A U.S. Environmental Protection Agency (USEPA) inter laboratory method validation study was conducted on USEPA Method 515.1, "Determination of Chlorinated Acids in Water by Gas Chromatography with an Electron Capture Detector. "This method is one of the 6 pesticide methods developed for the USEPA National Pesticide Survey (NPS). Method recovery and precision for analyses of sub-ppb to low-ppb concentrations of chlorinated acids were determined in reagent water and finished drinking waters. The analytes evaluated in the study included the 12 pesticides that were quantitatively measured in the National Pesticide Survey (bentazon, 2,4-D, 2,4-DB, 3,5-dichlorobenzoic acid, DCPA-diacid, dicamba, dichlorprop, 5-hydroxydicamba, pentachlorophenol, picloram, 2,4,5- T, and 2,4,5-TP) and 5 pesticides (acif luorfen, chloramben, dalapon, dinoseb, and 4-n it ro phenol) that were only qualitatively assessed in the National Pesticide Survey because of recognized method imprecision. The study design was based on Youden's nonreplicate plan for collaborative tests of analytical methods. The waters were spiked with 17 chlorinated acids, each at 6 concentration levels, prepared as 3 Youden pairs. Eight laboratories extracted the spiked test waters at pH <2 with ethyl ether, performed a solvent exchange with methyl tert-butyl ether, prepared methyl esters of the extracted acids using diazomethane, and analyzed an aliquot of each derivatized extract by gas chromatography with electron capture detection. The submitted data were analyzed using a USEPA computer program, which measured recovery and precision for each of the 17 compounds and compared the performance of the method between water types. Method 515.1 was judged acceptable for the 12 NPS analytes recovered quantitatively; mean percent recoveries at 10-15 times the method detection limits ranged from 79 to 105% in reagent water and from 75 to 123% in finished drinking water. In reagent water, overall precision (reproducibility relative standard deviation, RSDR) ranged from 9.6 to 34.2% and in finished drinking water, the RSDR ranged from 11.9 to 37.0%. Single-analyst precision (RSD for repeatability, RSDr) ranged from 5.8 to 17.7% in reagent water and from 4.6 to 27.9% in drinking water. Results for the 5 other NPS analytes were too inaccurate or imprecise and, for these compounds, supported use of the method for qualitative measurements only; the 5 compounds are not included in the adopted method. The method has been adopted first action by AOAC INTERNATIONAL for determination of residues of 12 chlorinated acids in finished drinking water.

A joint U.S. Environmental Protection Agency (EPA)/AOAC International collaborative study was conducted on EPA National Pesticide Survey Method 4, Determination of Pesticides in Finished Drinking Water by High Performance Liquid Chromatography with an Ultraviolet Detector, to determine mean recovery and precision for analyses of 18 pesticides and metabolites in reagent water and finished drinking waters. The study design was based on Youden's nonreplicate plan for collaborative studies of analytical methods. The waters were spiked with 18 pesticides and metabolites at 6 concentration levels, prepared as 3 Youden pairs. Ten volunteer laboratories extracted the spiked test waters with methylene chloride, performed a solvent exchange with methanol, and analyzed an aliquot of each extract by liquid chromatography using an ultraviolet detector at 254 nm. Results were analyzed using an EPA computer program that calculated recovery and precision for each of the 18 compounds and compared method performance between water types. The method was judged useable for all analytes tested. Recoveries were good for 16 of the analytes. The method exhibited poor recovery but acceptable between-laboratory precision for metribuzin DADK and metribuzin DK. The method performance statistics for carbofuran phenol and linuron were calculated to be significantly different for reagent water and finished drinking water. Similar statistical differences were observedfor metribuzin DADK, but these effects were not considered of practical significance due to the poor recovery (<25%) in both waters. In reagent water, the overall relative standard deviation (RSDR) range was 7.8-24.1% for all 18 compounds and 5.5-38.6% in finished drinking water. The single- analyst relative standard deviations (RSDr) range was 5.7-17.6% in reagent water and 4.0- 18.7% in finished drinking water. The method has been adopted first action by AOAC International.

A joint U.S. Environmental Protection Agency (USEPA)-AOAC interlaboratory method validation study was conducted on USEPA National Pesticide Survey (NPS) Method 6, "Determination of Ethylene Thiourea (ETU) in Finished Drinking Water by Gas Chromatography with a Nitrogen-Phosphorus Detector." The purpose of the study was to determine and compare the mean recoveries and precision for determination of ETU in reagent water and finished drinking waters. The study design was based on Youden's nonreplicate plan for collaborative tests of analytical methods. The waters were spiked with ETU at 6 concentrations levels, prepared as 3 Youden pairs. In the method, the test water is extracted by passing the sample through an absorbent matrix type tube. ETU is recovered from the tube with methylene chloride, the extract is solvent-exchanged to ethyl acetate, and an aliquot of each extract is analyzed by gas chromatography using a nitrogen-phosphorus detector. Twelve laboratories participated in the study. Data were analyzed using a USEPA computer program, which measured recovery and precision for ETU and compared the performance of the method between the 2 water types. Over the concentration range tested, the mean percent recoveries of ETU were 82-92% in reagent water and 85-98% in finished drinking water. The range of the betweenlaboratory relative standard deviations (RSDR) for the 6 concentrations was 5-24% in reagent water, but was only 4-9% in finished drinking water. The range of the within-laboratory relative standard deviations (RSDr) was 6-14% for reagent water and 6- 10% for finished drinking water. Results for the 2 water matrixes showed no statistically significant differences. The method has been adopted first action by AOAC International for determination of ETU in finished drinking waters.