Degradation of sodium 4-dodecylbenzenesulphonate photoinduced by Fe(III) in aqueous solution

Article abstract of DOI:10.1016/S0045-6535(99)00434-8

The Fe(III)-photoinduced degradation of 4-dodecylbenzenesulphonate (DBS) in aqueous solution was investigated. The mixing of DBS (1 mm) and Fe(III) (1 mm) solutions immediately led to the formation of a precipitate that contained DBS and monomeric Fe(OH)²⁺, the predominant Fe(III) species. Both species were also present in the supernatant. Irradiation of the supernatant solution resulted in a photoredox process that yielded Fe(II) and ·OH radicals. The disappearance of DBS was shown to involve only attack by ·OH radicals; the quantum yield of DBS disappearance is similar to the quantum yield of ·OH radical formation. A wavelength effect was also observed; the rate of DBS disappearance was higher for shorter wavelength irradiation. Five photoproducts, all containing the benzene sulphonate group, were identified. ·OH radicals preferentially abstract hydrogen from the carbon in the α position of the aromatic ring. The results show that the Fe(III)-photoinduced degradation of DBS could be used as an alternative method for polluted water treatment. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0045-6535(99)00434-8

Chemosphere 41 (2000) 363±370
Degradation of sodium 4-dodecylbenzenesulphonate
photoinduced by Fe(III) in aqueous solution
Gilles Mailhot, Ayaz Asif, Mich ꢀe le Bolte *
Laboratoire de Photochimie Mol eꢁ culaire et Macromol eꢁ culaire, Unit eꢁ Mixte de Recherche CNRS, Universit eꢁ Blaise Pascal n° 6505,
F-63177 Aubiere Cedex, France
Received 6 August 1999; accepted 21 September 1999
Abstract
The Fe(III)-photoinduced degradation of 4-dodecylbenzenesulphonate (DBS) in aqueous solution was investigated.
The mixing of DBS (1 mm) and Fe(III) (1 mm) solutions immediately led to the formation of a precipitate that con-
2‡
tained DBS and monomeric Fe(OH) , the predominant Fe(III) species. Both species were also present in the super-
natant. Irradiation of the supernatant solution resulted in a photoredox process that yielded Fe(II) and áOH radicals.
The disappearance of DBS was shown to involve only attack by áOH radicals; the quantum yield of DBS disappearance
is similar to the quantum yield of áOH radical formation. A wavelength e€ect was also observed; the rate of DBS
disappearance was higher for shorter wavelength irradiation. Five photoproducts, all containing the benzene sulpho-
nate group, were identi®ed. áOH radicals preferentially abstract hydrogen from the carbon in the a position of the
aromatic ring. The results show that the Fe(III)-photoinduced degradation of DBS could be used as an alternative
method for polluted water treatment. Ó 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Photodegradation; Dodecylbenzenesulphonate; Iron(III); Solar light
1
. Introduction
Over the last few years, Hidaka et al. (1992, 1994)
and Zhao et al. (1992) extensively reported on the rapid
photodegradation (k P 330 nm) of several kinds of an-
ionic, cationic and non-ionic surfactants catalysed by
Environmental contamination by toxic or carcino-
genic chemicals is a serious global problem. The surf-
actants that are widely employed in domestic and
industrial ®elds constitute one of the main families of
pollutants in surface and ground water (Faust, 1975);
they can cause a severe ecological problem due to their
very slow and inecient biodegradation (Swisher and
Pollut, 1963; Larson and Payne, 1981; Larson and Perry,
2
TiO semiconductors in aqueous aerated suspensions;
for aromatic surfactants, the aromatic ring was decom-
posed more rapidly than was the alkyl chain. The
mechanism involved the attack of the surfactants by
2
áOH radicals followed by further degradation to CO (g)
2
4
�
and SO ions.
1
981; Larson et al., 1983. Several days or even weeks are
Linear alkylbenzene sulphonate (LAS) is one of the
most important technical surfactant in world-wide usage
(Berth and Jeschke, 1989), the estimated consumption
of LAS was about 1.8 million metric tons. Among
the LAS-type surfactants used in household detergents,
4-dodecylbenzenesulphonate (DBS) is the most repre-
sentative.
required for the complete biodecomposition of surfac-
tants, which often leads to the formation of more per-
sistent and even more toxic metabolites.
