Structural elucidation of metolachlor photoproducts by liquid chromatography/high-resolution tandem mass spectrometry

Article abstract of DOI:10.1002/rcm.7382

Rationale Metolachlor is one of the most intensively used chloroacetanilide herbicides in agriculture. It has been detected in water; consequently, under UV-visible irradiation, it can be transformed in degradation products (TPs). The structures of TPs were elucidated by liquid chromatography/high-resolution tandem mass spectrometry (LC/HR-MS/MS). The potential toxicities of these TPs were estimated by in silico tests. Methods Aqueous solutions of metolachlor were irradiated in a self-made reactor equipped with a mercury vapor lamp. Analyses were carried out using high-performance liquid chromatography coupled to quadrupole time-of-flight (QTOF) mass spectrometer. High-resolution m/z measurements, MS/MS and isotopic labeling experiments allowed structural elucidation of metolachlor TPs. Their toxicities were estimated in silico, using the T.E.S.T. program. Results Ten major metolachlor photoproducts were characterized by LC/MS/MS after irradiation of metolachlor in aqueous solution. Elucidation of their chemical structures was identified using high-resolution measurements and MS/MS experiments. They resulted from the combination of dehalogenation, hydroxylation and cyclisation processes. The potential oral rat lethal dose (LD50) was assessed with QSAR tests for metolachlor and each photoproduct. Results indicate that most of the TPs are much more toxic than metolachlor. Conclusions UV-vis irradiation of metolachlor in aqueous solution leads to the formation of ten photoproducts. QSAR estimations show that the location of added hydroxyl group(s) is of key relevance as regards to biological activity and that routine water analysis should take into account the TPs are more toxic than the parent molecule.

Full text of DOI:10.1002/rcm.7382

Research Article
Received: 1 April 2015
Revised: 24 August 2015
Accepted: 30 August 2015
Published online in Wiley Online Library
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286
(wileyonlinelibrary.com) DOI: 10.1002/rcm.7382
Structural elucidation of metolachlor photoproducts by liquid
chromatography/high-resolution tandem mass spectrometry
*
Édith Nicol, Christophe Genty, Stéphane Bouchonnet and Sophie Bourcier
École Polytechnique, Laboratoire de Chimie Moléculaire, UMR 9168 CNRS, route de Saclay, 91128 Palaiseau cedex, France
RATIONALE: Metolachlor is one of the most intensively used chloroacetanilide herbicides in agriculture. It has been
detected in water; consequently, under UV-visible irradiation, it can be transformed in degradation products (TPs).
The structures of TPs were elucidated by liquid chromatography/high-resolution tandem mass spectrometry (LC/HR-
MS/MS). The potential toxicities of these TPs were estimated by in silico tests.
METHODS: Aqueous solutions of metolachlor were irradiated in a self-made reactor equipped with a mercury vapor
lamp. Analyses were carried out using high-performance liquid chromatography coupled to quadrupole time-of-flight
(QTOF) mass spectrometer. High-resolution m/z measurements, MS/MS and isotopic labeling experiments allowed
structural elucidation of metolachlor TPs. Their toxicities were estimated in silico, using the T.E.S.T. program.
RESULTS: Ten major metolachlor photoproducts were characterized by LC/MS/MS after irradiation of metolachlor in
aqueous solution. Elucidation of their chemical structures was identified using high-resolution measurements and
MS/MS experiments. They resulted from the combination of dehalogenation, hydroxylation and cyclisation processes.
The potential oral rat lethal dose (LD50) was assessed with QSAR tests for metolachlor and each photoproduct. Results
indicate that most of the TPs are much more toxic than metolachlor.
CONCLUSIONS: UV–vis irradiation of metolachlor in aqueous solution leads to the formation of ten photoproducts.
