By-products formation during degradation of isoproturon in aqueous solution. I: ozonation

Article abstract of DOI:10.1016/S0043-1354(00)00427-9

The degradation of the herbicide isoproturon during its ozonation in aqueous solution has been investigated with the aim of identifying intermediate as well as final by-products formed. At ambient temperature, phosphate-buffered (pH=7) isoproturon aqueous solutions (10, 10^-1 and 10^-3mg/l) were ozonated in a semi-batch reactor, under a continuous flow of ozonated air whose ozone concentration was 9 and 0.9mgO₃/l_air for the highest and the two lower herbicide concentrations respectively. Measured steady-state ozone concentrations during the two sets of experiments (i.e. the highest and the lower isoproturon concentration) were 1.9 and 0.7mgO₃/l. Under all of the above conditions, isoproturon was always completely removed in a period ranging between 5 and 15min, essentially by reacting with molecular ozone. High-performance liquid chromatography-mass spectrometry (HPLC-MS) analyses indicate that primary degradation by-products are formed either by introducing OH groups in the aromatic ring and/or in the side-chain substituents, or by breaking down the isopropyl alkyl chain. The results also show that these primary intermediates are successively degraded yielding low molecular weight compounds such as aldehydes, simple organic acids and α-oxo-acids, which have been identified by gas chromatography-electron capture detection (GC-ECD), ion chromatography (IC) and GC-MS, respectively. On the basis of the analytical results, a pathway for the degradation of isoproturon by ozone has been proposed. Copyright

Full text of DOI:10.1016/S0043-1354(00)00427-9

Wat. Res. Vol. 35, No. 7, pp. 1695–1704, 2001
2001 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
#
PII: S0043-1354(00)00427-9
0043-1354/01/$ - see front matter
BY-PRODUCTS FORMATION DURING DEGRADATION OF
ISOPROTURON IN AQUEOUS SOLUTION. I: OZONATION
1
GIUSEPPE MASCOLO *, ANTONIO LOPEZ , HUW JAMES and MIKE FIELDING
1
2
2
1
Istituto di Ricerca Sulle Acque, Consiglio Nazionale delle Ricerche, Via F. De Blasio, 5, 70123 Bari, Italy
and WRc, Henley Road, Medmeham, Marlow, Bucks SL7 2HD, UK
2
(
First received 5 August 1999; accepted in revised form 23 August 2000)
Abstract}The degradation of the herbicide isoproturon during its ozonation in aqueous solution has been
investigated witht he aim of identifying intermediate as well as final by-products formed. At ambient
ꢀ1
ꢀ3
temperature, phosphate-buffered (pH=7) isoproturon aqueous solutions (10, 10 and 10 mg/l) were
ozonated in a semi-batchreactor, under a continuous flow of ozonated air whose ozone concentration was
9
and 0.9 mg O
steady-state ozone concentrations during the two sets of experiments (i.e. the highest and the lower
isoproturon concentration) were 1.9 and 0.7 mg O /l. Under all of the above conditions, isoproturon was
3
/lair for the highest and the two lower herbicide concentrations respectively. Measured
3
always completely removed in a period ranging between 5 and 15 min, essentially by reacting with
molecular ozone. High-performance liquid chromatography-mass spectrometry (HPLC-MS) analyses
indicate that primary degradation by-products are formed either by introducing OH groups in the
aromatic ring and/or in the side-chain substituents, or by breaking down the isopropyl alkyl chain. The
results also show that these primary intermediates are successively degraded yielding low molecular weight
compounds suchas alde hy des, simple organic acids and a-oxo-acids, which have been identified by gas
chromatography-electron capture detection (GC-ECD), ion chromatography (IC) and GC-MS,
respectively. On the basis of the analytical results, a pathway for the degradation of isoproturon by
ozone has been proposed. # 2001 Elsevier Science Ltd. All rights reserved
Keywords }by-products, isoproturon, herbicide, water, ozonation, HPLC, mass-spectrometry
INTRODUCTION
discounted. From a chemical standpoint the com-
monly used chemical disinfectants (i.e., Cl
ClO , BrCl/HOBr, O , etc.) are all oxidant species
2
/HOCl,
The application of herbicides on agricultural soils is a
well established and effective practice to control weed
growth. Herbicides are applied to soils repeatedly
and because of their rather slow degradation
2
3
potentially able to react with herbicides causing their
removal and/or degradation (EPA, 1986). Therefore,
the degradation by-products should be expected, and
their identification is an important issue as little is
known about their nature, and their toxicity could be
even higher than that of the parent compounds.
With this aim, long-term investigations have been
carried out to assess whether herbicides react with
disinfectants, and if so, to identify the resulting by-
products (Mascolo et al., 1994, 1995). The present
paper is the first part of an investigation to study the
degradation of the herbicide isoproturon during its
reactions withtwo disinfectants: ozone (part I) and
chlorine (part II).
(Khadrani et al., 1999; Madhun and Freed, 1987;
Pieuchot et al., 1996), their occurrence in ground-
waters as well as surface waters has become more and
more frequent. There is, therefore, concern regarding
herbicide contamination of water resources (Fielding
et al., 1992; Lacorte et al., 1996) and also an interest
in reliable technologies for removing herbicides from
water. Even though this latter goal is commonly
achieved by appropriate physico-chemical methods
suchas adsorption onto activated carbon, membrane
filtration and advanced oxidation processes (Di
Giano, 1987; O’Brien et al., 1987; Camel and
Bermond, 1998; Ollis et al., 1991), the possibility
that herbicides can be removed and/or degraded even
during the disinfection step routinely carried out at
drinking water treatment plants should not be
Currently, ozone applications for water treatment
are numerous and widespread all over the world in
spite of its two major technological drawbacks
(
Rice and Netzer, 1984), viz. firstly, under conditions
common to water and wastewater treatment, ozone is
an unstable gas and it must, therefore, be generated
and applied at its point of use; secondly, as ozone is
*
Author to whom all correspondence should be addressed.