*
Corresponding author. Tel.: +33-4-73-40-71-69; fax: +33-
4
-73-40-77-00.
E-mail address: boltem@ciesum.univ-bpclermont.fr (M.
Bolte).
In our investigation, the photochemical behaviour of
DBS in the presence of Fe(III) was studied. DBS does
0045-6535/00/$ - see front matter Ó 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 9 9 ) 0 0 4 3 4 - 8

3
64
G. Mailhot et al. / Chemosphere 41 (2000) 363±370
not absorb solar light, but its transformation can be
photoinduced by Fe(III). With Fe(III), two main types
of photoreaction can be observed depending on the
nature of the pollutant: If the pollutant is a complexing
agent, such as nitrilotriacetic acid (NTA) (Stolzberg and
Hume, 1975; Andrianirinaharivelo et al., 1993), ethy-
lenediaminetetraacetic acid (EDTA) (Carey and Lang-
ford, 1973; Svenson et al., 1989; Kari et al., 1995) or
ethylenediaminetetra-(methylenephosphonic acid) (ED-
TMP) (Matthijs et al., 1989), a complex between the
pollutant and Fe(III) is formed, which absorbs light into
the visible spectral region; an intramolecular photoredox
process is generally observed that leads to the degrada-
tion of the pollutant. On the other hand, if Fe(III) is not
complexed by the pollutant, it can absorb the solar light
and undergo a photoredox process giving rise to Fe(II)
and áOH radicals (Baxendale and Magee, 1955; Dainton
and Tordo€, 1957). The pollutant is then attacked by
áOH radicals, which is a very strong oxidising agent. We
also examined the photochemical degradation of the 4-
and Lomb). The beam was parallel, and the reactor was
a cylindrical quartz cell (2 cm path length). The photon
¯ux was measured by ferrioxalate actinometry (Calvert
and Pitts, 1966).
A high-pressure Hg lamp (Philips HPW 125 W),
giving an emission at 365 nm was located at a focal axis
of an elliptical stainless steel cylinder used for kinetic
and preparative experiments. The reactor, a water-
jacketed Pyrex tube (d ˆ 2:8 cm) was centred at the
other focal axis. The reaction medium was well stirred.
1
5
The unit provided illumination (I
s
0
ꢁ 4:0 Â 10 photons
�
1
� 2
cm ) of a large volume (40 ml) of solution.
UV-visible spectra were recorded with a CARY 3
double beam spectrophotometer. EPR spectra were
obtained with a BRUKER ER-200D spectrometer at
9.30 GHz with a modulation ®eld of 100 KHz. A Xe±Hg
Hanovia lamp was used for irradiation in the EPR
spectrometer cavity (k P 305 nm).
HPLC chromatograms were carried out with a
WATERS 540 system equipped with a diode array UV-
visible detector (WATERS 990) and a 25-cm Lichro-
sphere RP18 Merck column (reverse phase). The eluent,
ethylbenzenesulphonic acid (EBS)
compound of DBS, in the presence of Fe(III).
, a model
�
1
at a ¯ow rate of 1 ml min , was a mixture of deionised
water (with 1& H PO ) and methanol, the composition
3
4
2
. Experimental
of which depended on the product investigated; conve-
nient compositions were 1.5:1 (v/v) for the characteri-
sation of DBS and 9:1 (v/v) for the characterisation of
the photoproducts. The kinetics of DBS disappearance
was evaluated from the integrated areas of the peaks of
the homologs.
2
.1. Reagents and solutions
Technical grade 4-dodecylbenzenesulphonic acid
sodium salt (DBS) (Aldrich) was used without further
puri®cation. This commercial product is generally
a mixture of various alkyl homologs (C10-C14) and
phenyl position isomers (2-CnLAS, 3-CnLAS; . . .) with
an average carbon chain length of approximately
n ˆ 12:4-acetyl-benzenesulphonic acid sodium salt (1)
The photoproducts were separated by preparative
HPLC (``Fast'' procedure, Gilson Co.) on a C-18 reverse
phase column (21:4 Â 50 mm), using successively, water
3 4
+1& of H PO and pure water as eluents. Negative
electrospray mass spectra (ESMS) were obtained from
``Service Central dÕanalyse'', CNRS, Vernaison, France.
