Kinetic model for the antiradical activity of the isolated p-catechol group in flavanone type structures using the free stable radical 2,2-diphenyl-1- picrylhydrazyl as the antiradical probe

Article abstract of DOI:10.1021/jf070689s

The time evolution of 2,2-diphenyl-1-picrylhydrazyl radical (DPPH ^?) concentration in four solvents (methanol, ethanol, propanol, and acetonitrile) during its reduction by three flavanones containing an isolated p-catechol group (taxifolin, eriodyctiol, and fustin) as well as the time evolution of the mass spectra of the reaction mixture has been determined by spectrophotometry and liquid mass spectrometry, respectively. In alcoholic solvents the reduction curves consisted of an initial short but fast kinetics step followed by a longer slow kinetics step; in contrast, in acetonitrile the reduction curves completely lacked the slow kinetics step. From the results, a kinetic model for the reaction of reduction of the DPPH^? by the isolated p-catechol group in flavanone type structures is proposed. According to this model, the p-catechol group rapidly transfers two hydrogen atoms to DPPH^?, through a fast rate constant k₁, yielding the corresponding o-quinone. Then, the intermediate o-quinone forms an adduct with the alcoholic solvent, through a slow rate constant fa, and regenerates the p-catechol group. The regenerated p-catechol group reduces additional DPPH ^? through a fast rate constant k₃, yielding the corresponding o-quinone, which can form a new adduct with the solvent to regenerate the p-catechol group, and so on. From the kinetics model, two explicit kinetics equations have been derived that fit very well the experimental data points acquired from all assayed compounds in all of the experiments carried out, thus allowing an accurate determination of the corresponding rate and stoichiometric constants.

Full text of DOI:10.1021/jf070689s

5512 J. Agric. Food Chem. 2007, 55, 5512−5522
Kinetic Model for the Antiradical Activity of the Isolated
p-Catechol Group in Flavanone Type Structures Using the Free
Stable Radical 2,2-Diphenyl-1-picrylhydrazyl as the Antiradical
Probe
JOSEÄ M. SENDRA,* ENRIQUE SENTANDREU, AND JOSEÄ L. NAVARRO
Instituto de Agroqu´ımica y Tecnolog´ıa de Alimentos, P.O. Box 76, 46100 Burjassot, Valencia, Spain
The time evolution of 2,2-diphenyl-1-picrylhydrazyl radical (DPPH•) concentration in four solvents
(methanol, ethanol, propanol, and acetonitrile) during its reduction by three flavanones containing an
isolated p-catechol group (taxifolin, eriodyctiol, and fustin) as well as the time evolution of the mass
spectra of the reaction mixture has been determined by spectrophotometry and liquid mass
spectrometry, respectively. In alcoholic solvents the reduction curves consisted of an initial short but
fast kinetics step followed by a longer slow kinetics step; in contrast, in acetonitrile the reduction
curves completely lacked the slow kinetics step. From the results, a kinetic model for the reaction of
reduction of the DPPH• by the isolated p-catechol group in flavanone type structures is proposed.
According to this model, the p-catechol group rapidly transfers two hydrogen atoms to DPPH•, through
a fast rate constant k1, yielding the corresponding o-quinone. Then, the intermediate o-quinone forms
an adduct with the alcoholic solvent, through a slow rate constant k2, and regenerates the p-catechol
group. The regenerated p-catechol group reduces additional DPPH• through a fast rate constant k3,
yielding the corresponding o-quinone, which can form a new adduct with the solvent to regenerate
the p-catechol group, and so on. From the kinetics model, two explicit kinetics equations have been
derived that fit very well the experimental data points acquired from all assayed compounds in all of
the experiments carried out, thus allowing an accurate determination of the corresponding rate and
stoichiometric constants.
KEYWORDS: Antiradical activity; kinetics; p-catechol; flavanone; DPPH•
INTRODUCTION
groups results in a new extended active antiradical group with
enhanced antiradical activity. Hence (see Figure 1), the antiradi-
cal activity of eriodyctiol (5,7,3′,4′-tetrahydroxyflavanone),
which contains an isolated p-catechol group in the B-ring, is
enhanced by the presence of a double bond between C2 and
C3 in the C-ring (luteolin, 5,7,3′,4′-tetrahydroxyflavone) and
even more enhanced by the additional presence of a hydroxyl
on C3 (quercetin, 3,5,7,3′,4′-pentahydroxyflavanone), that is, a
conjugated vinyl-alcohol (12, 13).
Most edible plants as well as foodstuff derived from plants
contain components (e.g., ascorbic acid, flavonoids, carotenoids,
anthocyanidins, hydroxycinnamic acid derivatives, etc.) that
exhibit antiradical activity (1-6). There is partial evidence that
many of these compounds could be bioactive against different
free radical mediated diseases, such as cancer, cardiovascular
diseases, arthritis, and diabetes as well as premature body aging
(7-10). Hence, the intake of antiradicals present in food is
thought to be an important health-protecting factor (11).
It is known that antiradical activity mainly depends on a
reduced number of “active antiradical groups”, having at least
a free hydroxyl group, which are contained within the chemical
structure of active compounds, rather than on their full chemical
structure. Among these active antiradical groups, the vinyl-
alcohol and the p-catechol (3,4-dihydroxybenzene) groups are
noticeable due to their intense antiradical activities. It is also
recognized that the conjugation of isolated active antiradical
Among the different published methodologies for determining
the antiradical activity of both isolated compounds and complex
mixtures of antiradicals, the 2,2-diphenyl-1-picrylhydrazyl
(DPPH•) assay, initially developed by Blois (14) and more
recently adapted by Brand-Williams et al. (15), has been widely
used due to its simplicity. This methodology is based on the
reduction of the free stable radical DPPH•, which strongly
absorbs at 515 nm, to the corresponding hydrazine, which is
almost transparent at this wavelength, by the transfer of
hydrogen atoms from the antiradical. Hence, the time evolution
of the absorbance, subsequently converted to DPPH• concentra-
tion, is the parameter monitored.
* Author to whom correspondence should be addressed (e-mail
jsendra@iata.csic.es; telephone +34 963 90 00 22; fax +34 963 63 63 01).
10.1021/jf070689s CCC: $37.00
© 2007 American Chemical Society
Published on Web 06/14/2007

Antiradical Activity of the p-Catechol Group
J. Agric. Food Chem., Vol. 55, No. 14, 2007 5513
kinetics model for the reaction of reduction of the DPPH• by
the isolated p-catechol group in flavanone type structures is
proposed. Moreover, two explicit kinetic equations have been
derived from the kinetics model, which fit very well the
experimental data points acquired from all of the assayed
compounds, thus allowing the accurate determination of the
corresponding rate and stoichiometric constants.
Figure 1. Chemical structures of the experimental and reference
compounds.
MATERIALS AND METHODS
Reagents and Standards. Spectrophotometric grade methanol,
ethanol, propanol, and acetonitrile were from Sigma (Sigma-Aldrich
Co., St. Louis, MO). Taxifolin and DPPH• (94.6% purity) were from
Fluka (Fluka AG Chemische, Buchs, Switzerland). Fustin was from
Roth (Carl Roth GmbH, Karlsruhe, Germany), and eriodyctiol was
from Extrasynthe`se (Genay, France). Anhydrous sodium sulfate was
from Panreac (Panreac Qu´ımica S.A., Barcelona, Spain).
