Subcritical Water Extraction Followed by Liquid Chromatography Mass

Due to the great potential of atrazine in contaminating groundwater, its use has been banned in several countries and often replaced by terbuthylazine...
0 downloads 0 Views 159KB Size
Environ. Sci. Technol. 1999, 33, 3271-3277

Subcritical Water Extraction Followed by Liquid Chromatography Mass Spectrometry for Determining Terbuthylazine and Its Metabolites in Aged and Incubated Soils A N T O N I O D I C O R C I A , * ,† ANNA BARRA CARACCIOLO,‡ CARLO CRESCENZI,† GIUSEPPE GIULIANO,‡ SUSANNA MURTAS,† AND ROBERTO SAMPERI† Dipartimento di Chimica, Universita` “La Sapienza”, Piazza Aldo Moro 5, 00185 Roma, Italy, and Istituto di Ricerca sulle Acque (IRSA), via Reno 2, Roma, Italy

Due to the great potential of atrazine in contaminating groundwater, its use has been banned in several countries and often replaced by terbuthylazine (CBET). Little is known on the fate of CBET in soil. The purpose of this work has been (1) to develop a general method for analyzing CBET and its degradation products (DPs) in soil and (2) to use this method for elucidating the fate of CBET incubated in both surface and subsurface samples of an agricultural soil which had been receiving repeated CBET spills. This method involves analyte extraction from soil at 100 °C by phosphate-buffered water. Analytes coming out of the extraction cell were collected by a graphitized carbon black extraction cartridge. After analyte elution with a suitable solvent mixture, the final extract was analyzed by LCMS. From an aged soil, our method extracted altogether quantities of CBET and its DPs respectively 2.1 and 1.4 times larger than those by two previously reported methods. For the analytes considered, limits of quantification (S/N 10) ranged between 0.22 and 5.5 ng per gram of soil. The laboratory CBET degradation experiment showed that (1) similarly to atrazine, remarkable amounts of hydroxylated metabolites were formed; (2) when the subsoil microflora was in the presence of rather large amounts of CBET, it degraded the herbicide with a rate similar to that of the topsoil microflora.

Introduction Over the last 35 years, s-triazine herbicides, mainly atrazine, have been the most extensively used for selective weed control of several types of crops. The fate and behavior of s-triazines in soil has raised environmental concern because these compounds and their degradation products (DPs) have been frequently detected in groundwater. Furthermore, s-triazine DPs exhibit herbicidal effects that, like parent compounds, prevent crop rotation in successive years (1, 2). Degradation by soil microflora is considered to be the major sink for s-triazines in the environment. The actual * Corresponding author phone: +39-06-49913752; fax: +39-06490631; e-mail: [email protected]. † Universita ` “La Sapienza”. ‡ Istituto di Ricerca sulle Acque (IRSA). 10.1021/es990130b CCC: $18.00 Published on Web 08/11/1999

