Le ruthénium, issu du retraitement des combustibles usés de type Uranium-Oxyde a une très faible solubilité dans les verres de conditionnement de déchets radioactifs. Il précipite dans ces verres à l'état liquide sous forme de particules de RuO2 polyédrique ou aciculaire. Parce que leurs morphologies et leurs dispersions peuvent influencer les propriétés physico-chimiques des verres, la connaissance et le contrôle de leur mécanisme de formation sont d'extrême importance. Tout l'enjeu de cette thèse est de déterminer les différents chemins réactionnels de transformation du ruthénium, présent au sein du déchet calciné, lors de l'élaboration des verres. Par une approche de simplification progressive nous étudions les interactions entre la fritte de verre et des composés simples (NaNO3-RuO2) et plus complexes (calcinat NaNO3-RuO2- Al2O3). Grâce à l'apport de la microscopie et du XANES in situ en température nous suivons l'évolution de la composition, la spéciation et la morphologie des phases intermédiaires contenant du ruthénium. Ces composés sont caractérisés à l'état solide par MEB, DRX, HRTEM et spectroscopie d'absorption X au seuil K du ruthénium. Cette approche combinée nous permet de montrer que la modification de la spéciation du ruthénium au cours de l'élaboration du verre est à l'origine du contrôle de la morphologie des particules de RuO2 dans le verre. En particulier, la formation d'un intermédiaire réactionnel (Na3RuO4) est une des étapes fondamentales à l'origine de la précipitation de RuO2 de morphologie aciculaire. La formation de polyèdres dans le verre résulte au contraire de l'interaction directe de particules de RuO2 avec le verre à l'état liquide. / Ruthenium, arising from the reprocessing of spent uranium oxide fuel, has a low solubility in glass melt. It crystallizes in the form of particles of RuO2 of acicular or polyhedral morphology dispersed in fission product and actinides waste containment glass. Since the morphology of these particles strongly influences the physico-chemical properties, the knowledge and the control of their mechanism of formation are of major importance. The goal of this work is to determine the chemical reactions responsible for the formation of RuO2 particles of acicular or polyhedral shape during glass synthesis. Using a simplification approach, the reactions between RuO2-NaNO3, and more complex calcine RuO2-Al2O3-Na2O and a sodium borosilicate glass are studied. In situ scanning electron microscopy and XANES at increasing temperatures are used to follow changes in composition, speciation and morphology of the ruthenium intermediate species. Those compounds are thoroughly characterised by SEM, XRD, HRTEM, and ruthenium K-edge X-ray absorption spectroscopy. This combined approach allows us to show that the ruthenium speciation modification during vitrification is the key of control of the morphology of RuO2 particles in the glass. In particular, the formation of a specific intermediate compound (Na3RuO4) is one of the main steps that lead to the precipitation of needle-shaped RuO2 particles in the melt. The formation of polyhedral particles, on the contrary, results from the direct incorporation of RuO2 crystals in the melt followed by an Ostwald ripening mechanism.

Les verres nucléaires sont synthétisés par réactions chimiques à haute température entre un précurseur vitreux (fritte de verre) et un déchet calciné (calcinat) dans un procédé de calcination-vitrification. Le déchet est d'abord séché et dénitré (calcination) avant d'être mélangé à la fritte de verre (vitrification). Une succession de processus physico-chimiques d'imprégnation, diffusion, cristallisation et dissolution, est mise en jeu afin d'intégrer les éléments présents dans le calcinat au sein du réseau vitreux. Ces réactions, dépendantes de la composition des précurseurs et des conditions d'élaboration, doivent être complètes afin d'assurer l'homogénéité du verre et garantir son comportement à long terme. Ce travail a pour objectif de déterminer les réactions chimiques entre les précurseurs et de quantifier les cinétiques réactionnelles afin d'identifier in fine les processus responsables de leur limitation. Un système simplifié constitué d'une fritte de verre de type alumino-borosilicate de sodium et d'un calcinat contenant du nitrate de sodium et de l'oxyde d'aluminium (composés majeurs présents dans les calcinats complexes) est complexifié progressivement afin de déterminer l'influence des éléments de faible solubilité, présents initialement dans les solutions de produits de fission à vitrifier. Les cas des oxydes de molybdène et de néodyme sont en particulier étudiés. Les conditions de formation (temps, température) des phases cristallines de type molybdates (sodium, calcium) et aluminates (sodium, néodyme) ainsi que leur domaine de stabilité dans les calcinats sont déterminés. Les cinétiques de dissolution de ces phases dans la fritte de verre sont modélisées. Il est montré que la dissolution du molybdène, mis en évidence sous forme Na2MoO4, est contrôlée d'une part par la solubilité thermodynamique du MoO3 dans le verre, indépendamment de la dissolution des aluminates de sodium. D'autre part, les cinétiques de dissolution de Na2MoO4 et des aluminates présentent un comportement arrhénien avec la température dont les valeurs des énergies d'activation sont proches de celles de la viscosité du verre. Ces travaux décrivent également les mécanismes de formation d'intermédiaires réactionnels à l'origine de la cristallisation de la " yellow phase " (riche en oxydes de molybdène, d'alcalins et d'alcalinoterreux) pouvant se former dans des verres plus complexes. / Nuclear waste glasses are produced by chemical reactions between a solid waste (calcine) and a glassy precursor (glass frit) through a high-temperature vitrification process. The waste is first dried and calcined (to lose water and nitrogen respectively), then mixed with the glass frit. A succession of physicochemical processes of impregnation, diffusion, crystallization and dissolution is involved in order to incorporate the radioactive elements within the glassy network. These reactions, which are dependent on the precursor composition and the synthesis conditions, must be complete to ensure the homogeneity of the glass and to guarantee its long-term behavior. The aim of this work is to determine the chemical reactions between the precursors and to quantify the reaction kinetics in order to identify