This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Author's personal copy
Available online at www.sciencedirect.com
Geochimica et Cosmochimica Acta 72 (2008) 4901–4915 www.elsevier.com/locate/gca
Characterization of uraninite nanoparticles produced by Shewanella oneidensis MR-1 William D. Burgos a,*, Jeffrey T. McDonough a, John M. Senko a, Gengxin Zhang a, Alice C. Dohnalkova b, Shelly D. Kelly c, Yuri Gorby d, Kenneth M. Kemner c a
Department of Civil and Environmental Engineering, The Pennsylvania State University, 212 Sackett Building, University Park, PA 16802, USA b Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, WA, USA c Biosciences Division, Argonne National Laboratory, Argonne, IL, USA d J. Craig Venter Institute, La Jolla, CA, USA Received 6 December 2007; accepted in revised form 25 July 2008; available online 6 August 2008
Abstract The reduction of uranium(VI) by Shewanella oneidensis MR-1 was studied to examine the effects of bioreduction kinetics and background electrolyte on the physical properties and reactivity to re-oxidation of the biogenic uraninite, UO2(s). Bioreduction experiments were conducted with uranyl acetate as the electron acceptor and sodium lactate as the electron donor under resting cell conditions in a 30 mM NaHCO3 buffer, and in a PIPES-buffered artificial groundwater (PBAGW). MR-1 was cultured in batch mode in a defined minimal medium with a specified air-to-medium volume ratio such that electron acceptor (O2) limiting conditions were reached just when cells were harvested for subsequent experiments. The rate of U(VI) bioreduction was manipulated by varying the cell density and the incubation temperature (1.0 108 cell ml1 at 20 °C or 2.0 108 cell ml1 at 37 °C) to generate U(IV) solids at ‘‘fast” and ‘‘slow” rates in the two different buffers. The presence of Ca in PBAGW buffer altered U(VI) speciation and solubility, and significantly decreased U(VI) bioreduction kinetics. High resolution transmission electron microscopy was used to measure uraninite particle size distributions produced under the four different conditions. The most common primary particle size was 2.9–3.0 nm regardless of U(VI) bioreduction rate or background electrolyte. Extended X-ray absorption fine-structure spectroscopy was also used to estimate uraninite particle size and was consistent with TEM results. The reactivity of the biogenic uraninite products with dissolved oxygen was tested, and neither U(VI) bioreduction rate nor background electrolyte had any statistical effect on oxidation rates. With MR-1, uraninite particle size was not controlled by the bioreduction rate of U(VI) or the background electrolyte. These results for MR-1, where U(VI) bioreduction rate had no discernible effect on uraninite particle size or oxidation rate, contrast with our recent research with Shewanella putrefaciens CN32, where U(VI) bioreduction rate strongly influenced both uraninite particle size and oxidation rate. These two studies with Shewanella species can be viewed as consistent if one assumes that particle size controls oxidation rates, so the similar uraninite particle sizes produced by MR-1 regardless of U(VI) bioreduction rate would result in similar oxidation rates. Factors that might explain why U(VI) bioreduction rate was an important control on uraninite particle size for CN32 but not for MR-1 are discussed. Ó 2008 Elsevier Ltd. All rights reserved.
1. INTRODUCTION The United States’ history of utilizing uranium resources for weapons and energy production has left an environmen*
Corresponding author. Fax: +1 814 863 7304. E-mail address:
[email protected] (W.D. Burgos).
