Actuellement, il y a une baisse du prix des ressources énergétiques distribuées, en particulier l'énergie solaire photovoltaïque, conduisant à la croissance significative de leur capacité d'installation dans de nombreux pays. D'autre part, les politiques encourageant l'efficacité énergétique ont favorisé le développement de charges DC dans les zones domestiques, telles que l'éclairage LED, les ordinateurs,, les téléphones, les téléviseurs, les moteurs DC efficaces et les véhicules électriques. Grace à ce changement, le système de distribution de microgrid DC devient plus attractive que le système de distribution à courant alternatif traditionnel. Les avantages principaux du microgrid DC sont l'efficacité énergétique plus élevée, plus facile à intégrer avec les sources d'énergie distribuées et le système de stockage. Alors que de nombreuses recherches se concentrent sur les stratégies de contrôle et la gestion de l'énergie dans le microgrid DC, sa protection reçoit une attention insuffisante et un manque de réglementation et d'expériences. La protection dans les réseaux DC est plus difficile que dans le réseau AC en raison de l'arc continu, de la valeur plus élevée du courant de courtcircuit et du taux de défaut de montée. En outre, dans les réseaux distribués à courant continu sont composés de nombreux dispositifs de commutation électroniques et semi-conducteurs, qui ne supportent le courant de défaut que quelques dizaines de microsecondes. Les disjoncteurs mécaniques, qui ont un temps de réponse de quelques dizaines de millisecondes, ne semblent pas satisfaire aux exigences de sécurité du microréseau à courant continu. L'absence d'un dispositif de protection efficace constitue un obstacle au développement du microgrid DC dans le système distribué. Cette thèse propose un disjoncteur DC auto-alimenté à courant continu utilisant normalement JFET SiC, qui offre un excellent dispositif de protection pour les microgrids DC grâce à son temps de réponse rapide et ses faibles pertes à l'état passant. La conception du disjoncteur DC à semi-conducteurs vise à répondre à deux objectifs: temps de réponse rapide et fiabilité. Les spécifications conçues et les énergies critiques qui entraînent la destruction du disjoncteur sont identifiées sur la base des résultats mesurés d'un JFET populaire dans le commerce. Un pilote de protection très rapide et fiable basé sur une topologie à convertisseur flyback avant est utilisé pour générer une tension négative suffisante pour tourner et maintenir le JFET SiC. Le convertisseur sera activé chaque fois que le disjoncteur détecte des défauts de court-circuit en détectant la tension de drain-source de JFET et crée une tension négative s'applique à la porte de JFET. Pour éviter une défaillance de la porte par surtension au niveau de la grille du JFET, la tension de sortie du convertisseur de retour vers l'avant est régulée à l'aide de la mesure coté primaire. Les résultats expérimentaux sur le prototype du disjoncteur DC ont validé les principes de fonctionnement proposés et ont confirmé que le disjoncteur DC à semi-conducteurs proposé peut interrompre le défaut en 3 μs. D'un autre côté, un modèle du JFET normalement activé dans l'environnement Matlab/Simulink est construit pour étudier les comportements du SSCB pendant une durée de court-circuit. L'accord entre la simulation et les résultats expérimentaux confirment que ce modèle JFET peut être utilisé pour simuler le fonctionnement d'un disjoncteur DC et dans l'étude du fonctionnement du microgrid DC pendant le processus de défaut et de compensation / Currently, there is a drop in the price of distributed energy resources, especially solar PVs, which leads to a significant growth of the installed capacities in many countries. On the other hand, policies encouraging energy efficiency have promoted the development of DC loads in domestic areas, such as LEDs lighting, computers, telephones, televisions, efficient DC motors and electric vehicles. Corresponding to these changes in sources and loads, DC microgrid distribution system becomes more attractive than the traditional AC distribution system. The main advantages of the DC microgrid are higher energy efficiency, easier in integrating with distributed energy sources and storage systems. While many studies concentrate on the control strategies and energy management in the DC microgrid, the protection still receives inadequate attention and lack of regulations and experiences. Protection in DC grids is more complex than AC grids due to the continuous arc, higher short circuit current value and fault rate of rising. Furthermore, the DC distributed grids are composed of many electronic and semiconductor switching devices, which only sustain the fault currents of some tens of microseconds. Mechanical circuit breakers, which have a response time in tens of milliseconds, seem not to meet the safety requirement of DC microgrids. The lack of effective protection devices is a barrier to the development of DC microgrids in the distributed systems. This thesis proposes a self-power solid state DC circuit breaker using normally-on SiC JFET, which offers a great protection device for DC microgrids due to its fast response time and low on-state losses. The design of the solid state DC circuit breaker aims to meet two objectives: fast response time and high reliability. The designed specifications and critical energies that result in the destruction of the circuit breaker are identified on the basis of the experiments of a commercial normally-on JFET. In addition, a very fast and reliable protection driver based on a forward-flyback converter topology is employed to generate a sufficient negative voltage to turn and hold off the SiC JFET. The converter will be activated whenever short-circuit faults are detected by sensing the drain-source voltage, then creating a negative voltage applied to the gate of JFET. To avoid gate failure by overvoltage at the gate of JFET, the output voltage of the forward-flyback converter is regulated using Primary Side Sensing technique. Experimental results validated the working principle of the proposed solid state DC circuit breaker with fault clearing time less than 3 μs. Additionally, a model of the normally-on JFET in Matlab/Simulink environment is built for exploring the behaviors of the solid-state DC circuit breaker during short-circuit faults. The agreement between the simulation and experimental results confirms that this JFET model can be appropriately used for the investigation of solid state DC circuit breaker operations and DC microgrids in general during fault evens and clearing fault processes