QSAR estimations show that the location of added hydroxyl group(s) is of key relevance as regards to biological activity
and that routine water analysis should take into account the TPs are more toxic than the parent molecule. Copyright ©
2015 John Wiley & Sons, Ltd.
The use of pesticides in agriculture is considered essential to
optimize yields, but their intensive use leads to
contamination of water and soils. Herbicides inhibit the
growth of weeds by interfering with bioenergetic
pathways.[1] The frequent use of herbicides with high
persistence in soil has resulted in systematic detection of these
compounds in surface water and groundwater. Consequently,
all the common water treatment processes such as hydrolysis,
biodegradation, chemical oxidation and UV irradiation are
still under development.[2–5] Even if increasingly efficient,
these treatments do not lead to full mineralization of
pollutants and produce transformation products which can
UV photolysis processes are commonly used as an alternative
to chlorine disinfection. Some efforts have been devoted to
pesticide direct and catalyzed photodegradation, including
the identification of photoproducts, kinetic studies and
establishment of reaction pathways, in order to integrate
these TPs in pesticide control.[3] These studies showed
that photolysis, in combination with other techniques
multiplying the ways of degradation, would target a large
range of pollutants. Independently of the treatment methods
used, the knowledge of the chemical structures of TPs would
permit analytical methods to be established able to target
those specific compounds and to predict their impact on
the environment.
contaminate the environment.[6] In
a 2013 review on
emerging pollutants, Aguera et al. stated that analytical
trends are no more limited to quantitation of listed
contaminants, but also involve identification of by-products
generated by water treatment processes.[7] Several studies
have established that some persistent transformation
products (TPs) could prove to be more toxic than the parent
molecules.[3,8,9] These new contaminants are released in the
nature and their stability and persistency have pointed out
their potential occurrence as environmental contaminants.
Metolachlor (Meto) is a chloroacetanilide herbicide which
has been widely used to treat corn, soybeans, peanuts and
potatoes before being banned in 2006. High concentrations
of metolachlor in surface water and groundwater have been
reported by several authors.[10–14] Different authors have
identified photoproducts using gas chromatography/mass
spectrometry (GC/MS).[15–17] Complementing these works,
we recently investigated the characterization of some other
metolachlor photodegradation products by GC/MS
coupling.[18] In this study, post-photolysis kinetic studies
showed that most of the photoproducts have a lifetime
allowing their persistence in the environment. This is of great
concern considering that some of them were evaluated to be
significantly more toxic than metolachlor.[19] With the aim of
covering a larger range of polarities, the chemical structures
* Correspondence to: S. Bourcier, École Polytechnique,
Laboratoire de Chimie Moléculaire, UMR 9168 CNRS,
route de Saclay, 91128 Palaiseau cedex, France.
E-mail: sophie.bourcier@polytechnique.edu
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286
Copyright © 2015 John Wiley & Sons, Ltd.

É. Nicol et al.
of some degradation products of chloroacetamides were
investigated using both GC/MS/MS and LC/MS/MS except
for metolachlor for which only TPs characterized by
GC/MS/MS have been reported.[19] Gutowski and
coworkers[20] have recently compared TPs formed by
photolysis of s-metolachlor and its commercial product
Mercantor Gold. They concluded that the TPs formed are
the same and that the formation of TPs from Mercantor Gold
is faster. In this work, the identification of the structures of
TPs was carried out using MetaPC software.
The present work had two goals; the first one was to
complete the characterization of the polar photoproducts
formed by direct photolysis of metolachlor using high-
resolution MS, MS/MS experiments and isotopic labelling.
The second one was to assess the potential toxicity
(developmental toxicity and lethal oral doses in rats) of
photoproducts using the software T.E.S.T.[21]
and non-sonicated aqueous solutions of metolachlor. Two
irradiation times were tested (5 and 60 min) to evaluate the
persistence of photoproducts.
LC/MS experiments
Liquid chromatography/tandem mass spectrometry experiments
were carried out with electrospray ionization (+ ion mode)
with a QTOF Premier instrument equipped with a Z-spray
electrospray source (Waters, Saint Quentin-en-Yvelines,
France). A model 2690 liquid chromatography module from
Waters was coupled to the QTOF premier mass spectrometer.
