Oxidation of polycyclic aromatic hydrocarbons catalyzed by iron tetrasulfophthalocyanine FePcS: Inverse isotope effects and oxygen labeling studies

Article abstract of DOI:10.1002/(sici)1099-0682(199809)1998:9<1269::aid-ejic1269>3.0.co;2-g

Iron(III) tetrasulfophthalocyanine (FePcS) was shown to catalyze the oxidation of polycyclic aromatic hydrocarbons by H₂O₂. Benzo[a]pyrene and anthracene were converted to the corresponding quinones while biphenyl-2,2′-dicarboxylic acid was the main product of phenanthrene oxidation. The mechanism of the anthracene oxidation by H₂O₂ in the presence of FePcS or by KHSO₅ with iron(III) mesotetrakis(3,5-disulfonatomesityl)porphyrin (FeTMPS) (see Figure 1 for catalyst structures) has been investigated in details by using kinetic isotope effects (KIEs) and ¹⁸O labeling studies. KIEs measured on the substrate consumption in the competitive oxidation of [H₁₀] anthracene and [D₁₀]anthracene by FePcS/H₂O₂ and FeTMPS/KHSO₅ were essentially the same, 0.75 ± 0.02 and 0.76 ± 0.06, respectively. These inverse KIEs on the first oxidation step can be explained by the sp²-to-sp³ hybridization change during the addition of an electrophilic oxoiron complex to the sp² carbon center of anthracene to form a σ adduct (this inverse KIE being enhanced by stronger slacking interactions between the perdeuterated substrate with the macrocyclic catalyst). Although the first oxidation step seems to be the same, different distribution of the oxidation products of anthracene and very different ¹⁸O incorporation into anthrone and anthraquinone in catalytic oxidations performed in the presence of H₂¹⁸O suggested that different active species should be responsible for anthracene oxidation in both catalytic systems. All the results obtained are compatible with an involvement of TMPSFe^V=O (or TMPS⁺Fe^IV=O), having two redox equivalents above the iron(III) state of the metalloporphyrin precursor, while PcSFe^IV=O (one redox equivalent above Fe^III state of FePcS) was proposed to be the active species in the metallophthalocyanine-based system.

Full text of DOI:10.1002/(sici)1099-0682(199809)1998:9<1269::aid-ejic1269>3.0.co;2-g

FULL PAPER
Oxidation of Polycyclic Aromatic Hydrocarbons Catalyzed by Iron
Tetrasulfophthalocyanine FePcS: Inverse Isotope Effects and Oxygen
Labeling Studies
Alexander Sorokin and Bernard Meunier*
Laboratoire de Chimie de Coordination du CNRS,
205 route de Narbonne, F-31077 Toulouse cedex 4, France
Fax: (internat.) ϩ 33-5/61553003
E-mail: bmeunier@lcc-toulouse.fr
Received April 24, 1998
Keywords: Oxidation / Catalysis / Phthalocyanine / Iron / Anthracene
Iron(III) tetrasulfophthalocyanine (FePcS) was shown to
catalyze the oxidation of polycyclic aromatic hydrocarbons
sp2 carbon center of anthracene to form a σ adduct (this
inverse KIE being enhanced by stronger stacking
by H2O2. Benzo[a]pyrene and anthracene were converted to interactions between the perdeuterated substrate with the
the corresponding quinones while biphenyl-2,2Ј-dicarboxylic macrocyclic catalyst). Although the first oxidation step seems
acid was the main product of phenanthrene oxidation. The
mechanism of the anthracene oxidation by H2O2 in the
presence of FePcS or by KHSO5 with iron(III) meso-
to be the same, different distribution of the oxidation
products of anthracene and very different 18O incorporation
into anthrone and anthraquinone in catalytic oxidations
tetrakis(3,5-disulfonatomesityl)porphyrin (FeTMPS) (see performed in the presence of H218O suggested that different
Figure 1 for catalyst structures) has been investigated in
active species should be responsible for anthracene oxidation
in both catalytic systems. All the results obtained are
details by using kinetic isotope effects (KIEs) and 18O
labeling studies. KIEs measured on the substrate compatible with an involvement of TMPSFeV=O (or
consumption in the competitive oxidation of [H10]anthracene
and [D10]anthracene by FePcS/H2O2 and FeTMPS/KHSO5
were essentially the same, 0.75 ± 0.02 and 0.76 ± 0.06, PcSFeIV=O (one redox equivalent above FeIII state of FePcS)
TMPS+FeIV=O), having two redox equivalents above the
iron(III) state of the metalloporphyrin precursor, while
respectively. These inverse KIEs on the first oxidation step
can be explained by the sp2-to-sp3 hybridization change
during the addition of an electrophilic oxoiron complex to the
was proposed to be the active species in the
metallophthalocyanine-based system.
Introduction
contamination process, along with photochemical oxidation
with dioxygen[6], microorganisms capable of rapidly and ef-
Polycyclic aromatic hydrocarbons (PAH) belong to en- ficiently mineralizing recalcitrant and potentially carcino-
vironmental pollutants because of their rather slow biodeg- genic PAHs needs to be found[1]. Chemical oxidation of
radation. PAHs can be metabolized by bacteria, fungi or PAH [7] needs elevated temperatures and/or powerful oxi-
algae. Three main degradation pathways have been ident- dants, e.g. chromium compounds which constitute environ-
ified for the microbial catabolism of polycyclic aromatic hy- mental problems. For example, phenanthrene has been oxi-
drocarbons[1]. Cytochrome P-450 monooxygenases catalyze dized to phenanthrenequinone by CrO3 at 55°C (yield
the formation of arene oxides, which are then transformed 80%)[8], by the system CeIV/H2SO4/(NH4)2SO4 at 50°C
to phenol or epoxy-diol derivatives. Ligninases are capable (yield 42%)[9], by KMnO4/Al2O3/H5IO6 (yield 25%)[10]. Ar-
of oxidizing PAH to quinones followed by ring cleavage[2]
.
enes can be oxidized to quinones by H2O2/hexafluoroace-
For example, benzo[a]pyrene was oxidized to a mixture of tone at 70°C[11]. In this latter case, biphenyl-2,2Ј-dicar-
1,6-, 3,6- and 6,12-quinones[3], pyrene to 1,6- and 1,8-qui- boxylic acid was the only isolated oxidation product of
nones[4] and anthracene to 9,10-anthraquinone[2]. The third phenanthrene[11]. Obviously, there is a need for environ-
enzymatic degradation process involves dioxygenase en- mental-friendly chemical catalysts to oxidize PAH. An
zymes to incorporate both atoms of molecular oxygen into alternative approach using biomimetic metalloporphyrin
the aromatic ring to form cis-dihydrodiols. After a dehydro- catalysts[12] has been shown to be useful in the oxidation of
genase-catalyzed rearomatization to produce catechols, a 2-methylnaphthalene[12a]. Phenanthrene and naphthalene
subsequent enzymatic cleavage by intradiol or extradiol di- were oxidized by a porphyrinruthenium/2,6-dichloropyri-
oxygenases gives cis,cis-muconic acid or 2-hydroxymuconic dine N-oxide system to give the corresponding quinones
semialdehyde as products of aromatic cycle cleavage, respec- with 40% and 29% yields, respectively[13]. Recently, the oxi-
tively[5]. Although biodegradation of PAHs is the major de- dation of naphthalene and methylnaphthalenes by peracetic
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ1948/98/0909Ϫ1269 $ 17.50ϩ.50/0
1269

