crystal faces are assumed, all sites are identical, and, at least Surface complexation models (SCMs) are commonly used to over the range of natural pHs, sites can assume two or three replicate observed acid/base and ion binding behaviors of mineral protonation states (10)(13)(15).
surfaces. A variety of models have been employed; the constructs
Potentiometric titration data, typically used to determine can differ in representation of both the surface and the interfacial the acid/base properties of oxides, can generally be reasonregion. Little effort has been made to assess how or why model ably well fit using many different SCMs (4, 5, 16). Among response depends on the model employed or the surface complexes more complex constructs, such as the TLM, several combiconsidered. We have examined three SCMs and determined how nations of parameters can often replicate such data equally model-predicted cation partitioning and proton release depend on well (16, 17).
pH, ionic strength, cation loading, model, model fit, and proposed surface complex. Significant differences were observed, between Researchers using SCMs to analyze metal binding by oxmodels, for comparable complexes. On the other hand, triple-layer ides have invoked a variety of sorbed surface species to model responses for several different complexes were almost idenreplicate observed behavior (see Fig. and Table ). In tical over certain ranges. Model properties that contribute to the some cases the choice of surface species was constrained by observed behaviors were considered.