La thèse s’inscrit dans le cadre d’un projet de recherche visant à élaborer une anode permettant le dégagement d’oxygène en milieu de cryolithe fondue à 960°C et dont l’utilisation supprimerait les émissions de dioxyde de carbone du procédé de production d’aluminium Hall-Héroult. Le type d’anode envisagée est un composite Cermet (CERamique – METal), généralement constitué d’une phase métallique à base de nickel, cuivre et fer et d’une ou plusieurs phase(s) oxyde(s) conductrice(s) du type NixFe3-xO4 et Ni1-xFexO. Deux types de cermet, biphasé et triphasé, ont été étudiés dans le cadre de ce projet. L’objectif principal a été de comprendre les mécanismes de dégradation de telles anodes durant les électrolyses. Les cermets ont été testés sous une densité de courant de 0,80 A/cm² pour différentes durées, puis analysés par microscopie optique et microscopie électronique à balayage couplée à une analyse dispersive en énergie. Les premières électrolyses, effectuées dans un électrolyte synthétique, sans aluminium métallique et sous atmosphère argon ont révélé à la fois une dégradation chimique et électrochimique. L’étude de la dégradation chimique a été réalisée grâce à des immersions dans l’électrolyte sans polarisation. Les résultats ont mis en évidence une dissolution de la phase spinelle, liée à un phénomène de substitution entre les ions Al3+ de l’électrolyte et Fe3+ du spinelle, entrainant la formation d’un aluminate du type NixFe3-x-yAlyO4. Lorsque x = 0 et y proche de 2, la phase spinelle est dissoute alors que lorsque x est proche de 1, la phase spinelle est stabilisée et sa dissolution ralentie. Les électrolyses menées jusqu’à la fin de vie des matériaux ont révélé l'attaque préférentielle de la phase métallique pour former des composés fluorés et oxydes. Grâce aux caractérisations micrographiques et aux calculs thermodynamiques (logiciel FactSage 7.1), un mécanisme de dégradation global des anodes a été proposé. La compréhension des mécanismes de dégradation chimique et électrochimique des matériaux a permis de proposer des voies d'amélioration concernant à la fois le matériau d'anode et la chimie du bain cryolithique. / The thesis is a part of a research project which develop an anode permitting the release of oxygen in molten cryolite medium at 960°C, in order to eliminate carbon dioxide emissions from the aluminium production process called Hall-Héroult process. The type of anode envisaged is a Cermet composite (CERamic - METal), generally consisting of a metallic phase based on nickel, copper and iron and one or more conductor oxide phase(s) such as NixFe3-xO4 and Ni1-xFexO type. Two types of cermet, two-phase and three-phase, have been studied in this project. The main objective was to understand the degradation mechanisms of such anodes during electrolysis. The cermets were tested at a current density of 0.80 A/cm² for different durations and analyzed by optical microscopy and scanning electron microscopy coupled with dispersive energy analysis. The first electrolyses, carried out in a synthetic electrolyte, without metallic aluminum and under argon atmosphere revealed both chemical and electrochemical degradations. Study of chemical degradation was carried out thanks to immersions in the electrolyte without polarization. The results showed a dissolution of the spinel phase, linked to a substitution phenomenon between the Al3+ ions from the electrolyte and Fe3+ from spinel, resulting in the formation of a NixFe3-x-yAlyO4 aluminate type. When x = 0 and y close to 2, the spinel phase is dissolved whereas when x is close to 1, the spinel phase is stabilized and its dissolution slowed down. Electrolyses conducted to the end of life of the materials revealed the preferential attack of the metal phase to form fluorinated compounds and oxides. Thanks to micrographic characterizations and thermodynamic calculations (FactSage 7.1 software), a global degradation mechanism of the anodes has been proposed. The understanding of the chemical and electrochemical degradation mechanisms of the materials made it possible to propose ways of improvement concerning both the anode material and the cryolithic bath chemistry.

