Determination of Pyrethrin and Pyrethroid Pesticides in Urine and Water Matrixes by Liquid Chromatography with Diode Array Detection

Transcription

1 1236 LOPER & ANDERSON: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 6, 2003 TECHNICAL COMMUNICATIONS Determination of Pyrethrin and Pyrethroid Pesticides in Urine and Water Matrixes by Liquid Chromatography with Diode Array Detection BOBBY L. LOPER and KIM A. ANDERSON 1 Oregon State University, Department of Environmental and Molecular Toxicology, Food Safety and Environmental Stewardship Program, 1007 Agriculture and Life Sciences Building, Corvallis, OR The following pyrethrin and pyrethroid pesticides were determined in urine and water matrixes by liquid chromatography with diode array detection (LC DAD): pyrethrin I, pyrethrin II, tetramethrin, baythroid, bifenthrin, fenvalerate, phenothrin, allethrin, resmethrin, cis-permethrin, and trans-permethrin. In addition, 3-phenoxybenzyl alcohol, a metabolite of various pyrethroids, was also successfully determined by the analytical method. The matrix extraction was simple, inexpensive, and fast, using only sodium chloride and acetonitrile. The acetonitrile extract was filtered and analyzed by LC DAD. The method detection limits for the pyrethrin pesticides in 5 ml urine were determined to range from to 0.04 g/ml, depending on the individual pyrethrin. Recoveries from spiked tap water ranged from 77 to 96%; recoveries from urine ranged from 80 to 117%. This method is especially well-suited to clinical investigations, in which rapid analysis of forensic samples is often required. Received June 19, Accepted by JS July 1, Author to whom correspondence should be addressed; kim.anderson@orst.edu. Natural pyrethrins were among the earliest pesticides discovered by agriculturists. It is believed that they were extracted from dried pyrethrin and chrysanthemum flowers (Chrysanthemum cinerariaefolium, C. coccineum) as early as the first century by the Chinese (1). Compared with other pesticides, their relative lack of terrestrial environmental persistence and their slow development of pest resistance have ensured their popularity for both agricultural and public health applications. Pyrethrins are among the most widely used classes of pesticides in the United States, and they are also one of the most used classes of pesticides globally (2). Therefore, the determination of pyrethrin pesticides in forensic matrixes is increasingly of great interest. In this paper, the term pyrethrins refers to the natural insecticides derived from chrysanthemum flowers; pyrethroids are the synthetic chemicals, and pyrethrin is a general name covering both. Two pyrethrins, pyrethrin I and pyrethrin II, are the most prominent. When demand outpaced the supply of pyrethrins, synthetic pyrethroids were developed. Pyrethroids first became commercially available in the 1980s. Some of the most common pyrethroids are bifenthrin, tetramethrin, baythroid, fenvalerate, phenothrin, allethrin, resmethrin, and cis-andtrans-permethrin. Because of their typical uses (animal [pet] flea/pest control, household insect control, human lice shampoo, broad agricultural use), pyrethrins are fast becoming a concern in human health investigations. In the state of Oregon in 2001, pyrethrin was cited in more confirmed pesticide poisoning reports than did any other pesticide, accounting for over one-third of all pesticide poisonings (3). In addition to direct poisonings, there were indications that some people have (near) lethal sensitivities to pyrethrin compounds or coextractants (e.g., sesquiterpene lactones; Sheldon L. Wagner, Oregon State University, personal communication, 2001). Few analytical choices for forensic applications are available that are quick and accurate, do not require specialized instruments, and have sufficient specificity to confirm a human/animal pesticide poisoning (4). Methods are needed that can quickly determine a wide range of pyrethrins and pyrethroids in forensic matrixes. These methods must accommodate small-to-modest samples volumes with adequate sensitivity. On the basis of animal studies, any amount of pyrethrin absorbed would be expected to be rapidly excreted. Pyrethrins I and II were excreted unchanged in feces (5). When oral doses of radiolabeled resmethrin (pyretheroid) at 10 mg/kg were given to laying hens, 90% of the dose was eliminated in urine and feces within 24 h (6). Researchers gave male and female rats a single dose of cyhalothrin (pyrethroid) and determined that the rats absorbed about 55%; during the following 7 days the rats excreted the dose in urine (20 40%) and feces (6, 7). Similar results were reported for other studies in which fluvalinate (8) and cyfluthrin (9) pyrethroids; were excreted via urine within days after exposure. Therefore, urine is a useful matrix to use for veterinary forensic investigations of pyrethrin exposure. In humans, pyrethroids are metabolized in the liver, and the metabolites are renally eliminated; the elimination half-life for the metabolites has been reported as about 6 10 h (2, 10). 3-Phenoxybenzyl alcohol, which is derived from the metabo-

