La réactivité des anodes de carbone avec le CO₂ est l'une des principales préoccupations des alumineries utilisant le procédé Hall-Héroult. Une telle réactivité n'est pas souhaitable car elle augmente la consommation nette de carbone et raccourcit ainsi la durée de vie des anodes. La surconsommation d'anode est affectée par la réactivité intrinsèque de l'anode et les phénomènes de transport de masse. Différents modèles mathématiques du processus de gazéification ont été développés pour différentes géométries et techniques : La première partie de ce travail se concentre sur la gazéification d'une seule particule d'anode de carbone avec du CO₂, en utilisant un modèle de réaction-transport détaillé, basé sur la cinétique intrinsèque de la réaction et le transport des espèces gazeuses. Le modèle comprend les équations de conservation de la masse pour les composants gazeux et les particules solides de carbone, ce qui donne un ensemble d'équations différentielles partielles non linéaires, résolues à l'aide de techniques numériques. Le modèle peut prédire le taux de génération de gaz, les compositions de gaz et le taux de consommation de carbone pendant la gazéification d'une particule de carbone. Différents modèles cinétiques ont été comparés pour décrire le comportement de gazéification des particules de carbone. Il a été constaté que le modèle de pores aléatoires (RPM) fournissait la meilleure description de la réactivité des particules d'anode. Le modèle a également prédit le retrait des particules pendant le processus de gazéification. Le modèle a été validé à l'aide de résultats expérimentaux obtenus avec différentes gammes de tailles de particules. Un bon accord entre les résultats du modèle et les données expérimentales a montré que cette approche pouvait quantifier avec succès la cinétique de gazéification et la distribution du gaz au sein de la particule anodique. De plus, le modèle Langmuir-Hinshelwood (L-H) est utilisé afin de capturer l'effet d'inhibition du monoxyde de carbone sur la réaction de gazéification. Dans la deuxième partie, la simulation du processus de gazéification de l'anode avec du CO₂, en tant que lit de particules d'anode a été considérée. Le modèle numérique de la méthode des éléments discrets CFD multi-échelles (DEM) a été développé sur la base d'un concept eulérien-lagrangien. Le modèle comprend une méthode des éléments finis eulériens (FEM) pour le gaz et les particules solides, et un DEM lagrangien pour la phase particulaire, cette dernière visant à capturer l'effet de retrait des particules (mouvement des particules lors de la gazéification). Les propriétés physiques des particules, telles que la porosité et la surface spécifique, et les propriétés thermochimiques des particules, telles que la chaleur de réaction, sont finalement suivies. Les changements géométriques des particules, le transfert de chaleur et de masse, le retrait des particules et les réactions chimiques sont pris en compte lors de la gazéification de l'anode avec du CO₂. Les profils dynamiques de concentration et de température du réactif et des gaz produits ainsi que la conversion solide ont été modélisés à la fois dans les vides entre les particules et les pores à l'intérieur de chaque particule. Pour valider le modèle, des tests expérimentaux ont été réalisés à l'aide d'un lit de particules anodiques. Dans la dernière partie, une simulation d'une dalle d'anode a été réalisée. Le modèle contient la masse et les équations de transfert de chaleur pour les composants gazeux et les particules solides de carbone, ce qui donne un ensemble d'équations différentielles partielles non linéaires, résolues à l'aide de techniques numériques. Le modèle peut prédire le taux de génération de gaz, les compositions de gaz et le taux de consommation de carbone, la chute de pression et la distribution de température pendant la gazéification d'une particule de carbone. / The reactivity of carbon anodes with CO₂ is one of the main concerns in aluminum smelters using the Hall-Héroult process. Such reactivity is not desirable because it increases the net carbon consumption and thus shortens the lifetime of the anodes. Anode overconsumption is affected by anode intrinsic reactivity and mass transport phenomena. Different mathematic models of the gasification process were developed for different geometries and technics: The first part of this work focuses on the gasification of a single carbon-anode particle with CO₂, using a detailed reaction-transport model, based on the reaction intrinsic kinetics and transport of gaseous species. The model includes the mass conservation equations for the gas components and solid carbon particles, resulting in a set of nonlinear partial differential equations, being solved using numerical techniques. The model may predict the gas generation rate, the gas composition, and the carbon consumption rate during the gasification of a carbon particle. Various kinetic models were compared to describe the gasification behavior of carbon particles. It was found that the Random pore model (RPM) provided the best description of the reactivity of anode particles. The model also predicted the particle shrinkage during the gasification process. The model was validated using experimental results obtained with different particle size ranges. Good agreement between the model results and the experimental data showed that this approach could quantify with success the gasification kinetics and the gas distribution within the anode particle. In addition, the Langmuir-Hinshelwood (L-H) model is used in order to capture the inhibition effect of carbon monoxide on the gasification reaction. In the second part, the simulation of the gasification process of anode with CO₂, as an anode particle bed, was considered. Numerical multiscale CFD-discrete element method (DEM) model was developed based on an Eulerian-Lagrangian concept. The model includes an Eulerian finite element method (FEM) for the gas and solid particles, and a Lagrangian DEM for the particle phase, the latter intending to capture the particle shrinkage effect (movement of particles during gasification). The physical properties of particles, such as porosity and specific surface area, and the thermochemical properties of particles, such as the heat of reaction, are ultimately tracked. Geometric changes in particles, heat and mass transfer, particle shrinkage and chemical reactions are considered during anode gasification with CO₂. The dynamic concentration and temperature profiles of the reactant and product gases as well as the solid conversion were modeled both in the voids between the particles and the pores inside each particle. To validate the model, experimental tests were performed using a bed of anode particles. In the last part, a simulation of the anode slab was carried out. The model contains the mass, and heat transfer equations for the gas components and solid carbon particles, resulting in a set of nonlinear partial differential equations, which are solved using numerical techniques. The model can predict the gas generation rate, gas compositions, and carbon consumption rate, pressure drop, and temperature distribution during the gasification of an anode slab.