There is abundant scientific literature dealing with the
determination of the antiradical activity of plant extracts, juices,
and isolated compounds by means of the DPPH• assay.
However, the explicit kinetic equations for these reactions are
not known, and so the fitting of the experimental data points to
a given ad hoc equation to extract kinetic data lacks reliability
and is hardly justifiable. As a consequence, the kinetic rate
constants remain unknown, the antiradical activity is quantified
by means of empirical parameters (i.e., the antiradical activity
EC50), the influence of the molecular structure and solvents on
this activity has been only qualitatively determined, and even
the determined total stoichiometric constants are largely de-
pendent on the experimental reaction time (i.e., the value
assigned to the true reaction asymptote) (15).
In an attempt to improve the determination of the antiradical
activity of citrus juices, Sendra et al. (6) determined the time
evolution curves of the DPPH• concentration during its reduction
by different antiradicals present in these juices. According to
the observed experimental kinetics of the reduction process in
methanol, these authors roughly grouped the antiradicals into
three main groups: fast kinetics, fast+slow kinetics, and slow
kinetics. The components belonging to the fast kinetics group,
which contain isolated vinyl-alcohols (e.g., ascorbic acid),
exhibit a single short but very fast kinetics step; the components
belonging to the fast+slow kinetics group, which contain at least
a p-catechol group (e.g., chlorogenic acid), exhibit an initial
short but fast kinetics step followed by a longer slow kinetics
step; finally, the components belonging to the slow kinetics
group, which contain isolated phenols (e.g., hesperitin), exhibit
only a single slow kinetics step.
The isolated p-catechol group can be found in several families
of naturally occurring compounds such as flavanones, isofla-
vanones, protocatechuic (3,4-dihydroxybenzoic) acid derivatives,
and chalcones among others. The scientific literature is rather
confusing about the antiradical activity of this group, which
could be due to the fact that its antiradical activity also depends
on the chemical structure of the family. For instance, the total
stoichiometric constant of the isolated p-catechol group was
determined to be about 4 in flavanone type structures (13) but
>5 in protocatechuic acid alkyl esters (16) when methanol was
used as solvent. In nonalcoholic solvents, such as ethyl acetate
or acetonitrile, the determined total stoichiometric constant was
about 2 in both cases. There seems to be consensus, however,
that the p-catechol group reduces the DPPH• by the transfer of
two hydrogen atoms and its subsequent conversion to the
corresponding o-quinone (16-18).
Determination of the Antiradical Activity. Sample Preparation.
For spectrophotometric determinations, the solvent to be used (methanol,
ethanol, propanol, or acetonitrile) was dried overnight over anhydrous
sodium sulfate, and the working solutions of the antiradical and DPPH•
were freshly prepared before analysis. A volume of the antiradical
solution (between 5 and 40 µL) was added in situ, using a chromato-
graphic syringe, into a thermostated (22 °C) and stirred (600 rpm) quartz
spectrophotometric cuvette (3.5 mL of capacity and 1 cm path length)
containing an appropriate volume of DPPH• to yield a final volume of
2 mL (the final concentration of DPPH• was around 100 µmol/L), and
the spectrophotometric cuvette was immediately end-capped again. The
analysis time commenced with the addition of the antiradical. As a
general rule, those samples yielding an asymptotic value of the DPPH•
concentration of <10% or >90% of its initial concentration were
discarded.
UV-Vis Analysis. Absorbance was measured using a model 8453
UV-Vis spectrophotometer (Agilent Technologies GmbH, Karlsruhe,
Germany) equipped with a diode array detector and a thermostated cell
holder with magnetic stirring. Operating conditions were as follows:
vis lamp, on; UV lamp, off; wavelength, 515 nm; slit width, 1 nm;
and data acquisition rate, 2.1 s/data point. Automatic acquisition of
data was stopped after a reaction time of 60-90 min, depending on
the speediness of the kinetics. Thus, each set of data contained over
1500 data points. All samples were analyzed in duplicate.
Prior to the experiments on antiradical activity, a calibration curve
of absorbance versus concentration of DPPH• in all of the assayed
solvents was obtained to determine the molar extinction coefficient (ꢀ)
of DPPH•. From the linear fitting of data, the values determined for ꢀ
were as follows: methanol, 1.09 × 104; ethanol and propanol, 1.08 ×
104; and acetonitrile, 1.06 × 104 L/(mol cm).
Determination of the Mass Spectra. Sample Preparation. For mass
spectrometry determinations, the working solutions of the antiradical
and DPPH• were freshly prepared using anhydrous methanol from a
recipient opened immediately before analysis. A volume of the
antiradical solution (between 2 and 5 µL) was added, using a
chromatographic syringe, into a cuvette containing an appropriate
volume of DPPH• to yield a final volume of 2 mL (the final
concentration of DPPH• was around 25 µmol/L). The cuvette was end-
capped immediately and shaken by hand, and then an aliquot of the
reaction mixture was transferred into the infusion syringe (500 µL
capacity) for analysis. Reaction time commenced with the mixing of
the antiradical and DPPH•. Mass spectra were manually acquired from
5 min of reaction time onward, at a sampling rate of 5 min/mass
spectrum. Data acquisition was stopped after a reaction time of about
80 min.
In the present work the time evolution of the DPPH•
concentration in four solvents (methanol, ethanol, propanol, and
acetonitrile) during its reduction by three flavanones containing
an isolated p-catechol group (taxifolin, eriodyctiol, and fustin)
has been determined by spectrophotometry, and the time
evolution of the mass spectra of the reaction mixture has been
determined by liquid mass spectrometry. From the results, a
Mass Spectrometry Analysis. Mass spectra were obtained using an
LCQ Advantage (Thermo Finnigan, San Jose, CA) mass spectrometer,
equipped with an electrospray ionization source and ion trap detector.
Instrument control and analysis of data were carried out using a PC
loaded with the LCQ Tune/Excalibur software. The sample was
introduced into the mass spectrometer by direct infusion (syringe) at a
flow rate of 5 µL/min. Operating conditions were as follows: mode,

5514 J. Agric. Food Chem., Vol. 55, No. 14, 2007
Sendra et al.
with the constraints
y2 ) y0 + ys - y1 ) yo - σ2a0
y0 - y1 ) σ1a0
(2)
y0 - ys ) (σ1 + σ2)a0
where y is the time-dependent concentration of DPPH•, y0 is
the initial concentration of DPPH•, a0 is the initial concentration
of the antiradical, t is the reaction time, k1 is the fast kinetics
rate constant, y1 is the asymptote that would be reached due
solely to the fast kinetics antiradical activity, F2 is the slow
kinetics pseudo-rate constant, y2 is the asymptote that would
be reached due solely to the slow kinetics antiradical activity,
ys is the experimental asymptote of the reaction, and σ1 and σ2
are the stoichiometric constants of the fast and slow kinetics,
respectively. The results are given in Table 1.
As can be seen in Figure 2, the fittings were excellent in all
cases (r2 > 0.999, nonlinear parametric fitting using SigmaPlot
8.02). The resulting curves consisted in a short initial fast step,
which is due to a fast kinetics with a rate constant k1, followed
by a longer slow step up to reach the asymptote of the reaction,
which is due to a slow kinetics with a pseudo-rate constant F2.
