Volume IV
            Chemical Methods Workshop

                        OFFICE OF PESTICIDE PROGRAMS AND
                                WASHINGTON, D.C.
  JULY 30 - AUGUST 1, 1975

                 Volume IV

           Chemical Methods Workshop


                          FREDERICKSBURC, VIRGINIA
N^JULY 30 - AUGUST 1, 1975

FIRST YEAR OF PROGRESS are published Ln four volumes. Volume I
contains speeches and discussion from the Plenary Session, the agenda,
and lists of participants and speakers. Volumes II, III, and IV cover the
Toxicological Methods and Genetic Effects Workshop, the Ecosystems/
Modeling Workshop, and the Chemical Methods Workshop,  respectively.

                       TABLE OF CONTENTS
 Wednesday. July 30, 1975
     Dr. H. Alison Moye	 1

     Dr. J. S. Warnefc	31
     Dr. Donald E. Johnson	67
 Thursday. July 31.  1975
     Dr. Barry Commoner	39
     Dr. Edward 0.  Oswald	117

     Kenneth R. Hill	127
     Dr. Robert Moseman	141
     Dr. Robert Lewis	147
     Dr. William  T. Colwell	159
     Dr. R. Fukuto	165

     Dr. Edward O.  Oswald	173

     Jack B. Dixon	205


                      MULTIRESTDUE METHODOLOGY
                            Dr. H, AnsonMoye*

  I feel free now to at least-present my list of what were last year the alterna-
tive chemical^ or substitute chemicals.  Our contract really was composed of
two parts,  the first of which involved gas chromatography methodology, the
second of which involved liquid chromatography or high-speed liquid chromatogra-
phy, if you will, methodology.  The idea in the gas chromatography portion was to
flesh out the available data on extraction efficiencies  and gas chromatographic
characteristics of these compounds.
  We knew from the literature and past experience that some of them had no hope
of being gas chromatographed.  Some could be gas chromatographed with chemical
derivatlzation.   So we included a short study on two chemical derivatives for those
compounds that seemed appropriate.
  We were to take these compounds — I will list them for you shortly — and try
gas chromatography.  We had at our disposal a Microtech or Tracor, If you will,
MT 220 equipped with dual electron capture detectors and the microelectrolytic
conductivity detector that Mr. Dixon just described, a Varian 1200 with the Tracor
flame photometric detector.  So we had the capability of detecting essentially every
element responsive  to any element-specific detector of which you might conceive.
  We were to attempt direct gas chromatography on these compounds.  For those
that could not be gas chromatographed we were to attempt to derivatize with two  '
reagents which I will shortly describe (Figure 1). If we were successful in obtain-
ing gas chromatographic responses,  either of the intact pesticide or the derivatlzed
pesticide, we were to apply a silica gel column cleanup which had been devised by
*Food Science Department, University of Florida, Gainesville, Florida

 Drs. Shafik and Sherma, EPA, which had in previous work appeared to be some-
 what superior to the Florisil cleanup that had been so long favored by the Food
 and Drug Administration and subsequently by EPA.
                               Figure 1
                     SILICA GEL
                    i COLUMN
                     FROM  SOIL  '
                     0.01 -10ppm
  If we could get thes.e materials through the silica gel column,  we were then to go
back, apply them, to soils at levels of 0, 0.01, 0.1,  1.0 and 10 parts per million;
perform extractions, do the column chromatography and then the derlvatization
when necessary, and then the gas chromatography.
  In Figure 2 we broke our list of 33 compounds into three classes, mainly for
bookkeeping mechanism whereby we could apply the element-specific detectors.
We had a number of chlorine-containing compounds that were somewhat ambiguous
such as trifluralln, which is not  a chlorine-containing compound, but we didn't
know where else to put it.
  The phosphorus-containing compounds we examined were azinphos-ethyl, azinphos-
methyl, Azodrin, demeton, malathion,  methyl parathion, parathion, and phorate.
The nitrogen-containing compounds were amitrole, atrazlne, Avadex,  benomyl,
bromacil, Bux, CIPC, diquat, dinitro-orthocresol (DNOC),  ferbam, Furadan, IPC,
monuron, simazine, Temik or aldlcarb, and Zectran.  As you see from the list,

                                                       Figure 2
                                                     33 Pesticides
    Cap tan
Phosphorus -Containing
    Azinophos Ethyl
    Azinophos Methyl
    Methyl Parathion
       *13 pesticides which could not be satisfactorily gas chromatographed intact.

those starred were the ones that we could not gas chromatograph Intact.  AU the
rest of them we could.

                             Figure 3
                              13 PESTICIDES
                             'low conversion;
                             small  peak
  On these 13 compounds we attempted two derivatization procedures.  One is
a Schotten-Baumann type reaction of 2,5-dichlorobenzenesulfonyl chloride which
will react,  we have shown, with essentially any N-methyl carbamate, although
there are exceptions.  The other reaction was a pentafluoropropionic anhydride
derivatization that was also developed by Dr. Shafik, which had been shown to be
applicable to any hydrogen-replaceable compound, such as N-methyl carbamate.
  Of these  13 pesticides that were not gas chromatographable (Figure 3), there
were essentially six that we could not gas chromatograph under any circumstances,
with the exception of Temik or aldicarb. Temik did give a response with the
2,5-diachlorobenzene sulfonyl chloride; however,  it formed In the methylamlne
mode, making it nonspecific.  In other words, you could not separate it from
materials like the methomyl.

  We worked with four gas chromatographlc columns, two of which were really
holdovers or old-fashioned FDA columns, one a 10 percent DC-200 and one a
10 percent QF-1 (Figure 4).  As you see, the 10 percent DC-200 appears to be
far superior in its separatory power for 8 of the chlorinated compounds.
  There are some ambiguities there.  Endosulfan and perthane cannot be separated;
that is, the high boiling isomer of endosulfan cannot be separated from perthane.
QF-1 looks almost hopeless; everything kind of overlaps.
  There were two mixed-phase columns which were examined (Figure 5).
These were mixed-phase column packings that had been worked out by EPA and
had been used in their community studies program, and,  as  you see, they do not
do a  much better job; however, the 4 percent SE-30, 6 percent OV-210 would be
passable, I suppose. It had less  retention than the DC-200 for everything essential-
ly, but the efficiency was high enough that some reasonable  separations could be
obtained.  The 1.5 percent OV-17, 1.95 percent OV-210 was not very good, to say
the least.
           Once we established the gas chromatographability of these materials,
we tested them on the silica gel absorption column.  Let me back up just a minute
here.  The four columns we tested had the theoretical plates calculated for a 6-foot
long by 3 millimeter I.D. glass column.  The DC-200, calculated on aldrin at
200° C, was only 1,700 theoretical plates (Table 1).  So we  really were not all that
good in our column  packing, but  it was significantly superior to any of the three we
packed in terms of efficiency.
   The compounds that we could get gas chromatographed were applied to the
activity number three silica gel  column. This was a column 2.5 cm on the Inside
diameter, packed with 16 grams of Woelm activity three silica gel.  This Is a
silica gel that Is very closely controlled in  Its activity.  It is supplied as activity
three as a 15 percent moisture packing.  It performed, I think,  a remarkable

Figure 4.  Chromatograms of an 8-compound chlorinated pesticide
          mixture.  Column operating parameters are shown in Table 2.
                            10* DC-200
                               Trifluralin 2  nti
                               Chlorbenzilate  10 ng
                               Methoxychlor  2  ng

Figure 5.  Chromatograms of an 8-compound pesticide mixture.
          Column operating parameters are shown in Table 3.



2 20
— 10

^ 60









4% SH-30/6'i OV-210





6 A A ^.^.^^^

B 1.5% OV-17/1.95", OV-21u

1. Trifluralin 1 nu
2. PCNR 1 tig
3. Captan 1 ng
4. Folpet 1 ng
15 5 .. Endosulf an 1 n^

6. Pcrtliane 1 lij*
7. Chlorbenzi late 5 nx
8. Methoxychlor 1 ng

.1! 5,6,4
| 1
pl^L~-**-/^"' — **=»-i'::wi*
*"""' '.y M 	 TO 	 ZD 	 30" "inr.c (." ini

                   Table 1:  Computed Efficiencies of Columns
                    (tritium foil EC detector, Packard GLC).
        Column                                               Efficiency(l)
        	                                            (Theor.  Plates)
        10% DC-200                                             1700 (2)
        10% QF-1                                                676
       4% SE-30/6% OV-210                         .            1156
       1.5% OV-17/1.95% OV-210                                256

      (l)Efficiency computed from total retention and base width of aldrin peak.
      (2)Value represents superior efficiency.

 fractionation of these materials.  Table 2 shows the nitrogen-containing compounds
 and the halogen-containing compounds that we could gas chromatograph. If you
 look closely, you will see that there is essentially no holdover from one fraction
 to the  other. This is quite remarkable, and I don't believe, to my knowledge,
 any other adsorption cleanup column can say this.
   We found very,  very little holdover. There are five fractions.  The first fraction
 is eluted with 50 milliliters of hexane. This gives pentachloronitrobenzene.  The
second fraction is eluted with 60 percent benzene hexane.  This gives the four that
you see. The third fraction is 5 percent acetonitrile-benzene.   This gives the seven
that you see.  The fourth fraction, 10 percent acetonitrile-benzene, and here we get
the bulk of the carbamate or nitrogen-containing materials.  Fraction five at 20 per-
cent acetonitrile-benzene gets bromacil and monuron.
   For the organophosphates listed in Table 3 once again a fractionation is observed.
For the hexane fraction there is no organophosphate eluted.  The 60 percent benzene
hexane gives phorate, methyl parathion, parathion, and Azodrin; 5 percent
acetonitrile-benzene, malathion, azinphos-methyl, azinphos-ethyl; and 10 percent

      Table 2: Outline of Halogenated and Nitrogen-Containing Pesticide
               Standard (1 mg each)  Eluted from Silica Gel Column
                        16 g silica gel,  14 g
Fraction I
Fraction n
60% Benzene/
Trifluralln (105%)
Fraction IE
5% Acetonltrlle/
Captan (102%)
Fraction IV
10% Acetonltrlle/
Atrazlne (100%)
Fraction V
20% Acetonltrlle/
Bromacll (98%)
       Perthane (97%)     Folpet (98.1
       Methoxychlor(92%) Chlorobenzilate
       Endosulfan (90%)    (91%)
                         CIPC (90%)
                         Dlfolatan (100%)
                         DNOC (90%)
                         Bux, IPC, Zectran
                                            Simazlne (100%)
                                            Furadan (90%)
                        Monuron (75%)
Note: Percent recoveries are presented after each pesticide.
         Table 3:  Outline of Organophosphate Pesticides Recovered from
                  Silt Loam Soil and Eluted from Silica Gel Column
                        16 g silica gel,  14 g

Fraction I


No pesticide
                   Fraction n
               60% Benzene/Hexane
               methyl parathion
Fraction in      Fraction IV
5% Acetonltrlle/   10% Acetonitrile/
 Benzene          Benzene
malathion         demeton

 acetonitrile-benzene, demeton.  There is no fraction five; fraction five gives nothing.
 All of our organophosphates come off in the first four fractions.

   The cleanup capability of this column, I think, is illustrated by Figure 6.  This
 was a crude soil extract, and I might back iip and say that we worked with three
 types of soil, and we performed all of our extractions at the various levels with
 all three types.  We had a sandy soil, a sandy loam soil, and a silt loam soil.
 This is the sandy soil,  which was taken from a cow pasture and had the highest
 organic  content.  It also gave the lowest percentage recovery for essentially all
 the compounds that we  were able to recover satisfactorily, although not that much
   You can see the crude extract on  a DC-200 column.  It is quite bumpy.  There is
 no room to analyze anything at subpart per million levels with this crude extract.
 By putting it through the silica gel column, the hexane fraction shows a large pro-
 portion of the interferences.  Pentachloronitrobenzene elutes in this fraction and
 then on the gas chromatogram, and the relative retention you see there.

  In the  60 percent benzene hexane fraction (Figure 7) the materials elute as you
see — trifluralin, endosulfan,  perthane, and methoxychlor.  There is a little bit
of clutter in the initial part of the  chromatogram; however, trifluralin and endosul-
fan could be analyzed at the 0.01  part per million level. As in the next fraction,
5 percent acetonitrile-benzene, the  elutions of CIPC, folpet, captan, chlorobenzilate,
and difolatan are all at fairly clear places  for the 0.01 ppm level.
  The E  fraction is still fairly clean in Figure 8. There is one major interference
peak which elutes at about 5 minutes.  Atrazine and simazine fall before that, just
after one minor interference peak.  The 20 percent acetonitrile fraction gives
bromacil, which elutes  in 5 minutes in almost a perfectly clean chromatogram.
  Let me point out that the reaction involved in this is a pentafluoropropriontc
anhydride derivatization. It is simply an acylation which takes place at room

Figure 6. Gas chromatograms of 50-mg equivalent crude soil extract and
          five silica gel column fractions from the total 100 g control
          methanol-benzene (1:1)  Soxhlet extract of silt loam soil (retention
          times of pesticides were exhibited in appropriate chromatograms).
    , ;
                                              TRACOR 22:, lie.
                                              6'xi/s"i .:•.
                                              F low ra tc :  SI'  ;.
                                            CRUUC  SOIL  EX.'RACT

                    Figure 7
                                          CO*. Bcnzor.o/r.cxanc-

                                          fract ion
                        PERT! IAN H
                               D   5^ Acetroni tri le  i r. henzer.o

                                Ml NUTUS

Figure 8
*— \

101 Acetroni t r i Ic ;.n benzene
_ /""""" _
20°. Acetroni tr: le ;n bcr.^cnc ,
f tact ion
"0 i> iU 15— -• 7TT 	 '

temperature or for one hour in the presence of hexane and a small amount of
pyridlne (Figure 9).

                          Figure 9
                                                           1 hour
                           R-O-C-N'  o
  The reaction of 2,5-dichlorobenzenesulfonyl chloride is also an alkaline media
base reaction (Figure 10). Here I don't show the conditions; however, it can be
run in acetone or with a carbonate buffer.  It can be run in hexane with a small
amount of pyridine also, as is the PFPA derivative.  Here I show it reacting with
carbaryl, giving the sulfonate from the phenolic portion.  One interesting thing
about this reaction is, if you run this reaction with pyridine and hexane, you do
not get a derivative for carbaryl.  You get nothing.  You don't get the sulfonamide,
which would be produced by the reaction of 2,5-dlchlorobenzenesulfonyl  chloride
and methylamine; however, when you do run this reaction with a pyridine medium
in hexane, you can get methylamine derivatives from both aldicarb and methomyl,
so the reaction seems to have a degree of specificity built into it, such that if you
wanted to, you could screen for methomyl and aldicarb.

  Figure 11  shows a reagent we have just begun to work with.  It has recently been
reported  on in an article in JAOAC on ethylene thiourea, I believe.  Anyway, it is

                          Figure 10
^  \O-C-NH-CH3


                            sulfonvl chloride
                      \J      °      u


pentaflurobenzoyl chloride.  It can be made to react quite quickly with

2-aminobenzimidazole, a metabolite of benomyl, and also it reacts with MBC, a

metabolite of benomyl, and benomyl itself to give the exact same product.

                          Figure 11

                A95°C,30  MIN,

  In Figure 12 we have a sandy soil spiked with benomyl at the one ppm level,
versus a five nanogram standard, which would be equivalent to one ppm.  You
can see that after having gone through the silica gel column there is a very good
recovery.  We didn't calculate that out, but it gives a nice, strong peak on a
DC-200 column at about 20 minutes.  We plan to work more with this reagent
this coming year.
                              Figure 12
10 % DC-200, 220°C
ATT.,  4 x|04
                                         A, 5ng  BENOMYL
                                         B, I  ppm SANDY SOI I
                                         C, CONTROL

  Bather than force you to undergo the reading of the data and all of these extraction
efficiencies,  I thought I would just summarize briefly. Using methanol benzene as
extraction solvent in a Soxhlet extractor, we had poor recoveries for these materials:
monuron,  Furadan, Bux,  Zectran, IPC, Temik or aldicarb, captan,  folpet,
Dlfolatan,  and somewhat a low recovery — it's 50 to 71 percent — of triflurailn.
  Obviously,  the methanol in the methanol-benzene extraction solvent is reacting
with the carbamates to give an transesterification product.  If you redo the extrac-
tions in acetone, however, you're not much better off. With monuron you get 11
to 30 percent; Furadan,  50 to 60 percent; Bux, 40 to 64 percent; Zectran, still not
detectable.  If you use benzene in a tumbling fashion,  you can obtain captan,
folpet, and Dlfolatan In 75-100, 85-100 and 80-82 percent recoveries throughout the
0.01 to 10 ppm range.
  In summary, the materials that we could not derivatlze are consequently gas
chromatographed. Zectran, Temik, amitrole,  Avadex, Azodrin, benomyl — all
those are questionable —  diquat and  ferbam, the latter two we had really no hope
for in the beginning.   We have extraction problems with monuron, Furadan,  Zectran,
IFC, and Temik or aldicarb.
  I would like now to jump right ahead and, if there are no questions on these ex-
traction studies, talk about something I think was a little bit more interesting, at
least from our standpoint. We wanted to see if we could  not flesh out the methodology
available for carbamates  since, as-you know,  they do decompose quite quickly
under gas chromatographic conditions if they are not derlvatlzed.  There have been
reported some techniques whereby large amounts of injected  sample have been
satisfactory; however, there Is a problem in doing this with detectors, such as
the microelectrolytic conductivity detector, of contamination of columns and
furnaces and things of that nature.
  We have been working with what we call a dynamic fluorogenlc labeling detector.
This is a device that is based around the idea that a high-speed liquid chromato-

 graphic separation can be achieved, and subsequent to the column a reaction can
 be made to occur that will convert a nonfluorescent material to a highly fluorescent
 material.  The basis of this revolves around two fundamental points.  One is that
 you can find a liquid phase in a column that will be compatible with a reaction that
 will take place to give a fluorometric response. The other is that you can find the
 reagent that will give a fluorometric response that has an almost instantaneous
 reaction and also has by-products and beginning products which do not themselves
 fluoresce and therefore interfere with the materials under investigation.
   We looked at a number of reagents, some on paper, two in the laboratory.  We
 spent almost all the contract year on one and essentially ran into a dead end.  It is
 a reagent called fluoram or  fluorescamine, which was developed at Hoffman-
 LaRoche.  It's a reagent that is specific for primary amines.  This materials reacts
 in milliseconds under pH 9 aqueous conditions at room temperature with primary
 amines to give a very strong fluorophore. Fluoram itself does not fluoresce, nor
 do the hydrolytic products which occur upon the exposure to pH 9 conditions.  The
 reaction product is  the only fluorescent material.
   This reagent had  been picked up by American Instrument Company  and made as
 the heart of their amino acid analyzer.  They had been working with it for about a
 year, and during the course  of our studies we consulted them and discovered that
 they had arrived at  the same conclusions we had, that is, that the requirement
 that fluoram be dissolved in  an aprotic organic  solvent was  such  that  when it was
 mixed with the pH 9 aqueous buffar which had to be added later (that is, you could
not mix or you could not add the fluoram to the buffer and have that react with your
 eluting compound) it caused extreme noise in the flow through cell and fluorometer
due to inhomogeneous mixing.  We went through a whole series of mixing devices,
heating devices,  tumbling devices, and found out that even after pumping in acetone
into a pH 9 buffer and giving as much as two minutes for mixing through these
various devices, we could not get rid of the noise.  American Instrument Company

discovered this also,  and they have now issued a disclaimer to this reagent as a
derivatization reagent for amino acids.
  One reagent with which we did work which was surprisingly productive was
orthophthalaldehyde.  This  is a reagent which has been around a while.  There was
a publication in 1971 by Mark Beth from a hospital in Geneva,  Switzerland, where
he reported that if you mix this reagent in the presence of mercaptoethanol under
conditions about pH 9 to 11, you get an extremely strong fluorophore.  This again
is a millisecond, if not a second or so reaction (Figure 13).

                             Figure 13
                            _.„ .     HSC,H.OH
                            RNH2     pH9-n  »  FLUOROPHORE
  After looking at this reaction in a static system, we went to a dynamic flow
scheme whereby we based our whole arrangement on the capability of methyl
carbamates to be hydrolyzed under very mild conditions quite rapidly (Figure 14).
This was quite surprising.   We knew that they could be hydrolyzed to give a phenolic
or hydroxyl product plus methylamine plus CO ,  but we were surprised at how
readily this could be made to occur; pH 11 is a somewhat mild hydrolytic condition.

                             Figure 14

         R-0-C-NHCH3   H2°H*0°C>   R-OH * NH2CH3 *  C02

  Using this information, we set up the arrangement illustrated in Figure 15.  The
heart of the dynamic fluorogenic labeling detector instrument was, of course, the

Waters 6000 pump, which allowed separations to be obtained on a micro CIS revers*
phase column, also obtained from Waters.  I might say that many long hours went
into selecting the right column for these carbamates.  We examined essentially
every adsorption, reverse phase and normal phase column, and even took a few
pot shots at ion exchange columns, thinking that they might have some  adsorptive
properties which were somewhat rare, and arrived at this  micro CIS column,  which
allowed us to perform a quite good separation for a number of N-methyl carbamates,
using 40 percent acetonitrile in water.

                            Figure 15
               MOBILE  PHASE
          I \v    1.0 rnl/min
T       WATERS 660

(>W '— •
25 u I

i , - ,.«., !
— pwui-aw^r1 —


fiTj- i^ ! r
• - c^ ,— ^*—j
ml/min REAGENT
1.0 mi/min

  I show dioxane there because we found that when we got the system going the
dioxane was incompatible with the orthophthalaldehyde.  Acetonitrile did give the
best separations; however, it has a tendency to decompose and give ammonia,
which reacts instantly with the orthophthalaldehyde and gives a high blank.
  Going over to the dioxane, we lost a little bit of efficiency in the column, but It
was exceptionally satisfactory in terms  of signal to noise and response.  After
going through the column, there is a mixing whereby pH 11 borate buffer is
metered in with the Milton Roy single piston displacement pump at a 0.5 ml per
minute.  It goes through a 35-foot mixing coll or reaction coil.  This sounds almost
impossible, but this was made 0.01 mm I.D. diameter. No bubbles were Involved,
as is the custom with Technicon; this was held to 80° C in a water bath,  went
through another mixing tee where orthophthalaldehyde reagent, along with the mer-
captoethanol, was metered in at a milliliter per minute. The whole system is of
Teflon; all of these tees and the tubing are of Teflon.
  It goes directly, with no delay,  through a flow-through fluorometric cell that
was 5 mm on a side.  We have a photometer and a readout on the  recorder and
then to waste.
  Figure 16 is a chromatogram of a mixture of five carbamates.  Although .Lannate
and Temik are really not carbamates, we always throw them in.  They do give
methylamlne upon the hydrolysis conditions that we have set down.  Observe the
relatively sharp peak shape.  We're talking about 50 nanograms per compound.
There is a 25 microliter sample, injection.  This is in water with a small amount
of dioxane, a few percent dioxane for solubilization.  Our experiments have shown
that this can be boosted to as much as 1550 microliters ~ that's a quarter of a
millimeter — on a one-foot analytical column without any degradation in

   This means that you can use very dilute samples and, therefore, theoretically
 approach very low limits  of detection.  This arrangement — notice Lannate comes

                                Figure 16
               50 ng
               25 ul of
         2000 PLATES
         uC18 COLUMN
         409/o DIOXANE-H-O
         1ml/mfn FLOW
out in about 7 minutes ~ gives 2,000 theoretical plates, calculated on Lannate.
The same solution injected upon the same column but with the Chromatronics 254
nanometer mode UV adsorption detector,  at 0.02  adsorbence units, gives 2,200
plates (Figure 17).  So we're talking about just a 10 percent increase in separatory

  Notice the additional noise at one microgram for each sample with the 25-
microliter injection.  We were able to run just a few analytical curves (Figure
18).  There is an almost linear response for Lannate from one nanogram up to
150 nanograms.  Baygon and Temik also give somewhat linear analytical curves.

  The limit of detection is approached at around one nanogram.  Figure 19 shows
Lannate in an analytical standard.  It's five or six times the noise.  The proof of

                               Figure 17
                                       220O PLATES
                                       U.V. DETECTOR
                                       25 ul of O.O4ng/ul
                                       uC18 COLUMN
                                       4O»/0 D!OXANE-H2O
the pudding is in the selectivity.  We spiked sandy soil, the soil you saw before,
with crude extracts, getting the large interfence peaks by DC-200 (Figure 20).
There's nothing there.  You get no solvent response.  This bump is kind of an
artifact in a detector; 0.05 ppm corresponds to 125 nanograms.  There are really
about three peaks, and these peaks seem to come and go in spiked soil samples,
which make  me think that there is some sort of decomposition going on.  You can
see the 0.5 ppm is way off scale.

      Figure 18
                o - BAYGON
                A- TEMIK
    10           100
      Figure 19
 25 ul of 0.04ng/ul

 2000 PLATES
 40 °/0 DIOXANE-H2O
 1ml/min FLOW

         0.0 ppm
                                   Figure 20
                             LANNATE IN SOIL
                               (SANDY LOAM)
              0.5 ppm
  This is without any cleanup at all.  Acetone was used to extract the soil. It was
evaporated off and then the residue picked up in water with a small percentage of
dioxane for solubilization. We spiked lettuce  with Lannate, refluxed it with 0.25
normal sulfuric acid for a half an hour, partitioned that into methylene chloride,
concentrated it down, and shot the raw extract; it was green (Figure 21).  You can
see the control.  There is a kind of a hole there; 0.2 ppm shows up quite nicely.
  I might add that you just don't go injecting these raw extracts on an analytical
column without some sort of treatment.  They were put through millipore filters
to rid them of suspensions of all particulate matter.

                               Figure 21
                          LANNATE IN LETTUCE
                    0.2 ppm
0.0 ppm

  Figure 22 shows the same sample injected on a UV detector.  Notice the 125-
nanogram standard which is equivalent to 0.05 ppm falls here in about 5 minutes,
which is completely obscured by interferences on 0.02 adsorbence unit scale of the
UV detector.  I think we have the basis with this device for doing some very good
separations and some high-sensitivity detections of these carbamates.  I might
add that we have not done  extensive interference studies.  We do know that N-phenyl
carbamates do not either hydrolyze or give the response, so we've run things like
IPC and CIPC and obtained nothing on this detector.

                                Figure 22
                             U.V. DETECTOR
                                     L ANN ATE
                                     125 ng
  We also know that substituted urea herbicides under these conditions give no
response, so it is fairly specific.  That, in a nutshell, has been our effort for this
past year on the contract extension of multiresidue methodology and, I suppose,
tomorrow you will hear some plans for the coming year on this contract.

   DR. ZWEIG: Do you believe that most of the N-methyl carbamates that you've
tested will undergo this hydrolysis reaction, if you can call it such, under these
conditions, pH 11, and the second point is what do you think is the by-product that
you seem to find in the small peak that you get there?
   DR. MOYE:  All of the N-methyl carbamates,  and I say  N-methyl carbamates,
will hydrolyze under these conditions almost 100 percent.  We're talking about 80
to 100 percent.  The by-products, I don't know.  There's something in the soil
(we were spiking the soil and immediately extracting it in just a few minutes).
It was dirty soil, if you will; it was not sterilized.  It had a high microbial
population, a high organic content. So there was something going on apparently
between the soil and,  in this case, Lannate to give some peaks that were not there
in either the check soil or the Lannate standards, and I cannot  even guess what
those products might be.
   DR. ZWEIG:  What is your retention in the detector and the flow rate of
   DR. MOYE: The delay in the 35-foot length of Teflon tubing is two minutes
and there is essentially no handspring, very, very little.
   DR. HILL: I know you'd rather talk about the fluorogenic labeling,  but I'd like to
go back to an earlier part of your presentation.  On the slide you showed  on the
various fractions of the silica gel cleanup,  what were the. volumes of eluant you were
getting for each fraction? I didn't see that on the slide, or maybe I missed It.
   DR. MOYE: That's a good question.  We eluted each fraction with 50 millillters.
   DR. HILL: These were essentially uniform or equal fractional volumes for  —
   DR. MOYE: They were equal.
   DR. HILL: You were changing the  polarity with the mixtures.

  DR. MOYE: Right. They were equal volumes, and they were concentrated down
to a volume of one milliliter, so that 0.01 ppm of the pesticides that were noted on
the chromatograms would give a measurable response.  la other words, under the
conditions of that separation in the 40 ml elution volume 0.01 ppm was detectable.
That's all I can say.
  DR. HILL: And then on the next set of slides where you showed the chromatograms
of the various soil extract fractions, was that  the same detector used for each of
those or —
                               " *
  DR. MOYE: It was electron capture.
  DR. HILL: It was EC for all of those?
  DR. MOYE: Yes.
  DR. ROSS: You mentioned silica gel. Did you look at any other column types?
For example, your Fisher aluminum, Woelm, and so forth?
  DR. MOYE: No, we didn't.
  DR. ROSS: And the consistency of separation per class of compounds for these?
  DR. MOYE: No, we did not.

