Key role of the phosphate buffer in the H2O2 oxidation of aromatic pollutants catalyzed by iron tetrasulfophthalocyanine

Article abstract of DOI:10.1006/jcat.2001.3269

The non-innocent role of the phosphate buffer has been established in the H₂O₂ oxidative decomposition of 2,4,6-trichlorophenol (TCP), a benchmark pollutant, catalyzed by iron(III) tetrasulfophthalocyanine (FePcS). The catalytic oxidation of several other substrates (3,5-dichloroaniline, tetrachlorocatechol, di-tert-butylcatechol and catechol itself) has been carried out, also demonstrating a crucial influence of the phosphate buffer in the decomposition of the chlorinated substrates. Three hypotheses have been studied: modification of the ionic strength, formation of a peroxyphosphate species, or catalysis by a peroxyphosphate-FePcS complex. Supports for the latter proposal have been obtained from several experimental results and attempts have been made to characterize this putative catalytic intermediate. This intermediate derivative has also been generated from the reaction of FePcS with peroxymonophosphoric acid (PMPA) and its catalytic activity has been checked on the decomposition of TCP in different reaction mixture. A short mechanistic study has allowed different reaction pathways to be proposed, dependent on the active species implicated.

Full text of DOI:10.1006/jcat.2001.3269

Journal of Catalysis 202, 177–186 (2001)
doi:10.1006/jcat.2001.3269, available online at http://www.idealibrary.com on
Key Role of the Phosphate Buffer in the H2O2 Oxidation of Aromatic
Pollutants Catalyzed by Iron Tetrasulfophthalocyanine
Muriel Sanchez, Anke Hadasch,1 Rainer T. Fell,2 and Bernard Meunier3
Laboratoire de Chimie de Coordination du CNRS, 205 route de Narbonne, 31077 Toulouse cedex 4, France
Received March 16, 2001; accepted April 30, 2001
erties with different chemical applications, for example in
the well-known Merox industrial process for the catalytic
oxidation of mercaptans in petroleum distillates (7). In the
field of catalytic oxidations, we have recently shown that
water-soluble metallophthalocyanines were also suitable as
catalysts for the degradation of recalcitrant contaminants
dissolved in water (8–14). For example, iron(III) tetrasul-
fophthalocyanine (FePcS) (Fig. 1) when activated by hydro-
gen peroxide, was able to perform the oxidative degrada-
tion of several pollutants such as polychlorinated phenols
and anilines or polycondensed aromatics (8–14). All the
reactions described were carried out in a phosphate buffer
medium initially employed to maintain the solution in an
optimized pH value for the catalytic conversion. Here, we
report the non-innocent behavior of the phosphate buffer
in the oxidation of some of these aromatic pollutants, in
particular the chlorinated derivatives, for which an active
species derived from the phosphate ions could be involved
in the catalytic oxidative process.
The non-innocent role of the phosphate buffer has been estab-
lished in the H2O2 oxidative decomposition of 2,4,6-trichlorophenol
(TCP), a benchmark pollutant, catalyzed by iron(III) tetrasulfoph-
thalocyanine (FePcS). The catalytic oxidation of several other
substrates (3,5-dichloroaniline, tetrachlorocatechol, di-tert-butyl-
catechol and catechol itself ) has been carried out, also demonstrat-
ing a crucial influence of the phosphate buffer in the decomposition
of the chlorinated substrates. Three hypotheses have been stud-
ied: modification of the ionic strength, formation of a peroxyphos-
phate species, or catalysis by a peroxyphosphate–FePcS complex.
Supports for the latter proposal have been obtained from several
experimental results and attempts have been made to character-
ize this putative catalytic intermediate. This intermediate deriva-
tive has also been generated from the reaction of FePcS with per-
oxymonophosphoric acid (PMPA) and its catalytic activity has
been checked on the decomposition of TCP in different reaction
mixture. A short mechanistic study has allowed different reac-
tion pathways to be proposed, dependent on the active species
c
implicated.
2001 Academic Press
Key Words: iron(III) tetrasulfophthalocyanine; phosphate buffer,
peroxyphosphate; pollutant oxidation; 2,4,6-trichlorophenol;
3,5-dichloroaniline; tetrachlorocatechol; 3,5-di-tert-butylcatechol;
catechol.
EXPERIMENTALS AND METHODS
Methods
INTRODUCTION
The conversions of TCP, 3,5-DCA, DTBC, and DCBQ
were monitored by HPLC (Waters) equipped with a
-Bondapak C18 column. The eluents were mixtures of
methanol/ammonium acetate buffer 50 mM pH 4 (7/3,
v/v) for TCP and DCBQ, acetonitrile/phosphate buffer
50 mM pH 7 (55/45, v/v) for 3,5-DCA and methanol/water
(8/2, v/v) for DTBC. The flow rates were 1 ml/min and
the detection was performed at 220 nm. The conver-
sion of catechol and TCC were monitored by UV–visible
spectroscopy with a Hewlett Packard 8452A spectropho-
tometer. 1H and 31P NMR spectra were recorded on
Bruker AC 200 and AM 250 spectrometers. The sub-
strates were purchased from Aldrich. Hydrogen peroxide
was supplied from Acros as a 35 wt% aqueous solution.
Curox (2KHSO5 KHSO4 K2SO4) was a gift from Peroxid-
Chemie, Germany. All solvents used were of analytical
grade and milliQ-water wasalwaysused to prepare aqueous
Phthalocyanines (abbreviated as Pc) are large macro-
cyclic molecules with an extended, delocalized -electron
system which confers to them a large number of unique
properties (1, 2). These molecules have found applications
in fields as diverse as chemical sensors, electrochromism,
nonlinear optics, liquid crystals, and photodynamic cancer
therapy, in addition to their traditional use as dyes (1–6).