The determined value of the fast kinetics rate constant [k1 )
60.9 × 103 L/(mol min)] was greater by far than that of the
slow kinetics pseudo-rate constant [F2 ) 937 L/(mol min)], but
the determined value of both fast and slow kinetics stoichio-
metric constants (σ1 ) 1.97 and σ2 ) 2.04, respectively) was
very close to 2, indicating that the value of the total stoichio-
metric constant (σt ) σ1 + σ2 ) 4.01) of the reaction was very
close to 4.
Influence of the Solvent on the Antiradical Activity of
Taxifolin. The influence of the solvent on the reaction kinetics
was determined by reducing the DPPH• with taxifolin in four
different solvents, namely, methanol, ethanol, propanol, and
acetonitrile. Panels B, C, and D of Figure 2 show the reduction
curves in ethanol, propanol, and acetonitrile, respectively, and
data on the determined rate and stoichiometric constants are
given in Table 1. For comparison purposes, Figure 3 shows
one reduction curve per solvent when using approximately the
same initial concentrations of DPPH• and taxifolin. As can be
seen, there was a remarkable difference between the alcoholic
solvents and acetonitrile. For all of the alcoholic solvents
assayed, the short initial fast kinetics step was always followed
by a longer slow kinetics step, but this later step was completely
lacking for acetonitrile. Data in Table 1 indicate that within
the alcoholic solvents, the values of both fast rate (k1) and slow
pseudo-rate (F2) kinetics constants were dependent on the protic
power of the alcohol, in the order k1(methanol) ) 60.9 ×
103 > k1(ethanol) ) 28.1 × 103 > k1(propanol) ) 8.97 × 103
L/(mol min) and F2(methanol) ) 937 > F2(ethanol) ) 286 >
F2(propanol) ) 88 L/(mol min). On the other hand, the value
of the stoichiometric constants of both fast kinetics (σ1 ) 1.97,
2.01 and 2.03 for methanol, ethanol, and propanol, respec-
tively) and slow kinetics (σ2 ) 2.04, 1.95, and 1.92 for methanol,
ethanol, and propanol, respectively) was close to 2 for all of
the assayed alcohols. In the case of acetonitrile, in contrast, the
fast kinetics step, which was slower than those from methanol
and ethanol, was composed of the sum of two components, each
of them having a different fast rate constant [k1(1) ) 9.22 ×
103 and k1(2) ) 1.25 × 103 L/(mol min)], but the same
stoichiometric constant with a value very close to 1 [σ1(1) )
σ2(1) ) 1.03)], whereas the value of its slow kinetics pseudo-
rate and stoichiometric constants was zero (F2 ) σ2 ) 0). It
•
Figure 2. Time evolution of the DPPH concentration in (A) methanol,
(B) ethanol, (C) propanol, and (D) acetonitrile during its reduction by three
different initial concentrations of taxifolin. (Inset) Zooming of the first minute
of the reaction. See Table 1 for the initial concentrations of DPPH and
•
taxifolin in the assayed solvents.
negative; scan mode, full scan MS (m/z 200-800); number of
microscans, 5; maximum inject time, 150 ms; sheath gas (N2) flow
rate, 20 units; auxiliary/sweep flow rate, 0 units; ionization spray
voltage, 3.9 kV; capillary temperature, 160 °C; capillary voltage,
-20 V; and tube lens offset, -50 V.
RESULTS AND DISCUSSION
Figure 1 shows the chemical structures of the assayed
flavanones (taxifolin, eriodyctiol, and fustin) as well as the
structures of some referenced compounds (luteolin, quercetin,
and protocatechuic acid).
Antiradical Activity of Taxifolin in Methanol. Figure 2A
shows the time evolution of the concentration of DPPH• during
its reduction by three different initial amounts of taxifolin, as
well as a zooming of the first minute of the reaction. The sets
of experimental data points were fitted using the kinetic equation
proposed by Sendra et al. (6) for the determination of the
antiradical activity of citrus juices (see later for the derivation
of this equation from the kinetic model)
y1(y0 - y1)
y2(yo - y2)
y - ys )
+
y1 - yo(1 - e(k /σ )y t
)
y2 - yo(1 - e(F /σ )y t
)
1
1
1
2
2
2
(1)

Antiradical Activity of the p-Catechol Group
J. Agric. Food Chem., Vol. 55, No. 14, 2007 5515
•
Table 1. Rate (Mean
±
SD) and Stoichiometric (Mean
±
SD) Constant from the Reduction of DPPH in Methanol, Ethanol, Propanol, and
Acetonitrile by Taxifolin
methanola
ethanola
propanola
acetonitrilea
•
initial concentration (
µ
mol/L)
DPPH
99.872 (
O
)
)
95.305 (
94.291 (
93.346 (
O)
4)
3)
85.168 (
86.165 (
85.430 (
O)
4)
3)
100.242 (
O
)
)
)
98.755 (
4
99.186 (
4
100.464 (
3
)
98.976 (
3
taxifolin
19.31 (
16.10 (
11.59 (
O
4
3
)
)
)
16.104 (
12.078 (
O
4
)
)
)
19.46 (
15.81 (
12.16 (
O
)
32.867 (
20.542 (
12.325 (
O)
4)
3)
4
)
)
8.052 (3
3
mean rate constant [L/(mol min) from eq 27]
mean stoichiometric constants
k1
F2
k1(1)
k1(2)
(60.9
±
29
7.2)
×
103
(28.1
±
37
4.1)
×
103
(8.97
88
±
4
0.3)
×
103
937
±
286
±
±
0
(9.22
(1.25
±
±
0.6)
0.3)
×
×
103
103
σ1
σ2
1.975
2.038
±
±
0.031
0.028
2.010
1.955
±
±
0.029
0.019
2.030
1.922
±
±
0.035
0.072
0
σ
σ
1(1)
1(2)
σt
1.029
1.028
2.058
0
±
±
±
0.013
0.013
0.025
4.013
0.0434
±
0.034
3.965
0.0127
±
0.044
3.952
(3.65
±
±
0.090
0.07)
-
-
3
mean values of k2[ROH] (min 1) and
k2[ROH]
±
0.0013
±
0.0019
‚
10
k
2 [L/(mol min)] from eq 19
-3
-
4
-4
k2
(1.76
±
0.05)
×
10
(7.39
±
1.12)
×
10
(2.73
±
0.06)
×
10
0
a Symbols refer to the corresponding curves in Figure 2.
bonding from the solvent. In contrast, Litwinienko and Ingold
(20) propose that this influence is due to the dielectric constant
of the solvent which modifies the pKa of phenols. The
mechanism of the influence of the solvent on the value of the
slow kinetics pseudo-rate constant (F2) is not so explicitly
documented, but the work by Saito et al. (16) on the antiradical
activity of the p-catechol group in protocatechuic (3,4-dihy-
droxybenzoic) acid alkyl ester gives conclusive information. The
intermediate o-quinone forms an adduct with the alcoholic
solvent, regenerates the p-catechol group, and allows the reaction
to go on.