A. Sorokin, B. Meunier
FULL PAPER
acid catalyzed by octanitrophthalocyanine MnIII and FeII
Results
complexes has been reported[14]
.
Products and Reactivity Order in PAH Oxidations by
FePcS/H2O2
We recently developed a new efficient catalytic system for
the oxidative degradation of polychlorinated phenols based
on the association of iron(III) or manganese(III) complexes
of tetrasulfophthalocyanine (FePcS, MnPcS) and the rather
inexpensive and “green” oxidant H2O2[15][16]. This biomi-
metic system is able to cleave the aromatic cycle of 2,4,6-
trichlorophenol with the formation of chloromaleic acid as
the main product[15]. Besides FePcS and MnPcS, CoPcS is
also highly efficient in the oxidation of catechols[17]. Having
on hands these efficient catalysts for the oxidation of recal-
citrant pollutants such as chlorinated phenols, we decided
to investigate the activity of these catalysts in the oxidation
of other poorly biodegradable molecules. Here we report
the catalytic oxidation of polycondensed aromatics (mainly
anthracene, but also phenanthrene and benzo[a]pyrene) by
hydrogen peroxide or potassium monopersulfate (KHSO5)
and FePcS as catalyst [FeTMPS, a water-soluble iron(III)
porphyrin, was also used for comparison]. Detailed mech-
anistic studies involving kinetic isotope effects (KIEs) and
18O labeling experiments are also reported.
We checked the oxidations of benzo[a]pyrene, phen-
anthrene and anthracene, performed in a MeCN/H2O mix-
ture (3:1, v/v) using KHSO5 and H2O2 as oxidants. The
final concentrations of oxidant and substrate were 5 m
and 1 m, respectively. The catalyst concentrations were 10
µ, 37 µ and 100 µ corresponding to 1, 3.7 and 10%
catalyst/substrate ratios, respectively. The data of Table 1
indicated that FePcS was also an efficient catalyst for the
oxidation of these polyaromatic hydrocarbons. The oxi-
dation of benzo[a]pyrene by FePcS/KHSO5 (run 1) and by
FePcS/H2O2 (run 4) at pH ϭ 7 was complete within in a
few minutes leading to an orange material isolated by pre-
cipitation with water. DCI-CH4 mass spectrometry analysis
gave peaks at 282 (Mϩ, 100), 254 [(M Ϫ CO)ϩ, 17.4], 226
[(M Ϫ 2 CO)ϩ, 17.4] indicating the final product was a
quinone (or a mixture of quinone isomers) with a molecular
weight equal to 282. Three components were isolated by
dry column chromatography (neutral Al2O3, benzene) after
oxidation of benzo[a]pyrene in conditions used for run 4 of
Table 1 and characterized by UV spectroscopy according to
ref.[18] (Scheme 1). The UV/Vis spectra for all three qui-
nones were in an excellent agreement with the data pre-
viously published[18]. Based on reported extinction coef-
ficients, the distribution of quinones was as follows: 6,12-
benzo[a]pyrenedione (32%, first eluted product from col-
umn), 1,6- benzo[a]pyrenedione (44%, second eluted prod-
uct) and 3,6- benzo[a]pyrenedione (24%, third eluted prod-
uct). The total yield for all three quinones was quantitative
Figure 1. Structures of iron(III) complexes of tetrasulfophthalocy-
anine (FePcS) and meso-tetrakis(3,5-disulfonatomesityl)porphyrin
Table 1. Oxidation of polyaromatic hydrocarbons PAH by H2O2
and KHSO5 catalyzed by sulfonated phthalocyanineiron(III)
FePcS[a]
Run Oxidant
pH
% Cat./Sub. Conversion [%]
1 min 5 min 60 min
Benzo[a]pyrene oxidation
1
2
3
4
KHSO5
KHSO5
H2O2
2
7
7
7
1
50
14
11
43
97
100
70
1
41
1
3.7
20
100
20
H2O2
100
Anthracene oxidation
5
6
7
8
H2O2
H2O2
H2O2
H2O2
4.4
7
7
10
10
3.7
1
12
40
96
100
57
74
28
86
32
7
15
Anthracene oxidation by FeTMPS/KHSO5
9
KHSO5
7
3.7
11
33
100
Phenanthrene oxidation
10
11
12
13
14
15
KHSO5
H2O2
H2O2
H2O2
H2O2
H2O2
7
3.7
3.7
10
10
10
10
28
40
64
35
66
64
60
80
5.3
5.3
5.5
5.9
4.4
10
24
n.d.
32
n.d.
n.d.
n.d.
46
45
24
[a]
n.d.: not determined.
The H2O2 oxidation of phenanthrene is pH-dependent.
Using a 10% catalyst/substrate ratio the highest conversion
of phenanthrene was obtained at pH ϭ 4.4 (conversion be-
1270
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281

Polycyclic Aromatic Hydrocarbons Catalyzed by Iron Tetrasulfophthalocyanine FePcS
FULL PAPER
Scheme 1. Oxidation products of benzo[a]pyrene by FePcS/H2O2
ing rather low at pH ϭ 7) (Figure 2). But it should be noted
Biphenyl-2,2Ј-dicarboxylic acid (product of cleavage of
that anthracene was quickly and completely oxidized within one aromatic cycle) was the main oxidation product of
1 min at pH ϭ 7 (Table 1, run 6) while 1 hour was necessary phenanthrene, 9,10-phenanthrenequinone being the second
to reach 96% conversion at pH ϭ 4.4 (the pH optimum for main product. Three other minor products of the oxidative
phenanthrene). At pH ϭ 7 the order of substrate reactivity degradation of phenanthrene were also identified by GC-
was in agreement with oxidation potentials: benzo[a]pyrene MS analysis: 2-formyl-2Ј-hydroxybiphenyl, 2-carboxy-2Ј-
> anthracene > phenanthrene (the ionization potentials be- hydroxybiphenyl and 2,2Ј-biphenylcarbolactone (Scheme
ing 7.2, 7.4 and 8.1 eV, respectively[2]).
3). So, the FePcS/H2O2 system is able to cleave the aromatic
cycle of phenanthrene and to perform oxidative decar-
boxylation.
Figure 2. pH dependence of phenanthrene oxidation by FePcS/
[a]
H2O2
Oxidation of Anthracene by FeTMPS/KHSO5
We also checked the oxidation of anthracene by
FeTMPS/KHSO5 performed in an MeCN/25 m phos-
phate buffer mixture (3:1, v/v) at pH ϭ 7. The concen-
trations of oxidant, substrate and catalyst were 5 m, 1 m
and 37 µ, respectively. The anthracene conversion was
complete after 60 min (run 9, Table 1). The yields of ox-
anthrone and anthraquinone were determined by HPLC.
Differing from the FePcS/H2O2 mediated oxidation, the
main product with FeTMPS/KHSO5 was oxanthrone (yield
by HPLC ϭ 60%), while only 3% of anthraquinone were
found. Isolation of reaction products from a large scale an-
thracene oxidation by FeTMPS/KHSO5 followed by NMR
analysis gave 43% yield for oxanthrone, 16% yield for
anthraquinone and 14% yield of non-identified compounds
(probably coupling products). A partial oxidation of ox-
[a]
Reaction conditions: [catalyst] ϭ 0.1 m, [oxidant] ϭ 5 m,
[substrate] ϭ 1 m, acetonitrile/50 m phosphate buffer at pH ϭ
7 (3:1, v/v). Reaction time 60 min.
The FePcS-catalyzed oxidation of anthracene by H2O2 anthrone to anthraquinone might have occured during the
resulted in the formation of anthrone (tautomeric form of sample preparation for NMR analysis, explaining the dis-
9-hydroxyanthracene), oxanthrone (tautomeric form of crepancy with the HPLC analysis of the reaction mixture.
anthrahydroquinone) and anthraquinone (Scheme 2). The This rather unexpected result, oxanthrone, an easily oxidi-
yields of oxanthrone (5%) and anthraquinone (25%) were zable compound, as main reaction product in the presence
determined after 1 h by HPLC using calibration curve and of strong oxidants, stimulated us for a detailed mechanistic
molar extintion coefficients obtained with authentic study of anthracene oxidation by both catalytic systems,
samples.
FePcS/H2O2 and FeTMPS/KHSO5.
Scheme 2. Oxidation products of anthracene by FePcS/H2O2 or FeTMPS/KHSO5
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281
1271