Infrared spectra were recorded on a Nicolet 20 SX FTIR
(
98%), 4-hydroxybenzenesulphonic acid sodium salt
dihydrate (3) (98%) and 4-sulphobenzoic acid potassium
salt (5) (95%) (Aldrich) and p-EBS acid sodium salt
1
13
spectrometer (KBr pellets). H and C NMR spectra
were recorded on a Bruker AC 400 MHz Fourier
transform spectrometer.
(
TCI Extra Pure) were used without further puri®cation.
Ferric perchlorate nonahydrate (Fe(ClO 9H O)
Fluka > 97%) was stored in a dessicator.
All solutions were prepared from deionised ultrapure
4
)
3
,
2
(
2.3. Analysis
water (q ˆ 18:2 MX cm). When necessary, the solutions
were deaerated by bubbling with Ar for 30 min at room
temperature. The pH was adjusted with NaOH and
measured with an ORION pH meter to Æ0:02 pH units;
pH ˆ 3:0 for a solution of ‰FeꢀIII†Š ˆ 0:3 mm. The ionic
strength of the solution was not controlled. Fe(III) so-
lutions were prepared by diluting stock solutions (2 mM
The concentration of Fe(II) was determined by
complexometry with o-phenanthroline, taking e510
ˆ
4
� 1
� 1
1; 118 Â 10
m
cm
for the Fe(II)-phenanthroline
complex (Calvert and Pitts, 1966). HCHO formation
was measured by colorimetry with chromotropic acid as
described by Bricker and Vail (1950).
Fe(ClO
.2. Apparatus and procedures
Monochromatic irradiations at 296, 313 and 365 nm
4
)
3
) to the appropriate concentration.
The method used to determine the concentration of
monomeric Fe(III) was a modi®cation of Kuenzi's
procedure Kuenzi, 1982: 1 ml each of 0.05 M 8-hy-
droxyquinoline-5-sulphonic acid (HQSA) and acetic
acid bu€er (pH ˆ 4.6) were poured into a 10-ml volu-
metric ¯ask; 8 ml of sample was added to this solu-
tion and rapidly mixed. Within 30 s of mixing, the
2
were carried out with a high-pressure Hg lamp (Osram
HBO 200 W) and a grating monochromator (Bausch

G. Mailhot et al. / Chemosphere 41 (2000) 363±370
365
Fe ‡ H O ꢀ FeꢀOH† ‡ H‡
3
‡
2‡
ꢀ1†
absorbance of the tris complex of 8-hydroxyquinoline-
-sulphonate with Fe(III) (Fe(HQS) ) was measured at
2
5
3
k ˆ 572 nm; the same solution with 8 ml of water instead
of sample was used as a blank. HQSA reacts much more
rapidly with monomeric Fe(III) species than with Fe(III)
3‡
‡
‡
Fe ‡ 2H
2
O ꢀ FeꢀOH† ‡ 2H
ꢀ2†
ꢀ3†
ꢀ4†
2
3
‡
O ꢀ FeꢀOH† ꢀaq† ‡ 3H‡
Fe ‡ 3H
2
dimers or polymers; Fe(HQS)
3
absorbs in a spectral
region in which these latter species or Fe(III) precipitates
3
2Fe ‡ 2H O ꢀ Fe ꢀOH† ‡ 2H‡
3‡
4‡
do not. Values of e572 of Fe(HQS)
3
have been reported
� 1
2
2
2
�
1
by Kuenzi (1982) (5180 M
�
cm ) and by Faust
and Hoign ꢁe 1990 (4970 M cm ); an average value of
1
� 1
Under our experimental conditions (0:1 6 ‰FeꢀIII†Š 6 1
4
2
‡
�
1
� 1
mM and 2:5 6 pH 6 4:0), the formation of Fe
2
2
ꢀOH†
5
075 M cm was used here.
The DBS concentration was determined by the
‡
can be neglected; it would appear that Fe(OH) is the
predominant monomeric Fe(III)-hydroxyl complex
Methylene-Blue-Active-Substance (MBAS) method
Apha et al., 1981), which depends on the formation of
(
Knight and Sylva, 1975; Wagman et al., 1982), with
3
(
‡
Fe representing ꢂ10% of the total Fe(III) concentra-
a blue salt or ion pair when methylene blue, a cationic
dye reacts with anionic surfactants such as LAS. The
results were in agreement with those obtained by HPLC
measurements.
