Kinetics of the reaction between the antioxidant Trolox and the free radical DPPH? in semi-aqueous solution

Article abstract of DOI:10.1039/b602147f

Reaction of the free-radical diphenylpicrylhydrazyl (DPPH) with Trolox (TrOH) was investigated in buffered hydroalcoholic media by using a stopped-flow system. DPPH was reduced to the hydrazine analogue DPPH-H with a measured stoichiometry of about 2. DPPH-H was characterized by an acid-base equilibrium (pK_a = 8.6). Time-resolved absorption spectra recorded with an excess of either TrOH or DPPH indicated that no significant amount of the TrO radical was accumulated. The TrO radical formed in a first step further reacted quickly with DPPH. For 1: 1 ethanol-buffer mixtures at pH 7.4, the bimolecular rate constants of the first and second steps were 1.1 × 10⁴ M^-1 s^-1 and 2 × 10⁶ M ^-1 s^-1, respectively. A significant increase of the measured rate constant was observed for ethanol-buffer solutions as compared to ethanol. The rate was also increased at higher pH. A deuterium isotopic effect of 2.9 was measured. These data are discussed with regards to mechanisms involving either electron or proton exchange as rate determining steps in the reaction of DPPH with Trolox. The importance of solvent acidity control in investigation of antioxidant properties is outlined. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b602147f

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
R
Kinetics of the reaction between the antioxidant Troloxꢀ and the free radical
DPPH• in semi-aqueous solution†
Ouided Friaa and Daniel Brault*
Received 13th February 2006, Accepted 25th April 2006
First published as an Advance Article on the web 12th May 2006
DOI: 10.1039/b602147f
•
R
Reaction of the free-radical diphenylpicrylhydrazyl (DPPH ) with Troloxꢀ (TrOH) was investigated in
buffered hydroalcoholic media by using a stopped-flow system. DPPH• was reduced to the hydrazine
analogue DPPH-H with a measured stoichiometry of about 2. DPPH-H was characterized by an
acid–base equilibrium (pKa = 8.6). Time-resolved absorption spectra recorded with an excess of either
TrOH or DPPH• indicated that no significant amount of the TrO• radical was accumulated. The TrO•
radical formed in a first step further reacted quickly with DPPH•. For 1 : 1 ethanol–buffer mixtures at
pH 7.4, the bimolecular rate constants of the first and second steps were 1.1 × 104 M−1 s−1 and 2 ×
106 M−1 s−1, respectively. A significant increase of the measured rate constant was observed for
ethanol–buffer solutions as compared to ethanol. The rate was also increased at higher pH. A
deuterium isotopic effect of 2.9 was measured. These data are discussed with regards to mechanisms
involving either electron or proton exchange as rate determining steps in the reaction of DPPH• with
R
Troloxꢀ. The importance of solvent acidity control in investigation of antioxidant properties is outlined.
phenol-like groups. This mechanism has gained support from
Introduction
the deuterium kinetic isotope effect8,9 and the retardation of the
Scavenging deleterious radical species and redox status regulation
reaction in solvents establishing a hydrogen bond with the H-atom
are essential processes in normal cellular life.1,2 Pathological states,
and to some extent ageing, appear to be related to a deficiency in
regulation leading to oxidative damage, a phenomenon known
as oxidative stress.3 Cells have set up various lines of defense,
in particular to protect their membranes. Vitamin E inhibits
chain peroxidation of polyunsaturated lipids by scavenging free-
radicals.4 It is present as a-tocopherol (a-TOH) in cell membranes
where it is anchored by a lipophilic chain. This antioxidant,
belonging to the hydroxychroman class, finds its efficacy in the
reactivity of its phenolic group according to the overall reaction,
of the phenol group.10 This mechanism, referred to as HAT in
recent literature,11,12 is likely to be favored in apolar media such as
the lipidic core of membranes, which can be mimicked by alkanes.10
Alternatively, antioxidants may react through electron transfer
concerted with, or followed by, proton transfer.13,14 Besides, owing
to its lower redox potential, the deprotonated phenol form is much
more reactive towards various oxidants.15,16 Then, an alternative
mechanism assumes that ionization of the phenol group precedes
electron exchange12,17,18 and was named11,19 sequential proton loss
electron transfer (SPLET). This mechanism, favored in ionizing
solvents, was established by the addition of acid to the reaction
medium in order to suppress phenol ionization and by kinetic
analysis.12,19 It was also pointed out that adventitious acids or
bases present in the solvent, or acidification by the antioxidant
itself when it bears carboxylic chains, have important effects
on the antioxidant reaction rate.12,17 Theoretical calculations
of bond dissociation enthalpy and ionization potentials as a
measure of hydrogen versus electron transfer capability of various
antioxidants place a-tocopherol not only as the best hydrogen
donor, but also in a good position as an electron donor.20 In the
a-TOH + LOO• (or L•) → a-TO• + LOOH (or LH)
(1)
and to the relative stability of the a-tocopherol radical formed.5
Indeed, a-TO• may recombine or disproportionate through reac-
tion with another a-TO• radical, or may react with another LOO•
radical. Thus, radical-chain peroxidation is stopped through the
formation of non-radical products. Alternatively, a-TOH may be
regenerated by reaction with another antioxidant, such as ascorbic
acid.6 Deleterious prooxidant conditions may develop when the a-
tocopherol radical is not inactivated.7 The importance of these re-
actions has prompted numerous studies to understand the reaction
mechanism of hydroxychromans and other antioxidants belonging
to the phenol and flavonoid classes. Antioxidants may react by a
direct hydrogen atom transfer involving the labile hydrogen of
R
case of Troloxꢀ, the ionization of the carboxylic group in a water
medium might favor electron transfer. Clearly, the study of such
reactions requires proton activity control.
