Вы находитесь на странице: 1из 8

Journal of Hazardous Materials 172 (2009) 13031310

Contents lists available at ScienceDirect

Journal of Hazardous Materials


journal homepage: www.elsevier.com/locate/jhazmat

Photocatalytic degradation of eight pesticides in leaching water by use of ZnO under natural sunlight
S. Navarro a, , J. Fenoll b , N. Vela c , E. Ruiz b , G. Navarro a
Departamento de Qumica Agrcola, Geologa y Edafologa. Facultad de Qumica. Universidad de Murcia. Campus Universitario de Espinardo. 30100 Murcia, Spain Departamento de Calidad y Garanta Alimentaria. Instituto Murciano de Investigacin y Desarrollo Agrario y Alimentario (IMIDA). C/ Mayor s/n. La Alberca, 30150 Murcia, Spain c Facultad de Ciencias de la Salud. Universidad Catlica San Antonio de Murcia. Campus de Los Jernimos, s/n. Guadalupe, 30107 Murcia, Spain
b a

a r t i c l e

i n f o

a b s t r a c t
Photodegradation of eight pesticides in leaching water at pilot plant scale using the tandem ZnO/Na2 S2 O8 as photosensitizer/oxidant and compound parabolic collectors under natural sunlight is reported. The pesticides, habitually used on pepper culture and belonging to different chemical groups were azoxyxtrobin, kresoxim-methyl, hexaconazole, tebuconazole, triadimenol, and pyrimethanil (fungicides), primicarb (insecticide), and propyzamide (herbicide). As expected, the inuence of the semiconductor used at 150 mg L1 on the degradation of pesticides was very signicant in all cases. Photocatalytic experiments show that the addition of photosensitizer strongly improves the elimination of pesticides in comparison with photolytic tests; signicantly increasing the reaction rates. The use of Na2 S2 O8 implies a signicant reduction in treatment time showing a quicker reaction time than ZnO alone. On the contrary, the addition of H2 O2 into illuminated ZnO suspensions does not improve the rate of photooxidation. The disappearance of the pesticides followed rst-order kinetics according to LangmuirHinshelwood model and complete degradation occurs from 60 to 120 min. The disappearance time (DT75 ), referred to the normalized illumination time (t30W ) was lower than 3 min in all cases. 2009 Elsevier B.V. All rights reserved.

Article history: Received 29 April 2009 Received in revised form 30 July 2009 Accepted 30 July 2009 Available online 8 August 2009 Keywords: Agrochemicals Groundwater pollution Photooxidation Semiconductor Sunlight irradiation

1. Introduction The environmental fate of pesticides depends a lot on their mobility in soils and their tendency to partition into other environmental compartments, such as air and water. Transfer processes move pesticides in the environment. Concretely, leaching (the movement of water and dissolved chemicals through the soil) of pesticides through the soil prole from agricultural practices is receiving increasing attention [1]. In the leaching process, the physicochemical properties of the agrochemicals used as well as soil properties (texture, clay content, organic matter content and permeability) play a decisive role [2]. The displacement of pesticides from soil to water mostly depends on the extent to which they are retained in soils [3]. It is generally accepted that adsorption of pesticides by soils is more closely related to the soil organic matter content than any other single property [4]. Many papers published in the last four decades have demonstrated a high correlation between the organic matter content and the distribution coefcient Kd in a variety of soils. The most common and today generally accepted quantitative measure of the

Corresponding author. Tel.: +34 968 367477; fax: +34 968 364148. E-mail address: snavarro@um.es (S. Navarro). 0304-3894/$ see front matter 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jhazmat.2009.07.137

sorption of organic pollutants by soils from aqueous solutions is the soil organic adsorption coefcient (KOC ). This chemical specic parameter plays a signicant role in the fate of pesticides in aqueous/soil environment, like bioaccumulation and leaching ability, which is dened as: KOC = Kd /FOC , where FOC is the organic carbon fraction of the soil and Kd is the distribution coefcient. KOC values are universally used as measures of the relative potential mobility of pesticides in soils and in fugacity models describing the partitioning of pesticides in soil/water/atmosphere systems [5]. In general, compounds with higher log KOC values will be less mobile than those with lower values. Especially for the hydrophobic pesticides (KOW 2), their mobility, and therefore the risk of their leaching into groundwater, has been correlated with weak sorption on the soil, as quantied by KOC [6]. Generally, pesticides with log KOC 3 are potentially leacher compounds although pesticides with log KOC 3 have been nd in groundwater and drainage water [7]. The worlds tremendous output of pollutants challenges the capacity of waterways to assimilate or ush away pollution. Concretely, Europe confronts enormous groundwater pollution problems being Agriculture the biggest polluter, even more so than industries and municipalities. European Union (EU) recent estimates reveal that about 1.5 million industrial and waste disposal sites exist which could have the potential to negatively inuence

