The kinetics and thermodynamics of quinone-semiquinone-hydroquinone systems under physiological conditions

Article abstract of DOI:10.1039/a807650b

The steady-state concentration of semiquinones (Q^·-) determined by EPR in the mixtures of eleven alkyl-, methoxy-and chloro-substituted 1,4-benzoquinones as well as 1,4-naphthoquinone (Q) with corresponding hydroquinones (QH₂) in aqueous buffer. pH 7.40, was used to calculate a constant for equilibrium (1) Q + QH₂?Q^·- + Q^·- + 2H⁺ (k₁; 2k_-1; K₁ = k₁/2k_-1). The rale constants for comproportionation between Q and QH₂, k₁, were calculated from the combination of K₁ determined in this work and 2k_-1 reported previously. The Nernst equation was applied to calculate the change in one-electron reduction potential ΔE₁ = E(Q/Q^·-) - E(Q^·-/QH₂) in equilibrium (1). The E(Q^·-/QH₂) values were calculated from ΔE₁ and the values of E(Q/Q^·-) known from the literature. The correlations between E(Q^·-/QH₂) and E(Q/Q^·-) as well as between ΔE₁ (k₁) and E(Q/Q^·-) are discussed. The values of ΔE₁ and k₁ are suggested to be the key factors governing the autoxidation of QH₂.

Full text of DOI:10.1039/a807650b

View Article Online / Journal Homepage / Table of Contents for this issue
The kinetics and thermodynamics of quinone–semiquinone–
hydroquinone systems under physiological conditions
Vitaly A. Roginsky,a Leonid M. Pisarenko,a Wolf Bors*b and Christa Michelb
a N. N. Semenov Institute of Chemical Physics, Russian Academy of Sciences, Kosygin St. 4,
117977 Moscow, Russia
b Institut für Strahlenbiologie, GSF Forschungszentrum für Umwelt und Gesundheit,
D-85764 Neuherberg, Germany
Received (in Cambridge) 1st October 1998, Accepted 16th February 1999
The steady-state concentration of semiquinones (Q Ϫ) determined by EPR in the mixtures of eleven alkyl-, methoxy-
᝽
and chloro-substituted 1,4-benzoquinones as well as 1,4-naphthoquinone (Q) with corresponding hydroquinones
(QH2) in aqueous buffer, pH 7.40, was used to calculate a constant for equilibrium (1) Q ϩ QH2
Q
Ϫ ϩ Q Ϫ ϩ
᝽
᝽
2Hϩ (k1; 2kϪ1; K1 = k1/2kϪ1). The rate constants for comproportionation between Q and QH2, k1, were calculated
from the combination of K1 determined in this work and 2kϪ1 reported previously. The Nernst equation was applied
to calculate the change in one-electron reduction potential ∆E1 = E(Q/Q Ϫ) Ϫ E(Q Ϫ/QH2) in equilibrium (1). The
᝽
᝽
E(Q Ϫ/QH2) values were calculated from ∆E1 and the values of E(Q/Q Ϫ) known from the literature. The correlations
᝽
᝽
between E(Q Ϫ/QH2) and E(Q/Q Ϫ) as well as between ∆E1 (k1) and E(Q/Q Ϫ) are discussed. The values of ∆E1 and
᝽
᝽
᝽
k1 are suggested to be the key factors governing the autoxidation of QH2.
benzoquinone/1,4-hydroquinone couple.15 Using the Nernst
equation, K1 may be converted into the difference in one-
electron reduction potential in equilibrium (1), ∆E1, that
represents the combination of E(Q/Q Ϫ) and E(Q Ϫ/QH2).
Introduction
The reactivity and thermodynamic properties of quinones
(Q) and their reduced forms, semiquinones (Q Ϫ) and hydro-
᝽
᝽
᝽
quinones (QH2), are related to many biological problems
including quinone cytotoxicity,1,2 application of quinones as
antitumor agents,2,3 electron transfer,4 and the functioning of
Eqn. (2) may be used to calculate E(Q Ϫ/QH2) from ∆E1
᝽
∆E1 = E(Q/Q Ϫ) Ϫ E(Q Ϫ/QH2)
(2)
the antioxidant defense system.5 There are several equilibria
᝽
᝽
Ϫ
involving Q, Q and QH2 in chemical and biological systems.
᝽
provided that E(Q/Q Ϫ) is known. While considerable attention
᝽
The equilibrium (1) and its constituents, disproportionation
was paid to the determination of E(Q/Q Ϫ), the values of
Ϫ
᝽
of Q (reaction (Ϫ1)) and comproportionation between Q
᝽
E(Q Ϫ/QH2) in aqueous solutions have been reported only for
᝽
a few QH2.16 Meanwhile, E(Q Ϫ/QH2) determines to a large
(1)
᝽
Q ϩ QH2
Q
Ϫ ϩ Q Ϫ ϩ 2Hϩ
extent the reactivity of Q Ϫ and QH2, and thus this parameter is
(1),(Ϫ1)
᝽
᝽
(Ϫ1)
᝽
of vital interest for Q/QH2 chemistry and biochemistry.
