JESZHCIDS .AITAI.yp-FS Bx GAS C3EOMATOGPAPHY

                                 AT THE LA::3 MICHIOA:i BASH; OFFICE*

                                             by

                          W. D, Johnson, F, D. Fuller and L. E. Scarce*'^
             Hecords of organic pesticide usage by farmers show that, in IS66,

        37 percent were herbicides^  29 percent insecticides, h percent fungicides,

        e.nd 8 percent miscellaneous  pesticides.  More than 150 million pounds of

        pesticides are purchased by  urban dwell er^j  hk to 109 pounds of LOT per

        acre was  used to control Dutch Elra disease in Wisconsin urban co.rj-iunit.Les,

        This high urbar  usage of pesticides contributes a large pollution load to

        lakes and streams.   The evolutionary growth of gas chromatography is meeting

        the Federal Water Pollution  Control Administration (FrfPCA) laboratories' ne.-^d

       .to measure low level and changing pesticide characteristics.

             The  PvJTCA's Lake Michigan Bab in Office (LMEO) has the responsibility

        for conducting pesticide surveillance program in the Great; Lakes and the

        Central Missis sitroi  River Basins  waters.   Assistance and consultation Is also
                                                                              i
        vided to  federal; strata,  and local agencies for pesticide analysis  by gs.s

        chromatography.

             This  paper  elaborates on analytical  methods employed for chlorinated

        end thiophosphated pesticides,  types of samples analyzed,  tyrleal con-

        centrations,  and the  significance of findings.
       •^Presented e.t the National Meeting of  vhe 20th Mid-Amerie'r  Symposiur1 on
        Spectroscopy? May 13; 19^9; ir^ 'Jiiicago, Illinois

       '^Orgardc Chemist» .Supervisor^ Che,mi?try Sect: on jana  Chi?:\,  Leboratc -y
         Services Branch, respective] y; U.S.  D&ps-^'c jsnt of the I~~crior_,  Federal
         Water Pollution Control Adniniso.-v.'.-M . La!-;e Jiicbig
-------
                                    -2-




                               INTRODUCTION




      The Lake Michigan Basin Office (LMBO) is responsible for conducting




 pesticide surveillance programs in the Great Lakes and the Central Mississippi River




 Basins waters. Assistance and consultation is also provided to federal, state,




 and local agencies for pesticide analysis "by gas chromatography.  Pesticide




 pollution of these regions presents a potential hazard for aquatic organisms,




 fish, birds, vildlife, and man.  The chlorinated pesticides are the most




 persistent,while the thiophosphates are more highly toxic.




    "  Records of organic pesticide usage by farmers show that,  in 19^6,




 37 percent were herbicides,  29 percent insecticides,  h percent fungicides,




 and 8 percent miscellaneous  pesticides.  More than 150 million pounds of




 pesticides vere'purchased by urban dwellers;  ^ to 109 pounds  of DDT par acre




 "was used to control Dutch Elm disease in Wisconsin urban coomunities.  This




 high urban usage of pesticides contributes a large pollution load to lakes




 and streams.  Total. U-.S.  production of pest control chemicals  in 1967 was




 approximately 1.25 billion pounds having a market value of about 800 million




 dolls;rs.   The total U.S.  consumption was over one billion pounds of active




-pesticide chemicals (1).   Of all the insecticides used today,  75$ is applied




 to less than 2$ of the land,  and of the lj-57 million acres of farmland in the




 United States,  it is estimated that only 15$  of total crop acreage receive




 pesticides  (2).




