VZJ sign up for etocs
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 24 January 2007
Published in Vadose Zone J 6:149-157 (2007)
DOI: 10.2136/vzj2006.0114
© 2007 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 Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhong, L.
Right arrow Articles by Deng, B.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Zhong, L.
Right arrow Articles by Deng, B.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Zhong, L.
Right arrow Articles by Deng, B.
Related Collections
Right arrow Redox Processes
Right arrow Vadose Zone Processes and Chemical Transport
Right arrow Ground Water Quality

ORIGINAL RESEARCH

Uranium Immobilization by Hydrogen Sulfide Gaseous Treatment under Vadose Zone Conditions

Lirong Zhonga,*, Edward C. Thorntona and Baolin Dengb

a Pacific Northwest National Lab., Richland, WA 99354
b Univ. of Missouri, Columbia, MO 65211
* Corresponding author (lirong.zhong{at}pnl.gov)

Received 15 August 2006.



   ABSTRACT
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Mobility of hexavalent uranium [U(VI)] in H2S-treated soils was investigated using laboratory column experiments to assess the potential of applying in situ gaseous reduction for U immobilization in the vadose zone. Soil from the Hanford Formation in the U.S. Department of Energy Hanford Site, Washington, was used in this study. The impact of water chemistry and soil treatment on U(VI) immobilization and the role of gas humidity on soil treatment were investigated. The study revealed that soil uptake of U(VI) from deionized water was much higher than that from the simulated Hanford groundwater. Nevertheless, gas-treated soil was still shown to have the potential for immobilizing U(VI) from the simulated groundwater. In addition, changes in H2S column breakthrough indicated that humidity enhanced the reduction of soil Fe. In the first 20 pore volumes, the soil treated with moisturized H2S gas can effectively immobilize >80% of the mobile U(VI). Primary mechanisms for U immobilization included U(VI) sorption to the sediments, reduction of U(VI) to insoluble U(IV), and enhanced adsorption of U(VI) to newly formed Fe oxides. Remobilization of U following reoxidation of the sediment was relatively insignificant under the experimental conditions applied, apparently owing to the enhanced adsorption of U to poorly crystallized hydrous ferric oxide products.

Abbreviations: DIW, deionized water • HFO, hydrous ferric oxide • ISGR, in situ gaseous reduction • KPA, kinetic phosphorescence analyzer • PRB, permeable reactive barrier • PV, pore volume • SGW, simulated groundwater


   INTRODUCTION
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
URANIUM CONTAMINATION is usually left behind by facilities operated for nuclear fuel and weapons production, and by nuclear power plants. Uranium exists in the environment primarily in the hexavalent U(VI) and tetravalent U(IV) oxidation states. Uranium(IV) generally forms insoluble mineral phases, such as uraninite [UO2(s)]. Uranium(VI) often exists in species with higher solubility such as Na-boltwoodite [(Na, K)(UO2)(SiO3OH)(H2O)1.5], uranophane [Ca(UO2)2(SiO3OH)2(H2O)5], soddyite [(UO2)2SiO4(H2O)2], schoepite [(UO2)8O2(OH)12(H2O)12], and rutherfordine [UO2CO3] (Finch and Murakami, 1999; Liu et al., 2004) and the uranyl ion, UO22+, and its aqueous and carbonate complexes. Although U contamination in the vadose zone normally does not impose a direct health risk to human beings as does contamination in groundwater, mobility of U(VI) in the subsurface environments is a major concern. Complexation by carbonate reduces adsorption and increases the solubility of possible precipitates. Therefore, U present in the vadose zone can serve as a long-term source of groundwater contamination, as the infiltration of rainwater and the fluctuation of the groundwater table transport U to underlying aquifers (Zachara et al., 2005). The U.S. Department of Energy (DOE) Hanford Site in Washington, for example, has a thick vadose zone interval significantly contaminated by U.