The analytical column used was a Pursuit XRs Ultra (2.8 μm
C18 50 × 2.0 mm; Varian, Les Ulis, France). HPLC solvents
were acetonitrile (A) and water (B) both with 0.1% FA. The
following program of linear gradient elution was applied:
40% of A for 11 min and from 11.1 to 20 min, 100% of A. The
column was reconditioned with 40% of A for 10 min between
injections. The effluent was introduced at a rate of 0.2 mL
minÀ1 into the Z-spray interface. Ion source parameters were
adjusted as followed: the cone voltage (Vcone) ranged from 20
to 80 V while the capillary voltage was set to 2.6 kV. Typical
values for the other source parameters were 2 V for the
extraction cone and 2.4 V for the ion guide. Source and
desolvation temperatures were set to 120 °C and 450 °C,
respectively. Nitrogen was used as both nebulizing and
desolvation gas. Gas flows ranged from 70 L.hÀ1 to 700 L.
hÀ1. The volume injected was 10 μL. Argon was used as
collision gas at a flow rate of 0.28 mL minÀ1 corresponding
to a pressure of ca. 4 × 10À3 mbar. To record the MS/MS
spectra, collision energies (Ecol) were scanned from 2 to 15 eV
for each compound in order to obtain the main characteristic
ions and their decomposition pathways. Three acquisition
modes were used to characterize each compound: (i) full scan
mode in V-mode acquisition to identify the protonated
molecules of the photoproducts; (ii) MS/MS of the precursor
ion; and (iii) full scan in W-mode acquisition for high
resolution. The concomitant analysis of metolachlor-D6
photoproducts was achieved in order to gain additional
structural information. Only the most abundant by-products
were taken into account as they are likely to be present in
environmental matrices. For accurate mass measurements,
spectra were acquired using both the W-mode and an
independent reference spray via the LockSpray interface.
Phosphoric acid was used as the lock mass at a flow rate of
10 μL minÀ1. The LockSray frequency was set to 0.1 Hz and
data for the reference compounds were averaged over 10
EXPERIMENTAL
Solutions
All chemicals except metolachlor-D6 were obtained from
Sigma-Aldrich (Saint Quentin Fallavier, France) and were
used as received (98% purity). Metolachlor-D6 was purchased
from C.I.L. Cluzeau Info Labo (Ste Foy-la-Grande, France).
1 mg mLÀ1 solutions were prepared in water and stored
at À20 °C. These solutions were then diluted at 10À6 M in
50:50 mixture of a H2O/CH3CN acidified with formic acid
(FA) (0.1%). The chemical structures of metolachlor and
metolachlor-D6 are displayed in Table 1.
Photodegradation experiments and sample preparation
A home-made photoreactor equipped with a high-pressure
mercury lamp (HPL-N 125W/542 E27 SG; Philips, 94856 Ivry
sur Seine, France) was used for photolysis experiments.
UV-visible irradiations were carried out in quartz tubes
containing 10 mL of an aqueous solution of metolachlor at
1 mg mLÀ1. The lamp was placed into the inner part of the
reactor and cooled by water circulation to avoid uncontrolled
heating of the irradiated solution; the temperature was thus
maintained at 25 2 °C. The luminous flux emitted from
the HPL-N lamp was reported by the manufacturer to be
6200 lm. The solution was stirred during the irradiation time
with a sonicator (Bioblock Scientific, Illkirch, France). No
noticeable effects were observed when analyzing sonicated
spectra minÀ1
. The accurate mass and the elemental
composition of all ions were obtained by using the MassLynx
software. The ion used for the mass correction was m/z
294.9385. The precision of mass measurement for the
determination of elemental composition was less than 5 ppm.
Table 1. Names and chemical structures of the studied
compounds
Structure
Name
X
MW
Computer-aided toxicity prediction
Metolachlor (Meto)
Metolachlor D6 (MetoD6)
H
D
283
289
The Toxicity Estimation Software Tool (T.E.S.T.) is an
Environmental Protection Agency computerized predictive
system available online with Quantitative Structure Activity
Relationships (QSAR) mathematical models.[21] T.E.S.T. has
a variety of toxicity endpoints used to predict chemical
toxicity values from physical properties of molecules. The
QSAR model uses a simple linear function of molecular
wileyonlinelibrary.com/journal/rcm
Copyright © 2015 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286

Elucidation of metolachlor photoproducts by LC/MS/MS
descriptors (such as the octanol-water partition coefficient,
molecular weight or the number of benzene rings, for
instance): Toxicity = ax1 + bx2 + c, where x1 and x2 are
independent descriptor variables and a, b, and c are fitted
parameters. Models for assessing toxicity solely from
molecular structure are based on information-rich structural
descriptors that quantify transport, bulk, and electronic
attributes of a chemical structure.
