The FP interferometer was found to be feasible for detecting changes in machining forces. The fiber optic sensor was able to detect increases in strain corresponding to force increases detected by a dynamometer. The FP interferometer system must progress in several ways. A better data acquisition and data analysis system must be developed. A robust sensor must be made to withstand the harsh environment of machining. Also a method for eliminating the affects of thermal strain must be created. Finally, the placement of the FP sensor must also be determined. The FP has the potential to effectively monitor machining forces without affecting the rigidity of a turning operation setup. / Master of Science

Interferometric optical fiber sensors have proved many orders of magnitude more sensitive than their electrical counterparts, but they suffer from limitations in signal demodulation caused by phase ambiguity and complex fringe counting when the output phase difference exceeds one fringe period. Various signal demodulation methods have been developed to overcome some of the these drawbacks with limited success. This thesis proposes a new measurement system for the extrinsic Fabry-Perot interferometer (EFPI) sensor. Using a wavelength modulated source and a novel extended-gap EFPI, some of the limitations of interferometric signal demodulation are overcome. By scanning the output wavelength of a multilongitudinal mode laser diode through current modulation, the EFPI sensor signal is scanned through multiple fringes. Gap movement is then unambiguously determined by monitoring the phase of the multiple fringe pattern. / Master of Science

An adaptation of an extrinsic Fabry-Perot interferometer (EFPI) strain sensor is described, which permits the state of cure of an epoxy matrix to be monitored, when the sensor is embedded in a polymeric matrix composite. By using a glass rod with a retroreflecting end for the target fiber in the EFPI sensor, the intensity of the light reflected depends on the refractive index of the host matrix, if a low coherence source is used. As the epoxy cross-links during cure, the refractive index of the epoxy will increase to a value exceeding that of the target fiber. The resulting increased loss in the fiber can be detected at the sensor output and correlated to the state of cure of the epoxy. After cure, the sensor may be operated as a conventional extrinsic Fabry-Perot interferometric strain sensor if a coherent source is used. Using the modified extrinsic Fabry-Perot sensor, we monitor the cure of Devcon® 5-Minute® Epoxy, and show that it cures in approximately 60 minutes. / Master of Science

Linewidth is a measure of the frequency stability of any kind of oscillator, and it is a defining characteristic of coherent lasers. Narrow linewidth laser technology, particularly in the field of fiber-based infrared lasers, has progressed to the point where highly stable sources are commercially available with linewidths on the order of 1-100 kHz. In order to achieve a higher level of stability, the laser must be augmented by an external frequency stabilization system. This paper presents the design and operation of a frequency locking system for infrared fiber lasers. Using the Pound-Drever-Hall technique, the system significantly reduces the linewidth of an input laser with an un-stabilized linewidth of 2 kHz. It uses a high-finesse Fabry-Perot cavity, which is mechanically and thermally isolated, as a frequency reference to measure the time-varying frequency of the input laser. An electronic feedback loop works to correct the frequency error and maintain constant optical power. Testing has proven the Pound-Drever-Hall system to be highly stable and capable of operating continuously for several seconds at a time. / Master of Science

Ce travail présente une caractérisation de la turbulence dans la région du Trapèze de la nébuleuse d’Orion. Les raies observées sont Hα, [OIII] et le doublet du [SII], ce dernier permettant aussi d’obtenir la densité électronique. Le traitement de données d’interférométrie de Fabry-Pérot permet d’obtenir une carte des vitesses radiales dans le gaz pour chaque raie. Une technique d’homogénéisation présentée par Lagrois & Joncas (2011) utilisant un filtre de type Zurflueh a permis de mettre en évidence les fluctuations de vitesse provenant de la turbulence. L’analyse statistique des vitesses à l’aide des fonctions d’autocorrélation et de structure ont permis d’obtenir l’exposant de Kolmogorov γ pour chacun des ions. En Hα et en [OIII], l’exposant (~ 1/2) est typique d’une turbulence supersonique avec ondes de choc tandis qu’en [SII], l’exposant (~ 1/3) est typique d’une turbulence subsonique sans choc. De plus, l’étude de l’histogramme des fluctuations de vitesse a permis de localiser des zones où on retrouve de l’intermittence dans la turbulence.