A limited number of methods have been developed for the remediation of U in the vadose zone. Soils can be washed using chemicals, such as citric acid and H2O2, to extract and remove U (e.g., Francis and Dodge, 1998; Choy et al., 2006). For contamination at relatively low concentrations and in soils of shallow depth, phytoremediation can be used to accumulate U in plants to achieve removal from the soil (e.g., Huang et al., 1998; Sharmasarkar and George, 2002).

In situ gaseous reduction (ISGR) treatment of vadose zone sediments with diluted H2S provides a possible means for immobilization of a variety of contaminants in the subsurface environment (Thornton and Amonette, 1999; Thornton, 2000). This technology uses diluted H2S gas as a reductant for immobilization of contaminants that show substantially lower mobility in their reduced oxidation states, e.g., Tc, U, and Cr. The rate of U(VI) reduction by H2S depends strongly on solution pH and carbonate concentrations (Hua et al., 2006). It is conceivable that the ISGR approach can be used in two ways: (i) to immobilize or stabilize preexisting contaminants in the vadose zone by direct H2S treatment; or (ii) to create a permeable reactive barrier (PRB) in which a gaseous mixture of H2S diluted in N2 is passed through an interval in the vadose zone to produce a volume of reduced sediment. The reduced phases (which contain ferrous oxyhydroxides and ferrous sulfide) would form a PRB that could immobilize possible future releases of contaminants from surface facilities or waste sites, such as during the processing of waste tank decommissioning at the Hanford Site.

The general reaction for H2S with Fe(III) hydroxide and Fe(III) oxides, in the absence of O2, can be expressed as follows (Cantrell et al., 2003; Davydov et al., 1998):

Formula 1[1]
For U immobilization, the Fe(II) generated in the treatment zone can act as the reductive reagent to reduce U(VI) to U(IV) when U-contaminated water infiltrates through the zone. Reductive immobilization of U(VI) by Fe(II) has been observed by several groups of researchers (e.g., Charlet et al., 1998; Liger et al., 1999; Livens et al., 2004; Sani et al., 2004; Behrends and van Cappellen, 2005). The efficiency and the lifetime of the PRB are dependent on the reductive capacity of the barrier, which is determined by the amount and the phase(s) of Fe (hydr)oxides and the H2S treatment duration.

In this study, batch tests, soil H2S treatment and U(VI) immobilization column experiments, and modeling studies were conducted to investigate whether H2S-treated sediment can effectively immobilize U in aqueous solutions under vadose zone conditions. The experimental parameters and conditions for soil treatment were similar to the ones used by Cantrell et al. (2003), and likewise, Eq. [1] was used to characterize the reactions between H2S and soil in this work. The influence of groundwater geochemistry on U immobilization and the influence of humidification of the H2S gas mixture during sediment treatment on soil reduction and its impact on U immobilization were evaluated.


   MATERIALS AND METHODS
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Materials
Uncontaminated Hanford Formation sediment was obtained from 200 East Area of the DOE Hanford Site, Washington, for use in this study. The sediment was collected from the ground surface during a hot and dry summer. The moisture content of the soil was determined to be 0.40%. The sediment was sieved to <2 mm before packing into the treatment columns.

The Hanford soil is a medium- to coarse-grained sand. X-ray diffraction analysis of the Hanford Formation fraction showed 30 to 50% quartz, 5 to 20% plagioclase, and <10% K-feldspar (Serne et al., 2002). Serne et al. (2002) also reported that the total Fe content of four Hanford Formation samples ranged from 3.46 to 9.64% (w/w) as Fe2O3. Iron was assumed to be in the Fe(III) forms including Fe(III) oxides, although some Fe(II) may be present in grain interiors.

One percent (1% mole/mole or 0.45 mmol/L) H2S (in N2) was mixed with pure N2 to prepare a 8.93 x 10–3 mmol/L (200 ppm) H2S gas mixture. Solutions containing U(VI) were prepared by dissolving uranyl nitrate hexahydrate dry chemical [UO2(NO3)2·6H2O] (Fluka, Buchs, Switzerland) in deionized water (DIW) and simulated groundwater. Simulated groundwater (SGW) that reflected the geochemistry of the Hanford Site groundwater was formulated with 3 x 10–4 M CaCO3, 2 x 10–4 M CaCl2, 2 x 10–4 M MgSO4, 4 x 10–4 M NaHCO3, and 6 x 10–5 M KHCO3. The pH of the U(VI) solutions in SGW and DIW were ~8.3 and ~7.7, respectively.

