EXTERNAL REPORT SCK•CEN-ER-120 10/PDC/P-9

Behaviour of Selenium in Boom Clay

Pierre De Cannière1, André Maes2, Steve Williams3, Christophe Bruggeman2, Thomas Beauwens1, Norbert Maes1, and Mark Cowper3

1 SCK•CEN 2 KULeuven 3 AEA Technology

SCK•CEN ref: CO 90 01 1467.01 1467 RP.W&D.037 NIROND ref: CCHO2004/00/00 DS251-A44/2.1

May, 2010

SCK•CEN RDD Boeretang 200 BE-2400 Mol Belgium

EXTERNAL REPORT OF THE BELGIAN NUCLEAR RESEARCH CENTRE SCK•CEN-ER-120 10/PDC/P-9

Behaviour of Selenium in Boom Clay

Pierre De Cannière1, André Maes2, Steve Williams3, Christophe Bruggeman2, Thomas Beauwens1, Norbert Maes1, and Mark Cowper3

1 SCK•CEN 2 KULeuven 3 AEA Technology

SCK•CEN ref: CO 90 01 1467.01 1467 RP.W&D.037 NIROND ref: CCHO2004/00/00 DS251- A44/2.1

May, 2010 Status: Unclassified ISSN 1782-2335

SCK•CEN Boeretang 200 BE-2400 Mol Belgium

This report can be cited as follows:

De Cannière P., Maes A., Williams S., Bruggeman C., Beauwens T., Maes N., and Cowper M. (2010) Behaviour of selenium in Boom Clay. Work performed under contract: SCK•CEN ref: CO 90 01 1467.01 1467 RP.W&D.037 – NIROND ref: CCHO2004/00/00 DS251-A44/2.1. External Report of the Belgian Nuclear Research Centre, SCK•CEN-ER-120. May 2010. PDF file available at: http://publications.sckcen.be/dspace/ http://publications.sckcen.be/dspace/simple-search

© SCK•CEN Studiecentrum voor Kernenergie Centre d’étude de l’énergie Nucléaire Boeretang 200 BE-2400 Mol Belgium

Phone +32 14 33 21 11 Fax +32 14 31 50 21 http://www.sckcen.be

Contact: Knowledge Centre [email protected]

RESTRICTED

All property rights and copyright are reserved. Any communication or reproduction of this document, and any communication or use of its content without explicit authorization is prohibited. Any infringement to this rule is illegal and entitles to claim damages from the infringer, without prejudice to any other right in case of granting a patent or registration in the field of intellectual property. SCK•CEN, Studiecentrum voor Kernenergie/Centre d'Etude de l'Energie Nucléaire Stichting van Openbaar Nut – Fondation d'Utilité Publique - Foundation of Public Utility Registered Office: Avenue Herrmann Debroux 40 – BE-1160 BRUSSEL Operational Office: Boeretang 200 – BE-2400 MOL

5/328

6/328

Table of Contents

Executive Summary ...... 11 Foreword ...... 11 Summary ...... 11 Redox disequilibrium...... 12 Unknown initial speciation of 79Se in the source term...... 13 Migration parameters selected for 79Se...... 13 Abstract...... 17

1. Introduction...... 21 1.1. Background...... 21 1.2. Objective...... 22 1.3. Main uncertainties and research strategy ...... 22 1.3.1 Complexity and uncertainties affecting the experiments with selenium ...... 23 1.3.2 Objectives and research strategy ...... 24 2. Thermodynamic calculations – AEAT ...... 29 2.1. Overview...... 29 2.2. Applied geochemical codes and reference database – AEAT...... 29 2.3. In situ Boom Clay conditions – AEAT...... 29 2.4. Expected speciation of selenium under Boom Clay conditions – AEAT...... 30 2.5. Calculated thermodynamical solubility of selenium under Boom Clay conditions – AEAT...... 32 3. Geochemical behaviour of selenium in Boom Clay ...... 37 3.1. Introduction...... 37 3.2. Sorption of selenium on Boom Clay and its components ...... 37 3.3. Kinetics of reduction of selenite ...... 41 3.4. Solubility of selenium compounds under reducing conditions ...... 42 3.5. Association of selenium with Boom Clay organic matter...... 43 3.6. Conclusions...... 45 4. Determination of migration parameters of selenium in Boom Clay...... 49 4.1. Introduction...... 49 4.2. Experimental...... 50 4.3. Results...... 51 4.4. Discussion...... 54

4.4.1 Pore diffusion coefficient (Dp) ...... 56 4.4.2 Porosity (η) ...... 56 4.4.3 Sorption – retardation (R) ...... 57 4.5. Conclusion ...... 58 4.5.1 Summary of the migration parameters for 79Se in Boom Clay...... 59 5. Summary and conclusions...... 63 5.1. Overview...... 63 5.2. Key uncertainties and abstraction for performance assessment...... 64 5.3. Summary of the transport parameters for the selenium species considered...... 68

7/328

5.4. Recommendations and future works...... 71 6. Acknowledgments ...... 75

7. References...... 79

APPENDICES...... 149

A1 General information on selenium ...... 153 A1.1 Overview...... 153 A1.2 Inventory and isotopic composition...... 153 A1.3 Uncertainties related to the value of selenium-79 half-life and recent changes...... 154 A1.4 Nuclear decay mode of 79Se and 75Se...... 157 A1.5 Selenium inorganic and organic chemistry ...... 157 A1.6 Chemotoxicity of selenium...... 160 A1.7 References...... 162 A2 Natural selenium in the environment and in Boom Clay...... 167 A2.1 Overview...... 167 A2.2 Primary sources of selenium in the earth crust and sediments...... 167 A2.3 Bioconcentration of selenium by a coccolithophorid, Emiliania huxleyi, and correlation selenium/calcium carbonate...... 169 A2.4 Environmental pollutions related to natural and industrial selenium sources...... 171 A2.5 Concentrations of natural selenium in the environment...... 171 A2.6 Concentrations of natural selenium in Boom Clay ...... 173 A2.6.1 Selenium in pyrite extracted from Boom Clay...... 175 A2.6.2 Selenium in Boom Clay water...... 176 A3 Selenium speciation in the source term ...... 181 A3.1 Selenium in spent fuel...... 181 A3.2 Selenium in vitrified high-level waste ...... 183 A3.3 Selenium in bituminised MLW...... 184 A3.4 Dissolution controlled by alpha radiolysis...... 184 A4 Selenium speciation behaviour in Boom Clay...... 187 A4.1 Overview...... 187 A4.2 Very slow reduction kinetics and derived uncertainties affecting the solubility value ...... 187 2– A4.3 Behaviour of selenate: SeO4 ...... 189 2– A4.3.1 Interaction of SeO4 with pyrite ...... 189 2– A4.3.1.1 AEAT: Interaction of SeO4 with pyrite (high concentration)...... 189 2– A4.3.1.2 KULeuven: Interaction of SeO4 with pyrite (low concentration)...... 190 2– A4.3.2 Interaction of SeO4 with Boom Clay – KULeuven...... 190 2– A4.3.3 Kinetics of reduction of SeO4 – KULeuven...... 191 2– A4.3.4 Solubility of SeO4 – KULeuven ...... 191 2– A4.3.5 Conclusions for SeO4 ...... 192 2– A4.4 Behaviour of selenite: SeO3 ...... 193 2– A4.4.1 Interaction of SeO3 with Boom Clay components (pyrite, OM)...... 195 2– A4.4.1.1 AEAT: Interaction of SeO3 with Boom Clay components (pyrite) ...... 196 2– A4.4.1.2 KULeuven: Interaction of SeO3 with Boom Clay components (pyrite, OM)...... 204 2– A4.4.2 Interaction of SeO3 with Boom Clay – KULeuven...... 211

8/328

2– A4.4.3 Kinetic of reduction of SeO3 onto pyrite – KULeuven...... 217 2– A4.4.4 Solubility of SeO3 – KULeuven ...... 218 2– A4.4.5 Conclusions for SeO3 ...... 218 A4.5 Behaviour of elemental Se(s)...... 220 A4.5.1 Interaction of elemental Se(s) with Boom Clay components (pyrite, OM)...... 220 A4.5.1.1 AEAT: Interaction of elemental Se(s) with Boom Clay components (pyrite, OM)...... 220 A4.5.1.2 KULeuven: Interaction of elemental Se(s) with Boom Clay components (pyrite, OM)...... 226 A4.5.2 Interaction of elemental Se(s) with Boom Clay ...... 227 A4.5.2.1 AEAT: Interaction of elemental Se(s) with Boom Clay ...... 227 A4.5.2.2 KULeuven: Interaction of elemental Se(s) with Boom Clay ...... 228 A4.5.3 Kinetic of reduction of elemental Se(s)...... 228 A4.5.4 Solubility of elemental Se(s) – AEAT...... 228 A4.5.5 Conclusions for elemental Se(s)...... 229 A4.6 Behaviour of selenide: HSe– ...... 230 A4.6.1 Interaction of HSe– with Boom Clay components (pyrite, OM) ...... 230 A4.6.1.1 AEAT: Interaction of HSe– with Boom Clay components (pyrite, OM) ...... 230 A4.6.1.2 KULeuven: Interaction of HSe– with Boom Clay components (pyrite, OM) ...... 233 A4.6.2 Interaction of HSe– with Boom Clay...... 233 A4.6.2.1 AEAT: Interaction of HSe– with Boom Clay...... 233 A4.6.2.2 KULeuven: Interaction of HSe– with Boom Clay...... 234 A4.6.3 Solubility of HSe– – AEAT...... 234 A4.6.4 Conclusions for HSe– ...... 235 A5 Immobilisation of selenium in the near-field ...... 239 A5.1 Immobilisation of selenium in cementitious buffer ...... 239 A5.2 Immobilisation of selenium by alteration and corrosion products ...... 239 A5.2.1 Uptake of selenium by spent fuel degradation products...... 239 A5.2.2 Sorption of selenium by iron oxy-hydroxides ...... 240 A6 Selenium background concentration in bentonite buffer materials ...... 243 A6.1 Introduction...... 243 A6.2 Experimental...... 243 A6.3 Results...... 243 A6.4 Comparison of Se concentrations in near-field and far-field conditions...... 245 A6.5 Conclusions...... 246 A6.6 References...... 247 A7 Sorption behaviour of selenite, selenate and sulfate on Fe and Al oxide surfaces...... 251 A7.1 EXAFS studies of selenite and selenate adsorption on goethite ...... 252 A7.1.1 Classical view of inner-sphere and outer-sphere surface complexes...... 252 A7.1.2 Evidence of inner-sphere complexes also implied for weakly sorbing species...... 253 A7.2 ATR-FTIR studies of selenate and sulfate adsorption on Fe and Al (hydr)oxide...... 253 A7.2.1 Sulfate adsorption on Fe oxyhydroxide...... 253 A7.2.2 Selenate and sulfate adsorption on Fe and Al (hydr)oxide...... 254 A7.2.3 Formation of outer-sphere versus inner-sphere surface complexes...... 255 A7.2.4 Effect of the nature of the mineral surface ...... 257 A8 Selenium and organic matter ...... 261

9/328

A9 Selenium migration behaviour in Boom Clay...... 269 A9.1 Overview...... 269 A9.2 Percolation Experiments ...... 270 75 2– A9.2.1 Percolation experiments with SeO3 ...... 270 A9.2.1.1 Percolation tests: experimental ...... 270 A9.2.1.2 Evolution of 75Se concentration in the percolation water...... 271 A9.2.1.3 75Se migration profile in the solid clay...... 273 A9.2.1.4 Modelling of the 75Se profile in the solid clay: questionable attempt ...... 277 A9.2.1.5 Comparison with the results of previous migration experiments made with clay plugs equilibrated with elemental and reduced 75Se...... 278 A9.2.2 Lessons learned during the updating of the Data Collection Forms (DCF’s) in 1999...... 279 A9.2.3 Conclusion of percolation experiments...... 282 A9.2.4 Percolation experiments with dual tracer: FeSe contacted with 14C-OM...... 283 A9.3 Electromigration experiments...... 286 A9.3.1 Electromigration: experimental setup...... 286 2– 2– A9.3.2 Electromigration experiments with oxidized Se sources (SeO4 and SeO3 ) (Beauwens et al., 2005) ...... 287 75 A9.3.3 Electromigration experiments with SeO3 – Boom Clay slurries...... 295 A10 Redox disequilibrium and reluctance of sulfate for reduction in deep clay formations ...... 299

A11 Behaviour of redox-sensitive elements in a nitrate plume associated with bituminized MLW – The selenium case study ...... 303

A12 List of Abbreviations...... 307

A13 List of Symbols ...... 323

A14 List of Physical Constants and Units ...... 327

10/328

Executive Summary

Foreword The main objective of the disposal of radioactive waste in deep geological formation is to guarantee the long-term safety by isolating the waste from man and the biosphere and by confining long-lived radionuclides as long as possible to delay and to attenuate their release towards aquifers and the biosphere.

In Belgium, geological disposal is studied since 1974 to assess the long-term safety of a deep repository of high-level waste and spent fuel in a clay formation, and since then, Boom Clay is studied as the reference host formation. Low permeability sedimentary clay formations are presently considered the best geological barriers providing both a physical (limited water flow) and a chemical (radionuclides retention) containment limiting the radionuclide transport. Because of the very low hydraulic conductivity and the small natural hydraulic gradient of the Boom Clay formation, molecular diffusion is considered as the dominant solute-transport mechanism.

This report presents the most up-to-date understanding and data dealing with the selenium transport in the Boom Clay formation.

With the re-estimation of the half-life of 79Se from 65 ka to 295 or 370 ka, the radioactive decay of 79Se in the clay barrier is negligible. 79Se is presently considered as the key mobile fission product for nuclear waste disposal in the Boom Clay. The correct understanding of selenium migration behaviour through the clay barrier is essential to underpin its transport parameters (apparent diffusion coefficient, Dapp; accessible porosity, η; retardation factor, R; and solubility limit, S) selected for the performance assessment calculations of a deep repository for spent fuel and high-level waste.

Summary 0 – FeSe, FeSe2, Se (solid phase) and HSe (aqueous species) are the thermodynamically stable selenium species expected under in situ reducing conditions in undisturbed Boom Clay at depth. HSe– migration should be limited by the low solubility of iron selenide or elemental selenium. Table 1 present the different interactions expected for all the inorganic species under which selenium could be released by the various waste forms to be disposed in Boom Clay.

11/328

Table 1: Matrix of different interactions expected for inorganic selenium species in Boom Clay. Oxidation Possible Aqueous Reduction Sorption OM State Solubility Species in Boom Clay Association Limiting Phase At equilibrium

2– +6 No SeO4 aq. Extremely slow Null None (+ solid solution (redox or very weak in cement) disequilibrium)

2– +4 CaSeO3 SeO3 aq. Easy Medium Association (+ solid solution (inner-sphere (+ reduction) in cement) complex) observed

(0) 2– – 0 Se SeO3 aq. / HSe aq. Slow — Colloid – colloid

– –1 FeSe2 , Fe×Sey HSe aq. — Unknown Not yet –2 FeSe (not considered) observed

A main feature observed during the different studies on the behaviour of selenium in Boom Clay is that its oxy-anionic species may suffer severe redox disequilibrium: indeed, selenate is very reluctant to reduction, while the sorption-reduction-precipitation of selenite is kinetically 2– controlled. For this reason, it is necessary to also consider the non-solubility limited SeO4 as a possible migrating species because of the large uncertainty on the speciation of selenium in the waste form.

To explain the choice of the migration parameters selected for 79Se in the present study, the two major uncertainties affecting the selenium chemical state in a deep disposal system are first summarized hereafter: (1) redox disequilibrium and (2) selenium original speciation in the source term (mainly spent fuel and vitrified HLW but also bituminized MLW).

Redox disequilibrium Under reducing conditions, at ambient temperature and in the absence of catalyst, or bacterial enzymatic activity, selenate is reluctant to reduction. In natural conditions, it is also often the case for the highest valence species of other elements (e.g., sulfate, perchlorate, arsenate, …), because of the multiple electron transfers needed in the reduction reaction. Under in situ conditions prevailing in Boom Clay, the selenate reduction is uncertain due to kinetic limitations. Another reason could be that electron donors are present in insufficient quantity in the system, not enough accessible, too less reactive, or have been consumed by oxidation reactions and water radiolysis (or other repository-induced perturbations).

A well known case of redox disequilibrium is the persistence of sulfate often observed in deep reducing sediments. Under strongly reducing conditions prevailing in Boom Clay, sulfur is

12/328

– expected to be present as HS aqueous species in equilibrium with FeS2 (pyrite) at a solubility -8 -7 –1 2– of about 10 to 10 mol·L . However, SO4 is detected in “undisturbed” Boom Clay porewater at the Mol site at higher concentration, typically of the order of 1 mg·L-1 (10-5 mol·L-1) or more, i.e., two to three orders of magnitude above the expected sulfide concentration. Although this might be due to traces of sulfate produced by pyrite oxidation, it could also reflect the true residual sulfate concentration present in the Boom Clay porewater.

The presence of higher concentration of sulfate in ancient porewater of Opalinus Clay (~ 180 Ma) and Callovo-Oxfordian Clay (~ 155 Ma) is another natural evidence that a fraction of sulfate can resist to reduction over geological time period and that hexavalent sulfur, S(VI), can still be present today and coexist with pyrite under strongly reducing conditions at depth.

Considering the chemical similarities between selenate and sulfate for their recalcitrance to reduction, we have thus to conclude that selenate could also subsist in deep geological formations under reducing conditions.

Unknown initial speciation of 79Se in the source term Presently, we ignore under which chemical form is 79Se present in spent nuclear fuel, vitrified HLW, or bituminized MLW waste (Eurobitum). It could be under a reduced form [Se(0), Se(-II)] as well as an oxidised species [Se(VI)]. Radiolytic effects (producing oxidising free radicals in water) are expected to favour the presence of selenate, while elemental selenium or selenide could be protected by U(4+) (reductant) present in the UO2 matrix of spent nuclear fuel (SF) or by other reductants (e.g., saccharose) added to the glass frit during the vitrification process of HLW. In the case of nitrate-bearing Eurobitum waste, the presence of 79Se under the selenate form can certainly not be ruled out because of the massive amounts of nitrate salts (up to 25 – 30 wt. %) present in this type of waste and the oxidizing conditions of the production process. As a consequence, in the absence of reduction of selenate, it is necessary to also take into account selenate along with the thermodynamically favoured selenide species expected for Boom Clay conditions.

Migration parameters selected for 79Se As conclusion, at least two sets of transport parameters are certainly needed to assess the long-term dose-to-man delivered by 79Se: a first set for selenate (without solubility limit), and a second one for selenide (with solubility limit). For selenate, a consistent set of parameters was derived from migration experiments, while for selenide, only the solubility limit was measured, the other parameters of selenide being taken in line with these of iodide. The parameter values are given in Table 2.

13/328

Table 2: Overview of the migration parameters selected for selenate and selenide in undisturbed Boom Clay. Values relevant for the Mol site under the present geochemical conditions, i.e., in the absence of perturbation. 2– SeO4 (divalent anion) Best Estimate Expert Range Source Range 2 –1 -11 -11 -11 Dapp (m s ) 3.17 (±0.98) × 10 2.2 – 4.2 × 10 1.5 – 7.3 × 10 η (—) 0.10 0.05 – 0.18 0.05 – 0.18 R (—) 2.3 1 – 5 1 – 5 Solubility (mol dm-3) unlimited unlimited unlimited

HSe– (monovalent anion) Best Estimate Expert Range Source Range 2 –1 -10 -10 -10 Dapp (m s ) 1.2 × 10 1.0 – 1.3 × 10 0.8 – 1.7 × 10 η (—) 0.16 0.14 – 0.18 0.12 – 0.18 R (—) 1 1 1 Solubility (mol dm-3) 2 × 10-8 1 × 10-9 – 1 × 10-7 1 × 10-9 – 1 × 10-7 Source range of a parameter is a range of values outside of which the parameter value is unlikely to lie, considering our current knowledge. Expert range of a parameter is the range of values within which experts expect the parameter value to lie.

In the present study only transport and retention of selenium in the Boom Clay formation have been investigated in detail. No retention with the corrosion products of carbon steel (siderite, magnetite, hematite, goethite, ferric hydrous oxide, green rust, …), or with cement hydrated phases in the concrete buffer of the SuperContainer, have been considered. While selenide is poorly soluble in the presence of Fe2+ released by iron corrosion products, in a cementitious buffer, selenite solubility may be limited by CaSeO3 in the presence of portlandite, Ca(OH)2. Selenate and selenite might also be incorporated in ettringite, AFt and AFm phases or be sorbed by calcium silicate hydrate (CSH) and layered double hydroxide (LDH) in the cement paste (Baur 2002; Baur and Johnson, 2003a,b, Bonhoure et al., 2006). These retention mechanisms, limited to the near-field of a deep repository, have been presently neglected and their role is also considered marginal for cemented medium-level waste (MLW).

For nitrate-bearing bituminised waste (MLW), as nitrate first undergoes reduction before sulfate or selenate (Oremland et al., 1999), no selenate reduction should be considered to remain conservative in safety calculations.

It must be noticed that only the mineral species of selenium have been accounted for in the present study. No volatile organic species of selenium, such as dimethyl selenide (DMSe) and dimethyl diselenide (DMDSe), or selenium associated with natural organic matter (NOM), have been presently considered. These organic species are often produced in nature by micro- organisms, planktonic algae and superior living organisms. They are only relevant if the exposure route of man to 79Se should be envisaged via the gas phase, i.e., in case where repositories galleries would be left open for centuries, or more, for the sake of retrievability of waste packages in the frame of a reversible repository. In the present study, the galleries of a

14/328

deep repository are considered immediately backfilled and closed as soon as possible after waste emplacement.

15/328

16/328

Abstract

With the re-estimation of the half-life of 79Se to 295 or 370 ka, the radioactive decay of 79Se in the clay barrier appears to be negligible. 79Se is thus presently considered as the key mobile fission product for nuclear waste disposal in the Boom Clay, the reference host formation presently studied in Belgium. Therefore, a good understanding of selenium migration behaviour through the clay barrier is essential to underpin its transport parameters (apparent diffusion coefficient, Dapp; accessible porosity, η; retardation factor, R; and solubility limit, S) selected for the performance assessment of a deep repository for spent fuel and high-level waste. Under the reducing conditions prevailing in Boom Clay at depth, selenide, [Se(-II)], is the predominant thermodynamically stable chemical form of selenium. HSe– is expected as the main aqueous species and its migration should be limited by the low solubility of iron selenide or elemental selenium. However, selenium may suffer of severe redox disequilibrium and experimental evidences presently suggest that selenate is very reluctant to reduction while a kinetically controlled sorption-reduction-precipitation behaviour is observed for selenite.

Due to the large uncertainty on the speciation of selenium in the waste form, and taking the stability of selenate versus reduction into account, it is also necessary to consider the non- 2– solubility limited SeO4 as a possible migrating species. However, when a lower oxidation state [Se(IV), Se(0), and Se(-II)] is present in the waste form, a solubility limit may also contribute to delay and attenuate the 79Se release from the source term and to spread it on a longer time period. Therefore, two sets of transport parameters are needed to assess the long- term dose-to-man associated to 79Se: a first set for selenate (without solubility limit) and a second set for selenide (with solubility limit). For selenate, a consistent set of parameters was derived from migration experiments, while for selenide, only the solubility limit was measured, the other parameters of selenide being unknown and considered in line with these of iodide.

Migration parameters selected for selenate and selenide in undisturbed Boom Clay. Values relevant for the Mol site under the geochemical conditions prevailing today. Transport Parameter 79SeO 2– H79Se– (Best Estimate, BE) 4

2 –1 -11 -10 Dapp (m s ) 3.17 (±0.98) × 10 1.2 × 10 η (—) 0.10 0.16 (divalent anion) (monovalent anion) R (—) 2.3 1

Solubility (mol dm-3) unlimited 2 × 10-8

17/328

Presently, the main uncertainty for the long-term safety assessment of 79Se clearly resides in the unknown speciation of selenium in the waste form. Indeed, spent fuel and vitrified high- level waste can accommodate both reduced and oxidized forms of selenium but experimental data are lacking.

Keywords: Se-79, selenium-79, selenium, selenate, selenite, selenide, sorption, reduction, precipitation, solubility, Boom Clay, redox-disequilibrium, reactive transport, fission product, diffusion, retention, transport parameters, performance assessment, uncertainty, spent fuel, high-level waste, HLW, radioactive waste disposal.

18/328

1. Introduction

19/328

20/328

1. Introduction

1.1. Background

Selenium-79 is a mobile fission product presently considered as the main contributor to the dose-to-man for the disposal of spent fuel (SF) and high-level waste (HLW) in Boom Clay (Marivoet et al., 1999; Mallants et al., 1999; Sillen and Marivoet, 2000; Marivoet and Weetjens, 2007). Selenium is a redox-sensitive element whose chemistry closely resembles that of sulfur. All its inorganic species are anionic and its more oxidised form (selenate) is very mobile in surface water body and in the near-surface environment (unsaturated soils: vadose zone) where oxygen is present. Only under strictly anoxic and reducing conditions, as those encountered at depth in organic rich shale or clay, selenium mobility of its more reduced forms (elemental selenium and heavy metal selenide) is limited by a low solubility.

Selenium is also a trace mineral that is essential to health but required only in small amounts (ODS–NIH, 2004). Since the years 1970–1980 the behaviour of selenium in the environment is the object of a large number of scientific publications reviewed in now classical text books (Frankenberger and Benson, 1994; Frankenberger and Engberg, 1998; Plant et al., 2004). The interest for selenium simultaneously arises from its high toxicity and from concerns for human health because important selenium deficits are also observed in dietary source in some part of the world. Selenium deficit can cause chronic diseases such as cancer, heart, endocrine and immune diseases, because selenium is a micronutrient indispensable to life. Indeed, selenium is incorporated into proteins to make selenoproteins, which are important antioxidant enzymes (e.g., glutathione peroxidase, Rotruck et al., 1973). The antioxidant properties of selenoproteins help prevent cellular damage from free radicals generated by the oxygen metabolism.

Although extensively studied in mineral and organic chemistry, biochemistry, physiology and medicine, with an overwhelming number of scientific publications, the behaviour of selenium in the environment, particularly in clay under reducing conditions remains complicated and sometimes difficult to predict.

Low but non-negligible amounts of selenium-79 produced by nuclear fission in electrical power plant have to be disposed of in geological repositories. The quantity of 79Se expected per current meter of underground gallery is of the same order of magnitude than for 129I: typically in the range of 0.1 to 1 mol of 79Se per current meter. Specific uncertainties deal also 79 with Se itself. Indeed, the half-life (T½) of selenium-79 has been revised several time in the last decade, varying on nearly two orders of magnitude (between 6.5 × 104 y and 1.1 × 106 y) and is presently estimated as ~ 3 × 105 y. Other uncertainties also exist on the chemical form

21/328

of selenium in the waste forms and on the inventory in vitrified high-level waste (HLW) because the loss of volatile selenium during reprocessing and vitrification operations. For more background information on selenium the reader can refer to Appendices A1–A3.

1.2. Objective

To better understand the processes and the mechanisms controlling the behaviour of selenium-79 in Boom Clay under reducing conditions, ONDRAF/NIRAS passed contract with three laboratories: SCK•CEN, AEA Technology, and KULeuven. The objective of this report is to present an overview of the main results of the experimental and modelling works performed by these three laboratories. The main methods used to understand the retention and diffusion of selenium in Boom Clay were speciation measurements, batch sorption tests, migration experiments, and geochemical modelling. The final aim pursued is to draw geochemical mechanisms and conclusions correctly describing the behaviour of 79Se in Boom Clay and that could be translated in terms of processes, concepts, models, and parameters values usable for the safety studies.

1.3. Main uncertainties and research strategy

Although the general bases of the inorganic chemistry of selenium in simple aqueous solutions (synthetic systems) are well established and that selenium chemistry closely resembles to that of sulfur, the behaviour a selenium in soils and reducing environments can be very complex and remains often difficult to decipher. Hereafter some reasons are mentioned why the study of the behaviour of selenium in Boom Clay under reducing conditions is a technical and scientific challenge.

1. First, the experimental window available to observe very low concentrations of selenium as found in natural waters is narrow and requires low-level detection techniques. The lower and upper working boundaries are respectively imposed by the limit of detection (~ 10-10 mol dm-3) of selenium in solution, and by the low solubility of most of the selenium mineral phases (~ 10-8 mol dm-3) under reducing conditions. The useful working range for sorption and solubility experiments is not wide and it is difficult to guarantee that sorption experiments can always be conducted below the limit of solubility of

reduced selenium forms. This is specially true at low Eh values when the higher valences of selenium are progressively reduced to let the place to less soluble species.

2. Kinetics limitations and redox-disequilibrium states hinder the reduction of the highest valences of selenium, increasing the complexity of observations in the laboratory.

22/328

3. Reduced forms of selenium are also very sensitive to oxidation: so, strictly anaerobic conditions have to be continuously maintained and monitored to prevent undesired perturbations of the studied systems and to guarantee the experiments validity. The nature and the purity of the solid phases supposed to control the selenium solubility are also critical and must be carefully controlled. Aqueous speciation techniques with enough sensitivity are also required to check under what chemical form is selenium present at very low concentration in water.

4. The complex biochemistry of selenium and its occurrence in the skeleton of natural organic molecules makes the studied systems even more complex.

5. And finally, as only abiotic selenium reduction pathways are considered for compact Boom Clay at depth, it is particularly important to avoid in the laboratory studies the development of undesired micro-organisms in the clay suspensions system.

1.3.1 Complexity and uncertainties affecting the experiments with selenium

The speciation of selenium-79 in the solid waste matrix is unknown. Selenium could be released by spent fuel (SF) and vitrified high-level waste (HLW) in a variety of redox states, 2– 2– including selenium oxyanions such as SeO3 or SeO4 . Therefore, experiments were set up in order to:

1. Check whether all the selenium species present are transformed into the thermodynamically stable species in the Boom Clay on short time periods;

2. Identify the thermodynamically stable species and to measure selenium solubility;

3. Assess the appropriate transport parameters (retardation factor, accessible porosity, apparent diffusion coefficient) to be used for performance assessment studies, and;

4. Control that selenium is not interacting with mobile organic matter, thereby increasing the selenium mobility within the clay layer.

The interpretation of data on the transport of selenium in Boom Clay and on the natural selenium background concentration in the clay formation have revealed unexpected uncertainties in the behaviour of selenium under reducing conditions.

Laboratory percolation experiments made during about one year on undisturbed Boom Clay 75 75 cores with a Se source, considered to be primarily composed of Na2 SeO3 first revealed the

23/328

presence of a mobile selenium species moving in the water at a concentration range of 10-9 – 10-8 mol dm-3. Simultaneously, the distribution profile of the selenium activity in the solid clay core was complex and very difficult to interpret.

Electro-migration experiments studying the diffusion of selenium in Boom Clay on shorter time scales (days to weeks), also revealed complex migration patterns in the clay cores and showed the existence of negatively charged species with different mobility.

Natural selenium is present in Boom Clay and appears to be primarily associated with pyrite

(FeS2). The concentration of natural selenium measured in Boom Clay porewater is about 2.4 × 10-8 mol dm-3, a value consistent with the solubility limit of a selenium-bearing solid phase. However, as a fraction of selenium in interstitial water is associated with natural dissolved organic matter (or even part of its molecular structure), the selenium concentration measured in porewater might not correspond to a true solubility.

The three experimental observations presented above are not straightforward to interpret because uncertainties subsist on selenium speciation, retention mechanisms (sorption or precipitation) and redox disequilibrium, a.o.: 1. Lack of data on selenium speciation in the water and in the solid phase: In all of the above mentioned experiments, it is unclear which selenium species is under investigation and whether radio-labelled selenium can interact with the natural Boom Clay organic matter, or not, and what type of association mechanism could be involved. 2. Uncertainty about the chemical retention mechanisms: The main question is that we do not know if the observed selenium concentrations are controlled by sorption or limited by solubility. 3. Uncertainty about kinetics of redox reactions: Many publications in the literature indicate that selenium oxidation/reduction reactions in natural media are not straightforward because of thermodynamical disequilibrium. So, often, several oxidation states may simultaneously coexist and achieving of the thermodynamical equilibrium seems to be kinetically controlled.

1.3.2 Objectives and research strategy

To overcome the afore-mentioned limitations and to reduce the uncertainties, an appropriate conceptual model describing the main processes at work and the corresponding parameters are needed. Therefore, a threefold strategy is elaborated to deliver scientifically sound information for performance assessment on the basis of (i) thermodynamic calculations, (ii)

24/328

laboratory batch interaction tests (sorption, reduction, solubility, association with OM), and (iii) migration experiments.

1. Geochemical modelling calculations are made to predict the speciation of selenium expected under Boom Clay conditions and to estimate its thermodynamic solubility.

2. Laboratory batch experiments are performed by KULeuven and AEA Technology, (details can be found in Appendix A4 (Selenium speciation behaviour in Boom Clay) and in following reports: Baker et al., 1997, 1998, 2000, 2002; Bruggeman et al., 2002, 2004, 2005, 2006, 2007; Cowper et al., 2003; Heath et al., 1997, 2000; Maes A. et al., 2004a,b, 2005) from different starting selenium solutions to achieve final Se equilibrium in conditions aiming to simulate these prevailing in situ in undisturbed Boom Clay. In order to predict the behaviour of selenium in the reducing conditions of Boom Clay, it is necessary to identify the mechanisms of interaction between selenium and the various components of Boom Clay and to understand the more relevant processes. Since selenium is a redox-sensitive element, it is clear that the reactive surface of reducing minerals

naturally present in Boom Clay, such as pyrite (FeS2) and siderite (FeCO3), may play a major role in its chemical behaviour. Pyrite surface is considered as the main reduction site because it is the redox-controlling phase and it might be very reactive towards the redox-sensitive selenium. Since organic molecules are also present as potential complexing and mobilising agents, their influence on the speciation of selenium also needs to be evaluated. To study the solubility of selenium in the reducing conditions of Boom Clay, two different experimental routes are followed from opposite initial oxidation states: on the one hand, starting from supersaturation with soluble oxidized selenium 2– 2– species (SeO3 and SeO4 ), and, on the other hand, starting from undersaturation with poorly soluble reduced elemental Se(s) and Fe(II) selenides. The purpose of using oversaturation is to examine if the solubility values measured under Boom Clay conditions effectively corresponds to the theoretical predictions based on redox equilibria and thermodynamic solubilities of the pertinent selenium solid phases. Conclusions with respect to speciation at equilibrium, kinetics of reduction, solubility and transport are drawn for each case.

3. Migration experiments are carried out to determine the best estimate of migration parameters needed by performance assessment (PA) studies for selenium and to understand the possible chemical-coupled processes (sorption, reduction, and precipitation) at work during its transport (details can be found in Appendix A9 on “Selenium migration behaviour in Boom Clay” and in following reports and papers: De Cannière et al., 1995, 1996; Maes N. et al., 2004e,f; Beauwens et al., 2005).

25/328

This strategy is aimed to globally understand the behaviour of selenium in Boom Clay, and to identify processes from which relevant parameters and associated uncertainties can be derived for PA studies. Also, unresolved key issues about some mechanisms and open questions would be identified to provide new recommendations for PA approach and further studies related to experimental research and modelling works.

26/328

2. Thermodynamic calculations

27/328

28/328

2. Thermodynamic calculations – AEAT

2.1. Overview

Geochemical modelling of inorganic selenium speciation and solubility have been performed for conditions representative of the Boom Clay using PHREEQC and the HATCHES 2– database. Under strongly oxidising conditions selenate (SeO4 ) dominates, under mildly- 2– oxidizing conditions, selenite (SeO3 ) oxidation state is stable, while under strongly reducing conditions, hydrogen selenide (HSe–) is the major aqueous species. Considering fixed pH values, the selenium solubility strongly depends on the redox potential. For elemental selenium, Se(s), the selenium concentration curves as a function of Eh exhibit a very characteristic V-shape valley profile. The Se(s) solubility is predicted to be strongly dependent on redox potential because its oxidation state in the solid, Se(0), is always different to that in solution, Se(-II), Se(IV) or Se(VI). The minimum in the solubility curves occurs where Se(s) is in equilibrium with both selenide and selenite species in solution.

2.2. Applied geochemical codes and reference database – AEAT Calculations made by Serco Assurance as part of the AEA Technology study were performed using the PHREEQC geochemical program version 2.8 developed by Parkhurst and Appelo (2003). Thermodynamic data were selected from the HATCHES database version NEA15 compiled by Bond et al. (1997). The relevant data in HATCHES were compared with these used by SCK•CEN to model selenium in Boom Clay. The SCK•CEN database contains data for selenium selected from the Lawrence Livermore National Laboratory (LLNL) database and data selected from HATCHES; there is no significant difference between the two databases concerning the major aqueous selenium species.

In the meantime, after the calculations presented here, thermodynamic data on selenium have been recently updated by Olin et al. (2005) in the frame of the Thermochemical Database Project (TDB) managed by the Nuclear Energy Agency (NEA). Complementary information on the chemical thermodynamics of compounds and complexes of Se (amongst other elements such as U, Np, Pu, Am, Tc, Ni and Zr) with selected organic ligands have also been newly compiled by Mompean et al. (2005). The present calculation results have thus been made prior to these two new updates simply mentioned her to be complete.

2.3. In situ Boom Clay conditions – AEAT The composition of synthetic Boom Clay water was calculated from the recipe used in the experiments (Dierckx, 1997). The resulting solution was predicted to equilibrate at a pH of 8.1 compared to a measured value of 8.3. Equilibration with a nitrogen atmosphere resulted

29/328

in a pH rise to 9.9, presumed to be due to loss of carbon dioxide. Modelling indicated that this was consistent with a drop in the total concentration of carbonate from 1.4 × 10-2 mol dm-3 to 9.6 × 10-3 mol dm-3. This lower carbonate concentration was applied in all subsequent calculations. A new detailed synthesis on the Boom Clay porewater composition has been compiled later by De Craen et al. (2004), but was not available at the time of the present study on selenium. However, the selenium speciation is not very sensitive to the minor differences in the estimated composition of the porewater.

For the mineralogical composition of Boom Clay, more detailed information is also provided by Baeyens et al. (1985a,b), Merceron et al. (1993, 1994, 1995), Griffault et al. (1996), Beaucaire et al. (2000), and more recently De Craen et al. (2004).

2.4. Expected speciation of selenium under Boom Clay conditions – AEAT The speciation of selenium was calculated for three sets of conditions over the pH range 6 to 11:

(a) pH + pe = 20.0 — strongly-oxidising conditions;

(b) pH + pe = 12.1 — mildly-oxidising (aerobic) conditions;

(c) pH + pe = 0.0 — strongly-reducing conditions.

Under strongly-oxidising conditions selenium is calculated to be present as the uncomplexed selenate ion across the pH range 6 to 11. Under mildly-oxidising to mildly-reducing conditions the selenite oxidation state is stable and the calculated speciation is shown in Figure 2.4.1 (AEAT). Under strongly-reducing conditions, the speciation is dominated by the hydrogen selenide ion, HSe–.

The effect of redox potential was studied in more detail by carrying out further calculations at fixed pH values (8.4 and 9.9) with the redox potential varied over the range -600 mV to

+600 mV vs Standard Hydrogen Electrode (SHE) (all subsequent redox potentials (Eh) are quoted vs SHE). The calculated selenium speciation as a function of Eh at pH 8.4 is shown in Figure 2.4.2 (AEAT). At low redox potentials selenide species dominate. The cross-over to selenite occurs at around -80 mV, and that from selenite to selenate is predicted at around +380 mV. Similar behaviour is expected at pH 9.9 but the oxidation state transitions are calculated to occur at lower Eh values.

30/328

1.2E-10

Total Se 1.0E-10

-3 8.0E-11 2- SeO3

6.0E-11

Concentration/ mol dm 4.0E-11

2.0E-11 - HSeO3

0.0E+00 67891011 pH

Figure 2.4.1 (AEAT): Predicted selenium speciation under mildly-oxidising conditions (pH + pE = 12.1).

1.2E-10

1.0E-10 Total Se

2- SeO3 - 2- HSe SeO4

-3 8.0E-11

6.0E-11

Concentration/ mol dm mol Concentration/ 4.0E-11

2.0E-11

- HSeO3

0.0E+00 -600 -400 -200 0 200 400 600 800 Eh/ mV

Figure 2.4.2 (AEAT): Predicted selenium speciation as a function of Eh at pH = 8.4.

31/328

2.5. Calculated thermodynamical solubility of selenium under Boom Clay conditions – AEAT The calculated solubility of FeSe(s) and elemental Se(s) in contact with synthetic Boom Clay water is shown as a function of pH in Figure 2.5.1 (AEAT). Under mildly-reducing conditions, selenium is predicted to be present as selenite in solution and the solubility curve for FeSe(s) shows a minimum at around pH 10. For elemental Se(s), the calculated solubility is strongly dependent on the redox conditions. The lowest solubility is predicted for Se(s) in equilibrium with selenide under mildly-reducing conditions. However, the solubility is calculated to increase greatly under strongly-reducing conditions as selenide is favoured with respect to Se(s). Similarly, as the redox potential is raised the solubility is enhanced due to the increasing stability of selenite (V-shape curve).

1.0E-03

1.0E-04 Se(s)/ selenide, strongly reducing

1.0E-05

1.0E-06 FeSe(s)/ selenide -3

1.0E-07

1.0E-08

1.0E-09

1.0E-10 Se(s)/ selenide, mildly reducing

1.0E-11 Se(s)/ selenite Total Se Concentration/ mol mol dm Concentration/ Se Total 1.0E-12

1.0E-13 Total Se with FeSe added. pH + pE = 0 Total Se with elemental Se added. pH + pE = 0 1.0E-14 Total Se with elemental Se added. pH + pE = 6.5 Total Se with elemental Se added. pH + pE = 8.4

1.0E-15 67891011 pH Figure 2.5.1 (AEAT): Total selenium concentration vs pH for synthetic Boom Clay water (SBCW) contacted with either 5 × 10-4 M FeSe(s) or 5 × 10-4 M elemental Se(s).

The calculated solubility of elemental Se(s) at pH 8.4 and pH 9.9 with variation in redox potential is shown in Figure 2.5.2 (AEAT). The reason for the strong dependence of the solubility of Se(s) on redox potential shown in Figure 2.5.1 (AEAT) is now clearer. The Se(s) solubility is predicted to be strongly dependent on redox potential because its oxidation state in the solid, Se(0), is always different to that in solution, Se(-II), Se(IV) or Se(VI). The minimum in the solubility curves occurs where Se(s) is in equilibrium with both selenide and selenite species in solution (V-shape valley profile).

32/328

1.0E+00

1.0E-01

1.0E-02

1.0E-03

1.0E-04

-3 Predicted Se(s) solubility (pH = 8.4) 1.0E-05

1.0E-06 Predicted Se(s) solubility (pH = 9.9) 1.0E-07

1.0E-08 selenide

1.0E-09 Concentration/ mol dm mol Concentration/ selenite 1.0E-10

1.0E-11 selenate

1.0E-12 Speciation prediction: the predominant 1.0E-13 oxidation state is shown only for pH 8.4 1.0E-14 -600 -400 -200 0 200 400 600 800 Eh/ mV

Figure 2.5.2 (AEAT): Predicted Se(s) solubility as a function of Eh in the presence of pyrite at pH = 8.4 and pH = 9.9 (inventory-limited concentration = 5 × 10-4 mol dm-3).

The predicted solubility of FeSe(s) with varying redox potential is shown in Figure 2.5.3 (AEAT) at pH 8.4, for a range of different assumptions regarding equilibration with other solid phases. Curve A shows the predicted FeSe(s) solubility for the assumption that it can be converted to Se(s) when the elemental form is predicted to be the most stable phase. In this case the predicted solubility of FeSe(s) is around 2 × 10-8 mol dm-3 but the solid is only predicted to be stable up to around -400 mV. Above this Eh value the solubility decreases due to the formation of Se(s) before rising as Se is oxidised to soluble selenite and selenate species. For the assumption that pyrite is in equilibrium with the solution (Curve B), the selenium solubility is predicted to be lower at low Eh values due to an increase in Fe(II) in solution. For the assumption that Fe(II) is oxidised to Fe(III) and will form crystalline hematite, an increase in Se solubility is predicted between Eh values of -300 mV and 0 mV (Curve C). Similar trends are predicted at pH 9.9.

33/328

1.0E+00

1.0E-01

1.0E-02

1.0E-03 Curve C 1.0E-04

-3 1.0E-05

1.0E-06

1.0E-07 Curve A

1.0E-08 Curve B 1.0E-09 Concentration/dm mol Predicted FeSe(s) solubility (pyrite present, Fe(OH)3 can precipitate) 1.0E-10

1.0E-11 Predicted FeSe(s) solubility (pyrite present, haematite can precipitate) 1.0E-12

1.0E-13 Predicted FeSe(s) solubility (no pyrite, all Se minerals can precipitate)

1.0E-14 -600 -400 -200 0 200 400 600 800 Eh/ mV

Figure 2.5.3 (AEAT): Predicted solubility of FeSe(s) as a function of Eh at pH = 8.4 (inventory-limited concentration = 5 × 10-4 mol dm-3).

34/328

3. Geochemical behaviour of selenium in Boom Clay

35/328

36/328

3. Geochemical behaviour of selenium in Boom Clay

3.1. Introduction

The fate of selenium in Boom Clay is governed by various abiotic and perhaps also microbially-mediated reactions. The main mechanisms expected to immobilise 79Se in Boom Clay deal with sorption, reduction and precipitation reactions. In the absence of catalyst or microbial activity, kinetic limitations and redox disequilibrium can affect the reduction of selenate and selenite species. The role of the selenium association with organic matter and bacterial reduction are also important and cannot be overlooked. However, the present experimental works on selenium in Boom Clay have only taken into account mineral reactions and neglected biochemical processes, amongst others responsible for the formation of volatile low molecular mass organic compounds of selenium.

3.2. Sorption of selenium on Boom Clay and its components

2– 2– – Inorganic selenium is always present in water as an anion: SeO4 , SeO3 and HSe . The non- specific electrostatic repulsion forces between these anions and the negatively charged surface of clay minerals (at neutral or slightly alkaline pH) hinder their sorption.

However, some oxy-anions, such as, a.o., borate, silicate, phosphate, molybdate, arsenate, and selenite, can form inner-sphere complexes at the surface of oxy-hydroxides of Fe3+ and Al3+. This sorption mechanism also often implies a ligand exchange when a functional group expulse, or replace another one. The common reaction scheme can be represented as follows:

2– – – SeO3 + =Fe–OH <—> =Fe–SeO3 + OH or, – – HSeO3 + =Fe–OH <—> =Fe–SeO3 + H2O where the selenite ligand attacks the Fe3+ nucleus and expels one hydroxyl group to take its place at the surface of the mineral. This reaction is analogue to the nucleophilic substitution reaction in organic chemistry. The expulsion of the –OH “leaving group” can be facilitated + under acidic conditions when the group is protonated as –OH2 . Indeed, the departure of a – neutral H2O molecule is energetically easier that this of an OH ion.

It is formally possible to write several similar reactions describing the complexation of selenite on hydroxylated surfaces. These reactions can imply the formation of monodentate or bidentate surface complexes with one or two metallic nuclei (here Fe3+). Therefore one will refer to monodentate, bidentate mononuclear, or bidentate binuclear, surface complexes. The

37/328

here abovementioned reaction refers to the formation of a simple monodentate mononuclear surface complex. More complex chemical drawing are required to adequately describe the molecular structure of bidentate mononuclear, or bidentate binuclear, surface complexes.

2– Under Boom Clay conditions, amongst the selenium oxyanions, only selenite (SeO3 ) forms stable inner-sphere complexes at neutral and slightly alkaline pH and can sorb on oxide surfaces by surface complexation reaction. Selenate sorption is much weaker than that of selenite and requires quite acidic conditions for the ligand exchange. As a consequence, selenate forms preferentially weak outer-sphere complexes and the formation of inner-sphere complexes is not favoured in Boom Clay.

The more efficient sorption sites expected are these of hydrous ferric oxides (HFO) (Dzombak and Morel, 1990) and oxy-hydroxide of aluminum. HFO sorption sites can be easily formed at the surface of pyrite when oxidation occurs. The aluminium-type sites are present as aluminol groups at the lateral edges of TOT (tetrahedron – octahedron – tetrahedron) clay minerals platelets where the gibbsite octahedral layers becomes accessible. In the case of the non- swelling TO (tetrahedron – octahedron) clay minerals, such as kaolinite, Al–OH groups could also be accessible at the outer octahedral basal plane of the clay platelets.

Only spectroscopic data can provide evidences confirming the existence of a particular surface complex. X-ray Absorption Spectroscopies (XAS) are tools of choice but they require very intense synchrotron light sources and large international infrastructures. Their application still remains delicate for alumino-silicate minerals because Al and Si atoms have a relatively small electron number (z = 13 and 14 respectively) and so poorly scatter X-rays. Effect of water radiolysis and the interactions of redox-sensitive elements with the intense X- ray beam can also complicate the studies: indeed, Bruggeman (2006) noticed the reduction of selenite into pink elemental Se0 during exposure of his sorption samples to the X-ray beamline. Finally, the mathematical interpretation of XAS experimental data to infer the molecular structure of surface complexes is also delicate. It must not be underestimated and requires the collaboration of skilled specialists in the field.

Hayes et al. (1987) were amongst the first to apply the Extended X-ray Absorption Fine 2– 2– Structure (EXAFS) spectroscopy to the study of the sorption of SeO4 and SeO3 on goethite 2– (FeO(OH); Fe: z = 26). They evidenced the formation of SeO3 inner-sphere bidentate binuclear surface complexes with goethite. Boyle-Wight et al. (2002) also studied the 2– adsorption of SeO3 on γ-Al2O3 with EXAFS but the weak X-ray scattering produced by 2– aluminum atoms (z = 13) did not allow them to successfully determine the structure of SeO3 surface complexes. Peak et al. (2006) also make use of EXAFS and X-ray Absorption Near- 2– Edge Spectroscopy (XANES) to unravel the SeO3 bonding mechanisms on a different Al- bearing materials.

38/328

In the frame of the present study, Bruggeman (2006) has performed XANES and EXAFS measurements on samples of illite du Puy contacted with selenite solution at different pH – values. XANES and EXAFS spectra closely reflected the spectrum of the HSeO3 species. New features observable in the Fourier-transformed radial structure function (RSF) of EXAFS spectra pointed out the formation of inner-sphere complexes with selenite on the clay platelet edges. The EXAFS spectra could be fitted up to the first and the second coordination shells, but not beyond. In addition, the EXAFS spectra were not recorded in the range of in situ pH value (8 – 8.5) for Boom Clay and were affected by non-negligible concentration of selenite remaining in solution. So the molecular structure of the different surface complexes could not be accurately determined and the corresponding sorption mechanisms remain uncertain. It was thus not possible to discriminate between monodentate, bidentate mononuclear, or bidentate binuclear, selenite surface complexes.

2– Bruggeman (2006) also performed sorption isotherms with SeO3 and sorption as a function of pH (so-called “sorption edge”) on illite du Puy. Adsorption was maximal in the neutral-to- acid pH range and sharply decreased from pH 6.0 to 8.5. Sorption isotherms were fairly linear and independent of the background electrolyte concentration. This latter feature is inherent to the inner-sphere sorption and thus confirmed the results obtained by EXAFS spectroscopy. Surface complexation modelling using the titration data determined by Bradbury and Baeyens (2005) for illite du Puy was also consistent with a monodentate inner-sphere surface complex with selenite.

For Boom Clay, a linear sorption behaviour has been observed for selenite only. Distribution 3 -1 ratios (RD) in the range from 5 and 5 000 dm kg have been measured. The distribution 2– ratios for SeO3 depend mainly on the solution-to-liquid ratios and on the contact time. The removal of selenite from solution is the faster for systems containing large quantities of solids 2– and low concentration of SeO3 .

However, the sorption studies of selenite on pyrite, or on Boom Clay containing pyrite, are 2– complicated by the progressive chemical reduction of SeO3 followed by the precipitation of elemental selenium, Se(0), or iron selenide, (FeSe, Fe×Sey, FeSe2). Then, the studied systems progressively evolves with time from a sorption mechanism to a slow chemical/precipitation process. At the end, the dissolved selenium concentration seems to be solubility controlled because it converges towards a value of ~ 3 × 10-9 mol dm-3 in all the studied systems. This makes difficult to interpret the experimental results and to translate them in a simple conceptual model directly usable for the safety studies. Aqueous selenium is removed from solution by a combination of sorption and precipitation whose proportion varies as a function of time.

39/328

2– In Boom Clay, no sorption has been observed in batch tests for selenate (SeO4 ) which is not known to form strong inner-sphere complexes with hydroxylated surfaces at in situ pH value representative of Boom Clay. However, EXAFS studies of Manceau and Charlet (1994) and recent surface complexation modelling from Fukushi and Sverjensky (2007) indicate that at lower pH, selenate could also form weak inner-sphere complex on the surface of iron oxide.

This could also explain the small retardation factor (R = 3.2 corresponding to a minute Kd of 0.4 L kg-1) recently determined for sulfate from breakthrough curves of percolation tests made 35 2– with SO4 . So, selenate is expected to migrate in Boom Clay as a very weakly sorbed or nearly unretarded species.

No sorption has been observed for selenide (HSe–), an anion lacking of oxygen atom, so a priori not expected to form inner-sphere complexes. However, it has only been studied in the frame of solubility experiments made with elemental selenium, Se(0), or iron selenide, FeSe. Studies of selenide in solution are also complicated by the low solubility of Se(0) and FeSe solid phases and the great sensitivity of solid selenides and dissolved HSe– to oxidation. 2– If oxidation occurs, the selenium concentration is dominated by small amounts of SeO3 or 2– SeO4 soluble species and no solubility value can be determined. At the present time, we do not have any experimental evidence proving the sorption of HSe– in Boom Clay, or on pyrite.

In addition, according to the principle of hard and soft Lewis acids and bases (HSAB) (Sposito, 1981, p. 76), no significant interaction with the surface hydroxyl groups (strong Lewis acid sites) located on the clay edges is expected for soft Lewis bases as iodide and selenide. So, based on the HSAB classification, in the absence of experimental evidence, HSe– (a large and polarisable single anion with a low electronegativity as I–) is tentatively considered as an unretarded species in Boom Clay as it is the case for I–.

In spite of this, one could imagine that a selenide anion could be exchanged with a sulfide anion at the surface of pyrite, but this process could not be demonstrated in the frame of the present work. So, for safety studies selenide is presently considered as a non-retarded species.

However, a paper of Liu et al. (2008) (Subatech, Ecole des Mines de Nantes, France) published after the completion of the present study mentions that dissolved selenide (HSe–) could be oxidized in elemental selenium (Se0) by the disulfide anion present in pyrite (which get reduced into sulfide), giving rise to an apparent distribution coefficient of ~ 67 L·kg-1 for selenium.

HSe– —> Se0 + H+ + 2 e– (oxidation of hydrogeno-selenide in elemental selenium) 2– + – 2– – – – S2 + H + 2 e —> S + HS (reduction of pyrite disulfide, Se–Se , in two sulfides) Fe2+ + S2– —> FeS (precipitation of one of the 2 sulfides with a ferrous ion)

– 0 – HSe + FeS2 —> Se + FeS + HS (global apparent reaction)

40/328

This reaction could be responsible for the possible retention of selenide in Boom Clay and should be experimentally verified by means of batch tests with Boom Clay pyrite suspensions and also with diffusion experiments to be made with H75Se–. However, in the absence of any experimental measurements, selenide is presently considered as non sorbed in Boom Clay.

2– 3 -1 To summarize, exception made of SeO3 , (RD: 5 – 5 000 dm kg ), no other dissolved selenium species has been found to significantly sorb onto Boom Clay or pyrite.

3.3. Kinetics of reduction of selenite

2– Only the kinetic of reduction of selenite (SeO3 ) has been studied in detail by Bruggeman et al. (2005). No results have been obtained in the present work for the other selenium species.

2– 2– Upon contacting SeO3 with pure pyrite (FeS2), a steady decrease in time (60 days) of SeO3 concentrations was observed, until a final concentration in solution of 3 × 10-9 mol dm-3 was reached. All investigated systems appear to follow a same rate law. According to Bruggeman et al., (2005), the decrease in Se(IV) concentration as a function of time seems proportional to the concentration of dissolved selenite and to the amount of solid FeS2 present in the system, and inversely proportional to the square root of the FeS2 occupancy by selenite. These 2– observations suggest that SeO3 reduction takes place through sorption onto FeS2 and that a selenium precipitate with a solubility of 3 × 10-9 mol dm-3 was formed.

d[]Se(+IV) [FeS ] = −k [][]Se(+IV) FeS 2 0 (eq. 3.3.1) dt 2 0 Se(+IV) []0

One also observes that the Se(IV) reduction rate slows down when clay minerals and dissolved Boom Clay organic matter (OM) are present in the system. It is interpreted as a 2– competition mechanism because SeO3 also remains associated with illite by an inner-sphere complex or with OM by an iron bridge.

2– The reduction of selenate (SeO4 ) in the Boom Clay conditions is not presently demonstrated and certainly very difficult, if existent, in the absence of catalyst. The kinetic limitations could likely be overcome by the enzymatic activity of sulfato-reducing bacteria (SRB) if they develop in the clay suspensions in the laboratory experiments. However, under in situ conditions with a high degree of clay compaction, due to space (and water) restrictions, the microbial activity is expected to be limited. According to a recent work of Cui et al. (2006),

41/328

green rust, a Fe(II)-mixed double layer hydroxide found in the corrosion products of iron under anaerobic conditions could be also able to reduce selenate.

The kinetic of reduction/dissolution of elemental Se(0) into HSe– has not been studied in detail. The reduction of Se(0) probably occurred at low Eh values in solubility tests when iron strips were added to the system to lower the redox potential. The HSe– ions released by the reduction/dissolution of Se(0) are expected to precipitate with Fe2+ to form FeSe. This precipitation of a less soluble phase simultaneously increases the dissolution of elemental selenium and the corrosion of metallic iron in the tests of AEAT with iron strips.

3.4. Solubility of selenium compounds under reducing conditions

The solubility of selenium mainly depend on the redox conditions. Under oxidizing conditions, selenium (VI) and (IV) salts are soluble and very mobile. No solubility limit are 2– 2– reached for SeO4 and SeO3 in Boom Clay water. Under reducing conditions, the mobility of selenium is limited by the solubility of poorly soluble compounds, such as elemental selenium, Se(0) and iron selenide(-II), FeSe. Under reducing conditions, selenium solubility also strongly depends on the value of the redox potential (see p. 33, Chapter 2, the V-shape curve on Figure 2.5.2).

The solubility of Se(0) and FeSe has been measured in synthetic clay water and in Boom Clay porewater, in the presence and in the absence of pyrite and iron strip, specially added to control the Eh value, the main parameter influencing the selenium solubility under reducing conditions (Figures 2.4.2 and 2.5.2).

In general, the solubility values obtained with stable commercial compounds of Se(0) and FeSe by AEAT are higher than these determined by means of solids synthesised on purpose by KULeuven and labelled with 75Se. Scanning electron microscope (SEM) images have revealed the undesirable presence of oxidized phases in the commercial synthetic products. Indeed, the purity of the solid phase is of the utmost importance. If oxidation products slightly 2– 2– contaminate the reagents, non-solubility limited species (SeO3 or SeO4 ) impose the selenium concentration in solution and it becomes strictly impossible to determine a true solubility value. To remove the soluble selenium salts, the reagent were successively washed with desionised water and the supernatant removed. After this “pre-leaching” operation, the solubility were determined again and much lower values were obtained. The main results of solubility experiments performed by AEAT and KULeuven (Appendix A4: Selenium speciation behaviour in Boom Clay) are summarised at Table 3.4.1.

42/328

Table 3.4.1: Main results of solubility (S) experiments performed with Se(0) and FeSe as solid phases.

Solid Phase Se(0) S (mol dm-3) FeSe S (mol dm-3)

Medium \ Lab AEAT KULeuven AEAT KULeuven

SCW (no OM) 1 × 10-7 – 3 × 10-7 2 × 10-8 – 9 × 10-8 1 × 10-5 – 3 × 10-5 4 × 10-10 – 4 × 10-9

-7 -7 SCW (+ FeS2) 1 × 10 – 4 × 10 — — —

BCW (+ OM) 2 × 10-7 – 6 × 10-5 — 6 × 10-7 – 5 × 10-6 —

BCW (+ clay) — 1.7 × 10-9 – 8 × 10-8 1 × 10-6 – 2 × 10-6 8 × 10-10– 2 × 10-9

Fe strips 1 × 10-6 – 4 × 10-6 — 3 × 10-8 – 5 × 10-8 —

SCW: synthetic clay water; BCW: Boom Clay water; OM: organic matter.

3.5. Association of selenium with Boom Clay organic matter

Natural selenium is often found in association with organic matter (OM) in soils and marine sediments. Boom Clay organic matter also contains traces of natural selenium (for more information see Appendices A2 and A8). This selenium is likely of organic nature and was incorporated in organic molecules by biochemical pathways where selenium occupies the place of sulfur and plays an important role in biochemical and enzymatic processes necessary to life. The incorporation of traces of natural selenium in Boom Clay organic matter by living organisms has occurred at the time of the clay sedimentation, 35–30 millions years ago.

Natural selenium is present in Boom Clay and appears to be primarily associated with pyrite

(FeS2). The concentration of natural selenium measured in Boom Clay porewater is about 2.4 × 10-8 mol dm-3, a value consistent with the solubility limit of a selenium-bearing solid phase. However, as a fraction of selenium in interstitial water is associated with natural dissolved organic matter (or may even be part of its molecular structure, in case of really covalent bond), the selenium concentration measured in porewater could not correspond to a true solubility.

The exact significance for geochemical calculations of the background levels of natural selenium concentrations measured in pyrite and in Boom Clay porewater remain also an open question. If the total selenium concentrations measured in Boom Clay porewater are partly ascribable to organo-selenium species, their values cannot be directly used in geochemical calculations to determine the solubility, or the saturation index (SI) of inorganic selenium

43/328

species such as iron selenide. Therefore, it is recommended to systematically analyse the natural selenium concentration in the purified clay fraction, pyrite, calcium carbonate, and in the Boom Clay organic matter (both solid (kerogen) and dissolved) and to determine the fraction of true inorganic selenium in Boom Clay.

All the common species of dissolved inorganic selenium in water are of anionic nature: 2– 2– – SeO4 , SeO3 and HSe . The non-specific electrostatic repulsion forces between these negatively charged entities and organic matter (a polydispersed polyelectrolyte structure bearing many negative functional groups) is also a priori not favourable to a direct association between inorganic selenium and OM.

2– Selenite (SeO3 ) is the only aqueous species of selenium clearly interacting with “dissolved” high molecular weight (MW) Boom Clay organic mater. A set of experiments were performed with homogeneous solutions, i.e., in absence of solid phases or mineral surfaces. A spike of 75 2– -6 -3 SeO3 (10 mol dm ) was added to synthetic and natural solutions containing “dissolved” organic matter (DOM) of different origin: OM extracted from Boom Clay and redispersed in synthetic water (128 mg dm-3), or OM-rich water from Gorleben aquifer (160 mg dm-3). After increasing contact times, the solutions were ultracentrifuged and 75Se and DOM measured in the supernatant respectively by γ-counting and readout of UV absorbency at 280 nm. A progressive removal of 75Se from solution was observed. After several months about 70–80 % of selenite introduced in the aqueous systems containing DOM was removed. So, an association or a reaction with the high MW centrifugeable OM was suspected. Gel permeation (GPC) and ion (IC) chromatography measurements were then performed to obtain information on the aqueous speciation of selenium in the supernatant recovered after centrifugation or ultrafiltration after contact with dissolved organic matter. The GPC results 75 2– revealed than the narrow peak of SeO3 observed at long elution time typical for small anions gradually decreased while a broad band of 75Se was simultaneously growing in the front of the chromatogram at the shorter elution times corresponding to the breakthrough of the larger OM molecules. These observations first suggest an association between selenite and OM, but the formation of colloidal elemental selenium by reduction of selenite by OM could also give a consistent explanation of the results. So, two possible mechanisms are proposed to explain these observations, but confirmation by advanced X-ray absorption spectroscopies (XAS) is still needed to raise all ambiguities.

In a first hypothesis, selenite could form inner-sphere complexes with Fe present in the humic 2– acids and be bound by an “iron bridge” (OM—Fe—SeO3 ) as suggested by Gustafsson and Johnsson (1994) in laboratory experiments on forested ecosystem. However, it is unclear if 2+ 2– 3+ Fe –humic complexes are able to bind SeO3 under reducing conditions as Fe –humic complexes do under air.

44/328

On the other hand, selenite is easily prone to reduction in the presence of organic agents such as ascorbic acid (Shaker, 1996). According to several authors (Matthiesen, 1994; Nakayasu et al., 1999; Struyck and Sposito, 2001) humic acids have also a certain redox capacity and 2– are able to reduce oxyanion in solution. So, SeO3 might be reduced by OM under the form of amorphous colloidal elemental Se(0). These selenium colloids could directly precipitate out of solution or might interact by hydrophobic forces with organic matter colloids, so that a weak physical association [Se colloids/OM colloids] could also be envisaged.

Finally, a third tentative association mechanism envisaged is the direct complexation of 2– SeO3 by positively charged functional groups present in the poorly characterised structure of humic acids such as, e.g., ammonium groups from amino-acids residues or peptide degradation products. However, to our knowledge little is known on positively charged sites in humic acids and such a hypothesis also requires the amine functions to be protonated what first needs a pH well below the pKa value of the functional group.

In contrast to selenite, no association of selenate with high molecular organic matter has been evidenced in Boom Clay. The same method used for selenite was also applied to selenate and no removal from solution was observed by ultracentrifugation or ultrafiltration. Moreover, GPC and IC measurements did not succeed to evidence an association, or a reduction, of 2– 2– SeO4 with OM. So, the interaction of SeO4 with OM is considered to be very weak (below detection capabilities of the techniques used) or negligible because selenate reduction is kinetically hindered and that selenate does not form inner-sphere complexes at slightly alkaline pH to establish strong chemical bonds with OM (e.g., via iron bridges).

Finally, up to now, no association has been observed between elemental selenium, Se(0), or iron selenide, FeSe, contacted with Boom Clay organic matter. No effect could be observed by ultrafiltration (cfr. AEAT results), suggesting the absence of small selenium colloids in the systems studied, or that Se-bearing particles were larger than 0.22 μm and so already eliminated at the first prefiltration step. Independent analyses with gel permeation chromatography and La3+ precipitation have not make possible to detect a significant association between Se(0), or FeSe, and colloids of Boom Clay organic matter.

3.6. Conclusions

The main mechanisms controlling the behaviour of inorganic selenium in Boom Clay are summarised in Table 3.6.1. They can be summarised as follows: Selenate is not or weakly sorbed, is very recalcitrant to reduction and could subsist under reducing condition. Selenate is not solubility limited, nor associated with organic matter (OM).

45/328

Selenite is moderately sorbed on Boom Clay and easily undergoes reduction. Selenite is soluble, except in cementitious buffer where it forms an insoluble calcium salt in the presence of portlandite (Ca(OH)2), or solid solutions with CSH, AFt (ettringite) and AFm cement phases (for more information, see Appendix A5.1: Immobilisation of selenium in cementitious buffer). An association is observed with OM, but it is unclear if the reason is due to the formation of inner sphere complexes involving the creation of iron bridges with OM, or if this association is caused by the formation of nano-colloids of elemental selenium (Se0) or iron selenide (FeSe×) intimately mixed with OM (weak hydrophobic interactions).

Table 3.6.1: Matrix of different interactions expected for inorganic selenium species in Boom Clay. Oxidation Species Sorption Reduction Precipitation OM State Association

2– +6 SeO4 (aq) Very low Extremely No solubility None reluctant limit

2– +4 SeO3 (aq) Medium Easy CaSeO3 Association (inner-sphere (+ solid solution (+ reduction) complex) in cement) observed

0 Se(s) — Slow Se Colloid – colloid

– –2 HSe (aq) Unknown — FeSe2 , Fe×Sey None (not considered)

Elemental selenium (Se0) is a solid phase which might control the solubility of selenium in Boom Clay and can also interact with OM by means of weak colloidal interactions. Elemental selenium can be further reduced in selenide, or at the contrary easily oxidized in selenite, giving rise to the famous V-shape solubility curve of reduced selenium.

Finally, selenide is poorly soluble in the presence of ferrous ions, but very few is known on its sorption in Boom Clay and this point still deserve experimental investigations. Although not impossible, no interaction has yet been observed with Boom Clay OM.

46/328

4. Determination of migration parameters of selenium in Boom Clay

47/328

48/328

4. Determination of migration parameters of selenium in Boom Clay

4.1. Introduction

The values of transport parameters of selenium in Boom Clay presently selected for performance assessment calculations only deal with selenate. The reason is that because of redox disequilibrium, this species could subsist in Boom Clay for an undetermined period of time and is also assumed to have the highest mobility. Because of its stability under laboratory conditions, well controlled migration experiments have been performed with selenate and sulfate, its chemical analogue. On the other hand, it is a particularly challenging task to study the transport of selenite and selenide in Boom Clay. Selenite is a more reactive species than selenate and undergoes a progressive reduction after sorption on the edges of clay minerals (inner-sphere complexes). As a consequence, it slowly precipitates during transport to form elemental selenium or iron selenide. Interpretations and modelling of several migration experiments with selenite affected by chemical-coupled transport did not succeeded to provide reliable migration parameters for selenium in Boom Clay (see more detail in Appendix A9: Selenium migration behaviour in Boom Clay). Experiments with selenide, the thermodynamical stable form of selenium in Boom Clay are also not easy and remain to be done. First, they require the complete chemical reduction of all selenium species in selenide, which is not a trivial operation starting with 75Se-labelled selenite or selenate sources. The reduction reaction or the chemical separation processes must be complete to isolate the only selenide species and to avoid contamination by other selenium species or residual chemicals. So, the chemical purity of the 75Se-labelled selenide must be carefully controlled to guarantee that no other selenium species are still present and that no chemical impurities that could affect experiments are left after the reduction operations. Finally, the very low solubility limit of iron selenide and the high sensitivity of dissolved HSe– to oxidation complicate the experiments and preclude the use of pulse injection test. Only well characterized pure FeSe solid sources confined between two undisturbed clay cores and imposing a constant solubility limit can be used in practice.

So, specific migration experiments have been performed to determine the transport 75 2– 35 2– parameters of selenate ( SeO4 ) and sulfate ( SO4 ) in Boom Clay. Selenate was studied by electro-migration experiments while sulfate was investigated with different conventional diffusion tests. Sulfate labelled with 35S chemically resemble to selenate and was chosen to understand the migration of divalent anions in Boom Clay, and especially to better understand the effect of the double charge on the anionic exclusion.

