Unexpected 2,4,6-trimethylphenol oxidation in the presence of Fe(III) aquacomplexes

Article abstract of DOI:10.1039/b514691g

2,4,6-Trimethylphenol (TMP) was efficiently oxidised by Fe(III) aquacomplexes. HPLC analysis was used to follow the kinetics of the redox process. Two degradation products were detected and identified: 2,6-dimethyl-4-(hydroxymethyl)phenol (P1) and 3,5-dimethyl-4-hydroxybenzaldehyde (P2) accounting for 100% of TMP degradation in the early stages of the reaction. The formation of the products was concomitant with the reduction of Fe(III) into Fe(II). The direct relation between TMP oxidation and the concentration of the monomeric species {Fe(H₂O)₅(OH)} ²⁺ gives evidence for the initial reaction to take place between TMP and this particular species. Moreover, the correlation between P2 formation and P1 disappearance during the reaction implies the sequence of reactions: TMP → P1 → P2. A mechanism leading to the two degradation products is proposed. the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006.

Full text of DOI:10.1039/b514691g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Unexpected 2,4,6-trimethylphenol oxidation in the presence of Fe(III)
aquacomplexes
Jean-Pierre Aguer,* Gilles Mailhot and Michele Bolte
`
Received (in Montpellier, France) 17th October 2005, Accepted 25th November 2005
First published as an Advance Article on the web 14th December 2005
DOI: 10.1039/b514691g
2
,4,6-Trimethylphenol (TMP) was efficiently oxidised by Fe(III) aquacomplexes. HPLC analysis
was used to follow the kinetics of the redox process. Two degradation products were detected and
identified: 2,6-dimethyl-4-(hydroxymethyl)phenol (P1) and 3,5-dimethyl-4-hydroxybenzaldehyde
(
P2) accounting for 100% of TMP degradation in the early stages of the reaction. The formation
of the products was concomitant with the reduction of Fe(III) into Fe(II). The direct relation
2
1
2 5
between TMP oxidation and the concentration of the monomeric species {Fe(H O) (OH)} gives
evidence for the initial reaction to take place between TMP and this particular species. Moreover,
the correlation between P2 formation and P1 disappearance during the reaction implies the
sequence of reactions: TMP - P1 - P2. A mechanism leading to the two degradation products
is proposed.
Introduction
dimethylbenzoquinone. From a mechanistic point of view, the
schemes of degradation are completely different. DMP degra-
dation involves the intermediate formation of the 2,6-di-
methylphenoxyl radical detected by EPR spectroscopy.
From this radical, the dimer can be formed, together with
Probe molecules are often used to evaluate the photodegrada-
tion processes in aqueous systems. For this purpose, 2,4,6-
trimethylphenol (TMP) was chosen to test the photoinductive
1
,2
properties of humic substances. The extension of its use to
other systems, especially systems used in heterogeneous photo-
catalysis, was then considered. Iron being very often proposed
to modify the properties of photocatalysts, more particularly
0 0
3
,3 ,5,5 -tetramethyldiphenoquinone obtained by further oxi-
dation of the dimer by Fe(III). 2,6-Dimethylbenzoquinone is
also formed by subsequent oxidation of the radical. In the case
5
of DSD, Wong-Wah-Chung et al. proposed an intermediate
TiO , it was necessary to investigate TMP behaviour in the
2
radical with one Fe(III) linked to the ethylenic carbon of DSD
through an oxygen atom. Then, from this radical, two path-
ways are possible: formation of the epoxide derivative or
reaction with another Fe(III) leading to the 5-amino-2-formyl-
benzenesulfonic acid.
presence of iron.
In addition, industrial processes used to prepare 2,6-di-
methylphenol (DMP), a starting compound for various synth-
eses, led to the formation of 2,4,6-trimethylphenol (TMP) as a
major by-product of little practical use. Methods involving
metal catalysts to transform TMP into DMP were investigated
So work on TMP oxidation induced by Fe(III) aquacom-
plexes was undertaken and is reported in this paper.