Soil Treatment Procedures
Soil was packed dry in glass columns with dimensions of 0.304 m long by 0.026-m inner diameter (Spectrum Chromatography, Houston, TX). The packed soil in each column was 300 ± 10 g. Column pore volume (PV) was calculated based on the column dimensions, measured sample weight, sediment moisture content, and an estimated particle density of 2700 kg/m3, and verified by the difference between the weight of the dry column and the weight of the column saturated with water. A 8.93 x 10–3 mmol/L (200 ppm) H2S in N2 mixture was directed downward through the column to treat the soil at a gas flow rate around 300 mL/min. Inflow and effluent H2S concentrations were measured to monitor the extent of soil treatment. Detailed soil treatment procedures are described elsewhere (Thornton et al., 2006).

To evaluate the effect of water moisture on soil reduction, treatment with moisturized gas was conducted in which the N2 gas was bubbled through a water column before mixing with H2S. The inflow gas mixture to the soil column was determined to be saturated with water vapor (i.e., 100% humidity) using an Omega RH82 moisture meter (Omega Engineering, Stamford, CT). The soil column effluent was simultaneously monitored for H2S concentration and water vapor content.

The H2S consumption by the soil during gas treatment tests was calculated based on the gas flow rate and the inflow and effluent concentration differences. The stoichiometry of the reaction between the soil and H2S, as indicated by Eq. [1], was used to estimate the amount of Fe(III) reduced during treatment. For this calculation, it was assumed that dissolved H2S is negligible since the sediment was nearly dry. Previous testing activities also indicated that adsorption of H2S by soil is very low (Thornton and Amonette, 1999).

Uranium Sorption Batch Studies
Sorption tests in both DIW and SGW were conducted to evaluate the role of water chemistry on U(VI) sorption to Hanford Site soil. One gram of pristine Hanford Site soil was placed into a 40-mL tube containing 30 mL of DIW or SGW with a range of 4.2 x 10–8 M (10 ppb) to 1.26 x 10–6 M (300 ppb) U(VI). The water–soil suspension was mixed on a rotator for 7 d. The soil–water mixture was then centrifuged to separate the phases. Uranium in the supernatant was analyzed using a kinetic phosphorescence analyzer (KPA) (Chemcheck Instrument, Richland, WA). A mass balance calculation was used to calculate the amount of U(VI) sorbed on the soil.

Sorption isotherms were fit with the Freundlich model. A sorption isotherm can be expressed using the following equation:

Formula 2[2]
where Cs (µmol/kg) is the sorbate concentration in soil; Cw (µmol/L) is the sorbate concentration in solution; Kf (µmol(1–n)Ln/kg) is the solid/water distribution ratio, and n is a measure of nonlinearity of the isotherm.

Uranium Immobilization Column Tests
Five column tests were conducted to evaluate the influence of soil treatment, H2S gas humidification, and groundwater geochemistry on U immobilization. Details of the treatment parameters for the five columns are summarized in Table 1. For U immobilization tests conducted with untreated soil, a glass column with i.d. = 0.026 m and length = 0.304 m was packed with pristine Hanford Site soil, and a 1.26 x 10–6 M (300 ppb) U(VI) solution was then flushed through the column. When H2S-treated soil was needed in an immobilization test, the soil column was first treated with the H2S–N2 gas mixture as described above. The treated soil pack was then directly flushed with a 1.26 x 10–6 M (300 ppb) U(VI) solution, without disassembly and repacking of the column. In all the immobilization tests, solution flow was controlled in an upward mode using a syringe pump (Kloehn Co. Ltd., Summerlin, NV). The columns were flushed with U solution in DIW or SGW at a flow rate of 0.4 mL/min. The pore velocity was between 3.22 and 3.62 m/d. The inflow solutions were open to the atmosphere. The O2 level in the column effluent was monitored and recorded using a LabView data acquisition system (National Instruments, Austin, TX) with an O2 probe installed in the effluent line. Calibration for the O2 probe was conducted daily. The column effluent was directed to a fraction collector for sampling. The collection time for each sample was 25 min. Samples were analyzed for U(VI) concentration using a KPA.