49/328

4.2. Experimental 35 2– Migration experiments with sulfate ( SO4 ) The different types of percolation experiments (core infiltration tests made on small “column” (height = 7 cm) of compact clay core submitted to a high hydraulic gradient) used to study the 35 - migration of sulfate (labelled with S: ß emitter, T½ = 86.7 d) in Boom Clay are schematically illustrated in Figure 4.2.1. The classification and the description of D2-pulse injection, C4-type percolation experiments and C3-type back-to-back diffusion experiment (same configuration as C4 but without hydraulic gradient), are given by Henrion et al. (1990) and Put et al. (1992) and also summarized in Appendix A9: Selenium migration behaviour in Boom Clay. Modelling of the breakthrough curves from the D2-pulse injection and the C4- percolation tests results in values for Dapp and ηR, while interpretation of the diffusion profile in the clay after post-mortem analysis of a C3 back-to-back diffusion experiment only provides the Dapp value.

RN ΔP injection

RBCW percolation Clay (1-2 MPa) “D2-pulse injection RN collection [RN]

RN breakthrough curve (for non or weak retarded RN)

ΔP RBCW t injection

Clay Clay

RN collection RN Source

[RN] [RN] “C4-percolation”

x t

RN distribution profile after RN breakthrough curve (for cutting (for retarded RN) solubility controlled RN – sigmoïd curve - and for non or weak retarded RN) Figure 4.2.1: Different types of percolation experiments used to determine the transport parameters of 35 2– sulfate with SO4 : D2-pulse injection and C4-type percolation experiments (see Put et al., 1992).

75 2– Electromigration of selenate ( SeO4 ) The experimental set-up used for the electromigration tests with 75Se-labelled selenate (75Se: - e capture (EC), T½ = 120 d) at different values of electrical field is described by Beauwens 75 2– 75 et al. (2005). Chemically pure SeO4 was not commercially supplied because Se is 75 2– usually conditioned under the form of selenite ( SeO3 ). Beauwens et al. (2005) prepared

50/328

pure 75Se-labelled selenate by oxidizing aliquots of labelled selenite with hydrogen peroxide

(H2O2).

Modelling of the diffusion profiles after post-mortem analysis of the clay cores provides apparent dispersion coefficients (Di) and apparent velocities (Vapp). By performing electromigration experiments at different electrical fields, hence different apparent velocities, the apparent diffusion coefficient (Dapp) can be obtained from following linear relationship:

Di = Dapp + αVapp

Where α stands for dispersion length (expressed in meter).

4.3. Results

The results obtained with D2-pulse injection and C4-type percolation experiments are discussed hereafter. Breakthrough curves of the percolation experiments (C4 and D2) are presented in Figure 4.3.1 and the corresponding recovery curves in Figure 4.3.2.

8 E+6 2 E+6 C4 – 7.12 D2 – 7.4 7 E+6 1 E+6

6 E+6 1 E+6 C4 – 7.11 5 E+6 1 E+6 C4 – 7.12 4 E+6 D2 – 7.2 8 E+5 C4 – 7.11 D2 – 7.4 3 E+6 6 E+5

2 E+6 4 E+5

1 E+6 2 E+5 7.4] [7.2 – (Bq/L) Conc. S-35 S-35 Conc. (Bq/L) [7.11 – 7.12] – [7.11 (Bq/L) Conc. S-35 D2 – 7.2 0 E+0 0 E+0 0 5 10 15 20 25 30 35 40 Σ Volume percolated water (ml)

35 2– Figure 4.3.1: Breakthrough curves of SO4 obtained with the C4 and D2 type of percolation experiments. The 35S recovery yield considerably varied for an unclear reason.

Both Figures 4.3.1 and 4.3.2 show a different recovery yield of 35S activity in the two types of percolation experiments. A first set of experiment presents an acceptable recovery yield

51/328

within the error bar of the activity measurements (100 to 110 %) while the second one is much lower (25 to 40 %) for an unclear reason.

120 D2 – 7.4

100 C4 – 7.12

80

60

C4 – 7.11 40

S-35 Recovery (%) D2 – 7.2 20

0 0 20406080100 Σ Volume percolated water (ml)

35 2– Figure 4.3.2: Recovery curves of SO4 obtained with the C4 and D2 type of percolation experiments. The 35S recovery yield considerably varied for an unclear reason.

The incomplete recovery affecting the two types of experiments could be due to an undetected experimental perturbation (e.g., sulfate precipitation induced by pyrite oxidation).

In case of pyrite oxidation, several poorly soluble sulfate salts could be formed, with whom 35 2– SO4 could coprecipitate, or undergo an isotopic exchange (solid solution): barite (BaSO4), celestite (SrSO4), gypsum (CaSO4 · 2 H2O), jarosite (KFe3(SO4)2(OH)6).

Despite the problem of incomplete recovery, consistent Dapp (and ηR) values were derived: -11 -11 -1 average Dapp = 2.7 × 10 (D2) and 3.8 × 10 m² s (C4).

35 2– Figure 4.3.3 presents the SO4 profile obtained in a clay core after a pure diffusion experiment (type C3 according Put et al., 2002) with sulfate (14 days duration). Only the apparent diffusion coefficient of sulfate was determined by means of this pure diffusion -11 2 -1 experiment: Dapp = 4.4 × 10 m s .

75 2– Figure 4.3.4 summarises the results of 7 electro-migration experiments made with SeO4 at different values of electrical field (Beauwens et al., 2005). The apparent diffusion coefficient

(Dapp) is then obtained by linear extrapolation to a zero apparent velocity (Vapp) (i.e., in the

52/328

75 2– absence of electrical field) from a plot of the dispersion coefficients (Di) of SeO4 as a function of the apparent velocity. More explanations on the electro-migration technique and the way to derive transport parameters from these experimental results are given by Maes et al. (1998; 1999; 2001; 2002).

800 35 2– SO4 Pure diffusion (C3) 600 experiment with 2 back-to-back Boom Clay cores 14 days 400

Bulk Activity Bulk (cps/g) 200

0 -40-30-20-100 10203040 Distance from source (mm)

35 2– Figure 4.3.3: Distribution profile of SO4 in a Boom Clay core after a pure diffusion experiment (duration: 14 days).

1,2E-10 Di (m²/s) 1,0E-10

8,0E-11

6,0E-11 y = 2,19E-04x + 1,73E-11 R2 = 9,01E-01 4,0E-11

2,0E-11

0,0E+00 0,0E+00 1,0E-07 2,0E-07 3,0E-07 4,0E-07 5,0E-07 V (m/s) app

Figure 4.3.4: Linear extrapolation of the apparent diffusion coefficient (Dapp) from dispersion coefficients (Di) obtained at different apparent velocity (Vapp) in electro-migration experiments performed with 75Se-labelled selenate at various values of electrical field.

53/328

Only the value of Dapp can be derived from the results obtained with the electro-migration technique. From the here above mentioned extrapolation, Beauwens et al. (2005) determined -11 2 -1 for selenate a value of Dapp = 1.7 × 10 m s .

2– In Table 4.3.1 the parameters derived for SO4 by means of conventional diffusion or 2– percolation experiments are compared to these of SeO4 obtained by electro-migration.

Table 4.3.1: Migration parameters derived for sulfate compared to these obtained for selenate.

Code Type Recovery ηR Dapp Deff = ηRDapp (—) (—) Yield (%) (—) (m2 s-1) (m2 s-1)

2– -11 -12 7.2–SO4 D2 ~ 25 0.29 2.8 × 10 8.1 × 10 2– -11 -12 7.4–SO4 D2 ~ 110 0.25 2.6 × 10 6.5 × 10 Average: 0.27 2.7 × 10-11 7.3 × 10-12

2– -11 -12 7.11–SO4 C4 ~ 40 0.15 3.7 × 10 5.6 × 10 2– -11 -12 7.12–SO4 C4 ~ 100 0.21 3.8 × 10 8.0 × 10 Average: 0.18 3.75 × 10-11 6.8 × 10-12

2– -11 PD–SO4 PD — — 4.4 × 10 —

2– -11 EM–SeO4 EM — — 1.7 × 10 —

All — Average: 0.23 3.17 × 10-11 7.29 × 10-12 Experiments — ± Std. Dev.: ±0.06 ±0.98 × 10-11 ±1.04 × 10-12

D2: pulse injection percolation experiment; C4: back-to-back percolation experiment; PD: pure Diffusion experiment; EM: electro-Migration experiment; Std. Dev.: Standard Deviation.

4.4. Discussion

The different applied experiments lead to a relatively consistent set of values for Dapp and ηR. For sulfate, the obtained values slightly depend on the kind of experiment (D2 or C4 percolation type). This is not only due to the nature of the transport and interaction processes

54/328

at work in each experiment, but also to limitations inherent to the modelling: it also thus depends on the conceptual model used, on the boundary conditions, on the basis hypotheses and simplifications introduced in the model itself.

-11 2 -1 Surprisingly, the value of the apparent diffusion coefficient (Dapp = 3.2 ± 1.0 × 10 m s ) is one order of magnitude lower than that normally expected in Boom Clay for the pore -10 2 -1 diffusion coefficient (Dp = 2 × 10 m s ) with the hypothesis that anions are not sorbed

(Dapp = Dp/R; if R = 1, —> Dapp = Dp). This could indicate that this hypothesis is not valid and that the studied oxyanions are sorbed (R > 1). As it can be observed on Table 4.3.1, to the highest Dapp values correspond also the lowest ηR values and vice et versa, but this relationship is not sufficiently significant and the parameters cannot be considered as correlated. The value of the effective diffusion coefficient (Deff = ηRDapp) is quite stable (7.0 ± 1.2 × 10-12 m2 s-1) but is also one order of magnitude lower than the value normally expected for anions in Boom Clay (~ 2 × 10-11 m2 s-1). The relationship between these different diffusion coefficients is given by the following set of equations:

Dapp = Daq / R Rf

Dp = Daq / Rf

Dapp = Dp / R

Deff = ηRDapp = ηDp

Where:

Dapp : apparent diffusion coefficient;

Daq : diffusion coefficient in pure water;

Dp : pore diffusion coefficient;

Deff : effective diffusion coefficient; η : diffusion accessible porosity; R : retardation factor;

Rf : rock factor depending on the tortuosity (τ) of the clay matrix and on the constrictivity of the pores.

The origin of the lower Dapp value for sulfate could be found in a lower Daq, or in a higher Rf or R value. In the absence of sorption (i.e., if R = 1), the Dapp value of divalent oxyanions – studied here is not compatible with the typical Dp value for monovalent anions such as I . The alternative is that some sorption could occur for oxyanions (R > 1) in Boom Clay. This could be the case because, in contrast to what was a priori expected, the ηR value (0.23) is higher than for the monovalent iodide (0.16). The increase of R would also compensate for a lower diffusion accessible porosity (η) expected for divalent anions because of the Donnan exclusion.

55/328

4.4.1 Pore diffusion coefficient (Dp)

The values of the diffusion coefficient of ions in pure water (Daq) can be compared for iodide, hydrogeno-sulfate and sulfate from data of Li and Gregory (1974):

– -9 2 -1 Daq (I ) = 2.06 × 10 m s ; – -9 2 -1 Daq (HSO4 ) = 1.37 × 10 m s (ratio iodide/hydrogeno-sulfate: 1.50); 2– -9 2 -1 Daq (SO4 ) = 1.07 × 10 m s (ratio iodide/sulfate: 1.93).

Only a factor < 2 is observed between the aqueous diffusion coefficients (Daq) of iodide (monovalent monoatomic ion) and of sulfate (divalent oxyanion) in water.

2 Dp = Daq/Rf = Daq/τ Where τ represents the tortuosity.

Dp depends on Daq, so a factor of 2 could be invoked.

Dp also depends on the rock factor (Rf) which is related to the tortuosity (τ): τ (HTO) ~ 3.1 vs. – – τ (I ) ~ 3.8 from experimental measurements (Rf (HTO) = 9.6; Rf (I ) = 14.7). However, the effect of the tortuosity is less important and will also results in a factor lower than 2: indeed, 2 – 2 the ratio τ (I )/τ (HTO) = 1.50. Thus, the Dp value expected for a divalent anion could be globally adapted by a factor of about 2 × 1.5 = 3, but a factor of ~3.33 still subsists to explain the difference of one order of magnitude with the observed Dapp. So, sorption seems also to have to be involved to fill the gap.

4.4.2 Porosity (η)

The porosity η cannot be directly determined from a migration experiment, only the product ηR (sometimes also referred to as α or ω, the rock capacity factor) is obtained by fitting. The Bruggeman equation (De Preter et al., 1992; Put et al., 1995) provides a relationship between the effective diffusion coefficient (Deff = ηRDapp) and the porosity (η):

-10 1.5 Deff = 6.84 × 10 × η (eq. 4.4.1)

2– 2– -12 As the average Deff value experimentally determined for SO4 and SeO4 is about 7 × 10 m2 s-1 it results in a porosity η = 0.05.

56/328

The value of the diffusion accessible porosity η calculated with the Bruggeman relationship for a divalent oxyanion is indeed smaller than for a monovalent anion. The η values similarly predicted for I– and HTO are respectively 0.10 and 0.25 while values of 0.16 and 0.37 have been experimentally determined. Taking this into account, the relative high ηR value (0.23) implies thus a small retardation factor.

4.4.3 Sorption – retardation (R)

Many studies (a.o., Bar-Yosef and Meek, 1987; Hayes et al., 1987; Kafkafi et al., 1988; Geelhoed et al., 1997a,b; Hiemstra and van Riemsdijk, 1996, 1999; Lefèvre and Fédoroff, 2006; Peak et al., 2006) indicate that oxyanion species (phosphate, arsenate, selenite) can sorb onto iron and aluminium oxy-hydroxide and on clay minerals edges surface by formation of inner- and outer- sphere complexes and also strongly depends on pH (sharp edge curves with pH). At low pH, when ≡S–OH groups present at the minerals surface are protonated and that the surface becomes positively charged, more outer- and inner-sphere complexes (weak and strong sites) are formed and oxyanion sorption increases. According to surface complexation modeling of hydrous ferric oxide surface from Dzombak and Morel (1990), at the slightly alkaline pH value of Boom Clay (8 – 8.5), no significant sorption is expected for the weak outer-sphere complexes formed with sulfate or selenate. Only a weak electrostatic interaction could probably be envisaged to explain a very small retardation of sulfate in compact Boom Clay. A more detailed information is given on the sorption mechanisms of selenium oxyanions on mineral surfaces in Appendix A7: Sorption behaviour of selenite, selenate and sulfate on Fe and Al oxide surface. Recent implications of a debate initiated by Manceau and Charlet (1994) who observed by EXAFS measurements the unexpected ability for the weakly sorbing selenate, to still form weak inner-sphere complexes at the surface of goethite and hydrous ferric oxide will be discussed there. However, what could be the exact molecular nature of this weak sorption mechanism, it does not change anything to the transport process at the macroscopic scale and has no impact on the values of the migration parameters: the sorption of selenate and sulfate in Boom Clay is extremely weak and probably below the limit of detection of conventional batch sorption tests as those performed by Bruggeman (2006).

Contrary to the well known behaviour of selenite able to form strong inner-sphere complexes at the surface of iron oxides (Hayes et al., 1987) and clay minerals (Bar-Yosef and Meek, 1987), sorption experiments made by Bruggeman (2006) with selenate onto pyrite and Boom Clay suspensions did not succeed to show any significant sorption of selenate. The infinitesimal decrease of concentration in the supernatant lying within the measurement error bar it makes extremely difficult to experimentally determine very small distribution coefficients (Kd of 0.2 –1 ml/g) corresponding to R values in the range 2 – 6.

57/328

4.5. Conclusion

To summarise the previously developed arguments, one can make the following observations:

• The Dp value of divalent sulfate and selenate needs to be decreased (by a factor of ~ 2) to reflect the difference between diffusion coefficients in pure water; • Because of Donnan exclusion, the porosity of doubled charged oxyanions should be smaller than for single charged anions, and; • It is necessary to also consider a weak sorption for oxyanions.

So, considering the following values of the transport parameters of sulfate:

-11 2 -1 -10 2 -1 2- Dp = 7.3 × 10 m s (versus 1.4 × 10 m s for iodide) – calculated from Daq (SO4 )

and considering the same rock factor as for iodide Rf = 14.7; -11 2 -1 Dapp = 3.17 × 10 m s ηR = 0.23 -12 2 -1 Deff = ηRDapp = 7.29 × 10 m s

The retardation factor (R) and the diffusion accessible porosity (η) for sulfate can be calculated:

R = Dp/Dapp = 2.3 η = (ηR)/R = 0.10

The corresponding distribution coefficient (Kd) for sulfate can also be calculated:

R = 1 + (Kd × ρd/η) or

Kd = (R-1) × η/ρd

-3 where ρd is the dry density for Boom Clay: ρd = 1.7 g cm

2– -1 for η = 0.10 (diffusion accessible porosity of the SO4 anion), Kd = 0.076 L kg . -1 for η = 0.38 (total porosity filled by water, η of HTO), Kd = 0.291 L kg .

A consistent dataset of migration parameters is then obtained for sulfate and selenate. The small retardation factor of 2.3 so calculated for sulfate corresponds to a Kd value of less than 0.3 L kg-1 which remains to be independently verified with a delicate and adequately designed sorption experiment on a dilute suspension of Boom Clay dispersed in porewater under in situ conditions. This dataset is proposed for selenate for the performance assessment calculations

58/328

if this selenium species does not undergo reduction in Boom Clay at very long term because of redox-disequilibrium.

4.5.1 Summary of the migration parameters for 79Se in Boom Clay

As conclusion, because of possible severe redox-disequilibrium, it is necessary to also take into account selenate along with the expected thermodynamically favoured selenide for Boom Clay conditions. Therefore, two sets of migration parameters including both selenate and selenide have been selected for safety calculations and performance assessment and are given hereafter in Table 4.5.1.

Table 4.5.1: Migration parameters selected for selenate and selenide in undisturbed Boom Clay. 2– (*) – ($) Parameter (unit) SeO4 HSe

2 -1 -11 -10 Dapp (m s ) 3.17 (± 0.98) × 10 1.2 × 10

ηR (—) 0.23 (± 0.06) 0.16

2 -1 -12 -11 De = ηRDapp = η Dp (m s ) 7.3 × 10 1.9 × 10

(#) 2 -1 -9 -9 Daq (m s ) 1.07 × 10 1.73 × 10 2 -1 Dp = Daq /Rf (m s ) -11 -10 – 7.28 × 10 1.2 × 10 using Rf (I ) = 14.7

R = Dp/Dapp (—) 2.3 1.0 0.10 0.16 η =ηR/R (—) (divalent anion) (monovalent anion) 2– (*) SeO4 is assumed to migrate very slightly retarded, and a consistent set of migration parameters is derived from these of sulfate. ($) HSe– is assumed to migrate unretarded as a monovalent anion, and, in the absence of experimental results, its – -10 -1 migration parameters are considered to be in line with those of iodide. The Dapp (I ) = 1.4 × 10 m² s and – – – the ratio between Daq(HSe )/Daq(I ) = 0.83, and the Dapp for HSe is corrected to be in line with this ratio. 2– – 2– – (#) Daq for SeO4 and HSe are respectively taken equal to the corresponding values of SO4 and HS (Li and Gregory, 1974).

59/328

60/328

5. Summary and conclusions

61/328

62/328

5. Summary and conclusions

5.1. Overview

The inorganic chemistry of selenium is well known and the uncertainties on its thermodynamic constants used to calculate its aqueous speciation and its solubility under Boom Clay conditions are fairly limited. However, selenium behaviour in the subsurface environment, and in particular its interactions with sediment components and organic matter under reducing conditions, are complex and often difficult to unravel as illustrated by Table 5.1.1 summarising the behaviour of the different inorganic selenium species in Boom Clay as a function of their oxidation state.

Table 5.1.1: Matrix of abiotic interactions observed for inorganic selenium species in Boom Clay. Oxidation Possible Aqueous Reduction Sorption OM State Solubility Species (eq.) in Boom Clay Association Limiting Phase

2– +6 No SeO4 aq. Extremely slow Very weak Not observed (+ solid solution (redox in cement) disequilibrium)

2– +4 CaSeO3 SeO3 aq. Easy Medium Association (+ solid solution (inner-sphere (+ reduction) in cement) complex at clay observed platelet edges)

(0) 2– – 0 Se SeO3 aq. / HSe Slow — Colloid – aq. colloid

– –1 FeSe2 , Fe×Sey HSe aq. — Unknown Not observed –2 FeSe (not considered)

One of the main reasons of this complexity is that selenium is a redox-sensitive element and that its reduction rate can be hindered because it involves the transfer of multiple electrons along with multiple oxygen atoms between its various oxidation states (VI, IV, 0, -I, -II). The reduction rate strongly varies with the oxidation number of the central atom in an oxyanion, and the higher the oxidation number, the slower the reaction (Shriver et al., 1990). As a consequence, selenate can be very recalcitrant to the reduction process and can subsist in metastable conditions, far from the thermodynamical equilibrium for undetermined periods of time. As for sulfur, the inorganic speciation of selenium in water is of anionic nature, but selenium can also be incorporated in natural organic compounds by biochemical transformations where it follows the pathways of sulfur. So, the selenium cycle and its behaviour in nature are complex and remain not fully elucidated (Shrift, 1964; Nriagu, 1976).

63/328

The selenium speciation in the source term is presently unknown, and selenate could be present in the waste form. If selenate is not reduced to a lower valence because of redox- disequilibrium (or more simply by lack of electron donors, or oxidizing conditions affecting the clay barrier, e.g. in a nitrate plume developing around bituminized MLW galleries), the geochemical predictions based on thermodynamical models are no longer valid for performance assessment calculations. Selenate is not solubility limited, nor significantly retarded under Boom Clay conditions. If selenate subsists on the long term, it implies for performance assessment that no more solubility limit can be taken into account for 79Se. So, two scenarios have to be considered: a first one with selenate (worse case: redox- disequilibrium) and a second one with selenide (best case: thermodynamic equilibrium). In both case, selenium is considered unretarded, and a solubility limit can only be applied in the case of selenide if Fe2+ is also present in the system.

From batch interaction experiments, it appears that selenite is the only reactive selenium species with Boom Clay components: sorption on pyrite and illite edges and interaction with organic matter (OM), finally leading to its reduction as elemental selenium or iron selenide. Because of this reactivity, selenite is unstable under Boom Clay conditions and is not expected to subsist in the formation on the long term, so, selenite is a transient species not considered for performance assessment (PA) calculations.

Natural selenium is also present in Boom Clay, mainly in pyrite and in the mobile and immobile organic matter and could also possibly be correlated to carbonate (Carignan, 2008; Personal Communication; Gaucher and Tournassat; 2008; Personal Communication). Two questions arise thus: 1. What is the real significance of the measured total selenium concentration in porewater ? Is it possible to estimate a “natural” solubility limit for selenium released by pyrite in Boom Clay water if a non-negligible fraction of dissolved selenium consists of organo- selenium incorporated in the molecular structure of organic matter ? 2. Is there a possible interaction (e.g., by isotopic exchange) between 79Se and natural selenium in Boom Clay (3 components in the system: pyrite, kerogen and carbonate) ?

5.2. Key uncertainties and abstraction for performance assessment

To draw robust conclusions and recommendations for performance assessment and to address open questions in future researches, it is important to clearly identify the remaining uncertainties and our knowledge gap. The main features of the behaviour of selenium in Boom Clay are thus summarized in Table 5.2.1 and commented hereafter.

64/328

Table 5.2.1: List of uncertainties dealing with selenium to be accounted for in performance assessment studies. Properties Known Uncertain Comments

Half-life of 79Se ± ± Present best estimate circa 295 – 377 ka

Inventory in the source term z Poorly known: volatilisation in reprocessing

Speciation in the source term z Lack of experimental measurements

Abiotic processes in Boom Clay Known Uncertain Comments

Sorption 2– – SeO4 z Very weak sorption for selenate 2– – SeO3 z Selenite prone to inner-sphere complexes – HSe– z Unknown for selenide, but expected weak

Reduction 2– – SeO4 z Very slow reduction: redox-disequilibrium 2– – SeO3 z Quite fast reduction for selenite

Precipitation 2– – SeO4 z No solubility limit for selenate 2– – SeO3 z Selenite poorly soluble in cement – 2+ – HSe z FeSe2: very poorly soluble in clay (+ Fe )

Abiotic association with OM 2– – SeO4 z No observed association for selenate 2– – SeO3 z Very clear association for selenite – HSe– z Unknown for selenide (see also kerogen)

Microbial transformations Known Uncertain Comments

Near-field z Microbial metabolism not excluded

Far-field z Low probability of microbial activity

Performance Assessment Known Uncertain Comments

Transport Parameters 2– 2– – SeO4 z η, R, Dapp taken as for SO4 2– – SeO3 z η, R, Dapp unknown for labile selenite – – – HSe z η, R, Dapp unknown and taken equal as for I S = 10-9 – 10-7 M

Microbial transformations are vast and could account for the reduction and precipitation of selenium as colloidal elemental selenium or iron selenide, or for the methylation of selenium and its volatilisation as dimethyl selenide (DMSe) or dimethyl diselenide (DMDSe).

65/328

79Se is one of the rare radionuclides whose half-life has considerably evolved in the recent years with non-negligible consequences for safety calculations. The flux of 79Se at the Boom Clay / aquifer interface is very sensitive to the values selected for its half-life. The half-life of Se-79 has been re-estimated several times in the last 20 years and more particularly since 2001, as schematically illustrated by Figure 5.2.1. With the shortest half-life value of 65 ka, a 79 non-negligible fraction of Se could decay in the clay barrier but with the longer T½ values recently determined, it is no more the case. A difference of about 25 % still affects the two 79 more recent values of Se T½: 295 ka (Jiang Song-Sheng, 2001; Singh, 2002; Bé et al., 2005) and 377 ka (Bienvenu et al., 2007). More details on the recent redetermination of 79Se half- life are given in Appendix A1: General information on selenium.

1 200 T½ = 1 100 ka

79

T½ (ka) Evolution of Se half-life these last 20 years 800

T½ = 650 ka

Best Estimate T½ = 377 ka 400 T½ = 356 ka T½ = 295 ka T½ = 65 ka 0 1 990 1 995 2 000 2 005 2 010 Year Figure 5.2.1: Recent evolution of 79Se half-life in the last 20 years. For numerical data and discussion, see also: Appendix A1: General information on selenium.

2– Except for selenite (SeO3 ), the other anionic species of selenium are not significantly sorbed in Boom Clay.

2– Selenate (SeO4 ), which does not undergo any observable reduction in Boom Clay at the laboratory space and time scale, migrates unretarded, or very weakly retarded (R = 2.3; -1 Kd < 0.3 L kg ), in Boom Clay as evidenced by electromigration experiments (Beauwens et al., 2005) and by deduction from comparable results of percolation tests made with 34S- labelled sulfate.

66/328

The migration behaviour of selenide (HSe–) in Boom Clay has not yet been studied but selenide is presently considered unretarded (R = 1; Kd = 0) because it is a soft Lewis base deprived of oxygen ligands to form surface complexes (it is also the case for iodide (I–) another soft Lewis base). Under the reducing conditions prevailing in Boom Clay and in the 0 – presence of a controlling selenium solid phase (Se , FeSe2, FeSe) selenide (HSe ) mobility is -8 -3 only restricted by a solubility limit (S ≃ 5 × 10 mol dm ). However, this assumption is only valid if selenate effectively undergoes reduction in Boom Clay.

The behaviour of selenite in Boom Clay is complex. Selenite is an oxy-anion forming inner- sphere complexes at the surface of iron and aluminium oxy-hydroxides groups located on the edges of the clay minerals (Bar-Yosef and Meek, 1987; Bruggeman, 2006). The principal experimental difficulty encountered in sorption tests with selenite under reducing conditions 2– occurs from the fact that after its sorption on the clay minerals surface, selenite (SeO3 ) is progressively reduced into elemental selenium (Se0) or selenide (HSe–). Then selenium precipitates as Se0, or FeSe, if Fe2+ is also present in solution, or adsorbed on the minerals surface. In these conditions, with continuous changes of speciation caused by the progressive reduction of selenite, it is not easy to distinguish between sorption and precipitation.

Therefore, it is difficult to determine a distribution coefficient (Kd), or to estimate a retardation factor (R) for pure selenite. Anyway, the reduction / precipitation of selenium is favourable to its retention, but it complicates the interpretation of the system because the mechanisms of immobilisation are difficult to decipher.

2– Despite the fact that it is well evidenced for selenite (SeO3 ), reduction in Boom Clay is very 2– unlikely and not proven for selenate (SeO4 ) in the absence of microbially-mediated reduction (Oremland, 1994; Oremland, et al., 1989, 1994, 1998). So, we are presently not able to demonstrate that selenate will be reduced in compact Boom Clay, even at very long term, if the “far-from-equilibrium” system is kinetically hindered. Then the system is no longer controlled by thermodynamics but by kinetics limitations. The worse case scenario for performance assessment studies is to consider that all selenium present in the source term (spent fuel, vitrified HLW, or bituminised waste) occurs as selenate because of the oxidizing – conditions locally imposed in the waste matrix by alpha and gamma radiolysis (or by NO3 present in bituminised waste). In this case, selenium is expected to be in a chemical form totally soluble and not sorbed.

67/328

5.3. Summary of the transport parameters for the selenium species considered

Based on the experimental results and simple calculations, two sets of transport parameters are proposed in Table 5.3.1 to assess the long-term dose-to-man delivered by 79Se: 2– • a first set for selenate (SeO4 , divalent anion, no solubility limit), and; • a second set for selenide (HSe–, monovalent anion, with a solubility limit).

Table 5.3.1: Overview of the migration parameters selected for selenate and selenide in undisturbed Boom Clay. Values relevant for the Mol site under the present geochemical conditions, i.e., in the absence of perturbation. 2– SeO4 (divalent anion) Best Estimate Expert Range Source Range

2 –1 -11 -11 -11 Dapp (m s ) 3.17 (±0.98) × 10 2.2 – 4.2 × 10 1.5 – 7.3 × 10

η (—) 0.10 0.05 – 0.18 0.05 – 0.18

R (—) 2.3 1 – 5 1 – 5

Solubility (mol dm-3) unlimited unlimited unlimited

HSe– (monovalent anion) Best Estimate Expert Range Source Range

2 –1 -10 -10 -10 Dapp (m s ) 1.2 × 10 1.0 – 1.3 × 10 0.8 – 1.7 × 10

η (—) 0.16 0.14 – 0.18 0.12 – 0.18

R (—) 1 1 1

Solubility (mol dm-3) 2 × 10-8 1 × 10-9 – 1 × 10-7 1 × 10-9 – 1 × 10-7 Source Range (SR) of a parameter is a range of values outside of which the parameter value is unlikely to lie, considering our current knowledge. Expert Range (ER) of a parameter is the range of values within which experts expect the parameter value to lie. Best Estimate (BE) value of a parameter is the selected value corresponding to the best current knowledge and proposed by experts as nominal value for the performance assessment calculations.

For selenate, a consistent set of transport parameters is derived from the interpretation of migration experiments made with selenate and sulfate. For selenide, only the solubility limit was measured. In the absence of experimental data for determining its transport parameters, HSe– is tentatively considered as a species unretarded in Boom Clay. The same values than these of the transport parameters of I– in Boom Clay summarised by Bruggeman et al., (2010) are attributed to HSe– on the basis of the resemblance of their electronic structures and related chemical properties. Indeed, according to the theory of hard and soft Lewis acids and bases (HSAB) (for more details, see Sposito, 1981, p. 76), no significant interaction (i.e., sorption) is expected between strong Lewis acid sites (S–OH groups located on the clay platelets edges) and soft Lewis bases such as iodide and selenide (large and polarisable monoatomic anion with a low electronegativity).

68/328

The rationale underpinning the selection of the parameter ranges (Best Estimate – BE values; Expert Range – ER, and Source Range – SR) presented in Table 5.3.1 is summarised hereafter.

2– Selenate – SeO4 Concentration / solubility limit, S: 2– • BE/ER/SR: no solubility limiting solid phase is considered for SeO4 in Boom Clay. Retardation factor, R: • BE: from migration experiments an average ηR = 0.23 was obtained (higher than 0.16 for -11 -11 iodide suggesting a slight sorption). BE R = Dp/Dapp (7.28 × 10 / 3.17 × 10 = 2.3) -11 -1 with Dp taken equal to 7.28 × 10 m² s (cfr. Table 4.5.1, p. 59). • ER: lower limit assuming non-retarded transport; the upper limit is the calculated value taking ηR = 0.23 (diffusion experiment) and the lower limit for η = 0.05. • SR: taken equal to ER.

Diffusion accessible porosity, η: • BE: from migration experiments an average ηR = 0.23 was obtained, and BE R = 2.3, thus η = ηR/R = 0.23/2.3 = 0.1. • ER: lower limit η calculated from the Bruggeman equation (eq. 4.4.1, p. 56) using an -12 -1 – average Deff = 7 × 10 m² s . Upper limit taken equal to η value of I . The selected range encompasses the standard deviation measured for ηR in the diffusion experiments. • SR: taken equal to ER. Values relevant for the Mol site under the present geochemical conditions, i.e., in the absence of perturbation. No effect of ionic strength is taken into account for the diffusion accessible porosity of anions.

Apparent diffusion coefficient, Dapp: • BE: average of the apparent diffusion coefficients values determined from the migration 2– 2– experiments made with SeO4 and SO4 . • ER: BE value ± the standard deviation calculated from the experimentally determined values.

• SR: since Dapp = Dp/R, the variation in Dapp must reflect the variation in R. Therefore SR upper and lower limits are adjusted with respect to BE to account for R range of 1–5 (with BE R = 2.3).

Selenide – HSe– Concentration / solubility limit, S: • BE: the thermodynamic solubility calculated by AEA Technology for FeSe(s).

69/328

• ER: lower and upper limit experimentally measured values considered to correspond to Se(-II) solubility. • SR: taken equal to ER. Retardation factor, R: • BE/ER/SR: assuming non-retarded transport, with selenide similar to iodide.

Diffusion accessible porosity, η: • BE: same as for I–. • ER/SR: taken identical to I–. SR also taken equal to ER. Values relevant for the Mol site under the present geochemical conditions, i.e., in the absence of perturbation. No effect of ionic strength is taken into account for the diffusion accessible porosity of anions.

Apparent diffusion coefficient, Dapp: – – • BE/ER/SR: for HSe taken equal to these of I , but corrected for the difference in Daq. – – – Ratio Daq(HSe )/Daq(I ) = 0.83. For more information on I transport parameters in Boom Clay, see Bruggeman et al. (2010).

For the sake of completeness, Table 5.3.2 gives an overview of the evolution of the 79Se transport parameters that have been selected for safety assessment calculations since the first PAGIS (1986) study. For more information on the changes of these parameters previously, see also Appendix A9: Selenium migration behaviour in Boom Clay.

Table 5.3.2: Evolution of 79Se transport parameters used for performance assessment calculations in Boom Clay. Project / PAGIS PACOMA EVEREST SPA SAFIR 2 Study & UPD 90 DCF: DS1 DCF: DS2 (Year) (1986) (1989-1990) (1994) (1998) (1999)

2 -1 -10 -10 -10 -10 -10 Dp (m s ) 3.20 × 10 2.00 × 10 2.00 × 10 2.00 × 10 2.00 × 10

R (—) 320 10 10 300 1

η (—) 0.33 0.10 0.10 0.10 0.13

S (mol dm–3) — — 1.00 × 10-8 1.50 × 10-8 5.50 × 10-8

2 -1 -12 -11 -11 -13 -10 Dapp (m s ) 1.00 × 10 2.00 × 10 2.00 × 10 6.67 × 10 2.00 × 10

ηR (—) 105.6 1 1 30 0.13

2 -1 -10 -11 -11 -11 -11 Deff (m s ) 1.06 × 10 2.00 × 10 2.00 × 10 2.00 × 10 2.60 × 10 UPD 90: Updating 90; DCF: Data Collection (Form) exercise; DS: Data Set.

70/328

5.4. Recommendations and future works

Several unresolved issues remain to be addressed, more particularly by order of decreasing importance: – What is the selenium speciation in the source term (spent fuel and vitrified HLW)? – The accurate reappraisal of the selenide and selenate transport parameters in Boom Clay by means of better controlled migration experiments with 75Se sources already at chemical equilibrium prior to start the tests, i.e. without labile selenite, and with sources of pure 75 - 75 2– H Se and SeO4 species; – And finally, a less crucial question: what is the role of solid carbonates (aragonite, calcite, siderite, …) as possible overlooked sinks of natural selenium in Boom Clay?

These questions are developed more in detail hereafter.

1. Selenium speciation in the source term is important to determine the fraction of selenium initially present as mobile selenate recalcitrant to reduction in Boom Clay. The experimental study of this question is difficult because a relative small activity of 79Se is mixed with large amounts of other fission products and minor actinides in the spent fuel

UO2 matrix (Comte, 2001; Comte et al., 2000; 2001; 2002a,b; 2003). The speciation of selenium in the high-activity vitrified waste could likely be more easily understood thanks to the extensive knowledge existing in the glass manufacturing industry which uses selenium as an additive to control the glass colour. In this perspective, the real redox conditions prevailing in the glass matrix should be assessed at least by a review of the literature and of the conditions in which the high-level waste is vitrified (what is the effect of sugar added as a reductant to the liquid waste before the calcination and vitrification 2– processes to minimize the ruthenium volatility ?). Under what chemical form (SeO4 , 2– 0 2– SeO3 , Se , or Se ) is selenium present in the spent fuel and in the glass matrix ?

2. Proposal of new migration experiments avoiding the kinetically controlled sorption / reduction of selenite present in sources not in chemical equilibrium prior to start the tests. In the measure of the possible, it would be advisable to perform: – Migration experiments with a pure H75Se- source at low concentration (below the 0 solubility limit of Se , FeSe2, and FeSe) to determine the migration parameters of reduced selenium forms, and; 75 2– – Migration experiments with a pure SeO4 source (i.e., without selenite !) to confirm the values of accessible porosity (η) and retardation factor (R) presently selected for selenate but only derived from a deductive reasoning based on the results obtained from migration experiments with sulfate.

71/328

3. Under what chemical form is natural selenium present in Boom Clay ? Is selenium only associated with pyrite (in solid solution) and organic matter (as degradation products of seleno-proteins present in the kerogen), or does it exist other sources of natural selenium

in Boom Clay: e.g., also in the CaCO3 deposited by coccolithophores ? Is natural selenium also associated with illite and smectite clay minerals as oxyanions ? To answer to these questions, it could be useful to perform analyses of natural selenium in Boom Clay porewater and clay samples along with routine measurements already carried out for other trace elements. It would also be worth to determine the natural selenium concentration in the purified clay fractions, in pyrite, in organic matter (both solid and dissolved), and in carbonate fractions isolated from Boom Clay. So, it could be useful to apply to Boom Clay the selenium sequential extraction methods developed by Kulp and Pratt (2004) and Oram et al., (2008). Finally, the study of the vertical distribution profile of natural selenium in the Boom Clay formation could allow to identify a possible correlation with

kerogen or CaCO3.

Finally, a particular attention should be paid in the future to any modifications regarding the half-life of selenium-79. In the same way, the values of the half-life used for 79Se should always be systematically mentioned along with the results of all the future safety calculations.

72/328

6. Acknowledgments

73/328

74/328

6. Acknowledgments

This work is undertaken in close co-operation with, and with the financial support of, ONDRAF/NIRAS, the Belgian Agency for Radioactive Waste and Enriched Fissile Materials. Parts of the research presented here have also been funded by the European Commission in the framework of the TRANCOM-I (Contract N° FI4W-CT95-0013) and TRANCOM-II (Contract N° FIKW-CT-2000-00008) projects. Lian Wang, Jan Marivoet, Xavier Sillen and Eef Weetjens are acknowledged for their contributions to the safety calculations, and for their numerous exchanges of ideas on the diffusion of selenium in clay. The support and the fruitful discussions with Cherry Tweed, Nick Pilkington, Robert Gens, and Ann Dierckx have been highly appreciated. Marc Aertsens made the modelling of the Se diffusion experiments, while Véra Pirlet provided data on the leaching of Se from glass exposed to conditions expected in the near-field. Hugo Moors, Marc Van Gompel, Louis Van Ravestyn, and Jacqueline Van Cluysen are also gratefully acknowledged for their assistance during the experiments.

We are also grateful to our colleagues Achim Albrecht, Hélène Pauwels and Elie Valcke for the many informal scientific exchanges on the question of the reduction of selenate and other redox-sensitive elements in the frame of the Mont Terri Bitumen Nitrate (BN) experiment.

Finally, we wish also to thank Pierre Henrion and Martin Put, our predecessors in charge of the diffusion experiments with Boom Clay, for their pioneer contributions to the Belgian Geological Waste Disposal project since its start more than three decades ago.

75/328

76/328

7. References

77/328

78/328

7. References

Abrams M.M., Burau R.G., and Zasoski R.J. (1990) Organic selenium distribution in selected California soils. Soil Science Society of America Journal 54(4), 979–982.

Aertsens M. (2000) Results from modelling the selenium diffusion experiments. Unpublished internal SCK•CEN note. 5 pp. Two annexes. Annex 1: The model Dfit38_prec. Annex 2: The model Dissflow. Personal Communication. .

Aertsens M., Put M., and Dierckx A. (1999) An analytical model for pulse injection experiments. In Proceedings of the Conference “Modelling of transport processes in soils at various scales in time and space” held in Leuven, Belgium, 24-26 November 1999, Leuven. 67–76.

Aertsens M., Wemaere I., and Wouters L. (2004) Spatial variability of transport parameters in the Boom Clay. Applied Clay Science 26, 37–45.

Aertsens M., Dierckx A., Put M., Moors H., Janssen K., Van Ravestyn L., Van Gompel Mc., and De Cannière P. (2005a) Determination of the hydraulic conductivity, the product ηR of the porosity η and the retardation factor R, and the apparent diffusion coefficient Dp on Boom Clay cores from the Mol-1 drilling. Final report for Task 2-7 for NIRAS/ONDRAF, contract N° CCHO-98/332; SCK•CEN contract N° KNT 90 98 1042. SCK•CEN report R-3503. Mol, Belgium. 22 pp. without appendices.

Aertsens M., Dierckx A., Put M., Moors H., Janssen K., Van Ravestyn L., Van Gompel Mc., Van Gompel Ma., and De Cannière P. (2005b) Determination of the hydraulic conductivity, ηR and the apparent diffusion coefficient on Ieper Clay and Boom Clay cores from the Doel-1 and Doel-2b drillings. Final report for Task 2-71 and 2-73 covering the period 1998-1999 for NIRAS/ONDRAF, contract N° CCHO-98/332; SCK•CEN contract N° KNT 90 98 1042. SCK•CEN report R-3589. Mol, Belgium. 18 pp. without appendices.

Aguerre S. and Frechou C. (2006) Development of a radiochemical separation for selenium with the aim of measuring its isotope 79 in low and intermediate nuclear wastes by ICP-MS. Talanta 69(3), 565–571. doi:10.1016/j.talanta.2005.10.028.

Akira Kitamure, Masahiro Shibata, and Hideo Kitao (2004) Solubility measurement of iron- selenium compounds under reducing conditions. Mat. Res. Soc. Symp. Proc. (MRS) 807, 609– 614.

79/328

Albrecht Achim (2005-2009) Personal communication. Selenate reduction could be hindered by the presence of nitrate salts conditioned in the bituminised waste (Cogema, Eurobitum). Contact with Oremland. Informal scientific exchanges in the frame of the Mont Terri Bitumen Nitrate (BN) experiment. See the Mont Terri Technical Discussion protocols and the mission reports of Andra – SCK•CEN exchange meetings.

Albrecht A. (2008) The oxidation-reduction potential in the near-field of a high level waste disposal site and the possible impact on Se speciation and migration. Presentation. Bioprota International Forum on Se-79 in the Biosphere, 5 and 6 May 2008, Nagra Wettingen (CH). In: Smith K. (Ed.) (2008) Report of Se-79 in the Biosphere. BIOPROTA Workshop, hosted by Nagra, Wettingen, 5–6 May 2008, V2.0, Final. BIOPROTA – http://www.bioprota.com.

Amouroux D., Liss P.S., Tessier E., Hamren-Larsson, M. and Donard, O.F.X. (2001) Role of oceans as biogenic sources of selenium. Earth and Planetary Science Letters 189 (3-4), 277– 283.

Anderson D.R. (1993) Use of conceptual models in performance assessments of repositories for disposal of high-level and transuranic radioactive waste. pp. 139–143 in: OECD (1993) The role of conceptual models in demonstrating repository post-closure safety. Proceedings of a NEA Workshop, Paris, 16-18 November 1993. Nuclear Energy Agency (NEA). OECD Documents. 190 pp. ISBN 92-64-14429-3.

Andersson K. and Allard B. (1983) Sorption of radionuclides on geologic media: a literature survey. Chalmers University of Technology, Göteborg, KBS Report TR 83-07.

Andra (2005a) Dossier 2005 Argile – Tome Evolution phénoménologique du stockage géologique. Rapport Andra n° C.RP.ADP.04.0025B. Collection « Les rapports », Andra (France), 520 pp. Voir: sélénium, p. 449, Encadré 10.7: La solubilité et la sorption des radionucléides dans les argiles gonflantes; Voir: sélénium, p. 453, Encadré 10.9: La solubilité et la sorption des radionucléides dans les matériaux cimentaires; Voir: sélénium, p. 463, Encadré 10.11: La solubilité et la sorption des radionucléides dans les argilites du Callovo-Oxfordien.

Andra (2005b) Dossier 2005 Argile – Tome Référentiel du site Meuse/Haute-Marne – Tome 1. Collection « Les rapports », Andra (France), 714 pp.

Andra (2005c) Dossier 2005 Argile – Tome évaluation de sûreté du stockage géologique. Collection « Les rapports », Andra (France), 737 pp. Chap 5 – Evaluation des performances à

80/328

long terme du stockage. Voir: 79Se, p. 231, Tableau 5.3-5: Valeurs des paramètres de transport et de rétention chimique dans le béton des colis de déchets B – calcul de référence. Voir: 79Se, p. 238, Tableau 5.3-10: Valeurs des paramètres hydrauliques, de transport et de rétention chimique dans les zones fracturées et microfissurées – calcul de référence.

Andra (2005d) Dossier 2005 – Référentiel de connaissance et modèle d'inventaire des colis de déchets à haute activité et à vie longue; Rapport Andra n° C.RP.AHVL.04.0006/A.

Andra (2005e) Dossier 2005 – Référentiel "Comportement des radionucléides et des toxiques chimiques d'un stockage dans le Callovo-Oxfordien jusqu'à l'homme" – Site de Meuse/Haute Marne; Rapport Andra n° C.RP.ASTR.04.0032.B.

Arai Y. and Sparks D.L. (2001) ATR–FTIR spectroscopic investigation on phosphate adsorption mechanisms at the ferrihydrite–water interface. Journal of Colloid and Interface Science 241(2), 317–326.

Araie H., Obata T., and Shiraiwa Y. (2003) Metabolism of selenium in a coccolithophorid, Emiliania Huxleyi. J Plant Res. 116, 119.

Atkins P.W (1982) Physical Chemistry. Oxford University Press. Second edition.

ATSDR (2003) Toxicological profile for selenium (update September 2003). Selenium CAS # 7782-49-2. Agency for Toxic Substances and Disease Registry (ATSDR). Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service. ToxFAQs address: http://www.atsdr.cdc.gov/toxfaq.html.

Avoscan L., Carrière M., Jehanneuf F., Collins R., Carrot F., Covès J., Gouget B. (2005) Ralstonia metallidurans CH34 resistance to selenium oxyanions: Growth kinetics, bioaccumulation and reduction. Human and Environmental Toxicology. Pierre Süe Laboratory UMR 9956 CEA-CNRS, France.

Baes C.F. and Mesmer R.E. (1976) The hydrolysis of cations. 489 pp. Wiley Inc. New York.

Baeyens B. (1982) Strontium, cesium and europium retention in Boom Clay: a potential repository site for nuclear waste. Ph.D. Thesis, Katholieke Universiteit Leuven, Belgium.

Baeyens B., Maes A., Cremers A., and Henrion P.N. (1985a) In situ physico-chemical characterization of Boom Clay. Radioactive Waste Management and the Nuclear Fuel Cycle 6(3-4), 391–408.

81/328

Baeyens B., Maes A., Cremers A., and Henrion P.N. (1985b) Aging effects in Boom Clay. Radioactive Waste Management and the Nuclear Fuel Cycle 6(3-4), 409–423.

Baeyens B. and Bradbury M. (1991) A physico-chemical characterisation technique for determining the porewater chemistry in argillaceous rocks. PSI-Bericht Nr. 103. Paul Scherrer Institut, Villigen.

Bradbury M.H. and Baeyens B. (1997a) Derivation of in situ Opalinus Clay porewater compositions from experimental and geochemical modelling studies. PSI-Bericht Nr. 97-14. Paul Scherrer Institut, Villigen.

Baeyens B. and Bradbury M.H. (1997b) A mechanistic description of Ni and Zn sorption on Na-montmorillonite. Part I. Titration and sorption measurements. Journal of Contaminant Hydrology 27, 199–222.

Bradbury M.H. and Baeyens B. (1998) A physicochemical characterisation and geochemical modelling approach for determining porewater chemistries in argillaceous rocks. Geochimica et Cosmochimica Acta 62(5), 783–795.

Baeyens B. and Bradbury M.H. (2004) Cation exchange capacity measurements on illite using the sodium and cesium isotope dilution technique: Effects of the index cation, electrolyte concentration and competition modelling. Clays and Clay Minerals 52(4), 421– 431.

Bagnall K.W. (1966) The chemistry of selenium, tellurium, and polonium. 200 pp. Elsevier Amsterdam, New York. Topics in inorganic and general chemistry. Monograph 7 (with 11 illustration and 25 tables).

Baker S., Baston G.M.N., Green A., Heath T.G., Ilett D.J., Pilkington N.J., Tweed C.J., and Williams S.J. (1997) Radionuclide solubilities in Boom Clay – 2nd Progress Report, Issue 1 (September 1997).

Baker S., Baston G.M.N., Green A., Heath T.G., Pilkington N.J., Tweed C.J., and Williams S.J. (1998) Radionuclide solubilities in Boom Clay – Task 2 and Task 3 Report. AEA Technology Report AEAT-3313, March 1998.

Baker S., Baston G.M.N., Boult K.A., Brownsword M., Ilett D.J., Maning M.C., Pilkington N.J., Tweed C.J., and Williams S.J. (2000) Radionuclide solubilities in Boom Clay – Phase II, Part 1: 2nd Interim Report, AEAT/R/NS/0055 Issue 1 (February 2000).

82/328

Baker S., Baston G.M.N., Boult K.A., Brownsword M., Cowper M., Green A., Heath T., Ilett D., Manning M., Pilkington N., Tweed C., and Williams S. (2000) Radionuclide solubilities in Boom Clay. Final Report, Part 2 (AEA Technology, July 2000).

Baker S., Cowper M.M., Maning M.C., and Williams S.J. (2002) Plutonium and selenium studies – First Progress Report, AEAT/R/NS/0553 Issue 1 (March 2002).

Balistrieri L.S. and Chao T.T. (1987) Selenium adsorption by goethite. Soil Science Society of America Journal 51(5), 1145–1151.

Balistrieri L.S. and Chao T.T. (1990) Adsorption of selenium by amorphous iron oxyhydroxide and manganese-dioxide. Geochimica et Cosmochimica Acta 54(3), 739–751.

Bañuelos G., Terry N., LeDuc D.L., Pilon-Smits E.A.H., and Mackey B. (2005) Field trial of transgenic Indian mustard plants shows enhanced phytoremediation of selenium-contaminated sediment. Environment Science and Technology, online publication, ASAP: doi:10.1021/es049035f.

Bar-Yosef B. (1987) Selenium desorption from C-kaolinite. Communications in Soil Science and Plant Analysis. 18(4), 771–779. ISSN 0010-3624.

Bar-Yosef B. and Meek D. (1987) Selenium sorption by kaolinite and montmorillonite. Soil Science 144(1), 11–19. ISSN 0038-075X.

Bar-Yosef B., Afik I., and Rosenberg R. (1988) Fluoride sorption by montmorillonite and kaolinite. Soil Science 145(3), 194–200. ISSN 0038-075X.

Bar-Yosef B and Kafkafi U. (1978) Phosphate desorption from kaolinite suspensions. Soil Science Society of America Journal 42(4), 570–574.

Bar-Yosef B, Kafkafi U, Rosenberg R, and Sposito G. (1988) Phosphorus adsorption by kaolinite and montmorillonite: I. effect of time, ionic strength, and pH. Soil Science Society of America Journal SSSJD4 52(6), 1580–1585. November-December 1988. 6 Fig., 1 Table, 26 ref. US-Israel Binational Agricultural R & D Grant I-401-82.

Baur Isabel Keller Regula (2002) The immobilisation of heavy metals and metalloids in 2– 2– cement stabilised wastes: a study focussing on the selenium oxyanions SeO3 and SeO4 . A dissertation submitted to the Swiss Federal Institute of Technology Zürich for the degree of Doctor of Natural Sciences. 112. Dissertation ETH-Zürich N°. 14 840. Zürich 2002.

83/328

Baur I. and Johnson C.A. (2003a) The solubility of selenate-AFt

(3 CaO · Al2O3 · 3 CaSeO4 · 37.5 H2O) and selenate-AFm (3 CaO · Al2O3 · CaSeO4 · x H2O). Cement and Concrete Research 33(11), 1741–1748.

Baur I. and Johnson C.A. (2003b) Sorption of selenite and selenate to cement minerals. Environmental Science and Technology 37(15), 3442–3447.

Bazer-Bachi F., Tevissen E., Descostes M., Grenut B., Meier P., Simonnot M.-O., and Sardin M. (2006) Characterization of iodide retention on Callovo-Oxfordian argillites and its influence on iodide migration. MIGRATION 2005, The 10th International Conference on the Chemistry and Migration of Actinides and Fission Products in the Geosphere. Physics and Chemistry of the Earth, Parts A/B/C 31(10-14), 517-522. doi:10.1016/j.pce.2006.04.015.

Bé M.-M., Chisté V., and Dulieu C. (2005) Périodes radioactives: Table de valeurs recommandées. Half-lives: Table of recommended values. CEA, Laboratoire National Henri 79 Becquerel. Note Technique DETECS/LNHB/2005-8. Edition Fév. 2005. See T½ of Se = 2.95 × 105 anno on p. 3.

Bé M.-M., Chisté V., and Dulieu C. (2006) Périodes radioactives: Table de valeurs recommandées. Half-lives: Table of recommended values. CEA, Laboratoire National Henri Becquerel. Note Technique DETECS/LNHB/2006-58. Edition Oct. 2006. 79 5 See T½ of Se = 3.56 × 10 anno on p. 3.

Beaucaire C., Pitsch H., Toulhoat P., Mottellier S., and Louvat D. (2000) Regional fluid characterisation and modelling of water-rock equilibria in the Boom Clay Formation and in the Rupelian aquifer at Mol, Belgium. Applied Geochemistry 15, 667–686.

Beauwens T., De Cannière P., Moors H., Wang L., and Maes N. (2005) Studying the migration behaviour of selenate in Boom Clay by electromigration. Engineering Geology 77, 285–293. doi:10.1016/j.enggeo.2004.07.019 (paper presented at the EREM’03 Conference).

Bell M.J. (1973) ORIGEN – The ORNL isotope generation and depletion code. Oak Ridge National Laboratory. Report ORNL-4623.

Belzile N., Chen Y.W., and Xu R.R. (2000) Early diagenetic behaviour of selenium in freshwater sediments. Applied Geochemistry 15(10), 1439–1454.

84/328

Benjamin M.M. and Bloom N.S. (1981) Effects of strong binding of anionic adsorbates on adsorption of trace metals on amorphous iron oxyhydroxide. In: Adsorption from aqueous solution. Tewari P.H. (ed.) (1981) Plenum Press, New York, pp. 41–60.

Berner U. (2002) Project Opalinus Clay: radionuclides limits in the near-field of a repository for spent fuel and vitrified HLW. PSI Report (PSI Bericht Nr 02-22, ISSN 1019-0643) (December 2002).

Berner U. (2002) In: The use of thermodynamic databases in performance assessment, Workshop Proceedings, OECD-NEA, Paris 2002, pp. 139–149.

Bertolino F.A., Torriero A.A.J., Salinas E., Olsina R., Martinez L.D., Raba J. (2006) Speciation analysis of selenium in natural water using square-wave voltammetry after preconcentration on activated carbon. Analytica Acta.

Bethke C.M. (2000) The Geochemist’s Workbench. A user’s Guide to Rxn, Act 2, Tact, React and Gtplot. Release 3.1. University of Illinois.

Bienvenu P., Cassette P., Andreoletti G., Bé M., Comte J., and Lépy M. (2007) A new determination of 79Se half-life. Applied Radiation and Isotopes 65 (3), 355-364. doi:10.1016/j.apradiso.2006.09.009.

Bienvenu P., Ferreux L., Andreoletti G., Arnal N., Lépy M., and Bé M. (2009) Determination of 126Sn half-life from ICP-MS and gamma spectrometry measurements. Radiochimica Acta 97 (12), 687–694. doi:10.1524/ract.2009.1673.

BIOPROTA (2008) International Forum on Se-79 in the Biosphere, 5 and 6 May 2008. Hosted by Nagra, Wettingen, Switzerland. http://www.bioprota.com (short citation).

BIOPROTA (2008) Key issues in biosphere aspects of assessment of the long-term impact of contaminant releases associated with radioactive waste management. Report of an International Forum on Se-79 in the Biosphere. 5–6 May 2008. Hosted by Nagra, Wettingen, Switzerland. Editor: K. Smith (2008) Version 2.0, Final, June 2008. http://www.bioprota.com. (full citation).

Blaylock M.J. and James B.R. (1993) Selenite and selenate quantification by hydride generation atomic spectrometry, ion chromatography, and colorimetry. Journal of Environmental Quality 22(4), 851–857.

85/328

Blaylock M.J., Tawfic T.A., and Vance G.F. (1995) Modeling selenite sorption in reclaimed coal mine soil materials. Soil Science 159 (1), 43–48.

Bolan N.S., Syers J.K., and Tillman R.W. (1986) Ionic strength effects on surface charge and adsorption of phosphate and sulphate by soils. European Journal of Soil Science 37(3), 379– 388. doi:10.1111/j.1365-2389.1986.tb00371.x.

Bond K.A., Heath T.G., and Tweed C.J. (1997) HATCHES a referenced thermodynamic database for chemical equilibrium studies. Nirex Report NSS/R379.

Bonhoure I., Baur I., Wieland E., Johnson C.A., and Scheidegger A.M. (2006) Uptake of Se(IV/VI) oxyanions by hardened cement paste and cement minerals: An X-ray absorption spectroscopy study. Cement and Concrete Research 36 (1), 91–98.

Bostick B.C. and Fendorf S. (2003) Arsenite sorption on troilite (FeS) and pyrite (FeS2), Geochimica et Cosmochimica Acta, 67(5), 909–921.

Boulanger D. (2006) Characterization of irradiated fuel assemblies. Part 2:Technical Note. Report BelgoNucléaire, BN ref. 0204246 / 221 – “A”.

Boult K.A., Cowper M.M., Heath T.G., Sato H., Shibutani T., and Yui M. (1998) Towards an understanding of the sorption of U(VI) and Se(IV) on sodium bentonite. Journal of Contaminant Hydrology 35 (1-3), 141–150.

Boyle-Wight E.J., Katz L.E., and Hayes K.F. (2002) Spectroscopic studies of the effects of selenate and selenite on cobalt sorption to gamma-Al2O3, Environmental Science & Technology 36, 1219–1225.

Bradbury M.H. and Baeyens B. (1997) A mechanistic description of Ni and Zn sorption on Na-montmorillonite. Part II: modelling, Journal of Contaminant Hydrology 27, 223–248.

Bradbury M.H. and Baeyens B. (2005) Modelling titration data and the sorption of Sr(II), Ni(II), Eu(III) and U(VI) on Na-illite. 71 pp. PSI Nuclear Energy and Safety Research Department – Laboratory for Waste Management, Villigen, Switzerland. ISSN 1019-0643.

Brennetot R., Pierry L., Atamyan T., Favre G., and Vailhen D. (2008) Optimisation of the operating conditions of a quadrupole ICP-MS with hexapole collision/reaction cell for the analysis of selenium-79 in spent nuclear fuel using experimental designs. Journal of Analytical Atomic Spectrometry 23 (10), 1350-1358. Retrieved January 8, 2010, from doi:10.1039/b802820f.

86/328

Brook L.S. (1952) The vapour pressures of tellurium and selenium. Journal of the American Chemical Society 74, 227–229.

Brookins D.G. (1988) Eh – pH diagrams for geochemistry. Springer-Verlag. Berlin Heidelberg New York London Paris Tokyo. 176 pp., 98 Figures and 61 Tables. For selenium, see, Fig. 4: Eh – pH diagram for part of the system Se-O-H, pp. 18 and 19.

Bros R., Hiroshi H., Gento K., and Toshihiko O. (2003) Mobilization and mechanisms of retardation in the Oklo natural reactor zone 2 (Gabon) – inferences from U, REE, Zr, Mo and Se isotopes. Applied Geochemistry 18 (12), 1807–1824. doi:10.1016/S0883-2927(03)00113- 6.

Brower F.M. and Graham E.L. (1958) Some chemical reactions of Colorado oil shale kerogen. Industrial and Engineering Chemistry 50 (7), 1059–1060. doi:10.1021/ie50583a043.

Brown G.E. Jr., Catalano J.G., Templeton A.S., Trainor T.P., Farges F., Bostick B.C., Kendelewicz T., Doyle C.S., Spormann A.M., Revill K., Morin G., Juillot F., and Calas G. (2005) Environmental interfaces, heavy metals, microbes, and plants: applications of XAFS spectroscopy and related synchrotron radiation methods to environmental science. Physica Scripta 115, 80–87. doi:10.1238/physica.topical.115a00080.

Bruggeman C., Vancluysen J., and Maes A. (2002) New selenium solution speciation method by ion chromatography plus gamma counting and its application to FeS2–controlled reducing conditions. Radiochimica Acta 90(9-11), 629–635.

Bruggeman C., Maes A., Vancluysen J., and Vandemussele P. (2004) Selenite reduction in

Boom Clay: effect of FeS2, clay minerals and dissolved organic matter. Annex 8 to WP 3, pp. 18, of the extended Final Scientific and Technical Report of the EC Trancom-II Project. In: Maes N. coordinator et al. (2004e,f) Migration case study: transport of radionuclides in a reducing clay sediment (TRANCOM-II). SCK•CEN-BLG-988 (04/NMa/P-50).

Bruggeman C., Maes A., Vancluysen J., and Vandemussele P. (2005) Selenite reduction in

Boom Clay: effect of FeS2, clay minerals and dissolved organic matter. Environmental Pollution 137(2), 209–221. URL: http://www.sciencedirect.com/science/journal/02697491. DOI: doi:10.1016/j.envpol.2005.02.010.

Bruggeman C. (2006) Assessment of the geochemical behaviour of selenium oxyanions under Boom Clay geochemical conditions. 118 pp. PhD Thesis N° 733 (2006). Dissertationes de Agricultura. Katholieke Universiteit Leuven, Faculteit Bio-ingenieurswetenschappen,

87/328

Department Microbiële en Moleculaire Systemen, Centrum voor Oppervlaktechemie en Katalyse.

Bruggeman C. and Maes A. (2007) Outlines of selenium redox geochemistry. Mobile fission and activation products in nuclear waste disposal. International workshop. L’Hermitage, La Baule – France, January 16-19, 2007. http://mofap07.in2p3.fr/. Proceedings to appear as a Nuclear Energy Agency (NEA, OECD) publication.

Bruggeman C., Aertsens M., Maes N., and Salah S. (2010) Iodine retention and migration behaviour in Boom Clay – Topical report, First Full Draft. External Report SCK•CEN- ER119, SCK•CEN, Mol, Belgium. PDF available at http://publications.sckcen.be/dspace/ http://publications.sckcen.be/dspace/simple-search

Bruggeman C., Maes N., Aertsens M., and De Cannière P. (2009) Tritiated water retention and migration behaviour in Boom Clay – SFC1 level 5 report: First Full Draft, ONDRAF/NIRAS report NIROND-TR-2009-16.

Bruggeman C., Maes N., Martens E., Govaerts J., Jacops E., Van Gompel M., and Van Ravestyn L. (2010) Technetium retention and migration behaviour in Boom Clay – Topical report, First Full Draft. External Report SCK•CEN-ER101, SCK•CEN, Mol, Belgium. PDF available at http://publications.sckcen.be/dspace/ http://publications.sckcen.be/dspace/simple-search

Bruggeman C., Salah S., Maes N., Wang L., Dierckx A., and Ochs M. (2008) Outline of the experimental approach adopted by SCK•CEN for developing radionuclide sorption parameters. Mol, Belgium: SCK•CEN, 18 pp. External Report of the Belgian Nuclear Research Centre; ER-73; CCHO 2004-2470/00/00, DS251-A44/2.1. – ISSN 1782-2335. http://publications.sckcen.be/dspace/handle/10038/908 or, http://hdl.handle.net/10038/908

Bruno J., Bosbach D., Kulik, D., and Navrotsky A. (2006) Preparation of guidelines for the evaluation of thermodynamic data for solid solutions: A state-of-the-art report. In preparation, 264 pp.

Bueno M. and Potin-Gautier M. (2002) Solid–Phase extraction for the simultaneous preconcentration of organic (seleno-cysteine) and inorganic [Se(IV), Se(VI)] selenium in natural waters. Journal of Chromatography A 963, 185–193.

Buerge-Weirich D., Behra P., and Sigg L. (2003) Adsorption of copper, nickel, and cadmium on goethite in the presence of organic ligands. Aquatic Geochemistry 9, 65–85.

88/328

Burns P.C. and Finch R. (eds.) (1999) Uranium: mineralogy, geochemistry and the environment. Reviews in Mineralogy, volume 38. 679 pp. ISBN 0-939950-50-2. Series editor: Paul H. Ribbe. Mineralogical Society of America. Washington DC. ISSN 0275-0279. See uranyl selenites, Table 17, p. 136–137. See also piretite reported by Vochten et al. (1996).

Cadelli N., Cottone G., Bertozzi G., and Girardi F. (1984) PAGIS, summary report of phase 1: a common methodological approach based on European data and models. EC, Luxembourg, Report EUR 9220 EN.

Cai Y. and Braids O.C. (Editor) (2004) Biogeochemistry of environmentally important trace elements. 452 pp. (ACS Symposium Series). American Chemical Society. ISBN: 0841238057.

Carignan J. and Wen H. (2007) Scaling NIST SRM 3149 for Se isotope analysis and isotopic variations of natural samples. Chemical Geology 242 (3-4), 347–350.

Carignan J. (2009) Personal webpage with publications of Jean Carignan on selenium geochemistry. CNRS Research Engineer, CRPG, Nancy University. Retrieved on January 5, 2010, from http://www.crpg.cnrs-nancy.fr/Equipes/FR/carignan.html.

Carvalho K.M. and Martin D.F. (2001) Removal of aqueous selenium by four aquatic plants. Journal of Aquatic Plant Management 39, 33–36.

Cassette P., Chartier F., Isnard H., Fréchou C., Laszak I., Degros J., et al. (2010) Determination of 93Zr decay scheme and half-life. Applied Radiation and Isotopes 68 (1), 122–130. doi:10.1016/j.apradiso.2009.08.011.

Catalano J.G., Zhang Z., Fenter P., and Bedzyk M.J. (2006) Inner-sphere adsorption geometry of Se(IV) at the hematite (100)-water interface. Journal of Colloid and Interface Science 297(2), 665–671.

CEA (2006) Le Journal du Centre CEA de Saclay, 2eme Trimestre 2006. N° 32. Dossier Le 79 Temps, Mesurer des durées. See T½ of Se on p. 5.

Chabroullet Christophe (2005) Étude de la remobilisation d’éléments traces à partir d'un sol de surface contaminé: Influence du vieillissement des composés organiques du sol sur la remobilisation du sélénium. Journées des thèses IRSN, 19-21 septembre 2005. Début de thèse: 20 Octobre 2003 (2ème année de thèse). Ecole doctorale Terre Univers Environnement (TUE); Université Joseph Fourier, Grenoble. Directeur de Thèse: Jean-Paul Gaudet LTHE

89/328

(Université Joseph Fourier, Grenoble). Responsable de Thèse IRSN/DEI/SECRE/LRE: Arnaud Martin-Garin.

Chao, T.T., and Sanzolone, R.F. (1989) Fractionation of soil selenium by sequential partial dissolution. Soil Sci. Soc. Am. J. 53, 385–392.

Charlet L., Scheinost A.C., Tournassat C., Greneche J.M., Géhin A., Fernández-Martinez A., Coudert S., Tisserand D., and Brendle J. (2007) Electron transfer at the mineral/water interface: Selenium reduction by ferrous iron sorbed on clay. Geochimica et Cosmochimica Acta 71 (23), 5731–5749. doi:10.1016/j.gca.2007.08.024.

Charlson R.J., Lovelock J.E., Andreae M.O., and Warren S.G. (1987) Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature 326, 655–661. doi:10.1038/326655a0.

Chaves L.H.G. (2005) The role of green rust in the environment: a review. Rev. bras. eng. agríc. ambient. [online], 9(2), 284–288. Available from: ISSN 1415-4366. doi:10.1590/S1415-43662005000200021. Accessed on: 24 Nov 2006.

Chen F., Burns P.C., and Ewing R.C. (1999) 79Se: Geochemical and crystallo-chemical retardation mechanisms. Journal of Nuclear Materials 275, 81–94.

Chen F., Burns P.C., and Ewing R.C. (2000) Near-field behavior of 99Tc during the oxidative alteration of spent nuclear fuel. Journal of Nuclear Materials 278, 225–232.

Christl M., Wacker L., Lippold J., Synal H., and Suter M. (2007) Beam Interactions with Materials and Atoms: Protactinium-231: A new radionuclide for AMS. Nuclear Instruments and Methods in Physics Research Section B 262 (2), 379–384. doi:10.1016/j.nimb.2007.05.017.

Christensen B.T., Bertelsen F., and Gissel-Nielsen G. (1989) Selenite fixation by soil particle-size separates. Journal of Soil Science 40, 641–647.

Claret F., Lerouge C., Laurioux T., Bizi M., Conte T., Ghestem JP., Wille G., Sato T., Gaucher E.C., Giffaut E., Tournassat C. (2010) Natural iodine in a clay formation: implications for iodine fate in geological disposals. Geochimica et Cosmochimica Acta 74, 16–29.

