Analyst, June, 1968, Vol. 93, $9. 363-367 363 Extraction of Organochlorine and Organophosphate Insecticides from Lake Waters BY H. B. PIONKE, J. G. KONRAD, G. CHESTERS AND D. E. ARMSTRONG (Department of Soils, University of Wisconsin, Madison, Wisconsin 63706, U. S.A .) A quantitative method for the unified extraction of organochlorine and organophosphate insecticides contained in waters at microgram per litre concentrations is described. Recoveries of parathion-methyl, diazinon, malathion, azinphos-methyl, .)I-BHC, heptachlor epoxide, aldrin, dieldrin, endrin, @’-TDE, pfi’-DDT and methoxychlor added to seven lake waters and distilled water ranged from 93.9 to 102.4 per cent. Recoveries of hepta- chlor ranged from 81-0 to 88.0 per cent., apparently because of degradation of the insecticide in aqueous systems.METHODS for the extraction of insecticides from waters must essentially fulfil certain criteria. It is particularly important that the method is sensitive, that it quantitatively extracts all the insecticide present and is sufficiently rapid and simple to be applicable to routine analysis. The advent of gas - liquid chromatography, with either an electron-capture detector for organochlorine, or potassium chloride thermionic detectors for organophosphorus, insecti- cides, coupled with concentration techniques, has provided the necessary sensitivity for detecting these compounds in concentrations commonly found in naturally occurring waters. The criteria of quantitative extraction are necessary because of the heterogeneity of natural water with respect to the kinds and amounts of contaminants present that could alter insecticide extractability.The preferred methods for extracting organochlorine and organophosphate insecticides from waters have been solvent-extraction techniques because of their ready applicability to routine analysis. Recoveries of organochlorine insecticides by solvent extraction are usually high and, in many investigations, 85 to 103 per cent. of heptachlor,l DDT, aldrin, dieldrin and y-BHC192 in “spiked” water samples at pg per litre concentrations were recovered by extraction with hexane, diethyl ether or chloroform. In contrast, recoveries of organo- phosphate insecticides are generally lower and, in one instance, five successive extractions of waters containing pg per litre concentrations of parathion and diazinon with either a 1 + 1 mixture of light petroleum and ether or chloroform recovered only 90 per cent.of the added insecticide.2 Another worker achieved a more efficient extraction, in which 90 per cent. of parathion-methyl and parathion contained in waters at pg per Litre concentrations was recovered in a single hexane extraction.* However, there is little information available to indicate that an adequate extraction method for organophosphate insecticides in water has been obtained. Because the previously described batch extractions are less efficient in the ng per litre concentration range, continuous solvent-extraction technique^,^ or activated carbon filters,l s2 s6 are often used at this level of sensitivity to obtain detectable amounts of organochlorine or organophosphate insecticides.However, the use of activated carbon is limited because of suspected degradation or irreversible adsorption of organochlorine or organophosphate insecticides .1 0 SAC and the authors364 INSECTICIDES- yBHC (y-isomer of 1,2,3,4,5,6-hexachlorocyclohexane) was obtained from Hooker Electrochemical, Niagara Falls, New York, with a purity of greater than 99 per cent. Heptachlor (1,4,5,6,7,10,10-heptachloro-4,7,8,9-tetrahydro-4,7-methyleneindene) and heptachlor epoxide (the 2,3-epoxide of heptachlor) were obtained from Velsicol Chemical Corp., Chicago, Illinois, with a purity of 98.6 per cent. for heptachlor and 97.9 per cent. for hept achlor epoxide. Aldrin (1,2,3,4,10,l0-hexachloro-l,4,4a,5,8,8a-hexahydro-exo-l,4-eutdo-5,8-dimethanonaph- thalene), dieldrin (the 6,7-epoxide of aldrin) and endrin (the 1,4-endo-endo-stereoisomer of aldrin) were obtained from Shell Chemical Corp., New York, with purities of greater than 99 per cent.pp'-TDE [l,l-dichloro-2,2-di(4-chlorophenyl)ethane], also referred to as fi+'-DDD, was obtained from Rohm and Haas Chemical Corp., Philadelphia, Pennsylvania. The purity was not stated. pp'-DDT [ 1 ,l,l-trichloro-2,2-di(4-chlorophenyl)ethane] was obtained from Geigy Chemi- cal Corp., Ardsley, New York, with a purity of 99.9 per cent. @'-Met hox ychlor [ 1 , 1 , 1 -t richloro-2,2-di (4-met hoxyphen yl) et hane] was obtained from Geigy Chemical Corp., with a purity of greater than 99 per cent. Parathion-methyl (dimethyl (4-nitrophenyl phosphorothionate) was obtained from Shell Chemical Corp., with a purity of greater than 99 per cent.Diazinon (diethyl(2-isopropyl-6-methyl-4-pyrirnidinyl phosphorothionate) was obtained from Geigy Chemical Corp. labelled at the 4-position of the ring with carbon-14. Malathion (S- [ 1,2-di (ethoxycarbonyl) ethyl] dimethyl phosphorothiolothionate ] was ob- tained from Nuclear Chicago Corp., Chicago, Illinois, labelled with carbon-14 at the 2 and 3- positions of the succinic acid moiety. Azinphos-methyl (Guthion) { S- (3,4-dihydro-4-oxobenzo [dl - [ 1,2,3]-triazin-3-ylmethyl) di- methyl phosphorothiolothionate) was obtained from Chemagro Corp., Kansas City, Missouri, with carbon-14 labelling in the carbonyl group. PIONKE et al. : EXTRACTION OF ORGANOCHLORINE AND EXPERIMENTAL [Analyst, Vol.93 LAKE WATERS- Samples were chosen on the basis of wide variation in composition of the bottom sediments as described earlier.6 The sediments were removed by centrifuging the waters before analysis. INSTRUMENTAL- A Packard, model 7620, gas-liquid chromatograph was used for analysis of organo- chlorine insecticides and parathion-methyl. Gas-chromatographic conditions for chlorinated hydrocarbons were : carrier gas, nitrogen with flow-rate of 125 ml per minute; tritium-foil electron-capture detector 200" C, 50 volts; column, 2 metres long x 4-mm i.d., of 10 per cent. DC-200 on 60 to 80-mesh Gas Chrom Q; column temperature 195" C; inlet temperature 235" C ; outlet temperature 225" C. Gas-chromatographic conditions for parathion-methyl were: carrier gas, helium with flow-rate of 60 ml per minute; potassium chloride thermionic detector, 200" C, 300 volts; column, 2 metres long x 4-mm i.d., of 10 per cent.DC-200 on 80 to 90-mesh Chromosorb W; column temperature, 200" C; inlet temperature, 240" C; outlet temperature, 240" C. The instrument incorporates the use of glass columns and on-column injection to avoid sample degradation resulting from contact of the sample with metal surfaces. A Packard, model 3365, liquid-scintillation spectrometer was used for radioactivity measurements. PPO (2,5-diphenyloxazole) and dimethyl POPOP (1,4-bis- [2-(4-methyl- 5-phenyloxazole)]-benzene) were used as the primary and secondary fluors, respectively. For benzene extracts, a solution of 0.5 per cent.of PPO and 0.03 per cent. of dimethyl POPOP in toluene was used. Corrections for sample quenching were determined by external standardisation. SOLVENTS- were used for "spiking" and extracting the water samples. DESCRIPTION OF METHOD Benzene and acetone, re-purified by glass-distillation with a 3-ball Snyder column,June, 19681 ORGANOPHOSPHATE INSECTICIDES FROM LAKE WATERS 365 PROCEDURE- A method for the extraction of organochlorine and organophosphate insecticides from waters, with benzene as the extractant, was found to be quantitative, sufficiently rapid and simple for use as a routine procedure. The method is described as follows. A 250-ml water sample is extracted with 25 ml of benzene in a single extraction by shaking in a separating funnel for 2 minutes, and the