Over the past 10 years, a series of substituted

Transcription

1 1172 ISHIMITSU ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 4, 2001 RESIDUES AND TRACE ELEMENTS Determination of Clethodim and Its Oxidation Metabolites in Crops by Liquid Chromatography with Confirmation by LC/MS SUSUMU ISHIMITSU 1,AKIKO KAIHARA,KIMIHIKO YOSHII,YUKARI TSUMURA,YUMIKO NAKAMURA, and YASUHIDE TONOGAI National Institute of Health Sciences, Osaka Branch, Division of Food Chemistry, Hoenzaka, Chuo-ku, Osaka, , Japan A method was developed for determination of the herbicide clethodim (C0) and its oxidation metabolites clethodim sulfoxide (C1) and clethodim sulfone (C2) in agricultural products. Upon extraction, both C0 and C1 were oxidized to C2 by m-chloroperoxybenzoic acid, and C2 was determined by liquid chromatography (LC). The C2 peak was confirmed by liquid chromatography/mass spectrometry (LC/MS) with electrospray ionization (ESI). Recoveries of C0 from radish, tomato, onion, sweet potato, kidney bean, carrot, cabbage, and lettuce ranged from 91 to 118% following fortification at ppm. The detection limit of C2 in crops was 0.01 ppm (S/N > 3). The fortified samples of onion, sweet potato, kidney bean, and carrot were confirmed by LC/MS (ESI), and the peak of C2 was detected. Over the past 10 years, a series of substituted 1,3-cyclohexanediones (clethodim and sethoxydim) have been introduced as postemergence grass herbicides for use in broadleaf crops. These compounds exhibit a remarkably similar spectrum of phytotoxicity on both annual and perennial grasses without affecting broadleaf plants (1 4). Clethodim (C0) is gradually oxidized into clethodim sulfoxide (C1) or clethodim sulfone (C2) in the field (5). The Ministry of Health and Welfare (MHW) in Japan has already set the maximum residue limits (MRL) in crops for 199 pesticides under the Food Sanitation Law. The MHW intends to set the MRL and the official analytical method for C0 in crops in the near future. Formerly, the Japanese Environment Agency imposed the tolerance of residual C0 and the analytical method for legumes, sugar beet, sweet potato, onion, and carrot on August 24, 1999 (6). However, there are some objections to this analytical method, including the use of toxic dichloromethane, and time-consumption for sample preparation. We propose that our improved method, presented here, be adopted as the Japanese official method for determination of C0 residues in crops. Received June 5, Accepted by JS September 19, Author to whom correspondence should be addressed. Materials and Methods Samples Radish, tomato, onion, sweet potato, kidney bean, carrot, cabbage, and lettuce were purchased from markets in Osaka, Japan. Reagents (a) Acetone, acetonitrile, n-hexane, methanol, and ethyl acetate. (Wako Pure Chemical Industries, Ltd., Osaka, Japan.) Pesticide residue analytical grade. (b) Sulfuric acid, m-chloroperoxybenzoic acid, sodium thiosulfate, diatomaceous earth Celite 545, sodium chloride, trichloroacetic acid, and sodium sulfate. Special grade (Wako Pure Chemical Industries). (c) Sodium sulfate. Activated at 120 C for 12 h. (d) Standard materials. C0 (10 µg/ml) in acetonitrile was obtained from Ehrnstorfer GmbH, Augsburg, Germany. C1 and C2 were obtained from Tomen, Ltd., Tokyo, Japan. The chemical structures of these compounds are shown in Figure 1. (e) Pesticide standard solution. Standard solutions of C1 and C2 (1000 µg/ml) were prepared by dissolving each pesticide in acetonitrile. For recovery experiments, the standard solution was diluted with acetonitrile (10 ppm). (f) Liquid chromatography (LC). Mobile phase A, 0.01% trichloroacetic acid in distilled water; mobile phase B, acetonitrile. (g) Liquid chromatography/mass spectrometry LC/MS. Mobile phase A, 5% acetic acid in distilled water; mobile phase B, acetonitrile. Apparatus (a) Cartridge columns. Bond Elut SAX, PRS, NH2, SCX, and PSA (Varian, Harbor City, CA). Sep-Pak Plus Alumina B and silica cartridge columns (Waters Corp., Milford, MA). (b) Electric homogenizer. Nihonseiki (Tokyo, Japan). (c) Rotary evaporator. Shibata Scientific Technology Ltd. (Tokyo, Japan). LC Analysis (a) Apparatus. A Hewlett Packard HP Series 1100 LC (Hewlett-Packard, Palo Alto, CA), equipped with a degasser G1322A, binary pump G1312A, thermostatted column compartment G1330A, autosampler G1329A, UV detector