Various methods have been designed to measure antioxidant
power.21 Some involve production of transient radical species
under steady state22 or pulse conditions.13,23 Besides, stable organic
radicals leading to color changes upon reaction with antioxidants
have received much attention.24–26 The commercially available 2,2-
diphenyl-1-picrylhydrazyl radical (DPPH•) is becoming widely
used.12,17,27 It is reduced by antioxidants to the hydrazine form
(DPPH-H). The colorimetric DPPH• method is particularly useful
Laboratoire de Biophysique Mole´culaire Cellulaire et Tissulaire (BIOMO-
CETI) CNRS UMR 7033, Universite´ Pierre et Marie Curie, Genopole
Campus 1, 5 rue Henri Desbrue`res, 91030, Evry cedex, Paris, France.
E-mail: dbrault@ccr.jussieu.fr; Fax: +33 1 69 87 43 60
† Electronic supplementary information (ESI) available: Absorption spec-
tra taken in pH 6.4 and 8.4 buffers and in pure ethanol. See DOI:
10.1039/b602147f
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2417–2423 | 2417
©

View Online
to determine stoichiometric data, i.e. the number of radicals
inactivated by one molecule of antioxidant. However, as protection
against oxidative stress relies on a dynamic competition between
the scavenging of radicals and their harmful reactions, kinetic data
are most precious.28
R
Troloxꢀ, a water-soluble a-tocopherol analogue with a car-
boxylic group replacing the lipophilic tail23 is frequently used as an
antioxidant standard. Obviously, sound knowledge of its reaction
mechanism is required in order to use it as a reference.
In a continuation of the above-mentioned studies aimed to
clarify the reaction mechanism of phenol-type antioxidants, we
R
report on a UV-visible study of the reaction of Troloxꢀ with
DPPH• in ethanol–buffer mixtures allowing pH control. Focus
is given on kinetic aspects. The influence of water content and
deuterium isotopic effects are examined.
Fig. 2 (a) Absorption spectra of DPPH-H (8.3 × 10−5 M, 1 : 1
ethanol–buffer solutions) at various pH: 6.1 (bottom), 6.4, 6.7, 6.9, 7.2,
7.5, 7.8, 8.2, 8.4, 8.7, 9.0, 9.3, 9.6 (top). (b) Absorbance at 426 nm versus
pH. The full line corresponds to the theoretical fit (see text) for a pK value
of 8.59.
Results
Solution properties of compounds under study
Absorption spectra of TrOH, DPPH• and DPPH-H in 1 : 1
mixtures of ethanol with water buffered at pH 7.4 are reported
in Fig. 1. Similar spectra were recorded for 1 : 1 mixtures buffered
with water at pH 6.4 and 8.4 except for DPPH-H, which displays
a pH dependent spectrum as reported below. The spectrum of
DPPH-H at pH 8.4 is also depicted in Fig. 1.
appears. Nice isosbestic points are observed at 365 and 302 nm.
These changes suggest an acid–base equilibrium according to:
DPPH-H ꢀ DPPH− + H+
(2)
The absorbance at 426 nm (Abs) was fitted by using non-linear
curve method according to the equation:
Abs = [AbsAH + AbsA × 10(pH–pK)]/(1 + 10(pH–pK)
)
where AbsAH and AbsA are the absorbances of the pure acid
(DPPH-H) and basic (DPPH−) forms, respectively. The fit dis-
played in Fig. 2b allowed us to derive a pK of 8.59 0.03, a value
in good agreement with literature data.29
Reaction of DPPH• with TrOH
Either TrOH or DPPH• were used in excess in order to analyze
kinetics under pseudo-first order conditions.30
TrOH in excess
Fig. 1 Absorption spectra of TroloxꢀR (· · ·), DPPH• (—), DPPH-H (---)
in 1 : 1 ethanol–pH 7.4 buffer mixture and DPPH-H (— —) in the same
solution, except that the buffer was adjusted to 8.4.
