Environ. Sci. Technol. 2009, 43, 3148–3154
Sunlight Nitrate-Induced Photodegradation of Chlorotoluron: Evidence of the Process in Aquatic Mesocosms ´ L I E U , * ,†,‡ SYLVIE NE FRANC ¸ O I S P E R R E A U , †,‡ ´ DE ´ RIQUE BONNEMOY,§ FRE MARTINE OLLITRAULT,| DIDIER AZAM,| LAURENT LAGADIC,⊥ J A C Q U E S B O H A T I E R , §,# A N D J A C Q U E S E I N H O R N †,∇ INRA, UR 258 Phytopharmacie et Me´diateurs Chimiques, 78000 Versailles, France, INRA, UR 251 Physico-chimie et Ecotoxicologie des Sols d’Agrosyste`mes Contamine´s, 78000 Versailles, France, CNRS, UMR 6023 Laboratoire Microorganismes: Ge´nome-Environnement, Universite´ B. Pascal, 63177 Aubie`re, France, INRA, UE 1036 Unite´ Expe´rimentale d’Ecologie et d’Ecotoxicologie Aquatique, 35000 Rennes, France, INRA, UMR 985 Ecologie et Sante´ des Ecosyste`mes, 35000 Rennes, France, Faculte´ de Pharmacie, Laboratoire de Biologie Cellulaire, 63000 Clermont-Ferrand, France, and INRA, UR 501 Biologie Cellulaire, 78000 Versailles, France
Received April 17, 2008. Revised manuscript received February 24, 2009. Accepted March 4, 2009.
The nitrate-induced photodegradation of chlorotoluron was demonstrated to occur efficiently in natural water through two series of experiments in outdoor aquatic mesocosms. During the first campaign, it was shown that the pesticide degradation kinetics was clearly dependent on nitrate concentration. This parameter also influenced the accumulation of the firstand second-generation byproducts, including predominantly N-terminus oxidation products and nitro-derivatives of the phenyl ring. The latter compounds, specific to the NO3- -induced photoprocess, appeared particularly abundant as compared to laboratory-simulated sunlight irradiation conditions. During the second campaign, a dual day-night sampling was achieved, which demonstrated the almost exclusive role of photodegradation versus biodegradation.
Introduction Phenylurea herbicides are widely used and often contaminate natural waters. Among them, chlorotoluron (CTU) is applied for pre- and postemergence control against many broadleaved and grass weeds, in particular on winter cereal crops. * Corresponding author phone: + 33-1-30 83 36 13; fax: + 33-1-30 83 31 19; e-mail:
[email protected]. † INRA, UR 258 Phytopharmacie et Me´diateurs Chimiques. ‡ INRA, UR 251 Physico-chimie et Ecotoxicologie des Sols d’Agrosyste`mes Contamine´s. § CNRS, UMR 6023 Laboratoire Microorganismes: Ge´nomeEnvironnement. | INRA, UE 1036 Unite´ Expe´rimentale d’Ecologie et d’Ecotoxicologie aquatique. ⊥ INRA, UMR 985 Ecologie et Sante´ des Ecosyste`mes. # Faculte´ de Pharmacie, Laboratoire de Biologie cellulaire. ∇ INRA, UR 501 Biologie Cellulaire. 3148
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 9, 2009
Its presence has been detected in surface and ground waters at the µg L-1 level (1). In soils and natural waters, biodegradation of CTU leads to the formation of 3-chloro-4methylaniline, N-dealkylation, and oxidation of the ringmethyl group (2-4). Some of the byproducts exhibit a higher toxicity than the pesticide (4). Chlorotoluron does not absorb solar light, but its phototransformation in natural waters can be induced by different absorbing species also present in the aquatic medium. For instance, humic substances (HS) can oxidize phenylureas via an initial electron-transfer step generated from their excited triplet state (5). Such mediated phototransformation was shown to be efficient in lake water supplemented with fulvic acids (6). However, depending on their nature and concentration, HS may also behave as a source or a sink for hydroxyl radicals, thus enhancing or limiting any induced photodegradation process (7). The ability of nitrate and nitrite ions to induce the phenylurea photodegradation has been demonstrated under simulated sunlight (8-11). Furthermore, diuron and chlorotoluron were found to be degradable under natural sunlight as well, in various NO3-/NO2- aqueous systems (12). Light excitation of NO3- and NO2- ions in water result in the formation of • OH radical and nitrogen reactive species (NO•, NO2•, N2O3, and N2O4) through reactions that have been extensively studied (13 and references therein). The primary phototransformations and subsequent reactions may be summarized as in eqs 1 to 5: NO-3+H2O + hν f NO•2 + •OH + OH-
(1)
3 NO-3 + hν f NO2 + O( P)
(2)
NO-2+H2O + hν f NO• + •OH + OH-
(3)
• • NO2 + OH f NO2+OH
(4)
2 NO•2 f N2O4
(5)