      Thousands'of pounds  of  chlorinated,  thiophosphated, and other pesticides




 run off into lakes and streams yearly.   The application of pesticides has




 been  so extensive that DDT is found in  antarctic penguins and  arctic life




 forms,  such  as lichens.   Chlorinated  pesticides  are so persistent that




 toxaphene was found ten years after application  in  Wisconsin lakes  (3)•

-------
                                    -3-

      The concentration of pesticides  are  determined by  several factors  such

 as kind of pesticide,  sorption  ability  and  climate.  As a general statement.,

 the concentration  of pesticides in  waters is "based  upon the  amount directly

 received,  that  portion from runoff, and the stability of molecules towards

 physicochemical '  effects.-  For example,  aldrin  decomposes to  form its

 epoxide,  dieldrin;  DDT generally decomposes to its  isomers/  pp'-DDT,  op-DDT,

 DDE,  etc,;  and  heptachlor is converted  to heptachlor epoxide.
     >

      The Initial, higher concentrations  of chlorinated and thiophosphated

 pesticides after spraying,  dusting  or runoff have the most detrimental

 effect  on fish  and wildlife. Dosages necessary  to  produce immediate  kills

 vary from species  to species'but generally  0.1 mg/1 will seriously sffg-st

 or kill most game  fish and benthic  fauna.   Numerous fish and fowl kills

 have been  reported in  and around Lake Michigan. (4).

      This  paper presents  some highlights  on how  pesticides are routinely

 eoaalyzed at the Lake Michigan Basin Office  by gas chromatography  with

 electron capture and microcoulometric titration  detection, thin layer .

 chromatography,  and positive identification by infrared spectroscopy  (iR).

     "FWPCA's Lake Michigan  Basin  Office routinely analyzes for the following

 chlorinated pesticides:   lindane, heptachlor, aldrin, heptachlor  epoxide,

 dieldrin,  endrin, op-DDT  and pp'  DDT.  Analyses  have "been performed according to

 the U.S. Public  Health  Service  revised methods (5)".

      The Lake Michigan Enforcement  Conference Pesticides  Committee recommends th

following compounds be determined in  the water, fish and  clams of Lake

Michigan:_ DDT,  dieldrin, DDD, DDE,  methoxychlor, chlordane,  and  endrin  (6).

-------
                                   -4-




                              SAMPLING PROCEDURE




     The type of samples analyzed at ths 1,-aka iiiehigaa Basin Office ej.-e;  vater,




bottom sediments, algae or aquatic plants, soils,and fish.



     The vater samples are generally grab samples collected in a modified




Kemmerer Sampler.  The bottom samples are taken vith an Efcasit  or Peters en




dredge or by core sampling.  Algae or small aquatic plants are collected




through the use of a plankton net.  Soil samples are taken from top soil in




areas having been treated vith pesticides.




      The containers are h to 10 liter glass bottles and quart jars cleaned




 vith dicihr ornate cleaning solution and rinsed vith distilled water, alcohol,




 acetone] ether,  then chloroform.  The caps are lined vith aluminum foil (3 )•
        I


                                                                &

      Carbon adsorption filters  may be used as a means  of collecting  or con-
                  "~        *_.-_-».   _ v          _ _ _ _


 centrating pesticides  from vater. This  method involves  the filtering of




 300 to  5,000 gallons of vater through a  cartridge  (3  x 18 inches  in  size)




 packed  vith granular  carbon  at  the rate  of 0.03-0.5 gallons per minute.  The



 cartridge  is filled vith 4.5  inches  of 4 x 10 mesh  carbon,  folloved  bylnine




 inches  of  30 mesh carbon and  finally 4.5 inches  of  4 x 10 mesh  carbon,




    . A  rapidly analyzed grab  sample  of vater,  bottom  sediments, algae,  aquatic




 plants  and fish may be more reliable since it permits  less degradation of




 pesticides.   Grab sampling also offers a savings in manpower and  equipment.

-------
                                   ,-5-



      Coraposite  sampling of the above may give a more representative picture




 of the pesticide  content over a period of 2 or 3 days.  However, some




 pesticides,  such  as thiophosphated and carbemates, may be degraded before




 the analysis can  be performed.



      All types  of samples, except  water, may be preserved for a few days or




 months by freezing depending upon  type of pesticide.  However, when collecting




 water samples in  the winter months, it is necessary to add sodium chloride or




 ethyl alcohol to  the water sample  to prevent freezing and breakage of the




 sample containers while in the field.  Water samples are stored at 5°C




 until analyzed.




