Modelling of the cation binding properties of fulvic acids: an extension of the RANDOM algorithm to include nitrogen and sulphur donor sites

The original RANDOM algorithm (, and ) for modelling cation binding to pedogenic fulvic acids, along with certain other discrete site models, possesses a major advantage through its ready incorporation into speciation packages, such as MINTEQA2 (Allison et al., 1991) or JESS. A limitation of the original RANDOM, however, is its restricted applicability to soil solutions. This paper describes a new development, namely, an extension of RANDOM to all types of fulvic acid including both pedogenic and aquagenic, which is motivated by the authorsʹ interest in modelling metal speciation in seawater. As with the original version, the new RANDOM2 is based on a random assignment of aromatic rings and functional groups to a carbon backbone. The resulting statistically synthesised structures are used to identify possible monodentate and bidentate binding sites for cations as well as concentrations of these sites. Whereas the original RANDOM is limited to oxygen-donor ligands, the new RANDOM2 incorporates not only oxygen but also nitrogen and sulphur ligands. RANDOM2, accordingly, is particularly useful for modelling cation binding to marine fulvic acids. RANDOM2 has been validated (in conjunction with MINTEQA2) against experimental data for protonation of and copper complexation by Suwannee River fulvic acid as well as observed complexation of copper by marine organic matter.