To assess the potential risk caused by a substance in an aquatic system, it is necessary to take into account its degradability and distribution characteristics under conditions as close as possible to nature. For this purpose, mesocosms offer realistic conditions to perform simultaneous studies on the fate of pollutants and their biological effects (14). Thus, mesocosm studies may include very complex interactive systems, in contrast with single-species toxicity tests conducted in laboratory (15). Up to now, investigations carried out in aquatic mesocosms on the fate of organic pollutants have led to observations dealing with their dissipation through direct photolysis (16), biodegradation (17), adsorption on sediments (18, 19), or accumulation in organisms (20). Transformation products resulting from biodegradation and/or abiotic reactions have been monitored in few cases (18, 21), but those studies did not attempt to correlate the presence of the byproducts with their effects on the ecosystems. The aim of this work was to investigate the effectiveness and importance of the nitrate-induced photodegradation process in aquatic mesocosms taking chlorotoluron as a model. Ecotoxicological effects issuing from the byproducts formed in situ were investigated in parallel (to be reported elsewhere). Concentration conditions chosen are realistic concerning the nitrate ion and only semirealistic for the CTU starting concentration. A relatively high value of the latter 10.1021/es8033439 CCC: $40.75
2009 American Chemical Society
Published on Web 03/30/2009
was estimated necessary to (i) achieve the monitoring of byproducts, even minor ones at very low levels in water samples (and also sediments), and (ii) be able to detect significant biological effects using our models. It has to be pointed out that the sunlight NO3- -induced degradation pathway of CTU does not seem to be different at lower concentrations of the herbicide (12).
Experimental Procedures Chemicals. Chlorotoluron (CTU, 3-(3-chloro-4-methylphenyl)-1,1-dimethylurea) > 99.5% purity (Cluzeau, Sainte-Foyla-Grande, France) was used in laboratory studies, whereas ca. 99% (Phyteurop, Montreuil-Bellay, France) was used for mesocosm contamination. 3-(5-chloro-4-methyl-2-nitrophenyl)-1,1-dimethylurea) was prepared by conventional nitration of CTU. NaNO3 (ACS reagent grade) was obtained from Aldrich (Germany) and organic solvents (methanol RPE and acetonitrile RS plus) were obtained from Carlo Erba (Valde-Reuil, France). LC-grade UP water was prepared by purification of reverse osmosis water in an Elgastat UHP system (Elga, High Wycombe, UK). The phenylurea stock solutions for laboratory study were prepared in UP water by ultrasound activated dissolution of the pesticide, followed by filtration on a 0.45 µm membrane (Millipore, Milford, MA). Standard solutions for HPLC were prepared in acetonitrile. For mesocosm contamination, 800-L solutions were prepared by careful dissolution of 14 g of CTU in tap water using a milk-tank equipped with mechanical stirring for 24 h. After filtration (press-filter), solutions were transferred to reservoirs covered with a black polypropylene sheet until being used and quantified by HPLC-UV within 24 h. Simulated Solar Light Irradiations. Polychromatic irradiations were performed in the laboratory using a cylindrical device equipped with six lamps emitting within 300-450 nm with a maximum emission at 365 nm (TLD 15 W, Philips). Light intensity measured by a radiometer (Vilber Lourmat, Marne-la-Valle´e, France) was 0.18 and 3.1 mW cm-2 at 312 and 365 nm, respectively (2.8 and 57.8 × 1014 photons cm-2 s-1). Aqueous solutions (300 mL) containing both chlorotoluron and nitrate were placed for the experiments in a 3 cm i.d. tubular pyrex reactor at the center of the lamp housing. The solutions were thoroughly mixed with gas sparging (air or nitrogen; 150 mL min-1; other details as in ref 9). Aquatic Mesocosm Experiments. Experimental Units. Mesocosm experiments were conducted at the Institut National de la Recherche Agronomique (INRA) aquatic experimental platform (Rennes, France, latitude 48°06′ N, altitude 75 m above sea level), using outdoor circular tanks of 3 m diameter and 0.7 m high (from a 24-mesocosm subunit distributed in 3 parallel lines), filled with sediment (depth 4-5 cm, 220 L/350 kg of 90/10 Loire River sand/sediment taken from a local pond), and 3500 L of water (depth 0.5 m, including 100 L of pond water in 2006 and 50/50 tap water/ pond water in 2007). The introduced plant and animal populations were: phytoplankton, zooplankton, macrophytes (Glyceria maxima), and a mollusc gastropod (Lymnaea stagnalis), the latter in 2006 only (35/mesocosm). The mesocosms were left for maturation during 2 months in 2006 before being contaminated, whereas in 2007 this phase was limited to 3 weeks (to limit heterogeneity between mesocosms). Contamination. For the 2006 campaign, eight formulations (three replicates each) were assigned randomly to the 24 mesocosms: CTU 1 or 0 mg L-1 and nitrate 5, 1, 0.2, or 0 mM. Four of those formulations were reproduced during the 2007 campaign (with three replicates, thus using 12 mesocosms only): CTU 1 or 0 mg L-1 and nitrate 1 or 0 mM. For both campaigns, the CTU aqueous solution at the desired concentration (see preparation above) was gently introduced under the water surface on T0 - 1 day and allowed to