2 ISHIMITSU ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 4, Figure 1. Chemical structures of clethodim and its oxidation metabolites. Figure 2. LC chromatogram of clethodim and its oxidation metabolites. Peaks: 1 = Clethodim sulfoxide (40 ng); 2 = clethodim sulfone (40 ng); 3 = clethodim (40 ng). G1314A, and ChemStation, was used for analysis of clethodim and its oxidized metabolites. The system was equipped with a stainless steel column (4.6 mm id 250 mm) packed with L-column ODS. (b) Operating parameters. The mobile phase flow rate was adjusted to 1.2 ml/min during analysis. The system was equilibrated at 30% mobile phase B in mobile phase A; a 15 min linear gradient to 100% mobile phase was begun and held for 5 min. When the gradient was completed, the mobile phase was returned to 70% A, 30% B and held for 8 min to re-equilibrate the column. The other conditions were as follows: temperature for column separation, 40 C, and wavelength of UV detection, 254 nm. LC/MS Analysis (a) Apparatus. A Shimadzu LCMS-QP8000 LC/MS (Shimadzu Corp., Kyoto, Japan), equipped with a degasser DGU-14AM, binary pump LC-10AD VP, thermostatted column compartment GTO-10AC VP, autosampler SIL-10AD VP, and UV detector SPD-10AV PG was used for analysis of C2. The system was equipped with stainless steel column (2.0 mm id 150 mm) packed with Wakosil II-3C18 HG. (b) Operating parameters. The mobile phase flow rate was adjusted to 0.2 ml/min during analysis. The system was equilibrated at 30% mobile phase B in mobile phase A; a 20 min linear gradient to 100% mobile phase was begun and held for 2 min. When the gradient was completed, the mobile phase was returned to 70% A, 30% B and held for 5 min to re-equilibrate the column. Other conditions were as follows: temperature for column separation, 50 C, and wavelength of UV detection, 254 nm. (c) MS conditions. Analytical mode, electrospray ionization (ESI; negative); drying gas (N 2 ); flow, 4.5 L/min; probe voltage, 4.5 kv. The selected ions for monitoring were m/z 358 (C0), 374 (C1), and 390 (C2). Extraction Ground control samples (10 g kidney bean and 20 g radish, tomato, onion, sweet potato, carrot, cabbage, and lettuce) were placed in a stainless steel cup to which 100 ml acetone was added. The mixture was homogenized for 3 min, and then filtered through filter paper with 7 g Celite 545 (10 mm thickness) into 300 ml round-bottom flask. The extract was rinsed and filtered with 50 ml acetone, and evaporated to a concentration of ca 30 ml at 40 C. The extract was transferred to a

