SUBSTITUTE
CHEMICAL
PROGRAM - THE FIRST YEAR OF PROGRESS
PROCEEDINGS OF A SYMPOSIUM
Volume IV
Chemical Methods Workshop
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF PESTICIDE PROGRAMS AND
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D.C.
JULY 30 - AUGUST 1, 1975
-------
SUBSTITUTE
CHEMICAL
'ROGR AM - THE FIRST YEAR OF PROGRESS
PROCEEDINGS OF A SYMPOSIUM
Volume IV
Chemical Methods Workshop
SHERATON-FREDERICKSBURG MOTOR INN
FREDERICKSBURC, VIRGINIA
N^JULY 30 - AUGUST 1, 1975
-------
The Proceedings of the SUBSTITUTE CHEMICAL PROGRAM -- THE
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
MULTIRESIDUE METHODOLOGY
Dr. H. Alison Moye 1
IDENTIFICATION OF IMPURITIES IN
TECHNICAL-GRADE PESTICIDES
Dr. J. S. Warnefc 31
SENSORY.-CHEMICAL PESTICIDE
WARNING SYSTEM
Dr. Donald E. Johnson 67
Thursday. July 31. 1975
'ANALYSIS OF PESTICIDES AND "
PESTICIDE METABOLITES BY
HYPERFINE LABELING
Dr. Barry Commoner 39
MASS SPECTROMETRY METHODS DEVELOPMENT
Dr. Edward 0. Oswald 117
AUTOMATED CLEANUP AND SPECIFIC DETECTOR
SYSTEM FOR PESTICIDE RESIDUE ANALYSIS
Kenneth R. Hill 127
MULTIRESIDUE METHODOLOGY
Dr. Robert Moseman 141
PESTICIDES IN AMBIENT AIR
Dr. Robert Lewis 147
IN SITU METHOD FOR ORGANOPHOSPHATE
INSECTICIDES
Dr. William T. Colwell 159
TOXIC POTENTI ATORS AS BY-PRODUCTS OF
ORGANOPHOSPHORUS INSECTICIDES
Dr. R. Fukuto 165
RESEARCH PROGRAMS OF THE CHEMISTRY
BRANCH
Dr. Edward O. Oswald 173
SPECIAL INTEREST TOPICS AND GENERAL
DISCUSSION .' 185
MICROELECTROLYTIC CONDUCTIVITY
DETECTOR
Jack B. Dixon 205
-------
Wednesday
-------
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
GAS
CHROMATOGRAPHY
v^
f
i
DERIVATIZATION
SILICA GEL
i COLUMN
CHROMATOGRAPHY
EXTRACTION
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
CO
_L
Chlorine-Containing
Cap tan
Chlorobenzilate
Difolatan
Endosulfan
Folpet
Methoxychlor
PCNB
Perthane
Trifluralin
L
Phosphorus -Containing
Azinophos Ethyl
Azinophos Methyl
Azodrin
Demeton
Malathion
Methyl Parathion
Parathion
Phorate
*13 pesticides which could not be satisfactorily gas chromatographed intact.
Nitrogen-Containing
Amitrole*
Atrazine*
Avadex
Benomyl*
Bromacil
Bux*
CIPC
Diquat*
DNOC*
Ferbam*
Furadan*
IPC*
Monuron*
Simazine*
Temik*
Zectran
-------
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
AMITROL
AZODRIN
BENOMYL
DIQUAT
FERBAM
TEMIK
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.
i
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
PCNB
Captan
Folpet
Endosulfan
Perthane
ng
ng
Chlorbenzilate 10 ng
Methoxychlor 2 ng
oDTime[rain)
-------
Figure 5. Chromatograms of an 8-compound pesticide mixture.
Column operating parameters are shown in Table 3.
1
90
SO
7U
60
50
40
30
2 20
u
— 10
w
Jg70
c
es.
^ 60
SI
40
30
2C
i
C
.
•
•
•
A
4% SH-30/6'i OV-210
i
2
5,4,5
7
5
1
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
J K
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
1
Fraction I
Hexane
PCNB(94%)
*
Fraction n
60% Benzene/
Hexane
Trifluralln (105%)
I
Fraction IE
5% Acetonltrlle/
Benzene
Captan (102%)
t
Fraction IV
10% Acetonltrlle/
Benzene
Atrazlne (100%)
1
Fraction V
20% Acetonltrlle/
Benzene
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%)
Temik
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
I
Fraction I
Hexane
No pesticide
Fraction n
60% Benzene/Hexane
phorate
methyl parathion
parathlon
Azodrln
Fraction in Fraction IV
5% Acetonltrlle/ 10% Acetonitrile/
Benzene Benzene
malathion demeton
azinphos-methyl
azlnphos-ethyl
-------
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
less.
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
10
-------
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).
10
41
j
30
•
20
in
, ;
TRACOR 22:, lie.
6'xi/s"i .:•.
i%.:
F low ra tc : SI' ;.
CRUUC SOIL EX.'RACT
10
MINUTES
11
-------
Figure 7
20
10
CO*. Bcnzor.o/r.cxanc-
fract ion
rRIFLURALIN ENDOSULFAN
PERT! IAN H
METHOXYCMLOR
D 5^ Acetroni tri le i r. henzer.o
fraction
40
•
..:
.--
30
CAPTAN CHLOROBENZILATE
20
FOLPET
10
CIPC
^JU'
DIFOLATAN
10
Ml NUTUS
12
-------
Figure 8
30
20
*— \
|io
u
to
J
1
u,
:/:
340
at
K
0
^
E
101 Acetroni t r i Ic ;.n benzene
fraction
ATRAZINE 5 SIMAZINE
i
i
_ /""""" _
^
20°. Acetroni tr: le ;n bcr.^cnc ,
f tact ion
BROMACIl
^^J,w>»-1*>*i'v^
^^jj5n^VW''t*"'
"0 i> iU 15— -• 7TT '
MINUTES
13
-------
temperature or for one hour in the presence of hexane and a small amount of
pyridlne (Figure 9).