A. Sorokin, B. Meunier
FULL PAPER
Scheme 3. Oxidation products of phenanthrene by FePcS/H2O2
Stability of Anthrone, Oxanthrone and Anthraquinone in the
Presence of FePcS/H2O2 and FeTMPS/KHSO5
GC-MS analyses and substrate conversions were obtained
by HPLC analyses (see Experimental Section). Equations
to calculate KIEs depend on a reaction order in sub-
strate[19]. When the reaction is first order in substrate KIEs
In these experiments we used the same reaction con-
ditions as those for anthracene oxidation: a MeCN/50 m
phosphate buffer pH ϭ 7 reaction medium (3:1, v/v), [cata-
lyst] ϭ 37 µ, [oxidant] ϭ 5 m, [substrate] ϭ 1 m, except
for anthraquinone (its concentration was reduced to 0.5 m
because of its low solubility in the reaction mixture). Under
these reaction conditions the conversions of anthrone in the
presence of FePcS/H2O2 and FeTMPS/KHSO5 were 23%
and 38%, respectively, after 1 h, indicating that anthrone
oxidation was slightly faster with FePcS/KHSO5. Ox-
anthrone was easily oxidized to anthraquinone by FePcS/
H2O2: the conversion was 90% after only 5 min. Again,
oxanthrone was relatively stable with FeTMPS/KHSO5: the
conversions were only 18 and 23% after 30 and 60 min,
respectively. Finally, we checked that anthraquinone was
stable in both catalytic systems. There was no observable
conversion after 60 min with FeTMPS/KHSO5 and only
10% with FePcS/H2O2. These results on stability of prod-
ucts of anthracene oxidation were in good agreement with
the composition of the reaction products for both catalytic
systems.
o
can be determined from equation (1) where aHo, aD and
aH, aD denote initial and current substrate concentrations
of H-labeled and D-labeled substrates, respectively.
o
kH/kD ϭ log (aH/aHo) / log (aD/aD
)
(1)
Determination of the Reaction Order for the Substrate in
Anthracene Oxidation by FePcS/H2O2
The reaction order in anthracene was determined with
concentrations of H2O2, anthracene and FePcS being 5 m,
1 m and 37 µ, respectively. The reactions were carried
out in an MeCN/50 m phosphate buffer mixture at pH ϭ
7 (3:1, v/v). Under these conditions the catalytic reaction
follows a first-order kinetic in substrate at least for 2 times
the reaction half-time and KIEs could be then determined
according to equation (1).
KIEs in Anthracene and Phenanthrene Oxidation by FePcS/H2O2
and FeTMPS/KHSO5
Method of Determination of Kinetic Isotope Effects
Kinetic isotope effects (KIEs) can be determined in intra-
Analysis of the reaction mixture after 3Ϫ5 min of reac-
or intermolecular competitive experiments by the analysis tion (conversions being 40Ϫ60%) indicated a preferential
of the ratios of reaction products resulting from CϪH ver- consumption of [D10]anthracene corresponding to an in-
sus CϪD bond transformations. Such analyses are reliable verse isotope effect. The kinetic isotope effect in anthracene
when the oxidation products, different only in their isotopic oxidation by the FePcS/H2O2 system was determined to be
substitution, are primary oxidation products and are stable equal to 0.75 ± 0.02. Oxidation of anthracene by FeTMPS/
in the reaction conditions. Since this is not a case in PAH KHSO5 presented the same isotope effect of 0.76 ± 0.06
oxidations by FePcS/H2O2 system we decided to determine (Table 2). In contrast, oxidation of phenanthrene by both
the KIE values by analyzing of the competitive consump- catalytic systems proceeds with no significant isotope effect:
tion of the deuterated versus the non-deuterated substrate. 0.96 ± 0.04 and 1.02 ± 0.01 for FePcS/H2O2 and FeTMPS/
In this case intermolecular KIE values of the initial oxi- KHSO5, respectively (Table 2).
dation reaction can be obtained. In order to determine
In order to find the origin of this inverse isotope effects
these KIEs concentrations of labeled and unlabeled sub- we prepared 9,10-dideuterioanthracene and determined the
strates should be determined before and after the oxidation KIE value in the competitive oxidation of the mixture of
reaction by monitoring the isotopic composition of the in- 9,10-dideuterioanthracene and [H10]anthracene. In this lat-
itial reaction mixture, the substrate conversion and the iso- ter case the KIE value were 0.90 ± 0.05 and 0.95 ± 0.01
topic composition of the remaining substrate after reaction. with the FePcS/H2O2 and FeTMPS/KHSO5 systems,
Isotopic compositions of substrates were determined by respectively.
1272
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281

Polycyclic Aromatic Hydrocarbons Catalyzed by Iron Tetrasulfophthalocyanine FePcS
FULL PAPER
Table 2. Kinetic isotope effect[a] in anthracene and phenanthrene eled water. In control experiments we checked that light
oxidation by FePcS/H2O2 and FeTMPS/KHSO5 at 20°C
oxygen atoms of anthraquinone did not exchange with
heavy oxygen atoms of 18O-labeled water under the used
KIE in the competitive oxidation of
reaction conditions, i.e. a mixture of acetonitrile/50m
Catalytic
system
[H10]anthracene/
[D10]anthracene[b]
([H10]anthracene/9,10-
[D2]anthracene)
[H10]phenanthrene/
[D10]phenanthrene[c]
phosphate buffer (3:1, v/v) at pH ϭ 7. Also, KHSO5 itself
does not exchange its oxygen atoms with water[20c]. FePcS-
mediated oxidations of anthracene were performed under
standard conditions (run 7 of Table 1) either under air or
under nitrogen, using a 50:50 mixture of [H10]anthracene
and [D10]anthracene. The oxygen atom distribution was
analyzed in both deuterated and non-deuterated products
for each experiment. In all cases, the oxygen atom distri-
bution for a defined product was exactly the same. For ex-
ample, the 18O incorporation into anthraquinone was fol-
lowed by determination of the molecular ion percentages at
208/216 [M], 210/218 [M ϩ 2] and 212/220 [M ϩ 4] corre-
sponding to the incorporation of none, one or two 18O oxy-
gen atoms in [H8]anthraquinone and [D8]anthraquinone,
respectively. In both cases, under air or under nitrogen,
anthrone, the primary oxidation product, contained only a
light oxygen atom. This result indicates that this oxygen
atom originated from H2O2, the non-labeled oxidant. In
contrast, both 16O and 18O atoms were incorporated into
anthraquinone. When the reaction was performed under air,
the molecular cluster consisted of M (68%) and M ϩ 2
(32%) peaks, meaning that the 18O atom was unambigu-
ously only located in one carbonyl group of this quinone. If
18O would occupied both carbonyl positions, the molecular
cluster would have consisted on three peaks: M, M ϩ 2 and
M ϩ 4. So, anthraquinone contained 100% of 16O in one
carbonyl position, 36% of 16O and 64% of 18O in another
one (Figure 3). When the reaction was performed under
nitrogen, again, only two peaks were found in the molecular
cluster: M (56%) and M ϩ 2 (44%). Consequently, 16O in-
corporation in one carbonyl position was complete, another
carbonyl position was enriched with heavy oxygen: 12% of
16O and 88% of 18O (Figure 3). This more significant incor-
poration of 18O at the second carbonyl position produced
in the absence of dioxygen suggests a possibility of O2 in-
volvement into oxidation pathway from anthrone to anthra-
quinone.
FePcS/H2O2
FeTMPS/
KHSO5
0.75 ± 0.02 (0.90 ± 0.05) 0.96 ± 0.04
0.76 ± 0.06 (0.95 ± 0.01) 1.02 ± 0.01
[a]
These data are the mean values obtained from 3 independent
[b]
oxidation reactions. Ϫ
Reaction conditions: [catalyst] ϭ 37 µ,
[oxidant] ϭ 5 m, [substrate] ϭ 1 m, MeCN/50 m phosphate
buffer pH ϭ 7 (3:1, v/v). Ϫ [c] Reaction conditions: [catalyst] ϭ 100
µ, [oxidant] ϭ 5 m, [substrate] ϭ 1 m, MeCN/33.3 m phos-
phate buffer pH ϭ 4.6 (1:1, v/v).
Analysis of Isotopic Composition of Products of Anthracene
Oxidation
Anthrone, oxanthrone and anthraquinone were identified
as oxidation products of anthracene by NMR and GC-MS
techniques. GC-MS analysis of the reaction mixture indi-
cated that anthrone, the primary oxidation product, was en-
riched in the deuterated derivative as expected from the in-
verse isotope effect on the first step of anthracene oxi-
dation. In contrast, anthraquinone, a second oxidation
product, was enriched in unlabeled product indicating a
KIE value > 1 during the oxidation of anthrone to this
quinone. Oxanthrone was unstable under GC-MS analysis
conditions and its isotopic composition could not be deter-
mined. The molecular peak of deuterated anthrone formed
after 3 min of reaction was 202 corresponding to a [D8]-
containing molecule, not a [D9]-molecule as expected for 9-
hydroxyanthracene. This result suggested that the equilib-
rium between the two tautomeric forms, anthrone and 9-
hydroxyanthracene, was rather fast (Scheme 4). Conse-
quently, a putative σ adduct with a FeϪOϪarene bond,
which should stabilize the 9-hydroxyanthracene tautomer,
could not be a long-lived intermediate as previously ob-
served in the metalloporphyrin-catalyzed oxidation of 2-
methylnapthalene[12a]
.
18O-Labeling Experiments
We also studied the 18O incorporation from labeled water
Oxidations of organic substrates catalyzed by metallo- into the oxidation products of anthracene mediated by
porphyrin[12][20][21] and metallophthalocyanine[16] com- FeTMPS/KHSO5. In this latter case, the mass spectrum of
plexes in the presence of 18O-labeled water or dioxygen pro- anthrone indicated the incorporation of 18O atom, unlike
vided valueable mechanistic information. In order to inves- with the FePcS/H2O2 oxidation. The percentage of 18O
tigate the mechanism of these catalytic PAH oxidations we found in anthrone was 63%, 16O incorporation being only
performed the oxidation of anthracene by FePcS/H2O2 and 37% (Figure 3). The MS molecular cluster of anthraqui-
FeTMPS/KHSO5 in reaction mixtures containing 18O-lab- none consisted of 3 peaks: M (16%), M ϩ 2 (59%) and M
Scheme 4. Deuterium exchange during the formation of anthrone in aqueous medium
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281
1273

