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

-------