JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 25 January 2007
Published in J Environ Qual 36:363-372 (2007)
DOI: 10.2134/jeq2006.0058
© 2007 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Stewart, B. D.
Right arrow Articles by Fendorf, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stewart, B. D.
Right arrow Articles by Fendorf, S.
Agricola
Right arrow Articles by Stewart, B. D.
Right arrow Articles by Fendorf, S.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Other Environmental Contamination

TECHNICAL REPORTS

Heavy Metals in the Environment

Quantifying Constraints Imposed by Calcium and Iron on Bacterial Reduction of Uranium(VI)

Brandy D. Stewart, Jim Neiss and Scott Fendorf*

Dep. of Geological and Environmental Sciences, Stanford Univ., Stanford, CA 94305

* Corresponding author (fendorf{at}stanford.edu)

Received for publication February 11, 2006.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Uranium is a redox active contaminant of concern to both human health and ecological preservation. In anaerobic soils and sediments, the more mobile, oxidized form of uranium (UO22+ and associated species) may be reduced by dissimilatory metal-reducing bacteria. Despite rapid reduction in controlled, experimental systems, various factors within soils or sediments may limit biological reduction of U(VI), inclusive of competing electron acceptors and alterations in uranyl speciation. Here we elucidate the impact of U(VI) speciation on the extent and rate of reduction, and we examine the impact of Fe(III) (hydr)oxides (ferrihydrite, goethite, and hematite) varying in free energies of formation. Observed pseudo first-order rate coefficients for U(VI) reduction vary from 12 ± 0.60 x 10–3 h–1 (0 mM Ca in the presence of goethite) to 2.0 ± 0.10 x 10–3 h–1 (0.8 mM Ca in the presence of hematite). Nevertheless, dissolved Ca (at concentrations from 0.2 to 0.8 mM) decreases the extent of U(VI) reduction by ~25% after 528 h relative to rates without Ca present. Imparting an important criterion on uranium reduction, goethite and hematite decrease the dissolved concentration of calcium through adsorption and thus tend to diminish the effect of calcium on uranium reduction. Ferrihydrite, in contrast, acts as a competitive electron acceptor and thus, like Ca, decreases uranium reduction. However, while ferrihydrite decreases U(VI) in solutions without Ca, with increasing Ca concentrations U(VI) reduction is enhanced in the presence of ferrihydrite (relative to its absence)—U(VI) reduction, in fact, becomes almost independent of Ca concentration. The quantitative framework described herein helps to predict the fate and transport of uranium within anaerobic environments.

Abbreviations: {Delta}Grxn, change in Gibb's free energy of reaction • DMRB, dissimilatory metal reducing bacteria • PIPES, 1,4-piperazinediethanesulfonic acid • S. putrefaciens, Shewanella putrefaciens • TSB, tryptic soy broth


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
MINING and nuclear enrichment activities throughout the last century have created a legacy of uranium and co-contaminants in the environment. In many cases, uranium, a contaminant of concern due to its toxic effects and radioactivity, has migrated from waste disposal sites to surrounding soils and sediments where its fate is often controlled by redox processes, and in particular the proportion of U(VI) to U(IV). In its oxidized form, U(VI), uranium exists as the relatively mobile uranyl ion (UO22+), particularly in carbonate-bearing solutions where it forms carbonato aqueous complexes. Conversely, under reducing conditions, uranium resides in a lower oxidation state, U(IV), and commonly forms relatively insoluble solid phases such as uraninite (UO2) (Cochran et al., 1986). The process of U(VI) reduction and concurrent immobilization is often facilitated by dissimilatory metal-reducing bacteria (DMRB) that couple the oxidation of carbon or H2 with the reduction of oxidized metals (Gorby and Lovley, 1992; Liu et al., 2002; Lovley et al., 1991). Owing to a general decrease in uranium solubility, and hence mobility, reduction processes plays an important role in the cycling, and in some cases natural attenuation, of uranium (Lovley et al., 1991; Brooks et al., 1999; Fredrickson et al., 2000). In fact, microbially mediated reduction has been studied as a remediation strategy for uranium-bearing ground waters.

