Radionuclide Dissociation from Bentonite Colloid Systems A thesis submitted to the University of Manchester for the degree of Doctor of Philosophy (PhD) in the faculty of Engineering and Physical Sciences 2015 Nicholas Sherriff School of Chemistry List of Contents Abstract 8 Declaration 10 Copyright Statement 11 Acknowledgements 12 Thesis description and layout 14 Chapter 1 Introduction 15 1.1 Introduction to Nuclear Power 16
1.1.1 The Nuclear Fuel Cycle 16 1.1.2 The “front end” of the nuclear fuel cycle 16 1.1.3 The “back end” of the nuclear fuel cycle 17 1.1.4 High Level Nuclear Waste Disposal 20 1.1.5 Deep geological disposal in other EU countries 21 1.2 Introduction to BELBaR 24 1.3 Multiple barrier system 25 1.4 Backfill material and purpose 26 1.5 Bentonite properties and characteristics 27 1.6 Bentonite used in the thesis 28 1.6.1 Type of Bentonite and chemical formula 28 1.6.2 Bentonite used in study characterisation 30 1.7 Colloids 32 1.7.1 Colloids defined 32 1.7.2 Colloid stability 32 1.7.3 The colloid ladder 33 1.8 Previous radionuclide/bentonite binding studies 35 35 1.8.1 Bulk bentonite studies 2
1.8.2 Modelling 35 1.8.3 Bentonite colloids and reversibility 37 1.8.4 Dissociation Rates 41 1.8.5 Colloid formation other than erosion 43 1.9 Radioisotopes used in this study 44 1.9.1 Europium 44 1.9.2 Uranium 45 1.9.3 Thorium 46 1.9.4 Americium 48 1.10 Aims and objectives 49 1.11 References 50 Chapter 2 Scientific Techniques 55 2.1 Experimental Techniques 56 2.1.1 Batch experiment approach 56 2.1.2 Centrifugation 57 2.1.3 Colloid Generation 58 2.1.4 Ultrafiltration 59 2.1.5 232U(VI) and 228Th(IV) purification 60 2.2 Solution Speciation Modelling 62 62 2.3 Experimental analysis techniques 63 63 2.2.1 PHREEQC 2.3.1 ICP-­‐AES 3
2.3.2 Liquid scintillation counting 64 2.3.3 Gamma Ray Spectrometry 66 2.3.4 Scanning Electron Microscopy (SEM) 67 2.4 References 69 Chapter 3 Reversibility in radionuclide/bentonite bulk and colloidal ternary systems 71 Chapter 4 Uranium dissociation from bentonite colloids, a kinetic investigation 97 Chapter 5 An investigation into thorium and americium dissociation from bentonite colloids 121 Chapter 6 Conclusions made from thesis and suggested further work 152 6.1 Conclusions 152 6.2 Further Work 157 159 Appendix: Thermodynamic speciation data 4
List of Figures Chapter 1 1.1 Schematic of the nuclear fuel cycle 19 1.2 Schematic of a proposed deposition hole 26 1.3 Structure of bentonite clay 27 1.4 XRD of bentonite 30 1.5 The colloid ladder 33 1.6 Hydrolysis constant correlation 37 1.7 Desorption graphs for Eu(III) and Th(IV) 40 1.8 Radionuclide within colloid problem 43 Chapter 2 2.1 Typical ultrafiltration cell 59 2.2 U(VI) and Th(IV) LSC alpha spectra 61 2.3 Schematic of ICP-­‐AES 63 2.4 Schematic of the components involved with LSC analysis 65 2.5 Schematic of an SEM 67 3.1 Dissociation plot of Eu and bulk bentonite 79 3.2 Colloid and Eu(III) concentrations in each size fraction 84 3.3 Natural log dissociation plot of Eu and colloidal bentonite 87 Chapter 3 Chapter 4 4.1 Colloid suspension viewed via SEM 106 4.2 A single colloid particle viewed via SEM 106 4.3 Colloid and U(VI) concentrations in each size fraction 107 4.4 Natural log dissociation plot of U and colloidal bentonite 110 5
Chapter 5 5.1 Colloid and Th(IV)/Am(III) concentrations in each size fraction 133 5.2 Dissociation plot of Th and colloidal bentonite 138 5.3 Th dissociation after 1 day Cellphos contact 139 5.4 Plot of the Am(III)/colloid dissociation experiment 141 6
List of Tables Chapter 1 1.1 Geodisposal in other EU countries 22 1.2 XRF elemental data for bentonite 31 1.3 List of estimated dissociation rate constants 41 Chapter 3 3.1 Dissociation data for Eu/bulk bentonite 81 3.2 Eu speciation in the bulk system 83 3.3 Eu speciation in the bulk system with EDTA 83 3.4 Eu speciation in the colloid system 85 3.5 Dissociation data for Eu/colloidal bentonite 88 109 111 Chapter 4 4.1 Uranium speciation data 4.2 Dissociation rate data for U/colloidal bentonite Chapter 5 5.1 Thorium speciation data 135 5.2 Americium speciation data 135 5.3 Saturation indices 136 Chapter 6 6.1 Summary of findings 153 6.2 Colloid surface area 153 7
Abstract Deep geological disposal is a method of managing high level, long-­‐lived nuclear waste. It is a concept that many countries are exploring for the possibility of managing nuclear waste generated from power production. For deep geological disposal to be viable then areas where problems may surface have to be explored. Bentonite clay has been proposed as the material to be used for the backfill of the repositories. Its swelling properties ensure that it will expand to plug the bore holes that will be made for the waste, its impermeable nature restricts contact between groundwater and the waste package and its stability on a geological timescale all make it desirable as a backfill material. This project looks at the role that colloids formed from the bentonite clay could have in facilitating radionuclide transport away from a nuclear waste repository. Several radionuclides (Eu(III), U(VI), Th(IV) and Am(III)) have been considered in this research, and information from these studies will be used in the BELBaR project’s outputs, which will eventually support a disposal safety case. Ternary systems of 152Eu(III), bulk bentonite and EDTA ([Eu] = 7.9 x 10-­‐10 M; pH = 6.0 – 7.0) have been studied. Without EDTA, there was slow uptake in a two-­‐stage process, with initial rapid sorption of Eu(III) (96%), followed by slower uptake of a smaller fraction (3.0 % over a period of 1 month). The reversibility of Eu(III) binding was tested by allowing Eu(III) to sorb to bentonite for 1 – 322 days. EDTA was added to the pre-­‐equilibrated Eu bentonite systems at 0.01 M. A dissociation rate constant of approximately 4.3 x 10-­‐8 s-­‐1 (values in the range 2.2 x 10-­‐8 – 1.0 x 10-­‐7 s-­‐1) for pre-­‐equilibration times ≥ 7 days was measured. Eventually, the amount of Eu(III) remaining bound to the bentonite was within error of that when EDTA was also present prior to contact (4.5 % ± 0.6). Eu interactions with colloidal bentonite were studied, and the dissociation rate constant measured by a resin competition method. A dissociation rate of 8.8 x 10-­‐7 s-­‐1 and a range of 7.7 x 10-­‐7 – 9.5 x 10-­‐7 s-­‐1 were measured. For both bulk and colloidal bentonite slow dissociation was observed for Eu(III), but there was no evidence for ‘irreversible’ binding. The interactions of 232U(VI) with bentonite colloids ([U] = 5.43 x 10-­‐10 M; pH = 8.8 ± 0.2) have been studied using a resin ion exchange competition technique. The reversibility of the interaction was studied by allowing U(VI) to sorb to bentonite colloids for periods from 1 – 35 days. A fraction of the U(VI) was removed from the solution instantaneously (28-­‐50 %), and after 3 days, the amount of U(VI) remaining on the bentonite colloids was 17-­‐ 25%. With time, the amount of U(VI) retained by the bentonite colloid is reduced further, with a first order dissociation rate constant of 5.6 x 10-­‐7 s-­‐1. Whilst the dissociating fraction was small (24% (+34; -­‐12 %)), complete dissociation was not observed. Although slow dissociation was observed for U(VI), there was no convincing evidence for ‘irreversible binding’ of the radionuclide by the colloid. The interactions of 228Th(IV) ([Th] = 3.79 x 10-­‐12 M; pH = 8.8 ± 0.2) and 241Am(III) ([Am] = 3.27 x 10-­‐9 M; pH = 8.8 ± 0.2), with bentonite colloids have been studied using an ion exchange competition technique. Th(IV) was not fully associated with the bentonite colloids, and filtration showed that the uptake after 1 week was 78.3% (± 2.7%). Am(III) was weakly associated to the bentonite colloids, the uptake after 1 week was 20.1 % (± 5.2 %). Cellulose phosphate was added to the radionuclide/bentonite colloid systems (1 g for Th(IV), 0.2 g for Am(III)), an amount that was sufficient to retain the radionuclide when no bentonite colloids are present. A fraction of the Th(IV) is initially removed by the Cellphos (75-­‐93 %), and after 7 days the amount of Th(IV) remaining on the colloids is 1 -­‐ 3 %. Over the time of the experiment, the amount of Th(IV) retained by the bentonite colloid appears to remain level and the amount bound to the bentonite colloid at the end of the experiment is 2.1 % ± 0.88 % which is within experimental error of the steady state equilibrium of the system. A fraction (48-­‐94 %) of the Am(III) is also initially removed by the Cellphos, after 7 days the 8
amount of Am(III) remaining on the colloids is 1.2 – 9.3 %. However, after 35 days of contact time with the cellulose phosphate it appears that Am(III) is released back into the system, preventing dissociation rates from being calculated in this case. Studies of the association of Eu(III) to the clay colloids and its subsequent dissociation in this thesis follow similar trends to those described elsewhere in the literature (Missana et al. (2008), Bouby et al. (2011)). The Eu/bentonite colloid dissociation rate calculated here (8.8 x 10-­‐7 s-­‐1 (± 9.1 x 10-­‐7 s-­‐1)) is within error of the dissociation rates for trivalent ions estimated by Wold (2010) (Am(III) 5.6 x 10-­‐7 s-­‐1 Cm(III) 1.7 x 10-­‐6 s-­‐1). The U(VI) studies in this thesis show a dissociation rate of 5.6 x 10-­‐7 s-­‐1 (± 4.2 × 10-­‐7) which is within error of the U(VI) dissociation rate estimated by Wold (2010) (8.3 X 10-­‐7 s-­‐1). Reliable dissociation rates could not be obtained from the Am(III) and the Th(IV) studies in this thesis, other studies (e.g. Bouby et al. (2011) showed signs of irreversible binding of Th(IV) to bentonite colloids, however, no irreversible binding was observed in this thesis. Am(III) did not appear to be a close analogue of Eu(III) in these systems. All of the isotopes studied in this thesis showed no evidence of irreversible binding to bentonite or bentonite colloids. As such, the role that bentonite colloids will have in the facilitated transport of radioisotopes away from a repository is likely to have only a limited impact, at most, on the environmental safety case. 9
Declaration No portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning. 10
Copyright Statement 1) The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the “copyright”) and he has given The University of Manchester certain rights to use such copyright, including for administrative purposes. 2) Copies of the thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. 3) The ownership of certain Copyright, patents, designs, trademarks and any or all other intellectual property (the “Intellectual Property Rights”) and any reproductions of copyright works in the thesis, for example graphs and tables (“Reproductions”), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property Rights and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property Rights and/or Reproductions. 4) Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see http://www.campus.manchester.ac.uk/medialibrary/polices/intellectual-­‐property.pdf), in any relevant Thesis restriction declarations deposited in the University Library’s regulations (see http://www.manchester.ac.uk/library/aboutus/regulations) and in The University’s policy on presentation of Theses. 11
Acknowledgements I never in my wildest dreams thought that someone such as myself would be able to gain a qualification with such prestige as a PhD, I of course could not have done this without help and this is where I let the people close to me know how their support, influence and tolerance have shaped this project. Firstly my children Sam and William, the reason I have persevered with University at all, let alone the PhD, their smiles and laughter is enough to fuel even the weariest mind, I love them both dearly and this is ultimately for them. Nick Bryan who gave me the chance and enabled me to undertake this project and has shown patience and understanding in all that I’ve done, I could never thank him enough for his guidance and help. Francis Livens who has been a great supervisor has helped me immensely, and took me on when he really didn’t have to (and treats me to lots of cake and coffee!) and Sarah Heath for her friendship and reassurance and laughter and also Gareth and Katie for their endless support, kind words and guidance. All of the people I’ve met in radiochemistry but especially Adam, Matt, Sean, Tamara, Tucker, Tony, Kurt and Jones for the laughs I’ve had and the caffeine I’ve consumed with them. Also a mention to Rosie, Adam, Nick and Graham who sits in our windowless office, many a laugh has been had. My family and friends, especially, my Mother, whose influence has shaped who I am, and although she died whilst I was undertaking this project I know she would have been proud and keen for me to see it through to the end. Ken, Richard, Pat, Tony and Andy who have all helped either financially or looking after the kids. Special mention to my oldest friend Will who has always been there for me and never stopped believing that I could do this. 12
My beloved Manchester City for providing me with endless banter and occasional bragging rights in the office, also the Monday night footy lot for giving me something to kick. And mostly I want to thank my wife Rebecca, who without, none of this could have been achieved, ultimately she deserves as much credit for this as I do. She has been my counsellor, teacher, friend and support for so many years now it is difficult to think that any kind of achievement could be met without her. Her love has kept me going and her strength and understanding have kept me grounded and sane, I will always be in her debt. Thank you all. 13
Thesis description and layout This thesis has been submitted in an alternative format, which means that, although it still has chapters, several of those chapters are publications (or in publication style). Chapters 1 and 2 follow a more traditional approach; chapters 3, 4 and 5 are separate publications that describe the research conducted in this thesis. Chapter 6 concludes the thesis and offers suggestions for further work. 14
Chapter 1 Introduction The content of Chapter 1 will be used for the BELBaR report and will present data which could support a safety case, it will introduce the reader to the principle of the project and past literature which is relevant to project. 15
1.1 Introduction to Nuclear Power 1.1.1 The Nuclear Fuel Cycle The Nuclear Fuel Cycle is the term used when describing the production of electricity via nuclear fission in its entirety, that is to say from mining the uranium needed, to the treatment and eventual disposal of radioactive waste. The cycle can be broken down in to two distinct areas, the “front end” which encompasses the mining, enriching and usage of nuclear fuel, and the “back end” which encompasses the treatment/disposal of spent nuclear fuel, and the decommissioning of obsolete facilities and the treatment/disposal of mining waste. 1.1.2 The “front end” of the nuclear fuel cycle Uranium occurs in most rocks with a concentration of around 2 – 4 ppm and, almost always, isotopic composition atom 99.2175 % 238U, 0.72 % 235U and 0.0055 % as 234U. The mined ore is milled and leached to extract the uranium and the uranium is recovered by solvent extraction. The wastes from mining and ore processing contain over 99 % by mass of the material mined, including all of the daughter products from the uranium, and must be treated accordingly (Tolenko, 2012). The uranium product is known as “yellowcake” (U3O8), which is ready to move onto the next stage of the cycle, which is enrichment. As 235U is the isotope that is needed for nuclear fission and makes up such a small percentage of the mined uranium then enrichment is needed to make the fuel usable. Before enrichment the U3O8 must be converted to uranium hexafluoride (UF6), it is in this form that enrichment can take place, conventionally in one of two ways. 16
Diffusion can be used to enrich the fuel since the diffusion rates of 235UF6 and 238UF6 through a membrane are very slightly different; this requires many stages of diffusion and uses a lot of energy to perform (Tolenko, 2012). Alternatively there is gaseous centrifugation, which whilst using less energy can be more challenging technically (Wilson, 1996). The enriched UF6 can then be processed to UO2, which can then be pressed into ceramic pellets, and assembled within a nuclear fuel assembly ready for use. 1.1.3 The “back end” of the nuclear fuel cycle Once the nuclear fuel within a reactor has had its useful energy extracted it is referred to as spent nuclear fuel, and has to be removed from the reactor and either has to be stored, reprocessed or disposed of. Most nuclear power is based on an open fuel cycle i.e. once the fuel is spent, it is removed as nuclear waste (Turner, 2012), but there are some closed or partially closed cycles (e.g. the UK has operated a closed cycle where spent fuel is reprocessed using the PUREX process). Spent fuel consists predominantly of uranium oxide (96 %), with smaller amounts of other actinides such as plutonium and americium (1 %) and fission products such as cesium and strontium (3 %) (Turner, 2012), and the initial treatment for spent fuel is within fuel ponds for both thermal and radioactive cooling. The waste produced from the nuclear fuel cycle can be categorised into three tiers (which may slightly vary from country to country), as defined by the IAEA (International Atomic Energy Agency). The definitions for the UK waste tiers are as follows: LLW (Low Level Waste) – is defined as radioactive contamination from day to day operations and procedures such as clothing, floor sweepings, papers and plastic. Low level waste cannot 17
exceed 4 GBq tonne-­‐1 of alpha radioactivity, or 12 GBq tonne-­‐1 of beta/gamma radioactivity (ONR, 2013). The LLW produced from licensed sites in the UK are disposed at the Low Level Waste Repository (LLWR) near Drigg in Cumbria. ILW (Intermediate Level Waste) – generally consists of contaminated items such as filters, fuel cladding and reactor components. ILW ranges in radioactivity but comprises waste that exceeds the activities thresholds for LLW but ultimately does not generate enough heat for special methods to be required in it disposal/storage. HLW (High Level Waste) – generally consists of spent fuel from nuclear reactors. HLW is defined as waste that generates heat from its radioactivity, this heat significantly raises the temperature of the waste and its surroundings. In the UK HLW is managed by incorporation into a glass matrix (vitrification) and is currently stored at Sellafield. 18
Figure 1.1 shows a schematic of a typical nuclear fuel cycle outlining the most important steps. Figure 1.1. A simple schematic of the nuclear fuel cycle adapted from Turner (2012) 19
1.1.4 High Level Nuclear Waste Disposal As nuclear power moves into the 21st century and is being accepted more by the UK public (YouGov surveys show increasing support for nuclear power (YouGov, 2013) and all three main political parties support nuclear power) as a viable energy source that eliminates the emissions of greenhouse gasses. There still remains the problem of what to do with the HLW that is produced by a nuclear reactor, specifically, how to dispose of it in the long-­‐term. The World Nuclear Association lists previous methods of disposal that have been investigated (WNA, 2013): -­‐ Long-­‐term above ground storage; Disposal into outer space; Deep borehole disposal; Rock melting; Disposal at subduction zones; Ocean disposal; Sub seabed disposal; Disposal in ice sheets; Direct injection; Deep geological disposal. Whilst all these methods have been considered in the past by various governments and countries, this work is primarily based on the deep geological disposal concept which is the option that has been adopted by the U.K. government, and many other nations (e.g. Sweden, Belgium, France etc.). 20
1.1.5 Current deep geological disposal concepts from other EU countries In Europe the most advanced programs are those in Sweden and Finland. In these, sites are already approved and there has been significant testing from as far back as the 1970’s (SKB 1983, NDA 2012). Both Sweden and Finland are using the Swedish KBS-­‐3 concept for geo-­‐
disposal and permit applications are in progress (Morris et al. 2011). Since these countries operate an open fuel cycle, the Swedish and Finnish concept involves packaging and direct placement of the spent nuclear fuel rods into the disposal facility (the majority of the waste by mass or volume will be UO2). UO2 is not very soluble so a safety case for direct disposal can be made and further treatment/handling is not required (Karnbranslehantering 2010, POSIVA 2012). The spent fuel is placed in a cast iron insert which in turn is placed in a copper canister. This is surrounded by compacted bentonite (the buffer) and then the tunnels leading to the boreholes will be filled with more bentonite (the backfill) providing layered tiers of protection (this is the concept of the multi barrier system in section 1.3). Geological disposal is internationally accepted as the best way to dispose of higher activity radioactive waste. There are major differences from country to country due to the nature and volumes of individual nations’ waste inventories. This reflects, among other things, the scale of each nation’s programme and whether or not that country has operated a closed or open fuel cycle. Other differences will arise from host geologies (many countries don’t have sites selected yet), and the diposal packages used (impacted by both pre-­‐treatment of the waste and the cost of materials). Table 1.1 highlights several EU countries and their requirements/progress for deep geological disposal. 21
Country Sweden Stage Site Waste types Very Forsmark – advanced stable crystalline bedrock at a depth of 400 – 700 m Spent fuel Finland Very Olkiluoto – advanced stable crystalline bedrock at a depth of 420 m Spent fuel France Advanced Potential site Vitrified HLW Meuse-­‐
Haute-­‐Marne Long lived ILW Long lived ILW Waste package Spent fuel Compacted rods encased bentonite/backfill in copper of bentonite canisters As per Sweden KBS-­‐
3 As per Sweden KBS-­‐3 Vitrified HLW in unalloyed steel containers Bentonite and steel plugs (no buffer)/backfilled with excavated clay and sealed with bentonite ILW conditioned in bitumen or concrete and encapsulated in steel Belgium Advanced No site decided Vitrified HLW Vitrified HLW in carbon Long lived steel super ILW containers ILW conditioned in bitumen or concrete in steel drums Switzerland Very No site advanced decided Spent fuel Long lived ILW Buffer/Backfill Spent fuel in carbon steel canisters ILW conditioned in bitumen or concrete in steel ILW placed in shafts, sealed with concrete and plugged with clay Void between super container and tunnel filled with cementitious material ILW placed in waste emplacement caverns Compacted bentonite clay/backfilled with a mixture of bentonite, gravel, dry-­‐stone and a bentonite/sand mixture 22
waste drums ILW placed in concrete lined tunnels/backfill will be a gas permeable mortar Germany Advanced Gorleben salt dome – salt rock to a depth of >1000 m, repository would be at approx. 870 m Spent fuel Spent fuel encased in massive POLLUX containers (inner part stainless steel, outer part nodular cast iron) UK Concept No site still being decided realised Vitrified HLW Vitrified HLW in stainless Long lived steel ILW canisters, disposed in KBS3-­‐V No buffer/ crushed salt backfill will be used for boreholes, tunnels backfilled with crushed rock Still developing the expertise and considering all concepts Long lived ILW in cement in steel canisters Table 1.1, Example EU countries and their current progress on deep geological disposal, adapted from Morris et al. (2011) 23
1.2 Introduction to BELBaR BELBaR (Bentonite Erosion: effects on the Long-­‐term performance of the engineered Barrier and Radionuclide transport) is a European Union 7th framework project that uses resources from several European institutions to investigate the engineered barrier system (EBS) for a bentonite back-­‐filled repository. Specifically, the project is studying the bentonite backfill itself and investigating whether: 1. The generation of colloids would significantly degrade the EBS; 2. Colloid transportation of radionuclides would reduce the efficiency of the natural barrier (i.e. the colloids would travel in groundwater away from the repository site). This project’s main aim is to investigate and understand the interaction of radionuclides with bentonite and bentonite colloids and to develop an understanding of the kinetics involved, so that long-­‐term predictions of radionuclide behaviour may be made. For this, we first need to understand the nature of the interaction between radionuclides and the bentonite structure. 24
1.3 Multiple barrier system When designing nuclear waste repositories styles vary from country to country, but most designs have similarities. The waste is placed in some kind of container/canister, outside of which is the ‘backfill’. The backfill fills the space between the container/canister and the ‘host geology’ (the geological formation where the repository is placed). All repositories use the principle of a ‘multiple barrier system’, the main points are: •
The waste is treated to be resistant to dissolution; •
The container/canister acts as the first physical barrier; •
If the container/canister is breached then the waste would have to cross the backfill (second physical barrier); •
The final barrier is the host geology, which is considerable as the repository will be several hundred metres below ground. The backfill has safety functions, that is, the safety role of the backfill will change over geological time periods. The backfill is initially designed to protect the canister that holds the waste. After a significant amount of time when the canister has degraded, the safety function changes to that of containment to slow/stop the movement of any waste that may leach out of the container/canister. In deep geological disposal the two commonist types of backfill proposed are; 1. Cementitious or high pH backfills; 2. Clay/bentonite backfills. 25