L’aluminium est produit industriellement grâce à l’électrolyse. Ce procédé a lieu dans une cuve d’électrolyse et il consiste à injecter un courant électrique pour transformer l’oxyde d’aluminium en aluminium métallique et en dioxyde de carbone. Les anodes permettent le passage du courant à travers la cuve et fournissent également le carbone nécessaire pour la réaction électrolytique. Un incident anodique se produit lorsqu’une déformation se développe sur la surface inférieure d’une anode ou lorsque l’anode est placée trop basse dans la cuve, ce qui cause un court-circuit partiel à la position de l’anode affectée. Les incidents anodiques ont un impact négatif sur l’efficacité du courant de la cuve. La détection et la correction hâtives des incidents anodiques sont donc avantageuses d’un point de vue économique. L’objectif de cette étude est de concevoir un système qui est capable de détecter les incidents anodiques en temps réel, et ce plus rapidement que la technique standard actuelle. Pour ce faire, l’Analyse en Composantes Principales a été utilisée pour entraîner des modèles prédictifs développés à partir des signaux individuels de courant d’anodes et du signal de voltage de la cuve, dans le but de classifier les anodes selon l’erreur de prédiction au carré et la statistique T2 de Hotelling. Cette stratégie semble permettre de détecter des incidents anodiques grâce aux signaux individuels de courant. Toutefois, ce n’est pas le cas avec le signal de voltage de la cuve. La surveillance de la moyenne ou l’écart-type des signaux de courant à haute fréquence à l’aide d’un modèle de régression logistique semble aussi faciliter la détection des incidents anodiques. / Aluminum metal is produced industrially in electrolysis cells, in which an electric current is used to transform aluminum oxide into metallic aluminum and carbon dioxide. Anodes are used to carry the current across the electrolysis cells and provide the carbon source necessary to drive the electrolytic reaction forward. Anodic incidents occur when an anode develops a spike or other deformation on its bottom surface or when the anode is set too low in the cell, causing the electrolysis cell to partially short circuit at the affected anode position. Anodic incidents have a deleterious effect on the cell’s current efficiency, making early detection and correction of anodic incidents economically advanta-geous. The objective of this study is to develop a real-time anodic incident detection system capableof identifying problematic anodes faster than the standard contemporary technique. Principal Component Analysis models were trained with individual anode current signals and cell voltage signals, and were subsequently used to classify anodes according to the squared prediction error and Hotelling’sT2 statistic. This strategy appears to enable anodic incident detection with individual anode current signals, but not with the cell voltage signal. Monitoring the signal mean and standard deviation of high-frequency anode current signals with a logistic regression model also appears to facilitate anod icincident detection.