A. Sorokin, B. Meunier
FULL PAPER
Figure 3. 18O incorporation into anthrone and anthraquinone during the oxidation of anthracene by FePcS/H2O2 and FeTMPS/KHSO5
in the presence of H218O
ϩ 4 (25%), indicating that 18O was incorporated in both anthracene (KIE ϭ 0.75 and 0.76, respectively) or phen-
carbonyl groups of the quinone (the amount of 16O and anthrene (KIE ϭ 0.96 and 1.02, respectively) (Table 2).
18O being 45 and 55%, respectively) (Figure 3). From these
Mechanistic Significance of Inverse Isotope Effects
labeled water experiments we concluded that active species
involved in the oxidation of anthracene by FePcS/H2O2 or
by FeTMPS/KHSO5 must be different.
The observation of a significant inverse isotope effect in
the anthracene oxidation by FePcS/H2O2 or FeTMPS/
KHSO5 indicated that these KIEs could not be interpreted
as primary KIEs involving a CϪH(D) bond cleavage in the
Discussion
We found that FePcS associated with H2O2 can be useful rate determining step of the initial step of the oxidation
in oxidation of polycyclic aromatic compounds. Anthracene reaction. Consequently, these inverse KIEs precluded an
was used as substrate to perform kinetic isotope studies and oxygen rebound mechanism which involves a hydrogen
oxygen labeling experiments. Then we compared the results atom abstraction from substrate by an electrophilic high-
of the catalytic oxidations of anthracene by FePcS/H2O2 valent oxoiron complex to give a short-life radical inter-
with those obtained with FeTMPS/KHSO5 in order to gain mediate followed by rapid recombination step[23]. Inverse
a more complete view on the mechanisms of these oxidation KIEs are usually associated with secondary KIEs when no
reactions catalyzed by two related iron macrocyclic com- CϪH(D) bond is cleaved or formed, but this bond should
plexes. High-valent oxometal complexes have been pro- modify some parameters in the transition state (TS). This
posed as active intermediates in hydrocarbon oxygenations is usually the case when a CϪH(D) bond is located at the
(hydroxylations and epoxidations) using KHSO5, NaOCl or reaction site itself, containing the leaving group (α second-
mCPBA as oxidants and porphyriniron or -manganese ary KIE) or at the adjacent carbon atom (β secondary
compounds as catalysts[20][21][22]. We recently proposed a KIE). In the case of the intermolecular [H10]anthracene/
nucleophilic peroxoiron(III) complex of tetrasulfophthalo- [D10]anthracene competitive oxidation an α secondary KIE
cyanine, PcSFeIIIOOH, as active species responsible for the is observed. Generally, the magnitude of an α-secondary
oxidative cleavage of quinones generated during the oxi- KIE is determined by an inverse stretching and a normal
[15][16]
dation of 2,4,6-trichlorophenol by FePcS/H2O2
.
bending contribution to the KIE value[24][25]. For example,
In the present work, several important differences must in SN2 reactions, a tighter TS, with short nucleophile α-
be noted in anthracene oxidations performed by FePcS/ carbon and/or α-carbon leaving group bond lengths, the
H2O2 or FeTMPS/KHSO5. These differences are (i) a dif- KIE values would be inverse or normal and small[24]. Based
ferent composition of oxidation products depending on the on calculations with model systems, the α-secondary KIEs
oxidation system used (Scheme 2), (ii) a different reactivity have been interpreted as reflecting the changes in the out-
of oxanthrone, which is relatively stable in the presence of of-plane bending motion, i.e. hybridization changes or dif-
FeTMPS/KHSO5, but very quickly oxidized by FePcS/ ferences in the loose/tight character of the TS[26]. The acti-
H2O2, (iii) and very different 18O incorporations into vation energy for sp2 Ǟ sp3 change in hybridization is lower
anthrone and anthraquinone in experiments performed in for the deuterated substrate compared to the non-deuter-
the presence of H218O (Figure 3). On the other side, both ated analogue, creating then an inverse KIE[27]. When the
catalytic systems, FePcS/H2O2 and FeTMPS/KHSO5, TS of the reaction occurs very late along the reaction coor-
showed essentially the same KIE values in the oxidation of dinate, i.e. the TS exhibits essentially a fully sp3-hybridized
1274
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281

Polycyclic Aromatic Hydrocarbons Catalyzed by Iron Tetrasulfophthalocyanine FePcS
FULL PAPER
carbon atom like in the reaction product, the kH/kD value weak agostic interactions between CϪH bonds of the al-
will be more pronounced, the maximum KIE expected for kane and the metal center. p-Xylene can interact with the
a sp2 Ǟ sp3 rehybridization will be 0.71[27]. Values of kH/ metal center of an porphyriniron(III), the smallest FeϪC
˚
˚
kD ranging from 0.8 to 0.9 correspond to TS geometries distance was 2.94 A and the iron atom was drawn 0.07 A
somewhere between sp2 and sp3 and are more frequently out of the mean plane of the porphyrin core indicating a
observed. Inverse secondary KIEs have been measured for possible weak agostic interaction[41]
.
the addition of a variety of electrophilic reagents to olefins
In the present case of the oxidation of anthracene cata-
and aromatics, e.g. for the reaction of 1-naphthylcarbene to lyzed by phthalocyanine- or porphyriniron complexes, an-
toluene (kH/kD ϭ 0.52)[28]. An other example of an inverse thracene, a polyaromatic substrate, should have a pro-
α-secondary kH/kD equal to 0.83 has been recently observed nounced tendency to form a stacking complex with FePcS
during of the [2 ϩ 2] photocycloaddition of arylalkenes to or FeTMPS. A rather polar reaction medium (an aceto-
C60[29]. Inverse KIEs have been reported for styrene epoxi- nitrile/water mixture) should favor stacking interactions be-
dations catalyzed by cytochrome P-450 (kH/kD ϭ 0.93) or tween two large planar hydrophobic molecules. Then a
performed with m-chloroperbenzoic acid (kH/kD
ϭ
question arises: can the difference in CϪH/CϪD substi-
0.82)[27a] or by hexacyanoferrate(III) in acid medium (kH/ tution on anthracene generate different stacking interac-
kD ϭ 0.80)[27b]. The cytochrome P-450 catalyzed oxidation tions between [H10]anthracene or [D10]anthracene with the
of isotopically labeled chlorobenzenes provides also an in- macrocyclic plane molecule of FePcS or FeTMPS? In other
verse KIE (0.95)[30]. This result has been explained by an words, are the observed KIEs only related to hydrophobic
initial attack of an activated triplet-like oxygen atom on the interactions between the aromatic substrate and the macro-
π-system of substrate to generate a tetrahedral intermediate cyclic catalyst? The differences observed on inverse
followed by rearrangement to phenol directly or via epoxide KIEs during the [H10]anthracene/[D10]anthracene and
or ketone intermediates[30]. The aromatic hydroxylation of [H10]anthracene/[D2]anthracene competitive oxidations
o- and p-xylene, catalyzed by cytochrome P-450, showed (Table 2) could be due to possible differences in stacking
also an inverse α-secondary isotope effects (0.83Ϫ0.94) interactions between [D10]anthracene and [D2]anthracene
which were consistent with an addition reaction of a with the macrocycle of the catalyst resulting from a slight
oxoiron species onto an sp2-hybridized carbon atom[31]
.
difference of the nature of CϪH bonds compared to CϪD
The inverse nature of KIEs was anticipated for sp2-to-sp3 bonds. So, the stacking interactions should increased with
rehybridization during the covalent modification of human the deuterium content of anthracene. This is the case with
5α-reductases by finasteride (a specific inhibitor) by a nucle- both catalytic systems: the KIE is 0.90 for [D2]anthracene
ophilic N addition to the double bond of this steroid-type and 0.75 with [D10]anthracene with FePcS/H2O2 (0.95 and
molecule[32]. The reaction of the nitroethane anion (the α- 0.76 with FeTMPS/KHSO5, respectively) (Table 2). Conse-
carbon atom is sp2-hybridized) with -amino acid oxidase quently, the value of the inverse KIE due to the sp2-to-sp3
proceeds with an inverse kH/kD ϭ 0.84 which was indicative hybridization change should be enhanced in a highly
of significant sp2-to-sp3 rehybridization in the TS[33]. Apart stacked substrate-catalyst intermediate during the addition
from the sp2-to-sp3 rehybridization, inverse isotope effects of an electrophilic oxoiron complex to the sp2-carbon atom
can be due to the difference in CϪH and CϪD distances, of anthracene to give a σ adduct. The data on 18O labeling
CϪD bond being slightly shorter than corresponding CϪH experiments are also consistent with this proposal (see be-
bond (d ϭ 1.112 and 1.107 A, respectively)[34]. The racemi- low).
˚
zation of 2,2Ј-dibromo-4,4Ј-dicarboxybiphenyl[35] and 9,10-
The quasi-absence of KIE observed in the phenanthrene
dihydro-4,5-dimethylphenanthrene[36] showed inverse KIEs oxidation by FePcS/H2O2 or by FeTMPS/KHSO5 (0.96 and
of 0.85 and 0.88, respectively. The origin of a large inverse 1.02, respectively) can be due to the reduction of stacking
secondary KIE in the bromination of a sterically congested interactions compared to anthracene and also to the differ-
olefin (0.36Ϫ0.65) has been attributed to the fact that CϪD ent nature of C9ϪH and C10ϪH bonds in phenanthrene.
bonds are shorter and less sterically demanding[37]. Large The C9ϪC10 bond of phenanthrene exhibits a relatively
inverse equilibrium isotope effects have also been observed high bond order, indicating that it possesses a considerable
for the oxidative addition of H2 to trans-W(PMe3)4I2 (KH
/
/
degree of double bond character (K region bonds of polyar-
omatics), while anthracene bonds show essentially an aro-
eq
KD ϭ 0.63)[38a] and cyclohexane to CpRh(CO)2 (KH
eq
eq
KD ϭ 0.1, at 163Ϫ193 K)[38b]. Recently, a substantial in- matic character (meso region bonds)[42]. As a result, the
eq
verse isotope effect of kH/kD ϭ 0.48 has also been published main product of phenanthrene oxidation was biphenyl-2,2Ј-
in CoIII-catalyzed polymerization of ethylene[39] and ex- dicarboxylic acid. A similar cleavage of the double bond of
[16]
plained by the presence of
a
β-agostic Co···H···C styrene was observed when oxidized by FePcS/H2O2
.
(Co···D···C) interaction, the inverse KIE being associated
with the release of the agostic bond. Related to this in-
terpretation of an inverse KIE, it must be noted that com-
Mechanistic Significance of 18O Labeling Experiments
In aqueous media oxo(porphyrin)metal(V) complexes
plexation of n-heptane with double A-frame porphy- have an axial hydroxo ligand. In order to explain the ob-
riniron(III) has been recently observed in crystals[40]. An served 50% 18O incorporation from H218O in epoxides dur-
˚
out-of-plane displacement of the iron center of 0.26 A and ing olefin epoxidations by a metalloporphyrin/KHSO5 sys-
an Fe···C distance of 2.5 A strongly suggest moderate to tem[20a][21b][21c], hydroxylation of DNA deoxyriboses[20b] or
˚
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281
1275