Dans cette thèse, nous initions la démonstration, à échelle réduite, de la faisabilité du photodétachement presque total, par laser, d’un jet d'ions négatifs d'hydrogène en cavité optique Fabry-Perot pour les futurs injecteurs de neutres destinés au chauffage des plasmas des réacteurs de fusion nucléaire.Nous élaborons une nouvelle méthode de mesure d’une section efficace de photodétachement, dont la connaissance à la longueur d’onde d’excitation est requise pour le dimensionnement de la cavité Fabry-Perot, basée sur l’observation de la saturation en régime d’éclairement impulsionnel. Le calcul analytique de l’accroissement du signal de détachement produit lors de l’éclairement d’un jet d’ions par une impulsion laser supposée gaussienne, fait apparaître une contrainte mathématique sur le flux requis pour transiter vers le régime saturé. Cette contrainte est une caractéristique de la transition vers la saturation pour toutes les expériences réalisées en faisceau gaussien et pour tous les processus d’interaction lumière-matière linéaires. Avec cette méthode, nous déduisons une section efficace de photodétachement de H- à 1064 nm – longueur d’onde sélectionnée pour les futurs injecteurs de neutres – en léger désaccord avec les prédictions théoriques.Pour réduire les exigences technologiques sur la cavité et le laser, nous étudions les résonances de Landau qui apparaissent dans le spectre de photodétachement en champ magnétique. S’asservir sur l’une de ces résonances permettrait d’augmenter la probabilité de photodétachement à un flux donné.Par ailleurs, nous présentons nos mesures des affinités électroniques du phosphore, du sélénium et de l’étain, réalisées avec le microscope de photodétachement. L'expérience de microscopie de photodétachement du phosphore est la première où l’atome neutre est laissé dans un terme excité. / In this thesis, we initiate the demonstration, at a reduced scale, of the feasibility of the almost-complete laser photodetachment of negative hydrogen ion beams in a Fabry-Perot optical cavity for future neutral beam injectors for the heating of fusion power plants plasmas.We develop a new method to measure a photodetachement cross section, the knowledge of which at the exciting wavelength is needed to scale the Fabry-Perot cavity, based on the observation of the saturation in a pulsed lighting regime. The analytical calculus of the detachment signal growth produced while illuminating a negative ion beam with a Gaussian laser pulse bring out a mathematical constraint on the required flux to pass through the saturated regime. This constraint is the signature of the transition toward the saturation for all experiment carried out in Gaussian beam and for all linear light-matter interaction processes. With this method, we measure the photodetachement cross section of H- at 1064 nm – selected wavelength for future neutral beam injectors – in slight disagreement with theoretical predictions.To reduce the technological requirement on the Fabry-Perot cavity and the laser, we study Landau resonances which appear in the photodetachement spectrum. Locking the laser on one of these resonances would allow increasing the photodetachment probability at a given flux.We also present our phosphorus, selenium and tin electron affinity measurements carried out with the photodetachment microscope. The photodetachment microscopy experiment of phosphorus is the first one where the neutral atom is left in an excited term.

L’imagerie hyperspectrale, qui consiste à acquérir l'image d'une scène dans un grand nombre de bandes spectrales, permet de détecter des cibles là où l'imagerie couleur classique ne permettrait pas de conclure. Les imageurs hyperspectraux à acquisition séquentielle sont inadaptés aux applications de détection en temps réel. Dans cette thèse, nous proposons d’utiliser un imageur multispectral snapshot, capable d’acquérir simultanément un nombre réduit de bandes spectrales sur un unique détecteur matriciel. Le capteur offrant un nombre de pixels limité, il est nécessaire de réaliser un compromis en choisissant soigneusement le nombre et les profils spectraux des filtres de l'imageur afin d’optimiser la performance de détection. Dans cet objectif, nous avons développé une méthode de sélection de bandes qui peut être utilisée dans la conception d’imageurs multispectraux basés sur une matrice de filtres fixes ou accordables. Nous montrons, à partir d'images hyperspectrales issues de différentes campagnes de mesure, que la sélection des bandes spectrales à acquérir peut conduire à des imageurs multispectraux capables de détecter des cibles ou des anomalies avec une