Though economically favorable when compared to other renewable energy storage technologies, thermal energy storage systems for concentrating solar thermal power (CSP) plants require additional cost reduction measures to help transition CSP plants to the point of grid-parity. Thermocline packed bed storage is regarded as one potential low cost solution due to the single tank requirement and low cost storage media. Thus sensible heat storage (SHS) and latent heat storage (LHS) packed bed systems, which are two thermocline varieties, are frequently investigated. LHS systems can be further classified as single phase change material (PCM) systems or cascaded systems wherein multiple PCMs are employed. This study compared the performance of SHS, single PCM, and cascaded PCM direct storage systems under the conditions that may be encountered in utility-scale molten salt CSP plants operating between 565°C and 288°C. A small-scale prototype SHS packed bed system was constructed and operated for use in validating a numerical model. The drawbacks of the latent heat storage process were discussed, and cascaded systems were investigated for their potential in mitigating the issues associated with adopting a single PCM. Several cascaded PCM configurations were evaluated. The study finds that the volume fraction of each PCM and the arrangement of latent heat in a 2-PCM and a 3-PCM system influences the output of the system, both in terms of quality and quantity of energy. In addition to studying systems of hypothetical PCMs, real salt PCM systems were examined and their selection process was discussed. A preliminary economic assessment was conducted to compare the cost of SHS, single-PCM LHS, cascaded LHS, and state-of-the-art 2-tank systems. To the author's knowledge, this is the first study that compares the cost of all three thermocline packed bed systems with the 2-tank design. The SHS system is significantly lower in cost than the remaining systems, however the LHS system does show some economic benefit over the 2-tank design. If LHS systems are to be viable in the future, low cost storage media and encapsulation techniques are necessary.

La toxine diphtérique (DT) une toxine bactérienne sécrétée par Corynebacterium diphtheriae. Lors de l'intoxication d'une cellule, le domaine de translocation (T) de la DT s'insère dans la membrane à pH acide et assiste la translocation du domaine catalytique (C) dans le cytosol. Le domaine T adopte une conformation en molten globule (MG) et devient compétent pour l'interaction membranaire. Nous avons identifié par mutagenèse les résidus du domaine T dont la protonation favorise la formation de l'état MG en solution et l'interaction membranaire. Les résultats montrent que la protonation concertée des six histidines du domaine T est impliquée dans la formation de l'état MG en solution. La paire His223-257 et l'His251 ont un effet prépondérant dans la formation de l'état molten globule en solution, alors que la paire His322-323 (mais également His251) est davantage impliquée dans l'interaction avec la membrane, et en particulier la liaison à la membrane.<br />Nous avons étudié les changements de conformation des deux domaines C et T dans une protéine CT correspondant à la DT tronquée de son domaine R, afin de comprendre les effets du lien covalent sur les conformations respectives de C et T en fonction du pH. Les mutants CTW50/153F et CTW206/281F ont permis de suivre la fluorescence des Trp de chaque domaine dans la protéine CT. Les<br />résultats montrent que le domaine T dirige la translocation de C dans les premières étapes de la translocation (formation de l'état MG, liaison et insertion membranaire), et qu'il pourrait jouer un rôle de chaperon en stabilisant l'état MG du domaine C.

A high temperature molten salt test loop that utilizes FLiNaK (LiF-NaF-KF) at 700ºC has been proposed by Oak Ridge National Laboratory (ORNL) to study molten salt flow characteristics through a pebble bed for applications in high temperature thermal systems, in particular the Pebble Bed – Advanced High Temperature Reactor (PB-AHTR). The University of Tennessee Nuclear Engineering Department has been tasked with developing and testing pressure instrumentation for direct measurements inside the high temperature environment. A nickel diaphragm based direct contact pressure sensor is developed for use in the salt. Capacitive and interferometric methods are used to infer the displacement of the diaphragm. Two sets of performance data were collected at high temperatures. The fiber optic, Fabry-Perot interferometric sensor was tested in a molten salt bath. The capacitive pressure sensor was tested at high temperatures in a furnace under argon cover gas.

