Mechanistic sorption models: Species, Thermodynamic, Application

During the last two decades mechanistic sorption models not only continued their development and parameterization, but also gained ground for application in real-world scenarios such as in the long-term safety analysis of potential nuclear waste repositories. This was only possible because fundamentals such as a proper identification of surface species (their numbers, stoichiometries, structures & denticity) could be based on combinations of spectroscopic experiments, thermodynamic modelling and quantum chemical calculations. Similar progress can be reported for the mineral characterization (specific surface area, binding sites, protolysis reactions). Based on realistic species set and mineral properties, respective formation constants can be derived from batch sorption experiments, also providing information about temperature dependence and kinetics (namely reversibility). Nowadays, mechanistic sorption models are not only a synonym for surface complexation models (SCM), but ideally also account for additional phenomena such as ion exchange or surface precipitation, as well as the formation of secondary phases (smart Kd-values).
The above sketched developments are illustrated for a recent case study about Np(V) and U(VI) sorption onto components of Gorleben overburden sediments. The talk presents the analysis of the on-site data situation. Then own measurements to complete the thermodynamic data base including species identification by means of time-resolved laser-induced fluorescence and attenuated total reflection Fourier-transform infrared spectroscopies are addressed, together with the fit procedure to obtain SCM parameter sets. Next, the scheme utilized for smart Kd computation (including its implementation into reactive transport codes) is explained, and results from an uncertainty and sensitivity analysis are discussed.
Conclusions will incorporate a strategy to join international expertise (and man power) aiming at a comprehensive sorption raw data re-evaluation. This would allow to derive an internally consistent (with respect to EDL definition, mineral characteristics and species set) data set for the computation of smart Kd-values. This strategy covers data assembly, evaluation, processing and storage into an appropriate data structure.