We propose to use techniques of bacterial physiology, biochemistry and molecular genetics, combined with environmental scanning electron microscopy (ESEM) and X-ray microscopy (XRM) and X-ray spectroscopy (XRS) to examine: [1] mechanisms of direct reduction of Cr⁶⁺ and U⁶⁺; [2] mechanisms of immobilization of Cr⁶⁺ or U⁶⁺ to both biologically and abiotically formed metal oxides; [3] mechanisms of remobilization of both adsorbed and precipitated ACT and TM. The work will test specific hypotheses regarding direct reduction of TM and ACT under anoxic conditions and indirect immobilization onto metal oxides under oxic conditions. The thesis is that there are two different pathways that can be followed for removal from solution, and that the resulting precipitates will require very different approaches for their bioremediation. Understanding the formation, properties, and stabilities of insoluble forms of TM and ACT should allow one to examine environmental samples and plan remediation efforts in a systematic and intelligent way that has heretofore not been possible because of the lack of understanding of precipitated ACT and TM.

The long term goals of the proposed work are aimed at development of methods for bioremediation of transition metal (TM) actinide (ACT) pollutants. The methods we develop will be applicable to polluted materials from soils, sediments or water systems, and will involve an understanding of the way that metal oxidizing (MOB) and metal reducing (MRB) bacteria interact both directly and indirectly with immobilized TM and ACT. To accomplish these long-term goals, we put forward a series of hypotheses that will be systematically tested as short term objectives. With answers to these questions, we should be able to move forward in accomplishing the long term goals of bioremediation. The basic idea of our work is that the nature of TM and ACT pollutants will be dependent upon the conditions under which these pollutants are deposited, and that understanding these differences is critical to developing methods for remediation – e.g. all TM and ACT pollutants are not alike! If this is so, then immobilized TM and ACT may be very different with regard to chemical and biological stability, and these differences will be critical to understand in terms of handling and bioremediating waste materials. The hypotheses to be tested thus involve aspects of immobilization of TM and ACT (by direct and indirect mechanisms), and ways in which the resulting insoluble forms can be preserved and/or transformed for remediation purposes (Figures 1 & 10).

If the history of the system is anoxic, then the following hypotheses are proposed:

1. Direct (cell or enzyme catalyzed) reduction of Cr and U will and does occur.

2. Enzymatic based delivery systems can be used to catalyze these reactions in cell-free remediation systems.

3. The resulting insoluble Cr and U species will be stable under oxic conditions.

If the history of the systems is oxic, then the following hypotheses are considered:

1. Adsorption of U and Cr onto metal (Mn and Fe) oxides will be rapid and efficient.

2. The redox state of the adsorbed Cr or U will remain as the hexavalent state.

3. Adsorption will be substrate (metal oxide) dependent.

4. Biologically formed oxides will be different from synthetic oxides in adsorptive ability.

5. Several factors will be important in controlling the rate and extent of adsorption:

i. temperature

ii. time of addition of TM or ACT

iii. counterions present (organic and inorganic)

iv. Eh/pH

6. The stability of the bound TM or ACT will vary with counterions, Eh, pH, and the ability of metal reducing bacteria (MRB) to solubilize metal oxides and/or directly reduce the TM or ACT.

Transition metals (TM) and actinides (ACT) are pollutants in many soils, sediments, and water systems. For example, more than 1/3 of the 91 waste sites surveyed at 18 DOE sites were heavily contaminated with TM and ACT (58). There are many unknowns in these systems, including those related to the chemical speciation and stability of TM and ACT in the polluted sites. For example, the solubility of both U and Cr are controlled in two distinctly different ways (Fig.1): [1] under anoxic conditions, they can be chemically or biologically reduced to insoluble forms (U⁶⁺ to U⁴⁺ and Cr⁶⁺ to Cr³⁺); and [2] under oxic conditions, they can be adsorbed to insoluble Mn or Fe oxides and oxyhydroxides (U⁶⁺ or Cr⁶⁺ + Me-OOH ---à U-Me-OOH). In both cases, the TM or ACT is removed from solution, but the stability of the precipitate will depend on formation conditions (Fig.1 & 10). Thus, before handling polluted samples one needs to know: [1] the nature of the contaminating TM or ACT; and [2] the stability of immobilized TM or ACT. If these two items can be determined, the probability of success in bioremediation will be markedly increased.