A. Sorokin, B. Meunier
FULL PAPER
CϪH bond hydroxylation of 4-isopropylbenzoic acid[20c] it Scheme 5. Oxidation products of anthracene and tautomeric equ-
ilibria for anthrone and oxanthrone
was proposed that the active hydroxo(oxo)(porphyrin)met-
al(V) complex undergoes a “redox tautomerism” (Figure 4).
The 18O incorporation depends on the reaction con-
ditions[21b], primarily on relative kinetics of oxidation step
and ligand exchange processes between the hydroxo ligand
and bulk water and the nature of the catalyst. The rate of
the oxo-hydroxo interchange in the oxo(porphyrin)man-
ganese(V) complex was estimated to be about 103 sϪ1[21c]
.
Figure 4. Redox tautomerism in a hydroxo(oxo)metal(V) species
having a water-soluble porphyrin ligand (O ϭ 16O, black O ϭ 18O)
The large 18O incorporation into anthrone (63%) and
anthraquinone (55%) observed during FeTMPS/KHS16O5
anthracene oxidation in the presence of H218O is consistent
with the involvement of oxo(porphyrin)iron complex as ac-
tive species. By contrast, the absence of 18O incorporation
into anthrone, produced in the oxidation by FePcS/H2O2,
could be explained by the involvement of an active oxo-
metal species unable to undergo such an oxo-hydroxo equi-
librium (in particular if there is no hydroxo ligand in the
axial position trans to the oxo group).
give a σ complex is probably the initial oxidation step (Fig-
ure 5). The inverse isotope effect, indicative of an sp2-to-
sp3 rehybridization in the TS containing both substrate and
catalyst stacked and the observed 18O incorporation into
anthrone, the first oxidation product, typical for oxoiron
complex mediated oxidation in aqueous medium strongly
suggested the formation of a new carbonϪoxygen bond to
form a σ complex A. This intermediate A can be trapped
by O2 to produce oxanthrone and anthraquinone via an
intermediate radical B. An intramolecular electron transfer
in intermediate A can lead to an intermediate carbocation
C, which can undergo either a proton elimination to give a
σ complex D or a nucleophilic attack of water to produce
an intermediate E, providing a reaction pathway for ox-
anthrone formation without intermediate formation of
anthrone. The hydrolysis of the FeIIIϪOϪaromatic bond in
both intermediates D and E gave 9-hydroxyanthracene
(quickly tautomerized to anthrone) and 9,10-dihydro-9,10-
dihydroxyanthracene F. Then F can be further oxidized by
a 2eϪ process with the release of 2Hϩ, producing ox-
anthrone with an increased 18O content, or by oxidation of
a CϪH bond by an oxygen rebound mechanism. The GC-
Tautomeric Equilibria in Oxidation Products of Anthracene
The sequence of anthracene oxidation to anthraquinone
and the relation between tautomeric forms of products are
depicted in Scheme 5. The first oxidation product, 9-
hydroxyanthracene is in fast keto-enol equilibrium with
anthrone, which is the predominant form. The anthrone/9-
hydroxyanthracene ratio determined by NMR in CDCl3 at
25°C was estimated to be 80:20[43]. 9,10-Dihydroxyanthra-
cene exists in solution in equilibrium with oxanthrone
(Keq ϭ 0.13) and is the major tautomer (89%)[44]. However,
the equilibrium is reached very slowly. The rate constants
for the 9,10-dihydroxyanthracene-oxanthrone interconver-
sion were determined to be kϪ2 ϭ 3 ϫ 10Ϫ6 sϪ1 and k2 ϭ
4 ϫ 10Ϫ7 Ϫ1[45]. It means that in 1 m solution only 1%
s
of oxanthrone will be transformed into 9,10-dihydroxyan-
thracene after 1 h at room temperature.
Mechanism of Anthracene Oxidation by FeTMPS/KHSO5
The oxidation of anthracene by FeTMPS/KHSO5 was MS analysis of [H10]anthracene/[D10]anthracene oxidation
significantly faster than those of anthrone and oxanthrone. products after 6 min of reaction revealed that anthraqui-
Under the same conditions the conversions of anthracene, none was enriched with nondeuterated material: 57% of
anthrone and oxanthrone after 30 min were 100, 38 and [H8]anthraquinone and 43% of [D8]anthraquinone, indicat-
18%, respectively. These data indicated that there was an ing a KIE value > 1 on the anthrone-to-anthraquinone oxi-
oxidation pathway to give directly oxanthrone, the main dation step. The 18O content in anthraquinone (55%) was
oxidation product, without intermediate formation of not increased compared to that in anthrone (63%). Both
anthrone.
findings suggest that oxidation of intermediate F proceeds
An electrophilic attack of a high-valent oxo(porphyrin)- by an oxygen rebound mechanism followed by H2O elimin-
iron complex at a central sp2-carbon atom of anthracene to ation to generate oxanthrone. The fact that the metallopor-
1276
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281