1.4 Backfill material and purpose The backfill that will be used in this study is bentonite clay. The spent fuel will be placed in a container that will be buried and surrounded by the backfill, as shown in Figure 1.2. Loose Bentonite backfill Compacted bentonite Spent nuclear fuel canister Figure 1.2. Schematic of a proposed deposition hole adapted from SKB (1983) The function of the backfill is two-­‐fold; initially it will provide protection to the canister from potentially corrosive materials that could be in the groundwater. Secondly, the backfill will provide protection from within by stopping the outward leaching of radionuclides to the surrounding groundwater (SKB 1983, Karnland 2010). 26
1.5 Bentonite properties and characteristics Bentonite is a naturally occurring clay that can be found in deposits around the world. Clay minerals are hydrous aluminium silicates and as such are classified as phyllosilicates (Stankovic et al. 2011). Bentonite clay is a mixture of minerals with a generic name. All bentonite contains montmorillonite as its primary ingredient, and it is the montmorillonite component of bentonite that gives the characteristics desirable for the function of the backfill. The remaining components of bentonite vary due to geological formation and location (Stankovic et al. 2011). Because of this, bentonite can contain a number of different minor minerals, such as quartz, feldspars, gypsum, calcite, pyrite, and various iron oxides/hydroxides (Karnland et al. 2010). The montmorillonite structure comprises of sheets, with an octahedral layer made up from Al2O3 or MgO flanked on the top and bottom with SiO2 tetrahedral layers. It is between these sheets, that metal ions are retained (usually Na, Ca, K etc.) (Teoh et al. 2013). Figure 1.3. The typical structure of the montmorillonite component in bentonite clay (Teoh et al. 2013) 27
Bentonite has been chosen as a backfill in many deep geological disposal concepts, due to its favourable characteristics in and around a repository, such as its low permeability, its swelling properties, its cation exchange abilities and its general availability (SKB 2003, Karnland et al. 2010, Stankovic et al. 2011). As bentonite has been chosen as the backfill in many concepts for deep geological disposal, and makes up the entirety of the BELBaR project, the sorption/desorption studies within this thesis use bentonite rather than pure montmorillonite. This gives a more realistic representation of radionuclide interactions with a candidate backfill material, rather than with just one component of it. Given the proposed application of bentonite as a backfill, the presence of minor components such as quartz and pyrite needs to be considered (as they lack properties such as swelling and cation exchange) and, if pure montmorillonite is used, then the effects of other components will be excluded. Finally, since the work described in this thesis is part of the larger BELBaR programme, the use of bentonite is necessary to allow comparison with results provided by other consortium members. 28
1.6 Bentonite used in this study 1.6.1 Type of bentonite and chemical formula The bentonite clay used in this study is from the Wyoming deposits and as such is known as ‘Wyoming Bentonite’ or MX-­‐80. The bentonite studied here is a sodium bentonite, because its main cation is sodium. It is difficult to define a chemical formula for bentonite since it is a natural mixture, but Wyoming bentonite has montmorillonite as its main component, and can contain other minor components such as illite/mica, quartz, calcite, goethite etc (Karnland et al. 2010, Koch et al. 2008). As such, we can have an ideal formula for that main component which is (Stankovic et al. 2011): Na0.33 (Al1.67Mg0.33)Si4O10(OH)2 Whilst bentonite is stable on a geological timescale, recent studies have suggested that the generation of colloids may take place at the interface between the bentonite and the rock (Missana et al. 2003, Kunze et al. 2008). 29
1.6.2 Characterisation of MX80 bentonite Karnland (2010) and Carlson (2004) characterised a bentonite similar to that used in this study in detail. The material used in this thesis was characterised using powder X-­‐ray diffraction (XRD), X-­‐ray fluorescence (XRF) and BET adsorption measurements, in order to confirm that this material was similar to that described by Karnland (2010) and Carlson (2004). The XRD pattern (Figure 1.4) shows that the sample contains mineral phases consistent with bentonite samples previously described in the literature (Carlson, 2004; NDA, 2014). The principal component is montmorillonite, while quartz, albite, biotite, calcite, mica, gypsum and goethite, all typically found in bentonite clays, are also present. Figure 1.4. XRD pattern of the bentonite sample used in this study fitted to the minerals expected. Sample was run on a Bruker D8 Advance with a step size of 0.2° and a scan rate of 6°min-­‐1 from 4 – 70° 2θ with a copper x-­‐ray source 30
Table 1.2 shows the elemental composition of the bentonite clay determined through XRF. Element Weight % (this oxide study) Na2O 1.88 (± 0.1) MgO 2.45 (± 0.2) Al2O3 19.24 (± 0.5) SiO2 57.28 (± 0.4) K2O 0.50 (± 0.1) CaO 1.18 (± 0.007) TiO2 0.15 (± 0.03) Fe2O3 3.90 (± 0.08) Weight % (Karnland 2010) 2.25 2.61 21.2 67.4 0.55 1.46 0.17 4.14 Weight % (Carlson 2004) 2.08 2.44 19.9 61.3 0.57 1.36 0.16 3.84 Table 1.2. Elemental composition of bulk bentonite determined by XRF on an Axios Sequential X-­‐ray Fluorescence Spectrometer with a rhodium x-­‐ray source, using the PANalytical quantification software package. Data from Karnland (2010) and Carlson (2004) are shown to allow comparison Table 1.2 shows the elemental composition of a representative sample of the MX-­‐80 bentonite used in this study. The sample is comparable with previous studies (e.g. Karnland 2010). The clay shows Al2O3 (weight % 19.24) and SiO2 (weight % 57.28) as a total weight % of 76.5, this will be the main component of the clay (montmorillonite) with some possible impurities (quartz and biotite). The data also shows the major elements that will make up the other minerals that can be seen in the XRD pattern such as Fe2O3 (goethite) and CaO and SO3 (gypsum) with CaO and CO2 (calcite). The data shown from the XRF analysis is consistent with the mineralogical composition determined by XRD analysis and with other studies (such as Karnland, 2010 and Carlson 2004). Carlson 2010 reports that the surface area of bentonite clays lie in the range of 20 – 100 m2/g. The surface area of the clay used in this study was calculated using Brunauer, Emmett and Teller (BET) analysis using a Micrometrics Gemini V after the sample was purged with He2 for 18 hours before adsorption with N2. The surface area of the sample was found to be 29.86 ± 0.57 m2/g. 31
It was not practical to make direct measurements of the cation exchange capacity (CEC) of the bentonite used here. However, since its mineralogy and bulk composition are so similar to those previously described, it is likey to have a comparable CEC, in the range 74 – 110 meq/100g (Carlson (2004)). 1.7 Colloids 1.7.1 Colloids defined Colloids are defined as particles with a size range of 1 – 1000 nm and due to their small size Brownian motion can keep them in suspension over geological time scales (Wold, 2010). Whilst a range of colloids, both organic and inorganic, are known, our interest is solely in the formation of clay colloids, specifically bentonite. 1.7.2 Colloid stability The stability of clay colloids is of course very significant to the long-­‐term performance of a bentonite back-­‐filled repository, since they can act as vectors for radionuclides. There are several parameters that greatly affect the stability of the colloids: the ionic strength of the groundwater; the pH of the groundwater; the dominant cation in the clay (whether monovalent or divalent). The stability of the colloids will be maximised at low ionic strength and pH >8 (Laaksoharju 2005, Wold 2010). 32
1.7.3 The colloid ladder The importance of colloids in accelerating transportation of radionuclides is not clear-­‐cut. The type of clay and the geological evolution of the area, such as glacial melt water altering the groundwater chemistry, could affect the importance of colloid mediated transport (Wold, 2010). To simplify the matter, the so-­‐called Colloid Ladder (Mori et al. 2003) has been proposed and this is shown schematically in Figure 1.5. Figure 1.5. The colloid ladder, Mori et al. (2003). 33
For colloids to be significant all of the conditions have to be met. •
Are colloids present? -­‐ If colloids are not present in the system, then they will not be able to promote radionuclide mobility, but if they are, then their effect on radionuclide movement needs to be considered. •
Are the colloids mobile? – If the colloids are not mobile, then they will not be able to promote radionuclide mobility. •
Are the colloids stable? -­‐ If the colloids are unstable, then any associated radionuclides would be removed from the mobile phase (the solution) as the colloids are destabilised. •
Is there radionuclide uptake? – If radionuclides do not bind to the colloids at all, then the colloids cannot promote transport. •
Is uptake irreversible? – If the binding of the radionuclide to the colloid is reversible, then colloids should not be significant for radionuclide transport, because the rock itself should be able to compete for binding of the radionuclide, thus making the previous points irrelevant. If the binding were to be irreversible however, then transport of the radionuclide would continue until the colloid stopped moving. For colloids to be significant in the transport of radionuclides, then the answers to each of the questions listed above must be ‘yes’. 34
1.8 Previous radionuclide/bentonite binding studies 1.8.1 Bulk bentonite studies Simple uptake on bentonite and similar materials has been studied fairly widely. Comans (1987) studied Cd sorption onto illite, whilst Comans et al. (1991) and Koning and Comans (2004) studied Cs sorption on illite. Bradbury and Baeyens (1999) studied the sorption of Zn and Ni onto montmorillonite, whilst Bradbury and Baeyens (2006) studied the interactions of Am(III), Np(V) and Pa(V) with montmorillonite. Morton et al. (2001) investigated the sorption/de-­‐sorption of copper on montmorillonite and Kowal et al. (2004) studied the interaction of U(VI) on montmorillonite. Ochs et al. (2003) studied the uptake of Cs, Ra, Am and Pb onto bentonite. Guo et al. (2009) studied Eu(III) adsorption/desorption onto bentonite. In all these examples, simple Kd values were measured in sorption and desorption experiments, and only small differences in the Kd values were found. Several technical reports also discuss/study uptake experiments onto bentonite clay including SKB (1983) and Karnland et al. (2010). The interaction of bentonite (studied in its entirety, from bulk material, colloids and gel phase) with radionuclides has been reviewed thoroughly by Wold (2010). Dissociation rate constants were calculated using Kd values and measured association rate constants found within the literature. These will be discussed in more detail in section 4.4. 1.8.2 Modelling Surface complexation models can help to explain sorption/desorption behaviour and allow the prediction of the extent of surface complexation between the radionuclide and the montmorillonite/bentonite (e.g. Kowal et al. 2004). The modelling work by Bradbury and 35
Baeyens has been particularly successful (e.g. Bradbury and Baeyens 1999; Bradbury et al. 2005). They have developed the two-­‐site protolysis non-­‐electrostatic surface complexation and cation exchange model (2SPNE SC/CE). The modelling is expressed using surface complexation (≡S) with the bound metal (An), 3 sites are considered for the modelling calculations, a strong site and two weak sites (≡SsOH, ≡Sw1OH and ≡Sw2OH) which are defined by the sites binding capacity (2.0 × 10-­‐3 mol kg-­‐1, 4.0 × 10-­‐2 mol kg-­‐1 and 4.0 × 10-­‐2 mol kg-­‐1 respectively). This is a quasi-­‐mechanistic model, which uses 3 inner sphere surface complexes for the trivalent radionuclides, ≡SO-­‐An2+, ≡SO-­‐An(OH)+ and ≡SO-­‐An(OH)2. In addition, the radionuclides are also allowed to sorb to the surface by a process of cation exchange. This model has been applied successfully to the binding of Eu(III)/Cm(III) on montmorillonite and illite (Bradbury and Baeyens, 2005) and Am(III), Np(V) and Pa(V) on montmorillonite (Bradbury and Baeyens, 2006). For this model, good correlation can be seen between the modelled equilibrium constants and the radionuclide hydrolysis constants (Bradbury and Baeyens, 2006). For example, Figure 1.6 shows the correlation between the stability constant for the strong site and the corresponding first hydrolysis constant. 36
624
M
for strongly hydrolysing metals. F
urements carried out under simil
demonstrated that Eu(III) is a go
Am(III).
Acknowledgment. The authors would
A. Schaible and N. Verde for their con
work. The constructive comments from
also gratefully acknowledged. Partial fi
by NAGRA.
References
"
!
1. NAGRA: Project Opalinus Clay: S
disposal feasibility for spent fuel,
long-lived intermediate-level wast
gra Technical Report NTB 02-05,
Fig. 4. Correlation of surface complexation constants of species
(2002).
sorbing
on the cstrong
sites of montmorillonite
with thesorbing corres- on the 2. sNEA:
Figure 1.6. Correlation of surface omplexation constants of species trong Using Thermodynamic Sor
ponding hydrolysis constants. (The stability constants are corredioelement Distribution Coefficien
lated when y = x − 1, see Sect. 3.3) Transition and heavy metals
NEA, Paris (2001).
233
), 237
Np(V) ( ), 228 Th(IV)
( ) and
U(VI) ("):
( ), 241
sites of montmorillonite (sAm(III)
K) with (the corresponding hydrolysis constants (OHK). Taken from 3. NEA:
NEA Sorption Project Pha
241
239
Bradbury and Baeyens [8]. The actinides Am(III) ( ), Np(V)
diction of Radionuclide Sorption
( ) and 233 Pa(V) ( ) measured in this study fit well on the
Radioactive Waste Disposal Usi
Bradbury & BS Kaeyens (2006). OH K x , R = 0.99)
linear correlation
line (log
y = 7.99 + 0.89 log
Models. OECD/NEA, Paris (2005
and are consistent with previous sorption measurement/modelling
4. Baeyens, B., Bradbury, M. H.: A
results.
and Zn sorption on Na-montmorillo
tion measurements. J. Contam. Hy
5. Bradbury, M. H., Baeyens, B.: A
and Zn sorption on Na-montmo
high sorption values and indicates that metals sorb acJ. Contam. Hydrol. 27, 223 (1997)
1.8.3 Bentonite colloids and to
reversibility cording
their hydrolysis behaviour in a systematic way.
6. Bradbury, M. H., Baeyens, B.: Mo
Metals which hydrolyse very readily also sorb very strongly,
Ni on Ca-montmorillonite. Geoch
and conversely, weakly hydrolysing metals exhibit weak
(1999).
7. Bradbury, M. H., Baeyens, B.: Sor
As discussed above, tsorption.
he uptake and binding of radionuclides to bentonite and montmorillonites: Experimental
Thus the LFER demonstrates a chemically consistent picwith cation exchange and surface
ture
involving
the
sorption
model
concepts
and
the
associmontmorillonite bulk clay has been extensively studied, but the same cannot be said fmochim.
or Acta 66, 2325 (2002).
ated fixed model parameters (Table 2), the strong site surface
8. Bradbury, M. H., Baeyens, B.: Mo
stability constants and the hydrolysis behaviour of a metals
Co(II), Ni(II), Zn(II), Cd(II), Eu(
binding to bentonite-­‐derived colloids, and dissociation from bentonite (bulk oNp(V)
r and U(VI) on montmorillon
with valencies
between
II and VI. kItinetics is rather
remarkable
ships
and estimates of surface bind
that such a relatively simple sorption model is capable of
heavy metals and actinides. Geoch
colloidal) have been squantitatively
tudied very little (Wold, the
2010). describing
sorption of such a wide variety
(2005).
of metals. A notable feature of the LFER between strong
9. Chapman, N. B., Shorter, J.: Advan
lationships. Plenum Press, London
site surface complexation constants and the corresponding
hydrolysis constants
is that
it bapplies
over ∼ 40
of
Models based on instantaneous equilibrium have een proposed for orders
the purpose o10.
f Dugger, D. L., Stanton, J. H., Irby,
mings, W. W.: The exchange of twe
magnitude.
acidic silanol groups of silica gel. J
predicting radionuclide transport through bulk bentonite (e.g. Ochs et al. 2003). However, 11. James, R. O., Healy, T. W.: Adsor
ions at the oxide-water interface. J
(1972).
kinetic processes are 4.
expected to determine whether bentonite colloid facilitated 12.
transport Summary
and conclusions
Schindler, P. W., Furst, B., Dick, R.
of surface silanol groups. I. Surface
Sorption edge data for the actinides 241 Am(III), 239 Np(V)
Cu2+ , Cd2+ and Pb2+ . J. Colloid In
of radionuclides is significant (Mori have
et al. 2003). and 233 Pa(V)
been
measured on SWy-1 montmoril13. Davis, J. A., Leckie, J. O.: Surface
the oxide/water interface. II. Surfac
lonite. To the best of the author’s knowledge, this is the first
233
oxyhydroxide and adsorption of m
sorption data set measured and modelled for Pa(V) on
Sci.
Batch experiments wmontmorillonite.
ere performed wAll
ith b
entonite c
olloids t
hat h
ad b
een s
piked w
ith a 67, 90 (1978).
of the actinide sorption data could be
14. Dzombak, D. A., Morel, F. M. M.:
quantitatively described with the 2SPNE SC/CE model. The
elling: Hydrous Ferric Oxide. Wile
241
233
forequilibration Am(III), 239
Np(V) band
Pa(V),
Wersin, P., Schwyn, B.: Project O
radionuclide cocktail new
and lmeasurements
eft for different periods efore they were i15.
ntroduced proach for the development of ge
and earlier measurements for 241 Am(III), 237 Np(V), 228 Th(IV)
safety Assessment, Nagra Technica
233
and aU(VI),
very well
theradionuclides same linear correlation
to fracture filling material, nd the follow
dissociation of the was measured. Using Wettingen, Switzerland (2004).
between surface complexation constants of species sorbing
16. Bradbury, M. H., Baeyens, B.: Nea
on the strong sites of montmorillonite and the corresponding
MX-80 bentonite for
centrifugation it was shown that, even before contact with the fracture fill material, Ucompacted
(VI), a high level radioactive waste rep
hydrolysis constants (LFER) established previously. This
rock., PSI Bericht Nr. 03-05, Nag
work demonstrates clearly, through the LFER, that there is
Institut, Villigen PSI, Switzerland
a quasi thermodynamically consistent quantitative picture of
17. Baeyens, B., Bradbury, M. H.: A qu
why metals exhibit moderate or high or very high sorption.
tion of37Ni, Zn and Ca sorption o
For example, very high Rd values are fully to be expected
Physico-chemical characterization
Tc(VII) and Np(V) were not bentonite colloid associated. However, the tri and tetravalent radionuclides (Th(IV), Pu(IV) and Am(III)) were associated with bentonite colloids (Bouby et al. 2010; Huber et al. 2011). It was found that the Th(IV), Am(III) and Pu(IV) colloid bound concentrations, before any fracture fill material contact was within the range of 84 – 100%. Dissociation occurs once the fracture fill material has been added and colloids have been exposed to it for a short period (100 – 300 h). The authors were unsure of the reason for this apparent induction period. Dissociation continues until at least 1000 h, and there are indications the system may have reached equilibrium after 7500 h (313 days). In this fracture filling experiment (Bouby 2010) bentonite colloids were formed in Grimsel Groundwater (I = 0.964 mmol L-­‐1; pH 9.6): the concentration of colloids was 25.56 ppm. These fracture filling material studies show that there is some reversibility, as the radionuclides eventually dissociate from the colloids after fracture filling material is introduced to the system. Transport experiments through a natural fracture within a granodiorite core (Bouby, 2010) were performed by spiking a bentonite colloid suspension with radionuclides and recovering them after migration. The colloids were spiked with Eu(III), Tb(III) and Th(IV). The Th(IV) shows a strong association to the bentonite colloids, compared to both the Eu(III) and the Tb(III). It was also found that recovery decreases with increasing residence time. Missana (2008) performed migration experiments through a fracture in a granodiorite core using Eu(III) and Pu(IV). The bentonite colloids were produced in Grimsel groundwater (pH 9.5). It was found that over 80% of Eu(III) and Pu(IV) were colloid bound before being introduced to the fracture. The migration experiment involved introducing the colloid/radionuclide suspension to the fracture and then recovering the colloids at the end of the fracture. The Eu(III) shows reversibility in the binding of the radionuclide to the bentonite colloid. The evidence is from dissociation whilst the colloids are in transit in the 38
fracture (i.e. less radionuclide comes out of the fracture than was initially introduced). For the Pu(IV) system however, recovery could be predicted directly from the colloid recovery, this could be evidence of a slower dissociation rate for Pu(IV). Dissociation rates were 4.84x10-­‐4 s-­‐1 for the Eu(III), but were not calculable for Pu(IV) due to there being no indication of Pu(IV) dissociation. The high dissociation rate for the Eu and absence of evidence for dissociation for Pu(IV) could be due to the small residence contact time (2.55 hours) of the suspension within the fracture. Bouby et al. (2011) performed a series of batch experiments using bentonite colloids prepared in Grimsel groundwater (I = 10-­‐3 mol/L, pH 9.6) with a concentration of 20 ppm. These were spiked with Cs(I), Eu(III), Th(IV) and U(VI). It was found that Cs(I) and U(VI) showed nearly no interaction with bentonite colloids but the trivalent and tetravalent radionuclides Eu(III) and Th(IV) did associate with the colloids, as suggested by the previous work (Bouby et al. 2010). It was shown that radionuclides tend to bind to the smaller colloids, presumably due to their higher specific surface area. Competition for the binding of the radionuclides was introduced in the form of humic acid. The addition of humic acid causes an instantaneous partial dissociation of the radionuclide from the bentonite colloid. The binding of Eu(III) is completely reversible in the presence of humic acid over a 3 year period. For Th(IV), even after 3 years of contact time, a significant amount is still bound to the bentonite colloids. This could be evidence for irreversibility in the Th(IV), but no rate constants or reaction half times were recorded for this work. Figure 1.7 shows the results from the experiments performed by Bouby et al. (2011): the top graph shows the binding of Th(IV) to bentonite colloids and humic acid over time, and the bottom graph shows the binding of Eu(III) to bentonite colloids and humic acid over time. The equilibrium case in each graph is from simultaneous mixing of humic acid, bentonite colloids and radionuclides that are sampled at the same time as the regular data set. 39
Cs, Eu, Th and U interaction with clay colloids in presence of HA: A FFFF study
3875
Desorption time (months)
100
Colloid bound Fraction (%)
10
15
20
25
30
35
40
Th
HA-bound
60
40
Bentonite colloids-bound
20
100
Colloid bound Fraction (%)
5
80
0
0
0
150
300
450
600
750
900
1050
1200
HA-bound
80
60
Eu
40
Bentonite colloids-bound
20
0
0
150
300
450
600
750
900
1050
1200
Desorption time (days)
HA-bound
UF data
Bentonite-bound
CT 1 day
CT 2 weeks
CT 3 months
CT 5 months
CT 7 months
CT 10 months
CT 12 months
CT 3 years
equilibrium case, Set 5
Fig. 6. Evolution of Eu and Th partitioning between humic acid (HA) and bentonite colloids with time. Eu and Th have been aged with HA
Figure 1.7. Desorption graphs for Eu(III) nd Th(IV), hich sThe
how the represent
irreversible nature for
of Eu
the or bentonite
colloids
before
mixing during
a certain
time
(CT: a
contact
time) inwGGW.
values
the recoveries
and Th as
determined after analysis of Set 4 DT samples (HA acting as competing ligand). The values indicated are reproducible within less than 10%
uncertainty. Full square symbols represent the fraction of elements desorbed from bentonite and bound to the competing humic acid. The
Th(IV) binding to bentonite colloids. Bouby et al. (2011) open square symbols represent the fraction of elements remaining bound to the bentonite colloids. The crossed-symbols are additional ultra
filtration data. The stars represent the partitioning between the two colloidal fractions for the so-called equilibrium system (Set 5). The curved
arrows represent
subjective envelopes.
The reasons for this ‘irreversibility’ are unclear, but the authors speculate that surface The desorption experiments for Eu(III), therefore, show
of the Supplementary data file, see Appendix A). Furtherthat bentonite colloids cannot compete with humic acid and
more, the result of these experiments resembles closely that
precipitation of Th(IV) could be taking place and that this could be the reason for the long-­‐
the Eu sorption to bentonite colloids is fully reversible even
obtained for the “HA-RN” solution without bentonite colafter 3 years aging.
loids added (data shown on Fig. S4 of the Supplementary
binding tA).
o the colloids. In contrast, Th(IV) appears to bind stronger to the clay
data file, lived see Appendix
Thebentonite combination
of all these
colloids than Eu(III). The equilibrium distribution between
data clearly shows that for both Eu(III) and Th(IV) the
humic acid and clay colloids is obviously not obtained withchemical equilibrium is shifted towards the formation of
in 3 years. In this case we observe a clear difference in the
humate complexes. This is also the case for UO22+ (data
final Th(IV) speciation as a function of the addition seonly shown in the Supplementary data file, see Appendix
and S5), where almost no surface attachment
quence. Adding Th(IV) to the humic acid first and then clay
A, Figs. S2
colloids results in a clear predominance of Th(IV)–humate
to clay colloids could be observed in all experiments.