and 60 min (Fig. 1(c)). The phototransformation process is
very fast, since most of the TPs appear after only 5 min of
irradiation. After 60 min of irradiation, the chromatographic
peaks corresponding to most of the TPs are more intense than
that of metolachlor, only 5% of the initial amount of
metolachlor being still detected. The protonated molecular
ions of the photoproducts were easily identified in the
corresponding mass spectra and confirmed in most cases by
the presence of a sodium ion adduct MNa+.
Table 2 lists the retention times and m/z ratios of the
protonated molecule MH+ and CID (collision-induced
dissociation) product ions of these protonated molecules.
Letters and numbers in the second line of Table 2 refer to the
ionic structures displayed in Fig. 2, which summarizes the
dissociation pathways typically observed for chloroacetamides.
These dissociation pathways were established on the basis of
previous works devoted to photolysis of acetochlor.[22] All the
protonated molecules of metolachlor-D6 photoproducts have
m/z values corresponding to those obtained with photolyzed
metolachlor shifted by 6 u. The comparison of our MS/MS data
RESULTS AND DISCUSSION
Characterization of the degradation products of
metolachlor
The first step of TP identification consisted in comparing the
chromatograms of irradiated and non-irradiated solutions of
metolachlor and metolachlor-D6. Figure 1 displays the
comparison between the reference solution (metolachlor)
(Fig. 1(a)) and the same solution irradiated for 5 min (Fig. 1(b))
(a)
(b)
(c)
Figure 1. LC/MS chromatograms recorded for the metolachlor reference
solution (a, * impurities), the same solution irradiated for 5 min (b), and the
same solution irradiated for 60 min (c).
Table 2. Main CID mass spectra of MH+ (Vcone = 20 V, Ecol = 15 eV) obtained by LC/MS using ESI-P ionization mode for
metolachlor and its transformation products (TPs)
Compound
RT (min) MH+
Ions (m/z)
A1 A1, 1 A2 A2, 1 A2, 2 B1 B2 B3 B1, 1 B1, 1, 2 AB AB1 AB2 AB3 AB3, 1
A
B
TP1
TP2
TP3
0.7
0.7
0.9
300 268 282
284 252 266 194
284 252 266
211 193
264
248
B1, 2
206
250 232 222 193 165
234 216
234 216
177 149
TP4
TP5
TP6
TP7
TP8
TP9
TP10
Meto
0.9
1.0
1.1
1.3
1.5
3.2
3.5
13.3
314 282 296
298 266 280
284 252 266
284 252 266 194
266 234 248
298 266
266 234
284 252
278 268
264 246 236 207 179
248 230 220
234 216 206 177 149
209 191 181
195 177
238
234
216
206 177
177
149
206
224 206
176
176
234
177 248
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286
Copyright © 2015 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

É. Nicol et al.
Figure 2. Decomposition pathways of protonated molecule of photoproducts (Y = H
for TP1, TP3–TP6, TP8 and TP9; Y = OH for TP2, TP7 and TP10).
with those of Gutowski and coworkers[20] shows that many
fragment ions are common but not all the molecular ions.
