&EPA
            United States
            Environmental Protection
            Agency
            Environmental Research
            Laboratory
            Athens GA 30605
EPA-600 4-80-008
January 1980
            Research and Development
Polyurethane  Foam as
Trapping Agent for
Airborne Pesticides
           Analytical  Method
           Development



-------
                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 MONITORING series.
This series describes research conducted to develop new or improved methods
and  instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                                EPA-600/4-80-008
                                                January  1980
        POLYURETHANE  FOAM AS  TRAPPING AGENT
               FOR AIRBORNE PESTICIDES
           Analytical Method  Development
                         by

         James D. Adams and Joseph H.  Caro
Soil Nitrogen and Environmental Chemistry Laboratory
      Beltsville Agricultural Research Center
           U.S. Department of Agriculture
             Beltsville, Maryland 20705
        Interagency Agreement No. 78-D-X0449
                  Project Officer

                  William R. Payne
           Environmental Systems Branch
         Environmental Research Laboratory
               Athens, Georgia 30605
        ENVIRONMENTAL RESEARCH LABORATORY
        OFFICE OF RESEARCH AND DEVELOPMENT
       U.S. ENVIRONMENTAL PROTECTION AGENCY
               ATHENS, GEORGIA 30605

-------
                                 DISCLAIMER

     This report has been reviewed by the Environmental Research Laboratory,
U.S. Environmental Protection Agency, Athens, Georgia, and approved for pub-
lication.  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.

-------
                                  FOREWORD
      Environmental protection efforts are increasingly directed towards
preventing adverse health and ecological effects associated with specific
compounds of natural or human origin.  As part of this Laboratory's research
on the occurrence, movement, transformation, impact, and control of environ-
mental contaminants, the Environmental Systems Branch studies complexes of
environmental processes that control the transport, transformation, degrada-
tion and impact of pollutants or other materials in soil and water; assesses
environmental factors that affect water quality; and develops new techniques
for measuring contaminants.

      Pesticides transported in the air from application sites on forest and
agricultural lands are receiving increased attention because of their po-
tentially adverse health effects on humans and animals.  A major problem in
efforts to measure ambient air concentrations of pesticides has been the lack
of an effective means of collecting samples for analysis.  This report evalu-
ates the use of polyurethane foam as a trapping agent for selected airborne
pesticides.

                                      David W. Duttweiler
                                      Director
                                      Environmental Research Laboratory
                                      Athens, Georgia
                                    iii

-------
                                    ABSTRACT

    This research program was initiated to develop and evaluate a method  for
determining levels of organochlorine, organophosphorus and N-methylcarbamate
insecticides in air, using polyurethane foam as the trapping agent.

    A multipesticide method was developed, using 4.4 cm-diameter plugs of
polyurethane foam as traps and a modified Sherma-Shafik multiresidue procedure
for analysis of foam extracts.  With this method, the minimum detectable  air
concentrations for vapors of 17 of the 18 organochlorine and organophosphorus
pesticides tested was 0.1 ng/m^ or less.  Six carbamate pesticides did not
volatilize in sufficient amounts to allow analysis.

    The experimental program consisted of measurements of efficiency of foams
for trapping vapors and aerosols, tests of foam compression during air samp-
ling, evaluations of techniques for extracting pesticide residues from the
foams, evaluations of extraction solvents, comparisons of the performance of
the ester and ether forms of polyurethane foam, evaluations of the steps  in
the Sherma-Shafik analytical procedure, evaluations of GLC columns and instru-
mental parameters, and tests of derivatization procedures for electron-capture
gas chromatography of the N-methylcarbamates.

    A 10-cm depth of foam was found to be an efficient trap for vaporized
pesticides.  However, there were indications that the foam may be less
effective in trapping pesticides contained in airborne aerosols.  Foams
compressed during air sampling, necessitating continuous monitoring of flow
rates for most accurate measurement of volume of air sampled.

    Five-cycle Soxhlet extraction with 1:1 hexane-acetone was the best pro-
cedure for removal of pesticides from foam plugs.  Excessive Soxhlet treatment
produced lower recoveries of organophosphorus insecticides.  Other solvents
gave low recoveries or unacceptable levels of coextracted contaminants.  The
ester form of polyurethane gave better results than the ether form.  GLC
columns other than those recommended by Sherma and Shafik gave improved per-
formance in multiresidue analysis of extracts.  Carbamates (in aerosols) were
best analyzed by derivatization with 0.1 ml pentafluoropropionic anhydride.

    This report was submitted in fulfillment of Interagency agreement No.
D6-0136 (78-D-X0449) by the Soil Nitrogen and Environmental Chemistry Labora-
tory (formerly the Agricultural Chemicals Management Laboratory), SEA-AR,
U. S. Department of Agriculture, Beltsville, MD, under the sponsorship of the
U. S. Environmental Protection Agency.  This report covers the period Septem-
ber 30, 1976, to January 15, 1979.  Owing to the time required for recruitment
of personnel, work began on June 30, 1977.  Work was terminated on Jan-
uary 15, 1979, because of the impending transfer of the principal investigator
to the U.  S. Environmental Protection Agency.

                                      iv

-------
                                 CONTENTS

Foreword 	   iii
Abstract 	    iv
Acknowledgements 	    vi

    1.  Introduction 	     1
    2.  Conclusions 	     2
    3.  Recommendations 	     4
    4.  Materials and Equipment 	     5
            Pesticides 	     5
            Solvents 	     5
            Gas Chromatographs 	     5
            Gas-Liquid Chromatography Columns 	     5
            Polyurethane Foam Plugs 	     9
    5.  Experimental Program and Results 	    10
            Investigation of PUF Trapping Efficiency 	    10
            Evaluation of Pesticide Extraction Methods 	«;    27
            Comparison of Ester and Ether Forms of PUF 	    43
            Evaluation of GLC Columns and Procedures 	    49
            Effect of Foam Structure on Air Flow Rates 	    66
            Analysis of N-Methylcarbamate Insecticides 	    67

References 	    75

-------
                                ACKNOWLEDGEMENTS

    We gratefully acknowledge the contributions of Mr. David P. T. Chen,
Mr. D. L. Brower and Mr. H. P. Freeman, who shared in the conduct of the
experiments.  Mr. Chen is a National Urban League Fellow, assigned to the Soil
Nitrogen and Environmental Chemistry Laboratory on a temporary basis as a
research chemist; Mr. Brower is a soil scientist and Mr. Freeman is a chemist,
both on the permanent staff of the Laboratory.  We also thank Dr. Henry F.
Enos, Deputy Director, Environmental Research Laboratory, U. S. Environmental
Protection Agency, Athens, GA, for advice and guidance during the formative
stages of the research program.
                                     vi

-------
                                   SECTION 1

                                  INTRODUCTION
    Every year in the United States, millions of pounds of pesticides  are
applied to agricultural and forest lands, 65% by aerial application  (1).
During application, a significant fraction remains in the air as vapor, as
particulate matter, or in aerosols.  Also, some of the material that does
reach the ground can later be introduced into the air as vapor by
volatilization, or as adsorbate on fine particles by wind or by mechanical
agricultural operations.  The movement of pesticides in air is a major mode  of
their translocation over both short and long distances.  These airborne
pesticides may have epidemiological consequences for both humans and animals.
As a result, the Congress has directed the EPA to establish a national system
to monitor ambient air for concentrations of pesticides.

    A major obstacle to the implementation of a national system is that the
methods that have been used for trapping pesticides from air suffer  from one
or more problems.  Among these are poor trapping efficiency, limited air flow,
degradation of trapped pesticides, difficulty in recovery of trapped
pesticide, instability of the trapping medium, and difficulty with cleanup of
the recovered pesticide to permit its unambiguous quantitation.

    Several of these effects do not occur with polyurethane foam (PUF).  That
is, PUF is not hygroscopic; since it is semi-solid, it is not entrained in air
streams; it does not volatilize; and the polymer is relatively nonreactive and
therefore unlikely to degrade labile pesticides.  Moreover, it has been
reported to perform efficiently as a trap for certain pesticides (2, 3).

    The present study was designed to provide an in-depth evaluation of PUF  as
a trapping agent for a broad spectrum of airborne pesticides, focusing on
efficiency of trapping, ease of recovery of trapped pesticides, relative
performance of the ether and ester forms of PUF, and compatibility of  PUF with
the multipesticide analytical scheme of Sherma and Shafik (4).  The  latter had
shown promise for use in a national monitoring system, except that the
trapping procedure, air flow through a column of ethylene glycol, exhibited
unacceptable shortcomings.

-------
                                   SECTION 2

                                  CONCLUSIONS
    The results of laboratory experiments using an airflow apparatus indicated
that PUF is an efficient trap for vapors of organochlorine and organophos-
phorus pesticides, even when substantial amounts of air are sampled.  A  10-cm
depth of ester-form PUF was adequate for trapping vapors of the 18 volatile
compounds tested when as much as 55 m^ of air was passed through the
apparatus.

    A finding having important implications for ambient monitoring was that
pesticides were not trapped as well by PUF when they were sampled as airborne
aerosols rather than as molecular vapors.  The aerosol tests included a  number
of carbamate insecticides that had not volatilized in earlier airflow
experiments.

    During extended air sampling with plugs of PUF, the plugs compressed
somewhat and air flow rates decreased.  In several 24-hour airflow tests, the
flow rate decrease ranged from 3 to 9%.  Accurate measurement of volume  of air
passing through PUF plugs in ambient air sampling will require monitoring of
flow rates.

    Experiments showed that Soxhlet extraction was superior to both solvent
elution and immersed plug compression as a technique for removal of pesticides
from PUF plugs.  However, excessive Soxhlet extraction should be avoided.
Tests showed that extractions carried beyond five Soxhlet cycles may cause
degradation and decreased recovery of the organophosphorus pesticides.   The
best solvent for extraction was 1:1 hexane-acetone.  Less polar solvents gave
low recoveries; more polar solvents coextracted interfering nonpesticidal
components.

    Soxhlet extracts of PUF plugs were compatible with the Sherma-Shafik
multipesticide analytical procedure (4).  Extracts could be analyzed, with low
background levels, at concentrations down to 0.1 ng/m^ Or lower for all
tested pesticides except for carbophenothion, which could be measured down to
0.3 ng/m3.


    No significant differences between the ester and ether forms of PUF  were
observed with respect to the trapping or extraction of pesticides.

    In evaluations of GLC columns, improved performance was obtained by  using
columns other than those recommended by Sherma and Shafik (4) for analysis of

-------
the three mixtures of pesticides resulting from their cleanup procedure.
Optimum results were obtained with use of these columns:   (a) 1.5% OV-17  and
1.95% OV-210 on Chromosorb W HP; (b) 3% OV-1 on Chromosorb W; and (c)  3%
DC-200 on Gas Chrom Q.

    Tests of the derivatization of the carbamate insecticides for analysis  by
electron-capture gas chromatography showed that the pentafluoropropionic
anhydride (PFPA) procedure of Sherma and Shafik (4) gave the most satisfactory
results, except that response was enhanced by increasing the amount  of PFPA
from the recommended 0.025 ml to 0.100 ml.  Trifluoroacetic anhydride,
1-fluoro-2,4-dinitrobenzene and heptafluorobutyric anhydride were also tested
as derivatization reagents, the first two giving unacceptable results  and  the
last offering no improvement over PFPA.

    Experiments with 2,4- and 2,6-diaminotoluene, which are products of PUF
degradation, showed that their PFPA derivatives were removed during  extract
cleanup and, therefore, would not interfere with the quantitation of carbamate
insecticides.

-------
                                 SECTION 3

                              RECOMMENDATIONS

      The stability of pesticides trapped on polyurethane foam should be
confirmed with respect to humidity and photochemical oxidants to evaluate
suitability for use in areas of high or low humidity or high smog levels.
The recommended procedure should be field tested by sampling ambient air in
an urban, an industrial, and an agricultural environment.

      If polyurethane foam is to be used in a multiple-site air monitoring
program for pesticides, it is recommended that, to trap pesticides in both
vapor and aerosol form, the column of foam be 25 cm deep and at least 4.4 cm
in diameter.  All of the foam should be hydrophobic and from one manufacturing
lot.  The foam should be thoroughly cleaned before use and be well protected
from contamination at all times.  All GLC columns to be used for quantitation
should be from a single source to ensure uniformity in measurement of residue
levels in samples collected from different sites.

      Further research is necessary to confirm the penetration of polyurethane
foams by aerosols and to determine the reasons and conditions for aerosol
penetration.  If the mechanism for this penetration is understood, it may be
possible to modify the foam or sampling conditions to minimize the effect.
Previous reports on sampling pesticides in air have been concerned with vapors,
but aerosols may well be a major form in which pesticides occur in air. Accord-
ingly, it will be necessary to examine trapping efficiency for aerosols in
evaluations of pesticide adsorbants other than polyurethane foam, such as the
reticular porous polymers.

-------
                                   SECTION 4

                            MATERIALS AND EQUIPMENT
PESTICIDES

    Because one objective of the program was to evaluate the feasibility of
applying the Cherma-Shafik analytical procedure to PUF extracts, the
pesticides in the original paper by Sherma and Shafik (4) were selected for
use in this study.  All pesticides used were analytical-grade standards,
obtained from commercial sources:  a- and 3-BHC from Applied Science
Laboratories, Inc.; the remainder of the compounds from Supelco, Inc.  The
list of pesticides, along with chemical names and equilibrium vapor pressures
as available, is shown in Table 1.

SOLVENTS

    All solvents used were glass-distilled, pesticide grade, obtained from
Burdick and Jackson Co.

GAS CHROMATOGRAPHS

    Four gas chromotographs were used in the program:  (1) a Tracer Model 222,
equipped with a ^Hi electron-capture detector (ECD), for analysis of
organochlorine and carbamate insecticides; (2) a Tracer Model 560, with
ECD, also for organochlorine analysis; (3) a second Tracer Model 222, with a
flame photometric detector (FPD), for analysis of organophosphorus
insecticides; and (4) a Varian Model 2100, with an alkali flame ionization
detector (AFID), also for organophosphorus analysis.

GAS-LIQUID CHROMATOGRAPHY COLUMNS

    Ten glass, packed GLC columns were used for pesticide analysis in the
various phases of the program.  Four of these were obtained from commercial
sources, four were packed in our laboratory using  packing material we had
prepared ourselves, and the remaining two were packed in our laboratory using
commercial packing material.