Hexavalent uranium is readily reduced to U(IV) by DMRB in aqueous systems dominated by uranyl-carbonato complexes such as UO2(CO3)22– and UO2(CO3)34– (Lovley et al., 1991; Clark et al., 1995; Brooks et al., 2003), which, in the absence of Ca, account for ~95% of U(VI) in solution in a typical ground water (Abdelouas et al., 1998a)—a pH of ~6.9, temperature of 16°C, and HCO3– and U(VI) concentrations of 1 mM and 4 µM. Inclusion of Ca, however, results in two ternary calcium-uranyl-carbonato species, CaUO2(CO3)32– and Ca2UO2(CO3)3, emerging as dominant aqueous species, accounting for 99.6% of U(VI) in solution (Bernhard et al., 1996; Kalmykov and Choppin, 2000; Bernhard et al., 2001; Kelly et al., 2003) and alters microbial reduction of U(VI). Brooks et al. (2003), for example, observed a 40% decrease in U(VI) reduction by Shewanella putrefaciens over a 30-h period upon the addition of 0.45 mM Ca, as compared with reduction without Ca.

In soils and sediments subject to anaerobic conditions, competing terminal electron acceptors (TEA), such as nitrate and ferric iron (the latter in the form of Fe(III) (hydr)oxides), can diminish both the rate and extent of microbial U(VI) reduction (Abdelouas et al., 1998b; Wielinga et al., 2000). Iron has an unusual impact on uranium in that it can serve as an oxidant or reductant depending on the specific geochemical conditions (Ginder-Vogel et al., 2006), and can even switch between being an oxidant and reductant during the course of a single incubation (see, for example, Wan et al., 2005). Redox couples of Fe(III/II) and U(VI/IV) are comparable and thus the specific species of either element along with the chemical gradients dictate which serves as the oxidant and reductant. Depending on the specific species of uranyl, the U(IV/IV) redox couple can change more than 200 mV, with UO22+ leading to the highest and Ca2UO2(CO3)3 the lowest—a shift in U(VI) speciation toward the calcium ternary complex thus makes U(IV) most susceptible to oxidation. Ferrihydrite has the highest redox potential of the Fe(III) (hydr)oxides and consequently has the greatest likelihood to compete as an electron acceptor in microbial respiration and serve as an oxidant of U(IV). Wielinga et al. (2000) for example, observed a 52% decrease in uranyl reduction by Shewanella alga, strain BrY, in the presence of ferrihydrite, compared with only uranyl in solution, over a period of 10 h, while, in contrast, neither goethite nor hematite impacted uranyl reduction (Wielinga et al., 2000).

Limiting our ability to predict U reduction is the absence of rate information describing the impacts of uranyl speciation, particularly those involving ternary complexes of Ca, or soil/sediment matrix effects. The aim of this study is thus to provide a quantitative framework on bacterial reduction of U(VI) at varying concentrations of dissolved calcium and to determine the impact of Fe(III) (hydr)oxides, varying systematically in free energy of formation, on reduction rates. The results of this study provide an understanding of geochemical limitations on microbial uranium reduction imposed by Ca and modifications induced by Fe (hydr)oxides within the mineral matrix of soils and sediments.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Cell Culturing and Harvesting
Shewanella putrefaciens, strain CN32, a facultative, dissimilatory metal-reducing bacterium, was used to evaluate the effect of calcium on biological uranyl reduction in batch studies involving various Fe(III) (hydr)oxide-coated quartz sands. The culture was obtained from the American Type Culture Collection (ATCC catalog number BAA-453), cultured aerobically in tryptic soy broth (TSB) until late log phase, and then frozen in 20% glycerol at –80°C where it was stored until needed.

