodecomposition. After thermal equilibration at 37 °C of the
oxidizable substrate, the appropriate amount of AAPH (final concentra-
tion of 1.7 × 10-2 M) was injected into the cell at the beginning of
data collection. After a few minutes, an ethanol solution of the
appropriate phenolic acid or ester was added to obtain a final
concentration of 5 × 10-6 M.
Figure 1. Structures of compounds 1−8 and Trolox.
U.K.), and styrene was was percolated over aluminum oxide to remove
the stabilizer. The azo compounds 2,2′-azobis(2,4-dimethylvaleronitrile)
(AMVN) and 2,2′-azobis(2-amidinopropane)dihydrochloride (AAPH)
were from Wako Pure Chemicals Ind. Ltd. and were stored at -20 °C.
Caffeic acid (1), caffeic acid phenethyl ester (2), protocatechuic acid
(3), chlorogenic acid (4), dihydrocaffeic acid (5), tert-butylcatechol (7),
5-di-tert-butyl-4-hydroxycinnamic acid (9), R-tocopherol (8), and Trolox
were commercially available (structures shown in Figure 1), whereas
methyl dihydrocaffeate (6) was synthesized from dihydrocaffeic acid
by Fischer esterification following the procedure described in the
literature (15). All other compounds were commercially available and
were used as received.
Autoxidation Experiments in Homogeneous Solution. Autoxida-
tion experiments were performed in a two-channel oxygen uptake
apparatus, based on a Validyne DP 15 differential pressure transducer,
that has been described elsewhere (16). The entire apparatus was
immersed in a thermostated bath ensuring a constant temperature within
(0.1 °C. In a typical experiment, a small amount of a methanol solution
of the antioxidant was added to an air-saturated solution of styrene
(final concentration of 5 × 10-6 M) and equilibrated with the reference
solution containing an excess of R-tocopherol (1 × 10-4 M) at 30 °C.
This instrumental setting allowed us to have the N2 production and the
oxygen consumption derived from the azoinitiator decomposition
already subtracted from the measured reaction rates. After equilibration,
a concentrated chlorobenzene solution of AMVN (final concentration
of 5 × 10-3 M) was injected in both reference and sample flasks, and
the oxygen consumption in the sample was measured from the
differential pressure between the two channels recorded as function of
time. The apparatus was calibrated before each series of experiments.
Induction period lengths (τ) were determined by the intersection
between the regression lines to the inhibited and the uninhibited traces.
Initiation rates, Ri, were determined in preliminary experiments by the
inhibitor method using R-tocopherol as reference antioxidant: Ri )
2[R-TOH]/τ.
RESULTS AND DISCUSSION
Autoxidation Experiments in Homogeneous Solution. The
antioxidant activity of caffeic acid (1), caffeic acid phenethyl
ester (2), dihydrocaffeic acid (5), methyl dihydrocaffeate (6),
protocatechuic acid (3), and 4-tert-butylcatechol (7) was deter-
mined by following the oxygen consumption in a closed vessel
during the autoxidation of an air-saturated chlorobenzene
solution of styrene (4.3 M), in the absence and in the presence
of one of the phenols or R-tocopherol (R-TOH), used as
reference antioxidant, each at 5 × 10-6 M concentration. All
samples were added with a small amount (0.1 M) of methanol
to improve the solubility of the inhibitors.