1304

S. Navarro et al. / Journal of Hazardous Materials 172 (2009) 13031310

soil and groundwater, the later being the major source for drinking water [8]. The consumption of pesticides by weight is decreasing, but it is the toxicity of an individual pesticide, not necessarily the amount used, which determines its potential for environmental damage. A further major threat to groundwater is the inappropriate storage of pesticides. Although a lot of effort is being made by EU countries in investigating the situation of pesticide pollution, the awareness of pesticides causing problems in groundwater is still increasing as indicated in their State of Environment reports [9]. Accession countries face similar or even more serious groundwater impacts and are supposed to follow EU environmental quality standards in the near future [10]. The protection of groundwater resources is a priority of EU environmental policy. Thus, on October 2000, the EU Water Framework Directive (WFD, 2000/60/EC) was adopted and 3 years later, on September 2003, the European Commission adopted a proposal for a new Directive to protect groundwater from pollution and deterioration (2006/118/EC) in whose Annex I, the quality standards for pesticides and metabolites are listed. Under certain conditions, some pesticides may leach to groundwater from normal eld applications [6]. Although there is large literature concerning groundwater pollution, predicting pesticide behaviour in subsurface geosystems remains a complex scientic and practical problem. Pesticide residues have been retrieved in groundwater bodies all over the world and their levels frequently exceed the drinking water limit set by the EU. The Drinking Water Directive (98/83/EC) requires pesticide concentration in drinking water not to exceed 0.1 g L1 for a single pesticide and 0.5 g L1 for total pesticides. Groundwater pollution not only affects the health of humans being as it is being used for drinking purpose, but also can act as a source of contamination for food chain, when used for irrigation. Unfortunately, a great number of pesticides are bio-recalcitrant, non-biodegradable. Therefore, biological process is not the ideal process and other more effective technologies such as advanced oxidation technologies (AOTs) have been proposed for treatment of polluted water by pesticides [1113]. These technologies allow pesticides to be removed by mineralization. Photolysis is one of the major transformation processes affecting the fate of pesticides in the aquatic environment. If the exciting energy employed is from the sun, the process is called solar photocatalysis [14]. Sunlight photoalteration processes are well now to play and important role in the degradation of pesticides and other contaminants in water by generation of highly reactive intermediates, mainly hydroxyl radical ( OH), a powerful non-specic oxidant (E = 2.8 V). Photocatalytic oxidation by semiconductor oxides is an area of environmental interest for the treatment of polluted water, particularly relevant for Mediterranean agricultural areas, where solar irradiation is highly available making this process quite attractive [1519]. Due to its stability and non-toxicity, titanium dioxide (TiO2 ) has been demonstrated to be an excellent catalyst and its behaviour is very well documented in the literature [20], although the photocatalytic effect of other semiconductors like zinc oxide (ZnO) is not so well known. ZnO is a very interesting wide band gap semiconductor material, because of its direct band gap, large excitation binding energy, and piezoelectric properties. The band gap of this semiconductor material is ca. 3.2 eV which corresponds to radiation of wavelength around 390 nm. Therefore, an UV light with wavelength shorter than 380 nm is needed to excite the electrons in valence band to conduction band. The electron/hole (e /h+ ) pairs thus generated serve as the oxidizing and reducing agents. Intrinsic ZnO is essentially pure semiconductor material while extrinsic ZnO can be formed from intrinsic ZnO by added impurity atoms to the crystal in a process known as doping. One of the main remaining obstacles, limiting the development of ZnO, is the difculty in achieving p-type doping of this material while

n-type is easily realized via excess Zn or with Al, Ga or In doping [21]. Although new management practices as integrated crop management (ICM) have evolved in the last years as response to the need to reduce dependence of pesticides, the use of agrochemicals is still very important in many areas. Concretely, during pepper cultivation many pesticides can be deposited on the soil surface by spray drift after pesticide application, wash-off from plants by rain or direct treatment to the soil. Some pesticides commonly used during the crop development can be leached through the soil prole as we have observed under laboratory and eld conditions (unpublished data). For this reason in this work we have studied the photosensitization effect of ZnO on the degradation of eight pesticides, commonly used for pepper protection, in leaching water under natural sunlight. 2. Experimental 2.1. Pesticides and reagents Pesticide analytical standards were obtained from Dr. Ehrenstorfer GmbH (Augsburg, Germany). They were higher than 98% pure. The commercial formulations used were purchased from Fitodolores SL (Murcia, Spain). The structures of the active ingredients are shown in Fig. 1 and their main physicalchemical properties in Table 1. Experimental values of octanol/water partition coefcient (KOW ), soil/organic partition coefcient (KOC ), aqueous solubility (SW ), aqueous hydrolysis, and GUS (Groundwater Ubiquity Score) index were taken from The Pesticide Properties DataBase (PPDB) [22]. Zinc oxide (99.9%, metals basis; <200 mesh particle size) and sodium peroxydisulfate (98%) was purchased from Alfa Aesar (Karlsruhe, Germany) and Panreac Qumica (Barcelone, Spain), respectively. Pesticide grade dichloromethane, ethyl acetate, cyclohexane and n-hexane were supplied by Scharlau (Barcelone, Spain). Stocks solutions (1000 g mL1 ) of each pesticide standard were prepared in ethyl acetate/cyclohexane (1/1, v/v), protected from light and stored at 5 C. A pesticide intermediate standard solution was prepared by dilution in the same solvent to obtain a concentration of 10 g mL1 . Several standard solutions, with concentrations of 0.052 g mL1 , were injected to obtain the linearity of detector response and the detection limits of the pesticides studied. The correlation coefcient was found to be >0.99 and the limits of quantication (LOQ, obtained at signal-to-noise ratio 10) ranged from 0.01 to 0.3 g L1 . 2.2. Leaching water The water used in the photodegradation studies was obtained from 8 lysimeters (3.5 m 4 m 1 m) from an experimental greenhouse located in Campo de Cartagena (SE Spain). A clayloam soil (pH = 8.7 and OM = 0.22%) was used. The soil was irrigated every 4 days by three dripperlines (45 min per day and 50 mL min1 per emitter). About 8 L per day were collected from each lysimeter. The water had pH = 8.41, EC = 4.32 dS m1 , TOC = 130 mg L1 , NO3 = 547 mg L1 , and NO2 = 0.12 mg L1 . The leaching water (1000 L) was collected and transported to the storage tank. Finally, the water was spiked with commercial products and stored in dark at ambient temperature for 1 week. 2.3. Solar photocatalysis experiment The experiment was carried out in a pilot plant placed in Murcia, SE Spain (latitude 37 59 N, longitude 1 08 W) using natural sunlight irradiation during July, 2008. The values (mean SD) of photosintetically active (400700 nm), UVA (315400 nm) and UVB