The biogeochemical cycling of iron and manganese play important roles in the chemistry of U (and other ACT) and Cr (and other TM). Many organisms oxidize (14,21,49,50) and/or reduce (33,52) Mn and/or Fe, and a few interact directly with both Cr and U, either reducing the hexavalent states (19,23,34-36,59) or oxidizing the lower valence states to the soluble hexavalent forms (11,14). These abilities circumscribe virtually all of the processes shown in Figures 1 and 10; if biotic and abiotic aspects of these processes were more clearly defined, planning bioremediation efforts could be methodological and predictable.

Metal Oxidizing Bacteria (MOB): Many bacteria and fungi catalyze Mn oxidation, converting soluble Mn²⁺ to MnO₂ (14,49,50). In a few cases, enzymes have been partially purified and studied (1), and in recent studies, genes coding for Mn oxidizing proteins have been identified, cloned, and sequenced (8,18). We will focus our studies on two different Mn oxidizing systems, SG-1 (54), and BG-18 (25,26). SG-18 is a marine bacillus isolated from a manganese-encrusted sand grain (54) that produces spores that can catalyze the oxidation of Mn²⁺ under different conditions (27,37,62), without the need for vegetative growth or metabolism. Figure 2A shows the effect of adding SG-1 spores to a solution of Mn²⁺, and 2B shows a transmission electron microscope (TEM) image of a manganese-encrusted SG-1 spore (53). This system requires molecular oxygen, but does not require cell growth or addition of metabolic substrates. Mandernack et al. showed that SG-1 spores produced many different kinds of Mn oxides depending on temperature, Mn²⁺ concentration, and time of incubation (37).

BG-18, a Pseudomonas-like organism, originally isolated from freshwater (25,26), is an efficient Mn oxidizing bacteria. Mn oxidation is coupled to cell growth and metabolism of heterotrophic substrates (2,25,26). Figure 3 shows a precipitate of MnO₂ formed by BG-18 cells, with the cells imbedded in the Mn oxides they form. This image of unfixed, unstained cells, was obtained using the x-ray microscope at LBNL.

Manganese and Iron Oxides as Scavengers: Mn and Fe oxides have been referred to as the "scavengers of the sea" (22), because of their ability to complex and remove from solution a variety of metals (Cd, Co, Cr, Pb, U, Pu, Th, etc.;4-6, 61,). These abilities are well known through many chemical studies of metal binding to freshly prepared Mn and Fe oxides (4-6), as well as environmental geochemical studies in which the distribution of various TM, ACT and other metals are correlated with the distribution of Mn and Fe oxides (70,71). The adsorptive abilities of the metal oxides has led to proposals for their use for the clean-up of metal-polluted waters (24), as well as the removal of radionuclides from polluted systems (40,44); Moore used Mn-oxide-coated acrylic fibers to collect radium from seawater (41,42). This information, while useful, is lacking in several key parameters necessary to understand if such systems can be used for bioremediation:

1. What is the redox state of the adsorbed TM or ACT on the metal oxides?
2. What factors control the rate(s) and extent(s) of adsorption of TM or ACT onto metal oxides?
3. How do biologically formed metal oxides compare to synthetically prepared metal oxides in terms of adsorptive abilities?
4. What are the effects of added counter ions (either organic or inorganic) in the system during adsorption?

For example, Arnold et al. (2) reported that biologically formed Mn oxides were much more efficient at binding heavy metals than were synthetically formed oxides.

Metal Reducing Bacteria (MRB): Since the late 1980’s MRB have been recognized as bacteria that grow anaerobically, using Mn³⁺, Mn⁴⁺, or Fe³⁺ as oxygen substitutes for respiration (33,52). This results in a dissimilatory reduction of the metals, with accompanying oxidation of organic carbon and growth of the bacteria. Two major groups of MRB are characterized, the Geobacter group and the Shewanella group (32). The former are anaerobes with a close affiliation to the sulfur- and sulfate- reducing bacteria, while the latter are facultative aerobes with a close relationship to Vibrios and E. coli (32); in addition, several new groups of reducers are known from recent studies which are distinct from either of these two major groups (32,51). Since the proposed work focuses on utilizing S. putrefaciens, the background literature reviewed will focus on results obtained with Shewanella species. Many of the findings are similar to those found in studies of Geobacter-like organisms, although a significant difference involves the ability of the Shewanella group to tolerate, and even use, molecular oxygen, so that introducing these organisms to an aerobic system would not stress them as it would some of the obligately anaerobic bacteria of the Geobacter group.

