Radiostrontium sorption on natural glauconite sands of the Neogene-Paleogene formations in Belgium.

The Neogene-Paleogene glauconite sands are investigated for radionuclide sorption in the framework of the Belgian radioactive waste disposal program. This study was set up to measure the adsorption of radiostrontium (85Sr) on the sands and on glauconite fractions to identify factors explaining variable sorption among different formations. Batch 85Sr sorption experiments were set up with 45 different glauconite sands and glauconite fractions (125-250 µm) in a background solution of 1 mM CaCl2.H2O and 0.5 mM KCl. The distribution coefficients (KD) for 85Sr2+ ranged 23-65 L kg-1 for the intact sands and ranged 50-144 L kg-1 for the glauconite fractions. The KD values strongly correlated with the CEC (R2 = 0.62 for sands and 0.82 for glauconite fractions) and corresponded well with CEC based predictions based on two existing models calibrated to soils. The KD on the complete sand is proportional to the glauconite content and the KD of the glauconite fraction if no other clay minerals are present in significant amounts. Sorption equilibrium was reached within 48 h in the complete sands, in milled complete sands, in glauconite fractions and in milled glauconite fractions, suggesting no diffusive boundaries in the glauconite pellets. It is concluded that glauconite sands have a suitably high retention of radiostrontium and the sorption strength is in line with that of other geological barriers when judged from the CEC.

Radiocaesium sorption on natural glauconite sands is unexpectedly as strong as on Boom Clay.

The Neogene-Paleogene glauconite sands of Belgium cover the Boom Clay deposits that are candidate host for radioactive waste disposal. It is unclear if the highly permeable sand formations may act as an additional barrier for radiocesium (137Cs) or could be added as a complementary sorption sink in a surface disposal concept. Glauconite is an Fe-rich phyllosilicate that is mainly present as 250-125 µm sized pellets in sand, it is unknown to what extent and how fast these pellets may bind 137Cs. Pelletized clays embedded in sand may have poorly accessible high affinity sites for 137Cs. The 137Cs sorption on 11 different glauconite sands was measured in batch in a background solution of 0.1 M CaCl2 and 0.5 mM KCl. The log transformed 137Cs distribution coefficient Kd (L kg-1) after 30 days reaction ranged 3.4-4.3, surprisingly close to the Kd of the Boom Clay (3.5). Isolated glauconite fractions exhibited similar 137Cs sorption potentials (log Kd 4.1-4.3) as the reference Illite du Puy (4.4). The small Kd variation among the Neogene-Paleogene sands was explained by its glauconite content (r = 0.82). The 137Cs sorption kinetics (1-57 days) of milled pellets (<2 µm) confirmed slower reaction with intact pellets than with milled samples. Additionally, the Kd values of milled samples (57 days) sorption are 1.1-1.5 fold larger than the corresponding intact pellets, suggesting that not all Cs binding sites are accessible in intact pellets. Strongly weathered pellets exhibited cracks (visible with SEM). In these pellets the Kd was similar for milled and intact pellets suggesting that cracks increase the accessibility of the inner sorption sites. After 8.5 months the Kd values were 1.6-1.8-fold above corresponding 1 month data and these long-term reactions were more pronounced as total sand K content was larger. An adsorption-desorption experiment illustrated that 137Cs sorption is not fully reversible.