The corrosion of current collector materials in MoltenCarbonate Fuel Cells (MCFC) is investigated. The essential aimsof this investigation were to study the corrosion behaviour ofdifferent materials, in varying cathode and anode MCFCenvironments, and to study the contact corrosion resistancesbetween the MCFC current collector and electrodes. For thesepurposes, pure iron, iron-chromium binary alloys and severalcommercial steels were investigated in molten carbonate meltswithin the pot-cell laboratory set-up. In addition, the contactcorrosion resistances, between an AISI 310 current collectorand two cathodes (NiO and LiCoO2), were studied in a laboratory fuel cell.Post-tests were done to study the corrosion products formed atthe surfaces. In cathode environments, corrosion potential increased overtime as a protective corrosion layer slowly formed. Eventually,the potential reached a stable value close to the cathodeoperating potential. The main cathode reaction, as corrosionpotential increased, changed from water reduction to oxygenreduction. Corrosion rate under the operating cathode conditiondepended on the chromium content; the higher the concentrationof chromium, the lower the corrosion rate. The corrosion ratesof ferritic steels, with high chromium content, and AISI 310were higher at the so-called outlet operating condition incomparison to the standard and so-called inlet conditions. Thecorrosion rate was higher at the beginning of the exposure,which resulted in a relatively fast corrosion layer growth thatslowed as the protective layer was formed. It was shown thatthe corrosion layers, formed on iron-chromium alloys, AISI 310and ferritic high chromium-containing steels, consisted of twolayers. The outer layer was porous and iron rich, while theinner layer was quite compact and rich in chromium and/oraluminiumTherefore, the corrosion behaviour was dependent onthe corrosion layer structure at the metal surface. In anode environments, the beneficial behaviour of aluminiumin ferritic alloys, with high aluminium contents, was due tothe formation of aluminium oxide and/or lithium aluminium oxideat the surface. The corrosion rates at the standard and outletconditions were of the same order of magnitude, while thecorrosion rates at the inlet conditions were considerablyhigher. The lower temperatures and higher carbon dioxideconcentrations in the inlet conditions appeared to result in asurface layer deficient in aluminium. A modified theoreticalmodel was developed to evaluate the corrosion current densitiesfrom experimental polarisation curves or linear polarisationresistance measurements in anode environments. The fittingswere found to be very good. An experimental method was developed forin-situmeasurements of the contributions to the totalohmic losses at the cathode in a laboratory scale MCFC. Thecontact resistance between the cathode and current collectorcontributed quite a large value to the total cathodepolarization. The corrosion layer, formed between the LiCoO2cathode and AISI 310 current collector, wasiron-rich and more porous, and contained a small amount ofcobalt. This was deemed to consist of a two-phase oxide, whichresulted in a lower conductivity. The corrosion layer, formedbetween the NiO cathode and AISI 310 current collector, wasrich in nickel. The corrosion layers on the AISI 310, incontact with the cathode, had a different composition comparedto samples immersed in carbonate melts. <b>Key words</b>: molten carbonate fuel cell (MCFC), corrosion,current collector, contact corrosion resistance.

The objective of this study is to synthesize nanomaterials by mixing molten salt (alkali nitrate salt eutectics) with inorganic nanoparticles. The thermo-physical properties of the synthesized nanomaterials were characterized experimentally. Experimental results allude to the existence of a distinct compressed phase even for the solid phase (i.e., in the nanocomposite samples). For example, the specific heat capacity of the nanocomposites was observed to be enhanced after melting and re-solidification - immediately after their synthesis; than those of the nanocomposites that were not subjected to melting and re-solidification. This shows that melting and re-solidification induced molecular reordering (i.e., formation of a compressed phase on the nanoparticle surface) even in the solid phase - leading to enhancement in the specific heat capacity. Numerical models (using analytical and computational approaches) were developed to simulate the fundamental transport mechanisms and the energy storage mechanisms responsible for the observed enhancements in the thermo-physical properties. In this study, a simple analytical model was proposed for predicting the specific heat capacity of nanoparticle suspensions in a solvent. The model explores the effect of the compressed phase – that is induced from the solvent molecules - at the interface with individual nanoparticles in the mixture. The results from the numerical simulations indicate that depending on the properties and morphology of the compressed phase – it can cause significant enhancement in the specific heat capacity of nanofluids and nanocomposites. The interfacial thermal resistance (also known as Kapitza resistance, or “Rk”) between