  DR. ROSS: In other words, this was the only procedure you chose. You just
took the silica gel procedure as published by Shaflk and Sherma and tried thatl
  DR. MOYE: Yes. We scaled it up somewhat. They were using it to separate
materials in air, trapped air samples,  and they had a microcolumn arrangement
where they used, if I recall,  about one  gram in a five millimeter diameter chroma-
tographic column.  We used the two and a half centimeter, which required scaling
up to 16 grams of silica  gel and a subsequent  increase in the volume of solvent.
  DR. ROSS: What do you predict as far as enhancing sensitivity in the interfacing
of the LC with the fluorometer system?

   DR. MOYE: Maybe I didn't understand that.  Would you repeat that?
   DR. ROSS: Do you predict anything in the future about the enhancing sensitivity
for this particular type of detector system?
   DR. MOYE: For what particular compounds  or any of them?
   DR. ROSS: Any of them in the group or a particular class, particularly with the
heat labile compounds.
   DR. MOYE: We are able to see one nanogram of N-methyl carbamates.  The
people at American Instrument Company are seeing about 0.01 nanogram of amino
acid, using this orthophthalaldehyde.  So we're  at about a hundred-fold difference
somewhere along the line, and I don't think we're  going to pick up that additional
sensitivity by fine tuning the reagents and the conditions of the reaction. I don't
know what the difference in sensitivity is, whether it's due to the inherent response
of amino acids versus methyl-amine or not, but the point is that with this limit of
sensitivity for the N-methyl carbamates and the large Injection volumes that can be
handled under these reverse phase conditions on the analytical column,  there Is no
bottom to the limit of detection, I think,  in terms  of parts per million from samples
like soil in particular and water in particular.

                          Dr. J. S. Warner*

   During the past year, Battelle's Columbus Laboratories have been involved in a
program aimed at identifying impurities in some of the candidate substitute pesticides.
The objective is to acquire enough information about what's present to avoid the pos-
sibility of selecting a substitute pesticide which has some toxic impurity that makes
the substitute material as  hazardous or more hazardous than the material it is intended
to replace.  Perhaps the most dramatic example of a toxic impurity is the tetrachlorodi-
benzodioxin found in 2,4,5-T.  This impurity, which went undetected for many years,
is so extremely toxic, teratogenic, and mutagenic that its presence in the technical-
grade pesticide must be kept below one ppm.  Naturally, we would like to know whether
any candidate substitute pesticide contains any unknown contaminant having a toxicity
similar to that of tetrachlorodibenzodioxin.
   The task of identifying  every impurity present in each  technical-grade pesticide
and determining the biological effects of those impurities is a monumental task indeed.
From a practical standpoint, it may well be an impossible task.  Nevertheless,  there
is merit in studying candidate substitute pesticides to obtain as much information as
reasonably possible about impurities present and their potential hazards.
   There are at least three possible approaches to studying the impurities present
in technical-grade pesticides. One approach involves studying the manufacturing
process,  identifying the main reactions and the numerous possible side reactions
that may occur, listing the potential impurities that may result, and estimating the
the likelihood of their actually being present at ppm or even subppm levels in the
final product.  A  second approach Involves taking an actual sample of the technical-
grade pesticide and subjecting it to the many separation and analysis  schemes available
*Battelle's Columbus Laboratories, Columbus, Ohio

         Figure 1.  Main Reactions in the Preparation of Parathion
          Step 1.
T ^2°5
penta sulfide
                                   2(RO)2 P(S)SH

Step 2.
          Step 3.
                     (R0)2 P(S)SH
 (RO)2 P(S)C1

          Step 4.   Cl
                      benzene             chlorobenzene

                                C1  \ ^/ N°2
                                p-nitro chloro-
          Step 5.   ci
                       \	/
Step 6.     (R0)2 P(S)C1
Table 1:  Reactions Occurring During Production of
' 16.

inOTT IT*" i

(RY)2P(Y)SR 4 R'OH:
frtY*i TWinrt t
•/DVli D/V1CIJ J. V f\
. (±ti J_ r( i jail T •WoW
(RY)2P(Y)YH 	 ^
(RY)2P(S)OR' T 	
(RY)2P(S)SH 	 »-
(RY)2PSH4 R'OH •—
(RY)2PSH + R'OH —
Generalized Reaction^*'
2(RO)2P{S)SH + H2S
	 ^{RY)2P(Y)SR' +
> f^^Y^ r*fv\OTT i n
(RY)2P(Y)YP(Y)(RY)2 4
:*• (RY)2P(O)SR'
(RY)2PSH + S
	 >- (RY)_POR' 4 H.S
£» £»
	 *• (RY)2PSR' 4 H20


(RY)2PSH — i^L. (RY)2P(S)SP(YR)2
(RY).PSH 4 Cl, 	
6 b
l*t * /-> *^V'«3 )« ** « Cl— ™"
2 2
t .
2S Cl + 2H 0 	
ArH 	 " ArCl
ArH 	 *• ArNO2
ArCl NaOH, A, ox.
-*-(RY)2P(S)SP(YR)2 +
k /nV^ TV* I'1'" TV" W"
22 2
*- SO2 4 4HC1 4-S

*. ArOH


      Table 1. (Continued)

ArCl + ROH -
ArCl + ArONa
(RY)2P(Y)C1 +
Generalized Reaction'4'
	 >• ArOR + HC1
	 >• ArOAt + NaCl
	 " O^rJP} + 2NaCl
ArONa 	 *• (RY)2P(Y)OAr + NaCl
+ ArONa 	 >- (RY)(ArO)P(Y)OAr + RONa
(RY)3PY + ArONa 	 »- (RY)2P(Y)YNa + ArOR
(RY)2P(Y)C1 +
(RY)2P(Y)C1 +
fT>Y) pc l^J
(RY),PY + H,C
iJ f»
RSH — tP-L. R<
(RY)2P(Y)YNa 	 >- (RY)2P(Y)YP(Y)(YR)2 + NaCl
ROH 	 >- (RY) (OR)PY + HC1
•*- (RY),PO
) 	 ^ (RY)2P(Y)OH + RYH
     (») R » »lkyl or iryl group; Ar » wyl group; Y • oxygen or sulfur.

   The Identification of impurities present in technical-grade parathion will serve
as an example.  The manufacturing process is outlined in Figure 1.  Steps 1 and 2
for the preparation of a chlorophosphorodithionate can potentially give rise to a myriad
of exchange and disproportionation reactions that are typical of organophosphorus and
sulfur chemistry.  Steps 3 to 5 for the preparation of the p-nitrophenoxide are included
because of the various chlorinated, nitrated, or phenolic compounds that may be formed
by side reactions.  A list of the various generalized reactions that may occur during
production and storage of parathion are shown In Table 1.  In addition to the main
intermediates, the reactions can give rise to all types of phosphate esters, phosphites,
pyrophosphates, ethers, and even dlbenzodloxins.  A list of many of the possible
contaminants that could potentially be formed by these reactions is  given in Table 2.

Table 2:  Possible Contaminants in Parathion

2 (RO^P^SH
3 (RO)2(RS)PS
4 (RO)3PS

8 (HO)(RS)2PS
8 (RO)(RS)P(OR)2
11 (RO)2P(S)SSP(S)(OR)2
12 (ROfePSB
13 (RO)3P
14 (HO)2(HS)P
18 (HO)2P(S>SP(OR)2
18 (RO)2(RS)PO

17 (RO)3PO
18 (HO)2P(S)Cl
19 (RO)(RS)P(S)d
20 (RO)2P(0)C1
21 (RO)(RS)P(O)C1
22 3,01,
23 S
24 C8H8
Mf* W C*\
ft S
26 C H Cl
S 4 2
2T O H /Nft \Ol
28 C6H3(N02)2Cl
29 C6H3(N02)C12
30 C6H2(NO2)2C12
31 CaH.(OH)a
o 4
32 CgH4(N02)OH
8 + 7
16 + 17
18 + 17
Likelihood o(
Being Present In
Table 2:  (Continued)








C H (NO )C10C B NO
632 6
64 642
I O T I O 1
C§X ®~N02
O N-^T )^^-NO
2 O/^x^
63 22
(BO) P(0)OC H NO
A v ^ *

27 + 32




18 +32
18 + 34
3 or 52°



22 or 27
22 or 27
22 or 27.

22 or 27
22 or 27
7 +6
Likelihood of
Being Present in
Step(d> Final Product(e>
5 o
5 o
5 +
5 o
5 o
5 0
5 0

5 0

6 o
6 o
8 "^*
6 +

6 o
6 o
6 Parataion
6 +
6 o
6 *
; * **
"S "^""^
ft ^A
6 or St
ft or St +
St o
St • +
St +
Q4 .,
St +
a +
(a)  »"C2H5

(b)  Precursor refers to the contaminant numbered in this table.  SM - starting material.
(o)  See Table 1.
(d)  The step (Figure 2) during which the contaminant might be formed.  St indicates
    possible formation during storage.
(e)  o • likely present only at levels below 1 ppm
    + • likely present at levels above 1 ppm but below 100 ppm
   ++ »likely present at levels above 100 ppm.

                  Figure 2.  Gas Chromatogram of Technical-Grade Parathion
9    12    15    18    21    24    27    30    33    36    39    42    45    48    51    54    57   CO
                                   Retention time in minutes

   The next step was to obtain a gas chromatogram of an actual technical-grade sample.
Since we were not interested in determining the amount of parathion present, but only
interested in detecting as many impurities as possible,  we used a very concentrated
solution, 10 percent, and ope rated the gas chromatograph at as high a sensitivity as
practical.  The resulting gas  chromatogram showing 19 components is given in
Figure 2. A GC-MS run was  then made in an effort to identify as many of these com-
ponents as possible on a one-shot basis.  The results are shown in Figure 3. We
identified ten components,  all of which were included in the list in Table 2.  The
phosphoryl chloride is an intermediate which we had expected would be more completely
hydrolyzed.   The three ethyl phosphates were expected as by-products from the reaction
of P S  with ethanol.  The p-nitrophenol intermediate was  expected, of course, and
    & o
the p-nitrophenetole was formed by reaction of the chloronitrobenzene with ethanol.
One of the parathion isomers, probably the second one,  is undoubtedly the thiolo
compound.  The first isomer may be an ortho-nitro compound. We did not have
reference samples of the impurities or model compounds to permit us to determine
which isomer was which.  The dithio analog was also an expected impurity.  The
amounts of these  impurities present range in general from a few percent to about
                 Figure 3.  GC-MS of Technical-Grade Parathion
                               '.'"'••••:-i'".	;"•'••",	; i  'i"i' .  .--i  -,- .--i . -
                          S3  1CB 110 123 133 1M 1S3 1SJ !TO ISO 135 2C3 210
  or isomer
                                                            3 C23
                                                                      5- !C3 -. •! 2 C IBS 2t3

              Figure 4.  GC-MS of Technical-Grade Methyl Parathion
 o I-
                                                             METHYL PARATHION
                                                                or isomer
I' 'I'"I-"!""!""'""!	)....l.ii.|i.i.|....1i..|i...|iii.|.n.|	[	
S3  60 73 80  M  103  US 123 133 110
        13  Z3  35 10
                                              Tsj 110 iso
& zia 'za & 'zm
100 ppm.  In this cursory study, no attempt was made to identify contaminants present
at low ppm levels.
   Similar results were obtained in the study of methyl parathion.  The manufacturing
process for methyl parathion, of course, is  essentially  the same as that for parathion
except that methyl alcohol is used in place of ethyl alcohol.  Seven of the impurities
analogous to those found in parathion were identified.  (See Figure 4.)
   Another common organophosphate pesticide, malathion, is prepared by a somewhat
different scheme as outlined in Figure 5.  In this case,  in addition to the various
possible phosphorus-containing impurities, there is the very likely possibility of
contamination by maleate intermediates and  hydrogen sulfide or methyl mercaptan
adducts.  The GC-MS results are given in Figure 6.  The eight components identified
include two methyl phosphates, a pyrophosphate, diethyl maleate,  and the hydrogen
sulfide and methyl mercaptan adducts of diethyl maleate.  The malathion isomer is
again probably the thiolo isomer.

                        Figure 5.  Synthesis of Malathion
       Step 3b
        Step 1
Step 2a   II    \>   +
         CH- C
and      anhydride

Step 3a   II    * 3
Step 2b   II    0
                    +  Vs

                            4-  (CH30)2P(S)SH
         maleic       "0,0-dimethyl
         anhydride    phosphorodithioate
                 (CH 0) P(S)SCH-CS
                   J  *     i     r
                              +   2C H OH
              0,0-d inethyIphosphorod i thioy1
              'succinic  anhydride

                                                 phosphorodithioate    sulfide

                                                     euccinic  anhydride
                                                  32     I     25+ H,0
   The synthesis scheme for the organophosphate, phorate, also called Thtmet, is

given in Figure 7.  In addition to the formation of the usual organophosphate by-

products,  it is quite likely that the hydroxymethyl intermediate will react with a

second molecule of the dithioate to give a bis phosphoryl compound or react with the

ethyl mercaptan in Step 3 with the loss of water instead of hydrogen sulfide and form

the ethoxy analog.  The GC-MS results are shown in Figure 8.   Four ethyl phosphate!

were detected, a pyrophosphate,  the ethoxy analog,  and the bis phosphoryl compound.

                Figure 6.  GC-MS of Technical-Grade  Malathion
                                                    Figure 8,   GC-MS of Technical-Grade Phorate

                                      I	•;""|"'H"ii|iLii|i.ii|yi|iiii(.iii[Miniiinlii.|,ii,|i.i,r''[	iini-iiii.!!!!!!!.^...!^!!.^..!!!!''!!"'!'''!!"''!""'"''!"''!"''!"''!"''!"''!"-!"''!"''!"''!"''!"''!""!""!"''!"''!"''!"''!

                        0    10  20   3d   K3   SO  60   70   BO • 90   100  118  120  130  110  ISO  160  170  180  130  230 210  220  230  210  2S3  2CO  270

                                 flLM NLM5EB

   A fifth organophosphate studied was Dursban (chlorpyrifos).  The synthesis scheme
is shown in Figure 9. An alternate method of preparing the intermediate 3,5,6-
trichloro-2-pyridinol is shown in Figure 10.  The various possible side reactions
are listed in Table 3. In addition to the usual side reactions, there is the possibility
of getting various substituted pyridines and also dipyridodioxins.  The impurities
actually identified by GC-MS are given in Figure 11. These include two ethyl phosphates,
a pyrophosphate, the intermediate trichloropyridinol, and the tetrachloro analog of
   In addition to the  above five organophosphates, we also studied some other classes
of pesticides.  One of these, PCNB, is one of the most pure pesticides available.
The synthesis scheme is shown in Figure 12. The primary by-products expected
are the various chlorinated and/or nitrated benzenes.  It is also possible to get trace
amounts of sulfonic acids,  biphenyls, biphenylenes, dibenzodioxins, dibenzofurans,
and diphenyl ethers.  None of these latter possible contaminants were found.  As
shown in Figure 13,  four different chlorinated and/or nitrated benzenes were found.
   Another aromatic nitro  compound studied, trifluralin, is prepared according to
the scheme in Figure 14.  Here there are possibilities for all sorts of aromatic by-
products having chloro, fluoro,  nitro, methylol,  carboxyl, aldehyde, hydroxyl,
ether, amino,  and/or substituted amino functionalities.  In fact,  we came up with
a list of 146 possible contaminants without taking into account positional isomers.
The gas chromatogram (Figure  15) showed 14 components. In the GC-MS analysis,
as shown in Figure 16, only two chloro compounds, in addition to four trifluralin
isomers, were found.
   The final pesticide that I'll report on Is  folpet.  This material is prepared from
phthalimide and perchloromethyl mercaptan according to the  scheme in Figure 17.  The
various reactions that might occur during the production or storage of folpet are listed
in Table 4.  In addition to derivatives of phthallc anhydride,  derivatives of  maleic anhy-
dride,  and oxalic acid  formed during the initial oxidation are possible.   Various sulfenyl,

                    Figure 9.   Synthesis of Dursban.
Step 1
           pyridine   chlorine
                                                  + 2HCI
                             2, 6-dichlorop/ridine    hydrogen chloride
Step 2
                          +  NaOH
                                150-155 C^

2,6-dichloropyridine    sodium     t-butyl
                       hydroxide   alcohol
Step 3
   HO  N   Cl

6-chloro-2-pyridinol   chlorine
                    «» '
                                            3, 5, 6-trichloro-
+ 2HCI

Step 4
                                   O,O-diethyl phos-
Step 5
Step 6
  O,O-diethyl phos-
      O,O-diethyl chloro-

      (C,H,0), P(S)   ,

                                              O,O-diethyl chloro-
                               v»  .
                           3, 5, 6-trichloro-

            O, O-diethylO-(3, 5, 6-trichloro-
            2-pyridinyl) phosphorothioate

        Figure 10.  Alternate Synthesis of 3,5,6-Trichloro-2-Pyridinol
                                           ci             ci
       Step Al
       Step A2
sulfinyl, and sulfonyl Impurities are possible,  as well as acids, amides,  imides,
chloroamides, chloroinides, and maleic adducts.  These are included in the list
of possible contaminants shown in Table 5.  The major contaminants found in the
GC-MS study (Figure 18) were phthalic anhydride and phthalimide.  The N-chloro-
oxamic acid found was one of the least expected contaminants.  We also found this
compound in captan.
   In conclusion, I want to emphasize again that so far we have only just skimmed
the surface by identifying some of the more major impurities.  Undoubtedly, most,
if not all, of the impurities identified have been known to industry researchers for
some time.  There are undoubtedly hundreds of trace impurities which were not
studied at all.  The amount of work needed to identify all of the impurities that are
present and may potentially be as hazardous as a tetrachlorodibenzodioxin is dif-
ficult to comprehend.  A full man-year of effort on each pesticide could be very
useful,  but even then some of the lesser impurities would be missed.
   Based on our experience so far in this program,  several recommendations can
be made. One recommendation, which I mentioned at the beginning,  is to make as
much use as possible of the information that has already been accumulated by the
manufacturers and to build on that.

                                Figure 11.  GC-MS of Technical-Grade Dursban
                                                           (C2H50) 2P (S > °P (S} (OC2H
     Figure 12.  Three Methods for the Synthesis of PCNB
                                                 HN03 + H2SO4
  c> ^^^ ci
                  Figure 13.  GC-MS of Technical-Grade PCNB
-)-TT^tTTjT|nnr^mi-1-im1mili.n|Mil|lni|	nm1!""!""!""!""'"''!''''''^'"1"''!'"''1"''!""'''''^'
 S3 2G3 r?6  2«3 238  353 31S 323 323 310 333 3CI  373 333 330
                                                                         120  130
   Another recommendation is that of making greater use of high-pressure liquid chroma-
tography (HPLC) for the separation of impurities.  By using HPLC, the less volatile and
less thermally stable impurities, which may be overlooked in GC separation, can be.
readily isolated and studied by a variety of analytical methods.
   A third recommendation is to use HPLC to o b tain fractions for toxicological studies.
In this way, a fraction containing a number of unknown components, each present at very
low levels in the pesticide, might be shown to be harmless without requiring more extensive
and costly analytical work.

   One message that comes out of this work loud and clear is that a great deal more
effort will be necessary before we can state with any degree of assuredness that sub-
stitute pesticide X hazardous trace  impurities.

Step 1
SCep 2
Step 3
 Step 5
             Figure  14.  Synthesis of Trifluralin
                              + Cl,

step 4     TQ
             Ci      Sulfur Tetrafluoride
                     4- NH(C3H7)2
             1       Diisopropylamlne

                         p-Chlorobenzoic  Acid

                                                 4-Chloro-3,5-DlnItrobenzoic Acid
                                                   Recry'd ETOH



Figure 15.  Gas Chromatogram of 10 Percent Technical-Grade Trifluralin
                  12    15    18    21   .24   27   .30

                    Retention time in minutes
33   36   39   42   45
        Figure 16.  GC-MS of Technical-Grade Trifluralin
                                                      or  isotner
     3   13  M  30  10 SO  .-J 7B  iO 33  I(T3  :!0
                                                              IUO 2OJ 2IO ZXO

    Figure 17.   Main Reactions in the Preparation of Folpet
 Step l
  [01   v
           naphthalene      o-xylene
Step 2
       0   +  RH
                                            200-220 C
Step 3   2CS2    +    5C12 •   2  .  2CC13SC1      +   S2C12

         carbon     chlorine      trichloromethane-   sulfur
         disulfide               sulfcnyl  chloride   mor.ochloride
Step 4
 NH  +  CC1  SCI  -jj-^
^         J     HjO
            phthalimide       trlchloromethane-      F-OLPET
                             sulfenyl chloride
                                                            KSCC13 + Had

           Figure 18.  GC-MS of Technical-Grade Folpet
                   PHTHALIC ANHYDRIDE
         10 SO  09 79 M  » IB 110 13 IB 1« ,U !« I7J IU l» 350 218 SB 238 210 ffia !M TO J« iai MS JUJ

Table 3:  Reactions Occurring During Production of Dursban
                                 Generalized Reaction!')
              (x»l to 5)
                 + NaOH
   (xaO to 4)
                                                   +  NaCl
HO   N   Cl
4 C.H.OH + P,S. — •- 2(C,K.O),P(S)SH
* 3 * 3 * 3 £
(RY),P(Y)SH + R'OH -^(RYLjitYJSR' +
2 - Z
(RY),P(Y)SH + R'OH — *-{RY),P(Y)OR' +
2 2
(RY)2P(Y)SR + R'OH -^- (RY)2P(Y)OR- •»•
(PYJ p(Y)rtP P2SS^ (j^Y) P(Y)SR
(RY)2P(Y)SH f H20— »~
 Table 3:  (Continued)

                                                            . .

                                          G«n«r»tued Reaction'*'






           JRY)2P(S)SH -*-2POR' +  H S
           (RY) PSH + R'OH -«-(RY) PSR' + HO
          i(RY)2P(S)SH  t  C»2-»~ (RY)2P(S)£SP(S)(YR)2
           (RY)2P(S)SH +  C12
                            (x « 1 to 4)
          (C H O) P(S)0
                        (x « 1 to 4)
                        (x » I to 4)
                                           >.  (C2H50)2P(S)OH
                                                                              * K
Table 4: Reactions Occurring During Production of Folpet



1 ^^N. T X™*\ 1 1 0 *
Generalized Reaction
rOr ^ rOr ^ C«- ra' i«i
(RCO)20 + HjO 	 •
(RCO)20 + NH3 — — <
(RCO>2KH + KH3 	
RCOOH + NH3 	 £-—
2CS2 + 3C12 — 1^-
csci2 + ci2 	 «-
CSC12 + 3C12 + 2K2
2S2C12 + 2H20 	
4CC13SC1 Fe«- S2C
2CC13SC1 + C12 	
CCljSCl + HjO 	
CCljSOH + H20 	
CCljSOH •»• CCljSCl
2cci3soscci3 	 •
CC1 S02SCC13 + HjC
CC13S02C1 + H20 -
cci3sci + (Rco)2>:i
CC1,SO_C1 + (RCO).
3 2 .
CSC12 + (RCO)2NH-
(RCO)2NH + S2C12 •
(RCO)2NSCC13 + H2
^^ c ^^ r ™i
0 0 °
^l^ o,-§v
•-RCONHj + RCOOH v v (RCO)2NH + H20
•*• 2 RCONH2
RCONHj + H20
2CSC12 + S2C12
0 	 ^-CC13S02C1 + 4HC1
*-3S + AHC1 + S02
12 + 2CC1, + CC13SSCC13
-— 2CC14 + S2C12
— CC13SOH + HC1
— C02 + S + 3HC1
NaOU. cci3soscci3
-cci3so2scci3 + cci3sscci3
) 	 "-CCljSOjH + CC13SH
— »-CCl.SO.H + HC1
— -— (RCO) 2NCSN(COR) 2
— — — (RCO) 2NSSN(COR) 2
— -— R(CC13S)CHCHC1R
Cl 	 »- RCHC1CHRCC13 + SOj,
0 •• (RCO)2NH H- CCljSOH
(RCO)2KH + 2CC13SC1 (RCO)2NC1 + CC13SSCC13 + HC1
RCON-H2 + 2CC13SC1
 a.  R * alkyl or aryl group.

                 Table 5:  Possible Contaminants in Folpet
                         Likelihood of Bed
 Pre-    Reaction          Present in Final
cursor3  Humber^    Stepc       Product^

                                                      —     1
                                                      3     1 or 4
                                                      3     1- or 4
            4     2
            6     2
                                                      *     2
                                                      2S     s

Table 5. (Continued)
^— ••

Pre- Reaction
Contaminant cursor3 Number**
1! ,NH

Likelihood of Being
Present in Final
Stepc Product*1
4. or St
4 or St
4 or St


+ .

Table 5.  (Continued)







\/~~\\ ^

CH- Cx
CH-C * •>
s^ c c /^
Table 5.  (Continued)
ii yssx' ii
0 0
9 9 ^.
cn-c xC^/Ox
|l JISSK^ ("j
*™ V T *"""^'l>^ ^^
0 0 ^sx'^
0 .
1 XNS02CC13
v vncci'Csnr.i jcrcicoNiL
Likelihood of Being
pre_ Reaction Present in Final
cursor* Munberb Stepc Product11
9+16 22 4 +
12+16 . 22 4 0
9+12+15 22 4 0
12+17 23 4 0
27+17 23 4 0
36+17 19 4
29+17 23 4 0
6+17 23 4 0
11+17 23 4 0
13+17 23 4 0

Table 5.  (Continued)




1 >

Likelihood of Being
Pre- Reaction Present in Final
cursor* Number0 Stepc Product**
12+18 24 4 0
27+18 24 4 0
42+17 19 4

1 11 0
3 21
44 42 +
45 42 0
45 28 4 +
12 27 4 +

9 27 4 - +
».  Precursor refers to the contaminant numbered in this table.   SM - starting material.
b.  See Table 1.

e.  n.t step (FtRure. 1) during which the contaninant night be formed.  St indicates
    possible formntion during scorane.

<•   J - UVclv nrcscnt only at levels  below 1 PPm.   "
     + - Uki-ly present at  levels above 1 ppm but below 100 ppm.
    •*+ • JJkcly present at  levels above 100  ppm.

  DR.  LEWIS:  Scott,  you pointed out that you probably missed a lot of things
that would not go through the gas chromatograph, but a lot of the compounds you
showed, from the standpoint of polarity at least, look like the kinds of things
that you wouldn't expect to go through a gas chromatograph.  What sort of columns
did you use, and do you have any feel as to what kind of efficiency the columns you
used might have had for passing some of the polar components?
  DR.  WARNER: This was all done on a packed column, glass columns with on-
column Injection, two  millimeter I.D. packed with three percent OV-17.  So,
It's about as nonpolar  a column as you can get to try to avoid holdup of more polar
materials.  It's true that a lot of these materials are quite polar. That's the rea-
son why we need to go to as high a concentration as possible, to overcome some
of the absorption effects of the column.
  DR.  ZWEIG:  Dr. Warner, I want to commend you.  This  Is a very Interesting
piece of work.  I have a few comments and a very few questions.  Number one, I
believe, and Dr. Ross might correct me,  that you made a statement In your open-
Ing remarks that commercial 2,4,5-T contains less than one part per million of
the dloxln.  I believe It's less than a tenth of a part per million.  Ami right, Dr.
  DR.  ROSS: Yes, you're right.
  DR.  WARNER: I agree.
  DR.  ZWEIG:  Let me first make another comment. Dr. Fukuto, who unfortu-
nately  cannot be here  tomorrow, actually Is presenting some very Interesting
work which follows exactly your line of work. He shows that some of these trl-
alkylphosphorothloates and dlthloates which have been found, for example, In
malathlon and phenthoate, actually are synerglstlc and potentiate the toxlclty of
the parent compound.  Therefore, several preparations of malathlon, which  In
itself has an LD5Q for rats of something like 2500 mg/kg In  the presence of very

small quantities of these trialkyl phosphates and phosphorothioates, the antichol-
Inesterase compounds which you have actually found,  are synergized to give an
LD50 of a thousand or 500.  So I think your observations are correct and very  in-
  The problem is not only to find out the toxicity of the impurities, but also the
synergistlc and potentiating effect.
  PCNB is made by different manufacturers and I don't know which particular
sample you got,  but it is known that certain manufacturer's  samples do contain
a sizable quantity of hexachlorobenzene, and indeed you did  find some in your
particular sample.  Is that correct?
  DR. WARNER: Yes.
  DR. ZWEIG: My last point is, technical-grade methyl parathlon is sold as 80
percent material.  Now, there ought to be 20 percent of a lot of impurities, and
I know part of it may be xylene straight solvent, but you should have found a lot
of impurities. Would you  speculate on what those might be?
 ' DR. WARNER: They're the aromatic solvent and are in the solvent front. We
didn't attempt to identify those.
  COMMENT: This is not a technical comment on your presentation,  but just
for  Information purposes,  the FAO-WHO joint meeting has examined PCNB and
it's been under review, and the existence of the hexachlorobenzene Impurity, for
example, that was just mentioned,  is one of the more serious problems In the
residues in food.
  A lot of the HCB that probably shows  up in certain European cheeses and dairy
products is suspected to come from these sources.  So, there is an obvious cor-
relating problem here with these impurities. It's recognized in this case that  it
could be as much as a 4 percent Impurity in some grade of PCNB. There are

efforts to try to pass International regulatory guidelines to limit this to some
much lower value, hopefully.  So there are some efforts being made where these
impurity problems are recognized.
  DR. WARNER: The hexachlorobenzene that we found was under one percent.
  DR. HANSON:  On the last paper, I think at the end, he suggested that there may
be a better method of going at  it,  and I know when I attended the Denver meeting,
I was little chagrined at the  thought that you were going to start to look for impuri-
ties in these types of products  at a level down among the ten parts per million, be-
cause we have invested lots  of money into one of our compounds, something about
five man years within the five  years looking only at levels down to 0.05 percent.
  For the risk involved and  the benefits, there's got to be a better way of doing
it, and I think he did suggest that toxlcity work, acute studies, possibly other sim-
ple type mechanisms  have got to be a better way than doing It this way.  Some-
body's got a lifetime project for a hundred guys here.  I'm not blaming Battelle  —
that must be nice for them.
  But, there's got to be a way other than analytical to get at this question, and
I think, analytically, It's a nightmare.  It's an absolute nightmare, and who is to
say that the hundred parts per million is where you stop?  The thrust of the whole
thing is what it does lexicologically, and being a chemist and not a toxlcoiogist,  I
can't comment on that, but I will.  I think that some sort of screening work,  to
get at the answers to this, is the better way.  And, of course,  we do  this type of
screening work in that technical products are used for most of the acute battery
and most of the chronic battery of tests.
  Now, granted some of the metabolism work Is done on pure compounds, because
you can see what kind of mess you'd get into if you're trying to do metabolism on a
hundred compounds at once.  So,  I don't mean to say that the problem Is all solved
before we start — I'm just saying that  we do a lot of work that says that at least

the product,  as we know it,  lexicologically is acceptable or not acceptable.  Now,
I'm not talking about bioaccumulation of trace elements and everything.
  But, I think we have to look at the risk benefit of the whole thing, and decide
whether we can afford to go this way.  Because, I don't frankly think we can. I
think we're going to bog down.
  QUESTION: I'd like to comment on your statement.  I think, in the original
work scope to this project there was to be a theoretical study of the candidate
compound and to postulate or predict as many by-products that one could conceive
within the reaction pot, most of which could be exonerated just through the likeli-
hood of forming.  Secondly,  I thought in this theoretical part there should have
been both a toxlcologist as well as a teratologist working with your theoretical
chemist.  Should  they find a potential impurity, an impurity which might be a
threat lexicologically,  they should decide at which levels by a quick scanning
mechanism which Is  Part II.
  And then the In-depth study that's  Part III is to come.  Now, I see that you did
a quick GC scan,  and also you did the GCMS scan.  Now,  have you actually done
an in-depth study, even though you put the many products that could conceivably
be formed in the preparation of parathion itself?  Have you actually done an in-
depth study of all of those?  It looks very complex with the five slides that you
showed of the by-products, but yet, how many products have you actually identi-
fied and really looked for in the technical compound?  In what levels?
  Certainly we can't go to every candidate compound and look for impurities at
0.01 part per million — I think it was from 0.1 to 0.01 in the original proposal.
I think that's been changed somewhat since the original proposal work scope came
out.  Most of those compounds you can exonerate initially through the theoretical
study. Can you not?  Would you like to comment on that?