S. Navarro et al. / Journal of Hazardous Materials 172 (2009) 13031310

1305

Fig. 1. Chemical structures of the active ingredients used (CAS RN are in brackets): (1) azoxystrobin [131860338], (2) hexaconazole [79983714], (3) kresoximmethyl [143390890], (4) pirimicarb [23103982], (5) propyzamide [23950585], (6) pyrimethanil [53112280], (7) tebuconazole [107534963], and (8) triadimenol [55219653].

(280315 nm) radiation were recovered with a portable photoradiometer Delta Ohm HD 2102.2 (Caseelle di Selvazzano, Italy). The mean values of PA, UVA and UVB at 14 h were 901.2 52.3, 26.1 2.4, and 1.83 0.16 (all in W m2 ), respectively while 100,750 5794 lx were recorded. Fig. 2 shows the mean values (n = 6) recorded from UV radiation during the sampling. To facilitate the comparison with other photocatalytic experiments, the normalized illumination time (t30W ) was used instead of exposition time (t). The t30W has been calculated according to the following equation: t30W,n = t30W,n1 + tn = tn tn1
Table 1 Physicalchemical characteristics of the pesticides used in this study. Commercial names Ortiva 25% (w/v), SC (Syngenta) Daltonex 3% WG (Nufarm) Stroby 50% WG (BASF) Aphox 50% WG (Syngenta) Kerb Flo 40% (w/v) SC (Dow) Scala 40% (w/v) SC (BASF) Folicur ME 10% WG (Bayer) Baydan 25% (w/v) EC (Bayer)
a b c d

where tn is the exposition time for each sample, UV the average solar ultraviolet radiation measured during t, Vi the total irradiated volume, and VT the total volume of the reactor (illuminated and unilluminated). In this expression, used by several authors working on this topic [17,18], time refers to a constant solar UV power of 30 W m2 , typical solar UV power on a perfectly sunny day around noon.

tn (UV/30)

Vi VT

(a) Solar photocatalytic plant. The solar pilot plant used in this experiment is based on compound parabolic collector (CPC) technology [23]. This small prototype consists of one photoreactor module (1.27 m2 ) with ve borosilicate tubes (200 cm length 4 cm i.d.) mounted on a curved polished aluminium reectors (0.9 cm radius of curvature) running in the East-West

Active ingredienta AzoxystrobinF Hexaconazole()F Kresoxim-methylF PirimicarbI PropyzamideH PyrimethanilF Tebuconazole()F Triadimenol(A + B)F

Molecular formula C22 H17 N3 O5 C14 H17 Cl2 N3 O C18 H19 NO4 C11 H18 N4 O2 C12 H11 Cl2 NO C12 H13 N3 C16 H22 ClN3 O C14 H18 ClN3 O2

Molecular weight 403.4 314.2 313.4 238.3 256.1 199.3 307.8 295.8

Log KOW 2.5 3.9 3.4 1.7 3.1 2.8 3.7 3.2

SW b 7 18 2 3100 9 121 36 72

Aqueous hydrolysisc Stable Stable 34 Stable Stable Stable Stable Stable

Soil sorption log KOC 2.6 3.0 2.5 1.9 3.3 2.5 2.9 2.4

GUS indexd 2.5 2.0 1.8 4.0 1.8 2.6 2.0 3.7

F: fungicide; H: herbicide; I: insecticide. Water solubility (mg L1 ). DT50 (days) at 20 C and pH = 7. Groundwater Ubiquity Score (GUS) Index.