 1999 American Chemical Society

knowledge of the degradation mechanisms of s-triazines mostly derives from studies involving atrazine. There is a body of evidences showing that the primary mechanism of atrazine degradation is oxidative N-dealkylation, mainly due to fungi (3). Nonbiological and biologically mediated (4) dehalogenation processes convert intact atrazine as well as its dealkylated metabolites to corresponding hydroxylated forms. Recently, a soil bacterial isolate has been shown to be capable of cleaving and partially mineralizing the triazine ring of atrazine (5). The great potential that atrazine has of contaminating groundwaters has urged issuing of regulations aimed at restricting its use or its prohibition. Terbuthylazine (CBET) has replaced atrazine in many areas where atrazine use has been banned. Unlike atrazine, relatively little is known on the fate and behavior of CBET in the environment. Those few authors who studied the mode of dissipation of CBET in soil found that (1) CBET was more stable and less mobile in soil than atrazine (6-9), (2) it was not mineralized by soil microorganisms (10, 11), and (3) the only metabolite detected in soil by CBET degradation was desethylterbuthylazine (9, 10-12). For analyzing s-triazines and their DPs in soils, the most widely used extraction method is organic solvent extraction by Soxhlet (13-16), batch extraction (17-19), and microwaveassisted extraction (MAE) (20). Generally, these methods make use of large volumes of toxic, flamable, and expensive solvents. As an alternative, supercritical fluid extraction with CO2 modified with methanol has been proposed to extract atrazine and its DPs from soil samples (21-23). However, unsatisfactory low recovery ranging around 25% of atrazine and its DPs in freshly spiked solid matrices was obtained (24). MAE with water near its boiling point has been used by Steinheimer (25) to extract atrazine and two related dealkylated metabolites from spiked agricultural soils. Field et al. (26) used water at 50 °C and 200 bar with in situ trapping on an anion exchanger disk for exhaustively and rapidly extracting two acid metabolites of a herbicide in aged soils. Hawthorne et al. (27) showed that water efficiently extracted medium-polar compounds, such as chlorophenols, from sand at 50 °C, while polycyclic aromatic hydrocarbons and n-alkanes remained adsorbed on the matrix. Over organic solvents, one of the advantages of using moderately heated water as extractant is that substantially cleaner extracts can be obtained, thus avoiding lengthy purification procedures of crude soil extracts. Very recently, we succeeded in rapidly extracting 18 pesticides from soils by water at 90 °C (28). The extraction equipment used by us was similar to that devised by Hawthorne (27), except that analyte collection was performed by a graphitized carbon black (Carbograph 4) solid-phase extraction (SPE) cartridge set on line with the extraction cell instead of liquid-liquid partitioning. Pesticides in the final extracts were quantified by liquid-chromatography-mass spectrometry (LC-MS) with an electrospray (ES) ion source. The purpose of this work has been 2-fold. One has been to develop a sensitive and efficient method based on hot water extraction with sorbent collection followed by LC-MS for monitoring CBET and its DPs in soils. The second objective has been that of evaluating, under laboratory conditions, the degradation rate and transformation pathway of CBET in both surface and subsurface soil layers taken from a field which had been treated repeatedly with this herbicide for several years. VOL. 33, NO. 18, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3271

TABLE 1. Physical and Chemical Characteristics of the Soils Used in This Study soil

sand, %

silt, %

clay, %

organic content, %

pH

Cadriano Paliano Camporeale

45 23 55

24 44 27

31 33 18

1.4 2.3 4.2

6.8 6.1 6.2

Experimental Section Reagents and Chemicals. Terbuthylazine (N2-tert-butyl-6chloro-N4-ethyl-1,3,5-triazine-2,4-diamine) (CBET), desethylterbuthylazine (CBAT), debutylterbuthyilazine (CEAT), desethyldebutylterbuthylazine (CAAT), 2-hydroxyterbuthylazine (OBET), 2-hydroxydesethylterbuthylazine (OBAT), 2-hydroxydebutylterbuthylazine (OEAT), sec-buthylazine (secBA), and 2-hydroxydesethyl-sec-buthylazine (DEOH-sec-BA) were purchased from Alltech, Sedriano, Italy. The latter two compounds were used as internal standards. Individual standard solutions of the nonhydroxylated compounds were prepared by dissolving each compound in methanol to obtain 1 mg/mL concentration, except for CAAT (0.1 mg/mL). Standard solutions of the hydroxylated species were prepared by dissolving them in methanol acidified with HCl, 5 mmol/ L, to obtain 0.5 mg/mL concentrations. A composite working standard solution of the target compounds was prepared by mixing the above solutions and diluting with methanol to obtain analyte concentrations of 2 µg/mL. A solution containing the two internal standards was prepared in the same way. When unused, all solutions mentioned above were stored at 4 °C. Methanol “Plus” of gradient grade was obtained from Carlo Erba, Milano, Italy. All other solvents and chemicals were of analytical grade (Carlo Erba) and were used as supplied. Sand (crystobalite, 40-200 mesh) was obtained from Fluka AG, Buchs, Switzerland. Extraction cartridges filled with 0.5 g of Carbograph 4 were supplied by LARA, Rome, Italy. They were prepared and used as previously reported (28). Soil Samples for Recovery Studies. For recovery studies, three analyte-free soil samples from various locations were considered. Table 1 gives some details of their compositions. Soils were heated at 110 °C until they were free flowing and sieved to obtain a particle size range between 37 and 420 µm. Soil samples amended with CBET and its six DPs at the level of 30 ng/g were prepared and laboratory-aged as previously reported (28). Selection of the Site for CBET Degradation Studies. Soil samples were taken at a site (October 1997, Manerbio, Italy) where maize had been intensively cropped over several years with repeated CBET applications. A coring technique developed by the Italian “Istituto di Ricerca sulle Acque” was used to obtain the samples. A steel tube (12 cm in diameter and 65 cm long) was forced vertically into the ground. The steel tube containing the core was sealed at both ends, stored at 5 ((2) °C, and sent to the laboratory within 12 h after sampling. Two portions of the core, respectively in the upper part (5-25 cm of depth) and the lower part (40-60 cm of depth), were considered for CBET degradation experiments. The chosen intervals are representative of the geopedologic horizons A (soil) and BC (subsoil). Abiotic and biotic features of the two soil samples are reported in Table 2. Soil Incubation Studies. After some manual homogenization of the soil samples, 350-g (dry weight) aliquots of them were poured in four flat dishes (two for the topsoil and two for the subsoil). One sample of the topsoil and one of the subsoil were amended with CBET dissolved in sterile water (8.5 mg/L) to obtain a herbicide concentration of 1.5 mg/kg. The remaining two soil samples received equivalent volumes of pure sterile water (control soils). Final soil moisture was 3272

