Reduction of U(VI) in goethite (α-FeOOH) suspensions by a dissimilatory metal-reducing bacterium

Abstract

Dissimilatory metal-reducing bacteria (DMRB) can utilize Fe(III) associated with aqueous complexes or solid phases, such as oxide and oxyhydroxide minerals, as a terminal electron acceptor coupled to the oxidation of H₂ or organic substrates. These bacteria are also capable of reducing other metal ions including Mn(IV), Cr(VI), and U(VI), a process that has a pronounced effect on their solubility and overall geochemical behavior. In spite of considerable study on an individual basis, the biogeochemical behavior of multiple metals subject to microbial reduction is poorly understood. To probe these complex processes, the reduction of U(VI) by the subsurface bacterium, Shewanella putrefaciens CN32, was investigated in the presence of goethite under conditions where the aqueous composition was controlled to vary U speciation and solubility. Uranium(VI), as the carbonate complexes UO₂(CO₃)_3(aq)⁴⁻ and UO₂(CO₃)_2(aq)²⁻, was reduced by the bacteria to U(IV) with or without goethite [α-FeOOH_(s)] present. Uranium(VI) in 1,4-piperazinediethhanesulfonic acid (PIPES) buffer that was estimated to be present predominantly as the U(VI) mineral metaschoepite [UO₃ · 2H₂O_(s)], was also reduced by the bacteria with or without goethite. In contrast, only ∼30% of the U(VI) associated with a synthetic metaschoepite was reduced by the organism in the presence of goethite with 1 mM lactate as the electron donor. This may have been due to the formation of a layer of UO_2(s) or Fe(OH)_3(s) on the surface of the metaschoepite that physically obstructed further bioreduction. Increasing the lactate to a non-limiting concentration (10 mM) increased the reduction of U(VI) from metaschoepite to greater than 80% indicating that the hypothesized surface-veneering effect was electron donor dependent. Uranium(VI) was also reduced by bacterially reduced anthraquinone-2,6-disulfonate (AQDS) in the absence of cells, and by Fe(II) sorbed to goethite in abiotic control experiments. In the absence of goethite, uraninite was a major product of direct microbial reduction and reduction by AH₂DS. These results indicate that DMRB, via a combination of direct enzymatic or indirect mechanisms, can reduce U(VI) to insoluble U(IV) in the presence of solid Fe oxides.

Introduction

A variety of anaerobic bacteria can catalyze the reduction of soluble species of U(VI) to insoluble U(IV) forms. Some of these microorganisms, including those capable of S (Lovley et al., 1993) and Fe (Lovley et al., 1991) respiration, do so by a direct enzymatic process coupled to the oxidation of organic compounds or H₂ Gorby and Lovley 1992, Lovley and Phillips 1992b. The product of this microbial reduction reaction is typically fine-grained uraninite [UO_2(s)] Gorby and Lovley 1992, Lovley and Phillips 1992b. The formation of some uranium ore deposits is believed to involve direct microbial reduction of U(VI) Lovley et al 1991, Mohagheghi et al 1985, as opposed to abiotic reduction by reduced species such as sulfide. Anderson, (1987) reported that U(VI) is not reduced to U(IV) by H₂S in anoxic seawater but rather diffuses into sediments where it is reduced to U(IV), possibly by anaerobic bacteria (McKee and Todd, 1993).

The purpose of this research was to investigate the reduction of U(VI), either as an aqueous species [UO₂(CO₃)_3(aq)⁴⁻] or as metaschoepite [UO₃ · 2H₂O_(s)], by the dissimilatory metal-reducing bacterium S. putrefaciens in the absence and presence of an Fe(III) oxide as an alternative electron acceptor during the metabolism of lactate. The influence of anthraquinone-2,6-disulfonate (AQDS) as a humic acid analog and electron shuttle (Lovley et al., 1998) on the microbial reduction process was investigated, as well as the potential for microbially generated reduced AQDS (AH₂DS) and Fe(II) to directly reduce U(VI). Soluble and extractable forms of U(VI) were measured and U X-ray absorption near edge structure (XANES) spectroscopy was used to determine the oxidation state of U associated with the solid phase and to quantify microbial reduction of U in complex mineral suspensions.