1306

S. Navarro et al. / Journal of Hazardous Materials 172 (2009) 13031310

dark with collectors covered by a black awning. Finally, the photosensitizer (ZnO) and oxidant (Na2 S2 O8 ) were added at 150 and 100 mg L1 and the cover removed after 15 min. Several samples (0, 15, 30, 60, 120, 240, and 480 min) were taken during the photoperiod (8 h), from 10 to 18 h. Periodically, air was injected in the tank. A parallel blank assay, without ZnO and Na2 S2 O8 (photolysis experiment), was carried out. In both cases, assays were replicated 3 times. (c) Photocatalytic kinetics. As reviewed by Konstantinou and Albanis [20], the photocatalytic oxidation of several organic pollutants-including pesticides over illuminated catalyst tted the LangmuirHinshelwood (LH) kinetics given by the following equation: r=
Fig. 2. Mean values (n = 6) for UVA and UVB radiation in the different sample points of the photoperiod (480 min).

kKC dC = + KC 2 dt

line. The water ows directly from one to another tube connected in series and nally to the reservoir tank (250 L) and a centrifugal pump (0.55 kw) then returns the water (45 L min1 ) to the collector tubes in a closed circuit. The reaction system was continuously stirred to achieve a homogeneous suspension and thermostated by circulating water to keep the temperature at 25 2 C. The illuminated volume was 12.55 L and the dead volume of PVC tubes about 6.5 L. Storage tank, owmeter, sensors (pH, O2 and T), pipes, and ttings complete the installation. A scheme of the pilot plant used is shown in Fig. 3. (b) Photocatalysis design. At the beginning of the assay, the leaching water (150 L) was mixed with commercial products (Table 1) to reach a spiking level of about 0.5 mg L1 of each one homogenizing the mixture for 20 min to constant concentration in the

where r is the mineralization rate of pesticide, C is the concentration of the pesticide, k is the rate constant, and K the adsorption coefcient of the pesticide. When the initial concentration C0 is small (ppbppm range), many researchers has approximated the LH kinetics to rst-order expression just to easily obtain the parameter involved in the LH equation according to the following equation: r= dC = k C = kKC dt

Its integration results in the following equations: Ct = C0 ek t or ln C0 C = kKt = k t

where k (in units of time1 ) is the apparent rst-order rate constant.

Fig. 3. Scheme of the pilot plant used for the photocatalytic experiments.

S. Navarro et al. / Journal of Hazardous Materials 172 (2009) 13031310

1307

The half-life time given by the equation t1/2 = (1/k)ln 2 is the amount of time required for 50% of the initial pesticide concentration to dissipate. Unlike the half-life, the dissipation time does not assume a specic degradation model (e.g., rst-order degradation). 2.4. Analytical determinations (a) GC/MS analysis. An Agilent (Waldbronn, Germany) HP 6890 gas chromatograph equipped with a 5973N mass spectrometer and automatic splitsplitless injector Agilent 7683 was operated in electron impact ionization mode with an ionizing energy of 70 eV, scanning from m/z 50 to 500 at 3.21 s per scan. The ion source temperature was 230 C, and the quadrupole temperature was 150 C. The electron multiplier voltage (EM voltage) was maintained at 1300 V, and a solvent delay of 4.5 min was employed. An HP-5MSI fused silica capillary column (30 m 0.25 mm i.d.) and 0.25 m lm thickness, supplied by Agilent Technologies, was employed. The column temperature was maintained at 70 C for 2 min and then programmed at 25 C/min to 150 C, increased to 200 C at a rate of 3 C/min, followed by a nal ramp to 280 C at a rate of 8 C/min, and held for 10 min. One microliter of samples was injected in splitless mode. Analysis was performed with selected ion monitoring (SIM) mode using primary and secondary ions. The target and qualier abundances were determined by injection of individual pesticide standards under the same chromatographic conditions using full scan with the mass/charge ratio ranging from m/z 50 to 500. Pesticides were conrmed by their retention times, the identication of target and qualier ions, and the determination of qualier-to-target ratios. Retention times had to be within 0.1 min of the expected time, and qualier-to-target ratios had to be within a 10% range for positive conrmation. The concentration of each compound was determined by comparing the peak areas in the sample to those found for mixtures of pesticide standards of known concentration. (b) Extraction procedure. To remove ZnO particles the water samples were ltered through a funnel of porous plate No. 4. A liquidliquid (LL) microextraction method was used for the isolation of pesticides. Water samples (20 mL) were extracted by sonication with 40 mL of n-hexanedichloromethane 1:1 mixture solvent. Finally, pesticide residues were determined by GC/MS in SIM mode as previously specied. Under these conditions, the recoveries obtained for the pesticide ranged from 76.5 to 106% with a relative standard deviation lower than 7% in the most unfavourable case. (c) Total organic carbon (TOC). An Analyzer Shimadzu TOC Vcsh (Kyoto, Japan) provided with an NDIR detector (680 C combustion catalytic oxidation technique) was used (LOD = 4 g L1 and accuracy <2%). 3. Results and discussion 3.1. Preliminary studies Laboratory experiments (n = 5) were conducted in order to determine the leaching potential of the tested compounds by use of disturbed clayloam soil columns (unpublished data). Soil columns offer good possibilities to conduct such tests, because they constitute closed systems, with the control of water leaching through the soil. All experiments were performed according to the OECD guidelines [24]. In all cases, 100 g of each pesticide were added to 150 g of soil (0.66 g g1 ). The compounds were leached with 0.01 M CaCl2 in order to minimize soil mineral balance disruption which was applied at a rate of 60 mL daily during 10 days