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 18, 1999

TABLE 2. Main Abiotic and Biotic Characteristics of the Soil Samples Taken from the Manerbio Site horizon depth of sample (cm) pH (soil/H2O 1:1, w/v) coarse sand (%) fine sand (%) coarse silt (%) fine silt (%) clay (%) organic carbon (%) texture class CEC (meq/100 g) moisture content (%) no. bacteria/g

soil

subsoil

A 5-25 7.2 37 16 9 20 19 1.8 sandy loam 14 18 1.2 × 108

Bc 40-60 7.4 38 15 12 12.5 22.5 0.14 sandy clay loam 11 16 2.1 × 107

25-27% gravimetric moisture content (wet weight basis). Soils were thoroughly stirred with a spatula to distribute homogeneously the herbicide and left in contact with CBET for 24 h. Fifteen gram aliquots of contaminated and control soils were then poured in sterilized glass bottles which were sealed and maintained in the dark at 15 °C. These conditions were chosen in order to simulate as close as possible those existing in a subsoil. At intervals, soil samples were taken and analyzed in duplicate following procedures reported in this section. In any case, single measurements of analyte and bacteria concentrations did not differ from the mean values by more than 10%. Microbial Analysis. Microbial analysis was performed by suitably modifying a previously reported procedure (29). This procedure uses a fluorescent agent to distinguish bacteria (that appear in a luminescent blue color) from nonliving bacterium-sized particles (that appear yellow). Briefly, bacteria were extracted from 1 g of soil with 10 mL of water containing 1.5% formaldehyde and 0.5% Tween 20. After shaking for 20 min, the suspension was left for 24 h in order to allow larger particles to settle out. Then, the 4′,6-diamino2-phenylindole fluorescent agent was added to an aliquot of the supernatant in the dark at 4 °C. After 30 min, the solution was filtered through a 0.2 µm filter pad, and bacteria were counted by epifluorescence microscopy. Extraction Procedure. A homemade extraction apparatus equal to that used in a previous study (28) was used for extracting the analytes from soil samples. Three grams of dried and sieved soil was first mixed with 2 g of sand and then poured into the extraction cell. Any void space remaining after packing the soil was filled with sand. The soil-containing tube was then put into the oven and heated at 100 °C for 5 min. A total of 10 mL of phosphate-buffered water (0.5 mol/ L, pH 7.5) was used to extract the analytes. Water was initially passed through the cell at a flow rate of 0.3 mL/min for 7 min, and then the flow rate was increased to 1 mL/min in 2 min. The Carbograph 4 cartridge was then disconnected and fitted into a sidearm filtering flask. After passing 100 mL of distilled water through the cartridge at a flow rate of ca. 80 mL/min, the cartridge was turned upside down, and analytes were re-extracted by following a previously reported procedure (30). Before solvent removal, 10 µL of the solution containing the two internal standards was added to the eluate. The eluant phase was eliminated at 40 °C in a water bath with a gentle nitrogen flow. The residue was reconstituted with 150 µL of a water/methanol solution (40:60, v/v) and acidified with HCl (pH 3), and 15 µL of the final extract was injected into the LC column. LC-ES/MS Analysis. The LC-MS equipment and several of the instrumental conditions used in this work were the same as reported elsewhere (30). For fractionating the analytes, the phase A was methanol and the phase B was