La production d'aluminium par le procédé de Hall-Héroult est influencée, entre autres, par la variation des propriétés de l'anode précuite en carbone. Les propriétés de celles-ci sont plus difficiles à contrôler en raison de la dégradation de la qualité et l'augmentation de la variabilité des propriétés des particules de coke de pétrole utilisées pour fabriquer les anodes en carbone. Actuellement, les propriétés du coke sont mesurées en laboratoire, sur des petits échantillons et sur une base peu fréquente. D'où l'importance de développer une méthode rapide et non destructive qui pourra être intégrée à la chaîne de production d'anodes et qui mesurera en temps réel les propriétés des particules de coke permettant d’ajuster la formulation de la pâte d’anode, telles que la densité apparente et la porosité ouverte des particules de coke. Dans une étude précédente, un montage d'analyse acoustique a été conçu et testé pour estimer la densité et la porosité des particules de coke en utilisant la transformée de Fourier (FFT) pour la décomposition en fréquences des signaux sonores, couplée à des méthodes de régression multivariées pour l'estimation de modèles prédictifs. Cette étude vise à améliorer les résultats de prédiction par l’utilisation de la transformée en ondelettes continue (CWT) et à appliquer la méthode proposée à des mélanges de particules de coke ayant une plus grande plage de variations dans leurs propriétés. Cette méthode de décomposition temps-fréquence permet d’obtenir une signature acoustique plus spécifique des particules de coke de différentes tailles et provenant de plusieurs fournisseurs. Le potentiel de l’approche proposée semble très prometteur en raison de la bonne capacité prédictive de la densité apparente et de la porosité ouverte des particules obtenues jusqu’à présent, sur des échantillons non-mélangés et mélangés provenant de différentes fournisseurs et tamisés en plusieurs fractions de tailles granulométriques. / The production of primary aluminum using the Hall-Héroult process is influenced by the variations in pre-baked carbon anode properties. Controlling their quality is a challenge due to the raw material quality degradation and variability. Petroleum coke is one of the main raw materials used to produce the carbon anodes. Currently, coke properties are measured in laboratory by using small samples and on an infrequent basis. Therefore, it is important to develop a fast and non-destructive sensing method that can be integrated into the anode production line in order to measure in real time coke particle properties allowing to adjust the anode paste formulation, such as the apparent density and open porosity of the coke particles. In a previous study, an experimental set-up based on an acoustic analysis was developed and tested to estimate the density and the porosity of coke particles. Fourier Transform (FFT) was used for frequency decomposition of the acoustic signals, and multivariate regression methods to analyze the FFT signal features and predict coke properties. This study aims at improving the predictive performance of the model by using the Continuous Wavelet Transform (CWT) and apply the proposed technique to mixtures of coke particles having a wider range of properties. This time-frequency decomposition method allows capturing the acoustic signature of coke particles of different sizes obtained from several suppliers. So far, the approach shows promising results for predicting the apparent density and the open porosity of mixed and unmixed coke particles obtained from different suppliers and sieved into several size classes.

Des anodes de haute densité et de qualité uniforme sont d'un grand intérêt dans la production d'aluminium primaire. La variation dans les propriétés des matières premières ainsi que le grand nombre de variables est reconnue comme étant un grand défi et conduit à des anodes de qualité très variable. Dans ce projet, une d'essai de comportement au compactage de la pâte d’anode a été développée. Le comportement au compactage de la pâte d'anode a été utilisé comme indice de qualité intermédiaire pour réduire le nombre de variables. Différentes combinaisons de temps et de températures de mélange ont été utilisées pour faire de la pâte d’anode afin de comprendre les effets des variables de mélange. Les pâtes ont été compactées, cuites et caractérisés par la suite. Le mélange effectué à 178 C pendant 10 minutes a été celui permettant d’obtenir la distribution la plus homogène de coke et de brai et la densité maximale des échantillons verts et cuits. Il est admis que le comportement à la déformation de la matrice liante (particules fines de coke+brai) contribue fortement au processus de compactage. Les mélanges des matrices liantes et des pâtes d’anodes, avec différents ratios de brai et de particules de coke, ont été compactés à des taux de déformation différentes. La compaction de la matrice liante et de la pâte d'anode, avec les compositions classiques utilisées dans ce projet, n’est pas significativement dépendante de la vitesse de mise en forme. L'effet de la forme et de la densité des particules sur la densité de la pâte compactée a été étudié. La densité apparente des particules, la densité apparente (Scott) et la densité apparente vibrée du lit de particules ont été mesurées pour différentes fractions de cinq sources de coke éponge. Des pâtes d’anode ont été produites à partir des cinq cokes et ensuite compactées. Il a été suggéré que la densité apparente vibrée n'est pas un facteur suffisant pour déterminer les propriétés de compaction et la densité de la pâte compactée et aussi que les facteurs de forme et la densité des particules sont des paramètres importants qui devraient être considérés.

