EPA-800/1-77-829 May 1977 Environmental Healtii Effects Research Series EXTENSION OF MULTI-RESIDUE METHODOLOGY: I. Determining Multiclass Pesticide Residues in Soil by Gas Chromatography II. Dynamic Fluorogenic Labelling Detector for Carbamates Health Effects Research Laboratory Office of Research and Development U.S. Environmental Protection Agency Research Triangle Park, North Carolina 27711 ------- RESEARCH REPORTING SERIES Research reports of the Office.of Research and Development, U.S. Environmental Protection Agency, have been grouped into nine series. These nine broad cate- gories were established to facilitate further development and application of en- vironmental technology. Elimination of traditional grouping was consciously planned to foster technology transfer and a maximum interface in related fields. The nine series are: 1. Environmental Health Effects Research 2. Environmental Protection Technology 3. Ecological Research 4. Environmental Monitoring 5. Socioeconomic Environmental Studies 6. Scientific and Technical Assessment Reports (STAR) 7. Interagency Energy-Environment Research and Development 8. "Special" Reports 9. Miscellaneous Reports This report has been assigned to the ENVIRONMENTAL HEALTH EFFECTS RE- SEARCH series. This series describes projects and studies relating to the toler- ances of man for unhealthful substances or conditions. This work is generally assessed from a medical viewpoint, including physiological or psychological studies. In addition to toxicology and other medical specialities, study areas in- clude biomedical instrumentation and health research techniques utilizing ani- mals — but always with intended application to human health measures. This document is available to the public through the National Technical Informa- tion Service, Springfield, Virginia 22161. ------- EPA-600/1-77-029 May 1977 EXTENSION OF MULTI-RESIDUE METHODOLOGY I. DETERMINING MULTICLASS PESTICIDE RESIDUES IN SOIL BY GAS CHROMATOGRAPHY II. DYNAMIC FLUOROGENIC LABELLING DETECTOR FOR CARBAMATES by H. Anson Moye, Sujit Witkonton and Gordon Cash University of Florida, IFAS. Food Science Department Pesticide Research Laboratory, Gainesville, FL 32611 Contract No. 68-02-1706 Project Officer Robert F. Moseman Environmental Toxicology Division Health Effects Research Laboratory Research Triangle Park, North Carolina 27711 Health Effects Research Laboratory Office of Research and Development U.S. Environmental Protection Agency Research Triangle Park, N.C. 27711 ------- DISCLAIMER This report has been reviewed by the Health Effects Research Laboratory, U.S. Environmental Protection Agency, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the U.S. Environmental Protection Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. 11 ------- FOREWORD The many benefits of our modern, developing, industrial society are accompanied by certain hazards. Careful assess- ment of the relative risk of existing and new man-made environmental hazards is necessary for the establishment of sound regulatory policy. These regulations serve to enhance ttye quality of our environment in order to promote the public health and welfare and the productive capacity of our Nation's population. The Health Effects Research Laboratory, Research Triangle Park conducts a coordinated environmental health research program in toxicology, epidemiology, and clinical studies using human volunteer subjects. These studies address problems in air pollution, non-ionizing radiation, environ- mental carcinogenesis and the toxicology of pesticides as well as other chemical pollutants. The Laboratory develops and revises air quality criteria documents on pollutants for which national ambient air quality standards exist or are proposed, provides the data for registration of new pesticides or proposed suspension of those already in use, conducts research on hazardous and toxic materials, and is preparing the health basis for non-ionizing radiation standards. Direct support to the regulatory function of the Agency is provided in the form of expert testimony and preparation of affidavits as well as expert advice to the Administrator to assure the adequacy of health care and surveillance of persons having suffered imminent and substantial endanger- ment of their health. This report represents a research effort to extend and improve analytical methodology for the determination of a variety of pesticide residues in our environment. The emphasis is on less persistent chemicals which are being used in place of the more environmentally stable pesticides. John H. Knelson, M.D. Director, Health Effects Research Laboratory 111 ------- ABSTRACT Cf all the forty pesticides under investigation, twelve (methyl parathion, parathion, malathion, phorate, azinphos methyl, p azinphos ethvl, Dursban , diazinon, dimethoate, phosphamidon, p Azodrin and demeton) were detected by the FPD-P detector, two p (Temik and methomyl) by the FPD-S detector, fourteen (captan, p chlorobenzilate, endosulfan, Avadex , folpet, methoxychlor, PCNB, p p Perthane , trifluralin, atrazine, bromacil, CIPC, Difolatan and simazine) by EC detector. Twelve have to be derivatized by PFPA R R (bux, carbofuran, monuron, Zectran , Landrins and IPC), PFBC (benomyl, monuron and amitrole) or DAM (DNOC, DNBP, and dicamba) and then detected by the EC detector. The 4% SE-30/6% OV-210 column exhibited superior separation efficiency for most pest- icides than other tested columns. The usefulness of the nitrogen- specific Hall detector was limited by its erratic performance. It was caused by the repetitive injection of unvented solvent and subsequent acid contamination in the detection system. The silica gel column was proven to be consistent and reliable for soil sample cleanup and separation of multi-class pesticides. Some impurities from blank silica gel columns were present in various eluted fractions; however, this drawback was offset by their effectiveness in cleanup and separation of var- ious pesticides from soil co-extractants. Good recovery results were obtained by the Soxhlet pro- cedure for most of the. organophosphate and halogenated pesticides, IV ------- except for demeton, bromacil, endosulfan and trifluralin. Also, captan, folpet and Difolatan reacted with methanol during Sox- hlet extraction; however, they were subsequently extracted and p recovered by ambient tumbling or Polytron ultrasonic extraction, Most of the heat-labile ^-containing pesticides were extracted by the Polytron and recovered in good yield except amitrole, Temik and methomyl. Figure 29 summarized the pro- cedure for detection of forty multi-class pesticides in soils. A dynamic fluorogenic labelling detector was designed and characterized for the high-pressure liquid chromatographic analysis of six N-methylcarbamate and two carbamoyl oxime pest- icides. Lannate, Matacil, Temik, Baygon, carbofuran, Sevin and Mesurol could be extracted from sandy soil at the 0.01 ppm level with recoveries ranging from 83 to 115%. Somewhat lower recoveries were experienced for sandy loam and silty loam soils. Zectran could not be reproducibly chromatographed under the liquid chromatographic conditions chosen for the separations. No cleanup was required prior to the liquid chromatography of the soil extracts; no significant interferences were observed for the unclean extracts. This report was submitted in fulfillment of Contract No. 68-03-1706 by Pesticide Research Laboratory, Food Science Department, IFAS, University of Florida under the sponsorship of the U.S. Environment Protection Agency. This report covers the period June 14, 1974, to June 26, 1976, and work was com- pleted as of June 26, 1976. ------- CONTENTS Foreword iii Abstract iv Figures viii Tables xiii Extension of Multi-residue Methodology. I. Determining Multiclass Pesticide Residues in Soil Introduction 1 Experimental 9 Apparatus 9 Reagents 11 Pesticide Standard Solutions 13 Pesticide Derivatization 13 Preparation of Spiked Soil Samples and Extraction Procedure 16 Polytron Ultrasonic Extraction Procedure 16 Silica Gel Column Chromatography 17 Results and Discussion 18 Screening of Pesticides for GLC responses .... 18 Derivatization of Pesticides 48 GLC Detection of S-containing Pesticides 70 Loss of Pesticides in the Concentration Steps . . 71 Silica gel Column Chromatography 74 Co-extractive Interferences from Crude Soil Extracts and Cleanup by Silica Gel Column Chromatography 79 Recoveries of Pesticides from Soils by Soxhlet Extraction 89 Recoveries of Some Organophosphate and Thermally Labile N-containing Pesticides from soil by Polytron Ultrasonic Extraction and Comparison of Results with Soxhlet Ex- traction 103 Stimultaneous Analysis of Forty Multi-class Pesticides in Sandy Soil 119 References 130 vx ------- Extension of Multi-residue Methodology. II. Dynamic Fluorogenic Labelling Detector for Carbamates Introduction 132 Experimental 134 Apparatus and Reagents 134 HPLC Column Study 136 Studies of Fluram 140 Studies on o-Phthalaldehyde 142 Conclusions 162 References 167 vii ------- FIGURES Extension of Multi-residue Methodology. I. Number Page la Chromatograms of 14 EC-sensitive pesticides on 4% SE 30/6% OV-210 column 24 Ib Chromatograms of 14 EC-sensitive pesticides on 1.51 OV-17/1.95% OV-210 column 25 Ic Chromatograms of 14 EC-sensitive pesticides on 5% OV-210 column 26 2 Analytical curve of Avadex 27 3 Analytical curve of captan, folpet and Difolatan . 28 4 Analytical curves of Perthane, chlorobenzilate and CIPC 29 5 Analytical curves of trifluralin and bromacil ... 30 6 Analytical curves of simazine and atrazine .... 31 7 Analytical curves of PCNB, endosulfan and methoxychlor 32 8 Analytical curves of Azodrin 41 * 9 Analytical curves of Dursban, dimethoate and phosphamidon 42. 10 Analytical curve of diazinon 43 11 Analytical curves of phorate and dimeton 44 12 Analytical curves of malathion and parathion ... 45 13 Analytical curves of azinphos ethyl, azinphos methyl and methyl parathion 46 14 Analytical curve of PFPA-IPC 54 15 Analytical curves of carbofuran, bux, Zectran, Landrins (PFPA) 55 16 Analytical curve of PFBC-amitrole 63 viii ------- FIGURES Number 17 Analytical curve of PFBC-benomyl 64 18 Analytical curve of PFBC-monuron 65 19 Analytical curves of dicamba, DNOC and DNBP (DAM) 66 20 Analytical curves of Temik and methomyl (4% SE 30/6% OV-210) 72 21 Analytical curves of Temik and methomyl (5% Carbowax 20 M) 73 22 Chromatogram of standard PFPA-derivatives of six carbamate pesticides (a), and after they were heated at 100°C in a sealed test tube for 6 hr in 1:1 benzene/MeOH (b) 78 23a Outlines of halogenated and organophosphate pest- icide standards (1 mg each) eluted from silica gel column 80 23b Outline of multi-class pesticide standard (1 mg) eluted from silica gel column '. 81 24a Gas chromatograms of crude control sandy soil extract (10 mg soil equivalent) 83 24b Gas chromatograms of control and 0.01 ppm PCNB from sandy soil (10 mg soil equivalent) 84 24c Gas chromatograms of 0.01 ppm in sandy soil of Avadex (3), trifluralin (39), endosulfan (21), methoxychlor (79), and Perthane~X35) in 60% benzene/methanol frac- tion and the respective control (50 mg soil equiv- alent) 85 24d Gas chromatograms of 0.01 ppm in sandy soil of CIPC (12), captan (10), folpet (22) , chlorobenzilate (11) and" Difolatan ~(T6) and the respective control (10 mg soil equivalent^" 86 24e Gas chromatograms of 0.01 ppm in sandy soil of atra- zine (2) and simazine (37) and the respective control (50 mg soil equivalent). 87 IX ------- FIGURES Number Page 24f Gas chromatograms of 0.01 ppm in sandy soil of bromacil (8) and the respective control (50 mg soil equivalent) 88 25a Gas chromatograms, Top; control (500 mg soil equivalent). Middle; 0.01 ppm OP pesticides in fraction II (60% B/H). Bottom; fraction III (5% A/B) 90 25b Chromatograms of 0.01 ppm OP pesticide in frac- tion IV (10% A/B), fraction V (20% A/B) and fraction VI (100% A/B) 91 25c Gas chromatograms of 6 FPD (P) sensitive pest- icides from 4% SE 30/6% OV-210 column 92 25d Gas chromatograms of 6 FPD (P) sensitive pest- icides from 4% SE 30/6% OV-210 column 93 26 Chromatograms of control crude sandy soil (2 mg equivalent) and recoveries of 0.01 ppm of dicamba (LS), DNOC (18) and DNBP (1£) by diazomethane derivatization 95 27a Chromatograms, left; of 1 ppm Zectran (40) and the respective control, right; 1 ppm IPC~(24) and its control (1 mg equivalent), PFPA deriv- atives . k 96 i 27b Chromatograms of 0.1 ppm bux (9) and its respec- tive control (1 mg soil equivalent), PFPA der- ivative 97 27c Gas chromatograms, of Landrins (25,26) and the respective control (1 mg soil eqUivaTent) PFPA derivatives 98 27d Chromatograms of 0.1 ppm carbofuran and its control 99 28a Chromatograms of PFBC-control (5 mg soil equiva- lent) (83%) of 0.01 ppm monuron from sandy soil . . 100 28b Chromatograms of PFBC-control (10 mg soil equiva- lent) and recovery of 0.1 ppm benomyl from sandy soil 101 29 Procedure for detection of forty pesticides in soil . _ 129 ------- Extension of Multi-residue Methodology. II. Figure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Isocratic separation os six carbamate mix on uC,g, 40% acetonitrile - 60% H20 Isocratic separation of early eluting carbamates on yCN, 12.5% acetonitrile - 87.5% H20 Isocratic separation of late eluting carbamates on CDS, 12.5% acetonitrile - 87.5% H20 Relative fluorescence of N-methylcarbamates as a function of time Relative fluorescence of N-methylcarbamates as a function of time Relative fluorescence of N-methylcarbamates as a function of time Relative fluorescence of N-methylcarbamates as a function of time Modular dynamic fluorogenic labelling liquid chromatograph Chromatograms , Top; of sandy soil, spiked at 0.01 ppm. Program No. 10 for 13 min. 25% dioxane - H20 to 40% dioxane - H20, 1 ml/min. Bottom; check sandy soil Chromatogram of sandy loam soil Chromatogram of silty loam soil Chromatogram of sandy soil spiked at 1.0 ppm . . . Chromatogram of sandy loam soil Chromatogram of silty loam soil Analytical curves for six N-methylcarbamate pesticides Page 138 139 139 144 144 145 145 147 149 149 150 150 151 151 154 16 Effect of withholding OPA (substituting pH 10 buffer) on response of celery spiked at 0.2 ppm with 1, Lannate; 2, Temik; 3, Baygon; 4, carbo- furan; and 5, Sevin 156 xi ------- Figure • Page 17 Effect of withholding NaOH (substituting pH 7 H-O) on response of lettuce spiked at 0.2 ppm with: 1, Lannate; 2, Temik; 3, Baygon; 4, carbofuran; 5, Sevin; and 6, Mesural 156 18 Plot of 4 a (peak width at base) for contri- butions of hydrolysis coil and of fluorometer cell, as a function of flow through each 157 19 Aminco Aminalyzer, adapted to N-methylcarbamate analyses 163 20 Aminalyzer chromatogram of: 1, Lannate; 2, Temik; 3, Baygon; 4, carbofuran; 5, Sevin; and 6, Mesurol 164 21 Aminalyzer extended range analytical curve of Lannate 164 xii ------- TABLES Extension of Multi-residue Methodology. I. Number Page 1 List of pesticides under multi-residue methodology investigation 2 2 Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 10% DC-200 19 3 Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 10% QF-1 20 4 Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 4% SE 30/6% OV-210 21 5 Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 1.5% OV-17/1.95% OV-210 22 6 Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 5% OV-210 .' 23 7 Computed efficiencies of five GLC columns 36 8 Retention times and response values relative to parathion of eight. OP pesticides 37 9 Retention times and response values relative to parathion of twelve organophosphate pesticides on 4% SE-30/6% OV-21.0, 200°C 38 10 Retention times and response values relative to parathion of twelve organophosphate pesticides on 1.5% OV-17/1.95% OV-210, 200°C 39 11 Retention times and response values relative to parathion of twelve organophosphate pesticides on 5% OV-210, 200°C 40 12 Retention times and response values relative to atrazine, of Nitrogen-containing pesticides on 5% SE-30 column, 180°C (Hall detector) 49 Xlll ------- TABLES Number Page 13 Retention times and response values relative to aldrin of PFPA derivatized pesticides on two columns (Packard 7820, EC) . . - 50 14 Retention times and response values relative to aldrin of PFPA derivatized pesticides on two columns (Tracor 222, EC) 51 15 Retention times and response values relative to aldrin of 2,5-dichlorobenzenesulfonyl chloride (pyridine system) derivatized pesticides on two GLC columns 52 16 Retention times and response values relative to aldrin of PFBC derivatized pesticides (pyridine system) on 10% DC-200 column, 205°C 57 17 Retention times and response values relative to aldrin of PFBC derivatized pesticides (pyridine system) on 4% SE-30/6% OV-210 column 58 18 Retention times and response values relative to aldrin of PFBC derivatized pesticides (pyridine system) on 1.5% 0V-17/1.95% OV-210 column .... 59 19 Retention times and response values relative to aldrin of PFBC derivatized pesticides (0.1 M NaOH system) on 1.5% 0V-17/1.95% OV-210 60 20 Retention times and response values relative to aldrin of PFBC derivatized pesticides (0.1 M NaOH system) on 4% SE-30/6% OV-210 61 21 Retention times and response values relative to aldrin of BPFT derivatized pesticides on two columns 67 22 Retention times and response values relative to aldrin of diazomethane derivatized pesticides on 4% SE-30/6% OV-210 68 23 Retention times and response values relative to aldrin of selected derivatization techniques on twelve multi-class pesticides 69 24 Percent recovery of 500 yg of pesticides each in the Kuderna-Danish concentrator step concentrated from 250 ml 1:1 benzene/methanol to 10 ml .... 75 xiv ------- TABLES Number Page 25 Effect of paraffin oil keeper on pesticide retention 76 26 Recovery of 100 yg of nitrogen containing pest- icides after heating for 6 hr (100°C) in a capped test tube with 2 ml benzene/methanol (1:1) . 