Antiradical Activity of Eriodyctiol and Fustin in Metha-
nol. To quantify the individual influence of the removal of the
hydroxyl on C3, in the C-ring, and the hydroxyl on C5, in the
A-ring, the reduction curves of DPPH• by different amounts of
eriodyctiol and fustin, respectively, were determined. The
resulting curves are shown in Figure 4, and the results are given
in Table 2. As can be seen, the removal of the hydroxyl on C3
(eriodyctiol) slightly increased the value of the fast kinetics rate
constant [k1 ) 70.2 × 103 L/(mol min)] and appreciably
decreased the value of the slow kinetics pseudo-rate constant
[F2 ) 330 L/(mol min)], but left unchanged the values of both
fast and slow kinetics stoichiometric constants (σ1 ) 2.04 and
σ2 ) 2.00, respectively). The removal of the hydroxyl on C5
(fustin) left unchanged the values of the fast kinetics rate [k1 )
59.5 × 103 L/(mol min)] and stoichiometric (σ1 ) 2.02)
constants, but appreciably increased both the values of the slow
kinetics pseudo-rate [F2 ) 1335 L/(mol min)] and stoichiometric
(σ2 ) 2.21) constants. It is unclear how a hydroxyl on the A-ring
can significantly affect the antiradical behavior of the p-catechol
in the B-ring, but it seems that in fustin there exists a limited
formation of a second adduct between a further intermediate
o-quinone and the solvent, with a partial regeneration of the
p-catechol group and the subsequent reduction of additional
DPPH•. This would explain the greater values of both the slow
kinetics pseudo-rate and stoichiometric constants of fustin when
compared with those of taxifolin. In fact, this second adduct
formation was already observed by Saito et al. (16) for the
p-catechol group in protocatechuic acid alkyl esters during the
reduction of the DPPH• in methanol. In the latter case, however,
the second adduct formation seems to be much more favored
•
Figure 3. Time evolution of the DPPH concentration in methanol (
O
),
ethanol (
taxifolin, at approximately the same initial concentrations of DPPH
94.3 mol/L) and taxifolin ( 12 mol/L).
4
), propanol (
3
), and acetonitrile (
]
) during its reduction by
•
(≈
µ
≈
µ
must be pointed out that the reduction curves for acetonitrile
were initially fitted using the single-term kinetics equation
proposed by Sendra et al. (6) for components belonging to the
fast-kinetics antiradical group (one rate/one stoichiometric
constant), but the fittings were not as good as they should be
(r2 < 0.99), indicating the presence of two different but close
rate constants. In contrast, the fittings using eq 1 were excellent
for all curves (r2 > 0.999). It seems that the theoretically
expected antiradical nonequivalence of both hydroxyls of the
p-catechol group, which could not be detected in alcoholic
solvents, becomes detectable in acetonitrile and, in general, in
nonalcoholic solvents. This could probably be due to the fact
that in acetonitrile all of the experimental data points strictly
belong to the fast kinetics, because there is no contaminating
slow kinetics, and that the fast kinetics step was significantly
slower than those from methanol and ethanol. The conjunction
of both factors probably allows the fitting to be much more
discriminating.
It is known that the solvent influences the value of the fast
kinetics rate constant (k1) in the order k1(methanol) > k1(ethanol)
> k1(propanol), although two different mechanisms have been
proposed to explain this influence. According to Valgimigli et
al. (19), the influence of the solvent is due to a differential
stabilization of the charge in the DPPH• radical by hydrogen

5516 J. Agric. Food Chem., Vol. 55, No. 14, 2007
Sendra et al.
adduct formed at the ionization source than a true reaction
product. If c was mainly a true reaction product, it would transfer
two hydrogen atoms to the DPPH•, thus decreasing with time
and yielding a significant amount of c* (c* ) c - 2, m/z 361
and 345 for taxifolin and fustin, respectively). However, because
a small amount of c* was present in both mass spectra, it could
not be disregarded that a small amount of c was the true reaction
product. To clarify this question, the ionic abundance of c* was
plotted against reaction time as shown in Figure 6. From the
result, it seems clear that the ionic abundance of c* remains
practically constant with time for taxifolin but slightly increases
for fustin, indicating that in this latter case at least a small
amount of c was the true reaction product. This result is in
accordance with the previous finding that both the slow kinetics
pseudo-rate and stoichiometric constants of fustin are greater
than those of taxifolin.
Kinetic Model. From all results mentioned above and
considering the available bibliographic information, the kinetic
model depicted in Figure 7, for the reaction of reduction of
DPPH• by the isolated p-catechol group in flavanone type
structures, is proposed.
•
Figure 4. Time evolution of the DPPH concentration in methanol during
For a clearer and more concise discussion on the proposed
model, it can be considered that the reaction of reduction is
composed of successive steps as follows.
its reduction by (A) eriodyctiol and (B) fustin. See Table 2 for the initial
concentrations of DPPH , eriodyctiol, and fustin, and the meaning of the
symbols.
•
First Step. According to this model, the first step of the
reaction is a rather fast transfer of two successive hydrogen
atoms from the p-catechol group a to the DPPH• [fast rate
constants k1(1) and k1(2)] and the subsequent transformation of
a into its corresponding o-quinone a*. It is postulated that the
antiradical nonequivalence of both hydroxyls of the p-catechol
group is not experimentally detectable in alcoholic solvents but
becomes detectable in nonalcoholic solvents. Hence, in alcoholic
solvents a cumulative fast kinetics rate constant [k1 ) k1(1) +
k1(2)] and a fast kinetics stoichiometric constant with a value
of 2 (σ1 ) 2) should experimentally be determined (i.e.,
reduction of DPPH• in alcoholic solvents by taxifolin, eri-
odyctiol, and fustin). In nonalcoholic solvents, such as aceto-
nitrile, two fast kinetics rate constants [k1(1) and k1(2)] and two
stoichiometric constants with value 1 [σ1(1) ) σ2(1) ) 1] should
experimentally be determined (i.e., reduction of DPPH• in
acetonitrile by taxifolin). Moreover, and due to the influence
of the solvent, it should experimentally be determined that k1-
(methanol) > k1(ethanol) > k1(propanol) (i.e., reduction of
DPPH• in methanol, ethanol, and propanol by taxifolin).
than in flavanone type structures, because the total stoichiometric
constant of the p-catechol group was about 5, indicating that
the value of the slow kinetics stoichiometric constant was about 3.
