Assessing the bioavailability of the radionuclides technetium-99, selenium-79 and uranium-238 in contaminated soils using the diffusive gradients in thin-films (DGT) technique

Abstract

The increasing global inventory of 99Tc, 79Se and 238U as components of nuclear waste, coupled with the legacy of contaminated sites, means that increasing attention has been centred on understanding the environmental behaviour of these radionuclides. Over the last 15 years, the technique of diffusive gradients in thin-films (DGT) has emerged as a promising tool for assessing bioavailability based on its demonstrated ability to successfully predict plant uptake for a range of metals and metalloids in soil (Zhang and Davison, 2015). It is therefore utilised in this thesis to address a substantial knowledge gap regarding the bioavailability and aging of Tc, Se and U in soils. Initially, the performance of a Chelex-ferrihydrite mixed binding layer (MBL) DGT for the novel simultaneous measurement of Se and U was investigated to validate its suitability for subsequent use. An assessment of the availability and aging of all three elements was centred on an 18-month laboratory incubation of a suite of spiked soils, throughout which time a series of DGT deployments were made. Lastly, the ability of DGT to predict ryegrass uptake of Tc across a range of soil types was investigated. The experimental work presented in this thesis reveals that the MBL DGT is an acceptable method for determining labile Se and U in soil, although at higher pH (> 7) and with increasing concentrations of HCO3 - , uptake of both elements was impaired. The availability and aging of Se and Tc within soil is governed by soil organic carbon, in addition to the Al and Fe oxide content for Se. Quantitatively, the aging of Tc could be best described by a pseudo-secondorder model, yet a natural exponential function provided the best fit for Se. U was found to be particularly resilient to aging within soils exposed to a fluctuating wet-dry moisture regime, where changes in pH were hypothesised to alter the solubility of key U binding phases, namely Fe and Al oxides and dissolved carbonate ligands. DGT was unable to reliably predict the extent of uptake of Tc within ryegrass across a range of soil types, although in situ deployments whilst the ryegrass was actively growing did yield encouraging results. Remedial efforts to reduce the soil-to-plant radionuclide transfer following a contamination event should look to utilise the complexing capacity of Corg in the case of Tc and Se, or prevent the exposure of the soil to wetting/drying events in the case of U. Where possible, DGT measurements of bioavailability within Tc-contaminated soils should be made in situ, although in a field setting this is logistically challenging and introduces significant uncertainties to the data. In practice, to reliably ascertain the bioavailability of Tc, it would seem that the best approach is to directly measure the accumulated activity within the plant.