U(VI) Adsorption to Heterogeneous Subsurface Media:  Application of a Surface Complexation Model

Abstract

The pH-dependent adsorption of U(VI) onto three heterogeneous, subsurface media from the Department of Energy (DOE) Oak Ridge Reservation, Savannah River Site, and Hanford Reservation was investigated. The three materials contained significant quantities of iron and manganese oxides with nearly identical extractable iron oxide contents (25.3-25.8 g/kg). A model independently developed for the adsorption of U(VI) to synthetic ferrihydrite (Waite, T. D.; Davis, J. A.; Payne, T. E.; Waychunas, G. A.; Xu, N. Geochim. Cosmochim. Acta 1994, 58, 5465-5478) was able to predict the major features of the pH-dependent U(VI) adsorption to the materials under the assumption that all the dithionite-citrate-bicarbonate extractable iron oxide was present as ferrihydrite. Further experiments with the Oak Ridge soil as a function of carbonate and U(VI) concentration indicated that the model could predict pH-dependent U(VI) adsorption to within a root mean square error of 0.163-0.408, even under conditions outside of those for which the model was developed. These results indicate that this model could be used as a first approximation in predicting U(VI) adsorption and transport in the subsurface. U(VI) adsorption also decreased at pH >10, even in the absence of carbonate, which is of potential importance to U(VI) mobility in extreme environments present in the subsurface at some DOE facilities. The pH-dependent adsorption of U(VI) was fundamentally different in systems with a constant CO2 partial pressure as compared to a constant total carbonate concentration. Experiments at constant CO2 partial pressure may not be representative of the conditions present in the subsurface, and a constant carbonate concentration does not always result in decreased U(VI) adsorption at higher pH values.

Sulfur Isotopes as Indicators of Amended Bacterial Sulfate Reduction Processes Influencing Field Scale Uranium Bioremediation

“…though sulfide concentrations and sulfur isotopes indicate increasing BSR activity. This suggests that maximum uranium removal may have been achieved in association with FeS precipitation (44), a connection supported by the sulfur isotope data, though the influence of uranium desorption as a consequence of elevated pH could contribute to the subsequent rebound in U(VI) concentrations during sulfate reduction (45,46) (29), suggesting that the majority of the sulfate in the system was reduced in the up-gradient area between the injection gallery and M-21. Coupled with hydrologic data suggesting preferential flow paths across the well gallery, measurable concentrations of sulfate in M-21 (average 1.2 mM) and M-24 (average 2.3 mM) during this time are thus attributed to a combination of fast flow paths in which sulfate can pass unaffected through the reduced zone and mixing of groundwater external to the reduced zone through the sampling process.…”

Hexavalent Uranium Diffusion into Soils from Concentrated Acidic and Alkaline Solutions

“…The presence of calcite is predicted to have a strong influence on U solubility in the AL systems because of Ca 2 UO 2 (CO 3 ) 3 formation, based on its reported stability (8), but the magnitude of this effect remains to be determined (27). In the initially acidic solutions, calcite rapidly buffered the reservoir to pH ≈ 8, and Ca 2+ levels are predicted to enhance U(VI) solubility through formation of aqueous Ca 2 UO 2 (CO 3 ) 3 . Similar effects apply for the alkaline U solution exposed to the AL soil, although slightly high equilibrium pH and presumably slightly lower Ca 2+ concentrations are attained.…”

“…Although exposed to initially pH 2 or pH 11 U(VI) solutions, strong buffering by the soils (especially the calcareous AL soils) resulted in diffusion at nearly constant pH after short neutralization periods. Diffusion of U(VI) into the calcareous AL soil is relatively efficient regardless of the initial solution pH, probably because of the stability of aqueous Ca 2 UO 2 (CO 3 ) 3 and consequent weaker sorption in these calcite-buffered systems. Diffusion into the OR soil under acidic conditions was also significantly faster than expected from sorption data obtained at lower concentrations.…”

“…The D o value for UO 2 2+ is applicable in dilute, acidic (pH < 5) solutions where it is the dominant aqueous U(VI) species. The generic uranyl carbonate D o will be used in the neutral to alkaline range, to represent aqueous uranyl carbonates, (UO 2 ) 2 (OH) 3 3 are all assigned a common value of D o , our diffusion calculations will be insensitive to uncertainties associated with speciation discussed later. The effective diffusivity of a solute in a saturated porous medium, D e , is usually only associated with transport in the water-filled pores.…”

Mathematical Formulation Requirements and Specifications for the Process Models

“…Based on a measurement of the relative amounts or surface areas of each mineral present in the soil or sediment, sorption by the mixture of phases can be predicted by an equilibrium calculation, without any fitting of experimental data for the mixture. CA model predictions are sometimes made by assuming that one mineral component dominates sorption (Barnett et al, 2002;Davis et al, 2004;Payne et al, 2004;Zhang et al, 2009), allowing a straightforward equilibrium calculation, if the exposed surface area of that mineral component in the soil or sediment can be quantified.…”