Desenvolvemos neste trabalho, um formalismo alternativo de \"dois níveis\" sugerido por Cattani, para o cálculo dos alargamentos e deslocamentos de linhas espectrais produzidos por colisões, quando é valida a aproximação de impacto. Este formalismo foi usado para calcular alargamentos e deslocamentos eletrônicos de algumas linhas do átomo de hélio neutro em um plasma, para as quais, foi possível aplicar a teoria de \"dois níveis\". A contribuição das colisões iônicas aos alargamentos e deslocamentos das linhas espectrais foi avaliada, utilizando a aproximação quase estática de Griem, Baranger, Kolb e Oertel. Os resultados dos cálculos foram comparados com os resultados experimentais de Kelleher, Berg e colaboradores, Wulff e Diatta e com as previsões dos formalismos, convergente e de corte. / In this work we develop an alternative formalism of \"two levels\" suggested by Cattani to the calculation of widths and shifts of spectral lines produced by collisions, in the impact approximation. This formalism was used to calculate in a plasma the electronic widths and shifts of neutral helium lines for which is valid the \"two levels\" approach. To take into account the ions contribution to the widths and shifts of the lines, we have . used the quasistatic approximation of Griem, Baranger, Kolb and Oertel. Our theoretical predictions are compared with the experimental results of Kelleher, Berg and collaborators, Wulff and Diatta and also with the predictions of the convergent and cutoff formalisms.

A contribuição dos eventos extremos, chamados \\emph{bursts}, relacionados com estruturas coerentes que se propagam na turbulência, parece ser parte da explicação para o transporte anômalo que acontece em plasmas confinados. Porém, ainda não existe um modelo adequado no quadro da teoria neoclássica de transporte, para explicar adequadamente a contribuição destes eventos extremos. Assim, a caracterização do transporte de partículas e energia, devido a esses eventos extremos, e a sua estrutura interna, são questões relevantes no estudo do plasma de fusão. Para isso, medições locais da temperatura dos elétrons são necessárias, pois o transporte de partículas e energia sofre correções devido à temperatura. Por terem boa resolução espacial e serem relativamente fáceis de construir, as sondas eletrostáticas são comumente utilizadas em tokamaks para medir a temperatura local na região da borda de plasmas confinados magneticamente. Os métodos de varredura de tensão e a configuração de sonda tripla estão dentre os métodos mais usados para medir a temperatura local na borda do plasma com sondas eletrostáticas. Neste trabalho, é apresentada uma comparação entre as medições de temperatura usando esses dois métodos no tokamak TCABR. Para esse fim, foram utilizadas descargas de plasma em condições padrão, onde a corrente de plasma e a densidade são mantidas praticamente estáveis e há baixa atividade MHD, isto é, a fase estacionária da descarga, com os parâmetros macroscópicos de equilíbrio aproximadamente constantes. As medições da temperatura pelos dois métodos têm boa correspondência na região da sombra do limitador, mas existem discrepâncias quando as medidas são feitas dentro da coluna de plasma. Verificou-se que esta discrepância se deve ao efeito de expansão da região de transição entre o plasma e a superfície da sonda, conhecida como bainha de plasma. Portanto, foram modificadas as equações do método da sonda tripla a fim de considerar, de maneira adequada, este efeito. Esta modificação conduz a medições de temperatura compatíveis entre os dois métodos. Também se estudou as estruturas coerentes que se propagam na turbulência da borda (bursts), que são detectados como picos no sinal de corrente de saturação iônica. Para esse fim, foi usada a técnica da análise condicional. Os resultados obtidos parecem ser compatíveis com estruturas que apresentam uma inclinação no plano radial-poloidal. Esta inclinação implica que o método comumente usado para medir a velocidade de propagação das estruturas, baseado no atraso entre a detecção do pico entre dois pinos próximos, pode levar a resultados completamente errados. Os resultados para a evolução média da temperatura durante os bursts são incompatíveis entre os dois métodos. De fato, os resultados fornecidos pelo método de sonda tripla aparentam estar fisicamente errados, uma vez que indicam um alto gradiente de temperatura dentro dos bursts. Constatou-se que este problema é devido aos altos gradientes de potencial dentro dos bursts que inviabilizam a hipótese de potencial uniforme na região entre os pinos da sonda tripla. / The contribution of the extreme events related to propagating structures in the turbulence, the so-called bursts, seems to be part of the explanation to anomalous transport in confined plasma. But, up to now, there is not an adequate model in the frame of neoclassical transport theory to properly account the contribution of these extreme events. Therefore, the characterization of the particle and energy transport due to the extreme events and the burst internal structure are an important issue in fusion plasma research. However, local electron temperature measurements are necessary to investigate those topics, because particles and energy transport have temperature corrections. Due to their good spatial resolution and relatively easy construction, electrostatic probes are often used to measure temperature in the plasma edge of tokamak. The triple probe configuration and voltage sweep technique are among the most common techniques used with electrostatic probes to measure local temperature in the plasma edge. We present a comparison between the temperature measurement using these two techniques in the tokamak TCABR. For this purpose, it was used stationary standard plasma discharges (it is, discharges with almost stable plasma current and density and with low MHD activity). For temperature measurement, the two methods have good correlation in the Scrape-Off-Layer region but there are discrepancies between their values inside the plasma column. We found that this discrepancy is due to the sheath expansion effect. Therefore, the triple probe equations were modified to properly consider this effect. The modification leads to compatible average temperature measurements between the two methods. We also studied coherent structures propagating in the edge turbulence (detected as bursts in the saturation current) by using the conditional analysis technique. The results seem to be compatibles with structures tilted in the radial-poloidal plane. This tilt implies that the common method to measure the propagation velocity using the delay time of detection between two probes, could produce wrong values. The results for the average temporal behavior of temperature during the bursts are inconsistent between the two methods. Indeed, the triple probe results seems to be physically wrong, once it indicates a strong temperature gradient inside the bursts. We found that this problem is due to the strong potential gradients inside the burst, breaking the assumption that the pins used in the triple probe configuration are in a homogeneous plasma.