Coget, Fernand (1966) Etude des précipités anodiques en électro-raffinage du cuivre. Université Catholique de Louvain (UCL). SC/Laboratoire de Chimie Minérale. Thèse de

90/328

doctorat. Promoteur R. Breckpot. Communication personnelle: les boues anodiques contiennent du sélénium.

Combs G.F, Jr. and Gray W.P. (1998) Chemopreventive agents: Selenium. Pharmacol. Ther. 79, 179–92.

Comte J., Brochard E., Bienvenu P., Genet M. (2000) Dosage du sélénium et de l'étain par ICP/MS. (DCC/DESD/SCCD/LARC, CEA) Congrès/colloque: Journées Scientifiques Forum LABO – 2000, 28/03/2000 – 31/03/2000, Paris, France.

Comte J. (2001) Dosage des radionucléides à vie longue 79Se et 126Sn dans les solutions de produits de fission issues du traitement des combustibles nucléaires. (DEN/CAD/DED/SAMRA/LARC, CEA). Thèse de l’Université Paris XI – UFR Scientifique d'Orsay. 19/11/2001. 191 pages.

Comte J. et al. (2001) Dosage des radionucléides 79Se et 126Sn dans les solutions de produits de fission issues du traitement des combustibles usés. Congrès/colloque: 3ème Université d'Eté: Mesures et Analyses des Substances Radioactives – 2001. 17/09/2001 – 21/09/2001, Méjannes le Clap, France.

Comte J., Bienvenu P., Point C., Excoffier E., et al. (2002a) Détermination de l'étain 126 par ICP/MS et spectrométrie gamma dans les solutions de produits de fission issues du traitement des combustibles nucléaires. Radiochimica Acta.

Comte J. et al. (2002b) Détermination du 79Se par ETV-ICP/MS dans les solutions de produits de fission issus du traitement des combustibles nucléaires. Journal of Analytical Atomic Spectrometry.

Comte J., Bienvenu P., Brochard E., Femandez J.-M. Andreoletti G. (2003) Determination of selenium-79 in solutions of fission products after pre-treatment by ion exchange chromatography and ETV-ICP-MS. Journal of Analytical Atomic Spectrometry 18(7), 702– 707. doi:10.1039/b209253k. ISSN 0267-9477.

Coppin F., Chabroullet C., Martin-Garin A., Balesdent J. and Gaudet J.P., (2006) Methodological approach to asses the effect of soil ageing on selenium behaviour: first results concerning mobility and solid partition of selenium. Biology and Fertility of Soils 42, 379– 386.

Cordfunke E.H.P. and Konings R.J.M. (1988) Chemical interactions in water-cooled nuclear fuel: a thermochemical approach. Journal of Nuclear Materials 152, 301–309.

91/328

Corvilain B., Contempre B., Longombe A.O., Goyens P., Gervy-Decoster C., Lamy F., Vanderpas J.B., and Dumont J.E. (1993) Selenium and the thyroid: How the relationship was established. American Journal of Clinical Nutrition 57 (2 Suppl.), 244S–248S.

Cowper M.M. and Williams S.J. (2003) Plutonium and selenium studies – third Progress Report, AEAT/R/NS/0697 (September 2003).

Crusius J. and Thomson J. (2003) Mobility of authigenic rhenium, silver and selenium during post-depositional oxidation in marine sediments. Geochimica et Cosmochimica Acta, 67 (2), 265–273.

Cui D. and Spahiu K. (2002) The reduction of U(VI) on corroded iron under anoxic conditions. Radiochimica Acta 90 (9-11), 623–628. doi:10.1524/ract.2002.90.9-11_2002.623.

Cui D., Spahiu K., and Wersin P. (2003a) Redox reactions of iron and uranium dioxide in simulated cement pore water under anoxic conditions. Mat. Res. Soc. Symp. Series Vol 757 Scientific Basis for Nuclear Waste Management XXXVI, p. 427.

Cui D., Low J., Lundstrom M., and Spahiu K. (2004) Spent fuel leaching under anoxic conditions and the effect of canister materials. Scientific Basis for Nuclear Waste Management XXVII. Kalmar; Sweden, 15-19 June 2003. pp. 89–94.

Cui D., Scheidegger A., Puranen A., Wersin P., and Spahiu K. (2006) On the interactions between iron canister materials and fission product 79Se under simulated deep repository conditions. 30 pp. Progress report to the Near-Field Processes (NF-Pro) European project. NFPro/RTDC2/WP5. http://project.nf-pro.org/workspaces/rtdc2/deliverables/ Fe-Se-Deliverable 2[1].5.7.doc

Cullen W.R. and Reimer K.J. (1989) Arsenic speciation in the environment. Chemical Reviews 89 (4), 713–764. doi:10.1021/cr00094a002.

Curti E., Kulik D.A., and Tits J. (2005) Solid solution of trace Eu(III) in calcite: thermodynamic evaluation of experimental data over a wide range of pH and pCO2. Geochimica et Cosmochimica Acta 69, 1721–1737.

Cutter G.A. (1978) Species determination of selenium in natural waters. Analytica Chimica Acta 98(1), 59–66.

92/328

Cutter G.A. (1982) Selenium in reducing waters. Science 217(4562), 829–831.

Cutter G.A. and Bruland K.W. (1984) The marine biogeochemistry of selenium: a re- evaluation. Limnology and Oceanography 29 (6), 1179–1192.

Cutter G.A. and Cutter L.S. (2004) Selenium biogeochemistry in the San Francisco Bay estuary: changes in the water column behaviour. Estuarine, Coastal and Shelf Science 61, 463–476.

Dacey J.W.H. and Wakeham S. (1986) Oceanic dimethylsulfide: production during zooplankton grazing on phytoplankton. Science 233 (4770), 1314–1316. doi:10.1126/science.233.4770.1314.

Danbara A. and Shiraiwa Y. (1999) Requirement of selenium for the growth and selection of adequate culture media in a marine coccolithophorid, Emiliania huxleyi. Bulletin of the Society of Sea Water Science, Japan 53 (6), 476–484.

Danbara A. and Shiraiwa Y. (2007) The requirement of selenium for the growth of marine coccolithophorids, Emiliania huxleyi, Gephyrocapsa oceanica and Helladosphaera sp. (Prymnesiophyceae). Plant and Cell Physiology 40 (7), 762–766.

DAMRI / DTA / CEA (1991) Radionucléides. pp. 190. DAMRI, Département des Applications et de la Métrologie des Rayonnements Ionisants (CEA, Direction des Technologies Avancées). Edition Elaboration, Plessis Robinson. Reédition Octobre 1991. ISBN: 2-906483-03-6. Recueil regroupant les données principales relatives aux schémas de désintégration des radionucléides les plus utiles. Voir 75Se p. 34.

Darcheville O., Février L., Haichar F.Z., O. Berge, Martin-Garin A., P. Renault., (2007) Aqueous, solid and gaseous partitioning of selenium in a sandy soil under different microbiological states. Journal of Environmental Radioactivity.

Davis J.A. and Leckie J.O. (1980) Surface ionization and complexation at the oxide/water interface. III. Adsorption of anions. Journal of Colloid and Interface Science 74, 32–43.

Davis J.A. and Kent D.B. (1990) Surface complexation modeling in aqueous geochemistry. pp. 177 – 260. Chapter 5 in: Hochella M.F. Jr. and White A.F. Editors (1990) Mineral-water interface geochemistry. 603 pp. Volume 23 of Reviews in mineralogy. Series Editor: Ribbe P.H. (1990) Mineralogical Society of America. Washington, D.C. See Figures 13c and 14 on p. 201: sorption edges of selenite and selenate oxyanions on ferrihydrite.

93/328

Davis J.A., Kent D.B., Rea B.A., Maest A.S., and Garabedian S.P. (1993) Influence of redox environment and aqueous speciation on metal transport in groundwater: preliminary results of trace injection studies. pp. 223–273. In: Metals in groundwater. Edited by Allen H.E., Perdue M.E., and Brown D.S. Lewis Publishers.

DeBose J.L., Lema S.C., and Nevitt G.A. (2008) Dimethylsulfoniopropionate as a foraging cue for reef fishes. Science 319 (5868), 1356. doi:10.1126/science.1151109.

De Cannière P. (1989) Etude de la radiolyse α et γ du carbonate de calcium. Contribution à la datation par résonance paramagnétique électronique. Thèse de doctorat. Université Catholique de Louvain, UCL, Louvain-la-Neuve, Belgique. See: the effect of nitrate incorporated in calcite single crystals doped with 210Po: very typical triplet fingerprint dominating the ESR spectra. Concentration up to 3 000 ppm of nitrate were measured in ultrapure analytical grade of synthetic calcite. So, calcite could incorporate selenite as observed for nitrate.

De Cannière et al. (1990 – 1999) Migration Studies. In: Geological disposal of conditioned high-level and long lived radioactive waste. Reports R-2901/R2922 (1991), R-2948/R-2976 (1992), R-2968 (1993), R-3026/R-3053 (1994), R-3080 (1995), R-3088 (1996), R-3135 (1996), R-3240 (1998), R3380 (1999), SCK•CEN, Mol, Belgium (

De Cannière P., Wang L., Moors H., and Put M.J. (1995) Migration behaviour of 75Se in Boom Clay under in situ conditions. Scientific communication presented as a poster at the Fifth International Conference on the: Chemistry and migration behaviour of actinides and fission products in the geosphere. Migration’95 Conference held in Saint-Malo (France), September 10 – 15, 1995. Book of Abstracts, pp. 30–31, Abstract N° PB2-01, Unpublished.

De Cannière P., Moors H., Lolivier P., De Preter P., and Put M. (1996) Laboratory and in situ migration experiments in the Boom Clay. 51 pp. European Commission, Nuclear Science and Technology. Work carried out under a cost-sharing contract with the European Atomic Energy Community in the framework of its fourth R&D programme entitled “Management and storage of radioactive waste” (1990-1994) Part A, Task 4: “Disposal of radioactive waste”. Contract N° FI2W-CT90-0039 with the European Commission, Nuclear Science and Technology. Final report EUR 16 927 EN. See section 3.2.7 Selenium and Table 3.12: results of the migration experiments with selenium under reducing conditions, pp. 26–29.

De Cannière P., Bruggeman C., Maes A., Wang L., Maes N., Beauwens T., Moors H., Aertsens M., Cowper M.M., and Williams S.J. (2003) Solubility, sorption and migration behaviour of selenium in Boom Clay. Scientific communication presented as a poster at the Ninth International Conference on the: Chemistry and migration behaviour of actinides and

94/328

fission products in the geosphere. Migration’03 Conference held in Gyeongju (Korea), September 21 – 26, 2003. Book of Abstracts, pp. 140–141, Abstract N° PB1-6, Unpublished.

De Cannière P., Aoki K., Arcos D., Bath A., Boisson J.-Y., Courdouan A., Degueldre C., Fernández A.M., Fierz T., Gäbler H-E., Gaucher E., Gautschi A., Griffaut L., Hernán P., Mäder U., Mazurek M., Mettler S., Pearson F.J., Schippers A., Scholtis A., Schwyn B., Sergeant C., Stroes-Gascoyne S., Tournassat C., Turrero M.J., Vinsot A., Waber H.N., Wersin P. (2008) Chapter 6: Geochemistry and microbiology experiments. pp. 69–86. In: Paul Bossart and Marc Thury (editors) (2008) Mont Terri Rock Laboratory – Project, Programme 1996 to 2007 and Results. Berichte der Landesgeologie – Rapports du Service géologique national – Rapporti del Servizio geologico nazionale – Reports of the Swiss Geological Survey N° 3. 443 pp. – Swiss Geological Survey, Wabern, 2008. ISSN 1661- 9285. ISBN 978-3-302-40016-7. http://www.swisstopo.ch.

De Craen M., Van Geet M., Wang L., and Put M., (2004) High sulphate concentrations in squeezed Boom Clay pore water: evidence of oxidation of clay cores. Physics and Chemistry of the Earth 29(1), 91–103.

De Craen M., Wang L., Van Geet M., and Moors H. (2004) Geochemistry of Boom Clay pore water at the Mol site – Status 2004. 179 pp. Scientific Report SCK•CEN-BLG-990 (04/MDC/P-48). September 2004. Unclassified report: ISSN 1379-2407. Waste and Disposal Department.

Delécaut G. (2004) The geochemical behaviour of uranium in the Boom Clay. Ph.D. Thesis of Catholic University of Louvain (UCL) made in the laboratories of SCK•CEN Mol. Louvain- la-Neuve, academic year 2003-2004. pp. 215. UCL Promoter P. Sonnet, SCK•CEN mentor P. De Cannière. See a.o., Section 7.2.1.1 Pyrite preparation, p. 143.

Deliens M., Piret P., and Comblain G. (1981) Les minéraux secondaires du Zaïre. Editions du Musée Royal de l’Afrique Centrale, Tervuren, Belgium. 113 pp.

Deniau I. (2002) Caractérisation géochimique du kérogène associé à l'argile de Boom (Mol, Belgique) et évolution sous divers stress thermiques, 213 pp. Thèse. Université Paris VI.

Deniau I. (2004) Étude cinétique et moléculaire des produits gazeux et liquides générés par la matière organique insoluble de l'argile de Boom sous divers stress thermiques. Rapport de synthèse, IRSN, Janvier 2003 – Décembre 2004, 50 pp.

Deniau I., Derenne S., Beaucaire C., Pitsch H., and Largeau C. (2001) Morphological and chemical features of a kerogen from the underground Mol laboratory (Boom Clay Formation,

95/328

Oligocene, Belgium): structure, source organisms and formation pathways. Organic Geochemistry 32, 1343-1356.

Deniau I., Derenne S., Beaucaire C., Pitsch H., and Largeau C. (2004) Occurrence and nature of thermolabile compounds in the Boom Clay kerogen (Oligocene, underground Mol laboratory, Belgium). Organic Geochemistry 35, 91-108.

Deniau I., Derenne S., Beaucaire C., Pitsch H., and Largeau C. (2005a) Simulation of thermal stress influence on the Boom Clay kerogen (Oligocene, Belgium) in relation with long-term storage of high activity nuclear waste. I – Study of generated soluble compounds. Applied Geochemistry 20, 587-597.

Deniau I., Behar F., Largeau C., De Cannière P., Beaucaire C., and Pitsch H. (2005b)

Determination of kinetic parameters and simulation of early CO2 production from the Boom Clay kerogen under low thermal stress. Applied Geochemistry 20, 2097-2107.

Deniau I., Lorant F., Pitsch H., De Cannière P., Beaucaire C., Behar F., and Largeau C. (2006) Molecular and kinetic study of gaseous and liquid compounds generated by the insoluble organic matter (kerogen) of the Boom Clay under various thermal stresses. Final report for the post-doc of Isabelle Deniau. Collaboration between IRSN, ENSCP, IFP, and SCK•CEN with the financial support of ONDRAF/NIRAS.

Deniau I., Devol-Brown I., Derenne S., Behar F., Largeau C. (2008) Comparison of the bulk geochemical features and thermal reactivity of kerogens from Mol (Boom Clay), Bure (Callovo–Oxfordian argillite) and Tournemire (Toarcian shales) underground research laboratories. Science of the Total Environment 389(2-3), 475-485. doi:10.1016/j.scitotenv.2007.09.013.

De Preter P., Put M., De Cannière P., and Moors H. (1992) Migration of radionuclides in Boom Clay. State-of-the-art report, June 1992. Study performed under cost sharing contract for the European Commission. Contract FI2W/0039. Report NIROND 92-07. See description of the C4 type percolation experiments in Appendix 1: Different types of migration experiments, pp. 1.7, 1.8 (type C4) and Fig. 1.1.1 (type C4). Bruggeman equation: see p. 4.

Descostes M., Mercier F., Beaucaire C., Zuddas P., and Trocellier P. (2001) Nature and distribution of chemical species on oxidized pyrite surface: Complementarity of XPS and nuclear microprobe analysis. Nuclear Instruments and Methods in Physics Research B 181, 603–609.

96/328

De Silva V., Woznichak M.M., Burns K.L., Grant K.B., and May S.W. (2004) Selenium redox cycling in the protective effects of organoselenides against oxidant-induced DNA damage. Journal of the American Chemical Society 126 (8), 2409–2413.

Deverel S.J (1986) Distribution and mobility of selenium and other trace elements in shallow ground water of the western San Joaquin Valley, California (Regional aquifer-system analysis). 12 pp. U.S. Geological Survey. ASIN: B00070YG3A.

Devillanova F.A. (Editor) (2007) Handbook of chalcogen chemistry: new perspectives in sulfur, selenium and tellurium. 740 pp. Royal Society of Chemistry. 1st edition. ISBN: 0854043667.

Devivier K., Devol-Brown I., and Savoye S. (2004) Study of iodide sorption to the argillite of Tournemire in alkaline media. Proceedings of the Reims 2002 Andra International Workshop on Clays in Natural and Engineered Barriers for Radioactive Waste Confinement. Applied Clay Science 26(1-4), 171–179. doi:10.1016/j.clay.2003.07.010. de Wouters R. (2005) Characterization of irradiated fuel assemblies. Part 1: Radiological inventory. Issued September, 2005. Belgatom, avenue Ariane, 4 – B-1200 Brussels. Technical Note. Report BN ref. 0204246 / 221 – “A” – 1.

Dierckx A., Maes A., Henrion P., and De Cannière P. (1997) Potentiometric titration of humic substances extracted from Boom Clay. pp. 573 578 in: Drozd J., Gonet S.S., Senesi N. and Weber J., eds. (1997) The role of humic substances in the ecosystems and in environmental protection. PTSH – Polish Society of Humic Substances. Polish Chapter of the International Humic Substances Society. Wroclaw, Poland (1997). Proceedings of the 8th meeting of the International Humic Substances Society, Wroclaw, Poland, September 9-14, 1996.

Dierckx A. (1997) Boom Clay in situ porewater chemistry. SCK•CEN Report BLG-734. 3 pp. February 1997.

Dierckx A., Put M., De Cannière P., Wang L., Maes N., Aertsens M., Maes A., Vancluysen J., Verdickt W., Gielen R., Christiaens M., Warwick P., Hall A., and van der Lee J. (2000) Transport of radionuclides due to complexation with organic matter in Clay formations (Trancom-Clay). European Commission, Nuclear Science and Technology. Contract N° FI4W-CT95-0013. Final Report EUR 19 135. 170 pp.

Dong K., He M., Jiang S., Wong H., Qiu J., Guan Y., et al. (2007) Beam Interactions with Materials and Atoms: Measurement of trace 129I concentrations in CsI powder and organic

97/328

liquid scintillator with accelerator mass spectrometry. Nuclear Instruments and Methods in Physics Research Section B 259 (1), 271–276. doi:10.1016/j.nimb.2007.01.232.

Dowdleand P. and Oremland R. (1998) Microbial oxidation of elemental selenium in soil slurries and bacterial cultures. Environmental Science and Technology 32, 3749–3755.

Drotar A., Fall L.R., Mishalanie E.A., Tavernier J.E., and Fall R. (1987) Enzymatic methylation of sulfide, selenide, and organic thiols by Tetrahymena thermophila. Applied and Environmental Microbiology 53(9), 2111–2118.

Duc M., Lefevre G., and Fedoroff M. (2006) Sorption of selenite ions on hematite. Journal of Colloid and Interface Science 298(2), 556–563.

Duc M., Lefevre G., Fedoroff M., Jeanjean J., Rouchaud J.C., Monteil-Rivera F., Dumonceau J., and Milonjic S. (2003) Sorption of selenium anionic species on apatites and iron oxides from aqueous solutions. Journal of Environmental Radioactivity 70 (1-2), 61–72, http://ejournals.ebsco.com/direct.asp?ArticleID=N69JL1D1R64AEVGCR68P

Dungan R.S. and Frankenberger W.T. (1999) Microbial transformations of selenium and the bioremediation of seleniferous environments. Bioremediation Journal 3 (3), 171–188. doi:10.1080/10889869991219299

Dungan R.S. and Frankenberger W.T., Jr. (2000) Factors affecting the volatilization of dimethylselenide by Enterobacter cloacae SLD1a-1. Soil Biol. Biochem. 32, 1353–1358.

Dungan R.S. and Frankenberger W.T. Jr. (2001) Biotransformation of selenium by Enterobacter cloacae SLD1 a-1: Formation of dimethyl selenide. Biogeochemistry 55, 73–86.

Dungan R.S., Yates S.R., and Frankenberger W.T. Jr. (2002) Vadose zone processes and chemical transport. Volatilization and degradation of soil-applied dimethylselenide. Journal of Environmental Quality 31, 2045–2050.

Durand B. and Nicaise G. (1980) Procedures for kerogen isolations. In: Durand B. (Eds.), Kerogen. Insoluble organic matter from sedimentary rocks. Technip, Paris, pp. 35-53.

Dworkin M. (Editor-in-Chief), Falkow S., Rosenberg E., Schleifer K.H., and Stackebrandt E. (Editors) (2006) The prokaryotes. Third edition. A handbook on the biology of bacteria. Volume 2: Ecophysiology and biochemistry. 1107 pp. Springer. New York, Singapore. ISBN- 10: 0-387-25492-7. doi:10.1007/0-387-30742-7.

98/328

Dzombak D.A. and Morel F.M.M. (1990) Surface complexation modeling – Hydrous ferric oxide, 393 pp. A Wiley-Interscience Publication. John Wiley & Sons Inc. New York – Chichester – Brisbane – Toronto – Singapore. ISBN 0-471-63731-9. See tables and figures from: 7.6 Sulfate, pp. 219 – 220; 7.7 Selenate, pp. 221 – 225; 7.8 Selenite, pp. 226 – 227; and 7.9 Thiosulfate, pp. 228 – 229. See also Figure 10.3 pp. 310.

EC (1979) European catalogue of geological formations having favourable characteristics for the disposal of solidified high-level and/or long-lived radioactive wastes. Volume 2: Belgium. EC, Luxembourg.

Elleouet C., Quentel F., and Madec C. (1996) Determination of inorganic and organic selenium species in natural waters by cathodic stripping voltammetry. Water Research 30(4), 909–914.

Ellis A.S., Johnson T.M., Herbel M.J., and Bullen T.D. (2003) Stable isotope fractionation of selenium by natural microbial consortia. Chemical Geology 195, 119 – 129. doi:10.1016/S0009-2541(02)00391-1.

Elrashidi M.A., Adriano D.C., Workman S.M., and Lindsay W.L. (1987) Chemical equilibria of selenium in soils: a theoretical development. Soil Science 144(2), 141–152.

Elson C.M. and MacDonald A.S. (1979) Determination of selenium in pyrite by an ion exchange electrothermal atomic-absorption spectrometric method. Analytica Chimica Acta 110(1), 153–156.

Ewing R.C. (2001) Corrosion of spent nuclear fuel: the long-term assessment. Interim Report. Environmental Management Science Program. Nuclear Engineering and Radiological Sciences, Materials Science & Engineering, Geological Sciences, University of Michigan. Project I.D. No.: 73751 (previously 59849). Grant Number: DE-FG07-97ER14816. Project Duration: 9/15/2000 to 9/14/2003.

Fernandez-Martinez A. and Charlet L. (2009) Selenium environmental cycling and bioavailability: a structural chemist point of view. Review in Environmental Science and Biotechnology 8, 81–110.

Ferry C., Poinssot C., Cappelaere C., Desgranges L., Jegou C., Miserque F., Piron J.P., Roudil D., Gras J.M. (2006) Specific outcomes of the research on the spent fuel long-term evolution in interim dry storage and deep geological disposal. Journal of Nuclear Materials 352, 246– 253. doi:10.1016/j.jnucmat.2006.02.061.

99/328

Ferri T. and Sangiorgio P. (1999) Voltammetric study of the interaction between Se(IV) and dissolved organic matter in environmental aqueous matrices. Analytica Chimica Acta 385, 337–343.

Février L., Martin-Garin A., and Leclerc-Cessac E. (2007) Variation of the distribution coefficient Kd of selenium in soils under various microbial states. Journal of Environmental Radioactivity 97, 189–205.

Filius J.D., Hiemstra T., and van Riemsdijk W.H. (1997) Adsorption of small weak organic acids on goethite: Modeling of mechanisms. Journal of Colloid and Interface Science 195, 368–380.

Filius J.D., Meeussen J.C.L., and van Riemsdijk W.H. (1999) Transport of malonate in a goethite–silica sand system. Colloids and Surfaces A-Physicochemical and Engineering Aspects 151, 245–253.

Finch R. and Murakami T. (1999) Systematics and paragenesis of uranium minerals, pp. 91 – 179. Chapter 3 in: Burns P.C. and Finch R. (eds.) (1999) Uranium: mineralogy, geochemistry and the environment. Reviews in Mineralogy, volume 38. 679 pp. ISBN 0-939950-50-2. Series editor: Paul H. Ribbe. Mineralogical Society of America. Washington DC. ISSN 0275- 0279. See uranyl selenites, Table 17, p. 136–137. See also piretite reported by Vochten et al. (1996).

Fio J.L. and Fujii R. (1990) Selenium speciation methods and application to soil saturation extracts from San Joaquin Valley, California, Soil Sci. Soc. Am. J. 54, 363–369.

Foster A.L., Brown G.E., Jr., and Parks G.A. (2003) X-ray absorption fine structure study of As(V) and Se(IV) sorption complexes on hydrous Mn oxides. Geochimica et Cosmochimica Acta 67(11), 1937–1953.

Fox P.M., LeDuc D.L., Hussein H., Lin Z.Q., and Terry N. (2003) Selenium speciation in soils and plants. Biogeochemistry of Environmentally Important Trace Elements 835, 339– 354.

Frankenberger W.T. Jr. (1989) Microbial volatilization of selenium at Kesterson Reservoir. Interim report. Dept. of Soil and Environmental Sciences, University of California. ISBN: B000736YFA.

Frankenberger W.T. Jr. and Benson S. (eds.) (1994) Selenium in the environment. pp. 456 Marcel Dekker, Inc., New York, Basel, Hong Kong. ISBN 0-8247-8993-8.

100/328

Frankenberger W.T. Jr. and Engberg R.A.. (eds.) (1998) Environmental chemistry of selenium. pp. 711. Marcel Dekker, Inc., New York, Basel, Hong Kong.

Frankenberger W.T. Jr and Arshad M. (2001) Bioremediation of selenium-contaminated sediments and water. Biofactors 14, 241–254.

Frankenberger W.T. Jr. (2001) Environmental chemistry of arsenic. CRC: Books in Soils, Plants, and the Environment. ISBN: 0824706765.

Frankenberger W.T. Jr. (ed.) (2002). Environmental chemistry of arsenic. pp. 410. Marcel Dekker, Inc., New York, Basel, Hong Kong.

Friel J.K., Mercer C., Andrews W.L., Simmons B.R., Jackson S.E., and Longerich H.P. (1996) Laboratory gloves as a source of trace element contamination. Biological Trace Element Research 54, 135–142. See Table 2 p. 138: the very high contamination concentration of selenium amongst the biologically important elements. Vinyl gloves are identified as an important source of selenium contamination in this study. Other main contaminants are Mg, Zn and Cu. The contamination source could be the filler used as load in the vinyl gloves, or the talc.

Frigg R. and Hartmann S. (2006) “Models in Science”, The Stanford Encyclopedia of Philosophy (Winter 2006 Edition), Zalta Edward N. (ed.), URL: http://plato.stanford.edu/archives/win2006/entries/models-science/.

Fukushi K. and Sverjensky D.A. (2007) A surface complexation model for sulfate and selenate on iron oxides consistent with spectroscopic and theoretical molecular evidence. Geochimica et Cosmochimica Acta 71(1), 1–24. doi:10.1016/j.gca.2006.08.048.

Garrels R.M. and Christ C.L. (1965) Minerals, solutions and equilibria. 453 pp. Harper and Rowley, New York.

Gaucher E., Blanc P., Bardot F., Braibant G., Buschaert S., Crouzet C., Gautier A., Girard J.- P., Jacquot E., Lassin A., Négrel G., Tournassat C., Vinsot A., and Altmann S. (2006) Modelling the porewater chemistry of the Callovian-Oxfordian formation at a regional scale. Comptes Rendus Geosciences 29, 55–77.

Gaucher E.C., Tournassat C., Jacquot E., Altmann S., and Vinsot A. (2007) Improvements in the modelling of the porewater chemistry of the Callovo-Oxfordian Formation. [Oral presentation: O/04/2]. In Proc. Clays in natural and engineered barriers for radioactive waste

101/328

confinement – Lille 2007 – 3rd International Meeting. Organised by Andra in Lille (France), 17 – 20 September 2007.

Geelhoed J.S., Findenegg G.R., and van Riemsdijk W.H. (1997) Availability to plants of phosphate adsorbed on goethite: experiment and simulation. European Journal of Soil Science 48, 473–481.

Geelhoed J.S., Hiemstra T., and van Riemsdijk W.H. (1997a) Phosphate and sulfate adsorption on goethite: Single anion and competitive adsorption. Geochimica et Cosmochimica Acta 61, 2389–2396.

Geelhoed J.S., Hiemstra T., and van Riemsdijk W.H. (1998) Competitive interaction between phosphate and citrate on goethite. Environmental Science & Technology 32, 2119–2123.

Geelhoed J.S., van Riemsdijk W.H., and Findenegg G.R. (1997b) Effects of sulphate and pH on the plant-availability of phosphate adsorbed on goethite. Plant and Soil 197, 241–249.

Geering H.R., Cary E.E., Jones L.H.P., and Alloway W.H. (1968) Solubility and redox criteria for the possible forms of selenium in soils. Soil Science Society of America Proceedings 32, 35–40.

Gens R. (1998) Radionuclide solubilities in Boom Clay / Phase II – Part 1, Reference RG/AV/98-2368, fax from R. Gens of ONDRAF/NIRAS to N.J. Pilkington of AEA Technology, 13 July 1998.

Ghosh-Dastidar A., Mahuli S., Agnihotri R., and Fan L.-S. (1996) Selenium capture using sorbent powders: mechanism of sorption by hydrated lime. Environmental Science & Technology 30 (2), 447–452.

Gillespie M.R. (2001) Potential for secondary iron oxy-hydroxide minerals to retard radionuclide migration: a literature review and assessment. Nirex Report NSS/R294.

Glaus M.A., Hummel W., and Van Loon L.R. (1997) Experimental determination and modelling of trace metal-humate interactions. PSI-Bericht Nr. 97-13. Paul Scherrer Institut, Villigen.

Gmelin (1990) Gmelin handbook of inorganic and organo-metallic chemistry. 8 edition. Se. Selenium. Supplement Vol. B 1. Compounds with hydrogen, oxygen, nitrogen. IX, 343 S. Springer Berlin. ISBN 3-540-93437-5.

102/328

Goh K.H. and Lim T.T. (2004) Geochemistry of inorganic arsenic and selenium in a tropical soil: effect of reaction time, pH and competitive anions on arsenic and selenium adsorption. Chemosphere 55(6), 849–859.

Goldberg S. and Glaubig R.A. (1988) Anion sorption on a calcareous, montmorillonitic soil – selenium. Soil Science Society of America Journal 52(4), 954–958.

Goldhaber S.B. (2003) Trace element risk assessment: essentiality versus toxicity. Regulatory Toxicology and Pharmacology 38, 232–242.

Gómez-Ariza J.L., Sánchez-Rodas D., Morales E., Herrgott O., and Marr I.L. (1999) Inorganic and organic selenium compound speciation with coupled HPLC-MWHG-AFS, Applied Organometallic Chemistry 13, 783–787.

Griffault L., Merceron T., Mossmann J.R., Neerdael B., De Cannière P., Beaucaire C., Daumas S., Bianchi A. and Christen R. (1996) Acquisition et régulation de la chimie des eaux en milieu argileux pour le projet de stockage de déchets radioactifs en formation géologique. Projet « Archimède-Argile ». Rapport Final EUR 17454FR. Commission Européenne, Sciences et Techniques Nucléaires. 176 pp. Travail coordonné par l’ANDRA dans le cadre du contrat F12W-CT92-0117 avec la Commission Européenne.

Grimes R.W. and Catlow C.R.A. (1991) The stability of fission products in uranium dioxide. Philosophical Transactions of the Royal Society of London A 335, 609.

Gruebel K.A., Davis J.A., and Leckie J.O. (1995) Kinetics of oxidation of selenite to selenate in the presence of oxygen, titania, and light. Environmental Science and Technology 29(3), 586–594.

Gudowski W., Gonzales E., Greneche D., Boucher L., Marivoet J., Zimmerman C., von Lenza W., and Vokal A. (2006) Impact of partitioning, transmutation and waste reduction technologies on the final nuclear waste disposal. Proceedings of the FISA 2006 Conference, Luxembourg, March 13-16, 2006, European Commission (in preparation).

Guo L., Jury W., and Frankenberger W.T. Jr. (2001) Coupled production and transport of selenium vapor in unsaturated soil; Evaluation by experiments and numerical simulation. Journal of Contaminant Hydrology 49, 67–85.

Gustafson G. and Olsson O. (1993) The development of conceptual models for repository site assessment. pp. 145–158 in: OECD (1993) The role of conceptual models in demonstrating

103/328

repository post-closure safety. Proceedings of a NEA Workshop, Paris, 16-18 November 1993. Nuclear Energy Agency (NEA). OECD Documents. 190 pp. ISBN 92-64-14429-3.

Gustafsson J.P. and Johnsson L. (1994) The association between selenium and humic substances in forested ecosystems – Laboratory evidence. Applied Organometallic Chemistry 8, 141–147.

Gysemans M. and Moors H. (2000) Determination of Se-75, Zr-95, Np-237 and Am-241 activities in Boom Clay samples from laboratory migration experiments using gamma-ray spectrometry. Applied Radiation and Isotopes 53(1-2), 209–213.

H3 (1992) Research and development on geological disposal of high-level radioactive waste, first progress report – H3. PNC TN1410 93-059.

H12 (2000) H12: Project to establish the scientific and technical basis for HLW disposal in Japan. Second progress report on research and development for the geological disposal of HLW in Japan. Project overview report. JNC Report TN1410 2000-001. Japan Nuclear Cycle Development Institute. April, 2000. See selenium parameters in Appendix K: Datasets for sensitivity analysis, pp. K-4 and K-5: Table K-4: Solubilities used in sensitivity analysis [mol l-1 at 25 °C]. Table K-5: Effective diffusion coefficients for the buffer used in sensitivity 2 -1 analysis [m s at 60 °C]. Table K-6: Distribution coefficients (Kd) used in sensitivity analysis [m3 kg-1 at 25 °C]. http://www.jaea.go.jp/04/tisou/english/h12report/e_h12index.html

H17 (2005) Development and management of the technical knowledge base for the geological disposal of HLW. JNC Report TN1400 2005-022. Japan Nuclear Cycle Development Institute. September, 2005. http://www.jaea.go.jp/04/tisou/english/pdf/H17_KM_Report_E.pdf

Hanjie W., Carignan J., Yuzhuo Q., and Shirong L. (2006) Selenium speciation in kerogen from two Chinese selenium deposits: Environmental implications. Environmental Science and Technology 40 (4), 1126–1132.

Hansen, H.C.B. (2002) Environmental chemistry of iron(II)-iron(III) LDHs (green rusts). Chapter 13, in: Rives, V. (Ed.) Layered double hydroxides: present and future, pp. 413–434, Nova Sci. Publ., NY.

Hatfield D.L., Berry M.J., and Gladyshev V.N. (Editors) (2006) Selenium: its molecular biology and role in human health. 444 pp. Springer. 2nd edition. ISBN: 0387338268.

104/328

Haudin C.S., Renault P., Hallaire V., Leclerc-Cessac E., Staunton S., (2007) Effect of aeration on mobility of selenium in columns of aggregated soil as influenced by straw amendment and tomato plant growth. Geoderma 141, 98–110.

Haudin C.S., Fardeau M.L., Amenc L., Renault P., Ollivier B., Leclerc-Cessac E., Staunton S., (2007) Responses of anaerobic bacteria to soil amendment with selenite. Soil Biology and Biochemistry 39, 2408–2413.

Haudin C.S., Renault P., Leclerc-Cessac E., Staunton S., (2007) Effect of selenite additions on microbial activity and dynamics in three soils incubated under aerobic conditions. Soil Biology and Biochemistry 39, 2670–2674.

Hayes K.F. (1987) Equilibrium, spectroscopic, and kinetics studies of ion adsorption at the oxide/aqueous interface. Ph.D. Thesis, Stanford University, Stanford, CA, USA.

Hayes K.F., Roe A.L., Brown G.E., Hodgson K.O., Leckie J.O., and Parks G.A. (1987) In situ x-ray absorption study of surface complexes: selenium oxyanions on α-FeOOH. Science 238, 783–786.

Hayes K.F., Papelis C., and Leckie J.O. (1988) Modeling ionic strength effects on anion adsorption at hydrous oxide/solution interfaces. Journal of Colloid and Interface Science 78, 717–726.

He M., Jiang S., Peng B., Ruan X., Dong K., Guan Y., et al. (2007) Beam Interactions with Materials and Atoms: 99Tc measurements with accelerator mass spectrometry at CIAE. Nuclear Instruments and Methods in Physics Research Section B 259 (1), 708–713. doi:10.1016/j.nimb.2007.01.209.

Heath T.G., Ilett D.J., Tweed C.J., and Williams S.J. (1997) Radionuclide solubilities in Boom Clay – Draft Task 1 Report, AEA Technology Report, February 1997.

Heath T.G., Ilett D.J., Maning M.C., Pilkington N.J., Tweed C.J., and Williams S.J. (1997) Radionuclide solubilities in Boom Clay – 1st Progress Report, (March 1997).

Heath T.G., Ilett D.J., Tweed C.J., and Williams S.J. (2000) Radionuclide solubilities in Boom Clay. Final report, part 1, AEAT/NS/R/0180 part 1, AEA Technology, UK.

Heath T.G., Ilett D.J., Tweed C.J., and Williams S.J. (2000) Radionuclide solubilities in Boom Clay. Final report, part 2, AEAT/NS/R/0180 part 2, AEA Technology, UK.

105/328

Hem J.D. (1985) Study and interpretation of the chemical characteristics of natural waters. USGS, US Geological survey water-supply paper 2254. Third edition. 264 pp. The sulfur cycle, p. 113.

Henrion P.N., Monsecour M., Fonteyne A., Put M., and De Regge P. (1985) Migration of radionuclides in Boom Clay. Radioactive Waste Management and the Nuclear Fuel Cycle 6(3-4), 313–359.

Henrion P., Put M., and Monsecour M. (1990) Synthesis report on: Transport of radionuclides in Boom Clay – State-of-the-Art. 87 p.p. SCK•CEN report coordinated by P. De Regge. R-2863.

Henrion P. and De Cannière (1990) Semi-annual report of the second semester 1990.

Hens M. and Merckx R. (2002) The role of colloidal particles in the speciation and analysis of “dissolved” phosphorus, Water Research 36, 1483–1492.

Herbel M.J., Blum J.S., Oremland R.S., and Borglin S.E. (2003) Reduction of elemental selenium to selenide: experiments with anoxic sediments and bacteria that respire Se- oxyanions. Geomicrobiology Journal 20(6), 587–602.

Herbel M., Johnson T., Oremland R., and Bullen T. (2000) Fractionation of selenium isotopes during bacterial respiratory reduction of selenium oxyanions. Geochimica et Cosmochimica Acta 64(21), 3701–3709. doi:10.1016/S0016-7037(00)00456-7.

Hockin S.L. and Gadd G.M. (2003) Linked redox precipitation of sulfur and selenium under anaerobic conditions by sulfate-reducing bacterial biofilms. Applied Environmental Microbiology 69, 7063–7072.

Hiemstra T., De Wit J.C.M., and van Riemsdijk W.H. (1989a) Multisite proton adsorption modelling at the solid/solution interface of (hydr)oxides: A new approach. II. Application to various important (hydr)oxides. Journal of Colloid and Interface Science 133(1), 105–117.

Hiemstra T., van Riemsdijk W.H., and Bolt G.H. (1989b) Multisite proton adsorption modelling at the solid/solution interface of (hydr)oxides: A new approach. I. Model description and evaluation of intrinsic reaction constants. Journal of Colloid and Interface Science 133(1), 91–104.

Hiemstra T. and van Riemsdijk W.H. (1996) A surface structural approach to ion adsorption: The charge distribution (CD) model. Journal of Colloid and Interface Science 179, 488–508.

106/328

Hiemstra T. and van Riemsdijk W.H. (1999) Surface structural ion adsorption modeling of competitive binding of oxyanions by metal (hydr)oxides. Journal of Colloid and Interface Science 210(1), 182–193.

Hiemstra T., Venema P., and van Riemsdijk W.H. (1996) Intrinsic proton affinity of reactive surface groups of metal (hydr)oxides: The bond valence principle. Journal of Colloid and Interface Science 184, 680–692.

Hogg D.R. (Editor) (1975) Organic compounds of sulphur, selenium and tellurium. Royal Society of Chemistry. ISBN: 0851862799.

Holland H.D. and Turekian K.K. (executives editors) (2004) Treatise on Geochemistry. Collection of 10 volumes. Elsevier Pergamon Oxford. ISBN (set): 0-08-043751-6. http://www.sciencedirect.com/science/referenceworks/0080437516. Selenium: search indexes p. 157, Vol. 10.

Howard J.J. (1977) Geochemistry of selenium. Formation of ferroselite and selenium behaviour in the vicinity of oxidizing sulfide and uranium deposits. Geochimica et Cosmochimica Acta 41, 1665–1678.

Hug S.J. (1997) In situ Fourier transform infrared measurements of sulfate adsorption on hematite in aqueous solutions. Journal of Colloid and Interface Science 188, 415–422.

Hummel W., Berner U., Curti E., Pearson F.J. and Thoenen T. (2002) Nagra/PSI Chemical Thermodynamic Data Base 01/01 (2002). Nagra Technical Report NTB 02-16, Nagra, Wettingen, Switzerland, and Universal Publishers/uPublish.com, Parkland, Florida.

Hummel W. (Chairman), Anderegg G., Puigdomenech I., Rao L., and Tochiyama O. (2005) Chemical thermodynamics of compounds and complexes of U, Np, Pu, Am, Tc, Se, Ni and Zr with selected organic ligands. Chemical Thermodynamics Series Volume 9. Elsevier Science.

Ihnat M. (1989) Occurrence and distribution of selenium. 368 pp. CRC. ISBN: 084934932X.

Ishihara Y., Ishiguro K., and Umeki H. (1999) Effect of change in the half-life of Se-79 on the safety of the HLW geological disposal system. JNC TN8400 99-086 (in Japanese).

Jacobs L.W. (1989) Selenium in Agriculture and the Environment. pp. 233: Proceedings (Special Publication, No 23, Soil Science Society of America, SSSA,) ISBN: 0891187898. (1989-05-01).

107/328

Jadhav R.A., Agnihotri R., Gupta H., and Fan L.-S. (2000) Mechanism of selenium sorption by activated carbon. Canadian Journal of Chemical Engineering 78 (1), 168–174.

Janssen K. (2003) μ-Xanes cartography of the spatial distribution of bio-accumulated selenium in onion bulbs grown in selenium-rich solutions. Seminar given at SCK•CEN by Prof. Koen Janssen (University of Antwerp, UA, campus Drie Eiken).

Jiang Song-Sheng, He Ming, Diao Li-Jun, Guo Jing-Ru, and Wu Shao-Yong (2001) Re- measurement of the half-life of 79Se with the projectile X-ray detection method. Chinese Physical Letters 18, 746–749. doi:10.1088/0256-307X/18/6/311.

Jiang S-S., He M., Diao L-J., Li C., Guo J., and Wu S. (2002) Re-measurement of the half-life of 79Se. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 489(1-3), 195–201.

Jian-ming Z., Wei Z., Hai-bo Q., Zhi-gang F., Bao-shan Z., and Hong-can S. (2008) An investigation on the source of soil Se in Yutangba, Enshi: evidence from native selenium. Acta Mineralogica Sinica.

Johnson L.H. and Smith P.A. (2000) The interaction of radiolysis products and canister corrosion products and the implications for radionuclide transport in the near-field of a repository for spent fuel. Nagra Technical Report NTB 00-04, Wettingen, Switzerland.

Johnson T.M. (2004) A review of mass-dependent fractionation of selenium isotopes and implications for other heavy stable isotopes. Chemical Geology 204 (3-4), 201–214. http://ejournals.ebsco.com/direct.asp?ArticleID=MW1H7VLGW5YX03E6W1CM

Johnson T.M. and Bullen T.D. (2003) Selenium isotope fractionation during reduction by Fe(II)-Fe(III) hydroxyde-sulfate (green rust). Geochimica et Cosmochimica Acta 67 (3), 413– 419.

Kafkafi U, Bar-Yosef B, Rosenberg R, and Sposito G. (1988) Phosphorus adsorption by kaolinite and montmorillonite: II. organic anion competition. Soil Science Society of America Journal SSSJD4 52(6), November-December 1988. 4 Fig., 2 Tables, 28 ref. US-Israel Binational Agricultural R&D Grant I-401-82.

Kashefi K. and Lovley D.R. (2000) Reduction of Fe(III), Mn(IV), and toxic metals at 100 °C by Pyrobaculum islandicum. Applied Environmental Microbiology 66(3), 1050–1060.

108/328

KBS-3 (1983) Final Storage of Spent Nuclear Fuel. Report by the Swedish Nuclear Fuel Supply Co. Stockholm, Sweden.

Killops S.D. and Killops V.J. (1993) An introduction to organic geochemistry. Longman Scientific & Technical, Essex, England, 265 pp.

Klayman D.L. and Günther W.H.H. (1973) Organic selenium compounds: their chemistry and biology. Wiley-Interscience, New York. ISBN: 0471490326.

Kleykamp H. (1985) The chemical state of fission products in oxide fuels. Journal of Nuclear Materials 131, 221–246.

Kleykamp H. (1993) The solubility of selected fission products in UO2 and (U, Pu)O2. Review article. Journal of Nuclear Materials 206, 82–86.

Klonowska A., Heulin T., and Vermeglio A. (2005) Selenite and tellurite reduction by shewanella oneidensis. Applied and Environmental Microbiology 71 (9), 5607–5609. doi:10.1128/AEM.71.9.5607-5609.2005

Köhler S.J., Bosbach D., and Oelkers E.H. (2005) Do clay mineral dissolution rates reach steady state? Geochimica et Cosmochimica Acta 69(8), 1997–2006.

Köhler S.J., Dufaud F., and Oelkers E.H. (2003) An experimental study of illite dissolution kinetics as a function of pH from 1.4 to 12.4 and temperature from 5 to 50°C. Geochimica et Cosmochimica Acta 67(19), 3583–3594.

Koopal L.K., Nederlof M.M., van Riemsdijk W.H., and Barneveld P.A. (1997) Semianalytical methods to determine first-order rate constant distributions. Langmuir 13, 961–969.

Kucha H. (1981) Precious metal alloys and organic matter in the Zechstein copper deposits, Poland. Mineralogy and Petrology 28 (1), 1–16.

Kulp T.R. and Pratt L.M. (2004) Speciation and weathering of selenium in upper cretaceous chalk and shale from South Dakota and Wyoming, USA. Geochimica et Cosmochimica Acta 68 (18), 3687–3701. doi:10.1016/j.gca.2004.03.008.

Ladrière Jean (1969) Etude de quelques réactions d'oxydo-réduction intervenant en électrolyse du cuivre. Université Catholique de Louvain (UCL). SC/Laboratoire de Chimie Minérale. Thèse de doctorat. Promoteur R. Breckpot. Collation 109. Communication personnelle: le sélénite est un poison pour l’électro-affinage du cuivre.

109/328

Ladrière J., Dussart F., Dabi J., Haulotte O., Verhaeghe S., and Regout J. (2009) Mössbauer study of the Boom clay, a geological formation for the storage of radioactive wastes in Belgium. Hyperfine Interactions 191 (1-3), 1–9. doi:10.1007/s10751-009-9977-9. http://www.springerlink.com/content/0575157242196551/.

Laenen B. (1997) The geochemical signature of relative sea-level cycles recognized in the Boom Clay. Thesis of the Catholic University of Louvain (KULeuven), 396 pp.

Lakshtanov L.Z. and Stipp S.L.S. (2004) Experimental study of europium(III) coprecipitation with calcite. Geochimica et Cosmochimica Acta 68, 819–827.

Lasaga A.C and Kirkpatrick R.J. (eds) (1981) Reviews in mineralogy. Volume 8: Kinetics of geochemical processes. 398 pp. Series editor: Ribbe P.H., Mineralogical Society of America. ISBN 0-939950-08-1.

Lasaga A.C (1997) Kinetic theory in the earth sciences. 811 pp. Princeton series in geochemistry, edited by Holland H.D. Princeton University Press, Princeton, New Jersey, USA. ISBN 0-691-03748-5.

Latimer W.M. (1952) The oxidation states of the elements and their potentials in aqueous solutions. Prentice-Hall Inc, New York.

Lauber M., Baeyens B., and Bradbury M.H. (2000a) Physico-chemical characterisation and sorption measurements of Cs, Sr, Ni, Eu, Th, Sn and Se on Opalinus Clay from Mont Terri. 92 pp. Waste Management Laboratory (LES) PSI, Switzerland. PSI Bericht Nr. 00-10. December 2000. ISSN 1019-0643. See particularly: 3.11 Se(IV) Sorption data p. 68; 3.11.1 Se kinetics and isotherm results 68–72.

Lauber M., Baeyens B., and Bradbury M.H. (2000b) Physico-chemical characterisation and sorption measurements of Cs, Sr, Ni, Eu, Th, Sn and Se on Opalinus Clay samples from Mont Terri. Nagra Technical Report NTB 00-01, Wettingen, Switzerland.

Leckie J.O, Benjamin M.M., Hayes K.F., Kaufman G., and Altmann S. (1980) Adsorption/coprecipitation of trace elements from water with iron oxydroxide. EPRI RP-910- 1. Electric Power Research Institute, Palo alto, California.

Lederer C.M., Hollander J.M., and Perlman I. (1967) Table of Isotopes, sixth edition. For 75Se and 79Se isotopes data, see p. 30, p. 210, p. 212, p. 214, and p. 527. John Wiley & Sons, Inc., New York • London • Sydney.

110/328

Lefèvre G. and Fédoroff M. (2006) Sorption of sulfate ions onto hematite studied by attenuated total reflection-infrared spectroscopy: Kinetics and competition with other ions. Proceedings of MIGRATION 2005, the 10th International Conference on the Chemistry and Migration of Actinides and Fission Products in the Geosphere. Physics and Chemistry of the Earth, Parts A/B/C 31 (10-14), 499–504. doi:10.1016/j.pce.2006.04.001.

Lemly A.D. (2002) Selenium assessment in aquatic ecosystems: a guide for hazard evaluation and water quality criteria. 161 pp. Springer, series on environmental management. New York. ISBN 0387953469.

Levander O.A. (1997) Nutrition and newly emerging viral diseases: An overview. Journal of Nutrition 127, 948S–950S.

Li C., Guo J., and Li D. (1997) A procedure for the separation of 79Se from fission products and application to the determination of the 79Se half-life. Journal of Radioanalytical and Nuclear Chemistry 220(1), 69–71.

Li Y.-H. and Gregory S. (1974) Diffusion of ions in sea water and in deep-sea sediments. Geochimica et Cosmochimica Acta 38, 703–714.

Lide D.R. (ed.) (1995) CRC Handbook of Chemistry and Physics. Table of the isotopes, T½ of 79Se = 6.5 × 104 years, see p. 11–53. 76th Edition (1995 – 1996). CRC Press, The Chemical Rubber Company. Boca Raton, New York, London, Tokyo.

Lindemer T.B., Besmann T.M., and Johnson C.E. (1981) Thermodynamic review and calculations – Alkali-metal oxides systems with nuclear fuels, fission products, and structural materials. Key review article providing thermodynamic data on various selenium and tellurium systems. Journal of Nuclear Materials 100, 178–226.

Liu X., Fattahi M., Montavon G., and Grambow B. (2008) Selenide retention onto pyrite under reducing conditions. Radiochimica Acta 96 (8), 473–479. doi:10.1524/ract.2008.1514.

Longnecker M.P., Taylor P.R., Levander O.A., Howe M., Veillon C., McAdam P.A., Patterson K.Y., Holden J.M., Stampfer M.J., Morris J.S., Willett W.C. (1991) Selenium in diet, blood, and toenails in relation to human health in a seleniferous area. American Journal of Clinical Nutrition 53, 1288–1294.

Lorant F., Largeau C., Behar F., and De Cannière P. (2008) Improved kinetic modeling of the early generation of CO2 from the Boom Clay kerogen. Implications for simulation of CO2

111/328

production upon disposal of high-activity nuclear waste. Organic Geochemistry 39(9), 1294- 1301. doi:10.1016/j.orggeochem.2008.05.008.

Losi M.E. and Frankenberger Jr. W.T. (1997) Reduction of selenium oxyanions by enterobacter cloacae strain sld1a-1: reduction of selenate to selenite. Environmental Toxicology and Chemistry 16, No. 9, 1851–1858. doi:10.1897/1551-5028(1997)016<1851:ROSOBE>2.3.CO;2

Lovley D.R. (1993) Dissimilatory metal reduction. Ann. Rev. Microbiol. 47, 263–290.

Loyaux-Lawniczak S., Refait P., Lecomte P., Ehrhardt J.-J., and Genin J.-M.R. (1999) The reduction of chromate ions by Fe(II) layered hydroxides. Hydrol. Earth Syst. Sci. 3, 593–599.

Loyauz-Lawniczak S., Refait P., Ehrhardt J.-J., Lecomte P., and Genin J.-M.R. (2000) Trapping of Cr by formation of ferrihydrite during the reduction of chromate ions by Fe(II)– Fe(III) hydroxysalt green rusts. Environ. Sci. Technol. 34, 438–443.

Lu Z., Fehn U., Tomaru H., Elmore D., and Ma X. (2007) Beam Interactions with Materials and Atoms: Reliability of 129I/I ratios produced from small sample masses. Nuclear Instruments and Methods in Physics Research Section B 259 (1), 359–364. doi:10.1016/j.nimb.2007.01.180.

Luo Shen Kai (1990) Master thesis on selenium made at SCK•CEN under the supervision of Pierre Henrion. In some gel permeation chromatography measurements the peak of 75Se and of OM were weakly correlated giving to think that a weak association could exist.

Luther G.W., Kostka J.E., Church T.M., Sulzberger B., and Stumm W. (1992) Seasonal iron cycling in the salt-marsh sedimentary environment – The importance of ligand complexes with Fe(II) and Fe(III) in the dissolution of Fe(III) minerals and pyrite, respectively. Marine Chemistry 40(1-2), 81–103.

Luttrell G. (1959) Annotated bibliography on the geology of selenium (Geological Survey bulletin). U.S. G.P.O. ASIN: B0007EP16M.

Macalady D.L., Ahmann D., Westall J., Garbino J., and Meyer J. (2002) Redox transformations, complexation and soil/sediment interactions of inorganic forms of As and Se in aquatic environments: Effects of natural organic matter. Tailings and Mine Waste ’02, Bakema, 13–15.

112/328

Macalady D.L., Ahmann D., and Garbino J. (2003) Redox transformations, complexation and soil/sediment interactions of inorganic forms of As and Se in aquatic environments: Effects of natural organic matter. Mid-year report for EPA agreement number R829515C003. Colorado State University, Fort Collins, Colorado.

Macdonald D.D., Roberts B., and Hyne J.B. (1978a) Corrosion Science 18, 411.

Macdonald D.D., Roberts B., and Hyne J.B. (1978b) Corrosion Science 18, 499.

MacGregor R.A. (1997) The geochemistry of selenium in sedimentary environments: examples from the UK and Jordan. University of Reading. Thesis submitted for the degree of Doctor of Philosophy in May 1997. Internet address as accessed on 2004-04-29: http://ourworld.compuserve.com/homepages/robmac/front.htm

Maes A., Bruggeman C., Van Geet M., Deniau I., and Largeau C. (2003) Topical report on methodologies developed to identify the source of mobile Boom Clay organic matter. 29 pp. Project Trancom-II: Migration case study: transport of radionuclides in a reducing clay sediment. SCK•CEN-BLG-967, SCK•CEN, Mol, Belgium.

Maes A., Bruggeman C., and Vancluysen J. (2004a) Selenium behaviour in Boom Clay. First progress report (07/2004). Work performed for ONDRAF/NIRAS under contract CCHO- 2004-004862. Katholieke Universiteit Leuven, Faculteit bio-ingenieurs-Wetenschappen, Departement Microbiële en Moleculaire Systemen (m2s), Laboratorium voor Colloidchemie.

Maes A., Bruggeman C., and Vancluysen J. (2004b) Selenium behaviour in Boom Clay. Second progress report (01/2005). Work performed for ONDRAF/NIRAS under contract CCHO-2004-004862. Katholieke Universiteit Leuven, Faculteit bio-ingenieurs- Wetenschappen, Departement Microbiële en Moleculaire Systemen (m2s), Laboratorium voor Colloidchemie.

Maes A., Bruggeman C., and Vancluysen J. (2005) Selenium behaviour in Boom Clay. Third progress report (07/2005). LCC-R.Se-07/2005. Work performed for ONDRAF/NIRAS under contract CCHO-2004-004862. Preliminary Report. Katholieke Universiteit Leuven, Faculteit bio-ingenieurs-Wetenschappen, Departement Microbiële en Moleculaire Systemen (m2s), Laboratorium voor Colloidchemie.

Maes N. (2004c) Natural concentration of selenium measured in Boom Clay water with high resolution ICP-MS (Royal Museum of Central Africa, Tervuren, Belgium). Decrease of selenium concentration in water after flocculation of humic acids by water acidification. Personal Communication.

113/328

Maes N., Moors H., De Cannière P., Aertsens M., and Put M. (1998) Determination of the diffusion coefficient of ionic species by electromigration: feasibility study. Radiochimica Acta 82, 183–189.

Maes N., Moors H., Dierckx A., De Cannière P., and Put M. (1999) The assessment of electromigration as a new technique to study diffusion of radionuclides in clayey soils. Journal of Contaminant Hydrology 36 (3-4), 231–247.

Maes N., Moors H., Dierckx A., Aertsens M., Wang L., De Cannière P., and Put M. (2001) Studying the migration behaviour of radionuclides in Boom Clay by electromigration. pp. 35– 1 to 35–31. In: EREM 2001, 3rd symposium and status report on electrokinetic remediation, (C. Czurda et al. ed., 2001) Schriftenreihe Angewandte Geologie Karlsruhe, 63. ISSN 0933- 2510.

Maes N., Moors H., Wang L., Delécaut G., De Cannière P., and Put M. (2002) The use of electromigration as a qualitative technique to study the migration behaviour and speciation of uranium in the Boom Clay. Radiochimica Acta 90, 741–746.

Maes N., De Cannière P., Sillen X., Van Ravestyn L., and Put M. (2004d) DS 2.6 Migration of iodide in backfill materials containing an anion getter. Phase 2: Performance tests on active carbon under geological conditions – In search of an alternative. Final Report. Restricted SCK•CEN report R-3896 published on 2004-08-12.

Maes N., Wang L., Delécaut G., Beauwens T., Van Geet M., Put M., Weetjens E., Marivoet J., van der Lee J., Warwick P., Hall A., Walker G., Maes A., Bruggeman C., Bennett D., Hicks T., Higgo J., and Galson D. (2004e) Migration case study: transport of radionuclides in a reducing clay sediment (TRANCOM-II). Final scientific and technical report of the EC TRANCOM-II project. SCK•CEN-BLG-988 (04/NMa/P-50). Contract N° FIKW-CT-2000- 00008. SCK•CEN, Waste and Disposal Department – R&D Geological Disposal, Mol, Belgium, September 2004. Natural concentration of selenium measured in Boom Clay: see Annex 14, pp. 26 – 27 and p. 35.

Maes N., Wang L., Delécaut G., Beauwens T., Van Geet M., Put M., Weetjens E., Marivoet J., van der Lee J., Warwick P., Hall A., Walker G., Maes A., Bruggeman C., Bennett D., Hicks T., Higgo J., and Galson D. (2004f) Migration case study: transport of radionuclides in a reducing clay sediment (TRANCOM-II). Final report EUR 21022 EN. European Commission, Nuclear Science and Technology, Luxembourg. Euratom 5th Framework Program period (1998 – 2002). Contract N° FIKW-CT-2000-00008.

114/328

Maes N., Delécaut G., Wang L., Cachoir C., Salah S., and Lemmens K. (2005) Uranium retention and migration behaviour in Boom Clay. Topical Report. SCK•CEN.

Malen K.A. Release of tellurium and cesium from UO2 in LWR fuel rods during irradiation. Unreferenced report from Studsvik Energyteknik A.B., Nyköping, Sweden.

Mallants D., Sillen X., and Marivoet J. (1999) Geological disposal of conditioned high-level and long-lived radioactive waste. Consequence analysis of the disposal of vitrified high-level waste in the case of the normal evolution scenario. R-3383, Restricted report to ONDRAF/NIRAS for contract CCHO-95/268 – KNT 90.94.601, Annex 6, Task 6.1. Waste and Disposal Department SCK•CEN, Mol Belgium. December 1999. 82 pp.

Manceau A. and Charlet L. (1994) The mechanism of selenate adsorption on goethite and hydrous ferric oxide. Journal of Colloid and Interface Science 168, 87–93.

Marin L., Lhomme J., and Carignan J. (2001) Determination of selenium concentration in sixty five reference materials for geochemical analysis by GFAAS after separation with thiol cotton. Geostandards Newsletter, 25 (2/3), 317–324.

Marcus P. (1995) Sulfur-assisted corrosion mechanisms and the role of alloyed elements. In: Corrosion mechanisms in theory and practice. Marcus P. and Olefjord I., eds. Marcel Dekker. New York. p. 239.

Marivoet J. and Bonne A. (1988) PAGIS, Performance Assessment of Geological Isolation Systems for radioactive waste. Disposal in clay formations. CEC, Luxembourg, Report EUR 11 776 EN.

Marivoet J. and Zeevaert T. (1991) PACOMA – Performance assessment of the geological disposal of medium level and alpha waste in a clay formation in Belgium. CEC, Luxembourg, Report EUR 13 042 EN.

Marivoet J. (1991) UPDATING 1990. Updating of the performance assessments of the geological disposal of high-level and medium-level waste in the Boom Clay Formation. ONDRAF/NIRAS report BLG-634.

Marivoet J., Volckaert G., Labat S., De Cannière P., Dierckx A, Kursten B., Lemmens K., Lolivier P., Mallants D., Sneyers A., Valcke E., Wang L., and Wemaere I. (1999a) Values for the near-field and clay parameters used in the performance assessment of the geological disposal of radioactive waste in the Boom Clay formation at the Mol site (volume 1 and 2). Report to NIRAS/ONDRAF. Geological disposal of conditioned high-level and long-lived

115/328

radioactive waste. Contract CCHO-98/332 – KNT 90.98.1042 Task 6.1. SCK•CEN Report R-3344 (July 1999).

Marivoet J. and Weetjens E. (2007) The importance of mobile fission products for long-term safety in the case of disposal of vitrified high-level waste and spent fuel in a clay formation. Mobile fission and activation products in nuclear waste disposal. International workshop. L’Hermitage, La Baule – France, January 16-19, 2007. http://mofap07.in2p3.fr/. Proceedings to appear as a Nuclear Energy Agency (NEA, OECD) publication.

Martinez M., Gimenez J., de Pablo J., Rovira, M., and Duro L. (2006) Sorption of selenium(IV) and selenium(VI) onto magnetite. Applied Surface Science 252 (10), 3767– 3773. Unique item number: RN182415520. Shelfmark: 1580.082000. ISSN: 0169-4332.

Masscheleyn P.H., Delaune R.D., and Patrick W.H. Jr (1990) Transformations of selenium as affected by sediment oxidation-reduction potential and pH. Environmental Science and Technology 24(1), 91–96.

Masscheleyn P.H., Delaune R.D., and Patrick, W.H. (1991) Biogeochemical behaviour of selenium in anoxic soils and sediments – an equilibrium thermodynamics approach. Journal of Environmental Science and Health part A – Environmental Science and Engineering & Toxic and Hazardous Substance Control 26(4), 555–573.

Matthiesen A. (1994) Evaluating the redox capacity and the redox potential of humic acids by redox titrations. In: Senesi N. and Miano T.M. (Eds.) (1994) Humic substances in the global environment and implications on human health. Amsterdam, Elsevier.

Matzke Hj. (1995) Oxygen potential measurements in high burnup LWR UO2 fuel. Journal of Nuclear Materials 223, 1–5. (key article).

Mazurek M., Alt-Epping P., Gimmi T., Waber H.N., Bath A., Buschaert S., and Gautschi A. (2007a) Tracer profiles across argillaceous formations: A tool to constrain transport processes. In Proc. 12th International Symposium on Water-Rock Interaction (WRI-12) held at Kunming (Yunnan province, Popular Republic of China) from July 31 to August 05, 2007. http://www.wri12.org/. 4 pp.

Mazurek M., Alt-Epping P., Bath A., Buschaert S., Gautschi A., Gimmi T., and Waber H.N. (2007b) CLAYTRAC project: evaluation of tracer profiles across argillaceous formations. [Oral presentation: O/01/2]. In Proc. Clays in natural and engineered barriers for radioactive waste confinement – Lille 2007 – 3rd International Meeting. Organised by Andra in Lille (France), 17 – 20 September 2007.

116/328

Mazurek Martin, Peter Alt-Epping, Adrian Bath, Thomas Gimmi, and H. Niklaus Waber with contributions by Stéphane Buschaert & Agnès Vinsot, Mieke De Craen, Isabelle Wemaere & Pierre De Cannière, Andreas Gautschi, Sébastien Savoye, and Laurent Wouters (2009) Natural tracer profiles across argillaceous formations: The CLAYTRAC project - review and synthesis Report. NEA Report N° 6253, 362 pp. Clay Club, Radioactive Waste Management, Nuclear Energy Agency (NEA), Organisation for Economic Co-operation and Development (OECD). Publication date: 31 Mar 2009. Language: English. (OECD Code: 66 2009 02 1 P). Price: EUR 75. ISBN 978-92-64-06047-0. N° 56717 2009. http://www.nea.fr | http://www.oecd.org. OECD Publishing, 2 rue André-Pascal, F-75775 Paris Cedex 16. Printed in France. http://www.oecdbookshop.org/oecd/display.asp?sf1=identifiers&st1=978-92-64-06047-0 http://www.oecdbookshop.org/oecd/get-it.asp?REF=6609021E.PDF&TYPE=browse

McAninch J.E., Bench G.S., Freeman S.P.H.T., Roberts M.L., Southon J.R., Vogel J.S., and Proctor I.D. (1994) PXAMS – Projectile X-ray Accelerator Mass Spectrometry: X-ray yields and applications. 13th international conference on the application of accelerators in research and industry; 7-10 Nov 1994; Denton, TX (United States). (Lawrence Livermore National Lab., CA (United States). Center for Accelerator Mass Spectrometry). Report Number(s) UCRL-JC-118859; CONF-941129—13. DOE Contract N° W-7405-ENG-48. See pp. 5 and 6, the need to remeasure the half-life of 79Se. See also the citation [8] of Singh B. (1993) Nuclear Data Sheets 70, 437. Question on the validity of the half-life of 79Se.

McKenzie R.C., Rafferty T.S., Beckett G.J. (1998) Selenium: an essential element for immune function. Immunology Today 19, 342–345.

McKnigh, D., Cory R., Hannigan R., Dunagan S., Gilmore M., Varekamp J., et al. (2005) Oxidation/reduction reactions involving dissolved humics as electron shuttlers in lake sediments (p. 07). Presented at the AGU Fall Meeting Abstracts.

Mees D.R., Pysto W., and Tarcha P.J. (1995) Formation of selenium colloids using sodium ascorbate as the reducing agent. Journal of Colloid and Interface Science 170, 254–260.

Mehra O.P. and Jackson M.L. (1960) Iron oxide removal from soils and clays by a dithionite- citrate system buffered with sodium bicarbonate. Clays and Clay Minerals 7, 317–327.

Merceron T., Mossmann J.R., Neerdael B., De Cannière P., Beaucaire C., Daumas S., Bianchi A. and Christen R. (1993) Archimedes – Clay Project: acquisition and regulation of the water chemistry in clay environment: 4th progress report, December 1993. Work coordinated by ANDRA in the frame of the CEC contract F12W-CT92-0117.

117/328

Merceron T. (1994) Characterisation of the geochemical environment of the Boom clay at Mol: Archimedes – Clay Project. Paper presented at the MIRAGE meeting on the Migration of Radionuclides in the Geosphere – Mirage project – 3rd Phase. Final meeting 1994, held in Brussels on the 15, 16 and 17th of November 1994.

Merceron T. (1995) Archimedes-Clay project: Acquisition and regulation of water chemistry in clay. In "Migration of radionuclides in the geosphere" (Mirage project - Third phase) - Proceedings of a progress meeting (work period 1992), Brussels, 7 and 8 October 1993. EUR 15914.

Mills K.C. (1974) Thermodynamic data for inorganic sulphides, selenides and tellurides. Butterworths London.

Mitchell J.K. (1993) Fundamentals of soil behaviour. 2nd edition. John Wiley & Sons, Inc.

MOFAP (2007) Mobile fission and activation products in nuclear waste disposal. International workshop. L’Hermitage, La Baule – France, January 16-19, 2007. http://mofap07.in2p3.fr/. Proceedings to appear as a Nuclear Energy Agency (NEA, OECD) publication.

Moffatt W.G. (1976) Binary phase diagrams handbook. See phase diagrams for the Li-S, Li- Se, Li-Te and K-Se systems. General Electric, Schenectady, NY.

Molinski V.J., and Leddicotte G.W. (1965) Radiochemistry of selenium. 49 pp. (Nuclear Science series / National Academy of Sciences-National Research Council, NAS-NS 3030, rev.). Publisher: the Clearinghouse for Federal Scientific and Technical Information, National Bureau of Standards, U.S. Dept. of Commerce; Rev. 1965 edition. Springfield, Va. ASIN: B0006ETRJ0.

Mompean F., Illemassène M., and Perrone J. (2005) Chemical thermodynamics of compounds and complexes of U, Np, Pu, Am, Tc, Se, Ni, and Zr with selected organic ligands. Chemical Thermodynamics Series, Volume 9. Elsevier Science. ISBN: 0-444-51402-3, 1132 pp.

Montavon G. and Grambow B. (2006) Subatech (Rennes, France) Funmig EC project.

Morel F. (1983) Principles of aquatic chemistry. Wiley-Interscience, New York.

Morgenstern U., Fifield L.K., Tims S.G., and Ditchburn R.G. (n.d.) Beam interactions with Materials and Atoms: Progress in AMS measurement of natural 32Si for glacier ice dating.