homogenize during 24 h before the introduction of nitrate (T0). Both campaigns were performed at the theoretical period of maximum sunlight intensity (T0 at summer solstice), during 3 months and 1 month in 2006 and 2007, respectively. Weather conditions are presented in the Supporting Information (SI-1). The water level in mesocosms was adjusted if necessary every 2 weeks using tap water to compensate for evaporation. Sampling. For CTU and byproduct analysis, water samples were taken from four different places 5 cm below the surface and pooled in a 1 L bottle. Then a 200 or 400 mL aliquot was kept in an amber PET bottle and stored at -20 °C before analysis. This sampling was performed once a day (at determined dates) at 10 a.m. in 2006 and twice a day at 7 a.m. and 10 p.m. in 2007 (close to sunrise and sunset). In 2006, sediment samples were also collected using prepositioned 200 mL aluminum containers. Analysis of Chlorotoluron and Byproducts. HPLC-UV analyses were performed on a Dionex system (Voisins-le Bretonneux, France), including an ASI100T autosampler, a P580 pump, a STH585 column oven, and a UVD380S UVphotodiode array detector. Separation was conducted on a 250 × 4.6 mm 5 µm Nucleodur C8 end-capped reversedphase column (Macherey-Nagel, Du ¨ ren, Germany) at 20 °C, using a gradient of water and acetonitrile at 1 mL min-1 flow rate (80/20 to 40/60 v/v). Samples were kept at 15 °C and protected from light in the autosampler before being injected. Quantifications were carried out at 240 nm using the available external standards (CTU for all compounds except the nitroderivatives and CTU-NO2b for nitrated ones). LC-MS-MS analysis was carried out using an HPLC system (Alliance 2695, Waters) coupled to a Quattro LC triple quadrupole mass spectrometer (Micromass, Manchester, UK) with an electrospray interface. Data acquisition and processing were performed by MassLynx NT 4.0 system. Quantitation was done using two MRM transitions in the positive mode (SI-2). Sample Treatment. Water samples were analyzed directly after 0.45 µm membrane filtration. Sediment samples (5 g/unit) were extracted three times by water/acetonitrile 90/ 10 v/v. After centrifugation and 2-fold dilution with water, the obtained solutions were analyzed by polystyrenedivinylbenzene SPE-LC-MS-MS. Other Characterizations in Mesocosms. A number of parameters were determined continuously or regularly (pH, water temperature, conductivity, dissolved oxygen concentration, rain). Dissolved oxygen concentration and turbidity were measured at the same frequency using a luminescent dissolved oxygen probe HQ 10 and a turbidimeter 2100N, respectively (Hach, Noisy-le-Grand, France). Ammonium and orthophosphate ions were analyzed by photometric methods using Spectroquant kits (Merck, Darmstadt, Germany) and a Photolab S12 photometer (WTW, Champagne-au-Mont d’Or, France). The concentration of nitrate and nitrite ions was determined by HPLC using ion-pairing chromatography (22). Dissolved organic carbon (DOC) was measured using a Shimadzu TOC-VCSH/CSN (Duisburg, Germany) analyzer after 0.45 µm filtration of the water samples.
Results
NO3- -Induced Photodegradation of Chlorotoluron under Simulated Solar Light. NO3- -induced photodegradation of chlorotoluron was investigated first in laboratory conditions irradiating 1 mg L-1 UP water solutions with 300-450 nm lamps (see spectra in SI-3). Dissipation of CTU was found to be effective in the presence of nitrate, even at low concentrations (see evolution curves in SI-4a). The kinetics of disappearance of the phenylurea followed a pseudo firstorder, according to [CTU] ) [CTU]0 × e-k′t with [CTU]0 the initial concentration, k′ the pseudokinetic constant, and t VOL. 43, NO. 9, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3149
SCHEME 1. CTU and its NO3--Induced Photodegradation Pathways
the time. k′ values were determined for
