ROLE OF TRACE ELEMENTS IN THE 226-RADIUM

Subscriber access provided by Iowa State University | Library

Article

ROLE OF TRACE ELEMENTS IN THE 226-RADIUM INCORPORATION IN SULFATE MINERALS (GYPSUM AND CELESTITE) Leslie Lestini, Catherine Beaucaire, Thomas Vercouter, Marine Ballini, and Michael Descostes ACS Earth Space Chem., Just Accepted Manuscript • DOI: 10.1021/ acsearthspacechem.8b00150 • Publication Date (Web): 16 Jan 2019 Downloaded from http://pubs.acs.org on January 17, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Earth and Space Chemistry

ROLE OF TRACE ELEMENTS IN THE 226-RADIUM INCORPORATION IN SULFATE MINERALS (GYPSUM AND CELESTITE)

Leslie Lestinia,c,d, Catherine Beaucairea,c, Thomas Vercouterb,c*, Marine Ballinid, Michaël Descostesd

aDEN-Service

d’Etude du Comportement des Radionucléides (SECR), CEA, Université Paris-Saclay, F-91191 Gif-sur-Yvette, France

bDEN-Service

d’Etudes Analytiques et de Réactivité des Surfaces (SEARS), CEA, Université ParisSaclay, F-91191 Gif-sur-Yvette, France

cLaboratory

Analysis and Modelling for Biology and Environment, Evry University, CNRS, CEA, UMR 8587, Bd. Francois Mitterrand,F-91025 Evry, France

dORANO

Mining- R&D Dpt, Tour AREVA, 1 Place Jean Millier, 92084 Paris La Défense, France

* Corresponding author. E-mail address: [email protected]

1

ACS Paragon Plus Environment

ACS Earth and Space Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Incorporation of

226Ra

within gypsum (CaSO4.2H2O(s)) and celestite (SrSO4(s)) was assessed through

dedicated batch experiments monitored over hundreds of days. Dissolution/recrystallization and coprecipitation experiments were carried out to investigate a range of chemical conditions, close to or far from equilibrium conditions. These data were used to establish the ability of gypsum and celestite to incorporate

226Ra.

On the contrary,

Celestite shows a high 226Ra

226Ra

incorporation, with a partition coefficient around 105.

is not significantly incorporated into pure gypsum. A very low value of

(8.8±1.5).10-4 has been estimated, which is much lower than the values reported in the literature. This may explain why 226Ra incorporation in gypsum is enhanced when Sr impurities are present (above 0.1 molar %). Under such conditions, the radium distribution coefficient is around 0.15± 0.09. This behavior can be explained by an ion-exchange mechanism between 226Ra and Sr, consistently with the existence of a solid solution between celestite and radium sulfate, similarly to the barite (BaSO4(s)) and radium sulfate solid solution. These results highlight the key role of trace compounds in the incorporation of 226Ra

in sulfate bearing minerals and bring new insights in our understanding of the 226Ra behavior in

the environment. This is illustrated by the radium content observed and modelled within the porewaters of uranium mining tailings.

Keywords: radium; gypsum; celestite; solid solution; distribution coefficient, mining tailings

2


Page 2 of 32

Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1.

INTRODUCTION

As a decay product of the

238U

decay chain,

relative high half-life (1620 y),

226Ra

226Ra

is naturally present in the uranium ore. Due to its

is predominant among other Ra isotopes. It is commonly

encountered in groundwaters1,2, deposits in oil pipelines3 or in desalinization plants of seawater4. According to its half-life, and its high specific activity, 226Ra is one of the main contaminants of concern (COC) in the uranium mining industry5,6. The abundance of

226Ra

is generally so low, in the range of

ppb, that no proper mineral phase has ever been found in natural environments. As a trace element, its mobility can then only be lowered by interactions with surrounding mineral phases present in its environment. It is actually well known that its behaviour is mainly controlled by surface interactions onto clays7,8, carbonates9, metal oxides10or amorphous phase11 and altered primary minerals. The fate of 226Ra

can also be controlled by solid solution formation processes12. As sulfate minerals are very

common natural phases found in most sedimentary environments, they give rise to a particular interest in their ability to fix radium or not. 226Ra was reported to be associated with sulfate minerals: gypsum (CaSO4.2H2O(s)) in the case of uranium mill tailings13, barite (BaSO4) and celestite (SrSO4) in brines from evaporation ponds of desalinisation plant14, or barite in potential high-level radioactive waste repositories12. Barite is the main sulfate phase that forms solid solution with 226Ra as first reported by Marie Curie15, and investigated by several researchers12,16-19. 226Ra

can be incorporated in the barite lattice either by co-precipitation or recrystallization processes, to

form a solid solution compound, Ba(1-x)RaxSO4(s). Such solid solutions were studied by different experimental approaches12,16,19. Although more differences (e.g. ionic radius, solubility of SO4 bearing minerals…) exist between Ra2+ and other divalent alkaline-earth elements like Sr2+, Ca2+, Mg2+, solid solution formation has been considered as well. A few studies by co-precipitation experiments with gypsum have been reported for the Ca-Ra-SO4-H2O system13,20,21, and to our knowledge, only two experimental studies reported the co-precipitation of 226Ra and Sr in the Sr-Ra-SO4 system22,23. Recently, different authors24,25,26 refined the thermodynamics properties of the ternary solid solution – aqueous solution system ((Sr,Ba,Ra)SO4, H2O). These thermodynamic data were comforted through long duration experiments carried out at 90°C. These conditions were chosen to evaluate the radium uptake

3



Page 4 of 32

in the context of the geological disposal of nuclear waste, which is somewhat different with mining environment. Interactions between

226Ra

and minerals of the barite isostructural family (MeSO4) are

usually described by the distribution coefficient DRa, mineral according to:

226Ra

([ ])solid Ra, mineral = [ ( ])aqueous [Me]

D226

(1)

226Ra [Me]

Where ([226Ra]/[Me])solid and ([226Ra]/[Me])aqueous represent the molar ratio in the solid and in the solution, respectively, between 226Ra and Me, the major element substituted. The experimental studies of the RaSO4-BaSO4 system provided DRa, barite values between 0.3 and 1.8, the highest values being obtained via co-precipitation experiments17,18, and the lowest value, obtained recently via recrystallization experiments12,19. For the RaSO4-CaSO4.2H2O system, DRa, gypsum values are ranging from 0.0320 to 0.321. A much higher DRa, celestite value of 340 has been reported by Goldschmidt for the RaSO4-SrSO4 system from one co-precipitation experiment22. However, a recent work reported experimental DRa, celestite values from 43 to below 1.0 from co-precipitation experiments23. The apparent discrepancy between the DRa values for each chemical system makes it difficult to establish a satisfactory model of 226Ra behavior with time. It is worth noting that too little attention has been given to

226Ra

interaction with celestite, for which the distribution coefficient DRa, celestite seems to be high

according to Goldschmidt’s report22. In the context of uranium mining and more particularly in the frame of remediation or environmental survey, the consolidation of the existing thermodynamic data related to the main minerals able to immobilize

226Ra

is necessary. Typically,

226Ra

activity is in the range of 1-10 Bq/L in mining waters

and 25-50 Bq/g in mill tailings5,13. Among the well-known phases that may trap

226Ra

such as clay-

minerals and ferric oxy-hydroxide, sulfate bearing minerals are ubiquitous in mining context (use of sulfuric acid and sulfate reactants in water treatment). They are also known to concentrate radioactivity. Hence, the purpose of our study is:

4


Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1) Re-examining the distribution coefficient of 226Ra between aqueous solutions and both gypsum and celestite, by conducting recrystallization experiments under near equilibrium conditions. Moreover the fact that minerals are rarely pure led us to consider additional co-precipitation experiments where 226Ra incorporation is studied in sulfate minerals in presence of impurities, such as Sr often present in gypsum. 2) Illustrating the environmental context by calculating the range of aqueous 226Ra activity at equilibrium with different solid-solution systems present in mining context. 2.