The present work is devoted to the EPR determination of
and QH2 (reaction(1)), are the most fundamental. Knowledge
of this equilibrium constant, K1 = k1/2kϪ1, along with the rate
constants for elementary reactions (Ϫ1) and (1), 2kϪ1 and k1,
opens up many opportunities to predict the reactivity of Q,
K1 from a steady-state concentration of Q Ϫ in the mixtures of
᝽
Q and QH2 for eleven Q/QH2 couples presented in Scheme 1.
These data were used to calculate k1, ∆E1, and E(Q Ϫ/QH2) and
᝽
Q
Ϫ, and QH2 and the behavior of these species in various
to establish the correlation between various one-electron
reduction potentials.
᝽
chemical and biological systems.
The value of 2kϪ1 determines to a significant degree the
Ϫ
᝽
stability of Q
and its steady-state concentration. Other
Experimental
things being equal, the lower 2kϪ1 the more significant becomes
the role of other reactions with participation of Q Ϫ. Much
The quinones and hydroquinones studied in this work are
presented in Scheme 1. Q 1, Q 5, Q 6, Q 7, Q 10, and Q 11 were
purchased from Aldrich; Q 2 and QH2 2 from Merck; QH2
4 and QH2 11 from Fluka, Q 8 from Lancaster, Q 9 from Sigma.
Q 3 and Q 4 were prepared via the oxidation of QH2 3 and QH2
4 with K3Fe(CN)6 in the 1:1 mixture of benzene and diethyl
ether. QH2 1, QH2 5, QH2 6, QH2 7, QH2 8, QH2 9 and QH2 10
were prepared by the reduction of corresponding Q by Zn
powder in acetic acid followed by removing the solvent with
a rotary evaporator and further extraction of QH2 with an
appropriate organic solvent. Both purchased and synthesized Q
and QH2 were purified by recrystallization, sublimation under
vacuum or using a silica gel (40–100 µm) column with CHCl3 as
an eluent. Sodium phosphates, NaH2PO4 and Na2HPO4, of
highest grade used to prepare buffer solutions, were purchased
from Merck. Other reagents were of the highest available grade.
᝽
attention has been given to the determination of 2kϪ1, basically
using pulse radiolysis combined with UV-Vis spectro-
photometry (refs. 6–9 and references therein). Surprisingly,
the quantitative information on the disproportionation of
substituted 1,4-benzosemiquinones was until recently very
restricted though the kinetics of this process with Q Ϫ produced
᝽
from substituted naphthoquinones and anthraquinones and
Q
with more complex structures have been studied in
detail. Our recent work9 has partly eliminated this gap. K1 was
previously reported for many Q/QH2 couples but only a few of
them were determined at physiological pH.10–13 When K1 and
2kϪ1 are known, this allows us to calculate the rate constant
for reaction (1), a parameter which significantly governs the
oxidizability of QH2 by molecular oxygen.14 Previously a k1
value has been reported only for the non-substituted 1,4-
J. Chem. Soc., Perkin Trans. 2, 1999, 871–876
871

View Article Online
O
O•
OH
R4
R3
R1
R2
R4
R3
R1
R2
R4
R3
R1
R2
O
O–
OH
Q
Q•–
QH2
R1
R2
R3
R4
1
2
3
H
Me
Et
H
H
H
H
H
H
H
H
H
4
5
6
7
8
9
10
But
Me
Me
Cl
OMe
Me
Cl
H
H
H
H
H
OMe
Cl
H
H
Pri
Cl
H
H
Me
H
H
OMe
H
Fig. 1 Time dependence of [Q Ϫ] in 50 mM phosphate buffer,
᝽
pH 7.40, at 37 ЊC for the mixtures of 0.5 mM Q 9 and 1 mM QH2 9 (plot
1); 25 µM Q 10 and 60 µM QH2 10 (plot 2); 62 µM Q 7 and 60 µM QH2
7 (plot 3).
OMe
Cl
Cl
O
O
O•
OH
O–
Q•– 11
OH
QH2 11
Q 11
Scheme 1 The structures of quinones, hydroquinones and semi-
quinones studied.
Aqueous solutions were prepared with doubly distilled water.
Experiments were performed at 37 ЊC with 50 mM phosphate
buffer, pH 7.40 0.02, (unless otherwise indicated), which was
prepared by mixing fifty millimolar solutions of NaH2PO4 and
Na2HPO4 without adding any acid or base. Solutions of the
individual phosphates used for the buffer preparation were
purged from traces of transition metals by Chelex-100 resin
(Bio-Rad) using a batch method.17 Stock solutions of Q and
QH2 were prepared, depending on solubility, with water or
Fig. 2 Plots of K1 against DMSO concentration for equilibrium (1)
determined in the mixtures of Q 2 with QH2 2 (plot 1); Q 9 with QH2 9
(plot 2); Q 4 with QH2 4 (plot 3); in 50 mM phosphate buffer, pH 7.40,
at 37 ЊC.
aqueous dimethyl sulfoxide (DMSO).
Ϫ
Steady-state concentration of Q in the mixture of Q and
᝽
QH2 used to calculate K1 was determined by EPR in a flat
quartz cell with a Varian E12 spectrometer (Varian, USA)
equipped with a TE104 dual cavity and temperature controller.
Solutions containing Q and corresponding QH2 were prepared
by adding a certain volume of stock solutions of Q and QH2.