When substituted by an appropriate metal, these ph-
thalocyanine derivatives can exhibit interesting redox prop-
1
´
Current address: L’OREAL, Centre de Chevilly, 188 rue Paul
Hochart, 94152 Chevilly-Larue cedex, France.
2 Current address: Clariant, Ludwig-Hermann-Str. 100, 86368
Gersthofen, Germany.
3 To whom correspondence should be addressed. E-mail: bmeunier@
lcc-toulouse.fr.
177
0021-9517/01 $35.00
c
Copyright
2001 by Academic Press
All rights of reproduction in any form reserved.

178
SANCHEZ ET AL.
vacuum and dissolved in deuterated dimethyl sulfoxide for
NMR analysis. Then, 3 l of CHCl3 (32.5 mol) was added
as internal standard to quantify the oxidation products. 1H
NMR ([D6] DMSO): (ppm) 7.44 (s, 2H, DCBQ, 21% ),
7.62–8.10 (br, 2H, coupling products, 13% ).
General Catalytic Procedure for 3,5-DCA Degradation
The reaction mixture containing 4.0 mol of 3,5-dichlo-
roaniline (80 l of a 50 mM stock solution in acetonitrile),
buffer pH 7 (400 l of a 500 mM stock solution), 0.12 mol
of FePcS (1.5 ml of a 80 M aqueous solution) was adjusted
to 4 ml with 240 l of CH3CN and 1.7 ml of water. The cat-
alytic oxidation was initiated by the addition of 24.0 mol
of H2O2 (20 l of a 1.2 M aqueous solution), stirred at room
temperature and monitored by HPLC.
FIG. 1. Structure of the iron(III) tetrasulfophthalocyanine (FePcS);
only one of the four possible regioisomers is depicted here.
solutions. Iron(III) tetrasulfophthalocyanine FePcS was
prepared according to previously published modifications
of the method of Weber and Busch (9, 15) and peroxy-
monophosphoric acid PMPA was synthesized following the
procedure described by Springer et al. (16).
General Catalytic Procedure for TCC Degradation
The reaction mixture of 2 ml contained 20.0 mol of
tetrachlorocatechol (5.14 mg in 1 ml of CH3CN), buffer
pH 7 (500 l of a 500 mM stock solution), 0.8 mol of
FePcS (500 l of a 1.6 mM aqueous solution). The catalytic
oxidation was initiated by the addition of 120 mol of H2O2
(10 l of a 35 wt% aqueous solution), stirred at room tem-
perature and monitored by UV–vis (464 nm).
General Catalytic Procedure for TCP Degradation
The reaction mixture contained 10.0 mol of 2,4,6-
trichlorophenol (2.5 ml of a 4 mM stock solution in ace-
tonitrile), 500.0 mol of buffer pH 7 (1 ml of a 500 mM
stock solution), 0.4 mol of FePcS (50 l of a 8 mM aque-
ous stock solution), and water until 10 ml. The catalytic
oxidation was initiated by the addition of an aqueous so-
lution of the oxidant (50 l of a 1.2 M aqueous solution of
H2O2, 15.4 mg of KHSO5 in 100 l of water or 180 l of a
3.8 wt% solution of PMPA in HNO3 65 wt% ). When using
PMPA as oxidant, the buffer concentration was increased
until 365 mM by adding 7.3 ml of a 500 mM Tris/CH3COOH
buffer solution and without diluting with water. The reac-
tion mixture was stirred at room temperature and aliquots
were taken at defined times, diluted with eluent to stop the
oxidation reaction, and analyzed by HPLC.
General Catalytic Procedure for DTBC Degradation
The reaction mixture of 2 ml contained 20.0 mol of 3,5-
di-tert-butylcatechol (4.45 mg in 500 l of CH3CN), buffer
pH 7 (500 l of a 500 mM stock solution), 0.8 mol of FePcS
(500 l of a 1.6 mM aqueous solution). The catalytic oxi-
dation was initiated by the addition of 120 mol of H2O2
(10 l of a 35 wt% aqueous solution), stirred at room tem-
perature and monitored by HPLC.
General Catalytic Procedure for Catechol Degradation
The reaction mixture of 2 ml contained 20.0 mol of cat-
echol (500 l of a 40 mM aqueous solution), buffer pH 10
(500 l of a 500 mM stock solution), 0.8 mol of FePcS
(500 l of a 1.6 mM aqueous solution). The catalytic oxi-
dation was initiated by the addition of 120 mol of H2O2
(10 l of a 35 wt% aqueous solution), stirred at room tem-
perature and monitored by UV–vis (500 nm).
1H NMR Analysis of the Oxidation Products of TCP
Catalyzed by the FePcS/PMPA System
A catalytic oxidative degradation of TCP was performed
with 32.9 mg of TCP (167.0 mol) dissolved in 42 ml of
CH3CN, 122 ml of a 500 mM Tris/CH3COOH buffer solu-
tion pH 7 (61.0 mmol), 835 l of a 8 mM FePcS aqueous so-
lution (6.7 mol), and 3.0 g of a 3.8 wt% solution of PMPA
in HNO3 65 wt% (1.0 mmol). The reaction mixture was
General Catalytic Procedure for DCBQ Degradation
The reaction mixture containing 10.0 mol of 2,6-
stirred 30 min at room temperature and the excess perox- dichloro-1,4-benzoquinone (250 l of a 40 mM stock so-
ide was neutralized by reduction with 174.0 mg of Na2S2O4 lution in CH3CN), buffer pH 7 (1 ml of a 500 mM stock
(1 mmol). The solution was adjusted to pH 1 with 7.5 ml of solution), 0.4 mol of FePcS (50 l of a 8 mM aqueous so-
conc. HCl saturated with NaCl. The organic products were lution) was adjusted to 10 ml with 2.25 ml of CH3CN and
then extracted with diethyl ether (3 150 ml) and the com- 6.45 ml of water. The catalytic oxidation was initiated by the
bined organic layers were dried under sodium sulfate. After addition of 60 mol of H2O2 (50 l of a 1.2 M aqueous solu-
evaporation of diethylether, the residue was dried under tion), stirred at room temperature and monitored byHPLC.