Although the study of dissimilatory microbial metal reduction is a new field, some general features are emerging. Several metals can be reduced by MRB, including Mn, Fe, U, Cr, and S^o (51) many of which may be environmentally significant for bioremediation. Whether catalysis reactions are specific and/or independent is, in general, not known, although for S. putrefaciens, mutants have been obtained that are deficient in Fe but not Mn reduction, and vice versa, thus implying that the terminal reductases for the two processes are different (13,52,63). Similar mutants have not been reported for other metals, or for other metal reducing strains. Growth coupled to the reduction of metals other than Fe or Mn has not been extensively studied.

Burdige et al., (7) showed that a culture of S. putrefaciens growing on manganese oxides with different mineralogies gave very different rates of reduction that were generally proportional to metal oxide surface area, so that a highly crystalline oxide like pyrolusite (with a very low surface to volume ratio) was reduced very poorly: almost not at all. Roden and Zachara (60), studying iron oxides, concluded that rate and extent of reduction were controlled by surface area and site concentration of the solid phase. From studies such as these, it seems likely that the major controlling factor in solid metal reduction is surface area, with other factors including crystal structure, morphology, free energy, and particle aggregation participating to lesser degrees (60).

Yield studies of bacteria on solid substrates vary because of bacterial attachment to the oxides (Figs 4 and 5). When growing on manganese oxides (Fig. 4), S. putrefaciens forms a layer of extracellular polymer that obscures individual cells when viewed by environmental scanning electron microscopy (ESEM). The nature of this polymer is not yet elucidated, but it is likely to be a polysaccharide based on the fact that it is difficult to visualize unless samples are dehydrated and viewed by standard scanning electron microscopy (Fig. 4). Interestingly, when the same cells are grown on iron oxides, no extracellular polymers are conspicuous (Fig. 5) when examined with ESEM.

Biochemistry of Metal Reduction: As of this writing, iron or manganese reductases have not been purified or characterized, although high levels of both activities have been observed both in whole cells and in cell free extract. Myers and Myers (47,48) reported that cytochromes and iron reductase activity of S. putrefaciens are located in the outer membrane, consistent with the observations that metal oxides are solids and cell contact is required for metal reduction. Morris (43) and Pealing and colleagues (56,57) studied a variety of different cytochromes from another strain of S. putrefaciens, including a low potential flavocytochrome with fumarate reductase activity. Tsapin et al. (73) purified a small (12 kDa) tetra-heme cytochrome c₃ of very low potential (-233 mV) with a high sequence similarity to the cytochrome c₃ of Desulfovibrio desulfuricans, and showed that this cytochrome in its reduced state could

reduce Fe³⁺. Mutants missing this cytochrome are unable to reduce Fe³⁺ and several other electron acceptors (Nealson, unpublished), but the physiological role of the cytochrome has not yet been elucidated. However, the gene is now cloned and sequenced (72), so it should be possible to generate insertional mutants and specify a function for this cytochrome. In addition, 5 other c-type cytochromes have been purified from S. putrefaciens. Two have low redox potentials consistent with metal reductases and one is membrane bound (Tsapin, Meyer, and Nealson, unpublished results). Gorby, Lovley and others have shown that cells or a partially purified cytochrome c₃ from Desulfovibrio vulgaris that could reduce U⁶⁺ and/or Cr⁶⁺ (23,34-36)

In many cases of metal reduction, the possibility that indirect reactions occur must be considered. Manganese is a particular problem because it is so easily reduced by other reductants (65-68), including Fe²⁺ or HS^-, both produced by S. putrefaciens (46). Thus, any sulfur- or iron- reducing bacterium can appear to be a Mn reducer via indirect reactions, and in fact, with catalytic amounts of iron or S^o, a biogeochemical cycle resulting in Mn reduction can be established.

Regulation of Metal Reduction: Several studies of regulation of metal reduction by S. putrefaciens indicate that regulation occurs at several levels. Arnold et al. (3) and DiChristina (12) demonstrated a physiological competition between iron and other electron acceptors, with oxygen or nitrate capable of inhibiting iron reduction. Similarly, Myers and Nealson (45) showed that oxygen or nitrate inhibited reduction of Mn⁴⁺ by S. putrefaciens, but that less electropositive electron acceptors like fumarate or sulfite did not. In general, in the presence of oxygen or nitrate, no metals are "breathed", but in the absence of other good electron acceptors, bacteria respire iron and/or manganese.