0016-7037/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.gca.2008.07.016
tal hazard that poses a present and future threat to the integrity of soil and groundwater (Riley and Zachara, 1992). Uranium in contaminated aquifers is most prevalent in a soluble VI oxidation state (Langmuir, 1978). Microbiologically mediated reduction of U(VI) to a relatively insoluble U(IV) (UO2(s)—uraninite) is a strategy for immobilizing U (Langmuir, 1978; Taylor, 1979), and may be
Author's personal copy
4902
W.D. Burgos et al. / Geochimica et Cosmochimica Acta 72 (2008) 4901–4915
stimulated in contaminated subsurface environments by the addition of a physiological electron donor (Lovley et al., 1991). However, this strategy is only effective if the U(IV) remains reduced, since U(IV) may be oxidized back to U(VI) resulting in continued mobility of U (Finneran et al., 2002; Sani et al., 2005; Senko et al., 2005; Ginder-Vogel et al., 2006). In order for enzymatic uranium reduction to be implemented, the potential oxidation of immobilized U precipitates should be considered. A factor that could contribute to the stability of U after it has been bioreduced is the nature of the U(IV) precipitates. Biogenic U(IV) phases may range in size from monomeric U(IV) (Suzuki et al., 2002; Boyanov et al., 2007) to individual UO2 (uraninite) nanoparticles ranging in size from 0.9 to 5 nm (Suzuki et al., 2002; Sani et al., 2005; Marshall et al., 2006; Senko et al., 2007) to large aggregates of up to 100 nm (Abdelouas et al., 1999). Similarly sized uraninite nanoparticles have been produced through the abiotic reduction of U(VI) with sulfide (Beyenal et al., 2004) and green rust (O’Loughlin et al., 2003). The cellular location, particle size and aggregation state of biogenic U(IV) precipitates undoubtedly influence their susceptibility to oxidative remobilization (Beller, 2005; Marshall et al., 2006; Senko et al., 2007) and the potential for transport of nanoparticulate U(IV) in groundwater (Suzuki et al., 2002). Indeed, work by Fredrickson et al. (2002) suggested that extracellularly deposited U(IV) produced by Shewanella putrefaciens CN32 could be oxidized by Mn(III/IV) oxides, while periplasmic deposition of U(IV) could ‘‘protect” it from reaction with Mn(III/IV) oxides. Furthermore, larger, highly aggregated U(IV) nanoparticles (produced via relatively slow rates of U(VI) bioreduction by CN32) were more resistant to oxidation than relatively small, and less aggregated U(IV) nanoparticles (produced via relatively rapid rates of U(VI) reduction), suggesting that the rate of U(VI) reduction influences the nature of U(IV) particles and therefore their susceptibility to re-oxidation (Senko et al., 2007). The rate of U(VI) reduction is by no means the only factor controlling the nature of biogenic U(IV) precipitates. The geochemical conditions under which U(VI) reduction takes place, and the physiological state of the microorganisms mediating U(VI) reduction (among other factors) will undoubtedly exert control on the nature of biogenic U(IV) phases. For example, the presence of dissolved calcium may inhibit U(VI) bioreduction (Brooks et al., 2003), and calcium incorporation into U(IV) minerals (Burns, 1999) could influence their susceptibility to re-oxidation. Shewanella species are a relatively well-characterized group of U(VI)-reducing bacteria that contain a complex and apparently redundant system of electron transport for the reduction of U(VI) and other metals (e.g., Fe(III), Mn(III/IV)) (Wade and DiChristina, 2000; Beliaev et al., 2001; Shi et al., 2006; Bretschger et al., 2007). This redundancy may explain the retention of some metal-reducing activity and the differences in cellular location of U(IV) phases (i.e., periplasmic or extracellular) upon deactivation of genes whose products are known to reduce U(VI) (Marshall et al., 2006). Furthermore, the conditions under which Shewanella species are cultured appear to strongly influence outer mem-
