-------
Section 7K
this Manual. In addition, TLC may be used as a screening procedure
followed by confirmation and quantitation using GC, or the quantitation
can be carried out by TLC if a gas chromatograph is not available or if
the pesticide of interest is unstable during GC. Extraction, cleanup,
and concentration steps normally precede TLC determination. Often more
stringent cleanup is required for TLC than for GC if streaked zones are
to be avoided. For example, the 15% diethyl ether fraction from the
Florisil column cleanup of a fat sample contains a large amount of lipids.
Although adequate for GC, further cleanup prior to TLC is required (EPA
PAM, Section 12,B,V). TLC has also been occasionally used for cleanup
of extracts prior to determination by GC (100).
Major advantages of TLC are simplicity, rapidity, and low cost. Sensi-
tivity ranges from about 5-500 ng for most pesticide detection methods.
Rapid semi-quantitative estimation can be achieved by visual comparison
of sample and standard spot sizes and/or intensities, and more accurate
quantitation can be carried out by in situ scanning of spots with a
spectrodensitometer.
This section will briefly survey important aspects of TLC for screening
and quantitation of pesticide residues. General techniques of TLC were
described in detail in an extensive treatise (101), while specific
procedures for pesticide TLC were covered in several papers (97, 102).
Applications of TLC to pesticide analysis have been reviewed (103, 104).
7K PRACTICAL CONSIDERATIONS IN TLC
Spots are applied to the thin layer using simple disposable capillaries,
GC syringes, or automatic multiple spotting devices. All initial zones
should be of small, uniform size, and only enough pesticide is spotted
to allow for detection after the run. Care should be taken that the
spotting pipet does not penetrate the surface layer. Standard solutions
must be spotted on the same plate as the sample, preferably on both sides
of the sample spot.
Layers are hand-coated with a commercial adjustable spreading device
(Figure 7-B) or purchased pre-coated on glass, plastic, or aluminum
backing. Analytical layers are usually 250 ym thick. Pre-coated layers
are of high purity and uniformity and are used almost exclusively in
most laboratories, especially for in situ quantitation by densitometry.
Substitution of one brand of adsorbent for another or pre-coated for hand-
coated plates often cannot be directly made. For example, silica gels with
differing polarities or surface hardness (binders) may require modified
solvent systems or detection reagents if similar results are to be
obtained. Pre-coated silica gel plates, especially those prepared with
an organic binder, are generally used as received from the manufacturer
without activation.
Silica gel and alumina layers usually give the best results, but polyamide,
microcrystalline cellulose, kieselguhr, zinc carbonate (105), and magnesium
-265-
-------
Section 7K.
oxide, among other adsorbents, have also been used. For reversed phase
TLC, hydrophobic C,ft chemically bonded silica gel plates are commercially
available. *•*
Figure 7-B Desaga/Brinkmann Adjustable Applicator For Coating
Regular or Gradient Layer Plates, Brinkmann Instru-
ments, Inc. - >>,,,.,,
Chromatography is carried out in a development chamber, most often a
rectangular, glass, paper-lined tank saturated with solvent vapors
(Figure 7-C). Low volume "sandwich" chambers are also used. Both
saturated and unsaturated atmospheres have teen used to advantage and
should be tested for optimum results in any,'particular application.
Ascending development for a distance of 10-20 cm is typical. It is
important to follow exactly all stated conditions when attempting to
reproduce a separation. The temperature, development chamber design
and equilibration, and water content of the adsorbent are probably
the most frequent sources of variation among laboratories.
The technique termed "high performance thin layer chromatography"
(HPTLC) has become increasingly important for separations and in situ
quantitative analysis in the recent past. HPTLC is carried out on
10 x 10 cm, 7.5 x 7.5 cm, or 5 x 5 cm pre-coated layers of silica gel
with a smaller particle size and a narrower particle size distribution
than in conventional TLC plates, thereby giving improved resolution and
sensitivity of detection. Volumes no larger than 1 PI must be spotted
for these advantages to be fully realized. For manual application,
spotting is usually done with a Pt-Ir tipped Nanopipet (or equivalent),
or this type of pipet is used with an automatic spotting device that
controls both the pressure of the pipet tip on the layer and the duration
of contact. Solvent development is carried out in a miniature glass
-266-
-------
Section 7K
rectangular chamber or in a commercial, automatic U-chamber device pro-
ducing radial.zones (106) (a special radial scanner is needed to quanti-
tate these separations). High resolution is achieved rapidly with short
development distances (i.e., 5 cm or less).
In a typical residue analysis, it is virtually impossible to apply the
whole cleaned-up sample extract or an appropriate, accurate aliquot, as
a spot of 1 yl or less, so HPTLC has not yet been widely used for actual
samples. New approaches are appearing that may solve this problem by
allowing a larger sample to be applied without sacrificing the benefits
of the HP layer. One proposed solution utilizes a two-section plate
with a high performance analytical layer above a spotting region; initial
development concentrates the diffusely applied sample into a narrow
zone at the interface of these layers (107). Another possibility is
the use of programmed multiple development (apparatus from Regis Chemical
Co.)» which causes large initial spots to be narrowed during migration
on the HP layer (108). HPTLC plates are available from Merck and What-
man, and HPTLC equipment from Camag, Inc. and Fotodyne Corp.
Figure 7-C. Desaga Rectangular TLC Tanks, Brinkmann Instruments, Inc.
The following solvent systems have proved tov be generally useful for
separation of a wide range of pesticides oh silica gel thin layers:
benzene mixed with varying amounts of ethanol for polar compounds or
with hexane for those which are less polar; and a mixture of hydro-
carbon plus acetone plus chloroform, with the addition of methanol for
more polar pesticides. Examples include pentane-acetone-chloroform
(65:30:5 v/v) or pentane-acetone-methanol-chloroform (70:15:10:5 v/v).
The purpose of the chloroform is to control the evaporation of acetone
in the atmosphere of an unsaturated tank. Proportions of the com-
ponents are changed to suit the requirements of specific separations.
After development and air drying of the layer, spot detection may be
achieved in a number of ways. Few pesticides are naturally colored, but
-267-
-------
Section 7L
colored derivatives may be made prior to spotting, e.g., dyes, formed
from aromatic amine moieties of urea herbicides by coupling with
N-ethyl-1-naphthylamine (FDA PAM, Vol. II, Sec. 120.216). Colorless
spots can be detected by applying a chromogenic reagent, either by
spraying or dipping. A commercial aerosol spray device is shown in
Figure 7-D. Dipping is the preferred method of application, if feasible,
because of the uniformity achieved and the hazards involved in careless
spraying of corrosive, toxic, or carcinogenic reagents. A Thomas-
Mitchell dipping tank is recommended. Sometimes the reagent can be
incorporated in the layer prior to development or included in the
developing solvent. Naturally fluorescent spots can be detected under
short (254 nm) or long (366 nm) wave DV light, or fluorescence may be
induced by application of fluorogenic reagents after development or
preparation of fluorescent derivatives (e.g., dansyl compounds) prior
to spotting (109). Spots that absorb UV light are detected as quenched
(dark) spots on layers containing phosphor activated by UV light (usually
254 nm). Radioactive (labeled) pesticides are detected by autoradiography
and some fungicides are detected by direct bipautography.
Figure 7-D. TLC Aerosol Sprayer, Brinkmann Instruments, Inc.
TL QUANTITATIVE TLC
Quantitatlon of separated spots may be achieved by "eyeball" comparison
between sample and standard spots run on the same plate or by some inde-
pendent analytical method (e.g., spectrophotometry or GC) after scraping
the spot, collecting, and eluting the pesticide from the adsorbent.
Manual elution is simply carried out by scraping the area containing the
pesticide spot, collecting the scrapings in a vial or tube, adding
solvent and agitating (vortexing), filtering the adsorbent, and concen-
trating the filtrate containing the pesticide. An automated elution
system is available from Camag, Inc. (110). Radioactive spots can be
quantitated by scintillation counting after scraping or by automatic
scanning of radioactivity on the layer.
-268-
-------
Section 7L
Colored, fluorescent, or quenched spots may be scanned on the layer when
a spectrodensitometer is available. Quantitation is achieved by scanning
sample and standard spots in the optimum instrumental mode and treating
the resultant peaks, representing the amount of light absorbed or
emitted, in the same manner as GC peaks for calculations. A versatile
densitometer is capable of scanning in single or double beam and re-
flectance or transmission modes, and has monochrometers or filters
for selection of the best wavelengths of incident and emitted (for
fluorescence) energy. Important considerations for densitometry are
adequate extract cleanup (111), precise and accurate spotting, uniform
layers, Rj? values between 0.3 and 0.7, uniform application of detection
reagents, and optimum' use of a good densitometer. •
Fluorescence densitometry .has proven to be the most advantageous mode
for pesticide analysis in terms of sensitivity and selectivity. If
the compound is naturally fluorescent (e.g., benomyl, Maretin, quino-
methionate), the procedure is usually straightforward and measurements
can be made immediately after separation. Sensitivity and reproduci-
bility are usually very high because no reagents are added or sprayed
on the chromatogram to produce background fluorescence.. For the majority
of pesticides that are not fluorescent, however, some kind of treatment
is required. Possibilities for producing fluorescence include treatment
of the plate with heat, acid, base, inorganic salts, or a combination of
these; preparing a derivative in solution before spotting; or applying
a fluorogenic reagent to the layer after separation. All of these
options are included in the papers cited in Table 7-2.
Manual spotting is best performed with 1 or 2 yl Microcap disposable
pipets, using repeated spotting with drying in between for larger volumes.
It has been shown that sample delivery errors below 1% are feasible with
Microcaps (112). Larger volumes of sample extracts are conveniently and
reproducibly spotted with a device such as the Kontes automatic spotter
that applies milliliter volumes of one to six samples or standards in
small, uniform zones with little operator attention. Solutions are
loaded into 5 ml capacity glass tubes and are delivered onto the layer
through Teflon coated needles, the rate of flow and spot size being
controlled by a stream of nitrogen or air focused onto the spotting
location. The Kontes spotter is pictured in Figure 7-L and described
in reference (113). Manual spotting of larger volumes onto commercial
plates with an inert -preadsorbent spotting area is quickly and re-
producibly done with a Drummond digital microdispenser. Samples and •-•
standards, including total extracts, so applied are narrowed to a common,
small initial zone size at the silica gel interface (114).
Pre-coated layers are recommended for quantitative TLC since it is very
difficult to hand-coat layers with adequate uniformity. They are
generally purified before use by a predevelopment with chloroform-
methanol (1:1 v/v) followed by evaporation of the solvent in a dust
free atmosphere. Uniform application of chromogenic or fluorogenic
reagents is better achieved by dipping than by spraying. However,
dipping is not always possible; its use depends on the reagent solvent,
adsorbent, and type of compounds on the layer.
-269-
-------
Figure 7-E.
Section 7L
Fiber Optics Thin Layer Scanner and Automatic Spot Applicator,
Kontes Glass Company, Inc.
When a new densitometric method is developed, the spots of interest should
be scanned in all possible modes and directions and at a variety of wave-
lengths in order to obtain the best signal to noise ratios and selec-
tivities. The optimum conditions are then used to obtain the calibra-
tion curve (linear range) and perform the analysis. Samples and
bracketing standards should always be chromatographed on the same plate.
Thin layer densitometry is capable of precision of 1-2% on a routine
basis and can rival GC and HPLC for determination of certain pesticide
residues in the hands of an experienced operator. A book covering the
principles and experimental details of thin layer densitometry, including
a chapter on pesticide analysis, has been published (115). Table 7-2
contains some recent, selected applications of thin layer densitometry.
A fiber optics scanner specifically designed for pesticide analysis (116)
is available from Kontes (Figure 7-E).
-270-
-------
Section 7L
TABLE 7-2
Compounds
Acidic herbicides
Bayrusil
Benomyl
Benomyl, carben-
dazin, and 2-AB
Captan, captafol
Carbaryl
Carbaryl
Chloranben
Chlorophenoxy
acid herbicides
Coumaphos
Coumaphos
Coumaphos and
0-analog
DDT
DDT
Fenitrothion
Fenitrothion,
breakdown
products, and
related compounds
Gibberellins
PESTICIDES QUANTITATED BY THIN
Sample matrix Scanning mode
.standards only
foods
cucumber
fiuita, vegetables
apple, potato
potato
apples, water,
lettuce
bean, tomato
water
water
water
eggs
water
water
water
standards
apple pulp
fluorescence
fluorescence
fluorescence
quench
fluorescence
visible
visible
visible
visible
fluorescence
fluorescence
fluorescence
visible
visible
fluorescence
fluorescence
fluorescence
LAYER DENSITCttffiTRY
Detection" method
4-bromoethyl-7-
methoxycoumarin
heating
—
fluorescent layers
NaC103
£-nitrobenzenedi-
azonlum fluoborate
p_-nltrobenzenediazo-
nium fluoborate
Bratton-Karshall
reagent
AgH03
heating
heating
heating
AgN03
AgN03
SnCl^fiuorescanlne
fluorescamine
H,SO.
Reference
'(125)
(134)
(144)
(118)
(117)
(135)
(142)
(133)
(122)
(127)
(141)
(131)
(135)
(137)
(121)
(145)
(140) -
-271-
-------
Section 7L
TABL6 7-2 (Continued)
CoirpoundB
SampJe matrix
Scanning mode Detection method
Reference
Glyphosate (via
H-nitroso
derivative)
Herbicide* contain-
ing KHj or OH groups
Karetln
HCPA and
Tcrbicil
OC1 pesticides
OF insecticides
OF pesticides
OF pesticides
Qulnocethlonate
Thiabcndazole
Thiourea
B-Trinzines
Triasines
shoots and
roots
water, soil
milk,, eggs
apples
human autopsy samples
standards only
vater
tissues
crops
fruits
citrus fruits
standards only
vater
fluorescence
fluorescence
fluorescence
UV absofbance
visible
visible
fluorescence
visible
fluorescence
fluorescence
UV absorbance
quench
visible
fluorescamine
dansyl chloride
AgN03
AgNOj or enzyme
inhibition
hydrolysis/dansyl
chloride
palladium chloride
fluorescent layers
iodine
(138)
(119)
(129)
(143)
(123)
(126)
(128)
(124)
(130)
(132)
(136)
(120)
(139)
Earlier analyses are reviewed by J. D. MacKell and R. W. Frei in J. Chromatogr. Sci., 13, 279 (1975).
-272-
-------
Section 7H
7M THIN LAYER SYSTEMS
a. Chlorinated Pesticides
Extracts of fatty and nonfatty foods cleaned-up on a Florisil
column are chromatographed on prewashed alumina layers developed with
heptane (for the 6% diethyl ether-petroleum ether Florisil eluate) or '
2% acetone in heptane (15% diethyl ether fraction). Detection is pro-
vided by spraying with AgN03-2-phenoxyethanol reagent in ethanbl or
acetone and exposing to high intensity short-wave UV light to produce
brown to purplish-^black spots. The construction of a DV light
apparatus containing four 15 watt lamps for rapid color development
and allowing a variable distance between the TLC plate and the light
source is described in the Canadian PAM, Section 14.10* Thin layer
media must be very low in chlorine content, and other precautions and
care must be taken to prevent large areas of the plate from turning
brown or gray, thereby reducing the contrast of the spots with back-
ground. A sensitivity in the 5-500 ng range is possible with AgNOs
reagent, with a light steaming before spraying often aiding the detection.
Conventional 20 cm x 20 cm glass plates, commercial pre-coated TLC sheets,
or 3-1/4 inch by 4 inch microslides, may be employed. Complete details of
these methods plus Ry values for numerous compounds in the aforementioned
two solvent systems, as well as for an alternative system consisting of
immobile dimethylformamide on alumina and isooctane as the developing
solvent, are given in Sections 410, 411, and 413 of the FDA PAM. Silver
nitrate has been incorporated into acid-washed alumina before the plates
are coated so that only exposure to UV light is required for spot visualiza-
tion (FDA PAM, Section 412). The AgN03 detection method has recently been
studied in detail for the determination of chlorinated insecticides and
herbicides (146).
0 • • •
Similar TLC procedures are described in detail in the EPA PAM, Section
12,B for the determination of chlorinated pesticide residues in. serum
and adipose tissue. An extract from 50 g of serum, cleaned-up on Florisil
and concentrated to 100 yl before spotting, will produce a visible spot
at 2 ppb, assuming that 10 ng of pesticide is detectable. An adipose
tissue extract from a 5 g sample, concentrated to 500 Ul,. will give a
readable spot at 10 ppb. The method involves TLC of the 6% and 15%
Florisil column eluates as above, with additional prior cleanup of the
15% fraction on an alumina micro-plate developed with acetonitrile.
Silica gel layers developed with hexane, 1% acetone-hexane, 10-50%
benzene-hexane, or 1% ethanol-hexane are recommended for screening
chlorinated pesticides in foods at 0.1 ppm levels (Canadian PAM, Pro-
cedures 9.1 and 12.4). Complete details of plate preparation, extract
concentration, and visualization with silver nitrate reagent are given
in this source, along with figures of spot locations for eleven common
pesticides in four mobile solvents.
-273-
-------
Section 7M
For complex pesticide mixtures, tvo dimensional or multiple development
techniques may be helpful. The former was used to identify organochlorine
pesticides in blood and tissues (147) and the latter (148) for the separa-
tion of 13 common pesticides.
Extensive listings of additional solvent systems, corresponding Rp values,
and detection reagents for chlorinated pesticides will be found in
references (103, 123, 149, and 150). In reference (123), 26 solvent
systems and 14 chromogenic reagents are evaluated for the determination
of 12 organochlorine pesticides in blood, urine, and tissue samples.
b. Organophosphorus (OP) Pesticides
Cleaned-up extracts may be developed with methylcyclohexane on DMF-
coated alumina layers and detection made by spraying with tetrabromo-
phenolphthalein ethyl ester, AgN03, and citric acid. This reagent reacts
only with thiophosphoryl compounds to give blue or magenta spots (FDA
PAM, Section 431; EPA PAM, Section 12,B). Thio and nonthio organophos-
phates are developed on silica gel layers with isooctane-acetone-chloro-
form (70:25:5 v/v) and .detected as blue or magenta spots by treatment
with 2^nitrobenzyl pyridine and tetraethylpentamine spray (FDA PAM,
Section 432).
A two dimensional procedure (IDA PAM, Section 614.11; 151) has the
significant advantage of specificity, obtained by bromine vapor oxidation
of the OP pesticides before development in the second direction. Silica
gel layers with toluene, 25 percent heptane in ethyl acetate, or ethyl
acetate as developing solvents were used along with the Storherr charcoal
column cleanup procedure and enzymatic detection with commercial ho-rse
serum cholinesterase and indoxyl acetate to identify 18 pesticides in
crops at 0.01 ppm levels. A sandwich type chamber is specified for
development to obtain the requisite resolution and sensitivity. The same
procedure should be well suited to OP pesticides in human and environ-
mental samples after appropriate cleanup.
Enzyme inhibition techniques are important for the selective and sensitive
(pg-ng amounts) detection of enzyme inhibitors such as OP and carbamate
insecticides and metabolites. These compounds inhibit esterases and
thereby prevent hydrolysis of a chromogenic substrate. Procedures include
separation by TLC (on silica gel layers sometimes as thick as 450 Jim),
optional treatment with bromine vapor or UV light, and spraying of the
layer with enzyme and substrate solutions. Areas corresponding to
inhibitors are visible as white spots on an Intensely colored background;
i.e., inhibited enzyme is surrounded by enzyme free to hydrolyze the sub-
strate and thus produce color. While many OP pesticides are inhibitors
per se, bromine or UV treatment is required to convert others to active
inhibitors. For carbamates, UV or bromine treatment may produce no change
or increased or decreased inhibition, depending on the compound. Sample
extracts often require minimal cleanup prior to TLC analysis with enzymatic
detection; for example, hexane extracts of many foods can be directly
chromatographed. Section 9.2 of the Canadian PAM provides procedural
-274-
-------
Section 7M
details, tables of sensitivities and effects of bromine and UV treat-
ment for.OF and carbamate pesticides, and diagrams of mobilities with
hexane-acetone (8:2 v/v), a generally useful development solvent for
TLC on 450 Vim silica gel layers. The preparation of these layers is
detailed in Section 12.4 of the Canadian PAM. Several different
esterases have been compared for the detection of 65 OF and carbamate
pesticides in vegetables and fruits (152).
TLC enzyme inhibition methods and applications to pesticides have been
reviewed (153-155) as have the merits of TLC for analysis of residues
(156). The separation and detection of 42 phosphate compounds using
five ternary solvent systems on three adsorbents and three selective
chromogenic sprays have been reported (157). Twenty five solvent
systems and several visualization reagents were evaluated for detection
of 12 OF insecticides in tissues (124).
c. Chlorophenoxy Acid Herbicides
Extracts containing methylated chlorophenoxy acids are cleaned-up on
a Florisil column and chromatographed on alumina layers using hexane
saturated with acetonitrile as the developing solvent. Cleaned-up
extracts containing free acids are developed for a distance of 3.5 cm
on a pre-coated silica gel sheet with cyclohexane->acetic acid (10:1 v/v),
then the sheet is dried and developed for 15 cm in the.same direction
with benzene-petroleum ether (3:1 v/v). Spraying with silver nitrate
. chromogenic reagent produces black spots with a sensitivity of ca 50 ng
for the esters and 100-500 ng for the free acids. Details of both
methods and % values are given in Sections 421 and 422 of the FDA PAM.
Other detection reagents for these pesticides include Rhodamine B and
Bromocresol green indicators (158).
d. Other Pesticide Classes
The TLC of other classes of pesticides including carbafflates, ureas,
phenols, dithiocarbamates, triazines, and organomercurials was reviewed
in references (103) and (104). Applications, solvent systems, detection,
and quantitation are covered in these references. TLC is particularly
applicable to herbicides, many of which are polar and not susceptible
to gas chromatographic analysis without derivative formation. Studies
have been reported for the TLC of triazine herbicides on silica gel
(120, 159) and polyamide (160); determination of 11 urea herbicides in
water (161); detection of dithiocarbamate fungicides with Congo red (162);
separation of carbamate and phenylurea pesticides on polyamide (163);
comparison of six reagents for detection of carbamate and phenylurea
pesticides on EPTLC plates (164); and separation and identification of
carbamate pesticides in post mortem material (165).
The TLC of five dithiocarbamate residues in chloroform extracts of leaves
is detailed in Section 9.3 of the Canadian PAM. Silica gel layers developed
-275-
-------
Section 7N
with benzene for dimethyldithiocarbamates or acetic acld-methanol-benzene
(1:2:12 v/v) for ethylenebisdithiocarbamates are used, with detection as
yellow, brown, or green spots after a cupric chloride-hydroxylamine
hydrochloride spray.
7N REFERENCES
(1) Eisenbeiss, F., and Seiper, H., J. Chromatogr.. 83. 439 (1973).
(2) Larose, R. H., J. Assoc. Off. Anal. Chen., 75_, 1046 (1974).
(3) Summermann, W., Rohleder, H., and Korte, F., Z. Lebensm. Unters.
Forsch., 166. 137 (1978).
(4) Mittelstaedt, W., Still, G. G., Durbeck, H., and Fuhr, F., J. Agr.
Food Chem.. 25,, 908 (1977).
(5) Snyder, L. R., J. Chromatogr. Sci.. 15, 441 (1977).
(6) Baumann, F., Pecsar, R. E., Wadsworth, R., and Thompson, B., Amer.
Lab., p. 97, September (1976).
(7) Horvath, C., Melander, W., Molnar, I., and Molnar, P., Anal. Chem.,
14, 2295 (1975).
(8) Majors, R. E., J. Chromatogr. Sci.. 15, 334 (1977)..
(9) Saunders, D. L., J. Chromatogr. Sci.. 15, 372 (1977).
(10) Bradley, T., Spectra-Physics Chromatography Review. 4^(3) , 8 (1978).
(11) Horvath, C., and Melander, W., J. Chromatogr. Sci.. 15, 393 (1977).
(12) Greenwood, J. V... Spectra-Physics Chroma tography. Review. J3 (2), 1 (1977),
(13) Snyder, L. R., and Saunders, D. L., J. Chromatogr. Sci.. 7_t 195 (1969).
(14) Snyder, L. R., J. Chromatogr.. 92. 223 (1974).
(15) Bakalyar, S. R., Amer. Lab., p. 43, June (1978).
(16) Dlebold, G. J., and Zare, R. N., Science. 196. 1439 (1977).
(17) Rucki, R. J., Talanta. Z7, 147 (1980).
(18) Kissinger, P. T., Anal. Chem.~. 49(4). 447A (1977).
(19) Lores, E, M., Bristol, D. W., and Moseman, R. F., J. Chromatogr. Sci.,
16, 358 (1978). ''—
-276-
-------
Section 7N
(20) Willmott, F. W., and Dolphin, R. J. , J. Chromatogr. Sci., 12,
695 (1974).
(21) Ettre, L. S.,' J. Chromatogr. Sci.. 16, 396 (1978).
(22) Julian, B. G., Vandenbom, H. W., and Kirkland, J. J., J. Chromatogr.,
112, 443 (1975).
(23) Krull, I. S., Goff, U., and Ashworth, R. B., Amer. Lab., p. 31,
August (1978).
(24) Chamberlain, W. J., Walters, D. B., and Chortyk, 0. T., Anal. Chin
Acta, 2i» 213 (1975).
(25) Rabel, F. M.. Amer. Lab.. p. 81, January (1980).
(26) Kirkland, J. J. «• Chromatographia. £, 661 (1975).
(27) Tseng, P. K., and Rogers, L. B., J. Chromatogr Sci., 16, 436 (1978).
(28) Berg, J., Varian Instrument Applications, 11(2), 5 (1977).
(29) Bakalyar, S. R., and Henry, R. A., J. Chromatogr.,. 126. 327 (1976).
(30) Lawrence,- J. F., and Turton, D., J. Chromatogr., 159. 207 (1978).
(31) Pribyl, J., and Herzel, F., J. Chromatogr., 166, 272 (1978).
(32) Conlon, R. D., Anal. Chem.. 41(4). 107A (1969).
(33) Munk, M. N., J. Chromatogr. Sci.. 8. 491 (1970).
(34) Veening, H., J. Chem. Educ.. 41(9),A549 (1970).
(35) Taggart, W. P., Industrial Research, p. 76, February (1974).
(36) Karasak, F. W., Research/Development, 24_(6), 52 (1973).
(37) Brown, P. R., High Pressure Liquid Chromatography, Academic Press,
N.Y., 1973.
(38) Parris, N. A., Instrumental Liquid Chromatography, Elsevier
Scientific Publishing Co., N.Y., 329 pp. (1976).
(39) Jackson, M. T., Jr., and Walker, J. Q., J. Chromatogr. Sci.. 15,
490 (1977).
(40) Creed, C. G., Research/Development. 27(9), 40 (1976).
(41) Ivie, K., Chapter 2 in Analytical Methods for Pesticides and Plant
Growth Regulators, Vol. XI, Zweig, G. and Sherma, J., eds., Academic
Press, NY (1980).
-277-
-------
Section 7N
(42)
(43)
(44)
(45)
(46)
(47)
(48)
(49)
(50)
(51)
(52)
(53)
(54)
(55)
(56)
(57)
(58)
(59)
(60)
Moye, H. A., J. Chromatogr. Sci.. 13, 268 (1975).
Schooley, D. A., and Quistad, G. B., Prog. Drug Metab.r 3_, 1-113 (.19 9).
Self, C., McKerrell, E. H., and Webber, T. J. N., Proc. Anal. Div.
Chem. Soc.. 12, 288 (1975). ~~"~"—~~
Hoodless, R. A., Sidwell, J. A., Skinner, J. C., and Treble, R. D.,
J. Chromatogr.. 166, 279 (1978).
Werner, W., GIT Fachz. Lab.. 22_, 785 (1978).
McFadden, W. H., Bradford, D, C., Games, D. E., and Gower, J. L.,
Amer. Lab., p. 55, October (1977); McFadden, W. H., Schwartz, H. L.,
and Evans, S., J. Chromatogr.. 122, 389 (1976).
Arpino, P. J., and Guiochon, G., Anal. Chem.. 51(7). 682A (1979).
Vestal, M. L., NBS Spec. Publ. 519 (Trace Org. Anal.; New Front.
Anal. Chem.), 647 (1979).
Oehler, D. D., and Holman, G. M., J. Agr. Food Chem.. 23, 590 (1975).
Moye, H. A., and Scherer, S. J., oral presentation at the 91st AOAC
Meeting, October 17-20, 1977, Washington, DC, Abstract No. 79.
Lawrence, J. F., and Leduc, R., J. Agr. Food Chem., 25 » 1362 (1977).
Aten, C. F., and Bourke, J. F., J. Agr. Food Chem.. 25. 1428 (1977).
Farrow, J. E., Hoodless, R. A., Sargent, M., and Sidwell, J. A.,
Analyst. 102, 752 (1977).
Mtmdy, D. E., and Machin, A. F., J. Chromatogr.. 139. 321 (1977);
Mundy, D. E., Quick, M. P., and Machin, A. F., J. Chromatogr..
121, 335 (1976).
Paschal, D. C., Bicknell, R., and Dresbach, D., Anal. Chem.. 49,
1551 (1977). —
Bjornsson, T. D., Blaschke, T. F., and Meffin, P. J., J. Pharm.
Sci., 66, 142 (1977).
Kvalvag, J., Ott, D. E., and Gunther, F. A., J. Assoc. Off. Anal.
Chem.. .60, 911 (1977). ~"~
Farrington, D. S., Hopkins, R. G., and Ruzicka, J. H.'A., Analyst.
102, 377 (1977). —
Selim, S., Cook, R. F., and Leppert, B. C., J, Agr. Food Chem.. 25.
567 (1977). a ~ —
-278-
-------
Section 7N
'(61) Byast, T. H., J. Chromatogr., 134. 216 (1977).
(62) Fasco, M. J., Piper, L. J., and Kaminsky, L. S., J. Chroma togr.. 131,
365 (1977).
(63) Lawrence, J. F.. J. Agr. Food Chen., 25, 211 (1977).
(64) Brinkman, U. A. Th., DeKok, A., DeVries, G., and Reymer, H. G. M.,
J. Chromatogr., 128, 101 (1976).
(65) Lawrence, J. F., J. Assoc. Off. Anal. Chem., 59, 1066 (1976).
(66) Maeda, M., and Tsuji, A., J. Chromatogr.. 120, 449 (1976).
(67) Pryde, A., and Darby, F. J., J. Chromatogr.. 115, 107 (1975).
(68) Pfeiffer, C. D., J. Chromatogr. Sci.. 14, 386 (1976).
(69) Byast, T. H., and Cotterill, E. G., J. Chromatogr., 104, 211 (1975).
(70) Smith, A. E., and Lord, K. A., J. Chromatogr., 107, 407 (1975).
(71) Whitacre, D. M., Atallah, Y. H., Forrette, J. E., and Suzuki, H. K.,
in Analytical Methods for Pesticides and Plant Growth Regulators
Volume X, Zweig, G. and Sherma, J., eds., Academic Press, OT, p. 367
(1981).
(72) Cabras, P., Meloni, M., and Pirisi, F. M.. J. Chromatogr.. 176, 473
(1979).
i
(73) Ernst, G. F., and Verveld-Roeder, S. Y., J. Chromatogr., 174, 269 (1979)
(74) Krause, R. T., J. Chromatogr. Sci., 16_, 281 (1978).
(75) Mbring, S. E., and McChesney, J, D., J. Assoc. Off. Anal. Chem., 62,
774 (1979).
Wong, L. T., Gnanaprakasam, S., and Thomas, B. H., J. Chromatogr.,
135, 149 (1977).
(76)
(77)
(78)
(79)
(80) Pflugmacher, J., and Ebing, W., J. Chromatogr., 151, 171 (1978).
Thean, J. E., Fong, W. G., Lorenz, D. R., and Stephens, T. L.,
J. Assoc. Off. Anal. Chem., 61, 115 (1978).
DiPrima, S. J., Cannizzaro, R. D., Roger, J.-C., and Ferrell, D.,
J. Agr. Food Chem., 26_, 968 (1978).
Lawrence, J. F., and Leduc, R., J. Assoc. Off. Anal. Chem., 61, 872
(1978).
-279-
-------
Section 7N
(81) Newsome, W. H., J. Agr. Pood Chen,. 26_, 1325 (1978).
(82) Bowman, M. C., Holder, C. L., and Bone, L. I., J. Assoc. Off. Anal.
Chem.. ,61, 1445 (1978). ~
(83) Lawrence, J. F., and Leduc, R., J. Chromatogr.. 152, 507 (1978).
(84) Ott, D. E., J. Assoc. Off. Anal. Chem.. 62, 93 (1979).
(85) Cochrane, W. P., and Lanouette, M., J. Assoc. Off. Anal. Chem.. 62,
100 (1979). —
(86) Brealey, C. J., and Lawrence, D. K.. J. Chromatogr.. 168, 461 (1979).
(87) Moye, H. A., and Hheaton, T. A., J. Agr. Food Chem.. .27(2), 291 (1979).
(88) Crathorne, B., and Watts, C. D., J. Chromatogr.. 169, 436 (1979).
(89) Stoeber, I., Z. Anal. Chem.. 293. 370 (1978); Stoeber, I., and
Reupert, R., Vom Wasser, 51, 273 (1978).
(90) Otsuki, A., and Takaku, T., Anal. Chem.. J51(7), 833. (1979).
(91) Rus, V., Rev. Chim. (Bucharest). 29(11), 1088 (1978).
(92) Ramsteiner, K. A., and Hoerman, W. P., J. Agr. Food Chem., 27(5)
934 (1979).
(93) Lord, K. A., Cayley, G. R., Smart, L. E., and Manlove, R., Analyst.
105, 257 (1980). ' *—^
(94) Mills, P. A., J. Assoc. Off. Agric. Chem.. 42, 734 (1959); 44, 171 (1961).
(95) Mitchell, L. C., J. Assoc. Off. Agric. Chem.. 40, 999 (1957).
(96) Sherma, J., and Zweig., Paper Chromatography. Academic Press, NY,
Chapter 12 (Pesticides), pp. 355-396 (1971).
(97) Getz, M. E., Advances in Chemistry Series 104. ACS, Chapter 8, (1971).
(98) Getz, M. E., Residue Reviews. 2.* 9 (1963).
(99) Coffin, D. E., J. Assoc. Off. Anal. Chem.. 49, 1018 (1966).
(100) Heatherington, R. M., and Parouchais, C., J. Assoc. Off. Anal. Chem..
53, 146 (1970). ~~ "
(101) Stahl, E., Thin Layer Chromatography. 2nd ed., Springer-Verlag, NY (1969)
(102) Kovacs, M. F., Jr., J. Assoc. Off. Agric. Chem.. 46, 884 (1963): 47,
1097 (1964). ~
-280-
-------
Section 7N
(103) Sherma, J., Chapter 1 In Analytical Methods for Pesticides and
Plant Growth Regulators, Vol. XI, Sherma, J., and Zweig, G., eds.,
Academic Press, NY (1980); also Chapter 1 in Volume VII of the
same series (1973).
(104) Sherma, J., Critical Reviews of Analytical Chemistry, CRC Press,
pp. 333-338, August (1973).
(105) Reichling, J., 2. Anal. Chem., 281, 139 (1976).
(106) Vitek, R. K., and Kent, D. M.. Amer. Lab., p. 71, January (1978).
(107) Halpaap, H., and Krebs, K.-F., J. Chromatogr.. 142, 823 (1977).
(108) Faber, D. B.. J. Chromatogr.. 142, 422 (1977).
(109) Lawrence, J. F., and Frei, R. W., Chromatogr. Rev., 18, 253 (1974).
(This is a review of theory, factors affecting emission, and appli-
cations of fluorometric derivatization for pesticide analysis by
HPLC and TLC.)
(110) Vitek, R. K., Seul, C. J., Baier, M., and Lad, E., Amer. Lab..
p. 109, February (1974).
(Ill) Sherma, J., Sample Preparation for Quantitative TtC, in Thin Layer
Chromatography; Quantitative Environmental andjClinical Applica-
tions, Touchstone, J. C., and Rogers, D., eds., Wiley-Interscience,
NY, in press.
(112) Emanuel, C. F., Anal. Chem.. 45, 1568 (1973).
(113) Getz, M. E., J. Assoc. Off. Anal. Chem.. 54, 982 (1971).
(114) Sherma, J., Amer. Lab., p. 105, October (1978).
(115) Touchstone, J. C., and Sherma, J., Densitometry j.njlh.i'a Layer
Chromatography. Wiley-Interscience, NY (1979).
(116) Beroza, M., Hill, K. R., and Norris, K. H., Anal. Chea., 40, 1608
(1968).
(117) Francoeur, Y., and Mallet, V., J. Assoc. Off. Anal. Chem., _60_, 1328
(1977).
(118) Polzhofer, K., Z. Lebensm. TJnters. Forsch., 163, xu9 (1977).
(119) Pribyl, J., and Herzel, F., Z. Anal. Chem.. 286. 95 (1977).
(120) Jork, H., and Roth, B.. J. Chromatogr.. 144, 39 (1977).
(121) Berkane, K., Cassie, G. E., and Mallet, V. N., J. Chromatogr.. 139_,
386 (1977).
-281-
-------
Section 7N
(122) Sherma, J., and Koropchack, J., Anal. Chim. Acta, 91, 259 (1977).
(123) Tewarii S. N., and Sharma, I. C., J. Chromatogr.. 131, 275 (1977).
(124) Tewari, S. N., and Harpalani, S. P., J. Chromatogr.. 130. 229 (1977).
(125) Dunges, W., Chroma tographia. j), 624 (1976).
(126) Stefanac, Z., Stengl, B., and Vasilic, Z., J. Chromatogr.. 124,
127 (1976).
(127) Volpe, Y., and Mallet, V. N., Anal. Chim. Acta. 81, 111 (1976).
(128) Lawrence, J. F., Renault, C., and Frei, R. W., J. Chromatogr.. 121,
343 (1976).
(129) Zakrevsky, J. G., and Mallet, V. N., J. Agr. Food Chem.. 23, 754
(1975) . — —•*.
(130) Francoeur, Y., and Mallet, V. N., J. Assoc. Off. Anal. Chem., 59,
172 (1976). ~~—
(131) Zakrevsky, J. G., and Mallet, V. N., J. Assoc. Off. Anal. Chem.. 58,
,554 (1975).
(132) Ottender, H., and Hezel, U., J. Chromatogr.. 109, 181 (1975).
(133) Sherma, J., and Touchstone, J. C., Chromatographia, _8, 261 (1975).
(134) Zakrevsky, J. G.,. and Mallet, V. N., Bull. Environ. Contain. Toxicol.,
13, 633 (1975). ~~~
(135) Sherma, J., Klopping, K., and Getz, M. E., Amer. Lab.. p. 66,
December (1977).
(136) Mandrou, B., Brun, S., and Kingkate, A., J. Asspc. Off. Anal. Chem.,
60, 699 (1977). =""
(137) Sherma, J., and Bloomer, K., J. Chromatogr.. 135, 235 (1977).
(138) Young, J. C., Khan, S. U., and Marriage, P. B., J. Agr. Food Chem.,
25, 918 (1977).
(139) Wislowska, E., and Kostowska, B., Chem. Anal.(Warsaw). 22, 975 (1977).
(140) Winkler, V. W., in Analytical Methods for Pesticides and Plant Growth
Regulators, Vol. X, Zweig, G., and Sherma, J., eds., Academic Press,
NY, pp. 545-559 (1978).
(141) Mallet, V. N., and Volpe, Y., J. Chromatogr.. 97., 415 (1978).
(142) Sherma, J., Kovalchick, A. J., and Mack, R., J. Assoc. Off. Anal.
Chem., 61, 616 (1978).
-282-
-------
Section 7N
(143) Polzhofer, K., Z. Lebensm. Uhters. Forsch., 167,- 162 (1978).
•
(144) Masao, M., Yasuo, I., and Hiroaki, H., Bull. Aerie. Chem. Insp.
Stn., 17, 43 (1977).
(145) Zakrevsky, J. G., and Mallet, V. N., J. Chromatogr., 132, 315 (1977).
(146) Ludwick, A. G., Lau, A. N. K., and Ludwick, L. M., J. Assoc. Off.
Anal. Chem., 60, 1077 (1977).
(147) Eliakis, C. E., Coutselinis, A. S., and Eliakis, D. C., Analyst, 93.,
368 (1968).
(148) Szokolay, A., and Madaric, A., J. Chromatogr.. 42, 509 (1969).
(149) Walker, K. C., and Beroza,.M., J. Aasoc. Off. Agric. Chem., 46, 250
(1963).
(150) Thielemann, H., Z. Chemie Lpz.. 14, 292 (1974).
(151) Gardner, A. M..'J. Assoc. Off. Anal. Chem.. 54., 517 (1971).
(152) Ernst, G. F., Pieterse, C., and Marterns, L. J. H., J. Chromatogr..
133. 245 (1977).
(153) Mendoza, C. E., Residue Reviews. 43, 105 (1972) and 50_, 43 (1974);
J. Chromatogr.. 78. 29 (1973).
(154) Garner, A. M., FDA By-Lines, 2(3), 173, January'(1972)„
(155) Villeneuve, D. C., Advances in Chemistry Series 104. ACS, Chapter
3 (1971).
(156) Abbott, D. C., and Thompson, J., Residue Reviews. 11, 1 .(1965);
Abbott, D. C., and Egan, H., Analyst. 92, 475 (1967).
(157) Getz, M. E., and Wheeler, H. G., J. Assoc. Off. Anal. Chem.. 51, 1101 (1968)
(158) Thielemann, H., Z. Anal. Chem., 272, 286 (1974).
(159) Van den Heede, M., and Heyndricks, A., Meded. Fac. Landbouwvet.
Rilksuniv. Gent. 41, 1457 (1976) (Analytical Abstracts, abstract
5G20, November, 1977; Chem. Abst. 86. 84568c, (1977).
(160) Reichling, J., and Fischer, H., Z. Anal. Chem.. 115_, 670 (1975).
(161) Deleu, R., Barthelemy, J.-P., and Copin, A., J. Chromatogr.. 134.,
483 (1977).
(162) Ragab, M. T. H., Anal. Lett., £, 295 (1976).
-283-
-------
Section 7N
(163) Reichling, H., and Fischer, H., Z. Anal. Chem.. 276, 301 (1975),
(164) Davies, R. D., J. Chromatoer.. 170, 453 (1979).
(165) Tewari, S. N., and Singh, R., J. Chromatogr.. 172, 528 (1979).
-284-
-------
Section 8
SAMP LING, EXTRACTION, AND
PROCEDURES IN PESTICIDE ANALYSIS
This section treats a .number of miscellaneous topics important in residue
analysis. These include general considerations for collection and ex-
traction of extracts. Specific procedures for extraction and cleanup of
pesticide and metabolite residues are discussed in Section 9.
SAMPLE COLLECTION, PREPARATION, AND STORAGE
8A GENERAL CONSIDERATIONS IN SAMPLING
Special considerations must be given to the procurements storage, and trans-
portation of samples to be analyzed for pesticide residues. Procedures
should ensure, as well as possible, that the pesticides originally present
have not undergone degradation or concentration and that potentially inter-
fering impurities have not been added. Plastics must be rigidly avoided
as containers for samples to be examined by electron capture GC because
minute traces of materials such as polyethylece may produce spurious re-
sponses. Similarly, metal containers may contain trace impurities such as
oil films, lacquers, or rosin from soldered joints that will cause inter-
ference in GC analysis. In general, glass jars or bottles with aluminum
foil or Teflon-lined lids are the most suitable sample containers, although
it is sometimes possible for pesticides in stored extracts to be adsorbed
onto the glass surfaces. Glass containers should be carefully precleaned
as outlined in Subsection 3L in Section 3. Aluminum foil can be cleaned
by agitating it in analytical reagent grade acetone followed by several
rinsings with pesticide grade ethyl acetate and hexane. Plastic containers
may be used, if necessary, only when non-interference with the subsequent
analysis has been proved at its limit of detection. Important variables
in the sampling and storage processes include the size of the sample,
source, stability, contamination, intended use, behavior of the pesticides,
and the temperature and time of storage.
Readers interested in a more exhaustive discussion of sampling and storage
procedures than provided in the following sections of this chapter are
referred to the publication "Guidelines on Sampling and Statistical
Methodologies for Ambient Pesticide Monitoring" (Monitoring Panel, Federal
Working Group on Pesticide Management, Washington, DC, 1974), This 60-page
-285-
-------
Section 8B, 80
manual contains chapters on .statistics and study design, air, soil, the
hydrologic environment, estuaries, fresh water fish, wildlife, foods and
feeds, and human tissues. The 20-page booklet entitled "Guidelines for
Sampling and Transporting Samples for Pesticide Residue Analysis" (Federal
Interdepartmental Committee on Pesticides Check Sample Program, London,
Ontario, Canada, April, 1979) contains detailed information on dry feeds,
plants and soils, food products, wildlife, tissues, forest substrates,
water, and fish. The influence of sampling methods on residue analytical
results, sampling criteria, and statistics of sampling data have been de-
scribed (1).
8B REPRESENTATIVE VS. BIASED SAMPLING
Samples collected for the purpose -of assessing tolerance infringements> such
as with agricultural and food products, should be random and representative.
To the contrary, most environmental samples are deliberately chosen to be
biased in nature. For example, a sample of water to be analyzed for the
highest possible pollution in a stream or lake would best be "taken as a grab
sample from the point of maximum pollution introduction (such as an effluent
pipe from a factory) rather than from the center of the river where it might
be most representative. If,, on the other hand, the objective is an average
residue profile of the entire body of water, the final sample would preferably
be a composite of a number of subsamples taken at various locations and water
depths. Analysis of a sick bird or fish in the middle of a metabolic cycle
would usually be more useful for determining any pesticide contamination than
a dead specimen that is likely to contain only metabolites. Similarly, human
stomach washings (lavage) taken at an early stage are more likely to contain
parent pesticides and to be useful for indication of pesticide poisoning.
It is important that the analyst be aware of these considerations and that
he be consulted when the sample to be collected is decided so that it is
valid for the purpose of the analysis and valuable time is not wasted on a
worthless sample.
8C SAMPLE CONTAINERS
Section 2 of the EPA Pesticide Analytical Manual specifies suitable sample
containers for various sample types. These include wide mouth glass bottles
with Teflon or aluminum foil lined screw caps for autopsy tissue samples of
less than 25 g, glass vials of at least 7 ml capacity for blood (avoid rubber
or cork caps), empty pesticide-grade solvent bottles for water samples, and
pint or quart capacity mason jars for larger environmental or agricultural
samples. Sample collection glassware should be scrupulously cleaned as out*-
lined in Section 3 of this Manual. Special precautions must be taken in
preparing glass containers and caps and taking samples for PCP analyses
because of the ubiquity of the chemical. These are outlined in Section
5,A,(A),Ca),IV of the EPA PAM. Specimens intended for organoehlorine com-
pound analysis are never wrapped directly in paper, cardboard, or plastic.
-286-
-------
Section 8C
It is common practice in some laboratories to wrap tissue or other samples
in aluminum foil prior to analysis. Figure 8-A, part a, shows a gas chro-
matogram of a pentane rinse of the shiny side of commercially available
aluminum foil. The amount of rinse injected corresponded to 2 sq. cm of
foil. The GC conditions included the use of a 10% OV-210 on Gas-Chrom Q
column and a 63Ni EC detector, typical of those used for analysis of pesti-
cide's and PCBs. Part b shows the corresponding rinse of the dull side of
the same foil. In general, the amount of interfering material was found
to vary with the brand and lot of foil. However, the risk of contamination
from this source dictates that aluminum foil not be used for packaging
samples without a thorough acetone prerinse (2).
Figure 8-A. Gas chromatogram of pentane rinse of aluminum
foil on OV-210 column with 63Ni electron capture
detector. Amount injected corresponds to 2 sq. cm
of foil, a « shiny side of foil; b « dull side of
foil (2).
-287-
-------
Section 8D
It is good procedure to clearly label collected samples with all pertinent
information such as a code number, date and time of collection, type of
sample, place and method of collection, description of collection site, size
of sample, etc. All samples that are perishable are shipped to the laboratory
in styrofoam containers with dry ice. A detailed description of systematic
procedures used for receiving, numbering, and storing environmental samples
at the National Monitoring and Residue Analysis Laboratory, Gulfport, MS,
has been published (3). A strategy for documenting the chain of custody
of samples that has the potential for being used as evidence in a legal
proceeding or agency enforcement action is detailed in Section IV of the
EPA National Enforcement Investigations Center Pesticide Product Laboratory
Procedures Manual (see Section 3E of this Manual).
8D SAMPLE COMPOSITING
After collection of a valid gross sample, compositing or reduction to an
analytical size sample may be required, especially for agricultural and food
samples. The general requirement is that the small analytical sample must
be fully representative of the gross sample collected. The exact steps in
the compositing procedure will depend on the particular sample involved.
Figure 8-B shows typical steps in reduction of a gross sample of an agri-
cultural product collected in the field, during processing, or at the'market.
Figure 8-B. Typical steps in reduction of a gross sample
Remove peal or husk (if necessory) and
reduce size of large units by cutting or chopping
1
Peel or husk
i
JMix and quarte
|
Subsample
sib
0
1
Fiesh or kernel!
|
Mix and quarter
i
Subsample
6lb
if necessary, reduce
size of large units by
cutting or chapping
I
| Mix and quarter I
Subdivide
Subdivide
Subdivide
-288-
-------
Section 8E
8E STORAGE OF SAMPLES
As a general rule, samples should be analyzed as soon as possible after
their collection. If storage is necessary, it should be under prescribed
conditions that preserve the integrity of the original sample. Samples other
than water are ordinarily stored in a freezer below 0°C, but, even then,
physical and chemical changes may occur in either the sample or in the
residues sought. Because many pesticides are photodegradable, it is
advisable to protect samples and any solutions or extracts from needless
exposure to light.
Tissue samples that are to be extracted within 24 hours may be held at normal
refrigerator temperature (+2 to +4°C). If extraction is not to be carried
out within this time, the. samples should be deep frozen at -12 to -18°C. If
tissues are stored in a "self-defrosting" freezer in unsealed containers,
the weight can markedly decrease due to desiccation. If the tissues were
not weighed prior to freezing, or if they are to be subdivided at a later
time, this desiccation may make it impossible to relate the amount of sub-
stance determined analytically to its original concentration in the tissue
(2). A related problem occurs when samples experience repeated freezing
and thawing. Adipose tissue in particular has a tendency to "leak" lipid
when the cell membranes are disrupted by a freeze-thaw cycle. In a series
of experiments in which such cycles were deliberately applied to a collection
of samples of adipose tissue from a rat, the apparent lipid content of the
tissue (mg per gram of tissue) decreased by an average of 10% after three
freezings. This loss was only apparent, and was not observed if the tissue
was extracted in the original storage container (2).
Blood samples that are to be separated for subsequent analysis of the serum
should be centrifuged as soon as possible after drawing. If the serum is
to.be analyzed within a 3-day period, storage at +2 to +4°C is suitable. If
storage is to be for longer periods, it is preferable to deep freeze at
-12 to -18°C. Otherwise, DDT may degrade in contact with broken red blood
cells (hemoglobin).
Agricultural or environmental samples that are to be analyzed for organo-
phosphates should be placed in tight containers and stored in deep freeze as
soon as possible after sampling unless sample preparation is to be conducted
within a very few hours. No difference was found in measured residue levels
for a series of OP pesticides when food samples were extracted immediately or
after storage at -17°C for several months (4).
Water samples should be extracted at once, if all all possible, or stored in
the dark at 4°C to avoid rupture of the container as a result of freezing.
Pesticides can be adsorbed on the glass container during storage, so the
container should be rinsed with solvent if the extraction is not made in the
container itself. For carbamates, the sample is acidified immediately after
collection with sulfuric acid and 10 g of sodium hydroxide are added for
each liter of sample. Maximum storage is 24 hours for all compounds except
chlorinated hydrocarbons, which can be held for up to 30 days.
