L’essor des télécommunications par fibre optique nécessite l’insertion en sortie des lasers d’un isolateur optique intégré protégeant celui-ci des réflexions qui le déstabilisent. Ce composant existe à l’heure actuelle sous forme massive, mais son intégration sur la même plaquette que le laser pose problème du fait de la difficulté à intégrer les bons matériaux magnétooptiques sur les substrats usuels de l’optique guidée. Dans cette perspective, l’intégration de nanoparticules magnétiques dans un sol-gel déposé sur les guides optiques est une voie prometteuse, développée par le laboratoire Hubert Curien. Cette thèse a eu pour but d’étudier de manière plus systématique le composant non-réciproque qui entre dans la fabrication des isolateurs à conversion de mode, à savoir le rotateur non-réciproque. Deux études poussées, l’une théorique, l’autre expérimentale, recoupées entre elles par des modèles numériques, ont été mises en oeuvre au cours des années de cette thèse. L’étude théorique a permis de tenir compte tous les paramètres ayant une influence sur l’état de polarisation de la lumière dans un guide magnétooptique, y compris les dichroïsmes souvent négligés. L’étude pratique, à partir d’échantillons sur verre réalisés en collaboration avec l’IMEP-LAHC et le laboratoire PHENIX, a abouti à une caractérisation quasi complète des effets magnétooptiques — longitudinaux et transverses — dans les guides et de l’influence des paramètres de fabrication sur ceux-ci. Au final, ces résultats nous ont donné une compréhension plus complète du fonctionnement des guides magnétooptiques, et nous ont permis de prédire les paramètres optimaux qu’il faudra mettre afin de fabriquer, dans un futur proche, l’isolateur complet sur une seule plaque de verre / The development of optical-fiber telecommunications requires the insertion of optical isolator between lasers and fibers, in order to protect them against perturbating reflexions. This component is currently inserted in a bulk form, but the goal is to integrate it on the same wafer than the laser; nevertheless, this is problematic due to the difficulty to integrate good magnetooptical materials on usual substrates as glass or silicon. One of the promising way to achieve this, developped by the Laboratoire Hubert Curien, is the embedding of magnetic nanoparticles into a sol-gel matrix deposited above the optical guides. This thesis aimed at studying more deeply the main non-reciprocal component of integrated mode conversion optical isolators: the non-reciprocal rotator. A theorical and a practical study have both been performed, with numerical simulations to confront them. The theorical study aimed at describing the evolution of propagation in magnetooptical waveguides with respect to all effects, even absorption and dichroïsm. Then a practical study was performed on glass samples engineered in collaboration with IMEP-LAHC and the PHENIX laboratory, and lead to a full measurement of longitudinal and transverse magnetooptical effects, and their evolution related to the fabrication parameters of the samples. Finally, these results gave us a comprehensive view of how magnetooptical waveguides behave, and we were able to predict the good parameters to choose in order to construct, in a close future, a glass-integrated optical isolator

La photonique du silicium a connu un développent massif pendant les dix derniers années. Presque toutes les briques technologiques de base ont été réalisées et ont démontrées des performances remarquables. Cependant, le manque d’une source laser intégrée en silicium a conduit les chercheurs à développer de composants basés sur l’intégration entre le silicium et les matériaux III-V.Dans cette thèse je décris la conception, la fabrication et la caractérisation des lasers hybrides III-V sur silicium basés sur cette intégration. Je propose un coupleur adiabatique qui permet de transférer intégralement le mode optique du guide silicium au guide III-V. Le guide actif III-V au centre du composant fourni le gain optique et les coupleurs, des deux cotés, assurent le transfert de la lumière dans les guides silicium.Les lasers mono longueur d’onde sont des éléments fondamentaux des communications