Tel.: +39-080-5820519; fax: +39-080-5313365; e-mail: only partially soluble in water, its introduction
mascolo@area.ba.cnr.it into water or wastewater requires a very efficient
1
695

1696
Giuseppe Mascolo et al.
gas–liquid contact to maximise the mass transfer of 1992). It can be biodegraded, like other phenyl-urea
ozone from the gas phase to the liquid phase. From herbicides (Cullington and Walker, 1999; Esposito
the chemical point of view, in aqueous solution et al., 1998; Vroumsia et al., 1996). It is also broken
molecular ozone can remain as O , or it can down by advanced oxidation processes (De Laat
3
decompose to form very reactive species suchas the
ꢁ
et al., 1996; Meijers et al., 1995; Prados et al., 1995;
hydroxyl radical (OH ) which is a stronger oxidising Roche and Prados, 1995) and several degradation by-
agent
(
(E
0
=2.80 V)
than
molecular
ozone products have been identified (Allemane et al., 1995;
E =2.07 V). Factors affecting ozone decomposition Aguer and Richard, 1996; Pramauro et al., 1993;
0
include: pH, presence of specific water constituents Richard and Bengana, 1996). However, isoproturon
i.e., initiators, promoters, and inhibitors of free degradation by-products have been identified only
radical reactions), ionic strengthand temperature for the system O /H O (Allemane et al., 1995), i.e.,
Langlais et al., 1991). The chemical effects of ozone when the predominant oxidant species should be the
(
3
2
2
(
ꢁ
in water are then a result of the following: its direct OH radical (Glaze, 1987; Glaze et al., 1987).
reaction withdissolved compounds; its decomposi-
Furthermore, in this study (Allemane et al., 1995)
ꢁ
tion into secondary oxidants (e.g. OH ); the sub- highly polar by-products would not have been
sequent reactions of these secondary oxidants with identified, as only GC-MS analyses without any
solutes (Hoigne
´
and Bader, 1976).
Molecular ozone, being a weaker oxidising agent volatile non-polar compounds) were carried out.
derivatization (which would only detect relatively
ꢁ
than OH , is a rather selective oxidant and because Therefore, the specific aim of the work reported in
of this, most water and wastewater treatment this paper was to identify the main by-products,
processes are carried out withrelatively low ozone including polar compounds, formed during the
dosages (Rice, 1997). Larger amounts of more reaction between isoproturon and ozone alone at
ꢁ
From the kinetic point of view, as OH is a much 1987; Glaze et al., 1987).
reactive oxidants suchas OH
due to their consumption in side oxidation reactions. species should be molecular ozone (O ) (Glaze,
ꢁ
reaction rate constants are about nine orders of under laboratory conditions which were analytically
would be required neutral pH, i.e., when the predominant oxidant
3
stronger and less-selective oxidant than ozone, its
By-products identification was first performed
magnitude greater than those of ozone for the same convenient
organic compound (Hoigne and Bader, 1983). tion=10 mg/l; transferred ozone dose ozo-
Referring to reaction mechanisms, molecular ne=45.9 mg/l) even though these conentrations
ozone can react withorganic substrates ei te hr were far more extreme than those which would
(isoproturon
initial
conccentra-
´
through cyclo-addition to unsaturated bonds or actually be encountered at real water treatment
electrophilic and nucleophilic reactions (Langlais plants. Follow-up work involved searching for the
et al., 1991). Electrophilic reactions take place on by-products identified during this preliminary step
molecular sites witha hi ghelectron density suchas when experimental conditions closer to reality were
those found in aromatic compounds substituted with used (i.e., isoproturon initial concentration=100 and
electron donor groups (–OH, NH , etc.). Conversely, 1 mg/l; transferred ozone dose ꢂ2 mg/l).
2
nucleophilic reactions occur on molecular sites
showing an electron deficit and, more frequently,
on carbons carrying electron-withdrawing groups
–NO , –Cl, etc.). As for the hydroxyl radical, it
reacts with organics by three major mechanisms:
hydroxy addition, hydrogen abstraction, and elec-
tron transfer (Huang et al., 1993; Legrini et al., 1993).
EXPERIMENTAL SECTION
(
2
Chemicals
Isoproturon
[3-(4-isopropylphenyl)-1,1-dimethyl-urea]
9% purity, (Polyscience, Niles, L, USA) was used without
9
further purification. Water and methanol used for liquid
In practice, even though pesticides removal by chromatography were HPLC-grade (Rathburn, Walkerburn
Scotland). Methanol and methylene chloride used for solid–
ozonation is often achieved successfully, the complete
liquid extractions were pesticide-grade from Rathburn.
destruction of these contaminants (i.e. the total
mineralization of organic carbon to carbon dioxide)
Diazomethane was produced from N-methyl-N-nitroso-p-
toluenesulphonamide (Aldrich, Milwaukee, WI, USA)
is seldom obtained. Therefore, the formation of according to a published procedure (Aldrich, 1993). All
several ozonation by-products is usually to be other chemicals were analytical grade from Aldrich.
