engleski

Sorption of arsenic on clay minerals - a theoretical approach

INTRODUCTION The mobility of arsenic species in the environment is largely controlled by solid phase sorption reactions (Langmuir, 1997 ; Wilkie and Hering, 1998). Quantitative evaluations of the solid phase/water partitioning of many toxic metals, including arsenic, are best accomplished by a surface complexation approach (Cherry et al. 1986 ; Sadiq, 1997). Equilibrium-based thermodynamic modelling is currently one of the most appropriate methods to evaluate the competitive geochemical processes that affect the transport and toxicity of arsenic, including predictions regarding arsenic persistance and mobility in the environment. NUMERICAL MODELLING The numerical computer code PHREEQC was used to simulate arsenic surface complexation from a small watershed with naturally high levels of arsenic on clay mineral components (kaolinite, montmorillonite and goethite) of a stream sediment. Surface complexation mass-action coefficients were obtained from the literature in the generalized two-layer model form, or from linear free energy relationships. A reasonable simulation of expected arsenic concentrations is obtained using competitive complexation, literature-derived sorption constants and the PHREEQC model. The USGS computer code PHREEQC (mainframe version 1.6 - released in January 1998) was used for all simulations. The WATEQ4F thermodynamic database formed the core to which surface complexation parameters were added. The code was used for the calculation of saturation indices, sensitivity analysis of parameters such as Eh, pH and temperature, modelling the mixing of stream water of different compositions. The surface complexation routine was used in the generalized two-layer model and competition between arsenic and other ions for sorbing phases for a finite number of sites was allowed. Theoretical mineral assemblages were allowed to come to equilibrium by simulating the flushing of many pore volumes through the sediment-mineral surface assemblage. The model output was evaluated using the ratio "R" of modeled vs. expected arsenic concentrations. One important limitation of the PHREEQC code is that its version 1.6 does not consider coprecipitation. It is possible to model nonideal solid solutions with two components or ideal with any number of components. The main drawback in the formulation is that complete thermodynamic equilibrium is assumed. At each point in reaction the entire solid is in equilibrium with the solution. The code was run on a CRAY T3E mainframe computer of the John von Neumann-Institute for Computing (NIC), at the Research Center Juelich, Juelich, Germany. SIMULATION EXAMPLE Adsorption isotherms in solutions (approximating an average stream water) with ionic strengths of 0.001 at 25 deg C were obtained by the simulation, for an arsenite and arsenate concentration range of 10-7 to 10-3 M and a pH range of 4-10. At low concentrations these isotherms obey equations of the Langmuir type. At higher concentrations the isotherms are more linear, indicating thze existence of more than one type of surface site on the amorphous iron hydroxide (goethiteamorph.) adsorbent. Removal of arsenite and arsenate by amorphous Fe hydroxide throughout the selected concentration range were determined as a function of pH. Thermodynamic data were taken from literature (Pierece and Moore, 1982) and verified through the WATEQ4F database. CONCLUSIONS AND IMPLICATIONS  Small differences in the conceptual model and data aquisition techniques can have a large effect on the simulation error. The error of neglecting competition by common compounds such as bicarbonate or silicic acid can equal or exceed the bias resulting from inappropriate choice of mineral phases.  A careful selection of the relative concentration of arsenic and amorphous iron hydroxide and pH, it should be possible to achieve high removal rates (approaching 90%) of arsenic from As-containing waters.  The continued development of field and analytical protocols to support surface complexation modelling is clearly needed.

arsenic; clay minerals; sorption

nije evidentirano