L'objectif de cette thèse s'oriente principalement sur l'étude du de puisclaquage post-radiatif des MOSFET sances irradiés aux ions lourds. Nous avons pu constater, pour la première fois, une réduction des paramètres de fiabilité et de durée de vie des MOSFET de puissance irradiés aux ions lourds grâce à un protocole de test combinant les connaissances sur les effets des rayonnements et des tests électriques accélérés. Les MOSFET ont été irradiés principalement en absence de polarisation, dans le but de discriminer tout effet dû aux polarisations. Grâce à des irradiations réalisées à différentes valeurs d'énergie, nous nous sommes intéressés aux effets de l'énergie des particules et des pertes d'énergie associées. Nous avons pu constater que la dégradation de la fiabilité des MOSFET ne peut pas être corrélée uniquement à l'énergie perdue par excitation électronique (LET), ou pour des ions avec la même valeur de LET obtenue pour deux énergies différentes, le pire cas a été observé à l'énergie la plus faible. Cette dégradation est même plus importante que celle obtenue à la valeur maximale de LET (au pic de Bragg). Appuyé par des résultats obtenus grâce à des irradiations aux neutrons, nous avons pu proposer une hypothèse qui est basée sur une corrélation entre l'effet des pertes d'énergie associées aux passages des ions et les mécanismes de claquage des diélectriques. / The goal of this thesis is oriented mainly on the study of Post Irradiation Gate Stress (PiGS) of power MOSFETs irradiated with heavy ions. We have seen, for the first time, a reduction of reliability parameters and lifetime of power MOSFETs irradiated with heavy ions using a test panel combining the knowledge of the effects of radiation and accelerated electrical test. MOSFETs were irradiated mainly with no polarization, in order to discriminate any effect attributed to the polarizations. Using irradiation performed at different energy values, we investigated the effects of the energy and energy lost by ionizing and non ionizing process. We have seen that the reliability degradation of MOSFETs can't be correlated only to the energy lost by electron excitation (LET), or ions with the same LET value obtained for two different energies, the worst case was observed at the lowest energy. This degradation is even greater than that obtained with the maximum value of LET (the Bragg peak). Supported by results obtained by neutrons irradiation, we could propose a hypothesis that is based on a correlation between the effect of energy lost associated with the passage of ions and mechanisms of dielectric breakdown.

L’émission électronique par effet de champ est un domaine qui concerne de nombreuses applications techniques. Dans ce travail, nous avons réalisé une étude essentiellement expérimentale des différents paramètres susceptibles d’avoir une influence sur l’émission électronique. En première partie, nous exposons les résultats obtenus pour un champ électrique homogène correspondant aux faibles intensités de courant. Le rôle de la distance inter-électrodes à champ constant et l’influence de la rugosité de surface sur l’émission électronique ont été étudiés. Nous discutons la méthode classique de Fowler-Nordheim utilisée pour le dépouillement des mesures en y portant un regard critique. Un modèle simple visant à prendre en compte l’échauffement des sites émetteurs est proposé. La seconde partie concerne l’effet de l’injection de gaz sur l’émission électronique, effet qui se traduit par une diminution du courant émis quand on augmente la pression de 10⁻⁶ Pa à 10⁻² Pa à champ macroscopique constant. Nous exposons des résultats montrant un effet de seuil concernant l’apparition de l’effet du gaz sur l’émission électronique. Nous présentons également des résultats pour différents matériaux de cathode et pour différents gaz (He, H₂, N₂, Ar). Une réversibilité de cet effet est montrée après le pompage pour redescendre à 10⁻⁵ Pa. La décroissance de courant par effet de gaz est interprétée par la diminution de la valeur du facteur d’accroissement local du champ électrique (β) au niveau des émetteurs à cause du bombardement de ces sites par les ions créés à leur proximité. Un calcul du flux d’ions bombardant un site émissif a permis d’estimer le temps nécessaire pour modifier un émetteur de façon cohérente avec les observations expérimentales. La théorie de la migration des atomes en surface de l’électrode en présence d’un champ électrique est proposée pour expliquer la réversibilité de l’effet de gaz observée qui est, selon cette théorie, liée à l’augmentation de la valeur de β au niveau des émetteurs. / Field electronic emission is a domain which concerns numerous different technical applications. In this work, we have taken an essentially experimental approach to study various parameters having influence on field emission. In the first part of the thesis, we have described the results obtained with a homogeneous electric field with relatively weak field emission. The role of the inter-electrode distance at constants field as well as that of the cathode surface roughness on field emission are studied. The classical method of Fowler-Norheim was then used for the analysis of the measurements. A simple model aiming to take into account the effect of the heating of the emission sites is then proposed. The second part of the theses concerns the effect of the injection of gas on the field emission; this effect being to significantly reduce emission intensity when the gas pressure is raised from 10⁻⁶ to 10⁻² Pa at constant field. A threshold value of emission intensity is shown to be necessary for the observation of this gas effect. The effect of different gas types (He, H₂, N₂, Ar) and cathode materials are also described. The gas effect is shown to be reversible upon lowering of the gas pressure to 10⁻⁵ Pa. The reduction in current is interpreted by a lowering of the field enhancement factor (β) of emission sites by ionic bombardment by ions created locally (within distances on the order of microns) near the cathode surface. A calculation of the flux of bombarding ions is used to estimate the time necessary to modify an emission site in a way corresponding to the observations. The phenomenon of surface migration in the presence of intense electric field is then proposed to explain the reversibility of the gas effect, increasing the local field enhancement factor.