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

  		Table 1. Parameters for immobilization column tests.

 
Extraction of Immobilized Uranium
To study the mechanism of U immobilization, the sediment in a treated column with immobilized U was extracted for U(VI) and total U. The column was disassembled in an anaerobic chamber. Before the soil was removed from the column, the column was evenly divided and marked into seven sections. Each section was 4.3 cm long. The soil in these sections was then removed from the column and separated. Samples for extraction were taken from sediments in Sections 1 (influent end), 3, 5, and 7 (effluent end). One molar NaHCO3 solution prepared in the anaerobic chamber was used to extract U(VI) (Duff et al., 1998; Kohler et al., 2004), and 0.5 M HNO3 was applied to extract total U in the sediments. The extraction was conducted in the chamber. The mixture of HNO3 and sediment was oxidized by bubbling air through for 12 h before the solid and aqueous phase were separated by centrifugation. This step was used to convert U(IV) to U(VI) so that the total U could be determined by a KPA.

Modeling
Geochemical modeling of aqueous speciation and adsorption reactions was undertaken using Geochemists Workbench (Bethke, 2004, p. 29–39, 47–61) to provide an explanation for the results of the batch tests reported in this study. The development of the aqueous speciation model for U involved consideration of the appropriate complexation reactions and the stability constants associated with these equilibria (Table 2). The source for the uranyl hydroxide and carbonate complexation constants are those presented in the Nuclear Energy Agency's database for U (Grenthe et al., 1992, p. 107, 313; see also Waite et al., 1994; Langmuir, 1997, p. 376–381, 551–552; Davis et al., 2004; Guillaumont et al., 2003, p. 157–292). More recently, it has been recognized that calcium uranyl carbonate complexes are also important in groundwater systems (Brooks et al., 2003; Bernhard et al., 2001). Thus the equilibria and constants involving the Ca2UO2(CO3)30 and CaUO2(CO3)32– complexes as presented in Bernhard et al. (2001) were incorporated into the modeling activities. Adsorption modeling was performed using the diffuse-layer or two-layer surface complexation model (Dzombak and Morel, 1990, p. 1–41), where it was assumed that hydrous ferric oxide (HFO) is the predominant sorbate. Goethite ({alpha}-FeOOH) was assumed to be the HFO sorbing surface and was present at a concentration of 10–3 mol/L with a sorbent surface area of 600 m2/g. Surface complexation constants reported by Dzombak and Morel (1990, p. 1–41, 305) were used in the model.


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

  		Table 2. Stability constants (logK at an ionic strength I = 0) for selected reactions involving U(VI) at 25°C.

 

   RESULTS AND DISCUSSION
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
Soil Reduction
Normalized column H2S effluent concentrations from the dry gas (Col-III) and the 100% humidity (Col-IV) treatments are plotted as a function of PV in Fig. 1 . Also shown are data from Thornton et al. (2006) for two dry gas treatments (MST-2 and MST-5) at different flow rates (200 and 500 mL/min). For the dry gas treatments, slower gas flow rates resulted in later breakthrough of H2S, indicating kinetic constraints for the reaction of dry H2S gas within the column at the flow rates used. For Col-IV treated with humidified H2S gas, H2S breakthrough was retarded the most, even though the flow rate was faster (245 mL/min) than the slowest dry gas flow rate (200 mL/min), indicating that moisture enhanced the reaction. The amount of H2S consumed increased from 38.4 to 60.1 mmol between the dry gas (Col-III) and the humidified gas (Col-IV) treatments. Likewise, using Eq. [1], the estimated amount of Fe(III) reduced increased from 2.35 mmol for the dry gas to 3.78 mmol for the humidified treatment.


Figure 1
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 1. Soil treatment column effluent normalized H2S concentration. Q is the flow rate of the gas mixture through the column.