tion. For the rest of the discussion, we will only consider
2‡
3‡
Fe(OH) ; Fe , the minor monomeric species, does not
absorb at k P 310 nm and does not interfere in the
photochemical behaviour of the system. The concen-
The formation of the photoproducts was followed by
HPLC with detection at 220 and 260 nm. The determi-
nation of the quantum yields of DBS disappearance and
Fe(II) formation were calculated from the concentration
as a function of irradiation time at conversion percent-
age lower than 10%.
2
‡
tration of Fe(OH) rapidly decreased after the disso-
lution of Fe(ClO in water, this disappearance has
4 3
)
been attributed to the possible formation of soluble
aggregates of Fe(III) (Knight and Sylva, 1975). It ap-
2‡
pears that the concentration of Fe(OH) , the most
photoactive species (Faust and Hoign ꢁe , 1990), strongly
depends on the age of the ferric solution (before and
after dilution) and on the starting concentration. We
3. Results
2‡
were able to determine [Fe(OH) ] with the HQSA
method.
3.1. Characterisation of DBS and Fe(III) in aqueous
solution
3.1.1. Evolution of the mixture of DBS and Fe(III) in
aqueous solution
Commercial DBS is a mixture of about 30 homologs
The addition of 1 mm Fe(III) to the 1 mM DBS re-
sulted in ¯occulation; the FTIR analysis of the precipi-
tate showed the presence of both Fe(III) and DBS.
Centrifugation and elimination of the precipitate left a
2‡
supernatant solution that contained Fe(OH) and DBS
with the concentrations of 0.23 and 0.40 mM, respec-
tively. 60% of starting DBS was eliminated from the
solution.
among which the species with n ˆ 9 and the phenyl ring
bonded on carbon-2 (2-C12LAS) is the major one. The
UV-Visible spectrum exhibits a maximum at 223 nm
�
1
� 1
�
(
e ꢁ 10 300 M cm ) and a structured band centred at
3.2. Photochemical behaviour
1
� 1
2
60 nm (e ꢁ 350 M cm ). Adachi and Kobayashi
(
1984) reported that there is no direct photodegradation
All attempts to detect any complexation between
DBS and Fe(OH) in the supernatant failed; the UV-
visible spectrum matched the sum of the three compo-
2‡
of DBS upon irradiation at k > 290 nm. The concen-
trations in DBS used in this work (0.1 or 1 mM) were
lower than the critical micelle concentration (CMC) re-
ported in the literature (1.6 mM at 25°C) (Fendler and
Fendler, 1975; Furton and Norelus, 1993).
2
‡
3‡
nents (Fe(OH) , Fe and DBS). At k P 310 nm, where
neither DBS nor Fe absorb, the spectrum only corre-
3‡
2‡
sponded to that of Fe(OH) as established by com-
parison with the spectrum reported by Faust and
In order to interpret accurately the absorption spec-
trum and photochemical behaviour, it is crucial to un-
derstand the hydrolytic speciation of Fe(III), which is
described by the following equilibria in dilute aqueous
solution (Escot, 1973). For simplicity, coordinated water
molecules are not explicitly included in the chemical
�
1
Hoign ꢁe (1990), e being equal to 1760 and 270 mol
l cm at 313 and 365 nm respectively. The fact that the
�
1
absorption at long wavelengths was only that of
Fe(OH) ruled out any signi®cant formation of ion-
2‡
2‡
pairs between DBS and Fe(OH) . Upon irradiation,
2‡
2‡
2‡
formula; thus, Fe(OH) refers to FeꢀOH†ꢀH
2
O†5
.