brane composition and excretion of extracellular components (DiChristina et al., 2002; Fang et al., 2006; Ruebush et al., 2006; Biju et al., 2007; Neal et al., 2007), some of which are redox-active proteins (Marshall et al., 2006) and components of electrically conductive bacterial nanowires (Gorby et al., 2006). Therefore, the conditions under which a microorganism is cultured could influence the production and location of U(VI)-reducing enzymes, which, in turn, could influence the cellular location, particle size and aggregation state of U(IV) precipitates. In light of these observations, we assessed the nature of U(IV) precipitates produced by S. oneidensis MR-1 grown under carefully controlled conditions (Hill, 2007) designed to ‘‘poise” that organism for the production of bacterial nanowires that are believed to play a significant role in the reduction of U(VI) and other metals (Gorby et al., 2006). This is in contrast to previous experiments in which U(IV) precipitates were produced via U(VI) reduction by S. putrefaciens CN32 that was grown in a complex medium (Senko et al., 2007). We manipulated rates of U(VI) reduction by S. oneidensis MR-1 in bicarbonate buffer and an artificial groundwater by varying the cell density and the incubation temperature. Biogenic U(IV) precipitates were characterized by X-ray absorption spectroscopy (XAS) and electron microscopy. The reactivity of these U(IV) precipitates was characterized by their susceptibility to oxidation by dissolved oxygen. 2. EXPERIMENTAL SECTION 2.1. Background electrolytes All experiments were conducted in one of two background electrolytes: 30 mM NaHCO3 (pH 6.8, prepared under an 80:20% N2:CO2 atm; referred to hereafter as NaHCO3 buffer) or a 10 mM Piperazine-N,N0 -bis-(2-ethanesulfonic acid) (PIPES)-buffered artificial groundwater (pH 6.3, prepared under an 85:15% N2:CO2 atm; referred to hereafter as PBAGW buffer). Both buffers also included 5 mM sodium lactate and 1.2–1.6 mM uranyl(VI) acetate. The components of PBAGW (aside from PIPES) were selected to simulate the groundwater from Area 2 at the Oak Ridge National Laboratory Field Research Center (Brooks, 2001). The concentrations of all components of PBAGW buffer are presented in electronic annex EA-1. Critical component concentrations of PBAGW with respect to uranyl speciation included 5 mM NaHCO3, 4.2 mM CaCl2, and 0.01 mM KH2PO4. The PBAGW buffer also contained macronutrients for bacterial growth whereas the NaHCO3 buffer did not. 2.2. Cell cultivation Shewanella oneidensis MR-1 (referred to hereafter as MR-1) was cultured in a chemically defined medium similar to PBAGW except with the addition of vitamins, trace metals (Tanner, 1997), arginine, glutamine, serine (Myers and Nealson, 1988), and ferric(III) nitrilotriacetic acid, and with the exclusion of uranyl acetate (electronic annex EA-1). The PBAGW-defined growth medium was prepared under oxic
Author's personal copy
Uraninite nanoparticles produced by S. oneidensis MR-1
4903
conditions, pH adjusted to 6.8–7.1 with NaOH, and dispensed into serum bottles (165 ml average total volume). Fifty-four milliliter of medium were added to each bottle and autoclaved, then 1 ml of filter-sterilized ferric NTA (5 mM) was added to establish a 2:1 air:liquid ratio (vol:vol) in all bottles. A frozen cell suspension of MR-1 was thawed, transferred to a culture tube containing Tryptic Soy Broth-without dextrose (TSB-D) and incubated aerobically for 24 h at 30 °C and 50 rpm. This culture was used to inoculate the PBAGW defined growth medium, which was then incubated in the sealed bottles for 15.5 h at 30 °C and 50 rpm. Under these conditions, oxygen in the headspace limited growth (Hill, 2007). Cell growth reached late-log phase at ca. 16 h. TEM images of MR-1 cells at different incubation times revealed that these culture conditions induced the production of extracellular appendages that grew in size and abundance between 16 and 24 h (electronic annex Fig. EA-1-1). An incubation time of 15.5 h was selected to harvest cells before this extracellular material was produced.
the previous bioreduction incubation, and were conducted in 120 ml serum bottles sealed with thick butyl stoppers and aluminum crimp tops. After a bioreduction incubation time of 30–50 days, whereupon soluble U(VI) was removed from solution by MR-1 to less than 19–33 lM U(VI) under the four conditions, cell-uranium precipitates were pasteurized (70 °C for 30 min) to deactivate biological activity. No growth was observed when pasteurized cell-uranium precipitates were used to inoculate TSB-D. Cell-uranium precipitates were added to NaHCO3 buffer to achieve a total U concentration of 150–200 lM (initially predominantly U(IV)). Oxygen was provided to incubations by flushing the headspace of anoxic NaHCO3 buffer-containing serum bottles with filter-sterilized air. To maintain the pH of the bicarbonate buffered medium, 20% of the air headspace was removed and replaced with an equivalent volume of CO2. No change in medium pH resulted from these manipulations.