For all the photoproducts, the loss of CH3OH from MH+ in
mass spectrometry, combined with the fact that none of the
six deuterium atoms are eliminated, demonstrate that the
ether function and the adjacent alkyl chain were not affected
by irradiation. Isotopic distributions show that the chlorine
atom was removed from all photoproducts. This is in good
agreement with literature data according to which the
photolysis decomposition of chlorinated compounds in
aqueous solution usually begins by dechlorination.[15,18]
Transformation products seem to result mainly from hydroxy
additions; they can be divided into five groups according to
their molecular weight. The first group consists of TPs with
Table 3. Mass-to-charge ratios and elemental compositions of protonated molecules and attributed structures for the
transformation products of metolachlor
Compound
TP1
TP2
TP3
TP4
TP5
TP6
Elemental composition
C15H26NO5
300
C15H26NO4
284
C15H26NO4
284
C15H24NO6
314
C15H24NO5
298
C15H26NO4
284
MH+ (m/z)
Structure
Compound
TP7
C15H26NO4
284
TP8
C15H24NO3
266
TP9
C15H24NO5
298
TP10
C15H24NO3
266
Elemental composition
MH+ ( m/z)
Structure
wileyonlinelibrary.com/journal/rcm
Copyright © 2015 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286

Elucidation of metolachlor photoproducts by LC/MS/MS
a raw formula of C15H23NO3 (MW 265) corresponding to
[metolachlor–Cl+H+O]: TP8 and TP10. The second group
includes TPs with a raw formula of C15H25NO4 (MW 283)
corresponding to [meto–Cl+3H+2O]: TP2, TP3, TP6 and
TP7. The third group includes TPs with a raw formula of
C15H23NO5 (MW 297) corresponding to [meto–Cl+H+3O]:
TP5 and TP9. Finally, the raw formulas of TP1 (C15H25NO5,
MW 299) and TP4 (C15H23NO6, MW 313) correspond to
[meto–Cl+3H+3O] and [meto–Cl+H+4O], respectively.
The chemical structures attributed to the photoproducts are
given in Table 3. Only TP10, which results from direct
substitution of the chlorine atom by a hydroxyl group, has
been previously reported in literature.[14,15,18] Ions in Table 2
show that the fragmentation of TP10H+ is logically similar
to that of MetoH+. For the other TPs, the loss of CH3OH from
MH+ competes with that of H2O. If one excepts the CID mass
spectra of MetoH+, TP2H+and TP10H+, the CID mass spectra
of all TPH+ include odd m/z fragment ions (see Table 2)
showing evidence that the nitrogen atom has been eliminated
in the dissociation process. This is confirmed by exact m/z
measurements performed in the W-mode. These ions are not
mass-shifted when irradiating metolachlor-D6. This means
that no deuterium atom is retained in their chemical
structures. Given the structure of metolachlor, NH=C(CH3)2
can be expelled from [TP1H–CH3OH]+ only if the photolysis
leads to a cyclization process involving the aromatic ring
and the CH2CO group. The possible structures arising from
this cyclization process are presented in Fig. 3. We recently
performed
a similar study about the photolysis of
acetochlor-D11.[22] For this particular labeled compound, all
the hydrogen atoms of both the cycle and its alkyl
substituents are replaced by deuterium atoms, and we
showed that the corresponding degradation products
retained all the deuteriums. Only the five-membered ring
structures C and D (Fig. 3) allow deuterium atoms to be kept.
Mechanisms for the formation of monohydroxylated species
including the structure C are proposed in Fig. 4. Photo-
induced elimination of the chlorine atom through homolytic
cleavage of the C–Cl bond can be followed by direct hydroxy
substitution leading to the compound referred to as TPa
monoOH in Fig. 4. The elimination of Cl. can also be followed
by the formation of a five-membered ring, as mentioned
above, leading to three isomeric forms of [meto–Cl]., which
result from electron delocalization on the aromatic ring. By
reaction with water, three monohydroxylated TPs can be
formed, which are named as TPx monoOH with x = b, c or d
in Fig. 4. In turn, these TPs can be photolysed to provide
compounds including one or two additional hydroxyl groups.
As an example, the NH=C(CH3)2 elimination pathway from
[TP1H–CH3OH]+ is given in Fig. 5. Depending on the TP
considered, this elimination (À57 u) is observed from A ions
to form A2 ions for TP1, TP5, TP6 and TP8, from AB ions to
form AB3 ions for TP1, TP3, TP4, TP6 and TP7, from B1,1 ions
Figure 3. Possible structures allowing the consecutive
eliminations of MeOH and NH=C(CH3)2 from MH+ in mass
spectrometry for TPs resulting from irradiation-induced
cyclisation.
Figure 4. Proposed photodegradation pathways leading to the formation of
monohydroxylated species (TPs monoOH) from metolachlor.
