Combining Surface Complexation and Linear Regression Models to Predict Cd, Ni, and Pb Adsorption on Guizhou Yellow Soils
This study investigates the adsorption characteristics and mechanisms of Cd, Ni, and Pb on yellow soils collected from multiple sites in Guizhou Province, China. Soil physicochemical properties, potentiometric titration, adsorption edge experiments, and a 1-site/2-pKa surface complexation model (SCM) were integrated to derive acid-base parameters and metal adsorption constants. Linear regression models established quantitative relationships between SCM parameters and soil properties, enabling prediction of adsorption behavior for new soil samples. Results showed that adsorption capacities increased with pH and followed the order Pb > Ni > Cd, influenced by hydrated radius, hydrolysis constant, and electronegativity. Soils with higher surface site concentration (Hs) and lower point of zero charge (pHpzc) exhibited greater metal adsorption. The SCM fitted adsorption edges with R ≥ 0.94, confirming its validity. pH was the dominant factor controlling metal complexation constants (lgKSOMe), acid-base equilibrium constants, and surface site density (Ds), with influence order Pb > Ni > Cd. Free iron oxide correlated negatively with deprotonation constant (pKa2). Ds was also affected by cation exchange capacity and specific surface area. Validation using a separate set of soil samples yielded good agreement between predicted and measured adsorption (R² = 0.75–0.82, RMSE = 0.1–0.51). This combined modeling approach simplifies experimental procedures and offers a robust tool for assessing heavy metal environmental risks and remediation strategies.