This thesis explores the development of efficient engineered composite alternative anodes and the use of ethanol as a fuel for Solid Oxide Fuel Cells. SOFCs can in theory operate with fuels other than hydrogen. However, this requires the design of efficient alternative anode material that do not catalyze carbon formation and that are tolerant to redox cycles. An innovative concept has been developed that relies on the impregnation of perovskites into porous YSZ structures to form the anode functional layer. Catalysts are added to provide sufficient catalytic activity. Cells with anodes containing LSCM and Ce/Pd have displayed excellent performance. At 800°C, and with a 65 μm thick electrolyte, the power outputs were above 1W/cm² in humidified hydrogen and 0.7 W/cm² in humidified methane. These electrodes have shown the ability to reduce CO₂ electrochemically with an efficiency that is similar to that which can be achieved for H₂O electrolysis and the anodes could operate on pure CO₂. The importance of incorporating an efficient catalyst was demonstrated. The use of 0.5 wt% of Pd is sufficient to dramatically improve the performance in such electrodes. The microstructure of impregnated LSCM-YSZ composites plays an important role in the high performance obtained. A layer of LSCM nanoparticles covering the YSZ is formed upon reduction, offering a great surface area for electrochemical reactions. The fabrication method presented in this thesis is a powerful tool for designing microstructures in situ. Among the various fuels under consideration for SOFCs, ethanol offers outstanding advantages. Half cell measurements have been performed to characterize the performance of different types of anodes when operated on ethanol/steam mixtures. Steady performance was achieved on LSCM-CGO anodes. Carbon deposits from gas phase reactions have been evidenced and were found to be responsible for the performance enhancement when the cell is operated in diluted ethanol as compared to hydrogen. At high steam content, polarization resistances of LSCM-CGO-YSZ anodes in ethanol/ steam mixtures were shown to be below 0.3 Ω.cm² at 950°C.

>Magister Scientiae - MSc / A comparative study of electrolysers using different anodic materials for the electrolysis of brine (sodium chloride) for the production of sodium hypochlorite as a source of A comparative study of electrolysers using different anodic materials for the electrolysis of brine (sodium chloride) for the production of sodium hypochlorite as a source of available chlorine for disinfection of rural water supplies has been undertaken. The electrolyser design used was tubular in form, having two chambers i.e. anode inside and cathode outside, separated by a tubular inorganic ceramic membrane. The anode was made of titanium rod coated with a thin layer of platinum and a further coat of metal oxide. The cathode was made of stainless steel wire. available chlorine for disinfection of rural water supplies has been undertaken. The electrolyser design used was tubular in form, having two chambers i.e. anode inside and cathode outside, separated by a tubular inorganic ceramic membrane. The anode was made of titanium rod coated with a thin layer of platinum and a further coat of metal oxide. The cathode was made of stainless steel wire. An assessment of these electrolysers was undertaken by studying the effects of some variable parameters i.e. current, voltage and sodium chloride concentration. The flow rate was kept unchanged at 50ml/h anolyte and 140ml/h catholyte since it was found to be optimum flow rate for chlorine generation. Figures of merit of the electrolysers were calculated on the basis of three sets of measurements. Analytical methods used for the determination of sodium hypochlorite concentration were iodometric and N, N-Diethyl-p- Phenylenediamine (DPD) titration methods. The DPD titration method was used to determine the chlorine concentration of less than 1mg/L, while the iodometric titration method was used to determine chlorine concentration of ImgIL and above. Sodium chlorate present in the hypochlorite solution was also determined using a spectrophotometric method. The cobalt oxide electrolyser has been shown to be superior as compared to the ruthenium dioxide and manganese dioxide electrolysers in terms of hypochlorite generation. Sodium chlorate was present but at concentration levels not hazardous for use in dosing water for drinking purposes. Analysis of hydroxyl radicals was undertaken since there were claims that these are produced during brine electrolysis. Hydroxyl radical analysis was not successful, since sodium hypochlorite and hypochlorous acid interfere using the analytical method described in this study.