the processes responsible for their limitation. A simplified system consisting of a sodium-aluminum borosilicate glass frit and a calcine containing sodium nitrate and aluminum oxide (the principal oxides present in complex calcines) is progressively complexified to determine the influence of low solubility elements initially present in the fission product solutions to be vitrified. The cases of molybdenum and neodymium oxides are the focus of attention. The formation conditions (time, temperature) of crystalline molybdates (sodium, calcium) and aluminates (sodium, neodymium) and their range of stability in the calcines are determined. The dissolution kinetics of these phases in the glass frit is modeled. It is shown that the dissolution of molybdenum, as Na2MoO4, is controlled by the thermodynamic solubility of MoO3 in the glass. It is independent of the sodium aluminate dissolution. For both, Na2MoO4 and sodium aluminates, dissolution reactions present an Arrhenian behaviour and the activation energies are close to that of the viscous flow. This work also describes the formation mechanisms of intermediate phases which can lead to the crystallization of the "yellow phase" (enriched in molybdenum, alkali and alkaline-earth oxides) that can form in more complex glasses.

This project was specifically aimed at using increased acceleration separation, as a method to optimize the recovery of gold in an ore body mainly consisting of hematite. The specific gravity of gold is much higher in comparison to the carrying material, making it possible to separate gold from other materials such as silica, base metals and their oxides, usually associated with gravitation–gold–recovery processes. The ore body investigated in this project originated from a mined gold reef containing a large proportion of gold locked inside the gold pyrite complexes. In the mine's processing plant a gold pyrite concentrate was produced by means of a flotation process. The roasting process that followed, oxidized the pyrite to iron oxide (hematite) and sulphur dioxide. The gold particles which were locked up inside the pyrite gold complex were exposed or liberated, allowing the chemicals to penetrate the complex and dissolve the gold. After the cyanide gold extraction process, the material was pumped on to a mine reserve dump, referred to as tailings or tailings reserve dump. The tailings usually contain iron oxides, free gold, gold associated with iron oxides and gold associated with silica, and free silica, commonly referred to as calcine. The gold content on the calcine dump was significantly lower than the flotation concentrate before the extraction of the gold and it was no longer viable for the mine to process the tailings further. As the volume of the mine reserve dump increased over the years, it became viable to recover the gold in a high volume low grade plant. Several attempts were made to recover the gold in this dump, but due to the high cost of processing and milling the material, it was not done. The norm in the mining industry is that it is impossible to concentrate the gold by means of gravity separation techniques where the average particle sizes are smaller than 50 um in diameter and upgrading with inexpensive gravity separation techniques was ruled out by the mine, because the average particle sizes were too small. The dump investigated in this project differed from other reserve dumps in that the main phase of material in this dump was hematite and not silica. A suspension of this material would have different fall–out properties than other mine reserve dumps, because of the hematite's high specific gravity compared to silica. This property of the material birthed the idea that the material will respond positively to high acceleration separation, although the particle sizes were too small for effective upgrading according to the norm in the mining industry. Using acceleration concentration as a first stage in the gold recovery process the production cost per gram of gold produced could be reduced significantly. Firstly, the volume of concentrated material to be treated in the chemical extraction process was reduced ninety percent and secondly, the gold concentration was increased significantly. If the gold could be concentrated to more than twenty grams of gold per ton, it could be extracted economically with an aggressive chemical processes. This was not possible with low grade material contained in the dump. The theoretical principle, on which this project was based, was to make use of high acceleration separation to establish separation between the particles associated with the gold, and the particles not associated with gold. Applying a high gravitational force would have an influence on the velocity by which the particles would fall–out in a suspension. As the acceleration force increased the fall–out velocity would also be increased and the particles with higher specific gravity would be affected more. A factor that was equally important was the particle size and weight distribution. A large hematite particle would compete with a small gold particle due to the similarity in weight. This could cause loss in small gold particles or retention of hematite particles with no gold content. Very little scientific information was available on the material investigated and in order to assemble a concentration plant setup, the head grade and particle size distribution for both the dump and bulk sample were determined accurately. Thereafter, chemical analyses and mineralogical examination were done on a representative sample of the bulk sample, determining the chemical composition of the material. The results obtained thereof were evaluated and used to configure a pilot plant. A large bulk sample was processed in the pilot plant and from the analytical results the efficiency could be evaluated. The results at optimum acceleration forces applied, resulted in a recovery of 5% of the mass, with a gold concentrate of 90 g/t Au, which represented 58% recovery of the gold. The hematite with high specific gravity as main phase positively influenced the high acceleration separation process. It proved that if the specific gravity of particles in a suspension were increased, high acceleration separation could be applied effectively to separate much smaller particle sizes. / Thesis (M.Sc. Engineering Sciences (Chemical and Minerals Engineering))--North-West University, Potchefstroom Campus, 2012.

