Long-term dynamics of uranium reduction/reoxidationunder low sulfate conditions

The biological reduction and precipitation of uranium in groundwater has the potential to prevent uranium migration from contaminated sites. Although previous research has shown that uranium bioremediation is maximized during iron reduction, little is known on how long-term iron/uranium reducing conditions can be maintained. Questions also remain about the stability of uranium and other reduced species after a long-term biostimulation scheme is discontinued and oxidants (i.e., oxygen) re-enter the bioreduced zone. To gain further insights into these processes, four columns, packed with sediment containing iron as Fe-oxides (mainly Al-goethite) and silicate Fe (Fe-containing clays), were operated in the laboratory under field-relevant flow conditions to measure the long-term (>200 day) removal efficiency of uranium from a simulated groundwater during biostimulation with an electron donor (3 mM acetate) under low sulfate conditions. The biostimulation experiments were then followed by reoxidation of the reduced sediments with oxygen. During biostimulation, Fe(III) reduction occurred simultaneously with U(VI) reduction. Both Fe-oxides and silicate Fe(III) were partly reduced, and silicate Fe(III) reduction was detected only during the first half of the biostimulation phase while Fe-oxide reduction occurred throughout the whole biostimulation period. Mössbauer measurements indicated that the biogenic Fe(II) precipitate resulting from Fe-oxide reduction was neither siderite nor FeS0.09 (mackinawite). U(VI) reduction efficiency increased throughout the bioreduction period, while the Fe(III) reduction gradually decreased with time. Effluent Fe(II) concentrations decreased linearly by only 30% over the final 100 days of biostimulation, indicating that bioreducible Fe(III) in the sediment was not exhausted at the termination of the experiment. Even though Fe(III) reduction did not change substantially with time, microorganisms not typically associated with Fe(III) and U(VI) reduction (including methanogens) became a significant fraction of the total microbial population during long-term biostimulation, meaning that most acetate was utilized for biological processes other than Fe(III) and U(VI) reduction. This corresponds with an electron donor/acceptor mass balance showing that the amount of Fe(III), U(VI) and SO4 2 reduced accounted for very little (<2%) of the acetate consumed after day 104 of bioreduction. Selected columns were reoxidized after 209 days by discontinuing acetate addition and purging the influent media with a gas containing 20% oxygen. Uranium reoxidation occurred rapidly with a very large uranium spike exiting the column (7–8 times higher than the original influent concentration) which resulted in 61% of the precipitated uranium resolubilized and transported out of the column after 21 days and virtually all of the uranium being removed by day 122. During the first 21 days of reoxidation, the Fe(III) and U(VI) reducing microbial population, as measured by quantitative PCR, remained at pre-oxidation levels (even though the gene transcripts that represent the methanogen population decreased by 99%) indicating that short-term disruptions in biostimulation (equipment failure, etc.) may not negatively affect the uranium reducing microbial population. 2008 Elsevier Ltd. All rights reserved. 0016-7037/$ see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.gca.2008.05.040 * Corresponding author. Fax: +1 609 258 2799. E-mail address: [email protected] (P.R. Jaffé). www.elsevier.com/locate/gca Available online at www.sciencedirect.com Geochimica et Cosmochimica Acta 72 (2008) 3603–3615