The reactions were initiated by thermal decomposition of
AMVN at 30 °C and were followed by means of a gas
absorption recording apparatus using a commercial differential
pressure transducer as detector (16). Under these conditions,
the slopes of the inhibited period of the oxygen consumption
traces provide the rate constants for the reaction of the
antioxidants with peroxyl radicals, kinh, by means of eq 1:
d[O2]
dt
kp[styrene]Ri
-
)
(1)
nkinh[antioxidant]
The rate constant for the propagation reaction of styrene at 30
°C is known to be kp ) 41 M-1 s-1 (20); the rate of initiation,
Ri, was measured in preliminary experiments as described under
EPR and Bond Dissociation Enthalpy (BDE) Measurements. The
BDE of the phenolic O-H bond of 3,5-di-tert-butyl-4-hydroxycinnamic
acid (9) was determined by using the radical EPR equilibration method
described elsewhere (17, 18). Briefly, a deoxygenated benzene solution
containing a mixture of phenol 9 and 2,4,6-trimethylphenol in a ratio
ranging between 1/20 and 1/50 and di-tert-butyl peroxide (10% v/v)
were sealed under nitrogen in a Suprasil quartz EPR tube. The sample
was inserted at room temperature in the cavity of an EPR spectrometer
and photolyzed with the unfiltered light from a 500 W high-pressure
mercury lamp. Relative radical concentrations, determined by computer
simulation, afforded the equilibrium constant between the two phenol/
phenoxyl couples, from which the ∆BDE(OH) for 9 was calculated
with respect to the known BDE(O-H) value for 2,4,6-trimethylphenol
(81.6 kcal/mol) (17, 19) on the assumption that the entropic term can
be neglected.
Materials and Methods, and the stoichiometric factor, n, that
is, the number of peroxyl radicals trapped by each antioxidant
molecule, was obtained from eq 2 and found to be equal to 2,
Riτ
n )
(2)
[antioxidant]
except for compound 3, for which the oxygen consumption trace
does not show a definite induction period. This value is
consistent with the mechanism of inhibition of catechols, which
are known to react with two peroxyl radicals to give o-quinone
derivatives (21).
In the case of protocatechuic acid, the rate constant of
inhibition, kinh, was instead determined by using eq 3 (22), which
Preparation of Large Unilamellar Vesicles. The appropriate
amount of a chloroform solution of PC was placed in a round-bottom
tube, and the solvent was carefully removed with a gentle nitrogen
stream. The thin film was vortex-stirred for 7 min with 0.6 mL of
buffered isotonic solutions (phosphate-buffered saline for pH 5, 6, 7.2,
and 8 or 0.1 M potassium hydrogen phthalate for pH 4) containing 1
Rox,0
Rox
Rox
nkinh[antioxidant]
-
)
(3)
Rox,0
2k R
x
t
i
holds when a distinct inhibited period is not observed. The terms

2934 J. Agric. Food Chem., Vol. 54, No. 8, 2006
Amorati et al.
Table 1. Measured Rate Constants (kinh) for the Reaction of Phenolic
Antioxidants with Peroxyl Radicals at 30
the Thermally Initiated Autoxidation of Styrene and Estimated Bond
Dissociation Enthalpies (BDE) for the O H Bond
°C in Chlorobenzene during
−
-
-1
antioxidant
kinha (M 1 s
)
BDEb (kcal/mol)
1
2
3
4
5
6
7
8
2.9
3.0
6.5
×
×
×
105
105
104
78.7
78.7
82.2 c
78.7
79.0
79.0
78.8
2.9
3.5
4.3
3.2
×
×
×
×
105
105
105
106
Figure 3. Mechanism for the reaction of catechols with peroxyl radicals.
d
77.1
±
0.1e
withdrawing COOH group directly bonded to the aromatic ring
has the opposite effect.
a In the presence of 0.1 M MeOH; mean of three measures, error within 10%.
b BDE values calculated by using the additive rule with the revised BDE(PhO
H)
−
Because it is well established that in phenolic antioxidants
every substituent contributes to the overall BDE value and that
this contribution is approximately constant and independent of
the nature of the other substituents, except when they interact
sterically with each other, the O-H bond strengths of the various
derivatives can be easily estimated (17-19, 25). The group
contributions can be used to calculate the BDEs of 7 (78.8 kcal/
mol), of 5 and 6 (79.0 kcal/mol, under the reasonable assumption
that the alkyl chain has the same effect of a methyl group), and
of 3 (82.2 kcal/mol). To calculate the O-H bond strength in
caffeic acid and its esters, the unreported contribution of the
acrylic substituent, CHdCHsCOOH is needed. Thus, the O-H
BDE value in 3,5-di-tert-butyl-4-hydroxycinnamic acid (9) was
measured by means of the EPR radical equilibration technique
as described under Materials and Methods. From the resulting
value (79.7 kcal/mol), a contribution of -2.0 kcal/mol can be
calculated for a para acrylic substituent, when it is considered
that the additive term of two tert-butyl substituents ortho to the
OH group is -4.8 kcal/mol (18). Thus, the BDE value for
caffeic acid 1 and its esters 2 and 4 is estimated as 78.7 kcal/
mol.