MATERIAL AND METHODS

Two types of experiments were performed: firstly, recrystallization experiments aim to estimate the rate of

226Ra

incorporation in a pure solid (CaSO4.2H2O or SrSO4), and secondly co-precipitation

experiments to evaluate the potential role of impurities on 226Ra incorporation. All the experiments were performed at 22°C. All solutions were prepared with deionized water (Milli-Q) and chemicals of ACS reagent grade. More details are given in the following sections for each type of experiments.

2.1.

Recrystallization experiments

The incorporation experiment of 226Ra in a solid near equilibrium conditions consists of following the evolution of 226Ra activity over time in the aqueous solution in contact with the solid phase initially free of 226Ra. The methodology has followed the one by Curti et al.12. The trace element is incorporated in the solid through a dissolution/recrystallization process, characteristic of the dynamic equilibrium between the solid and the solution. As the trace element incorporation in the solid phase depends on the dissolution/recrystallization rate, a kinetic monitoring of such processes was necessary. This is achieved by following the uptake of a radioisotope of the major element, *Me, substituted to Me in the solid. In fact, introducing this radioisotope in the solution creates an isotopic disequilibrium that will be balanced by incorporation of an amount of the radioisotope in order to reach the dilute isotopic equilibrium. So the same process as in the radium incorporation study is monitored, and gives access to the amount of solid recrystallized on a time scale. 45Ca and 85Sr have been selected to monitor the recrystallization of gypsum and celestite, respectively.

5



Page 6 of 32

In the case of such recrystallization experiments, Curti et al. 12 showed that it is possible to apply two different models. If the equilibrium is limited to a surface layer of the solid particles, then the heterogeneous recrystallization model must be considered. In this case the incorporation of the radiotracer, i.e. the dilute isotopic balance between the solid phase and the aqueous phase can be described by an exponential law12:

A(t)/A0(t) = exp– ([S/L.Ω.σ.t]/[Me]saturation)

(2)

where A(t) et A0(t) represent the activity of radio labeled metal *Me (i.e. 45Ca,85Sr) in the solution and the initial activity remaining in solution at the sampling time t, respectively. The activities were corrected for the natural decay of *Me. S/L represents the solid:liquid ratio (g.L-1), Ω is the dissolution/recrystallization rate (mol.m-2.day-1), σ is the specific surface of the solid (m².g-1), [Me]saturation is the Me (Ca, Sr) concentration in the solution equilibrated with the solid, and t is the interaction time (day). Concurrently, if the equilibrium is fully reached between aqueous solution and the solid, the homogeneous recrystallization model can be used. In this case, the incorporation of the radiotracer is described as follows:

A(t)/A0(t) = [Me]saturation/([S/L.Ω.σ.t]+[Me]saturation)

(3)

The S/L and [Me]saturation parameters are assumed to be constant during the experiment; σ is also assumed constant with time. Whatever the recrystallization model, the recrystallization rate Ω is then obtained by fitting the *Me activity data with time. On the one hand, incorporation experiments were run with synthetic gypsum, precipitated from a supersaturated solution, and with a commercial celestite (Sigma-Aldrich; pure at 99%). Analysis of the minerals showed that celestite is particularly rich in Ba (1000 ppm), while gypsum contains only Sr impurities (70 ppm). On the other hand, incorporation experiments were run with natural gypsum (Italy), containing around 270 ppm of Sr, and 20 ppm of Pb. All solid phases were characterized by powder

6


Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


XRD analysis (Inel XRG 3000 diffractometer) and by BET (Micromeritics ASAP 2010) surface experiment with N2 prior to the experiments. Specific surface of synthetic, natural gypsum and celestite are respectively found at 0.6, 2 and 3 m2.g-1. For each system studied, i.e. the

226Ra-gypsum

system and the

226Ra-celestite

system, the

226Ra

incorporation experiments and the recrystallization experiments have been run in separate batch solutions. For the 226Ra-gypsum system, two sets of ten batches of 10 mL each, with a solid:liquid (S/L) ratio of 32 g.L-1, have been prepared: one spiked with

226Ra

tracer (IPL source), the other one spiked

with 45Ca tracer (Cerca source). The slurries were then mounted on rotating end-over-end shakers to provide a continuous mixing. At regular time intervals, the batch solutions were centrifuged during 1h at 20 000 rpm. After filtration (Millipore filters, 0.2 m), 226Ra was analyzed by gamma spectrometry, using an N-type high coaxial purity germanium detector (ITECH-Instruments) in airtight plastic container (50 mL) at 186.21 keV. The InterWinner©6.0 software was used to calculate the activities taking into account the parent-daughter decay law. The detection limit is estimated at 10 ± 5 Bq or 1.10-12 moles. 45Ca

was measured by liquid scintillation (Packard TRI-CARB 2500TR). For each set, 3 standards have

been prepared as controls with only the solution in equilibrium with gypsum and spiked in the same way as the suspensions. The solutions were diluted with UltimaGold LT in a ratio 1/5. For the 226Ra-celestite system, incorporation experiments were carried out in a single batch of 0.6 L in presence of synthetic celestite, with a S/L ratio of 0.07 g.L-1 and spiked with

226Ra.

Recrystallization

experiments were done with ten batch solutions of 20 mL each, with a S/L ratio of 1.5 g.L-1, and spiked with 85Sr (Cerca source). 85Sr was measured by gamma counting (WIZARD 3”1480). The experimental conditions are summarized in table 1.