Fe(OH) can act as a photoinitiator of degradation

3
66
G. Mailhot et al. / Chemosphere 41 (2000) 363±370
through internal electron transfer. Its charge transfer
band strongly overlaps the solar UV spectrum (290±
The quantum yields of DBS disappearance (U) were
determined in the presence of oxygen for irradiation at
296, 313 and 365 nm with triplicate runs. The results,
which are summarised in Table 1, show that the quan-
tum yields of DBS disappearance are strongly wave-
length dependent and are similar to the values of the
quantum yield of áOH radical formation as reported by
Benkelberg and Warneck (1995). This agreement leads
to the conclusion that at least at the beginning of the
reaction all the photolytically generated áOH react with
DBS.
4
00 nm) and its photolysis eciently produces áOH
radicals and Fe(II) (Baxendale and Magee, 1955;
Dainton and Tordo€, 1957; Faust and Hoign ꢁe 1990;
Matsuura and Smith, 1970).
FeꢀOH† ‡ ‡ hm
2
k P 290 nm
!
Fe ‡ áOH
2‡
ꢀ5†
Benkelberg and Warneck (1995) measured the quantum
yield of áOH radicals formation (/) as a function of
excitation wavelength (Table 1). They attributed the
increase in the quantum yield when the excitation
wavelength is decreased to the kinetic energy required to
achieve the ejection of áOH radicals from the solvent
cage.
3
.2.2. HPLC experiments
The decrease in DBS concentration was measured by
HPLC and complexometry (MBAS) for irradiation at
65 nm; the results obtained by both methods were in
good agreement (Fig. 2); the rate of DBS disappearance
3
The áOH radicals, being very strong oxidation agents,
are able to attack DBS; irradiation at 365 nm e€ected
the complete removal of DBS.
2‡
decreased when the initial concentration of Fe(OH)
was decreased. The appearance of peaks mixed with the
injection peak was observed from the start of the irra-
diation; several photoproducts with very high polarity
3
.2.1. Kinetics measurements
Irradiation of the supernatant containing Fe(OH)
2‡
2 3
were formed. With the use of the 9:1 (v/v) H O/CH OH
and DBS at 365 nm, resulted in a continuous decrease in
absorbance across the spectrum, re¯ecting the disap-
pearance of Fe(III) and DBS (Fig. 1); a similar result
was obtained for irradiation at 296 and 313 nm.
eluent, the formation of ®ve main photoproducts was
observed. The rate of formation and degradation of
the photoproducts also increased with increasing
2‡
[Fe(OH) ]. At long irradiation times, the absorbance of
DBS and of the photoproducts at 220 nm disappeared.
In the presence of 1% (v/v) 2-propanol as a scavenger of
áOH radicals, the degradation of DBS was totally in-
hibited. This result con®rmed that DBS degradation is
only due to the áOH radicals produced upon irradiation
of Fe(III) in aqueous solution.
Table 1
Quantum yields of DBS disappearance and quantum yields of
OH radical formation
a
b
k (nm)
U
dis: DBS
k (nm)
/ (áOH)
296
313
365
0:33 Æ 0:03
0:17 Æ 0:01
0:08 Æ 0:02
290
310
370
0:288 Æ 0:025
0:195 Æ 0:033
0:074 Æ 0:015
3
.2.3. Formation of Fe(II)
In all cases, Fe(II), as determined by complexometry,
a
Our work.
(
was formed. [Fe(II)] quickly rises, then reaches a
constant value upon continued irradiations (Fig. 3); the
b
30).
Fig. 2. Decrease in DBS concentration upon irradiation at
365 nm measured by HPLC (n) and complexometry with
Fig. 1. Evolution of Fe-DBS UV-visible spectrum upon
irradiation at 365 nm.
2‡
MBAS (s). ‰FeꢀOH† Š ˆ 0:03 mm; ‰DBSŠ ˆ 0:08 mm.

G. Mailhot et al. / Chemosphere 41 (2000) 363±370
367
plateau is due to the reoxidation of Fe(II) species. The
value of the [Fe(II)] at the plateau is a positive function
quartet instead of a triplet of doublets was due to the
identical value of the coupling constants a and a . In
the presence of 1% of 2-propanol, no signal was detected.