Double-distilled water was used for preparing all aqueous
solutions.
expected (Reynolds et al., 1989) as found for other
pollutants (Brambilla et al., 1993; Ku et al., 1998;
Prados and Ciba, 1997; Yao et al., 1998).
Isoproturon is a herbicide belonging to the class of
phenyl-urea derivatives, widely used in agriculture to
OZONATION EXPERIMENTS
control broadleaf and grassy weeds in cereals and
other crops (Tomlin, 1997). Because in Europe the
annual consumption of isoproturon is estimated to
be over 1000 tons, it has been included in the
A 10 mg/l isoproturon aqueous solution was prepared by
introducing appropriate amount of isoproturon into a
5
00ml flask and dissolving it (withstirring) in p ho sp ha te
buffer solution (10 mM) at pH=7. The solution was then
placed into a Normag reactor (Normag, Hofheim am
3
European ‘‘black list’’ of pesticides (Fielding et al., Taunus, Germany) for ozonation. Ozonated air (9 mg O /l)

By-products formation during degradation of isoproturon
1697
was produced by a Fisher 502 ozonator (Meckenheim,
Germany) and bubbled into the Normag reactor at a flow
isocratic 100% A for 5 min, then programmed from 100%
A to 100% B over 25 min, then isocratic until time=35 min.
rate of 60 l/h. Experiments at lower isoproturon concentra- For the determination of other organic acids, 20 ml of water
tion (100 and 1 ppb) were carried out witha flow rate of 6 l/h sample were acidified to pH=2 withsulp hu ric acid and
of ozonated air (9 mg O
/l). Ozone output was monitored liquid–liquid extracted with diethyl ether (2 Â 3 ml). The
before eachexperiment by determining (by titration with extracts were dried with sodium sulphate, methylated with
3
sodium thiosulphate) the amount of free iodine liberated diazomethane and their volume reduced to 1 ml for gas-
from a potassium iodide solution. Ozone transferred during chromatography/mass-spectrometry (GC-MS) analyses.
the reaction was determined by the following equation:
GC-MS analyses were carried out witha VG (Manc he ster,
UK) TS-250 trisector magnetic mass spectrometer equipped
withan electron impact (EI) source and a PDP-11 micro-
computer (Digital, Maynard, MA, USA) for data acquisi-
tion. The mass spectrometer was interfaced to a 5890 gas-
chromatograph (Agilent Technologies, Palo Alto, CA,
O3 transferred ¼ ðO3consumed þ O residual in waterÞ
3
¼
ðO in the ðinÞgas ꢀ O in the ðoffÞ gasÞ
3
3
Samples were withdrawn from the reaction mixtures at
scheduled times, quenched with sodium thiosulfate, and
analysed by HPLC. Residual ozone concentrations in the
aqueous solutions were determined by the Indigo method
USA). The column used was
.25 mm i.d. and 0.25 mm film thickness) from Supelco
Bellefonte, PA, USA) directly interfaced to the ion source
a SPB-5 (30 m length,
0
(
(Standard Methods, 1999).
through a transfer line heated to 2808C. One microlitre of
the sample was injected into a splitless injector at 2008C.
The operating conditions were: carrier gas: helium, flow
rate: 1 ml/min, initial temperature: 408C (1 min), program-
ming rate: 158C/min up to 2808C, then isothermal for 5 min.
ANALYTICAL DETERMINATIONS
The concentration of unreacted isoproturon and the EI mass spectra were obtained at 70 eV electron energy with
production of isoproturon by-products were monitored by
the ion source at 1808C. The magnetic analyser was scanned
HPLC-UV with a 1050 chromatographic system (Agilent from 40 to 400 a.m.u. over a period of 1 s, witha resolution
Technologies, Palo Alto, CA, USA) equipped with a
Spherisorb C-8, 5 mm (250 Â 4.6 mm i.d.) column from
of 500 (10% valley). The background-subtracted mass
spectra were matched against those in the NIST mass
Thames Chromatography (Windsor, England) and a 1050 spectra library and interpreted on the basis of the observed
series multiple wavelengthdetector, set at 240 nm. Samples,
injected via a 10 ml Rheodyne loop, were eluted with a
methanol/50 mM ammonium acetate 40/60 v/v solution at a
flow rate of 1 ml/min. HPLC-MS analyses were carried out
by ionspray-mass spectrometry withan API 165 mass
spectrometer (Applied Biosystems/MSD Sciex, Foster City,
CA, USA) equipped withan ionspray interface. Teh
fragmentation.
Determination of aldehydes was carried out by derivati-
zation with pentafluorobenzylhydroxylamine hydrochloride
(PFBHA, Aldrich) according to a published procedure (Le
Lacheur et al., 1993). The derivatives were analysed by gas-
chromatography/electron capture detection (GC-ECD) with
a HP 5890 GC (Agilent Technologies, Palo Alto, CA, USA).
ionspray conditions (positive ions) were: nebulizer gas One microlitre of the of sample was injected into a
(air)=1.2 l/min, curtain gas (nitrogen)=1 l/min, needle
voltage=5500 V, orifice voltage=20 V and ring volt-
30 m  0.25 mm DB-1 column (J&W, Folsom, CA, USA).