Shewanella putrefaciens was grown aerobically by placing 1 mL of frozen stock culture in 100 mL of TSB. The culture was placed on a shaker at room temperature for 12 h (late log phase) before it was transferred into the same medium and incubated while shaking for an additional 12 h. It was then centrifuged at 4500 rpm for 12 min, resuspended in 100 mL of bicarbonate buffer, and centrifuged a second time. Finally, the culture was suspended in 100 mL of PIPES buffered artificial ground water medium for 30 min until the start of the experiment. The artificial ground water medium consisted of the following ingredients (in mg L–1): KHCO3, 380; KCl, 5; MgSO4, 50; NaCl, 30; NH4Cl, 0.95; KH2PO4, 0.95; and 1 mL Wolfe's mineral solution. Sodium lactate was added to provide a final concentration of 3 mM. The media was buffered with 10 mM PIPES (1,4-piperazinediethanesulfonic acid) at pH 7 and made anoxic by boiling and cooling under a stream of O2–free N2/CO2 (80:20) gas.

Synthesis of Iron (Hydr)oxide-Coated Sands
Ferrihydrite was prepared according to the method described by Brooks et al. (1996). A solution of ferric chloride was rapidly titrated with sodium hydroxide over a period of approximately 10 min until a pH of 7 was reached. Chloride and Na were then removed from the ferrihydrite with sequential rinses.

Goethite was prepared using a slightly modified version of Atkinson et al. (1967). Ferric nitrate was combined with concentrated sodium hydroxide in the absence of CO2. An effort was made to keep CO2 out of the product during initial synthesis but not during the dialysis purification stage; CO2 adsorbed during product synthesis can significantly impact the zero point of charge of the product (Van Geen et al., 1994). Sodium hydroxide was slowly pumped into the ferric nitrate solution over a period of several hours until a pH of 12 was reached during continuous stirring. The slurry was then placed in a 60°C oven for 24 h, and finally salts (and in particular, residual nitrate) were removed by dialysis for a period of approximately 10 d.

Finally, hematite was prepared following the method described by Schwertmann and Cornell (2000) in which a concentrated ferric nitrate solution is added gradually over a period of 4 h to boiling, distilled water. The solution was stirred constantly. The product was then cooled overnight and purified by dialysis for approximately 7 d to remove excess salts (in particular nitrate). X-ray diffraction analysis was performed on all three Fe (hydr)oxides to ensure purity of product.

The iron (hydr)oxides were prepared individually and then used to coat quartz sand as reported previously (Brooks et al., 1996; Hansel et al., 2003). Briefly, the iron (hydr)oxide slurry was poured over the quartz sand and mixed thoroughly by hand. The mixture was then left to dry for 48 h before being washed with deionized water. Iron concentration on the sands is approximately 10 g kg–1 (1% by weight).

Uranium(VI) Reduction Reactions
To investigate the effects of aqueous calcium concentration on uranium reduction, we assembled batch systems containing S. putrefaciens, Fe(III) (hydr)oxide-coated sand, artificial ground water medium, uranyl acetate, and varying concentrations of CaCl2, under anoxic conditions in a glovebag (Coy Laboratory Products) with a N2 (95%)/H2 (5%) atmosphere. Lactate was provided in the ground water medium as a carbon source and electron donor at a concentration of 3 mM. All solutions were made anoxic by boiling and cooling under a stream of N2 (80%)/CO2 (20%) gas. Each 125 mL serum vial contained 1.0 g of Fe (hydr)oxide-coated sand, {approx}107 cells of S. putrefaciens, and 105 mL of ground water media. Anoxic stock solutions containing U-acetate, CaCl2, and KHCO3 were allowed to equilibrate overnight before being sterilely injected into the batch bottles. Each batch system began with 0.168 mM U, added in the form of uranyl acetate, UO2(C2H3O2)2, and a Ca2+ concentration of either 0 mM, 0.4 mM, 0.6 mM, or 0.8 mM (added as CaCl2). The batch systems were assembled in a glovebag and then shaken at room temperature outside the glovebag between samplings; bottles were brought into the glovebag during sampling. A separate experiment was run for each of the three Fe (hydr)oxides (goethite, ferrihydrite, and hematite) as well as an experiment with no Fe (no-Fe control). All of the aforementioned biotic systems were conducted in triplicate and error values were calculated as standard deviation of the three measurements divided by the square root of the number of measurements. Abiotic controls for each of the iron (hydr)oxides, and one without iron, were conducted in duplicate.