N
H
2
‡
of the initial concentration of Fe(OH) . Any involve-
ment of a Fenton-type reaction in the early stages of the
reaction can be ruled out; no H
system.
2
O
2
was added to the
3.2.5. Photoproduct analyses
Irradiation of Fe(III)-DBS solutions yields ®ve main
photoproducts; (1), (3) and (5) were identi®ed with au-
thentic compounds commercially available. The char-
acterisation of photoproducts (2) and (4) was realised
from additional experiments with 4-EBS acid, which
contains a short alkyl chain; the same ®ve photoprod-
ucts were formed upon irradiation of EBS.
3
.2.4. EPR spectroscopy upon irradiation (kP305 nm)
For spin-trapping EPR experiments, 5,5-dimethyl-1-
pyrroline-1-oxyde (DMPO) (1 mg/ml), was added to the
Fe(III) and DBS solution. Upon irradiation, a 1:2:2:1
quartet (a ˆ 14:6 G) was obtained, a signal that is
characteristic of the stable adduct with áOH radicals
As a matter of fact, even at high concentrations
6 mM Fe(III) and 4 mM EBS), the immediate forma-
(
(
Fig. 4) (Janzen et al., 1978). The observation of a
tion of a precipitation when the Fe(III) and EBS aque-
ous solutions were mixed in aqueous solution was never
observed, rendering the separation and identi®cation of
the common photoproducts easier. The photoproducts
(
2) and (4) were separated by preparative HPLC without
any damage, but some thermal reactions occurred when
they were thoroughly dried. In general, these species
were kept wet and negative electrospray mass spect-
rometry was performed directly on the aqueous solution;
2
exchange in D O was necessary prior to the recording of
their NMR spectra.
The characteristics of the ®ve photoproducts are
summarised in Table 2.
The photodegradation of 4-acetylbenzenesulphonate
(
1), the major photoproduct, was also investigated in the
presence of Fe(III). Photoproducts (3) and (4) were also
obtained upon irradiation of (1) at 365 nm; (2) and (5)
were not observed under these conditions.
Fig. 3. Formation of Fe(II) as a function of irradiation time at
3
65 nm in aerated solution; at t ˆ 0 ‰FeꢀOH† ‡Š ˆ 0:11 mM:
2
‰
FeꢀIII†Š ˆ 0:3 mM, ‰DBSŠ ˆ 0:1 mM, pH ˆ 3:0.
0
0
3
.2.6. Oxygen e€ect
The quantum yield of DBS disappearance was not
signi®cantly changed in the absence of oxygen. In con-
trast, the absence of oxygen a€ected the formation of the
photoproducts very strongly. The rate of formation of
the ®ve main photoproducts was divided by 10 and we
observed the formation of the new photoproducts, cer-
tainly some alcohols as already observed with tributyltin
in the absence of oxygen (Mailhot et al., 1999). In the
same time, we observed an important increase of the rate
of formation of Fe(II), multiplied by 5. These observa-
tions can be explained by the redox reaction with Fe(III)
for the termination reaction of radicals.
á
R ‡ FeꢀIII† ! ROH ‡ FeꢀII†
4
. Mechanism and discussion
The primary and secondary photoproducts that were
isolated all contain the sulphonate group, indicating that
the carbon in the a position of the aromatic ring is the
principal site of attack by OH radicals that leads to the
cleavage of the alkyl chain.
Fig. 4. Spin-trapping experiments with DMPO (1 mg per 1 ml
of solution) by EPR spectroscopy upon irradiation (k P 305
nm) : ‰FeꢀIII†Š ˆ 0:3 mM, ‰DBSŠ ˆ 0:1 mM, pH ˆ 3:0.
á
0
0

3
68
G. Mailhot et al. / Chemosphere 41 (2000) 363±370

G. Mailhot et al. / Chemosphere 41 (2000) 363±370
369
2
yields RO , peroxy radicals
á
á
2
Sehested et al. (1975) studied the di€erent sites of
Reaction of R with O
attack of áOH radicals on methylated benzene and
concluded that in mildly acidic solution benzyl radicals
were formed via path A.
that can undergo a head-to-head termination reaction to
form intermediate tetroxides R±O±O±O±O±R (Sonntag
and Schuchmann, 1991); their decomposition to give
molecular oxygen and two alkoxy radicals is a well-
known reaction. The b-scission of the resulting alkoxy
radicals is the typical decomposition pathway reported
in Carlson and Wiles (1969).
