Kinetics and mechanism of the base-catalyzed oxygenation of 1H-2-phenyl-3-hydroxy-4-oxoquinolines in DMSO/H2O

Article abstract of DOI:10.1016/j.tet.2013.05.117

The oxygenation of 4′-substituted 1H-2-phenyl-3-hydroxy-4- oxoquinolines (PhquinH₂) in a DMSO/H₂O (50/50) solution leads to the cleavage products at the C2-C3 bond in about 75% yield at room temperature. The oxygenation, deduced from the product compositions, has two main pathways, one proceeding via an endoperoxide leading to CO-release, and the other through a 1,2-dioxetane intermediate without CO-loss. The reaction is specific base-catalyzed and the kinetic measurements resulted in the rate law -a...?[PhquinH₂]/a...?t=kOH ^_- [OH^-] [PhquinH₂] [O₂]. The rate constant, activation enthalpy, and entropy at 303.16 K are as follows: kOH^_-=(2.42±0.03)×10³mol ^-2L²s^-1; ΔG^?=73. 13±4.02 kJ mol^-1; ΔH^?=70.60±4. 04 kJ mol^-1; ΔS^?=-28±2 J mol ^-1 K^-1. The reaction fits a Hammett linear free energy relationship for 4′-substituted substrates, and electron-releasing groups make the oxygenation reaction faster (ρ=-0.258). The EPR spectrum of the reaction mixtures showed the presence of the organic radical 1H-2-phenyl-3-oxyl-4-oxoquinoline and superoxide ion due to single electron transfer from the carbanion to dioxygen. The pathway via 1,2-dioxetane could be proved by chemiluminescence measurements.

Full text of DOI:10.1016/j.tet.2013.05.117

Tetrahedron 69 (2013) 6666e6672
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Kinetics and mechanism of the base-catalyzed oxygenation
of 1H-2-phenyl-3-hydroxy-4-oxoquinolines in DMSO/H2O
a
,
y
a
,
a
b
*
ꢀ
ꢀ
ꢀ
Miklos Czaun , Gabor Speier , Jozsef Kaizer , Nadia El Bakkali-Taheri ,
Etelka Farkas c
Department of Chemistry, University of Pannonia, Egyetem u. 10., 8200 Veszprem, Hungary
a
ꢀ
b
ꢀ
ꢀ
ꢀ
BiosCiences, Institute des Sciences Moleculaires de Marseille, UMR 7313, CNRS Aix-Marseille Universite, Service 342, Campus de Saint-Jerome,
13397 Marseille cedex 20, France
c Department of Inorganic Chemistry, University of Debrecen, 4032 Debrecen, Hungary
a r t i c l e i n f o
a b s t r a c t
The oxygenation of 40-substituted 1H-2-phenyl-3-hydroxy-4-oxoquinolines (PhquinH2) in a DMSO/H2O
(50/50) solution leads to the cleavage products at the C2eC3 bond in about 75% yield at room tem-
perature. The oxygenation, deduced from the product compositions, has two main pathways, one pro-
ceeding via an endoperoxide leading to CO-release, and the other through a 1,2-dioxetane intermediate
without CO-loss. The reaction is specific base-catalyzed and the kinetic measurements resulted in the
Article history:
Received 22 February 2013
Received in revised form 17 May 2013
Accepted 28 May 2013
Available online 2 June 2013
rate law ꢀd½PhquinH2ꢁ=dt ¼ kOH [OHꢀ] [PhquinH2] [O2]. The rate constant, activation enthalpy, and
ꢀ
Keywords:
ꢀ2
2
entropy at 303.16 K are as follows: kOH ¼ ð2:42 ꢂ 0:03Þ ꢃ 103 mol L sꢀ1
;
D
Gz¼73.13ꢂ4.02 kJ molꢀ1
;
ꢀ
Oxoquinolines
Endoperoxide
1,2-Dioxetane
Oxygenation
Degradation
D
Hz¼70.60ꢂ4.04 kJ molꢀ1
;
D
Sz¼ꢀ28ꢂ2 J molꢀ1
K
ꢀ1. The reaction fits a Hammett linear free energy
relationship for 40-substituted substrates, and electron-releasing groups make the oxygenation reaction
faster (
r
¼ꢀ0.258). The EPR spectrum of the reaction mixtures showed the presence of the organic radical
1H-2-phenyl-3-oxyl-4-oxoquinoline and superoxide ion due to single electron transfer from the carb-
anion to dioxygen. The pathway via 1,2-dioxetane could be proved by chemiluminescence
measurements.
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
with CO release, namely, the flavonol 2,4-dioxygenase7,8 produced
by Aspergillus flavus9 and Aspergillus japonicus,7,10 the prokaryotic
The biological oxygenation of natural substances by oxygenases
is an important part of oxidative metabolic processes.1 These are
mainly catalyzed by metalloenzymes containing usually iron or
copper ion at their active sites.2 The redox behavior of these tran-
sition metal ions seems to be significant for the catalytic ability, to
the activation of the substrates or molecular oxygen.3 Among the
dioxygenases the so-called CO-releasing dioxygenases form an
extremely interesting group since during the catalytic process two
CeC bonds are cleaved with concomitant formation of carbon
monoxide.4 It is known that mammalian microsomal heme mon-
ooxygenase system forms CO and biliverdin IXa from iron pro-
toporphyrin IX (1) in the presence of dioxygen and NADPH.5,6
Furthermore four more enzymes are known which, in an analo-
gous manner, catalyze the oxidative cleavage of two CeC bonds
aci-reductone oxidase from Klebsiella pneumoniae,11 the 1H-3-
hydroxy-4-oxoquinaldine 2,4-dioxygenase,12 and 1H-3-hydroxy-
4-oxoquinoline 2,4-dioxygenase,13 respectively. The substrates (2,
4) of these dioxygenases show common features as shown in
Scheme 1, since they possess a double bound between the C2 and
C3, a hydroxyl group at C3, and a carbonyl group at C4.