-289-
-------
Section 8F
Whole fish can be stored for up to six months if an even temperature of at
least -26°C is maintained with a good glaze on the sample and rapid initial
freezing. Homogenized samples require less storage space, but these samples
should be monitored for stability of the compounds of interest if held
longer than one month.
If lengthy storage is required prior to analysis, a good alternative to
storage of sample is to extract the sample at once, remove most or all of
the solvent, and store the extract at a low temperature. Decomposition in
samples that must be stored can be evaluated by storing spiked controls
along with the samples. Organophosphorus pesticides field-extracted with
chloroform from water were successfully preserved for three weeks upon re-
frigeration. Of the 16 compounds tested, only EPN and malathion were not
stable (5).
If freezing is not possible, wildlife and fish samples may be preserved in
formalin or alcohol. Because analytical results are usually in terms of
wet weight, the wet or "fresh" weight of the sample before it was preserved
should be recorded, as well as the volume of preservative used in each jar.
Specimens preserved in formalin or alcohol must be accompanied by a "control"
jar. This jar must contain the same mixture used in preserving the specimens,
and must be prepared (i.e., rinsed and sealed) in the same manner as the jars
containing specimens. This may not be equivalent to freezing for storage
of samples, however. For example, Abate was partially converted to Abate
sulfoxide in fresh samples stored in formalin or formalin plus 5% acetic
acid, but not in frozen samples (6). Formaldehyde should be checked for the
presence of PCB contamination prior to use as a sample preservative (7).
Comments pertinent to collecting samples of 'different types will be pre-
sented in the Subsections 8F to 8K. Methods, for the analysis of the various
sample types are surveyed in Section 9 of this Manual.
8F SAMPLING OF AGRICULTURAL AND FOOD PRODUCTS
Procedures for sampling, sample preparation, sample compositing, and sample
reporting, as required by Federal law, for all commodity types are outlined
in detail in Sections 140-143 of the FDA Pesticide Analytical Manual, Volume
I. Section 3 of the Canadian Department of National Health and Welfare
Analytical Manual for Pesticide Residues in Foods gives guidelines for
systematically obtaining representative samples of processed and packaged
foods, bulk foods, and field crops and for handling, shipping, and storing
samples. Recommended minimum sizes are tabulated for different samples, with
a general sample requirement of n product units, where n equals the square
root of the total but need not exceed 10-15 separate units.
Section 4 of the same Canadian PAM covers laboratory preparation of analytical
samples from gross samples of fresh, frozen, and canned vegetables, fruits,
and juices; dry cereal grains, flakes, dehydrated fruits and vegetables;
animal tissues; eggs; butter and margarine; milk and cream; cheese and nuts;
fats and oil; and fish and fish products.
-290-
-------
Section 8G
It is suggested that readers interested iii analysis of sample substrates
of this type for legal compliance to tolerance levels should refer to these
two excellent -sources of information. Sampling methods for trace organic
analysis of foods have also Been described by Horwitz and Howard (8) . If
the purpose of an analysis is to obtain information on maximum residue
levels in a particular situation, biased sampling would be used, e.g., the
lower perimeter of fruit would be sampled from certain trees most likely
to have received a higher dose of pesticide spray.
8G SAMPLING OF BIOLOGICAL MATERIALS
Adipose tissue, blood, and urine samples from live and autopsy animal and
human subjects are commonly analyzed for pesticide residues. The amounts
of sample required, the time of collection, and the compound to be detected
are determined by the nature of the pesticide (s) of interest. Pesticides
that degrade or .are metabolized readily may be absent in a particular
sample, but their original presence can be deduced by determination of
metabolites such as alkyl phosphates from OP pesticides, phenols from
chlorophenoxy acid herbicides or carbamate insecticides, or DBA from DDT.
If body tissues or fluids are analyzed quickly in cases of high exposure,
the chance of finding the parent pesticide is greatly enhanced. If exposure
is low or a long time has elapsed after exposure, the analyst must be
familiar with pesticide metabolism in order to choose appropriate samples
and metabolites to determine. For example, the highest concentration of
organophosphorus pesticide urinary metabolites will be found from four to
eight hours after the donor's exposure (EPA PAM, Section 6, A, 2, (a) ,V). When
concentrations of pesticides or metabolites are expected to be small, samples
must be larger, e.g., morning urine samples or 24-hour pooled specimens.
The majority of human adipose tissue samples are taken during autopsy by an
attending physician. Samples should be placed in a clean glass container
with a foil- lined (never rubber- or cardboard -lined) screw cap. The aluminum
foil should be prerinsed with acetone. Plastic bags or bottles must be
avoided since they can contribute traces of impurities such as phthalates to
the sample, causing spurious GC peaks when the final concentrate is examined
by EC-GC or GC-MS (EPA PAM, Section 5,A,(l),(a) ,V). Up to 2% of radiolabeled
DDT was found to be lost by "extraction" into the plastic when liver samples
were stored in polyethylene bottles at 4°C overnight. This radioactivity was
not removed when the bottles were washed, so that the loss for one sample could
constitute a contamination for the next sample stored in the same bottle (2) .
Whole blood samples are. transferred to glass vials with Teflon or foil lined
screw caps* and the required serum aliquot is removed after a period of
settling in a refrigerator and subsequent centrifugation. Serum 'is stored
in a refrigerator at 2-5°C if the analysis is to be performed within 24 hours
or in a deep freeze (-15 to -25°C) for longer storage periods. The analysis
of chlorinated pesticides is not adversely affected by such storage for
periods up to six months (EPA PAM, Section 5 ,A, (2) , (a) »
-291-
-------
Section 8H
8H AIR SAMPLING
The EPA has in the past operated a nationwide air monitoring program in
order to gather information on the extent of human exposure to airborne
pesticides. This program utilized Greenburgh-Smith impingers containing
ethylene glycol for trapping organophosphorus and halogenated hydrocarbon
insecticides both in the vapor phase and as dusts. The air was drawn
through the impingers by means of a vacuum pump, the amount sampled (cu. m)
being.controlled by means of a flow meter and timer. The ethylene glycol
sampling procedure did not prove acceptable in terms of,convenience or
reliability, and the EPA national air sampling program was discontinued.
However, the ethylene glycol impinger continues to be used by some labora-
tories in local monitoring programs (3, 9, 10).
Robert G. Lewis of the U.S. EPA Health Research Effects Laboratory has .
written an extensive review of sampling methods for airborne pesticides"(11).
Included are discussions of many types of accumulative samplers (e.g., im-
pactors, bubblers, liquids supported on solid substrates, polymer foams,
etc.), reactive samplers, continuous and sequential samplers, and grab
samplers. Recent reports of pesticide recovery from air have included the
use of tubes containing XAD-2 resin for trapping 2,4~D acid, and its ester
and amide derivative .with 86-96% efficiency (12); hexylene glycol contained
in glass scrubbers for recovery of dieldrin and heptachlor at 0.1 ng/cu. m
(13); and XAD-2 resin for organothiophosphates (14).
One of the most promising approaches to the sampling of air involves use of
polyurethane foam. The updated review of air sampling methods in Section
8,A of the EPA PAM contains a discussion of this method in addition to other
approaches and apparatus recommended by the EPA for high volume ambient air
sampling, indoor air sampling, crop re-entry monitoring, and workplace and
personnel monitoring. Polyurethane foam vapor traps following a particle
filter have been evaluated (15) for sampling of pesticides, PCBs, and poly-
chlorinated naphthalenes. Collection rates up to 360 cu. m of air per 24
hours and sensitivities as low as 1 cu. m for some compounds can be achieved.
The filters and plugs were Soxhlet extracted with hexane-ethyl ether (95:5 v/v)
at 4 cycles per hour for 16-24 hours, and OC1 pesticides were determined by
EC-GC after alumina column cleanup and OP pesticides by FPD-GC without clean-
up. Collection was generally satisfactory but was poor for the more-volatile
OC1 compounds. Recovery was ca 75% for OP pesticides. It was shown that a
second trap in series with the first did not necessarily improve recovery
values. The collection of dieldrin, lindane, trifluralin, dacthal, chlordane,
and heptachlor on polyurethane foam was studied and optimum plug size and
shape for any chosen sampling rate were given. Trapping efficiency depended
on pesticide vapor pressure and the flow rate of air. The quantitation
limit was ca 0.1 ng/cu. m in a 5 cu. m air sample. It was crucial that the
plugs were carefully protected from laboratory contamination and Soxhlet ex-
tracted with pesticide-grade acetone and hexane prior to use if clean blanks
and highest sensitivity were to be achieved (16). Other collection efficiency
data for pesticides and PCBs are included in Section 8,B of the EPA PAM.
Because of possible pesticide degradation, air sampling apparatus should
be shielded from light during sample collection.
-292-
-------
Section 81
81 WATER SAMPLING
The design, of a comprehensive pesticide sampling program for environmental
waters is a specialized topic that is covered in publications available
from the Water Quality Control Division of the USEPA, National Environ-
mental Research Center, Cincinnati, Ohio. Important considerations include
the objective of the study, frequency of sampling, location of sampling
stations as related to hydrologic conditions, and the 'selection of sampling
methods. The following is a brief review of some important selected
factors in a sampling program.
a. Grab Samples
Water can be collected by taking one instantaneous ("grab") sample from
a given location, directly filling the sample container. The usual technique
is to submerge the container a few inches below the water surface during
filling to avoid skimming off any floating film that would be least repre-
sentative of the vertical water column. Several collections should be taken
at various depths and locations to provide a more representative sample.
Care should be exercised to avoid disturbing bottom sediment. Discrete
samples from various depths can be obtained with standard samplers consisting
of a metal outer container with a glass sample collection bottle inside
(e.g*., Precision and Esmarch samplers, EPA PAM, Section 10,A,II). Grab
sampling is often sufficient for lakes, reservoirs, etc., that are not subject
to rapid transitional changes.
Grab samples less than 2 liters are collected in wide mouth glass bottles,
and samples of one gallon or more in the glass bottles in which pesticide
quality solvents are supplied. All bottle caps should be Teflon lined. The
sample size is dictated primarily by the expected residue levels, the sensi-
tivity of the analysis, and the need to run duplicate, spiked, and background
analyses. A 500-1000 ml sample may suffice from water where pesticide levels
are expectedly high, while 2 liters or more may be needed for a surveillance
program where no high levels are anticipated. Rainwater is collected in
clean glass containers rather than metal or plastic. Samples should include
information that will help the analyst choose a proper analytical method and
interpret the results. This includes the location of sampling, depth, sus-
pected contaminants, type of sample (surface water, waste discharge, etc.),
and agricultural activity or spills in the immediate area or upstream.
Many pesticides are unstable in water, so samples should be analyzed as soon
as possible after collection, ideally within a few hours. If this is im-
practical because of distance from the sampling site to the laboratory and/or
the laboratory work load, store the sample in a refrigerator or freezer.
Samples being examined solely for organochlorine residues may be held up to
a week under refrigeration at 2 to 4°C with no adverse effect. Those samples
to be analyzed for organophosphorus or carbamate pesticides should be frozen
immediately after drawing the sample because of rapid degradation in aqueous
media (Table 1, Section 10,A of the EPA PAM shows data for the degradation
rate of 29 pesticides in water at ambient temperature in sealed containers
(I?)). pH adjustment is required for some samples immediately after collection
(e.g., adjustment to pH 2 with sulfuric acid for phenoxy acid herbicides).
Holding time and storage conditions must be reported along with the analytical
results and corrections made if rates of pesticide degradation are known.
Exposure of samples to sunlight should be avoided.
-293-
-------
Section 81
Every effort should be made to perform the solvent extraction step at the
earliest possible time after sampling, irrespective of the classes of
pesticides suspected of being present. Especially unstable pesticides can
be extracted immediately in the field. The resulting extracts can be safely
stored for periods up to three or four weeks at -15 to -20°C before pro-
ceeding with subsequent cleanup and determinative steps. One disadvantage
of glass sample bottles is possible breakage in shipment, and care should
be exercised in proper packaging to avoid this. Another disadvantage is
the already-mentioned possibility of pesticide adsorption on glass surfaces.
Reduced recovery (>90Z to 46-681) of DDT in water analysis upon storage has
also been noted due to adsorption on suspended matter in the sample (18).
The assumption made Is that a grab sample is at least representative of the
immediate water mass from which it was taken and somewhat representative of
the water that will pass the sampling point during some limited future time
interval. The grab sample is amenable to use in both random and nonrandom
sampling programs. The number,, frequency, and distribution of samples
collected will depend on the study objectives and the variability within the
"population" being sampled.
After sampling, pesticides are extracted from water, cleaned up and concen-
trated as necessary, and determined by GC or an alternative method. Pesti-
cides in clean water (e.g., drinking water) can be detected at 5-500 ppt
levels by electron capture GC without the need for extensive extract clean-
up. Impurities in "dirty" samples will require additional cleanup steps,
and background problems will cause difficulty in analyzing these low levels
accurately.
b. Continuous Samplers
Continuous and automatic devices are often used for sampling flowing
bodies of water such as rivers and streams. Activated carbon filters, have
been widely used for adsorption of pesticides and other kinds of organlcs
in natural waters since they were developed and Introduced by the tijS, Public
Health Service in 1951 (19). The technique involves passage of a continuous,
constantly controlled volume of water through a column of activated carbon
followed by desorption by means of elution or by Soxhlet extraction with a
suitable solvent or combination of solvents. The variable efficiency and
consistency of pesticide adsorption and desorption from the adsorbent prior
to determination, ease of contamination with extraneous organic substances,
and bacterial and oxidizing attack on the sorbed pesticides have caused
problems with carbon columns (20, 21).
filter materials which have been recommended as alternatives to carbon for
collection of pesticides (usually chlorinated insecticides) from natural
waters include reversed liquid-liquid partition systems (a hydrophobia phase
coated on a support) and other adsorbents. Carbowax 4000 (5 g) and ii-undecane
(22), silicones chemically bonded to diatomaceous earth support (23),
covalently bonded aromatic and alkyl chlorosilanes on Celite (24), porous
polyurethane foam columns (for pesticides and PCBs) (25, 26), polyethylene
film (20-25 pm thickness) (27), and polyurethane foam coated with selective
adsorbents (28) have all been used with varying success.
-294-
-------
Section 81
The XAD macroreticular adsorbent resins (XAD-1, -2, -4, and -7) have.been
used to collect organlcs from both potable (29, 30) and sea (31) water.
Optimum conditions for use with XAD-4 resin were found to be 2 g of adsor-
bent, a flow rate through the resin of 8 ml/minute, and 100 ml hexane-
diethyl ether (10:1 v/v) as elutlng solvent. Among 10 chlorinated
insecticides studied, only aldrin and p,p'-DDE were not quantitatively re-
covered, and recovery of PCBs was 76% (32). Details for use of XAD-2 and
-4 resins for many classes of trace organic water contaminants have been
published (33) and recoveries between 81 and 96% were reported for 20 ppt
levels of atrazine, lindane, dieldrin, DDT, and DDE (47% for aldrin). An
EPA report (34) recommends XAD-2 resin for routine monitoring of sea water
for chlorinated insecticides and PCBs. Average recovery for XAD-2 ex-
traction of fortified natural waters collected across Canada was 85% for
the 10-100 ng/liter levels of ten OC1 pesticid.es (recovery of mirex was
unacceptably low) and 82% for 250 ng/liter levels of PCBs; blanks from the
resin were a low 4 ng PCBs/liter (35). Concentrations as low as 0.1 ppt of
PCBs and organochlorinated pesticides were detected by recovery from water
on small XAD-2 columns (36), and ng levels of carbamates were recovered
(86-108% at 0.01-1 ppm levels) with the same adsorbent (37). Amberlite
XAD-4, porous polyurethane foam, and undecane plus Carbowax 4000 on Chromo-
sorb were comparable for extracting ten OC1 insecticides from environmental
water samples (22).
Continuous liquid-liquid extractors are an alternative to the filter-adsor-
bent .processes preferred by some analysts. A multi-chamber extractor with
internal solvent renewal replenishing (38) allowed extraction of 135 liters
of water at rates of 0.5-1.0 liter/hour and recovered greater than 97% of
ppb levels of pesticides. Subsequently, a similar modified apparatus per-
mitted use of both heavier- and lighter-than-water solvents (39). A simple
and rugged field version of the Kahn and Wayman apparatus (38)" excluded
solvent recycling and was based on mixed settling (40). -This apparatus,
which consisted of an extraction unit, magnetic stirrer, and pump, provided
quantitative recovery of pesticides and PCBs at levels of 0.1-1.0 ng/liter
of river water.
More recently, a similar in situ apparatus designed to solvent-extract large
amounts of sea and river water continuously while situated at a desired depth
at the sampling site has been described (41). A Teflon helix, continuous
liquid-liquid extractor, plus a continuous evaporative concentrator re-
covered pg to ng per liter amounts of OP pesticides from river and sea water
or from secondary sewage effluent with >80% efficiency (42). A comparative
study of recoveries from river water by continuous extraction and activated
carbon filters showed that the recoveries were similar but the former was
less costly (43).
The theory for extracting chlorinated pesticides continuously from water
with a stationary immiscible solvent is discussed in reference (40).
See Section 10,A,III-V of the EPA PAM for updated material on water sampling.
-295-
-------
Sections 8J, 8K
8J SAMPLING OF HOUSE DUST, SOIL, AND STEEAM BOTTOM SEDIMENT
House dust is collected with a vacuum cleaner, air dried, and sieved prior
to analysis. Soil is sampled By collecting cores or Borings of a known
diameter cut to a depth of ca. 3-4 inches or more from the centers of plots
1 sq. m In size. Ten to twenty cores representing a surface area of at
least 200 sq. m are recommended. The first 2 inches of core, containing
the grass or crop cover and roots, are separated from the underlying soil.
Corings representing each layer of soil are combined, quartered, and
divided into 2 IB samples for analysis. Soils are analyzed in an air-dry
state after sieving to remove foreign material. Another reported procedure
for soil sampling (3) involves collection of cores 1-3 inches deep and
3 inches in diameter with a hand-operated auger; on a 1/4 acre site
C10S feet x 105 feet), sampling Begins 7.5 feet from the Border of the site,
and a core is collected every 15 feet until 7 cores are oBtained. The
process is repeated along parallel lines separated By 15 feet from the
original sampling line, until a total.of 49 cores are collected. The cores
are sieved through a hardware cloth screen into a 3 gallon galvanized pail
and thoroughly mixed. The sample is transferred to two one-half gallon cans
with lids for shipment to the laBoratory. There is no way to collect a
truly representative soil sample, and reproduclBility of results on different
samples taken from the same area is often expectedly poor.
Sediment from the Bottom of a Body of water provides information concerning
the degree of pollution resulting from pesticides, particularly those that
are not readily degradaBle. This information comBined with residue .data on
the water and resident Biological life gives an overall pesticide cohtaminar
tion profile of the Body of water. Bottom sediment varies with respect to
Both particle size composition (surface adsorptive power) and organic content.
Therefore, sample sites should Be selected at random in an effort to collect
samples representing a rang'e of variation. In some cases consultation with
an oceanographer can indicate where one would Be likely to find the maximum
amounts of pollution from considerations such as currents and industrial
effluent discharges.
ABout a quart of sediment is a typical sample size. Actual collection is
accomplished "i±th one of a variety of core samplers or dredges. A, diagram
of a dredge-type device for collecting sediment samples has Been puBlished
(3). The dredge is thrown into the water at least 10 times to collect
samples, which are transferred each time to a. galvanized pail. The total
sample is mixed and transferred to one-half gallon cans (with a hole in each
lid to release any gas Buildup from organic matter- in the sample) for ship-
ment to the laBoratory. A simple Bottom sediment collector composed of a
steel can attached to the end of an aluminum pole has also Been described
(44). Samples may Be preserved with formalin or a variety of other steri-
lants provided they do not affect the analyses to Be run. Samples are air
dried and ground prior to analysis. They are stored, if necessary, in a
freezer if volatile compounds such as 2,4-D ester may Be present.
8K MARINE BIOLOGICAL AND WILDLIFE SAMPLES
A problem sometimes encountered when collecting plankton and Bottom
organisms is oBtaining the minimum weight necessary for successful analysis.
-296-
-------
Section 8L
As a general mle, a minimum of about 10 g will be required. Collected
organisms can be frozen at once or preserved with 5-10% formalin or 70%
ethanol, prepared with distilled water rather than the water from which the
collection was made. This eliminates the possibility of pesticides In the
water concentrating in the organisms over a period of time. Any added pre-
servative must be extracted and analyzed to determine if exchange of pesti-
cides from the organisms to the preservative has occurred.
Sufficient masses of plankton are collected by use of a tow net behind a
boat or by pumping water through a net. Bottom fauna are collected with
dredges or dip nets. Samples are washed through a screen and organisms
are hand picked from the remaining debris.
Fish are collected utilizing seines, gill nets, traps, electrocution de-
vices, otter trawls, or angling. Wrapping the fish in aluminum foil and
preservation by .quick freezing in dry ice is most desirable. When this is
not possible, liquid preservatives are used. Larger fish should be injected
with preservative from a syringe to prevent decomposition of internal organs.
Fish stored in formalin plus 5% ^28303 showed no loss of Abate (temephos)
residues (>1 ppb) for up to three weeks (45).
Fish can be analyzed whole to yield data on gross contamination, or the fish
can be sectioned to obtain information on edible and non-edible parts.
Analyses of individual organs and tissues yield information on distribution
of pesticides in the"fish. Analysis of blood from a dying fish may be
valuable for determining probable cause of death where pesticide exposure is
suspected. The blood is obtained by cutting the tail at the caudal peduncle
and collecting and freezing the blood in a small vial.
Invertebrate samples are collected in pitfall traps, as described by Wojicfc
et -JL!. (46). Bird samples are collected using Japanese mist nets placed
near a water source or in a cove where the net Is not visible. Traps baited
with peanut butter or some other foodstuff are employed for sampling mammals.
These traps, which are available in a variety of sizes, have a trap door that
closes when the animal enters to take the food. Non-crop vegetation samples
are obtained with shears, sickles, pocket knives, etc., usually from the
same sampling area as soil samples. All of these samples are sorted, wrapped
in aluminum foil with the shiny side out, tagged, and placed in a plastic
bag for shipping (3).
Some of the material In the sections on sampling was adapted from an EPA
training course manual (47). A review that Includes some of the above
sampling procedures and additional methods for collection of environmental
samples has been published (3).
8L CONTROL OF PROCEDURES FOR EXTRACTION OF RESIDUES
Specific procedures for the extraction and cleanup of pesticide multi-
residues In many sample types are surveyed in Section 9 of this Manual.
This subsection discusses general considerations of pesticide extraction
from collected samples.
-297-
-------
Section 8L
Many solvents are employed for extracting residues, depending on the
polarity of the pesticide and the amount of co-extractives expected from
the particular substrate. Solvents range from hexane or petroleum ether
for nonpolar organochlorine and organophosphorus compounds to methylene
chloride (dichloromethane) for polar carbamates. Chloroform, diethyl ether,
ethyl acetate, benzene, acetonitrile, methanol, acetone, and various "two-
and three-component mixtures of these have all been widely used. Addition
of acid to the organic solvents may aid extraction of acidic pesticides such
as 2,4-D herbicide. Acetonitrile is an excellent general purpose extraction
solvent for low fat-content samples (acetonitrile plus ca 35% water for low
moisture samples), and hexane/acetonitrile systems are widely recommended
for partition cleanup.
Although it has been shown in some cases that recovery of pesticides from
tissues during extraction does not necessarily correlate with recovery of
lipids, it is usually desirable to use an extraction solvent that will
quantitatively extract lipids with the pollutants for reporting purposes
(2). A study (48) has compared the recovery of lipid by nine solvent
mixtures from human adipose tissue for pesticide determination. Extraction
procedures should always be validated for each class of compounds in each
type of sample matrix to which it is applied. In addition to the nature of
the analyte, the toughness, water content, and lipid content of the sample
matrix will Influence the effectiveness of a given extraction procedure (2).
Different techniques are employed for bringing, the extraction solvent and
sample into contact. The best extraction is obtained, in general, by
achieving the most intimate contact between the two, although the type of
residue is an important distinction. When emulsions result from vigorous
shaking or mixing during extraction procedures, centrifugation will usually
be effective In separating solvent layers. A surface residue can usually
be extracted by a simple washing procedure, while the more common internal
residues can be extracted only after fine maceration of the sample. Some
soil samples tenaciously bind pesticides and require long periods (e.g.,
8 hours) of Soxhlet extraction rather than shorter periods of blending as
is common with plant materials. Blending of the sample plus solvent in a
Waring blender (Figure 8-C), Omni-Mixer, Wiley mill, or Hobart food chopper
is probably the most usual extraction procedure in use today, especially
for biological, plant, and food samples. Some additional sample subdivision,
such as cutting, chopping, or grinding usually precedes the blending operation.
A 5 minute period of blending at a moderate speed is typical for many samples.
A special device for aiding formation of a homogenous sample has been des-
cribed (49). The device, consisting of a handle and shaped aluminum sheet,
fits inside a blender jar and serves to gently push bulky samples into the
cutting blades during the blending operation. A liquid-nitrogen cooled
freeze grinder for biological materials containing labile pesticides has
also been devised (50).
Blending with a solvent followed by filtering or centrifuging is particularly
efficient for most vegetable samples. The water in the sample may give
-298-
-------
Section 8L
Figure 8-C. Waring Aseptic Dispersail Model AS-1 (Shown on 702-CR Base)
rise to emulsions with nonpolar solvents, and this can often be avoided by
use of a drying agent such as anhydrous Na2S04 or 2-propanol together with,
or before, the solvent. Meat samples containing too much connective tissue
for a blender to deal with effectively should be first comminuted by a
grinder. Simple heating of minced sample in a beaker on a steam bath with
solvent can be effective, possibly after grinding the sample with Na2S04
and sharp sand to help break down some connective tissue. More volatile
pesticides (e.g., lindane) might be lost in this way.
A comparative study of the efficiencies in the extraction of carbofuran
from radishes was made using three blenders, a Polytron ultrasonic homo-
genizer, a Lourdes blender, and a Waring blender. The least efficient
blender extracted 90% as much as the most efficient, and all three were con-
sidered useful for accurate pesticide analysis (51).
In some cases, more exhaustive extraction of residues from difficult samples
can be obtained by Soxhlet extraction for periods up to 12 hours or longer
with a solvent such as methanol-chloroform (1:1 v/v) (52). Soxhlet thimbles
may require exhaustive extraction prior to use so they do not contribute
-299-
-------
Section 8L
interferences to the analysis (53). Preliminary steps such as drying,
grinding, or chopping normally precede Soxhlet extraction, but care must
be exercised since some pesticides have been shown to be unstable in the
presence of homogenized samples (54). Even Soxhlet extraction may not give
complete extraction in all cases, and only studies with samples to which
radioactive tracers have been applied can indicate the absolute extraction
efficiency in any particular case. The usual evaluation procedure of
spiking a sample with pesticide and looking for quantitative extraction
is less reliable than the radiotracer method because the spiked chemical
will not be naturally incorporated in the same matrix as would the tracer.
Radiotracers are not always available or feasible to use, however. The
most important factor in preparation of a valid spiked sample to accurately
indicate recovery of endogenous compound may be the solvent in which the
spike is dissolved. In one study of the extraction of mirex from fish
muscle, recovery varied from 41 to 89% with a common extraction procedure
but different spiking solvents (2).
Water samples (100-500 ml) are generally extracted by shaking with an
appropriate solvent (3 x 100 ml) in a separatory funnel (55). Soils are
extracted by a variety of methods such as shaking, soaking, blending,
Soxhlet or Goldfisch extraction, or refluxing. -Two 15 minute extractions
in an ultrasonic generator were found comparable to a 24 hour Soxhlet
extraction for removal of s-triazine herbicides from fortified soils (56),
and a 30 second extraction technique using a Brinkmann Polytron ultrasonic
generator gave better recoveries of several chlorinated insecticides from
soil than did 8 hours of Soxhlet extraction (57).
An apparatus that simultaneously Soxhlet extracts pesticides and concen-
trates the resulting extract has been designed (58). Advantages of this
cyclic extraction-evaporation system are that distillation of solvents prior
to extraction can often be omitted, and excess solvent is re-utilized for
extraction.
The most efficient solvent and parameters for extraction of pesticides from
water can be determined using the ^-values originally suggested by Beroza
and co-workers for use in residue confirmation (Subsection 10F in Section
10). The £-value is the fraction of total pesticide.that is distributed
into the nonpolar phase of an equivolume immiscible pair of solvents. This
approach was used to study the extraction of OP pesticides from water (59),
and the best solvents were benzene, ethyl acetate, or diethyl ether for
diazinon and diazoxon at pH 7.4, ethyl acetate for malathion at pH 6, and
diethyl ether or ethyl acetate for fenthion (Baytex) at pH 3.4. ^-Values
can also be used to theoretically select water-to-solvent ratios and the
optimum number of extractions for maximum recovery of a pesticide in water
(60). As a practical example (61), diethyl ether or ethyl acetate was
found best for extraction of 2,4-D acid and esters and benzene for 2,4,5-T
-300-
-------
Section 8M
acid and esters. A 99% recovery of 2,4-D from one liter of aqueous solution
was obtained by a two stage serial extraction with 200 ml and 50 ml of ethyl
acetate under conditions predicted by p_-values.
8M CONTROL OF METHODOLOGY FOR CONCENTRATION OF SAMPLE SOLUTIONS AND
FRACTIONAL COLUMN ELUATES
The concentration of cleaned-up sample in the injection or spotting solution
is one important factor that determines if sufficient residue is available
for detection by GC, LC, or TLC. The analyst must determine this and con-
centrate final solutions according to the least sensitive pesticide in the
method's scope.
Purified extracts or eluate solutions containing even somewhat volatile com-
pounds are concentrated with minimum losses to a volume of ca 5-10 ml using
a Kuderna-Danish evaporative concentrator flask fitted on top with a 3-ball
Snyder reflux column and a collection tube on the bottom (Figure 8-D).
Figure 8-D. Kuderna-Danish; J
Evaporative Concentrator, Kontes
Glass Co. No. K-570000.
The tube is heated in steam water bath in a hood. The apparatus should be
mounted or held so the lower rounded flask surface is bathed in steam. Flasks,
which range in size from 125-1000 ml, should be initially charged with
40-60% of their nominal volume, and the column should be pre-wet with ca 1 ml
of solvent before beginning concentration to prevent possible initial small
loss of pesticides. Refluxing is continued until the final concentrate is
-301-
-------
Section 8M
collected in the lower tube. Boiling chips are required for smooth operation
of the K-D evaporator, and carborundum, checked for absence of contamination,
is recommended in preference to porcelain, vanadium, or glass chips, A
Snyder column modified By putting a distillation trap below the Vigreaux
bubble condensing system increased the degree and consistency of recovery
of nanogram amounts of HCE isomers upon Kuderna-Danish evaporation (.62).
For concentration from 5 ml to smaller volumes (as low as 50-100 pi), the
concentrate is cooled, the collection tube is removed from the K-D flask,
and a fresh chip is added. A micro-Snyder reflux column (Figure 8-E) is
fitted directly to top of the tube, and evaporation is begun by holding the
bottom of the tube in a steam or hot water bath. Evaporation is continued,
with care to avoid bumping, to slightly below the desired volume. The tube
is withdrawn from the water when boiling agitation becomes too vigorous;
immersion and withdrawal are alternated based on observation of boil agita-
tion. The apparatus is cooled 3-5 minutes, and condensate is allowed to
drain down into the tube before the column is removed. The sides of the
tube and column joint are rinsed with solvent to avoid hang-up of pesticides
on upper glass surfaces. A 1-2 ml syringe is useful for performing this
rinse. Finally, further fresh solvent is added to dilute up to the desired
volume, if necessary.
A special rack that simultaneously agitates and evaporates solutions in six
concentrator tubes fitted with micro-Snyder columns in a time equal to a
single tube is described in the EPA FAM, Section 5,A,3,a.
Figure 8-E. Semi-Micro Kuderna
Danish Apparatus, Kontes Glass
Co., No. K-569250.
-302-
-------
Section 8M
Extracts containing fats, oils, or plant extractives, or purified extracts
to which "keeper solution" has been added, can be evaporated on a rotating
vacuum type evaporator with the water bath at, or just slightly above,
room temperature (Figure 8-F). A double-reservoir rotoevaporation vessel
facilitating collection, concentration, and final volume calibration of
column eluates and eliminating a number of manual transfer steps has been
designed (63).
Figure 8-F. Rotary Evaporator,
Kontes Glass Co., No. K-570160
Extracts contained in a beaker or a centrifuge tube immersed in a water
bath at 40°C can be evaporated under a stream of nitrogen adjusted to cause
gentle depression on the surface of the solution. The nitrogen should be
passed through well maintained scrubber tubes to remove contaminants that
could cause pesticide degradation. Warming a tube by holding it in the
hand is a useful, gentle evaporation aid during nitrogen blow-down.
Figure 8-6 shows the Organomation Associates, Inc., N-Evap apparatus
that is widely used for evaporation by nitrogen blow-down.
An evaporation assembly combining an evaporative concentrator tube, a
Ruderna-Danish flask, and a rotary vacuum evaporator (Figure 8-F) is shown
in Section 10,A of the EPA PAM, Figure 1. The concentrator tube is not
immersed in a high temperature water bath as usual, but rather in a 35°C
water bath to minimize degradation of heat labile pesticides. This apparatus
confines the concentrated extract to one container, thereby eliminating the
need for transfer. One hundred ml of methylene chloride can be reduced to
5 ml in ca 20 minutes with a vacuum of 125 mm of mercury.
-303-
-------
Section 8M
Figure 8-G. Model 111 12-position
N-Evap* apparatus, Organomation
Associates, Inc., Northborough, MA.
Another multitube apparatus for nitrogen evaporation is available from
Kontes Glass Co. (64). Concentration is rapid until the solution reaches
0.5-1.0 ml, at which point evaporation slows markedly because this last
volume is below the heating zone of the evaporator block. Thus, losses
of pesticides from inadvertent evaporation to dryness (65) are avoided,
and a minimum of analyst attention is required (Fig. 8-H).
Figure 8-H. Ebullator,
Kontes Glass Co.
-304-
-------
Section 8M
A distillation column is fixed on top of the tube holding the sample, and
small bore stainless steel needlestock tubing is fitted through the column
down into the tip of the tube to direct a stream of micro bubbles of nitro-
gen through the solution to Initiate and maintain ebullation. Recoveries
of seven chlorinated pesticides after concentration for 2 hours in this
apparatus were greater than 94% with both hexane and benzene solvents.
It is important to avoid pesticide loss or decomposition during evaporation
steps. Numerous reports have been made (e.g., 65, 66) of severe pesticide
loss during concentration steps, even in the presence of sample coextrac-
tives. There was no correlation between the amount of coextractives and
evaporative losses, but apparently the nature of the coextractives may be
important. In most situations, organochlorine and organophosphorus pesti-
cides can be concentrated to small volumes without loss ,by the K-D evapo-
rative procedures described at the beginning of this subsection. Some
recoveries from 100- and 1000-fold concentrations carried out in K-D
assemblies are shown in the following table. The recoveries are quite
acceptable when concentrating to 1 ml, but when concentrating to 100 yl
without a keeper, recoveries become marginal. Using a keeper, such as a
paraffin oil, helps retain the compounds and greatly reduces losses. How-
ever, the keeper may interfere with some analysis, especially by flame
ionization detection or mass spectrometry.
TABLE 8-1
Losses of pesticides on evaporation in Kuderaa-Danish concentrators (67)
Pesticide
Original
amount in
100 ml hexane
% Recovery-on concentration to *
10 ml '
1 ml
0.1 ml
0.1 ml with
"keeper"
Diazinon
Aldrin
Malathion
Parathion
Dieldrin
_p_,jg;-DDT
40
1.0
40
10
2.5
5.0
102 (2.4)
103 (2.8)
85 (2.4)
93 (4.4)
103 (5.6)
96 (9.2)
85 (4.4)
85 (4.0)
91 (5.2)
84 (4.0)
92 (4.0)
91 (5.6)
71 (1.8)
69 (1.2)
77 (3.0)
70 (1.2)
78 (0.6)
78 (3.0)
83 (1.8)
81 (0.6)
88 (0.6)
82 (2.4)
90 (1.2)
90 (2.4)
Averages of six determinations for each pesticide. Standard deviations
given in parentheses. Gentle stream of nitrogen used to assist concentra-
tion below 10 ml.
-305-
-------
Section 8M
Evaporation to dryness should never occur. If the complete removal of a
particular solvent is required, solvent exchange can be carried out so that
the sample never gets to dryness. For example, hexane can be completely
removed by boiling-down to a low volume and adding small volumes of acetone
as evaporation continues until all hexane is eliminated.
The use of air for concentration of an extract should best be avoided.
Satisfactory recoveries are obtainable when the residue levels are rela-
tively high, but significant losses have been documented of even the more
stable pesticides at low concentration levels (65).
A commercial tube heater that avoids evaporation to dryness with micro K-D
apparatus was originally described by Beroza and Bowman (68) (Figure 8-1).
Six extended-tip K-D concentrator tubes are accommo.dated, and simultaneous
evaporation to less than 1 ml can be carried out without attention.
Figure 8-1. Tube Heater,
Kontes Glass Co., K-720000
Other reports of pesticide loss include dieldrin and DDT when an extract
was evaporated in the presence of light (69), mirex upon evaporation of
aqueous solutions (70), and carbamate pesticides when evaporated in a K-D
apparatus (71). In the latter case, rotary vacuum evaporation (Figure 8-F)
at 50-55°C with addition of a keeper, solution was recommended. Many
carbamates can be successfully evaporated under a nitrogen stream without
loss after adding a keeper. A satisfactory general purpose keeper is 5
drops of 1% paraffin oil in hexane. Solutions containing the herbicide
Balan (benefin) cannot be evaporated in a current of air without loss of
pesticide, whereas rotary evaporation at a temperature of 50°C or less is
successful. All evaporation, and concentration steps should be checked with
spiked samples if any question of pesticide loss should arise*
-306-
-------
Section 8N
The Importance of clean glassware in all parts of a pesticide analysis has
been stressed several times earlier in this Manual. The special importance
of clean,glassware to be used for concentration of solutions to small
volumes cannot be overemphasized.
The final solution to be used for the determinative step must be composed
of a solvent appropriate for the particular analytical procedure. Choice
of a volatile solvent for partition and column cleanup procedures is
advantageous because evaporation to an appropriate volume can be carried
out quickly enough to be practical. If a different solvent is required
for the final sample solution, solvent exchange can be carried out by
taking up the nearly dry residue in the new solvent after evaporation.
Solvents for GC and LC are restricted by the selectivity of the detector,
while for TLC almost any volatile solvent is useful for the solution to be
spotted. Chlorinated solvents cannot be present in the injected solution
when an EC or the Cl modes of the MC or electrolytic detectors are to be
used. Acetonitrile has an adverse effect on the response of the EC
detector, while aromatic and halogenated compounds and acetonitrile increase
the response of the thermionic detector. The most volatile solvent possible
should be used to shorten the venting period and minimize loss of early
eluting pesticides for those detectors that require solvent venting (e.g.,
FPD and CCD). A solvent free of UV absorption is required for the detection
by the ultraviolet LC monitor.
8N REFERENCES .
(1) Ambrus, A., Adv. Pest.' Sci.. Plenary Lect. Symp. Pap. Int. Congr. Pestic.
Chem.. 4th. 1978. j3» 62° (1979).
(2) Albro, P. W., Ann. NY Acad. Scj., 320, 19 (1979).
(3) Ford, J. H., McDaniel, C. A., White, F. C., Vest, R. E., and Roberts,
RJ E., J. Chromatogr. Sci.. 13, 291 (1975).
'H ' d
(4) Dick, G. L., Heenan, M. P., Love, J. L., Udy, P. B., and Davidson, F.,
N. Z. J. Sci.. 21, 71 (1978).
(5) Bourne, S. . J. Environ. Sci. Heath B.. 13(2), 75 (1978).
(6) Mies, J. W., Dole, W. E., and Churchill, F. C., Arch. Environ. Contain.
Toxicol., 5., 29 (1976).
(7) Kurtz, D. A., Pestic. Monit. J.. U.(4), 190 (1978).
(8) Horwitz, W., and Howard, J. W., NBS (US) Spec. Publ. 519:231 (1979).
(9) Arthur, R. D., Cain, J. D., and Barrentine, B. F., Bull. Environ.
Contain. Tosicol., 15, 129 (1976).
(10) Kutz, F. W., Yobs, A. R., and Yang, H. S. C., Chapter 4 in Air
Pollution from Pesticides and Agricultural Processes!, Lee, R. E., Jr.,
ed., CRC Press, Boca Raton, FL.
-307-
-------
Section 8N
(11) Lewis, R. G., Sampling and Analyses of Airborne Pesticides, in Air
Pollution from Pesticides and Agricultural•Processes. Lee, R. E., Jr.,
ed., CRC Press, Boca Raton, FL, pp. 52-94 (1976).
(12) Johnson, E. R., Yu, T. C., and Montgomery, M. L., Bull. Environ. Contam.
Toxicol., 1J, 369 (1977). ~~"~
(13) Taylor, A. W., Glotfelty, D. E., Turner, B. C., Silver, R. E.,
Freeman, H. P., and Weiss, A., J. Agr. Food Chem.. 25, 542 (1977).
(14) Kaminsky, F., and Melcher, R. G., Am. Ind. Eve. Assoc. J.. 39(8),
678 (1978). — '
(15) Lewis, R. G., Brown, A. R., and Jackson, M. D., Anal. Chem.. 49.
1668 (1977). —
(16) Turner, B. C., and Glotfelty, D. E., Anal. Chem.. £9, 7 (1977).
(17) Eichelberger, J. W., and Lichtenberg, J. J., Environ. Sci. and Technol.,
5, 541 (1971). : '
(18) Wilson, A. J., Bull. Environ. Contam. Toxicol.. 15, 515 (1976).
(19) Braus, H., Middleton, F. M., and Walton, G., Anal. Chem.. 23, 1160
(1951).
(20) Sproul, 0. J., and Ryckman, D. W., J..Water" Pollut. Contr. Fed., 33.
1188 (1961). ' ~'
(21) Ruzicka, J. H. A., and Abbott, D. C., Talanta. .20, 1261 (1973).
(22) Musty, P. R., and Nickless, G., J. Chromatogr.. 120, 369 (1976).
(23) Aue, W. A., Kapila, S., and Hastings, C. R., J. Chromatogr.. 73,
99 (1972). *- —'
(24) Ito, T., Water Resource Research Institute of the University of North
Carolina, Report No. 54, August (1971).
(25) Gesser, H. D., Chow, A., Davis, F. C., Uthe, J. F., and Reinke, J.,
Anal. Lett., ±, 883 (1971); Gesser, H. D., Sparling, A. B., ChowT A.,
and Turner, C. W., J. Am. Wat. Wks. Ass., 65. 220 (1973).
(26) Musty, P. R., and Nickless, G., J. Chromatogr... 100. 83 (1974).
(27) Weil, L., Quentin, K.-E., and Gitzowa, S., Gas -u. WassFach. 113,
64 (1972). '
(28) Uthe, J. F., Reinke, J., and Gesser, H. D., Environ. Lett.. .3, 117
(1972).
-308-
-------
Section 8N
(29) Burnham, A. K., Calder, G. V., Fritz, J. S., Junk, G. A., Svec, H. J.,
and Willis. R.. Anal. Chem., 44. 139 (1972).
(30) McNeil, E. E., Otson, R., Miles, W. F., and Rajabalee, F. J., J_.
Chromatogr., 132, 277 (1977).
(31) Riley, J. P., and Taylor, D., Anal. Ghim. Acta. 46, 307 (1969).
(32) Musty, P. R., and Nickless, G., J. Chromatogr., 89, 185 (1974).
(33) Junk, G. A., Richard, J. J., Grieser, M. D., Witiak, D., Witiak,
J. L., Arguello, M. D., Vick, R.,.Svec, H. J., Fritz, J. S.,*and
Calder, G. V., J. Chromatogr., 99, 745 (1974).
(34) Harvey, G. R., U.S. EPA Report R2-73-177, 32 pp (1973).
(35) Coburn, J. A., Valdmanis, I. A., and Chau, A. S. Y., J. Assoc. Off.
Anal. Chem.. 60_, 224 (1977).
(36) Rees, G. A. V., and Au, L., Bull. Environ. Contam. Toxicol., 22(4/5)
561 (1979).
(37) Sundaram, K. M. S., Szeto, S. Y., and Hindle, R., J. Chromatoer.. 177(1),
29 (1979).
(38) Kahn, L., and Wayman, C. H.', Anal. Chem., 36. 1340 (1964).
(39) Goldberg, M. C., DeLong, L., and Sinclair, M., Anal. Chem., 45, 89
(1973).
(40) Ahnoff, M., and Josefsson, B., Anal. Chem.. 46. 658 (1974).
(41) Ahnoff, M., and Josefsson, B., Anal. Chem., 48, 1268 (1976).
(42) Wu, C., and Suffet, I. H., Anal. Chem., 4£, 231 (1977).
(43) Brodtmann, N. V., J. Am. Wat. Wks. Ass., 67, 558 (1975).
(44) Miles, J. R. W., Pestic. Monit. J., 10, 87 (1976).
(45) Miles, J. W., Dale, W. E., and Churchill, F. C., Arch. Environ.
Contam. Toxicol.. 5_, 29 (1976).
(46) Wojcik, D. P., Banks, W. A., Hicks, D. M., and Plumley, J. K., Florida
Entomologist. 55_, 115 (1972).
(47) Pesticide Residue Analysis in Water. U.S. EPA, Office of Water Programs,
distributed by NTIS (No. PB-238 072) September (1974).
-309-
-------
Section 8N
(48) Mes, J., and Campbell, D. S., Bull. Environ. Contam. Toxicol.. 16,
53 (1976). ' —
(49) Analytical Methods for Pesticide '•Residues in Foods, Canadian Depart-
ment of National Health and Welfare, Section 14.4.
(50) Howells, K. J., Shaw, T. C., Rogers, P. P., and Galbraith, K. A.,
Lab. Pract.. 23, 248 (1974).
(51) Wheeler, W. B., Thompson, N. P., Edelstein, R, L., and Krause, R. J.,
Bull. Environ. Contam. Toxicol.. 23(3), 387 (1979).
(52) U.S. FDA Pesticide Analytical Manual, Vol. I, Section 253.
(53) Telling, G. M., Sissons, D. J., and Brinkman, H. W., J. Chromatogr..
137, 405 (1977). *~
(54) Nutahara, M., and Yamamoto, M., J. Pestic. Sci.. 3/2), 101 (1978).
(55) Agemian, H., and Chau, A. S. Y., J. Assoc. Off. Anal.'Chem.> 60, 1070
(1977). —
(56) Hill, B. D., and Stobbe, E. H., J. Agr. Food Chem.. 22, 1143 (1974).
(57) Johnsen, R. E., and Starr, R. I., J. Agr. Food Chem.. 20, 48 (1972).
(58) Voss, G., and Blass, G., Analyst, 98. 811 (1973).
(59) Suffet, I. H., and Faust," S. D., J. Agr. Food Chem.. 20, 52 (1972).
(60) Suffet, I. H., J. Agr. Food Chem.. 21, 288 (1973).
(61) Suffet, I. H., J. Agr. Food Chem.. 21, 591 (1973).
(62) Malaiyandi, M., J. Assoc. Off. Anal. Chem.. 61, 1459 (1978).
(63) May, T. W., and Stalling, D. L., Anal. Chem., j>l(l), 169 (1979).
(64) Beroza, M., Bowman, M. C., and Bierl, B. A., Anal. Chem.. 44, 2411
(1972). —
(65) Burke, J. A., Mills, P. A., and Bostwick, D. C., J. Assoc. Off. Anal.
Chem., 49, 999 (1966). '. ""
(66) Chiba, M., and Mbrley, H. V., J. Assoc. Off. Anal. Chem., 51, 55 (1968).
(67) Lewis, R. G., Accuracy in Trace Organic Analysis, in Accuracy in Trace
Analysis: Sampling, Sample Handling, and Analyses. Vol. 1, La Fleur, D.,
ed., NBS Special Publication 422, Washington, DC, pp. 9-33 (1976).
-310-
-------
Section 8N
(68) Beroza, M., and Bowman, M. C., Anal. Chem., 39_, 1200 (1967).
(69) McKinley, W. P.^ and Savary, G., J. Aar. Food Chem.. 10,, 229 (1962)
(70) Stein, V. B., and Pittroan, K. A., Bull. Environ. Contain. Toad-col.,
19(6), 755 (1978).
(71) Storherr, R. W., J. Assoc. Off. Anal. Chem.. 55, 283 (1972).
-311-
-------
Section 9
MULTIRESIDIE EXTRACTION AND ISOLATION PROCEDURES FOR
PESTICIDES AND METABOLITES AND RELATED OTOUNDS
This section presents brief descriptions 'and quality control aspects
of widely used multiresidue analytical procedures for different sample
substrates. A few methods for important individual residues are also
included. Many of the problem areas are treated in a general manner
elsewhere in this Manual, but they are high-lighted again here in re-
lation to the specific methods. References are given in each case
to sources of detailed methodology. Control of procedures for
collection of samples is covered in Subsections 8A-8K in Section 8
and for sample extraction and extract concentration in Subsections 8L
and 8M in Section 8.
CHLORINATED PESTICIDES
9A TISSUE, FAT, AND FOOD ANALYSIS BY THE MILLS, ONLEY, GAITHER PROCEDURE
a. Analysis of Tissue and Fat
The modified Mills, Onley, Gaither method described in Section
5,A,(l),(a) of the EPA Pesticide Analytical Manual has been determined
by a number of interlaboratory collaborative studies to yield very
acceptable precision and accuracy for the analysis of a number of
chlorinated pesticides and metabolites in human or animal fatty tissues.
However, many polar OP and carbamate pesticides are not recovered.
(This method involves dry maceration of a 5 gram sample with sand and
anhydrous Na2S04, isolation of fat by repeated extraction with petroleum
, ether, extraction of residues into acetonitrile, and then partitioning
back into petroleum ether after adding 2% NaCl, drying by elution through
a column of Na2S04, concentration of the eluate, cleanup on a Florisil
column, and EC-GC after reconcentration of column eluates.) If necessary,
further cleanup of the 15% ether-petroleum ether Florisil eluate is carried
out on a MgO-Celite column. Pooled blood serum can be analyzed by the MOG
Florisil procedure after extraction with a hexane-acetonitrile solvent
system [EPA PAM, Section 5,A, (3),(a),VIII].
-312-
-------
Section 9A
(1) Some analysts, with hope of saving time, have combined 6% and 15%
ethyl ether-petroleum ether Florisil column fractions and have then
attempted gas'chromatography on the mixture. With some luck this
approach might prove successful, but there is a good chance that it
could lead to erroneous conclusions. For example, in one documented
instance, an analyst reported the presence of aldrin in a human fat
sample. Other collaborators on the sample analysis found the same
peak in the 15% eluate, making its identification as aldrin impossible
since this compound elutes wholly in the 6% fraction. By combining
the fractions, the analyst inadvertently neglected the use of se-
lective adsorption as a valuable identification tool.
(2) The polarity of the ethyl ether-petroleum ether eluting solutions
exerts a profound effect on the elution pattern of several pesti-
cidal compounds. The amount of ethanbl, a relatively polar solvent,
in the ethyl ether is a critical factor as illustrated in Figure
4-A in Section 4. As indicated in this figure, with nd ethanol,
dieldrin would be expected to yield only 87% recovery in Fraction
II with the balance being retained on the column. If twice the
proper amount of ethanol is present, approximately 7% should elute
in Fraction I, giving a 93% recovery in Fraction II. If 2% ethanol
is present and all the dieldrin still does not elute in Fraction II,
the presence of moisture in the system may be the cause. An excess
of moisture may result in all or most of the dieldrin eluting in the
6% fraction. . "
(3) The "activity characteristics of Florisil may vary somewhat from lot
to lot. Each lot, when received at a laboratory, should be care-
fully evaluated to be certain,the compound elution characteristics
are satisfactory.