)
86.5 kcal/mol (19) and the group contributions (in kcal/mol) reported in refs 17,
18, and 25. c Value obtained with the COOH group in the para position. d From
ref 26. e Experimental value (17, 19).
Figure 2. Oxygen consumption traces recorded during the autoxidation
of styrene 4.3 M in PhCl at 30
-
3
°C initiated by AMVN (5 × 10 M) in the
absence of any antioxidant (0) and in the presence of the antioxidants
From Table 1 it is seen that both BDE values and inhibition
rate constants can be divided into three main groups: the first
one contains only protocatechuic acid and is characterized by
the highest O-H BDE value (82.2 kcal/mol) and a lower
reactivity (kinh ) 6.5 × 104 M-1 s-1); the second one includes
all of the remaining catecholic compounds that show intermedi-
ate and similar reactivities (2.9 × 105 M-1 s-1 e kinh e 4.3 ×
105 M-1 s-1) and bond strengths (78.7 e BDE e 79.0 kcal/
mol); the third one contains the very reactive R-tocopherol (kinh
) 3.2 × 106 M-1 s-1) (26), which shows also the lowest BDE
value (77.1 kcal/mol). The fact that the antioxidant activity of
phenolic acids and esters follows an inverse dependence on the
magnitude of their O-H BDE values provides evidence that in
weakly polar organic media, such as the mixture styrene/
chlorobenzene used in the present work, the key mechanism of
the chain-breaking action is the hydrogen atom transfer (HAT)
from the phenolic OH to peroxyl radicals (Figure 3).
-6
5.0
× 10 M.
Rox,0 and Rox represent the initial oxygen consumption rates in
the absence and in the presence of the inhibitor, respectively.
The termination rate constant for styrene, 2kt ) 4.2 × 107 M-1
s-1 at 30 °C (20), was used, and n was assumed to be equal to
2, similarly to the other phenolic acids.
The small amount of methanol present in the sample is
expected to slightly reduce the kinh value because of the
formation of a hydrogen bond complex between the alcohol
and the phenolic OH. However, this effect should be nearly
constant along the series of catechols, so that the relative
reactivity should be the same (23).
Table 1 reports the kinh values measured by analyzing the
oxygen consumption traces of Figure 2. It can be seen that, in
a homogeneous solution, the examined antioxidants possess a
remarkable antioxidant activity with the only exception of 3.
This strong activity depends on the presence of the catechol
moiety and is due to the formation, by hydrogen atom
abstraction, of a protonated semiquinone radical stabilized by
an intramolecular hydrogen bond stronger than that in the parent
catechol (24). This determines a low value of the BDE of the
catecholic O-H bond and a high rate of hydrogen atom transfer
to the chain carrying peroxyl radicals. In most of the investigated
compounds, the O-H BDE value of catechol undergoes a
further reduction due to the presence of ring substituents capable
of inducing an additional weakening of the O-H bond. The
only exception is protocatechuic acid (3), for which the electron-
Autoxidation Experiments in Large Unilamellar Vesicles.