Time Evolution of the Mass Spectra of the Reaction
Mixture in Methanol. Figure 5 shows the mass spectra
(electrospray, negative mode, sample introduction by infusion)
of the reaction mixture acquired at 5, 20, and 60 min during
the reduction of DPPH• in methanol by taxifolin and fustin (T5,
T20, and T60 and F5, F20, and F60, respectively). Because the
reaction must be carried out into the infusion syringe, reliable
data could only be acquired from about 5 min of reaction time
onward. At this initial time, the concentration of the parent
molecule has decayed to zero and, consequently, its correspond-
ing ion (m/z 303 and 287 for taxifolin and fustin, respectively)
could not be detected. It must be observed that the ion
corresponding to the reduced DPPH• (DPPH - H, m/z 394) as
well as minor irrelevant ions corresponding to the formation of
dimmers at the ionization source (m/z 603, 633, 635, 665, and
727 for taxifolin; m/z 571, 601, 633, 663, and 695 for fustin)
has been removed from the mass spectra. If the ion correspond-
ing to the parent molecule (m/z 303 and 287 for taxifolin and
fustin, respectively) is denoted as a, it is known that the first
step of the reduction process is a fast transfer of two hydrogen
atoms from a to DPPH•. Hence, a is transformed into a* (a* )
a - 2), which is clearly present in both mass spectra (a* )
301 and 285 for taxifolin and fustin, respectively). The ion a*
decreases with time to give its adduct b with methanol (b ) a*
+ 32), which is also very well observed in both mass spectra
(b ) 333 and 317 for taxifolin and fustin, respectively). The
ion b decreases with time more slowly than a* because it is
continuously being formed from a* but also rapidly destroyed
by the fast transfer of two hydrogen atoms to the DPPH• and
its subsequent transformation into b* (b* ) b - 2), which is
clearly seen in both mass spectra (b* ) 331 and 315 for taxifolin
and fustin, respectively). The ion b* increases exponentially
with time, reaching an asymptote. Apparently, at this point the
reaction is completed because the ion c, corresponding to the
adduct between b* and methanol (c ) b* + 32, c ) 363 and
347 for taxifolin and fustin, respectively) behaves more as an
Second Step. In nonalcoholic solvents, such as acetonitrile,
the reaction of reduction is already completed, and hence there
is no further slow kinetics step (i.e., reduction of DPPH• in
acetonitrile by taxifolin). In alcoholic solvents, in contrast, the
intermediate o-quinone a* quantitatively reacts with the solvent
(slow rate constant k2) and fully regenerates the p-catechol group
b. Hence, the total concentration of b, namely [b], coincides
with the total concentration of a and a*, that is, [b] ) [a*] )
[a]. According to Saito et al. (16) the alkoxy group enters on
the position C2′, in the B-ring.
Third Step. The regenerated p-catechol group b rapidly
transfers two hydrogen atoms to the DPPH• [fast rate constants
k3(1) and k3(2)], yielding the corresponding o-quinone b*. It is
evident that this step cannot experimentally be detected as a
fast kinetics step, with a rate constant k3 [k3 ) k3(1) + k3(2)],
but as a slow decreasing step with a rate constant of value close
to k2, because this slow rate constant is the limiting one. If the
reaction of reduction is completed at this point, then a
stoichiometric constant of value 2 (σ2 ) 2) should experimen-
tally be detected (i.e., reduction of DPPH• in methanol by

Antiradical Activity of the p-Catechol Group
J. Agric. Food Chem., Vol. 55, No. 14, 2007 5517
•
Table 2. Rate (Mean
±
SD) and Stoichiometric (Mean
mol/L)
±
SD) Constants from the Reduction of DPPH in Methanol by Eriodyctiol and Fustin
•
•
DPPH
eriodyctiola
DPPH
fustina
initial concentration (
µ
97.785 (
O
)
)
18.04 (
O
)
)
93.678 (
93.772 (
94.710 (
O)
4)
3)
13.80 (
10.35 (
O
4
)
)
)
97.952 (
4
13.53 (4
6.90 (3
average rate constant [L/(mol min)]
average stoichiometric constants
k1
F2
(70.2
±
5
1.5)
×
103
(59.5
1335 ±
±
4.2)
49
×
103
330
±
σ1
σ2
σt
2.042
1.996
4.037
±
±
±
0.052
0.001
0.052
2.026
2.206
4.232
±
±
±
0.057
0.054
0.076
a Symbols refer to the corresponding curves in Figure 4.
•
Figure 5. Mass spectra of the reaction mixture at 5, 20, and 60 min during the reduction of DPPH (24.23
and T60, respectively) and fustin (3.21 mol/L, F5, F20, and F60, respectively).
µmol/L) by taxifolin (3.04 µmol/L, T5, T20,
µ
given for the third step, a slow kinetics rate constant of value
close to k4 and a stoichiometric constant of value smaller than
2 (σ3 < 2) should experimentally be detected. However, taking
into account that it seems rather difficult to extract accurate
values for k2, k4, σ2, and σ3 by means of a fitting, the
experimentally determined value of the slow kinetics pseudo-
rate constant, F2(exp), should be, in these cases, close to the
sum of the corresponding F2 and F4 [F2(exp) ) F2 + F4], and
the experimentally determined value of slow kinetics stoichio-
metric constant, σ2(exp), the sum of σ2 and σ3 [σ2(exp) ) σ2 +
σ3 ) 2 + σ3] (i.e., reduction of DPPH• in methanol by fustin).
It must be noted that the value of σ3 can be any number between
0 and 2 (0 e σ3 e 2) because this value depends on only [c],
that is, on the extent of the adduct formation between b* and
the solvent. In fact, the value of σ3 [σ3 ) σ2(exp) - 2] allows
an approximate determination of [c], given that [c] ) (σ3/2) ×
[a] ) [(σ2(exp) - 2)/2] × [a].
Figure 6. Time evolution of the abundance of the ion c* during the
•
reduction of DPPH by taxifolin (
b
, m/z 361) and fustin (
O
, m/z 345).
taxifolin and eriodyctiol; reduction of DPPH• in ethanol and
propanol by taxifolin). On the other hand, because the value of
the slow rate constant k2 depends on the nucleophilic power of
the alcohol in the order methanol > ethanol > propanol, and
taking into account that F2 and k2 are directly correlated (see
later), it should experimentally be found that F2(methanol) >
F2(ethanol) > F2(propanol) (i.e., reduction of DPPH• in metha-
nol, ethanol, and propanol by taxifolin).
Fourth Step. A small amount of the intermediate o-quinone
b* reacts with the solvent (slow rate constant k4) and partially
regenerates the p-catechol group c. Because the extent of this
reaction is very limited in flavanone type structures, it is fulfilled
that [c] < [a].
Sixth Step. For structures other than flavanones, the possible
formation of a further and final adduct between c* and the
solvent to regenerate the p-catechol group, with the subsequent
reduction of additional DPPH•, cannot be disregarded.
Finally, it is important to emphasize that although this model
envisages many rate and stoichiometric constants, the antiradical
activity of the isolated p-catechol group in flavanone type
structures behaves experimentally as a very simple “active
antiradical group” exhibiting two rate and two stoichiometric
constants [k1, F2(exp), σ1, and σ2(exp)]. This is of particular
importance because it will allow the deduction of two different
but equivalent explicit kinetic equations for the antiradical
activity of this group against DPPH•.
Fifth Step. The partially regenerated p-catechol group c
transfers two hydrogen atoms to the DPPH• [fast rate constants
k5(1) and k5(2)], yielding the corresponding o-quinone c*.
Taking into account that [c] < [a] as well as the arguments

5518 J. Agric. Food Chem., Vol. 55, No. 14, 2007
Sendra et al.
•
Figure 7. Kinetic model for the antiradical activity of the isolated p-catechol group in flavanone type structures using DPPH as the antiradical probe.