(4)
Storage and holding temperature of Florisil are critical. The oven
used for holding this (and other adsorbents) should be confined
exclusively to this usage and not used as an all-purpose drying
Florisil will readily pick up air-borne contaminants that
oven.
may result in spurious chromatographic peaks. If the oven tempera-
ture varies more than +_ 1°C, considerable influence may be observed
in the retention characteristics. The recommended activation tempera-
ture is 130°C.
(5) Anhydrous Na2S(>4 used to top the Florisil column, even AR grade,
frequently contains sufficient impurities to result in spurious
peaks in the blank eluates. Because of the prevalence of this
j situation, it is good practice to Soxhlet extract all lots of
'this salt before use.
(6) The presence of peroxides in ethyl ether can result in extremely
low recoveries of organophosphorus compounds and also poses a
serious safety hazard. Methods have been set forth for the re-
moval of peroxides from ether but have not proven wholly satis-
factory. The purity of petroleum ether is also critical and may
exert a profound effect on the recovery of certain of the organo-
phosphorus compounds.
-313-
-------
Section 9A
(7) Glassware must be meticulously cleaned to remove electron capturing
contaminants. Reagent blanks must be run with each set of samples.
(8) Most chlorinated pesticides should be recovered in the range of
85-100%. HCB is an exception because of an unfavorable partition
ratio in the acetonitrile-petroleum ether solvent system. An
aldrin spike can be added to the minced fat at the start of the
procedure if this pesticide is known to be absent. Recovery of
this spike should not be less than 70%.
If improper Florisil fractionation occurs during an analysis, the
following points should be considered: Florisil that is too retentive
could result from (a) improper activation temperature, (b) improper
percentage of ethyl ether in petroleum ether, and ethyl ether that does
not contain the required 2% ethanol (read the label on the container
carefully). Florisil that appears insufficiently retentive might re-
sult from (a) or (b) above, or from residual amounts of a polar solvent
in the sample or standard being placed on the column. Likely possi-
bilities are acetonitrile from the sample partition cleanup step (if
drying steps are not performed properly) or incomplete removal of benzene
(or other solvent more polar than hexane) from a standard solution placed
on the column. Other sources of Florisil problems are. undoubtedly
possible. See also Subsection 9M for further comments on pesticide
elution from Florisil.
b. Analysis of Fatty and Nonfatty Foods Using Florisil Cleanup
The Mills, Onley, Gaither column method for determining nonionic
chlorinated pesticides in fatty foods is similar to that outlined in Sub-
section 7Aa and is described in detail in Sections 211, 231, and 232 of
the FDA PAM. Eluants are 6, 15, and 50% ethyl ether in petroleum ether.
The method for nonfatty foods (FDA PAM, Sections 212 and 232) involves
extraction of pesticides with acetonitrile or water-racetonitrile and
partition into petroleum ether prior to Florisil column chromatography
and EC-GC, The FDA PAM lists pesticides recovered through these pro-
cedures [results for some 200 pesticides and other chemicals are given
in Table 201-A and over 300 compounds have been tested (1)], samples to
which they are applicable, and supplemental cleanup procedures for the
Florisil column fractions. This AOAC multiresidue method is currently
official for 26 OC1 and OP pesticides and PCBs in various groupings of
42 nonfatty and 4 fatty foods (1). The problem areas are the same as
those given in Subsections 9Aa and 9M. The elution pattern of more than
150 pesticides from the U.S. FDA Florisil column eluted with 6, 15, 20,
30, 50, and 65% diethyl ether in petroleum ether is tabulated in Section
7.2(b) of the Canadian PAM. Free fatty acids in high quantity are not
sufficiently removed by this FDA/AOAC procedure to prevent interference
with pesticide determination using electron capture, KC1 thermionic, and
Hall electrolytic conductivity detectors (2). A potassium permanganate/
dilute sulfuric acid oxidation procedure was developed to supplement
Florisil chromatography for cleanup of chlorinated pesticide residues in
vegetable extracts. Twelve chlorinated pesticides were completely re-
covered, and only aldrin was lost via decomposition (3).
-314-
-------
Section 9A
Thirteen chlorinated pesticides were determined in milk by GC after Florisil
column cleanup. Of the several systems tested, extraction of milk with
20 ml hexane plus 5 ml acetonitrile plus 1 ml ethanol produced the highest
pesticide recoveries and lowest fat extraction (4).
In order to obtain more efficient cleanup of extracts of fatty foods and
recovery of additional pesticides of higher polarity (e.g., organo-
phosphates), a new elution system consisting of three different mixtures
of methylene chloride, hexane, and acetonitrile was devised as replace-
ment for the traditional diethyl ether-petroleum ether eluants. These
eluant mixtures are methylene chloride-hexane (20:80 v/v); methylene
chloride-acetonitrile-hexane (50:0.35:49.65 v/v); and methylene chloride-
acetonitrile-hexane (50:1.5:48.5 v/v). At least 50 pesticides and re-
lated chemicals have been recovered, in groupings different from the
mixed ether'systems, with these new solvents (5). Table 201-A of the
.FDA PAM also includes data on the elution characteristics of compounds
using the methylene chloride/hexane/acetonitrile system (FDA PAM, Section
252). A silver nitrate-coated Florisil column has provided cleanup of
fatty and vegetable sample extracts and fractionation of chlorinated pesti-
cides and phthalate esters prior to their simultaneous analysis by gas
chromatography (6).
Malathion and some other organophosphorus pesticides require 50% diethyl
ether-petroleum ether for elution from Florisil. This elution, which
must be preceded by elution with the 6% and 15% eluants, has been found
occasionally to be inconsistent. OP pesticides can be lost through de-
gradation on the Florisil column and during subsequent evaporations, or
when water dilution of the acetonitrile extract for residue transfer to
petroleum ether is carried out. Recoveries are tested by carrying known
amounts of pesticides through the procedure in the absence of crop sub-
strate. Only 23 of 70 OP pesticides and metabolites tested through the
MOG procedure were recovered, and not all recoveries were complete. The
AOAC has validated the procedure only for carbophenothion, diazinon,
ethion, malathion, methyl and ethyl parathion, and ronnel in 18 fruit and
vegetable crops (7-9). ,
Beckman and Garber (10) recommended the solvent series benzene, diethyl
ether-benzene (1:2 v/v), acetone, and methanol for elution of Florisil
columns. The elution pattern and recovery of 65 OP pesticides were studied,
but sample extracts were not tested. This system was later found to be
applicable to the determination of methyl and ethyl parathion, malathion,
malaoxon, and paraoxon residues in apples and lettuce, although "all-Florisil"
columns were not generally recommended as the best choice for cleanup of OP
pesticides (11).
A novel use of Florisil was the development of a partition column consisting
of acetonitrile on Florisil for the separation of some pesticides from fish,
beef, and butter fat (12). The technique was useful for cleanup of pesti-
cides having favorable ^-values (Section 10F) in a hexane-acetonitrile
system, which included dimethoate, temephos, methyl parathion,, fenitrothion,
•.rufomate, malathion, and parathion.
-315-
-------
Sections 9B, 9C, 9D
A multiresidue method for organochlorine,. organophosphorus, dinitrophenyl,
and carbamate pesticides in applies and other high-water crops was devised
using Florisil for cleanup (13). Carbamates were eluted from one column
with toluene-acetone (19:1 v/v) and acetylated with trifluoroacetamide for
EC-GC determination. Organochlorine and organophosphorus compounds were
eluted from a separate column with toluene-acetone (49:1 v/v) and determined
by GC. Dinitrophenyl compounds were then eluted from this column with 95%
ethanol, cleaned-up by solvent partitioning, methylated, and determined by
EC-GC. Most recoveries were greater than 75%, even for polar compounds.
9B HCB AND MIREX IN ADIPOSE TISSUE
Section 5,A,l,(b) of the EPA PAM describes the determination of hexachloro-
benzene (HCB) and mirex in fatty tissue with confirmation of HCB by formation
of Msj-isopropoxytetrachlorobenzene. The sample is dissolved- in hexane
and applied directly to a Florisil column. The HCB and mirex residues are
eluted with hexane and determined by direct EC-GC of the concentrated eluate.
HCB is then reacted with 2-propanol, and the BITS derivative is chromatographed
to provide confirmation of HCB. Mirex does not survive this reaction. Other
common pesticides, some of which are altered by the reaction, are all sepa-
rated from the HCB derivative on the OV-17/OV-210 column used (14).
9C HUMAN OR ANIMAL TISSUE AND HUMAN MILK ANALYSIS BY THE FLORISIL MICROMETHOD
If the size of the available tissue sample is so small as to make the macro
MOG method unsuitable, a micromethod is described in Section 5,A,(2) of the
EPA PAM requiring as little as 200-500 mg of sample. The sample is extracted
with acetonitrile, pesticides are partitioned into hexane, fractionated on
a 1.6 gram Florisil column (eluate I: 12 ml of hexane plus 12 ml..of 1%
methanol in hexane; eluate II: 12 ml of 1% methanol in hexane), and concen-
trated fractions determined by EC-GC. Several pesticides, including O-BHC,
lindane, diazinon, DDD, and toxaphene, split between fractions. Florisil
columns must be conditioned at 130°C at least overnight before xteing. Pre-
cautions concerning use of Florisil are similar to those outlined in Sub-
section 9Aa. Virtues of the micromethod include a low background level and
savings in the volume of solvent required.
Miniaturization of Florisil column cleanup has been reported in several papers
(15, 16). One procedure has been successfully studied by several labora-
tories (17).
9D HUMAN BLOOD OR SERUM
A 2 ml aliquot of serum is extracted with 6 ml hexane for 2 hours on a slow
speed rotary mixer. After concentration, the hexane layer is analyzed by
EC-GC [EPA PAM, Section 5,A, (3),(a)]. The procedure involves no cleanup,
but, if carefully handled, it is capable of yielding recoveries of chlorinated
pesticides comparable to that obtained from a full MOG cleanup technique
(see Tables 2-4 to 2-9 in Section 2). Since all pesticides will be present
in one extract, a GC column must be chosen that will separate the expected
-316-
-------
, Section 9E
pesticides. Certain serum samples will yield a very late eluting extraneous
peak (probably a phthalate) that is sometimes large enough to distort a
following chromatogram if time is not allowed for its elution from the
column. Blood samples should never be stored in containers with polyethylene
or rubber caps. Hexane was proven superior to hexane-formic acid for
extraction of dieldrin, lindane, and DDT from serum (18). Microcoulometric
GC determination after sulfuric acid extraction was successfully applied to
24 organochlorine pesticides in blood at 1 ppb levels with no cleanup (19).
Blood samples are not always analyzed without cleanup steps. Monitoring
of fur seal blood for OC1 pesticides and PCBs required chromatography of
the hexane extract on a 2.3 gram Florisil column prior to EC-GC (20),
Another monitoring study of pesticides in human blood was carried out by
hexane extraction of acidified samples followed by cleanup on a 1 gram
Florisil column and EC-GC (21).
Hexane-acetone (9:1 v/v) was a better extractant for DDT and BHC isomers
in human blood than was pure hexane. Maximum recoveries occurred when
serum was treated with formic acid before extraction. Ease of extract-
ability decreased in the order: f-BHC > a-BHC > g-BHC > £,£.'-DDE > p_,p_'-DDT >
o^'-DDT (22).
9E PENTACHLOROPHENOL (PCP) IN BLOOD AND URINE
Acidified blood is extracted with benzene on a Roto-Rack for 2 hours
followed by methylation of PCP and determination by EC-GC [EPA PAM,
Section 5,A,(3),(b)]. Urine is acidified, and hydrolysis carried out for
one hour to free conjugated PCP. PCP and phenol metabolites of PCP and
HCB are extracted with benzene, methylated with diazomethahe, the methylated
phenols are cleaned up and fractionated on an acid alumina- column, and
determination is carried out by EC-GC [EPA PAM, Section 5,A, (4),(a)]. The
following comments pertain to these methods:
a. The alkylating reagent diazomethane is a hazardous chemical and must
be handled with extreme caution.*
— Diazomethane and related alkylating reagents (e.g., diazoethane, diazo-
pentane) have been widely used in pesticide residue analysis and are
cited in several procedures in this Manual and the EPA PAM. These com-
pounds and their precursors are toxic and carcinogenic and are irritating
to the skin. Solutions have been known to explode inexplicably. It is
recommended that safer substitutes be found for these reagents whenever
possible, for example BF3-methanol for methylation of acid herbicides
and acetic anhydride for acetylation of pentachlorophenol (Section 9Ac).
Substitution of one reagent for another, however, can require a large
amount of effort to check the validity of the procedure with the new
reagent. If diazolkane reagents must be used to reproduce established
analytical procedures, take care to keep from direct contact with the
- skin. Wear disposable vinyl gloves and safety goggles, and avoid
breathing of vapors. Work behind a safety shield in an efficient hood
or inside a radiological glove box. Do not prepare or store reagents
in ground glass stoppered or etched glassware. Avoid strong light.
-317-
-------
b.
d.
Section 9E
A 1.5% OV-17/1.95% QF-1 column is not recommended since the relative
retention values for 2,4-D methyl ester and PCP methyl esters are
identical and these pesticides would not be differentiated.
All reagents including distilled water must be pre-extracted with
hexane to remove interfering materials. Reagent blanks should be
carried through the entire procedures with each set of samples and
standards.
Glassware should be washed with dilute NaOH followed by deionized
water and acetone.
Contact between wooden or paper materials and glassware should not
be permitted as some of these materials have been found to contain
significant levels of PCP.
Other ether derivatives (e.g., ethyl, propyl, amyl, etc.) can be pre-
pared and characterized for confirmation of PCP identity.
Fortified samples should be analyzed along with each series of actual
samples to verify adequate recovery of PCP and the other phenols of
interest. Because of the ubiquity of PCP, the "blank" used for forti-
.fication must be analyzed, and a correction must be made for the
amount of PCP found.
h. "A reagent blank consisting of 5 ml of pre-extracted distilled water
should also be carried through the entire procedures along with
samples.
f.
g-
i.
Confirmation of PCP is based on chemical ionization mass spectrometry
or extraction j^-values.
In the methods described above (23, 24), phenols were chromatographed on
conventional GC columns after derivatization to a more easily chromato-
graphed compound. The derivatization step exposes the analyst to a toxic
derlvatizing reagent and increases the possibilities of error. It has
been demonstrated that support coated polyester columns are suitable for
determining free chlorinated phenols in urine at subnanogram levels with-
out the need for derivatization (25).
A method for monitoring PCP in fish and other environmental samples with a
± 2% accuracy and precision has been described (26). PCP was extracted
from fish tissue and converted to pentachloroanisole (PCA) by means of
alkylation in the presence of potassium carbonate as a condensing agent.
After adding pentachlorophenetole as internal standard, determination was
carried out by electron impact mass fragmentography monitoring 280 m/e for
PCA and 294 m/e for the internal standard.
-318-
-------
Sections 9F, 9G
97 BIS(2-CHLOROPHENYL)ACETIC ACID (j5,2.!-DBA) IN HUMAN URINE
The excretion level of this metabolite is a sensitive indicator of exposure
to £,£f-DDT. Urine is extracted three times with an equal volume of 2%
acetic acid in hexane, the combined extracts are evaporated to remove
residual water or acetic acid, DDA is converted to its methyl ester by
reaction with BF3-methanol reagent, and the ester is extracted with hexane
and determined by GC with microcoulometric or EC detection [EPA PAM,
Section 5,A, (4),(b)].
Microcolumn Florisil cleanup (Subsection 9C) is required when the poorly
selective EC detector is used. DDA should elute completely in Fraction
II. Concentration and injection volumes depend upon the sensitivity of
the detector employed. A column of 5% OV-210 at 175-180°C will separate
DDA from p,£*-DDE (which usually is also present in high exposure donors),
whereas 4% SE-30/6% QF-1 or 1.5% OV-17/1.95% QF-1 columns at 200°C will not
resolve these compounds.
A very similar procedure involving diazomethane methylation, no cleanup,
propyl ester internal standard, a 1% QF-1 column at 190°C, and a 63Ni
pulsed EC detector has been reported. The calibration curve was linear
up to 1.5 yg DDA per liter, the coefficient of variation was ca 8%, the
absolute detection limit was 0.05 ng, and 20-30 samples could be run per
day (27).
96 2,4-D AND 2,4,5-T IN URINE
A method is described in the EPA'PAM, Section 5,A,(4),(c), for determining
these herbicides and their degradation products 2,4-dichlorophenol and
2,4,5-trichlorophenol in human and animal urine. Phenolic conjugates are
hydrolyzed in acid, and free phenols and acids are extracted with benzene
and ethylated with diazoethane. Cleanup and fractionation of derivatives
is carried out on a silica gel column (1 gram, containing 1.5% water), and
determination of concentrated eluates by EC-GC on a 4% SE-30/6% OV-210
column.
Deactivated silica gel (Subsection 4Ad in Section 4) columns should be
prepared just prior to use. Because of the differences in temperature and
humidity from one laboratory to another, silica gel elution parameters
should be established by each analyst under local conditions. The per-
centage water added for deactivation should be increased if the compounds
of interest elute in a later fraction than that indicated in the detailed
procedure, or, the percentage of benzene in the benzene-hexane eluant can
be increased. Early elution would be remedied by less deactivation or less
polar solvents. Spiked control urine, rather than standard compounds,
should be used to determine the elution pattern. See the footnote on
page 6 concerning the hazards associated with the ethylating reagent.
Alkylated standards are stable for one month if stored in a freezer
(-18°C) when not in use.
-319-
-------
Section 9H
A multiresidue scheme for phenol metabolites and including 2,4-D, 2,4,5-T,
and silvex is discussed in Subsection 9R. A method for monitoring 2,4-D
in the urine of pesticide spray operators at 0.1 ppm involved cleanup on
XAD-2 resin, quantitation by GC of the methyl ester, and confirmation by
trans-butylation to the ii-butyl ester. Recovery was 94 ± 6% for five
fortified samples (28).
9H KEPONE IN HUMAN BLOOD FOR ENVIRONMENTAL SAMPLES
The determination of Kepone in human blood, air, river water, bottom
sediments, and fish is described in the EPA PAM, Section 5,A,(5),(a).
This is based on the research of Moseman gt al. (29), Hodgson et al. (30),
and Earless ^£ al. (31). Samples are extracted, and the extracts are
cleaned up by chromatography on a micro Florisil column, base partitioning,
or gel permeation chromatography. Kepone is determined by EC-GC with
multiple columns. Confirmation is by chemical conversion to mirex
followed by further cleanup prior to EC-GC (32); detection with a Hall
conductivity detector in the Cl-mode; or chemical ionization mass
spectrometry (31).
It is mandatory to use 1-2% methanol in benzene for all sample and standard
solutions injected for EC-GC to obtain the maximum reproducible response.
Sufficient control and spiked reference materials should be utilized to
ensure the validity of analytical results for all sample types. Elution
patterns for the Florisil columns should be carefully established by each
analyst by eluting standard Kepone under local laboratory conditions.
Analytical standards should be validated by cross-reference analysis of
difrerent preparations of analytical grade Kepone with agreement within
i 5% of the established purity.
The analysis of field-collected avian tissues and eggs for Kepone residues
has been reported (33). Samples were extracted with benzene-isopropanol
(2:1 v/v) and extracts cleaned up with fuming H2S04-concentrated H2S04
(1:1 v/v). Separation of Kepone from OC1 pesticides and PCBs was obtained
on a 10 gram 130°C-activated Florisil column eluted with 100 ml of benzene-
acetone (95:5 v/v) followed by 200 ml of benzene-methanol (90:10 v/v); the
second eluate contained the Kepone. Determination was by EC-GC on a 4%
SE-30/6% QF-1 column and confirmation by GC-MS. Recoveries averaged 86%
at 1 ppm. Procedures for determination of Kepone in serum, plasma, urine
and fat have been reported. After addition of #2804, samples were extracted
with hexane-acetone (17:3 v/v), extracts were evaporated, and the residue
dissolved in benzene-methanol (99:1 v/v). The extraction was modified for
feces and bile. Programmed temperature GC with pulsed EC detection on a
4% SE-30/6% QF-1 column provided linear calibration curves for 10 pg-100 ng
of Kepone (5 ppb-50 ppm/gram sample) (34). Determination of Kepone in eels
(35); blue fish or shrimp (36); finfish, shellfish, and crustaceans (37);
water and sediment (38); and soil and mullet (39) using gas chromatography
have also been reported in the literature.
-320-
-------
Section 91
91 GEL PERMEATION CHROMATOGRAPHY
a. Gel Permeation Chromatographic Cleanup of Adipose Tissue
(1) Theory
Gel permeation chromatography (GPC) is a form of liquid chromatog-
raphy by which compounds are separated according to molecular size. It is
particularly useful in separating very large molecules such as lipids and
cholesterol found in adipose tissue samples from the smaller molecules of
pesticides, PCBs, etc. The method is as effective as the MOG procedure
for cleanup in pesticide residue analyses (4QJ and has the added advantages
that removal of fat is more complete and recoveries of pesticides are nearly
quantitative. Hence, it is the ideal choice for GC-MS analyses, where
maximum detectability of pesticides is needed and minute quantities of lipid
materials can cause serious interferences.
Porous polymer beads (e.g., BioBeads SX-3) are used as gel
particles and organic solvents (e.g., toluene, ethyl acetate, or cyclo-
hexane) are used for the mobile phase. The elution process is very simple
(isocratic only); the same solvent system is used for column preparation,
elution, and washing. Macromolecules cannot permeate the porous gel and
are rapidly eluted or "dumped" from the column. Molecules that can enter
the pores of the beads are temporarily retained to greater or lesser
extents depending on their molecular volumes. Hence, large-volume pesti-
cides such as mirex elute first (in this case, following shortly after
cholesterol), while small-volume pesticides such as HCB elute last. Since
molecular volume rather than molecular weight dictates the order of elution,
all equatorial B-BHC elutes after the other BHC isomers.
(2) Equipment
The gel permeation chromatograph is an AutoPrep Model 1001
(Analytical Biochemistry Laboratories, Inc., Columbia, MO), equipped with
a 2.5 cm id x 60 cm glass column (Chromaflex3* R-422350/6025, Kontes, Vine-
land, NJ, or equivalent) packed with 200- to 400-mesh BioBeads SX-3 (BioRad
Laboratories, Richmond, CA).
(3) Column Preparation and Operation
(a) Prepare a slurry of ca 60 g of BioBeads SX-3 in pesticide
quality (or equivalent) toluene-ethyl acetate (1:3 v/v). This will be
sufficient to pack a column about 25 cm long.
(b) Add small volumes of resin and solvent to the column. "Each
addition of resin must be in contact with enough solvent to swell the resin
before the next addition.
(c) After the resin is transferred to the column, compress the
gel to approximately 25 cm, allowing solvent to flow out of the column exit.
-321-
-------
Section 91
(d) Add only ethyl acetate-toluene (3:1 v/v) to the solvent
reservoir. Addition of other solvents to the system via sample introduction
will change the gel swelling ratio and must be kept to a minimum (i.e.,
< 5% v/v of aliquot injected).
(e) Install the column and start the pump. The pump operating
pressure should be 5-7 psi (not to exceed 10 psi).
(f) Adjust the pumping rate to approximately 5 ml/minute with
the pump vernier control valve.
(g) Set the timer to collect for 20 minutes and check the actual
pumping rate.
(h) The GPC elution pattern of the pesticides of interest should
be established for standards before introduction of biological samples
into the gel permeation chromatograph,
(4) Procedure for GPC Cleanup
(a) Start up the GPC instrument and elute the column with ethyl
acetate-toluene (3:1 v/v) until it is purged of entrained air.
(b) Introduce
-------
Section 91
mini-alumina column for improved purification of pesticide extracts from
fat samples. In many cases, GPC fractions require no further cleanup prior
to determination of residues by GC.
d. Application of GPC
The original GPC system consisting of BioBeads SX-2 crosslinfced poly-
styrene gel and cyclohexane was designed by Stalling et al. (41) for re-
moval of lipids from extracts of samples such as fish before EC-GC determina-
tion of commonly occurring pesticide and PCB residues. This excellent
method was later improved considerably by the use of BioBeads' SX-3 and
ethyl acetate-toluene. A broad range of OC1 and OP pesticides can be
recovered in good yields from fats and oils (42, Subsection b above). An
evaluation of the GPC system (43) with different sample types indicated
that ca 98% of the fat or oil content of the extract is generally eluted
prior to the pesticide fraction and that this cleanup may be superior to
that achieved by acetonitrile partition and Florisil adsorption. However,
although recoveries were higher by GPC than by Florisil adsorption, pre-
cision was poorer with the former method. Analyses can be automated since
an important feature of GPC is that the same column can be, used repeatedly
over long periods without significant change in elution volumes or recoveries.
With the GPC procedure described in Subsection a above, organochlorine
pesticides have been determined and confirmed in human tissue and milk
(EPA PAM Section 12,A). Samples, are extracted and cleaned up by a modified
Mills, Onley, Gaither procedure. After further cleanup of Florisil fractions
by GPC, determination is carried out by GC on a Carbowax 20M column with a
Hall electrolytic conductivity detector.
Recoveries ranging from 88-106% were reported for disulfoton, diazinon,
methyl parathion, malathion, parathion, dichlorvos, and fensulfothion in
an evaluation study of the automated gel permeation chromatographic cleanup
techniques using BioBeads SX-3 gel and an ethyl acetate-toluene (3:1 v/v)
elution solvent (42). A solvent composed of cyclohexane-methylene chloride
(85:15 v/v) with BioBeads SX-3 provided adequate cleanup for EC-GC (no
liquid partitioning) of 9 OP pesticides and 2 metabolites and 16 nonionic
OC1 pesticides in vegetable oils at 0.05-1.0 ppm. Vegetables, fruits, and
crops were analyzed for 26 organophosphorus pesticides and metabolites at
0.05-0.10 ppm levels using automated GPC for cleanup followed by FPD-GC.
Recoveries of 7 compounds from 12 sample types were in the range 83-103%,
and 8 compounds could be determined simultaneously (44). Carbamate and
organophosphorus compounds in several plant crops were recovered at levels
between 82 and 104% by automated GPC (45).
Gel chromatography on Sephadex LH-20 has also been reported (46, 47) for
cleanup of organochlorines and organophosphates prior to GC, but this
approach is now of minimal importance in residue analysis.
-323-
-------
Sections 9J, 9K
9J DETERMINATION OF CHLOROPHENOXY HERBICIDES IN
-------
Sections 91, 9M
9L CLEANUP ON SILICA GEL
Silicar CC-4 silica gel (50) has been widely used for cleanup and fraction-
ation of OC1 insecticides in various monitoring programs (51). For
example, in a study of duck wing contamination (52), a 15 g column was
eluted with 60 ml of petroleum ether (HCB, mirex recovered), 350 ml of
petroleum ether (PCBs, some DDE), and 150 ml of methylene chloride-
hexane-acetonltrile (80:19:1 v/v) (remainder of DDE, IDE, DDT, and other
OC1 compounds). The same elution sequence was used to determine OC1
residues in herons (53). A modified sequence with four eluants, used
to assess contamination of Bald Eagles, allowed collection of dieldrln
and endrin in a discrete fraction: 80 ml petroleum ether (HCB and mirex);
320 ml petroleum ether (PCBs, PBBs, DDE); 275 ml hexane-methylene chloride
(85:15 v/v) (OC1 compounds, except endrin and dieldrin); 200 ml methylene
chloride-hexane-acetonitrile (80:19:1 v/v) (endrin and dieldrin) (54).
9M CLEANUP ON DEACTIVATED FLORISIL AND SILICA GEL (see also Section 9G)
The method of Osadchuk et_ al. is described in the Canadian PAM, Section
7.2, Deactivated Florisil is prepared as outlined in Subsection 4Ac in
Section 4 of this Manual. The elution behavior of over 50 pesticides on
Florisil deactivated, with 2% water has been determined for use after
extraction and partition cleanup of residues. A 30 cm'x 2.5 cm id column
containing 15 cm of adsorbent is eluted with 300 ml portions of the
appropriate eluting mixture(s) ranging from pure hexane to 5-30% methylene
chloride in hexane to 5-30% ethyl acetate in hexane (Table 9-1). If the
analyst wishes to screen a sample extract for a larger number of pesti-
cides in one or two GC injections, the less polar eluants may be by-passed
and only the more polar used. However, some sample types may be inadequately
cleaned-up by this procedure or mutually interfering residues may occur in
the same fraction.
The following factors affect the success of this Florisil procedure:
a. Pesticides containing a mercaptan function are oxidized on the
Florisil column. For example, phorate, captan, carbophenothion,, chloro-
benside, disulfoton, and demeton have losses ranging from 20-100%. The
oxidation proceeds to the sulfoxide and then to the sulfone. Therefore,
non-detection of such pesticides does not guarantee they were not
originally present in the sample. The degree of oxidation by Florisil
increases with a lower extent of water deactivation (greater adsorbent
activity) or a greater time of contact with the column and may also be
affected by the pH of the particular Florisil used.
b." Oxygen analogs of organophosphorus pesticides are strongly adsorbed
on Florisil and cannot be completely eluted even with very polar solvents.
c. The 2% deactivated Florisil column can tolerate up to one gram
of fat or oil (30% methylene chloride in hexane or less polar eluants)
without extraneous EC-GC response.
d. Up to two grams of fat or oil can be applied directly to the column
and eluted with 10% methylene chloride in hexane to recover BHC isomers, the
DDT group, PCBs, and HCB.
-325-
-------
Table 9-1
Section 9M
ORDER OP EU'TION OF PESTICIDES FHOM FLORISIL PARTIALLY
DEACTIVATED WITH & WATER USINO 300 mi VOLUME OF ELVENTS
(?rom tht Canadian 7AM)
PESTICIDES
Hexoho
in
EtOAo InHexnne
Parc»nt
"Reeoverioo
Aroclor 125^ PCB
Chlordani
Toxaphene
Strobar.e
Chlordane
Aldrin
Hexachlorobenzfin*
Heptachlor
p.p'-DDS
o,p«-DDT
Kirex
Iiobenzan
p.p'-RDT
a-BHC
Porthane
p.p'-DDD
Chlorbonside M
PCNB
TCKB
P-BKC
f-BHO
Dlcofol
Konnol OP
Kcpachlor epoxide
Dicltlofonthion OP
Phorat* 6^M
Carbophcnothicn O^M
Endosulfan I
Dlcldrir.
Chlorpyrifoo OP
Endrin
Hethoxychlor
Parathion OP
Ethion2 OP
2.4-D eiethyl esttr
2A5-T ncthyl ester
Anilazina
Ovex
Fenitrothion OP
Tetradifon
Diazinon OP
Chlorothalonil .
Methyl Parathion^ OP
Sulphenone
Dioxathion OP
Malathion OP
Atrazine*
K
OP
OP
Endosulfan II
Captan
Fhoamet
DCPA
Arinphonnothyl.
S(60J«)
S(50JC)
+
s(75J«)
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>9S
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>95
>90
>95
/v80
>95
>90
>95
Not« A 30^ CH?Cl2 fraction v/an eluted prior to all ethyl acetate fractions. All others were «incl«
elution*.
+ - mostly *lutcs in first 250 ml
• - larce aisount in 250-300 ml fraction
S - some (ao percent)
Footnoea*: 1. OP « organophosphorus; H - mercaptan; PCB - polychlorlnated biph*nyl
2. Higher recoveries are obtained by elutloa with more polar eluents
3. Beoainlng methyl parachion eluCes In another 50 ml of 51 EtOAc
4. Detected by alkali flame detector
-326-
-------
Sections 9N, 90
The approach in d above has been used to determine HCB and mirex in fish
and butterfat by elution with acetonitrile from a column composed of the
fat or oil distributed on unactivated Florisil. This procedure has been
collaboratively studied (55) and adopted by the AOAC as official final
action (56). Care must be taken in analyses for HCB not to use plastic
wash bottles, since this compound was found as a contaminant in 30 of 34
such bottles tested (57).
A one-step Florisil column cleanup described by Langlols et al. (58) has
been widely used to isolate organochlorlnes and PCBs. It is similar to
the method just described but employs deactivated Florisil. Activated
Florisil is"equilibrated with 5% water, and 1 g of fat from fish or other
extracts is thoroughly mixed with 25 g of this Florisil. The adsorbent
is placed on top of a second 25 g portion of conditioned Florisil in a
25 mm id column. The column is eluted with 300 ml of hexane-methylene
chloride (4:1 v/v) (59).
Silica gel deactivated with 30% water has been used to isolate organo-
chlorines from lipids (60). A micro column of this silica gel eluted
with petroleum ether has been shown (61) to yield especially pure eluates.
Small columns of precisely deactivated silicic acid (3 g, 3.3% water) were
found to separate £,£f-DDT, cis- and trans-chlordane, 2.>£.'~DDE» *a& PCBs
from the majority of toxaphene components. This fractionation greatly
simplified the analysis of the pesticides (62).
9N LOW TEMPERATURE PRECIPITATION
This procedure (Canadian PAM, Section 7.4) is used to separate fats, oils,
and water from acetone-benzene-acid extracts of biological samples by pre-
cipitation at -78°C. The special low temperature cleanup apparatus is
described in detail (Canadian PAM, Section 14.5). Many apolar and polar
residues and metabolites (e.g., DDT, 2,4-D acid and ester, parathion, and
paraoxon) are retained in the acetone supernate and can be determined by
EC-GC. Forty pesticides have been quantitatively (80+ percent) recovered
from a variety of plant and animal products at levels greater than 0.05 ppm.
Freeze-out has been recently employed for the removal of lipids prior to
Florisil chromatography and EC-GC in the determination of methoxychlor
residues in microsamples of animal tissues and water at 10 ppb and 1 ppb
levels, respectively (15), and for cleanup of human milk samples for ECB
and other chlorinated pesticides by EC-GC (63).
90 CLEANUP ON ALUMINA
Hexane extracts of animal tissues are cleaned-up and prefraetionated on
narrow bore columns dry-packed with partially deactivated alumina and
silica gel by the method of Holden and Marsden (64). The initial alumina
column eluted with hexane provides removal of lipids, while the second
column affords pre-GC separation of residues plus further cleanup. Table
9-2 shows the elution order of chlorinated insecticides with hexane and
10% diethyl ether-hexane eluants. Alumina is activated at SOO°C and silica
-327-
-------
Section 90
gel at 150°C before deactivation with 5% (w/w) water. Interferences con-
tributed by columns in the Holden-Marsden method have been removed by
methylene chloride treatment of the columns. Basic alumina was recommended
for easier control of activity and faster pesticide elution (65).
Another alumina-silica column scheme (66) was devised for separation of
17 OC1 residues in 4 eluates, each containing pesticides separable on a
4% SE-30/6% OV-210 GC column. Microcolumns deactivated with 3-4% water
were used.
Table 9-2
ORDER OF ELUTION OF ORGANOCELORifoES FROM DEACTIVATED SILICA GEL
ACCORDING TO THE METHOD OF HOLDEN AND MARSDEN (64)
Eluted in order
by hexane
Eluted in order by 10%
diethyl ether in hexane
Hexachlorobenzene
Aldrin
PCBs
j3,_p_f-DDE
Heptachlor
p_,£'-MDE (DDMU) .
Endrin
Chlordane
p_,p_'-DCBP
Toxaphene
p_,p_f-TDE
Telodrin
Heptachlor epoxide
a-BHC
Perthane
B-BHC
Kelthane
Y-BHC
Dieldrin
Methoxychlor
Organochlorine insecticide residues in fatty foodstuffs were determined (67)
by using a cleanup technique based on a single 22 g column of activity-4
basic or neutral alumina. Concentrated hexane extracts of samples, con-
taining 0.4-0.5 grams of fat, were transferred to the column, and pesti-
cides were eluted with 150 ml of hexane prior to determination by EC-GC.
-328-
-------
Section 9P
Recoveries of 15 Insecticides from vegetable oil samples spiked at levels
of 5-250 yg/kg were between 70-124%. Routine determinations were carried
out for cyclodienes, EEC isomers, and HCB at the 5-10 yg/kg level and
DDT-type compounds at the 20-30 yg/kg level. Results of collaborative
studies were reported. If PCBs were present, the column was eluted with
10 ml and then 150 ml of hexane. The first fraction contained all the
PCBs and all or most of any residues of aldrin, heptachlor, HCB, £,£f-DDE
and j>,£f-DDT. The second fraction contained all the BHC isomers, heptachlor
epoxide, dieldrin, endrin, £,£*-DDD, methoxychlor, and Endosulfan A. Com-
pounds splitting between fractions included methoxychlor, toxaphene, per-
thane, chlordane, and strobane. Further collaborative study (68) of the
method found it satisfactory for determining residues of hexachlorobenzene
and g-HCH in butterfat and mutton fat; a-HCH, Y-HCH, £,£»-DDT, and £,£f-DDE
in chicken fat; g-HCH, dieldrin, hexachlorobenzene, and TDE In pork fat;
DDT isomers in eggs; and other OC1 insecticides in these and other .samples
of animal origin.
A microcolumn of 2.0 g of Woelm basic alumina deactivated with 11% water has
been used for cleanup of water extracts and fractionation of residues.
Petroleum ether (5 ml) eluted HCB, a- and Y-BHC, heptachlor epoxide (10%),
£,£f-DDE, £,£'-DDT, TDE, £,_p_'-DDT, telodrin, isodrin, aldrin, and heptachlor.
Subsequent elution with 10 ml of petroleum ether-ethyl ether (80:20 v/v)
recovered &-BHC, heptachlor epoxide (90%), dieldrin, and endrin (69).
In a comparative study (70), basic alumina was found to retain lipids
better than Florisil, which in turn held more than silicic acid. It was
also found that deactivation and elution with less polar solvents gave a
superior separation of organpchlorine pesticides from lipids than activated
adsorbents and more polar eluants. Saponification with ethanolic NaOH
followed by alumina column chromatography provided efficient removal of
lipids prior to GC determination of several OC1 insecticides (DDT was con-
verted to DDE) (71). A procedure for evaluation of the fat capacity of an
aluminum column has been described (68).
9P MISCELLANEOUS MULTIRESIDUE CLEANUP PROCEDURES
Other multiresidue procedures include the following: The method of de
Faubert Maunder (72) employs partition with dimethylformamlde (DMF) to
diminish the amount of fat carried over with the pesticides from fatty
samples. A hexane extract of the sample is extracted three times with
hexane-saturated DMF; ,the combined DMF phases are washed with a DMF-
saturated hexane and then shaken with a large volume of 2% Na2SO^ solution.
On standing, a hexane layer containing chlorinated pesticides forms on top
of the solution, this layer is separated, and residues are cleaned-up on
an alumina column and determined by EC-GC. Like the AOAC-MOG procedure
(Section 9A), this method does not give good recoveries of hexachlorobenzene
from fatty samples.
Wood (73) proposed a rapid method for small samples using dimethyl sulfoxide
(DMSO). This is a good solvent for chlorinated pesticides that dissolves
only low amounts of oil or fat. The fatty sample is mixed with Celite
-329-
-------
Section 9P
(1:15 w/w) and packed into a small column, and the pesticides are eluted
with DMSO. The eluate is adsorbed directly onto the top of a larger Florisil
column and the residues then eluted. with hexane from the Florisil. The
method does not seem to he widely used.
The de Faubert Maunder and Wood methods have been compared with the
standard FDA-AOAC Florisil procedure for analysis of chlorinated pesticides
in a variety of foodstuffs (74). No gross general differences were found
in results, but one method might be advantageous for a particular sample
type.
A rapid DMSO-petroleum ether partitioning cleanup method employing test
tubes and syringes in place of separatory funnels was found to recover
60 OC1, OP, and carbamate pesticides at levels > 50%. Losses were found
to be consistent, so the use of correction factors was proposed. Crops
containing 0.1-10 ppm levels were tested for analysis by GC (EC and FPD
detectors), TLC, and HPLC (75).
A reuseable, macroporous silica gel column provided fractionation and
88-105% recoveries of 0.1-1 ppm levels of different classes of pesticides
when eluted with a series of solvents of increasing polarity (76).
Thin layer chromatography (TLC) on 1-5 mm layers can provide cleanup if
a minimal amount of fatty material is present in the extract. Sample is
applied as a streak and developed along with standard marker compounds
on the same plate to allow location of the pesticide zones. These bands
are removed by scraping and are extracted to recover the separated pesti-
cides. Modified layers have been devised with capability for increased
sample loading, e.g., multiband or wedge-layer chromatoplates (77). With
the latter, cleanup and determination can be combined on the same layer
without intervening elution.
The use of ion exchange resins for cleanup of ionic pesticides has" been
reviewed (78). For example, acidic residues such as chlorophenols and
phenoxy acids in extracts of organic tissues, soil, and water will bind
under alkaline conditions to a strong base anion exchange resin. After
washing out impurities, the residues can be eluted from the resin column
by an acidic eluant and determined by EC-GC after appropriate derivatization
reactions (79).
The results of international cooperative studies of OC1 pesticide, PCS, and
Hg residues in wildlife have been reported (80). The analytical methods
were based on extraction, cleanup, and GC determination, but no two labora-
tories used exactly the same procedure. Nonetheless, there was reasonable
agreement among laboratories in analysis of test samples, the coefficient
of variation for different chlorinated compounds ranging from 10-17%.
Collaborative testing of a multiresidue method for chlorinated hydrocarbon
and other fumigant residues among 8 foreign laboratories was successfully
completed, and results were reported (81).
-330-
-------
Section 9Q
ORGANOPHOSPHORUS PESTICIDES
9Q DETERMINATION OF METABOLITES OR HYDROLYSIS PRODUCTS IN HUMAN URINE,
BLOOD, AND OTHER TISSUES.
The determination of intact organophosphorus pesticides in tissue or
blood from suspected poisoning victims is described in Section 6,A,(1)
of the EPA PAM (82). However, in cases of low exposure or in high
exposure cases after several hours, the probability of detecting parent
compounds is greatly reduced because of rapid metabolism (83). In most
instances, the determination of alkyl phosphate metabolites in urine
provides a measure of the extent of human exposure to the parent OP
pesticide. Section 6,A, (2),(a) of the EPA PAM and reference (84) contain
a sensitive and selective analytical procedure for alkyl phosphate and
phosphonate metabolites (hydrolysis products) of important pesticides.
OP metabolites in urine are extracted quantitatively with an anion-exchange
resin after addition of acetone in a 10:1 ratio to precipitate some inter<~
fering compounds. The compounds are eluted from the resin, derivatized
with diazopentane (see footnote on page 6 for precautions when using this
reagent), and the derivatives determined by FPD (P-mode)-GC. If very low
levels of alkyl phosphate metabolites are present, further cleanup on a
2.4 gram silica gel column deactivated with 20% water is•carried out.
Confirmation is by FPD-GC using both the P and S detector modes (recall
that the S-mode is 5 to 10 times less sensitive). Analysis can be made
at the 0.1 ppm level, so that the excretion of alkyl phosphates in urine
can be detected at pesticide levels much lower than those that result in
cholinesterase inhibition. The general class of organophosphate pesticide
(but not the exact compound) involved in the exposure may be deduced by
characterizing the metabolite(s) excreted. These analytical methods have
been applied to the analysis of the urine of rats exposed to a group of
aromatic and aliphatic Of and phosphonate pesticides (85). A
Because of the complexity of this method, routine analyses should be
validated by simultaneous analysis of spiked SPRM's. As outlined in
Section 3, one SPRM is analyzed along with each unknown if only occasional
analyses are performed, or the ratio of SPRM to routine analyses is at
least 10% when larger numbers are involved. Because of the possible in-
stability of urine samples spiked with alkyl phosphates, large samples of
SPRM should not be prepared ahead of time for periodic analyses. A
method for preparation of individual SPRM as needed is detailed in
Section 6,A,(2),(a),XI of the EPA PAM. However, it has been shown (86)
that dialkyl phosphate metabolites do not break down or disappear in urine
samples frozen for up 20 weeks prior to analysis.
Underivatized compounds may accumulate on the GC column after periods of
extended use. Injection of 1 Ul of diazopentane .solution should be made
every two weeks to react with these compounds. If peaks appear following
this injection, the column should be reconditioned (Subsection 41 in
Section 4). Further confirmation of any particular metabolite can be
accomplished by preparing its hexyl derivative.
-331-
-------
Section 9R
Reproducibility of this method is not as good as is desirable for a re-
liable, routine analytical method. This can be seen in Table 3 in EPA
PAM Section 6,A,2,(a), where recovery variabilities of 15% or greater are
reported for six analyses at the two highest spiking levels. In the study
of freezer storage of alkyl phosphate metabolites described above (86),
the method was found to give both low and highly variable recoveries.
Because of the unreliable quantitation obtained, the method currently
described in Section 6,A,2,(a) of the EPA PAM should be considered only
semi-quantitative.
In addition to alkyl phosphates, significant amounts of the corresponding
mono- and dicarboxylic acids are found in the urine of humans exposed to.
malathion. A silica gel cleanup FPD-GC method for determining these
acids as a measure of exposure to malathion has been devised (87). Urine
is extracted$ the extract is alkylated, and derivatized carboxylic acids .
are cleaned up according to a previously published (88) alkyl phosphate
method. Additional cleanup by solvent partitioning with ether and sili'ca
gel chromatography [elution with benzene followed by ethyl acetate-benzene
(10:90 v/v), collected as one fraction]-is also employed, Derivatized MCA
and DCA are determined on a 4% SE-30/6% QF-1 column at 200°C.
A reportedly simple and rapid method for quantitation of the metabolites of
malathion and other OP pesticides has been published (89). The omission of
an extraction at low pH and the mild condition of anion-exchange chromatog-
raphy on QAE-Sephadex prevented degradation of a malathion metabolite that
takes place under strongly acid conditions. Disadvantages of the commonly
used partition fractionation of malathion and malaoxon metabolites were
discussed.
Another new method also employing an ion exchange resin for determination
of mono- and diportic alkyl and aryl phosphates, phosphonates, and thio
analogs in human urine has been reported to have a detection limit of less
than 2 pmole for each of these classes of compounds. The acids were
protonated by passing through a hydrogen-form cation exchange resin.
Benzyl esters were formed by refluxing the column effluent with 3-benzyl-
l-£-tolyltriazene in acetone, partitioned into cyclohexane, and determined
by GC (5% OV-210 column) with a P-mode FPD. Inorganic (^-phosphate did not
interfere, but could be removed by calcium hydroxide precipitation if
desired (90).
Urinary dialkyl phosphate metabolites have also been determined using
l-(4-nitrobenzyl)-3-(4-tolyl)triazene as derivatizing reagent. Urine was
lyophilized, dialkyl phosphates were derivatized, and cleanup was carried
out by anhydrous nickel sulfate adsorption and silica gel chromatography.
GC analysis determined the metabolites at levels as low as 0.01 ppm (91).
9R DETERMINATION OF £-NITROPHENOL (PNP) AND OTHER PHENOLS IN URINE
Urinary PNP, the phenolic metabolite of ethyl and methyl parathion, EPN
(0-ethyl 0-jj-nitrophenyl (phenylphosphonothioate) nitrofen, etc., can be
-332-
-------
Section 9S
measured as an indicator of exposure to these organophosphorus pesticides.
A small volume of urine is hydrolyzed with HC1 to form free PNP, then made
alkaline and cleaned-up by extraction with .benzene-ether, and finally re-
acidified and extracted with benzene-ether to remove PNP. An aliquot of
dried extract is analyzed by EC-GC with on-column conversion of PNP to the
volatile trimethylsilyl derivative [EPA PAM, Section 6,A,(2),(b)].
A multiresidue analytical procedure for halo- and nitrophenols from a range
of biodegradable pesticides (organophosphates, phenoxy acids, organohalides)
is also useful for determining exposure to these pesticides (92, 93). A
one to five ml sample is treated with a 1/5 volume of concentrated hydro-
chloric acid and the mixture refluxed at 100°C for one hour. The phenols
are extracted with diethyl ether, ethylated by reaction with dlazoethane,
and the ethyl-ethers chromatographed on a silica gel column (2 grams, 2%
water deactivation). (See the footnote on page 6 concerning precautions
when using diazoalkanes)* Elution with various concentrations of benzene
in hexane purifies and fractionates the phenolic ethers, which are finally
determined by EC-GC.
Ten phenols, including the pesticides pentachlorophenol- and DNOC
(4,6-dinitro-o_-cresol), plus the herbicides 2,4-D, 2,4-,5-T, and silvex can
be determined by this scheme on one sample. All halogenated phenols are
eluted with 20% benzene-hexane, while nitrophenols and phenoxy acids elute
in the 60 and 80% fractions. The phenoxy acids are detected intact along
with 2,4-dichlorophenol and 2,4,5-trichlorophenol, their potential mammalian
metabolites.
A method for the determination of residues of the herbicide DNBP
(2-sec-butyl-4,6-dinitrophenol) in feed, blood, urine, feces, and tissues
by EC-GC has been devised in the EPA Health Effects Research Laboratory (94).
After extraction, the sample is reacted with diazomethane (see footnote on
page 6 concerning precautions when using diazoalkane) to produce the methyl
ether of DNBP."' Cleanup and recovery of the derivative is obtained on acid
alumina column eluted with hexane-benzene (40:60 v/v). Average recoveries
of greater than 85% were obtained from samples fortified at 0.1-30 ppm
levels.
9S SWEEP CO-DISTILLATION
Sweep co-distillation has proven to be a simple time saving cleanup technique
that eliminates the need for specialized adsorbents and large volumes of
purified solvents (8, 95-100). The technique can be used for OC1 and OP
residues in fruits and vegetables, or fats and oils. The procedural details
are different for the two sample types; however, the cleanup principle is
essentially the same. The concentrated sample, in an organic solvent, is
injected into a heated tube swept with 600 ml N2/min. Sample extractives
remain in the tube while volatilized components are swept into a simple
condensing train. After a 30 minute sweep time, the transfer lines are
disconnected and condensed pesticides are rinsed with organic solvent into
the sample tube. After volume adjustments, the sample may generally be
-333-
-------
Section 9S
analyzed by GC without further cleanup. If sensitivity levels in the low
part per billion range are desired, an auxiliary cleanup is•recommended.
The combination of sweep co-distillation and the micro Florisil column
[EPA PAM Section 5,A,(2)] has proven to be a thorough cleanup for fat
samples. A sulfuric acid/Celite column can be adapted as an optional
automatic cleanup step (101).
Figure 9-A is a schematic diagram of the apparatus as originally used for
cleanup of fruits and vegetables for determination of OP residues. The
glass wool-packed tube was placed inside a heated copper tube. A nitrogen
sweep of 600 ml/min was used. Two gram aliquots of sample were injected
followed by ethyl acetate injections every three minutes. See FDA PAM
Section 232.2 for method details.
OC1 and' OP residues in a variety of edible fats and oils have been
determined by a modified version of the sweep co-distillation cleanup
system (102,- 103). A tube packed with glass wool, sand, and glass beads
is operated in a vertical position with the injection port on bottom.
The cleanup is effected by the 250° heat and the nitrogen carrier gas
distributing the oil upward through one-half to three-fourths of the glass
bead packed column with a percolation type action. Pesticides are volatilized
and swept into the collector trap. Recent study of sweep co-distillation of
fats has shown that follow-up injections of solvent are not necessary. After
initial injection of the sample, the equipment may be left unattended for the
30 min sweep operation.
Figure 9-B shows the appearance of a commercial version of the "Sweep
Co-distiller" (Kontes Glass Co., Vineland, NJ). This apparatus permits
Simultaneous cleanup of four samples with a 30 min sweep time. The 30 cm
tube allows efficient cleanup for OC1 or OP residues in samples of fats,
oils, milk, and crops (operated in vertical position at 250°C with 600 ml
N2/min) (104). The tube for fat cleanup may be purchased prepacked, but
packing in the laboratory is preferable for consistent tube uniformity.
The empty tube may also be prepared for fruit and vegetable cleanup by
packing withC'15 cm glass wool in the injection end with remaining space
filled with glass beads. The oven would be swiveled to a horizontal position
for the fruit and vegetable cleanup. Operational parameters for the latter
application may be found in the FDA PAM, Section 232.2.
In a preliminary evaluation of the Kontes apparatus by Watts, common
organochlorine pesticides were quantitatively recovered from chicken fat,
and the fat residue was reduced to less than 1% of the original sample.
