Kinetics and mechanism of the stoichiometric oxygenation of the ionic zinc(II) flavonolate complex [Zn(fla)(idpa)]ClO4 (fla = flavonolate; idpa = 3,3′-iminobis(N,N-dimethylpropylamine))

Article abstract of DOI:10.1016/S1381-1169(01)00130-3

The preparation and characterization of [Zn^II(4′R-fla)(idpa)]ClO₄ (fla = flavonolate; idpa = 3,3′-iminobis(N,N-dimethylpropylamine) complexes were described. Oxygenation of Zn^II(fla)(idpa)]ClO₄ in dimethylformamide (DMF) solution at elevated conditions gave Zn^II(idpa)(O-bs)]ClO₄ (O-bs = O-benzoylsalicylate) and CO. The oxygenolysis of [Zn^II(fla)(idpa)]ClO₄ in DMF was followed by electronic spectroscopy and the rate constants were determined according to the rate law -d[{Zn^II(fla)(idpa)}]ClO₄]/dt = k_obs [{Zn^II(fla)(idpa)}ClO₄][O₂]. The rate constant, activation enthalpy and entropy at 373 K were k_obs/M-sec =(3.11 ± 0.12) × 10^-3, ΔH (kJ/mol) = 75 ± 5, @Delta;S (J/mol-K) = -96 ± 13. The reaction fitted a Hammett linear free energy relationship for 4′-substituted flavonols, and electron-releasing groups made the oxygenation reaction faster. In the course of the reaction O₂^·- was formed suggesting a SET mechanism. [Zn^II(fla)(idpa)]ClO₄ did not catalyze the oxygenation of flavonols.

Full text of DOI:10.1016/S1381-1169(01)00130-3

Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
Kinetics and mechanism of the stoichiometric oxygenation
of the ionic zinc(II) flavonolate complex
[Zn(fla)(idpa)]ClO4 (fla = flavonolate;
idpa = 3,3ꢀ-iminobis(N,N-dimethylpropylamine))
László Barhácsa, József Kaizera,b, Gábor Speiera,b,∗
a
Department of Organic Chemistry, University of Veszprém, 8201 Veszprém, Hungary
b
Research Group for Petrochemistry, Hungarian Academy of Sciences, 8201 Veszprém, Hungary
Received 16 November 2000; accepted 8 March 2001
Abstract
The preparation and characterization of [ZnII(4ꢀR-fla)(idpa)]ClO4 (fla = flavonolate; idpa = 3,3ꢀ-iminobis(N,N-dimethyl-
propylamine) complexes are described. Oxygenation of [ZnII(fla)(idpa)]ClO4 in dimethylformamide solution at elevated
conditions gives [ZnII(idpa)(O-bs)]ClO4 (O-bs = O-benzoylsalicylate) and carbon monoxide. The oxygenolysis of [ZnII(fla)
(idpa)]ClO4 in DMF was followed by electronic spectroscopy and the rate constants were determined according to the rate
law −d[{ZnII(fla)(idpa)}ClO4]/dt = kobs [{ZnII(fla)(idpa)}ClO4][O2]. The rate constant, activation enthalpy and entropy at
373 K are as follows: kobs (M−1 s−1) = (3.11 0.12) × 10−3, ꢀH (kJ mol−1) = 75 5, ꢀS (J mol−1 K−1) = −96 13.
The reaction fits a Hammett linear free energy relationship for 4ꢀ-substituted flavonols, and electron-releasing groups make the
oxygenation reaction faster. In the course of the reaction O2•− was formed suggesting a SET mechanism. [ZnII(fla)(idpa)]ClO4
does not catalyze the oxygenation of flavonols. © 2001 Elsevier Science B.V. All rights reserved.
‡
‡
Keywords: Flavonol 2,4-dioxygenase; Quercetinase; Flavonol; Zinc complexes; Oxygenation; Kinetics; Dioxygenase
1. Introduction
these reactions is of primary interest. Fungi such as
Aspergillus or Pullularia species oxidatively degrade
quercetin (1a) and related flavonols into a depside
and carbon monoxide (Eq. (1)) [6–9]. The copper-
containing metalloenzyme flavonol 2,4-dioxygenase
is responsible for this reaction [10].
Dioxygenation of various substrates with molecular
oxygen are selectively catalyzed by different types of
metalloenzymes. Oxygenases are metal-containing
proteins and a fair number of them utilizes copper at
their active sites [1–5]. Since oxidation reactions in
biological systems as well as in the industry play an
important role, the understanding of the nature of
∗ Corresponding author. Tel.: +36-88-422-022;
fax: +36-88-427-492.
E-mail address: speier@almos.vein.hu (G. Speier).
(1)
1381-1169/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S1381-1169(01)00130-3

118
L. Barha´cs et al. / Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
In previous studies the intermediate formation
oxygen. Model oxygenation reactions on quercetin
and the parent compound flavonol have been car-
ried out in order to understand the enzymic reaction.
