Molten salt reactors (MSRs) using chloride-based salt-matrixes as coolants or fuels are a promising option for advanced nuclear reactors, but the extreme temperatures and corrosivity of molten salts pose a challenge for implementation. Molten MgCl2-NaCl-KCl is a viable candidate for MSRs that is considered in this work. Thermochemical properties are derived from electrochemical tests that aid in characterizing the properties of salts. To study these properties, some work has proposed using a three-electrode system with a reference electrode housed in a ceramic membrane. This research aims to develop a stable high-temperature reference electrode using a ceramic membrane that is then applied to develop an on-line monitoring system of Ni-Cr alloy corrosion in chloride salt. A mullite tube used as the membrane of a Ni(II)/Ni reference electrode in molten MgCl2-NaCl-KCl is studied. The performance of two different membrane thicknesses (1.325mm and 0.255mm) was studied in temperature ranges from 635oC to 835oC and data collected on the calculated formal potential of the Ni(II)/Ni system. Tests indicated that the results were stable and repeatable, and the formal potential for both systems differed from the previous experimental data by 0.12V at most, indicating that the system can be applied as an effective reference electrode. Using the reference electrode, on-line monitoring the corrosion of Ni-15wt.%Cr, Ni-20wt.%Cr, and Ni-30wt.%Cr was studied for 120 hours in MgCl2-NaCl-KCl. The on-line measurements showed the concentration changes of dissolved Cr and Ni by corrosion in the bulk molten salt. This work confirms that Ni(II)/Ni reference electrodes with a mullite tube membrane are stable and effective in molten chloride salt systems, particularly MgCl2-NaCl-KCl. The mullite membrane prepared by the manufacturer may be used directly for electrochemical applications without polishing, simplifying the reference electrode manufacturing process, and making it easier to replicate. The use of a Ni(II)/Ni reference electrode provides an avenue to study a different range of salt systems than previous reference electrodes allowed, particularly alloys in chloride salts at high temperatures. This work also confirms that the mullite tube may be used to perform on-line analysis of alloy corrosion in high temperature molten chloride salts. The study of Ni-Cr alloys in chloride salts better prepares the nuclear industry to select coolant salts and alloy containers with the best set of thermochemical and corrosion resistant characteristics for MSRs. / Master of Science / The United States receives approximately 18% of its energy from nuclear technology. Many of the reactors supplying this energy are at the end of their lifecycle and the decommissioning of some of these plants has already begun. In order to replace this older generation of nuclear reactors, a safer and cheaper option has been suggested: Molten Salt Reactors. Molten salt reactors (MSRs) using high temperature salts as a fuel or coolant are a promising option, but the extreme conditions of molten salts pose a challenge for construction and use of MSRs. Molten MgCl2-NaCl-KCl is a salt being considered for MSR application, and is considered in this work. Properties of the salts considered for MSRs are being studied diligently before implementation of these reactors. Electrochemical tests are used to study and monitor these properties. These electrochemical tests use a three-electrode system with a reference electrode housed in a membrane. In this work, a mullite tube is used as a ceramic membrane for a reference electrode in molten MgCl2-NaCl-KCl. The performance of two different membrane thicknesses (1.325mm and 0.255mm) was studied in temperature ranges from 635oC to 835oC. Results indicate that the system is an effective reference electrode. Using this innovative reference electrode, a method of monitoring on-line corrosion of Ni-15wt.%Cr, Ni-20wt.%Cr, and Ni-30wt.%Cr alloys was studied for 120-hour time periods during exposure to MgCl2-NaCl-KCl. This work confirms that reference electrodes with a mullite membrane may be used for electrochemical applications when studying molten chloride salts. The use of a Ni(II)/Ni reference electrode with a mullite membrane provides an avenue to study a different range of salt systems than previous reference electrodes and ceramics allowed, particularly chloride salts. Additionally, this mullite membrane Ni(II)/Ni reference electrode system may be used for monitoring on-line corrosion of Ni-Cr alloys in chloride salt systems.

No description available.