77 27 Average percent recoveries of phorate and para- thion from sandy soil which were spiked at 0.00, 0.01, 0.10, 1.00 and 10.00 ppm 102 28 Soil characterization 104 29 Average percent recoveries of halogenated pest- icides from silty loam soil which were spiked at 0.01, 0.10, 1.00 and 10.00 ppm 105 30 Average percent recoveries of halogenated pest- icides from sandy soil which were spiked at 0.01, 0.10, 1.00' and 10.00 ppm 106 31 Average percent recoveries of halogenated pest- icides from sandy loam soil which were spiked at 0.01, 0.10, 1.00 and 10.00 ppm 107 32 Average percent recoveries of organophosphate pesticides from sandy soil which were spiked at 0.01, 0.10, 1.00 and 10.00 ppm 109 33 Average percent recoveries of organophosphate pesticides from sandy loam soil which were spiked at 0.01, 0.10, 1.00 and 10.00 ppm 110 34 Average percent recoveries of organophosphate pesticides from silty loam soil which were spiked at 0.01, 0.10, 1.00 and 10.00 ppm Ill 35 Comparison of solvents for ultrasonic-Polytron extraction of carbofuran from sandy soil 112 36 Effect of Polytron extraction time on recovery of carbofuran from sandy soil 114 37 Effect of Polytron motor speed on recovery of carbofuran from sandy soil 115 xv ------- TABLES Number 38 Average percent recoveries of pesticides from sandy soil samples which were spiked at 0.01 and 0.10 ppm (Polytron extraction) 116 39 Average percent recoveries of pesticides from sandy soil samples which were spiked at 0.01 and 0.10 ppm (Soxhlet extraction) 117 40 Average percent recoveries of pesticides from sandy soil samples which were spiked at 0.01 and 0.10 ppm (Polytron extraction) 120 41 Average percent recoveries of pesticides from sandy soil samples which were spiked at 0.01 and 0.10 ppm (Soxhlet extraction) 121 42 Multi-residue analysis of fourteen EC-sensitive pesticides in duplicate sandy soil samples which were spiked at 0.01 and 0.10 ppm of forty multi- class pesticides 122 43 Multi-residue analysis of twelve FPD-P sensitive pesticides in duplicate sandy soil samples which were spiked at 0.01, and 0.10 ppm of forty multi- class pesticides 125 44 Multi-residue analysis of twelve multi-class/ nitrogen containing pesticides in duplicate sandy soil samples which were spiked at 0.01 and 0.10 ppm 126 Extension of Multi-residue Methodology- II. Number Page 1 Lower limits of detection for seven carbamates in static Fluram system 141 2 Soil recoveries for seven carbamates for silty loam, sandy loam and sandy soils 159 xvi ------- Extension of Multi-residue Methodology. I. Determining Multiclass Pesticide Residues in Soils by Gas Chromatography H. Anson Moye and Sujit Witkonton Introduction The exceptionally large number of pesticides, represent- ing a diversity of compound classes, presents a severe prob- lem to those charged with their analyses. Efforts to extend the existing multi-residue methodology, largely compiled by EPA and FDA, to those compounds and substrates presently not included would certainly be in order for a number of reasons, which will not be elaborated here. Multi-residue methodology, due to its inception by the FDA, historically has concerned itself with the analysis of foods, primarily utilizing a gas chromatographic determinative step. Introduction of newer pesticides, along with a growing need to know of their presence in non-target systems (the en- vironment) has limited the body of multi-residue methodology in terms of its applicability, partly because these compounds have not been thoroughly tested through the multi-residue methods and partly because those that have been tested have been difficult to detect gas chromatographically. A large number, if not all, of those compounds listed in Table I would fc -\ certainly fall into one or both of these categories. ------- Table 1. List of Pesticides* under Multi-residue Methodology Investigation Number Pesticide Number Pesticide Number Pesticide 1 Amitrole 2 Atrazine 3 AvadexR (diallate) 4 Azinphos ethyl 5 Azinphos methyl 6 AzodrinR (monocrotophos) 7 Benomyl 8 Bromacil 9 BuxR (metalkamate) 10 Captan 11 Chlorobenzilate 12 CIPCR (chlorpropham) 13 Demeton 14 Diazinon 15 Dicamba 16 DifolatanR (captafol) 17 Dimethoate 18 DNOC 19 DNBP 20 DursbanR (chlorpyrifos) 21 Endosulfan 22 Folpet 23 FuradanR (carbofuran) 24 IPCR (propham) 25 2,3,5-LandrinR (trimethylphenyl methylcarbamate) 26 3,4,5-LandrinR 27 Malathion 28 Methorny1 29 Methoxychlor 30 Methyl parathion 31 Monuron 32 Parathion 33 Phosphamidon 34 PCNB 35 PerthaneR (ethylan) 36 Phorate 37 Simazine 38 TemikR (aldicarb) 39 Trifluralin 40 ZectranR (mexacarbate) *Where trade names are used, the common name appears in parentheses. Certain trade names are probably more familiar to many readers than the designated common names. For this reason, some pesticides will be referred to by their trade names through out the text. ------- Although techniques other than gas chromatography have been used in the determinative steps of pesticide methodology schemes, they usually lack sensitivity and selectivity. Usually the inability to perform a gas chromatographic analysis on an organic pesticide can be attributed to thermal instability or lack of volatility- These problems are well known for the car- bamates, carbamoyl oximes, phosphonates and some substituted ureas, and prevent full utilization of the commercially avail- able detectors, many of which would otherwise provide adequate sensitivity and selectivity. The use of chemical derivatives obviates these problems in many cases (1-5). The multi-residue method of FDA, later used and extended by EPA to non-food samples, is largely based upon the extraction and cleanup procedure reported by Mills, Onley and Gaither in 1963 (6). This procedure could be applied to those foods of 21 fat content or less; however, subsequent modifications have covered essentially all food types (7-10). The Pesticide Ana- lytical Manual (PAM) of the Food and Drug Administration con- tains a fairly comprehensive list of those pesticides which have, or have not, been tested through what has become known as the "Mills" procedure. Most of the compounds listed in Table I either were not completely recovered through the procedure or were simply not tested. Since the main responsibilities of the FDA have not in- cluded soil, no information concerning its extraction and clean- up is found within the PAM. Extraction methods in large variety, ------- however, have been reported in the pesticide analytical liter- ature utilizing the full spectrum of solvents and techniques from simple swirling to lengthy Soxhlet extraction. Probably the most extensive review of the field to date is that of Chiba (11) in Residue Reviews, which presents somewhat conflicting data on soil extraction efficiencies. It would be hoped that since this project concerns itself with the extension of existing methodology, those solvents used prior to the Mills Florisil column, would perform adequately as extractants of soil. Unfortunately, according to those works reviewed by Chiba, neither acetonitrile nor petroleum ether (both used in the Mills procedure) are satisfactory. Chiba found that solvent pairs, one nonpolar and ther other polar, were more satisfactory than any other arrangement for the extraction of air dried sandy loam soil. A great many factors contribute to the success with which soils may be extracted, including soil type, moisture content, agitatidn, contact time and solvent. These must all be considered in any efficiency study. Success in gas chromatographing involatile or thermally unstable compounds can often be achieved with one or both of the following approaches: (1) injection of relatively large amounts of pesticide, approaching one microgram, so that chem- ically active sites on the column are momentarily masked, re- sulting in a higher percentage of the pesticide reaching the detector, or (2) chemical derivatization usually at an ionizable function group, so that volatility and thermal stability are ------- improved. The first approach is made more feasible by the recent improvements in element specific detectors, such as the flame photometric, the electrolytic conductivity and the micro- coulometric detectors. These types of detectors possess such large inherent selectivity that relatively large amounts of sample can be injected without interferences becoming pro- hibitive. Selectivity can also be achieved in the second approach by making derivatives which are not only characteristic of certain functional groups or molecular structures but also give responses with element selective detectors or the electron capture detector. To meet the above objective, sequences and scope of re- search work were designed as follows: 1. Screening of pesticides listed in Table 1 for gas chromatographic response. The halogen containing pesticides were checked with the tritium electron capture detector (EC), the phosphorus and sulfur containing pesticides with the flame photometric detector (FPD) and the nitrogen containing pesti- cides with the electrolytic conductivity detector. Five columns of 10% DC-200, 10% QF-1, 4% SE-30/6% OV-210, 1.5% OV-17/1.95% OV-210, and 5% OV-210 and their operating conditions for chlori- nated hydrocarbons which were specified by PAM were tried. 2. Pesticide derivatization. Pesticides which gave poor or no response to gas liquid chromatography (GLC) were ini- tially derivatized by applying the perfluoroacylation method (12) and halogeriated-benzene-sulfonates formation (13). GLC con- ditions that are normally used for the nonderivatized pesticides ------- were used. 3. Soil sample preparation. Pesticides which gave accep- table GLC response were studied for recovery in sandy, sandy loam and siIty loam soils. Triplicate samples of 50-100 g of soil were spiked at 0.00, 0.01, 0.10, 1 and 10 ppm, extracted after curing for 30 minutes, then analyzed. 4. Recovery of pesticides from soil. One hundred gm of each soil sample was extracted by Soxhlet extraction with 300 ml of benzene-methanol (1:1). 5. Cleanup of soil extracts. Soil extracts were cleaned up by silica-gel column chromatography. Alternate solvent systems were examined in order to obtain maximum recovery of pesticides. Results are presented in average percent recoveries with specific accompanying chromatograms to show degree and ex- tent of co-extractive interferences. 6. Special efforts were directed at finding suitable der- ivatives for those pesticides which during the past year's (1974-75) work have been unsuccessfully gas chromatographed (Table 1). Nine additional pesticides were requested by EPA to be included in the 1975-76 multi-residue investigation. For those compounds which appeared appropriate (possess appropriate functional groups), the following derivatization reagents were tested: (1) N-methyl-bis-(heptafluorobutyryl) amide, (2) 9- chloromethyl anthracene, (3) pentafluorobenzoyl chloride, (4) 0- (2,3,4,5,6-pentafluorobenzyl) hydroxylamine hydrochloride, (5) acetic acid-N-hydroxysuccinimide ester, (6) a-bromo-2,3,4,5,6- pentafluorotoluene, (7) diazomethane, (8) pentafluoropropionic 6 ------- anhydride, and (9) N-methyl-bis-(trifluoroacetamide). When reagents became available during the course of the contract period and appeared to offer advantages over those listed above, they were also tested. Standard procedures were used to attempt derivatization on a micro scale. Aliquots of the appropriately manipulated re- action solution were injected onto three gas chromatographic columns (4% SE-30/6% OV-210, 1.5% OV-17/1.95% OV-210, and 5% 0V- 210), specified by EPA, installed in a Tracor 222 gas chroma- tograph equipped with dual tritium EC detectors. For those pesticides which gave responses, analytical curves were made and limits of detection established (three times noise). Re- producibility of derivatization and gas chromatography was also established. Once it had been determined which problem compounds were amenable to gas chromatography by derivatization, a soil ex- traction study was undertaken. Emphasis was on determining the feasibility of utilizing an ultrasound extraction device (29), the Willems Polytron PT 20, and comparing its performance to that of Soxhlet extraction. Since the three soil types in the 1974-75 work behaved similarly in terms of extraction effi- ciencies, only the soil giving the most gas chromatographic interference was studied, i.e. the sandy soil. Preliminary to extracting all the problem compounds, one compound was selected to use for optimizing the extraction technique for the Polytron. Once the optimum extraction tech- nique had been established, the problem compounds were applied 7 ------- as standards to the silica gel columns so as to determine in which eluted fraction they appeared. Subsequently each compound was added to soil, aged for 30 min, extracted with the Polytron, derivatized, if necessary, and analyzed gas chromatographically. Percent recoveries were compared to those already obtained by Soxhlet extraction. However, for those compounds which could be derivatized with one of the proposed reagents, Soxhlet extrac- tions were made for comparison. ------- Experimental Apparatus Electron Capture Detector. Packard model 7820 gas chrom- atograph equipped with two tritium foil electron capture de- tectors, two independent recorders and gas chromatographic oven Packard model 802, was operated under the following conditions: Two 1.8 m x 4 mm I.D. coil columns packed with 10% DC-200 and 10% QF-1 on Gas Chrom Q, 80-100.mesh; inlet temperature 220°C; outlet, 230°C; detector, 220PC; column, 195°C. Electron Capture Detector. Tracer 222 gas chromatograph equipped with two tritium foil electron capture detectors and two independent Soltec recorders was operated under the following conditions: Two 1.8m x 2 mm I.D. U-shaped glass columns packed with either pair of 10% DC-200, 10% QF-1, 4% SE-30/6% OV-210, 1.5% OV-17/1.95% OV-210 or 5% OV-210 on Gas Chrom Q, 80-100 mesh; inlet temperature 230°C; detector, 220°C; column, 205°C. Flame Photometric Detector. Varian Aerograph series 1200 equipped with Tracer flame photometric detector, Soltec re- corder 1 mv, high voltage supply and electrometer of Keithley instruments, was operated under the following conditions: 1.8 m x 2 mm I.D. coil-shaped glass columns packed with the same column packings as for the EC detectors. The packing was ------- topped with 2.5 cm of 10% Carbowax 20 M on Chromosorb W (HP), 80-100 mesh; inlet temperature, 230°C (300°C for Temik and methorny1); detector, 165°C; nitrogen carrier gas flow rate, 50 ml/min; C^, 15 ml/min; air, 50 ml/min; H2, 150 ml/min; high voltage setting, 850 volts. Phosphorus or sulfur mode. The FPD was modified by reversing the H~ and air/02 inlets so that the flame would not be extinguished by solvent after each injection. The sulfur mode of FPD was modified in order to increase its sensitivity by the introduction of S09 (25). The S09/N9 Lt £* £* (1.46 ppm of S02) auxiliary gas was mixed through the H~ line by means of a T-connector located approximately 15 cm before the detector. The auxiliary gas pressure was adjusted to 50 psig. and the flow was controlled at 50 ml/min with a metering valve. Hall Electrolytic Conductivity Detector, Model 310. Tracer 222 gas chromatograph was equipped with one Hall electrolytic \ conductivity detector and a Soltec recorder. The gas chroma- tograph was operated under the following conditions: 1.8 m x 2 mm I.D. U-shaped glass column packed with 5% SE- 30 on Chromosorb W (HP), 80-100 mesh; inlet temperature, 220°C; column, 180°C; transfer line, 270°C; Helium gas flow rate, 67 ml/min. Hall detector settings: Furnace temperature, 820°C; Mode: Nitrogen; Reaction gas: Hydrogen, 33 ml/min; cell flow, 0.5 ml/min. attenuation, 1 x 10. Chromaflex column, size 50-25, Kontes K-422250. Kuderna-Danish evaporative concentrator, 500 ml. 10 ------- Culture tube, 5 ml, Scientific Products, catalog no. T1346- 5. « Microliter pipets, 10, 25, 50, 100, 200, and 300 pi. Eppen- dorf. Soxhlet extraction apparatus; Kimax brand (24005), flask, 500 ml. Fisher Scientific Co. catalog no. 9-551C. Extraction thimbles, glass. Fisher Scientific Co. catalog no. 9-653C. Kerr wide mouth mason jars, 12 quarts. Disposable pasteur pipets. Unitized Extraction Assembly (Precision 65500). Fisher Scientific Co. catalog no. 9-678. Nitrogen evaporator with hot plate maintained at 40-50°C. p Fisher Isotemp dry bath, model 145. Automatic shaker of Eberback Corporation, Ann Arbor, Mich. Automatic tumbling assembly. Maxi-Mix model M-16715, series 16700, Thermolyne Sybron Corporation. Willems Polytron, model PR-20, Brinkmann Instruments, Westburg, New York. Reagents Burdick § Jackson's "distilled in glass" or redistilled solvents: hexane, benzene, acetonitrile, methanol, acetone. Anhydrous sodium sulfate, acetone-extracted. Pyridine, redistilled. Silica gel dry column, activity III, (20% H20) ICN Pharma- 11 ------- ceuticals, Inc., Life Sciences Group, Cleveland, Ohio, catalog no. 404526. Pentafluoropropionic anhydride, PCR, Inc., Research Chem- icals Division, Gainesville, FL, catalog no. 13670. 2,5-dichlorobenzenesulfonyl chloride, Eastman Kodak Co., Rochester, NY 14650. Recrystallized three times from redis- tilled isooctane. N-methyl-bis-(heptafluorobutyryl) amide, N-methyl-bis- (trifluoroacetamide). Regis Chemical Co., 8210 N. Austin Ave., Morton Grove, Illinois. Pentafluorobenzoyl chloride, 98% pure, a-bromo-2,3,4,5,6- pentafluorotoluene, 99+% pure, 9-Anthracene-methanol, Aldrich Chemical Co., 3355 Lenox Rd., N.E., Suite 750, Atlanta, Georgia. Thionyl chloride, Mallinckrodt Chemical Works, St. Louis, Missouri. Acetic acid N-hydroxysuccinimide ester, ICN Pharmaceu- ticals, Inc., Life Science group, Cleveland, Ohio. A 0-(2,3,4,5,6-Pentafluorobenzyl) hydroxylamine, Applied Science Laboratories, Inc., State College, Pennsylvania. Diazomethane in ether, prepared from "Diazald," Aldrich Chemical Co., Inc. (Synthesis procedure obtained from Aldrich's bulletin). Phosphate buffer (pH 7): 3 g NaOH and 17 g KH2P04 in 200ml water. Keeper solution (1% paraffin oil in benzene). 12 ------- Pesticide Standard Solutions Ten mg of each analytical grade pesticide (provided by EPA) was dissolved in 10 ml of redistilled benzene and kept as con- centrate stock solution (1 mg/ml). For those compounds which could not be dissolved readily, benzene/acetone, 1:1, was used. The stock solution was used for derivatization or diluted to obtain on-scale GLC responses for non-derivatized pesticides. Pesticide Derivatization 1. Perfluoroacylation method (12). In the 15 ml culture tube, add 200 yg pesticide, 2 ml hexane, 50 yl pyridine, 50 yl pentafluoropropionic anhydride (PFPA). Mix well and allow reaction to proceed for one hr at room temp- erature. Add 3 ml phosphate buffer pH 7 and mix on Maxi-Mix to stop reaction. Add 3 ml hexane, 50 yl acetonitrile, and mix again. After separation of aqueous and organic phases, remove the bottom layer of aqueous phase with a disposable pipet and discard. Add 2 ml distilled water, mix and aspirate bottom layer; repeat washing two more times. Dry the hexane solution with sodium sulfate. Make final volume to 10 ml with hexane and adjust concentration for on-scale GLC response. 2. Pentafluorobenzoylation (20). 2.1. Pyridine System. In 15 ml culture tube, add 200 yg pesticide, 4 ml benzene, 10 yl of 0.1% pyridine in benzene, 2-5 yl of neat pentafluorobenzoylchloride (PFBC). Mix well and allow reaction to proceed for 30 min by refluxing in dry 13 ------- bath (95°C). Add 4 ml of pH 7 phosphate buffer and mix. After separation of aqueous and organic phases, remove the bottom layer of aqueous phase. Add 4 ml distilled water, mix and as- pirate bottom layer; repeat washing two more times. Dry the organic solution with sodium sulfate. Make final volume to 10 ml and adjust concentration for on-scale GLC response. 2.2. Aqueous Sodium Hydroxide System (0.1 M). In 15 ml culture tube, add 200 yg pesticide, 4 ml 0.1 M NaOH, 2 yl PFBC. Mix well and allow reaction to proceed for 30 min by heating in a dry bath (95°C). After the reaction, add 4 ml of benzene to extract derivatized compound from aqueous phase. This was done in the same way as described in the pyridine system. 3. The derivatization of pesticides with the following compounds was similar to the pentafluorobenzoylation procedure (pyridine system) except that 10 yl neat pyridine was used: acetic acid N-hydroxy succinimide ester (AAHSE), N-methyl-bis- (heptafluorobutyryl) amide (MHFBA), N-methyl-bis-(trifluoro- acetamide) (MBTFA), and 9-chloromethyl-anthracene (CMA) 9-chloromethyl-anthracene was synthesized by refluxing 9- anthracene-methanol with thionyl chloride in benzene for 3 hr (21). 4. Pentafluorobenzylation (17). In 15 ml culture tube, add 200 yg pesticide, 2 ml acetone, 1 ml pH 12 buffer, 10 yl a-bromo-2,3,4,5,6-pentafluorotoluene (BPFT). Mix well and allow reaction to proceed for 2 hr by refluxing in dry bath (100°C). Add 2 ml benzene and mix by shaking. After separation of aqueous and organic phases, remove the aqueous phase and 14 ------- wash the organic phase twice with distilled water as previously described. 5. 0-(2,3,4,5,6-Pentafluorobenzyl) hydroxylamine (PFBHA) derivatization (19). Make a stock solution of the reagent, PFBHA, in pyri4ine to contain 50 mg per ml. In 15 ml culture tube, add 200 yg pesticide in 200 yl ben- zene. Remove benzene under a stream of N~. Add 0.2 ml of the PFBHA stock solution. Vortex the tube to assure contact with the pesticide, stopper and heat in a dry bath at 85°C for half an hour. Remove the pyridine with a stream of N~. Add one ml of hexane, mix, follow with 1 ml of distilled water. Shake the tube and allow the phases to separate. Transfer most of the hexane layer to another tube using a Pasteur pipet, and add anhydrous sodium sulfate to dry the hexane before GLC analysis. 