TABLE 3. Time-Scheduled Multiple-Ions SIM Conditions for Detecting Terbuthylazine and Its Potential Metabolites and Limits of Detection (LOD) compound OEAT CAAT DEOH-secBA(I.S.) OBAT CEAT OBET CBAT sec-BA(I.S.) CBET

channel mass, m/z (relative abundance)

cone retention voltage, window, LOD, V min pg

86 (90), 114 (50), 156 (100) 104 (30), 146 (100), 148 (25) 128 (80), 184 (100)

40 40 30

0-14 500 14-17.5 300 17.5-25

128 (100), 184 (40) 132 (30), 174 (100), 176 (25) 156 (100), 212 (30) 146 (100), 148 (25), 202 (40) 174 (40), 176 (10), 230 (100), 232 (25) 174 (100), 176 (25), 230 (40), 232 (10)

30 30 30 30 30

17.5-25 20 25-31 140 25-31 50 31-35 30 35-40

30

35-40

30

water. Both solvents contained formic acid, 0.2 mmol/L. The initial composition of the mobile phase was 0% A that was first increased to 15% after 10 min and then to 100% after 30 min. Structure significant fragment ions of the pseudomolecular ions were generated by the in-source collisioninduced decomposition (CID) process, after suitably adjusting the skimmer cone voltage. Conditions followed to detect target compounds in the time-scheduled multiple-ion selected ion monitoring (SIM) mode are reported in Table 3. Quantitation. Analytes were quantified by the external standard quantification procedure. Standard solutions were prepared at nine levels by using appropriate volumes of the working standard solution. For each analyte, the peak area vs injected amount chart was obtained by measuring at any injected amount the peak area resulting from the sum of the ion currents relative to parent and fragment ion(s) and relating this area to that of the internal standard (note that for the nonhydroxylated analytes the reference internal standard was sec-BA, while for the hydroxylated ones was DEOH-sec-BA). The response of the ES/MS detector was linearly related to injected amounts of the analytes up to 50-70 ng. Thus, the initial large quantities of CBET in the incubated soils were quantified by the UV detector set at 220 nm wavelength.

Results and Discussion Optimization of the Extraction Conditions. Currently, absorption into the soil organic matter, mainly humic acids, is considered the primary mechanism by which xenobiotics are sorbed to soil. Since humic acids are in a sort of solid state, this generally accepted theory depicts sorption of xenobiotics as molecules entrapped into humic acid clots essentially by nonspecific (hydrophobic) interactions. Hydroxytriazines are weakly basic compounds. Starting from this consideration and on the basis of several indirect experimental evidences, Lerch et al. (31) elaborated a mixedmode sorption model involving simultaneously hydrophobic interactions and cation exchange with both soil mineral components and humic acids coating the soil particles. They individuated dissociated soil silanols and carboxylated groups of humic acids, both of them being weak cation exchangers, as sites responsible for specific binding of hydroxytriazines to soil. Following this sorption model, they tailored an extractant composed of phosphate-buffered water (0.5 mol/ L, pH 7.5)/CH3CN (3:1, v/v) and succeeded in extracting at room temperature hydroxytriazines reversibly bound to an aged soil. In the opinion of the above authors, the phosphate buffer plays the role of detaching hydroxytriazines from the exchange sites by deprotonating them, while CH3CN serves to enhance solubility of hydroxytriazines in the extractant. Initial experiments conducted by us showed that, at a given