A typical set of spectra recorded after mixing solutions of DPPH•
(7.7 × 10−5 M after mixing) and TrOH (7.75 × 10−4 M after
mixing) in 1 : 1 ethanol–buffer mixtures at pH 7.4 is displayed in
Fig. 3a. The reaction was completed within 2.4 s. It was found to
correspond to the consumption of all the DPPH• present to yield
the DPPH-H form, as ascertained by comparison with spectra of
the pure compounds.
Solutions made with buffers at pH 6.4 yielded about the same
sets of spectra (see the ESI). The spectral changes recorded
with buffer at pH 8.4 (see the ESI) show a larger absorbance
increase of around 420 nm. This was due to the contribution
of the deprotonated form of the hydrazine, DPPH−, according
to the above-mentioned acid–base equilibrium (eqn 2). This is
believed to be established quickly, as is usual for proton exchange
reactions.31
Solutions of TrOH, DPPH• and DPPH-H were found to obey
Beer’s law at least up to 2 × 10−3, 1.4 × 10−4 and 1.1 × 10−4 M,
respectively. Beer’s law was also obeyed in all the other ethanol–
water mixtures. The absorption maximum of DPPH• was found
at 524 nm with a molar extinction coefficient, e, of 10700
500 M−1 cm−1 in 1 : 1 ethanol–buffer mixtures at all pH values.
In pure ethanol the maximum was at 515 nm (e = 10600
R
200 M−1 cm−1). Troloxꢀ presents a band at 290 nm (e = 3000
100 M−1 cm−1) in all the 1 : 1 ethanol–buffer mixtures.
The spectral dependence of DPPH-H solutions in 1 : 1 ethanol–
water mixtures between pH 6.1 and 9.6 is illustrated in Fig. 2a.
At the lowest pH, DPPH-H presents a band at 323 nm. When
the pH is increased, a slight decrease and blue shift of this band
is observed and, more prominently, an intense band at 426 nm
Difference spectra obtained by subtracting the first recorded
spectrum from subsequent spectra are shown in Fig. 3b. They
allow better inspection of isosbestic points that appear to be clearly
2418 | Org. Biomol. Chem., 2006, 4, 2417–2423
This journal is
The Royal Society of Chemistry 2006
©

View Online
Fig. 3 Time dependence of DPPH• absorption spectra upon reaction with TrOH in 1 : 1 ethanol–pH 7.4 buffer mixture. (a–b) DPPH•: 7 × 10−5 M,
TrOH: 7.75 × 10−4 M (excess). Times after mixing: 0.006, 0.018, 0.030, 0.042, 0.055, 0.067, 0.079, 0.103, 0.152, 0.201, 0.310, 0.407, 0.492, 0.809, 1.00,
1.50, 2.43 s. (c–d) DPPH•: 7.5 × 10−5 M (excess), TrOH: 7.8 × 10−6 M. Times after mixing, 0.049, 0.146, 0.243, 0.340, 0.438, 0.535, 0.730, 1.02, 1.51,
2.09, 3.06, 4.04, 5.01, 10.07 s. Difference spectra in panels (b) and (d) are derived from absolute spectra by subtracting the initial spectrum to subsequent
spectra. The arrows indicate direction of the absorption changes.
conserved. For 1 : 1 mixtures buffered at pH 7.4 these points are
observed at 345 and 439 nm. Similar findings were obtained for 1 :
1 mixtures buffered at pH 6.4 and 8.4 as well as for pure ethanol
(see the ESI).
The kinetics of the reaction were analyzed in more detail at
specific wavelengths. The decrease of DPPH• absorption at 524 nm
in a 1 : 1 mixture of ethanol and buffer (pH 7.4) is shown in
Fig. 4.
In this experiment, TrOH (9.06 × 10−4 M) was in large excess as
compared to DPPH• (5.33 × 10−5 M). The process appears to be
mainly monophasic and was tentatively fitted by an exponential.
The fit and residuals (curve fit minus experimental data) are
displayed in Fig. 4. Agreement was found to be fairly good,
although not perfect as shown by some fluttering observed on
the residuals. The fluttering was somewhat increased for lower
TrOH : DPPH• ratios. The observed rate constant derived from
monoexponential fit was found to linearly depend on the TrOH
concentration as shown in Fig. 5a. As depicted in Fig. 5b, these
slopes increase with pH. For instance, they were found to be
17200 700, 18200 800 and 31200 3400 M−1 s−1 at pH 6.4,
7.4 and 8.4, respectively.