6. Diazomethane derivatization (18). Pipet a benzene solution of an appropriately diluted pesticide standard into a 15 ml culture tube. Remove benzene under a stream of N~. Add one ml of diazomethane reagent and let stand 10 min with occa- sional shaking. Evaporate solvent with N2 and add 5 ml of ben- zene for GLC analysis. 7. Halogenated benzene-sulfonate formation (13). Reflux soil extract of pesticide with 2 mg of 2,5-dichlorobenzene- sulfonyl chloride in 2 ml of acetone and 50 yl pyridine for 30 minutes. The reaction solution is cooled and 5 ml of hexane and 2 ml buffer pH7 are added. Shake well and separate the organic phase. Wash the organic phase three times with 2 ml 15 ------- distilled water each time. Dry the organic phase with sodium sulfate and adjust volume for GLC analysis. Preparation of Spiked Soil Samples and Extraction Procedure Spike triplicate samples of 100 g (10% moisture) at 0, 0.01, 0.10, 1 and 10 ppm of pesticides and extract by Soxhlet extractor with 300 ml 1:1 benzene/methanol for 12 hours. Evaporate benzene/methanol extract by Kuderna-Danish evaporator with repetitive additions of benzene to eliminate methanol and water. Dilute the samples with benzene to obtain appropriate volume for direct GLC analysis. For further silica gel column cleanup, evaporate benzene almost to dryness and take the residue up in 1 ml hexane. As an alternative procedure, tumble or shake one hundred grams of each soil sample for one hour with 300 ml of benzene, acetone or benzene/methanol (1:1), filter, evaporate to dryness on hot plate (50°C) under nitrogen, and take the residue up in 1 ml hexane. * Polytron Ultrasonic Extraction Procedures Spike air-dried sandy soil samples of 50 g at 0, 0.01, and OilO ppm of pesticides and extract with the Polytron. The generator which is immersed in the soil sample is equipped with a saw-tooth cutting head. To the soil sample add 100 ml of benzene/methanol (1:1), unless stated otherwise. With the Polytron generator immersed in the sample, extract the soil twice for 60 sec. at maximum power unless otherwise noted. Filter the extracted soil and solvent through Whatman No. 42 16 ------- filter paper. Rinse the filter paper and the soil twice with 50 ml of the extracting solvent. Evaporate combined filtrate to dryness under a stream of N~ at room temperature (for captan, folpet, Difolatan, and methorny1), or by Kuderna-Danish evaporator, chromatograph the residue on the silica gel column (as below) and analyze gas chromatographically. Silica Gel Column Chromatography Transfer the hexane solutions of soil extracts to silica gel chromatographic columns. Prepare the column by packing 16 g of silica gel topped with 14 g of reagent grade Na-SO^,. Add one ml of soil extract (equivalent to 100 g of soil) to the top of the bed with a disposable pipet. Wash the container with 2, 3 and 4 ml of hexane and deliver to the column respec- tively. Elute the column successively with 50 ml of the fol- lowing solvent systems: (1) hexane, (2) 60% benzene in hexane, (3) 5% acetonitrile in benzene, (4) 10% acetonitrile in benzene, (5) 50% acetonitrile in benzene, and (6) acetonitrile. Evap- orate these fractions and adjust to volume with benzene for GLC analysis. 17 ------- Results and Discussion Screening of pesticides for GLC responses 1. Electron capture detector. Results for fourteen halo- genated pesticides on five GLC columns which gave relatively good responses are presented in Tables 2, 3, 4, 5 and 6. Gas chromatograms of these compounds on three GLC columns are illus- trated by Figures la, b, and c. Their analytical curves for quantitation of soil samples are presented in Figures 2-7. Retention times and peak areas of fourteen halogenated pesticides relative to aldrin (RRT and RPA respectively; see footnotes Table 2) on a 10% DC-200 column are presented in Table 2. All pesticides are presented according to their elution orders from lowest to highest RRTs. Multiple peaks of Avadex indicated the presence of impurities in the originally furnished pesticide standard. A subsequently furnished standard pro- duced only a single GLC peak (Tables 4, 5 and 6). Overlapping peaks on 10% DC-200 (Table 2) are: simazine and atrazine; captan and folpet; endosulfan and Perthane. Their decreasing degree of sensitivities (from highest RPA to lowest RPA) are: e trifluralin, endosulfan, PCNB, Avadex, bromacil, methoxychlor, Perthane, Difolatan, chlorobenzilate, folpet, captan, atrazine, simazine and CIPC. RRTs and RPAs of fourteen halogenated pesticides on 10% i QF-1 are presented in Table 3. Overlapping peaks are: PCNB, 18 ------- Table 2. Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 101 DC-200, 195°C (Att. = 3 x 10" , N7 flows = 60 ml/min. Packard 7820, EC). L Pesticide Avadex CIPC Trifluralin Simazine Atrazine PCNB Bromacil Captan Folpet Endosulfan Perthane Chlorobenzilate Difolatan Methoxychlor Aldrin RRT1 0.23, 0.29 0.37, 0.39, 0.57 0.29 0.34 0.39 0.40 0.48 0.92 1.18 1.19 1.50, 2.03 0. 1.98 2.36 3.11 4.60 1 (12.5 min) Computed as retention time relative 2Dt>A _ peak area of pesticide/amount Rp.2 Analytical Curve (Figure No.) 0.500 0.002 2.300 0.015 0.017 0.530 0.302 0.050 0.055 570, 0.083 0.101 0.078 0.088 0.240 1.000 to aldrin. of pesticide 4 5 — — 7 5 — — 7 4 4 — 7 — peak area of aldrin/amount or aldrin Only major peak of endosulfan was used for analytical curve construction. Note: Major peaks are indicated by underlining. 19 ------- Table 3. Retention times and response values relative to aldrin of fourteen EC-sensitiveQpesticides on 10% QF-1, 195°C (Att. = 1 x 10~ , N? flow = 60 ml/min. Packard 7820, EC). Pesticide Avadex CIPC PCNB Simazine Atrazine Trifluralin Perthane Endosulfan Folpet Bromacil A Captan Chlorobenzilate Methoxychlor Difolatan Aldrin RRT1 0.50 0.63 0.77 0.79 0.79 1.00 1.69 2.31, 4.00 2.92 3.08 3.23 3.54 5.00 8.17 1 (2.5 min) RPA2 0.017 0.001 0.039 0.002 0.013 0.394 0.065 0.390, 0.185 0.292 0.358 0.306 0.083 0.173 0.167 1.000 1^2See footnotes of Table 2. Note: Major peaks are indicated by underlining. 20 ------- Table 4. Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 4% SE-30/6% OV-210, 205°C (Att. = 2 x 10 , N9 flow = 67 ml/min. Tracer 222, EC). L Pesticide CIPC Avadex Simazine Atrazine Trifluralin PCNB Bromacil Endosulfan Folpet Captan Perthane Chlorobenzilate Methoxychlor Difolatan Aldrin RRT1 0.38 0.46 0.52 0.54 0.54 0.62 1.48 1.75, 2.56 1.76 1.76 1.91 2.53 4.30 4.40 1 (7.5 min) nr>A2 Analytical Curve KWV (Figure No.) 0.004 0.019 0.004 0.007 0.247 0.430 0.230 0.700, 0.210 0.136 0.105 0.008 0.025 0.242 0.105 1.000 — 2 6 6 — — — — 3 3 — — — 3 — " C/i« £ r\t*+-t\r\+- **c- nf T»^V.1» "J Note: Major peaks are indicated by underlining, 21 ------- Table 5. Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 1.5% OV-17/1.95% OV-210, 205°C (Att. 2 x 10 , N flow = 53 ml/min. Tracer 222, EC). Pesticide Trifluralin CIPC Avadex PCNB Atrazine Simazine Chlorobenzilate Endosulfan Folpet Cap tan Perthane Bromacil Difolatan Methoxychlor Aldrin RRT1 0.45 0.45 0.53 0.70 0.70 0.70 1.55, 3.32 1.97, 3.58 2.70 2.70 2.70 3.25 7.35 7.80 1 (2.8 min) RPA2 0.204 0.002 0.031 0.420 0.009 0.008 0.012, 0.019 0.660, 0.130 0.070 0.071 0.013 0.061 0.026 0.139 1.000 footnotes of Table 2. Note: Major peaks are indicated by underlining. 22 ------- Table 6. Retention times and response values relative to aldrin of fourteen EC-sensitive pesticides on 51 OV-210, 205°C (Att. 2 x 10 , N? flow = 53 ml/ min. Tracer 222, EC). Pesticide Avadex CIPC Atrazine Simazine PCNB Trifluralin Perthane Chlorobenzilate Endosulfan Folpet Cap tan Bromacil Methoxychlor Difolatan Aldrin RRT1 0.59 0.69 0.85 0.85 0.88 1.00 1.65 1.78, 5.25 2.22, 3.75 3.00 3.16 3.69 4.60 7.10 1 (2.3 min) RPA2 0.014 0.002 0.008 0.008 0.680 0.230 0.008 0.011, 0.029 0.860, 0.180 0.039 0.049 0.129 0.108 0.009 1.000 * See footnotes of Table 2. Note: Major peaks are indicated by underlining. 23 ------- to CTil to I 4% SE 30/6% 0V 210, 205°C, ATT. 2 x 102 0.1 ng ea. EC O) 0 5 10 15 20 25 30 MIN. FIGURE 1 a Chromatograms of 14 EC-sensitive pesticides. Pesticide mixtures were grouped to prevent peak coincidence; num- bers correspond to pesticides listed in Table 1. ------- 01 ro N 0.1 ng ea. (NJ 1.5% 0V 17/1.95% OY 210 205°C, ATT. 2 x 10 , EC IT) rtr O1 bO (H uo P vO MIN. FIGURE 1 b TfT Chromatograms of 14 EC-sensitive pesticides on 1.5% OV-17/ 1.95% OV-210 column. ------- 5% 0V 210, 205°C ATT. 2 x 102, EC 0 10 2 15 20 FIGURE 1 c Chromatograms o£ 14 EC-sensitive pesticides 50 MIN. ------- ts) fj W w 80- 70- 60- 50- 40- 30- 20-1 10- SE-30/6! OV-210, 205°C, ATT. 4 x 10 EC DETECTOR 0.1 0.2 0.3 NG AVADEX 0.4 FIGURE 2 Analytical curve of Avadex. 0.5 ------- tsj 06 160. 140. 120. 100 S 80- w w 60. 40 20 4% SE-30/6% OV-210, 205°C EC DETECTOR, ATT.=4 x 10 FOLPET DIFOLATAN O CAPTAN 0.05 0.10 0.15 0.20 0.25 NG CAPTAN (X 2 = NG, FOLPET; X 10 = NG, DIFOLATAN) FIGURE 3 Analytical curves of Captan, Folpet and Difolatan. ------- to 140 120 E 100 H S 80. w ac w 60. 40 20 10% DC-200, 205°C, ATT. 4 x 10' EC DETECTOR CHLOROBENZILATE PERTHANE 5 10 15 20 NG PERTHANE OR CHLOROBENZILATE (x4 = NG, CIPC) FIGURE 4 Analytical curves of Perthane, Chlorobenzilate and CIPC ------- -J O HH Pi s 60 . 50 . 40- 30 w H ac 20J HH < [•?•[ m = * 10-1 w 10% DC-2-&0, 205°C, ATT. 4 x 10' EC DECTECTOR TRIFLURALIN ROMACIL .02 .04 .06 NG TRIFLURALIN (x25 = NG BROMACIL) FIGURE 5 Analytical curves of Trifluralin and Bromacil. .08 ------- O4 M 140 120-1 100 80 g 60j S 40 20 41 SE-30/6% OV-210, 205°C, ATT. 4 x 10 EC DETECTOR SIMAZINE TRAZINE 1234 NG SIMAZINE (x5 = NG, ATRAZINE) FIGURE 6 Analytical curves of Simazine and Atrazine. ------- Mi W ac w 120.1 10% DC-200, 205°C, ATT. 4 x 10' EC DETECTOR 100. 80 60 40 20 ^ , .01 .02703704T057(h>.6? .08 NG PCNB (x5 = NG, ENDOSULFAN, xlO = NG, METHOXYCHLOR) FIGURE 7 Analytical curves of PCNB, Endosulfan and Methoxychlor. ------- simazine, atrazine; folpet and bromacil. Their decreasing degree of sensitivities are: trifluralin, endosulfan, bromacil, captan, folpet, methoxychlor, Difolatan, chlorobenzilate, Perthane PCNB, Avadex, atrazine, simazine and CIPC. Table 4 presents RRTs and RPAs of these compounds on the 4% SE 30/6% OV-210 column. Overlapping peaks are: simazine (5_7) , atrazine (2) and tri- fluralin (59); endosulfan (21), folpet (22) and captan (10); chlorobenzilate (11) and endosulfan (21), methoxychlor (29) and Difolatan (16) (see Figure 1 a for chromatograms). Their de- creasing degree of sensitivities are: endosulfan, PCNB, tri- fluralin, methoxychlor, bromacil, folpet, captan, Difolatan, chlorobenzilate, Avadex, Perthane, atrazine, simazine and CIPC. RRTs and RPAs of fourteen halogenated pesticides on 1.5% OV-17/1.951 OV-210 are presented in Table 5. Overlapping peaks are: trifluralin (39) and Avadex (3); PCNB (34), atrazine (2) and simazine (37); folpet (22), captan (10) and Perthane (55) (see Figure 1 b for chromatograms). Their decreasing degree of sensitivities are: endosulfan, PCNB, trifluralin, methoxychlor, captan, folpet, bromacil, Avadex, Difolatan, chlorobenzilate, Perthane, atrazine, simazine and CIPC. Table 6 presents RRTs and RPAs of these compounds on the 5% OV-210 column. Overlapping peaks are: atrazine (2), sim- azine (5_7), PCNB (54.), and trifluralin (59) ; folpet (22) and captan (10); endosulfan (21) and bromacil (8) (Figure 1 c). Their decreasing degree of sensitivities are: endosulfan, PCNB, trifluralin, bromacil, methoxychlor, captan, folpet, chloroben- zilate, Avadex, Difolatan, atrazine, simazine, Perthane and CIPC. 55 ------- From Figure 1 a, the overlapping peaks of atrazine (2), simazine (37) and trifluralin (2.9) on the 4% SE 30/6% OV-210 column could be separated by the 1.5% 0V-17/1.95% OV-210 column (Figure 1 b). None of the columns under investigation could be used to separate atrazine (2) and simazine (57). The best column for the separation of captan (10), folpet (22) and en- dosulfan (21) was the 5% OV-210 column (Figure 1 c). The over- lapping peaks between chlorobenzilate (11) and endosulfan (21), methoxychlor (29) and Difolatan on the 4% SE 30/6% OV-210 column (Figure 1 a) could be separated by 5% OV-210 (Figure Ic). Most of the analytical curves for the halogenated pesticides (Figure 2-7) exhibit good linearity and they intercept the point of origin except for folpet, Difolatan and captan (Figure 3), CIPC, chlorobenzilate and Perthane (Figure 4) and simazine (Figure 6), suggesting minor loss of these pesticides in the GLC system. This might be due to some degradation of these pest- .» icides on the GLC column under elevated temperature or loss by absorption on GLC liquid phases. However, quantitation of these pesticides was still possible and fairly accurate results could be obtained because consistent and linear analytical curves were observed. Unsatisfactory results were obtained for most nitrogen- containing pesticides and carbamate type compounds with the EC detector. Except for the tailing peak of bromacil and low sen- sitivity of CIPC, those pesticides giving moderately sensitive : peaks were atrazine, bromacil, Difolatan and simazine (Tables 2-6). Zectran gave inconsistently unacceptable low sensitivity. 34 ------- Those pesticides producing inconsistent multiple peaks or no responses were amitrole, benomyl, monuron, DNOC, DNBP, dicamba, Landrins (2,3,5 § 3,4,5), Zectran, IPC, bux and carbofuran. Theoretical plates were obtined from aldrin peaks for five tested GLC columns (Table 7). 4% SE 30/6% OV-210 gave the highest theoretical plates; therefore it exhibited superior separations (Figure 1 a). 2. Flame photometric detector (phosphorous mode). Re- tention times and peak areas of organophosphate (OP) pesticides relative to parathion on five GLC columns are presented in Tables 8-11. They are presented according to their elution order (from lowest to highest RRTs). Analytical curves are presented in Figures 8-13. Table 8 presents RRTs and RPAs of eight OP pesticides on 10% DC-200 and 10% QF-1 columns. The overlapping peaks on 10% DC-200 are phorate, Azodrin and demeton; on 10% QF-1 methyl parathion and malathion do not separate. No GLC peaks were obtained from the injection of Azodrin and azinphos methyl on 10% QF-1. Their decreasing sensitivities on 10% DC-200 are: azinphos methyl, parathion, methyl paration, phorate, malathion, azinphos ethyl, demeton, and Azodrin, on 10%- QF-1: parathion, malathion, phorate, methyl parathion, azinphos ethyl, demeton, azinphos methyl, and Azodrin. Table 9 presents RRTs and RPAs of twelve OP pesticides on a 4% SE 30/6% OV-210 column. Overlapping occurs for dimethoate and Azodrin; methyl parathion and malathion; parathion and phos- phamidon. Their decreasing degree of sensitivities are: diaz- 35 ------- CM o\ Table 7. Computed efficiencies of five GLC columns (Tracer 222, tritium foil EC detector, column temperature = 205°C. Att. 4 x 102) . r_, __ Efficiency L01umn (Theoretical Plates) 4% SE-30/6% OV-210 1.5% OV-17/1.95% OV-210 10% DC-200 5% OV-210 10% QF-12 2700 2300 1936 1849 676 N, Flow (mx/min) 67 53 67 53 60 Efficiency computed from total retention and base width of aldrin peak. 2 Result was obtained from Packard 7820, EC. ------- C/4 -J Table 8. Retention times and response values relative to parathion of eight organo- phosphate pesticides on two GLC columns (Att. 8 x 10~8, Varian 1200, FPD-P) 10% DC-200: 200° C, 50 ml/min Pesticide » Phorate Azodrin Demeton Methyl parathion Malathion Parathion Azinphos methyl Azinphos ethyl RRT1 0.36 0.37 0.39 0.68 0.89 1 (4.4 min) 5.45 7.25 •'•Computed as retention time relative 2RPA = Peak area of pesticide/amount RPA2 0.63 0.01 0.11 0.71 0.26 1.00 1.06 0.20 to parathion. of pesticide 10% QF-1: 200 RRT1 0.19 no peak 0.34 0.78 0.76 1 (3.8 min) no peak 6.10 0 C, 50 ml/min RPA2 0.53 — 0.11 0.42 0.64 1.00 — 0.29 peak area of parathion/amount of parathion responses when 1 yg of pesticides were injected ------- Table 9. Retention times and response values relative to parathion of twelve organophosphate pesticides on 4% SE-30/6% OV-210, 200° C (Att. 8 x 1Q-8, N2 flow = 50 ml/min. Varian 1200, FPD-P). , 7 Analytical Curve Pesticide RRT1 RPA^ (Figure ^ Demeton 0.17, 0.32 0.300, 0.500 (11) Phorate 0.26 1.590 (11) Diazinon 0.29 1.840 10 Dimethoate 0.48 0.414 9 Azodrin 0.50 0.100 8 Dursban 0.64 1.700 9 Methyl parathion 0.77 0.696 (13) Malathion 0.82 0.555 (12) Parathion 1.00 (4.9 min) 1.000 (12) Phosphamidon 0.74, 1.045 0.047, 0.285 9 .» Azinphos methyl 5.26 0.278 (13) Azinphos ethyl 6.50 0.370 (13) 1^2See footnotes of Table 8. Parentheses indicate other GLC conditons were used; see these conditions from respective figures. ------- Table 10. Retention times and response values relative to parathion of twelve organophosphate pesticides fi on 1.5% 0V-17/1.95% OV-.210, 200° C (Att. 8 x 10 ~° N2 flow = 50 ml/min. Varian 1200, FPD-P). Pesticide Pemeton Phorate Diazinon Dimethoate Azodrin Phosphamidon Dursban Methyl parathion Malathion Parathion Azinphos methyl Azinphos ethyl RRT1 0.16, 0.32 0.24 0.32 0.54 0.59 0.68, 0.97 0.76 0.77 0.88 1 (2.6 min) 9.65 12.00 RPA2 0.395, 0.762 ; .- 1 1.240 0.620 0.830 0.162 0.034, 0.149 1.270 1.490 1.030 1.000 1.070 0.780 1^2See footnotes of Table 8. 39 ------- Table 11. Retention times and response values relative to parathion of twelve organophosphate pesticides on 5% OV-210, 200° C (Att. 32 x 10"9, N? flow = 50 ml/min. Varian 1200, FPD-P). Pesticide Demeton Phorate Diazinon Dursban Dimethoate Azodrin Malathion Methyl parathion Parathion Phosphamidon Azinphos methyl Azinphos ethyl RRT1 0.12, 0.31 0.17 0.19 0.43 0.52 0.67 0.74 0.79 1 (3.04 min) 1.32 5.10 6.20 RPA2 0.292, 0.780 0.500 1.610 1.285 0.450 0.083 0.525 0.875 1.000 0.136 0.233 0.212 152 See footnotes of Table 8. 40 ------- 240 200- 4% SE-30/6% OV-210, 210°C, ATT. 8 x 10 (FPD, PHOSPHORUS MODE) -8 w X 160 120 80 40 20 40 60 80 NG AZODRIN FIGURE 8 Analytical curve of Azodrin. 100 ------- ts) W a: w a, 120 100 80 60 40 20 4% SE-30/6% OV-210ao200°C ATT = 8 x10 ' (FPD, PHOPHORUS MODE], DURSBAN 10 HOSPHAMIDON 40 50 20 30 NG FIGURE 9 Analytical curves of Dursban, Dimethoate and Phosphamidon ------- •p* tri H a: w a: w a, 100- 80. 60 40 20 41 SE-30/6% OV-210. 200°C, ATT, 8 x 10~8 (FPD, PHOSPHORUS MODE) 2 4_ 6 NG DIAZINON FIGURE 10 Analytical curve of Diazinon 10 ------- 140 10% DC-200, ATT. = 8 x 10 (Phosphorus Mode) DEMETON (column = 200° HORATE (column = 180°C) 2 34567 NG PHORATE (x5 = NG DEMETON) FIGURE 11 Analytical curve of phorate and demeton. ------- on w ac 100 80 . 60 . 40 . w 20 . 10% DC-200, 200°C, ATT, 8 x ID'8 (Phosphorus Mode) MALATHION PARATHION 2 4 6 8 10 NG MALATHION OR PARATHION 12 FIGURE 12 Analytical curves of malathion and parathion ------- -pa. ON 160 120 80 W S 40. 1.5% OV-17/1.95% OV-210, ATT. = 10 "8 (Phosphorus Mode) AZINPHOS ETHYL* (Col. = 240°C) AZINPHOS METHYLO (Col. = 220°C) METHYL PARATHIOND (Col. = 200°C) 1 234567 NG METHYL PARATHION (xlO = NG, AZINPHOS ETHYL OR AZINPHOS METHYL) FIGURE 13 Analytical curves of Azinphos Ethyl, Azinphos Methyl and Methyl Parathion ------- inon, Dursban, phorate, parathion, methyl parathion, malathion, demeton, dimethoate, azinphos ethyl, phosphamidon, azinphos methyl and Azodrin. table 10 presents RRTs and RPAs of twelve OP pesticides on 1.5% 0V-17/1.951 OV-210 column. The overlapping peaks are: demeton and diazinon; dimethoate and Azodrin; Dursban and methyl parathion. Their decreasing degree of sensitivities are: methyl parathion, Dursban, phorate, azinphos methyl, malathion, parathion, dimethoate, azinphos ethyl, demeton, diazinon, Azodrin and phosphamidon. Table 11 presents RRTs and RPAs of twelve OP pesticides on 5% OV-210 column. The overlapping peaks are: phorate and diazinon; malathion and methyl parathion. Their decreasing degree of sensitivities are: diazinon, Dursban, parathion, methyl parathion, demeton, malathion, phorate, dimethoate, azin- phos methyl, azinphos ethyl, phosphamidon and Azodrin. It was uncertain whether the minor peak for phosphamidon was an impurity or derived from degradation on the GLC column (Tables 9-10). However, the reproducibility of the major peak was good and only this peak was used for recovery computation. All of the analytical curves for the organophosphate pesticides (Figures 8-13) exhibit good linearity; and intercept the point of origin except for phosphamidon (Figure 9), suggesting in- creasing loss of phosphamidon when larger amounts were injected. 