0
·
The scission of the C±R bond (route A) directly leads
to the formation of 4-acetylbenzenesulphonate [1] as
n
a major photoproduct with the homologs 2-C LAS.
Hydroxy-4-acetylbenzenesulphonate [4] results from
the subsequent attack of [1] by áOH radicals.
In route B, the scission of the C-phenyl bond gives
At lower pH, the rate of the direct formation of the
benzyl radical via path B increased and became com-
·
·
rise to the phenyl radical , and through reaction with
2
petitive with the elimination of H O from the áOH ad-
2
O or Fe(III) leads to the formation of 4-hydroxy-
duct to the benzene ring in path A. We cannot rule out
the formation of the corresponding phenol from the
áOH adduct with the aromatic ring of DBS; but we never
observed it in our experiments.
benzenesulphonate [3].
The formation of 4-formylbenzenesulphonate [2] can
be explained by considering the presence of isomers
with the phenyl group attached to the ®rst carbon
Considering that the pH of the Fe(III) ± DBS solu-
á
(
n
1-C LAS) in commercial DBS. In this case, [2] could
tion is ꢂ3, one would expect the tertiary radical (R ),
be obtained via the b-scission of the C±C bond of the
corresponding ROá alkoxy radical. This type of rad-
ical can also isomerise to a carbon centred radical
which reacts with oxygen and after elimination of
which should have a greater stability then the benzyl
radical, to be formed by both paths A and B.
2
HOá give rise to a long chain keto compound.
Photoproduct [5] would be obtained by further oxi-
dation of [2].
From the above results and observations the fol-
lowing mechanism can be put forward.
Photoproducts [3] and [4] are also produced by the
Fe(III)-photoinduced degradation of [1]. It is worth
noting that this mechanism is not only valid for the
di€erent homologs present in commercial DBS but also
for EBS, as evidenced by the similarity of the photo-
products.
In the absence of oxygen, the attack of áOH radicals
on DBS leads to ROH and Fe(II) by a redox process
between Rá and Fe(III).
5
. Conclusion
This work shows that the degradation of DBS in
aqueous solution when photoinduced by Fe(III) is a very
ecient process, which could be used for the destruction

3
70
G. Mailhot et al. / Chemosphere 41 (2000) 363±370
of surfactants in polluted waters. The precipitation ob-
served when DBS and Fe(III) are mixed is also an in-
teresting example of the physical elimination of DBS
from water that could also be fruitfully employed in
water treatment. After removal of the precipitate,
Escot, M.T., 1973. Contribution ꢀa l' ꢁe tude des complexes
hydroxyl ꢁe s du fer ferrique. Th ꢀe se de docteur en Pharmacie,
Universit ꢁe de Pharmacie de Clermont-Ferrand, France.
Faust, S.D., 1975. In: Fate of pollutants in the air and water
environments. In: Su€et I.H. (Ed.), Advances in Environ-
ment and Technology, Part 2, vol. 8. Wiley, New York,.
Faust, B.C., Hoign ꢁe , J., 1990. Atmos. Environ. 24A, 79.
Fendler, J.H., Fendler, E.J., 1975. Catalysis in Micellar and
Macromolecular Systems. Academic Press, New York,
p. 20.
2
‡
Fe(OH) is the predominant species of Fe(III) present
2‡
in the solution. The photolysis of Fe(OH) produces
áOH radicals; the degradation of DBS is due to attack by
áOH radicals followed by secondary reactions that in-
volve O
2
. The primary step of the decomposition of the
Furton, K.G., Norelus, A., 1993. J. Chem. Educ. 70 (3), 254.