  DR. WARNER: Yes. That's true.  Most of them you can exonerate,  and we did
not go into a more in-depth study than what I showed.
  We did not have a toxicologist look them over and decide which ones can be
exonerated and which ones can't.  Now, ITm sure, just off the top of your head,
you can decide that some of them look like they might be pretty bad species, like
some of the analogs of dibenzodioxins, for example, the  dipyridodioxins.
  It could be a bad actor  at 10 ppm or even 1 ppm. There's nothing else on our
lists that might be as bad as a tetrachlorodibenzodloxin but with the dipyridodioxln
we might be in the same ball park.  It might require a major effort to prove that
it's not present at a toxic level.

                           Dr. Donald E. Johnson**
   Restrictions on the use of persistent organochlorine insecticides have led to
increased use of organophosphate and carbamate insecticides as substitutes.  Many
of these substitutes are highly toxic substances, manifesting their toxicity through
inhibition of cholinesterase which permits the accumulation of acetylcholine to reach
toxic levels.  Thus, the danger in the use of certain organophosphate and carbamate
insecticides is inherent.  Conditions for their safe use must be determined,  and
rules which assure that safe conditions prevail during use must be implemented.
   Although nearly all of  the cases  of severe poisoning or death have been among
persons in the mixing,  loading,  or application phases of the use of these insecticides,
more than 600 workers in California alone have suffered illness to some degree from
organophosphate insecticides during the years 1949-1973,  according to Spear ej^al.
(1).  Namba and co-workers (2) have given a higher estimate, the figure being 950
cases of poisoning due to organophosphates in the period 1957-1962.  Bogden and
associates (3) determined the cholinesterase activity in  a sample of migrant workers
in New Jersey and found that approximately one-third had activities below the lower
limit of normal.  Recognition of the danger of premature reentry has been given by
many others over the past 25 years, notably Quinby and Lemmon  (4) and Quinby et_
al_. (5).  Various Government agencies, especially EPA, have emphasized the
seriousness of the threat and have  taken action to reduce it.
   Danger in the use of organophosphates and carbamates might be reduced substan-
tially if a warning system were in  effect in a treated area during the time that the
insecticide residue remains at a health-threatening level.  The purpose of the work
undertaken was to develop a warning system based on the incorporation, or simulta-
neous but separate application, with the insecticide spray of chemical agents which were
odoriferous or visible.   Ideally, these agents would have volatility characteristics
 *Co-authors are J. D. Millar, L. M. Adams, S. W.  Seale, Jr., R. J. Prevost, and
  G. Zweig
 **Southwest Research Institute, San Antonio, Texas

 such that when they were no longer detectable,  either by smell or sight, then the
 level of residual insecticide would be low enough to permit safe reentry.  Other
 desirable characteristics of the agents would be low toxicity and low cost.
   The three insecticides involved in this program were methyl parathion, car-
 bofuran, and azinphos-methyl.  These were chosen because they are in wide use,
 are representative of their chemical types, and have vapor pressures which are
 spread more than one order of magnitude.
 Literature Search
   The principal objectives of the literature search were to identify odoriferous
 analogs of the insecticides of interest,  other compounds with outstanding odor
 properties (e.g., unusually bad or good odors,  low odor thresholds, long-lasting),
 and compounds which fluoresce,  phosphoresce, or change colors.
   Computerized searches for these compounds were made through the following
faculties:  CHEMCON; DOD Technical Searching Facility; NASA Scientific and
 Technical Information Facility; and EPA Abstract Search Center at Research
 Triangle Park. However, extensive manual searching of books, reference works,
and Chemical Abstracts was the major search effort.  Additional smaller searches
were required periodically in the program for technical guidance and evaluation.
  During the search period,  certain facts and theoretical aspects of the problem
had to be considered. Items receiving consideration are mentioned in the follow-
ing paragraphs.
  Volatility information on the three insecticides was of prime importance.   This
information is given in Table 1.

            Table 1:  Volatility Information on Three Insecticides

                                           Concentration in
                     Vapor .Pressure,       Saturated Vapor,        Boiling Point,
  Insecticide         20°,  mm Hg               20°C, ppb           760 mm He. °C
  Methyl parathion     0.97x10   (6)             12.8                 366**
  Carbofuran           0.52x10  *               10                   419**
  Azinphos-methyl     2.2x10  (6)                0.03                500**
  *'Extrapolated value based on vapor pressure data at higher temperatures supplied
   by manufacturer.
  **Extrapolated values based on information from reference (6) and manufacturer.
    Decomposition probably occurs before these temperatures are reached.

   Requirements for ideal odor agents were envisioned to be as  follows:
   • Vapor pressures should be close to those of the insecticides;
   • Molecular weights should not exceed 300 and, preferably,  should be below 250.
This requirement is based on the observation by Stoll (7) that no odorous compound
is known with a molecular weight above 300 and that the limit of perception,  for
many people,  begins around 250;
   • Odor perception thresholds should be below the saturated vapor concentrations
shown in Table 1 for the three insecticides.   Preferably, the thresholds should be
much below these values since the atmosphere sniffed in actual  field conditions
would never be saturated;
   • Chemical characteristics should be similar to the three insecticides since the
disappearance time of an insecticide is thought to be the sum of losses  from  such
factors as evaporation,  hydrolysis,  water solubility, photolysis, absorption, micro-
organism action,  oxidation, and isomerization.  Evaporation from surface deposits
is thought to be a major route of disappearance for many insecticides (8);
  •  The agents should be relatively nontoxic  originally and when they degrade;
  •  Odors of  the agents should evoke a sharp human response;
  •  The agents should not leave unacceptable residues.

 Laboratory Disappearance Tests
   The first experimental effort was to determine the disappearance times of the
 three insecticides of concern, potential odor agents,  and potential'visual agents.
 The objective was to match the disappearance times of the three insecticides with
 odor agents and visual agents.
   This work encountered the problem of uneven distribution of the substance under
 study.  A dilute solution of a substance in a volatile solvent does not deposit the sub-
 stance uniformly if permitted to evaporate unmolested on a flat surface.  Uneven
 distribution leads to unequal disappearance times for the areas of relatively heavy
 and light deposits.  Miniature spray devices were found difficult to control when
 spraying the small areas needed in the studies, and the stray spray droplets-required
 safety precautions which were onerous.
   Eventually, a system was developed which gave a fairly uniform distribution of
 insecticides and sensory agents  on glass plates which were then exposed to controlled
 conditions in the laboratory. A  4-inch frosted glass square was rotated on a turntable
 at 175 rpm while a selected volume of agent or insecticide in solution in toluene was
 applied to the  center of the plate.  Perfect reproduction of area covered and perfect
 uniformity of the deposit were not achieved, but the system gave fair-to-good repro-
 ducibility and  was very useful.   When the amount of solution deposited is 50 micro-
liters, the area covered  is approximately 30  cm  .  When all factors Involved except
 concentration  are constant,  then the amount of substance deposited per cm  becomes
dependent upon the concentration of substance in the solution.  Disappearance tests
were conducted at a temperature of 23° C (73° F) and with an air velocity over the
plates of 8 kph (5mph).
   In order to  follow the disappearance of the insecticides,  an analytical capability of
adequate sensitivity was  established using a gas chromatograph equipped with a flame
ionization detector.
      Carbofuran was found to chromatograph satisfactorily on a 20-inch, 1/8-inch
diameter SS tube packed with 10  percent UC-W98  on Chromosorb  WAW-DMCS (80-100

mesh) at an oven temperature of 150° C.  Other conditions are: carrier gas flow = 30
Ttd/min.; H  flow = 30 ml/min.; O2 flow = 240 ml/min.; injection port temperature =
2oo° C; and detector temperature = 200° C.  Methyl parathion was analyzed on the
same column at the conditions above, except an oven temperature of 160° C was
preferable.  Azinphos-methyl was also chromatographed on this column but at
temperatures of 250° c for the injection port and detector and 200° C for the oven.
  In order to determine the amount of insecticide present on the glass plates at the
desired intervals after application, a plate was rinsed with a stream of dichloro-
jjiethane which  was collected as it ran from the corner of the plate in a small glass
beaker,  The rinse was evaporated without heating, under a stream of nitrogen, to
predesignated volume and compared with a standard solution by gas chromatographic
  The odor agents were evaluated by two or more  Individuals sniffing the  plates.
The visual agents were evaluated by two or more individuals observing the plates
under short wave (254 nm) ultraviolet light.
yield Tests
   Outdoor disappearance tests were conducted for the three insecticides  by spraying
  everal plants  of the ornamental species Euonymus japonica and exposing them to the
  nvironment in the vicinity of the laboratory.  Periodically,  leaves were removed
for the purpose of determining the amount of residual insecticide.  The plants were
  prayed with emulsifiable formulations commonly used in agricultural practice.  The
    ay Was applied until the plant leaves were dripping wet; then the excess was
  emoved by tilting the plants and gently shaking until obvious pools disappeared.
   At the same  time, sensory agents were evaluated by exposing them by  incorpora-
tion in ^e sPray» on glass plates, in plastic membranes, and on paper strips.  The
   rpose of these tests was to ascertain whether or not the laboratory tests conducted
   rlier gave valid information and to prepare  for a field test later on involving
             crop spraying.

   Methyl parathion E-4 was applied at a theoretical coverage of 5.0 Mg/cm  (0.44
Ib. /acre). During the day, the plants were kept outdoors.  At night they were taken
into a greenhouse, since freezing at night was a possibility. The outdoor temperatufl
extremes during the exposure periods of the tests were 4 to 24° C (40 to 76° F).
Sunshine was plentiful and the relative humidity was generally low.  Wind velocity
varied from calm to 24 kph (15 mph).  The greenhouse temperature varied from
25 to 29° C (78 to 85° F), and the relative humidity was high.  Two tests were
conducted, each of 3 days duration.
   Guthion 2L (azinphos-methyl) and Furadan 4 Flowable  (carbofuran) were applied
at a theoretical coverage of 10 Mg/cm  (0.88 Ib. /acre).  The plants sprayed with
these products were kept outside except at night and when rain threatened.  During
these times, the plants were placed under a shelter which consisted of a protective
roof without sides.  Thus, the plants were sheltered from direct rainfall but were
otherwise exposed to the other atmospheric factors.  Shielding from direct rainfall
was  decided upon so that the disappearance time experienced would probably be a
maximum one under the test circumstances, rather than a minimum one which would
result if a downpour of rain was found to wash the plants free of the pesticide.  Even
so, the plants were wetted often with condensed moisture during foggy weather and
when the dew point was reached.   During the 55-day test period, the following general
statements about the weather apply:
  •   Temperature extremes were 0.50 C (33° F) and 32<> C (90° F).
  •   Wind speeds were most often 16-40 kph (10-25 mph) with occasional gusty
periods where the wind speed reached as high as 72 kph (45 mph).
  •   The relative humidity was usually moderate; however, extremes were
  •   Cloudy skies prevailed more than 50 percent of the  time.
  Subsequent to the above outdoor tests, a limited field test was conducted on
a cotton farm near Batesville, Texas.  The cotton plants, in the early blooming stages

were sprayed by airplane with methyl parathion at an intended dosage of 17 ^g/cm
(1.5 Ib. /acre) and Galecron at an intended dosage of 0.95 Mg/cm  (0.083 Ib. /acre).
The main purpose of the test was to follow the disappearances of the methyl parathion
from the cotton leaves and of visual and odor agents placed alongside of the field.
Leaf samples were taken at 20 minutes, 2, 4, 8, 22.5, and 48 hours after application.
Observations of the sensory agents were made at the 4,  8, 22.5, and 48-hour points.
   Spraying was completed at approximately 8 a.m. At that time, the wind was
blowing  at an estimated 30-40 kph (18.5-25 mph).  During that day and the remainder
of the test period, the wind blew steadily at that velocity, or slightly higher, with
frequent gusts to 56 kph (35 mph), except for the period near dawn when the velocity
dropped to the 8-16  kph (5-10 mph) range.  The days  were sunny with only scattered
light clouds.  The temperature span during the test period was from 25.6 to 34.4° C
(78 to 94° F).  The relative humidity ranged from high in the mornings to moderately
low in the afternoons. No rainfall occurred during the test period.
   Prior to the spraying, a small table was erected approximately 150 meters into
the field so that the table top was essentially at cotton plant height.  Twelve glass
plates,  5 x 5 cm  , were fastened to the table top by placing them on strips of tape
having adhesive on both sides.  This was  done  so that the turbulent air streams from
the spraying airplane could not dislodge the plates from the table top.  The  purpose
of the plates was to collect spray for analysis and comparison with leaf samples  and
the sensory agents on glass plates.  After the first samples were taken, the table top
was removed from the supporting legs and kept in the shade of the cotton plants.  It
had been found in the earlier preparatory tests that exposure in the shade more closely
matched the disappearance of a substance from a  glass plate with the disappearance
 of that same substance from a plant in the sun.
   At the time that the table with the glass plates was set up, 10 cotton plants sur-
 rounding the table (none closer  than 3 meters, none farther away than 20 meters)
 were tagged with white cloth strips.  This was done so that sample leaves would always
be taken from the tagged plants or the ones on either side of them.

   After spraying was completed, sampling consisted of removal of two of the glass
 plates and one canopy leaf from the windward side of each of the designated plants
 (or the ones on either side of them).  The upper sides (exposed to the aerial spraying
 of the plates were washed thoroughly with a small stream of dichloromethane (DCM),
 and the washings were collected in a single glass bottle for transport to the laboratot
 for analysis.   The leaves were placed  on each other, making a stack 10-high, on a
 polyethylene-covered board. A No.  15 laboratory cork borer was  used to cut through
 the leaves, giving a circular sample of each leaf of about 2 cm in diameter.  This
 sample was ejected from the borer into a glass beaker,  and  the process was repeated
 two times, with each set of leaf samples being put into a different beaker. Each set
 of 10  circles was washed 3 times with 15 ml of  DCM, using a small stainless steel
 spatula to make sure that agglomerates of leaf circles were  broken up and each
 circle contacted by the solvent.  The washings from each set of 10 circles were
 bottled in glass bottles and transported to the laboratory for analysis.
  At the laboratory,  each bottle of DCM washings was reduced to the standard
 volume of 3 ml under a stream  of nitrogen without heating, and the concentrate was
 analyzed by gas chromatography under conditions given above.  The concentration of
 methyl parathion in each sample was calculated from peak height measurements  on
 the chromatograms of the samples and a standard solution of methyl  parathion present
 on a per-square-centimeter basis of upper leaf surface or upper plate surface.
 Samples of leaves taken before  spraying showed no interference on the chromatogram*
 at the points of interest.
   During the hour preceding the spraying, two  visual agents and three odor agents
were put under test in a shaded area beside the cotton field.  The visual agents were
 anthracene and phenanthrene.  The odor agents were skatole, 2-phenylethanol, and
 /3-phenylethylphenylacetate.  The visual agents  were applied to frosted glass plates
 only.   The odorant skatole  was  applied to a frosted glass plate and also to paper  strip*
 The other two odorants were applied only to paper strips. Observations were made
 during the test by the two men conducting the test. Attempts were made to procure

disinterested observers from the population in the vicinity, but the remoteness of
the field made such arrangements unworkable.  As a consequence, an opinion survey
was taken, at another site, of the response of five field workers to the sensory items.
Cost Analysis
   The first step taken in the cost analysis was to develop concepts of packaging the
warning items and methods of field use.  These concepts were restricted to systems
to be used with methyl parathion, since warning systems devised for this widely used
insecticide seem to be most readily applicable for field use.  With packaging and use
concepts in mind,  cost estimates were prepared for three methods of field application
as being the sum of these cost factors:  1) product development,  2) manufacturing
cost,  3) distribution cost, 4) installation cost,  and 5) operational costs. Information
on the usage and cost of methyl parathion was obtained from pertinent literature and
by telephone interviews with knowledgeable persons at several universities and
agricultural extension services, and incremental additional costs of the use of the
warning system were estimated.  Cost estimates are based on data for similar mass-
produced items using financially conservative figures, i.e., benefits were given
minimal values and costs were given maximum values in terms of 1975  dollars.

Literature Search
   The literature search did not reveal any candidate odor compounds from the same
chemical families as the insecticides.  Use of the modifiers "odor," "smell," or
"odoriferous" in connection with names of the insecticides, or their families, always
resulted in zero citations, whether the search was by computer or manual.  These
negative results offer the options for believing that there are no such compounds or,
if such compounds exist,  their odoriferous characters are not cited as key pieces
of information. The insecticides of interest have appreciable odors emanating from
the emulsifiable concentrates,  but manufacturers' representatives are of the opinion
that most of the odors are from low molecular weight, volatile compounds present as
minor impurities.  Hence, the odors dissipate rather rapidly from the areas treated

                                                      Table 2:  Candidate Odor Agents

                                                            Boiling Point(15)

artificial musk, 1-tertbutyl -3 -methyl -
2,4,6 -trinitrobenzene
typical synthetic macrocyclic musk,
musk ambrette
musk xylene
musk ketone
muscone-active principle in natural
musk, 3-methylcyclopentadecanone
civetone, 9-cycloheptadecen-l-one

2 -isopropyl -3 -methoxypyrazine
2 -propyl -3 -methoxypyrazine
alpha -ionone, beta-ionone


ethyl vanillin
deca-trans, trans-2,4-dienal
2 -methoxynaphthalene







t. ° C


pr. , Odor Thresh. ,
mm Hg ppb (v/v)(^)


1. x 10"5
12 0.0 1.3
atm. 0.016
atm. 0. 1 1
                                                                                                              Odor Character
                                                                                                       like natural musk

                                                                                                       like natural musk

                                                                                                       like natural musk
                                                                                                       like natural musk
                                                                                                       like natural musk
                                                                                                       natural musk
                                                                                                       disgustingly obnoxious,  becoming
                                                                                                       pleasant in extreme dilutions
                                                                                                       like fresh bell pepper
                                                                                                       like cedar wood in strong dilutions;
                                                                                                       like violets in extreme dilutions.

                                                                                                       like vanilla

                                                                                                       like vanilla
                                                                                                       like fried chicken
                                                                                                       like old nerolin

Table 2:  Candidate Odor Agents (cont.)
     Boiling Point

methyl coumarin
methyl anthranllate
1 . 7, 7-trimethylbicyclo[4. 4. 0 J-
1, 7, 7-trimethylblcyclo[4. 4. OJ-
11 -acetate
" -propionate
" -butyrate
" -acrylate
alpha-(4 -hyd roxycyclohexy Imethyl)
1 -(phenylethoxy)adamantane
gamma -decalactone
gamma -undecalactone
phenylethyl phenylacetate

t, °C
pr. ,
mm Hg
                              Odor Thresh.
     Odor Character
like vanilla

like vanilla
like concord grapes

can be used in perfumes
                                              like peppermint

                                              like peppermint
                                              can be used in perfumes
                                              like peach
                                              like peach
                                              like peach


                                                           Table 2:  Candidate Odor Agents (cont. )
                                                              Boiling Point
benzyl ctnnamate
hexyl cinnamic aldehyde
benzyl benzoate
benzyl salicylate
o -b romophenol
alpha and beta-santalol
t, t>C
pr. ,
mm Hg
                                                                                       Odor Thresh. ,
                                                                                        ppb (v/v)<16>
                                                                                       0. 001
         Odor Character
sweet odor of balsam

faint, pleasant, aromatic
like sandalwood
                Note:  Compounds for which no odor threshold data are presently available were included In the above list on the
                      basis of actual or estimated high boiling points and indicated odorous character.

 with the insecticide sprays.  Methyl parathion does release methyl mercaptan very
 slowly by hydrolysis, but a satisfactory warning odor level is not reached under use
   At this point in the program, with the theoretically most promising chemical
 categories having been eliminated,  attention was turned toward compiling a list of
 odoriferous compounds with emphasis on those boiling above 275° C and having, or
 estimated as having,  low odor thresholds.  This list is given in Table 2. While a
 number of very odoriferous substances are listed in Table 2, only some members
 of the musk family, benzyl salicylate, and benzyl cinnaxnate have vapor pressures
approaching those of the three insecticides.  On this basis alone, these compounds
appeared to be the best odor candidates in this program.  However,  this turned out
not to be the case for reasons given below.
  Table 3 lists  fluorescent compounds having boiling points of Interest,  Several
compounds from this list were selected for disappearance testing, as mentioned below.
Laboratory Disappearance  Tests
  Working with emulsifiable concentrates commonly used, disappearance times were
determined for methyl parathion E-4,  Furadan 4 Flowable (carbofuran), and Guthion 2L
(azinphos-methyl).  Two levels of deposit of each insecticide were applied in this work,
         2                          2
10 Mg/cm   (0.88 Ib. /acre)  and 5 Mg/cm   (0.44 Ib. /acre), simulating typical applica-
tions in the field. At both levels of application, only about 3 percent of the methyl
parathion remained on the plate after 24 hours had passed.  The high level of Furadan
showed about 20 percent remaining after  15 days and the medium level about 3  percent
remaining after 12 days.  The high level  of Guthion showed about 60 percent remain-
ing after 30 days, with this dropping to around 12 percent after 60 days. The medium
level of Guthion was not followed beyond the 28-day mark, at which point about 20 per-
cent remained.   The results are shown graphically in Figures 1,2, and 3.
  Fluorescing agents  were usually screened first at application levels of 10, 1.0,
and 0.1 Mg/cm  .  Then the application level was adjusted upward or downward  until

              Table 3;  Candidate Visual Agents
 trans -stilbene
 1,4 -diphenylbutadiene
 4 -methoxybenzophenone
 triphenyl phosphite
 l-naphthyl-2-tolyl ketone
 diphenylene disulfide
Boiling Point, °C
    305 (720 mm Hg)
    350 (720 mm Hg)
    364 (explodes)

           Table 3:  Candidate Visual Agents (cont.)
N -benzylsuccinimide
4,4f -dibromobenzophenone
4,4f -ditoly sulf one
9 -phenylanthrac ene
0,0' -quaterphenyl
1,2, -benz anthracene
1,8 -dinitronaphthalene
1,2,6 -trihydroxyanthraquinone
1,2,7 -trihydroxyanthraquinone
Boiling Point, PC
   405 (714 mm Hg)
   445 (decomposes)
   459 (717 mm Hg)

Figure 1.  Disappearance of methyl parathion
            from glass plates
                  Temperature.................23  C
                  Air velocity over plates ..,.,..8 km per hr (5mph)
                  Methyl parathion per cm2	10 ^g	

              Figure 2.  Disappearance of Furadan from glass plates
Temperature	 23° C
Air velocity over plates	8 km per ar (5 mpn)
Furadan per cm2	10 ^g	
bfl  T
           2    3
                             Time, days
                                                11'   li   13   U   IS
             Figure 3. Disappearance of Guthion from glass plates
                                 Temperature	23° C
                                 Air velocity over plates	 8 km per hr ( 5 mph)
                                 Guthion per cm  .	10 j*g	

                                                             r- 0—
                                                      After 60 days, the
                                                      amount remaining was
                                                      about 12
                           ii  14   15   IB
                             Time,  days


           Table 4:  Typical Data for the Disappearance of Fluorescing Agents
                                  from Glass Mates
                                             Visual Fluorescing Strength*
Fluorescing Agent and Level
of Application,  Mg/cm













*S = strong; M = moderate; W = weak; O = none
 disappearance times that were thought to reasonably match or bracket the disappear-
 ance time of an insecticide were attained.  Table 4 shows some typical disappearance
 data acquired while trying to match the laboratory disappearance time of methyl
 parathion which was 24 hours or slightly more.  Disappearance data were acquired
 on the following compounds in addition to the three shown in Table 4:  carbazole,
 triphenylmethane, trans-stilbene, 6-methylcoumarin, 2,5-diphenyloxazole,  chrysene,
 pyrene,  o-quaterphenyl,  and p-terphenyl.  Disappearance times in excess of 25 days
 were attained with carbazole, one of the outstanding long-term fluorescent compounds.
 Chrysene, another long-term agent, was eliminated from the study because it is
 weakly carcinogenic (10).  These tests indicated that the laboratory disappearance
 times of all three of the insecticides could be matched under laboratory conditions

and that anthracene and phenanthrene were as good as or better than the other com-
pounds for use with methyl parathion.
  In the course of the screening of visual agents,  the fluorescent compounds were
sometimes mixed with insecticides. Occasionally, partial or complete quenching
of the fluorescence  resulted.  The possibility of putting an insecticide over a fluor-
escing substance, which could then not be  seen until the insecticide disappeared,
was intriguing as a  concept for  a sensory warning system.  However, this phenomenon
was found to be erratic and uncontrollable in the limited tests which were conducted
to explore it.
  Candidate odor agents were screened initially as were the visual agents.  At
first, they were applied at a coverage of 10 Mg/cm .  However, none of the odorants
tested gave much residual odor after 24 hours had passed.  By increasing the amount
applied by a factor of 2 or more,  residence times of 1 to 3 days could be achieved with
compounds such as  the musks,  vanillin, and skatole.  The simultaneous application of
fixatives, such as mineral oil,  were not of any  substantial help. However, residence
times could be increased drastically by mixing the odorant in a solution of a polymer,
then casting the solution in thin films.   The polymers used in this application were
Ethocel (Dow ethyl cellulose - 48 to 49.5 percent ethoxy content - 10  ops) and Vinylite
VYHH (Union Carbide copolymer of 86 percent vinyl chloride and 14  percent vinyl
acetate).  This was the technique devised  by which the long residence times of
carbofuran and azinphos-methyl could possibly be matched.
  Screening tests were conducted on the compounds listed in Table 5.  The candidates
which had boiling points nearest those of the insecticides (the artificial musks,  benzyl
salicylate, and benzyl cinnamate) were judged unacceptable because  of their mild odor
character. Strongly odoriferous  compounds, when applied by themselves, did not
have adequate residence times.  The best position to take appeared to be a compromise
position in which the desired high boiling point characteristic is traded  off in favor
of increased odor intensity.  However,  in order to use such substances, which are
quite volatile relative to the pesticides carbofuran and azinphos-methyl, the evaporation