3 1174 ISHIMITSU ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 4, 2001 Table 1. Recoveries of clethodim sulfone from clethodim by sulfide oxidation with m-chloroperoxybenzoic acid Recovery of clethodim sulfone, % Clethodim standard, ppm Standard Added to sample (Radish) 75.7 (Tomato) (Onion) (Sweet potato) 73.2 (Kidney bean) (Carrot) 79.7 (Cabbage) (Lettuce) 300 ml separatory funnel, and rinsed with 30 ml ethyl acetate. A 1 ml volume of 3M sulfuric acid and 50 mg m-chloroperoxybenzoic acid were added to the extract and warmed at 50 C for 1 min. A 30 ml volume of 10% sodium thiosulfate was added to stop the sulfide oxidation, and the sample was shaken vigorously for 5 min. The extract was transferred to a 300 ml separatory funnel, added to 70 ml water, 20 g NaCl, and 70 ml ethyl acetate, and shaken vigorously for 5 min. Another 100 ml ethyl acetate was added, and the solution was shaken again for 5 min. The organic layers were collected in a 300 ml Erlenmeyer flask, dehydrated with ca 20 g anhydrous Na 2 SO 4, and allowed to stand for 30 min. They were then filtered through filter paper to separate anhydrous Na 2 SO 4. The flask was then rinsed with an additional 20 ml ethyl acetate, and evaporated to dryness under vacuum at 40 C. Cleanup First procedure. The residue was dissolved in 5 ml 15% acetone in n-hexane and charged onto Bond Elut SAX plus PRS cartridge columns. The cartridge columns were rinsed with 20 ml 15% acetone in n-hexane, which was then discarded, followed by elution with 20 ml 50% acetone in n-hexane, and evaporated to dryness under vacuum at 40 C. The Bond Elut SAX plus PRS cartridge columns were conditioned with 10 ml n-hexane before use. Second procedure. The residue was dissolved in 10 ml ethyl acetate and charged onto a Sep-Pak Plus Alumina B cartridge column. The cartridge column was rinsed with 10 ml ethyl acetate, which was then discarded, followed by elution with 20 ml 5% methanol in ethyl acetate, evaporated to dryness under vacuum at 40 C, and then filled to 2 ml with acetonitrile. The Sep-Pak Plus Alumina B cartridge column was conditioned with 10 ml ethyl acetate before use. Quantification The sample solution was automatically injected into the LC system for residue analysis. The concentration of C2 was calculated based on a peak area calibration curve. The calibration curve was constructed with C2 generated from C0 by sulfide oxidation with m-chloroperoxybenzoic acid. Each injection was performed 3 times to test reproducibility. For routine work, 20 ng C2 standard was injected into the LC system twice per day. Recovery Test Chopped samples were fortified with ppm C0. Recovery data represent 3 replications. Table 2. Elution patterns of clethodim sulfone from several cartridge columns used for first or second cleanup Recovery, % b First cleanup Second cleanup Eluting solvent a NH2 PSA SAX SCX PRS Silica Alumina B a Successively eluted with 20 ml of each eluting solvent: 1, n-hexane; 2, acetone in n-hexane (3 + 17); 3, acetone in n-hexane (1 + 1); 4, acetone; 5, ethyl acetate; and 6, methanol in ethyl acetate ( ). b Clethodim sulfone (20 ng) was added.

4 ISHIMITSU ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 4, Figure 3. Typical LC chromatograms of crop extracts. Radish, tomato, onion, sweet potato, kidney bean, carrot, cabbage, and lettuce were fortified with clethodim at 1.0, 1.0, 0.5, 0.2, 0.2, 0.1, 0.1, and 0.05 ppm, respectively. Arrows indicate peaks of clethodim sulfone.

5 1176 ISHIMITSU ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 4, 2001 Figure 4. Selected ion monitoring (SIM) chromatograms (m/z = 390) of fortified sample by LC/MS (ESI). Onion, sweet potato, kidney bean, and carrot were fortified with clethodim at 0.5, 0.2, 0.2, and 0.1 ppm, respectively. Arrows indicate peaks of clethodim sulfone. Results and Discussion LC Conditions, Linearity, and Limit of Detection Optimal conditions for determinations of C0, C1, and C2 were investigated. An LC chromatogram of the standard is shown in Figure 2. Retention times of C0, C1, and C2 are approximately 15, 8.5, and 10 min, respectively. In the present study, trichloroacetic acid was used as an ion-pair reagent; 1-heptanesulfonic acid could not be used because C0, C1, and C2 were not dissolved from the column. The linear dynamic range of the detector response for C2 was examined and appeared to be from 5 to 100 ng injected on-column. The detection limits of the analytes in crops were 0.01 ppm for C2 (S/N > 3).