2.3. U(VI) bioreduction experiments
Soluble U(VI) was measured after centrifugation for 10 min at 14,100g and 20 °C. Total U(VI) (including soluble and solid-associated U(VI)) and total U (including U(VI) and U(IV)) were measured as described by Elias et al. (2003). Briefly, samples of well-mixed suspensions were removed from incubations by syringe in an anoxic glovebag (95:5% N2:H2 atm; Coy Laboratory Products Inc., Grass Lake, MI) and placed in 1 M anoxic NaHCO3 (pH 8.4). After extraction for 5 min, solids were removed by centrifugation and U(VI) was measured in the supernatant. To measure total U, suspension samples were placed in oxic 10% HNO3 under air for at least 12 h to oxidize all U(IV). U(VI) in all operational procedures was measured by kinetic phosphorescence analysis on a KPA-11 (ChemChek Instruments, Richland, WA) (Brina and Miller, 1992).
Bioreduction experiments were conducted in 1-L Pyrex bottles equipped with anoxic sampling caps. An anoxic cap consisted of a Balsh tube (bottom cut off) inserted through a cored thick black stopper fitted into the bottle top and secured with a plastic screw cap (center hole removed), and the protruding Balsh tube was sealed with a thick butyl stopper and aluminum crimp top. Reactors were filled with ca. 500 ml of anoxic buffer (either NaHCO3 or PBAGW) containing 5 mM sodium lactate, uranyl acetate stock solution was added to achieve an initial U(VI) concentration of 1.2–1.6 mM, and reactors were purged with a N2:CO2 gas mix corresponding to the buffer. Experiments were initiated by the addition of washed cell suspensions. Cell density was measured by absorbance at 600 nm and correlated to cell number by acridine orange direct counts. The A600 value at the time of cell harvest was 0.182 ± 0.021 based on over 500 trials and was continually validated. Cells were centrifuged (15 min at 3500g and 20 °C) under anoxic conditions and resuspended and washed three times in anoxic buffer (either NaHCO3 or PBAGW). Cell density and incubation temperature were varied to manipulate the rates of U(VI) bioreduction. With either buffer, ‘‘fast” bioreduction conditions were achieved using an MR-1 density of approximately 2.0 108 cell ml1 maintained at 37 °C, and ‘‘slow” bioreduction conditions were achieved using an MR-1 density of 1.0 108 cell ml1 maintained at 20 °C. All reactors were kept in darkness with no mixing except for sample collection. After cells were added, samples were periodically removed with sterile needle and syringe (in anoxic chamber) and soluble and total U(VI) concentrations were quantified as described below. 2.4. U(IV) oxidation experiments All biogenic uraninite oxidation experiments were conducted in 30 mM NaHCO3-buffered solution (pH 6.8, under 80:20% N2:CO2 atm) regardless of the buffer used in
2.5. U(VI) measurements
2.6. Electron microscopy For SEM, samples were prepared in a glove box following a previously published procedure (Zhang et al., 2007). Briefly, cell-mineral suspensions were fixed in anoxic 2.5% glutaraldehyde, placed on a glass cover slip, and mineral particles were allowed to settle onto the cover slip for 15 min. The particle-coated cover slips were gradually dehydrated in an ethanol series followed by critical point drying (CPD). All sample preparation, except CPD, was performed in an anoxic glovebag to minimize the exposure of samples to O2. Cover slips were mounted onto a SEM stub and Au coated for observation using a Zeiss Supra 35 FEG-VP SEM at an accelerating voltage of 10–15 kV. A short working distance (6–10 mm) and low beam current (30–40 mA) were used to achieve the best image resolution. A longer working distance (8.0 mm) and higher beam current (50–70 mA) were used for qualitative energy dispersive spectroscopy (EDS) analysis. For whole mount TEM, cells were fixed in 4% glutaraldehyde and placed on a 400 mesh carbon-sputtered formvar-coated copper grid. After 5 min, excess suspension was removed, the cells were negatively