Table 1: Experimental conditions of recrystallization experiments. Solid

Spike

S/L

Volume

Duration

[Ca]sat

[Sr]sat

(g.L-1)

(mL)

(days)

(mmol.L-1)

(mmol.L-1)

Recrystallization experiment I-GS1 Synth. gypsum

45Ca

32

10

200

15

I-GS2 Synth. gypsum

45Ca

32

10

200

15

7



I-GN Nat. gypsum

45Ca

32

10

200

I-CS4 Synth. celestite

85Sr

1.5

20

60

226Ra

Page 8 of 32

15 0.7

recrystallization experiments

I-GS1 Synth. gypsum

226Ra

32

10

200

15

I-GS2 Synth. gypsum

226Ra

32

10

200

15

I-GN Nat. gypsum

226Ra

32

10

200

15


226Ra

0.07

600

66

0.7


226Ra

0.07

600

129

0.7


226Ra

0.07

600

159

0.7

2.2. Co-precipitation experiments Co-precipitation experiments were carried out in order to test the role of Sr impurities in sulfate solids on the

226Ra

incorporation rate. In the case of co-precipitation in a closed system where the

concentrations change, heterogeneous model is generally considered as more representative of the trace element uptake. The partition coefficient Tr, MeSO4 is defined as16:

Ln([Tr]end/[Tr]ini) = Tr,MeSO4 Ln([Me]end/[Me]ini)

(4)

Where [ ]ini and [ ]end represent the initial and final concentrations, respectively, of major (Me) and trace (Tr) elements. The homogeneous partition coefficient D as defined in eq. (1) can also a good approximation to model the Tr uptake in co-precipitation experiment when considering that the system is closed to the equilibrium with an homogeneous composition of solid, or when the composition of solution does not change drastically during the experiment. However it is important to keep in mind that the estimation of the partition coefficient is highly dependent on the precipitation rate and is indirectly related to the state of over-saturation27-30. Generally, the D values decrease as the precipitation rate increases. At high rate of precipitation, the substituting ions tend to incorporate in the solid in the same ratio of ions than in aqueous solution, therefore the partition coefficient would tend to unity. In the following experiments, we tried to limit the over-saturation as possible to limit the deviation of coefficient partition from its equilibrium value.

8


Page 9 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The 226Ra incorporation was studied in the (Ca,Srtr)SO4.2H2O system. Batch experiments were obtained by mixing a solution of CaCl2 (>99.0%, FLUKA®) and SrCl2,6H2O (>99.0% ALDRICH®) with a solution of Na2SO4 (>99.0% ALDRICH®) in a final volume of 20 mL.226Ra was spiked to the solution at an initial activity of around 300 Bq/mL just before the addition of the Na2SO4 reactant. The influence of the state of over-saturation -from 1.18 to 1.4 with respect to the more soluble solid phase (CaSO4.2H2O)- was investigated by the choice of different concentrations of the major element (Ca) in the solution. Such low saturation indexes are typically found in mining environments. Different concentrations of trace element (Sr) were investigated at constant Sr/Ca molar ratio (~ 0.003). The upper Sr concentration was chosen so as to avoid too large over-saturation with respect to celestite. The quantities of precipitated gypsum were calculated from the initial and final concentrations of Ca measured in batchs. Analyses of cations were realized by ion chromatography (IC Metrohm Vario 850). The uncertainty is estimated at 2%. Sr at trace level was analyzed by ICP-MS (Varian 810). The detection limit is estimated at 10-11 mol/L with an uncertainty of 5%. The 226Ra activity was counted by γ-spectrometry as previously indicated. In order to perform observations on Ca-Sr-SO4 solids (X-rays diffractometry), some experiments were carried out in the absence of 226Ra. Experimental conditions are summarized in supporting information (table SI-1). 3.

RESULTS

3.1.

Recrystallization experiments

Gypsum The data set obtained with the 45Ca incorporation experiments in gypsum shows a significant decrease of the radioisotope solution activity with time, consistent with the on-going dissolution/recrystallization processes occurring in the solid (Fig. 1). Heterogeneous and homogeneous models were tested successively. The experimental data could be fitted with the heterogeneous model considering two successive trends, suggesting two different recrystallization trends. The homogeneous model using eq. (3) with a unique rate of recrystallization was also able to satisfactorily fit the decrease of 226Ra activity, leading to the estimation of a rate of recrystallization of 8.0.10-6 mol.m-2.d-1.

9



Homogeneous model Heterogeneous model

1,0

A(45Ca)(t) / A0(45Ca)(t)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 32

0,8

0,6

0,4

0,2

0,0 0

50

100

150

200

Duration (days)

Fig. 1. Evolution of 45Ca activity normalized to the initial activity (A0) with time in solution at equilibrium with gypsum. Fit according to heterogeneous and homogeneous models of recrystallization are also plotted.

10


Page 11 of 32

350 300

A(226Ra)(t) (Bq.mL-1)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


250 200 150 100 50

Fit DRa, gypsum = 0.3 Fit DRa, gypsum = 0.03

0 0

50

100

150

200

Duration (days)

Fig. 2. Evolution of 226Ra activity with time in solution at equilibrium with gypsum. Simulation of 226Ra incorporation in CaSO4.2H2O is given for aDRa, gypsum values of 0.0320(dashed line) and 0.321 (dash-dotted line).

These values have been obtained considering a constant solid to liquid ratio (S/L) of 31.4 g.L-1, a specific surface of 0.5 m².g-1 and [Ca]saturation of 1.5.10-2 M. However, a better agreement between the experimental data and the modeling is observed after 10 days of interaction. As previously mentioned31, microscopic observations showing dehydration features at the solid surface led to the conclusion that the prolonged time of the solid in the stove during the preparation stage, has conducted to its dehydration and formation of bassanite (CaSO4.0,5H2O). This was confirmed by XRD characterization. The observed change in slope on the experimental curve is then assigned to the solid rehydration during the first days of the experiment.

11



Page 12 of 32

The amount of newly formed solid (recrystallized) [Ca]solid in moles during the experiment can be estimated as follows:

[Ca]solid = Ω.σ.m.t

(5)

where m represents the mass of solid introduced in solution (g). After 200 days of experiment, about 0.04 g of gypsum has been recrystallized (14 % of the initial solid). Similar conclusions were drawn when using natural gypsum and additional synthetic gypsum compounds (data not shown). The incorporation of 226Ra remains negligible despite the pretty high quantities of recrystallized gypsum of up to 50 % of the initial solid. The 226Ra activity in the solution A(226Ra) shows no evolution with time considering the uncertainty of the measurement (Fig. 2). Neither the amount of

226Ra

incorporated in the solid nor the distribution

coefficient could be determined with the following equation:

[226Ra]solid = DRa, gypsum.[226Ra]solution.[Ca]solid/[Ca]solution

(6)

An estimation of the distribution coefficient is better obtained from the co-precipitation experiments as detailed in the 3.2 Section. As it will be further discussed, we also plotted on Fig. 2 the simulated values of the final aqueous 226Ra activity in our experiments according to the D values taken from the litterature20,21. Our results clearly indicate a DRa, gypsum inferior or at least equal to the lowest value (i.e. 0.03) used reported in the literature.

Celestite The data collected via the recrystallization experiment run with 85Sr on the celestite show a significant decrease of 85Sr solution activity with time, suggesting a fast incorporation in the solid (Fig. SI-1). The dissolution/recrystallization rate, obtained by fitting the data and considering a specific surface of 3 m2.g-1 and a solid:liquid ratio of 1.48 g.L-1, is 3.16 mol.m-2.d-1. We estimated that 16 to 28 % of the

12


Page 13 of 32

solid has recrystallized over the period of time of the experiment depending on the experimental set considered. Unlike in the Ra-Gypsum system, the 226Ra uptake experiments run on celestite shows a significant and rapid decrease of

226Ra

solution activity with time, well reproduced in the three experiments. The

incorporation stabilized at about 40% of the initial

226Ra

present in the solution (instead of less than

0.03% for gypsum). This threshold was reached after 10 to 20 days (Fig. 3).