                           ANALYTICAL PROCEDURES




 Sanrole and ReaŁent_Prep_aration.



      All water  samples analyzed for pesticides by the Leke Sichlgan Basin Office are




 subjected to  liquid-liquid extraction, using redistilled chloroform as




 the  solvent.  One  gallon of sample is placed in a ^--liter separatory




 funnel and 25 ml of saturated sodiun sulfate solution and 5 ^l of 1:1




 hydrochloric  acid  are added for each liter of sample.  The sample is




 extracted three times using 100 ml of redistilled chloroform for each




 extraction.   The extract is dried by pouring over a 2-inch column of




 anhydrous sodium sulfate.  The extract is cleaned-up by passing through




 a column of Florisil topped with 1-inch of anhydrous sodium sulfate.



 The  extract is then evaporated to a volume of 0.5 ml.  Further clean-up may




be accomplished by thin-layer chromatography separation or trapping




of specific peaks by gas chromatography.

-------
      Bottom sediments  are  spread in  aluminum lined pans  approximately




 12 x 2h x 2 inches,,  covered with gauze,  then air dried by forcing air




 currents across  the  surface of the sediment  until dry enough to grind to




 approximately  30 mesh.   This requires from one  to four days.  Pesticides




 sorted on particulate matter may "be  deserted by a 2k-  hour continuous  Soxhlet




 extraction of  a  25-200 gram sample with  chloroform.   The bottom sediments




 extracts are cleaned- up  in the same  manner described  in  the  paragraph above,




      Algae,  aquatic  plants and soil  samples  are dried, ground,  extracted




 end" the extracts cleaned-up in the same  manner  as  bottom sediment samples.




      Fish and  other  aquatic organisms are ground in s. blender end. an  aliquot




 removed for analysis.  The aliquot is dried vith anhydrous sodiun sulfate,




 extracted with chloroform and the extract cleaned-up  in  the  same manner  as




 described above.




      Carbon filter samples are dried at  40°C, then extracted vith redistilled




 chloroform in  a Soxhlet  apparatus for 35 hours.  The  chloroform extract  is




 then  concentrated and cleaned-up  in  preparation  for further  analyses  (7).




     All solvents are redistilled or commercially  available  redistilled




 solvents  are employed.  Reliability  of solvent purity  is  checked by running




 a blank.




Analysis  of  Samples




     Water  analyses are generally performed according to procedures in the




Federal Water Pollution Control Administration' s Interim Official ''isthods for
Chj.orina.ted I^drj3carbon_ .Pesti^cid.es in Water and Wjyy^'^at&r by Gas_ Chro^atosjranhy.




es recommended by the Pesticide Committee of the Lake Michigan Enforcement




Conference

-------
                                    -7-


     Bottom sediment samples are analyzed In accordance with the procedures


described in the F^PCA Great Lakes Region's Ch^mistr,/ Jghorator^ Manual -


Bottom §e3j.ments_ (9)-


     The analyses of fish samples are performed according to procedures


described by the Food and Drug Administration (lO).


     Samples of water, bottom sediments, algae, aquatic plants and fish are


analyzed by thin layer chromatography and gas chromatography with electron


capture and microcoulonetric titration.  The Lake Michigan Basin Office


•utilizes gas ehromatographs with an electron capture Nickel-63 detector,


flame ionization and microcoulometric halogen and sulfur titration cells.


The operating parameters of the gas ehromatographs and microcoulometrlc

                        o
titration have bean as follows:
            -3

     1.  Electron Capture


         a.  Volume injected:  0.5-4.0  ul   -


         b.  Inlet temperature: 225 °C


         c.  Column temperature: 175°c


         d.  Detector temperature:  195°C

                                      '  2
         e.  Attenuator setting: 32 x 10


     2.  Microcoulometrlc  Titration


         a.  Volume injected:  5-0  to 50.0 |al


         b.   Column temperature: 195"c


         c.   Inlet temperature:  250°C


         d.  Furnace  temperature:  850°G


          e.   Sensitivity setting:  500  ohms

-------
                                 ,  --8-


      The following  columns  are  used,  in gas  chromatography vith  electron

 capture detector.

 Material                       Suport
l) Aluminum   5 ft.    1/4"    100/120 Mesh Gas Chrom Q   7^ OV-17

                                                          9^ QF-1

2) Aluminum   5 ft. -   1/4"    80/100 Mesh Gas Chrom Q   10$ DC-200

                                                         1556 QP-1

For Biicrocoulometric detection, the gas chromatography column used is as follows;

Stainless Steel  5 ft  1/4"    70-80 M8sh IXDusted Gas   5$ DC-200
                                     Pack


«OV-17 - Methyl Phenyl Silicone
 QP-1  = Pluorosilicone
 DC-200 s Dow Corning Silicone Oil 200 (l2, ^00 centistokes)


     Thin layer chroratography involves spotting microliter portions of an

extract from a pesticide sample on a 200 x 200 mm glass plate coated with a

0.25 mm silica gel or alumina layer impregnated with silver nitrate.  This

procedure is used to semi -quant it ate, separate, or clean-up a sample. " The

prepared plate is developed in a closed developing tank containing a 1 cm

depth of carbon tetrachloride .  The plate is permitted to develop to a 10 cm

finish line.  The plate is removed and placed under strong ultraviolet (U.Y.)

light to develop the chlorinated pesticides  which, if present,  react with

the silver nitrate to precipitate silver as  "black spots.   If TLC is  employed

for clean-up,  plain silica gel or alumina plates are developed with  rhodamine 3

solution and the spots are identified by the JL-< value (ratio of spot travel

to the travel of the solvent front).  Samples  are semi -quant it at ed by comparing

visually with standards .

-------
                                   -9-





I^truinentati^on




     Infrared spectrophotornetry is used for identification  vhen pesticides



are present in relatively high concentrations.





      Gas  chromatography  results may-be'coupled with mass  spectroscopy (MS),




IR, UV, Nuclear Magnetic Resonance (111®) or   Differential Thermal  Analysis(DTA)




The resulting data can be fed into a  computer for  extended studies and data




processing.  This multiple instrumentation is available commercially  for




use vhen  tnalysis or studies require  further  confirmation.








                              DISCUSSION




     Clean-up of. samples, is a major problem, in pesticide  analysis.  Water




samples relatively free of oil may be cleaned-up by passing through a  k-6 inch




Fierisil-1-inch-anhydrous sodium sulfate column after extraction.  If the




sample is oily^  acetonitrile partitioning is  used to separate the pesticides




from the oil.  Bottom sediment,  algae, aquatic plants, and fish extracts are




cleaned-up in the sane manner as oily water samples.  If further clean-up is




necessary, the extracts are passed through a Florisil-arihydrous sodium sulfate




column.  After concentrating the cleaned-up samples,  containing a drop of




0.0015$ paraffin oil (a pesticide holding agent),  the samples are ready for




analysis by gas  chromatography,  thin layer chromatography or infrared




spectroscopy (3).

-------
                                  -10-





     Gas chromatography with electron capture is used as a rapid, quantitative




method for screening pesticides at the Lake Michigan Basin Office.  "By employing




shorter columns (approximately 2~5 feet), the retention time has been shortened;




however, if better resolution is desired, longer columns are employed.  Two or




more columns of different polarities are used to change the retention time




of the standard and unknown, thereby obtaining a means for tentative




identification




     There are numerous problems associated with the gas chromatograph.  The




electron capture detector may give improper response due to low standing




current.  This condition may be corrected by cleaning the detector,  unless




this effect is due to a slow bleeding column; this latter condition  may be




corrected by replacing tHe old column with one that has an immobile  phase with




a higher boiling point.  Carrier gas filter,  poor voltage,  defective detector,




and gas leaks in the system cause low standing voltage.





     Poor resolution can be corrected by employing the proper column substrate,




temperature,  replacing defective columns, correcting gas flow and improving




or correcting injection techniques.




     Thin layer chromatography is a useful tool that L?ISQ employs to isolate,




clean-up, and semi-quantitatively or qualitatively analyze for pesticides.




This is one of the most rapid methods of analysis and clean-up procedures.   However,




the sensitivity of this procedure is much less than by gas chromatography




and less quantitative.

-------
                                   -11-

     Since TLC is approximately 1,000 times less sensitive than gas chrorta-

tography (GO), TLC is useful only when pesticide concentrations are present in

microgram quantities.  Visual estimation of concentrations'makes' this method

only semi-quantitative.  However,.what -seems to "be a problem "becomes an asset

if the purified spots are employed in GC or IR for confirmatory or positive

identification.

      Infrared spectroscopic  identification is  employed to give positive

 identification of questionable  pesticides.   This method relies upon obtaining

   purified saaples so  that there  are a minimum number of  overlapping

 absorption peaks  in the  fingerprint.   Approximately  50 to 100  ug  of sample

                                                                          i.
 are required for  infrared identification.   A purified sample is obtained

 from a column-chromatography fraction, thin layer  chromatography  s-oot or gas

 chromatography trapping.  The large  sample requirement prevents the maximum

 use of ordinary infrared spectroscopy, since many  water saaples have less

 than 50 ug of material after preparation and purification.

      Infrared spectroseopy  is one of the most  povcrful scientific tools  for

 giving positive identification, but  presents the problem  of  obtaining

 10,OOO times the  amount  of  residue required for gas  chromatography,  even when

 employing a beam  condenser  and  scale expansion.  For good fingerprints the

 sample extract must be thoroughly cleaned-up and separated into relatively

 pure compounds prior to  analysis.

-------
                                   -12-



      Mierocoulometric tltration  after gas  chromatogra-phy is  a specific method



 for determining halogenated or thiophosphated pesticides.  The sensitivity



 of the method is approximately 1/3 to 1/20 of that  obtained  in gas



 chromatography with  electron capture. -



      Microcoulometric gas  chronatography combines the principles of gas



 chromatography,  combustion,  and  eoulometry into one operation.  The problem



 of separation  of extract into pure fractions is handled  through the use of



 an appropriately packed column of  the correct length and gas flow to give



 good resolution.  The problem of incomplete combustion is minimized by



 adjusting furnace temperature, oxygen or hydrogen and nitrogen flow rate.



 Also,  resistance is adjusted 'so that one nanograin of sulfur or chloride
                *       o


 produces at 3,east 5 percent of a full recorder scale deflection during



 microcoulometric  titration.



      Maintaining a uniform  amount  of electrically generated silver ion in



 the coulometer  cell is insured by maintaining the correct amount of silver



 plate on the electrodes, .flushing  electrodes vith electrolyte, and quickly



 cleaning with dilute  nitric acid and rinsing vith distilled water.  When



 sulfur dioxide  is being titrated with the triiodide ion,  the platinum



 electrodes may become  coated, reducing the sensitivity.  This difficulty can



 be  corrected by  cleaning the electrodes with dilute hydrochloric acid followed



 by  a  rinsing with distilled water,  and/or changing the electrolyte.



                                SUMMARY

                                                          *

      The lake Michigan Basin Office has  the responsibility for conducting



pesticide surveillance programs in the Great Lakes and the Central Mississippi



 River Basins waters.

-------
                                   -13-

     The detection of pesticides at low levels in water is made

possible through the use of gas chromatography and other sophisticated

analytical instrumentation,

     Grab sampling and analysis of water, "bottom sediment,, algae, aquatic

plants, soils, and fish are used for the rapid assessment of pesticide

3.evels in streams and lakes.

     Composite sampling over a period of 2 or 3 clays, 'however,  is more

representative of pesticide loadings.

     After sample preparation,  extraction, clean-up and concentration,
       i
samples are analyzed by the following methods:

          Gas chromatography vith a HI-63 electron capture detector and micro-

          coulornetric titration is employed with special columns to separate

          end analyze pesticide samples.