Base-catalyzed oxygenation of quercetin and related
flavonols under aqueous [21], and nonaqueous condi-
tions [22,23], and reactions with superoxide [24] have
been investigated. Metal complexes of cobalt [25–27]
and copper [28–30] have been found to act as cata-
lysts for the oxygenation reaction. Copper(I) [31] and
copper(II) [32–35] flavonolate complexes were also
successfully used for the oxygenation of flavonol.
We have previously reported that [CuII(fla)(idpa)]-
ClO4 as a model for the ES complex is a good mimic
for the flavonol 2,4-dioxygenase enzyme [17,35]. On
the bases of kinetic measurements of the stoichiomet-
ric and catalytic oxygenation reaction we assumed that
the redox active copper ion in the enzyme, and also in
the model reactions, acts as an electron buffer, due to
their ability to accept or donate electrons. In this sense
the flavonolato ligand is transformed to a radical, and
the copper(I) formed can activate dioxygen. In this
paper we report details for the synthesis and character-
ization of [ZnII(fla)(idpa)]ClO4, which is a model for
the ES complex with a nonredox metal zinc, and spec-
trophotometric studies aimed at elucidating the kinet-
ics and mechanism of the stoichiometric oxygenation
of [ZnII(fla)(idpa)]ClO4, to be compared with other
model systems with redox metal ions in order to clear
the possible role of the redox behavior of the metal ion.
of binary (ES) and ternary species (ESO2) has
been proposed at the oxygenation by flavonol
2,4-dioxygenase as shown in Eq. (2) [11–13]. Many
reports have appeared for the formation and structures
of ES from both enzyme and model studies, but until
now there has been only few mechanistic evidence for
the formation of ESO2 species. There are two types of
(2)
discussions on the structure of ES. One concerns
whether flavonol coordinates to the metallic center as
a bidentate ligand [11] or dissociates from the copper
ion forming free flavonolate ion in the activation step
[5]. The other concerns the change of the character of
the flavonolate ligand, i.e. how the flavonolate ligand
is activated for oxygenation: whether the flavonolate
ligand can be regarded as a radical as shown in (A)
or has an anionic character as in (B) (Fig. 1).
Extensive studies report on the coordination chem-
istry of flavonols of various metal ions. Recent
crystallographic studies of flavonolato complexes of
copper(I) [14], copper(II) [15–17], cobalt(III) [18],
and zinc(II) [19,20] disclosed the coordination mode
of the flavonolate ligand, geometries around the metal
ions, and their influence on the delocalization of
␲-electrons in the flavonolate ligand. These com-
plexes are important structural and functional models
in studies mimicking dioxygenases.
Apart from the coordination mode of flavonolate
ligands, it is important to know how the flavonolate
ligand is activated for the reaction with molecular
2. Experimental
2.1. Instrumentation
IR spectra were recorded in either Nujol or KBr
pellets on a Specord IR-75 (Carl Zeiss) spectrome-
ter. Electronic spectra were measured on a Shimadzu
UV-160 spectrometer using quartz cells. NMR spec-
tra were recorded on a Varian Unity-300 spectrome-
1
ter. H NMR shifts are reported as values downfield
from an external standard of Me4Si. GC analyses
were performed on HP 5830A and HP 5890 gas chro-
matographs equipped with a flame ionization detector
(CP SIL 8CB column), and a thermal conductiv-
ity detector (molecular sieve 5A column). GC–MS
measurements were recorded on a HP 5890II/5971
GC/MSD at 75 EV.
Fig. 1. Probable modes of coordination of flavonol in a flavonol–
metal binary intermediate.

L. Barha´cs et al. / Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
119
2.2. Materials and methods
(nujol) 3235, 1564, 1520, 1341, 1273, 1253, 1240,
1197, 1104, 1065, 982, 946, 919, 863, 833, 766, 690,
1
633, 553, 499 cm−1; H NMR ((CD3)2CO, 22◦C) δ
Solvents [diethyl ether and DMF (HPLC grade
with water content less than 0.03%)] were purchased
from Aldrich in analytically pure quality. Diethyl
ether was distilled from K–Na amalgam under Ar
atmosphere. Flavonol [21], 4ꢀ-methoxyflavonol [21],
4ꢀ-methylflavonol [36], and 4ꢀ-chloroflavonol [36]
were prepared by literature methods. The com-
pounds 3,3ꢀ-iminobis(N,N-dimethylpropylamine) and
Zn(ClO4)2·6H2O were obtained from Aldrich and
used without further purification. Diazomethane [37]
was freshly prepared according to the literature in
ether and immediately used for the methylation
reactions. Gaseous oxygen from Messergriesheim
was 99.6% and passed through P2O5 and Blaugel in
order to remove traces of water and other impurities.
7.3–8.8 (m, 8H, ArH), 4.4 (s, 1H), 3.2–3.4 (m, 4H),
2.85–3.1 (m, 4H), 2.65 (s, 12H), 2.4 (s, 3H), 2.05–2.15
(m, 4H). Anal. calcd. for C26H36N3O8ZnCl: C, 51.75;
H, 6.01; N, 6.96. Found C, 50.94; H, 5.93; N, 6.88.