1,0

A(226Ra)(t)/A0(226Ra)(t)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


I-CS1 I-CS2 I-CS3

0,8

0,6

0,4

0,2

0,0 0

50

100

150

200

Duration (days)

Fig. 3. Evolution of 226Ra activity normalized to the initial activity (A0) with time in solution at equilibrium with celestite.

3.2. Co-precipitation experiments Ca-226Ratr-SO4 Whatever the initial conditions and experiment duration, the initial 226Ra concentration remains constant with time, confirming the results previously obtained with recrystallization experiments. However, in order to better evaluate the DRa, gypsum value, the quantity of radium adsorbed on solid was determined by alpha spectrometry over a long time of counting. The measured total activity of

13


226Ra

in gypsum is


13.3±2.7 Bq/g. The DRa,

gypsum

Page 14 of 32

value is therefore estimated at (8.8±1.5)x10-4, lower than the value

estimated in evaporitic natural samples4 (i.e. 0.003), and than the ones determined in previous works13,20,21 (see Fig.4).

Ca-Srtr-226Ratr-SO4 Experimental results of Sr co-precipitation with gypsum are summarized in the supporting information (table SI-2). Various over-saturation states of the initial solutions with respect to gypsum (from 1.18 to 1.4) and initial Sr concentration (from 0.2 to 3.0 mmol/L) were tested. These saturation indexes are relatively low and typical of the mining environment. The Sr incorporation amount in gypsum is varying from 3 to 35 % of initial Sr content. Values of DSr, gypsum are found between 0.1 and 0.45 and are weakly correlated with the Sr/Ca molar ratio in gypsum (see figure SI-2). For a same initial oversaturation state with respect to gypsum, a higher initial Sr concentration gives a higher Sr/Ca in gypsum. These values are close to the observations reported in previous experimental studies carried out in brine conditions and at different rate of crystallization33, with an estimated value of DSr, gypsum around 0.1 at equilibrium. Contrarily to the experimental run in absence of Sr,

226Ra

is slightly more incorporated in gypsum in

presence of Sr impurities. The quantity of precipitated solid was calculated from the mass balance measured on calcium. In the same way the amount of radium incorporated in the solid is calculated from the mass balance measured on radium. The data set obtained with the

226Ra

uptake experiments by

gypsum in presence of Sr impurities show that, as previously observed in the Ca-Sr-SO4 system, the DRa, gypsum

is variable (from 0.03 to 0.56, with a mean value of 0.15± 0.09 in the 5-day experimental run, table

2).. The Sr/Ca in gypsum remains lower than 1 molar %. However, for similar initial chemical conditions, we note that the variability of DRa,

gypsum

values is reduced after 5 days of equilibration

compared to 24 hours of equilibration. For each of these co-precipitation experiments, partition coefficients Ra, gypsum and sr, gypsum related to the heterogeneous model of precipitation, were also calculated. They are close to the relative homogeneous distribution coefficients (see table 2 and table SI-2). The data were then interpreted with the homogeneous model in the following.

14


Page 15 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2: Summary of experimental results obtained for 226Ra incorporation in gypsum in presence of Sr impurities (co-precipitation experiments), for 1 day (1-6) and 5 days (1a-6a) of interaction. S.I.: saturation index. Run

[Ca]aq ini.

[Ca]aq. end

Casolid

[226Ra]aq.ini.

[226Ra]aq.end

226Ra solid

(mmol/L)

(mmol/L)

(mmol)

(Bq/mL)

(Bq/mL)

(mmol)

S.I.

DRa,

Ra,

gypsum

gypsum

1

55

33.02

0.65

317.9

230.4

2.1e-10

1.35

0.56

0.63

2

56

21.6

0.68

314.0

307.5

1.6e-11

1.35

0.01

0.02

3

167

100.41

1.33

312.3

238.0

1.80e-10

1.46

0.47

0.53

4

165.5

93.19

1.44

303.3

280.1

5.7e-11

1.46

0.10

0.14

5

275

164.8

2.21

311.5

286.2

6.2e-11

1.40

0.13

0.16

6

826

592.4

4.65

307.6

279.2

6.9e-11

1.18

0.26

0.29

1a

55.5

18.7

0.73

308.8

261.7

1.1e-10

1.35

0.09

0.15

2a

58.5

17.1

0.82

309.3

262.3

1.1e-10

1.36

0.07

0.13

3a

166

104.7

1.21

310.2

268.3

1.0e-10

1.46

0.27

0.31

4a

165.5

104.3

1.21

311.9

272.2

9.7e-11

1.46

0.25

0.29

5a

273.5

118.2

3.10

313.3

299.8

5.6e-11

1.40

0.03

0.05

6a

825

536.6

5.74

317.5

290.8

6.5e-11

1.18

0.17

0.20

4.

DISCUSSION

4.1.

Behavior of Sr and 226Ra as trace in (Ca,Tr)SO4

The experimental results are reported on a diagram showing the evolution of 226Ra solution activity with time. They indicate no decrease of this activity over 200 days (Fig. 2). From the definition of the distribution coefficient according to the homogeneous model of solid solution, it is possible to express the final concentration of Ra in solution, [226Ra]final, as a function of D:

226

[ Ra]final =

( (

[226Ra]

ini nCa

× V

D226Ra,gypsum [Ca]final

+

)

(7)

)

V nCa

Where [ ]ini and [ ]final (Bq/L) represent the initial and final aqueous concentrations, respectively; V (L) is the solution volume and nCa (mol), the quantity of Ca incorporated in the solid.

15



Page 16 of 32

The experimental results are compared with simulations obtained for a the two values of DRa, gypsum of 0.03 and 0.3, indicating that the experimental DRa, gypsum value is more likely to be lower than 0.03. Other sorption/desorption and co-precipitation experimental results have shown no evolution of 226Ra activity with time neither in the solution nor in the solid, suggesting that these two other mechanisms are not more efficient than dissolution and recrystallization in reducing

226Ra

mobility by interactions with

gypsum. Several arguments can be proposed to explain this absence of interactions between 226Ra and gypsum: the two main criteria governing element substitution are the ion charge and the ionic radius. Whereas calcium and radium have the same ion charge (+2), the difference between their ionic radii in eightcoordination (Ca coordination number in gypsum) is significantly high: 1.12 Å for Ca2+ and 1.7 Å for Ra2+. Therefore, Ra2+ be too large to substitute Ca2+ in the gypsum crystal lattice. In the case of binary (Me-Tr) solid solutions, equilibrium is defined by two mass action equations, one for each end-member of the solid solution. In the case of Me-Tr-SO4, we have the two following equilibria:

(Me).(SO4) = Ks, MeSO4.aMeSO4 = Ks, MeSO4.fMeSO4.(1-xTr)

(8)

(Tr).(SO4) = Ks, TrSO4 . aTrSO4= Ks, TrSO4 .fTrSO4 .xTr

(9)

Where Ks,MeSO4 and Ks,TrSO4 are the solubility products of the two end-members, respectively, and fMeSO4 and fTrSO4 are the activity coefficients of Me and Tr, respectively in the solid solution, and ( ) the activities of free aqueous ions in solution. xTr is defined as the molar fraction of TrSO4 in the solid solution. By combining the two equations (8) and (9), it is admitted in the case of isomorphous substitution mechanisms in solid solution, at equilibrium, that the theoretical values of the Tr partition coefficient in a host mineral can be expressed as:

( )solid K = mineral = ( )aqueous K [Tr]

DTr,

[Me]

[Tr]

s,MeSO4 γTr fMeSO4

(10)

s, TrSO4 γMe fTrSO4

[Me]

16


Page 17 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


This relationship lies on the fact that the two end-members crystallize in the same system. In this case, Tr better incorporates into the MeSO4 host solid when the TrSO4 end-member solubility is lower than the one of the host phase. The two respective end-members RaSO4 and SrSO4 crystallize in orthorhombic system whereas gypsum crystallizes in the monoclinic system. When the two end-members do not crystallize in the same structural group, it is always possible to define a hypothetical end-member such as RaSO4,2H2O or SrSO4,2H2O in the same crystallographic system than the one of gypsum. In this case, their thermodynamic properties are unknown, but may be determined by molecular simulations25 or by linear free energy relationship for isostructural minerals and aqueous ions33-34. In the case of dihydrate sulfate minerals, in our knowledge, lack of data invalidates this last approach. However we can assume that the solubility of the RaSO4,2H2O or SrSO4,2H2O end-members would also be very high (such as that of gypsum) and therefore would not favor incorporation of Ra and Sr in gypsum, as was experimentally confirmed in the present work. In the case of the Sr incorporation in gypsum, the measured distribution coefficients, comprised between 0.1 and 0.5 (see section 3.2), are very low compared to a theoretical value calculated from the solubility products ratio (i.e. around 120), implying that solid solution highly departs from ideality. Incorporation of different cations (Sr, Mg, Na, K) in gypsum was shown to occur through interstitial positions created by the hydration water molecules33. However according to XRD pattern obtained on Sr-doped (around 5% molar) biogenic gypsum2635, a significant shift to lower diffraction angle was observed when Sr concentrations increased, indicating that Sr ions are incorporated into the crystal lattice structure of gypsum crystals. In the present study, we did not observe any XRD peak shift, maybe due to the very low amount of Sr incorporation (less of 1%). It remains that, under our experimental conditions, we can relate without ambiguity the incorporation of 226Ra

in gypsum to the presence of Sr with a partition coefficient DRa, gypsum around 0.15±0.09, likely

involving a cation exchange process. In these experiments the Sr contents are ranging from 750 to 14650 ppm. Note also that for lower Sr content (around 270 ppm) as observed in natural gypsum we did not relate any incorporation of 226Ra.

17



Fig. 4. Comparison of DRa, gypsum in literature with values obtained in this study. Note that Sr impurities concentrations are comprised between 2000 and 20000 ppm in this study, against 70 ppm in pure gypsum and 270 ppm in natural gypsum. Data considered from the literature is: Somot et al.13, from sequential leaching of tailings; Gnanapragasam and Lewis20, co-precipitation of gypsum with [Ca]final from 0.04 to 0.06 mol/L in presence of seeds (gypsum/or hemihydrate/or anhydrite) in contact with 226Ra-bearing acid solution oversaturated with respect to gypsum, during 4, 6 or 8 weeks, S/W = 0.2 g/L; Yoshida et al.21, coprecipitation experiments: [Ca]=[SO4]=0.005 mmol/L; [Ra] =7.2.10-10 and 3.7.10-9 mol/L, during 168 hours; Rosenberg et al.14, estimation of DRa, gypsum from evaporitic ponds.

In conclusion, 226Ra appears to be not significantly incorporated into pure gypsum. A very low value of (8.8±1.5).10-4 has been estimated for the homogeneous distribution coefficient of radium into gypsum, which is much lower than the values reported in the literature and attributed to gypsum incorporation. In the presence of Sr impurities (0.1 to 1 molar %), 226Ra can be quantitatively incorporated into gypsum, leading to distribution coefficient mainly comprised between 0.03 and 0.3 (see table 2). These values are in the same order of magnitude as those previously determined in the literature through co-

18


Page 18 of 32

Page 19 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


precipitation experiments20,21 (fig. 4), but lower than the highest values estimated through sequential leaching13.

4.2.

Behavior of

226Ra

as trace in (Sr,Tr)SO4

As for gypsum, the amount of newly recrystallized celestite, nrecrystallized (moles; equation 5), was calculated to get the molar fraction xRa(ss) of 226RaSO4(s) in the solid according to:

xRa(ss) =

[226Ra]0 - [226Ra]t

(11)

nrecrystallized V

Where [226Ra]0 and [226Ra]t are the Ra concentrations in solution at the beginning of the experiment and at the sampling time t respectively, V being the solution volume. The homogeneous empirical distribution coefficient of 226Ra between celestite and an aqueous solution is given by the following equation:

DRa, celestite = (

xRa(ss) [Sr]

(12)

1 - xRa(ss)) [226Ra]t

The DRa, celestite calculated are very high, and at the first stage of the experiments, they happen to be close to the thermodynamic value, given by the equation (13)

DRa,

Ks,SrSO4 γRa fSrSO4

celestite

(13)

= Ks,RaSO4 γSr fRaSO4

with log Ks,SrSO4 = -6.63 and log Ks,RaSO4 = -10.26. In the case of ideal solid-solution, the activity coefficients in the solid-solution, fSrSO4 andfRaSO4, equal to unity. According to this formula, DRa, celestite would take a high value of 4266. This indicates that, while a pure exchange mechanism can occur between

226Ra

and Sr in the crystal lattice in the first steps of

interactions between 226Ra and celestite, the further processes are more complex and probably involve

19



Page 20 of 32

structural modifications of the crystals while integrating 226Ra in the structure. It is then no more an ideal case, and the modelling equation must be adapted. As defined in equation (13), the distribution coefficient is not constant and depends upon both the compositions of aqueous solution and solid solution. Under conditions close to equilibrium as in the case of recrystallization experiments, the D variation is mainly due to the evolution of the solid solution composition and is a function of the rate of Tr-Me substitution. The activity coefficients of RaSO4 and SrSO4, fRaSO4 and fSrSO4 are calculated by applying the Henry and Raoult’s law respectively, in the case of regular solid solution (see supplementary information). This calculation is valid here considering the large dilution of Ra in celestite, with xRa< 0.001. Hence, lnfRa is approximated to a constant a0, and lnfSr is approximately 0. In the present case, the non-ideality parameter, a0, has been graphically determined at 3.7 (see figure SI3), and is close to the value calculated with GM Selektor36 for the solid solution Sr-Ba-SO4. ao can be related to the Margules parameter according to:

W = R T a0

(14)

where R is the perfect gas constant (8.314662 J.mol-1.K-1) and T, the temperature in K. This value led to an estimation of the Margules parameter W of 8.7 kJ/mol, close to the value estimated through the semiempirical correlation established by Zhu18 (i.e. 8.1 kJ/mol). Consequently, and considering equal the activity coefficients Ra and Sr in solution, the equilibrium distribution coefficient DRa, celestite is estimated according to the following relation12:

(

a0 = ln

Ks,SrSO4

)

(15)

Ks,RaSO4 DRa, celestite

A value of DRa, celestite of 105 was found, for xRa(ss)~0.001, which is significantly high, but lower than the values from the literature, i.e. 280 from Goldschmidt’s study22, and 228 from Zhu’s study18.