Similar results have been obtained by Luke with both the Kontes and labora-
tory-assembled sweep-co-distillation units. An oven temperature of
227-230°C was used, and no solvent injections were made after the sample
was applied. Reproducible, quantitative recoveries were obtained for
organophosphorus and organochlorine pesticides from beef fat and butterfat
(105).
The operating principle of sweep co-distillation has been presented diagram-
matically, and recoveries of 36 OP pesticides in 20 substrates (0.03-0.5 ppm)
and 30 OC1 pesticides in 14 substrates (0.003-0.05 ppm) are tabulated (106).
-334-
-------
Section 9S
Figure 9-A. Sweep co-distillation apparatus, schematic diagram.
^Insulation
.-Asbestos
-Heating tape
To pyrometer
Septum .
Scrubber
tube
Water
and ice
bath
Storherr tube
containing 5-6 inches
silanized glass wool
Nitrogen
-Coding in|e*
coil
_ Glass wool
(silanized)
—4cm Anakrom
-Adapter
-Glass wool
(silanized)
-J 19/22
"'Concentration
tube
-Beaker
of water
Figure 9-B. Sweep co-distillation apparatus, (oven positioned for
fruit and vegetable cleanup), Kontes Glass Co.,
K-500750.
-335-
-------
Sections 9T, 9U
Heath and Black (107) recommended the following modifications for faster
and more convenient cleanup of organochlorine residues in animal fat: no
solvent introduction; 230°C distillation temperature; 600 ml/minute nitrogen
flow; 6.7 mm distillation tubes with simplified packing; and incorporation
of a U-tube condenser that allows direct introduction, onto a Florisil column
for secondary cleanup.
A literature review of the applications of sweep co-distillation including
a comparison to other cleanup methods has been published (108).
9T CHARCOAL CLEANUP OF NONFATTY FOOD EXTRACTS
A general determinative method for organophosphorus pesticide residues in
nonfatty foods is based on the FDA acetonitrile (or water/acetonitrile)
extraction procedure followed-by dilution with methylene chloride to
separate water, cleanup on a short charcoal column, and analysis by GC with
a P-selective detector. The chromatographic tube (300 mm x 22 mm id) is
packed dry with a one gram layer of Celite 545 followed by 6 grams of
adsorbent mixture (acid-treated Norit SG-X or Nuchar C-190 charcoal-hydrated
magnesium oxide-Celite 545, 1:2:4 w/w) and finally glass wool topping, and
the column is eluted with acetonitrile-benzene (1:1 v/v); The satisfactory
recovery of 41 pesticides and alteration products from kale and 9 typical
pesticides from other low and high sugar content crops was demonstrated
(109). A collaborative study (110) of this method for residues of six OP
compounds in apples and green beans verified recoveries between 86 and
125% when either a thermionic or FPD detector was employed. The method
is described in the FDA PAM, Section 232.3., and recoveries of 51 pesticides
and related chemicals are listed in Table 201-H of the FDA manual. Sections
4Ae and f of this Manual describe procedures for purification of Celite
and carbon adsorbents, 'respectively.
9U ACETONE EXTRACTION .,"
The FDA PAM contains details of a procedure for determination of polar
organophosphate and organonitrogen pesticides in nonfatty samples (FDA PAM
Sections 232.4 and 242.1). Samples are blended with acetone and filtered,
pesticides are extracted from the aqueous filtrates into petroleum ether-
methylene chloride, and an aliquot of concentrated extract is determined
by GC with a P- or N-selective detector. Lack of a column cleanup step
allows determination of many polar compounds that would not be recovered
from adsorbents such as Florisil or charcoal, but a specific detector,
rather than electron capture, must be used. Repeated injection of impure
extracts can shorten column life, so that packing material at the head of
column will need to be replaced often. A short (0.6 or 0.9 m) column of
& polar phase, such as DEGS or Carbowax 20M, will probably be advantageous
for the chromatography of polar compounds. If it is desired to examine
some pesticides with the electron capture detector, cleanup of acetone
extracts of nonfatty foods is carried out on a Florisil column (FDA PAM
Section 212.2). A list of pesticides recovered through these procedures,
with and without Florisil cleanup, is given in the FDA PAM, Table 201-1.
_
-336-
-------
Section 9V
9V MISCELLANEOUS MULTtRESIDUE CLEANUP PROCEDURES
Nine extraction procedures were compared for efficiency of removal of six
OP pesticides and metabolites from field treated crops. Soxhlet extraction
of the finely chopped crops with chloroform-methanol (90:10 v/v) proved
most reliable and efficient (111).
Alumina has not proven totally satisfactory;for cleanup of OP compounds
since recovery of the more polar compounds is not complete (112). Using
alumina (activity II to III) and petroleum ether and petroleum ether-
acetone (97:3 v/v) as eluants, Renvall and Akerblom (113) eluted only 13
of the 31 OP compounds they tested. However, many residue analyses are
based on alumina column cleanup, e.g., the determination of carbophenothion
in goose tissues (114) and monocrotophos in tobacco (115) by FPD-GC.
The Abbott et al. method (116), involving cleanup by solvent partition with-
out column chromatography, has proven adequate for analyses of seven types
of foods for 39 pesticides and metabolites when detection was made with a
thermionic detector. Finely chopped sample is mixed with anhydrous sodium
sulfate and extracted with acetonitrile. The extract is diluted with a
large volume of aqueous sodium sulfate, and the pesticides are extracted
into chloroform. The chloroform solution is dried and concentrated for
GC. Other determinations without column cleanup have been reported.
Methyl parathion, diazinon, malathion, and phorate were,determined in plant,
animal, water, and soil samples by EC-GC following only hexane extraction
and partition with aqueous acetonitrile (117). Azinphosemethyl and
dimethoate residues in apple leaves were determined by FPD-GC following
ethyl acetate extraction and cleanup by methylene chloride-water and hexane-
kcetonitrile partitionings (118). A multiresidue analysis of 14 pesticides
'?,n natural waters at ppb levels involving extraction and concentration
before FPD-GC has been reported (119).
The elution pattern of a series of representative OP pesticides from a
column (Kontes, Size 22) containing one gram of Wbelm silica gel deactivated
with 1.5% water and prewashed with 8 ml hexane before applying the sample
mixture is as follows:
Eluant
7 ml hexane
8 ml 60% benzene-hexane
8 ml benzene
8 ml 8% ethyl acetate-benzene
8 ml 50% ethyl acetate-benzene
Pesticides Eluted
carbophenothion
ethyl parathion
malathion; diazinon
paraoxon
-337-
-------
Section 9V
Silica gel or silicic acid columns have been used for cleanup of animal,
plant, soil, and water extracts prior to GC determination of OP pesticides
(120-122) and to separate OP pesticides and metabolites into groups to
facilitate their identification by GC (123). A tandem column of silica
gel and alumina was used to separate leptophos and its oxygen and 2,5-
dichlorophenol analogs prior to determination by FPD-GC (124).
A rapid, simple approach has been developed for approximately the total
residues of pesticides such as fenthion, disulfoton, and phorate, which
may consist of the parent pesticide and up to five metabolites formed by
oxidation of thionophosphate and sulfide groups in each molecule. The
insecticides and any metabolites are oxidized to t;he oxygen analog sulfone
with m-chloroperbenzoic acid, followed by removal of the acid on an alumina
column and determination of the sulfone by ITD-GC. Quantitative re-
coveries of parent pesticides and metabolites from corn, milk, grass,
and feces have been demonstrated (125). Metasystpx-R and its sulfone
were determined in plant and animal tissues and water at 10 ppb levels as
the sulfone after oxidation by KMnO^(126). '
A method for 40 organophosphorus pesticide residues in plant material
involved extraction with acetonitrile, partition into methylene chloride,
and GC with a P-selective thermionic detector (127). The same authors
used this extraction and cleanup procedure for plant material subsequent
to oxidation with potassium permanganate to convert organophosphorus
pesticides containing thioether groups (e.g., demeton, disulfoton, phorate)
to sulfones (128).
A collaborative study by 12 laboratories of the methods of Abbott et al.
(116; see above, this Subsection), Watts ££ al., and Sissions and Telling
was conducted for OP pesticides in fruits and vegetables (129). The method
of Abbott _et al. was found satisfactory for determination of malathion,
dichlorvos, dimethoate, omethoate, and parathion in 6 fruit and vegetable
crops (>90% average for all pesticides and crops at 0.5-2 mg/kg) and was
judged widely applicable to the determination of many other nonpolar and
medium-polarity OP pesticides and to a wider range of samples. The method
of Watts et al. (130), involving ethyl acetate extraction and cleanup on
a column of activated charcoal-magnesium oxide-Celite eluted with ethyl
acetate-acetone-toluene (an early version of the procedure described in
Section 9T), was found satisfactory for the same pesticides plus azinphos-
methyl in 6 crops (>90% average recovery at 0.5-2 mg/kg) and was also Judged
to be much more widely applicable. The method of Sissions and Telling (131),
employing cleanup by batch addition of charcoal followed by hexane and
hexane-acetone (98:2 v/v) elution through an activity -5 alumina column
was not successful for the more polar pesticides studied. Details and
modifications of these methods are discussed in the report of the collabora-
tive study.
-338-
-------
Section 9W
The methods of Abbott et al. (.116) and Sissons and Telling (112) and the
sweep co-distillation method (Section 9S) were compared for determination
of different OP pesticide residues in various vegetable crops. There was
no significant difference for most pesticide-crop combinations, except
that sweep co-distillation tended to give lower results for polar compounds
such as omethoate (132).
A tabulation has been made (133) of the validated applicability of five
multiresidue analytical methods to the .determination of some 50 OP
insecticides, acaricides, and nematocides. These procedures were the
AOAC (llth ed.) 29.001-29.027 general Florisil cleanup method for OC1
and OP pesticides; the AOAC (llth ed.) 29.028-29.033 multiple residue
carbon column cleanup method for OP pesticides; the AOAC (llth ed.)
29.034-29.038 single sweep oscillographie polarographic confirmatory
method; the Abbott\et al. method for total diet studies (116); and an
undescribed German procedure (134). In addition, individual determina-
tions of some of the compounds by other special methods were reviewed.
It was stated that, in general, the multiresidue methods were not usually
suitable for metabolites, requiring separate analysis for the parent
and metabolite; each method should be compound validated in the worker's
own laboratory; and that differences in results were more likely to arise
from sampling problems than from the analytical methods' themselves.
The use of a selective detector sometimes allows determination of OP
pesticides with no cleanup. Por example, a collaborative study of the
analysis of wheat for chlorpyrifos methyl, fenitrothion, malathion,
methacrifos, and pirimiphos methyl involved only methanol extraction for
40 hours followed by GC of an aliquot using a FPD or alkali flame ioniza-
tion detector (135).
CARBAMATE PESTICIDES AND METABOLITES AND MISCELLANEOUS. HERBICIDES
9W 1-NAPHTHOL IN URINE
Humans exposed to the N-methyl carbamate insecticide carbaryl excrete in
urine relatively large quantities of the metabolite 1-naphthol conjugated
as either the sulfate or glucuronide. Determination of 1-naphthol is
made by subjecting 5 ml of urine to acid hydrolysis under reflux to break
conjugates, extracting the 1-naphthol with benzene, and derivatization
with chloroacetic anhydride solution. After cleanup on a small silica
gel column (1 gram, 1.5% water), the derivative is quantitated by EC-GC
against a peak from standard 1-naphthol similarly derivatized. Details
are found in Section 7,A of the EPA PAM.
Elution patterns from the silica gel column must be established at the
temperature and humidity conditions prevalent in each laboratory. Spiked
control urine treated in the same manner as routine samples is used for
this purpose. Traces of water can affect the derivatization reaction and
must be avoided. Derivatized standards are stable for about 6 months
if stored in a refrigerator.
-339-
-------
Sections 9X; 9Y
9X ANALYSIS OF AMINE METABOLITES IN URINE
A method for determination, of amine metabolites from anilide, urea, and
carbamate pesticides was developed in the EPA Research Triangle Park
Laboratories (136). Pentafluoropropionic anhydride was the preferred
derivatization reagent for the aniline compounds, with cleanup on 1 gram
deactivated (3) percent water) silica gel columns. Determination was by
EC-GC on a 3% OV-1 column. Recoveries ranged from 85-90% at 1.0 and
0.1 ppm.
9Y OTHER INDIRECT (DERIVATIZATION) METHODS OF ANALYSIS
Numerous derivatization methods have been used for the indirect measure-
ment of residue levels of parent carbamate insecticides in a variety of
agricultural crops and other substrates. These have involved derivati-
zation of the amine or phenol moieties of the pesticides after hydrolysis,
or, less often, the intact insecticide. These derivative methods include
reaction of intact insecticides with bromine, silylating reagents, acetic
anhydride, and trifluoroacetic anhydride. Phenols resulting from alkaline
hydrolysis of the parent insecticides have been reacted with bromine
(with or without simultaneous esterification), silylating reagents,
mono- and trichloroacetyl chloride, pentafluorobenzyl bromide, and
l-fluoro-2,4-dinitrobenzene. The latter reagent is used for derivatization
of carbamate insecticides in the method for water analysis (Section 9A,C)
discussed in this Manual and described in detail in the EPA PAM, Section
10, A.
Amine hydrolysis products of carbamate insecticides have been reacted with
l-fluoro-2,4-dinitrobenzene and 4-bromobenzoyl chloride. These and other
reactions have been surveyed in a review article (137) in which pertinent
references are given.
GC methods for phenyl substituted urea and carbamate herbicides are usually
based on hydrolysis followed by determination of the corresponding aniline.
Anilines have been derivatized with halogen, 4-chloro-a,a,a-trifluoro-3,5-
dinitrotoluene, l-fluoro-2,4-dinitrobenzene, and pentafluoropropionic
anhydride. These reactions are also reviewed in reference (137).
A fluorogenic labeling derivatization reaction with dansyl chloride has"
been combined with EPLC for the determination of N-methylcarbamate insecti-
cides in soil and water. No preliminary cleanup was required, and de-
tection limits were 1-10 ng/4 pi injection (138).
Ten triazine herbicides were determined in vegetables at levels of
0.13-0.86 ppm by preparation of heptafluorobutyryl derivatives. Compared
to the parent compounds, the products were at least 300 times more sensi-
tive to electron capture detection and 5-10 fold more sensitive to Cl-mode
electrolytic conductivity detection (139).
-340-
-------
Sections 9Z, 9A,A
9Z DIRECT METHODS OF ANALYSIS
Determinations of intact, underivatized N-methylcarbamate insecticides
are hampered by their decomposition on GC columns under ordinary operating
conditions (140). Losses can be minimized by the use of specially prepared,
conditioned, and maintained .columns. Presilanized supports do not provide
sufficient deactivation to prevent degradation of carbamates, so it is
necessary to employ in situ silanization both during initial conditioning
and thereafter to restore column performance. Examples of direct analyses
of crop extracts include a multiresidue method (141) on 5-6% DC-200 after
acetonitrile partition and charcoal cleanup as for OP pesticides (109),
determination of carbofuran and other carbamates on 20% SE-30 (142), and
determination of 0.2-15 ng of carbaryl and 1-naphthol on a short column
of 3% SE-30 (143). Highly deactivated GC columns prepared from acid
washed Chromosorb W support that is surface modified with Carbowax 20M
have also been successfully used for chromatography of intact N-methyl-
carbamates without degradation on the column. Such columns, which are
extremely promising for performing analyses without required derivatization,
are described in Sections 4J and 5Lb of this Manual.
Urea and N-arylcarbamate herbicides are, in general, more thermally stable
than carbamate insecticides, and' are, therefore, more amenable to direct
determination by GC, For example, columns of 5% E-301 methyl silicone
at 150°C (144), 10% DC-200/15% QF-1 (1:1) at 160°C (145), and'5 and 10%
DC-200 (146) have been successfully used, the former for multiresidues of
urea herbicides and the latter two for carbamate herbicides in foods. '
However, decomposition of compounds on these column types has been noted
under certain conditions, and determinations are therefore often made via
thermally stable derivatives of hydrolysis products or directly on Carbowax
20M-treated columns. As an example of the latter, carbamate insecticides
and herbicides have been directly chromatographed on Carbowax 20M modified
(Ultra-Bond) supports containing 1-3% of a liquid phase such as OV-17,
OV-101, or OV-210. The Hall electrolytic conductivity detector was used,
and determinations in soil were demonstrated (147),
s-Triazine herbicide residues were determined in urine by hexane extraction
from a sample at pH 12, drying of the extract by passage through a sodium
sulfate column, concentration of the extract, and GC using a N-mode Hall
conductivity detector (148). Similar N-specific GC methods involving
cleanup were used to monitor triazines in European streams (149).
9A.A
ANALYSIS OF PLANT AND FOOD MATERIALS
Extraction of urea and carbamate pesticides from plant materials usually
involves blending with methylene chloride, acetone, chloroform, acetoni-
trile, or an alcohol (or these solvents plus anhydrous Na2S04). If the
presence of conjugates of hydroxy metabolites is suspected, hydrolysis
with an acid during extraction may be included (Section 9A,L).
Cleanup steps include solvent partition and/or liquid column chromatog-
raphy, the exact nature of which are pesticide- and sample-dependent.
-341-
-------
Section 9,A,B
For example, a column of 4:1 MgO-cellulose was used for cleanup of
carbamate herbicide residues from a variety of foods (145), while Florisil
was employed after acetonitrile-petroleum ether partition for the multi-
residue, multiclass determination of carbamate, urea, and amide residues
(150). Methods for extraction, cleanup, and GC of carbamates, ureas,
and other classes of herbicides (triazines, uracils, phenols) have been
reviewed (151-153). A multiresidue method for twelve triazine herbicides
in crops, water, and soils involving methanol extraction, alumina column
cleanup, and gas chromatography with a Carbowax column and thermionic,
microcoulometric, FPD, and electrolytic conductivity detectors has been
reported (154). Residues of 15 organonitrogen herbicides and fungicides
were screened in foods by acetone extraction, partition and Florisil
(2Z water) column cleanup, and CCD-GC determination (155). Herbicides
of different types were determined in crops at tolerance levels with no
column cleanup prior to GC with N- and Cl-mode conductivity detection
(156). Total .residues of Mesurol and dts sulfoxide and sulfone metabolites
in plant and animal tissues were determined by oxidation of the extract
with KMn04 to convert all residues to the corresponding sulfone, which
was detected at a limit of 0.03 ppm by a S-mode FPD (157).
9A,B AIR ANALYSIS
Section 8,B of the EPA PAM contains details of analytical methods for
chlorinated, organophosphorus, and N-methylcarbamate insecticides
collected by one of the procedures described in Section 8,A of the
EPA PAM or Section 8H of this Manual. The sampling medium is extracted
with hexane-diethyl ether (95:5 v/v). Chlorinated pesticides and PCBs
are measured by EC-GC after column chromatographic cleanup on alumina.
PCBs are separated from technical chlordane and other pesticides by
column chromatography on silicic acid deactivated with distilled water.
Organophosphorus pesticides are determined by direct injection of an
aliquot of extract into a chromatograph equipped with a flame photometric
detector. Carbamate pesticides are determined directly by GC using a
N-mode Hall detector and a 3% OV-101/Ultra Bond 20M column. As an
alternative for carbamates, derivatization is carried out with o-bromo-
2,3,4,5,6-pentafluorotoluene. The derivatives are cleaned-up and
fractionated on a column containing-1 g of deactivated silica gel and
determined by EC-GC. Collection efficiencies of OC1 and OP pesticides
and PCBs using different samplers and collection media as determined
with these analytical procedures are tabulated in Section 8,B of the
EPA PAM.
The air analysis method reported earlier in the EPA PAM was a multiclass,
multiresidue procedure (158) for residues collected in ethylene glycol,
in which prefractionation was carried out on a 1 g column of silica gel
deactivated with 20% water. A still earlier method for air analysis
(159) included Florisil column chromatography and was the basis of the
former EPA National Pesticide Monitoring Program (160). Details of
these can be found in earlier editions of the EPA PAM, but use of the
current, more widely applicable and tested procedures, described above,
is generally recommended.
-342-
-------
Section 9A,C
9A,C WATER ANALYSIS
A broadly applicable multiresidue, multiclass method for the monitoring
of water samples for pesticides is presented in Section 10,A of the
EPA PAM (161). Recovery studies were conducted on 42 halogenated com-
pounds, 38 OP compounds, and 7 carbamates. Recoveries of >80% were
achieved for 58 of the 87 compounds, 60-80% recovery for 13 compounds,
and <60% for the remaining 16 compounds (concentration levels 0.09-400 ppb).
Pesticides are extracted from water with methylene chloride, and the con-
centrated extract is chromatographed on a 1 gram deactivated (20% water)
silica gel column with four different solvents of increasing polarity to
separate the pesticides into groups. OC1 compounds are determined by
EC-GC, OP compounds by FPD-GC, and carbamates by EC-GC after conversion
to 2,4-dinitrophenyl ether derivatives. Low recoveries were in most
cases traced to losses during the silica gel chroma tography step.
Evaporation of solutions by air blowdown should not be used because losses
of all three classes of pesticides may occur. Concentrations are carried
out under a gentle stream of nitrogen. It is important to apply the con-
centrated extract to the silica gel column at the exact moment the last
of the hexane prewash reaches the top surface of the column. The total
0.5 ml extract plus the 1.0 ml hexane rinse must be transferred to the
column without loss to minimize the recovery error. Solvents contained
in several eluate fractions from the silica gel column may interfere in
the GC and carbamate derivatization steps. It is critical to follow the
directions for solvent removal and exchange outlined"in Section 10,A of
the EPA PAM. Sufficient silica gel should be activated (at 175°C) to
provide only a one-week supply, and deactivation should be carried out
only on the amount required for a 2 or 3 day. period. Longer storage
periods may result in a shift of the pesticide elution pattern of the
final deactivated columns. Each lot of silica gel should be tested for
the proper elution pattern with representative pesticide standards
eluting in each fraction. A number of the OP compounds require con-
siderable column pre-conditioning by repetitive injection of high-
concentration standards in order to obtain linearity of response and
accurate quantitation. Confirmation of pesticide identity should be
made by several techniques outlined in Section 10.
Section 10,B of the EPA PAM describes the determination of some free
acid herbicides (e.g., MCPA, 2,4-D, 2,4,5-T) in water. The water is
adjusted to pH 3 and extracted with methylene chloride. The extract is
taken to dryness, pesticides are esterified with 10% BCls in 2-chloroethanol,
and the resulting esters are extracted with hexane, concentrated, and
determined by EC-GC. If cleanup is required, chromatography on silica
gel deactivated with 20% water is employed. This procedure is a further
extension of the multiclass, multiresidue procedures described directly
above. When preparing the BC^-chloroethanol esterification reagent,
work in an efficient exhaust hood and wear disposable vinyl gloves because
2-chloroethanol is toxib by dermal contact or when inhaled.
The reagent is stable for at least thirty days if kept stoppered and
refrigerated. As usual, spiked reference material containing the same
pesticides at comparable concentrations as in the sample (if these are
-343-
-------
Section 9A,C
known) should be analyzed in parallel. Other aspects of quality control
are as discussed in the preceding paragraph. BCl3-methanol was also
chosen in another study (162) as the best derivatization reagent for
the determination of 8 phenylalkanoic acid herbicides in water
(0.01-2.5 yg/L); solvent partition and silica gel (5% water deactivated)
minicolumn cleanup and EC-GC with an OV-17/QF-1 column were employed.
The 1979 Analytical Methods Manual of the Inland Waters Directorate,
Water Quality Branch, Environment Canada, Ottawa, contains detailed
methods for the analysis of organochlorinated pesticides and PCBs,
organophosphorus pesticides (two procedures), phenoxy acid herbicides
(two procedures), pentachlorophenol, and N-methyl carbamates in waters.
The method for organochlorines, employing benzene extraction, Florisil
column cleanup, and EC-GC, has detection limits ranging from 0.001-0.01 ppb.
The first OP procedure determines dimethoate, fenitrothion, and phosphamidon
and the second determines 14 other OP pesticides, all at 0.005-0.1 ppb
levels by FPD-GC without cleanup. Phenoxy acid herbicides (2,4-D; 2,4,5-T;
Silvex) are extracted with chloroform from acidified water and converted
to their methyl esters utilizing BF3-methanol prior to cleanup on a
Florisil column and EC-GC determination at 0.01 ppb levels. A second
procedure determines 8 phenoxy acid herbicides at 0.01-2.5 yg/L levels
by extraction of acidified water with ethyl acetate, back extraction of
the polar herbicides into KHCOg, further concentration of acids by methy-
lene chloride extraction to a final volume of 1 ml, esterification with
BCl3/2-chloroethanol reagent, and EC-GC of the resultant 2-chloroethyl
esters. A separate procedure for MCPA (4-chloro-2-methylphenoxyacetic
acid) and MCPB [4-(4-chloro-2-methylphenoxy) butyric acid] in natural
water at 0.1-0.2 yg/L levels is based on extraction from an acidified
sample with methylene chloride, derivatization to pentafluorobenzyl esters,
cleanup and fractionation on a silica gel column, and EC-GC determination.
PCP is detected at 0.01 yg/L by benzene extraction from acidified water,
partition into potassium carbonate solution, acetylation with, acetic
anhydride, partition into hexane, and EC-GC. Five N-methyl carbamates are
determined at 0.10-1.0 yg/L. levels by extraction from acidified water with
methylene chloride, partition with base to remove phenols and acids present
in the extract, hydrolysis with methanolic KOH to the respective phenols,
extraction of the phenols with methylene chloride and derivatization with
penafluorobenzyl bromide, cleanup and fractionation of the ether derivatives
on a silica gel microcolumn, and EC-GC of the column eluates.
Cleanup is often not required for EC-GC analysis of surface water samples
(163) and is usually not required for any type of water if a selective GC
detector is employed. For example, the multiresidue analysis of 14 OP
pesticides in natural waters has been carried out at ppb levels by
extraction, concentration, and direct GC with a FPD detector in the P- and
S-modes (119). Results of an interlaboratory study of the analysis of 15
water samples for 10 OC1 pesticides without any column cleanup have been
reported (164). Where needed, cleanup and separation of common chlorinated
and OP insecticides extracted from water have been successfully carried
out in silica gel microcolumns (165, 166) and columns of deactivated
(5-20% H20) silica gel (above) and alumina (167).
-344-
-------
Section 9A,D
Extracts of water, sediment, sludge, sewage, and soil often contain
large amounts of elemental sulfur, which interfere in the GC analysis
of early eluting pesticides with the EC or FPD detectors. Chemical
desulf urization with Raney copper powder (168) or copper ribbon (169) ,
precipitation with metallic mercury (170) , reaction with CN (171),
and treatment with tetrabutylammonium sulf ite to produce an ion pair
with sulfur as 8203" (172) have been used to remove such interference.
See also Subsection 9A,D) .
Polar phosphorus, urea, and carbamate pesticides are extracted from
water with more polar solvents such as chloroform or methylene chloride.
Extraction of acidic or basic compounds is aided by adjusting the water
sample to a controlled pH value. An XAD macroreticular resin can also
be used for residue isolation and collection. Determination by GC is
carried out using an appropriate selective detector after extract con-
centration and any required cleanup and /or derivatization steps. As
an example, carbaryl and 1-naphthol have been determined in natural
water at 2.5-10 ppb levels (82-102% recovery at 5 ppb) by -EC-GC after
methylene chloride extraction, cleanup on an XAD-8 column, and derivatiza-
tion with heptaf luorobutyric anhydride reagent (173) . Sixteen organo-
phosphorus pesticides were determined. in drinking water at ng/liter levels
by extraction with Amberlite XAD-2 resin, elution from the resin with
hexane-acetone (85:15 v/v), and GC of the concentrated effluent using a
nitrogen-phosphorus selective detector (174) .
Chlorophenoxy herbicides and their esters have been determined by adjusting
the water sample to pH 2, extracting with benzene or diethyl ether,
methylating the acids with diazome thane or BF3-methanol, followed by
gas chromatography with an electron capture or microcoulometric detector
(175) (see the footnote on page 6 concerning the hazards of diazome thane) .
PCP has been determined in marine biota and sea- water by EC-GC of the
amyl diazohydrocarbon derivative after Florisil cleanup (0.002 ppb) and
by HPLC of the free phenol without cleanup (2 ppb) (176).
TLC determinations of carbamate, urea, triazine, and uracil herbicide
residues in water have been reviewed (137, 177), as have the extraction,
cleanup, GC determination, and confirmation of chlorinated insecticides
in water and soils (178).
9A,D SOIL, HOUSE DUST, AND BOTTOM SEDIMENT
The analysis of soil and house dust for organochlorine pesticides is
described in Section 11, A of the EPA PAM. Homogenized samples are
Soxhlet-extracted with acetone-hexane, extract is concentrated in a
K-D evaporator, and cleanup carried out on successive aluminum oxide
and Florisil columns. Eluates are concentrated as required and deter-
mined by EC-GC. A similar AOAC method has been declared official final
action for residues of aldrin, £,.p_'-DDE, jgi^'-DDT, o.,£*-DDT, £,£f-TDE,
dieldrin, endrin, heptachlor, heptachlor epoxide, and lindane (179).
Section 11, C of the EPA PAM references a procedure (180) for direct
GC determination of carbamate pesticides in soils using Carbowax 20M-
modif ied supports and the Hall electrolytic conductivity detector.
This method is now being investigated by the EPA for possible future
inclusion in the EPA PAM.
-345-
-------
Section 9A,D
Sediment samples are partially air dried, mixed with sodium sulfate,
and packed into a chromatographic column. The pesticides are extracted
from the column by elution with hexane-acetone (1:1 v/v). The extract
is washed with water to remove acetone, and the pesticides extracted
from water with 15% methylene chloride in hexane. The extract is dried
with sodium-sulfate, concentrated to a suitable volume, and cleaned-up
on a Florisil column. After desulfurization with copper, determination
of organochlorine pesticides is by EC-GC. Details of the entire pro-
cedure are presented in Section 11, B of the EPA PAM. Air drying of the
sample required 1-3 days, depending on the soil type. Such samples will
contain at least 50% water. Pesticide concentrations are expressed on
a "dry" basis, requiring determination of the dry weight of sediment by
weighing a separate, air-dried sample before and after heating overnight
at 100-110°C. Storage of soils in light can cause formation of artifacts
of OC1 pesticides (181). Moistening of dried soil with water (e.g.,
80ml/30Qml) may increase extraction of pesticides by solvents such as
hexane-isopropanol (3:1 v/v) (182).
Sediment samples may contain elemental sulfur that will be recovered
through the normal extraction and cleanup procedures for organochlorine
and organophosphate pesticides and detected by the EC,-FPD (P- or S-modes>,
and conductivity detectors. With the recommended GC columns and operating
parameters, sulfur can completely mask the chromatogram from the solvent
peak through the aldrin peak. The technique described in Section 11,B,VX
of the EPA PAM for desulfurization employs vigorous agitation for one
minute with bright metallic copper. Some pesticides may be degraded by
this treatment (e.g..^ OPs, heptachlor), but these are not likely to be
found in routine sediment samples because of breakdown in the aquatic
environment. The procedure should be carried out if the presence of sulfur
is indicated by an exploratory injection from the final extract concentrate
or if sulfur crystallizes out when the 6 and 15% ethyl ether eluates from
the Florisil column are concentrated. During determination of atrazine
residues in soil containing-high levels of ammonium nitrate fertilizers,
the response produced by the N-thermionic detector was not constant for
standards and samples due to the presence of the fertilizer in the sample
extracts (183.).
Part 5 of the 1979 Environment Canada Analytical Methods Manual, Inland
Water Directorate, Ottawa,. Canada, contains a method for organochlorine
pesticides and PCBs in sediment and fish. Nineteen compounds are determined
at 0.001-0.05 mg/kg levels by extraction of previously frozen samples with
acetonitrile, partition: with petroleum ether after appropriate dilution
with water, and cleanup and separation into four fractions on a Florisil
column. Each fraction is determined by EC-GC. Sulfur is removed by
precipitation with copper powder or mercury.
Nine chlorinated insecticides were determined by a modified GC procedure
(184) with recoveries of 75-99% from suspended sediment and bottom
material. Extraction was with acetone and hexane added separately,
coextractlves (including PCBs) were Isolated by alumina and silica gel
column chromatography, and EC-GC was used to analyze the various column
eluates. Some soil analyses have been carried out by EC-GC with tut
required column cleanup (185), but this is not common.
-346-
-------
Section 9A,D
Shake (blending) Soxhlet, and column extraction methods were compared
for efficiency in removing some twenty chlorinated insecticides from a
sandy loam soil. There was no statistical difference aiaong the three
methods for the majority of pesticides, but shake extraction was sig-
nificantly more efficient for BHC isomers (186). The shake extraction
method with hexane-acetone after moistening the soil with 0.2 M
was studied collaboratively using standard AOAC analytical methods
(Jlorisil cleanup and EC-GC) (187) and found to give excellent recoveries
for six insecticides in three different soils (188),
Soil residues of chlorfenvinphos, chlormephos, disulfoton, phorate, and
pirimiphos-ethyl were determined by GC with thermionic detection. Ex-
tracted compounds were cleaned-up on a carbon-cellulose column. Re-
coveries ranged from 95-101% (189). Another group of OP pesticides was
determined in soil by GC with the thermionic detector, following ex-
traction with acetone-hexane-benzene (1:1:1 v/v). Florisil was used to
clean up and fractionate the residues, Dichlofenthion, chlorpyrifos,
ethion, fonofos, and leptophos were eluted with benzene-hexane (9:1 v/v)
and parathion, diazinon, chlorfenvinphos, malathion, phosmet, azinphos
-methyl, diazoson, and paraoxon with hexane-acetone (95:5 v/v) (190).
A multiresidue GC procedure for the herbicides dichlobenil, dinitramine,
triallate, and trifluralin in soils was described by Smith (191). Ex-
traction was carried out with acetonitrile-water (9:1 v/v) in a Sonic
Dismembrator, herbicides were partitioned into hexane, and aliquots
injected directly into an EC chromatograph. Recoveries were 92-107%
from three soils at 0.05-0.5 ppm levels. Acetonitrile-water mixtures
have proven to be especially efficient solvents for residues of herbicides
of different chemical classes (192). Anilide herbicides- were determined
by GC after extraction from soil by blending with acetone (193). Urea
and carbamate herbicides were recovered from soils by shaking with
methanol (194) or acetone (195) and by alkaline hydrolysis and steam
distillation (196). lodinated (196) and 2,4-dinitrophenyl (195) deriva-
tives were used for EC-GC determination of the herbicides. Triazines
were extracted with diethyl ether from soil treated with ammonia (197)
and uracils with 1.5 N HaOH (198). Nineteen acidic, neutral, and basic
herbicides have been determined in soils by two dimensional TLC (199).
Carbofuran residues in soil were determined at the 0.1 mg/kg level with-
out cleanup by EC-GC after ammonium acetate extraction and formation of
the dinitrophenyl-ether derivative (200). Uracils have been recovered
by elution with water from a column prepared by mixing soil with Celite
and Ca(OH)2J the eluate was acidified and extracted with CHCl^, and uracil
determination was by RbCl thermionic-GC (201).
The electrolytic conductivity detector has been used to determine nitrogen-
containing residue* in crude soil extracts. A detector maintenance
program for decontamination of the transfer lines and vent valve pro-
vided reliable operation with little "down time"" even though lengthy
extract cleanup was not carried out (202).
The drying and storage of soils can have an effect on residue analysis.
For example, the extractable atrazine content of soil samples was reduced
-347-
-------
Sections 9A,E, 9AF
by drying at 45°C for 24 hours. Dried samples originally containing
1 ppm of atrazine showed no further significant loss when stored up to
180 days at room temperature, but there was significant loss between
180 and 360 days. Dried samples originally containing 10 ppm of atrazine
showed significant loss after 90 days of storage (203).
The analysis of pesticides of many classes in soils and plants has been
reviewed (204). Results of the U.S. EPA National Soils Monitoring
program employing Florisil cleanup of extracts prior to EC- or FPD-GC
for OC1 and OP pesticides, and partition cleanup of extracts prior to
GC determination of atrazine with an N-selective thermionic detector
have been published (205).
POLYCHLORINATED BIPHENYLS (PCBs), OTHER COMPOUNDS
9A,E PESTICIDE-PCB MIXTURES
PCBs are among the most ubiquitous and persistent chlorinated pollutants
found today in the environment. The residue analyst is concerned not
only with the detection and quantitative estimation of PCBs but with
their effect on the reliable determination of pesticide.residues. PCB
interference may occur with most common chlorinated pesticides in residue
analysis, and the- residue chemist must be aware of the nature of this
interference with respect to the GC columns being used and their opera-
ting parameters. Interference in routine analysis is possible with
ja,£f-DDT, jo,jp_'-DDT, £,2>DDD, and £,.p_f-DDE, as well as with early eluting
pesticides such as BEC isomers, aldrin, heptachlor, and heptachlor
epoxide, since prominent PCB peaks have retention times similar to these
pesticides on the recommended GC columns.
PCBs are frequently detected in human adipose tissues, often at concentra-
tions similar to those of chlorinated pesticides, and interference with
pesticide analysis can be significant, depending upon the columns and
operating parameters used. These interferences demonstrate the non-
specificity of the electron capture GC detector and the need for careful
confirmation by use of at least two GC columns, TLC, chemical reactions,
etc. (Section 10).
9A,F APPEARANCE OF PCB CHROMATOGRAMS
Whenever an analyst observes a conglomerate of chromatographic peaks upon
injection of a biological substrate into an EC detection system, the
possibility of the presence of PCBs should be considered. For example,
Figure 9-C shows a chromatogram resulting from the injection of 10 ng
Aroclor 1254 on a 4% SE-30/6% QF-1 column operated at 200°C with a
carrier flow of 70 ml/minute. The first isomer peak of consequence has
an absolute retention of about 6 minutes and the final peak about 38
minutes. Figure 9-D represents the chromatogram of 6 ng Aroclor 1260
under the same conditions, and major peaks ranging from 8 minutes to
nearly one hour are seen. Aroclors 1254 and 1260 have shown up most
widely in a variety of environmental and tissue samples.
-348-
-------
Section 9A,F
Figure 9-C. Aroclor 1254. Column 4% SE-30/6% QF-1,
200°C, carrier flow 70 ml/min.
1
8 12 16 20 24 28 32' 36 40
Figure 9-D. Aroclor 1260. Column 4% SE-20/6% QF-1,
200°C, carrier flow 70 ml/min.
16 '24 32 40
Retention, minutes
48
56
-349-
-------
Section 9A,F
The type of confusion evident when pesticides and PCBs are present in
the same substrate is illustrated in Figure 9-E, showing Aroclor 1254
co-chromatographed with a mixture of eight chlorinated insecticides.
Aldrin (peak 1), £,,p_'-DDE (3), £,£»-DDD (5), £,_p_'-DDT (6), Dilan I (7),
and methoxychlor (7) are seen to overlap PCB peaks so closely that
differentiation would be impossible. Heptachlor epoxide (2) and dieldrin
(4) (in large quantities) are partially separated, while Dilan II is
fairly well separated. A co-chromatogram of Aroclor 1260 with the same
pesticide mixture would show good separation of aldrin and Dilan II,
partial separation of heptachlor epoxide and Dilan I or methoxychlor,
appearance of disproportionately large Aroclor peaks at the retention
locations of chlorinated pesticides should alert the analyst to the
possible presence of these OC1 pesticides in the PCB sample.
Figure 9-E.
Aroclor 1254 (solid line) and pesticide mixture
(dotted line). Column 4% SE-30/6% QF-1, 200°C,
carrier flow 70 ml/min.
1 AldVjn
2 H«p».
3 p.p'-DDE
4 Di«Wrin
5 p.p'- DOO
6 p. p" DDT
7 Dilonl 4.M»lhorychlor
8 12 16 20
Retention, minutes
24
28
32
Confusing chromatograms also result when PCBs are mixed with the multi-
peak pesticides chlordane or toxaphene. Figure 9-F shows the co-chromato-
gram of chlordane and Aroclor 1254. The only clean separation is the
first peak of the earliest major pair of chlordane peaks, while partial
separation is obtained for the second peak of the third pair. -The early
minor chlordane peaks are well separated but are of little value for
quantitation of chlordane. Aroclor 1260 does not interfere as seriously
with chlordane under these same chromatographic parameters since the
first PCB peak does not elute until after first two major chlordane
peaks.
-350-
-------
Section 9A,F
Figure 9-F. Aroclor 1254 (solid line) and chlordane (dotted Line)
Column 4% SE-30/6% QF-1, 200°C, carrier flow 70 ml/min.
4Retenlion , minutes
Figure 9-G shows a mixture of Aroclor 1254 with toxaphene. Analyses of
toxaphene, chlordane, and PCBs are further confused because the chromato-
grams of environmental samples never exactly resemble those of standards,
Chlordane is not very widespread in environmental samples, so its mutual
analysis with PCBs is less likely to be a problem.
Figure 9-G.
Aroclor 1254 (solid line) and toxaphene (dotted'line)
Column 4% SE-30/6% QF-1, 200°C, carrier flow 70 ml/min.
12 *>
Retention, minutes
28
34) 40
-351-
-------
Section 9A,G
The actual effect of PCBs on quantisation of chlorinated pesticides is
highly dependent on the levels involved, the pesticide of interest, and
the attenuation Tfeing used. For example, if the ratio of PCBs to pesti-
cides is 10 ppm to 3 ppm, an attenuation can be used that will give an
adequate peak for DDE while DDT (for example) and PCBs will hardly be
seen. If quantitation of DDT is required, however, a lower attenuation
will be required (because of its lower response) to give an adequate
peak size, the DDE peak will be off-scale, and PCB peaks will be more
noticeable. At a ratio of 25 ppm PCB to 3 ppm pesticide, quantitation
of DDT will definitely be affected, and with 100 ppm PCB to 3 ppm pesti-
cide and attenuation to keep DDE on scale, determination of the latter
would be affected.
9A,G METHODS FOR SEPARATION AND ANALYSIS OF PESTICIDES AND PCBs
a. Published Procedures and Data
The EPA PAM contains macro and micro methods for determining
PCBs in human milk in Sections 9,B,(1) and 9,B,(2), respectively. In
the macro method, the milk sample (4-24 grams) is extracted with acetone
and hexane, PCBs are transferred to the hexane layer by adding sodium
sulfate solution, and the hexane is dried by passage through a sodium
sulfate column. Part of the sample is used for a lipid determination,
and the rest is partitioned with acetonitrile and then fractionated on
an activated Florisil column 10 cm in height. Identification and
quantitation of PCBs is carried out by EC-GC and confirmation by use
of different GC columns, and the electrolytic conductivity detector
(Cl-mode), chemical derivatization by perchlorination, and GC-MS of
pooled samples.
In the micro method, a 0.5 gram sample of milk is extracted with ace- '
tonitrile, residues are partitioned into hexane, the hexane is concentra-
ted, and the PCBs are eluted through a 1 gram deactivated (3% water)
Florisil column. The eluted PCBs are further separated from chlorinated
pesticides on a micro silicic acid column. Chemical derivatization by
perchlorination to yield decachlorobiphenyl (DCB) followed by EC-GC
is used to confirm PCBs. Neither the macro nor micro methods are
capable of accurately identifying or quantitating absolute levels of
PCBs, but they provide semi-quantitative results.
Filter paper, glass wool, and sodium sulfate are likely sources of PCB
contamination in the macro method, and these materials must be thoroughly
precleaned with pesticide grade solvents as described in Section 3K.
Each sample analyzed requires a total volume of ca 2000 ml of solvent,
and care must be taken in concentration of this large volume to the
final 1-5 ml for analysis. One blank and one fortified goat's milk sample
should be run with every set of 10 human milk samples for both the macro
and micro methods. Details for preparing these samples are described
in Section 9,B,(1), XIV and 9,B,(2), X of the EPA PAM. The amount of
Florisil needed for a proper elution pattern should be determined for
-352-
-------
Section 9A,G
each different lot by elution of analytical standards. Proper separation
of PCBs and .pesticides on the silicic acid column should be checked by
chromatographing standard compounds and analyzing-both eluate fractions.
The Aroclor standard providing a chromatogram most closely resembling
that of the sample should be used for quantitation of that sample.
Analysts inexperienced with the method should be guided through the pro-
cedures at least four times by a person experienced with the procedure,
using duplicate samples already analyzed by the experienced worker.
Then the analyst should be required to demonstrate proficiency on an
additional set of four spiked samples without aid before handling actual
samples. ,
The EPA Manual also describes the separation .of PCBs from DDT and its
analogs by the method of Armour and Burke (206) (Section 9,C), and a
thin layer method for semiquantitative estimation of PCBs in adipose
tissue (Section 9,D). Section 9,E illustrates chromatograms of different
Aroclors on 4% SE-30/6%- OV-210 or QF-1 and 1.5% OV-17/1.95% QF-1 GC
columns, and Section 9,F tabulates relative retention values and re-
sponse values of six Aroclors on OV-17/QF-1, SE-30/QF-1, and OV-210
columns. Retention indices have been calculated for-all 210 possible
individual PCBs on 13 GC phases, and recommendations were made for the
best phase combinations for separations (OV-210, Apiezon L, and OV-225
were among the best single columns; OV-3 + CHDMS and OV-3 or OV-25 + poly
MPE were ,the most discriminating pair) (207). HPLC and capillary column
GC have also been used to separate PCB mixtures (208, 209).
Crist and Moseman (210) of the EPA reported a simplified micro perchlorina-
tion method for determination of PCBs in biological samples. A sample
was cleaned-up by the modified MOG procedure (Section 7Aa), and the
PCBs were perchlorinated with SbCl5 to decachlorobiphenyl (DCB), which
was cleaned-up by hexane partitioning and chromatography on a 1.6 gram
column of activated 4j.orisil. Details are in Sectio^ 9,B, (2),IX of the
EPA PAM. The presence of impurities in SbCls reagent that can cause
erratic recoveries of PCBs was noted by Trotter and Young (211), and
DCB impurity was detected in various brands of the reagents used in the
Crist and Moseman procedure (210).
b. PCB Cleanup and Separation Systems
Depending upon the particular pesticides and PCBs present,
the amounts of each, and the purpose of the analysis, it may or may not
be necessary to separate PCBs and pesticides present in the same extract
before the determinative step. Some combinations may permit quantita-
tion of each without prior separation, others will require a separation
before determination, and still others may require a separation pro-
cedure that destroys or converts some, of the compounds to permit quantita-
tion of those remaining unchanged.
PCBs are eluted with 6% ethyl ether-petroleum ether in the modified MOG
procedure described in the EPA PAM, Section 5,A,(1),(a) and in the FDA
-353-
-------
Section 9A,G
PAM multiresidue procedures, Sections 211 and 212. They elute with
eluant 1 of the alternative methylene chloride elution system (Section
9A,B of this Manual and Section 252 of the FDA PAM). A study by Lieb
and Bills (212) found that the storage temperature of Florisil aftdr
initial activation (overnight, 130°C) influenced the GC pattern obtained
for Aroclor 1254 separated from lipids on a column of the Florisil. To
avoid selective adsorption of some PCB components and erroneous PCB
analyses, storage of activated Florisil at room temperature was
recommended. This, however, is in opposition to the procedure
recommended for routine pesticide work (continuous storage at 130°C
until use) and should be studied further. Hydroxy PCB metabolites
extracted from cow's milk were cleaned up by extraction with aqueous
alkali and re-extraction of the acidified aqueous solution with organic
solvent prior to further TLC cleanup and GC-MS determination (213).
The method of Armour and Burke (206) has been mo$t used for pesticide-
PCB separation. The 6% ethyl ether-petroleum ether Florisil column
eluate or eluate 1 of the alternative procedure (Section 9A,b of this
Manual) is concentrated to an appropriate volume and a 5 ml or smaller
aliquot applied to a column of partially deactivated silicic acid and
Celite, standardized before use to effect the best possible separation
between _p_,_p_f-DDE and Aroclor 1254. Petroleum ether followed by ace-
tonitrile-hexane-methylene chloride (1:19:80 v/v) are used to elute the
column, both fractions being collected in a K-D evaporation flask. The
eluates -are concentrated and subjected to EC-GC. Mixed results have been
reported with this silicic acid separation system. PCBs and polychlorinated
terphenyls split between the two fractions (EPA PAM, Section 9,C, Table 1)
as do the pesticides aldrin .and 2.,2.'-DDE (Canadian PAM, Section 7.5).
Polychlorinated naphthalenes (214) and dioxlns elute in the first fraction
and most other chlorinated pesticides (e.g., chlordane, toxaphene, DDT,
heptachlor, lindane) in the second. Because of the division of some
compounds between the two eluates, GC columns must be carefully chosen
to separate the components present in each fraction. The tables of
relative peak heights and peak retentions in the EPA PAM can help in this
selection. The chemist running this procedure for the first time should
perform a sufficient number of recovery trials with spiked samples to
gain confidence in its reliability. Impurities present in silicic acid
adsorbent batches, their effect on separations, and means for their
removal have been described (215). Pesticide-PCB separations were
found reproducible only for individual batches of adsorbent. Porter
and Burke have reported the separations of TCDD from PCBs on acidic,
basic, or neutral aluminum oxide; PCBs were eluted with hexane-methylene
chloride (99:1 v/v) and di, tri-, and tetrachlorodibenzo-p-dioxins with
hexane-methylene chloride (80:20 v/v) (216).
A slightly modified version of the Armour-Burke method is detailed in
the Canadian PAM Section 7.5, and the method has been miniaturized for
determining chlorinated pesticide and PCB residues in fish. In the latter
method, the sample is dried with Na2SO^ and packed into a column, which
is eluted with petroleum ether. Cleanup and separation of the extract
is on 1 cm (id) Florisil and silica gel columns (217).
-354-
-------
Section 9A,G
Other colunsns used in various multiresidue cleanup procedures provide
at least partial separations of organochlorine pesticides from PCBs.
These include columns of activated alumina (67; Section 90); silica
plus alumina (64; Section 90); 60 A° silica gel eluted with pentane
(elutes PCBs and mirex) and benzene (recovers DDE, DDT, TDE, BHC,
dieldrin) (218); and charcoal (219, 220). Elution of a charcoal column
with diethyl ether-acetone (3:1 v/v) removes OC1 pesticides while PCBs
are then eluted with benzene (59, 221). A Norit C-170 charcoal-poly-
urethane foam (40:60 w/w) mixture is especially useful for separation
of mirex and photomirex from PCBs (222).
Section 251.2 of the FDA PAM describes derivatization.and micro-column
chromatography procedures -for removal of DDT and related compounds
from extracts containing PCBs. Cleaned-up extracts are treated with
alkali to convert DDT to DDE and TDE to its olefin; PCBs are unchanged.
Subsequent oxidation of the solution with chromium trioxide in acetic
acid converts DDE and TDE olefin to dichlorobenzophenone, but again
PCBs remain intact. PCBs are then separated from polar dichloro-
benzophenone by elution with petroleum ether from a micro activated
Florisil column. Dichlorobenzophenone is eluted, if required, with ethyl
ether-petroleum ether (1:1 v/v). Recoveries of Aroclors 1242, 1254, and
1260 ranged from 77-100% (0.4-56 ug amounts), while DDT, DDE, and TDE
were recovered (as dichlorobenzophenone) between 5-86% (2-33 ug). The
same reactions used in this GC determinative procedure have been applied
to TLC estimation (Subsection d.) and confirmation (Subsection e.) -of
PCBs.
Other techniques for separating PCBs from DDT and its analogs by chemical
derivatization and column chromatography include: dehydrochlorination
with l,5-diazobicyclo(5.4.0)undec-5-ene reagent (223); sodium dichromate
in acetic acid plus suifuric acid for conversion of DDE to dichloro-
benzophenone without affecting DDT, TDE, or PCBs (224) ; oxidation by
chromic acid glacial acetic acid reagent followed by silica gel column
chromatography (225); a silica gel tube with MgO catalyst for conversion
of DDT to DDE and TDE to DMU without effect on PCBs (226) ; heating
cleaned-up fish or serum extracts with KOH/ethanol to convert OC1 pesti-
cides to alkenes, oxidation with Cr203 to more polar compounds, and
separation from PCBs on a Florisil column (227) ; and reaction of fish
tissue extract with fuming nitric acid followed by separation of nitrated
PCBs from mirex analogs by chromatography on a micro Florisil column (228) .