97% des articles publiés sur les études climatiques racontent que le réchauffement climatique est entièrement causé par les activités humaines. Les gaz émis lors de la production d'énergie électrique ainsi que d'autres gaz rejetés par les voitures ont un réel impact sur l'atmosphère. Une solution consiste à mettre au point des composants présentant des pertes de conduction plus faibles et des caractéristiques de claquage plus élevées qui pourraient être utilisés dans des centrales nucléaires, des cellules de commutation à haute puissance, des voitures hybrides (électriques), etc.De nos jours, les composants à base de silicium contrôlent environ 95% des dispositifs électroniques. Le carbure de silicium SiC et le nitrure de gallium GaN sont actuellement à l’étape de R&D, et commencent à être intégrés dans certains circuits électroniques. D'autres matériaux tels que Ga2O3, AlN ou le diamant sont encore à l’étape de recherche. Les derniers sont connus sous le nom de matériaux à bande ultra large et semblent être la solution requise pour les faibles pertes de puissance. Le diamant est reconnu comme le matériau ultime pour la prochaine génération de composants de puissance en raison de ses propriétés physiques exceptionnelles telles qu'un champ de claquage élevé (>10 MV/cm) permettant d'utiliser le dispositif pour une commande de puissance élevée, une mobilité de porteurs élevée (2 000 cm^2/V.s pour les trous), une vitesse de saturation élevée, une conductivité thermique élevée (22 W/cm.K) pour une parfaite dissipation de chaleur et une faible constante diélectrique. Théoriquement, le diamant est le semi-conducteur offrant le meilleur compromis entre résistance à l'état passant et tension de claquage. En particulier, en raison de l'ionisation incomplète des dopants, il est encore plus efficace à haute température. Diverses diodes Schottky en diamant (SBD) avec de bonnes performances à l’état passant et bloqué (7,7 MV/cm) ont été rapportées. En plus des SBDs, des transistors à effet de champ (FET) ont également été étudiés à travers des oxyde-métal semi-conducteur FETs (MOSFETs) utilisant une surface hydrogénée avec des densités de courant élevées à l'état passant ou des surface oxygéné avec de bonnes caractéristiques de blocage. Pour les composants de haute-tension, il est nécessaire de changer l’architecture de l’électrode afin d’éviter un claquage prématuré due à l’encombrement du champ électrique aux bords. Dans ce but, les techniques de terminaison de bord sont utilisées pour atteindre les caractéristiques idéales. La tâche évidente avant toute fabrication de composant est la partie simulation qui prédit l’optimisation de l’architecture et les caractéristiques attendues. Une bonne prédiction nécessite la connaissance des paramètres du matériau. Les paramètres importants pour le claquage sont les coefficients d'ionisation par impact. Plusieurs coefficients ont été publiés pour le diamant. Toutefois, ils ont été extraits en « fittant » des structures non optimisées, d'où un manque de précision.Dans cette étude, deux structures de terminaisons de bord pour des diodes Schottky, appelées plaque de champ et anneaux à champ flottant, ont été étudiées. Leur efficacité de distribution du champ de surface par analyse de courant induit par faisceau d'électrons (EBIC) a été observée. De plus, des FETs ont été fabriqués et caractérisés, un MESFET et un RB-MESFET. Les FETs présentent un claquage élevé, jusqu’à 3 kV et une faible résistance. Le développement des transistors est indissociable de la diode Schottky, car ils sont tous deux nécessaires à la fabrication de cellules de commutation. Et enfin, les coefficients d'ionisation par impact pour les électrons ont été mesurés à l'aide d’EBIC pour un champ >0,5 MV/cm dans une région sans défaut. Les valeurs mesurées sont (sous l’equation de Chynoweth) an = 971 /cm et bn=2,39x10^6 V/cm. Ces valeurs sont proches des coefficients mesurés expérimentalement et rapportés dans la littérature. / 97% of the published climate studies articles agree with the fact that recent global warming is entirely caused by human activities. The gases emitted to produce electrical energy plus other gases rejected by cars impact considerably on the atmosphere by greenhouse effect (without referring other factors). A solution to this problem is the development of components with lower power conduction losses and higher breakdown characteristics that could be used in nuclear power plants, high power commutation cells, hybrid (electric) cars and so on.The choice of the material to reach low power conduction losses and higher breakdown is of great importance. Nowadays, silicon-based devices control about 95% of all electronic components. Silicon carbide SiC and gallium nitride GaN are at present under research and development and start to be integrated into some electronic circuits. Other materials like Ga2O3, AlN or diamond are under research for power electronic application. The last ones are known as ultra wide bandgap materials and they seem to be the required solution to low power losses. Diamond is recognized as the ultimate material for the next next-generation of power devices owing to its exceptional physical properties such as high breakdown field (>10 MV/cm) to use the device for high power control, high carrier mobility (2000 cm^2/V.s for holes) for fast switching and high frequency devices, high saturation velocity, high thermal conductivity (22 W/cm.K) for a perfect heat dissipation and low dielectric constant. Theoretically, diamond is the best semiconducting material showing the best trade-off between on-resistance and breakdown voltage. Especially, due to the incomplete ionization of the dopant, it is even more efficient at high temperature. Various diamond Schottky barrier diodes (SBDs) with good forward and reverse performances (7.7 MV/cm) were reported. In addition to SBDs, switches diamond field effect transistors (FETs) were also investigated through metal-oxide-semiconductor FETs (MOSFETs) using either an H-terminated diamond surface with high current densities in on-state or an O-terminated one with high blocking characteristics. For the high blocking voltage devices, one needs to properly terminate the edge of the electrode at the surface in order to avoid premature breakdown of the devices due to electric field crowding at the borders. In that aim, edge termination (ET) techniques are used to push the limit of the devices and reach ideal features. The obvious task before any device fabrication if the simulation part that predicts the device optimization and expected characteristics. A good device prediction requires knowledge of the material parameters. Important parameters for device breakdown in the off-state are the impact ionization coefficients. At present, several ionization coefficients were reported for diamond, however, they were extracted by fitting non-optimized structures and hence there is a lack of accuracy.In this study, two edge terminations structures for Schottky barrier diodes called field plate (FP) oxide and floating field rings were investigated. Their effectiveness in surface field distribution via electron beam induced current (EBIC) analysis was observed. In addition, normally-on FETs were fabricated and characterized, a MESFET and a reverse blocking (RB)-MESFET. The FETs exhibited a high BV, up to 3 kV and a low on-resistance. The development of transistors is inseparable from the Schottky diode since both are required to fabricate commutation cells. And finally, impact ionization coefficients for electrons were measured using EBIC for a field >0.5 MV/cm in a defect-free region. The measured values are (in a Chynoweth form) an = 971 /cm and bn = 2.39x10^6 V/cm. These values are close to the experimentally measured coefficients reported in the literature.