Polycyclic Aromatic Hydrocarbons Catalyzed by Iron Tetrasulfophthalocyanine FePcS
FULL PAPER
Figure 5. Proposed mechanism for anthracene oxidation by FeTMPS/KHSO5 (O ϭ 16O, black ᭹ ϭ 18O)
phyrin-catalyzed oxidation of anthracene gave oxanthrone, enes, e.g. the addition of alkyllithium reagents to anthra-
not the more oxidizable 9,10-hydroxyanthracene tautomeric cene, has been published[46], such a peroxoiron(III) complex
form, explains the rather low yield of anthraquinone, since is probably not nucleophilic enough for a direct attack onto
oxanthrone is relatively stable under the reaction conditions an sp2-carbon atom of a PAH. Another possible active
(oxanthrone conversion being only 23% after 60 min).
metal-oxygen species could be a high-valent oxoiron com-
plex, having two redox equivalents above the FeIII state.
Several lines of evidence indicate that it is not the case for
the FePcS/H2O2 system[15b]. The fact that the oxygen atom
Active Species in FePcS/H2O2 Oxidations
In the FePcS/H2O2 mediated oxidation of trichloro- in anthrone originated from H2O2 (Figure 3) also indicated
phenol[15][17] a nucleophilic PcSFeIIIOOϪ complex was pro- that an oxoiron(V) complex was not involved in anthracene
posed to be the active species, able to perform two key oxidation by FePcS/H2O2. With such an oxoiron(V) com-
steps: (i) a nucleophilic addition to the electron-deficient plex we could expect both 16O and 18O incorporations due
double bond of intermediate quinones to provide epoxy to redox tautomerism, as observed in anthracene oxidation
quinones and (ii) a nucleophilic attack on the carbonyl by FeTMPS/KHSO5. The other possible formulation of the
group of the different quinones, leading to their ring cleav- active species involved in PAH oxidations by FePcS/H2O2
age[16]. Although the addition of nucleophiles to polyar- is a PcSFeIVϭO species[15b]. It can be formed by a 1eϪ oxi-
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281
1277

A. Sorokin, B. Meunier
FULL PAPER
Figure 6. Generation of peroxoiron(III) and oxoiron(IV) species by absence of redox tautomerism for PcSFeIVϭO explained the
100% incorporation of 16O into anthrone during the FePcS/
activation of FePcS with hydrogen peroxide
H2O2 oxidation of anthracene performed in heavy water.
Mechanism of Anthracene Oxidation by FePcS/H2O2
The first step of the FePcS/H2O2 mediated oxidation of
anthracene resembles that described in the FeTMPS/
KHSO5 oxidation (Figure 7). The nearly identical inverse
KIE (0.75) found for the first reaction step suggested also
the formation of a σ complex, intermediate AЈ which can
dation of PcSFeIIIOH complex (Figure 6, pathway A) or by be trapped by O2 to produce oxanthrone and anthraqui-
the intermediate formation of a µ-peroxo dimer (pathway none via a radical intermediate BЈ. The significant 16O in-
B). In the former case 18O from water should be found in corporation in anthraquinone when the oxidation was per-
anthrone obtained in the presence of H218O. The absence formed under air compared to that obtained under nitrogen
of 18O incorporation into anthrone during the oxidation of (Figure 3) was strongly indicative of this pathway 1. Hy-
anthracene by FePcS/H2O2 in H218O eliminated this pos- drolysis of the FeIIIϪOϪaromatic bond followed by proton
sibility. Consequently, PcSFeIVϭO should be formed by elimination and 1eϪ oxidation generated 9-hydroxyanthra-
pathway B by the homolytic cleavage of a µ-peroxo dimer cene which rapidly tautomerized to the thermodynamically
(we proposed the pathway A in ref.[15b] before having on favored anthrone. Successive 1eϪ oxidation and proton
hands the data of these labeling studies). This PcSFeIVϭO elimination steps gives rise to carbocation IЈ, which can re-
species is neutral and does not require the presence of an act with water (pathway 2) to produce oxanthrone and
anionic axial anion ligand (e.g. HOϪ) to compensate a anthraquinone. PcSFeIVϭO complex is not such a strong
charge. Consequently, no redox tautomerism is possible in oxidant to perform the abstraction of an H atom of a ali-
the absence of HOϪ ligand trans to the oxo entity. This phatic or aromatic CϪH bond. However, PcSFeIVϭO is
Figure 7. Proposed mechanism for anthracene oxidation by FePcS/H2O2 (O ϭ 16O, black ᭹ ϭ 18O)
1278
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281

Polycyclic Aromatic Hydrocarbons Catalyzed by Iron Tetrasulfophthalocyanine FePcS
FULL PAPER
probably a sufficiently good 1eϪ oxidant to perform 1eϪ
oxidations of intermediate products or of oxanthrone by
successive 1eϪ oxidations and proton eliminations to pro-
vide anthraquinone.
Materials: All chemicals used were of reagent grade. Potassium
monopersulfate was a gift from Interox (CuroxR) and is the triple
salt 2KHSO5·KHSO4·K2SO4. Hydrogen peroxide was obtained
from Janssen Chimica as a 35% aqueous solution. H218O (98 atom-
%) was supplied by Eurisotop (Gif-sur-Yvette, France). Lithium
aluminum deuteride was purchased from Fluka (minimum 99
atom-% D). (Tetrasulfophthalocyanine)iron (FePcS) was prepared
according to the method of Weber and Busch.[47] The iron complex
of octasodium meso-tetrakis(3,5-disulfonatomesityl)porphyrin
(TMPS) was synthesized according to ref.[48]. Anthracene (97%),
The rather low yields of anthrone and anthraquinone
after 30 min (5 and 21%, respectively) might be due to rad-
ical intermediates capable of undergoing side reactions, e.g.
coupling reactions. Accordingly, the presently proposed
mechanism, the first oxygen atom incorporated into the dif-
ferent products originates from PcSFeIVϭO (16O in labeling phenanthrene (98%), benzo[a]pyrene (98%), [D10]anthracene (> 98
experiments), and the value of 18O incorporation of the sec-
atom-% D), [D10]phenanthrene (97 atom-% D), anthrone (96%),
ond oxygen atom of anthraquinone from H218O will depend
anthraquinone (97%), phenanthrenequinone (99%), 9,10-dibromo-
anthracene (98%) were purchased from Aldrich. Oxanthrone was
on the relative rates of the different pathways indicated on
prepared from anthrone via 10-bromoanthrone according to ref.[45]
.
Figure 7. Very significant 18O incorporation into the second
oxygen position of anthraquinone under air (64%) and, es-
pecially under nitrogen (88%), strongly suggest that path-
way 2 is the major reaction pathway.
NMR (CD2Cl2): δ ϭ 2.47 (d, 1 H, J ϭ 9.5 Hz, OH), 5.78 (d, 1 H,
J ϭ 9.5 Hz, H10), 7.54 (td, 2 H, J ϭ 7.5 and 0.7 Hz, H3), 7.73 (td,
2 H, J ϭ 7.5 and 1.5 Hz, H2), 7.90 (dd, 2 H, J ϭ 7.5 and 0.7 Hz,
H4), 8.25 (dd, 2 H, J ϭ 7.5 and 1.5 Hz, H1); (CD2Cl2 ϩ small
amount of D2O): the same signals, except doublet at δ ϭ 2.47 dis-
appeared and the doublet at δ ϭ 5.78 became a singlet.
Conclusion
The (tetrasulfophthalocyanine)iron/hydrogen peroxide
oxidation system, based on an easy-to-prepare catalyst and
a “clean” oxidant, is capable of oxidizing polycyclic aro-
matic hydrocarbons, like anthracene, phenanthrene and
benzo[a]pyrene. Data obtained from product distribution,
kinetic isotope effects and 18O labeling studies, as well as
a comparative study of anthracene oxidation by FeTMPS/
KHSO5 allowed us to propose that an oxo(phthalocyani-
ne)iron(IV) complex, is probably responsible of these PAH
oxidations. Consequently, taking in consideration the pre-
viously reported mechanistic studies on quinone cleavage
by the same FePcS/H2O2 system, two kinds of the active
species can be generated from FePcS in the presence of
H2O2: (i) a nucleophilic peroxoiron complex PcSFeIIIOOϪ
responsible for double carbonϪcarbon bond cleavage in the
deep oxidation of trichlorophenol[15][16] and (ii) an electro-
philic PcSFeIVϭO complex capable of oxidizing aromatic
hydrocarbons by the formation of a σ adduct. This versa-
tility makes (tetrasulfophthalocyanine)iron in association
with hydrogen peroxide a suitable catalytic method for the
chemical degradation of different pollutants.
Preparation of 9,10-Dideuterioanthracene: The procedure was ad-
apted from ref.[49]. Dibromoanthracene (1.008 g, 3 mmol) and Li-
AlD4 (504 mg, 12 mmol) were added under nitrogen to 15 ml of
anhydrous dimethoxyethane. The reaction mixture, contained in a
50-ml round-bottom single-neck flask and maintained under nitro-
gen, was partly submerged in a ultrasonic laboratory cleaner Bran-
son 2200 (working frequency 47 kHz, 205 W) at 35Ϫ40°C. The
reaction was monitored by GC-MS. After 13 h, all dibromoanthra-
cene was reduced and the reaction mixture contained 9,10-dideut-
erioanthracene accompanied by minor amount of 9-bromo[D1]-
anthracene.The reaction mixture was cooled at Ϫ40°C and 1 ml of
D2O (99%) was slowly added. The temperature was allowed to re-
turn to 20°C and hydrolysis of the excess of LiAlD4 was performed
for 30 min. Then the resulting mixture was pourred into 30 ml of
water and extracted with diethyl ether (4 ϫ 40 ml). The combined
organic extracts were washed with saturated NaCl solution, dried
with Na2SO4 and concentrated. The white precipitate was sepa-
rated by filtration and recrystallized from diethyl ether to give 148
mg (27% yield) of the pure 9,10-dideuterioanthracene material. The
isotopic purity was determined by NMR and GC-MS methods.
The material contained 72.8% of 9,10-dideuterioanthracene and
27.2% of 9-deuterioanthracene.
General Catalytic Procedure for PAH Oxidation: Oxidation of
anthracene was carried out in a mixture of acetonitrile/50 m phos-
phate buffer at pH ϭ 7 (3:1, v/v) at 20°C under aerobic conditions.
Oxidations of phenanthrene and benzo[a]pyrene were performed in
acetonitrile/33.3 m phosphate buffer at pH ϭ 4.6 (1:1, v/v). The
reaction mixture (2 ml) contained 2 µmol of substrate introduced
as a 4 m acetonitrile solution and 0.074 or 0.1 µmol of catalyst,
A. S. is indebted to ELF-Atochem for a postdoctoral fellowship.
´
The authors are grateful to Jean-Louis Seris (ELF-Atochem, Lacq)
and Prof. Jean Bernadou (LCC-CNRS) for fruitful discussions.
Experimental Section
Instrumentation: Gas chromatography-mass spectrometry (GC-
MS) was performed with a Hewlett-Packard 5890 instrument using introduced as a 1.28 m solution in water, and the required
electron-impact ionization at 70 eV. The carrier gas for GC-MS amounts of acetonitrile and stock buffer solution to obtaine the
was He, and a 12 m ϫ 0.2 mm HL-1 (Crosslinked Methyl Silicone desired organic solvent/water ratios. Oxidations were started by ad-
Gum) capillary column was used. Ϫ High performance liquid chro- dition of 10 µmol of oxidant (10 µl of 3.5% H2O2 solution or 3.1
matography (HPLC) analyses were performed with a Waters-486 mg of potassium monopersulfate). The final concentrations of the
Liquid Chromatograph equipped with a µ-Bondapak C18 column,
catalyst, substrate and oxidant were 37 (or 100) µ, 1 m and 5
m, respectively. The reactions were followed by HPLC and GC-
eluent being a mixture of methanol and water (85:15, v/v) at a flow
rate of 1 ml minϪ1, detection at 242 nm for anthracene, 252 nm for MS techniques. Anthraquinone and anthrone were identified by
phenanthrene and 280 nm for benzo[a]pyrene analyses. Ϫ UV/Vis
comparison of its GC-MS behavior (retention time and mass spec-
spectra were recorded with a Hewlett-Packard 8452A diode array trum) with those of an authentic sample. Oxanthrone was identified
spectrophotometer. Ϫ 1H NMR spectra were recorded with Bruker by comparison of its NMR spectrum with that of an authentic
WM 250 spectrometer.
sample.
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281
1279

