Diffusion of Np through Illite du Puy

Illite is a clay mineral, which is present in many natural clay formations that are considered as potential host rock for nuclear waste repositories. As part of the geological barrier and in connection with safety assessments, it is therefore essential to know its retention capability with respect to radionuclides with long half-lives and to quantify their potential migration. In this context, diffusion experiments in combination with reactive transport simulations are a powerful tool to unravel the movement of radionuclides through porous clay systems.

In this study, first long-term investigations of Np diffusion through compacted Illite du Puy (1600 kg/m3) are presented. Two experiments, referred to as Np-1 and Np-2, were conducted under ambient conditions at pH 4.7±0.2 using a well-established set-up. The clay samples with a thickness of about 10 mm and a diameter of about 25.4 mm were placed in diffusion cells between two stainless steel filter plates. On the upstream side, each diffusion cell was connected to a source reservoir, respectively, each containing the background electrolyte (0.1 mol/L NaClO 4, 1·10-4 mol/L KCl, 5·10-3 mol/L 2-(N-morpholino)ethanesulfonic acid pH buffer, 1·10-3 mol/L NaN3) and 237 Np(V) with an initial concentration of 1.5·10 -5 mol/L. ... mehrOn the downstream side, a receiving reservoir containing just the background electrolyte was connected to each cell, respectively. In that way, a concentration gradient inside the clay samples was established.

Within the first 230 days of the experiment, the source reservoirs were regularly sampled and the receiving reservoirs regularly replaced. In the solutions, the 237 Np activity was determined by liquid scintillation counting. Based on these activity data, the diffusive flux and the accumulation of activity in the receiving reservoir was calculated. Applying the Fick’s law approach, values for the effective diffusion coefficient, De [m2/s], and the distribution coefficient, Kd [m3/kg], were determined. After termination of the diffusion experiments, the diffusion samples were pushed out of the diffusion cells and cut in segments under anaerobic conditions. Their total 237 Np content was determined via acid extraction (1 mol/L HNO 3). The fractions of Np(V) and Np(IV) were determined by solvent extraction of clay segment aliquots.

Np-1 was terminated after about 1059 days. At this time, the diffusing Np in the sample had reached an almost equalization of the concentration gradient. Particularly at the clay sample edges, high amounts of Np(IV) were detected. This could be due to microbial growth in the reservoir solutions, since organic matter can deliver organic complexing compounds and quite efficient reductants. Alternatively, the negative clay layer charge, caused by isomorphic substitutions by Fe (or Mg), in connection with the thermodynamic preference of the sorption of Np(IV) could have shifted the redox equilibrium to Np(IV).

In the case of Np-2, a fresh receiving reservoir containing 0.1 mol/L NaCl was connected after 1016 days with the aim to change the 237 Np diffusion profile and provide applicable data for reactive transport modeling. Np-2 was terminated after 1050 days. The profile of Np-2 differed significantly from the one of Np-1. One interesting feature was a quite sharp redox zone in the center of the sample, where increased amounts of Np(IV) were detected. Probably, higher C amounts of a Np(V) reductant such as Fe were present. A second feature was a decrease of the 237 Np profile towards both edges, showing that a 237 Np out-diffusion took place in downstream as well as upstream direction. Various hypotheses, which would lead to the observed Np diffusion profiles in each cell, will be tested with reactive transport simulations using surface complexation models to quantify sorption as a function of the geochemical conditions.

The comparison of Np-1 and Np-2 showed that both were in line in regard to the decrease of 237 Np(V) concentration in the source reservoir solutions with progressing experiment. However, the evolution of 237 Np(V) activity in each receiving reservoir solution was different. From both experiments, similar De values of about 1·10 -10 m2/s were deduced. However, the Kd values were different with Kd(Np-1) of 0.1±0.02 m3/kg and Kd(Np-2) of 0.06±0.005 m3/kg. This range of Kd values is in agreement with the findings of Schacherl et al., who conducted batch sorption experiments with Illite du Puy in 0.1 mol/L NaCl.

The Np(V)/Np(IV) ratio in the Illite du Puy samples ranged from 0.55 to 6 for Np-1 and from 1.8 to 13 for Np-2 with increased Np(IV) levels especially present at the edges of the clay samples. Schacherl et al. detected a partial reduction of Np(V) to Np(IV), particularly at pH 5 in batch sorption experiments under anaerobic conditions. These observations indicate a partial reduction of Np(V) to Np(IV) during the diffusion experiment with Illite du Puy even when connected to diffusion reservoir solutions under aerobic conditions.

This study delivers valuable long-term data on redox-sensitive radionuclide migration through a clay mineral present in clay host rocks and provides a contribution to safety assessment models.