Geological formations are being considered as host media for nuclear waste disposal. The occurrence of natural U in rocks provides a possibility to test the radionuclide migration models used in safety studies of the disposal over comparable time periods. Here we study U accumulation into boulders as a process analogue for matrix diffusion; the boulders were found in glacial till in Hämeenlinna, southern Finland. Based on the glacial history of the site, matrix diffusion simulations, and independent U-series disequilibria (USD) dating, the U accumulation was interpreted to originate from the end stage of the latest glaciation, i.e. the system age is about 10 000 years1,2. The known time scale offers a rare opportunity for quantitative model testing; normally the time scale is difficult to determine for a single process in a natural analogue.
The U accumulation was earlier1,2 interpreted to be due to matrix diffusion and sorption. The postulated accumulation history consists of short in-diffusion and out-diffusion stages, as well as a longer chain decay stage. The in-diffusion was caused by U-rich waters discharging on the boulders at the end stage of the glaciation. The subsequent partial out-diffusion represents the period the boulders were temporarily submerged in the Yoldia sea during the early stage of the Baltic Sea. The final isolated radioactive chain decay stage began when the boulders, and their surroundings, rose above the sea level due to land uplift.
In this paper we report the first radiochemical results of a new larger boulder from the same area as the one studied earlier1; qualitatively, also the U distribution appears to be the same. Due to the larger dimensions, we can sample the inner zone of the boulder which matrix diffusion can not have reached within the postulated time, i.e. the state of the boulder before the U accumulation. The large amount of sample material containing almost only the recently accumulated U provides an opportunity to experimentally approach the kinetics of U fixation in situ. Understanding the long-term U fixation is essential in natural analogue studies, because the matrix diffusion model only has fast reversible adsorption (based on Kd) as the fixation process. Attempts to separate and quantify sorbed U in natural analogues have been reported elsewhere3.