2.5. [ZnII (4ꢀCl-fla)(idpa)]ClO4 (3c)
Yield: 4.35 g, 70%; mp 241–243◦C; UV–VIS
(λmax, DMF) 262.5 nm (log ε 4.18), 419.5 (4.30); IR
(KBr) 3247, 1553, 1537, 1473, 1411, 1351, 1318,
1233, 1214, 1177, 1144, 1102, 1063, 1041, 1002,
964, 908, 843, 752, 681, 638, 548, 495 cm−1 1H
;
NMR ((CD3)2CO, 22◦C) δ 7.3–8.8 (m, 8H, ArH),
4.4 (s, 1H), 3.2–3.4 (m, 4H), 2.85–3.1 (m, 4H),
2.65 (s, 12H), 2.05–2.15 (m, 4H). Anal. calcd. for
C25H33N3O7ZnCl2: C, 48.13; H, 5.33; N, 6.74. Found
C, 47.52; H, 5.18; N, 6.38.
2.3. [ZnII (4ꢀR-fla)(idpa)]ClO4 (3) (general
procedure)
Zn(ClO4)2·6H2O (3.724 g, 10 mmol) and 3,3ꢀ-imi-
nobis(N,N-dimethylpropylamine) (2.3 cm3, 10 mmol)
were dissolved in 150 cm3 toluene–ethanol (1:1) and
stirred under an atmosphere of argon for 1 h, then
flavonol (2.38 g, 10 mmol) was added and the solu-
tion refluxed for 3 h. The solution was allowed to
cool to room temperature and a yellow solid which
deposited was then filtered off and recrystallized from
toluene–ethanol (1:1) to give yellow crystals of (3a)
(1.75 g, 30%): mp 231–233◦C; UV–VIS (λmax, DMF)
260 nm (log ε 4.17), 419 (4.32); IR (KBr) 3250, 1565,
1526, 1486, 1442, 1417, 1362, 1318, 1259, 1213,
1186, 1156, 1115, 1081, 1042, 999, 915, 757, 715,
692, 669, 629, 551, 497 cm−1; 1H NMR ((CD3)2CO,
22◦C) δ 7.3–8.8 (m, 9H, ArH), 4.4 (s, 1H), 3.2–3.4 (m,
4H), 2.85–3.1 (m, 4H), 2.65 (s, 12), 2.05–2.15 (m, 4H).
Anal. calcd. for C25H34N3O7ZnCl: C, 50.95; H, 5.81;
N, 7.13. Found C, 50.43; H, 6.09; N, 6.94. Complexes
[ZnII(4ꢀR-fla)(idpa)]ClO4 (3b–d) were prepared by the
same method as that described above for the prepa-
ration of (3a).
2.6. [ZnII (4ꢀOMe-fla)(idpa)]ClO4 (3d)
Yield: 4.95 g, 80%; mp 225–227◦C; UV–VIS (λmax
,
DMF) 266 nm (log ε 4.212), 427 (4.32); IR (KBr)
3247, 1556, 1502, 1492, 1416, 1363, 1316, 1256,
1217, 1176, 1156, 1116, 1043, 920, 836, 762, 683,
629, 576 cm−1; 1H NMR ((CD3)2CO, 22◦C) δ 7.3–8.8
(m, 8H, ArH), 4.4 (s, 1H), 3.84 (s, 3H), 3.2–3.4 (m,
4H), 2.85–3.1 (m, 4H), 2.65 (s, 12H), 2.05–2.15 (m,
4H). Anal. calcd. for C26H36N3O8ZnCl: C, 50.41; H,
5.86; N, 6.78. Found C, 49.75; H, 5.76; N, 6.55.
2.7. Oxygenation of [ZnII (fla)(idpa)]ClO4 (3a)
0.295 g (0.5 mmol) [ZnII(fla)(idpa)]ClO4 was dis-
solved in 15 cm3 DMF and stirred at 100◦C under
dioxygen atmosphere for 10 h. The formation of
[ZnII(idpa)(O-bs)]ClO4 (4a) from [ZnII(fla)(idpa)]-
ClO4 (3a) requires dioxygen but no apparent oxygen
uptake was observed, because the absorption of dioxy-
gen and the liberation of carbon monoxide compensate
each other. The GC analysis of the gas phase showed
0.14 mmol CO (28% conversion). Approximately,
0.14 mmol of dioxygen has been consumed and as
much carbon monoxide has been evolved during
the reaction. 0.5 cm3 water and 2 cm3 diazomethane
2.4. [ZnII (4ꢀMe-fla)(idpa)]ClO4 (3b)
Yield: 4.53 g, 75%; mp 197–199◦C; UV–VIS
(λmax, DMF) 261.5 nm (log ε 4.106), 422 (4.208); IR

120
L. Barha´cs et al. / Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
solution (in diethyl ether) was added to 1 cm3 of
the reaction mixture at room temperature and the
conversion of the coordinated flavonolate (35%) and
the yields of the products N,N-dimethylbenzamide
(14%), benzoic acid (20%), salicylic acid (34%) and
O-benzoylsalicylic acid (32%) were determined by
GC.