Both stock solutions and buffer were argon-bubbled prior to
mixing. The reaction mixture was immediately transferred
using a microsyringe into a flat EPR cell flushed with argon.
10 µM solution of the aminoxyl stable radical TEMPO in
benzene placed into one of the cavities was used as a reference
standard for the determination of the absolute concentration.
Instrument settings were as follows: microwave power, 5 mW;
modulation frequency, 12.5 kHz; modulation amplitude, 0.63 G
one hour as is exemplified by plot 1 in Fig. 1. This demonstrates
Ϫ
that Q and QH2 are the only products of Q dispropor-
᝽
tionation and thus this reaction is completely reversible. By
Ϫ
contrast, the concentration of Q formed in the Q 7/QH2
᝽
7 and Q 10/QH2 10 systems decreased dramatically with time
(plots 2 and 3, Fig. 1) suggesting that reaction (Ϫ1) in these
cases is not the only pathway of Q Ϫ decay.
᝽
A constant of equilibrium (1), K1, was calculated from
[Q Ϫ] by using eqn. (3), where [Q]0 and [QH2]0 are initial
᝽
K1 = [Q Ϫ]2/([Q]0 Ϫ 0.5[Q Ϫ]) ([QH2]0 Ϫ 0.5[Q Ϫ]) (3)
᝽
᝽
᝽
(for determination of [Q Ϫ]) or 0.05 G (for determination of
concentrations of the reagents. Typically, K1 was determined
in four or more separate runs at several concentrations of
[Q] and [QH2]. The K1 value was found to be independent of
᝽
hyperfine splitting parameters). The absolute concentration
Ϫ
of Q was calculated by double integrating EPR spectrum
᝽
Ϫ
[Q] or [QH2]. With the Q 7/QH2 7 couple, the concentration of
of Q and normalizing the obtained value to the intensity of
᝽
Ϫ
Q
extrapolated to zero time was used to calculate K1. With
the standard. The protocol we followed for EPR determinations
᝽
the Q 10/QH2 10 mixture, the starting concentration of Q Ϫ was
close to the sum of [Q] and [QH2]; an exact value of K1 could
not therefore be calculated.
has been reported in more detail elsewhere.13,18 A standard error
᝽
in the determination of [Q Ϫ] was typically within 15%.
᝽
In some cases K1 was determined in aqueous buffer contain-
ing a small amount of DMSO that was added to increase the
solubility of Q. As is exemplified by Fig. 2, K1 increased nearly
linearly with [DMSO]. The K1 values presented in Table 1 were
determined either in solution without DMSO or by using linear
extrapolation of the measured K1 values to zero concentration
of DMSO as shown in Fig. 2. For several Q/QH2 couples these
values may be compared with those reported in ref. 10 (Q 1/QH2
1, Q 2/QH2 2 and Q 11/QH2 11) and ref. 11 (Q 3/QH2 3).
The reported values differ from those determined in this
Results and discussion
EPR determination of K1
When Q and QH2 were mixed in deaerated buffer, well-resolved
multicomponent EPR spectra attributed to Q Ϫ were observed.
᝽
Hyperfine splitting parameters of these spectra were in reason-
able agreement with those reported in the literature19,20 and are
therefore not reported here. With most Q/QH2 couples the
intensity of the EPR spectrum remained constant for at least
872
J. Chem. Soc., Perkin Trans. 2, 1999, 871–876

View Article Online
Ϫ
Table 1 Parameters of equilibrium (1) Q ϩ QH2
mM sodium phosphate buffer
Q
Ϫ ϩ Q ϩ 2Hϩ (K1 = k1/2kϪ1) determined by EPR (K1) and pulse radiolysis (2kϪ1) in 50
᝽
᝽
a
Q/QH2
K1 b at 37 ЊC, pH 7.4
∆H1/kJ molϪ1 c
K1 at 37 ЊC, pH 7.4
2kϪ1/108 MϪ1 sϪ1 f
k1/MϪ1 sϪ1
1
2
(8.1 1.4) × 10Ϫ6
(3.3 0.6) × 10Ϫ6
(3.1 0.7) × 10Ϫ6
(8.5 2.3) × 10Ϫ8
(4.4 0.9) × 10Ϫ7
(7.9 2.2) × 10Ϫ7
(5.5 0.7) × 10Ϫ2
(2.6 0.4) × 10Ϫ5
(2.6 0.5) × 10Ϫ6
>1
50 4 (39)
50 5 (49)
2.4 × 10Ϫ6 d
1.1 × 10Ϫ6 d
1.6 0.2
1.35 0.02
0.91 0.04
0.35 0.17
1.15 0.20
0.38 0.08
nd
0.32 0.03
0.54 0.03
nd
1300 400
450 90
290 80
~3
50 20
30 14
nd
800 200
140 35
nd
3
54
50
nd
nd
64
46
54
nd
6
5
4
2.2 × 10Ϫ8 e
5
6
7
8
4
5
8
9
10
11
(5.2 1.4) × 10Ϫ6
nd (57)
1.0 × 10Ϫ5 d
2.76 0.10
1400 400
nd — Not determined. a The structures of Q/QH2 are given in Scheme 1. b Values of K1 mean SD from four or more independent experiments
conducted at various concentrations of Q and QH2. c ∆H1 in parentheses were reported in ref. 10. d Reported in ref. 10 at 25 ЊC and recalculated to
37 ЊC using ∆H1 determined there. e K1 reported in ref. 11 at 22 ЊC and pH 7.0 and recalculated to 37 ЊC and pH 7.4 using ∆H1 = 50 kJ molϪ1 and
d(log K1)/d(pH) = 2. f The averaged values determined by pulse radiolysis of Q and QH2 at room temperature in our previous work9 (see text for more
detail).