Fig. 4. Cumulative breakthrough curves (BTCs) of the pesticides applied to soil columns. Error bars have been avoided for better visualization.

with a peristaltic pump, corresponding to 210 mm precipitation. All pesticides were found in leachates after in different proportions as can be seen in Fig. 4 where the cumulative amount leached is shown. The breakthrough curves (BTCs) show that pirimicarb and triadimenol behave as leacher compounds since 48% and 41% of the initial amount, respectively was found in leachates, while kresoxim-methyl (1%), hexaconazole (1%), azoxyxtrobin (2%), tebuconazole (2%), pyrimethanil (6%), and propyzamide (12%) have transient properties (i.e., borderline leacher) according to Groundwater Ubiquity Score (GUS) Index [25]. The concentrations found in the total volume of water leached were higher than the drinking water limit set by the EU (0.1 g L1 ) in all cases, ranging from 1.72 to 96.3 g L1 for hexaconazole and pirimicarb, respectively. Field studies corroborated the results obtained in laboratory conditions. On the other hand, different blank experiments were performed prior to the photocatalytic procedure. Compounds and solar UV spectra do not overlap (or light absorption above 300 nm, beginning of sunlight spectrum at earth surface, is very scarce) showing that capacity for absorption of solar photons cannot produce photodegradation after being exposed to sunlight. Also hydrolysis experiments at the same concentration as the photocatalysis experiments were performed to evaluate that the results obtained during photocatalytic assays were not due to hydrolytic degradation. Hydrolysis tests were done at pH = 9 and 25 C. For all compounds, hydrolysis was irrelevant after 1 week. 3.2. Photolysis and photocatalysis kinetics Previous to the beginning of each assay the collectors were covered and water circulated during 15 min to homogenise the solution. The initial pesticide concentrations ranged from 0.33 to 0.55 mg L1 . Fig. 5 shows the comparative degradation of pesticides in leaching water by use (photocatalysis) or not (photolysis) of the selected photosensitizer (ZnO) during the studied photoperiod. Irradiation of water in the absence of ZnO shows that photolytic decomposition of the pesticides occurs at a much slower rate with percentages remaining at the end of the experiment ranged from 75 to 90% of the initial amount. Furthermore, blank experiments showed that the addition of the catalyst without solar radiation had a negligible effect on pesticide degradation. This small fall in pesticide concentration can be possibly due to the presence of some ions in leaching water such as nitrite (0.12 mg L1 ) and mainly nitrate (547 mg L1 ). The presence of ions either enhances or decreases the reaction rate depending on

1308

S. Navarro et al. / Journal of Hazardous Materials 172 (2009) 13031310

Fig. 5. Disappearance kinetics of the studied pesticides by photolysis () and photocatalysis ( ) with ZnO during the photoperiod (as t30W ). Error bars denote standard deviation. Observed versus predicted kinetics according to a rst-order model for photocatalysis experiments are shown in the inserted graphics (vertical dash lines show DT75 ).

