Iron Reduction and Radionuclide Immobilization: Kinetic, Thermodynamic and Hydrologic controls & Reaction-Based Modeling - Final Report

Abstract

Our research focused on (1) microbial reduction of Fe(III) and U(VI) individually, and concomitantly in natural sediments, (2) Fe(III) oxide surface chemistry, specifically with respect to reactions with Fe(II)and U(VI), (3) the influence of humic substances on Fe(III) and U(VI) bioreduction, and on U(VI) complexation, and (4) the development of reaction-based reactive transport biogeochemical models to numerically simulate our experimental results. We have continued our investigations on microbial reduction of Fe(III) oxides. Modeling our earlier experimental results required assumption of a hydrated surface for hematite, more reactive than predicted based on theoretical solubility (Burgos et al.2002). Subsequent studies with Shewanella putrefaciens and Geobacter sulfurreducens confirmed the rates of Fe(III) bioreduction depend on oxide surface area rather than oxide thermodynamic properties (Roden,2003a,b;2004; Burgos et al,2003). We examined the potential for bioreduction of U(VI) by Geobacter sulfurreducens in the presence of synthetic Fe(III) oxides and natural Fe(III) oxide-containing solids (Jeon et al,2004a,b) in which more than 95% of added U(VI) was sorbed to mineral surfaces. The results showed a significant portion of solid-associated U(VI) was resistant to both enzymatic and abiotic (Fe(II)-driven) reduction, but the rate and extent of bioreduction of U(VI) was increased due to the addition of anthraquinone-2,6-disulfonate (AQDS). Wemore » conducted long-term semicontinuous culture and column experiments on coupled Fe(III) oxide/U(VI) reduction. These experiments were conducted with natural subsurface sediment from the Oyster site in Virginia, whose Fe content and microbial reducibility are comparable to ORNL FRC sediments (Jeon et al, 2004b). The results conclusively demonstrated the potential for sustained removal of U(VI) from solution via DMRB activity in excess of the U(VI) sorption capacity of the natural mineral assemblages. Jang (2004) demonstrated that the hydrated surface of nano-particles of hematite (prepared according to well-cited recipes and confirmed to be 100% hematite by Mossbauer spectroscopy and XRD) exhibited the solubility of hydrous ferric oxide (HFO). Jang (2004) also demonstrated that the sorptive reactivity of hematite and HFO were identical except for different specific surface area and pHzpc, and that the reduction of U(VI) by sorbed Fe(II) in the presence of the two phases was also similar in spite of theoretical predictions of large differences in Nernst potential. These results are consistent with the modeling of hematite bioreduction experiments where the thermodynamic potential of hematite had to be adjusted to represent a more disordered surface phase in order to accurately model bioreduction kinetics (Burgos et al, 2002, 2003). We have demonstrated that humic substances enhance solid-phase Fe(III) bioreduction via both electron shuttling and Fe(II) complexation(Royer et al, 2002a, b). We have found that humic substances were shown to inhibit the bioreduction of dissolved U(VI) and that soluble humic-U(IV) complexes were likely formed (Burgos et al, 2004). Kirkham (2004) measured and modeled complexation of U(VI) by humic substances as a function of pH, pCO2, U(VI) concentration, and humic concentration, and demonstrated that humic substances can complex U(VI) even at neutral pH values and in the presence of high (ca.30 mM) carbonate concentrations. Jang(2004) measured the abiotic reduction of U(VI) by Fe(II) sorbed to Fe(III) oxides in the presence/absence of humic substances and demonstrated that humic substances inhibited the heterogeneous reduction of U(VI). We have recently developed, validated, and documented a series of diagonalized reaction-based reactive transport computer models (HYDROGEOCHEM; Yeh et al,2004a,b). We demonstrated that parallel kinetic reactions could be modeled if separate experiments are used to independently measure each contributing kinetic reaction (Burgos et al, 2003). We have demonstrated the use of a reaction-based reactive transport model (HYDROGEOCHEM) for the simulation of biological iron reduction in natural sediment columns (Burgos and Yeh, unpublished results). Finally, we have developed a preliminary reaction-based model of coupled Fe(III) oxide/U(VI) reduction that has been employed in numerical simulations of U(VI) bioreduction in bench-scale (Roden, 2003d) and field-scale (Roden and Scheibe, 2003;Roden, 2003c) systems.« less