          Thin layer chromatography is a practical vay to semi-quanti-

          tatively screen pesticide samples; this method may also be

          employed to clean-up or purify samples for infrared spectroscopy

          end gas chromatography.

          Infrared spectroscopy is employed as a more absolute method for

          identifying unknown pesticides.

-------
                                CONCLUSIOKS




1.   Pesticide concentrations in lakes and streams may "be kept under




     surveillance through the use of gas chromatography with electron




     capture and microcoulometric detection, vith infrared and




     thin layer chromatography as associated or hack-up methods.




2«   The analysis ofvater, bottom sediments, algae, aquatic plsxrbs, soils,




     and fish gives information required to adequately assess pesticide




    • levels of lakes and streams.
Disclaimer:



Mention of products and manufacturers is for- identification only and does




not imply endorsement by the Federal Water Pollution Control Administration




or the U, S. Dept. of the Interior.

-------
                                    -15=
                                 REFERENCES




1,  D, E. H. Frear, Pesticide Handbook .- Entoma, 20th Ed,, College Science




    Pub., State College, Perm. (1968),, pp. 27-31.




2,  R. F. Gould.  Advance in Chemist.ry Series, Organic Pesticides in the




    Environment, (1966), p. 3*




3.  W. D. Johnson, G. F. Lee and D« Spyridakis,  Persistence of Toxaphene




    in Treated Lakes. (Thesis, Univ. Wisconsin) (1965) and Air and Water




    Pollution, Int. J. Pergarnon Press 10, (1966) pp. 55-560.  Printed in




    Great Britain; also CA 66sl8l6b, (196?).




li;  J. J. Hiekey, J, Keith and F. B» Coon.  An Exploration of Pesticides




    in a Lake Michigan Ecosystem.  J. Appl.  Ecol. 3, (1966), pp.  liu-153.




5?o  H. P. Burchfield, D, W. Johnson and E. E. Storra.  Guide to the




    Analysis of Pesticide Residues, Yols. I  and II.   U. S. Dept.  of




    Health,  Education, and Welfare, Washington, D. C. Supt.  of Documents,




  •  U. S. Government Printing Office, I,  (l)-VIir.S,3.d«,(l). (196Ł).'




6.  U, S. Dept. of the Interior,  F,=/PCA.   Report on Insecticides in L^ke




    Michigan,  prepared by Pesticide Com.'nittee of the Lake  Michigan Enforce-




    ment Conference (Donald I. Mount, Ph.D.,  Chairman,  National Water




    Quality Laboratory,  Duluth, Minn.),  (1968), pp.  1-kk.




7.  A. W. Briedenbach', J.  J.  Lichtenberg,  C.  F. Henke,  D,  J.  Smith,




  "  J. W. Eichelberger,  Jr.,  and  H. Stierl.   The  Identification and




    Measurement of Chlorinated Hydrocarbon Pesticides in Surface Waters,




    U.S.  Dept.  of the Interior, WP-22, (1966),  pp. 1-70.

-------
                                    -16-
 8.  (Organic Pesticide Subcommittee on Methods Validation and Analytical

     Quality Control.  FWPCA Interim Official Methods for Chlorinated
     i
     Hydrocarbons Pesticides in Water and Waste Water by Gas Chromatography,

     (1968), pp. 1-21.

 5>.  U. S. Dept. of the Interior, F/JPCA, GLR, LMBO.  Great La.kes Region

     Committee on Analytical Methods, Chemistry Laboratory Manual, Bottom

     Sediments, (1963) pp. $0-73.

10.  R. E. Duggan, H. C. Barry, H, Y. Johnson, and Sidney 'v'illiarr.s.  Pesticide

     Analytical Manual, Vol. I:  Methods Which Detect Multiple Residues)

     Vol. lit  Methods for Individual Pesticide Residues, U.S. DKEW, Food
     a'nd Drug Administration, (revised 1?67 and 1968),

11.  A1, A. Rosen.  Chemical Analysis - A Weapon Against Water Pollution,

     J. Anal. Cheni. 39,  (196?) pp. 26A-33A.

-------