Les producteurs d’aluminium sont des acteurs majeurs dans l’économie du Québec. Le métal gris est apprécié pour sa légèreté et sa résistance à la corrosion. Bien qu’il possède des qualités indéniables, sa production consomme une quantité considérable d’électricité. L’amélioration de l’efficacité énergétique du procédé est donc primordiale d’un point de vue économique et environnemental. L’étude du contact électrique entre le carbone et la fonte à l’intérieur de l’ensemble cathodique fait partie de la liste des paramètres pour optimiser la consommation énergétique de la cellule Hall-Héroult. Ce contact doit offrir une résistance minimale au passage du courant nécessaire pour la réaction d’électrolyse. Au cours du temps, la qualité du contact se dégrade en raison de la transformation physique et chimique des matériaux. Cette thèse se concentre sur l’étude de la dégradation chimique de la surface de la fonte. La première partie étudie la pénétration des composés chimiques provenant du bain d’électrolyse à l’intérieur du bloc de carbone. Le gonflement sodique suit généralement la diffusion du sodium, un sous-produit de la réaction cathodique, et du bain. Le gonflement causé par le sodium a été mesuré directement à l’aide de LVDT alors que la diffusion du bain électrolytique a été déterminée par microtomographie à rayons X. La seconde partie évalue le mécanisme de la dégradation du contact électrique entre le carbone et la fonte. Des travaux en laboratoire ont été réalisés pour quantifier l’impact des paramètres d’opération. Les résultats obtenus ont été comparés par la suite à des échantillons industriels provenant de deux technologies pour évaluer leur degré de dégradation. Un modèle numérique a été calibré à partir de ces résultats pour estimer l’effet de la dégradation de la fonte sur la chute de voltage cathodique. Les résultats démontrent que les paramètres d’opération de la cellule d’électrolyse ont des effets sur la vitesse de pénétration des espèces chimiques dans le bloc de carbone. Un bain plus riche en sodium ou une densité de courant cathodique plus élevée augmente la vitesse de pénétration. La présence d’une nappe d’aluminium au démarrage de l’électrolyse au contraire divise le gonflement et la pénétration du bain de moitié. La vitesse de dégradation de la fonte suit la même tendance. De plus, une augmentation de température de 50 °C provoque une fusion partielle de la surface de la fonte. Ces résultats intégrés au modèle numérique montre que la dégradation du contact entre le carbone et la fonte augmente la chute de voltage cathodique mais aussi change la distribution du courant à la surface du bloc de carbone. La détermination du mécanisme de dégradation de la fonte par des essais en laboratoire combinée avec la pénétration des espèces chimiques constitue l’apport original de cette thèse et ceci permet d’éclaircir le processus d’évolution de ce contact électrique sous condition d’opération.

The purpose of this work was to determine the liquidus surface of the cryolite-rich portion of the ternary system Na₃AlF₆-CaF₂-AlF₃ and to establish the effect of Al₂O₃ on the operation of the Hall cell electrolysis. A series of isotherm of the cryolite-rich portion were graphed. It was shown that pseudo-binary phase diagrams of Al₂O₃ and bulk composition in the cryolite-rich portion of the Na₃AlF₆-CaF₂-AlF₃ system were found to be simple eutectic. The temperatures and the alumina contents of the double solubility limit, two important parameters for the Hall cell, of the joins 95 Na₃AlF₆/5 AlF₃-Na₃AlF₆, 90 Na₃AlF₆/ 10 NaCaAlF₆ and 85 Na₃AlF₃/15 AlF₃-NaCaAlF₆ were determined. The cryolite liquidus temperature of the quaternary system Na₃AlF₆-CaF₂-AlF₃-Al₂O₃ was found to be expressed by: T_Liq.. (C) = 1009.4 + 4.059(CaF₂) - 1.167(CaF₂)² + 0.968 x (CaF₂)(AlF₃) - 0.105(CaF₂)(AlF₃)² + 0.073 x (CaF₂)²(AlF₃) + 0.002(CaF₂)² (AlF₃)² - 4.165 x (AlF₃) - 0.054(AlF₃)² - 5.33(Al₂O₃) for CaF₂ 3.8~11.25%, AlF₃ 5~20%. / M.S.