efficacité de détection proche de celle obtenue avec une résolution hyperspectrale. Nous développons conjointement un démonstrateur constitué d'une matrice de 4 filtres de Fabry-Perot accordables électroniquement en vue de son implantation sur un imageur multispectral snapshot agile. Ces filtres sont développés en technologie MOEMS (microsystèmes opto-électro-mécaniques) en partenariat avec l'Institut d'Electronique Fondamentale. Nous présentons le dimensionnement optique du dispositif ainsi qu'une étude de tolérancement qui a permis de valider sa faisabilité. / Hyperspectral imaging, which consists in acquiring the image of a scene in a large number of spectral bands, can be used to detect targets that are not visible using conventional color imaging. Hyperspectral imagers based on sequential acquisition are unsuitable for real-time detection applications. In this thesis, we propose to use a snapshot multispectral imager able to acquire simultaneously a small number of spectral bands on a single image sensor. As the sensor offers a limited number of pixels, it is necessary to achieve a trade-off by carefully choosing the number and the spectral profiles of the imager’s filters in order to optimize the detection performance. For this purpose, we developed a band selection method that can be used to design multispectral imagers based on arrays of fixed or tunable filters. We use real hyperspectral images to show that the selection of spectral bands can lead to multispectral imagers able to compete against hyperspectral imagers for target detection and anomaly detection applications while allowing snapshot acquisition and real-time detection. We jointly develop an adaptive snapshot multispectral imager based on an array of 4 electronically tunable Fabry-Perot filters. The filters are developed in MOEMS technology (Micro-Opto-Electro-Mechanical Systems) in partnership with the Institut d'Electronique Fondamentale. We present the optical design of the device and a study of tolerancing which has validated its feasibility.

A doença de Fabry (DF, OMIM 301500) é uma doença lisossômica de depósito de herança ligada ao cromossomo X causada por mutações no gene GLA que levam à tradução da enzima lissomal α-galactosidase A (α-gal A, EC 3.2.1.22) com atividade deficiente ou nula. Como consequência, os substratos dessa enzima (majoritariamente globotriaosilceramida – Gb3 - e globotriaosilesfingosina – liso-Gb3) acumulam-se em diversos tecidos e fluidos corporais dos pacientes com DF. Estudos indicam que inflamação, estresse oxidativo e nitrosativo podem estar envolvidos na fisipatologia da DF, que ainda não está completamente esclarecida. Foi descrito previamente um aumento de parâmetros inflamatórios e próoxidantes em pacientes com DF durante a terapia de reposição enzimática (TRE), todavia se faz necessário avaliar tais parâmetros antes do início da TRE. Através de ensaios in vivo e in vitro em amostras de sangue e de urina de pacientes com DF, avaliamos alterações de estado redox, dano oxidativo a biomoléculas (incluindo lipídeos, proteínas e DNA) e níveis urinários de Gb3. No intuito de investigar os efeitos biológicos do mais recente biomarcador proposto para a DF, o liso-Gb3, realizamos ensaios in vitro em células embrionárias de epitélio renal humano (HEK-293T). Demonstramos, pela primeira vez, aumento de dano ao DNA em pacientes com DF. Ainda, observamos aumento de geração de espécies reativas nas mesmas amostras (através da sonda diclorofluoresceina - DCF) e, através do ensaio cometa com endonucleases, verificamos que o dano ao DNA é de natureza oxidativa em purinas nesses pacientes. Para avaliar a capacidade de reparo de DNA frente a um insulto oxidativo, realizamos o ensaio de desafio com peróxido de hidrogênio (H2O2), o qual demonstrou que os pacientes possuiam reparo aumentado (provavelmente como resposta ao estresse crônico), porém não suficientemente eficaz para reduzir o dano oxidativo basal a purinas a nível de indivíduos saudáveis. Observamos também que pacientes com DF antes de iniciar a TRE já possuíam níveis aumentados de lipoperoxidação (através da medida de espécies reativas ao ácido tiobarbitúrico – TBARS - e malondialdeído – MDA) e de estresse nitrosativo (medido através de nitrato e nitrito urinários), bem como desequilíbrios no sistema redox da glutationa (GSH - principal antioxidante intracelular, avaliado através da medida de GSH e da atividade das enzimas glutationa redutase, GR, e glutationa peroxidase, GPx). Após a TRE de longa duração (aproximadamente cinco anos), a lipoperoxidação e o estresse nitrosativo continuaram aumentados, porém o metabolismo da