3
,4
in which iron complexes were used to catalyse the reaction.
Actually, Fe(III) aquacomplexes were proved to be able to
promote oxidation of organic compounds in aqueous solution.
Experimental
5
Wong-Wah-Chung et al. observed a very fast reaction be-
Reagents
0
0
tween 4,4 -diaminostilbene-2,2 -disulfonic acid (DSD) and
Fe(III). Two oxidation products corresponding to the oxida-
tion of the stilbene double bound were identified: the DSD
epoxide derivative and 5-amino-2-formylbenzenesulfonic acid.
In the investigated system, the redox process essentially de-
pends on the concentration of monomeric species {Fe
2,4,6-Trimethylphenol was an Aldrich product (99%) used
without further purification. Ferric perchlorate nonahydrate
[Fe(ClO ) ꢀ 9H O; purity 97%] was a Fluka product kept in a
4
3
2
dessicator. 3,5-Dimethyl-4-hydroxybenzaldehyde was an
Aldrich product (95%). The Fe(III) solutions were prepared
2
1
ꢂ3
ꢂ1
(
H
2
O)
5
(OH)} which appears to be the major oxidative agent
by diluting
a
stock solution [2.0
ꢁ
10
mol
L
in the reaction. Under the same conditions, Mazellier and
6
Bolte observed a redox process between 2,6-dimethylphenol
Fe(ClO ) ꢀ 9H O] to the appropriate Fe(III) concentration.
4
3
2
All the solutions were prepared with ultrapure aerated water
(Millipore aQ, resistivity = 18.2 MO cm). Deoxygenated
solutions were obtained by bubbling with argon for 30 min
at room temperature.
(
(
DMP) and Fe(III): the reaction requires an excess of {Fe
0
21
0
H O) (OH)} . The major product is the 3,3 ,5,5 -tetra-
2
5
methyldiphenoquinone together with DMP dimer and 2,6-
Apparatus
Laboratoire de Photochimie Mole´culaire et Macromole´culaire (CNRS
UMR 6505), Universite´ Blaise Pascal, Ensemble Universitaire des
Ce´zeaux, F-63177 Aubie`re Cedex, France. E-mail:
J-Pierre.AGUER@univ-bpclermont.fr; Fax: þ33 4 73 40 77 00;
Tel: þ33 4 73 40 71 70
UV-visible spectra were recorded on a Cary 3 double beam
spectrophotometer. HPLC were carried out using a Merck
L6200 chromatograph equipped with a Hewlett Packard 1050
This journal is ꢃc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 191–196 | 191

View Article Online
detector and using a Waters chromatograph equipped with a
Waters 996 photodiode array detector. The eluent was a
mixture of methanol–water (60 : 40) acidified by H PO
3
4
ꢂ1
(
0.1%). The flow rate was 1 mL min and the column was
a ZORBAX Rx C8 (5 mm) of 25 cm length. Liquid chromato-
graphy-mass spectra analyses were obtained on a Thermo Ion
Trap DECA XP Ion Max apparatus by using APCI negative
mode. NMR spectra were recorded with a Bruker Advance
4
00 MHz, using solutions in deuterated water. Electron para-
magnetic resonance (EPR) spectra were obtained using a
Brucker ER-200D spectrometer at 9.30 GHz with a modula-
tion field of 100 KHz. An Xe–Hg Hanovia Lamp was used for
irradiation in the EPR spectrometer cavity (l > 345 nm).
Analysis
Fe(II) concentration was determined by complexometry with
ꢂ1 ꢂ1
ꢂ4
ꢂ1
21
0
7
cm ) 8600]. The
Fig. 1 UV-visible spectra of TMP (1.0 ꢁ 10 mol L ), Fe(OH)
2
,2 -bipyridine at 520 nm [e (L mol
monomeric concentration of Fe(III) species was determined
ꢂ4
ꢂ1
21
3.0 ꢁ 10 mol L ), the mixture TMP–Fe(OH) and the sum of the
(
21
spectrum of TMP and the spectrum of Fe(OH)
.