A espectrometria de emissão óptica com plasma induzido por laser (LIBS) constitui um método alternativo para a determinação simultânea de macro e micronutrientes em pastilhas de materiais vegetais. No entanto, pouca informação com relação às condições de fluência e de focalização do laser mais apropriadas para a análise de pastilhas de materiais vegetais são disponíveis na literatura. O objetivo deste trabalho foi avaliar os efeitos da fluência, do diâmetro de focalização do laser e do tamanho das partículas das amostras nas análises de pastilhas de tecidos vegetais por LIBS. O arranjo experimental foi composto por um laser pulsado de Nd:YAG a 1064 nm (pulsos de 5 n\'delta\', 360 mJ, 10 Hz) e os sinais de emissão foram coletados por um telescópio acoplado por fibra óptica ao espectrômetro Echelle com detector ICCD. Os parâmetros instrumentais foram ajustados em 20 pulsos acumulados, 2,0 \'mü\'s de atraso e 5,0 \'mü\'s de integração. Amostras laboratoriais consistiram de pastilhas preparadas com folhas moídas criogenicamente (90 % das partículas < 50 \'mü\'m). Maiores intensidades das linhas Ca I 442,554, Mg I 277,983, P I 213,618, Al I 309,271, B I 249,773, Cu I 324,755, Fe II 261,187, Mn II 257,610 e Zn II 206,200 nm foram obtidas com fluências entre 35 e 60 J cm-2 e diâmetros de focalização entre 600 e 1050 \'mü\'m. O uso de fluências e diâmetros de focalização maiores concorreu para melhorar os coeficientes de variação (CV) das medidas. Fixando-se o diâmetro de focalização em 750 \'mü\'m, observou-se que, para a maioria dos analíticos, o CV das medidas foi reduzido 2 vezes quando a fluência foi alterada de 20 para 50 J cm-2. Nestas condições, estes resultados foram atribuídos à maior massa removida e, conseqüentemente, à amostragem mais representativa. Os coeficientes angulares das curvas de calibração também aumentaram quando a fluência foi acrescida. No entanto, não foram verificadas diferenças significativas nos coeficientes de correlação das curvas de calibração e nos limites de detecção empregando-se 25 e 50 J cm-2. Embora as intensidades dos sinais de emissão aumentem com a fluência, a emissão de fundo também aumenta. Por outro lado, o tamanho das partículas das amostras afeta as eficiências de atomização e excitação no plasma induzido por laser. Quando partículas maiores são parcialmente vaporizadas, os sinais de emissão resultantes dos processos de atomização e excitação das espécies constituintes são menores do que os provenientes de partículas menores. Nesta dissertação, melhorias significativas da intensidade dos sinais de emissão e dos CVs das medidas foram observados quando pastilhas preparadas com partículas < 75 \'mü\'m foram analisadas. Não foram observadas diferenças significativas nas intensidades dos sinais de emissão e no CV das medidas nas análises de pastilhas preparadas com partículas entre 20 e 75 \'mü\'m. Pastilhas preparadas com partículas > 75 \'mü\'m apresentaram problemas de coesão e, após ablação, apresentaram crateras não uniformes. A análise destas pastilhas resultou em CV das medidas geralmente > 15 %. Os resultados obtidos neste trabalho confirmaram que é recomendável utilizar amostras de calibração e amostras-teste com distribuição do tamanho das partículas semelhante para evitar efeitos físicos de matriz, que alteram as interações laser-amostra / Most recently, it was demonstrated that laser induced breakdown spectrometry (LIBS) constitutes an alternative for the simultaneous determination of macro- and micronutrients in pellets of plant materials. However, there is no information regarding the most appropriate fluence and laser focusing for analysis of pellets of plant materials. The aim of this work was to evaluate the effect of laser focusing and fluence on LIBS analysis of plants. The effect of particle size on LIBS analysis of pellets of plant materials as well as sample preparation approaches were also discussed. The experimental setup was designed by using a Q-switch Nd:YAG laser (5 n\'delta\', 360 mJ, 10 Hz, _ \'lâmbda\' = 1064 nm) and the emission signals were collected by lenses into an optical fiber coupled to an echelle spectrometer equipped with a high-resolution ICCD. Instrumental parameters consisted of 20 accumulated laser pulses, 2.0 \'mü\'s delay time and 5.0 \'mü\'s integration time gate. Pellets prepared from cryogenically ground plant leaves (90 % particles < 50 \'mü\'m; median = 10 \'mü\'m) were used as laboratory samples and 6 laser spot diameters (from 180 to 1050 \'mü\'m) at different laser fluences were evaluated. Results indicate that elements emission intensities increased with both laser fluence and spot diameter. Higher intensities for Ca I 442.554, Mg I 277.983, P I 213.618, Al I 309.271, B I 249.773, Cu I 324.755, Fe II 261.187, Mn II 257.610, and Zn II 206.200 nm emission lines were observed with fluences in the 35 - 60 J cm-2 range and spot diameters between 600 and 1050 \'mü\'m. Higher spot diameters improved mass removal and minimized repeatability drawbacks. It was observed that the coefficients of variation (CV) of site-to-site measurements decreased by using higher laser focusing diameters. By fixing the laser spot diameter at 750 \'mü\'m, the CV of