3. Hall electrolytic conductivity detector. Table 12 presents retention times and response values relative to atra- zine of some nitrogen-containing pesticides. Most of the 47 ------- tested pesticides did not give acceptable GLC peaks, which could be due to the unsuitable GLC column or thermal decomposition of pesticides. 28% Pennwalt 223/4% KOH and 5% Poly A-135 on 80/100 mesh Gas-Chrom R, which are used primarily for the separation of simple amines or anilines, were tested for these nitrogen- containing pesticides. Unfortunately, no GLC peaks from any of the tested pesticides were observed (column temperature 160°- 180°C). It might be worthwhile to note that benomyl produced a peak which appeared immediately after the hexane solvent peak (5% SE-30 column). This peak might not be the intact benomyl peak because one would not expect it to have such a short re- tention time. It is assumed that benomyl may have undergone thermal decomposition to give butylisocyanate. Some solvents such as ethylacetate, methylene chloride or other halogen-containing compounds would severely contaminate the Hall detection system and consequently create noisy base- lines. Hexane seems to be the best solvent for this detector; however, it was observed that a periodic injection of pyridine (1 ul) corrected the noisy baseline. Pyridine probably neutra- lized the excess acids in the GLC system generated by repet- itive sample injections. Derivatization of pesticides Derivatization of nitrogen-containing pesticides was at- tempted by using pentafluoropropionic anhydride (PFPA) (Tables 13-14) and 2,5-dichlorobenzenesulfonyl chloride (DBSC) (Table 15). Most of the pesticides which have a proton attached to the nitrogen and an aromatic structure gave positive results 48 ------- Table 12. Retention times and response values relative to atrazine, of Nitrogen- containing pesticides on 5% SE-30 column, 180° C CHall detector, att. = 1 x 10; furnace, 820° C; Ho flow, 33 ml/min; cell flow, 0.5 ml/min; He flow, 67 ml/min; inlet, 225° C; transfer line, 280° C). Pesticide Atrazine CIPC Carbofuran Simazine IPC Zectran RRT 1 (1.87 min) 0.76 1.00 1.05 0.333 0.28, 1.38 RPA Remarks 1 9 mm^/ng, 141 Full scale response 0.206 0.035 4.200 2.600 0.48, 0.30 two peaks DNOC (500 ng) Temik (1 yg) Benomyl (1 yg) Avadex (1 yg) Bux (1 yg) Amitrole (1 yg) Azodrin (1 yg) Monuron (1 yg) no peak inconsistent tailing peak no peak except enlarged solvent peak noisy multiple peaks multiple tailing peaks no peak multiple peaks inconsistent tailing peaks ------- Table 13. Retention times and response values relative to aldrin of PFPA derivatized pesticides on two columns (Packard 7820, EC). tn o 10% DC-200; 195° C, 60 ml/min Pesticide IPC CIPC Simazine Atrazine Monuron Furadan Bux Zectran Aldrin Att. 3 RRT1 •*> 0.08 0.16 0.26 0.28 0.28 0.29 0.34, 0.41 0.38 1 (13 min) x 10"9 RPA2 0.230 0.088 0.640 0.240 0.087 0.158 0.167, 0.0206 0.0111 1.0000 10% QF-1; 195° C, 60 ml/min Att. RRT1 0.31 0.38 0.62 0.62 1.00 0.62 0.54, 0.69 0.62 1 (2.5 min) 1 x 10"9 RPA2 0.159 0.0336 0.346 0.236 0.0727 0.150 0.184, 0.0455 0.018 1.000 152 See footnotes of Table 2. The pesticides which were also derivatized and produced no responses were Avadex Benomyl, Bromacil, Difolatan, Amitrole, Diquat, DNOC, Ferbam, Temik and Azodrin (PFPA of Azodrin was also checked by FPD, no response was observed). ------- Table 14. Retention times and response values relative to aldrin of PFPA derivatized pesticides on two columns (Tracor 222, EC, Att. 4 x 102, xolumn at 195°C). Pesticide Aldrin 2,3, 5-Landrin 3,4,5-Landrin 2-AB3 DNBP Dicamba 4% SE-30/6% N0 flow = 6 RRT1 1.000 0.350 0.425 UD UD UD OV-210 7 ml/min RPA2 1.000 0.063 0.030 — — — 1.51 No RRTl 1.000 0.270 0.346 — — — OV-17/1.95% OV-210 flow =53 ml/min RPA2 1.000 0.218 0.087 — — — See footnotes of Table 2. 2-AB = 2-aminobenzimidazole, a degradation product of benomyl. UD = underivatizable. ------- Table 15. Retention times and response values relative to aldrin, of 2, 5-dichloro- benzenesulfonyl chloride derivatized pesticides (Pyridine system) on two GLC columns (Tracer 222, EC,-Att. 4 x Z en tsJ Pesticide Carbofuran Temik3 DNOC Monuron Alpha-naphthol Bux Aldrin 41 SE-30/6% 220° C, 60 RRT1 *r 0.50 0.69 0.75 1.00 4.00 5.00, 6.00 0 1 (1.25 min) OV-210 ml/min RPA2 0.013 0.040 0.400 0.067 0.040 .020, 0.003 1.000 1.51 OV-1 220° C RRT1 3.75 0.75 0.75 0.50 6.00 3.00, 4.00 1 (0.63 min) 7/1.95% OV-210 , 60 ml/min RPA2 0.014 0.051 0.700 0.030 0.034 0.01.5, 0.008 1.000 See footnotes of Table 2. Similar results were obtained for methomyl. When carbaryl was derivatized in buffer pH 12 system, peaks of RRT 4.0 and 0.69 appeared. The pesticides which were also derivatized and produced no responses were Amitrole, Avadex, Benomyl, Azodrin, Simazine, Atrazine, Bromacil, CIPC, IPC, Zectran Seviri and Difolatan. ------- (Table 13-14). Linear analytical curves are presented in Figures 14 and 15. The thermal instability of N-methylcarbamates and their reaction with acylating agents were briefly discussed by Khalifa and Mumma (5). Damico and Benson (14) postulated that the proton attached to nitrogen is readily transferable to the phenolic oxygen when the carbamate is subjected to elevated temp- eratures or under mass spectroscopic conditions. When the proton attached to the nitrogen is replaced, the products have greater thermal stability (15). Moye's original derivatization techniques (13) of DBSC with carbamates were modified by substituting the pH 12 buffer with a pyridine system (Table 15). This relatively high pH system may split- the intact carbamates before they react with DBSC. From comparison of the results obtained from carbaryl and alpha- naphthol in these two reaction systems, one might be able to postulate the reaction mechanisms of carbaryl and other similar carbamates with DBSC. In the pH 12 buffer system, carbaryl was split into two reactive parts, namely alpha-naphthol and methyl- amine or methylisocyanate which subsequently reacted with DBSC to yield l-naphthyl--2,5-dichlorobenzene sulfonamide (RRT = 4.00) and N-methyl 2,5-dichlorobenzene sulfonamide (RRT = 0.69) re- spectively (Table 15). In either the pH 12 or pyridine system, l-naphthyl-2,5-dichlorobenzenesulfonate (RRT = 4.00) was ob- tained when alpha-naphthol was substituted for carbaryl. In the pyridine system, carbaryl is likely to stay intact because no GLC peaks were produced. However, DBSC derivatives of Temik CRRT = 0.69) and methomyl (RRT = 0.69) in the pyridine system 53 ------- en H 5 HH w w DH 280. 240 . 200 . 160. 120. 80 40 SE-30/6% OV-210, 160°C. EC DETECTOR ATT. = 2 x 10 0.02 0.04 0.06 NG PFPA-IPC 0.08 FIGURE 14 Analytical curve of PFPA-IPC. 0.10 ------- tn tn 120 f 10° v_/ s 80 ' I—1 K 60 . w ^ 40 . 20 • 4% SE-30/6% OV-210, 205°C,?EC DETECTOR, ATT. = 4 x 10^ LANDRIN-3,4,5 FURADAN BUX LANDRIN-2,3,5 ZECTRAN 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 NG PFPA-DERIVATIZED PESTICIDES FIGURE 15 Analytical curves of Furadan, Bux, Zectran, Landrins (PFPA) ------- both gave identical RRT peaks, and they correspond to N-methyl- 2,5-dichlorobenzene sulfonamide. Retention times and response values relative to aldrin of pentafluorobenzoyl chloride (PFBC) derivatized pesticides in pyridine and O.I M NaOH systems are presented in Tables 16-18 and Tables 19-20 respectively. The best results obtained in the pyridine system were for amitrole, benomyl and monuron, although the resulting derivatives were far from perfect. One of the serious short-comings was the interference of excess PFBC which could not be eliminated by simple solvent partition. The PFBC GLC peak fused together with the solvent peak and caused serious tailing for about 15 min. Therefore, from Table 17, the short RRT (0.316) of amitrole was masked by excess PFBC, although it created no serious problem for the longer RRTs of benomyl (14.3) and monuron (2.98). A disadvantage of PFBC-amitrole was the tailing peak, which leads one to suspect its structural stabil- ity under normal GLC ^conditions. PFBC-benomyl seemed to be ideal for GLC detection except for its somewhat low sensitivity (RPA = 0.077) and long RRT (14.3). PFBC-monuron offered the best PFBC-derivative which could be applied to routine analyses. Attractive qualities were seen in the moderately long RRT (2.98) which separated it from most of the soil impurities and the good EC-GLC response (RPA = 0.73). The non-reactivity of PFBC with most soil impurities made it even more attractive for application to routine residue analysis. The linear ana- lytical curves of PFBC-derivatives of amitrole, benomyl and monurpn are presented in Figures 16-18. 56 ------- Table 16. Retention times and response values relative to aldrin of PFBC derivatized pesticides (pyridine system! on 10% DC-200 column, 205°C (Att. = 4 x 10 , N2 flow - 80 ml/min. Tracer 222, EC). Pesticide Amitrole Avadex Azodrin Benomyl Bux DNOC Carbofuran IPC Monuron Temik Zectran RRT 0.188 UD2 (2.97, 0.37, 0.56)1 1.00 10.400 0.375, 0.690 UD2 0.580 0.500 1.310 UD2 0.810 RPA 1.960 (0.033)1 0.018 0.110 0.004, 0.005 — 0.024 0.003 1.07 — 0.003 Data represents original structure of avadex, only major peak (underlining) was computed for RPA. UD = Underivatizable 5MBC and 2-AB PFBC-derivatives give similar RRT. 57 ------- Table 17. Retention times and response values relative to aldrin of PFBC derivatized pesticides (pyridine system) on 4% SE-30/6% OV-210 column, 205° C, (Att. = 4 x 10 , N9 flow = 80 ml/min. Tracor 222, EC). Z Pesticide I Amitrole Avadex Azodrin Benomyl Bux Dicamba DNBP DNOC Carbofuran IPC 2,3 , 5-Landrin 3,4, 5-Landrin Monuron Met homy 1 Phosphamidon Temik Zectran RRT 0.316 UD (4.404, 0.456, 0.670)1 0.350 14.300 0.825 UD UD UD 0.825 UD 0.705 0.872 2.980 UD UD UD 1.020 RPA 1.235 (0.027)2 0.012 0.077 0.004 — — — 0.001 — 0.008 0.023 0.730 — — — 0.001 (underlining) was computed for RPA. ; maj 'MBC and 2-AB PFBC- derivatives gave similar RRT. 58 ------- Table 18. Retention times and response values relative to aldrin of PFBC derivatized pesticides (pyridine system) on 1.5* OV-17/1.95% OV-210 column, 205° C (Att. 4 x 10 , N9 flow = 80 ml/min. Tracor 222, EC). L Pestic^e Amitrole Avadex Azodrin 2 Benomyl Bux Dicamba DNBP DNOC Carbofuran IPC 2,3,5-Landrin 3,4,5-Landrin Monuron Me thorny 1 Phosphamidon Temik Zectran RRT 0.522 i UD (0.24, 0.57, 0.74)1 — 19.00 0.784 UD UD ; UD 0.74 0.61 0.523 0.670 2.65 UD UD UD 0.822 RPA 0.072 i (0.041)1 — 0.196 0.001 — — — 0.055 0.001 0.010 0.039 0.480 — — — 0.014 Data represents original -structure of avadex; only major peak (underlining) was computed for RPA. 2MBC and 2-AB PFBC-derivatives gave similar RRT. 59 ------- Table 19. Retention times and response values relative to aldrin of PFBC derivatized pesticides (0.1 M NaOH system) on 1.5% OV-17/1.95% OV-210 column, 205° C (Att. = 4 x 10 , N7 flow = 80 ml/min. Tracer 222, EC). L Pesticide RRT RPA Amitrole Avadex Azodrin Benomyl Bux Dicamba DNBP DNOC Garbofuran IPC 2,3,5-Landrin 3,4,5-Landrin Monuron Methomy1 Phosphamidon Temik Zectran UD UD UD UD UD UD UD UD UD 0.792 0.580 0.700 1.57, 2.55 UD UD UD 13.25 0.017 0.340 0.445 0.001, 0.005 0.035 60 ------- Table 20. Retention times and response values relative to aldrin of PFBC derivatized pesticides (0.1 M NaOH system) on 4% SE-30/6% OV-210 column, 205° C (Att. = 4 x 10% N9 flow = 80 ml/min. Tracer 222, EC). L Pesticide RRT RPA Amitrole Avadex Azodrin Benomyl Bux Dicamba DNBP DNOC Garbofuran IPC 2,3,5-Landrin 3,4,5-Landrin Monuron Methomyl Phosphamidon Temik Zectran UD UD UD UD UD UD UD UD UD 0.870 0.710 0.840 1.76, 5.00 UD UD UD 19.4 0.016 0.212 0.181 0.001, 0.004 0.031 61 ------- In the 0.1 M NaOH system, no responses were obtained from amitrole, benomyl and most other pesticides (Table 19). Two low response GLC peaks were obtained from monuron which might be the result of its thermal degradation. Moderate sensitiv- ities (RPAs 0.2-0.4) were obtained for Landrin (2,3,4 § 3,4,5) (Tables 19-20). Retention times and response values relative to aldrin of a-bromo-2,3,4,5,6 -pentafluorotoluene (BPFT) derivatized pest- icides on two columns are presented in Table 21. The best re- sult was obtained for dicamba (RPAs = 0.7-0.9). DNBP and monuron gave moderate sensitivities (RPAs = 0.13-0.20) and un- acceptably low sensitivities were obtained from other BPFT derivatized pesticides (RPAs = 0.03-0.09). Derivatization of DNOC, DNBP, and dicamba was attempted successfully by using diazomethane (Table 22). The relatively simple derivatization procedure (18) and clean reagent blank A make it the most promising technique for the above pesticides. In addition, methylated derivatives of these compounds gave relatively high sensitivities, linear analytical curves, and good separation of GLC peaks (Figure 19 and Table 22). Retention times and responses relative to aldrin of selected derivatization techniques for twelve multi-class pesticides are compiled from previous results in Table 23. Among the three selected derivatization techniques (PFPA, DAM and PFBC), PFPA was found to be the most suitable derivatization reagent for most of the tested carbamate pesticides (bux, car- bofuran, IPC, Landrin -2,3,5, Landrin -3,4,5 and Zectran). 62 ------- w 100 . 80 • 60 40 20 1.5% OV-17/1.95% OV-210, 180°C, EC DETECTOR, ATT. 2 x 102. 0.1 0.2 0.3 NG, PFBC-AMITROLE FIGURE 16 0.4 0.5 Analytical curve of PFBC-Amitrole ------- 240. 200. 1.5% OV-17/1.95% OV-210, 230°C, EC DETECTOR, ATT. 2 x 102 w W 160 120' 80' 40- 1234 NG PFBC-BENOMYL FIGURE 17 Analytical curve of PFBC-Benomyl ------- Ol 60 50 40- s 30 W 20 10. 4% SE-30/6% OV-210, 205°C, EC DETECTOR, ATT. = 4 x 102 0.1 0.2 0.3 NG PFBC-MONURON 0.4 FIGURE 18 Analytical curve of PFBC-Monuron 0.5 ------- 120 H ac w ac 4% SE-30/6% OV-210, 205°C, EC DETECTOR, ATT. 4 x 10 0.01 0.02 0.03 > 0.04 NG DICAMBA (X2 = NG, DNOC OR DNBP) FIGURE 19 0.05 Analytical curves of Dicamba, DNOC and DNBP (Diazomethane Derivatives) ------- Table 21. Retention times and response values relative to aldrin of BPFT derivatized pesticides on two columns (Att. = 4 x 10 , Tracer 222, column temp. 195°C, EC, pH 12 buffer system). Pesticide Azodrin Dicamba DNBP DNOC 3,4,5-Landrin Monuron Methomyl Temik 4% SE-30/6% N2 flow = 6 RRT 2.800 1.560 6.175 5.350 3.975 0.740 4.025 4.025 OV-210 7 ml/min RPA 0.028 0.770 0.150 0.059 0.069 0.128 0.086 0.053 1.5% OV-17/1. N2 flow = 53 RRT 2.100 1.540 4.730 4.500 2.770 0.690 2.830 2.770 95% OV-210 ml/min RPA 0.041 0.950 0.156 0.110 0.098 0.192 0.098 0.098 67 ------- Table 22. Retention times and response values relative to aldrin of Diazomethane-derivatized pest- icides on 4% SE-30/6% OV-210 column, 205°C (Att. = 4 x 10 , N9 flow = 67 ml/min, Tracer 222, EC). £ Pesticide DNOC DNBP Dicamba RRT1 0.900 1.300 0.300 RPA2 0.230 0.190 0.610 * See footnotes of Table 2. Note: Diazomethane-derivatized pesticides which produced no GC responses were 2,3,5-Landrin, carbofuran, Zectran, and Bux. 68 ------- Table 23. Retention times and response values relative to aldrin of selected derivatization techniques on twelve multi-class pesticides. Pesticide RRT1 RPA2 Analytical Curve3 (Figure No.) PFPA Derivatization (Ref. IPC Landrin-2,3,5 Landrin-3,4,5 Bux Carbofuran Zectran 0.142 0.321 0.420 0.43, 0.54 0.462 0.530 I)4 0. 286 0.100 0.290 0.142, 0.047 0.148 0.105 15 16 16 16 16 16 DAM Derivatization (Ref. 5)4 Dicamba 0.300 0.368 14 DNOC 0.900 0.230 14 DNBP 1.300 0.190 ' 14 PFBC Derivatization, pyridine system (Ref. 5)4 Amitrole Monuron Benomyl 0.522 2.980 12.500 0.072 0.730 0.033 17 19 18 162 q See footnotes of Table 2. 3 See Figures for GLC operating conditons. 4 PFPA = Pentafluoropropionic anhydride; DAM = Diazomethane; PFBC = Pentafluorobenzoyl chloride. 69 ------- PFPA-derivatives of the above carbamates gave relatively high sensitivities with the EC detector (RPA between 0.10 and 0.29, Table 23) and also demonstrated a high degree of reproducibility along with relatively linear analytical curves (Figures 14 and 15). A disadvantage of PFPA-derivatized pesticides over other methods (DAM and PFBC) was their relatively short retention times (RRT = 0.14-0.53, Table 23), and they were eluted from GLC columns too close together, although they were slightly dif- ferent in RRT and could ultimately be differentiated from each other. Problems were encountered when these carbamate pest- icides were studied for low level (0.1 and 0.01 ppm) soil re- covery; this will be discussed in detail later. Negative derivatization results were obtained with the fol- lowing reagents: 9-chloromethylanthracene (CMA), acetic acid N-hydroxy succinimide ester (AAHSE), N-methyl-bis-(heptafluor- obutyryl) amide (MHFBA), and N-methyl-bis-(trifluoroacetamide) (MBTFA). « GLC Detection of Sulfur-Containing Pesticides None of the tested reagents produced suitable derivatives of Temik or methomyl for either the sulfur mode of the FPD or the EC detector. However, the non-specific methylamine deriva- tization of Temik and methomyl by DBSC was possible as pre- viously mentioned. Figure 20 presents analytical curves of Temik and methomyl which were obtained by the utilization of ir> the multi-residue, 4% SE-30/6% OV-210 (130°C) column and de- tected by the sulfur mode of the FPD. Lack of sensitivity on the above column led to the selection of a_ second rather polar 70 ------- column (24), 5% Carbowax 20 M, (Figure 21), on which sensitiv- ities of Temik and methomyl could be increased two and fifteen fold respectively. The analytical scheme was based on the ther- mal degradation of Temik or methomyl to nitrile derivatives at an injection port temperature of 300°C (24, 26). Sulfur dioxide added to the hydrogen gas of the FPD increased the sensitivity of Temik and methomyl by a factor of 5 (Figure 21) . Zehner and Simonaitis (25) claimed an 8-fold increase in the sensi- tivity of Temik and malathion when S02 was introduced. Loss of Pesticides in the Concentration Steps (Kuderna-Danish concentrator or evaporation of solvents under room temperature in hood) The stability and recovery of nitrogen-containing pesti- cides in the Kuderna-Danish concentrator step were briefly in- vestigated (Table 24) before any soil residue studies were undertaken. Only captan and methomyl were lost in unaccep- table amounts (73 and 84% respectively) in the Kuderna-Danish concentration step. The interaction between methanol and cap- tan under elevated temperature was believed to be the major factor contributing to the disappearance of captan. Surpris- ingly, methomyl could be recovered quantitatively after it was heated at 95°C in 1:1 benzene:methanol for 30 min (see footnote of Table 24) in a sealed test tube. Paraffin oil acted as a useful residue keeper as demon- strated in Table 25. Thirty-five and 22 percent of Avadex could be kept by 1 ml of 1% paraffin oil in benzene when the dry res- idue was subjected to continuous evaporation under a hood for 15 min and 16 hr respectively. Fifty percent of Zectran could 71 ------- --J tsl E O I-H W 160 - 140 - 120 • 100 • 80 60 40 201 100 200 300 400 500 600 NG, TEMIK, w/ and w/o S02> NG METHOMYL w/ S02 (xlO = NG, METHOMYL w/o S02) FIGURE 20 Analytical curves of Temik and Methomyl. SE-30/6% OV-210ifi130°C, ATT. = 8 x 10 (FPD, SULFUR MODE) TEMIK w/ SO METHOMYL w/o SO TEMIK w/o S02 METHOMYL w/ S02 ------- E-" OS 200 160 -J 120 £ 80 J 40 • 5% Carbowax 20M, ATT. 16 x 10 (FPD, SULFUR MODE) -9 METHOMYL (col. 200°C, w/S02) TEMIK (col. 120°C w/S02) TEMIK (col. 120°C, w/o S02) METHOMYL (col. 200°C, w/o S02) 50 100 150 200 250 300 350 400 NG, TEMIK OR METHOMYL FIGURE 21 Analytical curves of Temik and Methorny1. ------- be preserved by paraffin oil even when its dry residue was sub- jected to continued evaporation under a hood for 48 hr. Table 26 and Figure 22 present the degradation of seven carbamate and substituted urea pesticides in benzene/methanol (1:1) for 6 hr at elevated temperature (100°). The loss ranged from 15% for Landrin - 3,4,5 to 83% for carbofuran. Silica gel column chromatography Pesticide standards were studied for recoveries and elution patterns from a silica gel column prior to soil recovery studies (Figures 23 a and 23 b). Halogenated and organophosphate pest- icides outlined in Figure 23 a were recovered quantitatively (79-105%) in a single fraction. Recoveries of 79 to 90% were obtained for endosulfan, Avadex, Dursban, phorate, methyl para- thion, parathion, CIPC, diazinon, demeton, dimethoate, phospham- idon, and Azodrin. PCNB, trifluralin, Perthane, methoxychlor, captan, folpet, Difolatan, chlorobenzilate, malathion, azinphos methyl, azinphos ethyl, atrazine, simazine and bromacil were recovered at 90 to 105%. Elution patterns of multi-class pesticides are outlined in Figure 23 b. Pesticides which could be recovered from the column quantitatively (over 75%) in a single fraction were DNBP, benomyl, DNOC, bux, IPC, Zectran, carbofuran, monuron, and amitrole. Pesticides which could be recovered quantitatively (over 75%) from more than one elution fraction were 2,3,5-Landrin (46% fraction III; 38% fraction IV), and 3,4,5-Landrin (62% fraction III; 20% fraction IV). The dicamba standard was lost on the silica gel column; however, DAM dicamba could be eluted 74 ------- , m Table 24. Percent recovery of 500 yg of pesticides each in the Kuderna-Danish concentrator step concentrated from 250 ml 1:1 benzene/methanol to 10 ml. Pesticide Avadex Captan Bux Carbofuran Zectran Monuron Temik ? Methomyl Detection Techniques EC, EC, EC, EC, EC, EC, FPD, FPD, non-derivatized non-derivatized PFPA4 PFPA4 PFPA4 PFBC S-mode S-mode % Recovery 85 27 95 97 93 96 103 16 Sample of avadex standard (3 g) in 1 ml of 1:1 benzene/methanol was heated at 95° C for 48 hrs in a sealed Teflon-capped test tube, 86% of the heated Avadex was recovered; while samples of captan, folpet and Difolatan underwent the same treatment, no recoverable residues were observed. 