The ultimate goal of this study is to examine the feasibility of extracting silicon from silica through electrolysis. The objective of the thesis was to evaluate the physico-chemical properties of a cryolite-lithium fluoride mixture as an electrolyte for the electrolysis process. A study of 86.2wt%Cryolite and13.8wt%Lithium fluoride melt with silica concentration varying from 0-4wt% and temperature range of 900-1000°C was done. Three properties were measured using two sets of experiments: 1) Dissolution Behaviour Determination, to obtain a) solubility limit, b) dissolution rate (mass transfer coefficient) and 2) density using Archimedes’ Principle. The study concluded that solubility and dissolution rate increases with temperature and the addition of LiF to cryolite decreases the solubility limit but increases the rate at which silica dissolves into the melt. With addition of silica, the apparent density of electrolyte first increases up to 2-3wt% and the drops.

The ultimate goal of this study is to examine the feasibility of extracting silicon from silica through electrolysis. The objective of the thesis was to evaluate the physico-chemical properties of a cryolite-lithium fluoride mixture as an electrolyte for the electrolysis process. A study of 86.2wt%Cryolite and13.8wt%Lithium fluoride melt with silica concentration varying from 0-4wt% and temperature range of 900-1000°C was done. Three properties were measured using two sets of experiments: 1) Dissolution Behaviour Determination, to obtain a) solubility limit, b) dissolution rate (mass transfer coefficient) and 2) density using Archimedes’ Principle. The study concluded that solubility and dissolution rate increases with temperature and the addition of LiF to cryolite decreases the solubility limit but increases the rate at which silica dissolves into the melt. With addition of silica, the apparent density of electrolyte first increases up to 2-3wt% and the drops.

The ultimate goal of this study is to examine the feasibility of extracting silicon from silica through electrolysis. The objective of the thesis was to evaluate the physico-chemical properties of a cryolite-lithium fluoride mixture as an electrolyte for the electrolysis process. A study of 86.2wt%Cryolite and13.8wt%Lithium fluoride melt with silica concentration varying from 0-4wt% and temperature range of 900-1000°C was done. Three properties were measured using two sets of experiments: 1) Dissolution Behaviour Determination, to obtain a) solubility limit, b) dissolution rate (mass transfer coefficient) and 2) density using Archimedes’ Principle. The study concluded that solubility and dissolution rate increases with temperature and the addition of LiF to cryolite decreases the solubility limit but increases the rate at which silica dissolves into the melt. With addition of silica, the apparent density of electrolyte first increases up to 2-3wt% and the drops.

The ultimate goal of this study is to examine the feasibility of extracting silicon from silica through electrolysis. The objective of the thesis was to evaluate the physico-chemical properties of a cryolite-lithium fluoride mixture as an electrolyte for the electrolysis process. A study of 86.2wt%Cryolite and13.8wt%Lithium fluoride melt with silica concentration varying from 0-4wt% and temperature range of 900-1000°C was done. Three properties were measured using two sets of experiments: 1) Dissolution Behaviour Determination, to obtain a) solubility limit, b) dissolution rate (mass transfer coefficient) and 2) density using Archimedes’ Principle. The study concluded that solubility and dissolution rate increases with temperature and the addition of LiF to cryolite decreases the solubility limit but increases the rate at which silica dissolves into the melt. With addition of silica, the apparent density of electrolyte first increases up to 2-3wt% and the drops.