The antioxidant activities of 1, 2, 3, and 5-7 were also
investigated in large unilamellar vesicles of PC, where interfacial
interactions, molecular packing, and dynamics of the lipid phase
can be envisaged to be similar to those of natural membranes,
by using the water-soluble azo compound AAPH as initiator of
radical peroxidation and Trolox, a hydrophilic analogue of
vitamin E, as reference antioxidant. AAPH, owing to its
hydrophilicity, gives rise to radical chain initiation at a constant
rate in the aqueous phase and is commonly used for studying
the effectiveness of hydrophilic and lipophilic peroxidation
inhibitors against the attack of oxygen radicals to biomembranes

Antioxidant Activity of Phenolic Acids
J. Agric. Food Chem., Vol. 54, No. 8, 2006 2935
Table 2. Reactivities of Phenolic Antioxidants toward Peroxyl Radicals
in PC Unilamellar Vesicles at 37
°C and pH 7.2 and pKa Values for
the Hydroxyl Groups of Each Compound
a
antioxidant
nkinh(antioxidant)/2kinh(Trolox)
pKa
1
3.40
4.4b
8.5b
11.2b
2
3
0.70
0.70
8.4c
11.4c
4.4d
8.9d
10.8d
5
0.78
4.4b
9.2b
-
2
11.4b
Figure 4. Oxygen uptake traces during AAPH (1.7
peroxidation of PC (1.5
×
10 M) induced
C and pH
7.2 in the absence of inhibitor (control) and in the presence of one of the
-
2
× 10 M) unilamellar vesicles at 37 °
6
0.82
0.72
1.00
9.2b
11.1b
-
6
various phenolic antioxidants, each at the same concentration (5
M). The arrow shows antioxidant injection.
× 10
7
9.4e
13.8e
Trolox
4.4f
11.9g
from the external water environment. Trolox is also largely
partitioned in the water phase and in the polar interface of model
membranes.
a Mean of three determinations, error within 20%. b From ref 13. c Assumed to
be identical to the pKa values of methyl caffeate (13). d From ref 27. e From ref
28. f Assumed to be identical to the pKa of other carboxylic acids (13). g From ref
29.
The oxygen uptake traces (Figure 4), obtained at 37 °C and
pH 7.2, indicate that 1 behaves as a very effective antioxidant
showing both a longer inhibition time and a higher rate constant
for the reaction with peroxyl radicals than those measured with
the same amount of Trolox. These results are similar to those
reported by Laranjinha et al. (7) when assaying the antioxidant
activity of 1 in low-density lipoproteins by monitoring the
fluorescence decay of cis-parinaric acid previously incorporated
into the particle. On the other hand, in the presence of equivalent
amounts of 2, 3, and 5-7, the rate of oxygen uptake was
somewhat reduced with respect to control experiments, but no
evident inhibition time occurred. These results are in striking
contrast with what was observed in homogeneous solution,
where caffeic acid is characterized by about the same antioxidant
activity as the other phenolic acids.
We report, as a measure of the antioxidant activity, the ratio
nkinh(antioxidant)/2kinh(Trolox) (Table 2), where the two terms were
obtained by means of eq 3 from the slopes of the oxygen
consumption traces as described above. This was necessary
because in aqueous media the stoichiometric coefficient, n, could
not be determined for the majority of phenolic acids. Actually,
at pH 7.2 only caffeic acid showed a clear induction period
ranging between 4 and 5.
The data of Table 2 show that the reactivity of caffeic acid
toward peroxyl radicals is ∼5 times higher than that of the other
phenolic acids and even larger than that of Trolox. Because the
examined phenols contain two or three hydroxyl groups
characterized by different pKa values, autoxidation experiments
were carried out at various pH values of the buffered solution
used to prepare PC vesicles, to test if there is any pH dependence
of the antioxidant activity of these compounds.
The results of the measurements at pH 4 (Figure 5A) show
that none of the tested phenolic acids, including caffeic acid,
give any recognizable induction period. They only retard the
autoxidation, with the exception of Trolox, which behaves as
in neutral media. At pH 8 (Figure 5B), on the other hand, all
tested phenolic acids behave as good inhibitors with an activity
better than or comparable to that of Trolox. They give rise to
lag phases, where oxygen consumption is strongly reduced,
lasting for periods ∼2-3 times longer than that of Trolox. The
less efficient antioxidant at pH 8 is 2. On the basis of these
results, it may be inferred that, under acidic conditions (for
-
2
Figure 5. Oxygen uptake traces during AAPH (1.7
×
10 M) induced
10 M) unilamellar vesicles at 37 C in the
absence of inhibitor (control) and in the presence of one of the various
-
2
peroxidation of PC (1.5
×
°
-6
phenolic antioxidants, each at the same concentration (5
× 10 M). The
arrow shows antioxidant injection. Experiments were performed using (A)
acid (pH 4.0) or (B) basic (pH 8.0) buffered isotonic solutions.
example, of gastric lumen), the tested phenolic compounds are
not efficient radical scavengers. On the other hand, where the
pH can reach 7-8, such as in the intestinal tract, in blood, in
the extracellular fluid, and within cells, these diphenolic
compounds are expected to behave as very good antioxidants.