Author's personal copy
4904
W.D. Burgos et al. / Geochimica et Cosmochimica Acta 72 (2008) 4901–4915
stained with 2% uranyl acetate, air dried and examined with a JEOL 1200 EX II TEM at 80 kV. For thin-section TEM, all sample preparation was performed in a Coy anoxic glovebag. Cell-uranium precipitates were fixed in anoxic 2.5% glutaraldehyde followed by 3 washes in anoxic water, gradual dehydration in an ethanol series and infiltration in LR White resin. Polymerized samples were sectioned to 70 nm on a microtome (Leica Ultracut UCT), and sections were mounted on 200 mesh copper grids coated with formvar support film sputtered with carbon. Sections were examined using JEOL 2010 high resolution transmission electron microscope (HR TEM) equipped with LaB0 6 fila˚ . Element operating at 200 kV with a resolution of 1.9 A mental analysis was performed using an Oxford EDS system equipped with a SiLi detector coupled to the TEM, and analyzed with ISIS software. Images were digitally collected using a Gatan CCD camera and analyzed using Gatan Digital Micrograph. d-spacings of selected area diffraction (SAED) ring patterns were evaluated by Desktop Microscopist software. 2.7. X-ray absorption spectroscopy XAS measurements were made at the Materials Research Collaborative Access Team (MR-CAT) sector 10ID beamline (Segre et al., 2000) of the Advanced Photon Source at Argonne National Laboratory. The XAS spectra were collected in transmission mode using quick-scanning of the monochromator. X-ray absorption near edge structure (XANES) spectra from the cell-uranium precipitates were used to determine the average valence state within the samples. The X-ray absorption edge energy was calibrated by collecting the reference spectrum from hydrogen uranyl phosphate (U(VI) Std) during the collection of each spectrum. All data sets were accurately aligned in energy using the derivative of the edge of the U(VI) standard. A linear combination fitting (LCF) of the U(VI) Std and a biogenic nanoparticulate UO2 standard (U(IV) Std) (O’Loughlin et al., 2003) was used to determine the approximate U(VI) to U(IV) ratio in the sample XANES spectra. This approach assumes that the standard spectra represent the uranyl species in the samples. This assumption can lead to a systematic uncertainty for these samples because we do not know the U(VI) species. As a result we used our solid U(VI) phosphate standard and an aqueous uranyl(VI) carbonate standard, as likely U(VI) components of the samples. The valence state of the samples was verified by using the number of axial oxygen atoms of the uranyl (Oax) as determined by analysis of the EXAFS data. This method is less model-dependent because it does not assume prior knowledge of the uranyl species. At the end of the bioreduction incubation period, cell uranium-precipitates were stored at 4 °C before shipment to ANL. Samples for XAS were referred to their buffer and bioreduction condition as follows: NaHCO3 ‘‘fast”, NaHCO3 ‘‘slow”, PBAGW ‘‘fast”, and PBAGW ‘‘slow”. Samples (approximately 40 mg cell-uranium precipitates) were mounted as moist pastes on filter paper or as centrifuged pastes mounted in a plastic holder, covered with Kapton film and sealed with Kapton tape. All sample prep-
