Within the DGR project in the Czech Republic, the local Mg/Ca bentonite named BaM (Bentonite and Montmorillonite) has been tested with the aim to assess its sorption qualities with regard to cations necessary for the performance assessment of the repository.Sorption batch experiments present a tool to estimation of sorption coefficients which provide useful information on the transport properties of barrier materials. Depending on the requirements of the performance assessment and of the preferred transport models, various sorption models might be applied on the experimental data, which enable us to draw conclusions on different levels of understanding.
Hereby, data from batch experiments with Sr and Cs on Czech natural Mg/Ca bentonite which were performed in two background electrolytes, CaCl2 and NaCl respectively, of the same ionic strength (Is = 0.1 mol·L-1) are presented. Several values of solid-to-liquid ratios (m/V, [g·mL-1]) and a range of tracer total concentrations (C0, [mol·L-1]) were tested. The total concentration was controlled by non-active carrier (SrCl2·6H2O or CsCl) and the distribution between solid and liquid phase was assessed with the use of radioactive tracer (85Sr or 137Cs), concentration of which was negligible. On the basis of previously performed kinetic experiments, the equilibrium time was set to be 3 days for Sr and 7 days for Cs. The experimental vials were placed on the horizontal shaker for the desired contact time after which samples were centrifuged (966 g for 10 minutes) and 2 mL of supernatant were measured in a well-type NaI(Tl) scintillation detector. The average pH values at the end of experiments were 7.5 ± 0.1 in case of CaCl2 background electrolyte and 8.0 ± 0.1 in case of NaCl background electrolyte.
The sorption of Sr showed linear sorption isotherm in the whole range of tested concentrations. However, there appeared to be a dependence of distribution coefficient (Kd, [L·kg-1]) on the value of m/V, especially pronounced in NaCl background electrolyte. This dependence might be described by power function. After such recalculation, sorption data are applicable for the use in the modelling of Sr transport in the compacted bentonite. Sr Kd values measured in NaCl background electrolyte were generally higher than those in CaCl2 background electrolyte for the given experimental conditions.
The sorption of Cs was non-linear due to Cs dual sorption mechanism on the clay material. The major mechanism is ion exchange and the minor mechanism is specific sorption on the frayed edge sites (FES). The sorption on the FES is dominant in the lower Cs concentration range whereas it is superimposed by ion exchange in higher concentration range. This dual sorption mechanism results in the large dependence of Kd values measured for each Cs initial concentration, which ranged from tens to thousands L·kg-1. The Kd values were higher in lower concentration range and the non-linearity was more pronounced when CaCl2 was used as a background electrolyte. No significant influence of m/V ratio was observed, even though certain trends might be identified. The sorption isotherm of Cs was described by two-site Langmuir isotherm.
Because the main mechanism of Sr sorption onto clay material is ion exchange, which is supported by observed fast sorption kinetics and the linearity of sorption isotherm, the geochemical sorption model used was one-site ion exchange model realized in the PhreeqC3 programming environment. The model was described by five ion exchange reactions. The ion exchange sites initial concentration and composition was based on the real composition of bentonite BaM. Equilibrium constants for exchange reactions of four major exchangeable cations were taken from literature (Bradbury and Baeyens, 2002) and equilibrium constants for tracer cations were fitted.
For the case of Cs, two-site ion exchange/surface complexation model was applied. The fitted equilibrium constant for ion exchange reaction was found to be significantly higher than those for other cations. Surface complexation was described by single reaction. Because of such simplified approach, the fitted constants of surface complexation reaction differed for each background electrolyte used with that for CaCl2 background electrolyte being higher which is in agreement with the result of empirical sorption model.