optiques. Je décris les différentes solutions permettant d’obtenir un laser mono-longueur d’onde hybride III-V sur silicium. Des lasers mono longueur d’onde ont été fabriqués et caractérisés. Ils démontrent un seuil de 21 mA, une puissance de sortie qui dépasse 10 mW et une accordabilité de 45 nm. Ces composants représentent la première démonstration d’un laser accordable hybride III-V sur silicium. / Silicon photonics knew an impressive development in the last ten years. Almost all the fundamental building blocks have been demonstrated and reveal competitive performances. However, the lack of an efficient silicon integrated laser source has led the researchers to develop heterogeneous integration of III-V materials on silicon.In this thesis I describe the design, the fabrication and the performances of these hybrid III-V on silicon lasers. I propose the use of an adiabatic coupler that totally transfers the optical mode between the III-V and the silicon waveguides. The active waveguide on III-V materials at the center of the device provides the optical gain, while, on both side, adiabatic couplers allow a loss-less transfer of the optical mode to the silicon waveguide. Single wavelength emitting lasers are fundamental elements for high bandwidth optical links. I review all the effective solutions enabling single waveguide hybrid III-V on SOI lasers. DBR, microring based, DFB and AWG laser solutions were analysed. Single wavelength operating lasers have been fabricated and characterized. A laser threshold of only 21 mA, an output power of more than 10 mW and tunability over 45 nm with a SMSR of 45 dB have been measured. These devices represent the first demonstration of a monolithically integrated hybrid III-V/Si tunable laser made by wafer bonding technique.

The rapid development of optofluidics, the combination of microfluidics and integrated optics, since its formal conception in the early 2000's has aided in the advance of single-molecule analysis. The optofluidic platform discussed in this dissertation is called the liquid core anti-resonant reflecting optical waveguide (LC-ARROW). This platform uses ARROW waveguides to orthogonally intersect a liquid core waveguide with solid core rib waveguides for the excitation of specifically labeled molecules and collection of fluorescence signal. Since conception, the LC-ARROW platform has demonstrated its effectiveness as a lab-on-a-chip fluorescence biosensor. However, until the addition of optical multiplexing excitation waveguides, the platform lacked a critical functionality for use in rapid disease diagnostics, namely the ability to simultaneously detect different types of molecules and particles. In disease diagnostics, the ability to multiplex, detect and identify multiple biomarkers simultaneously is paramount for a sensor to be used as a rapid diagnostic system. This work brings optofluidic multiplexing to the sensor through the implementation of three specific designs: (1) the Y-splitter was the first multi-spot excitation design implemented on the platform, although it did not have the ability to multiplex it served as a critical stepping stone and showed that multi-spot excitation could improve the signal-to-noise ratio of the platform by ~50,000 times; (2) a multimode interference (MMI) waveguide which took the multi-spot idea and then demonstrated spectral multiplexing capable of correctly identifying multiple diverse biomarkers simultaneously; and, (3) a Triple-Core design which incorporates excitation and collection along multiple liquid cores, enabling spatial multiplexing which increases the number of individual molecules to be identified concurrently with the MMI waveguide excitation. In addition to describing the development of optical multiplexing, this dissertation includes an investigation of another LC-ARROW based design that enables 2D bioparticle trapping, the Anti-Brownian Electrokinetic (ABEL) trap. This design demonstrates two-dimensional compensation of a particle's Brownian motion in solution. The capability to maintain a molecule suspended in solution over time enables the ability to gain a deeper understanding of cellular function and therapies based on molecular functions.