a nanoparticle and the surrounding solvent molecules (for these molten salt based nanomaterials) is estimated using Molecular Dynamics (MD) simulations. This exercise is relevant for the design optimization of nanomaterials (nanoparticle size, shape, material, concentration, etc.). The design trade-off is between maximizing the thermal conductivity of the nanomaterial (which typically occurs for nanoparticle size varying between ~ 20-30nm) and maximizing the specific heat capacity (which typically occurs for nanoparticle size less than 5nm), while simultaneously minimizing the viscosity of the nanofluid. The specific heat capacity of nitrate salt-based nanomaterials was measured both for the nanocomposites (solid phase) and nanofluids (liquid phase). The neat salt sample was composed of a mixture of KNO3: NaNO3 (60:40 molar ratio). The enhancement of specific heat capacity of the nanomaterials obtained from the salt samples was found to be very sensitive to minor variations in the synthesis protocol. The measurements for the variation of the specific heat capacity with the mass concentration of nanoparticles were compared to the predictions from the analytical model. Materials characterization was performed using electron microscopy techniques (SEM and TEM). The rheological behavior of nanofluids can be non-Newtonian (e.g., shear thinning) even at very low mass concentrations of nanoparticles, while (in contrast) the pure undoped (neat) molten salt may be a Newtonian fluid. Such viscosity enhancements and change in rheological properties of nanofluids can be detrimental to the operational efficiencies for thermal management as well as energy storage applications (which can effectively lead to higher costs for energy conversion). Hence, the rheological behavior of the nanofluid samples was measured experimentally and compared to that of the neat solvent (pure molten salt eutectic). The viscosity measurements were performed for the nitrate based molten salt samples as a function of temperature, shear rate and the mass concentration of the nanoparticles. The experimental measurements for the rheological behavior were compared with analytical models proposed in the literature. The results from the analytical and computational investigations as well as the experimental measurements performed in this proposed study – were used to formulate the design rules for maximizing the enhancement in the thermo-physical properties (particularly the specific heat capacity) of various molten salt based inorganic nanomaterials. The results from these studies are summarized and the future directions are identified as a conclusion from this study.

The thesis is focussed on the performance of the fuel cell and the design of the cell for operation with natural gas and renewable fuels, e.g. biogas or gasified biomass. The performance is one of the important issues for the development and commercialisation of fuel cell stacks. In order to operate fuel cell on renewable fuels, without preceding reforming of the fuel, a high temperature fuel cell is needed, i.e. a solid oxide fuel cell (SOFC) or a molten carbonate fuel cell (MCFC). At present, the latter fuel cell type is much more mature when regarding the technical aspects than is the solid oxide fuel cell. The German company MTU has up to date installed about thirty MCFC plants, mainly in Europe and the USA but also in Japan. Moreover the European Commission has decided that the use of renewable fuels must increase at the expense of fossil fuels. This decision is one step towards a smaller dependence on fossil energy sources and limited emissions of greenhouse gases. The objective of this work is to better understand the factors that influence the cell performance: to determine the kinetic parameters of the hydrogen oxidation and the carbon monoxide oxidation and to get more information about the reaction mechanism, even when dealing with gases of low hydrogen content. The latter is of special importance when operating the cells on biogas or gasified biomass. These fuels also contain higher concentrations of carbon monoxide and carbon dioxide. It was found that the hydrogen mechanism proposed by Jewulski and Suski describes the anode performance even at lower concentrations of hydrogen, i.e. gases corresponding to gasified biomass. Furthermore, the carbon monoxide reaction will only slightly influence the anode performance but if the rate of the shift reaction is small the influence of direct oxidation of carbon monoxide will increase. Experimental investigations have shown that mass transfer limitations in the gas phase exist. By mathematical modelling it was found that the current collector has a larger affect on the concentration gradients than the porous electrode. The concentrations gradients in the current collector are caused by the shift reaction that mainly takes place at the electrode. However, if the gas corresponds to equilibrium at the current collector the profiles will become almost uniform. Furthermore the influence of wetting properties, the pore structure and pore size distribution have also been investigated in this thesis. The outcome of this thesis may be used for electrode development and design, as well as for input to reliable cell and stack models for system simulations. / QC 20101008

