Science of the Total Environment 407 (2009) 4297–4302

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r.com/ locate /sc i totenv

Interactions of diuron with dissolved organic matter from organic amendments

Mathieu Thevenot a,⁎, Sylvie Dousset b, Norbert Hertkorn c, Philippe Schmitt-Kopplin c, Francis Andreux d

a UMR 7618 INRA-CNRS-Université, Laboratoire de Biogéochimie et Ecologie des Milieux Continentaux, Centre INRA Versailles Grignon, Thiverval-Grignon, Franceb Nancy-Université, département Sciences Terre, CNRS, LIMOS – BP 239 – 54506 Vandœuvre-les-Nancy, Francec Institute of Ecological Chemistry, GSF Research Center for Environment and Health, Ingolstaedter Landstrasse 1, D-85764 Neuherberg, Germanyd UMR 1229 CMSE INRA-Université, UFR Sciences Terre et Environnement, 6 bd Gabriel, 21000 Dijon, France

⁎ Corresponding author. Tel.: +33 1 30 81 53 62; fax:E-mail address: Mathieu.Thevenot@grignon.inra.fr (

0048-9697/$ – see front matter © 2009 Elsevier B.V. Adoi:10.1016/j.scitotenv.2009.04.021

a b s t r a c t

a r t i c l e i n f o

Article history:Received 21 January 2009Received in revised form 6 April 2009Accepted 19 April 2009Available online 2 May 2009

Keywords:PesticideDissolved organic matterFacilitated transportsDiuron

Diuron is frequently detected in some drinking water reservoirs under the Burgundy vineyards, where organicamendments are applied. The environmental effect of these amendments on pesticide transport is ambiguous:on the one hand it could enhance their retention by increasing soil organic carbon content; on the other hand,dissolved organic matter (DOM) could facilitate their transport. Elutions were performed using columnspacked with glass beads in order to investigate DOM–diuron interactions, and the possible co-transport ofdiuron and DOM. Four organic amendments (A, B, C and D)were tested; C and Dwere sampled at fresh (F) andmature (M) stages. An increase in diuron leaching was observed only for A and DF amendments (up to 16%compared to theDOM-free blank samples), suggesting a DOMeffect on diuron transport. These results could beexplained by the higher DOM leaching for A and DF compared to B, CF, CM and DM increasing diuron–DOMinteractions. These interactions seem to be related to the aromatic and aliphatic content of the DOM,determining formation of hydrogen and non-covalent bonds. The degree of organic matter maturity does notseem to have any effect with amendment C, while a reduction in diuron leaching is observed between DF andDM. After equilibrium dialysis measurement of diuron–DOM complexes, it appeared that less than 3% of thediuron applied corresponded to complexeswith amolecularweight N1000 Da. Complexes b1000Da could alsotake part in this facilitated transport.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

Diuron is a frequently used herbicide in vineyards and wasfrequently found in natural waters (DIREN, 2003; DIREN et al.,2005) at concentrations exceeding the European regulatory limit fordrinking water of 0.1 µg L−1 (ECC, 1998).

The addition of organic amendments to soil to increase theamount of soil organic matter can modify the environmentalbehaviour of pesticides. Amendments may increase the amount ofdissolved organic matter (DOM) in the aqueous phase. Indeed,several studies suggest that a high concentration of DOM couldfavour the formation of a third phase (micro-emulsion) in solutionbetween the hydrophobic compounds in the water phase, influen-cing their mobility (Marschner et al., 2005; Fauser and Thomsen,2002). Moreover, recent studies have suggested that DOM mayfacilitate pesticide transport through soil. Indeed, competitionbetween pesticides, such as diuron, and DOM molecules for sorptionsites and pesticide–DOM interactions can both account for enhanced

+33 1 81 54 97.M. Thevenot).

ll rights reserved.

pesticide leaching (Cox et al., 2007; Williams et al., 2000). On theother hand, other studies suggest that adding organic amendmentsto soil decreases pesticide leaching (Cox et al., 2001; Albarrán et al.,2004). These contradictory results could be due to the differences inthe nature of soils, DOM, and/or pesticide characteristics. Bybypassing the soil system, interactions between pesticide and DOMfrom exogenous sources, which need to be clarified, can be betterstudied. This can be achieved in simplified laboratory experimentsby using glass bead systems rather than soil (De Smedit andWierenga, 1984; Qureshi et al., 2003).

Our objectives were to evaluate diuron affinity with variousorganic amendments. We were particularly interested in potentialdiuron–DOM interactions at a molecular level and the subsequentpotential diuron co-transport. Four amendments were chosenfor their organic matter composition and DOM content. The effectof organic matter maturity on diuron leaching was also tested onsome of the amendments used, which were sampled at fresh andmature stages of composting. Leaching experiments were performedon columns filled with glass beads to assess interactions betweendiuron and DOM from organic amendments. Analysis by NMR wasperformed on each DOM leachate in order to determine its composi-tion and gain molecular level insight into diuron–DOM interactionmechanisms.