40
1.8.4 Dissociation Rates First order dissociation rate constants have been provided by Huber et al. (2011). For Am(III), the rate constant is in the range of 0.0037 – 0.009hr-­‐1 (1 -­‐ 2.5x10-­‐6 s-­‐1). For Pu(IV), the rate constant is in the range of 0.0014 – 0.0085hr-­‐1 (3.9x10-­‐7 – 2.4x10-­‐6 s-­‐1). Wold (2010) has estimated first order rate constants for a range of radionuclides (Table 1.3). Forward rate Radionuclide (kf h-­‐1 cm3 g-­‐1) Pu(IV) 435 Dissociation rate constant 4.35x10-­‐3 hr-­‐1 (1.2x10-­‐6 s-­‐1) Am(III) 435 2x10-­‐3 hr-­‐1 (5.6x10-­‐7 s-­‐1) Np(IV) 1x10-­‐2 4.6x10-­‐7 hr-­‐1 (1.2x10-­‐10 s-­‐1) Cm(III) 100 6x10-­‐3 hr-­‐1 (1.7x10-­‐6 s-­‐1) U(VI) 3 3x10-­‐3 hr-­‐1 (8.3x10-­‐7 s-­‐1) Tc(IV) 2.5x105 – 1x104 0.63 -­‐ 15 hr-­‐1 (1.75x10-­‐4 – 4.2x10-­‐3 s-­‐1) Table 1.3. Table of estimated dissociation rate constants; Wold (2010) It should be noted that these values were estimated from measured sorption rate constants (kf and Kd values) and calculated assuming Kd=kf/kb where kf and kb are the association and dissociation rate constants, respectively. As such, they are estimates only, not least because they assume that the reaction with bentonite is a single reaction. Alternative approaches to calculating dissociation rates, demonstrated by Wold (2010), highlight the importance of directly measuring dissociation rates experimentally. Chapters 3, 41
4 and 5 will show that, in practice, there could be several parallel dissociation pathways, each with a characteristic rate constant. Estimations like the ones in table 1.3 assume that there is instantaneous dissociation of the metal, and as such would not take into account the separate steps that may be involved in the dissociation of the metal from the bentonite clay. Direct measurement allows the researcher to observe any differences in dissociation that occur directly from the experiment and as such provides a much more accurate representation of the dissociation of the metals from the bentonite clay. It should also be noted that attempts to replicate the calculations by the author of this thesis with the data given in Wold (2010) have been attempted with no success, suggesting there is some missing factor in the calculations made in the original work by Wold (2010). 42
1.8.5 Colloid formation other than erosion Colloids can be generated in ways other than erosion, for example when bentonite pore water mixes with the local background groundwater. Kunze et al. (2008) discovered that changes in the ionic strength caused by the addition of ground water could be sufficient to form bentonite colloids by precipitation in the mixing zone of the waters. Furthermore, they suggested that Th(IV), Eu(III) and Cm(III) could become associated with the colloids within the mixing zone of these waters. This is relevant to radionuclide transport, as there is the possibility that radionuclides could become internally bound within the colloid during this process. Internally bound radionuclides are more likely to have very slow dissociation rates than simple surface bound radionuclides, as competition for the radionuclide would be physically blocked by the overlying shell of the colloid. Figure 1.8 illustrates the potential limitations on reversibility of radionuclide binding if it is bound within the colloid. Figure 1.8. Colloids with externally bound radionuclides (left) and with internally bound radionuclides (right). Due to the physical barrier presented by the colloid shell, reversibility is more likely for the externally bound radionuclides. 43
1.9 Radioisotopes used in this study 1.9.1 Europium Europium is a rare earth element in the lanthanide series. Although elemental europium is not found in nature, it is commonly found within other minerals such as monazite and bastnaesite. It’s naturally occurring isotopes are 151Eu and 153Eu. It is most commonly found in the +III oxidation state, although under strongly reducing conditions, Eu(II) can exist (Lide, 2005). Europium is commonly used as a neutron absorber within nuclear reactors (to help control the rate of fission); these are usually in the form of control rods containing europium oxide. Radioactive isotopes of europium (152Eu and 154Eu) are formed through neutron irradiation of the 151Eu and 153Eu and to a lesser extent as a fission product of 235U (El Abd et al. 2012, Klochkov et al. 2002). The europium that is used in this study is 152Eu(III), which has a half-­‐life of 13.5 years and has multiple gamma emissions ranging from 120 to 1410 keV (Klochkov et al. 2002). This isotope has been used for a number of reasons: 1. It is a strong gamma emitter, and because of this it can be analysed relatively easily using gamma ray spectroscopy. 2. It is used as an analogue of Am(III), due to the chemical similarities between the two ions. It can also be used as an analogue for other trivalent lanthanides and actinides. 44
1.9.2 Uranium Uranium is a metallic element in nature, uranium can be found as a mixture of 3 isotopes (238U, 235U and 234U), with 238U making up 99.275% of the total uranium content, while 235U and 234U make up 0.72% and 0.005%, respectively (Holden, 1977). 235U undergoes fission through the bombardment of thermal neutrons. Thus it is the fuel for virtually all current nuclear reactors. The other isotopes of uranium will also be present in the fuel. As a result, a lot of research has been performed on uranium and its isotopes (Grenthe et al. 2011). The isotope of uranium used here is 232U(VI). The 232U used in this study is in the form of uranyl (UO22+), which is highly soluble and thus highly mobile within groundwater systems (Grenthe et al. 2011). 232U has a half-­‐life of 68.9 years and is an alpha emitter. The use of this isotope is beneficial for several reasons: 1. The ease of analysis for alpha emissions using liquid scintillation techniques makes measuring the activity in a given sample relatively simple. 2. It is representative of any uranium radioactive waste forms and any results from it can be applied to all forms of uranium waste. 3. Its immediate decay product is 228Th, which can be chemically separated and can be used in further studies in this project. 4. Most importantly, its short half-­‐life allows the study of systems with very low uranium mass concentrations, which prevent artefacts such as precipitation from occurring. 45
1.9.3 Thorium Thorium is another metallic element in the actinide series. Like all actinides, it has several isotopes with various stabilities, but the most common form of thorium is 232Th, which has a very long half-­‐life of 1.405 x 1010 years and is naturally occurring (Mathias, 2011). 232Th is not a fissile isotope, but can however be converted into the fissile 233U via neutron bombardment. This is the basis of the ‘Thorium Cycle’. One product of this cycle is 232U, which decays to 228Th (Seaborg et al. 1947, Katzin 1952). 228
Th(IV) is the isotope that will be used in this study. It is the immediate decay product of 232
U, and it can be obtained by chemical separation of the 232U. It has a half-­‐life of 1.9 years and decays via alpha emission. As an alpha emitter, this isotope can be analysed using liquid scintillation counting (much the same way as 232U), although due to the presence of short-­‐
lived product isotopes, equilibration of the sample is needed. The decay chain of 228Th is: 228
Th è 224Ra (3.7d) è220Rn (57s) è216Po (0.15s) è212Pb (no alpha decay) (Smith et al., 1988) The 228Th decay products, until 212Pb, are also alpha emitters. As the daughters are chemically different, they may not necessarily be in equilibrium with the 228Th when the sample is prepared. Therefore, some time is needed to ensure that the daughters ‘grow into’ the sample, so that the rate of alpha emission produced is directly proportional to the activity of 228Th. This study uses a 28-­‐day equilibration time between sampling of the thorium and analysis of the samples. This isotope is beneficial for several reasons: 1. As a direct product of 232U purification it is readily available. 2. The 228Th(IV) can be used as an analogue of Pu(IV) and other tetravalent actinides (such as U(IV)). The advantage is that Th only has a single stable oxidation state in aqueous systems 46
(IV), and so redox control is not required. We can also be sure that the oxidation state is not changing during the experiment. 3. With the recent surge in thorium reactor research, thorium could be a component of future waste. 47
1.9.4 Americium Americium is a by product of nuclear fission created in nuclear reactor operation, formed by beta decay of 241Pu, which is itself formed through multiple neutron captures in 239Pu. The most important isotopes of this element are 241Am and 243Am due to their long half lives (433 and 7380 years respectively) (Runde and Wallace, 2011). 241Am is a common source of low energy gamma rays, but it is also an alpha emitter, because of this it is primarily used in smoke detectors due to its cheap cost and long half-­‐life (Runde and Wallace, 2011), this study will use 241Am. 241
Am decays by alpha emission, accompanied by gamma emission. This means it can be analysed via liquid scintillation counting or by gamma ray spectroscopy (in this study it was measured by gamma ray spectroscopy). Like all of the isotopes mentioned there are benefits to using 241Am. 1. 152Eu is frequently used as an analogue for 241Am, so the validity of using this analogy can be tested within the same systems. 2. It is a gamma emitter, and because of this it can be analysed relatively easily using gamma ray spectroscopy. 48
1.10 Aims and objectives The aims of this project are to explore the behaviour of bentonite clay backfill in a nuclear waste repository, and more specifically, to determine if radionuclides could become more mobile in the presence of bentonite colloids. This will be assessed by performing association/dissociation experiments to explore the binding strengths of key radionuclides to bentonite clay. The objectives of this project are: •
To explore the association/dissociation of Eu(III) to bulk bentonite clay, and to measure the dissociation kinetics of Eu(III) from bulk bentonite clay. •
To explore the association/dissociation of Eu(III), U(VI), Th(IV) and Am(III) to colloidal bentonite clay, and to measure the dissociation kinetics of these isotopes from colloidal bentonite clay The data collected from these experiments will be used to test whether radionuclide binding is reversible. This will be presented over 3 experimental chapters. 49
1.11 References Bouby. M, Filby. A, Geckeis. H, Geyer. F, Götz. R, Hauser. W, Huber. F, Keesmann. S, Kienzler. B, Kunze. P, Küntzel. M, Lützenkirchen. J, Noseck. U, Panak. P, Plaschke. M, Pudewills. A, Schäfer. T, Seher. H & Walther. C (2010) Colloid/nanoparticle formation and mobility in the context of deep geological nuclear waste disposal (Project KOLLORADO-­‐1; Final report), T. Schäfer & U. Noseck (Eds.), FZKA Wissenschaftliche Berichte, FZKA 751. Bouby. M, Geckeis. H, Lützenkirchen. J, Mihai. S & Schafer. T (2011) Interaction of bentonite colloids with Cs, Eu, Th and U in presence of humic acid: A flow field-­‐flow fractionation study, Geochimica et Cosmochimica Acta, 75, 3866–3880. Bradbury. M. H & Baeyens. B (1999) Modelling the sorption of Zn and Ni on Ca-­‐
montmorillonite, Geochimica et Cosmochimica Acta, 63, 325–336. Bradbury. M. H, Baeyens. B, Geckeis. H & Rabung. T (2005) Sorption of Eu(III)/Cm(III) on Ca-­‐
montmorillonite and Na-­‐illite. Part 2: Surface complexation modelling, Geochimica et Cosmochimica Acta, 69, 5403–5412. Bradbury. M. H & Baeyens. B (2006) Modelling sorption data for the actinides Am(III), Np(V) and Pa(V) on montmorillonite, Radiochim. Acta, 94, 619–625. Carlson, L. (2004) Bentonite mineralogy. Part 1: Methods of investigation a literature review. Part 2: Mineralogical research of selected bentonites. POSIVA working report 2004-­‐02. Comans. R. N. J (1987) Adsorption, desorption and isotopic exchange of cadmium on illite: Evidence for complete reversibility. Wat. Res, 21, 1573-­‐1576. Comans. R. N. J, Haller. M & Preter. P. De (1991) Sorption of caesium on illite: Non-­‐
equilibrium behaviour and reversibility. Geochimica et cosmochemica, 55, 433-­‐440. 50
El Abd. A, Mostafa. M & El-­‐Amir. M. A (2012) Production of 152, 154 Eu mixed sources for calibrations of gamma ray spectrometers. J. Radioanal. Nucl. Chem., 293, 255-­‐260. Grenthe. I, Drozdzynski. J, Fujino. T, Buck. E. C, Abrecht-­‐Schmitt. T. E & Wolf. S. F (2011) Chemistry of actinides and transactinides Vol 1-­‐6. Chapter 5 Uranium, section 5.1, p253-­‐255. Grenthe. I, Drozdzynski. J, Fujino. T, Buck. E. C, Abrecht-­‐Schmitt. T. E & Wolf. S. F (2011) Chemistry of the actinides and transactinides Vol 1-­‐6. Chapter 5 Uranium, section 5.3, p257. Guo. Z, Xu. J, Shi. K, Tang. Y, Wu. W & Tao. Z (2009) Eu(III) adsorption/desorption on Na-­‐
bentonite: Experimental and modelling studies. Colloids and Surfaces A: Physiochem. Eng. Aspects, 339, 126-­‐133. Holden. N. E (1977) Isotopic composition of the elements and their variation in nature – A preliminary report. BNL-­‐NCS-­‐50605, Pure Appl. Chem. 52, 2731. Huber. F, Kunze. P, Geckeis. H & Schäfer. T (2011) Sorption reversibility kinetics in the ternary system radionuclide–bentonite colloids/nanoparticles–granite fracture filling material. Appl. Geochem., 26, 2226–2237. Karnbranslehantering. S (2010) Design and production of the KBS-­‐3 repository. SKB technical report, TR-­‐10-­‐12 Karnland. O (2010) Chemical and mineralogical characterization of the bentonite buffer for the acceptance control procedure in a KBS-­‐3 repository, SKB Technical Report, TR-­‐10-­‐60. Katzin. L. I (1952) Production and separation of U-­‐233. Collected papers, Natl. Nucl. En. Ser. Div. (IV), 17B, report TID-­‐5223, USAEC. Koch. D (2008) European bentonites as an alternative to MX-­‐80, Science & Technology series., 334, 23-­‐30. 51
Klochkov. E. P, Risovangi. V. D, Vaneev. Yu. E & Dorofeev. A .N (2002) Radiation characteristics of europium-­‐containing control rods in a SM-­‐2 reactor after long-­‐term operation. Atomic energy, Vol 93, No.2, pp 656-­‐660. Koning. A & Comans. R. N. J (2004) Reversibility of radiocaesium sorption on illite. Geochemica et Cosmochimica, Acta, 68, 2815-­‐2823. Kowal – Fouchard. A, Drot. R, Simoni. E & Ehrhardt. J (2004) Use of spectroscopic techniques for uranium(VI)/montmorillonite interaction modelling. Environ. Sci. Technol., 38, 1399-­‐
1407. Kunze. P, Seher. H, Hauser. W, Panak. P. J, Geckeis. H, Fanghänel. Th & Schäfer. T (2008) The influence of colloid formation in a granite groundwater bentonite porewater mixing zone on radionuclide speciation. J. Contam. Hydrol. 102, 263–272. Laaksoharju. M & Wold. S (2005) The colloid investigations conducted at the Aspo Hard Rock Laboratory during 2000 – 2004 SKB Technical Report, TR-­‐05-­‐20. D. R. Lide (2005) CRC Handbook of Chemistry and Physics, Internet Version 2005, CRC Press, Boca Raton, FL, 2005 section 4 p12. Missana. T, Alonso. U, García-­‐Gutiérrez. M, & Mingarro. M (2008) Role of bentonite colloids on europium and plutonium migration in a granite fracture. Appl. Geochem. 23, 1484-­‐1497. Mori. A, Alexander. W. R, Geckeis. H, Hauser. W, Schafer. T, Eikenberg. J, Fierz. Th, Degueldre. C & Missana. T (2003) The colloid and radionuclide retardation experiment at the Grimsel Test site: Influence of bentonite colloids on radionuclide migration in a fractured rock. Colloids and Surfaces A: Physiochem. Eng. Aspects. 217, 33 – 47. Morris. K, Law. G. T. W, Bryan. N. D (2011) Geodisposal of higher activity wastes. Issues in Environmental science and Technology, 32, Ch – 6. 52
Morton. J, Semraw. J & Hayes. K (2001) An X-­‐Ray absorption spectroscopy study of the structure and reversibility of copper adsorbed to montmorillinite clay. Geochemica et Cosmochimica, Acta, 65, 2709-­‐2722. NDA (2010) Geological disposal; steps towards implementation. NDA Report no. NDA/RWMD/013. NDA (2012) Geological disposal; concept selection process. NDA technical note no. 16764837. NDA (2014) Geological disposal: A review of the development of bentonite barriers in the KBS-­‐3V disposal concept. NDA technical note no. 21665941. Ochs. M, Lothenbach. B, Shibata. M, Sato. H & Yui. M (2003) Sensitivity analysis of radionuclide migration in compacted bentonite: a mechanistic model approach. J. Contam. Hydrol. 61, 313– 328. ONR (2013) Management of radioactive materials and radioactive waste on nuclear licensed sites. NS-­‐TAST-­‐GD-­‐024 Revision 4 POSIVA (2012) Safety case for the disposal of spent nuclear fuel at Olkiluoto – design basis 2012. ISBN 978-­‐951-­‐652-­‐184-­‐1; ISSN 1239-­‐3096. Runde. W. H & Schulz. W. W (2011) Chemistry of the actinides and transactinides Vol 1-­‐6. Chapter 8 Americium, section 8.3, p1265 -­‐ 1267. Seaborg. G. T, Gofman. J. W & Stoughton. R. W (1947) Nuclear properties of U-­‐233: A new fissionable isotope of uranium. Phys. Rev. 71, 378. SKBF/KBS Swedish Nuclear Fuel Supply Co/Division KBS, Final (1983) Storage of Spent Nuclear Fuel – KBS-­‐3. 53
Smith. M. R, Lautensleger. A. W & Laul. J. C (1988) A new method for the determination of radium-­‐228, thorium-­‐228, and radium-­‐224 in groundwaters via thoron (radon-­‐220). J. Radioanal. Nucl. Chem. 123, 107-­‐119. Stankovic. N, Logar. M, Lukovic. J, Pantic. J, Miljevic. M, Babic. B & Radosavljevic-­‐Mihajlovic. R (2011) Characterization of bentonite clay from “Greda” deposit. Process. Appl. Ceram. 5, 97-­‐101. Teoh. W. T, Takuma. T & Kazunori. S (2013) Sorption of Pb(II), Cd(II), and Ni(II) Toxic Metal Ions by alignate-­‐bentonite. J. Environ. Prot. 4, 51 – 55. Tolenko. J. S (2012) Nuclear Energy, Selected Entries from the Encyclopedia of Sustainability Science and Technology, Springer, Ch-­‐8, p216-­‐221. Turner. D. R (2012) Nuclear Energy, Selected Entries from the Encyclopedia of Sustainability Science and Technology, Springer, Ch-­‐9, p223-­‐243. Wickleder. M. S, Fourest. B & Dorhout. P. K (2011) Chemistry of the actinides and transactinides Vol 1-­‐6. Chapter 3 Thorium, section 3.1, p52-­‐55. Wilson. P. D (1996) The nuclear fuel cycle: From ore to waste, Oxford Science, Publications, Oxford. Wold. S (2010) Sorption of prioritized elements on montmorillonite colloids and their potential to transport radionuclides, SKB Technical Report, TR-­‐10-­‐20. World Nuclear Association (2013) <http://world-­‐nuclear.org/.> YouGov (2013) http://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/vud89m3rd5/YG-­‐
Archive-­‐nuclear-­‐power-­‐results-­‐141013.pdf 54
Chapter 2 Materials and Methods The content of Chapter 2 will be used for the BELBaR report and will present data which can be used to support a safety case. It will explain in detail the techniques and methodology used throughout the project. The section will cover both the experimental approaches and also the data analysis for all experiments conducted. 55
2.1 Experimental Techniques 2.1.1 Batch experiment approach All of the original research in this thesis uses a batch experiment approach as this is the easiest way of performing long term, time dependent sorption studies. An outline of the generic approach for all experiments is described in this section. All batch experiments were designed to be conducted in 10 ml of solution at pH 8.8, at which bentonite colloids were stable. Bentonite colloid suspensions were prepared, and radionuclide spikes (Eu, U, Am and Th), also pH balanced to 8.8 were added as required. Once the spike is added to the bentonite colloids, this represents time zero in these batch experiments, from this main stock, at certain pre-­‐equilibrium times (day 1, day 7 etc.), 3 x 10 ml samples could be removed for the next part of the experiment. To each 10 ml sample, a pre-­‐determined amount of ion exchange resin was added, then the sample was rocked for 1 hour before being centrifuged (see centrifuge section 2.1.2 for details). Upon completion of the centrifuge step, samples were removed for analysis. Eu and Am were measured via gamma ray spectroscopy (section 2.3.3) and U and Th were measured via liquid scintillation counting (section 2.3.2) to determine the radionuclide content. 56
2.1.2 Centrifugation Calculations of centrifugation time and particle size were performed using an Excel spreadsheet. The formula used in the spreadsheet is based on Stokes’ law, and includes assumptions, on spherical particle shape and initial size distribution. When dealing with bentonite colloids it has to be appreciated that, due to the layered nature of the clay, colloids of this type will not be spherical (as seen in images 4.1 and 4.2 in chapter 4 of this thesis). Most calculations based on Stokes’ law assume the settling particle is spherical so that use of Stokes’ law to interpret data from centrifuge separation of the bentonite colloids will be an approximation. Stokes’ law expresses the relationship between the settling velocity of a particle within a liquid, and the variables affecting it (Soukup et al. 2008), V=g(sp-­‐s1)D2/1.8η where: V = Terminal particle velocity (cm s-­‐1); s1 = Liquid density (g cm-­‐3); g = Acceleration due to gravity (980 cm s-­‐2); sp = Particle density (g cm-­‐3); D = Spherical diameter of particle (cm); η= Viscosity of liquid (cm-­‐1 s-­‐1). Stokes’ law has been adapted over the years for different applications, one of them being predicting size separation in a centrifuge. The formula used here is an integrated modification of Stokes’ law, identical to the one from Weiner et al. (1995). In this approach, the required centrifuge time (tsec) is given by, tsec=(18 η ln (Rs/S)) / (ω2 D2 Δp) where: Rs = Distance from centre of centrifuge to settling distance (cm); S = Distance from centre of centrifuge to top of water column (cm); ω = Angular velocity (rad s-­‐1); Δp = density of solid – density of media (g cm-­‐3). 57
The Rs and S values are specific to particular tubes. Δp depends on the materials being used, as this is always bentonite clay, then Δp is a constant in this study. The reference value for density was taken from Gokalp et al. (2011), which gives Δp = 1.741 g cm-­‐3. 2.1.3 Colloid Generation The clay used for the colloid generation is Wyoming bentonite, a ‘sodium bentonite’, since sodium is the dominant cation. The experimental techniques have been adapted from Bouby et al. (2011). The clay was sieved to <63 μm. 10 g of bentonite clay was soaked in 1 L of deionised water for 10 days, with the beaker covered in aluminium foil to ensure no contamination and no loss of water through evaporation. To aid the generation of colloids through abrasion, the suspension was kept stirring through the entire 10 days. The resulting bentonite slurry was evenly distributed into centrifuge tubes (50 ml x 20), even slurry distribution was achieved by keeping the slurry stirring and extracting the required 50 ml using a syringe. Each tube was centrifuged (4000 rpm, RCF = 2683 g, 11 minutes Rs = 10 cm, S = 5 cm), to ensure no bentonite colloids remaining in solution were larger than 500 nm. Each tube had the supernatant fluid decanted, and was then refilled with deionised water (50 ml) and sonicated in a sonic bath for 10 minutes to re-­‐suspend the clay. This process of centrifugation and sonication was repeated a further 3 times, with the supernatant being discarded. The supernatant fluid remaining after the 4th centrifugation step constituted the colloidal stock, which was stored in the dark. 58
2.1.4 Ultrafiltration Ultrafiltration can be used to filter nano-­‐scale particles from a solution. Depending on the particle shape and the density difference between the solid and the liquid phases, sometimes ultrafiltration can be the only way to separate a very small particle from solution. Ultrafiltration is widely used to separate proteins from solution, as such the pore size of the filters is in Daltons (atomic mass units) rather than nm, where the size separation is based on a size and weight correlation for proteins. As such it is difficult to apply a 2 dimensional size (i.e. nm) to the non molecular weight cut-­‐off limit (Daltons) of the filters (Pitois et al. 2008), size approximations can however be made. The filters in this study were polyethersulphone (PES) membrane filters with nominal size ranges of 3 KDa, 10 KDa and 100 KDa, approximate pore sizes of 1.5 nm, 3 nm and 10 nm respectively (assuming spherical particles). The ultrafiltration stirring cell has colloid suspension added (10 ml), the cell is sealed and pressurised with argon (1.5 bar), and the resulting filtrate is recovered for analysis. Ultrafiltration was used as part of the size characterization of the colloids in Chapter 3. Argon inlet Stirrer bar Polyethersulfone filter membrane Filtrate outlet Figure 2.1. Typical ultrafiltration cell 59
2.1.5 232U(VI) and 228Th(IV) purification 232
U is useful in radiometric studies due to its high specific activity, this makes it easily detectable with scintillation counting even at low mass concentrations. 228Th is a decay product of 232U and will therefore be generally present in a 232U solution. Both isotopes decay via alpha emission with relatively similar energies (5.52 MeV and 5.31 MeV respectively), so that differentiation between the two isotopes by liquid scintillation counting is not possible. In order to measure these isotopes, purification is essential and is achieved using a 2 ml UTEVA resin column (Eichrom). The column is supplied pre-­‐
conditioned in HCl (0.1 M), and has to be re-­‐conditioned with HCl washes (5 M) before use. Before separation, the 232U(VI) was in equilibrium with 228Th(IV). The untreated 232U(VI) (2 ml, 8 kBq ml-­‐1) sample was adjusted to an HCl concentration of 5 M (with a total final volume of 20 ml). This solution was passed through the column. At this HCl concentration, the 228
Th(IV) eluted through the column and the 232U(VI) was retained. The 232U(VI) was then eluted using HCl (0.1 M) to give separate 232U(VI) and 228Th(IV) fractions (the latter rapidly reaching equilibrium with 224Ra, 220Rn, 216Po and 212Pb), for use in the batch experiments in Chapters 4 and 5 respectively. The 232U will only be usable for approximately 3 months before 228Th in-­‐growth starts to affect alpha liquid scintillation counting analysis. Figure 2.2 shows the spectra of 232U and 228Th when analysed using liquid scintillation counting, and illustrates why separation of the isotopes is needed for effective analysis. 60
Figure 2.2. Alpha liquid scintillation spectra for U(VI) and Th(IV) fractions 61
2.2 Solution Speciation Modelling 2.2.1 PHREEQC PHREEQC is a computer program developed by the United States Geological Survey (Parkhurst et al. 1999) which can be used for the estimation of species distribution in solution. PHREEQC modelling was performed for all of the systems present in the experiments and the predicted speciation is shown in detail in Chapters 3, 4 and 5. The calculation in this thesis used PHREEQC Interactive (version 2.17.5.4799, released September 2010) and the Specific Interaction Theory Database ‘ThermoChimie v.7.b’, developed by Amphos 21, BRGM and HydrAsa for ANDRA (Duro et al 2006, Grive et al, 2010). Speciation modelling is useful in predicting the speciation of metal ions in solution within the systems (i.e. that fraction not bound to the bentonite). It also gives allows calculation ofsaturation indices and thus shows whether the solution will be super saturated and which phase(s) are most likely to precipitate. This allows attempted rationalisation of experimental results of the type described in this thesis. All speciation modelling used in this thesis assumes equilibrium with both atmospheric CO2 and montmorillonite clay. The thermodynamic data set used for calculations can be seen in the appendix at the end of this thesis. 62