The GC oven temperature programme was as follows:
age=200 V. The column and operating conditions were isothermal at 308C for 3 min; 30 to 3008C at 88C/min;
the same as used for the HPLC-UV analyses. The flow from isothermal for 10 min. The injector and detector were kept
the HPLC-UV was split to allow just 50 ml/min to enter the at 250 and 3208C, respectively.
ionspray interface. Quenched samples (isoproturon starting
concentration=10 mg/l) to be analysed by HPLC-MS for
by-products identification were concentrated 150 times
using the following solid-phase extraction (SPE) procedure:
after conditioning a 500 mg Bondelut C-18 solid phase
cartridge (Varian, Harbor City, CA, USA) by loading
sequentially methylene chloride/methanol (1/1 v/v solution),
methanol and water, 150 ml of quenched sample was passed
through the cartridge. After a 5 min drying step, the
RESULTS AND DISCUSSION
Figure 1 shows the isoproturon decay curves
recorded during ozonation carried out in aqueous
solutions at pH=7, and at three different initial
ꢀ
1
ꢀ3
isoproturon concentrations (10, 10 and 10 mg/l).
cartridge was eluted using methanol. The volume of the These decay curves demonstrate that, regardless of
extract was reduced to 1 ml using a gentle stream of purified
the conditions, ozone is very effective in removing
air. The formation of by-products was followed by HPLC-
isoproturon, and it is only the time (5–15 min)
MS, operating in selected ion monitoring (SIM), by
+
required to do so that is affected by the conditions.
monitoring the [M+H] ions of previously identified by-
products. In these determinations, quenched samples with In Fig. 1, as reported in the experimental section, the
an initial concentration of 10 mg/l isoproturon were applied ozone dose in water per minute was 18 and
analysed without any pre-concentration step, while those
1
.8 mg/l for the highest and for the two lowest
withan initial isoproturon concentration of 0.1 and
.001 mg/l were concentrated (100 and 10,000 times,
respectively) by SPE using Carbograph-4 cartridges (0.5
herbicide concentrations, respectively. As for the
transferred ozone [O3 transferred=(O3 consumed
0
and 1 g, respectively, obtained from Lara s.r.l., Italy). The +O
SPE procedure used was the same as that for the C-18
3
residual in water)=(O
off)gas)], in the case of the highest herbicide
3
in the (in)gasꢀO
3
in the
(
cartridges.
The quantitative determination of formic, acetic and
oxalic acids was carried out by ion chromatography using a
concentration (10 mg/l) consumption was measurable
only during the first 15 min, as afterwards the
Dionex 4000i (Dionex, Sunnyvale, CA, USA) LC system difference (O
3
in the (in)gasꢀO
3
in the (off)gas)
working in suppressed ion mode, equipped witha con-
ductivity detector. Samples, injected via a 50 ml loop, were
eluted at a flow rate of 1 ml/min through an OmniPac PAX-
became lower than the analytical experimental error.
During this 15 min period, the measured amount of
transferred ozone was 45.9 mg/l (44 mg/l as con-
sumed and 1.9 mg/l as residual), i.e., the ozone
transfer efficiency was 17% and the ratio (mg/l of
5
00 analytical column (250 Â 4 mm i.d.) witht he following
gradient (solvent A=0.75 mM NaOH with1.25% met ha -
nol; solvent B=50 mM NaOH with1.25% met ah nol):

1698
Giuseppe Mascolo et al.
Fig. 1. Isoproturon structure and ozonation decays at the Fig. 2. HPLC-UV chromatogram of isoproturon by-pro-
pH=7. Experimental conditions: *: [ozone]air=9 mg/l at
ducts after 3 min of ozonation at pH=7. Initial herbicide
l/min, ozone consumed=44 mg/l, residual ozone concen- concentration=10 mg/l. By-products identification numbers
1
tration are in water=1.9 mg/l; &: [ozone]air=9 mg/l at 0.1 l/
min, ozone consumed 5 1 mg/l, residual ozone concentra-
tion in water=0.7 mg/l; m: [ozone]air=9 mg/l at 0.1 l/min,
ozone consumed 51 mg/l, residual ozone concentration in
water=0.7 mg/l.
as reported for chemical structures in Fig. 5. Other
experimental conditions are in the text.
To monitor for degradation by-products, the
approachused was to initially apply the identification
techniques to the samples in which the highest
ozone consumed)/(mg/l of herbicide removed) was isoproturon concentration was used (i.e., 10 mg/l),
4/10=4.4. and then to search for the by-products so identified in
The experimental error associated with the mea- the samples where lower herbicide concentrations
4
ꢀ1
ꢀ3
surement of the consumed ozone when the lower (i.e., 10 and 10 mg/l) were present.
herbicide concentrations were used, renders such a
As is evident in the HPLC-UV chromatogram
measurement meaningless. In these cases, however, shown in Fig. 2, which refers to the highest herbicide
some reasonable assumptions can be applied. Assum- concentration, after a reaction time for isoproturon
ing that the stoichiometry of the reaction between withozone of 3 min several by-products w oh se
ozone and isoproturon is the same, even for the retention times are all shorter than that of isoprotur-
ꢀ
1
ꢀ3
lowest herbicide concentrations (10 and 10 mg/l), on are formed. This means that all the detected by-
the ozone consumed in these cases [4.4 Â mg/l of products are more polar than isoproturon, and that
isoproturon)] should be less than 1 mg/l. Considering they still maintain the aromatic structure as all are
that the residual ozone measured after 15 min in these UV detectable. In order to identify the by-products
cases was 0.7 mg/l, the amount of transferred ozone detected by HPLC-UV, HPLC-MS analyses of
falls into a range lower than 2 mg/l, which represents quenched samples were carried out. Table 1 reports
that actually measured at real water treatment plants the molecular weight (MW) of each detected by-
(Bonnelye and Richard, 1997). When the highest product as well as the major ions in their ionspray
herbicide concentration was used (i.e., 10 mg/ mass spectra. Examining Table 1 it is possible to note
l ffi 50 mM), given that the kinetic constants of ozone that by-products 1–3 have the same MW (222 a.m.u.)