Depuis le début des années 2000, les composants en carbure de silicium (SiC) sont présents sur le marché principalement sous la forme de diodes Schottky et de transistors FET. Ces nouveaux semi-conducteurs offrent des performances en commutation bien supérieures à celles des composants en silicium (Si) ce qui se traduit par une diminution des pertes et une réduction de la température de fonctionnement à système de refroidissement identique. L’utilisation de composants SiC ouvre donc la possibilité de concevoir des convertisseurs plus compacts ou à une fréquence de commutation élevée pour une même compacité. C’est avec cet objectif d’augmentation de la fréquence de commutation qu’a été menée cette étude axée sur l’utilisation de composants SiC au sein d’un onduleur triphasé. Le convertisseur sur lequel se base l’étude accepte une tension d’entrée de 450V et fournit en régime nominal un courant de sortie efficace par phase de 40 A. Le choix des composants SiC s’est porté sur des transistors JFET Normally-Off et des diodes Schottky SiC car ces composants étaient disponibles à la vente au début de ces travaux et offrent des pertes en commutation et en conduction inférieures aux autres structures en SiC. Les transistors FET possèdent une structure et des propriétés bien différentes des IGBT habituellement utilisés pour des convertisseurs de la gamme considérée notamment par leur capacité à conduire un courant inverse avec ou sans diode externe. De ce fait, il est nécessaire de développer de nouveaux outils d’aide au dimensionnement dédiés à ces composants SiC. Ces outils de calculs sont basés principalement sur les paramètres électriques et thermiques du système et sur les caractéristiques des composants SiC. Les premiers résultats montrent qu’en autorisant la conduction d’un courant inverse au sein des transistors, il est possible de diminuer le nombre de composants. Basées sur ces estimations, une maquette de bras d’onduleur a été développée et testée. Les premiers thermiques montrent que pour une puissance de 12kW, il est possible d’augmenter la fréquence de commutation de 12 kHz à 100 kHz. / Since 2000, Silicon Carbide (SiC) components are available on the market mainly as Schottky diodes and FET transistor. These new devices provide better switching performance than Silicon (Si) components that leads to a reduction of losses and operating temperatures at equivalent cooling system. Using SiC components allows to a better converter integration. It is in this context that ECA-EN has started this thesis dedicated to using SiC devices in a three-phase inverter at high switching frequency. The converter object of this study is supply by a input voltage of 450V and provides a current of 40A per phase. The components used for these study are SiC Normally-Off JFET and Schottky Diodes because these devices were commercialized at the begining of this thesis and offer better switching performance than others SiC components. FET transistors have a different structure compared to traditionnal IGBT especially their capability to conduct a reverse current with or without body diode. So it is necessary to develop new tools dedicated to the design of converters built with SiC components. These tools are based on the electrical properties of the converters and the statics and dynamics characteristics of the transistor and the diode. The results show that when the transistors conduct a reverse current, the number of components/dies can be reduced. According to data, a PCB board of an inverter leg has been built and tested at ECA-EN. The thermal measurement based on the heatsink shows that the switching frequency of an inverter leg can be increased from 12 to 100 kHz for an ouput power of 12kW.