Figure 9
O
R-O-C-NHCH3
C2F5<
\>
o
HEXANE
PYRIDINE
25°C
1 hour
,CH3
R-O-C-N' o
c*
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
14
-------
Figure 10
o
^ \O-C-NH-CH3
O
SEViN
pH12
2,5-dichlorobenzene-
sulfonvl chloride
cl
O
II
o-s
II
\J ° u
1-naphthyl-2,5-dichloro-
benzenesulfonate
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
2-AB
.0003%PYRIDINE
IN
BENZENE
A95°C,30 MIN,
C-NH-C
15
-------
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
oo
JO
*/•»
50
10
EC DETECTOR
10 % DC-200, 220°C
ATT., 4 x|04
A, 5ng BENOMYL
B, I ppm SANDY SOI I
C, CONTROL
MINUTES
-------
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-
17
-------
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
18
-------
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
o-phthalaldehyde
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
2min.
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
19
-------
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
25%DiOXANE-K2O
j50%DIOXANE-H2O
T WATERS 660
SOLVENT PROGRAMMER 35ft.
FLUOROMETER
(>W '— •
^ jjC1S COLUMN
SAMPLE
INJECTION
VALVt
25 u I
0.02
0.5
1
_
5O'C
i , - ,.«., !
— pwui-aw^r1 —
DELAY COIL
MILTON ROY
PUMPS
^
_J
A
fiTj- i^ ! r
• - c^ ,— ^*—j
FLOW CELL VA,.AS--
7
Jbuu.
RECORDER
M NaOH OPA
ml/min REAGENT
1.0 mi/min
20
-------
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
2
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
efficiency.
This means that you can use very dilute samples and, therefore, theoretically
approach very low limits of detection. This arrangement — notice Lannate comes
21
-------
Figure 16
50 ng
25 ul of
2ng/ul
2000 PLATES
uC18 COLUMN
409/o DIOXANE-H-O
1ml/mfn FLOW
O
10
MINUTES
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
power.
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
22
-------
Figure 17
0
220O PLATES
U.V. DETECTOR
25 ul of O.O4ng/ul
uC18 COLUMN
4O»/0 D!OXANE-H2O
10
MINUTES
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.
23
-------
Figure 18
o-LANNATE
o - BAYGON
A- TEMIK
10 100
ng/25ul
1000
Figure 19
1ng LANNATE
25 ul of 0.04ng/ul
2000 PLATES
uC-,8 COLUMN
40 °/0 DIOXANE-H2O
1ml/min FLOW
0
5
MINUTES
10
24
-------
0.0 ppm
Figure 20
LANNATE IN SOIL
(SANDY LOAM)
O.OSppm
0.5 ppm
0
10
10
10
MINUTES
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.
25
-------
Figure 21
LANNATE IN LETTUCE
0.2 ppm
0.0 ppm
MINUTES
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.
26
-------
Figure 22
U.V. DETECTOR
L ANN ATE
125 ng
CHECK
LETTUCE
25Omg
10
MINUTES
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.
27
-------
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
solvent?
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.
28
-------
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?
29
-------
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.
30
-------
IDENTIFICATION OF IMPURITIES IN TECHNICAL-GRADE PESTICIDES
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
31
-------
Figure 1. Main Reactions in the Preparation of Parathion
Step 1.
•»KVJ«
ethanol
T ^2°5
phosphorus
penta sulfide
2(RO)2 P(S)SH
O,O-diethyl
phosphorodi-
thioate
H2S
hydrogen
sulfide
Step 2.
Step 3.
(R0)2 P(S)SH
Cl,
(RO)2 P(S)C1
O,O-diethyl)
chlorophos-
phorothioate
Step 4. Cl
benzene chlorobenzene
HNO3
C1 \ ^/ N°2
p-nitro chloro-
benzene
Step 5. ci
\ /
Step 6. (R0)2 P(S)C1
-
-------
Table 1: Reactions Occurring During Production of
Parathion
Reaction
Number
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
' 16.
18.
19.
20.
21.
inOTT IT*" i
(RY)2P(Y)SH4 R'OH
(RY)2P(Y)SH4 R'OH
(RY)2P(Y)SR 4 R'OH:
P2S5
frtY*i TWinrt t
c.
•/DVli D/V1CIJ J. V f\
. (±ti J_ r( i jail T •WoW
(RY)2P(Y)YH ^
[oi
Z
(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' +
2
> f^^Y^ r*fv\OTT i n
(RY)2P(0)SH
(RY)2P(Y)YP(Y)(RY)2 4
(RY)?P(Y)SSP(Y)(RY)2
:*• (RY)2P(O)SR'
(RY)2PSH + S
>- (RY)_POR' 4 H.S
£» £»
*• (RY)2PSR' 4 H20
H20
RSH
[2S
(RY)2PSH — i^L. (RY)2P(S)SP(YR)2
(RY).PSH 4 Cl,
6 b
l*t * /-> *^V'«3 )« ** « Cl— ™"
2 2
(RYLP(S)SSP(S)(YR).
t .
2S Cl + 2H 0
C12
ArH " ArCl
HXO3
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
2HC1
•R)242HC1
33
-------
Table 1. (Continued)
Reaction
Number
22.
23.
24.
25.
26.
27.
28.
29.
30.