    Of the four commercial columns, three were 180-cm long, 2-mm I.D. coiled
columns for use in the Tracor 560 gas chromatograph.  The packings in these
were:   (a) 3% OV-210 on 80- to 100-mesh Supelcoport (Supelco, Inc.); (b) 5%
SE-30 on 80- to 100-mesh Supelcoport (Supelco, Inc.); and (c) 3% DC-200 on
100- to 120-mesh Gas Chrom Q (Applied Science Laboratories, Inc.).  The fourth

-------
                                 TABLE 1.  TEST PESTICIDES
   Common or trade name
                                Vapor pressure
mm Hg   at   T,  C
                           Chemical designation
Aldrin


Aminocarb

a-BHC

3-BHC

Carbaryl

Carbofuran


Carbophenothion


p,p'-DDD

p,p»-DDE

o,p'-DDT


p,p'-DDT

Diazinon


Dieldrin
 6 x 10
                                   -6
<5 x 10
       -3
3.05 x 10
         -7
1.8 x 10
        -7
25
26
20
3-D x 10*"'''      25

8.U x 10~5      20
25
1,2,3,^,10-10-hexachloro-l,k,ha.,5,8,8a-hexahydro-
   U-endo,exo-5,8-dimethanonaphthalene

U-dimethylamino-m-tolyl N-methylcarbamate

a-isomer of 1,2,3,U,5»6-hexachlorocyclohexane

g-isomer of 1,2,3,U,5,6-hexachlorocyclohexane

1-naphthyl N-methylcarbamate

2,3-dihydro-2,2-dimethyl-7-benzofuranyl
   N-methylcarbamate

0,0-diethyl-S-[[(p-chlorophenyl)thiojmethyl]
   phosphorodithioate

1,1-di chloro-2,2-r-bi s (p^chlor ophenyl) ethane

1,l-dichloro^2,2-bis(p^ehlorophenyl)ethylene

1,1,l-trichloro^2-(o-chlorophenyl)-2~(p-chloro-
   phenyl)ethane

1,1,l-trichloro-2,2-bis(p-chlorophenyl)ethane

0,0-diethylr-0- (2-isopropylr-l4.^methyl^6-pyrimidinyl)
   phosphorothioate

1,2,3, ^, 10, lO^hexachloror-6,7-epoxy-l, U, Ua, 5,6,7,8,
   8a-octahydro^-l, U-endo, exo^-5, 8i-dimethanonaph-
   thalene

-------
                                        TABLE 1.  (continued)
   Common or trade name
                                 Vapor pressure
nm Hg
at
               T, °C
Chemical designation
Endrin



Ethion


Heptachlor


Heptaehlor epoxide


2,3,5-iandrin.

Lindane

Malathion


Methiocart

Methoxychlor

Methyl parathion

Mexacarbate
1.5 x 10'
9.U x 10

1.25 x 10'
9.7 x 10
                  20
                  20

                  20
                  20
                    1 , 2 , 3 , k , 10 , 10-hexachloro-6 , 7-epoxy-l , h , ha , 5 , 6 , 7 ,
                      8 , 8a^oct ahydro^l , Wendo , endo^-5 , S^dimethanonaph-
                      thalene

                    0,0,0' ,0f»-tetraethyl S,S'r-methylene"bisphosphoro-
                      dithioate

                    1,^,5,6,7,8, 8-hept achloro-3a , h , 7 ,7a^tetrahydro-
                      U , 7-endo^-methanoindene
                                                       -3a , k , 7 , 7a-
                             1,^,5,6,7,8, 8^
                               tetrahydro-4 , 7-methanoindene
                   '2,3,5^tro,etju; ! jemu; H^methylcarbamate

                    Y-isomer of l,2,3,^,5»6-hexachlorocyclohexane

                    0 , 0-dimethyl S^ ( 1 , 2-dicarbethoxyethyl )
                      phosphorodithionate

                    li^methylthio-3 , 5-xylyl N-methylcarbamate
                                                          *
                    1 ,1 ,l*-trichloro-2 , 2-bis (.p-methoxyphenyl ) ethane

                    0 jO^dimethyl O^p-nitrophenyl phosphorothioate

                    U^dimethylamino-3 , 5-xylyl N-methylcarbamate

-------
TABLE 1  (continued)
Vapor pressure
Common or trade name

Parathion
oo
Propoxur
Ronnel
mm Hg at
-6
5-7 x 10

1.0 x 10~2
8 x 10~2
T, °C

20

20
25
Chemical designation

0,0-diethyl 0-p-nitrophenyl phosphorothioate

o-isopropoxyphenyl N-methylcarbamate
0,0-dimethyl 0-(2,U,5-trichlorophenyl) phosphoro-
                  thioate

-------
commercial column was a  180-cm  long, 4-mn  I.D.   U-tube,  packed with 3% OV-1 on
80- to 100-mesh Chromosorb W  (Tracor,  Inc.).

    To prepare packings, we used  80- to  100-mesh Chromosorb W as  support
material.  The procedure was  as follows, using  SE-30  as  an example:   0.50 g
SE-30 was dissolved in 35 ml  chloroform  in a  250-ml  filter flask,  and 9.50 g
of the support material was added.  Vacuum was  then  applied until  the solution
boiled briefly.  The vacuum was released and  the slurry  was swirled gently.
Evacuation, release and  swirling  were  repeated  until  all the  free  liquid had
evaporated.  The remaining slurry, which had  the consistency  of damp sand, was
subjected to fluidized drying in  a fluidizer  ("HI-EFF",  Applied Science^Labor-
atories, Inc.) with nitrogen  flowing through.   The  fluidizer  was  placed on a
hot plate at about 100°C surface  temperature  until  the material became free-
flowing.

    Packings prepared by this procedure were:   (a)  5% SE-30,  as above; (b) 3%
OV-1, from 0.30 g OV-1,  35 ml chloroform and  9.70 g  Chromosorb W;  (c) 5%
OV-210,  from 1.021 g OV-210,  70 ml chloroform and 19.40  g Chromosorb W; and
(d) another 5% OV-210, but treated with  Carbowax 20M by  the procedure of Ives
and Giuffrida (5).

    To pack columns, we  poured  the dry,  coated  support material into one end
of a  180-cm long, 4-mm I.D. U-tube, while  applying  a  vacuum to the other end
and simultaneously tapping the  column.   The column was then conditioned by
heating  to 250°C for 18  hours in  a flow  of nitrogen.  We packed columns with
the four self-prepared packings and with two  commercial  packings:   (a) 3%
DC-200 on 100- to 120-mesh Gas-Chrom Q (Applied Science  Laboratories, Inc.),
and (b)  a mixture of 1.5% OV-17 and 1.95%  OV-210 on  Chromosorb W  HP (Supelco,
Inc.) .

POLYURETHANE FOAM PLUGS

    Polyurethane foam, in both  ester (gray) and ether (white) forms, was pur-
chased in 5-cm-thick sheets from  a commercial source, Read Plastics, Inc.,
Rockville, MD.  Both of  these foams were flexible,  reticulated and hydrophobic
types.   Cylindrical plugs, 45-mm  diameter  by  5  cm thick, were cut from the
sheets with a special tool mounted in  a  drill press.  The tool was fabricated
to our specifications from thin-wall stainless  steel  tubing in our Instrument
Design Shop.

-------
                                   SECTION 5

                        EXPERIMENTAL PROGRAM AND RESULTS
INVESTIGATION OF PUF TRAPPING EFFICIENCY

    Laboratory evaluations of pesticide vapor trapping efficiency by various
media have generally involved the spiking of some type of  source with  small
quantities of the pesticides to be tested; passing a given volume of air
through the source and through the trapping medium in a suitable apparatus;
analyzing the source, trapping medium, and chamber washings  for pesticide  con-
tent; and calculating trapping efficiency by a material balance (6).   The
factors that control the rate at which a given pesticide is  volatilized in
these tests include the vapor pressure of the pesticide, air temperature,
airflow rate, energy of adsorption of the pesticide by the source material,
and surface area of the source.

    The material balance approach is valid only if the pesticides are  stable
during volatilization, trapping and analysis.  Finding a suitable method for
introducing thermally unstable pesticides with very low vapor pressures into
air in such a system is a major experimental obstacle.  Because of  the latter
restriction, we conducted experiments on PUF trapping efficiency not only  with
pesticides as molecular vapors, but also as components of  aerosols.

First Vapor Trapping Experiment

    One microgram of each of 16 pesticides was applied to  a  glass-fiber filter
mounted in an all-glass apparatus, and two ester-form PUF  plugs, 5  cm  deep and
45 mm in diameter, were placed in series in the apparatus  as traps.  Air was
drawn through the source and plugs for 22 hours at an average flow  rate of
2.52 m /hr, a total volume of 55.4 m-*.  At the end of that time, the plugs
were extracted individually by Soxhlet extraction with 1:1 hexane-acetone, and
the source was extracted by washing with acetone.  The extracts were concen-
trated, separated into pesticide groups by silica-gel column chromatography
according to Sherma and Shafik (4), and quantitated by GLC.   The analytical
results are listed in Table 2.

    The results  showed  that the N-methylcarbamates  (carbofuran,  carbaryl,
methiocarb) did  not volatilize from  the  source,  that  a  5-cm  depth  of PUF  was
insufficient to  trap  the more volatile compounds  quantitatively,  and that some
degradative  loss  of parathion  and  methyl parathion  apparently occurred.   The
remainder  of the  compounds were quantitatively  volatilized and  trapped.
                                      10

-------
Second Vapor Trapping Experiment

    To aid pesticide volatilization and concurrently reduce chemical degrada-
tion, a trapping experiment was conducted at an elevated temperature and with
decreased volume of air flow.  Glass wool (1.65 g) was placed in a 44-mm-I.D.

                 TABLE 2.   PUFa TRAPPING EFFICIENCY FOR 1-yg
                     DOSES OF PESTICIDES IN 55 M3 OF AIR

Residues, % of applied pesticides
Compound
Aldrin
p,p'-DDE
p.p'-DDD
p,p'-DDT
Lindane
Heptachlor
Heptachlor epoxide
Dieldrin
Methoxychlor
Methyl parathion
Parathion
Diazinon
Ma lath ion
Carbofuran
Carbaryl
Methiocarb
Remaining
on source
0
0
0
3
0
0
0
0
42
0
0
4
29
100
100
91
On first
PUF plug
98
100
84
98
69
71
74
100
45
70
75
64
64
0
0
0
On second
PUF plug
23
0
0
17
3
33
0
0
0
0
0
24
0
0
0
0
Total
recovered
121
100
84
118
72
104
74
100
87
70
75
92
93
100
100
91
a Ester-form polyurethane
                                      11

-------
glass tube and held in place with a brass screen.  To this, 1 ml of hexane  so-
lution containing 1 pg of each of 24 pesticides was applied slowly.  After
evaporation of the hexane, two 5-cm, ester-form PUF plugs were placed  in  the
tube, and the tube was wrapped with heating tape.  Air was drawn through  the
apparatus for 2 hours at 4.3 m^/hr.  The temperature in  the chamber was
45°C during the experiment.  The plugs (analyzed together) and the wool were
then subjected to 5-cycle Soxhlet extraction with 1:1 hexane-acetone.  The  ex-
tracts were concentrated, chromatographed on silica gel, and quantitated  by
GLC.  The analytical results are listed in Table 3.

    Except for a-BHC, the organochlorine insecticides were quantitatively
trapped on the plugs.  Certain of the organophosphorus compounds did not
volatilize completely, but the portions that did volatilize were
quantitatively trapped.  The carbamates again did not volatilize from  the
source to any appreciable extent.  Attempts to detect the carbamates on the
plugs were confounded by GLC interference in the form of a large injection
peak.  The low recovery of mexacarbate, a carbamate, may have been caused by
oxidation or thermal breakdown on the source.
           TABLE 3.   ESTER-FORM PUF TRAPPING EFFICIENCY AT 45°Ca
                    	Residues,  % ofapplied pesticides	
     Compound              On source             On plugs               Total
a-BHC
Aldrin
p,p'-DDE
o.p'-DDT
p , p ' -DDD
p.p'-DDT
*
Lindane
3-BHC
Heptachlor epoxide
Dieldrin
Endrin
Methyl parathion
0
0
0
0
0
0
0
0
0
0
0
0
76
98
100
110
92
86
93
102
88
101
95
113
76
98
100
110
92
86
93
102
88
101
95
113
                                     12

-------
                              TABLE 3  (continued)
     Compound
                                   Residues,  %  of  applied  pesticides
On source
On plugs
Total
Ronnel
Parathion
Ethion
Carbophenothion
Diazinon
Malathion
Propoxur
Carbaryl
Carbofuran
Aminocarb
Methiocarb
Mexacarbate
0 119
0 106
0 118
36 45
112 0
10 91
94
93
89
87
100
57
119
106
118
81
112
101
94
93
89
87
100
57
a 8.6 vr of air passed  through  system
Third Vapor Trapping Experiment

    A further experiment on vapor  trapping was  conducted under  intensified
conditions.  Chamber temperature was  increased  again,  to 52<>C,  and  air was
passed through  the apparatus  for 8 hours  at 3.88 m^/hr, a  total volume of
31 m3.  The air was prefiltered through two ester-form PUF plugs before
entry into the  sampling chamber.  Two similar plugs were again  used as traps.
The source and  plugs were extracted,  chromatographed and quantitated as
before, and the results are listed in Table 4.

    Total recoveries of all organochlorine and  organophosphorus pesticides
were quantitative or nearly so, showing that a  10-cm depth of PUF  is an  effic-
ient trap for these compounds.  For the most part, the carbamates  again  re-
mained on the source.  However, the small amounts of carbofuran and methiocarb
                                      13

-------
TABLE 4.  ESTER-FORM PUF TRAPPING EFFICIENCY AT 52°Ca

Compound
a-BHC
Aldrin
p , p ' -DDE
o,p'-DDT
p , p ' -ODD
p,p'-DDT
Lindane
g-BHC
Heptachlor epoxide
Dieldrin
Endrin
Methyl parathion
Ronnel
Parathion
Ethion
C ar b oph eno th ion
Diazinon
Malathion
Propoxur
Carbaryl
TABLE 4 (continued)
Residues,
On source
0
0
0
0
0
2
0
5
0
0
3
0
5
0
0
0
5
3
82
78

% of applied pesticides
On plugs
92
106
99
101
88
104
95
98
88
83
98
95
102
88
96
82
90
90
10
0


Total
92
106
99
101
88
106
95
103
88
83
101
95
107
88
96
82
95
93
92
78

                          14

-------
                               TABLE 4 (continued)
                                    Residues, % of applied pesticides
      Compound               On source             On plugs               Total
Carbofuran
Aminocarb
Methiocarb
Mexacarbate
88
68
91
62
8
5
3
2
96
73
94
64
 a  31.0  w?  of air passed through system

that did evaporate were apparently trapped essentially quantitatively,  sugges-
ting that PUF is also a suitable trapping agent for vapors of these carbamates.

Aerosol Generation

    Since we had not been able  to volatilize the carbamate insecticides  in
amounts sufficient to allow .PUF evaluation, several experiments were performed
with pesticide solutions injected into air as aerosols.  To do this, a  test
chamber was  fabricated  from aluminum sheet.  The chamber and associated  equip-
ment are illustrated in Figure  I/ and a closeup of the aerosol generating as-
sembly  is shown  in Figure 2.  The pesticide solutions were contained in  the
graduated glass  chamber and the aerosols were generated by aspiration with a ,
stream  of nitrogen through the vertical nozzle.  The aerosol spray at the
bottom  aperture  of the  test chamber moved upward, along with air flowing up
through the  chamber and with PUF plugs mounted in a glass holder at
the top of the chamber  to trap  the flowing aerosols.  A major shortcoming  of
this apparatus was that the sample air could not be prefiltered  to remove  con-
taminants .