A jazida do granito Madeira, associada à fácies albita granito, é um depósito de classe mundial com minério disseminado de Sn, Nb, Ta e F (Y, ETR, Li, U, Th) e, em sua parte central, contém um depósito de criolita maciça com 10 Mt (teor de 38% de Na3AlF6). O objetivo do trabalho foi compreender que contexto geológico permitiu a formação desta associação rocha-minério única no mundo. Para tanto, foram efetuados estudos isotópicos (Sm-Nd, Rb-Sr e Pb-Pb) e estudos tectônicos, enfocando o granito Madeira, seus correlatos e as rochas regionais. Durante uma primeira fase extensional, formaram-se as rochas vulcânicas do Grupo Iricoumé (1.890 a 1881 Ma), constituindo um complexo de caldeiras, e os corpos graníticos associados da Suíte Intrusiva Mapuera, ambos gerados a partir de fontes mantélicas. Concomitantemente aos estágios finais do vulcanismo iniciou-se a sedimentação na bacia Urupi (possivelmente um rift), acompanhada por um segundo pico de vulcanismo há 1.825 Ma. Fluidos mantélicos migraram para a zona afetada pela extensão regional, ascenderam acompanhando as isotermas e iniciaram a fenitização da crosta. Na continuidade deste processo, durante uma segunda fase extensional, rochas até refratárias tornaram-se fusíveis e originaram 5 magmas diferentes, todos com assinatura de fonte crustal e mantélica, que se posicionaram, entre 1.839 e 1.824 Ma, em estruturas geradas na fase anterior, formando os 3 corpos graníticos da Suíte Madeira. Numa terceira fase tectônica, desta feita transtensiva, fluidos mantélicos, possivelmente de natureza carbonatítica, fenitizaram rochas de nível crustal mais alto, enriquecidas em Sn, e nelas introduziram F, Nb, Y, ETR, U e Th em concentrações anômalas. Da fusão destas rochas resultou o magma do albita granito que se alojou, há 1.822 Ma, dentro do granito Madeira, mas com uma orientação N-S discordante da orientação geral NE-SW do granito Madeira e da estrutura que o aloja. / The deposit of the Madeira granite, associated with albite granite facies is a world-class deposit with disseminated ore of Sn, Nb, Ta and F (Y, REE, Li, U, Th), and its central part contains a deposit of massive cryolite with 10 Mtons (containing 38% of Na3AlF6). The objective was to understand the geological context to the formation of ore-rock association unique in the world. Therefore isotopic studies were performed (Sm-Nd, Rb-Sr and Pb-Pb) and tectonic studies focusing on the Madeira granite, its related and regional rocks. During a first extensional phase volcanic rocks of the Iricoumé Group (1890 to 1881 Ma) was originated forming a caldera complex and granitic bodies associated with Mapuera Intrusive Suite, both generated from mantle sources. At the same time the final stages of volcanism began the sedimentation in Urupi basin (possibly a rift), followed by a second peak of volcanism in 1825 Ma ago. Mantle fluids migrated to the area affected by regional extension rose following the isotherms and started the fenitization crust. Continuing this process in a second extensional phase , rocks become refractory and fuses originating 5 different magmas, all with crustal signature and mantle source, which is positioned between 1839 and 1824 Ma, in structures generated in previous phase, forming 3 granitic bodies of Madeira suite . In a third tectonic phase,, this time transtensive, mantle fluid, possibly of a carbonatitic fenitizated rocks from higher crustal level , enriched in Sn, and introduced F, Nb, Y, REE , U and Th in anomalous concentrations. The fusion of these rocks resulted in the albite granite magma that has positioned, there in 1822 Ma, within the Madeira granite, but with a NS orientation ,discordant of the general NE-SW of Madeira granite and the structure that it was contained.