4298 M. Thevenot et al. / Science of the Total Environment 407 (2009) 4297–4302

2. Materials and methods

2.1. Chemicals

Diuron [3-(3,4-dichlorophenyl)-1,1-dimethylurea] was purchasedfromDr Ehrenstorfer GmbH (Augsburg, Germany) with N98% certifiedpurity. Diuron 14C-labelled on the carbonyl group (InternationalIsotope Society, Munich), with a specific activity of 567×103 Bqmmol−1was used in the adsorption experiment. The half-life of diuronis 3 to 6 months, its water solubility is 36.4 mg L−1 at 25 °C and itsadsorption coefficient (Koc) is 400 L kg−1 (Tomlin, 1997). Bromide,used as a non-reactive tracer, was obtained from water dissolution ofKBr powder purchased from Prolabo (Fontenay-sous-Bois, France)with N99% certified purity.

2.2. Organic amendments

Four organic amendments (A, B, C, D) were selected for theirvariability in organic carbon and DOC content. The main physico-chemical properties of the amendments are listed in Table 1.Amendments A and D were produced commercially and arecommonly used in agriculture. The raw materials were of plant origin(cocoa-, olive- and grape-cakes) and of plant-animal origin (sheepmanure, wool dust). Amendments B and C were made by wine-growers in and for their vineyards, from plant (bark, vine shoots andgrape residues) and animal wastes (manures). One amendment fromeach group (C and D) was sampled within the first 48 h of composting(CF and DF) and at the final stage, at least 6 months later (CM and DM).At the end of the composting maturation period, dried OMwas addedto DM before its commercial distribution. All six samples were dried at30 °C for 3 to 5 days. The resulting product was grounded to particlessmaller than 5mm. Each sample wasmanually mixed in a large tray tohomogenize it before re-sampling. Elemental analysis was performedby the SADEF laboratory (Aspach-le-bas, France), the OC content wasdetermined by dry combustion (NF ISO 10694). Extractable DOC foreach amendment was measured using aqueous extraction with anadapted method (Zsolnay, 1996). A 10 g subsample of each amend-ment was agitated with 400 mL of distilled water (1:40) for 12 h in a500 mL polycarbonate bottle, then centrifuged at 15000 g. Twosuccessive extractions were performed then pooled to evaluate thetotal DOC value of each sample.

2.3. Experimental set-up

The diuron leaching experiment used columns filled with 2-mm-diameter glass beads (Abscisse, France). Each columnwas a 6-cm-longPVC pipe with an internal diameter of 6.8 cm, packed with 220 g ofglass beads overlying a 0.1 mm nylon mesh filter. The diuronadsorption on the glass beads and mesh filter was measured using a

Table 1Physico-chemical properties of the organic amendments.

Composition (%) pH ECa

(mS m−1)OCb

(g kg−1DM)cDOCc (g kg−1 DM)d

(extract 1:40)C/N

A Vegetable oil cakes (100) 6.2 607 454 69.7 17B Manures+bark+grape

residue (70/15/15)8.8 260 301 7.2 15

CF Manures+vine shootsgrape residue (25/50/25)

6.9 105 466 17.1 44

CM Composition similar to CF 7.7 75 315 4.9 28DF Manure+vegetable oil

cakes (60/40)6.4 826 383 34.9 12

DM Composition similar to DF 5.7 650 438 32.8 18

a EC: electric conductivity.b OC: organic carbon content.c DOC: dissolved organic carbon content.d g kg−1 of dry matter.

batch equilibriummethod and was small compared with the total lossmeasured during the column experiments (b3% of the initialconcentration for the glass beads and b4.5% for the mesh filter). Theequivalent of 10 g of dried product (31 t ha−1), re-humidified bymixing with 10 mL of ultra-pure water in order to reduce OMhydrophobicity, was placed above the beads in order to obtainamended columns (CDOM). A column with DOM-free dried productwas made up in the same manner for each sample (CFDOM). DOM wasextracted using the method described previously. After DOM extrac-tion, the solid residue was dried at 35 °C for 24 h, re-humidified andplaced on glass beads. For each amendment, the leaching experimentwas replicated three times, on amended columns (CDOM) and onDOM-free columns (CFDOM) which correspond to the so-called “blanksamples”.

2.4. Leachate collection

Diuron leaching experiments were performed in the laboratory at20 °C. Twenty-four hours before starting the elution, 5 mg of diuronwas applied to each column by pipetting 2.5 mL of an ethanol solutionon to the surface, equivalent to 14.3 kg ha−1 active ingredient. Ethanolwas used prior to water in order to dissolve diuron and uniformlyapply it on the surface of the columns. Moreover, this high diuron ratewas applied in order to limit pesticide analysis uncertainties andsubsequent risk of mis-interpretation of the results. Five minutesbefore the elution experiment, 5 mg of bromide was applied bypipetting 2.5 mL of an aqueous solution onto the column surface. Afiberglass filter was placed on the surface of each column to ensurehomogeneous water distribution. A total of 2 L of distilled water wasapplied to each column over 10 h at a constant flow rate of 200mL h−1

(57 cm of water per column) using a peristaltic pump. The high watervolume was used in order to elute a maximum of DOM and diuron,and to establish valid recovery rate. Effluents were collected every15 min in 250 mL glass bottles using an automatic fraction collector(Foxy 200, Teledyne ISCO, USA).