S. Navarro et al. / Journal of Hazardous Materials 172 (2009) 13031310

1309

the mechanism. Most of the anions and cations in leaching water such as chloride 446 (mg L1 ), sulphate (1.498 mg L1 ), sodium (344 mg L1 ) or potassium (165 mg L1 ) are transparent to solar irradiation. However, some metal cations, nitrate and nitrite show any absorbance. Nitrite ( max = 355) and nitrate ( max = 303) absorb light and undergo homolysis to produce OH radicals and nitrogen reactive species (NO , NO2 , N2 O3 , and N2 O4 ) leading to the degradation of pesticides [2629], although OH may further be scavenged by NO2 to form NO2 [30]. Also, the presence of other ions such as calcium (460 mg L1 ) and magnesium (157 mg L1 ) can accelerate de photodegradation as reported by Larson and Weber [27], and the presence of some traces of Fe2+ in leaching water can contribute to the degradation following photo-Fenton reaction [31]. Although only traces of manganese (0.02 mg L1 ) and copper (0.04 mg L1 ) were found in leaching water those small amounts of Mn2+ and Cu2+ ions could have catalytic effect in the degradation of the pesticides by trapping photogenerated electrons and thereby reducing e /h+ recombination [15]. On the other hand, it is known that humic substances (HS), the recalcitrant constituents of natural organic matter, can exhibit photocatalytic effect to pesticide residues in aqueous environment. Total organic carbon (TOC) was measured at the end of each assay. The mean TOC value recorded at that time for photocatalysis assays was 39 7 mg L1 . The initial TOC concentration could not be done with precision, as some of the pesticides usually adhere to the reactor walls due to hydrophobic components. For photolysis experiment, after 480 min 102 12 mg L1 of TOC were measured which it is indicative of a slow mineralization in absence of the photosensitizer. In a previous study [32] we have veried for leaching water a quick degradation (3) in comparison with drinking water (initial TOC = 1.1 mg L1 ). The photocatalytic effect of dissolved organic matter (DOM) is contradictory in the literature. The sunlight absorbance by DOM could provide a rich variety of photochemicals reactions resulting reactive transients produced mainly by HS which could participate in energy and electron transfer and free radical generation [33]. On the contrary, HS can reduce the pesticide photodegradation because the sensitization effect is hidden by the strong lter effect (quenching). They could act as one of the important sunlight-absorbing (especially on UV range), they undergo photolysis under incident UVVIS light and they can also quickly react with OH radicals [34]. Additionally, sorption protects pesticides from photolysis by competitive light attenuation, by migration of the pesticide into regions where light does not penetrate, or by quenching of the excited states of substrates by constituents of the particles [27]. The role of photocatalyst is to accelerate the rate of the reaction without being consumed in it. The amount of catalyst depends on the type and dimensions of the photoreactor used and the kind of pollutants to degrade [15]. In our case the optimal dose was 150 mg L1 . Higher dose did not signicantly increase the yield of the process. As can be seen in Fig. 4, for photocatalysis experiment

more than 75% of the pesticides amount initially present in leaching water was degraded after 1 h of illumination (t30W 3 min). Knowledge of kinetics is required to assess the efciency of systems used for the photooxidation of pesticides. The photolysis kinetic parameters of pesticides are shown in Table 2 where apparent rate constants and half-lives are listed. In all cases, the kinetics of dissipation followed an apparent rst-order degradation curve with R2 ranging from 0.981 to 0.995 and standard error of estimate lower than 6.5 in the most unfavourable case. The carbamic insecticide pirimicarb was quickly degraded. The amount of time required for 50% of its initial pesticide concentration to dissipate was about 10 min (t30W = 0.47 min). Similar behaviour was observed for pyrimethanil (anilinopyrimidine) and propyzamide (benzamide) where 20 min (t30W = 1 min) were necessary to dissipate 50%. The half-lives for strobilurin compounds (azoxystrobin and kresosim-methyl) were about 24 min (t30W = 1.2 min) while for triazole fungicides, the calculated values ranged from 25 to 35 min (t30W = 1.21.8 min) for tebuconazole and triadimenol, respectively. Obviously, the high rate degradation of all pesticides in the presence of ZnO can clearly be ascribed to the catalysts activity. The ability of the semiconductor to act as sensitizer and consequently to enhance the photodegradation of the pesticides studied can be attributed to its electronic structure which is characterized by lled valence band and an empty conduction band. When ZnO is illuminated with energy higher than its bandgap energy, electron/hole pairs are produced: ZnO + hv(<380 nm) h+ + e- . These species can either recombine or react with electron donors or electron acceptors which are adsorbed on the semiconductor surface or trapped within the surrounding electrical double layer of the charged particles and to produce hydroxyl radicals ( OH) which can promote the oxidation of pesticides. One practical problem in using ZnO or TiO2 as photosensitizers is e /h+ recombination, which in the presence of proper electron acceptor is extremely efcient. For this purpose, air was introduced in the tank at regular intervals to maintain the O2 concentration around 810 mg L1 . Molecular oxygen plays an important role in the photooxidation process. The principal role of dissolved oxygen in the photodegradation process is to act as an electron sink, although some authors [35] suggest that it can act as inner lter or scavenger because molecular oxygen has absorption bands at 185 and 254 nm. However, sunlight spectrum at earth surface begins at 300 nm. Therefore, the lter effect was not important in our case. An important parameter in the photocatalysed degradation of pesticides is the pH of the solution because it inuences the surface charge of the semiconductor. In our case, the experiments were carried out at an initial pH of 88.4. During the photodegradation process, there was a slow decrease in the pH (no more than 0.6 units). According to some authors [16,36], the maximum effectiveness for the ZnO system occurs between pH 6 and 8. At acidic pH, ZnO can react with acids to produce the corresponding salt and

Table 2 Kinetic parameters (rate constants and half-lives) for photocatalysis and photolysis of the studied pesticides in leaching water under natural sunlight. Pesticide C0 (%)a Photocatalysis Azoxystrobin Hexaconazole Kresoxim-methyl Pirimicarb Propyzamide Pyrimethanil Tebuconazole Triadimenol
a b

