1 Department of Environmental Engineering, Technical University of Denmark
The adsorption of phosphate, arsenate (As(V)) and arsenite (As(III)) onto synthetic calcite was studied in a series of batch experiments. The adsorption of the three ions was studied separately followed by studies of the competition between arsenate and phosphate. The experimental data was utilized to set up models for sorption of arsenate and phosphate onto calcite. This was done to enhance the understanding of the governing processes controlling adsorption as well as to develop a tool to predict the fate of arsenate and phosphate in a calcareous environment. The adsorption was studied in eleven different calcite-equilibrated solutions that varied in pH, PCO2, ionic strength and activity of Ca2+, CO3 2- and HCO3 -. To avoid the precipitation of phosphate or arsenic-containing minerals the experiments were conducted using a short reaction time (generally 3 h) and a low concentration of phosphate (≤ 50 μM) and arsenic (≤ 33 μM). The results show that little or no arsenite adsorbs onto calcite within 24 h at initial arsenite concentration of 0.67 μM. In contrast, both arsenate and phosphate adsorbs readily and quickly onto calcite, with arsenate adsorbing faster than phosphate (adsorption is complete after 1 and 2-3 hours, respectively). Also desorption is fast and complete for both ions within 0.5 h. The reversibility of the sorption process indicates that neither arsenate nor phosphate is readily incorporated into the calcite crystal lattice under our experimental conditions. The phosphate adsorption affinity for calcite is greater as compared to arsenate and the phosphate sorption isotherms are more strongly curved. However, the amount of both arsenate and phosphate adsorbed varied with the solution composition in the same manner. In particular, adsorption increased as the CO3 2- activity decreased (at constant pH) and as pH increased (at constant CO3 2- activity). The dependency on the carbonate activity indicates competition for sorption sites between carbonate and arsenate/phosphate, whereas the pH dependency is likely a response to changes in arsenate and phosphate speciation. The primary effect of the ionic strength on phosphate sorption onto calcite is its influence on the activity of the different aqueous phosphate species. For the adsorption of arsenate onto calcite, the effect of the ionic strength is more pronounced and cannot fully be accounted for by changes in the aqueous activity of the different arsenate species. Studies on competitive adsorption of arsenate and phosphate onto calcite show that the adsorption of arsenate onto calcite is strongly reduced by the presence of phosphate, whereas phosphate adsorption is only slightly reduced by arsenate addition. Simultaneous and sequential addition (3 hours apart) yields the same reduction in adsorption, underlining the high reversibility of the system. The reduction in adsorption of both arsenate and phosphate is most likely due to competition for the same sorption sites at the calcite surface, considering the similarity in sorption edges, pKa’s and geometry of the two anions. The adsorption of arsenate and phosphate in the single sorbate systems was modelled successfully using either the constant capacitance model (CCM) for calcite or the CD-MUSIC model for calcite. Generally the models capture the variation in arsenate and phosphate adsorption onto calcite as a function of solution composition. However, it was necessary to include two types of sorption sites in the CCM to reproduce the convex shape of the sorption isotherm, with the fraction of strong sites being greater in the phosphate model compared to the arsenate model. By combining the models for single sorbate systems the competitive adsorption of phosphate and arsenate onto calcite in a binary sorbate system could be predicted with the CD-MUSIC model. This is in contrast to the CCM (also based on the single sorbate systems) which under-predicted the competitive effect. This clearly shows the importance of competition studies in validating multicomponent models. Extrapolation of the experimental results to calcite bearing aquifers suggests a large variability in the mobility of arsenic. Under reduced conditions, arsenite, which does not adsorb onto calcite, will dominate and, hence, arsenic will not be affected by the presence of calcite. In contrast, when conditions are oxidizing, arsenate is the predominant species and, because arsenate adsorbs onto calcite, arsenic mobility will be significantly lower. However, the presence of competing ions must be considered. Phosphate is often present at concentration levels sufficient to significantly reduce arsenate adsorption and this suggests that adsorption of arsenate onto calcite is of minor importance in most groundwater aquifers. This underlines the importance of including competitive effects when estimating the mobility of an ion in the environment.