separated benzene phase is analysed directly by gas-liquid chromatography with an electron-capture detector for organochlorine insecticides, or potassium chloride thermionic detector for organophosphates. If both organochlorine and organo- phosphate are present in the sample, simultaneous analysis can be accomplished by splitting the column effluent and by using a dual electron-capture potassium chloride thermionic detector.If emulsification is encountered, anhydrous sodium sulphate can be used to remove the water. RESULTS AND DISCUSSION ORGANOCHLORINE INSECTICIDES- The evaluation of the above extraction procedure for organochlorine insecticides was based on “spiking” water samples with a mixture of the insecticides as follows. Amixture of the insecticides in acetone (20 p1) containing y-BHC, heptachlor, heptachlor epoxide, aldrin, dieldrin, endrin, $9’-TDE, @’-DDT and methoxychlor was added to the 250-ml water samples.The “spiked” sample, contained in a separating funnel, was shaken end-over-end continuously for 12 hours to obtain maximum dissolution of the insecticides. TABLE I RECOVERY OF ORGANOCHLORINE INSECTICIDES FROM LAKE-WATER SAMPLES Recoveries of insecticides, per cent., at concentrations of- A I \ 3-28 pg per litre 6.40 pg 8.90 pg 32.6 pg per litre hepta- per litre chlor per litre per litre pp’- $p’- meth- 1 -60 pg 1-23 pg per litre 2-08 pg hepta- 3.98 pg 5-02 pg per litre per litre per litre Lake waters yBHC chlor aldrin epoxide dieldrin endrin TDE DDT oxychlor Distilled water . . . . 96.5 85.6 101.8 98.3 100.0 99-2 97.0 97.2 98.5 Allequash ... . 96.5 83.6 102.3 99.0 100-3 99.0 98.6 96.9 94.4 Content,. .. . . 96.3 81.0 100.2 96.2 99.0 98-9 96.6 96.7 97.1 Crystal . . .. . . 97-2 88.0 102.0 98.6 98-5 99.1 96.0 94.4 97.0 Pickeral . . . . 95-6 84.5 101.7 96.6 97.5 96.8 97.2 95.0 97.0 Plum . . .. . . 96-8 82.0 102.4 97.0 98-4 97.2 98.0 96.8 97-2 Tomahawk .. . . 97.5 81.7 102.1 96.6 99-0 98.7 98.3 95.3 98-5 Mean recoveries . . 96.7 84.6 101.8 97.4 98.7 98.4 97.3 96.1 97.3 Average range of dupli- cates . . .. . . 2.0 2.9 2.7 1.0 1.7 1.3 2-8 0-7 2.0 Standard deviation be- Little Arbor Vitae . . 97-1 82-0 101.8 96-8 97.0 98-1 97.0 96.3 98-6 tween water samples 0-60 2.60 0.62 1-03 1-12 0-92 0-89 1.02 1.36 With the exception of heptachlor, the recoveries of organochlorine insecticides from lake waters and distilled water ranged from 94.4 to 102 per cent.at pg per litre insecticide concentrations, as seen in Table I. These results are the average of duplicate determinations conducted at different times. The average range of recoveries between duplicate deter- minations i.e., the summation of the range between duplicates divided by the number (i.e., 8, of water samples) for each insecticide was 0.7 to 2.9 per cent. (including heptachlor). This error term combines the errors associated with extraction and the analytical technique, indicating that the method of extraction is reproducible even for heptachlor, which displayed low recovery values. A second error term, the standard deviation, relates the quantitative reliability of the method between lake-water samples.Average extractability ranged from 96.1 to 101.8 per cent. (except for heptachlor), with a standard deviation of 0-60 per cent. for y-BHC to 1.36 per cent. for methoxychlor. The average ranges and standard deviations were deemed sufficiently low to substantiate the general applicability of this method for con- sistently quantitative recoveries of the insecticides from water at pg per litre concentrations.366 [Artalyst, Vol. 93 Heptachlor recoveries were invariably low (81 to 88 per cent.), even when the “spiked” waters were re-extracted with benzene. The consistent reproducibility of heptachlor recovery from a given water, as shown by the low average range between duplicates (2.9 per cent.) and the absence of heptachlor in the second benzene