2.5. Leachate analysis

All the samples were filtered at 0.4 µm with a polycarbonate filter(Sartorius, 25 mm) before analysis. Diuron and bromide measure-ments were performed using a Waters HPLC with a Diode ArrayDetector (Waters, Milford, MA, USA). Diuron was analyzed using a25 cm×4.6 mm C18-column packed with Kromasil 5 µm. The mobilephase was an acetonitrile–water mixture (70:30, v/v) and the flowrate was 0.8 mL min−1. Bromide detection was performed using a15 cm×4.6 mm IC Pack Anion HC. The mobile phase was acetonitrile-phosphate buffer (KH2PO4, pH 2.8) (20:80, v/v) and the flow rate was2 mL min−1. UV detection was performed at 251 nm for diuron and at200 nm for bromide. The detection limits were 0.1 mg L−1 for diuronand 0.25 mg L−1 for bromide.

DOC concentration in effluentswasmeasured by high-temperaturecombustion and catalytic oxidation on a platinised alumina catalyst,using a Shimadzu TOC-Analyser 5000A (Milton Keynes, UK). DOC wasmeasured by infrared detection of the amount of CO2 produced.Sampleswere acidified to pH 2with HCl 2 N andwere purged for 1minwith ultra-pure air at a 50 mL min−1 air-flow in order to removetrapped CO2. The detection limit for DOC was 2 mg L−1.

2.6. Nuclear magnetic resonance spectroscopy

For each amendment, 1H NMR and 13C NMR analyses wereperformed for each DOM. Ten grams of re-humidified compost wasleachedwith2 Lof distilledwater throughglass beads columns over 10hin order to obtain one sample for each amendment. Leachates werefiltered at 0.4 µmwith a polycarbonate filter prior to lyophilisation.

Table 3Distribution of carbon among the main structural groups of DOM of each compost asdetermined from their one-dimensional 13C nuclear magnetic resonance (NMR)spectra.

Range of δ(13C) (ppm) Assignment Proportion of carbon in structural groups (%)

A B CF CM DF DM

0–47 CHn 22.7 14.9 23.8 17.7 26.4 28.147–90 O(Nfunc)–CHn 51.8 20.9 42.1 28.1 37.8 25.590–108 O–C–O 1.8 5.8 4.6 6.2 5.0 6.5108–145 Car–C.H 6.4 21.2 3.5 15.3 8.9 10.1145–167 Car–C.N 1.6 12.0 0.6 5.8 3.7 3.2167–187 (CfO)–X– 15.4 19.2 24.4 21.0 16.5 13.9187–220 CfO 0.2 6.5 1.1 5.7 1.6 2.6

4299M. Thevenot et al. / Science of the Total Environment 407 (2009) 4297–4302

1HNMR and 13CNMR spectra (Fig. S1 in the Supporting Information)were acquired with a Bruker (Rheinstetten, Germany) DMX 500 NMRspectrometer at 303 K from samples dissolved in 0.1 N NaOD. Theinternal reference for 1H NMR was (H3C)3Si–CD2–CD2–COONa(−0.14 ppm), and for 13C an external reference (CH3OH in D2O:49 ppm) was used. 1H NMR spectra were acquired with a 5 mm z-gradient TXI cryogenic probe using the first increment of the presat-NOESY sequence (90° excitation pulses, solvent suppression withpresaturation and spin-lock, 5.2 s acquisition time, 14.7 s relaxationdelay,1Hzexponential linebroadening).13CNMRspectrawere acquiredwith a 5-mm broadband probe, using an inverse gated widebandalternating-phase low-power technique for zero residual splitting(WALTZ-16) decoupling (7.5 s relaxation delay) with an acquisitiontime of 500 ms and an exponential line broadening of 35 Hz.

Integration of the 1H NMR spectra of each DOM assigned accordingto specific regions of chemical shift (Hertkorn et al., 2002), providesthe non-exchangeable proton amounts shown in Table 2. Five regionswere determined: aliphatic protons [0–1.95 ppm], functionalisedaliphatic protons [1.95–3.1 ppm], carbohydrate [3.1–4.6 ppm], acetal[4.7–6.0 ppm] and aromatic protons [6.0–10.0 ppm]. The resonanceintegrals of seven common spectral regions (Hertkorn et al., 2002)were also performed on each 13C NMR spectra (Table 3): aliphaticcarbons [0–47 ppm], carbohydrate [47–90 ppm], doubly heteroatomsubstituted [90–108 ppm], aromatic carbons [108–145 ppm and 145–167 ppm], carbonyl-derivatives (carboxylic acids, ester and amides)[167–187 ppm] and carbonyl groups [187–220 ppm].