First Aerosol Trapping  Experiment

    The first two tests of aerosol trapping were conducted with  an
experimental aerosol bomb in lieu of the aerosol generator of Figure 2.   In
the first test,  the bomb was loaded with a hexane solution containing  5 yg of
each of 24 pesticides,  plus about 100 ml of Freon 12,  and was then screw-jack
mounted on its side for spraying upward into the test  chamber.   The  assembly
was placed outdoors, at 24°C air temperature and 80% relative humidity.  One
5-cm-thick ester-form PUF plug  was installed as the trap.  Five-second  bursts
of aerosol were  sprayed into the chamber every 15 seconds, for  about  15
minutes.  The Freon gas was vented for an additional 15 minutes  without
                                       15

-------
aerosol formation, then air was passed continuously  through  the  system  for one
hour.  Total air flow (1.5 hours) was 8.5 w?.  The chamber was washed down
with 300 ml acetone, the bomb residues were  recovered by washing with acetone,
and the plug was Soxhlet-extracted as usual.  Each of the three  extract
solutions was concentrated, separated into three  fractions on a  silica  gel
column, and quantitated by GLC.  The results are  listed in Table 5.

    With the exception of certain of the carbamates, recoveries  of  pesticides
from the system were low, so that the effectiveness  of the PUT trap could not
be adequately judged.  Substantial amounts of material remained  in  the  bomb,
and there were residues on the chamber walls for  all but the most volatile
compounds.  In general, recoveries were poorest for  the organophosphorus
compounds, best for the carbamates.  Results for  the carbamates  suggested that
ester-form PUF may be an effective trap for  these pesticides when in airborne
aerosols.  Volatilization and inefficient trapping of vapors of  the most
volatile compounds may have contributed to the poor material balance in  this
experiment, as suggested by the particularly low  recovery for a-BHC.  Loss of
a liquid chromatographic fraction (Fraction  I) during analysis may  also  have
                                           in.

            Figure  1.   Apparatus  for aerosol trapping experiments

                                       U

-------
contributed to this.  Other factors that may have decreased pesticide recovery
include oxidation, hydrolysis, and degradative reactions catalyzed by the
aluminum surface of the test chamber.  If there had been no volatilization nor
degradation during the experiment, all the pesticides should have had a
constant ratio of bomb:plug:chamber-wash residues.  It is obvious that they
did not.

                                                          '
                   Figure  2.   Closeup of aerosol generator

-------
               TABLE 5.  THE TRAPPING OF PESTICIDE AEROSOLS BY A
                          5-CM ESTER-FORM PUF PLUGa
Residues, % of applied pesticides
Compound
a-BHC
Aldrin
p.p'-DDE
o,p'-DDT
p,p'-DDD
p,p'-DDT
Lindane
B-BHC
Heptachlor epoxide
Dieldrin
Endrin
Methyl parathion
Ronne 1
Parathion
Ethion
Carbophenothion
Diazinon
Malathion
Propoxur
In
bomb
9.0
17.8
19.8
24.6
18.6
45-7
12.0
18.5
13.6
14.3
20.7
15.2
13.0
16.3
15.1
6.3
8.9
1.9
18.0
In chamber
washings
—
-
-
-
-
-
5.4
3.7
10.6
7.4
23.3
4.9
4.3
4.2
1.9
1.2
2.8
0.6
20.0
In PUF
trap
24.8
39.7
29.4
32.0
25.7
43.9
46.2
40.0
33.8
28.3
43.5
27.4
32.9
27.3
24.8
17.2
13.2
2.5
25.0
Total
133.8
157.5
149.2
156.6
144.3
189.6
63.6
62.2
58.0
50.0
87.5
47.5
50.2
47.8
41.8
24.7
24.9
5.0
63.0
TABLE 5 (continued)
                                      18

-------
                              TABLE 5  (continued)
                                      Residues,  % of  applied  pesticides
Compound
Carbaryl
Carbofuran
Aminocarb
Methiocarb
Mexacarbate
In
bomb
17.0
27.0
-
14.5
34.5
In chamber
washings
2.4
19.6
21.2
4.4
8.0
In PUF
trap
25.0
57.0
87.0
22.0
76.0
Total
44.4
103.6
108.2
40.9
118.5
a 8.5 m-* of air passed  through  system

Second Aerosol Trapping Experiment

    The preceding experiment was repeated, except  that  two ether-form  (rather
than ester form) PUF plugs were placed  in  series as  traps, flow rate was
4.82 m^/hr, temperature was 25°C, and relative humidity was 35%.  The
analytical results are  listed in Table  6.

    Carbofuran and aminocarb could not  be  quantitated because of  a  large
interfering peak in the chromatograms,  possibly a  result of a contaminant
introduced during filling of the aerosol bomb.  In general, total recoveries
of the remaining pesticides were low.   Substantial amounts of all pesticides
were trapped on the second plug, strongly  suggesting that at least  part of  the
poor recoveries was caused by passage of a portion of the aerosols  through
both plugs.  This is in contrast to  the results with molecular vapors, which
were adequately trapped on two  plugs (Tables  2, 3  and 4).  The ratio of
residue levels on the second plug to those on the  first was relatively
constant (last column, Table 6), suggesting that pesticide degradation was  not
responsible for the low residue levels  on  the first  plug.  These  levels were
generally somewhat lower than those  found  on  the single plug of the previous
experiment (Table 5), which may be a result of the use  of ether-form rather
than ester-form PUF, or may have been caused  by the  lower relative  humidity in
the second experiment.

Third Aerosol Trapping Experiment

    Following the experiment above,  two additional tests were conducted  on  the
N-methylcarbamate insecticides  alone, one  outdoors and  one with the apparatus
contained in a large environmental chamber.   In these,  the bomb used  in  the

                                      19

-------
              TABLE 6.  THE TRAPPING OF PESTICIDE AEROSOLS BY TWO
                          5-CM ETHER-FORM PUF PLUGS*
Residues, % of applied pesticides
Compound
a-BHC
Aldrin
p,p'-DDE
o,p'-DDT
p.p'-DDD
p.p'-DDT
Lindane
3-BHC
Heptachlor epoxide
Dieldrin
Endrin
Methyl parathion
Ronnel
Parathion
Ethion
Carbophenothion
Diazinon
Malathion
In
bomb
10.5
17.0
15.2
19.3
16.0
23.6
20.6
11.1
13.0
10.2
14.8
15.2
13.0
16.2
15.1
6.3
7.0
15.3
In
chamber
washings
1.0
2.2
3.7
5.3
4.2
6.7
1.9
14.4
18.4
13.5
24.2
5.0
4.3
4.2
1.8
1.2
2.4
3.9
On
first
plug
17.1
26.8
21.3
24.1
21.7
27.8
36.5
34.6
34.5
30.4
47.5
27.5
33.0
27.4
24.8
17.2
9.8
23.8
On
second
plug
6.4
11.7
14.2
16.5
14.6
18.9
12.9
26.7
25.7
19.7
29.8
16.8
19.4
18.9
15.7
11.7
5.6
15.4
Total
35.0
57.7
54.4
65.2
56.5
77.0
71.9
86.8
91.6
73.8
116.3
64.5
69.7
66.7
57.4
36.4
24.8
58.4
Residue
ratio on
plugs:
2/1
0.37
0.44
0.67
0.68
0.67
0.68
0.35
0.77
0.74
0.65
0.63
0.61
0.59
0.69
0.63
0.68
0.57
0.65
TABLE 6 (continued)
                                      20

-------
                              TABLE 6. (continued)
Residues, % of applied pesticides


Compound
Propoxur
Carbaryl
Carbofuran
Aminocarb
Methiocarb
Mexacarbate


In
bomb
15.7
12.8
-
-
12.0
10.0

In
chamber
washings
6.2
2.4
-
-
3.6
20.0

On
first
plug
28.0
24.0
-
-
18.5
26.5

On
second
plug Total
11.0 60.9
15.2 54.4
-
-
11.5 45.6
9.0 65.5
Residue
ratio on
plugs:
2/1
0.39
0.63
-
-
0.62
0.34
a 7.2 m3 of air passed through system

first two aerosol experiments was replaced with the aerosol generator
(Figure 2).  Neither experiment yielded usable results, owing to contamination
of chromatograms that prevented quantitation of the carbamates.

    A successful experiment was then performed, with the apparatus moved to a
second environmental chamber, using a mixture of organochlorine and organo-
phosphorus pesticides.  Three high-flow ester-form PUF plugs were used  in
series as traps.  Ten ml of hexane containing 5 pg of each of the 18 compounds
was aspirated into the chamber by a stream of nitrogen, and washed into the
air stream in the chamber with 3 x 15 ml hexane.  The conditions were:
29°C, 75% R.H., and 2.81 m3/hr flow rate.  The pesticides were introduced
into the chamber in about 15 minutes and the air sampling was continued for 30
minutes, so that total air volume sampled was 2.1 m3.  Extracts of the
various substrates were concentrated, separated by silica-gel liquid chromato-
graphy, and quantitated by GLC.  Results are shown in Table 7.

    Large losses of material occurred throughout this experiment.  The  first
PUF plug generally had the highest residue levels, but there were measurable
quantities of all the pesticides on the second and third plugs.  The presence
of residues on the third plug after passage of only 2.1 nr of air contrasts
sharply with the quantitative trapping of molecular vapors by two ester-form
PUF plugs (10 cm) after passage of 55 m3 of air (Table 2).  This again
implies that PUF, at least under certain conditions, may not trap aerosols  as
readily as vapors.
                                       21

-------
           TABLE 7.  THE TRAPPING OF PESTICIDE AEROSOLS BY THREE
                        5-CM ESTER-FORM PUF PLUGS*
Residues, % of applied
Compound
oe-BHC
Aldrin
p,p'-DDE
o,p'-DDT
p.p'-DDD
p,p'-DDT
Lindane
6-BHC
Heptachlor epoxide
Dieldrin
Endrin
Methyl parathion
Ronnel
Parathion
Ethion
Carbophenothion
Diazinon
Malathion
In
chamber
washings
0.8
1.4
10.4
13.7
10.0
13.7
3.0
7.3
10.0
17.4
14.3
7.1
5.5
10.8
9.6
30.2
15.4
16.2
On
first
plug
19.3
24.6
10.5
8.2
4.8
4.8
30.2
16.7
15.4
14.2
13.3
13.1
17.0
9.3
3.8
4.4
18.5
7.4
On
second
plug
0.5
1.0
2.0
2.7
2.4
3.3
1.5
1.2
2.5
2.1
1.7
22.1
12.2
1.3
2.3
2.2
20.8
1.7
pesticides
On
third
plug
0.6
0.7
2.5
2.9
2.9
3.4
2.0
2.3
3.4
5.6
2.7
2.4
18.5
3.1
3.9
4.4
2.2
3.9

Total
21.2
27.7
25.3
27.6
20.1
25.2
36.7
27.5
31.3
39.3
32.0
44.7
53.2
24.5
19.6
41.2
56.9
29.2
2.1 m3 of air passed through system
                                    22

-------
Fourth Aerosol Trapping Experiment

    To compare directly the  performance  of  the  two  forms  of PUF,  a replicate
of the previous experiment was  run,  with the  exception that ether-form,  rather
than ester-form, plugs were  used.  The conditions were: 29°C,  75% R.H.,  and
2.88 m^/hr flow rate.  The analytical results are listed  in Table 8.

    Poor recoveries  again occurred.  The degree of  trapping of the organo-
chlorine insecticides agreed well with that of  the  previous experiment,
whereas trapping of  some of  the organophosphorus compounds  was somewhat  less
efficient.  In general, the  two forms of PUF  apparently do  not differ  greatly
in their ability to  trap pesticides  in aerosols.
             TABLE  8.   THE  TRAPPING OF PESTICIDE AEROSOLS  BY THREE
                          5-CM ETHER-FORM PUF PLUGS3
Residues, % of applied
Compound
a-BHC
Aldrin
p,p'-DDE
o,p'-DDT
p,p'-DDD
p,p'-DDT
Lindane
e-BHC
Heptachlor epoxide
Dieldrin
Endrin
Methyl parathion
Table 8 (continued)
In
chamber
washings
0.4
1.5
13.8
16.7
13.7
13.9
4.7
8.7
18.2
24.8
23.4
11.4

On
first
plug
16.0
28.5
12.8
9.3
4.0
5.3 .
27.7
16.3
14.1
12.4
12.4
9.8

On
second
plug
n.d.
0.4
1.1
1.7
1.3
2.2
0.4
0.6
0.8
1.7
3.4
0.8

pesticides
On
third
plug
0.2
0.3
0.9
1.4
0.9
1.9
0.4
0.2
0.6
1.6
1.4
0.7


Total
16.6
30.7
28.6
29.1
19.9
23.3
33.2
25.8
33.7
40.5
40.6
22.7

                                       23

-------
                               TABLE  8  (continued)
                                   Residues,  %  of applied pesticides


Compound
In
chamber
washings
On
first
plug
On
second
plug
On
third
plug


Total
Ronne1

Parathion

Ethion

Carbophenothion

Diazinon

Malathion
10.5

14.4

13.5

16.6

37.1

30.2
16.2

 6.5

 3.3

 3.6

12.8

 5.5
0.6

0.7

1.3

1.6

0.6

0.8
0.7

0.8

1.3

1.4

0.5

0.8
28.0

22.4

19.4

23.2

51.0

37.3
a 2.2 m-' of air passed  through  system
Fifth Aerosol Trapping Experiment

    Because pesticide recoveries were  consistently  poor  in  the  aerosol  tests
conducted  in the  large aluminum test chamber, a  smaller  glass-chamber appar-
atus was fabricated  in this Laboratory by Mr. H. P. Freeman.  This  was  a modi-
fied Rappaport-Weinstock apparatus, consisting of a 44-mm ID  glass  tube with a
glass cup  on the  bottom, a stainless-steel  1.6-mm OD  tube connecting an ex-
terior sample solution reservoir to a  nitrogen-gas-driven aerosol generator
located above the cup, and space at the  top of the  tubular  chamber  for  five
5-cm-thick PUF plugs.  Dry nitrogen gas  was introduced under  slight pressure
through a  tubulation in the center of  the bottom cup  and a  vacuum pump  was
used to equalize  pressure in the chamber.   This  apparatus allowed the genera-
tion and trapping of aerosols without  exposure to air, so that  oxidation,
hydrolysis, or contamination from ambient vapors or aerosols  could  not  occur.