measurements improved at least 2-fold for all analytes when 50 J cm-2 was applied in comparison to 20 J cm-2. These results can be attributed to a larger mass removal obtained at higher laser fluences and better laser sampling representativeness. Sensitivity did also increase with laser fluence but no significant differences were observed in the detection limits using 25 and 50 J cm-2. Although emission intensities increased with laser fluence the background emission and noise increased as well. On the other hand, particle size distribution of powdered samples affects the atomization and excitation efficiencies by laser induced plasmas. When larger particles are incompletely vaporized the resulting emission signals were lower than those obtained from smaller particles. In this work, significant improvements on emission intensities and CV of measurements were observed when pellets made with particles < 75 \'mü\'m were analyzed. No significant differences were observed (emission intensities and CV of measurements) in the analysis of pellets prepared with particles in the 20-75 \'mü\'m range. Pellets prepared with particles > 75 \'mü\'m presented cohesion drawbacks and resulted non uniform craters after laser ablation. The CV of measurements for most analytes in these materials was > 15 %. Therefore, similar particle size distribution between calibration and test samples is recommended to avoid physical matrix effects on laser-sample interaction

A cultura da cana-de-açúcar no Brasil é uma das atividades mais relevantes na agroindústria nacional devido ao alto potencial de produção e elevada importância econômica de seus subprodutos como e.g. açúcar e etanol. Embora o Si não seja considerado um elemento essencial, a adubação silicatada tem proporcionado aumentos apreciáveis na produtividade da cana-de-açúcar e na resistência a algumas pragas e doenças. Ainda não existem métodos bem estabelecidos para a determinação de Si em material vegetal visando à diagnose foliar. O objetivo deste trabalho foi desenvolver um método para determinação de Si em pastilhas de folhas de cana-de-açúcar por espectrometria de emissão óptica com plasma induzido por laser (LIBS). O arranjo experimental foi composto por um laser pulsado de Nd:YAG a 1064 nm (pulsos de 5 ns, 360 mJ, 10 Hz) e os sinais de emissão foram coletados por um telescópio acoplado por fibra óptica ao espectrômetro Echelle com detector ICCD. Amostras laboratoriais consistiram de pastilhas preparadas com folhas de cana-de-açúcar moídas criogenicamente por 40 min. Melhores resultados foram obtidos quando a linha de emissão Si I 212,412 nm foi selecionada e os parâmetros instrumentais foram ajustados em 25 pulsos acumulados, 50 J cm-2, diâmetro de focalização de 750 \'mü\'m, 2,0 \'mü\'s de atraso e 4,5 \'mü\'s de integração. Os resultados obtidos por LIBS na análise de pastilhas de 20 amostras laboratoriais foram comparados com os obtidos por espectrometria de emissão óptica com plasma acoplado indutivamente (ICP OES) após decomposição alcalina e por espectrometria de micro-fluorescência de raio-X por energia dispersiva (\'mü\'ED-XRF) após análise direta das pastilhas. Aplicando-se o teste t de Student pareado, não foram verificadas diferenças significativas ao nível de 95% de confiança nas determinações de Si por LIBS, ICP OES e \'mü\'ED-XRF. Os resultados obtidos nesta tese indicam a viabilidade de LIBS para determinação direta de Si em pastilhas de folhas de cana-de-açúcar e abrem uma expectativa para futuras explorações de LIBS em análises em campo / The high production of sugarcane and its subproducts (e.g. sugar and ethanol fuel) make this crop one of the most important for the Brazilian agro-industrial economy. Although Si is not considered an essential element, the Si-based fertilizers have provided considerable improvement in productivity of sugarcane and resistance to some pests and diseases. At moment, there are no well-established methods for the determination of Si in plant materials aiming at nutritional diagnosis. The aim of this work was to develop a method for Si determination in pellets of sugarcane leaves by laser-induced breakdown spectroscopy (LIBS). The experimental setup was designed by using a Q-switched Nd:YAG laser (5 ns, 360 mJ, 10 Hz, \'lâmbda\' = 1064 nm) and the emission signals were collected by lenses into an optical fiber coupled to an Echelle spectrometer equipped with an ICCD. Pellets prepared from cryogenically ground sugarcane leaves were used as laboratory samples. Best results were obtained when the Si I 212.412 nm emission line was selected and the instrumental parameters were adjusted at 25 laser shots, 50 J cm-2, 750 \'mü\'m laser spot size, 2.0 \'mü\'s delay time and 4.5 \'mü\'s integration time gate. The results obtained by LIBS in the analysis of pellets of 20 laboratory samples were compared with those from inductively coupled plasma optical emission spectrometry (ICP OES) after alkaline digestion and by microenergy dispersive X-ray fluorescence (\'mü\'ED-XRF) after direct analysis of pellets. By applying a paired t-test at 95 % confidence level, there was no significant difference in the Si determination by LIBS, ICP OES and \'mü\'ED-XRF. The findings of this thesis indicate the feasibility of LIBS for the direct determination of Si in pellets of sugarcane leaves and open expectations for future investigations of LIBS for field analysis

