Model-based analysis of mixed uranium(VI) reduction by biotic and abiotic pathways during in situ bioremediation

•Mixed U(VI) reduction by biological and chemical pathways was integrated into a model.•The biotic and abiotic U(VI) reduction was investigated under realistic geologic systems.•Thermodynamics was integrated into a model of uranium reduction and immobilization.•Model-based analysis can formulate/confirm hypotheses for experimental investigations.•The approach offers a potential solution to an in-depth analysis of in situ bioremediation.

Given the numerous unknowns and uncertainties in sediment systems, understanding of the mechanisms of U(VI) reduction is still at the stage of improvement. Recent studies have shown that reductive immobilization of U(VI) in the subsurface is not caused by purely biological or purely abiotic reactions but rather a group of interconnected abiotic-biotic pathways (e.g. via mackinawite and biomass). These new findings necessitate an update of the existing mathematical models that make simplifications typically involving a single reducing agent (e.g. indigenous bacteria) for in situ bioremediation of uranium-contaminated groundwater. In this study, a comprehensive model was constructed based on new experimental observations, including mixed U(VI) reduction by chemical and enzymatic reactions. Thermodynamic analysis was done to predict the feasibility of the potential pathways that affect mackinawite formation under field conditions. Model simulations indicate that low concentrations of the reactant species make the reaction of homogeneous Fe(II) oxidation coupled to U(VI) reduction unfavorable in the field. Instead, FeS precipitation is an important Fe(II) sequestration reaction once sulfate metabolism dominates. The subsequent reduction of U(VI) by FeS (mackinawite) contributes to the total U(VI) removal under a variety of U(VI) concentrations encountered at field sites, which is in accordance with experimental observations. The model suggests the potential for both competition and coordination between chemical and biological pathways on the cell surface, providing a possible explanation as to why U(VI) can be efficiently reduced at either low or high sulfate concentration during the process of in situ bioremediation. Further increase in the resolution of the model (e.g. across multiple scales such as genome-, micron- and pore-sale) is necessary for better understanding of the interactions between biotic and abiotic pathways.