GSH foi normalizado a nível de indivíduos saudáveis (provavelmente também como resposta adaptativa ao estresse). Os níveis urinários de Gb3 diminuíram com a TRE, porém permaneceram ainda significativamente aumentados em relação aos controles. Concentrações de liso-Gb3 similares às encontradas no plasma de pacientes com DF tiveram efeito citotóxico significativo nas células renais, causaram dano ao DNA (incluindo dano oxidativo a purinas e pirimidinas) e induziram aumento na atividade da enzima antioxidante catalase (CAT) e na expressão da poli(ADP-ribose) polimerase 1 (PARP-1, enzima-chave no reparo de dano ao DNA por excisão de bases - BER). Por fim, analisando em conjunto os dados in vitro, podemos sugerir um mecanismo de ação para o liso-Gb3 no qual o H2O2, espécie reativa de conhecido papel sinalizador, atua como mediador. O H2O2 também ocupa papel central no mecanismo sugerido através dos estudos in vivo, de modo que os achados são complementares entre si e fornecem novos dados acerca da fisiopatologia da DF que podem ser úteis a estudos futuros na busca de terapias complementares à TRE, necessárias para a melhora clínica e de qualidade de vida dos pacientes com DF. / Fabry disease (FD, OMIM 301500) is an X-linked lysosomal storage disorder caused by mutations in GLA gene that lead to translated lysosomal enzyme α-galactosidade A (α-gal A, EC 3.2.1.22) with deficient or absent activity. Consequently, the enzyme’s substrates (mainly globotriaosylceramide – Gb3 and glotriaosylsphingosine – lyso-Gb3) accumulate in various tissues and body fluids of FD patients. Studies have pointed that inflammation, oxidative and nitrosative stress may be involved in FD pathophysiology, which is not completely known. It was previously described an increase of inflammatory and prooxidative parameters in FD patients during enzyme replacement therapy (ERT), although it is necessary to evaluate these parameters before the beginnig of ERT. Using in vivo and in vitro assays in blood and urine samples of FD patients, we evaluated abnormalities of redox status, oxidative damage to biomolecules (including lipids, proteins and DNA) and Gb3 urinary levels. In order to investigate the biological effects of the latest described biomarker for FD, lyso-Gb3, we performed in vitro assays in human embryonic kidney epithelial cells (HEK-293T). We demonstrated, for the first time, increased DNA damage in FD patients. Then, we observed increased reactive species generation (by dichlorofluorescein – DCF - probe) in the same samples and, by comet assay with endonucleases, we verified that DNA damage has an oxidative origin in purines in these patients. In order to evaluate repair capacity towards an oxidative insult, a challenge assay with hydrogen peroxide (H2O2) was performed, which demonstrated that patients had increased repair (probably as response to chronic stress), but not sufficiently effective to reduce basal oxidative damage in purines to healthy individuals’ levels. We also observed that FD patients even before initiating ERT already had increased lipid peroxidation levels (by thiobarbituric acid reactive species - TBARS - and malondialdehyde - MDA content) and nitrosative stress (by urinary nitrate and nitrite), as well as imbalances in glutathione (GSH) redox system (GSH – the main intracellular antioxidant, evaluated by GSH content and glutathione reductase, GR, and glutathione peroxidase, GPx, enzyme activities). After long-term ERT (approximately five years), lipid peroxidation and nitrosative stress remained increased, although GSH metabolism was restored to the level of healthy subjects (also probably as an adaptive response to stress). Gb3 urinary levels decreased after ERT, but remained already increased when compared to controls. Lyso-Gb3 levels similar to that found in plasma of FD patients caused significative cytotoxic effect on kidney cells, caused DNA damage (including oxidative damage to purines and pyrimidines) and induced increase in the antioxidant enzyme catalase (CAT) activity and poly(ADP-ribose) polymerase-1 (PARP-1) expression (key enzyme in DNA base excision repair). To conclude, analysing together the in vitro data, we could suggest a mechanism of action to lyso-Gb3 in which H2O2, reactive specie with well known signaling function, acts as a mediator. H2O2 has also a central role in the mechanism proponed by the in vivo studies, so that the findings are complementary and provide new data about FD pathophysiology that could be useful to future studies looking for complementary therapies to ERT, necessary to improvement in clinical aspects and life quality of FD patients.

