Ion binding to natural organic matter: General considerations and the NICA-Donnan model

The general principles of cation binding to humic matter and the various aspects of modeling used in general-purpose speciation programs are discussed. The discussion will focus on (1) the discrimination between chemical and electrostatic interactions, (2) the binding site heterogeneity, (3) the models for the electrostatic interactions, such as the Donnan model and (4) the complications inherent to ion binding in multi-component systems, including a different stoichiometry for different ions. The basic NICA (non ideal competitive adsorption) model addresses issues (1), (2) and (4) and ensures thermodynamic consistency. The electrostatic interactions (3) can be included using the ion activity adjacent to the binding sites. To compute this activity it is necessary to use an electrostatic model that relates the particle charge and the characteristic particle potential. Several simplifications to arrive at such a relation are analyzed. The Donnan model is proposed as the most convenient model. A short derivation of the NICA-Donnan model is presented to illustrate that the derivation is quite straightforward once it is clear which assumptions should be made. The main parameters of the NICA model are the site density, the median affinity constants for the different ions, the ion stoichiometry and the width of the generic affinity distribution. All these parameters are intrinsic properties that are independent of pH, salt concentration or metal concentration. The key parameter of the Donnan model is the Donnan volume, for which an empirical relation with one adjustable parameter is advised. The NICA-Donnan model has been used in several studies. To illustrate the capabilities of the model some results obtained for purified humic acid are presented together with model predictions. Key features of the model are its ability to adjust to different heterogeneities and binding stoichiometries and its predictive abilities with respect to metal ion-proton and metal-metal ion competition. With the application the model should be regarded as semi-empirical because assumptions have been made for its derivation that are not met in practice. Both for this reason and since it is difficult, if not impossible, to find a unique set of parameters for a given system the interpretation of the parameter values may be difficult. A brief discussion is given on the generic NICA-Donnan description of fulvic and humic acids presented by Milne et al. [C.J. Milne, D.G. Kinniburgh, E. Tipping, Environ. Sci. Technol. 35 (2001) 2049; C.J. Milne, D.G. Kinniburgh, W.H. van Riemsdijk, E. Tipping, Environ. Sci. Technol. 37 (2003) 958]. Finally a brief literature review is presented on the use of the NICA-Donnan model for other heterogeneous natural sorbent systems. (c) 2005 Elsevier B.V. All rights reserved.