2 Methomyl was recovered quantitatively when it was heated at 95°C for 30 min in a sealed test tube in 1:1 benzene/methanol. 3 See Figure 21 for GLC operating conditions. 4 Samples of pesticides should be free from methanol before PFPA derivatization, since methanol inhibited PFPA-pesticide reaction. 75 ------- Table 25. Effect of paraffin oil keeper on pesticide reten- tion. Pesticide Time after complete % pesticide recovered (5 yg) dryness of residue paraffin ^ para££ijl2 Avadex 15 min 65 100 Avadex 16 hr 22 44 Zect'ran 48 hr 50 100 Avadex was selected as a model pesticide because it is the most volatile pesticide under investigation. 2 One ml of 1% paraffin oil in benzene was added to each sample, No interfering GLC peaks for paraffin oil were observed. 76 ------- Table 26. Recovery of 100 yg of nitrogen containing pesticides after heating for 6 hr (100°C) in a capped test tube with 2 ml benzene/methanol (1:1). Pesticide IPC Landrin, 2,3,5 Bux Carbofuran Landrin, 3,4,5 Zectran Monuron RRT 0.18 0.35 0.40, 0.45 0.41 0.42 0.51 0.55 2 % recovery 74.0 70.5 28.0 17.3 85.0 51.5 48.0 LGLC conditions: 4% SE-30/6% OV-210, 205°C, Att. = 2 x 10 , Tracer 222, EC. 'Average of duplicate samples. 77 ------- (a) 9,23,26 STD (PFPA) 0.04 ng ea, (b) A6 hr. 100°C in 1:1 benzene/MeOH ) 5 10 15 20 MIN. FIGURE 22 Chromatograms of standard PFPA-deriv- atives of six carbamate pesticides (a) and after they were heated at 100°C in a sealed test tube for 6 hr in 1:1 benzene/MeOH (b). Conditions as in Table 26. 78 ------- quantitatively (85%) in fraction II. PFBC-benomyl could be re- covered in fraction II (801) and fraction III (15%), and PFBC- amitrole was eluted in fraction IV (50%) and fraction V (48%). Interference due to the PFBC reagent were removed by the silica gel column, The major percentage (-70%) of the unreacted PFBC reagent was observed in fraction I and some (~25%) was carried over to fraction II. All residue keeper (paraffin oil) was eluted in fraction I from the silica gel column. Fairly good yield (84%) of methomyl was observed while a lower yield (66%) was obtained for Temik. Co-extractive interferences from crude soil extracts and clean- up by silica gel column chromatography. Most gas chromatograms were derived from the multiple re- coveries of those forty pesticides (Table 1) in a single sandy soil sample (0.01 ppm) by Soxhlet extraction. Relatively low sensitivity pesticides which could not be presented in the same chromatogram because of interferences from other more sensitive pesticides are illustrated by broken line peaks. 1. Halogenated pesticides (EC). Chromatograms for a crude control sandy soil extract and soil spiked with fourteen EC-sensitive pesticides in five silica gel fractions and their respective controls are presented in Figures 24a-24f. Figure 24a presents the background interference of crude benzene/methanol sandy soil extract for 10 mg soil equivalent. Complete recovery of 0.01 ppm PCNB from sandy soil is seen in Figure 24b (Fraction I, Hexane). In fraction II, Avadex, Trifluralin, methoxychlor, and endosulfan were quantitatively 79 ------- Figure 23a. Outline of Halogenated and Orgariop'hosphate Pesticide Standards (1 mg each) eluted from the silica gel column (figures immediately after pesticide indicate percent recoveries). Fraction I Hexane PCNB (94%) Fraction II 60% Benzene/Hexane trifluralin (105%) Dursban (85%) Perthane (97%) phorate (79%) methoxychlor (92%) methyl parathion (88%) endosulfan (90%) parathion (87%) Avadex (79%) Fraction III 5% Acetonitrile/Benzene CIPC (90%) diazinon (82%) Silica gel captan (102%) malathion (95%) (16 g) folpet (98.5%) azinphos methyl (100%) Na7SO. flj ,4 Difolatan (100%) azinphos ethyl (94%) chlorobenzilate (91%) i Fraction IV 10% Acetonitrile/Benzene atrazine (100%) demeton (80%) simazine (100%) Fraction V 20% Acetonitrile/Benzene bromacil (98%) dimethoate (81%) phosphamidon (79%) Fraction VI Acetonitrile Azodrin (90%) 80 ------- Figure 23b. Outline of Multiclass pesticide standard (1 mg) eluted from silica gel column (figures immedi- ately after pesticide indicate percent recov- eries) . Fraction I Hexane No pesticides (paraffin oil) Fraction II 60% Benzene/Hexane DNBP (83%) PFBC-benomyl (80%) DAM-dicamba (85%) Fraction III 5% Acetonitrile/Benzene benomyl (75%) PFBC-benomyl (15%) DNOC (83%) bux (95%) 2,5,4-Landrin (46%) IPC (92%) Silica gel 3,4,5-Landrin (62%) Zectran (90%) (16 g) Fraction IV 10% Acetonitrile/Benzene Na SO H4 „} 2,5,5-Landrin (38%) Carbofuran (88%) 3,4,5-Landrin (20%) PFBC-amitrole (50%) Temik (66%) Fraction V 20% Acetonitrile/Benzene methomyl (84%) PFBC-amitrole (48%) monuron (95%) Fraction VI Acetonitrile amitrole (100%) Unrecoverable dicamba Some pesticides-were eluted in more than one fraction (under- lined) ; addition of those fractions represents total recov- eries; recovery values were estimated by single analysis of pesticide standards without addition of soil. 81 ------- recovered from the. sandy soil sample (91=110%) (Figure 24c) - t The chromatogram for Perthane was obtained from the recovery study of the individual compound because endosulfan, when in the same soil sample, masked the less sensitive Perthane peak. ' In fraction III captan, folpet, and Difolatan were re- covered quantitatively (76-89%) using the Polytron extractor with benzene/methanol (1:1) as the extracting solvent (Figure 24d). CIPC, fortified at 0.10 ppm, was partially obscured by one sandy soil coextractive peak which prevented accurate es- 4 timation. Complete recovery (100%) was obtained for chloro- benzilate added to a sandy soil sample which evidenced a minor interference peak. (Figure 24d) . Simazine and atrazine were both recovered in fraction IV (Figure 24e). No GLC column under investigation could be used to separate these two compounds, although they could be recovered from soil quantitatively (83-97%). High recovery (95%) was ob- tained for bromacil in fraction V (Figure 24f). 2. Organophosphate pesticides. No interference peaks from the column blank, clean-up or crude control soil samples (500 mg soil equivalent) were observed with the FPD (Figure 25a). In fraction II (Figure 25a), recovery results were fair (66-77%) for Dursban, methyl parathion, and parathion and poor (37%) for phorate. In fractions III-VI, shown in Figures 25a and 25b, poor recovery results (40-41%) were obtained for diazinon and ~4 . demetqn. Good recoveries (76-100%) were obtained for malathion, azinphos ethyl, and azinphos methyl (Fraction III); demethoate and phosphamidon (Fraction V); and Azodrin (Fraction VI). Gas 82 ------- 00 SE-30/6% 0V-210, 205°C ATT. 2 x 10% EC 0 30 FIGURE 24 a Gas chromatogram of crude control sandy soil extract (10 mg soil equivalent). ------- FRACTION I CONTROL 34 FRACTION I PCNB, 0.01 ppm 0 5 10 15 20 MIN. FIGURE 24 b Gas chromatograms of control and 0.01 ppm PCNB from sandy soil (10 mg soil equivalent). G.C conditions as for Fig. 24 a. 84 ------- 00 en CONTROL Fraction II 0.01 ppm 10 15 20 25 30 MIN. FIGURE 24 c Gas chromatograms of 0.01 ppm in sandy soil of Avadex (3), trifluralin (59) , endosulfan (21) , methoxychlor (29), and Perthane (35) and the respective control (50 mg soil equiv- alent) . G7C. conditions as for Fig. 24 b. ------- 00 10 15 20 MIN. 25 30 FIGURE 24 d Gas chromatograms of 0.01 ppm in sandy soil of CIPC (12) captan (10), folpet (22) , chlorobenzilate (11) and DiTo- latan (15J and the respective control (10 mg soil equiv- alent) . G.C. conditions as for Fig. 24 a. ------- 00 Fraction IV, 0.01 ppm 0 5 10 15 20 25 30 MIN. FIGURE 24 e Gas Chromatograms of 0.01 ppm in sandy soil of atrazine (2) and simazine (57) and the respective control (50 mg soil equivalent). G.C. conditions as for Fig. 24 a. ------- oo 00 CONTROL Fraction V, 0.01 ppm MIN. FIGURE 24 £ Gas chromatograms of 0.01 ppm in sandy soil of bromacil (8J and the respective control (50 mg soil equivalent). G.C. conditions same as for Fig. 24 a. ------- chromatograms of 12 OP pesticide standards are presented in Figures 25c and 25d. 3. Derivatization of pesticides. Chromatograms of DAM- control crude sandy soil sample (2 mg soil equivalent) and of 0.01 ppm of dicamba, DNOC and DNBP extracted from soil are presented in Figure 26. Recoveries were 59%, 73% and 103% re- spectively. Figure 27 (a-d) presents the chromatograms of 0.1 or 1 ppm of six carbamate pesticides recovered from sandy soil samples by the Polytron extractor and clean up by a silica gel column. Figures 27a-27c present the chromatograms and recoveries of some carbamate pesticides eluted in fraction III (5% aceton- itrile/benzene). Figure 27d represents the chromatograms for control soil and soil fortified with 0.1 ppm and eluted in fraction IV (20% acetonitrile/benzene). High background inter- ferences were observed for all PFPA-derivatized carbamate pest- icides. Chromatograms of PFBC-derivatized monuron and benomyl re- covered from sandy soil are presented in Figures 28a and 28b respectively. Monuron (0.01 ppm) was recovered in relatively low yield (50%) whereas benomyl could not be detected below the 0.1 ppm level due to coextractive interference (Figure 28b), although 90% recovery was obtained at the 0.1 ppm level. Recoveries of Pesticides from Soils by Soxhlet Extraction Soxhlet extraction for 12 hr exhibited superior efficiency to 1 hr tumbling for recovery of phorate and parathion from soil (Table 27). Therefore, most of the recovery studies were carried 89 ------- CONTROL (CRUDE EXTRACT) Fraction II. 0.01 ppm Fraction III. 0.01 ppm 0 30 35 MIN. FIGURE 25 a Top, control (500 mg soil equivalent) middle, chromatogram of 0.01 ppm OP pesticides in fraction II (60% B/H) and, bottom, fraction III (5% A/B). Peaks are for Phorate (36), Dursban (20), methyl parathion (30), para- thion (32), diazinon (14), malathion (27), azinphos methyl (5), and azin- phos ethyl (4) for Fig. 24 a, G.C. conditions as 90 ------- to F. IV. 0.01 ppm F. IV. 0.01 ppm F. IV. 0.01 ppm 5 10 15 20 MIN. FIGURE 25 b Chromatograms of 0.01 ppm OP pesticide in fraction IV (10% A/B), fraction V (20% A/B) and fraction VI (100% A/B). Peaks are demeton (13), dimethoate (17) phosphamidon (33), and Azodrin (6). G.C. conditions as in Fig. 24a. 91 ------- NJ 5 ng ea. SE 30/6% 0V 210, 200°C ,-9 ATT. 15 x 10 FPD(P) 0 5 10 15 20 25 30 MIN. FIGURE 25 c Gas chromatograms of 6 FPD (P)-sensitive pesticide standards. Peaks are demeton (13), phorate (36), Azodrin (6), methyl parathion (30), parathion (32), and azinphos ethyl (5). ------- 10 5. ng ea. 4% SE 30/61 0V 210, 200°C ,-9 ATT. 16 x 10 FPD(P) o f—I »\ •^h 30 35 10 15 20 MIN. FIGURE 25 d Gas chromatograms of 6 FPD (P)-sensitive pesticide standards. Peaks are diazinon (14), dimethoate (17), Dursban (20), malathion (27), phosphamidon (33), and azinphos ethyl (4). ------- out by the Soxhlet extraction method using 1:1 benzene/methanol as the extracting solvent. For those compounds unrecoverable by the above method, alternative techniques were used (Polytron), and their conditions were noted. Soil samples were screened, air-dried and characterized (Table 28). Air dried soils were adjusted to have 10% moisture content before the addition of pesticide standards for subsequent recovery studies. i »1. Halogenated pesticides. Table 29 shows that recoveries were above 70% for all compounds insiltyloam soil except bro- macil (63% at 1 ppm), endosulfan (68% at 0.1 ppm) and triflur- alin (50% at 0.01 ppm and 67% at 0.1 ppm). Captan, folpet and Difolatan reacted with methanol during Soxhlet extraction; this phenomena was proven by simply re- fluxing these compounds in methanol for one hour. After this treatment, no captan, folpet or Difolatan peaks could be detected. When benzene-tumbling was substituted for Soxhlet k «• extraction of these compounds, the recoveries were found to be 70 to 100%. In sandy soil (Table 30), some low recovery results were obtaimed from soil samples for bromacil (61.3% at 0.1 ppm and 56% at 1 ppm), endosulfan (63% at 0.1 ppm and 1 ppm), and tri- fluralin (59.3, 50.0, 52.5% at 0.01, 0.10 and 1.0 ppm respec- tively) . In sandy loam soil (Table 31), only trifluralin gave low recovery results for 0.01, 0.1 and 1 ppm (47-49%). The consistently low recovery of trifluralin from soils when the pesticide was present below 10 ppm might be an indication that trifluralin was tightly bound by soil particles. When the soil 94 ------- tn y^ww—>v_^ CONTROL 4% SE-30/6% OV-210, 205°C, Att.=2xlO' EC, 0.01 ppm 0.02 ng ea. 10 15 20 25 30 MIN. FIGURE 26 Chromatograms of control crude sandy soil (2 mg soil equivalent) and recoveries of 0.01 ppm of dicamba (15), DNOC (18) , and DNBP (19) by diazomethane derivatization. ------- FRACTION III 4% SE-30/6% OV-210, ,2 Att. 4x EC. 180°C CONTROL FIGURE 27 a Chromatograms, left, of 1 ppm Zectran (4^T) , and the respective control, right, 1 ppm IPC (24) and its control (1 mg soil equivalent), PFPA derivatives. ------- FRACTION III CONTROL 41 SE-30/6% OV-210, Att. 4xl02. EC 205°C BUX, 0.1 ppm 0 5 10 15 20 25 30 MIN. FIGURE 27 b Chromatograms of 0.1 ppm bux (9) and its respective con- trol (1 mg soil equivalent), PFPA derivative. ------- FRACTION III CONTROL LANDRIN, 2,3,5 1 ppm LANDRIN, 3,4,5 1 ppm 0 5 10 15 20 MIN. FIGURE 27 c Gas chromatograms of Landrin (2,3,5 and 3,4,5 isomers) (25, 26) and the re- spective control (1 mg soil equivalent), PFPA derivatives. G.C. conditions as in Fig. 27 b. 98 ------- CONTROL (PFPA) FRACTION IV CARBOFURAN 0.1 ppm 0 5 10 15 20 MIN. FIGURE 27 d Chromatograms of 0.1 ppm carbofuran and its control (1 mg soil equivalent), PFPA derivative. G.C. conditions as in Fig. 27 b. 99 ------- FRACTION V 0.01 ppm MONURON 5 10 15 20 MIN. FIGURE 28 a Chromatograms of PFBC-control (5 mg soil equivalent) and recovery (83%) of 0.01 ppm monuron from sandy soil G.C. conditions as in Fig. 27 b. 100 ------- CONTROL FRACTION III 0.1 ppm BENOMYL 0 10 15 20 MIN. FIGURE 28 b 25 30 Chromatograms of PFBC-control (10 mg soil equivalent) and recovery (80%) of 0.1 ppm benomyl fronusandy soil (1.5* OV-17/1.95 OV-210, 220°C Att. 2 x 10 . ------- Table 27. Average percent recoveries of phorate and para- thion from sandy soil which were spiked at 0.01, 0.10, 1.00 and 10.00 ppm. Comparison of tumbling and soxhlet extraction methods (without cleanup, FPD). % recovery of pesticide from ppm spiked Tumbling Phorate Parathion Soxhlet Phorate Paration 0.01 ppm 0.10 ppm 86.6 46.6 81.0 71.3 96.6 94.0 93.3 99.0 1.00 ppm 10 ppm 28.0 62.7 37.6 67.3 69.0 90.3 73.6 90.6 Average percent recovery of triplicate samples. 102 ------- particles were treated wi,th trifluralin at 10 ppm, it was re- covered from soil more readily (70.5-91.5% from all types of soils). The low recoveries of bromacil and endosulfan from sandy soil (Table 30) might be attributed to the higher per- centage of organic carbon in the s^ndy soil. 2. Organophosphate pesticides. Average percent recov- eries of seven organophosphate pesticides from crude sandy and sandy loam soil extracts are presented in Tables 32 and 33 re- spectively. Considerably lower recoveries were obtained for demeton treated sandy loam soil (68-82%) and sandy soil samples (60-73.3%). Average percent recoveries of other pesticides ranged from 73.3% (1 ppm malathion) to 103% (0.10 ppm azinphos ethyl) for sandy loam soil and 67% (0.01 ppm azinphos ethyl) to 102.6% (0.01 ppm methyl parathion) for sandy soil. When crude silty loam soil extracts of organophosphate pesticides were cleaned up by silica gel columns (Table 34), there were no significantly different recovery results than those non-cleaned up soil extracts (Tables 32 and 33). Therefore, it might be concluded that silica gel retained negligible amounts of the tested organophosphate pesticides. No major difference of per- cent recoveries from various soil types was observed (Tables 32- 34). Recoveries of some Organophosphate and Thermally Labile Nitrogen- Containing Pesticides from Soil by Polytron Extraction and Com- parison of Results with Soxhlet Extraction. 103 ------- Table 28. Soil Characterization' Class and particle size distribution (%) Sand (2-0.05 mm) very coarse (2-1 mm) coarse (1-0.5 mm) medium (0.5-0.2'S mm) fine (0.25-0.10 mm)_ very fine (0.10-0.05 mm) Silt (0.05-0.002 mm) Clay (0.002 mm) Sandy 84.0 0.1 3.7 31.0 42.5 6.7 12.0 4.0 Soil Type2 Sandy Loam 92.6 0.1 4.8 38.8 41.0 7.9 1.5 5.9 Silt Loam 54.2 0.0 0.1 1.5 33.0 19.6 24.3 19.9 Organic Carbon (percent) 2.64 0.23 0.33 pH H20 (1:1) CaCl2 (0.01 M, 1:2) KC1 (1 N, 1:1) 4.8 3.7 3.5 5.2 4.4 4.2 4.7 4.3 3.5 Soil samples were analyzed by Soil Characterization Laboratory, McCarty Hall, IFAS, University of Florida 'Sandy and sandy loam soil were collected from Gainesville, Florida, silt loam soil was provided by EPA. 104 ------- Table 29. Average percent recoveries of halogenated pesticides from silty loam soil which were spiked at 0.01, 0.10, 1.00, and 10.00 ppm [Soxhlet extraction except where otherwiase indicated, with silica gel column clean- up, EC detector]. Pesticide Atrazine Bromacil Captan Chlorobenzilate CIPC Difolatan3 Endosulfan Folpet3 Methoxychlor PCNB Perthane Simazine Trifluralin % recovery 0.01 ppm 87.3 79.7 100.0 ND2 84.0 82.0 94.7 100.0 108.0 73.0 85.0 95.0 50.0 of pesticide 0,10 ppm 79.0 85.8 70.3 90.3 70.3 80.3 68.0 92.7 86.3 84.7 87.0 82.3 67.0 from ppm 1.00 ppm 90.0 63.0 87.0 92.7 77.0 79.7 95.0 96.0 96.2 85.0 101.7 86.0 83.7 spiked 10 ppm 89.3 76.3 75.0 76.3 92.0 80.0 104.0 85.0 95.2 88.7 85.3 91.3 71.3 Average percent recovery of triplicate samples. 