Anaerobic respiration in bacteria is controlled by a variety of regulatory systems (fnr , arcBA, etc.), which sense oxygen, or some product of oxygen, and control metabolic pathways (like the tricarboxylic acid cycle) and specific reductases at the level of biosynthesis (74,75). Mutants in these control systems characteristically result in pleiotropic phenotypes, with a number of different anaerobic processes being simultaneously affected. Several such pleiotropic mutants have been isolated from strains of S. putrefaciens (63,64), and in one case the gene controlling the mutation was isolated, sequenced, and found to be an analogue of the fnr gene from E. coli (63), suggesting that similar control mechanisms operate to regulate anaerobic respiration of this metal reducer. Similar studies of regulation at the molecular level have not been reported for other metal reducers.

Metals can also be removed via direct reduction by the MRB. While iron and manganese are solubilized, other metals are converted to insoluble forms upon reduction. Of note are chromium (Cr⁶⁺) and uranium (U⁶⁺), which are soluble in the oxidized form, but insoluble as the respective Cr³⁺ and U⁴⁺, reduced species. As discussed above, reduction of both metals can be accomplished by either whole cells, or cell free extracts containing cytochrome c₃ (cc-3). The work with cc-3 and other cytochromes will constitute one of the early focal points of our study, and the recent cloning of cc-3 from S. putrefaciens (72) could offer a ready supply of the protein for such work, as well as a good model system for studying the specific mutants of the cc-3 system. S. putrefaciens can reduce Cr⁶⁺ but no detailed studies have been presented (Nealson, unpublished). As with uranium, the removal of Cr⁶⁺ should be possible using either intact cells, or cell free systems of MRB.

Stability of the immobilized TM or ACT: In terms of stability of immobilized pollutants, two areas are considered. First the stability of reduced metals immobilized by conversion to the reduced (insoluble) form. In this case, there have been few microbiological studies, and little is known. Second, stability of metal oxides that contain adsorbed pollutants, and here, a considerable amount is known.

Release of adsorbed pollutants: When iron or manganese oxide reduction occurs in municipal water systems, not only is the water fouled by excess soluble manganese and/or iron, but trace components bound to the metal oxides may also be released (15-17). Such reactions may account for the distribution of trace metals (6,61,70,71) and radionuclides like uranium in sediments (10,39,69) and anoxic water columns (38). The work of McKee et al., (38,39,69) may offer an environmental example of what one may expect with adsorbed TM and ACT. They reported that U accumulates in anoxic environments (Amazon Basin sediments, or reducing zones of fjords), but escapes (probably via release from metal oxide particles) to the overlying water. We propose that this reaction will operate to release TM and ACT bound to metal oxides. However, some subtle controls may be important. For instance, if the Eh and pH conditions are appropriate, the TM or ACT may be reduced and immobilized in a different form, rather than being released. It is these conditions that need to be defined for various metal oxides and microbes to control immobilized contaminants.

Confocal Laser Scanning Microscopy (CLSM) and Environmental Scanning Electron Microscopy (ESEM): We propose to use the microscope facility and the chemical and microbiological expertise of Dr. Brenda Little at the NRL (Stennis Space Center) to document microbial pollutant interactions. The utility of CLSM for location of bacteria in natural samples is well known, and this tool has been used to demonstrate that MRB are closely associated with iron and manganese oxides, and in fact distributed throughout the porous structures (30). The ESEM has revealed new information about MRB (Figures 4 and 5), including close attachment of S. putrefaciens to metal oxides, formation of a polymeric layer during Mn reduction, and detailed structure of the metal oxides, with and without bacteria.

Electroscan Corporation (Wilmington, MA) recently introduced a new development in SEM technology, the ESEM. ESEM coupled with energy dispersive x-ray spectroscopy (EDS) provides fast, accurate (50 Å resolution) images of microorganisms within biofilms, their spatial relationship to the substratum and elemental composition. Microorganisms that colonize surfaces produce polymers and form a gel matrix that is central to the structural integrity of the biofilm (9). The gel is known to immobilize water at the metal/biofilm interface and to trap metal species. Preparation of biological material for SEM requires extensive manipulation, including fixation, dehydration, and either air drying or critical-point drying because the SEM operates at high vacuum. Artifacts resulting from SEM sample preparation including solvent extraction of bacterial extracellular material and bound metals have been documented (29). Non-conducting samples including biofilms must be coated with a conductive film of metal before the specimen can be imaged. Uncoated non-conductors build up local concentrations of electrons, referred to as "charging," that prevent the formation of usable images. EDS can be used to determine the elemental composition of surface films in the SEM, but EDS analyses must be completed prior to deposition of the thin metal coating. EDS data are typically collected from an area; the specimen, removed from the specimen chamber and coated with a conductive layer, and returned to the SEM. The operator attempts to relocate and photograph the precise area from which the EDS data were collected.