118/328

Nuclear Instruments and Methods in Physics Research Section B. doi:10.1016/j.nimb.2009.10.019.

Mukai S., Kitao H., and Kataoka S. (1998) Advective diffusion experiment using I, Se, and Cs in fracture rocks. 1998 Fall meeting of the Atomic Energy Society of Japan 3, p. 842 (in Japanese).

Muller John (2006) Étude du cycle biogéochimique du sélénium dans l’environnement. Directeur de thèse: Abdelouas Abdesselam. Exposé Heures Thésards Subatech, 04/12/2006, Ecole des Mines de Nantes.

Muller J. (2008) Étude du cycle biogéochimique du sélénium dans l’environnement. PhD Thesis. Subatech, Ecole des Mines de Nantes.

Myneni S.C.B. and Tokunaga T.K. (1997) 213 ACS national meeting, Washington DC, P. 1077 Paper GEOC 141, 2904 pp.

Myneni S.C.B., Tokunaga T.K., and Brown G.E. (1997) Abiotic selenium redox transformations in the presence of Fe (II, III) oxides. Science 278 (5340), 1106–1109.

Nagra (1985a) Project Gewähr 1985. Endlager für hochaktive abfälle: bautchnik und betriebsphase. NGB 85-03. Nagra, Baden, Switzerland.

Nagra (1985b) Project Gewähr 1985. Nuclear waste management in Switzerland: feasibility studies and safety analyses. NGB 85-09. Nagra, Baden, Switzerland.

Nagra (1994a) Kristallin-I. Geology and Hydrogeology of the Crystalline Basement of Northern Switzerland. Nagra Technical Report, NTB 93-01. Nagra, Wettingen, Switzerland.

Nagra (1994b) Kristallin-I. Safety Assessment Report. Nagra Technical Report, NTB 93-22. Nagra, Wettingen, Switzerland.

Nagra (2002a) Projekt Opalinuston – Synthese der geowissenschaftlichen untersuchungsergebnisse. Entsorgungsnachweis für abgebrannte brennelemente, verglaste hochaktive sowie langlebige mittelaktive abfälle. Nagra Technical Report NTB 02-03. Nagra, Wettingen, Switzerland.

Nagra (2002b) Project Opalinus Clay: Safety report. Demonstration of disposal feasibility for spent fuel, vitrified high-level waste and long-lived intermediate-level waste (Entsorgungsnachweis). Nagra Technical Report NTB 02-05. Nagra, Wettingen, Switzerland.

119/328

Nagra (2008) Bioprota workshop. International Forum on Se-79 in the Biosphere. 5–6 May 2008. Hosted by Nagra, Wettingen, Switzerland.

Nakayasu K., Fukushima M., Sasaki K., Tanaka S., and Nakamura H. (1999) Comparative studies of the reduction behavior of chromium(VI) by humic substances and their precursors, Environmental Toxicology and Chemistry 18(6), 1085–1090.

Naveau A., Monteil-Rivera F., Guillon E., and Dumonceau J. (2007) Interactions of aqueous selenium (−II) and (IV) with metallic sulfide surfaces. Environmental Science and Technology 41 (15), 5376–5382. doi:10.1021/es0704481.

Newman D.K., Ahmann D., and Morel F.M.M.(1998) A brief review of microbial arsenate respiration. Geomicrobiology Journal 15, 255–268.

Nicoll S., Matzke Hj., Grimes R.W., and Catlow C R.A. (1997) The behaviour of single atoms of Mo in urania. Journal of Nuclear Materials. 240, 185.

Nirond (2004) A review of corrosion and material selection issues pertinent to underground disposal of highly active nuclear waste in Belgium. A report for ONDRAF/NIRAS prepared by the Corrosion Study Panel. Nirond (2004-02). See, Section 3.2 Sulfur-water system: p. 11– 12. See also, Section 4.2 The sulfur/water systems p. 55–56.

Noläng B. ed. (2005) In Book: Chemical thermodynamics of selenium. Chemical Thermodynamics Series, Volume 7. Elsevier Science. ISBN: 0-444-51403-1, 851 pp. Edited by Mompean F.J., Perrone J., and Illemassène M., OECD, Nuclear Energy Agency, Data Bank, Issy-les-Moulineaux (France).

Nriagu J.O. (Editor) (1976) Environmental biogeochemistry: carbon, nitrogen, phosphorus, sulfur and selenium cycles. (Volume 1). 423 pp. Ann Arbor Science. ASIN: B000I3YTLA.

NuDat 2.1 (2005) Nuclear Database version 2.1 from the US National Nuclear Data Center (NNDC), Brookhaven National Laboratory (BNL). URL: http://www.nndc.bnl.gov/nudat2 as seen on 02-Feb-2005.

Obata T., Araie H., and Shiraiwa Y. (2003) Kinetic studies on bioconcentration mechanism of selenium by a coccolithophorid, Emiliania huxleyi. Plant Cell Physiology 44, S43.

120/328

Obata T., Araie H., and Shiraiwa Y. (2004) Bioconcentration mechanism of selenium by a coccolithophorid, Emiliania huxleyi. Plant and Cell Physiology 45 (10), 1434–1441. doi:10.1093/pcp/pch164.

Obata T. and Shiraiwa Y. (2005) A novel eukaryotic selenoprotein in the haptophyte alga Emiliania huxleyi. Journal of Biological Chemistry 280 (18), 18462.

Ochs M., Lothenbach B., and Giffaut E. (2002) Uptake of oxo-anions by cements through solid-solution formation: experimental evidence and modelling. Radiochimica Acta, 90 (9), 639–646. doi:10.1524/ract.2002.90.9-11_2002.639.

ODS – NIH (2004) Dietary supplement fact sheet: selenium. 9 pp., 66 references. Document last updated: 08/01/2004. http://ods.od.nih.gov/factsheets/Selenium_pf.asp. Office of Dietary Supplements (ODS). National Institutes of Health (NIH). Bethesda, Maryland 20892 USA. http://ods.od.nih.gov.

OECD/NEA (1991) Review of safety assessment methods, disposal of radioactive waste. A report of the Performance Assessment Advisory Group of the Radioactive Waste Management Committee.

OECD/NEA (1992) Systematic approaches to scenario development. A report of the NEA working group on the identification and selection of scenarios for performance assessment of radioactive waste disposal.

OECD/NEA (1993) The role of conceptual models in demonstrating repository post-closure safety. Proceedings of a NEA Workshop, Paris, 16-18 November 1993. Nuclear Energy Agency (NEA). OECD Documents. 190 pp. ISBN 92-64-14429-3.

OECD/NEA (1997) Safety assessment of radioactive waste repositories – Systematic approaches to scenario development – An international database of Features, Events and Processes. Draft report (24/6/1997) of the NEA working group on development of a database of Features, Events and Processes relevant to the assessment of post-closure safety of radioactive waste repositories.

OECD/NEA (1999) Actinide and fission product partitioning and transmutation. status and assessment report.

Oger P.M., Daniel I., Cournoyer B., and Simionovici A. (2004) In situ micro X-ray absorption 2– near-edge structure study of microbiologically reduced selenite (SeO3 ). Spectrochimica Acta Part B 59, 1681–1686.

121/328

Oldfield J.E. (1999) Selenium Se world atlas. Publisher: Selenium-Tellurium Development Association. ASIN: B000H5RMK4.

Olin Å. (Chairman), Noläng B., Öhman L-O., Osadchii E.G., and Rosén E. (2005) Chemical thermodynamics of selenium. Chemical Thermodynamics Series, Volume 7. Elsevier Science. ISBN: 0-444-51403-1, 851 pp. Edited by Mompean F.J., Perrone J., and Illemassène M., OECD, Nuclear Energy Agency, Data Bank, Issy-les-Moulineaux (France).

Olyslaegers G. (2008) Concise mission report. SCK•CEN Report N° 125-14140 (2008-05- 07). Participation to the BIOPROTA International Forum on Se-79 in the Biosphere, 5 and 6 May 2008. Nagra, Wettingen (CH).

Olyslaegers G. (2008) In: Smith K. (Ed.) (2008) Report of Se-79 in the Biosphere. BIOPROTA Workshop, hosted by Nagra, Wettingen (CH), 5–6 May 2008, V2.0, Final. BIOPROTA – http://www.bioprota.com.

ONDRAF/NIRAS (2001) SAFIR 2 Report, coordinated by De Preter P., Lalieux P., and Cool W. (eds.). Safety Assessment and Feasibility Interim Report 2. Four volumes, 13 Chapters. ONDRAF/NIRAS, Belgian Agency for Radioactive Waste and Enriched Fissile Materials. NIROND 2001-06 E. December 2001.

ONDRAF/NIRAS (2004) Multi-criteria analysis on the selection of a reference EBS design for vitrified high-level waste. NIROND, Brussels. Report NIROND 2004-03 E.

Oppenheimer J.A. (1984) Speciation of selenium in groundwater. 29 pp. United States Environmental Protection Agency, Office of Research and Development. ASIN: B0006YT0T2.

Oram L.L., Strawn D.G., Marcus M.A., Fakra S.C., and Möller G. (2008) Macro- and microscale investigation of selenium speciation in Blackfoot River, Idaho sediments. Environmental Science and Technology 42 (18), 6830–6836. doi:10.1021/es7032229.

Oremland R.S. (1991) Selenate removal from waste water. Patent.

Oremland R.S. (1994) Biogeochemical transformations of selenium in anoxic environments. In: Frankenberger W.T., Jr., and Benson S.N. (Eds.) (1994) Selenium in the Environment: Marcel Dekker. New York, N.Y. pp. 389–419.

122/328

Oremland R.S., Blum J.S, Culbertson C.W., Visscher P.T., Miller L.G., Dowdle P., and Strohmaier R.E. (1994) Isolation, growth, and metabolism of an obligately anaerobic, selenate-respiring bacterium, strain SES-3. Applied and Environmental Microbiology 60(8), 3011–3019.

Oremland R.S., Blum J.S., Bindi A.B., Dowdle P.R., Herbel M., Stolz J.F. (1999) Simultaneous reduction of nitrate and selenate by cell suspensions of selenium-respiring bacteria. Applied and Environmental Microbiology 65(10), 4385–4392. Retrieved from http://aem.asm.org/cgi/content/abstract/65/10/4385.

Oremland R.S., Herbel M.J., Blum J.S., Langley S., Beveridge T.J., Ajayan P.M. (2004) Structural and spectral features of selenium nanospheres produced by Se-respiring bacteria. Applied and Environmental Microbiology 70(1), 52–60. Retrieved from http://aem.asm.org/cgi/content/abstract/70/1/52.

Oremland R.S., Hollibaugh J.T., Maest A.S., Presser T.S., Miller L.G., and Culbertson C.W. (1989) Selenate reduction to elemental selenium by anaerobic bacteria in sediments and culture: biogeochemical significance of a novel, sulfate-independent respiration. Applied and Environmental Microbiology 55(9), 2333–2343. Retrieved from http://aem.asm.org/cgi/content/abstract/55/9/2333.

Oremland R.S., Kulp T.R., Blum J.S., Hoeft S.E., Baesman S., Miller L.G., et al. (2005) A Microbial arsenic cycle in a salt-saturated, extreme environment. Science, 308(5726), 1305– 1308. Retrieved from http://www.sciencemag.org/cgi/content/abstract/308/5726/1305.

Oremland R.S., Steinberg N., Maest A., Miller L., and Hollibaugh J. (1990) Measurement of in situ rates of selenate removal by dissimilatory bacterial reduction in sediments. Environmental Science and Technology 24(8), 1157–1164.

Oremland R.S., Steinberg N.A., Presser T.S., and Miller L.G. (1991) In situ bacterial selenate reduction in the agricultural drainage systems of western Nevada. Applied and Environmental Microbiology 57(2), 615–617. Retrieved from http://aem.asm.org/cgi/content/abstract/57/2/615.

Oremland R.S. and Stolz J.F. (2003) The ecology of arsenic. Science 300(5621), 939–944. doi:10.1126/science.1081903. http://www.sciencemag.org/cgi/content/abstract/300/5621/939.

Oremland R.S., Stolz J., and Lovley D. (1998) “Green rust” in the lab and in the soil. Science, 281(5380), 1111. Technical comment.

123/328

Oremland R.S., Stolz J., Lovley D., Myneni S.C., Tokunaga T.K., Brown G.E., et al. (1998) “Green rust” in the lab and in the soil. Science, 281(5380), 1111. Retrieved from http://www.sciencemag.org.

Oremland R.S., Stolz J., and Hollibaugh J. (2004) The microbial arsenic cycle in Mono Lake, California. FEMS Microbiology Ecology 48(1), 15–27.

Oremland R.S., and Zehr J.P. (1986) Formation of methane and carbon dioxide from dimethylselenide in anoxic sediments and by a methanogenic bacterium. Applied and Environmental Microbiology 52(5), 1031–1036. Retrieved from http://aem.asm.org/cgi/content/abstract/52/5/1031.

Ortiz L., Volckaert G., and Mallants D. (2002) Gas generation and migration in Boom Clay, a potential host rock formation for nuclear waste storage. Engineering Geology 64, 2-3, 287– 296. doi:10.1016/S0013-7952(01)00107-7. See: Microbial activity in Boom Clay.

Parfitt R.L. and Smart R.S.C. (1977) Infrared spectra from binuclear bridging complexes of sulphate adsorbed on goethite (α-FeOOH). Journal of the Chemical Society Faraday Transaction 1, 73, 796–802.

Parfitt R.L. and Smart R.S.C. (1978) The mechanism of sulfate adsorption on iron oxides. Soil Science Society of America Journal 42, 48–50.

Parfitt R.L. Atkinson R.J., and Smart R.S.C. (1975) The mechanism of phosphate fixation by iron oxides. Soil Science Society of America Journal 39, 837–841.

Parfitt R.L., Russell J.D., and Farmer V.C. (1976) Confirmation of the structures of goethite (α-FEOOH) and phosphated goethite by infrared spectroscopy. Journal of the Chemical Society Faraday Transaction 1, 72, 1082–1087.

Parker G.W., Creek G.E., Hebert G.M., Lantz P.M., and Martin W.J. (1949) Oak Ridge National Laboratory Classified Report ORNL-499, December 1949 (unpublished). Cited as reference “ParkG49a”. In: Lederer C.M., Hollander J.M., and Perlman I. (1967) Table of Isotopes, sixth edition, p. 30, p. 210, p. 212, p. 214, and p. 527. John Wiley & Sons, Inc., New York • London • Sydney.

Parker J. (2004) Feasibility study for the reduction of colour within the glass furnace. Written by the Glass Technology Service (GTS) and the University of Sheffield, Engineering Materials Department for the Waste and Resources Action Programme (WRAP).

124/328

Project code: GLA0023. ISBN: 1-84405-090-4. http://www.wrap.org.uk/downloads/ColourReduction.b4a2fccd.pdf

Parkhurst D.L. and Appelo C.A.J. (1999) User’s guide to PHREEQC (Version 2) – A computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations. U.S. Geological Survey Water-resources Investigations Report 99- 4259. United States Geological Survey website as viewed on 23 September 2004: http://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqc/index.html, Note: “wwwbrr”, without a dot after www, is exactly written as displayed on the USGS web site. With a dot, the IP address of the web site cannot be found.

Parrington J.R., Knox H.D., Breneman S.L., Baum E.M., and Feiner F. (1996) Nuclides and isotopes. Fifteenth edition. Chart of the nuclides. Revised 1996 Courtesy of Knolls Atomic Power Laboratory (KAPL Inc.) a Lockheed Martin Company – General Electric Nuclear 5 79 79 Energy. T½ = 6.5 × 10 years for Se. Se: Z = 34 protons; N = 45 neutrons. See top of p. 26.

Pascal P. et Lumbroso H. (1977) Nouveau traité de chimie minérale. Compléments 7. Selenium. Par H. Lumbroso. XII, 290. Masson, Paris. ISBN 2-225-42988-X.

Patai S. and Rappoport Z. (1986) The chemistry of organic selenium and tellurium compounds. Wiley, Chichester. ISBN: 0471904252.

Pauwels H., Dictor M.-C., et Garrido F. (2009) Etude expérimentale du comportement biogéochimique des nitrates – Essais en présence d’échantillons du site du Mont Terri (projet BN-nitrates). Rapport final. BRGM/RP – 57387-FR. Juin 2009. 29 pp.

Pauwels H., Dictor M.-C., and Garrido F. (2010) Experimental study of the biogeochemical behaviour of nitrates – Tests in the presence of samples from the Mont Terri site. Final report of the lab supporting microbial tests made at BRGM for the Mont Terri Bitumen Nitrate (BN) experiment. Mont Terri Technical Note (TN). 29 pp.

Peak D., Ford R.G., and Sparks D.L. (1999) An in situ ATR-FTIR investigation of sulfate bonding mechanisms on goethite. Journal of Colloid and Interface Science 218(1), 289–299.

Peak D., Saha U.K., and Huang P.M. (2006) Selenite adsorption mechanism on pure and coated montmorillonite: an EXAFS and XANES spectroscopic study. Soil Science Society of America Journal 70, 192–203.

125/328

Pearson J.F.J. and Berner U. (1991) NAGRA Thermochemical Data Base I. Core data. Nagra Technical Report NTB 91-17. NAGRA, Wettingen.

Pearson J.F.J., Berner U., and Hummel W. (1992) NAGRA Thermochemical Data Base II. Supplemental data 05/92. Nagra Technical Report NTB 91-18. NAGRA, Wettingen.

Pearson F.J., Scholtis A., Gautschi A., Baeyens B., Bradbury M., and Degueldre C. (1999) Chemistry of porewater. In: Mont Terri Rock Laboratory: Results of the hydrogeological, geochemical and geotechnical experiments performed in 1996 and 1997 (Ed. Burkhalter, R.). Landeshydrologie und -geologie. Geologische Berichte Nr. 23. 129–147.

Pearson F.J., Arcos D., Bath A., Boisson J.-Y., Fernández A.M., Gäbler H.-E., Gaucher E., Gautschi A., Griffault L., Hernán P., and Waber H.N. (2003) Mont Terri Project – Geochemistry of water in the Opalinus Clay Formation at the Mont Terri Rock Laboratory – Synthesis Report. Berichte des BWG, Serie Geologie – Rapports de l'OFEG, Série Géologie – Rapporti dell'UFAEG, Serie Geologica – Report of the Swiss Federal Office for Water and Geology (FOWG), Geology Series. N°5, 319 pp. ISBN 3-906723-59-3. Bern, Switzerland.

Peitzsch Mirko and Kersten Michael (2008) Microbiological alkylation and volatilization of inorganic selenium from Se-LDH and ferroselite. Poster presented at the 2nd International Workshop on: Mechanisms and modelling of waste/cement interactions. October 12-16, 2008, Le Croisic (France). http://cement08.in2p3.fr.

Pejova B. and Grozdanov I. (2001) Solution growth and characterization of amorphous selenium thin films – Heat transformation to nanocrystalline gray selenium thin films. Applied Surface Science 177(3), 152–157.

Perkins R. and Foster A. (n.d.) Mineral affinities and distribution of selenium and other trace elements in black shale and phosphorite of the Phosphoria Formation. Life Cycle of the Phosphoria Formation: From Deposition to Post-Mining Environment 251–295.

Peters M.G., Maher W.A., Barford J.P., and Gomes V.G. (1997) Selenium association in estuarine sediments: redox effects. Water, Air and Soil Pollution 99, 275–282.

Peynet V. (2003) Rétention d’actinide et de produits de fission par des phases solides polyminérales. 269 pp. Voir Chap. III et IV: Adsorption et rétention du césium, de l’américium (III) et du sélénium (IV) en traces par la montmorillonite, la goethite et la calcite, en phase pure ou en mélange. Thèse de doctorat de l’Université Paris 6, spécialité Chimie Analytique. Thèse soutenue le 10 juillet 2003 pour l’obtention du grade de Docteur de l’Université Paris 6.

126/328

Pfennig G., Klewe-Nebenius H., and Seelmann-Eggebert W. (1995) Karlsruher Nuklidkarte, 4 79 6. Auflage 1995. T½ < 6.5 × 10 years for Se. Edited by Forschungszentrum Karlsruhe (FZK) GmbH. 79Se: Z = 34 protons; N = 45 neutrons. See 4th flaps.

Pickering I.J., Prince R.C., Salt D.E., and George G.N. (2000) Quantitative, chemically specific imaging of selenium transformation in plants. Proceedings of the National Academy of Science of USA 97(20), 10717–10722. Published online 2000 September 12. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=27089.

Pickering I.J., Wright C., Bubner B., Ellis D., Persans M.W., Yu E.Y., George G.N., Prince R.C., and Salt D.E. (2003) Chemical form and distribution of selenium and sulfur in the selenium hyperaccumulator astragalus bisulcatus. Plant Physiology 131, 1460–1467. 10.1104/pp.014787. http://www.plantphysiol.org/cgi/content/full/131/3/1460.

Pierce M.L. and Moore C.B. (1980) Adsorption of arsenite on amorphous iron hydroxide from dilute aqueous solution. Environmental Science and Technology 14, 214–216.

Pilkington N.J. (1998) Minutes of progress meeting on radionuclide solubilities in Boom Clay, held on 28 April 1998.

Pirlet V. (2005) Determination of the mobile leached concentrations of 79Se in near-field conditions. In: Characterization and compatibility with the disposal medium of Cogema and Eurochemic reprocessing waste forms (Tasks VM-6 and GV8 of NIRAS/ONDRAF contracts CCHO-90/123-1 and CCHO-90/123-2 – vitrified waste). The importance of the glass composition and the near-field for the mobile concentrations of radionuclides. Summary of the topical report for WP4 of RP.WD.008 (January 2000 – June 2003).

Plant J.A., Kinniburgh D.G., Smedley P.L., Fordyce F.M., and Klinck B.A. (2004) Chap. 9.02 Arsenic and selenium, pp. 17 – 66. In Environmental Geochemistry (ed. Lollar B.S.) Vol. 9 Treatise on Geochemistry (eds. Holland H.D. and Turekian K.K., executives editors) (Vol. 9, ISBN: 0-08-044344-3). Elsevier Pergamon Oxford. ISBN (set): 0-08-043751-6. http://www.sciencedirect.com/science/referenceworks/0080437516.

Plotnikov V.I. (1958) Coprecipitation of small quantities of selenium with ferric hydroxide. Russian Journal of Inorganic Chemistry 3(8),1761–1766 [pp. 56–64 in English translation].

Plotnikov V.I. (1960) Coprecipitation of selenium and tellurium with metal hydroxides. Russian Journal of Inorganic Chemistry 5, 351–354.

127/328

Plotnikov V.I. (1964) Coprecipitation of small amounts of selenium and zinc with metal hydroxides. Russian Journal of Inorganic Chemistry 9, 245–248.

PNC (1992) Research and development on geological disposal of high-level radioactive waste, first progress report – H3. PNC TN1410 93-059.

Poinssot C., Baeyens B., and Bradbury M.H. (1999) Experimental and modelling studies of caesium sorption on illite. Geochimica et Cosmochimica Acta 63(19/20), 3217–3227.

Pointeau I., Hainos D., Coreau N., and Reiller P. (2006) Effect of organics on selenite uptake by cementitious materials. Waste Management, 26 (7), 733–740. doi:10.1016/j.wasman.2006.01.026.

Pointeau I., Coreau N., Hainos D., and Reiller P. (2006) Influence of organic ligands on the radionuclides uptake [Se(IV) and U(VI)] by cementitious materials. CEA Saclay/DEN/DANS/DPC/SECR/L3MR, lecture given to the 2006 Summer School on Geochemistry and Migration of Actinides — Actinide Behaviour in Natural Environment (Actinet Network of Excellence).

Pourbaix M. (1966) Atlas of electrochemical equilibria in aqueous solutions. 645 pp. Pergamon Press. Oxford.

Pourbaix M. (1976) Atlas of electrochemical equilibria. NACE, Houston, TX.

Presser T., Piper D., Bird K., Skorupa J., Hamilton S., Detwiler S., et al. (2004) The Phosphoria Formation: A model for forecasting global selenium sources to the environment. Life cycle of the Phosphoria Formation: from deposition to the post-mining environment, 299.

Prussin S.G., Olander D.R., Lau W.K., and Hansson L. (1988) Release of fission products

(Xe, I, Te, Cs, Mo, and Tc) from polycrystalline UO2. Journal of Nuclear Materials 154, 25– 37.

Put M., De Cannière P., and Moors H. (1992) Migration of radionuclides in Boom Clay. State-of-the-art report 1992. SCK•CEN report to ONDRAF/NIRAS. See description of the C4 type percolation experiments in Annex 1: Different types of migration experiments, pp. 1.7, 1.8 (type C4) and Fig. 1.1.1 (type C4). See also De Preter et al. (1992) for the final report to the European Commission (EC).

128/328

Put M., Aertsens M., De Cannière P, Dierckx A., and Moors H. et al. (1997) Geological disposal of conditioned high-level and long-lived radioactive waste. Progress report to NIRAS/ONDRAF for the first semester of 1997, R-3240 (Mol, April 1998).

Rajan S.S.S. (1979) Adsorption of selenite, phosphate and sulphate on hydrous alumina. Journal of Soil Science 30, 709–718.

Reddy K.J., Zhang Z.H., Blaylock M.J., and Vance G.F. (1995) Method for detecting selenium speciation in groundwater. Environmental Science and Technology 29(7), 1754– 1759.

Redman A.D., Macalady D.L., and Ahmann D. (2002) Natural organic matter affects arsenic speciation and sorption onto hematite. Environmental Science and Technology 36, 2889– 2896.

2– Refait P., Simon L., and Génin J.-M.R. (2000) Reduction of SeO4 anions and anoxic formation of iron(II)-iron(III) hydroxy selenate green rust. Environmental Science and Technology 34, 819-825.

Rietra R.P.J.J., Hiemstra T., and van Riemsdijk W.H. (2001) Comparison of selenate and sulfate adsorption on goethite. Journal of Colloid and Interface Science 240, 384–390. doi:10.1006/jcis.2001.7650.

Román-Ross G., Cuello G.J., Turrillas X., Fernández-Martínez A., and Charlet L. (2006) Arsenite sorption and co-precipitation with calcite. Chemical Geology 233 (3-4), 328–336.

Rose A.L. and Waite T.D. (2003) Kinetics of iron complexation by dissolved natural organic matter in coastal waters. Marine Chemistry 84(1-2), 85–103.

Rotruck J.T., Pope A.L., Ganther H.E., Swanson A.B., Hafeman D.G., and Hoekstra W.G. (1973) Selenium: biological role as a component of glutathione peroxidase. Science 179, 588– 590.

Roussel-Debet S., Adam C., and Beaugelin-Seiller K (2005) Fiche radionucléide: Sélénium 79 et environnement. 27 pp. IRSN, Institut de Radioprotection et de Sûreté Nucléaire. Direction de l’environnement et de l’intervention – Service d’étude du comportement des radionucléides dans les écosystèmes.

129/328

Rouxel O., Ludden J., Carignan J., Marin L., and Fouquet Y. (2002) Natural variations of Se isotopic composition determined by hydride generation multiple collector inductively coupled plasma mass spectrometry. Geochimica et Cosmochimica Acta 66 (18), 3191–3199.

Rouxel O., Ludden J., Fouquet Y., and Carignan J. (n.d.) Natural variations of selenium isotopes determined by multicollector plasma source mass spectrometry: Application to seafloor hydrothermal systems. Presented at the Journal of Conference Abstracts 5, 858.

Runchal A.K. (1997) PORFLOW, a software tool for multiphase fluid flow, heat and mass transport in fractured porous media. User's manual, Version 3.07 (ACRi, Bel Air, California, USA, 1997).

Rusch U., Borkovec M., Daicic J., and van Riemsdijk W.H. (1997) Interpretation of competitive adsorption isotherms in terms of affinity distributions. Journal of Colloid and Interface Science 191, 247–255.

Ryden J.C., McLaughlin J.R. and Syers J.K. (1977a) Mechanisms of phosphate sorption by soils and hydrous ferric oxide gel. Journal of Soil Science 28, 72–92.

Ryden J.C., McLaughlin J.R. and Syers J.K. (1977b) Time-dependent sorption of phosphate by soils and hydrous ferric oxides. Journal of Soil Science 28, 585–595.

Ryser A.L., Strawn D.G., Marcus M.A., Johnson-Maynard J.L., Gunter M.E., and Möller G. (2005) Micro-spectroscopic investigation of selenium-bearing minerals from the Western US Phosphate Resource Area. Geochemical Transactions 6(1), 1–11. doi:10.1186/1467-4866-6-1. (http://www.geochemicaltransactions.com/ published online 28 January 2005).

SAFIR 2 Report (2001) coordinated by De Preter P., Lalieux P., and Cool W. (eds.) SAFIR 2 Report, Safety Assessment and Feasibility Interim Report 2. Four volumes, 13 Chapters. ONDRAF/NIRAS, Belgian Agency for Radioactive Waste and Enriched Fissile Materials. NIROND 2001-06 E. December 2001.

Sagoe-Crentsil K.K., and Glasser F.P (1993) “Green rust”, iron solubility and the role of chloride in the corrosion of steel at high pH. Cement and Concrete Research 23, 785–791.

Sahin F., Volkan M., Howard A.G., and Ataman O.Y. (2003) Selective pre-concentration of selenite from aqueous samples using mercapto-silica. Talanta 60 (5), 1003–1009.

Sarathchandra S.U. and Watkinson J.H. (1981) Oxidation of elemental selenium to selenite by Bacillus megaterium. Science 211 (4482), 600–1.

130/328

Sato H. (1997) Diffusion behavior of Se in compacted sodium bentonite under reducing conditions. PNC TN 8410 97-075.

Sato H., Yui M., and Yoshikawa H.(1995) Diffusion behaviour for Se and Zr in sodium bentonite. Material Research Society Symposium Proceedings Vol 353, Materials Research Society, pp. 269–276 (in Japanese).

Scheidegger A.M., Grolimund D., Cui D., Devoy J., Spahiu K., Wersin P., Bonhoure I., and Janousch M. (2003) Reduction of selenite on corroded iron surfaces: a micro-spectroscopic study. Journal de physique IV 104, 417–420. doi:10.1051/jp4:20030112.

Séby F., Potin-Gautier M., Giffaut E., and Donard O.F.X. (1998) Assessing the speciation and the biogeochemical processes affecting the mobility of selenium from a geological repository of radioactive wastes to the biosphere. Analusis 26, 193–198.

Seby F., Potin-Gautier M., Giffaut E., Borge G., and Donard O.F.X. (2001) A critical review of thermodynamic data for selenium species at 25 °C. Chemical Geology 171(3), 173–194.

Selenium (2007, March 4) In: Wikipedia, The Free Encyclopedia. Retrieved 23:56, March 5, 2007, from http://en.wikipedia.org/w/index.php?title=Selenium&oldid=112629566.

Shaker A.M. (1996) Kinetics of the reduction of Se(IV) to Se-sol. Journal of Colloid and Interface Science 180, 225–231.

Sharmasarkar S. and Vance G.F. (1995) Fractional partitioning for assessing solid-phase speciation and geochemical transformations of soil selenium. Soil Science 160(1), 43–55.

Sharmasarkar S. and Vance G.F. (2002) Selenite-selenate sorption in surface coal mine environment. Advances in Environmental Research 7(1), 87–95.

Sheppard M.I, Beals D.I, Thibault D.H., and O'Connor P. (1984) Soil nuclide distribution coefficients and their statistical distribution. AECL, Pinawa, Report AECL-8364.

Shrift A. (1964) A selenium cycle in nature ? Nature 201, 1304–1305 (Mar 28).

Shriver D.F., Atkins P.W., and Langford C.H. (1990) Inorganic chemistry. 706 pp. Oxford University Press. Oxford, Melbourne, Tokyo. ISBN 0-19-855231-9. See pp. 240 – 241: Electron transfer, atom transfer; and also p. 392: sluggish reduction of selenate and tellurate. 2+ – p. 416: aqueous solutions containing Fe and ClO4 are stable for many months in the

131/328

absence of dissolved oxygen because oxyanions reduction becomes slower as the oxidation number increases.

Shuh D.K., Kaltsoyannis N., and Bucher J.J. (1994) Environmental applications of XANES: Speciation of Tc in cement after chemical treatment and Se after bacterial uptake. Conference: Spring meeting of the Materials Research Society (MRS), San Francisco, CA (United States), 4–8 Apr 1994.

Sigg L. (1999) Redox potential measurements in natural waters: Significance, concepts and problems, in Redox: Fundamentals, processes and applications. J. Schüring, et al. Ed., Springer, Berlin.

Sillen X., and Marivoet J. (2002) Spent fuel performance assessment for a hypothetical repository in the Boom Clay at the Mol site (Belgium). BLG-877. Open report to ONDRAF/NIRAS for contract – KNT 90.95.656.02 and to the European Commission for contract EC FI4W-CT96-0018. Waste and Disposal Department SCK•CEN, Mol Belgium. February 2002. 138 pp.

Sindeeva N.D. (1964) Mineralogy and types of deposits of selenium and tellurium. Interscience publishers. 1st edition. ASIN: B000CMY78M.

Singh B. (1993) Nuclear Data Sheets Update for A = 79. Compilation 79Cu, 79Zn, 79Ga, 79Ge, 79As, 79Se, 79Br, 79Kr, 79Rb, 79Sr, 79Y; compiled, evaluated structure data. Nuclear Data Sheets 70, 437. 1993SI28. (comment: question on the validity of the half-life of 79Se), citation in McAninch et al. (1994).

Singh B. (2002) Nuclear Data Sheets for A = 79. Compilation 79Cu, 79Zn, 79Ga, 79Ge, 79As, 79Se, 79Br, 79Kr, 79Rb, 79Sr, 79Y, 79Zr; compiled, evaluated structure data. Nuclear Data Sheets 96(1), 1–176 (comment: see for the new half-life of 79Se = 2.95 E+5 years). NuDat 2.1: http://www.nndc.bnl.gov/nudat2/reCenter.jsp?z=34&n=45

Singh B. (2002) Nuclear Data Sheets for A = 79. Nuclear Data Sheets 96(1), 1–176.

SKB (1999) SR 97 – Post closure safety. Main Report. Volumes I & II. SKB Technical Report TR-99-06, Stockholm, Sweden.

Skoog D.A. and West D.M. (1976) Fundamentals of analytical chemistry. Third Edition. Holt, Rinehart and Winston. New York. ISBN 0-03-089495-6. See p. 365: sources of error in – + iodometric methods: 4 I + O2(g) + 4 H <=> 2 I2 + 2 H2O. In the index of keywords on page

132/328

797, see also: Iodide ion, air oxidation of, p. 58, 363, 365. See also: Iodine solution(s) preparation and standardization of, p. 358, 360, 747.

Sladkov V., Fourest B., David F., Venault L., Lecomte M. (2003) Application of capillary electrophoresis for inorganic selenium speciation in the frame of high-level waste management. Analytical and Bioanalytical Chemistry 376, 455–459.

Small, J., M. Nykyri, M. Helin, U. Hovi, T. Sarlin, and M. Itävaara (2008) Experimental and modelling investigations of the biogeochemistry of gas production from low and intermediate level radioactive waste. Applied Geochemistry in press.

Small, J. (2008) Biogeochemical modelling of a bituminous waste storage cell and impact on nitrate transport; Rapport Nexia Solutions N° Andra C RP FSTR 08.0008.

Smedley P.L. and Kinniburgh D.G. (2002) A review of the source, behaviour and distribution of arsenic in natural waters, Applied Geochemistry 17, 517–568.

Smith, D.W. (2006) “Phenomenology”, The Stanford Encyclopedia of Philosophy (Winter 2006 Edition), Zalta Edward N. (ed.), URL = http://plato.stanford.edu/archives/win2006/entries/phenomenology/.

Smith K. (Ed.) (2008) Report of Se-79 in the Biosphere. BIOPROTA Workshop, hosted by Nagra, Wettingen, 5–6 May 2008, V2.0, Final. BIOPROTA – http://www.bioprota.com.

Smith L. and Craig B. (2005) Practical corrosion control measures for elemental sulfur containing environments. Paper N° 05646 submitted to NACE Corrosion 2005. 20 pp.

Song-Sheng J., Jingru G., Shan J., Chunsheng L., Anzhi C., Ming H., Shaoyong W., and Shilin L. (1997) Determination of the half-life of 79Se with the accelerator mass spectrometry technique. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, Elsevier Science 123, 405–409.

Song-Sheng J., Ming H., Li-Jun D., Jing-Ru G., and Shao-Yong W. (2001) Remeasurement of the Half-Life of 79Se with the Projectile X-Ray Detection Method. Chinese Physics Letters, Institute of Physics Publishing 18, 746–749.

Spackman L., Hartman K.K.D., Harbour J.D., and Essington M.E. (1990) Adsorption of oxyanions by spent western oil shale: II. Selenite. Environmental Geology 15(2), 93–99. doi:10.1007/BF01705096.

133/328

233 Spahiu K. (2005) The effect of dissolved hydrogen on the dissolution of U doped UO2 (s), high burn-up spent fuel and MOX fuel. SKB TR-05-09.

Spahiu K., Cui D., and Lundström M. (2004) The fate of radiolytic oxidants during spent fuel leaching in the presence of dissolved near field hydrogen. Radiochimica Acta 92 (9-11), 625– 629. doi:10.1524/ract.92.9.625.54990.

Spahiu K., Devoy J., Cui D., and Lundström M. (2004) The reduction of U(VI) by near field hydrogen in the presence of UO2(s). Radiochimica Acta 92 (9-11), 597–601. doi:10.1524/ract.92.9.597.54985.

Spahiu K., Werme L., and Eklund U.-B. (2000) The influence of near field hydrogen on actinide solubilities and spent fuel leaching. Radiochimica Acta 88 (9-11), 507. doi:10.1524/ract.2000.88.9-11.507.

Sposito G. (1981) The thermodynamics of soils solutions. 223 pp. Oxford Clarendon Press. Oxford. ISBN 0-19-857568-8. See p. 75–77, Hard Soft Lewis Acid Base (HSAB) principle.

Sposito G. (1984) The surface chemistry of soils. Oxford University Press. New York.

Sposito G., Skipper N.T., Sutton R., Park S.-H., Soper A.K., and Greathouse J.A. (1999) Surface geochemistry of the clay minerals. Paper presented at the National Academy of Sciences colloquium “Geology, Mineralogy, and Human Welfare” held November 8-9, 1998 at the Arnold and Mabel Beckman Center in Irvine, CA. Colloquium Paper. Proceedings of the National Academy of Science USA 96(7), 3358–3364. http://www.pnas.org/cgi/content/full/96/7/3358.

Sposito G. and Levesque C.S. (1985) Sodium-calcium-magnesium exchange on Silver Hill Illite. Soil Science Society of America Journal 49, 1153-1159.

Stachowiak J., Gorecka D., and Zarakowska W. (2000) Sorption of selenium by selected cereal bran. Polish Journal of Food and Nutrition Sciences 9 (4), 27–30.

Steinberg N. and Oremland R. (1990) Dissimilatory selenate reduction potentials in a diversity of sediment types. Applied and Environmental Microbiology 56(11), 3550–3557.

Steinberg N.A., Blum J.S., Hochstein L., and Oremland R.S. (1992) Nitrate is a preferred electron acceptor for growth of freshwater selenate-respiring bacteria. Applied and Environmental Microbiology 58(1), 426–428.

134/328

Steudel R. (2002) The chemistry of organic polysulfanes R−Sn−R (n > 2). Chemical Reviews 102 (11), 3905–3946. doi:10.1021/cr010127m.

Stevenson F.J. Editor (1982) Humus chemistry: Genesis, composition, reactions. Wiley & Sons, New York.

Stevenson F.J. (1982) Extraction, fractionation, and general chemical composition of soil organic matter. In: Stevenson F.J., Editor (1982) Humus chemistry. Genesis, composition, reactions. John Wiley & Sons, New York, pp. 26–54.

Stevenson F.J. (1985) Geochemistry of soil humic substances. Humic substances in soil, sediment, and water: geochemistry, isolation, and characterization. John Wiley & Sons, New York. pp. 13–52, 14 Fig., 8 Table.

Stolz J., Basu P., and Oremland R. (2002) Microbial transformation of elements: the case of arsenic and selenium. International Microbiology 5(4), 201–207.

Stolz J.F., Basu P., Santini J.M., and Oremland R.S. (2006) Arsenic and selenium in microbial metabolism. Annual Review of Microbiology 60, 107–130. http://micro.annualreviews.org. doi:10.1146/annurev.micro.60.080805.142053.

Stolz J.F. and Oremland R.S. (1999) Bacterial respiration of arsenic and selenium. FEMS Microbiology Reviews 23(5), 615–627.

Stork A.W., Jury W.A., and Frankenberger W.T. Jr. (1999) Accelerated volatilization rates of selenium from different soils. Biol. Trace Elements Res. 69, 217–234.

Stroes-Gascoyne S., Schippers A., Schwyn B., Poulain S., Sergeant C., Simonoff M., Le Marrec C., Altmann S., Nagaoka T., Mauclaire L., McKenzie J., Daumas S., Vinsot A., Beaucaire C., and Matray J.-M. (2007) Microbial community analysis of Opalinus Clay drill core samples from the Mont Terri Underground Research Laboratory, Switzerland. Geomicrobiology Journal, 24, 1–17. doi:10.1080/01490450601134275. ISSN: 0149-0451 print / 1521-0529 online.

Struyck Z. and Sposito G. (2001) Redox properties of standard humic acids. Geoderma 102, 329–346.

Stumm W. (1992) Chemistry of the solid-water interface – Processes at the mineral-water and particle-water interface in natural systems. 428 pp. John Wiley & Sons, Inc. New York (USA).

135/328

Stumm W., Kummert R., and Sigg L. (1980) A ligand exchange model for the adsorption of inorganic and organic ligands at hydrous oxide interfaces. Croat. Chem. Acta 53, 291–312.

Stumm W. and Morgan J.J. (1996) Aquatic chemistry — Chemical equilibria and rates in natural waters, 1022 pp. Third edition. A Wiley-Interscience publication. John Wiley & Sons Inc. New York – Chichester – Brisbane – Toronto – Singapore. See section 15.14 The sulfur cycle: figure 15.25, p. 932.

Su C. and Suarez D.L. (1997) In situ infrared speciation of adsorbed carbonate on aluminium and iron oxides. Clays Clay and Minerals 45, 814–825.

Su C. and Suarez D.L. (2000) Selenate and selenite sorption on iron oxides: an infrared and electrophoretic study. Soil Science Society of America Journal 64,101–111.

Swift R.S. (1996) Organic matter characterisation. Methods of Analysis. Part 3. Chemical Methods. SSA and ASA, Madison, WI, pp. 1011–1069.

Switzer-Blum J., Stolz J., Oren A., and Oremland R. (2001) Selenihalanaerobacter shriftii gen. nov., sp. nov., a halophilic anaerobe from Dead Sea sediments that respires selenate. Archives of Microbiology 175(3), 208–219.

Swizer-Blum J., Bindi A., Buzzelli J., Stolz J., and Oremland R. (1998) Bacillus arsenicoselenatis, sp. nov., and Bacillus selenititreducens, sp. nov.: two haloakiliphiles from Mono Lake, California, that respire oxyanions of selenium and arsenic. Archives of Microbiology 171(1), 19–30.

Synal H. and Wacker L. (n.d.) Beam Interactions with Materials and Atoms: AMS measurement technique after 30 years: Possibilities and limitations of low energy systems. Nuclear Instruments and Methods in Physics Research Section B. doi:10.1016/j.nimb.2009.10.009.

Tachi Y., Shibutani T., Sato H., and Yui M. (1998) Sorption and diffusion behaviour of selenium in tuff. Journal of Contaminant Hydrology 35 (1-3), 77–89.

Thomsen M., Heinemeier J., Hornshøj P., Nielsen H., and Rud N. (1988) Beam Interactions with Materials and Atoms: Accelerator mass spectrometry applied to 32Si. Nuclear Instruments and Methods in Physics Research Section B 31 (3), 425–432. doi:10.1016/0168- 583X(88)90342-4.

136/328

Thomson C.D. (2004) Assessment of requirements for selenium and adequacy of selenium status: a review. European Journal of Clinical Nutrition 58, 391–402.

Thomson B.M., Longmire P.A., and Brookins D.G. (1986) Geochemical constraints on underground disposal of uranium mill tailings. Journal of Applied Geochemistry 1, 335–344.

Thomson J., Nixon S., Croudace I.W., Pedersen T.F., Brown L., Cook G.T., and MacKenzie A.B. (2001) Redox-sensitive element uptake in north-east Atlantic Ocean sediments (Benthic Boundary Layer Experiment sites) Earth and Planetary Science Letters 184, 535–547.

Todd J. D., Rogers R., Li Y.G., Wexler M., Bond P.L., Sun L., et al. (2007) Structural and regulatory genes required to make the gas dimethyl sulfide in bacteria. Science 315 (5812), 666–669. doi:10.1126/science.1135370.

Tokunaga T.K., Brown G.E., Pickering I.J., Sutton S.R., and Bait S. (1997) Selenium redox reactions and transport between ponded waters and sediments. Environmental Science & Technology 31 (5), 1419–1425.

Tournassat C. (2003) Clay Surface Chemistry of Fe(II). University of Grenoble–I, France, Ph.D. Thesis.

Tournassat C., Ferrage E., Poinsignon C., and Charlet L. (2004) The titration of clay minerals II. Structure-based model and implications for clay reactivity, J. Colloid Interface Sci. 273, 234–246.

Tournassat C., Gaucher E.C., Fattahi M., and Grambow B. (2006) On the mobility and potential retention of iodine in the Callovian–Oxfordian formation. Physics and Chemistry of the Earth, Parts A/B/C. Article in Press, Corrected Proof. Accessed online on 2006-12-02. doi:10.1016/j.pce.2005.12.004.

Trombe J. C., Gleizes A., and Galy J. (1985) Structure of a uranyl diselenite, UO2Se2O5. Acta Crystallographica Section C 41, 1571-1573. doi:10.1107/S0108270185008599. Acta Crystallographica Section C, Crystal Structure Communications, Volume 41, Part 11 research papers (inorganic compounds).

Tuniz C., and Norton G. (2008) Beam Interactions with Materials and Atoms: Accelerator mass spectrometry: New trends and applications. Nuclear Instruments and Methods in Physics Research Section B 266 (8), 1837–1845. doi:10.1016/j.nimb.2007.12.045.

137/328

Vandenberghe N. (1978) Sedimentology of the Boom Clay (Rupelian) in Belgium. Verhand. Kon. Acad. Wetenschappen België, XL, 147, 137 pp. van der Lee J. (1998) Thermodynamic and mathematical concepts of CHESS. Technical report Nr. LHM/RD/98/39. Centre d’Informatique Géologique (CIG), Ecole des Mines de Paris, Fontainebleau, France. van der Lee J. (2000) CHemical Equilibrium Speciation with Surfaces (CHESS). Centre d’Informatique Géologique, ARMINES, France. http://chess.ensmp.fr/. van der Lee J., and Wang L. (2004) Topical report on: Speciation and solubility calculations for uranium, plutonium, and selenium under Boom Clay conditions. SCK•CEN Restricted Report R-3 400, 16 pp. Also available as Annex 1 to WP 1 of the extended Final Scientific and Technical Report of the EC Trancom-II Project. In: Maes N. coordinator et al. (2004a) Migration case study: transport of radionuclides in a reducing clay sediment (TRANCOM-II). SCK•CEN-BLG-988 (04/NMa/P-50). see section 3.3 selenium, pp. 8 – 9.

Van Geet M., Maes N., and Dierckx A. (2003) Characteristics of the Boom Clay organic matter, a review. Royal Belgian Institute of Natural Sciences, Geological Survey of Belgium, Professional Paper 2003/1, N°298, Brussels, Belgium, 23 pp. ISSN 0378-0902.

Van Iseghem P., Lemmens K., and Pirlet V. (2007) The leaching of Se, Sn, Zr, Pd from vitrified high-level waste in clay slurries. Mobile fission and activation products in nuclear waste disposal. International workshop. L’Hermitage, La Baule – France, January 16-19, 2007. http://mofap07.in2p3.fr/. Proceedings to appear as a Nuclear Energy Agency (NEA, OECD) publication.

Van Loon L.R. and Soler J.M. (2004) Diffusion of HTO, 35Cl–, 125I– and 22Na+ in Opalinus Clay: Effect of confining pressure, sample orientation, sample depth and temperature. PSI Report Nr. 04–03, Paul Scherrer Institut, Switzerland.

Van Loon L.R., Wersin P., Soler J.M., Eikenberg J., Gimmi Th., Hernán P., Dewonck S. and Matray J.-M. (2004) In-situ diffusion of HTO, 22Na+, Cs+ and I– in Opalinus Clay at the Mont Terri underground rock laboratory. Radiochimica Acta 92, 757–763.

Van Loon L.R., Yllera A., and Hernandez A. (2006) Long-term diffusion experiment (DI-A): through-diffusion of HTO, 36Cl-, I– and 22Na+ in Opalinus Clay parallel to the bedding plane under laboratory conditions. Mont Terri Project Technical Report TR 2003-01.

138/328

van Riemsdijk W.H., De Wit J.C.M., Mous S.L.J., Koopal L.K., and Kinniburgh D.G. (1996) An analytical isotherm equation (CONICA) for non-ideal mono- and bidentate competitive ion adsorption to heterogeneous surfaces. Journal of Colloid and Interface Science 183, 35– 50.

Velinsky D.J. and Cutter G.A. (1990) Determination of elemental selenium and pyrite- selenium in sediments. Analytica Chimica Acta 235(2), 419–425.

Venema P., Hiemstra T., Weidler P.G., and van Riemsdijk W.H. (1998) Intrinsic proton affinity of reactive surface groups of metal (hydr)oxides: Application to iron (hydr)oxides. Journal of Colloid and Interface Science 198, 282–295.

Verink E.D. (1979) Simplified procedure for constructing Pourbaix diagrams. J. Educ. Modules Mat. Sci. Eng. 1, 535–560.

Vermeer A.W.P., van Riemsdijk W.H., and Koopal L.K. (1998) Adsorption of humic acid to mineral particles. 1. Specific and electrostatic interactions. Langmuir 14, 2810–2819.

Vieno T. and Nordman H. (1999) Safety assessment of spent fuel disposal in Hästholmen, Kivetty, Olkiluoto and Romuvaara. TILA-99. 253 pages. Report from VTT Energy to Posiva OY. Posiva Report. Posiva 99-07. Helsinki Finland. March 1999. ISBN 951-652-062-6. ISSN 1239-3096. See page 20: Activity inventories / Half-life of Se-79. Mention of incorrect value commonly accepted for the half-life of 79Se.

Vila-Costa M., Simo R., Harada H., Gasol J. M., Slezak D., and Kiene R.P. (2006) Dimethylsulfoniopropionate uptake by marine phytoplankton. Science 314 (5799), 652–654. doi:10.1126/science.1131043.

Villalobos M. and Leckie J.O. (2001) Surface complexation modelling and FTIR study of carbonate adsorption to goethite. Journal of Colloid and Interface Science 235, 15–32.

Vochten R., Blaton N., Peeters O., and Deliens M. (1996)

Piretite, Ca(UO2)3(SeO3)2(OH)2 · 4 H2O, a new calcium uranyl selenite from Shinkolobwe, Shaba, Zaire. Canadian Mineralogy 34, 1317–1322.

Wackett L.P. (2001) Selenium, the environment and microbes an annotated selection of world wide web sites relevant to the topics in environmental microbiology web alert. Environmental Microbiology 3 (9), 604–604(1).

139/328

Walker C.T., Rondinella V.V., Papaioannou D., Van Winckel S., Goll W., and Manzel R.

(2005) On the oxidation state of UO2 nuclear fuel at a burn-up of around 100 MWd/kgHM. Journal of Nuclear Materials 345, 192–205. doi:10.1016/j.jnucmat.2005.05.010.

Wang L., Dierckx A., and De Cannière P. (2000a) Speciation and solubility of radionuclides in Boom Clay: calculations performed with “The Geochemist’s Workbench-3.1” and the database of LLNL, version 8, release 6. Technical report BLG-847, SCK•CEN, Mol, Belgium.

Wang L., Dierckx A., and De Cannière P. (2000b) Speciation and solubility of radionuclides in Boom Clay: calculations performed with “The Geochemist's Workbench-3.1” and the database of LLNL, version 8, release 6. Restricted report, R-3508, SCK•CEN, Mol, Belgium.

Wang L., De Craen M., Maes N., De Cannière P., and Put M. (2001a) Determination of natural uranium concentration in Boom Clay: Effect of different extraction techniques. In: Geological evidence and theoretical bases for radionuclide-retention processes in heterogeneous media. Proceedings of the OECD/NEA GEOTRAP 5 Workshop, Oskarsham, Sweden, 7-9 May 2001, NEA/OECD, Paris, pp. 257.

Wang L., Marivoet J., De Cannière P., Sillen X., Moors H., Aertsens M., Put M., and Dierckx A. (2001b) Data collection form: the link between migration studies and performance assessment – Belgian Case. In: The use of thermodynamic databases in performance assessment. Proceedings of the Workshop on the OECD/NEA Thermochemical Database Project, Barcelona, Spain, 29-30 May 2001, NEA/OECD, Paris, p. 181.

Wang L., Dierckx A., De Cannière P., and Maes A. (2002) Uranium release from Boom Clay in bicarbonate media. Radiochimica Acta 90, 515.

Wang L. (2004) Proposal for a study of sorption in Boom Clay: a position paper. Report in preparation, SCK•CEN, Mol, Belgium.

Wang L., Ju Y., Liu G., Chou C., Zheng L., and Qi C. (n.d.) Selenium in Chinese coals: distribution, occurrence, and health impact. Environmental Earth Sciences, 1–11.

Wang W., Guan Y., He M., Jiang S., Wu S., and Li C. (2009) A method for measurement of ultratrace 79Se with accelerator mass spectrometry. Nuclear Inst. and Methods in Physics Research, B.

Wang W., Ruan X., et al. (2009) Absolute determination of 79Se/80Se with AMS. Nuclear Inst. and Methods in Physics Research, B.

140/328

Wang W., Guan Y., He M., Jiang S., Wu S., and Li C. (n.d.) Beam Interactions with Materials and Atoms: A method for measurement of ultratrace 79Se with accelerator mass spectrometry. Nuclear Instruments and Methods in Physics Research Section B. doi:10.1016/j.nimb.2009.10.024.

Wang W., Ruan X., et al. (n.d.) Beam Interactions with Materials and Atoms: Absolute determination of 79Se/80Se with AMS. Nuclear Instruments and Methods in Physics Research Section B. doi:10.1016/j.nimb.2009.10.025.

Weast R.C. (ed.) (1968-1969a) CRC Handbook of Chemistry and Physics. Table of the 79 4 th isotopes, T½ of Se = 7 × 10 years, see p. B-21. 49 Edition (1968 – 1969). CRC, The Chemical Rubber Company. Cleveland, Ohio.

Weast R.C. (ed.) (1968-1969b) CRC Handbook of Chemistry and Physics. The average amounts of the elements in Earth’s crust. Reprinted with permission of the author from “Mason B. (1952) Principles of Geochemistry. John Wiley and Sons”. see p. F–144. 49th Edition (1968 – 1969). CRC, The Chemical Rubber Company. Cleveland, Ohio.

Weast R.C. (ed.) (1968-1969c) CRC Handbook of Chemistry and Physics. Elements present in solution in sea water excluding dissolved gases. Reprinted with permission of the authors from “Sverdrup, Johnson, and Fleming (1942) The Oceans. Prentice-Hall, Inc.”. see p. F–145. 49th Edition (1968 – 1969). CRC, The Chemical Rubber Company. Cleveland, Ohio.

Weetjens E. (2004) Migration case study: transport of radionuclides in a reducing clay sediment (TRANCOM-II). SCK•CEN-BLG-988. Contract N° FIKW-CT-2000-00008. Annex 19 to WP-9. PA case study for new parameters of Se and U (03/EWe/N-21, 8 pp. SCK•CEN, Waste and Disposal Department – R&D Geological Disposal, Mol, Belgium.

Wen H. and Carignan J. (2007) Reviews on atmospheric selenium: Emissions, speciation and fate. Atmospheric Environment 41 (34), 7151–7165.

Wen H. and Carignan J. (2009) Ocean to continent transfer of atmospheric Se as revealed by epiphytic lichens. Environmental Pollution 157 (10), 2790–2797.

Wen H., Carignan J., Qiu Y., and Liu S. (2006) Selenium speciation in kerogen from two Chinese selenium deposits: environmental implications. Environmental Science and Technology 40 (4), 1126–1132. doi:10.1021/es051688o.

141/328

Wen H., Carignan J., Hu R., Fan H., Chang B., and Yang G. (2007) Large selenium isotopic variations and its implication in the Yutangba Se deposit, Hubei Province, China. Chinese Science Bulletin 52 (17), 2443–2447.

Wen H. and Qiu Y. (1999) Organic and inorganic occurrence of selenium in Laerma Se-Au deposit. Science in China Series D: Earth Sciences 42 (6), 662–669.

West J.M., Christofi N., Philp J.C., and Arme S.C. (1986) Investigations on the populations of introduced and resident micro-organisms in deep repositories and their effects on containment of radioactive wastes. 62 pp. Commission of the European Communities, Nuclear Science and Technology, EUR series. Final report EUR 10405 EN. Contract N° 376- 83-7 WAS UK. Luxembourg. ISBN 92-825-6201-8. See specially pp. 6 and 7.

Westall J.C. (1982) FITEQL: a computer program for determination of chemical equilibrium constants from experimental data. Report 82-01. Department of Chemistry, Oregon State University, Corvallis.

WHO (2003) World Health Organisation, WHO/SDE/WSH/03.04/13. Selenium in drinking- water. Background document for development of WHO Guidelines for Drinking-water Quality. Originally published in Guidelines for Drinking-water Quality, 2nd ed. Vol. 2 Health criteria and other supporting information. World Health Organisation, Geneva, 1996.

Wijnja H. and Schulthess C.P. (1999) ATR-FTIR and DRIFT spectroscopy of carbonate species at the aged gamma-Al2O3/water interface. Spectrochimica Acta Part A 55, 861–872.

Wijnja H. and Schulthess C.P. (2000) Vibrational spectroscopic study of selenate and sulfate adsorption mechanisms on Fe and Al(hydr)oxides surfaces. Journal of Colloid and Interface Science 229, 286–297. doi:10.1006/jcis.2000.6960, available online at http://www.idealibrary.com.

Wijnja H. and Schulthess C.P. (2001) Carbonate adsorption mechanisms on goethite studied with ATR-FTIR, DRIFT and proton coadsorption measurements. Soil Science Society of America Journal 65, 324–330.

Wikipedia contributors. (2008a) Coccolithophore. In Wikipedia, The Free Encyclopedia. Wikimedia Foundation. Retrieved March 21, 2008, from http://en.wikipedia.org/w/index.php?title=Coccolithophore&oldid=194436315.

142/328

Wikipedia contributors. (2008c) Dimethylsulfoniopropionate. In Wikipedia, The Free Encyclopedia. Wikimedia Foundation. Retrieved March 21, 2008, from http://en.wikipedia.org/w/index.php?title=Dimethylsulfoniopropionate&oldid=199883496.

Wikipedia contributors. (2009) Emiliania huxleyi. In Wikipedia, The Free Encyclopedia. Wikimedia Foundation. Retrieved January 3, 2010, from http://en.wikipedia.org/w/index.php?title=Emiliania_huxleyi&oldid=334837333.

Wittebroodt C. (2009) Transfert de l’iode dans l’argilite de Tournemire. Ph. D Thesis made at IRSN and Montpellier University 2 (UM2). Thèse présentée le 12 mars 2009 en soutenance devant l’Université Montpellier 2 pour l'obtention du diplôme de Doctorat Géosciences.

Wolery T. (1992) EQ3/6: A software package for geochemical modelling of aqueous systems: Package overview and installation guide. Technical Report UCRL-MA-110662 PT, I ed., Lawrence Livermore National Laboratory, Berkeley, CA (USA).

Workman S.M. and Soltanpour P.N. (1980) Importance of prereducing selenium(VI) to selenium(IV) and decomposing organic matter in soil extracts prior to determination of selenium using hydride generation. Soil Science Society of America Journal 44(6), 1331– 1333.

Worsfold P.J, Townsend A., and Poole C.F. eds. (2005) Encyclopedia of Analytical Science. Second Edition, Elsevier.

Wright W.G. (1999) Oxidation and mobilization of selenium by nitrate in irrigating drainage. Journal of Environmental Quality 28(4), 1182–1187.

Xiang X. and Jian-qiu G. (2008) Selenium distribution and comprehensive utilization. Journal of Yangtze University (Natural Science Edition) Sci and Eng V.

Yllera de Llano A., Bidoglio G., Avogadro A., Gibson P.N., and Rivas Romero P. (1996) Redox reactions and transport of selenium through fractured granite. Journal of Contaminant Hydrology 21(1-4), 129–139.

Yoshihisa Iida, Tetsuji Yamaguchi, Shinichi Nakyama, Tomoko Nakajima, and Joshiaki Sakamoto (2001) The solubility of metallic selenium under anoxic conditions. Mat. Res. Soc. Symp. Proc. 663, 1143–1149.

Yuan-Hui Li and Sandra Gregory (1974) Diffusion of ions in sea water and in deep-sea sediments. Geochimica et Cosmochimica Acta 38, 703–714.

143/328

Zehr J.P. and Oremland R.S. (1987) Reduction of selenate to selenide by sulfate-respiring bacteria: experiments with cell suspensions and estuarine sediments. Applied and Environmental Microbiology 53(6), 1365–1369. Retrieved from http://aem.asm.org/cgi/content/abstract/53/6/1365.

Zhang P.C. and Sparks D.L. (1990) Kinetics and mechanisms of sulfate adsorption/desorption on goethite using pressure-jump relaxation. Soil Science Society of America Journal 54, 1266–1273.

Zhang P.C. and Sparks D.L. (1990) Kinetics of selenate and selenite adsorption/desorption at the goethite/water interface. Environmental Science and Technology 24, 1848–1856.

Zhang H.Y., Hu Z.Q., and Lu K. (1995) Transformation from the amorphous to the nanocrystalline state in pure selenium. Nanostructured Materials 5(1), 41–52.

Zhang H., Wu Z., Yang C., Xia B., Xu D., and Yuan H. (2008) Spatial distributions and potential risk analysis of total soil selenium in Guangdong province, China. Journal of Environmental Quality 37 (3), 780–787. doi:10.2134/jeq2007.0154.

Zhang Y. and Moore J.N. (1996) Selenium fractionation and speciation in a wetland system. Environmental Science and Technology 30, 2613–2619.

Zhang Y. and Moore J.N. (1997) Interaction of selenate with a wetland sediment. Applied Geochemistry 12(5), 685–691.

Zhang Y. and Frankenberger W.T. Jr. (2000) Formation of dimethylselenonium compounds in soil. Environmental Science and Technology 34, 776–683.

Zhang Y. and Frankenberger W.T. Jr. (2003) Characterization of selenate removal from drainage water using rice straw. Journal of Environmental Quality 32, 441–446.

Zhang Y., Amrhein C., and Frankenberger W.T. Jr. (2005a) Effect of arsenate and molybdate on the removal of selenate from an aqueous solution by zerovalent iron. Science of the Total Environment 350(1-3), 1–11.

Zhang Y., Wang J., Amrhein C., and Frankenberger W.T. Jr. (2005b) Removal of selenate from water by zerovalent iron. Journal of Environmental Quality 34(2), 487–495.

144/328

Zhao Y.H., Lu K., and Liu T. (2004) EXAFS study of mechanical-milling-induced solid-state amorphization of Se. Journal of Non-Crystalline Solids 333, 246–251.

Zhong S. and Mucci A. (1995) Partitioning of rare earth elements (REEs) between calcite and seawater solutions at 25 °C and 1 atm., and at high dissolved REE concentrations. Geochimica et Cosmochimica Acta 59, 443–453.

Zhu J., Zuo W., Liang X., Li S., and Zheng B. (2004) Occurrence of native selenium in Yutangba and its environmental implications. Applied Geochemistry 19 (3), 461–467.

Zingaro R. (Editor) (1974) Selenium. 835 pp. Krieger Pub Co. ISBN: 0442295758.

Zingaro R.A., Dufner D.C., Murphy A.P., and Moody C.D. (1997) Reduction of oxoselenium anions by iron(II) hydroxide. Environment International 23(3), 299–304.

Zuidema P. (1993) Conceptual models: the performance assessment point of view. pp. 27–36 in: OECD (1993) The role of conceptual models in demonstrating repository post-closure safety. Proceedings of a NEA Workshop, Paris, 16-18 November 1993. Nuclear Energy Agency (NEA). OECD Documents. 190 pp. ISBN 92-64-14429-3.

Zuidema P., Gautschi A., Smith P., and Vomvoris S. (1993) The treatment of conceptual model uncertainty in the Nagra programme: a few examples. pp. 39–53 in: OECD (1993) The role of conceptual models in demonstrating repository post-closure safety. Proceedings of a NEA Workshop, Paris, 16-18 November 1993. Nuclear Energy Agency (NEA). OECD Documents. 190 pp. ISBN 92-64-14429-3.

Special workshops on selenium and mobile fission products

MOFAP (2007) Mobile fission and activation products in nuclear waste disposal. International workshop. L’Hermitage, La Baule – France, January 16-19, 2007. http://mofap07.in2p3.fr/. Proceedings to appear as a Nuclear Energy Agency (NEA, OECD) publication.

BIOPROTA (2008) Key issues in biosphere aspects of assessment of the long-term impact of contaminant releases associated with radioactive waste management. Report of an International Forum on Se-79 in the Biosphere. 5–6 May 2008. Hosted by Nagra, Wettingen, Switzerland. Editor: K. Smith (2008) Version 2.0, Final, June 2008. http://www.bioprota.com. (full citation).

145/328

146/328

APPENDICES

147/328

148/328

APPENDICES

The appendices contain more detailed information on specific topics discussed in the body of the synthesis text but which are too heavy to be easily accessible in a first reading. They are provided hereafter for the sake of traceability and completeness.

Appendix A1: General information on selenium.

Appendix A2: Natural selenium in the environment and in Boom Clay.

Appendix A3: Selenium speciation in the source term.

Appendix A4: Selenium speciation behaviour in Boom Clay.

Appendix A5: Immobilisation of selenium in the near-field.

Appendix A6: Selenium background concentration in bentonite buffer.

Appendix A7: Sorption behaviour of selenite, selenate, and sulfate on Fe and Al oxide surfaces.

Appendix A8: Selenium and organic matter.

Appendix A9: Selenium migration behaviour in Boom Clay.

Appendix A10: Redox disequilibrium and reluctance of sulfate for reduction in deep clay formations.

Appendix A11: Behaviour of redox-sensitive elements in a nitrate plume associated with bituminized MLW – The selenium case study.

Appendix A12: List of abbreviations.

Appendix A13: List of symbols.

Appendix A14: List of physical constants and units.

149/328

150/328

A1. General information on selenium

151/328

152/328

A1 General information on selenium

A1.1 Overview Selenium-79 is a long-lived radioisotope produced by the nuclear fission of 235U and 239Pu and other fissile minor actinides. Their exist several natural and artificial selenium isotopes. The isotopic composition of selenium present in spent fuel considerably differs from its natural composition. Information is provided here on the main differences in Se isotopic composition that could be needed to calculate isotopic dilution and exchange between 79Se and the more abundant stable isotopes of selenium present in Boom Clay and in spent fuel (SF): mainly 78Se, 80Se and 82Se. The total selenium quantity present in ~ 5 000 tons of spent fuel generated by 40 years of nuclear electricity production in Belgium (7 power plants) is estimated at about 400 kg. The calculated selenium inventory is about 1 mol elemental Se per ton heavy metal (tHM). This estimation is of the same order of magnitude than for 129I. A quantity of about one mol of selenium is expected to be disposed of per current meter of underground gallery.

Their exist large uncertainties affecting the value of 79Se half-life. Two recent changes have been published in the last ten years. The value of 65 000 years commonly accepted in most of the tables of isotopes and nuclides charts since the years fifties has been discarded to adopt recently the value of 1.1 millions years. Now this value itself is again questioned and the 79 present best estimate of Se T½ is taken as 295 000 years.

Selenium chemistry is intermediate between that of sulfur and tellurium. Basic information on the main chemical forms of inorganic and organic selenium is also provided here to make easier the reading on the next chapters before to provide more detailed information in further specialised sections.

A1.2 Inventory and isotopic composition

Selenium is present in weak but ponderable amounts in the spent fuel and HLW. The total quantity of selenium to be disposed of for the present Belgian nuclear program is about 400 kg. Table A1.2.1 presents the quantity of the various selenium isotopes calculated with the Origen code for a spent fuel with high burn-up (50 GWd/tHM) and its isotopic composition compared to that of selenium in the nature (Plant et al., 2004).

153/328

Table A1.2.1: Isotopic composition of selenium in nature and in spent nuclear fuel. Selenium Natural Isotope Isotope in Spent Fuel Isotope in Spent Fuel Isotope Abundance@ (%) Abundance# (%) Quantity# (g/tHM) 74Se (stable) 0.87 % — — 75 Se (T½ 120 d) — — — 76Se (stable) 9.02 % 8.86 E-03 % 7.18 E-03 77Se (stable) 7.58 % 1.40 % 1.13 78Se (stable) 23.52 % 4.77 % 3.86 79 Se (T½ 295 ka) — 8.52 % 6.90 80Se (stable) 49.82 % 25.56 % 20.70 82Se (stable) 9.19 % 59.76 % 48.40 Sum all 100.00 % 100.00 % 80.997 @ From Table 1, p. 19, Chapter 9.02, Arsenic and selenium. Vol. 9 Environmental Geochemistry. In the Treatise of Geochemistry (Plant et al., 2004). # Lionel Boucher, CEA (inventories in spent fuel with a burn-up of 50 GWd/tHM used for the Red-Impact EC Project, Jan Marivoet, Personal Communication).

The total quantity of all selenium isotopes in the spent fuel (burn-up of 50 GWd/tHM) is about 1 mol Se per ton heavy metal (tHM) whose 79Se represents about 8.5 %, i.e., in the order of ~ 0.1 mol / tHM. For different geometries of disposal of vitrified HLW and spent fuel with various burn-up, it corresponds to a quantity of 79Se / m gallery comprised between 0.1 and 1 mol 79Se per current meter of disposal gallery. This quantity expressed as mol / m gallery is comparable to that of 129I / m gallery because both isotopes are the most frequently produced by the fission of uranium nuclei. In other type of spent fuel, the 79Se isotopic abundance can even be higher, up to 12 % such as in advanced reactor fuel (inventory from the Red Impact EC project, Marivoet, 2005, Personal Communication).