ꢀ
ꢀ1 ꢀ1
s
withisoproturon and OH
(
ions are 141 M
which corresponds to the chemical structure of
ꢀ
1
ꢀ1
De Laat et al., 1996) and 70 M
Hoigne, 1985), respectively, a rapid calculation hydroxylated by-products is not novel as they have
indicates that ozone (after its transfer into water) also been identified during oxidation reaction of
s
(Staehelin and hydroxylated isoproturon. The formation of such
`
will react more rapidly withisoproturon (k [iso-
ꢀ
other phenyl-urea herbicides (Aguer and Richard,
s ) than with hydroxide ion to 1996; Richard and Bengana, 1996). As shown in
3
ꢀ1
ꢀ
proturon] ffi 7 Â 10
ꢀ7 ꢀ1
at pH=6 and Fig. 3a, the presence in the mass spectrum of by-
s at pH=9). It follows that molecular product 3 of an ion of m/z=165 indicates that the
ꢁ
ꢀ
form OH (kOH [OH ] ffi 7 Â 10
s
ꢀ4
ꢀ1
7
 10
ozone is the main species responsible for isoproturon hydroxyl group (OH–) in this by-product is located
degradation because of the high initial concentration on the benzylic carbon of isoproturon structure. In
used. Further evidence of this is that the pH has a fact, 165 a.m.u. is the molecular weight of a hydro-
negligible influence on isoproturon decay because, as xylated isoproturon after the loss of the hydroxy-
3 2
long as isoproturon is present in the solution, propyl group [(CH ) C–OH]. The lack of the ion
ꢁ
OH
formation by reaction of ozone with (m/z=165) in the spectra of by-products 1 and 2
hydroxide ion is very low, whatever the pH. As reported in Fig. 3b, suggests they are isomers with the
isoproturon ozonations carried out in the pH range hydroxyl group located on the isoproturon aromatic
6–9 gave rise to similar decay curves, they have not ring. As for by-product 5, it has a molecular weight
been reported. (180 a.m.u.) which is lower than that of isoproturon

By-products formation during degradation of isoproturon
1699
Table 1. Mass spectral data for isoproturon ozonation by-products analysed by HPLC-MS. By-products identification numbers as reported
for chemical structures in Fig. 5
+ +
[M+Na] , [M+H] and main fragment ion masses (relative intensity)
Derivative
MW
Isoproturon
206
222
222
222
208
180
196
198
212
228
228
229(20), 207(100), 165(18), 72(37)
245(13), 223(100), 181(39), 72(5), 46(9)
245(13), 223(100), 181(32),72(5), 46(8)
245(10), 223(100), 205(7), 165(15), 134(4)
231(30), 209(100), 167(50)
203(18), 181(100), 72(50)
219(100), 197(58), 72(8)
219(100), 197(65), 72(15)
1
2
3
4
5
6
7
8
9
235(30), 213(100), 143(36), 72(9)
251(50), 229(100), 159(40), 89(32), 72(11)
251(65), 229(100), 89(21), 72(15)
1
0
the nitrogen bonded to the ring. This hypothesis is
substantiated by the fact that all these by-products
are UV-detectable. By-products
probably isomers because they have the same MW
228 a.m.u.). Isomer formation during ozonation of
9 and 10 are
(
aryl compounds is very likely because of the lack of
selectivity of hydroxyl radical reactions. However,
for these by-products another structure of the same
MW is also possible in which the aromatic ring has
been opened giving rise to a structure derived from
muconic acid. Considering by-product 4, its MW
(
208 a.m.u.) is compatible witha derivative w ho se
structure is that of isoproturon where a methyl group
CH –) has been replaced by a hydroxyl (OH–).
(
3
Whether the methyl group has been replaced from
the isopropyl [(CH ) CH–] or from N-dimethyl
3
2
3 2
[(CH ) N–] group of the isoproturon molecule, can
be determined by checking the mass spectrum of this
by-product reported in Table 1. The presence of the
ion m/z=167, which corresponds to a fragment
3 2
where the isopropyl group [(CH ) CH–] has been
lost, clearly points to the second alternative. For by-
product 7 (MW=198 a.m.u.), its structure is similar
to by-product
4
(CH ) CH–] has been lost and two more hydroxyl
where the isopropyl group
[
groups have been gained.
3
2
In order to assess whether the above by-products
were formed simultaneously or sequentially, their
formation and/or decomposition withrespect to time
was followed by HPLC-MS. Furthermore, to check
whether such by-products can be formed under
conditions typically used at water treatment plants,
their evolution was monitored using lower isopro-
Fig. 3. Ionspray mass spectra of by-product 3 (a) and 1 (b).
Instrumental conditions in the text.
(206 a.m.u.) suggesting that it arises from some turon initial concentrations (i.e., 0.1 and 0.001 mg/l)
structural breakdown of the parent molecule. In fact, as well as lower (10 times) applied ozone dosage. As
its molecular weight corresponds to that of an shown in Fig. 4, the results show that for all the
isoproturon derivative where the isopropyl side investigated isoproturon concentrations, the primary
chain, hydroxylated (as in the by-product 3) or not by-products numbered as 1–3 are those which are
(as in the isoproturon itself), is replaced by an quantitatively predominant, and that an initial rather
hydroxyl group. By-products 6 and 8–10 (two fast isoproturon oxidation is followed by a slower
isomers) have molecular weights that increase pro- primary by-products (1–3) degradation. Figure 4 also
gressively by 16 a.m.u. from that of by-product 5 demonstrates, as expected, that by lowering the
(180 a.m.u.). This suggests that they are formed by initial isoproturon concentration the concentration
oxidation of by-product 5 by successive introduction of eachprimary by-product maximizes at longer
of hydroxyl groups into the aromatic ring and/or on reaction times. Looking at Fig. 4 it is also possible to

1700
Giuseppe Mascolo et al.