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286
Copyright © 2015 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

É. Nicol et al.
tertiary carbocation). Elimination of water from TP10H+ and
TP9H+ is not observed: from TP10H+, it would result in a
primary non-stabilized carbocation whereas from TP9H+, it
would lead to a vinylic non-stabilized cation. The three
hydroxyl groups on TP9 were thus assumed to be added on
sp2 carbon atoms of the six-membered ring since protonation
of a hydroxyl group carried by a sp2 carbon atom usually
does not lead to elimination of water.[23] Figure 6 shows
how the CID mass spectra of [TPH–H2O]+ ions at m/z 266,
recorded with the same collision energy (10 eV) for TP2,
TP3, TP6 and TP7, may allow differenciation of the four di-
hydroxylated isomeric photoproducts since their CID spectra
both differ by the nature and relative intensities of the
fragment ions.
Supplementary Figs. S1 and S2 (see Supporting Information)
propose dissociation pathways for TP2H+ and TP3H+; the
nomenclature of ions is the same as in Fig. 2. In both cases,
the relative intensity of B ions at m/z 266 is high, consistent
with the formation of a more stable tertiary cation. Water
removal from the ions of m/z 266 to produce ions of m/z 248
occurs mainly for TP3. In the case of TP2, water elimination
after proton transfer on the hydroxyl group in the alpha
position of the ketone function would have led to an unstable
ion. In both cases, the B ion (m/z 266) loses a methanol
molecule leading to AB (m/z 234). In the case of TP2, the
formation of the m/z 176 ion from 2AB through CHOHCO
elimination confirms the unambiguous location of a hydroxyl
group in the alpha position of the ketone function. Cyclization
of 2AB leads to a structure allowing elimination of a water
molecule to generate the 2AB1 ion at m/z 216 (see
Supplementary Fig. S1, Supporting Information). In the case
of TP3, the 3B1 ion loses CH2CO to generate 3B1,2 (m/z 206);
this shows that the chlorine atom has been substituted by a
hydrogen atom during the photolysis process (see
Supplementary Fig. S2). The dissociation pathways of TP6H+
and TP7H+ are presented in Supplementary Figs. S3 and S4,
respectively. In both cases, the m/z 266 ion is barely visible in
Fig. 6 and the formation of m/z 234 by elimination of CH3OH
from [TPH–H2O]+ is observed. The hydroxyl group and the
double bonds promote cyclization of the resulting 6AB ion.
This ion can lose either H2O or CO leading to the formation
Figure 5. Mechanisms postulated for the consecutive
eliminations of CH3OH, NH=C(CH3)2 and H2O from [TP1H]+.
to form B1,1,2 ions for TP9 (see Table 2 and Fig. 2).
Consequently, TP1 and TP3 to TP9 were assumed to include
a ’C-type’ skeleton in their structure.
To attribute a structure to each TP, all the fragmentation
pathways had to be considered. Cleavages leading to
characteristic neutral losses of 32 u for CH3OH and 18 u for
H2O were assumed to result from the protonation of the
oxygen atom of either an O-CH3 or OH group; protonation
may directly occur on this oxygen atom, or result from
intramolecular proton transfer. The loss of 42 u (H2C=CO)
or 58 u (CH(OH)=CO) from A [MH–CH3OH]+ or B [MH–H2O]+
indicates the nature of the atom or group (H or OH) that has
replaced the chlorine atom after homolytic cleavage of the
C–Cl bond. In the case of TP2, the presence of the A1 ion
resulting from CH(OH)=CO elimination from A allows the
unambiguous location of one hydroxyl group at the initial
position of the chlorine atom. The loss of 28 u for CO indicates
the presence of a carbonyl function, high-resolution m/z
measurements allowing the distinction between CO and
C2H4 losses. The relative intensities of ions may also give
insight about the site of hydroxylation. For example, for TP1
to TP5 and TP8, the [MH–H2O]+ ion is of much greater
abundance than for the other photoproducts because the
resulting cation is significantly more stable (conjugated
Figure 6. Comparison of CID mass spectra of [TPH–H2O]+ m/z 266 (ion B) for
TP2, TP3, TP6 and TP7.
wileyonlinelibrary.com/journal/rcm
Copyright © 2015 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286

Elucidation of metolachlor photoproducts by LC/MS/MS
relevance as regards biologial activity. This study clearly
shows that routine water analysis should also take into
account some degradation products and not only the listed
environmental pollutants.