Fig. 4 Plots of K1 against pH determined in phosphate buffer at 37 ЊC
Fig. 3 Van’t Hoff plots of K1 determined in phosphate buffer, pH
for the following couples: Q 1/QH2 1 (᭹), Q 2/QH2 1 (᭝), Q 9/QH2
7.40, for the following couples: Q 4/QH2 4 (plot 1); Q 9/QH2 9 (plot 2);
9 (᭺).
Q 3/QH2 3 (plot 3); Q 1/QH2 1 (plot 4); Q 8/QH2 8 (plot 5).
radiolysis experiments with Q and QH2 solutions were found to
be somewhat different. For this reason and because of the fact
that both Q and QH2 are present in the system, 2kϪ1 values were
averaged for calculations of k1. Although the values of 2kϪ1
study typically by a factor of 2–4; this is not too significant
a difference, as it corresponds to the difference in absolute con-
centration of Q Ϫ of about 1.5–2 times.
᝽
The temperature effect was studied for several Q/QH2
Ϫ
used in these calculations were determined at ca. 22 ЊC rather
couples. A steady-state concentration of Q
and thus K1
᝽
than at 37 ЊC, it is unlikely that the difference in 2kϪ1 between
22 ЊC and 37 ЊC is significant. Previously k1 has been reported
only for the Q 1/QH2 1 couple (58 MϪ1 sϪ1 at 25 ЊC and pH
7.0).15 To compare this value of k1 with that determined in
the present work, it has to be recalculated for our conditions.
When passing from pH 7.0 to pH 7.4 (with d(log k)/d(pH) = 2,
see Fig. 4), k1 will increase 6.3 times; when passing from
25 ЊC to 37 ЊC, k1 (with ∆H = 50 kJ molϪ1 (Table 1)) will increase
2.2 times. Hence, the value of k1 reported in ref. 15, being
recalculated for our conditions, is expected to equal 58 ×
6.3 × 2.2 ≈ 800 MϪ1 sϪ1. The latter value is in reasonable agree-
ment with 1300 400 MϪ1 sϪ1 determined in the present study
(Table 1). As seen from Table 1, k1 in the series of methyl-
substituted 1,4-benzoquinones/hydroquinones decreases dram-
atically with the number of methyl groups, i.e. in the direction
increased with increasing temperature (Fig. 3). The determined
enthalpies of equilibrium, ∆H1, varied within a rather narrow
range from 46 to 64 kJ molϪ1 (Table 1). With Q Ϫ 1 and Q Ϫ 2,
᝽
᝽
it was possible to compare the ∆H1 values determined in this
study with those reported in ref. 10; they were in excellent
agreement with each other (Table 1). K1 was found to rise with
pH evidently due to the larger contribution of an ionized form
of QH2, QHϪ, to equilibrium (1). The linear plots of log K1
against pH with slopes of 2.00 0.04 (Q Ϫ 1); 1.93 0.06 (Q
Ϫ
᝽
᝽
Ϫ
2); 1.91 0.05 (Q 9) were observed (Fig. 4). The slope of
᝽
nearly 2 is in agreement with previous works (refs. 11, 12, 15)
and predicted by the theory for the case when pH is far from
the pK16 of QH2.
Determination of k1
of decreasing E(Q/Q Ϫ); k1 also decreases with the volume of
᝽
The rate constants for reaction (1) between Q and QH2 were
calculated from the combination of the values of K1 determined
in this study and 2kϪ1 previously reported, largely in ref. 9.
alkyl substituent (cf. Q 2 with Q 3 and Q 4). However, k1
increases when methyl groups are replaced by methoxy groups
(cf. Q 5 with Q 8) despite the decrease in E(Q/Q Ϫ).
᝽
Calculation of ÄE1 and mid-point potential E(Q Ϫ/QH2)
k1 = K1(2kϪ1
)
(4)
᝽
The change of the one-electron reduction potential in equi-
librium (1), ∆E1 (in mV), was calculated from K1 using the
The values of k1 calculated in this way are given in Table 1.