Aqueous Analyses
Aqueous samples were withdrawn using sterile syringes and filtered through 0.2-µm membranes inside a glovebag at hourly intervals for the first 8 h of the experiment. Subsequent samples were collected approximately every 12 to 24 h for the next 7 d. A final sample was collected on Day 22. All samples were analyzed by inductively coupled plasma–optical emission spectrophotometry (TJA/IRIS Advantage) to determine the concentration of total U, Fe, and Ca in solution. The ferrozine method (Stookey, 1970) was used to measure Fe(II).

Preliminary batch experiments indicated that an aqueous Ca concentration of approximately 0.9 mM strongly impacted the biotic reduction of U(VI); to determine progressive impacts on U(VI) reduction, we therefore varied Ca concentrations from 0 to 0.8 mM in 0.2-mM increments. All batch systems containing Fe (hydr)oxide sands were run at least 22 d, which appeared adequate to reach approximate steady-state conditions with respect to U(VI) reduction.

Uranyl speciation calculations were performed using the chemical speciation program Visual Minteq and sources of thermodynamic data specified in Table 1. Predicted speciation of uranyl was examined for Ca concentrations ranging from 0 to 1 mM given aqueous conditions of the experiments (pH = 7, [CO3]T = 3.8mM, [U(VI)] = 0.168mM). For the purpose of the thermodynamic calculations, biogenic uraninite is considered as UO2 (am), as its physical characteristics are more similar to UO2(am) than they are to UO2(c) (Wan et al., 2005).


View this table:
[in this window]
[in a new window]

  		Table 1. Redox reactions for various U(VI) and Fe(III) species coupled with lactate (ferrihydrite, Fe(OH)3; goethite, FeOOH; hematite, Fe2O3).

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Uranium(VI) Speciation
Formation of the ternary Ca2UO2(CO3)3 complex progressively increases, becoming the dominant uranyl species, over a Ca concentration range of 0 to 1 mM. At a Ca concentration of 0.2 mM, 23% of U in solution is associated with the Ca ternary complex at pH 7 with [CO3]T = 3.8mM and [U(VI)] = 0.168 mM (Table 2)—uranyl-carbonato complexes account for 74% of the U(VI) species under these conditions. An increase in Ca concentration to 0.8 mM results in 89% of the uranyl residing within the uranyl-calcium-carbonato complexes. This shift in speciation from a system dominated by uranyl-carbonato to one dominated by uranyl-calcium-carbonato complexes correlates with the observed decrease in U(VI) reduction by S. putrefaciens (Fig. 1). Reduction of the various uranyl species, when coupled with lactate oxidation, results in appreciable shifts in apparent energy yield but does not systematically decrease with the ternary Ca complexes (Table 1).


View this table:
[in this window]
[in a new window]

  		Table 2. Uranyl speciation at various Ca(aq) concentrations between 0 and 1 mM (pH = 7, [CO3]T = 3.8mM, [U(VI)] = 0.168mM, T = 25°C, P = 1 atm).

 

Figure 1
View larger version (30K):
[in this window]
[in a new window]

  		Fig. 1. Temporal changes in uranyl concentration induced by S. putrefaciens at Ca concentrations ranging from 0 to 0.8 mM in the presence of (A) ferrihydrite, (B) goethite, (C) hematite, and (D) Fe (hydr)oxide-free system.

 
Uranium(VI) Reduction
Reduction of U(VI) by S. putrefaciens transpires at all Ca concentrations and with all three types of ferric (hydr)oxides investigated; however, both the rate and the extent of U(VI) reduction are dependent on Ca concentration—with an inverse relation between reduction rate and Ca concentration (Fig. 1). With ferrihydrite, U(aq) is reduced to 24% of its original concentration of 0.168 mM after 528 h when no Ca is present in solution. In contrast, addition of Ca (initial concentrations of 0.4, 0.6, and 0.8 mM) results in a 50% reduction in U(aq) after 528 h. In the case of goethite, U(aq) is reduced to 2% of starting concentration after 528 h with no Ca, while with the addition of Ca reduction is diminished and thus 20 to 35% of the initial U (0.168 mM) remains in solution after 528 h. With hematite, reduction ranges from 100% (no Ca) to 80% depending on the Ca concentration (Fig. 1). The addition of Ca to the no-Fe control system has the most dramatic affect on U(VI) reduction. After 528 h of reaction, the extent of U reduction decreases from 90 to 55% of the initial concentration with the addition of Ca—consistent with the findings of Brooks et al. (2003).