6 ISHIMITSU ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 4, Table 3. Recoveries of clethodim added to agricultural products Sample Fortification level, ppm Recovery, % a Radish ± 5.0 Tomato ± 4.8 Onion ± 5.8 Sweet potato ± 6.8 Kidney bean ± 2.6 Carrot ± 6.6 Cabbage ± 6.1 Lettuce ± 7.6 a Average ± standard deviation of 3 determinations. Clethodim was recovered as clethodim sulfone. Extraction C0 in crops was extracted with acetone. The volume of acetone in the extraction step was adjusted to 150 ml to obtain good recovery reproducibility, particularly for C0. The concentrate was filtered under vacuum following the addition of 7 g Celite 545. The Japanese Environment Agency method used toxic dichloromethane as an organic solvent. Therefore, we tried to use ethyl acetate instead of dichloromethane. Extraction twice with ethyl acetate gave better results, and the addition of NaCl before the extraction process was also effective. An attempt to use ethyl acetate instead of dichloromethane was successful because of the good recovery of C2. The Japanese Environment Agency method also used a Celite 545 column for extraction; however, the column was not necessary for the quantitation of C2. Although this did not affect recovery, it did shorten the procedure. Sulfide Oxidation m-chloroperoxybenzoic acid was used to sulfonate sethoxydim in the Japanese official method. The reaction from sethoxydim to sethoxydim sulfone by sulfide oxidation with m-chloroperoxybenzoic acid was reported to be about 60% (7). Therefore, we examined the percent reaction from C0 to C2 by sulfide oxidation with m-chloroperoxybenzoic acid. The result is shown in Table 1. The calculated recoveries of C2 generated from C0 and/or crops added to C0, based on the corresponding C2 standard, were approximately 74 and 76%, respectively. The same result was obtained from C1 to C2 by sulfide oxidation with m-chloroperoxybenzoic acid. Consequently, a calibration curve should be constructed by sulfide oxidation of C0 the same as for the sample preparation. These results indicated that a C0 standard is necessary for the method. Cleanup The quantitation of C2 could not be conducted without cleanup of the sulfonated solution. The Japanese Environment Agency method used silica, alumina, and NH2 (for bulbs) cartridge columns for sample purification; however, the Japanese official method for MHW cannot use a silica cartridge column. Therefore, we evaluated the use of ion exchange cartridge columns such as Bond Elut SAX, PRS, NH2, SCX, and PSA cartridge columns for sample purification. The results are shown in Table 2. Bond Elut SAX and PRS cartridge columns gave the best recovery. However, cleanup was not satisfactory with these 2 columns; therefore, a second cleanup with another column was necessary. Sep-Pak Plus Alumina B and silica cartridge columns were tested for the second cleanup. With the Sep-Pak silica cartridge column, C2 from crops could not be measured because many interfering peaks appeared on the chromatogram. On the other hand, when the Sep-Pak Plus Alumina B cartridge column was used, the LC chromatograms had fewer interfering peaks. Typical chromatograms of the fortified test solutions are shown in Figure 3. The LC chromatogram of the sample solution of cabbage shows an interfering peak close to the retention time of C2 after cleanup with 3 cartridge columns. The detected peaks in the samples were confirmed by LC/MS (ESI). The selected ion for monitoring was m/z 390 (C2). Figure 4 shows fortified peaks of C2 in onion, sweet potato, kidney bean, and carrot. The peak of C2 in unfortified crops was not detected. The peak area and the added concentration were not proportional. The results indicated that LC/MS cannot be used for quantitative analysis, and the use of matrix-matched standards must be examined. Recovery Test Recoveries of clethodim in 8 crops fortified at ppm are shown in Table 3. Values are the means of triplicate determinations. In Japan, the common acceptable range of recovery is %. The recoveries of C2 exceeded 90% in all cases; the recovery for C2 at low fortification was higher, possibly because of the interfering peaks on the chromatogram, and was within the acceptable range. The coefficient of variation of the recovery for C2 was within 10%. Conclusions Clethodim and oxidation metabolites in crops were extracted with acetone, sulfonated by m-chloroperoxybenzoic acid, cleaned up with Bond Elut SAX plus PRS cartridge columns and a Sep-Pak Plus Alumina B cartridge column, and detected by LC with UV detector. The method is rapid and does not use toxic dichloromethane. Recoveries from radish, tomato, onion, sweet potato, kidney bean, carrot, cabbage, and lettuce were % by the proposed method. The fortified peaks in samples were confirmed by LC/MS (ESI). We propose that this improved method be adopted as the Japanese official method for analysis of clethodim and its oxidation metabolites in crops.

7 1178 ISHIMITSU ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 84, NO. 4, 2001 References (1) Rendina, A.R., & Felts, J.M. (1988) Plant Physiol. 86, (2) Chevron Chemical Co. (1986) Ortho SELECT Herbicide, Technical Information Bulletin, New York, NY (3) Ishikawa, I., Yamada, S., Hosaka, H., Kawana, T., Okunuki, S., & Kohara, K. (1985) J. Pestic Sci. 10, (4) Iwataki, I., & Hirono, S. (1979) Advances in Pesticide Science, Fourth International Congress on Pesticide Chemicals, Pergamon Press, Oxford, UK, pp (5) Falb, L.N., Bridges, D.C., & Smith, A.E. (1991) J. Assoc. Off. Anal. Chem. 74, (6) Notice No. 311 of Environment Agency, Official Gazette, August 24, 1999, extra No. 163 (7) Goto, M., & Kato, M. (1987) Analytical Methods of Pesticide Residues: Addendum, Softsciences, Toyko, Japan, pp