2.3 Experimental analysis techniques 2.3.1 ICP-­‐AES Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-­‐AES) is a technique that can be used for the elemental analysis of analytes typically in the parts per-­‐million range. In this study, ICP-­‐AES has been used extensively for the determination of bentonite colloid concentrations. The samples were acidified to 2 % HNO3 and analysed for their Al and Mg concentrations. For bentonite, the Al and Mg should be in a ratio of 5.02:1 and the correct ratio from ICP-­‐AES results indicates bentonite colloids in suspension. The ppm results can be converted to g L-­‐1 then dividing through the molecular weight of the element analysed and the ratio predicted will give an elemental value in mol L-­‐1, this can be multiplied with the molecular weight of bentonite and converted back to ppm giving a concentration for bentonite colloids (analysis method adapted from Laaksoharju et al. (2005)). For ICP-­‐AES, a sample is subjected to high temperatures (~104 K) which decomposes molecules into individual atoms and causes thermal excitation. As the atoms relax, they emit light in the UV or visible regions. These emissions occur at a specific wavelength, diagnostic of a particular element, and the intensity of the emission is proportional to the concentration of the atoms. The light is detected using photomultipliers (Boss and Fredeen, 1997). All ICP-­‐AES analysis performed as part of this thesis was carried out on a Perkin-­‐Elmer Optima 5300 dual view ICP-­‐AES in the School of Earth, Atmospheric and environmental Sciences. A simple schematic of the ICP-­‐AES instrument is shown in Figure 2.3. Sample sprayed as an aerosol Plasma Monochromator Detector Figure 2.3. A simple schematic of how ICP-­‐AES is performed 63
2.3.2 Liquid scintillation counting LSC (liquid scintillation counting) is an analytical method capable of detecting very small amounts of alpha and beta emitting radioisotopes (< 0.1 Bq) in a sample (Passo and Cook, 1994; Horrocks, 1974). LSC was used for in this study both to confirm the successful separation of 232U and 228Th and also to measure 232U (Chapter 4) and 228Th (Chapter 5) experiments. In LSC analysis, a sample is dissolved or dispersed in a scintillation cocktail (Scintisafe 3). A decay event causes a transfer of energy to the solvent in the scintillation cocktail, and relaxation of these excited species results in a photon of light being emitted, which can be detected by a photomultiplier tube. In practice, the sample is raised into shielded counting chamber, which has a pair of photomultiplier tubes that detect the light emissions from the sample. The emission is only registered if it is detected by each of the photomultiplier tubes at the same time, since this ‘coincidence circuit counting’ helps to eliminate erroneous detections and reduces background interference from natural background radiation (Horrocks, 1974). A schematic of an LSC is shown in figure 2.4. 64
PSA Coincidence circuit Amplifier Amplifier Sample Photomultiplier tubes Figure 2.4, Schematic of the components involved with LSC analysis. PSA is described below LSC can be used for both alpha and beta emitting radionuclides and can determine between the two types of decay through PSA (pulse shape analysis). An alpha-­‐induced ionization produces emissions that last longer than beta-­‐induced ionizations (Passo and Cook, 1994). The PSA function can be set to discriminate between the two types, though in this study both 232U and 228Th are alpha-­‐emitting isotopes so only alpha detection is used. In some scenarios with LSC, quenching of the sample can occur. This is where the emission from an excited state is reabsorbed and can occur via chemical quenching (where the emission is chemically absorbed) or by colour quenching (where the emission is absorbed before leaving the vial by a coloured sample) (Baillie, 1963). In these studies, the samples are colourless and contain no chemically quenching components. All LSC analysis performed in this thesis used 0.1 ml of a known sample in 10 ml of scintillation cocktail (Scintisafe 3), and samples were counted on a 1220 QUANTULUS ultra-­‐low level scintillation spectrometer. 65
2.3.3 Gamma Ray Spectrometry In radioactive decay, when a nucleus is formed in an excited state, relaxation of this state occurs by emission of a gamma ray. The energies of the gamma emissions are all well defined and characteristic of the emitting isotope. The intensity of the emission is proportional to the radionuclide activity (Gilmore, 2008). For analysis to be performed on a gamma ray detector the sample is simply placed on the detector end cap. The detector is a single crystal of high purity germanium (HPGe), which has to be cooled to liquid nitrogen temperature (~ 77 K). As no separation of the radionuclides needs to take place, and the process is itself non-­‐destructive, gamma ray spectroscopy can be a useful tool for environmental radiometric analysis. Gamma-­‐ray spectroscopy was used in this thesis for the analysis of 152Eu (using the characteristic peak of 121.78 keV) and for 241Am (using the characteristic peak of 59.5 KeV). For both radionuclides, 1.5 ml of the sample was placed on the detector and counted for 1200 s (Eu) and 1800 s (Am). The experiments in Chapter 3 and 5 used gamma-­‐ray spectroscopy, carried out on a Canberra 2020 coaxial HPGe gamma spectrometer with an Ortec 919E multi-­‐channel analyser. There are several considerations in gamma-­‐ray spectrometry, especially the positioning and volume of the samples themselves, which has to be reproducible and closely matched to the calibration used. 66
2.3.4 Scanning Electron Microscopy (SEM) SEM is a technique that can be used to view particles that are on the μm to nm scale, particularly for particles that cannot be viewed with any great detail using a light microscope. SEM was used here to image bentonite colloids. The sample is placed in a vacuum chamber where it is subjected to an intense beam of electrons from an electron gun, usually with a tungsten filament or tungsten tip. Rather than a stationary beam of electrons, the SEM is scanned across the sample area in the x and y directions in a procedure known as raster scanning. This causes the beam to cover a rectangular area over the sample and allows a much more broad view of the sample than other methods such as transmission electron microscopy. When a bulk sample is raster scanned, secondary electrons are released by the sample and directed into a detector. The image that is viewed in this mode is thus made of secondary electrons from each point on the sample. (Egerton, 2005). Figure 2.5 shows a schematic of a typical SEM. Figure 2.5. Schematic of an SEM (Egerton, 2005) 67
All electron microscopy performed in this thesis was carried out on bentonite colloid samples which were air-­‐dried on a Leit Tab (12 mm, purchased from Agar scientific) and viewed on a Philips XL30 FEG ESEM under low vacuum, results of this can be seen in Chapter 4. 68
2.4 References Baillie. L .A (1963) Determination of liquid scintillation counting efficiency by pulse height shift, Advances in tracer methodology, p86-­‐92. Boss. C. B & Fredeen. K. J (1997) Concepts, instrumentation, and techniques in inductively coupled plasma optical emission spectroscopy, second edition, PERKIN ELMER, Ch 1-­‐4. Bouby. M, Geckeis. H, Lutzenkirchen. J, Mihai. S & Schafer. T (2011) Interaction of bentonite colloids with Cs, Eu, Th and U in presence of humic acid: A flow field-­‐flow fractionation study, Geochimica et Cosmochimica Acta, 75, 3866–3880. Duro, L. Cera, E. Grive, M. Domenech, C. Gaona, X. & Bruno, J (2006) Development of the thermochimie thermodynamic database. Janvier. Andra report C. RP. 0ENQ. 06. 0001. P373. Egerton. R. F (2005) Physical principles of electron microscopy, Springer. Gilmore. G (2008) Practical gamma-­‐ray spectroscopy, Wiley-­‐Blackwell. Grive, M. Riba, O. Montoya, V. & Duro, L. (2010) Update of the thermochimie database: Reporting of new data selection 2010. June 2010. Gokalp. Z, Basaran. M & Uzun. O (2011) Compaction and swelling characteristics of sand-­‐
bentonite and pumice-­‐bentonite mixtures. Clay Minerals, 46, 449-­‐459. Horrocks. D. L (1974) Applications of liquid scintillation counting, Academic Press, New York. Laaksoharju. M & Wold. S (2005) The colloid investigations conducted at the Aspo hard rock laboratory during 2000 – 2004 SKB Technical Report, TR-­‐05-­‐20. 69
Parkhurst. D. L & 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: Denver, CO, USA, p 312 Passo. C. J & Cook. G. T (1994) Handbook of environmental liquid scintillation spectrometry, Packard. Pitois. A, Ivanov. P. I, Abrahamsen. L. G, Bryan. N. D, Taylor. R. J & Sims. H. E (2008) Magnesium hydroxide bulk and colloid-­‐associated 152Eu in an alkaline environment: Colloid characterization and sorption properties in the presence and absence of carbonate. J. Environ. Monit. 10, 315-­‐24. Soukup. D. A, Buck. B. J & Harris. W (2008) Methods of soil analysis chapter 2 preparing soils for mineralogical analyses, Pg 23. Weiner. B. B, Tschamuter. W & Fairhurst. D (1995) Accurate particle sizing of high density materials to 10nm using an x-­‐Ray disc centrifuge with a moving source/detector. Brookhaven Instruments Corporation. 70
Chapter 3 Reversibility in Radionuclide/Bentonite Bulk and Colloidal Ternary Systems The content of Chapter 3 will be used for the BELBaR report and will present data which can be used to support a safety case. It has been accepted for publication as part of the IDGTP-­‐
Geodisposal 2014 conference proceedings and is waiting to be published. The contributions by the author of this thesis were: all of the filtration and dissociation experimental work; the interpretation of results and calculation of dissociation rate constants; and the drafting and final writing up of the paper. All ICP-­‐AES was performed on a Perkin-­‐Elmer Optima 5300 dual view ICP-­‐AES in the School of Earth, Atmospheric and Environmental Sciences by Mr Paul Lythgoe. PHREEQC modelling was conducted by Dr Kurt Smith and interpreted by the author of the thesis. This paper studies the dissociation of Eu(III) from both bulk bentonite clay and colloidal bentonite clay. The dissociation o Eu from bulk bentonite was studied first, as the pilot study to refine the techniques that were used in the rest of the thesis. In subsequent experiments, the pH range and solid:solution ratio were changed from those used in the preliminary experiments. 71
Reversibility in Radionuclide/Bentonite Bulk and Colloidal Ternary Systems Nick Sherriff1, Ragiab Issa1, Katherine Morris2, Francis Livens1,2, Sarah Heath1 and Nick Bryan3* 1
Centre for Radiochemistry Research, School of Chemistry, University of Manchester. 2
Research Centre for Radwaste and Decommissioning, School of Earth, Atmospheric and Environmental Sciences, University of Manchester. 3
National Nuclear Laboratory, 5th Floor, Chadwick House, Birchwood, Warrington, WA3 6AE, UK *[email protected] 72
Abstract Ternary systems of 152Eu(III), bulk bentonite and ethylenediaminetetraacetic acid (EDTA) ([Eu] = 7.9 x 10-­‐10 M; pH = 6.0 – 7.0) have been studied. Without EDTA, there was slow uptake in a two stage process, with initial rapid sorption of Eu(III) (96%), followed by slower uptake of a much smaller fraction (3.0 % over a period of 1 month). The reversibility of Eu(III) binding was tested by allowing Eu(III) to sorb to bentonite for 1 – 322 days. EDTA was added to the pre-­‐equilibrated Eu bentonite systems at 0.01 M, a concentration that was sufficient to suppress sorption in a system where EDTA was present prior to the contact of Eu(III) with bentonite. A fraction of the Eu was released instantaneously (30 -­‐ 50 %), but a significant amount remained bound. With time, the amount of Eu(III) retained by the bentonite reduced, with a slow fraction dissociation rate constant of approximately 4.3 x 10-­‐8 s-­‐1 (values in the range 2.2 x 10-­‐8 – 1.0 x 10-­‐7 s-­‐1) for pre-­‐equilibration times ≥ 7 days. Eventually, the amount of Eu(III) remaining bound to the bentonite was within error of that when EDTA was present prior to contact (4.5 % ± 0.6), although in systems with pre-­‐equilibration times > 100 days, full release took up to 500 days. Europium interactions with colloidal bentonite were also studied, and the dissociation rate constant measured by a resin competition method. For the colloids, more Eu was found in the slowly dissociating fraction (60 – 70 %), but the first order dissociation rate constant was faster, with an average rate constant of 8.8 x 10-­‐7 s-­‐
1
and a range of 7.7 x 10-­‐7 – 9.5 x 10-­‐7 s-­‐1. For both bulk and colloidal bentonite, although slow dissociation was observed for Eu(III), there was no convincing evidence for ‘irreversible’ binding. 73
Introduction Studies suggest that for colloidal transport to be significant, colloids must bind radionuclides irreversibly, since any reversibly bound radionuclides would be quickly removed by the available rock surface binding sites, whereas ‘irreversibly’ bound radionuclides will be transported with the colloids (Mori et al. 2003). In fact, the situation is more complex than this, because there is no simple division between reversibly and irreversibly bound metal ions. Calculations have shown that the slow dissociation of radionuclides from colloids (i.e. short of ‘irreversibility’) can facilitate transport. The transport residence time only needs to be small compared to the half time for dissociation, with migration increasing with decreasing rate constant (Bryan et al. 2007). The interactions of radionuclides with bulk and colloidal clays are important, because they have been suggested as a potential backfill material for a radioactive waste repository (Mori et al. 2003), and so the sorption of radionuclides by bulk bentonite, its constituents and other bulk clays have been studied extensively (Wold 2010). However, only the reversibility of Cs(I) binding to bulk clays has been studied in detail (e.g., Comans, 1987; Comans et al., 1991; de Koning and Comans, 2004), although Ivanov et al. (2012) have reported no evidence for slow dissociation of uranyl from bulk bentonite. Radionuclide sorption to bentonite colloids has also received attention, however, the dissociation kinetics have been studied barely at all (Wold 2010). In conditions relevant to the vicinity of a clay buffered radioactive waste repository, Missana et al. (2008) found > 75 % colloid bound for Eu(III) and Pu(IV), whilst Schäfer et al. (2004) found that approximately 80 % of both Th(IV) and Eu(III) were colloid associated. Huber et al. (2011) have reported competition experiments between bentonite colloids and fracture filling material. Slow dissociation from the colloids was observed, and there was some evidence that the system took 7,500 hours (313 days) to reach equilibrium: for Am(III), the rate constants were in the 74
range 1 -­‐ 2.5 x10-­‐6 s-­‐1, whilst for Pu(IV), the range was 3.9 x10-­‐7 -­‐ 2.4 x10-­‐6 s-­‐1. Wold (2010) estimated representative first order dissociation rate constants from sorption rate constants and Kd values for: Pu(IV) 1.2x10-­‐6 s-­‐1; Am(III) 5.6x10-­‐7 s-­‐1; Np(IV) 1.2x10-­‐10 s-­‐1; Cm(III) 1.7x10-­‐6 s-­‐
1
; U(VI) 8.3x10-­‐7 s-­‐1; Tc(IV) 1.75x10-­‐4 -­‐ 4.2x10-­‐3 s-­‐1. Geckeis et al. (2004) found that Am(III) and Pu(IV) transport through fractures could only be explained by slow dissociation from colloids, although the interactions were eventually reversible on a time scale of months. Bouby et al. (2011) found that Cs(I) and U(VI) did not bind to bentonite colloids, but that the tri-­‐ and tetravalent f-­‐block ions (Eu(III); Th(IV)) were strongly associated. They also studied the competition between bentonite colloids and humic acid. In the case of Eu(III), they found that although dissociation from the colloids was slow, equilibrium was eventually attained, and most of the Eu(III) dissociated from the bentonite and bound to the humic acid. However, in the case of Th(IV), even after 3 years, the system had not reached equilibrium. In this study, we have determined the dissociation rate constants for Eu(III) from both bulk and colloidal bentonite. The study of the uptake of radionuclides by bentonite is relatively straightforward, and requires only that the radionuclide is added to the clay. However, to study the reversibility of the interaction, it is necessary to allow the radionuclide to interact with the bentonite for a known period of time before using a competing ligand or cation binding resin as a strong sink to ’pull’ the radionuclide from the bentonite, so that the dissociation rate may be measured (e.g. Monsallier et al. 2003). Previous tri-­‐valent f-­‐element dissociation experiments have not considered dissociation from bulk bentonite, and those with colloidal bentonite have used ligands such as humic acid or complex natural solid competitors, such as fracture fill material, that could interact with the colloids. Other literature values are estimates based on sorption rates and distribution coefficients. The aim of this work is to measure directly the dissociation rate constants for bulk and colloidal bentonite using sinks that do not interact with bentonite: such data are useful in the 75
prediction of radionuclide migration in environments where bentonite exists, such as deep geological disposal. Materials and Methods The clay used in this work is a sodium Wyoming bentonite. All water was deionised (18 MΩ) and all reagents were analytical grade. All experiments were repeated in triplicate. Bulk Eu dissociation experiments Bentonite clay (0.5 g) was added to tubes with deionised water (8 ml), NaClO4 (1 ml, 0.1 M) and 152Eu (1 ml, 1 kBq ) to give a total Eu concentration of 7.9 × 10-­‐10 M, the system was then adjusted to pH 7 (± 0.1). The clay was suspended (maximising contact with the Eu(III)) and the tubes were left on their sides, maximising contact between dissolved species and the clay. After a pre-­‐equilibration period between 1 and 332 days, EDTA sodium salt was added to the system. Before the addition of EDTA, the tubes were centrifuged on a BOECO C-­‐28A centrifuge (15 mins, 4000 rpm) to remove particles larger than 2 μm. Once this was complete, the top 4 ml of suspension in the tube were removed and distributed into 2 ml containers. These were then centrifuged (35 minutes, 14000 rpm): this removed all particles larger than 0.25 μm. Following centrifugation, 1 ml of the solution was removed and replaced by EDTA, which when introduced to the 10 ml system gave an EDTA concentration of 0.01 M. Once the EDTA was added, the tubes were rocked for 1 hour to re-­‐suspend the clay. Following the re-­‐suspension, and after a period of 1 day, the tubes were again centrifuged as before and, 1.5 ml were removed for analysis by gamma ray spectrometry (Canberra 2020 coaxial HPGe gamma spectrometer with an Ortec 919E multi-­‐channel 76
analyser). The aliquots were then returned to the sample tube, the clay re-­‐suspended and the sample stored for later analysis. In a separate experiment, the uptake of Eu on to bentonite was measured in the absence of EDTA, but under the same conditions. Also, an experiment was performed where EDTA was added to the Eu tracer solution, prior to any contact with the bentonite clay (this was to ensure that the EDTA could effectively compete with the clay for the Eu). Colloid studies A stock solution of colloids (147 ppm) was prepared and colloid concentrations measured according to the method described by Bouby et al. (2011). Speciation calulations Solution speciation was calculated with PHREEQCi (Parkhurst et al. 1999), using the ThermoChimie v.7.b database (Duro et al 2006, Grive et al, 2010). Size distributions Syringe filters and ultrafiltration membranes were used sequentially to filter the bentonite colloid stock (450, 200 and 100 nm PES filters). The filters and membranes were adjusted to the pH of the colloid suspension (8.8 ± 0.2) and pre-­‐equilibrated for 68 hrs prior to use. Ultrafiltration was performed under pressure (argon, 1.5 bar) through membranes (Millipore, polyethersulfone: 300, 10 and 3 kDa). In a separate experiment, 152Eu (3.5 ml, 35 kBq) was added to a sample of the bentonite colloid stock (3.5 ml) and left for 68 hrs to 77
equilibrate before filtration as above. The syringe filters and ultrafiltration membranes were pre-­‐treated in 10-­‐4 M stable Eu(NO3)3 solution to prevent sorption of the radiotracer, according to the method of Pitois et al. (2008). Dowex 50WX4-­‐200 (H-­‐form) resin was converted to its sodium form. 20 g of Dowex resin were added to a sintered column, and washed with: deionised water (150 ml); HCl (2 M, 500 ml); deionised water (500 ml); NaCl (3 M, 500 ml); NaOH (0.1 M, 500 ml); and finally deionised water (500 ml). Experiments showed that the colloid did not interact with the pH pre-­‐conditioned resin (data not shown). 152
Eu solution (30 ml, 30 kBq ) was added to a container, and its pH adjusted (8.8 ± 0.1). Following the pH adjustment, colloid suspension was added to the 152Eu. This gave a total volume of 152Eu/colloid suspension of 300 ml. At different pre-­‐equilibration times, aliquots of the 152Eu/colloid suspension was removed (10 ml, x3) from the main mixture, and conditioned Dowex resin was added (1.4 g). The resulting mixture was allowed to shake gently for 1 hr. After shaking, the sample was centrifuged (15 mins, 4000 rpm) and a 1.5 ml aliquot of the supernatant was removed for analysis for Eu content. After measurement, the aliquot was returned to the sample tube, the resin re-­‐suspended, and the sample stored for subsequent analyses using the same method. 78
Results and Discussion Bulk bentonite dissociation Powder X-­‐ray diffraction was used to show that the EDTA had no effect on the bentonite (data not shown). Figure 3.1 shows the dissociation of Eu(III) from bentonite with EDTA contact time as a function of pre-­‐equilibration time of the Eu with the bentonite prior to addition of EDTA. Figure 3.1. Plot of percentage of Eu bound to bentonite vs time (days), as a function of pre-­‐
equilibration time (pH = 7 ± 0.1; I = 0.1 M NaClO4). The full black horizontal line represents the equilibrium position, and the dashed lines represent the experimental uncertainty for the equilibrium position. The data are plotted as the percentage of Eu remaining bound to the bentonite. Eu(III) dissociation followed a similar pattern for all pre-­‐equilibration times. A large part of the Eu(III) (30 -­‐ 50 %) dissociated almost instantaneously from the clay. 79
For the experiments with pre-­‐equilibration times less than 115 days, the Eu distribution between bulk bentonite and EDTA was within error of that at equilibrium between 20 and 100 days. That is, the amount of Eu remaining bound to the bentonite reached the value that would have been observed if the EDTA had been present from the start of the experiment before the Eu was introduced to the clay: This point is shown as the horizontal black line in Figure 3.1. Hence, for these systems, there was no indication of irreversibility. Experiments with pre-­‐equilibration times above 115 days took longer to reach the equilibrium distribution, although all but one of them had at least one point within error of equilibrium (by the experiment end the final plot of pre-­‐equil day 332 had 6.3 % still on the bentonite). Therefore, there was no convincing evidence for irreversibility in these systems. Linear portions of the plots in Figure 3.1 indicate a single first order dissociation rate constant, which may be determined from the gradient. Systems with pre-­‐equilibration times greater than 1 day showed different behaviour to that of the 1 day system, where the dissociation was distinctly faster, with a smaller decrease in gradient as EDTA contact time progresses, compared to the other systems. The average dissociation rate constant for this system (taken from day 1 of EDTA contact until the system reaches apparent equilibrium) is approximately 10-­‐6 s-­‐1. For the longer pre-­‐equilibration times, each of the plots shows more than one gradient. There is faster dissociation at the start of the experiment, but after approximately 7 days EDTA contact, there is a distinct reduction in reaction rate. First order dissociation rate constants were calculated by regression for the portions of the plots beyond 7 days of EDTA contact, and the results are shown in Table 3.1. 80
Amount of Eu in slow dissociating Dissociation rate constant (s-­‐1) τ (Days) fraction with errors (%) -­‐7
-­‐8
7 1.01 X 10 (±6.23 X 10 ) 17.3 (+3.1; -­‐2.9) 79 21 4.19 X 10-­‐8 (±8.51 X 10-­‐8) 11.9 (+25.2; -­‐4.8) 192 -­‐8 -­‐8
65 3.93 X 10 (±1.35 X 10 ) 19.3 (+5.4; -­‐3.5) 204 115 2.17 X 10-­‐8 (±1.70 X 10-­‐8) 20.5 (+8.6; -­‐6.0) 370 -­‐8 -­‐8
220 2.61 X 10 (±1.14 X 10 ) 24.3 (+6.7; -­‐5.2) 308 332 2.56 X 10-­‐8 (±2.87 X 10-­‐8) 20.7 (+11.2; -­‐7.3) 314 Table 3.1. Dissociation rate constants, reaction half time data and amounts for the most Pre-­‐
equilibration Time/day slowly dissociating fraction for Eu and bulk bentonite (pH = 7 ± 0.1; I = 0.1 M NaClO4). Errors are 2σ based on the error determined during regression of the data in Figure 3.1 The amounts of Eu(III) bound to the bentonite in the most slowly dissociating fraction are also shown in Table 3.1. There are relatively small differences between the rates for the different systems. The average Eu(III) dissociation rate constant is 4.3 x 10-­‐8 s-­‐1, with a range of 2.2 x 10-­‐8 – 1.0 x 10-­‐7 s-­‐1. The amount of Eu in the slowly dissociating fraction increases rapidly over the first few days of contact, but beyond 65 days, the amount seems to have reached equilibrium, with around 19 – 25 % slowly dissociating. Virtually all of the Eu in this system (> 96 %) was bound to the bentonite before addition of EDTA: therefore, the other 75 – 81 % of the Eu was bound to the bentonite, but did not get transferred to the slow fraction. This cannot be because it was locked into a different fraction, because it was released instantaneously when EDTA was added. Therefore, in this system, there must be an equilibrium between the Eu in the slowly dissociating and instantaneously available fractions. There is an apparent decrease in the value of the rate constant obtained from regression as pre-­‐equilibration time increases up to 115 days: however, there are large errors on the rate data. 81
When Eu was first added to the bentonite, it bound quickly, and within the first few hours of contact 96 % of the Eu was bound to the clay. Over the following month, there was a subsequent increase in the amount bound by 3.0 % and an associated increase in the Rd by a factor of 4.1. It is clear that more than this 3.0 % of the Eu changes its interaction with the bentonite, because there is different dissociation behaviour for the system with a short pre-­‐
equilibration time and those that were equilibrated for more than 1 day. Therefore, it seems that there is rapid uptake to some fraction that is bound, but may be (relatively) easily removed if a stronger sink becomes available. However, over time it seems that there is a transfer of some of the bound Eu to a fraction that dissociates more slowly. The small increase in Rd suggests that the transfer may also be associated with an increase in thermodynamic stability. However, some of the material (over half in most cases) also remains available for instantaneous dissociation, even after several months of contact time. 82
Speciation calculations for the bulk system Speciation modelling of the aqueous phase was performed for the Eu(III) in the bulk dissociation experiment systems with and without 0.01 M EDTA, and the results are presented in tables 3.2 and 3.3 respectively. Species Eu(CO3)+ Eu+3 EuSiO(OH)3+2 Eu(OH)+2 Eu(OH)2+ Eu(HCO3)+2 Eu(CO3)2-­‐ % distribution of solution species 47.01 28.43 19.46 4.31 0.53 0.16 0.11 Table 3.2. Eu(III) speciation in solution for bulk experiments without EDTA. [Eu]=7.9 x 10-­‐10 M; pH 7 assuming equilibrium with atmospheric CO2 and montmorillonite Species Eu(EDTA) Eu+3 Eu(C03)+ Eu(OH)+2 Eu(OH)2+ Eu(HCO3)+2 Eu(CO3)2-­‐ EuSiO(OH)3+2 % distribution of solution species 100.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Table 3.3. Eu(III) speciation in solution for bulk experiments with EDTA. [Eu]=7.9 x 10-­‐10 M; pH 7 assuming equilibrium with atmospheric CO2 and montmorillonite Table 3.2 shows that, in the solution for the bulk experiments, 47 % of the Eu(III) is present as Eu(CO3)+, 28 % is present as free Eu3+, 19 % is EuSiO(OH)3+2 , 4 % is present as Eu(OH)+2, with small percentages of carbonate and hydroxyl species. It is important to note that this calculation is only applicable to the Eu(III) that is not bound to the bentonite, which is small (approximately 1 %) in this case. 83