Aroclor 1254 residues in blood have been determined by extraction with
hexane-saturated acetonitrile and cleanup on an alumina column. Eluates
were analyzed by EC-GC on an OV-1 column (229). PCT, PCB, and DDT
residues in blood (5-11 ppb) were determined by heating with ethanol
and KOH to dehydro chlorinate DDT, extracting with hexane, washing with
H2S04, and chromatographing on a mixed silica + Florisil + Na2S04 column.
The hexane eluate was concentrated and analyzed by EC-GC on a 2% OV-1
column, and confirmation was by MS (230).
-355-
-------
Section 9A,G
c. GC Quantisation of PCBs
One of the most difficult aspects of PCB quantitation is to
obtain a match between the sample and a standard.- Because individual
congeners in the original source are likely to be distributed differ-
ently due to varying volatilities, solubilities, and reactivities,
biological and environmental samples seldom have a GC peak pattern
that will exactly match any Aroclor standard, and even commercial
preparations of PCBs vary in abundance of minor components from batch
to batch. In addition, detector response of different PCB isomers can
vary by as much as 10,000 fold, so that any similarity between samples
and a known commercial PCB mixture is likely to be purely fortuitous.
For example, the upper chromatogram in Figure 9-H is that of a standard
PCB mixture, Aroclor 1242. Below it is the same mixture added to an
ambient air sample at a level equivalent to 100 ng/m^. The PCB mixture
Figure 9-H.
Gas chromatograms of Aroclor 1242. (A) standard
fortification solution diluted 10 times to .200 pg/yl;
(B) residue in upper foam trap after 24,hours at
225 L/minute. Numbers indicate peaks used for quantita-
tion.
-356-
-------
Section 9A,G
was added to the air stream as vapor while the sample was being
collected, resulting in alteration of the relative peak ratios because
the various components were trapped with efficiencies ranging from
40-95% and also because of contributions from background materials.
It would be difficult to identify the added PCBs in this sample from
GC data alone. Figure 9-1 shows an even more difficult, but no less
typical, case. Chromatogram A is Aroclor 1016, while B is what was
isolated from the brain of a 'rat that had been fed this PCB mixture
for one year (231). The problem of accuracy in PCB analysis is 'not
easily solvable. Because of the complexity'of commercial mixtures,
identification of individual components is not practical. Complete
separation by GC is impossible, even with capillary columns. The mass
spectrometer cannot usually distinguish between all PCB isomers.
The most widely used practical approach for PCB quantitation is to
compare the total area or height of detector response for the residue
peaks to the total area or height of response obtained under the same
conditions for a known weight of the commercial Aroclor standard with
the most similar pattern. Only those peaks from the sample that can
be attributed to chlorobiphenyls are used, and these peaks must also
be present in the chromatogram of reference materials. If the presence
of more than one Aroclor is clearly indicated, the residue may be
quantitated using mixtures of Aroclor standards judged appropriate
for different portions of the sample chromatogram. In one interlaboratory
study of PCB analysis using Aroclor -1254 as reference standard, quantita-
tion via the three specific peaks with retention times, relative to DDE,
of 127, 147, and 177 produced the best interlaboratory agreement (232).
Figure 9-1.
Electron capture gas chromato-
grams of (a) Aroclor 1016
standard and (B) PCB residue
extracted from brain of rat
fed on diet containing
Aroclor 1016 for one year.
2 4 « S
TIME, minutes
-357-
-------
Section 9A,G
The procedure of Webb and McCall (233) has an advantage in that residues
can be quantitatively measured on a GC peak-to-peak basis against a
series of reference Aroclors with known weight percentage compositions
for individual peaks. Reference Aroclors 1016, 1242, 1248, 1254, and
1260 have been characterized using a Hall electrolytic conductivity
detector for Cl measurement and chemical ionization MS with single
ion monitoring for molecular weights (234). A ten laboratory collabora-
tive study of PCB quantitation in synthetic standard mixtures, milk, and
chicken fat samples indicated that greater overall precision was possible
using the individual peak method compared to total area or height (235).
The Individual peak method has been made official first action by AOAC
as an alternative to the total area quantitation procedure (56).
All PCB components may be converted by perchlorination with SbCl5 to
a single derivative (decachlorobiphenyl), and the total PCB content may.
then be'measured as this compound (210, 236; Subsection 9A,Ea). Quantita-
tive data are not truly valid with this approach unless the average
chlorine content of the original PCBs in the sample before chlorination
is known. A related approach to quantitation is dehydrodechlorination
of PCBs to biphenyl by lithium aluminum hydride in dodecane, followed
by HPLC with a UV detector at 248 nm. The absolute detection limit was
100 ng, and DDT isomers, chloronaphthalenes, and PCTs were determined
simultaneously (237).
Other quantitation approaches that have been attempted include estimation
of the weight of PCB injected by dividing the retention time x peak
height for all PCB peaks by the product of peak height and retention time
for 1 ng jp^jfr'-DDE on the same GC column (238), and peak resolution and
matching by a computer (239). GC-MS with individual mass monitoring using
a minicomputer-controlled spectrometer has been reported (240). This
method provided sensitive qualitative and quantitative analysis of sediment
extract without the need for elaborate column adsorption separations prior
to GC. Beroza and Bowman's ;p_-values have been applied to quantitation of
£,p_f-DDT in the presence of non-resolved PCB peaks, and results within
11% of actual were reported (241). The USFDA approach to chemical pro-
filing of PCB content in a sample to select the most suitable quantitation
standard has been discussed (242).
d. PCB Estimation by TLC
The semiquantitative TLC procedure (243) for determination of
PCBs in adipose tissue utilizes the 6% eluate of the Florisil cleanup.
The concentrate is treated with KOH to dehydrochlorinate DDT and DDD
to their olefins, thereby eliminating the problem of separating the pesti-
cides from the PCBs. Any interfering DDE is then oxidized to jj,j3'-dichloro-
benzophenone, which has an % value different from the PCBs on a silver
nitrate-Impregnated alumina layer developed with 5% benzene In hexane.
The PCBs give one spot for all formulations, and this is quantitated
against a standard Aroclor 1254 or 1260 spot run on the same plate. The
best standard can be chosen after examining a preliminary GC trace. The
final values obtained by this method should be considered as approximations,
-358-
-------
Section 9A,H
with a precision of roughly ±50% indicated by recovery studies. The
minimum level of detection is ca 1 ppm by exposure of the layer to
UV light. An EPA human monitoring program for PCB levels in adipose
tissue has utilized this TLC procedure with confirmatidn by combined
GC-MS (244).
e. Confirmation of PCBs
Confirmation of PCBs has been obtained by perchlorina-
tion (210, 236), alkali treatment (245), and reaction with chromic acid
(chromium trioxide). The stability of PCBs to alkali is useful for
confirming the identity of PCB residues, and at the same time, con-
version of DDT to DDE by the alkali removes some interference to
quantitation of PCBs. Treatment with alkali also provides additional
cleanup for many sample types. Resistance to oxidation with chromic
acid-acetic acid reagent is also useful evidence for identifying PCBs
in the presence of reactive pesticides such as DDE and DDT (243) and
chlorinated naphthalenes (246). However, it has been reported (247)
that alteration of PCB chromatographic patterns can occur upon chromium
trioxide acid digestion of animal tissue extracts, including changes
in peak areas and disappearance of several PCB homologs.
Two-dimensional (248) or multi-development reversed phase (249) TLC
systems, which separate PCBs from DDE, DDT, and other pesticides, can
aid identification. PCBs are destroyed by DV irradiation, but many
pesticides may be altered as well (250). Toxaphene survives UV treat-
ment that destroys PCBs and can be confirmed in- mixtures in this way.
Mirex, a late eluting pesticide that usually is not interfered with by
PCBs, also withstands TJV irradiation and can thus be confirmed. Irradia-
tion with controlled UV wavelengths has provided identification and
determination of aldrin, dieldrin, heptachlor, and heptachlor epoxide
in mixture with PCBs. Photoisomerization reactions of the pesticides,
producing products with longer retention times, were induced with wave-
lengths >290 nm; subsequent irradiation with wavelengths >230 nm yielded
photodechlorinated products of PCBs with shorter retention times (251).
Most organochlorine pesticides are destroyed by reaction with HN03-H2S04
whereas PCBs and toxaphene are unaffected. Chlorinated pesticides were
selectively detected in the presence of PCBs by use of a modified
Coulson conductivity detector at 600°C with a hydrogen flow of 1-2 ml/min.
(252). Mirex and PCBs have been differentiated based on the low sensi-
tivity of the Hall detector for the latter (253). A collection of spectra
helpful in confirming isolated residues of PCBs has been published (254).
9A,H DETEBMINATION OF POLYBROMINATED BIPHENTfLS
Polybrominated biphenyls (PBBs) were manufactured for use as a fire
retardant from 1970 to 1973. Since the summer of 1973, when PBBs were
accidentally mixed with dairy feed resulting in the contamination of
livestock and food products, the sensitive determination of low levels
of PBB residues has been of interest to the PDA and the EPA. One
-359-
-------
Section 9A,I
commercial PBB fire retardant (Firemaster BP-6) has been chemically and
toxicologically evaluated; 13 different polybromobiphenyls were found
plus a brominated naphthalene contaminant, and biological effects were
ascribed to PBBs (255).
A rapid method has been developed for analysis of plasma, feces, milk,
and bile using all disposable glassware to reduce the amount of labora-
tory background and cross-contamination of samples (256). The authors
found that this type of equipment was necessary because methods that had
proven to be effective for decontamination of PCBs were not effective
for PBBs. The methodology involved multiple extraction of ethanol-
denatured sample (except for' feces) with petroleum ether-diethyl ether
(1:1 v/v) in a disposable test tube, followed by cleanup on a miniature
adsorption column packed in a 23 -cm disposable Pasteur pipet. The
column contained Florisil, silica gel, and sodium sulfate in different
proportions, depending on the sample. The column was eluted with 5 or
10 ml of petroleum ether-benzene (98:2 v/v) into a disposable screw
top vial. Determination of PBBs in the concentrated eluate was made by
EC-GC on a silanized 5Z OV-17 column. Mean recoveries for the six major
components of a commercial PBB mixture were approximately 96% for plasma,
59Z for feces, 98Z for milk, and 89% for bile at 0.05-50.0 ppm levels.
The maximum background level was 0.0007 ppm for the major hexabromobiphenyl
peak, corresponding to a minimum detectable limit of ca 0.001 ppm.
The separation and characterization of PBBs by chromatography and
spectroscopy have been studied (257). Columns containing 1% SE-30 or
2% OV-17 were used for GC-FID-MS, 5 pm silica gel 60 (Merck) columns
for HPLC (UV detection), and paraffin-coated kieselguhr for reversed
phase TLC. In addition to quadrupole MS, NMR and UV spectroscopy were
evaluated. Capillary GC has been used to separate PBBs, as well as
chlorinated dibenzofurans and anisoles (258).
-------
Section 9A,J
toluene (3:1 v/v). GC-MS analyses to confirm the polychlorinated ter-
phenyl residues were made using a 152 cm x 2 mm id glass column packed
with 32 OV-1 on Gas Chrom Q. The column oven was programmed from
150°C (1 minute) to 300°C at 4°C/minute. Mass spectra were acquired
over the range 420 to 720 amu at 6 seconds/scan for PCT confirmation
(261).
9A,J SEPARATION AND DETERMINATION OF DIOXINS
Section 9,6 of the EPA PAM contains sample preparation and capillary
column GC-MS techniques developed and currently applied by EPA labora-
tories for isolation, detection, quantitation, and confirmation of
2,3,7,8-tetrachlorodibenzo-£--dioxin (TCDD) residues "(262). Tissues,
milk, water, soil, and sediment samples are subjected to an "acid-base"
sample preparation Involving saponificatiori with hot caustic solution,
followed by extraction with hexane, washing with concentrated sulfuric
acid, cleanup by alumina column chromatography, and capillary column
GC-high resolution mass spectrometric multiple ion selection analysis
for TCDD residues. Fish tissue is subjected to a "neutral" cleanup
procedure in which extraction is carried out with acetonitrile, and
cleanup by solvent partitioning and Florisil column chromatography
precedes alumina column chromatography. ^'Ci-TCSD is added to all
samples as an internal standard or marker to monitor and determine the
cleanup efficiency. Sensitivity of the procedure is in the 0.02-100 ppt
concentration range. Extreme care and very clean laboratory practices
are mandatory for low ppt analyses.
The efficiency, accuracy, precision, and validity of ppt TCDD analyses
depend on an incorporated quality assurance program that is described
as part of the procedure in Section 9,G of the EPA PAM. It is important
that TCDD analyses be conducted only by trained personnel with strict
safety procedures in effect. The hazards and analysis of TCDD have
been reviewed (263).
Reports from laboratories that have conducted environmental monitoring
projects for TCDD and have developed and applied analytical cleanup
systems and mass spectrometric methods of analysis for ppt levels of
TCDD residues in environmental, biological, human, and agricultural
samples and chemical formulations are contained in references (264-275).
It has been shown (276) that analysis of environmental samples by low
resolution GC-low resolution MS alone is acceptable if suitable control
samples are available to show the absence of interferences. When
suitable controls are not available and when cleanup is nonspecific,
positive results for TCDD must be confirmed by high resolution MS,
preferably using mass fragmentography with single or multiple ion
detection and/or chemical ionization (277). A recent HPLC method (278)
shows promise of being specific for TCDD and eliminating the need for
high resolutions MS confirmation.
-361-
-------
Sections 9A,K, 9AL
Chlorinated benzyl phenyl ethers have been identified as a possible
serious interference in the GC-MS determination of chlorinated dibenzo-p_-
dloxins. These compounds, which have been extracted from 2,4,5-trichlo-
phenol, have retention times and MS responses similar to TCDDs (279).
9A,K DETERMINATION OF ETHYLENETHIOUREA (ETU)
Because of its toxicological significance, constant occurrence as a
terminal residue following crop treatments with ethylenebisdithiocarba-
mate fungicides, and its actual presence in technical ethylenebis-
dithiocarbamates, analytical methods for determination of ETU have
become extremely important. A method for ETU "in apples (280) was based
on reaction with benzyl chloride to give 2-benzylmercaptoimidazoline,
which is subsequently treated with trifluoroacetic•anhydride to yield
2-benzylmercapto-J$r-trifluoroacetylimidazoline. This derivative is
measured by EC-GO.
ETU residues were measured in various crops by methanol extraction,
alumina column cleanup, and derlvatization with 1-bromobutane in the
presence of DMF and sodium borohydride. The resulting 2-butyl-
mercaptoimidazoline was measured down to 0.01 mg/kg with an FPD detector
(281). A similar method that determines ETU in milk, fruits, and
vegetables as the same derivative has been collaboratively studied (282)
and recommended as an AOAC official first action method (283).
EC-GO as well as S-mode FPD-GC have been used to determine ETU residues
from crops after derivatization with nt-trifluoromethylbenzyl chloride
(284). The trifluoroacetylated S-benzyl derivative has also been used
to determine ETU residues on tomatoes (285) .> ETU residues on fruit and -
vegetable crops were determined at 0.01-0.1 ppm levels without derivatiza-
tion (286). After methanol extraction and cleanup by hexane/aqueous
NH4C1 partition and alumina column chromatography, GC was performed on a
3Z Versamid 900 column with S-mode FPD detection. Recoveries ranged from
62-95%.
The occurrence, chemistry, and metabolism of ETU and analytical methods
for its determination have been reviewed (287).
9A,L DETERMINATION OF CONJUGATED PESTICIDE RESIDUES
Pesticides and pesticide metabolites are known to form carbohydrate
(glycoside, glucuronide), amino acid, sulfate, alkyl, and acyl conjugates
in various plant, animal, and soil systems. Because of the potential
biological activity of many of these conjugates, their identification
and determination has become an important task for the pesticide analyst.
Because conjugates are, in general, more polar and less lipophilic than
the parent pesticides, analytical methods are designed to take into
account these differences. In addition, the lability of certain conjugates
-362-
-------
Section 9A,L
may dictate the analytical approach taken when isolating and identifying
the intact compound or a derivative, e.g., the need for protection of
labile moieties from hydrolysis during extraction or the choice of column
LC rather than GC for separation of thermally unstable or nonvolatile
conjugates. Analysis of enzymatic or chemical hydrolysis products is
useful confirmatory information for conjugates that, have been identified
intact or may serve for the quantitation of a conjugate residue.
Different types of mass spectrometry (Section 10), including electron
impact, chemical ionization, field desorption, and laser ionization,
are probably the most powerful and widely used tools for structural
analysis of conjugates. The field desorption method is especially
useful due to its applicability to polar materials. NMR, particularly
using proton nuclei with the sensitive Fourier transform technique,
is another important aid for structure elucidation (Section 10K), as
are traditional 1R and UV absorption spectrometry and micro-XR (Section
10J).
Specific isolation methods depend on the exact nature of the conjugate
of interest and the sample matrix. Most conjugates are extractable with
water, alcohol, and water-alcohol mixtures from insects, plants, or
tissues. Samples may be freeze-drled and pre-extracted with an organic
solvent to remove lipophilic materials. Purification, separation, and
concentration of conjugates have been carried out using simple solvent
partitioning, counter-current liquid-liquid distribution, extraction
with liquid anlon-exchangers, Amberlite XAD-2 polymer columns, silicic
acid columns, Porapak Q resin columns, Sephadex LH-20 gel columns,
DEAE-cellulose and DEAE-Sephadex anion-exchange columns, Sephadex 6
gel columns, Biogel F columns, cation-exchange resin amino-acid analyzer
columns, liquid anion-exchange paper chromatography, TLC, and GC of
conjugates either directly or after forming a volatile derivative.
Most analytical work on pesticide conjugates to date has been conducted
for structural identifications or metabolism studies. The usual radio-
tracer detection techniques are widely used in metabolism research. A
review of analytical methods for different conjugate types, including
many literature references, and examples of applications to different
research problems will be found in reference (288). This volume also
contains Information on the nature and analysis of "bound" or unextract-
able pesticide residues. One approach to the analysis of bound residues
was reported for chloroaniline bound to lignin fractions of plants based
on release by pyrolysis (289); pyrolysates containing intact chloro-
anilines were collected and derivatized as trifluoroacetanllides, which
were purified and determined by EC-GC.
Relatively little attention has been given to the recovery of pesticide
conjugates by analytical procedures designed to determine the parent
residues. When the problem is addressed, the usual approach is that
which was taken to determine PCP residues in urine (24). The analytical
procedure for intact PCP residues was modified to include an acid
hydrolysis, the purpose of which was to free the conjugated forms of
the pesticide and allow its derivatization along with the unchanged
parent compound (see Section 9E).
-363-
-------
Sections 9A,M, 9A,N
A few similar procedures for other pesticides have been published, the
hydrolysis step in some cases serving both to break the conjugate and
to hydrolyze the parent pesticide to a new form prior to determination
(e.g., hydrolysis of carbamate insecticides to the corresponding phenol,
which is derivatized, cleaned-up, and determined by GC). 3-Hydroxy-
carbofuran, the major carbofuran metabolite produced in animals, is
present as the water soluble glucuronide conjugate. A mild acid
hydrolysis was used to free the conjugated form of the metabolite and
allow its extraction with organic solvent along with the parent com-
pound (290).
Conjugates of 2,4,5-T in biological samples have been broken and the
free acids released by a basic hydrolysis step (291). Residues of
conjugated iodofenphos phenol metabolites were recovered from liver and
kidney tissue by. extraction with ethanol-water-lN sodium hydroxide
(90:10:1 v/v), hydrolysis with IN sulfuric acid, and hexane + ethyl
ether extraction (292).
9A,M REVIEWS OF ANALYTICAL METHODS FOR PESTICIDES, PCBs, AND OTHER
NON-PESTICIDE POLLUTANTS
See Subsection 1G in Chapter 1 for a general bibliography of important
books and reviews on the analysis of pesticides and related pollutants.,
9A,N REFERENCES
(1) McMahon, B. M., and Burke, J. A., J. Assoc. Off. Anal. Cheat., 61,
640 (1978). ~~ ~~~" "~
(2) Sonobe, H., Carver, R. A., and Kamps, L. R., J. Agr. Food Chem..
28(2), 265 (1980).
(3) Suzuki, T., Ishikawa, K., Sato, N., and Sakai, K. I., J. Assoc. Off.
Anal. Chem., 62(3), 685 (1979).
(4) Suzuki, T., Ishikawa, K., Sato, N., and Sakai, K. I., J. Assoc. Off.
Anal. Chem.. 62(3). 681 (1979).
(5) Mills, P. A., Bong, B. A., Kamps, L. R., and Burke, J. A., J. Assoc.
Off. Anal. Chem.. 55, 39 (1972); Mitchell, L, E., J. Assoc. Off.
Anal. Chem.. 59, 209 (1976). *~
(6) Suzuki, T., Ishikawa, K., Sato, N., and Sakai, K. I., J. Assoc. Off.
Anal. Chem.. .62(3), 689 (1979).
(7) Vessel, J. R., J. Assoc. Off. Anal. Chem., 50, 430 (1967).
(8) McCully, K. A., J. Assoc. Off. Anal. Chem.. 55_, 291 (1972).
(9) Pardue, J. R., J. Assoc. Off.Anal. Chem., 54. 359 (1971),
-364-
-------
Section 9A,N
(10) Beckman, H., and Garber, D., J. Assoc. Off. Anal. Chem., 52, 286
(1969). . ~~ ' '
(11) Versino, B,, van der Venne, M. Th., and Vissers, H., J. Assoc.
Off. Anal. Chem.. 54, 147 (1971).
(12) Dale, W. E., and Miles, J. W., J. Assoc. Off. Anal. Chem., 59,
165 (1976). "~~
(13) Johansson, C. E., Pestic. Sci.. £, 313 (1978).
(14) Crist, H. L., Moseman, R. F., and Noneman, J. W., Bull. Environ.
Contam. Toxicol.. 14. 273 (1975).
(15) Solomon; J., and Lockhart, W. L., J. Assoc. Off. Anal. Chem.. 60,
690 (1977).
(16) Crist, H. L., and Moseman, R. F., J. Assoc. Off. Anal. Chem.»
60, 1277 (1977).
(17) Stijve, T., and Brane, E., Dt. Lebensmitt. Rdsch.. 73, 41 (1977).
(18) Franken, J. J., and Luyten, B. J. M., J, Assoc. Off. Anal. Chem.,
59. 1279 (1976).
(19) Griffith, F. D., and Blanke, R. V., J, Assoc. Off. Anal. Chem.,
57, 595 (1974).
(20) Kurtz, D. A., and Kim, K. C., Pestic. Monit. J.. 10, 79 (1976).
(21) Reiner, E., Krauthacker, B., Stipcevic, M., and Stefanic, Z.,
Pest. Monit. J.. 11, 54 (1977),
(22) Gupta, R. C., Karnik, A. B., Nigam, S. K., and Kashyap, S. K.,
Analyst. 103, 723 (1978).
(23) Edgerton, t. R., Moseman, R. F., Linder, R. E., and Wright, L. H.,
J. Chromatogr., 170(2), 331 (1979).
(24) Edgerton, T. R., and Moseman, R. F., J. Agr. Food Chem.« 27(1),
197 (1979).
(25) Edgerton, T. R., and Moseman, R. F., J. Chromatogr. Sci.« 18(1),
25 (1980).
(26) Wu, A., Lech, J. J., Glickman, A., and Pearson, M. L., J. Assoc.
Off. Anal. Chem.. .61(5), 1303 (1978).
(27) Delia Fiorentina, H., De Graeve, J., Grogna, M., and De Viest, F.,
J. Chromatogr.. 157, 421 (1978).
-365-
-------
Section 9A,N
(28) Smith, A. E., and Hayden, B. J., J. Chromatogr.. 171, 482 (1979).
(29) Moseman, R. F., Cristj H. L., Edgerton, T. R., and Ward, M. K.,
Arch. Environ. Contain. Toxicol., £, 221 (1977).
(30) Hodgson, D. W., Kantor, E. J., and Mann, J. B., Arch. Environ.
Contain. Toxjcol.. 7_t 99 (1978).
(31) Earless, R. L., Harris, D. E., Sovocool, G. W., Zehr, R. D., Wilson,
N. K., and Oswald, E. 0., Biomed. Mass. Spectrom.. 5/3), 232 (1978).
(32) Moseman, R. F., Ward, M. K., Crist, H. L., and Zehr, R. D., J. Agr.
Food Chem.. 26, 965 (1978).-
(33) Stafford, C. J., Reichel, W. L., Swineford, D. M., Prouty, R. M.,
and Gay, M. L., J. Assoc. Off. Anal. Chem., 61, 8 (1978).
(34) Blanfce, R. V., Farris, M. W., Griffith, F. P., and Guzelian, P.,
J. Anal. Toxicoli. JL, 57 (1977).
(35) Meijs, A. W. H. M., and Ernst, G. F., J. Chromatogr.. 171. 486 (1979).
(36) Mady, N., Smith, D., Smith, J., and Wezwick, C., NBS (US) Spec. Publ.
519, 341 (1979).
(37) Carver, R. A., Borsetti, A. P., and Kamps, L. R., J. Assoc. Off.
Anal. Chem.. 61,, 877 (1978).
(38) Salek, F. Y., and Lee, G. F., Environ. Sci. Technol.. 12, 297 (1978).
(39) Borsetti, A. P., and Roach, J. A. G., Bull, Environ. Contain. Toacicol.,
20, 241 (1978).
(40) Johnson, L. D., Waltz, R. H., Ussary, J. P., and Kaiser, F. E;.,
J. Assoc. Off. Anal. Chem.. 59, 174 (1976).
(41) Stalling, D. C., Tindle, R. C., and Johnson, J. L., J. Assoc. Off.
Anal. Chem.. 55, 32 (1972).
(42) Johnson, L. D., Waltz, R. H., Ussary, J. P., and Kaiser, F. E.,
J. Assoc. Off. Anal. Chem.. 5£, 174 (1976).
(43) Griffitt, K. R., and Craun, J. C., J. Assoc. Off. Anal. Chem.. 57.
168 (1974).
(44) Ault, J. A., Schofield, C. M., Johnson, L. D., and Waltz, R. H.,
J. Agr. Food Chem.. 27(4), 825 (1979).
(45) Pflugmacher, J., and Ebing, W., J. Chromatogr., 160(1), 213 (1978).
(46) Pflugmacher, J., and Ebing, W., J. Chromatogr., 93, 457 (1974);
J. Chromatogr.. 151, 171 (1978).
-366-
-------
Section 9A,N
(47) Wolff, G., and Ebing, W., J. Chromatogr.. 147, 213 (1978).
(48) Bovey, R. W., and Diaz-Colon, J. P., Tex. Agric. Exp. Stn. Misc.
Publ.. 1381. 33 pp (1978).
(49) Brown, M. J., J. Agr. Food Chem.. 2!3, 334 (1975).
(50) Cromartie, E., Reichel, W. L., Locke, L. N., Basile, A. A.,
Kaiser, T. E., Lament, T. G,, Mulhern, B. M., Prouty, R. M., and
Swineford, D. M., Pestic. Monit. J.. 9/1), 11 (1975).
(51) Blus, L. J., and Lament, T. G., Pestic. Monit. J.. JL3_(2), 56 (1979).
(52) White, D. H., Pestic. Monit. J.. 13(1), 12 (1979).
(53) Ohlendorf, H. M., Elder, J. B., Stendell, R. C., Hensler, G. L.,
and Johnson, R. W,, Pestic. Monit. J.. 13_(3), 115 (1979),
(54) Kaiser, T. E., Reichel, W. L., and Locke, L. N., Pestic. Monit. J..
13(4), 145 (1980). ""•
(55) Bong, R. L,, J. Assoc. Off. Anal. Chem.. 60. 229 (1977).
(56) Burke, J. A., J. Assoc. Off. Anal. Chem.. 61(2), 359 (1978).
(57) Rourke, D. R., Mueller, W. F., and Yang, R. S. H., J. Assoc. Off.
Anal. Chem.. 60. 233 (1977). ~"~
(58) Langlois, E. B., Stemp, A. R., and Liska, J., J. Milk Food Technol..
27(7), 202 (1964). ~~" '
(59) Frank, R., Braun, H. E., Holdrinet, M., Dodge,. D. P., and Nepszy,
S. J., Pestic. Monit. J.. 12_(2), 69 (1978).
(60) .Steinwandter, H., and Schlueter, H., Z. Anal. Chem.. 286. 90 (1977).
(61) Steinwandter, H., and Schlueter, H., Dt. Lebensmitt. Rdsch.. 74,
139 (1978), ~~ ~ • . ".."•.
(62) Bidleman, T. F., Mathews, J. R., Olney, C. E,, and Rice, C. P.,
J; Assoc. Off. Anal. Chem.. 61, 820 (1978). . ,
(63) Brevik, E. M., Bull. Environ. Contam. Toxicol.. 19,'281 (1978).
(64) Holden, A. V., and Marsden, K., J. Chromatogr.. 44, 481 (1969),
(65) Duinker, J. C., and Hillebrandi M.. Th. J.. J. Chromatogr.. 150.
195 (1978). ' ~~".
(66) Wells, D. E., and Johnstone, S. J., J. Chromatogr.. 140. 17 (1977).
-367-
-------
Section,9A,N
(67) Telling, G. M,, Sissons, D, J,, and Brinknan, H. W., J. Chromatogr.,
137. 405 (1977).
(68) Panel on Determination of Organochlorine Pesticides in Foodstuffs
of Animal Origin, Ministry of Agriculture, Fisheries and Food
(Great Britain), Analyst. 104. 425 (1979).
(69) Wegman, R. C. C., and Greve, P. A., Pestic. Monit. J.. 12(3), 149
(1978). ~™~ ~*~
(70) Claeys, R. R., and Inman, R. D., J. Assoc. Off. Anal. Chem.. 57.
399 (1974).
(71) Lea, R., J. Assoc. Off. Anal. Chem., 59. 224 (1976).
(72) de Faubert Maunder, M. J., Egan, H., Godley, E. W., Hammond, E. W.,
Roburn, J., and Thomson, J., Analyst, 89, 168 (1964).
(73) Wood, N. F., Analyst. 94, 399 (1969).
(74) Smart, N. A., Hill, A. R. C., and Roughan, P. A., J. Assoc. Off.
Anal. Chem.. ;57, 153 (1974).
(75) Friestad, H.-O., J. Assoc. Off. Anal.. Chem.,. 60. 268 (1977).
(76) Getz, M. E., Talanta. 22, 935 (1975).
(77) Abbott, D_ C., and Thomson, J., Chem. Ind. (London), p. 481 (1964);
Analyst. 89_, 613 (1964).
(78) Calderbank, A., Residue Reviews. 12, 14 (1966).
(79) Renberg, L., Anal. Chem.. 46_, 459 (1974).
(80) Holden, A. V., Pestic. Monit. J.. T.* 37 (1973); Environ. Qual. Saf. .
Suppl.. ^3, 40 (1975).
(81) Report by the Panel on Fumigant Residues on Grain, Analyst. 99.
570 (1974>.
(82) Lores, E. M. > Bradway, D. E., and Moseman, R. F., Arch. Environ.
Health. 33(5), 270 (1978).
(83) Fournier,. E., Sonnier, M., and Dally, S.., Clin. TorLcol.. 12(4),
457 (1978). " i ~~
(84) Lores,. E. M., and Bradway, D. E., J. Agr. Food Chem.. 25, 75 (1977).
-368-
-------
Section 9A,N
(85) Bradway, D. E., Shafik, T. M., and Lores, E. M., J. Agr. Food
Chem.. 25, 1353 (1977).
(86) Ito, G., Kilgore, W. W., and Seabury, J. J., Bull. Environ. Contarn.
ToKicol.. 22, 530 (1979).
(87) Bradway, D. E., and Shafik, T. M., J. Agr. Food Chem.. 25, 1342
(1977). ~~
(88) Shafik, T. M., and Enos, H. F., J. Agr. Food Chem., 17, 1186 (1969);
Shafik, T. M., Bradway, D. E., and Enos, H. F., J. Agr. Food Chem..
19, 885 (1971).
(89) Houx, N. W. H*, Jongen, W. M. F., de Vries, A. W., Welling, 'W., and
Dekker, A., Pestic. Sci.. 10(3), 185 (1979).
(90) Daughton, C. 6., Cook, A. M., and Alexander, M., Anal. Chem.. 51(12),
1949 (1979).
(91) Takade, D. Y., Reynolds, J. M., and Nelson, J. H., J. Agr. Food Chem.,
27(4), 746 (1979).
(92) Shafik, T. M., Sullivan, H. C., and Enos, H. F., J. Agr. Food Chem..
21, 295 (1973).
(93) Kirby, K. W., Keiser, J. E., Groene, J., and Slach, E. F.,
J. Agr. Food Chem.. 27(4), 757 (1979).
(94) Edgerton, T. R., and Moseman, R. F., J. Agr. Food Chem., J26, 425
(1978).
(95) Hladka, A., Kovac, J., and Krampl, V., Z. Anal. Chem.. 274, 371 (1975),
(96) Kovac, J., Batova, V., Hankova, A., and Szokolay, A., Bull. Environ.
Contam. Toxicol., 13, 692 (1975).
(97) Storherr, R. W., and Watts, R. R., J. Assoc. Off. Anal. Chem.. 48.
1154 (1965).
(98) Watts, R. R., and Storherr,r R. W., J. Assoc. Off. Anal. Chem.. 48.
1158 (1965).
(99) Watts, R. R., and Storherr, R. W., J. Assoc. Off. Anal. Chem.. 50,
581 (1967).
(100) Storherr, R. W., and Watts, R. R., J. Assoc. Off. Anal. Chem.. 51,
662 (1968).
(101) Eichner, M.. Chem. Labor. Betrieb. 2!8, 299 (1977).
-369-
-------
Section 9A,N
(102) Storherr, R. W., Murray, E. J., Klein, I., and Rosenberg, L, A.,
J. Assoc. Off. Anal. Chen.. 50 605 (1967).
(103) Malone, B., and Burke, J. A., J. Assoc.. Off. Anal. Chem.. 52, 790
(1969). ~~~
(104) Gay, L., and Cerney, M., Mitt. Geb. Lebensmittelunters. u. Hyp..
66, 225 (1975).
(105) Luke, B., results of research reported by McCully, K. A. in
J. Assoc. Off. Anal. Chem.. .62(2), 385 (1979).
(106) Eichner, M., Z. Lebensm. Untera. Forsch.. 167(4), 245 (1978).
(107) Heath, A. B., and Black, R. R., J. Asaoc. Off. Anal. Chem.. .62_(4),
757 (1979).
(108) Pflugmacher, J., and Ebing, W. , Landwirtsch. Porsch.» 32(1-2),
82 (1979).
(109) Storherr, R. W., Ott, P., and Watts, R, R., J. Assoc. Off. Anal.
Chem., 54_, 513 (1971). .
(110) Laski, R. R., J. Assoc. Off. Anal. Chem.. 57, 930 (1974).
(Ill) Bowman", M. C., Beroza, M., and Leuck, D. B., J. Agr. Food Chem..
16, 79.6 (1968). .
(112) Sissons, D. J., and Telling, G. M., J. Chromatoer.. 47, 328 (1970).
(113) Renvall, S.,. and Akerblom, M., Residue Reviews, 34, 1 (1971).
(114) Jennings, D. M., Bunyan, P. J., Brown, P. M., Stanley, P. I., and
Jones, ]?. J. S.» Pestie. Sci., J5, 245 (1975).
(115) Domanski, J. J., Haire, P. L., and Sheets, T. J., Arch. • Jfoviron.
Contam. Toactcol>, £, 377 (1976).
(116) Abbott, D. C., Crisp, S., Tarrant, K. R., and Tatton, J. 0*6.,
Pestle. Scir.. 1., 10 (1970.
(117)i Kadoum, A* MT, Bull. Environ. Contam. Toxicol.. 3^, 247 (1968).
(118) Free, I>. J., Butler, E. P», Kimball, E. R., and Stewart, D. K. R.,
J> Econ. Entomol.. 69, 473 (1976).
(119) Ripley, B. F., Wilkinson, R. J. * and Chau, A. S. Y.» J. Assoc. Off.
Anal. Chem... 57. 1033 (1974).
(120) Kadoum, A, M., Bull. Environ. Gontanu Toxicol.. 3^, 65 (1968) and
2, 264 (1967).
-370-
-------
Section 9A,N
(121) Ivey, M. C., and Mann, H. D.><, J. Agr. Food Chem.. 23, 319 (1975).
(122) Ivey, M. C., J. Assoc. Off. Anal, Cheta.. .59, 261 (1976).
(123) Bowman, M. C., and Hill, K. R., J. Agr. Food Chen.. 19, 342
(1971).
(124) Aharorison, N., and Ben-Aziz, A., J. Agr. Food Chem.. 22. 704 (1974).
(125) Bowman, M. C., and Beroza, M., J. Assoc. Off. Anal. Chem.. 52.
1231 (1969).
(126) Thornton, J. S., Olson, T. J., and Wagner, K., J. Agr. Food Chem..
25, 573 (1977).
(127) Hild, J., and Thier, H. P., Deut. Lebensm. Rundsch.. 73. 330 (1977).
(128) Hild, J., and Thier, H. P., Z. Lebensa. Unters. Forseh.. 166, 9 (1978),
(129) Report of the Ministry of Agriculture, Fisheries, and Foods, Analyst,
102, 858 (1977).
(130) Watts, R. R. Storherr, R. W., Pardue, J. R., and Osgood, T.,
J. Assoc. Off. Anal. Chem., 52, 522 (1969).
(131) Sissons, D. J., and Telling, G. M., J. Chromatogr.. 47, 328 (1970);
48, 468 (1970). .
(132) Smart, N. A., Hill, A. R. C., and Roughan, P. A., Analyst» 103 (1228),
770 (1978).
(133) Frehse, H., Anal. Chem.. 59., 896 (1976).
(134) Becker, G., Method 58 in Deutsche Forschungsgemeinschaft (Lieferung
1974) Ruckstands Analytik von Pflanzenschutzmitteln. No. 3., Verlag
Chemie, Weinheim/Bergstr., W. Germany.
(135) Desmarchelier, J., Pestic. Sci.. 8, 473 (1977).
(136) Bradway, D, E., and Shafik, T» M,, J. Chromatogr. Sci., 15, 322 (1977),
(137) Sherma, J., Critical Reviews in Analytical Chemistry, (CRC Press)
1 (3), PP, 299-354 (1973).
(138) Frei, R. W., Lawrence, J. F,, Hope, J., and Cassidy, R. M.,
J. Chromatogr. Sci., 12> 40 (1974).
(139> Bailey, R., Le Bel, G., and Lawrence, J. F., J. Chromatogr.. 161(1).
251 (1978).
-371-
-------
Section 9A,N
(140) Strother, A., J. Gas Chromatogr.. £, 110 (1968).
(141) Storherr, R. W., J. Assoc. Off. Anal. Chem.. 55, 283 (1972).
(142) 'Cook, R. F., Stanovick, R. P., and Caasil, ;C. C., J. Agr. Food
Chem.. 17, 277 (1969).
(143) Lewis, D. L., and Paris, D. F., J. Agr. Food Chem.. 22, 148 (1974).
(144) McKone, C. E., and Hance, R. J., J. Chromatogr.. 36. 234 (1968).
(145) Onley, J. H., and Yip, G., J. Assoc. Off. Anal. Chem.. 54, 1366
(1971). '
(146) LasfcL, R. R., and Watts, R. R., J. Assoc. Off. Anal. Chem.. 56,
328 (1973). ~
(147) Hall, R. C., and Harris, D. E., J. Chromatogr.. 169. 245 (1979).
(148) Erickson, M. D., Fank, C. W., and Morgan, D. P., J. Agr. Food Chem..
27(4), 740 (1979). —^—
(149) Hoermann, W. D., Tournayre, J. C., and Egli, H., Pestle... Mbnit. J..
13_(3), 128 (1979). -^__
(150) Thier, H. P., Angew. Chem. Int. Ed.. 10, 860 (1971).
(151) Williams, I. H., Residue Reviews. J38, 1 (1971).
(152) Cochrane, W. P., and Purkayastha, R., lexicological and Environ-
mental Chemistry Reviews. JL, 137 (1973).
(153) Byast, T. H., Cotterill, E. G., and Hance, R. J., Methods of
Analysis for Herbicide Residues. Technical Report No. 15, 2nd
ed., Agricultural Research Council, Weed Research Organization,
Yarnton, Oxford, England (1977), 65 pp.
(154) Ramsteiner, R., Hermann, W. D., and Eberle, D. 0., J. Assoc. Off.
Anal. Chem.. 57, 192 (1974).
(155) Lawrence, J. F., and McLeod, H. A., J. Assoc. Off. Anal. Chem..
60, 979 (1977).
(156) Lawrence, J. F., J. Chromatogr.. 121, 85 (1976).
(157) Thornton, J. S., and Draeger, G., Int. J. Environ. Anal. Chem..
1, 229 (1973).
(158) Sherma, J., and Shafik, T. M., Arch. Environ. Contain. Toxicol.. 3,
55 (1975). "
-372-
-------
Section 9A,N
(159) Arthur, R. D., Cain, J, D., and Barrentine, B. F., Bull. Environ.
Contain. Toxicol.. 15, 129 (1976); Pestic. Mbnit. J.. 10, 168 (1977).
(160) Kutz, F. W., Yobs, A. R., and Yang, H. S. C., Chapter 4 in Air
Pollution from Pesticides and Agricultural Processes„ Lee, R. E.,
ed., CRC Press, West Palm Beach, FL, 1976.
(161) Thompson, J. F., and Reid, S. J., Arch. Environ. Contam. Toxicol.,
.6, 143 (1977).
(162) Agemian, H., and Chau, A. S. Y., J. Assoc. Off. Anal, Chem., 60,
1070 (1977).
(163) Engst, R., and Knoll, R., Nahrung. 17, 837 (1973).
(164) Aspila, K. I., Carron, J.* M., and Chau, A. S. Y., J. Assoc. Off.
Anal. Chem., 60, 1097 (1977).
(165) Johnson, L, G., Bull. Environ. Gontam. Toxieol.. 5_, 542 (1970).
(166) Leoni, V., J. Chromatogr.. JS2, 63 (1971).
(167) Law, L. M., and Goerlitz, D. F., J. Assoc. Off. Anal» Chem.. 53,
1276 (1970).
(168) Schutzmann, R. L., Woodham, D. W., and Collier, C. W., J. Assoc.
Off. Anal. Chem., 54, 1117 (1971).
(169) Johnson, R. E.,. and Munsell, L. Y., Bull. Environ. jCpntani. ToadLcol.,
17, 573 (1977).
(170) Goerlitz, D. F., and Law, L. M.. Bull. Environ, Contam. TcTtJcol.,
£, 9 (1971).
(171) Mattson, P. E., and Nygren,. S., J> Chromatogr.. 124, 265 (1976).
(172) Jensen, S., Renberg, L., and Reutergardh, L,, Anal. Chem., 49,
316 (1977).
(173) Nagawawa, K.r Uchiyama, H., Ogamo, A., and Shinozuka, T., J.
Chromatogr., 144v 77 (1977).
(174) La Bel, G. C., Williams, D. t., Griffith, G., and Benoit, F. M.,
J. Assoc. Off. Anal. Chem., 62_(2), 241 (1979).
(175) Devine, J» M,, and Zweig, G., J. Assoc. Off. Anal. Chem., 52, 187
(1969).
(176) Faas, L. F., and Moore, J. C., J. Agr. Food Chem.> 27(3), 554 (1979),
-373-
-------
Section 9A,N
(177) Sherma, J., "Thin Layer Chromatography: Recent Advances", Chapter
1 in Analytical Methods for Pesticides and. Plant Growth Regula-
tors. Volume VII, Sherma, J. and Zweig, G., editors, Academic
Press, N.Y., 1973, pp. 1-88.
(178) Pionke, H. B., and Chesters, G., Soil Sci. Soc. Am. Proc., 32.
749 (1968). —'
(179) Burke, J. A., J. Assoc. Off. Anal. Chem.. 59, 339 (1976).
(180) Hall, R. C., and Harris, D. E., J. Chromatogr.. 169. 245 (1979).
(181) Maini, P,", and Collins, A., Bull. Environ. Contain. Toxicol.. 14.
593 (1975). ~ ! —
(182) Gary, A, E., Douglas, P., Tai, H., Mitchell, W. G., and Wiersma,
G. B., Pestic. Monit. J.. 13(1), 17 (1979).
(183) Cotterill, E. G., Analyst. 104(1242), 878 (1979)'.
(184) Goerlitz, D. F., and Law., L. M., J. Assoc. Off. Anal. Chem.. 57.
176 (1974). -~
(185) Suzuki, M., Yamato, Y., and Watanabe, T., Pestic. Mbnit. J.. 11,
54 (1977). —-
(186) Nash, R. G., Harris, P. D., Ensor, P, D., and Woolson, E. A.,
J. Assoc. Off. Anal. Chem.. 56. 728 (1973).
(187) Official Methods of Analysis, (1980), 13th Edition, AOAC,
Washington, B.C., Sections 29.015 and 29.018.
(188) Woolson, E. A., J. Assoc. Off. Anal. Chem.. 57. 604 (1974).
(189) Suett, D. L., Pestic. Sci.. 6., 385 (1975).
(190) Miles, J. R. W., Harris, C. R., and Moy, P., J. Econ. Entomol..
1, 97 (1978). ;
(191) Smith, A. E., J. Chromatogr.. 97. 103 (1974).
(192) Smith, A. E., J. Chromatogr.. 129. 309 (1976).
(193) Bartha, R., J. Agr. Food Chem.. 16. 602 (1968).
(194) McKone, C. E., J. Chromatogr.. JJ4, 60 (1969).
(195) Cohen, I. C., and Wheals, B. B...J. Chromatogr.. 43. 233 (1969).
-374-
-------
Section 9A,N
(196) Baunok, I., and Gelssbuhler, H.. Bull. Environ. Contain. Toxicol..
J5, 7 (1968).
(197) Abbott, D. C., Bunting, Ji A., and Thomson, J., Analyst, 90, 357
(1965).
(198) Jolliffe, V. A., Day, B. E., Jordan, L. S., and Mann, J. B,,
J. Agr. Food Chen.. 15, 174 (1967).
(199) Smith, A. E., and Fitzpatrick, A., J. Chromatogr.. 57, 303 (1971).
(200) Caro, J. H., Glotfelty, D. E., Freeman, H. P., and Taylor, A. W.,
J. Assoc. Off. Anal. Chea., 56_, 1319 (1973).
(201) Caverly, D. J,, and Denney, R. C., Analyst. 10£, 576 (1977).
(202) Sarna, L. P., and Webster, G. R. B., J. Assoc. Off. Anal. Chem.,
57, 1279 (1974).
(203) Swain, D. J., J. Aer. Food Chem.. 27(4), 915 (1979).
(204) Thier, H. P.> Angew. Chem. Int. Ed., 13, 217 (1974)„
(205) Carey, A. E., Gowen, J. A., Tai, H., Mitchell, W. G., and Wiersma,
G. B., Pestic. Monit. J., 12(4), 209 (1978).
(206) Armour, J. A., and Burke, J. A., J. Assoc. Off. Anal. Chem., 53,
761 (1970); FDA PAM (Section 251.1.
(207) Albro, P. W., Baseman, J". K., Clemmer, T. A., and Corbett, B. J.,
J. ChromatQgr^, 136, 147 (1977).
(208) Krupcik, J., Kriz, J., Prusova, D., Suchanek, P., and Cervenka, Z»,
J. Chromatogr., 142. 797 (1977).
(209) Brinkman, U. A. Th., Seetz, F. L., and Reymer, H. G. M., J. Chromatogr.,
116. 353 (1976).
(210) Crist, H. L., and Moseman, R. F., J. Assoc. Off. Anal. Chem., 60_,
1277 (1977).
(211) Trotter, W. J., and Young, S. J. V., J. Asaoc. Off. Anal. Chem..
58, 466 (1975).
(212) Lieb, A. J., and Bills, D. D., Bull. Environ. Contam. Toxicol..
12, 328 (1974).
(213) Gardner, A. M., Righter, H. F., and Roach, J. A. G., J. Assoc. Off.
Anal. Chem.. 59, 273 (1976).
-375-
-------
Section 9A,N
(214) Armour, J. A. , and Burke, J. A. , J, Assoc. Off. Anal. Chem. . 54,
r "J
(215)
(216)
(217)
(218)
(219)
(220)
(221)
(222)
(223)
(224)
(225)
(226)
(227)
(228)
(229)
(230)
(231)
(232)
Hucklns, J. N. , Stalling, D. L. , and Johnson, J. L. , J. Assoc.
Off. Anal. Chem. . £9, 975 (1976). - ~~
Porter, M. L. , and Burke, J. A,, J. Assoc. Off. Anal. Chen., 54.
1426 (1971). — - - - — —
Erney, D. R. , J. Assoc. Off. Anal. Chen.. _57., 576 (1974).
Picer, M. , Picer, N. , and Ahel, M. , Pestic. Monit. J.. 12(3) , 102
(1978) ; Picer, M. , and Ahel, M. , J. Chromatopr. . 150, 119 (1978) .
Berg, 0. W. , Diosady, P. L. , and Rees, G. A. U. , J. Assoc. Off.
Anal. Chem.. 53, 761 (1970). - ""•
Holdrinet, M, , J. Assoc. Off. Anal. Chem.. 5_7(3) , 580 (1974).
Tiechman, J. , Bevenue, A., and Hylin, J. W. , J. Chromatoer. . 151.
155 (1978). ~ - &— -
Babjafc, L. J. , and Chau, A. S. Y. , J. Assoc. Off. Anal. Chem.. 62(5),
1174 (1979). --- — —
Miles, J. R. W. , J. Assoc. Off. Anal. Chem.. 55. 1039 (1972).
Jensen, S., and Sundstrom, G. , Ambio.. 3,, 70 (1974).
Underwood, J. C. , Bull. Environ. Contam. Toacicol. . 2.1(6), 787 (1979),
Luckas, B. , Pscheidl, H. , and Haberland, D. . Nahrung. 20, KL-K2
(1976). —
Trotter, W. J., J. Assoc. Off. Anal. Chem.. 58, 461 (1975).
Insalaco, S. E. , Makarewicz, J. C.", and Vestling, M. W. , J. Acr.
Food Chem.. 28(2), 261 (1980). ~ - 6—
Welborn, M. E. , Borchard, R. E. , Hansen, L. G. , and Metcalf, R. L.
J. Assoc. Off. Anal. Chem.. SJ, 1248 (1974),
Doguchi, M. , and Fukano, S, , Bull. Environ. Contam. Toxicol.. 13,
57 (1975). — - "" - —
Bruse, V. W. , Kimbrough, R. D., Villanueva, E. C. , Jennings, R. ,
Linder, E., and Sovocool, G. W. , Arch. Environ. Health. 29, 301 (1974)
Cochrane, W. P., and Whitney, W. , Advances in Pesticide Science.
Geissbuehler, H, , ed., (3), 664 (1979)"; ~~ - '
.
-376-
-------
, Section 9A,N
(233) Webb, R. W., and McCall, A. C., J. Chromatogr. Sci.. JU, 366 (1973).
(234) Sawyer, L. D., J. Assoc. Off. Anal. Chen.. 61, 272 (1978).
(235) Sawyer, L. D.. J. Assoc. Off. Anal. Chen.. 61, 282 (1978).
(236) Armour, J. A., J. Assoc. Off. Anal. Cheat.. 56_, 302 (1973).
(237) Seidl, G., and Ballschmiter, K., Z. Anal. Chem.. 296_(4), 281 (1979).
(238) Collins, G. B., Holmes, D.C,, and Jackson, F. J,, J. Chromatogr..
71, 443 (1972).
(239) Zobel, M. G. R., J. Assoc. Off. Anal. Chem., 57, 791 (1974).
(240) EicheJberger, J. W., Harris, L. E., and Budde, W. L., Anal. Chem..
46, 227 (1974).