extract, indicated that the low recovery of heptachlor was not caused by incomplete extraction.It seems more likely that the low recoveries were caused by heptachlor degradation during incubation in the water, as suggested by Lamar, Doerlitz and Law.l For heptachlor, the standard deviation between samples was greater than for the other insecticides, which is consistent with degradation losses. If degradation of heptachlor occurs in water, it is expected that degradation rates would vary among water samples because of differences in water composition. Because solvent evaporation was observed to cause large volatilisation losses of aldrin, acetone was chosen as the “spiking” solvent to avoid the evaporation step necessary for the removal of hexane or other water-immiscible “spiking” solvents.After “spiking,” the acetone concentration in the water samples was 80mg per litre, which was considered too small to affect insecticide extractability. The lake waters were chosen on the basis of differences in bottom sediment charac- teristics that are likely to affect the water composition. The sediment samples provided a range of organic matter (0.14 to 64.2 per cent.) and clay (4-4 to 36.1 per cent.) contents; the extent of oxidation of the organic matter as measured by methoxyl content and carbon- to-nitrogen ratio showed wide variability. Furthermore, benzene extracts of the lake waters when gas chromatographed showed no contaminants that would interfere with the detection of the nine organochlorine insecticides at pg per litre concentrations.The concentrations of organochlorine insecticides in water (Table I) were chosen to provide sufficient sample for the accurate evaluation of the extraction method without exceeding the maximum aqueous solubilities. Within these solubility limits, concentrations were selected to obtain about the same detector response for each insecticide, thereby facilitating simultaneous chromatographic analysis of the insecticides. Although the extraction method was not evaluated for insecticide concentrations below 1 pg per litre in water, the quantitative recoveries attained (assuming quantitative recoveries were achieved for heptachlor) are likely to be achieved at one tenth, or less, of the insecticide concentrations presented in Table I.If slight modifications are used to increase the sensitivity of the method, e.g., by extracting larger volumes of water, or by manually changing sensitivity settings of the gas chromatograph, sensitivities can be obtained that blanket the sensitivity range required for the determination of organochlorine insecticides in most natural waters.’ PIONKE et al. : EXTRACTION OF ORGANOCHLORINE AND ORGANOPHOSPHATE INSECTICIDES- The extractability of organophosphate insecticides from lake waters was evaluated on the same samples used for organochlorine insecticides, and the method of extraction used was the same. “Spiking” the waters with organophosphate was achieved in the following manner. The insecticide, in 1 ml of benzene, was added to 250 ml of water and the benzene evaporated by passing a stream of air over the water.Homogeneous solution of the insecti- cides was determined by removing three aliquots of an azinphos-methyl “spiked” water without shaking. Each aliquot was extracted with benzene and the carbon-14 activity of the extracts was 99.3 0.5 per cent. of the expected activity, indicating homogeneous distribution of the insecticide in the water. In a later investigation* it was confirmed by gas chromatography that organophosphate insecticide in benzene extracts of soil-insecticide systems remained in the intact form, indicating that organophosphates extracted from the lake water would not have undergone degradation. The extractability of parathion-methyl, diazinon, malathion and azinphos-methyl added to the seven lake-water samples and distilled