      Table 5: Candidate Odor Agents Subjected To Screening Tests
Benzyl cinnamate
Benzyl salicylate
Ethyl vaniUin
2 -Methoxynaphthalene
Methyl anthranilate
Ethyl anthranilate
Ethyl salicylate
Musk ketone
Musk ambrette
Musk xylol
Anisyl acetate
Methyl eugenol
Phenoxyethyl isobutyrate
Anisyl alcohol
Galaxolide  (musk)
6 -Methylcoumarin
alpha-B romo acetophenone
Dimethyl sulfide
Several proprietary
  perfume oils and odor
  masking agents

 had to be retarded by the polymers mentioned above.  In the case of methyl parathion,
 its disappearance time could be matched by adjusting the quantities of the odor agent
 applied to the glass plates or to paper strips.  Matching could also be achieved by
 incorporating the odor agent in the methyl parathion prior to spraying.
   Through these tests, good matches in disappearance times could be achieved
 with respect to methyl parathion,  and matching the disappearance times of carbo-
 furan and azinphos-methyl appeared possible.  The odorants chosen for the field
 work were skatole,  j8-phenylethylphenylacetate, 2-phenylethanol, eugenol, and
 « -bromoacetophenone.
 Field Tests
   The disappearance curve of methyl parathion from the ornamental species
 Euonymus japonica  is shown In Figure 4.  The time required for more than 90 per-
 cent to be gone is about three times that found in the laboratory (see Figure 1). This
 may be the result of the much colder average temperature to which the plants were
 exposed, since the test was conducted in mid-winter, as compared to the laboratory
 tests.  This Increase in disappearance time was matched by increased times for the
 sensory agents exposed on and beside the plants.  In this test,  anthracene, phenan-
 threne, and skatole functioned well as sensory agents.  Eugenol and «-bromoaceto-
 phenone were also tested as odorant  sensory agents at this time, but they were judged
 to be inferior to skatole.
   Disappearance curves for methyl parathion from the cotton leaves and glass
 plates during the Batesville test are given in Figure 5.   The three values for leaf
 samples at each sampling time were  averaged before that disappearance curve was
 constructed.  The methyl parathion disappeared a bit more rapidly from the cotton
 leaves than from the glass plates.  At 24 and 48 hours, approximately 3 percent and
 1 percent,  respectively, remained on the leaves, while about 14 percent and 4 percent
 remained on the plates at those time  points.
   The observations  made on the sensory items  by the two men conducting the test
are given in Table 6 in terms of whether or not  a definite visual or odor response
was noted and the intensity of that response or signal.  It may be noted that

    Figure 4.  The disappearance of methyl parathion
               from Euonymus plants
  0 •
                        Time, Hrs,
         Figure 5.  The disappearance of methyl parathion
              from cotton leaves and glass plates

A  —
                                           glass plates


5 t'o 15 20
• A 	
' o—
	 	 , 	 , 	 1 	 A__|
10 35 40 45 5>


 Table 6:  Observations Made on Sensory Items During Batesville Field Test




Skatole (on plate)
Cone. _

Time, Hours











Skatole (on paper)




• s



' S



strip lost)
S - strong; M - moderate; W - weak; VW - very weak; E - extinct; I  - irregular
pattern, 90% of material gone; I - irregular pattern,  75% of material gone;
/ - gives individual responses  wnen disagreement of the observers occurred.

disagreement occurred rather frequently between the two men with regard to the
odor evaluations.  This disagreement is noted in the table by slash lines which
separate the individual responses.
   Whatever time for safe reentry is chosen between the 8- and 48-hour points
in Figure 5,  the time is matched or bracketed by the  anthracene quantities used.
The phenanthrene quantities employed are useful only at the 8-hour point.  The
skatole on the plate matches the 22.5-hour point fairly well.  The  concentrations
of skatole on paper cover the time span with about the same options as the anthracene
The 2-phenylethanol concentrations used were too  strong and did not match well.
The lowest concentration of /3-phenylethylphenylacetate matches the 8-hour point
fairly well,  and the next concentration would probably have been in the useful range
but this remains a speculation since the test paper was blown away by the strong wind
   Disappearance curves for the outdoor tests with carbofuran (Furadan) and
azinphos-methyl (Guthion) are given in Figures 6 and 7.  At the end of the 55-day
test, only 3  to 4 percent of the initial deposit of carbofuran remained on the plant
leaves but between 25 and 30 percent of the azinphos-methyl remained.
   During this test,  the visual agents (anthracene and phenanthrene) were exposed
on glass plates, in Ethocel films on glass plates, and on glass fibers or paper
impregnated with Ethocel and Vinylite VYHH solutions of the agents.  The odor  -
agents (skatole, /3-phenylethylphenylacetate, and 2-phenylethanol)  were exposed
similarly.  The exposures on glass plates and of Ethocel films on  glass plates were
unsatisfactory means due to rapid disappearance of the agents from the glass plates
and the lack of durability of the Ethocel films on glass plates.  The items consisting
of the impregnated glass fibers or paper fared better but were not without serious
problems.  Skatole was found to be  unable to endure prolonged outdoor exposure
without suffering severe changes in the basic odor or premature loss of odor.
Changes in the fluorescent color and character of the anthracene and phenanthrene
strips  rendered them unsatisfactory within the first 1 to 2 weeks of the test.
/3-phenylethylphenylacetate and  2-phenylethanol retained their characteristic odors
during the entire 40 days they were exposed, but from the 3-week point on,  the

                disappearance of Furandan from Euonymus plants
  30      40
Time, days
         Figure 7.  The disappearance of Guthion from Euonymus plants

odors weakened in such a gradual manner that distinct expiration points were
difficult to perceive.  At the end of 40 days, virtually all odor had left the items.
  After the field tests were completed, a panel of five persons was exposed to
three warning devices and interviewed regarding their opinion of the potential use-
fulness of the devices.  Some of the panel did not detect the fluorescent light source
immediately and many indicated that the light source was dim and not easily noticed.
All recognized the color of the light source.  Most of the panel said that a warning
was indicated.   Concerning the unpleasant smelling device (skatole), all of the panel
detected a  smell and most indicated that the smell was strong.  All said that the
smell was  easily noticed and most said that a warning was indicated.   Relative to the
pleasant smelling device (2-phenylethanol),  all of the panel detected a smell and most
indicated the smell was weak.  Most said the smell was not easily noticed and approxl
mately one-half said that a warning was indicated.  Virtually none of the panel pre-
ferred the pleasant smelling device (2-phenylethanol). The panel was evenly distri-
buted between preferring the unpleasant smelling device  (skatole) and preferring the
visual device.  All indicated that the concept would be useful for children,  especially
use of the unpleasant  smelling device (skatole).
Cost Analysis
  The cost of methyl parathion applications  ranges from $2.50 to $6.00 per acre
with the norm being $3.00 to $5.00 per acre.  For conceptual Method 1, where
skatole would be mixed with methyl-parathion, the increased cost per acre is $90 or
18-30 times the present cost.  For Method 2, where skatole would be used in two
perimeter warning signs per field,  the incremental additional cost is  $11.00 per
application.  This cost as a percentage of the extremes of  costs of application varies
from 1 to 45 percent (see Figure 8).  For Method 3, where anthracene or phenanthrefl^
would be used in two perimeter signs per field, the total additional cost per appli-
cation is $12.00. However,  an additional cost is incurred in this method—the cost of
viewing the treated  sign with an ultraviolet light.  It is assumed that a minimum of
three readings  will  have to be made per application, resulting in a combined incre-
mental additional cost of $15 per application.  This total cost as a percentage of the
total application cost is presented in Figure 9 and varies from 1.5 to 60 percent.


                     Figure 8.  Incremental cost for odor device

 ti o>
 o> a
 S ~

                  $100,  11%

                       $200,  5.5%
10 -
              100   200              500

                        Total Cost of One Field Application
                                                               $1000, 1.1%

   The odor agents finally chosen for work in this program were skatole, /?-phenyl-
ethylphenylacetate, and 2-phenylethanol.  The disappearance time for methyl para-
thion was matched quite well with the proper concentrations of the first two of these
substances and not so well with the latter one.  These substances, and probably most
other useful odorant compounds,  tend to tail off so that the end point as determined
by human sniffing is not sharply defined.  The disappearance times  of Guthion and
carbofuran were not well matched,  generally.
   Although the quantitative aspects of sensory odor agents are not as precise as
desired, they do have a distinct advantage in that they are easily recognized by the
general population.   Even young children can be instructed that if they  smell a cer-
tain odor that this means danger; thus, they should avoid the area.  Some thought
was given to the incorporation of odor agents in the insecticide formulation as an aid
in preventing reuse of the formulation containers. This idea probably  has to be
rejected because of cost considerations and the difficulties which might be encountered
in getting Government approval of adding a new chemical to the formulation.
   In this program, attempts were made to develop sensory agents which would
assist in preventing premature reentry into sprayed fields by not only farmers and
field workers but also less knowledgeable persons such as young children.  This
program has clearly demonstrated that for insecticides with vapor pressures similar
to methyl parathion,  odor systems are feasible and provide a reasonable definition
of the times when it is safe to reenter sprayed fields. For insecticides like carbo-
furan and azinphos-methyl, the development of useful systems requires more work
but it does appear feasible.
  The visual agents of importance to the program were polycyclic,  aromatic hydro-
carbons which fluoresce under ultraviolet irradiation.  The compounds offering the
best potential to this program were anthracene and phenanthrene.  The disappearance
of these substances,  when used in properly selected quantities, matched the disap-
pearance of methyl parathion quite well but did not match the disappearance times of
azinphos-methyl and carbofuran.  The end points  of the visual agents tend to be
sharper and more easily read than the end points  of the odor agents.

  The visual agents provide a more accurate end point because people have less
variation in their sense of vision than they do for odors.  These agents must be
exposed to UV light in order to reveal their fluorescence.  This requires more know-
ledge on the part of the user, and thus, it would be less useful for the general popula-
tion.  This program has demonstrated that for insecticides with vapor pressures
similar to methyl parathion relatively accurate visual systems can be perfected.
Additional feasibility studies are needed to evaluate visual agents for insecticides
like carbofuran and azinphos-methyl.
  The fact that the opinion survey was indecisive as to whether the visual sensory
item or the unpleasant sensory item is to be preferred is compatible with the experi-
mental results obtained when these items were exposed beside plants treated with
methyl parathion. Only through additional experience could a distinct preference for
either device be identified by objective or subjective means.
  Warning systems, both visual and odorous, can probably be developed success-
fully for methyl  parathion. With  respect to the conceptualized methods  for warning
systems given in the preceding section, Method 1 is unacceptably expensive. Method 2
is less costly than Method 3.  However,  the deciding factor  on which of these methods
is best for general use is probably not a cost factor but,  rather, that which  evokes
human response in the most effective and reliable manner.  Both methods, from a
cost standpoint,  are acceptable for use in the cotton-growing regions where cost of
application of methyl parathion is in the lower part of the range given.  Probably
neither would be acceptable in the higher part of the cost range given.
   Comparison of the insecticide disappearance data from this program with that of
other investigators has been made despite the difficulties resulting from dissimilarities
in weather conditions, quantity applied,  length of tests, plant species involved, and
analytical procedure.  In this program, the amount of methyl parathion remaining 3
days after application was 4.8 percent of the amount applied on the Euonymus plants
and less than 1 percent on the cotton plants.  Extremes found in the literature for the
same time period were 23 percent (11) and 1.1 percent (12). In a similar comparison,
but for a 4-day period, 86 percent of the azinphos-methyl was found on the plants

 in this program,  whereas the high and low figures from the literature were 46 per-
 cent (13) and 16.7 percent (12).  For carbofuran, the comparison had to be made
 after 14 days had passed.  In this program, 39 percent was found to remain.  Only
 one value could be found in the literature,  and it was less than 0.4 percent (14).
 1.  Spear, R. C., etal., Environ.  Sci. Technol. 9; 308-313 (1975).
 2.  Namba, T., etjd., Amer. J. Med., J50:  475-492 (1971).
 3.  Bogden, J. D., ^jt jd., Bull.  Environ. Contam. Toxicol., 513-517 (1975).
 4.  Quinby, G. E. and Lemmon,  A. B., J. Am.  Med. Assoc. 166; 740-746 (1958).
 5.  Quinby, G. E., etjd., J.  Econ. Entomol. 51; 831-838 (1958).
 6.  Melnikov, N. N., "Residue Reviews", 36, Gunther & Gunther, eds.,  Springer-
    Verlag, N. Y. (1971).
 7.  Stoll, M., "Molecular Structure and Organolyptic Quality, " Soc. of Chem. Ind.
    Monograph No. 1, The Macmillan Co., N. Y. (1957).
 8.  Spencer, W.  F., etjd., Pesticide Volatilization, "Residue Reviews, " 49
    Gunther & Gunther,  eds.,  Springer-Verlag, N. Y. (1973).
 9.  Ridgeway, R.  L., "Methods of Integrated  Control in Cotton, " Document 8,,,
    International Cotton Advisory Committee  Meeting, Managua, Nicaragua,  1972,
    pp. 190-195. (
 10. Steiner,  P.  E.,  Cancer Research 15;  632-635(1955).
 11. Waldron, A.  C.  and Goleman, D. L., J.  Agr. Food Chem.  17; 1066-1069
 12. Ware, G. W., etjd_., Bull.  Environ.  Contam. Toxicol.  13; 334-337  (1975).
 13. Scott, W. P., etal.,  J. Econ. Entomol. 67;  408-410 (1974).
 14. Fahey, J. E., etjd,,, J. Econ.  Entomol. 03; 589-591(1970).
15. Appel, L., Am. Perfumer Cosmet. 79; 25-39 (1964).
16.  Dravnieks, A., Report No. HTRI-C8140-1 ET Research Institute Technology
    Center, Chicago, fll., March 10, 1969.
17.  Seifert, R. M.,  etjd., J. Agr. Food  Chem. 18; 246-249 (1970).
18.  Buttery,  R.  G.,  etjd.,  J. Agr.  Food Chem.  17; 1322-1327 (1969).

   DR. LEWIS: Don, did it happen to rain during any of these tests that you were con-
   DR. JOHNSON:  No.
   DR. LEWIS: I guess you didn't look into what might happen if it rained ?
   DR. JOHNSON:  As a matter of fact,  it did rain on one of our cotton plant studies,
but it washed off all the methyl parathion, so the farmer resprayed this field a week
or so later.
   DR. LEWIS: This is what I was about to get to.  I would think something like methyl
parathion might wash off much more easily than either your skatole or anthracene.
   DR. JOHNSON:  For completing the study of this type, say this Is a feasibility study,
if you were going to carry this to a product, you'd have to do  a good deal more cli-
matic studies, before you'd say you had what you wanted.
   DR.  LEWIS:  Here's just a comment for something  I thought of, for what it's worth.
One of your odoriferous agents coumartn could also be used as a fluorescence agent.
If you dropped a drop of aqueous KOH on it and then exposed it to UV light, you'd
get a pretty bright  fluorescence there, you see.  You have a double bl-functional method.
   DR. JOHNSON:  Yes.  Coumarln doesn't give a  very good odor, though, unfortunately.
It's not sharp enough for us.
   DR. ZWEIG: Are there any other questions?  Comments?  Someone on the feasi-
bility of this idea?  Anything new from Dr. Durham on reentry times?
   See, the problem with the test is that we really  do not know what the level of para-
thion is at which it is safe to enter, and I think Dr. Johnson made a pretty good guess
at saying that if 90 percent of the methyl parathion is gone, he presumes that it would
be safe to reenter.
   However, once  lexicological studies will have been  conducted and completed,  I
 think that ~ if I'm correct, Don — your visual or olfactory agent strip can be tailor-
 made for different reentry times.

   DR. JOHNSON: Yes, particularly for pesticides. This is particularly true for
the pesticides like methyl parathion.
   Obviously, we've carried our feasibility study much further for that one, so we
know that you can vary the disappearance times fairly accurately in those short time
intervals, up to a couple of days.
   So, even other compounds — say it's 12 hours or 36 hours — could be manu-
factured in fairly close tolerances.


                       BY ESR HYPERFINE LABELING
                         Dr. Barry Commoner *

   It is clear that the last paper was a test of an idea, and ours too is a test of an
 idea.  The idea was simply this,  that clearly in the Pesticide Program there was
 a serious problem of detecting and characterizing,  and if possible,  identifying
 the numerous complex and often unknown aromatic compounds — many of them
 quite similar to others  — isomers and so on.  The methodology for that has
 been based largely on such things as gas chromatpgraphy with mass spectrometry
 attached to it.  tt is cumbersome. It is difficult to  separate isomers.  Sometimes
_yarious kinds of errors come into it.	     .   .......  	
   Several years ago, we at Washington University  — Dr. Vithayathil; Dr.  Weissman,
 a physical chemist; and I — got the idea that it  might be possible to develop a whole
 new type of analysis, particularly for aromatic compounds by converting the com-
 pound of interest into a free radical.  That can be done by either adding one elec-
 tron or taking one away, by controlled reduction or oxidation. The moment you
 have a molecule with one unpaired electron, the paramagnetic properties of the
 unpaired electron can be detected by the technique called electron spin resonance.
 It is simply a form of spectrometry in which you place the material into a mag-
 netic field; the unpaired electrons, being like little  magnets will be aligned in  two
 orientations toward the field, one with a north pole  pointing  south and the other
 with a north pole pointing north.  The energy levels of these two populations dif-
 fer, and the energy gap is a function of the magnetic field. You can set up a
 resonance absorption of energy,  providing you  have the right frequency, which
 for a typical Magnet of  3,000 gauss turns out to be in the radar field.   Actually,
 it is hard to vary the frequency in this field, so we  do the reverse.  You fix the
 frequency of the radiation, then simply sweep through the magnetic field.  So if
 you  have an unpaired electron present — using  X-band radiation which is a typical

 * Center for the Biology of Natural Systems, Washington University, St. Louis,

pesticides of various kinds to animals and then looked for residues of the pesti-
cides in urine, blood, and feces.  We are nearly finished with this, and we are
in the reporting phase. I think you can get a pretty good estimation of what the
technique will do.
   Figure 1 shows the ESR spectra of various pesticide free radicals we have
tested.  Now, let me pick out a typical free radical, for example, that of
Azodrin.  The way the scan is achieved is that the magnetic field is varying slow-
ly in the horizontal direction. The signal which we see here is the first deriva-
tive of the absorption of the microwaves. In the  case of Azodrin it is mainly
just one signal here; that  is an unpaired electron, without much splitting.  We
call the splitting "hyperfine splitting. " That goes back to the history of  spectres^
copy, and hyperfine  splitting is indicative of the molecular structure.
   Now let us look at Figure 1 starting from the top.  The first compound is PCNB.
In other words, it is a free radical of a pure preparation of PCNB, obtained by
reduction with glyceraldehyde under alkaline conditions.  What we have here are
three sharp peaks which appear at set magnetic field intervals.   The next com-
pound is methyl parathion (reduced now by an electrochemical technique that I
am going to describe later on).  This is a very elaborate  series of hyperfine
splittings.  Reduction of trifluralin  by glyceraldehyde under alkaline conditions
yields a signal which has  a triple pattern which repeats itself.  Dinoseb  also
yields a signal with  several splittings upon reduction with glyceraldehyde.
Fenthion and captan yield distinctive signals uppn.oxidation with trifluoromethane
sulfonic acid (TFMS).
   Now, there is a group of compounds whose  free radicals are formed by  a
technique that we tried out, which involved irradiation with X-rays. If you put
X-rays through a material you can knock electrons out, or add them in, and you
are  never  sure whether you are going to oxidize  or reduce.  Normally this does
not give hyperfine splitting because if you work in solution, the free radicals
interact very quickly.  If  you work in a solid, the electron is smeared through

                                Figure 1
                     (Red., Glyc.)
                     Me. PARATHION
                      (Red., Elec.)
                      (Red.,  Glyc.)
                       (Red., Glyc.)
                       (OX., TFMS)
                       (Ox., TFMS)
     9  -  2.005
                                 10  Gauss
            (X-ray,  Ada.)
                                                                  (X-ray, Ada.)
\J  \ |
                                                                   (X-ray, Ada.)
             (X-ray, Ada.)
                                                                  (X-ray, Ada.)
                                                                   (X-ray , Dext.)
   (X-ray, Ada.)

g - 2.005

 the whole structure, and you don't see a very definite thing.  However, there is
 a unique technique in which it is possible to put a compound — molecule by
 molecule — into a cage and them X-ray it.  This can be accomplished by using a
 compound like adamantane, which is a three-dimensional hydrocarbon.  In
 practice you make a joint crystallization,  for example of bromacil and adamantane,
 and the bromacil molecules end up caged in the adamantane.  We then X-ray it,
 making free radicals of the bromacil molecules, but since they are caged, they
 can't interact. This is bromacil simply X-rayed in adamantane.   In other words,
 all you do is reprecipitate it  with the  adamantane, dry it,  X-ray it, and then look
 at the ESR signal.  In the same manner we have obtained signals from methoxychlor,
 simazine, propham, ferbam and dursban.  Using a slight modification of this
 technique,  i.e., X-irradiation in polydextrin, we were able to produce the free
 radical of Azodrin.  Polydextrin, which is a short chain starch,  is capable of
 making helical cages around compounds which will crystallize in the helix.
   These are simply to show that one can  make up an album of such signals which
 could be used to identify the compounds.  All these are down at levels of about
 0.5 to 1 microgram of material in the  instrument, and I must say, the instrument
 can be made much more sensitive than this.  These are quick runs.  If you did
 slower  runs of about 15 or 20  minutes  the sensitivity could be enhanced.
   In Figure 2 we see the free radicals of several metabolities of the pesticides
that we have been working with.  The first signal is that of p-nitrophenol, a metab-
olite of methyl parathion,  which turns  up in the urine of rats fed methyl parathion.
This signal has a characteristic pattern.  The second signal is from the urine  of
rats administered trifluralin.  This signal also has an  interfering signal super-
imposed on the signal due to the as yet unidentified metabolite of trifluralin.  The
third signal in Figure 2 is that of an unidentified metabolite of fenthion,  which we
find in the feces of rats that have been  fed fenthion.  We have not yet tried to
identify it.  The emphasis in our contract  was not on making final identifications,
but in characterizing them.   The last signal here  is  that of an unidentified


                             Figure 2
                                        METABOLITE OF Me. PARATHION •
                                        (Electrolytic Reduction)
                                        UNIDENTIFIED METABOLITE OF
                                        (Glyceraldehyde Reduction)
                                                      10  Gauss
                                         UNIDENTIFIED METABOLITE  OF
                                         FENTHION -
                                         (TFMS Oxidation)
          DINOSEB -
          (Electrolytic Reduction)

g = 2.005


 aviation type of radar — if you use that kind of radiation at 3,000 gaua, you will
 get a resonance absorption.
    The reason this is very useful as a means of identifying and characterizing
 complex compounds is that the magnetic moment of the electron interacts with
 the magnetic moments of certain nuclei.  For example, the nitrogen nucleus is
 also magnetic and, as the electron swims through the molecule, it senses those
 paramagnetic nuclei with which it comes in contact and gives characteristic
 splittings of the signal.  One can get very elaborate split signals,  and the pattern
 of  splitting,  the spacing,  the height, the number of them, etc.,  are so character-
 istic of certain types of structures that an experienced individual can even simply
 look at a structure and begin to say, "Well, yes, this is a  nitrogen and there are
 two hydrogens near the nitrogen, and so on. " Apart from  that,  of course, one
 can make known patterns and then recognize them very easily.   That  was the
 basic idea that we thought of — developing techniques for making free radicals
 out of substances of interest,  and then showing whether or not we could detect
 them and characterize them.
    One of the other advantages is that it's a very sensitive technique.  In our lab,
 we've obtained as low as lO-'', 10- molar free radicals and detected them.
 There is, on the horizon, a modification of the technique,  a Fourier Transform
 electron spin resonance instrument, which will probably improve things by two
 orders of magnitude. It is inherently a very sensitive thing. It is pretty fast,
 and the  most important thing perhaps is that we thought we might be able to use
 it to analyze mixtures without making any physical separation.  I'll explain that
 point when  we get to it. At any rate, this was the  basic idea.

   In discussions we had with the Pesticide Program it was decided that it would
 be a good test to see if we could find and characterize pesticide molecules in
 pure substances initially, but, in particular, in the bodies  of animals that had
 been fed pesticides, particularly to see if we could pick up metabolites, as well
 as the pesticides themselves.  Besides doing some basic studies, working with
pure substances, we did a series  of experiments in which we fed technical-grade


metabolite of dinoseb which you find In the urine of rats fed dinoseb.  It is not going
to be terribly difficult to identify either of these compounds as soon as we get
around to it. What this shows is that one can get characteristic, hyperfine signals
from pesticides and from these metabolites that we have picked up in pesticide-
fed rats.
   Now I want to talk about the electrochemical technique of producing free radi-
cals, because it turns out that this has certain remarkable advantages which
probably will make it the method of choice.  Incidentally, free radicals normally
interact very readily with oxygen,  so usually all of this work has to be done under
high vacuum.  For example,  when we reduce with glyceraldehyde under alkaline
conditions,  the reactions are carried out in an evacuated cell.  We make these
cells ourselves;  any glassblower can do it.  In other words, the reactions are
done under high vacuum, and the analyses are made under high vacuum except
for the adamantane.  The latter are usually quite stable and handle very well.
   Now, in addition to the chemical and X-ray techniques described above, the
oxidation-reduction reactions can be done electrochemically. Again, it has to be
under high vacuum, and the cell we use is shown in Figure 3.  There is a stopcock
and that goes to the vacuum line.  You can see there is a small gap between the
platinum and tungsten electrodes which is the part that goes into the ESR spectrom-
eter.  The cell also has a reservoir for the sample.  The method is very simple.
We simply dissolve the material in a solvent such as methanol along with a carrier
electrolyte such as tertiary butyl ammonium persulfate (TBAP), evacuate the cell,
and apply an electromotive force (EMF).  The EMF at which the free radical will
be formed is a function of its redox potential.  Since different substances have
different redox potentials, we thought there was the possibility of selectively form-
ing free radicals in the mixture by adjusting the EMF so that you could literally
titrate and bring in one free radical at one EMF and another one in at a separate
EMF and so on.  That is the great advantage of this and, I must say, we were
pleasantly surprised to find how well it worked.

            Figure  3
ESR Electrolytic Cell
                  -Stop Cock
                   Platinum Electrode
                  -Sample Reservoir
                                                       Tungsten Electrod*

   In Figure 4 we see the ESR spectra obtained from pure preparations of PCNB
and 2-amino-4-nitro-6-secbutylphenol.  The latter is a metabolite of dinpseb.	
Both of these compounds yielded characteristic spectra which could be easily dis-
tinguished from each other.  Figure 5 shows the result of electrochemical reduc-
tion of a mixture of these two same compounds. In other words, equimolar mix-
tures of these two substances are taken up, put in the electrochemical cell,  and
analyzed.  Initially  at-0.8 volts all you see is the PCNB free radical;  although
the other compound is in the solution, you don't see it.  We raise the voltage to
-1.4, -2.0 , -2.25 and -2.3 volts.  As you increase the voltage you can see that
the PCNB signal weakens and as we go from -2.25 to -2.3 volts, we suddenly
see the 2-amino-4-nitro-6-secbutylphenol free radical.  In other words, in this
solution, simply by altering the EMF, we can bring one free radical into play
and then the other one.
   Originally we thought that as we increased the EMF  we would get the second
signal  superimposed on the first, and we were prepared to do a computer sub-
traction.  But, in many cases, as you continue the redox process, the first free
radical goes over to a fully reduced or oxidized form and disappears.  And, in
this case it worked out that way quite nicely.   So, this is an artificial mixture to
show that the electrochemical separation really works.
   Another example of electrochemical separation is shown in Figures  6 and 7.
Here we are dealing with methyl parathion and  its metabolite p-nitrophenol.  Fig-
ure 6 shows the spectra of free radicals obtained from pure preparations of
methyl parathion and p-nitrophenol. Figure 7 shows the results of stepwise elec-
trochemical reduction of a mixture of these two compounds. At zero volts there
are no signals.  At -.25 volts the methyl parathion signal turns up.  As you go to
-. 75 volts you still  see it.  At -1.35 volts it disappears.  Now as you lower the
voltage you see the  p-nitrophenol signal.  So here, without separating them, just
by changing the voltage you see first one free radical and then the other.