20


Page 21 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


However, the recent studies25,26 describing the

226Ra

incorporation in ternary systems such as (Sr,Ba,

Ra)SO4 through molecular simulation of defect formation energies arrived at very different value of W(Sr, Ra) (i.e. 17.7 kJ/mol). Other values of interaction parameter were refined also for Ba-Ra (i.e. 2.47 kJ/mol) and for Ba-Sr (i.e. 4.95 kJ/mol). This last value is close to the previous Zhu’s estimation18. Simulation realized by these authors with these new values of W showed that for high Ra concentration, an admixture of RaSO4 causes a phase separation of solid solution into a Ra-Ba-rich phase and a Srphase. Nevertheless, at low Ra concentration, (Ba,Sr)SO4 solid solution is predicted to be completely miscible at ambient temperature, with a maximum of Ra uptake for a molar concentration of Sr of 10% in solid. For higher Sr amounts in solid, the Ra uptake decreases following the Sr increase in the solid solution. Experiments of recrystallization of the ternary (Sr,Ba,Ra)SO4 solid solution carried out at ambient temperature25 and at 90°C24 confirm these results. The last study performed at 90°C highlighted that process of recrystallization appears more complex, involving several metastable phases from a Baand Ra rich precipitate to a Sr-rich solid. Though, such experimental conditions were defined to highlight admixture process and therefore are difficult to compare with ours: 226Ra is present at higher concentration than in the present study and the initial solution is not equilibrated with the mineral, meanwhile we equilibrate solid phase with the solution before adding 226Ra. Application of the W(Sr,Ra) value taken from literature25 for pure celestite would lead to a lower distribution coefficient DRa, celestite around 10. Concerning co-precipitation experiments, other DRa, celestite values were also at different ionic strengths23. They are comprised between 43 for less than 10% of initial Sr removal in celestite to below 1 for more than 90% of Sr removal. The high discrepancy observed, for relatively short reaction times (48 hours), between the theoretical value corrected from Margules parameter (i.e. 237) and the measured value was attributed to the kinetic limit for

226Ra

inclusion in celestite. In co-precipitation experiments, the

observed distribution coefficient usually differs from the one established at equilibrium conditions, and principally for high degree of over-saturation. Therefore, it is generally observed that composition of solid solution is enriched in the more soluble component30,34. In conclusion of our experiments, celestite, which is a sulfate mineral with an intermediate solubility product between gypsum and barite ones, can be considered as a good host for 226Ra. The 226Ra-celestite

21



Page 22 of 32

system seems to behave as a non-ideal solid solution. For low xRa(ss) (~ 0.001) generally expected in natural environment, DRa, celestite is estimated at around 105. However, according to the recent studies on the 226Ra incorporation in (Ba,Sr)SO4 solid solution24, it appears difficult to affirm that equilibrium with solid-solution is completely achieved at ambient temperature after 200 days, when more than 3 years at 90°C are necessary to reach a total equilibrium. This study illustrated also that presence of impurities in mineral sulfate can drastically impact the behavior of

226Ra.

This is the case for gypsum with Sr impurities. Other sulfates can also incorporate

impurities such as very common Ba-Sr-SO4 solid solution. Through experiments of 226Ra incorporation in celestite, in presence of different amounts of Ba impurities (not shown here) we verified that the distribution coefficients of 226Ra in celestite are not influenced by Ba impurities (~up to 1mol% of Ba in celestite). Conversely, in natural brines it has been observed that even in presence of Sr, if baryum is present, barite being more insoluble than celestite, barite remains the principal reservoir of 226Ra4.

4.3. Retention of226Ra governed by solid solution and solubility

The overall retention of 226Ra in sulfate bearing minerals is driven on the one hand by its propensity to co-precipitate as previously discussed, and on the other hand, by the solubility product of the hosting mineral. This can be illustrated by the following comparison in a simplified case. When only a MeSO4(s) solid controls the solubility of Me2+ and SO42-, then their concentrations are equal and [Me]aq ~ (Ks,MeSO4)0.5 according to equation (8).The equation (16) can be derived from the equation (10):

[226Ra]aq. = (Ks,MeSO4)0.5 / DRa, MeSO4x ( [226Ra]MeSO4 / [Me]MeSO4 )

(16)

Figure 5 shows a comparison of the aqueous 226Ra concentration at equilibrium plotted as a function of (Ks,MeSO4)0.5 / DRa, MeSO4, for different solid phases. The calculation were performed considering 1 kg of MeSO4(s), a total volume of 1 L and [226Ra]solid = 25 Bq/g. Such specific activity is typically found in mining environment6, especially in U mill tailings5,13. The variation of the ratio [226Ra]MeSO4 / [Me]MeSO4 (in Bq per mol) is very low as can be observed from the linearity of the data dots.

22


Page 23 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


According to this representation, celestite, barite and anglesite may be considered as efficient sulfate sinks for gypsum=

226Ra.

On the contrary, gypsum with a high solubility and a low partition coefficient (DRa,

8.8.10-4) is not a relevant sink for 226Ra, while it becomes slightly more efficient when impurities

(Sr, Ba) are encountered (i.e. with a partition coefficient of DRa, gypsum= 0.15 for Sr impurities). One can notice that despite a relatively low partition coefficient for barite (DRa,BaSO4 = 0.72), its very low solubility and quasi ubiquitous occurrence in natural environment makes it an efficient trap of

226Ra.

Celestite gathers very high retention properties although it is less observed in natural environment than barite. Significant retention of 226Ra in celestite is still predicted when considering lower DRa, celestite from the litterature25. Such representation may be relevant to evaluate the 226Ra retention capacity of the different SO4 bearing minerals considering the solubility of the MeSO4 solid and also the distribution coefficient. It may be used in the framework of mining water treatment and extended to radioactive effluents treatment. However, rigorously, this selectivity is also function of the solid/solution ratio and the whole chemical reactions governing the stability of the natural system. Indeed, the considered geochemical system in this case is oversimplified and far from environmental conditions where other water/rock equilibria and chemical species are to be taken into account. Such calculations may be achieved with geochemical tools as illustrated in the next section.

23



1E+05

Gypsum (CaSO4.2H2O) 1E+04

[226Ra]aq. (Bq.L-1)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 32

1E+03

Gypsum with impurities (CaSO4.2H2O) 1E+02 1E+01 1E+00 1E-01

Barite (BaSO4)

1E-02 1E-6

Celestite-2 (SrSO4) Anglesite (PbSO4)

Celestite (SrSO4)

1E-5

1E-4

0,001

0,01

0,1

1

10

(Ks, MeSO4)0,5/DRa, MeSO4 Fig. 5. Overall retention of 226Ra in different SO4-bearing minerals. 226Ra aqueous concentration at equilibrium with the hosting phase is a function of the mineral solubility (Ks, MeSO4) and the partition coefficient (DRa, MeSO4). To extend the comparison, SrSO4 is also plotted with a DRa, celestine = 10 in agreement with Vinograd et al.25 (named Celestite-2). Additional data for anglesite (PbSO4) is also reported (data taken from [17]). Calculation were performed for 1 kg of MeSO4(s), a total volume of 1 L and [226Ra]solid = 25 Bq/g.