Hidaka, H., Nohara, K., Ooishi, K., Zhao, J., Serpone, N.,
Pelizzetti, E., 1994. Chemosphere 29 (12), 2619.
surfactant involves H-abstraction at the carbon atom in
the a position of the aromatic ring, giving rise to dif-
ferent benzenesulphonate derivatives. For long irradia-
tion times, the total disappearance of DBS and of the
photoproducts is observed. Our experimental condi-
tions, which are very well controlled in terms of pH,
concentration and water quality, permit a clear assess-
ment of the photochemical mechanism of DBS degra-
dation.
Hidaka, H., Zhao, J., Pelizzetti, E., Serpone, N., 1992. J. Phys.
Chem. 96, 2226.
Janzen, E.D., Nutter, D.E., Davis, E.R., Blackburn, B.J.,
Poyer, J.L., McCay, P.B., 1978. Can. J. Chem. 56, 2237.
Kari, F.G., Hilger, S., Canonica, S., 1995. Environ. Sci.
Technol. 29, 1008.
Knight, R.J., Sylva, R.N., 1975. J. Inorg. Nucl. Chem. 37, 779.
Kuenzi, W.H., 1982. Die hydrolyse von eisen (III) der ein¯uss
von chlorid auf bildung und zerfall von vernetzungsproduk-
ten. Ph. D. dissertation ETH n 7016, Eidgen o ssischem
Technischem Hochschule, Z u rich, Switzerland.
Acknowledgements
Larson, R.J., Payne, A.G., 1981. Appl. Environ. Microbiol. 41
(
3), 621.
We thank Professor Morton Ho€man (Boston Uni-
versity) for fruitful comments on this manuscript.
Larson, R.J., Perry, R.L., 1981. Wat. Res. 15, 697.
Larson, R.J., Vashon, R.D., Games, L.M., 1983. Biodeterio-
ration 5, 235.
Mailhot, G., Astruc, M., Bolte, M., 1999. Appl. Organometal.
Chem. 13, 53.
References
Matsuura, T., Smith, J.M., 1970. Ind. Eng. Chem. Fundam. 9
(2), 252.
Adachi, A., Kobayashi, T., 1984. Eisei Kagaku 30 (6), 377.
Andrianirinaharivelo, S.L., Pilichowski, J.F., Bolte, M., 1993.
Transition. Met. Chem. 18, 37.
Matthijs, E., DeOude, N.T., Bolte, M., Lemaire, J., 1989. Wat.
Res. 23 (7), 845.
Apha et al., 1981. In: Standard Methods for the Examination of
Water and Wastewater,15th ed. American Public Health
Association, Washington, DC, p. 530.
Sehested, K., Cor®tzen, H., Christensen, H.C., Hart, E.J., 1975.
J. Phys. Chem. 79 (4), 310.
Sonntag, C.V., Schuchmann, H.P., 1991. Angew Chem. Int. Ed.
Engl. 30, 1229.
Baxendale, J.H., Magee, J., 1955. Trans. Faraday Soc. 51, 205.
Benkelberg, H.J., Warneck, P., 1995. J. Phys. Chem. 99, 5214.
Berth, P., Jeschke, P., 1989. Tenside Surfactants Detergents 26,
Stolzberg, R., Hume, D., 1975. Environ. Sci. Technol. 9, 654.
Svenson, A., Kaj, L., Bjorndal, H., 1989. Chemosphere 18
(9/10), 1805.
7
5.
Bricker, C.E., Vail, W.A., 1950. Anal. Chem. 5, 720.
Calvert, J.G., Pitts, J.M. (Ed.), 1966. Photochemistry. Wiley,
New York, p. 783.
Swisher, R.D., Pollut, J.W., 1963. Control Fed. 35, 877.
Wagman, D.D., Evans, W.H., Parker, V.B., Schumm, R.H.,
Halow, I., Bailey, S.M., Churmey, K.L., Nuttall, R.L.,
1982. J. Phys. Chem. Ref. Data 11 (2), 392.
Zhao, J., Oota, H., Hidaka, H., Pelizzetti, E., Serpone, N.,
1992. J. Photochem. Photobiol. A Chem. 69, 251.
Carey, J.H., Langford, C.H., 1973. Can. J. Chem. 51, 3665.
Carlson, D.J., Wiles, D.M., 1969. Macromolecules 2, 597.
Dainton, F.S., Tordo€, M., 1957. Trans. Faraday. Soc. 666, 53.