 
The inflow gas mixture had a steady saturated relative humidity (100%); however, the effluent gas had the same humidity as the ambient air (33.0%) for the first 560 PVs as the column soil absorbed water from the gas mixture (Fig. 2 ). The effluent moisture content started to increase then, and reached steady saturation state (99.9%) at about 13 000 PVs. The total amount of water taken up by the soil was calculated to be 1.76 g based on mass balance of the inflow and outflow. At the end of the soil treatment, the column had a weight gain of 1.59 g, close to the calculated water capture.


Figure 2
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 2. Effluent H2S concentration and moisture from the column of soil treated with the moisturized gas mixture.

 
Likewise, water vapor enhanced Cr(VI) reduction by gaseous H2S (Hua and Deng, 2003), which was attributed to dissolution of Cr(VI) into the water film that formed on particle surfaces, followed by H2S transfer to the film, resulting in Cr(VI) reduction. This proposed mechanism is believed to enhance the soil reduction processes associated with this study. The water solubility of Fe(III) oxides is very low in neutral pH. When pH decreases one unit, however, the activity of Fe3+ increases 1000-fold (Lindsay, 1988). When humidified gas flowed through the soil column, a water film was formed on the soil particles and H2S would dissolve into the film. The pH of the water could be very low, resulting in enhanced solubilization of Fe(III) oxides in the film and, thereby enhance the reduction of soil Fe oxides by sulfide.

Dos Santos Afonso and Stumm (1992) also studied the dissolution of Fe(III) (hydr)oxides by H2S using aqueous-phase batch tests. A gas mixture of H2S in N2 was bubbled through a hematite suspension in their tests and the Fe(II) concentration was monitored. Formation of {equiv}FeS– and {equiv}FeSH complexes on the Fe (hydr)oxide surfaces and subsequent electron transfer was postulated as the reductive dissolution mechanism. A similar mechanism might also be behind the enhanced soil reduction reported here. Enhanced H2S consumption and soil reduction is consistent with the observation that the efficiency of Fe oxide scrubbers used for removal of H2S from waste gas mixtures is enhanced when the waste gas is humidified before treatment (e.g., Fe-based solid scavenger systems portrayed in Table 1 in Heguy and Bogner, 2005).

Uranium Immobilization
For the untreated Hanford soil, an estimated retardation factor (R) based on the U breakthrough curves in column test Col-II with SGW was 9 (Fig. 3A ), whereas no U breakthrough was observed in test Col-I with DIW during the monitoring period (up to 180 PVs; plot not shown). In Col-II, outflow U concentrations achieved inflow concentrations by ~30 PVs. These results indicate that water chemistry has a significant influence on U(VI) immobilization by adsorption (see discussion on geochemical modeling below).


Figure 3
View larger version (23K):
[in this window]
[in a new window]

  		Fig. 3. (A) Effluent U(VI) and fraction of U immobilized, and (B) effluent O2 concentration in column tests. Test Col-II: untreated Hanford Site soil; Col-III: soil treated with dry H2S–N2 gas; Col-IV: soil treated with the moisturized gas mixture. Uranium(VI) in simulated groundwater (SGW) at similar concentration was used as the influent in all tests; DIW is deionized water.

 
The U(VI) breakthrough was nearly immediate with the untreated soil (Test Col-II), while in the dry and moisturized H2S gas treated columns, U(VI) was substantially retarded with the greatest retardation observed with the moisturized gas stream (Fig. 3A). The difference in the effluent U(VI) concentrations between the test with untreated soil (Col-II) and treated soil (Col-III or Col-IV) at the same PV is attributed to U immobilization by interaction with the reducing products in the soil. By comparing the U(VI) effluent profiles, the performance of the H2S gas soil treatment for U immobilization can be evaluated as shown in Fig. 3A for the first 30 PVs. The soil treated with the moisturized gas mixture (Col-IV) had a higher immobilization capacity than the soil treated with the dry gas mixture (Col-II). This result is in agreement with the observation that, at the same gas flow rate and PV, the soil was more readily reduced when the moisturized gas mixture was applied during treatment (Fig. 1).

