Modeling sorption of divalent metal cations on hydrous manganese oxide using the diffuse double layer model

Manganese oxides are important scavengers of trace metals and other contaminants in the environment. The inclusion of Mn oxides in predictive models, however, has been difficult due to the lack of a comprehensive set of sorption reactions consistent with a given surface complexation model (SCM), and the discrepancies between published sorption data and predictions using the available models. The authors have compiled a set of surface complexation reactions for synthetic hydrous Mn oxide (HMO) using a two surface site model and the diffuse double layer SCM which complements databases developed for hydrous Fe (III) oxide, goethite and crystalline Al oxide. This compilation encompasses a range of data observed in the literature for the complex HMO surface and provides an error envelope for predictions not well defined by fitting parameters for single or limited data sets. Data describing surface characteristics and cation sorption were compiled from the literature for the synthetic HMO phases birnessite, vernadite and δ-MnO₂. A specific surface area of 746 m²g⁻¹ and a surface site density of 2.1 mmol g⁻¹ were determined from crystallographic data and considered fixed parameters in the model. Potentiometric titration data sets were adjusted to a pH_IEP value of 2.2. Two site types (≡XOH and ≡YOH) were used. The fraction of total sites attributed to ≡XOH (α) and pK_a2 were optimized for each of 7 published potentiometric titration data sets using the computer program FITEQL3.2. pK_a2 values of 2.35±0.077 (≡XOH) and 6.06±0.040 (≡YOH) were determined at the 95% confidence level. The calculated average α value was 0.64, with high and low values ranging from 1.0 to 0.24, respectively. pK_a2 and α values and published cation sorption data were used subsequently to determine equilibrium surface complexation constants for Ba²⁺, Ca²⁺, Cd²⁺, Co²⁺, Cu²⁺, Mg²⁺, Mn²⁺, Ni²⁺, Pb²⁺, Sr²⁺ and Zn²⁺. In addition, average model parameters were used to predict additional sorption data for which complementary titration data were not available. The two-site model accounts for variability in the titration data and most metal sorption data are fit well using the pK_a2 and α values reported above. A linear free energy relationship (LFER) appears to exist for some of the metals; however, redox and cation exchange reactions may limit the prediction of surface complexation constants for additional metals using the LFER.