Flavonol and 1H-3-hydroxy-4-oxoquinaldine 2,4-dioxygenase
cleave the C2eC3 bond in 2, 2a, and 5a with the incorporation of
both O-atoms of O2 into the substrates with release of CO (Scheme
2). The bacterial CO-forming dioxygenases that catalyze the N-
heterocyclic ring scission reactions in the metabolism of quinaldine
and 1H-4 oxoquinoline are small monomeric proteins, which cleave
their substrates with the help of dioxygen. Flavonol 2,4-
dioxygenase (quercetinase) from A. japonicus was reported to be
a homodimer containing one atom of CuII per monomer unit.10
The structure was recently determined by the X-ray measure-
ment of quercetinase isolated from A. japonicus and it was proposed
that the copper ion activates flavonol for dioxygen attack by for-
* Corresponding author. Fax: þ36 88 624469; e-mail addresses: speier@
almos.vein.hu, speier@almos.uni-pannon.hu (G. Speier).
y
Current address: Loker Hydrocarbon Research Institute and Department of
mation of
a
CuIIesubstrate complex.14 The base-catalyzed
Chemistry, University of Southern California, University Park Campus, Los Angeles,
CA 90089-1661, USA.
autoxidation15e17 and the copper ion assisted oxygenation of
0040-4020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2013.05.117

M. Czaun et al. / Tetrahedron 69 (2013) 6666e6672
6667
derivatives in DMSO/H2O solution at a pH value of 10.4, room
temperature, and atmospheric dioxygen pressure for 6 h resulted
in the dioxygenolytic cleavage of the N-heterocycle (Eq. 1). The O2-
uptake and GCeMS analyses of the products obtained after acid-
ification with HCl and then methylation with diazomethane gave
product distribution. Benzoin was used as internal standard. The
reaction products showed a diversity of products compiled in
Table 1. From the product composition it seems clear that their
formation can be deduced from endoperoxide and also from 1,2-
dioxetane intermediates. The formation of 1,2-dioxetane could
be proved by measurement of the chemiluminescence of the re-
action. 1,2-Dioxetanes readily decompose to excited carbonyl
compounds, which then emit light and give ground state car-
bonyls.27e29 Measuring the emitted light during the reaction in
DMSOewater at 20 ꢄC shows that in the first phase of the reaction
at optimal concentration there is a maximum of emittance as can
be seen in Fig. 1. It may suggest that the 1,2-dioxetane pathway is
more dominant in the first part of the reaction and later slowly
diminishes. Earlier similar work on flavonols did not give splitting
products due to 1,2-dioxetane formation. However, during model
studies on the oxygenation of copper(II) flavonolate complexes in
one case ketocarboxylatocopper complexes could be isolated and
characterized, without CO-loss, which hinted to the formation of
1,2-dioxetane.30 The data show also that the substituents in the 40
position influence only slightly the ratio of the two reaction
pathways. The solvents have a profound effect on the ratio of the
two pathways as has been shown earlier.26
O
5'
B
OH
4'
6'
1'
1
8
N
N
N
N
HO
9 O
7
3'
OH
2
Fe
2'
OH
A
C
6
3
10
4
5
OH
2
O
1
H
N
1
4
8
O
O
H
2
R
9
7
6
S
3
OH
OH
10
5
3
4
O
Scheme 1. Substrates for CO-releasing enzymes.
Scheme 2. Flavonol and 1H-3-hydroxy-4-oxoquinoline 2,4-dioxygenase action.
Ph4'R1
flavonols18 were found to provide a model for the reaction of fla-
vonol 2,4-dioxygenase. In contrast to quercetinase, where CuII may
act as a base or as an electron acceptor due to valence isomerism in
copper flavonolates,19 both 1H-3-hydroxy-4-oxoquinoline and 1H-
3-hydroxy-4-oxoquinaldine 2,4-dioxygenase do not contain any
metal or other cofactor.20 An amino acid residue may act as a base
catalyst, however the active site residue(s) proposed to activate the
N-heterocyclic substrates for O2 attack remain to be identified.
Recent findings on studies of these enzymes and their models with
the substrates 2e4 suggest that deprotonation leads to the corre-
sponding carbanions first and then direct electron transfer to O2
resulting in a radical radical pair.21 This can only happen if the
electron affinity of the anions is less than 20 kcal/mol22 and the
formed substrate radical is stabilized.23 Model reactions revealed
that the flavonoxyl radical is less persistent17 than the one from 2-
N
O
6
O
H
N
Ph4'R1
O
R1
CO2H
7
O
H
N
O2
OH
(1)
H
N
Ph4'R1
OH
O
O
CO2H
NH2
5a; R1 = H
8
phenyl-3-hydroxy-4(1H)-oxoquinoline
being
flavonol
2,4-
5b; R1 = Br
5c; R1 = OMe
5d; R1= NO2
dioxygenase a copper-containing dioxygenase. Suggestions for the
formation of the peroxy anion,24 and also a direct one step ionic
mechanism may also be possible for the reaction of carbanions with
dioxygen.25 Recent studies on the base-catalyzed oxygenation of 40-
substituted 1H-2-phenyl-3-hydroxy-4-oxoquinolines have shown
that there are two main reaction pathways (via endoperoxide or
1,2-dioxetane intermediates) and as a result of stepwise electron
transfer from the carbanions to dioxygen radicals are formed in
aprotic solvents.26 In order to elucidate the molecular mechanism
of these reactions we studied the kinetics of the base-catalyzed
oxygenation of 1H-2-phenyl-3-hydroxy-4-oxoquinolines and
looked for possible intermediates of the reaction in DMSO/H2O.