Table 3.3 is the calculated speciation for the same system in the presence of 0.01 M EDTA. It can be seen that all of the Eu(III) in this system is now complexed with the EDTA, thus reinforcing its effectiveness as a competing ligand for use in the bulk studies. In fact, when the experiment is performed in the reverse order (so Eu and EDTA are mixed before the introduction of bentonite) then the bentonite can’t compete with the EDTA, all of the Eu remains in solution complexed to the EDTA. Colloid Size Distributions The colloid and Eu/colloid mixture size distributions are shown in Figure 3.2: nominally, the <3 kDa fraction will contain particles no larger than ca. 1nm, and so this fraction can be considered as the ‘true solution’. 95.4 % of Colloid and 152-­‐Eu 100 80 76.4 60 40 23.6 20 1.2 0 1.9 0 1.5 0 500 -­‐ 450 nm 450 -­‐ 200 nm 200 nm -­‐ 3 kDa <3 kDa Fracgon size Figure 3.2. Bentonite colloid (blue columns) and 152Eu(III) (red columns) in each size fraction. The vast majority of the bentonite colloids were in nominal size fractions greater than 200 nm. For the 3kDa – 200 nm fraction, the ratio of Al : Mg (0.62 : 1, with Al = 0.065 ppm and Mg = 0.105 ppm) was far from that of bentonite theoretical value (5.02:1), and so this fraction does not seem to contain significant quantities of bentonite colloids. Also, the Al and Mg concentrations in this fraction were low compared to the others, and so a reliable bentonite colloid concentration could not be calculated. Given that the colloid stock solution had a concentration of 147 ppm, the material responsible for the Al and Mg in the 3kDa – 84
200 nm fraction represents a very small percentage of the total. The 152Eu(III) showed broadly similar behaviour to that of the colloid, and most of it was bound to the larger colloids (>450 nm; ≈ 95 %). Only 1.5 % of the 152Eu(III) remains in the true solution fraction (<3 kDa). Therefore > 98.5 % of the Eu is colloid associated. The Eu found in the colloidal fraction where no bentonite colloids were detected (10kDa – 200 nm) was probably due to Eu bound to low concentrations of smaller bentonite colloids that could not be detected or due to other trace colloidal material derived from the bentonite clay, such as quartz. These phases could generate colloids that could bind Eu, but they would not have the expected Al : Mg ratio. Even a very low mass of colloids in the smaller fractions would be particularly reactive, because of their very high specific surface area. Despite this, the amount of Eu observed in the range 3 kDa – 200 nm is small, and so the 152Eu(III) is largely associated with the bentonite colloids. Speciation calculations for the colloid system Speciation modelling of the aqueous phase was performed for the dissolved Eu(III) in the bentonite colloid dissociation experiment systems, and the results are presented in table 3.4. Species Eu(CO3)2-­‐ Eu(CO3)+ Eu(CO3)3-­‐3 Eu(OH)2+ EuSiO(OH)3+2 Eu(OH)+2 Eu+3 % distribution of solution species 88.77 9.95 1.06 0.11 0.09 0.02 0.002 Table 3.4. Eu(III) speciation in solution for colloid experiments (excluding colloid associated). [Eu]=7.9 x 10-­‐10 M; pH 8.8 assuming equilibrium with atmospheric CO2 and montmorillonite 85
88.77 % of the Eu in the solution is present as Eu(CO3)2-­‐, 9.95 % as Eu(CO3)+ and 1.06 % as Eu(CO3)3-­‐3 so that, overall, 99.78 % of the Eu in this system is present as a carbonate complex. As in the bulk experiments, filtration studies on the colloidal systems show that nearly all of the Eu(III) (99 %) is bound to the bentonite colloids although, unlike the bulk experiments, speciation modelling shows that the changed pH has affected the species distribution in solution. In the bulk studies Eu+3 comprises approximately 28 % of the solution species. By contrast, in the colloid studies, carbonate complexes dominate the solution composition. Nevertheless, there is essentially no difference in degree of association with the bentonite colloid (nearly all of the Eu(III) associates), which suggests that interaction between the Eu(III) ion and the bentonite binding site is strong. Eu(III) colloid dissociation Sodium perchlorate was not added to the colloid experiments, because this would have destabilised the colloids. Instead, the colloid solution was used without alteration. Figure 3.3 shows the dissociation of Eu(III) from the bentonite colloids as a function of Eu/colloid pre-­‐
equilibration time. 86
Figure 3.3. Natural log plot of the colloid dissociation experiment: ln(percentage bound to bentonite) vs time (days), as a function of pre-­‐equilibration time (pH = 8.8 ± 0.1). The black horizontal line represents the equilibrium distribution, and the dashed lines represent the experimental uncertainty (1.4 ± 1.1 %). In this system, when the Eu was added to the Dowex resin before the addition of colloids, 1.4 ± 1.1 % remained in solution: this represents the equilibrium position for the colloid experiments (horizontal line, Figure 3.3). For all samples, a similar pattern is observed. A significant part of the Eu(III) (approximately 30 -­‐ 40 %) dissociated almost instantaneously from the colloid. This was followed by slower dissociation over time. The dissociation observed for the colloids was very different to that of the bulk bentonite. More Eu is found in the slowly dissociating fraction, and there is no increase in the amount bound beyond 1 day pre-­‐equilibration time. Further, beyond the initial rapid dissociation, only a single rate constant is observed throughout the duration of the experiment. 87
First order dissociation rate constants and the amounts in the most slowly dissociating component were calculated by regression, and the results are shown in Table 3.5. Amount of Eu in slow Pre-­‐equilibration Dissociation rate constant System/day -­‐1
dissociating fraction with τ (Days) (s ) errors (%) 8.97 X 10-­‐7 (±2.27 X 10-­‐8) 1 7 69.8 (+3.1; -­‐2.9) 8.94 -­‐7
-­‐8
60.8 (+12.4; -­‐10.3) 10.5 -­‐7
-­‐7
64.9 (+22; -­‐17) 8.47 -­‐7
-­‐8
69.2 (+11; -­‐9.5) 8.87 7.66 X 10 (±5.63 X 10 ) 21 89 9.47 X 10 (±2.63 X 10 ) 9.04 X 10 (±2.55 X 10 ) Table 3.5. Dissociation rate constants, reaction half time data and amounts for the most slowly dissociating fraction (pH = 8.8 ± 0.1). Note, calculating an overall rate for all data in Figure 3.3 gives an average first order rate constant of 8.8 x 10-­‐7 s-­‐1. Errors are 2σ based on the error determined during regression of the data in Figure 3.3 The dissociation rate constants for the colloids are over an order of magnitude higher than for the bulk sample. The reasons for the differences are uncertain, but it could be due to the narrow size distribution of the bentonite colloids making the system more homogenous than bulk experiments (Figure 3.2). Beyond being more heterogeneous in terms of particle size, as a natural material, the bulk sample is also more chemically heterogeneous too. It seems likely that both of these factors contribute to the difference. With bulk bentonite it could be expected that a significant portion of the Eu would be bound within the cation exchange layers of the clay, simply because there is so much clay in these systems. Dissociation from these layers might be expected to be slower than dissociation from the surface of the clay and that can perhaps be seen in the dissociation rate measurements. For the bentonite colloids, due to their much smaller size, it could be expected that a significant portion of the Eu would probably be surface bound, due to the 88
much smaller size of the colloids and the greater specific surface area of the colloids. As such, the Eu would be more available for dissociation, as is maybe shown by the faster dissociation rate calculated for colloids. However, the amount of Eu in the most slowly dissociating fraction of the bentonite colloids (approximately 66%) is nearly 3 times higher than the amount in the same fraction for the bulk clay (approximately 19 %). In the dissociation of Eu from bentonite colloids, there is a two step process, with an initial instantaneous dissociation (approximately 30-­‐40 %) followed by a steady dissociation.Dissociation of Eu from bulk bentonite shows at least 3 dissociation steps: an initial instantaneous dissociation (presumably from the surface of the clay), a second dissociation which takes place on timescale of days, and a final, slower dissociation process. Long term release is represented by the linear portions of the data (from 7 days onwards) and, by this point, approximately 80 % of the Eu has already dissociated. Even though the proportion of Eu in the slowest dissociating fraction for the bulk dissociation experiments is 3 times smaller than that in the Eu colloid experiments, it is still an order of magnitude slower than dissociation from colloids. This could reflct the possibility that, in bulk clay, the Eu can become ‘trapped’ in the interlayers of the clay, thus making it less available for dissociation. Such a process would be less likely in the bentonite colloids simply because the colloid particles are so much smaller and thus the Eu cannot be so effectively isolated from the solution. The binding of Eu(III) to bentonite in this experiment was similar to that reported in the literature. Missana et al. (2008) and Bouby et al. (2011) found that it would associate to the clay, but in the presence of a competitor, it would dissociate. Bouby et al. (2011) used humic acid as a competitor and reported that on addition of humic acid an instantaneous, partial dissociation of the radionuclide from the bentonite occurred. The same was observed in 89
these experiments, since on initial addition of EDTA/resin at least 30 % of the Eu dissociates immediately. Wold (2010) estimated dissociation rate constants for some metal ions (see values above). These values were calculated from Kd values and association rates. We might expect some differences between the values reported here and those of Wold (2010), because Wold’s calculation assumes that all of the bentonite bound radionuclide represents a single fraction, whereas, the kinetic data for the bulk and colloidal experiments (Figures 3.1 and 3.3) show at least 3 and 2 fractions, respectively. Wold’s rate constant for Am(III) (5.6x10-­‐7 s-­‐1) is an order of magnitude higher than that recorded here for the bulk (4.3 x 10-­‐8 s-­‐1). The other comparable rate constant from Wold (2010), Cm(III) (1.7x10-­‐6 s-­‐1), is also higher than the Eu bulk experiment rate here (4.3 x 10-­‐8 s-­‐1). However, the colloid rate constants observed here fall within the range of the rate constants for trivalent metals reported by Wold (2010). Huber et al. (2011) have also provided dissociation rate constants for bentonite colloids from competition experiments using fracture filling material. For Am(III), the values were in the range 1 -­‐ 2.5 x10-­‐6 s-­‐1, whilst for Pu(IV), the range was 3.9 x10-­‐7 -­‐ 2.4 x10-­‐6 s-­‐1. The bulk rate constants in this study are lower than the data from Huber et al. (2011) (Table 3.1), but again the colloidal values are much closer (Table 3.2). At present, the mechanism responsible for the slow dissociation is uncertain. For a lanthanide ion, such as Eu(III), we would not automatically expect dissociation from a surface complexation site to be so slow. Therefore, it seems that there must be some other mechanism responsible for the slow dissociation. In the case of some radionuclides and concentrations, it has been suggested that surface precipitation could be responsible, particularly for tetravalent ions (Bouby et al. 2011). However, it seems less likely that is the case for a trivalent radionuclide, such as Eu(III), especially given the low concentrations of Eu and the ambient pH studied here. Metal ions do diffuse into the interlayer spaces of clay 90
structures, and it is possible that this is partly responsible for the slow dissociation: this might explain the difference between the bulk and colloidal Eu(III) rate constants, since the particles will be smaller for the colloid system, and so the extent to which a metal ion could get hidden inside the structure could be reduced, giving a faster dissociation rate. Clearly, additional work is required to deduce the mechanism of interaction more fully. 91
Conclusions For bulk bentonite, beyond approximately 100 days pre-­‐equilibration time, the amount of Eu in the most slowly dissociating fraction was effectively constant. However, the experiments suggested that there is a change in dissociation behaviour over pre-­‐equilibration times up to 1 week, with a reduction in the slowest dissociation rate constant of approximately an order of magnitude. In addition to a significant fraction that dissociates instantaneously and some material with intermediate rates, there does seem to be a most slowly dissociating fraction that has a characteristic rate constant of approximately 4.3 x 10-­‐8 s-­‐1. There was a difference between the colloid and bulk dissociation. The first order rate constant was significantly higher (i.e. faster reaction rate) for colloids, with an average dissociation rate constant of 8.8 x 10-­‐7 s-­‐1, over an order of magnitude faster than for bulk bentonite. There was also only a single rate constant in the colloid system, and the amount slowly dissociating did not increase beyond a pre-­‐equilibration time of one day. Previous work with humic substances has shown that in a system where some portion of colloid associated radionuclides dissociate slowly, this will dominate transport (Bryan et al. 2007). No evidence for ‘irreversible’ binding of Eu by bentonite has been observed. Although Eu(III) does bind both to bulk bentonite and bentonite colloids, and there were two different dissociation rates calculated, there was no evidence that the radionuclides were bound permanently. In groundwater systems, where bentonite colloids are stable and could move with the flow of the water, this observation is important. For colloid facilitated transport to be significant to the long-­‐term performance of a waste repository, then radionuclide association must take place, and this process must be irreversible (Mori et al. (2003)). If the association is reversible then, in the presence of a strong sink (in a repository this would be the host rock), 92
then it would be expected that the metal would dissociate from the colloid readily, and thus stop facilitated transport away from the repository. Whether or not the rate constants measured here mean that colloidal transport has any significance for facilitated transport would depend upon colloid residence times in the subsurface. 93
Acknowledgements The research leading to these results has received funding from the European Atomic Energy Community’s Seventh Framework Programme (FP7/2007-­‐-­‐-­‐2011) under grant agreement number 295487, The BELBaR project. NB would also like to thank the National Nuclear Laboratory for Strategic Research funding. 94
References Bouby. M, Geckeis. H, Lutzenkirchen. J, Mihai. S & Schafer. T (2011) Interaction of bentonite colloids with Cs, Eu, Th and U in presence of humic acid: A flow field-­‐flow fractionation study, Geochimica et Cosmochimica Acta, 75, 3866–3880. Bryan. N. D, Jones. D. L. M, Keepax. R. E, Farrelly. D. H, Abrahamsen. L. G, Warwick. P & Evans. N (2007) The role of humic non-­‐exchangeable binding in the promotion of metal ion transport in the environment, Journal of Environmental Monitoring, 9, 329-­‐347. Comans. R. N. J (1987) Adsorption, desorption and isotopic exchange of cadmium on illite: evidence for complete reversibility. Water Research, 21, 1573–1576. Comans. R. N. J, Haller. M & Depreter. P (1991) Sorption of cesium on illite – nonequilibrium behavior and reversibility. Geochemica et cosmochimica, Acta,, 55, 433– 440. Koning. A. De & Comans. R. N. J (2004) Reversibility of radiocaesium sorption on illite. Geochimica Et Cosmochimica Acta, 68, 2815–2823. Duro. L, Cera. E, Grive. M, Domenech. C, Gaona. X, & Bruno. J (2006) Development of the thermochimie thermodynamic database. Janvier. Andra report C. RP. 0ENQ. 06. 0001. P373. Geckeis. H, Schäfer. T, Hauser. W, Rabung. T, Missana. T, Degueldre. C, Möri. A, Eikenberg. J, Fierz. T & Alexander. W. R (2004) results of the colloid and radionuclide retention experiment (CRR) at the Grimsel test site (GTS), Switzerland -­‐impact of reaction kinetics and speciation on radionuclide migration. Radiochimica Acta. 92, 765-­‐774. Grive. M, Riba. O, Montoya. V & Duro. L (2010) Update of the thermochimie database: Reporting of new data selection 2010. June 2010. 95
Huber. F, Kunze. P, Geckeis. H & Schäfer. T (2011) Sorption reversibility kinetics in the ternary system radionuclide-­‐bentonite colloids/nanoparticles-­‐granite fracture filling material. Applied Geochemistry, 26, 2226–2237. Ivanov. P, Griffiths. T, Bryan. N. D, Bozhikov. G & Dmitriev. S (2012) The effect of humic acid on uranyl sorption onto bentonite at trace uranium levels. J. Environ. Monit., 14, 2968–2975. Missana. T, Alonso. U, García-­‐Gutiérrez. M, & Mingarro. M (2008) Role of bentonite colloids on europium and plutonium migration in a granite fracture. Applied Geochemistry, 23, 1484-­‐
1497. Monsallier. J. M, Schuessler. W, Buckau. G, Rabung. T, Kim. J. I, Jones. D, Keepax. R & Bryan. N (2003) Kinetic investigation of Eu(III)-­‐humate interactions by ion exchange resins, Analytical Chemistry, 75, 13, 3168-­‐3174. Möri. A, Alexander. W. R, Geckeis. H, Hauser. W, Schäfer. T, Eikenberg. J, Fierz. T, Degueldre. C & Missana. T (2003) The colloid and radionuclide retardation experiment at the Grimsel Test Site: influence of bentonite colloids on radionuclide migration in a fractured rock. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 217, 33-­‐47. Pitois. A, Ivanov. P. I, Abrahamsen. L. G, Bryan. N. D, Taylor. R. J & Sims. H. E (2008) Magnesium hydroxide bulk and colloid-­‐associated 152Eu in an alkaline environment: Colloid characterization and sorption properties in the presence and absence of carbonate. Journal of Environmental Monitoring, 10, 315-­‐24. Schäfer. T, Geckeis. H, Bouby. M & Fanghänel. T (2004) U, Th, Eu and colloid mobility in a granite fracture under near-­‐natural flow conditions. Radiochimica Acta. 92, 731-­‐737. Wold. S (2010) Sorption of prioritized elements on montmorillonite colloids and their potential to transport radionuclides, SKB Technical Report, TR-­‐10-­‐20. 96
Chapter 4 Uranium dissociation from bentonite colloids, a kinetic investigation The content of Chapter 4 will be used for the BELBaR report and will present data, which can be used in support of a safety case. It is to be submitted to a relevant journal in the near future. The contribution by the author of this thesis was: all of the filtration and dissociation experimental work; SEM imaging; interpretation of results and calculating of dissociation rate constants; and the initial and final writing up of the paper. All ICP-­‐AES for bentonite colloid determination was performed on a Perkin-­‐Elmer Optima 5300 dual view ICP-­‐AES in the School of Earth, Atmospheric and Environmental Sciences by Mr Paul Lythgoe. PHREEQC modelling was conducted by Dr Kurt Smith and interpreted by the author of the thesis. In a deep repository, it would be expected that the majority of the waste would be in the form of UO2, with the uranium in the +4 oxidation state (U(IV)). This chapter will however look at the binding of U(VI) to bentonite colloids. Chapter 5 studies the binding of Th(IV) to bentonite colloids, and this can be used as an analogue of U(IV). The study of U(VI) binding to bentonite colloids may also be of interest, if it is stabilised by possible pH changes or generated from radiolytic oxidation, since it is expected to be more mobile in a groundwater system than the U(IV). 97
Uranium dissociation from bentonite colloids, a kinetic investigation 1,3
1,2
Nick Sherriff , Francis Livens , Sarah Heath1, Katherine Morris2, Kurt Smith3 and Nick Bryan3* 1
Centre for Radiochemistry Research, School of Chemistry, University of Manchester. 2
Research Centre for Radwaste and Decommissioning, School of Earth, Atmospheric and Environmental Sciences, University of Manchester. 3
National Nuclear Laboratory, 5th Floor, Chadwick House, Birchwood, Warrington, WA3 6AE, UK *[email protected] 98
Abstract The interactions of 232U(VI), with bentonite colloids (171 ± 6.4 ppm) and ([U] = 5.43 x 10-­‐10 M; pH = 8.8 ± 0.2) have been studied using an ion exchange competition technique. Uptake of U(VI) onto the colloids was relatively weak, with uptake after 1 week 54 ± 1.5 %. The reversibility of the interaction was studied by allowing U(VI) to sorb to bentonite colloids for periods between 1 and 35 days. A fraction of the total U(VI) was extracted from the system instantaneously (28-­‐50 %), and after 4 days, the amount of U(VI) remaining on the colloids was 17-­‐ 25%. With time, the amount of U(VI) retained by the bentonite colloid decreases further, with a first order dissociation rate constant of 5.6 x 10-­‐7 s-­‐1 (values in the range of 3.1 x 10-­‐7 – 7.3 x 10-­‐7 s -­‐1). Linear regression shows the percentage of U(VI) in the dissociating fraction (24% (+34; -­‐12 %)), but complete dissociation of U(VI) from the colloids was not observed. The pre-­‐equilibration time of the 232U(VI) and bentonite colloids (1 – 35 days) did not affect either the amount of U(VI) bound to the bentonite colloid or the rate of dissociation in these experiments. Although slow dissociation was observed for U(VI), there was no convincing evidence for ‘irreversible binding’ of the radionuclide by the colloid. 99
Introduction For a High Level Waste repository, the multiple barrier method is usually employed. This involves treating the waste to be resistant to dissolution; placing the waste in a canister, which acts as the first physical barrier; then surrounding with a backfill, which acts as a second physical barrier. The backfill has safety functions, that is, the safety role of the backfill will change over geological time periods. Initially, the backfill will protect the canister that holds the waste, but after an amount of time during which the canister has degraded, then the safety function of the backfill changes to providing containment to limit the movement of waste components. Bentonite clay has been suggested as a backfill, due to its favourable characteristics, particularly low permeability, its swelling properties and its cation exchange abilities (SKB 1983, Karnland et al. 2010, Stankovic et al. 2011). For colloids to have a role in the transport of radionuclides away from a radioactive waste repository, then they have to be present, stable, mobile and be able to bind the radionuclide ‘irreversibly’ (Honeyman 1999; Ryan and Elimelech 1996; Missana et al. 2008). Bentonite is prone to the production of colloids and so, if it is to be used in a repository, then it is important to understand their interactions with radionuclides. Metal ion binding to bentonite/montmorillonite clay has been studied extensively (Comans 1987; Comans et al. 1991; Koning and Comans 2004), but there has been very little study of the interaction of colloidal bentonite with radionuclides. Most dissociation data comes from Wold (2010), where dissociation rates were calculated using Kd values and association rates. Models based on instantaneous equilibrium have been developed for radionuclide movement within bulk bentonite (e.g. Ochs et al. 2003). However, it is expected that kinetic 100
processes will determine whether the transport of radionuclides by colloids is significant for a repository safety case (Mori et al. 2003). Uranium is a major component of some nuclear waste and, as such, its association with and dissociation from bentonite colloids could be important. Previous studies have been undertaken using U(VI) and bentonite colloids. Batch experiments were performed on bentonite colloid/uranyl mixtures left for different equilibration periods to test competition between the colloids and granadiorite rock (originating from Grimsel in Switzerland). Even before introduction to the granadiorite rock, U(VI), Tc(VII) and Np(V) were predominantly not bentonite colloid associated (Bouby et al. 2010; Huber et al. 2011). Bouby et al. (2010) found that 5.8%, 4.4% and 11.7% of U(VI) ([U]=1 x 10-­‐7 mol L-­‐1) sorbed to the colloids. Huber et al. (2011) reported approximately 1.6 % of U(VI) ([U]=4.3 x 10-­‐7 mol L-­‐1) sorbed to the colloids. Bouby et al. (2011) performed batch experiments using bentonite colloids in a synthetic groundwater characteristic of the Grimsel test site in Switzerland. The colloids were spiked with Cs(I), Eu(III), Th(IV) and U(VI). U(VI) and Cs(I) showed the weakest interaction with the colloids, with an association of 6% for U(VI) ([U]= 4.2 x 10-­‐8 mol L-­‐1). Mori et al. (2003) performed transport experiments with several radionuclides, including U(VI). Again, it was found that the U(VI) was weakly associated to bentonite colloids, with 6% sorbed. However, sorption onto the colloids was stronger than sorption to bulk bentonite, and the Kd values increased over 12 weeks from 8 x 102 to 1.5 x 103 ml g-­‐1. From the transport experiment, however, no rate constant could be determined for U(VI) dissociation, as no breakthrough curve was observed, showing that the U(VI) was quantitatively removed from the colloids and became sorbed to the rock and infill during the transport experiment. Batch experiments were also performed on Grimsel granodiorite and fracture infill material with no bentonite colloids present. The U(VI) was more strongly 101
sorbed onto the granodiorite than on the fracture infill, but even after 12 weeks, equilibrium was not established, indicating strong kinetic control. Ivanov et al. (2012) studied the sorption of uranyl on bulk bentonite in the presence of humic acid. The humic acid limited the sorption to the bentonite, and was able to remove uranyl already sorbed. No slow dissociation from the bentonite was observed. First order rate constants have been calculated by Wold (2010) using known Kd values and first order association rate constants (kf), assuming kd=kf/kb where kb is the dissociation rate constant. For U(VI), a dissociation rate constant of 3 x 10-­‐3 hr-­‐1 (8.3 x 10-­‐7 s-­‐1) was obtained. The previous work has suggested that uranium does not associate with bentonite colloids strongly. However, dissociation rate constants have not been measured directly. The aim of this work was to measure dissociation rate constants for U(VI) from colloidal bentonite, using a sink that has no interaction with bentonite. Therefore, after a set equilibration time, a strong sink (Dowex ion exchange resin) was introduced to bentonite/uranyl mixtures to remove the metal ions from the colloids. This work follows on from a previous study of Eu(III) interactions with bentonite colloids (Sherriff et al. 2014). 102
Materials and Methods The clay used in this work is a sodium Wyoming bentonite. All water was deionised (18 MΩ) and all reagents were analytical grade. All experiments were repeated in triplicate. Experimental Methods Bentonite colloid generation and analysis The bentonite used in this study is Wyoming bentonite. Bentonite clay (10 g) was added to Milli-­‐Q (MQ) deionised water (1000 mL) and this suspension was sealed and stirred for 10 days. The slurry was evenly distributed into centrifuge tubes (50 mL x 20). Even slurry distribution was achieved by extracting the slurry whilst in motion using a syringe. The tubes were centrifuged on a BOECO C-­‐28A centrifuge (11 mins, 4000 rpm, RCF = 2683). The supernatant was removed from each tube, which was then refilled to the 50 mL mark with fresh deionised water to remove any soluble salts that might be in the raw clay. The tubes were then sonicated for 10 minutes to re-­‐suspend the clay. This process of centrifugation and sonication was repeated a further 3 times, with the supernatant being discarded. The supernatant after the 4th centrifugation step was the colloid stock (171 ppm), which was stored in a plastic bottle in the dark. This method is adapted from that of Bouby et al. (2011). Colloid suspension (1 ml) was added to deionised water (9 ml), and, to this, HNO3 (0.2 ml; 16 M) was added. The samples were analysed by ICP-­‐AES on a Perkin-­‐Elmer Optima 5300 dual view spectrometer, for Al and Mg concentrations to determine an overall concentration of bentonite colloids. Multiple standards in the range 0.05 -­‐ 10 ppm of Al and Mg were used for calibration. 103
Colloid suspensions were allowed to dry on a Leit Tab (12 mm, purchased from Agar scientific) and viewed on a Philips XL30 FEG ESEM. 232
232
U(VI) purification U(VI) was obtained through separation using a 2 ml UTEVA resin column purchased from Eichrom. The column is supplied pre-­‐conditioned in HCl (0.1 M), and was re-­‐conditioned with HCl washes (5 M, 3 x 5 ml). Before separation, the 232U(VI) was in radioactive with 228