The Dielectric Loaded Surface Plasmon Polariton Waveguides (DLSPPWs) have recently emerged as a possible solution to carry both optical and electrical signals on- chip. However, in the particular context of optical interconnects, advanced functionalities such as filtering, switching, and routing are required in order to replace in the future the equivalent electronic components which are too much power consumer and also to reduce their footprints. After presenting the interest and limitation of the leakage radiation microscopy method used all along this work, we show several active devices using thermo-sensitive polymers as the dielectric load driven electrically by Joule heating. Then we demonstrate the feasibility of all-optical systems by either doping the dielectric with metallic nanoparticles or by plasmo-thermal eect of a second plasmonic mode providing a localized heating of controlled shape. The dynamic activation of our thermo- optical devices is performed using a homemade fiber-to-fiber setup which allows us to investigate the response time of a plasmo-thermal heating as well as true datacom transmission. Some improvements of the original DLSPPWs performances are proposed by adding a metallic wall on one side of the polymer ridge. This system can act as a compact and athermal polarization converter

This PhD thesis is devoted to study electro-optic properties of Gemanium/Silicon-Germanium (Ge/SiGe) multiple quantum wells (MQWs) for light modulation, detection, and emission on Si platform. It reports the first development of high speed, low energy Ge/SiGe electro-absorption modulator in a waveguide configuration based on the quantum-confined Stark effect (QCSE), demonstrates the first Ge/SiGe photodiode with high speed performance compatible with 40 Gb/s data transmission, and realizes the first Ge/SiGe light emitting diode based on Ge direct gap transition at room temperature. Extensive DC and RF measurements were performed on each tested prototype, which was realized using the same epitaxial growth and fabrication process. Simple theoretical models were employed to describe experimental properties of the Ge/SiGe MQWs. The studies show that Ge/SiGe MQWs could potentially be employed as a new photonics platform for the development of a high speed optical link fully compatible with silicon technology.

SWIFTS (Stationary Wave Integrated Fourier Transform Spectrometer) est une nouvelle technologie innovante de spectrométrie qui permet une réduction radicale de la taille des spectromètres à Transformée de Fourier, tout en conservant, et même en améliorant leurs performances. Grâce aux avancées de l'optique intégrée et des nanotechnologies, SWIFTS repose sur une méthode de détection optique originale, sans aucune partie mobile, où des nanoplots métalliques échantillonnent directement le champ évanescent d'une onde stationnaire dans un guide d'onde.Dans cette thèse, nous proposons de présenter le cheminement complet qui a mené, en partant du concept original, au développement puis à la mise en pratique de la technologie SWIFTS. Le document illustre notamment les caractérisations optiques, les choix technologiques et les optimisations entrepris pour la réalisation de spectromètres fonctionnels dans le domaine visible et proche-infrarouge. Des procédures de calibrages novatrices et complémentaires, basées sur du multiplexage fréquentiel et sur de l'interférométrie à faible cohérence temporelle, ont été développées pour déterminer avec précision les différentes irrégularités de fabrication et de comportement de l'appareil complètement intégré. Les spectromètres calibrés permettent à présent d'aborder des applications diverses en industrie et en recherche, de la caractérisation hautes performances de lasers, à l'interrogation de capteurs fibrés à réseaux de Bragg, aux techniques de spectrométries Raman et LIBS, et de tomographie optique OCT, jusqu'aux sciences de l'Univers (géophysique, astrophysique).SWIFTS est une innovation de rupture qui, de part sa miniaturisation obtenue sans compromis avec de hautes performances d'analyse spectrale, a la capacité de faire passer la spectrométrie du stade de la mesure complexe en laboratoire à celle d'un simple composant intégré pour des applications exigeantes. / SWIFTS (Stationary Wave Integrated Fourier Transform Spectrometer) is a new innovative technology of spectrometry that allows a drastic reduction of the size of Fourier transform spectrometers, while maintaining, and even improving their performance. With advances in integrated optics and nanotechnology, SWIFTS is based on an original method of optical detection, without any moving part, where metallic nanodots directly sample the evanescent field of a standing wave in a waveguide.In this thesis, we propose to present the complete process that led, starting from the original concept, to the development and the applications of the technology. The document illustrates the optical characterizations, the technological choices and the optimizations made for the realization of functional spectrometers in the visible and near-infrared range. Innovative and complementary procedures of calibrations, based on frequency multiplexing and low coherence interferometry, have been developed to accurately determine the various irregularities of the manufacturing and of the behavior of the integrated device. The calibrated spectrometers allow to address various applications in industry and research, such as high performance characterization of lasers, interrogation of fiber Bragg gratings sensors, Raman and LIBS spectrometry, optical coherence tomography OCT, and sciences of the Universe (geophysics, astrophysics).SWIFTS is a breakthrough innovation in spectrometry, without trade-off between miniaturization and high performance, that opens the way for product development based on the most demanding applications currently performed in research laboratories.

