Speciation and matrix effects on arsenic removal by coagulation with ferric chloride and alum

Bench-scale studies were conducted with synthetic freshwaters and actual source waters to assess the applicability of coagulation processes to meet a lower standard for arsenic in drinking water. The parameters examined for their effects on arsenic removal include initial arsenic concentration and oxidation state, pH, coagulant type and dose, source water composition, background particulates, and natural organic matter.; Studies conducted in synthetic freshwaters indicated that arsenic removal by coagulation with ferric chloride or alum was generally consistent with removal being governed by adsorption onto freshly-precipitated amorphous iron(III) or aluminum hydroxides. This hypothesis is supported by observations that As(V) removal by ferric chloride was more efficient than As(III) removal and that, for both As(III) and As(V), removal was independent of initial As concentration (over a limited range) and increased with coagulant dose. However, the pH dependence of As(III) removal by coagulation with ferric chloride was not consistent with that of adsorption. In addition, the negligible As(III) removal by alum, as compared with moderate removal by ferric chloride, suggested adsorption alone could not account for observed removal. Partial oxidation of As(III) to As(V) at the iron(III) hydroxide surface might account for this discrepancy.; The addition of inorganic and organic solutes to synthetic freshwaters resulted in significant decreases in AS(III) removal by ferric chloride for sulfate addition at low pH (4 to 5) and for organic matter over the pH range 4 to 9. Removal as As(V) by ferric chloride or alum was unaffected by variations in solution composition below pH 8 for ferric chloride and at pH 6 and 7 for alum. Significant decreases in As(V) removal by ferric chloride at pH 9 were observed in the presence of organic matter. Although an anomalous decrease in As(V) removal by ferric chloride was observed at pH 6 with addition of higher concentrations of organic matter, this effect was eliminated by concomitant addition of calcium.; In general, arsenic removal from synthetic freshwaters (amended with organic matter, sulfate, and calcium) was comparable to that observed with actual source waters. Thus the poor As(V) removal observed with source waters at pH 9 could be attributed to the effects of organic matter (and to a lesser extent phosphate) and the poor AS(III) removal observed with source waters over the pH range 4 to 9 to the effects of sulfate (at low pH) and of organic matter. Under optimal pH conditions, however, excellent removal of As(V) could be achieved with either ferric chloride or alum, which suggests that, for influent As(V) concentrations typical of U.S. source waters, residual, dissolved As(V) concentrations {dollar}{lcub}<{rcub}2{bsol} {bsol}mu{dollar}g/L in product water can be achieved by coagulation.