8
by using the modified Kuenzi procedure with 8-hydroxy-
9
quinoline-5-sulfonic acid (HQSA).
The kinetics of TMP disappearance and of the formation of
oxidation products were determined by HPLC experiments
spectra of the components (Fig. 1). The difference between
the sum and the spectrum of the mixture was assigned to the
very fast reaction between Fe(III) and TMP (cf. Fig. 2). The
spectral evolution of the mixture was monitored as a function
of time. A typical evolution is represented in Fig. 2. HPLC
analysis gave evidence for TMP disappearance and the for-
mation of two products, together with Fe(II) formation de-
tected by complexometry.
(ldetection = 278 nm).
Results
Characterisation of the Fe(III)–TMP mixture
3
ꢂ1 10
)
TMP is soluble in water (1.01 ꢁ 10 mg L
and its UV-
(H O)
visible spectrum presents an absorption band at l
nm) 278 [e (L mol cm ) 1250].
The UV-visible spectrum of Fe(III) in aqueous solution
max
2
ꢂ1
ꢂ1
(
TMP degradation
The observed ‘‘thermal’’ (in the dark and at room tempera-
ture) degradation of TMP was directly related to the presence
of Fe(III) since TMP alone was stable in aqueous solution. In
order to estimate the influence of Fe(III) and of its speciation
on the process, two sets of experiments were carried out.
Firstly the concentration of Fe(III) was varied and taken as
continuously evolved. Under our starting experimental condi-
ꢂ5
tions of concentration ([Fe(III)] between 5 ꢁ 10 and 1.0 ꢁ
ꢂ3 ꢂ1
1
0
mol L and pH (4.2 > pH > 3.0) according to Fe(III)
concentration), the species present just after dissolving
Fe(ClO was the monomeric form {Fe(H O)
ꢀ 9H
OH)} , simply noted as Fe(OH) . The concentration of mono-
4
1
)
3
2
O
2
5
2
21
(
ꢂ4
ꢂ4
ꢂ4
ꢂ4
ꢂ4
0
3
.5 ꢁ 10 , 1.0 ꢁ 10 , 1.5 ꢁ 10 , 2.0 ꢁ 10 , 2.5 ꢁ 10 and
ꢂ4 ꢂ1 21
meric species rapidly decreased after the dissolution of ferric
perchlorate in water. As a result it is almost impossible at such
.0 ꢁ 10 mol L ; in all cases the percentage of Fe(OH)
ꢂ5
ꢂ4
ꢂ1
low concentrations (5 ꢁ 10 to 3 ꢁ 10 mol L ) to get
21
1
00% of Fe(OH) . The disappearance was attributed to the
1
1
possible formation of soluble aggregates. Using the 8-hydro-
xyquinoline-5-sulfonic acid (HQSA) method, we were able to
measure the percentage of monomeric species in the starting
Fe(III) solution:
2
1
21
%
Fe(OH) = 100 ꢁ {[Fe(OH) ]/[Fe(III)tot]}
where Fe(III)tot is the concentration of Fe(III) added in the
solution. The different percentages of monomeric species were
obtained by the use of Fe(III) stock solutions with different
ageing.
Two maxima at 280 nm and 298 nm, corresponding,
21
respectively, to TMP and Fe(OH) , were observed in the
UV-visible spectrum of a freshly prepared mixture of TMP
ꢂ1
ꢂ4
ꢂ1
21
ꢂ4
(
1.0 ꢁ 10 mol L ) and Fe(OH) (3.0 ꢁ 10 mol L ).
21
There was no detectable complexation between Fe(OH) and
TMP in our experimental conditions. At t = 0, the UV-visible
spectrum of the mixture was roughly the sum of the two
ꢂ4
ꢂ1
Fig. 2 UV-visible spectra of the mixture TMP (1.0 ꢁ 10 mol L )–
2
1
ꢂ4
ꢂ1
Fe(OH) (3.0 ꢁ 10 mol L ) as a function of time.