With several Q Ϫ, the 2kϪ1 values measured in ref. 9 via pulse
᝽
J. Chem. Soc., Perkin Trans. 2, 1999, 871–876
873

View Article Online
Table 2 One-electron reduction mid-point potentials (in mV) in the
system Q–Q Ϫ–QH2 at 25 ЊC and pH 7.0
᝽
Ϫ
c
Ϫ
E(Q Ϫ/QH2)
E(Q /QH2)c
a
b
Q/QH2
∆E1
E(Q/Q
)
᝽
᝽
᝽
1
2
Ϫ370
Ϫ391
Ϫ395
Ϫ485
Ϫ443
Ϫ428
Ϫ147
Ϫ337
Ϫ399
>0
ϩ78
ϩ23
ϩ448
ϩ459
ϩ460
ϩ414
ϩ395
ϩ453
ϩ363
ϩ358
ϩ617
ϩ187
~ϩ290
>ϩ650
ϩ240
3
0d
4
Ϫ3211
Ϫ80
ϩ489f
ϩ430
5
6
Ϫ70d
ϩ470e
Ϫ15011
Ϫ110g
ϩ650e
Ϫ140
7
6231
8
9
10
11
ϩ7261
ϩ2122
Ϫ380
a The structures of Q/QH2 are given in Scheme 1. b Recalculated from
data given in Table 1 using an experimental value of ∆H1 if available (or
∆H1 = 50 kJ molϪ1 when not available) and assuming that d(log K1)/
d(pH) = 2. c Taken from ref. 16, unless otherwise indicated. d Estimated
Fig. 5 The correlation between mid-point potential E(Q/Q Ϫ) in
based on the correlation of E(Q/Q Ϫ) with the structures of alkyl-
᝽
᝽
aqueous buffer, pH 7.0, (SHE as a reference electrode) and E(Q/Q Ϫ) in
substituted 1,4-benzoquinones reported in refs. 16, 21. e Estimated from
᝽
Ϫ
the correlation of E(Q/Q ) in aqueous buffer and that in MeCN21 (see
acetonitrile (SCE as a reference electrode) for 1,4-benzoquinones (᭹),
1,4-naphthoquinones (ᮀ), 9,10-anthraquinones (᭺) and miscellaneous
compounds (᭝). Data were taken from ref. 14 and 18, respectively.
᝽
below). f Calculated on the basis of data reported in ref. 11. g Estimated
from the correlation of E(Q/Q Ϫ) in aqueous buffer and that in
᝽
Numbers at symbols represent Q/Q Ϫ/QH2 structures as they are given
MeCN.22
᝽
in Schemes 1 and 2.
ln K1 = 0.0389∆E1
(5)
demonstrates, this correlation is also workable for a larger
assortment of Q including tert-butyl- and methoxy-substituted
benzoquinones, several 1,4-naphthoquinones (NQ), and 9,10-
anthraquinones (AQ) (see Scheme 2). However, hydroxy-
substituted NQ and AQ visibly do not fit this correlation
(Fig. 5). Without regard for hydroxy-substituted NQ and
Nernst equation. Eqn. (5) shows ∆E1 at the standard temper-
ature, 25 ЊC. The values of ∆E1 for the standard conditions
(25 ЊC, pH 7.0) are given in Table 2. As mentioned above, ∆E1 is
Ϫ
the difference between two one-electron potentials, E(Q/Q
)
᝽
and E(Q Ϫ/QH2)† (eqn. (2)). If E(Q/Q Ϫ) is known, eqn. (2)
᝽
᝽
allows us to calculate the second potential E(Q Ϫ/QH2) from
AQ, the correlation between E(Q/Q Ϫ) in aqueous buffer,
᝽
᝽
∆E1. As a rule, the values of E(Q/Q Ϫ) applied to calculate
pH 7.0 (standard hydrogen electrode, SHE, as a reference
᝽
E(Q Ϫ/QH2) were taken from ref. 16. The values of E(Q Ϫ/QH2)
Ϫ
electrode), E(Q/Q )aq, and that in acetonitrile (saturated
᝽
᝽
᝽
calculated from ∆E1 by eqn. (2) are listed in Table 2. While
calomel electrode, SCE, as a reference electrode), E(Q/
E(Q Ϫ/QH2) for Q Ϫ 1, Q Ϫ 7, and Q Ϫ 11 reported in refs. 16,
Ϫ
Q
᝽
)MeCN, is described by the eqn (6). Reduction potentials
᝽
᝽
᝽
᝽
23 and those determined in our work were in reasonable agree-
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
)
ment, the difference for Q 2, Q 4 and Q 5 was rather
᝽
᝽
᝽
E(Q/Q
)
aq = 650 ϩ 1.1 E(Q/Q
(6)
MeCN
᝽
᝽
significant (Table 2). It should be noticed that the E(Q Ϫ/QH2)
᝽
values reported in ref. 23 were calculated using a sophisticated
are given in mV. Such a correlation may be very useful in
Ϫ
protocol rather than directly. With Q Ϫ 3, Q Ϫ 6, Q Ϫ 8 and Q
Ϫ
Ϫ
᝽
᝽
᝽
᝽
estimating E(Q/Q
)
when E(Q/Q
)
is known. Nearly
aq
MeCN
᝽
᝽
9, the E(Q Ϫ/QH2) was determined in our work for the first
the same correlation may be suggested with E(Q/Q Ϫ) deter-
᝽
᝽
time.
mined in other organic solvents.