    In a three-hour experiment with this apparatus  at a  nitrogen flow rate of
3.43 m-Vhr, the trapping of eight organochlorine and  five organophosphorus
insecticides by five ester-form plugs  in series  (25 cm of foam)  was studied.
The plugs  were extracted and analyzed  individually  and the  results  for  the
organochlorines are  listed in Table 9.   Results  for the  organophosphorus com-
pounds are not tabulated because extracts for all but the first  plugs were
lost.
                                      24

-------
              TABLE 9.  THE TRAPPING OF PESTICIDE AEROSOLS BY FIVE
                          5-CM ESTER-FORM PUF PLUGS*
Compound
Aldrin
p,p'-DDE
o,p'-DDT
p.p'-DDD
p,p'-DDT
Heptachlor ep oxide
Dieldrin
Endrin
Residues, % of applied pesticides
In
chamber On plug No.
washings 12345
nd 102.9 2.8 1.8 1.4 1.0
3.6 81.6 2.7 1.9 1.0 2.2
5.1 92.8 3.8 2.7 2.9 2.1
7.1 65.7 nd nd nd 2.2
9.7 70.6 5.6 4.2 3.6 3.9
nd 95.7 1.7 1.3 1.0 1.4
nd 87.3 4.9 3.6 2.0 4.2
nd 83.6 4.8 3.0 1.5 2.6

Total
109.9
93.0
109.4
75.0
97.6
101.1
102.0
95.5
a 10.3 m^ of air passed through system
    Trapping of all the organochlorine pesticides was almost quantitative on
the first plug, but there were detectable residues on all the plugs.  This
indicates that under these conditions small amounts of the pesticide aerosols
penetrated the first plug and were trapped on later plugs.  In this experi-
ment, 25 cm of ester-form PUF was sufficient to trap organochlorine aerosols
essentially quantitatively.  There were substantial but less than quantitative
amounts (60 - 75%) of the organophosphorus pesticides trapped on the first
plug; presumably, succeeding plugs would also have contained measurable
residues of these compounds.
Summary and Analysis of Pesticide Trapping Experiments

    A number of conclusions may be drawn from the PUF trapping
(Tables 2-9).  It is apparent that a 10-cm depth of  foam  is an
for vapors of organochlorine and organophosphorus pesticides.
the test pesticides from ca. 10 nP of air, and of many of them
as 55 m^ of air, was essentially quantitative.  The  ester form
somewhat, but not significantly, superior to the ether form in
efficiency.  The N-methylcarbamate insecticides were extremely
experiments
efficient trap
Trapping of all
from as much
of PUF was
its trapping
difficult to
                                     25

-------
vaporize, but the small amounts  that did vaporize  were  apparently efficiently
trapped on PUF.

    When the pesticides were injected  into  the  air as aerosols,  there were
indications that PUF trapping ability  was substantially lower  than with
molecular vapors.  Although most of the residues were found  on the first plug,
measurable amounts of all the test pesticides occurred  on  backup plugs, even
when as many as 5 plugs (25-cm total depth  of PUF)  were placed in series to
trap the aerosols.  The pattern, high  residue levels on first  plug and much
lower levels on succeeding plugs, is consistant with three possible
interpretations, but the early termination  of the  experimental program did not
allow disclosure of which was operative.  The possibilities  are:   (a) most of
the generated aerosols had vaporized by the time they reached  the trapping
plugs, so that vapors were trapped on  the first plug and aerosols, less
efficiently trapped, passed through; (b) hexane solvent carried  into the
aerosols desorbed small amounts  of the pesticides  as it passed through the
plugs; or (c) plugs were contaminated  before the experiments with low levels
of the pesticides not removed in earlier experiments with  the  same plugs,  or
adsorbed from laboratory air.

    One possible reason for poor trapping efficiency of aerosols by PUF is
electrostatic repulsion.  Polymers such as  PUF  tend to  have  high dielectric
constants, thus can acquire significant electrostatic charges.   Moreover,
droplets disrupted from an organic liquid such  as  hexane by  atomization
acquire charges, and the charge  to mass ratio increases as the solvent
evaporates (7).  If the charges  on the droplet  and PUF  surfaces  were the same,
electrostatic repulsion would result and trapping  efficiency would
consequently diminish.  Humidity in air tends to dissipate electrostatic
charges, so that the charge effect would be accentuated under  conditions of
low humidity.

    The low recoveries of pesticides in the large-chamber  aerosol experiments
and the differences in recoveries between compounds may have been caused by
any of a number of physical and  chemical processes.  Among the latter are
oxidation, hydrolysis, rearrangement,  and catalytic degradation.   These are
especially important for the organophosphorus pesticides.  Hydrolysis would
cause a total loss from chromatograms  obtained  with the flame  photometric
detector (FPD), since the hydrolysis products are  nonvolatile  phosphate
acids.  Oxidation and rearrangement can give volatile products that would  not
have the same GLC retention times as the parent organophosphorus molecules.
For example, it has been shown that the oxidation  product  and  the
rearrangement product of parathion (paraoxon and S-ethyl parathion,
respectively) have shorter retention times  than the parent molecule on 3%
OV-1, and that the three compounds are separable (8).   Examination of the  FPD
chromatograms in some of the experiments did in fact show  a  number of new,
small, short-retention-time GLC  peaks.  That these chemical  transformations
were quite important to pesticide recovery  was  illustrated by  the high
recoveries in the one experiment conducted  under conditions  in which they
could not occur (Table 9).
                                       26

-------
    Among the physical processes  that may have  contributed  to  losses  are
imperfect injection of aerosol into  the chamber,  loss  of  droplets  through
chamber apertures, imperfect washing of residues  off chamber walls, and
aerosol passage through all the plugs because of  poor  adsorptivity, as
discussed above.  The contribution of each of these processes  to losses or
differences in recovery is unknown.
EVALUATION OF PESTICIDE EXTRACTION METHODS

Comparison of Three Extraction Techniques

    Considerable exploratory work in  analytical methodology was  conducted
throughout our laboratory program, looking  to  development  of  the optimum
method for use of PUF as a  trapping agent in a national  airborne pesticide
monitoring program.  An early phase of  this involved  a search for the best
technique for extracting a  wide variety of pesticides from PUF plugs.  Three
replicate experiments were  performed, in which extractions of the 24 test
pesticides by Soxhlet treatment, mechanical squeezing, and simple elution were
compared.  In each experiment, six ester-form  PUF  plugs  were  spiked directly
with all 24 compounds in hexane at 1  yg of  each compound per  plug.  The hexane
was allowed to evaporate, then two of the plugs were  Soxhlet-extracted with
250 ml of 1:1 hexane-acetone for five Soxhlet  cycles; two  were extracted by
squeezing each plug five times with a metal plunger in 50  ml  1:1
hexane-acetone and repeating this four  times with  fresh  solvent,  for a total
of 250 ml of extract solution; and the  remaining two  were  extracted by passing
250 ml of .1:1 hexane-acetone through  each plug at  a rate of about 4 ml/min.
All extract solutions were  concentrated, separated into  fractions by liquid
chromatography according to Sherma and  Shafik  (4), and analyzed  by GLC.
Results of the experiments  are summarized in Table 10.

    Several samples were lost during  the many  analytical manipulations
involved, but clear-cut findings were nevertheless obtained.   For both the
organochlorine and organophosphorus insecticides,  recoveries  were little
different among the three techniques, but Soxhlet  extraction  gave the most
consistent results for classes of compounds and, at least  for the
organochlorines, showed the least variation in replicate determinations of
individual compounds.  Results by the squeezing and the  elution  techniques
were not appreciably different from each other with respect  to both extent  and
precision of pesticide recovery^  The very  high and variable  recovery  of  the
early-eluting a-BHC by all  methods reflects the difficulty in quantitating  on
a sharply sloped baseline.  Results for the carbamates showed that recoveries
by the Soxhlet technique were not only  distinctly  higher than by the other
techniques, but were also the most consistent  for  the class.   Variations  in
recoveries of individual compounds, although large, were no greater  than  those
obtained by squeezing or elution.

    Cumulative results for  all 24 pesticides (last line, Table 10) indicated
that 5-cycle Soxhlet extraction with  1:1 hexane-acetone  gave  the highest  and
most consistent recoveries, and it was  consequently selected  as  the  method  of
choice in the other phases  of the experimental program.  The  limitations  to 5


                                      27

-------
cycles and the selection of 1:1 hexane-acetone as the preferred solvent were
based on additional exploratory work described in detail in the following
sections.
            TABLE 10.  COMPARISON OF THREE TECHNIQUES FOR EXTRACTION
                        OF PESTICIDES FROM PUF PLUGS3
% Recovery by
Soxhlet
extraction
Compound
a-BHC
Aldrin
p.p'-DDE
o , p ' -DDT
p , p ' -ODD
p.p'-DDT
Lindane
g-BHC
Hept. epox.
Dieldrin
Endrin
Organochlorines
Me parathion
Ronne 1
Meanb
123.0
96. 3C
106. lc
95. 6C
96. 7C
93. 9C
82.0
85.0
89.4
99.2
94.6
96.5
85.3
90.3
Standard
deviation
35.1
8.1
9.9
10.6
4.2
8.5
5.7
18.9
10.2
7.1
15.3
11.0
24.2
6.2
Mechanical
squeezing
Meanb
128.3
108.5
104.9
100.2
131.5
107.9
86.3
76. 7e
85. 4e
94. Oe
84. 6e
101.4
84. 6C
73. 2C
Standard
deviation
23.3
10.1
14.2
12.9
23.4
17.8
10.2
14.1
22.7
11.7
9.6
17.2
9.6
2.8
Elution
Meanb
128.9
105.4
101.1
96.3
124.0
105.9
86.9
73. 4d
73. 5d
95. 2d
96. 5d
98.8
79.1
68.8
Standard
deviation
27.6
18.6
15.3
9.7
26.8
16.1
8.2
9.9
16.6
12.6
11.0
17.6
13.0
8.8
Table 10 (continued)
                                      28

-------
                              TABLE 10 (continued)
% Recovery by
Soxhlet
extraction
Compound
Parathion
Ethion
Carbophenothion
Diazinon
Malathion
Organophosphates
Propoxur
Carbaryl
Carbofuran
Aminocarb
Methiocarb
Mexacarbate
Car hamate s
All pesticides
Meanb
76.6
99.8
99.1
91. ld
95. 5d
91.1
91. 9d
97. 8d
81. 9d
85.3d
102. 5d
93. ld
92.1
93.8
Standard
deviation
18.9
8.7
5.3
6.9
5.0
8.2
13.0
20.5
26.3
10.5
3.3
25.1
7.6
9.4
Mechanical
squeezing
Meanb
89. 2C
91. Oc
99. 8C
89. 4d
79. 6d
86.7
80. 8d
56. 6d
73. 4d
53. 9d
88. 5d
36. 9d
65.0
88.0
Standard
deviation
5.0
4.2
7.2
4.6
2.6
8.6
19.4
11.8
11.9
19.4
10.3
16.0
19.3
21.3
Elution
Meanb
91.4
96.8
102.5
91. 9d
87. 6d
88.3
77.3d
66. 7d
70.3d
89. Od
73. 8d
63. ld
73.4
89.4
Standard
deviation
7.8
11.2
15.9
7.1
10.3
11.3
24.0
26.0
28.6
22.9
18.9
28.1
9.2
17.2
a Ester-form polyurethane
  Six replicates, unless otherwise specified
c Five replicates
  Four replicates
  Three replicates
                                      29

-------
Development of the Soxhlet Procedure

    Because certain of the organophosphorus pesticides may  degrade  under
prolonged high-temperature conditions such as  those  encountered  in  Soxhlet
extraction, systematic study was made to determine the optimum number of
Soxhlet cycles for extraction of these compounds  from PUF plugs.  Twelve
ester-form plugs were precleaned by Soxhlet extraction with acetone for  850
cycles, then each plug was spiked with a hexane solution containing lyg  of
each of the 7 test organophosphorus pesticides.   The plugs  were  then dried and
placed individually in Soxhlet extractors.  All plugs were  extracted with
280 ml of 1:1 hexane-acetone, but the extractions were stopped after different
numbers of Soxhlet cycles ranging from 1 through  12.  The extracts  were
evaporated in Kuderna-Danish flasks with five  drops  of paraffin  oil keeper
solution, dissolved in 1 ml acetone, and quantitated on a 3% OV-1 column with
flame photometric detector.  Recoveries are listed in Table 11.

    Extraction of all the test compounds proceeded very rapidly, reaching
quantitative level in only 2 cycles.  At 3 cycles, recoveries  were  not only
quantitative, but also consistent for the class of compounds.  Recoveries  only
for the class remained essentially quantitative up to 9 cycles,  but the
variability increased with increased cycling,  a reflection  of  decreasing
recoveries of parathion and methyl parathion.  The data were subjected to
least-squares regression analysis, with the postulation that change in
recovery is a first-order process with respect to time, and thus with respect
to number of extraction cycles.  The operative relation is:

                                 In y = a + b  x

where y_ is the percent recovery, jc is the number  of  extraction cycles, b_ is
the slope of the regression line, and £ is the natural logarithm of the
intercept of the line.  The coefficients of this  equation are  shown in Table
12, along with correlation coefficients, r.  The  regression analysis
statistically confirmed the observation that parathion and  methyl parathion
were being lost with prolonged Soxhlet extration, as illustrated by the  values
of the slopes of the regression lines.  The negative values for  parathion  and
methyl parathion are significant at the 95% confidence level.

    In view of these results, Soxhlet extraction  for from 3 to 7 cycle.s
appeared to be the optimum range of duration of treatment,  and 5 cycles  was
consequently selected as the technique used in other phases of our
investigations.  Five cycles would not produce significant  loss  of  the labile
organophosphorus pesticides, while assuring quantitative recovery of other
organophosphorus compounds, carbamates, and organochlorine  pesticide residues
on the PUF plugs.

Choice of Extraction Solvent

    A large number of pure solvents and combinations of solvents can be  used
for the extraction of pesticide residues from  PUF.   However, to  keep the work
within practical bounds, it was necessary to choose  a small number  of likely
candidates for testing.  The solvents had to be commercially available in


                                      30

-------
TABLE 11.  THE EFFECT OF NUMBER OF SOXHLET EXTRACTION
           CYCLES ON RECOVERIES OF ORGANOPHOSPHORUS
           PESTICIDES FROM POLYURETHANE FOAM PLUGS

% recovery
Compound
Diazinon
Me parathion
Ronnel
Malathion
Parathion
Ethion
Carbophenothion
Average
Std. deviation
1
90
83
90
89
86
90
88
88
2.6
2
105
95
102
105
95
108
103
102
5-0
3
98
100
99
99
95
100
100
99
1.8
1*
97
85
97
101
91
97
100
95
5.6
. 5
110
98
109
100
9^
100
102
102
5.8
at Soxhlet cycle No.
7
103
95
106
107
96
91
99
100
6.0
8
103
90
105
107
8U
99
98
98
8.5
9
109
95
10 U
105
95
97
102
101
5.H
10
91
75
97
99
77
101
100
91
11.0
11
108
86
105
10U
8U
100
99
98
9.U
12
97
78
101
101
^
97
103
9".
9.8

-------
               TABLE  12.  REGRESSION AND  CORRELATION COEFFICIENTS
                    OF SOXHLET EXTRACTION DATA IN TABLE 11.
                          	Coefficients
    'Compound                     a                      b
Diazinon                    4.6377(103.3)a            -0.0019         -0.107

Methyl  parathion            4.6240(101.9)             -0.0185         -0.661*

Ronnel                      4.6220(101.7)             +0.0010         +0.095

Malathion                   4.6269(102.3)             +0.0007         +0.084

Parathion                   4.6022(99.7)              -0.0157         -0.701*

Ethion                      4.6250(102.0)    '         -0.0040         -0.353

Carbophenothion            4.6151(101.0)             -0.0006         -0.120
 a  Numbers  in parentheses  are  antilogs  of  a,  and  are  intercepts  expressed
     as  % pesticide  recovery

adequate purity for pesticide residue analysis, not be unduly hazardous, be
inert to the pesticides, be readily distilled and evaporated, and be
reasonably priced.  Alcohols, ethers and aromatic solvents were rejected
because they could not meet one or more of these criteria.  Hexane and
acetone, by contrast, were selected for study because they not only met  the
criteria, but also had a history of use as pesticide extractants.  Moreover,
they differed greatly with respect to polarity, so that they provided needed
contrast in dissolving power.  Hexane, being relatively nonpolar, is a less
powerful solvent than the more polar acetone.  Tests were conducted with
pesticide-grade" hexane, pesticide-grade acetone, and a 1:1 mixture of the
two.  In the interests of time, experiments were limited to extractions  of the
seven test organophosphorus insecticides from ester-form PUF plugs,
circumventing the need for the liquid chromatographic separation step of the  '
Sherma-Shafik procedure.