Nos últimos 20 anos, a espectroscopia de emissão ótica com plasma induzido a laser (LIBS) tem se tornado uma das mais promissoras ferramentas em química analítica, cujas aplicações são destinadas às análises multi-elementares em amostras nos estados sólido, líquido e gasoso. Suas aplicações são as mais diversas, pois sua instrumentação é relativamente simples e pode ser portátil. Em especial, a LIBS apresenta grande potencial de uso na agricultura, com diversas publicações que exploram, principalmente, a concentração de carbono em solos. Entretanto, a construção de modelos gerais de calibração é uma das maiores dificuldades da técnica, pois ela está suscetível aos efeitos de matriz que adicionam comportamentos não-lineares às intensidades das emissões. Com intuito de compreender quais são as principais causas dessa dependência, este trabalho avaliou relações entre propriedades físicas do plasma LIBS e as inclinações das curvas de calibração para cinco emissões de carbono em amostras sintéticas de diferentes potenciais de ionização, sendo elas KCl e H3BO3 com 1% de CuSO4 e concentrações entre 0 e 10 % de carbono. Constatou-se que temperatura e densidade de elétrons dependem inversamente da concentração de carbono nas amostras, o que provavelmente está associado às taxas de ablação de material da amostra e recombinação no interior do plasma. Ademais, esses parâmetros em plasmas originados das amostras de KCl + CuSO4 são maiores do que para a matriz de H3BO3 + CuSO4, além de que apenas plasmas originados da primeira matriz satisfizeram o critério de McWhirter para C I, Cu I e Cu II em todas as concentrações de carbono, indicando que eles devem apresentar condições de equilíbrio termodiâmico local. O mesmo não foi obtido para o caso da matriz de H3BO3 + CuSO4, o que pode explicar suas elevadas incertezas nos valores da temperatura e densidade de elétrons. Verificou-se ainda que as inclinações das curvas de calibração para o carbono foram maiores para o caso da matriz com elementos majoritários de menor potencial de ionização e que isso está diretamente relacionado às propriedades intrínsecas dos plasmas originados, o que corrobora a hipótese deste trabalho. Além disso, cálculos de correlação entre a área do pico de carbono em 247,89 nm e cada ponto dos espectros LIBS mostraram que a emissão de Cu II em 224,72 nm linearizou as curvas de calibração, minimizando os efeitos de matriz. Por fim, este trabalho contribuiu tanto com avanços no conhecimento até então existente a respeito dos efeitos de matriz quanto com um eficiente conjunto de técnicas analíticas para espectros, cujo potencial de aplicação é enorme. / In the last 20 years, laser-induced breakdown spectroscopy (LIBS) has been one of the main tools in analytical chemistry, whose applications are destinated to multi-elementar analysis in solid, liquid or gaseous samples. Its use is diverse, since its instrumentation is relatively simple and can be portable. In particular, LIBS has a great potential for use in agriculture, with many publications that explore mainly the carbon concentration in soils. However, the construction of general calibration models is one of the greatest difficulties of the technique, since it is susceptible to matrix effects that add non-linear behaviors to the emission intensities. In order to understand the main causes of this dependence, this study evaluated the relationships between physical properties of the LIBS plasma and the slope of the calibration curves for carbon emissions in synthetic samples with different ionization potentials. Those samples were made of KCl and H3BO3 with 1% of CuSO4 and carbon concentrations ranging from 0 to 10%. It was found that the plasma temperature and electron density are inversely proportional to the carbon concentration in the samples, which is probably associated with rates of sample ablation and recombination within the plasma. In addition, these parameters in plasmas originating from the samples of KCl + CuSO4 are larger than those originating from the matrix of H3BO3 + CuSO4 at all carbon concentrations, indicating that the first matrix must be in local thermodynamics equilibrium. However, similar results were not found for the other matrix which may explain its high uncertainties in the values of temperature and electron density. It was verified that the slope of the carbon calibration curves were higher for the matrix with elements of low ionization potential than for the other matrix and that this is directly related to the intrinsic properties the plasmas originated, which corroborates the hypothesis of this study. Furthermore, the analysis of correlation between the carbon peak area at 247.89 nm and each point of the LIBS spectra have shown that the Cu II emission at 224.72 nm linearized the calibration curves, minimizing the matrix effects. Finally, this study contributed with advances in the knowledge about matrix effects and with an efficient set of analytical tools for spectra whose application potential is huge.