CO2H
9
4'PhCO2H
10
Table 1
Product composition of oxygenation of 1H-2-phenyl-3-hydroxy-4-oxoquinolines
(5aed)
Products/Substratea
40-H
40-Br
40-OMe
40-NO2
2. Results and discussion
Benzoic acid
Phenylglyoxylic acid
Anthranilic acid
2-Phenyl-4H-3,1-benzox-azin-4-one
N-Benzoylanthranilic acid
O2-uptake
0.2478
0.1447
0.0914
0.1501
0.2307
0.66
0.2015
0.1526
0.1419
0.1831
0.2307
0.73
0.2631
0.1024
0.2301
0.1213
0.2026
0.79
0.2547
0.3017
0.0804
0.1422
0.1025
0.65
2.1. Oxygenation of 40-substituted 1H-2-phenyl-3-hydroxy-4-
oxoquinolines (5aed)
The oxygenation of 1H-2-phenyl-3-hydroxy-4-oxoquinoline
a
and its 40-bromo-, 40-methoxy, and 40-nitro-substituted
The products formed are given in mmols in relation to 1 mmol substrate.

6668
M. Czaun et al. / Tetrahedron 69 (2013) 6666e6672
possible. In order to check the presence of radicals and to estimate
the extent of the radical pathway of the overall reaction inhibition,
experiments were carried out by the use of excess 2,6-di-tert-butyl-
4-methylphenol and ascorbic acid. Kinetic runs with these in-
hibitors showed no significant decrease in the reaction rate; the
ꢀ
rate constant kOH at 303.0 K and a pH of 10.6 were found to be
0.50ꢃ103 and 0.35ꢃ103 molꢀ2 L2 sꢀ1 ðkOH ¼ 0:65 ꢃ 103
ꢀ
molꢀ2L2sꢀ1without inhibitorÞ: The tetrazolium blue test for the
superoxide ion was slightly positive.33 These findings suggest
a radical reaction pathway, however, due to the probably greater
stability of the 1H-2-phenyl-3-oxyl-4-oxoquinoline compared to
that of 2,6-di-tert-butyl-4-methyl-phenoxyl and ascorbic acid
radical no inhibition could be observed. Unfortunately this pre-
vented the determination of the possible ratio of non-radical
pathway of the reaction.
Fig. 1. Time dependence of chemiluminescence of PhquinH2 in alkaline DMSO/H2O
2.3. Electrochemistry
under O2.
The electrochemistry of the 40-substituted 1H-2-phenyl-3-
hydroxy-4-oxoquinolines (5aed) was carried out under an argon
atmosphere and the following conditions: [40R1PhquinH2]¼
7.00ꢃ10ꢀ3 molꢀ1, [ButOK]¼1.40ꢃ10ꢀ2 mol Lꢀ1, [NBu4ClO4]¼
0.1 mol Lꢀ1, scan rate¼20 mV/s, with a Pt working electrode in DMF
solution at room temperature, NBu4ClO4 as supporting electrolyte.
The potential values are relative to NHE using an Ag/AgCl reference
electrode and corrections taken into account as given in the liter-
ature.34 A typical CV of the deprotonated 5a can be seen in Fig. 3.
The other radicals derived from 1H-2-phenyl-3-hydroxy-4-
oxoquinolines (5bed) resulted in similar spectra (SFigs. 1e3). The
corresponding anodic oxidation and cathodic reduction potentials
are compiled in Table 2. From the shape of the first redox cycle of
the spectra it seems clearly that the one-electron oxidation/re-
The elimination of carbon monoxide during the reactions could
be detected and quantified by GLCeMS and volumetric measure-
ments. The amount of CO formed was also a measure for the en-
doperoxide pathway at the oxygenation reactions.
2.2. EPR measurements and inhibition experiments
It has been shown earlier that the oxygenation of flavonolates18
leads to radical species. The base-catalyzed oxygenation of 1H-2-
phenyl-3-hydroxy-4-oxoquinolines leads also to radical prod-
ucts.26 In the case of flavonols the flavonoxyl radical could only be
detected if the reaction was carried out in aprotic solvent.31 In
DMSO/H2O mixtures no radical species could be detected by EPR.32
At the oxygenation of 1H-2-phenyl-3-hydroxy-4-oxoquinolines in
aprotic solvents, such as DMF, MeCN, THF, and DMSO the presence
of persistent 1H-2-phenyl-3-oxyl-4-oxoquinoline radicals could be
shown.26 In DMSO/H2O mixture at various pH values the corre-
sponding 1H-2-phenyl-3-oxyl-4-oxoquinoline radical gave a well
resolved EPR spectrum with hyperfine structure, as shown in Fig. 2,
duction
of
the
deprotonated 1H-2-phenyl-3-hydroxy-4-
oxoquinoline exhibits good reversibility or quasi-reversibility. A
comparison with the CV spectra of the O-analogs (flavonols)17,35
show that the radicals formed from the 1H-2-phenyl-3-hydroxy-
4-oxoquinolines are much more persistent than those formed from
flavonolates.
0
with the parameters g¼2.0036, aN¼1.03, aH¼6.05, aH ¼1.15,
00
aH ¼0.9.
Fig. 2. The EPR spectrum taken during the base-catalyzed oxygenation of PhquinH2 at
room temperature in alkaline DMSO/H2O.