Solid Oxide Fuel Cells could play a key role in energy systems of the future because they can directly convert the chemical energy of fuels into electrical energy in a reliable and energy efficient way. The choice of materials for the components of fuel cells is crucial for the achievement of the high performance and the low price necessary to establish fuel cell technology in the energy market. Current state of the art anodes consisting of nickel and yttria stabilised zirconia (Ni/YSZ) offer good electrochemical performance but suffer from limitations like carbon deposition, redox instability and sulphur poisoning. This thesis explores the properties of composite fully ceramic anodes consisting of a skeleton of yttria stabilised zirconia (YSZ) or cerium gadolinium oxide (CGO) and a perovskite phase based on B-site doped lanthanum strontium titanate. The perovskite phase was fabricated in situ inside the pores of the skeleton material by the infiltration of an aqueous precursor and subsequent firing (impregnation method). Material characterisation of the composite anodes was carried out by X-ray diffraction and the microstructure investigated by electron microscope techniques. The electrochemical performance was tested by IV-curves and impedance spectroscopy. Particularly the investigation of the connection between the microstructure of the impregnated anodes and their electrochemical performance is a main objective of this work. The electrochemical performance of cells with a CGO skeleton and an impregnated lanthanum strontium titanate phase was found to be inferior compared to cells with a YSZ skeleton, even if the ionic conductivity of CGO is known to be higher than the ionic conductivity of YSZ. The difference was assigned to mass transport problems tightly connected to the different microstructure of the composite anodes. A significant improvement of the performance could be achieved by the utilisation of A-site deficient perovskites as impregnated phase in a YSZ skeleton. Cells with composite anodes of YSZ and La₀.₄Sr₀.₄Ti₀.₉₄Mn₀.₀₆O[sub](3-δ) show power densities of 156.2 mW/cm² at a measuring temperature of 750 °C compared to 58.5 mW/cm² measured in a similar cell with A-site stoichiometric LSTM, both cells having an electrolyte thickness of around 60 μm. The superiority of the performance of anodes with A-site deficient perovskites is mainly due to a lower ohmic resistance of only 0.5 Ω*cm², indicating better conductivity of the composite with A-site deficient perovskites. The investigation of the microstructure of composite anodes with A-site deficient perovskites showed the decoration of the surface with nanoparticles after reduction. These nanoparticles originate from exsolution of ions from the B-site of the perovskite and can't be found in A-site stoichiometric perovskites. The influence of fabrication parameters like firing temperature of the skeleton, firing temperature after impregnation or vacuum impregnation on the microstructure and electrochemical performance of the composite anodes was studied. Particularly the increase of the firing temperature of the skeleton from 1400 °C to 1500 °C resulted in an impressive improvement of total cell resistance and maximal power density.

The modification of anode materials is important to enhance the power generation of MFCs (microbial fuel cells). A novel and cost-effective modified anode that is fabricated by dispersing manganese dioxide (MnO2) and HNTs (Halloysite nanotubes) on carbon cloth to improve the MFCs' power production was reported. The results show that the MnO2/HNT anodes acquire more bacteria and provide greater kinetic activity and power density compared to the unmodified anode. Among all modified anodes, 75 wt% MnO2/HNT exhibits the highest electrochemical performance. The maximum power density is 767.3 mWm(-2), which 21.6 higher than the unmodified anode (631 mW/m(2)). Besides, CE (Coulombic efficiency) was improved 20.7, indicating that more chemical energy transformed to electricity. XRD (X-Ray powder diffraction) and FTIR (Fourier transform infrared spectroscopy) are used to characterize the structure and functional groups of the anode. CV (cyclic voltammetry) scans and SEM (scanning electron microscope) images demonstrate that the measured power density is associated with the attachment of bacteria, the microorganism morphology differed between the modified and the original anode. These findings demonstrate that MnO2/FINT nanocomposites can alter the characteristics of carbon cloth anodes to effectively modify the anode for practical MFC applications. (C) 2016 Elsevier Ltd. All rights reserved.

The service life and electrocatalytic activity of tantalum oxide/iridium oxide coated titanium plate and mesh anodes used in the electrolytic production of chromic acid were investigated by performing accelerated life tests, voltammetric and chronoamperometric measurements in chrome (VI) solutions. Experimental results showed that the service life for the coated mesh anode was 1059 hours, compared to 828 hours for the plate anode at a current density of 1.2 A cm-2. In addition, the coating failed earlier in higher chromic acid concentration. Physical analysis by SEM and EDS before and after accelerated life test confirmed that the deactivation was a result of corrosion of IrO2 followed by titanium substrate passivation. A simple and rapid method for assessing the electrocatalytic activity of iridium–tantalum oxide coating based on a chronoamperometric technique was developed.

We have investigated the use of aluminum based amorphous metallic glass as the anode in lithium ion rechargeable batteries. Amorphous metallic glasses have no long-range ordered microstructure; the atoms are less closely packed compared to the crystalline alloys of the same compositions; they usually have higher ionic conductivity than crystalline materials, which make rapid lithium diffusion possible. Many metallic systems have higher theoretical capacity for lithium than graphite/carbon; in addition irreversible capacity loss can be avoided in metallic systems. With careful processing, we are able to obtain nano-crystalline phases dispersed in the amorphous metallic glass matrix. These crystalline regions may form the active centers with which lithium reacts. The surrounding matrix can respond very well to the volume changes as these nano-size regions take up lithium. A comparison study of various kinds of anode materials for lithium rechargeable batteries is carried out. / Singapore-MIT Alliance (SMA)