Th(IV). The untreated 232U(VI) (2 ml, 8 kBq ml-­‐1) sample was adjusted to an HCl concentration of 5 M (with a total final volume of 20 ml). This solution was passed through the column. At this HCl concentration, the 228Th(IV) passed through the column and the 232
U(VI) was retained. The 232U(VI) was then eluted using HCl (0.1 M, 16 ml) to give a pure solution of 232U(VI) (16 ml, 1 kBq ml-­‐1). Uranium/bentonite colloid association 232
U(VI) (4 ml, 1 kBq ml-­‐1) was adjusted to pH 8.8 ± 0.2, colloid suspension (4 ml, 171 ppm) was added, and the 232U(VI)/colloid suspension (8 ml, 0.5 kBq ml-­‐1) was allowed to equilibrate for 24 hrs. Syringe filters (450 nm, 200 nm and 100 nm polyether sulphone (PES) membranes were used to filter the 232U(VI)/colloid suspension. The filters were first equilibrated for 68 hrs to the pH of the 232U(VI)/colloid suspension (8.8 ± 0.2). A separate filter was exposed to the 232U(VI)/colloid suspension to test whether the filters themselves sorbed the 232U(VI); less than 1 % of the 232U(VI) was sorbed by the filter. 104
Speciation calulations Solution speciation was calculated with PHREEQCi (Parkhurst et al. 1999), using the ThermoChimie v.7.b database (Duro et al 2006, Grive et al, 2010). Dissociation experiments Dowex 50W-­‐X4 200 mesh (H-­‐form) resin was converted to its sodium form. 20 g of Dowex resin were added to a sintered column, and washed with: deionised water (150 ml); HCl (2 M, 500 ml); deionised water (500 ml); NaCl (3 M, 500 ml); NaOH (0.1 M, 500 ml); and finally deionised water (500 ml). Experiments showed that the colloid did not interact with the pH pre-­‐conditioned resin (data not shown). 232
U(VI) solution (15 ml, 15 kBq) was added to a container and its pH was adjusted to 8.8 + 0.1). After pH adjustment, colloid suspension (135 ml, 171 ppm) was added to the 232U(VI) to give an overall 232U(VI)/bentonite colloid solution of 150 ml ([232U] = 100 Bq ml-­‐1, 5.43×10-­‐10 M; [colloid] = 171 ppm). Aliquots from this main solution were then removed at different pre-­‐equilibration times (3 x 10 ml), and to these conditioned Dowex resin was added (1.4 g). The samples are then allowed to rock gently for 1 hr before being centrifuged (15 mins, 4000 rpm, RCF = 2504) and a 0.1 ml aliquot was removed for 232U(VI) analysis. The samples were returned to storage and further aliquots were taken at intervals. 232
U(VI) analysis was performed via scintillation counting on a 1220 QUANTULUS ultra-­‐low level scintillation spectrometer. Scintisafe 3 (purchased from Fisher) was added to the samples (10 ml), followed by deionised water (0.9 ml), HCl (1 M, 1 ml), and finally the sample (0.1 ml). The mixtures were shaken vigorously for 30 seconds and allowed to settle for 24 hrs before analysis. 105
Results and Discussion Bentonite colloid SEM SEM images were taken of the colloid sample and are shown in Figures 4.1 and 4.2. Figure 4.1. Colloid suspension (171 ppm) allowed to dry on a Leit tab (12 mm) and viewed on a Philips XL30 FEG ESEM. The image shows several colloid particles. Figure 4.2. A single colloid particle (approximately 500 nm) on a Leit tab (12 mm) and viewed on a Philips XL30 FEG ESEM. 106
Figure 4.1 shows a patch of the Leit tab with several colloids < 1 μm. Figure 4.2 is a higher magnification image of a single colloid particle at approximately 500 nm in size. It is angular in places and some surface detail is visible, showing a rough looking texture. There were a smaller number of colloids between 250 -­‐500 nm, but acquiring a high-­‐resolution image of these was difficult as focusing the electron beam started to deform the colloids themselves. 232
U(VI)/Bentonite colloid association Previous studies (Bouby et al. (2010), Bouby et al. (2011), Mori et al. (2003), Huber et al. (2011)) have indicated that there is little or no association of U(VI) to bentonite colloids (U(VI) concentrations of 1×10-­‐7, 4.2×10-­‐8, 8.69×10-­‐7 and 4.3×10-­‐7 M, respectively). Filtration experiments were performed to measure the amount of U(VI) that was colloid bound. Sequential filtrations of the colloid stock and the 232U(VI)/colloid suspension were performed through 450, 200 and 100 nm PES filters. The results are shown in Figure 4.3. Figure 4.3. A comparison of U(VI) (3.38×10-­‐10 M) (blue columns) and bentonite colloid (171 ppm) (red columns) filtrates at pH 8.8 107
The colloid size distribution (red columns) shows 75 % in the greater than 450 nm fraction, 24 % in 450 – 200 nm, 1 % in the 200 – 100 nm fraction, and 0 % in the below 100 nm fraction, this is consistent with the SEM images (Figures 4.1 and 4.2) which show the majority of the colloids as being above 500 nm. U(VI) (blue columns) has 47 % in the >450 nm fraction, 4 % in the 450 -­‐200 nm fraction and 3 % in the 200 – 100 nm fraction, with 46 % in the <100 nm fraction. In the absence of bentonite colloids, no uranium is found in the >100 nm fraction, showing that approximately 54 % of the uranium is bentonite colloid associated. The 46 % in the <100 nm fraction cannot be associated with the bentonite colloids, as none were found in this size range. Bouby et al. (2010), Bouby et al. (2011) and Mori et al. (2003) found a much smaller amount of U associated to the bentonite colloids (5.8 %, 6 % and 6 % respectively). With their experiments they use U concentrations of 4.3 x 10-­‐7 M, 4.2 x 10-­‐8 M and 8.69 x 10-­‐7 M respectively and 20 ppm colloid suspensions, whereas in this study the U concentration is 3.38 x 10-­‐10 (two to three orders of magnitude less) and a 171 ppm colloid suspension (nearly 9 times the amount used in the other studies). The differences in experimental conditions could explain the differences seen in the association of U to bentonite colloids, this study has much lower mass concentration of U and a much higher concentration of bentonite colloids. It could therefore be possible that a saturation level is reached in the other studies, and this could explain the observation in previous studies of seemingly poorU/bentonte colloid binding. For the size fractions where bentonite colloids are present (> 100 nm), the amount of uranium bound per mass of colloid decreases with increasing colloid fraction size. This is expected, since the smaller colloids will have the higher specific surface area, and hence mass for mass more binding sites. 108
Speciation calculations Speciation modelling was performed for the U(VI) and the results can be seen in Table 4.1. The speciation calculations do not include any uranium that is colloid associated, in the true system the uranyl and bentonite colloids will be in equilibria. Species % distribution of solution species -­‐4
UO2(CO3)3 97.00 -­‐2
UO2(CO3)2 2.95 -­‐ UO2(OH)3 1.06 × 10-­‐2 UO2(CO3) 5.47 × 10-­‐03 UO2(OH)2 1.99 × 10-­‐03 +
UO2(OH) 2.67 × 10-­‐05 UO2(OH)4-­‐2 5.75 × 10-­‐06 UO2SiO(OH)3+ 2.34 × 10-­‐06 (UO2)2(CO3)(OH)3-­‐ 8.88 × 10-­‐07 +2
UO2 9.12 × 10-­‐09 Table 4.1. Speciation of U(VI) in solution; [U] = 5.43 x 10-­‐10 M, pH 8.8, assuming equilibrium with atmospheric CO2 and montmorillonite 97 % of the uranium in the solution is present as UO2(CO3)3-­‐4 and 2.95 % as UO2(CO3)2-­‐2. Overall, 99.95 % of the uranium species in this system is complexed with carbonate. The aqueous U(VI) species are in equilibrium with the U(VI) species bound to the colloid surface (S≡U↔Un+↔Un-­‐) and the equilibria of the system dictates that 54 % of the uranyl is colloid bound even when the amount of free uranyl is small (9.12 × 10-­‐09 %). As 54 % is still able to bind to the colloid then this could be an indication that the interaction of a uranyl ion and the bentonite binding site is strong. 232
U(VI) Dissociation from bentonite colloids The principle underlying the dissociation experiment was that, of the U(VI) added to the system, 54 % would bind to the bentonite colloid (as shown in the filtration studies) and 46 109
% would remain in the solution. Note that, when percentages are shown in all discussions it is referring to the total quantity of element in the system (i.e. the sum of element bound and element in solution) and dissociation is always discussed as relative to the total amount of U in the system. When the U/bentonite colloid system was interacted with Dowex, the first U removed onto the Dowex would be the 46 % in solution, and any removal over 54 % would reflect removal of U from the bentonite colloid. Figure 4.4 shows the dissociation of U(VI) from bentonite colloids as a function of U/colloid pre-­‐equilibration time. Figure 4.4. Natural log plot of the colloid dissociation experiment: ln(percentage ([U] = 5.43 x 10-­‐10 M) bound to bentonite) vs time (days), pH = 8.8 ± 0.2. The dashed black horizontal line represents the equilibrium distribution, and the thin lines represent the experimental uncertainty (1.7±0.2 %) In a separate experiment, U was added to the Dowex resin before the addition of bentonite colloid suspension. In that system, at equilibrium, 1.7 ± 0.2 % of the U was in solution. This 110
represents the position of equilibrium for the experiments shown in Figure 4.4 (the dashed horizontal line in Figure 4.4). In the absence of bentonite colloids, the uranyl is removed from the solution by the resin within 48 hrs. On initial contact with the Dowex resin there is an instantaneous reduction in the amount of U in solution (approximately 30-­‐50 %). The filtration experiment showed that approximately 46 % of U(VI) is not bound to the bentonite colloids, and the instantaneous reduction of U(VI) from the solution is consistent with that figure, so most of the initial rapid drop is probably due to removal from solution of the uranium not associated with the bentonite colloid. Analysis after 4 days of contact time shows 25.3 ± 0.80 %, 18.6 ± 0.9, 21.8 ± 0.4 % and 17.3 ± 0.4 % for pre-­‐equilibration times of 1, 7, 21 and 35 days, respectively, with an average of 20.8 % U remaining in solution. From this point on, the plots for the different equilibration times show broadly similar behaviour to each other. The gradient of dissociation is steeper between the initial dowex contact time and day 4, than the remainder of the experiment, this could be an indication of two dissociation rates. First order rate constants and the amount of uranium in the most slowly dissociating fraction were calculated by linear regression, and the results are given in Table 4.2. Pre-­‐equilibration System/day 1 7 21 35 Dissociation rate constant (s-­‐1) 7.8 × 10-­‐7 (± 5.8 × 10-­‐7) Amount of U in fraction (%) τ (Days) 27.9 (+24.3; -­‐13.0) 11 -­‐7
-­‐6
20.4 (+69.7; -­‐15.8) 25.8 -­‐7
-­‐7
24.9 (+7.2; -­‐5.6) 17.4 -­‐7
-­‐7
23.7 (+33.8; -­‐14.0) 11.7 3.1 × 10 (± 1.6 × 10 ) 4.6 × 10 (± 2.4 × 10 ) 6.9 × 10 (± 9.0 × 10 ) Table 4.2. Dissociation rate constants, reaction half time (τ ) data and amounts for the most slowly dissociating fraction (pH = 8.8 ± 0.1). Note, calculating an overall rate for all data in 111
Figure 4.4 gives an average first order rate constant of 5.6 x 10-­‐7 s-­‐1. Errors are 2σ based on the error determined during regression of the data in Figure 4.4. The average dissociation rate constant for the data in Table 4.2 is 5.6 × 10-­‐7 s-­‐1 (± 4.2 × 10-­‐7). The lack of kinetic data for U(VI) dissociation from bentonite colloids in the literature, makes comparisons difficult. Wold (2010) calculated a first order U(VI) dissociation rate constant from Kd values of 8.3 x 10-­‐7 s-­‐1. The experimental values in Table 4.2 are similar. The overall rate constant is within error of the value estimated by Wold (2010), and three of the individual values in the table are also within error (only the 21 day system is outside). Sherriff et al. (2014) measured a dissociation rate constant in a similar experiment using Eu(III) ((8.8 ± 9.1 )x 10-­‐7 s-­‐1). Again, the U(VI) rate constant measured here is within experimental error of this value. Sherriff et al. (2014) showed that beyond an initial rapid dissociation in their Eu(III)/bentonite colloid studies, only a single slow dissociation rate constant was observed. With U, however, there is evidence of a two-­‐step slow dissociation from the bentonite colloids. In the absence of bentonite colloids, the resin removes the U from solution within 48 hrs. There is an instantaneous removal of U(VI) from the system (30 – 50 %), which is taken to be free U(VI) that is not colloid bound (evidenced from the filtration studies). Then the first dissociation step that happens over a 4 day period (between time point day 1 to day 4) where approximately 30 % of U dissociates from the colloid, this is followed by a slow dissociation of the remaining U from the colloids. The average amount of U(VI) in the most slowly dissociating fraction is approximately 24 % of the U inventory. Sherriff et al. (2014) reported an average of Eu(III) amount of 66% (+ 12.1; -­‐9.9) in the most slowly dissociating fraction, but in that system, approximately 100 % 112
of the Eu(III) is associated with the bentonite colloids (76 % in the greater than 450 nm colloid fraction and 24 % in the 450 – 200 nm colloid fraction). Taking into account that 46% of the U(VI) is not colloid-­‐associated (the data in Figure 4.4 suggest that only 54 ± 6 % of the U is colloid associated) and that the rate was calculated from the linear portion of the graph (so after 4 days, when approximately 30 % of the U is removed from the system due to the first dissociation step) then this explains why there is less U in the most slowly dissociating fraction when compared to the similar Eu experiments from Sherriff et al. (2014). Hence, although a significant fraction of both metal ions is found in the most slowly dissociating fraction, it seems that the Eu has a higher affinity for bentonite than U, but this however does not appear to alter the speed of dissociation from the bentonite colloids. Because there is evidence of a two-­‐step dissociation, then it is likely that the uranyl ion is bound to the bentonite colloid in two different ways. The initial dissociation (day 1 – 4) observed is presumably U(VI) that is more available for dissociation (possibly surface bound). After 4 days, the slower dissociation (the long linear portion of the graph) dominates. It is possible that this colloid bound U(VI) is less available for dissociation because of a speciation change on the colloid surface or integration into the colloid layers. The dissociation is contradictory to other studies done with U(VI) sorption to bentonite colloids (Bouby et al. 2010, Bouby et al. 2011, Mori et al. 2003). In these other studies, the binding affinity was small (5.8 %, 6 % and 6 % respectively) compared to this study (54 %). However, in these other experiments, U concentrations of 4.3 x 10-­‐7 M, 4.2 x 10-­‐8 M and 8.69 x 10-­‐7 M respectively were used (two to three orders of magnitude more than used in this study), together with 20 ppm colloid suspensions (approximately 9 times less than used in this study). It is possible that some kind of saturation limit was reached, preventing measurement of dissociation rates for U(VI), as so little U(VI) was bentonite colloid bound to begin with. By contrast, this study used trace mass concentrations of U(VI) (5.43 x 10-­‐10 M) 113
in an excess of colloid suspension (171 ppm). In addition, this study did not measure any kind of transport, but rather studied direct dissociation from a stationary system, and as such was able to measure the direct dissociation kinetics of the U(VI). Sherriff et al. (2014) also studied the dissociation of Eu from bulk bentonite clay, in these experiments, up until 65 days of radionuclide/bentonite colloid pre-­‐contact time different amounts of Eu in the most slowly dissociating fraction were observed. After 65 days of pre-­‐
contact time no differences in the amount of Eu in the dissociating fraction were observed and the amount in the most slowly dissociating fraction gave an average of 21.2 %(+8.0 %; -­‐
5.5 %). This is much more similar to the amount of U in the most slowly dissociating fraction (approximately 24 %). It is worth noting however that the dissociation rate for bulk Eu-­‐
bentonite is an order of magnitude slower than the dissociation rate studied in this system. The mechanism responsible for the slow dissociation of the U(VI) from bentonite is unknown at present. Slow dissociation could reflect several possible effects, from surface precipitation (Bouby et al. 2011) to integration into the clay layers (Sherriff et al. 2014). The fact that both U(VI) and Eu(III), with their very different chemistries, show very similar rate constants could suggest that there is a common mechanism, but U shows at least 2 slow dissociating steps and Eu only shows 1. It has been suggested that slow dissociation of surface complexes could explain the slow dissociation. If this were the case, then we might expect that their dissociation rate constants would be different due to differing binding affinities to the bentonite and the chemistry of the different metals. It is clear that further study is required to provide information on the mechanism and where the ion actually binds to the bentonite colloid. 114
Conclusions Because 232U(VI) was used in the dissociation experiments, it only gave a small window of time to perform the experiments (approximately 65 days) before in-­‐growth of 228Th started to affect the data and therefore, no experiment reached steady state. However, all systems were trending towards the equilibrium position and as such, there is no evidence that the U(VI) in this study was bound to the bentonite colloids permanently (irreversibly). It is clear that the association to bentonite colloids is smaller for U(VI) compared to Eu(III) (this study suggests approximately 54 ± 1.2 %, with Eu(III) it is approximately 100 %). Of the 54 % of U(VI) that is colloid bound, approximately 30 % dissociates over a 4 day period and 24% (+34; -­‐12 %) dissociates slowly and this could have a significant effect on the transport of uranium if such behaviour were repeated in the environment. Other studies show that only a small percentage of the U interacts with bentonite colloids (Mori et al. 2003, Bouby et al. 2010, etc.) and as such no dissociation data could be obtained in these. This study uses a much smaller (two to three orders of magnitude) mass concentration of U and a much higher concentration of bentonite colloids (approximately 9 times more), it is possible that the other studies reached a saturation limit with the colloids that is not observed in this study. Even though the trivalent and tetravalent isotopes show an apparently stronger affinity for bentonite colloids (Mori et al. 2003, Bouby et al. 2010, Bouby et al. 2011) than U(VI), the dissociation rate constants are very similar. At least for these two ions, it seems that the chemistry of the metal ion has little effect on its dissociation rate, however, the observation of two dissociation steps is evidence for a difference in the way the metal binds to the clay. But even though the amount bound is lower than that for Eu(III), the controlling factor in the transport of radionuclides is the dissociation rate constant (Bryan et al. 2007). 115
These experiments have provided a measured value for the first order dissociation rate constant for uranyl dissociating from bentonite colloids ((5.6 ± 4.2 )× 10-­‐7 s-­‐1). When combined with other data, principally migration rate, this result will allow the significance of bentonite colloid transport of U to be assessed. 116
Acknowledgements The research leading to these results has received funding from the European Atomic Energy Community’s Seventh Framework Programme (FP7/2007-­‐-­‐‐2011) under grant agreement number 295487, The BELBaR project. NB would also like to thank the National Nuclear Laboratory for Strategic Research funding. 117
References Bouby. M, Filby. A, Geckeis. H, Geyer. F, Götz. R, Hauser. W, Huber. F, Keesmann. S, Kienzler. B, Kunze. P, Küntzel. M, Lützenkirchen. J, Noseck. U, Panak. P, Plaschke. M, Pudewills. A, Schäfer. T, Seher. H & Walther. C (2010) Colloid/nanoparticle formation and mobility in the context of deep geological nuclear waste disposal (Project KOLLORADO-­‐1; Final report), T. Schäfer & U. Noseck (Eds.), FZKA Wissenschaftliche Berichte, FZKA 751. Bouby. M, Geckeis. H, Lützenkirchen. J, Mihai. S & Schafer. T (2011) Interaction of bentonite colloids with Cs, Eu, Th and U in presence of humic acid: A flow field-­‐flow fractionation study, Geochimica et Cosmochimica Acta, 75, 3866–3880. Bryan. N. D, Jones. D. L .M, Keepax. R. E, Farrelly. D. H, Abrahamsen. L . G, Warwick. P & Evans. N (2007) The role of humic non-­‐exchangeable binding in the promotion of metal ion transport in the environment, J. of Environ. Monit., 9, 329-­‐347. Comans. R. N. J (1987) Adsorption, desorption and isotopic exchange of cadmium on illite: Evidence for complete reversibility. Wat. Res, 21, 1573-­‐1576. Comans. R. N. J, Haller. M & Preter. P. De (1991) Sorption of caesium on illite: Non-­‐
equilibrium behaviour and reversibility. Geochemica et cosmochimica, Acta, 55, 433-­‐440. Duro. L, Cera. E, Grive. M, Domenech. C, Gaona. X, & Bruno. J (2006) Development of the thermochimie thermodynamic database. Janvier. Andra report C. RP. 0ENQ. 06. 0001. P373. Grenthe. I, Drozdzynski. J, Fujino. T, Buck. E. C, Abrecht-­‐Schmitt. T. E & Wolf. S. F (2011) Chemistry of the actinides and transactinides Vol 1-­‐6. Chapter 5 Uranium, section 5.1, p253-­‐
255. Grive. M, Riba. O, Montoya. V, & Duro. L, (2010) Update of the thermochimie database: Reporting of new data selection 2010. June 2010. 118
Honeyman. B. D (1999) Colloidal culprits in contamination. Nature, 397, 23–24 Huber. F, Kunze. P, Geckeis. H & Schäfer. T (2011) Sorption reversibility kinetics in the ternary system radionuclide–bentonite colloids/nanoparticles–granite fracture filling material. Appl. Geochem. 26, 2226–2237. Ivanov. P, Griffiths. T, Bryan. N. D, Bozhikov. G & Dmitriev. S (2012) The effect of humic acid on uranyl sorption onto bentonite at trace uranium levels. J. Environ. Monit., 14, 2968–2975. Karnland. O (2010) Chemical and mineralogical characterization of the bentonite buffer for the acceptance control procedure in a KBS-­‐3 repository, SKB Technical Report, TR-­‐10-­‐60. Koning. A. De & Comans. R. N. J (2004) Reversibility of radiocaesium sorption on illite. Geochimica Et Cosmochimica Acta, 68, 2815–2823. Missana. T, Alonso. U, García-­‐Gutiérrez. M, & Mingarro. M (2008) Role of bentonite colloids on europium and plutonium migration in a granite fracture. Appl. Geochem. 23, 1484-­‐1497. Mori. A, Alexander. W. R, Geckeis. H, Hauser. W, Schafer. T, Eikenberg. J, Fierz. Th, Degueldre. C & Missana. T (2003) The colloid and radionuclide retardation experiment at the Grimsel Test site: Influence of bentonite colloids on radionuclide migration in a fractured rock. Colloids and Surfaces A: Physiochem. Eng. Aspects., 217, 33 – 47. Ochs. M, Lothenbach. B, Shibata. M, Sato. H & Yui. M (2003) Sensitivity analysis of radionuclide migration in compacted bentonite: a mechanistic model approach. J. Contam. Hydrol. 61, 313– 328. Parkhurst. D. L & 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: Denver, CO, USA, p 312 119
Ryan. J. N & Elimelech. M (1996), Review: colloid mobilization and transport in groundwater. Colloid Surface: A, 107, 1–56. Sherriff. N, Issa. R, Morris. K, Livens. F. R, Heath. S. L & Bryan. N. D (2014) Reversibility in Radionuclide/Bentonite Bulk and Colloidal Ternary Systems. Submitted to IDGTP-­‐
Geodisposal 2014 conference proceedings. SKBF/KBS Swedish Nuclear Fuel Supply Co/Division KBS, Final (1983) Storage of Spent Nuclear Fuel – KBS-­‐3. Stankovic. N, Logar. M, Lukovic. J, Pantic. J, Miljevic. M, Babic. B & Radosavljevic-­‐Mihajlovic. A (2011) Characterization of bentonite clay from “Greda” deposit. Process. Appl. Ceram. 5, 97-­‐101. Wold. S (2010) Sorption of prioritized elements on montmorillonite colloids and their potential to transport radionuclides, SKB Technical Report, TR-­‐10-­‐20. 120
Chapter 5 An investigation into thorium and americium dissociation from bentonite colloids The content of Chapter 5 will be used for the BELBaR report and will present data which can be used in support of a safety case. It is presented in the format of a research publication. The contributions by the author of this thesis were: all of the filtration and dissociation experimental work; the interpretation of results and calculating of dissociation rate constants; and the initial and final writing up of the paper. All ICP-­‐AES for bentonite colloid determination was performed on a Perkin-­‐Elmer Optima 5300 dual view ICP-­‐AES in the School of Earth, Atmospheric and Environmental Sciences by Mr Paul Lythgoe. PHREEQC modelling was conducted by Dr Kurt Smith and interpreted by the author of the thesis. 121
An investigation into thorium and americium dissociation from bentonite colloids 1
1,2
Nick Sherriff , Francis Livens , Sarah Heath1, Katherine Morris2, Kurt Smith3 and Nick Bryan3* 1
Centre for Radiochemistry Research, School of Chemistry, University of Manchester. 2
Research Centre for Radwaste and Decommissioning, School of Earth, Atmospheric and Environmental Sciences, University of Manchester. 3
National Nuclear Laboratory, 5th Floor, Chadwick House, Birchwood, Warrington, WA3 6AE, UK *[email protected] 122
Abstract The interactions of 228Th(IV) ([Th] = 3.79 x 10-­‐12 M; pH = 8.8 ± 0.2) and 241Am(III) ([Am] = 3.27 x 10-­‐9 M; pH = 8.8 ± 0.2), with bentonite colloids (171 ± 6 ppm) have been studied using an ion exchange competition technique. Th(IV) was not fully associated with bentonite colloids with uptake after 1 week of 78.3 ± 2.7%. Am(III) was weakly associated to the bentonite colloids, with uptake after 1 week of 20.1 ± 5.2 %. The reversibility of the interaction was studied by allowing the Th(IV) and Am(III) to sorb to bentonite colloids for 1 – 63 days and 1 – 49 days respectively. Cellulose phosphate (Cellphos) was then added to the radionuclide/bentonite colloid systems (1 g for Th(IV), 0.2 g for Am(III)), in an amount sufficient to retain the radionuclide when no bentonite colloids are present. A large fraction of the Th(IV) is initially sorbed by the Cellphos (75-­‐93 %) and after 7 days the amount of Th(IV) remaining on the colloids is only 1-­‐3 %. Over the time of the experiments, the amount of Th(IV) retained by the bentonite colloid appears to remain level and the amount bound to the bentonite colloid at the end of the experiment is 2.1 ± 0.88 % within error of the steady state of the system. A fraction of the Am(III) is also initially sorbed by the Cellphos (48-­‐94 %), and after 7 days, the amount of Am(III) remaining on the colloids is 1.2-­‐9.3 %. However, after 35 days contact time with cellulose phosphate, Am(III) is released back into the system, preventing dissociation rates from being calculated in this case. 123
Introduction For the disposal of higher activity radioactive waste, the concept of the multi barrier system is often employed. In this, initially, the waste is placed into a container/canister, and this is surrounded by the backfill. The backfill serves two main purposes. It protects the container/canister from potentially corrosive materials that may weaken it, but after a significant time for the canister to corrode, it will also stop the outward movement of radionuclides to the surrounding geology (SKB 1983, Karnland et al. 2010). Bentonite clay has been chosen as a backfill material due to its properties, specifically its cation exchange ability, its swelling properties and its availability ((SKB 1983, Karnland et al. 2010, Stankovic et al. 2011). With bentonite proposed as a backfill it is important to understand the interactions between bentonite colloids and radionuclides that could occur, as this could potentially affect the mobility of radionuclides from the waste repository. For colloids to facilitate in the transport of radionuclides from a repository they firstly have to be in the present in the system, they have to be stable, mobile and the radionuclide has to be able to sorb to the colloid ‘irreversibly’ (Honeyman 1999; Ryan and Elimelech 1996; Missana et al. 2008). Metal ion binding to bulk bentonite/montmorillonite clay has been studied extensively (Bradbury and Baeyens 1999, 2005; Morton et al. 2001; Kowal et al. 2004; Ochs et al. 2003; Guo et al. 2009). However, there has been little research into the interaction of metal ions and colloidal bentonite clay, and the kinetic data for this system is limited (Wold 2010). Most dissociation data in fact comes from Wold (2010), where rates have been calculated using association rates and Kd values, and it has been suggested that kinetics could be the determining factor in the transport of radionuclides (Mori et al. 2003). 124