Fig. 3. The cyclic voltammogram of deprotonated 1H-2-phenyl-3-hydroxy-4-
oxoquinoline (5a) in DMF under argon atmosphere.
However, the concentration of the 1H-2-phenyl-3-oxyl-4-
oxoquinoline radical formed in solution during the oxygenation
process is rather low, and in the UVevis spectrum no absorption at
532 nm26 attributable to it could be seen. From that we conclude
that the formation of the 1H-2-phenyl-3-oxyl-4-oxoquinoline
radical during the oxygenation of the deprotonated 1H-2-phenyl-
3-hydroxy-4-oxoquinoline indicates that a radical reaction path-
way occurs in the reaction, but that a non-radical pathway is also
Table 2
Redox potentials of the deprotonated 40R1PhquinH2 (5aed) in DMF
R1
Eox1 (mV)
Ered1 (mV)
Eox2 (mV)
H
MeO
Br
140.8
110.2
182.0
219.0
82.0
41.7
89.0
556.0
566.0
551.0
d
NO2
133.0

M. Czaun et al. / Tetrahedron 69 (2013) 6666e6672
6669
This is also supported by the good quality of the EPR spectra
with hyperfine structures and the persistence of the radicals for
days under the conditions.
2.3.1. Determination of the pK values of 40R1PhquinH2 (5aed). The
determination of the dissociation constants of 40R1PhquinH2 in 50/
50% DMSO/H2O at 25 ꢄC and I¼0.2 mol Lꢀ1 (KNO3) was carried out
with a Radometer pHM93 instrument and a Metrohm 60222 com-
bination electrode. The electrode was conditioned for 3 days in the
same solvent before the measurements were made. The titrationwas
done with a 0.2 mol Lꢀ1 KOH solution and its concentration was de-
termined by pH-metry using the Gran relation to be 0.1978 mol Lꢀ1
.
The titrations were carried out with a Metrohm 715 Dosimat auto-
matic burette. The readings of the instrument were converted to
concentrationsbytheIrvingmethod.36 Theautoprotolysisconstantof
water (pK) was found to be 15.412 under the conditions applied. The
dissociation constants of the 40RPhquinH2 derivatives were de-
termined by pH-metric method at 2 and 1 mmol Lꢀ1 analytical con-
centrations. The results were evaluated by the program
SUPERQUAD.37 The obtained pK values are shown in Table 3.
Fig. 4. Change in the electronic absorption spectra of PhQuinH2 in the presence of base
under dioxygen. [PhquinH2]¼2.00ꢃ10ꢀ3 mol Lꢀ1, [O2]¼2.69ꢃ10ꢀ3 mol Lꢀ1, pH¼10.4,
[Buffer]¼2.39ꢃ10ꢀ2 mol Lꢀ1, I¼0.1 mol Lꢀ1, DMSO/H2O, 20 ꢄC.
This means that the reaction is specific base-catalyzed and the rate
expression (3) describes the reaction velocities (STable 1).
Table 3
h
ih
i
The measured pK values 40R1PhquinH2 (5aed) in
1
reaction rate ¼ kOH OHꢀ 4’R PhquinH2 ½O2ꢁ
(3)
ꢀ
DMSO/H2O (50/50) at 25 ꢄC and I¼0.2 mLꢀ1 (KNO3)
R1
ꢀ
pK
The rate constant kOH at 303.16 K was found to be (2.42ꢂ0.03)ꢃ
103 molꢀ2. The influence of 40-substituted groups on the reaction
rate of the oxygenation showed a linear Hammett plot with a re-
H
MeO
Br
10.43 (1)
10.63 (1)
10.93 (4)
9.47 (1)
action constant of
r
¼ꢀ0.258 (Fig. 5, STable 2), indicating that
NO2
electron-releasing groups result in enhanced reaction rates. We
also investigated the dependence of the reaction rates on the first
anodic oxidation potential of the deprotonated substrates 5aed.
Fig. 6 shows that substrates with electron-releasing groups have
2.4. Kinetic measurements
The kinetic measurements on the oxygenation of 40-substituted
1H-2-phenyl-3-hydroxy-4-oxoquinolines in
(40R1PhquinH2)
DMSO/H2O were followed by UVevis spectroscopy under pseudo-
first-order conditions. The ionic strength I¼0.1 mol Lꢀ1 (KNO3),
the pH, the dioxygen concentration, and the temperature were kept
constant during the experiments. The rate constants kobs according
to Eq. 2 were determined.
h
i
reaction rate ¼ kobs 4’R1PhquinH2
(2)
Under these conditions a typical variation of the electron
spectrum in the range 290e500 nm as a function of time can be
seen in Fig. 4. During the reaction the band at 363 nm decreases
while that at 305 nm increases and there is an isosbestic point at
320 nm. The reactions were carried out at higher pH values too.
Without added base no oxygenation reaction occurs at all. The time
profile of the oxygenation reaction is shown in SFig. 4 and the log
[PhquinH2] versus time diagram (SFig. 5) resulted in a straight line,
which is characteristic for first-order dependence in the substrate.
This could be also supported by measuring the initial reaction rates
of the oxygenation reaction with varying initial substrate concen-
trations of PhquinH2. Plots of the reaction rate versus the initial
PhquinH2 concentration resulted also in a straight line (SFig. 6)
reinforcing the first-order dependence in PhquinH2. The rate de-
pendence on the dioxygen concentration was determined by
measuring the pseudo-first-order rate constants kobs at different
dioxygen concentrations. A straight line of the reaction rate versus
dioxygen concentration (SFig. 7) shows that the reaction has a first-
order dependence on the dioxygen concentration. The effect of the
pH on the reaction was evaluated by carrying out the reactions at
various pH under pseudo-first-order conditions. The resulted
straight line indicated a first-order dependence (SFig. 8), while
carrying out the measurements at one pH value and different buffer
concentrations (SFig. 9) showed no difference in the reaction rates.