2 LOPER & ANDERSON: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 6, Table 1. Recovery of pyrethrins from fortified tap water samples a Mean recovery, % (RSD, %) b Fortification level, g/ml Pyrethrin II Pyrethrin I Tetramethrin Baythroid Fenvalerate Phenothrin Allethrin Resmethrin cis-permethrin trans-permethrin (9) 92 (10) 82 (8) 81 (10) 84 (11) 77 (9) 96 (7) 103 (5) 93 (12) 86 (8) (7) 86 (7) 92 (3) 92 (3) 89 (3) 85 (8) 89 (3) 87 (7) 89 (3) 85 (4) DAD used = 235 nm, except for tetramethrin, baythroid, and resmethrin (for which DAD = 245 nm). n = 12. a b lism of various pyrethroids, is also an important diagnostic analyte for exposure (11, 12). However, pyrethroids may be more effectively metabolized by adults than by infants; therefore, diagnostic analyses should include both pyrethrin and a metabolite such as 3-phenoxybenzyl alcohol. The amount of pyrethrin in the urine depends on many variables, including age, dose, and exposure. Reported concentrations of pyrethrin and pyrethroid metabolites in urine have ranged from about 5 g/l to 30 mg/l (2, 13, 14). Many analytical methods do not include both pyrethrin and pyrethroid pesticides. Many pyrethrin methods were developed (and optimized) for the determination of the active ingredient (pyrethrins) in various formulations of the commercial product. Therefore, there were no issues of limited sample size or cleanup or requirements for good sensitivity or quick turn-around time. Because many methods were for the formulated product (typically 20 50% pyrethrin), sample cleanup was not a major obstacle, and samples were diluted (by a factor of 10 to >100) significantly before analysis. Methods using gas chromatography (GC; 15) or liquid chromatography (LC; 16, 17) have been reported that performed very well with formulated commercial products. Pyrethrin and pyrethroid residues in crops and foods have generated significant analytical development, which was recently reviewed by Chen and Wang (18). Many methods often focus on only one or 2 pyrethrins, and numerous approaches are available, although many are labor-intensive, typically requiring a liquid extraction step and a liquid partition step followed by solid-phase cleanup. Sensitivities are typically low g/ml (ppm) levels; however, lower sensitivities can be obtained with the addition of a derivatization step before GC analysis. Detection limits of g/ml have been reported (18). Good analytical methods for pyretheroid metabolites have been reported (19), but they tend to be more complicated (hydrolysis, solid-phase cleanup, and derivatization) and somewhat expensive, i.e., GC with high-resolution mass spectrometry (MS) with negative chemical ionization (20) and GC with tandem MS (21), and they focus solely on pyrethroid metabolites (2, 12). However, because some of the most severe poisonings have been reported in infants (22), who may not be able to effectively metabolize pyrethrum (23), it may be important to include the pesticide in any investigation. Overall, it is important that methods are developed for both the primary pesticide and a metabolite in an analytical approach that can provide quick turn-around time. This paper presents a method for the determination of pyrethrin pesticides (pyrethrin I, pyrethrin II, tetramethrin, baythroid, fenvalerate, phenothrin, allethrin, resmethrin, and trans-permethrin) and one pyrethroid metabolite (3-phenoxybenzyl alcohol). The method cleanup adheres to the principles of green-chemistry, avoiding the use of chlorinated or carcinogenic solvents and minimizing the use of hazardous reagent. The only reagents required for extraction are sodium chloride and acetonitrile. The method also minimizes the waste stream. Sample size requirements are minimal; only 1 15 ml urine is used, which is often a requirement in forensic investigations.