Molten carbonate fuel cells (MCFCs) have become an attractive emerging technology for stationary co-generation of heat and power. From a technical perspective, dynamic operation has a significant effect on the fuel cell life cycle and, hence, economic viability of the device. The scope of this thesis is to present an improved understanding of the system behaviour at transient operation that can be used to design a more robust control system in order to overcome the cost and the operating lifetime issues. Hence, a comprehensive multi-component multidimensional transient mathematical model is developed based on the conservation laws of mass, momentum, species, energy and electric charges coupled through the reaction kinetics. In essence, this model is a set of partial differential equations that are discretized and solved using the finite-volume based commercial software, ANSYS FLUENT 12.0.1. The model is validated with two sets of experimental results, available in open literature, and good agreements are obtained. The validated model is further engaged in an extensive study. First, the MCFC behaviour at high current densities or oxidant utilization, when the mass transfer becomes dominant, is investigated using peroxide and superoxide reaction mechanisms. In brief, both mechanisms predicted the linear region of the polarization curve accurately. However, none of these mechanisms showed a downward bent in the polarization curve. A positive exponent for the carbon-dioxide mole fraction is probably essential in obtaining the downward bent (“knee”) at high current densities which is in contrast to what has been reported in the literature to date. Next, a sinusoidal impedance approach is used to examine the dynamic response of the unit cell to inlet perturbations at various impedance frequencies. This analysis is further used to determine the phase shifts and time scales of the major dynamic processes within the fuel cell. Furthermore, numerical simulation is utilized in order to investigate the underlying electrochemical and transport phenomena without performing costly experiments. Results showed that the electrochemical reactions and the charge transport process occur under a millisecond. The mass transport process showed a comparatively larger time scale. The energy transport process is the slowest process in the cell and takes about an hour to reach its steady state condition. Furthermore, the developed mathematical model is utilized as a predictive tool to provide a three-dimensional demonstration of the transient physical and chemical processes at system startiv up. The local distribution of field variables and quantities are presented. The results show that increasing the electrode thickness provides a higher reaction rate, but may lead to larger ohmic loss which is not desirable. The reversible heat generation and consumption mechanisms of the cathode and anode are dominant in the first 10 s while the heat conduction from the solid materials to the gas phase is not considerable. The activation and ohmic heating have the same impact within the anode and cathode because of their similar electric conductivity and voltage loss. Increasing the thermal conductivity of the cathode material will facilitate the process of heat transport throughout the cell. This can also be accomplished by lowering the effects of heat conduction by means of a cathode material with a smaller thickness. In addition, a thermodynamic model is utilized to examine energy efficiency, exergy efficiency and entropy generation of a MCFC. By changing the operating temperature from 883 K to 963 K, the energy efficiency of the unit cell varies from 42.8 % to 50.5 % while the exergy efficiency remains in the range of 26.8% to 36.3%. Both efficiencies initially rise at lower current densities up to the point that they attain their maximum values and ultimately decrease with the increase of current density. With the increase of pressure, both energy and exergy efficiencies of the cell increase. An increase in this anode/cathode flow ratio lessens the energy and exergy efficiencies of the unit cell. Higher operating pressure and temperature decrease the unit cell entropy generation. / UOIT

In Hinblick auf den stetig steigenden Bedarf der chemischen Industrie an Grundstoffchemikalien und der teilweise unsicheren Versorgung mit Erdöl und Erdgas ist es notwendig alternative Rohstoffe und Verwertungspfade für die Bereitstellung von Basischemikalien zu finden. Ziel der vorliegenden Arbeit war die Untersuchung der Hydrogenolyse von Lignin in niedermolekulare Produkte unter Verwendung geeigneter Salzschmelzen. Es konnte gezeigt werden, dass Lignin in Zinkchlorid/Kaliumchloridschmelzen in niedermolekulare Produkte abgebaut werden kann. Hierbei erwiesen sich der Einsatz eines entsprechenden Eutektikums und einer Alternativschmelze mit niedrigem Schmelzpunkt als hilfreich. Durch den Einsatz verschiedener Apparaturen wurden Untersuchungen in statischer und dynamischer Atmosphäre durchgeführt. Es ergaben sich während der Untersuchung Abhängigkeiten der Hydrogenolyse von verschiedenen Reaktionsparametern. Optima der Umsetzung hinsichtlich der Reaktionsparameter Temperatur, Zeit