FIGURE 1. Concentrations of terbuthylazine (CBET) and its metabolites in an aged soil as measured by varying the extraction cell temperature. Extractant: 10 mL phosphate-buffered (0.5 mol/L, pH 7.5) water. See the text for acronym explanation. extraction temperature, phosphate-buffered water (0.5 mol/ L, pH 7.5) was indeed more efficient than water alone in extracting both hydroxylated and nonhydroxylated triazines from an aged soil. The extraction efficiency of water steadily increases as its temperature is increased. On the other hand, water extraction performed at too high temperatures is expected to give soil extracts containing large amounts of unwanted naturally occurring low-polar species. More important, chlorotriazines sorbed to soil can be slowly converted to hydroxytriazines by water near its boiling point (18). Therefore, we evaluated the optimum temperature at which maximum extraction yield of CBET and its metabolites could be achieved without decomposition of the chlorotriazines. For this purpose, we selected the same aged agricultural soil used for laboratory degradation studies of CBET (see the Experimental Section) but sampled in a former time. At each temperature considered, analyses were made in triplicate, and mean concentrations measured for each target compound were plotted against the extraction temperature (Figure 1). A typical mass chromatogram obtained by analyzing a Manerbio soil sample is shown in Figure 2. As a general remark, no trace of metabolites generated by deterbutylation of CBET was detected in soil extracts. Conceivably, removal of the tert-butyl group by microorganisms is a process greatly inhibited by steric hindrance, or this process cannot take place due to the absence of any hydrogen atom bonded to the tertiary carbon one. Unfortunately, no detailed reaction mechanism was found in the literature to support the latter hypothesis. Increasing the temperature had the expected effect of enhancing remarkably the extraction yields of CBET and its degradation products. As an example, the amount of CBET extracted by phosphate-buffered water at 100 °C was about nine times larger than that extracted at 30 °C. The amounts VOL. 33, NO. 18, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3273

TABLE 4. Concentrations (ng/g) of Terbuthylazine (CBET) and Its DPS in an Aged Agricultural Soil As Measured by This Method with Three Different Extractants and by Other Two Proposed Extraction Methods acetate phosphate buffer buffer (0.5 mol/L, (0.5 mol/L, compd pure water pH 4.5) pH 7.5) method Aa method Bb OEAT CAAT OBAT CEAT OBET CBAT CBET

nd nd 14(1.0)c nd 34 (2.4) 1.1 (0.12) 2.0 (0.18)

nd nd 19 (1.6) nd 45 (2.5) 1.5 (0.19) 2.5 (0.21)

nd nd 20 (1.5) nd 46 (2.7) 1.7 (0.20) 3.2 (0.23)

nd nd 6.5 (0.8) nd 25 (1.9) 0.7 (0.16) 1.6 (0.23)

nd nd 13 (1.3) nd 33 (2.3) 1.6 (0.18) 2.9 (0.20)

a Method A: Soxhlet extraction with methanol for 24 h (14). b Method B: batch extraction at room temperature with phosphate buffer (0.5 mol/L, pH 7.5)/CH3CN (3:1, v/v) (33). c Mean values from triplicate measurements and standard deviation (in parentheses).

FIGURE 2. Multiple ions SIM mass chromatograms of a standard solution (injected amount of each analyte: 2 ng) and of an aged soil extract. The concentrations of each analyte measured by us are indicated on the apexes of the corresponding peaks. See the text for acronym explanation. of analytes extracted did not further increase either setting the extraction cell temperature at 120 °C or increasing the extractant volume passing through the soil bed. Furthermore, no evidence was obtained for some decomposition of the chlorotriazines. At increasingly extraction temperatures, water leaving the extraction cell took a more and more intense brown color, this indicating that increasing quantities of humic acids were removed from the soil. The mechanism by which phosphate ions dissolve humic acids has been explained elsewhere (31). Capriel et al. (32) showed that those fractions of atrazine as well as its DPs not extractable with conventional organic solvents were mainly absorbed and/or adsorbed on soil humic acids. Presumably, extracting the soil at 100 °C with the phosphate buffer had the multiple effects of dissolving efficiently humic acid clots into which a part of s-triazines could be embedded, driving readily the cation exchange equilibrium back to free hydroxytriazines, and improving solubility of the analytes in water. Under the same experimental conditions, the Manerbio soil was also analyzed by extracting the analytes with both pure and acetate-buffered water (0.5 mol/L, pH 4.5). Results of triplicate experiments were compared with that using phosphate buffer as extractant (Table 4). Pure water was remarkably less efficient than phosphate buffer to recover both hydroxylated and nonhydroxylated triazines. Based on the observation of the solution color coming out of the extraction cell, both pure water and the acetate-buffer were much less efficient than the phosphate buffer in dissolving humic acids. Nevertheless, the acetate buffer recovered hydroxytriazines as efficiently as the phosphate buffer. The mixed-mode sorption model cited above assumes that hydroxytriazines are specifically bound to cation exchange sites in part located on the external layers of humic acid clots. Bearing this in mind, efficient removal of hydroxytri3274