In Test Col-III (dry H2S gas treatment), no O2 was detected in the first 14 PV (Fig. 3B) even though U(VI) had started to exit from the column (Fig. 3A). The same phenomenon was seen in Col-IV (moisturized H2S gas treatment) for the first 30 PVs. It can be deduced from this observation that when O2 and U(VI) compete for the reducing regent in the soil, O2 is reduced before U(VI). We also speculate that the reduction of U(VI) in the column is a kinetics-controlled process. The residence time of the U(VI) solution in the column was about 2.1 h. This contact time apparently was not long enough for U(VI) to be reduced, even though there was no O2 to compete for the reducing regent. If the residence time was longer, the breakthrough of U(VI) might be at later pore volumes. Future work should be undertaken to study the kinetics of U(VI) reduction by H2S-treated soil.

In the batch sorption isotherms (Fig. 4 ) much more U(VI) was sorbed from DIW relative to SGW, with Kf values of 25.57 and 6.28, respectively. Furthermore, a linear isotherm was observed for U(VI) sorption from DIW (n = 1 in Eq. [2]), whereas sorption from SGW was nonlinear (n = 0.76). Linear isotherms infer similar and infinite sorption sites in the concentration range assessed compared with nonlinear behavior, which infers a limitation of sites and sites of varying affinity. The equilibrium pH values were similar to the applied solutions for both DIW and SGW. The decreased sorption from SGW is consistent with the early breakthrough with SGW in the column studies.


Figure 4
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 4. Sorption isotherms for U(VI) in deionized water (DIW) and in simulated grounwater (SGW) sorbed to pristine Hanford soil. The symbols are measured data and the lines are the best fit using the Freundlich model.

 
To quantitatively compare the batch and column tests, retardation factors (R) were calculated from the isotherm fits using the following equation (Nkedi-Kizza et al., 1987):

Formula 3[3]
where {rho} is the soil bulk density (g/cm3), ne is the porosity, and Cw is the influent U(VI) concentration (µmol/L). The calculated R values for Col-I and Col-II were 149.6 and 32.0, respectively. The batch test results verified the column test observation, i.e., immobilization of U(VI) by adsorption from DIW is much stronger than that from SGW. The discrepancy between the R values estimated from column tests and those calculated based on batch tests might be due to the difference between batch and column test conditions, such as water/soil ratio and pH. For an example, the pH of the U in the DIW solution increased from 7.7 to about 8.5 in the effluent. This change might affect the U(VI) sorption behavior.

Aqueous Speciation and Adsorption Modeling of Batch Test Results
The measured sorption isotherms show that U(VI) adsorbed much more strongly from DIW than from SWG. The difference in U mobility observed between the batch tests conducted with DIW and SGW in contact with untreated soil in this study is significant and has implications for U transport in oxidized environments. Geochemical modeling of U(VI) aqueous speciation and adsorption was undertaken to gain insight into this observed behavior.

Aqueous speciation and saturation runs were performed initially for the DIW and SGW solution chemistries with 1.26 x 10–6 mol/L (300 ppb) total U. The formation constants of the primary uranyl aqueous species used in the speciation are identified in Table 2. The stability domains of various U(VI) species as a function of pH are presented in Fig. 5 for the batch test solution chemistries involved. The primary uranyl complex in DIW is UO2(CO3)22– under the experimental conditions, where it is assumed that the solution was in equilibrium with the atmosphere (pCO2 = 10–4.5 MPa and pO2 = 0.02 MPa) at a measured pH value of 7.7 and an activity of UO22+ of about 10–11 (as determined by the speciation calculations). The predominant uranyl aqueous complex in the SGW is interpreted to be Ca2UO2(CO3)30, based on the modeling calculations presented in Fig. 5B at a pH of 8.3, an activity of UO22+ of 10–14, pO2 = 0.02, aCa2+ = 10–3.3, aHCO3– = 10–3.1, and for the activities of the other major constituents involved. Thus, the activity of the UO22+ species was much less in the batch test conducted with SGW, primarily because of the formation of the Ca2UO2(CO3)3 aqueous complex. Curtis et al. (2004) and Davis et al. (2004) have previously reported that this species is the most important U(VI) solution species in modeling activities conducted with groundwater solutions in near equilibrium with calcite.