Fig. 4. Evolution of isoproturon by-products at pH=7. The curve numbers identify the detected
compounds whose identification numbers are those reported for chemical structures in Fig. 5.
Experimental conditions (a) and (a’): [isoproturon]
consumed=44 mg/l, residual ozone concentration in water=1.9 mg/l; (b) and (b’): [isoprotur-
on] =0.1 mg/l, [ozone]air=9 mg/l at 0.1l/min, ozone consumed 5 1 mg/l, residual ozone concentration
in water=0.7 mg/l, quenched samples concentrated 100 times; (c) and (c’): [isoproturon] =0.001 mg/l,
ozone]air=9 mg/l at 0.1 l/min, ozone consumed 51 mg/l, residual ozone concentration in water=0.7 mg/l,
quenched samples concentrated 10,000 times.
0
=10 mg/l, [ozone]air=9 mg/l at 1 l/min, ozone
0
0
[
conclude that during real drinking water treatment, methyl). This explains why the by-products identified
the most likely by-products to be formed are those in the present investigation are different from those
numbered from 4 to 10.
found in previous studies, in which the main
ꢁ
On the basis of the above results, the isoproturon degradation pathway was due to reaction of OH
ozonation pathway shown in Fig. 5 has been radical witheit he r a C–N or a C–H bond (Allemane
proposed. As previously explained, in the present et al., 1995; Prados et al., 1995). Moreover, it is
investigation the mechanism of isoproturon degrada- interesting to note that the chemical structures of the
tion is mainly based on the reaction with molecular primary by-products identified during previous in-
ozone. The pathway depicted in Fig. 5 indicates that vestigations (e.g., quinone-, aniline-, benzaldehyde-,
the initial attack of molecular ozone occurs either on nitrobenzene-, isopropylbenzene- and phenol-like
the aromatic ring or on an alkyl group (isopropyl and structures) are structurally simpler than those found

By-products formation during degradation of isoproturon
1701
Fig. 5. Proposed pathway for isoproturon oxidation by ozone.
during the present study (i.e., 1–4 in Fig. 5). This is previous studies (Allemane et al., 1995; Prados et al.,
ꢁ
not unexpected as OH is a stronger oxidant and 1995) which only utilised GC-MS.
more reactive than O . Also, it is important to point
out that during the present investigation because undergo further degradation that lead to their
HPLC-MS was used, it has been possible to identify disappearance. As withlonger reaction times no by-
polar by-products not detectable (if formed) in the products were detected using HPLC-UV, this is
As shown in Fig. 4, by-products 7, 9 and 10
3

1702
Giuseppe Mascolo et al.
probably be due to aromatic ring cleavage by formed (Corless et al., 1990). Accordingly, such
ꢁ
reaction withozone and OH . It is to be expected compounds were looked for and found as described
has already disappeared, dissolved ozone can react plots of detected carbonyl compounds, reported in
that at longer reaction times, when the isoproturon in the experimental section. The quantitative
ꢁ
ꢁ
withOH ion to give rise to OH radical. Ozone Fig. 6, show that formaldehyde, acetone and glyoxal
chemistry, supports such an assumption, as during are the main compounds formed, with their concen-
extensive ozonation of aromatic structures, low MW trations being far greater than that of the others. As
compounds suchas alde hy des, ketones and acids are
for acetone, its formation is probably due to the
carbon–carbon cleavage of the isopropyl group of
by-product 3. Furthermore, Fig. 6 shows that
aldehydes, once formed, are further degraded as
proved by the fact that their concentration regularly
decreases after the first 5 min. From a chemical
viewpoint, the most likely products of aldehyde
oxidation are carboxylic acids. Consequently, as
reported in the experimental section, ozonated
samples were analysed also for organic acids and
the results are reported in Fig. 7. This figure shows
that, during the ozonation, the formation of the
identified acids (i.e., formic, acetic, and oxalic) is
practically continuous, indicating that such products
are the real end by-products of the overall ozonation
reaction. This conclusion was expected considering
the very low values available in literature of the
kinetic constants of the reactions between ozone and
aliphatic carboxylic acids (Hoigne, 1987). The
´
measured concentrations of formic acid, much higher
than those expected from the formaldehyde concen-
Fig. 6. Aldehydes and acetone formation during isoprotur-
on ozonation at pH=7. Experimental conditions: [isopro-
turon]
0
=10 mg/l, [ozone]air=9 mg/l at 1 l/min, ozone trations reported in Fig. 6, prove that this acid is the
consumed=44 mg/l, residual ozone concentration in
water=1.9 mg/l.
end by-product of several previously formed inter-
mediates.
Finally, ozone chemistry also suggests that during
the investigated reaction the formation of a-oxo-
acids can occur. This was confirmed by GC-MS
analyses of methylated samples. As reported in
Table 2, a-oxo-acids withfive or six carbon atoms
were detected. As no authentic standards were
available, they could not be quantified. However,
the peak areas found for these compounds were very
low, suggesting that they were only present at low
concentrations. Presumably, suchcompounds are
formed by breakdown of the aromatic structures of
some isoproturon primary degradation by-products,
and can give rise (by decarboxylation) to lower
molecular weight derivatives.
