Search

Reduction of hexavalent uranium (U(VI)) to less soluble tetravalent uranium (U(IV)) through enzymatic or abiotic redox reactions has the potential to alter U mobility in subsurface environments. As a ubiquitous natural mineral, magnetite (Fe3O4) is of interest because of its ability to act as a rechargeable reductant for U(VI). Natural magnetites are often impure with titanium, and structural Fe(3+) replacement by Ti(IV) yields a proportional increase in the relative Fe(2+) content in the metal sublattice to maintain bulk charge neutrality. In the absence of oxidation, the Ti content sets the initial bulk Fe(2+)/Fe(3+) ratio (R). Here, we demonstrate that Ti-doped magnetites (Fe3 - xTixO4) reduce U(VI) to U(IV). The U(VI)-Fe(2+) redox reactivity was found to be controlled directly by R but was otherwise independent of Ti content (xTi). However, in contrast to previous studies with pure magnetite where U(VI) was reduced to nanocrystalline uraninite (UO2), the presence of structural Ti (xTi = 0.25-0.53) results in the formation of U(IV) species that lack the bidentate U-O2-U bridges of uraninite. Extended X-ray absorption fine structure spectroscopic analysis indicated that the titanomagnetite-bound U(IV) phase has a novel U(IV)-Ti binding geometry different from the coordination of U(IV) in the mineral brannerite (U(IV)Ti2O6). The observed U(IV)-Ti coordination at a distance of 3.43 Å suggests a binuclear corner-sharing adsorption/incorporation U(IV) complex with the solid phase. Furthermore, we explored the effect of oxidation (decreasing R) and solids-to-solution ratio on the reduced U(IV) phase. The formation of the non-uraninite U(IV)-Ti phase appears to be controlled by availability of surface Ti sites rather than R. Our work highlights a previously unrecognized role of Ti in the environmental chemistry of U(IV) and suggests that further work to characterize the long-term stability of U(IV) phases formed in the presence of Ti is warranted.

Hexavalent uranium (U(VI)) can be reduced enzymatically by various microbes and abiotically by Fe(2+)-bearing minerals, including magnetite, of interest because of its formation from Fe(3+) (oxy)hydroxides via dissimilatory iron reduction. Magnetite is also a corrosion product of iron metal in suboxic and anoxic conditions and is likely to form during corrosion of steel waste containers holding uranium-containing spent nuclear fuel. Previous work indicated discrepancies in the extent of U(VI) reduction by magnetite. Here, we demonstrate that the stoichiometry (the bulk Fe(2+)/Fe(3+) ratio, x) of magnetite can, in part, explain the observed discrepancies. In our studies, magnetite stoichiometry significantly influenced the extent of U(VI) reduction by magnetite. Stoichiometric and partially oxidized magnetites with x ≥ 0.38 reduced U(VI) to U(IV) in UO(2) (uraninite) nanoparticles, whereas with more oxidized magnetites (x < 0.38) and maghemite (x = 0), sorbed U(VI) was the dominant phase observed. Furthermore, as with our chemically synthesized magnetites (x ≥ 0.38), nanoparticulate UO(2) was formed from reduction of U(VI) in a heat-killed suspension of biogenic magnetite (x = 0.43). X-ray absorption and Mössbauer spectroscopy results indicate that reduction of U(VI) to U(IV) is coupled to oxidation of Fe(2+) in magnetite. The addition of aqueous Fe(2+) to suspensions of oxidized magnetite resulted in reduction of U(VI) to UO(2), consistent with our previous finding that Fe(2+) taken up from solution increased the magnetite stoichiometry. Our results suggest that magnetite stoichiometry and the ability of aqueous Fe(2+) to recharge magnetite are important factors in reduction of U(VI) in the subsurface.