To assess the antioxidant activity of phenolic acids at
intermediate pH values, additional experiments were carried out

2936 J. Agric. Food Chem., Vol. 54, No. 8, 2006
Amorati et al.
Figure 6. Antioxidant activity of caffeic acid with respect to Trolox
-
2
measured at various pH values during the AAPH induced peroxidation of
PC large unilamellar vesicles at 37 °C. Error bars: ± 20%. The line
Figure 7. Oxygen uptake traces during AAPH (1.7
×
10 M) induced
-
2
peroxidation of PC (1.5
×
10 M) unilamellar vesicles at 37 C and pH
7.2 in the absence of inhibitor (control) and in the presence of caffeic
acid (1), chlorogenic acid (4), and caffeic acid phenethyl ester (2), each
°
represents the percent molar fraction of the phenolate anion from caffeic
-
acid (12 ) calculated from the pKa values given in Table 2.
-
6
at the same concentration (5
injection.
× 10 M). The arrow shows antioxidant
with caffeic acid. The results are shown in Figure 6, reporting
the ratio nkinh(1)/2kinh(Trolox) as a function of pH.
The plot shows that the antioxidant activity of 1 undergoes
a sudden increase at pH ∼7, which is when the amount of the
species 12- (12- is the phenolate formed from 1, i.e., X )
CHdCHCOO- in Figure 3) is ∼4%. We attribute the greater
reactivity of caffeic acid and related compounds, observed at
high pH, to the formation of the phenolate ion (12-), the reaction
of which with peroxyl radicals must be faster than that of the
parent catecholic species. Consistent with this explanation is
the fact that the other catechols characterized by higher pKa
values and thus containing a lower percentage of the corre-
sponding phenolate at pH 7.2 are less reactive than 1, whereas
under more basic conditions (pH 8.0) they become almost as
reactive as caffeic acid (Figure 5B), independent on the nature
of their side chain. Moreover, also caffeic acid behaves as a
poor antioxidant at acidic pH, where the 12- species is almost
absent (Figure 5A).
These results imply that the anion 12- behaves as a better
antioxidant than the parent species 1-. A similar effect has been
observed by Mukai and co-workers (14) in the reaction between
aroxyl radicals and several catechins at several pH values. The
larger reactivity of the hydroxyl group in the ionized species
was attributed to the greater electron-donating capacity of the
O- group, which activates the adjacent OH.
An alternative explanation might be that in basic solution
the inhibition process involves a rapid electron transfer (ET) to
peroxyl radicals from the anion of the phenolic acid (Figure
3). The decrease of the reduction potential of phenols with
deprotonation to give phenolates [for instance, for catechol EpH7
) 0.53 and EpH13.5 ) 0.04 V vs NHE (29)] provides some
support for this reaction path. It should be pointed out that the
anomalous acceleration in alcoholic solvents of the reaction
between substituted phenols and 2,2-diphenyl-1-picrylhydrazyl
radical (DPPH), recently reported, has been explained on a
similar basis (30, 31), being attributed to partial ionization of
the phenol to give the corresponding phenoxide anion followed
by a very fast ET from the phenoxide to the DPPH radical.
A difficult aspect to be explained in terms of this model is
the behavior of 2, as shown in Figure 4. In fact, its pKa (8.4),
corresponding to the first deprotonation of the catecholic
function, is reported to be almost coincident with that of caffeic
acid (8.5), so that acid and related ester are expected to show
a similar reactivity. In contrast, 2 is considerably less reactive
than 1, both at pH 7.2 and at pH 8.0. A possible justification
for the different antioxidant activities of acid and ester might
be their different hydrophilicities, because the solubility of 2
in water is lower than that of 1 due to the substitution of the
carboxylic OH with a OCH2CH2Ph group. Different partition
of the inhibitors between the water and the lipid phase of the
PC vesicles is predictable, with 1 being highly partitioned in
the water phase and 2 in the lipid region of model membranes.