In addition to bacteria having an important role in the biogeochemical cycling of U, microbial processes can remove and concentrate U from contaminated ground and surface waters (Lovley and Phillips, 1992a) or from contaminated soil wash through reductive precipitation (Phillips et al., 1995b). Remediation of contaminated groundwater via wells and treatment ex situ is a relatively inefficient and costly process, while microbial U(VI) reduction can potentially be applied for the removal of U from solution in situ (Lovley, 1995).

Some DMRB can reduce solid phase Fe(III) oxides and oxyhydroxides including poorly crystalline phases such as ferrihydrite and crystalline phases such as goethite (Roden and Zachara, 1996), hematite (Zachara et al., 1998), and magnetite (Kostka and Nealson, 1995; Dong et al., 1999). S. putrefaciens and S. algae species appear to be particularly effective at reducing crystalline Fe(III) oxide phases although, to our knowledge, a systematic analysis of the ability of different DMRB to reduce different Fe(III) oxide minerals has not been undertaken. Fe(III) oxides are ubiquitous in nature and, in some aquifer sediments, they comprise the largest mass of oxidant (e.g., electron acceptor) for microbially catalyzed oxidation of organic compounds (Heron et al., 1994b). Hence, in Fe(III) oxide containing subsurface sediments with soluble U(VI), DMRB are presented with multiple electron acceptors that can be potentially linked to respiration. The actual pϵ where the valence transformations occur is strongly dependent on pH, reactant concentrations, aqueous speciation, and solid-phase distribution of the elements, and chemical potentials of reactants and products formed. In oxidized soils and sediments at circumneutral pH, Fe(III) is typically present as insoluble (hydr)oxides, while a significant fraction of the total U(VI) can be soluble carbonate or hydroxo complexes. From a thermodynamic standpoint (Francis et al., 1994) and because of much greater solubility, U(VI) should be reduced preferentially to Fe(III), which is poorly soluble in the oxidized form. However, differences in bacterial enzyme specificity and kinetics make it difficult to predict the sequence of reduction. Also, on a mass basis, Fe(III) is typically present in much higher concentrations than U. Therefore, it is difficult to predict, a priori, the fate of U(VI) in Fe(III) oxide containing sediments as microbial metal reduction proceeds.

Uranium may exist in contaminated soils in a variety of valence states (predominantly IV and VI) reflecting the nature of the source term and weathering extent Hunter and Bertsch 1998, Morris et al 1996. Uranium(VI) is the most stable valence form under oxidizing geochemical conditions Grenthe et al 1992a, Langmuir 1978. While U(VI) precipitates in sparingly soluble mineral phases including metaschoepite [UO₃ · 2H₂O_(s)] and various phosphates and silicates Grenthe et al 1995, Langmuir 1978, UO_2(aq)²⁺ forms a series of strong aqueous complexes with CO₃²⁻ [e.g., UO₂CO_3(aq)^o, UO₂(CO₃)_2(aq)²⁻, UO₂(CO₃)_(aq)⁴⁻] that greatly enhance U(VI) solubility in carbonate containing waters at circumneutral pH and above. Uranium(VI)–carbonate aqueous complexes are neutral or anionic in charge, and poorly reactive with mineral surfaces such as Fe(III) and Al oxyhydroxides Duff and Amrhein 1996, Hsi and Langmuir 1985, Waite et al 1994 and clays that typically control metal ion migration in soil and groundwater by adsorption reactions. Remediation technologies for U(VI) contamination often involve reduction of mobile U(VI) aqueous complexes to insoluble U(IV) precipitates Fiedor et al 1998, Gu et al 1998, where carbonate complexation has minimal effect on solubility (Casas et al., 1998). To be effective in the long term, however, remediation techniques for U(VI) must target both mobile aqueous species that are groundwater contaminants, as well as U(VI) precipitates that may be long term sources.