    In  the first experiment,  six ester-form  PUF  plugs were precleaned by
individual Soxhlet extraction with hexane  for  17 hours at about  5  cycles per
hour, then were dried by  drawing PUF-filtered  air  through for about  an hour.
Three of the plugs were each  spiked with  1 ml  of hexane containing 0.1 yg  of
each of the  seven OP pesticides, and the  other three were spiked with 1  ml of
hexane  containing 1.0 yg  of each pesticide.  The plugs were  placed
individually in size E Soxhlet extractors  (Kontes Glass Co.) and left open for
15 min. to evaporate the  hexane.  The  extractors were then fitted  to 300-ml
boiling flasks and 250 ml of hexane, acetone,  or 1:1 hexane-acetone  was  poured
through the plugs.  Temperatures of the Soxhlet heaters were adjusted to give

                                      32

-------
solvent distillation rates of  about  20  minutes  per extraction cycle.   After
five cycles, each solvent was  cooled to about 35°C and  poured into a
Kuderna-Danish (K-D) flask fitted with  a 15-ml  receiver tube.  A boiling chip
and 5 drops of a 1% paraffin oil-in-hexane  keeper  solution were added to each
flask, Snyder columns were fitted to the flasks, and  the flasks were  heated
gently in a steam bath until solvent loss through  the Snyder columns  stopped.
The flasks were removed  from the  steam  bath and, after  cooling, about 5  ml  of
residual solution remained in  each receiver.  This was  evaporated with gentle
heating in a stream of nitrogen  to 1 ml.  The solutions were quantitated with
the alkali-flame ionization detector, using a 3% OV-1,  180-cm column  at
200°C.  The elution times for  the organophosphorus compounds, as determined
with individual standards, were:  diazinon,  2.4; methyl parathion,  3.1;
ronnel, 3.5; malathion,  3.85;  parathion,4.15; ethion, 9.8;  and
carbophenothion, 11.1 minutes.

    The final concentrates obtained  from the acetone  and 1:1 hexane-acetone
extractions were yellow  oils that yielded chromatograms so contaminated  that
quantitation of the early peaks  of interest was not possible.  By contrast,
the hexane extracts were colorless and  yielded  acceptable chromatograms.
Results with hexane are  shown  in Table  13.   Recoveries  of 1 yg of the test
compounds were generally high  and consistent, whereas those at the 0.1 pg
level were considerably  lower  and more  variable.   Above-quantitative
recoveries of 0.1 ug of  malathion and ethion were  produced by background
interferences.

    The contaminated  extracts  obtained  with acetone and hexane-acetone in the
above  experiment were  evidence that  hexane-precleaning  of the PUF plugs  was
not adequate  except  for  further extraction with hexane.  Consequently, the
experiment was  repeated  with  plugs  that had been subjected to more intensive
pretreatment.   Six ester-form  PUF plugs were individually Soxhlet-extracted
with  acetone  for  170 hours  at  about  5 cycles per hour.   During this treatment,
the foam  plugs  swelled  somewhat,  showing that  the  more  polar acetone
penetrated  the polyurethane matrix  as well  as  passing through the open
channels  of  the PUF.   This  swelling  did not interfere with the subsequent
analytical process.  The cleaned plugs  were dried  by  drawing PUF-prefiltered
air through  the plugs  for  1 hour,  during which  time the plugs shrank to  their
original  dimensions.   The  procedure  of  the previous experiment was repeated,
and yielded  colorless  extracts and  usable chromatograms.  Thus, initial  plug
cleaning  by  Soxhlet  extraction with  acetone for 850 cycles was adequate,
whereas Soxhlet extraction with hexane  for 85  cycles  was not.  The results of
the repeat experiment  are  tabulated  in Table 14.

    The superiority of 1:1 hexane-acetone as a  pesticide extractant for  PUF
was apparent  in this experiment.  At the 1-jJg  spike level, recoveries of all
the test compounds were  essentially  quantitative with the mixed solvent,
whereas those with acetone were  somewhat more variable  and those with hexane
were not only more variable, but  also substantially less than quantitative.
At the 0.1-yg level, recoveries  with all the solvents were lower and  quite
variable, but the mixed  solvent  again was best. These  results led us to use
1:1 hexane-acetone as the solvent of choice in  the other phases of our
experimental program.
                                       33

-------
           TABLE 13.  RECOVERY OF ORGANOPHOSPHORUS INSECTICIDES FROM
               SPIKED PUF PLUGS BY SOXHLET EXTRACTION WITH HEXANE
                                          Percent recovery at
Compound                          1 vg per plug             0.1 yg per plug
Diazinon
Methyl parathion
Ronne 1
Malathion
Parathion
Ethion
Carbophenothion
115
85
102
99
88
92
91
86
60
66
155
57
120
101
Average                                96.0                        92.1

Standard deviation                     10.3                        36.0




 Performance of the Developed Method With PUF Extract Blanks

     Having identified the most suitable extraction technique,  duration of
 extraction, and solvent, we conducted a systematic study of the method of
 choice, 5-cycle Soxhlet extraction with 1:1 hexane-acetone, to determine
 whether the procedure would yield coextractives from PUF that  would interfere
 with analysis of any of the 24 test pesticides.  Each of the three fractions
 resulting from the silica-gel liquid-chromatographic cleanup step of the
 Shenna-Shafik analytical method (4) was examined.  Of the 24 compounds, 6
 organochlorines occur in Fraction I, 5 organochlorines and 5 organophosphates
 in Fraction II, and 2 organophosphates and the 6 carbamates in Fraction III.
 Specific compounds are indicated in the chromatograms that follow.  The
 chromatograms of PUF blank extracts and of extracts spiked with appropriate
 pesticide standard solutions can be used not only to illustrate the presence
 or absence of coextracted interferences to pesticide detection and
 quantitation, but also to evaluate the minimum detectability of each pesticide.

     Six ester-form PUF plugs were Soxhlet-extracted with acetone for 98 hours
 at about 5 cycles per hour (490 cycles).  To remove acetone, the solvent was
 decanted while the plugs were squeezed with a metal plunger.  Then, clean

                                       34

-------
           TABLE 14.   RECOVERY OF ORGANOPHOSPHORUS INSECTICIDES FROM
           SPIKED PUF PLUGS BY SOXHLET EXTRACTION WITH THREE SOLVENTS
Percent recovery at
Compound
Diazinon
Me parathion
Ronne 1
Malathion
Parathion
Ethion
Carbophenothion

Hexane
98
76
82
80
79
83
68
1 ug per
Hex-Acet
103
95
98
101
96
100
97
plug
Acetone
110
93
102
95
94
98
101

Hexane
105
51
72
90
38
82
55
0.1 jig per
Hex-Acet
91
82
65
121
73
111
95
plug
Acetone
128
37
85
105
65
81
89
Average
Std deviation
80.9
9-1
98.6
2.9
99.0
5.9
70.4
23.7
91.1
20.0
84.3
28.8
300-ml receiver flasks containing 280 ml  1:1 hexane-acetone were  fitted  to the
Soxhlets without removing the plugs, and  extraction was conducted  for 5
cycles.  The extract solutions were evaporated  to dryness with 5  drops of
keeper solution, taken up in 0.5 ml hexane, and chromatographed on silica
gel.  The resulting fractions were analyzed by GLC, spiked with appropriate
pesticide standard solutions, and analyzed again by GLC under the  same
conditions.

    The chromatograms for Fraction I, both blank and with spike,  are presented
in Figure 3.  Analysis was by 63flli electron-capture gas chromatography
(ECGC) on a 1.5% OV-17/1.95% OV-210 column.  The spike solution contained
0.1 yg of each of the 6 organochlorine  insecticides normally found in Fraction
I, plus 0.1 jig of lindane.  The latter  was included because it was
occasionally found in Fraction I as well  as in Fraction II, its normal
location.  Comparison of the two chromatograms  in Figure 3 shows  that Soxhlet
extraction of these plugs with acetone  for about 490 cycles constituted
adequate treatment for precleaning the  plugs.  No significant interferences  to
the quantitation of the organochlorine  compounds appear in Fraction I.   Each
of the peaks in Figure 3B represents 4  x  10~H  grams of the appropriate
pesticide.
                                      35

-------
                                                   B
                                                    1.   a-BHC
                                                    2.   Lindane
                                                    3.   Aldrin
                                                    k.   p,p'-DDE
                                                    5.   o.p'-DDT
                                                    6.   p,p'-DDD
                                                    7.   p,p'-DDT
Figure 3.  Fraction I chromatograms.   A.  PUF blank; B.   Spiked blank
                                    36

-------
    Because both, organochlorine  and  organopbosphorus pesticides occur in
Fraction II, PUF blanks were  chromatograpbed by both electron-capture and
flame-photometric GLC.  Chromatograms  typical of the six samples of each that
were run are presented in Figure 4.  In Figure 4A,  the column was 1.5%
OV-17/1.95% OV-210, and temperatures of inlet, column, and 63Ni-detector
were 250°, 205°, and  325°C, respectively.   Injection volume was 4 pi
from 15 ml of final extract.   The three small peaks that are shown at.
retention times  of  1.3, 1.8,  and 3.7 minutes appeared in the chromatograms  of
all six samples  at  about the  same level.   There was also a large peak,  not
illustrated, at  about 23 minutes. This peak did not occur in solvent blanks,
but did appear in all plug extracts, with  some variation in peak height.  In
figure 4B, the column was 3%  OV-1 and  5 yl of a 1-ml extract was injected.
There were no detectable peaks after the injection  peak.

    Two representative blank  extracts  were spiked with standard solutions
containing the five organochlorine and five organophosphorus pesticides  that
elute in liquid-chromatography Fraction II.  Spiked-blank solution volumes
were adjusted to the  original volumes  of the blank  extracts, then one solution
was analyzed by  ECGC  and the  other by  FPD-GC.  The  electron-capture chromato-
gram is reproduced  in Figure  5 and the flame-photometric chromatogram in
Figure 6.

    In Figure 5, each peak represents  0.27 ng of a  pesticide.   The first of
the small peaks  in  the blank  (Figure 4A) appears as a small shoulder at  the
leading-edge base of  the lindane peak, the second occurs between the peaks
for $-BHC and ronnel, and the third  is seen between the peaks for ronnel and
methyl parathion.   The large  late peak, designated  peak A, is illustrated in
this chromatogram.  Although  none of the background peaks interfere with the
peaks of interest,  the large  late peak does make it necessary to delay further
injections on the column until the peak has cleared the detector.  In
Figure 6, each of the peaks represents about 5 ng of the designated organo-
phosphorus pesticide. No interferences with quantitation are evident.

    A fraction III  PUF blank  eluate  off the silica-gel column was evaporated
to 0.5 ml, and 10 yl  of this  solution  was  analyzed  by flame-photometric  GLC.
In the chromatogram,  reproduced  in Figure  7A, there are two small, early
peaks.  A standard  solution containing diazinon and malathion, which would
normally appear  in  Fraction III, was added to the blank solution, the volume
was adjusted, and 10  yl of  this  solution was injected into the GC, with the
result depicted  in  Figure 7B. The first large peak represents 5 ng of
diazinon and the second, 10 ng of malathion.  Neither of the background peaks
interfered with  the pesticide peaks.

    The remaining five Fraction  III  blanks were evaporated with keeper solu-
tion and derivatized  with pentafluoropropionic anhydride (PFPA) according to
the method developed  for the  carbamate insecticides, as described in a later
section of this  report.  Electron-capture GC of the derivatized blanks gener-
ally showed a relatively broad injection peak and three small, early peaks.
One of the chromatograms, representing injection of 0.6 yl out of 5 ml of so-
lution onto a 3% DC-200 column,  is shown in Figure  8.  This sample, as may be
seen, produced several somewhat  larger peaks as well as a large, late peak,


                                       37

-------
Figure h.  Fraction II FUF blanks.  A.  Electron-capture chromatogram.
           B.  Flame photometric chromatogram

                                    38

-------
                                 1.  Lindane
                                 2.  g-BHC
                                 3.  Ronnel
                                 k.  Methyl parathion
                                 5.  Heptachlor  epoxide
                                 6.  Parathion
                                 7.  Dieldrin
                                 8.  Endrin
                                 9 -  Ethion
                                 10.  Carbophenothion
Figure 5.  Fraction II spiked blank—electron-capture
                             39

-------
                                       1.   Methyl parathion
                                       2.   Ronnel
                                       3.   Parathion
                                       h.   Ethion
                                       5.   Carbophenothion
Figure 6.  Fraction II spiked blank—flame photometer
                            40

-------
      f
r^^J
                                                    B
                                                             1.  Diazinon
                                                             2.  Malathion
        Figure 7«   Fraction III ehromatograms.
                   B.  OP-spiked blank
A. PUF "blank;
                                     41

-------
Figure 8.  Electron-capture chromatogram of PUF blank extract
           derivatized with PFPA
                              42

-------
and may be considered as  a worst  case.   A standard  solution containing the  8
compounds normally found  in Fraction III - 6  carbamates  and 2  organophos-
phates - was also derivatized with  PFPA and subjected  to ECGC  on a 3%  DC-200
column.  The chromatogram, representing an injection of  120 pg of each carba-
mate, is shown in Figure  9.  The  two organophosphorus  compounds produced  only
one peak, sufficiently  removed  that it  did not  interfere with  the carbamate
peaks.  To estimate  the possible  extent of interference  arising from the  PUF
background, the derivatized standard of Figure  9  and the derivatized blank  of
Figure 8 were simultaneously applied to the column  at  the same injection
volumes as before.   The chromatogram produced is  shown in Figure 10.   The
large, late peak does not interfere with any  of the carbamate  peaks, but  at
least two of the small  peaks do interfere.  Comparison of peak heights for  the
standard and spiked  blank shows that the peak sizes of aminocarb and
mexacarbate were affected by the  PUF blank.  The  aminocarb peak height was
increased about 48%, and  that of  mexacarbate  about  63%,  relative to the peak
heights in the standard.   Such  interferences, though random, do occur,  and
would have greatest  effect at very  low  carbamate  levels.  Consequently, false
positive results could  be obtained  unless identities of  the peaks were
confirmed by independent  means.