PHOSPHATE IN THE FePcS/H2O2 OXIDATION OF POLLUTANTS
179
TABLE 1
RESULTS AND DISCUSSION
Influence of the Buffer on the Oxidation of TCP Catalyzed
by FePcS/H2O2
Influence of the Phosphate Buffer on the Oxidative
Degradation of 2,4,6-Trichlorophenol
Catalyzed by FePcS/H2O2
Conversion at
60 min (% )
Run
Buffer
pH
2,4,6-Trichlorophenol (TCP) is one of the major chlori-
nated phenols produced by paper mills in the delignification
of wood pulp by chlorine bleaching (17). Because of its slow
biodegradation bymicroorganisms, TCP isa well-known re-
calcitrant pollutant (18). On the other hand, we have shown
in previouspapersthat TCP could be degraded in a fewmin-
utes by using 4 mol% of FePcS as catalyst, activated by the
“green” oxidant H2O2 (9, 11–14). This efficient catalytic ox-
idation generates aromatic ring cleavage products (mainly
C4 diacids) and oxidative coupling products (Fig. 2).
1
2
3
4
5
6
7
Phosphate
—
7
5–6
7
7
7
99
4
5
6
7
—
Tris/HCl
Tris/H2SO4
Tris/CH3COOH
Tris/H3PO4
7
7
6
91
Note. 1 mM TCP, 50 mM buffer (buffer/CH3CN = 1/3 v/v), 0.04 mM
FePcS (FePcS/TCP = 4% ), 6 mM H2O2.
The pH of the medium was stabilized to pH 7 with a
phosphate buffer in order to get mainly the more oxidiz- Surprisingly, no decomposition of TCP occurred with any
able phenolate form of TCP in solution ( pKa = 6.2) (19). of these buffers (Table 1, runs 4–6). On the other hand,
To facilitate a possible large scale application, we tried to when this buffer was acidified with o-phosphoric acid, the
eliminate the phosphate buffer. The results are summarized reactivity of TCP was entirely restored with levels of con-
in Table 1. Without buffering the solution, the conversion version similar to those obtained with a phosphate-buffered
of TCP falls to 4% after 60 min (Table 1, run 2) compared medium (Table 1, run 7). So, the presence of phosphate ions
to full conversion using phosphate buffer (Table 1, run 1). seems to be absolutely necessary to perform the oxidative
The replacement of phosphate buffer by the addition of decomposition of TCP catalyzed by FePcS/H2O2.
sodium hydroxide to the reaction mixture until the optimal
Subsequently, quantification of this strong ion influence
pH value of 7 was reached did not restore the decomposi- was undertaken through a kinetic study of the oxidation of
tion reaction of TCP (Table 1, run 3). The influence of pH is TCP with different phosphate buffer concentrations. The
unlikely to be the unique reason behind the lack of reactiv- experiments were carried out starting from a 1 mM so-
ity of TCP in an unbuffered medium (pH 5). However, since lution of TCP in an acetonitrile/buffered water mixture
the addition of sodium hydroxide did not stabilize the pH (v/v = 1/3) containing 4 mol% of FePcS catalyst. Note that
duringthe reaction in which diacidsmust be formed, supple- the presence of CH3CN is not essential to (but helps) the
mentary experiments were carried out with other buffers. solubilization of the poorly water-soluble TCP substrate
We used tris(hydroxymethyl)aminomethane (Tris) buffer and to the monomerization of the iron tetrasulfophthalo-
acidified with diverse acids: HCl, H2SO4, and CH3COOH. cyanine. The buffer, with a final concentration of 100 mM,
FIG. 2. Product distribution from the H2O2 oxidation of TCP catalyzed by FePcS.

180
SANCHEZ ET AL.
gives only 5% 3,5-DCA conversion after 60 min, whereas,
in a phosphate buffer, 70% ofthe substrate wasconverted in
the same time (Fig. 4, run 1), and the reaction rate slows to
give only 13% conversion when using a Tris buffer acidified
with sulfuric acid (Fig. 4, run 3). Again, for the degrada-
tion of 3,5-DCA catalyzed by the FePcS/H2O2 system, an
influence of the phosphate ions was observed.
Catechols. Subsequently, we studied the degradation of
catechols to evaluate whether this phenomenon is com-
mon to all the catalytic oxidation reactions initiated by
the iron tetrasulfophthalocyanine. Indeed, this complex has
also previously been proved to be able to catalyze the ox-
idation of diverse catechols, which are not pollutants but
tannin models used to test the efficiency of potential addi-
tives in future washing powders (23).