3 1238 LOPER & ANDERSON: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 6, 2003 Table 2. Recovery of pyrethrin from fortified urine samples Mean recovery, % (RSD, %) b Fortification level, g/ml Pyrethrin II Pyrethrin I Tetramethrin Baythroid Fenvalerate Phenothrin Allethrin Resmethrin cis-permethrin trans-permethrin (24) 93 (9) 86 (6) 86 (8) 89 (11) 81 (6) 88 (17) 95 (27) 91 (22) 81 (16) (6) 80 (5) 92 (3) 92 (3) 87 (4) 84 (8) 89 (2) 87 (6) 89 (4) 85 (4) Analyzed over 6 months; primary DAD used = 235 nm, except for tetramethrin, baythroid, and resmethrin (for which DAD = 245 nm). b n = 12. a Experimental Reagents (a) Acetonitrile. Optimal grade (Fisher Scientific, Pittsburgh, PA). (b) Sodium chloride saturated solution. Approximately 14 g NaCl (ACS reagent grade; Fisher Scientific) was added to a 100 ml volumetric flask and diluted to volume with type 1 water. (c) Pyrethrin standards. Pyrethrins and pyrethroids (Chem Service, West Chester, PA). Prepare standard solutions of pyrethrin (29.7% pyrethrin I and 22.2% pyrethrin II), baythroid, phenothrin, and tetramethrin, each at 400 g/ml; fenvalerate at 200 g/ml; and allethrin, resmethrin, cis-permethrin, and trans-permethrin, each at 400 g/ml. The Chemical Abstract Service Registry Numbers are as follows: pyrethrin I ( ); allethrin ( ); baythroid (synonym cyfluthrin; ( ); fenvalerate ( ); permethrin ( ); phenothrin (synonym sumithrin; ); resmethrin ( ); tetramethrin ( ); bifenthrin ( ); and 3-phenoxybenzyl alcohol ( ). Serial dilutions were prepared as appropriate from 0.1 to 50 g/ml, and all standards were prepared in acetonitrile. Primary standards were kept at 4 C and used before the manufacturer s expiration date; secondary standards were prepared fresh biweekly. (d) Urine samples. Samples were kept frozen until analysis. All clinical samples were provided by S.L. Wagner (National Pesticide Monitoring Program, Oregon State University). Apparatus LC system. Hewlett-Packard Model 1100 liquid chromatograph with diode array detector. Column: Phenomenex, LUNA C18(2), 3 m, mm; temperature, 30 C. Mobile phase: (A) type 1 water (18 M cm) and (B) acetonitrile. Flow rate: 0.5 ml/min. Gradient: start, 90% A and 10% B; from 1 to 35 min, gradient to 90% B; 90% B from 31 to 35 min; from 35 to 36 min, gradient to 100% B; 100% B until 46 min; from 46 to 47 min, gradient to 10% B. Injection volume: 40 L. Detector: 190 to 600 nm; signal A: 235 nm, signal B: 245 nm. Sample Preparation and Analysis A 1 15 ml sample (typically 5 ml) of urine or water was transferred to a 25 ml volumetric flask. Approximately 4 g NaCl, 3.5 ml acetonitrile, and 5 ml saturated NaCl solution were added. The flask was shaken for 1 min and allowed to stand until the acetonitrile layer separated. The acetonitrile layer was transferred to a 5 ml disposable centrifuge tube and centrifuged. The aqueous portion was removed and extracted again with 1 ml acetonitrile, which was added to the 25 ml volumetric flask (shake, stand, separate). The acetonitrile extracts were combined and adjusted to a known volume (typically 1.0 or 0.5 ml); the total was mixed and filtered (0.45 m) into an LC vial for analysis by LC with diode array detection (DAD). The unused portion can be stored in the dark at 4 C for future analysis.