1
92 | New J. Chem., 2006, 30, 191–196
This journal is ꢃc the Royal Society of Chemistry the Centre National de la Recherche Scientifique 2006

View Article Online
21
Fig. 4 Degraded TMP as a function of initial Fe(OH) concentra-
tion.
21
the reaction with respect to Fe(OH) was then determined.
The initial rate can be expressed as followed:
2
1
a
V = k([Fe(OH) ] ) ꢁ [TMP]
(1)
0
0
0
2
1
where V
0
is the initial rate, k is the rate constant, [Fe(OH)
]
0
and [TMP] are the initial concentrations of both products,
0
[
TMP] being kept constant. The value of a was evaluated to
.73 by plotting ln(V ) as a function of ln[Fe(OH) ] .
0 0
0
2
1
0
Fe(II) formation
The formation of Fe(II) was followed by complexometry,
giving evidence for the involvement of Fe(III) in a redox
reaction with TMP at room temperature in the dark (Fig. 5).
Fig. 3 TMP disappearance as a function of time under different
ꢂ1
ꢂ4
conditions, [TMP]
ꢂ4
0
= 1.0 ꢁ 10 mol L . (a) K [Fe(III)] = 0.5 ꢁ
ꢂ1
21
ꢂ4
ꢂ1
ꢂ1
1
0
mol L (90% of Fe(OH) ). m [Fe(III)] = 1.0 ꢁ 10 mol L
21
ꢂ4
(
90% of Fe(OH) ). . [Fe(III)] = 1.5 ꢁ 10 mol L (90% of
The plateau value of the Fe(II) concentration increased with
21
21
ꢂ4
ꢂ1
21
Fe(OH) ).E [Fe(III)] = 2.0 ꢁ 10 mol L (90% of Fe(OH) ). %
increasing the Fe(OH)
21
percentage. Moreover, the ratio
ꢂ4
ꢂ1
21
Fe(III)] = 2.5 ꢁ 10 mol L (90% of Fe(OH) ). ’ [Fe(III)] = 3.0
[
[Fe
]form./[TMP]disap. was calculated all along the process at
21
ꢂ4
ꢂ1
21
10 mol L (90% of Fe(OH) ). (b) K [Fe(III)] = 3.0 ꢁ 10 mol
ꢂ4
ꢁ
two percentages of Fe(OH) species, 40 and 90%. It reached
a common value of around 2 for the two percentages.
ꢂ1
21
(o5% of Fe(OH) ). m [Fe(III)] = 3.0 ꢁ 10 mol L (30% of
ꢂ4
ꢂ1
L
21 ꢂ1 21
ꢂ4
Fe(OH) ). . [Fe(III)] = 3.0 ꢁ 10 mol L (40% of Fe(OH) ).E
ꢂ4
ꢂ1
21
Fe(III)] = 3.0 ꢁ 10 mol L (60% of Fe(OH) ). ’ [Fe(III)] =
[
ꢂ4
ꢂ1
21
3
.0 ꢁ 10 mol L (90% of Fe(OH) ).
species was equal to 90 ꢄ 5% and the initial concentration of
ꢂ4
ꢂ1
TMP was equal to 1.0 ꢁ 10 mol L . Secondly, to confirm
2
1
the involvement of Fe(OH) in the reaction, the effect of the
21
percentage of Fe(OH) species was investigated at constant
ꢂ4
ꢂ1
Fe(III) total concentration (3.0 ꢁ 10 mol L ) and with the
ꢂ4
ꢂ1
same initial concentration of TMP (1.0 ꢁ 10 mol L ). The
starting percentage values were determined using the HQSA
method; the results are presented in Fig. 3a and 3b, respec-
tively. The disappearance of TMP was fast at the beginning,
slowed down and reached a pseudo plateau-value. The con-
centration of TMP at the pseudo plateau depends on both the
2
1
concentration of Fe(III) and the Fe(OH)
percentage. It
2
1
clearly appeared that the higher the Fe(OH) percentage or
Fe(III)tot concentration, the higher the abatement in TMP was.