Because the p-catechol group in flavanone type structures
behaves experimentally as an “active antiradical group” exhibit-
ing two rate and two stoichiometric constants, the kinetic model
can reasonably be simplified to
k1
y + a
9
8 a* + y - H
(3)
(4)
(5)
k2
98 b
a* + ROH
k3
y + b
98 b* + y - H
The set of the corresponding differential equations has no
analytical solution, and so it makes no sense to set it out. Instead,
our aim is to deduce an approximate but accurate explicit
analytical solution. To achieve this, it is important to observe
that the rate constant k1 is greater by far than the rate constant
k2 (k1 . k2). This allows uncoupling of reaction 3 (fast kinetics)
from the reactions 4 and 5 (slow kinetics), solving them
separately to obtain two terms (one for the fast kinetics and the
other for the slow kinetics), and then coupling both terms. To
uncouple reaction 3 from reactions 4 and 5, it is assumed that
they are not simultaneous but successive or, in other words,
that reaction 3 is so rapid when compared with reaction 4 that
the latter does not start until the former is completed. Hence,
reaction 4 starts after a very short time, t1, at which y in reaction
3 has already reached its asymptotic value y1. To couple both
solutions, t1 in the second term (slow kinetics) must tend toward
zero.
Figure 8. Time contribution and asymptote of the fast kinetics (small
open circles) and slow kinetics (small open triangles) terms describing
the fast kinetics (large open circles) and slow kinetics (large open triangles)
•
steps, respectively, of the reduction curve of DPPH (100.464
µmol/L) in
methanol by taxifolin (11.59
µmol/L).
Deduction of the Explicit Kinetic Equations from the
Kinetic Model. Figure 8 shows the reduction curve of the
DPPH• (100.46 µmol/L) in methanol by taxifolin (11.59 µmol/
L), where the curve has been approximately divided into its
fast (large blank circles) and slow (large blank triangles) kinetics
steps, and the asymptotes corresponding to the fast (y1), slow
(y2), and total (ys) kinetics are shown as long dash lines.

Antiradical Activity of the p-Catechol Group
J. Agric. Food Chem., Vol. 55, No. 14, 2007 5519
From eq 3
Taking into account that it is fulfilled
σ2a0 ) y1 - ys
eq 14 can be rewritten as
dy
) -k1ya
dt
(6)
with the boundary conditions at t ) 0, y ) y0, a ) a0, and a*
) 0; and at t ) ∞ (or t > t1), y ) y1, a ) 0, and a* ) a0. Let
σ1 be the stoichiometric constant of the reaction, then (6)
y - ys ) (y1 - ys) e-k [ROH](t-t )
(15)
2
1
which would be the uncoupled term describing the slow kinetics
step of the reaction.
y0 - y1 y - y1
y - y1
σ1
σ1 )
)
f a )
(7)
To couple reactions 3 and 4 and make them simultaneous,
the value of t1 in eq 15 must tend toward zero. As shown in
Figure 8, when t1 tends toward zero in eq 15, the value of its
y1 tends toward y0 and the value of its ys must tend toward a
value y2 (the asymptotic value of the small open triangles) such
that it must be fulfilled
a0
a
and eq 6 becomes
k1
) - y(y - y1)
σ1
dy
dt
(8)
y1 - ys ) y0 - y2 f y2 ) yo + ys - y1
and so eq 15 becomes
(16)
the solution of which, in its most explicit form, with the given
boundary conditions, is well-known (6)
y1(y0 - y1)
y - y2 ) (y0 - y2) e-k [ROH]t ) (y1 - ys) e-k [ROH]t
2
2
y - y1 )
(9)
y1 - y0(1 - e(k /σ )y t
)
1
1
1
(17)
with the constraint for the value of y2 given by eq 16. Hence,
eq 17 is the coupled term (small open triangles in Figure 8)
describing the slow kinetics step of the reaction.
and corresponds to the term (small blank circles in Figure 8)
describing the fast kinetics step of the reaction.
From eqs 4 and 5
Taking into account that for the full reaction
da*
dt
) -k2a*[ROH]
(10)
(11)
y0 - ys ) y0 - y1 + y1 - ys ) y0 - y1 + yo - y2 (18)
the approximate kinetic equation describing this reaction is given
by
dy
) -k3yb
dt
y1(y0 - y1)
+ (y1 - ys) e-k [ROH]t
where [ROH] is the molar concentration of the alcohol and with
the following boundary conditions: at t ) t1, a* ) a0, y ) y1,
and b ) b* ) 0 and at t ) ∞, a* ) b ) 0, y ) ys, and b* )
a0, ys being the experimental asymptote of the reaction. At any
time t g t1 it is fulfilled that a0 ) a* + b + b*, that is, b )
a0 - b* - a*. On the other hand, let σ2 be the stoichiometric
constant of the reaction. Then, at any time t g t1 it is fulfilled
that (a0 - b*) ) (y - ys)/σ2 and consequently
2
y - ys )
y1 - y0(1 - e(k /σ )y t
)
1
1
1
(19)
with the constraints
y0 - y1 ) σ1a0
y0 - y2 ) y1 - ys ) σ2a0
y0 - ys ) (σ1 + σ2)a0
b ) (y - ys)/σ2 - a*
(12)
The value of a* (a simple decaying exponential) can easily be
deduced from eq 10
Second Approximation. Taking into account that at any time
of the reaction process the concentration of b is smaller than
that of a*, depends directly on the concentration of a*, is almost
independent of the concentration of y, because y is in excess
and k3 . k2, and its time evolution is similar to that of a*, it
seems reasonable to assume that b ≈ Ka*, where K , 1.
Because the concentration of b is directly dependent on the rate
constant k2 but inversely dependent on the rate constant k3, K
must also fulfill these same dependences and be also de-
pendent on the solvent used, that is, K ) K(k2, k3, [ROH]).
Hence
a* ) a0 e-k [ROH]t
(13)
2
and substituted into eq 12 and the result substituted into eq 11,
but the resulting differential equation, due to the exponential
term, has no analytical solution. To overcome this difficulty
there are two fine approximations.
First Approximation. Taking into account that the concentra-
tion of b is very small at any time of the reaction process,
because it is continuously and slowly formed from a* (slow
rate constant k2) but continuously and rapidly destroyed by the
transfer of two hydrogen atoms to the DPPH• (fast rate constant
k3), it would be reasonable to assume that its concentration
remains practically constant with time, similarly to the concen-
tration of the enzyme-substrate complex in simple Michaelis-
Menten kinetics, and consequently db/dt ≈ 0. Derivation of eq
12 leads to the immediate solution
1
K
b ≈ Ka* f a* ≈
b
(20)
Substitution of eq 20 into eq 12 gives
(y - ys)
K
1 + K
b )
(21)
σ2
y - ys ) σ2a0 e-k [ROH](t-t )
(14)
2
1
and substitution of eq 21 into eq 11 gives

5520 J. Agric. Food Chem., Vol. 55, No. 14, 2007
Sendra et al.
antiradical kinetics, mainly modulated by the rate constant (k2)
of formation of the adduct, and thus fitted using a simple
decaying exponential, which corresponds to a pseudo-first-order
reaction. This allows the determination of the slow rate constant
of the adduct formation (k2) and the stoichiometric constant of
the pseudo-slow kinetics (σ2). However, this approximation,
although very close to “reality”, has the frustrating drawback
that it does not allow a direct comparison of the speediness of
both kinetics by direct comparison of the rate constants k1 and
k2, because they differ by >7 orders of magnitude.