4 LOPER & ANDERSON: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 6, Figure 1. Chromatograms of control urine and urine fortified with pyrenthrin I and pyrenthrin II (0.1 g/ml, final concentration). Confirmation by LC/MS LC/MS with electrospray ionization (ESI) in the positive ion mode was used to confirm the identity of some analytes (Dionex 600 liquid chromatograph and ThermoFinnigan quadruple mass spectrometer). The LC parameters were the same as those described above. The optimum ESI conditions included the following: probe temperature, 450 C; needle voltage, 1.5 kv; and source voltage, V; other operational parameters were based on the recommended tuning procedures used daily. Total ion chromatograms were typically collected from 100 to 500 m/z. At a source voltage (cone voltage of 10 V), the parent ion (M-H + ) was the dominant ion of the mass spectrum and was used for confirmation. Results and Discussion Calibration curves prepared from 3 4 standard solutions during the analysis of 20 analytical batches gave correlation coefficients (r) of The linearity of the method was >2 orders of magnitude, which minimizes the analytical difficulties encountered with other methods having a relatively narrow concentration range. Tables 1 and 2 show the recoveries of pyrethrin I, pyrethrin II, tetramethrin, baythroid, fenvalerate, phenothrin, allethrin, resmethrin, cis-permethrin, and trans-permethrin from fortified tap water and urine, respectively. The average recoveries from tap water ranged from 77 to 103% with relative standard deviations (RSDs) of 3 12% (n = 12). All recovery studies were performed over the course of 6 months. Urine samples were fortified with pyrethrin in acetonitrile at levels of g/ml (final concentrations) before extraction. The average recoveries from urine ranged from 81 to 117% with RSDs of 2 27%. During the studies, one analytical batch gave lower recoveries of approximately 55 70% for pyrethrin II, allethrin, resmethrin, and permethrin. If it were not for this single analytical run, the RSDs for all analytes in urine would have been 11% for all analytes tested. Overall, this is good precision for real-life matrixes. The wavelengths used were selected to provide the largest response with the fewest background interferences. All analyte data were collected at both 235 and 245 nm; the wavelengths used for calculations in the recovery studies are listed in Tables 1 and 2. In general, there was very little difference (<15%) in the accuracy or precision of the recoveries calculated for the 2 wavelengths. The signal at 235 nm was more responsive for most analytes. The ratio of the signal at 235 nm to the signal at 245 nm was approximately 1.5. The detection limits of the method were determined by the analysis of extracted reagent blank samples (n =5),basedon 3 times the signal-to-baseline noise ratio. The concentrations were then subsequently verified by analysis of urine samples. Based on 5 ml urine samples and a final volume of 0.5 ml, the method detection limits (MDLs) were g/ml for tetramethrin, allethrin, baythroid, resmethrin, cis-/trans-permethrin, and phenothrin; g/ml for pyrethrum I; 0.04 g/ml for pyrethrin II; and g/ml for fenvalerate. The higher MDL for pyrethrin II is due to a peak in the chromatogram of the extract (at about 34 min) very near the retention time of pyrethrin II; the full DAD signal ( nm) for the interference was distinguishable from that of pyrethrin II, but it did elevate/vary the baseline and caused an increase in the MDL (Figure 1). Chromatographic conditions could be improved by using the blank subtraction feature available on the HP ChemStation software. We used this feature for sensitive quantitation; however, Figure 1 shows the complete chromatogram without blank subtraction. The retention order was m-phenoxybenzyl alcohol (21 min), pyrethrin I (29.6 min), tetramethrin (31.4 min), allethrin (31.8 min), pyrethrin II (33.7 min), baythroid (34.3 min), resmethrin (35.2 min), fenvalerate (35.3 min), cis-permethrin (19; 35.7 min), trans-permethrin (36.3 min), phenothrin (36.4 min), and bifenthrin (37 min). All analytes