A jazida do granito Madeira, associada à fácies albita granito, é um depósito de classe mundial com minério disseminado de Sn, Nb, Ta e F (Y, ETR, Li, U, Th) e, em sua parte central, contém um depósito de criolita maciça com 10 Mt (teor de 38% de Na3AlF6). O objetivo do trabalho foi compreender que contexto geológico permitiu a formação desta associação rocha-minério única no mundo. Para tanto, foram efetuados estudos isotópicos (Sm-Nd, Rb-Sr e Pb-Pb) e estudos tectônicos, enfocando o granito Madeira, seus correlatos e as rochas regionais. Durante uma primeira fase extensional, formaram-se as rochas vulcânicas do Grupo Iricoumé (1.890 a 1881 Ma), constituindo um complexo de caldeiras, e os corpos graníticos associados da Suíte Intrusiva Mapuera, ambos gerados a partir de fontes mantélicas. Concomitantemente aos estágios finais do vulcanismo iniciou-se a sedimentação na bacia Urupi (possivelmente um rift), acompanhada por um segundo pico de vulcanismo há 1.825 Ma. Fluidos mantélicos migraram para a zona afetada pela extensão regional, ascenderam acompanhando as isotermas e iniciaram a fenitização da crosta. Na continuidade deste processo, durante uma segunda fase extensional, rochas até refratárias tornaram-se fusíveis e originaram 5 magmas diferentes, todos com assinatura de fonte crustal e mantélica, que se posicionaram, entre 1.839 e 1.824 Ma, em estruturas geradas na fase anterior, formando os 3 corpos graníticos da Suíte Madeira. Numa terceira fase tectônica, desta feita transtensiva, fluidos mantélicos, possivelmente de natureza carbonatítica, fenitizaram rochas de nível crustal mais alto, enriquecidas em Sn, e nelas introduziram F, Nb, Y, ETR, U e Th em concentrações anômalas. Da fusão destas rochas resultou o magma do albita granito que se alojou, há 1.822 Ma, dentro do granito Madeira, mas com uma orientação N-S discordante da orientação geral NE-SW do granito Madeira e da estrutura que o aloja. / The deposit of the Madeira granite, associated with albite granite facies is a world-class deposit with disseminated ore of Sn, Nb, Ta and F (Y, REE, Li, U, Th), and its central part contains a deposit of massive cryolite with 10 Mtons (containing 38% of Na3AlF6). The objective was to understand the geological context to the formation of ore-rock association unique in the world. Therefore isotopic studies were performed (Sm-Nd, Rb-Sr and Pb-Pb) and tectonic studies focusing on the Madeira granite, its related and regional rocks. During a first extensional phase volcanic rocks of the Iricoumé Group (1890 to 1881 Ma) was originated forming a caldera complex and granitic bodies associated with Mapuera Intrusive Suite, both generated from mantle sources. At the same time the final stages of volcanism began the sedimentation in Urupi basin (possibly a rift), followed by a second peak of volcanism in 1825 Ma ago. Mantle fluids migrated to the area affected by regional extension rose following the isotherms and started the fenitization crust. Continuing this process in a second extensional phase , rocks become refractory and fuses originating 5 different magmas, all with crustal signature and mantle source, which is positioned between 1839 and 1824 Ma, in structures generated in previous phase, forming 3 granitic bodies of Madeira suite . In a third tectonic phase,, this time transtensive, mantle fluid, possibly of a carbonatitic fenitizated rocks from higher crustal level , enriched in Sn, and introduced F, Nb, Y, REE , U and Th in anomalous concentrations. The fusion of these rocks resulted in the albite granite magma that has positioned, there in 1822 Ma, within the Madeira granite, but with a NS orientation ,discordant of the general NE-SW of Madeira granite and the structure that it was contained.

A jazida do granito Madeira, associada à fácies albita granito, é um depósito de classe mundial com minério disseminado de Sn, Nb, Ta e F (Y, ETR, Li, U, Th) e, em sua parte central, contém um depósito de criolita maciça com 10 Mt (teor de 38% de Na3AlF6). O objetivo do trabalho foi compreender que contexto geológico permitiu a formação desta associação rocha-minério única no mundo. Para tanto, foram efetuados estudos isotópicos (Sm-Nd, Rb-Sr e Pb-Pb) e estudos tectônicos, enfocando o granito Madeira, seus correlatos e as rochas regionais. Durante uma primeira fase extensional, formaram-se as rochas vulcânicas do Grupo Iricoumé (1.890 a 1881 Ma), constituindo um complexo de caldeiras, e os corpos graníticos associados da Suíte Intrusiva Mapuera, ambos gerados a partir de fontes mantélicas. Concomitantemente aos estágios finais do vulcanismo iniciou-se a sedimentação na bacia Urupi (possivelmente um rift), acompanhada por um segundo pico de vulcanismo há 1.825 Ma. Fluidos mantélicos migraram para a zona afetada pela extensão regional, ascenderam acompanhando as isotermas e iniciaram a fenitização da crosta. Na continuidade deste processo, durante uma segunda fase extensional, rochas até refratárias tornaram-se fusíveis e originaram 5 magmas diferentes, todos com assinatura de fonte crustal e mantélica, que se posicionaram, entre 1.839 e 1.824 Ma, em estruturas geradas na fase anterior, formando os 3 corpos graníticos da Suíte Madeira. Numa terceira fase tectônica, desta feita transtensiva, fluidos mantélicos, possivelmente de natureza carbonatítica, fenitizaram rochas de nível crustal mais alto, enriquecidas em Sn, e nelas introduziram F, Nb, Y, ETR, U e Th em concentrações anômalas. Da fusão destas rochas resultou o magma do albita granito que se alojou, há 1.822 Ma, dentro do granito Madeira, mas com uma orientação N-S discordante da orientação geral NE-SW do granito Madeira e da estrutura que o aloja. / The deposit of the Madeira granite, associated with albite granite facies is a world-class deposit with disseminated ore of Sn, Nb, Ta and F (Y, REE, Li, U, Th), and its central part contains a deposit of massive cryolite with 10 Mtons (containing 38% of Na3AlF6). The objective was to understand the geological context to the formation of ore-rock association unique in the world. Therefore isotopic studies were performed (Sm-Nd, Rb-Sr and Pb-Pb) and tectonic studies focusing on the Madeira granite, its related and regional rocks. During a first extensional phase volcanic rocks of the Iricoumé Group (1890 to 1881 Ma) was originated forming a caldera complex and granitic bodies associated with Mapuera Intrusive Suite, both generated from mantle sources. At the same time the final stages of volcanism began the sedimentation in Urupi basin (possibly a rift), followed by a second peak of volcanism in 1825 Ma ago. Mantle fluids migrated to the area affected by regional extension rose following the isotherms and started the fenitization crust. Continuing this process in a second extensional phase , rocks become refractory and fuses originating 5 different magmas, all with crustal signature and mantle source, which is positioned between 1839 and 1824 Ma, in structures generated in previous phase, forming 3 granitic bodies of Madeira suite . In a third tectonic phase,, this time transtensive, mantle fluid, possibly of a carbonatitic fenitizated rocks from higher crustal level , enriched in Sn, and introduced F, Nb, Y, REE , U and Th in anomalous concentrations. The fusion of these rocks resulted in the albite granite magma that has positioned, there in 1822 Ma, within the Madeira granite, but with a NS orientation ,discordant of the general NE-SW of Madeira granite and the structure that it was contained.