(241) Zelenski, S. G., Tashiro, J., Worthen, L. R., and Oln.ey, C. E.,
J. Chromatogr.. 84, 67 (1973).
(242) Cairns, '£., and Jacobson, R. A., J. Chem. Inf. Comput._Sci., 17,
105 (1977).
(243) Mulhern, B. M., Cromartie, E., Reichel, W. L., and Belisle, A. A.,
J. Assoc. Off. Anal. Chem.. 54, 548 (1970).
(244) Yobs, A. R., Abstracts of Papers WATR 18; Biros, F. J., Enos, H. F,,
Walker, A. C., and Thompson, J. F,., Abstracts of Papers WATR 19;
164th National ACS Meeting, New IJork City, August, 1972.
(245) Young, S. J. V., and Burke, J. A,, Bull. Environ. Contain. Toxicol..
2, 160 (1972).
(246) Holmes, D. C., and Wallen, M., J. Chromatogr.. 71» 562 (1972).
(247) Szelewski, M. J., Hill, D. R,, Spiegel, S. J., and Tifft, E. C.,
Jr.. Anal. Chem., 51, 2405 (1979).
(248) Fehringer, N. V., and Westall, J^ E., J. Chromatogr.. 57, 397 (1971)
(249) de Vos, R., and Peet, E. W., Bull. Environ. Contam. Toxicol., ,6_,
164 (1971).
Lewis, R. G., Hanisch, R. G., MacLeod, K., and Sovocool, G. W.,
J. Agr. Food Chem.. 24, 1030 (1976). .
(250)
(251) Mansour, M., and Parlar, H., J. Agr. Food Chem.. 26, 483 (1978).
-377-
-------
(252)
Section 9A,N
Dolan, J. W., Hall, R. C,, and Todd, T. M., J. Assoc. Off. Anal.
Chem.. 55, 537 (1972). - - "~~
(253) Ballet, D. J., Norstrom, R. J., Onuska, F. I., Comba, M. E., and
Sampson, R., J. Agr. Food Chem.. _24, 1189 (1976).
(254) MacNeil, J. D., Bull. Environ. Contain. Toxicol.. 15, 66 (1976).
(255) Hass, J. R., McConnell, E, E., and Harvan, D. J., J. Aer. Food
Chem.. 26, 94 (1978). "
(256) Willett, L. B.% Brumm, C..J., and Williams, C. L,, J. Agr. Food
Chem., 26, 122 (1978). ' •
(257) DeKok, J. J., J. Chromatogr.. 142_, 367 (1977).
(258) Farrell, T. J., J. Chromatogr. Sci.. 18(1), 10 (1980).
(259) Fehringer, N. V., J. Assoc. Off. Anal. Chem.. 58, 978 (1975).
(260) Chittam, B., Safe, S., Ruzo, L. 0., Hutzinger, 0., and Zitko, V.,
J. Agr. Food Chem.. 25_, 323 (1977).
(261) Wright, L. H., Lewis, R. G., Crist, H, L., Sovocool, G. W., and
Simpson, J. M., J. Anal. ToxLcpl.. .2(3), 76 (1978).
(262) Harless, R. L., Oswald, E. 0., Wilkinson, M. K., Dupuy, A. E., Jr.,
McDaniel, D. D., and Tai, H., Anal. Chem.. 52_(8), 1239-1244 (1980).
(263) Firestone, D., Ecol. Bull.. 27., 39 (1978).
(264) O'Keefe, P. W., Meselson, M. S., and Baughman, R. W., J. Assoc.
Off. Anal. Chem.. 61, 621 (1978). '• .
(265) Shadoff, L. A., and Hummel, R. A., Biomed. Mass. Spectrom. 5, 7
(1978). •*
(266) Shadoff, L. A., Hummel, R. A., Lamparski, L., and Davidson, J. H.,
Bull. Environ. Contam. ToadLcol. 18. 478 (1977).
(267) di Dimenico, A., Anal. Chem.. 61(6), 735 (1979).
(268) Buser, H. R., in Dioxin; lexicological and Chemical Aspects.
Cattabani, E., Cavallaro, A., and Galli, G., eds., Spectrum Publ.
Inc., NY, Chapter 4, pp. 27-41 (1978).
(269) Freudenthal, J., in Dioatin; ToxJcological and Chemical Aspects.
Cattabani, E., Cavallaro, A., and Galli, G., eds., Spectrum Publ.
Inc., NY, Chapter 5, pp. 43-50 (1978).
-378-
-------
Section 9A,N
Harless, R. L. , and Oswald, E. 0., in Dioxin; Toxicologies! and
Chemical Aspects, Cattabani, E., Cavallaro, A., and Galli,j3. ,
i ,~ Spectrum Publ. Inc., NY, Chapter 6, pp. 51-57 (1978)"
(270)
(271) Cruiranett, W.'B., Nature (London), 283/5745), 330 (1980).
(272) Hummel, R, A., J. Agr. Food Ghent. , 25_, 1049 (1977).
(273) Buser, H. R. , Anal. Chem. 49, 918 (1977).
(274) Buser, E. R.' . J. Chromatogr . 129. 303 (1976).
(275) Rappe, C. , and Marklund, S., Cheroosphere 3, 281 (1978).
(276) Hummel, R. A., and Shadoff, L. A., Anal. Chem.. 52(1), 191 (1980).
(277) Hass, J. R. , and Friesen, M. D. , N.Y. Aead. Sci. , Sci. Week,
June 21-30 (1978),
(278) Lamparsfci, L. L. , Nes trick, T. J. , and Stehl, R, H. , Anal. Chem. ,
51(9), 1453 (1979).
(279) Shadoff, L. A., Anal. Chem., ^O(ll), 1586 (1978).
(280) Newsome, W. H, , J. Agr. Food Chem.. 20, 967 (1972).
(281) Haines, L. D. , and Adler, I. L, , J. Assoc. Off. Anal. Chem., 56,
333 (1973).
(282) Yip, G. , J. Assoc. Off. Anal. Chem., 60, 370, 1105 and 1111 (1977).
;;(283) Newsome, W. H. , J. Assocl Off. Anal. Chem., 61(2), 361 (1978).
•o • ' l>' • •
(284) King, R. R. , J. Agr. Food Chem.. 25, 73 (1977).
(285) Newsome, W. H. , J. Agr. Food Chem.. 24, 999 (1976).
(286) Otto, S., Keller, W. , and Drescher, N. , J. Environ. Sci. Health.
Part B, 12, 179 (1977).
(287) Committee on Terminal Pesticide Residues Applied Chemistry
Division, ITIPAC, Pure Appl. Chem.. 49. 675 (1977).
(288) Kaufman, D. D. , Still, G. G. , Paulson, G. P., and Bandal, S. K. ,
Bound and Conjugated Pesticide Residues, ACS Symposium, Series
No. 29, 1976.
(289) Balba, H. M. , Still, G. G. , and Mansager, E, R. , presented at the
175th ACS National Meeting, Anaheim, California, March 13-17, 1978,
Pesticide Chemistry Section Paper No. 6.
-379-
-------
Section 9A,N
(290) Cook, R. F., Carbofuran, in Analytical Methods for Pesticides
and Plant Growth Regulators. Vol. VII. Shgt-™., .T. *~A 7~-*c
G., editors, 1973, p. 187; J. Agr. Food Chem.. 25, 1013 (1977).
(291) Nony, C. R., Bowman, M. C., Holder, C. L., Young, j. F., and
Oiler, W. L., J. Pharm. Sci.. 65. 1810 (1977).
(292) Ivey, M. C., and Oehler, D. D., J. Agr. Food Chem.. 24, 1049 (1976),
-380-
-------
Section 10
CONFIRM/VTORY AND OTHER DETERMINATIVE PROCEDURES
10A REQUIREMENTS FOR POSITIVE CONFIRMATION OF PESTICIDE IDENTITY
Obtaining convincing identification of a trace residue is-a major task
of.the pesticide analyst. The" identity of pesticide residues should
always be confirmed by a method different from that used in the initial
determination since interpretation of results (e.g., decisions of a
legal or health nature) as well as reliable quantitation .(selection
of standards) depend on correct identification. Multiresidue GC
analytical methods do not provide irrefutable identification since
interfering materials and artifacts are often observed, and metabolic
and decomposition products may be encountered.
A specific example of a serious identification problem is the determina-
tion of the PCBs, which are easily mistaken for pesticide residues such
as £,£T-DDE and j3,£r-DDT. Another important example concerns overlapping
peaks when foods are screened for tolerance levels: a 4% SE-30/6% QF-1
column may give peaks at essentially identical retention times for endrin
and .0,2.'-DDT, for Endosulfan I and £,_p_'-DDE, and for B-BHC and lindane.
Both DDT and DDE are very common pesticides with rather high tolerance
levels. Thus, if the analyst is unaware that endrin and endosulfan may
produce corresponding GC responses, he may conclude that observed peaks
indicate only insignificant quantities of DDT and DDE relative to
tolerance levels and that no further work is necessary. Unfortunately,
what appears to be insignificant response for DDT and DDE is very sub-
stantial response for endrin and Endosulfan I because of lower GC
sensitivity to these compounds and lower tolerance levels; therefore,
confirmation of identity is mandatory (1).
Confirmatory evidence is especially important with the relatively non-
specific EC detector. One difficulty is that determinations of very-
low pesticide concentrations are usually required, and many potentially
useful confirmatory methods (e.g., infrared spectroscopy) require a
greater quantity and/or purity of pesticide than might be available.
The techniques chosen for confirming various residues will depend on
the nature of the pesticide, the level found, the type and amount of
sample, and the presence of other residues. The lower the concentration
of pesticide present, the fewer or less certain are the available methods
for making positive identification. If larger amounts of residue are
found and can be isolated in a reasonably pure state, infrared (IR)
spectroscopy and mass spectrometry can provide unequivocal identification.
-381-
-------
Section 10A
Considerations of set theory (2) indicate that three independent "equivocal"
results are required in order to be confident of the identity of a pesticide
residue. These might be elution in a certain fraction from a liquid chroma-
tography cleanup column, a GC retention time, and a positive response of a
selective GC detector. Another possible combination that would be a basis
for confidence is the GC retention times from a polar column and a nonpolar
column plus an Rp value from paper chromatography (PC) or thin layer
chromatography (TLC) or an extraction £-value. Still another would be a
GC retention time, a PC or TLC RF value, and the GC retention time of a
derivative formed by a chemical or photochemical reaction.
The dependence or independence of measured values was studied by Elgar (3)
who reported that many widely used confirmatory methods may not give
truly independent evidence of identity since they are measuring the same
chemical or physical properties. Thus, care must be exercised when
deciding which methods to use in combination to avoid doing a great deal
of work without gaining additional useful information. Examples of highly
correlated (not independent) values include GC retention times on certain
stationary phases (Figures 5-A,A in Section 5); PC or TLC Rp values from
certain adsorbent/solvent systems; ^-values in different solvent pairs;
and PC, TLC, and ^-values. These combinations will not provide independent
information for confirming residue identity.
In Figure 10-A, the correspondence between extraction ^-values in hexane/
acetonitrile and isooctane/DMF solvent pairs (A), and ^-values in hexane/
acetonitrile and TLC Rp values with the system silica gel/hexane (B) is
shown by the generally straight line along which the plotted data points
lie. The independence of TLC and PC data [Figure 10-A, (C)] and GC and
TLC data (D) is illustrated by the scatter of the points. Clearly, many
combinations of alternative columns, selective detectors, £-values or PC
or TLC, and chemical derivatization can be applied for purposes of confirma-
tion.,,
o
Figure 10-A. Degree of correspondence between different types of data for
residue confirmation. A = extraction £-values, B = TLC vs
£-values, C = PC vs TLC, D = TLC vs GC
J.1.0
£0,8
,5
1 0.6
|o.Z
O
O
o
" 2?°
!/° . .
0,2 0.4 0.6 0.3 1.0
Ijoocuno/DMF
B Extraction p-valuc v
•"• thin-layer chromatography
l.u
50.8
,2
{0.6
?0.4
e
g
£0.2
O
00
o
^D *
0 0°
o
0 0.2 0~.
-------
Section 10B
When the analyst is making pesticide identifications, common sense is
necessary. An example of misapplied common sense would be reporting
methyl or ethyl parathion in human fat; metabolically it is virtually
impossible for parathion to persist per se and to appear in a tissue or
body fluid (except gastrointestinal). The.persistence of heptachlor
would also be very unlikely because body metabolism normally converts
it to heptachlor epoxide. Chromatography with EC detection of human
adipose tissue from the general population often produces peaks with
retention characteristics very close or identical to the RRT^ values for
ct-BHC and/or £,j>/-DDE. However, the presence of these compounds has
rarely, if ever, been confirmed. In these instances, the peaks in question
represent artifacts that happen to have the same retention times as these
pesticides, and careful confirmation by ancillary techniques would provide
the proper identification.
In summary, since all methods and tests regularly used in residue analysis
are presumptive in nature (the behavior of an unknown is compared to that
of a known, standard material), it is most desirable to use a number of
tests that measure different chemical or physical properties. The initial
GC method should have been proved to recover and detect the pesticide
residues of interest, and it is desirable that data are available on the
behavior of many pesticides and their metabolites and degradation products
in the various operations that comprise the method. The analyst should
be familiar with and capable of fully using and interpreting these data
and all other available information, including pesticide usage, the
chemistry and metabolism of residues, common artifacts from sample sub-
strates and reagents, and the possibility of interfering residues, such
as PCBs and phthalate esters. Analytical conclusions must be reached with
an open mind, common sense, and reasonable judgment. The extent of
confirmatory effort and the exact procedure chosen will depend on factors
such as the history and significance of the sample; nature and level of
the residues; sairple type; purpose of the analysis; and practical con-
siderations such as time, cost, number of samples, and available instrumen-
tation. Alternatives to confirmation of residues in all samples are
discussed in Section IE. "Unusual" residues should be verified in all
analyses, even at low levels, to support a decision to devote further
effort to tracing their origins. Confirmatory methods should yield
identical results with both the suspected sample residue and standard
reference material subjected concurrently to the same tests. Similar
concentrations of the sample and standard should be used in the comparative
testing to demonstrate quantitative as well as qualitative confirmatory
evidence (FDA PAM, Section 601). The following subsections discuss the
more widely.used confirmatory procedures, some of which are also useful
for residue quantitation.
10B GC RELATIVE RETENTION TIMES
In most laboratories, the initial, tentative identification of a pesticide
residue results from a multiresidue procedure involving extraction, cleanup,
-383-
-------
Section 10B
and gas chromatography. Tables of GC retention times for particular column-
detector combinations are normally used for the tentative identification.
The recovery of a residue through the preliminary cleanup steps should not
be overlooked as valuable, supplemental confirmatory evidence. This is
particularly true when such characteristic properties as the ability to
withstand acid or alkali treatment or elution in a particular fraction
from an adsorbent column is involved.
The following guidelines are useful for the proper utilization of
retention times in making compound identifications.
a. The use of relative retention or Kovats1 retention indicies (4)
rather than absolute retention is more reliable (Subsection 5N in Section 5).
b. Be highly suspicious cjf any peak with, a calculated relative re-
tention value (RRT) that does not precisely match that of the standard
or that of the tables [EPA PAM, Section 4,A, (6)]. A simple aid is to
co-chromatograph some pure standard of the suspect compound along with
the sample extract and observe the peak configuration compared to that
of the sample alone. If some distortion is evident in the configuration
of a given suspect peak, the identification can be safely negated.
c. If cleanup is used on the sample, always run the elution fractions
separately. Do not pool the elution cuts. Selective adsorption combined
with GC retention characteristics provides a valuable identification tool
for pesticide analysis.
d. NEVER rely on one GC column for positive identification. Use an
alternative column providing a completely different peak elution pattern.
As illustrated in Figure 5-A,A in Section 5, the combination of a nonpolar
DC-200 column with a slightly polar DC-200/QF-1 column (plot A) is not
very useful for confirmation. Another highly correlated pair of phases
is slightly polar 4% SE-30/6% QF-1 with slightly polar 1.5% OV-17/1.95%
QF-1. To the contrary, a combination of DC-200 with highly polar DECS
(plot B) or highly polar OV-210 with OV-17/QF-1 (plot C) would be a good
choice. Other complimentary pairs are SE-30/QF-1 with either DECS or
OV-210.
Specific examples (5) of the utility of at least two different GC columns
for sample diagnosis include the following. Identity of certain early
eluting BHC isomers, particularly the alpha isomer, may be hindered by
the presence of hexachlorobenzene. The latter is co-eluted with a-BHC
on silicone columns and with B-BHC on Apiezon, but all three compounds
are resolved on a polar cyano-silicone column. Dieldrin and £,£f-DDE
are difficult to resolve on a number of single phase silicone columns
but are separated on Apiezon, cyano-silicone, and trifluoropropyl silicone
(QF-1, OV-210, SP-2401). On Apiezon, dieldrin elutes before DDE while
the order is reversed on the cyano-silicone column. On the QF-1 or OV-210
-384-
-------
Section 10B
column, dieldrin elutes far later than £,j>/-DDE, to the extent of about
l.'Ax at 180°C column temperature. Figure 10-B illustrates the confirma-
tion of organochlorine pesticides.by comparison of relative retention
times on two columns of different polarities.
Figure 10-B. Dual column confirmation of pesticides by electron capture
detection* Pesticides (from left): lindane, aldrin,
dieldrin, o,£*-DDT, p_,_p_'-DDT, and unknown. Top chromatogram:
4% SE-30/6TOV-210 on Gas Chrom Q. Bottom chromatogram:
1.5% OV-17/1.95% OV-210 on Gas Chrom Q. Both columns:
6.3 mm x 183 cm glass. Carrier gas: nitrogen, 65 ml/min.
Oven temperature: 200°C. Chart speed: 1.^27 cm/min.
-385-
-------
Section IOC
IOC SELECTIVE GC DETECTORS
The EC detector, being rather non-specific, responds to any electron
capturing compounds injected in addition to pesticides. For this
reason, interpretation of results from EC-GC is facilitated if
additional chromatograms are run using one or more of the highly
selective detectors. The thermionic, flame photometric, or conductivity
detectors, described in Subsections 5E, 5F, and 5G, are especially useful
for confirmation. Because interference peaks may occur with even the
most selective detectors available, the absence of a peak is really
more conclusive than a positive response. For example, if a peak on
an electron capture chromatogram suspected of being a chlorinated
pesticide does not appear when the sample is injected into a
chromatograph with the same column and a Hall conductivity detector
in the Cl mode, this is convincing evidence that the original peak
was definitely not due to a chlorinated pesticide but most likely
an artifact with a coincident retention time. Appearance of the peak
in the conductivity chromatogram indicates that the peak was due to a
halogenated compound, but further confirmation is still required to
prove that the peak truly represents the pesticide of interest and not
an artifact.
Because of the- selectivity of its filters, the flame photometric
detector (FPD) may simplify confirmation of sulfur- and/or phosphorus-
containing residues. Identification of a thiophosphate is usually
unequivocal if (a) its retention ratios on at least two different .GC
columns of different polarity match with those of a standard, (b) the
compound elutes in the correct fraction from a cleanup column, (c) it
is detected by the FPD detector, and (d) the sulfur (394 nm) to phosphorus
(526 nm) response ratio of the FPD matches the standard (Subsection 5F).
Figure 10-C illustrates simultaneous chromatograms of parathion,
malathion, and diazinon generated by monitoring both phosphorus and
sulfur emissions with a dual flame photometric detector.
-386-
-------
Figure IOC
Figure 10-G.
Gas chromatograms of phosphorothioate pesticides obtained
simultaneously with a dual flame photometric detector.
I.;.;. ._^—:>- • .—-a)
UH
" i-!:
M..rp:.-u...;.^.:;::io•;,-'— 3 —»::• s
;r:.3 -_-:j::
«— Tla*/Mtn.
".: •••:?*—rj:.~i
•r.trirr-o-nu
• • F r ~*T
I'prl
Ul|- ,U
. I .
-387-
-------
Section 10D
.Relying on only one GC column may lead to incqrrect identification, even
with'the added information from the Florisil column elution and the FPD.
For example, phosalone and azinphosmethyl have the same relative retention
time on an OV-1 GC column; both elute in the third (hexane-acetone,
85:15 v/v) fraction* from the 2% deactivated Florisil column recommended
by the Canadian PAM (Section 9M, this Manual), and both respond equally
to the FPD since each compound has one P and two S atoms per molecule.
A second example is phosalone and phosmet (Imidan), both of which have
the same retention on OV-17, elute in the third Florisil column fraction*,
and have one P and two S atoms per molecule (6).
The selectivity of an EC detector can be improved if the products formed
in the detector are allowed to pass to a second column with another EC
detector; the resultant distinctive peak pattern can provide identifica-
tion of OC1 pesticides and PCBs (7).
When it is possible to use two gas chromatographic detectors, further
confidence in qualitative accuracy can be achieved. For example,
simultaneous analysis by electron capture and flame photometric gas
chromatography is very useful for confirmation of organophosphorus
pesticides (Figure 10-D). The Hall microelectrolytic conductivity and
nitrogen-phosphorus detectors are likewise very useful for dual-detector
confirmation.
10D THIN LAYER CHROMATOGRAPHY (TLC) RF VALUES
Experimental aspects of TLC and its use for screening and quantitation
of residues were covered earlier in Subsections J through M in Section 7.
TLC is perhaps the simplest confirmation technique for GC when levels''of
residues present are high enough. An aliquot of cleaned-up extract is
evaporated to near dryness, a suitable solvent is added, and a detectable
quantity of the sample is spotted on a thin layer plate together with.
appropriate standards. An agreement of about + 2 mm in migration distance
of the sample and standard spots is considered adequate since the movement
of the sample is likely to be affected by co-extractives despite cleanup
steps. If the sample contains several pesticides, different solvents
and/or adsorbents may be required before all are separated and matched
with standards. Mixing together the sample and a standard and observing
whether separation occurs (co-chromatography) is another procedure for
making comparisons.
It is best not to rely on published or previously determined Rj- values
for confirmations because differences in development conditions from run
to run cause these values to be non-reproducible. Standards and samples
A recently devised elution system for 2% deactivated Florisil columns,
modified from that described in Section 9M, includes four eluents:
hexane-methylene chloride•(95:5 v/v), hexane-methylene chloride
(70:30 v/v), hexane-acetone (85:15 v/v), and hexane-acetone (1:1 v/v
(6).
-388-
-------
Section 10D
Figure 10-D.
Simultaneous gas chromatograms of organochlorine and
organophosphorus pesticides using electron .capture and
flame photometric detectors. Columns: 6.3 mm x 183 cm,
4% SE-30/6% OV-210 on Gas Chrom Q. .Carrier gas: nitrogen.
Oven temperature: 200°C. Detectors: Pulsed 63Ni, 270°C;
flame photometric, 526 nm filter, 200°C. Chart speed:
1.27 cm/min.
should always be run on adjacent areas of the same plate if possible.
If Rp values must be used, the value relative to the Rp of a standard
compound X run on the same plate (Rg value) will be more reliable than
the absolute Rp value for many of the same reasons that relative GC
retention times are more reliable than absolute retentions. Chlorinated
pesticides are often referred to DDD, and phosphates to parathion, in
calculating R^ values.
Although TLC is very widely applied for pesticide confirmation, results
may not always be conclusive. TLC confirmation of many pesticides, such
as toxaphene and chlordane, is greatly influenced by the degree of clean-
up on the sample extract and the level of detection. Oils and waxes will
-389-
-------
Section 10D
particularly.interfere with TLC, causing streaked zones and/or distorted
RF values that may completely negate its value for confirmation. The
15% ethyl ether-petroleum ether Florisil column extract normally requires
further cleanup prior to TLC (FDA Pesticide Analytical Manual, Section
411.5).
Detection reagents yielding spots of different colors with different
pesticides are especially valuable for confirmation. Diphenylamine-
zinc chloride reagent provides such differentiation for chlorinated
pesticides; various shades of purple, grey, green, and reddish-orange
colors are produced on the layer after spraying and oven heating (FDA
PAM, Section 612). Identification of naturally fluorescent pesticides
is aided by heating the chromatogram, causing specific alterations in
recorded spectra (8). Th.is heating procedure may, however, increase
background fluorescence from co-extracted compounds also present in
the sample. TLC after fluorogenic labeling (9) of pesticide residues
is a. combination of chromatography with chemical derivatization (Sub-
section 10G) that can provide very specific detection of certain residues.
If sufficient pesticide is present in the thin layer spot, scraping,
collecting the adsorbent, and eluting the compound followed by mass
spectrometry (Subsection 10L) can provide unequivocal identification.
»
It was mentioned earlier in this section that if additional independent
information is to be gained by running PC plus TLC or TLC in more than
one system, the systems must be very carefully chosen to be truly
"different". The use of multiple Rp values for identification purposes
was studied by. Connors (10), who found that useful, uncorrelated data
can be obtained in several ways, such as by pairing aqueous with non-
aqueous systems, acidic with basic solvents or supports, aprotic with
protic solvents, polar with nonpolar solvents, hydrogen-bond donors with
hydrogen-bond acceptors, or reversed phase with normal phase systems.
The specific approach that might be successful depends on the chemical
nature of the pesticides to be confirmed. The important point is that
different thin layer and/or PC systems chosen at random will not
necessarily provide the analyst with any additional, independent evidence
of identity. Similar correlation studies were reported by Dale and
Court (11).
Permanent records of TLC plates for documentation should be made by one
or more of the following methods: Xeroxing the original plate, spraying
with plastic to preserve the plate, hand tracing or charting, densitometry,
or color photography (12). Where available, the latter appears to be the
preferred procedure.
Section 614.11 of the FDA PAM describes a method for confirmation of
organophosphorus pesticide residues by two-dimensional TLC. It is
applicable at levels as low as 0.01 ppm in nonfatty food extracts cleaned
up by carbon column chromatography. The pesticides are oxidized by
bromine vapor after the first development, and detection is made with
horse serum cholinesterase and indoxyl acetate substrate after the second
development. The system provides good specificity because it involves
chromatography of both the parent pesticides and their derivatives.
-390-
-------
Sections IDE, 10F
10E HIGH PERFORMANCE LIQUID CHROMATOGRAPHY (HPLC)
See Subsections 7A through 71 in Section 7 for a discussion of this topic.
HPLC has been used mainly for quantitation of residues in situations where
GC is either not applicable or not convenient to use. An HPLC retention
time can serve as evidence to confirm GC in the same way as a PC or TLC
Rp value. The liquid chromatographic system should be carefully chosen
to be "different" from the GC system (i.e., adsorption rather than
partition), and the independence of the data must be clearly established
if it is desired to use both PC or TLC and HPLC data for confirmation.
The variable wavelength UV detector allows determination of the wave-
length of maximum absorption for each pesticide. Detection of HPLC
effluents with a Cl-selective electrolytic conductivity detector (13)
can also provide useful confirmatory evidence.
f •
10F EXTRACTION £-VALUES
Extraction £-values (14-18) are a tool for identifying, pesticides at the
low ng level. The £-value is determined by equilibration of a solute
between volumes of two immiscible liquid phases, followed by the analysis
of one of the phases for the solute. The £-value, defined as the
fraction of total solute partitioning into the upper phase, can be
derived from a single distribution between the solvents or from a
multiple distribution, as in counter-current distribution. ^-Values
for most pesticides are appreciably different from those of normal
co-extracted contaminants. The determination of these values is
simplified since only relative, rather than absolute, data are required,
and sensitivity is at or only slightly above the level of EC-GC.
Details including experimental procedures, formulas for calculating
^-values and the fractional amount extracted after repeated extractions,
graphs for determining specificity in a given system, and ^-values for
131 pesticides in six binary solvent systems (hexane-90% DMSO, heptane-
90% ethanol, isooctane-80% acetone, hexane-acetonitrile, isooctane-DMF,
and isooctane-85% DMF) are given in Section 621 of the FDA PAM, reference
(15), and Section 12,C of the EPA PAM (data for 88 pesticides in the
latter). A device and method for determining ^-values with unequilibrated
solvents or unequal phase volumes are given in the FDA PAM, Section 622.1
and reference (18).
As mentioned earlier, the general technique of determining £-values has
much in common with the use of several GC columns, PC, and TLC in identi-
fication studies since all systems may share the same partition mechanisms.
Unless the analyst assures himself that the data are not correlate4, it is
best to use either a PC c>r_ TLC Rp value or an extraction p_-value as one
independent criterion of identity. The great advantage of ^-values over
PC or TLC is that the method is useful at levels amenable to quantitative
analysis by EC-GC where sufficient residue might not be available for
either of the former techniques.
-391-
-------
Section 10G
10G DERIVATIZATION (CHEMICAL REACTION) TECHNIQUES
Derivatives of pesticides are prepared for various reasons, such as to
decrease volatility or increase detectability for HPLC or TLC; to
increase volatility, stability, and/or detectability and avoid tailing
peaks for gas chromatography; for removal of interferences in residue
analysis (19); and to alter the structure to aid characterization.
It is this latter topic that will be discussed in this subsection.
Comparison of retention times on a given GC stationary phase before and
after chemical derivatization is a relatively r.ecent innovation that
is becoming increasingly important for corroboration of residue identity.
Desirable characteristics of any chemical derivatization technique include:
a. A specific product should be formed with at least as much or
more response to electron capture, or to other detection, compared to
the parent pesticide.
b. The product should have a different retention time than the
parent, preferably greater to differentiate it clearly from the background.
c. Reactions should be essentially quantitative, they should use
highly pure reagents and solvents, and they should be facile and rapid.
Reagents 'and equipment should be inexpensive, if possible.
d. A cleanup method should be available to remove any background
interferences introduced by the reaction.
e. If product structures and reaction mechanisms and limitations
are known, misidentifications can be avoided because the analyst can
elucidate the extent and probable sources of error in the procedure.
f. Sensitivity should be at least in the 0.01 to 0.1 ppm range in
terms of the parent pesticide, which is lower than the established
tolerance values for most pesticides.
g. The same reaction should occur, and to the same degree, in
both the sample extract and in a solution of the reference material of
the suspected compound at the same concentration. Matrix effects can
play an important role in the applicability of chemical derivatization
for quantitation purposes. A reaction might work very well for pure
standards but may fail when applied to samples due to the effects of
sample components.
h. The reaction should be safe to perform.
Derivatization reactions for gas chromatography are usually carried out
in solution, on the surface of a solid matrix, or in a GC precolumn.
Reactions in solution on a microscale are most common for residue level
-392-
-------
Section 10G
work. The reaction usually involves heating of the reactants in a small
sealed tube, after which the derivative is dissolved in a suitable solvent.
If direct injection into a gas chromatograph is not possible, cleanup
by solvent partitioning and/or column chromatography and concentration
steps may be applied. Solid matrix reactions are generally carried
out by introduction of dissolved compound onto a microcolumn composed
of solid support (e.g., alumina) mixed with reagent(s). After a
specified reaction time, solvent is added to elute the derivative for
GC determination. The advantages of this approach are simplicity,
reduced glassware needs, and ability to react many samples simultaneously.
However, the same derivative as formed in a solution reaction is nbt
always produced and/or eluted from the column in a solid matrix reaction
with the same active reagent. GC precolumns are usually composed of a
reagent-solid support mixture located in a heated area ahead of the
'analytical.column. The sample is injected into the precolumn, and the
derivative is formed and swept by the carrier gas onto the analytical
column for determination. Speed of operation is the greatest advantage
for those reactions that are rapid enough to be feasible by the precolumn
technique. The chromatograph is best fitted with a special injection
apparatus so injection can be made into the precolumn for derivatization
or directly into the analytical column for normal operation. In addition
to these three types, some derivatizations may be carried out in a hot
injection port or on the analytical column itself.
Derivatization aimed at increasing detectability in HPLC is usually carried
out "post column", or after separation of the parent molecules rather than
the derivatives. This is accomplished by inserting a mixing chamber at
the end of the column and pumping in reagent to mix with the column
effluent. The derivative is formed in a reaction coil, and is measured
subsequently in a suitable detector. This type of derivatization can be
easily automated for routine analysis. Derivatization for the purpose
of providing detection in TLC is generally carried out by spraying chemical
reagents, also "post chromatography". For both of these liquid chromatog-
raphy techniques, derivatization for confirmation of identity is usually
done in solution or "on column" (or "on-layer", for TLC), prior to
chromatography, as is the case for GC.
The following subsections review some procedures for confirming residue
identity by chemical derivatization. Table 651-A of the FDA PAM contains
an extensive further listing of derivatization methods for more than
100 pesticides and related compounds of many chemical types, including
comments on the level of applicability, yields, and 60 references to the
original papers. A review paper on chemical derivatization for GC and
HPLC has been published (20).
a. Organochlorine Pesticides
Most of the effort to date in the development of confirmatory
derivatization tests has been confined to the organochlbrine insecticides.
For these compounds, addition, oxidation, epoxidation, rearrangement,
-393-
-------
Section 10G
dechlorination, hydrolysis, reduction, and dehydrochlorination'are the
most commonly used reactions. Examples of specific tests are shown in
Tables 10-1, 10-2, and 10-3, as reviewed through 1978 by Cochfane
(20-22). Table 10-4 lists selected references for OC1 pesticide
derivatization methods published since 1978. Section 9A,G,e discusses
confirmation reactions suitable for PCB-pesticide mixtures. It must
be realized that these reactions destroy some pesticides (and Artifacts)
in addition to forming pesticide derivatives.
TABLE 10-1
CONFIRMATORY DERIVATIZATION TESTS FOR
PESTICIDE AND METABOLITE RESIDUES (21)
Pesticide
DDT
DDE
DDD
Methoxychlor
Reaction
(a) Dehydrochlorination
(b) Dechlorination (of
Oxidation
Dehydrochlorination
Dehydrochlorination
' -isomer)
Aldrin
Dieldrin
Endrin
Endosulfan
Heptachlor
Heptachlor epoxide
cis- and trans-Chlordane
Nonachlor
(a) Addition^-— Br2
^ tert-BuOCl
(b) Epoxidation
cleavage
Epoxide ^^— rearrangement
acetylation
(a) Epoxide rearrangement
(b) Dechlorination
Sulf ite reduction
acetylation
(a)
Ally lie ^- — hydroxylation
dechlorination
(b) Addition
(c) Epoxidation
Epoxide rearrangement
Dehydrochlorination
(a) Dechlorination
(b) Dehydrochlorination
-394-
-------
TABLE 10-1 (Continued)
Section 10G
Parent Pesticide
Heptachlor
trans-Chlordane
cia-Chlordane
cisj- and trans-
Chlordane
Endrin
Endosulfan
Metabolite
(a) Chlordene
(b) 1-Hydroxy-
chlordene
Group Reacted
(1) Allylic hydrogen
(2) Double bond
(1) Allylic hydrogen
(2) Double bond
(c) l-Hydroxy-2,3- (1) Hydroxyl
epoxychlordene
(2) Epoxide
2-Chlorochlordene iDouble bond or
3-Chlorochlordene (. gem-dichloro group
1,2-Dichloro- Chloro epoxide or
chlordene epoxide gem-dichloro group
Photo-endrin gem-Pichloro
Endosulfan diol Hydroxyl
Derivative-
1-Bromochlordene
Chlordene epoxide
Silyl ether
Chloroacetate
Epoxide
Silyl ether
Trihydroxy
chlordane
/Epoxide or
Ihexachloro
Chloroacetate
or heptachloro
Pentachloro
Acetate or silyl
ether
-395-
-------
Section 10G
TABLE 10-2
CONFIRMATORY TESTS FOR ORGANOCHLORINE PESTICIDES C22)
Pesticide Class Reagent or Reaction Type
General
Hexachlorobenzene (HCB)
BHC isomers
Cyclodiene insecticides
Mirex
Kepone
PCBs
Chlorobiphenyls and PGP
DDT
CrCl2 reduction (26)1
KOH denydrochlorination (46)1
Base/alcohol
KOH hydrolysis/diazomethane
NaOMe/MeOH or GC alkaline precolumn
Comparison of 3 methods (D)
10 various reactions (D)
BCl3/2~chloroethanol (D/E)
UV irradiation (D/E/H)
H2S04 or 60% KOH (E/M)
_t-BuOK/£-BuOH or CrCl2 (E/M) '
Acid or "base-Al203 microcolumn (C/E/H/T/M)
Base-catalyzed intramolecular cyclization
Silylation/acetylation (T/M)
UV dechlorination
KOH/esterification
LiAlH4/PCl5
SbCl5 perchlorination
Acetylation and butylation
Reduction and/or oxidation
•^•Figures in parenthesis indicate the number of pesticides studied,
letters indicate the particular pesticide(s) confirmed
C * chlordanes, D = dieldrin, E = endrin, H = heptachlor,
T » Thiodane (endosulfan) and M = and metabolites
-396-
-------
TABLE 10-3
Section 10G
CHEMICAL DERIVATIZATION OF ORGANOCHLORINE, PCB, PBB,
AND RELATED COMPOUNDS3 (20)
Pesticide/
Compound
Derivatisation
Procedure
Substrate
PCB/OC EihanoKc KOH
EthanottoKOH/
CrO3
Photoisomerisa-,
lion (OCs)
Photo-dechlprin-
ation(PCBs)
"MgO micro-
reaction"
PCB/OC/chlorin-
ated paraffins Photolysis
PCBs
Hydro.xylated
PCBs
PCB/Mirex
Mirex
Kepone
Kelvan
PBB
HCB
Endosulfan
Heptachlor and
Epoxide
Lindane
Toxaphene
Polychloro
naphthalenes
TCDD
TiO2 photode-
chlorination
Perchlori nation
Silylation
Photolysis
Photolysis
Hematin
dechlorination
Chlorination
Photolysis or
oxidation
Photolysis
2-propanol/KOH
Fish and fish prod.
Fish, serum
Environmental
samples
Fish
Fish
Aqueous media
Fish
Photolysis
Acetyiation
Photolysis
NiCl2/NaBH4
Acetyiation
Human tissue
and milk
Urine, feces.eggs
Model system
Blood, oyster
Potatoes
Feeds, dairy prod.
Adipox tissue,
human milk
Soils
Foods
Soils
MethanolicKOH —
Photolysis
Photolysis
Silica, soil
PCB = polychlorinated biphenyl; PBB » polybrominated biphenyl;
HCB = hexachlorobenzene; TCDD = tetrachlorodibenzodioxin;
OC = organochlorines.
-397-
-------
Section 10G
TABLE 10-4
CONFIRMATORY DERIVATIZATION REACTIONS FOR OC1
PESTICIDES PUBLISHED SINCE 1978
Compounds Studied
Reagent (sensitivity)
Reference
Chlordane and mirex
Mirex and PCBs
(in fish)
OC1 pesticides and PCBs
(Harp seal tissue)
Chlorophenoxy acid
herbicides
Cone. ^SO^-fuming HNOg .(1:1 v/v)
to remove PCBs and other
interferences
Reduction of mirex with
chromous chloride
Dechlorination using sodium
ethoxide
Pentafluorobenzyl bromide
(10 ppb in urine)
23
24
25
26
A confirmatory technique related to chemical derivatization is ultra-
violet degradation or photolysis (27, 28; Table 652-A of the FDA PAM).
Degradation products arising from UV treatment of chlorinated insecti-
cides and detected by EC-GC can provide identification of these pesti-
cides (28) at 75-100 pg levels. Depending on the length of irradation
(often ca 10 minutes), all of the parent pesticide may not be degraded.
Solvent and sample blanks should be run to prove if background is reacted
as well. Isooctane is a good solvent because it is little affected by
UV light.
Section 12,D,(1) of the EPA PAM gives details of a microscale alkali
dehydrochlorination method for use in multiresidue analysis. This
procedure produces derivatives for identity confirmation and provides
supplemental cleanup for some troublesome extracts after Florisil
chromatography. Section 651.12 of the FDA PAM describes the micro-
scale alkali treatment method that is"part of the AOAC official method
for perthane. Table 651.1 lists the behavior of about 40 compounds
-398-
-------
Section 10G
under these reaction conditions. Alkali reactions carried out on a GC
precolumn rather than in solution have proved advantageous in some
instances (29). Section 5,A,(l),Cb) of the EPA PAM describes the
confirmation of HCB in fatty tissue by formation of the disubstituted
ether derivative bis-isopropoxytetrachlorobenzene (30).
Section 11 of the Canadian PAM gives complete details for the following
tests:
Pesticide(s)
£,£f-DDT, £,£!-DDT, £,£f-TDE,
methoxychlor
£,£'-DDT, endrin
Dieldrin, endrin
Chlordane, heptachlor
epoxide
Aldrin, heptachlor,
£»£f -DDE
Aldrin
Endosulfan
Chlorophenoxy acid
herbicides
Captan
Reagent
Sodium methylate
Chromous chloride
BC'l- in 2-chloroethanol
K-tert butoxide/tert-butanol,
silylation
Chromic acid .
m-Chloroperbenzoic acid
Alcoholic KOH
jtt-Propanol
Resorcinol
A special two stage, mixed phase 180 cm column consisting of 165 cm of
4% OV-1/6% QF-1 and 15 cm of 3% OV-1/6% OV-225 at the injector end is
recommended in the Canadian Manual for resolving HCB, BHC isomers, sulfur,
and aldrin for confirmatory purposes, because they are not resolved on
the 4% SE-30/6% QF-1 working column.
One method of differentiating PCBs from organochlorine pesticides is
by treating the residues with a non-fuming HN03-H2S04 mixture. Organo-
chlorine pesticides are destroyed, whereas"PCBs (and toxaphene) are
unaffected. Various confirmation methods for PCBs are covered in Sub-
section 9A,G,e in Section 9 of this Manual, and perchlorination is
described in Section 12,D,(2) of the EPA PAM.
b. Other Pesticide Classes
Residues of organophosphorus pesticides may be confirmed by
alkaline hydrolysis followed by esterification of the resulting dialkyl
phosphates to trialkyl phosphates (31). This procedure does not distinguish
pesticides that produce the same hydrolysis product. According to McCully (32),
-399-
-------
Section 10G
the three most practical methods for confirmation of OP pesticides are
oxidation to oxygen analogs (33), pentafluorobenzyl bromide derivatiza-
tion of hydrolyzed phenols or thiophenols (34), and chromous chloride
reduction (35). The sodium hypochlorite oxidation method has the widest
applicability, but it suffers from low sensitivity, difficulty in
analyzing the analog products, and inability to distinguish analogs
originally present in samples. The CrCl2 method is simple but applicable
only to OP pesticides containing a nitro group. The pentafluorobenzyl
bromide procedure is intermediate in scope. These and other reactions
used to identify organophosphorus pesticides are listed in Table 10-5,
along with information on triazine, carbamate, and urea pesticides.
This table is from a review article (22) that gives the original
references for these reactions. Table 10-6 contains a selection of
more recent references. The pentafluorobenzyl bromide derivatization
procedure for OP pesticides is being collaboratively evaluated on
different substrates (36).
Triazine herbicides have been confirmed by silylation, methoxylation
(in sodium methoxide-methanol), methylation (CH3l-NaH), and hydrolysis-
DNFB reactions (37, 38), and linuron has also been confirmed by alkylation
(with alkyl halide - NaH) (37).
TABLE 10-5
CONFIRMATORY TESTS FOR ORGANOPHOSPHORUS, TRIAZINE, UREA, AND CARBAMATE COMPOUNDS (22)
Pentlcid. Clasa
Orgftnophospluccs
Triatinea
C«rban»te» and
ureas
Chlorophonoxy
•elds
Compound Type
Confii
! General
Phenol-generating compounds
Aryl-H0_ and Aryl-CN
containing compounds
d) P™ S compounds
e) -Nit *nd -Nllg containing oompounda
f) OH compounds (diatimm metabolites)
g) Crufornate
a) Chloro-a_-trlaEines
b) Hydroxy-s-triauineB
c) Cyanazlne and Metabolites
a) Intact compound
b) Phenol-generating compounds
o) Amlne-gonerating compounds
a) Esters
Hydrolysla/mothylatlon
HyUrolyalB/PFD ether formation
Reduction (CrClg.PdCl.,, Zn/HCl)
Oxidation (to P — 0)
i)Alkyation (Hall/Mel/DMSO)
ii)Ueainlnat Ion/methyl a tion
Silylation or alkylation
UV dechlorination
1) Alkylation
' Silylation
Methoxylation
HydrolyslB/DNP formation
Sllylation
Alkylation
Chlurination '
li
ill
iv
1
11
ill
Acid catalyzed cyollcation
11
ill
Acetylation
Silylation
Alkylation
Iv Perfluorlnatlon»
1 Bromination
11) Chloroacetylatlon
ill) Thlophosphorylation
ivi Silylation
v) Dichlorobenzene eulfonylation
•' DNT/DNP
Pentafluorobaneylation
Tn^l «n kl _..
___________
lodinatioii
Dromination
-Bromobenzoylation
p-Br
2^^-
Vil
1
11
ill
iv _.
v) DNT
vi) PentafluorobenKylatlon
(Amines in general)
1) Transesterlfloation
ii) - ' '
-., Dromlnation
Mncludea trifluoroaeetylatlon, pentafluoropropylation. and heptafluorobutylatlon
-400-
-------
TABLE 10-6
Section 10G
CONFIRMATORY DERIVATIZATION REACTIONS FOR PESTICIDES
OF VARIOUS CLASSES PUBLISHED SINCE 1975
Compounds Studied
Reagent or Derivative (Sensitivity)
Reference
-N0_-containing herbicides
and fungicides
Organonitrogen fungicides
and herbicides
Carbofuran and metabolites
OP pesticides
Sulfoxide-containing
pesticides
Carbaryl
Carbamate insecticides
Dimilin (TH 6040)
Thiabendazole
II-Aryl carbamates
S-Containing carbamates
Carbamate and urea
herbicides
Dimilin (TH 6040)
Azodrin
OP pesticides
s-Triazines
NT-Methyl carbamate
insecticides and
metabolites
CrCl3 reduction to -NH2 followed by CCD-GC (39)
(0.5-1.0 ppm)
Alkylation, methoxylation, trifluoroalkyla- (40)
tion (0.1 ppm)
Heptafluorobutyric anhydride plus tri- (41)
methylamine catalyst (10 pg)
In-block methylation with THAM (yg levels) (42)
Trifluoroacetic anhydride (1 ppm) (43)
N-mono- and trichloroacetyl, and IJ-nitroso (44)
derivatives
Heptafluorobutyryl derivatives (0.1 ppm) (45)
Trifluoroacetyl derivative (0.02 ppm) (46)
Pentafluorobenzyl chloride (0.01 ppra) (47)
Flash heater reaction with trimethylanilinium (48)
hydroxide (ng levels)
Trimethylphenylammonium hydroxide injected (49)
with compound into gas chromatograph
(20 ng)
Alkylation by NaH/CH3I (0.1 ppm) (50)
Conversion to N.N'-dimethyl analog with (51)
NaH/CH3I (0.25 ng)
Trifluoroacetylation (2 ppb) (52)
Oxidation with neutralized NaOCl (53)
(0.25-0.5 ppm in fruits and vegetables)
Trifluoroacetic acid plus a silylation (54)
reagent (20 pmol)
Post-column HPLC fluorometric labeling (55)
with o-phthalaldehyde after basic
hydrolysis
-401-
-------
Sections 10H, 101
10H SPECTROMETRY (SPECTROPHOTOMETRY)
Spectrophotometric methods for residue determination (quantitation)
usually are not as sensitive or selective as GC or TLC, and for this
reason they are.not as widely used as in the early days of pesticide
analysis before chromatographic methods were developed. The appli-
cability of spectrometry is especially limited for multiresidue determina-
tions or analyses of a parent compound, metabolites, and hydrolysis
products.
Spectrometry can be very valuable, however, in conjunction with chroma-
tography as a confirmatory tool, and it is this aspect that will be
stressed in the following subsections.
101 VISIBLE, UV, FLUORESCENCE, AND PHOSPHORESCENCE
Very few pesticides are naturally colored, so a chromophoric group must
be formed by a reaction.or added through derivatization before most
pesticides can be measured in the visible 'spectral region. The colori-
metric method then becomes specific to the color forming 'group involved.
The inferior sensitivity of direct and indirect visible spectrophoto-
metric methods limits their usefulness for confirmation in human and
environmental monitoring where residues are generally present at low
concentrations.
The correlation between UV spectra and pesticide structure and the
usefulness of'UV spectrophotometry in confirming identification have
been reviewed (56). Spectra-structure correlations can be of value to
the analyst in identifying chromophores and therefore making confirma-
tions, especially in conjunction with spectral information obtained by
other methods, such as IR, NMR, and MS. In some cases, extinction
coefficients (absorptivities) are sufficiently large to permit identifica-
tions at submicrogram levels. If a suitable absorption wavelength of a
pesticide can be chosen that is free of interference from contaminants
or solvents, UV spectrophotometry can be performed directly without
sample purification and at a greater saving of time. However, because
absorption of UV energy is quite common for most organic compounds,
rigorous cleanup may be required to remove any interferences that can
absorb in the spectral region where the pesticide will be measured. The
transparency of many functional groups (and often large segments of
complex molecules) in the near UV spectra'l range imposes a limitation
on interpretations of absorption bands in this region. Solvents must be
carefully chosen to be transparent at the wavelengths absorbed by the
pesticide. UV absorbing groups can be added by chemical derivatization
methods, and this procedure has been used to detect pesticides by HPLC
UV detectors and by TLC. UV spectra of 76 reference pesticides have
been published (57). Visible and UV Spectrophotometric methods for
pesticides have been reviewed (58), including development of color by
azo coupling and IT complexing and recent instrumental developments.
-402-
-------
Section 10J
If a pesticide is naturally fluorescent or can be made fluorescent by
derivatization, fluorescence spectrophotometry is likely to be more
selective and sensitive than either visible or UV absorption methods.
Concurrence of fluorescence excitation and emission spectra between
samples and standards, recorded either in solution or directly on
thin layer chromatograms, has served as' a valuable confirmatory aid
for certain pesticides. Fluorescence characteristics are dependent
on a number of experimental conditions which must be closely controlled,
e.g., solvent and pH effects (59). Removal of naturally occurring
fluorescent interferences from biological samples can-pose serious
cleanup problems. Fluorescence and phosphorescence methods for pesti-
cides have been reviewed by Argauer (60), and the phosphorimetry of
pesticides has been reviewed by Baeyens (61).
10J INFRARED (IR)
IR spectroscopy with micro sampling techniques is generally sensitive
at the 1 yg level but has been used as low as the -0.1 yg level in some
applications. It is thus considerably less sensitive than GC or TLC
and cannot be used unless enough sample is available to provide a
sufficient concentration of pesticides for IR observation. Sample
extracts require a stringent cleanup procedure (e.g., partition plus-
column adsorption chromatography) plus additional purification either
by GC or TLC. Thin layer spots are scraped and collected, and the
pesticide is eluted from the adsorbent with an appropriate solution.
Fractions can be collected from a gas chromatograph equipped with a
stream splitter: a small percentage of the effluent stream goes to
the detector for monitoring purposes while the remainder goes to a
collecting device.
Potassium bromide (KBr) micro-pellet techniques using a pellet of 1-2 mm
diameter are described in Section 12,E of the EPA PAM. These methods
were developed by R. C. Blinn of the American Cyanamid Co. The key to
their sensitivity is the ability to transfer the maximum amount of pesti-
cide to a very small amount of KBr to be pressed into the micro-pellet.
The equipment commonly used by Blinn for preparing micro-pellets by the
syringe method is shown in Figure 10-E, and the technique for transfer
of the sample-KBr mixture adhering to the syringe needle is pictured in
Figure 10-F. The method using a commercially available "wick stick"
may be the most reliable and foolproof for preparing micro KBr pellets
(Figure 10-G). The sample is applied to the wedge of potassium bromide,
which is then dipped at its base into a volatile solvent. The solution
migrates up the wedge to the tip where the solvent evaporates, and the
compound becomes concentrated at the tip. The tip is then cut or broken
from the wedge and pressed into a micro-pellet. A procedure using micro
pellets of 0.5 mm diameter to measure 1.4 ug of DDT has been reported (62).
-403-
-------
Section 10J
The Blinn microtechniques are sensitive and reliable, but considerable
experience is required to prepare pellets with a minimum of contamination.