Kk3
dy
dt
1
) -
y(y - ys)
(22)
1 + K σ2
If a pseudo-rate constant F2 is defined as
Kk3
F2 )
(23)
(24)
1 + K
eq 22 becomes
F
Third, the second approximation (eq 27) envisages the full
reaction of reduction as “it behaves as”, that is, as the
conjunction of two true antiradical kinetics, the former being
fast and the latter slow. Similarly to eq 19, the fast step of the
reduction curve is considered to be due to a true fast antiradical
kinetics with a fast rate constant k1, and thus fitted using the
mathematical function derived from a second-order reaction.
This allows the determination of the rate (k1) and stoichiometric
(σ1) constants corresponding to the fast kinetics. In contrast to
eq 19, the slow step of the reduction curve is also considered
to be due to a true slow antiradical kinetics with a slow pseudo-
rate constant F2, and thus fitted using the same mathematical
function. This allows the determination of the slow kinetics
stoichiometric (σ2) and the pseudo-rate (F2) constants. This
approach clearly does not reflect “reality”, because the p-
catechol group is considered to have two different and inde-
pendent “antiradical groups”, one exhibiting a fast kinetics and
the other a slow kinetics (due to the adduct formation, evidently),
and so the value of the slow pseudo-rate constant F2 does not
correspond to any of the reactions of the reduction process.
However, this approximation has an advantage of vital impor-
tance, because it allows the immediate comparison of the
speediness of both kinetics by direct comparison of their
corresponding rate (k1) and pseudo-rate (F2) constants. It must
be emphasized, on the other hand, that this direct comparison,
although quite important for isolated compounds, becomes
essential to evaluate the antiradical activity of mixtures of
antiradicals.
dy
dt
) - 2 y(y - ys)
σ2
which is similar to eq 8 and its solution, taking into account
the boundary conditions, is given by
ys(y1 - ys)
ys - y1 [1 - e(F /σ )y (t-t )
y - ys )
(25)
2
2
s
1
]
which is the uncoupled term describing the slow kinetics step
of the reaction.
As in the case of the first approximation, to couple reactions
3 and 4 and make them simultaneous, the value of t1 in eq 25
must tend toward zero. Hence, by following exactly the same
considerations as above, it can be deduced that the coupled term
is given by
y2(y0 - y2)
y2 - y0(1 - e(F /σ )y t
y - y2 )
(26)
2
2
2
)
and the approximate kinetic equation describing this reaction
is given by
y1(y0 - y1)
y2(y0 - y2)
y2 - y0(1 - e(F /σ )y t
y - ys )
+
y1 - y0(1 - e(k /σ )y t
)
)
1
1
1
2
2
2
(27)
As indicated previously, the values of F2 and k2 are directly
with the constraints
correlated. Table 1 gives the adjusted values of k2[ROH] (min-1
)
and k2 [L/(mol min)] for the reduction of DPPH• by taxifolin
in methanol, ethanol, and propanol (Figure 2A,B,C, respec-
tively), using eq 19 for fitting. The resulting data demonstrate
that there exists a linear relationship between F2 and k2[ROH]
(F2 ) mk2[ROH], m ) 21695 and r2 ) 0.9994). Moreover,
taking into account the definition of F2 (eq 23)
y2 ) y0 + ys - y1 ) y0 - σ2a0
y0 - y1 ) σ1a0
y0 - ys ) (σ1 + σ2)a0
which coincides with eq 1.
In this work, eq 27 has been preferred to eq 19 for fitting the
experimental data points, due to the reasons explained below.
First, both eqs 19 and 27 yield practically the same fitting in
all cases, with excellent coefficients of correlation (r2 > 0.999,
although the values from eq 27 are slightly better than those
from eq 19). The adjusted values of the parameters k1, σ1, and
σ2 are also practically identical from both fittings. However,
the adjusted values of the slow kinetics rate (k2, from eq 19)
and the pseudo-rate (F2, from eq 27) constants are extremely
different, differing by >5 orders of magnitude.
Second, the first approximation (eq 19) envisages the full
reaction of reduction as “it really is”. The fast step of the
reduction curve is considered to be due to a true fast antiradical
kinetics with a fast rate constant k1 and thus fitted using the
mathematical function derived from a second-order reaction.
This allows the determination of the rate (k1) and stoichiometric
(σ1) constants corresponding to the fast kinetics. The slow step
of the reduction curve is considered to be due to a pseudo-slow
Kk3
mk2[ROH]
) mk2[ROH] f K )
1 + K
k3 - mk2[ROH]
that is, K was directly dependent on k2[ROH] and inversely
dependent on k3, and its value, determined by assuming that k3
≈ k1, was 0.0157, 9.87 × 10-3, and 8.90 × 10-3 for methanol,
ethanol, and propanol, respectively, that is, K , 1 in all cases,
as was expected. On the other hand, the determined values for
the rate constant k2 [1.757 × 10-3, 7.391 × 10-4, and 2.726 ×
10-4 L/(mol min) for methanol, ethanol, and propanol, respec-
tively] were very small and in the range of the published values
of this rate constant for the adduct formation of quinomethanes
(21).
Finally, it must be emphasized that neither eq 27 nor eq 19
is a general kinetics equation for determining the antiradical
activity of the p-catechol group in any chemical structure. On
the contrary, they are rather specific and should be used only

Antiradical Activity of the p-Catechol Group
J. Agric. Food Chem., Vol. 55, No. 14, 2007 5521
to determine the antiradical activity of “active antiradical
groups” or “antiradical mixtures” that experimentally exhibit
two rate and two stoichiometric constants. This explains why
eq 27 works very well for determining the antiradical activity
of citrus juices (6), because due to their high content in ascorbic
acid, their cumulative antiradical activity experimentally behaves
as an antiradical group exhibiting two rate and two stoichio-
metric constants (average constants in this case). This explains
also why eq 27, or eq 19, should not be used for determining
the antiradical activity of the p-catechol group in flavone type
structures (e.g., luteolin). Preliminary but consistent data indicate
that the presence of a double bond between C2 and C3, in the
C-ring, influences the antiradical kinetics of the p-catechol group
in such a way that it experimentally behaves as an antiradical
group, exhibiting discernible three rate and three stoichiometric
constants.
σ1(1)
σ1(2)
σ2
stoichiometric constant () 1) of the first component
of the fast kinetics in nonalcoholic solvents
stoichiometric constant () 1) of the second compo-
nent of the fast kinetics in nonalcoholic solvents
fraction of the slow kinetics stoichiometric constant
() 2) due to the regenerated p-catechol group from
the first adduct
σ3
fraction of the slow kinetics stoichiometric constant
(0 e σ3 e 2) due to the partially regenerated
p-catechol group from the second adduct
σ2(exp) total stoichiometric constant of the slow kinetics
() σ2 + σ3 ) 2 + σ3)
σt
total stoichiometric constant [) σ1 + σ2(exp)]
LITERATURE CITED
ABBREVIATIONS USED
(1) 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.
(2) Natella, F.; Nardini, M.; Di Felice, M.; Saccini, C. Benzoic and
cinnamic acid derivatives as antioxidants: structure-activity
relation. J. Agric. Food Chem. 1999, 47, 1453-1459.