Thorium is not a major component of current nuclear wastes, but can be used however as an analogue of Pu(IV) and other tetravalent actinide ions (such as U(IV)). 241Am is a component of nuclear waste, with a half-­‐life of 433 years (Runde and Wallace, 2011). As such, studying dissociation of both of these radionuclides from bentonite colloids will provide valuable information to support a disposal safety case. Ochs et al. (2003) used a mechanistic model to explore the sensitivity of migration of Cs, Ra, Am and Pb in compacted bentonite to geochemical parameters. The scenario modelled considers interaction between ground water and compacted bentonite, and takes account of trace species in the bentonite and the groundwater. The speciation of Am was strongly dependent on both the pH and composition of the pore water, but the variability of the systems makes it difficult to select sorption and diffusion coefficients for their model. Mori et al. (2003) performed fracture transport and batch experiments at the Grimsel test site in Switzerland. A mixture of radionuclides, including Th(IV) ([Th] = 1.1 x 10-­‐8 mol L-­‐1) and Am(III) ([Am] = 5.4 x 10 -­‐9 and 1.15 x 10-­‐8 mol L-­‐1) were prepared, and the tracer cocktails were injected into one end of a natural fracture and recovered after migration through the fracture. This was done with and without the presence of bentonite colloids. Without bentonite colloids, 6 – 58 % of the Am(III) was already in the colloidal fraction (shown through ultracentrifugation), illustrating the presence of Am radiocolloids (sometimes referred to as ‘intrinsic’ radiocolloids), presumably as a result of hydrolysis. With the addition of bentonite colloids (20 ppm), the amount of Am(III) in the colloidal fraction rose to 99 %. Th(IV) showed 20 – 30 % in the colloidal fraction before the addition of bentonite colloids (20 ppm), after which the amount of Th(IV) in the colloidal fraction rose to 94 %. Transport experiments were performed through the fracture with Am(III), both with and without bentonite colloids. Without bentonite colloids 34 % of added Am was recovered from the fracture, indicating faster transport than would occur in non-­‐facilitated transport. 125
Ijima et al. (2008) performed sorption experiments with Am and bentonite colloids at pH 8 and 10. Even before the addition of bentonite colloids, they found that Am radiocolloids were formed, consistent with the observations described above for this work. Iijima et al. (2010) performed batch experiments with Cs and Am with granite and bentonite colloids in an N2 atmosphere at pH 9.6. They prepared 3 batches, with different orders of addition using bentonite colloids (C), granite (G) and the radionuclide (R). Batch 1 was CR+G (i.e. bentonite colloids and radionuclide were initially equilibrated, then granite was added after 100 days), batch 2 was GR + C and batch 3 was GC + R. They found the binding of Am to bentonite colloids to be fully reversible, but also demonstrated the formation of Am radiocolloids for both the CR + G and GC + R batches. They cautioned that the formation of Am radiocolloids could lead to over estimations of the amount of Am bound to bentonite colloids. Grandiorite core migration studies were performed by Bouby et al. (2010) with Eu(III), Th(IV), Tb(III) and U(IV) using natural Grimsel ground water (pH 9.6) and bentonite colloids (10 ppm). It was shown by ultracentrifugation that the Th(IV) ([Th] = 1 x 10-­‐7 mol L-­‐1) was 99 % associated with the bentonite colloids before the experiment started and that Eu(III) ([Eu] = 1 x 10-­‐7) ,included as an analogue of Am(III), was 96 % associated. Recovery of the radioisotopes was directly related to the transport residence time of each experiment through the grandiorite core, but a pattern was observed, with recovery of Th(IV) being greater in all experiments compared to Eu(III) and Tb(III). This could indicate a greater strength of sorption for Th(IV) to the bentonite colloids. Batch experiments were performed by Huber et al. (2011) on a mixture of radionuclides including Th(IV)/bentonite colloid mixtures and Am(III)/bentonite colloid mixtures, in the presence of grandiorite fracture filling material. It was found that, again, Th(IV) ([Th] = 7.3 x 10-­‐8 mol L-­‐1) sorption to bentonite colloids appeared to be stronger than Am(III) ([Am] = 1.5 x 126
10-­‐9 – 8 x 10-­‐9 mol L-­‐1) sorption. Even after 1 year, equilibrium was not reached for either the Am(III) or the Th(IV) systems. Bouby et al. (2011) performed batch experiments using bentonite colloids in a synthetic groundwater, characteristic of the Grimsel test site in Switzerland. The colloids were spiked with Cs(I), Eu(III), Th(IV) and U(VI), with humic acid included as a competitor. Both Eu(III) ([Eu] = 6.56 x 10-­‐8 mol L-­‐1) and Th(IV) ([Th] = 4.31 x 10-­‐8 mol L-­‐1) were found to be bentonite-­‐
colloid associated and, after 3 years, there was still a significant fraction of Th(IV) bound to the bentonite colloids. It is unknown whether this partial ‘irreversibility’ is due to the binding strength of Th(IV) with bentonite colloids or some surface precipitation of Th(IV). First order rate constants have been estimated by Wold (2010) using the average of known Kd values and first order rate constants (kf) assuming Kd=kf/kb where kb is the dissociation rate constant. The dissociation rate constant for Th(IV) was not calculated but the analogous Pu(IV) rate was, and found to be 1.2 x 10-­‐6 s-­‐1. The dissociation rate constant for Am(III) was 5.6 x 10-­‐7 s-­‐1. Both values were calculated from Kd values obtained in experiments by Geckeis et al. (2004), which had radionuclide/bentonite colloid contact times between 1 hour and several weeks. This study will use 228Th(IV) (as an analogue for Pu(IV)) and 241Am(III). To study reversibility of the interactions, a strong sink (Cellphos resin) is also introduced to remove the metal from the colloid. This work follows on from previous studies of Eu(III) interactions with bentonite colloids (Sherriff et al. (2014)) and U(VI) interactions with bentonite colloids (Sherriff et al. (2015)). 127
Materials and Methods The clay used in this work is a sodium Wyoming bentonite (MX-­‐80), used without any pre-­‐
treatment. All water was deionised (18 MΩ) and all reagents were analytical grade. All experiments were repeated in triplicate. Before measurement, 228Th(IV) samples were equilibrated for 28 days to allow daughter nuclides to return to equilibrium. Experimental Methods Bentonite colloid generation and analysis Bentonite clay (10 g) was added to Milli-­‐Q (MQ) water (1000 mL), and the suspension was sealed and stirred for 10 days. The slurry was evenly distributed into centrifuge tubes (50 mL x 20), with an even slurry distribution achieved by extracting the slurry whilst in motion using a syringe. The tubes were centrifuged on a BOECO C-­‐28A centrifuge (11 mins, 4000 rpm, RCF = 2683 g). Each tube had the supernatant decanted and was then refilled to the 50 mL mark with fresh MQ water to remove any salts that may be in the clay, then sonicated for 10 minutes to re-­‐suspend the clay. This process of centrifugation and sonication was repeated a further 3 times, with the supernatant being discarded. The supernatant fluid remaining after the 4th centrifugation step constituted the colloidal stock (171 ppm), which was stored in the dark. This method is adapted from Bouby et al. (2011). Colloid suspension (1 ml) was added to MQ water (9 ml) and then HNO3 (16 M, 0.2 ml) was added. These samples were analysed for Al and Mg using ICP-­‐AES to determine an overall concentration of bentonite colloids. ICP-­‐AES was performed on a Perkin-­‐Elmer Optima 5300 dual view spectrometer. Multiple standards in the range 0.05 -­‐ 10 ppm of Al and Mg were 128
used for ICP-­‐AES quantification. This is the same colloid stock used in the work of Sherriff et al. (2015) and detailed characterisation data can be found in that paper. 228
Th(IV) purification 228
Th(IV) was obtained through separation using a 2 ml UTEVA resin column (Eichrom). The column comes pre-­‐conditioned in HCl (0.1 M), and was re-­‐conditioned with HCl (5 M, 3 x 5 ml) washes. Before separation, the 228Th(IV) was in equilibrium with 232U(VI). The untreated 232
U(VI)/ 228Th(IV) (2 ml, 8 kBq ml-­‐1) sample was adjusted to a HCl concentration of 5 M (20 ml). This solution was passed through the column. At this HCl concentration, the 228Th(IV) passed through the column and the 232U(VI) was retained. A further wash with HCl (5 M, 15 ml) gave a U(VI) free 228Th(IV) solution (35 ml, 230 Bq ml-­‐1). Thorium/bentonite colloid association 228
Th(IV) (4 ml, 230 Bq ml-­‐1) was adjusted to pH 8.8 ± 0.2, colloid suspension (6 ml, 171 ppm) was added and the 228Th(IV)/colloid suspension (10 ml, 92 Bq ml-­‐1) allowed to equilibrate for 24 hrs. Syringe filters (450 nm, 200 nm and 100 nm PES filters) were used to filter the 228
Th(IV) /colloid suspension. The filters were first equilibrated for 68 hrs to the pH of the 228
Th(IV) /colloid suspension (8.8 + 0.2). Separate filters were exposed to the 228
Th(IV)/colloid suspension to see if the filters themselves competed for the 228Th(IV), and it was found that, of the 228Th(IV) added, 9.3 ± 6.1 % sorbed to the filter. 129
Americium/bentonite colloid association 241
Am(III) (0.1 ml, 10 kBq ml-­‐1) was added to DI water (0.9 ml) and the pH adjusted to 8.8 ± 0.2. Colloid suspension (19 ml, 171 ppm) was added to this and the 241Am(III)/colloid suspension (20 ml, 50 Bq ml-­‐1) was allowed to equilibrate for 24 hrs. Syringe filters (450 nm and 200 nm PES filters) were used to filter the 241Am(III)/colloid suspension. The filters were first equilibrated for 68 hrs at the pH of the 241Am(III)/colloid suspension (8.8 ± 0.2). Separate filters were exposed to the 241Am(III)/colloid suspension to see if the filters themselves competed for the 241Am(III), but <1 % of the 241Am(III) sorbed to the filters. Speciation calculations Solution speciation was calculated with PHREEQCi (Parkhurst et al. 1999), using the ThermoChimie v.7.b database (Duro et al 2006, Grive et al, 2010). Dissociation experiments Cellulose phosphate (Cellphos) was purchased from Sigma-­‐Aldrich. 20 g was added to DI water (100 ml) and equilibrated to pH 8.8 for 24 hours with stirring. The DI water was decanted and the Cellphos was placed on a watch glass and air-­‐dried before being stored in a plastic screw top container (Li, 2011). The colloid did not interact with the pre-­‐conditioned resin (data not shown). 228
Th(IV) solution (1.5 ml, 4 kBq) was added to a container and its pH was adjusted to 8.8 ± 0.1. After the pH adjustment, colloid suspension (148.5 ml, 171 ppm) was added to the 228
Th(IV) to give a 228Th(IV)/bentonite colloid mix of 150 ml (26.7 Bq ml-­‐1, 3.79 × 10-­‐12 M). 130
Aliquots from this main solution were then removed at different pre-­‐equilibration times (3 x 10 ml at each time point) and, to these, conditioned Cellphos was added (1 g). The samples were then allowed to rock gently for 1 hr before being centrifuged (15 mins, 4000 rpm, RCF = 2504) and a 0.1 ml aliquot was removed for 228Th(IV) analysis. The samples taken for analysis were stored for 28 days as to equilibrate with 228Th(IV) daughter products. 228
Th(IV) analysis was performed via scintillation counting on a 1220 QUANTULUS ultra-­‐low level scintillation spectrometer. Scintisafe 3 (Fisher) was added to the samples (10 ml), followed by DI water (0.9 ml), HCl (1 M, 1 ml), and finally the sample (0.1 ml). The mixtures were shaken vigorously for 30 seconds and allowed to equilibrate for 28 days before analysis. 241
Am(III) solution (1.5 ml, 15 kBq) was added to a container and its pH was adjusted to 8.8 ± 0.1. After pH adjustment, colloid suspension (148.5 ml, 171 ppm) was added to the 241Am(III) to give an overall 241Am(III)/bentonite colloid mix of 150 ml (100 Bq ml-­‐1, 3.27 × 10-­‐9 M). Aliquots from this main solution were then removed at different pre-­‐equilibration times (3 x 10 ml for each sample point) and conditioned Cellphos was added (0.2 g). The samples are then allowed to rock gently for 1 hr before being centrifuged (15 mins, 4000 rpm, RCF = 2504) and a 1.5 ml aliquot was removed for 241Am(III) analysis via gamma-­‐ray spectrometry. Due to the non-­‐destructive nature of this analysis, aliquots were returned to the experiment following analysis. 241
Am(III) analysis was performed on a lead/copper shielded gamma ray spectrometer (Canberra 2020 coaxial HPGe gamma spectrometer with an Ortec 919E multi-­‐channel analyser). The samples were placed in identical containers and counted for 30 minutes before being returned to the experiment. 131
Results and discussions 228
Th(IV) and 241Am(III)/bentonite colloid association Previous studies (Bouby et al. (2010), Bouby et al. (2011), Mori et al. (2003), Huber et al. (2011)) show that Th(IV) is strongly associated to the bentonite colloids. It is, however, likely that a significant proportion of Th(IV) actually forms radiocolloids before any interaction with bentonite colloids; for example Mori et al. (2003) measured 20 – 30 % of Th(IV) ([Th] = 1.12 × 10-­‐8 M) in the colloidal fraction before any interaction with bentonite colloids. The same studies have also shown that Am(III) (or, in the case of Bouby et al. (2010), Eu(III) as an analogue) is also strongly bound to the bentonite colloids, while Mori et al. (2003) measured 6 -­‐58 % of Am(III) ([Am] = 5.4 x 10 -­‐9 and 1.15 x 10-­‐8 mol L-­‐1) in the colloidal fraction before introduction of the bentonite colloids. Here, the colloid stock was filtered sequentially through 450, 200 and 100 nm PES filters, and each filtrate was then analysed for colloid content by ICP-­‐AES. 228Th(IV)/colloid and 241
Am(III)/colloid suspensions were filtered separately through the same size PES filters (Am(III) was only filtered down to 200 nm, as no bentonite colloids were detected below this size) and analysed for Th(IV) and Am(III) content. These data were then combined to show the amount of radionuclide that is colloid-­‐associated in the two systems, and the results are shown in Figure 5.1. 132
Figure 5.1. Comparison of Th(IV) (1.31 x 10-­‐12 M), Am(III) (3.27 x 10-­‐9 M) and bentonite colloid solution (171 ppm) filtrates at pH 8.8. As can be seen in figure 5.1, the bentonite colloid size distribution shows proportions of 75 % in the >450 nm fraction, 24 % in 450 – 200 nm and 1 % in the <200 nm fraction, consistent with previous studies by Sherriff et al. (2014; 2015), which showed that there were no detectable colloids below 100 nm. Th(IV) is distributed 77 % in the >450 nm fraction, followed by 1 % in the 450 -­‐200 nm fraction and 22 % in the <200 nm fraction. Approximately 78 % of the Th(IV) is found in the >450 – 200 nm fraction, which is where 99 % of the bentonite colloids are detected, so it is likely that the Th(IV) in these fractions is associated with bentonite colloids. As only 1 % of the bentonite colloids are below 200 nm (no bentonite colloids detected <100 nm), then the remaining 22 % of Th(IV) cannot be bentonite colloid associated. Mori et al. (2003) showed that, even before the addition of bentonite colloids 20 – 30 % of the Th(IV) was in the colloidal fraction. This could be an indication of Th precipitation or 133
formation of Th radiocolloids. Complexation studies were also performed by Ekberg et al. (1999) which showed that Th(IV) formed polymeric species in addition to monomeric species in NaClO4, consistent with the formation of a Th(IV) precipitate or Th(IV) radiocolloids. In this study, below 200 nm there is a very small bentonite colloid concentration (1 %), so it is possible that the Th(IV) in this fraction (22 %) is already colloidal, either as radiocolloids, a precipitate or some other polymeric species (Mori et al. 2003, Ekberg et al. 1999) and as such, could be unavailable for binding to the bentonite colloids. Am(III) has a distribution of only 6 % in the >450 nm fraction, followed by 15 % in the 450 -­‐
200 nm fraction and 79 % in the <200 nm fraction. Of the Am(III) in the suspension, if the same arguments are made for Am as for Th (99 % of the bentonite colloids are in the >450 – 200 nm fractions so the Am present in these fractions is bentonite colloid associated) then only approximately 21 % of the Am is bentonite colloid associated. Mori et al. (2003) showed that even before the addition of bentonite colloids, 6 – 58 % of the Am(III) was in the colloidal fraction, but offered no explanation for this large range. Ijima et al. (2010) showed that in certain systems, radiocolloids of Am are formed, and that the formation of Am radiocolloids affects the concentration of Am in some of their experiments, especially when Am is allowed to mix with bentonite colloids before the addition of a strong sink. As previously mentioned, in these studies, below 200 nm there is a very small bentonite colloid concentration (1 %). Since it is possible that Am can form intrinsic radiocolloids (Ijima et al. 2010) and the amount of Am that is already in the colloidal fraction before the addition of bentonite colloids can be as much as approximately 60 % in similar conditions (Mori et al. 2003), then this may explain the seemingly poor binding affinity of Am to bentonite colloids. The 79 % of Am which shows no interaction with the bentonite colloids may reflect the formation of Am radiocolloids. 134
Speciation calculations Speciation modelling was performed for the Th(IV) and Am(III) and the results can be seen in Tables 5.1 and 5.2 respectively. The speciation calculations do not include any metal that is colloid associated, in the true system the metal and bentonite colloids will be in equilibria. Species % distribution of solution species -­‐1
Th(OH)3(CO3) 59.6 Th(OH)2(CO3)2-­‐2 32.3 Th(OH)4 7.88 Th(OH)2(CO3) 0.14 -­‐2
Th(OH)4(CO3) 5.74 x 10-­‐02 Th(OH)3+1 3.35 x 10-­‐02 Th(OH)(CO3)4-­‐5 9.98 x 10-­‐03 Th+4 1.25 x 10-­‐17 Table 5.1. Thorium speciation in solution (excluding colloid associated), Th(IV) 5.43 x 10-­‐10 M, pH 8.8, assuming equilibrium with atmospheric CO2 and montmorillonite Species % distribution of solution species -­‐
79.7 Am(CO3) 17.8 Am(CO3)2 +
-­‐3
Am(CO3)3 1.23 +
Am(OH)2 0.80 +2
AmOSi(OH)3 +2
Am(OH) +2
Am(HCO3) 0.32 0.12 1.18 x 10-­‐02 Am+3 4.35 x 10-­‐03 Table 5.2. Americium speciation in solution (excluding colloid associated), Am(III) 3.27 x 10-­‐9 M, pH 8.8, assuming equilibrium with atmospheric CO2 and montmorillonite The Th(IV) species predicted in the aqueous phase are shown in Table 5.1. 59.6 % of the thorium species formed in the system is Th(OH)3(CO3)-­‐1, 32.3 % is Th(OH)2(CO3)2-­‐2 and, overall, 91.9 % of the thorium species in this system are hydroxy carbonate complexes, 7.9 % of the thorium is present as thorium hydroxide. 135
The predicted Am(III) species present in the aqueous phase in the experimental systems are shown in Table 5.2. 79.7 % of the americium species formed in the system is Am(CO3)2, 17.8 % is Am(CO3)1+ and 1.23 % is Am(CO3)3-­‐3. The PHREEQC modelling shows the equilibrium distribution of species that are not bound to the bentonite colloids. This exercise also provides insight into saturation, via calculation of saturation indices. The saturation indices are different for the different solid phases which may form and, in this study, those of interest are the colloidal (coll), crystalline (cr), microcrystalline (mcr) and amorphous (am) forms. Since the saturation index is the negative log of the relevant ion activity product, a negative value of the saturation index indicates under saturation and a positive value indicates super-­‐saturation. The formation of amorphous and microcrystalline actinide species will occur first, then only at higher temperatures and long timespans are amorphous species expected to transform into crystalline species (Yui et al. 2003). Saturation indices for relevant Th and Am species are shown in table 5.3, Eu (from Sherriff et al. (2014)) is included for comparison to Am. Species ThO2 ThO2 ThO2 Saturation form coll cr mcr Saturation value -­‐6.22 3.11 1.88 Am(CO
3)(OH) Am(OH)3 am am 2.84 1.39 3)(OH) Eu(CO
Eu(OH)3 Eu(OH)3 cr am cr 0.40 -­‐5.18 -­‐3.04 Table 5.3. Saturation indices for Th(IV) and Am(III) species in the aqueous phase. A positive value indicates super-­‐saturation. The Eu(III) saturation indices, taken from Sherriff et al. (2014) are shown for comparison to Am(III) 136
While the saturation index for colloidal ThO2 indicates undersaturation (SI = -­‐6.22) it is positive values for both crystalline and microcrystalline ThO2 SI = 3.11 and 1.88 respectively). Although the solution is saturated with respect to both these species, and more so with respect to crystalline ThO2 , the rate of formation will also be important and it is more likely that the microcrystalline species would form, at least initially. Such behaviour would have an effect on the binding of Th(IV) to the bentonite colloids, and would be consistent with the filtration data. The Am(III) in these studies shows a weak affinity for the bentonite colloids (21 % association), and the calculated solution speciation of the Am(III) in table 5.2 is very similar to the Eu(III) speciation performed in studies from Sherriff et al. (2014) but in that study the Eu(III) showed a strong affinity to the bentonite colloid (nearly 100 % association). Table 5.3 shows the saturation indices for the relevant Am(III) phases in this study and the saturation indices of the relevant Eu(III) phases from Sherriff et al. (2014). The solution is supersaturated with respect to both amorphous Am(CO3)(OH) and amorphous Am(OH)3 (saturation indices of 2.84 and 1.39 respectively). Amorphous products will form first (Yui et al. 2003) and it is possible that these amorphous products form even before addition of bentonite colloids, and their formation is not reversible, thus making the Am(III) incorporated in them unavailable for binding to the bentonite colloid. This does not happen in the case of Eu(III). Systems, which are undersaturated with respect to both amorphous and crystalline Eu(OH)3 (SI -­‐5.18 and -­‐3.04 respectively). The systems is, in principle, slightly supersaturated with respect to crystalline Eu(CO3)(OH) (SI 0.40) but, due to its crystalline nature is expected to take a long time to form (so not before the addition of bentonite colloids). This is because the formation of the crystalline species need time and an elevated temperature (yui et al. 2003), in these systems there is no heat applied and the bentonite colloids are added on the same day of solution preparation. The differences in 137
solid phase formation between Am(III) and Eu(III) provide a possible explanation for the low measured association of Am(III) to bentonite colloids. 228
Th(IV) dissociation from bentonite colloids. The principle underlying the dissociation experiment was that, of the Th(IV) added to the system, 78 % would bind to the bentonite colloid (as shown in the filtration studies) and 22 % would remain in the solution. Please note that, when percentages are shown in all discussions it is referring to the total element concentration in the systems (sum of element bound and element in solution) and dissociation is always discussed as the removal of the total Th % from the system. When the Th/bentonite colloid system was interacted with Cellphos, the first Th removed onto the Cellphos would be the 22 % in solution, and any removal over 22 % would reflect removal of Th from the bentonite colloid. Figure 5.2 shows the dissociation of Th(IV) from bentonite colloids for different Th/colloid pre-­‐equilibration times. Figure 5.2. Plot of Th(IV) bound to bentonite colloids ([Th] = 3.79 x 10-­‐12 M) vs time (days), (pH = 8.8 ± 0.2). The dashed black horizontal line represents the steady state distribution the dotted line above and below being the error range (1.5 ± 1.1 %) 138
The steady state position was determined in a separate experiment where Th was added to the Cellphos resin before the addition of bentonite colloid suspension. In that system, 1.5 ± 1.1 % of the Th remained in solution after 7 days equilibration, and this was taken to be the steady state position for the experiments shown in Figure 5.2 (the dashed horizontal line in Figure 5.2). On initial contact, the Cellphos resin retains 76 – 93% of the Th from the experiment. Analysis after 7 days of desorption contact time shows 1.12 ± 0.26 %, 1.65 ± 0.82, 3.05 ± 1.48 % and 2.17 ± 1.04 % remaining associated with bentonite colloids for pre-­‐equilibration times of 1, 7, 21 and 35 days respectively. The 1 and 7 day equilibrated systems (the dark blue and orange points in figure 5.2) initially behave differently from the other data sets. Figure 5.3 shows the dissociation behaviour after 1 day of Cellphos contact for Th/bentonite colloid equilibration times of 1, 7, 21, 35 and 63 days. Figure 5.3, Th dissociation from bentonite colloids after 1 day contact with Cellphos for Th/bentonite colloid pre-­‐equilibration times of 1, 7, 21, 35 and 63 days. 139
Figure 5.3 shows that the bentonite colloids retain 6.75 % and 11.94 % of the Th after Th/bentonite colloid pre-­‐equilibration times of 1 and 7 days respectively. All subsequent pre-­‐equilibration times (21, 35 and 63 days) retain approximately the same amount of Th (23.9, 22.2 and 25 % respectively). The data suggest that, following initial uptake on to the bentonite colloid, there is a slow change in speciation of the Th associated with the colloid, which renders it less available. This may also be some kind of slow Th precipitation on to the bentonite colloid surface as speculated in the studies by Bouby et al. (2011). The dissociation data in figure 5.2 and 5.3 show that this slow change on the bentonite colloid surface appears to complete within 21 days of Th/bentonite colloid pre-­‐equilibration. Figure 5.2 shows that after 7 days contact with Cellphos, dissociation of Th from the bentonite colloids changes no further (i.e. no further dissociation is observed) for any sample points. Bouby et al. (2011) performed batch desorption experiments with Th and bentonite colloids using humic acid as a competitor, over a period of 3 years. Some Th remained on the bentonite colloids (approximately 30 %), and this partial irreversibility was attributed to possible Th precipitation on the surface of colloids. In the studies described here, however, desorption of the Th from the bentonite colloids falls to within error of the steady state position within 7 days of contact time with the Cellphos resin, for all samples. In this system, there is no indication of irreversible binding to bentonite colloids. 140
241
Am(III) dissociation from bentonite colloids The principle of the dissociation experiment was similar, with the expectation that, of the Am(III) added to the system, 21 % would bind to the bentonite colloid (as shown in the filtration studies) and 79 % would remain in the solution. When the Am/bentonite colloid system was interacted with Cellphos, the 79 % in solution would rapidly be removed by the Cellphos and any removal after 79 % would reflect removal from the bentonite colloid. As for Th, when percentages are shown they are all relative to the total amount of element added to the systems. Figure 5.3 shows the dissociation of Am(III) from bentonite colloids as a function of Am/colloid pre-­‐equilibration time. Figure 5.4. Plot of Am bound to bentonite colloids ([Am] = 3.27 x 10-­‐9 M) vs time (days), (pH = 8.8 ± 0.2). The dashed black horizontal line represents the steady state distribution and the dotted line above and below is the associated uncertainty (0.7 ± 0.1 %) 141