3.1.
32.
ArCl + ROH -
ArCl + ArONa
rxxrxC1
(RY)2P(Y)C1 +
(RY)2P(Y)OAr
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
fi
•*- (RY),PO
3
) ^ (RY)2P(Y)OH + RYH
3SR
(») 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.
34
-------
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
'recursor^)
SM
1
2
19
2
3
18
3
2
7
2
6
2
2
2
2
12
12
12
12
4
9
21
16
20
11
8
8
9
11
22
SM
24
25
24
27
28
29
28
27
48
Reaction
Number<°>
1
2
29
3
4
29
S
5
6
8 + 7
8+7
9
10
18
11
12
13
14
IS
10
3
29
4
29
17
16 + 17
18+17
18 + 17
17
18
—
19
19
20
20
20
20
21
21
31
Likelihood o(
Being Present In
Step
-------
Table 2: (Continued)
No.
34
35
36
37
38
39
40
41
42
43
44
46
46
47
48
49
SO
51
52
53
54
55
56
57
58
59
60
61
Contaminant''^
C6H3(N02)(OH)C1
CftH2(N02)2(OH)Cl
C6H4(N02)OC6H4N°2
C6H3(N02)2°C6H4N°2
C H (NO )C10C B NO
632 6
C-H,ClOCaH NOa
64 642
^^^^^*^^^^
I O T I O 1
Q
C§X ®~N02
Q
O N-^T )^^-NO
2 O/^x^
CgH4(OH)Cl
CflH4(N02)OR
C.H (NO) OR
63 22
C8H3(N02)(OH)C1
C8H2(N02)2(OR)C1
(HO)2P(S)OC8H4N02
(RO)2P(S)OC8H3(N02)2
(BO)2P(S)OC8H3N02Cl
(BO)P(S)(OC8H4N02)2
(BO)(RS)P(S)OC8H4N02
(BO) P(0)OC H NO
A v ^ *
(RO)P(0)OC8H4N02
(BO)2P(0)SC8H4N02
(RO)(N02C6H40)P(S)OH
(HO)(N02C6H40)P(0)OH
(RO)2P(S)OH
(RO)2P(0)OH
RSH
RSSR
Precursor*'')
29
30
27 + 32
28+32
27+33
29+32
27+34
26+32
27+31
31
31+34
34
28
27
28
29
30
18 +32
18+33
18 + 34
48
19+32
20+32
48
21+32
48
48
48
56
48
58
3 or 52°
60
Reaction
Number*0*
21
21
23
23
23
23
23
23
23
24
24
24
22 or 27
22 or 27
22 or 27.
22 or 27
22 or 27
25
25
25
26
25
25
30
25
10
10
31
7 +6
31
7+6
31
32
Likelihood of
Being Present in
Step(d> Final Product(e>
5 o
5 o
5 +
5 o
5
5 o
5
5 0
5
5 0
5 0
6 o
*
6 o
8 "^*
6 +
6 o
6 o
6 Parataion
6 +
6 o
6 *
; * **
"S "^""^
St
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.
36
-------
Figure 2. Gas Chromatogram of Technical-Grade Parathion
—
3
12
11
13
d
-i—•—i—i—i—i—i—r
9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 CO
—t
63
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
(C2H50)2(C2H5S)PS
C6H4(N02)OH
1L.X-
'.'"'••••:-i'". ;"•'••", ; i 'i"i' . .--i -,- .--i . -
S3 1CB 110 123 133 1M 1S3 1SJ !TO ISO 135 2C3 210
PARATHION
or isomer
.A-
Dithio
analog
of
PARATHION
/
3 C23
5- !C3 -. •! 2 C IBS 2t3
38
-------
Figure 4. GC-MS of Technical-Grade Methyl Parathion
§
s.
8.
o I-
-'1
(CK30),(CH3S)PS
METHYL PARATHION
or isomer
Dithio
analog
of
METHYL
IPARATHION
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.
39
-------
Figure 5. Synthesis of Malathion
and
Step 3b
mechanol
Step 1
CH-
Step 2a II \> +
CH- C
6
maleic
and anhydride
CHCOoc.n.
Step 3a II * 3
diethyl
maleatc
,J
Step 2b II 0
CH-9X
0
+ Vs
phosphorus
pentasulfide
2(CH30)2P(S)SH
0,0-dinethyl
ethanol
diethyl
inaleate
4- (CH30)2P(S)SH
0,0-ditnethyl
phosphorodithioate
(CH30)2?(S)SH
maleic "0,0-dimethyl
anhydride phosphorodithioate
(CH 0) P(S)SCH-CS
J * i r
+ 2C H OH
ethanol
0,0-d inethyIphosphorod i thioy1
'succinic anhydride
MALATHIOH
H2S
hydrogen
phosphorodithioate sulfide
CHCOOCflHr
(CH30)2P(S)SCH-CN
0,0-direthylphcsphorodithioyl-
euccinic anhydride
(CH,0)-P(S)SCKCCOC0H-
32 I 25+ H,0
MALATHION
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.
40
-------
Figure 6. GC-MS of Technical-Grade Malathion
MAIATH10N
-------
Figure 8, GC-MS of Technical-Grade Phorate
PRORATE
to
8
r>
CD
sj
o
(C2H50)2(C2H5S)PS
(C2H50)2P(S)SP(S)(OC2H5)2
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
Dursban.
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,
43
-------
Figure 9. Synthesis of Dursban.
Step 1
2CI.