In contrast, biological specimens for examination with ESEM are attached to a Peltier stage maintained at 4°C and imaged in an environment of water vapor at 2-5 torr to maintain samples in a hydrated state. ESEM uses a unique secondary electron detector capable of forming high resolution images at pressures in the range of 0.1 to 20 torr. At these relatively high pressures, specimen charging is dissipated into the gaseous environment of the specimen chamber, enabling direct observation of uncoated, nonconductive specimens. Wet biofilms on mineral surfaces can be observed directly, and EDS data can be collected at the same time sample morphology and topography are photographed. ESEM images and EDS spectra were recently used to demonstrate that surfaces of iron and manganese oxides were structurally and chemically altered during microbial reduction (28,30,31,). .

X-ray Microscopy (XRM), X-ray Spectroscopy (XRS) and X-ray Photoelectron Diffraction (XPD): Development of XRM, XRS, and XPD has been proceeding at LBNL for several years, with the involvement of Dr. Brian Tonner. These tools have recently been used to study manganese oxidizing and reducing bacteria with exciting results that show how these tools can impact environmental studies. Soft x-ray spectro-microsocopy is a relatively new method of obtaining chemical composition information with high spatial resolution. For application to environmental problems, it is a tool for speciation of metal contaminants in solution and as precipitates. Closely related techniques of surface spectroscopy, namely x-ray photoelectron spectroscopy (XPS) and x-ray photoelectron diffraction (XPD) are used to determine the surface chemical structure of environmentally relevant substrates. The particular advantages of the "soft" x-ray spectrum for environmental analyses are: 1) it covers the important absorption edges of C, N, and O; 2) it overlaps the most chemically sensitive lines of the transition elements and actinides; and 3) it is the wavelength range most suited to high spatial resolution using zone-plate x-ray lenses.

We propose to use soft x-ray spectroscopic microscopy to study chemical interactions between ACT, TM, and metal oxides under a variety of controlled laboratory conditions related to environmental systems. X-ray spectroscopy, in the ‘hard’ X-ray region, is a reasonably well-known technique for chemical analysis. The capabilities of the ‘soft’ x-ray region (corresponding to photon energies below 1,000 eV) are less well known. We will make extensive use of these new tools for quantitative micro-analysis of the mechanisms for actinide bioremediation by direct and indirect pathways involving microbial metabolism (Fig. 1). In the indirect process, biologically produced metal oxides act as substrates for actinide adsorption. An x-ray micrograph of an unfixed, unstained culture of BG-18 is shown in Fig. 3 (above) -- this culture is in the process of oxidizing Mn and creating biological Mn oxide of the type we propose to study. In Fig. 6, a high resolution image of manganite crystals floating in water is shown, illustrating the ability of this technique to image hydrated materials at high resolution.

While the techniques are new, they are firmly grounded in quantitative physical methods. For example, Fig. 7 shows x-ray absorption spectra from a study of S. putrefaciens reduction of MnO₂ particles. The control spectra are on the right side of the figure. As Mn reduction proceeds, the spectrum for Mn³⁺ appears, suggesting that the first stage in reduction is a one electron transfer to Mn⁴⁺. In addition, a natural sample (Mn nodule from Oneida Lake, NY) was examined, showing that the surface is coated with a layer of Mn²⁺. This figure shows the potential value of the technique -- it can be applied directly to environmentally produced materials with quantitative interpretation. For our proposed work, the measurements can be applied to solution of some long-standing problems dealing with the metabolic pathway of Mn reduction as reflected in the distribution of Mn valences in large inorganic deposits of Mn oxides (nodules). The soft x-ray absorption spectrum of Mn, at the L edge (2p core level) is very sensitive to the valence charge of the metal atom. The intrinsic line-width of this core-level is very narrow, so that the multiplet splitting at both the 2p_3/2 and 2p_1/2 intensities of the multiplet lines is used to fingerprint the metal atom valence. It is also possible to identify the metal atom spin state, since the spin state and valence are closely related. In the specific cases shown in Figure 7, we can see directly that the surface of the MnO₂ particle is reduced to valence (III). From calibration standards and curve-fitting, we will be able to determine the thickness of the reduced layer, and the distribution of Mn in other valence states. Spectra from the Mn-nodule deposit show that surface layers of the particles are nearly entirely in valence (II). Further work is planned to probe deeper below the surface of the individual particles in nodules, to compare the oxidation state at the particle interior to that at the surface layer.