A1.3 Uncertainties related to the value of selenium-79 half-life and recent changes In most handbooks (e.g., CRC: Weast ed., 1968-1969a; Lide ed., 1995), Nuclides Charts (Pfennig et al., 1995), and tables of isotopes (e.g., Lederer et al., 1967), the half-life of Se-79 is approximately 6.5 × 104 years. This is also the value used in all inventory calculations as well as release and transport analyses of most of the performance assessment studies. However, in the last decade (nineties) two very different values have also appeared in the literature for the half-life of 79Se. The time span of the values published for 79Se half-life spreads over nearly two orders of magnitude (the ratio between the maximum and the minimum values is about 17). The main values available for the half-life of 79Se are the following: 65 000 y (Parker et al., 1949; Lederer et al., 1967); 650 000 y (likely a simple

154/328

typing mistake present in the chart of the nuclides of KAPL Inc.; see Parrington et al., 1996); 1.1 My (Singh, 1993; Vieno and Nordman, 1999), and more recently 295 000 y (Jiang et al. 2001, 2002; Singh, 2002). Before the TILA-99 report in which Vieno and Nordman (1999) put in question the commonly accepted value of 65 000 y for the 79Se half-life, all the performance assessment calculations dealing with Boom Clay were performed with this value. It was thus the case for the following PA studies: the PAGIS, PACOMA, Updating 90, Everest, and SPA studies, and other performance assessment calculations results (Mallants et al., 1999) used for the SAFIR 2 report published in 2001. After 2001, the value of 79Se 6 T½ = 1.1 × 10 years was used for the first time for performance assessment (PA) calculations at the Mol site in the SPA+ report by Sillen and Marivoet (2002). From the point of view of performance assessment, this change means that the activity inventory of 79Se would be about a factor 17 lower than with T½ = 65 000 years because in the ORIGEN calculations the number of atoms have been converted into Becquerels using the half-life of 6.5 × 104 years. On the other hand, in the release and transport analyses, the decay of 79Se in the clay barrier would, of course, be slower with the new half-life. Afterwards, and more recently, this value has been again reassessed for performance assessment studies made in 2004 in the frame of the European project Red Impact (Jan Marivoet, 2005, Personal Communication).

The half-life presently recommended for 79Se in the nuclear database NuDat 2.1 of the US Brookhaven National Laboratory (BNL) is 2.95 × 105 years, as determined by Song-Sheng et al. (1997; 2001) and Jiang et al. (2001, 2002), and selected in the Nuclear Data Sheets (NDS) by Singh (2002). The same value of 2.95 × 105 years for 79Se is also chosen by Bé et al. (2005) in the last version of the “Table of recommended values for half-lives” adopted by the Laboratoire National Henri Becquerel (LNHB) of CEA (CEA, 2006).

The recent recognition of the importance of the mobile fission products (MFP’s) as the main contributors to the dose-to-man for the deep disposal of HLW and SF has fostered a reappraisal of the half-life of MFP’s and activation products: a.o., 32Si, 79Se, 93Zr, 99Tc, 126Sn. The emergence of advanced accelerator mass spectrometry (AMS) to measure with high sensitivity and superior precision the small number of atoms present in clusters of less than 106 atoms has allowed a redetermination of the half-life of several critical MFP’s (see a special issue (2008) of Nuclear Instruments and Methods on “Beam Interactions with Materials and Atoms”, a.o., Tuniz and Norton, 2008; Synal and Wacker, 2008).

79 T½ for Se is trivially obtained (A = λ N; T½ = 1/λ = N/A) from the ratio of a small number of 79Se atoms (N) to their activity (A, in Bq) by measuring by AMS a bunch of highly purified 79Se atoms (N) and counting their activity independently with a high accuracy. The purity of the isolated radionuclide of interest is a key factor in the determination of its T½. It is thus necessary to extract and to separate very efficiently traces of MFP’s present in large amount of other radionuclides obtained from spent fuel dissolution. Therefore, it is not surprising that

155/328

the same teams in charge of reassessing the 79Se inventory in spent fuel also often provide its new T½ values (Comte 2001; Comte et al., 2001; 2002a,b; 2003; Bienvenu et al., 2007).

79 A new value of 377 ka for Se T½ has been recently determined by Bienvenu et al. (2007). This value is ~ 25 % larger than that of 295 ka (Jiang Song-Sheng, 2001; Singh, 2002; Bé 79 et al., 2005). Meanwhile, it seems that the new determinations of Se T½ converge towards the interval 295 – 377 ka.

The various values assigned to the half-life of 79Se, their main bibliographic sources, and the evolution of their use in PA studies as a function of time are summarized hereafter in Table A1.3.1. More information on the nuclear decay mode of 79Se and 75Se is also available in Section A1.4.

Table A1.3.1: Revised values of the half-life time of 79Se available in the literature and evolution of its use in performance assessment studies for Boom Clay. T½ values ranging on about two orders of magnitude.

Reference on Publishing T½ Value Used in Comments 79 79 T½ Se Year (year) PA Studies on T½ Se Parker et al. (1949) (1949) 65 000 SAFIR 2 ancient value SPA accepted till ~ 2 000 Parrington et al. (1996) (—) 650 000 never used typing mistake ? Singh (NDS, 1993) in: (1993) 1 100 000 SPA+ 1st major revision Vieno and Nordman (1999) (1999) Trancom-II used till 2 004 Song-Sheng et al. (1997; 2001) (1997) 295 000 Red Impact 2nd major revision Jiang et al. (2001, 2002), (2001) used after 2 004 Singh (2002); NuDat 2.1 (2002) Bé et al. (2005), and, (2005) 356 000 3rd major revision LNHB, CEA (2006) (2006) used after 2 004 Bienvenu et al. (2007) (2007) 377 000 — not yet used NuDat 2.1 (2005) Nuclear Database version 2.1 from the US National Nuclear Data Center (NNDC), Brookhaven National Laboratory (BNL). URL: http://www.nndc.bnl.gov/nudat2 as seen on 02-Feb-2005. NDS: Nuclear Data Sheets. Laboratoire National Henri Becquerel: LNHB. PA: Performance Assessment.

Because of the decrease in the activity inventory and because selenium is solubility-limited in reducing conditions, but sorbs rather weakly and thus migrates rapidly through the geosphere, the overall effect in most cases would be a decrease in the maximum release rates (Bq a-1) as stated by Vieno and Nordman (1999) in TILA-99.

156/328

A1.4 Nuclear decay mode of 79Se and 75Se

The respective nuclear decay modes of the two main selenium isotopes relevant for our studies (79Se, the long-lived radionuclide, and, 75Se, the short-lived tracer mimicking the former behaviour for laboratory tests) are the following:

1. Selenium-79 (safety studies) to be disposed of in the Boom Clay formation:

79 79 − − 34Se → 35Br (stable) + 1 e (β decay 100 %, T½ = 295 ky)

2. Selenium-75 commonly used as a tracer in the laboratory experiments:

75 75 34Se → 33As (stable) (EC 100 %, T½ = 120 days)

One can notice that 79Se (excess of neutrons) gains a proton after its beta decay to produce the next stable element, 79Br, while at the opposite, 75Se (deficit of neutrons) loses a proton after an electronic capture (EC) to decay to the previous stable element: 75As. More details on the full decay schemes of both radioisotopes and the energy of β particles and γ rays emitted are available in the Nuclear Data Sheets, Table of Isotopes (Lederer et al., 1967), or nuclear database (NuDat 2.1).

A1.5 Selenium inorganic and organic chemistry

Chemistry of selenium resembles that of sulfur: although the number of inorganic species found in natural systems is relatively small, many different types of organo-selenium compounds exist in nature. These substances are synthesized by micro-organisms, phytoplankton, such as coccolithophores (Emiliania Huxleyi, see a.o., Obata, 2003;2004), and many plants. The cycle of selenium in the ocean is as complex as this of sulfur and will briefly be treated in Appendix A2 on: Natural selenium in the environment and in Boom Clay.

Table A1.5.1 and Fig. A1.5.1 give a general overview on the main inorganic species of selenium considered in the present study and in the thermodynamic calculations dealing with selenium in natural aquatic systems.

157/328

Table A1.5.1: Oxidation states and chemical forms of major inorganic compounds of selenium. Oxidation State Inorganic Compound Aqueous Species Solid Phase

2– Se(VI) Selenate SeO4 CaSeO4 · 2 H2O

2– Se(IV) Selenite SeO3 CaSeO3

Se(0) Elemental selenium Se0 Se0

– – Se(-I) Diselenide Se–Se FeSe2, Fe3Se4

Se(-II) Selenide HSe– FeSe

1.2 - HSeO4

.8 H2SeO3(aq)

-- SeO4 HSeO- .4 3

Se(c)

-- Eh (volts) 0 SeO3 H2Se(aq) FeSe (c) - 2 HSe –.4 FeSe(c)

25°C –.8 0 2 4 6 8 10 12 14 pH

Figure A1.5.1: General pH–Eh diagram for selenium in water at 25 °C.

158/328

Table A1.5.2 also provides a non-exhaustive list of some organic species of selenium commonly found in nature and produced by living organisms.

Despite the fact that the organic forms of selenium are not presently considered in our thermodynamical speciation calculations, they are mentioned here simply to illustrate the frequent presence of selenium in natural systems in close association with organic matter. The reason is that selenium plays a very specific role as a micronutrient in most of the biological systems where its enzymatic activity protects cell membranes against oxidation. This feature must be kept in mind when discussing the relatively well documented association of selenium with natural organic matter (NOM) (MacGregor, 1997, PhD Thesis): selenium loosely associated with NOM in laboratory interaction batch tests must not be confused with selenium naturally present into the chemical structure of NOM (covalent bond). Kerogen is known to often contains selenium as it is the case in coal deposits in China (Wang et al., n.d.; Wen et al., 2006; 2007) and in the Western US (Kulp and Pratt, 2004; Oram et al., 2008).

Table A1.5.2: Some major organic compounds of selenium commonly found in nature and whose degradation products are incorporated in natural organic matter (NOM) by micro-organisms. Oxidation State Organic Compound Abbreviation Chemical Formula

Se(-II) Dimethylselenide DMSe CH3SeCH3

Se(-I) Dimethyldiselenide DMDSe CH3Se–SeCH3

Se(-II) Diallylselenide DASe Allyl-Se-Allyl

Se(-I) Diallyldiselenide DADSe Allyl-Se–Se-Allyl

+ + Se(-II) Trimethylselenonium TMSe (CH3)3Se

+ – Se(-II) Selenomethionine Se-meth. H3N CHCOO ·CH2CH2SeCH3

+ – Se(-II) Selenocysteine Se-cyst. H3N CHCOO ·CHSeH

Se(-II) Se-Glutathione GSH-Px Peroxidase enzyme (protein)

Se(-II) Selenocyanate — SeCN–

Dimethylselenide (DMSe) and dimethyldiselenide (DMDSe) are volatile compounds (Oremland and Zehr, 1986; Stork et al., 1999) that could contribute to man exposure to 79Se via the gas route in some particular conditions if sulfato-reducing bacteria (SRB) are at work in water-containing spaces near non-closed galleries (Karsten Pedersen, Personal Communication, May 2009) or in near surface installations (Peitzsch and Kersten, 2008).

Diallylselenide (DASe) and diallyldiselenide (DADSe) are produced by alliae plants (onions, garlic, …) while selenomethionine and selenocysteine are selenium containing amino-acids.

159/328

These amino-acids are the building blocks of complex proteins and enzymes such as glutathione peroxidase (GSH-Px), a free radical interceptor critical for cell life and preventing cancer. It explains why selenium is bioconcentrated by many organisms, from marine unicellular algae (coccolithophores, Emiliania Huxleyi, see a.o., Obata, 2003; 2004) to superior plants (broccoli, mustard, Brazil nuts, …) and easily accumulate in the food web.

Although selenocyanate (analogous to thiocyanate, SCN–) is an artificial compound where selenium is covalently bound to a nitrile (cyanide) group, it is also mentioned in Table A1.5.2 just to remind the strong tendency of selenium to be associated with organic carbon, like sulfur does.

As concluding remark, if sulfato-reducing bacteria (SRB), or other micro-organisms (such as some yeasts) capable of metabolizing selenium, are present in the clay suspension systems studied in the laboratory for solubility or sorption tests, the presence of organic molecules of selenium should not be ruled out: indeed, microbial activity could bias the results of some experiments or cause selenium loss to the atmosphere by the formation of gaseous organic compounds. The easiest way to overcome micro-organisms growth is to limit the duration time of the experiments to benefit of the initial time lag hindering their development (see the Technical Note of the Mont Terri BN experiment, lab supporting microbial tests made at BRGM by Pauwels et al., 2009). The possible presence of organic selenium in natural Boom Clay water samples should also be carefully controlled before attempting to determine inorganic selenium “true” solubility levels. More information is given in Appendix A2 on: Natural selenium in the environment and in Boom Clay.

A1.6 Chemotoxicity of selenium Although the aspects related to the chemical toxicity of selenium are not essential in the context of the present study, this section provides a short summary on the question of the very particular toxicity of selenium. The aim is not only to provide the interested reader with a general background on the subject, but also to preserve the information gathered during our reading on selenium. So, if in the future one must address the problems dealing with the chemo-toxicity of a deep disposal site, when stable toxic elements will be eventually released to the biosphere, the information compiled here on selenium will remain available for further works if needed.

Selenium is a trace mineral that is essential to good health but required only in small amounts (ODS–NIH, 2004; Thomson, 2004; Goldhaber, 2003). Selenium is incorporated into proteins to make selenoproteins, which are important antioxidant enzymes. The antioxidant properties of selenoproteins help prevent cellular damage from free radicals. Free radicals are natural by- products of oxygen metabolism that may contribute to the development of chronic diseases such as cancer and heart disease (ATSDR, 2003; Goldhaber, 2003; Combs and Gray, 1998).

160/328

Other selenoproteins help regulate thyroid function and play a role in the immune system (McKenzie et al., 1998; Levander, 1997; Arthur, 1991; Corvilain et al., 1993).

Selenium is an oligo-element necessary to mammals and humans in daily intake of a few tenths of micrograms per day. Because of its large oxido-reduction reactivity, selenium is a key player in the function of many proteins and enzymes, as, e.g. in the catalytic activity of the glutathione peroxidase which scavenges the peroxide free radicals in the intracellular liquid, protecting the phospholipids of the cell membranes against oxidation. It is also involved in the metabolic processes of iodo amino-acids, as tri-iodo-thyroxin, and is indispensable to the proper function of the thyroid gland. More than hundred different selenium-containing enzymes have been inventoried up to now without knowing all their functions. It could also be implied in our immune-protection and our defence against viruses.

Paradoxically, if the spectrum of activity of organo-selenium is very broad and probably largely unexplored, its concentration range between deficiency and toxicity is very narrow as illustrated by the values listed in Table A1.6.1. The daily selenium intake for a human adult cannot be lower than 40 μg day-1 to prevent severe deficiency, and higher than 400 μg day-1 to avoid its toxicity. Outside this small interval, human health is at risk. The recommended daily selenium intake for a human adult is even in a narrower range between 55 μg day-1 and 75 μg day-1. So, to paraphrase the citation of the Swiss physician Paracelsus “Dosis sola facit venenum”, the right dose of selenium differentiates the poison from the remedy.

Table A1.6.1: Deficiency and toxicity thresholds of selenium for human diet. Threshold Daily intake Daily intake (WHO) (μg day-1) (μmol day-1) Deficiency < 20 0.25 Minimum > 40 0.51 Adequate (woman) 55 0.70 Adequate (man) 75 0.95 Toxic > 400 > 5.06 Morbid – Lethal > 1 000 > 12.66 (source: subset of data from Table 12, p. 56 from Plant J.A. et al. (2004). WHO: World Health Organisation (WHO, 2003).

The recommended selenium daily intake is 0.9 μg kg-1 (11 nmol kg-1 ) of body weight in human adults (WHO, 2003). The maximum admissible dose of selenium for man per body weight is 5 μg kg-1 day-1 (60 nmol kg-1 day-1) for the US Environmental Protection Agency

161/328

(EPA). The maximum admissible concentration of selenium in drinking water is 10 μg dm-3 (1.27 × 10-7 mol dm-3) as set by the World Health Organisation (WHO, 2003).

Selenium deficiency: severe lacks of selenium in the human diet are relatively rare but have been reported in China where two endemic diseases affecting respectively the heart and the bone joins have been diagnosed in remote rural areas where soils are selenium-deficient: − the Keshan Disease (KD) is a cardiomyopathy, and; − the Kashin-Beck Disease (KBD) is an osteoarthropathy. These diseases can be corrected by re-equilibrating the diet of the local population and by external supply of food.

Selenium toxicity: the human symptoms of selenosis, the poisoning by too high selenium intakes, are hairs loss, nails deformity, and deficiency of the peripheral nervous system. Teratogenic and carcinogenic effects have not yet been clearly reported up to now for humans, but have been well recognised for fishes and aquatic birds whose selenium metabolisms is known to notably differ from that of mammals (Frankenberger and Benson, 1994). Water is rarely the main source of contamination.

A1.7 References

Arthur J.R. (1991) The role of selenium in thyroid hormone metabolism. Can. J. Physiol. Pharmacol. 69, 1648–1652.

ATSDR (2003) Toxicological profile for selenium (update September 2003). Selenium CAS # 7782-49-2. Agency for Toxic Substances and Disease Registry (ATSDR). Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service. ToxFAQs address: http://www.atsdr.cdc.gov/toxfaq.html.

Combs G.F, Jr. and Gray W.P. (1998) Chemopreventive agents: Selenium. Pharmacol. Ther. 79, 179–92.

Corvilain B., Contempre B., Longombe A.O., Goyens P., Gervy-Decoster C., Lamy F., Vanderpas J.B., and Dumont J.E. (1993) Selenium and the thyroid: How the relationship was established. American Journal of Clinical Nutrition 57 (2 Suppl.), 244S–248S.

Goldhaber S.B. (2003) Trace element risk assessment: essentiality versus toxicity. Regulatory Toxicology and Pharmacology 38, 232–242.

162/328

Levander O.A. (1997) Nutrition and newly emerging viral diseases: An overview. Journal of Nutrition 127, 948S–950S.

Longnecker M.P., Taylor P.R., Levander O.A., Howe M., Veillon C., McAdam P.A., Patterson K.Y., Holden J.M., Stampfer M.J., Morris J.S., Willett W.C. (1991) Selenium in diet, blood, and toenails in relation to human health in a seleniferous area. American Journal of Clinical Nutrition 53, 1288–1294.

McKenzie R.C., Rafferty T.S., Beckett G.J. (1998) Selenium: an essential element for immune function. Immunology Today 19, 342–345.

ODS – NIH (2004) Dietary supplement fact sheet: selenium. 9 pp., 66 references. Document last updated: 08/01/2004. http://ods.od.nih.gov/factsheets/Selenium_pf.asp. Office of Dietary Supplements (ODS). National Institutes of Health (NIH). Bethesda, Maryland 20892 USA. http://ods.od.nih.gov.

Thomson C.D. (2004) Assessment of requirements for selenium and adequacy of selenium status: a review. European Journal of Clinical Nutrition 58, 391–402.

163/328

164/328

A2. Natural selenium in the environment and in Boom Clay

165/328

166/328

A2 Natural selenium in the environment and in Boom Clay

A2.1 Overview Selenium is a very ubiquitous element in Earth's crust, but at trace concentration. Its chemical behaviour is intermediate between that of sulfur and tellurium. Selenium geochemical cycle is similar to that of sulfur, although its abundance is much lower. As sulfur, selenium also plays a critical role in most of the biochemical systems. This chapter first gives an overview of the distribution of selenium in igneous and sedimentary rocks, its main natural sources on Earth, and principal origins of man-made pollution linked to its industrial use. Then the general occurrence of selenium in the environment and water is discussed to put in perspective the concentration of natural selenium measured in Boom Clay and to illustrate the reasons of its preferred association with sulfides minerals (FeS2) and organic matter.

A2.2 Primary sources of selenium in the earth crust and sediments

Although no economically viable ore deposits of selenium pure minerals are presently known, selenium is an ubiquitous element present at trace concentrations in volcanic rocks and in many reducing sedimentary environments.

The primary source of selenium on earth surface probably arises from the volcanic activity in which selenium accompanies sulfur in volcanic effluents. Selenium can be found in volcanic rocks especially in basalts where it may be dissolved in volcanic glass (possible natural analogue of vitrified HLW). Selenium in volcanic areas can be used as a pathfinder in prospecting for volcanogenic ore deposits. Soils in the neighbourhood of volcanoes tend to have enriched amounts of selenium. After volcanic rocks weathering, oxidation, mobilisation, and transport, selenium may be reduced and trapped by precipitation in sulfide minerals.

Selenium whose chemistry is very close to that of sulfur, but with a larger ionic radius is found associated in reducing sediments and soils with heavy metal sulfides (pyrite, FeS2; sphalerite, ZnS; galena, PbS; chalcopyrite, CuFeS2, and chalcocite, Cu2S). Selenium exists in association with iron in two rare selenide minerals: ferroselite (FeSe2, see: Brookins, 1988) and its polymorph variety, dzharkenite (FeSe2, see: Ryser et al., 2005). Selenium occurs also in eucairite (CuAgSe), crooksite (CuThSe) and clausthalite (PbSe), but these minerals are too rare to be used as a major source of selenium. Today the main source of selenium in the world arises from the residues of the copper sulfides processing. Selenium is a by-product of copper refineries. It is mainly recovered from the anodic mud deposited during copper electrolysis (Coget, 1966; Ladrière, 1969; KULeuven–UCL, Laboratory of Prof. R. Breckpot, Leuven), and also in a less manner from flue dust of pyrometallurgical treatment of Cu2S.

167/328

Being a redox-sensitive element, the mobility of selenium is also analogous to that of uranium. Both elements are dissolved in their higher valence and very mobile when their primary bearing minerals are undergoing oxidation and dissolution (weathering in supergene conditions). At the opposite, they precipitate and become trapped within sediments when they encounter strongly reducing conditions in subsurface environments, e.g., in the presence of organic matter or sulfides. According to Deliens et al. (1981), selenium may occasionally be found in close association with uranium in some Se-bearing sulfide minerals such as Cu9S5 (digenite). Uranyl selenites occur where uranium-bearing seleniferous sulfides undergo oxidation, as reported by Vochten et al. (1996) who determined the structure of piretite,

Ca(UO2)3(SeO3)2(OH)2 · 4 H2O, a new calcium uranyl selenite from Shinkolobwe mine, Shaba, Democratic Republic of Congo. Although the occurrence of such an association in nature is not so common, it cannot be ruled out in the near-field of a repository for spent 79 nuclear fuel under oxidizing conditions. Indeed, the retention of Se on UO2 has already been observed in spent fuel alteration products (Trombe et al., 1985; Chen et al., 1999, 2000; Ewing, 2001) (see also Appendix A5: Immobilisation of selenium in the near-field).

Selenium is also present in coal deposits, in oil fields, and in organic matter-rich sediments like clays and shale at the source of oil reservoir. Selenium can be chiefly associated to inorganic or organic sulfides respectively. Two distinct routes of incorporation of selenium in organic matter-rich sediments can be envisaged.

On one hand, selenium simply accompanies inorganic sulfur in sulfide bearing minerals. Metallic sulfides are most frequently precipitated by sulfato-reducing bacteria (SRB) in sediments. Here, the only role of organic matter is to fuel the SRB bacteria: OM acts as electron donor in the system, and sulfate/selenate as electron acceptor. As a result, the formed sulfides/selenides coprecipitate with heavy metals.

On the other hand, organic matter itself can be very rich in organic-sulfur compounds originating from the maturation/degradation of proteins and peptides containing cysteine and methionine, the two main sulfur-based amino-acids. Selenium can take the place of sulfur in these amino-acids giving rise to seleno-cysteine and seleno-methionine (see Table A1.5.2).

Indeed, the biochemistry of selenium present in proteins is complex and far from being entirely explored. The first discovered Se-containing enzyme (Rotruck et al., 1973), the glutathione peroxidase (GSH-Px) acts as a critical anti-oxidant catalyst essential for the survival of living cells in mammals and humans. It protects phospholipids of the cell membrane against aggression of free-radicals.

So, it is not surprising to find selenium incorporated in the natural organic matter and kerogen of terrestrial and marine origins. Selenium may have been metabolised by a large variety of

168/328

organisms such as bacteria, fungi, algae, unicellular phytoplankton (coccolithophorids), or in more evolved organisms as plants, fishes and animals, and recycled in matured organic matter after the death of these organisms.

A2.3 Bioconcentration of selenium by a coccolithophorid, Emiliania huxleyi, and correlation selenium/calcium carbonate

The biochemical cycle of sulfur is very rich and can give interesting analogies and clues for understanding the behaviour of organic selenium in nature and its accumulation and distribution in sedimentary formations during geological ages. Beside the two sulfur- containing amino-acids (methionine and cysteine) and related proteins, intermediate low- molecular mass organic compounds (LMC’s) are synthesized by marine bacteria and planktonic organisms in the oceans. LMC’s are continuously released in the ocean water, emitted in the atmosphere (volatile compounds), or accumulated in marine sediments and dynamically participate to the global sulfur cycle.

Two examples of LMC’s molecules playing a non-negligible role in ocean and marine atmosphere chemistry are the dimethyl sulfide (DMS) gas and dimethylsulfoniopropionate

(DMSP); DMS (CH3SCH3), along with methanethiol (CH3SH), is itself a breakdown product of DMSP: • The dimethyl sulfide gas is the most abundant biological sulfur compound emitted to the atmosphere by marine organisms and one of the main components of the “smell of the sea”. It is produced by marine bacteria carrying the appropriate regulatory genes (Todd et al., 2007). Oceanic dimethylsulfide is also produced by phytoplankton and zooplankton (Dacey et al., 1986). When DMS is released to the marine atmosphere, it is oxidized by

O2 and UV light in various sulfur-containing compounds, a.o., dimethyl sulfoxide (DMSO), and ultimately to sulfuric acid. In their turn, sulfate aerosols above oceans act as cloud condensation nuclei and affect global climate by different feedback mechanisms (CLAW hypothesis proposed by Charlson et al., 1987). • Dimethylsulfoniopropionate, a secondary metabolite of many marine algae, also accounts for most of the organic sulfur fluxes from primary to secondary producers in marine microbial food chains (Vila-Costa et al., 2006). It also plays a role in the life of coral reef habitats as a signalling agent between algae other reef organisms.

In addition, selenium is an essential micronutrient necessary to the growth of marine phytoplankton and its biochemical cycle in oceans is likely governed by an even greater complexity than that of sulfur cycle. After active uptake by planktonic algae, organic selenium is incorporated in the soft tissues associated to their shells. After their death planktonic shells sink at the bottom of the sea where they form marine sediment deposits

169/328

which further evolve with diagenesis. Oceans represent thus a major biogenic source of selenium (Amouroux et al., 2001) that cannot be ignored to understand the selenium distribution in sedimentary geological formations such as Boom Clay.

The coccolithophorids – a group of unicellular, marine, planktonic, photosynthetic algae protected by a spherical envelop made of multiple CaCO3 disks (coccoliths) embedded in a gelatinous sheath – not only play a major role in the global carbon cycle, but also contribute to the selenium bioaccumulation in marine sediments. Chalk deposits not only sequesters CO2 but also selenium, a process whose importance was only recently recognised by marine biologists and geologists.

Coccolithophorids, such as Emiliania huxleyi, one of the most abundant phytoplanktonic species in the oceans, possess a very efficient bioconcentration metabolism of selenium to uptake this element essential to their growth (Araie et al., 2003; Danbara and Shiraiwa, 1999, 2007). Indeed, as experimentally determined by Obata et al. (2003, 2004), 75Se-labelled selenite is rapidly absorbed by coccolithophorids, transformed in a temporary pool of LMC’s, and finally incorporated in seleno-proteins and organic molecules associated to their calcareous shells.

During geological periods of hot climate, such as during the Upper Cretaceous epoch, massive seasonal algae blooms, and the preservation of selenium in degradation-resistant organic matter intimately associated to coccoliths, have caused the sedimentation of selenium- rich chalk and shale formations. This was particularly the case of large areas in the Rocky Mountains (Western USA) today affected by selenium-related environmental problems (Olivier Leupin, Nagra, personal communication, cfr., his Postdoc on Se at USGS Denver Colorado).

Selenium is considered by Presser et al. (2004) as a geochemical exploration tool that reveals an ancient productive biological environment.

So, in geological formations such as Boom Clay, selenium could not only be associated with pyrite or organic matter (kerogen), but also correlated with carbonate-rich sediment layers deposited by coccolithophorids. Such a selenium/carbonate correlation is presently search for by Jean Garignan (Nancy University, CRPG, personal communication, 2008) on vertical profiles in the Callovo-Oxfordian Clay formation at the Bure underground research laboratory (GdR ForPro-II team). A similar and unexpected correlation between iodide and aragonite shell debris has also been recently evidenced by Claret et al. (2010) (Tournassat and Gaucher, personal communication, March 2009; Claret et al. (2009) poster presented on the topic at the Migration’09 Conference in Kennewick, Washington, USA).

170/328

A2.4 Environmental pollutions related to natural and industrial selenium sources

Selenium main applications deal with glass industry where it is used at high concentration as a pigment for special ruby-coloured glasses, and at low concentration (pale pink) as a decolorizing agent for window glass to counterbalance the green colour of Fe2+.

Its sensitivity to light coupled to its p-type semi-conductor properties have promoted its used in Xerography (charges transfer in the Xerox photocopy process), solar cells, and AC-to-DC rectifiers (CuSe).

Most of the pollution by selenium occurs from metallic sulfides ores mining and processing, but also from the subsequent oxidation of mine tailings, mobilising large amounts of selenate in the run-off waters.

Oxidation also plays a key role in the natural weathering of organic-rich clay and shale such as Cretaceous Pierre Shale in the central west America (Wyoming, Colorado and South Dakota). When selenium occurs in alkaline soils (such as those of the Western US sites with a recent volcanic activity) and is oxidized as selenate, selenium becomes water-soluble. This form is highly toxic, easily leached from the soil, and available to plants. Discharge of agricultural water run-off from seleniferous soils has caused deep damages to the ecosystems and to wildlife (aquatic birds) as illustrated by the considerable number of studies dedicated to the remediation of the Kesterson reservoir (San Joaquin Valley, California).

Combustion of selenium-rich fossil fuels (oil, coal, bituminized shale) can also directly release to the atmosphere SeO2 gas similar to SO2 responsible for the acid rains, while oil refineries are faced with selenium problems in processes where selenium intimately tracks sulfur. Finally furnaces of glass factories may discharge significant amounts of volatile selenium in plant waste gases and dust.

A2.5 Concentrations of natural selenium in the environment

Selenium is a trace element less abundant than uranium and thorium in the Earth’s crust. According to the CRC Handbook of Chemistry and Physics (Weast ed., 1968-1969b,c), selenium average abundance in Earth’s crust is ~ 0.08 mg kg-1 about 50 times less than uranium (4 mg kg-1), and 150 less than thorium (12 mg kg-1). However, as mentioned above for uranium, because of its great sensitivity to redox conditions, it can be reconcentrated to very high level in reducing sedimentary rocks, especially in coal and organic-rich shale. This may pose an important concern for human and animal health in many areas in the world.

171/328

Table A2.5.1 summarizes some data taken from Plant et al. (2004) and presents an overview of the concentrations of natural selenium in the Earth’s crust and in various types of rocks and fossil fuel deposits. For the sake of comparison, mean values of selenium concentrations measured in Boom Clay by Maes et al. (2004a,b) are also reported in the same table, but will further be discussed.

Table A2.5.1: Natural concentrations of selenium in the Earth’s crust and several types of igneous and sedimentary rocks, oil, and coal, by increasing concentration ranges. Selenium Materials (mg kg-1)

Earth’s crust 0.05

Igneous rocks Granite 0.01 – 0.05 Volcanic rocks 0.35 Volcanic tuffs 9.15

Sedimentary rocks Sandstone < 0.05 Limestone 0.03 – 0.08 Marine carbonates 0.17 Phosphate (apatite) 1 – 300 Oil 0.01 – 1.4 Coal (USA) 0.46 – 10.7 Black shale (Western USA)* 1 – 675 Carbon shale (China) 206 – 280 Mudstone 0.1 – 1 500 Stone coal (China) up to 6 500

Boom Clay Extracted and fractionated pyrite# 12.2 – 33.0

Calculated mean value for Boom Clay (BC) 0.61 – 1.65 (considering 5 % pyrite in BC)

Source: Plant et al. (2004) subset of data from Table 7, p. 46, Chap 9.02, Arsenic and selenium. In Environmental Geochemistry (ed. Lollar B.S.) Vol 9 Treatise on Geochemistry (eds. Holland H.D. and Turekian K.K., 2004). # Values from Maes et al., (2004a,b), Table 4, p. 27 of Annex 14 to WP 5 & 8 Migration experiments and demonstration of model concept for trivalent radionuclides. * Cretaceous deposits also rich in coccolithophores and affected by volcanism.

172/328

The selenium content in pyrite extracted from Boom Clay (12.2 – 33.0 mg kg-1) is clearly higher than in most of the igneous rocks (granite, basalt) and sedimentary rocks (sandstone, limestone). The selenium concentration in Boom Clay pyrite is in the range of that observed in coal and organic-rich shale: both may contain a lot of seleniferous metallic sulfides, and also organo-selenium bound to organic matter. The mean selenium concentration calculated for the Boom Clay as a total rock on the basis of a content of 5 % of pyrite is in the range of 0.61 – 1.65 mg kg-1, a value on the lower side for the range of mudstone.

Table A2.5.2 presents concentrations of natural selenium commonly found in various water bodies in different regions of the world. The Se concentration range is spread on about five orders of magnitude, from about 10-10 to 10-5 mol dm-3. The highest values are reported for the South-West of the USA which have been extensively studied because of the weathering of the seleniferous Pierre Shale and the following environmental problems posed by selenate in surface water. However, most of the Se concentration values given in Table A2.5.2 lie in the range 10-9 to 10-8 mol dm-3. Se concentrations in clay porewater are lower than these of problematic catchments, such as these of the San Joachim Valley in California (USA), but one order of magnitude higher than in seawater (1.14 × 10-9 – 2.15 × 10-9 mol dm-3) or estuarine water.

The selenium concentration measured in Boom Clay porewater (2.41 × 10-8 mol dm-3) is about 11 times higher than in seawater if one refers to the average concentration of 2.15 × 10-9 mol dm-3 estimated for seawater by Thomson et al. (2001). The reason is that an important fraction of natural selenium in Boom Clay water is likely associated with the dissolved organic matter (DOM), and presumably incorporated in the chemical structure of humic acids, where it should occupy a position similar to that of sulfur.

A2.6 Concentrations of natural selenium in Boom Clay

Natural selenium concentrations have been measured by high resolution ICP-MS in pyrite separated from Boom Clay and in porewater collected from a piezometer installed in the HADES underground research facility (URF). A strong enrichment in selenium is observed in pyrite whose Se concentration is about four orders of magnitude higher than that of Boom Clay porewater. This is compatible with a redox-control of the mobility of selenium by pyrite as also observed for uranium in Boom Clay.

173/328

Table A2.5.2: Natural concentrations ranges of selenium in various water bodies in the world: rain water, river and lake water, seawater and estuaries, groundwater and porewater. Boom Clay porewater is given for comparison. Se concentration range Se concentration range Water body (Country) (μg dm-3) (mol dm-3)

Rain water Polar ice (—) 0.02 2.53 × 10-10 Various (—) 0.04 – 1.40 5.06 × 10-10 – 1.77 × 10-8

River and lake water Mississippi River (USA) 0.14 1.77 × 10-9 Amazon River (Brazil) 0.21 2.66 × 10-9 Colorado River !!! (USA) <1 – 400 <1.27 × 10-8 – 5.06 × 10-6

Seawater and estuaries Seawater (—) 0.09 – 0.17 1.14 × 10-9 – 2.15 × 10-9 San Francisco Bay (USA) 0.10 – 0.20 1.27 × 10-9 – 2.53 × 10-9

Groundwater East Midlands Triassic Sandstone (UK) <0.06 – 0.86 <7.59 × 10-10 – 1.09 × 10-8 Bengal Basin (Bangladesh) <0.50 <6.33 × 10-9 Punjab !!! (Pakistan) avg. 62 avg. 7.85 × 10-7 Colorado River catchments !!! (USA) up to 1 300 up to 1.65 × 10-5 San Joaquin Valley, California !!! (USA) <1 – 2 000 <1.27 × 10-8 – 2.53 × 10-5

Porewater Baseline, Lake Macquarie (Australia) <0.20 <2.53 × 10-9 Smelter-impacted Lake Macquarie (Australia) 0.30 – 5.00 3.8 × 10-9 – 6.33 × 10-8

Boom Clay Porewater#

-8 HADES EG-BS piezometer (Mol, Belgium) 1.90 2.41 × 10

Source: Plant et al. (2004) subset of data from Table 9, p. 48, Chap 9.02, Arsenic and selenium. In Environmental Geochemistry (ed. Lollar B.S.) Vol 9 Treatise on Geochemistry (eds. Holland H.D. and Turekian K.K., 2004). # Value from Maes et al., (2004a,b).

174/328

A2.6.1 Selenium in pyrite extracted from Boom Clay

Sulfide minerals are the most common sources of selenium in shale and organic-rich clay.

Pyrite (FeS2) being present up to 1 – 5 % wt. in Boom Clay, selenium has been specifically searched for in pyrite extracted from Boom Clay.

Natural selenium has been measured by Maes et al. (2004a, b) in pyrite extracted from Boom Clay by Delécaut (2004). Pyrite was separated from fresh Boom Clay samples disaggregated and dispersed in water. Suspensions of clay were wet sieved and four different fractions between 20 and 500 μm were collected along with some fragments > 500 μm. The heavy minerals of these fractions were extracted by sedimentation in heavy liquid (bromoform). The characterisation by x-ray diffraction (XRD) showed that the main constituents isolated were pyrite and quartz. Besides, minor amounts of carbonates and Ti-oxides were also present as well as clay minerals traces. The various fractions obtained were used without further treatment.

The selenium concentrations (mg kg-1) were measured in the different fractions of isolated pyrite by means of high resolution ICP-MS. The analysis results are presented in Table A2.6.1 as a function of the size fraction.

Table A2.6.1: Selenium content in different fractions of Boom Clay pyrite. Size Fraction Shape [Se] (μm) (—) (mg kg-1) 20-32 Framboïds 12.2 32-64 Framboïds 18.1 64-125 Framboïds 29.3 125-500 Aggregates 33.0 >500 Concretions 20.1 >500 Faecal pellets 17.0 Average value — 21.6 Standard Deviation — ± 7.9

The selenium content in pyrite extracted from Boom Clay is in the range 12 – 33 ppm and exhibits a maximum for the aggregates in the size fraction between 125 – 500 μm. No clear trend can be distinguished for the different morphologies of pyrite (framboïds, pellets, concretions). The average value of selenium in these Boom Clay pyrites is 21.6 ± 7.9 mg kg-1.

175/328

No chemical analyses are presently available for selenium concentrations in the clay mineral fraction cleared of pyrite.

A2.6.2 Selenium in Boom Clay water

Selenium was also measured in some water samples by means of high resolution ICP-MS at the Museum of Central Africa (Tervuren) (Maes et al., 2004a, b). A total selenium concentration of 1.9 μg dm-3 (1.9 ppb) was measured in water collected from the large extension gallery bottom shaft (EG-BS) piezometer. It corresponds to a total selenium concentration of 2.4 × 10-8 mol dm-3, a value in agreement with that of groundwaters listed in Table A2.5.2, and one order of magnitude higher than in sea water.

The selenium concentration found in the solid pyrite is about 11 400 times higher than in Boom Clay porewater. This ratio between solid:liquid concentrations of natural selenium is similar to the S:L ratio determined for uranium in Boom Clay, another element whose mobility strongly depends on redox conditions.

Maes (2004, pers. comm.) observed than after acidification of the water sample and subsequent flocculation and removal of part of the dissolved organic matter (DOM), the selenium concentration abruptly decreased in the analysed porewater. This suggests that a fraction of natural selenium is associated with DOM. One reason could be the presence of natural organo-selenium covalently bound in the chemical structure of organic matter. This is consistent with the biochemical cycle of selenium and its presence in many enzymatic sites in proteins and could be explained by the bioconcentration of selenium by coccolithophores such as Emiliania Huxleyi and the subsequent incorporation of organo-selenium in marine kerogen (Amouroux et al., 2001; Araie et al., 2003; Danbara and Shiraiwa, 1999;2007; Obata et al., 2003;2004). Recent studies also indicate that kerogen can contain high concentration of selenium (Hanjie et al., 2006; Carignan, 2008, Personal communication).

However, up to now, no specific chemical analyses are presently available for natural selenium concentrations in the solid and the dissolved organic matter of Boom Clay. The presence of high amounts of pyrite intimately associated with Boom Clay kerogen (cfr., works made at ENSCP and IFP respectively by Deniau et al., (2007) and Lorant et al., (2007)) make the occurrence of selenide and organo-selenium in Boom Clay kerogen highly probable, but this point still requires an experimental verification.

It is thus recommended to determine the natural selenium concentration, on the one hand in purified secondary mineral fractions of Boom Clay, especially pyrite and coccolithophore-

176/328

carbonate rich layers), and on the other hand in the Boom Clay organic matter (both sulfide- rich kerogen and dissolved OM).

More information on the existing correlation, or the possible association, between natural selenium and organic matter is given in Appendix A8 (Selenium and organic matter) where it is discussed in detail: − to know if the solubility limit of natural inorganic selenium in Boom Clay is reached, and; − to assess the possible interaction of long-lived selenium-79 with natural selenium associated to the natural organic matter.

177/328

178/328

A3. Selenium speciation in the source term

179/328

180/328

A3 Selenium speciation in the source term

Two types of waste forms are presently considered: the non-reprocessed spent fuel (SF) and the vitrified high-level waste (HLW) arising from the reprocessing of spent fuel at the Cogema plant of La Hague (France).

The exact speciation of selenium in the nuclear waste is unclear and the redox conditions in the different waste forms (spent fuel and HLW glass) are still debated. The difficulty to gather information on selenium in UO2 fuel is likely related to its low abundance in the spent fuel and also because it is not an important element for short-term nuclear accident scenario.

Almost all the chemical forms of inorganic selenium could in principle be accommodated in both waste matrices (SF and HLW glass), from the higher valences to the lowest oxidation 2– 2– 0 2– state, i.e., SeO4 , SeO3 , Se , and Se . Tetravalent uranium, U(IV) in UO2, the main component of light water reactor (LWR) fuel, could still impose reducing conditions in spent fuel, even at high burn-up, while 79Se is likely incorporated in the glass matrix under oxidizing conditions if no reducing agent is added to the calcination stream, or to the molten glass.

A3.1 Selenium in spent fuel

A commonly encountered idea is that in-pile fission reactions and alpha recoil create locally oxidising conditions in the spent fuel matrix, leading to oxidation of selenium and of a fraction of uranium. Inversely, a contradictory point of view argues that the two (or three) fission products issued from the fragmentation of one UO2 formula unit after fission will lack oxygen atoms and so would be implanted in a partially reduced form in the UO2 matrix. Although it seems consistent for one single fission reaction, or for 100 % burnup, this reasoning is no longer applicable to a small number of fission product atoms spread in the mass of the uranium oxide matrix. There exist a sufficient quantity of oxygen atoms available in the UO2 crystal lattice to recombine with the fission fragments. Moreover, considering a statistically significant number of fission reactions, a mass balance budget indicates that the number of oxygen atoms available per atom of fission product is not in deficit but at the contrary in net excess after fission occurs.

In fact, the fission products consume less oxygen than the quantity which is liberated by the fission of tetravalent U and Pu. Indeed, although two fission products are formed from each U (or Pu) atom, many of the fission products are noble metals (Mo, Tc, Ru, Rh, and Pd, with particularly high yields for fission of Pu), rare gases (Kr, Xe), and halogen elements (Cl, Br,

181/328

I), which do not combine with oxygen. In addition, many of fission and activation products are trivalent lanthanides and actinides. The dissolution of trivalent rare earth ions (and of Pu,

Am, and Cm) in tetravalent UO2 also requires 25 % less oxygen atoms per metal atom (La2O3

= LaO1.5).

The oxygen atoms left in excess after the UO2 fission reaction are released and causes thus the oxygen potential ΔG(O2) of the fuel to increase (in fact to become less negative) at constant oxygen-to-metal (O/M) ratio. Therefore, UO2 fuel was commonly assumed to oxidize with increasing burnup, in particular for Pu fission.

Changes in ΔG(O2) and in oxidation state of the fuel (its O/M ratio) affect many important properties and influence significantly the irradiation behaviour of the fuel; examples are clad internal corrosion, fuel thermal conductivity, creep and plasticity of the fuel. The knowledge of the development of ΔG(O2) and of the O/M ratio of the fuel as a function of burnup was therefore of large scientific and technological interest.

The extent of UO2 oxidation (i.e., the increase of the oxygen stoichiometric coefficient x in

UO2+x) was considered to be limited by several oxygen interception mechanisms, amongst others, the oxidation of molybdenum, an important FP produced at high yield, and by the reaction of oxygen with the inner wall of the zircaloy cladding of the fuel pin. The reaction

Mo + O2 = MoO2 was regarded as the main buffering mechanism of oxygen in the irradiated fuel matrix itself, while zirconium from the cladding was seen as an effective getter scavenging the remaining O2 inside the closed system of fuel rod. Nevertheless, oxidation of

UO2 was considered to increase at high burnup.

The question of oxygen potential (ΔGO2) in the fuel pin system is thus particularly important to assess the degree of oxidation of uranium and fission products in UO2 fuel and was the object of many studies. Matzke (1995) performed measurements of oxygen potential in high burnup light water reactor (LWR) UO2 fuel. Against all expectations none of the studied fuels showed a significant oxidation. These observations were in contrast to the commonly accepted ideas that UO2 fuel progressively oxidize due to burnup.

More recent works (Walker et al., 2005; Ferry et al., 2006) also show that uranium in UO2 fuel appear not to oxidize even at high burnup. The oxygen potential does not increase with burnup as previously thought. More surprisingly, according to Walker et al. (2005) molybdenum appears also not to buffer the system and Zr to only intercept 23 % of O2. The oxygen stoichiometry of the fuel does not evolved towards UO2+x (the oxidized form) but remains constant, or even exhibits a negative x in the stoichiometry (UO2-x). So, uranium remains tetravalent in the spent fuel. In these conditions, one could also in principle expect that selenium would be present in the spent fuel at the reduced state.

182/328

However, by analogy with the behaviour of tellurium with fuel burnup, it is not evident that selenium should only exist under a reduced form in spent fuel.

The reactivity and the chemical state of tellurium within the fuel are very complex although the oxygen potential of the Te/TeO2 equilibrium is a great deal higher than that of U-Pu oxides in fuel. Therefore, tellurium can form metallic phases with U, Pd and Sn, or can be constituent of multi-component fuel-fission product oxides depending on the local oxygen potential and can be dissolved in the oxide fuel. Tellurium has been found in the gap in association with other fission products forming stable oxides. Tellurium chemical behaviour is thus very versatile and both Te and BaTeO3 phases have been identified. Therefore, according to Kleykamp (1985), the multiplicity of the tellurium containing phases prevents a self-contained description of its chemical state in spent fuel. Quite similar conclusions can probably be drawn for selenium.

As a conclusion, it is not because uranium does not appear to oxidize in high burnup fuel that one can de facto conclude that selenium also does not oxidize. On the one hand, U4+ could be considered as a major redox buffer precluding selenium oxidation, while conversely on the other hand, oxidation of some more reactive FP, whose selenium, could help to protect U4+ if 4+ these FP and selenium have more affinity with O2 than U . Thermodynamic calculations are certainly needed to assess this question. However, no data are presently available for selenium

(Se/SeO2) in the thermodynamic database of the laboratory of high and medium activity (LHMA) of SCK•CEN and selenium does not figure on the principal Ellingham diagrams. Indeed, 79Se is a trace element in spent fuel and is not important in case of failed fuel pins or for short-term consequences of major nuclear accidents. So, it remains to gather the relevant thermodynamic data for selenium in spent fuel to perform the necessary calculations to first theoretically address the question of selenium speciation in spent fuel before to setup a specific experimental program in collaboration with LHMA.

A3.2 Selenium in vitrified high-level waste

During the reprocessing of spent fuel selenium is exposed to extremely oxidizing conditions. First, during the fuel dissolution step when it enters in contact with boiling concentrated nitric acid during a prolonged contact time. Then, during the evaporation/drying/calcination process of the liquid HLW prior to vitrification. If elemental selenium resists to oxidation, it will likely be rapidly volatilized and trapped in the scrubbers of the gaseous effluents of the reprocessing plant which are diluted and discharged in sea.

183/328

The fraction of less volatile selenium subsisting under oxidized forms would easily be 2– 2– incorporated in the molten silica glass. Indeed, SeO4 and SeO3 can substitute tetrahedral silicate or trigonal borate entities respectively in the glass network.

However, reduced forms of selenium are also soluble in glass as attested by their wide industrial use in glass decolourising process. Indeed, Se and Se2– are added to glass because their pink colour counterbalances the green colour of Cr3+ and Fe2+ impurities to obtain flint grey glass. It appears that in nuclear waste vitrification plants sugar is sometimes added as a source of carbon to the HLW stream in the calcination process to reduce ruthenium as metal in order to limit the volatile RuO2 oxide. Molten saccharose and carbon monoxide resulting from the sugar partial combustion could also reduce selenite and selenate at high temperature and then the incorporation of a fraction of the volatile Se and Se2– species in the glass cannot be excluded.

A3.3 Selenium in bituminised MLW

In the case of nitrate-bearing Eurobitum waste (MLW), the presence of 79Se under the selenate form can certainly not be ruled out because of the massive amounts of nitrate salts (up to 25 – 30 wt. %) present in this type of waste and the oxidizing conditions of the production process. Moreover, in the presence of nitrate, selenate reduction is drastically hindered and likely improbable (Wright, 1999; Oremland et al., 1999; Oremland et al., 2006). An interesting analogy with the behaviour of sulfate in petroleum geological reservoir could also be useful: indeed, to avoid sour oil problems related to microbially-mediated sulfato reduction, nitrate is often added to oil drilling fluids, or injected in oil fields in large amount, to suppress sulfate reduction. On the basis of reaction free energy, a common rule used to identify the redox reaction first involved in microbial respiration, is that the more powerful oxidant must be first consumed by the reductant before a weaker oxidant, also available in the – 2– 2– system, can be used: (O2 > NO3 > Mn(IV) > Fe(III) > SO4 > SeO4 > CO2).

A3.4 Dissolution controlled by alpha radiolysis

Finally, if selenium is released from nuclear spent fuel or HLW glass by a corrosion process controlled by alpha-radiolysis imposing locally oxidizing conditions, selenium could also dissolve in porewater as selenate. If reduction of selenate is kinetically not possible under Boom Clay conditions, selenium is expected to migrate as an unretarded (or a very weakly retarded) species (1 ≤ R ≤ 3) and its concentration would not be solubility limited. In such conditions, the amount of selenium present in the waste inventory and the corrosion rate could be the limiting factors for the dose-to-the-man.

184/328

A4. Selenium speciation behaviour in Boom Clay

185/328

186/328

A4 Selenium speciation behaviour in Boom Clay

A4.1 Overview Various batch experiments were independently performed by AEA Technology and KULeuven to determine the solubility and the extent of sorption of the different aqueous inorganic species of selenium, ranked hereafter by decreasing order of valence: selenate 2– 2– – (SeO4 ), selenite (SeO3 ), elemental selenium (Se(s)), and selenide (HSe ). In Appendix A4, the results of these experiments obtained by both laboratories are presented for each afore mentioned species. After a general survey of the methods used for the selenium compounds characterisation, the selenium-75 source preparation, and the speciation techniques, the behaviour of each species is studied by following its interaction with individual Boom Clay components (pyrite, FeCO3, organic matter, …) and with Boom Clay suspensions in porewater and synthetic solutions. The results of the solubility and sorption experiments are presented along with the kinetics of reduction/precipitation when possible (case of selenite). The possible association between selenium and Boom Clay organic matter is also studied by means of specifically developed methods. The overall picture of the behaviour of selenium in Boom Clay makes possible to compare the different and complementary approaches and to draw conclusions to support the results of migration experiments and to identify the processes and parameters relevant for performance assessment (PA) studies. Finally, the remaining uncertainties related to the main processes involved in the retention and the transport of selenium in Boom Clay are discussed.

A4.2 Very slow reduction kinetics and derived uncertainties affecting the solubility value As discussed in Chapter 2 (Thermodynamic calculations), selenium is a redox-sensitive element whose solubility essentially depends on Eh value. Therefore, the kinetics of reduction of selenate and selenite were studied by contacting these species with undisturbed Boom Clay and chemical reductants. Useful observations were made from batch experiments performed with clay suspensions by KULeuven and started with Se(+VI) and Se(+IV). They can be summarized as follows hereafter.

2– 1. The reduction of SeO4 is extremely difficult without catalyst.

2– The kinetic of reduction of SeO4 in Boom Clay and the different parameters involved 2– were not separately investigated. However, it seems that the reduction of SeO4 does not significantly occur at room temperature without catalyst (e.g., green rust in canister corrosion products, Cui et al., 2006) or is extremely slow: it might depend on the enzymatic activity of micro-organisms such as sulfato-reducing bacteria (SRB) possibly present in the clay suspensions.

187/328

2– 2. The reduction of SeO3 occurs in Boom Clay but slowly.

2– 2– When SeO3 is contacted with pyrite, a continuous decrease with time of SeO3 concentrations is observed. Chemical reduction is the main envisaged mechanism, but simultaneous sorption also plays a role. The reaction rate seems to be proportional to the concentration of Se(+IV) in solution and to the quantity of pyrite engaged in the system.

The reaction rate is also inversely proportional to the square root of the pyrite (FeS2) 2– occupancy by selenium. It suggests that SeO3 reduction occurs through sorption onto

FeS2. 3. Elemental selenium, Se(s) can be reduced and redissolved as HSe–

Se(0) can be reduced as Se(-II) with an higher concentration when the Eh value is sufficiently low in the presence of solid Boom Clay, or iron strip, as previously mentioned by AEAT in his thermodynamical calculations (see Figure 2.5.2, p. 33, Chapter 2). The amount of solid phase in the suspensions seems to be a critical parameter for the reaction kinetics in the system. In all batch tests, 75Se concentrations in solution tended to decrease, but the mixtures with the lowest initial Se concentration and the highest solid-to-liquid ratio reached a plateau concentration most rapidly. As conclusion of the overall observations, only the reduction of selenite was proven to be significant and contributed to remove this species from solution. However this process is slow and it seems to be more advantageous to directly start from elemental selenium, Se(s), and iron selenide, FeSe(s), to study the solubility of the reduced forms of selenium. Finally, a critical point related to the easy oxidation of these selenium compounds in all these studies is certainly the chemical purity of the reagents used, as supplied by a commercial manufacturer, or synthesised on purpose in the laboratory. The great sensitivity of the reduced 2– forms of selenium to oxidation should always be kept in mind. Selenite (SeO3 ) is 2– 75 progressively oxidised in selenate (SeO4 ) in the mother solutions of Se because of the oxidising species continuously produced by water radiolysis. Iron selenide (FeSe) is also often contaminated by elemental selenium, Se(0), which can alter its apparent stoichiometry 2– (Fe7Se8). Moreover, if iron selenide is partly oxidized, small amounts of SeO4 may be present at concentration far above the solubility limit of FeSe, compromising the results of delicate solubility experiments. Therefore, the chemical purity of all reagents should be carefully checked with the appropriate techniques (SEM/EDX, ion chromatography, …) prior 2– to use, and soluble oxidized species (such as SeO4 ) first removed by a sufficient washing of the poorly soluble solids (FeSe, Se0) before starting experiments for long-term equilibration periods.

188/328

2– A4.3 Behaviour of selenate: SeO4

Sodium selenate is a very soluble inorganic compound of selenium and no solubility limit is 2– expected for the SeO4 species in undisturbed Boom Clay porewater in the absence of heavy metals, or large concentrations of earth-alkaline cation such as Ba2+.

In the experimental conditions and at the time scale of the experiments we have up to now performed, we have not observed significant sorption, nor reduction/precipitation for selenate present in synthetic Boom Clay porewater (i.e., without prior addition of organic matter in water) contacted with pyrite and Boom Clay suspensions.

Therefore, selenate is expected to migrate in Boom Clay as a conservative tracer, i.e., without any retention. The worst case scenario to be considered for performance assessment (PA) studies is that all the 79Se inventory would be present in the source term as selenate and therefore would migrate without solubility limit and non-retarded in Boom Clay, as it is the case for another major contributor to the dose: 129I.

2– A4.3.1 Interaction of SeO4 with pyrite Synthetic Boom Clay water solutions (absence of dissolved organic matter) containing high (10-3 mol dm-3) and low (10-6 mol dm-3) concentrations of selenate were contacted with fresh crushed pyrite. No decrease of the selenate concentrations was observed at the time scale of the experiments for the total selenium, nor for 75Se spike.

2– 2– So, two possibilities may be considered: (i) SeO4 is not sorbed onto pyrite, or (ii) the SeO4 reduction/precipitation is extremely slow and could not be observed on the duration of experiments (up to 13 months). The effect of dissolved organic matter (DOM) on the kinetics of reduction of selenate by pyrite surface was not investigated in these experiments.

2– A4.3.1.1 AEAT: Interaction of SeO4 with pyrite (high concentration) The removal of selenate from solution by crushed pyrite and pyrite coupons was measured from a synthetic Boom Clay water in the absence of organic material. An initial selenate concentration of 1 × 10-3 mol dm-3 was used.

In all the experiments the total aqueous selenium concentration remained unchanged over a period of 13 months, showing that there was no obvious reduction/surface-precipitation of selenate at, or sorption to, the pyrite surface over this timescale in these conditions.

XRD analysis of pyrite powder from the experiments showed no bulk changes to the mineralogy and no secondary selenium phases were found. SEM examination of a pyrite

189/328

coupon showed a region of oxidation associated with a crack in the pyrite surface but no selenium was detected in this area, or on any other area away from the crack.

2– A4.3.1.2 KULeuven: Interaction of SeO4 with pyrite (low concentration) 2– The reduction kinetics of SeO4 in the presence of crushed pyrite (FeS2) were monitored using synthetic Boom Clay water (absence of organics) as background electrolyte. The 75 2– experiments were carried out in an oxygen-depleted glovebox and radioactive SeO4 (2.6 × 10-7 mol dm-3 and 6.0 × 10-8 mol dm-3 – present as minute fractions in the 75Se spike) were added to batch systems containing different FeS2 solution–to–pyrite ratio of 200:1 and 100:1. The selenium speciation and solution concentrations were followed as a function of time.

2– 75 2– No significant decrease in the stable SeO4 concentrations, nor in the active SeO4 concentrations, as a function of time was observed. Both species remained in the (+VI) redox state and in their oxy-anionic form. Concentrations in solution remained roughly equal to the starting concentration, although there were some indications that a slight decrease might be 75 2– occurring for the lowest SeO4 concentrations. Therefore we concluded that: 2– − at pH relevant for Boom Clay (pH > 8), SeO4 is not significantly sorbed onto pyrite

(FeS2), nor onto Fe hydroxides (if present at the surface), as it is also reported in literature by Goldberg and Glaubig (1988) and Rietra et al. (2001), and that; 2– − SeO4 abiotic reduction/precipitation by pyrite surface is very slow and has not been observed at the time scale of our experiments.

2– The possible interaction of SeO4 with dissolved Boom Clay organic matter was not specifically verified in independent tests, but batch interaction experiments carried out with Boom Clay suspensions suggest it is not detectable or insignificant in Boom Clay conditions.

2– A4.3.2 Interaction of SeO4 with Boom Clay – KULeuven The possible adsorption and reduction/precipitation of Se(+VI) in the presence of Boom Clay used as solid phase and Boom Clay organic matter was investigated. Batch experiments with selenate were performed on Boom Clay suspensions at a liquid-to-solid ratio of 4.18:1. Boom 2– -6 -3 -7 -3 Clay batches with different stable SeO4 concentrations (10 mol dm , 5 × 10 mol dm , and 10-7 mol dm-3) were measured as a function of time in an oxygen-depleted glovebox. Since only inactive selenate was used in this set-up, the qualitative and quantitative 2– determinations of SeO4 was only done by ion chromatography (IC).

190/328

2– No really convincing proof of reduction of SeO4 was observed in Boom Clay suspensions. 2– Indeed, after long equilibration times the systems were not yet in equilibrium and SeO4 species were still detected in solution.

2– The reduction of SeO4 at the clay minerals surface seems very difficult and kinetically hindered. Indeed, by comparing the adsorption of selenate and sulfate on the surface of goethite (Rietra, 2001), it is likely that at pH above 8, selenate only forms very weak outer-sphere complexes with the sorption sites available in the clay matrix (silanol, aluminol, or ferrol, groups only accessible on the lateral edges of the clay platelets). It would thereby be subjected to a quite extensive competition with the other anions present in the system and 2– would thus not significantly sorb (particularly compared to SeO3 ).

2– The association of SeO4 with dissolved Boom Clay organic matter was not explicitly investigated (e.g. by gel permeation chromatography, GPC).

2– A4.3.3 Kinetics of reduction of SeO4 – KULeuven As no convincing evidence of selenate reduction could be obtained in the present works, the 2– kinetics of reduction of SeO4 in Boom Clay and the different related parameters were not separately investigated.

We do not presently dispose on specific information on the reduction kinetics of selenate in Boom Clay. We ignore if this reduction is possible without the assistance of a catalyst: e.g., the surface of an active solid phase, such as green rust, a corrosion products of metallic iron, as suggested by the works of Cui et al. (2006), or the enzymatic activity of some biochemical processes.

2– However, we cannot rule out that the reduction of SeO4 might proceed extremely slowly, and could depend on the activity of micro-organisms developing with time in the clay suspensions, such as Sulfato-Reducing Bacteria (SRB).

2– The kinetics of reduction of SeO4 in Boom Clay is one of the major uncertainties that still deserves further experimental investigation.

2– A4.3.4 Solubility of SeO4 – KULeuven 2– The limit of solubility is far from being reached for the SeO4 concentrations range used in the Boom Clay experimental observation window. Indeed, according to Weast (1968), the -3 -3 Na2SeO4 solubility is about 84 g dm (0.445 mol dm ) in pure water.

191/328

2– A4.3.5 Conclusions for SeO4

2– After addition of high and low concentrations of SeO4 to suspensions of pure pyrite or fresh Boom Clay, no significant sorption could be evidenced over periods of a few months. This is consistent with the very weak interactions expected at pH > 8 for the outer-sphere complexes between selenate and the surface of pyrite and clay minerals.

However, it is difficult to draw definitive conclusions for a possible very slow and progressive reduction of selenate followed by the precipitation of a poorly soluble phase such as, e.g., elemental Se(0) or FeSe. The results of interaction tests between selenate and pyrite and Boom Clay suspensions can be interpreted in two opposite ways:

1. on the one hand, when selenate is contacted with FeS2, no rapid sorption is observed at 2– low or high SeO4 concentration, and we cannot obtain a clear indication of a slow reduction/precipitation at the time scale of these experiments; 2. on the other hand, when selenate is let to interact with Boom Clay suspensions, no sorption is a priori expected onto clay minerals at pH > 8, but on the long-term, some progressive removal of selenate from solution could be interpreted as a tiny indication for a very slow reduction/precipitation. However, this reduction could be microbially- mediated if micro-organisms develop in the clay suspensions with time.

Because selenate sorption is likely very limited (if existent) in the pH range 8 – 10 in Boom 2– Clay, we therefore conclude that SeO4 would diffuse almost unretarded through the clay layer.

The abiotic selenate reduction/precipitation is likely kinetically hindered. Selenate might subsist for an undetermined period of time under compact in situ Boom Clay conditions, if no microbial activity can develop to facilitate its reduction.

The association of selenate with Boom Clay organic matter has not yet been specifically 2– studied but is presently considered as relatively unlikely. Indeed, the interaction of SeO4 with OM is expected to be low because selenate reduction is kinetically hindered and that selenate does not form inner-sphere complexes to establish chemical bonds with OM (such as, e.g., iron bridges). Finally, the non-specific electrostatic repulsion forces between negatively 2– charged entities (SeO4 and OM polyelectrolyte) is also a priori not quite favourable to their association.

In conclusion, because of the remaining large uncertainty on the kinetics of reduction of selenate onto pyrite and Boom Clay, we cannot presently proof a significant retention (by reduction/precipitation mechanisms) for selenate in undisturbed and compact Boom Clay.

192/328

Thus, we cannot presently consider the reduction/precipitation of selenate as an established mechanism that could be taken into account in safety studies for the removal of selenate from Boom Clay porewater. This point is also further discussed in Section A9.3.2 dealing with the 2– results and interpretation of electromigration experiments with oxidized Se sources (SeO4 2– and SeO3 ).

2– So, the kinetics of reduction of SeO4 in Boom Clay is one of the major uncertainties that still deserves further experimental investigation. This is important if a large fraction of the 79Se inventory present in the source term would occur under the form of free selenate because of the radiolysis effects affecting the spent fuels and the vitrified HLW.

2– A4.4 Behaviour of selenite: SeO3

2– 75 Selenite (SeO3 ) in its stable form, or as tracer labelled with Se, was certainly the selenium species studied into the most details in the frame of the present work. Because its great sensitivity to experimental redox conditions, selenite has exhibited an intricate behaviour with results often contradictory and difficult to decipher.

2– The SeO3 concentration introduced as sodium salt in Boom Clay interstitial water is not solubility limited. The poorly soluble calcium selenite (CaSeO3), which could be formed in the near-field environment in the presence of cement, is not expected in situ in Boom Clay under undisturbed conditions.

2– However, when SeO3 is added to systems containing fresh Boom Clay suspensions, or 2– chemical reductants, SeO3 can be reduced (e.g. by adsorption onto FeS2) and poorly soluble 2– selenium (0) or (-II) phases precipitate. After complete disappearance of SeO3 , the selenium solution concentration is controlled by the solubility of the resulting selenium solid phases: elemental Se(s), iron selenides such as FeSe, or FeSe2, or solid solutions with iron sulfides such as [FeS2-×(Se)×].

Selenite is an oxy-anion known to be able to chemically sorb onto iron(III) oxy-hydroxides by a mechanism of ligand exchange. A strong inner-sphere complex may easily be formed between the oxygen atoms of selenite and an oxide surface bearing exchangeable hydroxyl groups. A bidentate bond would be stronger than mono-dentate link. In case of even partial oxidation of pyrite surface, the newly formed hydrous ferric oxides would be particularly favourable sorption sites for selenite. However, under undisturbed in situ condition, iron(III) oxides are not expected to be present in Boom Clay, and up to now have not yet been detected as far as we know. Under reducing conditions in marine sediments, Fe(III) is converted in Fe(II), and the most common thermodynamically stable phases of iron(II) identified in Boom

193/328

Clay are pyrite and siderite. Nevertheless, hydroxyl groups are present on the edges of clay mineral platelets, a.o, as aluminol groups on the lateral edges of the gibbsite layers. So, sorption sites might be available in undisturbed clay in octahedral aluminium side positions to 2– form inner-sphere complex with SeO3 .

Two main mechanisms could be at work to remove selenite from Boom Clay porewater: on 2– one hand, sorption of SeO3 onto pyrite surface or broken edges of clay minerals, and on the other hand, a slow chemical reduction of Se(IV) at pyrite interface, followed by the precipitation of elemental selenium, Se(0). Kinetic results are available for the reduction of 2– SeO3 at the surface of pyrite. The reduction rate appears to be directly dependent on the amount of solid FeS2 present in the system and inversely proportional to the concentration of selenite in water. The system with the highest pyrite-to-solution ratio and the lowest selenite concentration evolve the fastest. The reduction of selenite at the surface of pyrite seems to be retarded if clay minerals are also present in the system. This could be tentatively explained by 2– the competitive sorption of SeO3 between antagonist sites present on pyrite surface and clay platelet edges. The desorption of selenite from clay minerals has first to occur before that selenite can migrate to pyrite to accept electrons from its surface.

Sorption of selenite seems to occur first at the time scale of one month, followed by a slow and progressive move from sorption to a reduction/precipitation mechanism after more than one year. Some experiments indicate only linear sorption isotherms as observed by AEAT with pure pyrite, while others tend to give arguments in favour of a final solubility control (KULeuven experiments with clay suspensions).

However, the interpretation of the first series of experiments made with 75Se have been often 75 2– obscured by the initially unexpected contamination of selenate ( SeO4 ) produced by oxidising free-radicals in the concentrated spiking solutions because of intense water radiolysis.

2– Finally, it is worth to note that SeO3 is the only species of selenium for which an association with Boom Clay dissolved organic matter (OM) has been clearly established by gel permeation chromatography (GPC) and removal of OM from solution by La3+ precipitation. 2– Such an association could not be evidenced for the other selenium species studied, i.e., SeO4 , Se(0) and HSe– (from FeSe), after a prolonged contact time with Boom Clay porewater containing dissolved organic matter (DOM).

2– The exact interaction mechanism of SeO3 with OM remains unclear, but two routes could be 2– first envisaged. On one hand, SeO3 could form inner-sphere complexes with Fe(III) groups 2– present in the complex structure of humic substances: e.g., OM—Fe—SeO3 . It is the hypothesis of the so-dubbed “iron bridges”, a concept also introduced to explain the sorption

194/328

of humic acids onto the surface of iron oxides and clay minerals edges. On the other hand, 2– SeO3 could be reduced as colloidal elemental Se(0) which in his turn could interact with organic colloids. This is the assumption of weak hydrophobic interactions between colloids of different nature. However, presently we have no experimental evidence, nor even a clue, to identify the exact mechanism. This association Se–OM is also likely very different from that implied in the chemical binding of natural selenium in organic matter of shale and mudrocks (see Appendix A2 on Natural selenium in the environment and in Boom Clay, and Appendix A8 on Selenium and organic matter). There, complicated biochemical pathways are probably involved in the incorporation of organo-selenium molecules in the degradation products of complex enzymatic structures where selenium occupies the position of a sulfur atom.

2– A4.4.1 Interaction of SeO3 with Boom Clay components (pyrite, OM)

Batch sorption tests with selenite in contact with main Boom Clay components (pyrite and organic matter, OM) have been performed independently by AEA Technology and KULeuven on a range of selenite concentrations spreading on about 5 orders of magnitude (10-8 – 10-3 mol dm-3).