N
pyridine chlorine
+ 2HCI
N
2, 6-dichlorop/ridine hydrogen chloride
Step 2
IpX
rx.A:i
+ NaOH
150-155 C^
H2°
•f
2,6-dichloropyridine sodium t-butyl
hydroxide alcohol
N
6-chloro-2-
pyridinol
NaCl
sodium
chloride
Step 3
Cl-
HO N Cl
6-chloro-2-pyridinol chlorine
Cl
HO
O
«» '
N
Cl
3, 5, 6-trichloro-
2-pyridinol
+ 2HCI
hydrogen
chloride
Step 4
SH
ethanol
phosphorus
pentosulfide
O,O-diethyl phos-
phorodithioate
hydrogen
suLfide
Step 5
Cl,
Step 6
O,O-diethyl phos-
phorodithioate
Cl
O,O-diethyl chloro-
phosphorothioate
Cl
(C,H,0), P(S) ,
2
chlorine
Cl
O,O-diethyl chloro-
phosphorothioate
HCl
hydrogen
chloride
HO
Cl
Na2CO3
DMF
O
v» .
N
3, 5, 6-trichloro-
2-pyridinol
Cl
Cl
O, O-diethylO-(3, 5, 6-trichloro-
2-pyridinyl) phosphorothioate
(Dursban)
44
-------
Figure 10. Alternate Synthesis of 3,5,6-Trichloro-2-Pyridinol
ci ci
A
Step Al
pyrldlne
4C12
chlorine
2,3,5,6-tetrachloro-
pyridine
Step A2
2,3,5,6-tetrachloro-
pyrldine
sodium
hydroxide
3,5,6-trichloro-
2-pyrldinol
4HCi
hydrogen
chloride
NaCl
sodium
chloride
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.
45
-------
Figure 11. GC-MS of Technical-Grade Dursban
*>•
(C2H50)3PS
(C2H50) 2P (S > °P (S} (OC2H
5'2
(C2H50)2
-------
Figure 12. Three Methods for the Synthesis of PCNB
ci
'2/13/24
CI
HN03 + H2SO4
Ci
2b\HN03+H2S04
c> ^^^ ci
ci
Ic
HN03
-------
Figure 13. GC-MS of Technical-Grade PCNB
PCNB
8
-)-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 contains.no hazardous trace impurities.
48
-------
Step 1
SCep 2
Step 3
Step 5
Figure 14. Synthesis of Trifluralin
+ Cl,
Toluene
CH3
o
Cl
COOH
HN0
OOOH
step 4 TQ
SF,.
Ci Sulfur Tetrafluoride
CF3
4- NH(C3H7)2
1 Diisopropylamlne
Cl
p-Chlorotoluene
COOH
p-Chlorobenzoic Acid
COOU
4-Chloro-3,5-DlnItrobenzoic Acid
Recry'd ETOH
CF,
4-Chloro-3,5-Dinitrobenzocri-
fluoride
Recry'd-hexane-benzene
[3
Pr2N.HCl
Trifluralin
49
-------
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
TRIFLURALIN
or isotner
3 13 M 30 10 SO .-J 7B iO 33 I(T3 :!0
IUO 2OJ 2IO ZXO
50
-------
Figure 17. Main Reactions in the Preparation of Folpet
Step l
or
CH
rH
CH
[01 v
catalyse
naphthalene o-xylene
Step 2
9
0 + RH
200-220 C
ammonia
HH
phthallc
anhydride
phthalinide
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
51
-------
Figure 18. GC-MS of Technical-Grade Folpet
PHTHALIMIDE
N-CKLOROOXAMIC ACID
(HOOCCONHC1)
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
52
-------
Table 3: Reactions Occurring During Production of Dursban
Reaction
Number
Generalized Reaction!')
®
(x»l to 5)
N
xHCl
+ NaOH
(xaO to 4)
N
+ NaCl
HO N Cl
Cl
ZHCl
5
6
7
9
10
11
12
13
14
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)
Reaction
Numb«r
. .
G«n«r»tued Reaction'*'
15
16
17
18
1$
20
21
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
2HCI
(C2H50)2P(S)Cl
(x « 1 to 4)
(C H O) P(S)0
(x « 1 to 4)
(x » I to 4)
>. (C2H50)2P(S)OH
JAK
* K
-------
Table 4: Reactions Occurring During Production of Folpet
Reaction
Number
1
3
4
5
6
7
8
9
10
11
12
13
1/4
15
16
17
18
19
20
21
22
23
24
25
26
27
28
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 •
CCljSCl + RCH-CHR
RCH-CHR + CCl.SOj
(RCO)2NSCC13 + H2
(RCO)2NSCC13-M
^^ c ^^ r ™i
0 0 °
^l^ o,-§v
- 2 RCOOH
•-RCONHj + RCOOH v v (RCO)2NH + H20
•*• 2 RCONH2
RCONHj + H20
2CSC12 + S2C12
CCljSCl
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
,J!2?Ji^(Rco)2Nscci3
lia!Ja2!L.(Rco>2Hsoacci3
— -— (RCO) 2NCSN(COR) 2
— — — (RCO) 2NSSN(COR) 2
— -— R(CC13S)CHCHC1R
Cl »- RCHC1CHRCC13 + SOj,
0 •• (RCO)2NH H- CCljSOH
^(RCO)2NS02CC13
(RCO)2KH + 2CC13SC1 (RCO)2NC1 + CC13SSCC13 + HC1
RCON-H2 + 2CC13SC1
RCONHC1 + CC13SSCC13 -f KC1
a. R * alkyl or aryl group.
55
-------
Table 5: Possible Contaminants in Folpet
No.
Contaminant
Likelihood of Bed
Pre- Reaction Present in Final
cursor3 Humber^ Stepc Product^
9
6. HOOCCH-CHCOOH
COOH
SM.