'ND = non-detectable. These pesticides reacted with methanol when 1:1 Benzene/Methanol was used for Soxhlet extraction; therefore, 200 ml of Benzene was used with tumbling extraction for the spiked soil samples for 1 hour. 105 ------- Table 30. Average percent recoveries of halogenated pesticides from sandy soil which were spiked at 0.01, 0.10, 1.00 and 10.00 ppm [Soxhlet extraction with silica gel column cleanup, EC detector]. % recovery of pesticide from ppm spiked -v Pesticide Atrazine Bromacil Chlorobenzilate CIPC Endosulfan Methoxychlor PCNB Perthane Simazine Trifluralin 0.01 ppm ND 95.7 100.0 ND 78.3 82.7 69.0 77.3 91.0 .? 59.3 0.10 ppm 60.0 61.3 85.7 92.7 63.0 76.3 73.0 71.7 77.5 50.0 1.00 ppm 84.7 56.0 91.0 82.3 63.0 78.7 88.0 84.7 110.0 52.5 10 ppm 92.7 74.5 97.2 99.7 69.0 84.0 95.5 94.5 92.3 70.5 Average percent recovery of triplicate samples. 106 ------- Table 31. Average percent recoveries of halogenated pesticides from sandy loam soil which were spiked at 0.01, 0.10, 1.00, and 10.00 ppm (Soxhlet extraction with silica gel column cleanup, EC detector). % recovery of pesticide from ppm spiked Pesticide Atrazine Bromacil Chlorobenzilate CIPC Endosulfan Methoxychlor PCNB Perthane Simazine Trifluralin 0.01 ppm 83.7 96.3 99.7 106.0 81.3 99.0 70.0 96.0 97.0 47.0 0.10 ppm 72.7 88.0 93.3 88.0 77.0 83.3 76.0 75.7 79.0 54.0 1.00 ppm 92.3 91.0 96.0 61.0 83.0 89.7 77.0 77.0 87.2 49.7 10 ppm 95.0 93.0 80.5 83.0 86.0 91.0 92.5 87.5 97.0 91.5 Average percent recovery of triplicate samples. 107 ------- Carbofuran was selected for use in evaluating a number of extraction parameters which included solvent type, ultrasonic power, and extraction time. Benzene-methanol (1:1) was found to be the best extracting solvent (Table 35). No significant dif- ference was observed in the extraction of carbofuran from sandy soil by varying extraction time (30 to 90 sec. Table 36). How- ever, the best recovery of carbofuran was obtained when a max- imum motor speed of the Polytron was applied (Table 37). Later experience gained fr.om this investigation suggested use of medium speed, because the Polytron generator head was being ground away excessively by soil particles. Average percent recoveries of sixteen multi-class pesticides from sandy soil are presented in Table 38. Benzene-methanol (1:1) was selected as the solvent for general extraction, al- though methanol reacted with some pesticides during Soxhlet ex- traction (Table 39) and subsequently yielded poor or no recov- ery for carbamates, captan, folpet and Difolatan. Polytron , extraction (Table 38) apparently overcame the problem and gave reasonably good recoveries for those pesticides at 0.01-0.1 ppm levels (captan, 64-76%; Difolatan, 89-100%; folpet, 88-101%). Other pesticides in Table 38 which could be detected at 0.01- 0.1 ppm levels were: Avadex (70-75%), bux (89%), dicamba (77- 86%),,DNBP (74-75%), DNOC (60-67%), carbofuran (100%) and mon- uron (83-93%). Pesticides which could not be detected at 0.01- 0.1 ppm levels but were detectable at 1 and 10 ppm levels were benomyl (61-69%), IPC (71-73%), Landrin -2,3,5 (80-81%), Landrin -3,4,5 (73-82%) and Zectran (76-85%). The only pesticide which 108 ------- Table 32. r Average percent recoveries of organophosphate pest icides from sandy soil which were spiked at 0.01, 0.10, 1.00, and 10.00 ppm (Soxhlet extraction with out column cleanup, FPD detector). Pesticide Phorate Parathion Demeton Malathion Methyl parathion Azinphos ethyl 2 Azinphos methyl % recovery of pesticide from ppm spiked 0.01 ppm 96.6 93.3 60.0 70.0 102.6 67.0 .. 86.0 0 . 10 ppm 94.0 99.0 60.0 81.0 93.3 71.7 101.7 1.00 ppm 69.0 73.6 71.7 90.3 82.0 93.3 100.7 10 ppm 90.3 90.6 73.3 93.3 88.3 99.7 92.3 Average percent recovery of triplicate samples 'With silica gel column cleanup. 109 ------- Table 33. Average percent recoveries of organophosphate pest- icides from sandy loam soil which were spiked at 0.01, 0.10, 1.00, and 10.00 ppm (Soxhlet extraction without column cleanup, FPD detector). Pesticide Phorate ! Parathion Demeton Malathion Methyl parathion Azinphos ethyl Azinphos methyl % recovery of pesticide from ppm spike 0.01 ppm 77.3 99.0 68.0 92.6 102.6 83.3 101.6 0.10 ppm 102.6 84.6 73.0 84.0 83.0 103.0 99.3 1.00 ppm 89.0 75.3 73.3 73.3 84.0 102.0 78.3 10 ppm 80.0 81.0 82.0 80.6 72.6 92.0 83.3 Average percent recovery of triplicate samples. 110 ------- Table 34. Average percent recoveries of organophosphate pest- icides from silty loam soil which were spiked at 0.01, 0.10, 1.00, and 10.00 ppm (Soxhlet extraction, with silica gel column cleanup, FPD detector). Pesticide Phorate Parathion Demeton Malathion Methyl Parathion Azinphos ethyl Azinphos methyl % recovery 0.01 ppm 75.0 93.0 69.3 95.7 101.0 72.0 77.7 of pesticide 0.10 ppm 81.0 101.0 72.3 96.7 91.7 90.3 91.0 from ppm 1.00 ppm 75.0 85.7 69.3 96.7 93.3 86.7 89.3 spike 10 ppm 71.7 86.0 71.3 101.0 91.0 87.3 78.3 Average percent recovery of triplicate samples. Ill ------- Table 35. Comparison of solvents for ultrasonic-polytron extraction of carbofuran (1 ppm) from sandy soil. Solvent mean % recovery of duplicate samples and deviation from mean Acetone 71.5 ± 3.5 Acetonitrile 70.5 ±0.5 Methanol 83.0 ± 2.0 Hexane 24.8 ± 5.5 Benzene 74.3 ±3.8 Benzene-Methanol (1:1) 100 ± 0.0 GLC column and operating conditions: 4% SE-30/6% OV-210, 200°C, Att. = 4 x 10 (PFPA derivatization, EC). Soil and solvent: 50 g air-dried sandy soil extracted by 100 ml x 2 of solvent and washed with 50 ml x 2. Polytron: 1 min x 2, maximum speed. 112 ------- could not be detected even at the 10 ppm level, was amitrole. Due to high water solubility of amitrole, most conventional solvents used for extraction were water or an ammonium hydroxide- glycol (5+20) mixture (26). Groves and Chough (26) suggested that polar ppmpounds like amitrole may be bound in the clay- organip matter ion exchange systems similar to the binding of nutrient ions; that entrapment between atom-layers of expandable lattice clays may occur; that insoluble compounds or complexes may be formed; and that adsorption may occur due to electro- static or residual valence forces. The undetectability of benomyl at low residue levels (0.01-0.1 ppm) was attributed to the low sensitivity of the PFBC derivative and the high back- ground interference of the PFBC reagent as mentioned before. IPC, Landrins, and Zectran could not be detected at low levels (0.01-0.1 ppm) (Table 38) mainly due to the interference peaks of the PFPA derivatives and their short retention times which appeared in the same regions as most underivatized soil coex- tractives. The Soxhlet extraction (Table 39) procedure prevented the complete recoveries of captan, folpet, Difolatan, carbamate and substituted urea compounds due to pesticides-methanol inter- action or thermal degradation under prolonged (12 hr) elevated temperature. However, better recoveries were obtained with Soxhlet extraction than with the Polytron for the following heat stable pesticides: Avadex (73-100 vs. 70-75), dicamba (86- 97 vs. 77-86), DNBP (91-100 vs. 74-75), DNOC (99-104 vs. 60-67). 113 ------- Table 36. Effect of Polytron extraction time on recovery of carbofuran from sandy soil.* Extraction time, mean % recovery of duplicate sec samples and deviation from mean 30 x 2 71.5 ± 0.5 60 x 2 77.0 ± 1.0 90 x 2 72.0 ± 2.0 *GLC, column and conditions: same as Table 34. Soil and extraction solvent: 50 g air-dried soil extracted by 100 x 2 ml Benzene-MeOH (1:1) and washed with 50 x 2 ml. Polytron speed: 50% of maximum. 114 ------- Table 37. Effect of Polytron motor speed on recovery of carbofuran from sandy soil. Speed, % of maximum 30 50 70 100 mean % recovery of duplicate samples and deviation from mean 71.5 ± 71.5 ± 75.0 ± 100 ± 6. 6. 1. 0. 5 5 0 0 GLC, column and conditions: same as Table 34. Soil and extraction solvent: same as Table 35. Polytron extraction time: 1 min x 2. 115 ------- Table 38. Average percent recoveries of pesticides from sandy soil samples (triplicate) which were spiked at 0.01 and 0.10 ppm (Polytron extraction with silica gel column cleanup, EC detector)=l % recovery of pesticide from ppm spiked' r co i. -H-.H4C i T Amitrole Avadex 3 Benomyl Bux 4 Cap tan Dicamba Difolatan4 DNBP5 DNOC5 Carbofuran Folpet4 I PC Landrin-2,3,5 Landrin-3,4,5 Monuron Zectran 0.01 ppm (undetectable) 70.0 (69.3) undetectable 76.0 86.0 89.3 75.3 67.3 undetectable 88.3 (72.7) (79.7) (82.0) 83.0 (85.0) 0.10 ppm ' (undetectable) 75,0 (61.3) 89.5 64.3 77.0 100.0 74.0 60.0 100.0 101.0 (70.7) (81.0) (73.3) 93.0 (76.0) See Tables 21 and 22 for RRT, RPA and GLC operating conditions 'Parentheses imply undetectable at 0.01 and 0.10 ppm levels, subsequent attempts for 1 and 10 ppm were conducted and recorded in parentheses respectively. Amitrole and benomyl were derivatized before silica gel column cleanup. Solvents were evaporated by standing in a hood at room temp- erature (other pesticides by Kuderna-Danish concentrator). 'Analyzed in crude extract. 116 ------- Table 39. Average percent recoveries of pesticides from sandy soil samples (triplicate) which were spiked at 0.01, and 0.10 ppm (Soxhlet extraction, with silica gel column cleanup, EC detector). Pesticide Amitrole Avadex Benomyl Bux Captan Dicamba Difolatan DNBP DNOC Carbofuran Folpet IPC Landrin-2,3,5 Landrin-3,4,5 Monuron Zectran 1 recovery of pesticide 0.01 ppm (undetectable) 100.0 undetectable 64.3 0.0 85.5 0.0 90.7 104.3 undetectable 0.0 (60.1) (57.0) (50.3) 50.3 30.0 from ppm spiked 0.10 ppm (undetectable) 73.3 85.0 39.3 0.0 97.0 0.0 100.0 99.2 48.3 0.0 (65.0) (63.3) (64.3) 37.0 33.3 117 ------- In the Soxhlet extraction, benomyl was converted to Methyl N-(2- benzimidazolyl) carbamate (MBC) and 2-aminobenzimidazole (2-AB) more readily than in the Polytron procedure. The better recoveries with the Soxhlet procedure (85% at 0.1 ppm vs. 61 69% at 1-10 ppm) reflected the detection of the degradation products of benomyl, i.e., PFBC-MBC and PFBC-2AB. Average percent recoveries of five organophosphate pest- icides by Polytron and Soxhlet extraction from sandy soil are presented in Tables 40 and 41 respectively. High percent re- coveries were obtained for most pesticides (82-103%) by the Polytron extraction except for methomyl and Temik (Table 40). The inability to recover methomyl was partly due to loss during Kuderna-Danish evaporation, whereas the disappearance of Temik might be due to its strong binding with soil particles. There was no evidence of any breakdown or alteration of Temik or other pesticides by the Polytron extraction which agrees with the findings of Johnsen and Starr (28). The above indefinite conclusion led us to further inves- tigate the disappearance of these two compounds. No accept- able recovery results could be obtained for these compounds by room temperature solvent evaporation or by mixing the con- centrated control sandy soil sample with Temik and methomyl standards. Pesticide standard peaks also disappeared when they were injected simultaneously from the same syringe to- gether with the same amount of control soil extract. All the above evidence led us to believe that carbamoyl pesticides reacted with soil coextractives to produce other compounds under elevated temperature in the GLC system. 118 ------- Simultaneous analysis of forty multi-class pesticides in sandy soil Duplicate sandy soil samples were spiked with 0, 0.01 and 0.1 ppm of the forty multi-class pesticides listed in Table 1 and analyzed for their recoveries simultaneously. Table 42 presents the results of fourteen EC-sensitive pesticides. Ex- cellent recoveries were obtained for most pesticides at 0.01 ppm level (90-116%) except for Perthane, chlorobenzilate, captan, folpet and Difolatan. Perthane was masked by the more responsive endosulfan while chlorobenzilate suffered from in- terference by coextractives from sandy soil. Captan, folpet and Difolatan reacted with methanol during the Soxhlet extrac- tion procedure. At 0.1 ppm, all the halogenated pesticides gave excellent recoveries (86-115%) except for atrazine and simazine (72%). Data in the last column of Tables 42, 43, and 44 are derived from 0.1 ppm equivalent of pesticides (without soil) and underwent the entire procedure used in the extraction of pesticides from soil samples. The original purpose was to determine whether low recovery results were caused by the ana- lytical procedure itself or because of the soil binding. No endosulfan could be recovered when it was heated in the Soxhlet extractor without soil for 12 hours in contrast to 95% recovery of the same amount of pesticide spiked in soil (0.1 ppm). Con- siderably lower recoveries were obtained from non-soil samples as compared to their corresponding soil samples, i.e., PCNB (87 vs. 93%), trifluralin (73 vs. 86%), chlorobenzilate (80 vs. 119 ------- Table 40. Average percent recoveries of pesticides from sandy soil samples (triplicate) which were spiked at 0.01 and 0.10 ppm (Polytron extrac- tion with silica gel column cleanup, FPD; P- mode). 4 Pesticide Azodrin Diazinon Dimethoate Dursban Met homy 1 Phosphamidon Temik3 % recovery of pesticide 0.01 ppm 83.0 93.3 102.7 94.0 (undetectable) 88.7 (undetectable) from ppm spiked 0.10 ppm 81.7 87-0 98.3 96.7 (0.0) 90.3 (13.0) See Table 9 for RRT, RPA and GLC operating conditions. 7 See footnote of Table 38 for explanation. FPD operated on S-mode, see Figure 21 for analytical curves. Solvents were evaporated by Kuderna-Danish concentrator. 120 ------- Table 41. Average, percent recoveries of pesticides from sandy soil samples (triplicate) which were spiked at 0.01 and 0.10 ppm (Soxhlet extrac- tion with silica gel column cleanup, FPD; P- mode). Pesticide Azodrin Diazinon Dimethoate Dursban Methomyl Phosphamidon Temik % recovery of pesticide 0.01 ppm 91.0 61.0 91.0 66.0 0.0 76.0 0.0 from ppm spiked 0.10 ppm 98.0 99.5 102.2 104,0 0.0 100.0 0.0 121 ------- Table 42. Multi-residue analysis of fourteen EC-sensitive pesticides in duplicate sandy soil samples which were spiked at 0.01, and 0.1 ppm of forty multi- class pesticides in Table 1 (Soxhlet extraction, 4% SE-30/6% OV-210, 205°C, Att. = 2 x 10Z, N2 flow = 67 ml/min, Tracor 222, EC). Pesticides eluted in % recoveries from ppm spiked silica *el fractions 0.01 ppm " 0.10 ppm (0.10 ppm)1 Hexane PCNB 116 93 87 60% Benzene/Hexane Avadex 110 115 107 Perthane undetectable 93 80 Trifluralin 91 86 73 Methoxychlor 107 88 88 Endosulfan 94 95 O2 5% Acetonitrile/benzene CIPC " undetectable Captan underwent structural transformation Folpet in the Soxhlet extracting solvents (benzene/methanol, 1:1) Difolatan Chlorobenzilate undetectable 96 80 10% Acetonitrile/benzene Atrazine 90 72 64 Simazine 122 ------- Table 42 continued. Pesticides elutedin % recoveries from ppm spiked silica gel fractions 0.01 ppm 0.10 ppm (0.10 ppm) 20% Acetpnitrile/benzene Promacil 95 93 93 0.1 ppm equivalent of pesticides without soil underwent the same extracting procedure (Soxhlet), evaporation (Kuderna-Danish) and cleanup (silica gel). 2 Loss due to absence of Keeper. 123 ------- 96%). Identical results were obtained for methoxychlor (88%) i and bromacil (93%). Table 43 presents the results of multi-residue analysis of twelve FPD-P sensitive pesticides in duplicate sandy soil samples \ which were spiked at 0, 0.01, and 0.10 ppm of forty multi-class pesticides. At 0.01 ppm, pesticides which were recovered above 70% wdre: methyl parathion (77%), parathion (76%), malathion (100%) , dimethoate (91%), phosphamidon (76%) and Azodrin (91%). Those recoveries below 70% were: phorate (37%), Dursban (66%), diazinon (41%), and demeton (40%). At 0.1 ppm, all twelve OP pesticides were recovered at more than 78% (79-100%). The 0.1 ppm of the non-soil samples (last column of Table 43) gave very poor recoveries except for parathion (71%) and Azodrin (76%). The low recoveries (5-53%) of other OP pesticides, i.e., phorate (48%), methyl parathion (37%), diazinon (45%), malathion (32%) demeton (5%) and dimethoate (53%), might be attributed to their thermal degradation on the glass surfaces of the Soxhlet ex- tractor in the absence of soil coextractives. Soil coextrac- tives might act as protective agents in preventing pesticide breakdown on the glass surface. It is already known in GLC that regular injection of sample extracts or silanizing reagents will help prevent the degradation of some relatively heat-labile compounds. 'J'able 44 presents the results of multi-residue analysis of twelve nitrogen containing pesticides in duplicate sandy soil samples which were spiked at 0, 0.01 and 0.10 ppm. At 0.01 ppm, »' good recoveries were obtained from DNOC (73%) and DNBP (103%). 124 ------- Table 43. Multi-residue analysis of twelve FPD-P sensitive pesticides in duplicate sandy soil samples which were spiked at 0.01 and 0.10 ppm of forty multi- class pesticides in Table 1 (Soxhlet extraction, 41 SE-30/6% OV-210, 200° C, Att. = 16 x 10 , N? flow = 50 ml/min, Varian 1200, FPD-P). Pesticides eluted in silica gel fractions , 60% benzene/hexane Phorate Dursban Methyl parathion Parathion 5% acetonitrile/benzene Diazinon Malathion * ).01 37 66 77 76 41 100 recoveries from ppm ppm 0.10 ppm 94 79 100 88 90 106 spiked (0.10 ppm)1 48 72 37 71 45 32 10% acetonitrile/benzene Demeton 40 86 2 20% acetonitrile/benzene Dimethoate 91 98 53 100% acetonitrile/benzene Phosphamidon 76 100 2 Azodrin 91 100 76 >2See footnotes of Table 42. Note: Temik and methomyl were unrecoverable (detected by FPD-S). 125 ------- Table 44. Multi-residue analysis of twelve nitrogen con- taining pesticides in duplicate sandy soil samples which were spiked at 0.01, and 0.10 ppm of forty multiclass pesticides in Table 1 (Sox- hlet extraction, 4% SE-30/6% OV-210, 205° C, Att. = 2 x 10Z, N9 flow = 67 ml/min, Tra< EC). L Tracer 222, Pesticides eluted in silica gel fractions % recoveries from ppm spiked 0.01 ppm 0.10 ppm (0.10 ppm) Without column cleanup Dicamba (DAM) 59 DNOC (DAM) 73 t DNBP (DAM) 103 Benomyl (PFBC) undetectable 5% Acetonitrile/benzene 2,3,5-Landrin (PFPA) 3,4,5-Landrin (PFPA) Zectran (PFPA) IPC (PFPA) Bux (PFPA) 20% Acetonitrile/benzene Carbofuran (PFPA " Monuraon (PFBC) 50 100% Acetonitrile/benzene Amitrole (PFBC) " 103 112 117 90 undetectable 18 35 100 80 undetectable 52 38 undetectable •See footnotes of Table 41. 126 ------- Poor recoveries were obtained for dicamba (59%) and monuron (50%). Pesticides which were undetectable at all concentrations were: Landrins, Zectran, IPC, bux, carbofuran, and amitrole. Benomyl was recoverable at 0.1 ppm (90%) while monuron gave poor results (52%) at this level. Excellent results were ob- tained for dicamba (103%), DNOC (112%), and DNBP (117%). Poor or no recoveries (below 38%) were observed for most of these pesticides in the non-soil samples (last column of Table 43) except for DNBP and benomyl. Conclusion Aside from a rapid deterioration of the Polytron generator rotor and bearings, it offers a real advantage in extracting multiclass pesticides from soil compared to either Soxhlet ex- traction or tumbling. More specifically, of the forty com- pounds studied, the following, which could be successfully ex- tracted from soil with the Polytron, could not be extracted with the Soxhlet due to degradation: CIPC, captan, folpet, Difolatan, carbofuran, Zectran, bux and benomyl. Amitrole could neither be extracted by Soxhlet nor Polytron. It was not determined whether methomyl or Temik could be extracted by the Polytron (see part II for confirmation of this) since they could not be gas chromatographed in the presence of soil coex- tractives. Dicamba could not be eluted from the silica gel intact, however, it could be if it were methylated prior to chromatography. DNOC also could not be eluted at residue levels, 127 ------- however, it could- be determined in the crude extract after methylation. See Fig. 29 for a summary of the procedure util izing the Polytron. 