Figures 5, 6, and 7 illustrate the degradation products of
catechol and its tetrachloro and di-tert-butyl derivatives by
the FePcS/H2O2 system (24). These substrates were again
subjected to the same catalytic system in media buffered
either by the phosphate mixture KH2PO4/Na2HPO4 or by
Tris acidified with sulfuric acid to the desired pH value. Ac-
cording to the table in Fig. 5, the decomposition reaction of
tetrachlorocatechol(TCC) isveryfast, reachingcompletion
in less than 1 min with 4 mol% of catalyst in the phosphate
buffer (pH 7) (Fig. 5, run 1). Using the Tris/H2SO4 buffer,
FIG. 3. Influence of the phosphate buffer concentration on the con-
version of TCP catalyzed by FePcS/H2O2.
consisted of a mixture of phosphate buffer pH 7 and Tris/
CH3COOH buffer pH 7 with different molar ratios. The
reaction was initiated by the addition of 6 equivalents of
H2O2 and, 10 min later, the reaction mixture was analyzed
by HPLC (Fig. 3). According to Fig. 3, the concentration of
phosphate ions directly influences the conversion of TCP
in a linear manner. Doubling the phosphate buffer concen-
tration leads to a twofold conversion of TCP suggesting a
reaction order of one with respect to the phosphate con-
centration.
This influence of the phosphate buffer on the TCP degra-
dation is really surprising but it is not only confined to TCP
as this effect has been also observed in degradation studies
with other substrates as described in the next paragraph.
Influence of the Phosphate on the Conversion
of Different Substrates
Dichloroanilines. We have seen the influence of the
buffer on the decomposition of a pollutant, 2,4,6-
trichlorophenol, catalyzed by FePcS/H2O2. Dichloroani-
lines (DCA) are other polychlorinated pollutants used as
precursors on a large scale in the industrial synthesis of
pesticides, plastics, and dyes (20, 21). 3,5-Dichloroaniline
(3,5-DCA), the most toxic, is additionally a precursor in
the industrial synthesis of azo dyes (22). In a recent work,
we have shown that the FePcS/H2O2 catalytic system was
able to oxidize 3,5-DCA leading to the formation of (i) ring
cleavage products, (ii) compounds resulting from the oxida-
tion of the amine function, and (iii) species resulting from
oxidative coupling (Fig. 4) (8).
The experiments described were carried out at pH 7 in a
phosphate buffer medium. In view of the results described
above for TCP, we also tried to oxidize 3,5-DCA by the
FePcS/H2O2 system without buffering or with a Tris buffer.
In the table of Fig. 4, run 2, the nonbuffered experiment
FIG. 4. Influence of the buffer on the conversion of 3,5-DCA
catalyzed by FePcS/H2O2.

PHOSPHATE IN THE FePcS/H2O2 OXIDATION OF POLLUTANTS
181
FIG. 7. Influence of the buffer on the conversion of catechol with
FePcS/H2O2.
FIG. 5. Influence of the buffer on the conversion of TCC catalyzed
by FePcS/H2O2.
sition ofchlorinated aromatic compounds, which are known
to be more difficult to oxidize because of the electron-
withdrawing effect of the halogen atoms. For the latter, the
degradation cannot be achieved by the FePcS/H2O2 system
under “classical” conditions but necessitates the presence
of phosphate ions, which seem to play a key role. In order to
verify this hypothesis, we employed an oxidant more pow-
erful than H2O2 for the catalytic degradation of TCP, itself
chosen as a representative chlorinated aromatic substrate.
the catalytic oxidation is slower (<30% in 1 min) and the
conversion remainslow(<30% after 15min) indicatingthat
the decomposition of this substrate is also sensitive to the
presence of phosphate ions (Fig. 5, run 2).
We were very surprised to observe that the other catechol
derivatives were rapidly converted to degradation products
both in phosphate buffer and Tris buffer (Figs. 6 and 7). The
catalytic oxidation reactions were very fast using 4 mol%
of FePcS, activated with H2O2, and give rise to the same
degradation products whatever the buffer.
Utilization of KHSO5 as an Alternative Oxidant for the
Catalytic Decomposition of TCP with FePcS
From the various substrates tested under a variety of con-
ditions, it becomes evident that the phosphate buffer only
appears to play an important role in the catalytic decompo-
Potassium monoperoxysulfate (KHSO5), available com-
mercially as the triple salt 2KHSO5 KHSO4 K2SO4, is
used for example in Baeyer–Villiger oxidations (25) or
for the synthesis of highly reactive dioxiranes (26). It is
also an effective oxidant for the activation of metalloph-
thalocyanines and metalloporphyrins (8–14, 27). As desc-
ribed in previous papers, in phosphate buffer, the catalytic
oxidation of all the aromatic substrates cited here pro-
ceeds much faster using KHSO5 instead of H2O2 (8–14).
This raises the question of what is happening when using
KHSO5 as oxidant in a medium containing no phosphate
ions?
2,4,6-Trichlorophenol is fully converted by the FePcS/
KHSO5 system after 60 min in a phosphate buffer at pH 7
(Table 2, run 1). The same result is observed in a Tris/H2SO4
buffer at pH 7 (Table 2, run 2) or even in the absence of a
buffer (Table 2, run 3), despite the low pH of the mixture
(pH 2) caused by the acidity of KHSO5 ( pKa = 1). So, in
contrast to the FePcS/H2O2 catalytic system, activation of
the metallophthalocyanine with the powerful KHSO5 ox-
idant does not require the presence of phosphate ions for
the degradation of recalcitrant chlorinated substrates.
FIG. 6. Influence of the buffer on the conversion of DTBC catalyzed
by FePcS/H2O2.

182
SANCHEZ ET AL.