O presente trabalho tem como objetivo desenvolver um processo de fabricação de elementos micro-ópticos utilizando-se litografia por feixe de elétrons, empregando o resiste SU-8, negativo e amplificado quimicamente, sobre substrato de Si. Para tanto, é realizado o estudo dos parâmetros do efeito de proximidade a, b e h para se modelar e controlar os efeitos do espalhamento dos elétrons no resiste e no substrato, e se altera o processamento convencional do SU-8 para se obter um processo com baixo contraste. A determinação dos parâmetros do efeito de proximidade para o sistema de escrita direta e amostra SU-8 / Si é feita experimentalmente e por simulação de Monte Carlo. Particularmente, verifica-se a dependência dos mesmos com a profundidade do resiste. Primeiramente utilizando o software PROXY, obtêm-se a, b e h da observação de padrões de teste revelados. Chega-se a 4m para o parâmetro () que mede o retroespalhamento dos elétrons pelo substrato e 0,7 para a relação (h) entre a intensidade destes com aquela dos elétrons diretamente espalhados pelo resiste (alcance dado por a). Ainda, com esses dados, estima-se o diâmetro do feixe do microscópio eletrônico de varredura a partir da equação de aproximação de espalhamento direto para pequenos ângulos (a = 128nm na superfície do resiste) e se determina a resolução lateral do processo (a = 800nm na interface resiste/ substrato, para um filme de 2,4m). Em seguida, usa-se o software CASINO para se calcular os parâmetros de proximidade a partir da curva de densidade de energia dissipada no resiste obtida pela simulação da trajetória de espalhamento dos elétrons. Confrontam-se, finalmente, os valores obtidos pelos dois métodos. Em relação ao processamento do resiste SU-8, são determinadas as condições experimentais para a fabricação de estruturas tridimensionais por litografia de feixe de elétrons. Especificamente, busca-se desenvolver um processo com características (espessura, contraste, sensibilidade e rugosidade) adequadas para a fabricação de micro-dispositivos ópticos. Inicia-se com o levantamento das curvas de contraste e da sensibilidade do SU-8 para determinadas temperaturas de aquecimento pós-exposição. Obtém-se contraste abaixo de 1 para aquecimento pós-exposição abaixo da temperatura de transição vítrea do resiste, mantendo-se sensibilidade elevada (2C/cm2). Em seguida, mede-se a rugosidade da superfície do filme revelado para diferentes doses de exposição. Para finalizar, submete-se a amostra a um processo de cura e escoamento térmico, para melhorar a dureza e a rugosidade do resiste a ser utilizado como dispositivo final Consegue-se um valor de rugosidade (40nm) inferior a 20 vezes o comprimento de onda de diodo laser de eletrônica de consumo. Por fim, é produzido um dispositivo com perfil discretizado em 16 níveis como prova de conceito. / This work aims at developing an electron-beam lithography process for the fabrication of microoptical elements using the negative tone chemically amplified resist SU-8 on Si substrate. A study of the proximity effect parameters a, b and h is carried out to model and control the electron scattering both in the resist and in the substrate, and the SU-8 standard processing conditions are changed to achieve a low contrast process. The determination of the SU-8 / Si proximity effect parameters and its dependence with resist depth is done employing an experimental method and through Monte Carlo simulations. First, a, b and h are obtained comparing exposed patterns calculated by the software PROXY. b, the parameter which measures the backscattering of the electrons by the substrate, is equal to 4m and the value of h, the ratio of the dose contribution of backscattered electrons to that of the forward scattered (related to a), is 0.7. The extrapolation of exposed patterns data is used to estimate the scanning electron microscope beam diameter through the equation for low angle scattering (a = 128nm at the resist surface) and the lateral resolution of the process is determined (a = 800nm at the resist/ substrate interface, for a 2.4m film). With aid of the software CASINO, Monte Carlo simulations of the scattering trajectories of electrons in substrate and resist materials are calculated, recording the energy that they dissipate through collisions along their path. The