This project was specifically aimed at using increased acceleration separation, as a method to optimize the recovery of gold in an ore body mainly consisting of hematite. The specific gravity of gold is much higher in comparison to the carrying material, making it possible to separate gold from other materials such as silica, base metals and their oxides, usually associated with gravitation–gold–recovery processes. The ore body investigated in this project originated from a mined gold reef containing a large proportion of gold locked inside the gold pyrite complexes. In the mine's processing plant a gold pyrite concentrate was produced by means of a flotation process. The roasting process that followed, oxidized the pyrite to iron oxide (hematite) and sulphur dioxide. The gold particles which were locked up inside the pyrite gold complex were exposed or liberated, allowing the chemicals to penetrate the complex and dissolve the gold. After the cyanide gold extraction process, the material was pumped on to a mine reserve dump, referred to as tailings or tailings reserve dump. The tailings usually contain iron oxides, free gold, gold associated with iron oxides and gold associated with silica, and free silica, commonly referred to as calcine. The gold content on the calcine dump was significantly lower than the flotation concentrate before the extraction of the gold and it was no longer viable for the mine to process the tailings further. As the volume of the mine reserve dump increased over the years, it became viable to recover the gold in a high volume low grade plant. Several attempts were made to recover the gold in this dump, but due to the high cost of processing and milling the material, it was not done. The norm in the mining industry is that it is impossible to concentrate the gold by means of gravity separation techniques where the average particle sizes are smaller than 50 um in diameter and upgrading with inexpensive gravity separation techniques was ruled out by the mine, because the average particle sizes were too small. The dump investigated in this project differed from other reserve dumps in that the main phase of material in this dump was hematite and not silica. A suspension of this material would have different fall–out properties than other mine reserve dumps, because of the hematite's high specific gravity compared to silica. This property of the material birthed the idea that the material will respond positively to high acceleration separation, although the particle sizes were too small for effective upgrading according to the norm in the mining industry. Using acceleration concentration as a first stage in the gold recovery process the production cost per gram of gold produced could be reduced significantly. Firstly, the volume of concentrated material to be treated in the chemical extraction process was reduced ninety percent and secondly, the gold concentration was increased significantly. If the gold could be concentrated to more than twenty grams of gold per ton, it could be extracted economically with an aggressive chemical processes. This was not possible with low grade material contained in the dump. The theoretical principle, on which this project was based, was to make use of high acceleration separation to establish separation between the particles associated with the gold, and the particles not associated with gold. Applying a high gravitational force would have an influence on the velocity by which the particles would fall–out in a suspension. As the acceleration force increased the fall–out velocity would also be increased and the particles with higher specific gravity would be affected more. A factor that was equally important was the particle size and weight distribution. A large hematite particle would compete with a small gold particle due to the similarity in weight. This could cause loss in small gold particles or retention of hematite particles with no gold content. Very little scientific information was available on the material investigated and in order to assemble a concentration plant setup, the head grade and particle size distribution for both the dump and bulk sample were determined accurately. Thereafter, chemical analyses and mineralogical examination were done on a representative sample of the bulk sample, determining the chemical composition of the material. The results obtained thereof were evaluated and used to configure a pilot plant. A large bulk sample was processed in the pilot plant and from the analytical results the efficiency could be evaluated. The results at optimum acceleration forces applied, resulted in a recovery of 5% of the mass, with a gold concentrate of 90 g/t Au, which represented 58% recovery of the gold. The hematite with high specific gravity as main phase positively influenced the high acceleration separation process. It proved that if the specific gravity of particles in a suspension were increased, high acceleration separation could be applied effectively to separate much smaller particle sizes. / Thesis (M.Sc. Engineering Sciences (Chemical and Minerals Engineering))--North-West University, Potchefstroom Campus, 2012.