Le nitrure de gallium (GaN) est un semi-conducteur à grand gap dont les propriétés en font un bon candidat pour remplacer le Si dans le domaine de l’électronique de puissance. Dans cette optique, cette thèse s’est intéressée à l’une des étapes de la conception d’une diode de puissance Schottky à base de GaN épitaxié sur Si : la gravure. L’objectif était alors de développer des procédés permettant une reprise de contact sur une couche enterrée de GaN n+. Ils devaient combiner des performances de gravure compatibles avec des processus industriels et conduire à des profils et états de surface adaptés à la réalisation d’un contact ohmique. La réalisation d’une structure pseudo-verticale, au travers de la gravure par la face supérieure d’une couche de 5 à 6 μm de GaN a été étudiée avec quatre réacteurs plasma différents. Des analyses du plasma et du matériau après gravure ont mis en évidence que le meilleur compromis était obtenu pour des plasmas inductifs en chimie chlorée avec une tension d’auto-polarisation modérée. Il a également été déterminé que l’ajout d’un gaz fluoré dans un plasma de ce type permettait de générer une passivation à même de protéger la surface du GaN. Pour répondre à la chute de vitesse de gravure qu’elle entraîne, un procédé alternant des étapes de gravure et de passivation a été développé et étudié. D’autre part, des recherches ont été menées afin de créer une structure verticale. Afin d’atteindre le GaN n+, des gravures en face arrière de vias de Si d’une profondeur de 300 μm par les procédés cryogéniques standard et STiGer anisotrope ont été effectuées. Un procédé de révélation permettant l’étude des couches composant le tampon, situé entre le Si et le GaN, a été mis au point. Leur gravure par des plasmas chlorés s’est révélée efficace au travers ou non de vias de Si. / Gallium nitride (GaN) is a wide bandgap III-V semiconductor with interesting electrical properties in order to replace Si in the field of power devices. The subject of this thesis was then to study one of the essential steps to realize a GaN based Schottky diode : the etch. The aim was to develop etching processes allowing the realization of an ohmic contact on an embedded GaN n+ layer. They must combine GaN etching performances compatible with industrial requirements and etch profiles and surface states compatible with an ohmic contact. The etching of a 5 to 6 μm GaN layer by four different reactors was studied in order to realize a pseudovertical structure. Plasma and GaN surfaces analyses were performed during and after etching by five developed and optimized processes. These analyses showed that best compromise was obtained for inductive chlorine plasmas with a moderate bias. The creation of a passivation layer which is able to protect GaN surface, thanks to fluorine addition in plasma chemistries was also established. To overcome the etch rate decreases induced by fluorine addition, a time multiplexed etching process, alternating etch and passivation steps, was developed and studied. Researches were also performed to achieve a vertical diode. Etching of 300 μm depth Si vias by standard cryogenic and anisotropic STiGer processes were carried out. A revelation process was also developed in order to study buffer layers etching. Effective buffer etching by chlorine plasma was demonstrated with or without Si vias.