TABLE 2
was added). The reaction conditions were as follows:
TCP (1 mM) in an acetonitrile/phosphate buffer mixture
1/3 v/v (final concentration of phosphate buffer: 50 mM,
pH 7) and H2O2 (6 mM). No conversion of TCP was
observed (Table 3, run 2). This reaction was also fol-
lowed by 31P NMR spectroscopy; the chemical shift cor-
responding to the phosphate buffer ( = 6.7 ppm) re-
mained unchanged suggesting that peroxyphosphate was
not formed under these conditions. According to 31P NMR
spectroscopy, no reaction occurred between 1 equivalent
of phosphate buffer and 1 equivalent of H2O2 even af-
ter 1 h at room temperature. In contrast, this reaction
has been observed to be catalyzed by the iron tetrasul-
fophthalocyanine complex. Hence, FePcS was added to
the NMR tube but no new signal appeared in the spec-
trum (note: the commercial peroxydiphosphate K4P2O8
resonates at = 9.5 ppm). To the best of our knowl-
edge, there are no reports in the literature concerning
the formation of a peroxyphosphate salt from the reac-
tion of hydrogen peroxide with inorganic phosphate com-
pounds. On the other hand, several peroxyphosphorus in-
termediates have been postulated when reacting H2O2
with activated phosphorus compounds. This reaction was
first used in 1956 by Rosenblatt for the fast hydroly-
sis of a common insecticide: paraoxon (O,O-diethyl O-p-
nitrophenylphosphate) (29). More recently, Bunton et al.
described the cleavage of bis(4-nitrophenyl)phosphate and
4-nitrophenylphosphorochloridate by alkaline hydrogen
peroxide (30) while Yang et al. proposed the perhydrol-
ysis of O-ethyl S-(2-diisopropylamino)ethyl methyl phos-
phonothioate (VX agent) as an efficient decontamina-
tion method for this exceedingly toxic nerve agent (31).
In a last example, Kende et al. have activated hydrogen
peroxide using organophosphorus compounds by forming
more active peroxy species able to perform alkene epox-
idation (32). In our case, the phosphate ion is not elec-
trophilic enough to react with H2O2, but could act as
an axial ligand for the cationic iron(III) FePcS complex
(by replacing the initial hydroxyl group). This coordina-
tion of the phosphate to the metal center may enhance
its electrophilicity and thus allow its subsequent reaction
with H2O2. The resulting peroxyphosphate–FePcS com-
plex which can be drawn as the two possible isomeric
forms (a) and (b) would be then involved in the oxida-
tive degradation of the chlorinated substrates (Fig. 9). It
Influence of the Buffer on the Oxidation of TCP Catalyzed
by FePcS Activated with KHSO5 or H2O2
Conversion at 60 min (% )
Run
Buffer
H2O2 (pH)
KHSO5 (pH)
1
2
3
Phosphate
Tris/H2SO4
—
99 (7)
7 (7)
4 (5–6)
100 (7)
100 (7)
100 (2)
Note. 1.0 mM TCP, 50 mM buffer (buffer/CH3CN = 1/3 v/v), 0.04 mM
FePcS (FePcS/TCP = 4% ), 6 mM H2O2 or 5 mM KHSO5.
From all the above observations, it became apparent that
elucidation of the exact role of the phosphate buffer in
the catalytic oxidation of aromatic chlorinated compounds
with the FePcS/H2O2 system was necessary.
Role of the Phosphate Ions in the Catalytic Oxidation
of Aromatic Chlorinated Compounds
with the FePcS/H2O2 System
Ionic strength. As the phosphate buffer is composed of
a mixture of the potassium salt KH2PO4 (25 mM) and the
sodium salt Na2HPO4 (25 mM), a first simple hypothesis
was that a modification of the ionic strength could influ-
ence the course of the reaction where polar compounds
are formed. In run 1, Table 3, the ionic strength of a
Tris/CH3COOH reaction mixture was increased by adding
NaCl (25 mM) and KCl (25 mM). However, no improve-
ment in the conversion of TCP was observed, effectively
eliminating this hypothesis.
Peroxyphosphate hypothesis. According to the results
obtained, the phosphate ions play a direct role in these
catalytic oxidations. This led us to investigate whether the
phosphate ions themselves could form an active species re-
sponsible for the degradation of the substrates. It is known
that some salts like carbonate or borate react with hydro-
gen peroxide to afford hydroperoxy species showing strong
oxidative abilities (28). Peroxycarbonate and peroxyborate
have thus proved to be more efficient than H2O2 for the
activation of metallotetrasulfophthalocyanines in the cata-
lytic decomposition of catechol and its derivatives (24).
Thus, we carried out chemical and analytical investigations
in order to examine a putative intermediate involving the
phosphate moiety, such as a peroxyphosphate (Fig. 8).
First, the oxidative reaction of TCP was tried in
the presence of only phosphate and H2O2 (no FePcS
FIG. 8. Hypothetical formation of a peroxyphosphate from the reac-
tion of phosphate ions with H2O2.
FIG. 9. Formation of the peroxyphosphate–FePcS complex.

PHOSPHATE IN THE FePcS/H2O2 OXIDATION OF POLLUTANTS
TABLE 3
183
Control Reactions
Conversion at
60 min (% )
Run
1
Conditions
Tris/CH3COOH
9
+ NaCl + KCl
Phosphate + H2O2
FePcS + H2O2
2
3
4
9
6
6
FePcS + phosphate
Note. 1 mM TCP, 50 mM buffer (buffer/CH3CN = 1/3 v/v), 0.04 mM
FePcS (FePcS/TCP = 4% ), 25 mM NaCl, 25 mM KCl, 6 mM H2O2.
should be noted that the trans-coordination of a phosphate
moiety and hydrogen peroxide could also be postulated
as an active species but, in contrast to metalloporphyrin-
catalyzed oxidations, no enhancement of the catalytic ac-
tivity has been observed with metallophthalocyanines via a
trans-coordination effect with nitrogen-containing ligands.