With the combination of spectroscopy and microscopy one can perform detailed studies of the distribution of organic and inorganic compounds resulting from microbial metabolism and attachment. It is possible to image specimens through a thin layer of water, in a so-called ‘wet cell’. The penetrating power is good, so that whole cells can be imaged. Nearly simultaneous and spatially registered optical and x-ray micrographs show that the specimen wet cell retains live bacteria for at least the initial x-ray beam exposure. Spatial resolution ranges down to about 400 Angstroms.

By varying the incident wavelength (energy) of the x-ray beam, one can change the transparency of the sample to emphasize either the organic (microbial) or inorganic (precipitate) component in the specimen. An example of this is shown in Figure 8, in which two different incident energies are used to excite the same sample. In one case, the Mn oxide particles are enhanced, while in the other, the bacteria are visible.

Spectroscopy of liquids and liquid-solid interfaces: One of the essential steps in the development of the in-situ soft x-ray spectro-microscopy by our group was the demonstration of spectroscopic capabilities in liquids. Although work had been done in water soluble ions using hard x-rays, our work is the first to extend this to metal ions using soft x-rays. The importance of this lies in the fact that soft x-rays are needed to get the sharp spectral lines shown in Fig. 7 and the high spatial resolution of Fig. 6.

An example of the use of soft x-ray spectroscopy to study aqueous state transition metal ions taken from our recent research is seen in Figure 9. There are obvious differences in the spectra. The quality of spectra for solvated ions is very good, so that one can extract information about the valence, ligand geometry, and spin-state of the solvated ions. Two practical applications of liquid-state spectroscopy are: 1) differentiation of metal complexes that are solids from those in liquid states; and 2) measurement of valences of TM and ACT in liquid state, rather than only the average oxidation state as in traditional wet chemistry methods. The methods used to acquire spectra from solvated ions in aqueous solutions can be applied to a number of different problems. The immediate need is for studies as a function of pH on samples including environmentally important minerals, such as goethite and iron substituted clays, for which there are questions regarding the relative distributions of Fe(II) and Fe(III). The technique is not restricted to metals ions in aqueous solutions; it is also well suited to identifying organic compounds in solution.

Structure of environmental interfaces by photoelectron diffraction: We propose to study the interface formation of actinides and transition metal complexes using surface science techniques. These tools are high resolution x-ray photoelectron spectroscopy (XPS or ESCA), and a new technique called x-ray photoelectron diffraction (XPD). The latter method, XPD, is used for determining the chemical valence and adsorption sites of actinides on oxide surfaces. The project team includes experts in surface crystallography by x-ray photoelectron diffraction (XPD), which

combines very accurate experimental measurements with full quantum mechanical simulations, resulting in high precision atomic structures of interfaces. XPD gives a local picture of atomic bonding, allowing one to select those atomic species of importance. Such measurements result in a diffuse diffraction pattern characteristic of the local bonding surrounding a specific atomic species at the surface of the oxide layer. This diffraction pattern is then analyzed using quantum mechanical dynamical diffraction theory, to determine adsorption site symmetry and bond lengths of chemical species at the oxide surface. This body of work is very valuable in understanding certain fundamental problems in environmental science. For example, the removal of actinide pollution from solution by bacterially-produced transition metal oxides is a known phenomenon. We currently hypothesize that the attachment of actinide to the metal oxide is a physisorption step, without any change in valence. However, there is absolutely no information on model systems in this field, and consequently, no evidence one way or the other for this idea.

The ability to analyze wet samples, while making a sophisticated electronic structure measurement, is unique to x-ray spectroscopic imaging. Spectro-microscopy tools will be used to determine a number of important characteristics of liquid and solid samples. We will analyze individual particles in a complex mixture, obtaining the composition of each, and determining the distribution of charge states rather than the average composition. Importantly, this can be done for samples under water, eliminating the uncertainty of changes in chemical composition to small particles that may occur upon drying. Our laboratory currently has the highest spatial resolution for this kind of work at the energies required for actinide research. We will use these sophisticated surface and interface analysis tools to determine the adsorption kinetics of actinide compounds in solution as they become incorporated in the surface region of substrate oxides.