Similar kinetic traces were obtained in mixtures made with lower
water content. The slopes of the linear fit were significantly lower,
however. For 2 : 1 and 3 : 1 ethanol–water mixtures, slopes were
found to be 5700 100 and 4200 100 M−1 s−1 at pH 6.7, 8600
1300 and 7000 700 M−1 s−1 at pH 7.4, 10200 800 and 7600
300 M−1 s−1 at pH 8.4, respectively. In pure ethanol (data not
shown), the slope of the linear fit was much lower, 320 5 M−1 s−1.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2417–2423 | 2419
©

View Online
DPPH• in excess
Spectra were recorded after mixing solutions of DPPH• (7.5 ×
10−5 after mixing) and TrOH (7.8 × 10−6 M after mixing) in
1 : 1 ethanol–buffer mixtures at various pH. In this case, the
TrOH concentration is the limiting factor and DPPH• is only
partly consumed. Spectral changes obtained with solutions made
with pH 7.4 buffer are displayed in Fig. 3c along with difference
spectra calculated as above (Fig. 3d). Quite well-defined isosbestic
points were observed at 346 and 453 nm. These difference spectra
are similar to those recorded with excess TrOH, except that the
isosbestic points were somewhat shifted. Solutions made with
buffers at pH 6.4 and 8.4 yielded about the same sets of spectra
(see the ESI).
The stoichiometry of the reaction, n, was calculated by using
the relation n = D(Abs)/(De × l × [TrOH]) where D(Abs) is the
absorbance change recorded at the end of the reaction for a given
wavelength, De is the difference between the molar extinction
coefficients of DPPH• and DPPH-H, l is the optical length of
R
the optical cell and [TrOH] the Troloxꢀ concentration. Changes
in the DPPH-H spectrum with pH were taken into account in
the calculation. The stoichiometry was found to be 2.50 0.05,
2.10 0.05 and 1.8 0.07 at pH 6.4, 7.4 and 8.4, respectively.
These values correspond to the mean of four measurements
on solution containing DPPH• and TrOH in ratios between 10
and 20.
Fig. 4 (a) Absorbance decay (k = 524 nm) following mixing of DPPH•
(5.33 × 10−5 M after mixing) and TrOH (9.06 × 10−4 M) in 1 :
1 ethanol–buffer solutions (pH 7.4). (b) Residuals (curve fit minus
experimental data) from monoexponential fit.
Discussion
The reaction of the colored DPPH• radical is a popular method to
evaluate the strength of antioxidants. We have investigated kinetic
aspects of this reaction with a water-soluble analogue of vitamin E,
R
Troloxꢀ, commonly used as a standard. In keeping with current
knowledge from the literature, the reaction is described by the
scheme shown in Fig. 6.
DPPH• being a free radical, the initial step of the reaction with
R
Troloxꢀ leads to the formation of the one-electron oxidized form
of the antioxidant, the phenoxyl-type radical TrO•. This first step
could be reversible, as suggested for phenols.32 The TrO• radical
can react with remaining DPPH• (line (b) of the scheme in Fig. 6) or
disappears through dimerization or disproportionation reactions
involving resonant forms33 [line (c)].
The TrO• radical possesses an absorption band around 435 nm
in water23 very close to the isosbestic points observed in Fig. 3
(see also the ESI for the other pH values). We can estimate that a
variation of 10% of the absorbance at these specific wavelengths
should be visible on the difference spectra. We will take the
spectra shown in Fig. 3a–b as an example. In this experiment,
the maximum concentration of TrO• that could be formed is
equal to the DPPH• concentration (the limiting reagent), i.e.
7.7 × 10−5 M. An isosbestic point is observed at 439 nm with
an absorbance of 0.22. At this wavelength, the molar extinction
Fig. 5 (a) Linear dependence (solid lines) of the observed rate constants
on TroloxꢀR concentration at pH (᭜) 6.4, (ꢀ) 7.0, (+) 7.4, (᭺) 7.7, (×) 8.0,
(ꢀ) 8.4 and pD (ꢁ) 7.4, (ꢂ) 8.8. Dependence of simulated rate constants
(see text) on TroloxꢀR concentration at pH 6.4 (— —), 7.4 (- - -) and 8.4
(– – –). (b) Slopes of the linear plots shown in panel (a) versus pH.