4.4. Application to uranium mill tailings repository

The experimental data describing the interaction of 226Ra with gypsum, celestite (this study) and barite19 were used to explain the fate of 226Ra in the uranium mill tailings of Bellezane (France). These tailings were already studied and described elsewhere37,38. The chemical composition of the solid matrix and the interstitial waters are reported in the supplementary information (Table SI-4). Contrarily to the previous application example provided in this study, the water chemistry composition is relatively complex with other cations and anions. Mineralogical characterizations confirmed the presence of gypsum (about 9 % of the dry mass) and barite in trace, in addition to minerals inherited

24


Page 25 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


from the ore substratum. The average 226Ra content is 25 Bq/g of dry sediment and [226Ra] analyzed in pore waters is 0.59 Bq/L. Ca and SO4 porewaters concentrations are in equilibrium with the solubility of gypsum. Assuming that all the 226Ra is contained in the (Me,Ra)SO4 solid solution, calculations were performed to identify the hosting MeSO4 mineral (Me = Ba, Sr and Ca) able to reproduce the

226Ra

activities

observed in the tailings. The [226Ra] concentration in the pore water at equilibrium was then calculated using the partition coefficient DRa, MeSO4 and compared to the chemical analysis (see Table SI-5). As shown in Fig. 6, [226Ra]aq measured in the interstitial water (0.59 Bq/L) is better modelled by the (Ba,Ra)SO4 solid solution while considering the solubility with gypsum (0.16 Bq/L). Other (Me,Ra)SO4 solid solutions cannot reproduce such low 226Ra aqueous concentrations (from 15.9 to 546857 Bq/L for (Sr,Ra)SO4 and (Ca,Ra)SO4.2H2O, respectively). Additional calculations introducing thermodynamic data related to (Ba,Ra)SO4 and (Sr,Ra)SO4 solid solution (Ks,RaSO4, Ks,MeSO4 and a0) were achieved using previous geochemical model describing the porewaters30. Modelling were performed with PhreeqC for windows (V2.18.3-5570)39 and llnl.dat 2010-02-09 database. Similar results were found (0.17 and 5.8 Bq/L respectively for (Ba,Ra)SO4 and (Sr,Ra)SO4). These calculations were not possible for gypsum as no a0 value is available. Although gypsum is found in large quantities in the tailings, the partition coefficient is too small to identify this mineral as an efficient trap for 226Ra. Nevertheless, the coexistence of gypsum and (Ba,Ra)SO4, even in small amount, seems to maintain the 226Ra concentration at a low level in interstitial waters, and prevent from undesirable 226Ra migration.

25



107 (Ca,Ra)SO4.2H2O

106

546857

105

[226Ra]aq. (Bq/L)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 32

104

(Ca,Ra)SO4.2H2O + impurities

103

3281

2

10

(Sr,Ra)SO4

101 15.9

100 10-1

0.59 Bq/L

(Ba,Ra)SO4 0.16

10-2

(Me,Ra)SO4

Fig. 6. Modelling the 226Ra aqueous concentrations within the porewaters for Bellezane taillings when considering equilibrium with different (Me,Ra)SO4 solid solution (Me = Ba, Sr, Ca). Experimental aqueous concentration was determined at 0.59 Bq/L.

ACKNOWLEDGEMENTS This research was co-funded by Orano Mining R&D and CEA.

26


Page 27 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References (1) Beaucaire, C.; Toulhoat, P. Redox chemistry of uranium and iron, radium geochemistry, and uranium isotopes in the groundwaters of the Lodeve Basin, Massif Central, France. Applied Geochemistry 1987, 2, 417-426. (2) Beaucaire, C.; Criaud, A.; Michard, G. Contrôle des concentrations de certains éléments trace (As, Sb, Ge, U, Ra, Ba) dans les eaux du Cézallier (Massif Central, France). Chemical Geology 1987, 63, 85-99. (3) Gazineu, M.H.P.; Hazin, C.A. Radium and potassium-40 in solid wastes from the oil industry. Applied Radiation and Isotopes 2008, 66, 90-94. (4) Rosenberg, Y.O.; Metz, V.; Ganor, J. Radium removal in a large scale evaporitic system; Geochim. Cosmochim. Acta 2013, 103, 121-137. (5) Dejeant, A.; Bourva, L.; Sia, R.; Galoisy, L.; Calas, G.; Phrommavanh, V.; Descostes, M. Field analyses of

238U

and

226Ra

in two uranium mill tailings piles from Niger using

portable HPGe detector. J. Environ. Radioact. 2014, 137, 105-112. (6) Leermakers, M.; Phrommavanh, V.; Drozdzak, J.; Gao, Y.; Nos, J.; Descostes, M. DGT as a useful monitoring tool for radionuclides and trace metals in environments impacted by uranium mining: Case study of the Sagnes wetland in France. Chemosphere 2016, 155, 142-151. (7) Ames, L.L.; McGarrah, J.E.; Walker, B.A. Sorption of trace constituents from aqueous solutions onto secondary minerals. 2. Radium. Clays and Clay Min. 1983, 31(5), 335342. (8) Renoso-Maset, E.; Ly, J. Study of uranium(VI) and radium(II) sorption at trace level on kaolinite using a multisite ion exchange model. J. Environ. Radioact. 2016, 157, 136– 148.

27



(9) Jones, M.J.; Butchins, L.J.; Charnock, J.M.; Pattrick, R.A.D.; Small, J.S.; Vaughan, D.J.; Wincott, P.L.; Livens, F.R. Reactions of radium and barium with the surfaces of carbonate minerals. Appl. Geochem. 2011, 26, 1231–1238. (10) Sajih, M.; Bryan, N.D.; Vaughan, D.J.; Descostes, M.; Phrommavanh, V.; Nos, J.; Morris, K. Adsorption of Radium and Barium on Goethite and Ferrihydrite: A kinetic and surface complexation modelling study. Geochim. Cosmochim. Acta 2014, 146, 150-163. (11) Ames, L.L.; McGarrah, J.E.; Walker, B.A.; Salter, P.F. Uranium and radium sorption on amorphous ferric oxyhydroxide. Chem. Geol. 1983, 40(1–2), 135-148. (12) Curti, E.; Fujiwara, K.; Iijima, K.; Tits, J.; Cuesta,C.; Kitamura, A.; Glaus, M.A.; Müller, W. Radium uptake during barite recrystallization at 23 +/- 2 degrees C as a function of solution composition: An experimental Ba-133 and Ra-226 tracer study. Geochim. Cosmochim. Acta 2010, 74(12), 3553-3570. (13) Somot, S.; Pagel, M.; Thiry, J. Speciation of radium in the uranium mill tailings of Escarpière (Vendée-France). C.R. Académie des Sciences/Earth and Planet. Sci. 1997, 352, 111-116. (14) Rosenberg, Y.O., Metz, V., Ganor, J. Co-precipitation of radium in high ionic strength systems: 1. Thermodynamic properties of the Na–Ra–Cl–SO4–H2O system – Estimating Pitzer parameters for RaCl2. Geochim. Cosmochim. Acta 2011, 75(19), 5389-5402. (15) Curie, M. Radium and the new concepts in chemistry. Nobel Lecture 1910. (16) Doerner, H.A.; Hoskins, W.M.M. Co-precipitation of radium and barium sulfates. J. Amer. Chem. Soc. 1925, 47, 662-675. (17) Langmuir, D.; Riese, A.C. The thermodynamic properties of radium. Geochim. Cosmochim. Acta 1985, 49(7), 1593-1601. (18) Zhu, C. Coprecipitation in the barite isostructural family: 1- Binary mixing properties, Geochim. Cosmochim. Acta 2004, 68(16), 3327-3337.