Chambers21 reported a linear correlation between E(Q/Q
Ϫ
)
᝽
The correlation between various one-electron reduction potentials
determined in acetonitrile and the sum of the Hammett
substituent constants, Σσ, for substituted 1,4-benzoquinones
By contrast to aprotic organic solvents, direct determination of
and the related correlation for E(Q Ϫ/QH2) for substituted
one-electron reduction potentials, E(Q/Q Ϫ) and E(Q Ϫ/QH2),
᝽
᝽
᝽
1,4-hydroquinones. A parallel existence of these two linear
in an aqueous medium using a routine electrochemical tech-
nique (polarography or potentiometry) is almost impossible
Ϫ
correlations suggests a linear correlation between E(Q/Q
)
᝽
and E(Q Ϫ/QH2). The latter is given in Fig. 6. With a few
because of the instability of Q Ϫ. Under these circumstances,
᝽
᝽
exceptions, the values of E(Q Ϫ/QH2) and E(Q/Q Ϫ) demon-
the determination of E(Q/Q Ϫ) and E(Q Ϫ/QH2) in aqueous
᝽
᝽
᝽
᝽
strate the excellent correlation for various kinds of Q and QH2
that is described by eqn. (7).
solution requires much more complicated non-direct methods,
mostly pulse radiolysis and the combination of pulse radiolysis
and EPR technique using reference compounds with known
reduction potentials. This is probably the reason why the in-
formation on E(Q/Q Ϫ) and especially E(Q Ϫ/QH2) in aqueous
Ϫ
E(Q Ϫ/QH2) = Ϫ680 ϩ 0.81 E(Q/Q
)
(7)
᝽
᝽
᝽
᝽
solution is much more limited as compared to organic solvents.
Thus the approach using various correlations for prediction of
unknown one-electron reduction potentials in aqueous solution
looks very promising. Wardman24 has drawn attention to an
From the standpoint of quantum chemistry, the occurrence
Ϫ
of this correlation means that, when Q transforms into Q
᝽
and Q Ϫ transforms into QH2, an additional electron falls into
the same lowest uncoupled molecular orbital (refs. 21, 22 and
references therein). Fig. 7 depicts the same correlation for the
᝽
excellent correlation between E(Q/Q Ϫ) determined for methyl-
᝽
substituted 1,4-benzoquinones in aqueous buffer and those in
case of aqueous solution. Although the general tendency
remains the same—E(Q Ϫ/QH2) decreases when E(Q/Q
)
aprotic organic solvents and the application of the correlation
Ϫ
᝽
᝽
as a promising way to predict E(Q/Q Ϫ) in water. As Fig. 5
᝽
decreases—the quality of the correlation is considerably
worse, besides, it becomes non-linear. This is not a surprise
since E(Q Ϫ/QH2) depends on a prototropic equilibrium
Ϫ
† In principle, the form E(Q , 2H+/QH2) should be used instead of the
᝽
᝽
short form E(Q Ϫ/QH2). For simplicity, we use the short form ignoring
protonation in the text.
(characterized by pK) which varies significantly from one
QH2 to another; the latter results in a different contribution
᝽
874
J. Chem. Soc., Perkin Trans. 2, 1999, 871–876

View Article Online
O
O•
OH
R4
R3
R1
R2
R4
R3
R1
R2
R4
R3
R1
R2
O
O–
OH
Q
Q•–
QH2
R1
R2
R3
R4
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Me
Me
Me
Me
Ph
Cl
NMe2
COMe
OCOMe
CF3
NO2
CONH2
Cl
H
Me
H
Me
Me
H
H
H
H
H
H
H
H
Cl
Cl
Cl
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
CN
Me
Me
Me
H
H
H
H
H
H
H
Fig. 6 The correlation between E(Q Ϫ/QH2) and E(Q/Q Ϫ) deter-
᝽
᝽
mined in acetonitrile (SCE as a reference electrode) for 1,4-benzo-
quinones (᭹), 1,4-naphthoquinones (ᮀ), 9,10-anthraquinones (᭺) and
miscellaneous compounds (᭝). Data were taken from ref. 18. Numbers
H
H
Cl
CN
Cl
Cl
at symbols represent Q/Q Ϫ/QH2 structures as they are given in
᝽
Schemes 1 and 2.
R4
O
O
R4
O•
R4
OH
OH
R1
R2
R1
R2
R1
R2
R3
R3
O–
R3
R3
QH2
Q
Q•–
R1
R2
R4
27
Me
OH
H
Me
NH2
Cl
H
H
H
Me
H
H
H
OH
H
H
H
H
H
H
H
H
H
28
29
30
31
32
33
Cl
OMe
H
H
OMe
R6
O
R1
R2
R5
O•
R4
R1
O
R3
R6
Fig. 7 The correlation between mid-potential E(Q Ϫ/QH2) and E(Q/
᝽
Q
Ϫ) determined in aqueous buffer, pH 7.0, (SHE as a reference
Q
᝽
R6
OH
OH
R1
electrode) for benzoquinones (᭹), 1,4-naphthoquinones (ᮀ), 9,10-
anthraquinones (᭺) and miscellaneous compounds (᭝). Data were
taken largely from ref. 14 and partly from Table 2 of the present work.
R2
R2
Numbers at symbols represent Q/Q Ϫ/QH2 structures as they are given
᝽
R5
R5
in Schemes 1 and 2.