Deuterium isotopic effect
Deuterated hydroalcoholic solutions were prepared by mixing
deuterated ethanol with an equal volume of deuterium oxide
buffered to pD 7.4 or 8.8. Fairly good fits to the monoexponentials
were obtained in both cases and the observed rate constants thus
R
coefficient of the Troloxꢀ radical is 7000 M−1 cm−1 according to
published data.23 Thus, it can be estimated that change in the
concentration of the TrO• radical in the course of the reaction
does not exceed about 3 × 10−6 M, i.e. 5% of the maximum value.
The system can be described by the following set of differential
equations:
R
derived were found to linearly depend on Troloxꢀ concentration,
as shown in Fig. 5a. The bimolecular rate constants were
significantly lower than those for non-deuterated medium.
2420 | Org. Biomol. Chem., 2006, 4, 2417–2423
This journal is
The Royal Society of Chemistry 2006
©

View Online
Fig. 6 Postulated reaction scheme of the reaction between DPPH• and TroloxꢀR . In the pH range investigated, TroloxꢀR is present as its carboxylate form
as shown in the scheme.
d[DPPH•]/dt = −k1f × [DPPH•] × [TrOH] + k1r × [DPPH-H]
× [TrO•] − k2f × [DPPH•] × [TrO•]
constant was kept low enough and had no effect on the simulation.
The result of such simulation is illustrated in Fig. 5a for pH 7.4
solutions. A good agreement was found between simulated and
d[TrOH]/dt = −k1f × [DPPH•] × [TrOH] + k1r × [DPPH-H]
experimental data by using k1f = 1.1 × 104 M−1 s−1 and k2f
=
× [TrO•]
2 × 106 M−1 s−1. The value of the dimerization constant, 2 ×
kd = 4 × 105 M−1 s−1, was also found to be a reasonable estimate.
Similar agreement was found at pH 6.4 and 8.4 with the values
k1f = 1.0 × 104 M−1 s−1, k2f = 2 × 106 M−1 s−1 and k1f = 1.9 ×
104 M−1 s−1, k2f = 2 × 106 M−1 s−1, respectively. It can be noted
that the rate constant k2f is quite large. The stoichiometry of the
reaction measured with DPPH• in excess was found to be close
to the value of 2 in agreement with the sequence of reactions
described in line (a) and (b) of the scheme in Fig. 6.
d[DPPH-H]/dt = k1f × [DPPH•] × [TrOH] − k1r × [DPPH-H]
× [TrO•] + k2f × [DPPH•] × [TrO•]
d[TrO•]/dt = k1f × [DPPH•] × [TrOH] − k1r × [DPPH-H]
× [TrO•] − k2f × [DPPH•] × [TrO•] - 2 × kd × [TrO•]2
It was not possible to obtain an analytical mathematical solution
for this system. A short program based on these differential
equations and on rate constant estimates was thus written for the
As depicted in Fig. 5b, the observed rate constant for the
•
R
reaction of DPPH with excess Troloxꢀ increases with pH. The
•
R
Mathcad software. For each initial Troloxꢀ and DPPH concentra-
tions, the program yielded the TrO• and DPPH• concentrations as
a function of time for various sets of estimates for the 2 × kd, k1f and
k2f rate constants. The bimolecular rate constant, 2 × kd, has been
estimated to be 4 × 105 M−1 s−1 in pH 7.4 buffer.34 This value was
introduced as a constant in the program. The above-mentioned
limit for the change in the TrO• concentration was used as a
constraint to select valid rate constant sets. The curves simulating
the DPPH• concentration were fitted to monoexponentials. The
exponential factor obtained from the simulation was compared to
the experimental rate constant and the set of k1f and k2f constants
adjusted in order to obtain the best agreement over the whole
R
Troloxꢀ carboxylic chain with a pK of 3.6 in water35 is not likely
to be directly involved. In fact, in all cases, Troloxꢀ is expected to
R
be present as its carboxylate form as depicted in Fig. 6. It could be
R
hypothesized that a form of the Troloxꢀ radical protonated on the
phenol (TrOH•+) is first produced through reaction with DPPH•.