Fig. 5. Hammett plot for the oxygenation of 40R1PhquinH2. [40R1quinH2]¼
2.00ꢃ10ꢀ3 mol Lꢀ1, [O2]¼2.16ꢃ10ꢀ3 mol Lꢀ1, pH¼10.0, I¼0.1 mol Lꢀ1, DMSO/H2O (50/
50), 20 ꢄC.
Fig. 6. The dependence of the reaction rate on the first oxidation potentials of the
substrates 40R1PhquinH2.

6670
M. Czaun et al. / Tetrahedron 69 (2013) 6666e6672
a more negative anodic oxidation potential and also a higher re-
action rate. The linear correlation between the anodic oxidation
potential and the reaction rate supports that a higher electron
density on the deprotonated substrates 5aed makes the electron
transfer easier or enhances the nucleophilic character of the carb-
anions in their reaction with dioxygen. The kinetic and the EPR data
on the oxygenation reactions of the carbanions of the substrates
5aed show similar features to flavonolates,17 phenolates,38 and
carbanions.39
that the reaction proceeds by route c and/or by the SET reaction as
indicated in route b, while route a can be dismissed. The quanti-
tative measure on the concentration of the radical 13 by UVevis
spectroscopy showed only a small concentration during the re-
action. While this does not account for the less favored pathway b,
the relative stability of the radical 13 and the slow reaction between
13 and the superoxide anion hints to a smaller extent of this route.
According to Scheme 3 the rate Eqs. 4 and 5 can be deduced, where
ꢀ
the kOH may be expressed as shown in Eq. 6.
The reduction power of the carbanions formed from 5aed is
high and a direct one electron transfer to dioxygen gives radicals,
from which the radicals formed from 5aed exhibit relative good
stability (like the phenoxyl radicals in general),38 while those
formed from carbanions and flavonolates are much less stable. Since
the first electron reduction of dioxygen to superoxide ion is un-
favorable (Eꢄ¼ꢀ0.62 V vs NHE in DMF),40 the reduction potential of
phenolates,41 carbanions,42 flavonolates,17 and the carbanions from
5aed are more negative than that of O2, and so the electron transfer
is facilitated. The stability of the radicals formed is dependent on the
delocalization of the unpaired electron. According to that, the sta-
bility of the flavonoxy radicals is rather low while the phenoxy and
the 40R1PhquinHꢅ radicals are more stable. The stability depends of
course on the solvent used too. While water is considered in many
cases as a good quencher for radicals, the PhquinHꢅ radical shows
relatively good stability and persistence, and an EPR spectrum with
hyperfine structure as shown in Fig. 1 could be recorded. A detailed
study of the EPR spectra of the 40R1PhquinHꢅ radicals in aprotic
solvents will be shown in a forthcoming paper.
k4K1K2K300
½H2Oꢁ
.
ꢀ
ꢁ
ꢀ
ꢁꢀ
ꢁꢀ
ꢁ
ꢀd 40R1PhquinH2 dt ¼
OHꢀ R1PhquinH2 O2
(4)
k000 K1K2
½H2Oꢁ
ꢀ
ꢁ
ꢀ
ꢁꢀ
ꢁꢀ
ꢁ
ꢀd 40R1PhquinH2 =dt ¼
OHꢀ 40R1PhquinH2 O2
3
(5)
(6)
k4K1K2K300
k0300K1K2
kOH
¼
¼
½H2Oꢁ
½H2Oꢁ
The third-order rate expression (3) and the non-dependence of
the reaction rate on the buffer concentration shows that the re-
action is specific base-catalyzed and the reaction rate is dependent
on the anions 40R1PhquinHꢀ and dioxygen. The possible reaction
pathways, which satisfy the kinetic data are summarized in Scheme
3. In the first reversible step, the substrates 5aed are deprotonated
to the O-anions 11, which show valence isomerism to give the
carbanions 12. These steps are fast preequilibria and the concen-
tration of 11 and 12 together can be calculated for various pH values
from the corresponding pK values of 5aed determined potentio-
metrically as listed in Table 3. There are more routes from that
point, which satisfy the kinetic data. Route a concerns a possible
rate-limitingꢀ step as
a
single electron transfer between
40R1PhquinH and dioxygen leading to 40R1PhquinHꢅ and super-
oxide ion. We believe however that this route can be ignored, be-
cause this step cannot be rate determining since electron transfers
from electron-rich substrates to dioxygen are usually fast re-
actions.35 The fast reversible reaction of 12 with dioxygen (route b)
can lead to the radicals 13 and superoxide anion, and their reaction
in a rate-determining step (k4) results in the peroxidic compounds
14. These steps are also evidenced by detecting the radical 13a by
EPR spectroscopy and proving the presence of the superoxide ion
by the tetrazolium blue method.33 However, the concentration of
the radicals 13 and also that of superoxide anion were rather low.
The direct reaction of singlet carbanion with triplet dioxygen re-
quires a change in multiplicity.17 It occurs after electron transfer
from the anion to O2 since a change in electron spin occurs readily
in O2$ꢀ: Carbanions of radicals with electron affinities¼20 kcal/mol
react with ground state dioxygen while those of higher ones do not
because it becomes unfavorable.24 The peroxidic species 14 can also
be formed from 12 by route c, where a direct electrophilic attack of
dioxygen on the carbanion 12 is rate-limiting. Whether the carb-
anion reacts with 3O2 at all depends on the energy level of the
HOMO orbital (or redox potential) of the anion 40R1PhquinHꢀ.
