Calibration of a General Mineralogy-Constrained Sorption Model for Heavy Metals and Application to Evaluation of Reactive Transport Processes in Natural Soil Columns

Heavy metal transport in the subsurface is controlled by physical and geochemical equilibrium and/or non-equilibrium processes. We present a transferable reactive transport model for lead and cadmium incorporating both non-electrostatic surface complexation and ion exchange reactions on major soil mineral surfaces (iron and aluminum hydroxides, kaolinite, illite, and montmorrillonite). Metal surface reaction stoichiometries on the different mineral surfaces are based on molecular-scale surface coordination chemistry as determined from public spectroscopic studies. The model was calibrated to batch equilibration experiments carried out on a suite of soils over a range of pH (2-6) and metal concentrations (0.5-6 mM). Surface and exchange site densities were estimated from quantitative mineralogical determinations and cation exchange capacities. Calibration involved global optimization of surface reaction equilibrium constants for each metal onto the various mineral surfaces by fitting to experimental equilibrium dissolved metal concentrations. The transferability of the calibrated model was tested with a soil batch data set not used for calibration, and was found to adequately describe lead and cadmium sorption when differences in soil mineral abundances were taken into account. The calibrated model was then applied to evaluate soil column transport experiments in which a pulse of lead and/or cadmium was introduced and subsequently flushed. Good agreement between experimental and modeled breakthrough curves indicate that lead and cadmium mobility in the soil columns is controlled primarily by relatively fast sorption equilibria. Agreement between model predictions and experiments was improved at higher flow rates and was generally better for cadmium than lead, which showed some tailing during column flush-out, suggesting that lead transport is also partially controlled by non-equilibrium processes. The results of this study illustrate how a calibrated batch equilibrium model can be applied to elucidate reactive transport processes observed under dynamic flow conditions.