The overall reaction would be accelerated upon deprotonation
of this form. The pKa for the TrOH•+/TrO• couple has been
estimated36 to be 2.3. This value appears too low as compared
to the pH range investigated to sustain this hypothesis, that was
also rejected by Litwinienko.37 Alternatively, this effect may arise
from deprotonation of the phenol group, the pK of which has
been estimated15,23 to range between 11.9 and 11.7. As shown by
pulse radiolysis studies, phenolate ions react with various oxidizing
radicals much more quickly than phenols.15 The same effect was
found when singlet oxygen was used as an oxidizing agent.16 Based
on the higher reactivity of the phenolate form, a sequential proton
loss electron transfer (SPLET) mechanism was recently proposed
by Litwinienko11,19,37 and Foti.12 Unfortunately, the poor stability
R
range of the Troloxꢀ concentration. Fluttering of the residuals
similar to those depicted in experimental curves (see Fig. 4b) was
accepted. The reversibility of the reaction was also checked. In
fact, reversibility requires the introduction of a k1r constant with a
high enough value. However, this led to decay curves significantly
departing from the exponential, suggesting that the back reaction
is unimportant in our experimental conditions. Therefore, this
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2417–2423 | 2421
©

View Online
of DPPH• above pH 9.0 did not allowed us to investigate the pH
effect over a large range.
in the development of research on other antioxidants in these
directions.
Our value for the rate constant of the reaction between DPPH•
R
and Troloxꢀ in pure ethanol (320 M−1 s−1) is fairly similar to that
reported for methanol solutions38 (360 M−1 s−1). For comparison,
another tocopherol analog, pentamethylhydroxychroman, yielded
similar values in ethanol (410 M−1 s−1)37 and values ranging
between 350 M−1 s−1 and 1070 M−1 s−1 in methanol.14,37 An
important acceleration of the reaction is observed when ethanol–
water mixtures are used instead of pure ethanol, in favor of
an electron transfer process. Deuterium isotopic effect has been
investigated to shed some light on the mechanism. In fact,
deuterium substitution may have several consequences. First, in
a pure hydrogen abstraction process involving a linear transition
state, it is expected that the rate would be decreased by a factor
equal to the relative vibrational energy of the O–H and O–D bonds,
i.e. 9.8. This value was approached for the reaction of DPPH• with
phenols in pure benzene.39 In aqueous solutions, the ionization
of the phenol group must be taken into account as outlined
above. The acidity of the phenol group is likely to be reduced
in D2O. Typically, weak acids are weaker in D2O than in H2O by
about 0.4–0.5 pK units.40,41 In our experiments, the buffers were
adjusted to pD values of 7.4 and 8.8. Thus, the ionization states of
the phenol group, and eventually of reaction intermediates, would
be similar to those in non-deuterated mixtures at pH 7.0 and
8.4. The slopes of the curves reported in Fig. 5a were compared
accordingly. An isotopic effect of 2.9 0.3 was measured for the
two pH/pD values. Even if lower than that reported for a pure
organic solvent,39 this isotopic effect is significant. The reason of
this effect cannot be found in a difference of the initial ionization
Materials and methods
Materials
The stable free-radical 2,2-diphenyl-1-picrylhydrazyl (DPPH•),
95%
pure
and
6-hydroxy-2,5,7,8-tetramethylchroman-2-
R
carboxylic acid (Troloxꢀ, TrOH), 97% pure were purchased
from Sigma-Aldrich (Saint-Quentin Fallavier, France) and
1,1-diphenyl-2-picrylhydrazine, (DPPH-H), 98% pure from
Fluka Chemika (Buchs, Switzerland). Stock solutions of these
compounds were made in analytical grade ethanol from Merck
(VWR, Fontenay-sous-Bois, France) and stored at 4 ◦C for a
maximum of 2 d in the dark. DPPH• and DPPH-H dissolve slowly
in ethanol. Ethanol-d (99.5% deuteration) was purchased from
Sigma-Aldrich and deuterium oxide (99.9% D) from Eurisotop
(Saclay, France). Ultra-pure water was furnished by a Direct-Q3
apparatus from Millipore (Molsheim, France).
Hydroalcoholic mixtures were prepared by adding ethanol
to the desired amount (v/v) of water that was previously
buffered at the desired pH. Tris(hydroxymethyl)aminomethane
(TRIS) or bis(2-hydroxyethyl)iminotris(hydroxymethyl)methane
(BIS-TRIS) at a concentration of 10 mM in the aqueous compo-
nent were used in the pH range 7.4–9.6 and 6.1–7.2, respectively.
TRIS and BIS-TRIS were purchased from Merck and Sigma-
Aldrich, respectively. For deuterium isotopic experiments, the
heavy water component was buffered by using 10 mM of TRIS or
BIS-TRIS for “pH” 8.4 and 7.0, respectively. These “pH” values
correspond to the reading of the glass electrode standardized in the
usual way. The pD values were calculated according to Salomaa
by adding 0.4 units to these “pH” values.40 In all cases, solutions
were normally equilibrated with air.
R
state of Troloxꢀ, as outlined above. It can be hypothesized that
the rate of deprotonation of the phenol group could be a limiting
factor. This interpretation is in line with the SPLET mechanism.