A soja e os seus derivados são uma das mercadorias agrícolas mais valiosas e comercializadas no mundo. O Brasil é o segundo maior produtor mundial de soja, cujo complexo é o principal gerador de divisas cambiais para o país. Porém, um dos principais fatores que podem limitar os rendimentos da cultura e a qualidade dos produtos é a presença de doenças. Para evitar maiores prejuízos é importante identificar rapidamente qual doença acomete a planta e determinar o método mais eficaz de manejo. Atualmente, uma doença recém descoberta que causa preocupação aos produtores é a infestação pelo nematoide Aphelenchoides besseyi, causador da, popularmente conhecida, Soja louca II. Ainda não existem métodos que proporcionem um diagnóstico preciso para a doença, sendo este realizado por inspeção visual da plantação. Sabendo que plantas sadias e doentes apresentam perfis nutricionais diferentes, este trabalho vem propor o uso da técnica analítica LIBS, associada ao método livre de calibração (CF), como uma ferramenta alternativa para diagnóstico da doença. A técnica LIBS permite análises multielementares simultâneas, de maneira rápida e de baixo custo. Já o método CF tem por principal vantagem dispensar o emprego de qualquer padrão de matriz compatível, ou material de referência, para as quantificações em LIBS, fazendo-o com base nas características físicas do plasma formado. Assim, a associação de ambos se torna absolutamente adequada para determinações analíticas em amostras desconhecidas. Neste trabalho, foram conduzidos estudos de caracterização qualitativos, por meio de um sistema LIBS de baixa resolução, e avaliações semiquantitativa e quantitativa por meio de medidas LIBS em um sistema com detecção de alta resolução (DP LIBS). O primeiro estudo demonstrou que, qualitativamente, as variações nos macroelementos Ca, Mg e K presentes nas folhas de soja, permitiu alcançar, em média, 90% de acurácia na diferenciação entre amostras doentes e sadias, e inferir que a doença provoca relativa diminuição da concentração de Ca e Mg, e um pequeno aumento nos níveis de K. As quantificações obtidas pelo método CF foram consideradas satisfatórias, além de pioneiras nas análises de amostras de folhas de soja puras. A partir dos perfis nutricionais quantificados foram construídos classificadores (classificação via regressão associada à regressão por mínimos quadrados parciais), cuja acurácia alcançada foi de 94%, 92% e 90%, respectivamente para os dados de concentração obtidos pela técnica de referência, e DP LIBS+CF nas geometrias colinear e ortogonal. Na tentativa de eliminar o emprego de qualquer reagente, e ainda assim obter uma ferramenta diagnóstica precisa, foi conduzido um estudo semiquantitativo, também a partir da associação DP LIBS+CF. Baseado nas relações elementares entre os analitos que compõem a amostra, este apresentou acurácia diagnóstica superior a 90% para a geometria ortogonal. Portanto, este estudo permitiu o desenvolvimento de uma metodologia para diagnóstico que respeita 100% dos preceitos da Química Verde, e agrega muito ao sistema atual, quando o que se deseja é uma avaliação diagnóstica rápida para tomada de decisão imediata no campo. / Soybeans and their by-products are one of the most valuable and traded commodities in the world. Brazil is the second largest soybean producer in the world, whose soy complex is the main generator of foreign exchange for the country. However, one of the main factors that can limit crop yields and product quality is the occurrence of disease. To avoid further damage, it is important to quickly identify which disease affects the plant and determine the most effective method of management. Currently, a disease newly discovered and of concern to producers is the infestation by the nematode Aphelenchoides besseyi, which causes the disease known as GSFR (Green Steam and Foliar Retention). There are still no methods to provide a precise diagnosis for the disease, which is, nowadays, performed by visual inspection of the plantation. Knowing that healthy and diseased plants have different nutrient profiles, this work proposes the use of the LIBS technique, associated with the calibration free method (CF), as an alternative tool for the disease detection. The LIBS technique allows simultaneous multi-element analysis, quickly and inexpensively. The CF method has the main advantage of avoiding the use of any compatible standard matrix, or reference material, for LIBS quantifications, based on the physical characteristics of the plasma formed. Thus, the association of both becomes absolutely suitable for the analytical determination of unknown samples. Qualitative characterization studies using a low-resolution LIBS system, and semiquantitative and quantitative analysis using LIBS measurements in a high-resolution detection system (DP LIBS), were performed. The first study showed that it was possible to achieve, on average, 90% accuracy in the differentiation between diseased and healthy samples, and to infer that the disease provokes relative decrease in Ca and Mg concentration, and a small increase in K levels. Quantifications achieved by the CF method were considered satisfactory since it is a pioneer study in the analysis of pure soybean leaves. Classifiers were constructed from the nutritional profiles quantified by CF, applying classification via regression associated with partial least square regression, with accuracy of 94%, 92% and 90%, respectively, for the concentration data obtained by the reference technique, and DP LIBS + CF colinear and orthogonal. In attempt to eliminate the use of any chemical reagent and still obtain an efficient diagnostic tool, a semiquantitative study was conducted, also from the DP LIBS + CF association and based on the elementary relations between the analytes that make up the sample, with accuracy diagnosis greater than 90% for DP LIBS in orthogonal geometry. Therefore, this study allowed the development of a diagnostic methodology that respects 100% of green chemistry principles, and adds a lot to the current system, if the desire is a rapid diagnostic evaluation for immediate decision making in the field.

La contamination par métaux des eaux est un enjeu de santé publique important. Des contrôles et traitements de plus en plus drastiques sont ainsi effectués sur les eux destinés à la consommation humaine. Il est indispensable, pour cela, de posséder des outils analytiques fiables et sensibles, adaptés aux réglementations existantes et suffisamment souples d'utilisation. La technique de « Laser Induced Breakdown Spectroscopy » (LIBS), ayant fait ses preuves pour l’analyse des solides, y compris dans les explorations exo-terrestres, présente des avantages très intéressants pour les liquides dont, par exemple, sont caractère multi-élémentaire et la possibilité des mesures in-situ de la contamination des eaux par métaux.Un première volet d’étude de ce travail de thèse à permis d'étudier les potentialités de la spectroscopie sur plasma induit par laser (LIBS) pour l'analyse des métaux dissous ou en suspensions dans les eaux. Etude qu’a permis de constater un effet important lié à la taille des particules lors de l’analyse de suspensions. Un deuxième volet consisté à regarder les effets de matrice organique représentée par un acide humique et ceux des minéraux naturels représentés par la bentonite et des particules d’alumine. Effets matrice qu’ont étais corriges par une normalisation par étalon interne. / Metal contamination of water is a major public health issue. Controls and treatments are more drastic and performed on them for human consumption. It is essential for this to possess reliable and sensitive analytical tools adapted to the existing regulations and flexible enough to use. The technique of "Laser Induced Breakdown Spectroscopy" (LIBS), proven for the analysis of solids, including exo-terrestrial explorations, this very interesting advantages for liquids including, for example, are multi-character elementary and the possibility of in-situ measurements of water contamination by metals.A first part of this study thesis allowed to explore the potential (LIBS) for the analysis of dissolved metals or suspensions in water. Study has found an important effect related to the particle size in the analysis of suspensions. A second prong is to look at the effects of organic matrix represented by humic acid and those natural minerals represented by the bentonite and alumina particles. The matrix effect observed was corrected by normalization by internal standard.