REFERENCES
[1] P. Böger, B. Matthes, J. Schmalfuß. Towards the primary
target of chloroacetamides – new findings pave the way.
Pest. Manag. Sci. 2000, 56, 497.
[2] M. Petrovic, S. Gonzalez, D. Barcelo. Analysis and removal
of emerging contaminants in wastewater and drinking
water. TRAC, Trends Anal. Chem. 2003, 22, 685.
[3] M. L. Hladik, E. J. Bouwer, A. L. Roberts. Neutral
degradates of chloroacetamide herbicides: occurrence in
drinking water and removal during conventional water
treatment. Water Res. 2008, 42, 490.
[4] J. Restivo, J. J. M. Órfão, S. Armenise, E. Garcia-Bordejé,
M. F. R. Pereira. Catalytic ozonation of metolachlor under
continuous operation using nanocarbon materials grown on
a ceramic monolith Original. J. Hazard. Mater. 2012, 239, 249.
[5] S. Brosillon, L. Lhomme, D. Wolbert. Modelling of a falling
thin film deposited photocatalytic step reactor for water
purification: Pesticide treatment Original. Chem. Eng. J.
2011, 169, 216.
[6] S. Sanches, M. T. B. Crespo, V. J. Pereira. Drinking water
treatment of priority pesticides using low pressure UV
photolysis and advanced oxidation processes. Water Res.
2010, 44, 1809.
Figure 7. In silico predicted toxicity values for oral rat LD50
for metolachlor and its photoproducts.
of m/z 216 and m/z 206, respectively. The elimination of water
from m/z 266 is observed for TP6 due to the stabilization of the
carbocation formed. This ion obtained at m/z 248 can give by
elimination of CH3OH the ion at m/z 216. Two decomposition
pathways allow the formation of 6AB1 according a higher
relative intensity in the mass spectrum (Fig. 6). From 7AB
(Supplementary Fig. S4, see Supporting Information), we can
note that the elimination of CO is observed because the
hydroxyl group can stabilize the carbocation formed. The
elimination of H2O from AB is observed for TP6 whereas this
cannot happen from 7AB.
[7] A. Aguera, M. J. M. Bueno, A. R. Fernandez-Alba. New
trends in the analytical determination of emerging
contaminants and their transformation products in
environmental waters. Environ. Sci. Pollut. Res. 2013, 20, 3496.
[8] A.B.A. Boxall, C. J. Sinclair, Assessment of the
environmental properties and effects of pesticide
degradation products. Pesticide behaviour in soils and
water, British Crop Protection Council symposium
proceedings. 2001, 78, 113.
[9] T. C. Cavalier, T. L. Lavy, J. D. Mattice. Persistence of
selected pesticides in ground-water samples. Ground Water
1991, 29, 225.
[10] D. Zaruk, M. Alaee, E. Sverko, M. Comba. Occurrence of
triazine herbicides and metolachlor in the Niagara River
and other major tributaries draining into Lake Ontario.
Anal. Chim. Acta 1998, 376, 113.
In silico toxicity prediction
The potential oral rat lethal dose (LD50) was assessed with
QSAR tests for metolachlor and each photoproduct (Fig. 7).
The oral rat lethal dose 50 (LD50) indicates the amount of
chemical in mg/kg body weight that would cause the death
of 50% of a test population of rats after oral ingestion. With
the exception of TP10, which exhibits a potential toxicity
lower than that of metolachlor, all photoproducts are
expected to be much more toxic than metolachlor which is,
according to the toxicity scale classification of Hodge and
Sterner,[24] moderately toxic. These results are within QSAR
estimations but such a difference in predicted toxicity
between the parent compound and most of its by-products
suggest that ’real’ bioassays should be carried out.
[11] J. E. Barbash, G. P. Thelin, D. W. Kolpin, R. J. Gilliom. Major
herbicides in ground water: results from the National Water-
Quality Assessment. J. Environ. Qual. 2001, 30, 831.