The objective of this thesis is to use the Majorana description of a spin-1 system to give a geometrical construction of a maximal set of Mutually Unbiased Bases (MUBs) and Symmetric Informationally Complete Positive Operator Valued Measures (SIC-POVMs) for this system. In the Majorana Approach, an arbitrary pure state of a spin-1 system is represented by a pair of points on the Reimann sphere, or a pair of unit vectors (known as Majorana vectors or M-vectors). Spin-1 states can be of three types: those whose vectors are parallel, those whose vectors are antiparallel and those whose vectors make an arbitrary angle. The types of bases possible for a spin-1 system are thus geometrically much more varied than for a spin-half system or qubit, which is the standard unit of information storage in most quantum protocols. Our derivation of the MUBs and SIC-POVMs proceeds from a recently derived expression for the squared overlap of two spin-1 states in terms of their M-vectors and the minimal additional set of assumptions that are needed. These assumptions include time-reversal invariance in the case of the MUBs and the requirement of three-fold symmetry in the case of the SIC-POVMs. The applications of these results to problems in quantum information are mentioned.

This PhD thesis sets out to investigate novel Silicon Carbide (SiC) based field effect devices (Schottky and transistor structures), with gas sensitive layers for monitoring hydrogen and propene gases at high temperatures. The devices developed by the author were shown to exhibit sensitivities at least 1~2 orders of magnitude (voltage shift, ƒ¢V) higher than those reported in literature. Not only did the author seek to investigate the gas sensing potential of such devices, but also he set out to study, analyse and establish the gas interaction mechanism of these novel sensors. High temperature tolerant hydrogen and hydrocarbon sensors are required in numerous applications such as: aerospace, nuclear power plant, space exploration and exhaust monitoring in automobiles. Monitoring these gases in a reliable and efficient manner is of great value in these applications, not only from a safety point of view but also for economical reasons. Hence there is an absolute necessity for simple, efficient and high performance sensors not only for monitoring and leak detection but also to function as part of a safety device to prevent accidents. The proposed sensor structure of combining SiC with gas sensitive oxide layers allow them to be operated at high temperatures, making them extremely appealing for direct or in-situ monitoring applications. The microstructural analysis performed using Atomic Force Microscopy (AFM), Scanning Electron Microscopy (SEM), X-ray Photoelectron Spectroscopy (XPS) and Rutherford Backscattering Spectroscopy (RBS) provides no evidence of inter-diffusion between different layers, in spite of the sensors being annealing at 650‹ in O2, H2 and C3H6 atmospheres for approximately 50hrs. Samples in different conditions (as deposited, annealed and tested) were compared. The electrical properties of the MROSiC (current-voltage, I-V and capacitance-voltage, C-V characteristics) and MESFET (drain current-source drain voltage (ID-VSD) and transfer, (ãID-H2 concentration) characteristics) devices were measured in the presence and absence of H2 and C3H6. Several parameters such as barrier height, saturation currents, pinch-off voltages and channel conductance were determined from the electrical characteristics, and their influence on the device performance was studied. The authorfs proposed gas interaction model based on energy band diagram is well supported by the experimental data obtained.

