Application of a Two-Site Ideal Model for the Prediction of As-SO4-Cl Ion Exchange Equilibria

This paper was aimed at obtaining a deeper understanding of the binary equilibria of As(V)Cl− and SO4 2−/Cl− and of the ternary equilibria of As(V)/SO4 2−/Cl−. To this purpose, a series of batch tests were carried out at different values of the total solution normality. These data were combined with those obtained through continuous flow column tests performed to collect equilibrium data over the entire ionic fraction domain. The equilibrium data were then described using two different models based on the assumption of ideal behaviour of both the liquid and the adsorbed phases: in the first model the resin was considered to have only one type of binding-sites, where two types of binding sites were assumed by the second model, named double-selectivity model. Among these two models, the latter provided the best fitting of binary equilibrium data for both As chlorides and sulphate chlorides systems. However, the same model was unable to fit the experimental data in the ternary system.

Ion exchange processes are effective for the removal of arsenic (As) from drinking water. However, the As uptake capacity of ion exchange resins is affected by the presence of other anions such as sulphates and nitrates. As these ions are typically found in groundwater, the design of ion exchange process aimed at removing As from groundwater may be affected by their presence. Therefore, to properly design an ion exchange process for As removal, it is important to characterise the ion exchange equilibria of As in solution in the presence of competing anions. This paper was aimed at obtaining a deeper understanding of the binary equilibria of As(V)Cl- and SO42-/Cl- and of the ternary equilibria of As(V)/SO42-/Cl-. To this purpose, a series of batch tests were carried out at different values of the total solution normality. These data were combined with those obtained through continuous flow column tests performed to collect equilibrium data over the entire ionic fraction domain. The equilibrium data were then described using two different models based on the assumption of ideal behaviour of both the liquid and the adsorbed phases: in the first model the resin was considered to have only one type of binding-sites, where two types of binding sites were assumed by the second model, named double-selectivity model. Among these two models, the latter provided the best fitting of binary equilibrium data for both As chlorides and sulphate chlorides systems. However, the same model was unable to fit the experimental data of As in the ternary system As(V)/SO42-/Cl- with a satisfactory agreement probably due to the presence of non-ideality which the model did not account for.