                                 Figure 4
ESR Spectra of Reduced Pure PCNB and
2 -Amino-4 -nitro -6- sec. butylphenol
                                       Sfjn D..IK

                          Figure 5
Slepwise Electrochemical Reduction of a
Mixture of PCNB and 2-Amino-4-nitro-6-sec.butylphenol

                                Figure 6
ESR Spectra of Reduced Pure
Methyl Parathion andp-Nitrophenol

                          Figure 7
Stepwise Electrochemical Reduction of a
Mixture of Methyl Parathlon and p-Nitrophenol

    Next we will see how to use this technique in analyzing urine from rats that
 have been fed various pesticides.  Figure 8 shows the results obtained from, urine
 from a group of rats on a control diet.  First, let me tell you what the sample
 preparation procedure is.  The urine (25 ml) is extracted with hexane, the extract
.is taken to dryness, and the residue^is simply taken up in methanol with the carrier-
 electrolyte TBAP and placed in the cell.  Figure  8 shows an interference signal.
 This turns up in urine  and sometimes in feces. It is a well known free radical,
 ubiquinone7 which arises from the diet of the_rat.  So there it is and that is	
 at zero volts. Sometimes you get the free  radical form simply from the carrier
 solution.  As you make it more negative, this is -0.25, -1.5, -2.0 and -2.25, it
 gradually disappears.  So this is our background  signal, if you like, of urine from
 a control-fed rat.
    Figure 9 shows the  results from urine of dinoseb-fed rats.  At zero volts here
 is the ubiquinone  signal,  and as  you go down, the ubiquinone signal is still there
 but now a new signal is coming in at -2 volts and  -2.5 volts.  We have a pure sig-
 nal of this as yet unidentified metabolite of dinoseb.   In other words, the interfer-
 ence signal has disappeared yielding a rather clean signal of a metabolite of dino-
 seb.  We have no doubt that we will be able to identify the structure of this metab-
    Figure 10 shows  a similar run for urine from a rat that has been fed methyl
 parathion.  At first  we see the ubiquinone signal which disappears as you apply the
 EMF, and in comes p-nitrophenol, which is very  readily identified.  And this is
 simply a rat that has been administered this pesticide in the diet and whose urine
 has been collected and  analyzed in a rather simple way.
    In addition to the examples  given here,  we were able to obtain PCNB signals
 in the urine of PCNB-fed rats  and fenthion metabolite signals in the feces  of
 fenthion-fed rats.  To summarize, it looks to us as though this is a method which
 can be extremely useful in characterizing pesticides and their metabolic products
 in biological and environmental materials.  There is no reason why it shouldn't

                             Figure 8
   ESR Spectra of Urine Extracts from
   Rats Fed Control Diet —
   Electrochemical Reduction


                                                              C.   -1.5 V

                                      •V i
                                          q . 2.005
E.   -2.25 V

                              Figure 9
ESR Spectra ol Urine Extracts from
Rat* Fed Dinoseb Diet -
Electrochemical Reduction

                       Figure 10
ESR Spectra ol Urine Extracts from
Rat* Administered Methyl Parathion-
ElBctrochemlcal Reduction

work on soil,  on tissue samples, etc.  The remarkable thing about this is that it
takes rather little cleanup to get to the point where you can do a good analysis,
and, as I say,  it is inherently a very sensitive technique. As some of you know,
it is analagous to nuclear magnetic resonance,  and in that case that technique
was very much enhanced in its sensitivity by developing a Fourier Transformer
instrument. Several of us at the University'have been thinking about building
the first Fourier Transformer electron spin resonance machine, and in fact,
rudimentary parts of it are now  kicking around between our place, chemistry
and physics.  It  looks as though  it is going to work.  If it does, the entire sensi-
tivity of the process  at the physical end will be enhanced at least 50-fold,  probably
100-fold, which  means that the thing will become much more sensitive than it now
is and it is already down in the range of parts per billion anyway.
   Therefore,  we do feel that we have in hand an  idea which has been reduced to
practice, and that  it  is something that can be very useful in the Alternative Pesti-
cides Program.

                        Dr. Edward O. Oswald *

   The first area which I will discuss is mass spectrometry methods development.
Within the Environmental Toxicology Division of Research Triangle Park, we have
two extramural efforts on the Substitute Chemical Program, which have been awarded
during the later portion of FY-75.  One of these efforts is a feasibility study investi-
gating field lonization, mass spectrometry and Its applications to investigating phanna-
cokinetics, metabolism distribution, excretion of pesticides, and biological systems.
   The second area concentrates on interfacing a  mass spectrometry system with
high-pressure liquid chromatographic systems.  Within the framework of this con-
tract, which resides hi the Analytical  Chemistry  Branch,  this first of all concentrates
on the design — Interfacing hardware  to connect a high-pressure liquid  chromato-
graph with a chemical ionization mass spectrometer.  This particular award was made
to Flnigan Corporation, Sunnyvale,  California.  The project officer on behalf of
Finigan Corporation is Dr. Bill McFadden.

   DR. ROSS:  In respect to field absorption, I think if you just hit on what Is unique
about this particular technique,  and what you anticipate using it for, and on what
particular compounds. In other words, what can be done with field absorption that
cannot be done with chemical lonization, as well as electron Impact mass spectrom-
etry?  That's my first question. I have another one when you've answered that one.
   DR. OSWALD:  That I expected.  Numerous modes of lonization are utilized in
mass spectrometry. As you bombard the compound with a stream of electrons, you
have excess energy once you knock one electron.  In numerous Instances In the
analytical mode, you are unable to detect the molecular species as you analyze In
the molecular regions utilizing a high  energy, such as 70  EV.  In this Instance you
*Health Effects Research Laboratory, Research Triangle Park, North Carolina

 do have excessive energy which can bring about fragmentation of the molecules,
 minimize molecular species.
   In the case of field ionization, you are depositing a sample on an edge or point.
 A high-energy field is produced at the same time the sample is vaporized and ionized.
 In the case of chemical ionization, an indirect mode of ionization is involved in that
 a carrier gas, whether it's methane, isobutane, or helium, first of all is ionized.
 The charged species of these is then transferred by high pressure,  in this case in
 the ion source from the charged species to the solute, namely the compound of interest.
   In the case of both field ionization and chemical ionization, the level of energy of
 transfer for the ionization mode is much less than that of the direct electron impact
 mode of ionization.  So in this case you are looking at increased populations of the
 molecular species.
   In respect to  the field ionization contract, two compounds specifically considered
 for this program are parathion and trifluralin, both from the standpoint of metabolic
 investigations and inhalation studies.  From the metabolic investigations the normal
 sequence of events as expected would be administration of the compound,  analysis of
biological fluids, excreter and tissues, using the analytical technique  in this case,
field ionization mass spectrometry.
   I'll attempt later today to differentiate field ionization and field absorption.  In
the case of field ionization, there is  a requirement of volatilization of the material.
In field absorption there is not; it is  also a low energy process.  The requirement
for vaporization in field ionization has advantages and disadvantages.
   Going on further to elaborate  on the field Ionization contract, a difficult problem
in inhalation measurements is that of having an in situ mode of analysis of compounds
used in inhalation studies.
   DR. ROSS: The FI technique  has been used primarily for a  long chain, hydro-
carbons and things  of this nature, more so than other environmental compounds.

Since field ionization, field absorption is available commercially,  the uniqueness
of the contract at Stanford University is a feasibility study for pesticides, the in situ
studies as you were mentioning.
  DR. OSWALD: That's correct.  The other unique portion of this is development of
the multipoint source system.
  DR. ROSS: I am assuming that the purpose of this contract is to interface the
liquid chromatograph.  With these objectives you would like to be able to chromato-
graph those heat labile compounds.  Assuming that you accomplish this, and once the
compounds have been eluted along with the solvent to the source, do you know anything
about the source, the stripping process they're going to use to get the solvent away
from the compound,  and also the limitations that it has once it reaches the source?
Because one of the objectives of an interface system like this,  if you're going to do
a study with heat  labile compounds,  is  to have minimum contact of the surface area
once it reaches the ion source. And if I read the  proposal correctly, it seems as if
this is a derivative of the Scott method which used a heated wire.  There will probably
be maximum contact once it got into the source.   Can you elaborate on this?
   DR. OSWALD:  The general technique of interfacing with LC systems, then re-
moving the solvent from the eluting column prior to entrance into the ion source in-
volves the following processes.  A concentrating effect must be produced.  And with
most of these systems up to this point  in time, one percent or less of the LC effluent
enters the ion source.  One of the first criteria to be overcome and desirably to be
included in this program is to increase the amount of solute which enters the ion
   At this point in time a realistic figure is at least 30 percent.  From the standpoint
of concentrating the solute or  totally eliminating the solvent,  the mechanism used is
very similar to that initial work described by Scott, in that a moving wire — in the

 case of this particular contract, a modified ribbon or trough system — is Included.
 There are points of consideration which should be given in any system of this nature.
   You balance the volatility of the solvent with that of the solute.  As you go to very
 nonvolatile solvents, you'll always have a tendency of losing the solute.  At the same
 time we were talking about removing all of the solvent in this  mode of operation, not
 using the  solvent as a mode of ionization,  as was reported in other earlier investi-
   The flexibility of the system as proposed at this point in time is that from the
 Agency standpoint we felt that the system should be applicable to chemical ionization
 systems available within the Agency now,  either the Finigan system or the Varian
 systems.  The Finigan system, as proposed,  would have flexibility to operate in
 addition to the CI mode in the electron impact mode; without removing hardware the
 same system could also operate in the GC/EI mode, and a GC/CI mode,  in addition
 to the LC/EI/CI modes.
   I'll try  to clear up these abbreviations if there are any problems, without bogging
 down and using other time here.
   DR. ROSS:  With reference to the same question I just asked.   Selection of the CI
 mode — I think I understand, but several people here are not  well-informed.  The
 difference between electron impact and chemical ionization — there are  definite
 advantages of  one method over the other.  I think that maybe you should elaborate
 on the advantages of going to the CI mode in lieu of the electron Impact,  and since
the CI mode and pesticide residue analyses really are still at  a rather pioneering
 stage at this point, why we are pursuing this particular technique? '
   DR. OSWALD: To make the answer short,  flexibility within the system. To
 elaborate  on this, a system which is capable of operating in the mode of  chemical
ionization, namely an open, high-pressure source, can also be operated in an electron
 impact mode.  Electron impact spectra do complement chemical ionization spectra.

Now, without getting into the technicalities too deeply, polyfunctional compounds, as
I indicated earlier,  have a great tendency to have low concentration of molecular ions
formed in the direct electron impact ionization mode.  In that you are taking a stream
of electrons, bombarding a compound with this, knocking out an electron, and forming
an ionized species.
   In the case of a chlorinated species,  depending upon the structure of the compound,
you find a relatively high abundance of molecular species formed in an electron impact
mode.  In the case of certain biodegradables, where you've got polyfunctional nuclei,
such as nitrogen, phosphorous, oxygen, and others, the tendency to form a molecular
ion upon electron Impact mode of ionization is very low.  This means if you do not
know the characteristics of the molecular species, you are limited on how much you
can say,  or even quantitate about the total unknown.
   From the standpoint of considering why we approach the chemical ionization route
versus electron impact of the modes, as I indicated earlier, it is because of flexi-
bility.  If we can solve the chemical ionization problem, the means of getting the
solute into the ionization mode, it means minor adjustments in the reagent to simulate
the electron impact-like mode.
   For instance, a number of laboratories have taken just  the chemical ionization
mass spectrometer, using reagent gases such as methane or iscbutane.  In this
case you are dropping the total energy  level for the mode of ionization somewhere
in the range of 15 to 20 electron volts.   This means in order to break a carbon-
carbon bond, which requires the equivalent of 12 to 20 electron volts, you do not
have a large excess of energy available.
   As you choose your gas, you 'also choose the amount of electron excess energy.
If you want to look at it this way, you also increase the degree of ionization.  Taking
the other situation, you can substitute in chemical ionization  mass spectrometry
helium and get an electron impact-like spectrum,  in that comparing a true electron

impact mode where there is no reagent gas versus helium, you do see very useful
characteristics which mimic electron impact systems or characteristics.
   The flexibility and the complementation of chemical ionization and its capability
from a chemical ionization standpoint gives a lot of advantages,  more so than if we
would have considered only the electron impact system mode of  interfacing separately.
   DR. ROSS: What about the complexity between the two types of spectra?
   DR. OSWALD: If you read the literature — and I'm one not to generalize but
I've been involved In this particular area for a number of years  — and only after you
do the  experiment can you say in most cases the chemical ionization spectra are more
simple in that there is less excessive energy fragmentation.   From an interpretive
standpoint, there are normally less ions present in chemical ionization spectra than
there are electron impact spectra due to the level of excess energy available.
   At the same time this Is compound-dependent.  Numerous compounds such as PCB
have a threshold level for Ionization.  CI is not the most desirable mode of ionization
for these types of compounds.  You need more energy to virtually make a stable or
to form a stable Ion.
   DR. ROSS: I have three very quick questions, and maybe you've already answered
them.  I didn't get them.  One is, do you have to work in the mass spectrometry in
the atmospheric mode? In other words,  how do you Interface a liquid column and a
mass spectrometer?  Two, how do  you get rid of the solvent? And three, Isn't It
rather  unusual that the Agency went with a commercial company? That,  In a way,
may be commercializing the products.  Or is this a normal procedure?
  DR.  OSWALD:  Let me take the question concerning the  removal of the solvent.
The removal  of a solvent will be carried out by a stream of gas,  namely nitrogen,
and a temperature below that required to remove the solute but still sufficient to re-
move that of a solvent.

  In regard to the question concerning interfacing the LC to that of a mass spec-
trometer where you're talking about pressure differences, in this  case the interface
does require vacuum lock systems, so that the engineering techniques are required.
You're not working atmospheric pressure. You're working at comparable levels to
that of normal chemical ionization mass spectrometer of 10   tor, or in that range.
  In relationship to the contractor, at the time that this particular contract was
advertised, the idea — and the idea still remains — was that the expertise needed
to understand the hardware regardless of the system Is the most important thing.
You don't bring in a group which may have one expertise and try to fit it to a system
with which they're unfamiliar.  That's the hurdle on which we don't want to concentrate,
In this particular case we singled out two mass spectrometer systems common to the
Agency, which is not unusual.  In this case there are  approximately 35 or 40 systems.
  QUESTION:  Is that common?
  QUESTION:  Which are in the Finigan system?
  DR. OSWALD: A number of laboratories do have the Varian systems.  One of the
criteria which, by the way, is included on this work scope is once this  contract has
fabricated and evaluated this system, two systems will be released to the Agency at
the termination of this contract.  One will be evaluated in our laboratory, and one will
be evaluated in another one,  for application to environmental problems.
  QUESTION:  I think you just answered my question, but do I understand that once
the LC interface system is set up,  you have not created a dedicated mass spectrum?
That you can break that apart and use it for other purposes without a lot of cleanup
and so forth?
  DR. OSWALD: That's correct.  In fact, the proposal includes  that it will not
interrupt other modes of operation, namely GC electron impact,  administration of
the samples through the probe.  You still have the flexibility without taking the  hard-
ware off so that you can operate In the electron Impact, the chemical Ionization GC
modes, and also the LC modes.

   DR. HILL:  This isn't a question, it's mostly a response to some of the previous
questions as to a practical application of chemical ionization mass spec and pesticide
work. Many of you may already be familiar with this, but Dave Stalling out in
Columbia, Missouri is very interested in toxaphene in fish, and if you know the gas
chromatogram of toxaphene, you know what that looks like, and you probably all
heard of Cassda's efforts to resolve this into some 200 plus components.
   Chemical ionization mass spec, coupled with GC,  can take the seven fractions of
toxaphene, the early cleanup Cassda's fractions,  if you will.  I've seen Dave
Stalling's slides on this — they're beautiful.  It would be hopeless with conventional
mass spec to try to interpret this because each of these fractions  contains 20 to 25
components, but with chemical ionization you can clearly follow the loss of chlorine
or the numbers of chlorine items from the more early eluting compounds up to the
others, and you can get a fairly clear comparison of domestically produced toxaphene
versus other brands from other countries, which is becoming an important issue in
the toxicology of this material.  So this  is not a residue application but, after seeing
those slides, I think this is the only way you could ever make any  sense out of mass
spectra of toxaphene.
  DR. ROSS:  Thank you, Dr. Hill. I've talked with Dave Stall ings at the mass spec
conference the other week about his work, with reference to that.   I'm  aware of the
applications of chemical ionization.  There are numerous people who have had
questions related to chemical ionization and I thought this would be an opportunity
to differentiate among the various techniques. But one thing which I could have been
reading into what you  said earlier,  Ed,  in the development of this contract — you
said an interface system which would work for the common systems found among EPA
laboratories, which we named the Finigan and the Varian — since this  contract was
let to Bill McFadden of Finigan Instrument Corporation, I would find it very strange
that he would build an interface system that would also be adaptable to Varian

  DR. OSWALD:  The work scope reads either/and/or.
  DR. ROSS: Also there are two different types of instruments,  and I understand
that the interface system has to be specific,  either for your quadruple or for your
double focusing.
  DR. OSWALD:  That's correct. Now this is the other complexity as you become
involved with engineering of any system.  From the standpoint of the work scope as
proposed, it includes either the Finigan or the Varian systems.  We did not have a
company which would agree to tackle both with the available restricted level of

                            Kenneth R. Hill *
   This is a report on the status of an EPA-ARS Cooperative Agreement research
project entitled "The Development of a Semi-automated Cleanup and Analysis System
for Pesticide Multiresidue Analysis."
   Although existing methods (AOAC,  PAM, etc.) for the analysis of pesticide multi-
residues,  developed mainly for regulatory use on foods, have been very successful
for a variety of pesticide/substrate combinations, there remain a large number of
compounds (mainly OP and carbamate insecticides) which cannot be reliably mea-
sured by these techniques, especially in environmental samples.  For example, of
the P- and/or S-containing compounds in the priority list for the Substitute Chemical
Program,  only parathion, methyl parathion, and malathion have been determined by
the AOAC  multiresidue procedure and only the proceeding three plus azinphos-methyl
(and ethyl), demeton and phorate have been determined by the Abbott procedure (1).
   In each case the methods have only been proven for selected food groups and not
for a wide variety of materials such as soils, waters, aquatic organisms,  etc. An
urgent need therefore exists to extend present multiresidue methodology to the pesti-
cides  on the priority list and to search for and develop alternate multiresidue methods
which are  more universal in scope and offer greater simplicity and/or speed, espe-
cially in the cleanup stage.  This research project, based on  recent discoveries and
developments in the Analytical Chemistry Laboratory,  ARS, is directed to the latter
  . High-speed,  high-pressure liquid chromatography is limited in application by the
lack of selective detectors. One approach to overcoming this problem was made by
combining the P- and S-selective flame photometric  detector  with the  Pye traveling
wire system, thus enabling liquid effluents to be vaporized, detected,  and  measured
directly (2).  This apparatus and its capabilities will be described briefly.   Figure 1
 *Analytical Chemistry Laboratory, Agricultural Research Service, USDA, Beltsville,

              Figure 1

                                Figure 3
                              SAMPLE INJECTION
                                                              FLAME  PHOTOMETRIC
                                                                          GLASS TO METAL
                                                                          FITTING, 1/8" TC
                                        ALUMINO- V
                                        SILICATE  GLASS
shows a front view of the Pye traveling wire system with a flame photometric detector
installed in the back compartment.  The installation is rather straightforward and re-
quires only minor modifications to the Pye-Unicam system.  Figure 2 shows a top view
of the same apparatus.  Figure 3  shows a block diagram of the apparatus and its asso-
ciated liquid chromatographic columns.  Figure 4 shows the linearity achieved with
various concentrations of standard malathion over the range of 0.1 to 1 microgram per
milliliter eluted from a Corasil II glass bead column at  a flow rate of 4 milliliters per
minute.  Peak heights were linear with concentration with a correlation coefficient 'H1
of 0.99966 and a slope of 23.9 millimeters per nanogram. Figure 5 shows the
reproducibility that was obtained with repeated injections at the same concentration.
Figure 6 gives the names and structures of several of the pesticides used to test

    Figure 4.  Standard Malathion
                                     M- 23.93332 {SLOPE)
                                     i> • 2.01763  (INTERCEPT)
      0-5  I
                Figure 5

                              Figure 6
                : (DASANI7R )
          s   ^ o
   CH30    H2C-C-0-CH2
     ~ .""-..   "—. 0
CH3    0        O
the capabilities of this apparatus.  Figure 7 is a summary of the operations carried
out in low-pressure liquid chromatography on silica gel using step wise gradient elution
with increasingly polar solvents.  Figure 8 is a composite of a number of runs showing
the responses obtained for standard phorate and its 5 metabolites using the stepwise
elution shown in the preceding figure. Figure 9 gives a comparison, step by step, of
analysis by conventional GC and analysis by cleanup and direct detection. Figure 10
shows a liquid chromatogram of standard solutions of Dasanit and its metabolites and
Figure 11 shows the chromatogram of an extract of 10 grams of soil fortified with the
same concentrations of Dasanit and its metabolites.  Figure 12 presents the steps in
the liquid chromatograpbic analysis of oxamyl in alfalfa showing first the oxamyl
standard peak, second the extract of check alfalfa, and third the extract  of field-
treated alfalfa equivalent to a 10-gram sample.  Note that these liquid chromatograms
are aligned according to the time at which the acetone elution was started and not

                             Figure 7
  = S,S
                      I % ACETONE IN
                      BENZENE   50m!
                                           B    P=S,S02
                     5% ACETONE  IN
                     BENZENE   50rn!
                      7.5%ACETONE IN
                      BENZENE  50m!
                      10% ACETONE  IN
                      BENZENE  50m!
                         50 ml
                                           0    P=3,SQ
                                           E    P=0,S02
according to the injection time of the sample. This latter example was run in the
sulfur selective mode.  Figure 13 shows chromatograms of other pesticides to
illustrate the response to various sulfur, phosphorus or phosphorus- and sulfur-
containing compounds such as busulfan, coumaphos, and parathion.
   During the testing of this detector system with low-pressure cleanup type liquid
chromatography columns, it became obvious that the combination of cleanup column
and selective detector provided a new means of rapid screening for pesticide residues.
Excellent separation of pesticides and their metabolites was achieved by stepwise gradi-
ent elution and the extreme selectivity of the flame photometric detector for P and S
prevented interferences from most nonpesticidal components.  The major drawback was
the need to replace the silica gel column for each analysis.

                                Figure 8
      ICO       150
      EFFLUENT -  ml
 i % —i— 5 %	1— 75 %
    10 %~rACETON£
   The possible solution to this problem came quite by accident about 9 months ago
when it was discovered that EM silica gel SI-200A (mean pore diameter of 180 angstroms)
which is commonly used in aqueous gel permeation work could be programmed with
organic solvents to achieve almost any desired degree of separation of pesticides (3).
Impurities in organic extracts such as fats, waxes, and pigments were retained by
absorption near the top of the column but could be completely removed after analysis
by elution with pure acetone or methanol.  Such behavior is just the opposite of what
is normal for gel permeation cleanup based on molecular exclusion.
   In this project we will combine the Autoprep  1001 GPC cleanup apparatus with the
liquid chromatography flame photometric detector to devise a semi-automatic residue
analysis system for OP pesticides.  Since only about 1 percent of each sample reaches the

                                   Figure 9
                       GC Analysis
     Cleanup,  LC-FP
                      25  g.  sample
                    (1 pprn  = 25 vg.
         Weigh -
      25 g. sample
    (1 ppm » 25 ug.)
                   Solvent  extraction
   Solvent extraction
                  Filtration, partition
                        if  needed
                    Concentration by -
                  evaporation,  1-2 ml,
    Concentration by
  evaporation, 1-2 ml.
                  Cleanup -  silica gel,
                    alumina,  florisil
   Cleanup and measure-
      ment  in LC-FP
                    Concentration by
                  evaporation,  1-2 ml.
                     ul. injected  into
                     GC, measurement
fl/lOOth  of  eluant  sample ]
Breaches  detector.  250 ng.J

     1  ng =  0.004 ppm
               fl/lOOOtli of sample rcachesl
               (     detector.   25 ng.    J

                     1 ng. - 0.04 ppm
detector,  the remaining fraction can be collected for analysis of organochlorine pesti-

cides by GC, nitrogen-containing pesticides by GC or TLC, and the confirmation of

the OP pesticides by GC-MS, if needed or desired.  The research is  divided broadly

into three major phases.  First, combine the Autoprep 1001 and the LCFP detector

and investigate the performance of the combination for cleanup and screening analysis

of 5 P- or S-containing pesticides  selected from the priority list, using 5 substrates

as decribed later. Second,  determine the performance parameters and capabilities

of the SI-200A silica gel as a reusable cleanup and analytical column  packing in a

low-pressure chromatograph system employing the same pesticides and substrates

as in Part I. Third, if-the results of tests under Parts I and n are satisfactory,

                                     Figure 10
                                 PYROLYSIS OVEN TEMPERATURE
                          -UO'C-.U-320-C	,14	660'C 	
                       I 40


                                   10        5o
                               . ELUTICH TIMS-MINUTES
                                   -COLUMN  F.LUANT-
                         •BENZENEJ-25% ACETONE-f-10% ACETt-NE-
                                                            -ACETONE —
                     H 40
                     < 0
                                     Figure 11
                              PYROLYSIS ' OVEN TEMPERATURE
                                      '..-—>n	eeo*c—
                                             CHROMATO-RAM OF EXTRACT
                                             OF 10 GRAMS OF SOIL
                                             FORTIFIED WITH DASANIT
                                             AND ITS METABOLITES
I          — ^vw
-BENZENE-J--2.3V. Af.
                                  19        29
                               ELUTION TIME- MINUTES
                                   COLUMN F.LUANT	

combine I and n to create a semi-automatic residue analysis system and investi-

gate and define its optimum operating conditions.  Testing will be confined to the

following pesticides: aldicarb, phorate, captan,  demeton, and endosulfan.  All

significant metabolites (the oxygen-analogs,  sulfoxides, sulfones,  etc.) of these

compounds will be included.  Although not on the substitute chemicals list, chlor-

pyrifos (Dursban) will be included as a convenient reference compound since it

                                 Figure 12
                           3°r  FP-LC ANALYSIS
                               OF OXAMYL IN
                           eof-  ALFALFA
« o

I m •BL'NZENE u in ACETONE- /I EXTRACT OF CHECK ALFALFA / V 0.5ml O lOgromi -yJ x lr|ocr 40- — BENZENE —»{*- ACETONE- EXTRACT OF TREATED ALFALFA O.S ml « • 10 pr am» ; •BENZENE—*}*-ACETONE-*- -COLUMN ELUANT — ' contains chlorine, nitrogen, phosphorus, and sulfur. The substrates that will be used for testing cleanup efficiency are animal fat, milk, soil (Beltsville sandy loam), alfalfa, and one vegetable or fruit selected from the group of tomatoes, carrots, and apples. All work will be carried out using fortified samples except for some check analyses of field-incurred residues of phorate in soil and alfalfa which is undergoing testing this year at Beltsville. While waiting for the necessary apparatus to arrive, we have continued to modify and improve the Pye-Unicam flame photometric system by substituting stainless steel pyrolysis oven tubing in place of the silica glass tubing. This eliminates a fragile (and frequently broken) part of the apparatus and also permits improved heat control over the line leading from the pyrolysis tube to the detector. 136

 The Autoprep 1001 cleanup system has just been obtained and we are now in the
 process of connecting up the LCFP detector system to it.  Meanwhile, in a
 separate low pressure LC system, we are using phorate and its metabolites to
 optimize the columns and concentrations of hexane/acetone to use in the step
 gradient elution of SI-200A gel.
 1.  Frehse, H., andBracknell, U.K. Organophosphorus compounds.  Report
    to the IUPAC Residue Analysis Commission,  October 1974.
 2.  Hill, K.R., and Jones, W.M. Adaptation of the flame photometric detector
    to liquid chromatography of pesticides.  J. Chromatogr. Scl. In press.
 3.  Getz, M. E. A recycleabie cleanup column utilizing large pore silica gel.
    Talanta.  In press.
  DR. MOSEMAN: Ken, have you done anything or have you noted any effects of
water on the silica gel?
  MR. HILL:  No, we have not.
  DR. MOYE:  Would you comment on the SI-200 gel.  Is that apolyacrylamlde?
  MR. HILL:  I don't believe it is. I'm not too sure of the exact structure of it.
It's probably a silica gel.  It's one of a series of three gels they sell for this work.
Three different exclusion ranges.
  DR. MOYE:  Are you limited on your solvents in terms of molecular volumes?
  MR. HILL:  Yes.  That's why you don't see acetonitrile used In our work.
  QUESTION:  I was Interested In your comment to the effect that you really don't
care how much metal you have or what types of metal you have In the pyrolysls unit,
but you're trying to combust It  anyway.  This reminds me of some work I did years
ago where I was trying to pipe parathion through a length of stainless steel quarter-
inch tubing, and it was quite surprising to me how much of that material is simply

lost, irreversibly lost on surfaces.  Whether there is a decomposition to inorganic
phosphates or what, I don't know, but I'd be interested to hear what results you get.
   MR. HILL: I didn't really have time this morning to go into some of the perfor-
mance peculiarities of this system, but one thing with the old silica tubing system
(the reason we said we were skeptical was to whether pyrolysis was the mechanism)
was that if you start out at a low oven temperature, say 300 degrees,  and you in-
ject equal quantities of something like parathion and raise the oven temperature
stepwise at 50 degree intervals,  the signal starts to go up becuase the wire is going
very fast, 30 centimeters a second. The signal starts to climb, and then you would
expect, if pyrolysis was the mechanism, that when it reached a total pyrolysis
condition that it would simply plateau out.  In actual fact it starts down again.
There's an optimum temperature, apparently.  We assumed this meant that we had
actually been increasing the rate of molecular  vaporization and that when we reached
a pyrolysis condition a new species was being formed which the detector couldn't
   QUESTION:  What sort of band spreading or carry-over or memory effects might
you anticipate in the pyrolizer?
   MR. HILL: In the pyrolizer,  we're going to have to be very careful with that,  we
think.  I have a young,  new technician working on it right now.  We hope we can keep
him on for the entire contract.  He's strictly here for the summer.  He has been
looking into this and apparently is finding some hangup in this steel tube system.  He
finds if he keeps shooting in stuff fast enough he can plateau out very  quickly.
   QUESTION:  I would think that if you go to a system,  eventually, in some way,
shape, or form, a moving wire type of transfer system (transport system) would
severely limit you in terms of using buffered systems possibly.  I'm  just hypoth-
   MR. HILL: With the Pye-Unicam system as it's normally sold and used with a
flame detector, for example, the signal is very noisy due to hydrocarbon impurities

on the wire.  The man who originally helped develop this with me, Bill Jones, un-
fortunately died of a heart attack and he had developed a skill so he could use one
spool of wire with maybe ten passes — you know, wind it all the way up and reverse
it back and forth. We put a new man on it in the first week,  and I think he was
breaking wire about once a day.
   DR. R.  T. ROSS: In replacement of the conventional flame for the old FPD
system, with your pyrolysis here, does this eliminate your problems with fogging
of the quartz windows?  Are you using the same melpar photomultiplier tubes?
   MR. HILL: Yes.
   DR. ROSS: Are these mounted similarly?  I'm talking about the one that you see
in most of the Government laboratories, mounted on a Tracer instrument.
   MR.  HILL: That's a standard Tracer flame photomultiplier burner housing and
it is modified.  I didn't want to go into the details, but the publication on this will
indicate the modification consists of changing the carrier inlet from the bottom to the
side. You drill  a hole, tap it, and plug up the bottom.
   DR. ROSS: Is it necessary to use a venting system  or anything of this nature?
   MR.  HILL: No, essentially the solvent is being  removed from the wire from the
little solvent removing oven which is there, nitrogen flow plus heat. Then it goes'
into the pyrolysis oven at which point you should have mostly pesticide plus any
other crud coming along.
   DR.  ROSS: Do you project increasing the number of specific items that you can
detect,  for example use of more than just your photomultiplier tube and so forth,
trying to look at other elements?
   MR.  HILL: No, not in the immediate project.
   DR. ROSS: Is this projected?