A. Sorokin, B. Meunier
FULL PAPER
[8]
[9]
H218O Experiments: Reactions were carried out using H218O (98
atom-%) according to the following procedure. A mixture of 25 µl
of 0.5 M phosphate buffer pH ϭ 7 and 29 µl of catalyst stock
solution (1.28 m in water) was dried under vacuum. The dry resi-
due was then dissolved in 250 µl of H218O and 250 µl of
[H10]anthracene and [D10]anthracene stock solutions (50:50 mix-
ture, 4 m total concentration in MeCN) and 500 µl of MeCN
were added. The oxidations were performed under air and under
nitrogen. In the latter case the reaction mixture was submitted to
3 freeze-pump-thaw cycles and the reaction flask was filled up with
nitrogen. The reaction was started by the addition of 5 µl of 3.5%
H2O2 (for FePcS oxidation) or 1.6 mg of KHSO5 (for FeTMPS
oxidation). The substrate conversion was monitored by HPLC and
isotopic compositions of products were analyzed by GC-MS.
M. Juaristi, J. M. Aizpurua, B. Lecea, C. Palomo, Can. J. Chem.
1984, 62, 2941Ϫ2944.
E. V. Dehmlov, J. K. Makrandi, J. Chem. Research (S) 1986,
32Ϫ33.
[10]
[11]
D. Villemin, M. Ricard, Nouv. J. Chem. 1984, 8, 185Ϫ189.
W. Adam, P. A.Ganeshpure, Synthesis 1993, 280Ϫ282.
[12] [12a]
R. Song, A. Sorokin, J. Bernadou, B. Meunier, J. Org.
Chem. 1997, 62, 673Ϫ678. Ϫ [12b] G. J. Harden, M. M. Coombs,
J. Chem. Soc., Perkin Trans. 1 1995, 3037Ϫ3042.
T. Higuchi, C. Satake, M. Hirobe, J. Am. Chem. Soc. 1995,
117, 8879Ϫ8880.
[13]
[14]
S. V. Barkanova, V. M. Derkacheva, O. V. Dolotova, V. D. Li,
V. M. Negrimovsky, O. L. Kaliya, E. A. LukЈyanets, Tetra-
hedron Lett. 1996, 37, 1637Ϫ1640.
[15] [15a]
A. Sorokin, J.-L. Se´ris, B. Meunier, Science 1995, 268,
[15b]
1163Ϫ1166. Ϫ
B. Meunier, A. Sorokin, Acc. Chem. Soc.
1997, 30, 470Ϫ476.
[16] [16a] A. Sorokin, B. Meunier, Chem. Eur. J. 1996, 2, 1308Ϫ1317.
[16b]
Determination of Isotopic Composition of Substrates and Prod-
ucts: Mass spectra of [H10]anthracene and [D10]anthracene show
very strong molecular peaks accompanied with little fragmen-
tations. [H10]anthracene: MS; m/z (%): 178 (100) [Mϩ], 176 (18) [M
Ϫ 2 H]ϩ, 152 (8) [M Ϫ C2H2]ϩ, 89 (13) [M Ϫ C7H5]ϩ. [D10]anthra-
cene: MS; m/z (%): 188 (100) [Mϩ], 184 (13) [M Ϫ 2 D]ϩ, 160 (9)
[M Ϫ C2D2]ϩ, 94 (17) [M Ϫ C7D5]ϩ. So, under conditions of GC-
MS analysis, within experimental error, there is no isotope effect on
fragmentation and isotopic compositions of [H10]anthracene and
[D10]anthracene can be determined from the intensities of corre-
sponding molecular peaks. The mass spectra of phenanthrene and
products of oxidation of anthracene and phenanthrene showed the
same features and their isotopic compositions were determined
from the intensities of the molecular peaks of deuterated and non-
deuterated compounds.
Ϫ
A. Sorokin, S. De Suzzoni-Dezard, D. Poullain, J.-P.
Noel, B. Meunier, J. Am. Chem. Soc. 1996, 118, 7410Ϫ7411. Ϫ
¨
[16c]
A. Hadasch, A. Sorokin, A. Rabion, B. Meunier, New J.
Chem. 1998, 45Ϫ51.
[17]
[18]
[19]
A. Sorokin, L. Fraisse, A. Rabion, B. Meunier, J. Mol. Cat.
1997, 117, 103Ϫ114.
R. J. Lorentzen, J. Caspary, S. A. Lesko, P. O. TsЈo, Biochemis-
try 1975, 14, 3970Ϫ3977.
L. Melander, J. M. Saunders, Reaction Rates of Isotopic Mol-
ecules, Wiley, New York, 1980, p. 95.
[20] [20a]
J. Bernadou, A.-S. Fabiano, A. Robert, B. Meunier, J. Am.
Chem. Soc. 1994, 116, 9375Ϫ9376. Ϫ [20b] M. Pitie´, J. Bernadou,
[20c]
B. Meunier, J. Am. Chem. Soc. 1995, 117, 2935Ϫ2936. Ϫ
R. J. Balahura, A. Sorokin, J. Bernadou, B. Meunier, Inorg.
Chem. 1997, 36, 3488Ϫ3492.
[21] [21a]
W. Nam, J. S. Valentine, J. Am. Chem. Soc. 1993, 115,
[21b]
1772Ϫ1778. Ϫ
1997, 119, 1916Ϫ1922. Ϫ
J. Am. Chem. Soc. 1997, 119, 6269Ϫ6273.
K. A. Lee, W. Nam, J. Am. Chem. Soc.
[21c]
J. T. Groves, J. Lee, S. S. Marla,
B. Meunier, Chem. Rev. 1992, 92, 1411Ϫ1456.
[22]
KIE determinations based on [H10]anthracene/9,10-dideut-
erioanthracene oxidations were performed taking into account ad- [23] [23a] T. J. McMurry, J. T. Groves, In Cytochrome P-450 Structure,
Mechanism and Biochemistry (Ed.: P. R. Ortiz de Montellano),
mixture of monodeuterated anthracene, the presence of [M ϩ 1]
peaks due to 13C natural content and [M Ϫ 1] and [M Ϫ 2] frag-
mentations according to the following formalism:
[23b]
Plenum, New York, 1986; pp. 1Ϫ28. Ϫ
M. Newcomb, M.-
H. Le Tadic-Biadatti, D. L. Chestney, E. S. Roberts, P. F. Hol-
lenberg, J. Am. Chem. Soc. 1995, 117, 12085Ϫ12091.
R. A. Poirier, Y. Wang, K. C.Westaway, J. Am. Chem. Soc. 1994,
116, 2526Ϫ2533.
[24]
[25]
I178 ϭ [D0]-A ϩ FMϪ1·[D1]-A ϩ FMϪ2·[D2]-A
I179 ϭ [D1]-A ϩ FMϩ1·[D0]-A ϩ FMϪ1·[D2]-A
I180 ϭ [D2]-A ϩ FMϩ1·[D1]-A
(2)
(3)
(4)
S. Wolfe, C.-K. Kim, J. Am. Chem. Soc. 1991, 113, 8056Ϫ8061.
[26] [26a] P. A. Nielsen; S. S. Glad, F. Jensen, J. Am. Chem. Soc. 1996,
[26b]
118, 10577Ϫ10583. Ϫ
Soc. 1997, 119, 227Ϫ232.
S. S. Glad, F. Jensen, J. Am. Chem.
where I178, I179, I180 were intensities of peaks at m/z 178, 179 and
180, respectively; [D0]-A, [D1]-A, [D2]-A were percentages of corre-
sponding isotopomers of anthracene; FMϩ1, FMϪ1, FMϪ2 were rela-
tive intensities of [M ϩ 1], [M Ϫ 1] and [M Ϫ 2] peaks. The latter
intensities were determined in separate GC-MS experiments to be
FMϩ1 ϭ 0.15, FMϪ1 ϭ 0.11, FMϪ2 ϭ 0.18. KIEs were calculated
from the data obtained on conversions and the percentages of
[H10]anthracene and 9,10-dideuterioanthracene in an initial reac-
tion mixture and after several minutes when the substrate conver-
sion was near 50%.
[27] [27a]
R. P. Hanzlik, G. O. Shearer, Biochem. Pharm. 1978, 27,
[27b]
1441Ϫ1444 and references therein. Ϫ
A. K. Bhattacharjee,
M. K. Mahanti, Gazz. Chim. Ital. 1984, 114, 337Ϫ340.
G. W. Griffin, K. A. Horn, J. Am. Chem. Soc. 1987, 109,
4919Ϫ4926 and references therein.
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
G. Vassilikogiannakis, M. Orfanopoulos, J. Am. Chem. Soc.
1997, 119, 7394Ϫ7395.
K. R. Korzekwa, D. C. Swinney, W. F. Trager, Biochemistry
1989, 28, 9019Ϫ9027.
R. P. Hanzlik, K.-H. J. Ling, J. Am. Chem. Soc. 1993, 115,
9363Ϫ9370.
G. Tian, S.-Y. Chen, K. L. Facchine, S. R. Prakash, J. Am.
Chem. Soc. 1995, 117, 2369Ϫ2370.
K. A. Kurtz, P. F. Fitzpatrick, J. Am. Chem. Soc. 1997, 119,
[1]
C. E. Cerniglia, Biodegradation 1992, 3, 351Ϫ368.
K. E. Hammel in Degradation of Environmental Pollutants by
[2]
1155Ϫ1156.
Microorganisms and their Metalloenzymes. Metal Ions in Bio-
logical Systems (Eds.: H. Sigel, A. Sigel), Marcel Dekker, New
York, 1992, vol. 28, p. 41.
J. March, Advanced Organic Chemistry: Reactions, Mechanisms
and Structure, 4th ed., Wiley, New York, 1992, p. 20.
L. Melander, R. E. Carter, Acta Chem. Scand. 1964, 18,
1138Ϫ1149.
[3]
S. D. Haemmerli, M. S. A. Leisola, D. Sanglard, A. Fiechter, J.
Biol. Chem. 1986, 261, 6900Ϫ6903.
[4]
K. Mislow, R. Graeve, A. J. Gordon, G. H. Wahl, Jr., J. Am.
Chem. Soc. 1963, 85, 1199Ϫ1200.
K. E. Hammel, B. Kalyanaraman, T. K. Kirk, J. Biol. Chem.
1986, 261, 16948Ϫ16952.
[5]
R. W. Nagorski, H. Slebocka-Tilk, R. S. Brown, J. Am. Chem.
Soc. 1994, 116, 419Ϫ420.
J. D. Lipscomb, A. M. Orville in Degradation of Environmental
Pollutants by Microorganisms and their Metalloenzymes. Metal
Ions in Biological Systems (Eds.: H. Sigel, A. Sigel), Marcel
Dekker, New York, 1992, vol. 28, p. 243.
[38] [38a]
D. Rabinovich, G. Parkin, J. Am. Chem. Soc. 1993, 115,
353Ϫ354. Ϫ [38b] R. H. Schultz, A. A. Bengali, M. J. Tauber, B.
H. Weiller, E. P. Wasserman, K. R. Kyle, C. B. Moore, R. G.
Bergman, J. Am. Chem. Soc. 1994, 116, 7369Ϫ7377.
M. J. Tanner, M. Brookhart, J. M. DeSimone, J. Am. Chem.
Soc. 1997, 119, 7617Ϫ7618.
[6]
D. A. Lane in Chemical Analysis of Polycyclic Aromatic Com-
pounds (Ed.: T. Vo-Dinh), Wiley, New York, 1989, p. 31.
[7]
[39]
For a review on chemistry of polyarenes see: R. G. Harvey,
Polycyclic Aromatic Compounds, Wiley, New York, 1997, p. 96.
1280
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281