     In sum, our experiments show  that,  when the best analytical methodology is
used, pblyurethane foam does not  interfere with the quantitation of even  very
low  levels of airborne  organochlorine and organophosphorus insecticides,  but
that  some interference  with certain of  the N-methylcarbamates  may possibly
occur.
COMPARISON OF ESTER AND ETHER FORMS  OF PUF

    Experiments  were conducted to ascertain whether the  extractability  of  the
test  pesticides  from the two types of PUF differed  significantly.   In tests  of
recoveries from  pesticide aerosols,  the ester form  had yielded  slightly, but
not significantly,  superior results  (Tables 7 and 8),  which may have  reflected
differences  in extractability.

    Four  ether-form PUF plugs were spiked with all  24  pesticides,  two at
1.0 pg  of each compound per plug, and two at 0.1 pg per  plug.   In  addition,
two ester-form plugs were spiked at  the 0.1-yg level.  The six  plugs  were
extracted with 1:1  hexane-acetone for five Soxhlet  cycles  and  the  extracts
were  evaporated  with five drops of keeper solution, put  through liquid
chromatography on  silica gel, and quantitated by GLC.  The results for  the
0.1-pg  spikes are  recorded in Table  15 and those for the 1.0- yg spikes in
Table 16.  In the  latter table, the  results obtained earlier with  ester-form
plugs are shown  for comparison.

    Although average recoveries of all pesticides showed no significant
differences between the two forms of PUF at both spike levels,  close
examination of the  data reveals clear trends that indicate that extractability
of pesticides from  ester-form PUF is superior to that  from ether-form
material.  For example,  in Table 15, recoveries of  10  of the 24 compounds  from
ester-form PUF were in  the optimum range between 90 and  100%,  whereas
recoveries of only  3 compounds were  in that range with ether-form PUF.

                                      43

-------
                                1.  Propoxur
                                2.  Carbofuran
                                3.  Aminocarb
                                k.  Mexacarbate
                                5.  Carbaryl
                                6.  Methiocarb
                               OP.  Organophosphorus  peak
Figure 9-  ECGC chromatogram of Fraction III standard solution
           derivatized vith PFPA
                              44

-------
                                      1.  Propoxur
                                      2.  Carbofuran
                                      3-  Aminocarb
                                      h.  Mexacarbate
                                      5.  Carbaryl
                                     . £•_ Methipc_arb	
                                      A.  Background peaks
                                     OP.  Organophosphorus peak
Figure 10.  ECGC chromatogram of combined PUF blank and Fraction III
            standard, PFPA-derivatized
                                45

-------
        TABLE 15.   RECOVERIES OF PESTICIDE SPIKES (0.1 yG PER COMPOUND)
         FROM ESTER- AND ETHER-FORM PUF PLUGS BY SOXHLET EXTRACTIONS
Pesticide recoveries, %
Ester- form PUF
Compound
a-BHC
Aldrin
p , p ' -DDE
o.p'-DDT
p,p'-DDD
p,p'-DDT
Lindane
3-BHC
Hept. epox.
Dieldrin
Endrin
Me parathion
Ronnel
Parathion
Ethion
Carbophenothion
Diazinon
Ma lath ion
Propoxur
Repl. 1
84.0
86.9
93.0
100.0
87.4
109.7
97.3
87.8
82.4
86.0
57.9
88.2
94.9
97.2
94.8
77.3
94.4
104.8
80.0
Repl. 2
104.4
99.1
105.4
116.3
99.8
130.5
68.3
89.3
77.6
86.6
79.7
74.6
87.1
83.3
96.8
80.0
63.6
91.9
110.0
Avg.
94.2
93.0
99.2
108.2
93.6
120.1
82.8
88.6
80.0
86.3
68.8
81.4
91.0
90.3
95.8
78.7
79.0
98.4
95.0
Ether- form PUF
Repl. 1
145.1
163.4
187.7
190.0
187.5
201.0
88.5
84.6
76.2
83.5
70.1
87.5
100.0
86.6
88.9
88.9
-
-
187.5
Repl.
76.5
86.6
94.3
102.3
86.9
118.4
84.7
87.7
79.1
83.3
82.0
78.8
96.2
74.7
105.6
105.3
83.7
85.5
165.0
2 Avg
110.8
125.0
141.0
146.2
137.2
159.7
86.6
86.2
77.7
83.4
76.1
83.2
98.1
80.7
97.3
97.1
83.7
85.5
176.3
TABLE 15 (continued)
                                     46

-------
                              TABLE 15 (continued)
Pesticide recoveries, %
Compound
Carbaryl
Carbofuran
Aminocarb
Methiocarb
Mexacarbate
Average
Std. deviation
Ester-form PUF
Repl. 1 Repl. 2 Avg.
175.0 177.0 176.0
20.0 150.0 85.0
432.0 680.0 556.0
135.0 247.0 191.0
210.0 151.0 180.5
121.4
98.2
Ether-form PUF
Repl. 1 Repl. 2 Avg
32.5 125.0 78.8
80.0 45.0 62.5
350.0 412.0 381.0
73.0 175.0 124.0
233.0 262.0 247.5
121.9
69.1
a 5 cycles, with 1:1 hexane-acetone
        TABLE  16.  RECOVERIES  OF  PESTICIDE  SPIKES  (1.0 yG PER COMPOUND)
         FROM  ESTER- AND  ETHER-FORM PUF  PLUGS  BY SOXHLET  EXTRACTION3


                        	Pesticide  recoveries, %	
                             Ester-formEther-form PUF
   Compound                     PUF               Repl.  1  Repl.  2    Avg


 a-BHC                           123.0               176.6     160.6    168.6

 Aldrin                            96.3                91.5      77.8     84.7

 p,p'-DDE                        106.1                92.0      76.0     84.0

 o,p'-DDT                          95.6                93.3      78.6     86.0

 p,p'-DDD                          96.7                84.7      75.0     79.9

 TABLE 16 (continued)

                                      47

-------
                              TABLE 16 (continued)
Pesticide recoveries, %
Compound
p.p'-DDT
Lindane
3-BHC
Kept. epox.
Dieldrin
Endrin
Me parathion
Ronne 1
Parathion
Ethion
Carbophenothion
Diazinon
Malathion
Propoxur
Carbaryl
Carbofuran
Aminocarb
Methiocarb
Mexacarbate
Average
Std. deviation
Ester-form
PUFb
93.9
82.0
85.0
89.4
99.2
94.6
85.3
90.3
76.6
99.8
99.1
91.1
95.5
91.9
97.8
81.9
85.3
102.5
93.1
93.8
9.4
Ether-form
Repl. 1
94.3
67.8
99.8
76.4
101.1
83.5
77.8
78.0
80.7
97.1
86.2
80.3
90.9
97.5
101.0
97.0
88.0
100.0
100.0


Repl. 2
79.8
69.8
93.9
71.6
89.6
86.4
72.9
71.8
74.9
88.9
80.1
80.4
89.5
81.6
88.3
94.0
136.0
85.0
128.0


PUF
Avg
87.1
68.8
96.9
74.0
95.4
85.0
75.4
74.9
77.8
93.0
83.2
80.4
90.2
89.6
94.7
95.5
112.0
92.5
114.0
91.0
19.8
a 5 cycles, with 1:1 hexane-acetone
k Averages, from Table 10
                                     48

-------
Similarly, of the  24  compounds  in Table 16,  20 gave recoveries closer to 100%
with ester-  than with ether-form material,  and 14 of the 24 were closer in
Table 15.  If the  data for  aminocarb and mexacarbate,  which were shown in the
previous section to be interfered with by underlying background peaks, are
deleted, the average  deviations from 100% recovery at the 1.0-yg level are
8.6 +_ 1.5 and 17.3 _+  3.0%  for the ester and  ether forms, respectively.  We
therefore judge ester-form PUF  to be the superior material for an air
monitoring program.


EVALUATION OF GLC  COLUMNS  AND PROCEDURES

    One  of the most  important factors in the quantitation of pesticides by
gas-liquid chromatography  is the performance of the columns used.  The coating
of  the liquid on the  support material and the packing of the coated solid into
the columns  are critical  factors and are subject to considerable variability.
The quality  of a column is  manifest in the chromatograms obtained with it as
to  its ability to  separate the  compounds to  be analyzed, the sharpness of the
peaks, and the relation of peak height to quantity of compound.  On a poorly
prepared column, closely  eluting peaks will  be poorly separated and will
exhibit  tailing.   On a properly prepared column, the calibration curve, the
graph of peak height  vs.  quantity of injected material,  will be a straight
line that ends at  the origin.  When reactive compounds such as the
organophosphorus and  carbamate  pesticides are chromatographed on a poorly
prepared column, the  calibration graphs curve upward and may not pass through
the origin.  The usual cause for this phenomenon is retention and degradation
of  the compounds on  uncoated support material.

    We examined the  quality of  the columns that had been purchased or prepared
for use  in this study, particularly seeking  comparable or even improved per-
formance in  comparison to  that  reported by Sherma and Shafik in the original
publication  of the general analytical method (4).  These authors used a 5%
OV-210 column for  the organochlorine insecticides in liquid-chromatography
Fractions I  and II (except that a 5% SE.-30 column was used for dieldrin), a 5%
OV-210 Carbowax-treated column for the organophosphates in Fractions II and
III, and a 5% SE-30  column for  the carbamates in Fraction III.

    A 5% OV-210 column prepared in this laboratory was tested with the six
organochlorine pesticides  found in Fraction  I plus lindane, which is also
sometimes found in Fraction I.   The conditions were 180<>C column temperature
and 60 ml/min carrier gas  flow  rate, as specified by Sherma and Shafik.  In
the chromatogram,  the seven peaks were completely resolved except for slight
overlap between p,p'-DDD and p,p'-DDT.  The  chromatogram was very similar to
that of Figure 2 in  the Sherma-Shafik paper, and will  not be repeated here.
For comparison, a mixed column, 1.5% OV-17/1.95% OV-210, was also tested with
the same mixture of pesticides, but at 200OC column temperature and
36 ml/min carrier  gas flow  rate.   The chromatogram is  presented in Figure 11.
The retention times of the  compounds relative to aldrin are:  a-BHC, 0.48;
lindane, 0.63; aldrin,  1.00; p,p'-DDE, 2.29; o,p'-DDT, 3.24; p.p'-DDD, 3.52;
and p,p'-DDT, 4.28.   As may be  seen, the o.p'-DDT and p.p'-DDD peaks do
overlap  somewhat, but the peaks are adequately separated for quantitation of
all seven compounds,  and the overlap is no greater than the p,p'-DDD and


                                       49

-------
                                 1.   a-BHC
                                 2.   Lindane
                                 3.   Aldrin
                                 U.   p,p'-DDE
                                 5.   o,p'-DDT
                                 6.   p,p'-DDD
                                 7.   p,p'-DDT
                                   7
  0

Figure 11.
             8
          Mi nutes
12
16
ECGC chromatogram of 0.2  ng each of Fraction I
organochlorine insecticides, obtained with a
1.5* OV-17/1.95* OV-210 column.
                           50

-------
p.p'-DDT overlap on  the  5%  OV-210 column reported by Sherma and Shafik.   The
most significant difference in the chromatograms for the two columns is  the
shift of the p,p'-DDD  peak  away from the p,p'-DDT peak and toward the o,p'-DDT
peak.  Figure  11 indicates  that a mixed 1.5% OV-17/1.95% OV-210 column can be
a suitable alternative to a 5% OV-210 column for the analysis of
liquid-chromatography  Fraction I.

    A major advantage  of the mixed OV-17/OV-210 column over a 5% OV-210  is
that the former can  separate all 10 compounds found in Fraction II,  whereas
the latter does not  adequately separate dieldrin and methyl parathion, as
illustrated in Table III of Sherma and Shafik.   Figure 12 is an
electron-capture chromatogram for all five organochlorine and five
organophosphorus compounds  that occur in Fraction II.   The retention times of
the compounds  relative to that of aldrin are:  lindane, 0.63; 3-BHC, 0.72;
ronnel, 0.99;  methyl parathion, 1.29; heptachlor epoxide, 1.53;  parathion,
1.71; dieldrin, 2.37;  endrin, 2.91; ethion, 4.35; and carbophenothion, 4.68.
This chromatogram  suggests  that, at relatively high concentrations,  it may be
possible to quantitate the  organophosphorus compounds by electron-capture
detection at the same  time  that the organochlorine compounds are analyzed.

    The 5% OV-210  column treated with Carbowax 20M, recommended by Sherma and
Shafikj was tested for ability to separate the five organophosphorus
pesticides in  Fraction II and the two in Fraction III.  All pesticides within
a fraction were adequately  separated and the peak shapes were reasonably
symmetrical.   However, if all seven pesticides are applied to this column,
malathion and  methyl parathion overlap, as shown in Figure 13.   To obtain this
chromatogram,  all  seven pesticides at 5 ng per compound were injected
simultaneously, column temperature was 180°C, and flame photometric
detection was  used.  In a comparative test of a 3% OV-1 column with  the  seven
organophosphorus pesticides at 5 ng per compound, it was found that  the  peaks
were all separated.  A typical chromatogram with this column, showing
differences in order of elution as well as improved resolution,  is illustrated
in Figure 14.  Calibration  graphs were linear and the line passed through the
origin.  Figure 15 is  a calibration graph for diazinon and is typical of the
graphs for the organophosphorus compounds, although the others all had lower
slopes.  The 3% OV-1 column was used almost exclusively for the quantitation
of organophosphorus  compounds in the other phases of our experimental program
because tests  could  be run  with all seven compounds without the necessity for
prior separation by  liquid  chromatography.

    The SE-30  column prepared in this laboratory did not perform as  well as
that used by Sherma  and  Shafik for analysis of the N-methylcarbamates
(4, Fig. 7).   As an  illustration of our results, Figure 16 is a chromatogram
of the seven test  carbamates, at 600 pg per compound,  derivatized with PFPA
and using the  Sherma-Shafik GLC conditions, 195°C column temperature and
70 ml/min carrier  gas  flow  rate.  As may be seen, carbofuran and aminocarb
were not well  separated.  Separation was improved slightly by decreasing the
column temperature to  165°C, as illustrated in Figure 17.  By temperature
programming from 165°C to 195°C at 10°C/min. after a 5-minute hold at
165QC, the peak heights  for carbaryl and methiocarb were improved, as
illustrated in Figure  18.   However, the separation of carbofuran and aminocarb
was no better  under  these conditions.

