Search

Many countries intend to achieve the safe management of their radioactive wastes through geological disposal. In addition, radioactively contaminated land is of global concern. To address both of these technical challenges it is imperative to understand the behaviour and subsequent migration of radionuclides in the subsurface. This thesis addresses uncertainties in the behaviour of the long-lived, risk-driving radionuclides U and Np in their most mobile and environmentally relevant oxidation states, U(VI) and Np(V). The formation the U(VI) colloidal nanoparticles is identified under the high pH, low carbonate conditions expected within the near field of a cementitious Geological Disposal Facility (GDF). XAS, SAXS, and TEM have been used to characterise these U(VI) colloids as 60-80 nm clusters of 1-2 nm clarkeite-like (Na uranate) nanoparticles, which are stable in cement leachate for a period of at least 5 years. The reactivity of these U(VI) colloids towards a range of mineral phases was investigated. In the presence of the common rock-forming minerals biotite, orthoclase, and quartz, only limited reactivity was observed with > 80 % of the U(VI) remaining in the filtered fraction after up to 5 years of reaction. In contact with cement, > 97 % of the U(VI) was removed from solution within 1 month. Reversibility studies, luminescence spectroscopy, and XAS suggest that a large portion of the cement associated U(VI) is in a uranophane-like coordination environment, likely incorporated into the C-S-H interlayers or as a stable surface precipitate. Together, this suggests that while U(VI) colloids could form in high pH (> 13) cement leachate, providing an additional pathway for migration, many of them are likely to be removed from suspension by the presence of solid cement, although 2.4 % (1.0 IμM) U(VI) remained in the filtered fraction even after 21 months of reaction. The interaction of aqueous U(VI) and Np(V) with a range of environmentally relevant Mn minerals has also been studied under circumneutral to alkaline conditions. Here, extensive (up to 99 %) uptake of U(VI) and Np(V) was observed in systems containing δ-Mn(IV)O2, triclinic (Na)-birnessite [Na0.5Mn(IV/III)2O4 · 1.5H2O], hausmannite [Mn(III/II)3O4], and rhodochrosite [Mn(II)CO3]. The uptake of U(VI) by δ-MnO2 and hausmannite was found to be partially irreversible, suggesting that these minerals could be particularly important in determining radionuclide migration. XAS indicated that both U(VI) and Np(V) formed edge-sharing bidentate adsorption complexes on the surface of δ-MnO2 and hausmannite, implying that these complexes are responsible for the observed reversibility. These complexes were also identified on triclinic (Na)-birnessite; however, after 1 month of reaction U(VI) was found to have migrated into the triclinic (Na)-birnessite interlayer, replacing Na+. Reaction with all three investigated Mn oxide phases was rapid, with equilibrium being reached within at least 2 weeks. However, whilst U(VI) and Np(V) were both extensively removed from solution in systems containing rhodochrosite, these reactions were much slower, with equilibrium taking up to 4 months to be established. XAS suggested that this was due to the formation of a U(VI) or Np(V) containing precipitate on the rhodochrosite surface.

Exposure to vanadium (V) results in adverse health outcomes in humans and a plethora of species in the environment across trophic levels. V is released from rock, slag and mine tailings, where it is able to oxidize to vanadate (VV) (HnVO4 −3+n), which poses the greatest risk to human health. Vanadate is highly soluble unlike its VIII or VIV, and is readily taken up by well systems, resulting in human exposure in areas where water treatment is minimal. However, the geochemical controls that determine the mobility of V in the environment are poorly understood, making the behavior of V in the subsurface difficult to predict. This dissertation addresses this knowledge gap by investigating the abiotic reactions of V with common iron and manganese mineral phases. These mineral phases are already known to play an important role in the fate and transport of similar contaminants, but many aspects of their specific reactivity towards V are unknown. To address this knowledge gap, their surface capacity, adsorption affinity and their ability to retain vanadate via surface complexation is investigated. The vanadate adsorption capacity and surface affinity are both found to be inversely proportional to the crystallinity of the oxide examined. One manganese oxide, birnessite is also known to promote the oxidation of many trace metal contaminants. The oxidative capacity of birnessite towards VIV is much greater than its surface capacity for VV. At the aggregate scale, birnessite, V and Fe oxides coexist in close spatial proximity where rates of diffusion limit the transport of solutes. In such an environment, birnessite effectively oxidizes VIV and retains more VV than adjacent Fe oxide phases. Further, FeII-bearing oxides can reduce VV back to VIV. However, this pathway is slow compared to the oxidation rate of VIV by birnessite. This suggests that surface complexation of vanadate, even by birnessite is a greater attenuation pathway than abiotic FeII-mediated reduction. The result of this research emphasizes the role of abiotic surface processes in constraining the mobility of V within the soil matrix which will improve the reliability of models implemented in the management and reclamation of V contaminated sites.

The risk posed from incidental ingestion of arsenic-contaminated or lead-contaminated soil may depend on sorption of arsenate (As(V)) or lead (Pb(II)) to oxide surfaces in soil. Arsenate or lead sorbed to ferrihydrite, corundum, and birnessite model oxide minerals were used to simulate possible effects of ingestion of soil contaminated with As(V) or Pb(II). Arsenate or lead sorbed oxides were placed in a simulated gastrointestinal tract (in vitro) to ascertain the bioaccessibility of As(V) or Pb(II) and changes in As(V) or Pb(II) surface speciation. The speciation of As or Pb was determined using EXAFS and XANES analysis. The As(V) adsorption maximum was found to be 7.04 g kg-1, and 0.47 g kg-1 for ferrihydrite and corundum, respectively. The bioaccessible As(V) for ferrihydrite ranged form 0 to 5 % and for corundum ranged from 0 to 16 %. The surface speciation for ferrihydrite and corundum was determined to be binuclear bidentate. These results for As(V) sorbed to ferrihydrite and corundum suggest that the bioaccessibility of As(V) is related to the As(V) concentration, and the As(V) adsorption maximum. The bioaccessibility of As(V) sorbed to birnessite ranged from 16 to 28 % and the As(V) surface speciation was binuclear bidentate. The results for As (V) sorbed birnessite suggest that birnessite can reduce the bioaccessibility of As(V) in contaminated soil. Sorption of Pb(II) was significant for all the model oxides used. The sorption maximum was found to be 2.13 g kg-1 for corundum, 38.6 g kg-1 for ferrihydrite, and 127 g kg-1 for birnessite. The bioaccessible Pb(II) ranged from 53 to 88 % for ferrihydrite, Pb(II) concentrations greater than 200 mg kg-1 was greater than 85 % for corundum and below 200 mg kg-1 there was no bioaccessible Pb(II). There was no measurable bioaccessible Pb(II) for any concentration of Pb sorbed birnessite. The surface Pb(II) speciation was found to be: 1. Mononuclear bidentate for ferrihydrite, 2. Mononuclear bidentate and an outer-sphere complex for corundum, and 3. Binuclear bidentate, and trinuclear tridentate for birnessite. Ferrihydrite and corundum usefulness in Pb remediation would be limited when risk of incidental ingestion is of major concern.