De rapides avancées technologiques ont autorisé depuis le début des années 1980 un développement important des lasers de puissance et ont ouvert la voie aux régimes d’interaction laser-matière relativiste. L’accès aux intensités lumineuses supérieures à 1.1018 W.cm-2 a ainsi donné la possibilité à la communauté scientifique d’explorer une nouvelle physique riche d’applications. Bien que le principal moteur ait historiquement été constitué par les recherche sur la fusion par confinement inertiel, l’astrophysique de laboratoire, la génération de rayonnement (harmoniques, bétatron, X) ou la génération de particules énergétiques (électrons, ions) élargissent de plus en plus ce domaine d’étude.En effet, les très bonnes qualités des sources lumineuses et des sources d’ions créées par laser laissent fortement envisager qu’elles viendront un jour remplacer les sources conventionnelles comme les synchrotrons ou les accélérateurs qui sont des machines très coûteuses.Dans le cadre de cette thèse, une attention toute particulière a été portée à l’accélération d’ions qui a déjà montré son fort potentiel en termes de qualité des faisceaux accélérés. Malheureusement, ses applications sont encore limitées (radiographie, chauffage isochorique) à cause de paramètres limitants comme la divergence du faisceau, le spectre large ou l’énergie maximale atteinte par ces faisceaux. Au cours de ce travail de thèse, l’accent a plus particulièrement été mis sur l’augmentation de l’énergie maximale des faisceaux de protons dans le cadre des régimes à ultra haute intensité (supérieur à 1.1019W.cm-2). Cette recherche s’est orientée suivant deux axes principaux (impulsions longue et courtes), qui ont donné lieu à de nombreux échanges et au renforcement de la collaboration entre les laboratoires du LULI à l’École Polytechnique (France) et l’INRS-EMT (Canada).Dans le cadre des recherches menées au sein du LULI, des techniques innovantes ont pu être explorées afin de poursuivre la compréhension des mécanismes et d’améliorer les qualités de l’accélération d’ions à partir d’impulsion “longues” (entre 300 fs et 1,5 ps). Nous avons montré que l’utilisation de cibles ayant des dimensions transverses réduites permettait le confinement géométrique des électrons dans la zone d’impact du laser et augmentait ainsi significativement le taux de conversion de l’énergie laser vers les protons et l’énergie maximale atteinte par le faisceau. Par ailleurs, l’utilisation originale d’une optique plasma refocalisante a démontré son efficacité quant à réduire fortement la surface de focalisation du laser, conduisant à augmenter son intensité et donc l’énergie de coupure des faisceaux d’ions accélérés. Enfin, l’utilisation de deux impulsions laser a mis en évidence qu’une interaction entre les électrons accélérés par chaque impulsion était possible et qu’elle permet de modifier l’énergie et la typologie des faisceaux de protons.Les expériences réalisées au sein de l’INRS-EMT visaient quant à elle à améliorer la compréhension des régimes d’accélération femtoseconde, où peu d’études à ultrahaute intensité existaient au début de cette thèse, et à valider ou non la pertinence de ces régimes. Les nombreuses expériences menées ont clairement établi l’importance du contraste laser et la nécessité que ce derniers soit important pour que l’accélération de protons soit efficace dans ces régimes ultracourts. L’analyse systématique des faisceaux accélérés en face avant et en face arrière d’une cible mince montre que le processus d’accélération manifeste une certaine symétrie et prouve, qu’à énergie laser constante,l’accélération d’ions par laser n’est pas la plus efficace pour la plus courte durée d’impulsion. / Technological improvements since the 1980s have allowed for important developments in the field of high power lasers, thus paving the way