Polycyclic Aromatic Hydrocarbons Catalyzed by Iron Tetrasulfophthalocyanine FePcS
FULL PAPER
[40]
[41]
[45]
D. R. Evans, T. Drovetskaya, R. Bau, C. A. Reed, P. D. W.
Boyd, J. Am. Chem. Soc. 1997, 119, 3633Ϫ3634.
Z. Xie, R. Bau, C. A. Reed, Angew. Chem. Int. Ed. Engl. 1994,
33, 2433Ϫ2434.
S. A. Carlson, D. H. Hercules, Anal. Chem. 1973, 45,
1794Ϫ1799.
[46]
[47]
[48]
P. P. Fu, R. G. Harvey, Chem. Rev. 1978, 78, 317Ϫ361.
J. H. Weber, D. H. Busch, Inorg. Chem. 1965, 4, 469Ϫ471.
P. Hoffmann, G. Labat, A. Robert, B. Meunier, Tetrahedron
Lett. 1990, 31, 1991Ϫ1994.
[42]
[43]
See ref.[7], p. 26.
T. R. Criswell, B. H. Klanderman, J. Org. Chem. 1974, 39,
770Ϫ774.
[49]
B. H. Han, P. Boudjouk, Tetrahedron Lett. 1982, 23,
1643Ϫ1646.
[44]
K. Brederesk, E. F. Sommermann, Tetrahedron Lett. 1966, 41,
5009Ϫ5014.
[98126]
Eur. J. Inorg. Chem. 1998, 1269Ϫ1281
1281