5 1240 LOPER & ANDERSON: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 6, 2003 were resolved with only 2 exceptions, pyrethrin II and baythroid, as discussed above. Sample preparation was minimal, requiring only simple liquid liquid extraction and filtration; all manipulations were performed within a few minutes. Long-term performance under routine analytical conditions was tested in the analysis of >100 check standards, spiking solutions, blanks, control urine samples, fortified urine samples, and reagent blanks. The results were typical of the quality of the data reported in Tables 1 and 2. Conclusions This method demonstrated good accuracy and precision and good recoveries of pyrethrin pesticides and a pyrethroid metabolite in urine and water matrixes. The method uses standard LC DAD equipment. Some increases in sensitivity could possibly be obtained by reducing the acetonitrile volume after extraction in order to gain better concentration factors. Acknowledgments We thank Sheldon L. Wagner (Oregon State University) for helpful discussions and for providing clinical samples, and Jennifer Schaeffer, Eugene Johnson, and Wendy Hillwalker (Department of Environmental and Molecular Toxicology, Food Safety and Environmental Stewardship Program, Oregon State University) for helping to analyze some of the samples. References (1) Caswerett and Doull s Toxicology (1996) C.D. Klaassen (Ed.), McGraw-Hill, New York, NY (2) Heudorf, U., & Angerer, J. (2001) Environ. Health Perspect. 109, (3) Cole, M. (2002) in The Oregonian, May 7, pp B1, B4 (4) O Malley, M. (1997) Lancet 9059, (5) Elliot, M., Janes, N.F., Kimmel, E.C., & Casida, J.E. (1972) J. Agri. Food Chem. 20, (6) Christopher, R.J., Ivie, G.W., Beier, R.C., & Jenkins, W.L. (1989) J. Agri Food Chem. 37, (7) World Health Organization (1990) Cyhalothrin, Environmental Health Critera 99, Geneva, Switzerland (8) Kidd, H., & James, D.R. (Eds) (1991) The Agrochemicals Handbook, 3rd Ed., Royal Society of Chemistry Information Services, Cambridge, UK (as updated), pp 2 13 (9) Pesticide Residues in Food (1987) Evaluations Part II Toxicolog; Food and Agriculture Organization of the United Nations, Rome, Italy (10) Ray, D.E., & Forshaw, P.J. (2000) Clin. Toxicol. 38, (11) Hardt, J. (2001) Fresenius Z. Anal. Chem. 371, (12) Abu-Qwere, A.W., & Abou-Conia, M.B. (2001) J. Anal. Toxicol. 25, (13) Leng, G., Leng, A., Kuhn, K.H., Lewalter, J., & Pauluhn, J. (1997) Xenobiotica 27, (14) Arrebola, F.J., Martinex-Vidal, J.L., & Fernandex-Gutierres, A. (2000) Anal. Chim. Acta 401, (15) Nguven, K.T., Moorman, R., & Kuvkendall, V.J. (1998) J. AOAC Int. 81, (16) Kasaji, D., Rieder, A., Krenn, L., & Kopp, B. (1999) Chromatographia 50, (17) Wang, I., Subramanian, V., Moorman, R., Burleson, J., & Ko, J. (1997) J. Chromatogr. A 766, (18) Chen, Z.M., & Wang, Y.H. (1996) J. Chromatogr. A 754, (19) Colume, A., Cardenas, S., & Gallego, M. (2001) Rapid Commun. Mass Spectrom. 15, (20) Leng, G., Kuhn, K.H., & Dunemann, L. (1995) Int. J. Hyg. Environ. Med. 198, (21) Arrebola, F.J., Martinez-Vidal, J.L., Fernandez-Gutierrez, A. (2000) Anal. Chim. Acta 401, (22) Wagner, S.L. (2000) West.J.Med. 173,86 (23) Pesticide Information Profile, EXTONET (2002)