Whether this reaction proceeds via a radical pathway or through
ꢀ
a single (concerted) step to the adduct 40R1PhquinHO2 depends
$
mainly on the stability of the radicals 40R1PhquinHO2 and the
reaction conditions either supporting its stability or not. We believe
Scheme 3. The possible mechanisms of the base-catalyzed oxidation of 40R1PhquinH2.

M. Czaun et al. / Tetrahedron 69 (2013) 6666e6672
6671
The consecutive fast reaction steps of the peroxidic species 14
4. Experimental section
may follow two alternatives. In the one, due to intramolecular
nucleophilic attack on the C4 carbon, the endoperoxides 15 are
formed, which after simultaneous cleavage of the OeO bond and
concomitant elimination of CO result in the compounds 8, 9, and
benzoic acid.
That is the enzymatic pathway. The peroxidic species 14 may
also undergo an intramolecular nucleophilic attack on the C3 car-
bon, leading to a 1,2-dioxetane species, which after cleavage of the
OeO bond gives the compounds 7, 6, and phenylglyoxalic acid
without loss of CO. This represents a non-enzymatic pathway. All
attempts to isolate the endoperoxide 15 and 16 failed suggesting
that this type of endoperoxides is very unstable if compared with
other endoperoxides where CO elimination is not feasible.43,44 Until
now no endoperoxides of this general structure (18) could be
isolated.
4.1. General: materials and methods
All manipulations were carried out under pure argon or dini-
trogen atmospheres unless otherwise stated using standard
Schlenck-type inert gas techniques.48 Solvents used for the re-
actions were purified by literature methods49 and stored under
argon. 1H-2-Phenyl-3-hydroxy-4-oxoquinoline, 40-bromo-, 40-
methoxy-, and 40-nitro-1H-2-phenyl-3-hydroxy-4-oxoquinolines
were prepared according to literature methods.50 Diazomethane
was freshly prepared by a known procedure in ether and imme-
diately used for the methylation reactions.51 For the preparation of
the buffer for pH¼9.6 2.22 g KNO3 was dissolved in a mixture of
50 mL 0.1 mol Lꢀ1 sodium hydrogen carbonate and 5 mL 0.2 mol Lꢀ1
sodium hydroxide then the solution was diluted to 100 mL.52 The
UVevis spectra were recorded on a Shimadzu UV-160A and Agilent
8453 diode-array UVevis spectrophotometer and the IR spectra on
an Avatar 330 FT-IR Thermo Nicolet instrument and a Specord M80
(Carl Zeiss, Jena) infrared spectrophotometer. The EPR spectra were
taken on a JEOL JES- FE/3X EPR spectrometer of X band (9.64 GHz,
200
mW) at room temperature and in DMSOþwater mixture. The
product analyses were carried out on a Hewlett Packard 4890D gas
chromatograph. The mass spectra were recorded on a Hewlett
Packard 5890 II, 5971 GC/MSD (75 eV) mass spectrometer. The
chemiluminescent measurements were carried out on a TECAN
Infinite M200 instrument, microplate 96 well: in each well samples
added with automatic injection. The dioxygen concentration was
determined by the use of a Beckman Fieldlab Oxygen Analyzer. The
sensor was calibrated with dioxygen saturated DMF.32 For the po-
tentiometric titrations a Metrohm 715 Dosimat automatic burette
was used. Cyclic voltammograms were taken on a VoltaLab 10
potentiostat with Voltamaster 4 software for data processing. A
platinum working electrode, auxiliary electrode, and Ag/AgCl with
3 M KCl reference have been used. The potentials were referenced
versus the ferrocene/ferrocenium (Fc/Fcþ) redox couple. The CV
spectra were measured in argon-saturated DMF using 0.1 M tetra-
butylammonium perchlorate (TBAP) as electrolyte. Elemental
analyses were carried out at the University’s analytical unit.
1,2-Dioxetanes have been isolated and characterized, however
only with bulky groups on the carbon atoms.45,46 Only indirect
evidence for its presence could be found, e.g., by the photochemical
oxidation of quercetin and its derivatives47,48 where only the
cleavage products could be isolated. From the temperature de-
pendence of the reaction rates (SFig. 10) the activation parameters
at 303.16
K
D
as
D
Gz¼73.13ꢂ4.02 kJ molꢀ1
;
D
Hz¼70.60ꢂ4.04
kJ molꢀ1
;
Sz¼ꢀ28ꢂ2
J
molꢀ1 Kꢀ1
,
and EA¼73.23 ꢂ4.02
kJ molꢀ1could be calculated. These data also support, the negative
entropy of activation, a bimolecular step as the rate-determining
reaction in routes b or c.
3. Conclusion
From the results obtained the conclusion may be drawn that
the oxygenolysis of the 40-substituted 1H-2-phenyl-3-hydroxy-4-
oxoquinolines is a specific base-catalyzed reaction and the ki-
netic data revealed a first-order dependence with respect to the
1H-2-phenyl-3-hydroxy-4-oxoquinolines and dioxygen. The pK
values of the 1H-2-phenyl-3-hydroxy-4-oxoquinolines were de-
termined by potentiometric titrations. The product distribution
of the oxygenation reactions indicates endoperoxides and 1,2-
dioxetanes as intermediates. For the reaction mechanism two
routes are proposed. One (route b) involves an SET reaction from
the carbanions 12 to the dioxygen proceeding in a fast reversible
reaction yielding the 1H-2-phenyl-3-oxyl-4-oxoquinoline radi-
cals 13 and superoxide anion, which react in the rate-
determining step to the peroxidic species 9. These than un-
dergo an intramolecular nucleophilic attack either on the C3 or
C4 to 1,2-dioxetanes or endoperoxides, which break down to the
products in fast reaction steps. The other route may involve
a direct electrophilic attack of O2 on the carbaꢀnions 12 in the
rate-determining step leading to 40R1PhquinHO2 (9), which are
followed by fast consecutive reactions to the products. Route
b has been evidenced by the presence of the radical PhquinHꢅ
detected by EPR. Route a can be excluded due to data found in
the literature for fast electron-transfer reactions from carbanions
to dioxygen, while route c is possible and seemed to be the
dominant pathway in the case of flavonols in protic solvents. The
better stability of the radicals 40R1PhquinHꢅ may render these
reactions dualistic pathways.
