Effects of sorption competition on caesium diffusion through compacted argillaceous rock

We carried out a small-scale laboratory diffusion experiment on a disk-like sample of Opalinus clay from the Mont Terri underground laboratory (Switzerland) using 134Cs as tracer. A through-diffusion phase was followed by an out-diffusion phase where the tracer taken up by the sample was released again. Since the tracer concentration at both boundaries was monitored, careful mass-balance considerations were feasible.A first analysis of the experimental data was done in the frame of a single-species model accounting only for transport and non-linear sorption of caesium. The model could match the data of the through-diffusion phase, however only, when strongly reducing the sorption data based on batch sorption experiments. Yet, such a procedure was in strong contradiction with sorption measurements performed on dispersed and compacted systems. In addition, predictions concerning tracer out-diffusion and mass-balance considerations clearly revealed the shortcomings of this type of model.In a second attempt we applied a multi-species transport model where now the whole water chemistry and a sorption model for caesium were considered. First, the value for the diffusion coefficient was fixed to the best-fit value of the single-species model. But again, the sorption site densities had to be reduced strongly albeit the reduction factor was smaller. Only when fixing the sorption site densities to those values of the sorption model and letting the effective diffusion coefficient De free for the adjustment, could through-diffusion data be reasonably well fitted and out-diffusion as well as mass-balances be predicted in a satisfying manner. The main results are: (1) The best-fit could be achieved with a value for De of 1.8 × 10−10 m2 s−1 which is rather high but corroborated by results of a molecular modelling study. (2) If caesium arrives in the Opalinus clay sample potassium and sodium (calcium etc.) ions are released and caesium ions are sorbed. The released cations diffuse to lower concentration regions according to their individual concentration gradients. Since locally the cation concentration for potassium, (sodium and calcium) is increased, sorption of these cations is also locally enhanced, affecting in return the sorption behaviour of migrating caesium. Consequently, the sorption process of caesium in such diffusion experiments cannot be addressed by a non-linear isotherm formalism any longer. (3) A reasonable analysis of such single tracer diffusion experiments therefore requires the combined description of transport (diffusion) and sorption of many cations and the whole complex water chemistry of the system. Thus, single-species models can only be applied with care in the considered concentration ranges.