   MR. HILL: We have not tried other applications, such as hanging a Hall detector
in there, of course.  Anything you can wedge in should be useful.  Obviously at some
time I'd like to see somebody try that with our or somebody else's system using
nitrogen-selective detectors.  I must point out that we're not trying to make any
claim at all that this is a commercially viable  system.  As we have it now it's spread
all over a bench; it looks like a mass spectrometer.  It takes about as long to learn
to operate.  The Pye System was selected simply because it was available and worked
reasonably well.
   The other systems that were available, such as the Nuclear Chicago one, were
being phased off the market because of lack of sales. We bought one of their
continuous chain systems.  There is no reason it shouldn't work.  We hoped there
would be a commercial development of something similar to this.  That's the im-
portant thing.  The sensitivity is essentially that of the photometric detector, and
there you get a better split ratio than you achieve in normal analysis essentially.
We have not achieved the sensitivity predicted and we're not at all sure why.  We're
not seeing as much in our flame photometric as we know we should be able to.

                      MULTIRESIDUE METHODOLOGY
                            Dr. Robert Moseman

  The contract with the University of Florida deals with extension of multtresidue
methodology.  The principal investigator is Dr. H, Anson Moye. As Dr. Moye Indi-
cated yesterday, out of the list of approximately 35 compounds we were working
with,  as would be expected, we encountered some problems with a few of them.
What we would like to do this year is work on solving those problems we encountered
last year, and in addition to that, add a few compounds to the list, again working
with soil. We had some problems with the extraction of certain pesticides from
soil,  and we will attempt to overcome those problems.
  The cleanup system that was found last year to be quite effective was a silica
gel column,  and this year we will continue to pursue that avenue. We had problems
with the derivatization.  Last year we had limited  ourselves to two reagents,  the
dichlorobenzene sulfonyl chloride and the pentafluoropropionic anhydride.   In this
year's effort we intend to look at some other derivatizing reagents.
  Quantitative aspects of the efforts of this year will be limited to gas chromato-
graphy on standard GC columns using electron capture, flame photometric, and the
Hall detector.
   Some of the compounds that presented problems as far as the GC was concerned
last year are amitrole,  benomyl, monochrotophos or Azodrin, aldicarb or Temik,
and diallate also known as Avadex.  These compounds could not be handled intact
through the  GC, as would be expected if you look at the structures. We were also
unsuccessful in obtaining any kind of a reasonable derivative.
 * Office of Research and Development, Pesticides and Toxic Substances Effects
  Laboratory,  Research Triangle Park, North Carolina

   These are a few of the derivatizing reagents that we would like to take a look at:
the N-methyl-bis-(heptafluorobutyrl) amide will react with OH, NH ,  and NH func-
tional groups; the nine anthracene methanol will react with the carboxylic acid.  If
this material is refluxed with thionyl chloride, it can be converted to  nine anthracene
methanol chloride, which will react with OH,  NH2, or NH.  Pentafluorobenzoyl
chloride will react with OH, NHg, and NH.
   The pentafluorobenzoyl hydroxylamine hydrochloride will react with carbonyl .
Acidic acid -N- hydroxysuccinamide ester should react with NH  and the NH.
   In addition to these derivatizing reagents, if we find something during the contract
period or if anyone here has some suggestions,  we'd like to try those also.
   The compounds that we had problems with in the extraction from soil are captan,
folpet, Difolatan, Zectran,  carbofuran,  and monuron.  There were three different
types of soil that were fortified at four different levels.  These compounds gave us
either no recovery or low recovery, and Dr. Moye,  I think,  indicated yesterday that
possibly one other compound, IPC, also gave us some problems.
   Captan,  folpet, and Difolatan presented problems with recovery in that they were
decomposing in the boiling methanol -benzene that was used for Soxhlet extraction.
For the other compounds recovery was simply lower, and correct me if I'm wrong,
Anson, but you attempted to improve the recovery of captan, folpet, and Difolatan
by simply tumbling for extraction, but there again we had low recovery.
   DR. MOYE:  No,  we actually did quite well with tumbling in benzene for just these
three. The tumbling was not nearly as satisfactory with all of the other classes of
compounds.  But for those three we were able to get 90 to 100 percent recovery.
   DR. MOSEMAN:  In our effort to try to improve the extraction efficiency, what
we would like to do this year is use the polytron homogenizer to extract soils.  This
device combines ultrasonic energy with mechanical shearing. It's a high-speed
blending action and has been reported in the literature to be very good for recovery

of chlorinated hydrocarbon pesticides.  That is about the extent of the work that
has been done there. We would like to use this apparatus to improve our recoveries
for some of these compounds.
  There are  four organophosphates that we would like to add to the list:  Dlazinon
is pretty widely used; chlorpyrifos or Dursban; dlmethoate; and a compound of
similar structure, phosphamidon.
  Some additional compounds are methomyl or Lannate, which Is going to be a diffi-
cult compound; and Landrin with its two isomers, the 3,4,5- and the 2,3,5-lsomers;
dinoseb or DNBP; and dlcamba, the last two being herbicides.
  These compounds will be examined in the same manner used for the other com-
pounds. They will be fortified in soil,  extracted, cleaned up on silica gel, derivatized
if necessary, and carried through the standard GC quantitation.
   To summarize what we have in mind: first of all, we will try extraction of soil
with a polytron.   Results obtained will be compared to those we-got last year by
either Soxhlet extraction or tumbling with benzene-methanol.  We propose to fortify
soils at two spiking levels, 0.01 and 0.1 parts per million.  Extractions will be done
in triplicate, using, to begin with, benzene-methanol since we have a handle on  that,
but we can also look at other solvents for extraction.
   For cleanup and separation, the silica gel column system worked well last year
with the five  fractions.  Derlvatizatlon is going to be required.  As I mentioned
earlier, we're going to look at some other derivatizing reagents to see if we can
find a more universal derivatizing reagent to handle more compounds. Finally,
quantitation by standard gas chromatographic columns and electron capture, flame
photometric, and the Hall detector.

   DR. ROSS:  On your universal derlvatization agent,  do you have any speculations
 there on what you're going to use?  I mean, do you have some idea?  Certainly
 there should be suggestions.
   DR. MOSEMAN:  We want to look at those reagents that I showed, to screen
   DR. ROSS:  The acid chlorides.
   DR. MOSEMAN:  Yes.  And I think Dr. Moye has looked into that a little more
 closely.  Possibly the nine anthracene methanol reagent, which would cover a
 wider range of functional groups, may be an approach that we could use to give us
 broader coverage.
   DR. MOYE: Really, what we hope to accomplish here, I think, is a more potent
 acylating reagent that will perform with the carbamate-like materials such as aldi-
 carb and methomyl.  We were not able to derivatize those two materials last year
 with the anhydride.
   This N-methyl-bis-(heptafluorobutyrl) amide is a very potent acylating reagent.
 And this material has been shown to acylate compounds that have not been satis-
 factorily acylated with  the anhydrides. So that,  I would think, would be the prime
 prospect for those two  compounds, along with all of the N-methyl carbamates.
  DR. ROSS: In other  words,  by referring to a  universal derlvatization agent,
you're not talking about one agent that would derivatize all compounds.  It's kind
of a little bit overoptimistic.
  DR. MOYE: It is somewhat a little bit overoptimistic. Now, this nine chioro-
methyl anthracene, as Bob mentioned, is an acylating  agent, if you will, that will
react with many functional groups.  And possibly, this reagent can  be as effective
for things like aldicarb and methomyl  as well as for some of the materials that
have other functional groups.

  DR. MOYE: Yes.  I think we're going to concentrate on the difficulties with
the carbamates and the carbamyl oxime, that Is, methomyl and then the aldicarb
— also Zectran has been a problem — and see if we caa't solve those three with
the acylating agent. That will be a big step forward, I think.
  QUESTION: I also gather you're switching from Florlsil to silica gel. This is
somewhat different from yesterday's.
  DR. MOYE: No.
                                              V** •_.-•••••—•.
                                 • .':. •*,","  *     "
  DR. MOYE: No. We used silica gel this past year.
  DR. MOSEMAN: We can handle a lot more compounds with silica gel than we can
with the Florisll.  And the separation into  five fractions is quite helpful. It aids
in identification of the compounds.  You know which fraction It's going to come  In.
As far as the derivatization reagents, what we may have to do — I'm not saying
we're going to work on this, but I don't think there's one reagent around that Is
going to derivatize everything that  we're after,  but it would be nice — would be to
split the  cleaned up extract Into subportions  and use different derlvatlzing reagents,
depending on what compounds you're looking for.  That's  a possibility.  It becomes
more and more Involved as you try to handle more and more compounds with a
single system.
   DR. ROSS:  Bob, don't you think this would gradually progress In developing
these more sophisticated systems  such as Dr.  Hill has discussed,  and this Is what
we're going to eventually get away from, this derivatization?
   DR. MOSEMAN: Some of it, yes.  I think what Dr. Hill is doing is going to com-
plement what we're going to do very nicely.  There are certainly going to be com-
pounds we just cannot handle with standard,  or  what we refer to as standard, ex-
traction,  cleanup, derivatives, standard GC type approaches.  We're going to have
to go to something else. There's a limit to  that. And the LC system is a way to
go. The high-pressure LC system is the  way to go, a good way to go.
   DR. MOYE: We haven't abandoned derivatives.  We just have a lot of projects.

                        PESTICIDES IN AMBIENT AIR
                              Dr. Robert Lewis*

   Like most of the other programs you've heard about this morning, our programs
 which are directed toward the development of methodology for determination of
 pesticides in the air are just getting underway.  The effort is going to be largely
 extramural; however, we do have a relatively small in-house involvement.
   We began our extramural program last month by awarding a  contract to Southwest
 Research Institute in San Antonio,  Texas for evaluation of various collection media
 for low levels of pesticides  in air. Dr. Donald Johnson and John Rhodes at South-
 west will be responsible for this project.
   I think we need a little background at this point.  Several years ago one of the
 predecessor segments of EPA, the National Air Pollution Control Administration,
 had developed a high-volume air sampler for airborne pesticides under contract
 with Syracuse University Research Corporation.  Dr. Gunter Zweig was intimately
 involved with that project.
   Basically they modified the standard Hi-Vol air sampler, which has seen very wide use
for years by EPA for collection of particulate matter from air.  These  samplers are
used today in great numbers. If you look at Figure 1, you'd say it's a Hi-Vol sampler
because it is a shelter of a Hi-Vol sampler.  Inside you can see the differences.  The
modification essentially was the addition of the pesticide collection module at the top
of the sampler. In Figure 2, you can see that it consists  simply of a standard piece
of glass plumbing, 4 x 2-inch reducer,  affixed to the top of which is a glass fiber
filter mat, same type used in the Hi-Vol.  The filter is capable of collecting partic-
ulate matter down to at least three-tenths of a micrometer in diameter.
   This is followed by a bed of glass beads, which are coated with a liquid collec-
tion medium for entrapment of pesticide vapors. The sampler was designed to
*Health Effects Research Laboratory, Research Triangle Park, North Carolina

Figure 1.  Complete Airborne Pesticide Sampler

                                Figure 2.  Pesticide Collection Module Compartment with
                                                   Module Installed

 operate at 280 liters per minute (10 CFM) and the contractor found that cottonseed
 oil was the most efficient collection medium for the sampler operating under these
 conditions. The collection efficiencies which they determined for cottonseed oil
 were indeed quite impressive„

   There were, however, a number of drawbacks to the use of cottonseed oil as the
 collection medium for pesticides,  not the least of which is the rather involved isola-
 tion in cleanup methodology necessary before pesticides can be finally determined
 by GLC.  This is essentially the Mills-Olney-Gaither procedure for pesticides in
 fat.  Another problem is the fact that cottonseed oil is an  unsaturated material. It
 is readily oxidized, the extent of oxidation being very dependent on the condition of
 the atmosphere in which the sampling is taking place.  But it is frequently oxidized
 and sometimes polymerized during the sampling process, which further complicates
 the extraction of the pesticides from it.
   The sampler itself is a very good one for  ambient air monitoring.  It is simple in
 design,  very easy to operate,  has  very few maintenance problems, and is very inex-
 pensive. And I think present day costs for constructing one of these is only around
 $450.00.  With all the parts we have in the Agency for Hi-Vols,  it is probably less
than that.
  We thought it was well worthwhile,  therefore, to look for substitute collection
media for cottonseed oil for use in the SURC sampler; hence, the contract to South-
  The objectives of the contract are:
  1.  To evaluate other liquids as  substitutes for cottonseed oil in the SURC samplers.
  2.  To evaluate support-bonded sorbents as collection media for airborne pesticides.
  3.  To modify or design collection devices for most efficient use of collection media.

  We are particularly Interested in gas chromatographic liquid phases.  Since the
sampler only uses three milliliters of liquid, this is quite feasible from an economic
standpoint.  We have some preliminary data generated in-house that suggest certain
liquids like diethylene glycol succinate might have a lot of promise.  Silicons oils
will be evaluated and, of course, we're going to examine a number of others.
  The second objective  is to look at solid sorbents as substitutes for the cottonseed
oil, such things as support-bonded LC packings, Tenax, macroreticular solids,
such as Chromasorb-101 and -102, and polyurethane foam, which Bittleman and
Olney have found to be quite efficient for collection of PCBs and chlorinated insecti-
  The final objective is to modify or redesign the sampler as required to accommo-
date the new collection media.  We do  not anticipate that much modification will be
required.  The sampler seems to be well adapted to using a variety of types of
collection media.  We know it operates well with polyurethane foam, for example.
  The properties of liquid collection media that will be required are:
  1.  Proper viscosity to coat glass beads
  2.  Sufficiently low volatility
  3.  Melting point above 20  F
  4.  Chemical  stability
  5.  High absorptivity for wide ranges of organic compounds
  6.  Chemical  inertness
  7.  Acceptable recoveries of pesticides
  8.  Amenability to routine pesticide separation and cleanup, including new
These will be used as the criteria for the selection of candidate media.  The first
three criteria apply exclusively to liquids, the last five apply to solids as well as

liquids. We are not looking for new analytical methodology development, but for
procedures that are amenable to current analytical methodologies.
  Ideally we would like to minimize the number of steps between the extraction of
pesticide from the collection media and Its determination by gas chromatography.
  The-following is the list of study pesticides which Southwest has proposed to work
with. I think it Is an excellent list for the study.
        Chlorinated               Phosphates               Carbamates
        Alpha-BHC                Phosdrin                 Carbaryl
        Llndane                  Dlazinon                 Carbofuran
        Aldrin                    Methyl parathion
  We did not propose the study pesticides.  Not all of them are  substitute chemicals.
But all three important classes of pesticides are covered, along with a wide range
of volatility and polarity within the groups.  Methods will have to be developed for
generation of test atmospheres of these pesticides for Introduction Into the sampler
to determine collection efficiencies. The method that Southwest has proposed will
work quite well.  It's somewhat novel and I will not  elaborate on it.
  The sampler will also be field tested for a period of at least one month.  Of 13
respondents to our RFP,  Southwest was unique in proposing the generation of test
atmospheres of pesticides in the field during field testing, as well as testing the
sampler in the ambient air pool.  I think that sums up pretty well what our extra-
mural program is going to involve.
  I want to mention briefly one of two in-house projects that we have,  that is,
evaluation of the sampler shown In Figure 3,  which  is manufactured by the  Environ-
mental Research  Corporation, St.  Paul, Minnesota. It is a small, portable air
sampler, designed primarily for collection of organophosphorus compounds and

for short sampling periods. We view it as a "source-type" sampler, perhaps of
some value in sampling air In  recently treated fields for the purpose of reentry

                                 Figure 3
   The current price for the unit is $1500.00, which appears to be inflated.  If the
 lid,  which Is also the rain cover, is removed (Figure 4) with care not to break the
 plastic hinges, you can see  that It has two assemblies for holding filters.  Air is
 drawn through both filters simultaneously,   The filters consist of Poropack R sand-
 wiched between two surfaces of glass fiber filter.  They also are quite expensive,
 costing $20,00 a  piece. You are looking at $1,000.00 worth in Figure 4. The sam-
 pler uses two at one time and they are not reusable.
                             Figure 4
                                       -.  • •
   This particular model has two vacuum-cleaner motors (Figure 5); one for AC,
one for DC current.  Of course,  only one Is  operated at a time.  Air Is pulled
through both filters by one motor.  To convert from AC to DC, you simply rotate
the top portion 180 degrees so that the tubes which lead to the filter housing fall
into the proper sockets.  The sampler can operate at rates up to about 170 liters

per minute.  It is said by the manufacturer to be efficient for the collection of
malathion for periods up to about 3 hours.
                            Figure 5
  We are currently evaluating the sampler.  Work started yesterday, after rains
delayed the project in experimental tobacco plots which have been treated with
methyl parathion, both in the microencapsulated and emulsifiable concentrate forms,.
The work is being conducted as part of a cooperative effort with investigators from
North Carolina State University who are going to be looking at dislodgeable residues
at the same time.
  At the same time,  we are comparing the sampler with the SURC sampler and
with the MRI or MISCO sampler (Figure 6)0  There may be some people here who
are not familiar with the MRI sampler.  This is a sampler that was developed some
years ago by Miles, who was with the  Public Health Service,  was further developed
by Midwest  Research Institute, and is now available from Microchemical Specialties
Company (MISCO)0  It was used by EPA's Division of Pesticide Community Studies
for 3 years in  a rather extensive network throughout the United States to collect
pesticides from the air0 Dr. Anne Yobs was primarily responsible for the project.
The MRI sampler has, therefore, seen more use than any other sampler for sampling
airborne pesticides,  and is considered to be more or less the industry standard.,  So
we will be comparing all of our results to those obtained with the MRI sampler.

                                 Figure 6
   The MRI sampler uses impingers charged with ethylene glycol.  The clockwork
 shown on the Inside (Figure 7) is pretty well inaccessible while the sampler is oper-
 ating.  Some of the features relate to its unsuitability, in my opinion, as a sampler
for ambient air.  It is alow-volume air sampler, with a maximum sampling rate of
 about 20  liters per minute.  It is also expensive.  I imagine that the price is nearly
 $2,000,00 now.  This one cost over $1,600.00 about 3 to 4 years ago.  It is compli-
 cated to operate, bulky and heavy.  The impingers are fragile and  require a certain
 amount of skill to load and quantitatively transfer the collection medium to other
 vessels prior to analysis.  We're using the MRI sampler only because most of the
air data that are available were collected with this sampler,

                                     " ~Q ~~H
               T~^*^yig<^?^.p??y?^gy'^'«^^t'«p>ffy.. >\*HM #*„•**•, —T-ii|-irjuia.M.ii .
                                ••.-••" -' *wSiv. >v../        .' ~ ^*7!^*t^.yy7j

                        ^       :^          S       ^.
                • i  '• '•>  V' '      -v~ , »
                   .; ,;--;v
   DR. ZWEIG: Bob,  I think it is rather important to also determine in what state
the pesticides  exist in ambient air.  I remember when I was connected with the
project at Syracuse University several years ago, we were trying to distinguish
between pesticides and the  vapor phase as an aerosol, and absorbed on participate
matter,  I would think that  in your present studies these two questions could be

  If you  extracted the  filter, which we never did, this would give you an idea of
pesticides on participate matter that are  caught in the filter.  As far as vapors and
aerosols, this  I think  is indistinguishable from the liquid or even solid mediaa  I

think strictly from a basic point of view and also from a toxicological point of view,
it Is extremely important to find out in what shape and form pesticides exist in
ambient air.
  DR.  LEWIS: I agree with you 100 percent on that point.  But it is not so simple as
you describe. We have looked at what is on the particulate matter and we find very
little there. • When we have taken these filters which already had particulate matter
collected on them, and spiked them with a wide variety of pesticides (at least 20),
then drawn air through the filters for another 24  hours, we find that all of the pesti-
cides which we put on were stripped off (even DDT, to the extent of 95 percent).  So
a sampler designed to distinguish between pesticides in the particulate-absorbed state
and the vapor state is going to have to have a different design than this one. You have
to get the particulate matter, once it is collected,  out of the air stream and that is
going to involve a fairly complicated sampler.  We have,  in fact, in our objective
statements for this fiscal year, proposed a contract to answer this question; whether
or not  we will get funds for it is very  doubtful right now,
  DR.  ROSS: One comment, and a question.  First of all the comment. I was talking
with Midwest Research Institute not long ago, and they are still using a modified
version of the MRI collector which is  extremely  compact.  Now how expensive it is
I don't know, but somewhere in my files I can look up the person to get in contact
with.   I think it might be worth consulting these people because they are still using
this for measuring OP compounds around fill areas and so forth.
  DR. LEWIS: But as I say, we're looking primarily in this contract for ambient
air monitoring involving low levels of pesticides.  The smaller you go in impinger
systems, the more you're limited on'the amount of air you can sample in a given
period of time.
  DR. ROSS: Dr.  Zweig, do you know anything  about this system that Midwest
Research Institute is  using?  I forget the person's name momentarily.

   DR. ZWEIG: No,  all the people who were connected with the original project about
 six years ago have left.
   DR. ROSS: Well,  this person has modified it.  As a matter of fact, he had his
 first two systems working when I was there. It is very,  very small and I've forgot-
 ten the flow rate.  You see,  it was very high and  very effective.
   DR. LEWIS: It's too small.
   DR. ROSS: That's not what he was explaining to me.  I can give you specific—
   DR. LEWIS: You  can't get the flow rate needed for ambient air monitoring through
 such a small system.
   DR. ROSS: Then my question was back to using solids, you mentioned ethylene
   DR. LEWIS:  No, diethylefie glycol succlnate.
   DR. ROSS: Are you going to have to go through the long quality control extractions
 and everything as was done some years ago? Do  you have to  do this?
   DR. LEWIS:  No, we found that using unstabilized DEGS, we could extract pesti-
 cide from DEGS and  inject it directly into GC with very little  or no background
 problems.  However, you probably would have to  use at least a minimal cleanup
 because you are going to be collecting a lot of other compounds from ambient air
 in addition to pesticides.
  DR.  ROSS: So this will not have to be incorporated into major quality control
 programs to accompany this  particular study,  right?   There won't have to be a
big quality control program going on simultaneously?
  DR.  LEWIS:  I should think not.  If we  instituted an air sampling network,  then
 a quality control program would be necessary, however.
                                      •               '      •
   DR. ROSS: You mean you're not going to do that?
   DR. LEWIS: That's down the road a bit.

                         Dr. William T. Colwell*

  The goal of this new program is the development of a simple, rapid, on-site
method for the determination of residual organophosphate pesticides.  Most speci-
fically and practically the technique is designed to provide a field worker-usable
process for determining safe reentry levels of organophosphates.
  Figure 1 shows the three pesticides, parathion, methyl parathion, and Guthion,
that will be test compounds for the feasibility study.  As will be shown later, the
envisioned process depends on nucleophilic attack of the test reagent on phosphorus.
Base catalyzed hydrolysis of Guthion has been reported to occur primarily on the
methylene carbon.  Thus, it is possible that organophosphate pesticides of this class
will prove refractive.  If a usable portion of nucleophilic attack occurs on phospho-
rus or if the methylene moiety is sufficiently reactive for the ensuing  alpha-
elimination'reaction, then the  process could prove general.

                            Figure 1
                                    N       OCH3
 *Stanford Research Institute, Menlo Park,  California


   A general schematic of the first process to be investigated is shown in Figure 2.
 The dectector reagent, or a reactive salt, is allowed to react with the pesticide to
 produce an odoriferous compound, in this  case methyl isonltrile,  and detection is
 by smell.  Several points should be noted that are relevant to the rationale of the
   1.  A sample area of the field crop or a  collection of leaves, etc.,  in a container
 is all that needs be used.  Stated another way, only aliquots of the field are involved
 and the test material can be discarded.
   2.  The only equipment required is the preformed detector reagent and a field
 worker or workers of known sensitivity to the isonitrile.
   3.  Methyl isonitrile is an extremely potent and characteristic odorant. In previous
 work we had shown that the detection and recognition thresholds of the compound are
 about 10    g/liter of air.  Alternatively, fairly sensitive copper-benzidine air sam-
 pling tubes can be used for visual isonitrile detection.
                               1  Figure 2
     Detector         Organic          ,       Volatile         Subjective
      Reagent     +  Phosphate   	~	*   Compound 	>  Sensing
    (CH.NHCHO)      Pesticide   reaction      (CH.NC)           (Odor)
       o                                           o
  These  points can be amplified upon.  When finally developed,  the method will be
capable of giving a fairly absolute  measurement of remaining pesticide, as a
stoichiometric relationship exists between reagent and organophosphate.  Only a
small crop sample is used and the test odorant can be totally contained or confined
to a small area.  The test material and reaction product will not contact useful
crop, and additional toxicity testing and registration requirements will be avoided.
  There  are several possible odorants that can be tested, and reagents other than
methyl formamide. Figure 3 shows the details of this system, and it will be the
subject of our initial work.  A methyl formamide salt is generated that presumably
attacks the phosphate phosphorus.  The best leaving group, para-nitrophenolate.


is eliminated, giving a characteristic red-yellow color.  The remaining fragment
undergoes an alpha-elimination reaction, well-known in isocyanide chemistry, and
the odorant, methyl isonitrile,  is released.
                              Figure 3
                    CH3N-CHO  + BASE - »- CH,N-Qf
                      i                        "
              EtO-^-0-C-HCH,    +

                   EtO—P— SH     +   CsH-CH3

  We have demonstrated that the reaction does, In fact,  occur when we use a system
that is not suitable for the  simple field work detection system.'  A fresh solution of
formamide salt was prepared from the reaction of sodium hydride dispersion and
methyl formamide.  A drop of parathion was added to the solution and the character-
istic color of para-nitrophenolate rapidly appeared along with the readily detected
odor of methyl isonitrile.  The same reaction has been accomplished on a  glass
slide lightly  coated with parathion.
  The sodium salt preparation is impractical due to its hydrolytic instability.  An
obvious requirement for field use is the availability of a reasonably harmless, stable
preparation that is sufficiently reactive to generate the odorant at desired pesticide
  Figure 4 shows some other approaches we have taken to this  problem.  Methyl
formamide and tosyl chloride spontaneously generate product.  Reactions  involving

 these components thus serve only as negative blanks.  In work involving the more

 reactive phosphonofluoridates and amidates, we were able to generate methyl
 isonitrile using zinc chloride, bromide, and iodide complexes. Parathion,  however,

 proved insufficiently reactive. We have not yet investigated the reaction in  the

 presence of solvents such as hexamethylphosphoramide.  The ferric and aluminum
 chloride complexes also proved unreactive in this system.

               Figure 4:  Effects of Some Lewis Acid Catalyst
                    on the Generation of Methyl Isonitrile

       (1)  CH.NHCHO  +    TosCl    	>   CH.NC
              o                                   o
       (2)  CH.NHCHO  +    ZnCl_   +   TosCl   	>  CH_NC
              O                  ti                          O
       (3)  CH.NHCHO  +    ZnCl    +   parathion   	>   no detectable reaction
              O                  fi
       (4)  CH0NHCHO  +    FeCl -6H O  +  TosCl  	>   CH-NC
              3                  o    £                         o
       (5)  CH NHCHO  +    FeCl -6H-O  +  parathion   	>  no detectable reactil
              3                  o    /
       (6)  CH0NHCHO  +    A1CL   +  TosCl   	*>   CH.NC
              U                  O                         O
       (7)  CH-NHCHO  +    A1C1Q   +  parathion   	^  no detectable reaction
              o                  o

  Other more stable formamide salts will be investigated, such as the silver and

 calcium salts.  Also, as mentioned previously, solvent effects will be tested.
 Currently we have received some encouragement from preliminary reactions using
 ammonium bases  related to those used in phase-transfer reactions. Reactions of 40
 percent methanol  solutions of Triton B and its methoxide analog with parathion and

 formamide have given the yellow-red phenolate color, but questionable generation of
 isonitrile.  We are currently preparing hydrocarbon solutions of these fat-soluble

 ammonium hydroxide salts for testing.  An enclosed container of hydrocarbon,
 methyl formamlde, and the ammonium hydroxide should be hydrolytically stable and

 suitable for aerosol application.