The steady state position was determined in a separate experiment where Am was added to the Cellphos resin before the addition of bentonite colloid suspension. In that system, 0.7 ± 0.1 % of the Am remained in solution after 7 days equilibration, and this was taken to be the steady state position for the experiments shown in Figure 5.4 (the dashed horizontal line in Figure 5.4). On initial contact, the Cellphos resin retains 50 -­‐ 94 % of the Am from the experiment. Analysis after 7 days of contact time shows 1.23 ± 1.54 %, 9.31 ± 1.28, 1.95 ± 1.04 % and 1.94 ± 0.46 % Am that is bentonite colloid associated for pre-­‐equilibration times of 1, 7, 21 and 49 days respectively. From day 35 of the experiment there is a change in the behaviour of the Am that is bound to the Cellphos resin, with a portion being released from the Cellphos, and increasing the Am content of the aqueous phase (41.6 ± 5.6 %, 19.6 ± 18.6 %, 19.3 ± 33.62 % and 25.4 % ± 8.8 % for pre-­‐equilibration times of 1, 7, 21 and 49 days respectively). This release of Am by Cellphos was unexpected, since preliminary experiments with Am and Cellphos showed no such behaviour, but these were only performed over a 7 day period, while figure 5.4 shows that this behaviour is not observed until at least after 14 days. It is clear that the behaviour observed for Am is not the same as that in the Eu dissociation work of Sherriff et al. (2014). The Eu dissociation studies performed by Sherriff et al. (2014) used Dowex ion exchange resin whereas the Am studies performed here used Cellphos, the only other differences in the experimental setups is the concentration of metal added (Sherriff et al. (2014) used [Eu] 1.02 × 10-­‐10 M; These studies used [Am] = 3.27 × 10-­‐9 M). In both Eu and Am experiments, the pH remained unchanged, and both the Cellphos ion exchange resin used in these experiments and the Dowex ion exchange resin used by Sherriff et al. (2014) release sodium when metal exchange takes place. Cellphos is capable of 142
binding metal ions below pH 5 (Padilha et al. 1997) and its optimum binding range is pH 4.0 to 9.0 (Li et al. 2002) so this experiment is within the pH range of this study. ICP-­‐AES analysis was performed for dissolved phosphate to assess whether the behaviour seen in the Am experiments arose from degradation of the Cellphos resin (data not shown), but no dissolved phosphate was detectable, suggesting that degradation does not occur. Moreover, any degradation of Cellphos might reasonably be expected to lead to similar behaviour in the Th experiments, but this is not observed. Extended reviews of literature on Cellphos as an ion exchange resin also showed no evidence of Cellphos degradation, so this can be excluded as the cause of the Am behaviour. It could be that the Am does not behave as the Eu because, in these systems, their solution chemistries are different. Am radiocolloids have been observed in these types of experimental systems (Ijima et al. 2008; Ijima et al. 2010; Mori et al. 2003), and the association behaviour of Am to bentonite colloids from the filtration studies in this work suggests that radiocolloids or other polymeric complexes may form, rendering the Am unavailable for bentonite colloid binding. In the Eu work performed by Sherriff et al. (2014) no evidence of Eu radiocolloids were observed (and reviews of the literature find limited reference to the formation of Eu radiocolloids in other studies), in those same systems over 97 % of the Eu was bound to the bentonite colloid. It is unknown whether the Am that is released back into the aqueous phase is bentonite colloid bound present as some other species. At longer contact times, after the Am is released, it can be seen in figure 5.4 that the Cellphos resumes the Am uptake and the percentage of Am in the aqueous phase decreases again. These differences between Am and Eu may be important. Association/dissociation experiments of this type often use Eu(III) as an analogue of Am(III), the assumption being 143
that experimental observations for the Eu are directly transferrable to the Am because of their similar chemistries. The data from these studies suggest that the chemistries of Am and Eu appear to be different in some circumstances. It is important to identify whether these differences arise from differences in the elements chemistries, or from experimental artefacts. 144
Conclusions Thermodynamic speciation modelling shows that in these systems, at this pH (8.8), even with very low mass concentrations ([Th] = 3.79 x 10-­‐12M; [Am] = 3.29 x 10-­‐9M) super-­‐
saturation can still occur. This can limit the amount of metal in the system that is available for binding to the bentonite colloid. is the most probable solid species are radiocolloid or polymetric hydroxyl form, and confirmation of this would be beneficial for future studies. Th(IV) associates substantially with the bentonite colloids (approximately 78 %), while the remaining Th(IV) does not bind to the bentonite colloids, possibly due to formation of radiocolloids or polymeric Th hydroxy species. Mori et al. (2003) showed that, before addition of bentonite colloids, Th(IV) ([Th] = 1.1 x 10-­‐8 mol L-­‐1) already has 20 -­‐30 % in the colloidal fraction. This is consistent with the studies performed in this thesis, where 22 % of the Th(IV) is not bentonite/colloid associated. There is an initial uptake on to Cellphos of 76 – 93 % of Th and after 7 days all experiments are within error of the steady state. Bouby et al. (2011) found evidence of a partial irreversibility in their Th(IV) desorption experiments ([Th] = 4.31 x 10-­‐8 mol L-­‐1), although, they used a mass concentration of Th(IV) 4 orders of magnitude higher than in this thesis, and also used humic acid as a competitor. They speculated that Th(IV) surface precipitation occurred, accounting for the apparent irreversibility. The speciation calculations in this thesis for the Th(IV) systems show that some species are already supersaturated, even at such a low concentration. So it is possible that in the work performed by Bouby et al. (2011), at higher Th(IV) concentrations, that Th(IV) species could precipitate onto the colloid surface, explaining the apparent partial irreversibility in their studies. In the Th(IV) studies of this thesis, there is no evidence of irreversibility, but there is evidence of a change in Th speciation on the bentonite colloids over a period of 7-­‐21 days after initial uptake. 145
Am(III) shows limited association with the bentonite colloids in the experiments described in this thesis (approximately 20 %), which may arise from formation of Am radiocolloids as seen in previous studies (Ijima et al. 2008; Ijima et al. 2010; Mori et al. 2003). Mori et al. (2003) reported that, even before the addition of bentonite colloids, 6 -­‐58 % of the Am(III) ([Am] = 5.4 x 10 -­‐9 and 1.15 x 10-­‐8 mol L-­‐1) was already in the colloidal fraction, intrinsic radicolloid formation is suspected to be the reason. Speciation data of the Am(III) studies in this thesis shows that there are supersaturated Am(III) species in this system, which would support the proposition of radiocolloid formation and hence explain the poor binding affinity of Am(III) to bentonite colloids in this system. After 7 days, 90 % of the Am is on the Cellphos ion exchange resin, but after 35 days there is a significant release of Am from the Cellphos back into the system. Whether the released Am is bentonite colloid bound or exists as radiocolloids/polymeric hydroxides is unknown, but after the release into the system, Cellphos resumes uptake of Am. What is clear from the data is that in these systems, Eu does not predict the association/dissociation behaviour of Am. Further work clearly needs to be performed to understand the origins of the differences between Am and Eu, and the potential significance of radiocolloids for both elements. Whether these findings have any bearing on bentonite colloid facilitated transport would depend on the rates of transport processes, Th(IV) appears to be fully dissociating in these sytems in a short amount of time. However for Am(III), if the formation of radiocolloids renders the Am unavailable for binding, then this could mean that the formation of Am radiocolloids could facilitate transport without bentonite colloid vector. This clearly needs to be explored further. 146
Acknowledgements The research leading to these results has received funding from the European Atomic Energy Community’s Seventh Framework Programme (FP7/2007-­‐-­‐-­‐2011) under grant agreement number 295487, The BELBaR project. NB would also like to thank the National Nuclear Laboratory for Strategic Research funding. 147
References Bouby. M, Filby. A, Geckeis. H, Geyer. F, Götz. R, Hauser. W, Huber. F, Keesmann. S, Kienzler. B, Kunze. P, Küntzel. M, Lützenkirchen. J, Noseck. U, Panak. P, Plaschke. M, Pudewills. A, Schäfer. T, Seher. H & Walther. C (2010) Colloid/nanoparticle formation and mobility in the context of deep geological nuclear waste disposal (Project KOLLORADO-­‐1; Final report), T. Schäfer & U. Noseck (Eds.), FZKA Wissenschaftliche Berichte, FZKA 751. Bouby. M, Geckeis. H, Lützenkirchen. J, Mihai. S & Schafer. T (2011) Interaction of bentonite colloids with Cs, Eu, Th and U in presence of humic acid: A flow field-­‐flow fractionation study, Geochimica et Cosmochimica Acta, 75, 3866–3880. Bradbury. M. H & Baeyens. B (1999) Modelling the sorption of Zn and Ni on Ca-­‐
montmorillonite, Geochimica et Cosmochimica Acta, 63, 325–336. Bradbury. M .H, Baeyens. B, Geckeis. H & Rabung. T (2005) Sorption of Eu(III)/Cm(III) on Ca-­‐
montmorillonite and Na-­‐illite. Part 2: Surface complexation modelling, Geochimica et Cosmochimica Acta, 69, 5403–5412. Ekberg. C, Albinsson. Y, Comarmond. M. J & Brown. P. L (1999) Studies on the complexation behaviour of thorium(IV). 1. Hydrolysis equilibria, J. Solution Chem, 29, 1, 2000. Duro. L, Cera. E, Grive. M, Domenech. C, Gaona. X & Bruno. J (2006) Development of the thermochimie thermodynamic database. Janvier. Andra report C. RP. 0ENQ. 06. 0001. P373. Grive. M, Riba. O, Montoya. V & Duro. L, (2010) Update of the thermochimie database: Reporting of new data selection 2010. June 2010. Geckeis. H, Schäfer. T, Hauser. W, Rabung. T, Missana. T, Degueldre. C, Möri. A, Eikenberg. J, Fierz. T & Alexander. W. R (2004) results of the colloid and radionuclide retention 148
experiment (CRR) at the Grimsel test site (GTS), Switzerland -­‐impact of reaction kinetics and speciation on radionuclide migration. Radiochimica Acta. 92, 765-­‐774. Guo. Z, Xu. J, Shi. K, Tang. Y, Wu. Y & Tao. Z (2009) Eu(III) adsorption/desorption on Na-­‐
bentonite: Experimental and modelling studies. Colloids and Surfaces A: Physiochem. Eng. Aspects, 339, 126-­‐133. Huber. F, Kunze. P, Geckeis. H & Schäfer. T (2011) Sorption reversibility kinetics in the ternary system radionuclide–bentonite colloids/nanoparticles–granite fracture filling material. Appl. Geochem. 26, 2226–2237. Iijima. K, Shoji. Y & Tomura. T (2008) Sorption behaviour of americium on bentonite colloid. Radiochim. Acta, 96, 721-­‐730. Iijima. K, Tomura. T, Tobita. M & Suzuki. Y (2010) Distribution of Cs and Am in the solution-­‐
bentonite colloids-­‐granite ternary system: effect of addition order and sorption reversibility. Radiochim. Acta, 98, 729-­‐736. Karnland. O (2010) Chemical and mineralogical characterization of the bentonite buffer for the acceptance control procedure in a KBS-­‐3 repository, SKB Technical Report, TR-­‐10-­‐60. Kowal – Fouchard. A, Drot. R, Simoni. E & Ehrhardt. J (2004) Use of spectroscopic techniques for uranium(VI)/montmorillonite interaction modelling. Environ. Sci. Technol., 38, 1399-­‐
1407. Li. N. C. Y (2011) Interactions of complex metal ion-­‐humic and magnetite systems. PhD thesis, University of Manchester, p 67. Li. W, Zhao. H, Teasdale. P. R, John. R & Zhang. S (2002) Application of a cellulose phosphate ion exchange membrane as a binding phase in the diffusive gradients in thin films technique for measurement of trace metals. Analytica Chimica Acta, 464, 331 – 339. 149
Missana. T, Alonso. U, García-­‐Gutiérrez. M, & Mingarro. M (2008) Role of bentonite colloids on europium and plutonium migration in a granite fracture. Appl. Geochem. 23, 1484-­‐1497. Mori. A, Alexander. W. R, Geckeis. H, Hauser. W, Schafer. T, Eikenberg. J, Fierz. Th, Degueldre. C & Missana. T (2003) The colloid and radionuclide retardation experiment at the Grimsel Test site: Influence of bentonite colloids on radionuclide migration in a fractured rock. Colloids and Surfaces A: Physiochem. Eng. Aspects. 217, 33 – 47. Morton. J, Semraw. J & Hayes. K (2001) An X-­‐ray absorption spectroscopy study of the structure and reversibility of copper adsorbed to montmorillonite clay. Geochimica et Cosmochimica, Acta, 65, 2709-­‐2722. Ochs. M, Lothenbach. B, Shibata. M, Sato. H & Yui. M (2003) Sensitivity analysis of radionuclide migration in compacted bentonite: a mechanistic model approach. J. Contam. Hydrol. 61, 313– 328. Padilha. P. M, Rocha. J. C, Moreira. J. C, Campos. J. T. S & Federici. C. C (1997) Preconcentration of heavy metal ions from aqueous solutions by means of cellulose phospahate: an application in water analysis, Talanta, 45, 317 – 323. Parkhurst. D. L & 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: Denver, CO, USA, p 312 Sherriff. N, Issa. R, Morris. K, Livens. F. R, Heath. S. L & Bryan. N. D (2014) Reversibility in Radionuclide/Bentonite Bulk and Colloidal Ternary Systems. Submitted to IDGTP-­‐
Geodisposal 2014 conference proceedings. Sherriff. N, Livens. F. R, Heath. S. L, Morris. K, Smith. K & Bryan. N. D (2014) Uranium dissociation from bentonite colloids, a kinetic investigation. In preparation. 150
Stankovic. N, Logar. M, Lukovic. J, Pantic. J, Miljevic. M, Babic. B & Radosavljevic-­‐Mihajlovic. A (2011) Characterization of bentonite clay from “Greda” deposit. Process. Appl. Ceram. 5, 97-­‐101. Wold. S (2010) Sorption of prioritized elements on montmorillonite colloids and their potential to transport radionuclides, SKB Technical Report, TR-­‐10-­‐20. Yui. M, Rai. D, Ochs. M, Shibata. M (2003) Applicability of thermodynamic database of radioactive elements developed for the Japanese performance assessment of HLW repository. J. Nucl. Sci. Technol. 40:5, 356 – 362. 151
Chapter 6 Conclusions drawn from the thesis and suggested further work The content of Chapter 6 will be used for the BELBaR report and will present conclusions which can be used in support of a safety case. It will synthesise all of the research in the thesis and also offer suggestions where further work should be planned. 152
6.1 Conclusions This thesis has examined the dissociation and, where possible, the dissociation kinetics of selected radioactive metals from bentonite clay colloids (Eu(III) in chapter 3, U(VI) in chapter 4 and Am(III)/Th(IV) in chapter 5) all with the intention of producing data to support a disposal safety case. Specific key data from all chapters are summarised in table 6.1. Radionuclide Association to bentonite colloid % Dissociation rate constant Evidence of irreversibly bound radionuclide to colloid in this study 152
Eu(III) 97 8.8 x 10-­‐7 s-­‐1 No 232
U(VI) 54 5.6 x 10-­‐7 s-­‐1 No 228
Th(IV) 78 N/A No 241
Am(III) 21 N/A No Table 6.1. Summary of radionuclide/bentonite colloid findings for these studies (Eu(III)/bulk bentonite studies not included) The four radionuclides that were studied in this piece of work showed vast differences in initial sorption to bentonite colloids at the pH of bentonite colloid formation (8.8). The experiments with U, Th and Am used the same stock solution of bentonite colloids, enabling direct comparisons of binding affinities in relation not only to colloid size, but also estimation of the effects of colloid surface area. These data are shown in table 6.2. Colloid diameter Colloid amount Fraction 1: >450 nm 75 % Fraction 2: 450-­‐200 nm 24 % Fraction 3: <200 nm 1 % Surface area 6.55 cm2 10.16 cm2 0.75 cm2 U bound 47 % 4 % 49 % Th bound 77 % 1 % 22 % Am bound 6 % 15 % 79 % Table 6.2. Comparison of colloid diameter and surface area per fraction vs. radionuclide percentage bound. Surface area estimated assuming spherical particles and average diameters of 550 nm, 325 nm and 150 nm 153
As can be seen from table 6.2, the largest surface area is in the 450-­‐200 nm fraction that contains 24 % of the bentonite colloids with a surface area of 10.16 cm2. However, U and Th bind in the fraction that has the highest percentage of colloids rather than the highest surface area. A significant proportion of all three metals is in the >200 nm fraction (U = 49 %, Th = 22 % and Am = 79 %) but this fraction contains only 1 % of the colloids (below 100 nm no colloids are detected) and has the smallest estimated surface area (0.75 cm2). It is probable that the metals in this fraction are not bentonite colloid bound. In the case of Th and Am, it is possible that the metal which is not associated with bentonite colloids is in the form of either homogenous radiocolloids or as some type of polymeric hydroxyl species, while U will exist predominantly as a carbonate complex in solution. It has been possible to calculate first order dissociation rate constants using linear regression. Although the chemistries of Eu(III) and U(VI) are different, both Eu(III) and U(VI) have very similar dissociation rates (8.8 x 10-­‐7 s-­‐1 and 5.6 x 10-­‐7 s-­‐1 respectively). The nature of experimenting with 232U means that, although the very low mass concentration is helpful, there is only a small time frame (approximately 3 months) before in-­‐growth of 228Th will affect the data and, on this timescale equilibrium was not reached. However, dissociation up to this point was continuous and there was no evidence of irreversibility. For the Eu(III) colloid dissociation experiments, which were performed over a longer time period, all experiments were within error of steady state. The Th(IV) dissociation experiments in chapter 5 showed strong association to bentonite colloids (78 %), and the Cellphos removes the majority of Th from the system. Other studies have shown strong Th/bentonite colloid interactions with evidence of pseudo irreversibility 154
(e.g. Bouby et al. 2011), but in this system, after 7 days, in all experiments, dissociation is within error of the steady state for the system. This shows that in the presence of a strong sink, Th bound to bentonite colloids will dissociate readily. For the first two sample points (1 and 7 day Th/bentonite colloid equilibration time), uptake of Th by the Cellphos is larger (93 % and 88 % for pre-­‐equilibration times of 1 and 7 days respectively) than for the rest of the data set (approximately 75 %). This could be an indication of a slow speciation change on the bentonite colloid surface. Rate data could not be calculated for this system as the samples all reached a steady state before the second sample point. The Am(III) experiments in chapter 5 are surprising. Since Eu(III) is widely used as an analogue of Am(III) then the results from the Am(III) dissociation experiments should be comparable with the Eu(III) experiments from chapter 3, but they are not. The association of Am to bentonite colloids is low when compared to the analogous Eu experiments (Eu binds 97 %, Am binds 21 %). However, there is evidence that the Am will form radiocolloids (Ijima et al. 2008; Ijima et al. 2010; Mori et al. 2003) and it could be the formation of radiocolloids that makes the Am unavailable for binding to the bentonite colloid. No evidence for Eu radiocolloids has been observed in this study and saturation indices for both metal systems are different. The subsequent ‘release’ of Am by the Cellphos resin at day 35 was unexpected. Almost all of the sample points follow this pattern and the reasons for it are unclear. Following this release of Am, the Cellphos then appears to start to ‘re-­‐absorb’ the Am. 155
It is important to understand whether the differences observed here between the two elements reflect real chemical differences or arise from an unidentified experimental artefact. As a result of the erratic behaviour of Am in this system, no dissociation rates could be calculated. When dealing with elements that may form intrinsic radiocolloids or polymeric species, speciation modelling, as carried out here, cannot provide a full representation of the aqueous species present. The potential for formation of species which are not accounted for in the databases but could be thermodynamically dominant means that modelling results, as in chapter 5 for Th and Am, should be interpreted with care. However, the saturation indices from thermodynamic calculations can be very helpful when explaining the differences seen in some systems. In all the systems investigated there was no evidence of any irreversible binding to the bentonite colloids so, for these metals at least, even though the presence of bentonite colloids could facilitate movement away from a deep geological disposal facility, when faced with the strong sink and huge excess of binding sites represented by the host geology, the bound metals will eventually dissociate from the bentonite colloids. However, the rate of dissociation may be relatively slow if ‘internally bound’ radionuclides, held within the actual bentonite colloid rather than on its surface, are present. 156
6.2 Further Work The work conducted in this thesis provides data that may be used in support of a safety case. However, in a repository, the radionuclides would leach from the inside out and there is a chance that radionuclides could be bound within the colloid. This could affect the dissociation rate, as internally bound radionuclides could take longer to dissociate from the colloid, hence they could travel further away from the repository itself. Experiments could be conducted with clay samples with internally bound lanthanides/actinides, by generating colloids from this clay and performing dissociation experiments to assess the difference (if any) between internally and externally bound metal dissociation rates. The association of U to bentonite colloids in these studies is higher than that reported in the literature. This could arise from experimental differences (colloid and/or metal concentrations, fraction separation methods etc.) or from the use of different isotopes (this study uses 232U, other studies use much longer lived and therefore higher mass concentration 233U and 238U). Bentonite colloid association studies with an isotope like 236U would be beneficial not only to follow association but also to allow a longer timescale than is possible with 232U and thus ensure the experiment does reach steady state with a similar dissociation rate constant. Alternatively, the 232U experiments could be redesigned to extend their duration, for example with radiochemical separation of 232U from decay products. The nature of 228Th makes changing parameters in preliminary experiments challenging. , Every time a parameter is changed, there is a 28 day delay to allow equilibration of daughter products and this can be very challenging logistically. As previously stated in the conclusions there appears to be very quick dissociation to the steady state position, so experiments with different quantities of ion exchange resin and different durations may allow the rate of the dissociation to be measured. 157
The Am studied shows erratic dissociation results and unexpected association results, and the only apparent difference between the Eu and Am experiments is the type of ion exchanger used. Identical experiments should be conducted using Cellphos as the ion exchanger rather than Dowex for the Eu dissociation experiments. This would give invaluable data that would confirm whether or not the Am dissociation experiment anomalies are caused by using Cellphos as the competitor. For both Am and Th there is the possibility that formation of radiocolloids or polymeric hydroxyl species is occurring. Ultrafiltration could be performed on samples with the same environment (pH, metal concentration etc.) to confirm the presence of these species. The possible existence of Eu radiocolloids should also be explored. 158
Appendix List of the thermodynamic data used for the speciation calculations in the thesis, taken from the sit. Database. Eu(III) +1.000Eu+3 +1.000CO3-­‐2 = Eu(CO3)+ log_k 7.9 #95SPA/BRU delta_h 167.549 kJ/mol # # Enthalpy of formation: -­‐1113.013 kJ/mol NaEu(CO3)2:5H2O(s) NaEu(CO3)2:5H2O = +1.000Na+ +1.000Eu+3 +2.000CO3-­‐2 +5.000H2O log_k -­‐20.9 #05VER/VIT #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Eu+3 -­‐1.000H+ +1.000H4(SiO4) = EuSiO(OH)3+2 log_k -­‐2.62 #Original data 07THA/SIN and 96JEN/CHO #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Eu+3 -­‐1.000H+ +1.000H2O = Eu(OH)+2 log_k -­‐7.8 #95SPA/BRU delta_h 79.824 kJ/mol # # Enthalpy of formation: -­‐811.337 kJ/mol +1.000Eu+3 -­‐2.000H+ +2.000H2O = Eu(OH)2+ log_k -­‐15.7 #07NEC/ALT delta_h 144.521 kJ/mol # # Enthalpy of formation: -­‐1032.471 kJ/mol +1.000Eu+3 +1.000H+ +1.000CO3-­‐2 = Eu(HCO3)+2 log_k 12.43 #95SPA/BRU #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Eu+3 +2.000CO3-­‐2 = Eu(CO3)2-­‐ log_k 12.9 #95SPA/BRU #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Eu+3 +1.000Edta-­‐4 = Eu(Edta)-­‐ log_k 19.47 +1.000Eu+3 +3.000CO3-­‐2 = Eu(CO3)3-­‐3 log_k 14.8 #05VER/VIT #delta_h kJ/mol # # Enthalpy of formation: kJ/mol 159
U(VI) +1.000UO2+2 +3.000CO3-­‐2 = UO2(CO3)3-­‐4 log_k 21.84 #03GUI/FAN delta_h -­‐39.2 kJ/mol # # Enthalpy of formation: -­‐3083.89 kJ/mol +1.000UO2+2 +2.000CO3-­‐2 = UO2(CO3)2-­‐2 log_k 16.61 #03GUI/FAN delta_h 18.5 kJ/mol #92GRE/FUG # Enthalpy of formation: -­‐2350.96 kJ/mol +1.000UO2+2 -­‐3.000H+ +3.000H2O = UO2(OH)3-­‐ log_k -­‐20.25 #03GUI/FAN delta_h 148.06 kJ/mol #Estimated by linear correlations # Enthalpy of formation: -­‐1728.43 kJ/mol +1.000UO2+2 +1.000CO3-­‐2 = UO2(CO3) log_k 9.94 #03GUI/FAN delta_h 5 kJ/mol #92GRE/FUG # Enthalpy of formation: -­‐1689.23 kJ/mol +1.000UO2+2 -­‐2.000H+ +2.000H2O = UO2(OH)2 log_k -­‐12.15 #03GUI/FAN delta_h 111.16 kJ/mol # # Enthalpy of formation: -­‐1479.5 kJ/mol +1.000UO2+2 -­‐1.000H+ +1.000H2O = UO2(OH)+ log_k -­‐5.25 #03GUI/FAN delta_h 43.458 kJ/mol # # Enthalpy of formation: -­‐1261.372 kJ/mol +1.000UO2+2 -­‐4.000H+ +4.000H2O = UO2(OH)4-­‐2 log_k -­‐32.4 #03GUI/FAN delta_h 156.138 kJ/mol # # Enthalpy of formation: -­‐2006.182 kJ/mol +1.000UO2+2 -­‐1.000H+ +1.000H4(SiO4) = UO2SiO(OH)3+ log_k -­‐1.84 #03GUI/FAN #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +2.000UO2+2 -­‐3.000H+ +1.000CO3-­‐2 +3.000H2O = (UO2)2(CO3)(OH)3-­‐ log_k -­‐0.86 #92GRE/FUG #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000UO2+2 = UO2+2 log_k 0 # #delta_h kJ/mol # # Enthalpy of formation: -­‐1019 kJ/mol 160
Th(IV) +1.000Th+4 -­‐3.000H+ +1.000CO3-­‐2 +3.000H2O = Th(OH)3(CO3)-­‐ log_k -­‐3.7 #09RAN/FUG #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Th+4 -­‐2.000H+ +2.000CO3-­‐2 +2.000H2O = Th(OH)2(CO3)2-­‐2 log_k 8.8 #09RAN/FUG #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Th+4 -­‐4.000H+ +4.000H2O = Th(OH)4 log_k -­‐17.4 #09RAN/FUG delta_h 152.688 kJ/mol # # Enthalpy of formation: -­‐1759.319 kJ/mol +1.000Th+4 -­‐2.000H+ +1.000CO3-­‐2 +2.000H2O = Th(OH)2(CO3) log_k 2.5 #09RAN/FUG #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Th+4 -­‐4.000H+ +1.000CO3-­‐2 +4.000H2O = Th(OH)4(CO3)-­‐2 log_k -­‐15.6 #09RAN/FUG #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Th+4 -­‐3.000H+ +3.000H2O = Th(OH)3+ log_k -­‐11 #09GRI/RIB delta_h 125.623 kJ/mol # # Enthalpy of formation: -­‐1500.554 kJ/mol +1.000Th+4 -­‐1.000H+ +4.000CO3-­‐2 +1.000H2O = Th(OH)(CO3)4-­‐5 log_k 21.6 #09RAN/FUG #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Th+4 = Th+4 log_k 0 # #delta_h kJ/mol # # Enthalpy of formation: -­‐768.7 kJ/mol 161
Am(III) +1.000Am+3 +2.000CO3-­‐2 = Am(CO3)2-­‐ log_k 12.6 #recalculated from 03GUI/FAN #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Am+3 +1.000CO3-­‐2 = Am(CO3)+ log_k 7.9 #recalculated from 03GUI/FAN delta_h 158.156 kJ/mol # # Enthalpy of formation: -­‐1133.774 kJ/mol +1.000Am+3 +3.000CO3-­‐2 = Am(CO3)3-­‐3 log_k 14.6 #Recalculated from 03GUI/FAN #delta_h kJ/mol # # Enthalpy of formation: kJ/mol -­‐2.000H+ +1.000Am+3 +2.000H2O = Am(OH)2+ log_k -­‐15.1 #03GUI/FAN, 88STA/KIM, 94RUN/KIM, 83EDE/BUC, 83CAC/CHO, 92WIM/KLE delta_h 143.704 kJ/mol # # Enthalpy of formation: -­‐1044.656 kJ/mol -­‐1.000H+ +1.000Am+3 +1.000H4(SiO4) = AmOSi(OH)3+2 log_k -­‐2.31 #Original data 07THA/SIN, 05PAN/KIM and 97STE/FAN #delta_h kJ/mol # # Enthalpy of formation: kJ/mol -­‐1.000H+ +1.000Am+3 +1.000H2O = Am(OH)+2 log_k -­‐7.2 #03GUI/FAN, 88STA/KIM, 94RUN/KIM, 83EDE/BUC, 83CAC/CHO, 92WIM/KLE delta_h 78.411 kJ/mol # # Enthalpy of formation: -­‐824.119 kJ/mol +1.000H+ +1.000Am+3 +1.000CO3-­‐2 = Am(HCO3)+2 log_k 13.43 #03GUI/FAN #delta_h kJ/mol # # Enthalpy of formation: kJ/mol +1.000Am+3 = Am+3 log_k 0 # #delta_h kJ/mol # # Enthalpy of formation: -­‐616.7 kJ/mol 162