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 18, 1999

azines from the soil components by the acetate buffer supports the above theory instead of weakening it, considering that an acidic solution readily solubilizes cations bound to weak ion exchange groups by suppressing dissociation of the weakly acidic groups. CBET is about 1000 times less basic and more hydrophobic than hydroxytriazines. Therefore, it is conceivable to assume that the former compound is in part sorbed into humic acids by hydrophobic interactions. It follows that the partial failure of the acetate buffer to extract CBET could be related to its poor ability to dissolve humic acid clots into which the above compound is partially entrapped. Method Comparison. In term of extraction efficiency, we compared our extraction procedure with two other proposed procedures which were designed to extract atrazine and its DPs from soil. One (procedure A) is based on Soxhlet extraction with methanol for 24 h (14), while the other one (procedure B) is that prescribed by Lerch et al. (31) and involves double extraction at room temperature with the mixed-mode extractant cited above. For each extraction, the extractant volume-to-soil weight ratio was 2:1. It has to be pointed out that the latter procedure had not been optimized for quantitation of hydroxytriazines in soil, as the intent of the authors was that of demonstrating that hydroxytriazines are mainly bound to aged soils by the reversible mixed-mode sorption mechanism. For comparative evaluation, we measured original concentrations of weathered CBET and its DPs in the same aged Manerbio soil sample we used to optimize the extraction conditions (see above). To analyze soil extracts obtained by the extraction procedures A and B, the two extractants were first diluted with suitable water volumes so that the final organic solvent percentage was no larger than 5%. Second, analytes in the two solutions were extracted by Carbograph 4 cartridges, and finally the rest of the procedure reported in the Experimental Section was followed. Recovery studies conducted by adding known amounts of the analytes to the two water-diluted extractants ascertained that the performance of the extraction cartridge was not at all affected by the presence of the small percentage of either methanol or acetonitrile in water. With every method, extractions were made in triplicate, and results are shown in Table 4. As compared to 24-h extraction with hot methanol, the procedure B employing the mixed-mode extractant gave a better performance, in spite of the fact that the soil was extracted only two times at room temperature. In turn, the procedure developed by us proved to be more effective than the procedure B in extracting hydroxytriazines. Lerch et al. (31) reported that a third extraction step removed from the aged soil an additional amount of hydroxytriazines, which ac-

FIGURE 3. Concentration vs time plots of terbuthylazine (CBET) and its metabolites in the surface and subsurface of a soil amended with CBET. See the text for acronym explanation. Original concentrations of CBET in topsoil and subsoil were respectively 0.024 and 0.0065 nmol/g. M.B. ) mass balance. counted for about 28% of the total amount extracted. It cannot be excluded that other extraction steps could make the procedure B competitive with our procedure. In this case, however, the latter procedure offers advantages over the former one in terms of time and ease of analysis. Matrix Effect. It is generally accepted that the extraction yield of a given compound sorbed to an aged soil depends strictly upon the extent at which the performance of a given extraction procedure is affected by the type of soil or, better, by the soil organic matter content. We assessed whether the extractability of the analytes by our method was affected by the type of soil. For this study, three different soils characterized by different organic matter contents were selected (see Table 1). They were amended with the analytes at the level of 30 ng/g and laboratory-aged, as reported elsewhere (28). After aging, each soil was analyzed four times. In any case, analyte recoveries ranged between 95 and 103% with standard deviations no larger than 4%. Assuming laboratoryaged soils fairly mimicked naturally aged soils, we inferred our extraction procedure was not affected by the particular soil composition. Limits of Quantification. Under multiple-ion SIM conditions and based on the peak-to-peak noise measured on the baseline close to the analyte peak, the instrumental limits of detection (S/N ) 3) were calculated for the seven analytes by eluting them as reported in the Experimental Section and measuring peak heights against average background noise (Table 3). Considering this method involves extraction of 3 g of soil and injection of one-tenth of the final extract, limits of quantification (S/N ) 10) ranged between 5.5 ng (OEAT) and 0.22 ng (OBAT) per gram of soil. Laboratory CBET Degradation Studies. Degradation of a certain pesticide in topsoil is accomplished primarily by microflora specifically adapted to grow on it. Once this pesticide leaches beneath the upper horizon, however, only the ability of deeper soil layers to immobilize or degrade it can hinder groundwater contamination. The main intent of this study was that of evaluating the ability of microorganisms living under oligotrophic conditions of degrading CBET, as compared to that of topsoil microorganisms.