128 ------- Fraction I 50 ml hexane EC PCNB Fraction II 50 ml 60% B/H EC trifluralin Perthane methoxychlor endosulfan Avadex DNBP (DAM)3 . dicamba (DAM)1 FPD ~Dursban phorate methyl parathion parathion Forty Multiclass Pesticides in Soil (50 g) Polytron 100 ml benzene/methanol (I:l)(x2) wash with 50 ml (x2) benzene/methanol soil extract evaporate to dryness Silica gel chromatography: 16 g (14 g Na-SO,) Fraction III 50 ml 5% A/B EC CIPC captan folpet Difolatan chlorobenzilate benomyl (PEBC)2 DNOC (DAM)5 bux (PFPA),. IPC (PFPA)4 Zectran (PFPA)4 2,3,5-Landrin (PFPA)4 3,4,5-Landrin (PFPA)4 FPD diazinon malathion azinphos methyl azinphos ethyl Fraction IV 50 ml 10% A/B atrazine simazine carbofuran FPD demeton (PFPA) Fraction VI 50 ml Acetonitrile FPD Azodrin Fraction V 50 ml 20% A/B EC Erbmacil monuron (PFBC) FPD phosphamidon dimethoate Figure 29. Procedure for detection of forty pesticides benzene). in soils (A/B = Acetonitrile/ Derivative made prior to silica-gel cleanup. ^Detectable only at 1 ppm level. ^Analyzed in crude extract. 4Can only be analyzed at 1.0 ppm and higher. 5Detectable at 0.01 or 0.1 ppm unless otherwise specified. ------- References (Part I) 1. Argauer, R. J., J. Agr. Food Chem. 17_, 888 (1969). 2. Bowman, M. C., Beroza, M. J., J.A.O.A.C. 50, 926 (1967). 3. Crosby, D. G., Bowers, J. B., J. Agr. Food Chem. 16, 839, (1968). 4. Shafik, M.'l., Bradway, D., Biros, F- J., Enos, H. F., J. Agr. Food Chem. 1JJ, 1174 (1970). 5. Khalifa, S., Mumma, R. 0., J. Agr. Food Chem. 20, 632, (1972). 6. Mills, P. A., Onley, J. H., Gaither, R. A., J.A.O.A.C. 46, 186 (1963). 7. Johnson, L., J.A.O.A.C. 48, 668 (1965). 8. Porter, M. L., Young, S. J., Burke, J. A., J.A.O.A.C. 55, 1300 (1970). 9. Wessel, J. , J.A.O.A.C. 52^, 172 (1969). 10. Carr, R. L., J.A.O.A.C. S4> 525 (1971). 11. Chiba, M., Residue Reviews 50, 63 (1969). 12. Sherma, J., Shafik, T. M., Arch. Environ. Contain. Toxicol. ^, 55 (1975). 15. Moye, H. A., J. Agr. Food Chem. 23, 415 (1975). 14. Damico, J. N., Benson, W. R., J.A.O.A.C. 48, 344 (1965). 15. Fishbein, L. Zielinski, W. L., J. Chromatog. 20, 9 (1965). 16. Moye, H. A., J. Agr. Food Chem. 19, 452 (1971). 17. Kawahara, F. K., Anal. Chem. 40, 1009 (1968). 18. Yip, G., Howard, S. F., J.A.O.A.C. 51, 24 (1968). 130 ------- 19. Koshy, K. T., Kaiser, D. G., Van Der Slik, A. L., J. Chrom. Sci. 1^, 97 (1975). 20. Nash, R. G. , J.A.O.A.C. S7_, 1015 (1974). 21. Stewart, F. H. C., Aust. J. Chem. 15, 478 (1960). 22. Roth, M., Analy- Chem. 43_, 880 (1971). 23. Carey, W. F., Kenneth, H., J.A.O.A.C. 53, 1296 (1970). 24. Zehner, J. M., J. Chrom. Sci. 14, 348 (1976). 25. Knaak, J. B., Tallant, M. J., Sullivan, L. J., J. Agr. Food Chem. 14_, 573 (1966) . 26. Groves, K., Chough, K. S., J. Agr. Food Chem. 19, 840 (1971) 27- Fung, K. K. H., J. Agr. Food Chem. 23.* 695 (1975). 28. Johnsen, R. E., Starr, R. I., J. Agr. Food Chem. 20, 48 (1972). 29. Brown, B., Goodman, J. E., "High Intensity Ultrasonic" Van Nostrand, New York, N.Y., (1965) pp. 196-220. 131 ------- Extension of Multi-residue Methodology. II. Dynamic Fluorogenic Labelling Detector for Carbamates. H- Anson Moye an4 Gordon Cash Introduction Many pesticides now in common use are formally deriv- atives of methylcarbamic acid, i.e. carbamate esters or carbamoyl oximes. These pesticides are somewhat polar and extremely heat labile, making them poorly suited for con- ventional column cleanup techniques and separation by gas- liquid chromatography. Several approaches have been used to attempt to overcome these difficulties. One method is to form chemical derivatives which improve volatility and thermal stability (1,2). This method is often satisfactory if the analysis is directed toward only one compound at a time. Colorimetric procedures involving coupling with diazonium salts also have been utilized in similar situa- tions. Many reports of multi-residue methods for carba- mates have described separation by thin-layer chromatography and quantitation by various methods including chromogenisis, enzyme inhibition and, iri situ fluorometry. It would appear that high pressure liquid chromato- graphy (HPLC), since it can achieve high efficiency separ- ations at ambient temperatures could theoretically obviate 132 ------- the themal instability problem associated with the carba- mates. However, detection at the nanogram level normally required for sub-part-per-million analysis in foods and .en- vironmental samples, would seem to be unlikely with commer- cially available detectors. Preliminary experiments in this laboratory have demonstrated this to be the case that the carbamates as a class neither absorb U.V.-visible radiation nor fluoresce sufficiently (with the possible exceptions of carbaryl and benomyl) to allow for simple and straight for- ward extraction, cleanup and detection schemes. Recent success with post-column fluorogenic labelling of amino acids following a HPLC separation has offered en- couragement that this approach could be used for the anal- ysis of the carbamates in soil (3). The fluorogen used in p that study was Fluram , (4-phenylspiro [furan-2(3H), 1- phthalan]-33-dione) (4). This reagent is non fluorescent, nor is its hydrolysis product. Another promising fluorogenic reagent seemed to be o- phthalaldehyde (o-phthalic-dicarboxaldehyde) reported by Roth (5) to be extremely sensitive for primary amino acids. Like Fluram it reacts almost instantaneously, is itself non- fluroescent, nor are its hydrolytic products. Unlike Fluram, however, it is stable in protic solvents. Fluram is re- quired to be handled in aprotic solvents, such as acetone. Additionally, two other fluorogens specific for primary amines were investigated during the contract period, NBD-C1 133 ------- (4-chloro-7-nitrobenzo-2,l,3-oxadiazole) and dansyl chloride (l-dimethylaminonaphthalene-5-sulfonyl chloride). Eight commercially available, and widely used, carba- D mates were chosen for this study: Lannate (methomyl), R R R R Matacil (aminocarb), Temik (aldicarb), Baygon , Furadan R R R (carbofuran), Sevin (carbaryl), Mesurol and Zectran . Lannate and Temik are actually carbamoyl oximes but can be effictively hydrolyzed to methylamine and hence should be detectable under the proposed system. The three objectives for this HPLC portion of contract work, as specified in the contract were: (1) determine the optimum HPLC separation parameters for the above listed pesticides that would be compatible with the dynamic fluor- ogenic labelling detector, (2) determine the optimum para- meters for the dynamic fluorogenic labelling detector and (3) demonstrate the sensitivity and selectivity of the com- bined column-detector system for the analysis of the listed carbamates in soils. Additionally, a brief comparison was made between the assembled high pressure liquid chromatograph and a commer- cially available liquid chromatograph, the American Instru- p ment Company Aminalyzer (Silver Springs, MD). Experimental Apparatus and Reagents 1. Two Waters model 6000 solvent delivery systems (pumps) were programmed with a Waters model 660 solvent 134 ------- programmer (Waters Associates, Inc., Milford, Mass). 2. Sample injection valve Chromatronix HPSV with 25 ul sample loop (Spectra Physics, Berkeley, CA). 3. Columns, analytical HPLC, 30 cm x 4 mm I.D.; Waters: y Carbohydrate. 25 cm x 4 mm I.D.; pupont; Zorbax-ODS. 1 m x 4 mm I.D.: Dupont; CDS, ETH, SAX. Reeve Angel: Pellidon. Applied Sciences: Vydac Reverse Phase. 4. Kel-F "T"s (2), Laboratory Data Control (Riviera Beach, FL), model CJ3031. 5. Tubing, TeflonR, 0.062 in O.D. x 0.02 in I.D., Dixon Medical Products (Thornton, PA). Cat. No. 6-8-112. 6. Detector, U.V. absorption, 254 nm, Waters. 7. Water bath, Haake, model FK. 8. Reagent pumps (2), Milton Roy model 196. 9, Fluorometer American Instrument Co., model 125S, with model B16-63019 flow through cell and model 416-993 mercury-xenon lamp. Excitation, 315 nm; emission, 435 nm (for a-phthalaldehyde). Slits, 4 mm. Scale 1. 10. Recorder, Soltec model 211, 50 mv. 11. Solvents: scintillation grade dioxane, deionized water, pH 10 0.05 M borate buffer, 0.02 M NaOH. 12. o-phthalaldehyde (OPA o-phthalicdicarboxaldehyde), Aldrich. Stock solutions were made by dissolving 1 g. in 10 ml of ethanol adding 1 ml of mercaptoethanol and diluting with buffer to.v.olume. 13. Aminalyzer, American Instrument Co., equipped with 135 ------- a Waters yC,« column. 14. Carbamate pesticides obtained from E.P.A., Research Triangle Park, NC. HPLC Column Study Two types of columns were evaluated, reverse phase and adsorption, at either of two particle sizes, 10 y or 37 y diameter. Those 37 y reverse phase columns evaluated were: CDS, ETH, SAX, Vydac and Pellidon. The 10 y reverse phase columns were: yC,g> yCN, yNH2, yCarbohydrate and Zorbax-ODS. Only one adsorption column was evaluated, the yPorasil. No- tice that the SAX column is normally a strong.anion exchange •i column, but in this case was evaluated in the reverse phase mode. The adsorption mode for the separation of the carbamates was immediately discarded after severe tailing was noted for isooctane-methanol or isooctane-chloroform mobile phases. This could not be overcome by water saturation of the mobile phases or by deactivation of the silica by water impregnation. Much effort was expended in attempting an isocratic separation In the reverse phase mode using the 254 nm absorp- tion detector and ug amounts of carbamates. Several columns were immediately discarded as being unsatisfactory, either because no reasonable separations were achieved or because too little retention was observed: ETH, SAX, Vydac Reverse Phase and Pellidon. The Zorbax-ODS column exhibited over 5000 psig back pressure, causing the sample injection valve to 136 ------- leak and therefore was discarded as a candidate. Exhibiting about 5000 theoretical plates, the column appeared to be the most satisfactory giving adequate separation of the carbamate mix (Maticil was not included in this early work, Fig. 1), with 40% acetonitrile--water. At this percentage acetonitrile Baygon and carbofuran are not fully separated, although they can be easily quantitated; decreasing the acetonitrile percentage to 25% gave a complete separation at the baseline, their retentions being 29 and 33 minutes respectively- A flow of one ml/min was used through- out this work and was found to be nearly optimum while pro- viding for rapid analyses. The yCN column also performed well, giving about 2000 theoretical plates for a 1 foot x 4 mm I.D. column. How- ever, at high percentages of acetonitrile (>20%) it could not satisfactorily separate Baygon and carbofuran. By oper- ating this column at 12ig% acetonitrile - 87%% water Lannate, Temik, Baygon and carbofuran could be separated quite well (Fig. 2). By using the 12^% acetonitrile mobile phase the CDS column separated Sevin, Zectran and Mesurol quite well. The early eluters, while not shown (Fig. 3) elute immediately after the injection solvent. The Mesurol response shown is actually that of a conversion product produced by heating at 80°C for 15-min in ethanol; unchanged, it elutes too close to Zectran for adequate separation. 137 ------- (s) 00 W) yC1Q 1 ft. x 4 mm I.D. COLUMN 40t8ACETONITRILE - 601 WATER FLOW RATE I ml/min 0.16 A.U F.S. 0 16 MINUTES 24 32 FIG. 1 Isocratic separation of six carbamate mix on uC,g> 40% acetonitrile - 60% H20, ------- yCN 1 ft x 4 mm I.D. COLUMN 12.5% ACETONITRILE - 87.5% pH 8 BUFFER FLOW RATE 1 ml/min 0.16 A.U.F.S. t* LO 0 11 22 MINUTES FIG. 2 Isocratic separation of early eluting carbamates on yCN, 12.5% acetonitrile - 87.5% H20. ODS 1m x 4 mm I.D. COLUMN 12.5% ACETONITRILE - 87.5% pH 8 BUFFER FLOW RATE ml/min 0.16 A.U.F.S. 0 3 6 9 12 15 MINUTES FIG. 3 Isocratic separation of late eluting carbamates on ODS, 12.5% acetonitrile 87.5% H~0. 139 ------- The u NH7 and y Carbohydrate columns provided poor sep- Lf arations for all of the carbamates and were considered un- usable. Studies on Fluram Of the four fluorogenic reagents to be examined it became obvious that NBD-C1, which requires a non-aqueous re- action medium, would not be satisfactory, since separations on the adsorption column (u Porasil) were unachievable. Also, dansyl chloride was given low priority, since the reagent itself fluorcesces considerably. Of the remaining two reagents, Fluram was chosen for examination first, unfortunately so. A. Static Measurements. Since Fluram is unreactive toward secondary amines in general, it was expected that it would not react with the intact esters of methylcarbamic acid. This was verified experimentally by tests with Sevin and Lannate. Weigele ejt aJ (3), however, reported that the secondary amino acids, proline and sarcosine, were rendered reactive toward Fluram by preliminary treatment with aqueous bromine or N-chlorosuccinimide under strongly acidic con- ditions (pH 1) . By treating Sevin with N-chlorosuccinimide (or N- -3 bromosuccinimide, 10 M at 25°C for 2 hours) we were unable to observe fluorescence by adding Fluram (1.5 x 10 M) after buffering to pH9 with borate buffer. However, both Br9 t ^ and Iat 10 M could produce maximum fluorescence by incu- 140 ------- bating for 100 minutes before buffering and reacting with Fluram. Linear analytical curves for all the carbamates were obtined from 1 to 10 nanomoles/ml of final solution. Table 1 gives the lower limit of detection (LLD) in nano- moles per ml of each carbamate, assuming a signal to noise ratio of 3:1. Table 1 Carbamate LLD (nanomoles/ml) Carbofuran 0.95 Sevin 0.40 Zectran 0.30 Lannate 2.7 Temik 0.75 Baygon 1.3 Mesurol 0.71 B. Dynamic Measurements. A dynamic system was assem- bled separately from the column to simplify optimization of parameters unrelated to carbamate separation. A Harvard pulseless peristaltic pump was chosen for this early work in an effort to minimize noise from the fluorometric cell (flow through) as reagents were metered. Column effluent was simulated by passing an additional channel through the peristaltic pump. Zectran was used to optimize this system because it gave the greatest response in the static system. 141 ------- The dynamic system was assembled with a view toward simulating the manual operations performed with the static system. Unfortunately, the sensitivity of the dynamic system ne\rer approached that of the static system and was a result of a combination of thres factors: (1) since Fluram must be pumped in the aprotic solvent, actone, no fully elastic tubing could be found which did not rapidly deteriorate, (2) since rapid mixing is essential for the Fluram to provide maximum response when metered (in acetone) into the buffered solution of primary amine, no fully satis- factory mixing device could be designed which did not add considerable band spreading and, (3) since complete mixing immediately upstream of the fluorometric cell could not be accomplished, considerable noise, due to a Schlieren effect, could not be eliminated. Changing the solvent to dioxane did not improve the situation, nor did a series of elaborate in-line, mixing devices, the description of which is irrelevent here. Con- sequently after nine months effort with Fluram it was aban- doned and the remaining three months given to studying o- phthalaldehyde. Studies on o-Phthalaldehyde A. Static Measurements. With only three months re- maining in the contract period allotted to the HPLC work (no HPLC was to be done the second year of the contract period) it was decided to go immediately to a dynamic system 142 ------- and perform a rough optimization, based on the data of Roth (5). It was found (see next section, Dynamic Studies on o- phthalaldehyde) that flows for the liquid chromatograph with the dynamic fluorogenic labelling detector, as well as solvent an4 reagent compositions, were optimum under the following conditions: (1) column flow 25 to 40% dioxane- H20, 1 ml/min (2) hydrolysis reagent, 0.02N NaOH, 0.4 to 0.5 ml/min., (3) OPA reagent, 1 g./l., 1.0 to 1.3 ml/min. By simulating these conditions in a Teflon-capped culture tube a study was made to determine the extent of hydrolysis occurring in the 35 foot hydrolysis coil. To 2 ml of 25% dioxane - H-O in the culture tube the appropriate carbamate was added to 10 yl of dioxane to give 10 M final concentration; then 1 ml of 0.02N NaOH was added. The tube was capped and incubated at 73°C in a water bath, after which it was removed, cooled rapidly to room temperature and reacted with jl ml of the OPA reagent (1 g/1 + 1 ml mercaptoethanol). The tube was swirled and read immediately at 334 and 454 nm in the fluorometer. Figures 4 through 7 show the results of the various hydrol- ysis reactions. Note that even after 30 minutes the fluor- escent intensity is still climbing for all the carbamates, denoting incomplete hydrolysis at the intermediate times. When the fluorescence, for any of them, at 30 min is com- pared with that at 2.2 min (the residence time within the 143 ------- 8 7 PQ 4 2 1. LANNATE MATACIL 0 25 .5 10 20 30 MINUTES FIG. 4 Relative fluorescence of N-methylcarbamates as a function of time. See text for conditions. CO 0 10 20 30 MINUTES FIG. 5 Relative fluorescence of N-methylcarbamates as a function of time. See text for conditions. 144 ------- 5- 4. HHW 3- CARBOFURAN ===== SEVIN .25 .5 10 20 30 MINUTES FIG. 6 Relative fluorescence,of N-methylcarbamates as a function of time. See text for conditions. MINUTES FIG. 7 Relative fluorescence of N-methylcarbamates as a function of time. See text for conditions. 145 ------- hydrolysis coil of the detector) it can be seen that at most, only 20 to 50% of total hydrolysis has occurred. Note that Lannate and Temik are hydrolyzing in a two step fashion, as seen from an intermediate maximum in their curves. Recommendations as to how the observed incomplete hydrolysis ran be remedied are to be found in Section B. B. Dynamic Measurements. Dynamic measurements were made with the assembled dynamic fluorogenic labelling liquid chromatograph as diagrammed in Fig. 8. Separations were achieved with the yC,« reverse phase column. It was anticipated that with the addition of the hydrolysis coil (10.7 m x 0.5 mm I.D.) bandspreading might occur to the degree that the somewhat tenuous separations achieved for Baygon and carbofuran with 40% acetonitrile H^O might no longer be possible. This indeed proved to be the case, not to be demonstrated however, until acetonitrile was replaced with dioxane, necessary due to the extremely high background and noise experienced because of the decomposition of ace- tonitrile to ammonia which produces fluorescence with the a-phthalaldehyde. Methanol and isopropanol also proved to be compatible with the fluorogenic labelling reaction, however, neither was able to provide the column efficiencies achievable with dioxane. Methanol produced a ten percent reduction in theoretical plates and isopropanol a twenty percent reduction when compared to dioxane; this is consis- tent with their respective viscosities. 