As doenças raras são epidemiologicamente caracterizadas por ocorrerem entre 0,65-1:1.000 indivíduos. Houve importante incremento no surgimento de tratamentos específicos para muitas destas condições a partir da publicação de legislações específicas em diferentes países (sendo o pioneiro em 1983 nos EUA o chamado 'Orphan Drug Act'). Dentre as doenças que dispõem atualmente de tratamento específico está a Doença de Fabry, uma doença genética classificada erro inato do metabolismo do grupo das doenças Lísossômica se caracterizada pelo acúmulo de glicoesfíngol ipídíos no endotélio vascular, o que ocasiona problemas renais, cardíovasculares e neurológicos. Os problemas renais e cardíovasculares são os responsáveis pela maior morbi-mortalídade entre os portadores da doença. Esta ocorre em homens e mulheres, com manifestações heterogêneas entre indivíduos e diferentes gêneros. O tratamento específ ico da doença é através da chamada Terapia de Reposição Enzimática (TRE) havendo duas opções existentes: a agalsidase alfa e a agalsidase beta. Não há, porém, consenso sobre os resultados da TRE sobre desfechos clínicos relevantes. O alto custo associado a esta terapêutica (cerca de 200 mil dólares/ano/por paciente adulto) é o dificulta o acesso dos pacientes à mesma. Uma das formas se racionalizar a incorporação de tecnologias novas na área de saúde, que se caracteriza empreender uma avaliação objetiva de benefícios e custos associados é a chamada Avaliação de Tecnologias de Saúde. Este conjunto de métodos é cada vez mais empregado para auxiliar ou nortear decisões na área de saúde. Existem, no entanto , dificuldades associadas ao seu emprego no campo das doenças raras. Nossa proposta consiste em tendo como modelo a Doença de Fabry, obter um panorama da situação no Brasil a o acesso a tratamentos de alto custo para doenças raras, e, analisar a aplicabilidade dos princípios da avaliação de tecnologias de saúde no campo das doenças raras, tentando por fim colaborar no aprimoramento do processo de acessibilidade a estes tratamentos. OBJETIVOS Objetivo geral : Avaliar o acesso no Brasil a tratamentos de doenças raras e a aplicabilidade da avaliação de tecnolog ias em saúde no contexto específico destas. Objetivos específicos: 1) Caracterizar as políticas existentes no Brasil em relação ao acesso de medicamentos para doenças raras e avaliar as formas empregadas para este acesso. 2) Colaborar no aprimoramento do processo de incorporação destes tratamentos (de al to custo) no segmento das doenças raras, através da aplicação de principias de 'aval iação de tecnologias em saúde', usando como exemplo a Doença de Fabry. 3) Identificar e propor possíveis estratégias que possam colaborar para uma acessibilidade adequada e justa a tratamentos cl inicamente efetivos no campo das doenças raras. MÉTODOS 1) Foi realizada uma revisão da literatura caracterizando as politicas brasileiras na área de medicamentos e aval iando as formas de acesso utilizadas pelos pacientes; 2) Uma segunda revisão, de caráter sistemático, sobre os efeitos da TRE na DF, foca d a nos efe itos sobre a nefropatia e a cardiopatia (sistemas cujo comprometimento em pacientes com DF causa maior morbi-mortalidade) da doença foi feita. Baseado nos dados obtidos elaborou-se um modelo probabilístico (markoviano) para avaliar o efeito da TRE sobre a nefropatia da DF seguido de aval iação do custo-efetividade. A partir de uma análise destes resultados e da li teratura, foram identificados os aspectos que mais influenciariam a acessibilidade ao tratamento, seguida da elaboração de propostas para auxiliar a incorporação e a acessibil idade a estes tratamentos. RESULTADOS 1) A política de assistência farmacêutica brasileira atual é baseada em elencos de medicamentos divididos em atenção básica e do 'componente especiali zado'. Não existe uma política especifica direcionada aos tratamentos (de alto custo) dos portadores de doenças raras. 2) A revisão empreendida mostrou que dentre os dois sistemas aval iados de forma particular (pela morbi-mortalidade associada) havia mais dados quer sobre história natural , quer sobre efeitos da TRE em relação à nefropatia, tendo sido esta escolh ida então como foco da modelagem . O modelo construido foi capaz de identificar um subgrupo de pacientes com DF que quando tratado com TRE tem redução significativa (diferença absoluta de 10%) na probabilidade de ter progressão da nefropatia (evoluir a estágio dialítico). A despeito dos benefícios observados a análise mostrou ser esta opção não custo-efetiva no contexto existente (custo é superior ao limite preconizado pela OMS de s 3 vezes o PIB per capíta do Brasil). 