An experimental and ab initio study on the abiotic reduction of uranyl by ferrous iron

Abstract

It is important to understand the mechanisms controlling the removal of uranyl from solution from an environmental standpoint, particularly whether soluble Fe(II) is capable of reducing soluble U(VI) to insoluble U(IV). Experiments were performed to shed light into discrepancies of recent studies about precipitation of U-containing solids without changing oxidation states versus precipitation/reduction reactions, especially with respect to the kinetics of these reactions. To understand the atomistic mechanisms, thermodynamics, and kinetics of these redox processes, ab initio electron transfer (ET) calculations, using Marcus theory, were applied to study the reduction of U(VI)aq to U(V)aq by Fe(II)aq (the first rate-limiting ET-step). Outer-sphere (OS) and inner-sphere (IS) Fe–U complexes were modeled to represent simple species within a homogeneous environment through which ET could occur.Experiments on the chemical reduction were performed by reacting 1mM Fe(II)aq at pH 7.2 with high (i.e., 0.16mM) and lower (i.e., 0.02mM) concentrations of U(VI)aq. At higher U concentration, a rapid decrease in U(VI)aq was observed within the first hour of reaction. XRD and XPS analyses of the precipitates confirmed the presence of (meta)schoepite phases, where up to ∼25% of the original U was reduced to U4+ and/or U5+-containing phases. In contrast, at 0.02mM U, the U(VI)aq concentration remained fairly constant for the first 3h of reaction and only then began to decrease due to slower precipitation kinetics. XPS spectra confirm the partial chemical reduction U associated with the precipitate (up to ∼30%). Thermodynamic calculations support that the reduction of U(VI)aq to U(IV)aq by Fe(II)aq is energetically unfavorable. The batch experiments in this study show U(VI) is removed from solution by precipitation and that transitioning to a heterogeneous system in turn enables the solid U phase to be partially reduced.Ab initio ET calculations revealed that OS ET is strongly kinetically inhibited in all cases modeled. OS ET as a concerted proton-coupled ET reaction (ferrimagnetic spin configuration) is thermodynamically favorable (−35kJ/mol), but kinetically inhibited by concurrent proton-transfer (10−19s−1). OS ET as a sequential proton-coupled ET reaction is thermodynamically unfavorable (+102kJ/mol) as well as kinetically inhibited, where ET is the rate-limiting step (10−12s−1). In contrast, the reduction of U(VI)aq to U(V)aq by Fe(II)aq as an IS ET reaction is both thermodynamically favorable (−16kJ/mol) and kinetically rapid (108s−1); the IS ET rate is several orders of magnitude faster than the OS ET rate. Thus, reduction of U(VI)aq to U(V)aq by Fe(II)aq in a homogenous system could occur if an IS Fe–U complex can be achieved. However, the formation of IS Fe–U complexes in an homogeneous solution is predicted to be low; considerable thermodynamic and kinetic barriers exist to proceed from an OS ET reaction to an IS ET reaction, a process that needs to overcome dehydration of the first solvation shell (+96kJ/mol) and hydrolysis of Fe(II)aq. The computational results complement and further substantiate experimental results where the reduction of U(VI)aq by Fe(II)aq does not occur.