The lifetime of aluminium reduction cells is of considerable importance in the operating economics of an aluminium smelter. Not only are the cell materials costly, but cell life is also key in minimizing production downtime and the waste material generated in the cutting-out and relining of a reduction cell. This lifetime is increasingly driven by two components of the cell lining: the carbon cathode and the sidewall refractories which are the primary containment of the metal and electrolyte. Silicon nitride bonded silicon carbide (SNBSC) refractories are the current state-of-the-art sidewall materials. These materials are formed by encapsulation of SiC particles in a silicon nitride matrix, by nitridation of silicon at high temperatures under a nitrogen atmosphere. In this study the mechanisms of the corrosion of silicon nitride bonded silicon carbide refractories in an aluminium reduction cell environment were examined. Microstructural analysis of SNBSC materials by: X-ray diffraction, solid state NMR, Raman spectroscopy, XPS, and SEM, identified variations in α/β Si3N4 ratio and porosity in the binder phase, with higher porosity levels and β Si3N4 content found in the interior part of the brick. Unreacted metallic silicon was observed only as a crystalline phase inside SiC grains and is not associated with the binder phase. SNBSC samples from eight different commercial refractory manufacturers were tested in an industrial aluminium cell environment, both by immersion in industrial cells and in the laboratory. The samples showed extensive attack and considerable differentiation between the samples. The effects on the corrosion rate of porosity levels, amount of binder, α/β Si3N4 ratio, free Si content in the binder, and different factors in the environment were examined in the laboratory scale trials. A laboratory scale testing rig with some novel features, such as a rotating anode, was developed to examine corrosion resistance of samples under electrolysis conditions in molten cryolite at up to 1000oC. Corrosion test results, supported by thermodynamic calculations, identified the Si3N4 binder as the reactive phase in SNBSC materials, especially in the gas phase above the bath level. High porosity levels and high β Si3N4 fraction in the binder showed a statistically significant contribution to the corrosion rate in lab-scale corrosion trials. Commercial materials are also highly inhomogeneous in the distribution of the β Si3N4 phase as mentioned above, which complicates sample selection and the interpretation of corrosion test results. The crystal morphology of β Si3N4 is suggested as the reason for the high reactivity of these materials. This morphology characterised by elongated rod-like shape crystals with hexagonal cross section, presents a higher surface area compared to α Si3N4 crystals containing mainly flat matte crystals. In the light of this observation, the binder phase in selected samples was deliberately modified to convert a significant fraction of the α Si3N4 to β Si3N4. Corrosion rates measured from these modified samples in general support the argument that the crystallographic form of silicon nitride contributes to corrosion rate. The proposed corrosion mechanism for SNBSC materials in the aluminium reduction cell atmosphere is based on combination of oxidation of the binder followed by attack of corrosive gases to produce volatile SiF4. Thus the binder phase is initially passivated below the electrolyte level on the sidewall, where exposure to corrosive gases is limited, but occurs more rapidly in the area above the electrolyte/air interface. The intrusion of electrolyte into the refractory and capillary transport up the sidewall is a key in accelerating this reaction.

La mesure fiable et spatialement résolue du travail de sortie de matériaux et nanostructures est l'un des problèmes les plus importants en caractérisation de surface avancée pour les applications technologiques. Parmi les méthodes de mesure du travail de sortie disponibles, la microscopie à sonde de Kelvin (KFM), basé sur l'AFM, et la spectromicroscopie de photoémission d'électrons par rayons X (XPEEM) apparaissent comme les méthodes les plus prometteuses. Nous nous sommes intéressés dans cette thèse à l'investigation de la mesure locale du travail de sortie par KFM sous air ainsi qu'à l'étude de la complémentarité entre les mesures obtenues par KFM et XPEEM. Nous présentons tout d'abord une analyse des variations des mesures KFM avec la distance pointe-échantillon sur différents échantillons métalliques. Ces variations peuvent être expliquées par un modèle simple basé sur des transferts de charges de surfaces associés à l'inhomogénéité locale du travail de sortie. L'influence de la pointe, de l'environnement et des paramètres expérimentaux a été aussi étudié. Nous avons ensuite mis en évidence une augmentation de la résolution spatiale de l'imagerie KFM sous air grâce à la mise en oeuvre, sur le microscope commercial, d'un mode combiné AFM-KFM basé sur l‘acquisition simultanée des mesures KFM et de la topographie de surface. Enfin, nous nous sommes focalisés sur la caractérisation croisée par KFM et XPEEM de couches de graphène épitaxiées sur un substrat SiC(0001). Les images du travail de sortie obtenues par KFM sous air permettent de révéler qualitativement, avec une grande résolution spatiale, l'hétérogéneité de ces couches en surface. La mesure complémentaire par XPEEM spectroscopique au seuil de photoémission montre que cette hétérogéneité est liée à l'augmentation du travail de sortie local dû à une épaisseur de graphène variant entre 1 et 4-5 monocouches d'après la littérature. Ceci est corrélé qualitativement avec les intensités locales Si2p et C1s extraites des images XPEEM correspondantes. Des voies pour la compréhension de cette évolution atypique du travail de sortie d'un matériau, sont esquissées d'après la littérature disponible et les résultats micro-spectroscopiques, en terme de couplage électronique avec le substrat SiC.