The proposed three year program is focused on testing several specific hypotheses for TM or ACT bioremediation. We present a timetable for each collaborator. The program is complex, with several projects beginning at the onset of funding, and others beginning and ending during the course of the project. These will be discussed as individual components, but always in the context of the total project as presented in Figure 10.

A. The Formation of Fresh (Fe or Mn) oxides: This part of the program involves the study of two different organisms (SG-1 and BG-18) for their ability to form Mn oxides that can be used for adsorption studies of the kind presented in section I. Experiments will be initiated in the Nealson laboratory, optimizing rates and extents of Mn oxide formation by the two cultures. Work will begin with SG-1 spores, and as the conditions for oxide formation are established, the system will be prepared for study by the LBNL group. Minerals will be characterized by X-ray diffraction, and surface chemical properties, surface area, and high resolution TEM, ESEM, and EDS, in order to get detailed structural and chemical analyses of their structures before adsorption studies are undertaken. BG-18 will also be added to the studies as work with SG-1 progresses.

(SG-1 formation of Mn oxides/ Eh,pH,T, nutrients)

(U and Cr uptake by SG-1 oxides)

(BG-18 formation of biological Mn oxides)

(Binding of U and Cr to BG-18 oxides)

Experimental Details: SG-1 is a spore-forming marine bacillus, and the spores will be used for these experiments. Spores will be prepared as previously described (27,37), but in bulk preparations, so that milligrams of spores are available for each experiment. They will be introduced into vials with various levels of Mn²⁺, at a variety of Mn²⁺, T, Eh, pH and salinity values. They will also be tested with different levels of naturally occurring salts and ions, such as carbonate, sulfate, nitrate, and organics such as lactate, acetate, formate, and amino acids. As Mn oxide coatings begin to form they will be harvested, fixed and sent to Dr. Brenda Little for analysis. Previous studies by Mandernack et al. (37) indicate that a variety of Mn³⁺ and Mn⁴⁺ oxides can be formed by SG-1 simply by varying Mn²⁺ concentration and temperature. Careful attention will paid in these experiments to the times needed for efficient oxide formation.

As conditions are defined for oxide formation, they will be used to test uptake of TM and ACT during and after formation. That is, if the TM are continuously present, are they continuously adsorbed, or does this occur only after oxide formation is complete? If high levels are present, do they inhibit Mn oxide formation? Can TM or ACT be detected by standard EDS analysis after uptake experiments, even if they are taken up during early oxide formation? These experiments will be an interesting follow-up to old reports of Arnold et al. (2) who reported that biologically formed Mn oxides bound significantly higher levels of heavy metals than did synthetic oxides.

B. Binding of U and Cr to known metal oxides: A major part of the program is the use of new approaches (ESEM, XRM, and XRS) to collect images and data not previously obtained in studies of this sort. High resolution ESEM work will be initiated by Dr. Brenda Little at the Stennis Space Center, while XRM and XRS will be undertaken at the ALS at LBNL. These will involve analyses of structure and composition of metal oxides with TM and ACT bound. The initial work will use well-characterized oxides of known origin and mineralogy, and will set the stage for later work with biologically formed metal oxides. In these studies, important chemical variables like alkalinity will be carefully monitored and controlled, as they may be of great importance in predicting binding of TM and ACT to metal oxides.

(Binding of U & Cr to synthetic Mn oxides)

(Binding of U and Cr to synthetic Fe oxides)

(Binding of U and Cr to biological Mn oxides)

(Binding of U and Cr to biological Fe oxides)

Experimental Details: Experimental design will be straightforward in the sense that the questions can be addressed by standard methods of adsorption studies (4-6). Metal oxides (Mn and Fe) will be purchased or synthesized, characterized (X-ray, IR, ESEM, XRS, and surface area), and used in standard measurements of adsorption isotherms for U⁶⁺, U⁴⁺, Cr⁶⁺, and Cr³⁺ under various conditions. Once adsorption has been achieved, the surface chemistry of the oxides with adsorbed metals will be examined by both ESEM, XRM and XRS. In a second set of experiments, biologically formed Mn oxides will be tested for their ability to bind U and Cr. In these experiments, where the rates of formation of the oxides can be controlled, the ACT or TM will be added at various times to test for the importance of the time of incorporation on the stability of the complex. Oxides will be tested for metal content, metal speciation and morphology, using ESEM, XRS and XRM measurements.

The XPD work will use single-crystal oxide substrates, exposed to solutions of TM or ACT. We will determine adsorption sites and charge states of the interface layer, for substrate surfaces that expose different crystal faces and surfaces with varying degrees of defects. In parallel, we will be making microscopic measurements on the structure and morphology of environmentally produced substrates. This information will then be used to interpret adsorption rates for actinides found in the environment.