  Other compounds that will be tested are thioformamide salts. The thio-compounds

 are more nucleophilic and we hope-that stable salts of these compounds may prove

usable if  the oxy-compounds  fail.


   Our la-house experience with isonitriles prejudices us to this choice of odorant;
however,  easily detected isothlocyanates are generated from the reaction of thlo-
hydroxamic acids and phosphonofluoridates.  This reaction will be Investigated for
the case of the phosphate pesticides (Figure 5).

              Figure 5. Generation of Isothiocyanates f rpm Parathion
   A program will be carried on concurrently with the chemical work to determine
the sensitivity of a population of field hands to known concentrations of Isonitrile
and to determine their ability to discriminate this particular odor.

                           Dr. R. Fukuto*
  Enhancement of the activity of Insecticides and other biologically active compounds
by other chemicals is a well-known phenomenon, usually referred to as synergism or
potentiation.  In the case of insecticides, synergists have been used for many years
to increase the effectiveness of these materials in  controlling obnoxious insects.  A
classical case of the exploitation of synergists in promoting insecticidal efficacy is
the use of different methylenedioxyphenyl derivatives, e.g., piperonyl butoxide,
sesamex,  and related compounds, with the expensive, naturally occurring pyrethrins
and some  of their synthetic analogs.
  While synergism of Insecticides by the use of additives Is, at least In principle,
a useful Idea,  the same phenomenon In warm-blooded animals is obviously a matter
of considerable concern.  Numerous examples are known where enhancement of
toxicity has been observed when a combination of chemicals is administered to a
mammal.   This phenomenon has been termed potentiation.  Perhaps the first example
of this is the potentiation of the mammalian toxicity of malathion, a "safe" insecticide,
by EPN, which was observed over 20 years ago  and reported by Frawley in 1957.
Numerous other examples of enhanced toxicity have subsequently been reported, e. g.,
potentiation between Delnav and malathion, trichlorfon and malathion, trlchlorfon and
azinphos-methyl, and EPN and dimethoate.
  Although the examples cited above consider only cases of potentiation between
insecticidal chemicals, minor amounts of contaminants commonly present in technical
insecticides also have been observed to produce enhanced mammalian toxicity of cer-
tain organophosphorus insecticides.  An early example of this was reported by Casida
and Sanderson in 1963 from a study of a technical sample of dimethoate obtained
from Montecatini.   The acute rat oral LD.A of technical dimethoate varied from 215
"•University of California, Riverside.  Presented by Dr. Gunter Zweig

to 350 mg/kg while highly purified material gave values of 550 and 650 mg/kg to the
male and female rat, respectively.  Analysis of technical dimethoate showed that
potentiation was attributable to a petroleum ether Insoluble contaminant which, when
added to purified dimethoate,  caused a two-fold increase in toxicity at a concentration
of 10 percent (w/w).  Analysis of a typical sample of technical dimethoate showed
that it contained approximately 17 percent of the petroleum ether insoluble contaminant.
Based on its  physical and chemical properties, its impurity was believed to be re-
lated structurally to O-demethylated dimethoate. This was supported further by the
finding that treatment of purified dimethoate with a demethylating agent such as
O, O-dimethyl phosphorodlthioic acid gave a petroleum ether insoluble product which
showed potentiating action similar to that of the contaminant  isolated  from technical
dimethoate.  The structure of this contaminant, however, was not established.
   A more recent report by Pellegrini and SantL (1972) describes the  potentiating
activity of a variety of simple trialkyl phosphorothioate and phosphorodithioate esters
on the mammalian toxicity of such safe organophosphorus insecticides as phenthoate
(O, O-dimethyl S-(«-ethoxycarbonylbenzyl)-phosphorodithioate) and malathion. Table
1 presents data relating to the active ingredient content and rat oral toxicity of dif-
ferent technical and purified samples of phenthoate,  malathton, and M-1703.  M-1703
is O, O-dimethyl S-(«-isopropoxycarbonylbenzyl)-phosphorodithioate  and differs from
phenthoate only by the substitution of (an ethyl by an isopropyl moiety.
         Table 1:  Toxicity of Technical and Purified Phenthoate,
                  Malathion, and M-1703 to the Rat
                                       % Active           Rat Oral LD
	Compound	             Ingredient             (mg/kg)
Phenthoate, technical 1                  61.2                77.7
Phenthoate, technical 2                  78.7                118
Phenthoate, technical 3                  90.5                242.5
Phenthoate, purified                     98.5              4728
Malathion, technical                     92.2              1580
Malathion, purified                      98.2              8000
M-1703, technical                       83.2                205
M-1703, purified                        98                 2750

  The data reveal that substantial differences exist in the rat oral LI>50 values

between technical and purified samples.  The largest difference was between phen-

thoate technical sample 1 with an LD   of 77.7 mg/kg and purified phenthoate with
an LD n of 4728 mg/kg, a difference of over 60-fold.  Tests against four different
species of insects showed little, if any, variation between technical and purified


  Analysis by GLC and TLC of phenthoate revealed" the presence of at least six Impurities.

The structures of these impurities are presented in Figure 1.  Analogous impurities

also were present in malathion and M-1703.

                              Figure 1
    3  u   I
 phenthoate oxon

    O  u   I

        S -methyl phenthoate

 phenthoate acid

        II  .

           O, O, S-trimethyl phosphorodithioate

         0,S, S-trimethyl
   TLC and GLC analysis of phenthoate technical sample 3 (90.5 percent active

 ingredient)  gave the percentage of each impurity as indicated in Table 2.  The

 impurity detected in largest amount was O,O, S-trimethyl phosphorodithioate which

 was present at a level of about 2 percent.  Of considerable interest is the moderately

 high mammalian toxicity produced by some of the impurities compared to phenthoate
 itself.  For example, the rat oral LD   's of O,O,S-trimethyl phosphorothioate and
 O,S,S-trimethyl phosphorodithioate were 47 and 96 mg/kg, respectively.  This
 is quite surprising since neither of these compounds is believed to be a strong
 anticholinesterase.  A study of the mode of action of these compounds would be of
 considerable interest since they have, until recently, not been regarded as toxic

          Table 2: Percent Impurities in Technical Phenthoate
                            % Impurity               Rat Oral LD
                                                     of Impurity
  With the exception of phenthoate acid each of the Isolated impurities, when added
to purified phenthoate, caused significant potentiation.  Partial data are presented
in Table 3.  While considerable variation in the potency of potentiating ability is
evident,  there is little doubt potentiation occurs.  The most active potentiator was
O,S,S-trimethyl phosphorodithioate where as little as 0.0025 percent caused almost
a doubling in the toxicity of purified phenthoate to the rat (from 3100 to 1700 mg/kg).
  A better appreciation for the potentiating ability of the different impurities may
be obtained from Table 4 in which values for the amount of each impurity which causes
a two-fold increase in the toxicity of phenthoate are presented.  O, S,S-trimethyl
phosphorodithioate and O,O,S-trimethyl phosphorothioate clearly show the highest
potential activity and are approximately 50- to 100-fold more effective than S-Me
phenthoate or O, O, S-trimethyl phosphorodithioate.
phenthoate oxon
S-Me phenthoate
phenthoate acid
by TLC
by GLC

Table 3:  Potentiating Action of Impurities on Phenthoate Toxicity to the Rat

                            % Compound Added        Rat Oral LDg0
      Compound	     to Phenthoate            (mg/kg)
 phenthoate oxon                   0                       3100 (control)
                                  1                       1300
 S-Me phenthoate                   0.1                     2200
                                  0.5                       730
 (CH,0)0P(S)SCH_                  0.4                     1170
    323                  1.6                       650
 (CH S) P(O)OCH                  0.0025                  1700
                                  0.005.                    900
                                  0.01                      600
                                  0.1                       360
 (CH.O).P(O)SCH.                 0.0025                  1920
    323                 0.01                .    1240
                                  0.1                       620
     Table 4:  Percent Impurity Required to Produce a Two-Fold Increase
              in the Toxicity of Purified Phenthoate to the Rat
       	Impurity	            Percent Added

        (CHgS)2P(0)OCH3                         0.003

        (CHQ0) P(0)SCH,                         0.0055
           <5  &        O

        (CH30)2P(S)SCHg                         0.25

        S-Me phenthoate                           0.2

        phenthoate oxon                           0.75
   These results clearly demonstrate that trace amounts of simple trtmethyl phos-
 phorothioates and dithioates have a profound effect on the mammalian toxicjty of
 phenthoate and related esters.  However, they evidently have little effect on insectl-
 cidal activity.  Of some concern is the observation that some of these impurities are

themselves somewhat toxic to mammals.  This, coupled with their capacity to
potentiate the toxicity of such safe organophosphorus insecticides as phenthoate and
malathion, points out the need to examine other organophosphorus esters for impur-
ities and assess the effect of these impurities on mammalian systems.
   The preceding discussion was concerned primarily with the effect of Impurities
present in technical materials which are probably introduced during the manufac-
turing process.  In addition to this, there is the possibility that additional impurities
may be generated upon storage of technical materials, particularly when  subjected
to elevated temperatures. A good example of this is the decomposition of Diaztnon,
another "safe" insecticide, in the presence of trace amounts of water to monothio-
                             Figure 2
                                        heat or
   This  reaction,  which occurs after prolonged standing at room temperature or
much more rapidly at higher temperature, was responsible for the variation in
Diazinon toxicity   to mammals during the early days of its development as a useful
insecticide.  The problem has been solved by improved formulations.  There are
numerous other examples where an organophosphorus insecticide is converted upon
standing to highly toxic materials, e.g.,  the conversion of methyl demeton to a
sulfonium salt of extremely high mammalian toxicity and the conversion of dimethoate
by transesterification with cellsolves to give toxic products.

   QUESTION: I was wondering if any work has been done in this or other systems on
the pesticide properties of some of these other compounds?
   DR. ZWEIG: Yes, very good question. Now I have the answer to that one.  They
were found not to be insecticldal, and they were not potentlators of the Insecticidal
activity either.  This is very interesting.  They are potentiators of the mammalian
system, although the mechanism of the potentiatlon is really not known.
   QUESTION: To happen in solutions, all of them should be looked at for these
trace impurities.
   DR. ZWEIG: Yes.  Of course there is no contract, but we have a list of pesticides
that we wish the contractor to study, and these compounds are all included.  This is
correct; what Dr. Ross is saying is true.  Several years  ago there was found a
malathion product, a technical product, and it was analyzed and it was found by
analysis to contain TEPP. I'm just wondering if it didn't go through exactly the
same mechanism.
   COMMENT: It would follow the same mechanism.
   DR. ZWEIG: TEPP, to the best of our knowledge,  was not used in the formulating
plant.  It was not cross contamination and it certainly could have fallen right into the
same type of reaction.                                          •      «
   MR.  PROUTY:  My question to you and also perhaps to Dr. Warner is, are any
of these impurities produced by side reactions, are they likely to hang around in the
environment, or can we regard them as easily biodegradable or easily detoxified
and excreted by the animal?
   DR. ZWEIG: My answer would be, but I'd like Dr. Warner's answer also, that
these are quite water-soluble, and I do not think there's a problem. They are very
easily biodegradable and I donft think they will cause any problems In accumulation.
This would be my first guess. Dr. Warner, do you have any Idea?

   DR. WARNER:  I would agree with that, since these are phosphate esters like the
pesticide, they should have a similar type of degradation.
   DR. ZWEIG: I really don't think they're problems.  They are more problems of
the mammalian toxicity of the technical material. Very good point you're making.
Let me put it this way since you're from the Fish and Wildlife Service, is that correct?
I think someone ought to really study the potentiation system in fish.  You may
find that the same potentiation effect of these simple triachial esters on fish, the same
as in the warm-blooded mammalian system. That I don't think has been studied at
all, and I think this is worthwhile studying. But as far as accumulation, that I would
not worry about.  I think they are readily eliminated.

                          Dr. Edward O. Oswald*

  The last subject on our schedule prior to adjourning for lunch is a discussion of
the general program of the chemistry branch at RTF.  I'm going to make this very
brief, but at the same time try to give you some feel for our overall program and how
it fits into this particular Substitute Chemical Program. Within this general overview,
I will not go into great detail and data.  If you'd like some of this,  I can discuss it at
a later date.
   The Pesticides and Toxic Substances Effects Laboratory (PTSEL) chemistry branch
which is now a part of the Health Effects Research Laboratory at the Research Triangle
Park is divided into four sections: the quality assurance, methods development,
chemical characterization, and special projects sections.  During our presentations
you have had a chance to meet Dr.  Bob Lewis and Dr. Bob Moseman who are section
chiefs within the branch.  From the standpoint of the  quality assurance  programs, this
particular responsibility responds  to at least three areas: inter and intra quality
assurance programs as applied to the community studies laboratories of OPP,  to the
regions,  and to other state,  local,  and Federal agencies as a need or assistance  in
this  particular area of quality assurance is applied to monitoring.
  Within the quality assurance program, the inter and intra Q.  C. program has been
brought about by submission of standard reference materials  with a follow-up on
evaluation and analysis of the data generated from SRMs, or in  this case, standard
reference materials of various biological matrices, varying from simple complexity
of compounds and solvents to those including biological matrices such as adipose or
blood.  In addition to the inter and intra Q. C. programs, the quality assurance section
 *Health Effects Research Laboratory,  Research Triangle Park, North Carolina

 evaluates and standardizes methodology as applied to the Agency regional needs.  They
 also maintain a centralized electronics repair and calibration facility which was orig-
 inated for the community studies programs and now has been extended into the regions.
   The quality assurance program also includes the task of maintaining a working re-
 pository, presently of some 500 compounds,  through the assistance of the manufac-
 turers.  We ship analytical standards to any requesting organization at a present rate
 of 600 to 1000 samples per month, with approximately 550 or more laboratories
 throughout the world requesting information.
   One of the efforts within the Q. C.  program which has a direct bearing on this
 program but is not funded or associated immediately with the substitute chemicals
 is a program which we were involved with on evaluation of methodology for pesticides
 in water. This particular effort arose because very little could be said about meth-
 odology as it is stated in the Federal Register. We began this effort to see what could
 be accomplished. As this progressed through the stages of determining whether a
 methodology is capable of multiclass, mult ires idue analysis, we found that the exis-
 ting methodology could not handle a wide span of OP,  carbamates, and other substi-
 tuted pesticidal compounds.
   An effort is now under way within the branch to upgrade methodology  in water.  Of
 the 75 or more compounds, I'll list only a few from the Substitute  Chemical Program.
 These include atrazine, simazine,  captan, methoxychlor,  PCNB,  azinphos-methyl,
diazinon, disulfoton, parathion, paraoxon, ethyl,  and methyl,  including in this case
carbamates such as carbofuran and,carbaryl.  Within this particular process,  which
started out to be an evaluation of methodology, we found that the methodology as it
exists is not sufficient.  Therefore, we have now reversed the process back to develop-
ment of methodology and then reevaluation on the collaborative basis.
   For those of you who are not familiar with the process, the methodology is normally
developed through either contractural or internal processes; then it is put out on

collaborative evaluation to numerous laboratories.  Once something can be said about
this methodology, it is applied as needed to various monitoring programs.
  We can divide other efforts within the branch into several areas, one of which con-
cerns detector systems,  and we've heard some of the extramural efforts on this
earlier.  From the  standpoint of intramural efforts, we have the complementary of
mass spectrometry programs. I1!! touch very briefly only on one of these,  in this
case a comparative effort for organometallics, a class of compounds which is very
difficult to handle by any existing analytical methodology.  From this standpoint, the
idea was first developed  to see if field ionizatlon, or field absorption, coupled with
non-flame atomic absorption methodology,  could give a meaningful result.  This is
applied to such compounds as in this case the arsenlcals, cacodylic acid, eventually
shooting for an in situ analysis.
  From the standpoint of ionized species,  cacodylic acids and derivatives which are
very difficult to Isolate and/or the compounds which are chemically modified during
these processes,  we wanted to study the feasibility of using field absorption which
does not,  in this case, require a volatilization. If the feasibility were possible, this
technique would be  applied to quantltatlon.
  At this time we can say that field absorption and field ionizatlon are applicable to
both organometallics and organic trace analysis.  We have looked further into
feasibility of gas chromatography Interface with field ionization, which Is an extension
of other techniques you've heard about earlier and I won't go Into that.
  Other detector systems within the branch program include those of apply-
ing the Hall detector to nitrogen-containing pesticides,  both the carbamates, nltro-
contalnlng compounds, trazlnes, etc.,  optimizing conditions In numerous biological
substrates.  I think you gather from the presentations that you have heard that we
have research and  extramural efforts In water, soils, and heavily weighted In this
case in health effects-oriented substrates  Including human tissues, fluids,  and

   One of the program areas which involves a large amount of our time is developing
methodology for determination of index of exposure to various types, classes,  and
individual compounds of both pesticidal and nonpesticidal nature.  From the stand-
point of the organophosphate-containing compounds, the methodology which was
earlier reported by Dr. Shaf ik for determining alkyl phosphate metabolites in urine
has been upgraded and has been applied.  Numerous modifications have now been
   One of these, which is going to be reported at the ACS meeting, was developed  in
order to overcome some of the problems associated with isolation of these highly
polar metabolites from aqueous systems.  This encompasses using an ion exchange
res in to first of all extract the metabolites.  Derivatization is then done on the ion
exchanged resin and  the materials are then chromatographed by the normal sequence
of events.
   Other efforts concerning the biodegradable pesticides include those of looking at
the metabolites, either the substituted phenols,  the amines,  or the anilines.   Further-
more we investigated workable derivatization techniques applied to these  polar me-
   In addition, we have another area of research which has been directly associated
with the substitute or alternative chemicals program.  That area  is  evaluation of
chromatographic techniques  Involved in multiresidue procedures,  and you have heard
some of the extramural efforts on this earlier.  This particular effort involves look-
ing at the feasibility  and actual flexibility of chemical bondage GC phases for analysis
of intact pesticidal compounds. This involves preparation of chemically bonded
phases such as those of carbowax 20M, bonded chemically to numerous types  of
phases such as from the chromosorbs P, G and W, etc., and evaluating these chem-
ically bonded phases  for their chromatographic separation capabilities with such
compounds as those of phosphamidon, diazinon, atrazine, simazine,  disulfoton,
Dursban, methyl parathion,  and others, including trifluralin.

   In this particular effort the chemical bondage GC phases have resolving capabilities
much greater than those of the normal pack column characteristics.  Extension of
this, which is planned, is for applying this working GC column system to the other
detectors I've previously mentioned.
   Other areas within the branch programs include those of cleanup methodology, up-
grading, and modification* This is always a problem as you begin to consider the
multicomponent systems and biological substrates.  To us a biological substrate
could be anything from human adipose tissue to urine, feces,  water,  or air.  As we
all know, there are very few methods that apply to all of these. Each one has Its own
specific characteristics.  I have already mentioned the sequence of events that takes
place concerning methods development,  as we see it, within our branch. Normally
the method is developed and then collaboratlvely evaluated prior to use for monitor-
Ing purposes.
   I'd like to give you one last insight into an application which happened to be a com-
pound that by chance was  Included in the Substitute Chemical  Program.  It's a very
real portion of our program, and this is furnishing assistance from time to time to
Federal, state and local concerns in cases of suspected poisoning.  For Instance,  we
had received requests for technical assistance from  a group that had a suspected
'poisoning case, OP in this instance. The patient was overly intoxicated and had con-
sumed a material which was suspected of being an organophosphorus-containlng
pesticide.  In addition,  by unknown means, the formulation was contaminated with
turpentine and a number of other substances.  Through the assistance of a number
of the techniques which you've heard about earlier today,  namely GC, GCMS, cleanup
procedures, the formulation was first concluded to contain one of two compounds,
Azodrln or Bldrln.  In this case chromatographlc separation techniques would not
resolve this, but by mass spectrometry,  it was very easily discernible that this
particular formulation did contain Bldrin.  By translating this one step further to the
analysis of the patient's urine, It was found that this particular patient was excreting

approximately five parts per million of the dimethyl phosphate.  Therefore by cor-
relating this  information in an everyday working fashion, this gives you the type of
problems we are concerned with.  At the same time, there is always a need for in-
creasing the  specificity and sensitivity of methodology as it applies to multicompound,
multisubstrate systems.
   DR. MOYE: You mentioned the work of Dr. Shafik along the lines of index of ex-
posure, and he has developed, I think, some very s ignificant methodology for the
cholinesterase inhibitors, organophosphates.  Does your branch plan to extend this
to other classes of cholinesterase  inhibitors, that is,  to develop methodology for
index of exposure? I'm specifically thinking of the carbamates.
   DR. OSWALD:  Yes. We have within our programs at this time areas  which in-
clude both the intact and the transformation products of the carbamates.
   One of the problems which we have found, as I think you've already experienced,
is stability of compounds, both from a derivatization standpoint and from an ana-
lytical standpoint.
   In the case of the perfluorobutyl or the PNB derivatives, depending upon the com-
pound, we do find problems.  In relationship to the index of exposure as monitored
by urinary metabolites for OP compounds,  one of our biggest problems is not
necessarily due to instrumentation drawbacks.  It is due  mainly to reagents which
have nothing to do with what's in the urine,  but is really what we use to fabricate
the derivatives.
   DR. ROSS: I thought the analysis of pesticides and water was a primary
responsibility of Cincinnati.  What is your interfacing or coordination with  their
work? I'm talking about Lichtenberg's  manual notes  on organophosphates, chlorinated
phenoxides, and carbamates.

  DR. OSWALD:  We found,  like everyone else, that no one can do everything.  In
this particular effort we were requested to at least attempt to say something about
methodology; this  is on the Federal Register at this time.
  The type of involvement, whether it's water, air, soils, or tissues, in relation-
ship to our effort, I would have to put on a substrate basis. Our heaviest involve-
ments are those in health effects and human exposure systems,  and next to that
would be air on down to soils with water last.
  Ours was really a supplemental effort to that of Cincinnati.
  DR. ROSS: In other words, you're overlapping in methodology because it is the
responsibility of Cincinnati and not RTF to develop methodology for pesticides and
  DR. OSWALD:  It was supplemental.
  DR. ROSS: With reference to substitute chemical cleanup  methods, you were
investigating the feasibility of this some time ago.  What progress, if any,  has been
made on that?
  DR. OSWALD:  In relationship to cleanup methodology, one of the areas which I
didn't touch on in-depth — this Includes GPC as applied to multlclass, multi-
performance systems — we  can say at this time with existing methodology  that
modifications are required.
  DR. OSWALD:  Gel permeation chromatography.  You cannot take most of the
reported systems which were Initially developed for adipose  and apply them to
systems with different Interfering problems.  We have heard about some of these
from Dr. Hill and others. We're hoping to resolve some of these at a later date —
at least as It stands right now,  taking methodology for GPC for soils below a tenth of
a part per million. We do have  to come up with additional cleanup methodology.

   DR.  ROSS: Most of your methodology seems to be oriented toward soil chemistry
 rather than adipose tissue.  Is it weighed differently? Are the priorities greater
 now for soil than for tissue?
   DR.  OSWALD:  If you look at our responsibility for methodology,  for monitoring,
 especially from a regional standpoint and as we support the Office of Pesticide
 Programs community studies labs, you'll see that they've been diverse activities.
 In the regional situation, the only common involvement that one region may share
 with another is probably that in water.  We have found, through the clarity of eval-
 uations and the SRMs,  some regions  do not have any involvement in human-associated
 fluids.  One of the first reasons  for being involved with water was that all the  regions
 had a common need for methodology in water.  From the  standpoint of the community
 studies laboratories, their programs have been oriented toward that of human tissues,
 fluids,  air, and water.  The varying  amount of involvement is the reason you see a
 combination of all of these substrates in our involvements.  We do support all of
 these programs.
   DR. WARNER:  A number of years ago, the old chemical corps spent quite  a bit
 of effort on developing  colorimetric systems and field kits for determining the pres-
 ence of organophosphate CW agents.  Have these systems been looked at very
 thoroughly for the pesticide problem?                         .      t
   DR.  ZWEIG:  Yes, they have.  It turns out that many of the methods were based
 on cholinesterase inhibition.  We wanted to get away from stabilized enzymes so we
 were not happy.
   Secondly, most of the apparatuses  developed by the Army at that time were  ex-
 tremely costly,  very cumbersome, and very complicated, and when we checked out
 the sensitivity they were really not suitable for the  sensitivity that we were looking
at.  They weren't looking at tremendously large  contaminated areas.  We were talking
 about nanogram  quantities, and they were talking about milligrams per liter.

  We spent quite a bit of time at Edgewood talking to people like Cramer and Gamson.
None of their methods, with  the exception of the method that Stanford Research picked
up,  which was also developed on contract for the Army, seemed to be suitable.
  We were looking for simpler methods and, as I mentioned before, this effort on
our part was  for a much simpler method.
  There was a method that was developed, I think by Cornell Lab, also on the con-
tract with the chemical  corps, that is a so-called "ticket."  We are still looking at
this, and possibly this ticket might be funded by Research Triangle Park.
  Nobody can tell us what's behind the ticket.  It's a plastic tag and the idea is
similar to a dosimeter that you carry on your lapel. According to the Inventors,
who won't tell us what's inside the ticket, it turns from a green to a red depending
upon how much phosphate is present, but it's very nonspecific.  I think what they're
really depending on is the solvent carrier of the pesticide,  not the pesticide itself.
  When we thought about the system, we liked  the idea of a dosimeter that every
person carries with him, and if he is in danger, it will turn a different color.
   The ticket was Investigated by the Office of Pesticide Programs  and the Operations
Division under a project called "Operations Safeguard" several years ago when they
went to parathion.  It has not been found to be too useful, but it is still being Investi-
   To  answer your question, yes,  we have looked at all the  things that are declassified,
and we decided that with the exception of the one system that Stanford picked up, none
of the others seem to be suitable for our particular purpose.
   DR. LEWIS: I just wanted to point out that there are two aspects to the hazard In-
 volved with reenterlng  freshly sprayed fields.  One is air route  exposure of a person
 to the contaminated air In the area. But, I think that Is secondary to exposure from
 dislodgeable residues.   There has not been a satisfactory method developed for
 determining dislodgeable residues. All you can do is take  foliage samples and wash

 them off with water or various solvents. We are currently only looking at the air
 route of exposure, even though it may be the less important.
   COMMENT: We do have a very small effort that we did not mention here because
 it was not funded by the Substitute Chemical Program. It's exactly what you're just
 stating,  and that's a quickie method for determination of dislodgeable residue on
 fruit and leaves.  The thing looks like a nutcracker.  Basically what you do is put
 into this nutcracker device a leaf and a filter paper strip and as you press the handle
 together, you  slowly rotate the little wheel which wedges the leaf against the paper
 and actually rubs off 100 percent of the dislodgeable residue and everything else.
   It waxes the chlorophyll and the pigment onto the paper.  Then this paper is
 developed through a very simple solvent system and in a tiny test tube — a 10-minute
 development just like a paper chromatogram ~ and the paper has already been pre-
 treated  with DCQ coloring agent  and you  get a  color.  The color that is developed
 is a band; in a way it's semi-quantitative, directly proportional to  the concentration.
  This  is under development  now.  This was a very small contract, but it
 seems to be a very  simple field test  for  farm workers or somebody to actually
 measure the dislodgeable residue.  This has been written up as a  manuscript
 and will be published in the Bulletin of Toxicology  probably  in  the next  couple  of
  QUESTION: How's the sensitivity experiment?
  COMMENT: Within the range that we're looking at, Riverside has a contract to
study the overall concentration of pesticides in orange groves.  Part of it may be a
correlation between the cholinesterase levels and levels in the urine.
  It's a much  more comprehensive program.  This little bit where they developed
this method of dislodgeable residue Is within the actual range of what is found out in
the field after  one day, two days, and so on.   It's quite sensitive and,  as I said, it's
semi-quantitative. There are some problems. The coloring agent's not too stable.

You have to measure your color.  You have to heat the color.  There are certain
things that are not totally worked out, but they have shown the quantitative relation-
ship between the amount of residue and the concentration and the color.