Iron(III) Reduction
The concentration of Fe in the aqueous phase increased during the reaction period for ferrihydrite at all Ca concentrations (Fig. 2). The concentration of Fe generated after 528 h ranges from 0.11 mM in the absence of Ca, a concentration and rate comparable to that reported previously (Wielinga et al., 2000), to 0.29 mM in the presence of 0.8 mM Ca. Dissolved Ca appears to correlate with the concentration of Fe(II) produced, supporting the premise that Ca renders U(VI) less favorable as a terminal electron acceptor thus making Fe(III) the dominant electron acceptor. In contrast to the ferrihydrite system, negligible amounts (<0.2 µmol L–1) of aqueous Fe are detected in goethite or hematite systems (and in the no-Fe controls) at any Ca concentration; minimal Fe reduction transpired (less than the re-adsorption capacity of Fe(II) on goethite or hematite). The differences observed among the Fe (hydr)oxides are consistent with the surface area and free energy yields, where Fe is, at least initially, more available for bioreduction when present as ferrihydrite than either goethite or hematite (Lovley, 1991; Hansel et al., 2004).


Figure 2
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 2. Temporal changes in Fe(aq) concentration induced by S. putrefaciens in the presence of ferrihydrite.

 
Time-Dependence of Uranium(VI) Reduction
Uranium is removed from solution via reduction by S. putrefaciens with subsequent precipitation of biogenic UO2(s) in all systems studied but at varying rates. The rate of U removed from solution (expressed as the fraction of initial U) conforms to a first-order rate expression (Fig. 3). Pseudo first-order rate constants, k (h–1), are larger for systems without than those with Ca (Table 3), with the variation being most pronounced for the Fe (hydr)oxide-free system where the rate constant decreases from 13.7 ± 0.3 (x 10–3) h–1 in the absence of Ca to 0.8 ± 0.0 (x 10–3) h–1 when 0.8 mM Ca is added to the system. It follows that the rate constant in the presence of ferrihydrite changes the least, ranging from 4.1 ± 0.2 (x 10–3) h–1 for 0 mM Ca to 2.4 ± 0.2 (x 10–3) h–1 for 0.8 mM Ca. Pseudo first-order rate constants (Table 3) are linearly correlated to calcium concentrations between 0 and 1 mM (Fig. 4), albeit that relations result for systems having no iron, those with ferrihydrite, and those with goethite or hematite.


Figure 3
View larger version (28K):
[in this window]
[in a new window]

  		Fig. 3. First-order rate plots of uranyl depletion for (A) ferrihydrite, (B) goethite, (C) hematite, and (D) Fe (hydr)oxide-free system.

 

View this table:
[in this window]
[in a new window]

  		Table 3. Pseudo-first order rate coefficients (kobs) at various calcium concentrations for the different Fe(III) (hydr)oxide systems; equations describing the dependence of the overall rate coefficients on Ca concentration are provided.{dagger}

 

Figure 4
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 4. Relation between observed pseudo first-order rate coefficient and aqueous Ca concentrations for ferrihydrite, goethite, hematite, and Fe (hydr)oxide-free systems.