R4
R1
O–
Q•–
R3
R4
R3
QH2
Redox potentials and the kinetics of QH2 autoxidation
R2
H
R3
R4
R5
R6
Traditionally, the autoxidation of QH2 is considered to be
triggered by the direct interaction of QH2 with molecular
oxygen (eqn. (8)). This is a reason why repeated attempts have
34
35
36
37
38
39
40
41
42
H
H
H
H
H
H
Me
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
OH
SO3Na
QH2 ϩ O2 → Q Ϫ ϩ O2 Ϫ ϩ 2Hϩ
(8)
Me
Me
SO3Na
H
H
H
H
᝽
᝽
Me
SO3Na
H
H
H
been made to correlate the oxidizability of QH2 with the
SO3Na
SO3Na OH
OH
H
SO3Na
H
SO3Na
one-electron potential E(Q Ϫ/QH2)1,2 and the two-electron
OH
OH
᝽
reduction potential E(Q/QH2).25 These attempts had only
moderate success and many QH2 dropped out of the corre-
lation. Furthermore, reaction (8) is spin-restricted26 and is thus
expected to be extremely slow under physiological conditions.
In addition to this theoretical argument against reaction (8)
as a triggering step of QH2 autoxidation, experimental counter
arguments can be found in the literature. For many types of
QH2, e.g. 1,4-hydroquinone,14 1,4-naphthoquinones,27 and cate-
cholamines,28 QH2 autoxidation was reported to be a self-
accelerated autocatalytic process, with Q being a catalyst. It
was shown that the initial step of the oxidation of many QH2
H
Scheme 2 The structures of quinones, hydroquinones and semiqui-
nones taken into the correlations between various reduction potentials
(see Figs. 4–7).
of solvation energy to the reduction potential E(Q Ϫ/QH2).
᝽
Nevertheless, the correlation presented in Fig. 7 may be useful
for a rough estimation of unknown values of E(Q Ϫ/QH2) when
᝽
E(Q/Q Ϫ) is available.
᝽
J. Chem. Soc., Perkin Trans. 2, 1999, 871–876
875

View Article Online
very low. The latter may be roughly estimated. One-electron
O
O
O
O
O
reduction potentials E(DAsc/Asc Ϫ) and E(Asc Ϫ/AscHϪ) were
᝽
᝽
O
O
reported to be Ϫ174 mV31 and ϩ282 mV16, respectively; the
rate constant for Asc Ϫ disproportionation at pH 7.0 is as much
᝽
as 3 × 106 MϪ1 sϪ1 32
.
It is possible to calculate from these data
E(DAsc /Asc Ϫ) Ϫ E(Asc Ϫ/AscHϪ) = Ϫ456 mV and k1 = 0.7
O
᝽
᝽
MϪ1 Ϫ1. Such a low value of k1 could explain the main
s
43
44
45
46
features of AscHϪ autoxidation.
O
O
OH
O
In conclusion, the above observations strongly suggest that
the rate of reaction (1) and the value of ∆E1 are the key
factors controlling QH2 oxidizability.
C
CH2OH
HO
CH(OH)CH2OH
OH
O
O
HO
OMe
OH
Me
Acknowledgements
O
O
HO
The work was supported by the Volkswagen Stiftung (grant
1/71149), the Deutsche Forschungsgemeinschaft (grant 436Rus
113/245/0) and the Russian Foundation for Fundamental
Researches (grant 96-03-00103).
NH2
47
48
Scheme 3 The structures of quinones (43–47) and ascorbic acid (48)
taken into the correlations between various reduction potentials (see
Figs. 4–7).
References
was accelerated by adding Q.14,28 These observations suggest
reaction (1) between Q and QH2 resulting in the formation of
1 P. J. O’Brien, Chem. Biol. Interact., 1991, 80, 1.
2 A. Brunmark and E. Cadenas, Free Radical Biol. Med., 1989, 7, 435.
3 G. Powis, Free Radical Biol. Med., 1989, 6, 63.
4 P. Rich, Biochim. Biophys. Acta, 1981, 637, 28.
5 E. Cadenas, P. Hochstein and L. Ernster, Adv. Enzymol. Relat.
Areas Mol. Biol., 1992, 62, 97.
Q
Ϫ to be the main trigger reaction of QH2 autoxidation. If it is
᝽
the case, the efficiency of this process may be characterized by
either K1 , i.e. the difference ∆E1 = E(Q/Q Ϫ) Ϫ E(Q Ϫ/QH2), or,
᝽
᝽
to be more precise, by k1.
6 K. E. O’Shea and M. A. Fox, J. Am. Chem. Soc., 1991, 113, 611.
7 I. Wilson, P. Wardman, T.-S. Lin and A. C. Sartorelli, J. Med.
Chem., 1996, 29, 1381.
8 M. C. Rath, H. Pal and T. Mukherjee, J. Chem. Soc., Faraday Trans.,
1996, 92, 1891.
9 V. A. Roginsky, L. M. Pisarenko, C. Michel, M. Saran and W. Bors,
J. Chem. Soc., Faraday Trans., 1998, 94, 1835.
10 A. E. Alegria, M. López and N. Guevera, J. Chem. Soc., Faraday
Trans., 1996, 92, 4965.
11 J. K. Dohrmann and B. Bergmann, J. Phys. Chem., 1995, 99, 1218.
12 E. J. Land, T. Mukherjee and A. J. Swallow, J. Chem. Soc., Faraday
Trans., 1983, 79, 405.
13 V. A. Roginsky, G. Bruchelt and H. B. Stegmann, Biochemistry
(Moscow), 1998, 63, 240.
14 P. Eyer, Chem. Biol. Interact., 1991, 80, 159.
15 I. Yamazaki and T. Ohnishi, Biochim. Biophys. Acta, 1966, 112, 469.
16 P. Wardman, J. Phys. Chem. Ref. Data Ser., 1989, 18, 1637.
17 G. R. Buettner, J. Biochem. Biophys. Methods, 1988, 16, 27.
18 V. A. Roginsky and H. B. Stegmann, Free Radical Biol. Med., 1994,
17, 93.
19 K. B. Ulmschneider and H. B. Stegmann, in Landolt-Börnstein New
Series, ed. H. Fischer and K. H. Hellwege, Springer, Berlin, 1980,
9d1, 93.
20 D. Klotz, T. Jülich, G. Wax and H. B. Stegmann, in Landolt-
Börnstein New Series, ed. H. Fischer, Springer, Berlin, 1989, 17g, 69.
21 J. Q. Chambers, in The Chemistry of the Quinoid Compounds, Part 2,
ed. S. Patai, 1974, Wiley, London, p. 738.
22 M. Knüpling, J. T. Törring and S. Un, Chem. Phys., 1997, 219, 291.
23 K. Sugioka, M. Nakano, H. Totsune-Nakano, H. Minakami,
S. Tero-Kumota and Y. Ikegami, Biochim. Biophys. Acta, 1988, 936,
377.
24 P. Wardman, Free Radical Res. Commun., 1990, 8, 219.
25 Y. A. Ilan, G. Czapski and D. Meisel, Biochim. Biophys. Acta, 1976,
430, 209.
26 D. G. Graham, S. M. Tiffany, W. R. Bell and W. F. Gutknecht,
Mol. Pharmacol., 1978, 14, 644.
27 D. M. Miller, G. R. Buettner and S. D. Aust, Free Radical Biol.
Med., 1990, 8, 95.
28 T. Ishi and I. Fridovich, Free Radical Biol. Med., 1990, 8, 21.
29 V. A. Roginsky, T. K. Barsukova, G. Bruchelt and H. B. Stegmann,
Z. Naturforsch., Teil C, 1997, 52, 380.
30 V. A. Roginsky, T. K. Barsukova and L. M. Pisarenko, unpublished
work.
31 G. R. Buettner, Arch. Biochem. Biophys., 1993, 300, 535.
32 B. H. J. Bielski, A. O. Allen and H. A. Schwarz, J. Am. Chem. Soc.,
1981, 103, 3516.
To provide support for this view, a correlation between the
rate of QH2 autoxidation and ∆E1 or k1 is required. The major
problem is the evident shortage in the systematic and com-
parable kinetic information on the process under consideration.
As a rule, we have a chance to correlate the oxidizability of QH2
determined within a single work only. For this reason we restrict
our consideration to a few remarks and specific examples.
Doing this, we should take into account that the rate of QH2
oxidation is expected to depend not only on the rate of reaction
Ϫ
(1) but also on other factors including the reactivity of Q
᝽
towards oxygen in the equilibrium (9). If E(Q/Q Ϫ) > Ϫ150 mV,
᝽
Ϫ
Q
Ϫ ϩ O2
Q ϩ O2
(9)
᝽
᝽
equilibrium (9) is shifted to the left.29 The situation may be
altered by adding superoxide dismutase (SOD) that effectively
purges the system from O2 Ϫ. O’Brien1 reported the elevated
᝽
oxidizability of chloro-substituted 1,4-hydroquinones though
the values of E(Q Ϫ/QH2) for these QH2 are very high (Table 2).
᝽
The non-substituted 1,4-benzoquinone for which E(Q Ϫ/QH2)
᝽
is also very positive (Table 2) was reported to display rather
high oxidizability when SOD was added.14 The oxidizability of
methyl-substituted 1,4-hydroquinones decreases (with adding
SOD) with the increase of the number of methyl groups30
although E(Q Ϫ/QH2) becomes less positive in this direction
᝽
(Table 2). In the meantime, the oxidizability of methyl-
substituted 1,4-hydroquinones correlates reasonably with ∆E1
and k1.30 Besides, the elevated oxidizability of QH2 8 and QH2
111,2 is in line with a rather high value of k1 (Table 1). Elevated
1,2
oxidizibility of several other QH2 combines, as a rule, with
elevated values of ∆E1. 1,4,5,8-Tetrahydroxynaphthalene
(∆E1 = Ϫ95 mV), 2,3-dimethoxy-1,4-dihydroxynaphthalene
(∆E1 = Ϫ130 mV) and adriamycine (∆E1 = ϩ70 mV) are
examples of this.
This approach probably may be applied to the oxidation of
substrates other than QH2. Ascorbate was reported to oxidize
very slowly in the absence of a catalyst and not to display any
tendency for autoacceleration of this process.17,18 This suggests
that the rate of the reaction between ascorbate, AscHϪ and its
oxidized form, dehydroascorbic acid, DAsc, with the formation
of the ascorbyl radical, Asc Ϫ, (an analog of reaction (1)) is
Paper 8/07650B
᝽
876
J. Chem. Soc., Perkin Trans. 2, 1999, 871–876