2– -4 AEA Technology worked with inactive SeO3 at relatively high concentration (3.4 × 10 up -3 -3 75 2– to 1.3 × 10 mol dm ) while KULeuven used labelled SeO3 at much lower concentration (2 × 10-8 up to 5 × 10-6 mol dm-3).

A common and general trend comes clearly out of all the experimental results: the extent of selenite removal from solution is directly dependent on the mass of pyrite available (and hence, indirectly on the number of available sites on the pyrite surface) and is 2– inversely related to the initial amount of selenite: RD ~ pyrite mass / [SeO3 ]. The distribution ratio RD values calculated from the loss of selenium in solution show a general increasing trend with decreasing solution:pyrite ratio in synthetic Boom Clay water.

For the sorption experiments performed by AEA Technology onto pyrite with an initial selenite concentration of 1 × 10-3 mol dm-3 in synthetic Boom Clay water in the absence of organic matter, a simple surface complexation model provides a reasonable fit to the experimental data, reproducing the observed decrease in the concentration of sorbed material at lower aqueous selenium concentrations, and at higher water:pyrite ratios. This is in contrast to a solubility-controlled mechanism for selenium removal. If solubility control was to operate, the final selenium concentration would be expected to be independent of the initial concentration, and of the water:pyrite ratio. In this case the experimental data would be expected to lie on a vertical line in plots. Clearly the experimental data of AEA Technology

195/328

do not show such a trend and therefore solubility control is considered unlikely in these experiments. From the surface analysis data it is likely that sorption of selenite to pyrite occurs in oxidised areas.

A simple linear distribution (KD) mechanism was also observed from the sorption experiments performed by KULeuven with an initial selenite concentration of 1 × 10-6 mol dm-3 in synthetic Boom Clay water in the absence of organic matter (OM). A good consistency is observed between the results of sorption experiments at low and high concentration of selenite in synthetic water in the absence of OM.

The effect of OM on the sorption of selenite onto pyrite is less clear. Indeed, different effects are observed according to the origin and the preparation procedure of the extracted OM. On one hand, dissolved organic matter concentrated under oxidizing conditions seems to slow down the removal of selenite from solution in contrast to the absence of OM. On the other hand, Boom Clay organic matter extracted from the solid clay under anoxic conditions has no effect on the removal rate compared to the sorption tests made in the absence of OM. After 90 days, up to 25 % of the total selenium concentration in solution was observed to be associated with organic matter.

However, some of the long-term results of KULeuven obtained in the presence of organic matter (OM) seem to conflict with these of AEA Technology experiments made without OM.

Indeed, the mechanism of selenium removal from solution seems to apparently evolve as a function of time from a simple sorption behaviour to a solubility control as suggested by the overview plots shown by KULeuven. After a few days contact time, the selenium removal seems to follow a simple linear distribution KD mechanism. Then, the slope of the isotherm gradually increases as a function of time until the selenium concentrations converge towards two different values depending on the initial concentration and represented by vertical lines in the KULeuven plots. It seems to suggest that a solubility control could be at work on the long 75 2– term after ageing of the pyrite suspensions in contact with SeO3 . However, the two values of the final selenium concentrations (2.4 × 10-8 mol dm-3 and 1.35 × 10-7 mol dm-3 respectively) still depend on the initial concentrations and are about 40 times lower than these initial concentrations (1 × 10-6 mol dm-3 and 5 × 10-6 mol dm-3). This point is precisely in disagreement with a solubility control and represents a strong argument in favour of a sorption mechanism.

2– A4.4.1.1 AEAT: Interaction of SeO3 with Boom Clay components (pyrite) The removal of selenite from solution by crushed pyrite and pyrite coupons was first measured from a synthetic Boom Clay water in the absence of organic material. An initial

196/328

selenite concentration of 1 × 10-3 mol dm-3 was used at a solution–to–pyrite ratio of 50:1, in a similar manner to the experiments with selenate described above in Subsection A4.3.1.1.

In the presence of crushed pyrite at pH 10 and pH 8 there was evidence for the removal of up to approximately 10 % of the selenite from solution. At pH 10, concentrations of between 8.9 × 10-4 and 9.3 × 10-4 mol dm-3 were measured after 377 days’ equilibration compared to an initial concentration of 1.0 × 10-3 mol dm-3. At pH 8, the starting concentration of selenite was 9.3 × 10-4 mol dm-3 and this fell to between 8.4 × 10-4 and 8.5 × 10-4 mol dm-3 after 380 days. Although all the concentrations were within the ± 10 % uncertainty of the analytical measurements, the consistency of the decrease in concentration across all the experiments suggested it was genuine. The trend of the results with time tended to support the postulation that there is an initial fast removal process (probably sorption) within 100 days, followed by the achievement of a ‘steady-state’ aqueous selenium concentration. 3 -1 The results from the experiments gave distribution ratios (RD values) of 4 to 6 dm kg , if it was assumed that sorption was the only process to remove selenite from solution. No appreciable loss of selenium from solution could be established reliably in the experiments with pyrite coupons with the total selenium concentrations after 406 to 469 days being 9 × 10-4 to 1 × 10-3 mol dm-3. This would also be consistent with removal of selenite through sorption, as the available surface area is far less in the coupon experiments than in those with crushed pyrite.

Further experiments with crushed pyrite at solution-to-pyrite ratios of 50:1 and 5:1 and initial selenite concentrations of 1 × 10-3 to 3 × 10-4 mol dm-3 confirmed the removal of selenite from solution by pyrite in both synthetic Boom Clay water and in interstitial Boom Clay water. The results showed that there was an initial fast removal of aqueous selenite from solution within 1 month in both the synthetic Boom Clay water (no OM) and the interstitial Boom Clay water (with OM). Within this time, there also appeared to be dissolution of a surface-oxidised sulfur layer from the pyrite. The extent of selenium removal from solution was dependent directly on the mass of pyrite available (and hence, indirectly on the number of available sites on the pyrite surface) and the initial amount of selenite. RD values calculated from the loss of selenium for solution show a general increasing trend with decreasing solution:pyrite ratio in synthetic Boom Clay water (Table A4.4.1, AEAT).

197/328

2– Table A4.4.1 (AEAT): RD values for selenite (SeO3 ) sorption onto crushed pyrite in synthetic Boom Clay water in the absence of organic matter.

Initial Selenite Solution:Pyrite Equilibration RD Time Concentration -3 (days) 3 -1 (mol dm ) ratio (dm kg ) Synthetic Boom Clay Water (no DOM)

1.0 × 10-3 50:1 377 4 ± 11 6 ± 11 6 ± 11

9.3 × 10-4 50:1 380 5 ± 11 5 ± 11 5 ± 11

1.3 × 10-3 5:1 31 26 ± 6 93 21 ± 5 154 94 ± 20

6.5 × 10-4 5:1 31 180 ± 40 93 4 100 ± 820 154 1 100 ± 220

6.5 × 10-4 50:1 30 16 ± 13 92 22 ± 15 154 27 ± 16

3.4 × 10-4 5:1 31 4 900 ± 1 000 93 ≥ 3 000 154 570 ± 110

3.4 × 10-4 50:1 30 21 ± 14 92 27 ± 16 154 31 ± 16

Interstitial Boom Clay Water (with DOM)

4.9 × 10-4 5:1 29 470 ± 100 91 58 ± 13 145 14 ± 4

4.9 × 10-4 50:1 29 86 ± 27 91 41 ± 18 145 110 ± 30

RD: Distribution ratio. DOM: Dissolved Organic Matter.

198/328

XRD analysis of pyrite powder from the experiments showed no bulk changes to the mineralogy and no secondary selenium phases were found. Coupons recovered after equilibration with selenite solution for 13 months and for 23 months showed regions of oxidation associated with cracks in the surfaces. These oxidised areas contained significant concentrations of selenium as shown in Figure A4.4.1 (AEAT) for the coupon examined after 13 months. In contrast, the pyrite surfaces away from the cracks were found to have no associated selenium. For the coupon examined after 23 months, Figure A4.4.2 (AEAT) and Table A4.4.2 (AEAT) show that, at a short distance from a crack (spectrum 4), the surface was predominantly pyrite with a small amount of oxidation evident. As a grey peripheral alteration zone (spectrum 3) is traversed towards a dark central alteration zone adjacent to a crack (spectrum 2), O, Na, Si, Ca, and Se all increase significantly, while Fe and S simultaneously decrease.

Table A4.4.2 (AEAT): Elemental weight % analysis of spectra 2 to 4, surface of pyrite coupon exposed to selenite solution for 23 months. Spectrum O Na Al Si S Ca Fe Se Total

2 23.18 0.42 0.45 0.32 34.90 0.33 39.47 0.93 100

3 11.90 0.10 0.43 0.14 45.60 0.05 41.66 0.13 100

4 3.79 0.00 0.17 0.00 51.66 0.05 44.29 0.05 100

All results expressed as weight %. All elements analysed (normalised).

199/328

Pycoup Se4

Figure 13.4.1 (AEAT): scanning electron microscope (SEM) image of a pyrite coupon examined after 13 months equilibration with selenite solution.

Figure A4.4.1 (AEAT): Oxidised part of surface of pyrite coupon from experiment with selenite. EDAX spectrum from area of particle on surface. 200/328

08/10/2003 09:42:48 Py Coup Se4A - 4

Spectrum

Spectrum 2

Spectrum 3

Figure A4.4.2 (AEAT): Surface of pyrite coupon in selenite experiment. EDAX spectrum from area of crack (spectrum 2).

201/328

A sorption model was developed to test whether the experimental observations were consistent with a sorption mechanism for selenite. Given the lack of data characterising the pyrite surface and the narrow pH range of the experimental data, it was decided that a simple sorption model would be most appropriate. This model was based on the surface complexation approach, with sorption represented by the surface complexation reaction and equilibrium constant (KS):

2– 2– ≡Site + SeO3 <==> ≡Site-SeO3 (eq. A4.4.1) where ≡Site represents a sorption site on the pyrite or oxidised pyrite surface.

Neglecting activity corrections:

Csor Ks = CaqC(site) (eq. A4.4.2)

where: Csor is the concentration of selenium associated with the solid;

Caq is the concentration of selenium in solution, and;

C(site) is the concentration of uncomplexed sorption sites.

In such a system with a single type of sorption site, the sorbing species can be sorbed up to a maximum concentration corresponding to the total concentration of sorption sites, Ctot(site).

Substituting for C(site) (equal to Ctot(site) - Csor) and rearranging gives:

Caq 1 Caq = + (eq. A4.4.3) Csor Ctot (site)KS Ctot (site)

The concentration of sorption sites and the equilibrium constant can then be estimated from a plot of Caq/Csor versus Caq. From the experimental data obtained after three months for the two water:pyrite ratios, the fit parameters Ctot(site) and Ks, (together with the sorption site density determined from the fits) given in Table A4.4.3, were obtained.

Table A4.4.3 (AEAT): Sorption model parameters derived from simple surface complexation model fits to the three-month experimental data for selenite removal by pyrite (AEAT).

Experimental data sets fitted Concentration of sorption sites Log10Ks Site density (—) (mol dm-3) (—) (mol g-1)

5 cm3 g-1 water:pyrite data 1.1 × 10-3 6.1 5.3 × 10-6

50 cm3 g-1 water:pyrite data 5.5 × 10-4 3.1 2.8 × 10-5

202/328

The site density values determined in each case should, in principle, be equal. However, a lower site density value was determined for the lower water:pyrite ratio (5 cm3 g-1) data. This might be explained by a decrease in the accessibility of sorption sites at lower water:pyrite ratio data or, more likely, due to scatter in the experimental data.

Figure A4.4.3 (AEAT) shows that a reasonable overall fit to the data obtained after three -6 -1 months’ equilibration can be made using a site density of 7.0 × 10 mol g and a log10Ks value of 5.9. Although the experimental data at a water:pyrite ratio of 50:1 indicate a dependence of Csor on Caq, the model suggests that these values lie in the plateau region, in which sorbing species occupy all the available sorption sites, and that the apparent dependence of Csor on Caq is an artefact due to experimental scatter. Figure A4.4.3 also shows the data from experiments after a five months’ equilibration. Whilst these data show reasonable agreement with the overall fits determined for the three-month equilibration data, two of the experiments performed at a water:pyrite ratio of 5:1 show a significant increase in the final aqueous selenium concentration at longer times (from approximately 6 × 10-6 mol dm-3 to around 3 × 10-5 mol dm-3). This may be due to scatter or may indicate some release back into solution at longer times. Such behaviour would be difficult to explain although any oxidation of selenite to selenate would be expected to reduce the extent of sorption. An increase in the measured redox potential was observed in each case between the three-month and five-month measurements but it is not known whether this would have resulted in oxidation of selenite.

1.0E-02

-3 -3 prediction for Ci = 1.3 x 10 mol dm

1.0E-03 Expt. 5:1 data (3 months)

-3 -3 -3 expt at Ci = 1.3 x 10 mol dm Expt. 50:1 data (3 months) 5:1 overall model fit 50:1 overall model fit / mol dm

sor Expt. 5:1 data (5 months) C Expt. 50:1 data (5 months) 1.0E-04

Overall fit: site density = 7.0 x 10-6 mol g-1

log10 K = 5.90

1.0E-05 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03

-3 Cf/ mol dm

Figure A4.4.3 (AEAT): Overall surface complexation model fit to experimental three-month data for selenite removal by pyrite in synthetic Boom Clay Water (SBCW).

203/328

The experimental data for an initial aqueous selenium concentration of 1.3 × 10-3 mol dm-3 and at a 5:1 water:pyrite ratio are highlighted in Figure A4.4.3. Although these data lie close to the model curve, the calculated final aqueous concentration of selenite (also shown on Figure A4.4.3) is significantly lower than the experimentally-determined values. The decrease in the final concentration between three-month and five-month determinations is towards the model value so this may reflect a slow approach to equilibrium.

Despite these discrepancies the simple surface complexation model provides a reasonable fit to the experimental data, reproducing the observed decrease in the concentration of sorbed material at lower aqueous selenium concentrations, and at higher water:pyrite ratios. This is in contrast to a solubility-controlled mechanism for selenium removal. If solubility control was to operate, the final selenium concentration would be expected to be independent of the initial concentration, and of the water:pyrite ratio. In this case the experimental data would be expected to lie on a vertical line in plots such as Figure A4.4.3. Clearly the experimental data do not show such a trend and therefore solubility control is considered unlikely. From the surface analysis data it is likely that sorption of selenite to pyrite occurs in oxidised areas. Other studies have shown that oxidation of pyrite in such systems can give rise to an iron(III) oxide or hydroxide surface (Gillespie, 2001), which is known to sorb selenite in this pH region (Hayes, 1997; Hayes et al., 1987).

2– A4.4.1.2 KULeuven: Interaction of SeO3 with Boom Clay components (pyrite, OM)

Interactions of selenite with main Boom Clay components (pyrite, organic matter, and Fe2+) have been studied by KULeuven in several phases in the different Framework Programs (FP 4th, 5th, 6th) of the European Commission (EC). Results have been obtained for the TRANCOM-I and II projects dealing with the Transport of radionuclide by Complexation by Organic Matter.

TRANCOM-I project (FP-4)

In the framework of the TRANCOM-I European project (FP-4), several systems were set up in order to study the interaction of selenium with pyrite, both in the absence and the presence of Boom Clay organic matter, and with or without addition of Fe2+.

Three different crushed pyrite samples from two origins were used after sieving.

75 75 Se (T½ = 120 days) was used as tracer to study all the considered systems. Se was initially 75 2– expected to be in the SeO3 form as announced on the data sheet from the supplier, but this

204/328

was not verified. However, afterwards, we realised that oxidation reactions induced by water 75 2– radiolysis products might have converted a fraction of selenite in selenate ( SeO4 ).

75Se was added to three different systems: (i) pure pyrite only, (ii) pyrite with organic matter extracted from Boom Clay, and (iii) pyrite with addition of Fe2+ to precipitate iron(II) carbonate.

75 2– 1. SeO3 + pyrite system The interaction of selenite with pyrite was initially studied as a function of time in order to evaluate eventual kinetic effects. FeS2 was equilibrated for different periods of time with a synthetic clay water (SCW) solution containing 9 × 10-8 mol dm-3 of 75Se. After equilibration, the systems were centrifuged and the supernatants were used for measuring the 75Se activity, pH, and occasionally Eh. All the experiments were carried out under anaerobic conditions.

Between 7 days and 104 days of contact time, 75Se concentration remained constant around 4.4 × 10-8 mol dm-3, and no measurable concentration changes were observed. So, from these experimental results, in contrast to some literature data (Chao and Sanzolone, 1989), we conclude that the equilibrium concentration was reached after one week. A constant KD of about 100 dm3 kg-1 was obtained for six experimental durations.

75 2– 2. SeO3 + pyrite + organic matter system The influence of Boom Clay organic matter was investigated by pre-equilibrating pyrite with synthetic clay water (SCW) or with a Boom Clay extract (BCE). The overall starting concentration was 2.17 × 10-8 mol dm-3. The systems were allow to equilibrate for different time periods before phase separation and sampling.

No significant effect of the presence of humic substances on the “steady-state” selenium concentration (2 - 4 × 10-9 mol dm-3) was observed. Therefore, it was first considered that selenium was likely not associated with humic substances, in contrast to the frequent mentions in soil science literature (Christensen et al., 1989; Abrams et al., 1990; Fio and Fujii, 1990). This was also verified by gel permeation chromatography (GPC) of the 75 2– supernatant solutions. However, the possible contamination of the SeO3 spike by an 75 2– undetermined fraction of SeO4 could also explain these results.

3 -1 The KD values observed in this experiment (2 000 – 4 000 dm kg ) are higher than in the experiment on kinetic effects with pyrite only. It is not excluded that adsorption occurred onto the FeS2 surface irrespectively of the presence, or the absence, of Boom Clay organic matter.

The higher KD value can be ascribed to the higher specific surface area of the sample used in

205/328

this experiment (< 100 µm) compared to the kinetic experiments with pyrite only (> 0.84 mm).

75 2– 2+ 3. SeO3 + pyrite + [Fe / FeCO3 or Fe(OH)2] system Since both pyrite and siderite are present in the Boom Clay, it would be worth to study the interaction of selenium with both of them present at the same time. Pyrite was therefore contacted with synthetic Boom Clay water (SCW) to which different Fe2+ concentrations were added with the intention to precipitate increasing amounts of FeCO3 in the system. The purpose was to prepare a synthetic solution as close as possible to the interstitial water which 2+ is also at equilibrium with siderite (FeCO3) in situ in Boom Clay. After addition of Fe a precipitate was well visually observed, but unfortunately it was not analysed by XRD. It is therefore not possible to formally conclude that it was really FeCO3. The presence of a hydrated ferrous hydroxide, or green rust (mixed double layer) cannot be excluded.

The overall initial Se concentration was 4.97 × 10-7 mol dm-3. However, quite unexpectedly, the final selenium concentration was much higher than the predicted solubility of FeSe2. Despite the fact that the pyrite sample used in this experiment had a higher specific surface than the sample used to study the influence of Boom Clay organic matter the observed KD value (150 dm3 kg-1) was much smaller. Here again, the possible contamination of the 75 2– 75 2– SeO3 spike by an undetermined fraction of SeO4 (water radiolysis) could also explain these results.

75 2– 4. Possible interferences due to the presence of SeO4 in the three studied systems 2– The above-mentioned experiments seem thus to suggest that the interaction between SeO3 and FeS2 is dominated by adsorption effects, with no significant influence of organic matter.

However, the KD values observed in the experiments, are not in line with each other. Therefore, other unexpected mechanisms might be at hand. A first explanation of these results could be searched for in the poorly characterized chemical composition of the 75Se spike used: another selenium species than selenite might be present in the spike. The constant concentration of selenium (over time periods up to 100 days) observed in each experiment, 75 2– could be due to the presence of SeO4 , which is neither reduced nor adsorbed by the FeS2 surface (see Subsection A4.3.1). The difference in KD values observed could then entirely be 75 2– 75 attributable to the different amounts of SeO4 initially present in the Se spiking solution. Therefore, in the further experiments, an extensive use was made of the speciation techniques presented in Section A4.4 to discriminate selenite from selenate.

206/328

TRANCOM-II project (FP-5)

In the framework of the TRANCOM-II European project (FP-5), different sorption tests with 75 2– SeO3 were carried out by KULeuven on pyrite, used as a sorbent, in the presence and the absence of Boom Clay organic matter (OM).

75 2– 1. SeO3 + pyrite system Different amounts of ground pyrite (solution-to-pyrite ratios: 400:1 and 100:1 respectively) were contacted with synthetic Boom Clay water (SCW) in the absence of organic matter and spiked with different total concentrations (5 × 10-6 mol dm-3 and 10-6 mol dm-3) of 75Se (+ cold 75 2– carrier included) supposed to be in the form of SeO3 . The batches were allowed to equilibrate over different time periods up to two months before analysing. The selenium speciation and solution concentrations were measured over time.

2– A decrease in time of SeO3 concentrations was observed, depending mostly on the amount 2– of FeS2 present. Higher concentrations of SeO3 and lower solid-to-solution ratios of FeS2 2– resulted in higher end concentrations of SeO3 after a certain time. After relatively short equilibration times (up to one week), the systems follow a simple linear distribution 3 -1 3 -1 mechanism with a log KD of about 1.6 ± 0.2 dm kg (KD ~ 40 dm kg ). In two systems

(containing the highest amounts of FeS2), a final dissolved selenium concentration of about -9 -3 3 × 10 mol dm was reached; in the other systems (containing the lowest amounts of FeS2), equilibrium was not attained after two months. No other selenium species in solution apart 2– from SeO3 were detected, indicating that: 2– − SeO3 was merely sorbed and that no other species were formed, or; 2– – − selenium reaction products in solution (SeO4 or HSe , respectively left by oxidation or reduction reactions) were only present in minute concentrations below limit of detection.

So, these results summarised at Figure A4.4.4 suggest that one, or several, of the following reactions might have occurred at the pyrite surface: 2– − SeO3 is directly sorbed onto FeS2 surface, and/or; 2– − SeO3 is sorbed onto newly formed Fe-hydrous oxides, and/or; 2– − SeO3 is progressively reduced and precipitated as elemental Se(0), and/or; 2– – − SeO3 is progressively reduced in HSe , and that FeSe2 is then slowly precipitated in contact with Fe2+.

207/328

1.4E-06 ) 1.2E-06 1E-06 M Se 3x10-9 M 2,5 g/l FeS2 1.0E-06 5E-06 M Se 2,5 g/l FeS2 8.0E-07

6.0E-07

4.0E-07 1E-06 M Se KD

[Se] FeS2 solid phase (mol/g 10 g/l FeS2 2.0E-07 5E-06 M Se 10 g/l FeS2 0.0E+00 -1.0E-06 0.0E+00 1.0E-06 2.0E-06 3.0E-06 4.0E-06 5.0E-06 [Se(+IV)] solution phase (mol/l)

3 days 7 days 21 days 64 days Linear (3 days) Linear (7 days) Linear (21 days) Linear (64 days)

2– Figure A4.4.4 (KUL): Batch tests with SeO3 in contact with pure pyrite (FeS2) in synthetic Boom Clay water (in the absence of organic matter). Selenium sorbed/reduced/precipitated onto the FeS2 solid phase -1 2– -3 (mol g ) as a function of SeO3 concentration in solution (mol dm ) for two initial selenite concentrations, two amounts of pyrite and different contact times. For short contact times (3 and 7 days) a simple linear distribution ratio (KD) is observed. For longer contact time the selenium concentration in solution progressively decreases in all tubes to converge to the value of 3 × 10-9 mol dm-3.

2– XRD measurements on similar systems containing SeO3 and FeS, and a final selenium concentration in solution of 3 × 10-9 mol dm-3 observed in systems with different initial Se concentrations, suggest that a crystalline elemental Se(s) precipitate could also have formed in the present tests.

75 2– 2. SeO3 + pyrite + organic matter system -3 The afore-mentioned method was then extended to similar FeS2 (10 g dm ; solution-to-pyrite ratio 100:1) containing systems, but now contacted with synthetic Boom Clay water (SCW) in the presence of added organic matter (OM). Two sets of experiments were made with Boom Clay organic matter extracted in two different ways: − The first one was prepared with 100 ppm of OM from the TROM29 batch (dissolved OM concentrated from large volumes of Boom Clay porewater under air in the frame of the Trancom-II project, see Maes et al. (2004a,b) for more details). The TROM29 batch only contained mobile dissolved OM and was exposed to oxygen from air;

208/328

− The second set was prepared with better preserved OM from solid Boom Clay Extracts1 (BCE, 100 ppm) and a dilution of these extracts (30 ppm). The BCE batch also contained immobile OM from the solid and was always preserved under anoxic conditions. 75 2– -6 -3 -6 -3 Two initial SeO3 concentrations were added (5 × 10 mol dm and 1 × 10 mol dm ). Gel permeation chromatograms (GPC) were performed from the BCE samples to study the Se-OM interaction.

The following results were obtained: in all the batches a fraction of the added organic matter was sorbed onto the solid FeS2 phase following a linear sorption isotherm 3 -1 3 -1 (log KD OM = 4.26 ± 0.05 dm kg ) (KD OM = 18 200 dm kg ).

In the TROM29 batches (positive Eh values measured, indicating a possible oxidation), the decrease in total Se concentrations with time is smaller with respect to the results obtained in the absence of organic matter. After 320 days, total selenium concentrations of the order of -7 -3 2– 1 - 2 × 10 mol dm were measured, indicating that SeO3 reduction was probably restricted.

In the BCE batches (negative Eh values measured, indicating still reducing conditions), total selenium concentrations decreased with time with the same rate as if no organic matter was present, but remained constant after 14 days equilibration time (Figure A4.4.5). At this point, a small fraction of the total selenium in solution (< 10 %) was associated with the organic 2– matter. This “steady-state” condition corresponded to a sorption mechanism for SeO3 with a 3 -1 3 -1 log KD value of 3.49 ± 0.01 dm kg (KD = 3 090 dm kg ). After the setting of this sorption 2– “equilibrium”, SeO3 concentrations decreased only at a very slow rate, while selenium concentrations associated with organic matter remained constant or even slightly increased. 2– -6 -3 After 90 days in the systems with initial SeO3 concentration of 10 mol dm , up to 25 % of total selenium in solution was observed to be associated with organic matter, while in the 2– -6 -3 systems with initial SeO3 concentration of 5 × 10 mol dm this percentage was only 2 % –

17 %. At the end constant KD values were obtained for free selenium 3 -1 3 -1 (log KD = 3.57 ± 0.01 dm kg ) (KD = 3 715 dm kg ) in solution.

2– The SeO3 concentrations in solution rapidly decrease with time towards two constant values (indicated in Figure A4.4.5 by vertical lines drawn at 1.35 × 10-7 mol dm-3 and 2.4 × 10-8 mol dm-3 respectively), after which an apparent “equilibrium” is achieved. These 2– two concentrations are about forty times lower than the two initial SeO3 concentrations respectively introduced in the systems.

1 (KULeuven) Boom Clay extract was prepared by mixing 50 g of Boom Clay with 200 mL synthetic Boom Clay water (SCW) (liquid:solid ratio = 4:1). After 2 weeks equilibration time, samples were centrifuged (Beckman J2-21, rotor JA-14, ? RPM, 2 h), the supernatant was decanted and centrifuged again (same conditions). The liquid phase was then used as a Boom Clay extract (BCE). (KULeuven)

209/328

Up to this point, the specific interaction mechanism of selenium with dissolved Boom Clay 2– organic matter observed in experiments starting from SeO3 , remains unclear. We might envisage two possible explanations, but others could still exist:

2– 1. SeO3 anions could directly interact (e.g. through a ligand exchange mechanism) with dissolved natural organic matter (NOM). Such behaviour is already known and documented for phosphorus (Hens and Merckx, 2002) and arsenic (Smedley and Kinniburgh, 2002), or;

2– 2. SeO3 could be reduced to elemental Se(s) (e.g. through an adsorption-reduction reaction

at the FeS2 surface, or possibly on Fe(II)-bearing groups within the organic matter itself), and subsequently, the formed elemental Se(s) colloids could interact with the organic matter colloids.

[Se] FeS2 solid phase vs. [Se (IV)] solution phase

Boom Clay extract

5.0E-07 )

4.0E-07 30 ppm BCE 30 ppm BCE 1E-06 M Se 5E-06 M Se 3.0E-07 10 g/l FeS2 10 g/l FeS2 100 ppm BCE "free"[Se]sol = 1.35x10-7 M 5E-06 M Se 10 g/l FeS2 2.0E-07 "free"[Se]sol = 2.4x10-8 M

100 ppm BCE 1.0E-07

[Se] solid FeS2 phase(mol/g 1E-06 M Se 10 g/l FeS2

0.0E+00 -1.0E-07 0.0E+00 1.0E-07 2.0E-07 3.0E-07 4.0E-07 5.0E-07 6.0E-07 7.0E-07 8.0E-07 [Se(+IV)] solution phase (mol/l)

8 days 15 days 37 days 92 days

2– -3 Figure A4.4.5 (KUL): Batch tests with SeO3 in contact with 10 g dm FeS2 in the presence of Boom -1 Clay organic matter Extract (BCE). Selenium associated with the FeS2 solid phase (mol g ) as a function 2– -3 of the SeO3 concentration in solution (mol dm ) for two initial selenite concentrations, two 2– concentrations of BCE, and different contact times. SeO3 concentrations in solution rapidly decrease with time towards two constant values (indicated by vertical lines drawn at 1.35 × 10-7 mol dm-3 and 2.4 × 10-8 mol dm-3 respectively), after which an apparent equilibrium is achieved.

To conclude on the interactions of selenite with Boom Clay components (pyrite and OM), it seems that in the absence of organic matter, selenite removal follows a simple linear distribution coefficient (KD). In the presence of OM, for short contact times, selenite also initially exhibits a sorption (KD) behaviour, but at long-term, the situation becomes more

210/328

complex and the removal of selenite from solution could perhaps still be attributed to sorption, or to another mechanisms such as reduction / precipitation occurring during the ageing of the suspensions. Then, elemental Se(0) is suspected by KULeuven to be formed in colloidal suspension and to become weakly associated with the organic colloids.

2– A4.4.2 Interaction of SeO3 with Boom Clay – KULeuven The adsorption and reduction of Se(+IV) in the presence of solid Boom Clay and Boom Clay organic matter was investigated by KULeuven only.

In the framework of the TRANCOM–I European project (FP 4), numerous experiments with 75Se spikes were performed, starting mainly from conditions of oversaturation. At this time, it 75 75 was thought that the Se spike used consisted solely of Na2 SeO3. The general set-up of the experiments was performed as follows. In short centrifuge tubes a weighted amount of Boom Clay was first allowed to equilibrate with synthetic clay water (SCW). Then a small amount of SCW stock solution spiked with 75Se was added and the tubes were shaken for a chosen equilibration period. Phase separation was made by centrifugation. Measurements of pH, Eh, 75Se count rate and optical density (280 nm) were made on the supernatant solutions. A gel permeation chromatogram (GPC) of the supernatant was occasionally taken.

75 2– The results of these experiments with SeO3 and Boom Clay suspensions can be described by three main observations on the behaviour of selenium. (1) Unexpectedly, and contrary to the data from literature, the selenium removal from solution in contact with Boom Clay reached very quickly a steady-state. (2) In a series of experiments, the equilibrium selenium solution concentrations in all batches were about identical and independent of the liquid–to– solid ratio and of the equilibration time. This suggested that the final selenium concentration was solubility controlled. (3) In another series of experiments, different starting 75Se concentrations resulted in different equilibrium concentrations, the isotherms exhibiting a typical sorption mechanism with constant KD value. Moreover, in these experiments, gel permeation chromatograms made on supernatant solutions did not provide arguments indicating an association between selenium and organic matter.

1. Fast steady states Since selenium is known to show slow redox-reaction equilibria (Chao and Sanzolone, 1989), the interaction of selenite with Boom Clay was firstly studied over an extra long time period of up to about one year. Experimentally the interaction of selenite with Boom Clay was studied at only one SCW solution:Boom Clay ratio of ~ 3.9:1 and for a starting selenite concentration of 4.69 × 10-8 mol dm-3 Se. Six identical systems were set up. The observed selenium concentrations in the “equilibrium” solutions after one week and up to 43 weeks of equilibration were identical (3.9 × 10-9 mol dm-3) indicating that a steady-state was reached

211/328

within the short time period of one week in contrast to some literature data (Chao and Sanzolone, 1989). This rather quick reaction equilibrium was confirmed in experiments at different liquid–to–solid ratios with Boom Clay and also in experiments with pyrite. The gel permeation chromatograms taken on the extracts show one larger and one smaller UV peaks. The position of 75Se in the activity patterns coinciding with that of the second UV peak in the optical density pattern, suggests the presence of (small size) selenium oxyanions. Since the first large organic matter peak did not contain selenium, the initial working assumption was that selenium was not associated with organic matter in these experiments, in contrast to 75 2– literature data (Abrams et al., 1990). Afterwards, the presence of SeO4 was suspected in the 75Se spiking solution because of oxidation reactions induced by the water radiolysis. So, the observed 75Se behaviour was at least partly, if not entirely, imposed by selenate.

2. Indications of a solubility control In a second experiment, the distribution of 75Se was measured at different SCW solution:Boom Clay ratios and for different equilibration times using a starting concentration of 1.66 × 10-8 mol dm-3 Se. The equilibrium selenium solution concentrations in all experiments were about identical (6.8 × 10-9 mol dm-3) which is independent of the liquid–to– solid ratio and independent of the equilibration period. These results confirm the previous observations that equilibrium was reached after 7 days of contact time. The constant solution concentration of selenium versus an increasing concentration of removed Se per g clay at different liquid–to–solid ratio suggested that the selenium concentration was solubility controlled. Suspending Boom Clay at different liquid–to–solid ratios leads to different organic matter concentrations in the supernatant solutions. The constant value of the Se concentrations suggested (deduction) that selenium was not associated with organic matter since the presence of different humic substance concentrations in the supernatant solutions had no influence on the equilibrium Se concentration. The former conclusion was underpinned by a gel permeation chromatogram of the equilibrium solutions. Again, the large organic matter peak did not contain 75Se: this observation seemed to support the independence of the selenium level on the overall humic substance concentration. However, due to the 75 2– initially undetected presence of SeO4 in the spiking solutions submitted to an intense γ -radiolysis, the first interpretations of these experiments were inappropriate and had to be revised.

3. Indications of a sorption mechanism The solubility of selenium in the presence of Boom Clay was checked by contacting different 75Se concentrations (from 3.7 × 10-9 mol dm-3 to 3.7 × 10-7 mol dm-3) in synthetic clay water (SCW) with Boom Clay at a liquid–to–solid ratio of about 4:1. The systems were allowed to equilibrate for 1 week by end-over-end shaking. Different starting concentrations were used corresponding to different degrees of oversaturation. Unexpectedly, the observed equilibrium

212/328

selenium concentrations were different from each other (ranging from 3.75 × 10-10 mol dm-3 to 3.95 × 10-8 mol dm-3) and thus did not correspond to a precipitation mechanism. On the contrary, a plot of the selenium adsorption isotherm exhibits a typical sorption mechanism 3 -1 with a constant KD of about 34 dm kg .

The previous experiment (at different liquid–to–solid ratios) and the present data (constant liquid–to–solid ratio) thus each correspond to a different behaviour of selenium in the presence of Boom Clay. The first interpretation coming to the mind is that a precipitation and a sorption mechanism could maybe explain respectively the results of these two series of experiments.

In the final report of the TRANCOM-I European project (FP 4) (Dierckx et al., 1999), the discrepancy in the interaction mechanism (precipitation versus adsorption) observed in the different series of experiments with 75Se and Boom Clay, was attributed to the chemical composition of the two different batches of 75Se used as stock solution. However, since no clear difference between the two batches could be traced back with the techniques then applied, no answer to the problems was found. Since then, it became evident that an undetermined fraction of 75Se in the two stock solutions could have been present as selenate 75 2– ( SeO4 ) because of the progressive oxidation of selenite by the water radiolysis products. Selenate is not expected to react with any of the Boom Clay components at relatively short time scale. Therefore, the main results from the afore-mentioned TRANCOM-I experiments (both in the presence of Boom Clay and/or pyrite) are recapitulated hereafter in Table A4.4.4 for reassessment.

213/328

Table A4.4.4 (KUL): Overview of the main results of the different experiments with 75Se performed in the frame of the TRANCOM-I European project.

Batch Liq/Sol Initial Se Equil. Se Equil. Se conc. Log KD Glove Type of N° ratio conc. conc. as % of initial box experiment (—) (dm3 kg-1) (mol dm-3) (mol dm-3) Se conc. (dm3 kg-1) (atm.) (—)

-8 -8 1 3.9 6.30 × 10 2.80 × 10 ~ 44 % 0.69 N2/CO2 Succes. Extract. Boom Clay -8 -9 1 3.9 1.66 × 10 6.80 × 10 ~ 41 % 0.72 N2/H2/ Differ. L/S CO2 Boom Clay

-8 -9 2 3.9 4.70 × 10 3.90 × 10 ~ 8 % 1.60 N2/CO2 Long time exp. Boom Clay -7 -8 2 4.0 3.70 × 10 3.95 × 10 ~ 11 % 1.55 N2/H2 Differ. Conc. Boom Clay -9 -10 2 4.0 3.70 × 10 3.70 × 10 ~ 10 % 1.55 N2/H2 Differ. Conc. Boom Clay

-7 -7 1 100 4.97 × 10 1.92 × 10 ~ 39 % 2.20 N2/H2 Pyrite + Siderite -8 -8 1 100 9.20 × 10 4.30 × 10 ~ 48 % 2.04 N2/H2/ Kin. Effects CO2 Pyrite

-8 -8 2 400 2.17 × 10 2.51 × 10 ~ 11 % 3.50 N2/H2 Absence of OM Pyrite -8 -8 2 400 2.17 × 10 2.64 × 10 ~ 12 % 3.50 N2/H2 Presence of OM Pyrite OM: organic matter.

From data of Table A4.4.4, especially the ratio (%) given in the 5th column, it becomes now clear that, irrespective of the liquid–to–solid ratio, the nature of the solid, the presence of dissolved organic matter, and the type of glovebox used, a very characteristic feature emerges for all experiments after one week contact time: the fraction (expressed as %) of the initial Se concentration remaining in solution systematically corresponds to one of the two following values: ~ 44 % for batch 1 and ~ 10 % for batch 2. So, it appears that the selenium proportion left in solution at the end of the experiment was only dependent on the 75Se batch used, and did not decrease over longer contact times. Thus, the 75Se concentrations measured on the long term could no longer be considered as the result of an equilibrium situation, but rather as a simple steady-state.

Therefore, to explain this unforeseen finding, it was postulated that the selenium percentage 2– 75 remaining in solution correspond to the SeO4 fraction present in the original Se batch. For the 75Se batch 1, this Se(+VI) fraction amounts to about 44 ± 4 %, while for the 75Se batch 2 2– the Se(+VI) fraction is about 10 ±2 %. This SeO4 fraction is not adsorbed, nor reduced, nor

214/328

associated to organic matter, and thus rapidly dominates the total selenium concentration in 2– solution while the SeO3 fraction is either adsorbed, or reduced and precipitated (see Subsections A4.4.1 and later).

As a consequence, it is essential to control the selenium aqueous speciation in the initial 75Se spiking solution and in the final supernatants. A clear distinction between selenite and selenate is required to correctly interpret the experimental.

The KD values listed in Table A4.4.4 spread on three orders of magnitude and are comprised 3 -1 75 75 2– in the range 5 to 3160 dm kg . The KD obtained from the Se batch N°1 (richest in SeO4 ) lie in the lowest part of the range: from 5 to 160 dm3 kg-1. These from the 75Se batch N° 2 75 2– 3 -1 (with less SeO4 ) are the highest: from 40 to 3160 dm kg . This is consistent with the 2– absence of sorption observed for SeO4 in the conditions and at the time scale of these experiments.

In a new set of experiments, Boom Clay batches (Boom Clay + synthetic Boom Clay water, SCW) with three 75Se concentrations (1 × 10-6 mol dm-3, 5 × 10-7 mol dm-3, and 1 × 10-7 mol dm-3) were measured as a function of time. The speciation techniques described in Section A4.4 were used to distinguish the different selenium species in solution. Results from this set of experiments showed that the 75Se spike used was composed of both Se(+IV) and Se(+VI) species.

2– It was observed that selenite (SeO3 ) ions were removed from solution on a relatively short 2– time scale. However, equilibrium was not attained in less than one month. Selenate (SeO4 ) ions clearly remained in solution.

This confirms thus the hypothesis postulated to account for the contradictory results obtained in the experiments carried out for the TRANCOM-I project. It also shows the slow kinetics of 75 2– SeO3 interaction with Boom Clay: selenite can be reduced, but a constant concentration was not reached in less than one month.

In a last, long-term experiment, the speciation of selenium in the presence of Boom Clay was 75 2– studied starting again from oversaturation with SeO3 . Systems were prepared using two 75 2– liquid-to-solid ratios (20:1 and 4.76:1), as well as two SeO3 concentrations (5 × 10-6 mol dm-3 and 1 × 10-6 mol dm-3). These systems were allowed to equilibrate up to 9 months. The supernatant of the samples was analysed for total dissolved 75Se concentration, 75Se concentration left in solution after La3+ precipitation of dissolved humic substance (HS), optical density (OD) at 280 nm, Eh and pH. The solutions were also analysed by anion and gel permeation chromatography.

215/328

The experimental results indicated that the speciation procedures used were sufficiently 75 2– accurate for identifying the different Se species in the Boom Clay suspensions: SeO4 , 2– SeO3 and Se associated to the organic matter were readily determined. However, selenide (HSe–) could not be directly detected. It is also not excluded that dissolved selenide could have been immediately oxidised upon injection in the column of anion chromatography or gel 2– permeation chromatography, and so, were perhaps measured as SeO3 . The pH and Eh measurements indicated that the expected geochemical conditions were likely met for the 75 2– reduction of the spike of SeO3 in the Boom Clay suspensions.

75 2– Upon addition of SeO3 to the Boom Clay systems, a distribution of selenium between the liquid and the solid phases was first observed. Selenite sorption could be attributed to the formation of inner-sphere complexes at the surface of broken edges of the clay particles (also known as ligand exchange mechanism).

As presented on Figure A4.4.6, after a certain time (depending on the system studied) two 75 2– -7 -3 “stable” concentrations of SeO3 were observed at respectively 1.4 (± 0.2) × 10 mol dm and 2.4 (± 0.2) × 10-8 mol dm-3. These two concentrations coincide remarkably with the 75 constant values of Se measured in the experiments made with FeS2 and Boom Clay extract (BCE), whose description is given in Figure A4.4.5.

75 2– Afterwards, in some systems SeO3 concentrations in solution decreased again towards 3 × 10-9 mol dm-3, the experimental “solubility” value observed in experiments made with

FeS2 as a reductant. As already previously observed on pure pyrite systems, the amount of solid phase in the systems seems to be a critical parameter for the reaction kinetics. In all systems 75Se concentrations in solution tended to decrease, but the system with the lowest initial Se concentration and the largest solid–to–liquid ratio attained equilibrium most rapidly. In other words, the redox-systems evolution seems to follow the offer and demand in electron: when the number of sites (from the solid) giving electrons largely exceeds that of electron acceptors in solution, the redox-reaction rate considerably increases. The offer must be larger than the demand to accelerate the reaction rate.

Finally, in all the new systems, 75Se was also found to be associated with dissolved organic matter. This fraction was initially low (10 % – 30 %) relative to the total dissolved selenium concentration, but remained quite constant (although sometimes even slightly decreasing) throughout the experiment. However, after nine months, in some samples it increased and in these systems most of selenium in solution (40 % – 75 %) became associated with organic matter.

216/328

1.4E-07 5E-06 M Se

) 0,05 kg/L BC 1.2E-07

[Se]solution = 3x10-9 M 1.0E-07

8.0E-08 [Se]solution = 2.4x10-8 M 6.0E-08 [Se]solution = 1.35x10-7 M 1E-06 M Se 0,05 kg/L BC 4.0E-08 1E-06 M Se 0,21 kg/L BC

[Se] Boom Clay solid phase (mol/g 2.0E-08 5E-06 M Se 0,21 kg/L BC

0.0E+00 -5.0E-08 0.0E+00 5.0E-08 1.0E-07 1.5E-07 2.0E-07 2.5E-07

[Se(+IV)] Boom Clay solution phase (mol/l)

1 week 2 weeks 1 month 3 months 9 months

2– Figure A4.4.6 (KUL): Long-term batch experiments with SeO3 in contact with Boom Clay suspensions. Selenium associated with the Boom Clay solid phase (mol g-1) as a function of the 2– -3 SeO3 concentration in solution (mol dm ) for two initial selenite concentrations, two Boom 2– Clay solid-to-liquid ratios and different contact times. SeO3 concentrations in solution rapidly decrease towards two “constant” values (of 1.35 × 10-7 mol dm-3 and 2.4 × 10-8 mol dm-3 as 2– represented by vertical lines). After three months’ equilibration time, SeO3 concentrations in the two systems with the highest solid-to-liquid ratio again decreased towards the previously observed experimental selenium “solubility” value of 3 × 10-9 mol dm-3.

2– A4.4.3 Kinetic of reduction of SeO3 onto pyrite – KULeuven 2– 2– Upon contacting SeO3 with pure pyrite (FeS2), a steady decrease in time of SeO3 concentrations was observed, until a final concentration in solution of 3 × 10-9 mol dm-3 was reached. All investigated systems appear to follow a same rate law. According to Bruggeman et al., (2005), the decrease in Se(IV) concentration as a function of time seems proportional to the concentration of dissolved selenite and to the amount of solid FeS2 present in the system, and inversely proportional to the square root of the FeS2 occupancy by selenite. These 2– observations suggest that SeO3 reduction takes place through sorption onto FeS2 and that a selenium precipitate with a solubility of 3 × 10-9 mol dm-3 was formed. Based on a fit of the experimental data the following rate law is proposed:

d[]Se(+IV) [FeS ] = −k [][]Se(+IV) FeS 2 0 (eq. A4.4.4) dt 2 0 Se(+IV) []0

One also observes that the Se(IV) reduction rate slows down when clay minerals and dissolved Boom Clay organic matter are present in the system.

217/328

2– The competitive chemisorption of SeO3 onto the broken edges of clay minerals could be a possible explanation of the decrease of the selenite reduction rate in the presence of clay.

Indeed, if, on one hand, selenite can distribute freely between the solution and the clay 2– minerals edges, and, on the other hand, can sorb on pyrite surface, the adsorption of SeO3 onto the clay minerals could compete with that of pyrite, and so contribute to retard the 2– reduction of SeO3 by the FeS2 surface. This competition between two solid phases for selenite sorption is illustrated by the following combination of exchange reactions:

2– 2– Clay edge—SeO3 (ads.) <=> Clay edge + free SeO3

2– 2– free SeO3 + Pyrite <=> (ads.) SeO3 — Pyrite

2– 2– 2– Clay edge—SeO3 (ads.) <=> free SeO3 <=> (ads.) SeO3 — Pyrite

The effect of dissolved Boom Clay organic matter on the decrease of the selenite reduction rate is much less clear and remains presently not understood.

2– A4.4.4 Solubility of SeO3 – KULeuven 2– The SeO3 concentrations in Boom Clay are not limited by the solubility of metal selenite precipitates, as e.g. CaSeO3, a solid phase expected to be formed when selenite reacts with 2+ 2– Ca from cement. It could be worth to study more into detail the coprecipitation of SeO3 with calcium silicate hydrate (CSH) phases, CaCO3 (calcite), and FeCO3 (siderite), in cementitious environments as these expected for the SuperContainer concept, or the buffer to be used as backfill material for galleries of intermediate level waste.

2– However, when SeO3 is added to systems containing fresh Boom Clay suspensions, or 2– chemical reductants, SeO3 is reduced (e.g. by adsorption onto FeS2) and poorly soluble 2– selenium (0) or (-II) phases can precipitate. After complete disappearance of SeO3 , the selenium solution concentration is controlled by the solubility of the resulting selenium solid phases: elemental Se(s), iron selenides such as FeSe, or FeSe2, or solid solutions with iron sulfides such as [FeS2-×(Se)×].

2– A4.4.5 Conclusions for SeO3 The different systems studied with selenite in prolonged contact with suspensions of Boom Clay, or its different components, appear to be relatively complex and very sensitive to

218/328

experimental artefacts such as auto-oxidation. This makes interpretation of some results particularly difficult. However, one of the main cause of artefacts with 75Se radio-labelled 75 2– selenium has been now clearly identified and is due to the progressive oxidation of SeO3 in the spiking solution due to the intense water radiolysis. To avoid the interference of 75 2– SeO4 in the spike, the purity of the reagents has to be checked prior to experiments and speciation techniques are also mandatory to follow the evolution of 75Se in the supernatants after equilibration.

2– The removal of SeO3 from solution appears first to start as a linear sorption isotherm (KD approach), then, it seems to progressively evolve with time as a slow reduction/precipitation and the selenium concentration in solution converges towards a solubility limit (3 × 10-9 -3 -3 -1 mol dm ). The KD obtained so far are in the range 5 to 5 000 dm kg , but the extrapolation towards in situ conditions is not straightforward. The main uncertainty resides in the still poor knowledge of the retention mechanism and how to translate the obtained laboratory results in term of performance assessment (PA) calculations. Two options could be first envisaged: 2– − SeO3 simply sorbs without any reduction: pure KD approach without solubility limit, or; − Se(0) is the immobilised species after a complex combination of sorption/reduction. Sorption is not taken into account in the calculations and one should consider that the mobile species under reducing conditions is HSe– which is supposed to migrate without retardation (R = 1) but is solubility limited (S = 3 × 10-9 mol dm-3).

In other words, considering the thermodynamic calculations presented by AEA Technology at Figure 2.5.2 (p. 33, Chapter 2), if one assumes that the selenium species in equilibrium with

Boom Clay is the elemental Se(0), its solubility strongly depends on the Eh value. The dissolved selenium concentration as a function of Eh presents a clear V-profile with a shape of 2– young valley. If Eh increases towards less reducing conditions, SeO3 is expected to be the dominant aqueous species, that would migrate retarded (KD approach) but without solubility – control. At the contrary, if Eh decreases towards more reducing conditions, HSe is the species expected to dominate in solution and would migrate unretarded with a solubility limit. The same reasoning could be done for FeSe if necessary.

2– The reduction kinetics of SeO3 seems also to be proportional to the amount of pyrite introduced in the system and to the total selenite concentration present in solution. It is also inversely dependent on the square root of the selenite occupancy onto the FeS2 surface (Bruggeman et al., 2005).

2– A clear association of SeO3 with OM has been observed in contrast to all other selenium species studied in the present work.

219/328

A4.5 Behaviour of elemental Se(s) The solubility of elemental Se(0) has been studied independently by AEA Technology and KULeuven in synthetic and interstitial clay water, and in the presence of pyrite, iron strips, and with solid Boom Clay to vary Eh values and to see its influence on the Se(0) solubility. Ultrafiltration tests and gel permeation chromatography are also performed to observe a possible association between Se and organic matter.

A4.5.1 Interaction of elemental Se(s) with Boom Clay components (pyrite, OM)

A4.5.1.1 AEAT: Interaction of elemental Se(s) with Boom Clay components (pyrite, OM) The interaction of elemental selenium with crushed pyrite and pyrite coupons was first examined in a synthetic Boom Clay water in the absence of organic material. Sufficient selenium solid was added to give a selenium concentration of 5 × 10-4 mol dm-3 if dissolved completely. The solution–to–pyrite ratio was 50:1 in the experiments with crushed pyrite.

Although selenium concentrations of 1 × 10-5 mol dm-3 were measured in solution after one month’s equilibration, these concentrations decreased over the one year duration of the experiment. After 371 to 379 days’ equilibration, dissolved selenium concentrations were close to or below a detection limit of 1 × 10-7 mol dm-3 in these experiments at pH 8.3

(Eh = -100 mV) and 9.9 (Eh = -280 mV) respectively. Sampling of undisturbed experiments after 371 to 379 days’ equilibration gave selenium concentrations of 5 × 10-7 to -7 -3 -7 -7 -3 8 × 10 mol dm at pH 9.9 (Eh = -220 to -240 mV) and ≤ 1 × 10 to 4 × 10 mol dm at pH 8.4 to 8.5 (Eh = -100 to -120 mV) in the presence of crushed pyrite. Concentrations of ≤ 4 × 10-7 and ≤ 7 × 10-7 mol dm-3 were measured in the presence of the pyrite coupons.

XRD analysis of pyrite powder from the experiments showed no bulk changes to the mineralogy and no secondary selenium phases were found. Interestingly, XRD only identified pyrite, washing of the pyrite during sampling had separated the pyrite from the majority of the elemental selenium. Visual inspection of the coupon recovered from the experiment after 13 months’ equilibration prior to carbon coating for SEM showed the surface had a greenish coloration; this was not due to surface oxidation and may have been due to a very thin carbonate layer. Analysis revealed the surface to be coated with 1 to 5 µm crystal aggregates of selenium metal, individual selenium needles and circular oxidised areas (associated with drying spots from the evaporation of droplets of residual washing solution). Some of these oxidised areas had crystals of elemental selenium associated with them. No iron-selenium compounds (formed by reaction of selenium with the pyrite surface) were observed.

220/328

The dissolution of elemental selenium in interstitial Boom Clay water and 10 % interstitial Boom Clay water (obtained by diluting interstitial Boom Clay water with synthetic Boom Clay water) in the absence of Boom Clay, was examined. A mass of 1 g of elemental selenium was contacted with 50 cm3 of water. The individual experiments were then sampled after approximately 1, 2 or 3 months’ equilibration and filtered through 0.22 µm, 100 000 NMWCO and 10 000 NMWCO filters (Nominal Molecular Weight Cut-off).

In both waters the selenium concentrations in solution showed an increase with time but there were no discernible trends with the pore size of the filter. The selenium concentration in the 10 000 NMWCO filtrate of interstitial Boom Clay water rose from 7 × 10-7 mol dm-3 after 33 days (pH = 9.5), to 2 × 10-6 mol dm-3 after 68 days (pH = 9.4), and 6 × 10-5 mol dm-3 after 96 days (pH = 9.0). Measured redox potentials fell from -180 mV vs SHE after 33 days to -260 mV vs SHE after 68 days and 96 days. The overall trend in selenium concentration in

10 % interstitial Boom Clay water was similar. After 42 days (pH = 9.4, Eh = -230 mV) the selenium concentration ranged from 2 × 10-7 mol dm-3 (10 000 NMWCO) to -7 -3 7 × 10 mol dm (0.22 µm). However, after 66 days (pH = 9.4, Eh = -210 mV), the concentrations in solution (for all filter sizes) were below the detection limit of 3 × 10-7 mol dm-3. These may be anomalous results because selenium concentrations of -6 -3 4 × 10 mol dm were measured after 95 days equilibration (pH = 9.5, Eh = -210 mV). Although the results after 42 days’ equilibration might suggest some increase in selenium concentration with increasing filter pore size, this is not seen for the 95 day results.

After the initial sampling the three experiments in interstitial Boom Clay water were ‘re-started’. All the remaining solution was removed and the solids washed with deionised water. Fresh interstitial Boom Clay water was then added to each experiment. In addition, an iron strip (10 mm × 50 mm × 0.25 mm) was placed in two of the experiments. Within 2 to 4 days’ equilibration it was noted that the solutions in the experiments containing iron strips were black and opaque and a fine precipitate had also formed. The experiments were sampled after 18 to 61 days equilibration and the samples filtered through 0.22 µm, 100 000 NMWCO and 10 000 NMWCO filters. The experiment containing no iron strip was then ‘re-started’ for a further equilibration in the same manner as before and sampled after 45 days’ equilibration. Iron strips and selenium solids were recovered from the experiments for analysis by SEM and EDX.

Selenium concentrations of 4 × 10-6 mol dm-3 after 0.22 µm filtration, 4 × 10-6 mol dm-3 after 100 000 NMWCO filtration and 1 × 10-6 mol dm-3 after 10 000 NMWCO filtration were measured after re-equilibration with fresh interstitial Boom Clay water for 18 days in the absence of an iron strip (pH = 9.4, Eh = -210 mV). The subsequent re-equilibration with further fresh interstitial Boom Clay water for 45 days gave selenium concentrations of -4 -3 3 × 10 mol dm (pH = 9.8, Eh = -220 mV). The selenium concentrations above elemental

221/328

selenium re-equilibrated with fresh interstitial Boom Clay water and an iron strip for 44 and 61 days showed no effect of filter pore size and were 3 × 10-4 to 7 × 10-4 mol dm-3 at a redox potential of –290 to -330 mV and pH of 8.7 to 9.2. These latter concentrations are much higher than measured in the first initial equilibrations with interstitial Boom Clay water or synthetic Boom Clay water. It suggests that interaction with iron may result in increased concentrations of selenium in solution above elemental selenium. It is possible that this is due to more reducing conditions achieved in the presence of corroding iron (see Subsection 2.4).

EDX analysis of particles recovered from the experiments containing iron strips (Figure A4.5.1, AEAT) showed them to be mostly Fe, Se phases with small bright spots on the surface of the particles which were unreacted selenium. In addition, the presence of a coating on the iron strips was evident to the naked eye. After 40 days, this had the appearance of a uniform porous crust consisting of Fe, Se and O, with occasional trace Na and P. Upon magnification, the coating was seen to consist of equal-sized spheres embedded in a crystal matrix. The spheres were shown by EDX to be more oxidised and iron-rich than the crust which, although still Fe-rich, contained more selenium. In areas where the coating was absent, the surface of the iron strip was corroded. In comparison, iron strips from the iron selenide experiments showed no comparable corrosion (Figure A4.5.2, AEAT).

An additional type of coating was observed on the iron strip examined after 60 days’ exposure. This had the appearance of a dehydrated gel sitting directly on the iron surface, and due to the relationship to the other coating types, appeared to be the first coating to form. EDX showed that this was an iron oxide gel, with a high carbon content possibly from the organic content of the Boom Clay water, containing some phosphorus and sulfur. The oxidised porous crust containing spheres sat on top of this and was, as after 40 days, iron oxide spheres in a selenium-rich matrix. Adhering to this were large particles which are the most selenium–rich (Figure A4.5.3, AEAT). The solid analyses provide qualitative evidence that Fe/Se-containing phases may have been formed through the reaction of selenium with iron.

It is worth to note here that the reaction of elemental Se(0) with metallic iron might have potential implications for the corrosion of the metallic canisters and iron overpacks used to contain the high level waste. It is observed by AEA Technology that elemental Se(0) induces a general corrosion of iron strips surface immersed in interstitial Boom Clay water under reducing conditions. This mechanism should be related to the well known detrimental effect of elemental sulfur on iron corrosion (MacDonald D. et al., 1978; Marcus, 1995; Nirond, 2004). So, a new question arises: could the quantity of selenium present in the high-level waste (typically ~ 1 mol Se per meter current gallery) be a concern for the long-term corrosion resistance of the metallic barriers ?

222/328

Se0 + Fe strip 40 days solid – 1

Figure A4.5.1 (AEAT): Fe, Se solids recovered from elemental selenium + iron strip experiments after 40 days. EDAX from recovered solids after 60 days.

223/328

Fe strip 60 days in FeSe - 1

Fe strip 40 days in Se0 - 12

Figure A4.5.2 (AEAT): Comparison of iron strips showing surface etching in strip exposed to elemental selenium (lower figure) compared to that from iron selenide experiment.

224/328

13/10/2003 11:56:57 Fe strip 60 days in Se0 - 6

Processing option : All elements analyzed (Normalised).

O Na P S Fe Se Total

Spectrum 1 75.24 0.59 0.36 0.26 21.52 2.03 100.00 Spectrum 2 50.35 39.07 10.59 100.00 Spectrum 3 59.51 2.15 0.32 37.35 0.67 100.00 Spectrum 4 14.39 30.65 54.96 100.00

All results in Atomic Percent.

Figure A4.5.3 (AEAT): Coating on iron strip after 60 days’ exposure to solution in contact with elemental selenium and EDAX results from selected areas.

225/328

A4.5.1.2 KULeuven: Interaction of elemental Se(s) with Boom Clay components (pyrite, OM) As laid out in Section A4.2, reduced 75Se solid phases were prepared artificially. Chemical reduction systems provided amorphous red and crystalline grey-black elemental, Se(s), using 2– hydrazine (H4N2) and sodium sulfide (Na2S). Electro-reduction of SeO3 using a galvano- potentiostat was also used. Solid Se phases were precipitated on the surface of metallic (Pt, Ni) scraps. Amorphous and crystalline Se(s) solid phases were again identified.

The red amorphous 75Se0 and grey hexagonal 75Se0 phases from chemical reduction experiments were first contacted with synthetic Boom Clay water (SCW) (+ 2.37 × 10-5 mol dm-3 Fe2+) to measure the solubility of the prepared 75Se reduced solid phases. The solubility was found to be dependent on the reducing system and the solid phase. 75Se concentrations in solution fell in three solubility ranges, consistent with different solid phases i.e. amorphous Se(s) (2.2 × 10-8 mol dm-3 – 9.1 × 10-8 mol dm-3), and crystalline Se(s) (1.5 × 10-9 mol dm-3 – 8.1 × 10-9 mol dm-3). After removal of the supernatant SCW solution, half of each solid phase was re-contacted with SCW (+ 2.37 × 10-5 mol dm-3 Fe2+) and the other half was contacted with organic matter from Boom Clay Extract (BCE). The systems were allowed to equilibrate for 2 months and then were analysed for: − total 75Se concentration in solution; − 75Se concentration in solution after precipitation of dissolved humic substances using La3+; − Optical density in ultra-violet at 280 nm, and;

− Eh and pH. The total final 75Se concentrations ranged from 5.5 × 10-9 mol dm-3 to 2.0 × 10-7 mol dm-3. The results showed no direct relationship between the total 75Se concentration in solution and the dissolved humic substances. After addition of La3+ and removal of the precipitate, the 75Se concentration ranged from 1.5 × 10-9 mol dm-3 to 2.1 × 10-8 mol dm-3. In the presence and the absence of dissolved humic substances, the addition of La3+ resulted generally in the loss of 60 - 90 % of Se from solution, suggesting that Se colloids were present but not particularly associated with OM.

Equilibration of the electro-reduced 75Se0 with synthetic clay water (SCW, no OM) confirmed amorphous and crystalline Se solid phases. Equilibration of the metal scraps coated with electro-deposited 75Se in SCW showed indeed the formation of 2 different solid phases, having measured solubility ranges (after 12 days, centrifugation cut-off ~ 20 nm) approximately from 8.0 × 10-8 mol dm-3 to 2.0 × 10-7 mol dm-3 and from 4.0 × 10-9 mol dm-3 to 5.3 × 10-9 mol dm-3. Hereafter and similarly to the chemically reduced Se phases, half of the electro-reduced Se solid phases were contacted with SCW (+ 2.37 × 10-5 mol dm-3 Fe2+)

226/328

and the other half with organic matter from Boom Clay extract (BCE). Systems were allowed to equilibrate for 2 months before analysing.

Total 75Se concentration in solution, 75Se concentration in solution after 0.01 mol dm-3 La3+ precipitation of dissolved humic substances, optical density (OD) at 280 nm, Eh and pH were measured. The results again showed no important interaction of Se with dissolved humic substances, and concentrations in solution ranged from 3.1 × 10-9 mol dm-3 to 3.1 × 10-8 mol dm-3. After addition of La3+ and centrifugation, the concentrations lowered to the range of 1.2 – 3.3 × 10-9 mol dm-3. Filtration over a 0.02 µm filter yielded a similar low concentration range. The above results indicated again the existence of Se colloids.

A4.5.2 Interaction of elemental Se(s) with Boom Clay The solubility of elemental Se(0) has been study in synthetic and interstitial clay water, and in the presence of pyrite, iron strips, and with solid Boom Clay to vary Eh values and to see its influence on the Se(0) solubility.

A4.5.2.1 AEAT: Interaction of elemental Se(s) with Boom Clay The dissolution of elemental selenium in interstitial Boom Clay water in the presence of Boom Clay was examined. A mass of 1 g of elemental selenium was contacted with 50 cm3 of interstitial Boom Clay water and 2 g portions of Boom Clay. The individual experiments were then sampled after approximately 1, 2 or 3 months’ equilibration and filtered through 0.22 µm, 100 000 NMWCO and 10 000 NMWCO filters. After 35 days the measured selenium concentrations were 2 × 10-5 mol dm-3 after 0.22 µm filtration, 4 × 10-6 mol dm-3 after 100 000 NMWCO filtration and 2 × 10-5 mol dm-3 after 10 000 NMWCO filtration. These increased further to 3 × 10-4, 8 × 10-5 and 2 × 10-4 mol dm-3, respectively, after 75 days and 1 × 10-4, 1 × 10-4 and 9 × 10-5 mol dm-3 after 112 days. Redox potentials were in the range -240 to -290 mV vs SHE and the measured pH values were 9.3 to 9.4.

Thus, the presence of Boom Clay solids has a marked impact on the concentration of selenium in solution compared to the concentrations above elemental selenium in the initial equilibrations in interstitial Boom Clay water (Subsection A4.4.1). This is most pronounced after equilibration for one and two months where the concentrations are one to two orders of magnitude higher in the presence of the clay solids. After three months the difference is only a factor of two. This may suggest that the presence of Boom Clay, or a component thereof, enhances the rate of dissolution of selenium. The selenium concentrations in the presence of Boom Clay are similar to those in the second re-equilibration with interstitial Boom Clay (3 × 10-4 mol dm-3) and in the presence of metallic iron (3 × 10-4 to 7 × 10-4 mol dm-3). Figure 2.5.2 (p. 33, characteristic V-shape curve for the Se(s) solubility) and the discussion in

227/328

Subsection 2.5 showed that the concentration of selenium in solution above elemental selenium may be expected to be very sensitive to redox potential. This may be an explanation for the elevated concentrations found, especially because of the difficulty in measuring redox potentials in poorly-poised laboratory experiments and the possibility of redox disequilibrium.

A4.5.2.2 KULeuven: Interaction of elemental Se(s) with Boom Clay Reduced 75Se solid phases – prepared beforehand – were contacted with solid Boom Clay in order to study the Se behaviour starting from undersaturation. The same systems as described previously (on one hand different chemically reduced 75Se solid phases, on the other hand electro-deposited 75Se) were added to Boom Clay suspensions (two liquid-to-solid ratios of 20:1 and 4.76:1 respectively) and allowed to equilibrate up to two months. The supernatant of the samples was analysed for total 75Se concentration in solution, 75Se concentration in solution after La3+ precipitation of dissolved humic substances, optical density (OD) at

280 nm, Eh and pH.

75Se concentrations in solution ranged from 1.7 × 10-9 mol dm-3 to 8.4 × 10-8 mol dm-3, which was overall lower than the concentrations observed in batch experiments where the same 75Se solid phases were contacted with synthetic clay water (SCW, no OM) and Boom Clay Extract (BCE, with OM). This probably means that either reduced 75Se species interact with the Boom Clay solid phase (and systems are not yet in equilibrium because of kinetic limitations), or that during centrifugation, the reduced-Se colloids were carried away together with the clay particles present in the samples. However, no relation with the organic matter concentration was found, thus providing no evidence for HSe– complexation with Boom Clay humic substances. No extra 75Se removal was observed for increasing amount of solid Boom Clay giving no evidence of sorption.

A4.5.3 Kinetic of reduction of elemental Se(s) The kinetic of reduction/dissolution of elemental Se(0) into HSe– has not been studied in detail. The reduction of Se(0) probably occurred at low Eh values in solubility tests when iron strips were added to the system to lower the redox potential. The HSe– ions released by the reduction/dissolution of Se(0) are expected to precipitate with Fe2+ to form FeSe. This precipitation of a less soluble phase simultaneously increases the dissolution of elemental selenium and the corrosion of metallic iron in the tests of AEAT with iron strips.

A4.5.4 Solubility of elemental Se(s) – AEAT Solubility experiments were made with elemental selenium in synthetic Boom Clay water in the absence of pyrite to evaluate whether the presence of pyrite (Subsection A4.5.1) had any effect on the selenium concentration in solution. In the initial experiment, 1 dm3 of synthetic Boom Clay water was equilibrated with 0.8 g elemental selenium and sampled after one, four,

228/328

eleven and twelve months. Samples taken after one month and four months were filtered through 30 000 NMWCO filters (Nominal Molecular Weight Cut-Off), later samples were filtered through 0.22 µm, 100 000 NMWCO and 10 000 NMWCO filters. The volume of solution removed at each sampling was not replaced.

The concentration of selenium in solution at all samplings was ≤ 1 × 10-7 to 3 × 10-7 mol dm-3

(10 000 and 30 000 NMWCO filtered results) (pH 8.8 to 9.0, Eh = +360 to –230 mV). Because the determinations after one month and three months were below detection limits it was not possible to say definitely that the selenium concentration had reached a steady-state value within one month. However, there was no change in selenium concentration between samples taken after eleven and after twelve months. Filtration through 0.22 µm and 100 000 NMWCO pore filters had no effect on the selenium concentration.

It was thought possible that the surface of the selenium may have been oxidised and that such a layer might determine the concentrations of selenium in solution. To investigate this, the selenium solid was recovered from the experiment, washed and contacted with 40 cm3 of fresh synthetic Boom Clay water in triplicate. After 18 days’ equilibration, the concentrations -8 -7 -3 of selenium in solution were 4 × 10 to 3 × 10 mol dm (pH 9.1 to 9.3, Eh = +360 to +410 mV). These are similar to the concentrations of ≤ 1 × 10-7 to 4 × 10-7 mol dm-3 measured in the presence of crushed pyrite. Comparison with Figure 2.5.2 (p. 33, Chapter 2) suggests redox disequilibrium and poorly-poised experiments. It seems that the presence of pyrite has no effect on the dissolved concentration of selenium above elemental selenium over a timescale of one year. The concentrations measured are in the upper region of the range determined by KULeuven for amorphous elemental selenium.

From these results, a best estimate of the concentration of selenium in solution above elemental selenium would be 4 × 10-8 to 3 × 10-7 mol dm-3 at approximately pH 8 to 9 and under reducing conditions. It is also apparent that filtration in the range 10 000 NMWCO to 0.22 µm has no consistent effect on the concentration of selenium in the aqueous phase. This suggests that any selenium-containing colloids would be > 0.22 µm.

A4.5.5 Conclusions for elemental Se(s) The solubility of elemental Se(0) has been study in synthetic and interstitial clay water, and in the presence of pyrite, iron strips, and with solid Boom Clay to vary Eh values and to see its influence on the Se(0) solubility.

The solubility values determined by AEA Technology with commercial stable elemental selenium appear to be in general one order of magnitude higher than these determined by KULeuven using its own synthesised 75Se labelled Se(0). The Se(0) solubility range of AEAT

229/328

extents from 4 × 10-8 to 3 × 10-7 mol dm-3 while the solubility values determined by KULeuven vary from 1.7 × 10-9 to 8 × 10-8 mol dm-3.