Figure 3 shows concentrations of CBET and its biotransformation products vs time plots resulting from incubation of the herbicide in both surface and subsurface soils at 15 °C. Correspondingly, Figure 4 shows variations of the growth of the bacterial population in both control and CBETamended soils. After relatively short lag-phases, CBET degradation took place in both topsoil and subsoil. This result was expected as CBET was incubated in agricultural soils populated by CBET-adapted microorganisms. Previous studies have indicated that degradation rates of pesticides strongly decrease from the topsoil to deeper soil horizons and are to somewhat extent positively correlated with bacterial population density. In this study, bacterial concentration in the upper horizon was about six times larger than that in the lower horizon. Nevertheless, we found that half-lives of CBET in topsoil and subsoil were not profoundly different, being respectively 143 and 187 days. Our finding could be accounted for by considering that, as observed by some authors (33), the presence in some topsoils of particular readily available C sources can decrease the degradation rate of pesticides. After about 9 months from the beginning of the biodegradation experiment, approximately 27 and 34% of the initial CBET concentration was still present in topsoil and subsoil, respectively. Correspondingly, the bacterial concentrations in the two CBET amended soil samples became similar to those of the control soils. This may be due to the fact that residual CBET was sorbed to soil sites not easily accessible to the degraders or that the degraders were unable to proliferate on a low substrate concentration. Do¨rfler et al. (34) in their lysimeter experiment with C14ring labeled have found small amounts of CEAT amounting to 1.6% of the total extractable radioactivity. Conversely, no trace of metabolites coming from the loss of the tert-butyl group was detected in our experiment, even after 9 months of incubation. The degradation process of CBET produced significant amounts of two hydroxytriazines. Differently from these two metabolites, small but significant amounts of CBAT were formed even during the initial bacteria acclimation phase. As triazine dealkylation is mainly operated by fungi VOL. 33, NO. 18, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3275

FIGURE 4. Bacterial concentration vs time hystograms in the surface and subsurface of a soil amended (herbicide) and not amended (control) with CBET. (3), this suggests they need a short acclimation time. From our experimental data, it was not possible to deduce whether OBAT was generated by dealkylation of OBET or by dechlorination of CBAT. Only after about 5 months, an apparent lack of the mass balance was observed in the degradation experiment of CBET incubated in both surface and subsurface soils. Two previous studies (9, 11) have observed that CBET is not mineralized by the soil microflora. Radosevich et al. (5) reported that a bacterial culture isolated from an agricultural soil previously impacted with atrazine spills was capable of ring cleavage with production of urea and biuret, under both anaerobic and aerobic conditions. Therefore, it is feasible that the lack of mass balance observed in our degradation experiment was due to a phenotype able to open the CBET triazine ring 3276

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 18, 1999

after a long acclimation time. In conclusion, this laboratory CBET degradation study has highlighted that (i) microorganisms living subsoil, when in the presence of relatively large concentrations of CBET, have the capability of degrading this substrate similar to that of the microflora populating the topsoil and (ii) the degradation process of CBET produced hydroxylated metabolites in quantities similar to that of the dealkylated metabolite.

Literature Cited (1) McGlamery, M. D.; Slife, F. W.; Butler, H. Weeds 1967, 15, 3538. (2) Xu, Y.; Lorenz, W.; Pfister, G.; Bahadir, M.; Korte, F. Fresenius Z. Anal. Chem. 1986, 325, 377-380. (3) Kaufman, D. D.; Blake, J. Soil Biol. Biochem. 1970, 2, 73-80.