To check this point we investigated the antioxidant behavior
of chlorogenic acid (4), a derivative of 1 in which the carboxylic
group is esterified with the highly hydrophilic D-quinic acid.
These measurements, reported in Figure 7, show that 4 behaves
as a much better inhibitor than 2, although not so good as caffeic
acid. Thus, it seems that the different partitionings between water
and lipid phase is of considerable importance in determining
the antioxidant activity of these compounds: the more efficient
are those highly soluble in water such as caffeic acid and 4.
Finally, the observation that caffeic acid at pH 7.2 and the
other catecholic derivatives at basic pH give rise to inhibition
periods 2-3 times longer than that of Trolox deserves some
comments. Similar results have been previously reported by
different authors when studying the antioxidant activity of
polyphenols in micelles (32) or their oxidation in flow column
electrolysis (33). We spent considerable effort to detect and
identify the oxidation products of caffeic acid by analyzing the
reaction mixtures with ESI-MS spectrometry without any
appreciable success. Thus, the reason for these anomalous lag
times could not be clarified. According to Roginsky (32) this
behavior should be attributed to the formation of polymeric
oxidation products having antioxidant activities comparable to
those of the starting polyphenols.
In conclusion, the good antioxidant activity of caffeic acid
and related derivatives is due to the presence in these compounds
of the catechol moiety that gives rise, by hydrogen atom
abstraction, to semiquinone radicals highly stabilized by an
intramolecular hydrogen bond, thus determining a relatively
small O-H BDE value. The low O-H bond strength is further
reduced by the effect of ring substituents such as CHdCHs
COOR and CH2sCH2R. The acrylic group is of some relevance
in determining the good antioxidant properties of cinnamic acids.
The antioxidant activities measured in model membranes are
different with respect to homogeneous solution and strongly
dependent on the medium pH, this being attributed to the greater
reactivity of phenolate anions toward peroxyl radicals.
The results here reported show the importance of thermody-
namic and kinetic aspects in predicting the activity of chain-
breaking antioxidants. Moreover, this study confirms that
experimental measurements carried out in model membranes
at different pH values are crucial in the understanding of all

Antioxidant Activity of Phenolic Acids
J. Agric. Food Chem., Vol. 54, No. 8, 2006 2937
(17) Lucarini, M.; Pedrielli, P.; Pedulli, G. F.; Cabiddu, S.; Fattuoni,
C. Bond dissociation energies of O-H bonds in substituted
phenols by equilibration studies. J. Org. Chem. 1996, 61, 9259-
9263.
(18) Brigati, G.; Lucarini, M.; Mugnaini, V.; Pedulli, G. F. Deter-
mination of the substituent effect on the O-H bond dissociation
enthalpies of phenolic antioxidants by the EPR radical equilibra-
tion technique. J. Org. Chem. 2002, 67, 4828-4832.
(19) Mulder, P.; Korth, H.-G.; Pratt, D. A.; DiLabio, G. A.;
Valgimigli, L.; Pedulli, G. F.; Ingold, K. U. Critical re-evaluation
of the O-H bond dissociation enthalpy in phenol. J. Phys. Chem.
A 2005, 109, 2647-2655.
factors responsible for the protective activity of antioxidants
against peroxyl radical attack in biological systems.
ACKNOWLEDGMENT
We thank Dr. Veronica Mugnaini for technical assistance and
an anonymous referee for helpful suggestions.
LITERATURE CITED
(1) Shahidi, F.; Naczk, M. Food Phenolics. Sources, Chemistry,
Effects, Applications; Technomic: Lancaster, PA, 1995.