<p>When implementing syncronous filters with global clock nets, a substantial amount of energy is consumed in the clock net. Hence, it is interesting to reduce the size of the clock net, which in turn will reduce the total energy consumption. One way to acheive this, is to create a SIC (Structured Interfacing of Computational elements) since such a structure can be created without a global clock net. Our results show that a fifth order wavedigital filter with a sample rate of at least 10M samples/sec may be achieved</p>

Silicon Carbide (SiC) is a wide band-gap semiconductor material with excel-lent material properties for high frequency, high power and high temperature elec-tronics. In this work different SiC field-effect transistors have been studied using theoretical methods, with the focus on both the devices and the methods used. The rapid miniaturization of commercial devices demands better physical models than the drift-diffusion and hydrodynamic models most commonly used at present. The Monte Carlo method is the most accurate physical methods available and has been used in this work to study the performance in short-channel SiC field-effect devices. The drawback of the Monte-Carlo method is the computational power required and it is thus not well suited for device design where the layout requires to be optimized for best device performance. One approach to reduce the simulation time in the Monte Carlo method is to use a time-domain drift-diffusion model in contact and bulk regions of the device. In this work, a time-domain drift-diffusion model is implemented and verified against commercial tools and would be suitable for inclusion in the Monte-Carlo device simulator framework. Device optimization is traditionally performed by hand, changing device pa-rameters until sufficient performance is achieved. This is very time consuming work without any guarantee of achieving an optimal layout. In this work a tool is developed, which automatically changes device layout until optimal device per-formance is achieved. Device optimization requires hundreds of device simulations and thus it is essential that computationally efficient methods are used. One impor-tant physical process for RF power devices is self heating. Self heating can be fairly accurately modeled in two dimensions but this will greatly reduce the computa-tional speed. For realistic influence self heating must be studied in three dimensions and a method is developed using a combination of 2D electrical and 3D thermal simulations. The accuracy is much improved by using the proposed method in comparison to a 2D coupled electro/thermal simulation and at the same time offers greater efficiency. Linearity is another very important issue for RF power devices for telecommunication applications. A method to predict the linearity is imple-mented using nonlinear circuit simulation of the active device and neighboring passive elements.

The importance of simulation is increasing in the researchon semiconductor devices and materials. Simulations are used toexplore the characteristics of novel devices as well asproperties of the semiconductor materials that are underinvestigation, i.e. generally materials where the knowledge isinsufficient. A wide range of simulation methods exists, andthe method used in each case is selected according to therequirements of the work performed. For simulations of newsemiconductor materials, extremely small devices, or deviceswhere non-equilibrium transport is important, the Monte Carlo(MC) method is advantageous, since it can directly exploit themodels of the important physical processes in the device. One of the semiconductors that have attracted a lot ofattraction during the last decade is silicon carbide (SiC),which exists in a large number of polytypes, among which3C-SiC, 4H-SiC and 6H-SiC are most important. Although SiC hasbeen known for a very long time, it may be considered as a newmaterial due to the relatively small knowledge of the materialproperties. This dissertation is based on a number of MCstudies of both the intrinsic properties of different SiCpolytypes and the qualities of devices fabricated by thesepolytypes. In order to perform these studies a new full-bandensemble device MC simulator, the General Monte CarloSemiconductor (GEMS) simulator was developed. Algorithmsimplemented in the GEMS simulator, necessary when allmaterial-dependent data are numerical, and for the efficientsimulation of a large number of charge carriers in high-dopedareas, are also presented. In addition to the purely MC-relatedstudies, a comparison is made between the MC, drift-diffusion,and energy-balance methods for simulation of verticalMESFETs. The bulk transport properties of electrons in 2H-, 3C-, 4H-and 6H-SiC are studied. For high electric fields the driftvelocity, and carrier mean energy are presented as functions ofthe field. For 4H-SiC impact-ionization coefficients,calculated with a detailed quantum-mechanical model ofband-to-band tunneling, are presented. Additionally, a study oflow-field mobility in 4H-SiC is presented, where the importanceof considering the neutral impurity scattering, also at roomtemperature, is pointed out. The properties of 4H- and 6H-SiC when used in short-channelMOSFETs, assuming a high quality semiconductor-insulatorinterface, are investigated using a simple model for scatteringin the semiconductor-insulator interface. Furthermore, theeffect is studied on the low and high-field surface mobility,of the steps formed by the common off-axis-normal cutting ofthe 4H- and 6H-SiC crystals. In this study an extension of theprevious-mentioned simple model is used.

No description available.