results obtained representing the profile of the energy dissipated in the resist are used to determine the proximity effect parameters. The experimental method results are compared to that obtained by simulation. Regarding the SU-8 processing, the process parameters for the fabrication of three-dimensional structures by electron-beam lithography are determined. The process is designed to have specifications (thickness, contrast, sensitivity and surface roughness) suitable for microoptical elements fabrication. It begins with the determination of the SU-8 contrast curve and its sensitivity for specific post-exposure bake temperatures. A below the unit contrast process with high sensitivity (2C/cm2) is achieved postannealing the sample below the resist glass transition temperature. The film surface roughness is measured after resist development for different exposure doses, and a controlled hardbake (cure) and reflow is carried to enhance both the mechanical properties and the surface roughness of the structures that will remain as part of the final device. A RMS roughness of 40nm, lower than 20 times the wavelength of consumer electronics laser diode, is obtained. The electron-beam process designed is applied to the fabrication of a microelement with a 16-level profile discretization.

Les guides d’ondes plasmoniques à rubans dielectriques (DLSPPW) sont récemment apparus comme une des solutions possible pour le transport de signaux optiques et électriques sur puce. Néanmoins, dans le contexte particulier des interconnections optiques, des fonctionalitées avancées telles que filtrage, commutation, et routage sont nécessaires afin de remplacer dans le futur les composants electroniques équivalents trop gourmands en énergie et aussi réduire leur empreinte. Après une présentation des intérêts et limitations de la technique de micro- scopie à fuite radiative, nous montrons plusieurs composants actifs utilisant pour diélectrique des polymères thermo-sensibles controlés électriquement par eet Joule. Par la suite nous démontrons la faisabilité de systèmes tout optique que ce soit par dopage du polymère par des nanoparticules metalliques ou par eet thermo-plasmonique d’un second mode plasmon permettant un échauement localisé de forme choisie. L’activation dynamique de nos composants thermo-optiques est réalisée grâce à un montage fibre-à-fibre créé spécialement nous permettant d’investiguer le temps de réponse d’un chauage plasmonique ainsi que la transmission de signal télécom. Des améliorations de performances du concept DLSPPW original sont proposées par l’ajout d’un mur métallique sur le côté du ruban de polymère. Ce système peut alors fonctionner comme un convertisseur de polarisation compacte et athermique / The Dielectric Loaded Surface Plasmon Polariton Waveguides (DLSPPWs) have recently emerged as a possible solution to carry both optical and electrical signals on- chip. However, in the particular context of optical interconnects, advanced functionalities such as filtering, switching, and routing are required in order to replace in the future the equivalent electronic components which are too much power consumer and also to reduce their footprints. After presenting the interest and limitation of the leakage radiation microscopy method used all along this work, we show several active devices using thermo-sensitive polymers as the dielectric load driven electrically by Joule heating. Then we demonstrate the feasibility of all-optical systems by either doping the dielectric with metallic nanoparticles or by plasmo-thermal eect of a second plasmonic mode providing a localized heating of controlled shape. The dynamic activation of our thermo- optical devices is performed using a homemade fiber-to-fiber setup which allows us to investigate the response time of a plasmo-thermal heating as well as true datacom transmission. Some improvements of the original DLSPPWs performances are proposed by adding a metallic wall on one side of the polymer ridge. This system can act as a compact and athermal polarization converter