La modélisation de la cryolite, utilisée dans la fabrication de l’aluminium, implique plusieurs défis, notament la présence de discontinuités dans la solution et l’inclusion de la difference de densité entre les phases solide et liquide. Pour surmonter ces défis, plusieurs éléments novateurs ont été développés dans cette thèse. En premier lieu, le problème du changement de phase, communément appelé problème de Stefan, a été résolu en deux dimensions en utilisant la méthode des éléments finis étendue. Une formulation utilisant un multiplicateur de Lagrange stable spécialement développée et une interpolation enrichie a été utilisée pour imposer la température de fusion à l’interface. La vitesse de l’interface est déterminée par le saut dans le flux de chaleur à travers l’interface et a été calculée en utilisant la solution du multiplicateur de Lagrange. En second lieu, les effets convectifs ont été inclus par la résolution des équations de Stokes dans la phase liquide en utilisant la méthode des éléments finis étendue aussi. Troisièmement, le changement de densité entre les phases solide et liquide, généralement négligé dans la littérature, a été pris en compte par l’ajout d’une condition aux limites de vitesse non nulle à l’interface solide-liquide pour respecter la conservation de la masse dans le système. Des problèmes analytiques et numériques ont été résolus pour valider les divers composants du modèle et le système d’équations couplés. Les solutions aux problèmes numériques ont été comparées aux solutions obtenues avec l’algorithme de déplacement de maillage de Comsol. Ces comparaisons démontrent que le modèle par éléments finis étendue reproduit correctement le problème de changement phase avec densités variables. / The modelling of the cryolite bath, used in the smelting of aluminum, offers multiple challenges, particularly the presence of discontinuities in the solution and a difference in density between the solid and liquid phases. To over come these challenges, several novel elements were developed in this thesis. First of all, the phase change problem, commonly named the Stefan problem, was solved in two dimensions using the extended finite element method. A specially designed Lagrange multiplier formulation, using an enriched Lagrange multiplier solution, was implemented to impose the melting temperature on the interface. The interface velocity is determined by the jump in the heat flux across the interface and was calculated using the Lagrange multiplier values. Secondly, convection was included by solving the Stokes equations in the liquid phase using the extended finite element method as well. Thirdly, the density change between solid and liquid phases, usually neglected in the literature, was taken into account by the addition of a non-zero velocity boundary condition at the solid-liquid interface to maintain mass conservation in the system. Benchmark analytical and numerical problems were solved to validated the various components of the model and the coupled system of equations. The solutions to the numerical problems were compared to the solutions obtained using Comsol’s moving mesh algorithm. Theses comparisons show that the extended finite element method correctly solves the phase change problem with non-constant densities.