 
The effect of Ca on the reduction of uranyl is most pronounced in the absence of Fe (hydr)oxides but is appreciable in all cases studied (Fig. 1). Differences in behavior with respect to U(VI) reduction as a function of Ca concentration in the presence of ferrihydrite, goethite, and hematite are evident after 9 h of incubation (Fig. 5). By 97.5 h, all three Fe (hydr)oxide systems are responding similarly to each other and continue to do so until the final time point of 528 h where their behavior is again divergent (Fig. 5). After 9 h, Ca affects the ferrihydrite system most profoundly, as is illustrated by a larger Formula value (0.39 mM–1 Ca) for ferrihydrite compared with goethite and hematite (0.16 and 0.17 mM–1, respectively). However, by 97.5 h the slopes of the three systems are similar, implying that Ca is evoking a similar effect on uranyl reduction regardless of Fe (hydr)oxide type. By 528 h, less Ca is present in the aqueous phase for the goethite and hematite systems than the ferrihydrite system despite the same starting Ca concentrations (Table 4). This implies Ca sorption to goethite and hematite that serves to "shield" the microorganisms from the full impact of Ca in solution and the resulting shift in uranyl speciation. Calcium sorption on Fe (hydr)oxides thus decreases the aqueous concentration and concomitantly the proportion of the ternary complex. For instance, when the concentration of Ca in solution drops from 0.4 to 0.2 mM (Table 4) in the presence of hematite, the percentage of U(VI) associated with the ternary complex decreases from 53 to 35%. At Ca levels above 0.4 mM, uranyl reduction is similar for the Fe (hydr)oxide-free system as it is for the systems containing Fe(hydr)oxides suggesting saturation of Ca-suitable sorption sites. Additionally, after 528 h all systems have Formula values approaching zero suggesting near steady-state conditions.


Figure 5
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 5. Uranium remaining in solution as a function of calcium concentrations for Fe(III) (hydr)oxide type at (A) 9 h, (B) 97.5 h, (C) 171 h, and (D) 528 h.

 

View this table:
[in this window]
[in a new window]

  		Table 4. Calcium(aq) concentrations after 9 and 528 h of reaction for different Fe(III) (hydr)oxides.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Influence of Calcium on Uranium(VI) Reduction
There are several possible mechanisms by which Ca may diminish U reduction. First, Ca may render the reduction of uranyl less thermodynamically favorable. Second, Ca may kinetically limit reduction by yielding a dominant complex that either has poor-orbital overlap with the U reductase or hinders site accessibility. Finally, Ca could have a toxic effect on the microorganisms. Previous studies have eliminated the possibility of toxic effects of Ca or Ca-U-CO3 to the bacteria (Brooks et al., 2003). Thermodynamic considerations illustrate that there is less than 10% difference in energy gained between the reduction of uranyl-carbonato and uranyl-calcium-carbonato species (Table 5). The Gibb's free energy of reaction ({Delta}Grxn) for reduction of (UO2)2CO3(OH)3– (–157.3 kJ mol–1), the dominant aqueous species in the absence of Ca, and Ca2UO2(CO3)3 (–144.0 kJ mol–1), the dominant species in the presence of Ca, are only 8% different—and both are highly favorable. Moreover, even with a Ca concentration of 1 mM, the free energy of reaction for reduction of U(VI) is comparable to the energy yield from the reduction of ferrihydrite (under the conditions of this study). This gives rise to the hypothesis that the reduction of U(VI) by S. putrefaciens may have an unfavorable orbital symmetry, be sterically hindered, or have a high activation energy associated with uranyl-calcium-carbonato complex dissociation.


View this table:
[in this window]
[in a new window]

  		Table 5. Gibb's free energy of reaction ({Delta}Greaction) for Fe(III) and U(VI) reduction coupled with lactate oxidation at varying Ca(aq) concentrations. Values are calculated for dominant aqueous species as indicated in Table 2 and are adjusted for experimental conditions (pH = 7, [CO3]T = 3.8mM, [U(VI)] = 0.168mM).

 
Rate Determining Steps in Uranium(VI) Reduction
Although a multitude of U(VI) species reside under the conditions evaluated within this study, there are two dominant species, Ca2UO2(CO3)3, (UO2)2CO3(OH)3–, under the geochemical conditions of this study; a third species, UO2(CO3)22–, resides at appreciable fractions of the total U. Considering these three species, reduction reactions coupled with lactate oxidation (Table 1) are described by the stoichiometries of reactions 1Go through 3.

Formula 1[1]

Formula 2[2]

Formula 3[3]
Given that the reactions are unidirectional, the pH was maintained constant, and U(VI) reduction proceeds independent of lactate (enzyme substrate) concentration under the conditions and reaction period studied (Fig. 3), the operative rate determining, microbially-mediated, reduction reactions can be simply expressed:

Formula 4[4]

Formula 5[5]
where in reaction 4 (UO2)x(CO3)y represents the combined contributions of the carbonato species (UO2)2CO3(OH)3– and UO2(CO3)22–, both of which are present in appreciable proportions under the reaction conditions examined.