2.7. Diuron adsorption

Two diuron sorption experiments were performed. The firstadsorption experiment was performed in order to measure theamount of diuron adsorbed (QD, µg g−1) on amendments, inconditions close to those of the leaching experiment. Each sampleconsisted of 2 g of dried product placed in a 50 mL polycarbonatecentrifuge tube and treated with 0.25 mg of diuron in a 1 mL ethanolsolution. Twenty-four hours after the treatment, ethanol wasevaporated in the same manner as in the elution experiment, and10 mL of distilled water was added to each tube. The tubes wereagitated on a rotary shaker for 24 h at 20 °C in order to reachequilibrium, then centrifuged for 20min at 4000 g (Beckman-Avanti J-25 centrifuge maintained at 20±1 °C). Blanks prepared in the sameway, but without dried product, showed weak adsorption on thepolycarbonate tube. The amount of herbicide adsorbed by the sampleat equilibrium was determined by the difference between the initialand equilibrium herbicide concentrations in solution corrected usingthe blank adsorptionmeasurement. The experimentwas performed intriplicate.

The second adsorption experiment was performed in order toquantify the diuron adsorption coefficients in standardized condi-tions, to compare our results with those from the literature. Triplicatesamples of 2 g of dried amendment and 10 mL of 14C-diuron aqueoussolutions were shaken in 50-mL polycarbonate tube for 24 h at 20 °C.Diuron concentrations varied from 1 to 30 mg L−1. Following

Table 2Distribution of non-exchangeable hydrogen among different structural groups asdetermined from 1H nuclear magnetic resonance (NMR) spectra of DOM of eachcompost.

Range of δ(1H) (ppm) Assignment Proportion of hydrogen in structural groups (%)

A B CF CM DF DM

0.00–1.95 Cal–CHn 18.8 34.7 23.2 33.6 30.9 36.31.95–3.10 Cfunc–CHn 13.6 25.5 11.0 14.3 18.4 15.83.10–4.60 CHnO 56.2 19.5 53.3 38.5 44.4 40.74.70–6.00 Acetal 6.1 5.5 8.4 5.7 2.7 3.46.00–10.00 CarH 5.3 14.8 4.1 8.5 3.6 3.8

equilibration, the tubes were centrifuged for 20 min at 4000 g(Beckman-Avanti J-25 centrifuge maintained at 20±1 °C) and thediuron equilibrium concentration was determined by measuring theradioactivity in the supernatant by liquid scintillation counting (1900Tri-Carb, Packard, France); the amount of diuron adsorbed wasobtained by difference. Sorption isotherms were obtained by plottingx/m against Ceq and described using the Freundlich equation: x/m=KF×Ceqn , where KF (mg1−n Ln kg−1) and n (−) are empiricalsorption coefficients. The normalized adsorption coefficient (Koc) wascalculated as following Koc=KF/%OC×100. The amendment OCcontent was corrected for the amount of DOC in the supernatant.

2.8. Equilibrium dialysis experiment

In order to separate and quantify free diuron from diuron linked tothe DOMN1000 Da, an equilibrium dialysis was performed onleachates from CDOM columns amended with A and DF. In thosetreatments, the diuron concentration in the CDOM percolates wasgreater than in the CFDOM percolates. For each sample, dialysis wasduplicated on the first two percolate samples, where the highest DOMand diuron concentrations were found. The dialysis experiment wasperformed using a Dianorm equilibrium dialysis system (Dianorm,Munich – Germany). Cellulose ester dialysis membranes (Spectra/PorBiotech CE, Spectrum Laboratories), with amolecular weight cut-off of1000 Da, were used to separate a 10 mL Teflon® dialysis circular cellinto two halves. The water ionic force was maintained constant usingCaCl2 (from 2 to 4 mS m−1). For each percolate, 4 mL of eluted waterwas injected into one half-cell (hcell-A) and the CaCl2 solution intothe second (hcell-B). After 12 h of mechanical shaking (12 rpm) at23 °C, each half-cell was sampled and diuron and DOC concentrationswere determined by HPLC and TOC analysis, respectively.

2.9. Statistical analysis

The differences between CDOM and CFDOM in diuron and DOCpercolate recovery rates for a given sample, and between samples,were tested using a one-way analysis of variance (ANOVA; Fisher testat 95% confidence level). An ANOVAwas also performed on the diuronadsorption values obtained. These analyses were carried out withStatView 5.0.

3. Results and discussion

3.1. Organic amendment properties

The main physico-chemical characteristics of the organic amend-ments are given in Table 1. DOC concentrations vary greatly betweenamendments; the lowest are found for B, CF and CM and the highest forcommercial amendments A, DF and DM. This suggests that the DOCconcentration is linked to the nature of organic components and to thecomposting process, more than to the OC content of the amendments,

Table 4Diuron adsorption coefficients KF, n and normalized Koc valuesa,b.

QD (µg g−1) KF (mg1−n Ln kg−1) N (−) Koc

A 107.4 a [105.2–109.3] 42.4 ab [39.3–45.7] 0.76 a [0.70–0.82] 106.5 a [98.7–114.8]B 107.5 a [105.6–109.3] 47.3 b [43.8–51.0] 0.69 a [0.62–0.76] 160.3 b [148.5–172.9]CF 107.6 a [105.7–109.4] 40.5 a [38.2–43.0] 0.73 a [0.67–0.79] 89.1 c [84.1–94.6]CM 107.5 a [105.7–109.4] 38.9 a [35.3–42.8] 0.74 a [0.64–0.83] 125.5 a [113.9–138.1]DF 111.2 ab [109.3–113.0] 38.0 a [35.3–42.8] 0.79 a [0.73–0.84] 109.0 a [101.3–122.8]DM 112.9 b [111.0–114.8] 41.2 ab [38.3–44.4] 0.73 a [0.66–0.81] 101.7 ac [94.5–109.6]

a Values in parenthesis are 95% confidence intervals.b Values followed by the same letter for a given parameter do not differ at the 5% level.