3) A acessibilidade é ainda um obstáculo ao uso destes medicamentos. O custo é o limitante central da acessibilidade aos tratamentos específicos destas patologias. Atuar em aspectos associados a este e às pol íticas vigentes seriam formas de tentar mudar este cenário. CONCLUSÕES 1) Não há em nosso país políticas que sistematizem o processo de incorporação de tratamentos (de alto custo) para doenças raras. O acesso a estes é disperso, no caso da DF não existe disponibil idade da terapia de reposição enzimática via SUS. 2) A revisão sobre o efeito da TRE na nefropatia mostrou resultados muito heterogêneos. Foi possivel, no entanto, elaborar um modelo avaliando TRE na nefropatia da DM, o qual é o primeiro conhecido utilizando este desfecho especifico e elaborado dentro contexto brasileiro. A despeito de identificado um subgrupo de pacientes que poderia beneficiar-se significativamente com o tratamento especifico a estratégia não foi custo-efetiva para este desfecho. Os custos associados foram o limitante central. Novas opções em termos de efetivação da incorporação destas tecnologias devem ser consideradas. / BACKGROUND. According to the WHO definition rare diseases occur between 0.65-1/1,000 individuals. The U.S. 'Orphan Drug Act' in 1983', the first legislation with incentives for the development of therapies for rare diseases had an impact on the development of treatments for these diseases. Fabry Disease (FD) is a rare genetic disease characterized by accumulation of glycosphingolipids in vascular endothelium leading to systemic dysfunction (renal, cardiovascular, and neurologic disease). This disease has specific treatment available, with two options of recombinant enzymes (alfa or betagalsidase) for enzyme replacement therapy (ERT). There are few controlled trials evaluating their effects, indicating some improvements in neuropathic pain, in heart abnormalities and in globotriaosylceramide (GL-3) levels. Nevertheless most of the clinical benefits of ERT are still unclear. Another aspect is the high cost associated with this treatment which makes it not easily accessible. OBJECTIVES: General Objective: Evaluate access to rare disease treatments in Brazil and the applicability of health technology assessment (HTA) in the context of these diseases. Specific objectives: 1) Evaluate the Brazilian governmental policies for rare diseases treatment and how the access to the treatments is actually done. 2) Evaluate the applicability of health technology assessment (HTA) in the context of rare diseases, using the example of Fabry disease. 3) Identify and propose strategies that could contribute to a fair access to clinically effective treatments for of rare diseases. METHODS: An extensive literature review about the Brazilian policies in the area was performed. In the sequence a systematic review about the subject ERT and FD was conducted. After this, a model estimating the likelihood of nephropathy progression with or without ERT was built, followed by a cost-effectiveness analysis. Another literature review focused in the identification of obstacles to accessibility and possible strategies to overcome them was conducted. RESULTS: 1) There is no specific policy in Brazil regarding high cost drugs for the treatment of rare diseases. 2) ERT appears to slow the progression of nephropathy in the Fonly proteinuria subgroup '. However the cost associated is very high making this option not cost-effective. 3) The accessibility to innovative treatments for rare diseases is not adequate and the high cost of these therapies is a major obstacle to change this scenario. CONCLUSION: There are no policies in Brazil to systematize the access to the specific (high cost) treatments for rare diseases. The model evaluating TRE in FD nephropathy was able to identify benefits for a subgroup of patients. It was the first known model using this specific outcome and built focusing the Brazilian context. However the ERT strategy was not cost-effective for this outcome. The costs associated with these therapies are very high and an important limiting factor to the access. the central. New options should be considered to offer adequate access to the (effective) therapies.