Alternatively, a former study on the reaction of radicals with
organic reductants by Neta13 suggested that electron and proton
transfer involving the solvent could be concerted in the transition
state. A kinetic isotopic effect of about 2 was found in this study.13
A mechanism involving concerted proton transfer to the incipient
DPPH anion in the transition state is possible and would be
consistent with the measured isotopic effect.
Noticeably, the increase of the slope in Fig. 5 is about twofold
for one pH unit change, whereas a tenfold increase would
be expected for a process entirely controlled by phenol group
deprotonation. At the lowest pH, the slope tends to a constant
value of 1.75 × 104 M−1 s−1. It could be assumed that the
SPLET mechanism11,12,17,19,37 would be favored in the highest pH
range while proton-transfer-mediated electron transfer13 would
dominate at the lowest pH.
In conclusion, our results, taken together with literature
data,12,17 highlight the importance of the control of solvent acidity.
Obviously, this is better achieved by using mixtures of an aqueous
buffer with a miscible organic solvent. These systems would allow
studies of a large panel of antioxidants with various water solubil-
ity properties and should help the characterization of natural and
synthetic antioxidants.42,43 Furthermore, they may provide a better
insight into the complex mechanism of antioxidant reactions.
Also, kinetics should be considered in order to understand how
antioxidants could inhibit the development of the radical reaction
responsible for oxidative stress. The present study is likely to help
Instruments
The pH of the aqueous solutions was measured with a Radiome-
ter PHM240 apparatus (Radiometer Analytical, Villeurbanne,
France) equipped with a glass electrode and calibrated with
standards from Radiometer. UV-visible absorption spectra were
recorded with a UVIKON 923 spectrophotometer (Bio-Tek,
Kontron Instrument, Milano, Italy). An Applied Photophysics
(Leatherhead, UK) stopped-flow apparatus (model SX 18MV)
equipped with a 150 W xenon arc lamp was used for kinetic
measurements and to record time-resolved spectra. The mixing
time was 1.2 ms. Equal volumes of DPPH and Troloxꢀ solutions,
made with the same ethanol–water proportion, were mixed.
Absorption changes at a single wavelength were recorded by using
a conventional photomultiplier over total times ranging from 2.4
to 40 s. Typically, kinetic traces were composed of 2000 points and
at least 4 signals were averaged. Time-resolved absorption spectra
were collected in the range 200–740 nm by using a photodiode
array with a resolution of 2.17 nm. At the maximum scan speed
(2.56 ms between spectra) the first spectrum can be recorded
3.76 ms after the start of mixing. The slits were reduced to
0.8 mm in order to avoid DPPH• photobleaching. Experiments
were carried out at 20 ◦C.
•
R
2422 | Org. Biomol. Chem., 2006, 4, 2417–2423
This journal is
The Royal Society of Chemistry 2006
©

View Online
Data were fitted either by using the software furnished by
Applied Photophysics or Kaleidagraph from Synergy Software
(Reading, PA, USA). The Levenberg–Marquard algorithm was
used for non-linear curve fitting. The Mathcad software (Math-
soft, Inc., Cambridge, MA) was used for numerical simulations.
18 I. Nakanishi, K. Miyazaki, T. Shimada, Y. Iizuka, K. Inami, M.
Mochizuki, S. Urano, H. Okuda, T. Ozawa, S. Fukuzumi, N. Ikota
and K. Fukuhara, Org. Biomol. Chem., 2003, 1, 4085–4088.
19 G. Litwinienko and K. U. Ingold, J. Org. Chem., 2005, 70, 8982–8990.
20 M. Leopoldini, T. Marino, N. Russo and T. M., J. Phys. Chem. A, 2004,
108, 4916–4922.
21 K. Schlesier, M. Harwat, V. Bohm and R. Bitsch, Free Radical Res.,
2002, 36, 177–187.
22 V. W. Bowry and K. U. Ingold, J. Org. Chem., 1995, 60, 5456–5467.
23 M. J. Davies, L. G. Forni and R. L. Willson, Biochem. J., 1988, 255,
513–522.
Acknowledgements
This work was supported by the charity associations “Fondation
Marcel Bleustein-Blanchet pour la Vocation” and “Association
pour la Recherche contre le Cancer” (ARC) through PhD grants to
O.F. The stopped-flow apparatus was acquired thanks to subsidy
from ARC (grant # 7209).
24 M. S. Blois, Nature, 1958, 181, 1199–1200.
25 R. Re, N. Pellegrini, A. Proteggente, A. Pannala, M. Yang and C.
Rice-Evans, Free Radical Biol. Med., 1999, 26, 1231–1237.