Ce travail de thèse concerne l’utilisation des décharges électriques de haute tension (DEHT) en milieu aqueux comme méthode d’extraction des polyphénols à partir de pépins de raisin.Les arcs électriques produits en milieu aqueux provoquent une succession de phénomènes (ondes de choc, bulles de cavitation) qui ont pour effet de fragmenter toute matière première située à proximité de l’arc électrique. L’objectif de cette thèse est d’étudier ces phénomènes afin d’améliorer la compréhension et l’efficacité des DEHT en tant que méthode d’extraction.Dans un premier temps, une étude des conditions d’apparition de l’arc électrique dans l’eau a permis de montrer que l’arc apparaît initialement dans des bulles de vapeur générées à la surface de l’électrode à cause de l’échauffement du liquide par effet Joule. Des mesures électriques, des prises de vues à haute vitesse, ainsi qu’une simulation numérique du problème ont permis de vérifier cette hypothèse. Une étude paramétrique des phénomènes générés par l’arc électrique (onde de choc et bulle de cavitation) a été menée. Grâce à des mesures de la pression des ondes de choc, des mesures de la taille des bulles de cavitation, et grâce des mesures électriques précises (notamment de la résistance électrique de l’arc), il apparaît que l’amplitude des phénomènes dépendent essentiellement de l’énergie dépensée dans l’arc. Cette énergie doit être distinguée de l’énergie totale d’une impulsion électrique, dont une partie est dépensée avant le claquage, mais également de l’énergie disponible au moment du claquage, dont une partie importante est dépensée dans le circuit électrique. La partition de cette énergie entre l’arc et le circuit électrique dépend du rapport entre la résistance du circuit et la résistance de l’arc. Ainsi une méthode pour augmenter significativement l’amplitude des phénomènes étudiés, et donc l’efficacité du procédé est d’améliorer le rapport entre ces résistances. Il a par exemple été observé qu’en augmentant la longueur de l’arc électrique de 2.5 mm à 2 cm, la résistance de l’arc augmente de 40 m à 0.55, et l’amplitude de l’onde de choc augmente de 135%. Pour finir, une étude paramétrique sur l’efficacité des DEHT comme procédé d’extraction des polyphénols des pépins de raisin a été menée. Entre autres, les effets sur l’extraction de la conductivité du liquide, du rapport liquide-solide, du nombre d’impulsion, de l’énergie par impulsion, de la distance inter-électrode ont été étudiés. Ces études ont mis en évidence l’importance de la répartition de l’énergie totale d’une impulsion en énergie dépensée avant le claquage, énergie dépensée au claquage dans le circuit électrique et énergie dépensée dans l’arc électrique. Ces études ont montré comment cette répartition est influencée par ces différents paramètres, et comment cela influence l’efficacité d’extraction. L’influence de la distance inter-électrode, et donc de la longueur de l’arc, a été particulièrement été mise en évidence par les résultats d’extraction. / This thesis work focuses on the use of high voltage electrical discharges (HVED) in aqueous media as a method for extracting polyphenols from grape seeds. Electric arcs generated in an aqueous environment cause a succession of phenomena (shock waves, cavitation bubbles) that have the effect of fragmenting any raw material located near the electric arc. The objective of this thesis is to study these phenomena in order to improve the understanding and effectiveness ofHVED as an extraction method. First, a study of the conditions under which the electric arc appears in water showed that the arc initially appears in vapour bubbles generated on the electrode surface due to the heating of the liquid due to Joule effect. Electrical measurements, high-speed photography and a numerical simulation of the problem have allowed this hypothesis to be verified. A parametric study of the phenomena generated by the electric arc (shock wave and cavitation bubble) was carried out. Through measurements of shock wave pressure, of cavitation bubble size, and precise electrical measurements (including the electrical resistance of the arc), it appears that the amplitude of the phenomena depends essentially on the energy consumed in the arc. This energy mustbe distinguished from the total energy of an electrical pulse, part of which is spent before the breakdown. The energy spent in the electric arc must also be distinguishedfrom the energy available at electrical breakdown, as a significant part of breakdown energy is spent in the electrical circuit. The partition of breakdown energy between the arc and the electrical circuit depends on the ratio between the resistance of the circuit and that of the arc. Thus a method to significantly increase the amplitude of the studied phenomena (and therefore the efficiency of the process), is to improve the ratio between these resistances. For example, it has been observed that by increasing the length of the electric arc from 2.5 mm to 2 cm, the resistance of the arc increases from 40 m to 0.55, and the amplitude of the shock wave increases by 135%. Finally, a parametric study on the efficiency of DEHT as a process for extracting polyphenols from grape seedswas carried out. Among other things, the effects on the extraction of liquid conductivity, liquid-solid ratio, number of pulses, energy per pulse, and distance between electrodes were studied. These studies highlighted the importance of the distribution of the total pulse energy into energy spent before the breakdown, energy spent after breakdown in the electrical circuit and energy spent in the arc. These studies have shown howthis distribution is influenced by these different parameters, and how it influences extraction efficiency. The influence of the inter-electrode distance, and therefore the length of the arc, was particularly highlighted by the extraction results.