aration was performed in a Coy anoxic glovebag and all samples were stored in the chamber prior to analysis. Samples mounted in the holders were exposed to the atmosphere for less than 1 min before being mounted for XAS measurements in a free-flowing N2 environment to limit possible sample oxidation. These XAS sample holders have been shown to maintain anoxic integrity, when exposed to an oxic environment, for at least 8 h. Uranium XANES spectra were collected in 30-s intervals consecutively for 5 min. No radiation-induced changes to the XANES spectra were observed at the 30-s intervals of data collection. Extended X-ray absorption fine-structure (EXAFS) spectra were collected in 5-min intervals, averaged and then the background was removed using the programs Athena (Ravel and Newville, 2005) and IFEFFIT (Newville, 2001). The background removal parameter Rbkg ˚ (Newville et al., 1993). The effect of doublewas set to 1.0 A electron excitations like those described by Hennig (2007) were removed from these spectra as described elsewhere (Ravel and Kelly, in press). The EXAFS spectra were modeled using theoretical models from atomic clusters of uraninite (Wyckoff, 1960) as input for FEFF 7.02 (Zabinski et al., 1995). The EXAFS parameters were optimized to the measured spectra using the program FEFFIT (Stern et al., 1995). The EXAFS parameters are the amplitude reduction factor ðS20 Þ, an energy shift to align the theoretical and measured spectra (DE0), the number of atoms within a shell or coordination number (CN), distance between the uranium atom and the neighboring atoms (R), and the mean square displacement in that distance (r2). Details of the EXAFS model are provided in electronic annex EA-2. 3. RESULTS 3.1. U(VI) bioreduction kinetics The rate of U(VI) bioreduction by MR-1 was observed to be first-order with respect to the soluble U(VI) concentration (Fig. 1) according to: Rred ¼ d½UðVIÞ=dt ¼ k red ½UðVIÞðsolÞ
ð1Þ
where kred is the first-order bioreduction rate constant (d1). The rate of U(VI) bioreduction was certainly affected by other factors such as lactate concentration, cell/U(VI) ratio, cell/lactate ratio, temperature and (non-)mixing conditions but these factors are purposefully excluded from Eq. (1) for a simplified comparison. As designed, differences among the rates of U(VI) bioreduction were dependent on cell concentration and temperature (Table 1). For example, kred increased by a factor of 24 in NaHCO3 buffer when the cell concentration was doubled and the incubation temperature increased by 17 °C. Similarly, but to a lesser extent, kred increased by a factor of 7.6 in PBAGW buffer between the ‘‘fast” and ‘‘slow” bioreduction conditions. Liu et al. (2002) reported a first-order U(VI) bioreduction rate constant of 28.4 d1 for MR-1 in experiments conducted with 0.50 mM uranyl chloride, 5 mM Na-lactate, and 2.0 108 cell ml1 at 30 °C in 30 mM NaHCO3 buffer. While this is reasonably close to our reported rate constant of 11 d1 for the ‘‘fast” condition in NaHCO3 buffer, the
Author's personal copy
Uraninite nanoparticles produced by S. oneidensis MR-1
NaHCO3 "Fast"
PBAGW "Fast"
1. 6
NaHCO3 "Slow"
PBAGW "Slow"
Total [U(VI)] (mM)
Total [U(VI)] (mM)
1. 2
4905
0. 9
0. 6
0. 3
1. 2
0. 8
0. 4
0
0 0
2
4
6
8
10
0
12
10
20
NaHCO3 "Fast"
40
0. 9
PBAGW "Fast"
Soluble [U(VI)] (mM)
Soluble [U(VI)] (mM)
30
1. 2
1. 2
NaHCO3 "Slow"
0. 6
0. 3
PBAGW "Slow"
0. 9
0. 6
0. 3
0
0
0
2
4
6
8
10
12
0
10
Time (d )
20
30
40
Time (d )
Fig. 1. U(VI) bioreduction kinetics by Shewanella oneidensis MR-1 in (a and c) NaHCO3 buffer and (b and d) PBAGW buffer. With either buffer, ‘‘fast” bioreduction conditions were achieved using a cell concentration of 2.0 108 cell ml1 maintained at 37 °C, and ‘‘slow” bioreduction conditions were achieved using a cell concentration of 1.0 108 cell ml1 maintained at 20 °C. Measured triplicate data (symbols) and first-order rate models (lines).
Table 1 First-order rate constants for U(VI) bioreduction (kred, Eq. (1)) by Shewanella oneidensis MR-1, and subsequent first-order rate constants for uraninite oxidation (kox, Eq. (2)) by dissolved O2 Sample name
kred (d1)
r2
P-value
kox (d1)
r2
P-value
NaHCO3 ‘‘fast” NaHCO3 ‘‘slow” PBAGW ‘‘fast” PBAGW ‘‘slow”
11 0.45 0.99 0.13
0.97 0.87 0.98 0.86