Figure 5
View larger version (23K):
[in this window]
[in a new window]

  		Fig. 5. Activity diagram illustrating stability fields of U(VI) aqueous and solid species (shaded field) for pCO2 = 10–3.5 and under oxidizing conditions in (A) deionized water and (B) simulated groundwater. Symbol indicates the location of the solution used in this work.

 
To better understand the extent and nature of adsorption of U(VI) in the batch tests conducted in this study, sorption modeling was also undertaken to simulate interaction of the sediment and the DIW and SGW solution with U(VI) added. Modeling was performed using the diffuse-layer or two-layer surface complexation model (Dzombak and Morel, 1990, p. 1–41), where it was assumed that HFO is the predominant sorbate for U(VI). Goethite ({alpha}-FeOOH) was assumed to be the HFO sorbing surface and was assumed to be present at a concentration of 10–3 mol/L of solution and at a sorbent surface area of 600 m2/g. Total aqueous U was input at a concentration of 1.26 x 10–6 mol/L (300 ppb) and conditions were specified as oxidized at a pO2 of 0.02 MPa. The sorption database FeOH.dat associated with Geochemist Workbench was used for this exercise; however, the following additional sorption reactions were added to model the sorption of U(VI):

Formula 4[4]

Formula 5[5]
where (s) refers to a high-energy or strong site and (w) refers to a low-energy or weak site (Langmuir, 1997, p. 376–381; Dzombak and Morel, 1990, p. 1–41).

The results of adsorption modeling for the DIW solution with HFO indicated that nearly all of the U (>99%) was adsorbed, consistent with the experimental observations. Only about 12% of the U present in the SGW solution was predicted to be adsorbed. The primary reason for this behavior is that the activity of UO22+ was much lower in the SGW solution owing to the formation of the Ca2UO2(CO3)3 complex. In addition, about 9% of the HFO sorption sites were predicted to be occupied by Ca in the test involving the SGW solution. The results of the modeling exercise thus suggest that the higher mobility of U observed in the tests conducted with SGW is due to complexation of U(VI) with Ca in solution and, to a lesser extent, competition for adsorption sites.

Immobilization Mechanisms
The U and O2 effluent profiles for Test Col-V (data not shown) were similar to those in Col-IV. The difference in the degree of reoxidation between the inflow end and the effluent end of the column was revealed by the colors of the sediments. At the effluent end, a light greenish color indicated the presence of Fe(II), while at the inflow end, the color changed from greenish back to the pristine soil color. The color-changing front proceeded from the inflow end to the effluent end as more water was flushed through the column.

Test Col-V was terminated after partial reoxidation occurred and the sediments were extracted for U(VI) and total U. The U(IV) concentration was calculated from the difference between the total U and U(VI) concentrations. The concentrations of U(VI) and U(IV) in sediments at different sections of the column are plotted in Fig. 6 . Based on the extracted U concentrations, the total U in the sediments was calculated to be 0.53 mg. The immobilized U in this column was determined to be 0.54 mg based on the U mass balance between the inflow and effluent fluids. The close match between these two approaches indicated that the extraction with 0.5 M HNO3 can remove the total U immobilized in the soil by the means applied in this study. The extraction results indicated that more U was immobilized at the influent end than at the effluent end, and the change in the total U concentration was gradual across the column. It is clear that both U(VI) and reduced U were present as the immobilized form.


Figure 6
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 6. Immobilized U concentration in sediments across column. The distance from the column influent end was measured from the center of each section. Uranium(IV) was determined based on the measured total U and U(VI) concentration.