(2) Owen, R. W.; Giacosa, A.; Hull, W. E.; Haubner, R.; Spiegel-
halder, B.; Bartsch, H. The antioxidant/anticancer potential of
phenolic compounds isolated from olive oil. Eur. J. Cancer 2000,
36, 1235-1247.
(3) Halliwell, B.; Gutteridge J. M. C. Free Radicals in Biology and
Medicine, 3rd ed.; Oxford University Press: New York, 1999.
(4) Ivanov, V.; Carr, A. C.; Frei, B. Red wine antioxidants bind to
human lipoproteins and protect them from metal ion-dependent
and -independent oxidation. J. Agric. Food Chem. 2001, 49,
4442-4449.
(20) Howard, J. A. In Free Radicals; Kochi, J. K., Ed.; Wiley: New
York, 1973; Vol. 2, Chapter 12.
(21) Amorati, R.; Ferroni, F.; Lucarini, M.; Pedulli, G. F.; Valgimigli,
L. A quantitative approach to the recycling of R-tocopherol by
coantioxidants. J. Org. Chem. 2002, 67, 9295-9303.
(22) Chatgilialoglu, C.; Timokhin, V. I.; Zaborovskyi, A. B.; Lutsyk,
D. S.; Prystansy, R. E. Rate constants for the reaction of
cumylperoxyl radicals with Bu3SnH and (TMS)3SiH. Chem.
Commun. 1999, 405-406.
(23) Foti, M. C.; Barclay, L. R. C.; Ingold, K. U. The role of hydrogen
bonding on the H-atom donating abilities of catechols and
naphthalene diols and on a previously overlooked aspect of their
infrared spectra. J. Am. Chem. Soc. 2002, 124, 12881-12888.
(24) Lucarini, M.; Mugnaini, V.; Pedulli, G. F. Bond dissociation
enthalpies of polyphenols: the importance of cooperative effects.
J. Org. Chem. 2002, 67, 928-931.
(25) Lucarini, M.; Pedulli, G. F.; Guerra, M. A critical evaluation of
the factors determining the effect of intramolecular hydrogen
bonding on the O-H bond dissociation enthalpy of catechol and
of flavonoid antioxidants Chem. Eur. J. 2004, 10, 933-939.
(26) Burton, G. W.; Doba, T.; Gabe, E. J.; Hughes, L.; Lee, F. L.;
Prasad, L.; Ingold, K. U. Autoxidation of biological molecules.
4. Maximizing the antioxidant activity of phenols. J. Am. Chem.
Soc. 1985, 107, 7053-7065.
(27) Patel, V. M.; Joshi, J. D. Equilibrium study on the complex for-
mation of europium-, terbium-, dispropsium-, and thulium(III)
with some oxyacids, thioacids, and phenols. J. Indian Chem.
Soc. 1998, 75, 100-101.
(5) Rice-Evans, C. A.; Miller, N. J.; Paganga, G. Structure-
antioxidant activity relationships of flavonoids and phenolic
acids. Free Radical Biol. Med. 1996, 20, 933-956.
(6) Nardini, M.; D’Aquino, M.; Tomassi, G.; Gentili, V.; Di Felice,
M.; Scaccini, C. Inhibition of human low-density lipoprotein
oxidation by caffeic acid and other hydroxycinnamic acid
derivatives. Free Radical Biol. Med. 1995, 19, 541-552.
(7) Laranjinha, J. A.; Almeida, L. M.; Madeira, V. M. Reactivity
of dietary phenolic acids with peroxyl radicals: antioxidant
activity upon low-density lipoprotein peroxidation. Biochem.
Pharmacol. 1994, 48, 487-494.
(8) Raneva, V.; Shimasaki, H.; Ishida, Y.; Ueta, N.; Niki, E.
Antioxidative activity of 3,4-dihydroxyphenylacetic acid and
caffeic acid in rat plasma. Lipids 2001, 36, 1111-1116.
(9) Foley, S.; Navaratnam, S.; McGarvey D. J.; Land, E. J.; Truscott,
T. G.; Rice-Evans, C. A. Singlet oxygen quenching and the redox
properties of hydroxycinnamic acids. Free Radical Biol. Med.