SM
1
3
1
3
11
— 1
1
2
1
2
1
1
1
1
3 1 or 4
3 1- or 4
••
'"2
5
7
4 2
6 2
9.
26
* 2
2S s
56
-------
Table 5. (Continued)
^— ••
No.
.10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Pre- Reaction
Contaminant cursor3 Number**
@^
KOOCCH=CHCONH2
0
CH-C
1! ,NH
H2NOCCH=CKCONH2
cs2
csci2
s2ci2
cci3sci
cci3so2ci
s
cci3sscci3
CC13SOH
cci3soscci3
cci3>so2scci3
CC13S03H
9
8
4
6
11
12
11
SM
14
14
17
17
15
15
16
21
17
22
17
26
21+17
22
23
18
5
6
4
6
4
5
6
—
7'
7
11
12
8
9
10
14
11
16
13
25
15
16
17
18
Likelihood of Being
Present in Final
Stepc Product*1
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3.
4. or St
3
4 or St
4
4 or St
4
4
4
4
4-
0
0
0
0
0
+ .
0
-H-
+
0
0
•f
+
57
-------
Table 5. (Continued)
Ko.
25.
26.
27.
28.
29.
30.
31.
32.
Contaminant
CC13SH
|O T NSCC13
9
CH-C
II XNSCC13
CH~§
\/~~\\ ^
0
CH- Cx
II NSO CC1,
CH-C * •>
6
s^ c c /^
foJc)Ncs
-------
Table 5. (Continued)
Ko.
33.
34.
35.
36.
37.
38.
39.
40.
41.
Contaminant
?0
3
LJ^NSSN^XJ
-C C-CH
ii yssx' ii
CH-C C-CH
0 0
9 9 ^.
cn-c xC^/Ox
|l JISSK^ ("j
*™ V T *"""^'l>^ ^^
0 0 ^sx'^
0 .
ci3csc-c's
**-{*
p
Cl CSC-C
J | XNSCC13
p
ci3csc-cv
1 XNS02CC13
6
HOOCCH(SCC13)CHC1COOH
HOOCCH (SCC13) CHC1CONH,
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
59
-------
Table 5. (Continued)
No
42
43
44
45
46
47
48
9
_
Contaminant
.
ci3cc-cx
1 >
cic-s
0
C13CC-?
6
HOOCCOOH
HOOCCONH2
H2NOCCONH2
HOOCCOKHC1
0
II
af
0
II
nfs-
y^/
II
0
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.
60
-------
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.
Ross?
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
61
-------
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-
teresting.
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
62
-------
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
63
-------
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?
64
-------
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.
65
-------
SENSORY CHEMICAL PESTICIDE WARNING SYSTEM*
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
67
-------
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.
Methods
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.
68
-------
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
-5
Methyl parathion 0.97x10 (6) 12.8 366**
-5
Carbofuran 0.52x10 * 10 419**
-7
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.
69
-------
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-
2
liters, the area covered is approximately 30 cm . When all factors Involved except
2
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
70
-------
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
analysis.
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.
71
-------
2
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
2
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
experienced.
• 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
72
-------
2
were sprayed by airplane with methyl parathion at an intended dosage of 17 ^g/cm
2
(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
2
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.
73
-------
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.
Results
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
75
-------
C5
Table 2: Candidate Odor Agents
Boiling Point(15)
Compound
artificial musk, 1-tertbutyl -3 -methyl -
2,4,6 -trinitrobenzene
typical synthetic macrocyclic musk,
11-oxahexadecanolide
musk ambrette
musk xylene
musk ketone
muscone-active principle in natural
musk, 3-methylcyclopentadecanone
civetone, 9-cycloheptadecen-l-one
2-hexyl-3-methoxypyrazine
2 -isopropyl -3 -methoxypyrazine
2-isobutyl-3-methoxypyrazine
2 -propyl -3 -methoxypyrazine
alpha -ionone, beta-ionone
vanillin
ethyl vanillin
deca-trans, trans-2,4-dienal
2 -methoxynaphthalene
Mol.
Wt.
283
256
268
297
294
238
250
194
152
166
152
192
152
166
152
158
t. ° C
m.96-7
m.35
25
25
25
328
130
342
159
—
—
—
—
127
81
250
285
170
.25
—
272
pr. , Odor Thresh. ,
mm Hg ppb (v/v)(^)
0.001
—
2.5xlo"5
1. x 10"5
2.4X10"6
760
0.5
742
2
o.ooi<17>
0.002
0.0)2
0.0)6
12 0.0 1.3
1
760
atm. 0.016
15
1.7X10"4
o.o?o<18>
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
Compound
coumarin
methyl coumarin
methyl anthranllate
1 . 7, 7-trimethylbicyclo[4. 4. 0 J-
decan-3-one
1, 7, 7-trimethylblcyclo[4. 4. OJ-
decan-3-formate
11 -acetate
" -propionate
" -butyrate
" -acrylate
alphft-(2-hydroxycycloh«xylmethyl)
butyrolactone
alpha-(4 -hyd roxycyclohexy Imethyl)
butyrolactone
1 -(phenylethoxy)adamantane
gamma -decalactone
gamma -undecalactone
gamma-dodecalactone
benzophenone
phenylethyl phenylacetate
Mol.
Wt.
146
161
151
195
224
238
252
266
250
196
196
256
170
184
198
182
240
t, °C
298
139
305
135
87
85
90
100
115
86
203
203
176
281
286
170
305
324
pr. ,
mm Hg
atm.
5
760
15
0.001
0.01
0.01
0.02
0.05
0.01
3
3
1
atm.
atm.
11
760
760
Odor Thresh.
ppb
0.28
0.65
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
rose-hyacinth
-------
oo
Table 2: Candidate Odor Agents (cont. )
(15)
Boiling Point
Compound
benzyl ctnnamate
hexyl cinnamic aldehyde
benzyl benzoate
benzyl salicylate
o -b romophenol
alpha and beta-santalol
Mol.
Wt.
238
216
212
228
170
220
t, t>C
350
305
323
368
194
302
pr. ,
mm Hg
760
760
760
760
760
760
Odor Thresh. ,
ppb (v/v)<16>
0. 001
Odor Character
sweet odor of balsam
faint, pleasant, aromatic
pleasant
unpleasant
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
conditions.
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,
2
and 0.1 Mg/cm . Then the application level was adjusted upward or downward until
79
-------
Table 3; Candidate Visual Agents
Compound
trans -stilbene
anthracene
phenanthrene
acridine
alpha-benzytnaphthalene
para-benzylnaphthalene
1,4 -diphenylbutadiene
4 -methoxybenzophenone
carbazole
triphenylmethane
2,5-diphenyloxazole
1,1,2,2-tetraphenylethane
2-hydroxybenzothiazole
diphenyturethane
triphenyl phosphite
quinazolinone
benzimidazole
1,2-benzophenazine
l,l'-binaphthyl
1,3,5-trinitronaphthalene
l-naphthyl-2-tolyl ketone
m-terphenyl
diphenylene disulfide
phenothiazine
thioxanthone
fluoranthene
p-terphenyl
anthraquinone
diphenylsulfone
Boiling Point, °C
305 (720 mm Hg)
340
340
345-6
350
350
350 (720 mm Hg)
354-5
355
358-9
360
360
360
360
360
360
360
360
360+
364 (explodes)
365
365
366
371
372
375
376
377
379
80
-------
Table 3: Candidate Visual Agents (cont.)
Compound
melene
triphenylmethanol
retene
pyrene
N -benzylsuccinimide
4,4f -dibromobenzophenone
4,4f -ditoly sulf one
1,2-benzofluorene
9 -phenylanthrac ene
0,0' -quaterphenyl
tetraphenylethylene
triphenylene
9,10-benzophenanthrene
1,2-dihydroxyanthraquinone
tetraphenylmethane
1,2, -benz anthracene
1,8 -dinitronaphthalene
chrysene
2,2-binaphthyl
1,3,5-triphenylbenzene
1,2,6 -trihydroxyanthraquinone
1,2,7 -trihydroxyanthraquinone
11,12-benzofluoranthene
picene
Boiling Point, PC
380
380
390
393
395
395
405 (714 mm Hg)
413
417
420
420
425
425
430
431
435
445 (decomposes)
448
452
459 (717 mm Hg)
459
462
480
519
81
-------
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
82
-------
7CJ
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
56739
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
83
-------
Table 4: Typical Data for the Disappearance of Fluorescing Agents
from Glass Mates
Visual Fluorescing Strength*
Fluorescing Agent and Level
of Application, Mg/cm
Anthracene
Anthracene
Anthracene
Phenanthrene
Phenanthrene
Phenanthrene
Fluoranthene
Fluoranthene
5
3
2
50
40
20
5
3
Hours
p_
S
S
S
S
S
S
S
S
16
S
S
W
S
W
o
S
M
20
S
S
o
S
W
S
W
24
M
S
S
O
M
0
BQ
M
M
W
M
40
W
W
W
W
44
W
O
O
0
48
0
*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
84
-------
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
2
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
85
-------
Table 5: Candidate Odor Agents Subjected To Screening Tests
Benzyl cinnamate
Benzyl salicylate
beta-Phenylethylphenylacetate
Coumarin
Ethyl vaniUin
2 -Methoxynaphthalene
Santalol
Vanillin
Methyl anthranilate
Ethyl anthranilate
Ethyl salicylate
Eugenol—USP—prime
Sandela
Musk ketone
Musk ambrette
Musk xylol
Hexadecanolide
Heptaldehyde
2-Phenylethanol
Pentadecanolide
Anisyl acetate
Terpineol
Laurine
Jasmonyl
Methyl eugenol
Geraniol
Phenoxyethyl isobutyrate
Anisyl alcohol
Acetanisole
Galaxolide (musk)
6 -Methylcoumarin
gamma-Dodecalactone
Skatole
alpha-Chloroacetophenone
alpha-B romo acetophenone
Menthol
Dimethyl sulfide
Several proprietary
perfume oils and odor
masking agents
86
-------
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
87
-------
too
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
20
©
A —
glass plates
leaves
0
^-^_
5 t'o 15 20
Time,
• A
25
Hrs.
' o—
, , 1 A__|
10 35 40 45 5>
88
-------
Table 6: Observations Made on Sensory Items During Batesville Field Test
Compound
Anthracene
Phenanthrene
Skatole (on plate)
Cone. _
/ig/cm
3
5
7.5
10
40
50
60
40
Time, Hours
0
S
S
S
S
s
s
s
s
4
w
8
S
s
w
w
w
s
8
E
W
S
S
E
E
E
S/W
22.5
E
I
I2
W/E
48
E
E
E
ff/1
\
Skatole (on paper)
2-Phenylethanol
/2-Phenylethyl-
Phenylacetate
6
9
12
24
1
3
6
9
1
3
6
9
• s
s
•s
s
s
s
s
s
s
s
s
s
S/W
s
s
s
s
s
s
s
s
(paper
S
S
W/E
S/W
S
' S
S/W
S
S
S
W/E
E
E
W/E
M
M
M
M
M
E
E
VW
vw
M
M
M
strip lost)
S
S
M
M
M
M
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.
89
-------
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
90
-------
disappearance of Furandan from Euonymus plants
10
Z3
30 40
Time, days
50
70
Figure 7. The disappearance of Guthion from Euonymus plants
103
-------
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.
92
-------
Figure 8. Incremental cost for odor device
01
o
O
•a
§
38
ti o>
o> a
S ~
2
'I
$100, 11%
$200, 5.5%
10 -
100 200 500
Total Cost of One Field Application
$1000, 1.1%
$1000
60
0}
O
13
§
•»H
:S
-------
Discussion
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.
94
-------
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
95
-------
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).
References
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
(1969).
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).
96
-------
DR. LEWIS: Don, did it happen to rain during any of these tests that you were con-
ducting?
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.
97
-------
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.
98
-------
Thursday
-------
ANALYSIS OF PESTICIDES AND PESTICIDE METABOLITES
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
"V"tv
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,
Missouri
99
-------
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
100
-------
Figure 1
ESR HYPERFINE SPECTRA OF FREE RADICAL FORMS OF DIFFERENT PESTICIDES
PCNB
(Red., Glyc.)
Me. PARATHION
(Red., Elec.)
TRIFLURALIN
(Red., Glyc.)
DINOSEB
(Red., Glyc.)
FENTHION
(OX., TFMS)
CAPTAN
(Ox., TFMS)
9 - 2.005
10 Gauss
BROMACIL
(X-ray, Ada.)
METHOXYCHLOR
(X-ray, Ada.)
A
A
A-
.
\J \ |
V
(X-ray, Ada.)
/
PROPHAM
(X-ray, Ada.)
,
FERBAM
(X-ray, Ada.)
V
..
AZODRIN
(X-ray , Dext.)
t
I
DURSBAN
(X-ray, Ada.)
g - 2.005
101
-------
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
102
-------
Figure 2
ESR SPECTRA OF METABOLITES FROM RATS ADMINISTERED DIFFERENT PESTICIDES
p-NITROPHENOL
METABOLITE OF Me. PARATHION •
URINE
(Electrolytic Reduction)
UNIDENTIFIED METABOLITE OF
URINE
(Glyceraldehyde Reduction)
10 Gauss
UNIDENTIFIED METABOLITE OF
FENTHION -
FECES
(TFMS Oxidation)
T
UNIDENTIFIED METABOLITE OF
DINOSEB -
URINE
(Electrolytic Reduction)
g = 2.005
103
-------
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
104
-------
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.
105
-------
Figure 3
ESR Electrolytic Cell
-Stop Cock
Platinum Electrode
3mm.
1mm<
•^—8mm.
-Sample Reservoir
t
Tungsten Electrod*
106
-------
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.
107
-------
Figure 4
ESR Spectra of Reduced Pure PCNB and
2 -Amino-4 -nitro -6- sec. butylphenol
Sfjn D..IK
198
-------
Figure 5
Slepwise Electrochemical Reduction of a
Mixture of PCNB and 2-Amino-4-nitro-6-sec.butylphenol
10£
-------
Figure 6
ESR Spectra of Reduced Pure
Methyl Parathion andp-Nitrophenol
110
-------
Figure 7
Stepwise Electrochemical Reduction of a
Mixture of Methyl Parathlon and p-Nitrophenol
-111
-------
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.
i
*
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-
olite.
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
112
-------
Figure 8
z
ESR Spectra of Urine Extracts from
Rats Fed Control Diet —
Electrochemical Reduction
^A*^
C. -1.5 V
•V i
Ail
q . 2.005
-«.!-—I/
E. -2.25 V
113
-------
Figure 9
ESR Spectra ol Urine Extracts from
Rat* Fed Dinoseb Diet -
Electrochemical Reduction
114
-------
Figure 10
ESR Spectra ol Urine Extracts from
Rat* Administered Methyl Parathion-
ElBctrochemlcal Reduction
'^^
115
-------
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
4
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.
116
-------
MASS SPECTROMETRY METHODS DEVELOPMENT
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
117
-------
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.
118
-------
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
source.
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
119
-------
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-
gations.
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.
120
-------
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
121
-------
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.
122
-------
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.
123
-------
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
instruments.
124
-------
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
funding.
125
-------
AUTOMATED CLEANUP AND SPECIFIC DETECTOR SYSTEM
FOR PESTICIDE RESIDUE ANALYSIS
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
approach.
. 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,
Maryland
127
-------
•m
Wo
Figure 1
128
-------
Figure 3
SOLVENTS
SAMPLE INJECTION
VALVE
FLAME PHOTOMETRIC
DETECTOR
GLASS TO METAL
FITTING, 1/8" TC
1/8"
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
129
-------
Figure 4. Standard Malathion
1401
T-0-93356
M- 23.93332 {SLOPE)
i> • 2.01763 (INTERCEPT)
0-5 I
5 NANOGRAMS Y-MX+ b
Figure 5
130
-------
Figure 6
MALATHION
PHORATE
(THIMET*)
:FENSULFOTH!ON
: (DASANI7R )
OXAWIYL
(VYDATE*)
s ^ o
yP-S-C-C-0-CH2
CH30 H2C-C-0-CH2
~ .""-.. "—. 0
CH3CH20'
CH3CH2CL S
V
\tl
rP-S-CH2~S-CH2CH3
-CHS
CH3 0 O
NN-C-C=N-0-C-NHCH3
SCH3
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
131
-------
Figure 7
BENZENE
50m!
A
= 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!
ACETONE
50 ml
C
0 P=3,SQ
E P=0,S02
P=09SO
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.
132
-------
Figure 8
!00
E
c
10
2:
o
C-
CO
UJ
LiJ
Ct
<£..
O
P»S,S
__J
P=0,SO
50
ICO 150
EFFLUENT - ml
200
250
1-3ENZENE'
•ACETONE IN BENZENE —
i % —i— 5 % 1— 75 %
>00
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
133
-------
Figure 9
Conventional
GC Analysis
Cleanup, LC-FP
Analysis
Weigh-
25 g. sample
(1 pprn = 25 vg.
Weigh -
25 g. sample
(1 ppm » 25 ug.)
Solvent extraction
(blending)
Solvent extraction
(blending)
Filtration, partition
if needed
Filtration
1
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,
134
-------
Figure 10
PYROLYSIS OVEN TEMPERATURE
-UO'C-.U-320-C ,14 660'C
100
I 40
I-
1"
<
lu
n.
DASANIT
CH30MATOG3AM OF
STANDARD SOLUTION
OF DASANIT AND ITS
METABOLITES
OASANIT
0-ANALOG
EULFONE
10 5o
. ELUTICH TIMS-MINUTES
-COLUMN F.LUANT-
•BENZENEJ-25% ACETONE-f-10% ACETt-NE-
-ACETONE —
H 40
< 0
Ixl
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
(DASAfllT
3-ANALOG
. ACETONE
rj*AC£TONt—
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
135
-------
Figure 12
3°r FP-LC ANALYSIS
OF OXAMYL IN
eof- ALFALFA
60
« 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.
References
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
137
-------
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
see.
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-
esizing.
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
138
-------
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?
139
-------
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.
140
-------
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
141
-------
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-
A
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.
A
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
142
-------
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
i
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.
143
-------
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
them.
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.
144
-------
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.
145
-------
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
147
-------
Figure 1. Complete Airborne Pesticide Sampler
148
-------
Figure 2. Pesticide Collection Module Compartment with
Module Installed
to
-------
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-
west.
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.
150
-------
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-
cides.
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
methodologies.
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
151
-------
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
p,p'-DDE
p.p'-DDT
Mlrex
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
152
-------
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
assessment.
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
153
-------
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.
154
-------
Figure 6
*s
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,
155
-------
" ~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
answered.
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
156
-------
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.
157
-------
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
glycol.
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.
158
-------
IN SITU METHOD FOR ORGANOPHOSPHATE INSECTICIDES
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
02M-/QV-o-p~OEt
OEt
w—CH2-S-P-OCH3
N OCH3
*Stanford Research Institute, Menlo Park, California
159
-------
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
research:
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
-8
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.
160
-------
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 "
S
EtO-^-0-C-HCH, +
I
EtO
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
concentrations.
Figure 4 shows some other approaches we have taken to this problem. Methyl
formamide and tosyl chloride spontaneously generate product. Reactions involving
161
-------
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
i
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.
162
-------
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
Rearrange-
ment
Olt
EtC
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.
163
-------
TOXIC POTENTIATORS AS BY-PRODUCTS OF
ORGANOPHOSPHORUS INSECTICIDES
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
50
"•University of California, Riverside. Presented by Dr. Gunter Zweig
165
-------
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
166
-------
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
OU
an LD n of 4728 mg/kg, a difference of over 60-fold. Tests against four different
50
species of insects showed little, if any, variation between technical and purified
samples.
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
O
II
(CH0O)0PSCHCOOEt
3 u I
C6H5
CHgO
•P-SCHCOOEt
u
C6H5
phenthoate oxon
S
II
(CELO)0PSCHCOOH
O u I
C6H5
S -methyl phenthoate
S
,0)0PS(
(CHgO)2PSCH3
phenthoate acid
O
II .
(CH30)2PSCH3
O,O,S-trlmethyl
phosphorothioate
O, O, S-trimethyl phosphorodithioate
O
II
(CH3S)2POCH3
0,S, S-trimethyl
phosphorodithioate
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
167
-------
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
50
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
materials.
Table 2: Percent Impurities in Technical Phenthoate
% Impurity Rat Oral LD
of Impurity
63
250
4500
450
47
96
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.
Impurity
phenthoate oxon
S-Me phenthoate
phenthoate acid
(CH30)2P(S)SCH3
(CH30)2P(0)SCH3
(CHQS).P(0)OCH0
by TLC
0.03
0.4
1
2.25
0.015
0.0066
by GLC
0.024
0.32
-
2.01
0.019
0.0053
168
-------
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
169
-------
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-
TEPP.
Figure 2
moisture
heat or
prolonged
storage
>
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.
170
-------
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?
171
-------
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.
172
-------
RESEARCH PROGRAM OF THE CHEMISTRY BRANCH
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
173
-------
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
174
-------
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
excreter.
175
-------
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
made.
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-
tabolites.
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.
176
-------
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
177
-------
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.
178
-------
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
water.
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. ROSS: GPC?
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.
179
-------
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.
180
-------
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-
gated.
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
181
-------
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
months.
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.
182
-------
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.
183
-------
SPECIAL INTEREST TOPICS AND GENERAL DISCUSSION
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
185
-------
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
litigation.
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.
186
-------
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.
187
-------
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.
188
-------
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.
189
-------
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
190
-------
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.
191
-------
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,
192
-------
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.
193
-------
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
themselves.
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
194
-------
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-
tector?
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
195
-------
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
methods.
196
-------
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
there.
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 —
197
-------
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.
198
-------
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
199
-------
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.
200
-------
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.
201
-------
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.
202
-------
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.
203
-------
MICROELECTROLYTIC CONDUCTIVITY DETECTOR
Jack B. Dixon*
NOT AVAILABLE
FOR
PUBLICATION
*Department of Entomology, Purdue University, Layfayette, Indiana
205
------- |