for relativistic laser-matter interactions. With laser intensities higher than 1.1018 W.cm-2, the scientific community could explore a new physics, full of promising applications. Historically, laser–plasma interaction research initially aimed at exploring fusion by inertial confinement, but, with the breakthrough of ultra-high power lasers, the scope of research could now be broadened to laboratory astrophysics, radiation generation (harmonics, betatron, X) and the production of high energy particles (electrons, ions). Regarding the latter, radiation and ion sources are of such excellent quality that they might in the future replace current conventional sources like synchrotrons or accelerators, which require very expensive facilities. This thesis focuses particularly on laser-driven ion acceleration, whose accelerated beams have already demonstrated strong potential, e.g. in ultrafast imaging or warm dense matter generation.Within this domain, the present work focused on strategies developed to increase the ion beam energy in the ultrahigh intensity regime (higher than 1.1019 W.cm-2), exploiting as well moderate (400 fs) and short (25 fs) pulses facilities available as a result of the collaboration between the LULI laboratory in France and the INRS-EMT in Quebec. Innovative acceleration techniques have been explored at the LULI laboratory using moderately short laser pulses (400 fs to a few ps). This has been done first by improving our understanding of acceleration physics. Then, confinement of the laser-driven fast electrons that are at the source of the ion acceleration could be obtained by using reduced size targets. With such targets, electron confinement in the acceleration area could be achieved, inducing improvement of the laser to ions conversion efficiency, the ion beam cut off energy, and the ion beam quality. Another strategy that was exploited was to use refocalizing plasma optics to produce strongly reduced laser focal spot sizes. This induces laser intensity increase and thus improvement of the ion beam cut off energy. Finally, we used the combination of two laser pulses to have the electrons accelerated by each laser pulse interact together. When this was the case, we noted an increase of the ion beam cut off energy along with a modification of the beam typology. Complementarily, the experiments carried out using the 200 TW laser system in Quebec aimed at improving our understanding of femtosecond ion acceleration regimes, as only a fewexperimental studies were yet available, and to confirm the relevance of these regimes for ion acceleration. The results obtained with this laser clearly show the important role of the laser pulse contrast ratio, and the need for it to be extremely high to obtain efficient ion acceleration in this ultrashort regime. The systematic recording of accelerated ion beams has showed that a quasisymmetric acceleration from the target front and rear sides is possible. These results have alsoproved that the highest proton energy is not necessarily obtained with the shortest pulse duration when the laser energy is kept constant. Thus, we demonstrated that the shortest pulses available today (i.e. 25 fs) are not the most efficient to accelerate ion beams.

none

Thesis (Ph. D.)--University of Texas at Austin, 2002. / Vita. Includes bibliographical references. Available also from UMI Company.

Thesis (Ph. D.)--University of Washington, 2001. / Vita. Includes bibliographical references (leaves 164-166).

Thesis (Ph. D.)--University of Washington, 2003. / Vita. Includes bibliographical references (p. 133-135).

Thesis--M. Phil., University of Hong Kong, 1975.

Thesis (MTech (Mechanical Engineering))--Peninsula Technikon, 1999. / Word processed copy. Summary in English. Includes bibliographical references (leaves 66-68). Also available online.