K (min1 )a Photolysis 99.99 (1.21) 98.74 (1.58) 100.59 (1.91) 95.51 (2.10) 97.27 (1.57) 96.95 (2.22) 98.23 (2.06) 97.43 (2.03) Photocatalysis 0.6014 (0.0460) 0.4762 (0.0550) 0.5862 (0.0451) 1.4727 (0.1931) 0.7019 (0.0547) 0.7370 (0.0636) 0.5475 (0.0647) 0.3885 (0.0418) Photolysis 0.0051 (0.0011) 0.0045 (0.0013) 0.0132 (0.0018) 0.0066 (0.0018) 0.0057 (0.0013) 0.0076 (0.0019) 0.0087 (0.0019) 0.0053 (0.0017)

t1/2 (min)b Photocatalysis 1.15 1.46 1.18 0.47 0.99 0.94 1.27 1.78 Photolysis 135.9 154.0 52.5 105.0 121.6 91.2 79.7 130.7

100.59 (3.26) 97.47 (4.37) 100.35 (3.26) 98.93 (5.39) 98.10 (3.31) 99.14 (3.72) 98.69 (4.32) 96.21 (4.01)

Standard error in parenthesis. Referred to t30W (calculated according Section 2.3).

1310

S. Navarro et al. / Journal of Hazardous Materials 172 (2009) 13031310 [2] S. Navarro, N. Vela, G. Navarro, Spanish J. Agric. Res. 5 (2007) 357375. [3] R.D. Wauchope, S. Yeh, J. Linders, R. Kloskowski, K. Tanaka, B. Rubin, A. Katayama, W. Krdel, Z. Gerstl, M. Lane, J.B. Unsworth, Pest Manage. Sci. 58 (2002) 419445. [4] C.E. Clapp, M.H.B. Hayes, N. Senesi, N.P.R. Bloom, P.M. Jardin, Humic Substances and Chemical Contaminants, Soil Science Society of America, Madison, WI, 2001, pp. 502. [5] D. Mackay, S. Paterson, Environ. Sci. Technol. 25 (1991) 427. [6] M. Arias-Estvez, E. Lpez-Periago, E. Martnez-Carballo, J. Simal-Gndara, J.C. Mejuto, L. Garca-Ro, Agric. Ecosys. Environ. 123 (2008) 247260. [7] J.A. Elliott, A.J. Cessna, K.B. Best, W. Nicholaichuk, L.C. Tollefson, J. Environ. Qual. 29 (2000) 16501656. [8] EEA, Environmental Signals 2001. Environmental Assessment Report No. 8 (European Environment Agency, Copenhagen, Denmark, 2001). http://reports.eea.eu.int/signals-2001. [9] EEA, Pesticides in Groundwater. Indicator WHS1a (European Environment Agency, Copenhagen, Denmark, 2004). http://themes.eea.europa.eu/Specic media/water/indicators/WHS01a%2C2004.05. [10] M. Schamann, P. Menger, G. Prokop, Contaminated sites management in CEE countries, in: Remediation of Polluted Sites in CEE Countries: Current Status and Perspectives, ICS-UNIDO, Trieste, Italy, 2002. [11] E. Pelizzetti, E. Pramauro, C. Minero, N. Serpone, E. Borgarello, Photodegradation of organic pollutants in aquatic systems catalyzed by semiconductors, in: M. Schiavello (Ed.), Photocatalysis and Environment Trends and Applications, NATO ASI Series, Kluiver, Dordrecht, The Netherlands, 1988. [12] J.M. Hermann, Catal. Today 53 (1999) 115129. [13] J. Blanco, S. Malato, Solar Detoxication. UNESCO, Natural Sciences, World Solar Programme 19962005 (http://www.unesco.org/science/wsp/publications/solar.htm), 2001. [14] S. Malato, J. Blanco, J.M. Herrman, Catal. Today 54 (1999) 191377. [15] K. Kabra, R. Chaudhary, R.L. Sawhney, Ind. Eng. Chem. Res. 43 (2004) 76837696. [16] E. Evgenidou, K. Fytianos, I. Poulios, Appl. Catal. B: Environ. 59 (2005) 8189. [17] M. Hincapi, M.I. Maldonado, I. Oller, W. Gernjak, J.A. Snchez-Prez, M.M. Ballesteros, S. Malato, Catal. Today 101 (2005) 203210. [18] I. Oller, W. Gernjak, M.I. Maldonado, L.A. Prez-E strada, J.A. Snchez-Prez, S. Malato, J. Hazard. Mater. B138 (2006) 507517. [19] M.I. Maldonado, P.C. Passarinho, I. Oller, W. Gernjak, P. Fernndez, J. Blanco, S. Malato, J. Photochem. Photobiol. A: Chem. 185 (2007) 354363. [20] I.K. Konstantinou, T.A. Albanis, Appl. Catal. B: Environ. 42 (2003) 319335. [21] S.J. Pearton, C.R. Abernathy, M.E. Overbeg, G.T. Thaler, D.P. Norton, N. Theodoropolou, F. Hebard, Y.D. Park, F. Ren, J. Kim, L.A. Boatner, J. Appl. Phys. 93 (2003) 113. [22] Agriculture & Environment Research Unit (AERU) at the University of Hertfordshire, The Pesticide Properties DataBase (PPDB). http://www.herts.ac.uk/aeru/footprint, 2009. [23] S. Malato, J. Blanco, A. Vidal, C. Richter, Appl. Catal. B: Environ. 37 (2002) 115. [24] OECD. Organisation for Economic Cooperation and Development Guidelines for Testing of Chemicals. No. 312, Leaching in Soil Columns. Paris, 2007. [25] D.I. Gustafson, Environ. Toxicol. Chem. 8 (1989) 339357. [26] P. Warneck, C. Wurzinger, J. Phys. Chem. 92 (1988) 62786283. [27] R.A. Larson, E.J. Weber, Reactions Mechanisms in Environmental Organic Chemistry, CRC Press, Boca Raton, FL, 1994. [28] A. Torrents, B.G. Anderson, S. Bilboulian, W.E. Johnson, C.J. Hapeman, Environ. Sci. Technol. 31 (1997) 14761482. [29] J. Mack, J.R. Bolton, J. Photochem. Photobiol. A 128 (1999) 113. [30] M.V. Shankar, S. Nlieu, L. Kerhoas, J. Einhorn, Chemosphere 71 (2008) 14611468. [31] R.A. Larson, M.B. Schlauch, K.A. Marley, J. Agric. Food Chem. 39 (1991) 20572062. [32] J. Fenoll, N. Vela, E. Ruiz, P. Flores, G. Navarro, P. Helln, S. Navarro, Presented at 5th European Conference on Pesticides and Related Organic Micropollutans in the Environment and 11th Symposium on Chemistry and Fate of Modern Pesticides. Marseille, France, 2008 (unpublished data). [33] I.K. Konstantinou, A.K. Zarkadis, T.A. Albanis, J. Environ. Qual. 30 (2001) 121130. [34] M.E. Lindsay, M.A. Tarr, Environ. Sci. Technol. 34 (2000) 444449. [35] H. Shirayama, Y. Tobezo, S. Taguchi, Water Res. 35 (2001) 19411950. [36] E. Evgenidou, K. Fytianos, I. Poulios, J. Photochem. Photobiol. A: Chem. 175 (2005) 2938. [37] M. Hincapi, G. Penuela, M.I. Maldonado, O. Malato, P. Fernndez-Ibanez, I. Oller, W. Gernjak, S. Malato, Appl. Catal. B: Environ. 64 (2006) 272281.

at alkalic pH, it can react with a base to form some complexes. The result is the dissolution and photodissolution of ZnO in those cases and in consequence a decrease in the reaction rates. The use of other electron acceptors, such as hydrogen peroxide (H2 O2 ) and inorganic peroxides (S2 O8 2 ) has been demonstrated to notably enhance the rate of degradation of several organic pollutants [13] because they trap the photogenerated electrons more efciently than molecular oxygen. However, preliminary studies carried out by us using H2 O2 in combination with ZnO show that the addition of peroxide does not improve the result or even, it can have a negative effect if an excess of peroxide because it can act as OH scavenger. Similar ndings were found by Evgenidou et al. [36] for the photooxidation of dimethoate. On the contrary, the addition of Na2 S2 O8 (100 mg L1 ) to the tank was benecial because the oxidant strongly enhances the reaction rate for all pesticides. No signicant differences were observed in the reaction rate when increasing the concentration of the oxidant although unlike H2 O2 , high concentrations of peroxydisulfate were not detrimental to the reaction rate. As pointed by some authors [36,37], the peroxydisulfate enhancement effect is related to both scavenging of electrons and production of additional oxidising species. It traps the photogenerated electron and reduces the probability of recombination (mayor cause of ZnO photocatalysis quantum yield) with the positive hole, generating SO4 radicals, which are also a very strong oxidizing species (E = 2.6 V) and more hydroxyl radicals. This implies a signicant reduction in treatment time, from 4 to 6 times faster than ZnO alone. 4. Conclusions The use of solar photocatalysis in presence of ZnO as photosensitizer constitutes a very effective method for the reduction and even elimination of the selected pesticides in leaching water. A synergistic effect was observed with the addition of the oxidant (Na2 S2 O8 ) into illuminated ZnO suspensions by enhancing the efciency of the process. The amount of time required for 50% of the initial pesticide concentration to dissipate ranged from 10 to 35 min in our conditions (equivalent to t30W = 0.52 min) and complete mineralization was achieved after 2 h of treatment. Bearing in mind the extensive use of pesticides in some Mediterranean agricultural areas, and the groundwater pollution problems, photocatalysis offers a good technology as substitute to other conventional methods for water remediation by using a renewable source of energy, inexhaustible and pollution-free, like sunlight mainly in some areas as SE of Spain receiving more than 3000 h of sunlight per year. Acknowledgements The authors acknowledge nancial support from Instituto Nacional de Investigacin y Tecnologa Agraria y Alimentaria (INIA), of Spain, research project number RTA2005-00127-00-00. References
[1] M. Vanclooster, J.J.T.I. Boesten, M. Trevisan, C.D. Brown, E. Capri, O.M. Eklo, B. Gottesbren, V. Gouy, A.M.A. van der Linden, Agric. Water Manage. 44 (2000) 119.