146 ------- MOBILE PHASE 1.0 ml/min 25% DIOXANE - 40% DIOXANE H0 WATERS 660 SOLVENT PROGRAMMER 35 ft. 80°C COLUMN SAMPLE INJECTION VALVE 25 ul 0.02N NaOH 0.4 ml/min OPA REAGENT 1.3 ml/min FIG. 8 Modular dynamic fluorogenic labelling liquid chromatograph, 147 ------- While no careful study of mobile phase composition, nor flow rate was made, it soon became obvious that in order to elute Mesurol within an hour and yet maintain separations between Lannate and Matacil, as well as between Baygon and carbofuran, that solvent programming was required; this was due in part to the extra peak broadening caused by the hydrolysis coil (see section C. Detector Band Spreading Studies) and in part to the installation of a yC,g column of somewhat less efficiency than that used in section A. By programming from 25% to 40% dioxane for 13 minutes (curve #10, extremely concave) and thereafter holding isocratic, nearly complete separations were achieved for Baygon and carbofuran and complete separations for Lannate and Matacil. See Figs. 9 through 14, chromatograms of spiked soil extracts for typical separations. If 100% recoveries had been achieved 1.0 ppm represents 1.25 ug of each carbamate, 0.1 ppm rep- «• resents 125 ng, ancl 0.01 represents 12.5 ng, all for 25 ul injections. No detailed study of the fluorogenic labelling reaction was performed; however, by making large adjustments in the parameters involved and by bracketing techniques it was possible to determine ranges of acceptable conditions, which, to a first approximation, approached the optimum. Total flow through the 5 mm square fluorometer cell was determined by high noise on the baseline which appeared at about 5 ml/min at the high flow end, and by the low flow inadequacies of the reagent pumps at about 2.7 ml/min at 148 ------- o OS 3 C/) U4 8 12 16 20 24 MINUTES FIG. 9 28 32 52 56 Chromatogram of sandy soil, spiked at 0.01 ppm (top). Program No. 10 for 13 minutes, 25% dioxane - H20 to 40% dioxane - H20, 1 ml/min. Check soil (bottom). o OS 0 4 8 12 16 20 24 28 32 52 MINUTES FIG. 10 Chromatogram of sandy loam soil, conditions as in Fig. 9. 56 149 ------- O OS w s 0 4 8 12 16 20 24 28 32 52 56 MINUTES FIG. 11 Chromatogram of silty loam soil, conditions as in Fig. 9. \J V 0 4 8 12 16 20 24 28 32 52 56 MINUTES FIG. 12 Chromatogram of sandy soil spiked at 1.0 ppm. Other conditions as in Fig. 9. 150 ------- _J I—I u < w E- 0 8 12 16 20 24 28 32 52 56 MINUTES FIG. 13 Chromatogram of sandy loam soil, conditions as in Fig. 12. A o 4 8 12 16 20 24 28 32 52 56 MINUTES FIG. 14 Chromatogram of silty loam soil, conditions as in Fig. 12. 151 ------- the low end. Column flow appeared near optimum at 1.0 ml/ rain. When the NaOH flow approached the OPA flow, synchro- nous noise (blips, regularly spaced small peaks) appeared; consequently the NaOH flow was reduced as much as possible, consistent with reproducible pumping (0.4 ml/min). This allowed a maximum NaOH concentration of 0.02M; higher con- centrations began to shift the pH of the OPA buffer and reduced the response obtainable at pH 10. Consequently, almost 2 ml of OPA were possible before noise began to appear on the baseline (1.7 ml/min). This relatively high flow of OPA also diluted the dioxane in the final mix re- ducing "roll off" in the baseline when increasing amounts of dioxane appeared via programming. (see Fig. 20, a chrom- atogram from the Aminco Aminalyzer for a comparison). OPA concentrations less than 1 g/1 (with a propor- tionate reduction in mercaptoethanol) were workable. For example, 0.1 g/1 OPA produced less than a 50% reduction in response to the carbamate compared to the 1.0 g/1 concen- tration making small errors in preparation of the OPA re- agent insignificant. Bubble formation in the hydrolysis coil limited the hydrolysis temperature to 80°C, sufficient for a non-optimum but nevertheless sensitive response; the 2.2 minutes resi- dence kept the band spreading to a minimum. Even with this incomplete hydrolysis (see Figs. 4-7) seven serial injections of 50 ng Lannate standards gave a relative standard deviation of 2.7%, certainly comparable to most gas chromatographs. 152 ------- Analytical curves for the carbamates studied are shown in Fig. 15; excellent linearity was achieved, with the lowest practical working decade being 5 to 50 ng. About 1 ng is a realistic lower limit of detection for Lannate (signal/noise = 3). Recent work (antecedent to the contract period) has shown that by placing a restrictor after the hydrolysis coil, utilizing an integral debubbler fluorometer cell, super- heating the hydrolysis bath (135°C) and cooling before in- troduction of the OPA, a two to three fold increase in sen- sitivity can be realized, approaching 100% hydrolysis of the carbamates. An additional two fold increase in sensitivity can be realized by increasing the NaOH concentration to 0.2N along with an increase in borate buffer strength from 0.05 to 0.125M. A significant feature of the OPA fluorogenic labelling reagent approach is that check (blank) responses can be ob- tained by either: (1) substituting the OPA reagent with pH 10 buffer, or (2) substituting the 0.02N NaOH with pH6 distilled water. This was done for an unclean celery ex- tract (Fig. 16) showing that only the carbamate peaks dis- appear when the OPA reagent is withheld indicating that remaining peaks are all due to natural (underivatized) fluorescors. Similarly, by withholding the 0.02N NaOH hydrolytic reagent only the carbamate peaks disappear, in- dicating that the remaining peaks are in the free form and 153 ------- 200 - 160 g 120. cx, 80. 40. SEVIN LANNATE CARBOFURAN 4 MESUROL 5 BAYGON 6 TEMIK 0 20 40 60 80 100 120 ng/25 ul FIG. 15 Analytical curves for six N-methylcarbamate pesticides; 25 ul injections, conditions as in Fig. 9. 154 ------- do not require hydrolysis to fluoresce (Fig. 17). C. Detector Band Spreading Studies. By connecting the sample injection valve immediately upstream of the 5 mm square fluorometric cell and by making 25 ul injections of a natural fluorescor (Sevin) at various flow rates, and by performing a similar operation after moving the sample in- jection valve immediately upstream of the hydrolysis coil, an analysis can be made for the bandspreading contributions of both coil and cell individually and in tandem, applying the relationship for chromatographic band spreading: 2 22 a coil + cell = a coil + a cell where: a = standard deviation for a Gaussian distri- bution, and 4a = chromatographic peak width at base line. This was done for the OPA system; the plot of peak width versus flow obtained is shown in Fig. 18. For typical post column flows, a 1.4 ml/min flow through the coil pro- duces a 4a of 27 sec; in addition a 4a of 7 sec. is realized for 2.7 ml/min through the cell (1 ml/min, column + 0.4 ml/min, NaOH =1.3 ml/min OPA). Consequently, the total extracolumn bandspreading can be calculated to be 28 sec, (4a coil + cell). For a typical HSLC column of 5000 theo- retical plates, a peak at a retention of 10 min would be 34 sec. wide at the base; if the OPA detector's 27 sec. contri- bution is added to this quadratically, as it should, then the peak would be 44 sec. wide at the base, equivalent to a 155 ------- CELERY 0.2 ppm 50 ng W/0 OPA W/ OPA 0 3 6 9 12 15 18 21 24 MINUTES FIG. 16 Effect of withholding OPA (substituting pH 10 buffer) on response of celery spiked at 0.2 ppm with: 1, Lannate; 2, Temik; 3, Baygon; 4, carbofuran; and^S, Sevin 1 w/0 NaOH W/ NaOH LETTUCE 0.2 ppm 50 ng 9 12 15 MINUTES FIG. 17 Effect of withholding NaOH (substituting pH 7 H20) on re- sponse of lettuce spiked at 0.2 ppm with: 1, Lannate; 2, Temik; 3, Baygon; 4, carbofuran; 5, Sevin; and 6, Mesurol 156 ------- 4.7 4.3 - 3.9 - 3.5 - 3.1 - 2.7 - 2.5 - 2.9 - 1.5 - 1.1 - CELL + COIL COIL •CELL 12 16 20 24 28 32 36 40 SECONDS FIG. 18 Plot of 4a (peak width at base) for contributions of hydrol- ysis coil and of fluorometer cell, as a function of flow through each. 157 ------- column of about 3000 theoretical plates, a 40% reduction in efficiency. However, peaks eluting after the 10 min peak will experience less broadening and peaks eluting before will experience more. p. So:J4 Recoveries. The three soil types studied in Part I were spiked and analyzed to determine percent recov- eries for the eight carbamates. Zectran exhibited highly ir- regular behavior both in the chromatography of standards and in the analysis of soil extracts and hence is not shown in the chromatograms (Figs. 9-14), nor in the percent recovery table (Table 2). Soil was extracted and analyzed in the following manner: 1. 50 g of soil was placed in an appropriately sized Polytron jar, spiked with the appropriate carbamate standard mixture, and extracted with 100 ml. of beijzene - MeOH at maximum speed for 1 min. 2. The solvent was decanted into 500 ml Erlen- meyer, 100 ml additional benzene - MeOH added to the soil and again extracted as above, the ex- tract combined with the previous. 3. The jar was rinsed with 20 ml of benzene - MeOH and filtered through filter paper under vacuum. An additional 20 ml was used to rinse the jar. 4. The extract was evaporated to about 1 ml in a 250 ml Kuderna-Danish on a steam bath. 158 ------- TABLE 2 SOIL RECOVERIES Lannate Matacil Temik Baygon Carbofuran Sevin Mesurol Silty Loam, PPM .01 .02 .05 76 36 61 83 91 93 91 96 37 73 83 77 91 96 72 33 67 74 85 85 93 1.0 74 45 78 82 81 88 97 2.0 78 63 75 81 80 85 88 Lannate Matacil Temik Baygon Carbofuran Sevin Mesurol Lannate Matacil Temik Baygon Carbofuran Sevin Mesurol Sandy Soil, PPM .01 .02 .05 83 84 97 87 100 115 104 .01 11 13 21 79 75 88 95 98 109 96 104 105 105 93 Sandy Loam, .02 80 13 40 91 110 108 100 82 88 79 78 90 90 81 PPM .05 73 19 58 76 93 87 88 1.0 70 81 81 86 96 88 91 1.0 79 35 76 81 89 84 87 2.0 66 85 62 84 86 93 92 2.0 75 52 75 93 101 91 94 159 ------- 5. A small amount (3 ml) of benzene - MeOH was used to transfer the residue to a 15 ml culture tube, to which 2 ml acetonitrile and 4 ml 0.2N K2HP04 was added. 6. This was partitioned twice with 3 ml portions of methylene chloride, and the methylene chloride dried over crystalline sodium sulfate. 7. The methylene chloride was transferred to a clean culture tube with several rinsings and evaporated to dryness with N-- 8. The residue was dissolved in 0.2 ml of iso- propanol and diluted with 0.8 ml of deionized water and injected immediately on the liquid chromatograph (25 ul). It was discovered that the carbamate mixture degraded rapidly in dioxane. The best solvent seemed to be isopro- panol. In any solvent in which the higher concentrations were made up (dioxane, isopropanol, DMF) there was a dis- tinct autocatalytic effect causing rapidly diminishing re- sponses. This phenomena was noticed at the 1.0 and 2.0 ppm level (50 and 100 ng/ul) to a lesser extent at the 0.05 ppm level (2.5 ng/ul) and not at all at 0.01 and 0.02 ppm (0.5 and 1.0 ng/ul). From the data in Table 2 it can be seen that generally better recoveries are realized particularly at the lower levels, from the sandy soil as compared to the sandy loam or silty loam. Notice the somewhat poor recoveries for the 160 ------- lower levels of Lannate, Temik and Matacil from the sandy loam and silty loam soils. Undoubtedly these low recov- eries would have been more pronounced if the soils had been aged at all after spiking rather than analyzed imme- diately. Also notice the relative lack of interference in the three chromatograrts for the 0.01 ppm level (Figs. 9-11). At the 1.0 level the baseline is completely smooth with no trace of interference. Contrast these with the G.C. chrom- atograms in Part I; also notice the extremely low recoveries by G.C. for Lannate and Temik, even after column cleanup. No extensive interference study was conducted; however, it was determined that it took 10 ug of a-naphthol, a degra- dation product of carbaryl, to produce the same sized peak as 10 ng of carbaryl, even though a-naphthol is highly fluorescent at very close to the wavelengths of the a- phthalaldehyde fluorophore; it elutes about 1 minute prior to carbaryl under the programmed mobile phase. It has not been established whether the urea herbicides elute under these conditions or not, nevertheless, 10 ug of monuron or diuron produced no response whatsoever. E. Aminco Aminalyzer. An American Instrument Co. (Aminco) amino acid analyzer, the Aminalyzer, was converted for the analysis of the N-methylcarbamate pesticides by placing 25% dioxane - H-O and 60% dioxane - H90 in the L i* mobile phase reservoirs, placing 0.2N NaOH in the sodium hypochlorite reservoir, and substituting the ion-exchange 161 ------- column with a 30 cm x V I.D. uC,g reverse phase column. Flows were similar to the previously described modular instrument, however, two major differences existed (Fig. 19). Firstly, the Aminalyzer is capable only of step gradient operation, achieved through control of a four way valve between four solvent reservoirs and the pump by a digital clock. Only one step (two solvent mixtures) was used for this work. The results of this are seen on the chromatogram of a six carbamate mix (Fig. 20), as a step in the baseline caused by switching over to 60% dioxane at 20 minutes into the chromatogram. Sensitivity was good, however, as seen by the extended range analytical curve for Lannate (Fig. 21). Secondly, hydrolysis was performed at a "nominal" 100°C within a 15 ft. x 0.02 in. I.D. Teflon hydrolysis coil; whether this provided complete hydrolysis was hot determined. Conclusions The o-phthalaldehyde based dynamic fluorogenic labelling high speed liquid chromatograph offers real advantages for the analysis of N-methylcarbamate pesticides in soil, and more than likely, in other types of agricultural and environ- mental samples. For the previously described arrangement it can reliably analyze seven of the eight carbamates tested (Zectran excluded) at the 0.01 ppm level. It can be assembled from "off the shelf" components and requires no special ex- 162 ------- MOBILE PHASE 25% DIOXANE 60% DIOXANE DIGITAL HCONTROL LOGIC DELAY uClg COLUMN SAMPLE INJECTION VALVE OPA n PUMP 0.2N NaOH FIG. 19 Aminco Aminalyze^0', adapted to N-methylcarbamate analyses. RECORDER FLUOROMETER WASTE ------- 1 6 CARBAMATE MIX 50 ng STEP-25% to 60% DIOXANE AMINCO AMINALYZER 5 0 8 16 24 32 40 MINUTES FIG. 20 Aminalyzer chromatogram of: 1, Lannate; 2, Temik; 3, Baygon; 4, carbofuran; 5, Sevin; and 6, Mesurol. 10 w u 2 § 10 w o a 3 -3 O 4 9 H 10 CD PP H 10' LANNATE AMINCO AMINALYZER 0.1 1 10 100 NANOGRAMS FIG. 21 Aminalyzer extended range analytical curve of Lannate. ------- pertise in assembly. The key reagent, a-phthalaldehyde, can be obtained commercially in pure form (Aldrich Chem- icals) , does not need repurification, and is reasonably inexpensive ($14.50/25g). The reagent is stable in solu- tion (5°C) for at least a week, possibly longer. b Two yC,g analytical columns were employed during the one year contract period for this portion of work; one would have been sufficient if it had not been inadvertently plugged with unclean leafy vegetable extract. It was for- tunate that the uC,R column performed adequate separations of the carbamates because additional trials with the uCN and yNH2 columns with the dynamic fluorogenic labelling detector demonstrated that they continually bled a material which reacted with the o-phthalaldehyde producing extremely intense fluorescence. No such bleeding was observed with the uC,g columns. The extreme sensitivity and more importantly, selec- tivity of this detector has made it possible to live with the effective degrading of the column efficiency due to band spreading within the hydrolysis coil, and to a lesser extent, within the fluorometer cell. Analytical columns of higher efficiency and higher hydrolysis temperatures, allowing for shorter coil length, should moderate this situation somewhat. Needless to say, if a complete chromatogram of all of the carbamates tested in this study is not necessary the 165 ------- dioxane percentage in the mobile phase can be adjusted to provide a more rapid analysis of the compound of interest. Dioxane percentages of up to 80% should be possible in the dioxane - H-0 mobile phase without adversely affecting the labelling reaction. Although a spectrophotofluorometer, equipped with grating monochromators, was utilized in this study, it certainly should be possible, if not advisable, to employ a filter fluorometer, such as the Aminco Fluoromonitor, or others of such type. As demonstrated in Part I of this report the Polytron ultrasonic extraction device appears to offer a rapid, efficient means for the extraction of soil, not-withstanding the fact that the generator blades deteriorate rapidly due to the abrasiveness of the soil. Minimum cleanup, an ace- > tonitrile water - methylene chloride partitioning, is required for nearly interference free liquid chromatograms of soil extract and nearly complete recovery of most of the carbamates, even as low as 0.01 ppm. The Aminco Aminalyzer performed adequately on car- bamate standards, however, it appeared to offer no real advantages over the assembled instrument, other than it is fully cabinetted. 166 ------- References (Part II) i 1. W. P. Cochrane, J. Chromatogr. Sci. , 13, 246 (1975). 2. H. W. Dorough and J. H. Thorstenson, Ibid, 13, 212 (1975) 3. A. M. Felix and G. Terkelsen, Arch. Biochem. Biophys., 157, 177 (1973). 4. S. Udenfriend, S. Stein, P. Bohler, and W. Dairman, Science, 178, 871 (1972). 5. M. Roth, Anal. Chem., 43, 880 (1971). 6. M. Weigele, S. DeBernardo and W. Leimgruber, Biochem. Biophys. Res. Comm; 50, 352 (1973). 167 ------- TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) . REPORT NO. EPA-600/1-77-029 2. 3. RECIPIENT'S ACCESSION>NO. Jfc 4. TITLE AND SUBTITLE EXTENSION OF MULTI-RESIDUE METHODOLOGY. I. Determining Multiclass Pesticide Residues in Soil by Gas ChromatograShy^. II. Dynamic Fluorogenic Labelling Detector for Carbamatps. 5. REPORT DATE 1Q77 FORMING ORGANIZATION CODE 7. AUTHOR(S) H. Anson Moye, Sujit Witkonton and Gordon Cash 8. PERFORMING ORGANIZATION REPORT NO. 9. PERFORMING ORGANIZATION NAME AND ADDRESS University of Florida, I FAS Food Science Department Pesticide Research Laboratory Gainesville, FL 32611 10. PROGRAM ELEMENT NO. 1EA615 11. CONTRACT/GRANT NO. 68-02-1706 12. SPONSORING AGENCY NAME AND ADDRESS Health Effects Research Laboratory Office of Research and Development U.S. Environmental Protection Agency Research Triangle Park. N.C. 27711 13. TYPE OF REPORT AND PERIOD COVERED RTP,NC 14. SPONSORING AGENCY CODE EPA/600/11 15. SUPPLEMENTARY NOTES 16. ABSTRACT A multi-residue procedure was devised for the extraction, cleanup and determination of forty seven pesticides in fortified soil samples. Most of the compounds were determined by gas chromatography interfaced with either the electron capture or the flame photometric detector. Several nitrogen containing pesticides were quantitated using a fluorescence detector in conjunction with a high performance liquid chromatographic system. Various methods of extraction were compared for several pesticides. A silica gel column, eluted sequentially with various solvents, was used for cleanup and separation of compounds into groups. Several different derivatization procedures were used to render some pesticides more amenable to gas chromatographic detection. A dynamic fluorogenic labelling detector was designed and characterized for the high pressure liquid chromatographic analysis of six Y-methylcarbamate and two carbamoyl oxime pesticides. Lannate, Matacil, Temik 3aygon, carbofuran, Sevin and Mesurol could be extracted from sandy soil at the 0.01 ppm level with recoveries ranging from 83 to 115%. Somewhat lower recoveries were experienced from sandy loam and si 1 try loam soils. Zectran could not be ellKl* chromat°9raphed under the liquid chromatographic conditions chosen for ^uaranons. No cleanup was required prior to the liquid chromatography of the soil extracts; no significant interference WPPP ohcpryoH f9r the unclean extracts.— KEY WORDS AND DOCUMENT ANALYSIS a. DESCRIPTORS b.lDENTIFIERS/OPEN ENDED TERMS c. COS AT I Field/Group Pesticides sxtraction chemical analysis gas chromatography soil analysis 14 B 07 A 18. DISTRIBUTION STATEMENT RELEASE TO PUBLIC 19. SECURITY CLASS (This Report) UNCLASSIFIED 21. NO. OF PAGES 20. SECURITY CLASS (Thispage) UNCLASSIFIED JL8A 22. PRICE EPA Form 2220-1 (9-73) 168 ------- |