B’. Direct Reduction of Cr⁶⁺ and U⁶⁺ by S. putrefaciens and Cell Extracts: These experiments will be performed by the Nealson laboratory, and are aimed at optimizing the conditions, and perhaps the cell components, that catalyze Cr⁶⁺ and U⁶⁺ reduction by S. putrefaciens. This part of the project can be initiated immediately. It will involve defining redox conditions under which strain MR-4 will catalyze Cr and U reduction. Experiments will be carried out in a biostat chamber with O₂, CO₂, and pH control, and Cr or U monitored during the experiment. This will allow us to define the conditions that favor Cr or U reduction by whole cells. We will vary pH, Eh, and nutrient type and level.

These experiments will be closely linked to pathways C’ and C" in Fig. 10, because as conditions for Mn oxide reduction are defined, and as conditions for Cr and U reduction are defined, we anticipate (based on the environmental data of McKee et al. (10,38,39)) that we will be able to set up controlled environment conditions in which we can reduce the metal oxides and quantitatively release bound U. As for Cr, we are making no predictions, but the experimental protocol, as planned should allow prediction as data are produced.

(MR-4 reduction of U and Cr)

(cyt c₃ reduction of U and Cr)

(Release of bound Cr and U by Reduced Mn oxides)

(Control of fate of released Cr and U)

Experimental Details: Growth conditions in MR-4 experiments will be controlled via a biostat, to define cell number, Eh, pH, T, etc. optimal for reduction of Cr and U. Once these values are known, and those for reduction of Mn oxides are known, it should be possible to test the hypothesis that the reactions can be separated by controlling the redox conditions. The corollary of that hypothesis is that by making the conditions appropriate, it will be possible to avoid release of the pollutants because they will be rapidly converted to the insoluble form as Mn reduction proceeds.

For the cytochrome c₃ experiments, two approaches will be taken. First, pure cytochrome c₃ will be chemically reduced (e.g. with dithionite), separated from reductant, and titrated against a variety of different compounds. Alternatively, it can be electrochemically reduced if chemical reductants are difficult to remove (55). While its redox potential is low (-230 mV), its reactivity against a variety of different solid state compounds has not been systematically tested. Second, the system will be set up as a cell free reductase system, using hydrogenase (pure enzyme) as a coupling enzyme and looking for enzyme-linked U or Cr reduction. Finally, as the work progresses and some of the other low potential cytochromes from S. putrefaciens become available, they will be tested for their abilities to reduce U and/or Cr in the same way.

C: Stability of Reduced TM or ACT. This part of the study will be primarily a chemical one, with limited biological work. The questions addressed will relate to the chemical stability of uraninite (UO₂) under oxidizing and reducing conditions, and the relationship between stability and commonly encountered environmental variables, such as Eh, pH, alkalinity, organic carbon, etc. The oxidation state of the U in the mineral will be monitored as incubations are continued. The work will be done by the group at LBNL under controlled conditions, with monitoring of the precipitates at regular intervals.

C’: Direct Reduction of TM and ACT. This work is analogous to that described for section B’ in Fig.10, with the exception that it will combined with work on the reduction of the Mn or Fe oxides.

C": Reduction of Metal Oxides by S. putrefaciens: In these experiments, we have

substantial preliminary data from many studies, but we lack a defined set of conditions under which Mn and/or Fe oxides are reduced. We will thus establish a biostat culture system with controlled Eh and pH, and examine the rate and extent of Mn and Fe reduction under defined conditions. In these experiments, we will use a voltametric electrode system to monitor concentrations of Fe²⁺ or Mn²⁺ in the outflow of the reaction vessel. This should allow rapid assessment of the rate of metal reduction under various combinations of Eh, pH and nutrients and temperature.

(Optimization of Mn reduction conditions)

(Comparison of conditions with different Mn oxides)

(Optimization of Fe oxide reducing conditions)

(Comparison of conditions with different Fe oxides)

Experimental Details: The details of these experiments are rather simple. Cells and oxides will be placed in a continuous flow system where Eh, pH, and other variables can be controlled through a feedback system, and other key variables like hydrogen and CO₂ are continuously monitored. The outflow of the system will be monitored for reduced Mn or Fe, and the conditions optimized for these variables. We anticipate defining conditions for Mn reduction early in the project, and using these conditions for experiments discussed late in the project of section 3 above.