                                       51

-------
                                         1.   Lindane
                                         2.   g-BHC
                                         3.   Ronnel
                                         k.   Methyl parathion
                                         5.   Heptachlor  epoxide
                                         6.   Parathion
                                         7.   Dieldrin
                                         8.   Endrin
                                         9.   Ethion
                                        10.   Carbophenothion
1
0

1
4

i
8

i t
12 16
Minutes
i
20

Figure 12.  ECGC chromatogram of O.U ng each of Fraction II
            insecticides, obtained with a 1.5$ OV-17/1.95$
            OV-210 column.
                            52

-------
                                               1. Diazinon
                                               2. Ronnel
                                               3. Malathion plus
                                                    Methyl parathion
                                               k. Parathion
                                               5. Carbophenothion
                                               6. Ethion
Figure 13.  Flame-photometric chromatogram of 5 ng each of the test
            organophosphorus insecticides, obtained with a 5% OV-210/
            Carbowax column
                                  53

-------
                                                 1.   Diazinon
                                                 2.   Methyl parathion
                                                 3.   Ronnel
                                                 k.   Malathion
                                                 5.   Parathion
                                                 6.   Ethion
                                                 7.   Carbophenothion
Figure lU.  Flame-photometric chromatogram of 5 ng each of the test
            organophosphorus insecticides, obtained with a 3% OV-1
            coliomn
                                  54

-------
    20
    16
 £
 u
    12
 O)

 
-------
                                         1.   Propoxur
                                         2.   2,3,5-Landrin
                                         3.   Carbofuran
                                         U.   Aminocarb
                                         5.   Mexacarbate
                                         6.   Carbaryl
                                         T.   Methiocarb
Figure l6.  Electron-capture chromatogram of N-methylcarbamate
            insecticides, obtained with a 5$ SE-30 column at
            195°C
                             56

-------
                                               1.  Propoxur
                                               2.  2,3,5-Landrin
                                               3.  Carbofuran
                                               k.  Aminocarb
                                               5.  Mexacarbate
                                               6.  Carbaryl
                                               T.  Methiocarb
Figure IT.  Electron-capture chromatogram of N-methylcarbamate
            insecticides, obtained with a 5% SE-30 column at
            165°C
                             57

-------
                                        1.  Propoxur
                                        2.  2,3,5-Landrin
                                        3.  Carbofuran
                                        H.-  Aminocarb
                                        5.  Mexacarbate
                                        6.  Carbaryl
                                        7.  Methiocarb
Figure 18.  Electron-capture chromatogram of N-methylcarbamate
            insecticides, obtained by temperature programming
            as described in text
                             58

-------
    In surveying the columns  available in this laboratory,  we found that a 3%
DC-200 column gave excellent  response  with the pentafluoropropionate
derivatives of the carbamates.   This  is  illustrated in Figure 19,  which
represents an injection  of  400  pg  of  each compound.  The peak height for
propoxur was about double  that  found  for PFP-propoxur at 600 pg with SE-30
column (Figure 16).  However, the  peaks  for 2,3,5-Landrin and carbofuran were
not separated.  In a later  experiment, 20 pg of each of the compounds in
liquid-chromatography Fraction  III except 2,3,5-Landrin (that is,  diazinon,
malathion and the remaining six carbamates) were treated with PFPA and applied
to the 3% DC-200 column.  The chromatogram is reproduced in Figure 20.  Both
of the organophosphorus  compounds  elute  after the PFP-carbamates and the
latter are well separated,  with excellent sensitivity with  respect to noise
level.  Calibration graphs  for  the six carbamates are presented in Figure 21.
All six graphs are linear  and pass through the origin, indicating  that
compound degradation on  the DC-200 column is not a problem.

    In the Sherma-Shafik procedure, diazinon and carbaryl elute together from
the SE-30 column, and at some levels  of diazinon this compound must be removed
by an acid wash before carbaryl can be quantitated.  By contrast,  diazinon
does not interfere on the  DC-200 column.  Since 2,3,5-Landrin is a minor
constituent of Landrin and, moreover,  production of the latter has been
discontinued by the manufacturer,  2,3,5-Landrin was omitted in the other
phases of our studies so that the  superior DC-200 column could be  used for the
quantitation of the carbamates.

    To summarize, we have  found GLC columns that exhibit advantages in
performance over  those recommended by Sherma and Shafik for quantitation of
each of the classes of test pesticides.   Our selections include a 1.5%
OV-17/1.95% OV-210 mixed column for organochlorine insecticide analyses, a 3%
OV-1 column for the organophosphorus  compounds, and a 3% DC-200 column for the
N-methylcarbamates.  The use of alternate columns is also beneficial in that
retention times on more  than one column can provide evidence with respect to
confirmation of identity of individual pesticides.

EFFECT OF FOAM  STRUCTURE ON AIR FLOW  RATES

     In the vapor  trapping  experiments described earlier, the PUF plugs
appeared  to compress  somewhat during  air sampling and the flow rate seemed to
diminish with time.   Six experiments  were performed to confirm this effect,  to
determine the relationship of flow rate* to sampling time, and to determine if
the  change in flow  rate  was directly  related to the porosity of the foam.
High-flow (low  pressure  drop),  45-mm  PUF plugs were tested in three of the
experiments, medium-flow plugs  in the other three.

    For each experiment, a single plug was placed in a glass sampling holder
and air was drawn through  the plug by vacuum pump.  The pressure across the
plug was measured with a diaphragm vacuum gauge, and the flow rate was
measured with a rotameter.   Readings  were taken periodically over a 24-hour
period.
                                       59

-------
                                         1.  Propoxur
                                         2.  2,3,5-Landrin and Carbofuran
                                         3.  Aminocarb
                                         U.  Mexacarbate
                                         5.  Carbaryl
                                         6.  Methiocarb
Figure 19-  Electron-capture chromatogram of UOO pg each of
            N-methylcarbamates, obtained on a 3% DC-200 column
                                  60

-------
                                     1. Propoxur
                                     2. Carbofuran
                                     3. Aminocarb
                                     k. Mexacarbate
                                     5. Carbaryl
                                     6. Methiocarb
                                     7. Diazinon
                                     8. Malathion
Figure 20.  Electron-capture chromatogram of 20 pg each of
            Fraction III pesticides, obtained on a 3% DC-200
            column
                                61

-------
 12
  8
       1.  Propoxur
       2.  Carbofuran
       3.  Methiocarb
       h.  Mexacarbat e
       5-  Aminocarb
       6.  Carbaryl
              0.5
1.0
1-5
2.0
2.5
Figure 21.   Calibration graphs of N-methylcarbamate insecticides, obtained on
            a 3% DC-200 column
                                      62

-------
    With all plugs, flow rates decreased with  time  and  the  pressure  across  the
plugs increased slightly with time.  The flow  rates decreased  relatively
rapidly at the start of the experiments and  asymptotically  approached  the
final flow rates.  The decrease  in  flow rate ranged from  3.2 to  9.0% of the
initial value, and no relationship  was apparent  between porosity of  the plug
and degree of decrease in the flow  rate.   The  change in flow rate with time
for a typical experiment is presented  in Table 17.   The plug in  this
experiment was rated as high-flow,  and the total decrease in flow rate was
8.6%,of the initial value.

    Quantitation of the pesticide residue  levels in air depends  on the
accuracy of the flow rate measurement  as well  as on the analytical
procedures.  These experiments indicate that,  for maximum accuracy in
quantitation of pesticides  in air with PUF foam, the flow rate should  be
monitored continuously during the sampling.


              TABLE 17.  CHANGE  IN  AIR FLOW  RATE WITH TIME  OF AIR
                            PASSAGE  THROUGH A PUF PLUG
Hours of
air flow
0
0.1
0.5
1.5
2.5
3.5
4.5
6.5
23.5
24.5
Air flow rate,
m3/hr
5.22
5.18
5.04
4.99
4.92
4.88
4.88
4.84
4.77
4.77
ANALYSIS  OF  N-METHYLCARBAMATE INSECTICIDES

Evaluation of Carbamate Derivatization Procedures

     Several  problems  exist in the GLC determinations of N-methylcarbamates.
These  compounds  generally have low vapor pressures, so that GLC retention


                                      63

-------
times are long; they are thermally unstable and decompose  in  the GLC  column
under common gas chromatographic conditions; and their  sensitivities  to
detection by electron-capture and other detectors  is not high.  The
Sherma-Shafik procedure (4) attempts to solve all  these problems by
derivatization of the N-methylcarbamates with pentafluoropropionic anhydride
(PFPA) according to the method of Shafik et al. (9).  In this procedure,  the
nitrogen-bound proton of the carbamate is replaced by a pentafluoropropionyl
(PFP) moiety.  The resulting PFP-carbamate compounds are more volatile and
more stable than the parent carbamates and also have five  fluorine atoms  to
enhance electron-capture detection.  Several other derivatization procedures
have also been used to facilitate quantitation of  carbamates.  For example,
the phenolic moieties liberated from the carbamates have been derivatized by
l-fluoro-2,4-dinitrobenzene (FDNB) (10).  This method has  been used on soil
residues in this laboratory with excellent results (11).   Parent N-methyl-
carbamates can also be derivatized with trifluoroacetic anhydride (TFAA)  and
heptafluorobutyric anhydride (HFBA) in reactions similar to that for  PFPA.  We
performed a number of experiments to determine which method was best  for
analysis of the seven test N-methylcarbamate insecticides, and to optimize  the
analytical conditions.

Effect of PFPA Concentrations on GLC Peak Heights—
    The procedure of Shafik et al. (9) specifies that the  carbamates  in 2 ml
isooctane should be reacted with 0.025 ml PFPA, with one drop of pyridine as
catalyst.  After one hour at room temperature, the reaction is stopped by
addition of 3 ml phosphate buffer.  Three ml  isooctane  and 0.05 ml
acetonitrile are added and the aqueous layer  is aspirated  with a pipet.   The
organic  layer  is washed three times with 2 ml water, with  the wash  removed
each  time by aspiration, and is then dried by addition  of  0.5 g sodium
sulfate.  An experiment was conducted using this procedure, partly  for
familiarization but also to examine the effect of  increasing  the amount  of
PFPA  reagent.  Pyridine was redistilled (12, p. 175), and  phosphate buffer  was
prepared as specified by Sherma and Shafik  (4).  A standard solution  of  the
seven  carbamates and the two organophosphorus insecticides that elute in
liquid-chromatography Fraction III was prepared in hexane  at  1 Ug/ml  per
compound.  One drop of keeper solution and  1.00 ml of this standard were
placed into each of seven 15-ml Kuderna-Danish receiver tubes.  The contents
of  each  tube were evaporated at 50oc in a water bath by a  stream of nitrogen
gas.  The Shafik et al. procedure was -then  followed, except that 0.025,  0.05,
0-10, 0.15, 0.20, 0.25 or 0.30 ml PFPA was  added to a tube.   The derivatized
carbamates were analyzed by electron-capture GLC on a 5% SE-30 column (the
superiority of the 3% DC-200 column had not yet been discovered).  Figure 22
shows  the chromatogram, representative of the seven chromatograms,  for  the
standard treated with 0.10 ml PFPA.  The relatively poor resolution of  peaks,
previously shown for the 5% SE-30 column (Figure 16), was  apparent.   However,
the organophosphorus components of the standard solution did  not interfere
with  the carbamate peaks.

    The height of each peak in each of the  seven chromatograms was measured
and is recorded in Table 18.  Changes in the amount of  PFPA used in the
reaction did not affect the peak heights of the individual carbamates
equally.  Peak heights increased for all the compounds  as  the quantity  of PFPA


                                      64

-------
                                                 1.  Propoxur
                                                 2.  2,3,5-Landrin
                                                 3.  Carbofuran
                                                 h.  Aminocarb
                                                 5.  Mexacarbate
                                                 .6.  Carbaryl
                                                 7.  Methiocarb
0
8          12
   M i n.
16
20
 Figure 22.  Electron-capture chromatoeram of kOO pg each, of the
             N-methylcarbamates,  reacted with 0.1 ml PFPA
                               65

-------
             TABLE 18.  THE EFFECT OF CONCENTRATION OF PFPA REAGENT
                 ON GLC PEAK HEIGHTS OF CARBAMATE INSECTICIDES
                       	Peak heights, mm	
                                          PFPA volume added, ml
   Compound             0.025    0.05    0.10    0.15    0.20    0.25      0.30
Propoxur                  46      52      80      91      88      95        92

2,3,5-Landrin             46      50      74      77      78      80        74

Carbofuran                30      33      46      49      52      54        50

Aminocarb                 48      36      54      36      36      31

Mexacarbate               33      38      43      35      27      27        26

Carbaryl                   3       3       8       8      18      20        20

Methiocarb                 6       9      15      19      22      23        19
was increased up to about 0.10 ml, but  the  increase was much more  modest  for
aminocarb and mexacarbate than for the  other compounds.  Moreover,  the
aminocarb and mexacarbate peaks decreased in size  as  the amount  of PFPA
exceeded 0.1 ml, whereas the other peaks leveled off  or continued  to
increase.  At the highest levels of PFPA, the  carbofuran peak  tended to
obscure the aminocarb peak.  The data indicated that  an increase of the PFPA
from 0.025 ml up to 0.10 ml in the Shafik et al. procedure  would be beneficial.


Evaluation  of  a TFAA-derivative  Procedure—
     Seiber  (13) demonstrated  that  temperature  and  solvent  polarity strongly
affect rates of N-methylcarbamate  derivatization by peffluoroacyl anhydrides.
When the  reaction  is  performed  in  polar solvents,  shorter  reaction times
and/or lower  temperatures may be used  to effect  complete  derivatization.
Greenhalgh  et  al.  (14)  reported  that methiocarb  could be  trifluoroacetylated
in 15  minutes  at 100°c  in ethyl  acetate. We therefore performed an
experiment  to  compare results  from the  Greenhalgh  et al.  procedure with those
from the  Shafik et  al.  meth6d.

    A  1-ml  aliquot  of each  1-ppm carbamate  standard solution was placed into a
screw-cap vial, one drop of keeper solution was  added, and the solvent was
evaporated  at  50°C  in a stream of  nitrogen. Ethyl acetate (2 ml)  and
0.20 ml TFAA were  added and the vial was capped.   After two hours, the vial
was opened, excess  reagent  and solvent  were evaporated in  a stream of
                                      66

-------
nitrogen, and 1 ml ethyl  acetate  was  added.   This  solution was  analyzed by GLC
as in the Sherma-Shafik procedure.

    The injection peak was very large and tailed badly.   Moreover,  the
derivative peaks were relatively  small,  had  short  retention times,  and  were
not well separated.  TFAA, and perhaps ethyl-TFA (if formed from ethyl  acetate
by transesterification),  may  not  be  totally  removed in the evaporation  step;
if not, they would produce a  large  injection peak  such as that  observed.
Greenhalgh et al. used flame-photometric detection in the sulfur mode for
analysis of methiocarb.   In that  system, neither TFAA nor ethyl-TFA would
interfere.  The Greenhalgh et al. procedure  did not give  acceptable results
with electron-capture detection and was  consequently abandoned.

Comparison of PFPA With HFBA  for  N-Methylcarbamate Derivatization—
    Seiber (13) reported  that TFA, PFP and HFB derivatives of several
N-methylcarbamates had similar stabilities and rates of formation.   However,
the electron-capture response increased  two-fold for carbaryl and five-fold
for carbofuran from the TFA to HFB  derivatives.  We conducted an experiment to
determine if the HFB derivatives  would offer higher electron-capture response
and/or better separation  of GLC peaks than the PFP derivatives.

    One ml of the Fraction III standard  solution and one  drop of keeper
solution were placed in a 4-dram  screw-cap vial  and evaporated  in a stream of
nitrogen at 50°C.  Benzene (4 ml) was added  and the vial  was sonicated  for 2
minutes to dissolve the pesticides.   After addition of 0.10 ml  HFBA, the vial
was sealed with a teflon-lined cap  and heated at 100°C for 2 hr.   The
solution was diluted with 4 ml hexane and washed three times with 5-ml
portions of water.  It was then dried with sodium  sulfate and concentrated to
5.0 ml with a stream of nitrogen. This  solution was analyzed by GLC on a  3%
SE-30 column, resulting in the chromatogram  shown  in Figure 23.   In the
figure, each peak represents  approximately 200 pg  of the  designated
carbamate.  Although the  retention  times here were longer and the residue
levels lower by half, a comparison  of Figure 23 with Figure 22  shows that  the'
response with this HFBA procedure was no better than that with  the Shafik  PFPA
procedure.  There was also no advantage  in terms of effort or analysis  time.

