European journal of soil science vol:60 issue:4 pages:590-612
P>Prediction of the fate of metals in soil requires knowledge of their solid-liquid partitioning. This paper reviews analytical methods and models for measuring or predicting the solid-liquid partitioning of metals in aerobic soils, and collates experimental data. The partitioning is often expressed with an empirical distribution coefficient or K-d, which gives the ratio of the concentration in the solid phase to that in the solution phase. The K-d value of a metal reflects the net effect of various reactions in the solid and liquid phases and varies by orders of magnitude among soils. The K-d value can be derived from the solid-liquid distribution of added metal or that of the soil-borne metal. Only part of the solid-phase metal is rapidly exchangeable with the solution phase. Various methods have been developed to quantify this 'labile' phase, and K-d values based on this phase often correlate better with soil properties than K-d values based on total concentration, and are more appropriate to express metal ion buffering in solute transport models. The in situ soil solution is the preferred solution phase for K-d determinations. Alternatively, water or dilute-salt extracts can be used, but these may underestimate in situ concentrations of dissolved metals because of dilution of metal-complexing ligands such as dissolved organic matter. Multi-surface models and empirical models have been proposed to predict metal partitioning from soil properties. Though soil pH is the most important soil property determining the retention of the free metal ion, K-d values based on total dissolved metal in solution may show little pH dependence for metal ions that have strong affinity for dissolved organic matter. The K-d coefficient is used as an equilibrium constant in risk assessment models. However, slow dissociation of metal complexes in solution and slow exchange of metals between labile and non-labile pools in the solid phase may invalidate this equilibrium assumption.