Fig. 7. Formation of organic acids during isoproturon
ozonation at pH=7. Experimental conditions: [isoprotur-
CONCLUSIONS
on]
0
=10 mg/l, [ozone]air=9 mg/l at 1 l/min, ozone consu-
residual ozone concentration in
water=1.9 mg/l.
med=44 mg/l,
A laboratory-scale investigation has been carried
out to identify the degradation by-products of a
Table 2. Low molecular weight organic acids identified by GC-MS analyses after derivatization with diazomethane
Chemical structure
IUPAC name
CH
CH
CH
CH
3
3
3
3
(CH
CH(CH
CH CH
CH¼¼C(OCH
3
)C¼¼C(OCH
)CH¼¼C(OCH
COCH
)COOCH
3
)COOCH
3
2-butenoic acid, 2-methoxy-3-methyl, methyl ester
2-pentenoic acid, 2-methoxy-4-methyl, methyl ester
hexanoic acid, 3-oxo, methyl ester
3
)COOCH
3 3
2
2
2 3
COOCH
3
3
2-butenoic acid, 2-methoxy, methyl ester

By-products formation during degradation of isoproturon
1703
herbicide during ozonation. Isoproturon, a phenyl- Fielding, M., Barcelo, D., Helweg, A., Galassi, S.,
Torstensson, L., Van Zoonen, P., Wolter, R., Angeletti
urea derivative, was chosen as a model compound
G. (1992) Pesticides in ground and drinking water. Water
because of its widespread use.
On the basis of the results the following conclu-
Pollution Research Report 27, Commission of the European
Communities, Directorate-General for Science, Research
sions have been drawn:
and Development, Brussels, Belgium.
Glaze W. H. (1987) Drinking water treatment withozone.
Environ. Sci. Technol. 21, 224–229.
Glaze W. H., Kang J. W. and Chapin D. H. (1987) The
chemistry of water treatment involving ozone, hydrogen
peroxide and ultraviolet radiation. Ozone Sci. Engng 9(4),
335–342.
*
Under the fixed experimental conditions, some of
which are similar to those encountered at real
water treatment plants, isoproturon is completely
removed by ozone;
*
Hoigne
Water Treatment pp. 121–143. Plenum Press, New York.
Hoigne J. and Bader H. (1976) The role of hydroxyl radical
´
J. (1987) In Stucki S., Process Technologies for
the main reaction of isoproturon is with molecular
ozone;
during ozonation, several intermediates that
maintain the aromatic ring of the isoproturon
´
*
reactions in ozonation processes in aqueous solutions.
Water Res. 10, 377–386.
are formed as a result of primary isoproturon Hoigne J. and Bader H. (1983) Rate constants of reactions
´
of ozone withorganic and inorganic compounds in
water}I. Non-dissociating organic compounds. Water
Res. 17, 173–183.
Huang C. P., Dong C. and Tang Z. (1993) Advanced
chemical oxidation: its present role and potential future
in hazardous waste treatment. Waste Management 13,
degradation;
successive degradations of suchintermediates, by
*
ꢁ
reaction withozone and OH , lead to the break-
down of their aromatic structures and to the
formation of low molecular weight compounds
3
61–377.
suchas alde hy des, simple organic acids and
oxo-acids.
a-
Khadrani A., Seigle-Murandi F., Steiman R. and Vroumsia
T. (1999) Degradation of three phenylurea herbicides
(
chlortoluron, isoproturon and diuron) by micromycetes
From the drinking water quality point of view, an
isolated from soil. Chemosphere 38(13), 3041–3050.
Ku Y., Chang J. L., Shen Y. S. and Lin S. Y. (1998)
Decomposition of diazinon in aqueous solution by
ozonation. Water Res. 32(6), 1957–1963.
issue worthy to be assessed is whether the toxicity of
the identified intermediates is lower or greater than
that of isoproturon.
Lacorte S., Ehresmann N. and Barcelo D. (1996) Persistence
´
of temephos and its transformation products in rice crop
field waters. Environ. Sci. Technol. 30(3), 917–923.
Langlais B., Reckhow D.A. and Brink D.R., Ozone in
Water Treatment, Application and Engineering (pp. 12–
13). Lewis Publishers, Chelsea, MI, USA.
REFERENCES
Aguer J. P. and Richard C. (1996) Transformation of
fenuron induced by photochemical excitation of humic Legrini O., Oliveros E. and Braun A. M. (1993) Photo-
acids. Pesticide Sci. 46, 151–155.
chemical processes for water treatment. Chem. Rev. 93,
671–698.
Le Lacheur R. M., Sonnenberg L. B., Singer P. C.,
Christman R. F. and Charles M. J. (1993) Identification
of carbonyl compounds in environmental samples.
Environ. Sci. Technol. 27, 2745–2753.
Madhun Y. A. and Freed V. H. (1987) Degradation of the
herbicides bromacil, diuron and chlortoluron in soil.
Chemosphere 16(5), 1003–1011.
1
Aldrich(1993) Diazald
generators, technical information bulletin, No. AL-180.
Allemane H., Prados-Ramirez M., Croue J. P. and Legube
, MNNG and diazomethane
´
B. (1995) Identification of the initial isoproturon oxida-
tion by-products resulting from the ozone/hydrogen
peroxide system. Rev. Sci. l’Eau 8, 315–331.