4.2. The bulk oxygenation of 1H-2-phenyl-3-hydroxy-4-
oxoquinoline (PhquinH2)
PhquinH2 (237.25 mg, 1 mmol) in a DMSO/H2O mixture (10 mL),
at 25 ꢄC, 1 bar O2 pressure were oxygenated for 6 h in a thermo-
stated reaction vessel attached to a buret. The evolved CO was
determined by GCeMS from the gas phase and was found to be
0.86 mmol (86%). The reaction mixture was acidified and treated
with diazomethane and then analyzed by GC. Gas chromatogram
and the conditions with the results are shown in SFig. 11.
4.3. Kinetic measurements
The reaction of 1H-2-phenyl-3-hydroxy-4-oxoquinoline
(PhquinH2) with O2 was performed in DMSO/H2O (50/50) solutions.
In a typical experiment PhquinH2 was dissolved in DMSO under an
argon atmosphere in a thermostated reaction vessel and a buffer
solution was added to adjust the pH value. The reaction vessel was
equipped with an inlet for sampling with a syringe and connected to
a mercury manometer to maintain constant pressure. The solution
was then heated to the appropriate temperature. A sample was
taken by syringe and the initial concentration of PhquinH2 was de-
termined by UVevis spectroscopy measuring the absorbance of the
reaction mixture at 363 nm (lmax¼363 nm,
3
¼4885 molꢀ1 L cmꢀ1 of
PhquinH2). The argon was then replaced by dioxygen and the con-
sumption of PhquinH2 was analyzed periodically (ca. every 5 min).

6672
M. Czaun et al. / Tetrahedron 69 (2013) 6666e6672
ꢀ
18. Balogh-Hergovich, E.; Speier, G.; Argay, G. J. Chem. Soc., Chem. Commun. 1991,
551e552.
The rate of consumption of PhquinH2 was independent of the stir-
ring rate, excluding eventual diffusion control effects. Experimental
conditions are summarized in STable 1. The temperature was de-
termined with an accuracy of ꢂ0.5 ꢄC; the concentrations of
PhquinH2 were measured with a relative error of ca. ꢂ2%; the
pressure of the dioxygenwas determined with an accuracy of ꢂ0.5%.
ꢀ
19. Balogh-Hergovich, E.; Kaizer, J.; Speier, G.; Huttner, G.; Jacobi, A. Inorg. Chem.
2000, 39, 4224e4229.
20. Steiner, R. A.; Janssen, H. J.; Roversi, P.; Oakley, A. J.; Fetzner, S. PNAS 2010, 107,
657e662.
21. Frerichs-Deeken, U.; Ranguelova, K.; Kappl, R.; Huettermann, J.; Fetzner, S.
Biochemistry 2004, 14485e14499.
22. Fetzner, S.; Steiner, R. A. Appl. Microbiol. Biotechnol. 2010, 86, 791e804.
23. Fetzner, S. Nat. Chem. Biol. 2007, 3, 374e375.
24. Schmitt, R. J.; Bierbaum, M.; DePuy, C. H. J. Am. Chem. Soc. 1979, 101,
6443e6445.
Acknowledgements
We thank the Hungarian National Fund K67871 and K75783 and
COST Actions CM0905, CM1201, and CM1003 for financial support,
and Drs A. Rockenbauer and L. Korecz (Budapest) for the EPR
measurements.
25. Massey, V. J. Biol. Chem. 1994, 269, 22459e22462.
26. Czaun, M.; Speier, G. Tetrahedron Lett. 2002, 43, 5961e5963.
27. Gundermann, K.-D.; McCapra, F. Chemiluminescence in Organic Chemistry;
Springer: Berlin, 198737.
28. Kopecky, K. R.; Mumford, C. Can. J. Chem. 1969, 47, 709e711.
29. Kopecky, K. R.; Filby, J. E.; Mumford, C.; Lockwood, P. A.; Ding, J.-Y. Can. J. Chem.
1975, 53, 1103e1122.
Supplementary data
ꢀ
30. Balogh-Hergovich, E.; Kaizer, J.; Pap, J.; Speier, G.; Huttner, G.; Zsolnai, L. Eur. J.
Inorg. Chem. 2002, 2287e2295.
31. Kaizer, J.; Ganszky, I.; Csurka, K.; Speier, G.; Korecz, L.; Rockenbauer, A.; Giorgi,
Kinetic data and graphs, CV spectra, chemiluminescence mea-
surements, and product identification are described. Supplemen-
tary data related to this article can be found at http://dx.doi.org/
10.1016/j.tet.2013.05.117.
ꢀ
M.; Reglier, M.; Antonczak, S. J. Inorg. Biochem. 2007, 101, 893e899.
ꢀ
32. Balogh-Hergovich, E.; Speier, G. J. Org. Chem. 2001, 66, 7974e7978.
33. Greenwald, R. A. CRC Handbook of Methods for Oxygen Radical Research; CRC:
Boca Raton, 1984.
