In the context of future storage of nuclear material in deep geological layers, the transfer of selenium-79 from groundwater to biosphere through irrigation is one of the scenarios considered by the ANDRA (National Agency for Radioactive Waste Management). So, the soil would act as an interface between the geosphere and biosphere.
Actually the model adopted to evaluate the element mobility in soil is based on a simple representation of its distribution between the quantity adsorbed on the soil and the amount remaining in the solution (KD model). Such distribution is considered as instantaneous, reversible and linear with the concentration of contamination. This model has some inadequacies with respect to selenium because this latter can be present in different redox states that control its mobility and whose transformation kinetics among states are poorly known (Se(-II), Se(0), Se(IV) and Se(VI)). In order to improve predictions on the mobility of selenium in soil, selenate (Se(VI)) - which is the most mobile form - has been used to study its interactions with respect to two different soils (soil B and soil R). A kinetic model, alternative to the Kd model, has been developed to describe the evolution of stocks of Se(VI) in solution. This model considers that a fraction of selenium is associated with soil in a reversibly way (potentially mobile) and a portion of it is stabilized in soil (pseudo-irreversibly fixed). This model integrates on one hand, in the soil, kinetics of biotic and abiotic stabilization and on the other hand, in solution, a reduction kinetic.
With the goal of acquiring the parameters of the models, various experiments using dialysis bags have been effectuated, both in batch and with open-flow reactors. The parameter acquisition has allowed kinetic and Kd models to be compared in different realistic scenarios of contamination (chronic or sequential) of a surface soil with 79Se(VI). In addition, the sorption mechanisms of Se(VI) have been evaluated in the two soils in batch adding specific competitors (humic acid and calcium carbonates) that
can sorb on solid sites such as selenate. This study has been completed with the analysis of the sorption of Se(VI) on pure phases available on the market (silica, aluminum hydroxide, goethite, bentonite, calcium carbonate and humic acids) or extracted from a soil (humic substances). This investigation has been carried out at different concentrations of Se(VI) (10-8, 10-6 et/ou 10-3 mol/L). In addition, some pure reactive phases have been added to the soil in order to study the solid-solid interaction having a role on the sorption of Se(VI).
This study has underlined that in the R-soil Se(VI) was sorbed in the form of outer sphere complexes (OSC) at concentrations lower than 10-6 mol/L, whereas in B-soil, the majority were sorbed as inner sphere complexes (ISC). As the formation of OSC is reversible and instantaneous, in experiments with open-flow reactors, the use of Kd model was sufficient to describe the sorption of Se(VI) in R-soil. On the contrary, for soil B, the study has shown that the Kd model, unlike the kinetic model, was inadequate to describe the pseudo-irreversible sorption of Se(VI), caused by the formation of ISC. Due to the supply of nutrients for microorganisms, it has been shown that biotic mechanisms predominated in soil B, due to the utilization of cellulose dialysis bags. However, abiotic mechanisms took places in soil B, too.
The studies on pure phases have shown that only aluminum hydroxide (pH 5.2 and 8) and goethite (pH 5.2) could sorb Se(VI), respectively in a pseudo-irreversible and reversible way (for [Se(VI)] < 10-6 mol/L).
Moreover, it has been shown that, in both soils, the addition of some pure phases (goethite and aluminum hydroxide), could cause an increase or a decrease of the sorption of Se(VI) with respect to the one
expected (additivity reaction).
This study has shown that, knowing only the element composition of the soil, it is not sufficient to evaluate the sorption of selenate without any experimentation.