   DR.. OSWALD: During the presentations given yesterday and today we have
heard Discuss ion Qa_vj^rAoujj9Ptgs..^gg^"'dJtogGC'detector systems such as the Hall;
multiresidue methodology; the multiplicity of problems and complexities involved
in identification of impurities; problems associated with reentry; exquisite Instru-
mentation; newly planned efforts, including those of mass  spectrometry and auto-
mated cleanup methodology; continuation of multiresidue methodology; research
related to pesticides In air; means and detections of pesticides in the field, again
related to reentry problems; and potentiation of pesticide  as related to contami-r
nants or potential contaminants.
   We have discussed these areas to attempt to further scientific knowledge and
to answer the questions from the chemical standpoint to what is there, how much
is there, and what does It mean.
   Before we go  Into other topics which the audience may  have Interest In discuss-
ing,  I'd like for you to think back through the presentations and if there are any
comments concerning the plan Of the program, the design of the particular program,
its emphasis, specific information which may be pertinent to various compounds or
classes, now is the time you can discuss these. Feel free _to make any comments""
at this time.
   DR. J. ROSS:  One question I have concerns the list of compounds that are
used for the work in this program.  These compounds are not among the com-
pounds that are being considered for substitute compounds,  and there's a wide-
ranging list.  What is the basis of selection for these ?
   DR.  OSWALD: As was explained in the general session — and I think you'll
probably hear some more about this — there has been a coordinated effort to
determine on a use basis what compounds may serve as alternates or substitutes

for the following persistent chlorinatives: DDT,  aldrin, dieldrin, chlordane,
heptachlor,  substitutes for 2,4,5-T, the ethylene bisthiocarbamates, and certain
fungicides.  The matrix which has been prepared on these is associated with that
of a crop/use orientation.
   DR.  ZWEIG:  There was an article in C and E News — June 9 — which listed
all the compounds that are under consideration right now, which are potential
substitutes for DDT, aldrin, dieldrin, chlordane, heptachlor, 2,4, 5-T,  and EBDC.
This is  pretty much the list that we go  by.  The list will be expanding as we con-
sider,  in all probability,  chlorinated pesticides for Akton,  but you're totally up
to date as of yesterday.   I think you have the total list of pesticides that are now
being studied for litigation, already have been cancelled or suspended.
   As Dick Back pointed out yesterday during the opening session, a compound
like carbaryl was left out as a substitute for DDT.  This is not an all-inclusive
list; this was the best possible list that we drew up.
   Carbaryl —and this was not explained yesterday— was not a substitute chemical
as far as our internal bookkeeping was concerned because we had looked at it from
another point of view in an internal review.
   I should explain to you that the Agency is looking at substitute chemicals,  and
this is one thing, but we have other compounds for internal review as, for example,
chlordane/heptachlor.  These are the compounds that are being thought about for
   Carbaryl is a compound that we had inherited from many years back when there
was some question about its safety,  and it had just gone  into  the mill of the internal
review.  Actually we have looked at it and yet,  in a way, it should have really
been one of the prime candidates in the Substitute Chemical Program.  In most of
the multiresidue methods carbaryl is always chosen, and in the air pollution
studies and in all the other work that was reported here carbaryl  was always shown
as a prime compound,  representing the carbamates.

   I would say that this list is quite good, but it is not totally inclusive.  As we are
looking at other substitutes for other compounds to be considered for litigation,
undoubtedly this list will expand, but there's nothing particularly significant about
the choice of any of these.  These compounds came  out of a computer match-up,
as Dr. Oswald pointed out, between the use pattern — in other words,  the site and
the insect, the corn borer against corn.  If this had a registered use and this would
substitute for aldrin, then it would be considered as a substitute chemical.
   Most of our contractors and cooperators are adhering pretty much to this list,
but if they want to  choose additional compounds that are more  representative of
their chemical methodology, we have no particular  objection to it.   It would expand
the horizon of the methodology. Does that explain it?
   I  do have a question to add. We have discussed very sophisticated methods,
including ESR and  hyperfine labeling, mass spectrometry,  electron impact, and
chemical ionization.  I would like to  ask Dr.  Hill to tell us a little bit about the
simple method which is still very useful — TLC ~ and I know that he has some
efforts in his lab  right now.   Maybe he can tell us a little about that.  That's
still a good method. It's easy, very cheap, very simple,  and very sensitive.
   DR.  HILL: Unfortunately,  the TLC plates did not survive the local humidity
for two days, so I'm not going to show them.  We've been very fortunate in having
Professor Joe Sherma  attached to our lab  this summer as a faculty appointment.
We brought him in on a "do anything you please" basis, as long as it's in TLC.
   After he came onboard, we kicked around a few  ideas, and one thing that's
always been in the back of our minds for years is a quantitative TLC procedure
using optical scanning devices which is pretty well  along for many compounds,
using conventional color development, but we're talking about  samples down in
the  range of micrograms,  maybe one tenth if you're reasonably good,  a half a
 microgram if not.

    We all know that cholinesterase inhibition techniques for TLC are extremely
 sensitive, but the plates are usually messy.  You wind up with a cholinesterase
 inhibition white spot on a blue or some other color background, where you've
 inhibited some reaction.  Joe and Mr.  Getz decided to see  if they couldn't get a
 decent looking, scannable TLC plate,  using the cholinesterase inhibition approach,
 and in the first few weeks,  they met with success beyond their expectations.
    The first approach was very conventional, but they very quickly realized it
 (the white spot on a black or blue colored background) was not terribly useful,
• even if you  managed to clean up the background so it is a uniform color and not
 streaked or mottled or spotted. The white spots are simply areas — the concen-
 tration is a function of the area. There is no optical density to deal with here.
 It's either white or not at all, and so the idea was to get a color reversal plate,
 which they have managed to do now.  We get blue spots on a white background,
 and a very white background at that.
    This  spot is proportional to concentration with the usual plateau in the higher
 concentrations which is a characteristic of  saturation. This was done  by using
 horse serum,  and the acetothiocholine is hydrolyzed by  the cholinesterase enzyme
 to form  thiocholine and acetic acid. The thiocholine has the ability to reduce the
 blue redox dye (2,6-dichloroindophenol) to a colorless or yellow form..
    You dip your plate into the dye  and get a solid blue plate.  In about 30 seconds
 it just simply bleaches out white,  except where the pesticides are, which stays
 blue.  The area of inhibition is proportional to the amount of pesticide present,
 and we're down in the nanogram range.
    I had  hoped to be able to show you some of the actual plates.  As I  said, we
 made some  up Tuesday morning before coming  down here and kept them in the
 refrigerator all day with a cover glass, but  they didn't survive.

   QUESTION:  Did they survive long  enough for scanning?
   DR. HILL:  Yes, in the lab there's no problem.  You can keep them overnight
in a refrigerator and they're quite good.  It's just that, after coming down here
in a hot car, it was too much for them. What actually happened is we got another
reversal  back to white.  These were originally bright blue spots on a white back-
ground, but they faded out. This was  down at 32 nanograms—the one we were
hoping would survive and didn't — you see the dye is back again. This was a
white plate, again with blue spots,  and this is 2, 4,  8, 32 nanograms of phorate.
Then,  just for the fun of it, they put on carbofuran and the 3-hydroxycarbofuran
which also gave very good spots.  They didn't try the carbonate until this plate.
   This plate was one week old before I brought it down.  It had been in the refrig-
erator that long, but it immediately faded out after one day at room temperature.
They are about to wind this up, and I think they're going to get something interest-
ing published on this very shortly.
   QUESTION:  Are these commercially available plates?
   DR. HILL:  Right. They're silica gel plates. "Quantagram" Is the one plate
they're concentrating on for the moment.  They were using Merck plates before
that.   When it comes to  the dip technique, which is critical to this, the stability
of the plate coating is critical.
   The Merck plates were very hard to the touch ~ physically, you could scratch
off the Quantagram plate before the Merck plate.  When you get the thing wet (ap-
parently the binders get involved) in the dip  tank,  the  Merck plates slough off and
the other ones hang in there.
   QUESTION: How about the Eastman chromato-sheets?
   DR. HILL:  They have not tried any of those. They're not stiff enough to be
that reliable for scanning. That's one of the problems.

   QUESTION:  Is it by reflecting ?
   DR. HILL: This is by reflecting.  Yes, I didn't want to go into that because of
my personal involvement.
   QUESTION:  Is it area or optical density that determines the intensity of the
spot or response?
   DR. HILL: It's both.  The response of the instrument is designed and based on
reading both the area  and the optical density simultaneously.  The total signal that
you see — the integrated optical signal — is a result of both of those factors.
Either one alone will not give you a linear curve, but combined they will.  Of
course you have to use area — you cannot use peak height — because  that may
not have any relationship to concentration because of the shape  of the spot or
something.  We're using aKontes scanner which they lent us.
   QUESTION:  Is that a scanner that scans  across or can you position it in one
phase ?
   DR. HILL: Essentially you position your plate, and then the scanning head is
positioned by eye over the row of spots here, and scanned in this  direction. The
plate moves under the head with a cover glass on it, and the scanning  head rests
on the  cover glass,  which is part of the patent,  and you can scan six rows of spots
that way.
   To show you how well this really works,  we've just finished  a two-year study of
carbofuran in soil and alfalfa at Beltsville, with John Neal, an entomologist,  in
which we did all of our analyses by TLC.  This included acid hydrolysis as a first
step to break up the conjugates  and all that.
   QUESTION:  Can you give any  idea of the linearity and sensitivity?
   DR. HILL: The linearity for any TLC plate  is always the same. We will
always see something about like this, and it doesn't really matter much what your
units are  here.  You'll have about a decade  or a decade and a half of linearity

 before it flattens out, and this may very well be nanograms or it may be
 micrograms, depending on what kind of plate you're using.
    This type of thing is limited to about a decade or a decade and a half.  We can
 usually push it to a decade and a half with the cholinesterases — I've seen some
 of the plates that have been running anywhere from two nanograms or less up to
 32  before they begin to flatten out.
    QUESTION: You were not using plant extracts?
    DR. HILL:  These are standards, right.
    QUESTION: Do you have to activate the organophosphates with bromine?
    DR. HILL:  They were trying that.   I don't think they're doing that now.
   QUESTION:   Is the spotting technique for this scanner critical ?
   DR. HILL: Yes, we're using the Kontes spotting device which was developed
in our lab by Mr. Getz and Dr.  Beroza.  If you'll consider for a minute what
you're doing in pesticide residue analysis — you fill up these little glass tubes
with the syringe needles on the end  and then put them in this device,  and  it drains
out by gravity flow onto the TLC plate.  It's made in such a way that we have an
air manifold underneath through which you can bring in compressed dry air or
nitrogen,  and this evaporates your solvent at a fixed rate and controls the spot.
   It allows you to put on two or three milliliters of sample containing your pesti-
cide after you've concentrated it down. You can put two or three milliliters in
these tubes  and put it all on the plate, whereas if you do any other type of spotting,
usually manual, you have to concentrate down to a very small volume,  and then
transfer it with capillary pipettes.
   COMMENT: Two or three milliliters must take a little while.
   DR. HILL: No.  You can just put it on and go have a cup of coffee.  It can be
up to a half an hour or so, but it controls not  only the size and the quantity of the
material  you transfer, but  it also has  other benefits.

   Of course, we're spending a lot of time, you might say,  in an educational effort
for quantitative TLC in which we try to tell everybody who will listen that the
instrumentation is not the limiting factor, and we're not talking about the Kontes
equipment any more than we are anybody else's. The  machinery has been avail-
able to give you plus or minus one percent.  It's just a matter of learning the
operator skill and you would not expect any less of a man who was using a gas
chromatograpa.  It'sjust that TLC grew up in an era of qualitative, sloppy
technique, and it's just an educational thing.
   The quality of  the commercial plates has come along enormously in the last
five to eight years.  Initially, we only used homemade plates; now we trust the
ones that you buy. You still get an edge effect.  The Kontes spotter forces you
to work in toward the center of the plate, and you can't crowd it.  You are limited
because of the optical design to six rows.  Six rows is just  about optimum for one
analysis per plate.  You can have three concentrations of standard, your unknown
sample, and a control extract,.all on one plate for background.
   Incidentally, our motivation behind all of this is not to undermine GLC, but
TLC is still very much the method of choice in many countries.  Some countries
are constantly complaining that the United States is so preoccupied with  their
expensive equipment.  This is really an approach to get at the basic economics
of it, and  even the cost of the instrumentation is kept in mind because there are
many countries that are now getting into pesticide problems, and we're going to
have something.
   DR. ZWEIG:  I would like to raise another topic that has been either forgotten
or conveniently ignored. We've all  become very much interested,  as you will
hear later on this aftenoon in a talk by Dr. Phillips, in insect hormones and
pheromones, and this is a real  problem.  So far we chemists — I still include
myself among the chemists — have  been very fortunate that we're working with
compounds that have a marker — either they  have a nice nitrogen, a sulphur,

phosphorus, a chlorine or something. What do we do with things that just have
CH and O ?  I just would like to bring this up as a discussion point. Have any of
you had experience withjthe analysis of,  let's say,  the commercial insect hormone
(that would be Altosid) ~ that's the only one I can think of at this moment — and
the Thompson-Hay ward  compound, but that's not in the same class,  that's in the
pheromones like Gyplure and the insect hormones of the Altosid type ?  What are
we going to do about these?  These may become  more prevalent in the future,
and what are we going to do as residue chemists ?  How do these compounds fit
into the mult ire sidue scheme?
    DR. HILL:  Yes, I can address two areas of this.  First of all, it's not Gyplure,
but disparlure that  is undergoing extensive field  testing in Pennsylvania and other
places.  In one of the projects in Beltsville, they were concerned with an analyti-
cal method for disparlure to use this fall.
    Jack Plimmer's lab is in charge of this.  Mr. Holden from our lab  is on this
project,  and the only thing he or I were ever able to come up with was a derivatiza-
tion based on the epoxide functional group.  Essentially,  without going through the
whole molecule, they're simply looking at what to do with this.  They are going to
open this up and hang something on there,  something that will be electron-capturing.
The decision of what to do with this has not yet been made, but this is  the approach.
The other compound, which is not exactly in the category of the things you're talk-
ing about, but since you mentioned Altosid, the developmental compound TH-6040,
which is not a synthetic juvenile hormone, but a substituted urea, which,  if you
look  at it, you'd swear was a herbicide, but  it affects cuticle formation in insects,
and it has been considered as a possible larvicide in animal feeding.  We've been
looking into that and the method turns out to be  quite simple.  Liquid  chromato-
graphy with U. V. detector measures the absorption of these  aromatic things even
though fluorine and chlorine are in there; you can get by with virtually no cleanup.
    DR.. ZWEIG:  That's why I don't consider the TH compound the same thing.

   DR.  HILL: Everyone of these — a generalization,  I'm getting in trouble —
but virtually all of these pheromones must have something this way from which
you can get a derivative.  They're not completely inert hydrocarbons.  They will
either have a double bond or something like this.
   COMMENT: We have the same concern in terms of residue and tolerance set-
ting,  and fate and movement in the environment.
   DR.  HILL: Yes,  I realize you're obliged to do that in regulatory agencies.
Our interest in an analytical method for this — and when I say "our" I mean Jack
Plimmers' interest — is for scientific purposes only,  as a screening and survey
device to find out where this compound is.  He's not interested in developing
methods because he thinks there's a residue or a residue hazard, either one.
It's a purely scientific tool.
   QUESTION:  I haven't done  any work on these hormones, but you have similar
problems when you start looking at drugs, like marijuana or tetrahydrocannabinal-
type compounds.  There's no handle, and you have sensitivity and selectivity
problems,  and the  derivative route has been used.  It has problems in that you
derivatize other things in blood and urine, and in that case something like mass
spectrometry is necessary.  You end up using GC mass spec to get whatever
quantitation you can.  It's expensive, but this may be where you'll end up too.
   DR.  MOYE: I hate to predict things, but it seems to me that we may be over-
looking the fantastic  advances  in various types of column chromatography, namely
high-speed liquid chromatography,  in terms of the separately power of the columns
   I suspect that molecules of this  sort, that don't have any handles that we can
get hold of, we  may find by a combination of separatory techniques —  I'm
thinking in terms of a molecular-sized separation of a very narrow band width,
where we can pick out a fraction of molecular size, say from 200 to 210  molecular
weights, and be very sure that we have  that,  and then pump this fraction  into a

second column that will perform a separation in terms of partitioning or
absorption — we might be able to get away with a very nonspecific detector
such as a UV or a flame ionization detector.
   DR.  ZWEIG:  How are you going to distinguish between this and the natural con-
stituent, if you're into natural constitutents? What is your handle? What type of
UV absorbence do you have that's  specific? Frankly, if I may prophesize and
look into the  future,  I would say the only way to go about it is by a specific deriva-
tization, where you put the  handle on. You need a handle.  Otherwise, you just
have a long-chain fatty alcohol, or long-chain fatty aldehyde type compound.
Where do you know  it's going to come off the column with a nonspecific UV de-
   DR.  MOYE: The idea is to have a standard that you are convinced of what it is.
   DR.  ZWEIG: With a tracer?
   DR.  MOYE: Not necessarily with a tracer, but with large amounts of a material.
If you're talking about nanogram quantities, you may be hard put detector-wise,
but if you can get large amounts,  you can work on a macro scale, and work out
the separations.
    Once again, I think we may in  many ways be overlooking the extreme selecti-
vity of  some of these newer liquid chromatographic columns that are coming
along.  By parlaying, if you will, several separatory modes ~ molecular size,
absorption,  reverse phase, or partitioning — the degree of selectivity may be
somewhat surprising in that we can pick out a material such as these compounds
present in fairly small amounts,  along with large quantities  of not too different
 structure.  A difference has to be there, and that's what you're banking on.
    DR. MOSEMAN: To reinforce what  Ansonhad said, we were able to  separate
 acetonitrilic acid and citric acids extracted from water on a surface-bonded gas  •
 chromatographic column. This separation could not be done  on any commercially

prepared,  standard GC column packing, of -which we knew.  I think there's
promise in the resolving power of certain columns, particularly high-speed
liquid chromatography, and it can be done.
   QUESTION: Yes, but as the conference has pointed out, we're hung up on
specific detectors.  At least, that's the take-home lesson I'm going to go away
with. Here we don't have a specific detector. We're going back to the old flame
ionization or we're going back to UV detector — nonspecific detectors.  I haven't
had any experience with these compounds, but I do know that — the man isn't here
from Zoecon — when they came in to show the residue picture of this, they had
to go through all sorts of very difficult separations, and then in the end, they
had to use  a flame ionization detector.  It was very difficult for them to convince
us that the signal that they got was really due to Altosid or one of its metabolites.
   I think it's a real challenge to the residue chemist to develop  separation and
and detection techniques.
   COMMENT: If you don't have a way to get at an element-specific detector or
to use an element-specific detector, I'm simply saying that there is the possibility
of resolving components in a mixture,  and looking at them with a relatively non-
specific detector  system.  Granted  it's not the best way to do it, but if that's the
only thing you've  got —
   QUESTION: What about using the Beroza technique of sticking the insect? You
have a cage of the end of your gas chromatograph and you see how  the insect be-
haves.  As a matter of fact,  it has been used as a chemoreceptor for anti-feeding
compounds.  I'm  just throwing this out. That's where your specificity is,  in their
biological response and not their chemical response.  Maybe we're taking a step
backward when we talk about this, but let's not forget some of these biological

    COMMENT:  That's specific detector.  I'm not at all convinced that specificity
is as great as we wouldjiope either.  I think some of the work of Rolloff with the "_"_
 electroantennagrams show that these moth antenna can respond to a wide variety
 of closely related compounds that are not necessarily pheromones.  You  could be
 slightly fooled by it,  I think. I'm just speculating out loud.
    One other comment which is more political than scientific is that there seems '
'to be a neurotic preoccupation In EPA with sensitivity,  from my point of view,
 rather than selectivity.  In other words, when I promote within my  own Agency,
 the interest in something like this  LC flame ionization, I'm very careful to not
 spell out the possible advantages it might have  in sensitivity, even though it's
    I'm using the argument that the selectivity is improved.  I don't know if pure
 sensitivity by itself is necessarily a highly desirable thing when you're talking
 about something like these compounds,  and that's why we were satisfied to use
 UV detector with an LC for TH 6040 even though we  were looking for it in milk
 among other things.
    Until we get  some feedback that there is a real toxicological concern  about the
 compound,  the sensitivities are of secondary importance, not out of the  question,
 not out of the picture, but of secondary importance,
    COMMENT:  I don't think sensitivity is the hang up of the Agency. I think
 sensitivity is something that we as analytical chemists have been totally guilty of.
 I once wrote a chapter called "The Vanishing Zero."  We are the ones who have
 been pushing back the zero all the way  down, almost to the single molecule.   Obvi-
 ously, when EPA sees a method coming along  and they can only detect microgram
 quantities,  they say, "That's not good enough.  Why can't you detect picograms ?"
     COMMENT:  We're spoiled. I think that's it.  It's a riatural human phenomenon.
     COMMENT:  In other words, the analytical chemist has gotten himself into —

   COMMENT: We've gotten ourselves into a bind.  There's no doubt.  We always
say  selectivity and sensitivity, the two together,  ought to be considered as a
total factor.
   DR. JOHNSON: I was going to bring this up in general, but it does relate
to this particular type of compound.  Amino acid techniques are being used for
precisely this kind of application for drugs and other things where you have con-
siderable selectivity and sensitivity, and they are  very compound-specific.
   You're not going to have a general multiresidue type of thing.  I haven't seen
or heard much from EPA as far as sponsoring programs that might use this
technique for some of these newer compounds.
   DR. OSWALD:  In relationship to the question on amino acids, we do have
small efforts, at least within our division at this time.  One of the efforts was
directed  toward, if I remember correctly, metabolites of benomyl.  It was a
feasibility study to see what could be done.  The results were positive and as we
find more and more compounds which just do not fit any one scheme — when you
talk about handling or analyzing by some way  — the best way we can do this is to
consider chemical, instrumental, chromatographic, biological, and anything else.
We grasp for that extra  nonexisting solution.  At this time, that's what we have to
do .
   DR. MOYE: Dr. Zweig mentioned nanogram amounts  versus microgram.  I
think, quite frequently,  people working at residue levels limit their vision to
sample size,  that is to say,  the sample sizes  that  have historically been used in
terms of environmental work. In most cases we have the advantage that we can
dig as much dirt or pump as much water or as much air as we need, but when you
get to tissues — animal  tissue and that sort of thing — you're talking about some-
thing else.

   One thing that was driven home to me was the fact that on a liquid chromato-
graphic analytical column, you can inject, under the right conditions, as much
as one milliliter of extract, as opposed to one microliter that most gas chromato-
graphic columns can handle.  This,  relating back to the original sample size,
gives you an extreme sensitivity.
   DR. OSWALD: We have covered pheromones, juvenile hormones, and others.
Are there some other topics that should be discussed?
   DR. ZWEIG:  I may be in a better position to bring up these points because
these are things that we're facing every day, and you are the people  who, hope-
fully, are going to give us the answers.
   How are we doing with toxaphene?  I heard toxaphene mentioned once this
morning.  I know that the people at Cincinnati have developed a method for toxa-
phene, and it's based on mass spectrometry, but I cannot describe it to you.  It's
been published, and maybe one of you  read the bulletin in which it was described.
I'd like to throw this out for a general discussion — how are we doing with toxa-
phene ?
   DR. OSWALD: The methodology which Dr.  Zweig was referring  to for toxa-
phene water is basically a GC multiple ion detection method based on toxaphene
components.  I think most of you realize that the analysis of any multicomponent
pesticide in a biological substrate, whether it's toxaphene,  PCB,  or whatever,
is difficult.  If you look at the other techniques, you're comparing peak positions
on two or more conditions and hoping for the best. In that case, your specificity
is just about gone down the drain.  The same type of problem exists with PCBs.
You would like to convert all of these to one end product and,  depending upon the
conditions of the sample, the number  of other things, some of this is feasible.
   In the case of toxaphene, there are other problems.   The only thing I can
address, at this time,  toward a solution or at least with the type of  involvement
which we have in this particular category, is that we are looking for an index of

 exposure to this type of compound, not looking for the relatively nonpolar material,
 but looking more for the polar metabolite as an indication.  Anyone else like to
 comment on this ?
   DR.  ZWEIG:  In other words, you don't think much of the Buddie method?
   DR.  OSWALD: I didn't say that.  No, what I am saying, since you asked me
 to describe it,  it is a GCMS mass spectrometry multiple ion detection method.
 With some of this specificity built in, I think it's functional. I really can't com-
 ment anymore than I have — I haven't seen it in operation that much.
   There is  a general comment which I'll pass to  all of you. As we begin to look
 at an environmental sample  and ask the  question,  "What are trace levels?" and
 look at the techniques used,  whether it's chromatographic GCL or whatever, the
 first question that always comes up when you get what you feel is an analytical
 result in relationship to validation is how specific is the method?  This is some-
 thing we've talked about these two days.
   Specificity comes in a lot of camouflaged coats; some are easy because of the
 way the molecule's  built, others you have to build in specificity — either from
 the standpoint of cleanup,  detectors, or other techniques.
   The instance that you brought out on the pheromones, juvenile hormones,  you
 have to  build, or at least start from some point, and subfractionate,  and build in
 specificity just  by use of techniques alone, working from an unknown to something
you can give by polarity differences, chemical characteristics, and finally say,
 "Yes, this is something which,  even though the conditions are fairly exact, you
 still have to verify the other means."
   DR.  HILL: I have comments on the toxaphene  question. If you're talking about
toxaphene, you're talking about an American product — Hercules. If you're talk-
 ing about camphechlor, then you're talking about products that are made from
 Lord-knows-what-kind of pine oil to begin with, wherever the pine trees are
grown.  There are two questions here.

   If you're talking about toxaphene, the American product, and most of the
residue analytical chemists that I know—certainly those in my lab—can follow a
toxaphene signal on a GC on into the noise level. They've done this in many
instances of actual environmental samples, such as a massive toxaphene spraying
program in the Southwest, where we were part of a group of labs looking at water
and mud samples taken from reservoirs and rivers.  Even though we had some oil
from the pumping mechanism in this sample, we were easily  able to follow the
very characteristic toxaphene signal down to the parts per billion level.
   The question of toxaphene or~ca,mphechlor — camphechlor is another issue -.-.
arises from the international interest in this compound.  The toxicologists will
raise the questions,  "Well, they've seen the data on toxaphene, not the components,
but the mixture and, internationally, as this camphechlor is being circulated and
used, is it going to be identifiable ? Is the analytical residue signal going to be the
same or is it even relatively the same compound?"
   That question needed to be answered.  The problem is rounding up enough
samples.  I made an effort through my connections with FAO to have their field
offices send in samples.  We didn't get a tremendous response. We have three
foreign samples and even their history is a little dubious.  I  just received word
last week that  Professor Cassda will  accept these samples and do a mass spec
comparative analysis of that with the American-made product — look for and
 spell out any similarities or differences.  This is a big step  toward answering
the toxicological concerns about this  material more than the  analytical.
    There are some interesting things about toxaphene.  I'm surprised it isn't on
 the substitute chemicals lists instead of being on the other side of the  fence. It's
 obviously being used as a substitute for a lot of halogen compounds that are on
 the way out, and as Cassda says, it's biodegradable.

   Because there were fish kills associated with toxaphene spray dips on cattle
several years ago,  I ran an internal experiment at Beltsville in which a herd of
calves was simultaneously spray dipped with toxaphene, and every day and then
at periodic intervals thereafter, the calves, in pairs, were waded through a one-
foot deep rubber or plastic swimming pool, and the water was taken for analysis.
   We were trying to find some practical way of finding out how long this stuff
hangs in the hair of the cattle and can be washed off their legs when they go wad-
ing in the local creek.  We were having bioassays done by another lab on the fish.
   We were able to follow this toxaphene down to about 40  or 50 days,  at which
point it was still giving a reasonable amount of toxicity to  fish. We were having
a little problem with the curve, because the calves would lay down in the pool,
and you'd have a big peak.  They'd urinate in the pool, and you'd have a problem.
It was a very in-house thing.   It's never going to be published; it's a little bit too
crude for that. It was just to find out how long this stuff remains when the farmer
spray dips  and then lets them go in the field, and they wade around in the local
water supply.   It stays on those legs a long time.
   QUESTION: Can your chemists, when they get a total unknown source sample
or water sample, identify toxaphene among all the other peaks that they find?
   DR. HILL:   We could take one approach.  I won't say that we've actually done
this, but we've done the  research  to back it up, and that is, toxaphene  is a
fairly easy compound to destroy and to get by difference, that is, if you have a
signal which looks like toxaphene, you can treat with a reagent, and you can very
readily remove it entirely from the chromatogram or you  can bunch it up into one
corner with a  small number of peaks.  I think some photochemistry has been done
in the same vein.  Under reduction  toxaphene is very  sensitive to reducing agents.
It's not sensitive to oxidizing agents,  which is one of the big virtues in environment,
as far as its field treatment incorporation into soils or muds. It should be reduced
fairly readily  under these conditions.

   DR. LEWIS: Toxaphene like PCB,  except even worst by virtue of the fact that
it is made up of so many compounds, is a problem from the environmental moni-
toring standpoint.  All of these compounds are not going to stay together in  the
same ratios, and any method that depends  on Identification of toxaphene as an
intact mixture of compounds is going to have problems from a quantitive as well
as a qualitative analytical  aspect.

   QUESTION:  Do you find great changes in the pattern we have just experienced
here?  I'm talking about field water residues as toxaphene signatures.
   DR. HILL: Toxaphene  disappears,  but it doesn't change the pattern enough to
not be recognizable.
   DR. LEWIS:  Again you're looking at situations where you knew what was there
to begin with, and you have a little bit of built-in bias.  Using a relatively non-
specific detector — and speaking of specificity, electron capture is not much better
than flame ionization ~ those peaks that you're attributing to toxaphene components
in looking at ratios and patterns and whatever could be something entirely unrelated
to toxaphene.
   DR. HILL: The advantage we had ~ it's also the point I'm making — is that
since we knew it was toxaphene, we were able ~ right down to the point where we
couldn't see  it anymore ~ to still compare those signature pieces, if you will, and
the shape of the pattern and the ratios  were not so much different for us to be con-
vinced that there was any  drastic shifting going on there.


                          Jack B. Dixon*
                        NOT AVAILABLE


*Department of Entomology, Purdue University, Layfayette, Indiana