    The  same Fraction  III standard  was also  derivatized with 0.10 ml HFBA by
 the Shafik et'al. procedure  detailed above.   The derivatized solution  was
 analyzed  on  the  3%  DC-200 column at 165oc column temperature.  At  the  200-pg
 residue  level,  six  sharp  and  well-separated  peaks  were obtained for the seven
 carbamates.  Derivatization  of each carbamate individually with HFBA showed
 that  2,3,5-Landrin  and  carbofuran had such close retention times that  they
would appear as  one  peak  when both compounds were  present.  The six-peak
 chromatogram of  the  HFB-carbamates  was therefore essentially identical to  that
obtained  with  the PFP-carbamates  on the same column (Figure 19).

    A standard  solution  containing only the six test  carbamates, without
2,3,5-Landrin, was  prepared  at  1  ug/ml per compound.  One ml of  this solution
was derivatized  with 0.10 ml  HFBA by the Shafik et al. procedure and analyzed
on a  3% DC-200 column.  The  chromatogram is presented in Figure  24.  A
comparison of  this  figure with  the corresponding chromatogram  for  the
                                       67

-------
                                                 1.   Propoxur
                                                 2.   2,3,5-Landrin
                                                 3.   Carbofuran
                                                 k.   Amiriocarb
                                                 5.   Mexacarbate
                                                 6.   Carbaryl
                                                 7.   Methiocarb
Figure 23.  Electron-capture chromatogram of 200 pg each of the
            W-methylcarbamates, reacted with 0.1 ml HFBA
                                   68

-------
                                               1.   Propoxur
                                               2.   Carbofuran
                                               3.   Aminocarb
                                               k.   Mexacarb ate
                                               5.   Carbaryl
                                               6.   Methiocarb
Figiire 2k.  Electron-capture chromatogram of  six HFB-carbamate
            derivatives, obtained on  a  3% DC-200 column
                                69

-------
PFP-carbamates (not shown) demonstrates that there  is  little  difference  in
retention times or peak heights between HFB-carbamates and PFP-carbamates.
Our experiments have shown that HFBA can be regarded as a substitute  for,  but
apparently not an improvement on, PFPA in the Shafik et al. procedure.

Evaluation of l-Fluoro-2,4-dinitrobenzene for Derivatization  of  Carbamates—
    Holden (10) reported a procedure for the GLC determination of  several  of
the carbamates used in this study.  Reaction of l-fluoro-2,4-dinitrobenzene
(FDNB) with carbamates yielded ether derivatives of the phenolic moieties
liberated from the carbamates.  These derivatives were well-separated and
readily quantitated on a DC-200 GLC column.  We conducted an  experiment  to
determine if the Holden procedure could be applied  to  extracts from
polyurethane foam, and if it offered any advantage  over the PFPA
derivatization.

    A blank was prepared by Soxhlet-extracting a PUF plug with 280 ml 1:1
hexane-acetone for five cycles.  The solution was evaporated  and 5 ml 5%
aqueous borax  solution and 0.5 ml FDNB reagent (1.5 ml FDNB in 25  ml  acetone)
were added.  The contents were agitated in an ultrasonic bath, heated at
70°C for one hour in a water bath, and then extracted  with 5  ml  hexane.  The
extract was analyzed by GLC on a 3% DC-200 column,  resulting  in  the
chromatogram reproduced in Figure 25.  If a carbamate  were present in the
extract at the 1-ug level, this chromatogram would  represent  an  injection  of
40 pg of the carbamate.  This high background level was shown, by
derivatization of the carbamates by the same procedure and addition to the
blank, to be prohibitive for quantitation of the carbamates.  Consequently,
the method was abandoned.

Effect of Diaminotoluenes on Quantitation of N-Methy1carbamates

    When PUF degrades, two of the products are 2,4- and 2,6-diaminotoluene
(15).  These compounds would be detected by an electron-capture  detector if
they were present in the final extracts of the Sherma-Shafik  procedure and
thus might interfere with GLC analysis of organochlorine or carbamate
pesticides.  In the case of the carbamates, any diaminotoluenes  occurring  in
liquid-chromatogrpahy Fraction III would have to be derivatized  by PFPA  to
interfere with pesticide quantitation.  Each of the diaminotoluenes has  four
reactive amino protons.  If all four possible substitutions were accomplished
by PFPA, the diaminotoluenes would be detectable at very low  concentrations  by
electron-capture detection.  We conducted tests to  determine  the likelihood
and possible extent of the interference.

    For the following tests, 2,4- and 2,6-diaminotoluene were purchased  from a
commercial source (Aldrich Chemical Co.) and recrystallized three  times  from
absolute ethanol to remove impurities.  Individual  1000-ppm standard solutions
of both compounds were prepared by weighing 50 mg of the recrystallized
material into  a 50-ml volumetric flask and adding acetone.  These  standards
were serially  diluted to the 1 ug/ml level in hexane.  One ml of each of the
diluted standards was evaporated, with five drops of paraffin oil  keeper
solution, and  derivatized with PFPA.  GLC analysis  of  the derivatives on a 3%
DC-200 column  showed that the two peaks were slightly  separated  and that the
                                      70

-------
                          \
Figure 25.  Electron-capture chromatogram of a PUF-plug
            blank extract derivatized with FDNB
                            71

-------
retention times were very close to that of PFP-propoxur.  To  check  this,  a
solution containing 1 yg of propoxur was added to  1-ml aliquots  of  the  1-ppm
diaminotoluene standards.  These solutions were evaporated, derivatized and
analyzed on a 3% DC-200 column.  The chromatograms are presented in Figure
26.  The propoxur peak was slightly removed from that of 2,6-diaminotoluene,
but was not separated from the 2,4-diaminotoluene  peak.  Electron-capture
response of the PFP-diaminotoluenes was very high.  Each peak in Figure 26
represents an injection of 40 pg of material; the  peak heights of the
diaminotoluenes were almost four times that of propoxur.  The diaminotoluenes
have been reported to occur in concentrations as high as 400  ppm in PUF,  and
this test shows that they could interfere with quantitation of propoxur on 3%
DC-200 if they are present in Fraction III.

    A further test was conducted to determine if the PFP-diaminotoluenes  would
interfere with quantitation of carbamates on an SE-30 column.  One  ml of  a
standard solution containing 1 yg of each of the pesticides that occur  in
Fraction III was added to 1 ml each of the 1-ppm diaminotoluene  standards.
This was evaporated with 5 drops of keeper solution, derivatized with PFPA,
and quantitated by GLC on a 5% SE-30 column.  The  chromatogram is presented  in
Figure 27.  The two PFP-diaminotoluenes appear as  one peak that  is  adequately
separated from PFP-propoxur.

    After these tests, it remained to be determined whether the
diaminotoluenes would appear in any of the three liquid-chromatography  eluate
fractions.  One-mi aliquots of the 1-ppm diaminotoluene solutions,  evaporated
and resuspended in 0.5 ml hexane, were chromatographed on a silica-gel  column
as required by the Sherma-Shafik procedure.  Each  fraction was analyzed by
ECGC, Fraction III being derivatized with PFPA before analysis.   There  were no
unusual peaks in the chromatograms, so the diaminotoluenes apparently do  not
elute from the liquid-chromatography column.
                                      72

-------
                                                       1.  2,6-Diaminotoluene
                                                       2.  2,^-Diaminotoluene
                                                       3.  Propoxur
Figure 26.  Electron-capture chromatograms of diaminotoluenes and
            propoxur, obtained on a 3% DC-200 column
                                    73

-------
                                                   2.
                                                   3.
                                                   U.
                                                   5.
                                                   6.
                                                   7.
                                                   8.
2,U-Diaminotoluene
plus 2,6-Diaminotoluene
Propoxur
2,3,5-Landrin
Carbofuran
Aminocarb
Mexacarbate
Carbaryl
Methiocarb
Figure 27-  Electron-capture chromatogram of Fraction III pesticides plus
            diaminotoluenes, obtained on a 5$ SE-30 column
                                   74

-------
                                    REFERENCES

 1.  Fowler, D. L., and J. N. Mahan.  The Pesticide Review,  1977.   Agricultural
     Stabilization and Conservation Service, U. S. Department of Agriculture,
     Washington, D. C., 1978.  44 pp.

 2.  Turner, B. C., and D. E. Glotfelty.  Field Air Sampling of Pesticide
     Vapors With Polyurethane Foam.  Anal. Chem., 49 (1):7-10,  1977.

 3.  Lewis, R. G., A. R. Brown and M. D. Jackson.  Evaluation of Polyurethane
     Foam for Sampling of Pesticides, Polychlorinated Biphenyls and
     Polychlorinated Naphthalenes in Ambient Air.  Anal.  Chem.  49
     (12):1668-1672.

 4.  ,Sherma, J., and T. M. Shafik.  A Multiclass, Multiresidue  Analytical
     Method for Determining Pesticide Residues in Air. Arch. Environ.  Contain.
     Toxicol. 3(1):55-71, 1975.

 5.  Ives, N. F., and L. Giuffrida.  Gas-Liquid Chromatographic Column
     Preparation for Adsorptive Compounds.  J. Assoc. Offie. Anal.  Chemists
     53(5):973-977, 1970.

 6.  Rhoades, J. W., and D. E. Johnson.  Evaluation of Collection Media for  Low
     Levels of Airborne Pesticides.  National Technical Information Service
     PB-275,668, 140 p, 1977.

 7.  Mercer, T. T.  Aerosol Technology in Hazard Evaluation.  Academic  Press,
     New York, 1973.

 8.  Spencer, W. F., J. D. Adams, R. E. Hess, T. D. Shoup and R. C. Spear.
     Conversion of parathion to paraoxon on soil dusts and clay minerals as
     affected by ozone and UV light.  Submitted to J. Agric. Food Chem., 1979.

 9-  Shafik, T. M., D. Bradway and P. F. Mongan.  Electron Capture  Gas
     Chromatography of Picogram Levels of Aromatic N-methyl  Carbamate
     Insecticides.  Presented at 163rd ACS National Meeting, Boston,
     Massachusetts, April, 1972.

10.  Holden, E. R.  Gas Chromatographic Determination of  Residues  of
     Methylcarbamate Insecticides in Crops as Their 2,4-Dinitrophenyl Ether
     Derivatives.  J. Assoc. Offic. Anal. Chemists 56(3):713-717,  1973.

11.  Caro, J. H., D. E. Glotfelty, H. P. Freeman and A. W. Taylor.   Acid
     Ammonium Acetate Extraction and Electron Capture Gas Chromatographic
     Determination of Carbofuran in Soils.  J. Assoc. Offic. Anal.  Chemists
     56(6):1319-1323, 1973.

                                       75

-------
12.   Vogel, A.  I.  A Text-book of Practical Organic Chemistry.  John Wiley and
     Sons, New York, 1962.

13.   Seiber, J. N.  N-Perfluoroacyl Derivatives for Methylcarbamate Analysis
     by Gas Chromatography.  J. Agric. Food Chem.  20(2):443-446, 1972.

14.   Greenhalgh, R., W. D. Marshall and R. R. King.  Trifluoroacetylation of
     Mesurol (4-Methylthio-3,5-xylyl-N-methylcarbamate),  its Sulfoxide,
     Sulfone, and Phenol Analogs for Analysis by Gas Chromatography.   J.
     Agric. Food Chem. 24(2):266-270, 1976.

15.   Guthrie, J. L., and R. W. McKinney.  Determination of 2,4- and
     2,6-Diaminotoluene in Flexible Urethane Foams.  Anal. Chem.
     49(12):1676-1680, 1977.
                                     76

-------
                                  TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
 REPORT NO.
 EPA-600/4-80-008
2.
                              3. RECIPIENT'S ACCESSIONING.
.TITLE AND SUBTITLE
Polyurethane Foam  as Trapping Agent for Airborne
Pesticides:  Analytical Method Development
                              5. REPORT DATE
                               January  1980  issuing date
                              6. PERFORMING ORGANIZATION CODE
 AUTHOR(S)

 James D. Adams and Joseph H.  Caro
                              8. PERFORMING ORGANIZATION REPORT NO.
 PERFORMING ORGANIZATION NAME AND ADDRESS
 Soil Nitrogen and Environmental  Chemistry Laboratory
 Beltsville Agricultural Research Center
 U.S. Department of Agriculture
 Beltsville, Maryland  20705
                              10. PROGRAM ELEMENT NO.
                                A1VL1A
                              11. CONTRACT/GRANT NO.
                                Interagency Agreement
                                78-D-X0449
12. SPONSORING AGENCY NAME AND ADDRESS
  Environmental Research Laboratory—Athens GA
  Office  of Research and Development
  U.S.  Environmental Protection Agency
  Athens,  Georgia  30605
                              13. TYPE OF REPORT AND PERIOD COVERED
                                Final,  9/76-1/79
                              14. SPONSORING AGENCY CODE
                                EPA/600/01
15. SUPPLEMENTARY NOTES
16. ABSTRACT
       A  method for determining levels  of organochlorine, or'ganophospjiorus, and N-methy
  carbamate  insecticides in air was  developed  using 4.4 cm-dimeter  plugs of polyure-
  thane foam as  traps and a modified Sherma-Shafik multiresidue procedure for analysis
  of foam  extracts.   With this method, the minimum detectable air  concentrations for
  vapors of  17 of the 18 organochlorine  and organophosphorus pesticides tested was 0.1
  nanogram per cubic meter or less.   Six carbamate pesticides did  not  volatilize in
  sufficient amounts to allow analysis.

       The  experimental program consisted of measurements of efficiency of foams for
  trapping vapors and aerosols, tests of foam compression during air  sampling, evalua-
  tions of extraction solvents, evaluations of techniques for extracting pesticide resi-
  dues from  the  foams, comparisons of the performance of the ester and ether forms of
  polyurethane foam, evaluations of  the  steps in the Sherma-Shafik analytical procedure,
  evaluations of GLC columns and instrumental parameters, and tests of derivatization
  procedures for electron-capture gas chromatography of the N-methylcarbamates,

       A  10-cm  depth of foam was found  to be an efficient trap for vaporized pesticides
  Five-cycle Soxhlet extraction with 1:1 hexane-acetone was the best  procedure for re-
  moval of pesticides from foam plugs.	
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
a.
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                                               COSATI Field/Group
  Air Pollution
  Chemical analysis
  Gas chromatography
  Monitors
  Pesticides
                                               07C
                                               14B
                                               68A
                                               99A
 8. DISTRIBUTION STATEMENT

  RELEASE  TO PUBLIC
                 19. SECURITY CLASS (This Report}
21. NO. OF PAGES

      83
                 20. SECURITY CLASS (This page)
                   UNCLASSIFIED
                                            22. PRICE
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE 77
                                       4 U.S GOVERNMENT PRINTING OFFICE 1960-657-146/5568

-------