1999, 26, 1202-1208.
(10) Cuvelier, M.-E.; Richard, H.; Berset, C. Comparison of the
antioxidative activity of some acid-phenols: structure-activity
relationship. Biosci., Biotechnol., Biochem. 1992, 56, 324-325.
(11) Chen, J. H.; Ho, C.-T. Antioxidant activities of caffeic acid and
its related hydroxycinnamic acid compounds. J. Agric. Food
Chem. 1997, 45, 2374-2378.
(12) Moon, J.-A.; Terao, J. Antioxidant activity of caffeic acid and
dihydrocaffeic acid in lard and human low-density lipoprotein.
J. Agric. Food Chem. 1998, 46, 5062-5065.
(28) Kennedy, J. A.; Munro, M. H. G.; Powell, H. K. J.; Porter, L.
J.; Foo, L. Y. The protonation reaction of catechin, epicatechin,
and related compounds. Aust. J. Chem. 1984, 37, 885-892.
(29) Steenken, S.; Neta, P. One-electron redox potentials of phenols.
Hydroxy and aminophenols and related compounds of biological
interest. J. Phys. Chem. 1982, 86, 3661-3667.
(30) Litwinienko, G.; Ingold, K. U. Abnormal solvent effects on
hydrogen atom abstraction. 1. The reaction of phenols with 2,2-
diphenyl-1-picrylhydrazyl (dpph) in alcohols. J. Org. Chem.
2003, 68, 3433-3438.
(31) Foti, M. C.; Daquino, C.; Geraci, C. Electron-transfer reaction
of cinnamic acids and their methyl esters with the DPPH• radical
in alcoholic solutions J. Org. Chem. 2004, 69, 2309-2314.
(32) Roginsky, V. Chain-breaking antioxidant activity of natural
polyphenols as determined during the chain oxidation of methyl
linoleate in Triton X-100 micelles. Arch. Biochem. Biophys.
2003, 414, 261-270.
(33) Hotta, H.; Sakamoto, H.; Nagano, S.; Osakai, T.; Tsujino, Y.
Unusually large numbers of electrons for the oxidation of
polyphenolic antioxidants. Biochim. Biophys. Acta 2001, 1526,
159-167.
(13) Silva, F. A. M.; Borges, F.; Guimara˜es, C.; Lima, J. L. F. C.;
Matos, C.; Reis, S. Phenolic acids and derivatives: studies on
the relationship among structure, radical scavenging activity, and
physicochemical parameters. J. Agric. Food Chem. 2000, 48,
2122-2126.
(14) (a) Mukai, K.; Oka, W.; Watanabe, K.; Egawa, Y.; Nagaoka,
S.; Terao, J. Kinetic study of free-radical-scavenging action of
flavonoids in homogeneous and aqueous Triton X-100 micellar
solutions. J. Phys. Chem. A 1997, 101, 3746-3753. (b) Mukai,
K.; Mitani, S.; Ohara, K.; Nagaoka, S. I. Structure-activity
relationship of the tocopherol regeneration reaction by catechins.
Free Radical Biol. Med. 2005, 38, 1243-1256.
(15) Borges, M. F. M.; Pinto, M. M. M. Separation of the diastereo-
isomers of ethyl-esters of caffeic, ferulic and isoferulic acids by
thin-layer and high-performance liquid chromatography. J. Liq.
Chromatogr. 1994, 17, 1125-1139.
(16) Amorati, R.; Pedulli, G. F.; Valgimigli, L.; Attanasi, O. A.;
Filippone, P.; Fiorucci, C.; Saladino, R. Absolute rate constant
for the reaction of peroxyl radicals with cardanol derivatives. J.
Chem. Soc., Perkin Trans. 2 2001, 2142-2146.
Received for review December 16, 2005. Revised manuscript received
February 21, 2006. Accepted February 21, 2006. Financial support from
MIUR and the University of Bologna (Research Project “Free Radical
Processes in Chemistry and Biology: Fundamental Aspects and
Applications in Environment and Material Sciences”) is gratefully
acknowledged.
JF053159+