La plasmonique est un domaine de la nano-photonique qui étudie le comportement de la lumière à des échelles sub-longueurs d'ondes en présence de métaux. Les plasmons polaritons de surface (SPPs) sont des modes électromagnétiques qui se propagent à l'interface entre un diélectrique et un métal. Les SPPs trouvent des applications dans plusieurs domaines comme la communication et le traitement tout-optique du signal, la spectroscopie, la détection en biologie et en chimie. De nombreux composants plasmoniques (modulateurs, coupleurs, détecteurs ...) ont été démontrés ces dernières années. Cependant, leur l'intégration reste conditionnée par l'absence d'un générateur compact (pompage électrique, dimensions réduites) et par les grandes pertes ohmiques. Les techniques standards de génération de SPs nécessitent l'alignement d'un laser externe sur un prisme ou un réseau de diffraction afin d'adapter le vecteur d'onde incident avec celui du plasmon. L'approche que nous avons choisie est basée sur l'utilisation de lasers à semiconducteur ayant une polarisation transverse magnétique (TM) comme source d'excitation et de gain. Notre approche, permet d'obtenir des dispositifs compacts et facilement intégrables sur puce. Pendant ma thèse j'ai étudié expérimentalement et numériquement les performances d'un laser en fonction rapprochement du contact métallique à sa région active. La proximité du gain optique au métal est nécessaire pour la réalisation de dispositifs plasmoniques actifs. J'ai démontré la génération et l'amplification des plasmons de surface dans la bande télécom (λ=1.3µm), avec des dispositifs compacts, à base de semiconducteurs, fonctionnant par injection électrique et à température ambiante. Notamment, j'ai réalisé une architecture élégante, avec coupleur intégré, pour la génération de SPPs accessibles sur le sommet du dispositif. Un dispositif avec gaine superficielle ultrafine a permis de démontrer un mode hybride plasmonique avec une fraction consistante de champ électrique à l'interface métal/semiconducteur. Finalement, j'ai montrée que la structuration nanométrique du contact métallique réduit les pertes du mode laser. Les résultats sont renforcés par une nouvelle technique de imagerie de champ proche (SNOM) qui a permis de mesurer les SPPs à l'interface métal/or et à l'interface métal/ semiconducteur. Grâce aux mesures SNOM, il a aussi été possible de démontrer sans aucune ambiguïté l'effet de la structuration du métal sur le mode optique. / The field of plasmonics is experiencing a rapid development, due to the interest in studying the behavior of light at the nanometer scale. Key ingredients of plasmonics are the surface plasmons (SPs), electromagnetic modes localized at the interface between a metal and a dielectric. SPs rely on the interaction between electromagnetic radiation and conduction electrons at metallic interfaces or in "small" metallic nanostructures. The recent intense activity on plasmonics has been also enabled by state-of-the-art nano fabrication techniques and by high-sensitivity optical characterization techniques. These tools pave the way to promising applications (integration in electronics, chemical and biological detection...), which exploit the SP peculiarity of confining optical fields over sub-wavelength mode volumes. The number of publications concerning plasmonics has been continuously increasing over the last twenty years giving rise to a dynamic research context. Several plasmonic devices have been demonstrated during the last years (modulators, couplers, detectors ...). However their integration is limited by the absence of a compact generator (electrical pumping, small dimensions) and by the huge ohmic losses. Standard techniques for surface plasmon polariton (SPP) generation need an external alignment with a laser source on a prism or on a grating. Our approach is based on semiconductor lasers sources with a transverse magnetic (TM) polarization. Therefore, it is possible to obtain compact semiconductor devices suitable for the on chip integration. During my thesis I studied experimentally and numerically the performance of a diode laser as a function of the metal distance from its active region. The proximity of the gain to the metal is necessary to realize active plasmonic devices. I demonstrated the generation and the amplification of SPP in the telecom range (λ=1.3µm) with compact semiconductor based devices, operating at room temperature and by electrical injection. I realized an elegant architecture with an integrated coupler grating for the SPP generation. The SPPs are directly accessible at the device surface. An ultra-thin cladding device allowed the demonstration of a hybrid plasmonic laser with a consistent fraction of electric field at the metal/semiconductor interface. Finally I demonstrated that the metal patterning allows a loss reduction, decreasing the laser threshold. The results are strengthened by a new near-field technique (NSOM) which permitted to measure the SPPs at the metal/air interface and at the metal/semiconductor interface. Thanks to the NSOM we showed unambiguously the effect of the metal patterning on the optical mode.