They require the availability of a modern IR spectrophotometer including
a beam condenser and microcells. Contamination from such sources as
the sample, solvent, reagents, atmosphere, and handling is their major
source of error. The same "amounts of interferences that would be
inconsequential for macro-sampling techniques become a significant
percentage of a micro sample and contribute to the spectrum. Clean
gloves should always be worn when preparing micro-pellets, and only
purified solvents and reagents and carefully cleaned equipment should
be used. Inevitable losses due to handling and processing require that
the isolation procedure be started with sufficient sample to finally
achieve a useable spectrum.
Another method that is in effect in a micro-sampling technique involves
scale expansion, or electronic amplification of the signal from the
spectrometer. This method increases pen response without an increase
in the sample concentration, but the response to all interferences and
electronic noise is increased as well. All sources of interference
must, therefore, again be minimized.
IR microtechniques have been reviewed by Blinn (63), including discussion
of micro multiple internal reflectance. Advantages of internal re*-
flectance include ease of applying (by dotting or streaking) sample to
the surface of the reflectance plate (crystal), minimizing of inter-
ferences from handling and reagents, and ease of recovery after IR
evaluation (samples made into pellets are essentially lost for further
scrutiny). A disadvantage is lowered sensitivity compared to the KBr
micro-pellet method. Sensitivity is increased by spreading a very thin
film of sample over the effective sample area of a very thin reflectance
plate. The multiple reflectance method has been applied to the identifi-
cation of Thiram residues at 0.1 ppm on lettuce after extraction, Florisil
chromatography, and TLC (64). An alternative micro-KBr technique with
sensitivity levels similar to the method in the EPA PAM is detailed in
the FDA PAM, Section 631.
-404-
-------
Section 10J
Figure 10-E. Equipment for preparation of micro KBr pellets
(photo courtesy of R. C. Blinn)
Figure 10-F., Illustration of technique of Curry et al.
' (Fhoto courtesy of R. C. Blinn)
A,S Curry,, J F Read, C Srtwff, on4 *W Jftc*io$.
-405-
-------
Section 10J
Figure 10-G. Illustration of wick-stick method
STAINLESS STEEL
HOLDER
STAINLESS
STEEL CAP
SAMPLE CONCENTRATED
AT TIP
GLASS VIAL
«8r "WICK-STICK"
VOLATILE'SOLVENT
The FDA Manual (Section 632),also gives details of a qualitative micro
procedure for collection of GC fractions directly on powdered KBr fov
IR confirmation. Interpretation of IR spectra from collected fractions
must take into consideration the stability of the pesticide of interest
to GC conditions. The analyst should be sure he is measuring the spectrum
of unchanged pesticide rather than of a degradation product. In addition,
the specificity of the GC detector will often obscure elution of inter-
fering materials from the GC column, so that a fraction presumably con-
taining isolated pesticide could be totally unsuitable for IR characteriza-
tion. These interfering materials might be from the sample substrate or
bleed or breakdown products from the stationary phase of the column
packing. The column exit line should be heated at least to column
temperature to the point of trapping, otherwise condensates resulting
from previous samples may contaminate the trapped compound. Use of
splitters at the column exit is usually necessary'because of the high
sensitivity (detectors would be overloaded by the pg quantities for
IR) and the destructive nature of pesticide detectors.
-406-
-------
Section 10J
Several different types of IR detectors directly coupled with gas chroma-
tographs (65) have become available commercially but have not proven
especially useful for pesticide residue work because of various dis-
advantages. Trapping procedures have been used almost exclusively,
including the following methods:
a. Passing column effluent through solvent (66, 67).
b. Condensing effluent on a micro sodium chloride plate.
c. Condensing effluent on a thermo-electrically cooled capillary
plate for internal multiple reflectance IR.
d. Trapping fractions on column packing (68-70). This procedure
is very efficient, and fractions are easily collected for subsequent
IR evaluation; reagent interferences are possible.
e. Collecting on a TLC plate for further cleanup prior to IR (71).
f. Trapping on Millipore or siliconized filter material (72, 73).
g. Using vatious types of liquid nitrogen or dry ice cold surface
traps (74).
h. Using a cool or cold small internal diameter tubing at the
GC vent (75).
i. Trapping directly on KBr powder supported by pipe cleaner inside
capillary tubing (FDA PAM, Section 632). This procedure is probably the
most sensitive of any, tubes can be changed for each peak, and the
technique is free of sources of interferences.-
The choice of trapping procedure will depend on the amount of compound
available, IR technique to be used, purity of the compound eluted from
the GC column, and equipment available to the chemist.
IR spectra of over 400 reference pesticides have been published (76) to
aid the analyst in matching spectra of unknown pesticides. The ASTM
FIRST-1 computer search program (65) and similar computer retrieval
systems aid in matching sample and reference spectra when standards
cannot be easily chosen for a manual point-by-point comparison.
Reference Raman spectra of OC1,OP, and carbamate pesticides were
published (77). Giang has compiled a bibliography with 855 references
published to 1976 on the use of IR spectrophotometry in pesticide analysis
(78).
An important recent development in IR analysis that is capable of sensi-
tivity at subnanogram (79) residue levels is the Fourier transform (FT)
or interferometric method. In FT-IR, a Michelson interferometer is
used instead of the prism or grating and slits in a conventional spectro-
meter. The slitless spectrometer has an advantage in energy throughput,
-407-
-------
Section 10K
in addition to the so-called Fellget's or multiplex advantage that allows
all wavelengths to be detected simultaneously throughout the spectral
range. The signal to noise ratio increases with consecutive accumulation
of scans and is proportional to the square root of the number of scans.
Since each scan requires only a few seconds and instrument stability is
high, many cumulative scans can be made on each sample. The fast scan
capability is ideal for on-the-fly IR detection of GC effluents. FT-IR
spectroscopy has at least an order of magnitude greater resolving power,
greater wavelength accuracy, and a greater scan range than does con-
ventional dispersion IR spectroscopy. There is also a much smaller image
in the sample compartment without any special measures, making FT-IR
ideal for microsamples. A dedicated minicomputer, in addition to the
basic FT-IR optical equipment and detector, is required to collect, process,
and store the data. FT-IR methodology and equipment have been reviewed
(65). There is no doubt that much use will be madg of FT-IR spectroscopy
for pesticide determination and confirmation as the principles, techniques,
and instrumentation become more familiar.
10K NUCLEAR MAGNETIC RESONANCE (NMR)
NMR spectroscopy has had only limited application in residue analysis
"because of its low sensitivity relative to other analytical methods,
e.g., GC-MS, IR, and UV. Despite this drawback, it is one of the most
valuable tools available for structural analysis and identity confirma-
tion. Current pulsed Fourier transform NMR spectrometers (80) allow
routine acquisition of useful data on as little as 10 yg of a proton NMR.
sample in a few minutes of experimental time. The NMR sensitivity of
13C is lower; with current commercial instrumentation, a practical sample
size is greater than 20 mg, although 13C spectra of as little as 300 ug
have been obtained on modified instruments (81) . Useful information is
provided by NMR in many areas relevant to the analysis of pesticides,
their metabolites, and degradation products, such as identification and
structural characterization, molecular geometries, conformations and
stereochemistry, chemical kinetics and equilibria, complex formation
and binding, and electronic charge distributions.
Residues of £,£f-DDT and p^ja'-DDE isolated from adipose and liver tissue
samples have been analyzed by NMR (82) , with semi-quantitative determina-
tion of the relative concentrations of the pesticides. Included in NMR
studies of the metabolism, binding, and degradation of pesticides are
% spectra useful for identification of £,£'-DDT (83, 84), £,_p_'-DDA (85),
aldrin and dieldrin derivatives and other chlorinated pesticides (86) ,
rotenoids (87), and dithiocarbamates (88). Other IH reference spectra
of organophosphorus (89), diphenylme thane (DDT type) (90), and carbamate
(91) pesticides have been published and are useful for identity confirma-
tion. The application of NMR to pesticide analysis has been reviewed
(80, 92, 93).
Carbon-13 NMR spectra have been published for a-BHC (94) , for several
chlorinated biphenyls (95, 96), and for 30 chlorinated polycyclodiene
pesticides (97). Studies of technical chlordane components (98), mirex
(99, 100), and Kepone and its photo-products (101) have provided
-408-
-------
Section 10L
NMR data useful for confirmation and structural characterization. Chlorine
nuclear quadrupole resonance spectrometry has been used to study the
structures of several chlorinated pesticides including BHC, aldrin,
endrin, endosulfan, and dieldrin (102-104). 31P- NMR chemical shifts
have been correlated with structures of some organophosphorus pesticides
(105), and 31P Fourier transform NMR has been used for the determination
of malathion at ppm levels (106).
10L MASS SPECTROMETRY (MS)
The mass spectrometer is a very sensitive spectroscopic tool for pesticide
residue analysis, providing useful data on ng or pg residue levels. Ions
are produced from neutral sample molecules and are then sorted according
to their mass-to-charge ratio (m/z). The mass spectrum is a record of
these different ions and their relative abundance. A mass spectrum is
usually quite characteristic of an individual pesticide, sometimes even
providing data that will differentiate geometric isomers. Pesticide
identifications can be made by matching the mass spectrum of an unknown
sample with the mass spectrum of a known material. This comparison is
especially valuable because it is based on many different peaks character-
istic of the unknown compound. The composition of. an unknown compound can
be obtained without comparison to a reference material by making exact
mass measurements of the molecular ion and, other key fragment ions in
the spectrum. Because the exact mass of every element has a unique
fractional value on a scale compared to 12C = 12.0000, any combination
of these elements into a chemical formula will have a unique fractional
mass, specific for that combination of elements. The exact masses used
to determine the chemical composition of a compound can be obtained on
either a low or high resolution spectrometer. The advantage of a high
resolution instrument is the ability to separate ions with different
compositions that are at the same nominal mass (e.g., m/z 28 for CO, N2,
C2H4) and to obtain accurate mass values for these ions. The molecular
ion is the species .resulting from the removal of a single electron from
a molecule. After the recommendation of Benyon (107), the symbol M is used
to represent the odd-electron molecular ion formed from an even electron
molecule.
a. MS Instrumentation and Operation
(1) Introduction
Five components are common to most mass spectrometers: the
inlet system, the ion source, the mass analyzer, the detector, and the
readout system.' In ad'dition, a vacuum must be maintained throughout the
spectrometer from inlet to detector so that ions formed in the source will
not be lost from collisions with atomspheric gas molecules. A second
reason for maintenance of vacuum is to prevent oxidation of the filament
in the ion source and various other inside parts of the spectrometer and
-409-
-------
Section 10L
electron multiplier. A sample is introduced via the inlet system into the
ion source, where it is ionized. The function of the inlet system is to
transfer the sample from a high pressure (i.e., 1 atm) region into the
vacuum of the spectrometer without seriously unbalancing the spectrometer
operation. The generated beam of ions is focused and separated in the
mass analyzer according to the m/z ratios. The detection system senses
the mass-separated ion beams, and the readout device translates the
signal provided by the detection system into an output that can be
interpreted by the analyst. Several reviews of pesticide residue
analyses by MS have been written by Safe (108), Skinner and Greenhalgh
(109), and Ryan (110). In addition, detailed reviews of mass spectrometry
have been made by Burlingame et al. (Ill) and Alford (112).
(2) Inlet Systems: Direct Insertion Probe
Samples may be introduced directly into the ion source with
a direct insertion probe assembly. For example, the sample is loaded into
a short length of melting point capillary, placed in'the heater well at the
end of a probe, and inserted to within a few millimeters of the ion source
through a vacuum lock that maintains a vacuum-tight arrangement. The
temperature is then increased until the sample vaporizes and a spectrum
is obtained. The introduction of trapped GC fractions into an inde-
pendent mass spectrometer by these techniques was used for residue
analysis prior to the development of combined gas chromatography/mass
spectrometry (e.g., 113), but the latter procedure is now employed
almost exclusively. The direct insertion probe is reserved mainly
for samples that cannot be chromatographed, because of thermal instability,
low vapor pressure, and/or high polarity. However, combined with specific
ionization techniques, such as negative chemical ionization, direct
insertion can provide a sensitive, rapid screening method (113A).
Combined GC/MS
For Impure samples such as biological extracts, the gas chromatograph of
a coupled GC/MS instrument serves as an efficient inlet system for intro-
duction of samples into the spectrometer. The resolution provided by gas
chromatography offers extra sample cleanup in addition to any partition
and liquid chromatography steps. Temperature and sometimes flow rate
programming have proven useful for achieving high chromatography resolution
with the combined instrument. Column bleed can be a serious proble in
GC/MS, since bleeding liquid phase is also detected by the mass spectro-
meter and contributes spurious ions to the analytical spectra. Carefully
conditioned, low bleed columns that are stable at high temperatures should
be used whenever possible. Other approaches that alleviate problems from
column bleed include use of a short, bleed-absorbing column placed between
the analytical column and the GC/MS interface, programming the flow rate
of the carrier gas, and computer subtraction of background resulting from
the bleed (114).
-410-
-------
Section 10L
Compatability of the gas chromatograph and mass spectrometer is a problem
because of the large volume of carrier gas eluting from the chromatograph
and the need to operate the spectrometer at high vacuum (10~5 - l.0~6 mm Hg).
In the simplest approach, the two instruments are connected directly, and
a large pumping system is used to maintain the required vacuum in the mass
spectrometer. This approach has been used successfully with GC columns
having flow rates up to ca 5.0 ml/minute. Introduction of samples from
packed columns into the mass spectrometer requires removal of most of
the carrier gas by means of an interface between the two instruments.
At the same time, as much sample as possible should be retained so that
the gas flowing into the spectrometer is enriched in sample. Three basic
types of sample enriching devices or separators have widespread use in
modern GC/MS systems, namely effusion (Watson-Biemann; Brunnee), jet
(Ryhage), and membrane (Llewellyn~Littlejohn). Each has its own advantages
and limitations, and there appears to be no strong preference for one over
the other. In all cases, some carrier gas enters the source along with
the sample molecules, and broadening of GC peaks by the interface may
occur. In practice, most separators convey only 20-40% of the sample
in the GC effluent to the mass spectrometer. The theory and operation of
separators have been described in detail by McFadden (115, 116). The
mass spectrometer in a combined instrument must be able to scan through
an appropriate mass range, e.g., from mass 10 to mass 800, in a small
fraction of the time that it takes to elute the peaks from the gas
chromatograph. •
Combined LC/MS
GC/MS is sometimes limited by the volatility or heat sensitivity of the
compounds under study. To circumvent these difficulties, various methods
of interfacing a high pressure liquid chromatograph with a mass spectro-
meter have been explored. The demands on a LC/MS interface are much more
extreme than for GC/MS, because of the greater enrichment required
(usually 10^) and the possible adverse effects (e.g., background inter-
ference, chemical ionization effects, filament damage, etc.) of excess
solvent entering the ion chamber. Six methods have been used for LC/MS
interfacing. Three methods, namely the high capacity atmospheric pressure
ionization source, the semipermeable dimethyl silicone membrane, and
modification of sample at the interface by reduction to hydrocarbon, have
not been widely, accepted. Methods involving direct introduction with no
enrichment, direct introduction with jet enrichment, and mechanical
transfer using a moving wire with belt are most promising and are under
active development. All six methods have been described and compared
by McFadden (116), with appropriate original literature references.
The most common commercial liquid chromatograph/mass spectrometer inter-
face (117) (Figure 10-H) consists of a continuous belt that introduces
the LC effluent into a chamber at atmospheric pressure and then sequentially
passes it beneath an infrared heater and through two vacuum locks into a
vaporization chamber. Under optimum conditions the LC solvent is evaporated
from the belt by the heater and vacuum locks, leaving only a deposit of
the sample on the belt. The vacuum locks also accomplish the transition
-411
-------
Section 10L
from atmospheric pressure to the vacuum system of the mass spectrometer.
In the vaporization chamber, a second heater volatilizes the sample in
front of a nipple leading into the ion source. A third heater cleans
the helt before its return to the atmospheric chamber via the vacuum
locks.
Figure 10-H.
HPLC/MS interface developed under contract by
Finnigan Corporation* for EPA.
LC EFFLUENT
FLASH
VAPORIZER
INFRARED
REFLECTOR
ION
SOURCE
CLEAN-UP
HEATER
DRIVE
WHEELS
SPRING
LOADED
IDLER
WHEEL
This interface is able to accommodate most commonly used organic LC
solvents at optimum flow rates varying from about 0.2 to 1.5 ml/minute.
The use of water as an LC solvent generally requires ail LC effluent
splitter if reasonable LC flow rates are to be used, since the maximum
capacity of the interface for water appears to be about 0.1 ml/minute.
The LC/MS system has been successfully applied to the analysis of a
large number of carbamate pesticides (117).
The field of LC/MS is still under development. A second LC/MS interface
has been introduced by Hewlett-Packard, and other commercial interfaces
are anticipated in the near future.
(3) lonization Processes: Electron Impact (El)
The most widely used ionization source is the electron impact
type wherein gaseous molecules are ionized by electrons emitted from a
* Mention of commercial products does not constitute endorsement by the U.S.
EPA.
-412-
-------
Section 10L
glowing filament. These positive ions are accelerated into the analyzer
section. The El source is relatively stable and easy to operate, and
has high ionization efficiency. The overall quantity of positive ions
and the nature of the fragmentation process depend on the energy of the
ionizing electron beam.
Upon ionization of many compounds at low electron energy levels (0-20
electron volts or eV), a large fraction of the ion current tends to be
carried by unfragmented molecular ions. However, the absolute intensity
is relatively -low. At higher energy levels, fragmentation and rearrange-
ment are more prevalent, and the ion current is much higher. Molecules
are often cleaved to such an extent that the molecular ion is absent from
the.mass spectrum or is of very low intensity. (A great many other com-
pounds form only fragments even at low eV values.) Because mass spectra
are more reproducible when compounds are ionized by 60-80 eV electrons,
most, mass spectrometers are operated in this energy range. It is note-
worthy that the El source produces mass spectra that are quite repeatable
among instruments and distinctively characteristic of the compounds being
ionized. This has led to the collection of large libraries of mass
spectral data with which unknown spectra can be compared. Such comparisons
often permit rapid identification of the unknown pesticide.
Chemical Ionization (CI)
Chemical ionization spectra are obtained by adding methane, helium, or
other reagent gas (at relatively high pressures of about 1 mm or 130 Pa Hg)
to the sample either as the GC carrier gas or after removal of the GC
carrier gas by the separator. In the latter case, the CI reagent gas
is introduced into the mass spectrometer just ahead of the point at which
the effluent enters the ion source, or into the source itself. Electrons
produce reagent gas ions that subsequently ionize sample molecules by
chemical reactions, e.g., proton transfer, hydride abstraction, .ion
attachment, and resonance transfer. The mass spectra obtained with CI
are quite different from those formed on electron impact and are, in
general, simple and complementary to electron impact spectra for pesticide
confirmation. Although CI usually provides molecular ion (MT-) or
(M + H)+ or (M - H)+ peaks of high intensity, a study (118) of a series
of chlorinated and organophosphorus pesticides found no molecular ion
or ions in the molecular ion region produced from electron impact or
chemical ionization for a number of specific compounds. CIMS has sensi-
tivity at least as good as that of El (119) and offers the advantage of
allowing characterization of a sample's chemical reactivity through the
choice of the reagent gases. In addition to methane and helium, other
gases including isobutane, hydrogen, argon-water, ammonia, and nitric
acid have been used successfully to produce CI spectra (120). Positive
CI data for 29 OP insecticides and metabolites have been reported (121).
-413-
-------
Section 10L
Field lonization (FI)
Field ionization involves passing a gaseous compound between an anode
(usually a thin wire or sharp blade) and a cathode. An extraordinarily
high electric field, approximately 108 V/cm, is impressed on the anode,
permitting (as most commonly explained) valence electrons of the sample
to "tunnel" to the metal of the wire or blade. Charge separation can
take the form of electrons tunneling out of molecules, proton transfers
(facilitated by tuneling) between molecules, separation of the oppositely
charged ions in electrolytes, and so on (111). A positive ion results,
which can be separated according to mass-to-charge ratio and detected.
This is a relatively low energy ionization method that often produces
enhanced molecular ion intensities and a cleaner, spectrum for compounds
with poor thermal stability.
Fragmentations are less prevalent and different from those observed in
normal El spectra. The FIMS of a number of pesticides has been studied
(122), and FIMS has been combined sequentially with HPLC for the determina-
tion of trifluralin (123).
Field Desorption (FD)
Field desorption MS is a modification of FI in which the sample is applied
directly to a carbon or metallic filament anode. As with FIMS, field
desorption depends on application of very high electric fields (5000-10,000 V)
to this anode. Sample molecules in contact with the anode desorb as ions
into the source, where they are separated and mass analyzed. Like FI, field
desorption produces ions of low internal energy, and usually results in
minimal sample fragmentation. Unlike FI, field desorption has no require-
ment that the compound be volatile prior to ionization. Mass- spectra can
be obtained for samples that are thermally unstable or have no appreciable
vapor pressure, as for example, salts. The field desorption mass spectrum
of endrin and its El spectrum, which has a low abundance of the molecular
ion, are shown in Figure 10-1 (124). Strong molecular ion peaks are pro-
duced for most pesticides (118), including highly polar pesticides and
metabolites such as carbamates and ureas (124-126). Impurities may also
give only molecular ions, so interpretation of mass spectra is sometimes
simplified and the necessity of sample cleanup reduced. However, assign-
ment of molecular ions and interpretation of spectra in biological samples
can be complicated by the presence of (M + H)+, (M + Na)+, or other ion
adducts and cluster ions. Other disadvantages are that quantitative data
are difficult to obtain by FDMS, and valuable structural information pro-
.vided by fragmentation is lost.
Atmospheric Pressure lonization (API)
—12 —15
A novel method with high sensitivity (10 - 10 g) involves generation
of ions with an atmospheric pressure ionization source. The API instrument
-414-
-------
Section 10L
is essentially an electron capture detector with suitable interfacing to
a mass spectrometer so that its ions can be mass-identified. The source
uses °%i on gold foil to produce electrons that can interact with
nitrogen and water passing through the ionization chamber at atmospheric
pressure. Preheated carrier gas enters the API source just behind the
sample injection port. Both gas and sample pass through the *>%i source
block where ionization reactions take place. Just beyond the chamber
is a small aperture "through which the ions pass on their way to being
mass analyzed and detected. With certain samples, this source generates
more ions for a given quantity of sample molecules than any other ion
source; this is reflected in the reference to the API mass spectrometer
as the "femtogram machine" (127). The ^%i source has been replaced by
a corona discharge (128) , producing identical API mass spectra and limits
of detection but a greater dynamic response range. Qualitative and
quantitative applications of negative ion formation from pesticides in
the API mass spectrometer have been studied (129) .
Figure 10-1. Electron impact (El) and field desorption (FD)
mass spectra of endrin (124). •
"
210 230 250 270 290 310 330 350 370 390
i«h
40-
FD
300 310 330 330 3IO
370 360 390 400
-415-
-------
Section 10L
Negative Chemical lonization
An important new development in chemical ionization methodology is
simultaneously pulsed positive and negative CI mass spectrometry,
developed by Hunt et al. (130). In this method, both the positive
and negative ions produced in a CI source are alternately pulsed from
the source, with appropriate potentials, through a quadrupole analyzer
to two electron multipliers, one for positive and one for negative ions.
Positive and negative mass spectra are, thereby, measured "simultaneously".
Under favorable circumstances, negative CI can afford sensitivity two or
three orders of magnitude greater than that obtainable with positive CI
(131), making it a very relevant technique for residue analysis of
pesticides and dioxins.
The positive and negative methane (132) and isobutane (133) CI mass
spectra of selected polycyclic and aromatic chlorinated insecticides
of several types have been determined and published. The negative
CI spectra with isobutane as enhancement gas were exceptionally simple,
with the most abundant ion for almost all compounds studied being
(M + Cl)~ (133).
Negative CIMS with methylene chloride reagent gas was the basis of a
multiresidue screening procedure for OC1 residues in environmental
substrates at 1 ng levels (134). The four principle negative ion-forming'
reactions with methylene chloride as reagent gas were (1) resonance
capture of an electron to give M~ ; (2) chloride attachment to hydrogen-
bonding or carbon-bonding substrates to give (M + Cl)~; (3) deprotonization
or dissociative capture of an electron for relatively strong gas phase
acids to give (M - H)~ ; and (4) oxygen-chloride exchange to give
(M - CI +'0)~ (113A, 134).
Polychlorinated dibenzo-p_-dioxins were determined in biological samples
by methane negative CIMS, which was found to be as much as 1000-fold more
sensitive than methane positive CI and electron impact MS. The use of
oxygen with or without methane resulted in decreased sensitivity but
increased selectivity for the dioxins. Detection limits ranged from
100 to 500 pg for 2,3,7,8-TCDD down to ca 1-10 pg for 1,2,3,4,6,7,8-HpCDD
by selective ion monitoring (131). The highly toxic 2,3,7,8~TCDD can be
distinguished from other isomers by negative chemical ionization and
reaction with oxygen to form dichloroquinoxide ions (134A, 134B).
A discussion of 13 methods for ionization of organic compounds in MS
has been published, including detailed consideration of chemical ioniza-
tion and field ionization and pesticide spectra (135). Design con-
siderations of El, CI, FI, FD, and API sources have been described (136).
Five ionization methods were compared for producing positive and negative
ion mass spectra of typical organophosphorus pesticides. The negative
ionization techniques were much more sensitive for the 16 compounds
tested (137).
-416-
-------
Section 10L
(4) Mass Analyzer Systems
Low resolution magnetic analyzer systems depend on bending of
the ion beam in a magnetic .field. The magnetic field segregates the ions
into beams, each of a different m/z. To obtain the mass spectrum, the
magnetic field is varied, and each m/z ion from light to heavy is
successively brought to focus on the exit slit. Such analyzers are
referred to as single- or direction-focusing analyzers. High resolution
instruments have an analyzer region with an electrostatic sector for
velocity or kinetic energy focusing plus a magnetic sector for separation
of fragments according to m/z ratio.
Quadrupole analyzers are based on mass separation in a radio frequency
(RFX•electric field. This field is established on a set of four
precision parallel, usually circular, rods, with both a DC voltage
and an RF alternating voltage being applied to these rods. Ions are
accelerated gently (5-30 V) into the analyzer or filter region, and
begin to oscillate between the rods. At a given DC and RF level, ions
of a specified m/z value undergo stable oscillations and pass through
the length of the analyzer tube to the detector. Ions of lower or
higher mass will undergo increasingly erratic oscillations that eventually
result in their striking the rods or walls. The spectrum is obtained by
sweeping ^the applied RF voltage and DC ramp voltage and measuring the
detector, current as a function of time.
, (5) Resolution ,
. Resolution describes the performance of the mass analyzer
in terms of its ability to separate ions of different masses from one
another. Resolution is expressed in numerical form by the equation
M/AM where M and M + AM are mass numbers of two neighboring peaks of
equal intensity in the mass spectrum. The criterion for resolution is
a relative height of the valley between peaks of 10%, with.each peak
contributing 5% to the valley. . For example, an instrument would, have
a resolution of 100 if two peaks with a mass difference of 1 .part in
100 (e.g.,- m/e 100 and 101) were resolved to the 10% level. Low resolu-
tion mass spectrometers typically show maximum resolution values between
300 and 1000, while high resolution instruments are capable of attaining
resolutions well in excess of 104. The advantage of a high resolution
spectrometer is the capability of resolving ions with very little
differences in mass and obtaining the masses of these ions accurately
to 0.001 mass units or better. Exact masses are determined using a
computer coupled to the mass spectrometer or by peak matching known
marker peaks and unknown peaks on an oscilloscope (138). Once the
exact mass of a key ion (often the molecular ion) is known, the elemental
composition or formula of the molecular or fragment ion is obtained,
again by using a computer or by consulting tabulations of the mass.es
of different combinations of atoms. Elements .indicated to be present
by the mass spectral pattern or prior information about, the unknown sample
are often needed to correctly evaluate the data.
-417-
-------
Section 10L
Resolutions of the order of 1000 are attainable with low resolution
magnetic and quadrupole analyzer designs, although single-focusing
magnetic analyzers can attain higher resolutions with an extreme decrease
in sensitivity due to the narrow slits that must be used. Resolution in
excess of 8000 is considered high, since this is the amount usually
necessary to resolve most mass doublets. The extra focusing added in
& high resolution mass spectrometer reduces the overall number of ions
traversing the instrument, thus reducing the overall sensitivity. To
overcome such a reduction, the mass range is usually scanned at a slow
rate. To minimize the effects from slow scanning and decreased sensi-
tivity, only as much resolution as is necessary to perform the requited
analysis should be used, since the accuracy of an exact mass measurement
is independent of resolution as long as any mass doublets are separated.
References (108, 110, 139-141) review methods and applications of MS
and combined GO/MS to pesticide residue analysis, and references (111,
112) give a more general survey of GC/MS instrumentation, principles,
and techniques.
b. Examples of GC/MS Confirmation
Figure 10-J shows the electron capture gas chromatogram obtained
by injection of an aliquot of the 6% ethyl ether Florisil column eluate
from cleanup of a human adipose tissue extract (142). Figure 10-K shows
the total ion current chromatogram of the same eluate from GC-MS. Although
the curves are drawn to different scales and are not directly comparable,
it is evident that many more compounds are identifiable in the latter
because of the general response of the mass spectrometer. In general,
chromatograms traced by the total ion monitor are similar, but not
necessarily identical, in response and sensitivity to those traced
by a flame ionization detector. Differences exist in sensitivities
to some compounds, and broadening occurs in some peaks in the interface
to the mass spectrometer. Figure 10-L shows the mass spectrum of
standard j3,jj'-DDE, the major GC peak evident in both chromatograms in
Figures 10-J and 10-K.
The identification of pesticides from their mass spectra is often
complicated by the obscuring of low mass ions by impurity fragments,
especially in biological extracts. For this reason, extra cleanup
of extracts may be needed for GC-MS as compared to GC alone. For
example, alkaline hydrolysis has been used for the 15% ethyl ether
Florisil column eluate, while additional column adsorption cleanup
(e.g., alumina plus Florisil columns) or use of silica gel rather than
Florisil initially has been successful for the 6% ethyl ether eluate.
Gel permeation chromatography has also been successfully applied to the
6 and 15% fractions (143).
-418-
-------
Section 10L
Figure 10-J.
Electron capture chromatogram of human adipose tissue
extract, 6% ether Florisil column eluate
!
I
I?
INJECTS?!
Figure 10-K, Total ion current profile of the same human adipose
tissue extract
-419-
-------
Section 10L
c. The Mass Spectrometer as a GC Detector
There are a number of ways to use the mass spectrometer as a
sensitive and selective GC dete'ctor. These procedures require that the
analyst know what compound or compounds he/she is looking for and are
not applicable to totally unknown samples.
Selected ion monitoring (SIM), also called multiple ion detection (MID)
or multiple ion selection (MIS), involves automatic, continuous
monitoring of a few ions of different masses. Tracings of the selected
masses are recorded simultaneously as rapid switching is accomplished
in the spectrometer to bring each ion into the detector in turn for a
short period of time. Simultaneous recording of one or several compounds
can be achieved, with characterization of each being based on the
formation of one or more selected ions (144). To use SIM effectively,
one should know the kind of compound sought and its MS characteristics.
Sensitivity of detection for SIM can sometimes be extended to the sub-
picogram range, which is considerably more sensitive than conventional
scanning because of the longer sampling time at each selected mass.
Sensitivity for a particular compound is influenced by the extent of
fragmentation and the fraction of the total ion current carried by the
selected ions. Identification and quantitation of compounds can be
improved by exact mass measurement (e.g., to 0.001 amu) of the specified
ion, but only at the expense of sensitivity (144A). Figure 10-M shows
the m/z 405, 407, 409, and 411 ions of trans-nonachlor and isomers
monitored simultaneously in a human adipose tissue. Total ion current
profiles (TICP) cannot be generated by the SIM technique because data
from only certain masses are collected.
Compounds not resolved by gas chromatography can still be detected with
certainty if their molecular (or other characteristic) ions can be re-
solved by SIM. Recording the masses and relative intensities of several
ions formed from a single pesticide can increase the certainty of compound
identification. SIM has been applied to the detection of organophosphorus
insecticides (145) and to carbofuran and metabolites in crops (146).
Repetitive scanning through a narrow mass range generates quantifiable
spectral envelopes from several ions at once. This procedure, generally
sensitive at low ng levels, has been applied to pesticide analysis (147).
Reagent ion monitoring is an interesting variation of single ion monitoring,
wherein the intensity of reagent ions used in a chemical ionization source
is monitored as a function of time. The intensities of reagent ions de-
crease when they react with material eluted from the GC column, providing
a chromatogram that is distinctive from those produced by other detectors
(148).
-420-
-------
Section 10L
Figure 10-L. Total mass spectrum of £,pf-DDE
I00-|
80-
>,60-
(/)
§
J=j 40-
4>
_>
•§ 20-
o:
A^'-DDE
/T^\ /7
C'~\ /~?~\
\ f c X.
Cl Cl
1
. ISO
1
,= 105
75
I lib
ill fi i
li "i ' il I 1 II 1 i
.... . i . jllii Ji.ii.iilA.ajlLiylL Jli. J LuLt j. s J.L JjJi
1 1 ' 1 ' I ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 • 1 ' 1 ' IT")"!"")1 I p fT
50 100 150
~^\ ,
\>— i
^/
'6
lUt _ s
_j
6 2.
Q
gf
3_
K"
_ Q
3 -•
O
3
200 250 300
Figure 10-M. Selected ion monitoring applied to a human adipose tissue
extract. The four masses shown have been monitored as
the extract elutes from an-OV-17/OV-210 GC column at 180°C.
The largest peak is trans-nonachlor, the last eluting
peak is cis-nonachlor, and the peak preceding trans-
nonachlor is an isomeric nonachlor also observed to be
_ present in technical chlordane.
411.0
409.0
407.0
405.0
-421-
-------
Section 10L
d. Computerization of GC-MS
Combination of a computer with a GC-MS system can serve several
very useful functions.
(1) The computerized GC-MS data acquisition system permits rapid
processing of information from complex sample mixtures. The mass spectra
of specific compounds in the mixture can be experimentally obtained and
automatically matched with a library file of standard mass spectra. Com-
puter control of data acquisition may enable the operator to devise
relatively complex scanning procedures. For example, different mass
ranges may be sampled for different time periods, masses may be sampled
for times related to the intensities being measured, or several dis-
continuous mass ranges may be sampled.
(2) Column bleed and other background can be conveniently sub-
tracted by the computer.
(3) Continuous repetitive scans can be made during the entire
chromatographic separation; for example, a spectrum can be scanned every
2-4 seconds. In a typical GC/MS run, several hundred to more than a
thousand mass spectra may be acquired in this way, each one being a complete
spectrum over the mass range selected.
All spectra are stored, and chromatograms may later be reconstructed by
the computer by summing and plotting the total ion current detected in
each scan but excluding carrier gas ions or other interfering ions.
Reconstructed total ion current profile chromatograms (TICP) obtained
resemble those traced in real time by a conventional total ion monitor
of a magnetic deflection spectrometer. A typical reconstructed GC/MS
total ion current profile of an extract of human fat is shown in Figure
10-N, with some of the components identified (142).
(4) The computer can trace the intensities of selected character-
istic masses from among the great quantity of data acquired by continuous
repetitive scanning. The resulting mass chromatograms or extracted ion
current profiles (EICP) (149) resemble the single or selected ion profiles
described earlier and permit compounds and spectra of interest to be
located and the appropriate spectrum to be retrieved and plotted. EICPs
can be individual selected ion current profiles (SICP) or summed sets of
several masses, all extracted from scanned data. EICPs have an advantage
over SIM in that large numbers of ion profiles and complete spectra can
be examined rapidly after only one chromatographic separation, but this
computerized acquisition of .repetitively scanned spectra is of considerably
lower sensitivity (as much as 10 ) than SIM because of the longer integra-
tion time characteristic of the latter method (149). Reference (140)
illustrates computer-generated selected ion current profiles.
-422-
-------
Section 10L
Figure 10-N..
Computer reconstructed total ion chromatogram and mass
chromatograms of M = 237 (o,j>'-DDT and j),j>'-DDT) and
M s 405 (trans-nonachlor) from a composite human adipose
tissue extract. Column: 45.7 m SCOT column coated with
SE-30, programmed from 170-240°C at 2°C/minute (N.C. * not
chlorinated).
M-OOT
M« 237.0
100 1W
SCANNVMCR
M-408.0
1
The limited mass range chromatogram (148) is a variation of mass chroma-
tography that has proved especially valuable in the determination of
polychlorinated hydrocarbons. In this technique, the computer sums
ion intensities (collected from repetitive scanning) through a limited
mass range as a function of scan number or time. (The procedure has
also been termed selected ion summation analysis or SIS.) For example,
the molecular ion cluster of mirex, due to the contributions of 37ci
from each of the 12 chlorine atoms, is spread over a range of more than
27 V. Instead of treating a single ion (e.g., CiQ^di2+' > nominal
m/z 540), the entire cluster can be summed to provide increased sensi-
tivity with some sacrifice in specificity. The method has been used to
identify dieldrin and HCB residues in lake trout (150).
-423-
-------
Section 10L
(5) Quantitation of peak areas in the selected ion profiles and
ratios of these peaks can be provided.
Computer coupled GC/MS equipment is extremely expensive, and
highly qualified personnel are needed for operation, maintenance, and
interpretation of data. A significant amount of "down-time" is to be
anticipated because of the complex nature of the instrumentation. Computer-
ized data acquisition and processing for magnetic instruments, quadrupole
instruments, and selected ion monitoring have been described (110), as
have techniques available for computer identification of unknown mass
spectra using various retrieval systems (151).
e. Applications of GC/MS to Pesticide Analysis
Reference spectra and fragmentation data for pesticides of several
types and for related chemicals have been published (105, 108, 152-156).
Applications of GC/MS include confirmation of the 1-naphthyl chloroacetate
derivative of 1-naphthol (a carbaryl metabolite) extracted from urine
(157); 2,4-D, 2,4,5-T, and 2,4,5-TCP in urine (158, 159); organophosphorus
pesticides in blood and urine (160, 161) and food (162); multiple
chlorinated insecticides in human adipose and liver tissue (142, 143,
163), foods (164), and soils (165); toxaphene in human and biological
samples (166); Kepone in human and environmental samples (167); chlordane-
related residues in human samples (142, 168); thiabendazole and
5-hydroxythiabendazole in animal tissue (on-column methylation plus SIM)
(169); dimethoate residues in "wheat by SIM at m/z 87 (170); and mirex in
fish (171).
An important application of GC/MS has been mutual determination and
identification of PCBs in the presence of chlorinated pesticides (172).
Insecticides mixed with PCBs have been identified at levels below 10 ng
without complete separation on a GC column by peak monitoring' MS as
described earlier (173). GC/MS has been successfully.applied to the
detailed analysis of complex pesticide mixtures, such as technical
chlordane (168). Pesticides and PCBs have also been identified by
GC/MS using chlorine isotope ratios to reconstruct chromatograms that
are characteristic for the number of chlorine atoms found in repetitive-
scan spectra (174).
Special MS and GC/MS techniques that have been applied to the analysis
of simple and complex pesticides in a variety of sample substrates include
selected ion monitoring (175, 176), field ionization (177), and field
desorption MS (178). Methods have also been developed for the determina-
tion of carbamates and ureas by combined liquid chromatography/mass
spectrometry (117). References (108, 110-112, 179) contain reviews of
applications to residue analysis. Symbolism and nomenclature of mass
spectrometry have been reviewed (107).
-424-
-------
Section 10M
f. Mass Spectrometry/Mass Spectrometry (MS/MS)
Mass spectrometrists have, within the last few years, investigated
the possible elimination of any preseparation method, such as GC or LC,
for the analysis of complex mixtures. Instead, the mass spectrometer
itself is used as the separation device, followed by a second mass
spectrometric analysis of the sample. This technique is called mass
Spectrometry/mass Spectrometry. Techniques are available to perform
this method using quadrupole or magnetic sector instruments with positive
or negative ions. Operation commonly involves the separation of the ions
of a particular m/z value, characteristic of a given compound present in
a complex mixture, by a first mass spectrometer. This ion current then
encounters collisions with gas molecules, which impart considerable energy
to them through the process of collisional activation. The resulting
energetic ions may-then decompose into characteristic fragments, which
are then analyzed in a second or third mass analyzer region, as the case
may be. This method,holds promise as a rapid method of mixture analysis.
Hunt et al. (179A) have used MS/MS to analyze nitrophenols in sewage
sludge. However, recent studies show that artifacts can be created in
the analysis (111). Other references involving MS/MS' include (179B, 179C).
10M QUALITY ASSURANCE OF GC-LOW RESOLUTION MASS SPECTROMETRY
This section reviews procedures to be followed for quality assurance of
data derived from the mass spectrometer in the identification, confirma-
tion, and quantitative determination of chlorinated insecticides, PCBs,
hexachlorophene, and PBBs in human tissues and fluids. These methods
were developed at the Health Effects Research Laboratory, U.S. EPA
Research Triangle Park, NC (180) for use in the EPA National Human
Monitoring Program for adipose tissue and serum samples; The procedures
assure interpretable mass spectra of the highest experimentally obtainable
quality for compound identification as well as quantitative accuracy when
monitoring ion intensities (as by selected ion monitoring). The specific
pesticides of current interest are the following:
'-DDT
£»£.'
p_,p_'-DDT
£»£.' -DDE
£.»£* -DDE
£,£/ -DDD
p_,p_'-DDD
cc-BHC
g-BHC
Lindane (y-BHC)
5-BHC
Aldrin
Dieldrin
Heptachlor
Heptachlor epoxide
Endrin
Mirex ' '
Oxychlordane
trans-Nonachlor
Polychlorinated biphenyls
Hexachlorobenzene
Polybrominated biphenyls
Polychlorinated terphenyls
-425-
-------
Section 10M
a. Introduction to Quality Assurance Procedures
Correct identification of organic pollutants from gas chromatography-
mass spectrometry (GC/MS) data requires valid mass spectra of the compounds
detected. This is independent of the actual method of interpretation of
the spectra, i.e., an empirical search for a match within a collection of
authentic spectra or an analysis from the principles of organic ion frag-
mentation. A properly operating and well tuned GC/MS instrument is re-
quired to obtain valid mass spectra.
The purpose of the following procedure is to permit a check of the per-
formance of the total operating computerized GC/MS system. Thus, with
a minimum expenditure of time, an operator can be reasonably sure that
the GO column, the enrichment device, the ion source, the ion separating
device, the ion detection device, the signal amplifying circuits, the
analog to digital converter, the data reduction system, and the data
output system are all functioning properly.
An unsuccessful test requires the examination of the individual sub-
systems and correction of the faulty component(s). Environmental data
acquired after a successful system check are, in a real sense, validated
and of far more value than unvalidated data. Environmental data acquired
after an unsuccessful test may be worthless and may cause erroneous
identifications. It is recommended that the tests be applied often on
a working system, especially when there is a suspicion of a malfunction.
The procedure is written for a low resolution mass spectrometer such as
the Finnigan 3200 or the Hewlett Packard 5930A quadrupole-type mass
spectrometer, equipped with an automated data system such as the Finnigan
6000 or Hewlett-Packard 5933A system. However, the test is clearly and
readily adapted to any GC/MS system by suitable modification of the
detailed procedure. .';
There is a special need to closely monitor the performance of the quadrupole
mass spectrometer. Unlike the magnetic deflection spectrometer, the active
ion separating element of a quadrupole spectrometer (the rods) is directly
contaminated during operation, and after prolonged operation is subject to
severely degraded performance. Since degraded performance usually affects
the high mass region first, the test includes high mass end criteria.
High quality, high mass data are important since many environmentally
significant compounds have molecular and fragment ions in the 300-500 y
range.
A quadrupole mass spectrometer, which has been tuned to give a reference
compound spectrum that meets the criteria of this test, will, in general,
generate mass spectra of organic compounds that are very similar, if not
identical, to spectra generated by other types of mass spectrometers. Thus,
quadrupole mass spectra will be directly comparable to spectra of authentic
samples in collections that have developed over the years, mainly from
magnetic sector mass spectrometers.
-426-
-------
Section 10M
Assurance of mass spectral data is obtained through a set of two levels
of functionality tests. The first test requires establishment of pro-
duction, dispersion, and detection of ions from a reference compound,
perfluorotri-n-butylamine (PFTBA). Relative peak heights are adjusted
to conform to the known electron impact spectrum, with a slight biasing
toward increased transmission of ions higher than m/z = 200, which are
not commonly interfered with by tissue component fragments.
The second test of the GC-MS combination requires injection of a known
low-level standard sample while the operation is under computer control.
This is followed by periodic verification of the quality of spectra
compared to spectra of known ideal quality. Chemical compounds used
may be bis(perfluorophenyl)phenylphosphine(or decafluorotriphenyl phosphine,
DFTPP). Another set of compounds commonly used are aldrin and heptachlar
epoxide. Heptachlor epoxide is useful as a representative member of the
important chlordane series of pesticides and, more generally, because the
M-C1 ion, six-chlorine isotope cluster beginning at 351 m/z allows a test
of sensitivity and resolution at a very useful mass, not provided for by
PFTBA or many other mass calibration compounds, but quite relevant to
pesticide work. The appearance of the 351 m/z cluster may be examined
at 100, 10, and 1 ng levels, as instrument sensitivity requires, with
respect to appearance of the six-chlorine cluster versus statistical
appearance. Resolution of 13C isotope peaks and relative abundance
versus the 81 m/e peak may also be determined. Aldrin, injected as a
GC retention time test, also has its mass spectrum routinely compared
against the literature spectrum with respect to correctness of chloro
cluster statistics, sensitivity, and relative appearance of high and low
mass fragment ions. The retention time of heptachlor epoxide relative
to aldrin (1.59 + 0.02) on a 1.5% OV-17/1.95% OV-210 column at 185°C may
also be determined, along with GC column resolution. This test has the
advantage of providing a full functionality evaluation of the GC/MS system,
including sensitivity, data system acquisition, and recall of spectra.
b. Quality Assurance Procedures
(1) Using PFTBA (3M trade designation: FC-43) standard:
(a) Check on oscilloscope and/or light beam oscillograph
that 69, 131, 219, 264, 414, 502, and 614 m/z ions are present and in
reasonable relative abundance according to the following tabulation:
-427-
-------
Section 10M
General Desired Appearance of the Mass Spectrum of PFTBA
Mass
(m/z)
69
100
114
119
131
219
264
414
426
502
614
Relative
Abundance %
100.0
24.
9,
20.
70.
68.
16.
5,
2,
2.
0.3
Isotope Abundance Checks, Percentage Ratio of
Ion Signal Abundances
(70)/(69)
(220)/(219)
(503)7(502)
1.1%
4.4%
10.3%
Tune mass spectrometer as required, with respect to
resolution, optimum peak shape, sensitivity, and minimum mass falloff
(refer to appropriate instrument manual for instructions).
(c) Calibrate data system and verify the calibration by
examining a PFTBA spectrum acquired under data system control (refer to
appropriate data system manual for programs).
(2) Run aldrin and/or heptachlor epoxide and examine the re-
constructed total ion chromatogram and mass spectra.
(3) Perform DFTPP test (optional).
(4) Go on to sample runs.
c. Preparation of Aldrin and/or Heptachlor Epoxide Standards
Primary standards of aldrin and heptachlor epoxide can be obtained
from the Pesticide Repository, Health Effects Research Laboratory, EPA,
Research Triangle Park, NC.
-428-
-------
Section 10M
Carefully weigh out 20 mg of the pesticide and dissolve in 100 ml
of n-hexane (pesticide quality, or equivalent) in a volumetric flask. Keep
this stock solution under refrigeration. Replace every 6 months.
Prepare a working standard of 20 ng/vl concentration by diluting
1 ml of the stock solution to 10 ml in a volumetric flask. These working
solutions should be replaced at least monthly.
d. Preparation of Decafluorotriphenylphosphine (DFTPP) Standards
Prepare a stock solution of DFTPP at 1 mg/ml concentration in
acetone (pesticide quality, or equivalent). This stock solution has been
shown to be 97+ percent stable after 6 months, and indications are that
it will remain useable for several years.- Dilute an aliquot of the stock
solution to 10 yg/ml (10 nl/yl) concentration in acetone. The very small
quantity of material present in very dilute solutions is subject to
depreciation due to adsorption on the walls of the glass container,
reaction with trace impurities in acetone, etc. Therefore, this solution
may be useable only in the short term, perhaps 1-3 weeks.
e. Quality Assurance Test
(1) Adjust the GC column flow to normal operational level
(e.g., 30 to 45 ml/min) and set the desired oven temperature (e.g.,
185°C). The parameters should be adjusted to provide at least four
spectral scans during the elution of the aldrin, heptachlor epoxide,
or DFTPP standard.
(2) Set mass spectrometer at normal or high sensitivity as desired.
(3) Calibrate the instrument.
(4) Inject 40 ng of aldrin and/or heptachlor epoxide (or 20 ng
of DFTPP) on the GC column and note the time (or start stopwatch).
(5) After the solvent passes through the analyzer and the Vacuum
has recovered, turn on the ionizer and start scanning.
(6) Note the exact retention time of the standard as it elutes
from the column. This retention time can be used as a daily check of the
condition of the GC column and separator by comparing the values. The
retention times should not vary significantly from day to day under identical
operating conditions.
(7) Terminate the run, turn off the ion source and electron
multiplier, and reconstruct the gas chromatogram.
(8) Select a spectrum number on the front side of the GC peak as
near the apex as possible and select a background spectrum number immediately
preceding the peak.
-429-
-------
Section 10M
(9) Plot or display the mass spectrum and compare against a
reference spectrum- The spectrum obtained on the test system should
contain ion abundances within limits given for the key ions in the
following tables. Sensitivity is considered adequate if 40 ng or less
of either aldrin or heptachlor epoxide and 20 ng. or less of DFTPP pro-
vide good mass spectra.
Reference Aldrin Mass Spectrum
(5-chlorine cluster check)
Reference Heptachlor Epoxide Mass Spectrum
(6-chlorine cluster check)
m/z
261
262
263
264
265
267
269
271
Abund. (%)
61.5
4.7
100.0
7.8
65.0
21.1
3.4
0.2
(intensity of
fragment)
(intensity of
base peak)
(261)
(66)
45%
m/z
351
352
353
354
355
357
359
361
363
(intensity of
fragment)
(intensity of
base peak)
Abund. (%)
51.2
5.6
100.0
11.2
81.2
35.2
8.5
1.1
0.06
(351) m
A7ff/
(81)
Reference Mass Spectrum of DFTPP
51
68
70
127
197
198
199
275
365
441
442
443 (M+l)
444 (M+2)
s Abundance Criteria
30-60% of mass 198
Less than 2% of mass 69
Less than 2% of mass 69 (1.1% theoretical)
40-60% of mass 198
Less than 1% of mass 198
Base peak, 100% relative abundance
5-9% of .mass 198 (6.6% theoretical)
10-30% of mass 198
1% of mass 198
Less than mass 443
40-60% of mass 198 — this ion is very sensitive
to spectrum number chosen and condition of
equipment. If greater than 60%, equipment is
OK if all other criteria are met.
17-23% of mass 442 (19.8% theoretical)
1.86% (theoretical)
-430-
-------
Section 10M
f. Protocol for Analysis of Samples
(1) Sample Collection
Samples of human adipose tissue are obtained through
cooperating medical pathologists and medical examiners at hospitals in
cities selected according "to a proportionate, stratified-random design.
The conterminous 48 states were divided into 9 census divisions, according
to the 1970 census of the United States. A city within each census
division was selected from those already participating in the National
Human Monitoring Program as the collection site for special projects.
Blood sera samples are collected throughout the U.S. by
means of a .cooperative arrangement between EPA and the U.S. Public Health
Service. The PHS'program, called the Health and Nutritional Examination
Survey II (HANES II), provides blood specimens from a probability sample
of persons 12 to 74 years old, along with various medical and nutritional
parameters and some information regarding pesticide use by the indi-
viduals sampled. The blood is drawn into evacuated ampoules, allowed to
clot, and centrifuged, and the serum is decanted into a clean vial.