Nous avons étudié l'influence de la fabrication de guides d'ondes optiques par échange protonique doux(SPE) sur les cristaux de niobate de lithium (LN) polarisé périodiquement et nous avons montré que,dans certains cas, ce processus conduit à la création de nanodomaines en surface. Ces nanodomaines enforme d'aiguille peuvent être responsables de la réduction de l'efficacité de conversion non linéaireobservée dans les guides qui sont affectés. Nous avons également étudié l'influence de différents typesd'échange protonique sur la formation, par application d'un champ électrique, de domaines dans le LNcongruent. Cette étude montre que le seuil de nucléation peut être fortement réduit par la présence duguide d'onde et que l'apparition et le développement des domaines en forme de traits est fortementmodifiée. Elle montre également que la fusion des nanodomaines existants au voisinage des parois dedomaine aboutit à la formation de parois élargies et de domaines en forme de dendrites. En irradiantavec un faisceau d'électrons la surface Z- d'un échantillon de LN préalablement soumis à un échangeprotonique doux et recouvert d'une couche de résine électronique, nous avons réussi à former desdomaines avec des formes arbitraires. Par cette technique, nous avons fabriqué des domainespériodiques d'excellente qualité dans des cristaux présentant des guides canaux SPE. Des expériences degénération de deuxième harmonique dans ces guides nous ont permis d'obtenir des efficacités deconversion de 48%/W.cm2 ce qui est conforme aux prédictions ainsi que la forme des spectres d'accordde phase que nous avons observés. Ceci démontre tout l'intérêt de ce processus / The investigation of influence of the soft proton exchange (SPE) optical waveguide (WG) creation onperiodically poled lithium niobate (PPLN) has been done. It has been shown that the WG fabricationprocess can induce the formation of needle like nanodomains, which can be responsible for thedegradation of the nonlinear response of the WG created in PPLN crystals. The domain structure (DS)evolution has been studied in congruent lithium niobate (LN) crystals with surface layers modified bythree different proton exchange techniques. The significant decrease of the nucleation threshold fieldand qualitative change of domain rays nucleation and growth have been revealed. The formation of abroad domain boundary and dendrite domain structure as a result of nanodomains merging in front ofthe moving rays has been demonstrated. The formation of DS in LN with SPE by irradiation of coveredby electron resist polar surface of LN has been investigated. Formation of domains with arbitrary shapesas a result of discrete switching has been revealed. Finally, it has been demonstrated that electron beamirradiation of lithium niobate crystals with surface resist layer can produce high quality periodical domainpatterns after channel waveguide fabrication. Nonlinear characterizations show that the conversionefficiencies and the phase matching spectra conform to theoretical predictions, indicating that thiscombination presents a great interest for device fabrication. Second harmonic generation withnormalized nonlinear conversion efficiency up to 48%/(W cm2) has been achieved in such waveguides

Ce manuscrit de thèse s’intéresse à l’étude théorique et à l’observation d’effets quantiques résultantde la manipulation de photons en fréquence. Ainsi nous rapportons une expérience d’interférenceà deux photons, pour laquelle l’intrication et la manipulation des photons survient dans le domainedes fréquences. Nous montrons que cette figure d’interférence est sensible à la dispersion des deuxphotons jumeaux et que ce phénomène peut être compensé de manière non-locale. D’un autre coté,nous étudions la réalisation d’une marche aléatoire quantique établie par la modulation de phase.Nous mettons en évidence un comportement intéressant de ces marches et suggérons un schémaexpérimental. / This manuscript deals with a theoretical and experimental study of quantum effects taking place inthe frequency domain. On one side, we report a two photons interference experiment, where both theentanglement of the photons and their manipulation take place in the frequency domain. We showthat this interference pattern is sensitive to the dispersion of both photons and allows us to perform anonlocal dispersion cancellation. On the other side we study the implementation of a quantum walkbased on the phase modulation. We predict an interesting behavior of these quantum walks andsuggest a physical implementation.