Interaction of uranium(VI) and colloidal iron(III) oxyhydroxide in an abandoned uranium mine

Acid mine drainage and process water from the in-situ leaching by sulphuric acid applied on uranium mines usually contain high amounts of iron, uranium, and sulphate besides other elements such as aluminium, manganese, zinc, calcium and silicon. Flooding of abandoned U mines by shallow groundwater initiates de-acidification, oxidation, and hydrolysis reactions by which colloids are formed, dominated by X-ray amorphous Fe(III) oxyhydroxide phases. Previous investigation on a flooding scenario has given spectroscopic evidence that the high affinity of U(VI) onto this type of colloids is due to the formation of mononuclear, inner-sphere surface complexes on 2-line ferrihydrite (Fh) where we suppose two types of coordination: (i) edge-sharing linkage of the UO22+ cation to an Fe(O,OH)6 octahedron stabilized by a hydrogen bond between one axial oxygen atom and the Fh surface, occurring in absence of CO2 and other ligands, and (ii) double-corner sharing linkage of the UO22+ cation to two Fe(O,OH)6 octahedrons, occurring in the presence of atmospheric CO2 at pH ≤ 8 at the expense of the type (i) surface complex owing to competitive adsorption of carbonate on the Fh surface. Whereas a ternary sorption complex with monodentate linkage of carbonate to the adsorbed UO22+ cannot be ruled out at the tested conditions, ternary complexes with other ligands such as sulphate, nitrate, and silicate could be excluded. The retarding impact of colloids on the transport of U(VI) does not only depend on the structure, but also on the long-term stability of the covalent bond between U(VI) and the solid phase. We started a series of desorption experiments in a simulated groundwater medium and a dilute NaHCO3 solution to study the kinetics of U(VI) desorption from U-loaded Fh. On a longer time scale of decades to centuries, the transformation of Fh to more crystalline phases such as hematite and goethite must be considered as well. We use the common approach of gently heating the samples to accelerate the crystallization of Fh in order to quantify the amount of U(VI) that will be incorporated in the lattice during the aging process as shown for instance by Duff et al. Recent results of these experiments will be presented. In conclusion, our results show that the interaction of U(VI) and the highly reactive surface of ferric oxyhydroxide colloids plays a major role in lowering or delaying the release of radiotoxic uranium from flooded uranium mines into adjacent groundwater aquifers or surface water