28


Page 28 of 32

Page 29 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(19) Brandt, F.; Curti, E.; Klinkenberg, M.; Rozov, K.; Bosbach, D. Replacement of barite by a (Ba,Ra)SO4 solid solution at close-to-equilibrium conditions: A combined experimental and theoretical study. Geochim. Cosmochim. Acta 2015, 155, 1-15. (20) Gnanapragasam, E.K.; Lewis, B.A.G. Elastic strain energy and the distribution coefficient of radium in solid solutions with calcium salts. Geochim. Cosmochim. Acta 1995, 59(24), 5103-5111. (21) Yoshida, Y.; Nakazawa, T.; Yoshikawa, H.; Nakanishi, T. Partition coefficient of Ra in gypsum. J. Radioanal. Nucl. Chem. 2009, 280(3), 541-545. (22) Goldschmidt, B. Etude du fractionnement par cristallisation mixte à l'aide des radioéléments. Ph.D. Dissertation,Faculté des Sciences de l'Université de Paris, France, 1940. (23) Zhang, T.; Gregory, K.; Hammack, R.W.; Vidic, R.D. Co-precipitation of radium with barium and strontium sulfate and its impact on the fate of radium during treatment of produced water from unconventional gas extraction. Environ. Sci. Technol. 2014, 48, 4596-4603. (24) Klinkenberg, M.; Weber, J.; Barthel, J.; Vinograd, V.; Poonoosamy, J.; Kruth, M.; Bosbach, D.; Brandt, F. The solid solution–aqueous solution system (Sr,Ba,Ra)SO4+H2O: A combined experimental and theoretical study of phase equilibria at Sr-rich compositions. Chemical Geology 2018, 497, 1-17. (25) Vinograd, V.L.; Kulik, D.A.; Brandt, F.; Klinkenberg, M.; Weber, J.; Winkler, B.; Bosbach, D. Thermodynamics of the solid solution - Aqueous solution system (Ba,Sr,Ra)SO4+H2O: I. The effect of strontium content on radium uptake by barite. Applied Geochemistry 2018, 89, 59-74. (26) Vinograd, V.L.; Kulik, D.A.; Brandt, F.; Klinkenberg, M.; Weber, J.; Winkler, B.; Bosbach, D. Thermodynamics of the solid solution - Aqueous solution system

29



(Ba,Sr,Ra)SO4+H2O: II. Radium retention in barite-type minerals at elevated temperatures. Applied Geochemistry 2018, 93, 190-208. (27) Lorens, R.B. Sr, Cd, Mn and Co distribution coefficients in calcite as a function of calcite precipitation rate. Geochim. Cosmochim. Acta 1981, 45(4), 553-561. (28) Treserio, A.J. and Pankow, J.F. Solid solution partitioning of Sr2+, Ba2+ and Cd2+ to calcite. Geochim. Cosmochim. Acta 1996, 60(6), 1053-1063. (29) Prieto, M.; Astilleros, J.M.; Fernandez-Diaz, L. Environmental remediation by crystallisation of solid solutions. Elements 2013, 9, 195-201. (30) Prieto, M. Thermodynamics of solid solution-aqueous solution systems. Reviews in Mineralogy and Geochemistry 2009, 70, 47-85. (31) Lestini, L.; Beaucaire, C.; Vercouter, T.; Descostes, M. Radium uptake by recrystallized gypsum: an incorporation study. Proc. Earth and Planet. Sci. 2014, 7, 479-482. (32) Kushnir, J. The coprecipitation of strontium, magnesium, sodium, potassium and chloride ions with gypsum. An experimental study. Geochim. Cosmochim. Acta 1980, 44(10), 1471-1482. (33) Sverjensky, D.A.; Molling, P.A. A linear free energy relationship for crystalline solids and aqueous solution. Nature 1992, 356, 231-234. (34) Prieto, M.; Heberling, F.; Rodriguez-Galan, R.M.; Brandt, F. Crystallization behavior of solid solutions from aqueous solutions: an environmental perspective. Progress in Crystal Growth and Characterization of Materials 2016, 62, 29-68. (35) Pouria, A.; Bandegani, H.; Pourbaghi-Masouleh, M.; Hesaraki, S.; Alizadeh, M. Physicochemical properties and cellular responses of strontium-doped gypsum biomaterials. Bioinorg. Chem. Appl. 2012, 2012, 1-9.

30


Page 30 of 32

Page 31 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(36) Kulick, D.A. Geochemical thermodynamic modelling of ion partitioning. In Ion partitioning in ambient temperature aqueous systems, Prieto, M., Stoll, H., Eds.; Mineralogical Society of Great Britain & Ireland: 2010, pp 420. (37) Nos, J.; Boizard, A.; Peiffert, C.; Phrommavanh, V.; Cathelineau, M.; Descostes, M. Geochemical characterization of uranium mill tailings. Mineralogical Magazine 2013, 77(5), 1863. (38) Ballini, M.; Nos, J.; Phrommavanh, V.; Descostes, M. U and 226Ra mobility in the uranium mill tailings of Bellezane (France). Goldschmidt Conference 2017, Paris, France. (39) Parkhurst, D.L; Appelo, C.A.J. User’s guide to PHREEQC (version 2) – A computer program for speciation, batch-reaction, one dimensional transport, and inverse geochemical calculations. US Geological Survey Water-Resource Investigation Report 99-4259: 1999, pp 312. (40) Bruno, J.; Bosbach, D.; Kulick, D.A.; Navrosky, A. Chemical thermodynamics of solid solutions of interest in nuclear waste management (a state-of-the-art report). In Chemical Thermodynamics of Solid Solutions of Interest in Radioactive Waste Management, Mompean, F.J., Illemassene, M., Perrone, J., Eds; OECD-Nuclear Energy Agency: Issyles-Moulineaux (France):2007, pp 292.

31



Supporting information 

Experimental conditions of co-precipitation experiments (Table S.I-1)



Experimental results for Sr incorporation in gypsum by co-precipitation (Table SI-2)



Experimental data for 226Ra incorporation by recrystallization in celestite at equilibrium (Table SI-3)



Evolution of 85Sr activity normalized to the initial activity (A0) with time in solution at equilibrium with celestite (Fig.SI-1)



Evolution of DSr, gypsum with Sr/Ca molar ratio in gypsum (Fig. SI-2)



Description of the method for the calculation of non-ideality parameters



Modelling of

226Ra

in the Bellezane (France) mill tailings repository : Geochemical

composition of solid and porewaters, and selected partition coefficients used in the modelling

32


Page 32 of 32

ROLE OF TRACE ELEMENTS IN THE 226-RADIUM

Recommend Documents