FIG. 10. Synthesis and acid dissociation constants of peroxymono-
phosphoric acid (PMPA).
generate the peroxyphosphate–FePcS complex, in situ, by
adding PMPA to the phthalocyanine and subsequently re-
act it with TCP in a Tris-buffered medium. First, a 3.8%
Peroxyphosphate–FePcS complex hypothesis. Several solution of PMPA was prepared according to the proce-
observations support the “peroxyphosphate–FePcS com- dure described by Springer (16) by hydrolyzing tetrapotas-
plex” hypothesis. First, the control reactions showed that sium peroxydiphosphate in a solution of nitric acid in such
the three partners (FePcS + phosphate + H2O2) are indis- a manner that the final acid concentration was 2.2 mol/L.
pensable in ensuring good catalytic activities: the experi- Hydrolysis was conducted at 50 C for 30 min. This freshly
ments(phosphate +H2O2), (FePcS+H2O2), and (FePcS + prepared solution was then added to a reaction mixture
phosphate) did not lead to an oxidation of TCP (Table 3, containing TCP (1 mM) in an acetonitrile/Tris buffer (1/3,
runs 2–4). Finally, the peroxyphosphate–FePcS complex hy- v/v) and FePcS (0.04 mM, i.e., FePcS/TCP = 4% ). The
pothesis might explain the 31P NMR results obtained from final concentrations of Tris buffer and PMPA were 365
the reaction: FePcS + phosphate + H2O2. The absence of and 6 mM, respectively. The conversion of TCP was fol-
new signals in the spectrum could be rationalized: (i) by the lowed by HPLC (Table 4, run 1), which indicated that
fact that even if only a small amount of this complex (but after only 5 min, 89% of TCP had been decomposed, cor-
sufficient to catalyze the reaction) was formed, it will not responding to a catalytic activity of 4.5 cycles per minute.
be observable by NMR because of its paramagnetic char- Moreover, control reactions have shown that PMPA alone
acter and (ii) its lifetime would be too short versus that of was not able to oxidize TCP (Table 4, run 2) even in
the NMR experiment. The peroxyphosphate–FePcS com- the presence of a simple iron salt (Table 4, run 3). We
plex intermediate was not directly observable under reac- also checked that the precursor of PMPA, PDP, was in-
tion conditions tested so far, hence attempts were made to active (Table 4, run 4). This means that the metalloph-
support its involvement.
thalocyanine is really essential and does not just cat-
In the literature, it has been shown that peroxymono- alyze the production of PMPA from phosphate and H2O2.
phosphoric acid (PMPA) could be prepared by acid hy- These observations strongly support the formation of a
drolysis of peroxydiphosphate (PDP) (33) (Fig. 10). Since
its discovery in 1910 by Schmidlin and Massini (34), PMPA
has appeared in several reports concerning the oxidation
TABLE 4
Oxidation of TCP with PMPA or PDP
Conversion of TCP (% )
of diverse organic and inorganic substrates such as arylke-
tones and aromatic amines (35–37), amino acids (38, 39),
substituted indoles (40), and thiocyanate and hydrazinium
salts (41, 42). This strong oxidant has also found applica-
tions in industrial processes for the treatment of organics
in waste waters (43, 44) and for the bleaching of fabrics
(45).
Recently, Springer reported a new method for the delig-
nification of wood and kraft pulp using PMPA, which af-
forded strong bright pulps with minimal damage to cellu-
Run
Conditions
5 min
10 min
1
2
3
4
PMPA + FePcS
PMPA
89
0
100
5
PMPA + Fe(NO3)3
4
—
11
0
PDP + FePcS
Note. 1 mM TCP, 365 mM Tris/CH3COOH buffer (buffer/CH3CN =
1/3 v/v), 0.04 mM FePcS or Fe(NO3)3 (FePcS/TCP = 4% ), 6 mM PMPA
lose and hemicellulose (16, 46). As a result, we tried to or PDP.

184
SANCHEZ ET AL.
TABLE 5
TABLE 6
Product Distribution from the Oxidation of TCP Catalyzed
by FePcS/Phosphate/H2O2 and FePcS/PMPA
Influence of the Buffer on the Oxidation of DCBQ
Catalyzed by FePcS/H2O2
Oxidative
coupling
products
Ring
cleavage
products
Conversion at
60 min (% )
Run
Buffer
pH
1
2
3
Phosphate
Tris/H2SO4
—
7
7
5–6
76
61
39
FePcS/phosphate/H2O2
FePcS/PMPA
0%
13%
29%
21%
27%
0%
Note. 1 mM DCBQ, 50 mM buffer (buffer/CH3CN = 1/3 v/v), 0.04 mM
FePcS (FePcS/DCBQ = 4% ), 6 mM H2O2.
peroxyphosphate–FePcS complex as active intermediate
during the catalytic cycle.