In addition, there was a roughly linear relationship between
Fig. 5 Fe(II) formation during the reaction under different condi-
ꢂ4
ꢂ1
ꢂ4
ꢂ1
tions, [TMP]
0
= 1.0 ꢁ 10 mol L . . [Fe(III)] = 3.0 ꢁ 10 mol L
21 ꢂ1
ꢂ4
(90% of Fe(OH) ). % [Fe(III)] = 3.0 ꢁ 10 mol L (40% of
21
21
Fe(OH) ).
initial [Fe(OH) ] and degraded [TMP] (Fig. 4). The order of
This journal is ꢃc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 191–196 | 193

View Article Online
Fig. 6 Chemical structures and UV-visible spectra of P1 and P2.
Transformation products
concentration was determined by a calibration procedure.
Taking into account that only two products were detected
by HPLC, the quantitative conversion of TMP to P1 and P2
was admitted. With this hypothesis, P1 concentration was
calculated by subtracting P2 formation from TMP disappear-
ance:
The formation of two products P1 and P2 was detected by
HPLC when mixing Fe(III) with TMP. Both were more polar
than TMP; the retention time was 3.7, 5.8 and 8.6 min for P1,
P2 and TMP, respectively. The peaks were observed whatever
the percentage of monomeric species and the Fe(III) concen-
ꢂ4
ꢂ4
ꢂ1
tration (0.5 ꢁ 10 r [Fe(III)] r 3.0 ꢁ 10 mol L ). Product
1
was identified by classical analytical techniques: H NMR
[P1]formed = [TMP]disap. ꢂ [P2]formed
1
1
spectroscopy and mass spectrometry. The results of H NMR
The ratio [P2]/([P1] þ [P2]) for different concentrations of
in D O are gathered below:
2
monomeric species was determined at constant concentration
ꢂ4
ꢂ1
in TMP (1.0 ꢁ 10 mol L ). As shown in Fig. 7, the ratio
d (400 MHz; D O; Me Si): 2.32 (6H, s, Me),
H
2
4
[
P2]/([P1] þ [P2]) appears to be favoured when the concentra-
4.58 (2H, s, CH ), 7.13 (2H, s, Ph).
2
tion of monomeric species increases. In order to confirm this
tendency, TMP disappearance and P1 and P2 formation were
The analysis by LC-MS gave two major peaks at m/z: 151.2
ꢂ
ꢂ4
ꢂ1
investigated in a mixture of TMP (1.0 ꢁ 10 mol L ) and
(
M , 100%) and 149.2 (20%). From these results, product 1
P1) was assigned as 2,6-dimethyl-4-(hydroxymethyl)phenol.
2
1
freshly prepared Fe(III) (90% of Fe(OH) ) at a higher con-
(
ꢂ3
ꢂ1
centration c = 1.0 ꢁ 10 mol L . With the increase of the
reaction time, P1 concentration decreased whereas P2 con-
tinuously accumulated in the solution, implying that P2 results
from the oxidation of P1 by Fe(III). This hypothesis was
corroborated by plotting P2 concentration as a function of
P1 concentration (Fig. 8). A linear correlation was found
Product 2 (P2) was identified as 3,5-dimethyl-4-hydroxyben-
zaldehyde by comparison with a commercially available
authentic sample. The chemical structures and the UV spectra
of these products are given in Fig. 6. In the mixtures, P2
2
1
Fig. 7 Influence of Fe(OH) concentration on the ratio [P2]/([P1] þ
P2].
[
Fig. 8 Plot of P2 concentration as a function of P1 concentration.