Bonnelye V. and Richard Y. (1997) Changes in ozone
demand of water during the treatment process. Ozone Sci.
Engng 19, 339–350.
¨
Mascolo G., Lopez A., Foldenyi R., Passino R. and
Brambilla A. M., Calvosa L., Monteverdi A., Polesello S.
and Rindone B. (1993) Ozone oxidation of polyethoxy-
lated alcohols. Water Res. 27(8), 1313–1322.
Camel V. and Bermond A. (1998) The use of ozone and
associated oxidation processes in drinking water treat-
ment. Water Res. 32(11), 3208–3222.
Tiravanti G. (1995) Prometryne oxidation by sodium
hypochlorite in aqueous solutions: kinetics and mechan-
ism. Environ. Sci. Technol. 29(12), 2987–2991.
Mascolo G., Lopez A., Passino R., Ricco G. and
Tiravanti G. (1994) Degradation of sulphur containing
s-triazines during water chlorination. Water Res. 28(11),
2499–2506.
Corless C. E., Reynolds G. L., Graham N. J. D. and Perry
R. (1990) Ozonation of pyrene in aqueous solution. Meijers R. T., Oderwald-Muller E. J., Nuhn P. A. N. and
Water Res. 24(9), 1119–1123.
Kruithof J. C. (1995) Degradation of pesticides by
ozonation and advances oxidation. Ozone Sci. Engng 17,
673–686.
O’Brien R. P., Rizzo J. L., Schuliger and W. G. (1987) In
Control of Organic Substances in Water and Wastewater,
ed. Berger B. B., Noyes data Corporation, Park Ridge,
NJ, USA (Chapter 11).
Ollis D. F., Pelizzetti E. and Serpone N. (1991) Destruction
of water contaminants. Environ. Sci. Technol. 25(9),
1523–1529.
Pieuchot M., Perrin-Ganier C., Portal J-M. and
Schiavon M. (1996) Study of the mineralization and
degradation of isoproturon in three soils. Chemosphere
33(3), 467–478.
Cullington J. E. and Walker A. (1999) Rapid biodegrada-
tion of diuron and other phenylurea herbicides by a soil
bacterium. Soil Biol. Biochem. 31(5), 677–686.
De Laat J., Maouala-Makata P. and Dore M. (1996) Rate
`
constants for reactions of ozone and hydroxyl radicals
withseveral pe hn yl-ureas and acetamides.
Technol. 17, 707–716.
Di Giano F. A. (1987) In Control of Organic Substances in
Water and Wastewater, ed. Berger, B. B., Noyes data
Corporation, Park Ridge, NJ, USA Chapter 3.
EPA (1986) Municipal Wastewater Disinfection: Design
Manual. EPA/625/1-86/021, Cincinnati, OH, USA.
Esposito E., Paulillo S. M. and Manfio G. P. (1998)
Biodegradation of the herbicide diuron in soil by
indigenous actinomycetes. Chemosphere 37(3), 541–548.
Environ.
Prados M., Roche P. and Allemane H. (1995) State of the
art in the pesticides oxidation field. In Proceedings of the

1704
Giuseppe Mascolo et al.
1
2th Ozone World Congress, Lille, France, 15–18 May, Roche P. and Prados M. (1995) Removal of pesticides by
use of ozone or hydrogen peroxide/ozone. Ozone Sci.
Engng 17, 657–672.
Vol. 1, pp. 99–113.
Prados M. and Ciba N. (1997) Atrazine by-products
formation in xidation processes. In Proceedings of the Staehelin J. and Hoigne J. (1985) Decomposition of ozone
`
13th Ozone World Congress, Kyoto, Japan, 26–31
October, Vol. 2, pp. 651–656.
Pramauro E., Vincenti M., Augugliano V. and Palmisano L.
in water in the presence of organic solutes acting as
promoters and inhibitors of radical chain reactions.
Environ. Sci. Technol. 19(12), 1206–1213.
(
1993) Photocatalytic degradation of monouron in aqu-
eous TiO dispersions. Environ. Sci. Technol. 27(9), 1790–
795.
Standard Methods for the Examination of Water and
Wastewater, 20thed. (1999) American Public Health
Association (APHA), American Water Works Associa-
tion (AWWA) and Water Environment Federation
(WEF), Washington, DC, USA.
Tomlin, C. D. S. Ed. (1997) The Pesticide Manual. 11th
edition, BritishCrop Protection Council, Farn ha m, UK.
Vroumsia T., Steiman R., Seigle-Murandi J., Bennoit-
Guyod L. and Khadrani A. (1996) Biodegradation of
three substituted phenylurea herbicides (chlortoluron,
diuron and isoproturon) by soil fungi. A comparative
study. Chemosphere 33(10), 2045–2056.
2
1
Reynolds G., Graham N., Perry R. and Rice R. G. (1989)
Aqueous ozonation of pesticides: a review. Ozone Sci.
Engng 11, 339–383.
Rice R. G. (1997) Application of ozone for industrial
wastewater treatment}a review. Ozone Sci. Engng 18,
4
77–515.
Rice R. G. and Netzer A. (1984) Handbook of Ozone
Technologyand Applications , Vol. 2, Ann Arbor Science
Publisher Inc., Ann Arbor, MI, USA.
Richard C. and Bengana S. (1996) pH effect in the
photocatalytic transformation of a phenyl-urea herbicide.
Chemosphere 33(4), 635–641.
Yao J. J., Huang Z. H. and Masten S. J. (1998) The
ozonation of pyrene: pathway and product identification.
Water Res. 32(10), 3001–3012.