Actuellement, la majorité des composants à semi-conducteur pour l'électronique de puissance est réalisée à partir de silicium. Afin de répondre aux nouvelles contraintes électriques et thermiques imposées par la montée en tension et en densité de puissance des convertisseurs d'énergie électrique, une solution repose sur l'emploi d'un semi-conducteur à large bande interdite tel que le carbure de silicium (SiC), du fait de son champ électrique critique (EC) environ dix fois plus élevé que celui du silicium et de sa capacité à fonctionner à des températures supérieures à 200 °C. Une revue des nombreuses publications concernant des diodes en SiC de tenue en tension élevée, voire leur disponibilité commerciale (jusqu'à 10 kV), est présentée, qui montre les progrès réalisés grâce aux efforts portés sur l'amélioration du matériau et l'optimisation de la protection périphérique des composants. L'étape de passivation primaire reste une étape critique très souvent mentionnée. Beaucoup moins de travaux prennent en considération l'impact des matériaux de passivation secondaire et d'encapsulation. L'objectif de cette thèse est de contribuer à une meilleure connaissance des paramètres et des mécanismes de rupture impactant la tenue en tension à l'état bloqué de l'ensemble que forment la puce et son l'environnement isolant électrique. Ainsi, une étude expérimentale de l'influence de différents paramètres liés au semi-conducteur ainsi qu'aux matériaux de passivation et d'encapsulation présents en surface de la puce a été menée, à l'aide de diodes en SiC-4H avec protection périphérique par poche implantée, réalisées par la société IBS, dans le cadre du projet de recherche 'FilSiC'. Dans un premier temps, une étude par simulation numérique de l'ensemble de la structure (SiC, électrodes métalliques, isolants) a été effectuée à l'aide du logiciel Sentaurus Device (Synopsys). Elle a permis de quantifier les contraintes en champ électrique dans toute la structure pour une tension appliquée donnée, et leur sensibilité aux caractéristiques des matériaux isolants prises en compte. Cette étude a également servi au choix des paramètres liés au substrat épitaxié et à la géométrie latérale et en surface des diodes (les paramètres technologiques étant fixés par ailleurs), pertinents pour l'étude expérimentale de leur tension de rupture, dans une gamme de 1 kV à 6 kV. En parallèle, la caractérisation électrique, au sein de structures Métal-Isolant-Semi-conducteur, du matériau de passivation primaire utilisé (dépôt épais de dioxyde de silicium), dans une gamme de température jusqu'à 300 °C, a permis de déterminer ses propriétés électriques, dont la rigidité diélectrique. Le travail a ensuite porté sur la caractérisation à température ambiante de la tension de rupture de la structure complète des différentes diodes fabriquées, effectuée sous vide et sous azote à la pression atmosphérique. Les comportements expérimentaux visualisés sous vide au moment du claquage, et couplés aux informations issues des simulations, ont notamment permis d'estimer les valeurs des champs maximaux induits dans les différents matériaux isolants, et de corréler leur impact avec les valeurs de rigidité diélectrique connues pour ces isolants. Les résultats complémentaires sous azote ont permis de confirmer certains paramètres technologiques et mécanismes mis en jeu lors du claquage des diodes d'autre part. Plusieurs conclusions utiles pour l'optimisation des paramètres technologiques (épitaxie et poche) et des couches isolantes de passivation et d'encapsulation (épaisseur, permittivité) de la diode 'haute tension' en SiC ont pu être dégagées de ces travaux. / Nowadays, most of the semiconductor devices used in power electronics are silicon (Si) based devices. In order to meet the new electrical and thermal constraints induced by the demand in rising both the operating voltage and the power density of the electrical energy converters, the use of wide band gap semiconductors such as silicon carbide (SiC) may represent an adequate solution, thanks to their critical electric field (EC) which is about ten times higher than that of Si and to their ability of operating at temperatures beyond 200 °C. A state-of-the-art on the readily (commercially) available high voltage SiC diodes (10kV or higher) is presented, highlighting the progress made in improving the materials themselves as well as their peripheral protection. However, regarding the die insulating materials, the studies mainly focus on the primary passivation step, which was often mentioned as the most critical one. Obviously much less work is carried out on the impact of the secondary passivation and encapsulation materials. The goal of this study is to contribute to a better knowledge of the mechanisms involved in the SiC chips and electrically insulating environment breakdown while in a blocking state, as well as to the identification of the most relevant parameters acting on these phenomena. Thus, a study of the correlated properties of the semiconductor and the passivation and encapsulation materials present on the surface of the chip was carried out on SiC-4H bipolar diodes protected by junction termination extension (JTE), supplied by IBS society, within the framework of the 'FilSiC' research project. First, a numerical simulation study of the entire structure (SiC and insulating materials) was carried out using the Sentaurus Device software (Synopsys). This allowed for quantifying the electric field stresses throughout the structure for a given applied voltage as well as their dependence on the properties of the considered insulating materials. This study was equally used for choosing the appropriate parameters of the epitaxial substrate and for designing the lateral and the surface geometry of the diodes (the technological parameters being fixed elsewhere), in view of the subsequent experimental study of their breakdown voltage in the 1kV-6kV range. Concurrently, the electrical characterization of the primary passivation material used (thick silicon dioxide layer) was performed by using MIS (metal-insulator-semiconductor) type structures, in a temperature range of up to 300 °C. This allowed to determine its main electrical properties, particularly the dielectric breakdown voltage. The work then focused on the room-temperature characterization of the breakdown voltage of the full structures built around the different manufactured diodes. The tests were carried out both under high vacuum conditions and under nitrogen at atmospheric pressure. The behavior of the different study cases observed under vacuum conditions during the breakdown, coupled with the simulation results, allowed to determine the values of the maximum electric fields induced in the different insulating materials and to correlate them to their known dielectric breakdown values. On the other hand, additional results obtained under nitrogen atmosphere confirmed some technological parameters and mechanisms at play during the breakdown of the diodes. Several guidelines for the optimization of these technological parameters (epitaxy and JTE) and of the insulating passivation and encapsulation layers (thickness, permittivity) of the "high voltage" SiC diode were derived from this study.