Se(0) solubility value appears not to be influenced by the presence of organic matter. Indeed, no filtration effect is observed by AEAT using different NMWCO ultrafilters. No association Se(0) is also observed by KULeuven using gel permeation chromatography and La3+ precipitation.

The main observation is that the Se(0) solubility appears to increase in the presence of Boom

Clay components or iron strips because it makes the Eh value more negative and Se(0) dissolves as HSe– under strongly reducing conditions as illustrated by the characteristic V-shaped curves on Figure 2.5.2. (See p. 33, Chapter 2: Thermodynamic calculations). The effect of iron is particularly spectacular and SEM examinations have also revealed an important corrosion rate of the iron strip surface in the presence of Se(0) in contrast to FeSe. It appears that elemental Se(0) strongly increases the corrosion rate of iron and vice et versa, according to the following redox reaction:

Fe(0) —> Fe2+ + 2 e– Se(0) + 2 e– —> Se2– Fe + Se —> FeSe

A4.6 Behaviour of selenide: HSe–

The solubility of iron selenide (FeSe) has been studied independently by AEA Technology and KULeuven in synthetic and interstitial clay water, and in the presence of pyrite, iron strips, and with solid Boom Clay. Ultrafiltration tests and gel permeation chromatography are also performed to observe a possible association between HSe– and organic matter.

Iron sulfide (FeSe) appears to be very sensitive to oxidation, as iron sulfide in general. As a consequence, highly soluble oxidation products can interfere with the solubility measurements. Therefore, it is important to first control the purity of the reagents used and to wash FeSe prior to use to remove possible small amounts of very soluble oxidation products.

A4.6.1 Interaction of HSe– with Boom Clay components (pyrite, OM)

A4.6.1.1 AEAT: Interaction of HSe– with Boom Clay components (pyrite, OM) The interaction of iron selenide (FeSe) with crushed pyrite and pyrite coupons was first examined in a synthetic Boom Clay water in the absence of organic material. Sufficient iron

230/328

selenide solid was added to give a selenium concentration of 5 × 10-4 mol dm-3 if dissolved completely. The solution–to–pyrite ratio was 50:1 in the experiments with crushed pyrite.

The concentrations of dissolved selenium above iron selenide in the presence of crushed pyrite were 3 × 10-5 mol dm-3 after one month, 1 × 10-5 mol dm-3 after three months and seven months, and 1 × 10-5 to 3 × 10-5 mol dm-3 after one year at pH 10 and at pH 8. These concentrations were similar to those measured above samples of crushed pyrite equilibrated with synthetic Boom Clay water at pH 10. Sampling of undisturbed experiments after 376 to 379 days equilibration gave selenium concentrations of 9 × 10-6 to 3 × 10-5 mol dm-3 at pH 9.9 and 8 × 10-6 to 1 × 10-5 mol dm-3 at pH 8.6 to 8.7. The concentrations of dissolved selenium in the two experiments with pyrite coupons after 406 days’ equilibration were 1 × 10-5 mol dm-3 (re-sampling after 469 days gave a selenium concentration of 1 × 10-6 mol dm-3) and 6 × 10-6 mol dm-3. Therefore, the aqueous selenium concentrations above iron selenide in the presence of pyrite at pH 8 and 10 remained fairly constant over one year and were in the approximate range 1 × 10-5 to 3 × 10-5 mol dm-3. However, later results (see below Subsection A4.6.4) indicated that these concentrations were controlled by the dissolution of an oxidised layer on the surface of the iron selenide.

XRD analysis of pyrite powder from the experiments showed no bulk changes to the mineralogy and no secondary selenium phases were found. As in samples from the corresponding selenium experiments with crushed pyrite, XRD only identified pyrite. Washing of the pyrite during sampling had separated the pyrite from the majority of the iron selenide. Analysis of a sample of powder recovered from a coupon experiment showed the presence of Fe7Se8 and a trace of elemental selenium. These were also found to be present in the sample of FeSe as supplied and were not formed through interaction with the pyrite.

SEM/EDX analysis of a coupon from these experiments showed the surface to be predominantly pyrite with smaller primary (natural) inclusions of alumino-silicate minerals. Some alumina deposits from polishing of the coupon prior to the experiment were also observed. The surface of the pyrite was not oxidised and looked ‘fresh’, no oxygen was seen in EDX spectra. Small particles of FeSe adhering to the surface of the coupon were seen in the BSEM images (Back scattering Scanning Electron Microscopy), no other selenium compounds were observed.

The dissolution of iron selenide in interstitial Boom Clay water and 10 % interstitial Boom Clay water (obtained by diluting interstitial Boom Clay water with synthetic Boom Clay water) in the absence of Boom Clay, was also examined. A mass of 1 g of iron selenide was contacted with 50 cm3 of water. The individual experiments were then sampled after approximately 1, 2 or 3 months’ equilibration and filtered through 0.22 µm, 100 000 NMWCO and 10 000 NMWCO filters.

231/328

In interstitial Boom Clay water the dissolved concentration of selenium decreased from 5 × 10-6 mol dm-3 after 33 days’ equilibration to 6 × 10-7 mol dm-3 after 96 days with no effect of filter pore size. Redox potentials achieved –220 mV vs SHE in 33 days and decreased further to –300 mV vs SHE by the time of the final sampling with pH values of 9.3 to 9.4. The selenium concentrations in 10 % interstitial Boom Clay water were slightly lower than those measured in interstitial Boom Clay water, again with no effect of filter pore size. After 42 days, the concentration of selenium in solution was 1 × 10-6 mol dm-3, decreasing to ≤ 3 × 10-7 mol dm-3 after 66 days and 5 × 10-8 to 9 × 10-8 mol dm-3 after 95 days. The redox potentials in these experiments were –200 to –220 mV and the pH values were 8.9 to 9.4.

After the initial sampling the experiments in interstitial Boom Clay water were ‘re-started’. The remaining solution was removed and the solids washed with deionised water. Fresh interstitial Boom Clay water was then added. An iron strip was placed in two experiments. The black solution and fine precipitate observed in the ‘re-started’ selenium experiments containing iron strips were not seen in the equivalent iron selenide experiments. The experiments were sampled after 18 to 61 days equilibration and the samples filtered through 0.22 µm, 100 000 NMWCO and 10 000 NMWCO filters. The experiment containing no iron strip was then ‘re-started’ for a further equilibration in the same manner as before and sampled after 45 days’ equilibration.

After re-equilibration with fresh interstitial Boom Clay water, iron selenide gave selenium concentrations of ≤ 3 × 10-7 to 4 × 10-7 mol dm-3 at -200 mV vs SHE after 18 days and 5 × 10-7 to 9 × 10-7 mol dm-3 at -270 mV vs SHE after 45 days. There was no effect of filter pore size. These concentrations are similar to those measured after 96 days’ equilibration in the initial experiments. The selenium concentrations above iron selenide re-equilibrated with fresh interstitial Boom Clay water and an iron strip also showed no effect of filter pore size and were 3 × 10-7 to 1 × 10-6 mol dm-3 at –210 to -250 mV vs SHE. These concentrations are also similar to those measured after 68 to 96 days’ equilibration in the initial experiments.

SEM showed that iron selenide removed at the end of the experiments with the iron strips had a similar appearance to ‘as-supplied’ iron selenide analysed at the same time, both showed some evidence of oxidation, with Fe, Se and O present. However, in addition, the ‘as-supplied’ iron selenide showed a small amount of small highly-oxidised needle crystals, which were not present in the recovered iron selenide. The iron strips from the experiments showed no surface coating and EDX only showed the presence of iron.

232/328

A4.6.1.2 KULeuven: Interaction of HSe– with Boom Clay components (pyrite, OM) 75 0 0 Reduced metal – selenide solid phases were synthesised using Na2S + Fe(II), Fe and Zn as chemical reductants.

The iron-selenide precipitates were first contacted with synthetic Boom Clay water (SCW) (+ 2.37 × 10-5 mol dm-3 Fe2+) to measure the solubility of the prepared 75Se reduced solid phases. The solubility was found to be dependent on the reducing system and the solid phase. 75Se concentrations in solution were spread in general in a lower range than the solubilities measured above 75Se0 precipitates, i.e. 4.5 × 10-10 mol dm-3 – 4.3 × 10-9 mol dm-3. After removal of the supernatant SCW solution, half of each solid phase was re-contacted with fresh SCW (no OM) (+ 2.37 × 10-5 mol dm-3 Fe2+) and the other half was contacted with Boom Clay extract (BCE) containing OM. The systems were allowed to equilibrate for 2 months and were then analysed for: total 75Se concentration in solution; 75Se concentration in solution after precipitation of dissolved humic substances using La3+; optical density (OD) 75 -9 -3 at 280 nm; Eh and pH. The total final se concentrations ranged from 1.1 × 10 mol dm to 1.4 × 10-8 mol dm-3. The results showed no direct relationship between the total 75Se concentration in solution and the dissolved humic substances. After addition of La3+ and removal of the precipitate, the 75Se concentration ranged from 5.4 × 10-10 mol dm-3 to 4.0 × 10-9 mol dm-3. In the presence and the absence of dissolved humic substances, the addition of La3+ resulted generally in the loss of 70 - 95 % of Se from solution, suggesting that Se colloids were present but not particularly associated with OM.

A4.6.2 Interaction of HSe– with Boom Clay

A4.6.2.1 AEAT: Interaction of HSe– with Boom Clay The dissolution of iron selenide in interstitial Boom Clay water in the presence of Boom Clay was examined in a similar manner to elemental selenium (Subsection A4.5.2). A mass of 1 g of iron selenide was contacted with 50 cm3 of interstitial Boom Clay water and 2 g portions of Boom Clay. The individual experiments were then sampled after approximately 1, 2 or 3 months’ equilibration and filtered through 0.22 µm, 100 000 NMWCO and 10 000 NMWCO filters. After 36, 74 and 112 days the selenium concentrations in solution were 1 × 10-6 to 2 × 10-6 mol dm-3 with no significant trend with the pore size of the filters. The measured redox potentials were –220 to –240 mV. It is noted that these concentrations may be determined by the presence of a soluble surface layer on the iron selenide and lie within the range 6 × 10-7 to 5 × 10-6 mol dm-3 measured in the experiments in interstitial Boom Clay water.

233/328

A4.6.2.2 KULeuven: Interaction of HSe– with Boom Clay Reduced 75Se solid phases – prepared beforehand – were contacted with solid Boom Clay in order to study the selenium behaviour starting from undersaturation. The same systems as described previously (different chemically reduced 75Se solid phases) were added to Boom Clay suspensions (two liquid-to-solid ratios of 20:1 and 4.76:1 respectively) and allowed to equilibrate up to two months. The supernatant of the samples was analysed for total 75Se concentration in solution, 75Se concentration in solution after La3+ precipitation of dissolved humic substances, optical density (OD) at 280 nm, Eh and pH.

75Se concentrations in solution ranged from 8.2 × 10-10 mol dm-3 to 1.9 × 10-9 mol dm-3, which was overall lower than the concentrations observed in batch experiments where the same 75Se solid phases were contacted with synthetic clay water (SCW, no OM) and Boom Clay Extract (BCE, with OM). This probably means that either reduced 75Se species interact with the Boom Clay solid phase (and systems are not yet in equilibrium because of kinetic limitations), or that during centrifugation, the reduced-Se colloids were carried away together with the clay particles present in the samples. However, no relation with the organic matter concentration was found, thus providing no evidence for HSe– complexation with Boom Clay humic substances. No extra 75Se removal was observed for increasing amount of solid Boom Clay giving no evidence of sorption.

A4.6.3 Solubility of HSe– – AEAT Solubility experiments were made with iron selenide in synthetic Boom Clay water in the absence of pyrite to evaluate whether the presence of pyrite (Subsection A4.6.1) had any effect on the selenium concentration in solution. In the initial experiment, 1 dm3 of synthetic Boom Clay water was equilibrated with 1.34 g iron selenide and sampled after one, four, eleven and twelve months. Samples taken after one month and four months were filtered through 30 000 NMWCO filters (Nominal Molecular Weight Cut-Off), later samples were filtered through 0.22 µm, 100 000 NMWCO and 10 000 NMWCO filters. The volume of solution removed at each sampling was not replaced.

Over the whole duration of the experiment the concentration of selenium in solution was 8 × 10-6 to 2 × 10-5 mol dm-3 with no trend with time. Results for the later filtrations showed that filter pore size had no effect on the selenium concentration. These results are consistent with the data obtained in the presence of crushed pyrite (Subsection A4.6.1).

It was thought possible that the surface of the iron selenide may have been oxidised and that such a layer might determine the concentrations of selenium in solution. To investigate this, the iron selenide solid was recovered from the experiment, washed (the solid is thus ‘pre-leached’) and contacted with 40 cm3 of fresh synthetic Boom Clay water in triplicate.

234/328

Solution samples taken after 21 days’ equilibration gave concentrations of selenium in -8 -8 -3 solution of ≤ 3 × 10 to 5 × 10 mol dm (pH = 9.1, Eh = –220 to –270 mV). Thus, it seems that there was a trace of a soluble (probably oxidised) layer on the surface of the iron selenide as supplied and that this was responsible for the initial measured selenium concentrations. It should be noted that the XRD of the iron selenide solid recovered from the coupon experiment showed the presence of Fe7Se8 and a trace of elemental selenium. It is possible that these phases could remain in the ‘pre-leached’ solid and may control the aqueous selenium concentration. In such a case, the ‘true’ solubility of FeSe could be lower than that measured here.

At present the best estimate of the concentration of selenium in solution above FeSe from these results is ≤ 3 × 10-8 to 5 × 10-8 mol dm-3 at pH 9 and –200 to –300 mV. Although such values are consistent with Figure 2.5.1 (p. 32, Chapter 2), they are above the range of values measured by KULeuven (4.5 × 10-10 to 1.4 × 10-8 mol dm-3 in synthetic clay water (SCW, no OM) and Boom Clay Extract (BCE, with OM), including possible colloidal contribution).

A4.6.4 Conclusions for HSe– The solubility of FeSe has been study in synthetic and interstitial clay water, and in the presence of pyrite, iron strips, and with solid Boom Clay to vary Eh values and to see its influence on the FeSe solubility.

The first solubility values determined by AEA Technology with commercial FeSe appeared to be very high: 8 × 10-6 to 2 × 10-5 mol dm-3. The reason was that the surface of iron selenide was oxidised and that the resulting soluble layer determined the concentration of selenium in solution.

Therefore, the FeSe precipitate was pre-leached to remove the soluble impurities and the solubility experiments restarted with the washed samples. The new FeSe solubility measurements ranged then two or three orders of magnitude lower: from 3 × 10-8 to 5 × 10-8 mol dm-3. The solubility values determined independently by KULeuven were even lower and varied from 4.5 × 10-10 to 1.4 × 10-8 mol dm-3.

FeSe solubility value appeared not to be influenced by the presence of organic matter (OM). Indeed, no filtration effect was observed by AEAT using different NMWCO ultrafilters. No association between HSe– and OM was observed by KULeuven using gel permeation chromatography and La3+ precipitation.

As a conclusion, it is mandatory to check the purity of FeSe to avoid oxidation artefacts, leading to the presence of soluble salts imposing a selenium concentration well above the true

235/328

solubility limit. SEM images are also useful to control the presence of small amounts of contaminants below the limit of detection of XRD.

236/328

A5. Immobilisation of selenium in the near-field

237/328

238/328

A5 Immobilisation of selenium in the near-field

A5.1 Immobilisation of selenium in cementitious buffer

In the present Belgian concept for deep disposal of high-level waste, spent fuel elements and HLW glass canisters are surrounded by carbon steel overpacks protected against corrosion by high pH of cement (SuperContainer concept: ONDRAF/NIRAS, 2002). Because the thickness of the SuperContainer (~ 0.7 m) is small with respect to that of the clay formation (30 to 50 m according to the selected scenario), the contribution of the SuperContainer to the retention of selenium is presently not considered.

Several authors (Ochs et al., 2002; Baur I., 2002; Baur and Johnson, 2003a,b; Bonhoure et al., 2006) describe the immobilisation of selenite and selenate along with different cement phases (CSH, ettringite, AFt, AFm, carbonate, …) by coprecipitation, formation of solid solutions and sorption. Formation of solid solutions of poorly soluble selenate phases (selenate taking the place normally occupied by sulfate in ettringite, AFt and AFm), or sorption of selenite (inner-sphere complex) on various cement minerals (layered double hydroxides, LDH’s, e.g., hydrotalcite, bearing positively charged sites) are important processes for the immobilisation of selenium in cement. Calcium carbonate and calcium silicate hydrates (CSH) can also scavenge selenite as they do with nitrate a planar trigonal oxyanion isoelectronic with bicarbonate and of about the same ionic radius (De Cannière, 1989, Ph.D. Thesis). Similarly, selenate can exchange with sulfate or silicate and be incorporated in the crystal lattice of ettringite, AFt, AFm and residual gypsum during recrystallisation and ageing of cement.

A5.2 Immobilisation of selenium by alteration and corrosion products

A5.2.1 Uptake of selenium by spent fuel degradation products

Finally, in the near-field of a repository, corrosion and alteration products of spent fuel matrix, metal containers, and altered minerals in the oxidized zone could also contribute to somewhat delay the release of selenium towards the far-field. Indeed, the retention of 79Se on

UO2 is observed in spent fuel alteration products (Trombe et al., 1985; Chen et al., 1999, 2000; Ewing, 2001). This is corroborated by the natural association between uranium and selenium occasionally found by Deliens et al. (1981) in some uranium-bearing seleniferous copper sulfide. Vochten et al. (1996) also discovered and first determined the structure of piretite, Ca(UO2)3(SeO3)2(OH)2 · 4 H2O, a new calcium uranyl selenite found in oxidized uranium ores collected in a weathered zone of the Shinkolobwe mine (Katanga, Congo) (see also Appendix A2 on Natural selenium in the environment and in Boom Clay).

239/328

A5.2.2 Sorption of selenium by iron oxy-hydroxides

Furthermore, the sorption and reduction of selenite could occur onto corrosion products and green rust (or fougerite, a recently discovered LDH mineral) in the presence of hydrogen produced by iron anaerobic corrosion (Spahiu et al., 2000; Scheidegger et al., 2003; Cui et al., 2006). Sorption of selenite is also possible onto iron oxy-hydroxide produced by pyrite oxidation in the excavation disturbed zone (EDZ) around the galleries during the construction, operation and ventilation phases of the repository. Then, the main uncertainty concerns the nature of the redox-controlling phases in the near-field which could be reducing or locally oxidizing depending on the geochemical processes at work or the repository concept (ventilated galleries, or not; post-closure phase history, …).

240/328

A6. Selenium background in bentonite buffer materials

241/328

242/328

A6 Selenium background concentration in bentonite buffer materials

A6.1 Introduction

The studies of the leaching behaviour of radionuclides in backfill materials has been extended to 79Se, considered now as a mobile fission product critical for safety calculations, with the aim to determine 79Se steady-state concentration under near-field conditions. In the past, no attention was given to selenium in previous glass compatibility programmes.

A6.2 Experimental

The experimental programme consisted of leaching tests with SON68 and SM539 glasses doped with inactive Se. Leaching tests were performed at a temperature of 40 °C with glass powder with a surface area to volume ratio (SA/V) equal to 2 500 m-1 in order to increase the build-up of selenium concentration in the leachate. The glass was added as powder (125 - 250 µm fraction). The test medium ('Int-RIC') was a mixture with nominally 712.5 g

(wet) Boom Clay, 712.5 g (air-dry) M2 backfill mixture, 37.5 g (air-dry) Fe3O4 powder, and 37.5 g (air-dry) AISI 316 L stainless steel powder per litre of Boom Clay porewater. The M2 backfill mixture consists of 65 % FoCa-Clay, 30 % quartz sand and 5 % graphite (Timrex). The durations of the tests were 90, 180, 365, 540 and 720 days. The leaching tests were performed in duplicate.

Analyses and interpretation focus on the determination of the mobile Se concentration in the leachates obtained after ultrafiltration over membranes of 10 000 and 100 000 MWCO. The Se concentrations in the leachate after ultrafiltration were analysed by ICP-MS (analytical service of SCK•CEN). Because of difficulties to detect Se in the 540 days sample, high- resolution ICP-MS was used for the 720 days sample (Royal Museum of Central Africa, department of geology at Tervuren, Belgium).

A6.3 Results

The pH of the leachates slightly increases as a function of time from 7.9 to 8.3. The redox potential (Eh) measured while gently stirring the solution was negative varying from -41 to -268 mV (SHE) as expected from the characteristics of the reducing medium.

Using inactive isotopes of Se for practical reasons, no distinction is possible between the amount of Se leached from glass and selenium already present in the medium before dissolution of glass. An analysis of the initial medium was necessary to determine the

243/328

background concentration of Se in the solid and the liquid fractions of the medium (without glass addition). However, the determination of the background concentrations of selenium was done on an equivalent medium called Int-SIC2 and not on Int-RIC. The Se concentration in the solid fraction is given in the third column of Table A6.3.1. We assumed that the amount of Se measured in Int-SIC was representative of this present in Int-RIC. For the liquid fraction consisting of synthetic clay water, the result of the analysis is only indicative. Nevertheless, the Se concentration was also measured in the liquid fraction of Int-RIC by means of two blank tests (Table A6.3.2). Table A6.3.2 also gives the amount of Se in glass per g of the medium allowing a comparison of the amount of Se present already in the medium and the maximal amount which could be leached from glass. The amount of Se is higher in the background medium than in the whole glass.

Table A6.3.1: Background concentrations of Se in Int-SIC (mol dm-3). [Se] Solid clay medium μg Se in the glass Int-SIC* per g Int-SIC# (mol dm-3) (μg Se/g Int-SIC) (µg/g Int-SIC)

< 6.33 × 10-7 < 50 7.96

the uncertainty on the result is about 20 %. * the solutions were only filtered with a filter of 0.45 µm. # this amount is obtained by multiplying the amount of selenium as determined by ICP-MS after dissolution of the glass in µg/g glass by the amount of glass added in the experiment (g) and dividing then this value by the number of g of Int-SIC added.

Table A6.3.2: Background concentrations of Se in Int-RIC (mol dm-3) after 365 and 720 days in blank tests for two filtered fractions. Initial [Se] (Int- Solution (blank: Solution (blank: Solution Solution (blank: SIC) 365 d) 365 d) (blank: 720 d) 720 d) (0.45 μm) (0.45 μm) (YM10) (0.45 μm) (YM10) (mol dm-3) (mol dm-3) (mol dm-3) (mol dm-3) (mol dm-3) < 6.33 × 10-7 3.17 × 10-7 < 2.53 × 10-7 1.13 × 10-8 4.43 × 10-8

0.45 μm: after micro-filtration through a 0.45 μm membrane. YM 10: ultrafiltration through 10 000 MWCO membrane.

There is no tendency in the evolution of the Se concentrations as a function of time for the blanks. Because there is no simple way to select the background Se concentration in the starting medium, we decided to consider a range of concentrations taking the lowest and the highest value obtained after ultrafiltration through 10 000 MWCO membranes. This range of concentrations is used for comparison with the leachates concentrations of Se.

2 712.5 g (wet) Boom Clay, 712.5 g (air-dry) M2 backfill mixture, 37.5 g (air-dry) Fe3O4 powder, and 37.5 g (air-dry) AISI 316 L stainless steel powder per liter Synthetic Clay Water.

244/328

The Se concentration is stable with time as illustrated in Figure A6.3.1, around 2 × 10-7M for both types of glasses. The two glasses have a different behaviour regarding their boron release but the type of glass has obviously no influence on the Se concentrations measured in the leachates. There is no difference in concentration level before and after ultrafiltration through 10 000 or 100 000 MWCO membranes. The Se concentration measured in the leaching tests is in the range of the background concentrations measured in the initial backfill medium (red dashed line). So, the small amount of Se leached out of the glass cannot be easily distinguished from the total amount of Se already present in the near-field medium in contact with the glass. The Se speciation in the medium was not measured.

1E-6 1E-6

1E-7 1E-7

1E-8 1E-8

[Se] (mol/l) [Se] (mol/l) [Se] 1E-9 1E-9

SON68 SM539 1E-10 1E-10 0 100 200 300 400 500 600 700 800 0 100 200 300 400 500 600 700 800 Duration (days) Duration (days)

Figure A6.3.1: Concentration of Se in the leachates after ultrafiltration for different test durations for the SON68 and SM539 glasses. The background Se concentration in the medium is represented by dashed lines.

A6.4 Comparison of Se concentrations in near-field and far-field conditions

The values of the Se concentrations measured in the leaching tests (near-field) are higher than the Se concentrations measured in Boom Clay in percolation tests (far-field) and thermodynamical solubility values used in performance assessment (PA) for the selenium in the far-field (Table A6.4.1). The high Se concentration in the leaching tests is due to the Se content of the medium representing the near-field.

Selenium is a redox-sensitive element whose behaviour under reducing conditions can be difficult to interpret because of redox-disequilibrium. The Se speciation in the source term, i.e. in the glass, is unknown. If, in the far-field, the expected thermodynamic stable phases are 0 2– elemental Se , FeSe, FeSe2, in the HLW glass and in our laboratory tests, a mixture of SeO4 , 2– SeO3 are likely to be present, increasing the total Se concentration measured in solution. The near-field Se concentrations are indeed higher that what we should expect from the solubility

245/328

2– of the reduced forms of Se. The selenate, SeO4 , does not easily reduce to selenite and is unretarded and not solubility-limited in Boom Clay. The kinetics of reduction are extremely 2– slow for SeO4 which could also be present in the waste form and so selenate might remained unreduced in Boom Clay. The measured Se concentrations being higher than the solubility value used by PA or the Se concentration determined in the Boom Clay water ([Se] = 2.4 × 10-8 mol dm-3 measured by high resolution ICP-MS) and through percolation tests, it is 2– possible that Se speciation in the leaching tests was dominated by SeO4 . However, Se speciation measurements were performed, so no conclusions can be drawn.

Table A6.4.1: Comparison of selenium concentrations obtained from these glass leaching tests with values measured in the Boom Clay water, or used in PA calculations. Type of study (+ reference) [Se] (—) (mol dm-3)

Near-field medium (these leaching tests) 2.53 × 10-7

Thermodynamic calculations with: Se (0) as solubility limiting solid phase -8 – Best estimate PA values (DCF) 5.50 × 10 Marivoet et al., (1999) -9 – PSI calculations: Berner (2002) 5.30 × 10

Far-field: Boom Clay Migration experiments 5.00 × 10-9

A6.5 Conclusions

We conclude that the Se concentrations leached from glass powders are rather low ([Se] = 2.5 × 10-7 mol dm-3) and not significantly higher than Se background concentrations in the simulated near-field medium. Selenium concentration has likely reached a steady-state concentration in the near-field medium. The small selenium amount released by glass dissolution does not significantly increase the overall Se concentration already present in the near-field medium. Furthermore, there is no difference in Se concentration level after different types of filtrations.

Due to the great sensitivity of selenium on redox conditions and view the uncertainty on selenium redox state in our tests, it is difficult to give any conclusion for the selenium behaviour and its final concentration in the near-field medium. It is rather clear that information on the Se speciation in the glass would be worthy to study as Se is a relevant and sensitive element for PA.

246/328

A6.6 References

Berner U. (2002) Project Opalinus Clay: radionuclides limits in the near-field of a repository for spent fuel and vitrified HLW. PSI Report (PSI Bericht Nr 02-22, ISSN 1019-0643) (December 2002).

Marivoet J., Volckaert G., Labat S., De Cannière P., Dierckx A, Kursten B., Lemmens K., Lolivier P., Mallants D., Sneyers A., Valcke E., Wang L., and Wemaere I. (1999a) Values for the near-field and clay parameters used in the performance assessment of the geological disposal of radioactive waste in the Boom Clay formation at the Mol site (volume 1 and 2). Report to NIRAS/ONDRAF. Geological disposal of conditioned high-level and long-lived radioactive waste. Contract CCHO-98/332 – KNT 90.98.1042 Task 6.1. SCK•CEN Report R-3344 (July 1999).

Pirlet V. (2005) Determination of the mobile leached concentrations of 79Se in near-field conditions. In: Characterization and compatibility with the disposal medium of Cogema and Eurochemic reprocessing waste forms (Tasks VM-6 and GV8 of NIRAS/ONDRAF contracts CCHO-90/123-1 and CCHO-90/123-2 – vitrified waste). The importance of the glass composition and the near-field for the mobile concentrations of radionuclides. Summary of the topical report for WP4 of RP.WD.008 (January 2000 – June 2003).

247/328

248/328

A7. Sorption behaviour of selenite, selenate and sulfate on Fe and Al oxide surfaces

249/328

250/328

A7 Sorption behaviour of selenite, selenate and sulfate on Fe and Al oxide surfaces

The aim of this section is to discuss the general sorption mechanism of selenium oxyanion and sulfate at the water-oxide interface to explain and to predict their behaviour under natural conditions in clay. No pure iron and aluminium oxides are observed in Boom Clay, but it contains a large proportion of illite and interstratified illite/smectite clay minerals. Aluminium hydroxyl groups present in the octahedral layer are accessible on the surface of the lateral edges of these clay minerals. These groups are Lewis acid sites as those found in pure aluminium oxides, so they could be sorbent sites for oxyanions (Lewis bases).

2– The selenate anion (SeO4 ) is the thermodynamically stable selenium species under oxidizing conditions and adsorbs very weakly, or even not, to metal oxide surfaces, depending on pH 2– conditions. In contrast, the selenite anion (SeO3 ) is stable under slightly suboxic conditions and binds strongly to metal oxide surfaces.

Solid-water partitioning reactions on the iron oxide-water system are very well studied because iron oxyhydroxide is a very effective sorbent for many cations and anions and it is a relatively simple and convenient system for experimental and modelling works (Dzombak and Morel, 1990). Selenium oxyanions are good example of contrasted sorbing species. Indeed, selenite and selenate sorbed to iron oxide respond quite differently to changes in the ionic strength of the aqueous phase over a wide range of pH values. The selenite ion belongs to the general class of adsorbing ions that have a strong affinity for oxide surface hydroxyl sites and whose adsorption is relatively unaffected by changes in ionic strength. The selenate ion is a representative of the second class of adsorbing ions that bond weakly to such sites and whose adsorption is markedly reduced by increasing ionic strength. These ionic strength effects suggest that strongly bonded ions form inner-sphere coordination complexes with oxide surface oxygen's, whereas the more weakly bonded ions form outer-sphere, ion-pair complexes that retain their primary hydration sphere upon adsorption.

A strict definition and a clear distinction between inner- and outer-sphere surface complexes is important, particularly in the case of the adsorption mechanism of the weakly sorbing selenate which will be discussed later in the text. Sposito et al., (1999) give the following definition in a review on the surface geochemistry of the clay minerals: “An inner-sphere surface complex has no water molecule interposed between the surface functional group and the ion or molecule it binds whereas an outer-sphere surface complex has at least one such interposed water molecule. Outer-sphere surface complexes thus comprise solvated adsorbed ions”.

251/328

A7.1 EXAFS studies of selenite and selenate adsorption on goethite

X-ray absorption spectroscopy (XAS) provides direct structural information for adsorbed species at solid-liquid interfaces. Unlike infrared and Raman spectroscopy, XAS yields relatively accurate, direct measurements of average interatomic distance, coordination number, and the type of coordinating ligand for the nearest neighbours of the x-ray absorbing atom.

A7.1.1 Classical view of inner-sphere and outer-sphere surface complexes

Hayes et al. (1987) have performed in situ extended x-ray absorption fine structure (EXAFS) measurements of adsorbed selenate and selenite ions at the α-FeO(OH) (goethite) – water interface. Their observations indicate that selenate behaves as a weakly bonded, outer-sphere complex while selenite forms a strongly bonded, inner-sphere complex. The selenite ion is bonded directly to the goethite surface and creates a bidentate bridge with two iron atoms as illustrated on Figure A7.1.1 reprinted from Hayes (1987). Adsorbed selenate has no iron atom in the second coordination shell of selenium, which indicates that selenate retains its hydration sphere upon sorption.

2– Figure A7.1.1: Possible molecular structures for selenite (SeO3 ) coordinated with Fe atoms at the goethite surface: (a) bidentate chelating a single Fe nucleus; (b) monodentate mononuclear Fe association; (c) bidentate bridging two Fe nuclei (bidentate, binuclear complex). Reprinted from Hayes (1987) Ph.D. thesis, Stanford University; figure taken from Davis and Kent (1990).

252/328

A7.1.2 Evidence of inner-sphere complexes also implied for weakly sorbing species

Surprisingly Manceau and Charlet (1994) observed with EXAFS spectroscopy in their pH conditions that the weakly sorbing selenate ion forms a binuclear bidentate inner-sphere surface complexes on goethite and hydrous ferric oxide (HFO). This unexpected finding is in total contrast with the results of Hayes et al. (1987) indicating that selenate keeps its primary hydration shell when sorbed onto goethite and forms an outer-sphere (OS) surface complex. This observation also put in question the relevance of the ionic strength dependence of sorption isotherms as a criterion used to support the formation of outer-sphere surface complexes. Indeed, the effect of ionic strength on cation and anion adsorption on oxide surfaces does not provide reliable information on the exact way they are bound to mineral surfaces, but it certainly remains a good indication of their relative bonding affinity for surface hydroxyl groups.

How to explain this paradox observation affecting the sorption mechanism of a weakly sorbing species as selenate and how to reconcile the results of Hayes et al. (1987) with these of Manceau and Charlet (1994) ? Sposito (1984) previously yet suggested that anions with 2– 2– weak sorption affinity, such as SeO4 and SO4 , sometimes sorb as an outer-sphere complex but occasionally also as an inner-sphere complex on soil mineral surfaces. The key for understanding the sorption mechanism of weakly sorbing species likely resides in the pH effect in the formation of inner-sphere complexes. It will be discussed hereafter at the light of recent results of Raman and infrared spectroscopy with selenate and sulfate on Fe and Al- oxides.

A7.2 ATR-FTIR studies of selenate and sulfate adsorption on Fe and Al (hydr)oxide

2– 2– The in situ vibrational spectroscopy studies of SeO4 and SO4 adsorbed on Fe and Al (hydr)oxides provides also valuable information for the identification of the types of surface complexes of these anions on these minerals. The combined Raman and attenuated total 2– 2– reflectance Fourier-transform infrared (ATR-FTIR) spectral data show that SeO4 and SO4 have intermediate complexation behaviour of forming both inner- and outer-sphere surface complexes. The relative importance of the type of surface complex depends on the pH and type of mineral.

A7.2.1 Sulfate adsorption on Fe oxyhydroxide

The mechanism of sulfate adsorption on aqueous interface of various iron oxides was studied in situ at different pH and ionic strength by several authors using attenuated total reflectance

253/328

Fourier transform infrared (ATR-FTIR) spectroscopy. Hug (1997) investigated the sulfate adsorption on hematite (α-Fe2O3) in the presence of an aqueous phase between pH 3 and 5. He obtained infrared spectra consistent with the formation of a monodentate sulfate inner- sphere complexes on the wet hematite surface. However, spectral data suggest that after drying of the samples a bidentate inner-sphere complexes could also be formed with the surface Fe(III) sites. Peak et al. (1999) examined the mechanism of sulfate adsorption on goethite (α-FeOOH). They concluded that sulfate forms both outer-sphere and weak inner- sphere surface complexes on goethite at pH less than 6. At pH values greater than 6, sulfate adsorbs on the goethite only as an outer-sphere complex. The relative amount of outer-sphere sulfate surface complexation also increased with decreasing ionic strength.

A7.2.2 Selenate and sulfate adsorption on Fe and Al (hydr)oxide

2– 2– The coordination and speciation of selenate (SeO4 ) and sulfate (SO4 ) on goethite and Al oxide were studied by Wijnja and Schulthess (2000) using Raman and ATR-FTIR spectroscopy. This study also indicates that both inner- and outer-sphere surface complexes of 2– 2– SeO4 and SO4 occur on these metal (hydr)oxide surfaces. The spectral data show that 2– 2– SeO4 and SO4 have a similar complexation behaviour on the same adsorbent. On goethite, these oxyanions predominantly form weak monodentate inner-sphere surface complexes at pH < 6, while at pH > 6 they are essentially present as outer-sphere surface complexes. In contrast, on Al oxide, these anions exist mainly as weaker outer-sphere surface complexes, but a small fraction is also present as inner-sphere complex at pH < 6. A comparison of the spectral intensities of these anions on goethite and Al oxide shows that complexation of these anions with Al oxide is weaker than with Fe oxide.

2– Controversy still exists about the molecular structure of SeO4 surface complexes formed on hydrous metal oxides. Sposito (1984) already indicated that anions with a weak sorption 2– 2– affinity, such as SO4 and SeO4 , could exhibit a versatile behaviour between outer- and inner-sphere complexes on mineral surfaces. The macroscopic observation of decreasing 2– 2– SeO4 and SO4 adsorption on metal oxides with increasing concentration of the background electrolyte has commonly been interpreted as an indication for the formation of an outer- sphere surface complex of these anions (Hayes et al., 1987). However, the macroscopic nature of the ionic strength effects on anion adsorption is only an indirect technique for evaluation of sorption mechanisms at mineral surfaces. Conversely, the absence, or the presence, of a water molecule in the coordination shell of a surface complex is likely also not an entirely sufficient argument to conclude on the strength, or at the contrary the weakness, of the character of the bonding of a weakly sorbing species to a mineral surface.

254/328

Figure A7.2.1 taken from Hayes et al. (1988) and reprinted by Davis and Kent (1990) shows the classical anion “sorption edge” curves of selenite and selenate anions on ferrihydrite. The fraction of species adsorbed is represented as a function of pH. The batch sorption experiments were performed at three different values of the ionic strength: 0.013 M, 0.1 M, and 1 M NaNO3.

Figure A7.2.1: Effect of pH and NaNO3 concentration on Se(VI) and Se(IV) adsorption by 0.001 M Fe (as ferrihydrite), total Se = 10-4 M in each experiment. Data from Hayes et al. (1988). Se(VI): ×-boxes, 0.013 M NaNO3; open squares, 0.1 M NaNO3; filled squares, 1.0 M NaNO3; Se(IV): open-circles, 0.013 M NaNO3; filled circles, 1.0 M NaNO3.

As observed on Figure A7.2.1, selenate [Se(VI)] is much less sorbed than selenite [Se(IV)] at pH < 8 and selenate sorption is quasi nil at pH > 8. The selenate sorption is strongly affected by the concentration of the background electrolyte and decreases at high concentration of

NaNO3. Selenite sorption already starts at pH ~ 11 and gradually increases at lower pH. The effect of ionic strength is much less pronounced for selenite, but surprisingly, the selenite sorption even slightly increases at higher NaNO3 concentration.

A7.2.3 Formation of outer-sphere versus inner-sphere surface complexes

The non-specific sorption of oxyanions due to single electrostatic interactions is explained by the acido-basic functions of the hydroxyl groups present at the surface of iron oxides. At high pH when the ≡S–OH surface groups are largely deprotonated as ≡S–O– the electrostatic repulsion hinder, or even inhibit, the sorption of anions. As the pH is lowered below the zero-

255/328

point-of-charge (ZPC) of the hydroxylated surface, this latter becomes progressively positively charged by the protonation of the ≡S–OH groups. The anions adsorption increases + thus at low pH as the population of the ≡S–OH2 groups grows. The different variants of the electrical double layer (EDL) theory allow to explain the effect of the ionic strength on this non-specific electrostatic sorption: this latter decreases when the double layer is compressed at high ionic strength and that its thickness decreases (cfr., e.g., Davis and Kent, 1990).

The formation of both outer- and inner-sphere surface complexes on goethite is also treated by Peak et al. (1999) based on the pH-dependent protonation of the surface groups. It is + proposed that inner-sphere surface complexes are formed by ligand exchange with ≡Fe-OH2 groups, which are predominant at lower pH values. Indeed, a typical ligand exchange reaction for forming a monodentate mononuclear inner-sphere surface complex with selenite in neutral condition

2– – – SeO3 + ≡Fe—OH ⇌ ≡Fe—OSeO2 + OH (eq. A7.2.1) ligand neutral inner-sphere hydroxyl ion (selenite) surface complex (leaving group)

can be rewritten by adding a proton to the hydroxylated surface, so that a H2O molecule is expelled in the nucleophilic substitution reaction in place of an OH– ion:

2– + – SeO3 + ≡Fe—OH2 ⇌ ≡Fe—OSeO2 + H2O (eq. A7.2.2) ligand protonated inner-sphere water (selenite) surface complex (leaving group)

This basic example with a simplified ligand-exchange reaction explicitly implies the removal of one –OH group from the Fe(III) sites. As H2O molecules are better “leaving groups” than naked OH– hydroxyl ions, a pH lowering favours the ligand exchange process. At neutral pH + values and higher, the number of Fe-OH2 groups easily available for ligand exchange is much lower and the ligand exchange becomes more difficult. As the pH continues to raise, the formation of inner-sphere complex is progressively impeded and inhibited: only much weaker non-specific electrostatic bindings, or hydrogen bonds, remain possible through the formation of an outer-sphere complex with oxyanions. In their turn, these latter becomes more and more + rare as the pH increases and the population of ≡S–OH2 sites decays.

2– + + 2– SeO4 + ≡Fe—OH2 ⇌ ≡Fe—OH2 ··(H2O)SeO4 (eq. A7.2.3) ligand protonated outer-sphere (selenate) surface complex

256/328

A7.2.4 Effect of the nature of the mineral surface

The nature of the mineral surface also plays an important role on the strength of the sorption and on the ratio between the formation of inner- and outer-sphere complexes. Hydrous ferric oxide (HFO) are known to be stronger sorbent than aluminium oxide, themselves much stronger than silica surface. The surface reactivity decreases in the following order:

≡Fe–OH > ≡Al–OH >>> ≡Si–OH.

The predominance of weaker outer-sphere surface complexes on Al oxide may be the result of a lower complexation affinity of Al with oxyanions compared to that of Fe. Indeed, Al-sites are harder Lewis acids than Fe(III)-sites (see the hard soft acids bases (HSAB) principle explained in pp. 75–77 by Sposito, 1981) and they could more strongly retain their hydroxyl groups (hardest base). As a consequence the ligand exchange with oxyanions is hindered and the oxyanion sorption weaker. In a similar way, a lower sorption affinity is also expected for the aluminol groups present on the octahedral gibbsite layer accessible at the clay mineral edges. The proportion of the lateral surface of octahedral layer accessible in clay minerals is also much lower than in pure aluminium oxide and the proximity and steric hindrance of the two adjacent silica tetrahedral layers could also restrain the surface complexation reactions (see Figure A7.2.2).

Figure A7.2.2: Structure of 1:1 and 2:1 layer clay minerals with the ≡S–OH surface groups accessible on the lateral edges of the octahedral (gibbsite) and tetrahedral (silica) layers. Slightly modified from Sposito et al. (1999).

257/328

To conclude, in Boom Clay, only selenite is expected to form relatively strong inner-sphere surface complexes with aluminium hydroxyl groups located on the edges of illite and smectite clay minerals. Selenate is not assumed to significantly sorb onto clay minerals in Boom Clay.

Bruggeman (2006) has experimentally confirmed these assumptions by means of batch sorption tests (made with selenate and selenite on fresh Boom Clay suspensions under anaerobic conditions) and independent EXAFS measurements (with selenite on sodium- conditioned illite du Puy). No sorption is observed for selenate in Boom Clay suspensions while batch sorption tests and EXAFS show a significant sorption of selenite on illite with the formation of inner-sphere complexes.

However, recent diffusion experiments with sulfate in undisturbed compact Boom Clay cores have provided a (one order of magnitude) lower that expected average value for the apparent -11 2 -1 diffusion coefficient of sulfate (Dapp = 3.2 ± 1.0 × 10 m s ) while the corresponding diffusion accessible porosity (ηR = 0.23) for a divalent oxyanion was surprisingly higher than for a monovalent simple anion as iodide (ηR = 0.16). According to a deduction scheme proposed by Maes (2006) and a set of simple calculations based on the Bruggeman relationship, it could be consistent with a small retardation factor (R = 2.3) suggesting that sulfate is thus weakly sorbed in Boom Clay.

R values in the range 2 – 6 correspond to very small distribution coefficients (Kd of 0.2 – 1 ml/g), especially difficult to determine experimentally by means of classic batch sorption tests. Indeed, the decrease of concentration expected in the supernatant is extremely small and probably within the uncertainties of the measurement technique. This remains thus to be carefully verified by means of dedicated sorption tests adequately designed.

258/328

A8. Selenium and organic matter

259/328

260/328

A8 Selenium and organic matter

During the HR-ICP-MS analyses of selenium, the precipitation, or flocculation, of humic acid was observed by Maes (2004) after acidification of the solution. The consequence was a decrease of the Se concentration remaining in solution after filtration, or centrifugation. This observation suggests that a fraction of selenium could be associated with the natural organic matter present in Boom Clay. However, analyses of natural selenium concentrations in immobile and mobile organic matter of Boom Clay have not yet been performed and would deserve more attention in the future.

Selenium may occur in nature in many organic compounds where it is covalently bound to carbon atoms. Selenium may occupy the place of sulfur in many peptides and proteins and their subsequent degradation products commonly found in soils. Some major organo-selenium compounds are listed in Table A8.1 as illustration.

Table A8.1: Some major organic compounds of selenium commonly found in nature and whose degradation products are incorporated in natural organic matter (NOM) by micro-organisms. Oxidation State Organic Compound Abbreviation Chemical Formula

Se(-II) Dimethylselenide DMSe CH3SeCH3

Se(-I) Dimethyldiselenide DMDSe CH3Se–SeCH3

Se(-II) Diallylselenide DASe Allyl-Se-Allyl

Se(-I) Diallyldiselenide DADSe Allyl-Se–Se-Allyl

+ + Se(-II) Trimethylselenonium TMSe (CH3)3Se

+ – Se(-II) Selenomethionine Se-meth. H3N CHCOO ·CH2CH2SeCH3

+ – Se(-II) Selenocysteine Se-cyst. H3N CHCOO ·CHSeH

Se(-II) Se-Glutathione GSH-Px enzyme: complex protein

Se(-II) Selenocyanate — SeCN–

Selenium is also often found associated with organic matter (OM) in seleniferous rocks and contaminated soils.

The frequently observed correlation between selenium and organic matter in water and soil samples does not necessarily imply that inorganic selenium species are complexed by dissolved organic matter, or sorbed onto solid OM (kerogen fraction). Two well established

261/328

selenium pathways and many other hypotheses could explain this correlation between selenium and organic matter in environmental samples: 1. microbial sulfato-reduction can occur in shallow sediments, or soils, under reducing conditions in the presence of organic matter (OM). Traces of dissolved inorganic selenate are reduced along with sulfates (electron acceptors) and subsequently coprecipitated as selenide into sulfides minerals (e.g., pyrite). Organic matter is fuelling the bacterial activity as an electron donors, so seleniferous sulfide deposits are quite logically correlated to OM; 2. as previously mentioned, organo-selenium molecules might be present in residues of biodegraded “sulfur-like” protein structures left after maturation of organic matter in soils or sediments; 3. selenium could be linked to organic matter by means of an iron or aluminium bridge: Fe3+ or Al3+ cations could bind Se(IV) or Se(-II) and OM in dissolved Se–Fe–OM ternary complexes where the trivalent cation occupies a central position between both anionic entities;

+ 4. if positively charged amino-acids groups (H3N -CH-COOH) are present in the structure of OM, they could attract selenite or selenide anions; 5. and finally, a weak association of amorphous colloidal Se0 and OM by means of hydrophobic interactions cannot be ruled out. These different potential mechanisms are discussed more in detail hereafter in their corresponding sections respectively.

1. Sulfato-reduction and correlation between sulfide/selenide and organic matter A first pathway involves the process of sulfato-reduction where selenium follows sulfur and is incorporated in selenium-bearing pyrite present in the clay and shale formations with high content in organic matter. This organic matter (reductant) is needed as electron donor to fuel the sulfato-reducing bacteria (SRB) responsible for the reduction of sulfate and selenate present in the sediments. After reduction, the produced sulfides and selenides react with Fe2+ leading to the precipitation of selenium-rich pyrite. In case of sulfato-reduction, pyrite is correlated with OM, and so is also selenium at trace level.

2. Organo-selenium present in the organic matter structure A second pathway is based on the direct biological incorporation of organo-selenium compounds in the natural organic matter through two seleno-amino-acids: seleno-methionine and seleno-cysteine. Numerous organisms are able to incorporate selenium in peptides and proteins: bacteria, yeast, fungi, algae, phytoplankton, plants, animals, mammals, and finally humans. Selenium takes the place of sulfur in proteins and other organic molecules. Enzymes can selectively organise around selenium cofactors to catalyse specific oxidation-reduction

262/328

processes in the intra-cellular liquid (cytosol). The first discovered (Rotruck et al., 1973) and the best known protein is the selenium containing glutathione peroxidase protecting living cells and their phospho-lipidic membranes against oxidation from free radicals (peroxide: HO–O•, and HO•) produced by the respiration chain in the mitochondria.

Selenium also exhibits a strong bioaccumulation in the food chain. It is indispensable to mammals and humans, but there is little evidence that selenium is essential for plants. However, some plants, e.g., Indian mustard, cabbage, broccoli, cauliflower, sprout, behave like Se-bioaccumulators and present a surprising resistance to high levels of selenium, so that their use in selenium phytoremediation is envisaged using transgenic varieties (Bañuelos et al., 2005).

Crops of the alliae family (garlic, onion, …) can also accumulate a considerable amount of selenium in their bulb as illustrated by μ-XANES tomography made at University of Antwerp (UA) by Prof. Janssen (2003). Selenium can also take the place of sulfur in allyl sulfide (the volatile eye-irritant substance emitted by onions), and many plants can emit volatile organo- selenium compounds (selenium disulfide, diselenium disulfide).

Many biological pathways of incorporation of selenium into proteins exist and still remain to be discovered. After the death of marine plankton and terrestrial plants, maturation of organic matter occurs giving rise to humic matter and kerogen bearing evolved organo-selenium compounds. Organic matter maturation increases with burial depth and temperature leading to a great diversity of sulfur-organic molecules accompanied by seleno-organic molecules. As a consequence, organic-rich clay and shale, but also coal deposits and oil reservoirs may contains in some locations very high level of selenium.

In the case of Boom Clay, according to Deniau (2005) low sulfur content and S/C ratios around 0.025 are observed for the immature Boom Clay Kerogen (BCK). As a consequence, only a low level of organo-selenium is expected for Boom Clay organic matter.

3. Iron bridge [Se—Fe—OM] A hypothesis remains to be verified: the indirect association of selenium with organic matter (OM) by means of a third binding species.

If one excepts one very particular species of organo-selenium, the tri-methyl-selenonium + + 0 (CH3)3Se , or TMSe , which is a cation, and elemental Se , all inorganic dissolved species of 2– 2– – selenium are negatively charged anions (SeO4 , SeO3 , HSe ). As a consequence, these anionic species experience in a large range of pH a general non-specific electrostatic repulsion from the negatively charged humic acids (because the presence of dissociated carboxylic, R-COO–, and phenolic, Ø-O–, groups). This explains why the association between

263/328

radiolabelled 75Se and humic acids is very rarely observed, and if sometimes noticed, very weak.

To elaborate a satisfying theory accounting for a sorption of anionic selenium on the generally negatively charged organic matter, one should envisage mechanisms like chemisorption, inner-sphere complex, bidentate bonds, or ligand exchange with an intermediate metallic species acting as bridge between the two anionic antagonists. This could perhaps occur if amorphous hydrous oxides of iron (III) or aluminium (III) are present in the system. Colloidal particles of iron or aluminum hydrous oxides (goethite, gibbsite) could then play the role of bridge between the two negatively charged entities.

2– – It is possible to imagine that both the selenium species (as SeO3 or HSe ) and the organic matter would form inner-sphere complexes on the colloidal Fe(OH)3 particles. The mechanism would be similar to the sorption of selenite onto iron hydroxides and to the iron- bridges that strongly retain OM at the surface of goethite.

2– – OM + Fe(OH)3 + SeO3 —> OM—Fe(OH)—SeO3 (eq. A8.1)

This association is analogous to that of ternary complex of A type (surface-metal-ligand, ⊖ ⊕ ⊖) in surface complexation model (while ternary surface complexes, of B type follow the reverse order: surface-ligand-metal, ⊕ ⊖ ⊕), as mentioned by Buerge-Weirich et al. (2003) in one of their publications on adsorption of heavy metals on goethite in the presence of organic ligands.

The same hypothesis is also put forward by Redman et al. (2002) and Macalady et al. (2002) to tentatively explain the mobilisation of arsenic in the presence of iron oxide and organic matter. Indeed, the reason of the mobilisation of arsenic in the presence of organic matter and iron oxides under strongly reducing conditions seems more complicated than thought and could depend on other poorly understood processes. A consistent theory is also lacking to explain the anomalously high level of arsenic found in water wells in Bangladesh where arsenic poses a serious health problem to millions of persons.

So, although such an iron bridge [Se—Fe—OM] is perfectly conceivable, there is still a lack of evidences to clearly demonstrate this hypothesis in the context of Boom Clay. In the frame of our limited survey of the literature we have not found many convincing papers on the subject. This topic certainly deserves more attention and further research. If this mechanism exists, a supplementary correlation with iron should also be observed in natural samples containing selenium in association with OM.

264/328

4. Other weak interactions Other poorly established mechanisms are also sometimes suggested to explain an association between selenium and OM.

A first one sets forth the possible presence of protonated amino-acids in the structure of + organic matter: H3N -CH-COOH. At low pH, these amino-acids could bear positively charged ammonium groups that could interact with selenium anions.

A second one puts forwards a hydrophobic interaction between humic acids and colloidal elemental selenium, as recently suggested by Maes et al. (2004) to explain the association of selenium with organic matter in some Se/OM interaction experiments.

5. Possible implications A possible correlation (or association) between selenium and organic matter could be important for two reasons: − to know if the solubility limit of natural inorganic selenium in Boom Clay is reached, and; − to assess the possible interaction of long-lived selenium-79 with natural selenium associated to the natural organic matter.

(i) Impact of a possible confusion between total dissolved selenium and inorganic selenium on the determination of saturation index of different Se species in Boom Clay porewater

When measuring the total selenium concentration present in clay porewater, it is not easy to distinguish free-inorganic dissolved selenium (e.g., HSe– expected under strongly reducing conditions) from organo-selenium.

If a non-negligible fraction of selenium present in Boom Clay is well biologically incorporated in the organic matter since its sedimentation, the total dissolved selenium concentration in Boom Clay porewater does not represent only the sum of inorganic Se species. The speciation of dissolved selenium must be considered with prudence before to determine a saturation index, or to attempt to validate the results of thermodynamical solubility calculations. A clear distinction need to be made between inorganic (mineral) and organic dissolved selenium: [Seaq mineral] = [Seaq total] – [Seaq organic]. Considering that the total dissolved selenium corresponds to dissolved selenide would lead to an overestimation of the true dissolved selenide concentration.

(ii) Effect of organic matter on the mobility of 79Se: comparison with the case of 129I

What is the potential implication of organic matter on the transport of 79Se ? An increase of its mobility if 79Se would be associated to the dissolved organic matter (OM), or at the contrary a

265/328

higher retention if 79Se would be rather sorbed on the immobile OM, or isotopically exchanged with natural Se–OM groups present in the Boom Clay kerogen ?

The anionic character of dissolved inorganic species of selenium is in principle not favourable to their association with the negatively charged organic matter at slightly alkaline pH (8 –8.5). 2– Only a weak association between SeO3 and OM has been observed. No association could be – 2– established between HSe or SeO4 and OM.

If stable ternary complexes [Se—Fe(III)—OM] (a.k.a. iron bridge) should play a significant role in Boom Clay, they could influence the transport of 79Se.

Another open question deals with the presence of organo-selenium compounds in natural kerogen which could perhaps also contribute to retard the long-lived 79Se by isotopic exchanges.

Such a question was already formulated in studies dealing with a possible very weak retention observed for 129I in Opalinus Clay (OPA) and in the Callovo-Oxfordian Clay (COx). Indeed, organic matter of these clay formations is also naturally rich in organo-iodine compounds (a 3 -1 few mg iodine / g clay), and a low Kd value (~ 0.2 cm g ) corresponding to a small retardation factor R = 2 has been determined by several authors (Van Loon and Soler, 2004; Van Loon et al., 2006; Devol-Brown et al., 2003, Tournassat et al., 2005). A possible isotopic exchange between 129I and natural organo-iodine was suspected for the Callovo-Oxfordian Clay (Bure URL) and the Toarcian Clay (Tournemire tunnel), but the effect cannot be taken into account for safety calculations because it is very small and still disputed amongst different laboratories in France (Devivier et al. (2004), IRSN; Bazer-Bachi et al. (2006), CEA; Tournassat et al. (2006), BRGM; Reiler (2006) in GCA, CEA; Wittebroodt (Ph. D Thesis, 2009), IRSN). Moreover, according to Claret et al., (2009) at BRGM, natural iodine in the COx is not correlated with organic matter but well with calcium carbonate (aragonite of the fossil shells preserved in the COx). The same correlation could also apply for natural selenium in COx, but it remains yet to be proven (Carignan, 2008, Personal Communication).

266/328

A9. Selenium migration behaviour in Boom Clay

267/328

268/328

A9 Selenium migration behaviour in Boom Clay

A9.1 Overview

The results of the first migration experiments made with 75Se on compact Boom Clay cores have revealed very intricate profiles in the solid clay at the end of the tests. The initial interpretations of these results at SCK•CEN with the analytical Micof code (Henrion et al., 1990; Put et al. 1992; De Preter et al., 1992; De Cannière et al., 1996) have first lead to think that selenium was strongly sorbed onto Boom Clay with a retardation factor (R) of ~ 300. However, when considering that the 75Se source was initially not in chemical equilibrium with Boom Clay conditions and that the large activity engaged was much higher than the solubility limit of the species prevailing under reducing conditions, we realised that most of selenium had likely precipitated around the source after its progressive reduction along its migration path. Because the gamma counting method used to measure selenium-75 in the clay did not allow to distinguish between precipitated selenium and sorbed selenium, we realised that the first interpretations of the migration profiles were inappropriate and that precipitated selenium was mistakenly considered as sorbed. As a consequence, a highly overestimated retardation factor was derived from these experimental data and used for the performance assessment (PA) (Marivoet et al., 1999: Data Collection Forms, Feb 1998). When we realised the process really at work, we considered that most of selenium was precipitated near the source and we assigned new values to the migration parameters of 79Se (Marivoet et al., 1999: Data collection Forms, Feb 1999; data set 1). Since then selenium is considered as a non-retarded element in Boom Clay (R = 1). However it is expected to be solubility limited (S = 5 × 10-8 mol dm-3) under reducing conditions. Recent results obtained from electro- migration tests (Beauwens et al., 2005) made with 75Se selenite source containing selenate as impurity have revealed that if selenate is present in the system, the selenate species migrates unretarded, but contrary to selenite, it does not undergo reduction, nor subsequent precipitation within our experimental conditions, and at the time frame and the spatial scale of the investigated system. So, we realise that if selenium would be present in the source term of a HLW geological repository as selenate, it could be not solubility limited as expected according to thermodynamical calculations.

269/328

A9.2 Percolation Experiments

75 2– A9.2.1 Percolation experiments with SeO3

A9.2.1.1 Percolation tests: experimental The principle of percolation experiment with an undisturbed Boom Clay core is given in 75 Figure A9.2.1. A few µl of Se (T½ = 120 d) solution with a very high activity (0.68 milli- Curie: 2.5 × 107 Bq of 75Se) are pipetted onto a filter paper and dried in air. 75Se is not carrier- free and is accompanied by cold selenium because it was produced by neutron activation according to the nuclear reaction 74Se (n, γ) 75Se. All selenium is supposed to be initially under the form of selenite according to the technical specifications given by the supplier. However, the manufacturer also acknowledges that because of the high γ activity of the 75Se stock solution, and water radiolysis, a fraction of selenite could have been oxidized in selenate by oxidizing free radicals (•OH) produced by water radiolysis. At the time of these experiments, this point was not controlled by ion chromatography. The source is then transferred into a nitrogen atmosphere glovebox (≤ 1 ppm O2) and sandwiched between two Boom Clay cores of about 3.5 cm length (sampled by drilling perpendicular to the bedding plane). (This type of source configuration in sandwich is sometimes referred in the literature as “back-to-back” plugs). The whole is then pushed with a manual press in a stainless steel cell and confined between two stainless steel porous filters (Krebshöge). The screw-caps of the percolation cell are tighten with a torquemeter set at 150 Newton meter. After assembly, the stainless steel set-up is gas tight to prevent oxygen ingress in order to reproduce the anticipated in situ chemical conditions imposed by the clay. Boom Clay interstitial water in equilibrium with a 0.4 % CO2 atmosphere is percolated at constant flow rate through the core as illustrated in Figures A9.2.1 and A9.2.2. A typical flow rate of ~ 0.150 cm3 d-1 is achieved with a hydraulic gradient (ΔP) of 10 to 14.3 MPa m-1 (~ 10 bar on 7 cm). Water samples are collected at regular interval at the outlet of the cells and their 75Se concentration measured with an automatic NaI(Tl) gamma counter (75Se decays to 75As, emitting main γ at 136, 265 and 279 keV respectively, DAMRI, 1991).

clay water outlet

radionuclide source C

Figure A9.2.1: Schematic principle of a percolation experiment with the radionuclide source sandwiched between two clay cores. Also known as “back-to-back” or “C4 type” experiment according to the in-house classification of M.J. Put described by De Preter et al. (1992).

270/328

[RN] [RN]

x t

RN profile after cutting RN breakthrough curve (for retarded RN) (for non or weakly retarded RN) (for solubility controlled RN: sigmoid curve) RN = Radionuclides

Figure A9.2.2: Schematic representation of a percolation experiment: expected radionuclides profiles in clay and in the percolated water.

Two percolation experiments were performed in the surface laboratory between 1995 and 1998: • NRM010 A (from 08-March-1995 to 02-April-1997), and; • NRM010 B (from 08-March-1995 to 05-March-1998). These two percolation experiments are referred as “C4 type” according to an in-house classification defined by M.J. Put and described in the state-of-the-art report on the Migration studies by De Preter et al. (1992).

A9.2.1.2 Evolution of 75Se concentration in the percolation water The evolution with time of the concentrations of selenium measured in the water collected at the outlets of the percolation experiments is given on Figure A9.2.3 (linear scale) and Figure A9.2.4 (logarithmic scale). The value of the hydraulic conductivity of the percolation experiments varies from 2 × 10-12 m s-1 to 4 × 10-12 m s-1, a typical range for Boom Clay. After a very fast breakthrough peak of 75Se concentration at respectively about 8 × 10-7 mol dm-3 (NRM010 A) and 2.8 × 10-7 mol dm-3 (NRM010 B) a lower selenium concentration is observed in the percolate of both experiments at about 1.6 × 10-8 mol dm-3.

271/328

9.0 E-07 ] -1

8.0 E-07

7.0 E-07

6.0 E-07 Migration experiment NRM010A with Se-75

5.0 E-07 selenium concentration [mole.l

4.0 E-07

3.0 E-07

2.0 E-07

1.0 E-07

0.0 E+00 0 20 40 60 80 100 120 140 160 180 Volume of percolated water [ml]

Figure A9.2.3: Concentration of 75Se in the percolation water of the NRM010A experiment. A concentration plateau of 1.6 × 10-8 M is reached at the end of the experiment (linear concentration scale).

1.0 E-06 ] -1

1.0 E-07 Migration experiment NRM010A with Se-75 selenium concentration [mole.l concentration selenium

1.0 E-08

1.0 E-09 0 20 40 60 80 100 120 140 160 180 Volume of percolated water [ml]

Figure A9.2.4: Concentration of 75Se in the percolation water of the NRM010A experiment. A concentration plateau of 1.6 × 10-8 M is reached at the end of the experiment (log concentration scale).

As seen on Figure A9.2.3 (linear scale) and Figure A9.2.4 (logarithmic scale) showing the evolution of 75Se concentration in the percolated water, a breakthrough peak of mobile 75Se rapidly appears in the first milliliters of water collected at the beginning of the percolation. Although no chemical speciation analyses could be performed on these samples, selenium is 2– assumed to be mainly present in the first water samples as slightly retarded SeO3 or very 2– 75 mobile SeO4 . Then the Se concentration rapidly decreased to reach a plateau value of about 1.6 × 10-8 mol dm-3. At this level of concentration, we suppose that the selenium

272/328

concentration is solubility limited by metallic Se, or by FeSe, or FeSe2 (crystalline or amorphous phases). We interpret this low selenium concentration assuming that the mobile dissolved species of selenium is HSe– the species in thermodynamical equilibrium with reduced selenium solid phases under in situ conditions. In this hypothesis, we do not envisage 2– the dissolution of selenide by oxidation (then traces of SeO4 could also be present in solution). The fact that 75Se is permanently detected in the percolation water suggests that we are likely faced with a constant concentration source of poorly soluble selenium (Se, FeSe, or

FeSe2 ?). So, to summarize our interpretation, the sequence of elution of the different aqueous species of selenium present in the percolate is the following: small quantities of impurities of 2– SeO4 the more oxidized but also an almost unretarded species of selenium appears the first 2– in the water. Then a part of the bulk of SeO3 , the main species of selenium introduced at the source percolates out of the clay core without undergoing chemical reduction because of kinetics limitations. Finally HSe–, the selenium species at the lowest valence is released in the water when all selenate and the non reacted fraction of selenite have been totally leached out of the core. Selenide originates from the chemical reduction of selenite by pyrite or organic matter during the transport of selenite in the clay. However, no direct speciation measurements were performed on the percolate. The quantity of mobile selenium released by each clay core is low: about ~ 0.62 % of the initial inventory pipetted onto the paper filter for each percolation test.

A9.2.1.3 75Se migration profile in the solid clay At the end of the percolation experiments (2.07 and 2.99 years of percolation for experiments NRM010A (756 d) and NRM010B (1 093 d) respectively), the clay cores were progressively removed from the permeameter cells with a screw mechanical press and cut in thin slices of 0.5 – 1 mm thickness to allow the determination of the migration profile of selenium in the solid clay. 75Se was analyzed by means of an automatic NaI(Tl) counter.

The 75Se activity profile measured in the clay core after the percolation experiment for NRM010A is presented in Figures A9.2.5 and A9.2.6 (plotted on a linear or a log scale respectively). An effect of percolation due to advection is observable on the asymmetric activity profile and an increase of selenium concentration appears downwards to the source position, in the direction of the water flow.

273/328

600 000

NRM010A: migration of Se-75 500 000 in a vertical Boom Clay core

Fit result 400 000 Fitted experimental results Non fitted experimental points

300 000 Conc (Bq/cm)

200 000

Direction of percolation

100 000

5 4 3 2 1 0 -1 -2 -3 -4 Distance (cm)

Figure A9.2.5: Migration profile of 75Se in a vertical Boom Clay core (experiment NRM010A, percolation type C4). Linear scale. Fit on relative error.

10 000 000

NRM010A: migration of Se-75 1 000 000 Direction of percolation in a vertical Boom Clay core

100 000 Fit result Fitted experimental results Non fitted experimental points 10 000

Conc (Bq/cm) 1 000

100

10

1 5 4 3 2 1 0 -1 -2 -3 -4 Distance (cm)

Figure A9.2.6: Migration profile of 75Se in a vertical Boom Clay core (experiment NRM010A, percolation type C4). Log scale. Fit on relative error.

Figure A9.2.5 (linear scale) shows a well developed peak of 75Se whose the tails reach the baseline after about 1.2 cm. Very surprisingly, the activity increases again near the stainless steel porous plate confining the clay core at the water outlet ! Figure A9.2.6 gives a clearer representation of the 75Se migration profile on a logarithmic activity scale. Most of the measurements points of the central diffusion peak (xo ± 1.2 cm) are lying along a parabolic curve, if one excludes however a tenth of points with the highest activity just near the 75Se source. Beyond one centimeter on each side, the 75Se activity decreases less rapidly. At greater distance, the parabolic curve does no longer fit the measurement points whose activity

274/328

is clearly above the limit of detection for 75Se. A more or less constant background of 75Se is then observed. The migration is not symmetrical in both directions from the source: an unexpected increase of activity occurs downwards just beside the outlet stainless filter. Due to the complexity of the measured migration profile modelling attempts have only been performed with the experimental points lying on the parabolic curve. The other points have not been taken into account for the fitting. This complicated migration profile is a good illustration of chemical-coupled transport involving also reduction-precipitation reactions and chemi-sorption with surface complexation. Although the inorganic chemistry of selenium is very similar to that of sulfur and may appear relatively simple when looking at Eh - pH diagrams (Brookins, 1988), the whole situation is more difficult to interpret without thermodynamical sorption modelling, and without data on reduction-precipitation kinetics, or on interaction of selenium with organic matter (OM).