4300 M. Thevenot et al. / Science of the Total Environment 407 (2009) 4297–4302

which were quite similar between all the samples. The lower DOCcontents of B, CF and CM could be related to their OM nature, richer inhydrophobic compounds such as lignin of vine shoots and barks,compared to the commercial amendments containing various plant oilcakes (Table 1).

With regard to maturity, for amendment C, the C/N ratio, OCcontent and DOC concentration decrease between the fresh stage (CF)and the mature stage (CM) (Table 1). For amendment D, the C/N ratioand OC content increase between the fresh stage (DF) and the maturestage (DM), but the DOC concentration remains comparable (Table 1).Organic matter maturation decreases OC content, C/N ratio and thequantity of DOC in the organic amendment (Goyal et al., 2005). A DOCconcentration lower than 4 g kg−1 was suggested as a maturityparameter (Zmora-Nahum et al., 2005). These observations areconfirmed in the case of compost C. In the case of amendment D,the addition of dried oil-cake at the end of the maturation periodcould explain the increase in C/N ratio and OC content, as well as thehigh DOC concentration.

3.2. 1H and 13C NMR spectroscopy

Judging by the 1H-NMR integration (Table 2 and Fig. S1 in thesupplementary data) for DOM from each compost, less than 20% of thenon-exchangeable protons are present in acetal and aromatic groups[CarH]. The most represented regions are carbohydrates [CHnO] (from19.5% to 56.2%) and non-substituted aliphatic carbon [Cal–CHn] (from18.8% to 36.3%). These results are confirmed by 13C-NMR integration.Indeed, the three main regions observed on 13C-NMR spectracorrespond to carbohydrates [O(Nfunc)–CHn] (from 20.9% to 51.8%),aliphatic carbon [CHn] (from 14.9% to 28.1%) and carboxyl group[CfO)–X–] (from 13.9% to 19.2%) (Table 3 and Fig. S1 in thesupplementary data). The aromatic regions represent from 4.1% to14.8% of the hydrogen [CarH] and from 4.1% to 33.2% of the carbon instructural groups [Car–CH, Car–CN] (Tables 2 and 3).

1H-NMR integration indicates a higher proportion of aliphaticcompounds in the DOM of DF, DM, CM and B (N30%) than in that of Aand CF (18.8% and 23.2%). Carbohydrate patterns are more pro-nounced in the DOM of A and CF than in that of CM, DF and DM, thelowest being for the DOM of B (Tables 2 and 3). The carbohydrateresonance range may include resonance contributions from severalfunctional groups, such as aminomethine groups [–CH(NH–)] and/ormethylene groups [–CH2] bonded to amide functional groups [–CH2

(NHCO–)] (Montoneri et al., 2003). The proportion of aromatic carbonis higher in the DOMof B and CM (33.2% and 21.1%) than in that of A, CF,DF and DM (8.0%, 4.1%, 12.6% and 13.3% respectively). This could beexplained by higher amounts of more humified compounds, that arericher in lignin, in the amendments B and CM than in the amendmentsA, CF and D (Table 1). Each sample exhibits carboxyl group resonance,with a higher proportion for the DOM of B, CF and CM (from 19.2% to24.4%) than in that of A, DF and DM (from 13.9% to 16.5%) (Table 3,Fig. S1 in the supplementary data). Concerning the maturity effect onDOM, pronounced alterations across the carbohydrate region wereobserved between the fresh stage of the DOM (CF and DF) and themature stage of the DOM (CM and DM), in accordance with organicmatter degradation. Concurrent with the carbohydrate decrease, anincrease in non-substituted aliphatic and aromatic groups is observed(Tables 2 and 3). Similar evolution in aliphatic and aromatic structureswas observed in the DOM from municipal waste composts (Chefetzet al., 1996; Hartlieb et al., 2001).

3.3. Diuron adsorption

The amount of diuron adsorbed (QD) varies from 107.4 µg g−1 [A]to 112.9 µg g−1 [DM] (Table 4), for pH of the supernatant ranging from5.7 to 7.2. At these pH values, diuron exists as single neutral speciessuggesting no significant pH effect on diuron adsorption during our

experiment from one amendment to another (Fontecha-Cámara et al.,2007). Studies on diuron adsorption on solid organic substrates arescarce in the literature. These values are higher than those measuredfor vermi-composted vineyard wastes (42 to 52 µg g−1) (Romeroet al., 2006), but similar to the value obtained for a peat (93 µg g−1)(Madhun et al., 1986a). The amounts of diuron adsorbed are similarbetween amendments (Table 4), except for DM, which presents aslightly higher QD value than for A, B, CF and CM. This result suggeststhat diuron sorption was not related to the OC content of theamendments (Table 1). The fact that B and CM have a QD similar to theother amendments, despite having a lower OC content (Table 1),suggests that diuron sorption is related to the nature of the organicmatter, increasing the affinity for these two organic substances. Itmight be due to a higher content of lignin-like compounds (from barkand vine shoot/grape residues) and a higher degree of maturity.Indeed, diuron sorption could be enhanced by these compounds,notably due to their surface area and functional groups (aromaticrings, hydroxyl constituents and carbonyl groups…; Tables 2 and 3)increasing non-ionic interactions, such as van der Walls and p−pbonds (Romero et al., 2006; Ahangar et al., 2008).