Il est important pour l’industrie nucléaire de mener des études amont pour mieux appréhender les phénomènes en lien avec la corrosion des alliages chromino-formeurs. Cette étude a pour objectif de mieux comprendre les paramètres influant sur les mécanismes et la cinétique de croissance des couches de chromine formées par oxydation à haute température. Ce travail s’articule en deux parties complémentaires : une partie de modélisation qui est focalisée sur les aspects cinétiques, et une partie expérimentale axée sur la compréhension des mécanismes de croissance de la chromine. La partie modélisation a pour objectif de prendre en compte l’effet des courts-circuits de diffusion dans la description des cinétiques de croissance de la chromine. Le but est d’évaluer quantitativement comment une diffusion mixte par le volume et les joints de grains peut influencer la cinétique de croissance d’une couche d’oxyde dont la microstructure évolue avec le temps. Ce travail a été réalisé à l’aide de modèles analytiques et grâce à de nouveaux développements d’un modèle numérique existant (EKINOX). Plusieurs lois de croissance de grains ont été étudiées (croissance parabolique, croissance cubique et prise en compte de l’existence d’un gradient de taille de grains dans la couche d’oxyde). Une étude paramétrique a été réalisée pour étudier l’influence de la vitesse de croissance des grains sur la cinétique de croissance de la couche de chromine. Ces simulations ont permis d’évaluer la déviation de lacinétique d’oxydation par rapport à la loi parabolique simple suivant la vitesse de croissance des grains d’oxyde, choisie comme donnée d’entrée du modèle. Ces calculs permettent également d’évaluer les ordres de grandeurs de la durée du régime transitoire durant lequel l’évolution de la taille de grains de l’oxyde influence la cinétique d’oxydation de façon significative. Les cinétiques d’oxydation calculées pour les différents cas étudiés ont été extrapolées à l’aide des méthodes classiquement utilisées dans le domaine de l’oxydation à haute température : la méthode « log-log» et la méthode dite du « kp local ». La comparaison des extrapolations et des solutions exactes permet de proposer des « bonnes pratiques » à adopter pour interpréter les cinétiques d’oxydation expérimentales. Dans la partie expérimentale de ce travail, des essais d’oxydation d’un alliage modèle Ni-30Cr à 700°C sous atmosphère d’argon technique (10-5 atm de O2) ont été réalisés en thermobalance pour des durées allant de 30 minutes à 50 heures. La couche de chromine formée à l’issue de ces essais a été finement caractérisée afin de déterminer le mécanisme d’oxydation ainsi que l’étape cinétiquement limitante. Dans les conditions de l’étude, la couche d’oxyde formée a été identifiée comme étant de la chromine de structure corindon dont les grains adoptent une orientation préférentielle de croissance <0001>. La cinétique d’oxydation suit une loi parabolique au terme d’un régime transitoire de 3 h environ. La valeur de la constante cinétique parabolique kp a été évaluée pendant le régime stationnaire et est égale à 10-15 cm2 s-1. Des expériences d’oxydation séquencée et de marqueurs ont également été réalisées. Elles ont permis de mettre en évidence une croissance anionique de la couche de chromine. La croissance de l’oxyde apparait, selon les essais de marquage isotopique dans 18O2, contrôlée par la diffusion de l’oxygène par les joints de grains et le volume de l’oxyde. Les analyses complémentaires des couches d’oxyde, par photoélectrochimie, suggèrent que le défaut majoritaire permettant cette diffusion est la lacune d’oxygène. Enfin, la morphologie de la couche d’oxyde indique une croissance plus rapide de la couche au niveau des joints de grains de métal / This thesis aims to better understand the kinetics and mechanisms of growth of chromia scales. This work is articulated in two complementary parts, on the one hand a modeling part which studies the kinetic aspects, and on the other hand an experimental part centered on the characterization of chromia growth understanding of the mechanisms. The purpose of the modeling part is to take into account the effect of diffusion short circuits in chromia the growth kinetics. The aim is to understand how a dual diffusion by both volume and grain boundaries can impact oxidation kinetics. This approach helps to define "good practices" for experimental oxidation kinetics interpretation. This modeling work was done using analytical models EKINOX numerical model. Several grain growth laws have been studied (parabolic growth, cubic growth, and grain size gradient). In addition, a parametric study was carried out to study the influence of grain growth rate on the growth kinetics of chromia. Calculations of oxidation kinetics for the various cases treated made it possible to conclude that the taking into account diffusion short circuits and their proportion evolution in time generates a deviation of the oxidation kinetics from a simple parabolic law. The rate of oxidation kinetics calculated varies with the growth rate of the oxide grains chosen as input data of the modelling. When kinetics are treated using a kinetic law involving a growth rate that changes over time, it is possible to determine the time range for which the evolution of oxide grain size significantly influences the oxidation kinetics. The oxidation kinetics modeled for the different cases were extrapolated using the methods conventionally used in high temperature oxidation, the log-log method and the local kp method. To extrapolate experimental kinetics, it is recommended to use the local kp method beforehand in order to know if the steady state has been reached. For the second experimental part, oxidation tests of the model alloy Ni-30Cr at 700°C in technical argon atmosphere were carried out. The chromia layer formed at the end of these tests was finely characterized in order to determine the oxidation mechanism as well as the kinetically limiting step. Under the conditions of the study, the nature of the oxide layer was identified as chromia. The oxidation kinetics follows a parabolic law after a transient stage that lasts about 3 hours. The stationary kp value has been determined. The mechanism of chromia growth has been identified as the diffusion of oxygen by the grain boundaries and the volume of the oxide. The morphology of the oxide layer indicates a faster oxidation over the grain boundaries of the metal. Finally, the chromine grains adopt a preferential orientation in the <0001> direction of the corundum structure. Chromium depletion profiles have been determined in the metal substrate and indicate a more pronounced chromium depletion at the grain boundaries of the metal. The type of semiconduction was determined in chromine and reveals an n-type semiconduction for an oxidation time of 7 h and an n-type and insulating semiconduction for an oxidation time of 50 h.