Under conditions where the ternary Ca complex of U(VI) dominates the aqueous speciation, reduction may proceed either by reaction 5 or through the small proportion of carbonato complexes without calcium (reaction 4). In the latter case, inter-species conversion must then be described as well,

Formula 6[6a]
Given the excess concentration of bicarbonate and calcium relative to uranium, and the constant pH, the inter-species conversion simplifies to:

Formula 7[6b]
The reactions expressed in reaction 4 and 5 thus appear to represent the rate-controlling processes for uranyl reduction, with inter-conversion of uranyl species potentially important within Ca-containing solutions. On the basis of U(VI) reduction rates in the presence and absence of Ca, reaction 4 proceeds at a much faster rate than reaction 5 (i.e., k1 > > k2). With the selective removal of the carbonato complexes, reaction 6Go would proceed in the forward direction during reduction. Reaction 4 thus expresses the rate-controlling reaction for non-Ca-bearing solutions, while reaction 5 and the coupling of reactions 4 and 6Go would operate in series (reaction 7) as the rate influencing reactions in Ca-bearing solutions.

Formula 8[7]
Rate data for the various systems conform reasonably well to a pseudo first-order kinetic expression dependent on the total U(VI) concentration and having an observed (or overall) rate coefficient representing various factors.

Formula 9[8]
The observed rate coefficient, kobs (Table 3; Fig. 4), for U(VI) reduction in the absence of Ca is thus proportional to k1—the proportionality coefficient being the fraction of total U residing in the carbonato complex. With the exception of systems with ferrihydrite, which acts as a competing electron acceptor, the averaged k1 value (k1 = kobs/Kp, where Kp is the proportion of carbonato species and is equal to 1 in the absence of Ca) is 1.18 x 10–2 h–1, giving the following operative rate expression.

Formula 10[9]
In Ca-bearing solutions, the rate expression becomes more convoluted and would be described by an observed rate coefficient encompassing the initial proportion of the U-carbonato complex, its rate of regeneration, and the rate of Ca ternary complex reduction. The rate expression for the reduction of U(VI) is therefore

Formula 11[10]
The decrease in rate coefficient with increased Ca concentrations (Fig. 4) indicates that k1 > > k3, leading to the simplification for sustained reduction

Formula 12[11]
which represents contributions from reaction 5 and 7 to U(VI) reduction. Decreasing rate coefficients with increased Ca concentrations described in Table 3 thus represent the combined contributions of k2 and k3 and increased proportions (effective concentration) of the Ca ternary complex. For Eq. [10], the observed rate coefficient would be proportional to k2,3 (where k2,3 = k2 + k3), with the proportionality coefficient (Kp) again being the fraction of total U residing in the ternary complex (Kp = –1.15[Ca] + 0.97) and can be described as ln k2–3 = 30.5[Ca2UO2(CO3)3] – 7.18.

Impact of Iron(III) (Hydr)oxides on Uranium(VI) Reduction
Ferric (hydr)oxides impart an interesting complexity on U reduction. On the one hand, they can serve as competing electron acceptors of U(VI) and as oxidants of U(IV) (under specific reaction conditions). On the other hand, they can also regulate the dissolved concentration of Ca, decreasing the proportion of the Ca-U ternary complex, and enhancing U(VI) reduction.

Ferrihydrite has the highest redox potential of the Fe(III) phases investigated here, and indeed it is the only Fe(III) (hydr)oxide to decrease the extent of U(VI) reduction. As noted above, ferrihydrite may either compete as an electron acceptor in microbial respiration or may act as an oxidant of biogenic UO2. A series of recent reports indicate reoxidation of uraninite by Fe(III) at high concentrations of HCO3– resulting from bacterial respiration (Sani et al., 2005; Senko et al., 2005; Wan et al., 2005). However, under the conditions of our study (HCO3– < 6 mM, 0.168 U(VI), and pH 7), even the oxidation of biogenic UO2 to Ca2UO2(CO3)3 by ferrihydrite, the most viable reaction, would be thermodynamically favorable only at Fe(II) concentrations less than 0.025 mM Fe(II) during the initial stages of reduction and 0.050 mM at late stages. These Fe(II) values are exceeded during the initial stages of reaction (t < 100 h) and reoxidation of U(IV) therefore does not appear operative in this system. Instead, ferrihydrite appears to serve as a competing electron acceptor, which in the absence of Ca retards the rate of U(VI) reduction.