The diuron adsorption coefficients (KF) vary from 38 to 47 mg1−n

Ln kg−1 (Table 4). These KF values are lower than the value obtainedfor a peat (184 mg1−n Ln kg−1) and for similar vineyard residues asused for amendments B and C (77 to 158 mg1−n Ln kg−1) (Romeroet al., 2006; Madhun et al., 1986a). The diuron adsorption affinity issimilar for amendments A, CF, CM, DF and DM, but is slightly higher forB despite its lower OC content (Table 1). This result is in accordancewith the Koc value of amendment B, which is significantly higher thanthose of the other amendments. The Koc of CM was higher than that ofCF (Table 4), also suggesting an effect of the nature and the degree ofmaturity of the organic matter on diuron adsorption.

3.4. Bromide leaching

The shapes of the bromide elution curves are similar for both CDOMand CFDOM columns and for all sample types (Fig. S2 in thesupplementary data). For all organic amendments, the average recoveryrate of bromide in the leachates is not significantly different for CDOMcolumns and for CFDOM columns (Table 5). These results suggest goodreproducibility of elution between CDOM and CFDOM columns.

3.5. Diuron and DOM leaching

The cumulative amount of diuron recovered varies from 45.6% to72.6% of the herbicide applied (Table 5, Fig. S2 in the supplementarydata). The quantities of diuron leached through the CDOM columns aresignificantly higher for A and DF than for the other amendments(Table 5), despite their similar adsorption coefficient (Table 4). For Aand DF, a significantly higher proportion of diuron leached from CDOMcolumns (72.6% [A] and 64.9% [DF]) than from CFDOM (56.6% [A] and50.7% [DF]) (Table 4). In addition, the amounts of DOC leached from Aand DF were significantly higher than from B, CF, CM and DM (Table 5).Thus, potential diuron interactions with DOM from A and DF couldexplain the diuron leaching increase due to the formation of diuron–

Table 5Bromide, diuron and DOC recovery rates for each sample type on CDOM and Cblank columnsa,b.

Sample type Bromide (%) Diuron (%) DOC (g kg−1)

CDOM CFDOM CDOM CFDOM CDOM CFDOM

A 92.9±5.7 a 93.8±5.1 a 72.6±3.6 a 56.6±11.1 a⁎ 56.0±3.4 a 5.1±1.3 a⁎B 88.8±3.5 a 92.2±1.6 a 45.6±4.3 b 50.0±4.4 a 8.0±0.5 b 2.3±0.1 bc⁎CF 91.6±3.0 a 95.3±7.5 a 54.3±1.7 bc 49.9±2.3 a 11.5±0.9 c 2.2±0.5 bc⁎CM 91.0±0.9 a 90.1±6.8 a 55.7±6.9 cd 52.7±5.1 a 5.0±0.2 d 1.8±0.1 c⁎DF 97.9±4.3 a 92.6±3.0 a 64.9±2.8 ad 50.7±7.9 a⁎ 41.1±1.7 e 4.2±0.5 ab⁎DM 90.9±4.3 a 93.1±1.1 a 46.0±4.9 b 45.9±6.0 a 32.7±1.5 f 3.5±1.2 abc⁎Mean±SD 92.2±3.1 92.8±1.7 56.5±10.6 50.9±3.5 25.7±20.7 3.2±1.3

a Values followed by the same letter for a given column do not differ at the % level.b For a given amendment, values followed by an asterisk significantly differ at the 5% level between CDOM and CFDOM modality.

4301M. Thevenot et al. / Science of the Total Environment 407 (2009) 4297–4302

DOM complexes. These results are in agreement with other studies(Cox et al., 2007; González-Pradas et al., 1998; Madhun et al., 1986b).

The NMR data (Tables 2 and 3) show that carbohydrates andaliphatic carbon mainly influence the DOM composition of amend-ments A and DF. This result suggests that the increase of diuronleaching from A- and DF-amended columns may be explained by thelarger quantities of DOM leached (Table 5) and the formation ofhydrogen bonds between the carbonyl and/or amine functionalgroups of diuron with the acidic function of aliphatic compoundsand/or carbohydrates, as previously suggested (Gaillardon et al., 1980;Chefetz et al., 2004).