(ca. every 10 min). Experimental conditions are sum-
marized in Table 1. The temperature was determined
with an accuracy of 0.5◦C and the pressure of
dioxygen with an accuracy of 0.5%. The O2 con-
centration was calculated from literature data [38]
taking into account the partial pressure of DMF [39]
and assuming the validity of Dalton’s law.
2.8. Kinetic measurements
3. Results and discussion
Reactions of [ZnII(4ꢀR-fla)(idpa)]ClO4 complexes
with O2 were performed in DMF solutions. In a
typical experiment [ZnII(fla)(idpa)]ClO4 (3) was dis-
solved under argon atmosphere in a thermostatted
reaction vessel with an inlet for taking samples with
a syringe, and connected to mercury manometer to
regulate constant pressure. The solution was then
heated to the appropriate temperature. A sample was
then taken by syringe, and the initial concentra-
tion of [ZnII(fla)(idpa)]ClO4 (3) was determined by
UV–VIS spectroscopy measuring the absorbance of
the reaction mixture at 419 nm (log ε 4.32) {λmax of
a typical band of [ZnII(fla)(idpa)]ClO4}. The argon
was then replaced by dioxygen and the consumption
of [ZnII(fla)(idpa)]ClO4 was analyzed periodically
The kinetics of the oxygenation reaction in DMF
was followed by determining the concentration of
[ZnII(fla)(idpa)]ClO4 (3a) in the reaction solution as
a function of time by UV–VIS spectroscopy under
constant dioxygen pressure, determining the variation
of the absorbance of the flavonolate band at 419 nm.
The solvent had a very low water content (below
0.03%), however by rising the water concentration to
2.38×10−2 M by addition of water did not change the
reaction rate (Table 1; experiments 12). A simple rate
law for the reaction between [ZnII(fla)(idpa)]ClO4
(3a) and O2 is given in Eq. (2). In order to deter-
mine the rate dependence on the various reactants,
oxygenation runs were performed at various initial
Table 1
Kinetic data for the oxygenation of [ZnII(fla)(idpa)]ClO4 in DMF solution
Experiment
number
Temperature
(◦C)
103[O2]
(mol dm−3
104[Zn]
(mol dm−3
105k (s−1
)
R (%)b
103kobs
(s−1 mol−1 dm3)
108 d[Zn]/dt
(mol dm−3 s−1
)
a
)
)
1
2
3
4
5
6
7
100
100
100
100
100
100
100
7.70
7.70
7.70
7.70
1.54
4.50
5.50
3.42
4.83
9.81
14.13
14.13
14.25
14.58
2.58
2.10
2.20
2.54
0.60
1.50
1.59
99.85
99.79
99.50
99.53
99.28
99.48
99.68
3.35
2.73
2.86
3.30
3.89
3.33
2.89
3.11
5.20
9.95
19.87
73.19
3.10
0.05
0.07
0.08
0.10
0.18
0.10
0.08
0.12c
0.12
0.45
0.32
5.12d
0.05e
0.88
1.01
2.16
3.59
0.85
2.14
2.32
8
9
10
11
12
110
120
130
100
100
7.45
6.96
6.54
7.70
7.70
13.45
10.94
12.60
4.88
3.87
6.93
12.99
5.64
99.83
99.37
99.97
99.09
99.85
7.15
2.39
a In 50 cm3 DMF.
b Correlation coefficients of least-squares regressions.
ꢀ
ꢂ
ꢁ
ꢁ
c Mean value of the kinetic constant k and its standard deviation σ(k) were calculated as k =
iwiki/ wi and σ(k) =
i
ꢀ
ꢂ1/2
ꢁ
ꢁ
iwi(ki − k)2/(n − 1) wi
, where wi = 1/σi2.
i
d 48 cm3 DMF-2 cm3 py.
e In the presence of added water [H2O] = 2.38 × 10−2 M.

L. Barha´cs et al. / Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
121
substrate concentrations (Table 1; experiments 1–4)
the validity of Dalton’s law, the dissolved concen-
tration of O2 being 7.7 × 10−3 mol dm−3 at 100◦C
and 760 mm Hg O2 pressure) show that the dioxygen
concentration affects the rate of the reaction (exper-
iments 4–7; columns kꢀ and R in Table 1), and that
the reaction is first order with respect to dioxygen
concentration.
On the basis of the results above, one can conclude
that the oxygenation of [{ZnII(fla)(idpa)}ClO4] obeys
an overall second order rate equation with m = n = 1
(Eq. (2)), from which a mean value of the kinetic
constant kobs of (3.11 0.12) × 10−3 mol−1 dm3 s−1
at 373.16 K was obtained (Table 1). Addition of small
amount of pyridine to the reaction mixture accel-
erated the reaction rate to kobs = (7.32 0.51) ×
10−2 mol−1 dm3 s−1 with respect to [ZnII(fla)(idpa)]
ClO4 (experiment 11 in Table 1).
d[{ZnII(fla)(idpa)}ClO4]
−
dt
d[{ZnII(O-bs)(idpa)}ClO4]
=
dt
= k[{ZnII(fla)(idpa)}ClO4]m[O2]n
(2)
and at different dioxygen pressures (Table 1; experi-
ments 4–7 under pseudo first order conditions with a
constant dioxygen pressure). Eq. (2) can then have a
more simple feature as written in Eq. (3), kꢀ being the
pseudo first order rate constant.
d[{ZnII(fla)(idpa)}ClO4]
−
dt
= kꢀ[{ZnII(fla)(idpa)}ClO4]m
(3)
The activation parameters for the oxygenation
reaction were determined from the temperature depen-
dence of the kinetic constant kobs. The temperature-
dependent reaction rate measurements in the range of
373.16–403.16 K (experiments 1–7, 8–10 in Table 1)
resulted in a straight line in the Eyring plot, with a
correlation coefficient of 99.58% and activation para-
Measuring the time dependence of the change of
concentration of [ZnII(fla)(idpa)]ClO4 (3a) during the
oxygenation shows that plots of log[{ZnII(fla)(idpa)}
ClO4] versus time were linear in experiments 1–4, in-
dicating that the reaction is first order with respect to
substrate concentration. Columns kꢀ and R in Table 1
report slopes and the correlation coefficients obtained
by the least-squares method for these linear regres-
sions. A typical first order plot for experiment 3 shows
that the reaction remains first order for the whole time
in which the experiment was followed (58% conver-
sion, 6.5 h).
meters Ea = 78 5 kJ mol−1, ꢀH = 75 5 kJ mol−1
‡
‡
and ꢀS = −96
13 J mol−1 K−1, respectively.
Though activation parameters are often not discrim-
inating factors in recognizing the reaction pathway,
‡
the negative entropy of activation (ꢀS ), however,
clearly indicates an associative mode of activation in
the rate-determining step.
From variations of the reaction rates, plots of
−d[{ZnII(fla)(idpa)}ClO4]/dt versus [{ZnII(fla)(idpa)}
ClO4] were also linear in experiments 1–4 with a
correlation coefficient of 99.38%, reinforcing that the
reaction is indeed first order with respect to substrate
concentration.
Reaction rates on the oxygenation of 4ꢀ-substituted
flavonolato zinc complexes ([ZnII(4ꢀR-fla)(idpa)]ClO4
(3) Eq. (4)) under identical conditions were deter-
mined
(4)
Experiments made at different dioxygen concen-
trations (calculated from literature data assuming
(with various electron-withdrawing or releasing
substituents R) in order to find out electronic effects

122
L. Barha´cs et al. / Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
Table 2
Hammet data for the oxygenation of [ZnII(4ꢀR-fla)(idpa)]ClO4
Complex
R
σ
102k (s−1 mol−1 dm3)a
3a
3b
3c
3d
H
CH3
Cl
0.000
1.00
1.84
0.78
1.81
0.05
0.06
0.01
0.05
−0.170
+0.227
−0.268
OCH3
a In DMF at 120◦C.
on the reaction rate (Table 2). The Hammett plot ob-
tained is shown in Fig. 2. Since the rate constants of
all the substituents fall on a single line, we conclude
that there is no change in the mechanism over the
series of zinc flavonolates studied. The linearity in
Fig. 2 demonstrates a linear free-energy relationship
[40], and the reaction constant ρ was found to be
−0.83. The observation that ρ is large and negative
supports the contention that the substituents at the
para-position have a significant electronic effect at the
remote reaction center. Electron-donating substituents
increase the reaction rate, while electron-withdrawing
substituents slow down the reaction. These results
suggest that the dioxygen species at the transition
state should be electrophilic.
Fig. 2. Hammett plot for the oxygenation of [ZnII(4Rꢀ-fla)(idpa)]
ClO4 complexes (Table 2).
Deprotonation of flavonols enhances the reactivity
of the corresponding anions toward dioxygen [27].
On the bases of the kinetic, CV and EPR studies
a single electron transfer (SET) was assumed for
the oxygenation of potassium flavonolate. The key
step was thought to be the formation of a flavonoxy
radical in a bimolecular process followed by the re-
action of the flavonoxy radical with the superoxide
ion [23]. Flavonolate as a chelating ligand forms
a stable complex with zinc ion in the presence of
3,3ꢀ-iminobis(N,N-dimethylpropylamine). The Zn(II)
center is in distorted trigonal-bipyramidal environ-
ment. Due to coordination of flavonolate to the zinc
there are changes in the bond lengths of the pyranone
A reaction mechanism that fits the chemical, spec-
troscopic, kinetic and thermodynamic data is shown
in Scheme 1. Flavonol and quercetin are assumed to
be reactive substrates in the enzyme reaction [11–13].
However, it is also known that these substrates are
also inert towards dioxygen in their protonated forms.
Scheme 1.

L. Barha´cs et al. / Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
123
ring, which may be assigned to delocalization of
the ␲-system over the whole molecule [20]. Kinetic
studies on the oxygenation of the flavonolato zinc
complexes established a second-order overall rate
expression, indicating that the rate-determining
step must be bimolecular, namely the reaction of
[ZnII(fla)(idpa)]ClO4 (3a) with dioxygen. Reduction
of nitroblue tetrazolium (NBT) to blue formazan
was used as an indicator for superoxid anion forma-
tion and this test indicated its presence. Addition of
NBT to the reaction mixture, the yellow color turns
to intense green and a new intense band at 521 nm
arises. This absorption has been assigned to formazan
signaling the presence of superoxide anion [41].
On that basis we believe that the oxygenolysis of
flavonolate complexes with nonredox metals, such as
[ZnII(fla)(idpa)]ClO4, in aprotic solvents has a single
electron transfer (SET) mechanism. The second order
overall rate equation can be interpreted by two mecha-
nistic pathways with different rate-determining steps,
both satisfying the kinetic data (Scheme 1). In the first
case the slow reaction of [ZnII(fla)(idpa)]ClO4 with
O2 is an irreversible rate-determining electron trans-
fer from the substrate to O2 forming the flavonoxy
radical complex 5 and superoxide ion. This is fol-
lowed by a fast radical radical coupling reaction of
the flavonoxy radical (5) and superoxide ion to give a
deprotonated hydroperoxide species (6), which after
can be simplified then further to a single second order
•−
rate equation by assuming k2/k and K1[O2]/[O2
being much smaller than 1.
]
−1
d[{ZnII(fla)(idpa)}ClO4]
−
dt
−
k1k2[{ZnII(fla•)(idpa)}ClO4][O2][O2
]
•
=
(5)
−
•
k1[O2] + (k−1 + k2)[O2
]
d[{ZnII(fla)(idpa)}ClO4]
dt
−
K1k2[{ZnII(fla•)(idpa)}ClO4]0−[O2]
=
•
k2/k−1 + 1 + K1[O2]/[O2
]
= kobs[{ZnII(fla)(idpa)}ClO4]0[O2]
(6)
As a conclusion it can be said that the interac-
tion between the bidentate flavonolate ligand and the
Lewis acid zinc(II) ion leads to lowering the electron
density on the flavonolate ligand, which causes a de-
crease in reaction rate of the single electron transfer
from the Zn(II)-coordinated flavonolate to dioxy-
•−
gen or the radical radical coupling of 5 and O2
.
This is supported by the Hammett correlation and a
rate-accelerating effect of added pyridine on the reac-
tion. In the oxygenation reaction of the flavonolate in
various systems the reaction rates are in the following
order: n-Bu4Nfla > flaK > [ZnII(fla)(idpa)]ClO4 +
py > [ZnII(fla)(idpa)]ClO4 < [CuII(fla)(idpa)]ClO4
(Table 3). The small difference in ρ values observed
for Zn and potassium-containing systems suggests
that the mechanism of the two oxygenation reac-
tions are the same. The higher electron density on
the flavonolate increases the reducing power of the
enolate or carbanion form (Fig. 1) making the elec-
tron transfer easier or shifting the pre-equilibrium
more to the side of 5 and O2•−, that means that K1
becomes bigger. The difference in reaction rates ob-
served for [ZnII(fla)(idpa)]ClO4 in the presence and
absence of py indicates that there are two pathways
for the oxygenation, namely a slower process in DMF
where the idpa ligand is bonded to zinc, and a faster
one in the presence of an excess pyridine, where the
ligand idpa is probably replaced by the strongly lig-
ating pyridine, while the flavonolate remains bonded
to the zinc ion. That is supported by conductivity
measurements too, where the conductivity of the so-
lution (Λ0 = 35 cm2 mol−1 ohm−1) does not change,
excluding the formation of new ions.
=
an intramolecular An reaction on 4C O leads to an
unstable endoperoxide (7), and its decomposition
results then in the O-benzoylsalicylato zinc(II) com-
plex and carbon monoxide. All attempts to isolate
or detect the endoperoxide 7 failed. That suggest
that this type of endoperoxides (7) are very unstable
if compared to other endoperoxides [42] where the
elimination of CO is not possible. In the second case
it can be assumed that [ZnII(fla)(idpa)]ClO4 under-
goes electron transfer to O2 in a fast reversible step
resulting in the flavonoxy radical complex 5 and su-
peroxide ion. This pre-equilibrium lies very much on
the side of the starting reagents (k1/k being small).
−1
This is followed then by the rate-determining step
where [ZnII(fla•)(idpa)]ClO4 reacts with O2•−. All
consecutive steps are fast and identical with those out-
lined before. Applying steady-state treatment for this
mechanism, [ZnII(fla•)(idpa)]ClO4/dt = 0, the rate
Eq. (5) can be deduced. Rearranging Eq. (5) the rate
equation can have the form as shown in Eq. (6). This

124
L. Barha´cs et al. / Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
Table 3
Oxygenation rates and activation parameters for fla(n-Bu4N), flaK, [ZnII(4ꢀR-fla)(idpa)]ClO4 and [CuII(4ꢀR-fla)(idpa)]ClO4
kobs (s−1 mol−1 dm3)
ꢀH (kJ mol−1
)
ꢀS (J mol−1 K−1
)
ρ
‡
‡
Complex
fla(n-Bu4N)a
(5.18
(3.28
(7.32
(3.11
(6.13
0.16) × 10−1
0.10) × 10−1
5.12) × 10−2
0.12) × 10−3
0.16) × 10−3
flaKa
29
2
−161
6
−0.73
[ZnII(fla)(idpa)]ClO4
[ZnII(fla)(idpa)]ClO4
[CuII(fla)(idpa)]ClO4
b
c
75
64
5
5
−96
13
13
−0.83
−0.29
−120
a 40◦C, DMF.
b 100◦C, DMF + py.
c 100◦C, DMF.
In comparison of the two mechanism we believe
tha•t−the one with the rate-determining step of 5 and
O2
is operative. Recent data on the rate constant
of the electron transfer reaction from C60-anion rad-
ical (C60•−) to molecular oxygen was found to be
(3.7 0.2) × 102 mol−1 dm3 s−1 [43]. The relatively
slow rate was explained by the endothermic process
(ꢀGet > 0) based on CV measurements. In our case
the rate constant is three order of magnitude smaller
(3.11 0.12) × 10−3 M−1 s−1. Despite the fact that
the ꢀG value is even bigger in this case (based on
calculation of the CV values [23]) and so an even
slower reaction rate could be expected, we count
on a mechanism in which the complex 3a reacts
with O2 in a fast pre-equilibrium and the reaction of
Fig. 3. Spectral changes at the reaction of [ZnII(fla)(idpa)]ClO4
with excess flavonol. [Zn]0 = 1.5 × 10−4 M, [flaH]0 = 6.0 ×
10−4 M, T = 130◦C, 50 cm3 DMF, Ar.
3a is probably dismantled during the reaction, while
ZnII(fla)2 is formed and the basic ligand idpa depro-
tonates three further flavonol molecules. The species
with the flavonolate ions so formed are much less re-
active towards dioxygen than 3a.
•−
[ZnII(fla•)(idpa)]ClO4 with O2 is rate-determining.
Attempts to use [ZnII(fla)(idpa)]ClO4 as a catalyst
for the oxygenation of flavonol failed. The reason for
that can be either that the flavonol cannot be replaced
by the product O-benzoylsalicylic acid and so the cat-
alytic cycle is disrupted, or 3a reacts with an excess of
flavonol to an unreactive species, which is a pittfall in
the cycle. Spectrophotometric monitoring of the reac-
tion of 3a with flavonol under argon proved the second
case to be true. As shown in Fig. 3 the time profile of
the reaction indicates the formation of a new species
by increasing absorbance at 419 nm and decreasing
intensities at 342.5 nm with isobestic points at 375.5
and 288 nm. The increase of the band at 419 nm, char-
acteristic for the flavonolate ion, suggests that further
4 moles of flavonolate ions are formed at the end of the
reaction. This may be interpreted as shown in Eq. (7).
The above results indicate that the copper ion at the
active site of quercetinase and in model systems under-
goes valence tautomerism CuII(fla−) ↔ CuI(fla•) (in
a similar manner as in the case of CuII(catecholate) ↔
CuI(semiquinone) systems [44]), and the CuI(fla•)
species reacts with dioxygen to the superoxo cop-
per complex CuII(O2)(fla•). In model studies with
Zn(II) valence tautomerism is excluded, however
the coordinated flavonolate ligand in ZnII(fla) under-
goes electron transfer to O2 to •g−ive ZnII(fla•) and
free superoxide ion. The free O2 could be proved
but the detection of ZnII(fla•) failed by EPR due
to its in•s−tability. However, the reaction of ZnII(fla•)
with O2 and consecutive reactions (see Scheme 1)
lead to the endoperoxide and to the zinc complex
of O-benzoylsalicylic acid ZnII(O-bs)(idpa)+. In that
case a different mechanism is operative resulting in
the same products. It is also interesting to note in the
[ZnII(fla)(idpa)]+ + 4flaH
→ [ZnII(fla)2] + idpa43+ + 3fla− + H+
(7)

L. Barha´cs et al. / Journal of Molecular Catalysis A: Chemical 172 (2001) 117–125
125
case of the Zn model, contrary to its copper analogue,
the exchange of the product ligand O-bs by fla is not
possible in attempted catalytic reactions, which means
that ZnII(fla)(idpa)+ does not catalyze the oxygena-
tion of flavonol. This is explained by the formation of
new, stable, inactive zinc flavonolate complexes.
[18] W. Hiller, A. Nishinaga, A.Z. Rieker, Z. Naturforsch. 47b
(1992) 1185.
[19] T.A. Annan, C. Peppe, D.G. Tuck, Can. J. Chem. 68
(1990) 423.
[20] É. Balogh-Hergovich, J. Kaizer, G. Speier, G. Huttner, P.
Rutsch, Acta Crystallogr. C55 (1999) 557.
[21] A. Nishinaga, T. Tojo, H. Tomita, T.J. Matsuura, J. Chem.
Soc., Perkin Trans. 1 (1979) 2511.
[22] V. Rajananda, S.B. Brown, Tetrahedron Lett. 22 (1981) 4331.
[23] G. Speier, J. Kaizer, L. Barhács, J. Org. Chem. 65 (2000)
3449.
Acknowledgements
[24] M.M.A. El-Sukkary, G. Speier, J. Chem. Soc., Chem.
Commun. (1981) 745.
[25] A. Nishinaga, T. Tojo, T. Matsuura, J. Chem. Soc., Chem.
Commun. (1974) 896.
[26] A. Nishinaga, N. Numada, K. Maruyama, Tetrahedron Lett.
30 (1989) 2257.
Financial support of the Hungarian National
Research Fund (OTKA T-7443, T-016285 and
T-30400) is gratefully acknowledged.
[27] A. Nishinaga, T. Kuwashige, T. Tsutsui, T. Mashino, K.
Maruyama, J. Chem. Soc., Dalton Trans. (1994) 805.
[28] M. Utaka, M. Hojo, Y. Fujii, A. Takeda, Chem. Lett.
(1984) 635.
[29] M. Utaka, A. Takeda, J. Chem. Soc., Chem. Commun. (1985)
1824.
[30] É. Balogh-Hergovich, G. Speier, J. Mol. Catal. 71 (1992).
[31] É. Balogh-Hergovich, J. Kaizer, G. Speier, V. Fülöp, L.
Párkányi, Inorg. Chem. 38 (1999) 3787.
[32] É. Balogh-Hergovich, J. Kaizer, G. Speier, Inorg. Chim. Acta
256 (1997) 9.
[33] É. Balogh-Hergovich, J. Kaizer, G. Speier, G. Argay, L.
Párkányi, J. Chem. Soc., Dalton Trans. (1999) 3847.
[34] É. Balogh-Hergovich, J. Kaizer, G. Speier, J. Mol. Catal. 159
(2000) 215.
[35] G. Speier, J. Kaizer, J. Pap, L. Barhács, Inorg. Chim. Acta,
in press.
[36] M.A. Smith, R.M. Newman, R.A. Webb, J. Heterocycl. Chem.
5 (1966) 425.
[37] F. Arndt, in: A.H. Blatt (Ed.), Organic Syntheses, Vol. 2,
Wiley, New York, 1943, p. 165.
References
[1] B.G. Fox, J.D. Lipscomb, in: C. Reddy, G.A. Hamilton (Eds.),
Biological Oxidation Systems, Academic Press, New York,
1990, p. 367.
[2] K.D. Karlin, Z. Tyeklár, Bioinorganic Chemistry of Copper,
Chapman & Hall, New York, 1992.
[3] E.I. Solomon, M.J. Baldwin, M.D. Lowery, Chem. Rev. 92
(1992) 521.
[4] K.D. Karlin, Science 261 (1993) 701.
[5] T. Funabiki, Catalysis by Metal Complexes, Oxygenases and
Model Systems, Kluwer Academic Press, Dordrecht, The
Netherlands, 1997.
[6] D.W. Westlake, G. Talbot, E.R. Blakley, F.J. Simpson, Can.
J. Microbiol. 5 (1959) 62.
[7] F.J. Simpson, G. Talbot, D.W. Westlake, Biochem. Biophys.
Res. Commun. 2 (1959) 621.
[8] S. Hattori, I. Noguchi, Nature 184 (1959) 1145.
[9] H.G. Krishnamurty, F.J. Simpson, J. Biol. Chem. 245 (1970)
1467.
[10] T. Oka, F.J. Simpson, H.G. Krishnamurty, Can. J. Microbiol.
18 (1972) 493.
[11] E. Makasheva, N.T. Golovkina, Zh. Obsch. Khim. 43 (1973)
1640.
[12] M. Thomson, C.R. Williams, Anal. Chim. Acta 85
(1976) 375.
[13] K. Takamura, M. Ito, Chem. Pharm. Bull. 25 (1977) 3218.
[14] G. Speier, V. Fülöp, L. Párkányi, J. Chem. Soc., Chem.
Commun. (1990) 512.
[15] É. Balogh-Hergovich, G. Speier, G. Argay, J. Chem. Soc.,
Chem. Commun. (1991) 551.
[16] I. Lippai, G. Speier, G. Huttner, L. Zsolnai, Acta Crystallogr.
C53 (1997) 1547.
[17] É. Balogh-Hergovich, J. Kaizer, G. Speier, G. Huttner, L.
Zsolnai, Inorg. Chim. Acta 304 (2000) 72.
[38] A. Kruis, Landolt-Börnstein, Board 4, Teil 4., Springer,
Berlin, 1976, p. 269.
[39] G. Ram, A.R. Sharaf, J. Ind. Chem. Soc. 45 (1968) 13.
[40] J. March, Advanced Organic Chemistry: Reactions, Mecha-
nisms, and Structure, 4th Edition, Wiley, New York, 1992,
pp. 278–286.
[41] B.H.J. Bielski, H.W. Richter, J. Am. Chem. Soc. 110 (1988)
8085.
[42] E.L. Clennan, C.S. Foote, in: W. Ando (Ed.), Organic
Peroxides, Wiley, Chichester, 1992, p. 225.
[43] T. Konishi, M. Fujitsuka, O. Ito, Chem. Lett. (2000) 202.
[44] C.G. Pierpont, C.W. Lange, in: K.D. Karlin (Ed.), Progress
in Inorganic Chemistry, Vol. 41, Wiley, New York, 1994,
pp. 422–424.