 
There are several possible mechanisms behind U(VI) immobilization. The first one is the adsorption of U(VI) to the gas-treated soil. The second mechanism is reduction of U. Uranium(VI) is reduced to U(IV) and precipitates from the aqueous solution. Uranium(IV) was present in the soil throughout the column. The third mechanism is the enhanced sorption of U(VI) to reoxidized soil. The Fe(II) generated from soil reduction by H2S is reoxidized when O2 is introduced to the column. The initial oxidation product of Fe(III) is likely to be poorly crystallized HFO with a high surface area (Kukkadapu et al., 2006). The HFO or Fe oxides are reported to strongly adsorb U(VI), resulting in its immobilization (Zhong et al., 2005). In the soil pack treated with moisturized gas, more Fe3+ was reduced, therefore reoxidation generated more amorphous HFO, which provided enhanced U immobilization by sorption (Fig. 3). Because of enhanced sorption, there was more sorbed U(VI) in the column influent end and sorbed U(VI) gradually decreased toward the effluent end in Test Col-V (Fig. 6). Effluent O2 was not detected at the termination of the experiment, O2 availability was therefore limited. In the inflow, Fe(II) could first be oxidized by O2 dissolved in the flushing water, while O2 was depleted at the effluent end. More newly oxidized Fe oxides were formed toward the inflow end of the column and more U(VI) was sorbed.

It is also worth noting that remobilization of U from the sediment does not appear to be significant following O2 breakthrough (Fig. 3), suggesting that U release does not occur on reoxidation of the sediment under the tested experimental conditions. The reoxidized U is speculated to be sorbed to the newly formed Fe oxides, and remobilization is therefore prevented. Further work should be undertaken to assess the extent to which U may be remobilized on reoxidation of the H2S-gas-treated sediment.

Implications
Within the first 20 PVs of pumping U(VI) solutions through the moisturized H2S-gas-treated sediment, it was observed that >80% of the mobile U(VI) was immobilized. Both (enhanced) sorption and reduction contributed to U immobilization. During the U immobilization processes, breakthrough of the U(VI) occurred well before breakthrough of O2. That is, even though the sediment column was still under a reduced environment, consuming all the O2 traveling through, U(VI) was able to move through a reduced environment. This observation implies that the reduction of U(VI) may be a kinetically controlled process. It may also imply that the reducing potential at this stage was not significant enough for the reduction of U(VI).

This study on the immobilization of U(VI) by soil treated with gaseous reduction indicates that ISGR-treated Hanford soil is capable of effectively immobilizing U(VI) from SGW. The immobilization is further enhanced by soil treatment undertaken with a moisturized H2S gas mixture. It also appears that release of U on reoxidation of the sediment is low owing to the enhanced adsorption of U to the poorly crystallized HFO products. These results thus indicate that gaseous treatment of the vadose zone could potentially be undertaken to produce a horizontal reduced interval beneath a contaminated site. A permeable reactive barrier emplaced by this approach would immobilize U(VI) by reduction and adsorption if a waste solution from the site infiltrates vertically downward into the treated interval. Further evaluation of the potential lifetime of an ISGR vadose zone PRB is currently being undertaken through additional column testing and flow and reactive transport modeling activities.


   ACKNOWLEDGMENTS
 
Dr. Jim Szecsody at Pacific Northwest National Laboratory (PNNL) helped to set up the data acquisition system. The funding for this study was provided by the Environmental Management Science Program of the U.S. Department of Energy (DOE) (Grant no. DE-FG02-03ER63616). PNNL is operated by Battelle for U.S. DOE.


   REFERENCES
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Vadose Zone JHome page
E. C. Thornton, L. Zhong, M. Oostrom, and B. Deng
Experimental and Theoretical Assessment of the Lifetime of a Gaseous-Reduced Vadose Zone Permeable Reactive Barrier
Vadose Zone J., November 20, 2007; 6(4): 1050 - 1056.
[Abstract] [Full Text] [PDF]

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 Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhong, L.
Right arrow Articles by Deng, B.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Zhong, L.
Right arrow Articles by Deng, B.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Zhong, L.
Right arrow Articles by Deng, B.
Related Collections
Right arrow Redox Processes
Right arrow Vadose Zone Processes and Chemical Transport
Right arrow Ground Water Quality

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