For B, CF, CM and DM, the quantities of diuron percolated do notsignificantly differ between the DOM-free and DOM treatments,despite the variations of DOC quantities percolated between CDOM andCFDOM (Table 5). Thus, for these amendments, diuron leaching doesnot seem to be related to DOM content but rather occurs in free form,although the diuron–DOM interaction hypothesis cannot be fullyrejected. Moreover, the similar quantities of diuron leached from B, CF

and CM, are in agreement with their comparable QD values (Table 4).The fact that the quantities of diuron leached from compost DM andcomposts B, CF and CM are similar might be explained by the higher QD

value (Table 4) and the higher amount of DOC leached (Table 5).The results of the leaching experiment suggest that the DOC

concentration could be a key parameter to explain and understand thediuron leaching process. Moreover, we can suppose that the amount ofDOC in solution is also a limiting factor for diuron–DOM interactions(i.e. potentially fewer reactive sites and consequently less diuronremobilization) and its influence on diuron leaching.

There is apparently no maturity effect on diuron transfer foramendment C, as diuron leaching is similar for CF and CM (Table 5).This is in accordance with their KF values, but in contrast to thedecrease in DOM leached from CM (Tables 5). For amendment D,diuron leached in greater amounts from the fresh stage (DF) than fromthe mature stage (DM), in spite of their similar KF values, but inaccordance with the higher DOC amounts leached from DF (Table 5).The difference between the two amendments, C and D, could beexplain by specific variation in their compositionwith time, such as anincrease of lignin abundance, which could influence the diuronaffinity for the solid phase as suggested by the higher Koc value for CM

compared to CF (Table 4).Concerning the dialysis measurement, the equilibrium diuron

concentrations are identical in hcell-A and hcell-B (Table S1, seesupplementary data). One of the drawbacks of themethod used is thatdetection of diuron–DOM complexes N1000 Da is only feasible if thedifference in diuron concentrations between the two half-cells islarger than 3% of the diuron applied, differences smaller than 3% ofbound diuron escape detection. The amount of complexes formedmayreach 4–10% (Diuron–DOM complexes N2500 Da) and less than 1.5%of the amount applied (Napropamide–DOMN500 Da) (Cox et al.,2007; Williams et al., 2000). Thus diuron–DOMN1000 Da complexesmay be present in quantities inferior to 3% of the initial amount of

diuron applied. At the end of the experiment, diuron recovery rateswere less than 100% (55% and 75%) as shown for other pesticides (Leeand Farmer, 1989). Diuron loss may be attributed to sorption on thedialysis membrane and perhaps on the cell surface. It also appears thata significant quantity of DOM diffused through the membrane,suggesting the existence of diuron–DOM complexes b 1000 Da, asshown for napropamide (Williams et al., 1999). Although the small-sized complex hypothesis is not rejected, it was not measurable withthe technique used in this experiment. In conclusion, the dialysismethod used was not hugely efficient, notably due to weak recoveryrates and to the cut-off chosen. However, the results indicate that atleast 3% of the diuron was leached as diuron–DOM complex.

4. Conclusion

Tested organic amendments show different diuron leachingabilities: it is higher for A and DF and lower or similar for the others.In solution, the NMR data suggest that a part of the diuron could becomplexed with DOM, mainly by means of hydrogen bonds. Thehigher diuron leaching for A and DF was better explained by the higherDOC content for these amendments compared with the others.Otherwise the affinity between diuron and the solid phase of thestudied organic amendments could be related to non-ionic bonds,such as “van der Waals” and/or π−π bonds, related to the aromaticityof the amendments. The degree of maturity of the organic matter hada variable effect on diuron leaching: no effect was observed in the caseof compost C, yet an attenuation in diuron leaching was observed forcompost D, in accordance with a decrease of the amount of DOMleached. In conclusion, our results suggest that organic amendments,and their hydrosoluble fraction, could have a great impact on diurondissipation by influencing its adsorption and transport processesthrough various chemical interactions. However, the exact chemicalmechanisms need to be clarified in the future. If the DOM effect ondiuron leaching is confirmed in soils, agricultural practices such asmulching and grass cover, could, in some cases, facilitate pesticideleaching, and increase the risk of groundwater contamination.

Acknowledgements

The organic amendments were provided by Dr. L. Thuriès, C. Rionand J. Lacroute. Thanks to E. Barriuso, J.-C. Fournier, S. Gerber, S. Houotand V. Dumeny for their contribution, notably for the dialysis and 14C-adsorption experiments. Reviewcomments of this paper by C. Chateauand N. Nunan were also greatly appreciated. This research waspartially funded by a PhD grant from the Ministère de l'EducationNationale, de la Recherche et de la Technologie.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.scitotenv.2009.04.021.

4302 M. Thevenot et al. / Science of the Total Environment 407 (2009) 4297–4302

References

Ahangar AG, Smernik RJ, Kookana RS, Chittleborough DJ. Clear effect of soil organicmatter chemistry, as determined by NMR spectroscopy, on the sorption of diuron.Chemosphere 2008;70:1153–60.

Albarrán A, Celis R, Hermosín MC, López-Piñeiro A, Cornejo J. Behaviour of simazine insoil amended with the final residue of the olive-oil extraction process. Chemo-sphere 2004;54:717–24.

Chefetz B, Bilkis YI, Polubesova T. Sorption–desorption behavior of triazine andphenylurea herbicides in Kishon river sediments. Water Res 2004;38:4383–94.

Chefetz B, Hatcher PG, Hadar Y, Chen Y. Chemical and biological characterization oforganic matter during composting of municipal solid waste. J Environ Qual1996;25:776–85.