(4) Mandelbaum, R. T.; Wackett, L. P.; Allan, D. L. Environ. Sci. Technol. 1993, 27, 1943-1946. (5) Radosevich, M.; Traina, S. J.; Hao, Y. L.; Tuovinen, O. H. Appl. Environ. Microbiol. 1995, 61, 297-302. (6) Sahid, I. B.; Teoh, S. S. Bull. Environ. Contam. Toxicol. 1994, 52, 226-230. (7) Bowman, B. T. Environ. Toxicol. Chem. 1989, 8, 485-491. (8) Burkhard, N.; Guth, J. A. Pestic. Sci. 1981, 12, 45-52. (9) Klotz, D.; Do¨rfler, U.; Scheunert, I. Chemosphere 1997, 35, 8798. (10) Gerstl, Z.; Sluszny, C.; Alayof, A.; Graber, E. R. Sci. Total Environ. 1997, 196, 119-129. (11) Dousset, S.; Mouvet, C.; Schiavon, M. Pestic. Sci. 1997, 49, 9-16. (12) Funari, E.; Barbieri, L.; Bottoni, P.; Del Carlo, G.; Forti, S.; Giuliano, G.; Marinelli, A.; Santini, C.; Zavatti, A. Chemosphere 1998, 36, 1759-1773. (13) Sirons, G. J.; Frank, R.; Sawyer, T. J. Agric. Food Chem. 1973, 21, 1016-1020. (14) Xu, Y.; Lorenz, W.; Pfister, G.; Bahadir, M.; Korte, F. Fresenius Z. Anal. Chem. 1986, 325, 377-380. (15) McGlamery, M. D.; Slife, F. W.; Butler, H. Weeds 1967, 15, 3538. (16) Abia´n, J.; Durand, G.; Barcelo´, D. J. Agric. Food Chem. 1993, 41, 1264-1273. (17) Vanderheyden, V.; Debongnie, P.; Pussemier, L. Pestic. Sci. 1997, 49, 237-242. (18) Huang, L. Q.; Pignatello, J. J. J. Assoc. Off. Anal. Chem. 1990, 73, 443-446. (19) Cabras, P.; Spanedda, L.; Pellecchia, M.; Gennari, M. J. Chromatogr. 1989, 472, 411-415. (20) Molins, C.; Hogendoorn, E. A.; Heusinkweld, H. A. G.; van Harten, D. C.; van Zoonen, P.; Baumann, R. A. Chromatographia 1996, 43, 527-532.

(21) Robertson, R. Environ. Sci. Technol. 1994, 28, 346-350. (22) Lopez-Avila, V, Dodhiwala, N. S.; Beckert, W. F. J. Agric. Food Chem. 1993, 41, 2035-2040. (23) Papilloud, S.; Haerdi, W. Chromatographia 1994, 38, 514-519. (24) Papilloud, S.; Haerdi, W.; Chiron, S. Barcelo´, D. Environ. Sci. Technol. 1996, 30, 1822-1826. (25) Steinheimer, T. B. J. Agric. Food Chem. 1993, 41, 588-595. (26) Field, J. A.; Monohan, K.; Reed, R. Anal. Chem. 1998, 70, 19561962. (27) Hawthorne, S. B.; Yang, Y.; Miller, D. J. Anal. Chem. 1994, 66, 2912-2920. (28) Crescenzi, C.; D’Ascenzo, G.; Di Corcia, A.; Nazzari, M.; Marchese, S.; Samperi, R. Anal. Chem. 1999, 71, 2157-2163. (29) Porter, K. G.; Feig, Y. S. Limnol. Oceanogr. 1980, 25, 943-948. (30) Di Corcia, A.; Crescenzi, C.; Guerriero, E.; Samperi, R. Environ. Sci. Technol. 1997, 31, 1658-1663. (31) Lerch, R. N.; Thurman, E. M.; Kruger, E. L. Environ. Sci. Technol. 1997, 31, 1539-1546. (32) Capriel, P.; Haisch, A.; Khan, S. U. J. Agric. Food Chem. 1985, 33, 567-569. (33) Daugherty, D. D.; Karel, S. F. Appl. Environ. Microbiol. 1994, 60, 3261-3267. (34) Do¨rfler, U.; Feicht, E. A.; Scheunert, I. Chemosphere 1997, 99106.

Received for review February 5, 1999. Revised manuscript received June 30, 1999. Accepted July 1, 1999. ES990130B

VOL. 33, NO. 18, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3277