water is shown in Table 11.With the exception of parathion-methyl, the average recoveries were 98-4, 99.7 and 99.5 per cent., with ranges of 97.6 to 99-2, 98.3 to 100.6 and 97.1 to 100.9 per cent. for diazinon, malathion and azinphos- methyl, respectively. The average recovery for parathion-methyl was 95.3 per cent ., with a range of 93.9 to 97-7 per cent. This lower recovery value for parathion-methyl probably reflects the higher inherent error in gas-chromatographic determination than in radioassay by liquid scintillation spectroscopy. However, parathion-methyl was determined satisfac- torily at a concentration of 38pg per litre in the lake water, and no concentration of theJune, 19681 ORGANOPHOSPHATE INSECTICIDES FROM LAKE WATERS TABLE I1 RECOVERY OF ORGANOPHOSPHATES FROM LAKE-WATER SAMPLES 367 Recoveries of insecticides, per cent., at concentrations of- A f \ 38 pg per litre 40 pg per litre 60 pg per litre 66 pg per litre Lake waters parathion-methyl* diazinon malathion azinphos-methyl Distilled water .. .. .. 97.7 99-2 10045 98.6 Allequash .. .. .. 96.1 98.4 1004 100.8 Content . . .. .. .. 96.3 98-7 99.3 100.9 crystal .. .. . . 94.4 99.2 99.6 100.6 Little Arbor Vitae . . .. 94.6 99.0 100.6 97.1 Pickeral .. .. .. - 98.1 99.6 98.9 Plum .. .. .. 97.6 98.3 98.7 Mean recoveries .. .. 96.3 98.4 99.7 99.6 Average range of duplicates . . 1.4 1.7 1.6 1.6 Standard deviation between Tomahawk .. .. .. 93.9 97.2 99.7 100.2 water samples . . .. 1.44 0.76 0.78 1.36 * Determined by gas chromatograph, other insecticides determined from carbon-14 activity.benzene extracts was found to be necessary. Concentrating benzene extracts is complicated by the thermal instability of organophosphate insecticides. Parathion-methyl was found to degrade when the benzene extracts were concentrated by evaporation at 80" C, and similar decomposition of other organophosphates would be expected to occur if a concentration step were used. Therefore, concentration of the benzene extract is not recommended prior to gas-chromatographic determination of organophosphate insecticides, unless investigations on the stability of the insecticide are conducted simultaneously. Statistical analyses of the extractabilities of organophosphates were conducted in a manner similar to that used for the organochlorine insecticides.In a l l cases the average range of duplicates and standard deviations were similar to those obtained for ~rganochlorine insecticides, and were well within the limits of statistical error, substantiating the general applicability of the prescribed extraction method for consistent quantitative recovery of organophosphate insecticides from lake waters in the pg per litre range. The method described here is a rapid, reliable procedure that is suited to routine analysis for the simultaneous extraction and determination of organochlorine and organophosphate insecticides from lake waters in pg per litre concentrations. This work was supported in part by the US. Department of Agriculture, ARS Contract No. 12-14-100-8164(41) and the U.S. Department of the Interior, OWRR Project No. B-016-WIS. REFERENCES 1. 2. 3. 4. 6. 6. 7. 8. Lamar, W. L., Doerlitz, D. F., and Law, L. M., Geol. Surv. Wut.-Supply Pup., 1965, 1818-B: 1. Teasley, J. L., and Cox, W. S., J . Amer. Wut. Wks Ass., 1963, 55, 1093. Warnick, S. L., and Gaufin, A. R., Ibid., 1966, 57, 1023. Kahn. L., and Wayman, C. H., Anulyt. Cham., 1964, 36, 1340. Rosen, A. A., and Middleton, F. M.. Ibid., 1959, 31, 1729. Lotse, E. G., Graetz, D. A., Chesters, G., Lee, G. B., and Newland, L. W., J. Envir. Sci. & Tecknol., Faust. S. D., and Suffet, I. H., in Gunther, F. A., Editor, "Residue Reviews," Springer-Verlag, Konrad, J. G., Armstrong, D. E., and Chesters, G., Agron. J., 1967, 59, 691. Received January Sth, 1968 1968, May, 2. Berlin, Gottingen and Heidelberg, Volume 16, 1966, p. 44.