Cox L, Cecchi R, Celis R, HermosínMC, KoskinenWC, Cornejo J. Effect of exogenous carbonon movement of simazine and 2,4-D in soils. Soil Sci Soc Am J 2001;65:1688–95.

Cox L, Velarde P, Cabrera A, Hermosín MC, Cornejo J. Dissolved organic carboninteractions with sorption and leaching of diuron in organic-amended soils. Eur JSoil Sci 2007;58:714–21.

De Smedit F, Wierenga PJ. Solute transfer through columns of glass beads. Water ResourRes 1984;20:225–32.

DIREN Rhône-Alpes. État des lieux des pesticides dans les eaux de la région Rhône-Alpes, Campagne de Septembre 2001 à Décembre 2002, 2003.

DIREN, DRAF, DRASS, FREDON, Bourgogne. Réseau de suivi des pesticides dans les eauxen région Bourgogne. Rapport de présentation des résultats de l'année hydro-logique d'août 2003 à juillet 2004, 2005.

European Community Council. Directive on the quality of water for human consump-tion. Off J Eur Communities 1998;L330:32–60 (98/83/EEC).

Fauser P, Thomsen M. Sensitivity analysis of calculated exposure concentrations anddissipation of DEHP in a topsoil compartment: the influence of the third phaseeffect and Dissolved Organic Matter (DOM). Sci Total Environ 2002;296:89-103.

Fontecha-Cámara MA, López-RamónMV, Alvarez-MerinoMA,Moreno-Castilla C. Effect ofsurface chemistry, solution pH, and ionic strength on the removal of herbicides diuronand amitrole fromwater by an activated carbon fiber. Langmuir 2007;23:1242–7.

Gaillardon P, Calvet R, Gaudry JC. Adsorption de quelques phénylurées herbicides pardes acides humiques. Weed Res 1980;20:201–4.

González-Pradas E, Villafrancha-Sánchez M, Fernández-Pérez M, Sociás-Viciana M,Ureña-Amate MD. Sorption and leaching of diuron on natural and peat-amendedcalcareous soil from Spain. Water Res 1998;32:2814–20.

Goyal S, Dhull SK, Kapoor KK. Chemical and biological changes during composting ofdifferent organic wastes and assessment of compost maturity. Bioresour Technol2005;96:1584–91.

Hartlieb N, Marschner B, Klein W. Transformation of dissolved organic matter (DOM)and 14C-labelled organic contaminants during composting of municipal biowaste.Sci Total Environ 2001;20:1-10.

Hertkorn N, Permin A, Perminova I, Kovalevskii D, Yudov M, Petrosyan V, et al.Comparative analysis of partial structures of a peat humic and fulvic acid using one-and two-dimensional nuclear magnetic resonance spectroscopy. J Environ Qual2002;31:375–82.

Lee DY, Farmer WJ. Dissolved organic matter interaction with napropamide and fourother nonionic pesticides. J Environ Qual 1989;18:468–74.

Madhun YA, Freed VH, Young JL, Fang SC. Sorption of bromacil, chlortoluron, and diuronby soils. Soil Sci Soc Am J 1986a;50:1467–71.

Madhun YA, Young JL, Freed VH. Binding of herbicides by water-soluble organicmaterials from soil. J Environ Qual 1986b;15:64–8.

Marschner B, Winkler R, Jodemann D. Factors controlling the partitioning of pyrene todissolvedorganicmatter extracted fromdifferent soils. Eur J Soil Sci 2005;56:299–306.

Montoneri E, Savarino P, Adani F, Genevini PL, Ricca G, Zanetti F, et al. Polyalkylphenyl-sulphonic acid with acid groups of variable strength from compost. Waste Manage2003;23:523–35.

Qureshi S, RichardsBK,HayAG,TsaiCC,McBrideMB,BaveyeP, et al. Effectofmicrobial activityon trace element release from sewage sludge. Environ Sci Technol 2003;37:3361–6.

Romero E, Salido A, Cifuentes C, Fernandez JD, Nogales R. Effect of vermicompostingprocess on pesticide sorption capability using agro-industrial wastes. Int J EnvironAnal Chem 2006;86:289–97.

Tomlin CDS. The pesticide manual, 11th ed. British Crop Protection Council and theRoyal Society of Chemistry, UK, 1997, 1606 pp.

Williams CF, Farmer WJ, Letey J, Nelson SD. Molecular weight of dissolved organicmatter-napropamide complex as affected by napropamide-soil application meth-ods. J Environ Qual 1999;28:1429–35.

Williams CF, Agassi M, Letey J, FarmerWJ, Nelson SD, Ben-HurM. Facilitated transport ofnapropamide by dissolved organic matter through soil columns. Soil Sci Soc Am J2000;64:590–4.

Zmora-Nahum S, Markovitch O, Tarchitzky J, Chen Y. Dissolved organic carbon (DOC) asa parameter of compost maturity. Soil Biol Biochem 2005;37:2109–16.

Zsolnay A. Dissolved humus in soil waters. In: Piccolo A, editor. Humic Substances inTerrestrial Ecosystems. Amsterdam: Elsevier; 1996. 688 pp.