Reaction products and mechanism. The organic prod- the FePcS/PMPA complex is not able to perform the C–C
ucts of the reaction of TCP with FePcS + PMPA were ex- bond cleavage. The same observation was made for the
tracted following a method previously described (14). They reaction catalyzed by FePcS/KHSO5, where DCBQ was
1
were analyzed by H NMR spectroscopy and compared formed in 60% yield at pH 2 (13). Thus, the further oxi-
to the products obtained from the FePcS/phosphate/H2O2 dation of DCBQ by the FePcS/H2O2 system should not be
system. We have already mentioned that with this latter ox- expected to be sensitive to the presence of phosphate ions.
idant combination, the degradation products of TCP were This was probed by performing an oxidation of DCBQ us-
mainly ring cleavage products (maleic acid being the major ing the FePcS/H2O2 system in different media. In a phos-
1
product) and oxidative coupling products (Fig. 2). The H phate buffer, the conversion ofDCBQ was76% after 60min
NMR analysis of the reaction catalyzed by FePcS + PMPA (Table 6, run 1) and 61% usinga Tris/H2SO4 buffer (Table 6,
shows that under these conditions the major products were run 2). The lack of a significant difference between these
the compounds resulting from oxidative coupling (21% ) two experiments confirms the assumption that the phos-
and 2,6-dichloro-1,4-benzoquinone (DCBQ) (13% ) even phate ions are not essential after the formation of DCBQ.
if some minors unidentified compounds appeared also in Without buffering the solution, the conversion of DCBQ
1
the H NMR spectrum (Table 5). The total yield of identi- decreases to 39% , but in this case the solution has a lower
fied products was only 34% but as already mentioned for pH value (Table 6, run 3).
the H2O2 oxidation of TCP catalyzed by FePcS in phos-
phate buffer, it was underestimated because of the extrac- would be active in a step preceding the formation of
tion method before NMR analysis (13, 14).
DCBQ. It is interesting to note that DCBQ and the ox-
Thus, the proposed peroxyphosphate–FePcS complex
DCBQ has also been detected in the reaction catalyzed idative coupling products have the same precursor: a phe-
by the FePcS/phosphate/H2O2 system, but it rapidly un- noxy radical that results from the one electron oxidation
dergoes further degradration through Grob-type fragmen- of the phenolate form of TCP (Fig. 11). This phenoxy
tations, leading to ring cleavage products. It has been radical can be further oxidized to cationic species and
proven that this step could be accomplished either by H2O2 then to DCBQ (13). The different behavior of the three
alone or via FePcS+H2O2 catalysis (47), but apparently catalytic systems FePcS/phosphate/H2O2, FePcS/KHSO5,
FIG. 11. Proposed mechanism for the degradation of TCP using the different catalytic systems: FePcS/H2O2, FePcS/phosphate/H2O2,
FePcS/KHSO5, FePcS/PMPA.

PHOSPHATE IN THE FePcS/H2O2 OXIDATION OF POLLUTANTS
185
and FePcS/PMPA could result from the first degradation
step of TCP, which consists of a one-electron abstrac-
tion of the phenolate derivative. We propose that the
2. McKeown, N. B., “Phthalocyanine Materials: Synthesis, Structure and
Function.” Cambridge University Press, Cambridge, UK, 1998.
3. Gouloumis, A., Liu, S.-G., Sastre, A., Va´zquez, P., Echegoyen, L., and
Torres, T., Chem. Eur. J. 6, 3600 (2000).
4. de la Torre, G., Claessens, C.-G., and Torres, T., Eur. J. Org. Chem.
2821 (2000).
5. Ali, H., and van Lier, J. E., Chem. Rev. 99, 2379 (1999).
6. Engelkamp, H., Middelbeek, S., and Nolte, R. J. M., Science 284, 785
(1999).
7. Salazar, J. R., in “Handbook of Petroleum Processes” (Mayers, R. A.,
Ed.), Part 9. McGraw–Hill, New York, 1986.
8. Hadasch, A., and Meunier, B., Eur. J. Inorg. Chem. 2319 (1999).
9. Hadasch, A., Sorokin, A., Rabion, A., and Meunier, B., New J. Chem.
22, 45 (1998).
10. Sorokin, A., and Meunier, B., Eur. J. Inorg. Chem. 1269 (1998).
11. Meunier, B., and Sorokin, A., Acc. Chem. Res. 30, 470 (1997).
12. Sorokin, A., De Suzzoni-Dezard, S., Poullain, D., Noe¨l, J.-P., and
Meunier, B., J. Am. Chem. Soc. 118, 7410 (1996).
´
iron(III)hydroperoxo species (FeIII OOH), postulated as
–
the predominant intermediate in the activation of FePcS by
H2O2, is not able to perform this oxidation. On the other
hand, this could be achieved by the more active Fe(IV)=O
and Fe(III)-peroxyphosphate species postulated in the
course of the reactions of FePcS with KHSO5, FePcS with
phosphate+H2O2, and FePcS with PMPA. Note that the
behavior of the two catalytic systems activated either by
KHSO5 or by PMPA is very similar; we can suppose that
a peroxysulfate–FePcS complex may be transiently formed
=
in the former case as well as the corresponding Fe(IV) O
species that can also be obtained from the isomeric
(b) form of the peroxyphosphate–FePcS complex by ho-
molytic cleavage of the O–O bond. However, these highly
oxidizing species are not able to cleave the C–C bond of
13. Sorokin, A., and Meunier, B., Chem. Eur. J. 2, 1308 (1996).
14. Sorokin, A., Se´ris, J.-L., and Meunier, B., Science 268, 1163 (1995).
15. Weber, M. J. H., and Busch, D. H., Inorg. Chem. 4, 469 (1965).
the aromatic ring by a Grob-type fragmentation which can 16. Springer, E. L., J. Pulp Pap. Sci. 23, J582 (1997).
17. Pal, N., Lewandowski, G., and Armenante, P. M., Biotechnol. Bioeng.
only be achieved by H2O2 or an iron(III)–OOH species,
and the degradation reaction stops at DCBQ and oxidative
coupling products. This proposed mechanism is depicted in
Fig. 11.
46, 599 (1995).
18. Assmuth, T., and Vartiainen, T., Chemosphere 28, 971 (1994).
19. Tratnyek, P. G., and Holgne´, J., Environ. Sci. Technol. 25, 1596
(1991).
