SURFACE COMPLEXATION MODELING OF METAL ION ADSORPTION ONTO IRON OXIDES IN SINGLE AND BISOLUTE SYSTEMS

The development of predictive models to describe metal ion sorption has been a primary focus of research over the past several decades. Surface complexation models (SCMs) have emerged as promising tools for predicting sorption to oxide and clay materials. However, the reliability of SCMs is dependent on selection of appropriate surface complexation reactions and model parameters. This research focused on developing a strategy for developing a self-consistent database of Triple Layer Model (TLM) parameters for two iron hydroxides, ferrihydrite (HFO) and goethite a-FeOOH). Spectroscopic data were used to guide the selection of surface complexation model reactions, and to the extent possible, theoretically based parameter estimation techniques were employed.

The adsorption of two divalent metals, Cd(II) and Pb(II), onto a-FeOOH and HFO was modeled in single and bi-solute systems using the TLM. Selection of surface complexation reactions was supported with evidence from extended x-ray absorption fine structure spectroscopy (EXAFS). The surface site density values were based on tritium exchange experiments. The TLM parameters were optimized for single solute adsorption and then used, without any modification, to predict adsorption in additional single and bisolute systems.

HFO was modeled using the assumption of one average surface site while goethite was modeled with strong and weak sites. Cd(II) and Pb(II) adsorption were modeled using mononuclear inner-sphere bidentate and monodentate surface complexes onto HFO and inner-sphere bidentate surface complexes on both weak and strong surface sites onto goethite. The calibrated TLM predicted single-solute data and captured the competitive adsorption behavior reasonably well.

Selenite (Se(IV)) adsorption onto goethite was also modeled in single and bi-solute systems (with Pb or Cd as competing solute) using the calibrated TLM. Inner-sphere bidentate surface reactions were used to describe Se(IV) adsorption. While reasonable predictions of bi-solute data were observed for Cd(II) and Pb(II) in the presence of Se(IV), the model predictions for Se(IV) suggested that fewer sites were available for oxyanion adsorption compared to cation adsorption.