COMPARATIVE EFFECTIVENESS OF 14 SOLUTIONS FOR EXTRACTING ARSENIC FROM 4 WESTERN NEW-YORK SOILS

Four western New York soils differing in soil type and land use were subjected to (i) extraction of As by 14 solutions for soil-solvent mixtures shaken for 1, 3, 7, and 18 hours; (ii) a sequential extraction with solvents designed to identify water soluble-As (Ws-As) and As associated with Al (Al-As), Fe (Fe-As), and Ca (Ca-As); and (iii) a hot HNO2-H2SO4 digestion to determine 'total' As. Arsenic was determined by atomic absorption spectrophotometry employing arsine generation. In each soil sample the sequential extraction demonstrated that Fe-As >> Ca-As > Al-As > Ws-As; the total amount removed by the sequential extraction was only 34 to 41% of the total As in three low As (15.0 to 16.4 μg/g As) soil samples and 98% in a lead arsenate-sprayed orchard sample containing 52.1 μg/g As. Deionized distilled water, 1N NH4Cl, and 0.5M solutions of CH2COONH4 ('NH4Ac') NH4NO3, and (NH4)2CO3 generally removed < 0.1 μg/g As from the soils. For the more effective solvents, the amount of As extracted increased with shaking time but an 18-hour shake never exceeded 80% of that extracted sequentially for any sample. The relative effectiveness of the basic solutions in removing As was in the order of 0.5M NH4F ≃ 0.5M NaHCO3 < 0.5M (NH4)2CO3 < 0.5M Na2CO3 < 0.1M NaOH; for acid solutions the order was 0.05N HCl + 0.025N H2SO4 ≃ 0.5M HCl < 0.5M KH2PO4 < 0.5N H2SO4. Arsenic extraction data obtained in this study are consistent with the view that P and As react similarly in solutions containing acetate, sulfate, fluoride, bicarbonate, hydroxide, and hydrogen ions.