So, we are reduced to a purely speculative scenario to give a qualitative interpretation of the observed profiles. We thought to have introduced selenium onto the filter paper as selenite 2– (SeO3 ) according to the chemical specifications of the source provided by Amersham, but no direct measurements were performed to control the chemical form of selenium in the spike. However, because of the water radiolysis of the 75Se stock solution selenite was likely 2– accompanied by selenate (SeO4 ) as indirectly evidenced by Beauwens et al. (2004) during later electromigration tests. This hypothesis has been independently confirmed by Bruggeman et al. (2006) which applied specific characterization techniques to several selenium sources supplied by the same manufacturer. Under the low Eh conditions normally prevailing in the 2– undisturbed Boom Clay cores used, SeO3 was most likely slowly reduced to metallic selenium (Se), or further to selenide (HSe–). Metallic selenium is very insoluble while HSe– 2+ easily precipitates with Fe to form ferroselite (FeSe, or FeSe2), an iron selenide analogous to pyrite. So, perhaps that most of selenium precipitated close to the source in the first days of the percolation experiments. The rate of precipitation may be controlled, either by the redox properties (reducing capacity, state of freshness) of the clay surrounding the filter paper, either by a kinetic limitation, the reduction reactions involving multiple electrons transfers being often quite slow. On one hand, if the redox capacity of the first “section” of clay adjacent to the paper filter spiked with 75Se is exhausted by the reduction reaction of cold 2– selenium, SeO3 has to move to the next “section” to be reduced and to precipitate as Se, or 75 FeSe2. In this hypothesis, we are faced to an Eh gradient in the clay surrounding the Se source explaining partly the profile. On the other hand, if the redox capacity of the clay close to the source is sufficient to reduce immediately all the selenium and that selenium continues to move without immediate reduction, it may be due to kinetic limitations. The whole is still 2– complicated by the fact that the oxyanion SeO3 can easily be sorbed onto ferric and aluminium hydrous oxide by ligand exchange forming a stable monodentate or bidentate complex with the metallic ion (also called inner-sphere complex). This reaction is similar to

275/328

that of chemi-sorption of borate, silicate, phosphate and arsenate, onto Al or Fe oxy- hydroxides, schematically, for a monodentate surface complex:

2– – – SeO3 + >Al—OH —> >Al—SeO3 + OH (eq. A9.2.1) or for a bidentate complex:

2– + – 2 X–OH + SeO3 + H —> 2 X=SeO3 + OH + H2O (eq. A9.2.2) Surface Ligand Proton Inner-sphere Water hydroxyl group Bidentate (X= Surface) complex

So, selenite might already be retarded onto aluminium oxide groups located on the lateral edges of clay minerals (or also on the basal plane of kaolinite) before to be reduced by the more reactional clay constituents (pyrite, siderite, glauconite, sorbed Fe2+, and maybe organic 2– matter). If pyrite oxidation occurs, SeO3 might be sorbed onto the resulting ferrihydrite

[FeO(OH)], or jarosite [KFe3(SO4)2(OH)6], produced in the near-field of a ventilated gallery. The type of chemi-sorption reaction on iron oxy-hydroxides is the same than on aluminium oxides, but with a higher affinity for Fe3+ than for Al3+, as, e.g., in the case of a monodentate mononuclear complex:

2– – – SeO3 + >Fe—OH —> >Fe—SeO3 + OH (eq. A9.2.3) or for a bidentate binuclear surface complex:

(eq. A9.2.4)

As seen on Figures A9.2.5 and A9.2.6 the selenium bulk concentration outside the parabolic curve (log scale) is more or less constant if one excepts the slices near to the porous stainless steel filter at the outlet. This concentration corresponds to that of selenium measured in the percolated water (~ 1.6 × 10-8 mol dm-3 ).

How to explain the unexpected 75Se “peak” (or accumulation) observed downwards near to the stainless steel porous filter at the outlet of the percolation cell ? The best clue we have up 2– 2– to now is that a fraction of the soluble selenium (SeO3 accompanied by SeO4 at the beginning of percolation, followed later by HSe– in the steady state) is sorbed or precipitated onto a very thin layer of ferric hydrous oxide always present at the surface of sintered stainless steel. A direct reduction-precipitation of selenium by metallic iron (Fe-0) is also

276/328

another possible mechanism. After its uptake by the metallic porous filter, selenium is progressively released in the water and percolates outside, or diffuses back inside the clay core, the filter acting then as a secondary source of 75Se.

A9.2.1.4 Modelling of the 75Se profile in the solid clay: questionable attempt So, to summarize these explanations, several zones can be distinguished in the migration profile of 75Se in the clay core presented at Figures A9.2.5 and A9.2.6: 1. a region very close to the initial source (first precipitation ?); 2. a region inside a parabolic curve (mainly diffusion); 3. a region outside the parabolic curve with a low but measurable background of 75Se; 4. a region close to the stainless filter at the outlet acting as a secondary 75Se source (back-diffusion peak), and; 5. finally, the stainless steel outlet filter first taking up 75Se, and then releasing the accumulated 75Se.

An attempt of modelling with the Micof code (based on an analytical solution of the simple advection / dispersion model with linear and reversible sorption) was performed only using the measurement points of the second region which are properly aligned on a parabolic curve. All other points (very close to the 75Se source, in the background, or near the stainless steel filter at the outlet) have been omitted in the fitting to simplify the problem. The programme dfit38 was used. The results of the fitting presented at Figures A9.2.5 and A9.2.6 are the following:

ηR = 1 123 R = 11 230 if we assume η = 0.10 -14 2 -1 Dapp = 6.9 × 10 m s -11 2 -1 Deff = ηRDapp = 7.7 × 10 m s

What are the consequences of omitting the points around the origin during the fit ? First, the 5 fitted value for the total amount of selenium in the clay (Qmo = 1.18 × 10 Bq) is 200 times smaller than the real total amount of selenium effectively present in the clay core 7 (Qo = 2.52 × 10 Bq). Indeed, the major part of selenium in the clay is situated around the source (see Figure A9.2.5, linear concentration scale), and these points have been omitted. Then, by omitting these points, the value of the diffusion coefficient is higher than if these points should not have been omitted. As a consequence, the value of the diffusion coefficient obtained from the fit presented at Figure A9.2.6 is a conservative estimation for the migration of 75Se around the source.

277/328

However, for the positions around the edges of the clay core, the fitted value for the diffusion coefficient is no more conservative. According to this fit, no selenium could have diffused out of the clay. But, during this experiment, a small fraction (0.62 %) of selenium (1.55 × 105 Bq on the 2.52 × 107 Bq initially introduced) has been leached out of the clay. The profile measured outside the parabolic domain suggests a concentration limit controlled by a solid phase as, e.g., Se, FeSe, or, FeSe2.

As the transport of selenium may be coupled with various chemical reactions implying reduction, precipitation and chemisorption, two or three species of selenium with different 2– 2– migration behaviour (certainly SeO3 and SeO4 in the beginning, and then after its reduction: HSe–, or maybe HSe–Fe–OM) may contribute to the general profile measured in the clay and in the water. However, the programme “dfit38s” fits only one mobile species and is based on a conceptual model considering only diffusion, advection, sorption and radioactive decay, without taking into account reduction or precipitation phenomena. Another model (and various programmes for different chemical conditions) is needed to take into account the migration of two different species, as it was done by Put (1994) to attempt to fit the first four migration experiments made by Henrion and De Cannière (1990) with metallic Se and reduced Se (whose results are also summarized in Table A9.2.1). Three species seems a maximum for such a model to avoid the risk of over-parameterisation. However, a model with stable species migrating independently from each others is not sufficient. One of its main limitation is the difficulty to take into account the progressive transformation of one soluble 2– – species (SeO3 ) in another one (HSe ) precipitating, or sorbing differently, during the transport (need of kinetic constant, and solubility limit under in situ conditions as additional parameters).

A9.2.1.5 Comparison with the results of previous migration experiments made with clay plugs equilibrated with elemental and reduced 75Se

Four migration tests were performed by De Cannière and Henrion, (1990) with elemental selenium (reduction of selenite by hydrazine hydrochloride) and with reduced selenium (2 months in contact with a Boom Clay slurry) as two pure diffusion tests and two percolation tests. The detailed experimental procedure are reported in the semi-annual progress report of 1990 to 1992. Their principle is similar to these described more in detail in the previous section. Pure diffusion tests were not percolated with porewater. The main difference resided in the nature of the source and the way selenium was pre-equilibrated to match in situ chemical conditions. Table A9.2.1 allows for the comparison of the results of the percolation NRM010A experiment with those of the four previous migration tests performed under reducing conditions with pre-equilibrated sources as reported by De Cannière et al. (1996). The results have been sorted by increasing order of retardation according to the ηR value.

278/328

Table A9.2.1: Results of migration experiments with selenium in Boom Clay under reducing conditions.

(a) Selenium Migration Test ηR Dapp Deff = ηRDapp Cbulk Source Type (—) (m2 s-1) (m2 s-1) (Bq ml-1)

⎯Reduced • diffusion (C3) — 2.1 × 10-13 — 122 • percolation (C4) 54 1.9 × 10-13 1.0 × 10-11 849

⎯Metal • diffusion (C3) — 1.2 × 10-13 — 1 610 • percolation (C4) 215 1.3 × 10-13 2.8 × 10-11 11.300

2– -13 -11 ⎯SeO3 • percolation (C4) 1 123 0.69 × 10 7.7 × 10 —

(a): Cbulk : the calculated bulk concentration in the source at the end of the experiment, taking into account a counter yield of 33 %. The bulk concentration is the sum of the concentrations in solution and in the

solid phase (sorption + precipitation) per unit clay volume. Cbulk = ηR Caq (Caq : concentration in water).

An important remark must be made when analysing the reported migration parameters for these experiments. The experiments were at that time not modelled with a constant concentration boundary condition (all Se was considered to be soluble) which is clearly not correct. Simply speaking, the Dapp is taken from the width of the diffusion profile which develops in de clay around the source position. However, if the source is solid (solubility limited constant release), it is clear that the so-called distribution profile is very narrow leading to relative low Dapp values which in fact do not reflect true diffusion. Finally, a last question arises: what is the mechanism of retardation of selenium in Boom Clay ? It is difficult to distinguish true reversible sorption from specific chemi-sorption, or surface precipitation. We have no clue to know if HSe– is sorbed, or not. As anion without oxygen similar to iodide, a priori, we do not expect sorption because of the electrostatic repulsion with the negatively charged surface of the clay minerals (anion exclusion). Due to 3– the absence of oxygen in its molecular structure, chemisorption analogous to that of PO4 , or 2– – 2+ 3+ SeO3 is also not expected. However, HSe could strongly interact with Fe / Fe anywhere in the clay (sorbed iron, cation exchange pool, OM complex), or present at the surface of minerals as pyrite, siderite, or glauconite. If this interaction does not lead to an irreversible precipitation of FeSe, but to a reversible exchange, perhaps that HSe– could be retarded.

A9.2.2 Lessons learned during the updating of the Data Collection Forms (DCF’s) in 1999

In the course of the revision of migration parameters for selenium in the frame of the exercise of the Data Collection Forms (DCF’s) made in 1999 for the performance assessment, we realized a problem of interpretation and modelling with selenium percolation experiments. The selenium migration experiments were erroneously interpreted previously with the Micof

279/328

code. The analytical solutions used in the various programs designed for the different experimental configurations were very well adapted to the case of conservative tracers, i.e. non sorbed radionuclides as, tritiated water (HTO), halogenide anions (131I–, 82Br–, 36Cl–), or alkaline cations (22Na+, 134Cs+), with a simple chemistry, without complexation, without solubility limit, and without multiple oxidation states. However, precipitation processes (often initiated by redox reactions) were not taken into account in the conceptual model and in the code. So, precipitation in the Boom Clay cores labelled with 75Se, was not considered during the modelling calculations performed with the bulk activity in the clay cores to determine the migration parameters and was misinterpreted as a strong sorption.

The reasons for the confusion of precipitation with sorption in the modelling calculations is explained hereafter. After cutting the clay cores in thin slices, 75Se in each slice was counted by means of a NaI(Tl) crystal, but only the bulk activity could be directly measured in the lab without sequential extraction procedure:

Bulk activity = Σ activities = activities of the (soluble + sorbed + precipitated) fractions.

So, when measuring the total activity present in each clay slice, the distinction between sorption and precipitation was not possible. The analytical advection-dispersion model used took only into account linear and reversible sorption, not precipitation. The reason of the problem is that all activity measured in the solid phase was considered by the Micof model as sorbed. So, the large fraction of precipitated selenium was also considered as “sorbed selenium” in the calculations. The consequence was an overestimation of the retardation factors. To remain conservative for the safety assessment, these values were considered as suspect and all the soluble and mobile fraction of reduced selenium (HSe–) was considered as non retarded in the new set of parameter.

In fact, to correctly interpret diffusion or percolation tests in term of linear and reversible sorption with fast local equilibrium, the same precautions than for “Kd” measurements should also be taken as illustrated by Figure A9.2.7. In “Kd” experiments because one cannot also distinguish between sorbed species and precipitated phases, the first golden rule is to always work below the solubility limit. The second rule is to work at low loading capacity of the sorption sites to remain in the linear range if no sorption isotherm can be performed. So, the same rules also apply to migration experiments to remain in the valid range of the hypothesis underlying the conceptual model based on the hypothesis of linear and reversible sorption.

280/328

log Cs (mg/g)

precipitation sorption site saturation co-/surface- precipitation

ΓT

neither Kd nor isotherms isotherms

Kd solubility

log Cl (mg/L)

Figure A9.2.7: General principles of sorption isotherm and applicable working ranges to remain in the limit of validity of the models and to avoid precipitation processes.

The problem is the most critical with the redox-sensitive elements as Se, Tc, Np and U. The reason for the precipitation is that the 75Se sources were not in direct equilibrium with Boom Clay at the start of the experiments. 75Se was directly loaded in the clay in oxidized form 2– 2– (SeO3 accompanied by SeO4 produced by radiolysis) as delivered by the supplier. The initial selenium concentration in the source was relatively high because of the presence of cold selenium present as carrier. Moreover, because of the half-life of selenium (120 days) and to be able to detect 75Se during two or three years, a large activity was introduced at the start. The consequence was a slow precipitation in the source, or more worrying, in the clay during the transport process.

In conclusion, a source of a redox-sensitive radionuclide not in chemical equilibrium with the clay makes modelling and interpretations very difficult, if not impossible. Simple advection / diffusion transport models are no longer valid and complicated reactive transport models should be used to estimate the migration parameters. Moreover extra parameters as kinetic rates and specific area would also be needed. Indeed, the kinetics of reduction / precipitation are unknown and will depend amongst others on experimental conditions (pH, Eh, pCO2, ...), and on the mineral surface effectively accessible in the sample (could depend on the liquid-to- solid ratio). So, how to acquire correct data relevant for real in situ conditions ?

Reactive transport modelling is a very difficult task for redox-sensitive elements and remains an unresolved issue. Most often, the chemical-coupled models are only used for predictive

281/328

calculations, not for parameters estimation from experimental results. The knowledge of the exact mechanisms and reactions occurring during transport is a mandatory prerequisite. If multiple chemical reactions are simultaneously, or consecutively, involved, the system will be too complex to be modelled and a serious risk of over-parameterisation exists.

Recommendations So, to perform successful migration experiments with redox-sensitive elements in equilibrium with the clay, it is necessary to know the exact geochemical conditions prevailing in situ in 2– the clay (pH, Eh, pCO2, CO3 , ...) and to reproduce them effectively in the lab. After correct calculation of the expected speciation of the element under the relevant conditions a last challenge is to master and to control its speciation in the source to verify that the chemical equilibrium is reached before to start the migration tests.

A9.2.3 Conclusion of percolation experiments

However, because of conceptual model restrictions, and limitations of analytical measurement techniques, the determination of the migration parameters for selenium (diffusion accessible porosity, η; retardation factor, R; and apparent diffusion coefficient, Dapp) was very problematic with these total 75Se activity profiles. The analytical model used only considers diffusion, advection, linear reversible sorption and radioactive decay. The very high retardation factor values calculated from this “solid” profile (ηR = 1 123; R = 11 230) were completely incompatible with the low retardation factor value (R = 0) determined from the evolution of the concentration of selenium in the percolation water.

This apparent conflict can be resolved if selenite present initially on the filter paper is reduced by pyrite or organic matter in the Boom Clay and precipitates in the clay as metallic selenium or iron selenide. Considering the precipitated solid as “reversibly” sorbed in the mathematical analysis would give erroneous apparently high retardation factors. This conclusion, coupled with the solubility studies indicates that all or most of the selenite is reduced to Se(0) or Se(-2) in the Boom Clay. However, the work also highlights the need to avoid kinetically controlled precipitation reactions during future percolation experiments with redox-sensitive elements. Thus it is necessary to work with a source of the element chemically preconditioned to be in equilibrium with the clay under in situ conditions before starting the migration experiments, or to use concentrations below the solubility limit of the species occurring under in situ conditions.

282/328

A9.2.4 Percolation experiments with dual tracer: FeSe contacted with 14C-OM

In the framework of the EC 5th Framework TRANCOM-II project (Maes et al., 2004), percolation experiments were performed with so-called double-labelled tracer sources consisting of a tracer under a specific chemical form (here HSe– released by the FeSe solid phase) and in contact with 14C-labelled Boom Clay Organic Matter (BCOM). The aim is to verify if the mobile BCOM facilitates selenium migration by following both the fate of Se and the 14C-labelled BCOM.

To avoid technical problems in the preparation of reduced 75Se(-II) sources starting from 2– 74 76 77 78 SeO3 , it was decided to use stable selenium ( Se, 0.9 %; Se, 9.0 %; Se, 7.6 %; Se, 23.5 %; 80Se, 49.8 %; 82Se, 9.2 %) in the form of commercially available FeSe powder (Alfa Aesar).

A Nalgene tube was filled with 10 ml of synthetic Boom Clay water (SCW, prepared anaerobically, bubbled with Ar containing 0.4 % CO2 for 4 hours prior use) and spiked with an aliquot of 200 µl of the 14C-OM (prepared by Loughborough University; TROM 33-34, 5 316 mg C dm-3, 642.2 kBq cm-3). Subsequently, 0.061 g of FeSe (stored under anaerobic conditions) powder was added to this solution. The FeSe powder was used “as is” and was not washed nor purified.

The suspension is shaken and a small aliquot is taken as dual tracer source for the migration experiments (loading of the migration experiments was performed in a glovebox under anaerobic conditions in Ar containing 0.4 % CO2). The breakthrough of the stable isotopes of Se was followed by High Resolution ICP-MS (Museum for Middle Africa, Tervuren, Belgium) while the 14C-OM breakthrough was measured by liquid scintillation (LSC counting). Because of its much higher sensitivity high resolution ICP-MS was used to be able to measure concentrations lower than 10-7 mol dm-3 (corresponding to the detection limit of the normal ICP-Mass Spectrometer used in routine at SCK•CEN).

Two percolation experiments codename SeCOM 1 and SeCOM 2 were started under similar conditions but they exhibit different hydraulic conductivity: respectively 2.3 × 10-13 m s-1 and 7.1 × 10-13 m s-1.

For both experiments presented at Figures A9.2.8 and A9.2.9 we do not observe a clear relation between the breakthrough of 14C-OM and this of selenium. So, we have no obvious indication of an association between dissolved selenium (expected to be present under the form of HSe– in these experimental conditions) and 14C-labelled natural organic matter (NOM). However, a small fraction of Se percolates very fast out of the clay core. We attribute 2– it to the partial oxidation of FeSe in the source releasing mobile SeO4 . Indeed the solid FeSe

283/328

2– was used “as is” without any purification). The SeO4 species move unretarded through Boom Clay as evidenced by electromigration experiments with oxidized selenium sources (see Section A9.3.2). The presence of oxidised species was also observed during high resolution ICP-MS analyses of reference samples prepared from the same FeSe powder (Alfa Aesar).

80000

70000 SeCOM1 SeCOM2 60000

50000

40000

30000 Activity C-14 (Bq/l)

20000

10000

0 0 5 10 15 20 25 30 35 40 Average total percolated volume (ml)

Figure A9.2.8: Se-14C-OM percolation experiments: breakthrough curves of 14C-OM.

9,0E-07 SeCOM-1 8,0E-07 SeCOM-2 Se-RBCW 7,0E-07

6,0E-07

5,0E-07

4,0E-07

3,0E-07 Se concentration (mol/l) Se concentration

2,0E-07

1,0E-07

0,0E+00 0 5 10 15 20 25 30 35 40 Average total percolated volume of RBCW (ml)

Figure A9.2.9: Se-14C-OM percolation experiments: breakthrough curves of selenium. The fast 2– initial Se breakthrough, not associated to NOM, is due to the SeO4 . The horizontal line indicates the natural Se concentration in the Boom Clay porewater.

284/328

Furthermore, high resolution ICP-MS measurements of Boom Clay porewater samples made at Tervuren showed a concentration of natural dissolved selenium of 2 × 10-8 mol dm-3. This concentration is below the limit of detection of the standard ICP-MS used in the analytical chemistry laboratory of SCK•CEN and could not be observed previously because of the lack of sensitivity of this technique.

After the breakthrough of the oxidised Se species, the Se concentration decreases back to this natural background value. Surprisingly, the measured concentration of natural selenium in the water is higher than the value calculated from thermodynamic data for the solubility of Se solid phases [Se(0) or FeSe]. Because of the presence of this unexpected relatively high concentration of natural dissolved Se we cannot determine any migration parameters for aqueous selenium species from these experiments. We can only conclude that the different 14 soluble species of selenium released by FeSe(s) are not associated to the C-labelled NOM. If – the FeSe(s) source was not partially oxidized at the beginning, we expected HSe to be the dominant species in the percolation water of these two experiments.

The source of Se in the Boom Clay is presently unknown and deserves further study. Recent measurements on fractionated Boom Clay pyrite samples of different sizes and morphologies showed that pyrite contains between 12 – 33 ppm of Se (see Table A9.2.2). Another source for mobile natural Se might be the natural organic matter. It has been reported that Se is often associated to humic substances. This question also deserves more attention.

Table A9.2.2: Selenium content in different fractions of Boom Clay pyrite. Size Fraction Shape [Se] (μm) (—) (ppm) 20-32 Framboïds 12.2 32-64 Framboïds 18.1 64-125 Framboïds 29.3 125-500 Aggregates 33.0 >500 Concretions 20.1 >500 Faecal pellets 17.0

285/328

A9.3 Electromigration experiments

A9.3.1 Electromigration: experimental setup For electromigration experiments (more info on the electromigration technique can be found in Maes et al., 1998, 1999, 2001, 2002 and Beauwens et al., 2005), a clay core of approximately 10 cm in length is cut in two pieces. An aliquot (100 – 200 µl) of the radionuclide source is spiked on the surface of the transversal section and the cores are mounted in a Plexiglas electromigration cell. Electromigration is performed either in normal laboratory atmosphere or inside a glovebox under anaerobic conditions and with a controlled

Ar/0.4 % CO2 atmosphere. A constant electrical current of 5 – 20 mA is applied with a power supply between the electrodes for a time period of days up to weeks, and Boom Clay porewater is used as electrolyte solution. To avoid the development of acid and alkaline fronts in the clay core due to electrolysis reactions at the electrodes, the Boom Clay porewater is continuously pumped around to a buffer tank, taking care not to provoke a short-circuit as illustrated in Fig. A9.3.1. The electrical field in the clay core is continuously monitored. When the electromigration cell is dismantled, the clay core is cut in 1 mm thick slices which are then analyzed by radiometry to determine the activity profile of the radionuclide in the clay after diffusion (see Fig. A9.3.2).

DC power supply Cathode Anode - +

source position

clay core ceramic filter r ⊕

Electromigration cell Water compartment Water compartment

Peristaltic Acid-Base Peristaltic pump neutralization pump reservoir

Figure A9.3.1: Schematic principle of the set-up used for electromigration experiments.

286/328

ΔE - +

Cathode (–) Anode (+)

Electrolyte Clay Clay Electrolyte

RN source [RN]

cation anion x

Radionuclide (RN) distribution profile after cutting

Figure A9.3.2: Distribution profile of tracer in the clay core submitted to an electrical field.

2– 2– A9.3.2 Electromigration experiments with oxidized Se sources (SeO4 and SeO3 ) (Beauwens et al., 2005) The 75Se stock solution (37.86 MBq, ref. date 17-04-2002), was purchased from Amersham, -3 in the form of sodium selenite (Na2SeO3, containing about 50 mg dm of non radioactive selenite as carrier, 3.9 × 10-4 mol dm-3).

Two kinds of experiments were performed: the first one with the untreated diluted stock solution (EM_SeO3-series, source species: mainly selenite), and the second one with the diluted and oxidized stock solution (EM_SeO4 series, source species: mainly selenate).

Experiment with selenite: a 20 µl-aliquot of the 75Se stock solution was diluted in 2 000 µl BCW before its use in electromigration experiment (EM_SeO3/2).

Experiments with selenate: selenate was obtained by a first dilution step of 20 µl of the selenite stock solution with 36 µl hydrogen peroxide (H2O2, 30 % wt./vol.) and 144 µl of Milli-Q water. A second 10-fold dilution step in Boom Clay water gave the source solution used in electromigration experiments EM_SeO4/1, EM_SeO4/6 and EM_SeO4/7. For experiments EM_SeO4/2 through EM_SeO4/5, the second dilution step consisted in mixing 100 µl of the first dilution with 20 µl of NaOH 0.1 N and 880 µl Boom Clay water.

287/328

After electromigration, each clay core is cut in 1 mm thick slices, whose gamma activity is counted during 20 minutes in the 60 – 467 keV energy range with a Packard 5250 gamma counter.

It is clear from the activity profiles of electromigration experiments performed with the untreated selenite source solution (e.g. EM_SeO3/2, shown on Fig. A9.3.3) that two different selenium fractions are present in the source (likely due to a mobile and an immobile species), as they are distinctly separated in the Boom Clay by the electrokinetic processes. After two days of electromigration only, a small activity peak has moved away from the source position: we think that this mobile fraction is likely selenate. Most of the 75Se-activity remained at the source position, originating from the selenite initially present in the spike. If selenate moves, without undergoing reduction, in a practically unretarded way in the Boom Clay, the behaviour of selenite remains unclear. Selenite is either sorbed, reduced and precipitated as elemental selenium (Se0), or precipitated as iron selenide (FeSe). The presence of two different species in this experiment was a surprise for us, as we purchased a “chemically pure” selenite source. Selenate has however also been detected in the 75Se-labelled selenite spike used by Bruggeman et al. (2002). The presence of two different selenium fractions was also suspected from diffusion profiles in bentonite spiked with selenite (Garcia-Gutiérrez et 75 2– al., 2001). Apparently, radiolytic production of free radicals in a stock solution of SeO3 may cause a partial oxidation of selenite to selenate This information was confirmed by Amersham.

1000.00 activity (cps/g) 100.00

10.00 cathode anode

1.00

0.10 D.L.=0.07 cps/g

0.01 -50 -30 -10 10 30 50 distance from source (mm)

Figure A9.3.3: Experiment “EM_SeO3/2” with the untreated selenite source (note the logarithmic scale on the ordinate axis).

288/328

The identification of the fast moving species as selenate is confirmed by the activity profiles of electromigration experiments performed with the oxidized selenite sources. As shown on Fig. A9.3.4 (experiment EM_SeO4/1), after 18 hours of migration, a typical Gaussian diffusive profile is observed towards the anode. However, some activity remains at the source position, most likely because oxidation was incomplete for kinetics reasons. If oxidation occurred in the presence of NaOH, the relative amount of immobile species was even more pronounced (Fig. A9.3.4, experiment EM_SeO4/3). Apparently, the efficiency of H2O2 to oxidize selenite to selenate is lower at high pH. Following Yllera De Llano et al. (1996), the optimal conditions for the oxidation of selenite to selenate by H2O2 are met at neutral pH, and by bubbling O2 through the solution. In the present experiments, the volume of source prepared was to small to perform accurate pH- measurements.

Besides the assumption of incomplete selenite oxidation, another explanation could account for the persistence of some activity at the source position, i.e. that the clay is oxidized, either at the time of loading of the clay core, or during the experiment itself, thereby creating the opportunity for selenite to sorb onto newly formed iron oxy-hydroxides. To check the validity of such assumption, two last experiments were performed in anaerobic conditions (glovebox with Ar/0.4 % CO2 mixture) with the same selenate source. For experiment EM_SeO4/6, the oxygen contamination level inside the glovebox remained smaller than 8 ppm throughout the whole experiment, making it unlikely the formation of iron oxy-hydroxides or to deplete the reducing capacity of the Boom Clay. The activity profile is also shown on Fig. A9.3.4. As one can see, only one species is present in the activity profile. While this provides a further evidence for the absence of selenate reduction in Boom Clay at short time scales, even under anaerobic conditions, it does not resolve the issue raised by the presence of two species under ambient air conditions. Nevertheless, as oxygen contamination occurred during the loading of the second experiment in the glovebox (EM_SeO4/7, up to 400 ppm O2), the possible formation of iron oxy-hydroxides does not seem to exert an influence on the selenate behaviour in Boom Clay, at least during the short time scale of the present experiments. We then conclude that some selenite remained in the source at the time of loading experiments

EM_SeO4/1 to EM_SeO4/5, but was completely oxidized at the start of the two experiments performed in the glovebox (EM_SeO4/6 and EM_SeO4/7).

289/328

100 activity (cps/g clay) 90 EM_SeO4/3:163 h - 1mA 80 EM_SeO4/6:76 h - 5mA EM_SeO4/1: 18 h - 8mA 70 EM_SeO4/7:21 h - 12mA 60

50

40 cathode anode 30

20

10

0 -40-200 204060 distance from source (mm) Figure A9.3.4: Experiments with oxidized selenium source: EM_SeO4/1, EM_SeO4/3, EM_SeO4/6, and EM_SeO4/7. In case of incomplete oxidation an immobile species subsists beside the mobile one.

By performing a series of electromigration experiments with increasing electric fields, it is possible to estimate the apparent diffusion coefficient of selenate in the Boom Clay. Two different methods are used for this purpose, i.e. “curve fitting” or “hydrodynamic relationship”, and the “Nernst – Einstein” method. More details are given by Maes et al., (1999) and Beauwens et al., (2005).

The hydrodynamic relationship is a straightforward method, as both parameters, the apparent dispersion coefficient (Di) and the apparent velocity (Vapp) are determined for each electromigration experiment by fitting the measured activity profiles with a Gaussian equation. The method is based on the linear regression between the apparent dispersion coefficient Di (sum of advection and diffusion terms) and the apparent electromigration velocity Vapp, according to equation A9.3.1.

It is generally accepted that Di varies linearly with the advection velocity (Domenico and Schwartz, 1998), the dispersion length (α) is the linear coefficient. The apparent diffusion coefficient is extrapolated at zero electric field (as in the nature), and is thus given by the intercept of the Di - Vapp regression line with the ordinate axis. The value obtained for Dapp of selenate is 2.2 × 10-11 m2 s-1, as indicated on Fig. A9.3.5.

Di = Dapp +αVapp (eq. A9.3.1)

290/328

The Nernst-Einstein relationship makes the link between the diffusion coefficient in pure water and the ionic mobility (Atkins, 1983). To use this relationship in a porous medium, correction terms are introduced to account for sorption, tortuosity and electro-osmotic effects. For more details see Maes et al., (1999) and Beauwens et al., (2005).

The modified Nernst-Einstein relationship can be written as:

D0 ⎛ µeo ⎞ kT Dapp = = ⎜ µapp + ⎟ (eq. A9.3.2) Rτ 2 ⎝ Rτ 2 ⎠ Ze

Because the electrical double layer of clay minerals is enriched in cations, when a porous medium such as Boom Clay is submitted to an electric field, the water molecules associated to the cations present in the electrical double layer also accompany them towards the cathode: this displacement of water molecules under an electrical field gradient is known under the name of electro-osmosis. As anionic species move towards the anode under the action of the electric field, their migration is somehow counteracted by the electro-osmotic flow. The electro-osmotic mobility must then be added to the apparent mobility calculated from the activity peak's position in the clay.

As illustrated on Fig. A9.3.6, the “apparent” electromigration mobility (µapp) is given by the slope of the regression line obtained by plotting the apparent velocity (Vapp) as a function of the electric field (E) for a series of electromigration experiments. In order to calculate the

“true” electromigration mobility (µem), i.e., in the absence of electro-osmosis, we must add up -9 2 -1 -1 µapp and µeo. The value obtained for the slope of the regression line is 2.7 × 10 m s V , so the “true electromigration velocity” of selenate is here equal to 4.9 × 10-9 m2 s-1 V-1 as for the electro-osmotic mobility we take the value (2.2 × 10-9 m2 s-1 V-1) previously determined for tritiated water (HTO) by electromigration experiments in Boom Clay (Maes et al., 1999). Note that in equation A9.3.2 we assume that both HTO and selenate are unretarded, i.e. R = 1.

The apparent diffusion coefficient is then obtained by the Nernst-Einstein relationship given -11 2 -1 above. The value calculated by this method for Dapp is 6.2 × 10 m s , about three times higher than the value obtained with the hydrodynamic relationship. This is perhaps a consequence of the systematic correction term introduced for electro-osmosis. To overcome sampling variations inherent to the clay core itself, the electro-osmotic mobility should ideally be determined for each experiment by adding tritiated water to the source. It is also difficult to assert beforehand that both tracers (75Se and HTO) will always remain inside the clay core during the whole duration of the experiment and will never migrate into the reservoirs containing the electrolyte in which the electrodes are immersed. If a tracer comes out of the core and finally reaches one of the two electrolyte reservoirs, because of the continuous pumping made to neutralize the two compartments, it will inevitably re-enter the clay core

291/328

from the other side giving rise to some unexpected and very confusing migration pattern. Another point is that selenate could perhaps be slightly retarded, but here we assumed R = 1 for conservative reasons.

The values of Di and Vapp calculated by the two methods for each experiment apart are also given in Table A9.3.1. To do so, we used the previously published values for electro-osmotic mobility and dispersion length (Maes et al., 1999). Averaging over all experiments yields Dapp -11 2 -1 -11 2 -1 = 6.6 ± 1.8 × 10 m s with the Nernst-Einstein method and Dapp = 4.8 ± 2.0 × 10 m s with the hydrodynamic relationship method (see Table A9.3.1).

However the best way to derive a consistent Dapp is the method based on the linear regression between the apparent dispersion coefficient Di (sum of advection and diffusion terms) and the apparent electromigration velocity Vapp, according to equation A9.3.1. The apparent diffusion coefficient is extrapolated at zero electric field (as in the nature), and is thus given by the intercept of the Di - Vapp regression line with the ordinate axis. The value obtained for Dapp of selenate is 2.2 × 10-11 m2 s-1, as indicated on Fig. A9.3.5.

1.3E-10

Di (m/s²) 1.0E-10

7.5E-11

-4 -11 5.0E-11 y = 2.08×10 Vapp+ 2.19×10 R2 = 0.847

2.5E-11

0.0E+00 0.00E+00 1.00E-07 2.00E-07 3.00E-07 4.00E-07 5.00E-07

Vapp (m/s)

Figure A9.3.5: Estimation of the diffusion coefficient of selenate from the hydrodynamic relationship between the dispersion coefficient and the apparent advection velocity. Experiments EM_SeO3/2 (black square) and EM_SeO4/1 through EM_SeO4/7 (empty squares).

292/328

5.0E-07 Vapp (m/s)

4.0E-07

3.0E-07

V = 2.68 x 10-9 E 2.0E-07 app r2 = 0.874

1.0E-07

0.0E+00 0 20 40 60 80 100 120 140 160 180 Electric field (V/m)

Figure A9.3.6: Estimation of the electromigration mobility of selenate in Boom Clay from the relationship between advection velocity and electric field for experiments EM_SeO3/2 (black triangle) and EM_SeO4/1 to EM_SeO4/7 (grey squares).

293/328

Table A9.3.1: Experimental parameters and results of electromigration experiments made with selenite and selenate. (c) (d) Experiment Source Atmosphere I Immobile/Mobile Migration Electric Di Vapp Dapp Dapp Code Used Composition Se fractions Time Field × 10-11 × 10-7 × 10-11 × 10-11 (—) (—) (—) (mA) (%) (hours) (V m-1) (m² s-1) (m s-1) (m² s-1) (m s-1)

EM_SeO3/2 selenite Ambient air 8 93 – 7 47 74 6.9 1.5 5.3 1.5

EM_SeO4/1(a) selenate Ambient air 8 27 – 73 18 26 5.7 1.7 11.0 1.7

EM_SeO4/2(b) selenate Ambient air 8 69 – 31 21 67 4.8 1.6 6.0 1.

EM_SeO4/3(b) selenate Ambient air 1.2 69 – 31 163 11.5 3.8 0.3 5.6 0.26

EM_SeO4/4(b) selenate Ambient air 20 70 – 30 25 161 11.2 4.1 6.0 4.1

EM_SeO4/5(b) selenate Ambient air 15 57 – 43 24 123 10.2 3.7 6.6 3.7

(a) EM_SeO4/6 selenate 99.6 % Ar – 0.4 % CO2 5 0 – 100 76 45 3.0 1.3 6.4 1.3

(a) EM_SeO4/7 selenate 99.6 % Ar – 0.4 % CO2 12 0 – 100 21 97 6.8 2.6 6.2 2.6 Average ± std deviation 6.6 ± 1.8 4.8 ± 2.0 N-E H-R (a): selenite oxidized without NaOH addition (see text). (b): selenite oxidized with NaOH addition (see text). -9 (c): apparent diffusion coefficient calculated for each experiment apart using the Nernst-Einstein (N-E) method (equation A9.3.2), where µeo/Rτ² = 2.2 × 10 , and T = 298 K). (d): apparent diffusion coefficient calculated for each experiment apart by the hydrodynamic relationship (H-R) method, but using α = 8.47×10-5 m, as given by (Maes et al., 1999). N-E: Nernst-Einstein relationship. H-R: Hydrodynamic relationship. .

294/328

75 A9.3.3 Electromigration experiments with SeO3 – Boom Clay slurries

The aqueous Se species in equilibrium with Boom Clay is expected to be selenide (HSe–), whose concentration should be controlled by the solubility of solid phases such as FeSe2 or Se metal. These solids have a very low solubility in reducing conditions: 2 × 10-9 mol dm-3 for -7 -3 FeSe2 and 3 to 8 × 10 mol dm for Se(0) according to Maes et al., (2003). Batch experiments with selenite in Boom Clay slurries have shown that selenite undergoes a sorption/reduction process in contact with Boom Clay (Maes et al., 2003), while selenate remains in solution.

In order to assess the behaviour of selenite in equilibrium with Boom Clay by electromigration, a slurry was prepared in a controlled atmosphere (99.6 % Ar, 0.4 % CO2) by mixing 1.6 ml of 75Se stock solution (selenite) (composition previously given) in Boom Clay porewater with 2 g of fresh Boom Clay and 8.4 ml of Boom Clay porewater (solid-to-liquid ratio = 1/5) in a 15 ml Nalgene tube. The tube was left in the glovebox during one month under constant agitation. To get rid of selenate trace impurities present in the stock solution, the slurry was centrifuged (at 21 255 g during 30 min, T = 4 °C). The containing selenate supernatant was eliminated and the solid centrifugation slug was washed twice with Boom Clay porewater. The clay paste was then resuspended with 2 ml of Boom Clay porewater and the slurry formed was loaded in between two half clay cores, held in the middle of a 2 mm thick Teflon ring to avoid any slurry loss during the loading of the migration cell.

Two electromigration experiments were performed with the slurry as 75Se source. The experimental parameters are listed in Table A9.3.2, and the activity profiles of both experiments are shown on Fig. A9.3.7. As observed from these profiles, the main activity remains at the source position, we assume that selenite undergoes some kind of sorption followed by a reduction-precipitation process. The activity profile, more developed in the anode direction, indicates the gradual release of selenium, either by desorption of selenite, followed by sorption to the next sorbing sites, or by a solubility-limited release of selenide. If we assume that the solid phase controlling the selenium solubility is crystalline elemental selenium (Se(c)), it can however be calculated that the concentration of selenide in equilibrium with a redox potential of -350 mV (SHE) at a pH of 8.2 (Boom Clay in situ conditions) is as low as 10-14 mol dm-3. However, to detect such a low Se concentration, it is necessary to work with carrier-free 75Se produced in a cyclotron beam by a (p, n) nuclear reaction on a 75As target.

These experiments also demonstrate that no selenate is formed due to oxidation, despite the fact that these experiments took place in ambient atmosphere and lasted for relatively long times (see Table A9.3.2). The number of porewater renewal, i.e., the number of times water

295/328

moved by electro-osmosis through the whole length of each of the clay cores is given in Table A9.3.2 for indication.

75 Table A9.3.2: Number of porewater renewal in electromigration tests with SeO3 – Boom Clay slurries. Loaded Experiment I E Time Recovery Porewater 75Se activity Code # (mA) (V m-1) (hours) (%) Renewal (cps g-1)

EM_SeO3/5_BCS 647 8 60 ± 6 452 94 3.4

EM_SeO3/6_BCS 729 8 68 ± 5 675 85 4.7

1000 activity (cps/g clay) 100

10

cathode anode 1

0.1

0.01

0.001 -50 -40 -30 -20 -10 0 10 20 30 40 50 distance from source (mm) Figure A9.3.7: Activity profiles of electromigration experiments loaded with Boom Clay 75 2– slurries spiked with SeO3 . The experiments shown are respectively EM_SeO3/5_BCS (grey circles) and EM_SeO3/6_BCS (black squares). The dotted line represents the detection limit (0.06 cps g-1).

296/328

A10. Redox disequilibrium and reluctance of sulfate for reduction in deep clay formations

297/328

298/328

A10 Redox disequilibrium and reluctance of sulfate for reduction in deep clay formations

Under reducing conditions, at ambient temperature and in the absence of catalyst, or bacterial enzymatic activity, selenate is reluctant to reduction. In natural conditions, it is also often the case for the highest valence species of other common elements (e.g., carbonate, phosphate, arsenate, sulfate, perchlorate, …), because of the multiple electron transfers needed in the reduction reaction. Under in situ conditions prevailing in Boom Clay, the selenate reduction is uncertain due to kinetic limitations. Another reason could be that electron donors are present in insufficient quantity in the system, not enough accessible, too less reactive, or have been consumed by oxidation reactions and water radiolysis (or other repository-induced perturbations).

A well known case of redox disequilibrium is the persistence of sulfate often observed in deep reducing sediments. Under strongly reducing conditions prevailing in Boom Clay, sulfur is – expected to be present as HS aqueous species in equilibrium with FeS2 (pyrite) at a solubility -8 -7 –1 2– of about 10 to 10 mol L . However, SO4 is detected in “undisturbed” Boom Clay porewater at the Mol site at higher concentration, typically of the order of 1 mg L-1 (10-5 mol L-1) or more, i.e., two to three orders of magnitude above the expected sulfide concentration. Although this might be due to traces of sulfate produced by pyrite oxidation, a process virtually impossible to totally prevent during clay sampling and borehole drilling even under inert atmosphere, it could also reflect the true residual sulfate concentration present in the Boom Clay porewater. Because of the high content of Boom Clay sediments in natural organic matter (NOM), sulfate reducing bacteria (SRB) active at the sedimentation time (30 – 35 Ma) in the poorly consolidated clay below the seabed might have been able to transform most of the sulfate contained in the seawater into sulfide precipitated with Fe2+ as pyrite, but some sulfate might have also subsisted. At the beginning of the Quaternary era (2 Ma), hydrochemical changes from marine to fluviatile conditions have freshen the surrounding aquifers (Mazurek et al., 2008a,b; 2009). As a consequence most of seawater salts (mainly Cl– 2– and SO4 ) have diffused out of the clay formation in the fresh aquifers recharged by meteoritic water, leading to the low sulfate concentration observed today. However, elevated concentration of dissolved sulfate have been entrapped from sea water in other deep clay formations studied for radioactive waste disposal in Switzerland and in France. Higher concentration of sulfate have been measured in the Opalinus Clay (OPA) at Mont Terri Hard Rock Laboratory and in the Callovo-Oxfordian Clay (COx) at the Bure underground research laboratory.

In the porewater of well preserved undisturbed OPA Clay (half-diluted seawater, ~ 0.3 M Cl-), 2– – the observed SO4 to Cl concentration ratio is still this of seawater (Pearson et al., 2003;

299/328

De Cannière et al., 2008). In the COx Clay, due to diagenesis processes, sulfate concentration is now also determined by the solubility of celestite (SrSO4) (Gaucher et al., 2006). The presence of sulfate in ancient porewater of OPA Clay (180 Ma) and COx Clay (155 Ma) is a natural evidence that a fraction of sulfate can resist to reduction over geological time period and that hexavalent sulfur, S(VI), can still be present today and coexist with pyrite under strongly reducing conditions at depth. Another explanation, not incompatible with redox disequilibrium, is that an insufficient quantity of electron donor presently subsists in the system to completely reduce the sulfate inventory initially available in seawater.

Considering the chemical similarities between selenate and sulfate for their recalcitrance to reduction, we have thus to conclude that selenate could also subsist in deep geological formations under reducing conditions.

300/328

A11. Behaviour of redox-sensitive elements in a nitrate plume associated with bituminized MLW – The selenium case study

301/328

302/328

A11 Behaviour of redox-sensitive elements in a nitrate plume associated with bituminized MLW – The selenium case study

Oxidation and mobilization of selenium by nitrate is a known problem in irrigation drainage in Western US. Selenium (Se) can be oxidized by nitrate from irrigation on Cretaceous marine shale in the Western Colorado (Wright, 1999). Dissolved Se concentrations are – positively correlated with dissolved NO3 concentrations in surface water and ground water samples from irrigated areas. Redox conditions dominate in the mobilisation of Se in marine shale hydrogeological settings: dissolved Se concentrations increase with increasing Eh. Negative (i.e., favourable) ΔG (delta Gibbs free energies) values for the oxidation of Se by – – NO3 are obtained from theoretical calculations, indicating that NO3 can act as an electron acceptor for the oxidation of Se. Wright (1999) observed in laboratory batch experiments with shale suspensions under anaerobic conditions, in O2-free water, an increase of Se – concentrations over time with increased NO3 concentrations.

Concomitantly, Steinberg et al., (1992) and Oremland et al., (1999) have also observed that selenate can resist to microbial reduction in the presence of sufficiently high nitrate concentrations. With both ions present, nitrate reduction precedes selenate reduction. The presence of nitrate precludes the microbial reduction of selenate. Stock cultures of denitrifiers grow anaerobically on nitrate but not on selenate.

Nitrates are also commonly injected in oil fields for preventing sulfato-reduction by SRB bacteria and subsequent oil souring (H2S in petroleum) leading to oil degradation and severe corrosion-related problems in oil wells and transport pipelines.

These multiple evidences and independent lines of reasoning suggest that selenide could be oxidized or selenate be reluctant to reduction in Boom Clay affected by a nitrate plume released by the bituminized MLW (Eurobitum). Therefore, only the migration of the more mobile selenate species should be taken into account for safety calculations dealing with the disposal of Eurobitum MLW.

Specific batch and diffusion experiments with selenate and selenide in the presence of nitrate have thus to be performed to assess the compatibility of Eurobitum waste with Boom Clay. This topic is also presently being addressed in France by Andra (Achim Albrecht, personal communication) and discussed at the Mont Terri Rock Laboratory (Switzerland) in the frame of the Bitumen Nitrate (BN) experiment.

303/328

304/328

A12. List of Abbreviations

305/328

306/328

A12 List of Abbreviations

(am.) amorphous solid phase (cr.) crystalline solid phase (el.) elemental solid phase (met.) metallic solid phase “as is” as such, as supplied, untreated < d.l. below detection limit 14C-OM 14C-labelled Organic Matter A atomic mass number; A = 79 for selenium a.m.u. atomic mass unit a.o. amongst other AAS Atomic Absorption Spectrophotometry AC Anion Chromatography Alternate Current ADE Advection Dispersion Equation ADM Advection Dispersion Model ADS Accelerator Driven System Adduct Addition product AEA Atomic Energy Agency (Harwell, UK) AEAT AEA Technology, Ltd. (Harwell, UK) AECL Atomic Energy of Canada Limited AES Atomic Emission Spectrophotometry, Auger Electron Spectroscopy, or; Altered Evolution Scenario (see also NES and EES) AFM Atomic Force Microscopy AFS Atomic Fluorescence Spectrophotometry AHA Aldrich Humic Acid aka also known as AMD Acid Mine Drainage AMS Accelerator Mass Spectrometry Andra Agence Nationale pour la gestion des Déchets Radioactifs (France) AOM Amorphous Organic Matter Aq-SS Aqueous – Solid Solution ASIN Amazon Standard Identification Number ASR Alkali Silica Reaction ASV Anodic Stripping Voltammetry ATR Attenuated Total Reflectance

307/328

ATR-IR Attenuated Total Reflectance – Infrared Spectroscopy ATR-FTIR Attenuated Total Reflectance – Fourier Transformed InfraRed Spectroscopy ATSDR Agency for Toxic Substances and Disease Registry (US): http://www.atsdr.cdc.gov/toxfaq.html AVS Acid Volatile Sulfide B&B Baeyens and Bradbury BC Boom Clay BCE Boom Clay Extract: organic matter extracted from Boom Clay with synthetic Boom Clay water BCF Boom Clay Formation Biosphere Conversion Factor BCHA Boom Clay Humic Acid BCK Boom Clay Kerogen BCOM Boom Clay Organic Matter BCW Boom Clay Water = porewater = interstitial water BE Best Estimate BGR Bundesanstalt für Geowissenschaften und Rohstoffe (Deutschland) Federal Institute for Geosciences and Natural Resources (Germany) BGS British Geological Survey b.l.d Below limit of detection BLG Belgian open international report published by SCK•CEN BNL Brookhaven National Laboratory Bq Becquerel = 1 disintegration per second BRGM Bureau de Recherches Géologiques et Minières (Orléans, France) BSEM Back Scattering Scanning Electron Microscope (images) C2 Repository System Safety Function C2: Confinement; water tightness of waste package and overpack C3 Code used in the SCK•CEN migration laboratory to name a pure diffusion experiment with a radioactive source between two back-to-back clay cores C4 Code used in the SCK•CEN migration laboratory to name a percolation experiment with a radioactive source between two back-to-back clay cores ca circa, approximately, around, about CA Component Additivity CAM Component Additive Method (bottom up approach) CAS Chemical Abstract Service CCM Constant Capacitance Model

308/328

CD Charge Distribution CDB Citrate-Dithionite-Bicarbonate Fe(III) extraction buffer CE Capillary Electrophoresis CEA Commissariat à l’Energie Atomique (Paris, Fr) CEC Cation Exchange Capacity Cedra Société coopérative nationale pour l’entreposage de déchets radioactifs (Suisse) Ci Curie = 3.7 × 1010 Bq = activity of 1 g radium CIG Centre d’Informatique Géologique CNRS Centre National de la Recherche Scientifique (France) CRIEPI Central Research Institute of Electric Power Industry (Japan) CSC Conditional Stability Constant CSIC Consejo Superior de Investigaciones Científicas National Research Council of Spain cpm Count Per Minute cps Count Per Second CSH Calcium Silicate Hydrate D2 Code used in the SCK•CEN migration laboratory to name a percolation experiment with a radioactive source injected as a pulse Da Dalton (atomic mass unit) DAMRI Département des Applications et de la Métrologie des Rayonnements Ionisants (CEA, Direction des Technologies Avancées) DC Direct Current DCF Data Collection Forms DDLM Diffuse Double Layer Model DIC Dissolved Inorganic Carbon DL Double Layer DLM Double Layer Model DMDS DiMethyl-DiSelenide DMS DiMethyl-Selenide DOC Dissolved Organic Carbon DoE US Department of Energy, or; US Department of Environment DOI Digital Object Identifier: persistent identification number used by scientific publishers to univocally identify an electronic publication http://dx.doi.org/doi:number DOM Dissolved Organic Matter DR Diffuse Reflectance DRIFT Diffuse Reflectance Infrared Fourier Transformed spectroscopy

309/328

DS Design Specification; Data Set DSC Differential Scanning Calorimetry DTA Differential Thermal Analysis E Electrical field e.g. ex gratia, for example EBS Engineered Barrier System EC European Commission; Electrical Conductivity; Electronic Capture, e– capture EC 5thFP European Commission fifth Framework Program ed. Editor EDAX Energy Dispersive X-ray Analysis EDL Electrical Double Layer eds. Editors EDS Energy Dispersive x-ray Spectrometry EDX Energy Dispersive X-ray EDZ Excavation Damaged Zone EdZ Excavation disturbed Zone EES Expected Evolution Scenario (see also NES and AES) EFZ Excavation Failed Zone EG-BS Piezometer installed at Extension Gallery – Bottom Shaft

Eh redox potential measured with respect to Standard Hydrogen Electrode EIA Environmental Impact Assessment EM Electro-Migration EMP Ecole des Mines de Paris ENRESA Empresa Nacional de Residuos Radiactivos (Spain); Spanish radioactive waste management company EPA Environmental Protection Agency EPR Exchangeable Potassium Ratio; Electron Paramagnetic Resonance; European Pressurised Reactor eq. Equation ER Expert Range; External Report published by SCK•CEN ESR Exchangeable Sodium Ratio; Electron Spin Resonance ESRF European Synchrotron Radiation Facility (located in Grenoble, France) ETLM Extended Triple Layer Model

310/328

ETV Electro-Thermal Vaporisation ETV-ICP/MS Electro-Thermal Vaporisation coupled with Inductively Coupled Plasma Mass Spectrometry et al. et alii: and co-workers EVEREST EVEREST Project: Sensitivity Analysis of Geological Disposal Systems EXAFS Extended X-ray Absorption Fine Structure Spectroscopy FA Fulvic Acids FANC Federal Agency of Nuclear Control (in Belgium) FAO Food and Agriculture Organization of the United Nations FAP’s Fission and Activation Products FBR Fast Breeder Reactor FEMS Federation of European Microbiological Societies FEP’s Features, Events, and Processes FES Frayed Edge Sites FeSe Achavalite

FeSe2 Ferroselite FF Far Field Fig. Figure FNRS Fonds National de la Recherche Scientifique (Belgique) FoCa FoCa Clay from the Fourge Cahaine quarry in France FP Fission Products, or; European Framework Program (e.g., 5th FP, or FP-5) F-T Flow-Through FTIR Fourier Transform Infra Red spectroscopy FZK ForschungsZentrum Karlsruhe GmbH g Gravitational acceleration (g = 9.81 m s-2) GC Gas Chromatography Generalised Composite GC-MS Gas Chromatography – Mass Spectrometry GCA Generalized Composite Approach (top down approach) GDWQ Guidelines for drinking-water quality GEM Gibbs Energy Minimization gen. nov. Genus nova: new genus (microbiology) GF Graphite Furnace GPC Gel Permeation Chromatography GRS Gesellschaft für Anlagen- und Reaktorsicherheit GmbH (Deutschland) Institute for the safety of nuclear reactor (Germany) GSH-Px selenium-Glutathione peroxidase: antioxidant enzyme protecting the phospholipids membrane of living cells against peroxides free radicals

311/328

GSL Galson Sciences Ltd. GTA Groupe de Travail Architecture GTS Grimsel Test Site GWB The Geochemist’s Workbench (geochemical code) HA Humic Acids HATCHES Harwell thermodynamic database used by AEA Technology HFO Hydrous Ferric Oxide HG Hydride Generation HLW High Level Waste HPLC High Performance Liquid Chromatography HR High Resolution H-R Hydrodynamic Relationship HR-ICP-MS High Resolution Inductively-Coupled Plasma Mass Spectrometry HRL Hard Rock Laboratory (Aspö) HS Humic Substances (not recommended in sulfide-containing systems because very confusing with HS–, Hydrogeno-Sulphide, bisulphide). Use better HA or OM instead of HS. HS– Hydrogeno-Sulphide, bisulphide HSAB Hard and Soft (Lewis) Acids and Bases (principle) HTO Tritiated water HTR High Temperature Reactor I Current IAP Ion Activity Product ID Inner Diameter i.e. id est I/S Illite/Smectite (interstratified Mixed Layer, ML) IAEA International Atomic Energy Agency (Vienna, Au) IC Ion Chromatography ICP Inductively-Coupled Plasma ICP-AES Inductively-Coupled Plasma Atomic Emission Spectrophotometry ICP-OES Inductively-Coupled Plasma Optical Emission Spectrophotometry ICP-MS Inductively-Coupled Plasma Mass Spectrometry ICRP International Commission of Radio-Protection INAA Instrumental Neutron Activation Analysis INRA Institut National pour la Recherche Agronomique (France) IRF Instant Released Fraction (gap inventory in spent fuel) IRMM Institute for Reference Materials and Measurements IS Ionic Strength Interstratified Illite/Smectite

312/328

Inner-Sphere (complex) ISBN International Standard Book Number ISSN International Standard Series Number ISO International Standard Organisation ISO/CD ISO Committee Draft ITU European Institute for Transuranium elements (Karlsruhe, Germany) IUPAC International Union of Pure and Applied Chemistry JAEA Japan Atomic Energy Agency (ex JNC, ex PNC) JNC Japan Nuclear Cycle Development Institute JRC Joint Research Centre K Equilibrium constant (log K) KAPL Knolls Atomic Power Laboratory (KAPL Inc.) KBD Kashin-Beck Disease (endemic osteoarthropathy due to deficiency in selenium in China) KD Keshan Disease (endemic cardiomyopathy due to deficiency in selenium in China)

Kd Distribution coefficient, assuming linear and reversible equilibrium

between solid and solution phases (see also Rd)

Ks Equilibrium constant KULeuven Katholieke Universiteit van Leuven L Liquid; Litre (dm3) LC Liquid Chromatography LC-MS Liquid Chromatography-Mass Spectrometry LDH Layer Double Hydroxide; synonym: MDL, Mixed Double Layer LFER Linear Free-Energy Relationship LIBS Laser Induced Breakdown Spectroscopy LIPS Laser Induced Plasma Spectrometry LILW Low and Intermediate Level Waste LLNL Lawrence Livermore National Laboratory LLW Low Level Waste LNHB Laboratoire National Henri Becquerel (CEA, France) LoD Limit of Detection; Level of Detection L/S Liquid-to-Solid ratio (see also S/L) LSC Liquid Scintillation Counting LTE Long Term Evolution LU Loughborough University

313/328

LWR Light Water Reactor M/L Mixed Layer (interstratified Illite/Smectite Mixed Layer) e.g. hydrotalcite (Mg2+ /Al3+); or green rust (Fe2+ /Fe3+) M2 Praclay M2 Mixture: 65 % FoCa clay + 30 % sand + 5 % graphite MALDI Matrix Assisted Laser Desorption-Ionization TOF MS Time-of-Flight Mass Spectrometry MAS Magic Angle Spinning (see NMR) MC-ICP-MS Multiple Collection Inductively-Coupled Plasma Mass Spectrometry MDL Mixed Double Layer (interstratified Illite/Smectite Mixed Layer) e.g. hydrotalcite (Mg2+ /Al3+); or green rust (Fe2+ /Fe3+) synonym: LDH, Layer Double Hydroxide MICOF Migration Code in Fortran (with analytical solutions developed by Martin J. Put at SCK•CEN; collection of several programs used in the Migration team to fit data from migration experiments) MLW Medium Level Waste MMA Museum for Middle Africa, Tervuren, Belgium NMR Nuclear Magnetic Resonance

MOX (PuO2/UO2) Mixed OXide based nuclear fuel MRS Materials Research Society MS Mass Spectrometry MTP Mont Terri Project MUSIC MUltiple SIte Complexation MW Molecular Weight MWCO Molecular Weight Cut-Off NAA Neutron Activation Analysis n.a. not analysed, or, not accounted for Nagra National Cooperative for the Disposal of Radioactive Waste (Switzerland) Société coopérative nationale pour l’entreposage de déchets radioactifs (Suisse) Nagra Nationale Genossenschaft für die Lagerung radioaktiver Abfälle (Schweiz) n.d. not detected, or, not determined ? very ambiguous abbreviation if not clearly defined in its own context ! NDA Nuclear Decommissioning Authority n.d.a non destructive analysis n.f. not filtered n.m. not measured

314/328

NDS Nuclear Data Sheet N-E Nernst-Einstein relationship NES Normal Evolution Scenario (see also EES and AES) NAS National Academy of Sciences, or, Natural Analogue Studies NBS National Bureau of Standard NDA Nuclear Decommissioning Authority (UK) non destructive analysis NEA Nuclear Energy Agency (Paris, Fr) NERC National Environment Research Council (UK) NES Natural Evolution Scenario NF Near Field NIH National Institutes of Health (US) NIRAS Nationale Instelling voor Radioactief Afval en Verrijkte Splijtstoffen (België) Belgian agency for radioactive waste and enriched fissile materials NIREX Nuclear Industry Radioactive Waste Executive (United Kingdom) NIROND NIRAS/ONDRAF NIST National Institute of Standards and Technology NMWCO Nominal Molecular Weight Cut-off NNDC US National Nuclear Data Center NOM Natural Organic Matter NPP Nuclear Power Plant NRC National Research Council of Canada NTB Nagra Technischer Berichte NuDat NuDat 2.1 Nuclear Database from the US National Nuclear Data Center (NNDC), Brookhaven National Laboratory (BNL)URL: http://www.nndc.bnl.gov/nudat2, as seen on 02-Feb-2005. OD Optical Density; Outer Diameter ODS Office of Dietary Supplements (US): http://ods.od.nih.gov/factsheets/Selenium_pf.asp OECD Organisation for Economic Cooperation and Development (Paris, Fr) OES Optical Emission Spectrophotometry OM Organic Matter OMS Organisation Mondiale de la Santé ONDRAF Organisme National des Déchets Radio-Actifs et des matières Fissiles enrichie (Belgique) Belgian agency for radioactive waste and enriched fissile materials

315/328

OPG Ontario Power Generation ORNL Oak Ridge National Laboratory ORP Oxidation-Reduction Potential OS Outer-Sphere (complex) OSHA Occupational Safety and Health Administration (US) OSTI US Office of Scientific and Technical Information P Pressure PA Performance Assessment PACOMA Performance Assessment of the Geological Disposal of Medium-level and Alpha waste in a clay formation in Belgium PAGIS Performance Assessment of Geological Isolation Systems PAR Potassium Adsorption Ratio PDF Probability Density Function; Portable Document Format (Adobe) – pE -log [e ] = (F × Eh ) / 2.3 RT PEC Proton Exchange capacity PEEK Poly-Ether-Ether-Ketone PES Poly-Ether-Sulfone PD Pure Diffusion experiment (i.e., without advection due to a hydraulic gradient) pH -log [H+] = potential of hydrogen, hydrogen pondii PHREEQC pH Redox Equilibrium-C code PI Principal Investigator PIXE Proton Induced X-ray Emission pKa -log Ka = -log [equilibrium constant] POC Particulate Organic Carbon PORFLOW Numerical transport code developed by Runchal (Acri, California). Code used by the Performance Assessment team for the safety studies) POSIVA Finnish radioactive waste management company Organization responsible for radioactive waste management in Finland pp. pages PSI Paul Scherrer Institute (Switzerland) P&T Partitioning and Transmutation PWR Pressurised Water Reactor PXAMS Projectile X-ray Accelerator Mass Spectrometry Py Pyrolysis Py-GC-MS Pyrolysis Gas Chromatography – Mass Spectrometry PZC Point of Zero Charge PZNPC Point of Zero Net Proton Charge

316/328

PZSE Point of Zero Salt Effect Q quantity R Retardation factor; R-####: Restricted report number R2 Repository System Safety Function R2: geological barrier; Retention of radionuclide and spread in time RBCW Real Boom Clay Water = Boom Clay porewater RCW Real Clay Water (with organic matter)

Rd Distribution coefficient, not assuming equilibrium between solid and

solution phases (see also Kd) R&D Research and Development RDD Research, Development and Demonstration REV Representative Elementary Volume RIC Real Interstitial Clay Water RIS Research Information System (citation format for Refman and ProCite) RN Radionuclides RP Research Plan RSF Radial Structure Functions (terminology used for XANES and EXAFS spectroscopy) S atomic symbol of the sulfur element (z = 16; A = 32), or also; Solubility; Solid S/C Sulfur-to-Carbon ratio S/L Solid-to-Liquid ratio (see also L/S) SA Surface Area, Specific Area; Safety Assessment SA/V Surface Area / Volume SAFIR 2 Safety Assessment and Feasibility Interim Report 2 SAR Sodium Adsorption Ratio; Semi-Annual Report SBCW Synthetic Boom Clay Water = artificial preparation according a recipe (without organic matter) SC Super Container; Surface Complexation SCE Saturated Calomel Electrode SCK•CEN StudieCentrum voor Kernenergie – Centre d’Etude de l’énergie Nucléaire SCM Surface Complexation Modelling SCN Thiocyanate SCW Synthetic Clay Water (without organic matter)

317/328

SD Standard Deviation Se atomic symbol of the selenium element (z = 34; A = 79) SEC Size Exclusion Chromatography SEM Scanning Electron Microscope SEP Solid Extraction Phase SF Spent Fuel SFC Safety and Feasibility Case SHE Standard Hydrogen Electrode SI Saturation Index = log (Q/K); Système International d’unité, International unit System SIC Synthetic Interstitial Clay Water SMEP Solid Micro-Extraction Phase SKB Svensk Kärnbränslehantering AB Swedish Institute for the management of spent fuel SKI Swedish Nuclear Power Inspectorate SM Standard Methods sp. nov. Species nova: new species (microbiology) SPA Spent fuel Performance Assessment (European Project) SPA+ Spent fuel Performance Assessment Plus (SPA report updated for ONDRAF/NIRAS) SR Synchrotron Radiation; Source Range SRB Sulfato-Reducing Bacteria SRM Standard Reference Material SSRL Stanford Synchrotron Radiation Laboratory STG Science and Technology Group SXRF Synchrotron X-Ray Fluorescence T Temperature TD Through-Diffusion TDB Thermodynamic DataBase TDS Total Dissolved Solids, or Total Dissolved Salts TGA Thermo-Gravimetric Analysis TIC Total Inorganic Carbon TIMS Thermal Ionisation Mass Spectrometry TLM Triple Layer Model, see also ETLM TN Technical Note TOC Total Organic Carbon TOF-MS Time-of-Flight Mass Spectrometry TR Technical Report

318/328

TRANCOM TRANsport of radionuclides by Complexation with Organic Matter. European project: phase 1: 4th FP; phase 2: 5th FP. TRIS tris(hydroxymethyl)methylamine, or; tris(hydroxymethyl)aminomethane; TRIS buffer TROM TRancom Organic Matter (concentrated batch of OM used for labelling with 14C) TRU TRansUranic Waste TSM Thermodynamic Sorption Model TTK (Millipore ultrafilter unit code number) UA University of Antwerp UF Ultra Filtration u.f. ultra-filtered UK United Kingdom UKAEA UK Atomic Energy Authority UNO United Nation Organisation (New York, USA)

UOX-55 UO2 uranium oxide based nuclear fuel with a burn-up of 55 MW d/tHM URF Underground Research Facility URL 1. Underground Research Laboratory 2. Uniform Resource Locator: web address on the internet UPD90 Updating 1990, BLG-634 Report (Marivoet J., 1991) US United State USA United State of America USGS United State Geological Survey UV Ultra-Violet Vis Visible VITO Vlaamse Instelling voor Technologische Onderzoek (Mol, Belgium) VOC Volatile Organic Carbon vs versus VTT Technical Research Centre of Finland WAC Waste Acceptance Criteria WHO World Health Organisation of the United Nations WP Work Package XAFS X-ray Absorption Fine Structure XANES X-ray Absorption Near Edge Spectroscopy XAS X-ray Absorption Spectroscopy XRD X-ray Diffraction XRF X-ray Fluorescence XSW X-ray Standing Wave

319/328

YMP Yucca Mountain Project z atomic number = number of proton = number of electron in an atom; z = 34 for selenium zpc zero point of charge ZVI Zero-Valent-Iron: metallic Fe, Fe0

320/328

A13. List of Symbols

321/328

322/328

A13 List of Symbols

α dispersion length μ electro-mobility, or; micro- η diffusion accessible porosity of the species in the porous medium

μapp apparent electro-mobility

ρd dry density of the porous material

μeo electro-osmotic mobility A activity (Bq) C concentration of the considered species in the water accessible by diffusion

C(site) concentration of uncomplexed sorption sites

Caq concentration of aqueous selenium (aqua: dissolved in water)

Cbulk concentration in the bulk of the porous material

Csor concentration of selenium associated with the solid

D0 molecular diffusion coefficient

Dapp apparent diffusion coefficient obtained after fitting of data from diffusion experiments

Daq molecular diffusion coefficient of the species in pure water at 25 °C.

Deff effective diffusion coefficient determining the flux of the species across a

section. Deff = ηRDapp

Di apparent dispersion coefficient

Dp pore diffusion coefficient of the species, diffusion coefficient of the species in the porous medium

G geometrical factor G = 1/Rf : inverse of the rock factor, or formation factor K hydraulic conductivity

Ka association equilibrium constant 3 -1 Kd linear distribution coefficient (dm kg = 1 L/kg = 1 ml/g)

Kh hydraulic conductivity in the horizontal direction (parallel to bedding)

Ks stability constant for dissolution / precipitation equilibrium; dissolution constant

Kv hydraulic conductivity in the vertical direction (perpendicular to bedding) R retardation factor of the species in the porous medium. R = water velocity / species velocity

Rd distribution ratio: non linear distribution coefficient (dm3 kg-1 = L/kg = ml/g)

323/328

Rf rock factor, or formation factor: inverse of the geometrical factor (G), taking into account the tortuosity and the constrictivity of the porous 2 medium. Rf = τ S solubility limit; do not confuse with the atomic symbol of the sulfur element S (z = 16; A = 32)

T½ radioactive half-life; radioactive period

Vapp apparent velocity

VDarcy Darcy velocity wt. % weight percent x position coordinate σ sigma, standard deviation (see also Gaussian distribution, average)

σc consolidation pressure

σe effective stress σe = σc - σw

σw porewater pressure τ tortuosity

324/328

A14. List of Physical Constants and Units

325/328

326/328

A14 List of Physical Constants and Units

°C Celsius grade a anno = year = y A ampere A Atomic mass Å Angstrom = 10-10 m

5 -2 atm 1 atmosphere ≃ 10 Pa ≃ 100 kPa ≃ 1 bar ≃ 0.1 MPa ≃ 1 kg cm Bq Becquerel = 1 disintegration per second Ci 1 Curie = 3.7 × 1010 Bq = activity of 1 g radium cpm Count Per Minute cps Count Per Second d day Da Dalton (atomic mass unit) (see MWCO) dm deci-meter = 10-1 m dm3 cubic deci-meter = 1 L = 1 liter e charge of proton = 1.60219 × 10-19 coulomb eV electron Volt F Faraday constant = F = eL = 96 485 Coulomb mol-1 g gram, or, gravitational acceleration (g = 9.81 m s-2) GWd GigaWatt-day (energy produced by a nuclear power plant of 1 GW in one day) GWd/tHM GigaWatt-day/ ton Heavy Metals: unit expressing the nuclear fuel burn-up h hour k Boltzmann constant = 1.38066 × 10-23 J K-1 ka kilo anno = kilo years = ky kDa kilo-Dalton (atomic mass unit) (see MWCO) keV kilo electron Volt kg kilogram = 103 g L Avogadro number = 6.023 × 1023 mol-1 m meter M molar concentration (1 mol dm-3) N normal concentration (1 eq. dm-3) dm deci-meter = 10-1 m cm centi-meter = 10-2 m mm milli-meter = 10-3 m μm micro-meter = 10-6 m

327/328

nm nano-meter = 10-9 m pm pico-meter = 10-12 m Ma Mega Anno = Mega years = My MBq Mega-Becquerel meq. milli-equivalent MeV Mega electron Volt min. minute mL milli-litre = 1 cm3 mol mole: quantity containing Avogadro number (6.023 × 1023) of molecules N Number of neutron Pa Pascal = Newton / m2 – pe -log [e ] = (F × Eh ) / 2.3 RT ppb part per billion (10-9); 1 μg kg 1 ppm part per million (10-6); 1 mg kg-1 ppt part per trillion (10-12); 1 ng kg-1 R gas constant: R = kL = 8.31441 J mol-1 K-1 s second T absolute Temperature (Kelvin) tHM ton Heavy Metals (of nuclear fuel) V Volt W Watt = 1 joule s-1 z number of electrical charges Z number of proton

328/328