1
94 | New J. Chem., 2006, 30, 191–196
This journal is ꢃc the Royal Society of Chemistry the Centre National de la Recherche Scientifique 2006

View Article Online
Fig. 9 Evolution of Fe(II) concentration formed as a function of the
ꢂ4
ꢂ1
ꢂ3
reaction time, [TMP]
ꢂ1
0
= 1.0 ꢁ 10 mol L , [Fe(III)] = 1.0 ꢁ 10
0
21
mol L (90% of Fe(OH) ). m [Fe(II)] calculated. ꢁ [Fe(II)] mea-
sured.
between the two concentrations, supporting the following
sequence of reaction:
ꢂ1
Fig. 10 Spin trapping experiments with DMPO (1 g L ) by EPR
ꢂ3
spectroscopy on irradiation (l > 345 nm): [TMP]
0
= 1.0 ꢁ 10 mol
TMP - P1 (alcohol derivative) - P2 (aldehyde derivative)
ꢂ1
ꢂ3
ꢂ1
21
L
, [Fe(III)]
0
= 3.0 ꢁ 10 mol L (90% of Fe(OH) ).
2
1
The ratio [Fe ]form/([P1]form.þ [P2]form.) was calculated
throughout the reaction for the concentration of Fe(III) equal
ꢂ3
ꢂ1
to 1.0 ꢁ 10 mol L . At the beginning of the reaction,
corresponding to the higher concentration of P1, the ratio
took a value of around 2 (see Fe(II) formation). In addition,
throughout the reaction, Fe(II) concentration, calculated on
the basis of two Fe(II) formed for one molecule of P1 detected
and two additional Fe(II) to transform P1 into P2, fits the
measured Fe(II) concentration very well (Fig. 9).
signal detected by EPR spectroscopy was unchanged, in
agreement with the involvement of the phenoxyl radical.
Discussion
Reaction of TMP in the presence of Fe(III) was studied in
aqueous solution. Formation of two main products was ob-
served. The first one was identified as 2,6-dimethyl-4-(hydro-
xymethyl)phenol and the second one as 3,5-dimethyl-4-
hydroxybenzaldehyde.
Effects of oxygen and isopropanol
The effect of oxygen was investigated on a mixture of freshly
ꢂ4
ꢂ1
ꢂ4
Such products were also mentioned in the study of Li and
4
Liu in the case of experiments performed with TMP in the
prepared Fe(III) (3.0 ꢁ 10 mol L) and TMP (1.0 ꢁ 10
ꢂ1
mol L). TMP degradation was not affected by the absence of
oxygen. Isopropanol used as a hydroxyl radical scavenger was
also added: no effect was observed on the reaction indicating
that hydroxyl radicals were not involved in TMP transforma-
tion.
presence of iron salts in alcoholic solvent. If the products were
well described, the mechanism was unclear and did not fit our
observations very well. The authors proposed the radical
cation formed via electron transfer from 2,4,6-TMP to Fe(III)
as the first intermediate of the reaction. Loss of a proton gives
rise to the benzyl radical which reacts with oxygen to form a
peroxy species. This last species is reduced in the presence of
the Fe(II) formed in the first step of the reaction. In this
process, the Fe(III) species are totally regenerated. Our ob-
servations, showing the accumulation of Fe(II) and the lack of
oxygen effect, did not agree with such a mechanism. In our
case, and according to the set of results provided by the EPR
experiments, the redox reaction between TMP and Fe(III)
involves the phenoxyl radical in a primary step. As demon-
strated by the linear correlation between the concentration of
EPR spectroscopy
Because the phenoxyl radical was established as the first
intermediate in the oxidation of 2,6-dimethylphenol in pre-
6
sence of Fe(III) aquacomplexes, spin-trapping experiments in
0
EPR spectroscopy were undertaken. 5,5 -Dimethylpyrroline-
ꢂ1
N-oxide (DMPO, 1 g L ) was added to the aqueous solution
ꢂ3
ꢂ1
of Fe(III) (3.0 ꢁ 10 mol L with 90% of monomeric species)
ꢂ3
ꢂ1
and TMP (1.0 ꢁ 10 mol L ). A 1 : 2 : 2 : 1 quartet (with
coupling constant a = 15.0 Gauss) was observed (Fig. 10).