20. Struijs, J., and Rogers, J. E., Appl. Environ. Microbiol. 55, 2527
(1989).
21. Pothuluri, J. V., Hinson, J. A., and Gerniglia, C. E., J. Environ. Qual.
20, 330 (1991).
CONCLUSION
We have highlighted the key role played by the phos-
phate buffer in the H2O2 oxidative degradation of several
chlorinated aromatic substrates catalyzed by an iron(III)
tetrasulfophthalocyanine. The results of the experiments
and control reactions have led us to propose the tran-
sient formation of a peroxyphosphate–FePcS complex as
an active species involved in the oxidation pathway. A
similar intermediate could be obtained by the reaction of
the peroxymonophosphoric acid (PMPA) with FePcS. Us-
ing 2,4,6-trichlorophenol as a model compound, we were
able to establish that the influence of phosphate is ef-
fective only in the first reaction step, which corresponds
to the one-electron oxidation of the substrate. Correlat-
ing all the experimental results obtained has allowed us
to propose a mechanism involving Fe(III)–OOH, Fe(III)–
peroxyphosphate, Fe(III)–peroxysulfate, and Fe(IV)==O
active species, depending on the activation mode used.
22. Valentovic, M. A., Ball, J. G., Anestis, D. K., and Rankin, G. O., Toxic
In vitro 9, 75 (1995).
23. Hage, R., Iburg, J. E., Kershner, J., Koek, J. H., Lempers, E. L. M.,
Martens, R. J., Racherla, U. S., Russel, S. W., Swarthoff, T., van der
Vliet, M. R. P., Wamaar, J. B., van der Wolf, L., and Krijnen, B., Nature
369, 637 (1994).
24. Sorokin, A., Fraisse, L., Rabion, A., and Meunier, B., J. Mol. Catal.
117, 103 (1997).
25. Renz, M., and Meunier, B., Eur. J. Org. Chem. 737 (1999).
26. Adam, W., and Hadjiarapoglou, L., Top. Curr. Chem. 164, 45
(1993).
27. (a) Meunier, B., New J. Chem. 16, 203 (1992); (b) Meunier, B., Chem.
Rev. 92, 1411 (1992).
28. (a) Richardson, D. E., Yao, H., Frank, K. M., and Bennett, D. A.,
J. Am. Chem. Soc. 122, 1729 (2000); (b) Kleinschmit, P., Bertsch-Frank,
B., Lehmann, T., and Panster P., in “IndustrialInorganicChemicalsand
Products,” Vol. 4, p. 3469. Wiley–VCH, New York, 1998.
29. Epstein, J., Demek, M. M., and Rosenblatt, D. H., J. Org. Chem. 21,
796 (1956).
´
30. Mejia-Radillo, Y., Yatsimirsky, A. K., Foroudian, H. J., Gillitt, N. D.,
and Bunton, C. A., J. Phys. Org. Chem. 13, 505 (2000).
31. Yang, Y.-C., Berg, F. J., Szafraniec, L. L., Beaudry, W. T., Bunton, C. A.,
and Kumar, A., J. Chem. Soc. Perkin Trans. 2, 607 (1997).
32. Kende, A. S., Delair, P., and Blass, B. E., Tetrahedron Lett. 35,
8123 (1994).
33. Battaglia, C. J., and Edwards, J. O., Inorg. Chem. 4, 552 (1965).
34. Schmidlin, J., and Massini, P., Chem. Ber. 43, 1102 (1910).
35. Ogata, Y., Tomizawa, K., and Ikeda, T., J. Org. Chem. 43, 2417 (1978).
36. Panda, A. K., Mahapatro, S. N., and Panigrahi, G. P., J. Org. Chem. 46,
4000 (1981).
ACKNOWLEDGMENTS
The financial support of CNRS is gratefully acknowledged. R.T.F thanks
the Alexander von Humboldt Foundation (Feodor-Lynen grant) and the
European Community (TMR Grant ERB4001GT973571) for postdoc-
toral fellowships.
REFERENCES
37. Ogata, Y., Tomizawa, K., and Marikawa, T., J. Org. Chem. 44, 352
(1979).
1. Leznoff, C. C., and Lever, A. B. P. (Eds.), “Phthalocyanines: Properties
and Applications,” Vols. 1–4. VCH, New York, 1989–1996.
38. Panigrahi, G. P., and Paichha, R. C., Int. J. Chem. Kinet. 23, 345
(1991).

186
SANCHEZ ET AL.
39. Panigrahi, G. P., and Paichha, R. C., Indian J. Chem. 29A, 272
43. Santo, Y., Nakano, E., Ishidoshiro, H., and Santo, K., Patent JP
04235791, Aug. 1992.
(1990).
40. Ghanem, R., Carmona, C., Munoz, M. A., Guardado, P., and Balo´n, 44. Groisman, A. Sh., Belyaeva, Z. G., Khachaturyan, O. B., and Sugurova,
M., J. Chem. Soc. Perkin Trans. 2, 2197 (1996). I. Yu., Patent SU 1699953, Dec. 1991.
41. Mishra, D. K., Peter Amala Dhas, T., Bhatnagar, P., and Gupta, Y. K., 45. Hartmann, C. E., Lex, W., and Preidt, D., Patent WO 9520643, Aug.
Indian J. Chem. 31A, 91 (1992).
1995.
42. Peter Amala Dhas, T., Mishra, D. K., Mittal, R. K., and Gupta, Y. K.,
Int. J. Chem. Kinet. 23, 203 (1991).
46. Springer, E. L., Patent US 5529663, Jun. 1996.
47. Hadasch, A., Ph.D. thesis, Univ. of Toulouse, 1999.