(3) Jimenez-Escrig, A.; Jimenez-Jimenez, I.; Sanchez-Moreno, C.;
Saura-Calixto, F. Evaluation of free radical scavenging of dietary
carotenoids by the stable radical 2,2-diphenyl-1-picrylhydrazyl.
J. Sci. Food Agric. 2000, 11, 1686-1690.
(4) Sang, S. M.; Cheng, X. F.; Stark, R. E.; Rosen, R. T.; Yang, C.
S.; Ho, C. T. Chemical studies on antioxidant mechanism of tea
catechins: analysis of radical reaction products of catechins and
epicatechin with 2,2-diphenyl-1-picrylhydrazyl. Bioorg. Med.
Chem. 2002, 10, 2233-2237.
(5) Ka¨hko¨nen, M. P.; Heinonen, M. Antioxidant activity of antho-
cyanins and their aglycons. J. Agric. Food Chem. 2003, 51, 628-
633.
(6) Sendra, J. M.; Sentandreu, E.; Navarro, J. L. Reduction kinetics
of the free stable radical 2,2-diphenyl-1-picrylhydrazyl (DPPH•)
for determination of the antiradical activity of citrus juices. Eur.
Food Res. Technol. 2006, 223, 615-624.
(7) Trush, M. A.; Kensler, T. W. Role of free radicals in carcinogen
activation. In OxidatiVe Stress. Oxidants and Antioxidants; Sies,
H., Ed.; Academic Press: San Diego, CA, 1991; pp 277-
318.
(8) Haliwell, B.; Gutteridge, J. M. C.; Cross, C. E. Free radicals,
antioxidants and human diseases. Where are we now?. J. Lab.
Clin. Med. 1992, 119, 598-620.
(9) Mate´s, J. M.; Peres-Gomez, C.; Nun˜ez, de Castro, I. Antioxidant
enzymes and human diseases. Clin. Biochem. 1999, 32, 595-
603.
(10) Baghurst, K. Report: the health benefits of citrus juices. CSIRO
Health Sci. Nutr. 2003.
a
time-dependent concentration (µmol/L) of the anti-
radical
a0
DPPH•
initial concentration (µmol/L) of the antiradical
2,2-diphenyl-1-picrylhydrazyl
k1
cumulative fast kinetics rate constant [) k1(1) +
k1(2)] [L/(mol min) in alcoholic solvents]
k1(1)
k1(2)
k2
first component [L/(mol min)] of the fast kinetics
rate constant in nonalcoholic solvents
second component [L/(mol min)] of the fast kinetics
rate constant in nonalcoholic solvents
rate constant [L/(mol min)] of the first adduct
formation in alcoholic solvents [in eq 19, fraction
of the cumulative pseudo-slow kinetics rate con-
stant [L/(mol min)] due to the regenerated p-
catechol group from the first adduct]
k4
rate constant [L/(mol min)] of the second adduct
formation in alcoholic solvents [in eq 19, fraction
of the cumulative pseudo-slow kinetics rate con-
stant [L/(mol min)] due to the regenerated p-
catechol group from the second adduct
t
time (min)
y
time-dependent concentration (µmol/L) of DPPH•
initial concentration (µmol/L) of DPPH•
DPPH• concentration asymptote (µmol/L) that would
be reached due solely to the antiradical activity
of the fast kinetics
y0
y1
y2
DPPH• concentration asymptote (µmol/L) that would
be reached due solely to the antiradical activity
of the slow kinetics
(11) Parke, D. W. Nutritional antioxidants and disease prevention:
mechanisms of action. In Antioxidants in Human Health and
Disease; Basu, T. K., Temple, N. J., Garg, M. L., Eds.; CAB
International: Wallingford, Oxon, U.K., 1999; pp 1-13.
(12) Benavente-Garc´ıa, O.; Castillo, J.; Mar´ın, F. R.; Ortun˜o, A.; del
R´ıo, J. A. Uses and properties of Citrus flavonoids. Review. J.
Agric. Food Chem. 1997, 45, 4505-4515.
(13) Tsimogiannis, D. I.; Oreopoulou, V. Free radical scavenging and
antioxidant activity of 5,7,3′,4′-hydroxy-substituted flavonoids.
InnoVatiVe Food Sci. Emerg. Technol. 2004, 5, 523-528.
(14) Blois, M. S. Antioxidant determination by the use of a stable
free radical. Nature 1958, 181, 1199-1200.
ys
experimental DPPH• concentration asymptote (µmol/
L)
F2
in eq 27, first component [L/(mol min)] of the slow
kinetics pseudo-rate constant in nonalcoholic
solvents
F4
in eq 27, second component [L/(mol min)] of the
slow kinetics pseudo-rate constant in nonalcoholic
solvents
F2(exp) in eq 27, cumulative slow kinetics pseudo-rate
constant () F2 + F4) [L/(mol min)] in alcoholic
solvents
(15) Brand-Williams, W.; Cuvelier, M. E.; Berset, C. Use of a free
radical method to evaluate antioxidant activity. Lebensm. Wiss.
Technol. 1995, 28, 25-30.
σ1
total stoichiometric constant of the fast kinetics
[) σ1(1) + σ1(2) ) 2] in alcoholic solvents

5522 J. Agric. Food Chem., Vol. 55, No. 14, 2007
Sendra et al.
(16) Saito, S.; Okamoto, Y.; Kawabata, J. Effects of alcoholic solvents
on antiradical abilities of protocatechuic acid and its alkyl esters.
Biosci., Biotechnol., Biochem. 2004, 68, 1221-1227.
(17) Dangles, O.; Fargeix, G.; Dufour, C. One-electron oxidation of
quercetin and quercetin derivatives in protic and non protic
media. J. Chem. Soc., Perkin Trans. 2 1999, 1387-1395.
(18) Pannala, A. S.; Chan, T. S.; O’Brien, P. J.; Rice-Evans, C. A.
Flavonoids B-ring chemistry and antioxidant activity: fast
reaction kinetics. Biochem. Biophys. Res. Commun. 2001, 282,
1161-1168.
(19) Valgimigli, L.; Banks, J. T.; Ingold, K. U.; Lusztyk, J. Kinetic
solvent effects on hydroxylic hydrogen atom abstractions are
independent of the nature of the abstracting radical. Two extreme
tests using vitamin E and phenol. J. Am. Chem. Soc. 1995, 117,
9966-9971.
(20) Litwinienko, G.; Ingold, K. U. Abnormal solvent effects on
hydrogen atom abstractions. I. The reactions of phenols with
2,2-diphenyl-1-picrylhydrazyl (dpph•) in alcohols. J. Org. Chem.
2003, 68, 3433-3438.
(21) Dyall, L. K.; Winstein, S. Nuclear magnetic resonance spectra
and characterization of some quinone methides. J. Am. Chem.
Soc. 1972, 94, 2196-2199.
Received for review March 3, 2007. Revised manuscript received May
2, 2007. Accepted May 2, 2007. This research was supported by the
Ministerio de Educacio´n y Ciencia (Spain), Project AGL2006-05809ALI,
FEDER funds, and AGROALIMED (Conseller´ıa de Agricultura, Pesca
I Alimentacio´, Generalitat Valenciana, Spain).
JF070689S