Ferrihydrite, however, is not a static phase during Fe(II) production and instead undergoes a cascade of secondary reactions, dominantly leading to goethite or magnetite depending on the Fe(II) concentrations (Hansel et al., 2003). With progressive incubation, the ferrihydrite system responds similarly to goethite and hematite (Fig. 5), consistent with a shift in mineralogy from ferrihydrite to goethite during dissimilatory iron reduction (Hansel et al., 2004). Hansel et al. report a threshold limit of 0.4 mM (1mmol Fe(II)/g-ferrihydrite) below which ferrihydrite converts primarily to goethite and above which it converts to both goethite and magnetite (Hansel et al., 2003). Iron(II) levels produced in the ferrihydrite system do not exceed this threshold when Ca is present until 171 h (0.52 mM, 1.3mmol Fe(II)/g-ferrihydrite), and would therefore lead to pronounced generation of goethite.

In addition to serving as a competing electron acceptor, Fe(III) (hydr)oxides influence U(VI) reduction through adsorption of Ca, diminishing its aqueous concentrations and thereby decreasing the proportion of the apparently less reactive Ca2UO2(CO3)3 species. The impact is pronounced with goethite and hematite (Fig. 1). Thus, the inhibitory effect of Ca on U(VI) reduction may be diminished when competing sinks for Ca are present; ultimately, the dissolved concentration of Ca in equilibrium with the uranyl species is the controlling factor. Finally, all of the Fe(III) (hydr)oxides, and ferrihydrite in particular, have an intriguing impact on U(VI) reduction in the presence of Ca. While ferrihydrite acts as a competing electron acceptor for microbial reduction of U(VI) in the absence of Ca, as Ca concentrations increase, U(VI) reduction is promoted (rather than retarded) in systems with ferrihydrite as compared with those without. In fact, U(VI) reduction in the presence of ferrihydrite becomes nearly independent of the Ca concentration (Fig. 4). The mechanism by which ferrihydrite, and goethite and hematite to a lesser degree, maintain U(VI) reduction within increased Ca concentration is unclear but the impact is startling.

Implications for Uranium(VI) Reduction and Subsequent Immobilization
Biological reduction of U(VI) typically results in the sparingly soluble biogenic UO2 phase, retarding its migration within surface and subsurface environments. Prevailing geochemical conditions, however, can drastically modify the rate of microbial U(VI) reduction. Calcium categorically decreases the rate of reduction, with higher Ca concentrations promoting the fraction of Ca2UO2(CO3)3 and diminished reduction rate. Iron (hydr)oxides, in contrast, have nonlinear effects on U(VI) reduction. Goethite and hematite act as sorbents of Ca, and as a result decrease the proportion of the less reducible Ca2UO2(CO3)3 species. Ferrihydrite serves as a competing electron acceptor at low Ca concentrations, but promotes U(VI) reduction under conditions where the Ca2UO2(CO3)3 species predominates. In sum, dissolved Ca and associated complexes of U(VI) will decrease the rate of reduction, but offsetting geochemical factors, such as those imparted by Fe(III) (hydr)oxides, will help to ensure U(VI) reduction at a viable rate.


   ACKNOWLEDGMENTS
 
We would like to thank Matthew Ginder-Vogel for his constructive input on the manuscript and the valuable input from three anonymous reviewers and the associate editor. This work was funded by the Environmental Remediation Science Program, Office of Biological and Environmental Science, U.S. Dep. of Energy (grant number ER63609-1021814).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Stewart, B. D.
Right arrow Articles by Fendorf, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stewart, B. D.
Right arrow Articles by Fendorf, S.
Agricola
Right arrow Articles by Stewart, B. D.
Right arrow Articles by Fendorf, S.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Other Environmental Contamination

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome