Uranium Release from Sediment to Groundwater: Influence of Water Chemistry and Insights into Release Mechanisms Samrat Alam, Tao Cheng PII: DOI: Reference:

S0169-7722(14)00082-5 doi: 10.1016/j.jconhyd.2014.06.001 CONHYD 3008

To appear in:

Journal of Contaminant Hydrology

Received date: Revised date: Accepted date:

16 August 2013 28 May 2014 2 June 2014

Please cite this article as: Alam, Samrat, Cheng, Tao, Uranium Release from Sediment to Groundwater: Influence of Water Chemistry and Insights into Release Mechanisms, Journal of Contaminant Hydrology (2014), doi: 10.1016/j.jconhyd.2014.06.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

SC

RI

PT

ACCEPTED MANUSCRIPT

NU

Uranium Release from Sediment to Groundwater: Influence of Water Chemistry and

TE

D

MA

Insights into Release Mechanisms

AC CE P

Md. Samrat Alam and Tao Cheng*

Department of Earth Sciences, Memorial University St. John’s, Newfoundland & Labrador, A1B 3X5 Canada

* Corresponding Author. Phone: (709) 864-8924; Fax: (709) 864-7437, Email: [email protected]

1

ACCEPTED MANUSCRIPT Abstract

PT

Uranium (U) contamination in groundwater often results from natural geochemical processes such as mineral dissolution and desorption of adsorbed U from mineral surface.

RI

Although U adsorption and U mineral dissolution have been extensively studied, current

SC

knowledge of minerals and water chemistry conditions that control U release in uncontaminated soil and aquifers is still limited. Identification of these minerals and the knowledge of how water

NU

chemistry conditions influence U release is critical to better understand, predict, and manage

MA

geogenic U contamination in soil and groundwater. The objective of this study is to determine the extent and mechanisms of U release from a heterogeneous natural sediment under water

D

chemistry conditions relevant to natural soil water and groundwater. A sediment sample was

TE

collected and characterized by XRD, SEM-EDX and extraction methods, and examined using

AC CE P

laboratory leaching experiments. Our results show that Fe-Mn (oxy)hydroxides and silicate minerals are the major U hosting minerals, and a substantial fraction of U exists as adsorbed ions on minerals. We also found that U release is controlled by a number of interactive processes including dissolution of U-bearing minerals, U desorption from mineral surface, formation of aqueous U complexes, and reductive precipitation of U. Results from this study shed light on the important geochemical reactions that need be considered for developing a conceptual model that predicts U contamination in subsurface environment. Keywords: Uranium release, Sediment, Mineral dissolution, Desorption, Groundwater contamination

2

ACCEPTED MANUSCRIPT 1. Introduction Uranium (U) is a contaminant commonly found in groundwater that could pose a serious

PT

threat to human health. Dissolved U at very low concentrations is found in most natural waters

RI

(Mkandawire, 2013), and typical groundwater concentration of dissolved uranium is on the order of a few µg U/L (Herring, 2013; Wiedemeier et al., 1995). Although groundwater U

SC

contamination in some cases is caused by anthropogenic pollution such as uranium mining,

NU

processing of uranium ores, and production and disposal of radioactive materials, U in groundwater is more often introduced by natural geochemical processes: i.e., U is released to

MA

groundwater from its hosting rocks and sediments via mineral dissolution and/or U desorption from mineral surface (Chen et al., 2005; James and Sinha, 2006). Uranium is ubiquitous in the

D

crust with an average concentration of 2.76 mg/kg (Herring, 2013). Common naturally-occurring

TE

U minerals include: oxides (uraninite and pitchblende), silicates (coffinite, soddyite, uranophane,

AC CE P

and uranothorite), phosphates (autunite), and vanadates (carnotite). Besides discrete U minerals, a significant fraction of solid-phase U exists in the form of uranyl ion (UO2)2+ adsorbed to mineral surfaces under oxidizing conditions (Welch and Lico, 1998; Wiedemeier et al., 1995). In the absence of anthropogenic pollution, U concentration in soil water and groundwater is controlled by U release from minerals to water. The extent of U release is influenced by U hosting minerals, oxidation state of U, and water chemistry (Fanghanel and Neck, 2002). In addition to thermodynamic controls, U release is strongly influenced by kinetics of mineral dissolution and U desorption (Fox et al., 2012; Liu et al., 2009; Liu et al., 2004; Qafoku et al., 2005). Under reducing conditions, the oxidation state of U is +4, and the stable U (IV) phases are mainly uraninite and coffinite (Duff et al., 1999). Organic complexes of U (IV) associated with humic materials may also retain U (IV) in solid phases (Bednar et al., 2007). The solubility of U

3

ACCEPTED MANUSCRIPT (IV) minerals is extremely low, and reducing conditions effectively diminishes the movement of uranium in soils and groundwater (Duff et al., 1999; Wiedemeier et al., 1995). Reductive

PT

precipitation of U(VI) is an effective method to immobilize U (Abdelouas et al., 1998; Finneran

RI

et al., 2002; Fredrickson et al., 2000; Lovley and Phillips, 1992), while oxidative dissolution of U(IV) minerals is a major mechanism of U mobilization (Finch and Murakami, 1999; Finch and

SC

Ewing, 1992; Wiedemeier et al., 1995). Under oxidizing conditions, the predominant oxidation

NU

state of U is +6, and U(VI) mainly exists in the form of uranyl ion. The adsorption/desorption of uranyl ion to/from mineral surface is a major process that controls U mobility under oxidizing

MA

conditions. Important U(VI) adsorbing minerals include Fe (oxy)hydroxides, clay minerals, and organic matters (Bowman, 1997; Catalano et al., 2006; Kelly et al., 2003; Kelly et al., 2006;

TE

D

Wang et al., 2005; Wiedemeier et al., 1995). Water chemistry parameters that control U adsorption include: pH, redox potential (Eh), dissolved carbonate, phosphate, and natural organic

AC CE P

matter (Bednar et al., 2007; Cheng et al., 2004; Echevarria et al., 2001; Sanding and Bruno, 1992). Water chemistry influences U adsorption/desorption by changing surface charge and solubility of minerals, U oxidation state and speciation, as well as the speciation of aqueous and surface complexes (Bachmaf et al., 2008; Casas et al., 1998; Echevarria et al., 2001; Katsoyiannis, 2007; Wazne et al., 2003). U(VI) adsorption to sediments have been extensively studied (e.g., Giammar and Hering, 2004; Sharp et al., 2011; Ulrich et al., 2008; Wang et al., 2013), and U hosting minerals at some contaminated sites have been identified (Arai et al., 2007; McKinley et al., 2007; Stubbs et al., 2009). A number of studies have investigated dissolution of pure U minerals (e.g., Giammar and Hering, 2004), oxidative dissolution of uraninite (Campbell et al., 2011; Ulrich et al., 2008; Wang et al., 2013), and U release from contaminated sediments at elevated bicarbonate

4

ACCEPTED MANUSCRIPT concentrations (Fox et al., 2012; Kohler et al., 2004; Liu et al., 2009). It has been observed that high U concentration in water is often associated with oxidizing, carbonate-rich, and phosphate-

PT

free conditions (Catalano et al., 2006; Kelly et al., 2003; Kelly et al., 2006). There have been

RI

fewer systematic studies on how water chemistry variables (e.g., pH and redox potential) and aqueous species commonly found in natural water (e.g., citrate, bicarbonate, and natural organic

SC

matter) influence U release from uncontaminated natural sediments to water. It is not clear to

NU

what extent water chemistry conditions and these natural aqueous species will influence U release, and what U release mechanisms (desorption or mineral dissolution) are involved and/or

MA

dominate with the presence of these natural compounds under a wide range of water chemistry conditions, which are often found in natural water. Systematic studies on U release under

TE

D

environmentally relevant water chemistry conditions will advance our understanding of geogenic U contamination in soil and aquifers and improve our ability to predict and manage geogenic U

the world.

AC CE P

contamination in groundwater, which is an important source of drinking water in many parts of

The objective of this study is to quantify the extent of U release and identify U release mechanisms under a range of water chemistry conditions relevant to natural soil water and groundwater. A sediment sample was collected and characterized by XRD, SEM-EDX, and extraction methods. The sediment sample was subsequently examined using laboratory leaching experiments to investigate the effects of pH, citrate, redox potential (Eh), bicarbonate, and natural organic matter on U and major element release. Our results showed that Fe-Mn (oxy)hydroxides and silicate minerals are the major U hosting minerals, and substantial amounts of U exist as adsorbed ions. Overall, both desorption and mineral dissolution were found important for U release under the studied conditions, and the relative importance of these two 5

ACCEPTED MANUSCRIPT mechanisms was pH dependent. In addition, we found that under reducing conditions, aqueous U concentration was substantially lower compared to that under oxidizing conditions, which was

PT

caused by reductive precipitation of the U released (via both desorption and mineral dissolution).

RI

Bicarbonate, citrate, and natural organic matter were found to promote U release, however, the

2. Materials and methods

NU

2.1 Sample collection and characterization

SC

extent and mechanisms for the enhanced U release were dependent upon the specific compound.

MA

A sediment sample was collected on 30th August, 2012 from a site (latitude: N 46.89577624 and longitude: W 55.39293679) near St. Lawrence (Fig. 1), a town located in Burin

D

Peninsula, south coast of the Island of Newfoundland, Canada, where uranium concentrations are

TE

high in sediments and some water wells according to the Department of Environment and

AC CE P

Conservation – Government of Newfoundland and Labrador (Newfoundland and Labrador Water Resource Portal and maps.gov.nl.ca/water/mapbrowser/Default.aspx). The sediment sample was collected from a depth of 0.6 to 1 m below land surface, air dried and sieved through 0.053 mm sieve, well mixed, and stored in plastic buckets for use in all subsequent experiments. The cutoff size of 0.053 mm is recommended by the Till Protocol Working Group Canada for geochemical analysis because ore minerals are easily broken down to this size range over short distances and it contains phyllosilicates that will scavenge cations released during weathering (Lett, 1995; Levson, 2001; Nevalainen, 1989; Shilts, 1993; Tarvainen, 1995). Mineral composition of the sediment sample was determined using X-ray diffraction (XRD) and scanning electronic microscopy equipped with energy dispersive spectroscopy (SEM-EDX). XRD analyses were performed using Rigaku Ultima IV diffractometer with Cu K

6

ACCEPTED MANUSCRIPT α radiation and operating at 40 kV and 44 mA. The samples were scanned from 2θ = 5° to 90° at a scan rate of 1 degree per second. The results obtained were processed using the MDI Jade

PT

computer program and databases from International Centre for Diffraction Database (ICDD) and

RI

the Inorganic Crystal Structure Database (ICSD). SEM-EDX analyses were performed using FEI MLA 650 F scanning electronic microscope equipped with a Bruker EDX system with dual

SC

Xflash 5030 SDD x-ray detectors (Sylvester, 2012). The energy limit for the EDX display is

NU

limited in software at 20 kV, 4096 channels, and 5eV/channel. The acceleration voltage used was 25 kV, typical for mineral liberation analysis (MLA), and the beam current (spot size) was

for mineral identification and counting.

MA

adjusted at 10 nano Amps (nA). Data obtained by SEM-EDX were processed using FEI software

TE

D

The pH, organic carbon, and metal concentrations of the sample were determined using wet chemistry methods. Sediment pH was determined by mixing 5 mL 5 mM CaCl2 with 2.5

AC CE P

grams of dry sediment sample, and measuring the pH of the supernatant (Williams et al., 2003). Organic carbon content of the sediment sample was determined following the procedure described by Gregorich and Ellert (1993), i.e., a solution of 5 mM CaCl2 was use to extract organic carbon from the sediment, and the extracted dissolved organic carbon (DOC) in the CaCl2 solution was measured by a Shimadzu TOC-V analyzer. A one step total digestion method was used to determine total U concentration in the sediment. In this total digestion method, 0.1 g of dry sediment sample was mixed with 2 ml 8 M HNO3 and 1 ml concentrated HF. The mixture was refluxed overnight on a hot plate at approximately 70 °C. As the sample dissolved completely, the solution was evaporated to dryness, then 2 ml 8 M HNO3 and 1 ml 0.453 M boric acid were added to the residue and evaporated to dryness. This step was repeated before another 2 ml 8 M HNO3 was added to the

7

ACCEPTED MANUSCRIPT residue and heated gently at approximately 70 °C until the entire residue dissolved. The sample was then transferred to a clean dry 120 ml container, and 1.3 ml 0.22 M oxalic acid and 0.665 ml

PT

0.453 M boric acid were added to the container and made up to a final weight of 60 g with

RI

nanopure water. The solution was then measured by ICP-MS to determine U concentration. Duplicate sediment samples were run to determine total U concentration.

SC

To determine total adsorbed U concentration in the sediment, sodium (bi)carbonate

NU

solution was used as an extractant (Kohler et al., 2004). Ten (10) mL solution containing 1.44×10-2 M NaHCO3 and 2.8×10-3 M Na2CO3 was mixed with 0.5 g dry sediment in a

MA

centrifuge tube and agitated gently on a shaker table. After 12 days of mixing, samples were filtered through 0.45 µm filters and analyzed by ICP-MS for U concentration. Duplicate

TE

D

sediment samples were run to dertermine total adsorbed U concentration. A five step sequential extraction procedure (Tessier et al., 1979) was performed on the

AC CE P

sediment sample to determine the elements (i.e., U, Fe, Mn, Al, Si, Mg, and Ca) contained in each of the following five phases: exchangeable, weak acid extractable, Fe-Mn oxides, organics and/or oxidizable minerals, and residual (mostly silicate minerals), with the extracted solutions measured by ICP-MS. Residue from each extraction step was rinsed with deionized water before used in the next extraction step. To extract the exchangeable elements (i.e., elements that are very weakly adsorbed to mineral surface), half (0.5) gram of dry sediment was mixed with 8 ml of 1M MgCl2 (pH = 7) solution and agitated in a 50 ml centrifuge tube for 1 hour. To obtain weak acid extractable elements, eight (8) ml of 1 M NaOAc (adjusted to pH = 5 with HOAc) was added to the residue from the first step and agitated for 5 hrs. To dissolve Fe-Mn oxides, fifteen (15) ml of 0.04 M NH2OH.HCl (in 25% (v/v) HOAc) was added to the residue from the 2nd step, and the mixture was heated to 96 ± 3 ○C for 6 hrs with occasional agitation. Organics and/or

8

ACCEPTED MANUSCRIPT oxidizable minerals were extracted by adding 3 ml of 0.02 M HNO3 and 5 ml 30% H2O2 (adjust pH to 2 with HNO3) to the residue from the 3rd step, and the mixture was heated to 85 ± 2 ○C for

PT

2 hrs with occasional agitation. Finally, residue from the 4th step was transferred to a Teflon jar,

RI

and 2 ml of 8 M HNO3 and 1 ml concentrated HF were added. The jar was capped and heated on a hot plate until the residue completely dissolved. It should be noted that the five phases defined

SC

by Tessier et al.’s extraction method is of operational nature. The minerals dissolved in each step

NU

depend on sediment sample composition and may not always from the targeted minerals (Clark et al., 1996).

MA

2.2 Batch leaching experiments

Sodium chloride (NaCl), hydrochloric acid (HCl), sodium hydroxide (NaOH), sodium

TE

D

bicarbonate (NaHCO3), sodium citrate dihydrate (C6H9Na3O9.2H2O), and sodium ascorbate (C6H7NaO6) were all analytical grade and purchased from VWR Canada. Humic acid (natural

AC CE P

organic matter (NOM)) was obtained from Alfa Aesar. All the solutions used in our experiments were prepared by dissolving the chemicals in de-ionized water. To evaluate the effects of pH, citrate, Eh, bicarbonate, and natural organic matter (NOM) on mineral dissolution and uranium release from sediment to water, four types of batch leaching experiments were conducted: (i) pH experiments; (ii) Eh experiments; (iii) bicarbonate experiments; and (iv) NOM experiments. We carried out four pH experiments (pH = 3, 5, 8 and 10), four Eh experiments (Eh = +200 to +300 mV, pH = 10; Eh = +200 to +300 mV, pH = 3; Eh = +50 to -150 mV, pH = 10; Eh = +50 to -150 mV, pH = 3), two bicarbonate experiments (bicarbonate concentration = 0.01 M, pH = 8; bicarbonate concentration = 0.001 M, pH = 8), and two NOM experiments (humic acid concentration = 50 mg C/L, pH = 8; humic acid concentration = 20 mg C/L, pH = 8). To prepare a sample for leaching experiments, one gram

9

ACCEPTED MANUSCRIPT (1.000 g) air-dried, sieved sediment sample was mixed with 40 ml background solution in a 50 ml HDPE centrifuge tube. The background solution was 0.01 M NaCl solution for pH

PT

experiments, citrate or citrate + ascorbate in 0.01 M NaCl solution for Eh experiments (Table 1),

RI

0.01 M or 0.001 M bicarbonate in 0.01 M NaCl solution for bicarbonate experiments, and 20 mg C/L or 50 mg C/L humic acid in 0.01 M NaCl solution for NOM experiments. The pH of each

SC

sample was adjusted and maintained by adding 1 M NaOH and/or 1M HCl to the suspensions.

NU

The total volume of the NaOH + HCl solution used was very small (< 0.5 mL), so that the final volume of the solution was close to 40 mL. In preparing the background solutions used in our

MA

leaching experiments, we did not purge the solutions. Therefore, these solutions were not stripped of dissolved O2. As a result, in our low Eh experiments, dissolved O2 and ascorbate

TE

D

could coexist in the solution. Duplicate samples were prepared for all the pH, Eh, bicarbonate, and NOM experiments. The centrifuge tubes holding the samples were tightly capped and placed

AC CE P

on a shaker table. At pre-determined leaching time of 0, 1, 2, 4, 8, 16 days, duplicate sample tubes were sacrificed for each experiment. The tubes were taken off from the shaker and uncapped. The pH and Eh of the solution were immediately measured using Thermo Orion Kit Star A211 Ph Bt with pH electrode (8102 ROSS; Thermo Orion) and ORP electrode (Orion Sure-Flow Comb Redox Ele). Supernatant was withdrawn from each tube, promptly filtered through 0.45 µm nylon filters, and analyzed by ICP-MS for U, Fe, Mn, Al, Si, Mg and Ca. Reference solutions with mixed metals were used to the check the accuracy of the analysis. The typical recovery for Fe, Mn, Al, Si, Mg, and Ca for these referecne solutions ranged from 90% to 110%. For U, the recovery was from 94% to 105%. The limit of detection (LOD) for U, Fe, Mn, Al, Si, Mg and Ca was 0.56 µgL-1, 115 µgL-1, 0.17 µgL-1, 5.3 µgL-1, 250 µgL-1, 1.2 µgL-1, 126 µgL-1, respectively. Total dissolved inorganic carbon (TIC) for a selected number of filtered

10

ACCEPTED MANUSCRIPT supernatant samples from our pH and bicarbonate experiments were determined by a Shimadzu TOC-V analyzer.

PT

3. Results and discussion

RI

3.1 Characterization of sediment

SC

3.1.1 Mineralogical composition

X-ray diffraction (XRD) profile shows that quartz, albite and microcline constitute the major

NU

mineral phases in the sediment sample (Fig. 2). Many unidentified small peaks in the XRD

MA

profile suggest the existence of other minerals. These minerals could not be identified because their quantity was below the detection limit of XRD analysis (~ 1 to 5% of the total mass)

D

(Pecharsky and Zavalij, 2008). SEM-EDX analysis substantiated the presence of quartz, albite

TE

and microcline (potassium feldspar), and identified many other minerals (Table 2). Fe (oxy)hydroxides (0.59 wt.%) (Table 2 and Fig. 3B) and pyrites (FeS2) (0.14 wt.%) (Table 2 and

AC CE P

Fig. 3C) were identified in SEM-EDX analysis. SEM-EDX showed the presence of few coffinite (U(SiO4)1-x(OH) 4x) grains that contain U(IV) (Fig. 3A). No uranium dioxide (UO2) or minerals that contain U(VI) (i.e., uranyl ion) were positively identified. However, this does not exclude the possibility that uranyl ions are present in the sediment in dispersed forms (i.e., adsorbed ions on mineral surface and/or impurities in mineral structure) that are below the detection limit of SEM-EDX, which is ~ 0.01% of the total mass.

3.1.2 pH, organic carbon, and concentration and phase distribution of uranium and major elements The sediment has a pH value of 5.7, and organic carbon content of 0.01 wt%. The low organic content of this sediment suggests the majority of U (and all other elements) extracted in

11

ACCEPTED MANUSCRIPT the 4th step of our sequential extraction were from oxidizable minerals (e.g., pyrites, coffinite), rather than from organic matters. Total U concentration of the sediment sample measured by the

PT

one step total digestion method was 25.9 mg/kg (27.098 and 24.810 mg/kg respectively for the

RI

each of the duplicate samples). By summing U mass from each step of sequential extraction, total U concentration in the sediment sample was calculated as 26 mg/kg, very close to that

SC

measured by the one step total digestion method. The distribution of U in different phases is

NU

shown in Fig. 4. Weak acid extractable U (2nd extraction step) and U bound to Fe-Mn oxides (3rd extraction step) account for 35.4% and 21.6% of the total U, respectively. U (as well as other

MA

elements) extracted in the 2nd step of our sequential extraction were released due to desorption and/or dissolution of minerals under weakly acidic conditions. Although only small amount of Fe

TE

D

(oxy)hydroxides (0.59 wt%) was present in the sediment as shown by SEM-EDX, Fe (oxy)hydroxides are known to strongly adsorb U. Therefore, it is not surprising that Fe

AC CE P

(oxy)hydroxides is a major sink for U in this sediment. U is also found in exchangeable (weakly adsorbed) and oxidizable minerals, each of which amounts to 0.5% and 7.3% of the total U, respectively. Although coffinite (U(SiO4)1-x(OH) 4x) is the only U-bearing mineral identified by our SEM-EDX analysis, U was found in each step of our sequential extraction, indicating uranium in our sample exists as adsorbed ions on mineral surface and/or impurities in minerals. Both Fe (oxy)hydroxides and silicate minerals are known to absorb trace elements (e.g., U) (Chatain et al., 2005; Tessier et al., 1979). The residual phase (mostly silicate minerals) holds a significant portion of U (35.2%). U most probably exists in the residual phase as impurities in the structure of silicate minerals, which are not easily soluble and resistant to dissolution in the first 4 steps of extraction.

12

ACCEPTED MANUSCRIPT Our sodium (bi)carbonate extraction showed that total adsorbed U in the sediment was 10.5 mg/kg. While this single step extraction method is quick and simple, sequential extraction

PT

method provides additional information on the amount of U that is incorporated into the

RI

insoluble minerals (i.e., “non-labile” U) and U distribution in different soluble minerals (Clark et al., 1996). The exchangeable U, weak acid extractable U, and certain fractions of the U bound to

SC

Fe-Mn oxides determined by our sequential extraction method are presumably associated with

NU

minerals via adsorption. The sum of U mass of the exchangeable, weak acid extractable and FeMn oxides bound is 26 mg/kg × (0.5% + 35.4% + 21.6%) = 15 mg/kg, higher than the total

MA

adsorbed U (10.5 mg/kg) determined by our sodium (bi)carbonate extraction. These results indicate some of the U associated with Fe-Mn oxides is either incorporated into the structure of

are dissolved.

TE

D

the minerals or is very strongly adsorbed, and therefore could not be released unless the minerals

AC CE P

The total Fe, Mn, Al and Ca concentration in the sediment sample was measured as 36.4, 0.9, 88.1 and 5.8 g/kg respectively by summing each fraction from sequential extraction steps. Total Si and phase distribution of Si could not be determined due to the use of HF and formation of volatile SiF4 gas when extracting the residual phase. However, we concluded, based on our SEM-EDX results, that the majority of Si in the sediment sample is in silicate minerals (e.g., albite, clays, feldspar, chlorite, quartz). Due to the use of high concentration MgCl2 (1 M) in extracting exchangeable phase elements, the amount of exchangeable phase Mg could not be determined. As a result, total Mg and phase distribution of Mg could not be determined. The distributions of Fe, Mn, Al and Ca in different phases are shown in Fig. 4. Sequential extraction showed that the largest pool of Fe is in the residual phase (silicate minerals), accounting for 88.4% of the total Fe. SEM-EDX results confirmed Fe is present in a 13

ACCEPTED MANUSCRIPT number of silicate minerals including clays, chlorite, almandine and biotite. Sequential extraction showed that Fe bound to oxidizable minerals and Fe-Mn oxides amounts to 4.8% and 4.5% of

PT

the total Fe, respectively. SEM-EDX identified Fe minerals including Fe (oxy)hydroxides,

RI

ilmenite and hematite, as well as pyrite and sphalerite (Table 2). Sequential extraction showed small amounts of Fe were in weak acid extractable phase (1.0%) and exchangeable phases

SC

(0.8%). These could be Fe adsorbed to minerals such as clays and Fe-Mn oxides.

NU

Sequential extraction showed total Mn concentration was much lower than that of Fe (0.9

MA

g/kg vs. 36.4 g/kg), and the largest pool of Mn is in the residual phase (silicate minerals), accounting for 84.7% of the total Mn. Mn bound to exchangeable phase, oxidizable minerals,

D

and Fe-Mn oxides accounts for 5.1%, 3.9% and 3.8% of total Mn, respectively. A small amount

TE

of Mn was found in the weak acid extractable phase (1.0%).

AC CE P

Sequential extraction showed that the largest pool of Al is in the residual phase (silicate minerals) accounting for 93.3% of the total Al. SEM-EDX identified a number of silicate minerals including albite, almandine, biotite, clay, feldspar and kaersutite (Table 2), which are the hosts for Al in the residual phase. Sequential extraction showed Al bound to oxidizable minerals and Al bound to Fe-Mn oxides amounts to 3.2% and 2.7% of the total Al, respectively. Small amounts of Al were found in weak acid extractable phase (0.5%) and exchangeable phase (0.2%). Sequential extraction showed the largest pool of Ca is in the residual phase (silicate minerals), accounting for 74.0% of the total Ca. Ca was identified by SEM-EDX in silicate minerals including plagioclase feldspar and kaersutite. The second largest pool of Ca is in the oxidizable minerals, amounts to 14.9% of the total Ca. Whereas no Ca was found in Fe-Mn

14

ACCEPTED MANUSCRIPT oxides, 11.1% of total Ca was found in the exchangeable phase. No Ca was found in the weak acid extractable phase by sequential extraction, but a few grains of calcite were identified in

PT

SEM-EDX analysis (Table 2).

RI

Although the amount of exchangeable Mg could not be quantified by our sequential

SC

extraction method, Mg in other phases was determined by this method. The “total” (excluding exchangeable phase) concentration of Mg in the sediment sample was measured as 6.3 g/kg, by

NU

summing the amount of Mg found in the weak acid extractable phase, Fe-Mn oxides, oxidizable

MA

minerals, and the residual phase. The majority of Mg was found in the residual phase (silicate minerals), accounting for 74.9% of “total” (excluding exchangeable) Mg. SEM-EDX confirmed

D

the presence of Mg-bearing silicate minerals including biotite, chlorite, and kaersutite (Table 2).

TE

Sequential extraction showed that Mg in oxidizable minerals amounts to 12.7% of “total” (excluding exchangeable) Mg, and Mg bound to weak acid extractable phase and Fe-Mn oxides

AC CE P

amounts to 6.8% and 5.6% of “total” (excluding exchangeable) Mg, respectively.

3.2 Batch leaching experiments 3.2.1 Effects of pH

U release from the sediment to water increased with increasing pH (Fig. 5). At low pH, U release was very low: maximum U concentration in water was 4.3, 2.0 and 5.7 µg/L, respectively at pH 3, 5 and 8. At pH 10, release of U was much higher: maximum U concentration was 22.0 µg/L. At pH 3, U concentration in water reached its maximum at t = 1 day, gradually decreased from t = 1 to 4 days, and reached a steady state concentration of 1.9 µg/L after day 4. The decrease in U concentration during day 1 to day 4 is probably due to U re-adsorption to minerals. At pH 5, water U concentration increased during the first day of leaching and was steady at 1.8

15

ACCEPTED MANUSCRIPT µg/L after day 1. At pH 8, U concentration increased steadily during the whole leaching period, reached its maximum of 5.7 µg/L at t = 16 days. At pH 10, U concentration increased gradually

PT

until day 8 and became stabilized at ~22 µg/L afterwards.

RI

By examining the release profile of U, Fe and Si at each pH, we found that the pattern of

SC

U release was similar to that of Fe and Si, and that the amount of U release was correlated to the amount of Fe and Si release (except for Si at pH 3), as indicated by linear regression of U

NU

concentration vs. Fe (or Si) concentration (Appendix A, Fig A.1 to A.4). Results of linear

MA

regression of U concentration vs. Fe concentration were summarized in Table 3. The correlation between U and Fe release (r2), as well as the slope of linear regression (indicating concentration

D

of U release per concentration of Fe release) were lower at lower pH. At lower pH, a fraction of

TE

Fe release was due to Fe desorption (e.g., desorption of exchangeable Fe), which did not contribute to U release, leading to lower r2 and lower slope. At higher pH, adsorbed Fe became

AC CE P

more stable, therefore a higher fraction of Fe release was due to mineral dissolution, leading to higher r2 and higher slope.

The y-intercept of linear regression of U concentration vs. Fe concentration, which is the concentration of U released when no Fe is released, represents concentration of U released due to U desorption under conditions that (i) Fe released is mostly from mineral dissolution, and (ii) U desorption rate is high. In our sediment, the amount of exchangeable Fe was much lower compared to Fe in other mineral phases (Fig. 4), and mineral dissolution was substantial in our leaching experiments, therefore, it is reasonable to assume that most of Fe released in our leaching experiments was due to mineral dissolution. Previous studies have shown that U desorption from minerals is very fast, while dissolution of U-bearing minerals is much slower (Giammar, 2001; Giammar and Hering, 2001; Giammar and Hering, 2002). Therefore, we used 16

ACCEPTED MANUSCRIPT the y-intercept of linear regression of U concentration vs. Fe concentration to estimate U desorption, and found that U desorption occurred at all pH and increased with increasing pH

PT

(Table 3): at pH 3, 5, and 8, U desorption was very low (0.9 to 1.8 µg/L), while at pH 10 U

RI

desorption was significantly higher (9.3 µg/L).

SC

The non-linear increase in U desorption with increasing pH was due to the non-linear nature of U adsorption to the sediment. At high pH, formation of aqueous U complexes (e.g., U-

NU

carbonate complexes, U-hydroxide complexes) could drastically decrease U adsorption (Barnett

MA

et al., 2002; Davis et al., 2004) and therefore increase U desorption. Dissolved inorganic carbonate concentration increases substantially with increasing pH in solutions open to the

D

atmosphere (Allison and Moodie, 1965). Therefore, it is important to quantify dissolved

TE

inorganic carbonate concentration in our pH experiments in order to ascertain if the change in U release was solely due to pH change or due to the combined effects of pH and dissolved

AC CE P

inorganic carbonate. Total dissolved inorganic carbonate (TIC) concentration was measured for a selected number of samples in our pH experiments. The TIC concentration of the pH 3, 5, 8, and 10 experiments after 1 day of leaching was 0.033 0.030, 0.031, and 0.033 mmol/L respectively, and was 0.029, 0.030, 0.030, and 0.033 mmol/L respectively after 4 days of leaching. These concentrations were within the range of equilibrium TIC concentration for open systems at pH < 7 (0.01-0.07 mmol/L) calculated by Visual MINTEQ (Gustafsson, 2009). The TIC concentrations in our high pH experiments were much lower than the equilibrium TIC concentration for systems open to the atmosphere based on thermodynamic calculation (0.66 mmol/L at pH = 8, and 170 mmol/L at pH 10, calculated by Visual MINTEQ). The above results confirmed that in our pH experiments, inorganic carbonate concentration was very low at both low and high pH, indicating our leaching solutions were well isolated from atmospheric CO2.

17

ACCEPTED MANUSCRIPT Calculations using Visual MINTEQ showed that in our experiments the major aqueous U species were UO22+ (98.6%) at pH 3, UO22+ (68.3%) + (UO2OH)+ (28.2%) at pH 5, (UO2(OH)3)- (38.4%)

PT

+ UO2(OH)2 (aqueous) (43.5%) + UO2CO3 (aqueous) (7.0%) + (UO2(CO3)2)2- (6.5%) +

RI

(UO2OH)+ (3.8%) at pH 8, and (UO2(OH)3)- (97.6%) at pH 10. These calculations demonstrated that U-hydroxide complexes was the dominant aqueous U species at high pH in our experiments.

SC

Therefore higher U desorption observed in our pH 10 experiment was caused by formation of

NU

aqueous U-hydroxide complexes rather than U-carbonate complexes.

MA

U desorption at each of our experimental pH was simulated using a surface complexation model developed by Waite et al. (1994). Although this model was calibrated using data of U

D

adsorption to synthetic ferrihydrite, Barnett et al. (2002) found the model can provide good

TE

prediction of U adsorption to heterogeneous natural sediments by assuming that Fe oxides (ranged from 25.3−25.8 g Fe/kg sediment) governed U adsorption to these heterogeneous natural

AC CE P

sediments, and that in terms of U adsorption, all the amorphous and crystalline Fe oxides behaved as ferrihydrite. In our calculation, we made similar assumptions that Fe oxides (determined by our sequential extraction as: 36.4 mg/kg × 4.5% = 1.64 mg/kg) governed U adsorption/desorption to our sediment, and that Fe oxides adsorb/desorb U in the same way as ferrihydrite. We also assumed that total U available for adsorption/desorption in each leaching experiment was 262.5 µg/L, based on U extracted by sodium (bi)carbonate solution (10.5 mg U/kg sediment) (Kohler et al., 2004). In our calculation, the strong and weak site densities, equilibrium constants for aqueous U-hydroxide and U-carbonate complexes, surface acidity constants, and specific surface area etc. were taken from Waite et al. (1994) and Barnett et al. (2002). Computer program FITEQL 4.0 (Herbelin and Westall, 1999) was used to implement model calculations. At pH 5, 8, and 10, model calculated U desorption was 9.0%, 0.3%, and

18

ACCEPTED MANUSCRIPT 12.0% of the total available U (262.5 µg/L), close to or higher than the experimentally determined U desorption of 0.5%, 0.7%, and 3.5%. At pH 3, U desorption was largely over

PT

predicted by the model (99.5% by the model vs. 0.4% by our experiments). Over prediction of U

RI

desorption by the model is attributed to U adsorption to minerals other than Fe oxides. In our modeling calculations, we assumed Fe oxides governed U adsorption/desorption. However, Fe

SC

oxides concentration in our sediment was very low and much lower than that in Barnett et al.’s

NU

sediments (2002), therefore other minerals in our sediment might significantly adsorb U. Nonetheless, Waite et al.’s model (1994) gave reasonable approximation of U desorption at pH ≥

MA

5 for our experiments.

D

The amount of U release at each pH due to mineral dissolution can be estimated by the

TE

difference between total U release and U release due to desorption (Table 3, column #6 = column #5 - column #3). The ratio of U release due to mineral dissolution to U release due to desorption

AC CE P

at each pH was shown in Table 3 (column #7). At all pH but 5, this ratio > 1, indicating dominant U release mechanism was mineral dissolution. At pH 3, U release was correlated to Fe release (r2 = 0.71), but not Si release (r2 = 0.08) (Appendix A, Fig. A1), indicating dissolution of Fe (oxy)hydroxide rather than Fe-rich silicate minerals was the major contributor to U release. At pH 8 and 10, U release was correlated to both Fe and Si release: r2 = 0.92 (for U vs. Fe) and 0.95 (for U vs. Si) at pH 8 (Appendix A, Fig. A2); r2 = 0.90 (for U vs. Fe) and 0.90 (for U vs. Si) at pH 10 (Appendix A, Fig. A3), indicating dissolution of Fe-rich silicate minerals was the major contributor to U release. At pH 5, both Fe (oxy)hydroxide and silicate minerals were stable (as indicated by low Fe and Si release, Fig 6a and 6d), desorption replaced mineral dissolution as the more important U release mechanism.

19

ACCEPTED MANUSCRIPT When pH increased from 3 to 10, U release increased 5 folds, and Fe and Si release increased 5 folds and 2 folds respectively, while the release of all other major elements decreased

PT

(Fig. 6a to 6f). These results suggested that U released by mineral dissolution was related to the

RI

dissolution of Fe- and Si-rich minerals such as Fe (oxy)hydroxides and silicate minerals. At pH 3 and 5, Fe release was low: Fe released at t = 16 days was 271 and 491 µg/L, respectively. At pH

SC

8 and 10, Fe concentrations were much higher: Fe released at t = 16 days was 2,987 and 3,760

NU

µg/L, respectively (Fig. 6a). As indicated by our sequential extraction and SEM-EDX analysis, most Fe in the sediment is contained in certain silicate minerals (e.g., chlorite, Fe-rich clays) and

MA

Fe (oxy)hydroxides. At low pH, both Fe (oxy)hydroxides and Fe-rich silicate minerals could have dissolved and released Fe. In addition, desorption of exchangeable Fe (i.e., adsorbed Fe)

TE

D

could have contributed to Fe release at low pH (Davis and Leckie, 1978; Oliveira et al., 2003; Tombácz et al., 2004). However, the amount of exchangeable Fe is much lower compared to Fe

AC CE P

in Fe (oxy)hydroxides and silicate minerals (Fig. 4). At high pH, Fe (oxy)hydroxides and exchangeable Fe became stable, however, the solubility of silicate minerals increased (as shown by the increase in Si release at high pH, Fig. 6d), leading to high Fe and U release.

3.2.2 Effects of citrate and low redox potential (Eh) 3.2.2.1 Citrate Citrate significantly increased U release: 166.8 µg/L (at pH 10) and 122.9 µg/L (at pH 3) U were released in the presence of 0.03 M citrate (Fig. 7), whereas in citrate free experiments (Fig. 5), U release was only 22.0 µg/L (at pH 10) and 4.3 µg/L (at pH 3). Compared to citrate free experiments, Fe, Mn, Al, Si, and Mg release was higher in the presence of citrate (Fig. 8a8b). The higher major element release is attributed to formation of aqueous Fe- and Al-citrate

20

ACCEPTED MANUSCRIPT complexes (Engelmann et al., 2003; Francis and Dodge, 1993), which promotes dissolution of Fe (oxy)hydoxides and silicate minerals.

PT

By examining the release profile of U and major elements, we found that in the presence

RI

of 0.03 M citrate, U release pattern was similar to that of Fe/Al/Si, and there was a positive and

SC

reasonably high correlation between U release and Fe/Al/Si release (Appendix B, Fig. B.1 and B.2), suggesting U release was related to the dissolution of Fe (oxy)hydoxides and/or Fe-rich

NU

silicate minerals. U released due to desorption, estimated by the y-intercept of linear regression

MA

of U vs. Fe release, was 72.7 µg/L U (at pH 10) and 30.1 µg/L (at pH 3) in the presence of citrate (Table 3), much higher than that in the absence of citrate, which was 9.3 µg/L U (at pH 10) and

D

0.9 µg/L (at pH 3). Calculations using Visual MINTEQ showed that in the presence of 0.03 M

TE

citrate, major aqueous U species were (UO2-citrate)- (95.3%) and UO22+ (4.6%) at pH 3, while at pH 10, formation of aqueous U-citrate complexes was unimportant (< 0.1%), and the dominant

AC CE P

aqueous U species was (UO2(OH)3)- (97.5%) either in the presence and absence of citrate. The above results demonstrated that at acidic pH, formation of non-adsorbing aqueous U-citrate complex contributed to the higher U desorption compared to citrate-free experiments (Bailey et al., 2005; Huang et al., 1998; Kantar, 2001; Lozano et al., 2011; Pasilis and Pemberton, 2003). However, at high pH, higher U desorption compared to citrate-free experiments was due to citrate competition for surface adsorbing sites only (Kantar, 2001). U release due to mineral dissolution was estimated by the difference between total U release and U released due to desorption (Table 3, column 6). U released due to mineral dissolution was higher in the presence of citrate compared to those in citrate free experiments: 94.1 µg/L vs. 12.7 µg/L (at pH 10), and 92.8 µg/L vs. 3.4 µg/L (at pH 3). The higher U release due to mineral dissolution in the presence of citrate can be explained by citrate enhanced 21

ACCEPTED MANUSCRIPT dissolution of Fe oxyhydoxides and silicate minerals. The relative importance of desorption and mineral dissolution to U release is evaluated by the ratio of concentration of U release due to

PT

desorption to concentration of U release due to mineral dissolution (Table 3, column 7). At both

RI

pH 3 and 10, the ratio is greater than 1, indicating mineral dissolution was the more important U release mechanism. However, when pH increased from 3 to 10, the ratio decreased from 3.1 to

SC

1.3, indicating the importance of U desorption increased at higher pH.

NU

3.2.2.2 Low Eh

MA

The presence of 0.06 M ascorbate substantially reduced redox potential (Eh) and U release (Fig. 7). In the presence of 0.03 M citrate, Eh was in the range of +200 to +300 mV, U

D

released after 16 days of leaching were 166.8 µg/L (at pH 10) and 122.9 µg/L (at pH 3). In the

TE

presence of 0.03 M citrate + 0.06 M ascorbate, Eh reduced to the range of +50 to -150 mV, and

AC CE P

U released after 16 days of leaching reduced to 80.3 µg/L (at pH 10) and 36.4 µg/L (at pH 3). Mineral dissolution and major element release was influenced by Eh (Fig. 8c-8d). Release of all the major elements (except Al and Si at pH 3) was higher at low Eh. For example, when Eh = +200 to +300 mV, 14,046 µg/L (at pH 10) and 31,015 µg/L (at pH 3) Fe were released after 16 days, whereas when Eh = +50 to -150 mV, 23,763 µg/L (at pH 10) and 41,000 µg/L (at pH 3) Fe were released after 16 days of leaching. The higher Fe release at lower Eh is attributed to reductive dissolution of Fe-rich minerals (Fe (oxy)hydroxides and/or Fe-rich silicate minerals) (Ahmed et al., 2004; Bhattacharya et al., 1997; Nickson et al., 2000). Enhanced dissolution of Fe-rich minerals at low Eh also facilitated the release of other elements in these minerals, which explains the higher Al, Si, Mg, and Ca release at low Eh. The only exception is Al and Si at pH 3: with decreasing Eh, Al release decreased 15% and Si release decreased 11%.

22

ACCEPTED MANUSCRIPT U release, in contrast to most of the major elements, was substantially lower at lower Eh compared to that at high Eh (52% lower at pH 10 and 71% lower at pH 3). The higher

PT

dissolution of minerals at low Eh, especially Fe (oxy)hydroxides and silicate minerals, which are

RI

the main U hosting minerals in our sediment, is expected to promote U release. The lower U release observed at low Eh suggests processes that reduced aqueous U concentration (e.g., U re-

SC

adsorption, precipitation) occurred in the presence of 0.06 M ascorbate.

NU

By examining the release pattern of U and major elements, we found that in our citrate +

MA

ascorbate experiments, at pH 10, major element concentrations increased steadily during leaching. However, U concentration increased only during the first 2 days of leaching, peaked at

D

100 µg/L at day 2, after which decreased and leveled off after day 8 (Appendix B, Fig. B3). At

TE

pH 3, U concentration was high (100 µg/L) at the beginning of leaching, but decreased afterwards, while major element concentration in water increased steadily during the entire

AC CE P

leaching process (Appendix B, Fig. B.4). Even though in our citrate + ascorbate experiments, mineral dissolution was an important mechanism of U release, linear regression of aqueous U concentration vs. major element release at pH 10 was characterized by low r2 (Appendix B, Fig. B3), indicating geochemical processes other than mineral dissolution might have altered aqueous U concentration. Visual MINTEQ calculations showed that at pH 10, in our citrate + ascorbate experiments, reduction of U(VI) to U(IV) was thermodynamically unfavorable and the leaching solution was under-saturated with respect to U(IV) minerals. Therefore, the decrease in aqueous U concentration after the initial increase is unlikely caused by reductive precipitation of U(VI). Re-adsorption of the aqueous U is more likely. At pH 3, aqueous U concentration was negatively correlated to major element release (i.e., negative slope and high r2) (Appendix B, Fig. B4), indicating aqueous U was removed via mechanisms that were positively correlated to major

23

ACCEPTED MANUSCRIPT element release, which was controlled by reductive dissolution of Fe-rich minerals. Visual MINTEQ calculations showed that at pH 3, reduction of U(VI) to U(IV) was thermodynamically

PT

favorable and the leaching solution was over-saturated with respect to U(IV) minerals, both at

RI

the beginning and the end of leaching. For example, after 16 days leaching, aqueous U concentration was 36 µg/L and the saturation index for U4O9, UO2 (amorphous), and uraninite

SC

was 12.5, 3.2, and 6.4 respectively. Those calculations confirmed that in our low Eh

NU

experiments, at pH 3, U(VI) was reduced to U(IV) and precipitate as U(IV) minerals with very low solubility (Bargar et al., 2008; Moon et al., 2009; Tokunaga et al., 2005), leading to decrease

MA

in aqueous U concentration.

D

3.2.3 Effects of bicarbonate

TE

At a fixed pH 8, bicarbonate greatly enhanced U release from sediment to water. Uranium concentration in water after 16 days of leaching was 5.7, 10.8, and 114.8 µg/L,

AC CE P

respectively for the non-bicarbonate, 0.001 M bicarbonate, and 0.01M bicarbonate experiment (Fig. 9). Major element release was not significantly affected by bicarbonate concentration (Fig. 10a to 10e). When bicarbonate concentration increased from near zero to 0.01 M, U release increased 20 fold. However, major element release decreased either moderately (Mn and Si) or slightly (Fe, Al, and Mg). These results showed that bicarbonate does not promote mineral dissolution, consistent with previous report (Kohler et al., 2004). Total dissolved inorganic carbonate (TIC) concentration was measured for a selected number of samples in our bicarbonate experiments to determine if there were substantial changes in bicarbonate concentration during leaching. For the 0.001 M (1 mmol/L) bicarbonate experiment, TIC concentration after 1, 2, and 4 days of leaching was 0.9 0.7, and 0.8 mmol/L, respectively; for the 0.01 M (10 mmol/L) bicarbonate experiment, TIC concentration after 1, 2, 24

ACCEPTED MANUSCRIPT and 4 days of leaching was 8.0, 7.4, and 8.3 mmol/L respectively. Measured TIC concentration was negligible low for the non-bicarbonate experiment (0.031 and 0.030 mmol/L, respectively

PT

after 1 and 4 days of leaching). The above results confirmed that TIC concentrations in our

RI

bicarbonate experiments were relatively stable and close to the added bicarbonate concentration. Exchange of atmospheric CO2 with aqueous CO2 is a slow process, and our sample tubes were

SC

tightly capped, which further prevented exchange of CO2 between atmosphere and the solution.

NU

In addition, there were almost no carbonate minerals in our sediment (shown by our SEM-EDX and sequential extraction results), so the sediment was not a significant source of bicarbonate in

MA

our leaching experiments. All the above factors explain the stable TIC concentration in our bicarbonate experiments. In our 0.001 M and 0.01 M bicarbonate experiments, TIC

TE

D

concentrations were a little lower than the added TIC concentrations, presumably due to bicarbonate adsorption to minerals.

AC CE P

By examining the release profile of U and major elements, we found that in the nonbicarbonate experiment, the pattern of U release and that of major elements (except Ca, which was an outliner) was similar: both U release and major element release increased steadily all the way through the experiment (Appendix C, Fig.C1). Linear regression analysis of U release vs. major element release was characterized by high r2 and positive slope (except Ca, which was an outliner) (Appendix C, Fig.C1), suggesting U release was related to mineral dissolution. In the 0.001 M bicarbonate experiment, however, the pattern of U release and that of major elements was different (Appendix C, Fig. C.2), and linear regression of U release vs. major element release was characterized by low r2 and close to zero slope (Appendix C, Fig. C.2), indicating U release was not related to mineral dissolution. In the 0.01 M bicarbonate experiment, we found that U and major element release was much faster, and the pattern of U release and major

25

ACCEPTED MANUSCRIPT element release was similar: U concentration and the concentration of each major element increased rapidly at the beginning of leaching and reached its maximum at day 1, after which

PT

stayed around that concentration until the end of leaching (Appendix C, Fig. C.3). This pattern of

RI

U and major element release was very different from that in the non-bicarbonate experiment (rapid release vs. gradual release), suggesting different U release mechanisms. Under this

SC

condition, the correlation between U and major element concentration cannot be assessed

NU

appropriately by linear regression (Appendix C, Fig. C.3). However, the observation that Fe/Mn/Al/Si/Mg release in the 0.01 M bicarbonate experiment was very close to or even lower

MA

than that in the non-bicarbonate experiment (Fig. 10a to 10e), implies that the much higher and faster U release observed in the 0.01 M bicarbonate experiment compared to non-bicarbonate

TE

D

experiment (114.8 vs. 5.7 µg/L) was due to increased U desorption rather than mineral dissolution. At alkaline pH, bicarbonate promotes U desorption via formation of non-adsorbing

AC CE P

aqueous U-carbonate complexes (UO2CO3, (UO2(CO3)2)2-, and (UO2(CO3)3)4-) (Baborowski and Bozau, 2006; Barnett et al., 2002; Grenthe and Lagerman, 1991; Nguyen Trung et al., 1992; Regenspurg et al., 2009). High bicarbonate concentration was also found to increase U desorption rate (Liu et al., 2009). Calculations using Visual MINTEQ showed that the major aqueous U species in our 0.001 M bicarbonate experiments were (UO2(CO3)3)4- (70.6%) + (UO2(CO3)2)2- (28.0%), and in our 0.01M bicarbonate experiments were (UO2(CO3)3)4- (88.4%) + (UO2(CO3)2)2- (11.5%). These calculations demonstrated that the higher U desorption in our bicarbonate experiments was due to formation of aqueous U-carbonate complexes. U desorption for each of our bicarbonate experiment was simulated using the surface complexation model developed by Waite et al. (1994) (described in section 3.2.1). While experimentally determined U desorption was 0.7%, 2.6%, and 42.3% of the total available U for

26

ACCEPTED MANUSCRIPT the non-bicarbonate, 0.001 M bicarbonate, and 0.01M bicarbonate experiment respectively, model calculated U desorption was 0.3%, 18.7%, and 87.5% respectively. As discussed in

PT

section 3.2.1, over prediction of desorption was caused by U adsorption to minerals other than Fe

RI

oxides, which was an important mechanisms in our experiments but was not included in the model. Despite of the over prediction, Waite et al.’s model provided reasonable approximation of

NU

3.2.4 Effects of natural organic matter (NOM)

SC

the effects of bicarbonate on U desorption for our experiments.

MA

NOM promoted U release from sediment to water. After 16 days of leaching, 5.6, 12.4, and 21.2 µg/L uranium were released in the non-NOM, 20 mg C/L, and 50 mg C/L experiment,

D

respectively (Fig. 11). NOM also influenced major element release, but in a different manner:

TE

with increasing NOM concentration, Fe release increased marginally from 3000 to 3244 µg/L

AC CE P

(Fig 12a), and release of Mn/Al/Si/Mg decreased (Fig. 12b to 12e). The observed decrease in major element release (except Fe) with increasing NOM concentration is consistent with previous reports that NOM reduces silicate mineral dissolution (Jones and Tiller, 1999; Tombácz et al., 2004). The decrease in silicate mineral dissolution is due to adsorption of negatively charged NOM to mineral surface, which neutralizes the positive charges of metal ions on mineral surface and therefore reduces mineral solubility (Gu et al., 1994; Specht et al., 2000). Although NOM could reduce Eh and promote reductive dissolution of Fe(III) minerals (Gu et al., 2005), in our experiments Eh was not influenced by NOM (Appendix E, Fig. E.1). Therefore, reductive dissolution of Fe(III) minerals is unlikely to be the dominant U release mechanism in our NOM experiments. The increased U release observed in our NOM experiments was presumably due to NOM induced U desorption: (i) NOM competes with U(VI) for sorption sites on oxides and clay minerals, which increases U desorption (Bednar et al., 27

ACCEPTED MANUSCRIPT 2007), and (ii) NOM forms low-adsorbing aqueous complexes with U(VI), which also increases U desorption (Lenhart and Honeyman, 1999). The profile of U and major element release in our

PT

NOM experiments showed that U concentration and major element concentration reached its

RI

maximum during the first a few days of leaching, and stayed around that maximum concentration during the remaining leaching process (Appendix D, Fig. D.1 and D.2). This

SC

pattern of U and major element release was very similar to that in our 0.01 M bicarbonate

NU

experiment, suggesting similar U release mechanisms (i.e., U desorption) in the bicarbonate and NOM experiments. The ability of NOM to promote U desorption, however, was much lower

MA

compared to that of bicarbonate in our experiments. In the presence of 50 mg C/L NOM, only 21.2 µg/L U was released, while in the presence of 0.01 M bicarbonate, 114.8 µg/L U was

AC CE P

4. Conclusions

TE

D

released.

U in a heterogeneous natural sediment sample is associated with a number of different mineral phases including silicate minerals, Fe-Mn (oxy)hydroxides, oxidizable minerals and exchangeable phases (i.e., weakly adsorbed U). U release from these minerals to water is a complicated process and controlled by a suite of interactive geochemical reactions including: dissolution of U-bearing minerals (especially Fe (oxy)hydroxides and silicate minerals), U desorption from mineral surface, formation of aqueous U complexes, and reductive precipitation of U. Water chemistry conditions were found to significantly influence the extent and mechanisms of U release from minerals to water. In the pH range of 3 to 10, both mineral dissolution and U desorption were found important for U release. While the extent of U desorption increased with increasing pH, U release due to mineral dissolution was high at both high and low pH, but was the lowest at near neutral pH. The presence of bicarbonate and citrate 28

ACCEPTED MANUSCRIPT in water significantly increased U release, and NOM moderately increased U release. While bicarbonate and NOM increased U release by promoting U desorption via formation of aqueous

PT

U complexes, U release mechanisms in the presence of citrate was more complicated. Increase in

RI

mineral dissolution and increase in U desorption, both due to formation of aqueous citrate complexes, contributed to the increased U release. Under reducing conditions, mineral

SC

dissolution in our experiments was generally higher compared to that under oxidizing conditions

NU

due to reductive dissolution of Fe-rich minerals, however, aqueous U concentration was lower under reducing conditions, which was due to reduction of U(VI) and subsequent formation of

MA

U(IV) minerals with low solubility.

D

Acknowledgements

TE

This work was supported by the Research & Development Corporation of Newfoundland and

AC CE P

Labrador’s Ignite R&D Program (Project #: 5404.1354.101). We are grateful to Dr. Stephen Amor of the Newfoundland & Labrador’s Department of Natural Resources for his help in selecting sampling sites and for allowing us to use his sampling equipment. The constructive comments and suggestions of three anonymous reviewers led to great improvements of this manuscript.

5. References Abdelouas, A., Lu, Y., Lutze, W. and Nuttall, H.E., 1998. Reduction of U(VI) to U(IV) by indigenous bacteria in contaminated ground water. Journal of Contaminant Hydrology, 35(1–3): 217-233.

29

ACCEPTED MANUSCRIPT Ahmed, K.M., Bhattacharya, P., Hasan, M.A., Akhter, S.H., Alam, S., Bhuyian, M., Imam, M.B., Khan, A.A. and Sracek, O., 2004. Arsenic enrichment in groundwater of the alluvial

PT

aquifers in Bangladesh: an overview. Applied Geochemistry, 19(2): 181-200.

RI

Allison, L. and Moodie, C., 1965. Carbonate. Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties: 1379-1396.

SC

Arai, Y., Marcus, M.A., Tamura, N., Davis, J.A. and Zachara, J.M., 2007. Spectroscopic

NU

Evidence for Uranium Bearing Precipitates in Vadose Zone Sediments at the Hanford 300-Area Site. Environmental Science & Technology, 41(13): 4633-4639.

MA

Baborowski, M. and Bozau, E., 2006. Impact of former mining activities on the uranium distribution in the River Saale (Germany). Applied Geochemistry, 21(6): 1073-1082.

TE

D

Bachmaf, S., Planer-Friedrich, B. and Merkel, B.J., 2008. Uranium sorption and desorption behavior on bentonite. Uranium, Mining and Hydrogeology: 515-524.

AC CE P

Bailey, E.H., Mosselmans, J.F.W. and Schofield, P.F., 2005. Uranyl-citrate speciation in acidic aqueous solutions—an XAS study between 25 and 200 °C. Chemical Geology, 216(1–2): 1-16.

Bargar, J.R., Bernier-Latmani, R., Giammar, D.E. and Tebo, B.M., 2008. Biogenic uraninite nanoparticles and their importance for uranium remediation. Elements, 4(6): 407-412. Barnett, M.O., Jardine, P.M. and Brooks, S.C., 2002. U (VI) adsorption to heterogeneous subsurface media: Application of a surface complexation model. Environmental Science & Technology, 36(5): 937-942. Bednar, A., Medina, V., Ulmer-Scholle, D., Frey, B., Johnson, B., Brostoff, W. and Larson, S., 2007. Effects of organic matter on the distribution of uranium in soil and plant matrices. Chemosphere, 70(2): 237-247.

30

ACCEPTED MANUSCRIPT Bhattacharya, P., Chatterjee, D. and Jacks, G., 1997. Occurrence of Arsenic-contaminated groundwater in alluvial aquifers from Delta Plains, Eastern India: Options for safe

PT

drinking water supply. International Journal of Water Resources Development, 13(1): 79-

RI

92.

Bowman, R.S., 1997. Aqueous Environmental Geochemistry. Eos, Transactions American

SC

Geophysical Union, 78(50): 586-586.

NU

Campbell, K.M., Veeramani, H., Ulrich, K.-U., Blue, L.Y., Giammar, D.E., Bernier-Latmani, R., Stubbs, J.E., Suvorova, E., Yabusaki, S. and Lezama-Pacheco, J.S., 2011. Oxidative

MA

dissolution of biogenic uraninite in groundwater at Old Rifle, CO. Environmental Science & Technology, 45(20): 8748-8754.

TE

D

Casas, I., de Pablo, J., Giménez, J., Torrero, M.E., Bruno, J., Cera, E., Finch, R.J. and Ewing, R.C., 1998. The role of pe, pH, and carbonate on the solubility of UO2 and uraninite

2231.

AC CE P

under nominally reducing conditions. Geochimica et Cosmochimica Acta, 62(13): 2223-

Catalano, J.G., McKinley, J.P., Zachara, J.M., Heald, S.M., Smith, S.C. and Brown Jr, G.E., 2006. Changes in uranium speciation through a depth sequence of contaminated Hanford sediments. Environmental Science & Technology, 40(8): 2517-2524. Chatain, V., Bayard, R., Sanchez, F., Moszkowicz, P. and Gourdon, R., 2005. Effect of indigenous bacterial activity on arsenic mobilization under anaerobic conditions. Environment International, 31(2): 221-226. Chen, B., Roos, P., Borggaard, O.K., Zhu, Y.-G. and Jakobsen, I., 2005. Mycorrhiza and root hairs in barley enhance acquisition of phosphorus and uranium from phosphate rock but mycorrhiza decreases root to shoot uranium transfer. New Phytologist, 165(2): 591-598.

31

ACCEPTED MANUSCRIPT Cheng, T., Barnett, M.O., Roden, E.E. and Zhuang, J., 2004. Effects of phosphate on uranium (VI) adsorption to goethite-coated sand. Environmental Science & Technology, 38(22):

PT

6059-6065.

RI

Clark, S., Johnson, W., Malek, M., Serkiz, S. and Hinton, T., 1996. A comparison of sequential extraction techniques to estimate geochemical controls on the mobility of fission product,

SC

actinide, and heavy metal contaminants in soils. Radiochimica acta, 74: 173-180.

NU

Davis, J.A. and Leckie, J.O., 1978. Surface ionization and complexation at the oxide/water interface II. Surface properties of amorphous iron oxyhydroxide and adsorption of metal

MA

ions. Journal of colloid and interface science, 67(1): 90-107. Davis, J.A., Meece, D.E., Kohler, M. and Curtis, G.P., 2004. Approaches to surface

TE

D

complexation modeling of uranium (VI) adsorption on aquifer sediments. Geochimica et Cosmochimica Acta, 68(18): 3621-3641.

AC CE P

Duff, M., Hunter, D., Bertsch, P. and Amrhein, C., 1999. Factors influencing uranium reduction and solubility in evaporation pond sediments. Biogeochemistry, 45(1): 95-114. Echevarria, G., Sheppard, M.I. and Morel, J., 2001. Effect of pH on the sorption of uranium in soils. Journal of Environmental Radioactivity, 53(2): 257-264. Engelmann, M.D., Bobier, R.T., Hiatt, T. and Cheng, I.F., 2003. Variability of the Fenton reaction characteristics of the EDTA, DTPA, and citrate complexes of iron. Biometals, 16(4): 519-527. Fanghanel, T. and Neck, V., 2002. Aquatic chemistry and solubility phenomena of actinide oxides/hyroxides. Pure and Applied Chemistry, 74(10): 1895-1908. Finch, R. and Murakami, T., 1999. Systematics and paragenesis of uranium minerals. Reviews in Mineralogy and Geochemistry, 38(1): 91-179.

32

ACCEPTED MANUSCRIPT Finch, R.J. and Ewing, R.C., 1992. The corrosion of uraninite under oxidizing conditions. Journal of Nuclear Materials, 190: 133-156.

PT

Finneran, K.T., Anderson, R.T., Nevin, K.P. and Lovley, D.R., 2002. Potential for

RI

bioremediation of uranium-contaminated aquifers with microbial U (VI) reduction. Soil and Sediment Contamination: An International Journal, 11(3): 339-357.

SC

Fox, P.M., Davis, J.A., Hay, M.B., Conrad, M.E., Campbell, K.M., Williams, K.H. and Long,

NU

P.E., 2012. Rate‐limited U (VI) desorption during a small‐scale tracer test in a heterogeneous uranium‐contaminated aquifer. Water Resources Research, 48(5).

MA

Francis, A. and Dodge, C., 1993. Influence of complex structure on the biodegradation of ironcitrate complexes. Applied and Environmental Microbiology, 59(1): 109-113.

D

Fredrickson, J.K., Zachara, J.M., Kennedy, D.W., Duff, M.C., Gorby, Y.A., Li, S.-m.W. and

TE

Krupka, K.M., 2000. Reduction of U (VI) in goethite (α-FeOOH) suspensions by a

AC CE P

dissimilatory metal-reducing bacterium. Geochimica et Cosmochimica Acta, 64(18): 3085-3098.

Giammar, D., 2001. Geochemistry of uranium at mineral-water interfaces: rates of sorptiondesorption and dissolution-precipitation reactions. Ph.D. Dissertation, California Institute of Technology. http://resolver.caltech.edu/CaltechETD:etd-05212002-125705 Giammar, D.E. and Hering, J.G., 2001. Time scales for sorption-desorption and surface precipitation of uranyl on goethite. Environmental Science & Technology, 35(16): 33323337. Giammar, D.E. and Hering, J.G., 2002. Equilibrium and kinetic aspects of soddyite dissolution and secondary phase precipitation in aqueous suspension. Geochimica et Cosmochimica Acta, 66(18): 3235-3245.

33

ACCEPTED MANUSCRIPT Giammar, D.E. and Hering, J.G., 2004. Influence of dissolved sodium and cesium on uranyl oxide hydrate solubility. Environmental Science & Technology, 38(1): 171-179.

PT

Gregorich, E. and Ellert, B., 1993. Light fraction and macroorganic matter in mineral soils. Soil

RI

Sampling and Methods of Analysis. Lewis Publ., Boca Raton, FL: 397-407. Grenthe, I. and Lagerman, B., 1991. Studies on metal carbonate equilibria: 23. Complex

SC

formation in the Th (IV)-H2O-CO2 (g) system. Acta Chem. Scand, 45: 231-238.

NU

Gu, B., Schmitt, J., Chen, Z., Liang, L. and McCarthy, J.F., 1994. Adsorption and desorption of natural organic matter on iron oxide: mechanisms and models. Environmental Science &

MA

Technology, 28(1): 38-46.

Gu, B., Yan, H., Zhou, P., Watson, D.B., Park, M. and Istok, J., 2005. Natural humics impact

5268-5275.

TE

D

uranium bioreduction and oxidation. Environmental Science & Technology, 39(14):

AC CE P

Gustafsson, J., 2009. Visual MINTEQ ver. 2.61. Department of land and water resources engineering, KTH Royal Institute of Technology, SE-100, 44. Herbelin, A. and Westall, J., 1999. FITEQL: A Computer Program for Determination of Chemical Equilibrium Constants from Experimental Data, 1999. Department of Chemistry, Oregon State University, USA: Corvallis. Herring, J.S., 2013. Uranium and thorium resources, Nuclear Energy. Springer, pp. 463-490. Huang, F.Y., Brady, P.V., Lindgren, E.R. and Guerra, P., 1998. Biodegradation of uraniumcitrate complexes: implications for extraction of uranium from soils. Environmental Science & Technology, 32(3): 379-382.

34

ACCEPTED MANUSCRIPT James, L.J. and Sinha, A.K., 2006. Geochemical coupling of uranium and phosphorous in soils overlying an unmined uranium deposit: Coles Hill, Virginia. Journal of Geochemcal

PT

Exploration, 91(1-3): 56-70.

Phenanthrene by a Soil Humic Acid: 

RI

Jones, K.D. and Tiller, C.L., 1999. Effect of Solution Chemistry on the Extent of Binding of A Comparison of Dissolved and Clay Bound

SC

Humic. Environmental Science & Technology, 33(4): 580-587.

NU

Kantar, C., 2001. The role of citric acid in the transport of uranium (VI) through saturated porous media: The application of surface chemical models to transport simulations of benchexperiments.

Ph.D.

Dissertation,

MA

scale

Colorado

School

of

Mines.

http://laer.mines.edu/people/abstracts/ckantarabstract.pdf

TE

D

Katsoyiannis, I.A., 2007. Carbonate effects and pH-dependence of uranium sorption onto bacteriogenic iron oxides: Kinetic and equilibrium studies. Journal of Hazardous

AC CE P

Materials, 139(1): 31-37.

Kelly, S.D., Newville, M.G., Cheng, L., Kemner, K.M., Sutton, S.R., Fenter, P., Sturchio, N.C. and Spötl, C., 2003. Uranyl Incorporation in Natural Calcite. Environmental Science & Technology, 37(7): 1284-1287. Kelly, S.D., Rasbury, E.T., Chattopadhyay, S., Kropf, A.J. and Kemner, K.M., 2006. Evidence of a Stable Uranyl Site in Ancient Organic-Rich Calcite. Environmental Science & Technology, 40(7): 2262-2268. Kohler, M., Curtis, G.P., Meece, D.E. and Davis, J.A., 2004. Methods for estimating adsorbed uranium (VI) and distribution coefficients of contaminated sediments. Environmental Science & Technology, 38(1): 240-247.

35

ACCEPTED MANUSCRIPT Lenhart, J.J. and Honeyman, B.D., 1999. Uranium (VI) sorption to hematite in the presence of humic acid. Geochimica et Cosmochimica Acta, 63(19): 2891-2901.

PT

Lett, R.E., 1995. Analytical methods for drift. . In Drift Exploration in the Canadian Cordillera,

RI

British Columbia, (ed.) P.T. Bobrowsky, S.J. Sibbick, J.M. Newell and P.F. Matysek; British Columbia Ministry of Energy. Mines and Petroleum Resources: Paper 1995-2, p.

SC

215-228.

NU

Levson, V.M., 2001. Regional till geochemical surveys in the Canadian Cordillera: sample media, methods and anomaly evaluation. In Drift Exploration in Glaciated Terrain, (ed.)

MA

M.B. McClenaghan, P.T. Bobrowsky, G.E.M. Hall and S.J. Cook. Association of Exploration Geochemistry - Geological Society of London: Special Publication 185, 45-

TE

D

68.

Liu, C., Shi, Z. and Zachara, J.M., 2009. Kinetics of uranium (VI) desorption from contaminated

AC CE P

sediments: Effect of geochemical conditions and model evaluation. Environmental Science & Technology, 43(17): 6560-6566. Liu, C., Zachara, J.M., Qafoku, O., McKinley, J.P., Heald, S.M. and Wang, Z., 2004. Dissolution of uranyl microprecipitates in subsurface sediments at Hanford Site, USA. Geochimica et Cosmochimica Acta, 68(22): 4519-4537. Lovley, D.R. and Phillips, E.J., 1992. Bioremediation of uranium contamination with enzymatic uranium reduction. Environmental Science & Technology, 26(11): 2228-2234. Lozano, J., Blanco Rodríguez, P., Vera Tomé, F. and Calvo, C.P., 2011. Enhancing uranium solubilization in soils by citrate, EDTA, and EDDS chelating amendments. Journal of Hazardous Materials, 198: 224-231.

36

ACCEPTED MANUSCRIPT McKinley, J.P., Zachara, J.M., Wan, J., McCready, D.E. and Heald, S.M., 2007. Geochemical controls on contaminant uranium in vadose Hanford formation sediments at the 200 Area

PT

and 300 Area, Hanford Site, Washington. Vadose Zone Journal, 6(4): 1004-1017.

RI

Mkandawire, M., 2013. Biogeochemical behaviour and bioremediation of uranium in waters of abandoned mines. Environmental Science and Pollution Research: 1-28.

SC

Moon, H., Komlos, J. and Jaffé, P., 2009. Biogenic U (IV) oxidation by dissolved oxygen and

NU

nitrate in sediment after prolonged U (VI)/Fe (III)/(SO4)2− reduction. Journal of Contaminant Hydrology, 105(1): 18-27.

MA

Nevalainen, R., 1989. Lithology of fine till fractions in the Kuhmo greenstrone belt area, eastern Finland. Geological Survey of Finland., Special Paper 7: 59-65.

TE

D

Nguyen Trung, C., Begun, G. and Palmer, D.A., 1992. Aqueous uranium complexes. 2. Raman spectroscopic study of the complex formation of the dioxouranium (VI) ion with a variety

AC CE P

of inorganic and organic ligands. Inorganic Chemistry, 31(25): 5280-5287. Nickson, R.T., McArthur, J.M., Ravenscroft, P., Burgess, W.G. and Ahmed, K.M., 2000. Mechanism of arsenic release to groundwater, Bangladesh and West Bengal. Applied Geochemistry, 15(4): 403-413. Oliveira, L.C., Rios, R.V., Fabris, J.D., Sapag, K., Garg, V.K. and Lago, R.M., 2003. Clay–iron oxide magnetic composites for the adsorption of contaminants in water. Applied Clay Science, 22(4): 169-177. Pasilis, S.P. and Pemberton, J.E., 2003. Speciation and coordination chemistry of uranyl (VI)citrate complexes in aqueous solution. Inorganic Chemistry, 42(21): 6793-6800. Pecharsky, V. and Zavalij, P., 2008. Fundamentals of powder diffraction and structural characterization of materials, 69. Springer.

37

ACCEPTED MANUSCRIPT Qafoku, N.P., Zachara, J.M., Liu, C., Gassman, P.L., Qafoku, O.S. and Smith, S.C., 2005. Kinetic desorption and sorption of U (VI) during reactive transport in a contaminated

PT

Hanford sediment. Environmental Science & Technology, 39(9): 3157-3165.

RI

Regenspurg, S., Schild, D., Schäfer, T., Huber, F. and Malmström, M.E., 2009. Removal of

Applied Geochemistry, 24(9): 1617-1625.

SC

uranium (VI) from the aqueous phase by iron (II) minerals in presence of bicarbonate.

NU

Sanding, A. and Bruno, J., 1992. The solubility of (UO2)3(PO4)2 · 4H2O(s) and the formation of U(VI) phosphate complexes: Their influence in uranium speciation in natural waters.

MA

Geochimica et Cosmochimica Acta, 56(12): 4135-4145. Sharp, J.O., Lezama-Pacheco, J.S., Schofield, E.J., Junier, P., Ulrich, K.-U., Chinni, S.,

TE

D

Veeramani, H., Margot-Roquier, C., Webb, S.M., Tebo, B.M., Giammar, D.E., Bargar, J.R. and Latmani, R.B., 2011. Uranium speciation and stability after reductive

6510.

AC CE P

immobilization in aquifer sediments. Geochimica et Cosmochimica Acta, 75(21): 6497-

Shilts, W., 1993. Geological Survey of Canada's contributions to understanding the composition of glacial sediments. Canadian Journal of Earth Sciences, 30(2): 333-353. Specht, C., Kumke, M. and Frimmel, F., 2000. Characterization of NOM adsorption to clay minerals by size exclusion chromatography. Water Research, 34(16): 4063-4069. Stubbs, J.E., Veblen, L.A., Elbert, D.C., Zachara, J.M., Davis, J.A. and Veblen, D.R., 2009. Newly recognized hosts for uranium in the Hanford Site vadose zone. Geochimica et Cosmochimica Acta, 73(6): 1563-1576. Sylvester, P.J., 2012. Use of the Mineral Liberation Analyzer (MLA) for mineralogical studies of sediments and sedimentary rocks. In Quantitative Mineralogy and Micro-Analysis of

38

ACCEPTED MANUSCRIPT Sediments and Sedimentary Rocks (P. Sylvester, ed.) Mineral.Assoc. Can. Short Course Series 42: 1-16.

PT

Tarvainen, T., 1995. The geochemical correlation between coarse and fine fractions of till in

RI

southern Finland. Journal of Geochemical Exploration, 54(3): 187-198. Tessier, A., Campbell, P.G. and Bisson, M., 1979. Sequential extraction procedure for the

SC

speciation of particulate trace metals. Analytical Chemistry, 51(7): 844-851.

NU

Tokunaga, T.K., Wan, J., Pena, J., Brodie, E.L., Firestone, M.K., Hazen, T.C., Sutton, S.R., Lanzirotti, A. and Newville, M., 2005. Uranium reduction in sediments under diffusion-

MA

limited transport of organic carbon. Environmental Science & Technology, 39(18): 70777083.

TE

D

Tombácz, E., Libor, Z., Illés, E., Majzik, A. and Klumpp, E., 2004. The role of reactive surface sites and complexation by humic acids in the interaction of clay mineral and iron oxide

AC CE P

particles. Organic Geochemistry, 35(3): 257-267. Ulrich, K.-U., Singh, A., Schofield, E.J., Bargar, J.R., Veeramani, H., Sharp, J.O., BernierLatmani, R. and Giammar, D.E., 2008. Dissolution of biogenic and synthetic UO2 under varied reducing conditions. Environmental Science & Technology, 42(15): 5600-5606. Waite, T.D., Davis, J.A., Payne, T.E., Waychunas, G.A. and Xu, N., 1994. Uranium(VI) adsorption to ferrihydrite: Application of a surface complexation model. Geochimica et Cosmochimica Acta, 58(24): 5465-5478. Wang, Z., Lee, S.-W., Kapoor, P., Tebo, B.M. and Giammar, D.E., 2013. Uraninite oxidation and dissolution induced by manganese oxide: A redox reaction between two insoluble minerals. Geochimica et Cosmochimica Acta, 100(0): 24-40.

39

ACCEPTED MANUSCRIPT Wang, Z., Zachara, J.M., McKinley, J.P. and Smith, S.C., 2005. Cryogenic Laser Induced U(VI) Fluorescence Studies of a U(VI) Substituted Natural Calcite: 

Implications to U(VI)

PT

Speciation in Contaminated Hanford Sediments. Environmental Science & Technology,

RI

39(8): 2651-2659.

Wazne, M., Korfiatis, G.P. and Meng, X., 2003. Carbonate effects on hexavalent uranium

SC

adsorption by iron oxyhydroxide. Environmental Science & Technology, 37(16): 3619-

NU

3624.

Welch, A.H. and Lico, M.S., 1998. Factors controlling As and U in shallow ground water,

MA

southern Carson Desert, Nevada. Applied Geochemistry, 13(4): 521-539. Wiedemeier, T.H., Wilson, J.T., Kampbell, D.H., Miller, R.N. and Hansen, J.E., 1995. Technical

TE

D

Protocol for Implementing Intrinsic Remediation with Long-Term Monitoring for Natural Attenuation of Fuel Contamination Dissolved in Groundwater. Volume II, DTIC

AC CE P

Document.

Williams, L.E., Barnett, M.O., Kramer, T.A. and Melville, J.G., 2003. Adsorption and transport of arsenic (V) in experimental subsurface systems. Journal of Environmental Quality, 32(3): 841-850.

40

ACCEPTED MANUSCRIPT Figure 1: Sampling location. Figure 2: XRD profile of the sediment sample.

PT

Figure 3: SEM images and respective EDX spectra showing (A) coffinite; (B) Fe oxyhydroxide

RI

and (C) pyrite.

SC

Figure 4: Weight percentage of U, Mn, Fe, Al and Ca in various phases of the sediment sample

NU

determined by sequential extraction.

Figure 5: Effects of pH on U release. Each symbol indicates the average from duplicate

MA

experiments. Height of the error bars indicates the difference from duplicate experiments. Figure 6: Effect of pH on the release of U, Fe, Mn, Al, Si, Mg and Ca at t = 16 days. a: U and Fe;

TE

D

b: U and Mn; c: U and Al; d: U and Si; e: U and Mg and f: U and Ca. Each symbol indicates the average from duplicate experiments. Height of the error bars indicates the difference from

AC CE P

duplicate experiments.

Figure 7: Effects of redox potential (Eh) on U release. The squares and circles indicate data from citrate experiments, and the diamonds and triangles indicate date from citrate + ascorbate experiments. Each symbol indicates the average from duplicate experiments. Height of the error bars indicates the difference from duplicate experiments. Figure 8: Influence of citrate and ascorbate on major element release from sediment to water at t = 16 days. a: citrate free vs. 0.03 M citrate at pH 3; b: citrate free vs. 0.03 M citrate at pH 10; c. 0.03 M citrate vs. 0.03 M citrate + 0.06 M ascorbate at pH 3 and d. 0.03 M citrate vs. 0.03 M citrate + 0.06 M ascorbate at pH 10. Each symbol indicates the average from duplicate experiments. Height of the error bars indicates the difference from duplicate experiments.

41

ACCEPTED MANUSCRIPT Figure 9: Effects of bicarbonate on U release. Each symbol indicates the average from duplicate experiments. Height of the error bars indicates the difference from duplicate experiments.

PT

Figure 10: Effects of bicarbonate concentration on the release of U, Fe, Mn, Al, Si and Mg at t =

RI

16 days. a: U and Fe; b: U and Mn; c: U and Al; d: U and Si and e: U and Mg. Each symbol

SC

indicates the average from duplicate experiments. Height of the error bars indicates the

NU

difference from duplicate experiments.

Figure 11: Effects of NOM on U release. Each symbol indicates the average from duplicate

MA

experiments. Height of the error bars indicates the difference from duplicate experiments. Figure 12: Effects of NOM concentration on the release of U, Fe, Mn, Al, Si and Mg at t = 16

TE

D

days. a: U and Fe; b: U and Mn; c: U and Al; d: U and Si and e: U and Mg. Each symbol indicates the average from duplicate experiments. Height of the error bars indicates the

AC CE P

difference from duplicate experiments.

42

ACCEPTED MANUSCRIPT Table 1. pH and redox potential (Eh) of the Eh experiments with various reducing agents. Table 2. Mineralogical compositions of the sediment sample identified by SEM-EDX analysis.

PT

Table 3. Linear regression results of U release vs. Fe release for pH experiments and experiments

AC CE P

TE

D

MA

NU

SC

RI

with citrate.

43

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Figure 1

44

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Figure 2

45

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Figure 3

46

ACCEPTED MANUSCRIPT

PT

Figure 4

SC

RI

U

NU

Ca

MA

Al

Weak acid extractable Fe-Mn oxides Oxidizable minerals Residual

Fe 20

40

AC CE P

0

TE

D

Mn

Exchangeable

60

47

80

100 wt (%)

ACCEPTED MANUSCRIPT

PT

Figure 5

RI

25

SC

15

pH-10

NU

pH-8

10

MA

pH-5

0 2

4

6

8

10

TE

0

D

5

Time (Days)

AC CE P

U released (µg/L)

20

48

pH-3

12

14

16

ACCEPTED MANUSCRIPT Figure 6

Total Fe

1000 0

0

2

4

6

8

0

10

0

pH

c

20

D

8000 15

TE

Total U Total Al

10

4000

5 0 0

2

4

6

8

Al released (µg/L)

MA

12000

AC CE P

U released (µg/L)

25

Total Mn

5

NU

0

25

0

2

4

15000

10

5000

5 0 0

1200

Total U

400

Total Mg

25

U released (µg/L)

U released (µg/L)

10000

Total U Total Si

15

2

4

6

8

10

pH

800

5

10

d

20

10

15 10

8

0

0

Mg released (µg/L)

e

20

6 pH

pH

25

500

Total U

25000

f

20

20000

15

15000

10

10000 Total U Total Ca

5

5000 BDLa

0

0 0

2

4

6

8

0

10

0 0

2

pH a

4

6 pH

Below detection limit

*Release of Ca was below detection limit of 126 µg L-1 at pH 10 49

8

10

Ca released (µg/L)

5

10

Mn released (µg/L)

2000 Total U

1000

RI

10

15

SC

3000

b

Si released (µg/L)

20

1500

PT

4000

U released (µg/L)

U released (µg/L)

15

25

U released (µg/L)

a

20

5000

Fe released (µg/L)

25

ACCEPTED MANUSCRIPT

PT

Figure 7

Eh (+200 to +300 mV); pH-10

240

RI

Eh (+200 to +300 mV); pH-3 Eh (+50 to -150 mV); pH-10

200

SC

NU

160

MA

120 80

0 0

2

4

TE

D

40

AC CE P

U released (µg/L)

Eh (+50 to -150 mV); pH-3

6

8 10 Time (Days)

50

12

14

16

ACCEPTED MANUSCRIPT Figure 8

pH 10

a

PT

20000 citrate free 0.03 M citrate

18000

12000 10000 8000

NU

30000

14000

SC

40000

20000

10000

0 Mn

Al

Si

6000 4000 2000 0

Ca

Fe

Mn

25000

Elements released (µg/L)

AC CE P

60000

c 40000

20000

Si

Mg

Ca

0.03 M citrate + 0.06 M ascorbate 0.03 M citrate

0.03 M citrate + 0.06 M ascorbate 0.03 M citrate

Al

pH 10

30000

TE

pH 3

80000

Mg

D

Fe

Elements released (µg/L)

citrate free 0.03 M citrate

RI

16000

MA

Elements released (µg/L)

50000

pH 3

Elements released (µg/L)

60000

20000

d

15000

10000

5000

0

0 Fe

Mn

Al

Si

Mg

Fe

Ca

51

Mn

Al

Si

Mg

Ca

ACCEPTED MANUSCRIPT

0.01 M bicarbonate; pH-8

1000

RI

0.001 M bicarbonate; pH-8

SC

Blank; pH-8

NU

100

MA

10

2

4

6

8

10

12

TE

0

D

1

Time (Days)

AC CE P

U released (µg/L)

PT

Figure 9

52

14

16

ACCEPTED MANUSCRIPT Figure 10

Total U

700

Total Fe

0

8000

Total U

2000

U released (µg/L)

D TE

c

40

4000

Al released (µg/L)

6000 80

8000

d

AC CE P

Total Si

80

800

400 Total U Total Mg

0 0

0 0 0.005 0.01 Bicarbonate concentration (M)

1000

600

40

0

200

Mg released (µg/L)

U released (µg/L)

e

4000 Total U

0 0 0.005 0.01 Bicarbonate concentration (M)

120

12000

80

40

100

0 0.005 0.01 Bicarbonate concentration (M)

Total Al

0

Total U Total Mn

120

MA

120 U released (µg/L)

0

0 0 0.005 0.01 Bicarbonate concentration (M)

200

NU

0

40

300

0 0.005 0.01 Bicarbonate concentration (M)

*Release of Ca was below detection limit of 126 µg L-1 at pH 8

53

Mn released (µg/L)

1400

40

PT

a

80

RI

2100

SC

80

500 400

U released (µg/L)

2800

120

Si released (µg/L)

3500

Fe released (µg/L)

U released (µg/L)

120

ACCEPTED MANUSCRIPT

50 mg C/L; pH-8 20 mg C/L; pH-8 Blank; pH-8

RI

25

SC

20

NU

15

MA

10 5

2

4

6

8

10

12

TE

0

D

0

Time (Days)

AC CE P

U released (µg/L)

PT

Figure 11

54

14

16

ACCEPTED MANUSCRIPT

Total U Total Fe

8000

MA

c

20

6000

D

15

4000

TE

10

2000

5

Total U Total Al

0 0

10

20

30

40

600

10

400 Total U Total Mg

5

PT

20

200

0 10

20

30

40

0

12000

10

4000

5

Total U Total Si 10 20 30 40 50 NOM concentration (mg C/L)

50

NOM concentration (mg C/L)

*Release of Ca was below detection limit of 126 µg L-1 at pH 8

55

Total Mn

100

8000

0 0

Total U

15

Mg released (µg/L)

U released (µg/L)

15

200

10 20 30 40 50 NOM concentration (mg C/L)

1000 800

300

d

25

0

50

e

20

5

0

0

NOM concentration (mg C/L) 25

10

0

10 20 30 40 50 NOM concentration (mg C/L)

AC CE P

U released (µg/L)

25

15

NU

0

400

0

0

U released (µg/L)

0

20

0

Si released (µg/L)

1000

5

500

RI

2000 10

25

SC

15

U released (µg/L)

3000

Fe released (µg/L)

a

20 U released (µg/L)

4000

Al released (µg/L)

25

Mn released (µg/L)

Figure 12

ACCEPTED MANUSCRIPT Table 1

Sodium citrate

0.03

3

Sodium citrate

0.03

10

Sodium citrate + sodium ascorbate

0.03 + 0.06

3

+200 to +300

SC

0.03 + 0.06

10

AC CE P

TE

D

MA

Sodium citrate + sodium ascorbate

Eh (mV)

PT

pH

56

+200 to +300

RI

Concentrations (M)

NU

Reagents

-150 to +50

-150 to +50

ACCEPTED MANUSCRIPT Table 2

SC

1.81 1.32 0.65 0.72 0.59 0.19 0.32 0.20 0.24 0.24 0.15 0.16 0.14 0.11 0.05 0.02 0.01 0.02 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00

MA

D

TE

AC CE P

57

Number of grains 10342 10229 9585 5553

PT

RI

Weight (%) 35.46 25.89 21.64 10.03

NU

Minerals Quartz (SiO2) Albite (NaAlSi3O8) Potassium feldspar (KAlSi3O8) Fe-poor Clays Chlorite ((Mg,Fe)3(Si,Al)4O10(OH)2.(Mg,Fe)3(OH)6) Plagioclase feldspar excluding albite (NaAlSi3O8 – CaAl2Si2O8) Fe-rich Clays (Fe spotted on clays) Kaersutite (NaCa2(Mg4Ti)Si6Al2O23(OH)2) Fe (oxy)hydroxides Fine-grain-silicate Titano-Fe-oxide Titanite (CaTiSiO5) Rutile (TiO2) Zircon (ZrSiO4) Biotite (K(Mg,Fe)3(AlSi3O10)(F,OH)2) Ilmenite (FeTiO3) Pyrite (FeS2) Almandine (Fe3Al2Si3O12) Apatite (Ca5(PO4)3(F,Cl,OH)) Calcite (CaCO3) Sphalerite ((Zn,Fe)S) Galena (PbS) Monazite ((Ce,La)PO4) Pyrrhotite (Fe1-xS (x = 0 to 0.2)) Bastnasite ((Ce,La,Y)CO3F)) Hematite (Fe2O3, α-Fe2O3) Chalcopyrite (CuFeS2) Coffinite (U(SiO4)1-x(OH)4x) Cr-Spinel (Mg(Al,Cr)2O4) Kozoite ((Nd,La,Sm,Pr)(CO3)(OH)) CaOH-Fe

1754 1353 727 713 576 465 286 401 196 145 282 131 164 123 102 21 39 12 41 10 8 7 5 2 1 1 1

ACCEPTED MANUSCRIPT Table 3 r2

(µg/L)

Total U release at day 16

Difference between total U release at day 16 and y intercept

(µg/L)

(µg/L)

PT

y intercept

RI

Slope

SC

pH

Ratio of U released due to mineral dissolution to U released due to desorption

0.0011

0.9

0.71

4.3

3.4

3.6

pH 5

0.0015

1.3

0.68

2.0

0.7

0.6

pH 8

0.0013

1.8

0.92

5.7

3.9

2.1

pH 10

0.0032

9.3

0.90

22.0

12.7

1.4

pH 3 (with citrate)

0.0027

30.1

0.79

122.9

92.8

3.1

pH 10 (with citrate)

0.0079

72.7

0.71

166.8

94.1

1.3

AC CE P

TE

D

MA

NU

pH 3

58

ACCEPTED MANUSCRIPT Highlights Fe (oxy)hydroxides and silicate minerals are the major U-hosting minerals

RI

Formation of aqueous U complexes increases U desorption

PT

U is released by U desorption and mineral dissolution at all pH

AC CE P

TE

D

MA

NU

SC

Low redox potential decreases aqueous U concentration due to reductive precipitation

59