(2) Cleanup
Tissues are normally extracted and cleaned, up according to
a modified Mills-Onley-Gaither procedure (Subsection 9A) by laboratories
under contract to the National Human Monitoring Program. Concentrated
extracts, corresponding to the 6% and 15% ethyl ether in petroleum ether
fraction from the Florisil cleanup column, are then sent to the ACB/HERL-
RTP for GC/MS analysis. Composite samples, comprising 100-500 individual
samples, require additional cleanup before GC/MS analysis. The usual
method of choice is gel permeation chromatography (GPC) as described in
Section 91. Bloodjsera samples (Section 9D) may or may not need GPC
cleanup.
(3) Analysis
After cleanup, samples are concentrated by removal of solvent
at room temperature under a gentle stream of nitrogen. The final volume is
usually 100 yl, but it may be smaller if levels of compounds sought are
particularly low. Quantitative analysis is performed in the electron
impact (El) mode. Aliquots of 5 to 50 ul are co-injected with aldrin
(e.g., 250 ng in hexane) as an internal standard into the GC/MS system.
A total ion current profile is generated, and retention times relative
to aldrin are determined for each component of interest. Mass spectral
data are recalled from the computer for each component of interest and
analyzed against reference mass spectra obtained from various literature
references (e.g., 142, 143) or from a reference library such as the NIH-
EPA Chemical Information System (181), or, most preferably, generated
from authentic laboratory standards. Relative retention times are also
compared to those of the reference material for further confirmation.
-431-
-------
Sections ION, 100
After identification, quantitative analyses are usually performed by se-
lected ion monitoring (SIM). An authentic reference sample is used for
direct comparison. Identification may be further confirmed by chemical
ionization GC/MS, where available.
g. GC/MS Systems
Manufacturer's operating manuals should be consulted for descrip-
tions and detailed operating instructions for specific GC/MS systems. The.
previous edition of this Manual contained information on two GC-MS systems:
the Hewlett Packard 5930A quadrupole MS, 5700A gas chromatograph, and
5933A data system; and the Finnigan 3200 quadrupole MS and 9500 gas
chromatograph.
Another EPA Manual (182) contains specific information on the
Finnigan 1015 and 3000 quadrupole GC/MS systems coupled with a PDP-8
data system. This Manual includes 10 chapters covering the following
material: (1) introduction to broad spectrum organic analysis, routine
monitoring of large numbers of target compounds, and real time selected
ion monitoring; (2) detailed start-up and calibration procedures; (3)
preparation methods for water samples; (4) information on OUTPUT programs
for data analysis; (5) compound identification using PDP-8 software; (6)
specialized techniques such as single ion monitoring, open tubular
columns, chemical ionization, accurate mass measurement, standard
additions, and sample spikes; (7) miscellaneous auxiliary software programs
and housekeeping routines; (8) preventive maintenance; (9) trouble
shooting; and (10) selected bibliography up to 1978, mostly to information
from EPA laboratories.
ION BIOLOGICAL METHODS
Bioassay techniques, which include insecticidal activity, enzymatic, and
immunological methods, have been described as providing an independent
criterion of identity when combined with GC, chemical reactions, etc.
(2). These methods, which depend on the measurement of a physiological
response of a test organism induced by exposure to the pesticide, have
advantages of simplicity and sensitivity but are relatively non-specific
so that their utility for confirmation is rather poor. The insect
bioassay technique has been reviewed (183).
Specificity of enzyme inhibition is greatly enhanced by combination with
TLC for detection and confirmation of organophosphate and certain carba-
mate pesticides. The Rj? value plus biological response provide important
identity information at levels typically in the range of 500 pg to 10 ng
for these compounds.
100 POLAROGRAPHY (VOLTAMMETRY)
Voltammetry is the generic name for a group of electroanalytical methods
in which current-vs-voltage curves are recorded when a gradually changing
-432-
-------
Section 100
voltage is applied to a cell containing the solution to be analyzed, a
stable reference electrode, and a small-area working or indicator
electrode. In the special case where.the indicator electrode is a
dropping mercury electrode, the technique is called polarography. In
addition to classical DC polarography, in which the current is measured
for each drop as voltage is increased linearly with time, modern
variations include DC current sampled polarography, pulse polarography,
differential pulse polarography, linear sweep (rapid scan) polarography •,
and AC polarography. These newer methods differ in the type of voltage
signal applied and/or the manner in which the current is measured, and
they are generally more sensitive and/or selective than traditional DC
polarography.
The use of polarography as a confirmatory test is described in Section
12,F of the EPA PAM and Sections 640 and 641 of the FDA PAM: Procedures
and applications of polarography for both identification and determina-
tion of pesticide residues have been reviewed (184-186).
Polarographic identification of a pesticide residue is based on the
determination of the peak potential of the unknown in a cleaned-up
extract, and comparison with the potential of about, the same amount
of a reference standard under identical conditions. As a check,
addition of the standard compound to the unknown should result in an
increase in the wave height but not appearance of another wave. Mixtures
can be identified if the peak potentials of the components are sufficiently
separated. Trapped GC fractions may be subjected to polarography to
confirm identifications based on retention times. Instrumentation for
such modern voltammetric techniques as fast sweep oscillography provides
sensitivity comparable to colorimetry. Pesticides not containing an
oxidizable or reducible functional group can be made amenable to polarography
by formation of a suitable derivative (e.g., nitro, halogen, carbonyl, etc).
Most polarographic studies have been applied to phosphorus-containing
insecticides such as parathion, diazinon, malathion, and carbophenothion
(187). A collaborative study confirmed the usefulness of single sweep
oscillographic polarography for identifying such residues in non-fatty
foods (188). Nitrophenol metabolites of OP pesticides were determined
in urine by polarography (189). Thirty-eight herbicides have been
studied by single sweep derivative polarography (190), methylcarbamate
insecticides by AC polarography and cyclic voltammetry (191), and urea
herbicides by anodic polarography (192). Published voltammetric reduction
potentials for about 100 organochlorine insecticides, PCBs, and naphthalenes
(3 electrode potentiostat, DMSO solvent) are a useful aid in identification
of residues (193). Parathion and related insecticides and metabolites
were polarographically determined in blood without extraction (194),. The
voltammetry of 1,3,5-triazines (195), propachlor herbicide (in soil) (196),
dithiocarbamates (197, 198), dinitroaniline herbicides (199), thiourea-
containing pesticides (200), trifluralin (in soils) (201), azomethine-
containing pesticides (e.g., Cytrolane, Cyolarie, chlordimeform) (202),
PCP (203), and phosmet (in apples) (204) has been reported. Paraquat
can be directly determined in urine and serum by differential pulse
polarography at ca 0.04 yg/ml levels (205).
-433-
-------
Sections 10P, 10Q
10P MISCELLANEOUS CONFIRMATORY METHODS
a. Carbon Skeleton Chromatography ' . ,
Carbon skeleton chromatography (CSC) is useful in characterizing
insecticide residues in amounts down to 5-100 ng. Apparatus for CSC
consists of a precolumn containing a hot (ca 300°C) catalyst attached
to a gas chromatograph equipped with a flame ionization detector
(available from National Instruments Laboratory, Rockville, MD). the
compound to be identified is injected directly on the catalyst .bed
(e.g., 1% Pd on 60-80 mesh Gas-Chrom P) and is swept over the bed by
hydrogen.carrier gas.- Nitrogen is introduced through the normal instru-
ment inlet so that the detector yields optimum response. While in the
precolumn, all functional groups are stripped from the compound, and
any multiple bonds are saturated. The resulting hydrocarbons are
carried into the chromatographic column where they are separated and
identified by their retention characteristics relative to standards.
This identification method, which is in effect a deriyatization pro- .
cedure, has been applied .to heptachlor, heptachlor epoxide, chlordane,
aldrin, endrin, DDT and its analogs, and carbaryl. Sufficient residue
must be available for the method to be of value. Techniques, applica-
tions to many pesticide classes, and characterization of products of
CSC (as well as some other precolumn reaction confirmatory methods)
have been reported by Beroza and co-workers (206-209) and Asai et al.
(210, 211). Identification of 5-10 ng amounts of polychlorinated
biphenyls, terphenyls, naphthalenes, dioxins, and dibenzofurans in
biological samples has been demonstrated (212), and mixtures of
polychlorinated naphthalenes, PCBs, PCTs, and OC1 pesticides have been
analyzed (213).
b. Fragmentation Procedures
GC fragmentation procedures are similar to CSC except that the
reaction in the precolumn decomposes the pesticides, yielding character-
istic fragment peak patterns or fingerprint chromatograms helpful in
making identifications. A palladium catalyst at 300°C (210) and reagents
such as Na2C03, CuO, CdCl2» A1C13, and I^C^Oy at 240°C (214) have been
applied to chlorinated and OP insecticides with EC detection of the
reaction products. . .
Gas chromatograms of 33 organochlorine pesticides after ultra-
violet irradiation have been published. These characteristic photo-
decomposition patterns are also useful for conclusive residue confirma-
tion (215).
10Q REFERENCES ,
(1) Analytical Methods for Pesticide Residues in Foods, Department of
National Health and Welfare, Canada, 1973, Section 10.2.
-434-
-------
Section 10Q
.(2) Robinson, J., Richardson, A., and Elgar, K. E., Chemical Identity
in Microanalysis, presented at the ACS National Meeting, New York
City, September 11-16, 1966; Robinson, J., Chem. Br., 7., 472 (1971).
(3) Elgar, K. E., The Identification of Pesticides at Residue Concentra-
tions, Advances in Chemistry Series 104, Chapter 10, ACS, Washington,
D.C., 1971, page 151.
(4) Onuska, F. I., and Comba, Mi E., J. Chromatogr., 119. 385 (1976).
(5) Ruzicka, J. H. A., and Abbott, D. C., Talanta, 20, 1277 (1973).
(6) Bailey, R., Health and Welfare Canada, Health Protection Branch,
personal communication (1980).
(7) Aue, W. A., and Kapila, S., Anal. Chem.. 50., 536 (1978); Kapila, S.,
and Aue, W. A., J. Chromatogr.. 148(2). 343 (1978).
(8) Mallet, V., J. Chromatogr.. 2i> 217 (1973). .
(9) Lawrence, J. F., and Frei, R, W., Chromatogr. Rev.. 18, 253 (1974).
(10) Connors, K. A., Anal. Chem.. 46_, 53 (1974). .
(11) Dale, T., and Court, W. E.. Chromatographia. .13(1), 124 (1980).
(12) Heinz, D. E., and Vitek, R. K., J. Chromatogr. Sci.. 135 570 (1975).
(13) Dolan, J. W., and Seiber, J. N., Anal. Chem.. 49. 326 (1977).
(14) Beroza, M., and Bowman, M. C., J. Assoc. Off. Anal. Chem., 48, 358 (1965).
(15) Bowman, M. C., and Beroza, M... J. Assoc. Off. Anal. Chem.. 48, 943 (1965).
•0 '.''..
(16) Beroza, M., and Bowman, M. C., Anal. Chem., 37, 291 (1965).
(17) Beroza, M., and Bowman, M. C.s Anal. Chem.. .38, 837 (1966).
(18) Bowman, M. C., and Beroza, M.. Anal. Chem.. 38, 1427 (1966).
(19) Crist, H. L., Harless, R. L., Moseman, R. F., and Callis, M. H.,
Bull. Environ. Contam. Toxicol.. 24(1), 231 (1980).
(20) Cochrane, W. P., J. Chromatogr. Sci.. 17(3), 124 (1979).
(21) Cochrane, W. P., and Chau, A. S. Y., Chemical Derivatization Techniques
for Confirmation of Organochlorine Residue Identity, Advances in
Chemistry Series 104, Chapter 2, ACS, Washington, D.C., 1971, page 11.
(22) Cochrane, W. P., J. Chromatogr. Sci., 13, 246 (1975).
-435-
-------
Section 10Q
(23)
(24)
(25)
(26)
(27)
(28)
(29)
(30)
(31)
(32)
(33)
(34)
(35)
(36)
(37)
(38)
(39)
(40)
(41)
Holdrinet, M., Bull. Environ. Contain. Toxicol.. J21(l-2), 46 (1979).
Lusby, W. R., and Hill, K. R., Bull. Environ. Contam. Toxicol..
.22(4-5) , 576 (1979). "*"~ —
Rosewell, K. T., and Baker, B. E., Bull. Environ. Contam. Toxicol.,
21(4-5), 470 (1979).
De Beer, J. 0., Van Peteghem, C. H., and Heyndrickx, A. M., J. Assoc.
Off. Anal. Chem.. 61, 1140 (1978). ~~
Glotfelty, D. E., Anal. Chem.. 44, 1250 (1972).
Banks, K. A., and Bills, D. D., J. Chromatogr.. 33, 450 (1968).
Cochrane, W. P., and Maybury, R. B., J. Assoc. Off. Anal. Chem.,
5£, 1324 (1973). . """~ —-
Watts, R. W., Hodgson, D. W., Crist, H. L., and Moseman, R. F.,
J. Assoc. Off. Anal. Chem.. .63(5), 1128 (1980).
Shafik, T. M., Bradway, D., and Enos, H. F., Bull. Environ. Contam.
Toxicol.. 6., 55 (1971).
McCully, K. A., J. Assoc. Off. Anal. Chem.. J50, 376 (1977).
Singh, J., and Lapointe, M* R., J. Assoc. Off. Anal. Chem., 57.
1285 (1974). —
Coburn, J. A., and Chau, A. S. Y., J. Assoc. Off. Anal. Chem.. 57,
1272 (1974); Environ. Lett.. 10, 225 (1975).—
Forbes, M. A., Wilson, B. P., Greenhaigh, R., and Cochrane, W. P.,"
Bull. Environ. Contam. Toxicol.. 13. 141 (1975).
McCully, K. A., J. Assoc. Off. Anal. Chem.. .62(2), 385 (1979).
Cochrane, W. P., and Greenhaigh, R., Confirmation of Pesticide
Residue Identity by Chemical Derivatization, presented at 166th
National ACS Meeting, Chicago, August 29, 1973: Khan, S. U.,
Greenhaigh, R., and Cochrane, W. P., J. Agr. Food Chem.. 23, 430 (1975).
Lawrence, J. F., J. Agric. Food Chem.. 22, 936 (1974).
Lawrence, J. G., Lewis, D., and McLeod, H. A*, J. Agric. Food Chem.,
.25, 1359 (1977). ~ •
Lawrence, J. F., J. Aerie. Food Chem.. 2±, 1236 (1976).
Lawrence, J. F., Lewis, D. A., and McLeod, H. A., J. Chromatogr.
138, 143 (1977). &—
-436-
-------
Section 10Q
(42> Miles, J. W. „ and Dale, W. E., J. Aerie. Food Chen., 26i, 480 (1978).
(43) Greenhalgh, R., King, R. R., and Marshall, W. D., J. Agric. Food
Chem.. 26, 475 (1978).
(44) Magallona, E. D., and Gunther, F. A., Arch. Environ. Contam. Toxicol..
5> 185 (1977).
(45) Lawrence, J. F., J. Chromatogr.. 123, 287 (1976).
(46) Worobey, B. L., and Webster, G. R. B., J. Assoc. Off. Anal. Chem..
60, 213 (1977).
(47) Noshe, N., Kobayoshi, S., Tanaka, A., Hirose, A., and Watanabe, A.,
J. Chromatogr.. 130. 410 (1977).
(48) Wein, R. G". , and Tanaka, F. S., J. Chromatogr.. 130. 55 (1977).
(49) Bromilow, R. H., and Lord, K. A., J. Chromatogr.. 125. 495 (1976).
(50) Lawrence, J. F., J. Assoc. Off. Anal. Chem.. 59. 1061 (1976).
(51) Lawrence, J. F., and Sundaram, K. M. S., J. Assoc. Off. Anal. Chem..
59, 938 (1976).
(52) Lawrence, J. F., and McLeod, H. A., J. Assoc. Off. Anal. Chem.. 59.,
637, (1976).
(53) Singh, J., and Cochrane, W. P., J. Assoc. Off. Anal. Chem.. 62_(4),
751 (1979).
(54) Stoks, P. G., and Schwartz, A. W., J. Chromatogr.. 168(2), 455 (1979).
(55) Krause, R. J., J. Chromatogr.. 185. 615 (1979).
(56) Aly, 0. M., Faust, S. D., and Suffet, I. H., Ultraviolet Spectro-
photometry in Residue Analysis; Spectra-Structure Correlations,
Advances in Chemistry Series 104, Chapter 7, ACS,. Washington, D.C.,
1971, page 95.
(57) Gore, R. C., Hannah, R. W., Pattacini, S. C., and Porro, T. J.,
J. Assoc. Off. Anal. Chem.. 54, 1040 (1971).
(58) Cyr, T., Cyr, N., and Haque, R., Spectrophotometric Methods in
Analytical Methods for Pesticides and Plant Growth Regulators,
Zweig, G., and Sherma, J., eds., Volume IX, Chapter 3, Academic
Press, New York, 1977, page 75.
(59) MacDougall, D., Residue Rev.. JL, 24 (1962); 5., 119 (1964).
-437-
-------
Section 10Q
(60)
(61)
(62)
(63)
(64)
(65)
(66)
(67)
(68)
(69)
(70)
(71)
(72)
(73)
(74)
(75)
(76)
Argauer, R. J., Fluorescence Methods for Pesticides, in Analytical
Methods for Pesticides and Plant Growth Regulators. Zweig, G.,
and Sherma, J., eds., Volume IX, Chapter 4, Academic Press, 1977,
New York, page 101.
Baeyens, W., Pharm. Weekbl.. Ill, 1075 (1976).
Sloane, H. J., and Obremski, R. J. t Appl. Spectrosc.. 31, 506 (1977).
Blinn, R. C., The Identification of Pesticides at Residue Concentra-
tions, Advances in Chemistry Series 104. Chapter 6, ACS, Washington,
B.C., 1971, page 81.
Kettmann, R., Closset, J.-L., Copin, A'., Duculot, C., and Martens,
P.-H., Anal. Lett., £, 1013 (1973).
Cyr, T., Cyr, N., .and Haque, R., Infrared Spectrometry, in
Analytical Methods for Pesticides and Plant Growth Regulators.
Zweig, G., and Sherma, J., eds., Volume IX, Chapter 5, Academic
Press, New York, 1977, page 137.
Crosby, N. T., and Laws, E. Q., Analyst. 89, 319 (1964).
Neiswender, D. D., J. Gas Chromatogr., ,4, 426 (1966).
Shearer, D. A., Stone, B. C., and McGugan, W. A., Analyst, 88,
147 (1963). - *— —
Bierl, B. A., Beroza, M., and Ruth, J. M., J. Gas Chromatogr., 6,
286 (1968). •—
Scanlon, R. A., Arnold, R. G., and Lindsay, R. C., J. Gas Chromatogr.
6., 372 (1968). : —
Nano, G. M., Sancin, P., and Martelli, A., J. Gas Chromatogr.. 3,
85 (1965). —
Hannah, R. W., and Dwyer, J. L., Anal. Chem., 36. 2341 (1964).
Hammerstrand, K., Juntunen, J. M., and Henes, A. R., Anal. Biochem.
.27, 172 (1969).
Curry, A. S., Read, J. F., Brown, C., and Jenkins, R. W., J. Chromatogr.,
.38, 200 (1968). " 5—
Hartman, K. T., J. Assoc. Off. Anal. Chem., 50, 615 (1967).
Giang, P. A., Library of Infrared Spectra of Important Pesticides,
ia Analytical Methods for Pesticides and Plant Growth Regulators.
Zweig, G., and Sherma, J., eds., Volume IX, Chapter 6, Academic
Press, New York, 1977, page 153.
-438-
-------
Section 10Q
(77) Nicholas, M. L., J. Assoc. Off. Anal. Chem.. 59, 197, 1071, 1266
(1976).
(78) Giang, P. A., Agric. Res. Serv. NE-91. 95 pp (1978).
(79) Cournoyer, R., Shearer, J. C., and Anderson, D. H., Anal. Chem..
.49, 2275 (1977).
(80) Cyr, N., Cyr, T., and Haque, R., Nuclear Magnetic Resonance
Spectroscopyr in Analytical Methods for Pesticides and Plant
Growth Regulators. Zweig, G., and Sherma, J., eds., Volume IX,
Chapter 2, Academic Press, New York, 1977, page 51.
(81) Wilson, D. M., Olsen, R. W., and Burlingame, A. L., Rev. Sci.
Instrum.. 45. 1095 (1974).
(82) Biros, F. J., J. Assoc. Off. Anal. Chem.. 53, 733 (1970).
(83) Wilson, N. K., J. Am. Chem. Soc., 94, 2431 (1972).
(84) Ross, R* T., and Biros, J. F., Bjochem. Biophys. Res. Commun., 39,
723 (1970).
(85) Ross, R. T., and Biros, F. J., Mass Spectrometry and NMR Spectroscopy
in Pesticide Chemistry, Plenum Press, New York, N.Y., 1974, pages
263-272.
(86) McKinney, J. D., Wilson, N. K., Keith, L. H., and Alford, A. L.,
Mass Spectrometry and NMR Spectroscopy in Pesticide Chemistry.
Plenum Press, New York, .N.Y., 1974, pages 139-160.
(87) Crombie, L., and Lown, J. W., J. Chem. Soc.. 1962» 775.
(88) Wilson, N. K., J. Phvs. Chem.. 75, 1067 '(1971).
(89) Keith, L. H., Garrison, A. W., and Alford, A. L., J. Assoc. Off.
Anal. Chem.. 51, 1053 (1968).
C90)r Keith, L. H., Alford, A. L., and Garrison, A. W., J. Assoc. Off.
"" Anal. Chem., 52., 1074 (1969).
(91) Keith, L. H., and Alford, A. L., J. Assoc. Off. Anal. Chem.. 53,
157 (1970).
(92) Wilson, N. K., Nuclear Magnetic Resonance Spectroscopy in Pesticide
Chemistry, in Pesticide Analysis, Das, K, G., ed., Marcel Dekker,
Inc., New York, 1980.
(93) Keith, L, H., and Alford, A. L., J. Assoc. Off. Anal. Chem..53.
1018 (1970).
-439-
-------
Section 10Q
(94) Yamamoto, 0., Yanagisawa, M., Hayamizu, K., and Kotowycz, G.,
J. Magn. Reson.. j>, 216 (1973).
(95) Wilson, N. K., and Anderson, M., Mass Spectrometry and NMR
Spectroscopy in Pesticide Chemistry. Plenum Press, New York, N.Y.,
1974, pp. 197-218.
(96) Wilson, N. K., J. Am. Chem. Soc.. 97, 3573 (1975).
(97) Cox, R. H., and McKinney, J. D., Ore. Magn. Reson.. 11(11), 54i
(1978).
(98) Wilson, N. K., and Sovocool, G. W., Org. Magn. Reson.. 9., 536 (1977).
(99) Alley, E. G., Layton, B. R., and Minyard, J. P., Jr.; J. Org. Chem..
41, 462 (1976).
(100) Wilson, N. K., J. Phys. Chem., 83_, 2649 (1979).
(101) Wilson, N. K., and Zehr, R. D., J. Org. Chem.. 44. 1278 (1979).
(102) Roll, D. B., and Biros, F. J., Anal. Chem.. 41, 407 (1969).
(103) Gegiou, D., Anal. Chem.. 4£, 742 (1974).
(104) Gegiou, D., Talanta. 21, 889 (1974).
(105) Ross, R. T., and Biros, J. F., Anal. Chem.. 52, 139 (1970).
(106) Gurley, T. W., and Ritchey, W. M., Anal. Chem.. 48. 1137 (1976).
(107) Benyon, J. H., Recommendations for Symbolism and Nomenclature for
Mass Spectroscopy, 26th Annual Conference on Mass Spectrometry
and Allied Topics, IUPAC, May 28-Jutte 2, 1978, St. Louis, MO.
(108) Safe, S., Environmental Applications of Mass Spectrometry, in
Mass Spectrometry. Vol. 5, Johnstone, R. A. W., Sr. Reporter, (
The Chemical Society, London, 1979, Chapter 9, p. 234.
(109) Skinner, S. I. M., and Greenhalgh, R., Mass Spectra of Insecticides.
Herbicides and Fungicides and Metabolites. Chemistry and Biology
Research Institute, Agriculture Canada, Ottawa, Ontario, Canada, 1977.
(110) Ryan, J. F., Residue Analysis Applications of Mass Spectrometry,
in Analytical Methods for Pesticides and Plant Growth Regulators.
Zweig, G., and Sherma, J., eds., Volume IX, Chapter 1, Academic "-'-••'
Press, New York, 1977, p. 1.
-440-
-------
Section 10Q
(111)
(112)
(113)
(113A)
(114)
(115)
(116)
(117)
(118)
(119)
(120)
(121)
(122)
(123)
(124)
(125)
(126)
Burlingame, A. L. , Baillie, T. A., Derrick, P. J., and Chizov, 0. S.,
Anal. Chem.. 5.2(5), 214R (1980); (see also earlier reviews of mass
spectrometry in The Fundamental Reviews issues of Anal. Chem.,
issue number 5 each even year).
Alford, A., Biomed. Mass Spectrom..4. 1 (1977); 5_, 259 (1978).
Biros, F. J., Residue Rev.. 4jO, 1 (1971).
Kuehl, D. W., Whitaker, M. J., and Dougherty, R. C., Anal. Chem..
52., 935 (1980).
Grayson, M. A., J. Chromatogr. Sci.. 15. 539 (1977).
McFadden, W. H., Techniques of Combined Gas Chromatography/Mass
Spectrometry; Applications in Organic Analysis, John Wiley & Son,
New York, 1973.
McFadden, W. H., J. Chromatogr. Sci.. 17, 2 (1979).
Wright, L. H., and Oswald, E. 0., Liquid Chromatography/Mass
Spectrometric Methods of Analyses for Garbamate Pesticides,
presented at the 26th Annual Conference on Mass Spectrometry and
Allied Topics, St. Louis, MO, May-June', 1978.
Fales, H. M., Milne, G. W. A., Winkler, H. U., Beckey, H. D.,
Damico, J. N., and Barren, R., Anal. Chem., 47, 207 (1975).
Hunt, D. F., Stafford, G. C., Crow, F., and Russell, J. W.,
24th Annual Conference on Mass Spectrometry and Allied Topics,
Meeting Abstract No. AA5, 1976.
Stan,.H.-J., Z. Anal. Chem.. 287. 104 (1977).
Holmstead, R. L., and Casida, J. E., J. Assoc. Off. Anal. Chem..
57, 1050 (1974).
Damico, J. N., Barren, R. P., and Sphon, J. A., Int. J. Mass Spectrom.
Ion Phys., 2± 161 (1969).
Heck, H. , Dyer, R. L., Scott, A. C., and Anbar, M., J. Agric. Food
Chem.. 25, 901 (1977).
Schulten, H. R., and Beckey, H. D., J. Agric.'Food Chem.. 21, 372
(1973); 24, 743 (1976).
Schulten, H. R., 2. Anal. Chem.. 293(4), 273 (1978).
Yamato, Y., Suzuki, M., and Watanabe, T., Biomed. Mass Spectrom..
£(5), 205 (1979).
-441-
-------
Section 10Q
(127) Carroll, D. I., Dzidic, I., Stillwell, R. N., Horning, M. G., and
Horning, E. C., Anal. Chem.. 46. 706 (1974).
(128) Carroll, D. I., Dzidic, I., Stillwell, R. N., Haegele, K. D.,
and Horning, E. C., Anal. Chem.. £7, 2369 (1975).
(129) Horning, E. C., Carroll, D. I., Dzidic, I., Lin,, S.-N., Stillwell,
R. N., and Thenot, J.-P., J. Chromatogr., 142. 481 (1977).
(130) Hunt, D. F., Stafford, G. C., Crow, F. W., and Russell, J. W.,
Anal. Chem.. 4£, 2098 (1976).
(131) Haas, J. R., Friesen, M. D., Harvan, D. J., and Parker, C. E.,
Anal. Chem., 50. 1474 (1978).
•
(132) Biros, F. J., Dougherty, R. C., and Dalton, J., Qrg. Mass. Spectrom..
.6, 1161 and 1171 (1972). *~~~^ "
(133) Dougherty, R. .C.,- Roberts, .J. D., and Biros, F. J., Anal. Chem..
47, 54 (1975). ' ~ '
(134) Dougherty, R. C., and Piotrowska, K. K., J. Assoc. Off. Anal. Chem..
59_, 1023 (1976). —____^ .
(134A) Hass, J. R., and Friesen, M. D., Ore. Mass Spectrom.. 14
9 (1979). —* *— —'
(134B) Hunt, D. F., Harvey, T. M., and Russell, S. W., J. Chem. Soc. Chem.
Commun., 151 (1975). "^
(135) Milne, G. W. A., and Lacey, M. J., CRC Grit. Rev. Anal. Chem.. 4,
45-104 (1974). ~' ' —: ~
(136) Milberg, R. M., and Cook, J. C., Jr., J. Chromatogr. Sci., 17,
17 (1979). —
(137) Busch, K. L., Bursey, M. M., Haas, J. R., and Sovocool, G. W.,
Appl. Spectrosc.. 3^(4), 388 (1978).
(138) Abramson, F. P., Anal. Chem.. <44_(14), 28A (1972).
(139) Oswald, E. 0., Albro, P. W., and McKinney, J. D.. J. Chromatogr..
98, 363 (1974). £~
(140) Vander Veld, G., and Ryan, J. F., J. Chromatogr. Sci.. 13_, 322 (1975),
(141) Dougherty, R. C., in Biochemical Applications of Mass Spectrometrv.
First Supplementary Volume. Waller, G. R., and Dermer, 0. C., eds.,
Wiley-Interscience, New York, 1980, Chapter 32A.
-442-
-------
Section 10Q
(142) Sovocool, G. W., and Lewis, R. G., The Identification of Trace.
Levels of Organic Pollutants in Human Tissues: Compounds Related
to Chlordane/Heptachlor Exposure, in Trace Substances in Environ-
mental Health, IX, 1975, Hemphill, D. D., ed., University of
Missouri Press, 1976, pp. 265-280.
(143) Wright, L. H., Lewis, R. G., Grist, H. L., Sovocool, G. W.., and
Simpson, J. M., The Identification of Polychlorinated Terphenyls
at Trace Levels in Human Adipose Tissue by Gas Chromatography/Mass
Spectrometry, J. Anal. Toxicol., 2., 76-79 (1978).
(144) Weil, L., Frimmel, F., and Quentin, K.-E., Z. Anal. Chem.. 268,
97 (1974).
(145) Rosen, J. D., and Pareles, S. R., Mass Spectrometry and NMR Spectro-
scopy in Pesticide Chemistry. Plenum Press, New York, N.Y., 1974,
pp. 91-98.
(146) Chapman, R. A., and Robinson, J. R., J. Chromatogr.. 140, 209 (1977).
(147) Bergstedt, L., and Widmark, G., Chromatographia. 3_, 59 (1970).
(148) Fenselau, C. Anal. Chem.. j49_(6), 563A (1977).
(149) Budde, W. L., and Eichelberger, J. W., J. Chromatogr.. 134. 147 (1977).
(150) Kuehl, D. W., Anal. Chem.. 49. 521 (1977).
(151) .McLafferty, F. W., and Veiikataraghavan, R., J. Chromatogr. Sci.. 17,
24 (1979). '
(152) Damico, J. N., Barren, R. P., and Ruth, J. M., Organic Mass Spectro-
metry, JL, 331 (1968); Damico, J. N., J. Assoc. Off. Anal. Chem.. 49,
1027 (1966); Benson, W. R., and Damico, J. N., J. Assoc. Off. Anal.
Chem., 48, 344 (1965); 51, 347 (1968).
(153) Mestres, R., Chevallier, C., Espinoza, C., and Corner, R., Ann. Falsif.
Expert. Chim.. .70(751), 177 (1977).
(154) Zeman, A., and Woerle, R., Org. Mass Spectrom.. 13(1), 43 (1978).
(155) Leclercq, P. A., and Pacakova, V., J. Chromatogr.. 1.78(1), 193 (1979).
(156) Stan, H.-J., Abraham, B., Jung, J., Kellert, M., and Steinland, K.,
Z. Anal. Chem.. 287. 271 (1977).
(157) Biros, F. J., and Sullivan, H., cited in (113).
(158) Biros, F. J., Applications of Combined GC/MS to Pesticide Residue
Identification, presented at the ACS-Chem. Inst. Canada International
Meeting, Toronto, May, 1970.
-443-
-------
Section 10Q
(159) Van Peteghem, C. H., and Hendrickx, A. M., J. Agric. Food Chem..
24, 635 (1976). ~——*-
(160) Shafik, T. M., Biros, F. J., and Enos, H. F., J. Aerie. Food Chem..
18_, 1174 (1970). ' . .
(161) Biros, F. J., and Ross, R. T., Fragmentation Processes in the Mass
Spectra of Trialkylphosphates, Phosphorothionates, Phosphorothiolates,
and Phosphorodithiolates, Presented at 18th Conference on Mass
Spectrometry and Allied Topics, San Francisco, CA, June, 1970.
(162) Stan, H.-J., Z. Lebensm. Unters. Forsch.. 164. 153 (1977).
(163) Biros, F. J., and Walker, A. C., J. Agric. Food.Chem.. 18, 425 (1970).
(164) Bellman, S. W., and Barry, T. L., J. Assoc. Off. Anal. Chem.. 54,
499 (1971). 'i ~" ~" ~~ " ~~
(165) Suzuki, M., Yamato, Y., and Koga, M., Biomed. Mass Spectrom..
5.(9), 518 (1978). '."'-.
(166) Harless, R. L., and Oswald, E. 0., Gas Chromatography/Mass Spectro-
metric Methods of Analysis for Toxaphene and Dioxins in Human and
Biological Samples, presented at the 26th Annual Conference on
Mass Spectrometry & Allied Topics, St. Louis, MO, May-June, 1978.
(167) Harless, R. L., Harris, D. E., Sovocool, G. W., Zehr, R. D.,
Wilson, N. K., and Oswald, E. 0., Biomed. Mass Spectrom.. 5(3),
232 (1978).
(168) Sovocool, G. W., Lewis, R. G., Harless, R. L., Wilson,. N. K., and
Zehr, R. D., Anal. Chem.. 49, 734 (1977).
(169) Van den Heuvel, W. J. A., Wood, J. S., DiGiovanni, M., and Walker,
R. W., J. Agric. Food Chem.. 25. 386 (1977).
(170) Lee, Y. W., and Westcott, N. D., J. Assoc. Off. Anal. Chem.. 62(4).
782 (1979). ~~~~~ . ' ' ' |~~*'
(171) Laseter, J. L., DeLeon, I. R., and Remele, P. C., Anal. Chem..
50_(8), 1169 (1978). "~
(172) Bagley, E. G., Reichel, W. L. , and Cromartie, E., J. Assoc. Off.
Anal. Chem.. 5.3, 251 (1970)'. '
(173) Bonelli, E. J., Anal. Chem.. 44, 603 (1972). ,
(174) Canada, D. C., and Regnier, F. E., J. Chromatogr. Sci.. 14, 149 (1976).
-444-
-------
Section 10Q
(175) Neher, M. B., and Hoyland, J. R., Specific Ion Mass Spectrometric
Detector for Gas Chromatographic Pesticide Analysis, U.S. Environ-
mental Protection Agency, Washington, DC, Report No. EPA-660/2- .
74-004, January, 1974.
(176) Thruston, A. D., Jr., A Quantitative Method for Toxaphene by
GC/CIMS Specific Ion Monitoring, U.S. Environmental Protection
Agency, Washington, DC,. Report No. EPA-600/4-76-010, March, 1976.
(177) Dyer, R. L., Heck, H. d'A., Scott, A. C., and Anbar, M., Feasibility
of Applying Field lonization Mass Spectrometry to Pesticide Research,
U.S. Environmental Protection Agency, Washington, DC, Report No.
EPA-600/1-76-037, November/ 1976.
(178) Ryan; J. F., Harless, R.L., and Lewis, R. G., Application of Field
lonization Mass Spectrometry to Environmental Analysis, Proceedings
of, the 23rd Annual Conference on Mass Spectrometry and Allied
Topics, Houston, TX, May 25-30, 1975, pp. 46-48.
(179) Horning, E. C., Carroll, C. I., Dzidic, I., Stillwell, R., and
Thenot, J.-P,, J. Assoc. Off. Anal. Chem., 61, 1232 (1978).
(179A)'Hunt, D. F., Shabanowitz, J., and Giordani, A. B., Anal. Chem. .52.
386-390 (1980).
(179B) Yost, R. A., and Enke, C. G., Anal. Chem., 51(12), 1251A (1979).
(179C) Cooks, R. G., Amer. Lab., p. Ill, October (1978).
(180) Lewis, R. G., Research Report of the Analytical Chemistry Branch,
ETD, HERL, RTP Program Element No. 1EAG15, Sovocool, G. W., and
Wright, L. H., principal investigators.
(181) Heller, S. R., and Milne, G. W. A., Environ. Sci. Technol., 13(7),
798-803 (1979).
(182) Budde, W. L., and Eichelberger, J. W., Organic Analysis Using Gas
Chromatography/Mass Spectrometry, EPA 600/8-79-006, Environmental
Monitoring and Support Laboratory, Office of Research and Development,
U.S. EPA, Cincinnati, Ohio, March, 1979. Ann Arbor Science
Publishers (Wiley), Ann Arbor, MI, 241 pp (1979).
(183) Sun, Y. P., Analytical Methods for Pesticides and Plant Growth
Regulators, Zweig, G., ed., Vol. 1, Academic Press, New York, N.Y.,
1963, page 571.
(184) Allen, P. T., Analytical Methods for Pesticides and Plant Growth
Regulators, Zweig, G,, ed. Vol. V, Chapter 3, Academic Press,
• New York, 1967, page 67.
-445-
-------
Section 10Q
(185) Gajan, R. X., Residue Rev. 5, 80 (1964); £, 75 (1964).
(186) Smyth, M. R.., and Smyth, W. F., Analyst. 103(1227), 529 (1978).
(187) Smyth, M. R., and Osteryoung, J. G., Anal. Chim. Acta. 96, 335 (1978).
(188) Gajan, R. T., J. Assoc. Off. Anal. Chem.. 5_2, 811 (1969).
(189) Zietek, M., Mikrochim.'' Acta. j21(l-2) , 75 (1979).
(190) Hance, R. J., Pestic. Sci.. ^, 112 (1970).
(191) Booth, M. D., and Fleet, B., Talanta. 17, 491 (1970).
(192) Kutlukova, V. S., Toropov, A. P., and Lozovatskaya, M. A., Anal. Abstr..
.27, Abstract No. 2940 (1974).
(193) Farwell, S. 0., Beland, F. A., and Geer, R. D., Bull. Environ. Contam.
Toxicol.. 10, 157 (1973). " '. '
(194) Zietefc, M., Mikrochim. Acta. II (5-6), 549 (1976).
(195) Marchidan, S., Rev. Roum. Chim., 22^ 127 (1977).
(196) Filimonova, M. M., Zh. Anal. Khim., 32_, 140 (1977).
(197) Budnlkov, G. K., Zh. Anal. Khim., 32_, 212 (1977); 3£ 2275 (1975).
(198) Nangniot, P., Zenon-Roland, L., and Berlemont-Frennet, M.,
Analusis, 6(6), 273 (1978).
(199) Southwick, L. M., Willis, G. H., Dasgupta, P. K., and Kesztheli,
C. P., Anal. Chim. Acta, 82, 29 (1976).
(200) Osteryoung, J. G., Anal. Chem., 49, 2310 (1977).
(201) Filimonova, M. M., Zh. Anal. Khim., 32, 812 (1977).
(202) Smyth, M. R., and Osteryoung, J. G., Anal. Chem., 50(12), 1632 (1978).
(203) Wade, A. L., Hawkridge, F. M., and Williams, H. P., Anal. Chim. Acta.
105(1), 91 (1979).
(204) Davidek, J., Nemethova, M., and Seifert, J., Z. Anal. Chem., 287,
286 (1977). '
(205) Franke, G., Pietrulla, W., and Preussner, K., Z. Anal. Chem., 298(1),
38 (1979). ~~
-446-
-------
Section 10Q
(206) Beroza, M., and Inscoe, M. N., in Ancillary Techniques of Gas
Chromatography. Ettre, L. S., and McFadden, W. H., eds., Wiley-
Interscience, N.Y., 1969, pages 89-144.
(207) Beroza, M., and Goad, R. A., J. Gas Chromatogr., 4., 199 (1966).
(208) Beroza, M., J. Org. Chem.. £8,. 3562 (1963).
(209) Beroza, M., J. Gas Chromatogr.. 2., 330 (1964).
(210) Asai, R. I., Gunther, F. A. , and Westlake, W. E., Residue Rev.,
J.9, 57 (1967).
(211) Asai, R. I., Gunther, F. A., Westlake, W. "E., and Iwata, Y.,
J. Agr. Food Chem.. 20, 628 (1971).
(212) Zimmerli, B., J. Chromatogr., 88., 65 (1974).
(213) Cooke, M., Nickless, G., Prescott, A. M., and Roberts, D. J.,
J. Chromatogr.. .156(2) , 293 (1978) ; Prescott, .A.'M., and Cooke, M.,
Proc. Analyt.. Div. Chem. Soc., 16_(1), 10 (1979).
(214) Minyard, J. P., and Jackson, E. R., J. Agr. Food Chem., 13, 50 (1965),
(215) Erney, D. R., Anal. Lett.. 12(A5). 501 (1979).
-447-
-------
Section 11
TRAINING OF PESTICIDE ANALYTI01 CHEMISTS
This chapter is by far the shortest in the manual, and the reader nay
question the logic of devoting a special section to this subject. During
the years of operating the interlaboratory quality control program des-
cribed in Section 2, the editors observed overwhelming evidence that
participating laboratories with chemists who had formalt specialized
training demonstrated far superior analytical performance than did those
laboratories which lacked this advantage. We, therefore, regard training
as a highly important subject and deserving of special treatment.
Although many good programs are available in undergraduate and graduate
schools for the training of analytical chemists, few, if any, specifically
train pesticide analysts. An undergraduate or graduate student on a
research project with a professor interested in development of residue
analytical methods does receive valuable training and experience, but
such, professors are few and far between in American education institutions.
A number of companies in the private sector offer short courses particularly
designed for training users of company-produced equipment. A certain few
universities and private educational organizations run short courses
touching upon a few of the highlights of pesticide residue analysis.
Some governmental agencies operate similar short training courses.
The residue chemist must not only be familiar with the technique of trace
analysis in general and of residue analysis in particular, but he must be
able to perform routine service and adjustments and preventative main-
tenance, such as module replacements and replumbing, on his instruments.
In order to achieve these abilities, a generally trained analytical chemist
should be given on-the-job training by an experienced residue chemist when
he is hired, if at all possible. Since this is often not possible,
especially in smaller laboratories, this Manual is designed to substitute,
in small part, for such training and to help the analyst recognize certain
pitfalls and to better perform analyses of biological and environmental
media. There is, however, no really satisfactory substitute for intensive,
practical bench training of the type formerly provided by the EPA Perrine
Primate Laboratory Training Program, Perrine, Florida. During the years of
conducting the interlaboratory quality control program described in Section
2, it was very apparent that those laboratories which took most advantage
of the Perrine training facility recorded far better analytical performances
on round robin samples than laboratories not participating in the training
program. As a specific illustration of this, the reader is referred to
Table 11-1 (copied from Table 2-15 in Section 2) which lists the relative
performance ranking of 34 laboratories in one interlaboratory check sample
exercise.
-448-
-------
Section 11
The eight laboratories with top performance had previously sent personnel
to the Perrine training program. Of the 17 laboratories in the top half
of the table, 10 of these laboratories had Perrine-trained chemists. Of
the 17 laboratories in the lower quality half of the table, only one
laboratory near the top of the lower half had sent personnel for training.
All laboratories which had Perrine-trained personnel are check-marked
next to their identifying code numbers.
The editors feel that the data shown by this table provide most conclusive
evidence of the value of a proper training program in the potential
quality output of a pesticide analytical chemist. Unfortunately, however,
the agency saw fit to discontinue the Perrine training program, the only
one of its kind in existence, and, as stated in Section 2J, some recent
results on interlaboratory quality assurance fat check samples (see Table
2-23) indicate the need for a training program.
It is hoped that some educational institutions or governmental agency will
recognize the need and set up programs to provide such training, and that
laboratory supervisors will take advantage of these in urging their residue
chemists to obtain and refresh, on a continuing basis, their training and
knowledge in analytical and instrumental areas. Rapid developments in
instrumentation and new techniques, and the need to analyze at lower and
lower levels for an ever increasing number of pesticides and metabolites,
dictate a constant need for training and retraining in a field as highly
complex as that of residue analytical chemistry. Furthermore, recent
disclosures of pollution of the nation's air and water by a wide variety
of organic compounds, Including pesticides, point up the need for scientists
with a sound background of analytical expertise.
-449-
-------
TABLE 11-1
Section 11
RELATIVE PERFORMANCE RANKINGS
CHECK SAMPLE NO. 26, MIXTURE IN SOLVENT
Lab; Code
Number
/161.
/137,
/135.
/162.
/ 87.
/113A.
/113.
/ 85.
48.
130.
/ 66.
73.
/ 72.
84.
89.
88.
83,
96.
97.
164.
/ 68.
92.
93.
90.
53.
163.
95.
160.
45.
71.
52.
47.
69.
54.
Compounds
Missed
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
-1
1
1
2
2
0
3
2
3
4
4
False
Identifications
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.
0
. 0
0
1
1
- 0
0
1
0
0
1
6
1
0
0
1
0
0
4
No. of ..
Rejects -'
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
2
0
4
1
0
2
2
1
0
0
0
6
3
2
1
2
4
Total <,/
Score -'
198
198
197
197
197
197
196
196
. 195
195
195
194
194
192
192
189
189
187
181
169
168
. 168
164
159
158
157
146
133
128
127
123
115 "'
84
25
If Values outside confidence limits
2J Total possible score, 200 points
-450-
-------
a
AFID
AFS
API
AR
Section 12
ABBREVIATIONS*
Selectivity
alkali flame ionization detector
amperes full scale
atmospheric pressure ionization
analytical reagent
BGS
BHC
BHT
detector background signal
hexachlorocyclohexane
butylated hydroxytoluene
°C
CCD
CDEC
CI
cm
cone.
DCNA
DDA
CSC
CV
degrees centigrade
Coulson conductivity detector
sulfallate
chemical ionization
centimeter
concentrated
dichloran
bis Cp_-chlorophenyl) acetic acid
carbon skeleton chromatography
coefficient of variation
2,4-D •
DC or dc
DCS
ODD
DDE ,
DDT
DDMU
DEF
DECS
DEPP ,
DEPTP
DFTPP •
DMF
DMSO
DNBP
DNFB '
DNOC
2,4-dichlorophenoxyacetic.acid
direct current
decachlorobiphenyl
See TDE
dichlorodiphenyldichlo'roethylene
dichlorodiphenyltrichloroethane
£.»£.' -DDD» olefin
S,S,S-tributyl phosphorotrithioate
diethylene glycol succ/nate
(C2H50)2-PO-0-C6H5
decafluorotriphenyl phosphine
dimethylformamide
dimethyl sulfoxide
dinoseb
2,4-dinitrofluorobenzene
4,6-dinitro-o-cresol
EC
El
EICP
EPA
EPN
ETD
ETU
eV
electron capture
electron impact . .. • .
extracted ion current profile
Environmental Protection Agency
0-ethyl 0-p_-nitrophenyl phenylphosphonothioate
Environmental Toxicology Division
ethylenethiourea
electron volt
-451-
-------
Section 12
FD
FDA
FI
FID
FPD
fsd
FT
8
GC/MS
GC
GPC
field desorption
Food and Drug Administration
field ionization
flame ionization detector
flame photometric detector
full scale deflection
Fourier transform
gram
gas chromatography coupled with mass spectrometry
gas chromatography
gel permeation chromatography
HCB
HECD
HP
HPLC
HPTLC
Hz
hexachlorobenzene
Hall electrolytic conductivity detector
high performance
high performance liquid chromatography
high performance thin .layer chromatography
hertz
id
IR
•'•sat
inside diameter
infrared . . , .
maximum- current from a saturated detector
k'
K-D
kg
capacity factor
Kuderna-Danish
kilogram
1 or L
LC
liter
liquid chromatography
MC
MCPA
MCPB
m/z
mg
MID
MIS
ml
mm
MOG
MS
MT
molecular ion
microcoulometric
[(4-chloro-p_-tolyl)oxy] acetic acid
4-[(4-chloro-o_-tolyl)oxy] butyric acid
mass to charge ratio
milligram
multiple ion detection
multiple ion selection
milliliter
millimeter
Mills, Onley, Gaither
mass spectrometry
Microtek
-452-
-------
Section 12
N
ng
Ni
nm
NMR
N-P
number of theoretical plates
nanogram
nickel
nanometer
nuclear magnetic resonance
nitrogen-phosphorus
OC1
od
OP
organochlorine
outside diameter
organophosphorus
PAM pesticide analytical manual
PBB polybrominated biphenyl
PC paper chromatography
PCB polychlorinated biphenyl
POP pentachlorophenol
PCT polychlorinated terphenyl
PFTBA perfluorotri-n-butylamine
pg picogram
pH measure of acidity; negative log of H+ concentration
PID photoionization detector
PLOT porous layer open tubular
PM photomultiplier
PNP 4-nitrOphenol
ppb parts per billion
ppm parts per million
ppt parts per trillion
psi pounds per square inch
^-values partition ratio of a solute between immiscible solvents
QA
QC
quality assurance
quality control
R
RF
Rj«
RRT
RSD
resolution
radiofrequency
ratio of distance moved by TLC spot to distance of solvent front
relative retention time
relative standard deviation
Rp value relative to that of a standard compound
s or SD standard deviation
SCOT support coated open tubular
SEU standard error unit
SICP selective ion current profile
SIM selected ion monitoring
SIS selected ion summation
SPED sulfur-phosphorus emission detector
SPRM standard reference material
-453-
-------
.Section 12
T total error
2,4,5-T 2,4,5-trichlorophenoxyacetic acid
TC to contain
TCDD 2,3,7,8-tetrachlorodibenzo-p_-dioxin
TD to deliver
TDE DDD; 2,2-bis(£-chlorophenyl)-l,l-dichloroethane
THF tetrahydrofuran
TIC total ion current
TICP total ion current plot
TLC thin layer chromatography
P
VS
yi
ym
UV
V
vs.
WCOT
micron; also atomic mass units
microgram
microliter
micrometer
ultraviolet
volts
versus
wall coated open tubular
* For abbreviations, names, and formulas of pesticides not listed,
see the U.S. EPA Analytical Reference Standards Manual (EPA-600/9-78-012).
-454-
-------
Section 13
QUALITY CONTROL MANUAL REVISIONS
This manual will be revised biennially dnd all persons on the
mailing list will automatically be mailed copies of the revisions. The
question then for each manual holder is whether his name is in fact
on the list. Consider the following points:
1. If you received this manual or a set of revisions in response to a
mail or phone request, you. are definitely on the list. ... .
2. If you received the manual as a handout at some training course,
and your name and affiliation were not recorded, you are probably
not on the list and, therefore, will not automatically receive
revisions.
3. If you obtained your copy of the manual from some individual not
associated with the Laboratory at Research Triangle Park, NC, you
are probably not on the list and, therefore, will not automatically
receive revisions.
If, after reading the foregoing, there is a doubt that you may not
be on the mailing list, please clip off the section below, complete it
in full and mail it as shown to ensure that you will receive all future
revisions. ......
TO: Quality Assurance Section, Anal. Chem. Br. (MD-69)
Environmental Toxicology Division
EPA, Health Effects Research Laboratory
Research Triangle Park, NC 27711
This is to request that your record be reviewed to be certain the
undersigned is on your mailing list to receive copies of all future
quality control manual revisions.
(Print or type name and full business address)
455
*U.S. GOVERNMENT PRINTING OFP1CE: 19 9 2 . 618 . 0 0 3/10 7 9 0
-------
-------