34. Lever, A. B. P.; Milaeva, E.; Speier, G. In The Redox Chemistry of Metal-
lophthalocyanines; Leznoff, C. C., Lever, A. B. P., Eds.; Phthalocyanines; VCH:
New York, NY, 1993; Vol. 3, pp 1e70.
References and notes
ꢀ
35. Pap, J. S.; Matuz, A.; Barath, G.; Kripli, B.; Giorgi, M.; Speier, G.; Kaizer, J. J. Inorg.
1. Hayaishi, O. Molecular Mechanism of Oxygen Activation; Academic: New
York-London, NY, 1974.
Biochem. 2012, 108, 15e21.
36. Irving, H. M.; Miles, M. G.; Pettit, L. D. Anal. Chim. Acta 1967, 38, 475e488.
37. Gans, P.; Sabatini, A.; Vacca, A. J. Chem. Soc., Dalton Trans. 1985, 1195e1200.
38. March, J. Advanced Organic Chemistry, 4th ed.; Wiley-Interscience: New York,
NY, 1992.
2. Fox, B. G.; Lipscomb, J. D. In Biological Oxidation Systems; Reddy, C., Hamilton, G.
A., Eds.; Academic: New York, NY, 1990; Vol. 1, pp 367e388.
3. Funabiki, T. In Dioxygenases in Catalysis by Metal Complexes, Oxygenases and
Model Systems; Funabiki, T., Ed.; Kluwer Academic: Dordrecht, The Netherlands,
1997; pp 19e89.
39. Bordwell, F. G.; Bausch, M. J. J. Am. Chem. Soc. 1986, 108, 1985e1988.
40. Sawyer, D. T. Oxygen Chemistry; Oxford University: Oxford, 1991.
ꢁ
ꢁ
ꢀ
ꢀ
€
4. Fetzner, S.; Tshisuaka, B.; Lingens, F.; Krappl, R.; Huttermann, J. Angew. Chem.,
41. Mihalovic, M. L.; Cerkovic, Z. In The Chemistry of the Hydroxyl Group; Patai, S.,
Ed.; Wiley: New York, NY, 1971, Part 1.
Int. Ed. 1998, 37, 576e597.
5. Tenhunen, R.; Marver, H. S.; Schmid, R. J. Biol. Chem. 1969, 244, 6388e6394.
6. Maines, M. D. FASEB J. 1988, 2, 2557e2568.
42. Eberson, L. Electron-transfer Reactions in Organic Chemistry; Spinger: Berlin,
1987.
7. Tranchimand, S.; Brouant, P.; Iacazio, G. Biodegradation 2010, 21, 833e859.
8. Fetzner, S. Appl. Environ. Microbiol. 2012, 78, 2505e2514.
9. Oka, T.; Simpson, F. J. Biochem. Biophys. Res. Commun. 1971, 43, 1e5.
43. Konishi, T.; Fujitsuka, M.; Ito, O. Chem. Lett. 2000, 2, 202e203.
44. Clennan, E. L.; Foote, C. S. In Organic Peroxides; Ando, W., Ed.; Wiley: Chichester,
UK, 1992; pp 225e259.
45. Matsuura, T.; Matsushima, H.; Sakamoto, H. J. Am. Chem. Soc. 1967, 89,
6370e6371.
46. Matsuura, T.; Matsushima, H.; Nakashima, R. Tetrahedron 1970, 26, 435e443.
47. Bronstein, J.; Edwards, B.; Voyta, J. C. J. Biolumin. Chemilumin. 1989, 4, 99e111.
48. Shriver, D. F.; Drezdzon, M. A. The Manipulation of Air-sensitive Compounds; John
Wiley & Sons: New York, NY, 1990.
49. Perrin, D. D.; Armarego, W. L.; Perrin, D. R. Purification of Laboratory Chemicals,
2nd ed.; Pergamon: New York, NY, 1990.
€
10. Fusetti, F.; Schroter, K. H.; Steiner, R. A.; van Noort, P. I.; Pijning, T.; Rozeboom,
H. J.; Kalk, K. H.; Egmond, M. R.; Dijkstra, B. W. Structure 2002, 10, 259e268.
11. Wray, J. W.; Abeles, R. H. J. Biol. Chem. 1993, 268, 21466e21469.
12. Hund, H. K.; de Beyer, A.; Lingens, F. Biol. Chem. Hoppe-Seyler 1990, 371,
1005e1008.
13. Bott, G.; Schmidt, M.; Rommel, T. O.; Lingens, F. Biol. Chem. Hoppe-Seyler 1990,
371, 999e1003.
ꢀ
14. Speier, G. In Dioxygen Activation and Homogeneous Catalytic Oxidation; Simandi,
50. Hradil, P.; Jirman, J. Collect. Czech. Chem. Commun. 1995, 60, 1357e1366.
51. Arndt, F. In Org. Synth.; Blatt, A. H., Ed.; John Wiley & Sons: New York, NY, 1943;
Vol. 2, pp 165e167.
52. Dean, J. A. Lange’s Handbook of Chemistry; McGraw-Hill Book: New York, NY,
1973; pp 74e77; Section 5.
L. I., Ed.; Elsevier: Amsterdam, 1991; pp 269e278.
15. Rajanda, V.; Brown, S. B. Tetrahedron Lett. 1981, 22, 4331e4334.
16. Nishinaga, A.; Tojo, T.; Tomita, H.; Matsuura, T. J. Chem. Soc., Perkin Trans. 1 1979,
2511e2516.
ꢀ
17. Barhacs, L.; Kaizer, J.; Speier, G. J. Org. Chem. 2000, 65, 3449e3452.