The existence of a quartet instead of the triplet of doublets
usually observed with DMPO is due to the identical value of
the coupling constants of the electron with nitrogen aN and
21
21
Fe(OH) species and the degraded [TMP] (Fig. 4), Fe(OH)
appears to be the most active species in terms of oxidative
properties. The formation of two Fe(II) for one degraded TMP
argues for the P1 formation resulting from a two electron
with hydrogen a . In the presence of 1% of isopropanol the
H
This journal is ꢃc the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 191–196 | 195

View Article Online
Conclusion
Fast oxidation of 2,4,6-trimethylphenol is observed in the
presence of Fe(III) aquacomplexes. The efficient redox process
leads to the formation of two products: 2,6-dimethyl-4-(hy-
droxymethyl)phenol (P1) followed by 3,5-dimethyl-4-hydro-
xybenzaldehyde (P2), together with the formation of Fe(II).
This process essentially depends on the concentration of
2
1
monomeric species Fe(OH) , which appears to be the major
oxidative agent in the reaction. The correlation between the
decrease of P1 and the increase of P2 concentrations indicates
that P1 is primarily formed and then by a further oxidation
process gives rise to P2. An oxidation mechanism involving a
21
two electron transfer from TMP to Fe(OH) and then from
2
1
P1 to Fe(OH) was proposed. This redox reaction between
TMP and Fe(III) aquacomplexes raises the question of TMP
use as a probe molecule when iron is present in the tested
systems. Our results prove that iron traces could produce an
oxidative transformation of TMP and alter the final conclu-
sion.
Acknowledgements
We wish to thank Dr Patrick Mazellier (Universite
for his help in performing LC-MS experiments and Anne-
Sophie Martin (Universite Blaise Pascal, SEESIB Laboratory)
´
de Poitiers)
´
for the NMR spectra.
References
1
´
S. Canonica, U. Jans, K. Stemmler and J. Hoigne, Environ. Sci.
Technol., 1995, 29, 1822.
C. Richard, O. Trubetskaya, O. Trubetskoy, O. Reznikova, G.
2
Afanaseva, J.-P. Aguer and G. Guyot, Environ. Sci. Technol., 2004,
3
8, 2052.
K. Takehira, M. Shimizu, Y. Watanabe, H. Orita and T. Haya-
kawa, Tetrahedron Lett., 1990, 31, 2607.
Scheme 1
3
4
5
K.-T. Li and P.-Y. Liu, Appl. Catal., A, 2004, 272, 167.
P. Wong-Wah-Chung, G. Mailhot, J.-F. Pilichowski and M. Bolte,
New J. Chem., 2004, 28, 451.
oxidation. The nature of the product requires the addition of
an oxygen atom. Because of the lack of a molecular oxygen
effect on the rate of TMP disappearance, the redox process has
to be associated with a concerted mechanism of addition of an
hydroxide ion leading to P1. Taking into account that two
Fe(III) were once again necessary to form P2, the formation of
the aldehyde should be explained by a similar sequence of
reactions. The final step corresponds to a dehydration of the
resulting diol leading to the aldehyde. The overall mechanism
is presented in Scheme 1.
6 P. Mazellier and M. Bolte, Chemosphere, 1997, 35, 2181.
7
M. C. Goldberg, K. M. Cunningham and E. R. Weiner, J.
Photochem. Photobiol., A, 1993, 73, 105.
W. H. Kuenzi, PhD thesis, University of Zurich, Switzerland, 1982.
8
9 N. Brand, G. Mailhot and M. Bolte, Environ. Sci. Technol., 1998,
2, 2715.
3
1
0 W.-Y. Shiu, K.-C. Ma, D. Varhanickova and D. Mackay, Chemo-
sphere, 1994, 29, 1155.
1
1 R. J. Knight and R. N. Sylva, J. Inorg. Nucl. Chem., 1975, 37,
779–783.
1
96 | New J. Chem., 2006, 30, 191–196
This journal is ꢃc the Royal Society of Chemistry the Centre National de la Recherche Scientifique 2006