26 M. B. Arnao, Trends Food Sci. Technol., 2000, 11, 419–421.
27 W. Brand-Williams, M. E. Cuvelier and C. Berset, Lebensm.-Wiss.
Technol., 1995, 28, 25–30.
28 P. Goupy, C. Dufour, M. Loonis and O. Dangles, J. Agric. Food Chem.,
2003, 51, 615–622.
29 L. Constantin, P. Ionita and T. Constantinescu, Rev. Roum. Chim.,
1994, 39, 1141–1150.
References
30 C. Capellos and B. H. J. Bielski, Kinetic systems, Wiley-Interscience,
1 H. Sies, Eur. J. Biochem., 1993, 215, 213–219.
New York, 1972.
2 T. Finkel, Curr. Opin. Cell Biol., 2003, 15, 247–254.
3 B. Halliwell and J. M. C. Gutteridge, Free radicals in biology and
medicine, 3rd edn, Oxford University Press, Oxford, 1999.
4 G. W. Burton and K. U. Ingold, Acc. Chem. Res., 1986, 19, 194–201.
5 G. W. Burton, T. Doba, E. Gabe, L. Hughes, F. L. Lee, L. Prasad and
K. U. Ingold, J. Am. Chem. Soc., 1985, 107, 7053–7065.
6 J. E. Packer, T. F. Slater and R. L. Willson, Nature, 1979, 278, 737–738.
7 V. W. Bowry and R. Stocker, J. Am. Chem. Soc., 1993, 115, 6029–6044.
8 P. B. Ayscough and K. E. Russell, Can. J. Chem., 1965, 43, 3039–3044.
9 G. W. Burton and K. U. Ingold, J. Am. Chem. Soc., 1981, 103, 6472–
6477.
31 M. Eigen, W. Kruse, G. Maass and L. De Maeyer, in Progress in
reaction kinetics, ed. G. Porter, Pergamon Press, Oxford, 1964, vol. 2,
pp. 287–318.
32 J. S. Hogg, D. H. Lohmann and K. E. Russell, Can. J. Chem., 1961, 39,
1588–1594.
33 A. Kamal-Eldin and L. A. Appelqvist, Lipids, 1996, 31, 671–701.
34 E. Nunez Delicado, A. Sanchez Ferrer and F. Garcia Carmona,
Biochim. Biophys. Acta, 1997, 1335, 127–134.
35 J. Malyszko and M. Mechanik, Polish J. Chem., 2004, 78, 1575–1582.
36 M. J. Thomas and B. H. J. Bielski, J. Am. Chem. Soc., 1989, 111, 3315–
3319.
10 L. Valgimigli, J. T. Banks, K. U. Ingold and J. Lusztyk, J. Am. Chem.
37 M. Musialik and G. Litwinienko, Org. Lett., 2005, 7, 4951–4954.
38 S. Santosh Kumar, K. I. Priyadarsini and K. B. Sainis, Redox Rep.,
2002, 7, 35–40.
Soc., 1995, 117, 9966–9971.
11 G. Litwinienko and K. U. Ingold, J. Org. Chem., 2004, 69, 5888–5896.
12 M. C. Foti, C. Daquino and C. Geraci, J. Org. Chem., 2004, 69, 2309–
2314.
39 R. A. Bird, G. A. Harpell and K. E. Russell, Can. J. Chem., 1962, 40,
701–704.
13 P. Neta, R. E. Huie, P. Maruthamuthu and S. Steenken, J. Phys. Chem.,
40 P. Salomaa, L. L. Schaleger and F. A. Long, J. Am. Chem. Soc., 1964,
86, 1–7.
1989, 93, 7654–7659.
14 I. Nakanishi, T. Kawashima, K. Ohkubo, H. Kanazawa, K. Inami, M.
Mochizuki, K. Fukuhara, H. Okuda, T. Ozawa, S. Itoh, S. Fukuzumi
and N. Ikota, Org. Biomol. Chem., 2005, 3, 626–629.
41 L. Melander and W. H. Saunders, Reaction rates of isotopic molecules,
John Wiley & Sons, New York, 1980.
42 C. A. Rice-Evans, N. J. Miller and G. Paganga, Free Radical Biol. Med.,
1996, 20, 933–956.
43 M. Roche, C. Dufour, N. Mora and O. Dangles, Org. Biomol. Chem.,
2005, 3, 423–430.
15 S. Steenken and P. Neta, J. Phys. Chem., 1982, 86, 3661–3667.
16 R. Scurlock, M. Rougee and R. V. Bensasson, Free Radical Res.
Commun., 1990, 8, 251–258.
17 G. Litwinienko and K. U. Ingold, J. Org. Chem., 2003, 68, 3433–3438.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2417–2423 | 2423
©