[12] L. Guzzella, F. Pozzoni, G. Giuliano. Herbicide
contamination of surficial groundwater in Northern Italy.
Environ. Pollut. 2006, 142, 344.
[13] V. Roubeix, V. Fauvelle, J. Tison-Rosebery, N. Mazzella,
M. Coste, F. Delmas. Assessing the impact of
chloroacetanilide herbicides and their metabolites on
periphyton in the Leyre River (SW France) via short term
growth inhibition tests on autochthonous diatoms.
J. Environ. Monit. 2012, 14, 1655.
[14] J. Kochany, R. J. Maguire. Sunlight photodegradation of
metolachlor in water. J. Agric. Food Chem. 1994, 42, 406.
[15] C. L. Wu, H. Shemer, K. G. Linden. Photodegradation of
metolachlor applying UV and UV/H2O2. J. Agric. Food
Chem. 2007, 55, 4059.
[16] A. D. Dimou, V. A. Sakkas, T. A. Albanis. Metolachlor
photodegradation study in aqueous media under natural
and simulated solar irradiation. J. Agric. Food Chem. 2005,
53, 694.
CONCLUSIONS
Ten structures of metolachlor photoproducts were
characterized by LC/MS/MS. Two groups of compounds
have been found: (1) monocyclic photoproducts resulting
from dechlorination of metolachlor followed by addition of
a hydroxyl function; (2) bicyclic photoproducts arising from
dechlorination of metolachlor followed by addition of one to
four hydroxyl groups. Only some monohydroxylated
compounds could be identified by GC/MS in a previous
work.[19] The combined use of CID experiments and labeled
compounds allowed the identification of new cyclic
structures. The oral rat LD50 predicted values for these
compounds exhibit a greater toxicity than that estimated
for metolachlor, according to the results obtained with
standard ecotoxicity assays.[15] QSAR estimations show
that the location of added hydroxyl group(s) is of key
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286
Copyright © 2015 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

É. Nicol et al.
[17] H. H. Zheng, C. M. Ye. Identification of UV photoproducts
and hydrolysis products of butachlor by mass spectrometry.
Environ. Sci. Technol. 2001, 35, 2889.
[18] S. Coffinet, A. Rifai, C. Genty, Y. Souissi, S. Bourcier, M.
Sablier, S. Bouchonnet. Characterization of the photode-
gradation products of metolachlor: structural elucidation,
potential toxicity and persistence. J. Mass Spectrom. 2012,
47, 1582.
[19] Y. Souissi, S. Bouchonnet, S. Bourcier, K. O. Kusk, M. Sablier,
H. R. Andersen. Identification and ecotoxicity of
degradation products of chloroacetamide herbicides from
UV-treatment of water. Sci. Total Environ. 2013, 458, 527.
[22] Y. Souissi, S. Bourcier, S. Ait-Aissa, E. Maillot-Maréchal,
S. Bouchonnet, C. Genty, M. Sablier. Using mass
spectrometry to highlight structures of degradation
compounds obtained by photolysis of chloroacetamides:
Case of acetochlor. J. Chromatogr. A 2013, 1310, 98.
[23] S. Bourcier, C. Poisson, Y. Souissi, S. Kinani, S. Bouchonnet,
M. Sablier. Elucidation of the decomposition pathways of
protonated and deprotonated estrone ions: application to
the identification of photolysis products. Rapid Commun.
Mass Spectrom. 2010, 24, 2999.
[24] H. C. Hodge, J. H. Sterner. Tabulation of toxicity classers.
Am. Indust. Hygiene Assn Quart. 1949, 10, 93.
[20] L. Gutowski, O. Olsson, C. Leder, K. Kummerer.
A
comparative assessment of the transformation products of
S-metolachlor and its commercial product Mercantor Gold
(R) and their fate in the aquatic environment by employing
a combination of experimental and in silico methods. Sci.
Total Environ. 2015, 5056, 369.
SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of this article at the publisher’s website.
[21] http://www.epa.gov/nrmrl/std/qsar/qsar.html.
wileyonlinelibrary.com/journal/rcm
Copyright © 2015 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2015, 29, 2279–2286