und Ligninanteil in der Schmelze wurden herausgearbeitet. Die Ausbeute an gewünschten Flüssigprodukten wurde, im Untersuchungsbereich, an diesen Punkten maximiert. Gleichzeitig war die Rückstands- und Gasbildung eingeschränkt. Es konnten Erkenntnisse eines komplexen Systems der Abhängigkeiten der Ausbeuten an Reaktionsprodukten von den Parametern der Untersuchung gewonnen werden. Die Hydrogenolyse von Lignin führte zur Aromatisierung fester Residuen sowie zur Abreicherung von Sauerstofffunktionalitäten. Komplexe Reaktionsmechanismen bewirkten den Abbau von Methoxyl-, Carboxyl- und Hydroxylgruppen der Ligninstruktur. Carbeniumionmechanismen konnten als wichtige Reaktionen zur Spaltung von Ether Arylbindungen identifiziert werden. Die Freisetzung von Monomeren und die Polymerisation anderer Intermediäre sind durch Sekundärreaktionen denkbar. Die gebildeten Flüssigprodukte bestanden hauptsächlich aus Monoaromaten (v.a. Guajakole und Kresole) und wenigen Polyaromaten. Die Selektivität der Bildung einzelner Verbindungen war gering, d.h. die Flüssigprodukte sind eine heterogene Mischung mit geringen Konzentrationen der Einzelsubstanzen. Die Unterschiede in der Struktur der Ausgangslignine bildeten sich auch in der Zusammensetzung der Flüssigprodukte ab. / In view of the steadily increasing demand of the chemical industry to base chemicals and the partial uncertain supply of crude oil and gas, it is necessary to find alternative raw materials and conversion routes for the provision of basic chemicals. The aim of the present work was to investigate the hydrogenolysis of lignin in low molecular weight products using appropriate molten salt media. It could be demonstrated that lignin can be convert in low molecular weight products using zinc chloride/potassium chloride molten salt media. The use of an appropriate eutectic melt and of an alternative melt with low melting point proved helpful. By the use of different apparatus investigations in static and dynamic atmosphere could be carried out. During the investigation dependencies of the hydrogenolysis of various reaction parameters are submitted. Optima of the hydrogenolysis regarding to reaction temperature, time and lignin content at the melt could be identified. The yields were maximized at these points in the range of investigation. Concurrently formation of gases and residues were suppressed. Some evidence of a complex system of the dependencies of the yields of reaction products are obtained from the parameters of the investigation. Hydrogenolysis of lignin leads to aromatic solid residues and to a loss of oxygen containing structures. Structures containing methoxyl-, carboxyl- and hydroxyl groups are degraded by various complex reaction mechanisms. Mechanisms of the formation of carbonium ions were identified as important reactions of the cleavage of ether aryl bonds. Secondary reactions caused the liberation of monomers and polymerisation of some intermediaries. The resulting liquid products consist mainly of monoaromatics (guaiacols and cresols) and less of polyaromatics. The selectivity of the formation of single compounds was low, i.e. the liquid products constitute a heterogenous mixture with low concentrations of the single compounds. The structural differencies of the feedstock lignins also showed at the composition of the liquid products.

The Department of Energy funded a project to determine if the specific heat of thermal energy storage materials could be improved by adding nanoparticles. The standard thermal energy storage materials are molten salts. The chosen molten salt was a sodium nitrate and potassium nitrate eutectic, commercially called Hitec Solar Salt. Two nanoparticle types were chosen, alumina and silica. The nanoparticle composite materials were fabricated by mixing the components in an aqueous solution, mixing that solution for a set amount of time using a sonic mixer, then removing the water from the aqueous solution, leaving the composite molten salt behind as a fine white powder. The thermal properties of the composite and plain material were measured using two techniques: American Society for Testing and Materials (ASTM) 1269E and Modulating Differential Scanning Calorimetry (MDSC). These two techniques measured the specific heat and the heat of fusion of the plain and composite materials. The results of all the ASTM and MDSC measurements suggest that the addition of the nanoparticles using the given manufacturing technique increased the specific heat of the molten salt by approximately 20 percent, with both measurement techniques showing approximately the same level of increase. The silica and the alumina improved the specific heat by nearly the same amount over the base material. The heat of fusion did not seem to be significantly altered compared to the observed heat of fusion value of the unmodified material. It was also observed that the nitrate and silica composite material's specific heat decreased if the material was raised to a temperature above 400C. The specific heat was observed to decrease over time, even when the temperature was well below 400C. It is unknown why this occurred. The nitrate plus alumina composite and the plain nitrate were stable to a temperature of 450C for the test duration.